CD40L is a member of the TNF family of ligand which is normally expressed by T lymphocytes and interacts with CD40 expressed on B cell and antigen presenting cells . CD40L exerts complex anti- or pro-apoptotic effects in normal and transformed B lymphocytes, enhancing Fas mediated apoptosis [20, 21] but protecting against apoptosis induced by cytotoxic agents [15–17]. CD40 expression is not limited to cells of the hematopoietic system, and has been found detectable on a variety of human carcinomas, including bladder, breast, ovarian, lung as well as in melanoma cell lines [31–37]. The purpose of this study was to investigate the effect of CD40L on the apoptosis of CD40-expressing carcinomas and lymphoma cell lines induced by a variety of cytotoxic agents.
In NHL cell lines, CD40 ligand expressed on adjacent non tumoral cells was found capable 1) to inhibit the apoptosis induced by five different commonly used cytotoxic agents (DOX, ETO, CDDP, VIN, TAX), 2) to inhibit the activation of caspase-3/7 induced by DOX, ETO, VIN and TAX, 3) to inhibit drugs (DOX, ETO, VIN TAX) induced PARP cleavage by apoptosis protease like YAMA/CPP32/Apopain/Caspase-3 and 4) to partially reverse the antiproliferative effect of the five cytotoxic agents in the 5 NHL cell lines tested. The inhibition of caspase-3-like and 8-like activities, but not caspase-9-like activity, by permeable tetrapeptide inhibitors also blocked the apoptosis induced by doxorubicin in NHL cell lines, suggesting that the downregulation of caspase-3-like is an essential molecular mechanism of the protective effect of CD40L.
However, the modulation of caspase-3 is not the sole mechanism of the protective effect of CD40L in NHL cell lines. Raji was found to be resistant to reversion of apoptosis induction by TAX and ETO despite of modulation of caspase-3-like activation and PARP cleavage. In addition, resistance by CD40L to CDDP induced apoptosis was not associated with caspase-3/7 and PARP cleavage modulation in 4 of the 5 NHL cell lines tested. Apoptosis resistance mechanisms acting downstream of caspase-3 activation have been described in particular in the Raji cell line [55, 56] as well as in ovarian carcinoma cell lines treated with CDDP . Therefore, protection from drug-induced apoptosis by CD40L in NHL cells occurs through a caspase dependent pathway for anthracyclins, etoposide, paclitaxel and vinblastin, but through a caspase independent pathway for CDDP and for the resistant Raji cell line, presumably at a convergent point of apoptosis induction for all cytotoxic agents downstream of caspase-3.
In contrast, in CD40 expressing breast carcinoma cell lines, the protective effect of CD40L was found to be caspase-independent. Co-culture with CD40L L cells was found capable to protect breast (as well as renal, prostatic and colon) carcinoma cell lines against DOX induced apoptosis. Exposure of the 2 breast carcinoma cell lines tested here to doxorubicin, CDDP, Paclitaxel, Vinblastin or etoposide did not or only weakly increased caspase-3/7 activity while caspase-3/7 activity increased up to 28-fold in NHL cell lines tested after exposure to the same drugs. Finally, CD40L did not inhibit caspase-3/7 activation and PARP cleavage induced by any of the 5 cytotoxic agents in the 2 breast carcinoma cell lines tested. Actually, exposure to CD40L L cell was found to increase caspase-3/7 activity in drug-treated breast cancer cells, in marked contrast with what was observed for NHL cell lines. Consistent with these observations, cell permeable tetrapeptide inhibitors of caspase-3, 8 and 9-like activities failed to inhibit the apoptosis of breast carcinoma cell lines, in contrast to what is observed in NHL cell lines. These results indicate 1) that caspase-3 activation is not the single pathway required for the induction of apoptosis of breast cancer cell lines by the anti-cancer drugs tested considering the PARP cleavage but the weak caspase-3/7 activity modulation by drugs and 2) that the anti-apoptotic effect of CD40L on these cell lines do not involve a modulation of caspase-3/7 activity as it was shown by the absence of inhibition of caspase-3 activity and PARP cleavage.
Different mechanisms for CD40L protection against drug-induced apoptosis and drug anti proliferative effect could be suggested. Some studies indicated that the antiapoptotic function of CD40 is mediated by up-regulated expression of bcl-xL gene, an antiapoptotic member of the bcl-2 family of proteins and that up-regulation of Bcl-xL could be a key event in CD40-mediated survival in both normal tonsillar B cells and the immature B-cell lymphoma WEHI-231 cells [58–60]. More recently, Lee et al provided a crucial link in CD40-mediated antiapoptosis by linking the activation of the NFKB-signalling pathway to the up-regulation of Bcl-2 family members . The cytoplasmic domains of TNF receptor family like CD40 do not encode any enzymatic activity but some proteins, identified by the generic name TRAF, physically interact with them, providing some clues to the mechanisms of signal transduction . The CD40 cytoplasmic domain has been found to interact with TRAF2, TRAF3, TRAF5 and TRAF6 but TRAF2 is the best characterized of the TRAF proteins in term of signaling function. TRAF2 activates NFκB by means of a synergistic interaction with a novel protein called TANK . It has been suggested that TRAF2 would have to be released from CD40 in order for it to interact with TANK and activate NFκ B. Study of functional consequences, like NFkB activation, of preventing CD40-TRAF2 dissociation with the blocking antibody anti CD40L would provided information on mechanisms induced by CD40-TRAF2 interaction. Finally, the protective role of CD40L against drug-induced apoptosis may pass by the modulation of the cell cycle and the role of p53 family. Previous studies showed that CD40 can repress drug-induced apoptosis by, among others, B lymphoma cell cycle progression . Teoh et al showed, depending on multiple myeloma cells p53 status, that CD40 induced increase G1/S transition and cell proliferation or growth arrest with sub G1 phase cells and apoptosis . So many pathways remain to be explore in order to determine CD40/CD40L cellular mechanism.
Taken together, these results indicate that the protective effect of CD40L against cytotoxic agents involves caspase-dependent and independent pathways in NHL, and caspase independent pathways in breast carcinoma cells lines. Of note, similar observations were obtained in the colon carcinoma HCT116 cell line treated with doxorubicin, with no significant modulation of caspase-3/7 activity nor reversal of the cytostatic effect of doxorubicin (not shown). These results strongly suggest that the anti-apoptotic effect of CD40L involves different molecular mechanisms in the lymphoma and carcinoma cell lines tested in this study.
It has been reported that CD40L induces direct cytostatic effects on breast carcinoma cell lines [35, 39]. The lack of significant impact of CD40L on tritiated thymidine incorporation in the 2 breast carcinoma cell lines tested could hence have resulted from a combined "direct" inhibitory effect of CD40L on BCC proliferation and a reversal of drug-induced cytostatic effect by CD40L. This hypothesis is however unlikely since no direct anti-proliferative effect of CD40L expressing L cells was observed in the two breast carcinoma cell lines tested (not shown). The discrepancy between the present and previously published results  regarding the effect of CD40L on breast carcinoma cell proliferation could result either from a variability of response to CD40L in different carcinoma cell lines or alternatively to a different biological activity of membrane bound vs soluble CD40L. Indeed, soluble and membrane bound CD40L, as well as CD40 agonists, have been reported to exert opposite biological activities on lymphoma cell line proliferation and survival in previous studies [15–17, 23, 27, 28]. Tong et al reported apoptosis induction by CD40L in BCC in the absence of cytotoxic drugs, but moreover they used a recombinant molecule and not membrane bound CD40L . On the other hand, in agreement with our current finding Stumm et al have shown that CD40 stimulation in Breast carcinoma inhibited drug-induced apoptosis .
Ceramides have been reported to act as second messengers for the apoptosis induced by DNA damaging agents in some tumor cell lines [49–52]. A rapid intracellular ceramide increase has been observed after exposure to γ radiation or exposure to DNA damaging agents, resulting either from the activation of a sphingomyelinase or ceramide synthase, or from caspase-8 activation in different tumor models . To determine whether the protective signaI delivered by CD40L acts upstream or downstream of ceramide production, the protective effect of CD40L on the apoptosis induced by ceramides in BCC and NHL cell lines were compared. While cell permeable C2 and C6 ceramides induced apoptosis and blocked thymidine incorporation in the 5 NHL and 2 breast carcinoma cell lines tested, co-culture in the presence of CD40L L cells blocked the apoptotic signal induced by C2 and C6 ceramides in NHL and breast carcinoma cell lines. However, the molecular mechanisms involved in the induction of apoptosis by ceramides and cytotoxic agents were found to be different in breast carcinoma. While inhibitors of caspase-3 and 8-like activities partially prevented the apoptosis induced both by ceramide or doxorubicin in NHL cell lines, these inhibitors were active only for ceramide-induced apoptosis (and not for drug-induced apoptosis) in breast carcinoma cell lines. Since ceramide production in response to cytotoxic agents was not tested in the present study, it is not possible to establish whether CD40L affects ceramide increase in response to cytotoxic agents. However, ceramide and doxorubicin-induced apoptosis were modulated by different inhibitors in breast carcinoma suggesting that ceramide is at least not an exclusive mediator of doxorubicin induced apoptosis in this breast carcinoma cell line.
CD40L therefore may block the apoptosis of breast carcinoma cell lines independently of the modulation of caspase activities in these different models. Further studies on bcl-2 family protein like Bclxl/Bax, but also cell cycle, p53 and TRAF-2 would provide more information on mechanism of CD40L induced drug resistance.