Fusobacterium nucleatum facilitates cetuximab resistance in colorectal cancer via the PI3K/AKT and JAK/STAT3 pathways

Infectious pathogens contribute to about 20% of the total tumor burden. Fusobacterium nucleatum (Fn) has been associated with the initiation, progression, and therapy resistance in colorectal cancer (CRC). The over-abundance of Fn has been observed in patients with right-sided CRC than in those with left-sided CRC. While the KRAS/NRAS/BRAF wild-type status of the CRC conferred better response to cetuximab in patients with left-sided CRC than with right-sided CRC. However, treatment failure remains the leading cause of tumor relapse and poor clinical outcome in patients with CRC. Here, we have studied the association of Fn to cetuximab resistance. Our functional studies indicate that Fn facilitates resistance of CRC to cetuximab in vitro and in vivo. Moreover, Fn was found to target the PI3K/AKT and JAK/STAT3 pathways, which altered the response to cetuximab therapy. Therefore, assessing the levels and targeting Fn and the associated signaling pathways may allow modulating the treatment regimen and improve prognoses of CRC patients.


Introduction
CRC accounts for a high incidence and mortality rate, globally 1 , and arises due to complex interactions of genetic attributes, lifestyle, and environmental factors 2 . The treatment objectives in patients with metastatic CRC (mCRC) include, reduction of cancer load, cancer growth inhibition, prolonging survival time, and improving the quality of life. Monoclonal antibodies (mAbs) against the epidermal growth factor receptor (EGFR), such as cetuximab and panitumumab, bind the extracellular structural domain of the EGFR, and reinforce receptor intemalization and degeneration 3,4 . These anti-EGFR mAbs have been considered as targeted therapy for the treatment of mCRC patients harboring KRAS wild type (WT) status 5,6 . Furthermore, when administered as a monotherapy, cetuximab may confer persistent responses in 12-17% of the CRC patients 7 , and provide about 72% reaction rate as a combined chemotherapy 8 .
Several studies have described the mechanisms for intrinsic and acquired resistance to treatment with anti-EGFR mAbs, which we have reviewed previously 22 . In brief, these mechanisms mainly include mutations in the RAS, BRAF, PIK3CA, and EGFR, and 4 phosphorylation of the STAT3 and AKT, which induce resistance to the treatment with anti-EGFR mAbs, primarily via persistent activation of the EGFR downstream signaling pathways despite the EGFR blockade. Deregulation of these PI3K/AKT and JAK/STAT3 signaling pathways inhibit apoptosis, and promote multiple drug resistant in tumor cells 23,24 . However, the non-genetic and/or non-epigenetic mechanisms of resistance remain to be explored in details. Further, drug resistance leads to tumor recurrence, and the 5-year survival rate in unresectable mCRC patients remains less than 10% 25 . Additionally, patients rarely respond to immunosuppressive agents 26 .
Therefore, it is necessary to understand the mechanism of resistance to cetuximab in patients with CRC.
Recent reports have indicated the gut microorganisms to be associated with CRC 27-31 . Further, several studies have validated that tumor tissues and fecal specimens of CRC patients show high amounts of Fn than in normal controls [32][33][34][35][36][37] . Furthermore, the abundance of Fn was observed in patients with right-sided CRC than in those with leftsided CRC [38][39][40] . Moreover, CRC associated with Fn presents with worse clinical outcome 36,41,42 . Additionally, Fn enhances tumor growth 43 , protects tumors from immune attack 44 , and promotes chemo-resistance 42 . Taken together, Fn appears to be not only enriched in CRC but also to promote tumor initiation, growth, progression, and therapy resistance in patients.
To understand how cetuximab confers varied response in patients with right-sided and left-sided CRC, we hypothesized that the presence of Fn may contribute to therapy 5 resistance. A thorough literature survey indicated that the role of Fn in resistance to anti-EGFR mAbs remains to be explored in human CRC. Therefore, here, we have explored the association of Fn with resistance to cetuximab using an in vitro and in vivo approach. Our analysis implicates the PI3K/AKT and JAK/STAT3 pathways to confer resistance to cetuximab in response to Fn in CRC cells. Moreover, given the high proportion of cetuximab resistance in CRC patients, identification of such amenable targets should allow designing better treatment modalities and improve the survival outcome in patients.

Bacterial strain and cultured conditions
Fusobacterium nucleatum strain ATCC 25586 32 , were purchased from China General Microbial Species Preservation Center (CGMCC). Fn was cultured overnight under anaerobic conditions at 37℃ in brain heart infusion broth supplemented with Vitamin K1, hemin, and L-Cysteine as described before 45 .

Mutation analysis
DNA isolation from the CRC cell lines SW620, SW480, SW48, Caco2, LOVO, HT29, and HCT116 were performed making use of the QiAmp DNA-Isolation Kit (Qiagen, Hilden, Germany) in accordance with the instructions. KRAS and PIK3CA polymerase chain reactions for KRAS exon-2,3 and PIK3CA exon-9,20 augmentation were managed employing the primers listed in the table (See supplementary materials). The products of PCR were submited to direct sequencing. The results of sequencing were compared to the KRAS and PIK3CA sequence stored in GenBank databases.

Cell viability assay
The viability assays were determined in Caco2 and HT29 cells. CRC cells were seeded in 96-well plates at 2×10 3 cells per well with 100 μL culture medium.
The first approach: 24 h after seeding, cells were treated with Fn at various multiplicity of infection (MOI) of 100:1, 500:1, 1000:1. Cells were treated with medium as a negative control. Cell absorbance was detected by spectrophotometer in Caco2 and HT29 cells at indicated time.
The second approach: 24 h after seeding, cells were treated with cetuximab (0.1, 1, 10, 100, 1000 μg/ml) with or without Fn at MOI (100:1) for 48h. Cells were treated with medium as a negative control. Cell absorbance was detected by spectrophotometer in Caco2 and HT29 cells at indicated time.

Cell proliferation assay
Proliferation assays were determined in Caco2 and HT29 cells. 1×10 5 cells were seeded in per well of 6-well plates supplemented with 4 ml culture medium per well. The medium was replaced the second day. Then, cells were treated with cetuximab (100μg/ml) and/or Fn at MOI 100:1. Cells were treated with medium as a negative control. The numbers of cell were counted in indicated time using a hemocytometer.
Each experiment was repeated three times.

Colony formation assay
Cells were seeded at 500 cells per well in six-well plates. 24h incubation, and then the cells treated with cetuximab (100 μg/ml) and/or Fn at MOI (100:1), and then continuously incubated in fresh medium. Cells were treated with medium as a negative control. After 14-days incubation, the cells were washed twice with phosphate buffer solution, fixed for 20 min with 10% paraformaldehyde, and stained for 20 min with 0.05% crystal violet. The visual colonies were calculated.

Cell Cycle Analysis
Caco2 and HT29 cells were seeded at a density of 1x 10 5 cells per well in 6-well plates.
8 24 h incubation, and then cells were treated with the same manipulation as mentioned (colony formation assay) for 48h. The adherent fractions of cells were trypsinized and fixed overnight with 100% ice-cold ethanol at -20℃. The cells were washed twice with cold PBS and stained with RNase a and PI solutions in the dark at 37 ℃ for 30 minutes.
The cell cycle distribution was analyzed by flow cytometry (BD Biosciences, San Jose, CA). Each experiment was repeated three times.

Cell apoptosis
Caco2 and HT29 cells were seeded at a density of 1x 10 5 cells per well in 6-well plates.
24 h incubation, and then the cells were treated with cetuximab (100μg/ml) and/or Fn at MOI 100:1. Cells were treated with medium as a negative control. The extent of apoptosis was detected via Annexin V-FITC/PI apoptosis detection kit in line with the manufacturer's instruction. The cells were immediately analyzed by flow cytometry (BD Biosciences, San Jose, CA). Each experiment was repeated three times.

Tumor Xenograft Study
Four-week-old male athymic nude mice were raised under specific pathogen-free conditions provided with food and water provided ad will. Each mouse was subcutaneously injected with 1×10 7 Caco2 cells in the right axilla to set up a xenograft model. After 7 days of inoculation, the mice were randomly divided into 4 groups.
There were four groups: i) saline (Control); ii) Fn bacteria solution; iii) cetuximab (1 mg/mice); iv) cetuximab (1 mg/mice) and Fn bacteria solution. Fn (1x10 7 Clone forming unit (CFU)) was given by multipoint intratumoral injection 42 , twice a week for three weeks. Cetuximab treatment was given at a dose of 1 mg/mouse, intraperitoneal injection, twice a week for three weeks 46 .
Tumor volume and weight were calculated performed as previously described 42 . The research procedures were approved by the Institutional Animal Care and Use Committee of Tongji Hospital, Tongji Medicine, Huazhong University of Science and Technology.

Statistical Analysis
All results are expressed as means ± standard errors of means (SEM) of three independent experiments. Statistical analysis was carried out employing Graph-Pad Prism 6.0 (GraphPad Software, SanDiego, CA, USA). Statistical significance was reported if the value was <0.05 employing an unpaired Student's -test.

Statement
The research procedures were approved by the Institutional Animal Care and Use Committee of Tongji Hospital, Tongji Medicine, Huazhong University of Science and Technology, including any relevant details; We confirmed that all experiments were performed in accordance with the relevant guidelines and regulations.

Selection of multiplicity of infection (MOI) and IC50
To understand whether Fn participates in resistance to cetuximab, Fn was co-cultured with Caco2 and HT29 CRC cell lines using varied MOI. Our analysis suggested that Fn (MOI = 100) did not affect the proliferation of these cells ( Figure 1A and 1B).
To quantitatively assess the effect of Fn on resistance to cetuximab , we compared cetuximab-induced cell viability among the parental Caco2 and HT29 cells, and the Fnco-cultured Caco2 and HT29 cells, in the presence of different concentrations of cetuximab. Our analysis suggested the half maximal inhibitory concentration (IC50) to be 93.47 ± 6.978 μg/ml and 146.4 ± 9.430 μg/ml in the parental Caco2 cells and Fn-cocultured Caco2 cells, respectively ( Figure 1C p=0.0107). Whereas, the IC50 for parental HT29 cells and the Fn-co-cultured parental HT29 cells was 107.8 ± 10.91 μg/ml and 183.7 ± 12.28 μg/ml in the, respectively ( Figure 1D p=0.0099). Moreover, these differences in IC50 values for parental and Fn-co-cultured cells were statistically significant, indicating that Fn (MOI= 100) conferred cetuximab resistance in the Caco2 and HT29 CRC cells.

Fn induces cetuximab resistance in CRC cells in vitro and in vivo
While the Caco2 and HT29 cells are relatively sensitive to cetuximab treatment 46,48,49 , we confirmed their KRAS/PIK3CA (WT) status. Consistent with our hypothesis, cetuximab inhibited cell proliferation in parental Caco2 and HT29 cells but not in cells co-cultured with Fn (Figure 2A and 2B). Therefore, these data indicate that Fn inhibits the effects of cetuximab on cell proliferation in Caco2 and HT29 cells.
Next, we observed that Fn (MOI = 100:1) had no effect on the viability of Caco2 and HT29 cells than the untreated cells, but, it significantly reduced the effect of cetuximab on viability of these cells ( Figures 2C p=0.0001 and 2D p=0.0065). Further, Fn (MOI = 100) did not affect the cellular transformation of Caco2 and HT29 cells than the untreated cells ( Figure 2E p=0.1096 and 2F p=0.7899). While cetuximab reduced the cellular transformation in these cells, co-culturing them with Fn significantly reduced the effect induced by cetuximab ( Figure 2E p=0.0064 and 2F p=0.0214). Furthermore, Fn (MOI = 100:1) had no effect on apoptosis in Caco2 and HT29 cells than the untreated cells. However, while cetuximab induced apoptosis in Caco2 and HT29 cells, co-culturing them with Fn significantly reduced the effects induced by cetuximab ( Figure 2G p=0.0005 and 2H p=0.0005). Therefore, these results indicate that Fn inhibited the effects of cetuximab on cell viability, transformation, and apoptosis in Caco2 and HT29 cells.
Additionally, the Caco2 cells were inoculated into nude mice, and the mice were treated with cetuximab and/or Fn. Cetuximab significantly inhibited the tumor growth in vivo. While Fn alone had no effect on the growth of tumor cells, it inhibited the effect of cetuximab on tumor growth ( Figure 3A-3E, Figure 3D p=0.0044 and Figure 3E p=0.0008). Therefore, these results indicate that Fn confers resistance to reduction in tumor growth induced by cetuximab.
Taken together, the observations validate that Fn inhibits the effects of cetuximab on the CRC cell proliferation, viability, transformation, apoptosis, and tumor growth in vivo.

Fn promotes activation of the PI3K/AKT and JAK/STAT3 signaling pathways
Next, we set to understand the mechanism of how Fn influences resistance to cetuximab.
First, we found that Fn did not affect the cell cycle in Caco2 and HT29 cells after coculturing for 48 hours (Figure 4A and 4B). Here, we found Fn to promote resistance of CRC to cetuximab, in vitro and in vivo(Cetuximab was added and the results showed that it was present to increase apoptosis and reduced cell proliferation, viability and transformation even in cells cocultured with Fn. These results indicated that Fn was found to cause partial resistance to cetuximab.). The Fn-mediated activation of PI3K/AKT and JAK/STAT3 signaling pathways indicates the survival adaptation of cells regardless of inhibition of the EGFR.
While our analysis indicates induction of P-AKT and P-STAT3 to promote cetuximab resistance in Fn-infected CRC cells, we cannot neglect the involvement of mechanisms other than the PI3K/AKT and JAK/STAT3 signaling pathways. For example, Fn can enhance cancer formation in the CRC cells via interaction with E-cadherin and direct activation of WNT signaling pathway 43 , which has been implicated in resistance to cetuximab 46 . Moreover, the detailed mechanism by which Fn-mediated activation of the PI3K/AKT and JAK/STAT3 signaling pathway that confers cetuximab resistance remains to be studied. Therefore, strategies need to be devised to unravel the crosstalk between cetuximab and Fn. Nevertheless, the relation between the intestinal microbiota and therapeutic outcome appears complicated, multifactorial, and of special context. Finally, we anticipate that the tumor microenvironment (TME)-targeted therapy should enhance the effects of anti-EGFR mAbs, and improve patient outcome.    n.s., not significant, *P < 0.05, **P < 0.01, ***P < 0.001, by unpaired Student t test..
(C) Immunoblots of Caco2 cell lysates from Caco2 treated with cetuximab followed infected with Fn at indicated time. GAPDH served as the loading control.
(D) Immunoblots of HT29 cell lysates from HT29 treated with cetuximab followed infected with Fn at indicated time. GAPDH served as the loading control.