Growth-inhibitory effects of CD40 ligand (CD154) and its endogenous expression in human breast cancer

CD40 binding produces multifaceted growth signals in normal and malignant B cells, whereas its physiological role is less well characterized in epithelial cancers. We examined the growth outcome of CD40 ligation in human breast cancer cells, using CD40+ (T47D and BT-20) and CD40-negative (MCF-7, ZR-75-1) cell lines as defined by flow cytometric analysis, immunohistochemistry, and reverse transcription-PCR. Treatment with the soluble recombinant CD40 ligand (CD40L) molecules gp39 or CD40L-trimer significantly reduced [3H]thymidine uptake in BT-20 and T47D cells by up to 40%, but did not affect the growth of CD40-negative MCF-7 or ZR-75-1 cells. Similarly, significant growth inhibition was observed after co-incubation with CD40L-transfected murine L cells (55.0 +/- 8.9%, P < 0.001) that express membrane CD40L constitutively, or with paraformaldehyde-fixed, CD3+ CD40L+ PBLs from three different HLA-mismatched donors (39.7 +/- 3.7%, P < 0.01). Untransfected L cells and non-CD40L-expressing lymphocytes did not produce significant growth inhibition. The in vivo antitumorigenic effects of CD40L were examined using a s.c. severe combined immunodeficient-hu xenograft model. Pretreatment with two different soluble recombinant CD40L constructs (CD40L and gp39) produced similar xenograft growth-inhibitory effects [67 +/- 24% (n = 4), and 65 +/- 14% (n = 8) inhibition, respectively], which were reversed by co-treatment with the CD40L-neutralizing antibody LL48. In vitro analysis indicated that CD40L-induced growth inhibition was accompanied by apoptotic events including cell shrinkage, rounding, and detachment from the adherent T47D culture monolayer. Thirty-one and 27% of gp39-treated T47D and BT-20 cells underwent apoptosis, respectively, as compared with 56 and 65% from the same cell lines after treatment with the Fas agonistic antibody CH-11. An up-regulation of the proapoptotic protein Bax in T47D and BT-20 cells was observed, which indicated that this Bcl-2 family member may contribute to this growth-inhibitory effect. To explore the clinical relevance of CD40L-CD40 interaction, retrospective immunohistochemical analysis was carried to characterize in situ CD40- and CD40L-expression in breast cancer patient biopsies. All of the infiltrating ductal (5 of 5 cases tested) and lobular (4 of 4 cases) breast carcinomas, carcinomas in situ (6 of 6 cases), and mucinous carcinoma tested (1 case) expressed CD40. Varying proportions of tumor cells also expressed CD40L in the majority of infiltrating ductal (3 of 5 cases) and lobular (3 of 4 cases) carcinomas, and carcinomas in situ (4 of 6 cases), as determined by immunohistochemistry and validated by RT-PCR detection of the CD40L message in only CD40L positive-staining cases. Tumor infiltrating mononuclear cells from infiltrating carcinomas and carcinomas in situ expressed CD40 (10 of 10 cases), but less commonly CD40L (1 case of infiltrating lobular carcinoma, 2 cases of carcinoma in situ). Our findings indicate that the CD40 signaling pathway is active in human breast carcinoma cells. However, tumor-infiltrating lymphocytes from primary tumor tissues may be limited in their capacity to directly modulate tumor growth through the CD40L-CD40 loop.