A subset of patients with NSCLC, HNSCC, mCRC and pancreatic cancer are responding to therapy by several agents directed against the epidermal growth factor receptor (EGFR). Uniformly patients develop a papulopustular follicutis often accompanied by alopecia, xeroderma and changes in nails and eyelashes. To model this skin rash in a mouse model, EGFR was ablated in the epidermis in the litters derived from Keratin 5 driven Cre recombinase transgenic mice crossed with EGFR floxed mice. The skin of double transgenic mice reproduced the hallmarks of the skin lesions of patients undergoing chemotherapy with anti-EGFR agents: inflammation, pruritis, dry skin with neutrophilic pustules and infiltration of mast cells, macrophages and lymphocytes. Tissue samples isolated from skin of double transgenic mice contained high mRNA levels of a subset of inflammatory mediators namely TNF-alpha, IL-18, IL-1beta, IL-1ra, G-CSF, iNOS and CCL2. Higher levels of these mediators were also measured in plasma of double transgenic mice. EGFR ablated mice had normal body temperature but a slower heart rate that is in contrast with the smaller size typical of these animals. TGF-alpha, Amphiregulin and HB-EGF are downregulated in RNA extracted from skin of EGFR ablated mice while the ligand Epiregulin is highly upregulated. Attempts to ameliorate the skin phenotype by crossing EGFR ablated mice with knockout mice that were missing critical regulators of the highlighted inflammatory pathways such as TNFR1/2, and iNOS or by administering neutralizing G-CSF antibodies failed. Currently we are crossing double transgenic mice with MyD88, CCR2 and Rag1 knockout mice. We obtained plasma samples of 10 patients collected before and after one month of treatment with the anti EGFR drug Gefitinib. We observed a subset of inflammatory mediators to be elevated in at least 8 out of 10 patients after treatment with Gefitinib (CCL11, CCL22, CCL4, IL-1ra, IL-18). The other mediators examined (CCL2, CCL5, CCL17, CCL20, CXCL10, CXCL12, myeloperoxidase) have a variable outcome in all the different patients. Moreover IL-6 and CXCL8 are overall stable or decreased in the plasma of treated patients. Frozen sections from nonlesional skin before and after treatment were used to extract RNA encoding cytokines and chemokines. Patterns of expression are very variable within the patients and not always in the same direction of plasma changes. An interesting aspect of anti-EGFR therapy, particularly for NSCLC, HNSCC and mCRC, is the poor response in patients with an active mutant of the RAS oncogene. Last year we described our discovery of a 25 gene signature that characterizes keratinocytes transformed by oncogenic ras and either genetically or pharmacologically ablated for EGFR. Ingenuity analysis of the ras oncogene/EGFR ablation signature indicates 56% of this signature matches to one network, and the center node of the network is p38. Additional studies indicated that p38 is activated in keratinocytes genetically or pharmacologically ablated for EGFR and transformed with oncogenic ras. Activation of p38 was also seen in tumor orthografts of EGFR null/ras transformed keratinocytes. Using siRNA we found that the primary p38 isoform responsible for the activation is p38 alpha. Activation of p38 and gene expression changes seen in the 25 gene signature are also seen in drug treated human keratinocytes transformed by activated H-ras but not in drug treated non-transformed human keratinocytes. Additional microarray analyses of mouse keratinocytes genetically or pharmacologically ablated for EGFR without ras transformation revealed a 19 gene signature, of which 8 genes were down regulated and 11 genes were up-regulated. The Ingenuity analysis of the EGFR ablation signature in the absence of ras transformation reveals 57% of the signature genes that match to one network, with center nodes focusing on Erk1/2 and IL-1beta. These changes could be relevant for the inflammatory skin phenotype that develops during anti-EGFR therapy. One of the reproducible phenotypic changes observed during oncogenic ras transformation of keratinocytes is the loss of expression of suprabasal keratins K1 and K10 during differentiation. We now find that this previously unexplained response is due to NF-kappaB activity downstream of MyD88 and IL-1R. A 2.5kb fragment cloned from the K1 gene promoter contains an NFkappaB consensus binding site. When this sequence is used to drive a luciferase reporter, oncogenic ras inhibits reporter activity. However, MyD88 deficiency does not prevent this reduction in reporter activity. The K1 transcript 3UTR contains a well conserved site for miR203. Oncogenic ras opposes the induction of miR203 in differentiating conditions, but further work will be required to determine if this is mediated through NFkappaB. ROCK or Rho-associated kinase, a serine/threonine kinase, is an effector of Rho-dependent signaling and is involved in actin-cytoskeleton assembly as well as cell motility and contraction. To evaluate the role of ROCK activity in mouse keratinocyte differentiation, the effect of Y-27632, a ROCK-specific inhibitor, was tested. Inhibition of ROCK by Y-27632 induced keratinocyte differentiation markers K1 and K10 expression (mRNA and protein) in primary mouse keratinocytes cultured in 0.05 mM calcium medium. Y-27632 also enhanced the expression of K1 and K10 in keratinocytes grown in 0.12 mM calcium. Addition of Y-27632 to ras transformed keratinocytes also restored ras suppressed K1 and K10 expression. siRNA specifically targeting ROCK1 and ROCK2 reduced ROCK protein expression and enhanced K1 and K10 expression. Our study indicates that ROCK activities are important for maintaining the basal cell phenotype in cultured mouse primary keratinocytes. MyD88 deficiency in keratinocytes reduces the tumor yield in an initiation-promotion tumor induction experiment. Both initiation and promotion may be influenced by loss of MyD88 since the proinflammatory chemokines and cytokines elaborated by keratinocytes transformed by oncogenic ras or by treatment with 12-tetradecanoylphorbol-13-acetate (TPA) are attenuated with MyD88 ablation. To elucidate the role of keratinocyte-derived inflammation on skin cancers, a double transgenic mouse model (HGF-PKCalpha or DT) was used. K5-PKCalpha mice which overexpress PKCalpha in basal keratinocytes and develop a strong neutrophilic cutaneous inflammatory response upon topical TPA application were crossed with melanoma-prone MT1-HGF mice which overexpress HGF under a metallothionein promoter to create HGF-PKCalpha mice and their respective controls. While PKCalpha driven epidermal inflammation reduces melanomagenesis in DT mice, DT animals are very sensitive to squamous carcinogenesis. We hypothesize that in DT mice, keratinocyte-derived HGF synergizes with PKCalpha to drive tumor promotion and increase tumor growth. Primary keratinocytes derived from HGF or DT mice display upregulation of keratin 8 (K8) and downregulation of K1 and K10 mRNAs. The EGFR is transactivated in HGF and DT keratinocytes but not in PKCalpha or WT keratinocytes. The release of CXCL1 is augmented in ras-keratinocyte cell culture supernatants with DT ras-keratinocytes producing the most. Together these results suggest that the sensitivity of DT transgenic skin to squamous tumor induction is associated with the ability of HGF to transactivate EGFR signaling and PKCalpha to enhance tumor promotion.
Showing the most recent 10 out of 21 publications
