Fig 2: The upregulation of CXCR4 by KLF8 leads to a feed-forward activation of FAK upstream of KLF8(A) Overexpression of KLF8 causes CXCL12/CXCR4-dependent activation of FAK. Uninduced (U) and induced (I) 10A-iK8 cells were serum-starved for 24 hours followed by 3-hour treatment with AMD3100 (35 ng/ml) or mock treatment prior to CXCL12 (100 ng/ml) stimulation. Whole cell lysates were prepared for western blotting for FAK phosphorylation at Y397 (pFAK) and total FAK. (B) Overexpression of KLF8 induces CXCL12/CXCR4-dependent expression of endogenous KLF8. The 10A-iK8 cells were grown and treated similarly as in A except for inclusion of PF223, a FAK-specific inhibitor. Total RNA was prepared for qRT-PCR (top panel) and semi-quantitative RT-PCT (bottom panel). (C) Hypothetic model for the feed-forward signaling loop of KLF8 to CXCR4/CXCL12 to pFAK and back to KLF8.

Fig 3: Stromal FAKi priming reduces fibroblast-driven collagen remodeling and KPC cancer cell invasion into organotypic matrices.(A to C) Schematic of 3D organotypic invasion assay: Fibroblast-driven collagen contraction (A), cancer cell seeding (B), and invasion into the contracted matrix toward the chemotactic gradient, created by air-liquid interface [note the culture medium meniscus (where cells remain in contact with media)] (C). (D) Schematic representation of fibroblast-mediated collagen matrix contraction upon vehicle or FAKi priming. (E and F) Quantification (E) and representative images (F) (scale bars, 5 mm) of matrix contraction over time. (G) Schematic of the FAK-FRET biosensor. (H and I) Quantification (H) and representative intensity-merged maps of ECFP fluorescence lifetime (I) in fibroblasts during matrix remodeling. Scale bar, 100 µm. (J and K) Representative maximum intensity SHG projections in fibroblast-contracted matrices (J) (scale bars, 100 µm) with quantification of SHG signal intensity (K) at peak (left) and over depth (in µm; right). (L) Schematic representation of organotypic invasion assay with late epithelial treatment (during KPC cell invasion), FAKi priming (during matrix contraction), and chronic treatment (during matrix contraction and invasion). (M and N) Quantification of KPC cell invasion (M) (normalized cell number at >200-µm depth) with representative images of pan-cytokeratin–stained KPC cells (N) (dotted line, 200-µm invasion depth; scale bars, 50 µm). n = 3 biological repeats, with three matrices per repeat and three FOVs per matrix (D to N). n = 20 cells per condition per replicate (H and I). Results: means ± SEM. P values were determined using an ordinary one-way ANOVA with Tukey correction for multiple comparisons. Unless otherwise stated, all significance is compared to vehicle. ns, P > 0.05; *P < 0.05, **P < 0.01, and ***P < 0.001.

Fig 4: A FAK signature reveals a graded response to stromal and epithelial FAKi priming before gemcitabine/Abraxane.(A) Schematic representation of PDAC PDX, PDCL, and PDX TMA establishment. (B) Representative images of TKCC05 and TKCC10 PDX TMA samples stained with total and pTyr397-FAK. Scale bars, 50 µm; n = 3 cores per PDX. (C) Quantification of pTyr397-FAK levels in PDCLs [mass spectrometry, top; data obtained from (67)] and confirmation via Western blot (bottom), highlighting TKCC05 (orange; low active, high total FAK) and TKCC10 (purple; high active, low total FAK) cells. (D to G) Representative images (D and F) and quantification (E and G) (normalized cell number at >200-µm depth) of green fluorescent protein (GFP)–stained TKCC05 cells (D and E) and pan-cytokeratin–stained TKCC10 cells (F and G) invading into organotypic matrices primed with vehicle or FAKi (dotted line, 200-µm invasion depth). Scale bars, 50 µm. n = 3 biological repeats, 3 matrices per repeat, 3 FOVs per matrix. Results: means ± SEM. P values were determined using an ordinary one-way ANOVA with Tukey correction for multiple comparisons. Unless otherwise stated, all significance is compared to vehicle. (H) Timeline for TKCC05 and TKCC10 orthotopic studies. IVIS, In Vivo Imaging System. (I to L) Kaplan-Meier analysis of survival (I and K) and time to metastasis (J and L) in mice with TKCC05 (I and J) or TKCC10 (K and L) orthotopic tumors treated with vehicle/saline (TKCC05, n = 10; TKCC10, n = 8), vehicle/gemcitabine/Abraxane (TKCC05, n = 9; TKCC10, n = 8), FAKi/saline (TKCC05, n = 8; TKCC10, n = 7), or FAKi/gemcitabine/Abraxane (TKCC05, n = 8; TKCC10, n = 8 animals). (M) Percentage of TKCC10 mice with visible liver metastases at endpoint. Kaplan-Meier curves were compared using log-rank Mantel-Cox test. ns, P > 0.05; *P < 0.05, **P < 0.01, and ***P < 0.001.

Fig 5: Cortactin tyrosine to phenylalanine (3YF) mutation inhibits FAK-mediated cell migration and FA turnover.FAK-/- MEFs reconstituted with GFP-FAK WT, GFP-FAK R454, or GFP-FAK A712/713 were co-transfected with RFP-cortactin WT, or RFP-cortactin Y421, Y466, and Y482 mutated to phenylalanine (3YF) or mutated to glutamic acid (3YE). (A) In MEFs over-expressing the indicated RFP-cortactin constructs, FA lifetime was determined by counting the number of sequential frames with GFP-FAK FA-associated fluorescence above background. Data from live-cell spinning disc confocal microscopy after growth media supplemented with 50 ng/ml EGF stimulation (2 min intervals from a 60 min time-lapse) represents the mean lifetime of at least 50 leading edge-associated FAs ± SEM from at least 12 different cells within the indicated GFP-FAK reconstituted FAK-/- MEFs (***p<0.001). (B) Cell migration speed in µm/sec was determined by cell tracking (n = 12 per group, values are means ±SD, ***p<0.001).