SDF-1α mediates wound-promoted tumor growth in a syngeneic orthotopic mouse model of breast cancer

Abstract

Increased growth of residual tumors in the proximity of acute surgical wounds has been reported; however, the mechanisms of wound-promoted tumor growth remain unknown. Here, we used a syngeneic, orthotopic mouse model of breast cancer to study mechanisms of wound-promoted tumor growth. Our results demonstrate that exposure of metastatic mouse breast cancer cells (4T1) to SDF-1α, which is increased in wound fluid, results in increased tumor growth. Both, wounding and exposure of 4T1 cells to SDF-1α not only increased tumor growth, but also tumor cell proliferation rate and stromal collagen deposition. Conversely, systemic inhibition of SDF-1α signaling with the small molecule AMD 3100 abolished the effect of wounding, and decreased cell proliferation, collagen deposition, and neoangiogenesis to the levels observed in control animals. Furthermore, using different mouse strains we could demonstrate that the effect of wounding on tumor growth and SDF-1α levels is host dependent and varies between mouse strains. Our results show that wound-promoted tumor growth is mediated by elevated SDF-1α levels and indicate that the effect of acute wounds on tumor growth depends on the predetermined wound response of the host background and its predetermined wound response.

Figure 1. SDF-1α is elevated in wound fluid and increases tumor growth.

A. SDF-1α levels in wound fluid from BALB/c wildtype (WT) mice were higher than in wound fluid from BALB/c nu/nu mice (nu/nu) or plasma from BALB/c WT animals (cytokine microarray). B. SDF-1α levels in wound fluid from BALB/c WT animals increased during the course of wound healing and were higher in WT wound fluid than in wound fluid from nu/nu animals 9d after wounding (insert). n = 5 samples/time point (0.3d to 5d) and n = 3 samples/time point (5d to 14d); Triangle: mean; filled circle: individual data point. Insert: p = 0.05, n = 3, Mann Whitney test). C, D. Inhibition of SDF-1α/CXCR4 signaling by AMD3100 treatment abolished wound-promoted-tumor growth. C. Experimental design. D. Cumulative tumor volumes. p = 0.0027, n = 10 animals/group, Kruskal Wallis Test/Dunn’s Multiple Comparison Test, observation time: 21d, mean ±95% CI.

Figure 2. SDF-1α increases tumor growth by directly acting on tumor cells.

A. Experimental design. B. Pretreatment of 4T1 cells with wound fluid derived from WT animals but not wound fluid derived from BALB/c nu/nu animals increased tumor growth in vivo as compared to pretreatment of 4T1 cells with plasma. mean ±95% CI, p<0.0001, n = 15 animals/group, ANOVA/Bonferroni’s Multiple Comparison Test, observation time: 18d. C. Pre-treatment of 4T1 cells with SDF-1α and plasma in vitro resulted in increased tumor growth in vivo as compared to pretreatment of 4T1 cells with plasma only. p = 0.0015, n = 15 animals/group, unpaired t-test, observation time: 22d, mean ±95% CI. D. Inhibition of SDF-1α/CXCR4 signaling with AMD3100 (AMD) during pre-treatment of 4T1 cells with wound fluid abolished increased tumor growth observed after pretreatment of 4T1 cells with wound fluid. p = 0.0016, n = 15 animals/group, ANOVA/Bonferroni’s Multiple Comparison Test, observation time: 22d, mean ±95% CI.

Figure 3. Wound-induced SDF-1α/CXCR4 signaling in tumor cells alters tumor cell proliferation, stromal composition and vascularization of tumors.

A, C. BALB/c mice were inoculated with 4T1 cells that were pre-treated for 5d with wound fluid or mouse plasma. B, D. BALB/c mice were inoculated with 4T1 cells and underwent wounding or sham surgery 9 days later. CXCR4 signaling was systemically inhibited by AMD3100. A, B. Mitotic figures in tumor sections. A. Unpaired t-test, p<0.0001, n = 10 specimens/group, observation time: 18 days. B. Bonferroni’s Multiple Comparison test, p<0.0001, n = 8–12 specimens/group, observation time: 28, mean ±95% CI. C, D. Top: Collagen staining with Picrosirius red. Bottom: CD34-positive blood vessels in tumors. Numbers in the lower left corner of images represent the density of CD34-positive structures/mm³. C. p = 0.0173, Mann-Whitney test. D. p = 0.0155, ANOVA, mean ±95% CI).

Figure 4. Wound-promoted tumor growth and SDF-1α levels in wound fluid are dependent on the host background.

A,B. The host background influenced wound-promoted tumor growth. A. Wound-promoted tumor growth was assessed in female animals of the F1 generation derived from BALB/c mice bred with BALB/c (control), FVB/nJ, AKR/J, C57Bl/6JNcr, DBA/2J animals. B. Cumulative tumor volumes. Unpaired t-test. Mean ±95% CI. Representative of 2 independent experiments is shown. C,D. The host background influences SDF-1α levels in wound fluid. C. SDF-1α levels were analyzed by ELISA in healthy hosts 2 d or 9d after subcutaneous implantation of PVA sponges. D. SDF-1α levels in wound fluid (ELISA). Mann-Whitney test. N.A. : sample number not sufficient to perform statistical analysis. Bar: mean.