Unless otherwise noted, lectures are held the second Tuesday of each month at 4:00 PM in 341 Bardeen.

September 9th - Imaging Interkinetic Nuclear Migration and Neurogenesis in the Retina

Brian Link
Department of Cell Biology, Neurobiology and Anatomy
Medical College of Wisconsin

During retinal development, neuroepithelial progenitor cells divide in either a symmetric proliferative mode, where both daughter cells remain mitotic, or in a neurogenic mode, where at least one daughter cell exits the cell cycle and differentiates as a neuron. While the cellular mechanisms of neurogenesis remain unknown, heterogeneity in cell behaviors has been postulated to influence this cell fate. In this talk I will present studies addressing interkinetic nuclear migration, the apical-basal movement of nuclei in phase with the cell cycle, and the relationship of this cell behavior to neurogenesis. Using time-lapse imaging in zebrafish we found that various parameters of interkinetic nuclear migration were significantly heterogeneous among retinal neuroepithelial cells. We found that specific patterns of INKM precede a terminal mitosis. Finally, I will present recent findings on the signaling molecules that influence the relationship between nuclear position and neurogenesis. Overall, this data supports a novel model for neurogenesis where interkinetic nuclear migration differentially positions nuclei in neuroepithelial cells and therefore influences selection of progenitors for cell cycle exit based on apical-basal polarized signals.

October 7th - Extracting Quantitative Information from Biological Images
Note: Special Day

Anne Carpenter
Imaging Platform
The Broad Institute

Microscopy images contain abundant information about the properties of cells, organisms, or materials, but are rarely mined to their full potential. Automated image analysis can potentially produce rich, reproducible, quantitative results for hundreds of thousands of samples in an experiment. We are developing new methods for image analysis and data mining and releasing them to the scientific community via the CellProfiler open-source software project (www.cellprofiler.org).

Using CellProfiler’s flexible modules, researchers have set up automated image analysis pipelines to identify and measure the properties of a wide variety of biological “objects”, including mammalian, yeast, and Drosophila cells in culture, C. elegans worms, tumors in mice, yeast colonies, and yeast growth patches on agar. Multiple quantitative features of each object and each object’s sub-compartments are extracted, including size, shape, and the intensity and texture (smoothness) of each color channel in the original image.

Because the human visual system actually uses a combination of features to identify objects with complex or subtle appearances, we have created a tool (CellProfiler Analyst) to enable supervised machine learning based on multiple features of each object. In this approach, a researcher spends a few hours training a computer to recognize the objects of interest in images. Machine learning algorithms then distinguish objects of interest based on the rich set of hundreds of CellProfiler-measured features. We have used this approach to readily score dozens of complex phenotypes in images automatically and quantitatively.

November 13th - Mechanisms of Membrane Remodeling
Note: Thursday Lecture (Room 341 Bardeen at 4:00PM)

Jon Audhya
Department of Biomolecular Chemistry
University of Wisconsin-Madison

All eukaryotic cells contain an elaborate membrane system necessary for the transport and compartmentalization of various proteins and lipids. This architecture permits numerous biochemical and signaling processes to occur simultaneously within specialized organelles. While the core machinery necessary to direct vesicle movement has been largely defined, the regulatory mechanisms that modulate membrane trafficking remain poorly understood. In particular, we are interested in determining how the fates of membrane-associated proteins are regulated by developmental cues. Failure to respond efficiently to such signals can result in a variety of disease states including cancer, neurodegeneration, and diabetes. By combining high-resolution fluorescence microscopy, functional genomics approaches, and in vitro biochemistry, we use the nematode Caenorhabditis elegans to identify critical components necessary for membrane reorganization during development.

December 9th - Super-Resolution Microscopy In Vitro and In Vivo by Structured Illumination

Mats Gustafsson
Janelia Farm Research Campus
Howard Hughes Medical Institute

Periodically structured illumination light can extend the resolution of fluorescence microscopy beyond the classical limit through spatial frequency mixing. The amount of resolution extension, set by the spatial frequency of the illumination pattern, is normally about a factor of two, because the pattern frequency is limited by the diffraction in the same way as the conventional resolution.
Dramatically greater resolution extension is possible, however, if a nonlinearity can be introduced between the incoming illumination intensity and the outgoing emission rate, because such a nonlinearity can create harmonics of the illumination frequency. Reversible photo-switching of fluorophores constitutes one promising form of such nonlinearity.

Structured-illumination microscopy typically uses data reconstruction algorithms that assume that the entire data set represents a single unchanging structure. It has therefore been largely confined to fixed, unmoving samples. If a data set can be acquired in a time that is done short compared to sample motions, however, live imaging becomes possible. The time required per data set naturally scales with the number of axial planes required, and thus with the sample thickness. At the thin end of the thickness range lies TIRF microscopy, where the emitting region is thin enough to be considered 2D; there live imaging is possible with ~100 nm lateral resolution at multi-Hz frame rates for several hundred time points.

February 10th - Feeling Green: Visualizing Mechanical Signaling in Plants

Simon Gilroy
Department of Botany
University of Wisconsin-Madison

Due to the sessile nature of their lifestyle, plants have to respond to a wide range of signals to entrain their growth to the prevailing environmental situation. This can be seen in the sculpting of a tree to the prevailing wind or the tracking of a leaf as it follows the path of the sun across the sky. Plant responses to such environmental cues are largely expressed as changes in growth patterns, which in turn represent the coordination of growth across fields of cells in each organ. Plant cell growth itself represents a highly regulated balance between internal hydrostatic (turgor) pressure driving enlargement, and the modulation of the rigidity of the cell wall resisting this force. Turgor forces are large, 2-50 atmospheres being typical. Such forces allow plant organs to, for example, crack concrete as they grow. A precise and dynamic regulation of the properties of the cell wall, coordinated over the surface of each plant organ, is therefore required to restrain growth to the appropriate spatial and temporal patterns for the production of environmentally responsive developmental patterns. However, we still have a remarkably poor understanding of the cell and molecular control systems that regulate the direction and extent of cell growth. We have developed a range of fluorescence imaging approaches to monitor the dynamics of cytoplasmic and cell wall regulators, such as reactive oxygen species and protons. Using the growth of the root of Arabidopsis thaliana as our model, our analysis is revealing a highly dynamic and complex regulatory network involving rapid, fluctuating patterns of pH, chemical oxidation and classic signaling molecules, such as the calcium ion. In combination with molecular genetic approaches, we are beginning to piece together how these regulatory elements may be imposing control over the spatial and temporal components of organ system development.

March 3rd - In Situ, Image-Correlated Stable Isotope Analysis with the Wisc-SIMS Ion Microprobe
Note: Special Day

John Valley
Department of Geology and Geophysics
University of Wisconsin-Madison

The Wisc-SIMS Lab makes in situ isotope and trace element measurements from micron-scale spots correlated to imaging. The new CAMECA ims-1280 ion microprobe is a large radius, multicollector, secondary-ion mass-spectrometer that is optimized for high accuracy and precision of isotope ratios. Recent applications include: carbon isotope analysis of single bacteria; vital effects for oxygen isotope fractionation in carbonate tests of foraminifera; seasonal records of paleoclimate in stalagmites; Li isotope ratios in zircon that suggest oceans and weathering on the surface of the early Earth, 4.3 billion years ago; and analysis of samples from the comet Wild 2 collected by NASA’s Stardust Mission. Possible applications include non-radioactive isotope labels at the scale of single cells in plant or animal tissue.

April 14th - Insights into the Structure and Function of Type V Intermediate Filaments: Nuclear Lamins in Health and Disease

Robert Goldman
Department of Cell and Molecular Biology
Northwestern University

The nuclear lamins are the members of the intermediate filament family of proteins that are targeted to the nucleus. The lamins assemble into a variety of higher order structures both in the peripheral lamina and in the nucleoplasm of mammalian nuclei. Within these two nuclear compartments the lamins play key roles in DNA replication, chromatin organization, transcription, and in the mechanical properties of the nucleus. High resolution microscopic studies have revealed that A- and B-type lamins (LA/C, LB1 and LB2) form separate, but overlapping, meshworks in the nuclear lamina. Within the nucleoplasm, fluorescence correlation spectroscopy (FCS) reveals that lamins A and C are highly dynamic and appear to interact transiently with the more stable B type lamins. We are characterizing the role of these the lamin networks by silencing lamin expression using shRNA, and have found that silencing lamin B1 induces blebs at the nuclear surface. These blebs contain mainly A-type lamins and are devoid of B1 and B2-type lamins. Importantly, these blebs resemble those seen in cells from patients with various laminopathies, including Hutchinson-Gilford Progeria Syndrome (HGPS) and various muscular dystrophies. Studying both knockdown cells and cells obtained from progeria patients, we have found that blebs are euchromatin rich and contain high levels of activated RNA polymerase II (pol IIo) and acetylated histones. However, they are devoid of centromeres and trimethylated histone H4K20 (H4K20me3). As determined by “double replication labeling” experiments, these blebs contain exclusively early replicating or gene rich DNA. This finding is corroborated by chromosome painting analyses which reveal that primarily gene-rich chromosomes reside in the blebs. Using microlaser dissection we have been able to amplify the DNA within these blebs and carry out comparative genomic hybridization (CGH) analyses, and the results confirm the presence of gene rich chromosome regions within blebs. Thus, the lamin A rich blebs which form in nuclei of progeria patients, and in normal cells silenced for lamin B1 expression, are very similar and contain mainly euchromatin. The presence of euchromatin within the blebs demonstrates that chromatin organization is significantly altered, as gene rich chromosome domains are normally found in the interior of the nucleus, away from the lamina. The results also suggest that lamin B1 is primarily involved in anchoring and organizing heterochromatin in the lamina region, while lamin A plays a role regulating transcription. Taken together, our findings provide additional support to the hypothesis that the lamin A mutations causing the premature aging phenotypes of progeria patients are related to epigenetic defects. These data also shed important new light on the normal role of lamins in regulating nuclear architecture and function.

May 12th - Visualizing Neural Circuits in the Fly Brain

Julie Simpson
Janelia Farm Research Campus
Howard Hughes Medical Institute

The central nervous system of the adult fruit fly has approximately 100,000 neurons. Genetic tools allow us to reproducibly target different subsets of these neurons to assay the behavioral effects of altering their activity. We also drive fluorescent reporters in these neuronal subsets to visualize how these neurons project in order to assemble circuit maps. Primarily we use confocal microscopy on fixed whole-mount brains to explore the neural trajectories. We have participated in the development of an automatic alignment system that allows the expression patterns in many different brains to be compared on a common coordinate system; I will describe some results from this approach. Many of the expression patterns are complicated, with large numbers of neurons projecting in different and overlapping ways. I will discuss our attempts to adapt the Brainbow technique from mice to flies in order to visualize many different neurons simultaneously in different colors. Ultimately we hope to identify how neurons are synaptically connected to build behaviorally functional circuits; some progress in this endeavor can be achieved by showing which parts of neurons are input zones (dendrites or post-synaptic densities) and which parts are outputs (synapses). We have done some preliminary work in the central complex region using this approach. Many challenges for circuit mapping remain and I look forward to brainstorming about how to meet them.