EGF Receptor Signaling Inhibits Keratinocyte Apoptosis: Evidence for Mediation by Bcl-XL*

Oncogene (1998) 16, 1493 ± 1499 1998 Stockton Press All rights reserved 0950 ± 9232/98 $12.00 SHORT REPORT EGF receptor signaling inhibits keratinocyte apoptosis: evidence for mediation by Bcl-XL* Stefan W Stoll1, Mary Benedict2, Raj Mitra2, Annie Hiniker1, James T Elder1,3,4 and Gabriel NunÄ ez2,4 1Departments of Dermatology, 2Pathology, and 3Radiation Oncology and 4Comprehensive Cancer Center, University of Michigan Medical School, Ann Arbor, Michigan 48109, USA Signaling through the epidermal growth factor receptor 1997). An additional family of EGF-related growth (EGFR) has been primarily implicated in the growth of factors, termed neu dierentiation factors (NDFs), epithelial cells including keratinocytes. However, the further adds to the complexity of the ErbB signaling mechanism by which EGFR stimulation promotes system. NDFs are expressed from at least two genes as keratinocyte cell growth is poorly understood. Here we a number of isoforms generated by alternate splicing report that human keratinocytes undergo apoptosis when (Carraway et al., 1997). Initially, it was thought that incubated with the blocking EGFR monoclonal antibody NDF was the speci®c ligand for ErbB-2 (Holmes et al., 225 IgG, or PD153035, a highly speci®c EGFR tyrosine 1992). However, through much eort it is now clear kinase inhibitor. Endogenous mRNA and protein levels that a variety of asymmetric ErbB heterodimers of Bcl-XL, a member the Bcl-2 family which suppresses capable of binding ligands of both the EGF and apoptosis, were speci®cally inhibited by EGFR blockade. NDF families can form, including ErbB-1/-2, ErbB-1/- Furthermore, stimulation of EGFR signaling through two 3, and ErbB-2/-3 (Alimandi et al., 1997; Graus Porta et natural ligands, transforming growth factor (TGF)-a and al., 1997; Tzahar et al., 1996). Despite these emerging epidermal growth factor (EGF), increased the expression complexities, at least ®ve lines of evidence suggest that of Bcl-XL in quiescent keratinocytes and HaCaT cells. ErbB-1 is responsible for the bulk of autocrine Finally, ectopic expression of Bcl-XL in HaCaT cells keratinocyte activation through the ErbB system. increased survival after EGFR blockade when compared First, highly speci®c anti-EGFR monoclonal antibo- to untransfected cells or HaCaT keratinocytes trans- dies (mAb) abrogate proliferation and ligand-induced fected with empty vector. These results suggest that the tyrosine phosphorylation of a 170 kDa ErbB-1- anti-apoptotic protein Bcl-XL plays an important role in immunoreactive protein in normal human keratino- the maintenance of keratinocyte survival in response to cytes (Klein et al., 1992). Second, while NDF is EGFR signaling. expressed in the dermis in vivo, and in cultured keratinocytes in vitro, it is not mitogenic for Keywords: apoptosis; EGF receptor; Bcl-XL; keratino- keratinocytes in culture (Danilenko et al., 1995; cytes; signal transduction Poumay and Pittelkow, 1995). Third, keratinocytes do not detectably express ErbB-4 (Marikovsky et al., 1995). Fourth, ErbB-3 is essentially inactive catalyti- Keratinocytes appear to play an active role in cally (Guy et al., 1994). Finally, ErbB-2 has no known cutaneous immunoregulation and growth control via ligand and appears to respond catalytically to ligand the production of a number of cytokines and growth only as a heterodimer (Graus Porta et al., 1997). factors (Elder, 1995). Prominent among them are Activation of growth factor receptors transmit several members of the EGF family, including signals involved in mitogenesis but also in cellular transforming growth factor (TGF)-a, amphiregulin dierentiation and survival. For example, stimulation (AR), and heparin-binding EGF-like growth factor of IL-3 or erythropoietin receptors provides both (HB-EGF) (Coey et al., 1987; Cook et al., 1991; proliferation and survival signals to hematopoietic Hashimoto et al., 1994). It has been suggested that precursors (Boyer et al., 1992; Koury and Bondurant, these factors regulate growth and dierentiation of 1988; 1990; Silva et al., 1996). Similarly, signaling normal keratinocytes by serving as autocrine and/or through EGFR has been shown to promote cell cycle juxtacrine ligands for the EGF receptor (EGFR, ErbB- progression, dierentiation and survival of epithelial 1) (Barnard et al., 1994). However, three other cells (Merlo et al., 1995; Wu et al., 1995). Blockade of receptors have been identi®ed as belonging to the EGFR signaling on benign and tumor epithelial cell same subfamily of tyrosine kinases as ErbB-1 (termed lines with neutralizing antibodies directed against this ErbB-2, -3, and -4), and at least two of these, ErbB-2 receptor resulted in a cytostatic eect with or without and -3, are expressed by keratinocytes (Marikovsky et cell cycle arrest [(Wu et al., 1995), and references al., 1995; Poumay and Pittelkow, 1995; Xian et al., therein. In certain tumor cell lines, however, treatment with anti-EGFR antibodies induced apoptotic cell death suggesting that EGF stimulation is involved in Correspondence: G Nunez or JT Elder the regulation of cell survival (Wu et al., 1995). *This work was presented in part at the annual meeting of the However, the signi®cance of EGF-induced survival Society of Investigative Dermatology, Washington, District of Columbia, April 23 ± 27, 1997 for the overall function of EGFR remains unclear, as Received 17 July 1997; revised 17 October 1997; accepted 20 October does the mechanism by which EGFR activation 1997 promotes cell survival. EGFR mediates keratinocyte survival via Bcl-XL SW Stoll et al 1494 Apoptosis, a morphologically distinguished form of capable of extensive dierentiation upon transplantation programmed cell death, is critical not only during to nude mice (Boukamp et al., 1988). PD153035 and development and tissue homeostasis but also in the other low molecular weight inhibitors of ErbB tyrosine pathogenesis of cancer and a variety of diseases kinase activity are thought to owe their selectivity to (Thompson, 1995; Vaux and Strasser, 1996). Several interactions with structures in the kinase domain unique regulatory components of the apoptotic pathway have to the ErbB family (Fry et al., 1997; Hanks and Hunter, been identi®ed in various organisms including man 1995). Thus, PD153035 displayed 20-fold higher (Evan et al., 1996). The products of several members of inhibitory potency against ErbB-2 than against any of the Bcl-2 family of proteins including Bcl-2, Bcl-XL, ®ve other tyrosine kinases (Fry et al., 1994). While its Mcl-1, A1, Bcl-w share conserved regions termed Bcl-2 selectivity for ErbB-1 within the ErbB family has not homology (BH) domains 1, 2, 3 and 4 and function by been reported in detail, PD153035 was 106 times more repressing apoptosis (Chittenden et al., 1995; Yin et al., potent against ErbB-1 (as anity-puri®ed receptors from 1994). One member of this family, Bcl-XL, is expressed A431 cells, IC50=29 pM) than against ErbB-2 (as in many embryonic and adult tissues, and its overexpressed soluble cytoplasmic domains, importance has been revealed by analysis of mutant IC50=2.3 mM) (Fry et al., 1994). Nanomolar concentra- mice de®cient in Bcl-XL (Gonzales-Garcia et al., 1994; tions of PD153035 markedly inhibited HaCaT cell Motoyama et al., 1995). Unlike Bcl-2 itself, Bcl-XL is growth, with an IC50 of less than 30 nM (Figure 1a). At expressed at high levels in normal adult human these concentrations, PD153035 had no eect on keratinocytes (Wrone-Smith et al., 1995). Given the tyrosine phosphorylation by ErbB-2 cytoplasmic do- important role that Bcl-XL plays in the regulation of mains (Fry et al., 1994). In parallel experiments, HaCaT apoptosis and its expression in the cell type of interest, cells were incubated with increasing concentrations of we sought to determine if the EGFR pathway could 225 IgG, an anti-EGFR monoclonal antibody which regulate keratinocyte survival through Bcl-XL. blocks ligand binding without inducing tyrosine In order to examine the eect of EGFR blockade on phosphorylation (Sunada et al., 1986). Treatment with keratinocyte cell growth, HaCaT cells were incubated puri®ed 225 IgG but not with control IgG signi®cantly with PD153035, a highly selective inhibitor of the EGFR inhibited the growth of HaCaT cells (Figure 1b). tyrosine kinase. (Fry et al., 1994) HaCaT is a However, inhibition was not complete, despite an spontaneously-immortalized, nontumorigenic keratino- antibody concentration more than ®ve times that cyte cell line derived from normal-appearing back skin of required for EGF to saturate EGFR in normal a 62 year old male (Boukamp et al., 1988). Although this keratinocytes (1 mg/ml 225 IgG=6.6 nM, vs the EGFR- line carries mutations in both p53 alleles (Lehman et al., saturating concentration of 1.2 nM EGF (Nickolo et 1993) and displays a transformed phenotype in vitro,itis al., 1989)). Taken together, these results strongly Figure 1 EGFR blockade by the highly speci®c EGFR tyrosine kinase inhibitor PD153035 (Fry et al., 1994) (a) or by the anti- EGFR monoclonal antibody 225 IgG (Sunada et al., 1986) (b) reduces HaCaT cell numbers in culture. HaCaT cells, kindly provided by Prof Dr Fusenig (Heidelberg, Germany) were seeded at approximately 5000 cells/cm2 in Dulbecco's modi®ed Eagle's medium (DMEM) supplemented 10% fetal bovine serum (FBS). After 24 ± 48 h, cultures were incubated with the indicated doses of PD153035, 225 IgG or 1 mg/ml of the IgG isotype control MOPC21 (Organon Teknika, Durham, NC). Four days after treatment cells were either lysed with hexadecyldimethylammonium bromide (Eastman Kodak, New Haven, CT) or harvested by trypsinization. Counting of nuclei or cells was performed either by using a coulter counter (Coulter Electronics, Hialeah, FL) or a hemacytometer. Data are expressed as % of DMSO treated (a) or untreated controls (b) and are mean+s.e.m of three independent experiments each performed with triplicate plates. The data for the control IgG MOPC21 (b) represents mean+range (n=2).

Load more