mRNA microarray profiling identifies a novel circulating HTRA2 for detection of gastric cancer

Abstract Background mRNAs have been shown to be critical biomarkers or therapeutic targets for human diseases. However, only a few of them have been studied as blood‐based biomarkers for gastric carcinoma (GC) detection. Methods mRNA expression profiles for GC were screened using plasma samples from 10 GC patients with different TNM stages and 5 healthy individuals as controls. One candidate tumor‐related mRNA named HTRA2 was then evaluated in GC samples with quantitative real‐time polymerase chain reaction (qRT‐PCR). TCGAportal, UALCAN, and TISCH database were used to explore the function of HTRA2 in GC. Finally, the effect generated by HTRA2 expression on cell proliferating, invading, and migrating processes was assessed in vitro with knockdown and over‐expression strategies. Results HTRA2 displayed noticeable increase inside GC plasma compared with control cases. Higher expression of HTRA2 displayed a correlation to higher clinicopathological stage and worse prognosis. HTRA2 knocking down down‐regulated GC cells' proliferating, invading, and migrating states, while HTRA2 over‐expression exerted the inconsistent influence. HTRA2 protein, which may interact with PINK1, PARL, and CYCS, was mainly located in the mitochondria of cells and primarily involved cellular response and metabolic signaling pathway. Immune factors may interact with HTRA2 in GC, and HTRA2 was found noticeably linked with immunosuppressor such as CD274, IDO1, and TIGIT. Conclusion One plasma HTRA2 can be an emerging diagnosis‐related biomarker to achieve GC detecting process, but the particular regulatory effect still needs to be further explored.


| INTRODUC TI ON
Gastric carcinoma (GC) is the fifth most common carcinoma worldwide and the 3rd major factor causing carcinoma-associated death globally. 1 Particularly, the incidence ratio of GC has been increasingly risen in Eastern Asia. It will conduct the metastasizing in nearby tissues and organs via lymph nodes and the generation of more carcinoma cells via blood 2 that metastasis and recurrence are appearing under large ratios. 3 expressing state decreases inside several tumors and increases inside other tumors, [5][6][7] demonstrating that HTRA2 expressing state changes following tumor category. Currently, information regarding circulating HTRA2's implication inside GC is rare, and the expressing state exhibited by the HTRA2 in GC and the relevant correlations to clinicopathology-related characteristics are not clear.
Here, mRNA expression profiles for GC were screened using plasma samples covering 5 control groups, 5 cases with T3N1-3M0, and 5 cases with T3N0M0. As indicated from disease enrichment term, KEGG channel, and gene ontology (GO) analyses, numerous mRNAs had implication implied during carcinogenesis. Candidate tumor-related HTRA2 mRNA was then quantitated in GC tissues and plasmas, respectively. The high expression of HTRA2 mRNA positively contributes to cell proliferating, invading, and migrating processes. Therefore, HTRA2 may serve a novel circulating biomarker for detection of GC. cDNA under the labeling based on one fluorescent dye received the production with Eberwine's linear RNA amplifying approach. The mRNA array information received the analysis in terms of information summarizing process, normalizing process, and quality controlling process with the GeneSpring software V13.0 (Agilent).

| Cases and samples
We acquired peripheral blood (5 ml) of GC cases prior to the operating process. Next, the plasma received the isolation. By complying with the age-and-gender matching standard, fresh plasma samples which were not abnormal received the collection according to patients in Nanjing First Hospital from 2018 to 2020. In terms of overall peripheral blood samples, the anti-coagulant was ethylenediaminetetraacetic acid (EDTA). We acquired GC tissues and relevant samples according to cases carrying GC, having undergone surgery from 2018 to 2020 in Nanjing First Hospital. Prior to the operating process, the cases underwent no radiotherapeutic or chemotherapeutic processes. The mentioned tissue specimens received the immediate storing under −80°C based on one refrigerating element till using. Based on the standard, the paired neighboring non-tumor tissues had the confirmation to cover not any cells of tumor by pathologically related analyzing process and had the location of 5 cm from the GC edge.
The present work gained the approval from the Ethics Committee of Nanjing First Hospital, Nanjing Medical University. Prior to their engagement, all cases presented informed consent in a written form.

| RNA isolation, reverse transcription, and quantitative real-time polymerase chain reaction (qRT-PCR)
Overall RNA according to tissues under pairing received the extraction with TRIzol reagent (Thermo Fisher Scientific) and overall RNA inside plasma received the extraction with the use of TIANamp Virus RNA Tool by complying with the guidelines of the producer.

| Co-expression analysis
Co-expression analysis is based on correlations in mathematics and to find mRNA-non-coding RNA pairs whose expression profiles are similar from the gene expressing state. A large number of functionally relevant genes exhibit consistent expressing profiles based on related conditions, in particular genes under the common regulation using normal transcribing-related elements or genes whose product constitutes the same protein complex or genes which participate in the same regulatory path. Therefore, co-expression analysis and network construction based on co-expression results can help users to discover the possible relationship between mRNA and non-coding RNAs and can find mRNAs that affect the regulation of non-coding RNA expression.

| Plasmids constructing and transfecting process
Hongxin Biological Technology Co., Ltd. (China) offered the control plasmid vector pcDNA3, as well as the HTRA2 over-expression plasmid pcDNA3-HTRA2. Plasmid pcDNA3-HTRA2 over-expressed, using pcDNA3.1 (+) to be vector. 1 × 10 6 cells received the seeding process inside 6-well plates and the 24 h culturing process. After 12-16 h, serum-free medium received the switch as a complete medium supplemented by 10% FBS. The cells under the transfection can be employed to conduct the in-depth experimental processes.
The transfecting efficiency reached over 85%.

| Cell proliferation experiments
During the clonogenesis experiment, transfected cells were inoculated in 6-well plates at a density of 1000 cells per well and then cultured in RPMI 1640 medium containing 10% FBS. After 10 days, the cells were imfixed based on the use of methanol and then stained with GIMSA.
Finally, the colony is imaged and counted. In the ELL Counting Kit-8 (CCK-8) test process, MGC-803 and SGC-7901 cells were transfected and incubated at 37°C first. Next, ccK-8 solution (Biosharp) was introduced and incubated for 2 h in their respective wells. The absorbance received the measuring process at 0, 24, 48, 72, and 96 h at 450 nm.

| HTRA2 expression level analysis and clinicopathological analysis
TCGAportal was used to investigate the expression of HTRA2 in different tumor tissues and corresponding para-carcinoma tissues. The Human Protein Atlas database was used to examine HTRA2 expression in different tissues. UALCAN refers to a comprehensive and interactive web resource to delve into cancer OMICS data, which was

| Tools for HTRA2 location in cells and enrichment analysis of HTRA2
The Human Protein Atlas was utilized for obtaining HTRA2 location in cells. We used Metascape and STRING to create a network of interactions between HTRA2 and other important proteins and pathways. TISIDB, a web portal for tumor and immune system interactions, integrating a range of heterogeneous data, was adopted for delving into the spearman correlations between HTRA2 and immune-modulator expression.

| Statistics-related analyzing process
The continuing information received the comparative analysis by performing one individual t-testing process of the 2 groups. Statisticsrelated analyzing process was performed with SPSS (Version 22.0) and presented graphically in GraphPad Prism 8.0. A p value of 0.05 was considered to be statistically significant.

| Genome-wide profiles were produced for patient/control plasma samples to identify candidate mRNAs
High-throughput human mRNA microarray was conducted using plasma samples from 10 GC cases, including 5 cases with T3N0M0 (case 1) and the other 5 cases with T3N1-3M0 (case 2), and 5 normal controls (control) to assess the differences of mRNA expression profiles in GC progression. Heat map and scatter plot were constructed to show differentially expressed mRNAs (p < 0.05 and fold change >2) in three groups: case 1 vs. normal group, case 2 vs. normal group, case 2 vs. case 1 group, respectively ( Figures 1A,B, 2A

| HTRA2 is over-expressed in GC and has remarkable clinical significance
According to stage-related abnormal expression of mRNAs, we picked out one mRNA named HTRA2. Using qRT-PCR, we have detected HTRA2 expressing state in GC plasma samples and healthy controls. The results indicated HTRA2 was significantly over-expressed in GC plasma in contrast with control cases ( Figure 4A). To confirm HTRA2 was derived from GC tumor, we further detected in primary cancerous and neighboring noncancerous tissues from GC cases. Results showed HTRA2 expressing states were significantly higher than that of neighboring noncancerous tissues ( Figure 4B). The TCGA database shows results consistent with our conclusions ( Figure 4C). The expression level of HTRA2 mRNA in different types of cancer tissues indicated that the expression level of HTRA2 in GC tissues was also higher than that in control group ( Figure 4D). Subgroup analysis based on GC subtypes showed that HTRA2 mRNA expression was linked with age, grade, and individual cancer stages ( Figure 4E-H). To delve into the prognostic potential of HTRA2 in GC, Kaplan-Meier Plotter was used. As presented in Figure 4I-J, that higher HTRA2 expressing state displayed a correlation to shorter OS and PFS in contrast with lower expression `but not FP.

| HTRA2 plays a promoting role in GC cells in vitro
To clarify the physiology-related characteristic effect exerted by

| Genes and proteins co-interacted with HTRA2 are associated with cellular response and metabolic signaling pathway
According to the human protein atlas database, HTRA2 is located in the mitochondria of HEK293, RH-30, and U-2 OS cells ( Figure 7A).

Enrichment analysis of co-expression genes performed using
Metascape indicated that HTRA2 was primarily involved cellular response and metabolic signaling pathway ( Figure 7B,C)  and CYCS which can interact with HTRA2 ( Figure 7D).

| Research results of HTRA2 at single-cell level
We studied the expression of HTRA2 at the single-cell level. Singlecell cluster map of HTRA2 in STAD GES134520 databases showed the distribution of HTRA2 expression in different cell clusters ( Figure 8A). Violin and heat maps showed the specific expression of HTRA2 in different cells, which indicated that the HTRA2 mainly expressed at CD8T, plasma, DC, mast, fibroblasts, and malignant cell clusters ( Figure 8B,C). These results suggest that HTRA2 may also function in stromal cells or immune cells other than cancer cells.

| HTRA2 expression was correlated with immune factors
Existing studies have confirmed that the immune system is closely related to the occurrence and development of tumors. Therefore, we studied the relationship between the expression of HTRA2 and immune factors. As shown in Figures 9-11

| DISCUSS ION
Gastric carcinoma refers to one of the most prevalent causes of carcinoma-related death in the world. Cases usually diagnosed with advanced carcinoma due to the lack of noticeable symptoms. Hence, the development of GC screening tool would effectively reduce the overall mortality, and novel non-invasive biomarking elements with better sensitivities and specificities are urgently needed.
Circulating cell-free mRNA was first demonstrated in the 1990s in the plasma of cases with nasopharyngeal carcinoma. 8 The discovery of circulating mRNAs in carcinoma cases holds great promise for the use of mRNAs as distinctive, non-invasive carcinoma biomarking elements and therapeutic targets. Kopreski MS et al. 9 reported that 5T4 mRNA was reproducibly detected in 42% of carcinoma patient sera, including breast carcinoma patient sera and lung car- while also increasing mitochondrial apoptosis by up-regulating Omi/ HTRA2 expression. 17 Liu Xin et al. found that mitochondrial Omi/ HTRA2 promotes caspase activation through cleavage of HAX-1 in aging heart. 18 The basis of these studies has all verified one of our results, that is, HTRA2 is located in the mitochondria of cells. processing and increases neuronal vulnerability. In addition, processed HTRA2 released into the cytosol after ischemia contributes to neuronal injury via inhibition of XIAP. 21 The interactions between HTRA2 and other proteins are not described here one by one, and more research reports are expected in the future.
Existing studies have confirmed that the immune system is closely related to the occurrence and development of tumors.
Therefore, we studied the relationship between the expression of HTRA2 and immune factors. There was a strong correlation be- ALDH1A1 and HTRA2 expression was modulated in CCR2-deficient and CCR2-over-expressing cell lines. They found that ALDH1A1 and HTRA2 regulates CCR2-mediated breast cancer cell growth and cellular invasion in a CCL2/CCR2 context-dependent manner. 22 We also studied the expression of HTRA2 at the single-cell level and found that HTRA2 mainly expressed at CD8 T, plasma, DC, mast, fibroblasts, and malignant cell clusters. These results suggest that HTRA2 may also function in stromal cells or immune cells other than cancer cells. However, at present, there are few studies on the correlation between HTRA2 and cancer in immune cells or stromal cells.
It is expected that more studies will make breakthroughs in these aspects in the future.
In conclusion, we applied a systematic method to analyze several mRNAs capable of distinguishing cases with GC. We found HTRA2 was over-expressed in GC plasma in contrast with control cases.