Background
Metalloproteinases play important roles in tumor formation and development [
1]. Matrix metalloproteases (MMPs) represent the most prominent family associated with tumorigenesis [
2]. They are regarded to facilitate tumor progression by degradation of the extracellular matrix (ECM) and by promotion of cancer cell migration. The evolutionarily conserved ADAM (a disintegrin and metalloprotease)-family of cell-bound proteinases mediate the release of cell surface proteins such as growth factors. In particular, ADAM10 and ADAM17 appear to promote cancer progression by releasing HER/EGFR ligands. These proteases are even discussed as potential targets for cancer therapy [
3,
4].
Much less is known about the function and relevance of their close relatives, the ADAMTS (ADAMs with thrombospondin motifs) [
5]. These secreted proteins share several structural features with MMPs and ADAMs, but are additionally characterized by the presence of thrombospondin motifs which allow them to bind to the ECM. So far, nineteen members of this protease family have been identified in humans [
6]. Even though all are presumed to be proteolytically active, many of them are still marked as orphan ADAMTSs without known function or substrate. Some others were found to act as aggrecanases and versicanases and are thus involved in ECM degradation and connective tissue turnover [
7,
8].
In recent years, accumulating evidence suggests that ADAM/ADAMTS proteins might play an essentially important role in carcinogenesis [
9‐
12]. This multistep-process involves multiple genetic and epigenetic changes [
13], which cause gain of function or activation of oncogenes and loss-of function or inactivation of tumor suppressor genes. Changes in the methylation pattern are a major mechanism controlling the expression and activity of tumor related genes. DNA methylation at promoter and particularly transcription start sites as well as gene body DNA demethylation have been recurrently correlated with inactivation of tumor-suppressor genes [
14,
15]. Moreover, such epigenetic changes have been considered promising tools for the early diagnosis of cancer.
While only limited information has been published about potential epigenetic controls of
ADAM ectodomain sheddases, several
ADAMTS family members have been described as epigenetic targets and are presumed to act as tumor suppressors. The best described family member
ADAMTS1 was inter alia identified as epigenetically deregulated gene in colorectal and gastric cancer [
16‐
18].
ADAMTS9 shows high frequency of promoter methylation in esophageal, nasopharyngeal, gastric, colorectal, pancreatic cancer and multiple myeloma [
19,
20].
ADAMTS18 was found to be frequently epigenetically silenced in oesophageal, nasopharyngeal and multiple other carcinomas [
21,
22]. ADAMTS16 shows substantial structural similarity to ADAMTS18 [
23], however, little is known about its function or regulation [
24].
In this study, we report the evaluation of DNA methylation in genes of the ADAM and ADAMTS families in matched colorectal cancer (CRC), lung cancer (LC) and oral squamous-cell carcinoma (SCC) patient samples. Quite remarkably, ADAMTS16 promotor hypermethylation was found in all epithelial cancer subtypes analyzed. Moreover, ADAMTS16 protein expression was strikingly decreased in CRC patient samples. Finally, overexpression of ADAMTS16 in HT29 colorectal cancer cells dramatically decreased cell growth. Thus, our data suggest that ADAMTS16 may act as tumor suppressor in certain epithelial cancers.
Methods
Patient samples
CRC samples originated from the German National Genome Research Project “Integrated genomic investigation of colorectal carcinoma” were obtained from the Kiel BMB-CCC (biomaterial bank of the Comprehensive Cancer Center, University Hospital of Schleswig-Holstein, Campus Kiel, Germany). The samples were obtained from fresh unfixed surgical resectates, split by pathologists into tumor tissue and adjacent peri-tumoral non-malignant tissue (as controls), and were snap-frozen in liquid N2 and stored in the biobank at − 80 °C until further use. The tissue samples originated from various colon locations. In total, samples from 117 patients were investigated.
Matched LC tissue samples (tumor-free lung and tumor) were obtained from patients undergoing pneumectomy or lobectomy at the LungenClinic Grosshansdorf, Germany (n = 40) in the course of surgical treatment of previously diagnosed lung cancer.
Native tissue samples from patients suffering from oral lichen planus and/or oral squamous-cell carcinoma (n = 15) were collected from consultation hours for oral mucosa at the Department of Cranio-Maxillofacial Surgery, University Hospital of Schleswig-Holstein, Kiel Campus, Kiel, Germany. As control samples, non-inflamed tissue from the same patient was collected.
DNA methylation analysis
Genomic DNA extraction was done using DNeasy kit (Qiagen, Germany). DNA samples were bisulfite converted with the EZ DNA Methylation™ Kit (Zymo Research Corporation, USA) and afterwards measured for DNA methylation with the Infinium Human Methylation 450 k BeadChip (Illumina Inc., USA) according to the manufacturer’s protocol. The generated IDAT files were further processed with the Genome Studio Software (version 2011.1; Methylation Analysis Module version 1.9.0, Illumina) to derive the β-values. Thereby internal array controls and the default settings were used for data normalization. Methylation levels in Illumina Methylation assays are quantified using the ratio of intensities between methylated and unmethylated alleles. The β-values are continuous and range from 0 (unmethylated) to 1 (completely methylated) [
25].
Cell culture and transfection
Mycoplasma free HT29 cells were purchased from the American Type Culture Collection (ATCC), and grown in high glucose DMEM (Thermo Fisher Scientific) supplemented with 10% fetal calf serum (FCS) and 1% penicillin/streptomycin (Pen/Strep). Cells were transfected using Turbofect Transfection Reagent (Thermo Fisher Scientific) according to the manufacturer’s instructions. 24 h after transfection, cells were transferred to the X-Celligence device and in parallel approaches harvested for immunoblot analysis.
Impedance based xCELLigence proliferation assay
The xCELLigence invasion assay (ACEA Biosciences, USA) is based on changes in electrical impedance at the interphase between cell and electrode as migrating cells move through a barrier. These changes can be directly correlated with the proliferative capacity of seeded cells. The technique provides an advantage over existing standard proliferation assays, since the data is obtained continuously in real-time, when compared to end-point analysis in other methods. To analyze cell proliferation, HT29 cells were seeded at a density of 20,000 cells/well on E16 plates. The impedance value of each well was automatically monitored by the xCELLigence system for duration of 24 h and expressed as a CI (cell index) value. Averages of duplicates are shown derived from three independent experiments. The rate of cell growth was determined by calculating the slope of the line between the starting point and 24 h.
Western blot analysis
Cells were washed once with PBS and lysed in lysis buffer (5 mM Tris-HCl (pH 7.5), 1 mM EGTA, 250 mM saccharose, 1% Triton X-100) supplemented with cOmplete inhibitor cocktail (Roche Applied Science) and 10 mM 1,10-phenantroline monohydrate. Equal amounts of protein were loaded on 10% SDS-PAGE gels. The samples were electrotransferred onto polyvinylidene difluoride membranes (Hybond-P; Amersham) and blocked overnight with 5% skim milk in Tris-buffered saline (TBS). After incubation with anti-ADAMTS16 antibody (Santa Cruz, sc-50,490) in blocking buffer, the membranes were washed three times in TBST (TBS containing 0.1% Tween-20). Primary antibody was detected using affinity-purified peroxidase (POD)-conjugated secondary antibody (1:10,000) for 1 h at room temperature. Detection was carried out using the ECL detection system (Amersham). Signals were recorded by a luminescent image analyzer (Fusion FX7 imaging system; PEQLAB Biotechnologie). Equal loading as well as efficiency of transfer were routinely verified by reprobing the membrane for tubulin (DSHB clone E7).
Immunohistochemistry
Cryosections (7 μm) of the CRC samples were fixed with acetone. Slides were incubated in 3% H2O2 in PBS for 30 min. After blocking of the nonspecific binding (0.75% BSA in PBS), the sections were incubated with anti-ADAMTS16 antibody (Origene, dilution 1:100) over night. The staining was visualized by peroxidase-conjugated secondary antibody and diaminobenzidine (Vector labs). Finally, sections were counterstained by hemalum and embedded in Kaiser’s glycerol gelatine and photographed with an Axioplan microscope (Zeiss, Germany). The corresponding negative controls were stained omitting the anti-ADAMTS16 antibody.
Statistical analysis
Comparison of the DNA methylation status of patient matched tumor and peritumoral non-malignant DNA samples was performed using the script language R 3.2.2 (R foundation), Graphpad Prism 5.04 (GraphPad Software Inc., USA) and Excel 2010 (Microsoft, USA). CpGs were defined as differentially methylated if the difference of the mean β-values (∆βmean) was larger than 0.2 (|∆βmean| ≥ 0.2) compared to the control and significant after Wilcoxon signed-rank testing with Benjamini-Hochberg multiple testing correction for the 1145 tests performed (P < 0.05). CpGs which did not meet these criteria, but showed a methylation difference of 0.1 ≤ |∆βmean| < 0.2 (P < 0.05) were defined as intermediate methylated.
Discussion
CpG promoter hypermethylation has been demonstrated to be a frequent event during carcinogenesis. In this study, we aimed to find out whether members of the ADAM and ADAMTS family might represent novel gene targets epigenetically inactivated in epithelial tumorigenesis. Comparing malignant and non-malignant tissues of the same patients, we identified ADAMTS16 as a gene with cancer-specific promoter hypermethylation in CRC, LC and SCC patients.
Several ADAM family members, particularly ADAM9, ADAM10, ADAM12, ADAM15 and ADAM17, have been implicated in cancer formation and progression. ADAM10 and ADAM17 are even discussed as potential targets for cancer therapy [
3]. However, except for
ADAM12, we did not find relevant changes in the DNA methylation pattern in any of these tumor-associated proteases. The changes observed for
ADAM12 were located in the gene body and only found in CRC but not in SCC or LC patients. Overall, our findings indicate that differences in gene DNA methylation are unlikely to be responsible for the control of ADAM function in tumors. Instead, these enzymes seem to be rather controlled by posttranslational mechanisms. This assumption is in accordance with recent data stressing the relevance of protein maturation, localization and cell membrane changes for protease activation [
26,
27].
In contrast to the
ADAM family, epigenetic silencing and genetic inactivation in
ADAMTS family members has been frequently reported. This observation led to the concept that these protease family members could be important tumor suppressors.
ADAMTS15 is genetically silenced in human colorectal cancer [
28].
ADAMTS1 and
ADAMTS9 have been found to be epigenetically silenced in diverse malignant tumors [
16,
19].
ADAMTS12 has been identified as potential tumor suppressor in colorectal cancer [
29].
ADAMTS8 was shown to be differentially methylated in brain, thyroid, lung, nasopharyngeal, esophageal, gastric and colorectal cancers [
30]. Also
ADAMTS18 has recently been identified as tumor suppressor gene. Differential methylation has been reported in renal, gastric, colorectal, pancreatic, esophageal, and nasopharyngeal carcinomas [
21,
22].
ADAMTS16 shares conspicuous structural similarity with ADAMTS18 [
23]. However, ADAMTS16 is one of the least examined proteins from the whole ADAMTS family and little is known about its function. Today, the only known substrate of ADAMTS16 is α2-macroglobulin [
31], a general inhibitor of proteases. In this context, an involvement in the human ovarian follicle maturation has been proposed [
32]. The role of ADAMTS16 in tumorigenesis is not clear. So far, no epigenetic modifications have ever been reported for this protease.
Here, we identified
ADAMTS16 as commonly differentially methylated gene in three different types of epithelial cancers.
ADAMTS16 promoter hypermethylation at six CpGs immediately upstream of the transcription start site and hypomethylation in two CpGs in the gene body is very suggestive of decreased protein expression. To establish whether this would be the case, we analyzed CRC tumors and non-tumorous patient samples via immunohistochemistry. These analyses revealed that expression of ADAMTS16 is markedly decreased in CRC. The possibility that this might be causally linked to CpG-hypermethylation within the promoter region was supported through analysis of data provided by The Cancer Genome Atlas (TCGA,
http://cancergenome.nih.gov/, accessed on 05.02.2015) for a colon adenocarcinoma and rectum adenocarcinoma cohort (COADREAD,
n = 44 (ctrl),
n = 384 (canc)). These data are based on non-matched control and cancer samples. Gratifyingly, the same methylation changes in the 8 commonly differentially methylated CpGs that we described for CRC, LC and SCC patients were found. Gene expression analysis for the same TCGA COADREAD cohort (
n = 22 (ctrl),
n = 224 (canc)) revealed that
ADAMTS16 mRNA expression was significantly decreased from 0.29 in the control (ctrl) to 0.04 in the cancer tissue (canc) (
P < 0.0001). This decrease reflects a reduction of the
ADAMTS16 mRNA expression of 86.3%.
It became of immediate interest to investigate whether expression of ADAMTS16 might impact on a cellular function linked to carcinogenesis. Assessment of cell proliferation was chosen as a first approach in this direction. Overexpression of ADAMTS16 in HT29 colorectal cancer cells significantly reduced cell proliferation. These data are in accordance with data by Surridge et al., who showed that overexpression of ADAMTS16 in chondrosarcoma cells led to a decrease in cell proliferation and migration [
24]. However, further analyses of the ADAMTS16 effects on tumor cell migration and invasion are warranted in order to find out whether
ADAMTS16 might represent a novel tumor suppressor gene for CRC, LC and SCC.
Acknowledgements
The plasmid for ADAMTS16 expression was kindly provided by Ian M. Clark [
24]. The support of the technical staff of the molecular genetic and epigenetic laboratories of the Institute of Human Genetics in Kiel is gratefully acknowledged.