TNFR2 deficiency impairs the growth of mouse colon cancer

Objective: Tumor necrosis factor (TNF) receptor type II (TNFR2) is expressed by a wide spectrum of tumor cells including colon cancer, non-Hodgkin lymphoma, myeloma, renal carcinoma and ovarian cancer, and its exact role remains to be fully understood. In this study, we examined the effect of genetic ablation of TNFR2 on in vitro and in vivo growth of mouse MC38 and CT26 colon cancer cells. Methods: CRISPR/Cas9 technology was used to knockout TNFR2 on mouse MC38 and CT26 colon cancer cells. In vitro growth and colony formation of wild-type (W.T.) and TNFR2 deficiency of MC38 and CT26 cells, as well as the potential mechanism, was studied. The growth of W.T. and TNFR2 deficient MC38 and CT26 tumors in mice and intratumoral CD8 CTLs were also examined. Results: TNFR2 deficiency impaired in vitro proliferation and colony formation of cancer cells. This was associated with the inhibition of protein kinase B (AKT) phosphorylation and enhanced autophagy-induced cell death. Moreover, deficiency of TNFR2 also markedly impaired in vivo growth of MC38 or CT26 in the syngeneic C57BL/6 mice or BALB/c mice, respectively, accompanied by the decrease in soluble TNFR2 levels in the circulation and the increase in the number of tumor-infiltrating IFNγ+ CD8 cells. Conclusion: TNFR2 plays a role in the growth of mouse colon cancers. Our study provides further experimental evidence to support the development of TNFR2 antagonistic agents in the treatment of cancer.


Introduction
Aberrant high level of TNFR2 is expressed by a broad spectrum of human tumors, including colon cancer 1 , ovarian cancer 2 , breast cancer 3,4 , cholangiocarcinoma 5,6 , renal cell carcinoma 7 , colorectal cancer 8,9 , non-small cell lung cancer 10 , non-Hodgkin lymphoma 11 , and thoracic esophageal squamous cell carcinoma 12 . The clinicopathologic and prognostic implications of TNFR2 expression by tumor tissues were also reported. For example, TNFR2 expressed by breast cancer is associated with tumor size, clinical stage, and pathological grade 4 . TNFR2 expression by thoracic esophageal squamous cell carcinoma (ESCC) is positively related to invasion, clinical stage, and poor overall survival 12 . Furthermore, polymorphism of TNFRSF1B, the gene encoding TNFR2, is associated with the risk, survival, or response to the treatment of cancer patient [13][14][15][16] .
Consistent with the previous clinical investigations, studies based on human tumor cell lines also revealed the functional consequences of TNFR2 expression. For example, it was shown that TNFR2 promotes the growth of human SW480, COLO205, and HT29 colorectal cancer cells through PI3K-AKT signaling pathway or activation of STAT3 9,17 . Furthermore, in human MDA-MB-231 and MCF-7 breast cancer cells, as well as Kapras299 and L428 lymphoma cells, TNFR2 expression promotes the growth of tumor cells and protects malignant cells from DNA damage caused by chemotherapeutic agent adriamycin 18,19 . TNFR2 also mediates the effect Ivyspring International Publisher of TNF in promoting invasiveness of human CCKS1 cholangiocarcinoma cells 5 . In contrast, blockade of TNFR2 has been shown to induce the death of human OCCAR3 ovarian cancer cells 20 , and silencing of TNFR2 results in the inhibition of phosphorylation of AKT in HT29 cells 9 and inhibits the proliferation of Karpas299 cells 19 . These in vitro studies advanced our understanding of the role of TNFR2 in the development, growth, metastasis and survival of human tumor cells. However, the role of TNFR2 in the in vivo growth of the tumor and the underlying mechanism remains elusive.
In order to further define the role of TNFR2 on tumor development, we used CRISPR/Cas9 technology to knock out TNFR2 gene in MC38 and CT26 colon cancer cells (designated as MC38/TNFR2 -/and CT26/TNFR2 -/-, respectively). The result showed that the growth of resultant MC38/TNFR2 -/and CT26/TNFR2 -/cells was markedly inhibited in both in vitro and in vivo settings. Furthermore, the deficiency of TNFR2 markedly reduced the phosphorylation of AKT and promoted autophagy in cancer cells. Interestingly, genetic ablation of TNFR2 markedly reduced the blood levels of soluble TNFR2 and increased the number of tumors infiltrating IFNγ + CD8 CTLs. Furthermore, tumors in 15% MC38/ TNFR2 -/cell-bearing mice were completely regressed, and these mice had developed tumor-specific immunity. Therefore, our data support the notion that TNFR2 promotes tumor growth and thus represents a therapeutic target in the treatment of colon cancer.

Mice, cells, and reagents
8-12 weeks old, female BALB/c mice and female C57BL/6 mice were initially purchased from Jackson Laboratory and maintained under specific pathogen-free in the Animal Facility of the University of Macau. The animal research protocol was given approval by the Animal Research Ethics Committee of the University of Macau.
The CT26 colon cancer cell line and Lewis lung carcinoma (LLC) were purchased from the American Type Culture Collection (ATCC). The MC38 colon cancer cell line was purchased from iCell Bioscience Inc. (Shanghai, China). Penicillin/streptomycin (P.S.), HEPES, fetal bovine serum (FBS), L-glutamine, RPMI 1640 medium, and high glucose DMEM medium were purchased from Gibco/ThermoScientific (Waltham, MA, USA). CT26 colon cancer cells were cultured in RPMI 1640 medium supplemented with 10% FBS, 1% P.S., 10 mM HEPES and 2 mM L-glutamine at 37 o C and 5% carbon dioxide (CO2) in a humidified incubator. LLC and MC38 colon cancer cells were cultured in high glucose DMEM medium supplemented with 10% FBS and 1% P.S. at 37 o C and 5% CO 2 in a humidified incubator.

Design of TNFR2 sgRNAs and generation of TNFR2-CRISPR/U TM plasmids (based on CRISPR/Cas9 technology)
Guide RNA sequences for CRISPR/U TM were designed by Guangzhou Ubigene Biosciences Co., Ltd (China). Two synthesized DNA oligos were inserted into pX330 CRISPR plasmid. After transformation and propagation, pX330 containing gRNA targeting TNFR2 (TNFR2-CRISPR/U TM plasmid) were extracted with Plasmid Midi Kit (TransGen Biotech, Beijing, China) and verified with DNA sequencing.
Generation of TNFR2 knockout cell line with CRISPR/Cas9 technology TNFR2 knockout MC38 and CT26 cells were constructed by Guangzhou Ubigene Biosciences Co., Ltd (China). Briefly, the TNFR2-CRISPR/U TM plasmid was sequenced to confirm that the sequence inserted was correct. At final, insert oligonucleotides for mouse TNFR2 gRNA1 (F): ACAAGCGTGCC ACGCTGAAGAGG and TNFR2 gRNA2 (R): CTACTCAGTCCTCGCCAATGAGG were used in the following study. These two TNFR2 gRNAs were used to target the exon 2 of Tnfrsf1b gene ( Figure S1). 3,000,000 MC38 and CT26 colon cancer cells were placed in a sterile tube and centrifuged at 300 g for 4 min, respectively. After the supernatant was discarded, the cells were resuspended with 600 µL Buffer R and mixed with 30 µg endotoxin-free TNFR2-CRISPR/U TM plasmid. Add 3 ml Buffer E2 into the shock cup, and then put it into the card slot of the electrometer. The cells and plasmid mixture were extracted with 100 µL electrogun heads, inserted into the electroshock cup, and the electroshock was started under the electroshock condition. After the electrotransfer was completed, the cells were inoculated in a 6-well plate with preheated medium ready for further culture. 24-48 hours after transfection, the effect of electrotransfer was observed under the microscope, the wells with the best cell viability and percentage of green fluorescent cells were selected, and the medium containing puromycin was replaced for screening. After 2-3 days of screening, the cells were digested into a single-cell suspension for cell counting. A certain amount of cell suspension was diluted by the limited dilution method and inoculated into 96-well plates. The cells were incubated at 37 °C and 5% CO2 for static culture. Single-cell clones were screened out. The deletion of TNFR2 was confirmed by flow cytometry. The PCR products of the target site were ligated to T-Vector pMD™19 (Simple) (Takara) for DNA sequencing. Finally, two single-cell clones were expanded for inoculation in their respective in vivo tumor models.

RNA Extraction and RT-PCR
Total RNA was extracted using the TRlzol®. Reagents from the cultivated cells according to the manufacturer's instructions. In this way, quantitative measurement was done using a Spectrophotometer NanoDrop TM One Microvolume UV-Vis (Thermo-Fisher Scientific TM , USA). Additionally, cDNA was synthesized using PrimeScript™ RT Reagent Kit with gDNA Eraser (Perfect Real Time) (Takara Bio, USA) and C1000 Touch TM Thermal Cycler (Bio-Rad, USA), and the total RNAs were considered as the template. PCR was performed using a ViiA7 Real-Time PCR System (ThermoFisher Scientific TM , USA) with β-actin as an internal control.

Cell proliferation assay
MC38/WT and MC38/TNFR2 -/cells were counted and planted in 96-well plates with five replicates at 2,000 cells in 100 µL medium per well. CT26/WT and CT26/TNFR2 -/cells were counted and planted in 96-well plates with five replicates at 1,500 cells in 100 µL medium per well. Cell viability was determined at the time indicated by using MTT assay. Optical density (O.D.) values were recorded at 570 nm by using the enzyme-labeled instrument. The cell growth curves were drawn according to O.D. values. The experiment was repeated three times.

Colony formation assay
MC38/WT, MC38/TNFR2 -/cells, CT26/WT, and CT26/TNFR2 -/were counted and planted in 6-well plates at a density of 200 cells in 2 ml medium per well. The medium was refreshed every three days. This assay was terminated when two clones almost grew together. Cells were fixed using 4% paraformaldehyde for 20 min at room temperature and stained using crystal violet staining solution (Beyotime, Nantong, China). The quantities of clones were calculated by Image J. Colonies with > 50 cells were counted. The experiment was repeated three times.
Each sample was denatured with 4 × protein loading buffer (Takara) at 100 °C for 10 min. Then, equal amounts of protein (20 µg per sample) were separated by 12% SDS-PAGE gel and transferred to polyvinylidene difluoride (PVDF, Bio-Rad, USA) membranes. After blocking for 1 h in 5% skim milk powder in PBST at room temperature, the membranes were incubated with the primary antibodies overnight at 4 °C and secondary antibodies for 1 h at room temperature. Protein bands were visualized by a hypersensitive chemiluminescence kit (Thermo Scientific) on ChemiDoc XPS system (Bio-Rad). Protein expressions were measured using the Bio-Rad Image Lab software 5.2.1.

mCherry-GFP-LC3 assay
MC38/WT, MC38/TNFR2 -/cells, CT26/WT, and CT26/TNFR2 -/were seeded on the coverslips in 6-well plates. After 24 h, the cells were infected with 10 µL adenovirus Ad-mCherry-GFP-LC3 (multiplicity of infection (MOI) = 5) for 6 h according to the manufacturer's protocol. Cells were switched to fresh medium and incubated for an additional 24 h. Then, the cells were fixed and blocked for fluorescence detection. Analysis was performed using a laser confocal microscope (Leica SP8). For the cells infected with adenovirus Ad-mCherry-GFP-LC3, autophagosome and autolysosome status was evaluated by counting cells with GFP + /mCherry + -LC3 (yellow) and GFP − /mCherry + -LC3 (red) puncta, respectively. In the tandem mCherry-GFP-LC3 assay, when an autophagosome fuses with the lysosome to form an autolysosome under acidic environments, GFP fluorescence is quenched in the autolysosome, whereas mCherry fluorescence is more stable. Therefore, when an autophagosome has not yet fused with a lysosome or when the degradation's function of the lysosome with acidification is impaired, co-localization of both GFP and mCherry fluorescence shows GFP + /mCherry + -LC3 (yellow) puncta in the merged image. In contrast, mCherry alone (without GFP) fluorescence presents GFP − /mCherry + -LC3 (red) puncta, which corresponds to an autolysosome.

MC38/WT
and MC38/TNFR2 -/were subcutaneously injected into the right flank of C57BL/6 mice in single-cell suspension with 500,000 cells in 0.1 ml of PBS per mouse. CT26/WT and CT26/TNFR2 -/tumor cells were subcutaneously injected into the right flank of BALB/c mice in single-cell suspension with 200,000 cells in 0.1 ml of PBS per mouse. After indicated times, all the lymphoid tissue was pushed through a 70-µm pore size cell strainer to create a single-cell suspension. In some experiments, tumor-free mice were reinoculated with MC38/WT cells into the right flank, and the same number of LLC tumor cells were subcutaneously injected into the left flank. Tumor size was calculated by the following formula: (length×width 2 )/2. "Survival" represents the time to develop a 4 cm 3 tumor or a moribund state, a humane endpoint that triggers euthanasia.

ELISA for soluble TNFR2
Soluble TNFR2 levels in serum were determined using ELISA kit (Sino Biologicals Inc. PA, SEK50128). Samples were processed according to the manufacturer's instructions.

Flow Cytometry
After blocking FcR, cells were incubated with appropriately diluted antibodies and finally suspended in FACS buffer for cytometric analysis. The acquisition was performed by BD LSRFortessa and BD FACSAria™ Fusion flow cytometer. For intracellular cytokine staining, cells were restimulated with ionomycin, PMA, and BD GlogiPlug for 5 h. Data analysis was conducted by using FlowJo software (Tree Star Inc., Ashland, OR). FACS analysis was gated on the live cells only by using a LIVE/DEAD Fixable Dead Cell Stain Kit (Life Technologies).

Statistical Analysis
All data were presented as means ± SEM, and statistically significant differences between groups were performed by a two-tailed, paired Student t test, or ANOVA. All statistical analysis was performed with GraphPad Prism 7.0. A P value < 0.05 was considered to be statistically significant.

Generation of TNFR2 knockout MC38 and CT26 colon cancer cell lines
To generate CRISPR/Cas9 plasmid containing gRNAs specific targeting the second exon of the Tnfrsf1b gene (NCBI Gene ID: 21938), two sgRNAs were inserted into pX330-based plasmid and confirmed with DNA sequencing (data not shown). MC38/WT and CT26/WT cells were transiently transfected with TNFR2-pX330 plasmid by electrotransfer, and the cells were screened by puromycin. After selecting single-cell clones, we analyzed the nucleotide sequences of the PCR products of target DNA and confirmed the indel mutations were introduced into the genome (Supplementary Figure  S2). Deficiency in TNFR2 expression of these two clones was also confirmed by FACS ( Figure 1A) and real-time PCR ( Figure 1B). These clones were used in subsequent studies.

In vitro proliferation and clone formation of TNFR2-deficient cells are impaired
We next examined the effect of deficiency of TNFR2 on in vitro growth and clone formation of tumor cells. As shown in Figure 2A-B, TNFR2 ablation markedly decreased the proliferation of MC38 and CT26 cells (P < 0.05−0.001), and on day 4 or day 3 of culture, the proliferation was inhibited by 43% and 24%, respectively. Furthermore, colony formation of MC38 and CT26 cells was also reduced by 50% and 23% ( Figure 2C-F, P < 0.05−0.01). Therefore, deficiency of TNFR2 impairs the growth and clone formation of MC38 and CT26 cells.

Deficiency of TNFR2 results in the reduced phosphorylation of AKT
The PI3K/AKT is one of the critical dysregulated signaling pathways in different cancer cells, including colorectal cancer (CRC), and its abnormal activation can result in increased proliferation, invasion, and migration of cancer cells 21 . We found that the deficiency of TNFR2 in MC38 and CT26 cells caused the inhibition of the phosphorylation of AKT ( Figure  3A, Figure S6), whereas the phosphorylation of ERK, JNK, and P38 and total protein levels of them did not change (data not shown). Our data are consistent with the result of the previous study based on the silence of TNFR2 in HT29 cells 9 . Therefore, PI3K/AKT may be an important component of TNFR2 signaling pathway in promoting the proliferation of colon cancer cells.

The deficiency of TNFR2 activates autophagy
A previous study reported that the induction of autophagy and apoptosis could inhibit the growth of CRC cells 22 . It was also reported that PI3K/AKT/mTOR signaling pathway plays a critical role in regulating autophagy and apoptosis in CRC 23,24 . Although we could confirm that TNFR2 deficiency resulted in a marked decrease in the phosphorylation of AKT, we failed to observe that deficiency of TNFR2 increased the apoptosis of cancer cells (data not shown). To examine the effect of TNFR2 deficiency on autophagy of mouse colon cancer, we analyzed the autophagy-related hallmarks including Beclin1, LC3, Atg5, and p62/SQSTM1 by Western Blotting. Results showed that the deficiency of TNFR2 on MC38 and CT26 colon cancer cells markedly decreased p62/SQSTM1 expression while increased the expression of Beclin1, Atg5, and LC3-II/LC3-I ( Figure 3B, Figure S6). Further, the subcellular location of LC3, mainly in the autophagosomes, was detected by using a confocal microscope after infecting a report vector Ad-mCherry-GFP-LC3 in both W.T. and TNFR2 knockout MC38 or CT26 cells. The results demonstrated that MC38/TNFR2 -/and CT26/TNFR2 -/cells expressed GFP -/mCherry + -LC3 (red) puncta ( Figure  3C), indicating that autophagosome and lysosome were fused. The autolysosomes present red puncta because GFP fluorescence is more rapidly quenched by low lysosomal pH, whereas mCherry fluorescence is more stable. Thus, in addition to decreased AKT phosphorylation, the increased autophagy of TNFR2-deficient MC38 and CT26 colon cancer cells may also be attributable to the inhibition of growth of these cells.

Deficiency in TNFR2 impairs the growth of tumors in normal mice
To examine the in vivo effects of TNFR2 ablation on the development of MC38 and CT26 colon cancers, we subcutaneously inoculated MC38/TNFR2 -/and control cells (MC38/WT, 500,000 cells) into the right flank of C57BL/6 mice, or CT26/TNFR2 -/and control cells (CT26/WT, 200,000 cells) into the right flank of BALB/c mice, respectively. The result showed that, although there was no effect on the initial tumor formation, deficiency of TNFR2 resulted in marked impairment of tumor growth ( Figure 4A-F). On day 19 or 20 of inoculation, tumor sizes of TNFR2-deficient MC38 cell lines or CT26 cell lines were marked smaller ( Figure 4A-B, P < 0.01−0.001), and the tumor image was shown in Figure 4C-D. Accordingly, the weights of TNFR2-deficient MC38 tumors and CT26 tumors were reduced by 83-85% and 71-77% ( Figure  4E-F, P < 0.01−0.001), respectively. Furthermore, mice bearing TNFR2-deficient tumor cells survived longer significantly ( Figure 4G-H, P < 0.01−0.001). More importantly, about 15% of MC38/TNFR2 -/tumors were completely regressed within 28 days after inoculation. These tumor-free mice were alive for 80 days, whereas mice bearing W.T. tumor succumbed to tumor burden within 40 days after tumor inoculation ( Figure 4G). To investigate whether the tumor-free mice developed long-term MC38 tumor-specific immunity, at 80 days after the tumor inoculation, the surviving mice were reinoculated subcutaneously with MC38/WT tumor cells into the right flanks, and WT LLC tumor cells were inoculated into their left flanks. As controls, both WT LLC tumor cells and MC38/WT tumor cells were inoculated into untreated mice in the same manner (as schematically illustrated in Figure 5A). As expected, all mice (100%) in the control group developed measurable MC38 and LLC tumors by day five after inoculation, including LLC tumors in those surviving mice. However, none of surviving mice developed MC38 colon tumors ( Figure  5B-C). These results indicate that deficient in TNFR2 on tumor cells induced the development of long-term tumor antigen-specific immunity.

Serum levels of sTNFR2 in mice inoculated with TNFR2-deficient tumor cells are reduced
The elevation of soluble TNFR2 (sTNFR2) is correlated with the poor prognosis of tumor patients 1,25-27 . Since sTNFR2 levels were increased in the culture supernatant of MC38 and CT26 cells (Supplementary Figure S3) and the plasma of MC38 and CT26 tumor-bearing mice (Supplementary Figure  S4) in a time-dependent manner, we thus hypothesized that tumors might be a source of sTNFR2. To test this, we generated the same size of W.T. and TNFR2-deficient tumors (about 1000 mm 3 for MC38 tumor and 2000 mm 3 for CT26 tumor) by inoculating TNFR2-deficient cells 12 days before the injection of W.T. cells. On day 19 (MC38/WT) or 20 (CT26/WT) days after tumor inoculation, the levels of sTNFR2 in serum were determined. The results showed that the levels of sTNFR2 were markedly enhanced by greater than 2-3 folds in mice bearing W.T. tumors, compared to tumor-free normal mice ( Figure 6A-B, P < 0.001). Interestingly, sTNFR2 levels in mice bearing TNFR2-deficient tumors were markedly lower than that in mice bearing W.T. tumors (P < 0.001) and were close to that in tumor-free mice ( Figure 6A-B). These data suggest that tumors may be the major, if not solely, source of upregulating part of sTNFR2 in tumor-bearing individuals.

Increased infiltration of IFNγ + CD8 + CTLs in tumor deficient in TNFR2
sTNFR2 is a potent immunosuppressive mediator 28 . Our results also showed that etanercept, a soluble form of TNFR2 29 , could inhibit CD4 + T cell proliferation ( Figure S5). Thus, the reduced levels of sTNFR2 may result in the mobilization of antitumor immune responses. To test this, we examined Th1 cells present in the same size of W.T. and TNFR2 -/tumors (as described in 3.6). On day 19 (MC38/WT) or day 20 (CT26/WT) after tumor inoculation, the tumor-infiltrating leukocytes were analyzed. The result showed that the proportion of live CD45 + , CD45 + TCRβ + , CD45 + TCRβ + CD4 + , and CD45 + TCRβ + CD8 + cells were increased in TNFR2 -/tumor tissues ( Figure S7). More importantly, the proportion of tumor-infiltrating IFNγ-expressing CD8 + CTLs was increased by 2-4 folds in TNFR2-deficient tumors, compared with those in W.T. tumors (Figure 7, P < 0.01−0.001). Thus, our data indicate that deficiency of TNFR2 on MC38 and CT26 tumor cells induced potent CD8 + T cells mediated antitumor immune responses.

Discussion
It has been reported that TNF is expressed in both human and mouse colon cancer cells (such as HCT116 and CT26) 30 , and TNF has the capacity to promote the proliferation of colon cancer cells 31 . Further, it was shown that the increased expression of TNF is associated with advanced colorectal cancer stages 30,32 . In addition, in the presence of effector cells (macrophages) or active complements, anti-TNF treatment with infliximab markedly suppressed the survival of colon cancer cells 31 . Such an effect of TNF on tumor cells is likely mediated by TNFR2. In fact, the functional consequences of TNFR2 expression on mouse tumor cells were reported previously. For example, it was shown that knockdown of TNFR2 by shRNA rendered mouse Lewis Lung Carcinoma (LLC) more susceptible to TNF-induced apoptosis, accompanied by the down-regulation of Vegfa, Hgf, Plafg, and Cxcr4 expression, while the growth of such LLC cells in normal mice was normal 33 . In this study, the efficiency of TNFR2 knockdown was ~90%, while the remaining TNFR2 may still be able to support the in vivo growth of tumor cells. Furthermore, the surface expression levels of TNFR2 on LLC cells maintained in our own lab were much lower (data not shown) than that expressed by CT26 and MC38 tumor cells. That was why we decided to examine the functional implications of TNFR2 by genetically ablating TNFR2 on CT26 and MC38 tumor lines with CRISPR/Cas9 technology. The results of our study clearly showed that the deficiency of TNFR2 potently impaired the in vitro and in vivo growth of the tumor.
Previously, it was shown that overexpression of TNFR2 in tumor cells promotes cell proliferation and clone formation 9 . We thus also created TNFR2overexpressing MC38 and CT26 colon cancer cells by infecting Plvx-IRES-ZsGreen1-TNFR2 lentivirus vector. Plvx-IRES-ZsGreen1 and TNFR2 cDNA were digested by EcoRI and BamHI and then were ligated by Quick T4 DNA ligase to construct the Plvx-IRES-ZsGreen1-TNFR2 overexpression plasmid. Unexpectedly, overexpression of TNFR2 in MC38 and CT26 did not further promote the proliferation of tumor cells. One possible reason is that the original levels of TNFR2 have already rendered the maximal effect on the survival and proliferation of these two lines of cancer cells.
Once cleaved, membrane-bound TNFR2 becomes soluble TNFR2 (sTNFR2). It is well known that sTNFR2 is immunosuppressive and this notion could be verified by our experimental result showing that etanercept, a soluble form of TNFR2 29 , had the capacity to inhibit the proliferation of CD4 + T cells ( Figure S5). Therefore, the decreased levels of sTNFR2 in mice bearing TNFR2-deficient tumor was most likely attributable to the activation and expansion of tumor infiltrating CD8 CTLs.
Unlike TNFR2, TNFR1 is expressed by almost all types of cells except erythrocytes. With a death domain, TNFR1 signaling induces the cascade related to apoptosis 34 . In this study, we also determined the TNFR1 surface expression (using an anti-TNFR1 antibody, Serotec, HM104) by FACS and the soluble TNFR1 levels in the culture supernatant using ELISA kit. Our results showed that the TNFR1 was almost undetectable on the surface of MC38 and CT26 tumor cells. However, the mRNA level of TNFR1 was detectable by real-time PCR by these cancer cells. Furthermore, TNFR2 deletion did not affect the mRNA expression of TNFR1 (data shown in Figure  S8). This result was also similar to that reported by others in human colorectal cancer cell 35 , which indicated that the silencing TNFR2 expression did not affect TNFR1 mRNA level. Further, the soluble TNFR1 levels were much lower than sTNFR2 (data not shown). Nevertheless, elevated levels of sTNFR1 were found in the supernatant of tumor cell culture medium and in the serum of tumor-bearing mice (data not shown), which is consistent with clinical reports 36 . Thus, sTNFR1 may also be attributable to tumor immunosuppression, and this possibility should be addressed in future studies.
Aberrant activation of the PI3K/AKT signaling pathway facilitates a variety of cellular activities such as proliferation, cell cycle progression, aggressiveness, and chemoresistance in many cancers, especially in CRC 37 . Previous studies also revealed that TNFR2 neutralizing antibody could block the activation of PI3K/AKT and ERK signal pathway in cholangiocarcinoma cells 5 , and TNFR2 promoted the growth of human SW480, COLO205, and HT29 colorectal cancer cells through PI3K/AKT signaling pathway or activation of STAT3 9,35 . The immunohistochemistry (IHC) study showed that TNFR2 was positively associated with Ki67 expression in CRC tissues. Further, the western blot analysis found that TNFR2 promoted Ki67 expression in CRC cells via the PI3K/AKT signaling pathway 9 . Thus, the clinical evidence also supports the notion that TNFR2 promotes the growth of CRC. In this study, we found that deficiency of TNFR2 reduced the phosphorylation of AKT ( Figure 3A). In contrast, the phosphorylation of ERK, JNK, STAT3, and NF-κB was not inhibited by TNFR2 deficiency (data not shown), which was likely caused by intact TNFR1 signal, since knockout of TNFR2 did not affect the expression of TNFR1 (data shown in Figure S8). Our data indicate that the inhibition of PI3K-AKT signaling pathway may be a mechanism underlying the growth inhibition found in TNFR2-deficient tumor cells.
Autophagy occurs after nutrient deprivation or following chemotherapy in cancer cells 38,39 . The impact of autophagy on cancer progression is controversial, and whether the activation of autophagy in tumor cells actually causes death or represents a survival mechanism remains unclear. Autophagy also serves to remove damaged and potentially harmful organelles, thereby supporting cell survival. On the other hand, there is conclusive evidence that prolonged over-activation of the autophagosomal/ lysosomal pathway can lead to autophagic cell death 40 . A previous study showed that knockdown of RAB25 promotes autophagy and inhibits cell growth in ovarian cancer cells 41 . Further, autophagy induced by inhibiting the PI3K/Akt/mTOR pathway causes growth inhibition and death of cancer cells 42 . In the present study, deficiency of TNFR2 promoted autophagy, suggesting that the tumor cells underwent stress due to the deficiency of TNFR2 signaling.
Taken together, our study clearly shows that deficiency of TNFR2 on mouse MC38 and CT26 colon cancer cells could markedly impair the growth of the tumor, which may be attributable to the inhibition of AKT signaling pathway, and activation of autophagy and mobilization of antitumor Th1 responses due to reduction of sTNFR2. Our data suggest that the inhibition of tumor growth likely resulted from the direct effect on tumor cells as well as from the mobilization of anti-tumor immune responses. Our data provide further experimental evidence to support the notion that blockage of TNFR2 may represent a novel strategy in cancer immunotherapy.