Fatty acid synthase gene overexpression and copy number gain in prostate adenocarcinoma

Human Pathology (2006) 37, 401 – 409

www.elsevier.com/locate/humpath

Fatty acid synthase gene overexpression and copy number gain in prostate adenocarcinomaB Uzma S. Shah PhDa, Rajiv Dhir MDb, Susanne M. Gollin PhDc,d, Uma R. Chandran PhDb, Dale Lewisc,d, Marie Acquafondatab, Beth R. Pflug PhDa,d,* a

Department of Urology, University of Pittsburgh, Pittsburgh, PA 15232, USA Department of Pathology, University of Pittsburgh, Pittsburgh, PA 15232, USA c Department of Human Genetics, University of Pittsburgh Graduate School of Public Health, Pittsburgh, PA 15261, USA d University of Pittsburgh Cancer Institute, Pittsburgh, PA 15232, USA b

Received 19 September 2005; revised 18 November 2005; accepted 21 November 2005

Keywords: Prostate cancer; Fatty acid synthase; Cytogenetics; FISH; IHC; Gene copy number

Summary Cancer cells frequently exhibit a significant increase in overexpression and activity of fatty acid synthase (FASN). Elevated FASN pathway activity also occurs in prostate cancer, the second leading cause of cancer-related death in men in the United States. Studies show that genes associated with an increase in protein expression, such as HER2/neu in breast cancer, are associated with an increase in gene copy number as well as an increase in transcription. In the present study, we evaluated whether FASN follows a similar paradigm in prostate cancer. To date, elevated FASN expression in prostate cancer has not been correlated with gene copy number alterations. Using immunohistochemistry and fluorescence in situ hybridization analysis in paraffin-embedded tissue microarrays, we observed gene copy gain in 24% of all prostate adenocarcinoma specimens examined with concurrent increased FASN protein expression. Immunohistochemistry alone showed 59% of prostate cancer specimens in the same tissue microarray with high FASN expression. Increased FASN gene was observed in 53% of all prostate tissues expressing elevated FASN protein levels and in 2 of 5 prostate tumor cell lines tested. These findings suggest that FASN gene copy number increases may be involved in the resultant increase in FASN protein expression observed in prostatic disease. D 2006 Elsevier Inc. All rights reserved.

1. Introduction B This work was supported by the National Cancer Institute RO1CA095239 (BRP). * Corresponding author. E-mail addresses: [email protected] (U.S. Shah)8 [email protected] (R. Dhir)8 [email protected] (S.M. Gollin)8 [email protected] (U.R. Chandran)8 [email protected] (D. Lewis)8 [email protected] (M. Acquafondata)8 [email protected] (B.R. Pflug).

0046-8177/$ – see front matter D 2006 Elsevier Inc. All rights reserved. doi:10.1016/j.humpath.2005.11.022

Fatty acid synthase (FASN), a 250- to 270-kD cytosolic multifunctional polypeptide chain, contributes to the de novo synthesis of fatty acids. Under normal conditions, FASN activity remains minimal due to the supply of endogenous fatty acids by dietary fat [1]. However, in cancer cells, significant overexpression and concurrent increased activity of FASN represents one of

402 the most frequently observed phenotypic alterations [2]. Indeed, in prostate cancer (PCa), up-regulation of FASN correlates with progression as well as aggressiveness of the disease [3,4]. Fatty acid biosynthesis occurs in all living organisms and provides essential components for biologic membranes as well as a form of energy storage. Animal FASN is an androgen-regulated multifunctional enzyme that catalyzes the synthesis of long-chain fatty acids via sequential condensation of two-carbon units from malonyl-CoA, an intermediate derived from the carboxylation of acetyl-CoA. Fatty acid synthase is a homodimer of a multifunctional subunit protein that contains 7 distinct activities and a site for the prosthetic group 4V-phosphopantetheine (acyl carrier protein). Activity and expression of FASN significantly increase in tumorigenic tissue [3,5]. Studies have shown that the highest FASN expression occurs in poorly differentiated and metastatic androgen-independent cancers [6]. These findings suggest that FASN represents a potentially important target in aggressive prostate cancer. Furthermore, FASN may act as a metabolic oncogene for prostate cancer [2,5]. However, the exact role of FASN in prostate cancer progression or onset remains unclear. Many of the genes coding for the enzymes of the fatty acid biosynthetic pathway reside on human chromosome 17q: FASN (17q25), acetyl-CoA carboxylase (17q21) (ACACA), and ATP citrate lyase (17q12-21) (ACLY). The synthesis of malonyl-CoA is the first committed step of fatty acid synthesis and the enzyme that catalyzes this reaction, acetyl-CoA carboxylase, is the major site of regulation of fatty acid synthesis. Cytosolic ATP-citrate lyase is involved in converting citrate to acetyl-CoA, the precursor for the FASN reactions. This region of chromosome 17 is a common site for amplification of genes, including the ERBB2 oncogene (HER2/neu), as well as for mutations and/or deletions of genes, such as BRCA1, metastasis suppressor gene NME1 (NM23), NGFR (p75 NTR), and HPC2 [7-11]. Studies have shown that an increase in gene copy number correlates with protein expression in some genes. One such example is the ERBB2 (HER2/neu) oncogene, which is amplified and overexpressed in about 25% to 30% of primary breast cancer cases and is associated with a poor prognosis [12-14]. A strong correlation between HER2/neu protein expression and gene copy number is present. HER2 status is important as a predictive factor to identify breast cancer patients most likely to respond to therapeutic intervention using the HER2/neu antagonist, herceptin. To date, FASN protein expression and gene copy number have not been correlated in prostate cancer. Thus, to investigate if a correlation is present between FASN expression and gene copy number, we measured FASN protein levels by immunohistochemistry (IHC) and gene copy number by fluorescence in situ hybridization (FISH) assays in prostate cancer tissue and cell lines.

U.S. Shah et al.

2. Materials and methods 2.1. Cell lines Androgen-sensitive LNCaP, androgen-independent PC-3, and DU145 human prostate carcinoma cell lines were purchased from American Type Culture Collection (Rockville, Md) and cultured in RPMI 1640 (Gibco BRL, Grand Island, NY) with 10% fetal bovine serum (Hyclone, Logan, Utah) under standard culture conditions. The androgenindependent PPC-1 cell line was obtained from Dr Joel B. Nelson (University of Pittsburgh). A well-characterized human breast cancer cell line, SKBR3 established from the pleural effusion of a hormone-independent human breast cancer (estrogen receptor–negative and progesterone receptor–negative), was obtained from the ATCC. SKBR3 cells were maintained in 75-cm3 flasks in DMEM (Gibco BRL) supplemented with 10% FBS (Hyclone) and 2 mmol/L glutamine (Sigma Chemical Co, St Louis, Mo).

2.2. Patient material and tissue microarray construction Prostate cancer specimens from 166 patients were used for this study. The cases were obtained from urologic surgery files of the Urology Department of the University of Pittsburgh, all obtained from 1996 to 1998, and were selected after pathologist review (by Dr Rajiv Dhir) of the hematoxylin and eosin staining to determine Gleason score and grade of cancer. The patients were selected to provide adequate representation of different T stages, and thus, the study included consecutive T2/T3/T4 patients. The PSA failure array was constructed with the 28 PSA failure patients in our repository without positive margins. The array constructed was a bmultitumorQ array due to the various numbers of cases present on the array. High-density tissue microarrays (TMAs) were assembled using the manual tissue puncher/array (Beecher Instruments, Silver Springs, Md). Tissue cores were 0.6 mm in diameter and ranged in length from 1.0 to 3.0 mm depending on the depth of tissue in the donor block. Multiple replicate core samples of normal, high-grade PIN, and prostate cancer tissue were acquired from each case. Cores were inserted into a recipient block measuring 45  20  12 mm and spaced 0.8 mm apart. The TMA set included progression TMAs (88 cases of prostate cancer with different Gleason grades and prostate cancer T stage). The TMAs also contained foci of metastatic prostate cancer, PIN, normal adjacent to tumor (NAT), benign prostatic hyperplasia (BPH), and btrue normalQ prostatic tissue from organ donors (Table 1).

2.3. FISH assays The cultured cells were treated with Colcemid (0.1 mg/mL) for 5 hours before harvesting. After mitotic arrest, the cells were processed in accordance with standard cytogenetic

FASN overexpression in PCa

403

Table 1 Demographics of human prostate tissue samples for the TMA study Tissue type

No. of cases

Prostate carcinoma Stage II III IV Gleason scores 5-6 7 8-9 Metastases High-grade PINa BPH NAT Donor

88

a

37 25 26 11 43 34 16 21 23 24 16

Cases other than CA represented on this TMA.

laboratory procedures and prepared for cytogenetic analysis. Individual FISH assays were performed to investigate the amplification status of three genes: FASN (BAC clone RP13650J16), ACACA (BAC clone RP11-378E13), and ACLY (BAC clone RP11-156E6) purchased from Children’s Hospital Oakland Research Institute (Oakland, Calif). Clones were grown, extracted, and directly labeled with Spectrum Orange using Vysis Nick Translation kit (Vysis, Inc, Downers Grove, Ill). Each of these probes was cohybridized with a probe to the centromere of chromosome 17 (plasmid pZ1714) obtained from Dr Mariano Rocchi (Resources for Molecular Cytogenetics, University of Bari, Italy). Plasmid pZ17-14 was grown, extracted, and directly labeled with Spectrum Green using the Vysis Nick Translation kit (Vysis, Inc). Interphase nuclei (metaphase cells were also analyzed when available) were scored for copy number of each gene. One hundred cells for each sample were analyzed and the number of green signals (chromosome 17 centromere) and the number of orange signals (gene) were recorded for each cell. The ratio of orange signals to green signals was calculated. A ratio of 0.8-1.2 was considered to be within normal limits. A ratio of less than 0.8 was attributed to gene loss, and ratios greater than 1.2 were attributed to an increase in gene copy number in cell lines. BACs corresponding to the three genes were purchased from Children’s Hospital Oakland Research Institute, cultured, DNA isolated, and labeled with Spectrum Orange for FISH by Nick Translation using the kit from Vysis. For assessment of gene copy number, FISH analysis of the TMA slides was carried out using methods in place in the UPCI Cytogenetics Facility. Briefly, sections of the TMA were placed on slides by the tape transfer method. The slides were then deparaffinized, pretreated, protease treated, and fixed, and the DNA on the slides was denatured, chilled in ethanol, dehydrated, and dried in preparation for hybridization. The slides were then warmed, the probe applied, cover-slipped, sealed, and placed in a humid chamber at 378C to hybridize

for approximately 40 hours. The slides were then washed to remove excess unhybridized probe, counterstained with DAPI, dried, cover-slipped with antifade, and sealed. All FISH analyses were carried out using an Olympus BX51 or BX61 epifluorescence microscope (Olympus microscopes, Melville, Ky). The Genus software platform on the Cytovision System was used for image capture and analysis (Applied Imaging, San Jose, Calif). Orange spots were counted for at least 10 and as many as 100 cells per tissue core and the raw data were recorded. Copy number between 2 and 4 was considered copy number gain and copy number of at least 5 was considered amplification based on a neardiploid tumor karyotype.

2.4. Immunohistochemistry Tissue microarray slides were stained for FASN expression using a monoclonal antibody directed against FASN (Anti-FASN [M] antibody 34-6E7, FAS Gen, Inc, Baltimore, Md). Tissue microarray slides were soaked overnight in xylene. The following day, the slides were cleared in two fresh changes of xylene for 20 minutes each, hydrated in two changes of 100% ethanol for 20 minutes each, one change of 95% ethanol for 20 minutes, and one change of 70% ethanol for 20 minutes. Slides were then rinsed in several changes of distilled water. Slides were placed in a working solution of TBS buffer for 5 minutes before loading onto an Autostainer. Once on the Autostainer, slides were blocked with peroxidase for 5 minutes and then rinsed with TBS buffer. After proteinase K treatment for 8 minutes, the slides were rinsed with TBS buffer, incubated with FASN primary antibody (1:1000) for 30 minutes, and then rinsed again with TBS buffer. The slides were incubated with mouse labeled polymer for 30 minutes and then rinsed with TBS buffer followed by incubation with DAB and chromogen for 10 minutes and then rinsing with distilled water. The slides were then counterstained with DakoHematoxylin for 1.5 minutes and washed with warm tap water for 2 minutes before being dehydrated, cleared, and cover-slipped.

2.5. Data analysis 2.5.1. Immunohistochemical analysis Tissue microarray data sets were analyzed using a combination of three Prostate Cancer Tissue arrays: Progression Array one, Progression Array two, and PSA Failure Array. The TMA data set was a combination of PSA FAILURE, TMA1, and TMA2. It consisted of 647 tissue samples from 166 different patients with ages ranging from 15 to 85 years (Table 1). Immunohistochemical staining was analyzed by a pathologist (Dr Rajiv Dhir) and samples were scored from 0 to 4 depending upon the amount of staining present. The analysis scale was based on percentage of cells with cytoplasmic positivity. The scale used was 0 = no expression; 1 = cytoplasmic positivity seen in up to 10% of cells; 2 = cytoplasmic positivity seen in 10% to 25% of cells;

404

U.S. Shah et al.

Fig. 1 FASN protein expression in prostate tissue. Prostate TMA slides were examined for FASN protein expression by IHC. Donor (A), normal adjacent (B), high-grade PIN (C), adenocarcinoma (D), BPH (E), and metastases (F). Brown staining is indicative of FASN protein expression (magnification 60).

3 = cytoplasmic positivity seen in 25% to 50% of cells; 4 = cytoplasmic positivity seen in more than 50% of cells. For all high IHC scores, a cutoff of 3 or more was used. This number is more stringent than other studies, which used 2 or higher as a high IHC score. Because one patient was used for more than one tissue sample and these tissue samples were correlated, the values of these tissue samples on each slide were averaged. There were 11 donors, 11 non-prostate benign tissues, 14 benign tissues (BPH), 32 high-grade PIN, 42 NAT, 106 prostate adenocarcinoma (PRCA), 7 metastatic lymph node tissues, and 6 metastatic tissues. Mean values from each group were compared between prostate tissue groups using the Mann-Whitney rank sum test and P values of less than .001 were considered significantly different. Kruskal-Wallis 1-way analysis of variance on ranks compared mean values within Gleason scores and stages of prostate cancer. P values of less than .05 were considered Table 2

significant. All pairwise multiple comparison procedures were conducted using Dunn method. 2.5.2. FISH data analysis Tissue microarray data sets were analyzed using a Prostate Cancer Tissue array, Progression Array two. It consisted of 188 tissue samples from 44 different patients. Because one patient was used for more than one tissue sample and these tissue samples were correlated, the highest FISH score was used from each sample set. When high FISH scores were needed to correlate with IHC, a score of 3 or more was considered as high gene copy number. There were 16 donors, 10 non-prostate benign, 24 BPH, 24 high-grade PIN, 24 NAT, 88 PCa, 7 metastatic lymph node tissues, and 9 metastatic tissues. Three or more signals per cell in any sample were considered as a positive score signifying a gene copy gain. Each group was compared between prostate tissue

FASN protein expression in prostate tissue

Histology

n

Mean

SD

Adenocarcinoma NAT BPH High-grade PIN Metastatic Met LN Donor

113 45 22 33 7 7 16

1.988 0.370T 0.284T 1.679 2.061 1.844 0.781T

1.286 0.570 0.413 1.082 1.617 1.674 0.883

Median

25%

75%

2.0 0 0 2.0 1.67 2.25 0.625

1.0 0 0 0.75 0.745 0.165 0

3.0 0.50 0.67 2.35 3.575 3.375 1.125

NOTE. TMA slides were stained for FASN expression using a monoclonal antibody for FASN. Each core was then examined for expression and assigned a number from 0 to 4 dependent upon the intensity of FASN immunostaining as determined by a pathologist. T Significant difference from adenocarcinoma (P b .001).

FASN overexpression in PCa

405

groups using the Fisher exact test for positive or negative values and the Mann-Whitney rank sum test for highest FISH scores and P values of less than .05 were significantly different. Kruskal-Wallis 1-way analysis of variance on ranks compared mean values within Gleason scores and stages of prostate cancer. P values of less than .05 were considered significant. All pairwise multiple comparison procedures were conducted using Dunn method.

3. Results 3.1. FASN expression in prostate cancer To evaluate FASN protein expression in prostate cancer, prostate cancer cell lines as well as prostate tumor tissue were examined. Immunohistochemical analysis revealed that adenocarcinoma, metastases, and high-grade PIN lesions exhibit significantly higher levels of FASN protein than donor, BPH, and NAT prostate tissue (Fig. 1, Table 2). When FASN expression was analyzed according to stage of the cancer, the data demonstrated a significant up-regulation of FASN in stage II and IV tumors as compared with stage III tumors (Table 3). This observation indicates that cancer progression may not be directly related to FASN expression levels. Evaluation of FASN levels according to Gleason scores revealed that the differences in the median values among the various scoring groups lacked statistical significance and thus differences may have resulted from random sampling variability. Therefore, FASN expression is not associated with Gleason score (data not shown). Of note, FASN expression is elevated in high-grade PIN, suggesting that up-regulation of this pathway represents an early event in disease progression (Table 2).

3.2. FASN gene copy in prostate cancer To achieve high fatty acid synthesis pathway activity in cancer, FASN as well as precursor enzymes to the pathway must undergo coordinated up-regulation. Because FASN, ACACA, and ACLY are all located on 17q, we examined the possibility that these genes exhibit gene copy amplification.

Table 3 Mean FASN protein expression in prostate cancer according to stage Group Stage II Stage III Stage IV

n 40 40 32

Median a

2.150 1.250a,b 2.550b

25%

75%

1.300 0.500 1.835

3.275 2.000 3.500

NOTE. TMA slides were stained for FASN expression using a monoclonal antibody for FASN. Each core was then examined for expression and assigned a number from 0 to 4 dependent upon the intensity of FASN immunostaining as determined by a pathologist. Core identifications determining the stage of the adenocarcinoma were identified and samples compared and analyzed. a, b represent significant differences between groups ( P b .05).

Fig. 2 PC-3 and LNCaP prostate cancer cell lines demonstrated increases in ratio of the number of copies of FASN to the centromeric probe. ACACA showed amplification in DU145 and LNCaP cell lines while ACLY showed increases in PPC-1, DU145, LNCaP, and LAPC4. Coordinate increases in copy number of all three genes do not appear to be common to all prostate cancer cell lines tested as a means of up-regulating FAS pathway activity.

Gene copy number of FASN, ACACA, and ACLY in human prostate carcinoma cell lines, LNCaP, DU145, PC-3, PPC-1, and LAPC4 as well as in the human breast carcinoma cell line SKBr3 was assessed by FISH analysis. Previous studies demonstrated that the SKBr3 cell line has high HER2/neu expression through gene amplification and high FASN expression via transcriptional activation although FASN pathway activity remains low due to limiting ACACA [15,16]. The FISH analysis revealed that the FASN gene was present in all cell lines. In PC-3 cells, 95% of counted cells had increases in gene copy number per chromosome with 17% exhibiting 4 or more copies of the gene. In LNCaP cells, 33% demonstrated an increase in the

406 Table 4

U.S. Shah et al. Gene copy numbers of lipid metabolism enzymes located on chromosome 17q

FASN Acetyl-CoA carboxylase ATP citrate lyase

PC-3

PPC-1

DU145

LNCaP

LAPC4

SKBr3

95% (+) 17.5% N4 0 0

32% ( ) 0 25% (+)

25% ( ) 18% (+) 20% (+)

33% (+) 33% ( ) 30% (+) 54% (+)

39.5% ( ) 16% (+) 15% ( ) 30% (+)

95% ( ) 99% ( ) 66% (+)

NOTE. Prostate cancer cell lines and a control breast carcinoma cell line (SkBr3) were analyzed by FISH using probes for FASN, ACACA, and ACLY and a centromere for chromosome 17. Number of cells expressing high gene copy number per Chr17 (3 or more signals) was considered positive (+) and low or no scores were considered negative ( ). PC-3 and LNCaP cells are the only significantly increased FASN gene copy number prostate cancer cell lines.

FASN gene. In contrast, analysis of the other prostate carcinoma cell lines did not indicate an increase in FASN gene copy number (Fig. 2, Table 4). Fluorescence in situ hybridization analysis of ACACA and ACLY revealed that amplification of these genes do not occur in PC-3 cells. In the PPC-1 cell line, only amplification of ACLY (25%) was apparent whereas DU145 exhibited an amplification of both ACACA and ACLY by 18% and 20%, respectively. Thirty percent of the counted LNCaP cells exhibited amplification of ACACA and 54% of the cells showed amplification of ACLY. The LAPC4 cell line showed 30% of the cells with amplified ACLY and no change in ACACA. The human breast cell line SKBr3 has ACLY amplified in 66% of the counted cells but no increase in ACACA (Fig. 2, Table 4). These results indicate that increased copy number of all three genes in the fatty acid synthesis pathway is not coordinately amplified to up-regulate the FASN pathway activity in prostate cancer cells.

To further investigate FASN gene copy gain in prostate cancer, prostate cancer progression TMA slides were probed using the FASN gene probes by FISH analysis. In Fig. 3, photographs of the hybridized cells from donor, NAT, highgrade PIN lesions, adenocarcinoma, BPH, and metastatic tissue revealed a significant increase in FASN gene copy number in adenocarcinoma compared with BPH and a significant increase in metastases as compared with highgrade PIN prostate tissues. Although we do not see a significant increase in FASN gene copy number in adenocarcinoma in comparison with PIN, donor, or NAT tissues, a strong correlation exists between high protein expression as visualized by IHC and FASN gene copy number as revealed by FISH analysis (Fig. 4A and B). Indeed, 53% of all the prostate tissues examined with increased FASN protein expression (IHC+) also exhibit increased FASN gene copy number (FISH+) (Fig. 4A). Examination of IHC score versus FISH scores in the individual groups revealed that

Fig. 3 FASN gene copy number in prostate tissue. Prostate TMA slides were examined for FASN gene copy number by FISH analysis. Donor (A), normal adjacent (B), PIN (C), adenocarcinoma (D), BPH (E), and metastases (F). Inserts are magnified individual cells. Red punctuate staining is indicative of FASN gene copy and blue indicates nuclei (DAPI) (magnification 60).

FASN overexpression in PCa

407 Table 6 Percentage of FASN gene copy signals in benign versus tumor prostate tissue exhibiting high FASN protein expression Group

1 Signal

2 Signals

3 Signals

4 Signals

5 Signals

Tumor Benign

5 5

70 94

15 1

10 0

0 0

NOTE. All tissues found to have a high IHC score for FASN expression were evaluated for FASN gene copy signals. Note the increase in 3 or more signals in tumor as opposed to benign prostate tissue.

showed that 25% of high-scoring FASN prostate tumor tissues by IHC exhibit 3 or more FASN gene signals whereas only 1% of benign prostate tissue exhibited gene copy gain (Tables 5 and 6). Although significantly elevated FASN protein levels are demonstrated by IHC in 36% (8/22) of high-grade PIN lesions, no high-grade PIN specimens demonstrated increased FASN copy number, similar to the incidence of increased copies (1/20) in normal adjacent tissue whereas high FASN expression is infrequent by IHC in NAT as well (3/22). Fig. 4 A, Correlation between FASN gene copy number and protein expression in prostate cancer. Prostate TMA slides were examined for FASN protein expression via IHC and FASN gene copy number via FISH analyses. A correlation between FASN protein expression and gene copy number was examined. Each bar represents the percentage correlation for the complete set of prostate tissue core samples with high or low FASN protein (IHC+/ ) and high or low gene copy number (FISH+/ ), respectively. B, FASN gene copy number and protein expression in the diseased prostate. Prostate TMA slides were examined for FASN protein expression via IHC and FASN gene copy number via FISH analyses. A correlation between FASN protein expression and gene copy number was examined in patients with adenocarcinoma, NAT, donor normal, BPH, metastases, and high-grade PIN. Each scatter plot represents individual patients for the respective prostate tissue core samples with FASN protein as an IHC score and gene copy number by FISH score, respectively. The number of samples at each score is indicated by the size of the symbol and the degree of filling such that the larger or the darker the circle indicates an increased sample size for that particular coordinate. The data were graphed in SPLUS 6.1 (Insightful Corporation, Seattle, Wash).

adenocarcinoma and metastases do exhibit an increased incidence of high IHC and high FISH correlation (Fig. 4B). Analysis of the distribution of high FASN staining samples Table 5

4. Discussion Gross chromosomal rearrangements such as gene amplifications, deletions, and translocations are common phenomena in prostate cancer. Studies have shown that gene copy number increase or gene amplification of large genomic regions containing oncogenes plays an important role in tumor progression [17]. One possible target of gene amplification in prostate cancer is the FASN gene located on chromosome 17q25. We and others have shown that FASN protein as well as activity are up-regulated in prostate cancer [18]. Recently, DNA replication stress or telomere dysfunction was reported to provoke nonrandom regional amplification or deletion in cancer genomes [19-21]. At the same time, chromosome 17 is a common site for rearrangement as well as the location of many oncogenes, such as HER2/neu (ERBB2) [22]. In addition, chromosome 17 is also the location of other fatty acid metabolism genes such as ACACA and ACLY. To understand the mechanism by which this upregulation occurs in adenocarcinoma of the prostate, we have investigated gene copy alterations of the genes associated with the fatty acid synthesis pathway. Using

FASN gene copy signals in prostate tissue exhibiting high FASN protein expression

Group

1 Signal

2 Signals

3 Signals

4 Signals

5 Signals

No. of cells

Adenocarcinoma Metastases High-grade PIN NAT

99 24 13 13

1333 334 194 317

275 (15) 82 (16) 0 4 (1)

182 (10) 59 (12) 0 0

1 (0) 0 0 0

1890 499 207 334

(5) (5) (6) (4)

(71) (67) (94) (95)

NOTE. All tissues found to have a high IHC score for FASN expression were evaluated for FASN gene copy signals. The percentage of gene copies out of total cells evaluated is stated in parentheses. Note that PIN and NAT tissues exhibit little or no cells with 3 or more FASN gene copies.

408 prostate TMA slides, we show a significant up-regulation of FASN in PIN and PCa that correlates with tumor grade but not with progression in our prostate TMAs as others have previously described. We have also shown a significant increase in FASN gene copy numbers in prostate adenocarcinoma and metastases, but not in PIN lesions. When we correlate the amount of protein expression to the gene copy number, we find an association between high FASN protein expression and increased FASN gene copy number in 53% of all observed prostate tissues. The increase in immunohistochemical staining of FASN in tumor tissue correlated with a 25% increase in gene copy number (3 or more signals). In contrast, high IHC staining in benign tissue correlated with an increase in FASN gene copy number in only 1% of the cells. This novel finding suggests that like other genes on chromosome 17q such as HER2/neu in breast cancer, FASN genetic alterations may be contributing to FASN protein expression level increases in the diseased prostate. Despite an increase in FASN gene copy number in prostate cancer, it does not account for the total protein increase apparent in the diseased prostate. Transcriptional regulation plays an important role in protein expression of many gene products in cancer including FASN. Recent findings suggest that high-level expression of FASN in prostate cancer tissues is linked to phosphorylation and nuclear accumulation of Akt [23]. In addition, androgen regulates the FASN gene, and therefore, transcription factors that bind the androgen response element on the FASN gene may also increase transcription of FASN [24]. Heemers et al [24] demonstrated that coordinated stimulation of lipogenic gene expression by androgens is a common phenomenon in androgen-responsive prostate tumor lines and involves activation of the sterol regulatory element–binding protein (SREBP) pathway. Also related to the mechanism underlying androgen activation of the SREBP pathway, they showed that androgens induce a major increase in the expression of SREBP cleavage–activating protein, an escort protein that transports SREBPs from their site of synthesis in the endoplasmic reticulum to their site of proteolytical activation in the Golgi [24]. Similarly, growth hormone has been shown to increase transcription of FASN and contribute to its overexpression in prostate cancer also by way of increasing SREBPs [25]. Studies have also shown that transcriptional regulation of FASN is in part due to the PI3-kinase pathway. An inhibitor of the PI3k pathway caused a dramatic decrease in FASN protein expression and substantial effects were seen at the FASN messenger RNA level and at the level of transcriptional activity of FASN promoter-reporter constructs [26]. In breast cancer, the ERBB2 or Her2/neu receptor gene is regulated by both gene amplification and transcription factors. Interestingly, the Her2/neu receptor stimulates the MAPKinase pathways, which in turn increase expression of cofactors that serve as a positive feedback mechanism to further elevate protein expression of the receptor [27]. This

U.S. Shah et al. mechanism of transcriptional regulation suggests that gene amplification may be involved in the increased expression of the receptor initially, and owing to positive modulation by cofactors, Her2/neu expression is then further increased. The FASN enzyme mediates increased fatty acid synthesis, thereby producing numerous downstream effects within the cell, such as cell membrane production and activation of many signaling pathways. Thus similar to HER2/neu, a positive feedback loop for FASN transcription may exist, explaining why gene copy number gains correlate with only a fraction of the increase in gene product. Posttranslational FASN regulation as well as tissue fixation methods may also modulate FASN protein expression and thus contribute to a lack of correlation between gene amplification and protein expression in a portion of the cases. Therefore, it is necessary to further examine the regulatory mechanisms involved in FASN protein expression as FASN overexpression may serve as an important factor in prostate cancer etiology. In conclusion, the present study examines the relationship between FASN protein overexpression and FASN gene copy number in the diseased prostate. Our findings suggest that FASN protein overexpression may be attributed to a significant degree by an increase in gene copy number in prostate cancer cells and tissue. To date, no such relationship has been established and these novel findings suggest that transcriptional regulation of FASN in prostate cancer is not the only mechanism by which FASN protein overexpression may occur. This significant percentage of correlation suggests that genetic alterations of FASN may not result from a random event in prostate cancer. Findings in the literature suggest that DNA replication stress or telomere dysfunction may provoke nonrandom regional amplification or deletion in cancer genomes [19-21]. Further examination is necessary to determine how FASN gene alterations in the diseased prostate occur.

References [1] Diraison F, Yankah V, Letexier D, Dusserre E, Jones P, Beylot M. Differences in the regulation of adipose tissue and liver lipogenesis by carbohydrates in humans. J Lipid Res 2003;44:846 - 53. [2] Kuhajda F. Fatty-acid synthase and human cancer: new perspectives on its role in tumor biology. Nutrition 2000;16:202 - 8. [3] Rossi S, Graner E, Febbo P, et al. Fatty acid synthase expression defines distinct molecular signatures in prostate cancer. Mol Cancer Res 2003;1:707 - 15. [4] Swinnen JV, Roskams T, Joniau S, et al. Overexpression of fatty acid synthase is an early and common event in the development of prostate cancer. Int J Cancer 2002;98:19 - 22. [5] Baron A, Migita T, Tang D, Loda M. Fatty acid synthase: a metabolic oncogene in prostate cancer? J Cell Biochem 2004;91:47 - 53. [6] Swinnen JV, Vanderhoydonc F, Elgamal AA. Selective activation of the fatty acid synthesis pathway in human prostate cancer. Int J Cancer 2000:176. [7] Benowitz S. New gene finding may yield clues to prostate cancer complexity. J Natl Cancer Inst 2001;93:344 - 5. [8] Jarvinen T, Tanner M, Rantanen V, et al. Amplification and deletion of topoisomerase II alpha associate with ErbB-2 amplification and affect

FASN overexpression in PCa

[9]

[10]

[11]

[12]

[13]

[14]

[15] [16]

[17] [18]

sensitivity to topoisomerase II inhibitor doxorubicin in breast cancer. Am J Pathol 2000;156:839 - 47. Hendrickson B, Judkins T, Ward B, et al. Prevalence of five previously reported and recurrent BRCA1 genetic rearrangement mutations in 20,000 patients from hereditary breast/ovarian cancer families. Genes Chromosomes Cancer 2005;43:309 - 13. Jin W, Xu X, Yang T, Hua Z. p53 mutation, EGFR gene amplification and loss of heterozygosity on chromosome 10, 17 p in human gliomas. Chin Med J 2000;113:662 - 6. Terada R, Yasutake T, Nakamura S, et al. Clinical significance of nm23 expression and chromosome 17 numerical aberrations in primary gastric cancer. Med Oncol 2002;19:239 - 48. Borg A, Tandon A, Sigurdsson H, et al. HER2/neu amplification predicts poor survival in node-positive breast cancer. Cancer Res 1990;50:4332 - 7. Slamon D, Clark G, Wong S, Levin W, Ullrich A, McGuire W. Human breast cancer: correlation of relapse and survival with amplification of the HER-2/neu oncogene. Science 1987;253:177 - 82. Slamon D, Godolphin W, Jones L, et al. Studies of the HER-2/neu proto-oncogene in human breast and ovarian cancer. Science 1989; 244:707 - 12. Thompson B, Smith S. Lack of coordinated regulation of lipogenic enzymes in a human breast cell line, SKBr3. 1982;712:217-20. Oskuian B. Overexpression of fatty acid synthase in SKBr3 breast cancer cell line is mediated via a transcriptional mechanism. Cancer Lett 2000;149:43 - 51. Ozen M, Pathak S. Genetic alterations in human prostate cancer: a review of current literature. Anticancer Res 2000;20:1905 - 12. Pflug B, Pecher S, Brink A, Nelson J, Foster B. Increased fatty acid synthase expression and activity during progression of prostate cancer in the TRAMP model. Prostate 2003;57:245 - 54.

409 [19] Hellman A, Zlotorynski E, Scherer S, et al. A role for common fragile site induction in amplification of human oncogenes. Cancer Cell 2002:89 - 97. [20] O’Hagan R, Chang S, Maser R, et al. Telomere dysfunction provokes regional amplification and deletion in cancer genomes. Cancer Cell 2002:149 - 55. [21] Gollin S. Mechanisms leading to chromosomal instability. Semin Cancer Biol 2005;15:33 - 42. [22] Ittmann M. Loss of heterozygosity on chromosomes 10 and 17 in clinically localized prostate carcinoma. Prostate 1996;28:275 - 81. [23] Van-de-Sande T, Roskams T, Lerut E, et al. High-level expression of fatty acid synthase in human prostate cancer tissues is linked to activation and nuclear localization of Akt/PKB. J Pathol 2005;206: 214 - 9. [24] Heemers H, Maes B, Foufelle F, Heyns W, Verhoeven G, Swinnen J. Androgens stimulate lipogenic gene expression in prostate cancer cells by activation of the sterol regulatory element–binding protein cleavage activating protein/sterol regulatory element–binding protein pathway. Mol Endocrinol 2001;15:1817 - 28. [25] Swinnen J, Heemers H, Deboel L, Foufelle F, Heyns W, Verhoeven G. Stimulation of tumor-associated fatty acid synthase expression by growth factor activation of the sterol regulatory element–binding protein pathway. Oncogene 2000;19:5173 - 81. [26] Van-de-Sande T, De-Schrijver E, Heyns W, Verhoeven G, Swinnen J. Role of the phosphatidylinositol 3V-kinase/PTEN/Akt kinase pathway in the overexpression of fatty acid synthase in LNCaP prostate cancer cells. Cancer Res 2002;62:642 - 6. [27] Bosc D, Janknecht R. Regulation of Her2/neu promoter activity by the ETS transcription factor, ER81. J Cell Biochem 2002;86:174 - 83.