Projects

Exposure to stress during pregnancy can increase disease susceptibility in offspring, with effects that persist across generations. Using single-cell transcriptomics, multiplexed proteomics, and stereotaxic brain interventions in mice, I investigate how maternal stress reshapes immune function and brain development, and how these changes propagate to the next generation. The goal is to identify points of intervention that can reduce, if not reverse, the long-term health consequences of prenatal stress. This work is part of the Biology of Adversity Project at the Broad Institute. Figures : Top: developmental staging in mouse versus human. Middle: stereotaxic ablation of stress centers in the mouse brain, confirmed by reduced cortisol levels. Bottom: single-cell transcriptomic clustering of mouse brain cells.

MicroRNAs are small non-coding RNAs that regulate gene expression post-transcriptionally. Their discovery earned Gary Ruvkun and Victor Ambros the 2024 Nobel Prize in Physiology or Medicine. Ruvkun, at this institution, has spent decades unraveling how these molecules control development and disease. miR-92a, from the oncogenic miR-17-92 cluster, promotes colorectal tumor growth by suppressing pro-apoptotic and anti-proliferative targets, and modulates inflammatory signaling in the gut. Using knockout mouse models and high-dimensional immune profiling of the lamina propria, I investigated how its loss reshapes immune composition during colitis. In a separate collaborative project, we show that Type I interferon limits CNS autoimmunity through the miR-21–FOXO1 axis in pathogenic Th17 cells, providing a mechanistic basis for IFN-β therapy in Multiple Sclerosis. Figures: Top: RISC-mediated mRNA degradation and translational repression by miRNAs. Bottom: Reduced tumor susceptibility in miR-92a KO mice

Chemical synchronization of cells, the standard approach to cell cycle studies, perturbs the very biology under investigation. I developed a fully automated image analysis pipeline to stage individual cells in the cell cycle from high-resolution fluorescence microscopy images, eliminating this confound entirely. Combining DNA FISH, RNA FISH, and immunofluorescence, the pipeline enabled single-allele-resolution studies of gene expression and DNA damage responses. One key finding: the γH2AX peak in S phase after UV irradiation marks sites of active DNA replication, not the extent of damage itself, with direct implications for how DNA repair is interpreted in cycling cells. Figures: Top: Image-based cell cycle staging against DNA content by flow cytometry. Bottom: Cell Cycle-dependent response to UV-induced DNA damage.

As cells differentiate, genes reposition relative to nuclear compartments, including the lamina, heterochromatin, and interchromatin regions, in patterns that define cell identity and regulate expression. Whether these positional relationships hold during cell cycle progression, when chromatin undergoes large-scale reorganization, was less clear. Using DNA and RNA FISH at single-allele resolution, I showed that key genes are transcribed independently of their position relative to the nuclear lamina across cell cycle phases, challenging a simple repression model. In related collaborative work, CTCF-mediated genome architecture was shown to govern dosage of monoallelic autosomal gene expression. I extended these themes at the Children's Hospital of Philadelphia, using Hi-C to map cell cycle-dependent changes in three-dimensional chromosome organization. Figures: Top: Topological model of gene regulation. Bottom: Completely-automated, high-throughput image analysis pipeline for studying nuclear architecture-based gene regulation at single-allele resoltion.