Post-COVID-19 Effects on Chronic Gastritis and Gastric Cellular and Molecular Characteristics in Male Mice.

Results

SARS-CoV-2 Does Not Disseminate to the Stomach but Affects the Immune Cell Composition
In total, 106 plaque-forming units of SARS-CoV-2 were inhaled into K18-hACE2 mice, which were euthanized 2, 5, 7, and 14 days and 4 weeks after viral infection (Figure 1A). In line with our previous result,3 SARS-CoV-2 immediately spread throughout the lung parenchyma after infection, concomitant with acute respiratory diseases, for 7 days (Figure 1A and B). Although SARS-CoV-2 was eliminated from the lungs 14 days and 4 weeks after viral infection, fibrosis and mild inflammatory foci remained (Figure 1B). In contrast, SARS-CoV-2 infection did not induce disease progression in the stomach or disrupt glandular structures (Figure 1C). Even at 7 days post-infection (dpi), when the viral concentration and acute inflammation peaked in the lungs, pathologic signs and SARS-CoV-2 were devoid in the stomach (Figure 1B and C).
In contrast to severe COVID-19, which leads to death, mild clinical signs and a convalescent phase have been identified in mice that inhaled SARS-CoV-2. A slight reduction in body weight was observed in the K18-hACE2 mice at 7 dpi (Figure 1D). In addition, mild lymphopenia and neutrophilia were recapitulated at 7 dpi (Figure 1E). These clinical signs resolved rapidly at 14 dpi. Although body weight and blood immune cell changes were transient, immune cell changes in stomach tissues persisted for more than 2 weeks (Figure 1F and Figure 2). In particular, macrophage and mast cell numbers rapidly increased in the stomach tissue after viral infection, and these increases were maintained to some extent even after 4 weeks of infection (Figure 1F and Figure 2). Additionally, an increase in neutrophils was observed in the convalescence phase at 4 weeks post-infection (Figure 1F and Figure 2). However, SARS-CoV-2 did not alter CD4 T cell or eosinophil numbers in the stomach (Figure 1F and Figure 2).
Next, we analyzed cytokine expression, which could affect immune cells at 7 dpi (Figure 1G and Figure 2). Unexpectedly, most blood serum cytokine levels were unchanged or reduced in SARS-CoV-2-infected mice compared with those in uninfected mice (Figure 3A), suggesting that cytokine changes subsided at 7 dpi and that immune cells in the stomach could be induced at an early stage. Only CXCL13 expression was increased in response to SARS-CoV-2 infection at 7 dpi (Figure 1H). Overall, SARS-CoV-2 infection did not promote gastric disease but changed the stomach immune cell composition.

SARS-CoV-2 Infection Induces Dynamic Gene Expression Transitions in the Stomach
To analyze the effect of SARS-CoV-2 in more detail, we evaluated gene expression in uninfected and SARS-CoV-2-infected mice. Stomach specimens were obtained at 7 dpi, and RNA sequencing was performed (Figure 4A). Although there were no significant pathologic changes, the heat map and multi-dimensional scaling plot indicated that viral infection caused a profound change in RNA expression (Figure 4B and C). Levels of more than 1100 genes differed significantly between SARS-CoV-2-infected and uninfected stomach (Figure 4C). Additionally, similar changes in immune subsets were additionally observed in the Cibersort analysis based on transcriptomic data (Figure 3B).
Notably, the expression of several genes related to metaplasia, including Slc7all, Alpi, Dmp1, Cftr, Cldn4, Cdx2, and Tff3, was higher in SARS-CoV-2-infected stomach than in uninfected stomach (Figure 4D and E). In line with cytokine array results, Cxcl13 was one of the most upregulated genes in SARS-CoV-2-infected stomach compared with expression in uninfected stomach (Figure 1H and Figure 4D and E). In addition, other immune-associated genes, such as Arg1, Cxcl1, Il1rn, and Il1rl1 (ST2), were significantly different between the groups (Figure 4D). Among the downregulated genes following SARS-CoV-2 infection, growth factors, ECL cell markers, hormones, and metabolism-associated genes were identified. Gene set enrichment analysis (GSEA) of differentially expressed genes (DEGs) showed that gene sets related to metaplasia and early gastric cancer were enriched in SARS-CoV-2-infected stomach compared with levels in uninfected stomach (Figure 4F). To verify transcriptome results, we performed immunohistochemistry (IHC) for CDX2 and CLDN4, which are well-defined gastric carcinogenesis markers. Consistently, CDX2+ and CLDN4+ cell emergence was distinct in the corpus-antrum junctional cells of SARS-CoV-2-infected mice (Figure 4G).

SARS-CoV-2 Exerts a Negative Synergistic Effect With Hpylori-induced Gastritis
Based on the transcriptomic data, we investigated whether SARS-CoV-2 exerted a synergistic effect on gastric disease. Hence, we infected K18-hACE2 mice with SARS-CoV-2 before (pre-COVID) and after (post-COVID) H
pylori inoculation, and the mice were euthanized 20 weeks after H
pylori infection (Figure 5A). Immunostaining showed that spiral-shaped H
pylori successfully colonized the stomach 4 weeks after inoculation (Figure 5B). Twenty weeks after H
pylori infection, spontaneous chronic gastritis, accompanied by immune cell accumulation and parietal cell loss, was prominent in the corpus-antrum junction (Figure 5C). However, gastritis was more severe in both the pre-COVID and post-COVID groups than in H
pylori-infected mice (Figure 5C). In particular, several indicators involving a thickened mucosa, immune cell infiltration in the submucosa, and signet ring-like cell emergence were more prominent in post-COVID mice than in control mice (Figure 5C and D). Although the disease was more severe in pre-COVID mice than in H
pylori-infected mice, the worst outcomes were observed in post-COVID mice among the 3 models (Figure 5A, C, and D). However, in contrast to that in pre-COVID mice, there was no body weight loss due to SARS-CoV-2 infection in post-COVID mice (Figure 5E).
To determine whether SARS-CoV-2 affects H
pylori colonization, we compared the bacterial contents of the stomach specimens between mice infected with bacteria and those co-infected with bacteria and viruses. SARS-CoV-2 infection did not change this bacterial content, and H
pylori successfully colonized the stomach, as seen in control mice (Figure 5F). These results suggest that the increased gastritis severity in pre-COVID and post-COVID groups are not associated with the number of H
pylori residing in the gastric gland.

Gastric Lineage Changes in SARS-CoV-2- and Hpylori-infected Mice
To evaluate gastric disease severity in pre-COVID and post-COVID mice, we first conducted IHC for corpus lineage and proliferating cells affected by H
pylori infection. As expected, 20 weeks of H
pylori infection in the stomach elicited distinct histologic changes, including oxyntic atrophy (parietal cell loss), chief cell marker loss (MIST1), foveolar cell hyperplasia, and an increase in proliferating cells, in K18-hACE2 mice (Figure 6A–E). In post-COVID mice, ATP4A+ parietal cell numbers were significantly lower than those in H
pylori-infected mice, whereas there was no obvious change in pre-COVID mice (Figure 6A and B). Notably, parietal cell loss, which is primarily induced during chronic inflammation, differed depending on the timing of SARS-CoV-2 infection (Figure 6B). Meanwhile, MIST1+ cell ablation and an increase in MUC5AC+ foveolar cells were prominent in both post- and pre-COVID mice compared with observations in H
pylori mice (Figure 6A, C, and D). Overall, the degree of changes in corpus lineages was more radical in post-COVID mice than in pre-COVID mice. In contrast, SARS-CoV-2 infection caused a similar increase in MKI67+ proliferating cells in H
pylori-infected mice, regardless of the viral infection time (Figure 6A and E).

SARS-CoV-2 Strengthens the Chronic Inflammatory Response in Hpylori-infected Mice
We then conducted transcriptomics using stomach specimens from post-COVID mice to explore the altered molecular profiles. This analysis of H
pylori and post-COVID groups revealed that immune-related gene sets were altered and among the top 10 gene ontology terms (Figure 7A). This was consistent with the histologic features of post-COVID mice, characterized by more immune cells in the gland base and submucosal region (Figure 7C). Among the immune-related pathways, IL6/JAK_STAT3, IFNg, and IL2_STAT5 signaling, which can lead to chronic responses, were significantly and positively enriched in post-COVID mice compared with levels in control mice (Figure 8A, C, and D). In addition, the immunological gene set associated with preneoplastic metaplasia was also changed in post-COVID mice compared with levels in H
pylori-infected mice (Figure 7B).
Because the immune response following H
pylori infection can stimulate different inflammatory cells related to disease progression, we investigated which immune subsets were affected by SARS-CoV-2 in the H
pylori-infected gastritis model. As expected, H
pylori infection caused a dramatic increase in innate and adaptive immune cell numbers (Figure 7C–H, and Figure 9A and B). Notably, increased patterns of macrophages and mast cells, modulated upon infection with SARS-CoV-2 alone, were maintained in pre-and post-COVID mice (Figure 1F and Figure 7C, E, and H). In addition, CD4+ T cell numbers were higher in post-COVID mice than in H
pylori-infected mice (Figure 7C and D). However, SARS-CoV-2 did not affect neutrophils or eosinophils in the H
pylori-induced gastritis model (Figure 7F, G, and Figure 9A and B). Together, SARS-CoV-2 infection elicits a strengthened inflammatory response in H
pylori-infected mice.

Aggravated premalignant Condition in Pre- and Post-COVID Mice
During gastric carcinogenesis, H
pylori infection promotes stepwise progression involving oxyntic atrophy, chronic inflammation, preneoplastic metaplasia, and eventually cancer. To determine whether a strengthened immune response can contribute to disease progression in SARS-CoV-2-co-infected mice, we assessed premalignant conditions in H
pylori-infected pre- and post-COVID mice based on CD44v9 and GSII or MKI67, which are metaplasia and proliferation markers, respectively (Figure 10A). Immunofluorescence images indicated that CD44v9+/GSII+ metaplastic cells emerged from the basal region of the gland after H
pylori infection, and metaplastic cell numbers were remarkably elevated in post-COVID mice compared with those in H
pylori-infected mice (Figure 10A and B). Notably, preneoplastic cells expanded into the entire gland in post-COVID mice (Figure 10A). Albeit not to the same extent as that in post-COVID mice, metaplastic cell numbers in pre-COVID mice were slightly increased (Figure 10A and B). Consistently, GSII+/MKI67+ proliferating metaplastic cells were also increased in pre- and post-COVID mice compared with numbers in H
pylori-infected controls (Figure 10A and C).
Notably, disease progression-associated gene sets involving the epithelial-mesenchymal transition were significantly enriched in post-COVID mice compared with levels in H
pylori-infected controls (Figure 8B and Figure 10D). In addition, GSEA results showed positive enrichment scores in pathways associated with E2F targets and KRAS signaling in post-COVID mice compared with levels in H
pylori-infected controls, supporting the increase in metaplasia and proliferating cells (Figure 8B). In line with the transcriptomic analysis results, levels of p-STAT3 and p-ERK, which are pivotal for disease progression,33, 34, 35 were evident and elevated in the gastric lesions of co-infected mice compared with those of H
pylori-infected mice (Figure 8C and D). Previously, we described the role of WFDC2, which promotes preneoplastic metaplasia and serves as an effective prognostic marker and putative target in gastric cancer.19 Hence, we explored changes in Wfdc2 expression in K18-hACE2 mice. Co-infection with SARS-CoV-2 and H
pylori resulted in increased expression in the corpus-antrum junction, where gastritis and chronic inflammation were observed (Figure 10E and F). Collectively, our findings suggest that SARS-CoV-2 infection strengthens the immune response and promotes adverse premalignant conditions in response to H
pylori infection.

Discussion
Long COVID comprises a wide range of new, returning, or ongoing health problems experienced after infection with the COVID-19-causing virus. Emerging research suggests that the GI tract exhibits morbidities after the acute COVID-19 phase8,9,11 and that gut microbiomes might undergo crosstalk with SARS-CoV-2.36,37 Despite advances in our understanding of the post-acute sequelae of COVID-19, limited studies have examined the stomach. In this study, we performed RNA sequencing of stomach specimens from K18-hACE mice after SARS-CoV-2 inhalation. Transcriptomic data indicated that more than 1100 genes were significantly altered in the stomach during the acute COVID-19 phase. The most altered immune cells and genes were associated with the development of gastric lesions. Most importantly, SARS-CoV-2 and H
pylori co-infection exacerbated disease progression.
The relationship between non-inflammatory GI symptoms and COVID-19 remains controversial.26,27,38,39 However, here, SARS-CoV-2 infection exerted a negative synergistic effect on inflammation-prone disease in the stomach. In this regard, there are some possibilities that can be suggested. First, unlike previous studies, which have suggested that the microbiome can reduce H
pylori colonization,40 we observed that H
pylori colonization was not affected by SARS-CoV-2 co-infection. Second, although the co-infection of gastric epithelial cells by H
pylori and SARS-CoV-2 could theoretically lead to adverse outcomes, our data showed that SARS-CoV-2 does not directly infect gastric cells. Instead, it indirectly alters the epithelial cell molecular characteristics associated with pre-cancer progression. Third, whereas COVID-19 may influence translocation of the H
pylori CagA oncoprotein via the Type 4 secretion system (T4SS), the non-functional T4SS of the H
pylori SS1 strain used in our study suggests that this is unlikely in our model. However, macrophages and mast cells, which are distinct components of H
pylori-induced gastritis,22,41 were more abundant in the stomach following SARS-CoV-2 infection. In conclusion, we suggest that SARS-CoV-2 infection elicits a synergistic effect with H
pylori by promoting changes in the molecular characteristics of gastric cells and leading to an increase in the deposition of specific immune subsets in the stomach.
Most patients with COVID-19 improve within a few days to a few weeks after viral infection; therefore, at least 4 weeks after infection represents the time at which long COVID could be first identified. For extended long COVID-19 research on the stomach, we utilized the K18-hACE2 mice used in our previous research on the lungs.3,42 To avoid mortality and mimic the convalescence phase in human patients, we subjected animals to SARS-CoV-2 inhalation instead of using the previous intranasal method. Notably, we also divided the mice into pre-COVID and post-COVID groups, since human patients are rarely simultaneously infected with 2 different microorganisms. Recovery traits marked by a reduction in inflammatory foci were identified in the lungs at 4 weeks post-SARS-CoV-2 infection, but changes in immune subsets in the stomach remained even after viral particle removal. This suggests that the effects of SARS-CoV-2 infection on the stomach persist to some extent beyond the post-acute COVID-19 phase. Supporting this, pre-COVID mice exhibited significant alterations, including the loss of parietal cells, loss of chief cell markers, and an increase in metaplastic cells, compared with observations in H
pylori-infected control mice. Likewise, the immune subsets affected by SARS-CoV-2 infection were still enriched in the co-infected mice (Figure 1F and Figure 7). Because macrophages and mast cells can promote gastric inflammation and initiate gastric tumorigenesis, these pre-established subsets might accelerate gastric disease in co-infected mice.22,43,44
Over one-half of the global population is infected with H
pylori, meaning that not everyone with an H
pylori infection develops gastric cancer. Nevertheless, the importance of H
pylori in gastric carcinogenesis cannot be ignored.45,46 Various factors, including genetic diversity, dietary habits, and tobacco use, seem to define H
pylori pathogenicity. Also, these etiological factors appear to be associated with other microbial communities. Sgouras et al revealed that Lactobacillus spp. suppress chronic gastritis by inhibiting H
pylori colonization.40 EBV positivity is also associated with H
pylori-induced gastritis severity and gastric cancer development.31,32,47 Pryia Saju et al found that EBV can suppress the host response to H
pylori by mediating SHP-1 methylation. However, here, SARS-CoV-2 infection did not affect H
pylori colonization. Instead, SARS-CoV-2 induced profound alterations in the molecular profile related to gastric carcinogenesis. Previous results suggest that the most upregulated (Slc7a11, Dmp1, Cxcl13, Cftr, Cldn4) and downregulated gene sets (Chia1) in our transcriptomic data contribute to preneoplastic metaplasia, intestinalization, and cancer development, serving as prognostic markers for gastric cancer.16,48, 49, 50, 51 Especially, CDX2+ cell emergence in the corpus-antrum junction was noteworthy, because CDX2 was suggested to be a master transcription factor involved in pre-cancer development by enhancing intestinal markers, such as Muc2, Tff3, and Alpi, and subsequently promoting this disease with additional mutations.21,52 Hence, CDX2 likely plays a crucial role in accelerating premalignancy in these mice. However, whether CDX2 is the most crucial molecule involved in the dynamic and post-acute COVID-19-associated changes in the stomach is unclear. We assume that a robust inflammatory response derived from the primary infection site can initially and indirectly affect stomach cells.
This study has some limitations. Owing to the high prevalence of H
pylori and SARS-CoV-2 in South Korea, specimens without a history of infectious disease are extremely rare. Therefore, we were unable to provide comparative transcriptomic data for humans. In addition, we did not observe a direct interaction between gastric cancer and COVID-19. Although many researchers have attempted to establish gastric cancer-model mice, an appropriate cancer model resembling that of human patients has not been developed. Choi et al established transgenic mice expressing the Mist1-CreERT2-inducible mutated KRAS (G12D).34 They generated transgenic mice based on human data, and the mice developed dysplasia that effectively mimicked the molecular features shown in human patients,20 suggesting that KRAS has an oncogenic function during gastric carcinogenesis. However, these mice did not develop gastric adenocarcinoma. Furthermore, model mice must express K18-hACE2 to conduct COVID-19 research; however, we purchased all hACE2 male mice from the Jackson laboratory. Owing to issues related to interest and the lack of a gastric cancer model, we could not provide direct evidence for this and only could use male mice. However, we found that levels of KRAS, epithelial-mesenchymal transition, and E2F target gene sets were upregulated in the mice. Along with transcriptome changes, the number of metaplastic and proliferating cells was increased in post-COVID mice. Consistent with this, WFDC2, which was previously suggested to be related to gastric cancer prognosis, was prominently expressed in post-COVID mice.19
When looking at the overall phenomenon, stomach cells exposed to SARS-CoV-2 resemble a “loaded gun.” Protein levels do not seem to perfectly mimic RNA levels, which are remarkably elevated after viral infection. Nevertheless, cells positive for CDX2 and CLDN4 proteins were enriched in the corpus-antrum junction, where chronic gastritis primarily occurred. The potential for deterioration of stomach cells post-COVID seems to increase markedly when they encounter other triggers, such as H
pylori. Collectively, our findings provide a putative pathogenic mechanism in the stomach in the post-COVID era and imply that co-infection with SARS-CoV-2 and the pathogenic microbiome might be related to gastric disease progression and prognosis. Thus, this study provides significant evidence regarding the risk of severe chronic gastritis in the post-COVID-19 era.

Methods

Animals
Male K18-hACE mice (Tg [K18-ACE2] 2 Prlmn/J) with a C57BL6 background were purchased from Jackson Laboratory. Because only male mice are commercially available, all experiments were conducted using male mice. Animal experiments were performed at the Animal Biosafety Level 3 (ABL3) facility in accordance with the Public Health Service Policy on Human Care and Use of Laboratory Animals. Animals were randomly divided into 3 groups, using a computer-based random generator. All procedures were approved by the Institutional Animal Care and Use Committee (IACUC: 2020-0216) of Yonsei University College of Medicine and accredited by the Association for Assessment and Accreditation of Laboratory Animal Care International (#001071). In accordance with safety guidelines, all animal experiments were performed in the ABL3 facility, and all samples were allowed to leave the ABL3 facility following DNA and RNA protein extraction or fixation procedures to avoid SARS-CoV-2 activation. For quantification and obtain significance, we used at least 3 mice for each experiment, and there were no inclusion and exclusion criteria in this study.

Virus Production
Vero cells, a Vero African green monkey kidney cell line (KCLB 10081), were used to produce SARS-CoV-2. Culturing SARS-CoV-2 was obtained from the National Culture Collection for Pathogens of Osong, Korea (NCCP 43326, S type), and virus titer measurements were performed in accordance with our previous study.3 In brief, SARS-CoV-2 was cultured in Vero cells until a cytopathic effect was observed in >80% of cells. The supernatants were centrifuged at 380 × g for 15 minutes to remove cell debris, aliquoted, and stored at −80 °C. SARS-CoV-2 stocks and infected tissues were titrated using plaque assays. Vero cells were seeded in a 6-well plate the day before the assay. Serially diluted supernatants or homogenized tissues were then added, and the mixture was incubated for 1 hour with gentle agitation every 15 minutes. The cells were overlaid with Dulbecco’s modified Eagle medium containing 1% SeaPlaque agarose (Lonza), 2% fetal bovine serum (FBS), 100 units/mL penicillin, and 100 μg/mL streptomycin. After 3 days, the plaques were cleared, and the cells were fixed with 4% paraformaldehyde and stained with a 0.5% crystal violet–20% methanol solution. The enumerated plaques were used to determine the viral titers.

Mouse Infection With SARS-CoV-2 and Hpylori
For SARS-CoV-2 infection, a modified version of the Kent Scientific Chamber Nebulizer Delivery System was used to induce infection via inhalation. K18-hACE2 mice were exposed to 1 × 106 SARS-CoV-2 particles in an isolated chamber for 30 minutes. Uninfected control mice were treated with an equal volume of phosphate-buffered saline (PBS). For infection, H
pylori was cultured on antibiotic-supplemented Brucella-broth containing 10% FBS in a microaerobic jar with a Gas-Pack at 37 °C and a 5% CO2 incubator. The mice were starved for 12 hours before the first day of H
pylori infection, and 1 × 108 CFUs of H
pylori were orally inoculated 3 times at 2-day intervals. For the post-COVID model, 9-week-old mice were initially infected with H
pylori. After 4 weeks, SARS-CoV-2 was administered via inhalation. In the pre-COVID model, animals were subjected to SARS-CoV-2 inhalation 4 weeks before H
pylori infection. The total H
pylori infection period was 20 weeks. For blinding test, the primary investigator, who was the only person aware of the group allocation, infected virus and bacteria to randomized groups, and experiments involving measurement of body weight changes, immune cell changes, and histologic observation were conducted by other investigators who did not know the allocation.

Histopathologic Analysis
For histopathologic analysis, the mice were euthanized at each time point using a CO2 chamber, and collected tissues were fixed in 10% neutral buffered formalin (Sigma) for 24 hours and embedded in paraffin. Fixed samples were cut into 4 μm-wide slides using a microtome (Leica) for subsequent hematoxylin and eosin (H&E) staining and IHC. For H&E staining, sections were de-paraffinized through 3 immersions in xylene, followed by rehydration in a series of ethanol concentrations (100%, 95%, and 70%). The slides were then stained with 0.1% Mayer’s haematoxylin (Agilent) for 10 minutes, followed by immersion in 0.5% Eosin Y (Sigma) solution. The slides were then rinsed in distilled water until eosin streaking stopped and then dehydrated in ascending concentrations of ethanol (50%, 70%, 95%, and 100%) for 1 minute each. The slides were covered with a mounting solution (Thermo) and examined under a light microscope (Olympus BX43). Histopathologic analysis was performed by an experienced animal pathologist (K.T.N.).

In Situ Hybridization
An hACE2 and Wfdc2 RNA probe and RNAscope 2.5 HD Red Assay were purchased from ACD (Bio-Techne). The procedure was performed according to the manufacturer’s instructions. Briefly, paraffin-embedded slides were de-paraffinized in xylene, with 2 rounds of dehydration in 100% ethanol. After air-drying, the slides were treated with H2O2, dipped in boiling antigen retrieval buffer, and treated with protease K for 30 min. The RNA probe was incubated for 2 hours, and RNA signaling was amplified using the amplifying reagent from ACD, followed by detection with Fast Red reagent (ACD).

IHC
Paraffin-embedded samples were cut into 4-μm sections and de-paraffinized through 3 immersions in xylene, followed by rehydration using a descending graded series of ethanol. Antigen retrieval was performed using an antigen retrieval solution (pH 6.0; Agilent), with antigens retrieved through pressing and boiling in a high-pressure cooker for 15 minutes. After antigen retrieval, sections were cooled on ice for 1 hour and washed twice with Dulbecco’s PBS. Subsequently, sections were immersed in 3% H2O2 for 30 minutes to block endogenous peroxidase activity. After 2 additional washes with PBS, sections were incubated in a protein-blocking solution (Agilent) for 2 hours at room temperature in a humidity chamber. For mouse or rat primary antibodies, the vector M.O.M kit (Vector) was used before protein blocking. The slides were incubated with the primary antibodies overnight at 4 °C. The slides were then incubated with a horseradish peroxidase (HRP)-conjugated secondary antibody (Agilent) for 15 minutes or biotinylated anti-rat IgG (Vector) for 30 minutes, followed by incubation with ABC reagent (Vector) for 30 minutes at room temperature (RT). For the development of the HRP-labelled antibody on the sections, DAB (Agilent) was diluted and placed on the sections for the appropriate period to detect the signal. Mayer’s hematoxylin (Agilent) was used to counterstain the nuclei. After counterstaining, washing and dehydration steps were performed, and the slides were covered with a mounting solution (Thermo). For immunofluorescence staining, primary antibodies were visualized using secondary antibodies conjugated to Alexa488, Cy3, or Cy5 fluorophores. The used antibodies are listed in Table 1. Immunofluorescence images were acquired using 2 different microscopy systems, an EVOS-FL fluorescence microscope, and an LSM980 confocal microscope.

Hematologic Analysis
Peripheral blood samples were obtained from the hearts of euthanized mice using a 1-mL syringe. The collected blood was transferred into 1.5-mL microtubes supplemented with 20 μL of 0.5 M ethylenediaminetetraacetic acid to prevent clotting. A comprehensive blood count was performed using a hematology analyzer (BC-5000, Mindray Global).

Cytokine Array
Mouse serum cytokine and chemokine levels were measured using a mouse cytokine array kit following the manufacturer’s protocol. Mouse blood samples were centrifuged for 15 minutes at 2000 × g. The mean pixel intensity of each spot was quantified using ImageJ Fiji and compared with the background intensity. Thereafter, the mean intensity per duplicate was calculated and adjusted for differences in input amounts.

Transcriptomic and Bioinformatic Analyses
For the transcriptomic analysis, the same regions of the stomach were isolated using a biopsy punch and incubated in RNAlater to prevent RNA degradation. Total RNA was extracted using the TRIzol method, and RNA sequencing was conducted by Macrogen, Inc. Complete linkage and Euclidean distance, as similarity measures, were used for hierarchical clustering to analyze DEG sets. Gene enrichment, functional annotation, and pathway analyses of the significant genes were performed using gProfiler. GSEA (GSEA v4.0.3) was performed according to the manufacturer’s guidelines using the expression dataset and gene set files. The predictive analysis of the immune composition of the stomach was performed using CIBERSORTx, which is an analytical tool developed by Stanford University (https://cibersortx.stanford.edu/index.php).

RNA Extraction and Reverse Transcription-quantitative Polymerase Chain Reaction
Total RNA was extracted using TRIzol reagent (Invitrogen) according to the manufacturer’s protocol. cDNA was synthetised from 1 μg RNA samples using the ImProm-II reverse transcription system (Promega). Subsequent qPCR was performed with SYBR Green (Takara) using specific primers designed for the target gene. Results are presented as relative expression levels or fold changes compared with the control group. The primers used are listed in Table 2.

Bacteria Colonization Assay
To evaluate colonization with a SARS-CoV-2 infection, H
pylori was inoculated into K18-hACE2 mice. In the experimental group, the SARS-CoV-2 challenge was performed after 4 weeks of H
pylori infection. At 6 and 8 weeks post-H
pylori infection, stomach specimens were obtained from the same region of K18-hACE2 mice, and total DNA was isolated using a universal genomic DNA extraction kit (Takara). For the quantitative analysis, reverse transcription-quantitative polymerase chain reaction (RT-qPCR) was performed on total stomach DNA using primers for the host-specific Gapdh gene and H
pylori-specific UreB gene (Table 2). To estimate H
pylori colony-forming units (CFUs) in stomach specimens, pure microbial DNA was extracted from 1 × 108 CFUs of H
pylori, and 10-fold serially diluted DNA samples were utilized as a standard control for UreB. Similarly, a known amount of pure mouse genomic DNA was obtained from germ-free mice and used as a control for host Gapdh. Using this semi-quantitative method, we predicted the number of H
pylori CFUs and amount of host genomic DNA in each stomach specimen. Finally, the total number of H
pylori cells was normalized to the host genetic content acquired via RT-qPCR.

Statistical Analysis
Statistical analyses was performed using GraphPad Prism software version 9.0. Differences demonstrating statistical significance were assessed using unpaired Student t-test and 1-way or 2-way analysis of variance (ANOVA) with multiple comparison tests. All data and graphs were presented as the mean ± standard error of the mean (SEM). Statistical significance was set at P < .05.