Fig. 6: Brca2-deficient cancer cells shift CAF functions.

(a) Immortalized PSCs were seeded in Matrigel for 4 days. CM from PSCs or from KPC cells in which Brca2 was silenced by shRNA or nontargeting shControl (KPC) was then added for an additional 4 days. 3 nM of aglaroxin C or PBS control was added to the conditioned media. Pdl1 was measured by qRT-PCR. Statistical analysis was conducted via two way ANOVA. Data are presented as mean \(\pm\)SEM. For each condition n = 3-9 biologically independent culture domes combined from 3 independent experiments. (b–c) CD8⁺ T-cells were isolated from spleens of naïve C57BL/6 J mice and activated in the presence of CAFs isolated from either KPC-shControl or KPC-shBrca2 tumors. Subsequently, T-cells were subjected to FACS analysis for surface expression of CD69 and CD25. Statistical analysis was conducted via unpaired two tailed t test. Data quantified are presented as mean \(\pm\)SEM of 6 KPC-shControl and 6 KPC-shBrca2 mice. (d) Scores of ligand-receptor binding were calculated using the ICELLNET R package (see Methods) to predict potential differential interactions between ligands of CAFs derived from shControl vs shBrca2 tumors with immune checkpoint receptors on CD8⁺ T-cells. (e–g) CAFs were isolated from KPC-shControl (n = 3 mice) and shBrca2 (n = 3 mice) and stained with anti-CD155 (e), anti-Nectin2 (f), and anti-PD-L1 (g). Statistical significance was assessed by two tailed t-test. Data are presented as mean ± SEM (h-i) PSCs were treated with CM derived from KPC-shBrca2 or KPC-shControl organoids, or with growth medium as control for 4 days. Then, cultures were stained with Sirius red (SR) to assess collagen deposition (see Methods). Each point represents the average of shBrca2 or shControl normalized to the average of the growth medium control in each experiment. n = 3 independent experiments, each representing an average of 5 independent culture wells. Two tailed t-test was performed on normalized values. (i) Representative images of PSCs stained with SR following 4 days treatment with CM derived from KPC-shBrca2 or KPC-shControl organoids. Scale bar, 300 μm. (j–m) FFPE tumor sections from BRCA-mut and BRCA-WT PDAC patients were stained by double staining for αSMA and CLU and imaged using Second harmonic generation (SHG) imaging. (j) Representative images are shown. DRAQ5 was used to stain nuclei. Scale bar, 100 μm, or 25 μm (inset). (k–m) Quantification of matrix pattern using the TWOMBLI plug-in (see Methods) (n = 3 BRCA-WT and n = 3 BRCA-mut patients). The following parameters were analyzed: (k) Alignment—the extent to which fibers within the field of view are oriented in a similar direction; (l) Curvature- the mean change in angle moving incrementally along 40 µm mask fibers; and (m) Branchpoint—the number of intersections of mask fibers in the image. Statistical analysis was conducted via unpaired two tailed t test. Data are presented as mean ± SEM (n) Schematic representation of the proposed model. Secreted factors from BRCA-mutated cancer cells induce HSF1 activation in a subset of adjacent PSCs leading to their transcriptional rewiring into immune-regulatory CLU⁺ CAFs. Source data are provided as a Source Data file.