Identification of ZDHHC14 as a novel human tumour suppressor gene

Genomic changes affecting tumour suppressor genes are fundamental to cancer. We applied SNP array analysis to a panel of testicular germ cell tumours to search for novel tumour suppressor genes and identified a frequent small deletion on 6q25.3 affecting just one gene, ZDHHC14. The expression of ZDHHC14, a putative protein palmitoyltransferase with unknown cellular function, was decreased at both RNA and protein levels in testicular germ cell tumours. ZDHHC14 expression was also significantly decreased in a panel of prostate cancer samples and cell lines. In addition to our findings of genetic and protein expression changes in clinical samples, inducible overexpression of ZDHHC14 led to reduced cell viability and increased apoptosis through the classic caspase‐dependent apoptotic pathway and heterozygous knockout of ZDHHC14 decreased cell colony formation ability. Finally, we confirmed our in vitro findings of the tumour suppressor role of ZDHHC14 in a mouse xenograft model, showing that overexpression of ZDHHC14 inhibits tumourigenesis. Thus, we have identified a novel tumour suppressor gene that is commonly down‐regulated in testicular germ cell tumours and prostate cancer, as well as given insight into the cellular functional role of ZDHHC14, a potential protein palmitoyltransferase that may play a key protective role in cancer. © 2014 The Authors. The Journal of Pathology published by John Wiley & Sons Ltd on behalf of Pathological Society of Great Britain and Ireland.


Introduction
Tumourigenesis is a multistep process in which, elicited by the activation of oncogenes and the inactivation of tumour suppressor genes (TSGs), cells acquire increasingly malignant characteristics. TSGs prevent tumour initiation and progression by acting in signalling networks that regulate cell proliferation, differentiation, apoptosis, and other cellular and systemic processes. TSGs can be inactivated by deletions, point mutations [1] or by epigenetic mechanisms such as DNA methylation and histone modifications [2]. Among them, deletion is one of the most frequent genetic alterations detected in cancer cells and recurrent deletions detected in certain regions have been widely interrogated to identify TSGs. The development of comparative genomic hybridization and DNA microarray technologies has made it easy to identify genomic copy number alterations and has accelerated the discovery of TSGs. Single nucleotide polymorphism (SNP) arrays allow the determination of both genomic copy number changes and loss of heterozygosity (LOH), including copy number neutral LOH [3,4]. Many novel TSGs have been discovered in human malignancies using SNP array analysis, including PAX5 as a key target of genetic inactivation in B-cell acute lymphocytic leukaemia [5], PTPRD as a tumour suppressor in lung cancer and glioma [6,7], and IKFZF1 in acute lymphoblastic leukaemia [8].
Testicular germ cell tumours (TGCTs) are the most common malignancy in young men. Many genetic studies have been carried out in TGCTs and the ZDHHC14, a novel tumour suppressor gene 567 characteristic genetic change, gain of 12p [9], has been identified. Previous low-resolution genetic studies of TGCTs identified more genomic copy number gains rather than losses [9]. Interestingly, deregulation of important tumour suppressor genes such as RB1 , TP53 , APC , WT1 , MCC , and BRCA1 is rare in this type of cancer [10][11][12]. However, few array-based genomewide studies have been performed for TGCTs. Using the Affymetrix 10 K SNP array, we analysed a series of TGCTs to identify genome-wide copy number changes, in particular frequently occurring small genomic region gains and losses. We found frequent deletion of a small genomic region within 6q25.3 affecting only one known gene, ZDHHC14 . Further genetic studies revealed additional genomic alterations and reduced expression of ZDHHC14 not only in TGCTs, but also in prostate cancer. Using cell-based assays and xenografts, we performed functional studies that identified ZDHHC14 as a novel tumour suppressor.

Cell lines
Three human TGCT cell lines, 833 K, Susa, and GCT27 (obtained from Jennifer Parrington, John Masters, and Lloyd Kelland, respectively), five human prostate cancer cell lines, PC-3, DU-145, LNCaP, VCaP, and 22RV1 (ATCC, Manassas, VA, USA), and one immortalized human prostate epithelial cell line, PNT2-C2 (obtained from Norman Maitland), and the human embryonic kidney HEK293 (ATCC) and T-REx TM 293 (Invitrogen, Carlsbad, CA, USA) cell lines were used in this study. All cell lines were verified by STR profiling using the ABI AmpF/STR Identifiler kit (Applied Biosystems, Foster City, CA, USA).

Clinical samples
Fresh frozen TGCT and prostate cancer samples were collected and stored in ethically approved tissue banks, and used in this study under local ethics committee approval. Archived paraffin blocks of TGCT and prostate cancer samples were collected for tissue microarray (TMA) studies following approvals obtained from local ethics committees. Written informed consent was received from participants prior to inclusion in the study.

SNP array analysis
A standard phenol/chloroform method was used to extract DNA from cell lines and fresh frozen tissues with the application of micro-or macro-dissection for clinical prostate cancer samples. Affymetrix 10 K SNP arrays (Affymetrix, Santa Clara, CA, USA) were used according to the manufacturer's instructions, with standard quality control steps in the process. Standard Affymetrix criteria for final call rate and signal intensity were applied to define a successful hybridization.
Signal intensity data from SNP arrays were analysed using the commercial Partek Genomics Suite (Partek Incorporated, St Louis, MO, USA) and our in-house GOLF (V2.2.10) software [13][14][15]. SNP array data have been deposited in the Gene Expression Omnibus database (accession number GSE35216).

Quantitative reverse transcription PCR (Q-RT-PCR)
Total RNA was extracted using Trizol (Invitrogen) from cell lines and fresh frozen tissues and for clinical prostate cancer samples micro-or macrodissection was used to isolate high-purity tumour cells. Standard Q-RT-PCR was performed using predesigned Taqman  gene expression assays targeting ZDHHC14 (Hs00257233_m1) and the endogenously expressed GAPDH gene (Hs99999905_m1) (Applied Biosystems).

Generation of ZDHHC14 antibody and immunohistochemistry
A polyclonal rabbit antibody targeted to two peptides (H 2 N-CGLASQDSLHEDSVRG-COOH and H 2 N-CPRATPDEAADLERQID-COOH) of human ZDHHC14 was generated through Eurogentec (Liège, Belgium) using the First Class program. Immunohistochemistry was performed using the Vector ABC kit (Vector Laboratories Inc, Burlingame, CA, USA) and ZDHHC14 antibody, and protein expression scored as previously described [16].

Next-generation sequencing
The genomic locus of ZDHHC14 (292.42 kb) was amplified by long-range PCR using the Gene Amp XL PCR kit (Applied Biosystems) (with primers listed in Supplementary Table 1). We generated DNA libraries using the 454 GS FLX Titanium General Library kit (Roche, Basel, Switzerland) and sequenced them in a 454 Genome Sequencer FLX Instrument (Roche).

Promoter activity and DNA methylation analysis
HEK293 cells were transfected with the promotercontaining VISA plasmid and the empty VISA plasmid [17], and luciferase activity was measured using the One-Glo Luciferase Assay System (Promega, Madison, WI, USA). Genomic DNA was bisulphite-converted with the Epi-Tect Bisulfite kit (Qiagen, Venlo, The Netherlands). Biotin-labeled primers specific for three regions within the ZDHHC14 promoter region and one upstream CpG island were used to amplify the bisulphite-converted DNA (see primer sequences in Supplementary Table 4). The amplified DNA was used for pyrosequencing as previously described [18].

Cell viability, proliferation, and apoptosis assays
Cell viability and proliferation were assessed using the CellTiter 96  AQ ueous assay (Promega) following the manufacturer's instructions. 1500 cells were plated in each well of a 96-well plate. The following day, cells were treated with 1 µg/ml tetracycline to induce ZDHHC14 expression and after 72 h, cell viability and proliferation were assessed. Cell cycle distribution was established using a FACSCalibur flow cytometer (Becton Dickinson, Franklin Lakes, NJ, USA) as previously described [19], analysed using FlowJo 8.8.6 software. Apoptosis assays were performed using an FITC annexin V apoptosis detection kit I (Becton Dickinson) following the manufacturer's instructions.
ZDHHC14 knockout by zinc finger nucleases PNT2-C2 cells were transfected with ZDHHC14 CompoZr TM Knockout Zinc Finger Nucleases (CKOZFN23137-1KT, Sigma) using Lipofectamine and cultured for 3 days. Cells were split and used for the CEL-I surveyor assay (Transgenomic, Omaha, NE, USA) to confirm the action of the ZFNs, or seeded for clonal selection. One hundred clones were expanded and screened for the presence of ZFN-induced deletions by microsatellite analysis, using a 3730xl DNA analyser (Applied Biosystems). Positive clones were sequenced to confirm the deletion.

Colony formation assay
Five hundred cells were plated onto six-well plates and incubated for 1 week in 5% CO 2 at 37 • C, washed with PBS, fixed with 4% paraformaldehyde, and stained with 0.1% crystal violet solution. The number of visible colonies was counted.

Xenograft and bioluminescence studies
Male CD1 nude mice (Charles River Limited, UK), 6-8 weeks of age, were used in accordance with the UK Coordination Committee on Cancer Research guidelines and Home Office regulations. Mice were randomly allocated into two groups: one treated with tetracycline and the other treated with vehicle. One week after the treatment started, mice were injected subcutaneously with 2 × 10 6 293 T-REx TM clone Z14 3 luciferase-tagged cells into their right-side flank region and with the empty plasmid 4/TO luciferasetagged cells into their left flank. Mice were anaesthetized and bioluminescence imaging was performed after intraperitoneal injection of D-luciferin (Caliper Life Sciences, Hopkinton, MA, USA). The IVIS Vivo-Vision 100 Imaging System (Xenogen, Alameda, CA, USA) was used for bioluminescence imaging recording and post-processing. Quantification was performed using Living Image software 3.2.

Statistics
Statistical tests for data analysis included the chisquared test and Student's t-test, as described in each section. Statistical analyses were performed using the Prism 5.0b (GraphPad, La Jolla, CA, USA) statistical software package. p values of 0.05 or less were considered significant. Full experimental procedures are presented in the Supplementary materials and methods.

Frequent loss of 6q25.3 region affecting ZDHHC14 in TGCTs
To search for new genomic alterations, and in particular to identify novel putative tumour suppressor genes, we performed SNP array analysis on 36 TGCT clinical samples, including five with case-matched adjacent normal tissues, and three cell lines (833 K, Susa, and GCT27). The samples without case-matched normal tissues and cell lines were normalized against the average of the five normal tissues. We confirmed previously reported frequent genomic alterations such as gain on chromosomes 7, 8, 12p, and X, and loss of 4, 5, 11, and 13 [9]. In addition, there were a number of small regions (<1 mb) with genomic copy number gains or losses that were mostly detected in one or two samples, except a loss of a small genomic region on 6q25.3 that was detected in 48.7% (19/39) of the cases. In most samples, this loss was defined by only two contiguous SNPs (Supplementary Figure 1). In three of five cases with adjacent normal tissues, this 6q25.  The sub-microscopic deletion of 6q25.3 (indicated by circles) was found in three of these cases. In cases 74 T and 88T, as in most of the non-matched cases, the deletion was defined by only two contiguous SNPs. In case 87T, the SNP on the left of the two circled may also be deleted. Each dot represents an SNP and the ZDHHC14 locus is shown within the box. In each case, the middle line represents a log 2 ratio of 0 for SNP signal intensity; the top line a ratio of 1; and the bottom line a ratio of −1.
known gene, ZDHHC14 , which encodes a protein with an unknown function (Supplementary Figure 1).

ZDHHC14 expression is commonly down-regulated in TGCTs
After identifying the ZDHHC14 deletion, we investigated its consequence on expression of this gene. Firstly, we measured ZDHHC14 mRNA expression in 40 TGCT clinical samples and three cell lines and found dramatic underexpression when compared with three normal testis tissue samples (p < 0.0001) ( Figure 2A). We also studied ZDHHC14 protein expression by immunohistochemistry. As no commercial antibody was available, we generated an antibody against human ZDHHC14 and confirmed its specificity by western blot and immunocytochemistry ( Supplementary Figures 2A-2C), and demonstrated its efficiency in staining ZDHHC14 in normal germ cells (spermatocytes) (Supplementary Figure 2D). We analysed TMAs comprising 318 cases of TGCTs for ZDHHC14 expression and found that ZDHHC14 was significantly underexpressed compared with morphologically normal seminiferous tubules adjacent to the cancers (p = 0.0013). ZDHHC14 levels were divided into three categories: negative, weakly positive, and strongly positive. All normal testicular tissue samples were positive, whereas 21.4% of TGCT samples were negative ( Figures 2B and 2C).

ZDHHC14 expression is down-regulated in prostate cancer
Since TSGs often are involved in the genesis of more than one type of malignancy, and the Oncomine database revealed that ZDHHC14 mRNA was also down-regulated in lymphoma, liposarcoma, brain, kidney, lung, and colorectal cancers (data from the Oncomine database http://www. oncomine.org/resource/login.html), we investigated ZDHHC14 expression in prostate cancer. We detected a clear down-regulation of ZDHHC14 mRNA in 38 cancer samples compared with 21 benign prostatic hyperplasia (BPH) samples (p < 0.0001). ZDHHC14 mRNA was also underexpressed in the prostate cancer cell lines PC-3, DU-145, 22RV1, LNCaP, and VCaP ( Figure 3A). We next studied ZDHHC14 protein expression by immunohistochemistry, analysing 483 prostate cancer clinical samples, 73 with adjacent BPH or normal prostate glands, made into 11 TMAs and 50 BPH samples from non-cancer cases constructed in one TMA. Within the prostate gland, basal cells were strongly stained. The luminal epithelial cells in all normal and BPH samples were also positively stained. However, only 53% (256/483) of the cancer samples were positively stained (p < 0.001; chi-squared test) ( Figures 3B and 3C). This result taken with those from the Oncomine database suggests that ZDHHC14 down-regulation is a feature of many human cancers. We re-examined our published SNP array data [15,16] for genomic loss of the ZDHHC14 region, which was found in five cases from 71 clinical samples and six cell lines.   The coverage of the ZDHHC14 gene in the pools ranged from 78% to 92%, with the average depth for each sample from 11 to 145 reads. The quality (Phred score) was above 30 in all the samples (Supplementary  (Table 1). Although the missense mutations were scored as damaging using SIFT mutation prediction software (http://sift.jcvi.org/), the finding of mutations solely in cell lines indicates that ZDHHC14 mutation may not greatly contribute to its inactivation in those two tumour types. We also assessed promoter methylation status as a possible cause leading to ZDHHC14 downregulation. The promoter region was unmethylated in both non-malignant and malignant prostate and  testicular clinical samples (Supplementary Figure 4), ruling out promoter hypermethylation as the cause for ZDHHC14 underexpression in prostate cancer and TGCTs.

ZDHHC14 activates a classic caspase-dependent apoptotic pathway
To confirm the tumour suppressor role of ZDHHC14 and also to attempt to elucidate its cellular function, we generated a tetracycline-inducible ZDHHC14overexpressing stable 293 T-REx cell line and isolated three clones that overexpressed ZDHHC14 at different levels ( Figure 4A). Overexpression of ZDHHC14 decreased cell number in all three clones, with this decrease being directly proportional to ZDHHC14 expression levels ( Figure 4B). This result was due to ZDHHC14 overexpression inducing apoptosis, as determined by quantification of annexin V-positive cells and sub-G1 apoptotic cells by FACS analysis (Figures 4C and 4D), which were again proportional to ZDHHC14 levels. We also found cleavage of caspase-7 and PARP ( Figure 4E), showing that the classic caspase-dependent pathway was involved. We then performed transient overexpression of ZDHHC14 in 22RV1 prostate cancer cells, which led to an increase in sub-G1 apoptotic cells together with the cleavage of caspase-7 and PARP ( Figure 5). This result confirms the involvement of ZDHHC14 in inducing apoptosis and supports its role as a tumour suppressor in prostate cancer. In order to determine if ZDHHC14's apoptotic function was due to its palmitoyltransferase activity, we transiently transfected HEK293 cells with either wild-type ZDHHC14 or a mutated catalytic DHHC domain construct. We detected that the decrease in cell viability and increase in apoptosis were lower in the mutant construct transfected cells than in cells transfected with the wild type, although the mutated construct still had lower viability levels than the empty vector transfected cells, probably due to the inherent toxicity of overexpressing a protein (Supplementary Figure 5). These data demonstrate the involvement of the catalytic DHHC domain in inducing apoptosis.

ZDHHC14 heterozygous deletion increases colony formation
To further confirm the tumour suppressor role of ZDHHC14, we attempted its knockdown by siRNA in several cell lines, obtaining efficient mRNA knockdown, but no decrease at protein level. This was a consequence of the slow turnover rate of ZDHHC14 protein, as determined by persistent detection of ZDHHC14 expression after 6 days' treatment with the protein synthesis inhibitor cycloheximide (Supplementary Figure 6). shRNA stable knockdown was also attempted but without efficient knockdown even at the RNA level. Finally, we used zinc finger nucleases (ZFNs) to knock out ZDHHC14 in immortalized prostate epithelial PNT2-C2 cells. After screening 100 clones derived from cells transfected with ZDHHC14 ZFNs, we found a clone with an 11 bp deletion in one of the two alleles. This deletion generates a stop codon immediately after it ( Figure 6A). This heterozygous deletion decreased ZDHHC14 protein expression compared with a mock transfected clone and with the parental cells ( Figure 6B and Supplementary Figure  2). Colony formation assays showed both more and larger colonies in the ZDHHC14 heterozygous deleted cells than in parental and mock transfected cells ( Figures 6C and 6D), further confirming the tumour suppressor role of ZDHHC14 .

Increased ZDHHC14 expression suppresses xenograft tumour initiation
To assess the role of ZDHHC14 in tumour growth in vivo, we employed xenograft experiments using the 293 T-REx cells described above. We chose clone 3, with the highest ZDHHC14 levels, and the empty plasmid clone as a control. We randomly allocated ten mice into two groups and initiated the treatment by administrating tetracycline in drinking water, or vehicle for the control group. After 7 days, all mice in each group were injected subcutaneously with the ZDHHC14 plasmid-containing cells into the right flank and with the empty plasmid cells into the left flank, using each mouse as its own control. The presence of tumour cells and tumour growth was checked periodically, showing that tetracycline-dependent ZDHHC14 overexpression completely blocked tumour initiation, since none of these mice developed tumours in the right flank (Figure 7). In contrast, the same cells in the vehicle-treated mice grew to form a tumour in all cases (data not shown). As expected, cells transfected with empty plasmid developed tumours in all the mice in the tetracycline group ( Figure 7) and in the vehicle group (data not shown). These results demonstrate that ZDHHC14 has a tumour suppressor role in vivo.

Discussion
Tumour suppressor genes play critical roles in preventing tumourigenesis and are frequently inactivated during tumour development and progression. Deletion of 6q has been reported frequently in human tumours and one or more TSGs located within this region have been proposed [20][21][22]. Using SNP microarrays on TGCTs, we detected a frequent small deletion in chromosome region 6q25.3 containing only one known gene, ZDHHC14 . This region is also deleted in prostate cancer samples, although at a much lower frequency [15,16]. These data suggest that ZDHHC14 may be a putative TSG. This hypothesis has been further supported by the dramatic ZDHHC14 down-regulation found in TGCT and prostate cancer. Currently, there are only three studies on ZDHHC14, which paradoxically reported a putative oncogenic role for ZDHHC14 based on its up-regulation in gastric cancer [23], diffuse large B-cell and mantle cell lymphomas [24], and acute biphenotypic leukaemia and subsets of acute myeloid leukaemia [25]. Based on the different substrate specificities of DHHC proteins [26], it is plausible that ZDHHC14 has several substrates and, depending on the cellular type and context, can regulate different pathways. Therefore, ZDHHC14 might play oncogenic roles in some cell types and tumour suppressor roles in others. However, in publicly accessible databases, there is much evidence to support the tumour suppressor role of ZDHHC14 . ZDHHC14 is underexpressed in many human cancers, including lymphoma, liposarcoma, and brain, kidney, lung and colorectal cancers (data from the Oncomine database http://www.oncomine.org/resource/login.html). In addition, since the cancer genome sequencing efforts started, 36 mutations affecting the ZDHHC14 coding region have been described and ZDHHC14 gene mutation recurrence has been found in colon adenocarcinoma ( [28], colorectal cancer (2/72 samples) [29], and skin cutaneous melanoma (4/228 samples) (data from The Cancer Genome Atlas database http://cancergenome.nih.gov/). All these data support ZDHHC14 as a putative TSG that is commonly affected in many types of human cancer.
The inactivation of a TSG can be achieved by its complete switching off (two-hit mechanism), by haploinsufficiency, or even by its subtle downregulation (quasi-insufficiency) [30]. In this study, nearly all ZDHHC14 deletions were heterozygous and in most cases, ZDHHC14 was down-regulated but not completely lost. In addition, the limited number of missense mutations found means that in most of the cases, the other allele may remain functional. These results suggest that haploinsufficiency or quasi-insufficiency is enough to inactivate the tumour suppressor role of ZDHHC14. This mechanism is also supported by our functional studies, which showed a ZDHHC14 dose response effect in suppressing cell growth and activating apoptosis, and an increased colony formation ability by heterozygous knockout.
Regarding the mechanism leading to ZDHHC14 down-regulation, ZDHHC14 deletion has a pivotal role in TGCTs. However, some samples without the deletion expressed lower ZDHHC14 levels than samples with the deletion, indicating that other mechanisms may be involved. In both TGCTs and prostate cancer, our data showed that ZDHHC14 mutation and DNA methylation may not contribute much to its inactivation. Therefore, other transcriptional or posttranscriptional mechanisms should be the main cause of ZDHHC14 down-regulation in prostate cancer cases without genomic deletions.
Our final approach to confirm the tumour suppressor role of ZDHHC14 was to perform functional studies. While its cellular function has not been studied, ZDHHC14 belongs to the DHHC palmitoyltransferase family. It was recently discovered that the main function of DHHC proteins is protein palmitoylation, a reversible post-translational lipid modification that regulates membrane tethering and localization/function of key proteins in cell signalling and membrane trafficking [26,31]. Deregulation of certain palmitoyltransferases has been reported in several diseases including neurological disorders and cancer [32]. Although protein palmitoylation plays an important role in cellular function, it is poorly understood [31,33] and DHHC proteins remain understudied. There are 23 DHHC proteins in humans, of which a small number, including DHHC9 [34,35], HIP14/DHHC17 [36], Aph2/DHHC16 [37,38], and DHHC2 [39][40][41], have reported roles in cancer. We showed in vitro that ZDHHC14 heterozygous deletion increases colony formation and that overexpression suppresses cell growth by promoting apoptosis through the classic caspase-dependent pathway, which also explains why ZDHHC14 overexpression in xenograft tumours totally blocks tumour initiation. Our functional analysis using a DHHC domain mutated construct showed that palmitoylation is partially responsible for the pro-apoptotic role of ZDHHC14, which is further supported by the deletion found in exon 4 in Susa cells affecting the DHHC domain. There are also several examples linking palmitoylation and apoptosis [37,38,[42][43][44][45]. However, further studies are required to fully dissect the molecular mechanism linking ZDHHC14 and caspase-dependent apoptosis.
In summary, we have demonstrated that ZDHHC14 , a palmitoyltransferase DHHC family gene, is commonly inactivated in human cancers and that it may exert a tumour suppressor role through the induction of apoptosis. This is the first study showing the involvement of ZDHHC14 in a specific pathway, the classic caspase-dependent apoptotic pathway. These data, together with our in vivo data showing the blocking of tumour xenograft initiation, demonstrate that ZDHHC14 is a bona fide tumour suppressor gene.

SUPPORTING INFORMATION ON THE INTERNET
The following supporting information may be found in the online version of this article: Supplementary materials and methods. Table S1. List of primers used to amplify ZDHHC14 genomic locus for next-generation sequencing. Table S2. DNA pools used for the generation of next-generation sequencing libraries. Table S3. Next-generation sequencing coverage, depth, and Phred-score of the DNA samples. Table S4. List of primer sets used to amplify the bisulphite-converted DNA and perform pyrosequencing for DNA methylation analysis.