Surgical cancer tissue specimens and adjacent normal mucosa (≥5 cm away from the tumor margins) were obtained from 14 patients with colorectal cancer who had undergone surgery at the Third Xiangya Hospital of Central South University. Informed written consent was obtained from each patient, and the research protocols were approved by the Medical Ethics Committee of Xiangya Medical College. No patients had received preoperative adjuvant therapy. After collection, all tissue samples were immediately frozen in liquid nitrogen until use.

The full length TSC22D2 gene was cloned into the pIRESneo3-Flag vector with NheI and StuI (Takara, Dalian, China) sites. The TSC22D2 gene was amplified using the following primers: TSC22D2 sense primer (5′-ACG TGCTAGCGCCACCATGTCCAAGATGCCGGCCAA-3′), TSC22D2 anti-sense primer (5′-ACTGAGGCCTTTATGCTG AGGAGACATTCG-3′). Full-length PKM2 was generously provided by Professor Xianghuo He (Shanghai Medical College, Fudan University, Shanghai, China) and has been previously described (10).

The SW480 (human colorectal carcinoma) and HEK293 (human embryonic kidney) cell lines were cultured in Dulbecco’s modified Eagle’s medium (Life Technologies, Grand Island, NY, USA) with 10% fetal bovine serum (FBS) supplemented with 100 U/ml penicillin and 100 mg/ml streptomycin. The HeLa human cervical carcinoma cells were grown in 1640 medium (Life Technologies) supplemented with 10% FBS, 100 U/ml penicillin and 100 mg/ml streptomycin.

Cells were transfected using the Lipofectamine 3000 reagent (Life Technologies) according to the manufacturer’s instructions. HEK293, SW480 and HeLa cells stably expressing Flag-tagged TSC22D2 (Flag-TSC22D2) or the control vector (Flag-NC) were obtained by puromycin (600–1,000 ng/ml) (Life Technologies) selection for one month. All cells were maintained under standard culture conditions (37°C, 5% CO₂).

Total RNAs were isolated using the TRIzol reagent (Invitrogen, Carlsbad, CA, USA), and were converted to cDNA using the GoScript™ Reverse Transcription System (Promega, Madison, WI, USA) according to the manufacturer’s instructions. All real-time PCR reactions were performed with SYBR Green (Takara) using the Bio-Rad CFX Connect Real-Time system (Bio-Rad, Hercules, CA, USA). The following PCR program was used: denaturation at 95°C for 30 sec, followed by 40 cycles consisting of denaturation at 95°C for 5 sec, annealing at 60°C for 30 sec, and extension at 72°C for 30 sec. A melting curve analysis was applied to assess the specificity of the amplified PCR products. The amount of each target gene was quantified by the comparative C_T method using GAPDH as the normalization control. The following primers were synthesized from Life Technologies and used to amplify TSC22D2, CCND1 and GAPDH: TSC22D2 forward primer (5′-TGAT GGTGATGAAGACAGTGC-3′), TSC22D2 reverse primer (5′-GGGTTGGGAGTTGGGATAAT-3′); CCND1 forward primer (5′-CAACCTCCTCAACGACC-3′), CCND1 reverse primer (5′-CTTCTGTTCCTCGCAGAC-3′); GAPDH forward primer (5′-AACGGATTTGGTCGTATTGG-3′), GAPDH reverse primer (5′-TTGATTTTGGAGGGATCTCG-3′). All experiments were carried out in triplicate.

Nuclear and cytosolic extracts were fractionated using the NE-PER nuclear and cytoplasmic extraction kit (Thermo Scientific, Bremen, Germany) according to the manufacturer’s protocol. To isolate total protein, cells were harvested by scraping and transferred to SDS sample buffer supplemented with a protease inhibitor cocktail and the PhosSTOP phosphatase inhibitor (Roche, Pleasanton, CA, USA). Equal levels of protein from the samples were separated using SDS-PAGE gel electrophoresis and transferred to a PVDF membrane (Millipore Corp., Billerica, MA, USA). The membrane was blocked with PBST containing 5% skim milk for 1 h and then incubated overnight with the indicated primary antibodies at 4°C. The membrane was washed three times in PBST and subsequently incubated with an HRP-conjugated secondary antibody for 2 h at 37°C. The membranes were stripped of the primary antibodies and re-probed with additional antibodies as necessary. Bound antibodies were visualized using the enhanced chemiluminescence kit (Millipore).

The following antibodies were used in this study: monoclonal mouse anti-Flag (F1804, Sigma, St. Louis, MO, USA), anti-HA (H3663, Sigma), polyclonal rabbit anti-human TSC22D2 (AV39137, Sigma), polyclonal rabbit anti-human PKM2 (Cell Signaling, Danvers, MA, USA) and anti-CCND1 (Cell Signaling). An antibody against GAPDH (Cell Signaling) served as an endogenous control for equal loading, and anti-H3 (Beyotime, China) served as a nuclear control.

Cell proliferation was measured by counting viable cells using the Z2 Particle Counter and Size Analyzer Beckman Coulter (Miami, FL, USA). Briefly, SW480 and HeLa cells stably expressing Flag-tagged TSC22D2 (Flag-TSC22D2) or the control vector (Flag-NC) were seeded at 1.2×10⁴ cells per well in 24-well plates in quadruplicate, the number of viable cells in each well was measured at 0, 1, 2, 3, 4 and 5 days.

For the colony formation assays, SW480 and HeLa cells were seeded in a 6-well plate at a density of 500 and 200 cells per well respectively, and incubated at 37°C with 5% CO₂ for 2 weeks. The medium was changed every 3–4 days. At the end of the incubation, the medium was removed, and the cells were washed twice with PBS, fixed with 4% paraformaldehyde for 20 min, stained with crystal violet for 30 min at room temperature, washed and imaged. Colonies of >50 cells were identified using an inverted microscope (Olympus IX50; Olympus Corp., Tokyo, Japan) and counted.

SW480 and HeLa stable cells were plated at a density of 4×10⁵ cells per well in 6-well plates and cultured in medium without serum (starvation treatment) for 12 h to synchronize cell cycle progression. Cells were incubated in serum-containing growth medium for an additional 24–36 h and subsequently trypsinized, washed, fixed with 70% ice cold ethanol overnight at 4°C. DNA was stained by incubating the cells in PBS containing PI (50 μg/ml) and RNase A (50 μg/ml). The cell cycle distribution was determined by flow cytometric analysis using a MoFlo™ XDP High-Performance Cell Sorter (Beckman Coulter, Brea, CA, USA) and the data were analyzed using Summit v.5.2 software.

HEK293 Flag-NC, HEK293 Flag-TSC22D2, SW480 Flag-NC, SW480 Flag-TSC22D2 stably transfected cells and HEK293T cells transiently cotransfected with Flag-TSC22D2 and HA-PKM2 were seeded in 100-mm dishes. The cells were lysed in modified lysis buffer (50 mM Tris-HCl pH 7.5, 100 mM NaCI, 50 mM NaF, 1 mM Na₃VO₄, 30 mM sodium pyrophosphate, 0.5% NP-40 and 0.5 mM PMSF (Sigma) supplemented with an EDTA-free protease inhibitor cocktail (Roche). Lysates were incubated with 5 μg of primary antibody overnight at 4°C. Then, 40 μl of a 1:1 slurry of protein A/G Plus-agarose (Santa Cruz, CA, USA) and a specific antibody were added to the cells for ≥4 h at 4°C. The immunoprecipitates were washed four times with lysis buffer and boiled with sample loading buffer and then analyzed by western blotting.

Proteins were resolved on a 10% polyacrylamide gel, stained with Coomassie Brilliant Blue (R250) and subjected to mass spectrometry. Briefly, the stained SDS-PAGE gels were scanned, and stained bands were excised, cut into small (<1 mm³) pieces and washed three times by repetitive dehydration and hydration with acetonitrile and ammonium bicarbonate, respectively. Proteins were in-gel reduced in the presence of 25 mM DTT and 25 mM NH₄HCO₃ for 1 h at 56°C, immediately alkylated using 50 mM IAA and 25 mM NH₄HCO₃ for 30 min at room temperature in the dark and digested overnight at 37°C with 5 μg of trypsin in 25 mM NH₄HCO₃. Digested peptides were recovered, dried and resuspended in 50% CAN and 0.1% TFA. The peptide mixture was analyzed by nano-liquid chromatography-tandem mass spectrometry using an LTQ Velos-Orbitrap MS (Thermo Scientific, Waltham, MA, USA) coupled to an Ultimate RSLC nano-LC system (Dionex). Briefly, Raw MS data files were processed using the Proteome Discoverer v.1.4 (Thermo Scientific). Processed files were searched against the Swiss-Prot human database using the Sequest HT search engine. Mass tolerances for precursor and fragment ions were set to 10 ppm and 0.8 Da, respectively, in the database searches. Peptide identification with false discovery rates <1% (q-value <0.01) were discarded.

TSC22D2, HEK293, SW480 and HeLa cells grown on culture slides were maintained in a 35-mm dish. After cultured for 24 h, the cells were washed with phosphate-buffered saline (PBS), fixed with 4% paraformaldehyde in PBS for 10 min, permeabilized with 0.25% Triton X-100 for 30 min and blocked in 5% bovine serum albumin (BSA) for 1 h at room temperature. Fixed cells were then incubated with rabbit anti-human TSC22D2 antibody (1:1,000) at 4°C overnight, washed three times in PBS and stained with Alexa Fluor 488 donkey anti-rabbit IgG (Invitrogen, 1:1,000) for 2 h in the dark at room temperature. The cells were then incubated with DAPI to stain the nuclei and visualized using a confocal laser scanning microscope (Olympus Corp.).

For the colocalization assays, HEK293T transfected with Flag-TSC22D2 and HA-PKM2 were incubated with the anti-Flag mouse monoclonal antibody (1:1,000) and the anti-PKM2 rabbit monoclonal antibody (1:1,000) at 4°C overnight and subsequently stained with Alexa Fluor 488 donkey anti-mouse IgG (Invitrogen, 1:1,000) and Alexa Fluor 594 donkey anti-rabbit IgG (Invitrogen, 1:1,000) for 2 h at room temperature. After incubation with DAPI to stain the nuclei, the cells were visualized using a confocal laser scanning microscope (UltraView Vox; Perkin-Elmer, Waltham, MA, USA).

The experiments were repeated at least three times. All data are presented as the mean ± standard deviation. Paired Student’s t-test (two-tailed) and unpaired Student’s t-test (two-tailed) were performed to compare the means of two samples unless otherwise indicated. A p-value <0.05 was considered statistically significant. All statistical analyses were performed using the GraphPad Prism software (Graphpad Software, Inc.).

We focused on human colorectal cancer in this study because it was the most frequent malignancy in the multi-cancer family. We first evaluated the expression of TSC22D2 transcripts in colorectal cancer by interrogating the public gene expression GEO databases (GSE8671, GSE32323 and GSE24514) and found that the expression of TSC22D2 was lower in human colorectal cancer samples than in non-tumor samples (Fig. 1A). To confirm these observations, the expression of TSC22D2 was evaluated in 14 pairs of human CRC samples (primary tumor tissues and paired adjacent non-tumor tissues). TSC22D2 expression was reduced in 13 of 14 (92.9%) of colorectal tumors compared with the paired adjacent non-tumor tissues (Fig. 1B, p=0.015), consistent with the data provided by the GEO database. These results suggest that TSC22D2 plays a suppressor role in CRC tumorigenesis.

To investigate the function of TSC22D2 in colorectal cancer, the full-length TSC22D2 gene tagged with three Flag tags was cloned into the pIRESneo3 plasmid and stably transfected into SW480 and HeLa cells. The expression of TSC22D2 was confirmed by qRT-PCR and western blot analysis (Fig. 2A). Immunofluorescence assays demonstrated that TSC22D2 predominantly localized to the cytoplasm under steady state conditions (Fig. 2B). Cell proliferation assays and clone formation assays were conducted to investigate the effect of TSC22D2 on cell proliferation. TSC22D2-overexpressing cells exhibited a significantly slower growth rate (Fig. 3A) and a reduced number of clones (Fig. 3B) compared with control cells, indicating that TSC22D2 functions as a suppressor in CRC cell growth. Furthermore, flow cytometry (FCM) revealed that TSC22D2 induced a substantial increase in the proportion of cells at the G0/G1 phase and a concomitant decrease in the proportion of cells at the S phase of the cell cycle (Fig. 3C). These results indicate that TSC22D2 induces cell cycle arrest at the G0/G1 phase.

To explore the underlying mechanism by which TSC22D2 exerts its tumor-suppressive function, we performed co-IP assays (Fig. 4A) and LC-MS/MS analysis (Fig. 4B) in SW480 and HEK293 cells stably transfected with TSC22D2. We identified 90 and 79 candidate proteins in SW480 and HEK293 samples, respectively (Fig. 4C). There were 27 overlapping results for proteins associated with metabolism (PKM2, ATP5A1, LDH and ENO1), stress response and inflammation (STK39, HSPD1, HSP90AA1 and HSP90AB1), cell proliferation (NRBP1, PKM2, YWHAQ, LDHB, PCNA and PHB2), apoptosis (HSPD1, CAPN1, DNAJA1, HSP90AA1 and RPS3) and migration/metastasis (YWHAQ, HSPD1 and PHB2) between the 2 cell lines (Fig. 4D and Table I). These findings suggest that TSC22D2 plays a role in these biological pathways in cancer cells.

We selected PKM2, a glycolytic enzyme that catalyzes the conversion of phosphoenopyruvate (PEP) and ADP to pyruvate and ATP for further analysis due to its relatively high mass spectrometry score and its previously reported role in cancer cell growth and survival (7–9).

The interaction between PKM2 and TSC22D2 was further confirmed by IP analysis of HEK293T cells transfected with HA-tagged PKM2 and Flag-tagged TSC22D2 (Fig. 5A). To determine whether TSC22D2 and PKM2 co-localize in cells, we performed immunofluorescence staining with anti-Flag and anti-PKM2 antibodies in HEK293 cells. The results revealed that TSC22D2 and PKM2 co-localized primarily in the cytoplasm (Pearson’s correlation, 0.954) (Fig. 5B). Taken together, these data suggest that TSC22D2 physically associates with PKM2.

Accumulating evidence indicates that PKM2 is crucial for aerobic glycolysis and that it provides a growth advantage to tumors (11,12). To determine if the effect of TSC22D2 on cell growth is mediated by PKM2, we evaluated the expression of PKM2 and observed that there were no significant changes in PKM2 expression at the mRNA and protein level in cells overexpressing TSC22D2 (Fig. 6A).

PKM2 predominantly localizes to the cytosol and plays an important role in metabolic functions. Recently, several independent studies demonstrated that PKM2 can translocate to the nucleus and induce the expression of gene products required for tumorigenesis by activating multiple transcription factors (13–18). Thus, we asked whether TSC22D2 could affect the nuclear function of PKM2. A subcellular fractionation analysis was performed, and a slight but significant decrease in the nuclear levels of PKM2 was observed in TSC22D2-overexpressing cells compared with the control cells (Fig. 6B).

Cyclin D1, a key regulator required for the G1/S cell cycle transition, was reported to be a downstream gene of nuclear PKM2 (15,16). Therefore, we next investigated the effect of TSC22D2 on cyclin D1 expression. Intriguingly, TSC22D2 repressed the expression of cyclin D1 at both mRNA and protein levels (Fig. 6C and D), suggesting that TSC22D2 might inhibit cell growth by the influence on nuclear PKM2 and cyclin D1.