Fig 2: Breaking CXCL16/CXCR6 retention enhances protection against distant lung tumor metastasis(A) qPCR for the CXCR6 ligand Cxcl16 shows highest expression in the tumor.(B) Cxcl16 mRNA is similar in the distant mucosa from tumor-bearing mice compared to the mucosa from tumor-naive mice. Two-way ANOVA with Kruskal-Wallis test, p < 0.005. Bars represent means ± SEMs and symbols represent individual mice.(C) Flow cytometry validates CXCL16 expression on the surface of 4T1 tumor cells but not mucosal epithelial cells.(D) Confocal microscopy characterizes CXCL16 expression within the 4T1 primary tumor.(E) Experimental design for sample collection in the presence or absence of anti-CXCL16 blocking.(F and G) Global TCR-ß repertoire sequencing analysis compares clonotype sharing between the tumor and the pre-metastatic lung in the presence or absence of anti-CXCL16 blocking. These data are represented in all frequency categories (F) and by zooming into the top 200 high frequency clones (G).(H) Experimental procedure to monitor spontaneous lung metastasis after surgical removal of the primary tumor in the presence or absence of anti-CXCL16 treatment.(I) Analysis of the primary tumors. Bars represent means ± SEMs and symbols represent individual mice. Paired t test, p < 0.05.(J) Analysis of the number of metastatic tumor nodules.(K) Representative images of metastatic tumor lungs.(L) Representative H&E staining of anti-CXCL16 and isotype-treated lungs showing the area of the lung harboring tumor metastases (dark purple).(M) Quantification of metastatic tumor occupancy in the lung, performed by a third party in a 1-sided blinded manner. Data were measured by diving the area of tumor metastases by total lung area and represented as a percentage. Paired t test, p < 0.05. Bars represent means ± SEMs and symbols represent individual lung tissue sections. Data represent 2 individual experiments combined.

Fig 3: CXCR6- TEff/EMs are the precursor of distant TRMs(A) Experimental design of sample collection for TCR-ß repertoire sequencing.(B) Repertoire analysis shows substantial overlap between high and low frequency TCRs between tumor and distant TRMs.(C and D) Repertoire analysis shows that both CXCR6NEG and CXCR6POS subsets share TCRs with tumor TRMs; only the highly expanded clones in the CXCR6NEG population contributed to distant TRM formation.(E) Experimental design for TEff/EM precursor transfer.(F) Preference of CXCR6+ TEff/EM precursor cells to stay in the tumor while CXCR6- TEff/EM precursors egress to the distant mucosa. Data are represented as a ratio of recovered cells in the tumor divided by the recovered cells in the distant mucosa or LN. Mann-Whitney test, p < 0.05. Bars represent means ± SEMs and symbols represent individual mice. Data are pooled over 3 independent experiments.