Transcriptomic analysis reveals cell apoptotic signature modified by heparanase in melanoma cells

Abstract Heparanase has been implicated in many pathological conditions, especially inflammation and cancer, attributed to its degradation of heparan sulfate, a crucial component maintaining the integrity of the extracellular matrix. By silencing the heparanase gene (HPSE) in MDA‐MB‐435s melanoma cells, we investigated the impact of this protein on gene transcription. Transcriptome sequencing yielded a list of 279 differentially expressed genes, of which 140 were up‐regulated and 239 down‐regulated. The 140 up‐regulated genes were classified into a substantial set of gene ontology defined functions, for example, positive regulation of cell death, apoptotic process, response to cytokine, while 239 down‐regulated genes classify only into the two categories: nucleosome and nucleosome assembly. Our focus was drawn to an array of 28 pro‐apoptotic genes regulated by heparanase: real‐time PCR experiments further validated up‐regulation of EGR1, TXNIP, AXL, CYR61, LIMS2 and TNFRSF12A by at least 1.5‐fold, among which EGR1, CYR61, and TNFRSF12A were confirmed on protein level. We demonstrated significantly increased apoptotic cells by TUNEL staining upon HPSE silencing, mediated by activation of caspase 3/PARP1 pathway. The pro‐apoptotic gene expression and observation of apoptosis were extended to another melanoma cell line, MV3 cells, thus consolidating the anti‐apoptosis effect of heparanase in melanoma cells.

Although there is a great number of publications with focus ranging from cell signalling to clinical studies of heparanase, to date there is no report of gene expression profile correlated to altered HPSE expression. Using melanoma cells that express a high level of endogenous heparanase, we demonstrates that heparanase regulates a number of genes involved in a substantial set of biological functions.
Focusing on its regulation on cell apoptosis, we were able to validate the up-regulation of pro-apoptotic genes and display apoptosis in cell culture after silencing HPSE.

| Cell culture
T47D cells were from ATCC, MDA-MB-435s cells were from ATCC, and recently identified as melanoma origin. 19 Melanoma MV3 cells (from Gefan, Shanghai, China) were originally from ATCC. The cells were further confirmed by authentication testing preceding this study (Eurofins, Germany). All cells were cultured in DMEM supplemented with 10% foetal bovine serum (FBS) in 5% CO 2 at 37°C.
Medium was changed twice a week.

| SiRNA silencing of HPSE gene
Melanoma cells were seeded at a concentration of 1 × 10 5 /mL and maintained in complete medium for 24 hours. SiRNA silencing was achieved by the addition of Dharmafect 1 transfection reagent with an optimized smartpool of HPSE siRNAs, Silencer Select HPSE siRNA, or a control siRNA at a concentration of 30 nmol/L. After 24 hours culture media were refreshed, silencing of HPSE expression was confirmed by real-time PCR and Western blot after 48 hours.

| Real-time PCR
Total RNA was isolated from the cells by following the manufacturer's instructions. RNA (1 μg) was reversely transcribed to cDNA with a High-Capacity cDNA Reverse Transcription Kit and diluted to a final volume of 200 μL. A 20 μL reaction mixture containing 2 μL of cDNA template, 0.5 μmol/L primers and 10 μL SsoFast EvaGreen Supermix was added to a 96-well white-clear plate. RT-PCR was performed using a CFX384 TM RT-PCR system with the BioRad CFX manager software version 3.0. Conditions for amplification were 95°C for 30 seconds and 40 cycles of 95°C for 5 seconds followed by 56°C for 5 seconds. The fold change of mRNA was evaluated by the relative copy number and expression ratios of targeted genes normalized to the expression of the reference gene (18S rDNA). Ratios were calculated by the relative quantification method using the CFX manager software with the equation RCN = 2−ΔΔC t , where ΔC t = C t target − C t reference, and ΔΔC t = ΔC t test sample − ΔC t control sample. The sequences of primers used are described in File S3.
For TUNEL assay, cells were fixed in 4% PFA followed by permeabilization with 0.25% Titan ® X-100 for 10 min at room temperature. Fixed cells were subjected to TUNEL assay using the Click-iT ® Plus TUNEL System following the manufacture's instruction. For detection of activated caspase 3/7-positive cells, CellEvent TM Green Detection Reagent was added to the medium, and cells incubated for 30 minutes at 37°C in a humidified incubator. Thereafter, cells were fixed with 4% PFA and counterstained with DAPI before imaging by fluorescence microscopy.

| Statistics
Statistics for experimental data are expressed as mean ± standard deviation (SD) of three independent experiments. Unpaired Student's t test was used for statistical analysis. A p ≤ 0.05 was considered statistically significant.

| Subcellular location of heparanase and silencing of HPSE expression in melanoma cells
The HPSE promoter in normal cells and tissues is constitutively silenced by methylation [22][23][24][25] and the action of p53 26 except in placenta, activated immune cells and keratinocytes, in which heparanase is constitutively active. However, HPSE expression can be induced in a number of inflammation related pathological processes by mediators such as tumour necrosis factor (TNF) or interleukin 1β (IL-1β). [27][28][29] In this study, we screened breast cancer T47D cells, melanoma MDA-MB-435s cells and MV3 cells for heparanase expression. We found that melanoma MDA-MB-435s cells and MV3 cells expressed a remarkably high level of endogenous heparanase ( Figure 1A).
To characterize the subcellular location of heparanase in the melanoma cells, we performed cell fractionation to separate cell lysates into cytosolic, nuclear and nuclear binding protein fractions. By Western blot analysis, we found the majority (roughly 90%) of heparanase to be present in the cytoplasm and detected the presence of approximately 10% of heparanase in the nucleus in MDA-MB-435s cells, whereas 30% of heparanase was detected in MV3 nuclei ( Figure 1B). For both cell lines we were able to extract the nuclear binding proteins and identified nuclear heparanase protein mostly located in the nuclear binding protein fraction. Taking advantage of the smartpool of siRNAs targeting HPSE mRNA, we were able to eliminate heparanase to a substantial degree in both MDA-MB-435s and MV3 cells after 48, 72 and 96 hours confirmed on protein level by Western blot analysis ( Figure 1C). Furthermore, subcellular fractionation analysis further confirmed elimination of heparanase in both cytoplasmic and nuclear compartments by silencing HPSE expression in MDA-MB-435s cells as shown in Figure 1D.

| Transcriptomic analysis identifies differentially expressed genes
For RNA-sequencing, triplicate RNA samples from cells transfected with control siRNA and HPSE siRNA were prepared. cDNA libraries were constructed following the workflow shown in Figure  To identify the differentially expressed genes that were associated with HPSE expression, all sequenced genes were screened between the cells transfected with control siRNA and HPSE siRNA to remove genes with low counts by defining false discovery rate (FDR) > 0. By doing this, 13 644 genes were identified as a reference gene list for functional enrichment analysis. We used a criterion that marks genes for which the fold change (FC) of HPSE silenced over control silenced is ≥2, FDR ≤0.001, as up-regulated, and those for which the ratio is ≤0.5, FDR ≤ 0.001, as down-regulated. Applied to the data from six samples, this yielded a list of 279 differentially expressed genes, of which 140 were up-regulated and 239 down-regulated, presented in an MA-plot ( Figure 2B; File S1).
To visualize gene expression data, the expression raw counts were log2 transformed corrected by library size, and the differentially expressed genes (|log2 FC| ≥ 1, FDR ≤ 0.001, n = 3) were extracted and displayed as a heat map ( Figure 2C). Based on the similarity of their gene expression patterns, the heat map clusters genes that show biological signatures associated to the regulation by heparanase.

| Functional analysis of differentially expressed genes reveals up-regulation of an array of proapoptotic genes
To identify the potential functions of the differentially expressed genes, we used the online gene analysis tool PANTHER to perform a gene ontology (GO) term analysis on the list of the differentially expressed genes. These genes were classified into the categories: molecular function, biological process, and cellular component 21 .
The functional enrichments of up-and down-regulated genes are presented separately ( Figure 3A,B). In addition, we listed the GO classification and the complete GO enrichment analysis of differentially expressed genes in the File S2. Genes were annotated in GO terms using the terminology provided by PANTHER, primarily inflammatory response, extracellular matrix, cell adhesion, positive regulation of cell death for up-regulated genes, and nucleosome and nucleosome assembly for down-regulated genes. To our surprise, we found a substantial set of genes that were up-regulated in HPSE silenced cells, which suggests that heparanase could act as a negative regulator of transcription; those genes classify in the GO term of positive regulation of cell death, apoptotic process, response to cytokine, response to external stimuli, response to stimuli ( Figure 3A).
Many studies have detailed the involvements of heparanase in acute and chronic inflammation by modification of the extracellular matrix or direct regulation of inflammatory cell function. 30 As expected, genes related to inflammatory response  Figure 3C).
To verify the pro-apoptotic genes regulated by heparanase, we performed real-time PCR on the 28 genes comparing HPSE silenced cells using smartpool siRNAs to control cells. The results validated that among other genes the expression of EGR1, CYR61 and TNFRSF12A was consistently up-regulated in HPSE silencing cells as shown in Figure 3D. In parallel, Western blot analysis further confirmed the up-regulation of those genes on protein level as shown in Figure 3E.

| Silencing of HPSE expression in melanoma cells induces caspase 3/PARP1-mediated apoptosis
Heparanase was shown to promote tumour cell proliferation, migration and evasion of apoptosis. Earlier studies have shown that     16 To consolidate our finding of increased apoptosis, the cells were subjected to fluorescent staining for cleaved caspase 3/7 after 72 hours of gene silencing. Increased staining of cleaved caspase 3/7 was exhibited in HPSE silenced cells, compared to control cells ( Figure 4C). Furthermore, Western blot analysis of the whole cell lysates using antibodies against caspase 3, cleaved caspase 3 and PARP1, revealed fragmentation of caspase 3 and PARP1 ( Figure 4D) occurring in HPSE silenced cells from day 2, but nearly undetectable in control cells, suggesting involvement of caspase 3 in the apoptosis induced by elimination of heparanase.
The apoptosis of MDA-MB-435s cells was carefully consolidated by silencing HSPE expression using a different single siRNA ( Figure   S1A,B), as well as involving the activation of caspase 3/PARP1 pathway ( Figure S1C). A similar pattern of up-regulation of EGR1, CYR61 and TNFRSF12A was concomitantly observed on protein level shown by Western blots ( Figure S1D).
To extend the apoptotic phenotype observed in MDA-MB-435s cells, another melanoma cell line, MV3 cells were included and evaluated in this study. Apoptotic cells detected by TUNEL staining were consistently induced following 72 hours gene silencing targeting HPSE using both single siRNA and smartpool of siRNAs ( Figure   S2A,B), via activation of caspase 3/PARP1 ( Figure S2C). In order to provide potential mechanistic insight through linkage of differential expression of pro-apoptotic genes to melanoma cell apoptosis, we examined the expression of 28 pro-apoptotic genes in MV3 cells after silencing HPSE using smartpool of siRNAs. By real-time PCR, we were able to replicate the up-regulation of EGR1, TXNIP, AXL, CYR61, LIMS2 and TNFRSF12A as indicated in MDA-MB-435s cells ( Figure S2D). Accordingly, increased protein levels of EGR1, CYR61 and TNFRSF12A were also observed following HPSE gene silencing in MV3 cells shown by Western blots ( Figure S2E). leading to structural modification that loosens ECM barriers and enables cell dissemination. 36,37 However, the function of heparanase is not limited to the extracellular surroundings, which has been studied extensively.

| D ISCUSS I ON
Heparanase can also interfere with gene transcription directly by binding to nuclear DNA 38 or indirectly by controlling histone H3 methylation patterns. 39,40 The overexpression of heparanase in melanoma cell lines prompted us to study its role on cancer cells per se. 38,41 Accordingly, melanoma MDA-MB-435s cells and MV3 cells expressed remarkably high levels of endogenous heparanase ( Figure 1A). In both cell lines, there was easily detectable heparanase located in the nucleus, of which the majority was associated with the fraction of nuclear binding proteins ( Figure 1B). Using siRNAs targeting HPSE expression, we could eliminate a considerable level of heparanase after 48 hours from the cells, as well as from the nuclear compartment ( Figure 1C,D). Recently, the ability of heparanase to drive exosome secretion and alter exosome composition, has been linked to tumour progression. 43 And the presence of heparanase in autophagosomes confers the cells more resistance to stress and chemotherapy associated with increased autophagy. 44 Furthermore, heparanase was able to cooperate with Ras to enhance number and size of induced breast cancer lesions, suggesting pro-tumorigenic properties. 45 We have suggested that heparanase may modulate tumour cell apoptosis via direct interference with apoptotic pathways, thus altering the susceptibility to apoptosis-inducing factors possibly during carcinogenesis and anti-cancer therapy, although the apoptosis-modulating role of heparanase has not been studied sufficiently.
In this study, siRNA mediated silencing of HPSE expression in also confirmed on protein level ( Figure 3E). Of note, a similar effect was previously observed upon adding heparanase to rat cardiomyocytes, where adding exogenous latent heparanase was reported to down-regulate pro-apoptotic genes such as TNF superfamily 10 and its receptor TNFRSF10B. By acting directly on the apoptotic receptor and ligand, heparanase was able to provide protection of the cells against high glucose and H 2 O 2 induced cell-death. 46 The elucidation of transcription regulation by heparanase revealed its biological implication in apoptosis, and an array of proapoptotic genes modified by heparanase. It should be noted that although some canonical pathways and factors that trigger apoptosis have already been identified, the specific mechanisms that

DATA S HARING
Data are available on request from the authors.

CO N FLI C T O F I NTE R E S T
The authors declare that they have no competing interests.

AUTH O R S' CO NTR I B UTI O N S
Tianyi Song designed the research study, performed the experiments, analysed the data and wrote the manuscript; Dorothe Spillmann supervised the project and contributed with discussions, revised the manuscript.