Prlz, a Novel Prostate-Specific and Androgen-Responsive Gene of the TPD52 Family, Amplified in Chromosome 8Q21.1 and Overexpressed in Human Prostate Cancer

[CANCER RESEARCH 64, 1589–1594, March 1, 2004] Advances in Brief

PrLZ, a Novel Prostate-Specific and Androgen-Responsive Gene of the TPD52 Family, Amplified in Chromosome 8q21.1 and Overexpressed in Human Prostate Cancer

Ruoxiang Wang,1 Jianchun Xu,1 Outi Sarama¨ki,4 Tapio Visakorpi,4 William M. Sutherland,5 Jianguang Zhou,1 Buer Sen,1 So Dug Lim,2 Nicola Mabjeesh,3 Mahul Amin,2 Jin-Tang Dong,3 John A. Petros,1 Peter S. Nelson,6 Fray F. Marshall,1 Haiyen E. Zhau,1 and Leland W. K. Chung1 1Molecular Urology and Therapeutics, Department of Urology, 2Department of Pathology, and 3Winship Cancer Institute, Emory University School of Medicine, Atlanta, Georgia; 4Laboratory of Cancer Genetics, Institute of Medical Technology, University of Tampere and Tampere University Hospital, Tampere, Finland; 5Department of Cell Biology, University of Virginia School of Medicine, Charlottesville, Virginia; and 6Division of Human Biology, Fred Hutchinson Cancer Research Center, University of Washington, Seattle, Washington

Abstract We studied the LNCaP/C4–2 progression model for clues to genetic and expressional changes, which may cause the transition We report a previously unrecognized prostate-specific protein, PrLZ from androgen-dependent to androgen-independent status. For this (prostate leucine zipper), a new member of the Tumor Protein D52 report, we isolated and characterized a novel prostate-specific (TPD52) family. The gene for PrLZ was localized at chromosome 8q21.1, a locus most frequently amplified in human prostate cancer. transcript, PrLZ, based on its differentiated expression between Multiple tissue analyses demonstrated PrLZ predominantly in the LNCaP and lineage-related C4–2 cells. Our data suggest that prostate gland. Although its expression was enhanced by androgens in overexpression of PrLZ is associated with PCa progression. PrLZ androgen receptor-expressing cells, PrLZ was detected in all of the may function to promote prostatic epithelial proliferation and human prostate cancer cell lines, regardless of androgen receptor transformation. status. Monoclonal anti-PrLZ antibodies were produced and intense immunohistochemical staining of PrLZ was observed in prostate epi- Materials and Methods thelial cells in intraepithelial neoplasia and prostate cancer, whereas lower-level staining was detected in normal and benign epithelial Cell Lines and Reagents. PCa cell lines PC3 and DU145 were obtained components of the prostate gland. As the only prostate-specific gene from American Type Culture Collection (Manassas, VA). C4–2 and identified in the most frequently amplified genomic region in prostate C4–2B cells were derived from chimeric LNCaP cell tumors (1). ARCaP cancer, PrLZ may be the link between chromosome 8q amplification was from a PCa metastasis (4). Cells were maintained in T-medium and malignant transformation of the prostate epithelia. (Invitrogen, Carlsbad, CA) with 10% of fetal bovine serum and antibiotics (penicillin, 100 unit/ml and streptomycin, 100 ␮g/ml) at 37°C with 5% CO Introduction 2 in a humidified incubator. To study the effect of androgen, cells were plated Metastasis of prostate cancer (PCa) results from the malignant on a 10-cm dish at ϳ50% confluence, subjected to androgen starvation in progression of localized tumor (known as prostatic intraepithelial phenol red-free RPMI 1640 for 48 h, and treated with a synthetic androgen, neoplasia; PIN), and is the main cause of the morbidity and methyltrienolone (R1881; Perkin-Elmer Life and Analytical Sciences, Bos- ton, MA), in fresh phenol red-free RPMI 1640 with 10% dextran/charcoal mortality of this disease. The mechanism for the development and absorbed fetal bovine serum for 12 h. progression of PCa has yet to be elucidated. To study the devel- Microarray Expression Analysis. Total RNA samples from LNCaP and opment and progression of PCa with a lineage-related cell model, C4–2 cells were used for synthesizing fluorescence-labeled cDNA probes we established a series of LNCaP sublines, including the C4, to screen an array of 1500 unique sequences derived from the prostate C4–2, and C4–2B cell lines, through tumor-stroma interaction and expression database (5). The detailed protocol used for analyzing prostate xenograft selection (1). The LNCaP/C4–2 PCa model mimics the gene expression has been described previously (5). Two hybridizations progression of clinical PCa. Whereas growth of the LNCaP is were performed for each sample, and for each arrayed gene four data points androgen-dependent, proliferating only in intact but not in cas- were collected for statistical comparison. trated male hosts, its derivative C4–2 and C4–2B sublines are DNA Cloning and Sequencing. A cDNA library of the C4–2 cells was androgen-independent, capable of growing in castrated male hosts. constructed into the ZAPExpress phage (Stratagene, La Jolla, CA) by the These sublines express androgen receptor (AR), secrete prostate manufacturer’s recommended protocol. A 330-bp insert from the original expressed sequence tag clone, identified through cDNA microarray, was specific antigen (PSA) in the absence of androgenic hormones, and used as a probe to screen 2 ϫ 105 plaque-forming units of the library. metastasize to lymph node and bone. Furthermore, these cells Positive clones rescued into pBK-CMV phagemid were subjected to re- exhibit bone-like properties, expressing osteoblastic factors and striction mapping and DNA sequencing. A human genomic library in secreting bone matrix proteins (2, 3). bacterial artificial chromosome was screened by PCR with 5Ј-GCCT- GAACTGTTTGTACC TCTG-3Ј and 5Ј-GAGTAGGTGATCCGGGTG- Received 10/22/03; revised 1/15/04; accepted 1/19/04. GAGATG-3Ј as primers. Restriction fragments were subjected to nested Grant support: United States Department of Defense (DAMD 17–03-2–0033; deletion with the Erase-A-Base kit (Promega, Madison, WI). Manual DNA L. W. K. Chung, H. E. Zhau, R. Wang; and DAMD PC991274; P. S. Nelson) and sequencing was performed with the ⌬Taq Sequenase II kit (USB, Cleve- NIH (CA098912 and CA76620; L. W. K. Chung; CA82739, H. E. Zhau; and CA96994, J. A. Petros). land, OH), and automated sequencing was on an ABI sequencer (Applied The costs of publication of this article were defrayed in part by the payment of page Biosystems, Foster City, CA). charges. This article must therefore be hereby marked advertisement in accordance with Expression and Multiple Tissue Expression Assays. The PrLZ-specific 18 U.S.C. Section 1734 solely to indicate this fact. fragment was cloned to pGEM-T easy (Promega) after amplification of Requests for reprints: Ruoxiang Wang, Department of Urology, Emory University Ј Ј School of Medicine, 1365B Clifton Road, Suite B5204, Atlanta, GA 30322. Phone: the PrLZ cDNA with primers: 5 -GCCTGAACTGTTTGTACCTCTG-3 (404) 778-5116; Fax: (404) 778-3965; E-mail: [email protected]. and 5Ј-GAGTAGGTGATCCGGGTGGAGATG-3Ј. Similarly, coding se- 1589

Downloaded from cancerres.aacrjournals.org on October 1, 2021. © 2004 American Association for Cancer Research. A NOVEL PROSTATE-SPECIFIC PrLZ GENE IN PROSTATE CANCER quences of the PSA and prostate-specific membrane antigen (PSMA) were cloned with primers 5Ј-ATGTGGGTCCCGGTTGTCTTCCTCAC- CCTGTC-3Ј and 5Ј-TCAGGGGTTGGCCACGATGGTGTCCTTGATC-3Ј (PSA); and 5Ј-ATAGGATCCATGTGGAATCTCCTTCACGAAACC-3Ј and 5Ј-CATAAGCTTTTAGGCTACTTCACTCAAAGTCTC-3Ј (PSMA) from prostate RNA (Clontech, Palo Alto, CA). Manufacturer-recommended protocols were used in hybridizations to the Multiple Tissue Expression array, the Multiple Tumor Expression array, and the Human Total RNA Master Panel (Clontech). Immunohistochemical Staining. Polyclonal antibodies were produced by immunizing mice with synthetic peptides, followed by establishing and screening hybridoma cells for monoclonal antibody production (University of Virginia Hybridoma Core Facility). Antibody specificity was confirmed by Western blot and immunoprecipitation. Preimmune sera or control hybridoma fluid were used as negative control. Multiple tumor arrays prepared from 100 radical prostatectomies performed at Emory University were subjected to immunohistochemical staining with the anti-PrLZ antibody. The data were evaluated by two pathol- Fig. 1. PrLZ isolation. A, Northern blot hybridization to confirm differentiated ogists (S. D. L. and M. A.). PrLZ staining was set as low if 0–15% of the gene expression. After microarray screening, insert from selected clones was used as Ͼ a probe to confirm differentiated expression in LNCaP and C4–2 cells. In each panel, cells were stained stronger than the normal level, and high if 15% cells the specific hybridization of the gene being studied is shown at the top.Atthebottom, were strongly stained. The data were subjected to Fischer’s exact test for the same membrane was then hybridized to ␤-actin probe to determine equal loading. cross-table analysis, with P Ͻ 0.05 considered statistically significant. Except for PrLZ, other candidate genes are indicated with their array coordinates. B, PCR Analyses. All of the PCR reactions were initiated with incubation at schematic of the relationship between PrLZ and TPD52. Shaded area indicates coding region and slashed area shares identical sequence among cDNA clones. Probe (P) was 94°C for 2 min, followed by 30 cycles of 94°C, 30 s; 55°C, 30 s; and 72°C, 2 used to screen the cDNA library. PrLZ-502 is a fragment containing the PrLZ-specific min. Reactions were finished with a 72°C, 7-min extension. Primers used region and was used as probe in hybridizations. C, prostate-specific expression of were: 5Ј-CCTGAACTGTTTGTACCTCTGGGCCATATTGC-3Ј and 5Ј- PrLZ as determined on Multiple Tissue Expression array. PrLZ-502 was used to detect CAAATTTCTGAAGAGTAGGTGATCCGGGTGGAG-3Ј for exon 1 of the specific PrLZ expression. Predominant PrLZ expression was seen in prostate (coor- Ј Ј dinate E8), with minimal expression seen in the gastrointestinal system, including PrLZ gene; 5 -TCTAAAGTAGGGGGAACCAAGCCTGCTGGTGGTG-3 stomach (B5), duodenum (C5), jejunum (D5), ileum (E5), ilocecum (F5), appendix and 5Ј-ACTGATAGATGGAATTTATTAAGCTTTTCACATG-3Ј for exon 7 (G5), ascending colon (H5), transverse colon (A6), descending colon (B6), and rectum of the PrLZ gene; 5Ј-CAAGTTAACTGAGCTTTTTCTTAATTTCATTC (C6). Still lower levels of expression were seen in kidney (A7), pituitary (D3), salivary C-3Ј and 5Ј-GTTGGTACCTCCACAGAAGATGTTTATTTGATGTAAC-3Ј (E9), and mammary gland (F9). G12 and H12 represent 100 ng and 500 ng control Ј human DNA, respectively. Examined on the array were 64 normal adult human for 3 untranslated region of the HIF1A gene. Primers for amplification of the tissues, 8 embryonic tissues, and a series of cell lines. A list of these samples can mitochondrial DNA were 5Ј-AGTCAATAGAAGCCGGCG-3Ј and 5Ј-GGG- be found elsewhere.7 The same membrane was hybridized with ␤-tubulin probe GATTTAGAGGGTTCTGT-3Ј. Comparison of the expression of PrLZ and provided by the manufacturer. Ubiquitous and uniform signals were seen (data not TPD52 was done with gene-specific primer pairs of 5Ј-ATGGATTGTA- shown). GAGAGATGGACTTATATGAGG-3, 5Ј-TCACAGGCTCTCCTGTGTCTTT TCTGGAAGAGG-3Ј (PrLZ), and 5Ј-ATGGACCGCGGCGAGCAA-3Ј,5Ј- TCACAGGCTCTCCTGTGTCTTTTCTGGAAGAGG-3Ј (TPD52). Fluorescence in Situ Hybridization. The PrLZ-specific probe was labeled with either digoxigenin (Roche Diagnostics, Mannheim, Germany) or (Fig. 1B). Homology search through the GenBank revealed that Alexa Fluor 594-dUTP (Molecular Probes, Eugene, OR) by nick-transla- one represented a previously reported TPD52 (7), except the new tion, and hybridized together with FITC-dUTP (DuPont, Boston, MA) isolate had a longer 5Ј untranslated region and a shorter 3Ј un- -labeled chromosome 8 centromeric probe (pJM128) to metaphase chro- translated region. The other transcript contained a unique 502-bp mosomes as described (6). Slides were counterstained with 0.1 ␮M 4,6- sequence in its 5Ј region, completely divergent from that of diaminido-2-phehylindole in an antifade solution. Assignment of the gene to chromosome was based on 4,6-diaminido-2-phehylindole banding TPD52. It encoded a polypeptide with a 41-residue unique NH2 pattern. terminus, 35 residues longer than the TPD52 protein (Fig. 1B; Fluorescence in situ hybridization was used to determine PrLZ gene GenBank accession no. AF202897). We named this new isolate amplification in PCa specimens. Forty formalin-fixed, paraffin-embedded, PrLZ, and characterized its transcript and protein product. locally recurrent, hormone-refractory specimens (transurethral resection of The full cDNA for PrLZ was 2573 bp in size, encoding a 224 amino prostate) from Tampere University Hospital were analyzed (6). PrLZ acid residue polypeptide of 24.4 kDa, which was confirmed with in amplification was classified into three groups, no amplification (no in- vitro transcription coupled translation, with the full cDNA as template crease in PrLZ signal number), low-level amplification (3– 4 copies of signals per cell), and high-level amplification (Ն5 copies of signals or (data not shown). PrLZ protein would be acidic (pI 4.64) and hydro- clusters of signals per cell). philic. Many structural features of the TPD52 family were conserved in PrLZ. These included a coiled coil leucine zipper in the central Results region, two PEST domains flanking the leucine zipper, and multiple canonical serine/threonine phosphorylation sites, substrates for casein PrLZ Isolation. We studied 1500 arrayed genes for their ex- II kinase, protein kinase C, and cyclic AMP- and cyclic guanosine pression in LNCaP and C4–2 cells. Comparative analysis and 3Ј,5Ј-monophosphate-dependent kinases. PrLZ harbored two addi- subsequent Northern blot hybridization identified 8 clones with tional N-glycosylation sites and an NH -terminal citron motif in its Ͼ2-fold higher expression in C4–2 than in LNCaP (Fig. 1A). One 2 unique NH terminus. In cultured cells, the majority of the PrLZ- of the clones showed a 9-fold higher expression in C4–2 (PrLZ; 2 fusion protein was seen in cytoplasm, as determined by tagging PrLZ Fig. 1A). Using this expressed sequence tag clone (probe P; Fig. with green fluorescent protein and by Western blotting of the sub- 1B) as probe, we screened a cDNA library of C4–2 cells and isolated 24 positive clones. DNA sequencing analyses revealed cellular fractions (data not shown). that these clones were from two transcripts that shared a homologous sequence in their 3Ј half, but contained distinctive 5Ј regions 7 Internet address: http://www.clontech.com. 1590

Fig. 2. PrLZ gene amplification at chromosome 8q21.1. A, localization of PrLZ gene. Shown is a representative micrograph of the fluorescence in situ hybridization analysis (FISH), with red signals defining PrLZ at chromosomal 8q21.1 and green signals chromosome 8 centromeres. Inset is a magnification of the chromosome 8 and chromosome 8 ideogram. B, structure of the PrLZ gene. Genomic structure of PrLZ is shown at the top, with solid bars indicating exons. The corresponding region in cDNA is indicated in the lower portion. C, amplification of PrLZ gene in prostate cancer (PCa) specimens as detected by PCR. In this study, paired genomic DNA samples isolated from cancer cells (T), and unaffected glandular cells (N) of the same PCa specimen (depicted with case numbers), were used to detect the first and last exons of the PrLZ gene. Equal amounts (100 pg) of DNA were used in each sample. In the control lane (C), 10 ng of normal human genomic DNA (Clontech) was used. PrLZ gene was amplified in four of the seven cases tested (266T, 316T, 285T, and 320T). The equal amount of genomic DNA was controlled by PCR of 3Ј untranslated region of the hypoxia inducible factor 1␣ (HIF1A), which is rarely amplified in PCa (21). Mitochondrial DNA (MtDNA) was amplified for similar purposes. D, amplification of PrLZ gene in PCa specimens as detected by FISH. In contrast to the signal in normal prostate epithelial cells (top left panel), nuclei of PCa cases showed low-level amplification (top right panel) and high-level amplification of the PrLZ gene. Amplification is seen as scattered (bottom left panel) or clustered (bottom right panel) hybridization signals. Green signals outside the nuclei were due to autofluorescence, typical for FISH analysis with tissue sections.

Prostate Specificity of PrLZ. The first 502 bp of the PrLZ family, there were alternatively spliced messengers of the PrLZ cDNA, determined as unique and PrLZ-specific (probe PrLZ-502; gene. Distinct from other genes of the TPD52 family, we found Fig. 1B), was used as a probe to study the expression of PrLZ in that alternative splicing events were from unique locations of the various tissue and cell types. PrLZ was expressed predominantly in PrLZ gene, producing novel polypeptides (data not shown). human prostate (Fig. 1C), with only minimal expression in the Amplification of 8q is the most frequent genomic abnormality in gastrointestinal tract and a few other glandular tissues with secre- clinical PCa (8). We examined copy numbers of the PrLZ gene in PCa tory functions (i.e., pituitary, salivary, and mammary glands, pan- specimens by two methods. PCR was used to detect the first and the creas, and kidney). The expression in prostate was at least 12-fold last exons. A group of 7 paired samples from radical prostatectomy higher than in any other tissues. Importantly, PrLZ was not de- were examined. Genomic DNA was prepared from cancer cells iso- tected in the rest of the 52 different embryonic and adult human lated by laser capture microdissection. Matched control was from tissues, nor in a list of human cell lines of extraprostatic origin. morphologically normal glandular cells distant from the tumor- Similar prostate-specific expression of PrLZ was found in addi- affected area. This study detected varied levels of amplification in 4 tional studies with the Multiple Tumor Expression array and with of the 7 cases (Fig. 2C). Definitive confirmatory data were obtained the Human Total RNA Master Panel (data not shown). In contrast, by fluorescence in situ hybridization analysis of 40 PCa specimens TPD52 was expressed in numerous human tissues without evi- with the 502-bp PrLZ-specific probe. This study revealed that 27 of dence of tissue specificity. the 40 PCa cases (67.5%) harbored amplified PrLZ gene, with 9 Chromosomal Localization and Amplification of the PrLZ showing high copy number gain and 18 low-level amplification Gene. Taking advantage of the gene specificity of the 502-bp (Fig. 2D). fragment, we localized the PrLZ gene at human chromosome Regulated Expression of PrLZ. Under regular culture condi- 8q21.1 (Fig. 2A). The PrLZ gene was isolated in a 180-kb genomic tions (T-medium with 10% fetal bovine serum), PrLZ was ex- fragment in a bacterial artificial chromosome clone. Exons were pressed in all of the tested PCa cell lines, independent of the AR mapped and corresponding restriction fragments subcloned for status (Fig. 3A). This was in sharp contrast with PSA and PSMA, DNA sequencing. The PrLZ gene was also contained in the draft which were seen only in the AR-expressing LNCaP and its lineage sequence of 8q21.1 (GenBank accession no. NT_023700). PrLZ derivatives. To investigate the regulation of PrLZ by androgens, messenger was transcribed from 7 exons, which were scattered in we treated the PCa cells with an androgen analog, R1881 (1 nM), a 45-kb region (Fig. 2B). Similar to other members of the TPD52 after androgen deprivation. After culturing the PCa cell lines under 1591

Fig. 3. Expression of the PrLZ in prostate cancer cell lines. A, detection of PrLZ in prostate cancer cell lines by reverse transcription-PCR. In this study, the 502-bp gene-specific sequence in 5Ј of the PrLZ cDNA, and the full coding sequences of TPD52 and prostate-specific antigen were subjected to PCR amplification. The experiment was repeated at least three times; a representative result is shown. B, PrLZ expression is enhanced by androgen, as detected by Northern blot hybridization with PrLZ as probe (top panel) and reverse transcription-PCR detecting the 502-bp gene-specific sequence in 5Ј of the PrLZ cDNA (bottom panel). Expression of prostate-specific antigen (positively regulated by androgen) and PSMA (negatively regulated by androgen) was used to determine the effect of androgen. Full coding sequence of the TPD52 was amplified to show the distinctive response of PrLZ to androgen. The experiment was repeated three times; a representative result is shown. C, detection of PrLZ protein by Western blotting. In the top panel, control vector (lane 1), FLAG-tagged PrLZ (lane 2), and FLAG-tagged TPD52 (lane 3) were stably overexpressed in HEK293 cells, which did not express PrLZ. Whole cell lysates were subjected to immunoblotting with anti-FLAG (left) or anti-PrLZ antibody (right). Anti-PrLZ antibody specifically recognized the PrLZ protein (right). The bottom panel shows a representative result in which the anti-PrLZ antibody was used in immunoblotting to detect endogenous PrLZ expression in LNCaP and C4–2 cell lines in the presence (ϩ) or absence (Ϫ) of androgen (R1881; 1 nM).

this condition (phenol red-free RPMI 1640, fetal bovine serum- 4B). It was highly expressed in high-grade prostatic intraepithelial free, for 48 h), the PrLZ signal markedly declined. Nonetheless, neoplasia (PIN) (84.5% being high) and PCa (75% being high; Fig. androgen promoted the expression of PrLZ in LNCaP and its 4B). Gleason grade 4 tumors stained more than Gleason grade 3 lineage-derivative cells (Fig. 3B). Thus, PrLZ expression could be tumors. Remarkably, intense PrLZ staining was always limited to regulated by serum and androgen. In contrast, TPD52 was uni- malignant cells, with the neighboring unaffected epithelial cells formly expressed (Fig. 3A), its expression neither affected by stained at the normal level (Fig. 4B). Enhanced expression of PrLZ androgen deprivation nor by androgen stimulation (Fig. 3B). Com- was, therefore, tumor cell specific. pared with its regulation of PSA, androgen treatment seemed to have distinguishable effect on PrLZ expression. The effect of Discussion R1881 on PSA expression was seen only in LNCaP cells, whereas the constitutive expression in C4–2 and C4–2B was largely unaf- PrLZ as a Prostate-Specific Protein and New PCa Marker. fected (Fig. 3B). In contrast, androgen-enhanced PrLZ expression Data from all of our expression studies strongly supported thefind- was seen in both parental and LNCaP-lineaged cells, with the ing that PrLZ is predominantly expressed in prostate, with minimal highest up-regulation seen in C4–2B cells. The transition from expression seen in a few other glandular organs. One unique androgen-dependent to androgen-independent state seemed a het- feature of PrLZ expression is that, in contrast with other known erologous process: whereas LNCaP sublines lost the control for prostate-specific proteins, which are found only in AR-expressing PSA expression, these cells still maintain the capability of con- LNCaP and its lineaged cells, PrLZ expression was detected in trolling some other androgen responsive genes, such as PrLZ. all of the PCa cell lines regardless of their AR status (Fig. 3, A To study the expression of PrLZ at the protein level, we raised and B). polyclonal and monoclonal antibodies by immunizing mice with a PrLZ appears to be a new member of the TPD52 family, a group synthetic peptide, based on the unique NH2-terminal coding se- of homologous proteins identified due to overexpression of D52 in quence of PrLZ. The antigen specificity of the antibodies was tumors (7, 9, 10). Distinctively, PrLZ expression is prostate- assessed by Western blotting against the endogenous PrLZ protein specific and androgen-responsive, whereas TPD52 could be de- (Fig. 3C) and by immunoprecipitation analysis of the FLAG- tected in many tissue and cell types (10), and is not affected by tagged PrLZ (data not shown). androgen (Fig. 1C and Fig. 3B). Using a gene-specific probe, we PrLZ Expression in PCa Specimens. Using the specific anti- localized the PrLZ gene to chromosome 8q21.1 (Fig. 2A) and bodies, we determined the expression of PrLZ in association with determined the gene structure by studying a bacterial artificial clinical PCa. Specimens from 100 PCa cases were used. PrLZ was chromosome clone (Fig. 2B). Previously, other laboratories used generally low in unaffected secretory epithelia (75.3% being low) the full cDNA as a probe to localize the TPD52 gene on 8q, and in benign prostatic hyperplasia (BPH) (78.2% being low; Fig. proximate to the locus in which the PrLZ gene was localized (7, 1592

on AR presence (Fig. 2, B and C). Thus, enhanced PrLZ expression in PCa is correlated to malignancy rather than to AR activity. Oncogenic Property of PrLZ. The biological function of PrLZ

is unclear. Besides its unique NH2-terminal structure, PrLZ con- tains conserved domains of the TPD52 family. Members of the TPD52 family are involved in protein-protein interaction (11–13). Annexin VI (14), MAL2 (13), syntaxin I, and VAMP2 (15) have been identified as interacting partners of TPD52 proteins. 14–3-3, a crucial player in Ras signaling, vesicular transport, and cytoskel- etal organization, was revealed recently as another interacting partner (11). Also, TPD52 proteins could be modified post-trans- lationally with phosphorylation, N-glycosylation, or protease di- gestion. In a separate study we found that PrLZ was highly expressed in the developing prostate gland (data not shown) with a spatial and temporal expression suggesting that PrLZ may help regulate growth, morphogenesis, and cytodifferentiation of glandular ducts during development (16). The role of PrLZ in PCa development and progression has yet to be elucidated. The coding sequence of the PrLZ in PCa cell lines contained no mutations. PrLZ cDNA, when transfected to ARCaP cells, markedly stimulated their growth in immune-deficient mice (data not shown). Other TPD52 family members have been found to be associated with cancers in multiple tissues and organs (7, 10, 17). In addition, retroviral integration in avian TPD52 (R10) was accompanied by neuroepithelial proliferation (18). In a leukemic (HL-60 and K-562) cell differentiation model, expression of TPD52 members was associated with cell proliferation (9). PrLZ gene is located in 8q21. Amplification of this gene in PCa is independent from 8q23–24 amplification (19). Within the 8q21 Fig. 4. Abnormally enhanced PrLZ expression in prostate cancer (PCa) specimens. amplicon, we found previously that TCEB1 gene was ϳ6 mb pairs Representative data from immunohistochemical studies of 100 PCa specimens are centromeric to the PrLZ (20). Fluorescence in situ hybridization shown. A, PrLZ level in a normal, healthy, 42-year-old prostate, with low but discernible staining seen in glandular epithelia. B, PrLZ expression in PCa nodules analyses revealed that PrLZ and TCEB1 genes were coamplified in with adjacent BPH nodular region. Whereas the PCa nodules were heavily stained, PCa specimens (data not shown). Additional investigation would BPH in their vicinity was low for PrLZ, comparable with the normal level. C, PrLZ clarify whether both of these genes are involved in malignancies of in high-grade PIN with adjacent normal glands. Epithelia of the high-grade PIN exhibited a general up-regulation of PrLZ; adjacent normal or unaffected epithelial the prostate gland. cells displayed low level staining. D, PrLZ expression in PCa nodules with neigh- In summary, we have identified a prostate-specific and andro- boring normal or unaffected glands. In contrast to intense PrLZ staining in PCa cells, normal or unaffected glands in the immediate neighborhood showed low level gen-inducible gene, PrLZ, located at the most frequently amplified staining. E, uniformly high expression of PrLZ in PCa cells of Gleason score 3 tumor. 8q chromosome region in human PCa. PrLZ was seen in the F, representative specimens showed prevalently high PrLZ expression in PCa majority of the high-grade PIN specimens. Possibly up-regulation of typical Gleason score 4 tumors, with higher PrLZ staining than Gleason score 3 tumors. of PrLZ is an early sign of malignant transformation. PrLZ expression persisted in locally recurrent, hormone-refractory, and metastatic PCa. Additional elucidation of the function of PrLZ in 10). We have not been able to determine whether the transcripts for PCa cells could provide insight into the malignant progression of PrLZ and TPD52 are derived from the same gene by selective PCa cells. usage of promoters or arise from separate genes in close vicinity. We determined the genomic organization of the PrLZ by partially Acknowledgments sequencing a bacterial artificial chromosome clone, which contained all of the exons of PrLZ within a 50-kb span. We did not We thank Drs. Lara Harik, Fan Yeung, Yuanyuan Cui, and Zhihui Xie for technical assistance. We also acknowledge the technical support provided by find TPD52-specific exon in the PrLZ gene or within a 12-kb 5Ј Dr. Jay Fox of the Core Facilities at the University of Virginia for the chemical flanking region. In the draft sequences of the human genome, a synthesis of PrLZ peptides. TPD52-specific exon is located 107 kb upstream from the second exon of the PrLZ gene, raising the possibility that PrLZ and TPD52 References are splicing variants of one single gene. Both PrLZ and TPD52 are 1. Thalmann, G. N., Anezinis, P. E., Chang, S. M., Zhau, H. E., Kim, E. E., Hopwood, abnormally expressed in malignant tissues, with PrLZ seen specif- V. L., Pathak, S., von Eschenbach, A. C., and Chung, L. W. Androgen-independent ically in PCa and TPD52 in many tumors. Studying PrLZ may cancer progression and bone metastasis in the LNCaP model of human prostate elucidate the biological and pathophysiological function of the cancer. Cancer Res., 54: 2577–2581, 1994. 2. Keller, E. T. The role of osteoclastic activity in prostate cancer skeletal metas- TPD52 family. tases. Drugs Today (Barc.), 38: 91–102, 2002. Prostate specificity and enhanced expression in high-grade PIN 3. Taichman, R. S., Cooper, C., Keller, E. T., Pienta, K. J., Taichman, N. S., and and PCa make PrLZ an attractive diagnostic marker. Androgen- McCauley, L. K. Use of the stromal cell-derived factor-1/CXCR4 pathway in prostate cancer metastasis to bone. Cancer Res., 62: 1832–1837, 2002. independent, tumorigenic, and bone metastatic C4–2 expressed 4. Zhau, H. Y., Chang, S. M., Chen, B. Q., Wang, Y., Zhang, H., Kao, C., Sang, higher PrLZ levels than LNCaP in the LNCaP/C4–2 PCa progres- Q. A., Pathak, S. J., and Chung, L. W. Androgen-repressed phenotype in human prostate cancer. Proc. Natl. Acad. Sci. USA, 93: 15152–15157, 1996. sion model (Fig. 1A; Fig. 2, B and C; and Fig. 3B). Although it can 5. Lin, B., White, J. T., Ferguson, C., Bumgarner, R., Friedman, C., Trask, B., Ellis, be activated through liganded AR, PrLZ expression does not rely W., Lange, P., Hood, L., and Nelson, P. S. PART-1: a novel human prostate- 1593

specific, androgen-regulated gene that maps to chromosome 5q12. Cancer Res., interactions with TPD52-like proteins in the yeast two-hybrid system. Genomics, 60: 858– 863, 2000. 76: 81– 88, 2001. 6. Saramaki, O., Willi, N., Bratt, O., Gasser, T. C., Koivisto, P., Nupponen, N. N., 14. Thomas, D. D., Kaspar, K. M., Taft, W. B., Weng, N., Rodenkirch, L. A., and Bubendorf, L., and Visakorpi, T. Amplification of EIF3S3 gene is associated with Groblewski, G. E. Identification of annexin VI as a Ca2ϩ-sensitive CRHSP-28- advanced stage in prostate cancer. Am. J. Pathol., 159: 2089–2094, 2001. binding protein in pancreatic acinar cells. J. Biol. Chem., 277: 35496–35502, 7. Byrne, J. A., Tomasetto, C., Garnier, J. M., Rouyer, N., Mattei, M. G., Bellocq, 2002. J. P., Rio, M. C., and Basset, P. A screening method to identify genes commonly 15. Proux-Gillardeaux, V., Galli, T., Callebaut, I., Mikhailik, A., Calothy, G., and overexpressed in carcinomas and the identification of a novel complementary Marx, M. D53 is a novel endosomal SNARE-binding protein that enhances DNA sequence. Cancer Res., 55: 2896–2903, 1995. interaction of syntaxin 1 with the synaptobrevin 2 complex in vitro. Biochem. J., 8. Visakorpi, T., Kallioniemi, A. H., Syvanen, A. C., Hyytinen, E. R., Karhu, R., 370: 213–221, 2003. Tammela, T., Isola, J. J., and Kallioniemi, O. P. Genetic changes in primary and 16. Davies, J. A. Do different branching epithelia use a conserved developmental recurrent prostate cancer by comparative genomic hybridization. Cancer Res., 55: mechanism? Bioessays, 24: 937–948, 2002. 17. Balleine, R. L., Fejzo, M. S., Sathasivam, P., Basset, P., Clarke, C. L., and Byrne, 342–347, 1995. J. A. The hD52 (TPD52) gene is a candidate target gene for events resulting in 9. Byrne, J. A., Mattei, M. G., and Basset, P. Definition of the tumor protein D52 increased 8q21 copy number in human breast carcinoma. Genes Chromosomes (TPD52) gene family through cloning of D52 homologues in human (hD53) and Cancer, 29: 48–57, 2000. mouse (mD52). Genomics, 35: 523–532, 1996. 18. Proux, V., Provot, S., Felder-Schmittbuhl, M. P., Laugier, D., Calothy, G., and Marx, M. 10. Chen, S. L., Maroulakou, I. G., Green, J. E., Romano-Spica, V., Modi, W., Characterization of a leucine zipper-containing protein identified by retroviral insertion in Lautenberger, J., and Bhat, N. K. Isolation and characterization of a novel gene avian neuroretina cells. J. Biol. Chem., 271: 30790–30797, 1996. expressed in multiple cancers. Oncogene, 12: 741–751, 1996. 19. Nupponen, N. N., Kakkola, L., Koivisto, P., and Visakorpi, T. Genetic alterations 11. Boutros, R., Bailey, A. M., Wilson, S. H., and Byrne, J. A. Alternative splicing in hormone-refractory recurrent prostate carcinomas. Am. J. Pathol., 153: 141– as a mechanism for regulating 14–3-3 binding: interactions between hD53 148, 1998. (TPD52L1) and 14–3-3 proteins. J. Mol. Biol., 332: 675– 687, 2003. 20. Porkka, K., Saramaki, O., Tanner, M., and Visakorpi, T. Amplification and 12. Sathasivam, P., Bailey, A. M., Crossley, M., and Byrne, J. A. The role of the overexpression of Elongin C gene discovered in prostate cancer by cDNA coiled-coil motif in interactions mediated by TPD52. Biochem. Biophys. Res. microarrays. Lab. Invest., 82: 629– 637, 2002. Commun., 288: 56– 61, 2001. 21. Saramaki, O. R., Savinainen, K. J., Nupponen, N. N., Bratt, O., and Visakorpi, T. 13. Wilson, S. H., Bailey, A. M., Nourse, C. R., Mattei, M. G., and Byrne, J. A. Amplification of hypoxia-inducible factor 1␣ gene in prostate cancer. Cancer Identification of MAL2, a novel member of the mal proteolipid family, though Genet. Cytogenet., 128: 31–34, 2001.

1594

Downloaded from cancerres.aacrjournals.org on October 1, 2021. © 2004 American Association for Cancer Research. PrLZ, a Novel Prostate-Specific and Androgen-Responsive Gene of the TPD52 Family, Amplified in Chromosome 8q21.1 and Overexpressed in Human Prostate Cancer

Ruoxiang Wang, Jianchun Xu, Outi Saramäki, et al.

Cancer Res 2004;64:1589-1594.

Updated version Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/64/5/1589

Cited articles This article cites 20 articles, 8 of which you can access for free at: http://cancerres.aacrjournals.org/content/64/5/1589.full#ref-list-1

Citing articles This article has been cited by 11 HighWire-hosted articles. Access the articles at: http://cancerres.aacrjournals.org/content/64/5/1589.full#related-urls

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://cancerres.aacrjournals.org/content/64/5/1589. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.