Plasmatic Small Non-coding RNA For The Detection Of Gastric Cancer And Its Impact On Tumor Immunity

Gastric cancer is the second most common malignancy and the second leading cause of cancer death in China.Even with aggressive treatment,the 5-year survival rate in patients at advanced stage is still less than 30%.Cancer development and its response to treatments are strongly influenced by the immune system.The crosstalk between tumor cells and immune cells plays a central role in determining disease outcome.Dendritic cells(DCs)are initiator of immune response and they are capable of inducing either immune tolerance or activation.Mature DCs overexpressing costimulatory molecules and antigen presenting molecules promote T cell proliferation and differentiation,resulting in anti-tumor immune response.While immature DCs show antigen presentation dysfunction,secrete immunosuppressive factors,leading to tumor escaping from immune surveillance by induction of T cells anergy and regulatory T cells proliferation.Tumor cells are weak immunogenicity and secrete immunosuppressive factors to inhibit DCs maturation,thus induce tumor escape and promote tumor progression.Small non-coding RNA(snRNAs)regulates protein-encoding genes at post transcription or transcription level.Currently,the functional study of snRNAs mainly focuse on micro RNAs(miRNAs)and piwi-interacting RNAs(piRNAs).SnRNAs are highly stable.Expression profiles of miRNAs didn’t show much difference between fresh frozen and FFPE(formalin-fixed paraffin-embedded)samples dating back up to 10 years.Various cells secrete snRNAs,and these snRNAs circulate in body fluids as mediators of cell-cell communication.It was reported that pancreatic cancer derived exosomal miR-212-3p inhibited mRNA expression of MHC-II in DCs and induced immune tolerance,suggesting that tumor-derived mi RNAs regulate tumor immunity by acting on DCs in the tumor microenvironment.During the progression of gastric cancer,snRNAs expression profiles change intracellularly and extracellular.Studies revealed that mononuclear cell derived DCs showed less MHC-II expression in the peripheral blood of gastric cancer patients than those of healthy control,indicating maturation inhibition.While mature DCs infiltration was positively associated with overall survival and disease-free survival after initial surgery.Currently,little is known about the influence of gastric cancer cell-derived snRNAs on DCs maturation.Besides,circulating snRNAs serve as biomarkers for noninvasive diagnosis and surveillance.Efforts have been made to explore their feasibility for detection of gastric cancer with large samples stepwisely.However,most studies only focus on their expression in the peripheral blood of gastric cancer patients without tracing their original cells,so these biomarkers do not monitor tumor dynamics.In the present study,we screened the sn RNAs expression profile in plasma and FFPE samples of gastric cancer patients by deep sequencing,finding snRNAs secreted by gastric cancer cells,and verify their feasibility for the detection of gastric cancer,and also exploring their effects on DCs and tumor immunity to provide an informative reference for the detection and treatment of gastric cancer.Part 1 Investigation of cell-free non-coding small RNAs for the detection of gastric cancerObjectives: To investigate the characteristics of snRNAs in tissue and plasma samples,identify potential biomarkers for detection and establish a diagnosis model for the detection of gastric cancer.Methods: Plasma samples were collected from 82 patients with gastric cancer,and 82 healthy controls.FFPE samples were collected from 8 gastric cancer and their adjacent tissues.Firstly,the mixed plasma samples and the mixed tissue samples were sequenced respectively.Then,the candidate snRNAs were verified by qRT-PCR in 82 pairs of plasma samples.The relationship between sn RNAs expression and clinical stage was analyzed by Kruskal-Wallis test and Dunn-Bonferroni test.Finally,the diagnostic model of gastric cancer was established with Logistic regression analysis,and its diagnostic performance was evaluated by ROC curve.Results: 1 Spectra of non-coding small RNA 1.1 Length distribution of small RNA in FFPE and plasma samplesSmall RNAs with length of 18-35 nt were sequenced.In FFPE samples,small RNAs were mainly 20-36 nt in length,and were enriched near 22 nt and 32 nt to form two peaks,indicating miRNAs and piRNAs enrichment.While in plasma samples only a single peak appeared near 21 nt,indicating that only miRNAs were significantly enriched.1.2 Percentages of sequences and read counts for categorized small RNAs in FFPE and plasma samplesIn gastric cancer and adjacent specimens,the number of miRNAs was 6473 and 4236 respectly,accounting for 0.92% and 0.83% of the total RNAs in species,33.11% and 24.19% in quantity.In plasma samples,the number of miRNAs was 2466 and 2609,accounting for 0.98% and 0.92% of all the small RNA in species,40.36% and 54.79% in quantity.Indicating that mi RNAs are hilgly expressed though less in species compared with other small RNAs.The type and number of pi RNAs were rare.1.3 Differentially expressed small non-coding RNAsAccording to the different expression pattern in FFPE and plasma samples,snRNAs can be divided into four groups: elevated expressed in both FFPE and plasma samples(31 miRNAs);declined expressed in both FFPE and plasma samples(1 mi RNAs);elevated expressed in FFPE samples but declined expressed in plasma samples(3 miRNAs and a piRNAs): declined expressed in FFPE samples but elevated expressed in plasma samples(4 miRNAs).2 Validation of gastric cancer related small non-coding RNAs and their diagnostic performance 2.1 Validation of gastric cancer related small non-coding RNAsWe focused on the 31 miRNAs that were both elevated in FFPE and plasma samples of gastric cancer.Previous work had validated the expression levels of miR-17-5p,miR-127-3p,miR-379-5p,miR-433-3p,mi R-654-3p in 23 paired FFPE samples and 42 paird plasma samples by qRT-PCR.Results showed that the change of these miRNAs screened by deep sequencing was comparable with the qRT-PCR analysis.We then tested the expression of the five miRNAs for further validation using qRT-PCR with a total of 82 paired plasma samples of gastric cancer patients.Results showed that these five miRNAs were significantly upregulated compared with healthy controls,making them candidates for biomarkers for detection of gastric cancer.2.2 The relationship between the 5 mi RNAs and clinical stage of gastric cancerThe expression levels of mi R-17-5p,miR-379-5p,miR-433-3p and miR-654-3p at different stage of gastric cancer were significantly higher than those in healthy controls as confirmed by Kruskal-Wallis test,while no significant difference was found among different stages as confirmed by Dunn-Bonferroni test,indicating their quick response and durable expression during gastric cancer development.Compared with the healthy control,an increased expression level of miR-127-3p was found in stage I,stage II and stage III but not stage IV of gastric cancer plasma samples,suggesting that the miR-127-3p secretion was suppressed in stage IV.2.3 The diagnostic performance of single mi RNA for gastric cancerROC curve analyses were conducted to determine the diagnostic usefulness of the 5 selected mi RNAs for gastric cancer.Results showed that the AUC were 0.82(95% CI:0.75-0.88),0.78(95% CI: 0.71-0.85),0.81(95% CI: 0.74-0.88),0.78(95% CI: 0.71-0.85),0.76(95% C: 0.69-0.83)accordingly,with miRNAs-17-5p showed the best performance.2.4 Optimizing a panel of plasma mi RNAs biomarkers for diagnosis of gastric cancerTo optimize a panel of miRNAs biomarkers for gastric cancer with high accuracy,we used Logistic regression analysis.Results revealed that miR-17-5p and miR-127-3p in the plasma sample were risk factors of gastric cancer.The model showed: Y=110.98EmiR-17-5p+66.5EmiR-127-3p-2.73。The use of this model generated 67.10% sensitivity,96.30% specificity,and 0.84 AUC,with a cutoff value of 0.55,offering superior performance compared with single mi RNA.Conclusions:1 MiRNAs are in high expression level but in few category in FFPE and plasma samples of gastric cancer patients.2 Plasmic mi R-17-3p,miR-127-3p,miR-379-5p,miR-433-3p,mi R-654-3p are derived from gastric cancer cells and increase significantly at early stage of gastric cancer.3 Plasmatic miR-17-5p and miR-127-3p are risk factors for gastric cancer.Diagnostic formula for gastric cancer show: Y=110.98EmiR-17-5p+ 66.5EmiR-127-3p-2.73,with the cutoff value of 0.55.Part 2 Tumor-derived miR-17-5p inhibits dendritic cell maturationObjectives: To investigate the effect of gastric cancer derived miR-17-5p on DCs maturation.Methods:The expression level of miR-17-5p in BGC-823 cells and their supernatant was detected by qRT-PCR.Cy3 marked mi R-17-5p was transfected to BGC-823 cells,and then the conditioned medium was cocultured with DCs.Cy3 fluorescence signals in DCs were determined by flow cytometry and visualized by confocal microscopy.DCs were extracted from bone marrow of C57BL/6 mice,and purified by immunomagnetic beads.Flow cytometry was used to test the phenotype of DCs,and the uptake of FITC-dextran.The expression of cytokines in cells and supernatant were detected by qRT-PCR and ELISA.The allogeneic T cells were cocultured with pretreated DCs(with mitomycin C)and the T cell proliferation was assessed by BrdU ELISA,CD4+ CD25+ Foxp3+ Tregs(regulatory T cells)in cocultures were then quatified by flow cytometry.Results: 1 miRNAs-17-5p is derived from human gastric cancer cellsQRT-PCR was used to test the expression level of miRNAs-17-5p in BGC-823 cells and medium,and their relationship was evaluated by spearman rank correlation analysis.Results showed an increased tendency of miR-17-5p expression in the supernatant of BGC-823 cells in a time-dependent manner when cells were planted at lower concentration but not at other concentrations.A significant correlation was confirmed between medium concentration and plated cell number at 24 h(R2=0.96,P=0.00).The above data suggest that gastric cancer cells secrete miR-17-5p in a time-dependent and cell number-dependent manner,and high cell density inhibits its secretion.2 MiR-17-5p is shuttled from gastric cancer cells to imDCsBGC-823 cells were cultured alone(blank)or transfected with reagent(mock),unlabeled miR-17-5p or cy3-miR-17-5p for 24 h,and then the conditioned medium was cocultured with DCs for another 24 h.Flow cytometry was used to determine the cy3 fluorescence signals in DCs gating on CD11 c.Results showed that only a weak cy3 fluorescence signal was detected in groups of blank,mock and mi R-17-5p.Compared with these three groups,cy3-miR-17-5p group showed more cy3 fluorescence signals.Confocal microscopy revealed that cy3 signals were only seen in cy3-miR-17-5p group.The above data suggest that mi R-17-5p is derived from gastric cancer cells.3 MiR-17-5p inhibits LPS-induced maturation 3.1 Resistance of DCs-miR-17 to LPS-induced phenotypic maturationThe percentage of CD11c+ cells was 65% to 90% before purified,and it improved to more than 95% after immunomagnetic sorting.After LPS stimulation,the expression levels of CD80,CD86 and MHC-II were significantly increased compared with imDCs.DCs transfected with miR-17-5p mimics(DCs-mi R-17)before LPS stimulation,however,expressed lower levels of CD80,CD86 and MHC-II compared with those transfected with scramble control(DCs-N).The above data suggest a resistance of DCs-miR-17 to LPS-induced phenotypic maturation.3.2 DCs-miR-17 is resistant to LPS-induced suppression of endocytic capacityUpon LPS stimulation,endocytic capacity was impaired.DCs transfected with miR-17-5p mimics before LPS stimulation,however,showed enhanced endocytic capacity compared with DCs-N,but less potent than imDCs.The above data suggest that imDCs-miR-17 is resistant to LPS-induced suppression of endocytic activity.3.3 MiR-17-5p modulates cytokine profile secreted by DCs in response to LPS stimulationUpon LPS stimulation,the expression levels of IL-12 and TNF-α significantly increased compared with imDCs.However,upregulation of miR-17-5p caused a significant depression of IL-12 and TNF-α production of DCs.In response to LPS stimulation,mDCs produced less IL-10 than did imDCs,while DCs-miR-17 produced more IL-10 than did DCs-N.The above data suggest that mi R-17-5p modulates cytokine profile secreted by DCs in response to LPS stimulation 4 The regulation of T cells by miR-17-5p transfected DCs 4.1 Induction of T-cell hyporesponsiveness by DCs-mi R-17The percentage of CD4+ T cells was 24% to 30% before purified,and it improved to more than 95% after immunomagnetic sorting.DCs pretreated with mitomycin C were cocultured with CD4+ T cells,and Brd U ELISA was used to test the proliferation of T cells.Results showed that mDCs induced T cell hyperesponse compared with imDCs,while DCs-miR-17 suppress T cell proliferation compared with DCs-N.T cell proliferation when cocultured at 1: 20,1: 50,1: 100 with DCs was comparable to that with imDCs.DCs-miR-17 induced T cell hyporesponse compared with imDCs when cocultured at a ratio of 1:10(DCs: T).The above data suggest that mi R17-5p reduces the ability of DC to stimulate CD4+ T cell proliferation.4.2 DCs-miR-17 expanded TregsThe percentage of CD4+ CD25+ Foxp3+ Tregs in the coculture system was detected by flow cytometry.The results showed that Treg percentage was significantly increased in T cells cocultured with imDCs compared with that with mDCs.DC-mi R-17-induced Tregs were significantly more than that induced by DC-N or imDC.Indicating that mi R17-5p improve the ability of DC to induce CD4+ CD25+ Foxp3+ Tregs.Conclusions:1 MiR-17-5p is shuttled from gastric cancer cells to imDCs.2 MiR-17-5p inhibits LPS-induced maturation of imDCs.3 In the microenvironment of gastric cancer,mi R-17-5p inhibites CD4+ T proliferation and promotes Treg expansion by inhibiting DCs maturation.Part 3 Screening and Verification of miR-17-5p Target geneObjectives: Screening for the target of mi R-17-5p involved in the process of DC maturation resistance and elucidating its mechanism.Methods: The validated targets of miR-17-5p were queried in miRTarBase database for GO enrichment analysis and KEGG pathway analysis.Genes involved in the development of the immune system were submitted to the STRING database and filtered for interactions of high confidence(score>0.7).Interaction data were downloaded and processed with Cytoscape to find hub genes.The regulation of miR-17-5p on its target was verified by qRT-PCR.The function of miR-17-5p was verified by western blot.Results:1 Enrichment analysis of miR-17-5p target genesThere were 1142 validated target genes of miR-17-5p recorded by miRTarBase,among which 63 were identified by reporter assay.GO and KEGG pathway enrichment analyses were performed with these 63 derict targets of miR-17-5p.Results showed that genes were enriched in immune system process(42.86%)and developmental process(63.49%),as well as in pathways in cancer(28.57%)and miRNAs in cancer(22.22%).2 MiR-17-5p target genes involved in the developental process of immune systemA total of 23 genes were involved both in immune system process and developmental processes.STRING and Cytoscape were applied to analyse the interactions of proteins encoded by these genes.Results showed that the protein-protein interaction network consistsed of 19 nodes and 70 edges,with the maximum degree of 8,the minimum degree of 1,and the average degree of 3.68±2.31.Hub genes of the network were PTEN,BCL2,CDKN1 A,HIF1A and VEGFA.3 MiR-17-5p target genes involved in gastic cancer immunityAll of the hub genes were involved in pathways in cancer,among which BCL2 and VEGFA were associated with gastric cancer as recorded by KEGG DISEASE database.Above results reveale that BCL2 and VEGFA may affect the development of gastric cancer by the regulation of the immune system.4 MiR-17-5p inhibits the transcription of BCL2 in imDCsImDCs were transfected with miR-17-5p mimics or scramble control,and qRT-PCR was used to detect the expression of BCL2 and VEGFA at mRNA level.Results showed that the expression of BCL2 in DCs-miR-17 was significantly decreased,while no significant difference was found in the expression of VEGFA between DCs-miR-17 and DCs-N.5 MiR-17-5p promoted imDCs apoptosisSelf-antigens derived from apoptotic cells do not induce imDCs maturation,and they are required to maintain self-tolerance.ImDCs capture apoptotic DCs and resistant to LPS induced maturation.In this study,western blot was used to evaluate the expression level of apoptosis-related proteins in DCs.Results showed that the expression of anti-apoptotic protein Bcl-2 was significantly decreased followed by miR-17-5p mimics transfection,while the expression of pro-apoptotic protein Bax and cleaved caspase-3 were significantly increased.Results indicate that miR-17-5p promotes apoptosis of DCs.As DCs transfected with mi R-17-5p mimics showed enhanced endocytic capacity compared with DCs-N,we highly speculate that mi R-17-5p inhibit imDCs maturation by promoting apoptosis of DCs.Conclusions:1 Hub genes of the miR-17-5p regulatory network in the developental process of immune system were PTEN,BCL2,CDKN1 A,HIF1A and VEGFA.2 MiR-17-5p targeted BCL2 to inhibit its expression,and promote imDCs apoptosis,thereby inhibiting LPS induced imDCs maturation.