|Home | About | Journals | Submit | Contact Us | Français|
A clearer definition of the molecular determinants that drive the development and progression of prostate cancer (PCa) is urgently needed. Efforts to map recurrent somatic deletions in the tumor genome, especially homozygous deletions (HODs), have provided important positional information in the search for cancer-causing genes. Analyzing HODs in the tumors of 244 patients from two independent cohorts and 22 PCa xenografts using high-resolution single-nucleotide polymorphism arrays, herein we report the identification of CHD1, a chromatin remodeler, as one of the most frequently homozygously deleted genes in PCa, second only to PTEN in this regard. The HODs observed in CHD1, including deletions affecting only internal exons of CHD1, were found to completely extinguish the expression of mRNA of this gene in PCa xenografts. Loss of this chromatin remodeler in clinical specimens is significantly associated with an increased number of additional chromosomal deletions, both hemi- and homozygous, especially on 2q, 5q and 6q. Together with the deletions observed in HEK293 cells stably transfected with CHD1 small hairpin RNA, these data suggest a causal relationship. Downregulation of Chd1 in mouse prostate epithelial cells caused dramatic morphological changes indicative of increased invasiveness, but did not result in transformation. Indicating a new role of CHD1, these findings collectively suggest that distinct CHD1-associated alterations of genomic structure evolve during and are required for the development of PCa.
Although many prostate cancer (PCa) tumors are indolent and pose little health hazard, an important subset are aggressive and progress to disseminated disease, resulting in ~34 000 deaths in the United States every year (Siegel et al., 2011). Recent studies have highlighted the collaborative nature of multiple genomic alterations underlying the critical process of PCa progression (Carver et al., 2009; King et al., 2009; Ding et al., 2011).
Indeed, recent deep sequencing of the exome and the whole genome of tumor cells has revealed the extraordinary complexity of genomic alterations that characterize human cancers (Berger et al., 2011; Robbins et al., 2011). This complexity consists of various combinations of base substitution, translocations, gene fusion and copy-number alterations (CNAs). It is becoming increasingly clear that CNAs are a major component of the landscape of the PCa tumor genome (Taylor et al., 2010; Robbins et al., 2011; Kan et al., 2010). To assess the significance of these alterations in the development of PCa, it is necessary to distinguish alterations driving proliferation of cancer cells versus random changes or passengers. Beroukhim et al. (2007) have developed an important tool (Genomic Identification of Significant Targets in Cancer (GISTIC)) that has been used in the identification of a number of cancer genes (Beroukhim et al., 2010; Taylor et al., 2010).
After candidate regions of CNAs are identified, a major challenge is how to then further identify candidate cancer genes, as the size of deletions and amplifications are usually very large, covering many genes, especially in high-grade and metastatic PCa. Using GISTIC with stringent criteria can help to narrow the search to a target region with fewer genes. Even so, a majority of such regions in the tumor genome are hemizygous, with their in vivo biological effect on the growth advantage of cancer cells being difficult to infer because a second, unaltered allele is still present, notwithstanding established haploinsufficient or dominant cancer genes. The unequivocal loss-of-function associated with homozygous gene deletions (HODs) can simplify this process. As a result, identifying and characterizing HODs have led to the discovery of multiple recessive human cancer genes with PTEN being an important example in PCa.
Herein we report the assessment of HOD in the tumor genomes of 244 patients with primary PCa from two independent cohorts and 22 xenografts via genome-wide analysis of CNAs. While uncovering a number of new recurrent HODs using allele-specific analysis, we discover that CHD1 is the second, only to PTEN, most frequent homozygously deleted gene in PCa. We find that loss of both the alleles of CHD1 is significantly associated with additional, potentially targeted HOD in the tumor genome. Accordingly, chromosomal deletions associated with experimental knockdown of CHD1 provide evidence that CHD1 may have a causal role in prevention of deletion events. These data revealed a novel CHD1-associated coordinative network of alterations, and suggest a new role for this chromatin remodeler in the development of PCa.
Recurrent somatic deletions in the tumor genome, especially HODs, have been informative targets in the search for tumor-suppressor genes. To uncover the full spectrum of HODs in primary PCa tumors, we performed a comprehensive analysis of DNA CNAs in the tumor genomes of surgical specimens of cancer tissues from 244 PCa patients using Affymetrix high-resolution single-nucleotide polymorphism arrays (Affymetrix, Santa Clara, CA, USA). To identify CNAs that likely drive cancer growth, we first used GISTIC with a q = 0.01 (false discovery rate (FDR)) and a join-segment-size of 80 probes to identify the significant regions of deletions. The data revealed 20 significant CNA regions, including 15 and 13 deletions (Supplementary Figure 1) in the Johns Hopkins Hospital (JHH) and Swedish cohorts, respectively, as well as 5 and 7 amplifications. Although distinct CNAs unique to a specific cohort were observed, the overall patterns were remarkably similar. For example, all the 13 regions of deletion identified in the Swedish cohort overlapped with the regions of deletion identified in the JHH cohort with a positive agreement score of ~93% that is not statistically different from 100% agreement (P = 0.0781), although seven of the peak regions were different. Two distinct regions of deletion with peaks at 1q42.2 and 11q23.2 were observed in the JHH cohort but did not reach to the level of significance at q = 0.01 in the Swedish cohort. Among regions having the same peak between these two cohorts, four of them including 2q22.1, 3q13, 5q21.1 and 10q23.31 contain only one significant (q = 0.01) gene each: LRP1B, RYBP1, CHD1 and PTEN, respectively, as determined by GISTIC. We also carried out a minimum-overlap-region analysis (Supplementary Figures 2–5) with informative samples harboring deletions in these regions and confirmed our findings from the GISTIC analysis as described above.
Importantly, genome-wide allele-specific analysis revealed multiple occurrences of CHD1 HOD in both of the cohorts from JHH and Sweden, with frequencies of 7.1% and 10.7%, respectively (Table 1 and Supplementary Figures 6 and 7). In comparison, HOD frequency at PTEN was observed in ~13% and 16%, respectively, of the sample in these two cohorts. Thus, at this resolution, CHD1 is second only to PTEN as the most frequent homozygously deleted gene in the tumor genome of PCa.
In primary PCa tumors, the size of HODs affecting CHD1 ranged from ~138 kb to 2898 kb, in some instances covering RGMG, FAM174A and ST8SIA4, in addition to CHD1 (Supplementary Table 1). Three of the HODs eliminated the 5′ region of CHD1 (Figure 1a), while the majority removed the whole gene (Figure 1b). Three of the HODs removed the 3′-coding region of CHD1. We further analyzed CNAs among an additional 22 prostate tumor xenografts and identified four additional HODs at 5q21.1 that affected CHD1 (Supplementary Figure 7). One of the HODs removed at least 10 internal exons of CHD1 (Figure 1c). We confirmed the complete loss of CHD1 mRNA expression in all the four xenograft samples with CHD1 HOD (Figures 1d and e), but only a moderate loss of CHD1 mRNA in the samples with hemizygous deletion (Supplementary Figures 8A and B). Although one HOD (in Lu81) caused a contiguous loss of only an internal portion of CHD1 (Figure 1c), loss of this partial gene segment resulted in complete loss of the expression of all exons including the intact exons, as revealed by exon expression analysis using the Affymetrix human Exon 1.0 ST array (Supplementary Figure 8C). These results suggest that HODs affecting CHD1 typically result in complete absence of CHD1 expression, as opposed to hemizygous deletions of CHD1, which resulted in decreased but not absent CHD1 mRNA expression (Supplementary Figure 8A).
To explore the possible genomic and functional effects of CHD1 HOD in PCa, we began by characterizing the genome-wide DNA CNA profiles in tumor genomes with CHD1 HODs, initially comparing such cases with those harboring PTEN HODs and then comparing the cases harboring CHD1 or PTEN HODs with those harboring no CNAs at these two genes. As the frequency of HODs may be associated with general genomic instability as tumor cells evolve during cancer development and PTEN is the most frequent homozygously deleted gene in PCa, we therefore used tumors with PTEN HODs as ‘positive controls’ and tumors without deletions at these two genes as ‘negative controls’. In these comparisons, we observed a substantially higher frequency of HODs in other genomic regions among samples with CHD1 HOD. Cases harboring CHD1 HODs contained an average of 4.5 additional HODs per genome compared with an average of 0.89 per genome in cases harboring PTEN HODs (P = 0.0003, Wilcoxon two-sample test; Figure 2a) in the JHH cohort. We validated our findings using the Swedish cohort, where we observed ~4.64 additional HODs per genome among tumors harboring the CHD1 HOD, in contrast to 0.53 additional HODs per genome among tumors harboring the PTEN HOD (P = 0.0001). Combining these two cohorts we identified a total of 21 patients harboring CHD1 HODs with an average of 4.57 additional HODs per genome in the patients harboring CHD1 HOD, which is significantly higher than 0.71 additional HODs per genome in the patients harboring PTEN HODs (P = 1.15 × 10−7). In contrast to tumors harboring no CNAs at CHD1, based on allele-specific analysis, tumors harboring CHD1 HOD contained significantly more additional HODs in both the JHH and the Swedish cohorts, with P = 4.67 × 10−7 and P = 9.95 × 10−8, respectively (Figure 2a).
Analyzing these HODs, we noticed multiple small-sized HODs that clustered in the vicinity of particular regions on the same chromosome (Figures 1b and c). These clustered HODs may be derived from a single event that led to the initial genomic deletion/alteration (Stephens et al., 2011). To minimize any possible effects of these multiple, clustered HODs on calculations of the number of HOD events, we compared the number of cytological chromosomal bands (cytobands) affected by additional HODs in the tumors harboring CHD1 HODs versus tumors with complete loss of PTEN and with no CNAs at CHD1 and/or PTEN. The data revealed that a significantly larger number of cytobands were affected by these additional HODs in the genomes with complete loss of CHD1 than by those in the genomes containing either complete loss of PTEN or no CNAs at CHD1 and/or PTEN (Figure 2b).
We next analyzed the distribution of these additional HODs in tumors with complete loss of CHD1. As shown in Figure 3, in a comparison among these tumors from both the JHH and the Swedish cohorts, most of these additional HODs were not randomly distributed, but rather appeared preferentially located on chromosomes 2, 5 and 6, being absent from chromosomes 9, 11 and 14 through 22 among the tumor genomes with complete loss of CHD1. One exception was chromosome 4 that harbored HODs in the Swedish cohort (P = 9.91 10−10, Figure 3b), but not in the JHH cohort (Figure 3a). Some recurrent additional HODs targeted the same region or affected the same genes (Supplementary Figures 2, 6, 9, and Supplementary Table 2, respectively), while other additional HODs occurred only once in tumors harboring CHD1 HOD (Supplementary Figures 2, 3, 6, 7, 9 and 10).
To further evaluate the effects of loss of CHD1 on the changes in CNA signature among the tumor genomes in the JHH and Swedish cohorts, we first tested the association between any deletion (hemizygous and homozygous) at CHD1 and the other CNAs with their representative genes identified by GISTIC (Table 1). As shown in Figure 4, deletions of CHD1 was positively associated with deletions of LRP1B at 2q22.1, PDE4D at 5q11.2, MAP3K7 at 6q15 and gain of COL1A2 at 7q21.3 in both of the cohorts (P = 0.0079, Supplementary Table 3). We next compared the GISTIC signatures with or without the tumors harboring CHD1 deletions. As shown in the Supplementary Figure 11, removing the tumors harboring CHD1 deletions (right panels in A, B, C and D) substantially reduced the significance levels of signature peaks associated with deletions of LRP1B at 2q22.1, PDE4D at 5q11.2, MAP3K7 at 6q15 and gain of COL1A2 at 7q21.3 (marked by light blue ovals), in comparison to the significance levels of these signature peaks derived from all tumors including the ones harbored CHD1 deletions (left panels in A, B, C and D). These significant concurrences of CNAs suggest a novel collaborative CNA-network among these loci in the evolution of the PCa tumor genome.
In addition, we observed a significantly negative correlation between CNAs at CHD1 and TMPRSS2–ERG on 21q22 (Figure 4) in both the JHH and the Swedish cohorts (P = 0.0057 for the direction of association). Indeed, none of the 21 patients with CHD1 HOD harbored the deletion from 3′ of TMPRSS2 to 5′ of ERG, which is statistically significant (P = 7.34 × 10−4, Fisher’s exact test). These findings indicate that mutually exclusive selection of these two CNAs might occur during proliferation of PCa cells, resulting in two distinct subgroups of PCa.
Furthermore, we explored the association between loss of CHD1 and PCa aggressiveness. As the total number of CHD1 HOD is too small to perform a meaningful statistical analysis, we combined tumors with either hemizygous or HODs to explore the relationship between the loss of CHD1 and the Gleason score. We found that loss of CHD1 was significantly associated with tumors having a higher Gleason score <7 in the JHH cohort (P = 0.032) but not in the Swedish cohort. These findings warrant further investigation using additional larger cohorts.
The data presented above suggest that loss of CHD1 may predispose cells to genomic deletions, potentially at specific loci. To begin to address this question, we chose a well-established human cell transfection model, HEK293, for manipulating CHD1 expression and analyzed CNAs in cells with stable knockdown of Chd1 protein expression. We identified a deletion that affected LRP1B in a cell line stably transfected with CHD1 small hairpin RNA (shRNA; Supplementary Figure 12), which is consistent with our findings in the primary tumors, and suggests a causal relationship. The deletion at LRP1B was not observed in the cells either without transfection or stably transfected with control shRNA. We also observed one HOD each on chromosomes 3 and X, as well as hemizygous deletions on chromosomes 16 and 18 (Supplementary Figure 12) in the cell lines stably transfected with CHD1 shRNA but not in these two types of control cells.
To assess possible phenotypic alterations associated with CHD1 downregulation, we chose to use our mouse prostate epithelial cells (MPECs) model because the cells maintain progenitor cell characteristics over long-term culture and reflect the true nature of prostate epithelial stem cells in vivo (Barclay et al., 2008). We infected the MPECs using lentiviruses expressing CHD1 and control shRNAs. As shown in Figure 5b, CHD1 shRNA effectively reduced expression of Chd1 protein. In clonogenic analysis, we observed that MPECs expressing CHD1 shRNA formed more colonies than the cells expressing control shRNA (Figure 5a). Quantitative analysis of the clonogenic data revealed that cells with knockdown expression of CHD1 displayed a consistently higher (P = 0.0489) survival and proliferation rate (Figure 5b). This result suggests that silencing endogenous CHD1 expression might enhance cell clonogenicity or survivability. We further analyzed the morphological characteristics of MPECs growth, using silenced expression of endogenous CHD1 in a three-dimensional (3D) culture system. As expected, cell spheroids were formed by MPECs transfected with control shRNA (Barclay and Cramer, 2005; Figure 5c). Surprisingly, the MPECs transfected with CHD1 shRNA formed branch structures growing into the collagen matrix, with significantly (P < 0.0001) lengthening of branches and side branches (Figures 5d and e) that were not observed in cells transfected with control shRNA. These characteristics suggest that silenced CHD1 expression could significantly enhance cell invasiveness and/or developmental changes. However, renal-grafting these cells in mice did not result in tumorigenesis in our in vivo study (data not shown). These findings indicate that alterations of other genes, such as those were associated with CHD1 deletion in the primary tumors described above, may be needed to initiate the development of PCa.
Genetic alterations observed in a specific tumor genome at a particular stage are cumulative products of successive clonal expansion and mutational evolution (Liu et al., 2009; Anderson et al., 2011). However, HODs that do not affect cell survival could be maintained as random passenger mutations. When a HOD affects genes whose absence provides growth advantages, this HOD would be selected during cell proliferation and could be observed in a significant larger number of the tumor genomes in a large population of PCa patients. Analysis of HODs may reveal not only recessive cancer genes but can also uncover regions of genetic fragility (Bignell et al., 2010) sites predisposed to undergo DNA rearrangements in the tumor genome. Using an unbiased genome-wide analysis of tumor DNA, we uncovered CHD1, the chromatin remodeler, as one of the two most frequent homozygously deleted genes in the tumor genome of PCa. Importantly, inactivating mutations in the chromatin remodelers, CHD1 and CHD5, have been also reported recently in PCa (Berger et al., 2011; Robbins et al., 2011).
Among the three genes concurrently deleted with CHD1, LRP1B and PDE4D are apparently located at fragile sites on chromosomes of 2 and 5, respectively (Bignell et al., 2010). However, MAP3K7 seems not to be associated with inherent fragility and rather may function as a tumor-suppressor gene. Encoding TGF-β activated kinase-1, MAP3K7, has been reported to be frequently deleted in PCa, the occurrence of which is highly associated with high-grade disease (Liu et al., 2007). Recent work has demonstrated a tumor-suppressive role of MAP3K7 in liver cancer (Bettermann et al., 2010). As LRP1B and PDE4D are located within fragile sites of the genome, their functions as tumor-suppressor genes in the development of PCa are subject to further investigation. However, significant correlation between the loss of CHD1 and deletions of LRP1B and PDE4D rather than genes located at other fragile sites, suggests their concurrent losses provide advantages for selection and proliferation of cancer cells. Inactivation of LRP1B by mutation, deletion and methylation has been previously reported in the lung cancer (Ding et al., 2008; Kohno et al., 2010), glioblastoma multiforme (Yin et al., 2009), oral cancer (Cengiz et al., 2007) and gastric cancer (Lu et al., 2010). PDE4D, encoding cyclic adenosine monophosphate-specific phosphodiesterase 4D, has been previously identified as a potential tumor-suppressor gene in esophageal adenocarcinoma (Nancarrow et al., 2008). On the other hand, Rahrmann et al. (2009) identified PDE4D as a proliferation-promoting factor in PCa.
CHD1 encodes a chromodomain helicase DNA-binding protein known to have important roles in regulating gene expression in mammalian cells (Sims et al., 2007), gametogenesis in Drosophila melanogaster (McDaniel et al., 2008), and pluripotency of stem cells in mice (Gaspar-Maia et al., 2009). Multiple chromatin-remodeling genes have recently been implicated through mutational analyses of various cancer types, such as renal cell (Dalgliesh et al., 2010), ovarian (Jones et al., 2010), lung (Medina et al., 2008) and others (Weissman and Knudsen, 2009), emphasizing the importance of this epigenetic function in human carcinogenesis.
To explore the function of CHD 1 in the tumor genome of PCa, we compared the CNAs in tumors harboring HODs of CHD1 to either those harboring HODs of PTEN or those with no CNAs at CHD1 and/or PTEN. To our surprise, the tumors harboring CHD1 HOD had a significantly larger number of additional HODs in other locations of the genome. Most of these additional HODs are significantly located on chromosomal regions 2q, 5q and 6q where hemizygous deletions occurred and were associated with the loss of CHD1.
The consistent associations of CNAs found in clinical samples from two independent cohorts might be also due to selection for growth and proliferation advantage of the tumor cells during cancer development in the prostate, in addition to a possibly causal relationship. The cells transfected with shRNA against CHD1 obviously did not evolve as the tumor cells did in the clinical specimens. Multiple deletions, including those observed in clinical specimens, in the cells with knockdown expression of CHD1 provide evidence supporting the protective role of CHD1 and provide a basis for CHD1-associated CNAs if these cells could evolve and proliferate in the same way as those in clinical samples. The fact that experimental knockdown expression of CHD1 itself did not result in transformation and deletions on chromosomes 5 and 6 suggests that deletions on chromosomes 5q and 6q might be caused by (1) further genomic evolution of the cells with loss of CHD1 induced by tumor microenvironment, (2) a different mechanism other than loss of CHD1. From this point of view, the data from experimental knockdown expression of CHD1 is consistent with the findings in clinical samples from the two cohorts. More suitable in vitro and in vivo models with human prostate epithelial cells and tissue recombination could shed more light on mechanism of CHD1-associated CNAs.
Together with the results from HEK293 cells stably transfected with CHD1 shRNA, these data indicate that CHD1 may either have a role in protecting the genome from loss of DNA, or it may, in conjunction with the other associated CNAs, provide selection and/or growth advantages for tumor cells during the development of PCa. It is plausible that CHD1, a chromatin remodeler, may indirectly protect DNA because it facilitates deposition of H3 histone into chromatin (Konev et al., 2007; Sims and Reinberg, 2009). In addition, direct binding of CHD1 to DNA (Stokes and Perry, 1995) may provide an alternative mechanism of protection, however, it is unknown whether CHD1 is directly involved in DNA repair. Post-translational modifications to histones have been reported to influence DNA repair (Avvakumov et al., 2011). On the other hand, deletion at LRP1B in cells with knockdown expression of CHD1 but not in the controls could be due to chance, and its association with loss of CHD1 in clinical cohorts might be caused by selective advantage for cell proliferation. Extensive studies using a more suitable model with human prostate epithelial cells are needed to fully evaluate the effects of Chd1 knockdown on genomic alterations and tumorigenesis.
To further evaluate the effect of CHD1 loss in the development of PCa, we knocked down Chd1 expression in MPECs. These mouse prostate progenitor/stem cells maintain progenitor characteristics in long-term culture in vitro and are capable of fully differentiating into prostatic structures in vivo (Barclay et al., 2008). CHD1 is known for its role in development and in maintaining the pluripotency of embryonic stem cells. We speculate that loss of Chd1 in these stem cells may disrupt normal development of prostatic structures and lead to prostatic abnormality or eventually tumorigenesis. It is therefore better to use a developmental model with prostate progenitor/stem cells as the basis upon which to test our hypothesis. As prostate and tumor stem cells possess similar capabilities such as self-renewal and androgen independence, it has been speculated that tumor stem cells are derived from aberrant stem cells. Therefore, we believe that using this developmental model with prostate progenitor/stem cells to evaluate the role of CHD1 can provide critical insight into the mechanisms that drive the development of PCa. Although the MPECs expressing CHD1 shRNA generated only slightly more colonies than the cells expressing the control shRNA in the clonogenic assay, they produced significantly longer branching structures that extended into the collagen matrix in a 3D culture system. These morphological characteristics suggest invasive growth, as well as developmental changes (Barclay and Cramer, 2005) of the cells with down-regulation of Chd1. The fact that renal-grafting these cells with downregulated Chd1 in mice did not result in tumorigenesis in our in vivo study indicates that CHD1 itself may not be sufficient for tumor suppression although CHD1 HODs were frequently observed in PCa. It has been reported that transgenic mice expressing the TMPRSS2–ERG fusion gene also failed to develop prostatic intraepithelial neoplasia or prostate tumor (King et al., 2009). However, when these ERG overexpressing transgenic mice were crossed with PTEN-deficient mice, PCa was observed in their offspring (Carver et al., 2009; King et al., 2009). These studies demonstrated that alterations of multiple genes are required for a particular phenotype in the development of PCa. The genes concurrently altered with deletion of CHD1 that were identified in this study may represent a collaborative network that is acquired in cancer cells and cumulatively drives the development of PCa. This may explain why knockdown CHD1 in isolation failed to result in tumorigenesis experimentally. Future studies should reveal more details of mechanisms by which these genes (loci) interact in promoting the development of PCa.
The observation that loss of CHD1 is negatively correlated with the deletion at 21q22 that creates the fusion of TMPRSS2–ERG indicates tumors harboring these alterations may represent different subtypes of PCa. Berger et al. (2011) reported that the locations of DNA rearrangement breakpoints in tumors harboring TMPRSS2–ERG fusion inversely correlated with break-points in tumor lacking the ETS fusion. Their findings suggest a link between chromatin remodeling and genesis of genomic alterations. Our data provide further support for the role of a particular chromatin remodeler, CHD1, in the genesis of CNAs.
In summary, we have identified CHD1, as second only to PTEN, the most frequently homozygously deleted gene in PCa. We demonstrate that (1) complete loss of CHD1 was associated with increased number of additional HODs on 2q, 5q and 6q, and (2) loss of CHD1 was positively correlated with CNAs at LRP1B, PDE4D, MAP3K7 and COL1A2, but negatively associated with deletion at 21q22 that creates TMPRSS2–ERG fusion. Assessing the function of CHD1, we found that downregulation of Chd1 protein resulted in morphological changes in the growth of MPECs, but did not produce in vivo tumorigenesis. Although these findings suggest that collaborative alterations were acquired in a cumulative manner in the tumor genome and their interaction may drive the development of PCa, further study is warranted to illustrate how CHD1 interacts with other genes, including LRP1B, PDE4D and MAP3K7, to affect tumorigenesis and the development of PCa. It will also be important to evaluate whether loss of this chromatin remodeler directly causes aberrations of genomic structure such as deletions and amplifications, or indirectly lead to concurrent accumulation of these aberrations, especially at/near the specific fragile sites at 2q22.1 and 5q11.2. Although further study is necessary, these findings strongly implicate CHD1 as a critical factor in the dramatic genomic reorganization that accompanies prostate carcinogenesis.
Somatic tumor DNA from a total of 22 xenografts and 244 PCa patients undergoing radical prostatectomy for the treatment of clinically localized disease at two centers, one in the United States, (JHH) and the other in Sweden (Karolinska Institute) from 1988 to 2006 was used in this study. They were selected based on the availability of genomic DNA of sufficient quantity (<5 μg) and purity (<70% cancer cells for cancer specimens, no detectable cancer cells for normal samples). Tissue samples were obtained by macro-dissection of matched non-malignant (normal) and cancer-containing areas of prostate tissue as determined by histological evaluation of hematoxylin and eosinstained frozen sections of snap-frozen radical prostatectomy specimens. Amongst these 244 patients, 193 had normal control DNA, whereas 51 of them had no matched normal DNA available at the time of DNA analysis. Most of the 141 patients in the JHH cohort had a more aggressive form of PCa; 31%, 30%, and 46% of patients had pathologic Gleason score ≥8, pathologic stage ≥T3b and pretreatment serum PSA ≥10 ng ml−1, respectively. Most of the 103 patients in the Swedish cohort had a less aggressive form of PCa; ~51 and ~11% of these patients had a preoperative Gleason score ≤6 and a pathological Gleason score ≥8, respectively. Experimental procedures for assay of single-nucleotide polymorphism array, DNA copy-number analysis and identification of target genes are presented in the Supplementary Methods.
We isolated the total RNA from fresh frozen tissues of xenograft tumors using a TRIzol Reagent kit (Invitrogen, Grand Island, NY, USA). The amount and quality of total RNA were assessed using the Agilent 2100 Bioanalyzer (Agilent Technologies, Wilmington, DE, USA) before expression analysis. Expression microarray analysis was carried out according to the instructions in the Agilent 4 × 44 Whole Genome Expression Microarray System (Agilent Technologies) using an input of 500 ng of total RNA, and a 2-color design involving cohybrization of Cye-5-labeled test samples with a Cye-3-labeled common reference sample from benign prostate cells. The expression profile for each sample was represented as a normalized ratio of sample/reference for all entities represented on the array. Statistical analyses of the differentially expressed genes were performed using GeneSpring software (Agilent Technologies). In addition, these RNA samples were also analyzed for exon expression using the Affymetrix Human 1.0 ST array, according to the manufacturer’s instructions. We analyzed the hybridization intensity and exon expression levels using the Partek Genomic Suite 6.4 (Partek Inc., St Louis, MO, USA).
The CHD1 shRNAs, were purchased from OriGene (Rockville, MD, USA) and used for transfection of human cell lines HEK293 following the manufacturer’s instructions. These shRNAs include TI355777 (5′-ATG ATG GAG CTA AAG AAA TGT TGT AAC CA), TI355778 (CAC AAG GAG CTT GAG CCA TTT CTG TTA CG), TI355779 (AGT GTC AGA TGC TCC AGT TCA TAT CAC GG), and TI355780 (AAT GGA CAC AGT GAT GAA GAA AGT GTT AG-3′). Additional shRNAs were designed and constructed according to our previously published method (Sui et al., 2002; Sui and Shi, 2005) for lentiviral infection of the MPECs. We directly subcloned a control shRNA with a scrambled sequence of ‘5′-GGG CCA TGG CAC GTA CGG CAA G-3′’ and a CHD1 shRNA with a target sequence of ‘5′-GAA GAT GTG GAA TAT TAT AAT T-3′’ into a lentiviral vector pLU containing a puromycin-resistant gene that we previously described (Deng et al., 2009), in addition to the shRNAs against CHD1 and non-specific target from OriGene. The shRNA-containing lentiviruses were packaged following a standard protocol (Deng et al., 2007) and used to infect MPECs (Barclay et al., 2008). Two days post infection, 1.5 μg ml−1 of puromycin was added to select the transfected cells for at least 3 days and then used for downstream assays. We performed immunostaining and western blotting according to a previous protocol (Sui et al., 2002) with modifications (Supplementary Method) using a CHD1 antibody (Bethyl Laboratories, Inc., Montgomery, TX, USA).
Clonogenic analysis was performed as described previously (Cao et al., 2010). Briefly, cells infected by the lentivirus expressing the control shRNA, and CHD1shRNA were individually plated at different densities (125, 250, 500, 1000 and 2000 per dish) in 6-cm cell culture dishes. After 7–10 days, the cells were fixed in 10% formalin and stained by 0.1% crystal violet. Photoshop software was used to quantify the pixels of the stained cells.
According to a previously published method of 3D culture (Barclay and Cramer, 2005), MPECs infected by the control shRNA and CHD1 shRNA were individually trypsinized and resuspended in the collagen matrix at a density of 8 × 104 cells per ml. A 0.5 ml aliquot of this solution was dispensed into each well of a 24-well plate (4 × 104 cells per well). After the gel solidified at room temperature, 1.0 ml of culture medium was added to the top of the gel in each well. The cells were maintained in the 3-D culture for 8–14 days in a cell culture incubator with media changes every other day. At the end of the incubation period, the branch structures for pseudoductal morphogenesis were imaged using digital photomicrography, and the amount of outgrowth from spheroids over time was determined using Photoshop and digital photomicrography.
The significance in the number of additional HODs in the tumor genomes with complete loss of CHD1 in comparison to those in the tumor genomes either with PTEN HOD or without CNAs at PTEN and/or CHD1 was assessed using Wilcoxon’s two-sample test. We used Fisher’s exact test to test the associations between loss of CHD1 and other significant CNAs identified by the GISTIC across the whole genome. We used the binomial proportion test to assess the significance in the distribution of additional HODs in the tumors that harbored complete loss of CHD1. All the statistical analyses were performed using SAS software version 9.2 (SAS Institute Inc., Cary, NC, USA).
The study is partially supported by the National Institutes of Health Grants CA135008 and CA133066 (to W Liu and WB Isaacs), CA119069 (to J Xu), CA131338 and 133009 (to SL Zheng and WB Isaacs). We thank Tamara Adams for editing the manuscript. The support of William T Gerrard, Mario Duhon, Jennifer and John Chalsty and P Kevin Jaffe (to WBI) is gratefully acknowledged.
Conflict of interest The authors declare no conflict of interest.
Supplementary Information accompanies the paper on the Oncogene website (http://www.nature.com/onc)