The methods for cell culture and transfection with the TIG1-myc-his expression vector were as described previously [21]. Human melanoma A2058 and A375 cells were obtained from the Bioresource Collection and Research Center (BCRC, Hsinchu, Taiwan). The authentication of the A2058 and A375 cell lines was conducted by the Genomics life science company (https://www.genomics.com.tw/tw) using STR profiling and comparison with the DSMZ database. The certificate of certification can be found in the supplementary material. Additionally, cell lines were regularly analyzed by staining with the DNA dye DAPI to exclude mycoplasma contamination.

Cell viability and cell death were evaluated as described previously [22]. Briefly, A2058 or A375 cells (2 $\times$ 10${}^4$) were seeded in triplicate in 24-well plates and incubated at 37 °C in a medium containing 10% fetal bovine serum for 24 h. After transfection with empty vector or with TIG1-myc-his expression vector, cells were cultured for an additional 24–48 h in medium containing 1% fetal bovine serum. Cell culture supernatants were collected and the release of lactate dehydrogenase (LDH) was measured to assess cell death. In addition, cells were incubated in medium containing water-soluble tetrazolium 1 (WST-1) reagent (Roche Diagnostics, Mannheim, Germany) for 4–6 h to measure mitochondrial activity as an indicator of cell viability.

After transfection with empty vector or with TIG1-myc-his expression vector for 24 h, RNA was extracted from A2058 cells using TRIzol reagent (Invitrogen, Carlsbad, CA, USA). RNA-seq was then performed and the data analyzed by Biotools Microbiome Research Center (https://www.toolsbiotech.com/product_service_detail.php?id=390&cateId=1459). Raw reads obtained after sequencing were filtered for quality using Trimmomatic to obtain clean reads. Clean reads were aligned to the reference sequence using HISAT2, and gene expression levels (raw read counts) were calculated using feature Counts. Expression levels were normalized using RLE/TMM/FPKM. Differential gene expression analysis was performed based on biological replicates (triplicates) using DESeq2, with genes filtered based on a $|$Fold-Change$|$ $>$2 and corrected p-value 0.05.

A melanoma tissue cDNA array (MERT301) was obtained from Origene Technologies. After transfection of A2058 or A375 cells with empty vector or TIG1-myc-his expression vector for 24 h, RNA was extracted using TRIzol reagent. Reverse transcription was performed using GScript RTase enzyme (GeneDireX, Inc, Taiwan) according to the manufacturer’s instructions. Primers for amplifying TIG1 were as described in the literature [2], while the primer sequences for amplification of other genes were as follows: Schlafen family member 11 (SLFN11): forward primer: 5${}^{\prime }$-GCTGGAGCTTGAGAGGTCAG-3${}^{\prime }$ and reverse primer: 5${}^{\prime }$-GTGTAGCCAGAGTTCCCACC-3${}^{\prime }$. DERL3: forward primer: 5${}^{\prime }$-GGCCGACTTCGTCTTCATGT-3${}^{\prime }$ and reverse primer: 5${}^{\prime }$- GTAGATATGGCCCACCGCAA-3${}^{\prime }$. Homocysteine-responsive endoplasmic reticulum-resident ubiquitin-like domain member 1 (HERPUD1): forward primer: 5${}^{\prime }$-ACAAGGTGGCCCTATTGTGG-3${}^{\prime }$ and reverse primer: 5${}^{\prime }$-AACCGGCCTCGGTCTAAATG-3${}^{\prime }$. Heat shock protein family A member 6 (HSPA6): forward primer: 5${}^{\prime }$-AGCAAGATGAAGGAGACGGC-3${}^{\prime }$ and reverse primer: 5${}^{\prime }$-ACACCAGCGTCAATGGAGAG-3${}^{\prime }$. DNA damage-inducible transcript 3 (DDIT3): forward primer: 5${}^{\prime }$-TCAGAGCTGGAACCTGAGGA-3${}^{\prime }$ and reverse primer: 5${}^{\prime }$-CTGCCATCTCTGCAGTTGGA-3${}^{\prime }$. DnaJ homolog subfamily B member 9 (DNAJB9): forward primer: 5${}^{\prime }$-GACCGAGATTAGGGTGCGTG-3${}^{\prime }$ and reverse primer: 5${}^{\prime }$-CGCTCTGATGCCGATTTTGG-3${}^{\prime }$. HSPA5: forward primer: 5${}^{\prime }$-GAGAACACGGTCTTTGACGC-3${}^{\prime }$ and reverse primer: 5${}^{\prime }$-CTTTGGTTGCTTGGCGTTGG-3${}^{\prime }$. Quantitative real-time polymerase chain reaction (PCR) conditions were as previously described [21]. The expression level of the relevant genes was normalized to that of $\beta$-actin.

After transfection of A2058 and A375 cells with empty vector or TIG1-myc-his expression vector, cells were lysed with Radioimmunoprecipitation assay (RIPA) buffer (a lab-made solution). Western blotting was performed as previously described [21]. Antibodies recognizing myc-epitope (TIG1) or $\beta$-actin were purchased from Sigma-Aldrich (Saint Louis, MO, USA). The HERPUD1 antibody was purchased from Cell Signaling Technology (Beverly, MA, USA), while Binding immunoglobulin protein (BIP) and DNA damage-inducible transcript 3 (DDIT3) antibodies were purchased from Abclonal Inc (Woburn, MA, USA).

After transfection of A2058 and A375 cells with empty vector or TIG1-myc-his expression vector, cells were lysed with RIPA buffer. Caspase-3 activity was measured using a caspase-3 substrate as described previously [23]. Briefly, cell lysates were incubated in a buffer containing 25 mM HEPES [pH 7.5], 0.1% CHAPS, 5% sucrose, 5 mM DTT, 2 mM EDTA, and 2 mM caspase-3 substrate (Ac-DEVD-pNA; Sigma). Absorbance at 405 nm was then measured using an Enzyme Immuno Assay (EIA) reader (Infinite F200, Tecan, Durham, NC, USA).

Statistical analysis was carried out using SPSS version 24 (IBM Corp., Chiago, IL, USA). Each experimental data point represents the analysis of at least 3 replicates and is presented as the mean $\pm$ standard deviation (SD). Statistical analysis was performed using one-way analysis of variance (ANOVA) with Dunnett’s post hoc test. Correlations between continuous variables were assessed using Pearson correlation. Locally weighted scatter plot smoothing (LOWESS) was employed to visualize the relationship between significantly correlated variables based on mean smoothing. A p-value 0.05 was considered statistically significant.

Previous studies have revealed that TIG1 expression is downregulated in skin cancer tissue [21]. Therefore, we further analyzed the impact of TIG1 expression on cell growth in melanoma cell lines. Results from Fig. 1A demonstrate a significant decrease in cell viability and induction of cell death when A2058 melanoma cells express TIG1 for 24–48 h compared to the control group. Furthermore, as shown in Fig. 1B, the activity of caspase-3 increased in cells that expressed TIG1. This suggests that the cell death induced by TIG1 is associated with its promotion of apoptosis. Similar results were observed in A375 cells, where TIG1 was shown to reduce cell viability and promote apoptotic cell death (Fig. 1C,D). These findings suggest that the expression of TIG1 in melanoma cells inhibits cell growth.

Fig. 1.

The effect of Tazarotene-induced gene 1 (TIG1) on cell viability and death of A2058 and A375 cells. A2058 (A) or A375 (C) cells were transfected with empty vector or TIG1-myc-his expression vector and expressed for 24–48 h. Cell viability and death were measured by water-soluble tetrazolium 1 (WST-1) and lactate dehydrogenase (LDH) release, respectively. Caspase-3 activity was measured using a colorimetric substrate following the transfection of A2058 (B) and A375 (D) cells with empty vector or with TIG1-myc-his expression vector for 24 h. *p 0.05 compared with the control group.

To further explore the mechanism by which TIG1 inhibits melanoma cell growth, RNA-seq was used to investigate the impact of TIG1 expression on gene expression in the A2058 cell line. Results from the Bland-Altman (MA) plot show a significant increase in the expression of TIG1 and SLFN11 compared to the control group (Fig. 2A). Additionally, the Volcano plot results revealed significant upregulation of 100 genes in TIG1-expressing cells, and downregulation of one gene (Fig. 2B). The top 50 genes showing the most significant changes in expression are displayed using a heatmap (Fig. 2C).

Fig. 2.

RNA-seq analysis of genes that were differentially regulated by TIG1 in A2058 cells. A2058 cells were transfected with empty vector or TIG1-myc-his expression vector and expressed for 24 h. RNA-seq was used to determine gene expression profiles, and the Bland-Altman (MA) (A) and Volcano (B) plots are shown. Red dots indicate significant differences between groups transfected with the empty vector or the TIG1-myc expression vector. Genes with a more than two-fold difference in expression ratio between cells transfected with empty vector or TIG1 are presented as a heatmap for the top 50 genes (C). RNA-seq, RNA-sequencing.

Analysis of the gene ontology (GO) enrichment plot indicated that most of the genes regulated by TIG1 are associated with the cellular ER stress response (Fig. 3A). Among the genes regulated by TIG1, 25 were found to be associated with ER stress response and unfolded protein response, including DERL3, HSPA5 (BIP), DDIT3 and HERPUD1 (Fig. 3B).

Fig. 3.

The category of gene ontology (GO) biological process affected by TIG1 and derived from the RNA-seq data. Following transfection of A2058 cells with TIG1 expression vector, significant gene expression changes were mapped to cellular pathways, as represented by GO enrichment dot plots (A). Genes associated with Endoplasmic Reticulum (ER) stress response and unfolded protein response are shown (B).

To validate the results of RNA-seq, mRNA from A2058 and A375 cells transfected with empty vector or TIG1 expression vector was analyzed for the expression of genes related to ER stress response using real-time PCR. As shown in Fig. 4A, the expression of genes including DERL3, SLFN11, HSPA6, DNAJB9, HERPUD1, DDIT3 and BIP was significantly upregulated after A2058 cells were transfected with the TIG1 expression vector for 24 h, as compared to the control group. Similar results were observed in A375 cells, with significant upregulation of these genes following transfection with the TIG1 expression vector (Fig. 4B). HERPUD1, BIP and DDIT3 in particular showed high expression levels and were therefore selected for further study. The expression of SLFN11 was significantly increased following transfection with the TIG1 expression vector compared to the control group, consistent with the fold-change observed in the RNA-seq analysis. However, SLFN11 was excluded from further investigation due to its low expression level.

Fig. 4.

The effect of TIG1 on the expression of ER stress response-related genes in A2058 and A375 cells. A2058 cells (A) and A375 cells (B) were transfected with empty vector or with the TIG1-myc-his expression vector and expressed for 24 h. Total RNA was extracted and relative levels of the indicated mRNAs were measured by quantitative real-time polymerase chain reaction (PCR) after normalization to the expression of $\beta$-actin. *p 0.05 compared with the control group.

In addition to validating gene expression levels, we further analyzed whether the protein expression of these candidate genes also corresponds to the changes in mRNA expression. As shown in Fig. 5A, the protein levels of HERPUD1, BIP and DDIT3 in TIG1-expressing A2058 cells were increased compared to the control group. Similar trends in protein upregulation were observed in TIG1-expressing A375 cells (Fig. 5B). Increased expression of proteins related to the ER stress response is associated with the activation of intracellular caspase cascades [24, 25, 26]. As shown in Fig. 1B, the increased activity of caspase-3 in cells expressing TIG1 may be related to the upregulation of ER stress-related proteins induced by TIG1.

Fig. 5.

The effect of TIG1 on the expression of ER stress response-related proteins in A2058 and A375 cells. A2058 cells (A,C) and A375 cells (B) were transfected with empty vector or TIG1-myc-his expression vector and expressed for 24 h (A,B) or 6–48 h (C). Cell lysates were collected and the protein expression levels of HERPUD1, BIP and DDIT3 were measured by immunoblotting. The bar graphs quantify the indicated protein, with each sample normalized to the level of $\beta$-actin protein, based on three replicate western blot data. *p 0.05 compared with the control group.

To clarify the relationship between TIG1 expression and ER stress-related proteins, cell extracts from different time points were collected after transfection with the TIG1 expression vector and the expression level of these proteins was evaluated. As shown in Fig. 5C, TIG1 protein expression began 6 h after transfection with TIG1 expression vector, whereas upregulation of HERPUD1, BIP and DDIT3 protein expression occurred 12 or 24 h after transfection. These observations indicate that ER stress-related proteins are indeed regulated by TIG1 protein.

We next investigated the relationship between TIG1 expression and the expression of ER stress-related genes in skin cancer tissue using a melanoma cDNA array. Results from this analysis showed a positive correlation between the expression of TIG1 and that of HERPUD1, BIP and DDIT3 in melanoma tissue (Fig. 6). We also conducted LOWESS assay for TIG1 and HERPUD1, DDIT3, BIP, respectively. Pearson correlation analysis was used to study the association between TIG1 and these genes. Overall, significantly positive correlations (p 0.05) were between TIG1 and these genes. According to the LOWESS curve, the curves for the HERPUD1 and BIP genes show an increase from turning point to point (Supplementary Fig. 1).

Fig. 6.

Correlation of TIG1 gene expression with HERPUD1, BIP and DDIT3 gene expression in melanoma tissues. Real-time PCR was used to determine the expression levels of TIG1, HERPUD1, BIP and DDIT3 in commercial melanoma cDNA tissue arrays (MERT301). Pearson correlation analysis was performed to correlate the expression of TIG1 with that of HERPUD1, BIP, and DDIT3.

We also investigated whether the commonly used chemical chaperone TUDCA could mitigate the upregulation of ER stress-related proteins induced by TIG1 [27, 28]. Treatment of A2058 cells with a high dose of TUDCA (500 µM) prevented the TIG1-induced upregulation of HERPUD1, BIP and DDIT3 (Fig. 7A). TUDCA exhibited a similar attenuating effect on the upregulation of ER stress-related proteins by TIG1 in A375 cells (Fig. 7B).

Fig. 7.

The effect of Tauroursodeoxycholic acid (TUDCA) on TIG1-induced ER stress response-related protein expression in A2058 and A375 cells. A2058 cells (A) and A375 cells (B) were transfected with empty vector or with TIG1-myc-his expression vector and treated with different doses of TUDCA for 24 h. Cell lysates were then collected and the protein levels of HERPUD1, BIP and DDIT3 were evaluated by immunoblotting (A,B). The bar graphs quantify the indicated protein, with each sample normalized to the level of $\beta$-actin protein, based on three replicate western blot data. *p 0.05 compared with the control group. ${}^{\mathrm{\#}}$p 0.05 compared with the TIG1 group.

In addition to its effect on ER stress-related proteins, we also investigated whether TUDCA can alter the impact of TIG1 on cell growth. As shown in Fig. 8A, high-dose TUDCA treatment of A2058 cells attenuated the decrease in cell viability and increase in cell death caused by TIG1. Furthermore, TUDCA inhibited the increase in caspase-3 activity induced by TIG1 expression (Fig. 8B). The above results indicate that increased caspase-3 activity caused by TIG1 is associated with the TIG1 induction of ER stress-related proteins. Similar attenuation was observed in TUDCA-treated A375 cells (Fig. 8C,D). Taken together, these results indicate that TIG1 decreases the viability of melanoma cells and increases cell apoptotic death. These effects appear related to the upregulation of ER stress response proteins by TIG1.

Fig. 8.

The effect of TUDCA on cell viability and death of A2058 and A375 cells induced by TIG1. A2058 (A,B) or A375 (C,D) cells were transfected with empty vector or with TIG1-myc-his expression vector and treated with different doses of TUDCA for 24 h (B,D) or 48 h (A,C). Cell viability and death were then measured by WST-1 and LDH release, respectively. Caspase-3 activity was measured using a colorimetric substrate (B,D). *p 0.05 compared with the control group.