Twenty-three primary paired CRC samples containing adjacent tissues from the Second Affiliated Hospital of Xi’an Jiaotong University (Xi’an, China) were subjected to western blotting as well as RT–qPCR. Prior to sample collection, each patient signed an informed consent form. Permission for this study was obtained from the Clinical Research Ethics Committee of the Second Affiliated Hospital of Xi’an Jiaotong University (#2019022).

The TMAs (HColA180Su15, HCol-Ade180CS-01) with 164 adjacent and 186 CRC tissues were procured from Shanghai Outdo Biotech (China). The CRC tissues on the TMAs correspond to the clinicopathological parameters (age, gender, tumor diameter, tumor location, tumor invasion, lymph node metastasis, organ metastasis and AJCC stage) of the CRC patients. And 101 CRC tissues correspond to the follow-up information of CRC patients. Operative period was from January 2009 to October 2009, while the follow-up time was July 2015. Follow-up intervals were 6.7 to 7.2 years.

HT29, CaCo2, HCT116, SW480, DLD1 and HUVEC cell lines were acquired from the American Type Culture Collection (ATCC). NCM460 and HCoEpic normal human colon cell lines and Lovo cell line were acquired from the Cell Bank of the Chinese Academy of Sciences (Shanghai, China). PCR tests did not reveal any mycoplasma contamination. DLD1 and Lovo cells were grown in RPMI 1640 medium, and other cells were grown in high-glucose DMEM with 100 mL/L fetal bovine serum (FBS, Gemini, USA) in a humidified 5% CO₂ incubator at 37°C. To induce hypoxia, cells were cultured with 1% O₂, 94% N₂ and 5% CO₂ (vol/vol) at 37°C.

HCT116 cells were infected with the lentiviral vectors for TMEM100 overexpression or negative control lentiviral vectors (Shanghai Genechem Co., Ltd.). The efficiency of infection was determined by western blotting. Then, we used a medium with 2 μg/mL puromycin to establish stable overexpressing (LV-TMEM100) or negative control (LV-vector) HCT116 cell lines.

SW480 cells were transfected with TMEM100 specific siRNAs (GenePharma, Shanghai, China). The siRNA sequences were: si-TMEM100-1: 5`-CAGACUUUAUGUUCAUAGUUCUUCCUC-3`; si-TMEM100-2: 5`-CUUCCACAACUACAUAGGGUAUUGUUU-3`; the negative control sequence (si-NC) was UUC 5`-UCCGAACGUGUCACGUTT-3`. The lipofectamine 2000 reagent (Invitrogen, Carlsbad, CA) was supplemented to enable transfection, as instructed by the manufacturer. siRNAs were transfected for 6 h at a final concentration of 50 nM.

Paraffin-embedded sections and the TMAs were dewaxed with dimethylbenzene and gradient alcohol (100%, 95%, 85% and 75%). The antigen retrieval was done by microwaving sections in 0.01 M sodium citrate, pH 6.0. Then, incubation of slides was done for 20 min in the presence of 3% hydrogen peroxide after which they were incubated in the presence of goat serum at room temperature (RT) for 30 min. Then, overnight incubation of slides was done in the presence of primary antibodies (anti-TMEM100 (GeneTex, GTX83507); anti-Ki-67 (Servicebio, GB111499); anti-HIF-1α (Servicebio, GB13031-1); anti-CD34 (Servicebio, GB13013)) at 4°C followed by supplementation with biotinylated anti-IgG and incubation at 25°C for 1 h. After incubation for 30 min with streptomycin-HRP, sections were DAB-stained, counterstained with hematoxylin, washed with water, dehydrated by an alcohol gradient (80%, 90% and 100%) and dimethylbenzene. Lastly, neutral balsam and coverslips were used to seal the slides.

For the immunostaining intensity evaluation of TMAs, two pathologists that were not aware of clinicopathological features as well as patient outcomes independently scored the tissue microarrays. Immunoreactivity was grouped into 5 grades (% score) based on stained cells % as: 0, no staining; 1, 1-25%; 2, 26-50%; 3, 51-75%; and 4, >75%. Staining intensities were assigned into 4 grades (intensity score) as follows: 0, negative; 1, weak; 2, moderate; and 3, strong. Then, a final overall histological score was calculated by multiplication of the two values (18). An overall score of 0–12 was calculated and graded as low (score ≤ 5) or high (score>5) in order to estimate the relationship between the expression of TMEM100 and the clinicopathological parameters of CRC patients.

For the immunostaining intensity evaluation of murine subcutaneous tumors’ tissues, Image-Pro Plus 6.0 software (Media Cybernetics, Maryland United States) was used to calculate the ratio of the positively stained area to the area of the field of view, and the integral optical density from the five fields was taken.

Paraffin-embedded sections were dewaxed with xylene, dehydrated with gradient ethanol, stained with Harris’s hematoxylin for 5 min, differentiated with 0.1% hydrochloric acid alcohol, stained with 1% eosin for 2 min, dehydrated by gradient ethanol, cleared in xylene, mounted with neutral gum, and observed and photographed under a microscope.

Total proteins from cultured cells and clinical patient samples were extracted by the RIPA buffer (Beyotime, China) containing a protease suppressor cocktail (1:100, Bimake). For the western blotting assay, protein separation was achieved using 10% sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS–PAGE) after which they were transferred to polyvinylidene difluoride membranes. Membranes were blocked with Tris-buffered saline supplemented with 0.1 mL/L Tween 20 (TBST) and fresh prepared 100 g/L non-fat milk for 2 h at 25°C, followed by incubation at 4°C overnight in the presence of various primary antibodies (anti-TMEM100 (GeneTex, GTX83507); anti-CDK2 (Abcam, ab32147); anti-Cyclin D1 (Abcam, ab40754); anti-PCNA (Abcam, ab92552); anti-Bcl-2 (Abcam, ab182858); anti-caspase3 (Abcam, ab32351); anti-cleaved caspase3 (Abcam, ab32042); anti-PARP1 (Abcam, ab191217); anti-cleaved PARP1 (Abcam, ab32064); anti-E-cadherin (Abcam, ab40772); anti-Vimentin (Abcam, ab8978); anti-HIF-1α (Abcam, ab1) and anti-β-actin (Fdbio science, AP0060)). Then, they were washed using a TBST buffer followed by 1 h of incubation in the presence of goat anti-rabbit IgG-HRP antibodies (1:10000, Zhuangzhi, EK020) and goat anti-mouse IgG-HRP antibody (1:10000, Fdbio science, FDR007) at 25°C. Visualization of immunoreactivity was done using an ECL substrate (Bio–Rad, Hercules, CA) on the Gene Gnome XRQ System. ImageJ (National Institute of Health, MD) was used for densitometric assessments.

HCT116 cells were infected with LV-TMEM100 or LV-vector lentiviruses, and then cultured for 48 h under hypoxic conditions. Then total proteins from cultured cells were extracted by the RIPA buffer (Beyotime, China) containing a protease suppressor cocktail (1:100, Bimake). For the IP assay, first, anti-HIF-1α was mixed with Protein A/G PLUS-agarose (RuiSike Science & Technology Co., Ltd, China) at 4°C for 12 h. Then, antibody- agarose complex was incubated with the protein samples for 12 h at 4°C. After washing using a washing buffer, the antibody-agarose-protein complex was obtained and mixed with SDS-PAGE sample loading buffer (Beyotime, China). Then the mixture was heated in a Dry thermostat (Beyotime, China) at 95°C for 10 min. The protein solutions were further evaluated by the western blotting assay.

The TRIzol reagent (Invitrogen, USA) was used for total RNA extraction, as denoted by the manufacturer. Then, the Transcript First-Strand cDNA Synthesis Kit (Roche, Denmark) was used to reverse-transcribe 2 µg of total RNA to cDNA. RT–qPCR was done using a FastStart Universal SYBR Green Master Mix (Roche, Denmark) on a 7500 Fast Real-Time PCR System (Applied Biosystems, Foster City, CA, USA). Relative levels of target genes were evaluated by the 2^-△△Ct approach, with normalization to β-actin. Primer sequences for this assay were: β-actin (forward 5’-GGCACCACACCTTCTACAATGAGC-3’, reverse 5’-GATAGCACAGCCTGGATAGCAACG-3’), TMEM100 (forward 5’-GGAGAAGAGCCCCAAGAGTG -3’, reverse 5’- TGCAGCGGTAGCAGGAGA-3’), HIF-1α (forward: 5’-AGCTTCTGTTATGAGGCTCACC-3’, reverse: 5’-TGACTTGATGTTCATCGTCCTC-3’).

Cell viabilities were assessed by the Cell Counting Kit-8 (CCK-8, 7sea Pharmatech Co., Ltd). Cells were seeded in each well of a 96-well plate in 200 μL of medium (5 × 10³ cells/well) and then the CCK-8 reagent (10 μL) dissolved in 90 μL of serum-free medium was added to each well 24 h, 48 h, 72 h, and 96 h after plating. After the addition of the CCK-8 solution, cell incubation was done at 37°C for 30 min. Absorbance was measured at 450 nm.

Cells were cultured into 6-well plates (1×10³ cells/well) followed by incubation for 14 days at 37°C. Then, they were fixed in paraformaldehyde (4%) followed by crystal violet staining. Colony counts were done visually.

Cell migration assay was assessed by using a Transwell chamber with a pore size of 8 μm (Corning, USA). A total of 8×10⁴ cells were seeded into the upper chamber in 200 µL of serum-free DMEM. Then, 600 µL of DMEM with 200 mL/L FBS was added to the lower chamber. After incubation at 37°C with 50 mL/L CO₂ for 24 h, the cells that had migrated through the membrane were fixed in 40 mL/L paraformaldehyde and stained with crystal violet, while the cells in the upper chamber were carefully removed using a cotton swab. After drying, the cells that had migrated were counted with a microscope at 100× magnification in 5 random fields.

Cell invasion assay was assessed by using a Transwell chamber with a pore size of 8 μm (Corning, USA). For cell invasion assay, the upper chambers were covered with Matrigel (60 μL, 200 mg/mL, BD Biosciences, San Diego, CA, USA), and concreted for 4 h in the incubator. The next steps were same as the cell migration assay.

For cell cycle assessment, 2×10⁵ cells in each group were grown in 6-well plates in serum-free media for 36 h to maintain the same cell cycle. Then the media was replaced by high-glucose DMEM with 100 mL/L fetal bovine serum to culture cells for another 24 h. Adherent cells were digested into a centrifuge tube and washed gently to remove trypsin. Then 70% ice ethanol was added to fix the cells overnight at 4 °C. The next day, fixed cells were washed and resuspended in 500 μL PI/RNase Staining Buffer (BD Biosciences, Franklin Lakes, NJ, USA). After 15 min of incubation in the dark at 25°C, flow cytometry (BD Biosciences, Franklin Lakes, NJ, USA) was performed to assess cell cycle phases.

For cell apoptosis analysis, a PE Annexin V/7-amino-actinomycin Detection (7-AAD) Kit (BD Biosciences) was used according to the manufacturer’s instructions. To determine the apoptosis fraction, cells without PE Annexin V or 7-AAD staining were recognized as negative control. And cells staining with only PE Annexin V were in early apoptosis stage. Cells staining with PE Annexin V and 7-AAD were in late apoptosis stage.

The amount of cell-secreted VEGF proteins in the medium were evaluated by a VEGF ELISA kit, as instructed by the manufacturer (#KE00085, Proteintech, CA, USA).

This assay consisted of three groups: the control group (LV-vector group), the TMEM100 overexpression group (LV-TMEM100 group), and the TMEM100 overexpression with MG132 treatment group (LV-TMEM100 + MG132 group). HCT116 cells (20×10⁴) infected with LV-vector virus were inoculated in 6-well cell culture plates in the LV-vector group. HCT116 cells (20×10⁴) infected with LV-TMEM100 virus were inoculated into 6-well cell culture plates in the LV-TMEM100 group and the “LV-TMEM100+MG132” group. For the “LV-TMEM100 + MG132” group, MG132 was supplemented to the medium and cultured for 42 h before media collecting. The final MG132 concentration was 1 μM. All cells were grown in hypoxic environments for a total of 48 h. Then the media in the indicated groups were collected. Then HUVECs (1×10⁴ cells/well) were inoculated in 48-well cell culture plates precoated with polymerized Matrigel (BD Biosciences, San Diego, CA, USA) followed by incubation at 37°C for 4 h in conditioned media derived from the indicated cells under normoxic conditions. Changes in cellular morphology were microscopically observed. Total tube lengths in five random view-fields/well were determined by ImageJ software, and the mean value obtained. This assay was performed in triplicate.

The in vivo experiment was permitted by the Animal Care Committee of Xi’an Jiao Tong University (NO.XJTULAC2019-988). Athymic nude male mice (BALB/c, 4 weeks old) were acclimatized to laboratory conditions (12 h/12 h dark/light, 23°C, 50% humidity, and ad libitum provision of water and food) for 2 weeks before experimentation. The animals were maintained in specific pathogen-free environments. They were age-matched, randomized into groups (n=5/group), and randomly kept in various squirrel cages.

For xenograft tumor models, every mouse was subcutaneously administred with 7×10⁶ indicated cells into the left groin. The long diameter (R) and short diameter (r) of subcutaneous tumor were measured every five days using a vernier caliper. The tumor volume was calculated following the equation: Volume = (R×r²)/2. After 3 weeks, the mice were euthanized by an overdose of barbiturate (intravenous injection, 150 mg/kg pentobarbital sodium) to collect tissues. For lung metastasis models, the tail vein of each mouse was injected with 8×10⁶ of the indicated cells. After one month, mice were euthanized by an overdose of barbiturate to collect tissues. Murine tissues fixed in 10% formaldehyde were cleared in xylene, dehydrated with gradient ethanol, embedded in paraffin, and serially sectioned at 5 μm for HE staining and IHC. The numbers and sizes of metastatic tumors in the lungs from mice were evaluated by two pathologists who were blinded to mice grouping information. After which lung tissues were subjected to HE staining.

GraphPad Prism 8.0.1 (La Jolla, CA, USA) and SPSS 22.0 (United States) were used for data analyses. The hazard ratio and its 95% confidence interval were calculated by Cox proportional hazards regression. One-way ANOVA, chi-square test, or Fisher’s exact tests were used to assess the associations between TMEM100 levels and clinic-pathological features. The t test or the one-way ANOVA were used for analysis of continuous data. Kaplan–Meier and log-rank tests were used to estimate the overall survival rates by GraphPad Prism 8.0.1. Randomization and blinding were done in all assays. Data are shown as mean ± SD, p<0.05 denoted significance.

Compared to normal tissues (8.7%), mRNA levels of TMEM100 were downregulated in 21 out of 23 (91.3%) human CRC tissues ( Figure 1A ). In six pairs of CRC tissues, TMEM100 protein levels were markedly suppressed in CRC tissues, relative to normal adjacent tissues ( Figure 1B ). In comparison with that in normal colon epithelial cell line (NCM460), TMEM100 protein levels were downregulated in all 6 CRC cell lines (HT-29, HCT116, Lovo, SW480, CaCo2, and DLD1) ( Figure 1C ). Then, we further evaluated TMEM100 expression in 186 CRC and 164 adjacent tissues in tissues microarrays using IHC, and revealed that TMEM100 was dramatically suppressed in CRC tissues relative to adjacent tissues ( Figure 1D ). These findings imply that downregulation of TMEM100 plays a role in the pathogenesis of CRC.

The TMEM100 protein level was tested in 186 CRC tissues and 164 adjacent tissues in the tissue microarray using IHC. The higher and lower expression levels of TMEM100 were evaluated semiquantitatively by the staining intensity (high score: 6-12; low score: 1-5). Futher statistical analysis results showed the lower TMEM100 protein levels were associated with age, tumor diameter, tumor invasion, lymph node metastasis and metastasis ( Table 1 ). The Kaplan–Meier survival analyses revealed that lower relative TMEM100 protein levels in CRC patients correlated with worse overall survival compared with higher TMEM100 expression in CRC patients ( Figure 1E ). And the CRC patients with low TMEM100 level and positive regional lymph node metastasis indicates detrimental prognosis ( Figures 1F, G ). The univariate and multivariate Cox regression results indicated that lower TMEM100 protein levels is an independent risk factor for a detrimental prognostic outcomes in CRC patients ( Tables 2 , 3 ).

Malignant proliferation is one of the distinct features of CRC, and we investigated whether TMEM100 impedes CRC cell growth. In this study, HCT116 cells in the TMEM100 overexpression group exhibited a lower proliferation rate than those in the control group, whereas downregulating TMEM100 in SW480 cells led to increased cell proliferation rates ( Figure 2A ). In the colony formation assays, HCT116 cells in the TMEM100 overexpression group had a diminished ability for colony formation, relative to the control group ( Figure 2B ). The downregulation of TMEM100 enhanced the colony formation capacity of SW480 cells, relative to the control group ( Figure 2B ). In the xenograft mouse models, tumors formed by the TMEM100-overexpressing HCT116 cells were found to be slow-growing and distinctly smaller than those formed by the control cells ( Figures 2C, D, E ). The IHC staining revealed that the percentage of Ki67 positive cells of subcutaneous xenograft tumors were significantly lower in the TMEM100 overexpression group, relative to control group ( Figure 2F ). These findings imply that upregulation of TMEM100 inhibits CRC cell proliferations.

We sought to identify the pathomechanisms through which TMEM100 inhibits CRC cell proliferation. A cell cycle analysis was conducted to determine if upregulation of TMEM100 induces the suppression of CRC cell growth by arresting a specific phase of the cell cycle. Flow cytometry revealed that HCT116 cells in the TMEM100 overexpression group exhibited a markedly low abundance of cells in S phase, relative to the control group, and the percentage of cells in G₁ phase was markedly high ( Figure 3A ). Downregulating TMEM100 in SW480 cells enhanced the abundance of cells in S phase ( Figure 3A ). Thus, upregulation of TMEM100 seemed to inhibit G₁/S transition during cell cycle progression. To further clarify how TMEM100 suppresses G₁/S phase transition, key regulators related to G₁ phase were evaluated at protein levels. It was found that upregulation of TMEM100 markedly reduced the levels of cyclin D1, CDK2 as well as PCNA proteins in HCT116 cells; however, downregulation of TMEM100 evidently increased the level of these three proteins in SW480 cells ( Figure 3C ).

Meanwhile, the results of flow cytometry analysis showed that the percentage of apoptotic cells was dramatically increased in TMEM100 overexpressing HCT116 cells, but reduced in TMEM100 downregulated SW480 cells compared with control cells ( Figure 3B ). Cleaved-PARP1 and cleaved-caspase3 are the protein markers of apoptosis process, and Bcl-2 is a distinguished anti-apoptotic protein. The results of western blotting exhibited the protein level of PARP1, cleaved-PARP1, caspase3, cleaved-caspase3 and Bcl-2 ( Figure 3C ), which indicated that cell apoptosis was promoted by TMEM100 in CRC cells. These findings imply that TMEM100 inhibits cell proliferation via the arrest of G₁/S phase transitions and induces apoptosis in CRC.

Enhanced migration is another feature of CRC. Cell migration assay revealed that migrated and invasive HCT116 cells were markedly low in TMEM100 overexpression group, relative to control group ( Figures 4A, B ). For SW480 cells, the numbers of migrated and invasive cells were significantly high in si-TMEM100 groups, relative to si-NC groups ( Figures 4A, B ). Vimentin and E-cadherin are epithelial to mesenchymal transition (EMT) biomarkers, and western blotting analysis revealed that E-cadherin levels in HCT116 cells were elevated and Vimentin levels were low in TMEM100 overexpression group, compared to control group ( Figure 4C ). Consistently, E-cadherin as well as vimentin levels exhibited opposite patterns when TMEM100 was downregulated in SW480 cells ( Figure 4C ). Then, we evaluated the in vivo effects of TMEM100 on migratory capacities of CRC cells by injecting mice tail veins with HCT116 cells infected with LV-vector or LV-TMEM100 lentivirus. Histology showed that the abundance of metastatic tumors in lungs of TMEM100 overexpression group were low, relative to control group ( Figure 4D ). These findings imply that TMEM100 suppresses metastasis of CRC cells by regulating EMT.

We presented the potential association of TMEM100 with HIF-1α in the introduction. The excessive accumulation of HIF-1α further exerts its cancer-promoting effect under hypoxia. HIF-1α accumulation can be induced when cells were cultured under 1% O₂ condition for 12 h in vitro (13, 19). To establish the significance of TMEM100 in CRC, we evaluated HIF-1α levels in a tissue microarray with the same catalog number as that used to evaluate TMEM100 levels via IHC. TMEM100 protein levels were negatively associated with HIF-1α levels ( Figures 5A, B ). The mRNA levels of HIF-1α showed no changes when TMEM100 was overexpressed ( Figure 5C ). However, HIF-1α protein levels were markedly low under normoxic and hypoxic environments in the TMEM100 overexpression group, relative to the control group ( Figure 5D ). Vascular endothelial growth factor (VEGF) is among the HIF-1α target genes, and VEGF secreted by cells binds to its receptor and induces the generation of new blood microvessels. Overexpressions of TMEM100 markedly reduced VEGF release in HCT116 cells under normoxic and hypoxic conditions ( Figure 5E ). We then examined the influence of TMEM100 on HIF-1α protein content and microvascular density in subcutaneous xenograft tumors. The detection of endothelial cell markers on microvessels is one of the recognized methods to evaluate tumor angiogenesis. For example, CD34 has been used to evaluate angiogenesis in many malignancies. The IHC staining revealed that HIF-1α protein levels of subcutaneous xenograft tumors were significantly lower, and the microvascular density marked by CD34 was drastically suppressed in the TMEM100 overexpression group, relative to control group ( Figure 5F ). Therefore, we confirmed that the overexpression of TMEM100 led to a reduction in HIF-1α content and microvascular density.

Given that TMEM100 had no influence on the mRNA level of HIF-1α during normoxia and hypoxia, which indicated that TMEM100 had no effect on HIF-1α transcription, we assessed whether TMEM100 impacts HIF-1α degradation. Cycloheximide (CHX), a common reagent used to inhibit protein synthesis, blocks the elongation phase of eukaryotic translation. To further reveal whether TMEM100 regulated HIF-1α stability, we treated CRC cells with CHX to prevent protein synthesis. Overexpressed TMEM100 shortened the half-life of HIF-1α from 61 min to 40 min, which hinted that TMEM100 might be involved in HIF-1α degradation processes ( Figure 5G ). The ubiquitin–proteasome degradation system is the major pathway of HIF-1α degradation. HIF-1α was immunoprecipitated from cells, and the abundance of HIF-1α-linked ubiquitin evaluated by immunoblotting with an anti-ubiquitin antibody. The amounts of HIF-1α-conjugated ubiquitin from TMEM100-overexpressing cells were markedly low, compared with control cells during hypoxia ( Figure 6A ). We further tested whether TMEM100 inhibits HIF-1α degradation via the proteasome. TMEM100-overexpressing HCT116 cells were treated with MG132, an inhibitor of proteasome, under hypoxia. The results showed that overexpressing TMEM100 significantly decreased HIF-1α levels. However, in the presence of MG132, TMEM100 overexpression did not suppress HIF-1α levels, implying that TMEM100 inhibits HIF-1α accumulation in a proteasome-dependent manner ( Figure 6B ). Thus, TMEM100 promotes HIF-1α degradation via the ubiquitination–proteasome pathway in CRC cells.

We next verified the relationship of TMEM100, HIF-1α, angiogenesis and migration in CRC. HCT116 cells were grown in hypoxic conditions, and the TMEM100 overexpression group showed fewer migrating cells than the control group ( Figure 6C ). Then cells were treated with MG132 to block the ubiquitin/proteasome degradation of HIF-1α. Cells in the “TMEM100+MG132” group showed significant more migrating cells than those in the TMEM100 overexpression group, but considerable migrating cells relative to the control group ( Figure 6C ). In vitro tube formation assay revealed that cells in TMEM100 overexpression group showed weaker angiogenesis induction capacity, relative to the control group ( Figure 6D ). Cells in the “TMEM100+MG132” group showed significantly enhanced angiogenesis induction capacity compared with the TMEM100 overexpression group ( Figure 6D ). Thus, overexpressed TMEM100 suppressed the migration as well as angiogenesis induction capacity of CRC cells, but these effects were reversed by blocking the ubiquitin/proteasome degradation pathway process using MG132, which hinted that TMEM100 inhibited the migration and the angiogenesis induction capacities of CRC cells by impacting HIF-1α degradation via the ubiquitination/proteasome pathway ( Figure 6E ).