Article Text

Download PDFPDF

Original article
IRTKS is correlated with progression and survival time of patients with gastric cancer
  1. Li-Yu Huang1,2,
  2. Xuefei Wang3,
  3. Xiao-Fang Cui2,
  4. He Li3,
  5. Junjie Zhao3,
  6. Chong-Chao Wu2,
  7. Lingqiang Min3,
  8. Zhicheng Zhou4,
  9. Lixin Wan5,
  10. Yu-Ping Wang6,
  11. Chao Zhang7,8,
  12. Wei-Qiang Gao4,
  13. Yihong Sun3,
  14. Ze-Guang Han1,2
  1. 1 Key Laboratory of Systems Biomedicine (Ministry of Education) and Shanghai-MOST Key Laboratory for Disease and Health Genomics, Chinese National Human Genome Center at Shanghai, Shanghai Center for Systems Biomedicine, Shanghai Jiao Tong University, Shanghai, China
  2. 2 Key Laboratory of Systems Biomedicine (Ministry of Education) and Collaborative Innovation Center of Systems Biomedicine, Shanghai Center for Systems Biomedicine, Shanghai Jiao Tong University, Shanghai, China
  3. 3 Department of General Surgery, Zhongshan Hospital, General Surgery Research Institute, Fudan University, Shanghai, China
  4. 4 State Key Laboratory of Oncogenes and Related Genes, School of Biomedical Engineering and Med-X Research Institute, Shanghai Jiao Tong University, Shanghai, China
  5. 5 Department of Molecular Oncology, H. Lee Moffitt Cancer Center and Research Institute, Tampa, USA
  6. 6 Department of General Surgery, The First Hospital of Lanzhou University, Lanzhou, China
  7. 7 Institute for Computational Biomedicine, Weill Cornell Medical College of Cornell University, New York, USA
  8. 8 Department of Medicine, Division of Hematology and Medical Oncology, Weill Cornell Medical College of Cornell University, New York, USA
  1. Correspondence to Yihong Sun; sun.yihong{at}zs-hospital.sh.cn and Dr. Ze-Guang Han, Shanghai Center for Systems Biomedicine, Shanghai Jiao Tong University, China. ; hanzg{at}sjtu.edu.cn

Footnotes

  • L-YH, XW and X-FC contributed equally.

  • Contributors ZGH and LYH designed the experiments; LYH performed most of the animal and biochemical experiments; YS, XW, HL, JZ and LM collected and analysed patient samples; XFC, CCW and YPW performed some biochemical experiments; CZ analysed the biomedical information; LW, ZZ and WQG assisted in the analysis of some experiments. ZGH and LYH interpreted the data and wrote the manuscript.

  • Funding The study is supported from the National Natural ScienceFoundation of China (81402317, 81472621, 81672772, 81272271,81672324 and 31670806), and Science and Technology Commission ofShanghai Municipality (14ZR1429700 and 14140902500).

  • Competing interests None declared.

  • Patient consent Obtained.

  • Ethics approval Zhongshan Hospital.

  • Provenance and peer review Not commissioned; externally peer reviewed.

View Full Text

Statistics from Altmetric.com

Significance of this study

What is already known on this subject?

  • It has been reported that IRTKS could promote the cell proliferation of hepatocellular carcinoma cell lines via EGFR/ERK signalling.

  • During the in vitro ubiquitin assay, IRTKS promoted MDM2-mediated p53 ubiquitination and degradation in HT1080 and U2OS cells. IRTKS might be a negative regulator of p53.

What are the new findings?

  • IRTKS is overexpressed in gastric cancer samples and correlated with the survival time of patients with gastric cancer.

  • By using mouse model, we demonstrate that IRTKS could negatively regulate the tumour suppressor function of p53 in vivo. IRTKS deficiency could delay the tumour occurrence in p53+/−  mice and extend the survival time of p53+/−  mice.

  • Ser331 site of IRTKS protein might be a key point for p53 dissociation from MDM2 under DNA damage conditions.

How might it impact on clinical practice in the foreseeable future?

  • IRTKS, as a negative regulator of p53, provides a promising prognostic factor for patients with gastric cancer.

Introduction

Gastric cancer is one of the most deadly malignancy and the third most common leading cause of global cancer mortality.1 2 This heterogeneous disease arises from the complex interplay of host factors and environmental risk agents like Helicobacter pylori, smoke, dietary habits and so on. Moreover, there are emerging evidences showing that gastric cancer exhibits a wide spectrum of molecular aberrations.3 Copy number amplifications may alter gene expression levels and occur frequently in gastric cancer.4 By analysing the public The Cancer Genome Atlas (TCGA) database, we found that genomic DNA of IRTKS is frequently amplified in human cancers related to the digestive system, especially in gastric cancer.

In addition, in our previous work, we found that IRTKS may function as a novel negative regulator of p53 activity through promotion of MDM2-mediated p53 ubiquitination and degradation in HT1080 and U2OS cells.5 IRTKS is also frequently upregulated in human hepatocellular carcinoma (HCC) and promotes HCC cell proliferation.6 However, whether the contribution of IRTKS overexpression to the progression of human cancers is through regulation of p53 ubiquitination and degradation is not clear. Moreover, whether IRTKS overexpression that negatively regulates the p53 level is correlated with poor outcome in these patients with tumour should be further clarified.

Materials and methods

Patient samples

In total, 527 pairs of gastric cancer samples in a tissue array, surgical remnants provided by the Department of General Surgery, Zhongshan Hospital, were used with approval from each patient. H&E staining of each paraffin section and subsequent pathological diagnosis contributed to the classification of clinical samples. p53 antibody (DO-1, Santa Cruz Biotechnology, Santa Cruz, California, USA) at a 1:50 dilution was used to interact with the dewaxed paraffin sections of the 527 pairs of gastric cancer samples.

Mice

IRTKS-knockout mice were previously generated and were housed in Shanghai Model Organisms Center.7 Heterozygous IRTKS mice were mated with heterozygous p53 mice (Jackson Laboratory) to generate IRTKS +/− p53 +/− mice. Then, the IRTKS +/− p53 +/− mice were mated mutually to generate groups of mice with different genotypes. Nude mice and NOD-SCID mice were purchased from Shanghai Laboratory Animal Center, Chinese Academy of Sciences. Mice were fed with a regular diet and had free access to water and food. All the mice experiments were approved by the Animal Ethics Committee of the Shanghai Model Organisms Center.

Mouse embryonic fibroblasts

Mouse embryonic fibroblasts (MEFs) were isolated from E13.5 mouse embryos.

Plasmids

IRTKS S331A and S331D mutants were generated using a QuikChange XL Site-Directed Mutagenesis Kit (Stratagene, La Jolla, California, USA), according to the manufacturer’s instructions.

Antibodies

Anti-IRTKS rabbit polyclonal antibody was generated by our laboratory. Anti-p21, anti-p53 (DO-1), anti-Ub, anti-Chk2 and anti-actin antibodies were from Santa Cruz Biotechnology (Santa Cruz, California, USA). Anti-p53 (#2524), anti-cleaved caspase 3 and anti-BAX were obtained from Cell Signalling Technology (Beverly, Massachusetts, USA). Anti-MDM2 and anti-phosphorylated Ser/Thr antibodies were from Abcam (Cambridge, UK). Chk2 kinase (#7434) was from Cell Signalling Technology (Beverly, Massachusetts, USA).

Immunoprecipitation assay

Briefly, 1 mg of cell lysates was incubated with the indicated antibodies at 4°C overnight and then incubated with protein G sepharose beads at 4°C for another 2 hours. Each immunoprecipitate was washed four times with lysis buffer (50 mM Tris/HCl, pH 7.5, 150 mM NaCl, 1 mM EDTA, 0.3 mM DTT, 0.1% NP-40 and protease inhibitor cocktail (Sigma-Aldrich, St. Louis, Missouri, USA)) and then analysed by immunoblotting.

siRNA synthesis

Two siRNAs against IRTKS were used: siRNA-1 (sense, 5′-CCA GUC CCU UGA UCG AUA UTT-3′ and anti-sense, 5′-AUA UCG AUC AAG GGA CUG GTA-3′); siRNA-2 (sense, 5′-GCU UAA GCA AAU CAU GCU UTT-3′ and anti-sense, 5′-AAG CAU GAU UUG CUU AAG CAG-3′). Two siRNAs against Chk2 were used: siRNA-Chk2-1 (sense, 5′-GCA CGA GCU CUC ACA GUA UTT-3′ and anti-sense, 5′-AUA CUG UGA GAG CUC GUG CTT-3′); siRNA-Chk2-2 (sense, 5′-GCU UAU UGG GAA AGG CAA ATT-3′ and anti-sense, 5′-UUU GCC UUU CCC AAU AAG CTT-3′). All siRNAs were designed and chemically synthesised by Shanghai GenePharma.

In vitro kinase assay

Briefly, 1 µg of the indicated GST–IRTKS fusion proteins was incubated with 50 ng of commercially obtained recombinant active Chk2 kinase (Cell Signalling Technology, #7434), in the presence of 200 µM cold ATP in the kinase reaction buffer at 30°C for 30 min. The reaction was stopped by the addition of SDS-containing lysis buffer and resolved by SDS–PAGE. The phosphorylation of GST–IRTKS was detected by anti-phosphorylated Ser/Thr antibody.

MS analysis

For MS analysis, anti-Flag immunoprecipitations (IPs) were performed with the whole cell lysates derived from three 10-cm dishes of HEK293 cells co-transfected with Flag-IRTKS and HA-Chk2. The IP proteins were resolved by SDS–PAGE and were identified by Coomassie staining. The band containing IRTKS was in-gel digested with trypsin enzymes. The resulting peptides were extracted from the gel and were resuspended with 15 µL solvent A, respectively (A: water with 0.1% formic acid; B: ACN with 0.1% formic acid), separated by nano-LC and analysed by online electrospray MS/MS. The experiments were performed on an EASY-nLC 1000 system (Thermo Fisher Scientific, Waltham, Massachusetts, USA) connected to an Orbitrap Fusion mass spectrometer (Thermo Fisher Scientific, San Jose, California, USA) equipped with an online nano-electrospray ion source. MS/MS data were searched against the UniProt/SwissProt database (Taxonomy: Homo sapiens, 20 199 entries) using Mascot (Matrix Science, London, UK; V.2.3) and data analysis was performed using Proteome Discoverer software (Thermo Fisher Scientific, V.1.4.0.288). Peptides and modified peptides were accepted if they passed a 1% FDR threshold.

Protein half-life detection

MEF cells were plated in 6-well plates at 80% confluence and cultured in Dulbecco’s modified Eagle’s medium (Invitrogen, Carlsbad, California, USA) with 10% fetal bovine serum. After 24 hours of culture, cycloheximide (CHX, Sigma Chemical, St Louis, Missouri, USA) (100 µg/mL) was added. Cells were lysed in SDS buffer containing protease and phosphatase inhibitor (Sigma Chemical, St Louis, Missouri, USA) after CHX treatment at the indicated time points.

Cell cycle analysis

MEF cells were harvested and incubated with propidium iodide (Sigma) (50 µg/mL), RNase A (10 mg/mL) and Triton X-100 (0.2%) for 15 min at room temperature and then analysed by a FACS Calibur flow cytometer, CellQuest (BD Biosciences, USA).

Statistical analysis

Statistical evaluation of in vitro and in vivo experiments was calculated using Student’s t-test. Multiple group comparisons were analysed by one-way analysis of variance. Kaplan-Meier survival curves and log-rank (Mantel-Cox) tests were used for survival analysis of patients with gastric cancer and the genetically engineered mice with IRTKS and p53 deficiency. Different cut-off values, p<0.05 (*), p<0.01(**) and p<0.001 (***), were considered significant.

Results

IRTKS is upregulated in human gastric cancer and correlates with poor survival in patients with wild-type p53

To explore the role of IRTKS in human tumours, we analysed TCGA database and demonstrated that the genomic DNA of IRTKS is frequently amplified in human cancers related to the digestive system, including oesophageal and gastric cancers, where genomic DNA amplification of IRTKS was found in 7.4% of patients with gastric cancer (see online supplementary figure 1). We therefore examined IRTKS expression in 527 pairs of human gastric cancer samples in a tissue array. Quantitative immunohistochemistry (IHC) data showed that IRTKS expression was significantly higher in 398/527 (75.5%) tumour specimens than in adjacent non-cancerous tissues (p<0.0001) (figure 1A and  online supplementary table). Higher IRTKS expression was positively correlated with progression, particularly in late stages, of these gastric cancer samples (p=0.029) (figure 1B and online supplementary figure 2).

Supplementary Material

Supplementary Table 1
Figure 1

IRTKS is upregulated in human gastric cancer and correlates with poor survival in patients with wild-type p53. (A) IRTKS expression level in 527 paired gastric tumour samples and normal tissues. The expression scoring system was based on multiplication of the IRTKS immunohistochemistry staining intensity and extent. Staining intensity was classified as 0 (negative), 1 (weak), 2 (moderate) and 3 (strong). Staining extent was dependent on the percentage of positive cells (examined in 200 cells) and was divided as follows: 0 (<5%), 1 (5%–25%), 2 (26%–50%), 3 (51%–75%) and 4 (>75%). (B) IRTKS expression was positively correlated with late stages of gastric tumour samples. (C) Kaplan-Meier survival curves for 527 individuals with gastric cancer whose tumours showed negative (IRTKS−, black) or positive IRTKS (IRTKS+, red) staining. Of 527 samples, the individuals with wild-type p53 (n=206) were defined as negative nuclear staining.8–12 High IRTKS (IRTKS+) was defined as cytoplasmic staining scores ≥4, according to the Youden index analysis (online supplementary figure S2A).

We next investigated the relationship between IRTKS expression and the clinical outcome of patients with gastric cancer with different p53 status. Here, the status of p53 was characterised by IHC nuclear staining. Generally, it has been recognised that positive p53 IHC staining is largely ascribed to mutant p53, while the expression of wild-type p53 often becomes undetectable in gastric cancers bearing wild-type p53.8–12 Based on the above criteria, among the 527 tumour samples, 321 cases were defined as bearing mutant p53 and the remaining 206 individuals were defined as carrying wild-type p53. Remarkably, higher IRTKS expression (n=141 vs n=65) was significantly associated with poor outcome of patients with cancer carrying wild-type p53 (log-rank test: p=0.0153) (figure 1C), although there were no significant differences in the whole cohort (log-rank test: p=0.4983). While conversely in these patients with p53 mutant, better survival is observed in these patients with higher IRTKS expression although the p value was not significant (log-rank test: p=0.1773) (figure 1C).

We then performed an independent Western blotting analysis on randomly selected tumour specimens to evaluate the association of the IRTKS level with wild-type or mutant p53. Significantly, the result showed that the IRTKS level was reversely correlated with wild-type p53 expression (see online supplementary figure 3A and B), although no association between IRTKS and mutant p53 level was observed (see online supplementary figure 3C and D). These data suggested that IRTKS overexpression could exert an important role in tumorigenesis of gastric cancer by negatively regulating wild-type p53 activity.

IRTKS promotes cell proliferation in gastric cancer through downregulation of p53

To further validate the hypothesis that IRTKS may regulate wild-type p53 but not mutant p53, we transfected three types of plasmids encoding wild-type p53, as well as mutant p53H179C and p53R248W, the hotspot p53 mutants in cancer,13 14 into p53−/− mouse embryonic fibroblasts (MEFs). Significantly, ectopic IRTKS could induce a more intense decrease in wild-type p53 (36% remaining) than in mutant p53H179C and p53R248W (approximately 89% and 70% remaining) (figure 2A and B). Consistently, as shown in online supplementary figure 4, these p53 mutants p53H179C and p53R248W significantly attenuated the ability to bind IRTKS, as compared with that of wild-type p53.

Figure 2

IRTKS promotes cell proliferation in gastric cancer by negatively regulating p53. (A) p53 /− mouse embryonic fibroblast cells were transfected with the indicated IRTKS and p53 constructs. After 24 hours, the cell lysates were analysed by Western blotting. (B) The remaining amount of exogenous p53 was quantified by densitometry and normalised to the level of actin. (C) IRTKS and p53 expression levels were assessed by Western blotting. sh-IRTKS and sh-p53 were used to knock IRTKS and p53 down in AGS cells, respectively. (D) The colony formation of AGS cells treated with NC, sh-IRTKS, sh-p53 and combined sh-p53 and sh-IRTKS. (E) Quantification of the colonies shown in (D). (F) Results of Western blotting showing the expression levels of IRTKS and p53 in MKN45 cells treated with EV, IRTKS, p53 and combined p53 and IRTKS. (G) The colony formation of MKN45 cells treated as described in (F). (H) Quantification of the colonies shown in (G). Data are presented as the mean ± SD *p<0.05, **p<0.01 versus control. (I) The stable AGS cell lines were generated after transfection with NC, sh-IRTKS, sh-p53 or sh-p53 plus sh-IRTKS and were then injected subcutaneously into 6-week-old nude mice. A total of 5×106 cells were injected for each group. After 90 days of observation, tumour growth curves were constructed. (J) At 90 days post injection, the tumours were collected and weighed.

To explore the underlying mechanism by which IRTKS negatively regulates p53, we first examined the biological function of IRTKS in the gastric cancer cell lines AGS and MKN45, both of which express wild-type p53. As shown in  online supplementary figure S5A and B, when IRTKS was knocked down in AGS cells, whose IRTKS expression was relatively higher, the cell growth and colony formation of AGS cells were significantly suppressed. In contrast, the forced overexpression of IRTKS could strikingly enhance cell proliferation and colony formation of MKN45 cells (see online supplementary figure 5C and D). Interestingly, we found that, as IRTKS was knocked down in AGS cells, the endogenous wild-type p53 level was significantly increased (figure 2C). Furthermore, the suppressed colony formation mediated by IRTKS knockdown could be partially reversed by p53 knockdown induced by shRNA lentivirus infection (figure 2D and E), and these results were supported by the change in the p53 level after both IRTKS and p53 were knocked down (figure 2C). Meanwhile, when IRTKS was forcibly overexpressed in MKN45 cells, the endogenous or ectopic wild-type p53 level was significantly reduced (figure 2F), and the increased colony formation capacity induced by IRTKS overexpression was attenuated by ectopic p53 (figure 2G and H). Furthermore, in vivo tumorigenesis in nude mice also indicated that IRTKS knockdown significantly inhibited xenograft tumour growth of AGS cells. However, shRNA-induced p53 knockdown could abrogate the IRTKS knockdown-reduced tumour growth (figure 2I and J and online supplementary figure 6).

On the contrary, when IRTKS was knocked down or enforcedly overexpressed in the cell line BGC823 with p53 mutant, the cell growth and colony formation of BGC823 cells were not significantly changed (see online supplementary figure 5E-H).

These collective data suggested that IRTKS overexpression enhanced cell proliferation of gastric tumours possibly through influencing the wild-type p53 expression level.

IRTKS enhances mouse tumorigenesis in a p53-dependent manner

To understand the impact of IRTKS on the role of p53 in tumour suppression in vivo, we further crossed IRTKS-knockout mice7 with genetically inactivated p53 mice to generate various mouse genotypes. Among them, we monitored and observed six groups: p53+/+IRTKS+/+ , p53+/+IRTKS /− , p53+/−IRTKS+/+ , p53+/−IRTKS /− , p53 /−IRTKS+/+ and p53 /−IRTKS /− . All p53+/+ mice with or without IRTKS developed normally and had no difference in lifetime during the 24-month observation period (figure 3A). All p53 /−  mice with or without IRTKS also had no difference in tumour-free survival time and died around 6 months of age from lymphoma or sarcoma (figure 3B, and online supplementary figure 7A and B), indicating that IRTKS deficiency could not effectively block tumorigenesis and mortality in p53 /−  mice. However, we found that p53+/−  mice lacking IRTKS (p53+/−IRTKS /− ) displayed an increased median tumour-free survival (log-rank test: p=0.007) (figure 3A). The average lifetime of the p53+/−IRTKS /−  mice (n=11) was 20 months, significantly longer than that of p53+/−IRTKS+/+ mice (n=10), whose average lifetime was 16 months.15 Similar to p53 /−  mice, most of the p53+/−  mice died of lymphoma or sarcoma (see online supplementary figure 7A and B), suggesting that IRTKS may contribute to tumorigenesis in the case of p53 haploinsufficiency.

Figure 3

IRTKS deficiency inhibits tumorigenesis in a p53-dependent manner. (A) Tumour-free Kaplan-Meier survival curves of p53+/−IRTKS+/+ (n=10), p53+/−IRTKS /−  (n=11), p53+/+IRTKS+/+ (n=15) and p53+/+IRTKS /−  (n=15) mouse cohorts (log-rank Mantel-Haenszel test). (B) Tumour-free Kaplan-Meier survival curves of p53 /−IRTKS+/+ (n=13) and p53 /−IRTKS /−  (n=11) mouse cohorts (log-rank Mantel-Haenszel test). (C) Western blot analysis of p53 and IRTKS in the indicated mouse embryonic fibroblast (MEF) cells. (D-G) The indicated H-rasv12-transduced MEFs (1×107) were subcutaneously injected into NOD-SCID mice. The mice were sacrificed 3 weeks later and the tumours were weighed.

To confirm this observation, we further evaluated the in vivo tumorigenicity of MEFs derived from these six groups of mice (figure 3C). The p53 expression level in MEFs derived from p53+/+IRTKS /−  and p53+/−IRTKS /−  mice was significantly increased as compared with that from p53+/+IRTKS+/+ and p53+/−IRTKS+/+ mice, respectively, where the elevated p53 expression level in p53+/−  MEF cells with IRTKS depletion was comparable to that in p53+/+IRTKS+/+ MEFs (figure 3C). The MEF cells from these mice were transduced with a plasmid encoding the activated oncogene H-rasv12 and then were subcutaneously injected into immunocompromised NOD-SCID mice.16 As expected, the H-rasv12-transduced p53+/−IRTKS+/+ MEFs produced tumours in mice (figure 3D and E), whereas the p53+/−IRTKS /−  MEFs did not form visible tumours 3 weeks after injection. In agreement with experiments in the genetically engineered mice, both p53 /−IRTKS+/+ and p53 /−IRTKS /−  MEFs with ectopic H-rasv12 developed xenograft tumours in NOD-SCID mice, with no significant difference between the two groups (figure 3F and G). Meanwhile, all p53+/+ MEFs with H-rasv12 did not form tumours regardless of whether IRTKS was present (data not shown), implying that genomic instability caused by p53 haploinsufficiency may be a prerequisite for the contribution of IRTKS in tumorigenesis.

Together, the observation that the inhibitory effect of IRTKS deficiency on tumourigenicity exclusively occurred in p53-haploinsufficient mice suggested that the contribution of IRTKS overexpression to human tumours may depend on negative regulation of wild-type p53 function.

IRTKS deficiency increases cell cycle arrest and apoptosis through activation of p53 activity

To further evaluate the effect of IRTKS on p53 activity, we examined specific cellular behaviours, including cell cycle and apoptosis, in the p53-haploinsufficient MEFs or in mice. In untreated MEF cells at the same passage, we found that the cell growth of p53+/−IRTKS /−  MEFs was remarkably slower than that of p53+/−IRTKS+/+ MEFs (used as a control) (figure 4A). This result was further supported by the notably increased G2/M fraction of cell cycle progression in p53+/−IRTKS /−  MEFs (figure 4B and C). Moreover, we also found that cell apoptosis was significantly higher in MEFs and in spleen and thymus tissues of p53+/−IRTKS /−  mice than in the 2-month-old p53+/−IRTKS+/+ littermates, as shown by annexin V labelling, TUNEL assay and IHC staining of cleaved caspase 3 (figure 4D and E, online supplementary figure 8A,B). In the same experimental conditions, cell cycle distribution and apoptosis failed to show any statistical difference between p53 /−IRTKS+/+ and p53 /−IRTKS /−  MEFs or mice (see online supplementary figure 8A,B and see online supplementary figure 9A, B). These collective data from p53-haploinsufficient MEFs and mice suggested that IRTKS deficiency enhanced p53 function in cell cycle arrest and apoptosis.

Figure 4

IRTKS deficiency promotes cell cycle arrest and apoptosis of heterozygous p53 mouse embryonic fibroblasts (MEFs). (A) Cell growth rate analysis of p53+/−IRTKS+/+ , p53+/−IRTKS /− , p53 /−IRTKS+/+ and p53 /−IRTKS /−  MEFs at the same passage (P4). (B) Cell cycle analysis of p53+/−IRTKS+/+ and p53+/−IRTKS /−  MEFs at the same passage (P4). (C) Histogram columns representing the means of the cell cycle distribution shown in (B). (D) Eight-week-old p53+/−IRTKS+/+ and p53+/−IRTKS /−  mice were either untreated or exposed to 5 Gy of γ-irradiation, and then, the thymi from these mice were obtained 4 hours later. Single-cell suspensions of these thymocytes were prepared and stained with annexin V-FITC for FACS analysis. (E) The quantification of annexin V-positive cells in (D). (F) Western blot analyses of p53, p21 and BAX in the indicated MEF cells, where p53+/+IRTKS+/+ MEF cells were used as positive control. (G) Real-time PCR analysis of p53, BAX and p21 gene expression in p53+/−IRTKS+/+ and p53+/−IRTKS /−  MEFs. (H) Cell growth rate analysis of p53+/−IRTKS /−  and p53 /−IRTKS /−  MEFs transfected with either control or pCMV-Flag-IRTKS vectors. (I) Immunoblot assays of p53, p21 and BAX in p53+/−IRTKS /−  and p53 /−IRTKS /−  MEFs transfected with either control or pCMV-Flag-IRTKS vectors. (J) Real-time PCR analysis of p53, BAX and p21 gene expression in p53+/−IRTKS /−  MEFs transfected with empty vector or pCMV-Flag-IRTKS.

To confirm this hypothesis, we evaluated the p53 downstream pathway in the p53-haploinsufficient MEFs and mice. In agreement with the above cellular phenotypes, in the p53+/−IRTKS /−  group, the levels of p53 protein, but not mRNA, and the downstream target genes BAX and p21 were significantly higher than those in the p53+/−IRTKS+/+ group (figure 4F and G and online supplementary figure 8C). These data revealed that, in the absence of IRTKS, the p53 downstream signalling pathway was activated in p53-haploinsufficient cells.

To further evaluate the effect of IRTKS on p53 function, we performed a rescue experiment by transfecting a recombinant plasmid encoding IRTKS into p53+/−IRTKS /−  MEFs to restore IRTKS expression. Interestingly, compared with the empty vector control, ectopic expression of IRTKS significantly increased cell growth of p53+/−IRTKS /−  MEFs and decreased the p53 protein level, but not mRNA expression, as well as the levels of the downstream molecules BAX and p21 (figure 4H-J). However, in the absence of p53, ectopic IRTKS had no effect on cell proliferation or on BAX and p21 levels (figure 4H and I). The data indicated that IRTKS may regulate p53 activity by affecting the p53 expression level.

Furthermore, we examined p53+/+ cells and showed that cell apoptosis in p53+/+IRTKS /−  MEFs, as represented by the proportion of annexin V-positive cells, was significantly increased compared with p53+/+IRTKS+/+ MEFs when exposed to ultraviolet (UV) irradiation (see online supplementary figure 9C and D), supporting the notion that IRTKS functions as a negative regulator of p53 activity.

IRTKS promotes MDM2-mediated p53 ubiquitination and degradation in gastric cancer cells and MEFs

To determine the molecular mechanism by which IRTKS regulates the p53 level, we first evaluated p53 protein stability in IRTKS /−  MEFs. The data showed that the half-life of p53 in these MEFs lacking endogenous IRTKS (more than 60 min) was significantly longer than in cells expressing wild-type IRTKS (~15 min) (figure 5A and B), whereas the half-life of Mdm2 in these MEF cells without IRTKS was not obviously changed (figure 5A and B). We also found that binding of MDM2 to p53 was attenuated in IRTKS /−  MEFs, as shown by a reciprocal co-IP assay with anti-p53 and MDM2 antibodies (figure 5C and D). As a result, ubiquitination of p53 was decreased in the IRTKS-deficient MEFs (figure 5E). In contrast, when ectopic IRTKS was introduced into IRTKS-knockout MEFs, the interaction between MDM2 and p53 was clearly enhanced, and p53 polyubiquitination was increased, although the level of p53 was markedly decreased (figure 5F and G).

Figure 5

IRTKS promotes p53 ubiquitination and degradation in mouse embryonic fibroblast (MEF) cells and gastric cancer cells. (A) Same passage IRTKS+/+ and IRTKS /−  MEFs were treated with 100 µg/mL cycloheximide (CHX) and harvested at the indicated time points. Cell lysates were submitted to Western blotting. (B) The amount of p53 (left) and Mdm2 (right) was quantified by densitometry and normalised to the level of actin. (C and D) Co-immunoprecipitation (IP) experiments in IRTKS+/+ and IRTKS /−  MEFs with anti-p53 and anti-MDM2 antibodies, respectively. (E) Ubiquitination analysis of p53 in IRTKS+/+ and IRTKS /−  MEFs. Cells were incubated with MG132 (10 µM) for 9 hours before being harvested. (F) Co-IP experiments in IRTKS /−  MEFs transfected with empty vector or pCMV-Flag-IRTKS were performed using anti-p53 antibody. (G) Ubiquitination analysis of p53 in IRTKS /−  MEFs transfected with empty vector or pCMV-Flag-IRTKS. Cells were incubated with MG132 (10 µM) for 9 hours after 24 hours of transfection. (H) Co-IP experiments in AGS cells transfected with si-NC or si-IRTKS were performed using anti-p53 (DO-1) antibody. Mouse IgG was used as a control. (I) Ubiquitination analysis of p53 in AGS cells transfected with si-NC or si-IRTKS. (J) Co-IP experiments in MKN45 cells transfected with empty vector or pCMV-Flag-IRTKS with anti-p53 (DO-1) antibody. Mouse IgG was used as a control. (K) Ubiquitination analysis of p53 in MKN45 cells transfected with empty vector or pCMV-Flag-IRTKS.

In addition, we also examined the molecular interaction of IRTKS on Mdm2-mediated p53 ubiquitination in human gastric cancer cells. Expectedly, as with IRTKS knockdown in AGS cells, the association of MDM2 with p53 and p53 polyubiquitination, as shown by co-IP with anti-p53 antibody, was markedly decreased, although the p53 level was increased (figure 5H and I). By contrast, when IRTKS was forcibly overexpressed in MKN45 cells, the binding of MDM2 to p53 was significantly promoted and accompanied by increased polyubiquitination and degradation of p53 (figure 5J and K).

Meanwhile, we also examined the expression level of IRTKS in AGS and MKN45 cells after Mdm2 depletion, and we observed no significant change in both cell lines (see online supplementary figure 10A and B), which indicated that IRTKS stability might not be controlled by Mdm2.

These collective data revealed that IRTKS may form a tertiary complex with both Mdm2 and p53, and IRTKS overexpression promotes MDM2-mediated p53 ubiquitination and degradation in gastric cancer cells and MEFs, suggesting a new cancer-promoting mechanism role of IRTKS in tumorigenesis.

Ser331 phosphorylation of IRTKS is required for dissociation of p53 from the IRTKS/MDM2 complex on DNA damage

IRTKS is a member of the IRSp53/MIM (missing-in-metastasis) homology domain family and known to be tyrosine phosphorylated in response to insulin and EGF stimulation.6 7 17 However, when DNA is damaged, multiple proteins, including p53, undergo serine/threonine (Ser/Thr) phosphorylation by Ser/Thr protein kinase Chk1 or Chk2.18 Therefore, we first wanted to define whether the Ser/Thr phosphorylation status of IRTKS is involved in the MDM2-mediated p53 regulation under DNA damage conditions. Interestingly, we found that IRTKS was clearly Ser/Thr phosphorylated in MEFs after treatment with the genotoxic agent doxorubicin (DOX) or UV exposure (figure 6A and online supplementary figure 11A), along with an increased p53 level and phosphorylation of Chk2. Furthermore, an IP assay with anti-IRTKS antibody showed that Chk2 can be immunoprecipitated by endogenous IRTKS in MEFs, regardless of whether DNA damage is present (figure 6A and online supplementary figure 11A), suggesting that IRTKS could be a substrate of Chk2 kinase via Ser/Thr phosphorylation.

Figure 6

Ser331 is a key point for phosphorylation of IRTKS and dissociation from p53 on stress. (A) Co-immunoprecipitation (IP) experiments in IRTKS+/+ and IRTKS /−  mouse embryonic fibroblasts (MEFs) either left untreated or treated with 1.0 mM doxorubicin (DOX) for 4 hours were conducted with anti-IRTKS antibody. (B) Co-IP experiments in MEFs with wild-type IRTKS after Chk2 knockdown. (C) Chk1/2 substrate consensus analysis in IRTKS protein sequences across different species. (D) In the in vitro kinase assay, the indicated GST–IRTKS fusion proteins (1 µg each) were incubated with commercially obtained recombinant active Chk2 kinase (50 ng) at 30°C for 30 min. The phosphorylation of GST–IRTKS was detected by anti-phosphorylated Ser/Thr antibody. (E) Co-IP experiments in IRTKS /−  MEFs transfected with the indicated IRTKS constructs and left either treated or untreated with 1.0 mM DOX for 4 hours. (F) Ubiquitination analysis of p53 in IRTKS /−  MEFs transfected with the indicated IRTKS constructs. These cells were incubated with MG132 (10 µM) for 9 hours after 24 hours of transfection and then either treated or untreated with 1.0 mM DOX for 4 hours before harvesting. (G) The schematic model depicts IRTKS serine phosphorylation at the Ser331 site for p53 dissociation from the IRTKS/MDM2 complex in DNA damage conditions and the role of IRTKS overexpression in tumorigenesis.

To confirm this hypothesis, we knocked down Chk2 with RNAi and then examined the Ser/Thr phosphorylation status of IRTKS following DOX addition. Interestingly, with Chk2 knockdown, the Ser/Thr phosphorylation of IRTKS was significantly decreased in MEFs triggered by DOX (figure 6B), which indicated that Chk2 was an upstream Ser/Thr kinase of IRTKS.

Notably, we also found that the interaction between p53 and IRTKS was markedly diminished in MEFs after UV or DOX exposure, although the p53 level was significantly upregulated, whereas the interaction of MDM2 with IRTKS remained (figure 6A and online supplementary figure 11A). Therefore, we subsequently need to answer whether the Ser/Thr phosphorylation of IRTKS is responsible for dissociation of p53 from the IRTKS/MDM2 complex, which leads to upregulation of p53 due to the absence of MDM2-mediated ubiquitination. To define the Ser/Thr phosphorylation sites required for dissociation of p53 from the IRTKS/MDM2 complex, we analysed the consensus motif of the Chk1/2 substrate in IRTKS protein sequences from human, mouse and rat. Interestingly, we found that Ser331 of IRTKS is only one candidate site for Chk1/2 activity19–21 (figure 6C). Consistently, the MS analysis confirmed the existence of in vivo phosphorylation of Ser331 site within IRTKS (see online supplementary figure 11B). We then constructed IRTKS mutant that displaced Ser331 with alanine for in vitro kinase assay, and we found that the IRTKS phosphorylation triggered by active Chk2 kinase was significantly decreased on the IRTKSS331A mutant, as compared with that on wild-type IRTKS (figure 6D). Moreover, when wild-type IRTKS or the mutants were transfected into IRTKS /−  MEFs, the interaction between p53 and IRTKS was significantly attenuated in MEFs with ectopic IRTKSWT, but not IRTKSS331A, after DOX addition (figure 6E); by contrast, the p53 level was significantly reduced along with the markedly enhanced ubiquitination of p53 in MEFs expressing IRTKSS331A under DOX treatment (figure 6E and F). We further examined the interaction of IRTKS and p53 in gastric cancer MKN45 cells via a co-IP assay after transfecting the cells with these IRTKS vectors. The data showed that disruption of Ser331 phosphorylation of IRTKS facilitated the binding of IRTKS to p53 (see online supplementary figure 11C), whereas ubiquitination of p53 was obviously enhanced by IRTKSS331A (see online supplementary figure 11D).

To strengthen the hypothesis, we further constructed an IRTKSS331D mutant that displaced Ser331 with aspartic acid, which may function as a phosphorylation mimic. The result revealed that IRTKSS331D exhibited a somewhat weaker interaction with p53 than wild-type IRTKS (see online supplementary figure 11C), along with attenuated ubiquitination of p53 in MEFs expressing IRTKSS331D (see online supplementary figure 11D).

The collective data suggested that Ser331 phosphorylation of IRTKS is required for dissociation of p53 from the IRTKS/MDM2 complex on DNA damage.

Discussion

In this work, we uncovered that, when DNA damage is triggered by UV radiation or DOX treatment, IRTKS is phosphorylated at its Ser/Thr residues, among which Ser331 phosphorylation of IRTKS is required for p53 dissociation from the IRTKS/MDM2 complex (figure 6). Moreover, we confirmed that Chk2 acts as an upstream kinase required for IRTKS serine phosphorylation in response to DNA damage. In normal conditions, IRTKS may bind to p53 and MDM2 to balance MDM2-mediated p53 polyubiquitination and proteasome-dependent degradation, maintaining the stability of the p53 level for cell homoeostasis and genomic integrity. However, DNA damage induced Chk2 kinase-mediated IRTKS serine phosphorylation that prevents binding of IRTKS to p53 and simultaneously promotes p53 phosphorylation that blocks binding of p53 to MDM2, which, finally, attenuates p53 polyubiquitination and proteasome-dependent degradation, leading to p53 upregulation to initiate DNA repair or apoptosis (figure 6G). These suggest another new regulatory mechanism for Chk2-dependent p53 stabilisation by IRTKS phosphorylation.18 Also, it has been known that 14-3-3 can typically bind to the serine/threonine-phosphorylated residues at an RSxpSxP mode22; therefore, we speculate that the Chk2-mediated phosphorylation of IRTKS at Ser331 site (RSVpSVA) might be recognised by 14-3-3 molecule, which then blocks the interaction of p53 with IRTKS. The speculation is worthy of being further investigated.

In tumorigenesis, IRTKS overexpression due to DNA copy amplification or an epigenetic event may enhance the IRTKS/MDM2/p53 axis, resulting in p53 degradation, which could promote progression of gastric cancer. In this study, the survival data revealed that patients with higher IRTKS expression, together with wild-type p53, showed very poor clinical outcomes (figure 1C). In contrast, IRTKS deficiency reduced tumorigenesis and prolonged the survival time of p53 heterozygous mice (figure 3A). These data suggested that disruption of the IRTKS/MDM2/p53 axis due to genetic and epigenetic events may contribute to the pathogenesis of tumours by enhancing p53 degradation. Our findings provide new insight for understanding the molecular mechanism by which IRTKS overexpression plays an important role in pathogenesis of human gastric tumours.

Supplementary Material

Supplementary Figure 1

Acknowledgments

We gratefully thank Ping-Zhao Zhang (School of Life Sciences Fudan University) for providing p53 mutant vectors and Naoe Nihira (Harvard Medical School) for her critical reading of the manuscript.

References

View Abstract

Footnotes

  • L-YH, XW and X-FC contributed equally.

  • Contributors ZGH and LYH designed the experiments; LYH performed most of the animal and biochemical experiments; YS, XW, HL, JZ and LM collected and analysed patient samples; XFC, CCW and YPW performed some biochemical experiments; CZ analysed the biomedical information; LW, ZZ and WQG assisted in the analysis of some experiments. ZGH and LYH interpreted the data and wrote the manuscript.

  • Funding The study is supported from the National Natural ScienceFoundation of China (81402317, 81472621, 81672772, 81272271,81672324 and 31670806), and Science and Technology Commission ofShanghai Municipality (14ZR1429700 and 14140902500).

  • Competing interests None declared.

  • Patient consent Obtained.

  • Ethics approval Zhongshan Hospital.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.