Thesis Abstracts

Billy Connor Moore

Supervisor Name: Timothy Ravasi
Research Unit: Marine Climate Change Unit
Thesis: Physiological and Transcriptomic Response of Early-Life Stage Clownfish to Future Ocean Warming and Marine Heatwaves

Abstract:
Coral reef fish are currently under threat as anthropogenic emissions are driving long-term changes to mean sea surface temperatures, whilst also increasing the frequency and intensity of marine heatwaves. Increasing temperatures threaten to push reef fish beyond their natural thermal limits, with previous studies indicating this will elicit a range of adverse physiological and molecular responses. Nevertheless, the immediate, acclimatory and legacy effects of increasing temperature on early life-stage fish are unclear, and the molecular mechanisms underpinning many of the previously observed phenotypes remain unidentified. To address these knowledge gaps, I use a combination of genome sequencing, transcriptomics, and physiological measurements to investigate the response of early lifestage clownfish (Amphiprion ocellaris) to temperatures associated with ocean warming and marine heatwaves.
Firstly, I present two high quality chromosome-scale genomes for the anemonefish Amphiprion ocellaris and Amphiprion clarkii, that will aid future studies of these species. Then, via a series of aquaria-based experiments I show that A. ocellaris exhibit a range of physiological changes at +3°C, as larvae develop faster, and both larvae and juvenile clownfish display elevated metabolic rates. However, early exposure to +3 °C may also have an acclimatory effect, as the effect of acute temperature stress on metabolic rates are reduced with increasing developmental exposure. Furthermore, via stage- and tissuespecific transcriptome sequencing of larvae and juveniles I show widespread changes in gene expression at +3 °C. These genes encompass a range of biological functions including immune response, epigenetic reprogramming, neurotransmission, heat-stress, liver fibrosis and insulin signalling.
Overall, these results indicate that A. ocellaris will experience a very different developmental regime if they hatch and develop at +3 °C, with these changes likely to impact future population persistence. Moreover, the comprehensive transcriptomic sequencing conducted here, advances our understanding of how coral reef fish will respond to future climate change at the molecular level. Chapters one and two of this thesis have been peer-reviewed and published, and chapter three is currently in prep for journal submission.

Kojiro Suda

Supervisor Name: Keiko Kono
Research Unit: Membranology Unit
Thesis: The Role of Calcium Ion Dynamics and Inter-Organelle Communication During Plasma Membrane Damage-Dependent Cellular Senescence

Abstract:
Cellular senescence is a stable cell cycle arrest that contributes to a variety of physiological and pathological processes in vivo, including organismal aging, wound healing, and cancer. Accumulating evidence suggests that elimination of senescent cells ameliorates age-related pathologies. In vitro, various stresses, including oxidative stress, oncogene activation, telomere shortening, and DNA damage, induce cellular senescence via the DNA damage response. However, the triggers of cellular senescence in vivo remain controversial.
Here, I show that cellular senescence is induced by physiologically and pathologically occurring plasma membrane damaging stimuli such as mechanical injury and pore-forming toxins in normal human fibroblasts in vitro. I found that Ca2+ influx into the cytosol following plasma membrane damage is necessary and sufficient for the induction of plasma membrane damage-dependent senescence. Live cell imaging revealed that Ca2+ entering the cytosol is immediately incorporated by the endoplasmic reticulum (ER). Subsequently, mitochondrial Ca2+ levels rise steadily, suggesting Ca2+ transport from the ER to the mitochondria via their contact sites. I observed an increase in mitochondrial oxidative stress, and attenuation of oxidative stress with antioxidants suppressed plasma membrane damage-dependent senescence.
These results suggest that mitochondrial dysfunction due to Ca2+ accumulation induces plasma membrane damage-dependent senescence. I also found that Ca2+ transport from the ER to the mitochondria is necessary to maintain cytosolic Ca2+ levels and cell survival after plasma membrane damage. Proteomic analysis identified the proteins that mediate the ER-mitochondria contact. This study highlights an underappreciated subtype of cellular senescence, plasma membrane damage-dependent senescence, and provides mechanistic insights into Ca2+ dynamics based on plasma membrane damage-triggered inter-organelle communication.

Florian Lalande

Supervisor Name: Kenji Doya
Research Unit: Neural Computation Unit
Thesis: Planetary Systems Insights through Numerical Data Imputation Algorithms and Machine Learning

Abstract:
Since the first discoveries in the early 1990’s, the number of known exoplanets has exploded to reach over 5,500 as of December 2023. But the recorded information for each planets is sparse, with a lot of missing values, preventing from confidently drawing overarching conclusions. As most traditional data imputation methods provide a point estimate, they fail at capturing the complexity of multimodal data distributions and provide unreliable estimates in scenarios where data exhibits multiple modes. This calls for a new paradigm to model rich or complex numerical datasets.
This PhD thesis introduces the kNN×KDE, a numerical imputation tool which combines the flexibility of the k-nearest neighbors (kNN) and the simplicity of Kernel Density Estimation (KDE) to model the multi-dimensional distribution of missing data in datasets characterized by multimodality. This new method is tested against traditional and novel data imputation algorithms, and I show that the kNN×KDE not only provides better estimates, but also facilitates their interpretation. To demonstrate the practical significance of the kNN×KDE, I apply it to the NASA Exoplanet Archive – a dataset riddled with missing values, including both planetary radius and mass, and marked by pronounced multimodality. The analysis of the estimated distributions provided relevant insights into the demographics of the Exoplanet Population, potentially helping future missions to select interesting targets.
In addition, this PhD work includes two artificial neural network applications for planetary system analysis: a Convolutional Neural Network (CNN) to predict planetary system stability and a Graph Neural Network (GNN) to rediscover Newton’s Law of Gravitation and attempt to reproduce the scientific discovery of Neptune. Finally, this thesis features a Transformer model for Symbolic Regression applied to 120 real-world physics equations. These additional tools contribute to further characterize planetary systems evolution and understand the limits of Machine Learning for scientific discovery.

Zhehao Cheng

Supervisor Name: Yoko Yazaki-Sugiyama
Research Unit: Neuronal Mechanism for Critical Period Unit
Thesis: Neuronal Mechanism for Processing Complemental Information of Species Identity and Individual Variation in the Zebra Finch Higher Auditory Cortex

Abstract:
Animals transmit various information through vocal communications. Zebra finches, a songbird species extensively studied in laboratories, recognize and discriminate their own species identity as well as individual differences by hearing their songs. However, the neuronal mechanisms allowing them to detect individually unique songs in parallel with species common features have yet to be elucidated. Here, we found that zebra finch songs consist of acoustically similar common song elements overall, while they individually differ in element sequential arrangement from each other. We further revealed that each neuron in the higher auditory area, the caudal nidopallium (NCM), detected only a small subset of zebra finch songs, while NCM neuronal ensembles responded to all the zebra finch songs we presented by performing single unit electrophysiological recording in freely moving male zebra finches.
The NCM neurons selectively responded to specific song elements, which were similar in acoustical features between varieties of songs, while some of them were also sensitive to element sequential arrangement, which dramatically increases the capacities of variation with a limited number of element varieties. Taken together, our results suggest a neuronal mechanism for individual detection by sparse coding of individual NCM neurons to detect a small group of songs which are unique in sequence of elements but common in acoustic features.

Hung-Ju Chiang

Supervisor Name: Ichiro Masai
Research Unit: Developmental Neurobiology Unit
Thesis: Investigation of the Role of Male Germ Cell-Associated Kinase in Zebrafish Photoreceptor Ciliogenesis and Cell Survival

Abstract:
Photoreceptors are composed of highly specialized structures associated with phototransduction which takes place in the outer segments (OS), a specialized cilium of the cell. The connecting cilia play a pivotal role in mediating molecular transportation between the cell body and the OS. Defects in the ciliary structure can lead to various outcomes including cell death, however, the mechanisms regulating ciliary structure are not yet fully understood. Male germ cellassociated kinase (MAK) is a cilia-associated serine/threonine kinase, and its genetic mutation causes photoreceptor degeneration in mice and retinitis pigmentosa in humans.
Here, I report that zebrafish mak mutants show rapid photoreceptor degeneration during embryonic development. In mak mutants, both rod and cone photoreceptors completely lack the OS and undergo apoptosis. Interestingly, zebrafish mak mutants fail to generate axoneme during photoreceptor ciliogenesis, whereas the basal body and the transition zone are specified. These data suggest a specific role of MAK in axoneme development in zebrafish, which is opposite to mouse mak mutants showing the elongated axoneme of photoreceptors. Furthermore, the kinase activity of MAK plays a critical role in ciliary axoneme development and photoreceptor survival. Thus, MAK is required for ciliogenesis and OS formation in zebrafish photoreceptors to ensure intracellular protein transport and photoreceptor survival.

Maki Maeda

Supervisor Name: Sile Nic Chormaic
Research Unit: Light-Matter Interactions for Quantum Technologies Unit
Thesis: Design, Fabrication, and Characterization of Optical Nanofiber Cavities

Abstract:
Optical nanofibers (ONFs) pave the way for researchers to understand and control light-matter interactions at the nanoscale. Evanescent light fields at the waist of ONFs can be coupled to quantum emitters for various studies. In particular, significant experimental progress on quantum emitter coupling with an ONF-based cavity mode, for cavity quantum electrodynamics (cQED), has been achieved in the last two decades.
This thesis showcases the following topics: (i) the development of novel ONF cavity fabrication methods using a focused ion beam (FIB) which provides highly reflective cavity mirrors with more stable fabrication quality than conventional fabrication methods, (ii) the optical characterization of ONF cavities fabricated using the FIB technique, (iii) the investigation of the effects of laser annealing on the cavity modes, and (iv) the realization and characterization of an ONF-based cavity supporting higher-order modes (HOMs) and controlled excitation of the desired HOMs with polarization topology.
This is a significant step toward degenerate multimode cQED and potential access to evanescent field singular optics using a HOM-ONF. This PhD work advances ONF-based cavity systems, not only for cQED studies but also contributes to their fundamental understanding in quantum optics, structured light, and beyond.

Hsiao-Chiao Chien

Supervisor Name: Hiroki Ishikawa
Research Unit: Immune Signal Unit
Thesis: JunB is Essential for Chromatin Regulation in Pathogenic Th17 Cells

Abstract:
Epigenetic regulation and chromatin remodeling play critical role in governing gene expression during T helper (Th) cell differentiation. JunB, a member of the AP-1 transcription factor family, is required for autoimmune pathogenic Th17 transcriptional program, but its role in the chromatin regulation remains unknown. Using the assay for transposase-accessible chromatin sequencing (ATAC-seq), I found that JunB regulates chromatin remodeling in differentiation of diverse T helper subsets in vivo and in vitro.
Particularly, in pathogenic Th17 differentiation, JunB promotes chromatin accessibility at enhancer regions associated with a subset of Th17-related genes, including interleukin 17a (Il17a) and Il23 receptor (Il23r). JunB is necessary for BATF pioneering functions and/or the deposition of histone modifications, H3K4me1 and H3K27ac, through the recruitment of histone methyltransferase MLL4, and the recruitment of the BAF chromatin remodeling complex at these loci.
Moreover, JunB indirectly inhibits chromatin accessibility at loci related to fate switch from Th17 to Th1 cells, such as interferon gamma (Ifng), likely through inhibiting chromatin accessibility and expression of an immediate-early transcription factors, such as NF-kB, NR4A1, Ets and Runx family members. These results indicate that JunB is essential for enhancer activation of Th17-related genes, while indirectly inhibiting chromatin accessibility in diverse regions regulated by other key transcription factors.

Shukla Sarkar

Supervisor Name: Hiroki Ishikawa
Research Unit: Immune Signal Unit
Thesis: Analysis of the Function of JunB in Regulation of CD8+ T Cell Response

Abstract:
CD8+ T cells are essential for adaptive immune responses to eliminate virus-infected and cancer cells. T cell receptor (TCR)-stimulated naïve CD8+ T cells differentiate into cytotoxic effector CD8+ T cells. TCR-induced activator protein-1 (AP-1) subunit, BATF, and its interacting protein, IRF4, are required for clonal expansion of effector CD8+ T cell cells in response to acute infection of pathogens. BATF function is dependent on AP-1 dimeric partners, such as JunB. However, whether and how JunB regulates CD8+ T cell responses remains unknown. In this study, I found that JunB is induced upon TCR stimulation and essential for the clonal expansion of effector CD8+ T cells in response to Listeria Monocytogenes infection. JunB is required for survival and metabolic reprogramming to glycolysis in effector CD8+ T cells.
Furthermore, JunB promotes the expression of genes required for CD8+ T cell responses, such as Tcf7 and Runx3, while inhibiting the expression of a proapoptotic gene, Bcl2l11 (encoding Bim), and coinhibitory receptor genes, Pdcd1 and Havcr2 (encoding PD-1 and TIM3, respectively). JunB controls chromatin accessibility at genomic regions related to a subset of its target genes, including an enhancer region of Pdcd1. These results demonstrate that JunB is critical for transcriptional and epigenetic regulation for clonal expansion of effector CD8+ T cells.

Sergey Zobnin

Supervisor Name: Kenji Doya
Research Unit: Neural Computation Unit
Thesis: Investigation of the Bayesian Sensory-Motor Integration in the Cerebral Cortex

Abstract:
Animals need to sense their environment and differentiate sensory states to optimize behavior. For example, a force too weak may fail to pull a desired object closer, while too much effort will be unnecessarily exhausting. The prior experience interacting with similar objects can help generate the required action and produce an expectation of the sensory feedback. Combining prior expectation with the actual sensory observation is formalized with Bayes' rule. The neural mechanisms for such probabilistic computations remain unclear but, in mammals, can be attributed to the specific anatomical patterns of the cerebral cortex.
In particular, the pathways carrying bottom-up sensory signals mostly terminate in layer 4 and activate pyramidal neurons of the layer 2/3 of the sensory cortex, while the neurons of deeper layers 5 and 6 mostly receive projections from other cortical areas including the motor cortex. Based on this, I hypothesized that the pyramidal neurons of the superficial layers encode sensory evidence while the deeper layer neurons encode prior and posterior estimations. I conducted a series of lever-pulling experiments on mice, where the lever provided a variable tactile feedback. Calcium imaging of the S1 cortical area using a prism lens enabled me to analyze how matches or mismatches between action-dependent prediction and actual sensory inputs are encoded by neurons in superficial and deep layers.
I have identified functional asymmetry between superficial and deep cortical layers in the context of the probabilistic sensory-motor task. The deep layers were more associated with the prior information about the expected tactile stimulus in terms of the number of task coding neurons and the amount of information per given population size. In comparison with the deep layers, superficial layers were stronger associated with the sensory information about the actual tactile stimulus, but also encoded prior information. This research contributes to the study of the neural mechanisms underlying probabilistic estimation in the cerebral cortex.

Kota Ishikawa

Supervisor Name: Satoshi Mitarai
Research Unit: Marine Biophysics Unit
Thesis: Effects of Mean Flow and Turbulence on Planktivory by Anchored Garden Eels and Site-Attached Fish

Abstract:
Water flow is a key environmental factor that affects fish in coral reefs. Changes in flow properties, such as mean flow speed and turbulent fluctuation, have been expected to affect feeding of zooplanktivorous fish through its effects on the motions of both prey and predator. Flows are critical not only for freely swimming fish but also for anchored fish, such as garden eels, that feed while anchored to the sandy bottom by keeping the posterior parts of their bodies inside a burrow.
The objective of this study is to comparatively examine effects of flows on freely swimming reef fish and anchored garden eels to understand ecological traits, such as adaptation, habitat selection, and prey-predator interaction. To address this objective, I combined flow measurements in the field and flume experiments designed to examine effects of mean flow speed and small-scale turbulence on feeding and energy cost of reef damselfish (Chromis viridis) and garden eels (Heteroconger hassi). In situ flow measurements by acoustic Doppler velocimeter (ADV) indicated faster flow speed and stronger turbulence at damselfish habitat above corals compared with garden eel habitat on sandy flat bottom.Based on the in situ flow measurement, a range of mean flow speed and small-scale turbulence (e.g. dissipation rate) was reproduced in flumes to examine fish responses by label-free tracking of body points and 3D movement analysis. Relationship between feeding rate and flow speed showed an adaptation of damselfish to faster flow speed compared to garden eels. Energy cost and benefit model also indicated that the energetically optimal range of flow speed of damselfish was faster than that of garden eels. Detailed motion analysis revealed a unique strategy of garden eels against changes in flow speed, leading to the first foraging model I developed for this group of fish.
The anchored and site-attached fish also differed in their responses to small-scale turbulence: strong turbulence caused decrease in feeding rates under slow flows for damselfish and under fast flows for garden eels. The turbulence effect was associated with a reduction of the foraging area for damselfish and reduction of the search time for garden eels. By combining field flow measurements and flume experiments, my study revealed adaptations to hydrodynamic conditions in fish.

Markel Pardo Almanza

Supervisor Name: Yoshinori Okada
Research Unit: Quantum Materials Science Unit
Thesis: Fabrication and Spectroscopic Investigation of a Tunable Magnetic Material and its Heterostructures

Abstract:
The interplay between magnetism and the spin polarized surface states of topological insulators (TIs) can open a gap and give rise to the emergence of exotic states. Such states can be realized by the selective engineering of a magnetic material (MM)/TI heterostructure, inducing magnetic ordering via proximity effects. For this purpose, a tunable layered MM, Cr1+δTe2, is fabricated and investigated in this thesis.
A method to systematically control the δ content in molecular beam epitaxy grown thin films is developed. The effect of the δ content on the band structure is visualized by angle-resolved photoemission spectroscopy, revealing a rigid-band-like shift and the resulting tunable Hall properties. The magnetic properties are characterized, revealing the switch of the magnetic anisotropy with δ, explained by the increasing interlayer exchange interactions.
To conclude, Cr1.33Te2/Bi2Te3 heterostructures are fabricated and local variations on the magnetic proximity effect are investigated in real space, combining scanning tunneling microscopy/spectroscopy and cross-sectional transmission electron microscopy experiments. A correlation between a spatially inhomogeneous gap opening and distinct Bi2Te3 thicknesses is hypothesized. This thesis provides a universal insight into the direct exploration of quantum phenomena in magnetic TI heterostructures.

Sara Emad Mohamed Elagamy Abdelaal

Supervisor Name: Marco Terenzio
Research Unit: Molecular Neuroscience Unit
Thesis: Dynein Light Chain Roadblock 1 Regulates FMRP Axonal Transport and Degradation

Abstract:
The fragile X messenger ribonucleoprotein 1 (FMRP) is a multifunctional RNA binding protein (RBP) implicated in human neurodevelopmental and neurodegenerative disorders. FMRP mediates the localization and activity-dependent translation of its associated mRNAs through the formation of phase separated condensates that are trafficked by microtubule-based motors in axons. Axonal transport and localized mRNA translation are critical processes for long-term neuronal survival and are closely linked to the pathogenesis of neurological diseases. FMRP dynein-mediated axonal trafficking is still largely unexplored, but likely to constitute a key process underlying FMRP spatiotemporal translational regulation.
Here, we show that roadblock 1 (Dynlrb1), a subunit of the dynein complex, is a critical regulator of FMRP function in sensory neurons. In axons, FMRP associates with the dynein complex and is retrogradely trafficked in a Dynlrb1-dependent manner. Moreover, Dynlrb1 silencing induced FMRP granules accumulation and repressed the translation of Map1b, one of its primary mRNA targets. Our findings suggest that Dynlrb1 regulates FMRP function through the control of its transport and degradation.

Anjana Krishnadas

Supervisor Name: Yoshinori Okada
Research Unit: Quantum Materials Science Unit
Thesis: Spectroscopic Visualization of Surface Electronic State in High Temperature Superconducting Oxide Thin Films

Abstract:
The study of strongly correlated electron systems is one of the central topics in condensed matter physics. The phenomena occurring in these systems are so complicated that the unifying theory is far from complete, but their applications, such as high-temperature superconductors (HTSCs), have become an integral part of everyday life. The study of strongly correlated systems therefore bridges the gap between fundamental physics and the application of cutting-edge technology. Angle-resolved photoemission spectroscopy (ARPES) has proven to be a powerful tool to study surface electronic states.
Thus, we study the HTSCs- YBa2Cu3O7−x (YBCO) and the spinel superconductor LiTi2O4 (LTO) in thin őlms using ARPES. Despite their potential importance, the non-cleavable nature of both materials hinders the study of ARPES, and the electronic states on the surface have not yet been fully revealed.
For YBCO, we show that the doping of Y with Ca in YBCO exhibits anomalous carrier doping associated with the appearance of a folded chain-derived band dispersion, implying the possibility of controlling the surface state. On the other hand, the LTO has demonstrated its potential as an intriguing platform for exploring novel exotic states with geometrically frustrated pyrochlore lattices. Thus, our results open a rich playground for the future development of HTSC heterointerfaces with atomic design.

Sarah Yukie Nagasawa

Supervisor Name: Erik De Schutter
Research Unit: Computational Neuroscience Unit
Thesis: Stochastic Spatial Modeling of Vesicle Trafficking of AMPA Receptors to Understand Roles in Synaptic Plasticity

Abstract:
The regulation of AMPA-type glutamate receptors (AMPARs) in the synapse is central for maintaining the normal functioning of hippocampal synapses, and sustaining long-term storage of memories. Within post-synaptic neurons, an intracellular trafficking system of vesicles and endosomal structures exists that transports and delivers AMPARs to and from synapses. This trafficking system is believed to have an essential role in modulating synaptic plasticity.
However, it remains challenging to directly observe these fine processes in intact neurons during synaptic plasticity. In this thesis, I present a computational model that considers stochastic-spatial properties and biochemical signaling pathways of the intracellular trafficking system of AMPARs. I use the model to study vesicular-endosomal processes involved in the trafficking of AMPARs during synaptic plasticity.
Our model suggests (1) a time-delay in the contribution of endosomal-vesicle processes in enhancing synaptic strength, (2) an alternative source of AMPARs to endosomes is required to sustain increased synaptic strength, and (3) endosomes serve as storage sites to support multiple spines during synaptic plasticity.

Ianto Samuel Cannon

Supervisor Name: Marco Edoardo Rosti
Research Unit: Complex Fluids and Flows Unit
Thesis: Simulations of Multiphase Turbulent Flows

Abstract:
The flow of an incompressible fluid can be described exactly and succinctly using the Navier-Stokes equation. However, the nonlinearity of this equation leads to flow structures with detail at many length scales, known as turbulence. The only exact theory in turbulence was made in 1941 by Kolmogorov. In this thesis, we probe Kolmogorov's predictions in the case of multiphase systems by making direct numerical simulations of droplets, particles, and solid phases in turbulent flows.
Firstly, we show how the coalescence of droplets can reduce drag in a turbulent channel flow. Following on from this, in the remainder of the thesis we look more closely at the turbulent energy cascade using simulations of statistically homogeneous and isotropic flows. We show that isotropic and anisotropic particles couple to the cascade at different length scales. Next, we show that the plasticity of a fluid enhances its turbulent behaviour. Finally, we investigate the breakup and coalescence of droplets in the presence of the turbulent cascade.
The results presented in this thesis find cases in which Kolmogorov's predictions do not hold. In each case, we aim to explain why. These results can have wide-reaching implications for industry and the environment, including heart disease, micro-plastic dispersal, mudslides, and cloud formation.

Qiao Lu

Supervisor Name: Simone Pigolotti
Research Unit: Biological Complexity Unit
Thesis: Biophysical Modeling of Cas9 Target-searching and Recognition

Abstract:
In this thesis, I model the 1D diffusion and unbinding of Cas9 on/from DNA. Cas9 plays a key role in the CRISPR/Cas (CRISPR-associated protein) system. CRISPR (Clustered Regularly Interspaced Short Palindromic Repeats) are regions of a prokaryote DNA in which palindromic sequences are interspaced by sequences of foreign origins. These foreign sequences can be transcribed into guide RNAs. Cas9 is an enzyme that combines with guide RNA and afterwards can recognize and cleave DNA strands that are complementary to the guide RNA.
However, it is not clear how does Cas9 identify its target. The protospacer adjacent motif (PAM) is a “NGG” segment on DNA. The recognition of PAM is the initial stage of the target searching mechanism. Experiments suggest that Cas9 uses 3D diffusion combined with 1D diffusion along the DNA, a mechanism termed facilitated diffusion. My model explains the distribution of binding events observed in experiments and predicts biophysically relevant parameters. I then analyze the behavior of Cas9 on a generic DNA with disordered assortment of PAMs by using an analogy with Anderson localization in condensed matter physics. I then propose a model of the off-target behavior (specificity) of Cas9. From the measured rates, I determine the energy landscapes of on-target and off-target DNA sequences, and the thermodynamic parameters in double strand DNA and DNA-RNA hybrids. Finally, from a perspective of two-mode target recognition strategy, I investigate the effect of PAM and its binding energy on the efficiency of Cas9 in the facilitated diffusion process.

Vivek Pareek

Supervisor Name: Keshav Dani
Research Unit: Femtosecond Spectroscopy Unit
Thesis: Unraveling the Nature of Excitons and their Interactions through Time-Resolved Photoemission and Optical Spectroscopy

Abstract:
Unraveling the Nature of Excitons and their Interactions through Time-Resolved Photoemission and Optical Spectroscopy The exciton – a coulomb-bound electron-hole pair, was first conceptualized by Frenkel and Wannier in the 1930s. Since then, it has been integral to understanding the optoelectronic response in semiconductors, particularly in low-dimensional semiconductors. Therein, excitons have large binding energies due to quantum confinement and reduced dielectric screening, thus dominating the optical response of the material even at room temperature. Despite their impor- tance, a critical fundamental property of excitons remains inaccessible – the momentum of the constituent electrons and holes! Such a measurement would immediately reveal valuable information, such as their direct or indirect nature, their wave function, their size, and the nature of their interactions. Resolving the momentum coordinates of excitons requires the development of a new instrumentation platform that probes the excitons in time, energy, space, and momentum.
This thesis describes the need for such an instrumentation platform and its development, namely the development of time-resolved momentum microscopy. It then describes three studies on the nature of excitons and their interactions. First, we study the interlayer excitons in a WSe2/MoS2 heterostructure. Using time-resolved Momentum Microscopy, we resolve the momentum coordinates of the constituent electrons and holes within the interlayer exciton, directly measuring its size and confinement within the moir unit cell. Next, we demonstrate Floquet effects in monolayer WS2, in the absence of optical fields, resulting from the time-periodic oscillations in the electron self-energy due to excitons. The strong amplitude of the time-periodic perturbation allows us to observe the hybridization of the original band structure with the exciton-dressed one. Finally, we use traditional μ-TAS to study the exciton-exciton annihilation process in bilayer black phosphorus. We show that it is possible to alter the dimensionality of the exciton-exciton annihilation process from one dimensional-like to two dimensional-like by tuning the exciton density and temperature. In conclusion, this thesis answers some fundamental questions about excitons and their interactions in two dimensional semiconductors and paves the way for uncovering novel non-equilibrium phenomena in two dimensional materials.

Hunter Daniel Barbee

Supervisor Name: Keiko Kono
Research Unit: Membranology Unit
Thesis: Plasma Membrane Damage-Dependent Senescent Cells Accelerate Wound Healing In Vitro via Soluble Molecules and Increased Extracellular Vesicles

Abstract:
Cellular senescence is an essentially irreversible cell cycle arrest that contributes to physiological and pathological processes in vivo, including organismal aging, development, cancer progression, and wound healing. The three best-studied triggers of cellular senescence are telomere shortening, oncogene activation, and DNA damage. We previously showed that transient plasma membrane damage induces a novel subtype of senescence (PMDS) in normal human fibroblasts. However, the pathophysiological functions of PMDS cells remain unknown. Here, we investigated the paracrine functions of PMDS cells, via extracellular vesicles (EVs) and soluble molecules, and found that PMDS cells accelerate wound healing in vitro. We first measured the number of EVs and found that PMDS cells produce more EVs than DNA damage repair-dependent senescent (DDRS) cells do.
After co-culturing PMDS cells with young fibroblasts in a scratch assay, we observed an in vitro acceleration of wound healing, more significantly than with DDRS cell co-culture. We next separated the EV fraction and the soluble fraction of media from PMDS cells and tested their ability to accelerate wound healing. We found that these two fractions synergistically contribute to wound healing. Mass spectrometry analysis of PMDS cell-derived EVs revealed that wound healing regulators and Ca2+-signaling regulators are enriched in these EVs compared to EVs from young cells. Altogether, these results revealed that PMDS cells contribute to wound healing in vitro via EVs and soluble factors. Our findings raise the possibility that PMDS cells may accelerate tissue repair in vivo.

Anzhelika Koldaeva

Supervisor Name: Simone Pigolotti
Research Unit: Biological Complexity Unit
Thesis: Population Dynamics of Microorganisms in Spatially Structured Environments

Abstract:
Microbial populations live and grow in spatially structured environments. These structures lead to spatial patterns in populations and alter the course of their natural evolution. Such phenomena are theoretically studied using spatially explicit models. However, these models are still poorly understood due to their analytical and numerical complexity. In this thesis, we study two systems of microorganisms living and proliferating in different spatially structured environments. The first system consists of populations of Escherichia coli growing in rectangular microchannels with two open ends. We study such populations with a lattice model in which cells shift each other along lanes as they reproduce. The model predicts rapid diversity loss along the lanes, with neutral mutations appearing in the middle of the channel being the most likely to fixate.
These theoretical predictions are in agreement with our experimental observations. The second system is constituted by planktonic microorganisms that are transported by chaotic oceanic currents. To replicate their dynamics, we employ an individual-based coalescence model. The model predicts the effect of oceanic currents on the biodiversity of planktonic communities, as observed in metabarcoding data sampled from oceans and lakes around the world.

Barnaby Smith

Supervisor Name: Yoshinori Okada
Research Unit: Quantum Materials Science Unit
Thesis: Imaging Electronic States in the Charge Density Wave Material CeTe3

Abstract:
The study of condensed phases plays a crucial role in our understanding of physics. In particular, Charge Density Waves (CDWs) are an important phenomenon due to their close connection with a material's fermiology, in addition to their coexistence and competition with superconducting, magnetic, and other exotic phases. Rare earth tritellurides (RTe3) are a prototypical 2D material system which provide an excellent test platform for investigating these phases and the interplay between them.
In this thesis, I will present the first low temperature STM study of CeTe3, with the fist comprehensive discussion and simulation of quasiparticle interference (QPI) in any RTe3 compound. At 4.2 K, we confirm that CeTe3 shows the unidirectional CDW that is present at higher temperatures. We then use SX-ARPES to fit an interacting tight binding model and demonstrate that the observed QPI can be interpreted with JDOS simulations based on this model band structure. We reveal the unexpected result that the QPI arises mainly from scattering between the original bands and their shadow band replicas, with backscattering being relatively suppressed. This pioneering study can be powerfully extended in the future to use QPI measurements to understand the interaction of the CDW state and other phases.

Rachapun Rotrattanadumrong

Supervisor Name: Yohei Yokobayashi
Research Unit: Nucleic Acid Chemistry and Engineering Unit
Thesis: Machine Learning Guided Exploration of an Empirical Ribozyme Fitness Landscape

Abstract:
Fitness landscape of a biomolecule is a representation of its activity as a function of its sequence. Properties of a fitness landscape determine how evolution proceeds along the landscape. How functional variants are distributed and, more importantly, how connected these variants are within the sequence space are important scientific questions. Better understanding of these properties has important implications for evolutionary theory and biotechnology. Exploration of these spaces, however, is impeded by the combinatorial explosion of the sequence space. High-throughput experimental methods have recently reduced this impediment but only modestly. Better computational methods are needed to fully utilise the rich information from these experimental data to better understand the properties of the fitness landscape. In this work, I seek to improve this navigation process by combining data from massively parallel experimental assay with advanced computational techniques for smart library design. I focus on an artificial RNA enzyme or ribozyme that can catalyse a ligation reaction between two RNA fragments. This chemistry is analogous to that of the modern RNA polymerase enzymes, therefore, represents an important reaction in the origin of life.
In the first chapter, I discuss the background to this work in the context of evolutionary theory of fitness landscape and its implications in biotechnology. In chapter 2, I explore the use of processes borrowed from the field of evolutionary computation to solve optimisation problems using real experimental sequence-activity data. In chapter 3, I investigate the use of supervised machine learning models to extract information on epistatic interactions from the dataset collected during multiple rounds of directed evolution. I investigate and experimentally validate the extent to which a deep learning model can be used to guide a completely computational evolutionary algorithm towards distant regions of the fitness landscape. In the final chapter, I perform a comprehensive experimental assay of the combinatorial region explored by the deep learning-guided evolutionary algorithm. Using this dataset, I analyse higher-order epistasis and attempt to explain the increased predictability and navigability of the region sampled by the algorithm.
Finally, I provide the first experimental evidence of a large ‘neutral network’ connecting two functional ribozyme variants. This property of an RNA fitness landscape has been theoretically predicted, but no experimental evidence has been provided until now. Altogether, this work represents the most comprehensive experimental and computational study of the RNA ligase ribozyme fitness landscape to date, providing important insights into the evolutionary search space possibly explored during the earliest stages of life.

Osamu Horiguchi

Supervisor Name: Hiroshi Watanabe
Research Unit: Evolutionary Neurobiology Unit
Thesis: Identification and Functional Analysis of Group A bHLH Transcription Factor in Ctenophore Bolinopsis mikado

Abstract:
Animals have acquired various animal-specific cell types, such as neurons, during their early evolutionary phase. However, the evolutionary origin of these cells is largely unknown. In Bilateria and Cnidaria, the bHLH transcription factors, particularly those belonging to group A, are known as important transcription factors in cell differentiation including neurogenesis. On the other hand, the function of group A bHLH in Ctenophora, the earliest divergent extant animal bearing neurons, remains unexplored.
In this study, I identified and analyzed the function of group A bHLH in Ctenophora to get new insight into the evolutionary origin of neurons. I first attempted to identify group A bHLH from ctenophore. Molecular phylogenetic analysis identified six group A ctenophore bHLHs. Next, using Bolinopsis mikado, a species in which the gene knockdown method was established, I performed functional analysis of BmbHLH1 gene. BmbHLH1 is the B. mikado group A bHLH gene highly expressed during embryogenesis and larval development. Knockdown of BmbHLH1 resulted in expression changes in 106 genes. Among these genes, 28 were peptidases cathepsin. Cathepsin genes whose expression was reduced by BmbHLH1 knockdown were expressed in the digestive systems, where many neuropeptides are expressed, suggesting that they are involved in neuropeptide processing.
The results of this study may provide key insights into the evolutionary origin of peptide-expressing neurons, which are assumed to reside in ancestral neuronal states.

Thomas Francis Burns

Supervisor Name: Tomoki Fukai
Research Unit: Neural Coding and Brain Computing Unit
Thesis: Geometry and Topology in Memory and Navigation

Abstract:
Geometry and topology offer rich mathematical worlds and perspectives with which to study and improve our understanding of cognitive function. Here I present the following examples: (1) a functional role for inhibitory diversity in associative memories with graph- ical relationships; (2) improved memory capacity in an associative memory model with setwise connectivity, with implications for glial and dendritic function; (3) safe and effi- cient group navigation among conspecifics using purely local geometric information; and　(4) enhancing geometric and topological methods to probe the relations between neural activity and behaviour. In each work, tools and insights from geometry and topology are used in essential ways to gain improved insights or performance. This thesis contributes to our knowledge of the potential computational affordances of biological mechanisms (such as inhibition and setwise connectivity), while also demonstrating new geometric and topological methods and perspectives with which to deepen our understanding of cognitive tasks and their neural representations.

Shubham Deolka

Supervisor Name: Julia Khusnutdinova
Research Unit: Coordination Chemistry and Catalysis Unit
Thesis: Naphthyridines as Versatile Ligand Scaffolds For Metal-Metal Cooperative Bond Activation and Photocatalysis

Abstract:
In this thesis, I will discuss the application of naphthyridine-based ligands as binucleating or mononucleating ligands scaffolds for application in selective bond activation and photocatalysis. In the first part, I will discuss the design of new polynucleating ligands based on unsymmetrically substituted naphthyridines that were utilized for site-selective construction of bimetallic Pt/Cu or multimetallic Pd/Cu complexes. These complexes were then utilized for selective bond activation or transmetalation via metal-metal cooperation. I will also discuss the synthesis of high valent mononuclear Ni(III) complexes supported by simple unsubstituted or Me-substituted naphthyridines via aerobic oxidation and the application of these complexes in photocatalyzed C-H bond trifluoromethylation. In the final part, I will focus on the reactivity of Ni perfluoroalkyl complexes featuring longer perfluoroalkyl chains in oxidation and ligand-free perfluoroalkylation of C-H bonds.

Otis Davey Brunner

Supervisor Name: Satoshi Mitarai
Research Unit: Marine Biophysics Unit
Thesis: The Role of Connectivity in Structuring Community Composition and Diversity at Hydrothermal Vents Across the Northwest Pacific

Abstract:
Connectivity, or the movement of individuals among isolated habitat patches, promotes local and regional biodiversity, and its resilience to disturbances both natural and anthropogenic. Species associated with seafloor hydrothermal vent habitats are distinctly reliant on connectivity due to their spatial restriction to the point source of chemical energy from vent chimneys that fuels their chemosynthetic food web. Measuring connectivity among hydrothermal vents is particularly urgent in regions where mining of these ecosystems is imminent. Our understanding of connectivity is limited by the scarcity of observational data from these inaccessible deep-sea ecosystems. Modelling is a viable alternative to the study of connectivity as the dispersal that facilitates connectivity is mostly dictated by predictable ocean currents, which can be reliably simulated.
This thesis combines empirical observations of species’ distributions and environmental conditions at hydrothermal vents with simulations of dispersal, to model connectivity among vent sites in the Northwest Pacific. First, I curate the most comprehensive regional dataset of hydrothermal vent species distributions to infer connectivity in the form of a species assemblage network (Chapter 1). I then simulate how the planktonic larvae of vent species disperse among the vent sites in this region using Lagrangian particle tracking methods within an Ocean General Circulation Model (Chapter 2). Finally, I combine the among-site dispersal estimates with observations of local environmental parameters to create a simulated species assemblage network using a metacommunity model (Chapter 3). This metacommunity model accurately recreated the empirical observations from chapter 1 and gives crucial insight into the interacting effects of dispersal barriers and environmental niche on driving diversity and community composition patterns at hydrothermal vents.
Furthermore, I used the combination of observed and simulated connectivity results to quantitatively evaluate the relative role each individual hydrothermal vent plays in maintaining connectivity and biodiversity in the region. Such an evaluation has critical and timely implications for proposed mining and the spatial management of hydrothermal vents in this region. Lastly, we demonstrate that hydrothermal vents are natural laboratories for the advancement of metacommunity theory and conservation ecology due to their characteristic isolation and discrete nature.

Otis Davey Brunner

Theodoros Bouloumis

Supervisor Name: Sile Nic Chormaic
Research Unit: Light-Matter Interactions for Quantum Technologies Unit
Thesis: Metamaterial Plasmonic Tweezers for Enhanced Nanoparticle Trapping

Abstract:
Optical tweezers have gained significant attention in many research fields as the only technique that provides immobilisation (trapping) and manipulation of micro- and nanoparticles. Moving from the conventional, free-space configuration to plasmonic structures using strong near-field forces, resulted in many more avenues towards the exploration of the nanoworld. However, with that, many challenges also appeared, as is usually the case when pushing the boundaries of the unknown. In this thesis, we focus on how to achieve an efficient trap stiffness value for particles of just a few nanometres in size, such as colloidal quantum dots and gold nanoparticles. For this purpose, we investigate a novel metamaterial plasmonic design that exhibits a sharp plasmonic Fano resonance feature, which is very sensitive to refractive index changes of its environment. Three main projects are presented. In the first one, we work on the optimisation of the basic characteristics of the metamaterial, to ensure it has the desired plasmonic resonance and exhibits strong optical forces. We test its efficiency by trapping 20 nm polystyrene particles, yielding very high trap stiffness values. We also perform sequential trapping, revealing the ability of the structure for on-demand, particle nanopositioning. In the second project, we study the mechanism of self-induced back- action trapping. Under certain conditions, the particle can contribute to its own trap through an optomechanical coupling of its motion with the intracavity light intensity of the plasmonic nanocavity. For this experiment, gold nanoparticles were used and successfully trapped with extremely low laser intensities. Finally, the third project addresses the trapping of custom-synthesised organic quantum dots that can be tuned to the desired size and emission wavelength according to the expected application. Photoluminescence measurements are also performed and an overall evaluation of the applicability and potential uses of these quantum dots is discussed.

Nadine Wirkuttis

Supervisor Name: Jun Tani
Research Unit: Cognitive Neurorobotics Research Unit
Thesis: Social Interaction Under the Free Energy Principle: A Cognitive Robotics Approach

Abstract:
Although many investigate cognitive mechanisms of social behaviors, it is still not fully understood how individual and collective dynamics allow us to coordinate our actions in a wide range of social settings because of their complex interactive nature. Recently, the free energy principle has drawn large attention with an expectation that it could provide a unified theory of the brain to model action, perception, and learning. This work investigates how perception, action, and perception of actions of two agents can translate into a mutual, social context. I use a neurorobotics experimental setup to systematically study dyadic synchronized imitative interaction by extending the frame-works of predictive coding and active inference based on the free energy principle. In the proposed model, the top-down intention with the prior belief acts on the counterpart by predicting its future actions. On the other hand, the bottom-up inference for sensations of the actions generated by the counterparts updates the posterior beliefs by means of minimizing free energy. In a wide range of experiments, I explored how regulating free energy complexity, i.e., the divergence of the prior and posterior belief, during the robot learning phase as well as during robot interaction experiments, can guide behavior coordination in the robot imitative interaction context. The analysis of the individual as well as collective behaviors of robots allowed us to identify how individual dynamics can translate into three distinct interactive behavioral categories including: (1) one agent is leading and the other robot is following (and vice versa), (2) both agents are taking turns in leading and following, and (3) both agents are ignoring each other. The thesis clarifies the underlying mechanisms accounting for how these social behavior categories emerge and how turn-taking develops.

Stefano Pascarelli

Supervisor Name: Paola Laurino
Research Unit: Protein Engineering and Evolution Unit
Thesis: Protein Sequence, Structure, and Dynamics Reveal Insights in the Divergence of Protein Functions

Abstract:
Proteins participate in every important aspect of known living. The amino acid sequence of which a protein is composed contains information about the physicochemical properties, the three-dimensional structure, and its function. However, connecting protein sequence to function is still an open challenge, particularly for protein families with complex inter-relationships, i.e., heteromeric interactions. Such is the case of the Epidermal Growth Factor (EGF) receptor system, comprising four or more paralogs of the EGF receptor interacting with seven or more paralogs of the peptide ligand. In this thesis, I use the evolutionary history of the EGF receptor system to show how phylogenetic patterns of evolution relate to functional divergence at the protein sequence level. By combining measures of residue conservation and residue co-evolution I developed a method to identify residues responsible of a specific protein function. Mutations on the residues highlighted by this method altered the auto-phosphorylation level of the EGF receptor and affected cellular growth. Next, I studied a fish-specific gene duplication of the EGF receptor and used it to describe and model a rare pattern of sequence evolution. I showed that this pattern could be related to functional divergence, thus providing a way to identify the occurrence of the event and the residues responsible of it. Ultimately, I analyzed whole protein families using protein similarity networks. My results showed how the networks made from structural similarity of predicted 3D-models give a better representation of the protein functions compared to sequence similarity networks, thus supporting a paradigm shift from sequence-based to predicted-structure-based bioinformatics software. Overall, my thesis shows a deep interconnection between functional divergence and protein sequence evolution that can be exploited for prediction of function or identification of evolutionary events. The conceptual foundations of this study could be used in other fields where gene duplication and functional residues play an important part, as for example protein engineering and the study of copy number variation in cancer biology.

Tsung-Yen Huang

Supervisor Name: Hiroki Ishikawa
Research Unit: Immune Signal Unit
Thesis: Phosphoenolpyruvate Regulates the JunB-Dependent Pathogenic Th17 Transcriptional Program

Abstract:
Aerobic glycolysis, a metabolic pathway essential for effector T cell survival and proliferation, regulates the differentiation of autoimmune T helper (Th)17 cells, but the mechanism underlying this regulation is largely unknown. Here, we identify a glycolytic intermediate metabolite, phosphoenolpyruvate (PEP), as a negative regulator of Th17 differentiation. PEP supplementation or inhibition of downstream glycolytic enzymes in differentiating Th17 cells increases intracellular PEP levels and inhibits expression of Th17 signature molecules, such as IL-17A.
However, PEP supplementation does not significantly affect metabolic reprogramming, cell proliferation, and survival of differentiating Th17 cells. Mechanistically, PEP regulates JunB-dependent pathogenic Th17 transcriptional program by inhibiting DNA-binding activity of the JunB/BATF/IRF4 complex. Furthermore, daily administration of PEP to mice inhibits generation of Th17 cells and ameliorates Th17-dependent autoimmune encephalomyelitis. These data demonstrate that PEP links aerobic glycolysis to the JunB-dependent pathogenic Th17 transcriptional program, suggesting the therapeutic potential of PEP for autoimmune diseases.

Yuna Hattori

Supervisor Name: Pinaki Chakraborty
Research Unit: Fluid Mechanics Unit
Thesis: Self-similarity in a Boundary-layer Flow over a Dynamic Boundary: Flow of Air Induced by a Falling Soap Film

Abstract:
A wide range of dynamical phenomena in nature are self-similar. This remarkable property entails that scaled versions of a phenomenon conform onto themselves. It not only affords simplified mathematical analysis but also reveals the physical underpinnings of the phenomenon. In fluid flows, a textbook example of such phenomena is the boundary-layer flow over a rigid boundary---the Blasius boundary layer flow. In this thesis, we experimentally and theoretically study self-similarity in boundary-layer flow over a dynamic boundary, wherein the flow and the boundary are dynamically coupled. Our experimental setup is a soap-film channel, which is essentially a soapy waterfall---a planar film of soap-water solution falling under gravity. This setup has long been used to study quasi-two-dimensional flows in a laboratory setting.
Unlike previous experiments, however, where the focus is on the flow in the film, we train attention on what surrounds the film: air. The falling film drags the surrounding air, inducing flow in a thin layer of air adjacent to the film. This flowing air, in turn, resists the motion of the falling film; thus, the film-air interface is a dynamic boundary. We measure the velocity profile of the airflow in the boundary-layer of this interface using super-resolution Particle Image Velocimetry. (To our knowledge, these are the first experiments to measure airflow induced by a soap film.) The downstream evolution of the air velocity profile manifests self-similarity, which we analyze using the framework of boundary-layer theory. Surprisingly, we find that the conditions of self-similarity of the airflow also shed light on the downstream evolution of the film. Beyond air-film interaction, our findings may bear on a broader class of flows over dynamic boundaries, e.g. ocean-air interaction.

Friederike Metz

Supervisor Name: Thomas Busch
Research Unit: Quantum Systems Unit
Thesis: Machine Learning Applications for the Study and Control of Quantum Systems

Abstract:
Machine learning has achieved remarkable successes in recent years ranging from agents beating the best human players at games, to deep neural networks that locate and classify objects in images with unprecedented accuracies, and to powerful language models that compose texts indistinguishable to those written by humans. Hence, it is not surprising that machine learning is seeing increasingly diverse applications within numerous areas of the physical sciences such as quantum and condensed matter physics. In this thesis we consider the three main approaches of machine learning: supervised, unsupervised, and reinforcement learning, and explore how each can be employed as a tool to study or control quantum systems. To this end, we adopt classical machine learning methods, but also illustrate how present-day quantum devices and concepts from condensed matter physics can be harnessed to adapt the machine learning models to the physical system being studied.
In the first project, we use supervised learning techniques from classical object detection to locate quantum vortices in rotating Bose-Einstein condensates which enables the study of interesting out-of-equilibrium phenomena such as turbulence and chaos. We show that the machine learning model achieves high accuracies even in the presence of noise, which makes it especially suitable in experimental settings. Then, we move on to the field of unsupervised learning and develop a quantum anomaly detection framework based on parameterized quantum circuits to map out phase diagrams of quantum many-body systems. The proposed algorithm allows quantum systems to be analyzed on a quantum computer without any prior knowledge about the phases and phase transitions. Lastly, we consider two reinforcement learning applications for quantum control. In the first example, we use Q-learning to maximize the entanglement in discrete-time quantum walks, which can serve as a resource for quantum communication protocols. In the final project, we develop a novel approach for controlling quantum many-body systems by leveraging matrix product states for representing the quantum system and as a trainable machine learning ansatz for the reinforcement learning agent. This framework enables us to reach far larger system sizes than conventional neural network-based approaches while retaining the advantages of deep learning methods such as generalizability and robustness to noise.

Qiong Huang

Supervisor Name: Kenji Doya
Research Unit: Neural Computation Unit
Thesis: Multi-Agent Reinforcement Learning for Distributed Solar-Battery Energy Systems

Abstract:
Efficient utilization of renewable energy sources, such as solar energy, is crucial for achieving sustainable development goals. As solar energy production varies in time and space depending on weather conditions, how to combine it with distributed energy storage and exchange systems with intelligent control is an important research issue. In my thesis, I explore the use of reinforcement learning (RL) for adaptive control of energy storage in local batteries and energy sharing through energy grids depending on local energy demands.
I first extend the Autonomous Power Interchange System (APIS) from SONY to combine it with reinforcement learning algorithms in each house. I then consider different design decisions in applying RL: whether to use centralized or distributed control, at what level of detail actions should be learned, what information is used by each agent, and how much information is shared across agents. Based on these considerations, I implemented deep Q network (DQN) to set the parameters of real-time energy exchange protocol of APIS and tested it using the actual data collected from OIST DC-based Open Energy System (DCOES) and found that DQN agents outperform rule-based control on energy sharing and that prioritized experience replay further improves the performance of DQN. Simulation results also suggest that sharing average energy production, storage and usage within the community helps the performance. The results contribute to future designs of distributed intelligent agents and the effective operation of energy grid systems.

Aleksandra Bliznina

Supervisor Name: Nicholas Luscombe
Research Unit: Genomics and Regulatory Systems Unit
Thesis: Cross-genome Comparison of Global Oikopleura dioica Populations

Abstract:
Larvaceans represent the second most abundant zooplankton in all the world’s oceans, with key roles in marine food chains and global carbon flux. Oikopleura dioica is a free-swimming planktonic tunicate from the group and possesses the smallest animal genome with extremely dynamic organization: multiple genomic features such as transposon diversity, intron repertoire, gene content and order are altered in Oikopleura compared with other metazoans. Intriguingly, such genome reorganization has not affected the preservation of their ancestral morphology, since O. dioica maintains a chordate-like body plan throughout its life. Oikopleura dioica can be easily distinguished from other larvaceans mainly based on separate sexes and the presence of two subchordal cells on its tail. My research is focused on the cross-genome comparison of three O. dioica populations sampled from the Northern hemisphere: one in Europe (Barcelona/Bergen) and two in Japan (Osaka/Aomori and Okinawa/Kume). For each population, I generated high-quality genome assemblies using a combination of short- and long- read sequencing technologies, as well as chromatin conformation data, confirming preservation of three chromosome pairs. A pairwise comparison of populations revealed a striking degree of genome reshuffling that involves a vast number of synteny breaks and rearrangements. My research also shows that rearrangements mostly happen within individual chromosomes and generally preserve protein-coding features, such as genes and their constituent exons, although the gene order has been effectively randomized. Oikopleura dioica populations exhibit differences in repeats and gene content that affect even evolutionary conserved clusters, such as Hox genes. Consistent with an increased evolutionary rate, the accumulation of rearrangements in O. dioica appears to have happened much faster than in other animals and resulted in the divergence of multiple lineages of dioecious Oikopleura. The fact that their morphology stayed virtually identical makes O. dioica a perfect model to study genotype-phenotype correlation and the possible existence of unknown regulatory mechanisms. Overall, my thesis contributes new insights into the evolution of chordate genomes and, thus, may be interesting beyond the field of Oikopleura research.

Christopher Campbell

Supervisor Name: Thomas Busch
Research Unit: Quantum Systems Unit
Thesis: Supersymmetry and Nonequilibrium Quantum Dynamics

Abstract:
Supersymmetric algebras can be used in quantum mechanics to create and describe systems with very special properties and many examples of this exist in the area of quantum optics and ultracold quantum gases. The formalism of supersymmetric quantum mechanics factorizes a given Hamiltonian using a set of operators unique to it, and then constructs a supersymmetric partner Hamiltonian from these. Supersymmetric partner Hamiltonians turn out to possess almost the same eigenspectrum, leading to the existence of intertwining relationships between their respective Hilbert spaces. In this thesis I will present my work on using these intertwining relationships between partner Hamiltonians to study the quantum dynamics of different cold atomic systems. In the first project I explore the non-equilibrium dynamics following a quench to a many-body system between two supersymmetric partner potentials, and show how the similarity of the respective spectra affects the survival probability and the work distribution at zero and finite temperature. In the second project I apply the formalism to shortcuts to adiabaticity protocols and derive an intertwining relationship for the counterdiabatic terms in a hierarchy of Hamiltonians derived from supersymmetry.

Mohamed Atwa

Supervisor Name: Yoshinori Okada
Research Unit: Quantum Materials Science Unit
Thesis: Doping Evolution of Magneto-Transport Properties in the Layered Magnetic Semimetal Cr(1+δ)Te2

Abstract:
Layered magnetic materials are important both from the fundamental perspective of understanding charge, heat, and magnetism as well as from the technological perspective of magnetically enhanced thermoelectric energy generation. Cr (1+δ) Te 2 is a recently rediscovered magnetic transition metal chalcogenide (TMC) wherein controlling the fraction of Cr (δ) systematically tunes the electronic and magnetic properties of this simple binary system. This thesis reports mainly on modulating δ to tune the longitudinal thermopower S xx for different Cr 1+δ Te 2 compositions. We show that as the fraction of doped Cr(δ) increases between δ = 0.3 and δ = 0.68, the sign of S xx changes from positive to negative at a critical doping level of δC ≈ 0.5. The observed doping-dependent trend in the thermopower is consistent with the evolution of the semimetallic band structure in this material from ARPES, corroborating the electronic tunability of Cr (1+δ) Te 2 using multiple characterization techniques. Next, an anomalous enhancement in thermoelectric response is also reported around δC,
stemming from strong charge-spin coupling. Antiferromagnetic magnons are uncovered as the origin of this thermopower enhancement from analyses of the temperature-dependent magnetothermopower. This picture is further supported by the correspondence of the doping trend of the magnetothermopower with that of the magnetic anisotropy. The findings of this thesis point collectively to the critical nature of the doping level δC in magnetic semimetal Cr 1+δ Te 2. Around δC, near-Fermi-energy pseudogap formation and antiferromagnetic magnons combine to enhance thermoelectric energy conversion in Cr 1+δ Te 2.

Lewis Ruks

Supervisor Name: Thomas Busch
Co-Supervisor: Sile Nic Chormaic
Research Unit:Quantum Systems
Thesis: Wave Propagation and Light-Matter Interactions in Optical Nanofibers and Discrete Media

Abstract:
Building on more than 50 years of sustained progress, artificial systems of atoms and photons are now routinely controllable down to the nanoscale, which paves the way for simulators and processors powered by the light-matter interaction. In particular the rapid experimental progress made in platforms of nanoscale photonics and neutral atoms de- mands fresh computational studies along with more powerful theoretical tools in order to simulate these increasingly complex (quantum) optical systems. In this thesis I contribute to (1) the understanding of current state-of-the-art in experimental optical nanofiber sys- tems on one hand, and to (2) the general theory of emission into photonic lattices together with (3) quantum metrology in light-matter platforms. In the former I (1) systematically investigate light propagation in coupled optical nanofibers fibers and dispersion poten- tial mediated through these nanofibers for experimentally relevant parameters, shedding light on effects that may be observed in near-future setups. In the latter I (2) study hyperbolic lattices exhibiting strongly anisotropic emission, with results that may have applications in transporting and storing photons in nanoscale platforms. Additionally, in a collaborative work (3) a proposal is made for a metrological protocol consisting of quenching through a quantum phase transition to obtain quantum-limited precision in system measurements.

Kimberly Remund

Supervisor Name: Nic Shannon
Research Unit: Theory of Quantum Matter Unit
Thesis: Spin–1 Magnets and Their Excitations

Abstract:
Nature sometimes arranges itself in extremely curious ways, sowing the seed of very intriguing physics. Magnetic systems offer a rich variety of interesting features. They are traditionally studied in either their classical (S → ∞), or their extreme quantum limit (S=1/2). However, magnetic degrees of freedom in spin systems span within a 2 whole spectrum range and do not necessarily reduce to the specific case found at the extremities. Spin–1 magnets provide a good example of what happens to ground state and excitations properties for such instance. Indeed, a spin–1 is special, in the sense that, besides displaying dipolar degrees of freedom, a spin–1 can also exhibit on-site quadrupolar degrees of freedom, while retaining its quantum characteristics. Therefore, spin–1 systems are often used as examples to refer to spin-nematic order in magnetic insulators, Fe-based superconductors, or cold atoms. Unlike for spin-1/2, which in the classical limit can be represented by an O(3) vector, for spin–1, an O(3) vector does not completely describe all of what a spin–1 can do, namely intrinsically exhibiting quadrupoles. In this Thesis, I develop a united framework that enables us to treat dipolar and quadrupolar degrees of freedom of a spin–1 moment on an equal footing. My method is based on the extension of the usual su(3) algebra describing a quantum spin–1 into the u(3) algebra. Within the u(3) formalism, I derive equations of motion (EoM) for the objects living in the u(3) algebra. The u(3) approach enables the appropriate formulation for both classical and quantum derivations. Moreover, the EoM take a simple form suitable for numerical implementation. I illustrate this method by applying it to the well-known Bilinear-Biquadratic model on the triangular lattice for the ferroquadupolar state. This study is supported through classical low-temperature expansion in order to probe the thermodynamical properties, as well as quantum multi-bosons theory that allows to access dynamics. These results are validated by comparison with numerical simulations classical Monte Carlo (MC) and Molecular Dynamics (MD) respectively, both expressed in terms of u(3) objects. I show that at sufficiently low temperature numerical simulations can corrected for the classical statistics, the fully quantum zero-temperature analytical results are retrieved. Additionally, I confirm that our method is also applicable to anisotropic models, which is of experimental relevance. Finally, some new ideas, including the description of topological defects in spin–1 magnets and the generalization of the commonly used Self-Consistent Gaussian Approximation to the degrees of freedom of a spin–1 expressed within our u(3) framework are explored.

Manana Kutsia

Supervisor Name: Ichiro Masai
Research Unit: Developmental Neurobiology Unitt
Thesis: Single cell transcriptome analysis reveals heterogeneity and a dynamic regenerative response of quiescent radial glia in zebrafish adult brain

Abstract:
Neural regeneration in response to brain damage is an essential topic in medical science. In general, humans have a low regenerative capability. On the other hand, zebrafish show a remarkable ability to regenerate neural tissue in response to various types of brain injury. Radial glial cells(RG) that comprise the adult neural stem cell(aNSC) population in zebrafish telencephalon produce non-glial- neural precursor cells that potentially replace the lost neuronal population. However, there is no characterization of how diverse subpopulations of glial cells respond to the injury. We applied single-cell transcriptomics to RG in zebrafish adult telencephalon and identified five RG subtypes, which are classified into four quiescent RG (qRG) and one proliferating RG (pRG). One RG subtype shows high expression of ribosomal proteins, and its fraction is increased in response to brain damage. Consistently, the mTOR pathway is activated in RG near the injury site. It was reported that inflammatory response of brain-resident immune cells, microglia, is required for inducing regenerative response of RG in zebrafish. Genetical elimination of microglia not only suppressed damage-induced regenerative response of RG but also decreased fraction of the ribosomal expression-enriched RG. Lastly, our pseudo-time analysis revealed a lineage to generate the ribosomal expression-enriched RG. Our analysis reveal heterogeneity of qRG in zebrafish adult brain and their dynamic regenerative response to brain damage.

Maki Kohata Thomas

Supervisor Name: Satoshi Mitarai
Research Unit:Marine Biophysics Unit
Thesis: Role of Island Systems in Mangrove Biogeography

Abstract:
Studies on mangrove population connectivity have focused primarily on global to regional scales and have suggested a potential for long-distance connectivity, with archipelagos serving as steppingstones for trans-oceanic dispersal. To investigate the ecological roles of mangroves on island systems for their biogeography, we utilized network analysis to assign each location to biogeographical modules, and further analyzed for modularity roles in each area in the Indo-West Pacific region where >70% of mangrove species reside. All locations were placed into three modules in the region and were assigned with "non-hub" modularity roles which indicated that the mangrove is in general more isolated compared to other coastal species such as corals and seagrass. We further focused on one island system, the Ryukyu Archipelago in Japan, to investigate the spatiotemporal scale of propagule dispersal of Rhizophora stylosa, one of the widely distributed mangrove species in the Indo-West Pacific, utilizing population genetic methods and a release-recapture method employing GPS drifting buoys. The contribution of propagule dispersal to connectivity is still largely unknown, especially at the local scale. Evaluating fine-scale propagule dispersal patterns unique to individual island systems is important to understand their contribution to global species distributions, and to select appropriate sizes and locations for mangrove conservation in archipelagos. This study sought to quantify intra- and inter-island connectivity and to assess their contributions to oceanic scale dispersal of R. stylosa in the southern part of the Ryukyu Archipelago, which spans over 545 km in southwestern Japan. From 16 fringing populations on four islands, we identified three genetic populations, indicating distinct genetic structures comprising three distinguishable bioregions (genetic clusters). We found that the spatiotemporal scale of propagule dispersal is limited by the distance between islands (<200km), propagule viability duration, and fecundity. Overall, we found that demographic connectivity does not occur frequently enough to unify the genetic structure in the archipelago, and the Ryukyu Archipelago is isolated from the global mangrove distribution. In summary, mangrove habitats are more self-sustained and are further isolated among islands compared to other coastal species. Previously suggested roles of island mangrove habitats as a steppingstone appear to be site specific. Integrating different methods, numerical modeling, genetics, and physical oceanography, allowed evaluation in more depth. We concluded that the frequency of mangrove genetic exchange among populations is high enough to avoid harmful effects from the lack of connectivity although the spatiotemporal characteristic of dispersal is site-specific while it was found rare, stochastic, and localized in the island system.tant sequence features for signaling are also conserved between cnidarian and bilaterian neurexins. In the cnidarian model organism, Nematostella vectensis, delta form neurexins are expressed in neuron cell clusters that exhibit both peptidergic and neurotransmitter (glutamate/glycine) signaling ability. Knockdown of NvNrxnδ1 and NvNrxnδ2 genes does not affect peptide development, but normal behaviors are impaired. This experimental evidence suggests the likely functional importance of a presynaptic adhesion molecule in maintaining neuronal communication within a simple nervous system.

San To Chan

Supervisor Name: Amy Shen
Research Unit: Micro/Bio/Nanofluidics Unit
Thesis: Estimating Protein-Protein Interactions with High Aspect Ratio Plasmonic Nanopillars

Abstract:
Industries such as electronic packaging, 3D printing and food engineering require high speed and high precision dispensing of rheologically complex fluids. In other words, liquid bridges of complex fluids have to be broken quickly and cleanly. However, for complex fluids like epoxy, short liquid bridges have long breakup times. Hence, extension of liquid bridges is often required to speed up the dispensing process. Yet, this reduces the precision because the secondary droplets and a long capillary tail formed after breakup may fall randomly and contaminate the substrate. The industries that require dispensing of rheological complex fluids have hence been working hard to find the right solution to this speed-precision-tradeoff problem. This thesis aims to show that torsion can be an effective method to break up rheologically complex fluids quickly and cleanly, as opposed to extension. To this end, a simple experimental system which consists of two parallel plates is constructed, one of which can rotate unidirectionally, another one of which can move vertically. Fluids with different levels of rheological complexity are injected in between the two plates to test how torsion affects the thinning behavior of the liquid bridge. A critical finding is that as long as elasticity is present in the fluid, torsion can induce normal stress in the liquid bridge. As the level of torsion is high enough, the torsion-induced normal stress effectively creates an indent on the free surface of the liquid bridge, which propagates quickly towards the center and hence eliminates the need of any extension.

Ainash Garifullina

Supervisor Name: Amy Shen
Research Unit: Micro/Bio/Nanofluidics Unit
Thesis: Estimating Protein-Protein Interactions with High Aspect Ratio Plasmonic Nanopillars

Abstract:
Protein-protein interactions (PPI) are crucial for biochemical processes within or among cells, which make characterizing these interactions essential for understanding the fundamentals of every living organism. PPI occur between two or more protein molecules that come into physical contact with each other, and are caused by combinations of electrostatic forces, van der Waals interactions, hydrogen bonding, and hydrophobic effects determined by the geometry of these molecules. Due to various interactions taking place even within a single protein molecule, quantifying magnitude of PPI in terms of net attraction forces or affinity between different biomolecules remains a challenge, and protein characterization techniques aimed at measuring binding affinities are often time-consuming, require large volumes of analytes, or lack statistical significance. In this thesis, we developed a feasible high-throughput technique for estimating PPI in real-time based on the measurement of the equilibrium dissociation constant (KD) using a microfluidic platform containing high aspect ratio (≥10) plasmonic nanopillars. Plasmonic nanostructures have been proven to be extremely sensitive to the refractive index changes in the surroundings and thus allow highly accurate measurements of concentrations of protein molecules before and after equilibrium. First, we built a novel biosensing platform by shaping polystyrene polymer into nanopillars by using anodized aluminum oxide membrane as template. We then compared most common surface chemistry methods for biomolecule immobilization on non-spherical plasmonic nanostructures and identified that 11-mercaptoundecanoic acid linkers lead to the most reliable and reproducible biosensing results. Based on the detected wavelength shifts, we quantified fluctuations in biomolecule concentrations and calculated values of KD on the order of 10^-6, and successfully estimated magnitudes of PPI within C-reactive protein, SARS-CoV-2 spike protein 1, Biotin-Streptavidin protein systems with the LoD on the order of 10 pM range. Our opto-microfluidic platform with high aspect ratio plasmonic nanopillars can be applied to various biomolecular systems, laying foundations to high-throughput real-time detection and estimation of PPI based on protein binding affinity.

Mizuki Kato

Supervisor Name: Erik De Schutter
Research Unit: Computational Neuroscience Unit
Thesis: A Computational Model of Granule Cell Migration and Purkinje Cell Primary Dendrite Selection during Cerebellar Development

Abstract:
The aim of this project is to investigate the interrelationship between primary dendrite selection of Purkinje cells and migration of their pre-synaptic partner granule cells during cerebellar development. The developmental course of the cerebellum is conserved between humans and rodents. Deficits happening during this development can affect lifelong cognitive functions and cause severe mental disorders in both species. During development of the cerebellar cortex, its sole output neurons, Purkinje cells, grow more than three dendritic trees for each in an early stage, among which a primary tree is selected in later stages to develop further, whereas others completely retract. Earlier experimental studies on rodent cerebella have provided invaluable visualization of different neurons and suggested that the selection of Purkinje cell dendrites is coordinated by physical and synaptic interactions with granule cells under massive migration. However, technical limitations hinder a continuous observation on multiple populations of the cells during the development, rendering mechanism underlying the Purkinje cell dendrite selection process yet unclear. To answer this question, we constructed a computational model of Purkinje cell dendritic development and granule cell migration, using a new computational framework, NeuroDevSim, which allows consideration of both biological rules and cell-environment interactions during model generation. The constructed model in this thesis enables simultaneous simulation of dynamics during Purkinje cell dendritic development and interactions of the dendrites with migrating granule cells. The structure of the model comprises 48 Purkinje cells, 3,024 granule cells with extensions of their parallel fibers, and 119 Bergmann glia as a guide for granule cell migration in a (x, y, z) = (160, 140, 120)µm3 cube, calculated by assuming a cube of developing mice cerebellar cortex from postnatal days 0 to 10. Using this model cube, alterations of three factors to determine dendritic retractions are tested; fixed timing, dendritic maturation, or network activity. Then, impact of change in these factors on resulting dendritic morphology is analyzed. This thesis presents the first computational model that simulates both populations of Purkinje cells and the dynamics of surrounding granule cells, revealing the role of physical and synaptic interactions upon the dendritic development of Purkinje cells. This model can bridge the gap in unexplored developmental course of the early neonatal Purkinje cell dendrites from the aspect of cellular interactions, and expecting to provide new insight to track down how the cerebellar cortex develops into normal or abnormal structure.

Christine Joy Guzman

Supervisor Name: Hiroshi Watanabe
Research Unit: Evolutionary Neurobiology Unit
Thesis: Investigating the Role of Neurexins in the Early Evolution of the Nervous System

Abstract:
The appearance of the nervous system was an evolutionary event that allowed animals to control their physiology and behavior. In the nervous system, chemical signaling occurs either via volume transmission mainly by neuropeptides, or synaptic (wired) transmission by ‘classical’ chemical neurotransmitters. It is hypothesized that during the early evolution of animals, the diffusion-driven peptidergic signaling was the major mode of cell-cell communication, even for neuron-less animals. Later on, a faster and more targeted synaptic signaling was needed, as animal bodies and behaviors increased in complexity. But it remains largely unclear how the synaptic machinery evolved. One crucial event was the establishment of cell-cell contacts to specify and stabilize the communication between functionally distinct heterologous cell types like sensory neurons and muscles. In this study, I investigated Neurexins (Nrxns), a family of core presynaptic cell adhesion molecules with critical roles in bilaterian chemical synapse, using the non-bilaterian model Nematostella vectensis, a member of Cnidaria which is the closest outgroup to Bilateria. A series of gene expression and structure analyses indicated a non-neural origin of Nrxns. Functional analysis of the epithelial (classical) Nrxn in N. vectensis revealed its major role in cell adhesion, in particular in the maintenance of junction between ectodermal and endodermal epithelia. Neural Nrxns, named delta-Nrxns, are distinctly expressed in neuronal cell clusters that exhibit both peptidergic and classical neurotransmitter signaling abilities. Knockdown of NvNrxnδ1 and NvNrxnδ2 resulted in abnormal behaviors of N. vectensis polyps, involving muscle contraction. Interestingly, the knockdown of the neuropeptide- precursor gene co-expressed with delta-Nrxns did not show the same behavioral abnormalities, reflecting an independent role of delta-Nrxns from peptidergic signaling. Pharmacology experiments suggested that the delta-Nrxns are required for chemical neurotransmission (i.e., synaptic signaling). This study provides molecular, functional, and cellular insights into the ancestral non-neural function of Nrxns and may explain how and why this cell adhesion molecule family was employed in the synaptic machinery of the ancestral nervous system.

Sofiia Kosar

Supervisor Name: Keshav Dani
Research Unit: Femtosecond Spectroscopy Unit
Thesis: Characterization of Nanoscale Defects in Hybrid Perovskite Thin Films for Photovoltaic Applications

Abstract:
Hybrid halide perovskites have emerged as one of the most promising contenders for next generation, low-cost photovoltaic technologies. Thanks to the remarkable optoelectronic properties of hybrid perovskite absorbers, perovskite solar cells now achieve efficiencies comparable to conventional inorganic solar cells (Si, GaAs), despite being actively researched for only about a decade. The ability to be processed from solution and to be deposited on transparent and flexible substrates, makes them very attractive for various photovoltaic applications. However, before their wide commercialization, hybrid perovskites need to overcome important limitations. In particular, the presence of defects in hybrid perovskite thin films has been detrimental to material properties, and has been a critical reason preventing devices from reaching their full potential. To successfully deploy hybrid perovskites, we must understand the nature of the different types of defects, assess their potentially varied roles in device performance, and understand how they respond to passivation strategies. In this thesis, we have employed photoemission electron microscopy to directly image nanoscale defect clusters, and have uncovered the presence of multiple types of defects in state-of-the-art perovskite thin films. By adding time resolution to our photoemission electron microscopy measurements, we found that depending on their nature, these defects played surprisingly varied roles in charge carrier trapping – from highly detrimental to relatively benign. Further, we also found them to show varied response to passivation strategies, as seen from our photoemission measurements. With this work, by identifying the origins of various faults occurring in perovskite thin films and highlighting importance of designing meaningful and targeted strategies to overcome them, as well as demonstrating sophisticated yet greatly rewarding tools to detect these very nanoscale fault-rich sites, we hope to contribute to development of more viable and durable perovskite photovoltaics.

Aliya Mari Purser

Supervisor Name: Izumi Fukunaga

Research Unit: Sensory and Behavioural Neuroscience Unit

Thesis: Designing a Tractable Behavioral Paradigm for Investigating Olfactory Figure-ground Segregation

Abstract:

Odors naturally exist as mixtures in the environment. Detecting relevant cues amidst other signals and noise, a task called figure-ground segregation, is important for survival. Understanding mechanisms that enable animals to solve such a challenging task requires a paradigm that recapitulates key features of the task, yet ideally should be simple enough that allows mechanistic bases to be studied experimentally. In my PhD, I developed a behavioral paradigm using binary mixtures as a model for olfactory figure-ground segregation in mice. Ethyl butyrate (EB) was assigned as target odor and ten other background odors with differing degrees of chemical similarity to EB were included as part of a Go/No-Go task. The fact that the mixtures comprised only two odors made the number of possible odor combinations limited, which therefore made the paradigm tractable and ensured that all combinations can be presented exhaustively per behavioral session. Despite its simplicity, I demonstrate that the experimental paradigm can still impose a degree of challenge for mice through the use of a highly similar background odor. This captures recent findings that the degree of overlap between odor-evoked neural representations underlies figure-ground segregation difficulty. Additionally, it was determined that mice performing the binary mixture task can easily generalize when presented with a novel odor, which suggests that demixing is likely involved. Finally, two example cases are presented as examples of how the behavioral paradigm can be applied to investigate possible underlying neural mechanisms. Overall, despite its simplicity, this behavioral paradigm is a promising approach for investigating the olfactory figure-ground segregation.

Dongqi Han

Supervisor Name: Jun Tani

Research Unit: Cognitive Neurorobotics Research Unit

Thesis: Self-Organization of Action Hierarchy and Inferring Latent States in Deep Reinforcement Learning with Stochastic Recurrent Neural Networks

Abstract:

It has been a central question in cognitive science and artificial intelligence that what are the underlying mechanisms of learning to make highly adaptive and cognitive decisions in various challenging tasks. The current thesis proposes computational models of decision-making agents, most of which are simulated robots with sensors and motors, and discusses how the proposed models contribute to efficient learning and adaptive decision-making. Rather than the conventional usage of recurrent neural network (RNN) as function approximators in RL, the proposed results provide novel insights of how RNN models can be incorporated with RL to achieve more intelligent decision making.
Firstly, a novel, multiple-level, stochastic RNN model is proposed for solving hierarchical control tasks by model-free reinforcement learning (RL). It is shown that an action hierarchy, characterized by consistent representation for abstracted sub-goals in the higher level, self-develops during the learning in a challenging continuous control task. The emerged action hierarchy is also observed to enable faster relearning when the sub-goals are re-composed.
Then, the author introduces a variational RNN model for predicting state transitions in control tasks in which information of environment state is partially observable. By predicting new observations, the models learns to represent the underlying states of the environment that are important but probably not observable. A corresponding algorithm is proposed to facilitate efficient learning in partially observable environments.
Finally, the variational RNN model is extended to a predictive coding model that cooperate RL and active inference in the same network. It is investigated how RL is used for exploring the environment and avoiding punishment and goal-directed planning is then conducted in the framework of active inference. It is shown that a trained agent is able to select near-optimal actions to achieve a given goal by providing goal observation using backpropagation of prediction error in the variational RNN. The study offers novel insights about how RL and active inference can collaborate in a complementary manner for different purposes.

Takeshi Ricardo Tabuchi Yagui

Supervisor Name: Yohei Yokobayashi

Research Unit: Nucleic Acid Chemistry and Engineering Unit

Thesis: High-Throughput Screening of Cell-Free Riboswitches for Chemical Communication between Microdroplets

Abstract:

Riboswitches have recently attracted the attention of synthetic biologists as an alternative to transcription factors for genetic regulation, due to their engineerability, relative simplicity, and potential for responding to a wide array of chemical signals. However, biological constraints such as cell permeability, metabolic stability, and toxicity of the signaling molecules have prevented the development of some of those devices using conventional approaches with living cells. Cell-free systems are generally not subjected to such constraints and offer a unique platform for building biochemical and genetic systems that display complex functions without using living cells. Efforts to engineer regulatory components directly in cell-free systems thus far have been based on low-throughput experimental approaches, limiting the availability of basic components for building genetically programmed cell-free systems. Here, I report a high-throughput screening method for engineering cell-free riboswitches that respond to small molecules. Fluorescence-activated droplet-sorting (FADS) of riboswitch variants in a cell-free protein synthesis (CFPS) system rapidly identified cell-free riboswitches that respond to histamine and ciprofloxacin, compounds that are normally not amenable with conventional bacterial screening methods. Finally, histamine riboswitches obtained through this method were used to demonstrate chemical communication between microdroplets.

Saahil Acharya

Supervisor Name: Akihiro Kusumi

Research Unit: Membrane Cooperativity Unit

Thesis: SynGAP Condensates Recruit PSD95, and Selectively Retain Multivalent Receptors, Functioning as the Basic Platform for Generating Neuronal Excitatory Synapses

Abstract:

The synapse structures in both the post- and pre-synaptic plasma membranes consist of multiple nano-scale domains, precisely apposed to each other. Synaptic receptors in the post-synapse need to be assembled in the nano-scale domains, precisely opposite from the presynaptic neurotransmitter release sites, for proper synaptic function. We found that SynGAP forms phase-separated condensates through homophilic interactions mediated by its C-terminal coiled-coil domain as well as its intrinsically disordered region. SynGAP recruits PSD95 into these condensates, and this allows recruitment and immobilization of receptors such as Neuroligin and AMPAR (via TARPs), which have PDZ binding sites.
We also found that oligomerization of Neuroligin and AMPA receptors is essential for anchoring the receptors in the synapse. Functional dimeric Neuroligin and tetrameric TARP2 (linked to GluA1 subunits) can be captured as they diffuse through phase-separated condensates containing PSD95, while monomeric Neuroligin and TARP2 cannot. We revealed that the receptor oligomers enhance condensation, which in turn further recruit receptor oligomers.

Tim Keller

Supervisor Name: Thomas Busch

Research Unit: Quantum Systems Unit

Thesis:Controlling Superfluid and Insulating States in Interacting Quantum Gases

Abstract:

In this thesis I present two studies on controlling the state and properties of both single species and composite quantum gases by tuning the various interaction strengths. In the first work I derive a shortcut to adiabaticity (STA) for tuning a Feshbach resonance in repulsively interacting Bose-Einstein condensates (BECs) in the Thomas-Fermi regime. This shortcut mimics an adiabatic evolution and allows one to compress and expand a BEC without friction within an almost arbitrarily short time interval. I then use this technique to show how it can boost the performance of the so-called Feshbach quantum engine and also determine its limits and the instabilities it can lead to.
In the second part I show that a strongly correlated one-dimensional quantum gas in the Tonks-Girardeau (TG) limit that is immersed into a BEC can undergo a transition to a crystal-like insulator state without any externally imposed lattice potential. I develop a model that accurately describes the system in the pinned insulator state, both at zero temperature and for the situation where the TG gas is at a finite temperature. Additionally, I study the superfluid state that can persist in the gas for finite interactions away from the TG limit and uncover the full phase diagram of the system.

Dmitrii Koldaev

Supervisor Name: Eliot Fried

Research Unit: Mechanics and Materials Unit

Thesis:Mathematical Modeling and Numerical Analysis of Unstretchable Elastic Ribbons

Abstract:

Paper bends easily but cannot extend or contract much without tearing or creasing, and the same is true of more exotic materials like the ultra-thin bendable glass products sold by Corning and Nippon Electric Glass. When modeling such a material as a two-dimensional elastic body, its resistance to elongation and contraction can be incorporated considering only deformations under which the distances between material points are preserved. This constraint is accompanied by analytical and numerical challenges and is the source of many unanswered questions. The primary objective of this thesis is to develop robust and efficient numerical methods for finding stable equilibria of an unstretchable two-dimensional elastic material is bent so that its short edges are joined, with or without twist, to form a closed band. To find an equilibrium configuration of such a band, we minimize the bending energy. When the reference and deformed surfaces are parametrized as ruled surfaces, a dimension reduction is possible. This reduction converts the problem to one involving a system of ordinary differential equations for a pair of vector fields satisfying certain constraints that derive from the requirement that the material be unstretchable and periodicity or antiperiodicity conditions that incorporate the way in which the short edges of the reference strip are joined. We discretize this problem to obtain a multi-dimensional constrained optimization problem that is solved numerically. To incorporate the constraints on the discrete counterparts of the unknown vector fields, we introduce Lagrange multipliers and minimize an accordingly augmented version of the bending energy. One difficulty encountered in solving the optimization problem is to avoid saddle points that do not correspond to stable equilibria. To tackle this challenge, we developed a saddle-free Newton method for constrained optimization by generalizing ideas from the recent literature for a related class of unconstrained optimization problems. Additionally, we introduce an alternative constraint to ensure that the deformation is injective. The new constraint is bilateral and not only obviates the need to impose inequality constraints but also mollifies a certain singular feature of the energy density and circumvents symmetry assumptions that have been in imposed in all previous studies of this kind.

Alexandru Mihai

Supervisor Name: Eliot Fried

Research Unit:Mechanics and Materials Unit

Thesis: Soap Film Mediated 3D Self-Assembly: Suspended and Displacement Driven Geometries Using Centimeter-Scale Tiles

Abstract:

The self-assembly of polygonal pieces within a soap film is demonstrated. Gravity mediated as well as displacement driven assemblies of polygonal frames within a soap film are shown, namely five (platonic solid) geometries and 24 free-standing structures respectively. A model of axes-symmetric geometries is presented and compared to experimental results of 3D printed rings solely supported by a soap film membrane counteracting the gravitational force acting on the rings. The model is derived by both energy methods (using a calculus of variations approach) and a direct force balance approach. The explicit solution of the system extends the Goldschmidt limit of catenoid shaped minimal surfaces to include load bearing geometries corresponding to the maximal mass supported by a soap film spanning two rings of given radius. This model provides an approximation of the self-assembled polygonal frames suspended in the soap film.
Computational results using the FEniCS Project package confirm the analytical results for axes-symmetric and one-dimensional geometries. Eight prism and 16 pyramidal structures of varying base geometry and heights are constructed through a displacement driven approach on a base plate by the smooth deformation of a catenoid-like soap film until the film is no longer stable, resulting in a pinch-off phenomenon. Coupled with the edge-to-edge alignment of polygonal frames the pinch-off of the membrane results in free standing prism, pyramid, and octahedral geometries. Two effective radii corresponding to prism and pyramid structures respectively are derived, and found to predict the separation between the base and support needed for the pinch-off of the soap film to occur.

Shan Zou

Supervisor Name: Denis Konstantinov

Research Unit:Quantum Dynamics Unit

Thesis: Detection of the Rydberg States of Electrons on Superfluid Helium Confined in Microchannel Devices

Abstract:

The orbital and spin states of electrons on liquid helium can potentially serve as a viable resource for quantum information processing. In particular, an introduced spin-orbit interaction between spin of an electron and the Rydberg states of its vertical motion can open a new pathway towards building a scalable spin-qubit quantum computer.
Previously, the Rydberg state detection using mm-wavelength radiation was demonstrated for the bulk electron systems accumulated on a free surface of liquid helium. However, implementation of such electrons for qubits requires manipulation of individual electrons and their quantum states, thus implementation of some kind of microscopic devices. This thesis describes the experimental work with electrons on superfluid helium confined in fabricated microchannel devices, in particular the first observation of the Rydberg states of such electrons using both transport measurement and image-charge detection method.
This work should be considered as the first step towards potential detection and implementation of the spin-states of electrons on helium for qubits.

Agneesh Barua

Supervisor Name: Vincent Laudet

Research Unit: Marine Eco-Evo-Devo Unit

Thesis: The Evolutionary Genetics of Venoms: How Nature Created the Perfect Chemical Weapon

Abstract:

Venomous animals have fascinated humans for millennia. How nature shaped a simple biological secretion into a potent chemical weapon is a testament to evolution’s power and versatility. However, the early origins and genetic mechanisms of venom evolution are not clearly understood. Venoms consist of proteinaceous cocktails where each protein can be mapped to a specific gene; I utilized this genetic tractability to uncover the molecular and genetic mechanisms behind its evolution. Using a combination of quantitative genetics, transcriptomics, and phylogenetics, I have identified specific mechanisms that led to the origin of oral venoms in mammals and reptiles. Oral venoms originated from an ancient conserved gene regulatory network whose primary role was maintaining cellular homeostasis during increased protein production. This ancient system could tolerate high protein loads, facilitating the parallel recruitment of various diverse protein families into the ancient venom. Animals then increased complexity by sequence and copy number variation of toxins. High copy numbers contributed to this system’s phenotypic flexibility, which could further diversify through changes in evolutionary rates and by altering the combinations of toxins used. These features enabled evolution to refine venom cocktails to form optimal formulations. I provide the first unified and deep evolutionary model describing the early steps in forming a venom system and show how millions of years of evolution produced venom phenotypes in extant lineages. All chapters of this thesis have been peer-reviewed and published.

Makoto Tokoro Schreiber

Supervisor Name: Matthias Wolf

Research Unit: Molecular Cryo-Electron Microscopy Unit

Thesis: Diffraction-Based Experiments in Transmission Electron Microscopy: Lensing, Charging, and Amorphous Structures

Abstract:

In this thesis, three topics related to the electron beam interactions in various fields of electron microscopy are investigated.
In the first topic, one of the principal components of the microscope is investigated; the lens. The working principle of currently used electromagnetic lenses are discussed from the viewpoint of electromagnetic fields and electromagnetic potentials. A new physical basis for electromagnetic lensing is developed using the Aharonov-Bohm effect. The properties of this new type of lens are detailed and compared to existing lenses. Its applicability towards next-generation electron microscopes is explored. The lensing theory is tested using electron holography and Fresnel diffraction experiments.
The second topic investigates the charging behavior of vitrified ice films upon exposure to an electron beam through Fresnel diffraction experiments. This is important to understand the degradation of information contribution from the initial lowest electron-dose frames in a single-particle cryo electron microscopy dataset. In this work, the evolution of the charge distribution on the film is modeled spatially and temporally. The impacts of this on the resultant images and steps that can be taken to minimize information loss through computational and experimental means are discussed.
The third topic explores the possibility of applying the single-particle analysis methodology developed for cryo electron microscopy to solving short-range order structures in amorphous samples using nanobeam electron diffraction patterns. The direct applicability of currently existing algorithms and modifications required to suit the diffraction dataset are investigated.
The primary focus of the thesis will be the phase modification of the electron beam by electromagnetic potentials.

Ratnesh Kumar Gupta

Supervisor Name: Sile Nic Chormaic

Research Unit:Light-Matter Interactions for Quantum Technologies Unit

Thesis:Evanescent Field Mediated Interactions of Cold Rubidium Atoms with Optical Nanofiber Guided Light

Abstract:

The integration of optical nanophotonic structures and quantum emitters has been a major goal in atomic physics for its tremendous potential in quantum technology developments. Dielectric waveguide structures, specifically optical nanofibers (ONFs), have come up as a promising platform for such integration and have already been demonstrated in numerous applications when interfaced with quantum systems such as cold atomic ensembles and quantum dots.
ONFs enable tight transverse confinement of guided light over an extended length, typically many times longer than the Rayleigh length, offering strong atom-light interactions with potential long-range atom-atom interactions mediated by the guided light. This allows for better scalability in many quantum information applications than its corresponding free-space implementation including QED cavities.
We seek to extend the applications of ONFs in atomic systems. We present an observation of 5S1/2 → 4D3/2 electric quadrupole transition at 516.6 nm using only few μW of laser power propagating through ONF. Quadrupole transitions play an important role in atomic and molecular spectroscopy and can be useful for angular momentum transfer from light to atoms and high precision measurements of parity non-conservation. The steep gradient of the evanescent field of the light propagating through ONF is exploited to drive relatively inaccessible electric quadrupole transitions in laser cooled Rb87 atoms.
We study the polarization dependence single-frequency two-photon transition at 993 nm addressing 5S1/2 → 6S1/2 in 87Rb atoms. We developed a theoretical framework to study the two-photon transition rate as a function of helicity of excitation light. While the two-photon transition can be completely extinguished by varying the polarization of the excitation light in the case of paraxial Gaussian excitation, the same is not true for ONF-mediated excitation. This is an important result drawing differences between the excitations in the two cases.
Cold atoms have been trapped in the evanescent field of the ONF and have led to demonstrations of all-fiber optical memories and the observation of collective effects such as super-radiance. A major challenge is to maximize the number of atoms coupled to the common radiation field while preserving the inter-atomic phase correlation. It remains a challenging problem to capture a complete quantitative description of atomic dynamics during the loading of such fiber-based traps involving many-body interactions, intensity and field gradients and potentially complex scattering processes. We investigate optimization of such a fiber-based trap using an online optimization based on deep-learning leveraging the ability of machine learning to learn from experience by simply interacting with the process and optimizing a generally complex system.

Mai Omar Abdulrahman Ahmed

Supervisor Name: Ichiro Masai

Research Unit:Developmental Neurobiology Unit

Thesis:Investigating the Function of Strip1 in Ganglion Cell Survival and Neural Circuit Formation of the Developing Zebrafish Retina

Abstract:

In the vertebrate retina, an interplay between retinal ganglion cells (RGCs), amacrine and bipolar cells establishes a synaptic layer called the inner plexiform layer (IPL). This circuit conveys signals from photoreceptors to visual centers in the brain. However, the molecular mechanisms involved in its development remain poorly understood. Striatin-interacting protein 1 (Strip1) is a core component of the STRIPAK complex, and it has shown emerging roles in embryonic morphogenesis. Here, we uncover the importance of Strip1 in inner retina development. Using zebrafish, we show that loss of Strip1 causes defects in IPL formation. In strip1 mutants, RGCs undergo dramatic cell death shortly after birth. Amacrine and bipolar cells subsequently invade the degenerating RGC layer, leading to a disorganized IPL. Mechanistically, zebrafish Strip1 interacts with its STRIPAK partner, Striatin3, to promote RGC survival by suppressing Jun-mediated apoptosis. In addition to its function in RGC maintenance, Strip1 is required for RGC dendritic patterning, which likely promotes interaction between RGCs and ACs for IPL formation. Taken together, we propose that a series of Strip1-mediated regulatory events coordinates inner retinal circuit formation by maintaining RGCs during development, which ensures proper positioning and neurite patterning of inner retinal neurons.

Swathy Babu

Supervisor Name: Ichiro Masai

Research Unit: Developmental Neurobiology Unit

Thesis: Banp Regulates DNA Damage Response and Chromosome Segregation to Promote Cell-cycle Progression and Cell Survival in Zebrafish Retina

Abstract:

Btg3 associated nuclear protein (Banp) was initially identified as a nuclear matrix associated protein and is a tumor suppressor. Recently it was reported that Banp binds to the CGCG sequence enriched near the transcription initiation site of CpG island promoters, namely Banp motif, promotes the transcription in a DNA methylation dependent manner, and controls metabolic genes in pluripotent stem and differentiated neuronal cells. However, cellular roles of Banp in embryonic development remains to be elucidated. Here we report a novel role of Banp in cell-cycle progression and cell survival of zebrafish retinal progenitor cells. In zebrafish banp mutants, retinal progenitor cells showed mitotic arrest and subsequent apoptosis. DNA replication stress and tp53-dependent DNA damage response were activated in banp mutants. Inhibition of tp53 significantly rescued apoptosis but not mitotic arrest and DNA double strand break accumulation, suggesting that Banp is required for functional integrity of DNA replication and DNA damage repair. Furthermore, live imaging of mitosis in banp mutant retinas revealed that chromosome segregation was not smoothly processed from prometaphase to anaphase, leading to prolonged M phase. Bulk RNAseq analysis show that mRNA expression of two chromosomal segregation regulators, cenpt and ncapg, were decreased in banp mutants. Furthermore, ATACseq analysis showed that chromatin near their transcription start site was closed in banp mutants and indeed Banp motif was found in this chromatin-closed region, suggesting that Banp directly regulates cenpt and ncapg transcription to promote chromosome segregation during mitosis. Our findings reveal that Banp is required for cell-cycle progression and cell survival by regulating DNA damage response and chromosome segregation during mitosis.

Jigyasa Arora

Supervisor Name:Tom Bourguignon

Research Unit: Evolutionary Genomics Unit

Thesis:Functional Metagenomics and Evolution of Termite Gut Microbiome

Abstract:

Termites are amongst the most abundant terrestrial animals on earth, largely due to their ability to digest lignocellulose, the most abundant organic molecule on earth. Lignocellulose is broken down in the termite gut with the help of symbiotic bacteria. Studies using the 16S rRNA marker have shown that termites and their gut bacteria have had a complex coevolutionary history. Although many gut microbes are found nowhere else than in termite guts, bacterial communities do vary with termite diet. Up to now, studies have been focusing on termite species easy to sample or having a pest status. This sampling bias against early-evolving termite lineages and lineages feeding on substrates different than wood preclude a global understanding of the evolutionary history of termites and their gut microbes. To fill this gap, I sequenced whole gut metagenomes of 201 termite samples and Cryptocercus kyebangensis, a species of the genus sister to termites, sampled across the termite tree of life to represent termite phylogenetic and dietary diversity. Applying a combination of phylogenetic comparative methods and phylogenetic reconstructions, my thesis showcases that (i) the gut microbiome of all termites possess similar genes for carbohydrate breakdown and other metabolic pathways involved in the digestion of cellulose. The proportion of these genes vary with termite phylogeny and diet but the acquisition of a diet of soil from a wood-feeding ancestor was accompanied by changes in gene abundance rather than by the acquisition of new genes and pathways. Using ten single-copy protein-coding marker gene sequences, (ii) I studied the pattern of coevolution between termites and each of their gut bacterial phyla. I found significant cophylogenetic signals between termites and a dozen of gut bacterial lineages that were acquired by the common ancestor of all termites or by specific termite lineages. Finally, (iii) I investigated horizontal gene transfers among termite gut bacteria for genes involved in lignocellulose digestion. I found numerous horizontal transfers among microbial phyla. Overall, my Ph.D. thesis sheds new light on how gut microbiome has coevolved with their termite hosts since the inception of this nutritional symbiosis, some 150 million years.

Mohieldin Magdy Mahmoud Youssef

Supervisor Name:Tadashi Yamamoto

Research Unit: Cell Signal Unit

Thesis: Role of TOB in the Brain: An Insight into Stress Coping Machinery

Abstract:

Stress affects behavior and involves critical dynamic changes at multiple levels ranging from molecular pathways to neural circuits and behavior. Abnormalities at any of these levels leads to decreased stress resilience and pathological behavior. However, temporal modulation of molecular pathways remains poorly understood. Transducer of ErbB2.1, known as TOB, (TOB1) is involved in different physiological functions, including cellular stress and immediate response to stimulation. In this study, we investigated the role of TOB in the brain’s stress machinery at molecular, neural circuit, and behavioral levels. Interestingly, TOB protein levels increased after mice were exposed to acute stress paradigms. At the neural circuit level, functional magnetic resonance imaging (fMRI) suggested that intra-hippocampal and hippocampal-prefrontal connectivity was dysregulated in Tob knockout (Tob-KO) mice. Electrophysiological recordings in hippocampal slices showed increased postsynaptic AMPAR-mediated neurotransmission, accompanied by decreased GABA neurotransmission and subsequent altered Excitatory/Inhibitory balance after Tob deletion. At the behavioral level, Tob-KO mice show abnormal, hippocampus-dependent, contextual fear conditioning and extinction, and depression-like behaviors. On the other hand, increased anxiety observed in Tob-KO mice is hippocampus-independent. At the molecular level, we observed decreased stress-induced LCN2 expression and ERK phosphorylation, as well as increased MKP-1 expression. This study proposes that TOB serves as an important modulator in the hippocampal stress signaling machinery. Herein, we show a molecular pathway and neural circuit mechanism by which TOB contributes to expression of pathological stress-related behavior.

Larisa Sheloukhova

Supervisor Name: Hiroshi Watanabe

Research Unit: Evolutionary Neurobiology Unit

Thesis: Molecular Dissection of Ancestral Glia

Abstract:

Nervous systems of bilaterian animals generally consist of two cell types: neurons and glial cells. Glia participate in almost every process taking place in the nervous system of bilaterians, indicating a crucial role(s) of glial cells both in neurophysiological functions and for nervous system evolution. Therefore, tracing back the first glia and elucidating its ancestral function is important for understanding the evolution of the nervous system. Histological examinations have not so far revealed any morphological sign of glial cells in Cnidaria, the closest outgroup to Bilateria. This led to the hypothesis that glial cells appeared after the common bilaterian ancestor had branched off from Cnidaria. However, this view has not been well examined at the genetic level. In this work I sought to investigate gliogenic program conservation in non-bilaterian animals, i.e. basal metazoans. First, I analyzed bilaterian glial gene orthologs in basal metazoans using available genomes and transcriptomes. Second, I focused on Anthozoa (Cnidaria) as a class with the most conserved genetic program required for glial development among basal metazoans. Third, using Nematostella vectensis as an anthozoan model organism I explored the function of Gcm, the core gliogenic transcription factor (TF), by knocking it down and performing RNA-seq and RT-qPCR analyses. Among Nematostella Gcm targets are cell adhesion proteins, GABA and glutamate transporters, ion channels, metabolic and protein modifying enzymes, as well as zinc finger and Ets-related TFs. In addition, Gcm seems to control Notch-Delta signalling, which is one of the crucial neuro-gliogenic pathways in bilaterians. I also performed immunostaining of the Gcm-target proteins to get insights into the cell morphology. The major finding of my thesis is that Nematostella Gcm-expressing cells demonstrate characteristics of both neurons and glia, suggesting a dual nature of ancestral cells. This indicates the conservation of the gliogenic program intertwined with the neurogenic program which separated later in the animal evolution.

Xunwu Hu

Supervisor Name: Ye Zhang

Research Unit: Bioinspired Soft Matter Unit

Thesis: Developing Integrin-targeted Peptide Assemblies to Direct Cancer Cell migration

Abstract:

Advances in mechanistic understanding of integrin-mediated adhesion highlight the importance of precise control of ligand presentation in directing cell migration. The development of top-down nanofabrication techniques, such as polymer blending and nanolithography, achieved control over the spatial presentation of integrin ligand at the sub-micron resolution, which promoted the mechanistic study of integrin-mediated adhesions and inspired biomaterial innovations. We sought to enhance the spatial resolution beyond sub-micron resolution to understand the subsequent cellular response at the molecular level. To address this challenge, we developed a bottom-up nanofabrication strategy, reaching by far the highest spatial resolution of ligand presentation (48 ligands/100 nm2), which is beyond the submicron limit of the top-down technique. Via simple molecular engineering, we transformed a natural ECM-derived ligand into an assembling ligand. Co-assembly of the assembling ligand with non-functional motifs at various proportions, forms biocompatible nanofilaments presenting different ligand densities. Peptide assemblies possessing various ligand densities exert biphasic effects on cell migration, with fast migration occurring at low density because of promotion of nascent adhesion and lamellipodia formation, and inhibition occurring at high density due to the prevention of integrin and actin filament disassembly at the cell rear. Meanwhile, we illustrated the cellular response to extracellular super high-density ligands. When the cells are exposed to super high-density ligands, the stress-fiber-associated focal adhesions (FAs) slide inward, while the actin cytoskeleton together with the integrin receptors and adaptor proteins were stabilized at the cell rear, restricting the cell retraction and protrusion. By expressing vin 258, a mutant that possesses vinculin D1 domain exhibiting high affinity to talin and paxillin but lack of actin-binding domain, the cells successfully maintained the FA on the periphery but failed to preserve the actomyosin network and could not resume protrusion nor trailing edge retraction. Extra Rho activation preserves FAs at the cell edge and is associated with actomyosin bundles and eased the full disassembly of FAs facilitating trailing edge retraction but failed to resume the cell protrusion. By contrast, the constant activation of Tiam1/Rac1 signaling effectively rescued the cell migration restricted by the excessive binding interactions between integrins and the super high-density ligands. Together, this strategy may provide new insights in material design for manipulating and further understanding in ligand-density-dependent-modulation for manipulating and further understanding in ligand-density-dependent-modulation.

Evropi Toulkeridou

Supervisor Name: Evan P. Economo

Research Unit: Biodiversity and Biocomplexity Unit

Thesis: Automated segmentation of micro-CT images by deep learning and its application to comparative morphology

Abstract:

Semantic segmentation, i.e., association of each pixel of an image with a component label, is one of the most fascinating but challenging problems in the field of computer vision. During semantic segmentation, the image is divided into non-overlapping areas which are subsequently clustered together if found that they belong to the same object. A natural field of potential application is in organismal biology, where researchers are increasingly using three-dimensional (3D) scanning (e.g., confocal scanning or micro-computed tomography, micro-CT) which produces data-rich volumetric images for precise and comprehensive anatomical characterization.
To date, the segmentation of anatomical structures remains a big bottleneck to research, as it is commonly performed with highly tedious and time-consuming manual work. During recent years, however, machine learning (ML) methods are an emerging approach to overcoming this limitation, especially with the use of deep learning techniques such as convolutional neural networks (CNNs), which proved to be very efficient and, as such, promising candidates for image segmentation.

The main objective of this PhD project was to develop a pipeline for the fully-automated segmentation of anatomical structures in micro-CT images of insects using state-of-the-art deep learning methods. The restricted number of high-resolution 3D labeled images available necessitated the use of a CNN architecture that performs segmentation satisfactorily even with limited data; the U-Net architecture is such a convolutional network that has shown good performance in medical images using few annotated images.
Ant brains were selected for a test case, and a dataset was assembled of semi-manually segmented brain images of 94 ant species. It should be noted that both automated classification and segmentation tasks typically require big datasets for training and validation, which can be a challenge for researchers to produce for any given application. Since no dataset of micro-CT images of ant brains existed for the current case study, a new extensive dataset was created across a wide variation of ants. Its existence can be of great importance, as brain images of ants are not dissimilar to those of other insects; therefore, our dataset can act as a starting point for the development of an even bigger library of micro-CT images of insects, and work as a pre-training dataset for future CNNs.
The chosen species set was designed to be interesting for further evolutionary morphology analysis. In the second part of the study, we tested the social brain hypothesis for ants, i.e., the question of whether there is a connection on the brain investment and the sociality of the species. Further volumetric statistical analysis was performed combining the phylogenetic data of the species, leading to the rejection of the hypothesis.

Finally, it should be noted that our network is generalizable for segmenting the whole neural system in full-body scans, and works in tests on distantly related and morphologically divergent insects (e.g., fruit flies). The latter suggests algorithms such as our network can be applied generally across diverse taxa.

Masakazu Taira

Supervisor Name: Kenji Doya

Research Unit: Neural Computation Unit

Thesis: The Role of Serotonin Neurons in Mouse Reward-based Behaviors

Abstract:

Serotonin is an important neuromodulator in reward-driven learning and decision making. Particularly, serotonin neurons in the dorsal raphe nucleus (DRN) send dense projections throughout the brain and its implication in reward-based behaviors has been examined using various types of behavioral tasks. However, how DRN serotonin affects the computational processes of decision making remains unclear. Reinforcement learning is a theoretical framework to describe decision making process. Previous study based on the RL framework proposed hypotheses on the role of serotonin in decision making, such as temporal discounting and model-based value computation. The overall aim of this thesis is to behaviorally examine these hypotheses to understand the role of serotonin in reward-based behaviors.
Particularly I examined two hypotheses. First hypothesis is that in serotonin would control relative importance of future rewards. Previous behavioral studies showed that serotonin activation/suppression enhance/decrease patience to wait for future rewards. However, how serotonin regulate patience to act for future rewards remains unknown. In the first part of my thesis, I trained mice to perform a free-operant lever-pressing task, in which motor action rather than stationary waiting is required, and tested the effect of optogenetic activation and inhibition of DRN serotonin neurons on sustained motor actions for future rewards. I found that optogenetic manipulation of DRN serotonin neurons did affect neither persistent motor actions for future rewards nor response vigor, suggesting different role of DRN serotonin neurons in motor actions for future rewards compared to stationary waiting.
Second hypothesis is that serotonin would modulate model-based decision making. In model-based decision making, agents use their own internal model of action-outcome relationship to plan forward and select actions. Previous computational studies proposed the regulation of model-based decision making by serotonin neurons. However, direct behavioral evidence of how serotonin regulate the process is still limited. The two-step decision making task is an established behavioral task to understand model-based decision making not only in human subject but also in rodents. In the second half of my thesis, I trained mice to perform the two step decision making task and tested how optogenetic inhibition of serotonin neurons affect mice model-based decision making. I found the tendency that optogenetic inhibition of DRN serotonin neurons decreased the weight of model-based decision making in action selection.
These results revealed underlying computational role of DRN serotonin neurons regulating reward-based behaviors.

Yuka Suzuki

Supervisor Name: Evan P. Economo

Research Unit: Biodiversity and Biocomplexity Unit

Thesis: The effects of dispersal network structure on biodiversity pattern and stability in metacommunities

Abstract:

Biodiversity patterns in nature are often heterogeneous across space. Individual organisms can be influenced by spatial constraints via dispersal processes and interactions with the environment, and thus connectivity among patches is a key aspect of spatial structure that can influence the community processes controlling biodiversity patterns. Although the connectivity concept has fostered a large body of theory, many theoretical results are derived assuming simplified spatial structures, which are often not capturing the complexity of natural systems. Network metacommunity models, where patches are connected in potentially heterogeneous ways, can be used to build a theoretical basis for how different landscape and seascape structures may affect metacommunity dynamics. Likewise, recent technical advances permit a better estimation of connectivity in natural systems, such as networks of coral reefs or hydrothermal vents. To fill the knowledge gap about the role of space in ecology, this thesis investigates how the spatial structure drives metacommunity dynamics and biodiversity patterns and stability of metacommunities by utilizing computer simulations and marine connectivity data. In particular, I analyze the two aspects of metacommunity systems: biodiversity assembly and stability, both of which can contribute to biodiversity patterns. First, I examine the role of spatial topology (i.e. the pattern of dispersal linkages among patches) in the interplay between different metacommunity processes, including species sorting and mass effects. I find that network topology strongly mediates the balance of different dynamics in ways that are not captured by simplified spatial models that form the basis of metacommunity theory. Second, I examine the contribution of different aspects of spatial structure to metacommunity stability and find that the size of the network, rather than topology, is the dominant factor driving stability on the metacommunity level. Third, I combine these two simulation analyses with empirical marine connectivity networks and provide a better understanding of spatial processes applicable to natural systems. Overall, these analyses dissect the complexity in connectivity and reveal key aspects of connectivity in regulating biodiversity and stability. This study provides a better understanding of spatial processes and specifically reveals how and which spatial features control biodiversity patterns and metacommunity stability, contributing to a metacommunity theory that can ultimately inform conservation planning.

Menglin Wang

Supervisor Name: Tom Bourguignon

Research Unit: Evolutionary Genomics Unit

Thesis: Worldwide Historical Biogeography of Termites (Blattodea: Isoptera)

Abstract:

Termites are of crucial importance in terrestrial tropical and subtropical ecosystems where they play a key role in organic matter decomposition. Termites descent from a wood-feeding cockroach ancestor and diverged from the subsocial wood roach Cryptocercus more than 150 million years ago. Given that the origin of termites predates the breakup of the Pangaea, termites potentially acquired their current distribution pattern by a combination of vicariance, through continental drift, or oceanic dispersal, through over-water rafting or land bridges. The origin of termite distribution, called historical biogeography, can be studied by mean of reconstruction of ancestral distribution on time-calibrated phylogenetic trees. Although several studies have already reconstructed the global historical biogeography of termite lineages within the Neoisoptera, these studies did not investigate non-Neoisoptera termites and overlooked one biogeographic realm with unique fauna, Madagascar. In this project, I used complete mitochondrial genomes to build time-calibrated phylogenetic trees of termites and determine the precise dispersal events of yet unstudied termite lineages to and fro yet unstudied biogeographic realms. I sequenced the mitochondrial genomes of about 2500 samples including almost all termite species from North, South, and Central America and representatives of the termite diversity from Madagascar. In the chapters 1 and 2, I resolved the historical biogeography of Rhinotermitinae and the early-branching termite families Hodotermitidae, Stolotermitidae, and Archotermopsidae. In the chapter 3, I resolved the historical biogeography of termites from Madagascar. And in the chapter 4, I produced a near-complete phylogenetic tree of termites from the Americas that I used to study the diversification patterns of termites across the two continents that composed the New World. My thesis sheds light on the historical biogeography of termites at the global scale.

Ivan Mbogo

Supervisor Name: Hiroshi Watanabe

Research Unit: Evolutionary Neurobiology Unit

Thesis: The Evolution of Dual Functionality of β-catenin in Metazoans

Abstract:

The evolution of the multicellular body of animals from unicellular organisms is still a significant and long-lasting subject of interest in biology. Acquisition of cell-cell adhesion with cadherin, α- and β-catenin proteins is thought to be tightly coupled with the origin of animal epithelium and consequent evolutionary thrive of animals. Much research has shown, in a wide range of animal lineages such as bilaterians and cnidarians, that β-catenin associates with diverse intracellular proteins involved in gene transcription/translation and plays an essential role in the induction of the signalling centre (organiser) during animal embryogenesis. The pleiotropic and evolutionary conserved functions of β-catenin suggest deep evolutionary roots of the β-catenin complexes and involvement in the emergence of basic animal body plan. Recent progress in genomics has identified genes of the cell-cell adhesion complex and signalling machinery of β-catenin in genomes of early-branching animals, including Porifera (sponges) and Ctenophora (comb jellies). However, due to difficulties in applying molecular and genetic technologies in these non-model animals, the ancestral functions of β-catenin complexes remain largely to be explored.
In this study, I employed structural, proteomic, and functional approaches to understand evolutionarily conserved features of the β-catenin and its associated proteins. Structural analysis suggests a unicellular origin of the basic architecture of β-catenin protein, while amino acid residues critical in adhesive properties are conserved only within animals. To analyze evolutionarily conserved functional characteristics of basal animal β-catenins, I performed transphyletic studies where the basal animal β-catenins are expressed in Xenopus embryos. A series of proteomics analyses of β-catenin-associated proteins provided the first empirical evidence of the deep origin of the cadherin catenin complex. The transphyletic function studies, together with detailed sequence analysis, also revealed the β-catenin’s organiser-inducing function of Cnidaria, Porifera, but not Ctenophora. These data suggest that the primary function of ancestral β-catenin was to play adhesive roles, and its’ signalling properties were equipped later during the evolution of basal animals.

Po-Shun Chuang

Supervisor Name: Satoshi Mitarai

Research Unit: Marine Biophysics Unit

Thesis: From Polyps to Colonies: Applying Polyp Bail-Out to Study Coral Coloniality

Abstract:

Colonial lifestyles have been adopted by the majority of shallow-water stony corals (Cnidaria; Scleractinia), as they facilitate coral responses to environmental changes. Polyp bail-out, a coral stress response featuring colony dissociation and polyp detachment, offers a platform to study coloniality in stony corals. However, employing bailed-out polyps in coral research requires greater understanding of the biology of this stress response.
This thesis investigates the molecular basis of polyp bail-out in Pocillopora acuta, a branching coral that is common in the oceans around Okinawa, Japan, and examines bailed-out polyps to study the biological foundation of coral coloniality. First, I probe molecular mechanisms involved in hyperosmosis-induced polyp bail-out, based upon a P. acuta transcriptome assembly. Then, I monitor morphological and genetic changes of solitary P. acuta polyps after the induced bail-out response. Finally, I explore transcriptional profiles of bailed-out polyps and those of intact colonies in order to identify differences between corals at different levels of structural and social complexity.
Based on transcriptomic data, activation of TNF and FGF signaling pathways was identified during initiation of polyp bail-out, possibly linking them to the colony-dissociation and polyp-detachment processes in bail-out, respectively. Under ambient conditions, about half of bailed-out polyps displayed morphological recovery and genetic resumption of fundamental cellular processes within five days. When compared to recovered solitary polyps, healthy colonies showed activation of genes for neurological and circulatory system development and those with potential molecular transport regulatory functions. Furthermore, in response to environmental stresses, few genetic changes were shared by polyps and colonies, suggesting that coloniality promotes distinctively different stress responses, probably enhancing fitness in stony corals. Interestingly, transcriptomic data also revealed possible participation of the activin signaling system in development of coloniality.
This thesis presents a robust polyp bail-out induction protocol and develops a foundation for its application to coral research. Using this new research model, this thesis presents the first molecular-level study of coral coloniality and identifies genes potentially participating in functional integration of coral colonies, which are expected to be fruitful topics for future studies.

Tsung-Han Hsieh

Supervisor Name: Hiroki Ishikawa

Research Unit: Immune Signal Unit

Thesis: Deciphering the Role of AP-1 Transcription Factor JunB in CD4+ T Cells

Abstract:

IRF4 is critical for differentiation of various CD4+ effector T cells, such as T helper 1 (Th1), Th2, and Th17 subsets, through interaction with BATF-containing AP-1 heterodimers. A major BATF heterodimeric partner, JunB, regulates Th17 differentiation, but the role of JunB in other CD4+ effector T subsets is not fully understood. Here we demonstrate that JunB is essential for accumulation of Th1 and Th2 cells, as well as Th17 cells, both in vitro and in vivo. In mice immunized with lipopolysaccharide (LPS), papain, or complete Freund’s adjuvant (CFA), that induce predominantly Th1, Th2 and Th17 cells, respectively, accumulation of antigen-primed, Junb-deficient CD4+ T cells is significantly impaired. Loss of JunB decreases viability of cells activated under Th1-, Th2-, and Th17-polarizing conditions. RNA-sequencing (RNA-seq) and chromatin immunoprecipitation sequencing (ChIP-seq) reveal that JunB directly regulates expression of various genes that are commonly induced in priming of naïve CD4+ T cells, including a pro-apoptotic gene Bcl2l11 (encoding Bim), and genes that are specifically induced in Th1, Th2, and Th17 cells. Furthermore, JunB colocalizes with BATF and IRF4 at genomic regions for approximately half of JunB direct target genes. Taken together, JunB, in collaboration with BATF and IRF4, serves a critical function in differentiation of diverse CD4+ T cells by controlling common and lineage-specific gene expression.

Dong Cao

Supervisor Name: Ichiro Maruyama

Research Unit: Information Processing Biology Unit

Thesis: Investigation of Circular RNA Regulation by Cis and Trans Elements in Caenorhabditis Elegans

Abstract:

Circular RNAs (circRNAs) are regulatory molecules that show diverse functions. However, the regulation of circRNA formation is not well-understood. Through the large-scale isolation of neurons from the first larval stage of Caenorhabditis elegans, C. elegans neuronal circRNA profile was analyzed for the first time by RNA sequencing of the whole-transcriptome. Using circRNAs identified in this dataset, the regulation of circRNA by cis and trans elements was studied in vivo. It was found that reverse complementary matches (cis elements) found in introns that surround circRNA-forming exon(s) not only promote circRNA formation but also are involved in skipping of the exon(s). Through one-by-one mutagenesis of all the splicing sites and branch points required for exon-skipping and back-splicing in the zip-2 gene, it was uncovered that exon-skipping and back-splicing independently occur at the same time. Thirteen RNA binding proteins were screened as trans elements that regulate the formation of circRNAs in C. elegans. Among them, FUST-1, the homolog of FUS (fused in sarcoma), down-regulates both circRNA formation and exon skipping of the same pre-mRNA without affecting its cognate linear mRNA levels. In zip-2, the 5’ splice sites for back splicing and exon skipping seem to be important for FUST-1 to regulate exon-skipping and back-splicing, respectively. When two mutations (R446S and P447L), which are equivalent to natural mutations in the FUS nuclear localization signal (R524S and P525L) observed in amyotrophic lateral sclerosis, were introduced into FUST-1, both of the mutations dramatically down-regulated the formation of circRNAs. Moreover, an auto-regulation by FUST-1 "isoform a" was found to be required for the production of FUST-1 "isoform b", which has a different N-terminal sequence from "isoform a", by promoting the skipping of exon 5 of its own pre-mRNA. FUST-1, "isoform a" is a functional isoform for the regulation of circRNA formation, whereas "isoform b" is non-functional for the regulation exon-skipping or circRNA formation.

Joel Perez Urquizo

Supervisor Name: Keshav Dani

Research Unit: Femtosecond Spectroscopy Unit

Thesis: Terahertz Patch Antenna Microcavity Lasers with Integrated Beam Control

Abstract:

Terahertz (THz) refers to the region of the electromagnetic spectrum that lies in between the infrared and microwaves. This frequency range possesses great potential to host several applications in wide-ranging fields, such as wireless communications, astronomy, non-invasive imaging and security scanning. However, despite sustained progress over the past decade, THz technology has not yet reached the level of maturity and flexibility of the neighboring radiofrequency (RF) and optical range. One missing key aspect is the ability of integrating advanced beam control functionalities within a monolithic platform. A promising approach to achieve this goal is to combine within a single device two features of the neighboring ranges: optical microcavities, that can sustain efficient lasing operation; and antenna arrays, providing a high level of beam control. In this thesis, we investigate via simulations fabrication and characterization the emission properties of arrays of patch antenna-coupled microcavities embedding quantum cascade active regions. The geometrical configuration of the array allows independent and simultaneous tuning of the losses governing the microcavities as well as beam shaping by constructive interference in the far-field. We show that optimized arrays emit THz with unprecedented low beam divergence and robust lasing in single frequency and spatial mode. Additionally, we demonstrate polarization functionalization by coupling the patch antenna microcavities with plasmonic wires. This feature introduces an additional degree of freedom to adjust the relative emission from the cross-polarized modes of the patch, allowing the device to radiate with any coherent polarization state from linear to circular. Finally, we discuss how this design can further enable other advanced functionalities such as active beam steering and control of THz non-linearities. The successful implementation of integrated advanced functionalities and sources on-a-chip demonstrates the ability of our platform to replicate in the THz range the beam control concepts used in the RF and optics, thus paving the way towards establishing a mature technology in this range of the electromagnetic spectrum.

Kun-Lung Li

Supervisor Name: Hiroshi Watanabe

Research Unit: Evolutionary Neurobiology Unit

Thesis: A Study of Horizontally Transferred Glycosyl Hydrolase Family 6 Genes in Tunicate Genomes

Abstract:

Tunicates are the closest extant relatives of vertebrates. Tunicates produce cellulose-containing tunic and exhibit very different lifestyles among animals. Their unique ability to synthesize cellulose results from a horizontally transferred cellulose synthase gene (CesA). Interestingly, a Glycosyl Hydrolase Family 6 (GH6) hydrolase-like domain exists at the C-terminus of tunicate CesA but not in cellulose synthases of other organisms. This led to the identification of another independent GH6-encoding gene, GH6-1, in tunicate genomes. These GH6-encoding genes exist exclusively in tunicates within the animal kingdom. The existence of GH6-encoding genes and the combination of a GH6 and a cellulose synthase domain in tunicate genomes raised the question of the evolutionary origin and function of GH6s in tunicates. To answer these questions, I first examined the phylogenetic relationship of GH6-encoding genes by comparing their sequence signatures. Secondly, I examined the expression of tunicate CesA and GH6-1 genes in Ciona intestinalis type A, a model ascidian tunicate. The gene expression in embryos at early developmental stages was examined by quantitative reverse transcription PCR and in situ hybridization. The results were also compared with a set of single-cell transcriptome data provided by our collaborators. Finally, I generated GH6-1 knockout larvae of C. intestinalis type A to observe the phenotypes of function-less larvae. This study would help to address how tunicates evolved by obtaining their unique anatomy and ecology.

Andreas Thomasen

Supervisor Name: Nic Shannon

Research Unit: Theory of Quantum Matter Unit

Thesis: Topology of Band-Like Excitations in Frustrated Magnets and Their Experimental Signatures

Abstract:

The discovery of the integer quantum Hall effect, an inherently topological effect, heralded a new era in condensed matter physics, and the properties of topologically non-trivial bands of electrons have been extremely well-studied. Magnetic insulators also exhibit band like-excitations, and the topological properties of these bands are now an active frontier of research. In this Thesis we examine the topology of two different types of magnetic excitation, and some of the experimental signatures which result.
We consider first the triplon excitations of a quantum paramagnet on a bilayer breathing kagome lattice. We show that the most general model allowed by lattice symmetry can support topologically non-trivial triplon bands, and explore the connection with the Z2 topology found for electrons with spin-orbit coupling. In particular, we examine the conditions required for the stability of a Z2 invariant in the absence of a Kramers degeneracy. We also establish signatures for bands with finite Berry curvature coming from thermal and spin transport measurements.
We then turn to the magnon excitations of kagome-lattice system saturated by high magnetic field. We again consider the most general model allowed by the symmetry of the lattice, and examine the way in which symmetry dictates gaps between bands, and the associated topology of those bands. We also make explicit predictions for inelastic neutron scattering, and explore what can be learned about the topology of magnon bands from the changes seen in pinch-point and half moon features associated with a quadratic band touching.
These results shed light on some of the features which distinguish topological bands of magnetic excitaitons from topological bands of electrons, as well as the opportunities for observing them in experiment.

Jason Robert Ball

Supervisor Name: Denis Konstantinov

Research Unit: Quantum Dynamics Unit

Thesis: Investigating Color Centers in Diamond for Microwave Quantum Technologies

Abstract:

The rapid advancement of quantum technology in recent years has necessitated the de-velopment of many specialized microwave components such as the Josephson parametricamplifier (JPA), to be used in conjunction with superconducting qubits. In this thesis, Ipresent research into impurity spins in diamond, which include nitrogen (P1), nitrogen-vacancy (NV), and vacancy clusters, in particular (R5) centers, for use in these microwavequantum technologies. I first present the development of a 3D loop-gap microwave res-onator for use with these spin centers. We were able to demonstrate strong coupling withan ensemble of nitrogen-vacancy (NV) centers as well as that of nitrogen (P1) centers at 10mK. I next demonstrate two separate but related maser effects in the spin ensemble. Thefirst of which is a thermally-generated inversion of the NV centers, produced by an abruptcooling of the sample with a lifetime of several hours. Finally, I demonstrate another spin-based cavity amplifier, this time using the P1 centers and an active microwave pumpingscheme. This cavity amplifier has several desirable qualities, including a large gain andlow noise temperature. Such an amplifier may be promising for future applications.

Afshan Jamshaid

Supervisor Name: Yabing Qi

Research Unit: Energy Materials and Surface Sciences Unit

Thesis: Scanning Probe Microscopy Studies of Metal Halide Perovskite Materials

Abstract:

So far very little atomic-scale research has been done to understand the influence of additives on the structural and electronic properties of halide perovskite materials (e.g., MAPbI3=CH3NH3PbI3). Additives have been suggested to solve the thermal instability and ambient air induced degradation problems of perovskites. Among studied additives, Cl and KI were found to be eligible candidates for improving the power conversion efficiencies (PCEs) and degradation issues in perovskite solar cells. Instabilities and degradation occur in perovskite materials due to the interaction of water, oxygen, light, and temperature stimuli during solar cell operation. To solve instability and degradation problems in perovskite materials, it is imperative to study the fundamental processes of the instabilities and degradation of perovskite at the atomic scale. In this thesis, scanning tunneling microscopy (STM) and density functional theory (DFT) calculations provide a fundamental understanding of the origin of the Cl interactions with MAPbI3 and provide useful hints for the design of stable and high-performance perovskite solar cells. My results provide information about different Cl and (KI) concentrations that play an important role in MAPbI3. (1) I found that at low coverage, Cl does not incorporate in MAPbI3 surface lattice (corroborated by X-ray photoelectron spectroscopy measurements). (2) However, with increasing Cl concentration, it is possible to visualize Cl ions on MAPbI3 surfaces. With a concentration of [Cl]:[I] ~ 18% (MAPbI2.54Cl0.46), I observed that Cl ions were mostly incorporated close to the grain boundaries, (3) Additionally, Cl incorporation was also observed in the center regions of grains that have different structures (namely, the dimer and zigzag structures). (4) My study evidence that Cl can substitute I ions of the surface lattice structure and/or fill the surface I-vacancies in MAPbI3. (5) Furthermore, Cl incorporation alters the bandgap of MAPbI3 from 1.45 eV to 1.65 eV which was characterized by ultraviolet and inverse photoemission spectroscopy (UPS/IPES). (6) Finally, the Cl incorporation ratio was verified experimentally by performing Fourier transform infrared (FTIR) and XRD experiments on thick films, and I determined the optimal Cl incorporation ratio that leads to enhanced Cl doped MAPbI3 perovskite stability. It is important to point out that STM can be used to characterize mainly the surface (and possibly the subsurface) atomic structures of perovskite materials. Although it is challenging for STM to provide a complete picture (especially regarding the bulk structures of perovskites), my study does offer valuable insights into the surface/interface properties. For example, (i) Cl diffusion into the perovskite film (ii) beneficial effects of Cl at the surface/interface on charge extraction (iii) passivation of surface defects by Cl.

Mathias Mikkelsen

Supervisor Name: Thomas Busch

Research Unit: Quantum Systems Unit

Thesis: Quenched and Driven Dynamics of One-Dimensional Quantum Systems

Abstract:

In this thesis I use quenched and driven dynamics to explore the structure of few and many-body quantum systems, with a particular emphasis on the unitary dynamics of closed, one-dimensional quantum gases with short-range interactions.

As a first example I investigate a gas of strongly-interacting bosons in a one-dimensional ring lattice geometry, which is initially in an equilibrium phase and to which a constant rotation of the lattice is suddenly added. In a second project I explore the connection between non-equilibrium excitations described by the work statistics of a quench and the information scrambling defined by the squared commutator for relevant operators using analytic solutions for the paradigmatic example of two interacting particles in a harmonic trap. To be able to also describe larger systems, I then present an improved method of exact diagonalization in Fock space for multi-component few-body continuum systems incorporating an effective interaction and show some examples where this can be applied. Finally, I describe an investigation into the effects of nonlinear interactions in an open optomechanical setup, where a steady state is obtained due to the interplay between input and loss, using a Quantum Langevin formalism.

Seyedeh Sahar Seyed Hejazi

Supervisor Name: Thomas Busch

Research Unit: Quantum Systems Unit

Thesis: Atom-Light Interactions via Evanescent Fields

Abstract:

Newly developed techniques for controlling and measuring quantum systems have recently created an interest in exploring how the presence of dielectric surfaces affects atomic systems. In this thesis, I present results obtained by studying how the presence of evanescent modes or an evanescent field emerging from a dielectric medium can affect different quantum system, such as one or two atom systems, or even multi-component Bose–Einstein condensates.

Evanescent fields are exponentially decaying fields which typically appear on the surface of dielectric systems, such as flat half-planes, optical nanofibers or prisms. By bringing atoms close to the dielectic surfaces, they can couple to the evanescent modes, leading to new effects stemming from their guided nature or the spatial inhomogeneity. In the first project I present results on how the dipole-dipole interaction between two atoms can be enhanced close to the surface, which causes changes in various quantities, such as decay rates and frequency shift. In particular, the coupling between two different atoms depends on the orientation of their respective electric dipole moments and their relative location, which can lead to directional propagation of information between atoms and an oscillatory the decay rate. For multi-level atoms placed in the vicinity of an optical nanofiber, the spontaneous emission rates become a function of the magnetic sub-levels and orientation of electric dipole moments of atoms, in addition to the dependence on the optical modes of the fiber.

In a second project I consider the interesting feature of atom-fiber systems that is known chiral emission, and calculate the resulting chiral force on the atom. I show that it depends on various modes of fiber as well as on the orientation of electric dipole moment of atom.

When going beyond small systems and considering Bose-Einstein condensates, these fields can also act as artificial gauge fields and I show how they affect the phase-separation phase transition in two-component condensates.

Shoko Igarashi (Ota)

Supervisor Name: Kenji Doya

Research Unit: Neural Computation Unit

Thesis: Intrinsic Motivation in Creative Activity

Abstract:

Humans are born to be creative. Humans spontaneously make an effort to create something new and meaningful even without obtaining any extrinsic rewards. Intrinsic motivation is a desire for learning for its own sake (Berlyne, 1966) and a fundamental condition for intelligence and creativity (Boden, 1998).
Researches have found important drives for intrinsic motivation (Berlyne, 1966; Csikszentmihalyi 1997; Ryan and Deci, 2000), however, little is known about which factors of learning environment are essential for promoting intrinsic motivation when humans engage in creation.

Novelty, variety, and prediction errors are hypothesized as the key factors to explain the mechanisms of intrinsic motivation in learning (Barto et al., 2019; Oudeyer and Kalan, 2007; Schmidhuber, 1991), but those hypotheses are yet to be tested in human creative activities. Here, we propose a hypothesis that intrinsic motivation in creative activity is facilitated when humans can observe higher variety of expressions using simpler rules. Typical examples are composing haiku, solving problems and proving a theorem in physics, and inventing new technologies for new experiences. To examine the hypothesis, a novel human behavioral experiment was designed and conducted with an original computer game based on a framework of the Game of Life cellular automata (Conway, 1970).

The simplicity of a rule is controlled by parameters of state transition function and quantified by the complexity measures formulated in the theory of cellular automata. The variety of expression is quantified by the number of live cells, the number of cell state transitions, the entropy of the distribution of local patterns observed by participants, and empowerment using information theory. The degree of intrinsic motivation is measured by the subjective scores of enjoyment in the questionnaire after the task, the playing time before getting bored, the frequency of touching interaction. These measurements are compared between four rules (Rule 1 to 4) with different simplicity and variety of expression and tested with 42 participants.

The results of two-way ANOVA of the scores of enjoyment showed that participants are more motivated with a higher variety of expression and the simpler rule, while there is no interaction between two parameters. The results of multiple regression analyses also support a part of the hypothesis that the entropy of local patterns is related to intrinsic motivation. We further performed canonical correlation analysis of the variables related to intrinsic motivation versus behavioral measures and the result also supports our hypothesis that a variety of expression promotes intrinsic motivation. Classification of subjects by their preference of the entropy of local patterns provides a better fit of regression analysis, which suggests subtypes of participants favoring different components of our intrinsic motivation models. Comprehensive analyses show that subjects took different attitudes, such as watching autonomously evolving patterns or actively trying to identify the rules, which points to the possibility of further analysis with the distinction of participants based on their types of curiosity using clustering techniques.

Sandrine Burriel

Supervisor Name: Tadashi Yamamoto

Research Unit: Cell Signal Unit

Thesis: Functional analysis of LMTK1 in lung adenocarcinoma

Abstract:

Kinases are master regulators of key cellular processes, including those controlling cell proliferation and the cell cycle. Mutations, abnormal gene expression and aberrant activity of kinases are known to contribute to the development and progression of cancer.

The lemur tail kinase (LMTK) family was discovered 20 years ago but its function in health and disease, particularly in the context of cancer, remains largely unknown. The founding member of the family, LMTK1, also known as apoptosis-associated tyrosine kinase (AATK/AATYK), possesses two splice variants: a transmembrane protein (LMTK1B) and a cytosolic protein (LMTK1A). LMTK1 is predominantly expressed in the brain where it regulates axon and dendrite formation in neurons through its kinase activity. Independent of its function as a kinase, LMTK1 has also been found to act as a scaffold protein, recruiting protein phosphatase 1 (PP1) and SPAK to regulate the activity of the NKCC1 cotransporter. Recent work has established that, in several types of cancers, LMTK1 is downregulated due to DNA methylation of the associated AATK gene, which contributes to cancer progression.

In this project, we investigated the role of LMTK1 lung cancer progression, using a combination of computational analyses based on patient data from The Cancer Genome Atlas and experimental approaches, including RNA sequencing and biochemical analyses. Through patient data analysis, we confirmed that LMTK1 mRNA expression is downregulated in the tumors of non-small cell lung cancer patients, correlating with a moderate impact on survival for lung adenocarcinoma patients. In a lung adenocarcinoma cell model (A549 cells), restoring expression of both splice variants of LMTK1 reduces cell proliferation in a comparable manner, pointing to similar functions for both variants in the context of cell growth. Mutating either the kinase active site or the PP1 docking site showed that, in A549 cells, the kinase activity, rather than the protein scaffold function, is the main driver of reduced proliferation. After hydroxyurea mediated synchronization of cells in late G1 phase, cell cycle analysis showed that LMTK1A expression leads to slower progression from G1 through S phase, concomitant with decreased expression of Cyclin E1 and increased expression of p21. E2F1, a transcription factor known to regulate Cyclin E1 expression and itself target of YAP transcriptional activity, was also found to be decreased. Immunofluorescence showed that wild-type LMTK1A expression dramatically slowed nuclear translocation of YAP, and this effect was abrogated by the kinase-negative LMTK1A mutant. Our data suggest that this effect of LMTK1A might be independent of Rab11A activity, but that LMTK1A may mediate interactions between CD44, Merlin and Akt leading to stabilization of the actin cytoskeleton and enhanced regulation of YAP nuclear translocation. We hypothesise a novel role for LMTK1A kinase activity as a regulator of the Hippo pathway.

Shivani Sathish

Supervisor Name: Amy Shen

Research Unit: Micro/Bio/Nanofluidics Unit

Thesis: Surface-based Microfluidic Systems for Enhanced Biomarker Detection

Abstract:

Microfluidic surface based immunoassays have been evolving since 1980`s to serve as tools for disease diagnostics. These systems rely on detection of diagnostic biomarkers in solution when they bind to capture molecules attached to the surfaces of the microfluidic device. The “efficiency” of the microfluidic device is conventionally defined by the following features: (1) time taken to detect biomolecular binding events, (2) the volume of reagents needed, and (3) the limit of detection of the system. These features are directly influenced by the cumulative effect of several inter-related factors. This thesis explores the influence of two major fundamental factors, i.e., surface chemistry, and mass transport limitations, both independently and collectively, to ultimately design a versatile and “high efficiency” diagnostic tool.

As conventional microfluidic devices are fabricated with either glass or thermoplastic substrates, surface chemistries that enabled robust immobilization of capture biomolecules were investigated for each substrate. Microcontact printing of silanes on glass substrates was employed to achieve covalent biomolecule immobilization in microfluidic channels. Here, the silane-micropatterned glass substrates were chemically coupled with proteins and DNA aptamers using standard 1-ethyl-3-(3-(dimethylamino)propyl) carbodiimide (EDC) and N-hydroxysuccinimide (NHS) chemistry. On the other hand, highly reactive ionized air plasma was exploited to generate polar groups on poly(methyl methacrylate) substrates that enabled subsequent coupling of proteins using the same EDC—NHS chemistry. Both strategies facilitated effective covalent biomolecule immobilization in microfluidic channels, while preserving their molecular properties to serve as effective binding partners to complimentary biomarkers.

Subsequently, mass transfer analysis was performed to investigate the conditions that entail fast and sensitive detection of biomolecular binding events within microfluidic channels. To this end, rigid, 3D microfluidic devices were fabricated using the LightFab 3D Printer, to eliminate concerns of channel deformation due to flow. The microfluidic devices were biofunctionalized with capture biomolecules using the developed silane microcontact printing method. These devices were employed to visualize binding events between the capture biomolecules and complimentary fluorescently-labelled immunoglobulins (IgGs), via fluorescence microscopy. The effect of high shear flow rates on the concentration-dependent speed of the binding reaction and volume of reagents required for the microfluidic assay were described by dimensionless mathematical parameters. These parameters were derived by detailed scaling processes employing respective governing equations that describe the entire system, i.e., the Navier-Stokes equations, convection-diffusion equation and the Langmuir adsorption model. The conditions that enabled rapid detection of biomolecular binding events were extracted and employed to detect IgGs specific to Covid-19 antigen fragments in the fabricated devices. Subsequently, using these design rules, multiplexed microfluidic devices were developed to enable detection of IgGs specific to 3 different antigen fragments, in parallel.

Finally, an integrated, completely sealed, and disposable point‐of‐care testing prototype was developed by incorporating the refined rapid microfluidic immunoassay devices and 3D‐printing fabrication techniques. The palm‐sized modular prototype consists of a manually operated fluid handling device that allows precise mixing, filtration, and delivery of fluids to an on‐board microfluidic assay unit for subsequent detection of specific biochemical analytes, with a minimized risk of contamination.

Kamila Mustafina

Supervisor Name: Yohei Yokobayashi

Research Unit: Nucleic Acid Chemistry and Engineering Unit

Thesis: Engineering Synthetic Riboswitches for Mammalian Cells

Abstract:

Riboswitches are natural and artificial noncoding RNA elements capable of controlling gene expression in response to chemical signals without direct involvement of protein factors. One strategy for engineering synthetic riboswitches is to combine an aptamer –a short RNA sequence that specifically binds to a ligand– and a self-cleaving ribozyme to create an aptazyme whose self-cleavage activity is regulated by the aptamer ligand. These aptazymes can be embedded in the 3’UTR of mRNAs to chemically control gene expression in mammalian cells. This property of riboswitches opens a wide area of applications in biology and medicine. However, engineering riboswitches that function efficiently in mammalian cells remains challenging, partly due to the difficulties associated with generating and screening aptamers and aptazymes that function in the cellular environment rather than in test tubes.

In this thesis, I introduce two new ribozyme scaffolds for aptazyme engineering in mammalian cells. First, I identified highly active variants in mammalian cells from the twister and pistol ribozyme families. Then I used them as scaffolds for a new aptazyme architecture, where the aptamer is placed immediately upstream of the ribozyme in a tandem configuration. I optimized this design in mammalian cells, and then generated randomized libraries of 4096 aptazyme variants for high-throughput in vitro screening to identify switches with high ON/OFF ratios. Although the method allowed characterization of a large number of variants, their activities were not always reproducible when tested in cells.

Therefore, in addition to in vitro screening, I explored rational design approaches for the same tandem architecture. I fine-tuned the activity of the aptazyme by systematically varying the length of the inserted competing stem and introducing single-nucleotide mismatches and spacers. Using this method, I developed mammalian riboswitches with ON/OFF ratios greater than 6.0 for the twister scaffold, and greater than 5.0 for the circularly permuted pistol scaffold.

Lastly, learning from the experience of high-throughput in vitro screening and rational design in cells, I used high-throughput sequencing to directly screen for functional aptazymes in mammalian cells by quantifying the uncleaved fractions of aptazyme-embedded mRNAs. I verified this method with a small twister ribozyme library containing 256 variants and then applied it for a larger circularly permuted pistol ribozyme library consisting of 1024 variants.

This work expands both the tools and the methods available in the field of RNA engineering. Riboswitches based on the new ribozyme scaffolds provide compact and tunable tools for controlling gene expression. Rational and high-throughput design strategies developed in this thesis can be applied to generate other RNA devices for biomedical and synthetic biology applications.

Sakurako Watanabe

Supervisor Name: Jeff Wickens

Research Unit: Neurobiology Research Unit

Thesis: Interaction of multiple inputs in plasticity of the corticostriatal synapses

Abstract:

Spike timing-dependent plasticity (STDP) is a form of synaptic plasticity and physiologically relevant model of Hebbian learning. STDP depends on the relative timing of pre- and postsynaptic action potentials. In the brain, synaptic plasticity occurs in the context of concurrently occurring multiple inputs into the same neuron. Current theoretical and computational models assume that each presynaptic conditioning signal is independently processed in the neuron. However, in experimental STDP studies, different time points are measured in different cells rather than in the same cell. In the striatum, in particular spiny projection neurons (SPNs) receive many inputs including cortical and dopaminergic inputs during learning. Here, we investigated the plasticity of contiguous inputs from the cortex depending on their relative timing to firing in the SPN and tested whether dopamine could differentially modulate plasticity from those inputs. We performed whole-cell electrophysiological recordings in the SPNs of the dorsal medial striatum (DMS) of the acute brain slices of mice that express Drd1a/Drd2-eGFP or Drd1a-tdTomato to identify SPN cell type (D1 or D2). We also used Ai32 (RCL-ChR2(H134R)/EYFP) / DAT-Cre / Drd1a-tdTomato triple transgenic mice to enable temporally controlled release of dopamine with optogenetics. In order to test this selection ability, we stimulated cortical afferents from two independent sets of bipolar electrodes, each stimulating at different times in relation to SPN firing. We found that two presynaptic inputs interact with each other and show sublinear, linear or superlinear summation when stimulated simultaneously. In our STDP protocol, two near-simultaneous inputs were applied to a single dSPN; one input (S1) at positive timing and the other (S2) at negative timing in relation to postsynaptic firing. Surprisingly, both pre-post and post-pre pairing resulted in a decrease in normalized EPSPs, indicating long-term depression (LTD). This contrasts with the results of standard STDP protocols with single inputs, which show different direction of plasticity depending on the relative timing to firing. In addition, we tested whether these inputs were sufficient to induce plasticity in the absence of postsynaptic firing. When only S1 and S2 inputs were recorded without postsynaptic firing, LTD was observed. This suggests that EPSPs in close temporal proximity could initiate plasticity. We then applied dopamine two seconds after each pairing, which is known to induce LTP when tested with one presynaptic input. We found that interaction of multiple cortical inputs changes the outcome of corticostriatal STDP. In summary, the results suggest that multiple contiguous presynaptic inputs to the same neuron can induce plasticity independently, but they can also interact with each other and affect plasticity outcome. This is the first study demonstrating the ability of SPNs to differentiate multiple plasticity-inducing inputs from the cortex. The results with dopamine stimulation suggest that certain interaction of multiple cortical inputs favours LTP to occur with the right timing of rewarding dopamine application. This highlights the importance of a transient eligibility trace in the synapse to enable LTP and subsequent learning to occur.

Ayaka Usui

Supervisor Name: Thomas Busch

Research Unit: Quantum Systems Unit

Thesis: Control and measurement of non-classical properties of cold atomic and optical systems

Abstract:

In this work I consider various systems from the area of ultracold quantum gases and quantum optics to reveal non-classical correlations and features. First, I investigate the presence of a dynamical phase transition in a system of cold atoms trapped in a one-dimensional optical lattices. To the best of my knowledge this has been the first work to study dynamical phase transitions in a continuous model, and it revealed the relation between the dynamical phase transition and temporal orthogonality. Second, I consider an impurity coupled to a gas in a two dimensional lattice. This work has revealed the dynamics of the impurity and proposed an approach to probe the local excitation spectrum of the gas at the site that is coupled to the impurity. Third, I study two strongly interacting bosons with synthetic spin-orbit coupling. This work discovers the ground state beyond the mean-field regime and explores non-classical correlations in the ground state. Fourth, I present my contribution to a project on Bayesian estimation with continuous-variable systems, which are realised with optical setups. I explore what is the best probe state for heterodyne or homodyne measurements to estimate any phase rotation. Finally, I consider qubit and higher dimensional machines interfacing with an environment and monitoring a target system. The behaviour of the target is too complicated in general to reveal the dynamics and even the steady state. I suggest to use a virtual qubit as a tool to predict the quantitative behaviour of the target qubit. For this I generalise the idea of the virtual qubit to higher-level target systems so that one can design autonomous quantum machines beyond a few qubits.

Lin Li

Supervisor Name: Pinaki Chakraborty

Research Unit: Fluid Mechanics Unit

Thesis: Storm moisture and landfalling hurricanes

Abstract:

When a hurricane strikes land, it can cause immense destruction to humans and other terrestrial life. Despite its importance, few studies have focused on the dynamics of hurricanes past landfall. By contrast, how hurricanes originate and grow over warm oceans has been extensively studied. Over oceans, moisture from the ocean fuels the intense winds of a hurricane heat engine. Past landfall, the hurricane is severed from this heat source, and its intensity decays. It is generally thought that this decay is a non-thermodynamic process that is dominated by frictional drag with the land surface and where the moisture from the ocean plays no role. Challenging this notion, we argue that the "storm moisture" --- the stock moisture which a hurricane carries as it strikes land --- constitutes a source of heat that modulates the decay of intensity and shapes the internal structure of hurricanes past landfall. The critical, albeit unrecognized, role of storm moisture forms the leitmotif of this thesis, which consists of three parts. First, we analyze intensity data for North Atlantic landfalling hurricanes to show that the timescale of hurricane decay has increased in direct proportion to a contemporaneous rise in the sea-surface temperature over the past 50 years. Second, we show that a landfalling hurricane generates a cold-core in its eye that is distinct from the cold-core in an extratropical cyclone. Third, we show that as the cold-core grows past landfall, its competition with Ekman pumping restructures the flow in the hurricane and can lead to the splitting of the hurricane. In all three cases, we show, using a blend of theory, computational simulations, and field observations, that the storm moisture dictates the hurricane dynamics past landfall.

Christina Ripken

Supervisor Name: Síle Nic Chormaic

Research Unit: Light-Matter Interactions for Quantum Technologies Unit

Thesis: Micro- and Nanoplastics in Okinawa – Potential Impacts on Planktonic Microalgae and Endosymbiotic Dinoflagellates

Abstract:

Over the last century, anthropogenic influence on the marine ecosystem has increased and diversified, reaching from ocean acidification and temperature rise, to increased organic and inorganic pollution. Plastic pollution, as part of synthetic marine litter, has received a lot of attention, both scientific and societal, due to its high visibility across all marine ecosystems. Smaller plastic particles, though already discovered as early as the 1970s, have only recently gained recognition as a major problem for marine life.

The risk and challenge that the presence of micro- and nano-particles and -fibers poses to the various marine ecosystems are diverse and largely unquantified. This study aims to increase the knowledge about the interactions of these particles with microalgae of two different ecosystems. In addition, an initial assessment of the presence and distribution of micro- and nanoplastics around Okinawa is given.

The first part of this thesis is a field study of the coastal marine waters of Okinawa. No data existed about the micro- and nanoplastic pollution around Okinawa so far. This initial assessment offers, therefore, a first screen shot of abundance, distribution, type, and trace metal contamination of micro- and nano-sized plastic particles in this area. As Okinawa is located far from other big islands or continents, it was investigated if the plastic pollution patterns reflect population and industrial distribution of the island, as most marine plastic pollution comes from land. A risk assessment of the microplastics and associated trace metals was performed using the pollution load index and correlated with the marine pollution patterns around Okinawa.

The second part assesses the physiological effects which nanoplastics can have on micro-algal endosymbionts (the dinoflagellate Breviolum minutum, former Symbiodinium minutum) that are hosted by corals and sea anemones. These endosymbionts start free living and would get into contact with the plastic before they are taken up by their host. In this experiment the dinoflagellate are exposed to nanoplastics in a rolling tank for 10 days under optimal growth conditions. Changes in RNA expression patterns can be related to potential effects of nanoplastic on the symbiont-host dynamic, as certain expression patterns are needed for a successful integration of the symbiont into the host. Differential gene expression analysis can indicate how the symbiont-host dynamic might change in future due to the presence of nanoplastics in the reef ecosystem.

The third part of this thesis investigates the aggregate formation between microfibers or nanoplastics with marine planktonic primary producers (diatomes such as Skeletonema grethae or Odonetella aurita or the cyanobacterium Synechococcus elongatus), in order to understand how the biological carbon pump can be infiltrated by microfibers and nanoplastics in open ocean ecosystems. Phytoplankton (400,000 cells per mL) were exposed to different plastic concentrations ranging from 100 ng/L, 10 µg/L, 1 mg/L to 10 mg/L over a 7-day dark period under constant rotation in a rolling tank. Rolling tank incubations promote aggregate formation and simulate the falling of aggregates through the water column. Incorporation of micro-or nana- plastics into aggregates leads to the transport of these plastics to the seafloor. Effects of the plastic particles on formation and sinking velocities of aggregates were recorded, as well as photosynthesis rates, transparent exopolymeric particles (TEP) production and 3D imaging data.

Adrian David

Supervisor Name: Shinobu Hikami

Research Unit: Mathematical and Theoretical Physics Unit

Thesis: Higher-spin holography in de Sitter space: horizon modes, black holes, and the boundary partition function

Abstract:

In a strict sense, the problem of quantum gravity has been solved by string theory through the AdS/CFT correspondence in settings restricted to the spatial infinity of a universe with negative cosmological constant. Currently, there are outstanding questions of quantum gravity restricted to finite regions, as in the case of a positive cosmological constant where an observer only has access to partial information, as is the case for our observed universe.

De Sitter (dS) space is a natural toy model for quantum gravity in finite regions, however no satisfactory string-theoretic description of dS has been put forward. Recently, a new model for dS/CFT has been proposed with the Vasiliev bosonic higher-spin gravity as the bulk description, corresponding on CFT side to the Sp(2N) vector model.

In this body of work we review the recent progress concerning a higher-spin holographic duality over de Sitter space. The aim of the described approach is to construct a full holographic description of quantum gravity within the causal patch of a de Sitter observer. In particular, we focus on issues related to (i) cosmological horizon modes, (ii) black hole wordlines, and (iii) the boundary CFT partition function .

(i) We introduce [1] a spinor-helicity formalism to encode the data of massless fields of arbitrary spin on a cosmological horizon in de Sitter space. The evolution of free fields between past and future horizons (what might be called the free S-matrix in an observer’s causal patch) reduces to a simple Fourier transform in terms of these variables. We show how this arises via twistor theory, by decomposing the horizon-to-horizon problem into a pair of (more symmetric) horizon-to-twistor problems.

(ii) Similar to the static BPS black hole in AdS4 higher-spin theory [2], we solve the linearized Fronsdal equations with a source, and find the linearized version of the Didenko–Vasiliev black hole in de Sitter space. This is also shown to correspond holographically to the bilocal formulation of the boundary CFT.

(iii) We investigate the decomposition of the boundary CFT partition function in terms of spherical modes in the spinor-helicity basis. Further, we observe a discrepancy between the higher-spin-algebraic calculation of the partition function and the result of a direct calculation in the boundary CFT [3]; however, no such discrepancy arises at the level of n-point correlators, even when accounting for contact corrections. This paradox suggests a failure of locality in higher-spin theory, even on the boundary. A way forward from here is to introduce spin-locality as a replacement for spacetime locality, echoing recent developments in the bulk theory.

[1] A. David, N. Fischer, and Y. Neiman. Spinor-helicity variables for cosmological horizons in de Sitter space. Phys. Rev. D, 100:045005 (2019).

[2] V. E. Didenko, and M. A. Vasiliev. Static BPS black hole in 4d higher-spin gauge theory. Phys. Lett. B, 682:3, 305–315 (2009).

[3] D. Anninos et al. Cosmological shapes of higher-spin gravity. Journal of Cosmology and Astroparticle Physics, 2019:4 (2019).

Shijin Zhang

Supervisor Name: Ye Zhang

Research Unit: Bioinspired Soft Matter Unit

Thesis: Design and Synthesize Small Molecular Self-Assembling Peptides for Biomedical Applications

Abstract:

Molecular self-assembly is a spontaneous process of molecular association through non-covalent bond interaction. It is ubiquitous in biological system, for example, self-assembly of lipid bilayer and proteins. Imitating molecular self-assembly and creating novel functional molecular assemblies through synthetic chemistry are important issues in nanotechnology development. Small peptidic molecules are ideal candidates for imitating molecular self-assembly in biological system because of their biocompatibility and easy functionalization for bio-related applications including drug delivery, anti-cancer treatment and biosensors.

My thesis research focused on the design and synthesis of assembling peptides for novel nanostructure construction via self-assembly and co-assembly, and the exploration of their bio-related applications. The well-established building block naphthalene-phenylalanine-phenylalanine is selected for molecular engineering for pi-pi interaction and hydrogen bonding oriented self-assembly. Further molecule modifications by conjugating with light-responsive or enzyme-responsive motifs will endow peptidic molecules with regulated self-assembly behavior by controlling the external stimuli. Novel nanostructures in higher order can be reached either by multi-component co-assembly, or step-by-step assembly by triggering multiple external stimuli. Combination of such molecular design with medicinal functionalization is applied in the development of nanomedicine and biomedical engineering.

Andrew Winchester

Supervisor Name: Keshav Dani

Research Unit: Femtosecond Spectroscopy Unit

Thesis: Spatially and temporally resolved microscopy of traps in hybrid organic-inorganic perovskites

Abstract:

In recent years the class of materials known as hybrid organic-inorganic perovskite (HOIP) have received notable attention for use in photovoltaic applications, with record solar conversion efficiencies reaching other established thin film systems. Despite their rapid development, there are still ongoing issues related to heterogeneous film properties which limit device performance. It has been suggested that sites which capture charge carriers (traps) could be localized on a micrometer or smaller size scale, leading to regions of poor efficiency. Understanding the electronic properties of such regions, in particular on how they influence carrier recombination, will therefore provide crucial information about the carrier loss pathways in HOIP films, which will be essential for developing new strategies to minimize losses and create more efficient devices. In order to gain information about the ultrafast charge carrier recombination dynamics on nanoscale length scales, specialized techniques which can provide information with both high spatial and temporal resolution will be necessary. Here, we utilize time resolved photoemission electron microscopy (TR-PEEM) as a novel technique to study the nanoscale ultrafast properties of photo excited carriers and their relation to heterogeneous film properties in HOIP thin film materials. Following this overall theme, the work in this thesis will address several nanoscale properties and phenomenon. First, we will uncover the nanoscale distribution of carrier traps in a HOIP film which result in non-radiative losses. We then will describe in depth the ultrafast carrier trapping processes happening at nanoscale trap clusters. Following this, we will then discuss other novel information and studies on the traps in HOIP which can be realized using TR-PEEM, namely on effects of light treatments and morphological information. By gaining a deeper understanding in these directions, we hope to contribute to the broader goal of improving HOIP photovoltaic device efficiency and showcase TR-PEEM as a novel technique for studying photocarrier dynamics in semiconductor materials.

Krishna Priya Subramonian Rajasree

Supervisor Name: Síle Nic Chormaic

Research Unit: Light-Matter Interactions for Quantum Technologies Unit

Thesis: Rydberg Excitation and Other Multiphoton Processes in Cold Rubidium Atoms Near an Optical Nanofibre

Abstract:

Optical nanofibers (ONF) offer tight confinement and diffraction-free propagation over a long distance. When the fiber is smaller than the wavelength of the propagating light, a large evanescent light field extends beyond the fiber surface. This evanescent field, when coupled to surrounding atoms, provides a unique platform for studies on atom-light interactions for fundamental understandings of atomic physics and applications for quantum technologies. My thesis will present experimental investigations performed using an ONF embedded in a cold 87Rb atomic ensemble. Neutral atoms in highly excited states, namely Rydberg states, exhibit exotic features such as a long lifetime and a huge dipole moment making them strongly interacting systems. Rydberg atoms near an ONF are an important candidate for fiber-based quantum networks and the physics of Rydberg atoms at subwavelength distances from dielectric surfaces is unexplored. In my PhD work, two-photon excitation is used to excite atoms to a Rydberg level via evanescent field coupling. Successful creation of the Rydberg atoms next to the ONF paves the way for further studies on the Rydberg atom-ONF system. Aside from work on the highly excited Rydberg atoms, we have also investigated single-colour two-photon excitation at 993 nm addressing the 5S1/2 to 6S1/2 transition in ground state atoms mediated via an ONF. Interesting effects arising from multi-levels interacting with multiple light fields at different intensity regimes are studied.

Hong Huat Hoh

Supervisor Name: Tadashi Yamamoto

Research Unit: Cell Signal Unit

Thesis: Study on Alteration of Cellular Phenotypes and Processes in Cancer Using Exogenous Biological Agents

Abstract:

Cellular and genetic heterogeneity contributes to the complexity of cancer which poses challenges for cancer therapy. For instance, epithelial-to-mesenchymal transition (EMT) is a physiological phenomenon that was adopted to neoplasm to describe the possibility of carcinoma cells in acquiring mesenchymal traits, leading to invasion and metastatic dissemination. These EMT-induced tumor cells acquire cancer stem cells (CSCs) properties and contribute to heterogeneity within a tumor microenvironment. My thesis is set out to tackle the abovementioned phenomenon, with a common goal of investigating the potential of using exogenous biological agents in modifying the cellular phenotypes and processes of cancer. Firstly, I demonstrated the molecular mechanism and the ability of altering the stem cell property of breast cancer cells by downregulating CD44 molecules using exogenous miRNAs. Secondly, I investigated the potential of bioinspired integrin-targeted peptides to alter the microenvironment and subsequently cellular processes including metabolism of cancer cells for metastasis treatment in pancreatic cancer. Both studies utilized comprehensive molecular biology approaches in elucidating the functional changes and mechanisms behind the therapeutic effects of the biological agents using relevant cancer cell lines and animal xenotransplantation models. This thesis provides insights into cancer cell plasticity which can be harnessed for cancer therapy.

Sebastien Lapointe

Supervisor Name: Julia Khusnutdinova

Research Unit: Coordination Chemistry and Catalysis Unit

Thesis: Nickel Complexes of New Electron-Rich, Sterically-Hindered PNP Pincer Ligands

Abstract:

This thesis describes the synthesis, characterization and reactivity studies of nickel complexes of a new family of electron-rich, bulky PNP pincer ligands (PNP = 2,6-{R2PC(CH3)2}2-pyridine). The first chapter introduces the literature review of pincer complexes with a focus on nickel complexes and their applications. The second chapter introduces a new family of pyridine-based PNP pincer ligands that are based on the classic PNP framework, explores how the modified ligand influences the behavior of nickel complexes, and describes examples of mostly metal-based reactivity. This chapter mainly focuses on how this new family of PNP ligands can remain unreactive under specific reductive conditions, which allows the stabilization of complexes having unusual oxidation state of the metal center. The third chapter explores how it is possible to induce new types of ligand-based reactivity on the PNP pincer framework by blocking classical modes of metal-ligand cooperation, and how changes in the electronics and steric properties of the complex can lead to switching from metal-based reactivity to ligand-based reactivity.

Ankur Dhar

Supervisor Name: Tsumoru Shintake

Research Unit: Quantum Wave Microscopy Unit

Thesis: Imaging Monopoles in Spin Ice via Electron Holography

Abstract:

Originally proposed by Dirac, magnetic monopoles in vacuum have long remained elusive to detection, but recently emergent monopoles of the microscopic \textbf{H} field have been shown to exist in spin ice. As such, they present a valuable testing ground for the physics of magnetic monopoles which remain elusive as high energy particles. However, signatures of monopoles in spin-ice materials have only been indirect so far, and their direct observation has remained an open challenge since their discovery. One such technique that would make this direct observation a reality is electron holography, due to the electron's high sensitivity to magnetic fields via the Aharonov-Bohm effect. Currently the best holographic microscopes can achieve 3D spatial imaging of spin phenomena with sub-nanometer resolution.

In this thesis, I explore the possibility of imaging monopoles with electron holography through experimental measurements of monopole and spin ice analogs and computational simulation of how a monopole would appear in a pyrochlore spin ice thin film. My experimental work focused on measuring the phase signal from an elongated magnetic needle, as well as artificial spin ice formed from a 2D lattice of nanoscale magnets. My simulated results show for the first time what a monopole in pyrochlore spin ice would look like if imaged using electron holography. The experimental and simulation results together help define the technical requirements for what it would take to achieve direct observation of magnetic monopoles in spin ice via election holography.

Farzana Rahman

Supervisor Name: Kenji Doya

Research Unit: Neural Computation Unit

Thesis: Identifying the Evolutionary Conditions for the Emergence of Alternative Reproductive Tactics in Simulated Robot Colonies

Abstract:

Background and Aim

In many animals, individuals within one sex evolve with different behavior in order to increase their reproductive fitness. These evolved reproductive behaviors are called Alternative Reproductive Tactics (ARTs). Existing theoretical frameworks of Alternative reproductive tactics (mainly modeled by abstract mathematical models or individual based simulation) usually assume that there already exist two or three distinct phenotypes (with different tactic) and make prediction about the expected frequencies of different tactics in different situation. Also in these models agents have low embodiment with environment and lose important aspects of real life features between agents, such as- occupied body space of the agents, sensory information about the environment, physical interaction with other agents etc. By contrast, how distinct types emerge from initial continuous characteristics through evolution is a conceptually different question, which has been ignored by theoretical studies of ARTs, as they usually focus on how already present variations of reproductive behavior are maintained.

In this thesis, I developed a simulated robot evolution framework where agents have high embodiment with environment, along with both intersexual and intrasexual interaction, and investigated under which evolutionary condition alternative reproductive behavior spontaneously emerge from initially monomorphic population.

Materials and Methods

First, I designed and developed a robot simulator incorporating biological features (size, growth, mortality, sex, reproductive cost, etc.) with physical interaction, and a two layer robot controller, where the upper layer is a neural network that makes decision from the environmental input about choosing the behavioral module and the lower layer is four behavioral modules (Foraging, Mating or avoiding, Pushing or following, and waiting). Then I run evolution in this robot simulator where robotic agents can have intersexual and intrasexual interaction and thus evolve through both natural selection and sexual selection from initially monomorphic population.

After ensuring that the agent populations survive and evolve, I investigate the male and female agent behavior to detect any alternative reproductive tactics from the evolved population.

Second, I consider important environmental parameter (Food level, Mating cost, Competition level) and investigate for each parameter what kind of reproductive tactics evolve for each sex and how they are maintained for different level of each parameter. I investigate each parameter and also combination of them.

Then I investigate the evolutionary stability of the evolved phenotype of each sex and demonstrate how evolutionary stable point varies with the change of each environmental parameter.

Results

I have so far obtained the following results.

In the evolved generation, two types of genetically distinct phenotypes with different tactics emerged in males- Bourgeois and sneakers, where they compete to get access to female. The Bourgeois males have high growth rate (that incurs high growth cost), and invest more of their time to look for food to maintain big body size, so that they can get less interruption during mating (win the competition to get access to the female) and increase their fitness (chance of getting offspring). The Sneaker males have low growth rate (low growth cost) and invest most of their time looking for female. As they cannot win competition (get access to female) with the Bourgeois male they flock around the Bourgeois male and when a female comes near they sneak choosing the mating module. They increase their fitness by following and mating with female when bourgeois male invest time to look for food and sneaking female when they come to mate with bourgeois male. But mean fitness (number of offspring per male) of sneaker male is about the same or lower than bourgeois male as bourgeois male is more favored by female and less interrupted during mating.

Females were also observed to evolve two types of genetically distinct phenotype with different tactics, where they mainly differ by offspring number and quality. They are- Quality oriented female (QOF) and Number oriented female (NOF). QOF were big in size, less interested in mating and produced small number of high quality (high size and energy) offsprings, where NOF were small in size, highly interested in mating and produced high number of low quality (low size and energy) offspring.

Further, I observed how different population evolved in different Food level (different number of food resource) and competition Level (male female ratio in the population). I found that Female ART is regulating with food level and Male ART with competition level.

The findings for the variation of the parameters (Food level, competition level and mating cost) and evolutionary stable state of the evolved phenotype will be further discussed by the time of thesis submission.

Hemanta Sarmah

Supervisor Name: Tadashi Yamamoto

Research Unit: Cell Signal Unit

Thesis: Molecular and Physiological function of the mammalian CCR4-NOT subunit CNOT9

Abstract:

The multi-subunit eukaryotic CCR4-NOT complex, imparts gene expression control primarily via messenger RNA (mRNA) decay. Here, we present the role of subunit CNOT9 in target mRNA decay during embryonic development. CNOT9-/- mice appear normal during onset of gastrulation but exhibit growth and differentiation defects accompanied by extensive cell death by embryonic day 9.5. Sox-2 Cre conditional knockouts show marginal rescue and brief delay in phenotype emergence implying defects to be epiblast-dominant. Among various identified targets, we show that Lefty2 mRNA expression is most significantly post-transcriptionally regulated by CNOT9. Lefty2 3’-UTR containing mRNA has higher stability in cells expressing CNOT1-binding-mutant form of CNOT9 relative to cells expressing wild-type CNOT9-expressing cells. CNOT9 primarily localizes within cytoplasm and bridges interactions between the CCR4-NOT complex and miRNA-RISC complex in gastrulating embryos.

Margaret Mars Brisbin

Supervisor Name: Satoshi Mitarai

Research Unit: Marine Biophysics Unit

Thesis: Characterization of Acantharea-Phaeocystis photosymbioses: distribution, abundance, specificity, maintenance and host-control

Abstract:

Microbial eukaryotes (protists) are recognized as important contributors to marine biogeochemistry and play essential roles as both producers and consumers in marine ecosystems. Among protists, mixotrophs—or those thatuse both heterotrophy and autotrophy to satisfy their energy requirements—are especially important to primary production in low-nutrient regions where nutrient availability would otherwise limit primary production if not supplemented by heterotrophy. Acantharian protists (clades E and F) accomplish mixotrophy by hosting Phaeocystis spp. as algal endosymbionts. Despite their ecological importance, acantharians remain understudied due to their structural fragility and inability to survive in culture. In order to overcome these challenges and illuminate key aspects of acantharian biology and ecology—including distribution, abundance, and specificity and specialization of symbioses—single-cell sequencing techniques are necessaryand were used in combination with metabarcoding and high-throughput in situ imaging in this thesis. Major findings from this study were that i) adult acantharian cells (> 50 μm) are extremely rare below the photic zone, despite the relative abundance of acantharian sequences increasing with depth sampled, and that ii) while hosts simultaneously host multiple symbiont species, intra-host symbiont communities do not match environmental communities, demonstrating that symbionts are not systematically digested, and that iii) Phaeocystis cell division genes were significantly downregulated while photosynthesis genes were significantly upregulated in symbiosis. Mixotrophic acantharian sequences recovered fromdeep waters may, therefore, originate from detritus (or possibly reproductive cells), thus corroborating hypotheses that reliance on photosynthesis constrains mixotrophic acantharians to surface waters. In the photic zone, Phaeocystis symbionts serve acantharian hosts by producing organic carbon but do not seem permitted to reproduce, ultimately eliminating the possibility for mutualism. Interestingly, differential gene expression between colonial and solitary free-living Phaeocystis suggest some similarities between colonial and symbiotic cells, namely external stimuli sensing and signal transduction pathways inactivate cell division pathways. It is, therefore, possible that acantharians prevent symbiont overgrowth by manipulating Phaeocystis’ adaptations for colonial living. This thesis contributes new insight into the ecologically relevant photosymbioses between Acantharea and Phaeocystis and further illustrates the benefits of combining single-cell and imaging technologies to illuminate important microbial relationships in marine ecosystems.

Sona Rani Roy

Supervisor Name: Ye Zhang

Research Unit: Bioinspired Soft Matter Unit

Thesis: Design Integrin-targeted Molecular Self-assembling Peptides for Studying Cancer Migration Inhibition

Abstract:

The Yes-associated protein (YAP) is a major oncoprotein responsible for cancer metastasis. YAP’s oncogenic activity is regulated by both the Hippo kinase cascade and uniquely by a mechanical-force-induced actin remodeling process. Here, inspired by outside-in integrin activation via ligand-binding, we developed a molecular self-assembly (MSA) technology that selectively inhibit cancer cell migration by inactivating YAP through physical regulation mediated by actin cytoskeleton. Particularly, we designed a library of integrin-ligand-based peptides that upon binding with integrins, undergo self-assembly to form nanostructures specifically on plasma membrane of cancer cells. After characterizing the size and rigidity of the MSAs and confirming that their assembly does trigger formation of the integrin-talin-vinculin complexes that initiate actin cytoskeleton reorganization and inactivate YAP, we show in multiple cancer cell lines and in xenograft tumor models in mice that the MSAs exert potent, cancer-cell-specific, and dose-dependent anti-migration effects. Finally, we confirmed expected molecular consequences of the MSAs on YAP-related proteins like AMOT and on YAP target genes. Thus, our work illustrates how basic biochemical insights can be exploited as the basis for a nano-biointerface fabrication technology which links nanoscale protein activities at specific sub-cellular locations to both microscale and molecular biological activities to suppress cancer cell migration.

Nishtha Ranawat

Supervisor Name: Ichiro Masai

Research Unit: Developmental Neurobiology Unit

Thesis: Behavioral and transcriptomic analysis of embryonic microglia during colonization and degeneration of zebrafish retina

Abstract:

Microglia are the tissue resident macrophages of the brain that play critical role in establishment and maintenance of neuro-immunological interactions from early embryonic stages to adult life. The developing zebrafish retina provides a precise anatomical and molecularly characterized region of the forebrain ideal for studying migration of embryonic microglia. In this study we explored mechanisms underlying the colonization of developing zebrafish retina by microglia precursors. We also characterized microglia behavior in the context of retinal neurodegeneration. Migration of microglia precursors into retina was mainly studied using live imaging of transgenic fish and mutant fish models. Quantitative image analysis revealed that migration of microglia is a stepwise process depending on retinal blood vessel development followed by retinal neurogenesis. Initially, microglia precursors use emerging blood vessel system surrounding the lens to enter the space near basement membrane of retina. Thereafter, with propagation of waves of neurogenesis, these precursors begin to position themselves into neural retina with clear preferences towards regions that have undergone differentiation over undifferentiated ones. Furthermore, a significant reduction of microglia population is observed in delayed neurogenesis mutant. This suggests a possibility of molecular crosstalk between early born retinal neurons and embryonic microglia, thereby influencing microglia content within the retina. Lastly, with the help of single cell transcriptome analysis, we attempted to characterize microglia at the molecular level mainly focusing on targets/pathways contributing to migration into neural regions.

Collin Stecker

Supervisor Name: Yabing Qi

Research Unit: Energy Materials and Surface Sciences Unit

Thesis: Scanning Tunneling Microscopy and Photoelectron Spectroscopy Studies of Lead Halide Perovskite Surfaces, Defect Dynamics and the MAPbX 3 - CuPc Interface

Abstract:

Over the past decade, lead halide perovskites (PVKs) have emerged as a promising new light absorber material for thin film solar cells. Lab-scale perovskite-based photovoltaic devices have made impressive gains in power conversion efficiency (PCE) and are nearing the same efficiency as silicon-based solar cells. However, perovskite solar cells lack stability, and this is a major obstacle preventing commercialization. The interfaces between the different layers in a device have been implicated as potential areas of charge recombination and material degradation. Understanding the perovskite surface is crucial because it is involved in these interfaces and also because it is the layer that is in first contact with extrinsic species that may cause degradation. Defects in the perovskite material have also been identified as a potential cause of sub-optimal performance. Additionally, some strategies for improving stability have included using mixed halide perovskites, or perovskites containing cesium instead of or mixed with organic cations such as methylammonium (MA). Reports at the device engineering level are plentiful, but fundamental, atomic-scale understanding of the perovskite surface is scarce, especially from an experimental perspective. This thesis uses scanning tunneling microscopy (STM) to examine the perovskite surfaces of CsPbBr3 and mixed halide perovskites MAPbBr3-yIy and MAPbBr3-zClz, the surface defects of MAPbBr3 and their dynamics, as well a device-relevant perovskite/hole transport material (HTM) interface comprised of MAPbX3/CuPc, where X=I or Br. Furthermore, X-ray photoelectron spectroscopy (XPS) is used to characterize the sample composition and stability, and electronic properties are investigated by ultraviolet (UPS) and inverse photoemission spectroscopies (UPS and IPES). Where feasible, these experimental results are corroborated by density functional theory (DFT) calculations performed by collaborators. The goal of this thesis is to provide fundamental insight regarding perovskite surfaces, their defects and their dynamics, and their interfaces with other materials, which may help guide applied research toward creating devices with better performance and stability.

Noa Burshtein

Supervisor Name: Amy Shen

Research Unit: Micro/Bio/Nanofluidics Unit

Thesis: Flow Instabilities and Vortex Dynamics in Intersecting Flows

Abstract:

Inertial and elastic flow instabilities are initiated once critical conditions are met, often resulting in symmetry breaking of the flow field and in the formation of vortices. Predicting and controlling vortex formation and intensity are important for numerous industrial applications (i.e. engineering of bridges, pipelines, airplanes) and in smaller confined environments (i.e. around bends, obstacles and at intersections). However, due to the intermittent character of vortices, it is challenging to create and measure them under uniform and controlled conditions. The aims of this thesis are to experimentally study the effects of confinement and fluid properties on the onset of flow instabilities and vortex dynamics, under controlled and uniform conditions. For this purpose, stationary and stable vortices are formed within microfluidic intersections, imposing different levels of flow confinement as determined by their aspect ratio (α=d/w where d and w are the depth and width of the geometry, respectively). The channels configuration, allow measurements across streamwise planes at high spatial and temporal resolution, using a micro-particle imaging velocimetry (μ-PIV) system. By precisely controlling the Reynolds number (Re= ρUw/η, where U is the average flow velocity, ρ is the density and η is the viscosity of the fluid), symmetry breaking of 4-cells of Dean vortices is initiated at a low-moderate critical value (Rec), resulting in the merging of two co-rotating vortices into a single steady vortex. Similarly, by reducing the Re to Rec* the splitting of a single vortex is induced. The results indicate that the transition type depends on confinement, changing from subcritical to supercritical as α is increased. Merging and splitting of vortices are found to be exponential processes, with a rate that depends on the imposed Re. When imposing Re >> Rec intensification of the steady streamwise vortex occur, until a second critical value (Reus) is met. Above Reus, the flow field becomes unsteady as periodic fluctuations emerge. Additionally, we studied the effect of slight modifications of elasticity on the fluid by the addition of small quantities of flexible polymers. Our results show that polymer additive destabilizes the flow and symmetry breaking occurs at lower Rec. We also find that the stretched polymer’s torque acts counter to the vorticity, reducing vortex intensity. These experiments capture processes that govern turbulent flows and provide important insights into the mechanisms of turbulent drag reduction by polymers, used for reducing energy losses in pipe flows. The conclusions of this thesis are of general importance in fluid dynamics, offering new insights in the field of flow instabilities and vortex dynamics with an unprecedented degree of control.

Han Yan

Supervisor Name: Nic Shannon

Research Unit: Theory of Quantum Matter Unit

Thesis: Fracton States of Matter: From Holography to Frustrated Magnetism

The discipline of modern condensed matter physic has a lot of ambitions: to discover all possible quantum phases of matter, to study the exotic properties and applications of different matter states, and to realize them in experiments. A recent exciting development in this field is the discovery of the fracton states of matter. Featuring immobile excitations and gauged/ungauged subsystem symmetries, it is a phase of quantum many-body systems that transcend the traditional scenarios of Landau-Ginsberg symmetry breaking and topological quantum states.

This thesis is devoted to a few aspects of the fracton states of matter. First, we study a unique property of the fracton models: they mimic the quantum-informational fea- tures of gravity. This can be shown in the context of holographic principle or AdS/CFT duality: a fracton model in AdS space can be shown to satisfy the major properties of holography, the boundary entanglement entropy satisfies Ryu-Takayanagi formula, and the bulk reconstruction follows the Rindler reconstruction. Furthermore, the fracton model in hyperbolic space is known to be similar to various other toy models of holography including holographic tensor-networks and bit-threads model. The intriguing similarity between fracton models and gravity, as well as its implications, are discussed at length.

In the second half of the thesis, we explore possible experimental routes to realize the fracton phases. Here we focus on frustrated magnets on the pyrochlore lattice, one of the most versatile and experimentally fruitful framework to realize spin liquids. By analyzing the symmetry and the coarse-grained limit of the model, we find it possible to realize various versions of rank-2 U(1) gauge theory, in models motivated by real materials.

Tadashi Kozuno

Supervisor Name: Kenji Doya

Research Unit: Neural Computation Unit

Thesis: Efficient and Noise-Tolerant Reinforcement Learning Algorithms via Theoretical Analysis of Gap-Increasing and Softmax Operators

Abstract:

Model-free deep Reinforcement Learning (RL) algorithms, a combination of deep learning and model-free RL algorithms, have attained remarkable successes in solving complex tasks such as video games. However, theoretical analyses and recent empirical results indicate its proneness to various types of value update errors including but not limited to estimation error of updates due to finite samples and function approximation error. Because real-world tasks are inherently complex and stochastic, such errors are inevitable, and thus, the development of error-tolerant RL algorithms are of great importance for applications of RL to real problems. To this end, I propose two error-tolerant algorithms for RL called Conservative Value Iteration (CVI) and Gap-increasing RetrAce for Policy Evaluation (GRAPE).

CVI unifies value-iteration-like single-stage-lookahead algorithms such as soft value iteration, advantage learning and Ψ-learning, all of which are characterized by the use of a gap-increasing operator and/or softmax operator in value updates. We provide detailed theoretical analysis of CVI that not only shows CVI’s advantages but also contributes to the theory of RL in the following two points: First, it elucidates pros and cons of gap-increasing and softmax operators. Second, it provides an actual example in which performance of algorithms with max operator is worse than that of algorithms with soft- max operator demonstrating the limitation of traditional greedy value updates.

GRAPE is a policy evaluation algorithm extending advantage learning (AL) and re- trace, both of which have different advantages: AL is noise-tolerant as shown through our theoretical analysis of CVI, while retrace is efficient in that it is off-policy and allows the control of bias-variance trade-off. Theoretical analysis of GRAPE shows that it enjoys the merits of both algorithms. In experiments, we demonstrate the benefit of GRAPE combined with a variant of trust region policy optimization and its superiority to previous algorithms.

Leonidas Georgiou

Supervisor Name: Bernd Kuhn

Research Unit: Optical Neuroimaging Unit

Thesis: Astrocyte calcium activity mapping in behaving mice using anterograde axo-astrocytic AAV transfer

Abstract:

Astrocytes are more and more considered to be actively involved in information processing in the brain. This idea was primarily established through in vitro experiments. A series of controversies challenged these findings and highlighted the importance of studying astrocytes embedded in their physiological environment that is in awake animals. Astrocytes extend highly ramified processes that form functionally isolated microdomains where they exhibit a rich repertoire of localised calcium signals. How astrocyte microdomain [Ca2+]i signals relate to neuronal activity and behaviour is still unclear. My objective was to investigate circuit specific, single-astrocyte [Ca2+]i microdomain activity in mice during behavioural states and sensory stimuli.

We developed a method for sparse, high contrast labelling of astrocytes embedded in the thalamocortical circuit. We use this technique in combination with two-photon microscopy and genetically encoded calcium indicators (GECIs) in awake mice to study astrocyte [Ca2+]i signals during rest, running, whisker stimulation and thalamocortical axon activity.

Tracking AAV capsids with immunohistochemistry revealed that AAVs injected in the thalamus transfer to astrocytes and neurons in the cortex. This suggests anterograde intercellular transfer of AAVs via thalamocortical axons to cortical astrocytes and neurons. This AAV transfer can be used for the expression of a wide range of genetic tools.

Due to the low probability of AAV transfer we were able to sparsely label astrocytes with membrane tagged GCaPM6f. In combination with two-photon microscopy in awake head fixed mice free to run on a cylindrical treadmill with incorporated whisker stimuli we were able to record single-astrocyte microdomain [Ca2+]i activity. Unbiased analysis of these signals with automated, event based detection software revealed that [Ca2+]i signals in cortical astrocytes exhibit distinct feature changes across behaviour states (run, rest) but not during whisker stimulation.

By exploiting rAAV transfer properties we labelled thalamocortical axons and cortical astrocytes with GECIs. We investigated whether axon bouton calcium signals within the territory of astrocytes correlate with local astrocyte microdomain signals under physiological conditions during different behavioural states. We found no correlation between the calcium activity of boutons and nearby astrocyte microdomains reinforcing our previous observations, suggesting that neurons and astrocytes do not communicate with each other at fast time scales (1.5s).

Finally, we tested if astrocyte microdomain [Ca2+]i signals are random or if astrocytic maps or hotspots exist. Long recordings of single-astrocyte microdomain activity compared to random simulations revealed that astrocytes exhibit regions of higher activity represented as heatmaps. This suggests that astrocyte microdomain [Ca2+]i activity is non-random. Comparing the activity heatmap of the same astrocytes over days revealed that the heatmaps are stable suggesting the existence of astrocyte activity maps in the brain.

Shohei Takaoka

Supervisor Name: Tadashi Yamamoto

Research Unit: Cell Signal Unit

Thesis: Role of 5’ - 3’ exoribonuclease Xrn1 in energy expenditure and control of obesity

Abstract:

Gene expression of eukaryotes is regulated by various processes including mRNA degradation. Dysregulation of mRNA degradation is now considered to be possible cause of various disease including cancer, neurodegenerative disease, diabetes and obesity.

5’ - 3’ exoribonuclease Xrn1 functions at the last step of mRNA degradation. Recent studies showed that Xrn1 forms specific cytoplasmic messenger ribonucleoprotein (mRNP) granules in post-synapse to regulate local translation in neuron. However, there are no physiological studies that have investigated the function of Xrn1 in brain. Therefore, I generated brain-specific Xrn1 knockout mice for the first time and analyzed their phenotype. Interestingly, the knockout mice showed obesity and hyperphagia. Moreover, I found that dysregulated expression of appetite and energy homeostasis related genes in hypothalamus of knockout mice. I concluded that Xrn1 is involved in regulation of energy expenditure and obesity.

Viktoras Lisicovas

Supervisor Name: Keshav Dani

Research Unit: Femtosecond Spectroscopy Unit

Thesis: Improvements in optical techniques to investigate the behavior and neuronal network dynamics over long timescales

Abstract:

Developments in optical technology have produced a rapid shift in experimental neuroscience from electrophysiological neuron observation and stimulation to all optical solutions. The use of light offers distinct advantages to electrophysiological methods: 1) manipulation and observation are minimally invasive, especially when making use of two-photon techniques and near-infrared tissue windows 2) high-throughput capabilities allows gathering a lot more data quicker. Despite advances in optical imaging and stimulation, there are still few studies exploring neuronal activity dynamics at longer timescales (hours to days), where many interesting phenomena take place (i.e., decision making, learning, memory formation). Apart from high costs and expertise required, however, there are still technical challenges that stifle the wide adoption of these techniques. These include temporal resolution, photo-damage, and data processing, among others. In this thesis, I describe the development and implementation of optical solutions for the analysis of behavior, neuronal signaling, and environmental stimuli of C. elegans worms, which improve on the previous state-of-the-art.

I begin by reviewing advances in optical techniques applied in neuroscience for the study of networks of neurons, with emphasis on optogenetics and two-photon imaging. In particular, I focus on work done in C. elegans, which is especially suited for the study of neuronal information processing due to the tight integration of neuronal classes, availability of connectivity maps and ability to capture neuronal activity and behavior simultaneously.

Quantitative analysis of behavior offers an indirect view of the neuronal dynamics. I describe a behavioral tracking setup with the ability to stimulate and record over the course of hours and days. The setup enables timed multicolor stimulation of optogenetic proteins in combination with volatile chemicals throughout the area of the behavioral arena. I demonstrate the usefulness of this system by studying pain perception. Optogenetic activation of a nociceptive ASH neuron with or after an appetitive signal mediated through AIY interneuron results in a lasting inhibition of pain perception. I draw parallels with the gate control theory of pain inhibition.

Information flow in the neural network requires methods capable of simultaneously probing the activity of a group of neurons. I implement a two-photon temporal focusing microscopy setup and show significant improvements through the use of high power/high pulse repetition rate excitation system. I demonstrate that the increase in pulse repetition rate improves signal characteristics without added photobleaching burden. This implementation enables live imaging at high speed for extended periods of time. Live imaging of a subset of aversive sensory neurons expressing calcium indicator reveals that repeated activation of the nociceptive ASH neuron leads to a shift in activation after repeated stimulation. I model temperature increase during a live imaging scenario for different repetition rates at fixed peak intensities and find range centered around 0.5 to 5MHz to be optimal.

In summary, I develop techniques for the interrogation of animal behavior and inter-neuronal activity, over long timescales. These techniques can be implemented with a relatively modest investment and used to address a new dimension of scientific questions.

Stefan Pommer

Supervisor Name: Jeff Wickens

Research Unit: Neurobiology Research Unit

Thesis: The effect of serotonin receptor 5-HT1B on lateral inhibition between spiny projection neurons in the mouse striatum

Abstract:

Striatal spiny projection neurons (SPNs) make inhibitory synaptic connections with each other via collaterals of their main axon, forming a local lateral inhibition network. Previous studies have shown that serotonin, acting via the 5-HT1B receptor, modulates neurotransmitter release from terminals in the target nuclei of SPN projections. Despite this well accepted function, the role of 5-HT1B receptors in lateral inhibition locally among SPNs remains poorly understood. In this thesis, I investigate the role of 5-HT1B in lateral inhibition. To address this issue, whole-cell patch clamp recordings were made from SPNs in acute brain slices, while stimulating presynaptic SPNs expressing Channelrhodopsin. Inhibitory postsynaptic currents (iPSCs) mediated by GABA were measured before and after application of a 5-HT1B receptor agonist. Activation of 5-HT1B receptors significantly reduced the amplitude of iPSCs in SPNs for both, direct and indirect pathways. This was recovered by application of a 5-HT1B antagonist. Further analysis showed the effect was due to a reduced presynaptic release probability of GABA rather than a change in release size. Collectively, these results suggest a prominent role of serotonin in modulating lateral inhibition among striatal neurons and in general striatal function. The 5-HT1b receptor may therefore be a suitable target for future behavioral experiments investigating the currently unknown function of lateral inhibition in the striatum.

Hsieh-Fu Tsai

Supervisor Name: Amy Shen

Research Unit: Micro/Bio/Nanofluidics Unit

Thesis: Glioma on Chips Analysis of glioma cell guidance and interaction in microfluidic-controlled microenvironment enabled by machine learning

Abstract:

In biosystems, chemical and physical fields established by gradients are known to guide cell migration, which is a fundamental phenomenon underlying physiological as well as pathophysiological processes throughout the life cycle such as development, morphogenesis, and wound healing. Cells in the supportive tissue of the brain, glia, are electrically stimulated by the local field potentials from neuronal activities. How the electric field influence glial cells is yet fully understood. Furthermore, the cancer of glia, glioma, is not only the most common type of brain cancer, but the high-grade form of it (glioblastoma) is also aggressive with cells migrating into surrounding tissues (infiltration) and contribute to poor prognosis. In this thesis, I investigate how electric fields and chemical fields can affect the migration of glioblastoma cells in controlled microenvironment.

In part I of the thesis, I describe the engineered microsystems for studying glioblastoma migration and glioblastoma cancer biology. First, the physical and biochemical parameters of cell culture media are characterized for accurate simulation and physical control. Next, versatile microdevice fabrication and treatment methods are developed to create robust but flexible microsystems. Hybrid PMMA/PDMS microfabrication strategy is developed to create microdevices with advantages of streamlined workflow, reversible sealing, bubble-free, low cell consumption, and high experimental throughput. An open-source machine learning software, Usiigaci, is developed for semi-automated instance-aware segmentation, tracking, and data analysis of single cell migration under phase contrast microscopy, enabling high throughput label-free cell migration data analysis.

In part II, I present the studies of glioblastoma cancer biology in biomimetic tumor microenvironments. The glioblastoma cell electrotaxis, which is directional migration under electric field, is dependent of extracellular matrix and voltage gated calcium channels but no interaction with ligand-activated receptor signaling was found. Furthermore, the adhesion of glioblastoma cells to endothelium subjected to both shear flow and electric field is investigated in a quick-fit hybrid microdevice. In addition, the hypothetical role of the electrical stimulation in glioblastoma paracrine signaling is tested by studying the exosomal microRNA expression profiles by microarrays.

In conclusion, robust but flexible microsystems are employed for studying glioblastoma cancer biology with emphasis on cell migration. The microfluidic tools together with machine learning data analysis enable us to elucidate fundamental mechanisms in the field of tumor biology with high experimental throughput.

Dina Mostafa

Supervisor Name: Tadashi Yamamoto

Research Unit: Cell Signal Unit

Thesis: Functional Analysis of CCR4-NOT Complex in Pancreatic β Cell

Abstract:

Regulation of mRNA decay in the cytoplasm is important for proper gene expression and its dysregulation causes various disorders. The carbon catabolite repression 4 (CCR4)–negative on TATA-less (NOT) complex (CCR4-NOT complex), a major deadenylase conserved in eukaryotes, catalyzes mRNA deadenylation which is the rate limiting step in mRNA decay pathway. By virtue of its deadenylation activity, it governs mRNA stability. Loss of subunits of the CCR4-NOT complex results in serious abnormalities in embryonic development and tissue function, suggesting that understanding the biological activities of this complex may pave the way for eventual development of therapeutics. Accordingly, this dissertation aimed at understanding the function of CCR4-NOT complex in pancreatic β cells by generating mice lacking the Cnot3 gene, which encodes an essential CCR4-NOT complex subunit, in β cells. Suppression of CNOT3 in β cells caused β cell dysfunction and diabetes. This was associated with the decreased expression of β cell-specific genes and increased expression of genes specifically repressed in β cells, called “β cell disallowed genes”. By combining whole transcriptome and mass spectrometry analyses and subsequent validations using quantitative real time PCR (qRT-PCR) and Western blotting, I found that mRNA and protein expression patterns were largely different from normal β cells upon CNOT3 suppression, which was clearly relevant to the observed phenotypes. I also found that some β-cell disallowed genes were stabilized in the absence of CNOT3, suggesting that their expression was maintained at low levels under the control of the CCR4-NOT complex. Together, this study uncovered mRNA deadenylation by CCR4-NOT complex as a novel molecular mechanism by which β cell identity and function are regulated. Next, in order to understand how the complex catalytic subunits contribute to the complex function, I used primary mouse embryonic fibroblasts (MEFs) prepared from mice lacking CCR4-NOT complex subunits. We found that maintenance of cell viability is one of the fundamental roles of the CCR4-NOT complex, which is mediated mainly by catalytic activity of the CNOT7/8 subunits. Their vital importance in regulating global mRNA expression is clearly reflected in the results of RNA-seq and bulk poly (A) tail assay. In contrast, CNOT6/6L subunits are dispensable for complex formation and MEFs viability.

Paavo Parmas

Supervisor Name: Kenji Doya

Research Unit: Neural Computation Unit

Thesis: Total stochastic gradient algorithms and applications to model-based reinforcement learning

Abstract:

Optimizing via stochastic gradients is a powerful and flexible technique ubiquitously used in machine learning, reinforcement learning, control, operations research, etc. In many of these applications, the gradients are estimated through a stochastic sampling process, and the learning performance hinges on the accuracy of the estimated gradients. This thesis develops a collection of several novel statistical algorithms to acquire improved gradient estimation accuracy. The need to develop such algorithms was motivated from a model-based reinforcement learning (MBRL) scenario, where I observed that chaotic properties of the dynamics caused the gradient variance to explode when using standard gradient estimation techniques, such as reparameterization gradients. The new techniques sometimes improve the accuracy by 10^6 times and more. The methods rely on both new gradient estimators, as well as clever algorithms to take advantage of the graph structure of the computations to combine estimators in a statistically principled way. While the work started by trying to solve a specific problem related to MBRL, the proposed solutions are general and applicable to any other stochastic computation graph. The problems with chaos have recently been also observed in other tasks, such as meta-learning or protein folding software, and my solutions may prove useful in those domains as well. The main contributions are an 1) MBRL framework called PIPPS, which is similar to the PILCO algorithm, but lifts all of its restrictions by swapping the cumbersome moment-matching computations with a particle sampling approach while achieving the same learning performance with no down-sides, 2) the total propagation algorithm, which is a replacement for backpropagation that prevents the exploding gradient problem by combining gradient estimators in the backwards pass, 3) the probabilistic computation graph framework, which is an intuitive visual method to reason about total gradients on graphs, 4) new policy gradient estimators derived by using the probabilistic computation graph framework, 5) some theoretical discussion about control variates for gradients as well as a unified theory of reparameterization and likelihood ratio gradient estimators. The research has so far lead to two publications (ICML’18 and NeurIPS’18), but also includes yet unpublished work. I hope that this work could lead towards new software frameworks that go beyond backpropagation, and implement more advanced methods for estimating gradients.

James Schloss

Supervisor Name: Thomas Busch

Research Unit: Quantum Systems Unit

Thesis: Massively parallel split-step Fourier techniques for simulating quantum systems on graphics processing units

Abstract:

The split-step Fourier method is a powerful technique for solving partial differential equations and simulating quantum systems of various forms. In this body of work, I focus on several variations of this method to allow for simulations of one, two, and three-dimensional quantum systems, along with several notable methods for controlling these systems.

In particular, I use quantum optimal control and shortcuts to adiabaticity to describe the non-adiabatic generation of superposition states in strongly correlated one-dimensional systems, analyze chaotic vortex trajectories in two dimensions by using rotation and phase imprinting methods, and simulate stable, three-dimensional vortex structures in Bose-Einstein condensates through artificial magnetic fields generated by the evanescent field of an optical nanofiber.

I also discuss algorithmic optimizations for implementing the compressed split-step Fourier method for graphics processing units and multicomponent simulations. All variations present in this work are justified with physical systems where such techniques have been applied and have been incorporated into a state-of-the-art and open-source software suite known as GPUE, which is currently the fastest quantum simulator of its kind.

Mayank Aggarwal

Supervisor Name: Jeff Wickens

Research Unit: Neurobiology Research Unit

Thesis: The Role of Ventral Tegmental Area and Nucleus Accumbens in the Kamin Blocking Effect

Abstract:

The overall aim of the research described in this thesis is to identify neural substrates underlying the Kamin blocking effect. This phenomenon is crucial for understanding of the neural mechanisms of associative learning. Kamin blocking refers to the finding that learning of an association between a cue and an outcome is attenuated when they are paired in the presence of another cue which has previously been conditioned using that outcome. The blocking effect suggests that associative learning is driven by prediction errors, and is not based purely on temporal contiguity between events. In the context of appetitive classical conditioning, recent evidence suggests that the ventral tegmental area and the nucleus accumbens play a role in computing reward prediction error. The current study shows that blocking inhibition in the ventral tegmental area or inactivating the nucleus accumbens neurons during compound cue conditioning attenuates Kamin blocking. Inactivating the nucleus accumbens during single cue conditioning also attenuates Kamin blocking. Taken together, these findings suggest that inhibition in the ventral tegmental area, inhibitory output from the nucleus accumbens, and learning in the nucleus accumbens play crucial roles in the Kamin blocking effect. Previous studies show that dopamine transients track the theoretical reward prediction error during appetitive classical conditioning, and the reduction in the dopamine response evoked by the reward when it is expected has been suggested to play a role in the Kamin blocking effect. In support of this hypothesis, the current study also found that goal tracking rats, in which expected rewards have previously been shown to evoke a robust dopamine response, did not express the Kamin blocking effect. Conversely, sign trackers, in which expected rewards evoke a diminished dopamine response, expressed the blocking effect. These findings are discussed in relation to psychological theory of learning and the possible underlying neural mechanisms.

Bianca Sieveritz

Supervisor Name: Gordon Arbuthnott

Research Unit: Brain Mechanism for Behaviour Unit

Thesis: Ventral motor thalamic projections to prelimbic cortex in cost-benefit decision-making

Abstract:

Innervation patterns of VM projection neurons to cortex have been studied extensively in recent years and differ across cortical areas. Hence, the first aim of this thesis was to clarify the innervation pattern of VM projections to prelimbic cortex and to investigate if they target IT neurons and layer 1 inhibitory interneurons. Using anatomical tracing and immunohistochemistry in light microscopy sections, I confirmed that VM projections in prelimbic cortex target pyramidal neurons, IT neurons and layer 1 inhibitory interneurons. Quantification of contacts by unbiased stereology further indicates that most pyramidal neurons targeted by VM projections are IT neurons. The second aim of this thesis was to determine if ventral motor thalamic projection neurons to prelimbic cortex are involved in cost-benefit decision-making. To address this aim performance of rats on a benefit-benefit, cost-cost and cost-benefit decision-making task was compared between two conditions: With and without administering optogenetic inhibition to ventral motor thalamic axon terminals in prelimbic cortex. Optogenetic inhibition on the cost-benefit decision-making task significantly increased the preference of animals for a high cost/high reward as compared to a low cost/low reward option. On the benefit-benefit decision-making task animals showed a slight increase in preference for a high reward as compared to a low reward. The preference of animals on the cost-cost decision-making task remained unchanged.

Simon Peter Mekhail

Supervisor Name: Síle Nic Chormaic

Research Unit: Light-Matter Interactions for Quantum Technologies Unit

Thesis: Optical Fiber Probes for in-Vivo Neuronal Compressive Microendscopy and Mode Analysis in Nanofibers

Abstract:

Parkinson’s disease is a debilitating and potentially life-threatening disorder which manifests as a malfunction of dopamine producing cells in the basal ganglia of the brain. While the disease was first described over 200 years ago the reason for the cell death remains poorly understood. Advances in functional labelling and stimulation of neurons, which occurred since the turn of the millennium, have allowed researches to image and excite brain tissue by means of completely optical devices. Although this has greatly fostered research in neuroscience there remains a key hurdle in reaching the basal ganglia; these nuclei are deep in the brain. For light to reach the basal ganglia in a controlled way, we propose the use of thin fiber implants for imaging both in- and ex-vivo.

In this thesis we discuss three main experiments. The first involves imaging with a fiber bundle and graded index lens and is split into two parts. We firstly attempt to overcome the problem of low resolution inherent to fiber bundles by using compressive sensing, a data processing technique which allows sub-Nyquist sampling by exploiting sparsity in the data. Secondly, we adapt this method to in-vivo calcium imaging in behaving mice. Our Second experiment addresses methods by which the, relatively large, graded index lens implant might be replaced with a short section of multimode fiber. Where this technique has a smaller foot print for a given field of view the temporal resolution suffers a bit as scanning is required. Finally, the third experiment addresses measurement and excitation of specific modes in few mode fibers by means of a spatial light modulator. Whereas this technique was initially investigated for fiber imaging it was deemed unsuitable for imaging in a behaving mouse due to the bend sensitivity of the fiber. Rather the method was adapted for use in optical nanofibers for studies into cold atom interactions with the evanescent fields of LP01 and LP11 group modes.

Neil Dalphin

Supervisor Name: Bernd Kuhn

Research Unit: Optical Neuroimaging Unit

Thesis: Two-Photon Voltage Imaging of Supragranular Barrel Cortex in Mice: Oscillations and Responses

Abstract:

Layer 1 of cortex is theorized to be responsible for gain control of cortical output or attentional focusing. However, layer 1 is often neglected in research, due to the difficulty of recording from it. Here, I describe two voltage imaging methods which allow recording neural activity, response to sensory stimuli and oscillations, in the supra-granular cortical layers (including layer 1) under anesthetized and awake conditions, and show first results.

I developed a new method to load a newly synthesized, water soluble voltage dye (DiMethyl-ANNINE-6plus) into layer 1. This involved creating a space between the dura and brain, by applying dilute hydrogen peroxide to the open craniotomy, and then injecting the dye into this space. This allowed the dye to diffuse through layer 1 without damaging the brain tissue. Two photon microscopy allowed optical sectioning, so neural activity could be investigated within these upper cortical areas. This protocol is promising for its ease of use in investigating sensory stimuli in layer 1. Additionally all major neural oscillations up to 40Hz can be detected with this method, so we could observe how they change with brain state, and cortical depth.

Alternatively I made recordings in layers 1 and 2, with bulk-loaded voltage dye (ANNINE-6). Voltage responses showed a fast and slow response to whisker stimulation, with a 10-20ms initial response, followed by a prolonged depolarization (peak around 200ms following stimulus), which was larger (~1.03% Δf/f) when the mouse was anesthetized than when it was awake (~0.61% Δf/f). Looking further into voltage responses, gamma frequency oscillations were found through layers 1 and 2, which also showed a power increase following whisker stimulation, but mainly in awake conditions.

I combined ANNINE-6 voltage imaging with GCaMP-6f calcium imaging, to compare the average membrane potential collected by the voltage dye with intracellular calcium responses. Calcium responses showed almost no activity during anesthesia, which indicates a lack of thalamic inputs into layer 1, during anesthesia, but showed a very large response when the mouse was awake. In comparison, voltage changes were stronger in anesthetized than awake animals, although the fast initial response may get larger during wakefulness.

Together these data show fundamental changes in cortical processing from anesthesia to wakefulness, suggesting a thalamic role in cortical integration and local brain oscillations. Additionally the presence of gamma in layer 1 sets an interesting question into its origin.

Sho Kasumie

Supervisor Name: Síle Nic Chormaic

Research Unit: Light-Matter Interactions for Quantum Technologies Unit

Thesis: Hollow Whispering Gallery Mode Resonators: from Fabrication to Application

Abstract:

In this thesis work, a number of optical phenomena, such as nonlinear optics, optomechanics, sensing devices, and microlaser fabrication, have been studied using glass whispering gallery mode resonators (WGMRs). WGMRs are ultra-high Q optical resonators, with a Q-factor of at least 10^7 being easily achieved using a relatively simple fabrication platform. This ensures a strong interaction between light and the material of the resonator. Sensing application using WGMRs are based on detecting dispersion and dissipation changes to the cavity resonances.

We first show that dispersion and dissipation can be distinguished by measuring parameters associated with the signals obtained from the peak cavity ring-up spectroscopy (CRUS). As CRUS happens in a very short (nanosecond) timescale, sensing of transient phenomena is possible by applying the techniques we developed. Another form of light-cavity interaction can be seen via the evanescent field of cavity-coupled light. Two cavities, i.e. a microsphere and a microbubble, interact with each other through their evanescent fields to form a photonic molecule. Theory and experimental results give insight on the evanescent field coupling in the system.

Finally, the interaction between light and the cavity material is studied. Light coupled to a silica WGMR induces parametric oscillation. While the light-matter interaction is strong, the geometrical boundary condition of the cavity limits the parametric oscillation to occur only near the zero-dispersion wavelength (ZDW). This restriction is partially removed if the ZDW is adjustable. In this work, we use a microbubble resonator and we control the thickness of its wall during the fabrication process. This can change the geometrical dispersion of the cavity and enables adjustment of the ZDW. As a result, a four-wave-mixing frequency comb centred around 770 nm is generated. Stimulated Raman scattering is another form of interaction between light and the material. The theory developed as part of this thesis work shows that Raman scattering not only provides a source of light, but also has dynamical behaviour. In the simplest case, two Raman modes excited adjacent to each other are studied theoretically and experimentally, showing that a switching process between two modes is enabled by changing the pump lasser power. Finally, as a simple demonstration, we coat a microbubble with a thin solgel layer doped with erbium ions in order to create a tunable laser. Overall, the work contained in this thesis makes several contributions to the field of research using WGMRs and also impacts areas beyond this design of resonator through some of the phenomena studied and observed.

Jiabao Chen

Supervisor Name: Denis Konstantinov

Research Unit: Quantum Dynamics Unit

Thesis: Coherent control of charged particle systems strongly interacting with microwave photons

Abstract:

The thesis covers three projects about coherent control of particles using light in the strong-coupling regime.

The first project is a theoretical proposal to generate squeezed state and spin-squeezed state of a harmonic oscillators and of an ensembles of two-level-systems respectively, which is strong-coupled to a two-level-system.

I mainly discuss a special case of Jaynes-Cummings model driven by an external field and its analogous in which a two level system is coupled to a collective large spin. By adiabatically turning on a linear driving term on the oscillator or the spin, the eigenstates of the system change character and its ground state evolves into squeezed states of the oscillator or the spin.

The second is the experimental realization strong coupling between the cyclotron motion of a collection of electrons floating on the surface of liquid Helium and the microwave photons in a Fabry-Perot resonator and the detailed analysis using both classical and quantum formalism.

The agreement in classical and quantum formalism demonstrated the mean-value nature of the observed normal mode splitting phenomena and the deviation from the minimal coupled oscillator model because of the inhomogeneity of the cavity microwave field acting on the collection of electrons. In particular, the realistic model reproduced the unexpected observed avoided crossing cyclotron resonance passive mode.

In the third project, I study the surface electrons on helium when a magnetic field parallel to the liquid surface is present.

The tilted magnetic field leads to the coupling between the surface bound state of electron and its motion parallel to the liquid surface.

Such coupling brings significant eigenenergy level shift, and the changes in the structure of eigenstates. For example, we can use this coupling to introduce non-linearity to the cyclotron motion of electrons, and create other interesting quantum states. To support the theoretical study experimental results are presented to compare with the predictions.

Thomas Nieddu

Supervisor Name: Síle Nic Chormaic

Research Unit: Light-Matter Interactions for Quantum Technologies Unit

Thesis: Optical Nanofibers for Multiphoton Processes and Selective Mode Interactions with Rubidium

Abstract:

Optical nanofibers (ONF) have been proved useful tools to probe cold atomic systems. Due to the intense evanescent field at their waist, ONFs have been used to probe or even trap atoms. However, very little experimental work has been done on exploiting the higher order modes (HOM) of such devices. The HOMs feature inhomogeneous polarization distributions around the ONF’s waist and can lead to the guiding of light carrying orbital angular momentum (OAM), via selective excitation of modes. Using modal decomposition at the output of the ONF, a transfer matrix of the atom-nanofiber system can be calculated. This information, when combined with the amplitude extinction resulting from the scattering of the guided light by the cold atomic ensemble surrounding the waist, allows to non-destructively infer the modal excitation at the waist of the ONF. In addition, ONFs feature high field intensities capable of triggering nonlinear optical effects in the surrounding atomic medium, even for input powers lower than the micro-watt. This latter property has allowed us to study fiber-induced Autler-Townes splitting and stimulated emission using a single-color two-photon excitation.

Girish Beedessee

Supervisor Name: Noriyuki Satoh

Research Unit: Marine Genomics Unit

Thesis: Genomic insights on secondary metabolism in symbiotic dinoflagellates

Abstract:

Dinoflagellates are an important group of phytoplankton found in a wide range of environment reflecting a remarkable diversity in form and nutrition styles, along with extensive fossils records due to the formation of robust cysts. They are typically unicellular, photosynthetic, free-swimming and form part of freshwater, brackish and marine phytoplankton communities. Dinoflagellates also produce a wide variety of secondary metabolites including toxins that are dangerous to man, marine animals, fish and other member of food chains. At present, the only available genomes of dinoflagellates are that of family Symbiodiniaceae. Dinoflagellate large nuclear genomes have hampered our understanding of their secondary metabolism evolution. To this end, a genomic survey of the genes needed for the biosynthesis of polyketide secondary metabolites was undertaken in the late-diverging Symbiodiniaceae family. This analysis revealed an extensive diversification of secondary metabolism genes, establishing how the protein architecture guides substrate specificity. To compare if the biosynthetic architecture is conserved, I generated the genome of the early-diverging Amphidinium dinoflagellate and compare it with Symbiodiniaceae genomes. To investigate if secondary metabolite biosynthesis is influenced by nutrient as have been reported by several studies, I performed a transcriptomic analysis of Amphidinium cells under nitrate and phosphate stress. Additionally, Isoseq sequencing provided the first evidence of long single-molecule mRNA of secondary metabolite genes. Overall, I confirm that nutrient stress has no influence on secondary metabolism and that secondary metabolism multifunctional genes are present in the Amphidinium genome.

Yafei Mao

Supervisor Name: Noriyuki Satoh

Research Unit: Marine Genomics Unit

Thesis: Whole-genome sequence analysis of the evolutionary history of the reef-building coral genus Acropora (Scleractinia, Cnidaria)

Abstract:

A major goal of evolutionary biology is to understand how evolutionary processes lead to speciation and diversification, and myriad paths have led to diversification in different evolutionary lineages. Introgression and ancient whole-genome duplication (WGD), paleopolyploid, are regarded as impact evolutionary forces on rapid speciation enhancing diversification recently. As well, ‘ecological opportunity’ is regarded as a prerequisite for evolutionary diversification. Reef-building corals provide the structural basis for one of Earth’s most spectacular and diverse—but increasingly threatened—ecosystems. Modern Indo-Pacific reefs are dominated by species of the staghorn coral genus Acropora (Anthozoa: Acroporidae), one of most diverse genera with close to 150 species, yet the evolutionary and ecological factors associated with their diversification and rise to dominance are unclear.

Hence, I used genomic data of Acropora to investigate its evolutionary history showing what the roles of introgression, WGD, and ecological opportunity play in its diversification and rise to dominance. In the first chapter, I introduce recent studies on Acropora, introgression, and WGD, as well as describe the methods of how to study them. In the second chapter, I analyze the genomes of five Acropora to examine the roles of introgression in the diversification of Acropora. I found evidence for a history marked by a major introgression event as well as recurrent gene flow across species. In addition, I found that introgression genes are evolving faster than others, consistent with a role for introgression in spreading adaptive genetic variations. In the third chapter, I use demographic inference to examine the roles of climate change in the rise to dominance of Acropora. Demographic analysis showed that Acropora lineages profited from climate-driven mass extinctions in the Plio-Pleistocene, indicating that Acropora exploited ecological opportunity opened by a new climatic regime favoring species that could cope with rapid sea-level changes. Moreover, the origin of modern Indo-Pacific Acropora is suspected from polyploidy based on variable chromosome numbers and complicated reticular evolutionary history. And thus, in the fourth chapter, I analyze the five Acropora genomes with an Astreopora genome to investigate whether WGD occurred and the roles of duplicated genes in Acropora. Using comprehensive phylogenomic and dS-based approaches, I analyzed five Acropora genomes as well as an Astreopora (Scleractinia: Acroporidae) genome as an outgroup to show that a WGD event likely occurred around 28 to 36 Million years ago (Mya) in the most recent common ancestor of Acropora. Besides, I found that duplicated genes became highly enriched in gene regulation functions, some of which are involved in stress responses. The different functional clusters of duplicated genes are related to the divergence of gene expression patterns during development. Some gene duplications of proteinaceous toxins were generated by WGD in Acropora compared with other Cnidarian species. Finally, in the fifth chapter, I briefly summarize the results and discuss future study directions in correlation with this thesis. Collectively, this thesis suggests that introgression, climate change, and WGD play impact roles in the evolutionary history of Acropora.

Nilupaer Abudukeyoumu

Supervisor Name: Gordon Arbuthnott

Research Unit: Brain Mechanism for Behaviour Unit

Thesis: Cholinergic interneurons in striatal microcircuit dynamics studied with anatomical and behavioral methods

Abstract:

The thesis will describe two major experiments. One that demonstrated quantitatively that the loss of Colinergic interneurons in the striatum does not result in a loss of cholinergic terminals - indeed they increase for about a month after the damage to the cells. The other examines the consequences of this cell damage on the behaviour of the animals.

Haytham Mohamed

Supervisor Name: Tadashi Yamamoto

Research Unit: Cell Signal Unit

Thesis: Post-transcriptional Regulation of Circadian Rhythm: Involvement of the CCR4-NOT Complex

Abstract:

examine the role of the CCR4-NOT complex in regulating circadian clocks by focusing on CNOT1, the scaffold protein. CNOT1 exhibits a constantly high expression in the mousesuperchiasmatic nucleus (SCN) as well as a rhythmic protein and mRNA pattern in the mouse liver with peak expression at the early morning. CNOT1 haplodeﬁciency in mice results inelongation of circadian period, lower overall activity, and alteration in mRNA and protein expression patterns of various clock genes, mainly Per2. The recruitment of CNOT1 to Per2 mR mediated through Zfp36L1, which itself oscillates in phase with Per2 mRNA. Upon Zfp36L1 knockdown, Per2 is stabilized. Taken together, this suggests that CNOT1 plays a critical rolregulating the mammalian molecular circadian clock and circadian behaviour.

Irina Reshodko

Supervisor Name: Thomas Busch

Research Unit: Quantum Systems Unit

Thesis: State engineering in one-dimensional quantum gases

Abstract:

Understanding, controlling and engineering the quantum states of interacting systems is currently a challenge driven by experimental progress and interest in quantum applications. In this work I consider two specific models of one-dimensional ultracold atomic systems analytically and numerically to determine their eigenstates and accessible dynamical behaviour.

The first part of the work deals with the idea of creating a bosonic atom dispenser, which can deterministically separate any number of atoms from an interacting ultracold gas. By engineering an effectively three-level system with a dark-like state connecting the initial and the target Fock states, I show that such a process is robust and can potentially be experimentally implemented using radio-frequency traps.

In my second project I have derived an analytical single-particle solution for the arbitrary finite Kronig-Penney model. In this model the potential is given by an infinite square well which contains an arbitrary number of arbitrarily positioned point-like barriers of arbitrary heights. I then demonstrated that using certain free parameters of the model as extra (virtual) dimensions we can observe the emergence of higher-dimensional physics in this one-dimensional system.

In particular I studied the appearance of the edge states and the emergence of a Hofstadter butterfly-like momentum spectrum in various configurations of the model. Finally, I used the single-particle solutions to study many-body correlations in a gas mode of either infinitely repulsive bosons or non-interacting fermions.

Kazuto Kawamura

Supervisor Name: Ichiro Maruyama

Research Unit: Information Processing Biology Unit

Thesis: Forward genetic screen for Caenorhabditis elegans mutants with a progressive decline in adult locomotor function

Abstract:

Lifespan and healthspan may not be regulated by the same set of genetic and environmental factors. In some cases, a gene can have disproportionate effects on lifespan and healthspan, suggesting that there are healthspan assurance genes that promote healthspan without largely affecting lifespan. Although most healthspan-related genes have been identified from lifespan studies, carrying out genetic screens that focus on healthspan-related phenotypes may uncover new genetic regulators that specifically affect healthspan. In this study, we used C. elegans to carry out an unbiased forward genetic screen for a healthspan-related phenotype—locomotor function during adulthood. We isolated four mutants that show a shortened locomotor healthspan. In one of the mutants, we found that a nonsense mutation in elpc-2 causes a progressive decline in locomotor function during adulthood with only subtle effects on lifespan. Other C. elegans mutants with mutations in subunits of the Elongator Complex also showed a similar inability to maintain locomotor function in adulthood. The Elongator Complex may be a critical regulator of locomotor healthspan in C. elegans. We show that carrying out genetic screens for healthspan-related phenotypes can provide novel insights into the genetic regulation of healthspan.

Tosif Ahamed

Supervisor Name: Ichiro Maruyama

Research Unit: Information Processing Biology Unit

Thesis: Capturing the nonlinear dynamics of animal behavior with Applications to the Nematode C. elegans

Abstract:

Organisms behave by making complex changes in their shape and posture over time, which can be adapted with remarkable flexibility to suit the environment. This complexity leads to several challenges in measurement and analysis of behavioral data. Existing methods address them by treating behavior as a discrete time process, which is composed of transitions between a finite number of stereotyped motifs, such as walking or reaching. In addition to not addressing its continuity, this viewpoint also ignores the fact that most behavior is not stereotyped. Indeed, even the continuous postural dynamics that make up a particular motif have unpredictable fluctuations, making variability as fundamental a characteristic of behavior as stereotypy. There is, therefore, a need for a perspective that captures the continuous complexity of animal behavior, and offers detailed insights into general principles underlying its generation and continuous control.

In my Ph.D. thesis, I propose a new approach of analyzing behavior, based on the idea that it is fundamentally a continuous-time spatiotemporal dynamical system. Recognizing that the instantaneous posture of an organism is insufficient to distinguish between behaviors, I define its behavioral state as a short temporal sequence of its postures. Stacking the posture sequences over time results in the state matrix X, on which I perform singular value decomposition to obtain a low dimensional "behavioral state space". As an organism moves, the corresponding behavioral state traces out a continuous trajectory in the state space, such that the geometry and topology of the trajectories encode quantitative and qualitative properties of behavior. I characterize an organism's behavioral dynamics using topological invariants calculated from estimates of local Jacobians of the state space trajectories. The invariants capture essential aspects of behavioral dynamics, such as the number of degrees of freedom, symmetries in the governing equations of motion, and measures of predictability and variability. Importantly, they are coordinate-independent, and thus do not depend on how behavioral data is collected and represented.

I apply the above approach to the nematode C. elegans and show that the dynamics of freely moving worms lie on an attractor embedded in a 6D state space. The attractor is globally composed of three sets of cyclic trajectories that form the animal's basic stereotyped motifs: forward, backward, and turning locomotion. In contrast to this global stereotypy, I find variability at small scales where the trajectories are locally unpredictable. This unpredictability shows signs of deterministic chaos in the form of exponential divergence of neighboring trajectories. I use estimated local Jacobians to calculate the entire spectrum of six Lyapunov exponents, which are invariants that measure the rates at which volumes defined by bundles of neighboring trajectories grow or shrink along different directions in state space. Surprisingly, I find a symmetric Lyapunov spectrum with two positive exponents generating variability, balanced by negative exponents driving stereotypy. Symmetric Lyapunov spectra are a hallmark of systems derived from a Hamiltonian by the addition of viscous damping, and I hypothesize that these dynamics provide a way to implement an efficient yet flexible control strategy for the worm. I use the findings above to propose a simple linear control law, which is tested on data from escape response experiments in the worm.

In summary, by treating behavior as a continuous time dynamical system, the state space approach aims to capture all relevant movement dynamics irrespective of whether they are stereotyped or variable. Moreover, it offers a detailed, coordinate-independent characterization of how an organism generates and controls its behavior. While my focus in this thesis is on the posture dynamics of C. elegans, I expect the approach outlined above to be generally applicable to other organisms.

Masakazu Igarashi

Supervisor Name: Jeffery Wickens

Research Unit: Neurobiology Research Unit

Thesis: The role of interhemispheric cortico-cortical connections in bimanual coordination in the rat

Abstract:

The overall aim of the thesis research is to investigate the role of interhemispheric cortico-cortical connections in bimanual coordination in the rat. The coordinated use of two hands is crucial for the basic needs of daily life, such as gathering and handling food, and also for human creative achievements such as playing musical instruments. Such coordinated motor skill is highly developed in primates, where it has been most extensively studied. Rodents also dexterously coordinate forelimbs when feeding. Rats and mice demonstrate repertoires of paw usage during handling of hood item. However, rodents have been less commonly used in the study of bimanual coordination because of limited quantitative measuring techniques. In the thesis research a high-resolution kinematic tracking system for bimanual coordination in rats was developed. The system was used to quantify bimanual coordination in head-fixed rats during food handling and consumption, which enabled detection of movement symmetry and asymmetry in bimanual forelimb use. Automatic segmentation and classification algorithm applied to identify three categories of movements: unimanual, symmetric bimanual, and asymmetric bimanual. Quantification of the classified movements revealed that the feeding behavior is dominated by symmetric bimanual movements with less prevalent asymmetric bimanual and unimanual movement. The role of interhemispheric cortical connections in bimanual coordination was then investigated. Tracing connections revealed that commissures from cortical forelimb motor areas run through the anterior part of corpus callosum. The anterior corpus callosum was pharmacologically blocked during food handling behavior, and the effect on forelimb kinematics was recorded. An automatic classification method developed as part of the thesis research revealed that the frequency of symmetric bimanual movements was reduced by the anterior corpus callosum inhibition. In contrast, asymmetric bimanual movements were increased. Other parameters such as global scale of motor skills, such as mean food drop rate, and mean consumption time remained unchanged. Taken together, these results suggest that the symmetric dominance in bimanual movements in rodents is modulated by cortico-cortical connections via the anterior corpus callosum.

Hiroaki Hamada

Supervisor Name: Kenji Doya

Research Unit: Neural Computation Unit

Thesis: Serotonergic Control of Brain-Wide Dynamics

Abstract:

In this thesis, I performed two experiments in order to understand the serotonergic regulation of brain-wide dynamics. First, I studied short-term and long-term influences of a serotonergic antidepressant on the brain dynamics with functional resonance imaging for rodents. I found that a serotonergic antidepressant, escitalopram, reshapes cortico-limbic functional architecture along time-course of medication. Additionally, long-term serotonergic antidepressant promotes spontaneous behaviors but influence on anxiety-like behaviors showed rather context-dependence and higher individuality. The results imply long-term serotonergic antidepressant enhances intrinsic motivation but not anxiety.

Second, I conducted a pilot experiment to access to serotonergic modulation of brain dynamics by optogenetic stimulation. I found that optogenetic stimulation of serotonin neurons in the dorsal raphe nucleus (DRN) induces brain responses in the frontal cortical regions (the anterior cingulate cortex, the medial prefrontal cortex, and the insular cortex), the striatum, and the ventral tegmental area (VTA). From these experiments, this thesis delineates how serotonin system regulate brain-wide dynamics in short and long time scales.

Jessica Verena Schulze

Supervisor Name: Kenji Doya

Research Unit: Neural Computation Unit

Thesis: Spatial and Modular Regularization in Effective Connectivity Inference from Neural Activity Data

Abstract:

Previous studies of effective connectivity inference from neural activity data applied simple regularization approaches like L1-norm regularization for sparseness.

In this thesis I investigate the incorporation of other bio-physiologically plausible priors over spatial and modular organization of neural circuits to facilitate network inference from neural activity data of large populations of neurons.

First, I incorporate distance-dependent connectivity into known techniques that impose sparseness via L1/L2 regularization in a linear non-linear poisson (LNP) cascade model. We start with simple maximum likelihood estimation with gradient decent and Newton’s method and then extend the idea of distance-based regularization to a hierarchical maximum a posteriori estimation approach with Metropolis-Hastings sampling. In this approach a prior over the existence of neural connections based on euclidian distance between neurons is included into the regularization term.

Second, I formulate a modularity prior which clusters neurons into modules and applies different model parameters for connections within each module and between modules. Then I derive a Gibbs sampling algorithm that dynamically re-assigns module membership and estimates a weight matrix.

To validate these methods, I apply them to simulated data sets with known ground truth connections weights for from up to 400 neurons. I then compare the results to previously proposed methods like model-free and simple generalized linear models (GLM) with sparse regularization.

Furthermore, I apply these methods to an extensive experimental data set of optical calcium recordings of the Posterior Parietal Cortex (PPC) and the Posteromedial Cortex (PM) of the mouse brain including populations of up to 600 neurons.

Patricia Himeka Wepfer

Supervisor Name: Satoshi Mitarai

Research Unit: Marine Biophysics Unit

Thesis: Spatial genetic structure in the coral genus Galaxea (Euphyllidae) and their associated Symbiodiniaceae communities

Abstract:

The evolution and systematics of corals have been difficult to unravel despite being the fundament of one of the world's most charismatic ecosystems. Coral diversity and diversification processes are not well understood due to morphological plasticity, potential hybridization and generally high rates of dispersal. Both spatially and molecularly extensive studies are needed to improve our understanding of coral ecology and evolution, including spatial biodiversity processes in the host and their associated symbionts. This dissertation investigates coral evolution in three complementary studies using the genus Galaxea L. as a model. First, I asked whether endosymbiotic community composition and water depth differentiate morphologically cryptic genetic lineages in G. fascicularis. The Symbiodinium ITS2-sequence was metabarcoded using next generation sequencing (NGS) and community assembly was analyzed with joint distribution models. Symbiodinium communities were found to cluster into three regular community types, which were unrelated to host lineage but could partially be explained by host polyp size and water depth. A strong component for community assembly remained residual and may be due to potential species interactions between Symbiodinium strains. Some indications for vertical segregation exist in Thailand but could not be confirmed in Japan. Second, I assessed how spatial connectivity between geographic populations corresponds to neutral differentiation on the subspecies level using population genomic methods. Coral populations were sampled from several locations along the Ryukyu archipelago, the Daito islands, and the Ogasawara Islands, and were characterized by restriction site-associated DNA sequencing (RAD, single digestion EcoR1). High gene flow could be confirmed in the North-South direction along the Ryukyu archipelago, but was small to nonexistent in East-West direction between Okinawa, Daito and Ogasawara. This was consistent with dispersal pathways as predicted by ocean currents. Lastly, I investigated the evolutionary history in the genus Galaxea taking a phylogeographic approach using RAD. I asked whether the genetically well-differentiated and sympatric lineages in Okinawa maintain their separation over geographic space and to what extent their spatial distributions overlap in the genus distribution range. Galaxea field collections were gathered from across the Indo-Pacific, including from the Red Sea, Maldives, Chagos, Western Australia, Thailand, Japan, Hong Kong, eastern Australia, Guam and Samoa, and were complemented by museum specimens to increase geographical coverage. At the same time the relationship between genetic lineages and taxonomic species was evaluated based on five out of seven currently accepted species (G. fascicularis, G. astreata, G. cryptoramosa, G. paucisepta, G. horrescens) to discuss their species status from a phylogenetic perspective. Overall this study indicates spatial rather than ecological or symbiosis-related processes to drive diversification and that the current taxonomy does not reflect biological species in this genus.

Ray Xin Lee

Supervisor Name: Bernd Kuhn

Research Unit: Optical Neuroimaging Unit

Thesis: Nature and source of animal spontaneous behaviors: Insights from psychobehavioral development and neuronal population dynamics in mice

Abstract:

Awake animals switch between different behavioral states irregularly even in a homogenous and steady environment, especially obvious outside from any behavioral task when they are free to voluntarily behave. Complex dynamics of spontaneous behaviors show structured patterns but do not reflect random noise from environmental fluctuations. These irregular but structured patterns have been taken as a representation of internal states such as emotion, which is assumed to represent the underlying background brain activity and its dynamics. To date, the nature and source of animal spontaneous behaviors remain as a major conceptual challenge to academia, due to the lack of studies which systematically and quantitatively examine this fundamental process. To achieve insights about self-initiation of animal spontaneous behaviors, the research was conducted in two directions: (i) To interpret previously challenging and inconclusive behavioral development by re-evaluating animal spontaneous behaviors that represent emotionality, centered on the study of PTSD (post-traumatic stress disorder)-like internal psychological development in laboratory mice; (ii) To identify and examine collective properties of neuronal populations in the cerebral cortex that determine dynamical structures of behaviors and behavioral state transitions, centered on the study of pathological development of chronic neuropathic pain in laboratory mice. The results from these studies demonstrate the first system-level view of experimentally disentangled components, processes, and determinants of animal self-initiated behaviors and relevant internal developments, explaining the nature and source of animal spontaneous behaviors.

Jui-Yin Lin

Supervisor Name: Denis Konstantinov

Research Unit: Quantum Dynamics Unit

Thesis: Transport properties of strongly correlated 2D electrons confined in microchannels

Abstract:

Optical near-fields are generated when light passes through components with Wigner crystal is the solid phase of strongly correlated electrons. The main theme of this thesis is employing the two-dimensional Wigner solid (WS) on liquid helium in the regime of strong coupling with medium excitations to probe the effects of dynamic interaction of WS with varied external potentials imposed by microscopic structures.

We first characterize the transport properties of a homogeneous WS, a WS island and an inhomogeneous WS in a microchannel geometry (PRB 94, 195311, 2016). A further study of the WS size demonstrates how the energy dissipates from the edges of WS, which thus affects the breaking of the strong coupling with liquid helium surface excitations, i.e. ripplons.

Then, by introducing an external spatial periodic potential, we observe the suppression of WS-ripplon coupling when the commensurability of periodic potential and electron spacing is 4, i.e. four particles within a period of external potential. This finding could help to understand the Frenkel-Kontorova model of 4:1 commensurate state.

Finally, some interesting phenomena of this strongly-correlated electron flow has been studied in a T-shape channel geometry. It shows a breakdown of Drude model when the electron-electron interaction cannot be neglected.

Kalale Chola

Supervisor Name: Tsumoru Shintake

Research Unit: Quantum Wave Microscopy Unit

Thesis: Development of an SPH variant of implicit LES for studying wave energy transport

Abstract:

The main goal of this project was to develop a fluid flow solver based on the smoothed particle hydrodynamics, SPH method for studying the energy transport efficiency of breaking waves. Since SPH is a relatively new approach in computational fluid mechanics, important aspects of numerical methods including; consistency, stability, accuracy and conservation are not well understood. By developing SPH from a Lagrangian, conservation properties are now well understood. Therefore, the first aim was to address the problem of consistency.

Secondly, standard SPH models suffer from spurious high frequency numerical noise in the pressure field. One adopted approach led to the development of a $\delta-$SPH model where artificial diffusion terms are added into the continuity equation; violating the conservation of mass. Motivated by this approach, my second aim was derive a pressure equation from the energy conservation law to address this problem.

Thirdly, implementation of boundary conditions for complicated geometries remains a problem for standard SPH models. The most plausible approach is to simulate a solid boundary using boundary particles that exert repulsive forces on the fluid. Several models have been developed, with varying levels of effectiveness. Motivated by the need for a simple boundary force model that can be derived rigorously, a new boundary model has been proposed.

Finally, the proposed model is called implicit SPH or iSPH meaning that turbulence is not explicitly modeled, it is rather implicitly modeled. All Standard SPH models on the other hand require explicit turbulence modeling, a very difficult task. In particular standard SPH models follow explicit LES whereas the iSPH models follows the implicit LES or ILES. Most importantly, as SPH is based on a zeroth order deconvolution approximation our proposed iSPH model is an order n approximate deconvolution model.

The iSPH model was applied to various free surface flow benchmark problems. The results obtained where in good agreement with both theoretical and experimental predictions.

Tsai-Ming Lu

Supervisor Name: Noriyuki Satoh

Research Unit: Marine Genomics Unit

Thesis: Comparative genomic studies on Dicyema japonicum: the phylogenetic position of dicyemids and the genomic adaptations to parasitic lifestyle

Abstract:

Parasitism has independently occurred more than 200 times across 15 animal phyla, yet remains a topic of debate that how free-living ancestors evolved to parasitic organisms. Dicyemid mesozoans are microscopic endoparasites inhabiting the renal sacs of some cephalopods. They possess simplified body organization without differentiated organs and have long fascinated biologists because of their incompletely known lifecycles. Obtaining genomic data from enigmatic parasites would be essential to better comprehend parasitism evolution. Here we decoded the genome of Dicyema japonicum which is approximately 68 Mbp with an extraordinarily shortened intron size of 38.2 bp on average. Comparing among bilaterians, D. japonicum retained fewer genes in most KEGG pathways, and four parasite species from different phyla showed a convergent gene number reduction in the metabolism pathways. In contrast, D. japonicum exhibited multi-copy gene clusters associated with endocytosis and membrane trafficking, perhaps reflecting its specialized nutrient-uptake strategy. Up-regulated transcripts at dispersal larvae stage were over-represented on gene ontology terms of motor activity and response to the stimulus. Taken together with immunostaining signals of neurotransmitters and neuropeptides occurring on apical cells, dicyemids may have potential mechanisms to sense environmental cues and could actively approach new hosts. In summary, the dicyemid genome provides a resource to uncover mysterious lifecycle of dicyemids, as well as studying comparative genomics to gain insights into the parasitism evolution. Furthermore, genomes of parasites may adapt through eliminating genes which are not necessary for parasitic lifestyle or increasing gene copies corresponding to lineage-specific biological processes.

E Laine Wong

Supervisor Name: Keshav Dani

Research Unit: Femtosecond Spectroscopy Unit

Thesis: Ultrafast spatiotemporal control of photocarriers in doped semiconductors

Abstract:

Control of the spatial and temporal dynamics of photoexcited charge carriers at the surfaces and interfaces of semiconducting materials is pertinent to many of modern technologies such as solar cells, photodetectors and other optoelectronic devices. In inhomogeneous materials such as nanostructured materials, spatial variations in carrier dynamics are inherent via the material design and accordingly allow for improvements in device performance e.g. transport of electrons from donor to acceptor regions in photovoltaic devices. On the other hand, one can also create spatial inhomogeneity in the carrier dynamics in homogeneous systems by using a non-uniform photoexcitation profile. This would have the advantage that the control of the carrier dynamics could be more flexible and not tied down to the fixed material design. In principle, photoexcitation even with a simple Gaussian beam could result in nontrivial inhomogeneous carrier dynamics within the full-width-half-maximum (FWHM) of the optical spot. However, in typical ultrafast spectroscopy measurements performed thus far, one averages out the photocarrier response over the excitation spot and spatial variations therein are inaccessible.

In this thesis, we create nontrivial spatiotemporal dynamics of the photoexcited electrons in a homogeneous Zn-doped GaAs using the spatial intensity variation in a simple Gaussian photoexcitation beam. We image these dynamics using time-resolved photoemission electron microscopy (TR-PEEM) - a technique offering both high temporal and spatial resolutions. In particular, we demonstrate the spatial redistribution of the photoexcited electrons in two different regimes: (I) the early time delays where we control the vertical transport of electrons to the sample surface and (II) at long time delays where we manipulate the lateral distribution of the electrons along the sample surface.

In the first study, we achieve spatial variations in the screening process of the intrinsic surface field, which influences the vertical drift of photoexcited electrons from the bulk to the sample surface. Combined with the occurrence of Auger recombination in regions of higher intensity, we see a depletion in the electron population at the center, with a gain just away from the center. We show control of these processes and how they affect the electron distribution on the sample surface in nontrivial ways.

In the second study, we show that at long time delays the spatially varying screening process leads to the creation of lateral fields along the sample surface. These highly local and spatially varying lateral fields then act upon the photoexcited electrons, eventually pulling them apart into two distinct distributions. Using a simple model that explains the experimental data, we also show the possibility of generating near-arbitrary lateral fields and thus controlling electrons on the sample surface in more general ways.

In conclusion, this thesis demonstrates the capability to create and control nontrivial spatiotemporal dynamics of the photoexcited electrons even in a homogeneous semiconductor by exploiting the intensity variation of an ultrafast light pulse. This capability could lead to a promising new handle for use in high-speed optoelectronic devices.

Chris Reinke

Supervisor Name: Kenji Doya

Research Unit: Neural Computation Unit

Thesis: The Gamma-Ensemble - Adaptive Reinforcement Learning via Modular Discounting

Abstract:

Reinforcement learning allows artificial agents to learn complex tasks, such as playing Go on an expert level. Still, unlike humans, artificial agents lack the ability to adapt learned behavior to task changes, or to new objectives, such as to capture as many opponent pieces within a given number of moves, instead of simply winning. The Independent Gamma-Ensemble (IGE), a new brain-inspired framework, allows such adaptations. It is composed of several Q-learning modules, each with a different discount factor. The off-policy nature of Q-learning allows modules to learn several policies in parallel, each representing a different solution for the payoff between a high reward sum and the time to gain it. The IGE adapts to new task conditions by switching between its policies (transfer learning). It can also decode the expected reward sum and the required time for each policy, allowing it to immediately select the most appropriate policy for a new task objective (zero-shot learning). Additionally, this allows to optimize the average reward in discrete MDPs where non-zero reward is only given in goal states. The convergence to the optimal policy can be proven for such MDPs. The modular structure behind the IGE can be combined with many reinforcement learning algorithms and applied to various tasks, allowing to improve the adaptive abilities of artificial agents in general.

Dongrong Zhang

Supervisor Name: Gustavo Gioia

Research Unit: Continuum Physics Unit

Thesis: Spectral theory of turbulent flows

Abstract:

Spectral theory of mean-velocity profile in canonical wall-bounded turbulent flows; specifically, pipe flow, channel flow, and plane-Couette flow. Generalized Monin-Obukhov similarity theory for thermally-stratified flows and spectral theory of thermally-stratified plane-Couette flows.

Lashmi Piriya Ananda Babu

Supervisor Name: Tomoyuki Takahashi

Research Unit: Cellular & Molecular Synaptic Function Unit

Thesis: Functional roles of microtubules in a giant presynaptic terminal

Abstract:

Roles of microtubules (MTs) in cell division and organelle transport are well established. In neurons, axonal MTs are extended to terminals and transport mitochondria and various molecules in vesicles. However, it is unknown whether MTs play functional roles in presynaptic terminals. To address this, I first identified MTs at the calyx of Held giant terminal, using immunofluorescence, electron-microscopic and live-staining methods, and found that MTS were co-localized with synaptic vesicles (SVs). I then disrupted MTs using vinblastine and tested its effect on synaptic efficacy and properties, using whole-cell patch-clamp recording methods at the calyx of Held. Acute MT disruption caused (i) a prolongation of recovery time after short-term depression and (ii) a desynchronization of quantal release, but had no effect on the basal synaptic transmission efficacy or the magnitude or time course of SV exo-endocytosis, measured using presynaptic membrane capacitance measurements. These results suggest that presynaptic MTs contribute to SV transport for release site replenishment, and in parallel, promote synchronous transmitter release for high-precision neurotransmission at this fast synapse.

Daisuke Takahashi

Supervisor Name: Hirotaka Sugawara

Research Unit: Advanced Medical Instrumentation Unit

Thesis: Minimal Gauged U(1) Extension of the Standard Model with Classical Scale Invariance and Phenomenology

Abstract:

Although the Standard Model (SM) is the best theory in describing phenomena among elementary particles, it suffers from several problems, such as the gauge hierarchy problem, origin of the electroweak symmetry breaking, non-zero neutrino mass, and no candidate of dark matter. For solving these problems, we consider minimal U(1) extension of the SM with the classically conformal invariance, where an anomaly-free U(1) gauge symmetry is introduced along with a U(1) Higgs field and three right-handed neutrinos (RHNs) for the seesaw mechanism generating neutrino masses. With no mass term allowed by the classically conformal invariance, the U(1) gauge symmetry is broken through the Coleman-Weinberg mechanism, which subsequently triggers the electroweak symmetry breaking. We perform parameter scan and identify regions resolving the SM Higgs vacuum instability, while satisfying the LHC Run-2 bound on the U(1) gauge boson production and the naturalness constraint. We also investigate cosmological aspects of the model. Introducing Z2 parity, one RHN being an unique parity-odd particle in the model serves as dark matter. A successful inflation scenario is possible by identifying the U(1) Higgs boson with inflaton and introducing its non-minimal gravitational coupling. Interestingly, the LHC physics and cosmological observations are complementary narrowing down the model parameter space.

Keita Ikegami

Supervisor Name: Noriyuki Satoh

Research Unit: Marine Genomics Unit

Thesis: Comparative transcriptome analysis of basal deuterostomes and its implications for the phylotypic stage

Abstract:

The modification of embryogenesis in animals is an interesting research subject in evolutionary developmental biology. The phylotypic stage was originally described as a developmental period shared across vertebrates, where embryos resemble each other, andis represented by a narrowing waist of the developmental “hourglass” model. Recent studies of comprehensive gene expression have found support for this model in other metazoans, plants, and fungus. However, the existence of the phylotypic stage remains unresolved in deuterostomes. Thus, this study examines whether the phylotypic stage can be found across four deuterostome taxa, namely Ciona intestinalis from the Phylum Urochordata, Branchiostoma floridae from the Phylum Cephalochordata, Ptychodera flava from the Phylum Hemichordata, and Acanthaster planci from the Phylum Echinodermata. Comparison of gene expression profiles was carried out by microarray analysis of RNA across predetermined developmental time points. In this thesis, I developed a method of microarray probe design in A. planci, and proposed guidelines for the technology for use with other marine invertebrates. Next, I compared the gene expression profile with this microarray method. I found that the overall gene expression somewhat resembled each other in A. planci, P. flava and B. floridae, while C. intestinalis exhibited a unique pattern. The gene expression profiles of A. planci, P. flava and B. floridae showed narrowing waist-like pattern from the blastula to gastrula stages up to early larval stages; these stages were more conserved during embryogenesis in deuterostome taxa. However, I failed to find evidence for a typical vertebrate-like phylotypic stage in the four deuterostome taxa.

Nino Espinas

Supervisor Name: Hidetoshi Saze

Research Unit: Plant Epigenetics Unit

Thesis: rCBP-dependent regulation in rice innate immunity

Abstract:

Plant innate immunity against bacterial attacks is a two-tiered inducible system capable of defense responses at local and systemic areas.These systems are the PTI and ETI. During infection, PTI has the ability to recognize microbial signatures upon bacterial contact, while ETI recognizes microbial protein secretions called effectors delivered inside the cell.The activation of PTI and ETI confers systemic tissues of infected plants a broad-spectrum immunity against later pathogen attacks termed systemic acquired resistance (SAR). Defense priming is an adaptive component of SAR that regulates the molecular storage of defense memory for a more effective defense response.

The main aim of this work is finding a novel molecular defense signaling pathway that is controlled by acetylation at the infected (local defense) and systemic tissues (priming defense).

To investigate the role of histone acetyltransferase-dependent pathway in plant immunity, I have isolated transgenic and mutant lines of rCBP, [rice Cyclic adenosine monophosphate response element-binding protein (CREB) Binding Protein], under Nipponbare cultivar background using RNAi silencing and gRNA/Cas9-mediated genome editing. Animal CBP was initially described as both transcriptional coactivator and histone acetyltransferase. The rCBP-RNAi lines with mistargeting of the other members of CBP family are characterized by massive sterility and impairment of the number of effective grains. On the other hand, the CRISPR/Cas9 mutant lines have wild-type number of effective grains.

To profile the global acetylation of histone lysine-sites via rCBP, I performed mass spectrometry-based proteomics in data dependent acquisition (DDA) and parallel reaction monitoring (PRM) modes. My results showed that H3 lysine sites are possibly targeted by rCBP with very high acetylation specificity on H3K9.

To implicate the role of rCBP in rice innate immunity, I conducted a pathogenesis assay with bacterial pathogen, Pseudomonas syringiae pv. oryzae (Pso). Pathogenesisassay showed that rCBP-/-mutants are resistant to Pso infection compared to segregated wild-type control.

I also performed transcriptome analysis on local and systemic tissues with Pso to investigate the genome-wide effects of rCBP mutation and to identify factors with roles in both basal and systemic immune response. As a result, I have identified seven putative rCBP-dependent transcriptional repressors that possibly explain the resistance phenotype of rCBP mutant lines.

Overall, these data preliminary indicate that rCBP is both a positive regulator of developmental processes and a negative regulator of rice immunity. These data also suggest that rCBP may execute this dual regulatory function either through H3K9ac and/or co-transcriptional activity on target gene loci.

Cong Liu

Supervisor Name: Evan Economo

Research Unit: Biodiversity and Biocomplexity Unit

Thesis: Understanding the ecological and evolutionary processes shaping ant biodiversity across spatiotemporal scales

Abstract:

Ecological and historical evolutionary processes together generate biodiversity patterns across geographies and across the tree of life. However, understanding the relative importance of, and the interplay between ecological mechanisms and evolutionary processes in shaping biodiversity patterns is still a challenge due to the different spatiotemporal scales on which they are operating. Therefore, a fundamental goal of biodiversity research is to use different research approaches to investigate how the ecological processes (dispersal, competition, and environmental filtering), as well as long-term evolutionary processes (adaptation and speciation), contribute to the community assembly and biodiversity patterns. In this thesis, I investigate ant biodiversity patterns and underlying eco-evolutionary processes across multiple systems (tropical agroecosystem, complex mountainous landscape, and Pacific archipelago), providing a comparative framework to understand the eco-evolutionary processes driving biodiversity patterns. I first present results from an ant biodiversity survey in Xishuangbanna, Yunnan, China, where I found 213 species/morphospecies of ants from 10 subfamilies and 61 genera. Forty species represent new records for Yunnan province and 17 species are newly recorded for China. In addition, I describe one new species, Aenictus yangi. When examining the changes in taxonomic, functional, and phylogenetic ant biodiversity after conversion to rubber plantation, I found a sharp decline of species richness in rubber plantation with lower than expected taxonomic and functional beta diversity. This suggested a strong environmental filtering driving ant biodiversity in the rubber plantation. I then investigate the variation in taxonomic and phylogenetic ant diversity patterns along a geographic transect spanning 5000m in elevational range in the Hengduan mountains, where environmental gradients and spatial connectivity are intertwined as a complex process that might shape biodiversity patterns. I found that environmental gradients dominate variation in both alpha and beta diversity in this landscape, with alpha diversity strongly declining with elevation and beta diversity driven by elevational differences. Finally, I apply a comparative phylogeographic framework to examine the evolution of the hyperdiverse ant genus Strumigenys in Fiji archipelago using RAD sequencing. My results revealed the history of Strumigenys species that colonized to Fiji archipelagoes in Miocene (10.5-7.5 Ma), following by two independent radiations across the whole archipelago, leading to the emergence of 11 endemic species. The population structure and demographic history of each endemic species consistently support the idea of deterministic macroevolutionary processes that drive the diversity dynamic of ants in Fiji archipelago. Together, this study highlights the need for a pluralistic framework that integrates different approaches to understanding the eco-evolutionary drivers of biodiversity patterns across scales.

Caroline Starzynski

Supervisor Name: Mitsuhiro Yanagida

Research Unit: G0 Cell Unit

Thesis: Investigating ancient metabolic reactions contributing to G0 quiescence survival in fission yeast S. pombe

Abstract:

Since all living organism rely on assimilating environmental nitrogen (N) to promote cell divisions, an efficient system to deal with N scarcity is deterministic for survival. For example, the fission yeast Schizosaccharomyces pombe (S. pombe) withstands long-term N starvation by induction of growth arrest and quiescence entry (G0 phase). In past studies 89 S. pombe genes were found to be required for survival upon G0 phase. Because these are involved in diverse intracellular functions, a clear mechanism for quiescence was difficult to assign. Therefore, we developed a BLAST-based approach to generate phylogenetic profiles and characterize evolutionary conserved metabolic reactions by comparative analysis using prokaryotic databases of Escherichia coli and Bacillus subtilis. We report fourteen proteins fulfilling this study’s homology criteria, half of which (7/14) are localized to the mitochondria. Subsequent in silico analysis suggests the involvement of two functional response mechanisms which comprise oxygen-consuming and glutamate-metabolic reactions. Among these, ∆sod2 showed abnormally low oxygen consumption upon quiescence and loss of regenerative capability. We report that the conserved mechanisms to survive N starvation comprise enzymes which are associated with the regulation of oxygen and glutamateoxoglutarate metabolism.

Zafer Hawash

Supervisor Name: Yabing Qi

Research Unit: Energy Materials and Surface Sciences Unit

Thesis: Surface science studies of perovskite solar cells: spiro-MeOTAD hole transport material and perovskite absorber

Abstract:

The past few years have witnessed an emergence of an outstanding class of thin film solar cells, which are based on organic-inorganic light absorbers, namely, perovskite solar cells (PSCs). PSCs possess energy conversion efficiencies (PCEs) comparable to traditional silicon and other solar cell technologies. To achieve high efficiencies, typically perovskite materials are sandwiched between selective contacts, which significantly facilitate charge carries extraction. These contacts are made of either electron or hole selective material and are called electron transport material (ETM) and hole transport material (HTM). This thesis discusses surface science aspects of the doping mechanism of spiro-MeOTAD HTM, and an engineering approach of the perovskite/spiroMeOTAD HTM interface for enhanced energy level alignment, efficiency, and reproducibility.

In this thesis surface science techniques (i.e., photoemission spectroscopy (PES), atomic force microscopy (AFM), and scanning electron microscopy(SEM)) combined with currentdensity voltage (J-V) measurements on hole only devices and PSCs are employed. PES measurements revealed that ambient air-exposure results in the migration of the commonly used Li-salt dopants from the bottom to the bulk (including top surface) of the spiro-MeOTAD HTM film. AFM and SEM images revealed the presence of pinholes with an average diameter of ~135nm, with a density of ~3.72 holes/µm2, and these pinholes form channels wiggling across the doped spiro-MeOTAD film. Under controlled environments of H2O (relative humidity 90%) and dry O2, PES measurements revealed that H2O is the constituent component in ambient air that leads to the V redistribution of the LiTFSI dopants. In addition, the J-V measurement results on hole only devices revealed that H2O vapor exposure results in an irreversible enhancement of LiTFSI-doped HTM conductivity due to redistribution of the LiTFSI dopants across the HTM film, which was examined by PES measurements. On the contrary, O2 exposure results in a reversible enhancement of the HTM film under applied bias, in which this enhancement is mainly due to O2 doping, which was confirmed by PES measurements.

In addition, to achieve better energy level alignment between the HTM and the perovskite absorber, an intentional deposition of an ultrathin layer of methylammonium iodide (MAI) on top of a methylammonium lead iodide (MAPI) perovskite film was implemented. Using PES measurements, it was found that the deposition of small amount of MAI on top of MAPI results in an interfacial, favorable, energy-level tuning of the MAPI film. XPS measurements revealed that the enhanced energy-level tuning was from MAI dissociated species, not the MAI itself. Furthermore, the optimized energetics were verified using perovskite solar cells. Substantially enhanced stabilized-PCE and reproducibility was achieved (from 15 ± 2% to 17.2 ± 0.4%).

Rico Pohle

Supervisor Name: Nic Shannon

Research Unit: Theory of Quantum Matter Unit

Thesis: Signatures of Novel Spin Liquids in Kagome-like Lattices

Abstract:

The phenomenon of magnetism in solids aroused the curiosity of scientists already in ancient times. While quantum mechanical effects on a single–particle level are well understood, magnets offer phenomena caused by collective interactions between many electrons and provide the opportunity to find novel states of matter. In this context, frustrated magnets play a central role, since interactions between local magnetic moments on a crystallographic lattice cannot be satisfied at the same time. This can prevent the systems to order even at very low temperatures, creating a magnetic state similar to those of liquids, which gives them the name spin liquids. Within this field, the kagome lattice — a two–dimensional network of corner–sharing triangles — has played a particularly iconic role and continues to provide rich inspiration to theoreticians and experimentalists alike.

In this thesis, we first explore the thermodynamic properties and signatures of classical spin liquids on kagome–like lattices, by the use of complementary analytical Husimi tree and numerical Monte Carlo simulation techniques. The emerging phenomenon ofa Curie–law crossover, reflecting a crossover between a high–temperature paramagnet and a low–temperature collective paramagnet, turns out to be a powerful signature of exotic physics in classical spin liquids, and explains the difficulty of making a precise estimate of the Curie–Weiss temperature in experiments.

But spin liquids do not necessarily need to show just one Curie–law crossover. The anisotropic Ising model on the shuriken, or square–kagome lattice, shows a succession of multiple Curie–law crossovers due to a rich phase diagram with many disordered ground states. Hereby, low–and high–temperature regimes are less correlated than the intervening classical spin liquid, allowing to extend the definition of reentrant phenomena to disordered systems.

Furthermore, we also study dynamical properties of the nearest–neighbour Heisenberg model on the bilayer breathing kagome lattice, which has been motivated by recent experiments on Ca10Cr7O28. Using semi–classical molecular–dynamics simulations, we are able to reproduce many features seen by inelastic neutron scattering experiments and provide a first explanation of its spin–liquid origin. Surprisingly, we find that excitations encode not one, but two distinct types of spin liquids at different time scales. Fast fluctuations reveal a Coulombic spin liquid, as known from the classical kagome–lattice antiferromagnet, while slow fluctuations reveal a spiral spin liquid, which can be understood by a mapping onto an effective spin-3/2 honeycomb model.

Kenneth Baughman

Supervisor Name: Noriyuki Satoh

Research Unit: Marine Genomics Unit

Thesis: Decoding and Analysis of the Crown-of-Thorns Starfish Acanthaster planci Genome

Abstract:

Echinoderms are at the base of the deuterostome clade, yet have radial body plans, a water-vascular system, and exoskeletons. In order to investigate how genomes control development, I studied the “Crown-of-Thorns Starfish” (COTS) or Acanthaster planci genome. I made four discoveries from sequencing two COTS specimens, one from the Great Barrier Reef, Australia (‘GBR’) and the other from Okinawa, Japan (‘OKI’). Separate 384 megabase (Mb) assemblies containing ~24,500 genes were generated. First, I discovered that both genomes displayed unexpectedly low heterozygosity; reciprocal BLAST alignment of scaffolds revealed 98.8% nucleotide identity, consistent with a single pacific COTS clade undergoing a recent population expansion. Second, although the unique Hox gene order in sea urchins was hypothesized to be related to pentaradial body plans, I discovered that COTS Hox and ParaHox clusters resemble hemichordate and chordate clusters. The COTS Hox cluster shares with sea urchins the transposition of even-skipped (Evx), as well as posterior Hox reorganization. I thus proposed an evolutionary scenario for how shuffling of the Hox cluster in urchins may have arisen. Third, recent studies show that hemichordates possess a deuterostome-specific cluster of transcription factors associated with development of pharyngeal gill slits. Although extant echinoderms do not have pharyngeal gill slits, I found the cluster in the COTS genome, supporting an ancient origin for pharyngeal gill slits as a deuterostome-defining morphological feature. Fourth, using systems biology notation, I mapped COTS candidate genes for 1-methlyadenine (1-MA)-mediated oocyte maturation. This thesis confirms that the high quality of the COTS genome is biologically significant, and amendable to future studies. Although COTS are famous for decimating coral reefs, this thesis shows that COTS can also be used for genomic and evolutionary developmental research.

Dongxin Zhang

Supervisor Name: Fujie Tanaka

Research Unit: Chemistry and Chemical Bioengineering Unit

Thesis: Amine catalyzed functionalization of enolizable ketones

Abstract:

Development of efficient methods for the synthesis of biologically important molecules in safe, atom economical, and environmentally friendly ways is a significant goal of modern organic chemistry. In this thesis, efficient methods using amines as catalysts for functionalization of enolizable ketones and synthesis of potential biologically active molecules have been developed. First, 1,8-diazabicyclo[5.4.0]undec-7-ene (DBU) was indentified to be an efficient catalyst for fast regioselective aldol reactions. Molecules containing tertiary alcohols were concisely obtained through the aldol reactions. The developed DBU-catalysis was applied for the synthesis of spirooxindoles and trifluoromethyl-substituted alcohols. Although the DBU-catalyzed aldol reactions are not enantioselective, the enantiomerically pure forms of the aldol products derived from -keto esters were obtained by the resolution of the enamines of the aldol products with a homochiral amine. Second, deuteration studies were carried out to elucidate the mechanism of the regioselective formation of the products in the aldol reactions catalyzed by DBU and to understand the relationship between the formation of carbanion or its equivalents and the bond-formation. Finally, enantioselective oxa-hetero-Diels-Alder reaction of enones with aryl trifluoromethyl ketones catalyzed by a novel amine catalyst system was developed. Tetrasubstituted carbon centers bearing a trifluoromethyl group were concisely constructed with the formation of the tetrahydropyranone ring. The hetero-Diels-Alder products were further transformed to various trifluoromethyl-substituted tetrahydropyran derivatives.

Yi-Jyun Luo

Supervisor Name: Noriyuki Satoh

Research Unit: Marine Genomics Unit

Thesis: Insights into lophotrochozoan evolution and the origin of morphological novelties from brachiopod, phoronid, and nemertean genomes

Abstract:

Brachiopods, phoronids, and nemerteans are closely related lophotrochozoans, yet they carry distinct feeding apparatuses and lifestyles. They are poorly studied despite their importance in ecology, evolution, and paleontology. As a result, the genetic basis of their evolutionary origins and body plans have been obscure. Since the Cambrian explosion ~540 million years ago, animal forms have greatly diversified. One fundamental question of animal evolution is how these diverse morphologies are formed. While animals share many developmental toolkit genes, they also possess novel genes and expansion of gene families in a lineage-specific manner. How lineage-specific genes and changes of genomic features contribute to morphological novelties is still a challenge in understanding animal evolution. Also, whether common toolkit genes are involved in patterning these novelties at the genomic level is not well understood. Here I present the genomes of the brachiopod Lingula anatina, the phoronid Phoronis australis, and the nemertean Notospermus geniculatus, together with multiple transcriptomes, providing a comparative platform to understand the evolution of animal genomes and the origin of lophotrochozoans.

Using genomic, transcriptomic, and proteomic approaches, I show that although Lingula and vertebrates have superficially similar hard tissue components, Lingula lacks genes involved in bone formation, suggesting an independent origin of their phosphate biominerals. Several genes involved in Lingula shell formation are shared by molluscs. However, Lingula has independently undergone domain combinations to produce shell matrix collagens with epidermal growth factor domains and carries lineage-specific shell matrix proteins. Gene family expansion, domain shuffling, and co-option of genes appear to be the genomic background of Lingula’s unique biomineralization. Genome-based phylogenetic analyses place Nemertea sister to the group of Brachiopoda and Phoronida. Lophotrochozoans share many gene families with deuterostomes, suggesting that lophotrochozoans retain a core set of bilaterian gene repertoire rather than ecdysozoans or remaining spiralians. Comparative transcriptomics demonstrates that lophophores of brachiopods and phoronids have resemblance not only morphologically but also at the molecular level. Despite lophophores are dissimilar from head structures, lophophores highly express vertebrate head organizer and neuronal marker genes, probably indicating a common origin of bilaterian head patterning. Together, this study reveals a dual nature of lophotrochozoans in which bilaterian-conserved and lineage-specific features shape the evolution of their genomes.

Mart Toots

Supervisor Name: Ulf Skoglund

Research Unit: Structural Cellular Biology Unit

Thesis: Exploring the potential of cryo-electron tomography on protein nanocrystals for molecular structure determination

Abstract:

The three-dimensional structure of a protein molecule, challenging to determine and close to impossible to predict, plays a key role in understanding protein function and has implications in drug design. When it comes to structure determination, there exist many complementary methods, each with their specific advantages and disadvantages. Most of those methods rely on a combined signal from thousands of individuals and cannot be used for directly reconstructing an actual 3d volume as it appears inside the sample.

This thesis focuses on developing the methodology and providing proof of concept for a novel approach in structure determination by reconstructing small protein nanocrystals via cryo-electron tomography. Real-space imaging gets past the phase problem that is a challenging companion of conventional diffraction-based methods. With electron tomography we can reconstruct and visualize a 3d nanocrystal in its entirety and study the properties of small biological crystal from a new perspective.

Being a relatively unexplored territory, nanocrystal tomography sets several challenges, such as creating nanocrystals small enough for imaging with transmission electron microscope and developing algorithms for going from a tilt-series to a 3d structure. For a proof of concept we create, image and reconstruct nanocrystals of hen egg white lysozyme that, having molecular weight of only 15 kDa, is generally considered unfeasible for electron tomography. Nanocrystals make finding and determining the relative orientations of the individual molecules possible, symmetry relations help reduce the effects of missing information, and by averaging we are able to reconstruct a molecular structure at a medium resolution of around 13 Å. Using Fourier Transform (FFT) we get a direct objective measure of the resolution of details within the reconstruction in the form of a diffraction pattern and show that in specific directions the resolution reaches as high as 7Å in a single tomogram.

Additionally, this work explores two other tightly related ideas. First, we study the concept of extended field and show with extensive simulations that extending the reconstruction space in various regularized iterative reconstruction procedures helps reduce the overall error and prevent over-smoothing. Second, calculating FFT of an image comes at a computational cost, and when the image is not periodic, the discontinuity of the opposing edges causes undesirable strong artifacts in the FFT that could obstruct important details. In this project we implemented a simultaneous 2d FFT and edge artifact removal for real-time applications on a Field Programmable Gate Array (FPGA) reconfigurable computing system.

Faisal Mahmood

Supervisor Name: Ulf Skoglund

Research Unit: Structural Cellular Biology Unit

Thesis: Algorithmic and architectural developments for cryo-electron tomography

Abstract:

Molecular structure determination is important for understanding functionalities and dynamics of macromolecules, such as proteins and nucleic acids. Cryo-electron tomography (Cryo-ET) is a technique that can be used to determine structures of individual macromolecules, thus providing snapshots of their native conformations. Such 3D reconstructions encounter several types of imperfections due to missing, corrupted and low-contrast data. This thesis focuses on the algorithmic and architectural aspects of improving and accelerating reconstructions for Cryo-ET specifically and the tomographic image reconstruction in general. The thesis explores modern compressed sensing and graph-based approaches for noise removal and recovering the missing wedge which act as a proof-of-concept for the applicability of sparsity exploiting methods to tomographic image reconstruction. The thesis also explores, analyses and explains an extended field (EF)-based noise removal method. When used in conjunction with a variety of reconstruction procedures with a regularization capability it proved to be computationally efficient, reliable and stable. Through extensive empirical simulations it was shown that extending the reconstruction space reduces the error at a relatively lower regularization parameter thus allowing a better fit with the projections and preventing over-smoothing. Computational constraints are a major issue in speeding up tomographic reconstruction and refinement. One of the fundamental components, which often become a bottleneck in a variety of reconstruction procedures, is the fast Fourier transform (FFT), specifically for analytical reconstructions. Generally, FFTs suffer from edge artifacts or series termination errors, which stem from the fact that two opposing edges of an image are often not periodic. These artifacts can propagate to next stages of processing and appear as errors in reconstructions. This thesis also explores simultaneous 2D FFTs and edge artifact removal for real-time applications. This was accomplished on a multi-FPGA (Field Programmable Gate Array) reconfigurable computing system with a high-speed bus. The algorithmic optimization and architecture are general and can be replicated to a variety of different hardware setups.

Lee O'Riordan

Supervisor Name: Thomas Busch

Research Unit: Quantum Systems Unit

Thesis: Non-equilibrium vortex dynamics in rapidly rotating Bose-Einstein condensates

Abstract:

The work carried out during Lee's PhD deals with the dynamics of vortices in ultracold Bose-Einstein condensates in non-equilibrium situations. While the equilibrium form of a rotating BEC carries a perfectly triangular lattice of vortices, perturbing this in a controlled way is useful to understand the dynamics of the individual vortices. Lee has carried out two main projects, with the first one dealing with perturbations of the background gas and the second one introducing a method that can remove or add localised angular momentum at will. Both lead to fundamentally different dynamics and are discussed in detail.

Vortices are topological excitations and therefore long lived. This makes their quantised angular momentum a potential candidate for storing quantum information. At the same time such systems are prototypes for studies in quantum turbulence and Lee's work provides a step towards further studies in both directions.

A second aspect of Lee's thesis is the development of a software suit that is tailored towards solving the non-linear Gross-Pitaevskii equation using advanced graphics processing units (GPUE). The area of GPU computing is under fast development and Lee has made significant contributions towards making it more accessible for anyone working with the GPE.

Mark Daly

Supervisor Name: Síle Nic Chormaic

Research Unit: Light-Matter Interactions for Quantum Technologies Unit

Thesis: Light-induced interactions using optical near-field devices

Abstract:

Optical near-fields are generated when light passes through components with wavelength, or subwavelength features. The near-fields generated at the surfaces of devices are often neglected, in part because the far-fields have more applications and are more readily accessible. Near-fields, as one might expect, occur very close to the surface of the material through which the light passes. However, near-fields present an interesting method of overcoming Rayleigh's diffraction limit. For example, the evanescent field at the surface of a prism or ultrathin fibre rapidly decays, but can exist in sub-diffraction limited areas. Similarly, the field generated by a subwavelength aperture or a plasmonic particle can have local field distributions with minute dimensions, allowing one to confine light to areas otherwise unattainable, extremely close to the surface of the material in question. By exploiting this aspect of optical near-fields we apply them to problems in atom and particle trapping.

Our main focus is on ultrathin optical fibres. These fibres differ from telecommunications fibre due to their lack of cladding material and their wavelength-scale dimensions. These two factors combine to produce a significant evanescent field at their waist. This field is readily accessible and can be used to trap particle or atoms through the optical forces which arise in such light-matter interactions. We can also use such devices to passively collect light which is emitted into the available guided mode. Here, we demonstrate how an ultrathin fibre can be used as a probe to determine the temperature of a cold atom cloud.

Ultrathin fibres, while extremely useful, have some limiting factors related to the strength and distribution of their evanescent fields. To improve upon the design, we also investigated how one can nanostructure an optical fibre using focussed ion beam milling techniques or combine optical fibres with gold dimer arrays to produce localised field enhancements. We used nanostructured fibres to trap 100 and 200 nm dielectric spheres within the structured region. Various numerical techniques were employed to characterise both the nanostructured fibre and the plasmonic-enhanced fibre.

Aside from optical fibres, we also briefy discuss how an array of Fresnel microlenses can be packaged with other atom chip designs to produce a device which could trap atoms microns away from a gold surface. We discuss the theory and fabrication technique for such a Fresnel microlens array atom chip.

Po-Shun Chuang

Supervisor Name: Satoshi Mitarai

Research Unit: Marine Biophysics Unit

Thesis: From Polyps to Colonies: Applying Polyp Bail-Out to Study Coral Coloniality

Abstract: