Figure 2
Neuroendocrine pancreas in mouse development. (a) Pancreata from Wnt1-Cre x Rosa26-mT/mG reporter lines at E15 (left) and P15 (right) were harvested, processed using CLARITY, and stained for insulin. 3D histology enables organ-wide visualization of developing islets (white) and associated neural features (green) throughout the pancreas; imaging datasets were then automatically processed with the computational pipeline (see Fig. S2). Int: intestine, Panc: pancreas, Sp: spleen. All samples were embedded with A1B1P4 hydrogel and cleared for 1–2 weeks at 37 °C. Scale bars = 1000 μm. (b) Automated quantification reveals changes in islet size and number at various stages of pancreatic development. Note massive expansion of islet number shortly following birth (p < 0.05, unpaired t-test) through P6, followed by reduction. (c) Islet size is also non-monotonic with aggregate size reduction (P6–P15, p < 0.05, unpaired t-test) and growth (P15–42, p < 0.0005, unpaired, t-test). (d) CLARITY enables refined whole-population measurements, as illustrated by a population shift towards larger islets at P42 compared to other timepoints. (e) The presence of neural crest structures in the neighborhood surrounding and within pancreatic islets was quantified. The level of neural crest-islet interaction is dynamic with a significant decrease between P2 and P6 (p < 0.0005, unpaired t-test) and a significant increase between P15 and P42 (p < 0.005, unpaired t-test). (b–e) n = 3 pancreata from litter mates per developmental stage, error is s.e.m. (f) Optical 2D sections (5 μm thick; top row) and corresponding 3D CLARITY images (1000 μm thick, bottom row) across development. While 2D sections suffice to observe gross changes in islet size, neural structures (arrowheads) are sparse and difficult to quantify in 2D. CLARITY reveals high-resolution structure not captured by 2D images, enabling visualization of neural remodeling dynamics. Scale bars = 100 μm.
