Learning how immune cells communicate

Efforts in specific lab-based projects and larger consortia like the Human Cell Atlas or the first phase of immgenT mostly focus on cataloging all cell types in the human body. But we are not single-cell organisms! Understanding immune responses, tissue organization, and bodily functions requires understanding how cells communicate.

What are the different modes of communication? Are there independent channels? New channels? Can all cells talk to each other? We believe that a systematic analysis of cellular interactions is creating a new field in biology, with an impact on diagnosis and treatment. Many current medications already target cellular interactions — immune checkpoint blockade, biotherapies in autoimmunity — but none directly measure their impact on cell communication. Methods are largely lacking.

In the lab, we invent and use new methods from computational biology (modeling), systems biology, and synthetic biology to characterize and modulate cell communication in vivo and in situ in human tissues. Because T cells are fundamentally important, we dissect their network of cellular interactions across tissues and diseases.

We combine systems biology technologies to extract the maximum amount of information from human tissues. Single-cell RNA-seq remains a core tool for defining the cellular composition of lesions. Building on this, the lab has established spatial biology pipelines — single-cell spatial transcriptomics and multiplexed immunofluorescence — that let us measure cell–cell interactions directly in tissues.

These multimodal datasets (molecules, cells, and their interactions) enable us to construct mechanistic models of disease, which we then test in mouse models. Whenever possible, we translate these insights back into patient care, for example by informing diagnostics.

Inflammatory bowel disease

Patients with medically refractory ulcerative colitis often undergo total colectomy with an ileal pouch–anal anastomosis — a J-pouch — which restores intestinal continuity. In a substantial fraction of patients the pouch later fails, and there is no good way to predict who is at risk. In the lab’s first paper, we went back to the colon removed at surgery and combined careful histopathologic evaluation with single-cell spatial transcriptomics, asking whether the original disease already carries the signature of what happens next. We found cellular and molecular abnormalities in those specimens that track with subsequent pouch failure, pointing both to candidate mechanisms and to features a pathologist could eventually report.

Olivas, Ngai, Schahrer et al., Cell Mol Gastroenterol Hepatol 2025

Immune checkpoint blockade-induced colitis

Immune checkpoint inhibitors (ICI) have revolutionized cancer treatment. But they also trigger autoimmunity in up to 40% of patients, which can be fatal or force the discontinuation of a potentially curative treatment. Although the cause of the disease is clear — drug-induced — the exact process leading to the pathology is still unknown.

ICI-induced autoimmunity is both a significant clinical challenge and a rare opportunity: a chance to watch the onset of autoimmunity in humans, on a known trigger and a known clock. We are investigating why Tregs aren’t sufficient to prevent ICI colitis, and ways to manipulate them to prevent autoimmunity while maintaining anti-tumor immunity — a major objective of tumor immunotherapy.

Kawasaki disease

An acute vasculitis of childhood that can damage the coronary arteries. Working with the Arditi and Noval-Rivas labs, we use single-cell and spatial genomics in human heart tissue and a mouse model to dissect the immune interactions that drive the vascular lesion, implicating NLRP3 inflammasome signaling, autophagy-mitophagy, or STING pathways as intervention points.