This article is mentioned in:

Introduction

Ubiquitin-specific processing enzyme 22 (USP22) is a novel deubiquitinating enzyme that can cleave ubiquitin (Ub) from Ub-conjugated protein substrates (1). As the subunit of the human SAGA coactivator complex, USP22 is linked to the regulation of gene transcription by deubiquitinating histones H2A and H2B (2,3). In addition, USP22 deubiquitinates intracellular protein, including the shelterin protein telomeric repeat binding factor 1 (4), the histone deacetylase sirtuin 1 (5) and the far upstream element-binding protein 1 fructose-1,6-bisphosphatase 1 (4), and therefore performs an extensive physiological function. A murine study showed that USP22 also regulates embryonic stem cell differentiation (6). In humans, the USP22 gene is located on chromosome 17, consists of 14 exons, and is transcribed and produced broadly across various tissues (7). Of note, elevated levels of USP22 have been identified in numerous types of human cancer, including colorectal (8) lung (9) and breast cancer (10). USP22 has been indicated in tumorigenesis. Deletion of USP22 leads to the accumulation of cells in the G1 phase of the cell cycle (2). For these reasons, USP22 is a putative cancer stem cell marker. Reducing the rate of USP22 expression may be a suitable target for cancer therapy (11). However, the mechanisms that lead to USP22 transcriptional activation, particularly in the human tumor cells, remain unknown.

Previously, USP22 transcription was activated by mitogen stimulation or viral infection in normal T and B lymphocytes (12), suggesting the regulation of USP22 gene expression occurs mainly at the transcriptional level. However, the mechanism in which signal transduction pathways regulate USP22 transcription is unclear. It is well-known that activation of the mitogen-activated protein kinase (MAPK) pathways is the main downstream event in response to mitogen stimulation. Three activated subgroups of MAPKs: Extracellular signal-regulated kinases (ERKs), p38 MAPK and c-Jun N-terminal kinases (JNKs), regulate diverse cellular responses, including cell proliferation, differentiation, survival, the inflammatory response and even cell death (13).

In the present study, p38 MAPK was involved in the regulation of USP22 transcription, but ERKs and JNKs were not. The chemotherapeutic agent cisplatin suppressed the USP22 gene partly through p38 MAPK. These results provide novel insights on the molecular mechanisms underlying USP22 expression.

Materials and methods

Cell cultures

Human cervical carcinoma (HeLa) cells were obtained from the Type Culture Collection of the Chinese Academy of Sciences (Shanghai, China) and were cultured in Dulbecco's modified Eagle's medium from Invitrogen (Carlsbad, CA, USA) supplemented with 10% fetal bovine serum at 37°C in a 5% CO2 incubator. Cisplatin and MAPK inhibitors U0126, SB203580 and SP600125 were from Sigma-Aldrich (St. Louis, MO, USA).

Plasmid constructs

The USP22 promoter fragments inserted into pGL-3 were constructed as described previously (14). Site-directed mutagenesis was carried out within the USP22 basic promoter p-210/+52 according to the manufacturer's instructions for the MutanBEST kit (Takara Bio Inc., Otsu, Shiga, Japan). Mutagenic primer pairs used for the polymerase chain reaction (PCR) amplification included 5′-GTAGCGTAATCTCCGTCCGC-3′ for the CREB/ATF-binding site mutagenesis, 5′-CCTGTAGGCTCTGGGTAGAC-3′ for the MYB-binding site mutagenesis, 5′-GGATCGGTGCCTGCCTTGCA-3′ for the Sp1-binding site (−7/−12) mutagenesis (complementary reverse primers are not shown, and mutated nucleotides are underlined). All the mutations were confirmed by DNA sequencing. The MAPK kinase 6 (MKK6) expression plasmid and dominant negative MKK6 (DN MKK6) plasmid were provided by Professor Jiahuai Han of Xiamen University (Xiamen, Fujian, China).

Transfections and dual luciferase reporter assay

Cells (1×104) were plated in 24-well plates 12 h before transfection with 0.5 µg of various USP22 promoter constructs and 0.1 µg pRL-TK (Promega Corp., Madison, WI, USA) using Lipofectamine 2000 (Invitrogen) in each well. Twenty-four hours after transfection, cells were was hed in phosphate-buffered saline and lysed for 30 min at room temperature using the passive lysis buffer (Promega Corp.). Luciferase activity was determined using the dual luciferase reporter assay system (Promega Corp.). The normalized luciferase activity was expressed as the ratio of firefly luciferase activity to Renilla luciferase for each sample. All the transfection experiments were repeated four times.

Total RNA isolation and quantitative PCR (qPCR)

Total RNA from cells treated with agents was prepared using TRIzol according to the manufacturer's instructions (Invitrogen). RNA was reverse-transcribed with oligo-dT primers using an RNA PCR kit (AMV) ver. 3.0 (Takara Bio Inc.), and the cDNA fragments were analyzed by qPCR using the SYBR-Green PCR Master mix (Toyobo Co., Ltd., Osaka, Japan) on an ABI 7500 real-time PCR System (Applied Biosystems, Foster City, CA, USA). USP22 primer pairs were: Forward, 5′-ACCACCACGCTCACGGACTG-3′; and reverse, 5′-TTGGCTGAGTGTTCAAATCG-3′; P21 primer pairs were: Forward, 5′-GCAGATCCACAGCGATATCC-3′; and reverse, 5′-CAACTGCTCACTGTCCACGG-3′. GAPDH primers were: Forward, 5′-AGAAGGCTGGGGCTCATTTG-3′; and reverse, 5′-AGGGGCCATCCACAGTCTTC-3′.

Western blot analysis

Cells were lysed with 1X SDS sample buffer. Protein was separated by 10% SDS-PAGE and electroblotted onto nitrocellulose membrane. The membrane was subsequently incubated with anti-USP22 antibody or anti-GAPDH antibody (Santa Cruz Biotechnology, Inc., Dallas, TX, USA) and developed using the electrochemiluminescence system (Pierce Biotechnology, Inc., Rockford, IL, USA), according to the manufacturer's instructions.

Flow cytometry

Cells (1×105) were seeded in 6-well plates overnight and subsequently SB203580 or dimethyl sulfoxide (DMSO) was added. After 12 h, the medium was changed for medium containing cisplatin for a further 12 or 24 h incubation. The cells were harvested by trypsinization, was hed with 1X PBS and resuspended in 50 µl of 1X PBS. Cell death was assessed using flow cytometry (BD Biosciences, Franklin Lakes, NJ, USA) following staining with fluorescein isothiocyanate-Annexin V and propidium iodide.

Statistical analysis

Numerical data are presented as the mean ± standard error of the mean. Statistical differences between sample means were determined using the unpaired, two-tailed Student's t-test. The significance level was set at α<0.05, and therefore, P<0.05 was considered to indicate a statistically significant difference.

Results

p38 MAPK regulates USP22 promoter activity

Pharmacological reagents that inhibit MAPK pathways with different specificity were used to determine whether MAPK signaling is involved in the regulation of USP22 expression: U0126, SB203580 and SP600125. These selectively inhibit ERK1/2, p38 MAPK and JNK, respectively. HeLa cells were transfected with the reporter construct, pGL-210/+52, which contains the USP22 basic promoter region from −210 to +52 (14). The transfected cells were cultured for a further 12 h in the presence of 2 µM U0126, 2 µM SB203580, 10 µM SP600125 and 0.1% DMSO as blank controls, and luciferase activity was measured. As shown in Fig. 1A, incubation of 2 µM SB203580 resulted in an ~150% induction of luciferase activity relative to DMSO (P<0.05). However, treatment of SP600125 and U0126 did not affect the luciferase activity in the tested cells (P>0.05).

Figure 1. p38 MAPK regulated USP22 promoter activity. (A) Treatment with 2 µM U0126, 2 µM SB203580, 10 µM SP600125 or vehicle DMSO (−) was added to HeLa cells transfected with 0.5 µg USP22 promoter construct pGL-210/+52 and 0.1 µg pRL-TK. The cells were harvested and assayed for luciferase. The normalized relative luciferase activity for pGL-210/+52 treated with DMSO was set at 100. The p38 inhibitor SB203580 markedly enhanced (*P<0.05 compared to DMSO treated) pGL3–210/+52 activity. (B) HeLa cells were transiently transfected with reporter vectors containing series of 5′ terminal deletion constructs of the USP22 promoter and treated with 2 µM SB203580. After 12 h, cells were harvested and assayed for luciferase. The normalized relative luciferase activity for pGL-210/+52 treated with DMSO was set at 100. (C) HeLa cells were transfected with pGL-210/+52 and incubated with SB203580 for different periods. Cells were harvested and assayed for luciferase. The normalized relative luciferase activity for pGL-210/+52 treated with DMSO for 0 h was set at 100. (D) Effects of MKK6 and DN MKK6 on pGL-210/+52 reporter construct. Cells were transiently transfected with 0.5 µg of reporter construct plus 1.0 µg of pCDNA3.1 (empty vector), or MKK6 or DN MKK6. After 24 h, the transfected cells were analysed for luciferase activity. The normalized relative luciferase activity obtained in cells transfected with pGL-210/+52 and pCDNA3.1 was set at 100. MAPK, mitogen-activated protein kinase; USP22, ubiquitin-specific processing enzyme 22; HeLa, human cervical carcinoma; MKK6, MAPK kinase 6; DN MKK6, dominant negative MKK6.

To further confirm the induction effects and localize the region in the promoter responsible for this enhancing effect by SB203580, reporter vectors containing a series of 5′ terminal deletion constructs of the USP22 promoter were transfected into HeLa cells and treated with 2 µM SB203580 for 12 h. Luciferase activity showed that SB203580 enhancing USP22 promoter activity was observed with constructs pGL-2828/+52, pGL-886/+52 and pGL-210/+52, but not with construct pGL-7/+52 (Fig. 1B), suggesting a certain degree of promotion by SB203580 may be mediated through response elements located within the −210/−7 domain. Subsequently, pGL-210/+52 was transfected and incubated with SB203580 for different periods. As shown in Fig. 1C, SB203580 induced USP22 promoter activity but this was not an early event. Significant increases in promoter activity (~150%) occurred after 12 h incubation.

MKK6 phosphorylates and activates p38 MAPK (15). To confirm whether MKK6 regulates USP22 expression, DN MKK6 and pGL-210/+52 were transiently co-transfected into HeLa cells. A luciferase assay showed that expression of DN MKK6 increased transcription of USP22 by 150%, which was similar to the action of SB203580. By contrast, forced expression of a constitutively active MKK6 had a significant inhibitory effect on USP22 promoter activity (Fig. 1D). All these results demonstrated that p38 MAPK is involved in the regulation of USP22 promoter activity.

p38 MAPK regulates endogenous USP22 expression

As the p38 MAPK pathway is involved in regulating USP22 promoter activity, the effects of p38 MAPK on endogenous USP22 expression were assessed. As shown in Fig. 2A, treatment of HeLa cells with 2 µM SB203580 for 12 h significantly increased USP22 mRNA expression by ~160% as examined using qPCR. Consistent with the PCR results, USP22 protein levels were also increased by SB203580 incubation (Fig. 2B).

Figure 2. p38 MAPK regulated endogenous USP22 expression. (A) SB203580 and USP22 expression. USP22 transcriptional levels in response to 2 µM SB203580 as examined via RT-qPCR. Data are normalized to GAPDH and are representative of independent triplicate experiments. (B) USP22 protein expression levels in response to SB203580 examined via western blot analysis. (C) Dominant negative MKK6 or MKK6 on USP22 mRNA expression. (D) p21 transcriptional levels in response to 2 µM SB203580 examined via RT-qPCR. Data are normalized to GAPDH and are representative of independent triplicate experiments. (E) Induction of USP22 by SB203580 at the transcriptional level. HeLa cells were treated with SB203580 (2 µM), actinomycin D (10 µg/ml) or SB203580 plus ActD for different time periods. USP22 mRNA expression was determined by RT-qPCR and normalized to GAPDH expression. USP22, ubiquitin-specific processing enzyme 22; RT-qPCR, reverse transcription-quantitative polymerase chain reaction; HeLa, human cervical carcinoma; MKK6, MAPK kinase 6; DN MKK6, dominant negative MKK6.

MKK6 and DN MKK6 were transfected into HeLa cells and USP22 mRNA expression was quantitated. As shown in Fig. 2C, forced expression of MKK6 decreased endogenous USP22 mRNA and DN MKK6 increased the expression.

p21, a cyclin-dependent kinase inhibitor, is repressed by USP22 (4). For this reason, levels of p21 mRNA were measured in response to SB203580. As shown in Fig. 2D, SB203580 treatment for 12 h decreased p21 mRNA expression in HeLa cells.

To further determine whether p38 MAPK influences USP22 mRNA stability in HeLa cells, actinomycin D was used to block de novo mRNA transcription. The level of mRNA was determined at different points in time using qPCR. When cells were treated with actinomycin D, the half-life of USP22 mRNA was between 4 and 5 h. When SB203580 was added 12 h before actinomycin D treatment, the half-life of USP22 mRNA did not change significantly and was similar to that in the cells without SB203580 treatment (Fig. 2E), excluding the possibility that p38 MAPK influences USP22 mRNA stability.

p38 MAPK regulates USP22 via the Sp1 binding site

As mentioned above, the region responsible for enhancing USP22 promoter activity using SB203580 is located within the −210/−7 domain. TFSEARCH analysis showed there to be several potential transcription factor binding sites within this region, including an MYB, a CREB/ATF and a Sp1 binding site. To characterize the involvement of particular cis elements in response to SB203580, wild-type and mutant reporter gene constructs of the USP22 promoters were generated. After plasmids were transfected into HeLa cells, SB203580 was incubated for another 12 h. As shown in Fig. 3, mutation of the MYB and CREB/ATF binding sites did not disrupt the enhancing effect of USP22 promoters by SB203580. However, mutation of the Sp1 binding site (−7/−13) resulted in increased basal USP22 promoter activity and abrogated the enhancing promoter activity treated by SB203580 (Fig. 3), suggesting SB203580 elicits promotion of USP22 transcription through the Sp1 binding site.

Figure 3. Effect of mutated binding sites on the reporter assay. HeLa cells were transiently transfected with the c-MYB mutation, CREB/ATF mutation, Sp1 mutation derived from pGL-210/+52 and wild-type pGL-210/+52 for 12 h. The cells were subsequently treated with 2 µM SB203580. After 12 h incubation, cells were analyzed for luciferase activity. The normalized relative luciferase activity for pGL-210/+52 treated with DMSO was set at 100. SB203580 (2 µM) did not enhance promoter activity of Sp1 mutation (#P>0.05). HeLa, human cervical carcinoma; DMSO, dimethyl sulfoxide.

Cisplatin represses USP22 expression through p38 MAPK

A previous study has shown that p38 MAPK can be activated by several anticancer reagents, such as cisplatin (16). Whether cisplatin can suppress USP22 expression was explored. HeLa cells were treated with 30 nM cisplatin for 0, 6, 12 and 24 h and USP22 expression at the mRNA level was analyzed using qPCR. As shown in Fig. 4A, cisplatin can induce an extremely significant decrease in USP22 mRNA expression in HeLa cells. Significant changes can be observed after 6 h of treatment (52% inhibition).

Figure 4. Effect of cisplatin on USP22 expression. (A) USP22 mRNA levels were measured by RT-qPCR in HeLa cells incubated for the indicated time with 30 nM cisplatin. The normalized USP22 mRNA without cisplatin treatment was set at 100. (B) USP22 promoter activity in response to cisplatin and SB203580 examined via the luciferase assay. (C) USP22 transcriptional levels in response to cisplatin and SB203580 examined via RT-qPCR. USP22, ubiquitin-specific processing enzyme 22; RT-qPCR, reverse transcription-quantitative polymerase chain reaction.

Subsequently, HeLa cells were co-treated with cisplatin and SB203580 for 12 h. As shown in Fig. 4B, cisplatin decreased USP22 promoter activity and this suppression was mostly restored in HeLa cells by treatment with 2 µM SB203580. Similarly, SB203580 partially restored the cisplatin-induced decrease in USP22 mRNA (Fig. 4C). These results indicate that cisplatin can suppress USP22 expression at the transcriptional level and this suppression is possibly mediated by p38 MAPK.

SB203580 did not protect HeLa cells from cisplatin-induced apoptosis

As overexpression of USP22 may be a factor for therapeutic resistance, the role of p38 MAPK/USP22 signaling was determined in the survival of cancer cells exposed to cisplatin. HeLa cells were treated with cisplatin and SB203580 in two ways and were subsequently analyzed by flow cytometry. One group of cells was pretreated with SB203580 for 12 h and subsequently cisplatin was added and allowed to incubate for another 12 or 24 h. The other was co-treated with SB203580 and cisplatin for 12 or 24 h. As shown in Fig. 5A, pretreatment or co-treatment of HeLa cells with SB203580 did not result in less cisplatin-induced apoptosis as determined by Annexin V staining. These results suggest that SB203580-induced promotion of USP22 expression did not protect HeLa cells against cisplatin-induced apoptosis.

Figure 5. Effect of blocking p38 MAPK on cisplatin-induced apoptosis. HeLa cells were co-treated with SB203580 and cisplatin for 24 or 48 h or pretreated with SB203580 for 12 h and were subsequently treated with cisplatin for 24 or 48 h, and stained with PI and Annexin V. The relative numbers of apoptotic cells are indicated. MAPK, mitogen-activated protein kinase; HeLa, human cervical carcinoma; PI, propidium iodide.

Discussion

Considering the crucial role of USP22 in carcinogenesis and tumor progression, p38 MAPK regulation of USP22 may have important implications. Studying signaling pathways that regulate USP22 expression may lead to identification of new therapeutic targets for cancer therapy. The present study reported that p38 MAPK acts upstream of USP22 and plays a negative role in USP22 transcription. p38 MAPK was found to regulate USP22 transcription via a Sp1 binding site. In addition, cisplatin suppressed USP22 expression in part through p38 MAPK.

One of the findings of the present study is that USP22 transcription is controlled by p38 MAPK. Pharmacological inhibitors were used and results showed that the inhibitor of p38 MAPK enhanced USP22 promoter activity under culture conditions but ERKs and JNKs did not. The further multiple experimental results leading to this conclusion are as follows: i) SB203580 promoted endogenous USP22 mRNA expression, ii) transfection of the p38 activator MKK6 repressed USP22 promoter activity and endogenous mRNA expression, and iii) transfection of the p38 inhibitor DN MKK6 enhanced USP22 promoter activity and endogenous mRNA expression. All these results indicated that p38 MAPK acts as the upstream regulator in regulation of USP22 transcription. The p38 MAPK pathway is an important regulator of numerous cellular responses and the target genes regulated by activated p38 MAPK are complex. Initially, cytokine genes, such as tumor necrosis factor-α and interleukin-10 (IL-10), have been described as target genes promoted by p38 MAPK in T cells and monocytes following stimulation (17,18). Recently, p38 MAPK has received increasing attention as a tumor suppressor and its activation suppresses oncogenesis in cancer cells with different patterns (19). For example, BMI-1, one of 11 death signature genes as USP22, is reported to be downregulated by p38 MAPK by posttranscription modification (7). Cyclin D1, the oncogene, is also negatively regulated by p38 MAPK at the transcription level (20). In the present study, the possibility that p38 MAPK affects the stability of USP22 mRNA was excluded and p38 MAPK was confirmed to regulate USP22 at the transcriptional level. Results showed that in response to SB203580 treatment, expression of USP22 and the key regulator of cell cycle, p21, exhibited the inverse trend. A previous study has demonstrated that p38 MAPK participates in p21 upregulation, subsequently inhibiting cell growth (21). It is not fully understood how activated p38 MAPK can upregulate p21 expression. One study showed that p38 MAPK stabilizes p21 mRNA (22). Another study reported that p38 MAPK can enhance p21 expression by promoting the transcriptional elongation (23). USP22 acts upstream of p21 and represses p21 expression (4) and the present results indicate that expression of the p38 MAPK promoter p21 may take place partially through USP22 downregulation.

p38 MAPK regulation of gene expression is mediated by the activation of a wide range of protein kinases, transcription factors and other proteins. The present study showed that p38 MAPK-induced repression of USP22 is mediated by a Sp1 binding site. Bioinformatic analysis showed that the 5′ flank region of USP22 is a typical feature of the TATA-less promoter and several DNA motifs are within the basic USP22 promoter, including an Sp1, a CREB/ATF and a c-MYB binding site. Mutation of these sites has been confirmed to lead to USP22 promoter activity changes (data not shown). The present study showed only the Sp1 binding site (GGGCGG) ahead of the transcription start site (−7 to −13) to be responsible for SB203580 treatment, as mutations at this site enhance USP22 promoter activity and abolish the effects of SB203580 on USP22 promoter activity. Previous studies have shown that the GGGCGG sequence is involved in regulation of USP22 expression and is specifically bound by Sp1. Despite its history, transcription factor Sp1 is activated by phosphorylation in response to p38 MAPK. For example, it has been reported that p38 MAPK regulation of IL-10 promoters activity is via Sp1. The mechanisms by which p38 MAPK activates Sp1 to regulate USP22 transcription will be discussed in a future study.

p38 MAPK activation is also necessary for the suppression of cancer cell growth as initiated by a variety of anticancer agents, including cisplatin (24). In the present study, cisplatin was used to activate p38 MAPK in HeLa cells. Treatment of these cells with cisplatin suppressed production of USP22 mRNA, which was mostly accounted for by decreased promoter activity. Cisplatin exerts its cytotoxic properties by DNA strand-cross links, therefore activating downstream signaling cascades, which ultimately induce apoptosis. In the present study, cisplatin was found to repress USP22 expression via the p38 MAPK pathway; SB203580 treatment antagonized cisplatin to repress USP22 expression, and the decrease of the USP22 promoter by cisplatin was also abolished by disruption of the Sp1 binding site. It has been proved that p38 MAPK is the universal sensor for cisplatin presence and the activation of p38 MAPK is required for apoptosis. Considering the p38 MAPK downregulation of USP22, it is not difficult to understand that cisplatin treatment suppresses USP22 expression.

USP22 is considered as the putative cancer stem cell marker and its overexpression has been associated with therapy resistance. Therefore, it appears that enhanced USP22 expression by SB203580 may exert anti-apoptotic effects in theory. However, previous studies showed that the role of SB203580 on cisplatin-induced apoptosis in tumor cells remains controversial. Numerous studies showed that the p38 MAPK inhibitor, SB203580, protected cells from cisplatin-induced apoptosis. However, several studies showed that SB203580 did not reverse cell apoptosis but sensitized cells to apoptosis. In the present study, pretreatment or co-treatment of SB203580 was not observed to protect HeLa cells from cisplatin-induced apoptosis, suggesting that although SB203580 enhances USP22 expression, this is not sufficient to antagonize the cisplatin-induced apoptosis.

In conclusion, the present study provides evidence that p38 MAPK regulates USP22 promoter activity and its expression in a human cancer cell line, and cisplatin represses USP22 expression partly through the p38 MAPK pathway. These findings provided new insights on the molecular mechanisms underlying USP22 expression and may have indications for developing novel therapeutic strategies.

Acknowledgements

The present study was supported by the National Nature Science Foundation of China (grant nos. 31000581 and 81460172).

References