Fusobacterium nucleatum‐induced exosomal HOTTIP promotes gastric cancer progression through the microRNA‐885‐3p/EphB2 axis

Abstract Recent studies have reported that Fusobacterium nucleatum (Fn) is associated with gastric cancer (GC). Cancer‐derived exosomes contain key regulatory noncoding RNAs and are a crucial medium of intercellular communication. However, the function and regulatory mechanism of exosomes (Fn‐GCEx) secreted from Fn‐infected GC cells remains unclear. In this study, Fn‐GCEx enhanced the proliferation, migration, and invasion capacity of GC cells in vitro, as well as tumor growth and metastasis in vivo. HOTTIP was also upregulated in GC cells treated with Fn‐GCEx. Moreover, knockdown of HOTTIP weakened the effects of Fn‐GCEx in recipient GC cells. Mechanistically, HOTTIP promoted EphB2 expression by sponging microRNA (miR)‐885‐3p, thus activating the PI3K/AKT pathway in Fn‐GCEx treated GC cells. Overall, Fn infection induced the upregulation of exosomal HOTTIP from GC cells that subsequently promoted GC progression through the miR‐885‐3p/EphB2/PI3K/AKT axis. Herein, we identify a potential molecular pathway and therapeutic target for GC.

Fusobacterium nucleatum is a Gram-negative, nonspore-forming anaerobe bacterium; it exists as normal oral flora and is an opportunistic pathogen in periodontal diseases such as periodontitis. 3 Recent studies, including ours, detected Fn expression in GC patients. [4][5][6] Moreover, identification of Fn was demonstrated to be associated with significantly reduced overall survival rate in patients with diffuse Lauren type GC. 7 However, the mechanism through which Fn affects GC is yet to be explored.
Exosomes are extracellular vesicles with a diameter of 30-150 nm and are present in various fluids in the human system. 8 Exosomes have been demonstrated to be central to cellular communication and transport cargo to specific recipient cells, affecting tumor-related processes. [9][10][11] However, whether Fn infection with GC cells can promote the secretion of exosomes and the mechanism of exosomes in the pathogenesis of Fn-related GC remains uncertain.
In the present study, we hypothesized that Fn infection with GC

| Cell lines and bacterial strains
Human GC cell lines AGS, HGC-27, and MKN-28 and the normal human gastric epithelial cell line GES-1 were obtained from Type Culture Collection of the Chinese Academy of Sciences. The cells were cultured in RPMI-1640 (Gibco) or DMEM (Gibco) supplemented with 10% FBS (Gibco) with 5% CO 2 at 37°C. The two Fn strains (ATCC49256 and ATCC25586) were purchased from ATCC.
The strains were initially inoculated and cultured on Columbia blood agar in anaerobic conditions at 37°C.

| Cell transfection
The HOTTIP overexpression vector and the shRNA mediating HOTTIP knockdown were purchased from Hanbio Biotechnology.

| Isolation and identification of exosomes
Exosomes were isolated from the culture supernatant of GC cells by differential ultracentrifugation. Isolated exosome imaging was undertaken using transmission electron microscopy, and Nanoparticle tracking analysis was used to measure the size distribution and concentration of exosomes. Exosome biomarkers were detected by western blotting.

| RNA extraction and RT-qPCR assay
Total RNA was isolated from tissues and cells with TRIzol or TRIzol LS reagent (Invitrogen). For lncRNA and mRNA, reverse transcriptions were carried out using a Prime Script RT reagent kit (Vazyme).
For miRNA, reverse transcriptions were carried out using a commercial miRNA reverse transcription PCR kit (Accurate Biology). All qPCR analyses were carried out using SYBR Premix Ex Taq (Vazyme).
GAPDH or U6 were used as internal controls. The PCR primers used in this study were synthesized by BioSune Biotechnology and are listed in Table S1. The relative expression levels were calculated using the comparative 2 −ΔΔCt method.

| Microbial FISH analysis
Paraffin sections of GC tissue samples that were positive for Fn DNA by probe-based qPCR assay were selected for microbial FISH analysis. The sequence of the "universal bacterial" probe EUB338 was 5′-GCT GCC TCC CGT AGG AGT-3′. The Fn-targeting probe FUS664 sequence was 5′-CTT GTA GTT CCG C(C/T) TAC CTC-3′.

| Cell proliferation assay
An EdU assay was carried out using a Cell Light EdU DNA imaging kit Hoechst-stained cells (with blue fluorescence) was used to examine cell proliferation.

| Colony formation assay
Cells were seeded in a 6-well plate at a density of 500 cells per well and cultured at 37°C with 5% CO 2 . Two weeks later, the colony surface was stained with crystal violet and counted.

| Wound healing assay
Cells were plated in 6-well plates to grow at 90% confluence. A pipette tip was utilized to create a scratch on the cell layer surface.
The images of the remaining gaps were recorded at 0, 24, and 48 h.

| Tumor xenograft experiments
Four-week-old BALB/c nude male mice were maintained under pathogen-free conditions. For subcutaneous tumorigenesis, AGS and HGC-27 cells (7 × 10 6 ) were subcutaneously injected into the back of the mice (n = 6/group). Exosomes derived from Fn-infected or untreated GC cells were injected intratumorally every 3 days.
Tumor size was measured every 2 days and the tumor volume was calculated using the formula: 0.5 × length × width. 2 Finally, mice were killed, subcutaneous tumors were surgically removed and weighed, and immunohistochemical staining was carried out to detect Ki-67 expression. For tumor metastasis in vivo, AGS and HGC-27 cells (7 × 10 6 ) were injected into the spleen of the mice (n = 6/group), and exosomes were injected through the tail vein every 3 days.
After 7 weeks, liver tissue was removed, embedded in paraffin, and stained with H&E.

| Statistical analysis
Statistical analyses, such as Student's t-test, one-way ANOVA, and Pearson's correlation test, were undertaken using GraphPad Prism version 8.0.1 (GraphPad Software). A p value less than 0.05 was considered to be statistically significant.

| Exosomes derived from Fn-infected GC cells promote the proliferation, migration, and invasion of GC cells in vitro
As shown in Figure 1A, Fn has been demonstrated in GC cells after coculture, thus we constructed an Fn-infected GC cell model.

| Exosomes derived from Fn-infected GC cells promote tumor growth and metastasis in vivo
To further examine the tumor-promoting effect of Fn-GCEx, in vivo animal models were constructed ( Figure 2A). We observed that intratumoral multipoint injection of Fn-GCEx had significant positive effect on tumor growth ( Figure 2B). After 3 weeks, the mice were killed and tumors were removed ( Figure 2C). The final average tumor weight of the Fn-GCEx groups were significantly greater than that of GCEx group ( Figure 2D). The Ki-67 levels, indicating proliferation of tumor tissues in Fn-GCEx groups, were similarly much increased ( Figure 2E).
We further investigated the function of Fn-GCEx in tumor metastasis ( Figure 2F). More tumor nodules were present in the liver surface of mice in the Fn-GCEx groups ( Figure 2G) and H&E stained liver sections showed a larger area and number of tumor nodules in the two Fn-GCEx groups ( Figure 2H). Taken together, the results suggested that Fn-GCEx promoted tumor growth and metastasis in vivo.

| HOTTIP is increased in GC cells treated with Fn-GCEx
Previous cumulative studies have investigated the regulation of lncR-NAs in GC progression. 12,13 Similarly, in the present study, we used RNA sequencing to screen the related lncRNAs (GEO: GSE214817).
As shown in Figure 3A

| HOTTIP in Fn-GCEx promotes GC progression by sponging miRNA-885-3p
Long noncoding RNAs function as ceRNAs sponging miRNAs is one of the vital mechanisms in tumorigenesis. [14][15][16] The localization and abundance of HOTTIP in AGS and HGC-27 cells was confirmed to be primarily in the cytoplasm using FISH analysis ( Figure 4A).

| EphB2 is the target gene of miR-885-3p in Fn-GCEx treated GC cells
To identify the underlying mechanism of miR-885-3p in Fn-GCEx treated GC cells, starBase version 2.0 was used to predict its potential target genes among the upregulated mRNAs in Fn-GCEx treated AGS cells ( Figure 5A) (GEO: GSE214817). According to the prediction data, EphB2 was revealed to possess two separate binding sites, site 1 (nt 33-38) and site 2 in its 3′-UTRs ( Figure 5B). A dual-luciferase reporter assay was undertaken to confirm that

| EphB2 activates the PI3K/AKT pathway in Fn-GCEx treated GC cells
To further identify the potential mechanisms of EphB2 in Fn-GCEx treated GC cells, KEGG pathway enrichment analysis was carried out to interrogate the differentially expressed genes in Fn-GCEx treated AGS cells compared to in GCEx treated AGS cells ( Figure 7A). In doing so we observed that the PI3K/AKT pathway was significantly enriched.
To address the relationship between EphB2 and the PI3K/ AKT pathway in Fn-GCEx treated GC cells, we undertook western

| DISCUSS ION
Herein, we report for the first time that exosomes released from Fninfected GC cells can promote the progression of GC. Furthermore, the lncRNA HOTTIP was observed to be elevated in Fn-infected GC cell-derived exosomes and subsequently promoted the proliferation and metastasis of uninfected GC cells. We further reveal that EphB2 was the target gene in GC cells regulated by Fn through HOTTIP/ miR-885-3p, suggesting that EphB2 could be an efficacious therapeutic target for GC.
The study results of Hsieh et al. revealed an underrepresentation of HP among the gastric microbiota of the GC patients, whereas Fusobacterium and Clostridium were highly abundant, including Fn. 5 They hypothesized that these bacteria would opportunistically infect the gastric epithelium and colonize the gastric mucosa by replacing HP through microbial succession, and might participate in the occurrence and development of GC. In the present study, we observed that Fn could infect and enter GC cells, suggesting that Fn might exist in the stomach in this way and be involved in GC progression. A previous study had reported that Fn infection facilitated the metastasis of CRC by the exosomes carrying miR-1246/92b-3p/27a-3p and CXCL16 from CRC cells. 20 Our findings revealed that Fn-GCEx visibly enhanced tumor growth and metastasis through carrying HOTTIP.
Considering that the lncRNAs primarily function as ceRNAs competitively binding to miRNAs during the tumor development process, 16,21,22 we identified that miR-885-3p could be directly bound to HOTTIP by undertaking sequencing analysis of recipient GC cells and RNA pull-down assay. It was verified that HOTTIP could downregulate the expression of miR-885-3p and this suggested that Moreover, we identified that elevated EphB2 in GC cells after Fn-GCEx treatment was predicted to be the direct target of miR-885-3p. More importantly, distinct to HOTTIP, the EphB2 expression was upregulated in Fn-GCEx cocultured GC cells but was almost undetectable in Fn-GCEx. That suggested the expression change of EphB2 was not directly caused by Fn, but regulated by the exosomal HOTTIP after Fn-GCEx penetrated the uninfected GC cells. It is well described that EphB2 plays an essential role in the progression of various cancers. For instance, EphB2 regulates migration by inducing EMT in cervical cancer, 27 and high expression of EphB2 predicts poor overall survival and high mortality in lung adenocarcinoma. 28 In line with these studies, we found here that the interference of EphB2 inhibited the progression of GC. Interestingly, the negative regulatory effect of EphB2 silencing in GC could not be rescued by Fn-GCEx, indicating that EphB2 was regulated by the F I G U R E 8 Schematic model of the mechanism of the promotion of gastric cancer (GC) progression by exosomes derived from Fusobacterium nucleatum (Fn)-infected GC cells (Fn-GCEx). miR, microRNA. exosomal HOTTIP through miR-885-3p in recipient uninfected GC cells. Importantly, it has been reported previously that abnormal expression of EphB2 activated the PI3K/AKT signaling pathway. 17,29 In this present study, we showed that the PI3K-AKT signaling pathway was repressed after the silencing of EphB2. These results indicated that EphB2-targeted therapy could attenuate the malignant effects of Fn in GC and EphB2 inhibition could therefore be a potentially effective treatment strategy for GC.
We illustrate the possible mechanism of Fn-GCEx in promoting GC through a schematic model in Figure 8. In summary, our new mechanistic model shows that Fn-infected GC cells could increase the secretion of exosomal HOTTIP and promote the progression of GC in uninfected GC cells through the miR-885-3p/EphB2/PI3K/ AKT axis. Moreover, it has been reported that some markers, such as Fn DNA and HOTTIP, are increased in saliva and serum samples of GC patients. 6,30 This might provide a noninvasive method for GC detection.

ACK N OWLED G M ENTS
None.

CO N FLI C T O F I NTE R E S T S TATE M E NT
The authors declare no conflict of interest.

E TH I C S S TATEM ENTS
Approval of the research protocol by an Institutional Review Board: This study was reviewed and approved by the Ethics Committee of Qilu Hospital of Shandong University.