A retrievable immune scaffold reveals a 5–7 week presymptomatic window for tracking type 1 diabetes progression, long before blood sugar tests turn positive.
Study: Longitudinal monitoring of type 1 diabetes progression to disease onset. Image credit: Halfpoint/Shutterstock.com
A recent study published in Science Advances described the application of subcutaneous microporous scaffolds, viz., immunological niches (INs), for longitudinal immune monitoring of the progression of type 1 diabetes (T1D) before clinical onset in a preclinical mouse model.
Current markers fail to predict disease timing
T1D is a chronic autoimmune disease that leads to the destruction of pancreatic β cells, resulting in hyperglycemia. Current treatment requires exogenous insulin every day, which carries a heavy burden and does not mitigate long-term risks. Advances in cell replacement therapy have enabled pluripotent stem cell (PSC)-derived islet transplants and allogeneic islet transplants from deceased donors.
However, both require immunosuppression and carry the risk of the same immune cell–targeting that destroys native cells. Therefore, therapies to delay or prevent T1D onset are necessary to preserve function without the risk of recurrent autoimmunity. While autoantibodies that indicate an autoimmune response to native β cells and the risk of T1D progression have been identified, they do not denote proximity in time to symptom onset or enable continuous tracking of disease evolution.
Implantable immune niches act as pancreatic surrogates
INs consist of a subcutaneously implanted microporous poly(ε-caprolactone) scaffold. They have been demonstrated to identify transcriptomic changes in cancer, multiple sclerosis, and T1D by capturing phenotypic changes of immune cells in the IN over the course of a disease. The IN serves as a retrievable surrogate immune niche for disease monitoring in a native tissue that is too fragile or inaccessible for biopsy, reflecting systemic immune changes occurring in the pancreas rather than directly sampling β cells and enabling repeated, minimally invasive sampling over time.
Gene signatures distinguish risk, progression, and timing
In the present study, researchers investigated the potential of INs to delineate risk status, progression status, and the presymptomatic time frame of T1D progression through repeated longitudinal sampling. First, they investigated disease incidence using serial IN explants replaced at weekly intervals in non-obese diabetic (NOD) and non-obese diabetic–resistant (NOR) mice. INs were subcutaneously implanted at four weeks of age, and weekly explants began at six weeks of age until hyperglycemia in NOD mice and 30 weeks of age in NOR mice.
The NOR group did not develop hyperglycemia, but nearly 50% of the NOD group did (henceforth referred to as NOD progressors). Further, there were no significant differences in blood glucose levels between NOR mice and NOD non-progressors (which did not develop hyperglycemia). Samples from the three groups were subject to bulk RNA sequencing at different time points to assess immune dynamics over the disease course.
Of over 23,000 sequenced genes, 4,593 met the study thresholds for variance and average count, and were used in subsequent analyses. The researchers investigated whether healthy mice (NOR) could be distinguished from mice at risk of diabetes (NOD). Using elastic net regression, they identified a reproducible gene expression signature that distinguished NOD from NOR mice at all time points.
Next, the team assessed whether mice that progressed to diabetes could be distinguished from those that remained normoglycemic. Elastic net regression on week 6 samples from NOD progressors and non-progressors revealed a 13-gene signature separating the two groups. T1D progression in current clinical practice is identified by dysglycemia, indicated by a glucose tolerance test (GTT).
For NOD mice, an intraperitoneal GTT can be performed at approximately 14 weeks of age to detect dysglycemia. The researchers conducted an intraperitoneal GTT at eight weeks and found no differences between the two groups. This suggested that the gene signature identified disease risk well before measurable glucose intolerance, preceding the equivalent clinical standard in mice. In addition, an eight-gene signature was identified that separated week 6 progressor samples from all non-progressor samples across multiple time points.
Further, six of the eight genes were identified in a published single-cell RNA sequencing (scRNA-seq) dataset of NOD islets. This meant that genes with changing pancreatic expression levels also showed changes in the IN, indicating that the IN could reflect disease-associated immune activity in the primary tissue over time. Next, the researchers investigated temporal changes in mice that progress to diabetes.
Elastic net regression performed on all time points and normalized to week 6 revealed a transition from early to late time points, with samples from symptom onset and one (−1) and three (−3) weeks before hyperglycemia onset clustering together. Samples from five weeks before onset (−5) spanned between samples from seven weeks before onset (−7) and later time points (−3 through disease onset), suggesting that this time frame (−7 to −5) marked a transition period in disease progression rather than a single discrete event.
Further, elastic net regression on −7 and −5 time points identified a 13-gene signature separating the two groups. Using the same scRNA-seq dataset, 11 genes were found to be expressed in NOD islets. Finally, the team used this gene signature to develop a scoring system for all samples using unsupervised singular value decomposition and supervised random forests. The score for the −7 to −5 transition separated the −7 and early presymptomatic samples from later time points in a continuous progression rather than a binary split.
Immune monitoring may redefine early T1D detection
Taken together, IN transcriptomic analysis can stratify at-risk and non-risk groups, progressors and non-progressors, and identify a presymptomatic transition approximately five to seven weeks before disease onset in the NOD model. These findings provide a basis for a proof-of-concept, subcutaneous immune monitoring strategy for T1D that currently relies on functional loss for diagnosis, highlighting the potential to dynamically track autoimmune activity before irreversible β-cell damage occurs.
Identification of individuals who progress to symptomatic disease well before dysglycemia emerges would allow for the administration of preventive therapies to preserve cellular function, pending further validation and translation beyond preclinical models.
Nearly half of adults now live with heart disease as obesity and diabetes surge