In a sprawling five-year collaboration, scientists at the University of Florida Diabetes Institute and international partners have built a high-resolution map of the human pancreas, profiling roughly 16 million cells across thousands of islets. The images capture the immune system swarming insulin-producing beta cells well before people show clinical signs of diabetes, offering a rare window into the earliest stages of Type 1 disease. Researchers say the sheer scale and spatial detail could change how prevention trials and therapies are designed.
The team analyzed pancreatic tissue from 88 organ donors, including single and multiple islet autoantibody-positive individuals, recent-onset cases and people with long-standing disease. In all, they profiled about 12,000 islets and 16 million single cells using two 45-plex antibody panels (79 markers total), according to a preprint posted on bioRxiv. The dataset ranks among the largest human spatial-proteomic surveys of Type 1 diabetes to date. The tissue came from the UF-based Network for Pancreatic Organ Donors with Diabetes (nPOD), which supplies donor pancreases to many labs.
As reported by the Tampa Free Press, UF researcher Mark Atkinson said the work helps “identify the inner pieces of that puzzle of how and why Type 1 diabetes develops.” The outlet also noted the story was described as published in Nature Metabolism on July 2.
What the images revealed
The spatial maps showed pro-inflammatory myeloid cells, including macrophages and mature dendritic cells, parked along islet edges and right next to PD1+ memory CD4 and exhausted-like CD8 T cells. According to the team’s preprint, the patterns line up with the idea that macrophages act as early first responders that call in adaptive immune cells to go after beta cells.
The authors also flagged early beta-cell shifts that show up before outright insulin loss, including reduced islet amyloid polypeptide (IAPP) and increased MHC-I. Those changes hint at a beta cell under mounting stress long before a person would land in a clinic with obvious symptoms.
How this is different
Older tissue studies typically tracked only a few proteins at a time. Imaging mass cytometry, by contrast, can read dozens of markers in place, letting researchers see which cell types sit next to each other and how they might interact at single-cell resolution, as described in Nature Methods. Co-author Maigan A. Brusko told the Tampa Free Press, “It’s like seeing the other drivers on the road, but also where they’ve been and where they’re going,” a leap in precision that could expose new therapeutic targets.
Implications and next steps
The authors say this atlas should help pinpoint cell states and neighborhoods that are potentially druggable and trackable in prevention trials, and that nPOD and collaborating teams plan to share the maps as a community resource. Related multimodal efforts that integrate histopathology with spatial transcriptomics have already nudged researchers to rethink early events in Type 1 diabetes, underscoring how large spatial datasets are moving to the center of the field; see recent integrated pancreas histopathology coverage for context.
The work does not offer a ready-made therapy and cannot replace longitudinal sampling of living patients, but it does hand researchers a much clearer target map for interventions aimed at the macrophage-T-cell axis or at shoring up beta-cell resilience before symptoms appear. For clinicians and families watching closely for earlier detection or prevention strategies, the new atlas is a hopeful, if early, step forward.
