2019 Simons Society of Fellows Retreat

Start Time

Meeting Location

AV

Dress Code

Saturday Evening Dinner Attire

Naama Aviram
The Rockefeller University

Exploring Bacterial Immunological Memory in CRISPR-Cas Systems

In order for a cell to survive, it must develop means to protect itself from potential pathogenic threats. Indeed, each and every living organism has evolved an immune system that can defend it from the dangers lurking in its environment. It was long known that even simple prokaryotic organisms harbor innate immune systems, such as restriction enzymes, whose purpose is to fight viral infections. On the other hand, adaptive immune systems, in which immunological memory is formed, were thought to exist only in complex eukaryotic organisms. One of the major discoveries of the last decade is the existence of prokaryotic adaptive immune systems that protect bacteria from invading viruses – the CRISPR-Cas systems (Clustered Regularly Interspaced Short Palindromic Repeats – CRISPR associated proteins).

The acquisition of immunological memory, termed “adaptation”, is a fundamental step in CRISPR immunity, and is mediated by integration of viral-derived DNA segments into the bacterial genome. Interestingly, although the viral genome harbors thousands of such segments, only a fraction are consistently acquired as functional immunological memory. In his talk Aviram will describe the molecular and evolutionary implications of the adaptation process and will describe her findings regarding the cross-talk between CRISPR immunity and other fundamental cellular machineries, which together dictate its selectivity and functionality.

Mariana Cardoso
New York University

Impact of Non-sensory Signals in the Early Visual Cortex

Perception and behavior emerge from a complex integration of information from external sensory stimuli and internal states. Internal states include a broad range of conditions, such as how attentive one is. From a neuronal perspective, internal states can alter the patterns of neural activity independent of changes in external inputs. To understand the interaction between external and internal factors, we are using an experimental animal model of amblyopia (“lazy eye”). Amblyopia is a visual developmental disorder, in which inappropriate visual input early in development causes enduring deficits in visual performance. These deficits are not fully explained by changes in the response properties of neurons in the early stages of the visual system. Could the amblyopic behavior deficits be explained by changes in patterns of neural activity that are not directly related to the visual stimulus?

We study alert monkeys performing controlled visual tasks. The subject shows behavioral deficits when viewing through the amblyopic eye, relative to the non-amblyopic (“fellow”) eye. These deficits are more profound than the deficits seen in simple visual acuity, confirming that amblyopia is a complex syndrome. Next, we will record neural activity of groups of neurons in the visual system while subjects perform the visual tasks to investigate how patterns of neuronal activity might change depending on which eye information comes from. In summary, in her talk, Cardoso will evaluate the idea that the representation of visual information in amblyopes depends not only on information about external stimuli, but also on how the brain internally regards information coming from the amblyopic eye.

Rosemary Cater
Columbia University Irving Medical Center

Smuggling Drugs into the Brain? Something seems Fishy…

Omega-3 fatty acids are found in very high concentrations in the central nervous system and are essential for normal brain function. However, omega-3 cannot be synthesized in situ and thus, to meet its requirements, the brain relies predominantly on dietary sources such as fish and eggs. For dietary omega-3 to reach the brain it needs to cross a layer of tightly packed cells that separates the blood from the brain. This ‘blood brain barrier’ exists to protect the brain from toxins, however it also prevents entry of ~98% of therapeutics to the brain and is therefore a major bottleneck in the development of neurotherapeutics. At this barrier, molecular machines called ‘transporters’ that reside in the cell membrane shuttle select molecules from the blood into the brain with a certain degree of specificity. For example, there are specific omega-3 transporters that are very highly expressed within this blood brain barrier that move omega-3 into the brain.

As a structural biologist, Cater is interested in harnessing recent technological advancements to obtain a three-dimensional structure of this omega-3 transporter, so as to be able to understand how this molecule works in both physiological and pathological settings. Using single-particle cryogenic electron microscopy, Cater has recently obtained a high-resolution three-dimensional map of the transporter in a close to native membrane-like environment. This structure provides us with the first insight into what this molecular machine looks like at an atomic level of detail and will allow us to answer fundamental questions concerning how omega-3 is transported across the blood brain barrier. Furthermore, this research will lay the foundation for the rational design of therapeutics that can be selectively shuttled into the brain.

Logan Grosenick
Columbia University

Neural Assembly Dynamics During Behavior

Understanding how coordinated assemblies of neurons in the brain interact to support representation and recall of experiences is a fundamental challenge in neuroscience. Developments in fast computational imaging and genetically-encoded activity reporters have opened up the possibility of recording from large volumes of neurons with cellular resolution as fast as neural dynamics unfold. Grosenick will discuss recent progress in using such approaches to understanding coordinated cell assemblies and their relationship to context and behavior.

Dorri Halbertal
Columbia University

An Intimate Tour in Flatland – The Emerging Field of 2D Materials: The Science and Novel Ways to Explore It

Barely 15 years have passed since the first isolation of graphene, a single atomic layer of carbon, and ever since we are witnessing an explosion in the field of 2D materials science and 2D physics, with an emergence of a level of devices’ control like never before. In a rather straightforward approach, one can nowadays stack atomic layers of materials with building blocks of different physical properties including semi-metals, semi-conductors, dielectrics, magnetic and superconducting materials in order to reach a device’s behavior of choice. Furthermore, these materials show strong tunability by external stimuli, such as electrostatic gating and magnetic field.

Last year a new degree of freedom was introduced, of controlling the relative twist angle between two atomic sheets, and this resulted in some extraordinary outcomes. For instance, if one tunes two sheets of graphene to a very specific angle of 1.1 deg, named the magic angle, after some careful tuning, amazingly this new hybrid material becomes superconducting, losing its electrical resistance completely although its individual components do not show any hint of such behavior.

In parallel to the great progress made from the devices’ perspective, in the past few years some novel scanning probe techniques were introduced, ideal for studying 2D materials. In her talk, Halbertal will review the rapidly evolving field of 2D materials, and will focus on two novel techniques she uses in her research to study these materials: Scanning nano-SQUID (Superconducting Quantum Interference Device) magnetic and thermal microscopy (which she developed with colleagues during her PhD work at the Weizmann Institute of Science) and scattering type scanning nearfield optical microscopy (s-SNOM), which Halbertal uses today as part of her research at Columbia University.

David Hirsh
Columbia University

Working in Academia and Biotech

David Hirsh will compare experiences in academic research with those in a biotechnology companyl; he will describe starting and growing a biotechnology company in the 1980’s during the early days of recombinant DNA technology. Over a decade, the science grew from engineering microorganisms for commercial applications to developing biological molecules for treating human diseases. Hirsh will discuss several of Synergen’s projects, focusing on the development of the IL-1 receptor antagonist to treat inflammation and choosing the disease targets for the drug. Both academia and biotech offer the rewards of scientific discovery but differ significantly in the challenges of timelines, financial needs and endpoints.

John Morgan
Stony Brook University

The Unreasonable Effectiveness of Physics in Mathematics

In 1959 the physicist Eugene Wigner gave a talk entitled “The unreasonable effectiveness of mathematics in the natural sciences.” He stated, “The first point is that the enormous usefulness of mathematics in the natural sciences is something bordering on the mysterious and that there is no rational explanation for it.” (He proffered no convincing reason either.)

Morgan will briefly discuss this phenomenon and give a couple of the most important examples historically. But the primary focus of this talk is to describe how this flow of usefulness has been turned on its head in the last 35 years. In the 1940s mathematics and physics diverged with the introduction in physics of quantum field theory, a theory that even today has no rigorous mathematical definition nor foundation. Since then, advances in fundamental physics consist of an admixture of mathematical reasoning and physical intuition. In the 1970s and 1980s it was realized that, in fact, the two fields were studying, each for their own reasons, each from their own perspective, each in their own language, and each with their own tools, very closely related mathematical objects. Out of this realization came a new convergence of the fields that has driven, and continues to drive, rigorous mathematical progress in a wide swath of most advanced and exoteric aspects of geometry and topology. The interplay of the two subjects has produced mathematical questions and, even more surprisingly, mathematical statements that `must’ be true. Even though these questions and statements concern basic objects of geometry and topology that have been the focus of mathematical study for a long time, many are unlike any mathematical results that have been seen before. What accounts for this unreasonable effectiveness of the (non-mathematically rigorous) natural sciences in mathematics?

Krist E. Perks
Columbia University

Action in Perception: what Electric Fish Teach Us About How Motor Systems Engage Sensory Circuits

Much of the content of our perception is acquired through active motor sampling – touching, reaching, sniffing, looking, exploring. Traditional views of sensory processing that leave the motor system out entirely are, at best, incomplete. Perks’ research aims to understand how motor signals impinge on sensory processing by implementing electrophysiological and behavioral techniques in weakly electric mormyrid fish, a preeminent model system for such studies. Many animals can detect weak electric fields in the water – termed electroreception. An ancestral mode of electroreception — shared among elasmobranch fish, teleost fish, and monotremes – enables passive detection of low-frequency electrical signals in water. However, weakly electric mormyrid fish have evolved the additional ability to generate brief pulses of electricity, known as electric organ discharges (EODs), which enables active electrolocation and communication. Nearby objects cause local changes in EOD pulse shape. A specialized set of electroreceptors distributed across their skin encode these changes by the precise timing of neural spikes. A motor command drives EOD production via electrogenic muscles, but a copy of the motor command is also sent to primary electrosensory structures in the hindbrain. With Dr. Nathaniel Sawtell at Columbia University, Perks conducts experiments to determine how input from the motor system transforms sensory input even at the first stages of processing.

Antigoni Polychroniadou
J.P. Morgan AI Research

Privacy Preserving Predictions

While big data technology offers great promise, it introduces a challenge for privacy. Can we achieve the best of both worlds, extracting the benefits that big data offers while providing privacy to the data owner? In this talk, Polychroniadou will give an overview of privacy preserving computation and show how streaming data, machine learning and cryptography support powerful new ways of organizing predictive analytics.

Carlotta Ronda
Columbia University

A New Platform to Dissect the Interactions between Host and Microbiome

Over the past few years microbiome research has dramatically reshaped our understanding of human biology. It has become clear the crucial role that commensals play in mediating disease processes (i.e. inflammatory disease) and their strong associations with the onset of neurodevelopmental disorders, neurodegenerative diseases and alternation of brain chemistry (i.e. Alzheimer, with Parkinson’s disease, autism, and depression). However, the complex interactions between commensal microbes and host intestinal tissue networks remains still a challenge due to the inaccessibility of the system. Therefore, the ability to recreate and maintain the tissue architecture and diversity, yet allowing tight experimental control could help to dissect the molecular mechanisms that govern the interplay between host and microbiome. In order to unravel the complexity of this interaction, we have devised a new platform that aims to recapitulate the physiological state in the gut combining tissue engineering and microbial ecology.

Mijo Simunovic
Rockefeller University

Synthetic Human Embryology: Using Stem Cells to Model the First Steps in Human Development

The development of an embryo is one of the most dramatic examples of self-organization in all of biology. In the past decades, considerable progress has been made using animal models, which have revealed a highly complex and intricate gene network that regulates embryo patterning. However, how all these components work together to determine the shape of an embryo remains a mystery. Perhaps one of the most intriguing questions in developmental biology is how close relatives, such as the mouse and the human, despite nearly identical signaling inputs, develop into completely differently looking organisms. It is becoming increasingly clear that beyond biochemical signaling, there are physical and mechanical components that control early development. Thus, physics and engineering-based approaches are crucial to bring our knowledge of embryogenesis to a quantitative level. It is also of significant medical interest to reverse-engineer embryogenesis so to create ways of mimicking organ formation in vitro in a controlled and quantitative way. We combine stem cell biology and bioengineering to generate quantitative models of the human embryo. In particular, we used quantitative experimental modeling, live-cell microscopy, and CRISPR gene editing to reveal a molecular mechanism by which the human embryo breaks symmetry. Symmetry breaking events are crucial for the formation of the body axes and ultimately the animal’s body plan.

Xin Sun
Columbia University

How to Choose a Surface Uniformly at Random

One of the most fundamental facts in probability theory is that the large-scale limit of the simple random walk is Brownian motion. In this talk Sun will present a two-dimensional analog of this result where the random walk is replaced by random surface. The counter part of Brownian motion turns out to be related to 2D quantum gravity.

Lisa Tran
Columbia University

Shaping Nanoparticle Assemblies at Liquid Crystal Interfaces

Liquid crystals are ubiquitous in modern society. Whenever we text or check our emails, we are interacting with LCDs – liquid crystal displays. These materials are the basis of the modern display industry because of their unique properties. They can be manipulated with electric fields and are deformable because they are elastic fluids. These properties allow for liquid crystals to be engineered into a pixel. Despite these advances in their technological applications, the structures that liquid crystals can form are yet to be completely understood. Current research aims to elucidate these patterns to develop liquid crystals as biological sensors and as blue prints for assembling nanomaterials. Tran confines liquid crystals into geometries that can trigger these patterns and structures called defects – localized, “melted” areas of disorder that can lower the overall distortion in the system. She will present recent work in which defects are controlled with microfluidics to confine liquid crystals into spherical shells. Molecular configurations are controlled by varying the geometry of the system and the concentration of surfactants, soapy molecules with water-loving heads and oil-loving tails. Tran will then present ongoing experiments where nanoparticles are used in place of traditional surfactants to pattern them at the liquid crystal-water interface. This work has the potential to dynamically template nanomaterials for the enhancement of liquid crystal-based optical devices and sensors.

Yuri Tschinkel
Simons Foundation

Distribution of Primes

Professor Tschinkel will discuss basic problems and examples arising in the area of arithmetic statistics.

Robert Yang
Columbia University

Artificial Neuroscience

The nascent field of Artificial Neuroscience is the science of artificial neural systems. Yang will show that by training artificial neural networks and studying their properties, we can gain insights into biological neural systems.