Many times a second we engage the motor system: we evaluate evidence, plan options, decide, and act, all within 10’s of ms. Given fixed times for nerve conduction and synaptic delays there would appear to be little time for processing through multiple circuits. However, both primates and cats appear to have widespread and relatively unstructured lateral connectivity within their motor cortex. Such extensive local motor cortex connections may endow the system with flexibility needed to learn new tasks and coordinate a wide range of outputs. Our central question is how the local circuits in the primary rodent motor cortex process sensory information to produce a coherent, integrated, and specific pattern of corticospinal output. When faced with these less structured circuits we exploit genetics and develop mouse models where specific classes of neurons can be selectively excited or inhibited using light over both the timescale and landscape of the cortical area devoted to sensory-motor processing. The primary focus will be on the in vivo structure and function of single neurons and the local circuits they form in the rodent motor cortex M1, specifically in the forepaw representation that are associated with simple reaching tasks. Our hypothesis is that the motor cortex implements a decision making machine that operates in two main stages through the cortical lamina. The widely distributed connections in the superficial layers provide the circuits for broadcasting incoming patterns of sensory information for evaluation. The resulting decisions are then transmitted to the deep cortical layers, which coordinate the output patterns. While the basic input pathways and output neurons in the rodent forepaw representation of M1 are known, there are few data on functional properties of individual neurons, laminar differences in properties, or their collective activity. It is intended that the experiments outlined here will provide the functional and structural data for us to develop a biologically-realistic model of the mouse motor cortex.

Mechanosensation ranges from osmotic pressure sensation in bacteria to touch and hearing in humans. Compared to smell, taste, and vision, it remains poorly understood at the molecular level. For example, it is not known how force actually gates channels in most systems. A related but unanswered question is how forces that are different at the macroscopic level such as pokes and whole body bends, lead to different neuronal responses even though the forces may look similar at the molecular level. In other words, how can a channel that is essentially making a point measurement of stress distinguish force fields that vary over much longer length scales? I will use a culture of primary C. elegans neurons combined with a hybrid fluorescence/atomic force microscope to apply precise forces to mechanoreceptor neurons while monitoring their response with calcium imaging. Examining the spatial dependence of force sensitivity at the sub-cellular level will give insight into the gating mechanism, while comparisons between neurons that respond to harsh touch and those that respond to body bending will reveal how these neurons differ and may reveal general principles of mechanoreceptor specialization. This system will also enable novel investigations into the spatial distribution of different sensory modalities in polymodal sensory neurons by applying tip-attached chemicals to precise locations along neural processes or by local tip heating.

In response to nutrient starvation, yeast cells enter a quiescent state. This state is defined by a marked increase in the viscosity of the cell cytoplasm, a phenomenon we term “cell-freezing”. Here we propose a detailed structural and functional study of cell-freezing in fission yeast (S. pombe). Cell-freezing is a novel phenomenon discovered only recently through measurements of the viscoelastic properties of the cytoplasm of fission yeast cells. It leads to a dramatic immobilization of all visible sub-cellular structures and implies the presence of a mechanically stable network throughout the cell, we thus hypothesize that such a dense and homogeneous network develops when the cell enters the starvation state and that this network is sufficiently fine to immobilize all major cell components.

The four objectives of the proposed research are A: characterize details of cell-freezing, B: identify and dissect its molecular machinery, C: analyze its function, and D: search for cell-freezing in higher organisms. To address these objectives, we will follow an interdisciplinary approach that includes methods from cell and molecular biology, structural biology and (bio-)physics.

The presence of lesions in the DNA template often interferes with the progression of DNA replication and leads to stalling or collapse of replication forks. Restart of such forks is essential for completion of DNA replication and consequently for cell survival and maintaining genomic integrity. Translesion synthesis (TLS) is one of the two main pathways mediating replication restart. It employs specialized, low fidelity DNA polymerases that are able to synthesize DNA across the damaged region. Unfortunately, TLS polymerases frequently introduce mutations, which contribute to cell dysfunction, aging, and malignant transformation.
To employ the beneficial properties of TLS and limit potentially dangerous mutagenesis, cells strictly regulate the TLS process. TLS depends on mono-ubiquitylation of PCNA, which promotes lesion bypass by TLS polymerases on the damaged template. However, the molecular details of this process remain uncertain.
To improve our understanding of how TLS occurs and how it is regulated, we propose to use a unique Xenopus egg extract system to analyze a TLS event at a defined, damaged locus. This system allows us to precisely follow DNA intermediates of the lesion bypass reaction. By comparing these intermediates with the status of PCNA ubiquitylation, we will determine whether and when PCNA is ubiquitylated during TLS-mediate restart of the stalled replication fork. We will also address how inhibiting PCNA ubiquitylation affects lesion bypass. Additionally, using chromatin immunoprecipitation, we will analyze the binding of different TLS and replicative polymerases to the stalled and restarted replication fork.

Somatic nuclei can be reprogrammed to a pluripotent or embryonic stage. Somatic cell reprogramming has been shown to be possible via animal cloning, cell fusion, and ectopic expression of a few embryonic stem cell (ESC) factors. The reprogramming time of a cell nucleus is normally relatively long and it depends upon cell type, methods of reprogramming, and culture conditions used. These observations led us to envisage that the process according to which somatic cells change their fate involves many factors and sequential events. Thus, transcription factors, chromatin-modifying complexes and signalling molecules are likely to operate in a specific order to induce and maintain the reprogramming of a somatic nucleus. We have recently shown that the periodic activation of Wnt/beta-catenin signalling controls fusion-mediated somatic-cell reprogramming. Only this periodic and limited nuclear accumulation of beta-catenin allows ESCs to reprogramme somatic cells after fusion. Furthermore, in ESCs, either overexpression of Axin2, which is part of the beta-catenin destruction complex, or deletion of GSK3, which is devoted to the phosphorylation of beta-catenin, inhibits ESC-fusion-mediated somatic-cell reprogramming. In contrast, deletion of a specific beta-catenin modulator, in ESCs strongly enhances their reprogramming ability. Our hypothesis is that signalling dependent genes encode for key reprogramming factors, which we aim to identify. We will perform transcriptome, microRNAome and epigenome analyses of ESCs pre-treated with positive and negative modulators of the Wnt pathway. In addition, we will perform the same type of analysis using wild-type and mutant ESCs which carry the activated or inhibited Wnt signalling pathway. All of the collected data will be analysed by ARACNe, MINDy and MRA reverse engineering algorithms, to reconstruct the reprogramming regulatory network and to identify pathways and master reprogrammer genes. Finally, the reprogrammer candidates will be cloned and validated for their ability to induce fusion-mediated reprogramming or direct reprogramming.

How do neural networks produce behaviors, and how do neural networks evolve to produce new behaviors? The relative simplicity of fly courtship, its stereotyped and yet rapidly evolving nature, the limited number of neurons involved, and the ease with which we can manipulate these neurons in vivo make this the ideal system to study this problem. Genetic approaches have shown that nearly all steps of the innate male courtship sequence in Drosophila melanogaster are controlled by approximately 1200 neurons which express the male-specific form of the transcription factor fruitless (FruM). I propose to characterize the cis-regulatory elements (CREs) controlling FruM expression in Drosophila melanogaster. I will use these CREs to inactivate or hyperactivate subsets of FruM neurons in the adult fly and assay the effect on courtship behavior. Additionally, I will drive transgenic reporters to describe their neuro-anatomy (number, location, and connectivity). This will determine the functional organization and degree of modularity - both anatomical and behavioral - inherent in the courtship neural network. I will then use conserved fruitless CREs to assay function, expression, and neuronal morphology in fly species with divergent courtship behaviors in an effort to understand how the courtship neural network has evolved to produce new behavioral outcomes.

X chromosome inactivation is a dosage compensation strategy used by mammals to ensure that female (XX) and male (XY) cells transcribe equal levels of genes. Repression of genes on the inactive X chromosome (Xi) is a major example of stable silencing that is maintained for the life of the organism. Reactivation of the Xi occurs both in vivo and in vitro when female mouse somatic cells acquire a pluripotent state. The aim of this research is to decipher the molecular mechanism by which a human somatic cell reprograms towards pluripotecy, through investigating the X chromosome reactivation. With this aim, I will use the heterokaryon system, whereby human B-lymphocytes are reprogrammed via fusion with mouse embryonic stem (ES) cells, as this method allows the greatest accessibility to monitor changes occurring during reprogramming progression. First, I will identify and characterize the dynamics of molecular events that are associated with the X reactivation in the human nuclei. I will then investigate the role of transcription factors (Oct4, Sox2 and Nanog) and repressive complexes (PRC1/2) associated with pluripotency in establishing Xi reactivation and its maintenance. I will fuse human B-lymphocytes with mouse ES cells carrying inducible knock-out/down alleles of the factors and I will monitor the different events associated with X reactivation upon deletion of the factors at different times after fusion. This study will provide insights into the molecular mechanisms of reprogramming and will further our understanding of stem cell identity and differentiation.

I propose to advance statistical methods for the understanding of genetic variation in populations by developing a first-order probabilistic language for coalescent models. Coalescent models are routinely used to analyze genetic polymorphism data to, for example, identify recombination hotspots in the genome, or to test theories of gene flow in speciation. More recently, these models have been used in genome-wide studies to identify correlations between genetic polymorphisms and disease. Basic coalescent theory has been extended to account for recombination, selection, and a variety of mutation processes and population demographics. However, a significant barrier to the development of more sophisticated and accurate coalescent models, as well as to their more widespread use in genome-wide studies, is the difficulty of implementing reliable and computationally feasible inference for these models. Rather than design inference for a single model, I propose to investigate inference for a family of coalescent models by borrowing techniques developed for research on first-order probabilistic languages. My project will have two major impacts on science. First, it will make coalescent models more accessible to geneticists by allowing them to construct models for their own studies without having to implement inference from scratch. Second, it will lead to a more cohesive understanding of the inference challenges in coalescent theory, a necessary step to overcoming barriers in the development of more sophisticated models. This project is my first step in a career devoted to investigating the rich set of problems in population genetics, a field of growing importance and scope.

Cycle-Quant: Developing Cell Cycle Profiles to Classify Responses and Regulators

Readily detectable changes in DNA content and cellular morphology are used to assign conventional cell cycle phases. However, cell division requires coordination of cell size, organelle duplication, cytoskeletal rearrangements, and changes in key proteins. Here, an interdisciplinary team of researchers will develop new methods for advanced cell cycle profiling. Novel landmarks will be identified in the continuum of the mammalian cell cycle, and in cycles of organelles and cell cycle proteins. An integrated and multi-disciplinary team of biologists, chemists, and computer scientists will establish new genetic and chemical cellular labels and machine learning-based computational techniques applicable to time-lapse microscopy.

The main aims are:

1. To develop advanced cell cycle profiling using a standardised and expandable set of non-transformed human cell lines. Different classes of genetically encoded fluorescent biological markers track, directly or indirectly reflecting cellular changes related to the cell cycle, will be tracked by time lapse fluorescence microscopy. Acquired image data will be processed by new computer-based data extraction methods for automated image recognition and adequate description of a spectrum of ‘sub-cellular cycles’ such as those of dividing organelles.

2. To develop novel, and uniform tools to identify flexible key cell-cycle landmarks, bypassing the requirement for the elaborate, intrusive introduction of genetic markers. This would create unprecedented flexibility for cell cycle profiling in any given cell type. We will a) use rich information from non-fluorescent cell images and, b) develop and validate permeable fluorescent cell-cycle-markers.

3. We will test cell cycle profiling by quantitatively comparing functions of key regulators of the G2-M transition and classify new regulators in focussed pathways.

We will combine our expertises to re-define cell cycle progression in a quantitative way, with the goal of assigning complex cell-cycle progression profiles. This coordinated effort will provide unprecedented opportunities to quantitatively compare genetic and chemical perturbants, essential for assigning functions to newly identified cell cycle regulators, and the development of robustly predictive models of cell cycle progression.

Stem cells possess two defining features; the capacity to maintain their identity during extended cell division (referred to as self-renewal) and the capacity to alter their identity to form specialised cells (referred to as differentiation). Preservation of a functional stem cell population requires a balance between self-renewal and differentiation. Although functionality operates on the stem cell population as a whole, it is the choices of individual cells that must be controlled to maintain such functionality. Self-renewal of embryonic stem (ES) cells is controlled by the action of a gene regulatory network centred around the transcription factors Oct4, Sox2 and Nanog. However, the mechanisms controlling operation of this network are not well understood. Recently, we demonstrated an heterogeneity in the expression of Nanog, which is strongly and directly correlated with self-renewal efficiency. This leads us to hypothesise that the cross-regulatory interactions between Oct4, Sox2, Nanog proteins and their target genes are only partly stable. This may allow a dynamic range of states to be explored by ES cells, only some of which enable differentiation. Here, our objective is to test this hypothesis by watching how the Oct4/Sox2/Nanog transcription factor network operates in individual stem cells. To do this we will put differently coloured “reporters” into each of the transcription factor genes (e.g. red at Oct4, blue at Sox2 and yellow at Nanog). These reporters will tell us when, where and to what extent each gene is switched on. We will then use precision engineering to allow ES cells, the developmentally more advanced epiblast stem cells and early mouse embryos to be cultured in small chambers with a surface area of less than 1 mm2 over which we can flow culture fluid with varying levels of “switch” chemicals. These “switch” chemicals will be used to alter the action of modified transcription factor genes that we have also inserted into the cells. Changes in expression of the network components can then be followed microscopically using the coloured reporters in real time. These studies will deliver insight into the regulatory dynamics underpinning stem cell identity. They will allow us to see the dynamic changes as they occur in stem cells that are maturing during development and will enhance our ability to manipulate stem cells behaviour at will.

Chromosome segregation during mitosis is mediated by a microtubule-based apparatus known as the mitotic spindle. However, the requirement for a “spindle matrix” has also been proposed, but its functional role in mitosis remains highly controversial. The fidelity of chromosome segregation is controlled by the spindle assembly checkpoint (SAC), a signaling pathway that delays mitotic progression to ensure that kinetochores (KTs) from all chromosomes are correctly attached to mitotic spindle microtubules before anaphase onset. A single unattached KT produces a “wait anaphase” signal whose inhibitory action is restricted to the vicinity of the spindle by an unknown mechanism. Recent work from the Maiato lab showed that Megator and the SAC protein Mad2 form a conserved complex that localizes to a nuclear-derived spindle matrix in living cells. This interaction promotes the recruitment of Mad2 and its activating kinase Mps1 to unattached KTs, thereby mediating normal mitotic duration and SAC response. These data provided a new conceptual view of the spindle-matrix, not as a rigid structural scaffold as classically envisioned, but as a dynamic spatial determinant of the SAC. Here, we propose a synergistic approach to dissect the molecular mechanism by which the spindle matrix controls mitosis in time and space. We will use affinity-tagged spindle matrix and SAC proteins in Drosophila and human tissue culture cells to isolate and identify common molecular interfaces by mass spectrometry. The function of potential candidates will be analyzed using a combination of laser microsurgery and RNAi to inflict structural and/or molecular perturbations of the spindle matrix and other components of the mitotic apparatus, while assessing their role in mitotic progression and fidelity in living cells. These studies will be complemented by Fluorescence Correlation Spectroscopy (FCS), Image Correlation Spectroscopy (ICS) and Fluorescence Recovery after Photobleaching (FRAP) studies to investigate the physical properties of the spindle matrix and to determine the nature of its spatiotemporal interaction with SAC proteins in vivo. Ultimately, we will integrate these data into a mathematical model of SAC regulation by the spindle matrix. In total, these studies will provide important novel insights into the critical problem of mitosis by capitalizing on an interdisciplinary approach.

Cells of our immune system normally aid tumour destruction, but in some situations do the reverse and promote tumour growth and spread. This project will bring together novel techniques and tools together with expertise acquired during applicant's postdoctoral training to gain unique insights into the interactions between tumours and the immune system with a view to identifying what factors determine a positive versus negative outcome. In particular, we will use 2-photon microscopy to track fluorescently-labelled white blood cells termed neutrophils, dendritic cells and T cells in real time during tumour development. This will give us an unprecedented opportunity to ‘see’ how these cells interact with malignant cells at different stages of tumour development which could lead to novel therapeutic approaches.

Accurate chromosome segregation depends on proper attachment of kinetochores to microtubules. Mounting evidence suggests that mechanical properties of the kinetochore make fundamental contributions to faithful segregation of chromosomes. However, such kinetochore properties have not been experimentally analyzed. Merotelic kinetochore orientation, a major cause of aneuploidy, is an error in which a single kinetochore binds microtubules emanating from opposite spindle poles. During anaphase, merotelically attached kinetochores have a stretched appearance due to pulling forces exerted by microtubules emanating from both spindle poles. We will establish merotelic kinetochores as a model for studying structural and mechanical properties of the kinetochore in vivo. Specifically, we will take advantage of recent achievements in biophysics and use laser ablation to sever one of the microtubule bundles attached to a merotelic kinetochore in anaphase cells. We argue that the change of the kinetochore shape after laser severing should reflect its elasto-plastic properties. We will image fluorescently labeled kinetochore proteins representing major kinetochore subdomains to determine the molecular architecture of the merotelic kinetochore before and after laser ablation. We will perform these experiments both in the fission yeast Schizosaccharomyces pombe and in mammalian tissue culture cells. This will allow us to identify molecular and mechanical kinetochore features that are either distinctive or evolutionarily conserved. Ultimately, we aim at integrating mechanical properties of the kinetochore into current biochemical models of kinetochore function and chromosome segregation.

Pheromones are chemicals that are produced by individual organisms, and that elicit highly stereotyped behavioral and developmental responses in other individuals of the same species. The chemical composition of pheromones is often complex, consisting of many small molecules of related or unrelated structures. These molecules can elicit distinct responses when acting individually or in combination with other pheromone components. Pheromones are recognized by members of large families of seven transmembrane domain receptor proteins in vertebrates and invertebrates, the ligand specificities of only a small subset of which have been described. The challenge of isolating and defining the chemical composition of pheromones, together with the presence of large numbers of putative pheromone receptor proteins, has made it difficult to describe the coding strategy used to respond to these biologically and ecologically important signaling molecules.

The goal of this project is to generate a comprehensive map of pheromone-receptor interactions in the C. elegans model system. Over twenty-five years ago, it was demonstrated that C. elegans produces a mixture of small molecules called dauer pheromone, high concentrations of which trigger entry into the stress-resistant and long-lived dauer larval stage. Recently, we and others have begun to elucidate the chemical composition of dauer pheromone, and have identified two G protein-coupled receptors required for the responses to a subset of pheromone components. The ability to cultivate C. elegans in sufficiently large quantities to extract, identify and synthesize pheromone components, and the genetic and genomic power of C. elegans together with its compact sensory nervous system, makes the generation of a complete pheromone-receptor interaction map a feasible goal in this organism.

The specific aims of this collaborative project are: 1) to define and chemically synthesize all components of C. elegans dauer pheromone; 2) to identify the neurons and chemoreceptors required for the responses to each of the components in vivo; and 3) to characterize the ligand response properties, selectivity, and affinity of the receptors via heterologous expression assays. These interdisciplinary aims integrate the expertise of J. Clardy (USA) in small molecule chemistry, of P. Sengupta (USA) in neurogenetics of C. elegans, and of K. Touhara (Japan) in biochemical mapping of ligand-receptor interactions.

Membrane and secretory proteins that fail to pass quality control in the endoplasmic reticulum are discharged into the cytosol and degraded by the proteasome. Folding and stability are parameters that control protein behavior and function. At present, many of the mammalian components involved in quality control remain to be identified. We propose to exploit sortase-mediated transacylation technology to dissect the crucial protein-protein interactions for proper folding and quality control. Using predominantly an in vitro approach, I shall equip proteins of interest with a sortase tag that allows the installation of a variety of probes, affinity handles and crosslinkers as a bootstrap to identify new players in the complicated molecular machines that execute protein quality control. Covalent ligation or crosslinking of known partner proteins in a sortase-assisted reaction would help identify specific new binders that interact with the newly generated cross-linked products, but not with the constituent proteins individually. The relevance of specific protein interactors and especially of weak binders will be approached by high-confidence quantitative proteomics. Understanding protein quality control in the ER is an essential aspect of biology, and of considerable biomedical relevance, since aggregation of misfolded proteins that escape the cellular quality control machinery underlies a range of important diseases.

Animals walk and run with remarkable speed, robustness and economy. An ostrich can run at speeds up to 45mph, and small ground birds can run at full speed over obstacles up to 40% of leg length. In contrast, the robustness and economy of legged robots, prosthetic legs and powered exoskeletons remains limited. State-of-the-art devices commonly use 15-20 times as much energy as a human or animal, and operate over a small range of terrain conditions. We believe that robotic machines are limited by our understanding of locomotion rather than by limits of technology, and improved theoretical models of running dynamics will help robotic applications live up to their potential.

The goal of this research is to define a single mathematical model that can recreate the running dynamics and whole-body control strategies of ground birds over a range of terrain conditions. We expect that animals achieve robust and economic locomotion using a specific combination of passive dynamics, which is the behaviour of tendons or other mechanical components, and active control, which is any neural activation of muscles. Our mathematical model will explicitly differentiate between passive dynamics and active control.

We will combine animal experiments, computer simulation, and implementation on a running robot in an iterative process to create and validate our mathematical model. Experiments on ground birds and on our running robot will include steady-state running and a range of terrain disturbances. Successful performance of our model and robot will be based on whether they produce structurally identical ground reaction force profiles and body motions as running birds.

Our model will focus on whole-body dynamics and their relation to whole-leg actuation and compliance features; thus, it will apply equally well to animals of varying morphologies and to technological applications. The model will help us understand the interplay of leg design and neuromuscular control in running animals. It will also provide clear guidelines for the design of prosthetic limbs, powered exoskeletons for assisted human mobility, and autonomous legged robots. New prosthetics and other mobility aids that match human walking dynamics and perform well in varied terrain could vastly improve quality of life for those with limited mobility, including amputees and ageing individuals.

Neurons have to perform a wide variety of complex morphogenetic events in order to functionally wire the developing brain. In the adult brain, this process continues with changes in morphology that alters their functional connectivity upon learning and experience, or during the repair of injured neural tissues. Specific signaling events to the cytoskeleton are essential for the control of these processes and are likely to be highly regulated at fine spatiotemporal scales. Accordingly, novel cell biological tools such as fluorescent biosensors that enable dissection of these signaling events in time and space, have given novel insights in signaling complexity. The aim of this proposal is to visualize the dynamics of the activation of specific signaling molecules in a panel of prototypical neuronal behaviors with a strong emphasis on the axon. We will study: axonal specification (the establishment of an axon and a dendrite), axonal guidance (the response of a growth cone to attractive and repulsive cues) and inhibition of axonal regeneration after injury (the response of the axon to a glial scar). Our hypothesis is that these different morphogenetic behaviors involve distinct signaling programs with a tight crosstalk between the signaling molecules of interest. Comparing these in the prototypical cell behaviors should give essential insights in how a neuron can perform this diverse set of tasks using the same signaling machinery. We propose to engineer a new generation of highly sensitive fluorescent probes to monitor the spatio-temporal activation profiles of the specific signaling molecules in primary E18 embryonal hippocampal neurons. Microfluidic technology will be used to engineer highly defined extracellular environments to induce and orient the different axonal behaviors. This will evoke the asymmetric nature of extracellular cues observed in vivo, and recapitulate the associated signal amplification mechanisms that will lead to robust signaling. Finally, computer vision techniques will be devised that allow the pooling of measurements from many cells into statistically significant datasets for identification of the spatiotemporal relationships among different signaling programs and cell behavior. Multivariate time-series analysis methods will be devised to inform the model on the cross-coordination between the activation of the signaling molecules of interest and neuronal morphogenesis.

Noise, generally regarded as a nuisance in signal detection can also be a benefit. The phenomenon known to physicists as stochastic resonance (SR), exploits noise to boost weak signals. It has been postulated that similar principles might also operate within nervous systems. The aim of this proposal is to test a potential noise-generating system in the Drosophila olfactory system. Olfactory systems must detect odorants across a range of intensities. Weak signals generated by low intensity odorants must be amplified to allow reliable detection. Excitatory local interneurons (eLNs) inject non-specific activity into many, if not all, odorant receptor channels in Drosophila. I propose to explore a potential SR-like role of this excitatory offset in odor perception. First, I will test the requirement of eLNs in the perception of low intensity odorants, using a combination of electrophysiological and functional imaging approaches. Second, based on my physiology data, I will perform behavioral studies to correlate physiological responses with behavioral output. Finally, I will test whether the activity injected by eLNs fits the criteria of noise by injecting artificial noise directly into eLN targets while inactivating eLNs. If artificially injected noise can rescue the loss of eLN output in odor perception, it will support a role of SR in olfactory information processing. Importantly, it will also provide the first evidence for the existence of an internal noise-generating system within a nervous system.

Cancer is a major class of diseases that arise due to unrestrained proliferation of cells in our body, leading to abnormal cell growth, distant metastasis and eventually death. Indeed, one in eight deaths worldwide occur due to cancer. Intriguingly, cancers show high diversity -- i.e. no two cancers have an identical molecular profile, and heterogeneity -- i.e. cells within a single cancer are not identical. We are yet to fully realize the impact of this variability on cancer progression and prognosis. I am interested in studying diversity and heterogeneity of tumors at multiple levels: genome sequence, gene expression, interaction of genes in pathways, and cell types within cancers. I will investigate whether an estimate of diversity at one level (e.g. genome) can predict diversity at another level (e.g. pathways) and whether such information can be used to stratify cancer samples. I also aim to study whether the characterization of diversity can have any predictive power for disease progression and development of resistance. My study will span mathematical modeling, systems biology analyses and also experimental validation of the assumptions and predictions of my models. My work will be of fundamental interest for a broad audience because of its implications for understanding cancer genomes and treatment strategies for patients.

The architecture of cells and subcellular structures can show remarkable variability between tissues and organisms, but there is currently little understanding of the evolutionary basis of this diversity. Thus it is unclear why metaphase spindles in different Eukaryotes exhibit a range of morphologies and the volumes of these spindles vary over one thousand fold. Here we propose a comparative study to investigate the evolutionary forces shaping the spindle. Using one cell stage nematode embryos as a model system, we will characterize spindle structure in several strains of ~60 species exhibiting a range of cell sizes, karyotypes, and life histories. A more detailed analysis of microtubule dynamics and spindle architecture from ~10 select species will reveal the biophysical basis of variation in spindle morphology. This comparative data will be interpreted using phylogenies and quantitative models of phenotypic evolution. We will then test the developed biophysical and evolutionary models with perturbation. This unique comparative approach combines molecular methods, quantitative microscopy, and mathematical models, to gain insight into how the mitotic spindle is shaped through evolution by selection, drift, and biophysical constraints.