The inflammatory microenvironment of the aging prostate facilitates cellular proliferation and hypertrophy.

Benign Prostatic Hypertrophy (BPH, also known as benign prostatic hyperplasia or benign prostatic enlargement), is one of the most common benign proliferative conditions associated with aging in men and is pathologically characterized by the proliferation of fibroblast/myofibroblast and epithelial cell types in the prostate. Previous studies from our laboratory have shown that the CXC-type chemokines, CXCL5 and CXCL12, are secreted by aging prostate stroma and promote both proliferative and transcriptional responses from prostate epithelial cells. Using array-based gene expression profiling and quantitative reverse-transcriptase polymerase chain reaction, we now show that the transcriptome of the aging prostate stroma is characterized by the up-regulation of several genes that encode secreted inflammatory mediators, including secreted CXC-type chemokines (CXCL1, CXCL2, CXCL5, CXCL6, CXCL12), interleukins (IL11, IL33), and transcripts with cytokine homology (CYTL1). At the protein level, ELISA experiments demonstrated that CXCL1, CXCL5, and CXCL6 were secreted by primary prostate stromal fibroblasts explanted from aging prostate stroma. Dose-response assays confirmed that, like CXCL5 and CXCL12, CXCL1 and CXCL6 promote low-level proliferative responses from both prostate stromal fibroblasts and epithelial cells. Taken together, these data suggest that inflammatory mediators are secreted by prostatic stroma consequent to aging, that the levels of these mediators are sufficient to promote low-level increases in the proliferative rate of both epithelial and stromal fibroblast cell types. Moreover, these processes may account for the low-level, but cumulative, proliferation of both epithelial and fibroblastic/myofibroblastic cell types that characterizes the aging-associated development of benign prostatic hypertophy.

Reactivation of androgen receptor-regulated lipid biosynthesis drives the progression of castration-resistant prostate cancer.

Androgen receptor (AR) is a transcriptional activator that, in prostate cells, stimulates gene expression required for various cellular functions, including metabolisms and proliferation. AR signaling is also essential for the development of hormone-dependent prostate cancer (PCa) and its activity can be blocked by androgen-deprivation therapies (ADTs). Although PCa patients initially respond well to ADTs, the cancer inevitably relapses and progresses to lethal castration-resistant prostate cancer (CRPC). Although AR activity is generally restored in CRPC despite the castrate level of androgens, it is unclear whether AR signaling is significantly reprogrammed. In this study, we examined the AR cistrome in a PCa cell line-derived CRPC model using integrated bioinformatical analyses. Significantly, we found that the AR cistrome is largely retained in the CRPC stage. In particular, AR-mediated lipid biosynthesis is highly conserved and reactivated during the progression to CRPC, and increased level of lipid synthesis is associated with poor prognosis. The restoration of lipid biosynthetic pathways is partially due to the increased expression of AR splice variants. Blocking lipid/cholesterol synthesis in AR variants-expressing CRPC cell line and xenograft models markedly reduces tumor growth through inhibition of mTOR pathway. Silencing the expression of a fatty acid elongase, ELOVL7, also leads to the regression of CRPC xenograft tumors. These results demonstrate the importance of reactivation of AR-regulated lipid biosynthetic pathways in driving CRPC progression, and suggest that ADTs may be therapeutically enhanced by blocking lipid biosynthetic pathways.

Distinct areas of allelic loss on chromosomal regions 10p and 10q in human prostate cancer.

Utilizing tissue microdissection and PCR techniques, we have examined 35 prostate tumors paired with normal tissues from the same patients for allelic loss at 24 polymorphic loci spanning chromosome 10. Twenty-five tumors (71%) were deleted for at least one chromosome 10 locus. Of the total 35 tumors, 6 (17%) were deleted for 10p loci only, 5 (14%) for 10q loci only, and 14 (40%) were deleted for both 10p and 10q loci. The common region of deletion on 10p included loci D10S211-D10S89-D10S111. Fluorescence in situ hybridization of yeast artificial chromosome probes encompassing these loci demonstrated that the 10p region of deletion maps to 10p11.2. Losses involving 10p loci alone were most common in localized (5/14, 36%) and least common in metastatic (0/8) tumors. The common region of deletion on 10q included loci D10S219-D10S215, consistent with the major region of deletion recently defined for prostate tumors on 10q. Losses involving 10q loci alone were lowest in localized and locally invasive tumors (1/14 and 2/12, respectively) and highest in tumors metastatic to regional lymph nodes (2/8). These results suggest that 10p losses may define less invasive tumors, whereas 10q losses may play a role in the progression to more advanced tumor states in the prostate. Furthermore, this is the first report of allelic loss of a defined region on 10p potentially harboring tumor suppressor gene loci in human prostate cancer.

Fluorescence in situ hybridization analysis of 8p allelic loss and chromosome 8 instability in human prostate cancer.

Cytogenetic and molecular biological studies have demonstrated deletion of sequences that map to the short arm of chromosome 8 (8p) in tumors from several organ systems, including sequences that map within or near 8p22 in human prostate tumors. Recent studies in our laboratory have suggested that deletion of sequences on 8p may be concurrent with alterations in dosage for chromosome 8. In order to further investigate this finding, the present study has applied fluorescence in situ hybridization (FISH) techniques to determine the status of chromosome 8 in prostate tumors that have undergone deletion of sequences at 8p22. Dosage of 8p22 sequences was assayed utilizing unique sequence cosmid DNA probes by FISH and confirmed by amplification of microsatellite sequences by polymerase chain reaction (PCR). Chromosome 8 dosage was assayed by FISH utilizing both unique sequence cosmid probe DNA (specific to the 8q13.1-q13.3 chromosomal region) and pericentromeric probe DNA. FISH analysis of 10 specimens of normal or benign prostatic hyperplasia tissues paired with 9 tumor and one prostatic intraepithelial neoplasia tissues from the same patients for dosage at 8p, 8cen, and 8q, and PCR analysis for dosage at 8p, demonstrated that (a) FISH provided a more precise means of evaluating allelic loss than PCR in prostate tissue; (b) 8p22 sequence losses occurred frequently in prostate tumors; (c) 8p22 sequence losses were most often detected in the absence of 8cen or 8q sequence dosage alterations, although they were sometimes accompanied by gain or loss of 8cen or 8q sequences; and (d) the pattern of 8p22 sequence losses was most often widespread rather than focal. This study is the first to describe FISH analysis of interphase nuclei within tissue sections using cosmid probe DNAs.

Evidence for three tumor suppressor gene loci on chromosome 8p in human prostate cancer.

Allelic loss of human chromosome sequences is often equated with inactivation of putative tumor suppressor genes. Loss of sequences on the short arm of chromosome 8 (8p) has been observed in human cancers, especially of 8p22 in prostate tumors. By using PCR analysis of highly polymorphic microsatellite repeat markers at nine 8p loci in 135 tumors, we observed deletion of sequences at 8p22 and at two other proximal deletion domains. These novel deletion domains encompass the NEFL locus and D8S87-ANK1 loci, respectively. These data suggest that three 8p tumor suppressor gene loci may be independently deleted in human prostate cancers.

Activated in prostate cancer: a PDZ domain-containing protein highly expressed in human primary prostate tumors.

Critical events in prostate tumorigenesis and metastasis likely include the abnormal activation and expression of specific genes. Using RNA expression profiling techniques, we have identified a transcript originating from the activated in prostate cancer (AIPC) gene, the expression of which is preferentially up-regulated in several cultured prostate tumor cell lines and human primary prostate tumors. Sequence analysis revealed that the AIPC protein encodes six PDZ domains, which are protein-protein binding domains likely involved in protein clustering and scaffolding. Immunohistochemical analysis of a tissue microarray comprising 158 tumor, 18 high-grade prostatic intraepithelial neoplasia, and 91 normal prostate specimens with an anti-AIPC antibody demonstrated abundant AIPC protein expression in 75% of tumors, 83% of prostatic intraepithelial neoplasia lesions, and 3% of normal tissues (P < 0.0001). These data suggest that the accumulation of AIPC protein may be closely associated with the initiation or early promotion of prostate tumorigenesis.

Distinct regions of allelic loss on 13q in prostate cancer.

Loss of heterozygosity (LOH) involving the long arm of chromosome 13 has been reported to occur in as many as one third of primary prostate cancers. Candidate tumor suppressor genes on 13q that may be important in the development of prostate cancer include the retinoblastoma susceptibility gene (RBI) and a gene associated with inherited breast cancer (BRCA2). To define the pattern of allelic loss of 13q in prostate cancer, LOH analysis was performed using nine mapped polymorphic markers spanning the entire chromosomal arm. Nineteen (48%) of 40 prostate cancer cases obtained following radical prostatectomy demonstrated atllelic loss with at least one marker. Furthermore, 13 (33%) of 40 cases had evidence of allelic loss involving a region of 13q14 containing RB1. To test the hypothesis that RB1 is the targeted tumor suppressor gene in this region, 37 of 40 cases were assessed for expression of pRB, the protein product of RB1 using immunohistochemical techniques. By this analysis, 8 (22%) of 37 prostate tumors demonstrated no pRB expression. However, allelic loss at RB1, assessed with an intragenic marker, did not correlate with absent pRB expression (Fisher's exact test, P=0.375). Taken together, these data confirm that allelic loss of a common region of 13q14 occurs in approximately one third of prostate cancers. Lack of correlation of LOH at RB1 with absent pRB expression suggests the existence of another tumor suppressor gene in this region important in prostate cancer.

Allelic loss in locally metastatic, multisampled prostate cancer.

In order to determine whether retention or loss of potential tumor suppressor loci that map to 8p, 10q, or 16q reflect genetic relationships among prostatic intraepithelial neoplasias (PINs), multicentric primary prostatic cancers, and regional lymph node metastases or are associated with the metastatic phenotype, we analyzed 19 cases of locally metastatic prostate carcinoma (stage D1) utilizing polymerase chain reaction techniques. In each case, tissue samples from metastatic tumor, the (dominant) primary tumor, and nonneoplastic prostatic tissue were examined. In selected cases, allelic loss in additional tumor foci, separate from the dominant tumor nodule, and areas of PIN were examined. Allelic loss of sequences on 8p, 10q, and 16q were observed in 20-29% of PINs, 18-42% of primary tumors, and 8-25% of metastatic tumors. Discrepancies in sequence dosage between histological components were most pronounced for 8p sequences, especially between the dominant tumor nodule and metastatic deposits in cases in which > or = 3 separate tumor foci/gland were identified. These results suggest that putative premalignant lesions, moderately or poorly differentiated, geographically separate primary tumor foci, and metastases within morphologically "complex" prostates (those with > or = 3 foci/gland) are likely to be more discordant for sequence dosage at 8p than those within "simpler" glands (< 3 foci/gland). Also, our results suggest that lymph node metastases may be genetically related to either the dominant or additional primary tumor foci in more complex prostates and that accumulation of genetic aberration may differ in primary and metastatic lesions.

8pter-p23 deletion is associated with racial differences in prostate cancer outcome.

Deletions of chromosome sequences mapping to the short arm of chromosome 8 have been observed frequently in a variety of human cancers. A small number of studies have suggested that the terminal portion of the short arm of chromosome 8, 8pter-p23, may be deleted independently of other portions of 8p in human tumors, and that deletion of the 8pter-p23 region may be correlated with poor prognosis. The aim of the present study was to physically define the minimal region of 8pter-p23 deletion and to define the frequency and prognostic significance of 8pter-p23 loss in human prostate tumors. DNA was purified from normal and tumor tissues of 45 radical prostatectomy specimens and amplified for 15 highly polymorphic microsatellite sequences, 13 spanning 8pter-p23 and 2 proximal 8p markers. Allelic loss of 8p sequences was observed in 28 of 45 (62%) tumors examined. Of these, approximately half (12 of 28; 43%) demonstrated independent loss of the 8pter-p23 region, with several tumors defining a 5-cM minimal region of deletion spanning D8S264-D8S1824-D8S1781-D8S262-D8S1798. When serum prostate-specific antigen was used as a surrogate end point marker for survival, 8pter-p23 loss was significantly associated with reduced disease-free progression (log-rank P = 0.0426). Moreover, loss of the 8pter-p23 region was significantly associated with poor survival for American Caucasian (log-rank P = 0.0024) but not African-American (log-rank P = 0.5832) prostate cancer patients. These studies suggest that independent deletion of 8pter-p23 is differentially associated with disease recurrence and poor outcome in American Caucasian but not African-American prostate cancer patients.

Profiling and verification of gene expression patterns in normal and malignant human prostate tissues by cDNA microarray analysis.

cDNA microarray technology allows the "profiling" of gene expression patterns for virtually any cellular material. In this study, we applied cDNA microarray technology to profile changes in gene expression associated with human prostate tumorigenesis. RNA prepared from normal and malignant prostate tissue was examined for the expression levels of 588 human genes. Four different methods for data normalization were utilized. Of these, normalization to ACTB expression proved to be the most rigorous technique with the least probability of producing spurious results. After normalization to ACTB expression, 15 of 588 (2.6%) genes examined by array analysis were differentially expressed by a factory of 2x or more in malignant compared to normal prostate tissues. The expression patterns for 8 of 15 genes have been reported previously in prostate tissues (TGFbeta3, TGFBR3, IGFII, IGFBP2, VEGF, FGF7, ERBB3, MYC), but those of seven genes are reported here for the first time (MLH1, CYP1B1, RFC4, EPHB3, MGST1, BTEB2, MLP). These genes describe at least four metabolic and signaling pathways likely disrupted in human prostate tumorigenesis. Reverse transcriptase polymerase chain reaction (RT-PCR) and Northern blot analyses quantitated with reference to ACTB expression levels verified the trends in gene expression levels observed by array analysis for 14/15 and 8/8 genes, respectively. However, RT-PCR and Northern blot analyses accurately verified the "fold" differences in expression levels for only 6/15 (40%) and 7/8 (88%) of genes examined, respectively, demonstrating the need to better validate quantitative differences in gene expression revealed by array-based techniques.

Loss of expression of human spectrin src homology domain binding protein 1 is associated with 10p loss in human prostatic adenocarcinoma.

The gene encoding human spectrin Src homology domain binding protein 1, or Hssh3bp1, which is a marker of macropinocytic vesicles and a potential regulator of macropinocytosis, co-localizes to a YAC containing chromosome 10p sequences at loci D10S89 and D10S111 that are frequently deleted in prostate tumors. Expression of Hssh3bp1 was evaluated at the protein level in 17 paired normal and malignant prostate tumor samples using the monoclonal antibody 2G8 to Hssh3bp1. These experiments demonstrated that 4/6 tumors (67%) with 10p deletion failed to express Hssh3bp1 protein compared to 5/11 (46%) tumors with intact 10p. Thus, loss of Hssh3bp1 expression is concordant with allelic loss of adjacent 10p sequences in human prostate tumors. In addition, two prostate tumor cell lines contain an exon skipping mutation in the Hssh3bp1 gene that leads to the abnormal splicing of the mRNA and loss of a portion of Abl tyrosine kinase SH3 domain binding site in the protein. These data are consistent with a role for Hssh3bp1 as a candidate tumor suppressor gene inactivated during prostate tumorigenesis.

An approach to definition of genetic alterations in prostate cancer.

Growth patterns in prostatic cancer can reduce detectability of genetic alterations. Tumors show histologic grade heterogeneity, multifocality, interdigitation of benign and malignant glands, and varying amounts of stroma. These characteristics introduce sampling errors when one uses traditional methods for genetic analysis that depend on disaggregated cells [metaphase or interphase chromosome studies] or on tissue extracts [Southern blotting or polymerase chain reaction (PCR)] to detect molecular events. To circumvent these problems, we used two approaches to study paraffin-embedded tumors, which permit focused analysis of critical tissue components. Serial 4- to 5-microns sections are applied to slides in groups of three. Every second slide is hematoxylin and eosin stained to visualize areas of carcinoma, dysplasia, hyperplasia, and stroma; tumor-rich areas are circled with ink and used as templates to examine or excise the same areas from adjacent nonstained sections. PCR methods for quantitative and qualitative gene assay are effective in evaluating samples when alteration at a particular locus is suspected. Fluorescence in situ hybridization with chromosome-specific paracentromeric probes for detection of copy number of the relevant chromosome is applied to the adjacent section. Normal chromosome controls for both methods were demonstrated. This protocol enables us to correlate genetic alterations precisely with tumor extent and morphology.

Genetic characterization of immortalized human prostate epithelial cell cultures. Evidence for structural rearrangements of chromosome 8 and i(8q) chromosome formation in primary tumor-derived cells.

We have utilized a combination of conventional and spectral karyotyping (SKY) techniques and allelotype analysis to assess numerical and structural chromosome alterations in two cell lines derived from normal human prostatic epithelium, and three cell lines derived from human prostate primary tumor epithelium, immortalized with the E6 and E7 transforming genes of human papilloma virus (HPV) 16 or the large T-antigen gene of simian virus 40 (SV40). These studies revealed trisomy for chromosome 20 and rearrangements involving chromosomes 3, 4, 8, 9, 10, 16, 17, 18, 19, 21, or 22. In addition, the four HPV-immortalized cell lines exhibited extensive duplications or translocations involving the 11q chromosomal region. Interestingly, allelotyping data disclosed loss of 8p sequences in two of the three primary tumor-derived cell lines, and SKY data revealed that the loss of 8p sequences was directly due to i(8q) chromosome formation and/or other structural alterations of chromosome 8. This provides intriguing evidence that 8p loss in primary human prostate tumors may, in some cases, result from complex structural rearrangements involving chromosome 8. Moreover, the data reported here provide direct evidence that such complex structural rearrangements sometimes include i(8q) chromosome formation.

8p22 loss concurrent with 8c gain is associated with poor outcome in prostate cancer.

OBJECTIVES: A critical issue in the management of prostate cancer is the ability to distinguish patients at risk of disease recurrence. The aim of this study was to determine whether specific physical alterations of chromosome 8 may be associated with disease recurrence and poor outcome using postoperative prostate-specific antigen (PSA) values as surrogate end points. METHODS: To test this hypothesis, we examined paired normal and tumor radical prostatectomy tissues from 25 patients with prostate cancer for chromosome 8 alterations using dual fluorescence in situ hybridization with a fluorescein-labeled 8p22-specific (8p) cosmid probe and a rhodamine-labeled 8-centromere-specific (8c) probe. The probes were enumerated in 200 nuclei per tissue. RESULTS: Of the 25 tumors examined, 22 demonstrated distinct classes of genetic alterations, or nuclear types, including disomy for 8p and 8c (1 tumor), loss of 8p and disomy for 8c (10 tumors), or loss of 8p concurrent with gain of 8c (11 tumors). The presence of even a small population of tumor nuclei characterized by the loss of 8p concurrent with the gain of 8c was correlated with poor tumor grade (P = 0.009), preoperative PSA values 11 ng/mL or higher (P = 0.022), high tumor stage (P = 0.086), and detectable, rising postoperative PSA values (P = 0.086). These observations are consistent with the hypothesis that a gain of chromosome 8 is associated with poor outcome in prostate cancer. CONCLUSIONS: 8p loss concurrent with 8c gain may successfully predict disease recurrence and poor clinical outcome before the observation of detectable postoperative PSA values in patients with prostate cancer.

Allelic loss of 8p sequences in prostatic intraepithelial neoplasia and carcinoma.

OBJECTIVES: Previous work has suggested that prostatic intraepithelial neoplasia (PIN) may be a premalignant lesion important in tumorigenesis of the prostate. However, to adequately test this hypothesis at the genetic level, it is necessary to determine whether lesions in close proximity demonstrate similar genetic alterations and, hence, whether an "evolutionary" relationship might exist between PIN and tumor in the same prostate. Therefore, the purpose of this study was to examine at least two PIN lesions per prostate (one adjacent to and another distant from malignant lesions in the same prostate) for similarities or differences in the types and frequencies of genetic alterations. METHODS: To accomplish this goal, DNA was extracted from microdissected PIN, tumor, and normal epithelial tissue samples from 48 radical prostatectomies and amplified using polymerase chain reaction techniques at highly polymorphic microsatellite repeat sequences at proximal (D8S87, 8p12) and distal (NEFL, 8p21) loci on the short arm of chromosome 8. PIN specimens were either adjacent to (within one high-power microscopic field [HPF]) or distant from (separated by two or more HPFs) tumor specimens from the same patients. RESULTS: Similar fractional allelic loss frequencies were observed for informative tumor (10 [35%] of 29) and PIN (6 [21%] of 29) samples at the NEFL locus, but allelic loss at the D8S87 locus was observed only in tumors (8 [22%] of 36 informative samples). Moreover, allelic loss at the NEFL locus involved the same allele in 4 cases and different alleles in 3 cases. Interestingly, all 4 cases with the same allele loss were from adjacent PIN and tumor tissues, and all 3 with different allele loss were from distant PIN and tumor. CONCLUSIONS: These results suggest that PIN and invasive cancer share common genetic events (eg, deletion at the NEFL locus) along the same pathway of development in the prostrate.

Fluorescence In Situ Hybridization (FISH) to Metaphase and Interphase Chromosomes.

The unambiguous identification of human chromosomes became possible with the discovery and implementation of G-banding techniques (1). Almost immediately, investigators developed various methods to physically map specific DNA sequences to banded chromosomes. A commonly used early technique involved the hybridization in situ of radioactively labeled probes to heat-denatured human metaphase chromosomes (reviewed in 2). These techniques were efficient, yet costly, time-consuming, and technically difficult. Isotopic hybridization in situ was rapidly superseded by nonisotopic techniques-especially those utilizing fluorescently labeled probes (3-6). This chapter describes basic methodology for the accomplishment of metaphase and interphase fluorescence in situ hybridization (FISH).

Polymerase chain reaction detection of DNA sequence deletions.

Allelic loss of human chromosome sequences contributes to tumorigenesis through the inactivation of putative tumor-suppressor genes. The Knudson hypothesis proposes that deletion or mutation must affect both alleles of the gene in order to disable tumor suppression (1). As might be expected, the effect of "two hits" on tumor-suppressor gene integrity-e.g., deletion of one allele and mutation of the remaining allele-would disable the gene from encoding gene product. The von Hippel-Lindau (VHL) gene is an example of a tumorsuppressor gene that fulfills the Knudson hypothesis-e.g., one mutant allele is inherited in the germline, and the other is mutated or deleted somatically in many clear-cell renal cell carcinomas (recently reviewed in ref. 2).

Clinicopathologic significance of p53 immunostaining in adenocarcinoma of the breast.

Methanol/acetone-fixed frozen sections of 87 breast carcinomas were studied with a panel of three anti-p53 monoclonal antibodies that had specificities for wild-type, mutant, or combined wild-type plus mutant epitopes by using the avidin-biotin method. Nuclear staining was present in 13 (15%) of 87 cases with the mutant-specific antibody. The combined-specificity antibody stained 28 (32%) of 87 cases, including all but one of the tumors that was positive with the mutant-specific antibody. None of the cases reacted with the wild-type-specific antibody. Immunostaining for mutant form p53 was strongly correlated with adverse clinicopathologic factors, including poor differentiation, absence of estrogen receptor protein, nodal metastases, and large tumor size. In groups that were stratified by axillary node status, disease-free survival (52-month mean follow-up) was worse among cases with positive staining for either antibody. This difference was statistically significant in node-positive patients with the combined-specificity antibody (disease free, 22% [p53+] vs recurred, 57% [p53+]). We concluded that (1) immunostaining for mutant forms of p53 characterizes a clinically aggressive subset of breast tumors and may have prognostic utility in some patient populations, and (2) antibody-dependent-staining patterns for p53 may reflect epitope specificities of various mutant forms.

ADAM-mediated amphiregulin shedding and EGFR transactivation.

INTRODUCTION: The ectodomain shedding of epidermal growth factor receptor (EGFR) ligands, such as amphiregulin (AREG), by ADAMs (A Disintegrin And Metalloproteases) can be stimulated by G protein-coupled receptor (GPCR) agonists. Interactions between the CXCR4 GPCR and the CXCL12 chemokine have been shown to mediate gene transcription and cellular proliferation in non-transformed and transformed prostate epithelial cells, as well as motility/invasiveness in transformed cells. OBJECTIVES: In this report, we investigated the ability of CXCL12 to stimulate amphiregulin ectodomain shedding in non-transformed and transformed prostate epithelial cells that respond proliferatively to sub-nanomolar levels of CXCL12 and amphiregulin. MATERIALS AND METHODS: Non-transformed N15C6 and transformed PC3 prostate epithelial cells were assessed for amphiregulin shedding, ADAM activation, Src phosphorylation and EGFR activation using ELISA, immunoblot, and immunoprecipitation techniques, and for proliferation using cell counting after stimulation with CXCL12 or vehicle. RESULTS: The results of these studies identify CXCL12 as a novel inducer of amphiregulin ectodomain shedding and show that both basal and CXCL12-mediated amphiregulin shedding are ADAM10- and Src kinase-dependent in non-transformed N15C6 cells. In contrast, amphiregulin shedding is not amplified subsequent to stimulation with exogenous CXCL12, and is not reduced subsequent to metalloprotease- or Src kinase-inhibition, in highly aggressive PC3 prostate cancer cells. These data also show that CXCL12-mediated cellular proliferation requires EGFR transactivation in a Src- and ADAM-dependent manner in non-transformed prostate epithelial cells. However, these same mechanisms are dysfunctional in highly transformed prostate cancer cells, which secrete amphiregulin in an autocrine manner that cannot be repressed through metalloprotease- or Src kinase inhibition. CONCLUSION: These findings show that non-transformed and transformed prostate epithelial cells may employ different mechanisms to activate EGFR ligands and thereby utilize the EGFR axis to promote cellular proliferation.

Alterations in cellular gene expression without changes in nuclear matrix protein content.

Cell metabolism and function are modulated in part by cell and nuclear shape. Nuclear shape is controlled by the nuclear matrix, the RNA-protein skeleton of the nucleus, and its interactions with cytoskeletal systems such as intermediate filaments and actin microfilaments. The nuclear matrix plays an important role in cell function and gene expression because active genes are bound to the nuclear matrix whereas inactive genes are not. It is unknown, however, how genes move on and off the matrix, and whether these events require compositional protein changes, i.e., alterations in protein content of the nuclear matrix, or other, more subtle alterations and/or modifications. The purpose of this investigation was to begin to determine how nuclear matrix protein composition is related to gene expression. We demonstrate that gene expression can change without apparent changes in the protein composition of the nuclear matrix in MCF10A breast epithelial cells.