Joanne S. Finch

Expression of metalloproteinase genes in human prostate cancer

SummaryTwenty-five surgical specimens of malignant human prostate, 3 lymph nodes with metastatic prostate carcinoma, 11 normal human prostates, as well as 3 human prostate cell lines (DU-145, PC3 and LNCaP) were examined for the expression of the human matrix metalloproteinase-7 gene (MMP-7) from the human collagenase family (originally called PUMP-1 for putative metalloproteinase-1) [Quantin et al. (1989) Biochemistry 28:5327–5334; Muller et al. (1988) Biochem J 253:187–192; Matrisian and Bowden (1990) Semin Cancer Biol 1:107–115]. Northern blots were prepared using total RNA extracted from 18 prostate adenocarcinomas, 2 lymph nodes with metastatic prostate carcinoma and 11 normal human prostates. When the northern blots were hybridized with a32P-labeledMMP-7 cDNA probe, a 1.2-kb mRNA was detected in 14 out of 18 prostate adenocarcinomas, 1 out of 2 metastatic lymph nodes, and 3 out of 11 normal prostates. The 3 human prostate cell lines did not show any evidence of theMMP-7 transcript. In situ hybridization was conducted to localize theMMP-7 mRNA to individual cells using a35S-labeledMMP-7 cRNA. In situ hybridization was carried out on snap-frozen tissue sections of 7 prostate adenocarcinomas and 3 lymph nodes containing metastatic prostate adenocarcinoma using the same tissues previously probed by northern analysis as well as new samples. In situ hybridization revealed that theMMP-7 gene was expressed in the epithelial cells of primary prostate adenocarcinoma as well as in invasive and metastatic cells.MMP-7 expression was also seen focally in some dysplastic glands but not in stroma. Additional northern blot analysis was performed using probes to human type-IV collagenase, type-I collagenase and stromelysin I in human prostate adenocarcinoma as well as normal prostate tissue. Our results indicated that 6 out of 10 adenocarcinoma samples and none of the 4 normal samples were positive for type-IV collagenase transcripts. Tissue samples were also examined for the expression of type-I collagenase (9 adenocarcinomas and 4 normal) and stromelysin I (13 adenocarcinomas) by northern analysis. None of the tissues was found to express the transcripts of interest at detectable levels. These data suggest that certain metalloproteinases are present in prostatic adenocarcinoma and may play a role in invasion and metastasis.

Role of cyclic AMP responsive element in the UVB induction of cyclooxygenase-2 transcription in human keratinocytes

It has been shown that UVB irradiation induces expression of COX-2 and up-regulation of COX-2 plays a functional role in UVB tumor promotion. In this study, we examined the cis-elements in the human COX-2 promoter that may be responsible for the UVB induction of COX-2. Analyses with the COX-2 promoter region revealed that the cyclic AMP responsive element near the TATA box was essential for both basal and UVB induced COX-2 expression. This was further supported by studies using a dominant negative mutant of CREB, which strongly inhibited the activity of COX-2 promoter. Electrophoretic mobility shift assays indicated that CREB and ATF-1 were the major proteins binding to the COX-2 CRE. CREB and ATF-1 were phosphorylated upon UVB treatment, and SB202190, a p38 MAPK inhibitor, decreased the phosphorylation of CREB/ATF-1 and suppressed COX-2 promoter activity. In contrast, treatment with forskolin, an activator of adenylyl cyclase, led to phosphorylation of CREB and ATF-1 and activation of COX-2 promoter. Finally, enhanced binding of phospho-CREB/ATF-1 to the COX-2 CRE was observed after UVB induction. Thus, one signaling pathway for UVB induction of human COX-2 involves activation of p38, subsequent phosphorylation of CREB/ATF-1, and activation of the COX-2 CRE through enhanced binding of phosphorylated CREB/ATF-1.

Extracellular Signal-regulated Kinase 1/2-mediated Phosphorylation of JunD and FosB Is Required for Okadaic Acid-induced Activator Protein 1 Activation

Previously, we reported that in papilloma-producing 308 mouse keratinocytes, the tumor promoter okadaic acid, a serine-threonine phosphatase inhibitor, increased binding of activator protein 1 (AP-1) to a consensus 12-O-tetradecanoylphorbol-13-acetate-responsive element (Rosenberger, S. F., and Bowden, G. T. (1996) Oncogene 12, 2301–2308). In this study, we investigated the correlation between AP-1 DNA binding and transactivation and examined molecular mechanisms involved in this process. Using a luciferase reporter driven by region −74 to +63 of the human collagenase gene, we demonstrated induction of AP-1-mediated transcription following okadaic acid treatment. By performing in vitro kinase assays, we found elevated activities of extracellular signal-regulated kinase (ERK) 1/2, c-Jun N-terminal kinase, and p38 mitogen-activated protein kinase. The ERK-1/2-specific inhibitor PD 98059 completely abrogated okadaic acid-induced AP-1 transactivation without altering AP-1 expression, DNA binding, or complex composition. Phosphorylation analyses indicated that inhibition of ERK-1/2 decreased okadaic acid-elevated phosphorylation of JunD and FosB. To further examine the role of JunD and FosB in okadaic acid-induced AP-1 transactivation, we generated fusion proteins of the DNA-binding domain of the yeast transcription factor Gal4 and the transactivation domain of either JunD or FosB. Cotransfection experiments of these constructs with a Gal4-luciferase reporter demonstrated that both JunD and FosB are required for okadaic acid-induced transcription. Treatment with PD 98059 reduced JunD/FosB-dependent transactivation, suggesting that ERK-1/2-mediated phosphorylation is a critical component in this process.

Further purification of epidermal growth factor by high-performance liquid chromatography

Abstract Epidermal growth factor (EGF) purified by the method of Savage and Cohen ( J. Biol. Chem. 247, 7601–7611 (1972) using DEAE-cellulose chromatography as the final purification step was further resolved into two major uv-absorbing components by reverse-phase high-performance liquid chromatography (HPLC). Both components, referred to as α-EGF and β-EGF, competed with 125 I-labeled EGF for the EGF receptor, induced premature eye opening in neonatal mice, and had an amino acid composition similar to that published by Savage et al. ( J. Biol. Chem. 247, 7612–7621 (1972). β-EGF migrated slightly faster than α-EGF during sodium dodecyl sulfate-urea-polyacrylamide gel electrophoresis. α-EGF was fourfold more potent than β-EGF and was 10-fold more potent than DEAE-purified EGF in stimulating DNA synthesis in quiescent Rat-1 cells. HPLC purification of EGF can replace the DEAE-cellulose chromatography step currently used and produces a more potent and less heterogeneous EGF species.

JunB negatively regulates AP-1 activity and cell proliferation of malignant mouse keratinocytes

Abstract Objective: Previously we have shown that a malignant mouse keratinocyte cell line, 10Gy5, has elevated AP-1 transactivation and reduced JunB protein levels compared to its parental benign cell line, 308, and that the tumorigenicity in the 10Gy5 cells could be blocked by a dominant negative c-Jun mutant protein. We wished to determine whether the change in JunB protein levels could account for the elevated AP-1 activity and whether re-expression of JunB in malignant 10Gy5 cells altered their proliferative capacity. Design: In the current study, we reduced JunB expression in benign 308 cells with antisense oligonucleotides and increased JunB expression in malignant 10Gy5 cells by stable transfection of a JunB expression vector. Results: Increased AP-1 activity was detected after treatment of the benign 308 cell line with JunB antisense oligonucleotides that reduced JunB protein levels. Stably JunB-transfected clones of malignant 10Gy5 cells showed decreased AP-1 activity, slowed in vitro cell proliferation and reduced tumor growth when xenografted to athymic nude mice. Conclusion: These findings suggest that expression of JunB protein has a negative effect on malignant tumor cell proliferation in part through its ability to inhibit AP-1 transactivation.

Okadaic acid induces transcription of junB through a CCAAT box and NF-Y.

The shellfish toxin, okadaic acid (OA), is a potent tumor promoter that induces expression of the proto-oncogene junB in mouse keratinocyte 308 cells. Here we show, through deletion analysis of the junB promoter, that sequences near the TATA box conferred transcriptional induction by OA. Transient transfections of luciferase constructs bearing the junB promoter with single mutations in various cis elements demonstrated that a promoter containing a mutated CCAAT box could not be induced by OA. When this CCAAT box was inserted into a heterologous promoter construct, OA induction was dependent on an intact CCAAT box. Flanking cis elements located near the CCAAT box, although not required for OA inducibility, did play a role in the basal level of transcription. NF-Y was shown by EMSA to bind to the CCAAT box. OA induction from the junB CCAAT box was blocked by dominant negative NF-YA as well as the CCAAT box-dependent anticancer drug, ET-473. Expression of a lexA/NF-YA chimeric protein demonstrated that OA induction was dependent on the binding of NF-Y family members. These studies demonstrate that OA can mediate transcriptional activation of junB through the classical CCAAT box and that transcription factor NF-Y plays a functional role in the induction.

Binding, internalization and intracellular processing of 125I-epidermal growth factor purified by isoelectric focusing

Abstract When epidermal growth factor (EGF) which had been extensively purified by HPLC was subjected to iodination with sodium 125iodide, 5 major species of differing isoelectric points were produced. Some of these species bound to rat fibroblasts with different affinities but were internalized with equal efficiency. Examination of the internalized 125I-labelled molecules revealed processing of all the 125I-EGF species to macromolecules with more acidic isoelectric points. The 125I-EGF species with a pI of 4.5 corresponded in electrofocusing behavior with intact non-iodinated EGF. Other EGF species probably represented molecules which were covalently modified as a result of the iodination procedure.

Inhibition of UVA-induced c-Jun N-terminal kinase activity results in caspase-dependent apoptosis in human keratinocytes.

Abstract Inhibition of c-Jun N-terminal kinase (JNK) with the pharmacologic inhibitor SP600125 in UVA-irradiated HaCaT cells and human primary keratinocytes resulted in dramatic phenotypic changes indicative of cell death. These phenotypic changes correlated with caspase 8, 9 and 3 activations as well as cleavage of the caspase substrate polyADP-ribose polymerase (PARP). Morphologic analysis and analysis of sub-G0 DNA content confirmed apoptotic cell death in these keratinocytes after combination treatment. Addition of the general caspase inhibitor zVAD-fmk to combination-treated HaCaT cells was able to completely block caspase activation, PARP cleavage, the increase in sub-G0 DNA content and the classic morphologic features of apoptosis, indicating that this combination treatment resulted in caspase-dependent apoptotic cell death. zVAD-fmk treatment of primary keratinocytes was able to completely inhibit caspase activation and PARP cleavage, reduce morphologic apoptosis at lower concentrations of SP600125 and decrease the sub-G0 DNA content detected after UVA + SP600125 treatment. However, cell death and a significant amount of debris was still detected after caspase inhibitor treatment, particularly with 125 nM SP600125. At subconfluent conditions and low passage, primary keratinocytes were more sensitive to UVA irradiation alone than HaCaT cells. In conclusion, we have observed that inhibition of UVA-induced JNK activity with the pharmacologic inhibitor SP600125 resulted in caspase-dependent apoptotic cell death in both the immortalized keratinocyte cell line HaCaT and primary keratinocytes. However, the increased sensitivity of primary keratinocytes to experimental stress may have also resulted in direct cellular injury and caspase-independent cell death.

Heterogeneity of 125I-labeled epidermal growth factor

Highly purified epidermal growth factor (EGF) was iodinated, and the labeled product with the same isoelectric point as underivatized EGF was isolated by isoelectric focusing. When the 125I-labeled EGF was analyzed by reverse-phase chromatography, the resulting profile of 125I activity was much broader than the profile obtained with underivatized EGF. Rechromatography of 125I-EGF fractions indicated that our highly-purified labelled EGF was indeed heterogeneous. Analysis of each HPLC column fraction demonstrated that degradation of EGF had not occurred. The column fractions containing 125I-EGF were pooled into five groups for analysis of cell binding characteristics. Scatchard plot analysis of the five 125I-EGF pools revealed markedly different binding behaviors. In contrast, they had equal potency in stimulating DNA synthesis, within the sensitivity of our assay. Specific activity measurements indicated that the five HPLC pools of 125I-EGF had varying numbers of 125I atoms per EGF molecule. The heterogeneity of the highly purified 125I-EGF and the binding characteristics of the 125I-EGF subfractions pose serious implications for all workers who use iodinated ligands for receptor binding studies.

Loss of catalase increases malignant mouse keratinocyte cell growth through activation of the stress activated JNK pathway

A cell line that produces mouse squamous cell carcinoma (6M90) was modified to develop a cell line with an introduced Tet‐responsive catalase transgene (MTOC2). We have previously reported that the overexpressed catalase in the MTOC2 cells reverses the malignant phenotype in part by decreasing epidermal growth factor receptor (EGFR) signaling. With this work we expanded the investigation into the differences between these two cell lines. We found that the decreased EGFR pathway activity of the MTOC2 cells is not because of reduced autocrine secretion of an epidermal growth factor (EGF) ligand but rather because of lower basal receptor activity. Phosphorylated levels of the mitogen‐activated protein kinase (MAPK) members JNK and p38 were both higher in the 6M90 cells with low catalase when compared with the MTOC2 cell line. Although treatment with an EGFR inhibitor, AG1478, blocked the increased activity of JNK in the 6M90 cells, a similar effect was not observed for p38. Basal levels of downstream c‐jun transcription were also found to be higher in the 6M90 cells versus MTOC2 cells. Activated p38 was found to down‐regulate the JNK MAPK pathway in the 6M90 cells. However, the 6M90 cells contain constitutively high levels of phosphorylated JNK, generating higher levels of phosphorylated c‐jun and total c‐jun than those in the MTOC2 cells. Inhibition of JNK activity in the 6M90 cells reduced AP‐1 transcription and cell proliferation. The data confirm the inhibitory effects of catalase on tumor cell growth, specifically through a ligand‐independent decrease in the stress activated JNK pathway.