Premalignant progression and malignant conversion in experimental skin carcinogenesis is itself a multistage process with predictable sequential expression of markers and genetic/epigenetic changes defining advancing progression. A subset of benign tumors is marked for progression as they erupt (high risk tumors), and they have a specific genetic signature. Deficiencies in crosstalk between cell surface adhesion receptors and extracellular matrix molecules are characteristic of malignant conversion. Syndecan 1 is a cell surface proteoglycan that contributes to keratinocyte migration, adhesion and growth factor responses. Syndecan 1 is lost during malignant conversion of mouse and human squamous cell carcinomas. Syndecan 1 null mice have been subjected to experimental skin carcinogenesis. In mice genetically ablated for syndecan 1, benign tumor formation is reduced during experimental skin carcinogenesis, but malignant conversion is enhanced. The susceptibility to malignant conversion is reproduced in orthografts of ras-transduced syndecan null keratinocytes. Migration and interaction with integrins is altered in ras-transduced syndecan 1 null keratinocytes producing a benign tumor type characteristic of high risk for premalignant progression. This is associated with enhanced TGFbeta signaling previously noted to increase risk for malignant conversion. Syndecan 1 expression in human squamous tumors may have diagnostic value for predicting malignant progression. Previous studies on premalignant progression indicated that loss of p53 function was tightly linked to malignant conversion. In a search for effectors downstream of p53, we discovered CLIC4, a member of a family of proteins with chloride channel activity. CLIC4 is a dimorphic protein that exists in soluble or membrane bound forms regulated by redox potential. Cytoplasmic CLIC4 translocates to the nucleus under a variety of cell stress stimuli and is essential for growth inhibition or apoptosis induced by p53, c-Myc or other regulators of cell growth and viability. We have shown that CLIC4 is excluded from the nucleus of tumor cells and overall expression is reduced in a variety of human cancers and mouse skin cancers. In grafting studies in vivo, targeting CLIC4 to the nucleus with an adenoviral or transgenic vector inhibits growth of mouse and human tumor cells. Paradoxically, stromal CLIC4 is highly upregulated in human and mouse cancers and associated with myofibroblast conversion. Current studies have revealed that CLIC4 is nitrosylated on specific cysteine moieties that induces unfolding of the protein and enhances association with the nuclear import machinery and nuclear translocation. Studies with cysteine mutants of CLIC4 have underlined the importance of the native folding of the protein for both association with protein partners, subcellular distribution and protein stability. Nitric oxide synthases, particularly eNOS in keratinocytes and thioredoxin reductase regulate this activity. NO also affects CLIC4 protein level as inhibition of NOS activity using 1400W and LNAME reduces the level of the protein in keratinocytes. This regulation is likely dependent on proteasome-mediated degradation of CLIC4, as levels are restored by proteosome inhibitors. We hypothesize that changes in redox potential in tumor cells are responsible for the altered nuclear localization and reduction in levels of CLIC4. The function of CLIC4 in the nucleus has been clarified by our recent findings that CLIC4 is an essential regulator of TGFbeta signaling. TGFbeta enhances CLIC4 expression and association with Schnurri-2, a transcription factor in the TGFbeta signaling pathway. Association with Schnurri-2 is essential for TGFbeta dependent CLIC4 nuclear translocation. Nuclear CLIC4 enhances TGFbeta signaling through interaction with phospho-Smad 2 and 3 and preventing their dephosphorylation by PPM1a, thus prolonging the signal. We hypothesize that defective TGFbeta signaling in tumor cells may also contribute to the exclusion of CLIC4 from the nucleus. Preliminary results suggest that the mechanism by which TGFbeta enhances CLIC4-Schnurri-2 association is a redox regulated phenomenon. Treatment of keratinocytes with the antioxidant N-acetyl cysteine inhibits TGFbeta dependent CLIC4-Schnurri-2 interaction, CLIC4 nuclear translocation and TGFbeta signaling, thus joining redox alterations and defective TGFbeta signaling as combined regulators of CLIC4 alterations in tumor cells. The marked elevation of stromal CLIC4 in cancers is likely in response to TGFbeta secreted by tumor cells that mediates myofibroblast conversion characterized by high expression of alpha smooth muscle actin (SMA) in the stroma. This is supported by knockdown of CLIC4 in primary murine dermal fibroblasts that decreases basal levels of SMA and inhibits its induction by TGFbeta. Our studies revealed that CLIC4 might affect TGFbeta dependent fibroblast transdifferentiation through activation of ERK both by itself and additively with TGFbeta. Based on these and previous reports from the laboratory, CLIC4 could become a therapeutic target to block TGFbeta signaling under conditions where TGFbeta enhances tumor progression through the stromal compartment. To address additional questions regarding CLIC4 and cancer development, we have generated both conventional and conditional CLIC4 knockout mice. CLIC4 knockout mice are viable and young mice display no apparent gross phenotypes. However, with aging these mice develop ulcerative skin lesions. The floxed animals are being mated with K5 Cre and inducible K14 Cre to generate epidermal specific null mice. Such inducible acute removal of CLIC4 should potentially offset possible compensation by other CLICs which might otherwise occur in conventional knockout mouse model systems. We have constructed a GFP-knockin targeting vector that will be used to generate genetically modified mice that will express GFP from the CLIC4 promoter. These mice will be employed to study the expression of CLIC4 under normal conditions, during mouse development and in various cell types and tissues with disturbed homeostasis. The knockin mice will also be used to study the differential expression of CLIC4 between tumor epithelium and stroma during induced skin carcinogenesis. While pursuing these basic studies to define effectors of critical pathways that might be useful therapeutic targets, we are also examining drugs that are currently in clinical trials. PEP005 (ingenol-3-angelate) is a natural product in clinical trials for the topical treatment of premalignant and malignant epithelial skin tumors. It is an activator of PKCdelta and thus has potential to induce a death pathway in keratinocytes. However, PKCdelta is either not expressed or catalytically dead in skin cancer cells. Our studies reveal a different mechanism of action. PEP005 is a substrate for P-glycoprotein (P-gp) and is transported deeply into the subcutis of the skin where it activates PKC on endothelial cells, causing hemorrhage and death of tumor cells by an anti-angiogenic mechanism. Inhibitors of P-gp prevent the hemorrhagic response and the therapeutic response of PEP-005. As a substrate for P-gp, PEP005 blocked photoaffinity labeling of human P-gp with labeled Iodoaryazidoprazosin and inhibited P-gp mediated drug resistance in HCT-15 cells. The intracellular levels of PEP005 were significantly lower in P-gp expressing cells and treatment with XR9576 increased the levels to the level of control cells, demons [summary truncated at 7800 characters]
Showing the most recent 10 out of 23 publications
