E2F8-induced GRPEL2 promoted colorectal cancer progression via targeting TIGAR

Abstract

Background

Colorectal cancer (CRC) is the leading cause for cancer mortality across the world. GRPEL2 is a critical regulator of mitochondria’s function with an oncogenic role in different kinds of cancer. The exact function of GRPEL2 -driven mitochondrial regulation and CRC progression have not been elucidated.

Methods

RNA-seq data from TCGA database was analyzed to identify biomarkers and therapeutic targets of CRC. The gene expression profile was validated by quantitative real-time PCR on 68 paired tumor and non-tumor samples from CRC patients. Tumorigenesis regulated by GRPEL2 was tested through EdU staining, Transwell assay, in vivo tumor growth and in vivo metastasis. The function of Mitochondria mediated by GRPEL2 was evaluated by transmission electron microscopy, DCFH-DA staining, mitochondrial membrane potential detection, and Calcein staining. LC–MS/MS screening and Co-IP were performed to discover protein partners of GRPEL2. E2F8-mediated transcriptional regulation of GRPEL2 was verified via Luciferase reporter and ChIP assays.

Results

GRPEL2 was upregulated in CRC tumor tissues and cell lines. High expression of GRPEL2 was associated with poor prognosis of CRC and inhibition of GRPEL2 suppressed CRC proliferation and migration by inducing mitochondria injury. Meanwhile, TIGAR was shown to interact with GRPEL2 and overexpression of TIGAR rescued CRC progression in the presence of GRPEL2 inhibition. Moreover, E2F8 was the upstream regulator of GRPEL2, which positively induced GRPEL2 transcription and expression in CRC.

Conclusion

Our work illustrated the oncogenic role of GRPEL2 in CRC development and determined the molecular mechanism of E2F8/GRPEL2/TIGAR pathway. These findings will provide novel insights and promising therapeutic targets for CRC treatment in the future.

Highlights

• 1. GRPEL2 was upregulated in CRC.

• 2. GRPEL2 depletion suppressed proliferation and migration of CRC.

• 3. GRPEL2 knockdown induced mitochondrial structural and functional impairment.

• 4. GRPEL2 regulated mitochondrial function via TIGAR.

• 5. E2F8 promoted GRPEL2 transcription and expression in CRC.

Introduction

Colorectal cancer (CRC) is the leading cause for cancer mortality across the world, particularly high incidence rates in developed countries [1]. Despite of the rapid progress in the diagnosis and therapeutic treatments for CRC [2], the pathogenesis mechanism of CRC is still complicated. Consequently, efforts for the exploration of CRC development are needed to identify promising therapeutic approaches.

Functional mitochondria play a crucial role in tumor cell growth by modulating the reactive oxygen species (ROS) level and activating the apoptotic pathway [3]. The chaperon protein mtHsp70 (mitochondrial Hsp70) is critical for the transport of nuclear-encoded polypeptides into the mitochondrial matrix [4]. Once activated, the ATP bound to mtHsp70 is hydrolyzed to ADP and the nucleotide exchange factor (NEF) replaces ADP with ATP, leading to the transport of polypeptides into the mitochondrial matrix [5]. GRPEL2 is one of the NEFs with well characterized role in regulating mitochondrial function [5], while its contribution to CRC development has not been illustrated. The TP53-induced glycolysis and apoptosis regulator (TIGAR) is a target of p53 pathway that reduces fructose-2,6-bisphosphate (F-2,6-2P) levels and increases NADPH via the pentose phosphate pathway (PPP) [6]. TIGAR has diverse functions in maintaining cellular homeostasis [7] and is overexpressed in different tumors [8,9,10,11]. Considering the overlapping between GRPEL2 and TIGAR in regulating mitochondrial function, it’s worth investigating the potential interaction and mechanism for them in CRC development.

E2F8 belongs to the E2F transcription factor family that acts as both transcriptional activator [12] and transcriptional repressor [13]. Numerous studies have demonstrated that E2F transcription factors are crucial in the modulation of cell proliferation, differentiation, and apoptosis [14, 15]. In CRC, E2F8 was upregulated in cancer tissues and cell lines, controlling the level of cell cycle genes [16]. Inhibition of E2F8 suppressed cell proliferation of CRC cells by affecting the NF-KB pathway [17]. E2F8 was also considered a stemness gene marker,being overexpressed upon celecoxib treatment in HT29 and DLD1 cells [18]. However, whether E2F8 could regulate GRPEL2 expression in CRC remains unclear.

In this study, we found GRPEL2 was significantly upregulated in CRC tumor tissues and cell lines. Upregulation of GRPEL2 was related to poor prognosis of CRC and inhibition of GRPEL2 was shown to suppress CRC proliferation and migration through inducing mitochondria injury. Meanwhile, TIGAR was demonstrated to interact with GRPEL2 and overexpression of TIGAR could rescue CRC progression in the presence of GRPEL2 inhibition. Moreover, E2F8 was identified as the upstream regulator of GRPEL2 that positively induced GRPEL2 transcription and expression in CRC. Our work illustrated the oncogenic function of GRPEL2 in CRC and the molecular mechanism of E2F8/GRPEL2/TIGAR pathway. These findings will provide novel insights and promising therapeutic targets for CRC treatment in the future.

Materials and methods

CRC patients’ samples

The study was approved by the Ethics Committee of the hospital (Approval No. SBQLL-2020-005) and written informed consents were collected from 68 patients for the research application of tissues. All tumor and adjacent normal tissues samples were collected from the hospital. Samples were snap-frozen and stored at − 80 °C until use.

Bioinformatic analysis

RNA-seq data of CRC samples in TCGA database were obtained from the Xena Functional Genomics Explorer (https://xenabrowser.net/heatmap/). The R survminer package v0.4.8 was applied for OS analysis, and to draw the Kaplan–Meier survival curve. Identified GRPEL2 protein partners were analyzed via Kyoto Encyclopedia of Genes and Genomic (KEGG) (http://www.genome.jp/kegg/).

Cell culture and transfection

The CRC cell lines (LOVO, #TCHu82; SW620, #TCHu101; DLD-1, #TCHu134 and HCT116, #TCHu99) were purchased from National Collection of Authenticated Cell Cultures (Shanghai, China) and normal large intestine epithelial cell line FHC was purchased from American Type Culture Collection (#CRL-1831, Manassas, USA). Cells were maintained by the medium of RPMI-1640 (Invitrogen, Carlsbad, CA, USA) supplying 10% fetal bovine serum (FBS, Invitrogen) and penicillin–streptomycin (P.S.) solution (Invitrogen). Small interfering RNAs (siRNAs) targeting GRPEL2 (si-GRPEL2) or negative control (si-NC) and plasmids for the overexpression of TIGAR or E2F8 were purchased from Genechem (Shanghai, China). Transient transfections were conducted with Lipofectamine 2000 (Invitrogen) in the guidance of the instructions and further experiments were conducted 48 h after transfection.

Quantitative real-time PCR (qRT-PCR)

Cells or tissues were isolated by TRIzol reagent (Invitrogen) in the guidance of the manufacturer’s protocol. After quantification, the RNA was reversed transcript into cDNA by the iScript cDNA Synthesis Kit (Bio-Rad, USA). SYBR Green Supermixes (Bio-Rad) was used for qRT-PCR assay. The endogenous controls to normalize target genes and miRNA were GAPDH, respectively. Relative RNA expression was calculated by the 2^−ΔΔCt method. The PCR products of TIGAR were also electrophoresed on 2.5% agarose gel, followed by staining with Ethidium bromide and imaging. The primers sequences were listed below:

GRPEL2: 5′—GAGCCAAAACACCAAGCCTTA—3′ (forward) and 5′—GGAGTTTAATGCT GATGGACCTT—3′ (reverse).

TIGAR: 5′—GGCTTCGGGAAAGGAAATACG—3′ (forward) and 5′—AACCTGGAATACCGCTGTCTG—3′ (reverse).

E2F8: 5′—GACATGCCTAACACAGCAT-3′ (forward) and 5′—ATGCTGTGTTAGGCATGTC—3′ (reverse).

GAPDH: 5′—GAGCCAAAACACCAAGCCTTA—3′ (forward) and 5′—GGAGTTTAATGCT GATGGACCTT—3′ (reverse).

Western blot analysis

Cells or tissues after homogenization were treated with lysis buffer for 20 min at 4 °C. The supernatant was then collected after centrifugation to quantify protein concentration was by BCA assay kit (Beyotime Biotechnology, Shanghai, China). Samples then were boiled and separated by 10% SDS-PAGE, followed with transferring to PVDF membranes (Millipore, USA). After incubating with non-fat milk for 1 h, membranes were then treated overnight at 4 °C with the indicated primary antibodies from Thermo Fisher Scientific (MA, USA) at a 1:1000 dilution (GRPEL2: # PA5-54,723; Cytochrome C: # 45–6100; TIGAR: # PA5-29,151; E2F8: # PA5-142,438; GAPDH: # MA1-16,757). After washing, goat anti-rabbit or anti-mouse HRP-conjugated secondary antibody (#7074 or #7076, Cell Signaling Technology) were used to incubate the membranes at a 1:5000 dilution. The bands were developed by an ECL detection kit (Invitrogen, USA) and their intensities were analyzed by Image J software. Cycloheximide was used to inhibit protein synthesis when detecting the stability of existing proteins.

Cell proliferation assay

EdU assay kit (#ab219801, Abcam, Cambridge, England) was used to test cell proliferation. Cells were incubated with EdU solution for 2 h and then were fixed and incubated with permeabilization buffer for 15 min. After washing, cells were incubated with reaction mix for 0.5 h. The nuclear was stained by Hoechst 33,342 (#4082, Cell Signaling Technology) at 1 µg/mL for 5 min. Cells were then washed, and images were taken and analyzed by fluorescence microscope.

Transwell assay

After the indicated treatment, 5 × 10⁵ cells were suspended in serum-free culture medium for 12 h and then seeded into the upper chamber of a Transwell plate (8 µm, Corning Incorporated, USA). The lower chamber was filled with culture medium containing 10% FBS. After incubation for 48 h, the remaining cells in the upper chamber were removed by cotton swabs, and the migratory cells were fixed with 4% paraformaldehyde for 30 min. After washing, the cells were stained with a crystal violet solution and imaged by a brightfield microscopy (Olympus, Tokyo, Japan) and quantified.

Dual-luciferase reporter assay

HEK-293 T cells were seeded into 24-well plates and co-transfected with 200 ng of luciferase vector with wild-type or mutant GRPEL2 promoter region, and 100 ng of E2F8 overexpression or negative control plasmid, and the pRL-TK plasmid (Promega, Madison, WI). After 48 h, cells were lysed using the lysis buffer (Promega) and signals were determined by the Dual-Luciferase Reporter Assay System (Promega) according to the manufacturer’s protocol. All tests were performed three times.

Xenograft mice tumor model

Animal studies were performed in accordance with the institutional guidelines of the hospital with approved protocol by the Ethics Committee of the hospital (Approval No. 2020–020). Six NOD-SCID mice of 4–6 weeks old for each group were housed in a specific pathogen-free facility. DLD-1 or HCT116 cells (1 × 10⁷) with indicated treatments were suspended in 0.1 mL of PBS and mixed at 1:1 with Matrigel and then injected subcutaneously into the axilla of mice. The tumor volume was monitored every week. Mice were then sacrificed, and tumors were harvested and weighed after 4 weeks, followed by further examinations.

Evaluation of tumor metastasis in mice

For the determination of tumor metastasis, DLD-1 or HCT116 luciferase-expressing cells (1 × 10⁶) with indicated treatments were re-suspended in 0.1 mL of PBS and injected into the spleens of NOD-SCID mice of 4–6 weeks old to induce liver metastasis model. The mice were imaged 4 weeks post the injection by adding the luciferase substrate at a dose of 150 mg/kg using an IVIS200 system (Xenogene, USA). Then liver tissues were harvested for Hematoxylin & eosin (H&E) staining using the Hematoxylin and Eosin Staining Kit (#C0105S, Beyotime, China) according to the manufacture’s protocol.

Immunohistochemistry

Slides were blocked by 5% BSA (bovine serum albumin) and incubated with 1:100 diluted indicated primary antibodies (GRPEL2: # PA5-54,723; TIGAR: # PA5-29,151; Ki-67: #MA5-14,520, Invitrogen) at room temperature for 1 h, followed by incubation with a 1:2000 diluted secondary antibody (#ab205718, Abcam) and diaminobenzidine (DAB) staining. After counterstaining with Mayer’s hematoxylin, the sections were dehydrated, cleared, and mounted. Control slides were made without the treatment of primary antibody. Each section was observed by a light microscope (Zeiss, Shanghai, China).

Transmission electron microscopy

Cells after indicated treatment were seeded on glass coverslips in 6-well plates and then fixed by 2% glutaraldehyde solution in 0.1 M sodium cacodylate buffer (pH 7.4) for 0.5 h. Then the cells were washed by 0.1 M sodium cacodylate buffer (pH 7.4) and processed according to the standard protocol at Electron Microscopy Unit. Images were captured by the Jeol JEM-1400 transmission electron microscope. Mitochondrial major axis length or cristae width was measured by ImageJ software.

ROS detection

Cells after indicated treatments were seeded in 6-well plate and cultured with medium containing 10 μmol/L DCFH-DA (Sigma, San Francisco, CA, USA). After incubation at 28 °C for 0.5 h, the cells were washed with PBS and subsequently imaged by a fluorescent microscope and analyzed by ImageJ software.

Flow cytometry analysis of mitochondrial permeability transition pore (mPTP) opening

Cells after indicated treatments were pre-incubated with 250 nM Calcium-AM plus 1 μM cobalt for 25 min. After washing with fresh PBS, cells were further incubated with medium containing 0.15 mM catechins for 8 h. After that, cells were detected by a flow cytometer NovoCyte 1300 (ACEA, San Diego, CA, USA) within the FITC-channel. Ionomycin treatment was used as the positive control for this assay.

Mitochondrial membrane potential measurement

Cells after indicated treatments were seeded into 6-well plate and JC-1 (Sigma) was added at a final concentration of 5 μM for 0.5 h. Cells were then harvested and detected by flow cytometer. Green (JC-1 monomer) and red (JC-1 aggregates) fluorescence were detected within the FITC-channel and PE-channel, respectively. The mean fluorescence intensity (MFI) was detected, and then the MFI ratio of green/red was calculated.

Protein complex purification and LC–MS/MS

Cells were suspended in lysis buffer (0.5% IGEPAL, 50 mM HEPES, pH 8.0, 150 mM NaCl, 50 mM NaF, 1.5 mM NaVO₃, 5 mM EDTA, 0.1% SDS, 0.5 mM PMSF and protease inhibitors; Sigma), sonicated, treated with benzonase and loaded into spin columns (Bio-Rad) containing Strep-Tactin beads (IBA, GmbH). Liquid chromatography–mass spectrometry (LC–MS) analysis was conducted on an Orbitrap Elite ETD mass spectrometer (Thermo Scientific). Parameters were the same as previously described [19]. RAW files were searched with Proteome Discoverer 1.4 (Thermo Scientific) against SEQUEST search engine. All reported data were based on high confidence peptides assigned in Proteome Discoverer with a 0.01% FDR by Percolator. After determination of TIGAR binding sequence, three amino acids were randomly selected to generate three mutations for further validation.

Co-immunoprecipitation (Co-IP) assay

Cells were first dissolved in lysis buffer and 500 µg of samples were incubated with previously mentioned primary antibody for 12 h. Then Protein A/G PLUS-Agarose (#sc-2003, Santa Cruz Biotechnology) was added, following by the incubation of 2 h. After washing by immunoprecipitation lysis buffer for three times, samples were added with 2 × SDS sample buffer and boiled. Then western blot was used to separate and detect the obtained samples.

Chromatin immunoprecipitation (ChIP) assay

The ChIP assay was conducted by the EZ-ChIP Kit (Millipore, USA). Chromatin of cells was obtained and immunoprecipitated by antibody against E2F8 (# PA5-142,438, Invitrogen). Antibody against IgG (Abcam, USA) was utilized as the negative control. Real-time PCR was used to calculate and determine the fold enrichment as the ratio of E2F8/IgG.

Statistical analysis

GraphPad Prism 6 was utilized for statistical analysis. Experiments were performed three times or more. Data were shown as the mean ± SD and determined by one-way ANOVA followed by Tukey’s multiple comparison test or Student’s t-test. P < 0.05 was taken as statistically significant. Spearman’s correlation was applied to analysis the correlation between different genes expression. Kaplan–Meier survival analysis was applied to determine the correlation between GRPEL2 or TIGAR level and overall survival in CRC patients.

Results

GRPEL2 was upregulated in CRC and related with poor prognosis

To determine the potential regulators for CRC development, we analyzed the gene expression profile of GEO datasets GSE146587, GSE156355, GSE110223 and GSE113513, which all contained 6 paired tumor and non-tumor tissues from CRC patients. A total of 40 genes with significant upregulation in tumor tissues were identified (Figure S1 A), among which 5 genes (KLK8, KRT23, FOXQ1, GRPEL2 and SALL4) were shared by all 4 datasets (Figure S1B). Numerous of studies have demonstrated the functions of those 5 genes related to the development of various kinds of tumors [20,21,22,23,24,25]. Next, the expression of those 5 genes were further tested in 68 paired normal and tumor tissues from CRC patients by real-time PCR. Consistently, all those genes obviously increased in tumor tissues, especially for KRT23 and GRPEL2 (Figure S1 C). Considering the mitochondrial localization of GRPEL2, we speculated that it could regulate the function of mitochondrial. Then the expression of GRPEL2 was evaluated in TCGA database. As shown in Fig. 1A, it was shown to be significantly upregulated in tumor tissues from both colon adenocarcinoma (COAD) and rectum adenocarcinoma (READ) patients. Intriguingly, for the 68 patients, those with high GRPEL2 expressions exhibited shorter overall survival (Fig. 1B). The upregulated expression of GRPEL2 in CRC tumor samples were also demonstrated by IHC staining and was found to be positively correlated with the stages of tumors (Fig. 1C) and level of CRC biomarker CEA and CA199 (Fig. 1D). Moreover, compared to normal human large intestine epithelial FHC cells, GRPEL2 was found to be dramatically upregulated in both mRNA (Fig. 1E) and protein (Fig. 1F) levels in CRC cell lines. Taking together, those data strongly indicated that GRPEL2 was overexpressed in CRC with significant correlation to the poor clinical prognosis.

Fig. 1

GRPEL2 was highly expressed in CRC and associated with poor prognosis. A GRPEL2 expression profile in CRC generated from TCGA database. B Kaplan Meier analysis of 68 CRC patients divided based on GRPEL2 expression. C Tissue staining results of GRPEL2 in tumor tissues from patients under different stages. D Correlation between serum levels of CEA and CA199 and expression of GREPEL2 in 68 CRC patients. A–D, N = 68. E mRNA and F protein levels of GRPEL2 in CRC cell lines, E–F, N = 3, *P < 0.05, **P < 0.01, ***P < 0.001

Inhibition of GRPEL2 suppressed proliferation and migration of CRC cells

To identify the function of GRPEL2, it was knocked down by siRNAs (Fig. 2A, B) or overexpressed (Figure S2 A and B) in both DLD-1 and HCT116 cells. In vitro studies demonstrated that inhibition of GRPEL2 significantly suppressed proliferation (Fig. 2C) and migration (Fig. 2D) abilities of DLD-1 and HCT116 cells, which was promoted when GRPEL2 was upregulated (Figure S2 C and D) as determined by EdU staining and Transwell assay. Most importantly, the knock down of GRPEL2 also strongly inhibited tumor growth in both DLD-1 and HCT116 xenograft models, as shown in Fig. 2E–G. Tissue staining results indicated that tumors with decreased GRPEL2 expression exhibited significantly low level of Ki-67 (Fig. 2H). Meanwhile, the liver metastasis ability of CRC cells was dramatically impaired once GRPEL2 was downregulated, as determined by live imaging of liver resident tumor cells and liver tissue staining of metastasis nodes (Fig. 2I, J). Above all, these results comprehensively demonstrated that the inhibition of GRPEL2 could significantly decrease proliferation and migration abilities of CRC cells.

Fig. 2

Inhibition of GRPEL2 suppressed proliferation and migration of CRC. A mRNA and B protein levels of GRPEL2 in DLD-1 and HCT116 cells after indicated treatments. (C) EdU staining of cell proliferation after indicated treatments. D Transwell results of DLD-1 and HCT116 cells after indicated treatments. A–D, N = 3. E Tumor growth curve, F tumor weight and G tumor images of DLD-1 and HCT116 xenograft models after indicated treatments. H Images and statistical results of Ki-67 staining in xenograft tumor tissues. I Live imaging and J liver tissue H&E staining showing CRC cell metastasis. E–J, N = 6, *P < 0.05, **P < 0.01, ***P < 0.001

Inhibition of GRPEL2 promoted mitochondrial injury of CRC cells

Regarding the mitochondrial localization of GRPEL2, we then focused on the influence of mitochondrial function in response to GRPEL2 inhibition in CRC cells. Transmission electron microscopy (TEM) was used to observe the mitochondrial morphology of DLD-1 and HCT116 cells. As shown in Fig. 3A, the inhibition of GRPEL2 caused a server disruption of mitochondrial ultrastructure with disordered cristae and irregular intermembrane and intercristal space. Moreover, after the downregulation of GRPEL2, a significant increase of ROS production was observed (Fig. 3B), with a decrease of mitochondrial membrane potential (Fig. 3C) and a decrease of Calcein fluorescence signal which indicates an enhanced mPTP opening (Fig. 3D). In addition, an upregulation of cytoplasm Cytochrome c was demonstrated in DLD-1 and HCT116 cells when GRPEL2 was inhibited (Fig. 3E), due to leakage of mitochondrial contents caused by the disruption of mitochondrial ultrastructure and the opening of mPTP. Above all, our data firmly demonstrated that inhibition of GRPEL2 could facilitate the injury of mitochondrial in CRC cells.

Fig. 3

Inhibition of GRPEL2 induced mitochondrial injury. A TEM images of DLD-1 and HCT116 cells after indicated treatments. B Determination of ROS production by DCFH-DA staining in DLD-1 and HCT116 cells. C Determination of mitochondrial membrane potential of DLD-1 and HCT116 cells after indicated treatments. D Flow cytometry results of Calcein detection in CRC cells. E Cytoplasm level of Cytochrome c in DLD-1 and HCT116 cells after treatments. A–E, N = 3, *P < 0.05, **P < 0.01, ***P < 0.001

GRPEL2 interacted with TIGAR to regulate mitochondrial function

In order to study the mechanism of how GRPEL2 affected mitochondrial function, LC–MS/MS was conducted to identify the protein partners that interacted with GRPEL2. Totally 112 proteins associated with metabolism pathway by KEGG analysis were discovered, among which 7 proteins (TIGAR, DAP3, TOMM5, TOMM7, GRPEL1, DECR1 and CISD1) were functional related to mitochondrial based on previous reports [5, 8, 26,27,28,29,30] (Fig. 4A). Further validation was performed by Co-IP assay and GRPEL2 was shown to interact with TIGAR, DAP3 and GRPEL1 directly (Fig. 4B). Meanwhile, only TIGAR was found to be significantly upregulated in tumors from both COAD and READ patients (Fig. 4C), which was similar to that of GRPEL2. Next, three different mutations were generated in the potential binding region of TIGAR for GRPEL2 without any influence on the correct transcription (Fig. 4D, E). When exogenously expressed in HEK-293 T cells, only wild-type TIGAR was able to interact with GRPEL2 as determined by Co-IP assay, not the mutant ones (Fig. 4F). Furthermore, when GRPEL2 was inhibited in DLD-1 or HCT116 cells, the stability of TIGAR protein was impaired with decreased half-life (Fig. 4G). In sum, our data well demonstrated that GRPEL2 could interact with TIGAR directly, which was also significantly upregulated in CRC.

Fig. 4

GRPEL2 interacted with TIGAR and regulated its expression. A Potential protein partners for GRPEL2 by LC–MS/MS combined with KEGG analysis. B Co-IP results validated the protein partners interacting with GRPEL2. C Expression profile of target genes in CRC generated from TCGA database. D Scheme of wild-type and mutated TIGAR protein sequence. E Relative mRNA expression in cells after transfected with different TIGAR plasmids. F Co-IP results of the direct interaction between GRPEL2 and TIGAR. G Protein stability of TIGAR in DLD-1 and HCT116 cells after indicated treatments. B–G, N = 3, *P < 0.05, **P < 0.01, ***P < 0.001

Expression of GRPEL2 was positively correlated with TIGAR in CRC

Regarding the similar expression trend of GRPEL2 and TIGAR in CRC patients, we further analyzed if they were positively correlated. As shown in Fig. 5A, TIGAR was found to be significantly upregulated in different types of cancers, especially in COAD and READ. Meanwhile, in the previously mentioned 68 paired CRC tissues, we demonstrated that TIGAR was obviously overexpressed in tumor tissues compared to normal ones, exhibiting a positive correlation with GRPEL2 in tumor tissues (Fig. 5B, C). Tissue staining results also verified the significant upregulation of TIGAR protein in CRC tumor tissues in comparison with normal ones (Fig. 5D). Moreover, CRC patients with high expression of TIGAR exhibited decreased overall survival in compared with those having low expression of TIGAR (Fig. 5E). To determine if GRPEL2 could regulate TIGAR in CRC, we tested the level of TIGAR in DLD-1 and HCT116 cells in response to knocking down of GRPEL2 by siRNA. As shown in Fig. 5F, G, the inhibition of GRPEL2 dramatically suppressed the expression of TIGAR in CRC cells while overexpression of GRPEL2 significantly stimulated TIGAR’s expression. Our data indicated that GRPEL2’s expression was positively associated with TIGAR and GRPEL2 could regulate TIGAR in CRC.

Fig. 5

GRPLE2 expression level was positively correlated with TIGAR in CRC. A Expression profile of TIGAR in clinical samples generated from TCGA database. B Relative expression of TIGAR in 68 paired normal and tumor tissues of CRC patients. C Correlation of GRPEL2 and TIGAR expression levels in CRC tumors. A–C, N = 68. D Tissue staining results of TIGAR in normal and tumor tissues from CRC patients. E Kaplan Meier analysis of 68 CRC patients divided based on TIGAR expression. F mRNA of TIGAR in CRC cells after inhibition of GRPEL2 and G protein levels of TIGAR in CRC cells after inhibition or overexpression of GRPEL2. F–G, N = 3, *P < 0.05, **P < 0.01, ***P < 0.001

Overexpression of TIGAR rescued anti-tumor effect induced by GRPEL2 inhibition

To investigate if TIGAR was responsible for GRPEL2’s function in CRC, the knockdown efficiency of GRPEL2 and the overexpression efficiency of TIGAR were detected by qPCR and WB in DLD-1 and HCT116 cell lines, The results showed that knockdown GRPEL1 inhibited the expression of GRPEL2 and TIGAR, while overexpression of TIGAR promoted the expression of TIGAR, and overexpression of TIGAR could reverse the decrease of TIGAR caused by knockdown GRPEL1 (Fig. 6A–B and Figure S3 A–B). Subsequently, Control, si-GRPEL2#2 + oe-NC, oe-TIGAR and si-GRPEL2#2 + oe-TIGAR were selected as the research objects Mitochondrial injury was induced by the inhibition of GRPEL2 but not affected by the overexpression of TIGAR, as demonstrated by the TEM scanning and the evaluation of ROS production, mitochondrial membrane potential, Calcein and Cytochrome c levels. However, upregulation of TIGAR significantly alleviated the mitochondrial injury caused by GRPEL2 inhibition (Fig. 6C). Meanwhile, inhibition of GRPEL2 upregulated the ROS level, while overexpression of TIGAR inhibited the production of ROS and significantly alleviated the high ROS level caused by GRPEL2 inhibition (Fig. 6D and FIG. S3 C). Inhibition of GRPEL2 decreased the mitochondrial membrane potential, while overexpression of TIGAR increased the mitochondrial membrane potential (Fig. 6E and Fig. S3D). In addition, overexpression of TIGAR (Figs. 6F and S3E) and elevated cytochrome c levels in cytoplasm (Figs. 6G and S3 F) also blocked the enhancement of calcin fluorescence signal.. Intriguingly, despite the downregulation of GRPEL2 and the overexpression of TIGAR alone, increased expression of TIGAR strongly promoted the proliferation and migration of CRC cells in vitro (Figure S4 A and B). Most importantly, compared to GRPEL2 inhibition alone, combined upregulation of TIGAR significantly stimulated the tumor growth in DLD-1 and HCT116 xenograft models (Figure S4 C–E). Tissue staining results also demonstrated an obvious upregulation of Ki-67 level in tumor tissues induced by TIGAR overexpression (Figure S4 F). Meanwhile, TIGAR promoted liver metastasis for DLD-1 and HCT116 cells in vivo in the presence of GRPEL2 inhibition (Figure S4G and H). This data strongly indicated that GRPEL2 directly regulated TIGAR and upregulation of TIGAR could rescue anti-tumor effect caused by GRPEL2 inhibition in CRC.

Fig. 6

Overexpression TIGAR rescued the phenotypes caused by GRPEL2 inhibition in DLD-1 cells. A mRNA and B protein levels of GRPEL2 and TIGAR after indicated treatments. C TEM images of DLD-1 and HCT116 cells after indicated treatments. D Determination of ROS production by DCFH-DA staining in DLD-1 cells. E Determination of mitochondrial membrane potential of DLD-1 cells after indicated treatments. F Flow cytometry results of Calcein detection in CRC cells. G Cytoplasm level of Cytochrome c in DLD-1 cells after treatments. A–G, N = 3, *P < 0.05, **P < 0.01, ***P < 0.001

GRPEL2 was positively regulated by E2F8 in CRC

After illustrating the downstream pathway of GRPEL2 in CRC development, we then tried to identify the upstream regulator that controls GRPEL2 expression. Considering the critical role of E2F family in regulating gene expression [13], we carefully examined the potential mediator for GRPEL2 among this family proteins and found that E2F8 was predicted to regulate GRPEL2 by hTFtarget database, as shown in Fig. 7A. There were two potential binding sites of E2F8 in GRPEL2 promoter locus, as predicted by JASPAR database (Fig. 7B, C). Interestingly, TCGA database also indicated that E2F8 was significantly overexpressed in COAD and READ tumor tissues, which was the same as GRPEL2 (Fig. 7D). To verify if E2F8 could regulate GRPEL2, we cloned wild-type and binding site mutated GRPEL2 promoter into luciferase reporter system and found that overexpression of E2F8 could only simulate the luciferase signal with wild-type GRPEL2 promoter, not the mutant one (Fig. 7E). Furthermore, the ChIP assay demonstrated that E2F8 directly interacted with the endogenous promoter region of GRPEL2 (Fig. 7F). In the 68 paired CRC patients’ samples, E2F8 was significantly upregulated in tumor samples compared to normal ones, exhibiting a positive correlation with the level of GRPEL2 in tumor tissues (Fig. 7G, H). Then we used siRNA to inhibit E2F8 expression to determine if GRPEL2 was affected. As expected, in response to E2F8 inhibition, expression of GRPEL2 were dramatically suppressed in DLD-1 and HCT116 cells (Fig. 7I, J). Above all, our data suggested that E2F8 could positively regulate GRPEL2 expression in CRC cells.

Fig. 7

E2F8 promoted GRPEL2 transcription and expression in CRC. A Predicted interaction between E2F8 and GRPEL2 genes by hTFtarget database. B Predicted conserved binding motif of E2F8 by JASPAR database. C Predicted binding sites of E2F8 in GRPEL2 promoter region by JASPAR database. D Expression profile of E2F8 in CRC generated by TCGA database. E Luciferase reporter assay verified that E2F8 could regulate GRPEL2 expression. F ChIP assay for the interaction of E2F8 in GRPEL2 promoter. E–F, N = 3. G Relative expression of E2F8 in 68 paired CRC tissues. H Correlation of E2F8 and GRPEL2 expression levels in CRC tumor samples. G–H, N = 68. I mRNA and J protein levels of E2F8 and GRPEL2 in DLD-1 and HCT116 cells after indicated treatments. I-J, N = 3. *P < 0.05, **P < 0.01, ***P < 0.001

Overexpression of E2F8 enhanced CRC development in the presence of GRPEL2 inhibition

After confirming how E2F8 regulates GRPEL2 expression, we further examined if E2F8 was involved in GRPEL2’s function in CRC. The efficiency of E2F8 overexpression and GRPEL2 suppression was demonstrated by western blot and qPCR (Fig. 8A, B). After the overexpression of E2F8, a significant decrease in ROS production was observed (Figure S5 A), with an increase of mitochondrial membrane potential (Figure S5B) and an increase of Calcein fluorescence signal which indicates an enhanced mPTP closing (Figure S5 C). In addition, a downregulation of cytoplasm Cytochrome c was demonstrated in DLD-1 and HCT116 cells when E2F8 was overexpressed (Figure S5D), due to restore of mitochondrial contents loss caused by the disruption of mitochondrial ultrastructure and the opening of mPTP. Intriguingly, overexpression of E2F8 promoted the proliferation and migration of CRC cells, while EdU staining results indicated that upregulation of E2F8 significantly increased proliferation of DLD-1 and HCT116 cells in the presence of GRPEL2 inhibition (Fig. 8C). Meanwhile, increased expression of E2F8 dramatically promoted migration of DLD-1 and HCT116 cells when GRPEL2 was suppressed, as demonstrated by Transwell assay (Fig. 8D). These results indicated that the anti-proliferation and migration effects caused by GRPEL2 inhibition could be rescued upon E2F8 overexpression.

Fig. 8

Overexpression of E2F8 restored the tumorigenesis of CRC cells after GRPEL2 inhibition. A mRNA and B protein levels of E2F8 and GRPEL2 after indicated treatments. C EdU staining of cell proliferation after indicated treatments. D Transwell assay results of DLD-1 and HCT116 cells after indicated treatments. A–D, N = 3, *P < 0.05, **P < 0.01, ***P < 0.001

Discussion

Currently, CRC recurrence and metastasis due to the lack of effective treatment approaches lead to the poor clinical prognosis [31]. Therefore, it is critical to discover novel mechanisms to identify diagnostic markers, prognostic factors and therapeutic targets. Bioinformatics has now become a powerful tool to study the mechanism of tumor development, identify effective tumor biomarkers, and provide personalized treatment strategies [32]. In the present work, we discovered the function of GRPEL2 in CRC by analyzing the data from tissue microarray, internal experiments and TCGA database. Our results demonstrated that GRPEL2 was overexpressed in CRC with a predicted role for OS of CRC patients, suggesting its potential function in CRC development. To verify this idea, we studied the phenotype of CRC cells in response to GRPEL2 knock down. The results demonstrated that GRPEL2 deficiency suppressed cell growth and migration in vitro, exhibited anti-tumor effect in vivo with decreased tumor growth and inhibited tumor metastasis.

GRPEL2 regulates the mitochondrial homeostasis [5]. Mitochondrial is the source of ROS in cells, which is correlated with mitochondrial membrane potential. Accumulating ROS often leads to decrease of mitochondrial membrane potential, which is involved in the cell apoptosis process [33]. Our data proved GRPEL2 deficiency in CRC cells obviously triggered the disruption of mitochondrial ultrastructure, leading to increased ROS production and decreased mitochondrial membrane potential. These findings were consistent with the previous studies about GRPEL2’s anti-apoptotic role in liver cancer [25].

To discover the downstream targets of GRPEL2 in regulating CRC progression, an unbiased LC–MS/MS screening was performed, focusing on the potential protein partners involved in the mitochondrial homeostasis. TIGAR was then identified with consistent upregulation in CRC and proved to interact with GRPEL2 directly by Co-IP assay. TIGAR is related to ROS limitation by stimulating the oxidative pentose phosphate pathway [8]. TIGAR is involved in damage resolution of intestinal epithelium [34]. TIGAR is also upregulated in different kinds of tumors, acting as antioxidant for tumor progression [35]. Loss of TIGAR in mice has been shown to increase survival and decrease tumor growth in intestinal and lymphoma models [8, 36]. Our bioinformatic analysis demonstrated that TIGAR was also upregulated in CRC, with a positive correlation to the level of GRPEL2 in tumor tissues. Intriguingly, the inhibition of GRPEL2 in CRC cells was shown to significantly reduce the expression of TIGAR and overexpression of TIGAR could rescue the phenotypes caused by GRPEL2 suppression. Our data for the first time linked GRPEL2’s function in CRC to the positive interaction and regulation of TIGAR, underlying the molecular mechanism of how GRPEL2 contributed to CRC progression.

Despite the numerous studies about the downstream targets and pathways of GRPEL2 in different contexts, few works have illustrated the upstream regulator for this gene. By the bioinformatic approach, we predicted E2F8 could potentially bind into the promoter region of GRPEL2 and verified through different methods. E2F8 has a conserved DNA-binding domain to trans-repress its target gene promoters [37], which correlates with cell proliferation [38, 39]. Furthermore, E2F7 and E2F8 genes are under the regulation of E2F1 by the oscillation of their mRNA and protein during cell cycle. Consequently, the crosstalk between E2F7/8 and E2F1 is critical in regulating cell cycle and cell death [40]. In addition, E2F family proteins have been reported to regulate the expression of mitochondrial-associated genes and loss of such regulation results in severe mitochondrial defects [41]. For example, E2F1 is found to play an important role in regulation of glycolysis and fatty acid metabolism and its loss of function leads to improved utilization of glucose and insulin responsiveness [42]. Interestingly, our data demonstrated E2F8 positively regulates GRPEL2 expression in which the knock down of E2F8 in CRC cells decreased the expression level of GRPEL2. Meanwhile, mitochondrial injury in CRC cells induced by the inhibition of GRPEL2 could be significantly rescued by the overexpression of E2F8, indicating that E2F8 could regulate mitochondrial function via regulating GRPEL2. However, consistent with previous reports, E2F8 exhibited an oncogenic role by promoting CRC cells proliferation and migration in the presence of GRPEL2 deficient. Whether this transcription activation role for E2F8 on GRPEL2 is CRC-specific or GRPEL2-specific needs to be further elucidated in the future work.

In summary, our work discovered GRPEL2 significantly overexpressed in CRC tumor tissues in comparison with normal ones, as well as CRC cell lines. GRPEL2 upregulation was related to poor prognosis of CRC and inhibition of GRPEL2 was shown to suppress CRC proliferation and migration via causing mitochondria injury. Meanwhile, TIGAR was demonstrated to be regulated by GRPEL2 and overexpression of TIGAR could rescue CRC progression in the presence of GRPEL2 inhibition. Moreover, E2F8 was identified as the upstream regulator of GRPEL2 that positively induced GRPEL2 transcription and expression in CRC. Our work illustrated the oncogenic role of GRPEL2 in CRC development and the molecular mechanism of E2F8/GRPEL2/TIGAR pathway. These findings will provide novel insights and promising therapeutic targets for CRC treatment in the future.