Serum microRNA Expression Proling in Malaria Patients and Bioinformatic Analysis of Hsa-miR-106b-5p

Malaria, caused by Plasmodium, is one of the three major infectious diseases that se riously endangers public health. Resistance to an ti-malarial drugs and insecticides has made the prevention and control of malar ia shown little improvement in the last four years. This study aimed to explore the changes in microRNA (miRNA) expression proling of malaria patient and predict malaria-related miRNA by bioinformatics methods to provide theoretical basis for further verication of the correlation between specic miRNAs and immune regulation of malaria. and analysis were further visualized The results demonstrate that it plays a role in inhibiting tumor progression and the emergence of ART resistance in Plasmodium, which may be achieved by regulating vascular morphogenesis, endocytosis, and VEGFA. The study ndings will provide a systematic understanding of the biological processes associated with hsa-miR-106b-5p, and they will provide data support and theoretical basis for further studies to verify the link between malaria and tumors, design malaria vaccines, and regulate miRNA functions to achieve disease treatment.


Background
Malaria is one of the three major infectious diseases that seriously endangers public health, is mainly transmitted by a bite of Plasmodium-infected female Anopheles mosquitoes []. According to the World malaria report 2020, in 2019, approximately 229 million cases of malaria were reported worldwide, resulting in 409,000 deaths, most of which were children under 5 years of age in sub-Saharan Africa []. In the past two decades, tremendous progress has been made in preventing, controlling, and eliminating malaria. However, to date, the global mortality of malaria remains high. The spread of insecticide resistance [, ], emergence of antimalarial drug resistance [, ], and lack of safe, effective, and easily accessible malaria vaccine all have made the prevention and control of malaria challenging.
Lots of studies have indicated that microRNAs (miRNAs) perform a crucial biological function in the immune regulation of Plasmodium and its host [,,]miRNAs are endogenous conservative single-stranded non-coding RNAs containing approximately 22 nucleotides. miRNAs do not possess any coding function of their own, but they can bind to the 3' end (3'-untranslated region) of speci c target mRNAs in a complementary base pairing manner, thereby inhibiting translation of the transcript and regulating physiological functions [, ]. miRNAs are biomarkers for various diseases, and they are widely involved in diverse physiological processes such as cell proliferation, differentiation, and apoptosis [, , ]. Some studies have shown that miRNA has potential to be biomarkers for malaria []. Additionally, our group also have con rmed that the downregulation of miR-106b in mice is positively correlated with a decrease in the immune protection of malaria vaccine. For miRNA has a complex regulatory network, it is possible to improve the problem of preventing and controlling of malaria through regulating function of malariarelated miRNA. The study of miRNA is important for understanding gene regulatory networks, gene functions and the biological processes. However, current ndings of the relationship between miRNA and malaria, and the underlying mechanisms are only the tip of the iceberg. Predicting target genes and conducting systematic bioinformatic analyses of miRNA are essential for an in-depth investigation. Therefore, in this study, we analyzed the serum miRNA expression pro ling of malaria patients by bioinformatics methods to obtain a comprehensive and systematic understanding of the biological processes and miRNA interactions in malaria.

Sample collection
The study was performed on two groups, ve Plasmodium falciparum (P.falciparum ) infected patients (experimental group) and three healthy people (control group), all participants were from the countryside in Myanmar border area. Blood samples of the participants were collected using vacuum tubes.
Subsequently, serum was separated out naturally by stored at room temperature for 30 mins − 1 h, and then stored at -20°C for follow up miRNA expression pro ling experiment.
Total RNA isolation and reverse transcription This experiment was performed at the Biomedical Analysis Center of Army Medical University. First, the Trizol method was used for RNA isolation. Following this, 200µL serum was immediately mixed with 800µL Trizol in an Eppendorf tube. The Total RNA Isolation Kit (Life Technologies) was used for subsequent experiments, according to manufacturer's instructions. The concentration and purity of RNA were determined using electrophoresis by enzyme-labeling measuring instrument (Gen5-CHS1.09). Finally, E.coli poly-A-polymerase was used to generate polyadenylated tails to the 3'-end of all mature miRNA sequences. cDNA was synthesized using reverse transcriptase with the qScript Flex cDNA synthesis kit (Quanta Biosciences).
Mirna Qrt-pcr An Echo550 instrument was used to inject samples (mix, template, primers). The ViiA7 Real-Time PCR Thermocycles and SYBR green-based real-time quantitative PCR method were used to quantify the relative expression of mature miRNAs, as described previously []. Two duplicate holes were prepared for each sample. All primers and probes used for miRNA quanti cation were purchased from Quanta Biosciences, and a total of 341 miRNAs were detected. The 2 −ΔΔCt method was applied to analyze the relative expression levels of miRNAs.

Analysis Of Serum Mirna Expression Pro ling
The fold change (FC) approach was used to calculate serum miRNA expression levels between P.falciparum-infected and healthy control groups. The miRNA expression level in the infected group changed twice as compared to that in the healthy group was de ned as differential expression, and differentially expressed miRNAs were obtained. By controlling the false discovery rate to correct the false positive rate, and using the SAM library, we screened the differentially expressed miRNAs (P < 0.05, |logFC|>1) and selected the most signi cant one (adjusted.P.value < 0.05, |logFC|>1) for bioinformatic analysis. Finally, the predicted overlapping target gene set was imported into the STRING online database (ver 11.0) [, ] for Protein-protein interaction (PPI) analysis. The minimum required interaction score was set at a medium con dence (0.400), and disconnected nodes in the network were excluded to construct a PPI network. The analysis results of STRING were imported into Cytoscape for further visual analysis and the hub genes, which played a pivotal role in the whole network, were screened. Statistical analysis miRNA expression pro ling data were analyzed using SPSS (ver18.0) computer software (SPSS for Windows, SPSS Inc., 2009). Statistically signi cant difference was set at P < 0.05.

Analysis of serum miRNA expression pro ling in malaria patients
To identify a diagnostic serum miRNA signature for malaria, Vii7 and a real-time based high-throughput PCR array were used to compare the serum miRNA expression pro ling of ve P.falciparum infected patients and three healthy controls. Of the 341 miRNAs detected, 64 were differentially expressed in malaria patients (P < 0.05). Among these, 27 were up-regulated and 37 were down-regulated (Additional le 1). Additionally, the miRNA with the most signi cant difference was hsa-miR-106b-5p (adjusted.P.value < 0.01, |logFC|>1). Its expression level of Plasmodium infected group was up-regulated by 14.871 times compared with the control group, and it was selected for subsequent bioinformatic analysis.
Prediction of the target gene of hsa-miR-106b-5p Four target gene sets of hsa-miR-106b-5p, 415 2037 1384 and 1091 target genes, were predicated by TargetScan, DIANA-microT, miRDB, and miRTarbase respectively ( Fig. 1). A total of 78 overlapping genes were obtained using Venn analysis, accounting for 2.6% of the target genes predicted by all databases.
A PPI network of the predicted target genes of hsa-miR-106b-5p was constructed using STRING, consisting of 39 nodes and 43 edges (Fig. 5). Following analysis, 78 predicted overlapping target genes were used as input, while 39 genes, which were disconnected from any other node after analysis, were removed. It has been demonstrated that complex interactions such as co-expression, neighborhood, and gene fusion, exist between target genes. Furthermore, the ve hub genes (genes that interact with at least four other target genes) VEGFA, STAT3, RACGAP1, OCRL, and RBBP7, in the PPI network were speculated to play essential roles in the biological effects regulated by hsa-miR-106b-5p. Further analyses of the ve hub genes revealed that all of them played vital roles in tumorigenesis.

Discussion
Analysis of serum miRNAs expression pro ling between malaria patients and healthy controls revealed that 64 of the 341 detected miRNAs were differentially expressed. Among them, expression of the miRNA, hsa-miR-106b-5p, was signi cantly higher in malaria patients than in the healthy controls (the expression level is 14.871 times that of the control group), which had the smallest P value and adjusted. P.value, and largest logFC value. Thus, we con rmed that the downregulation of miR-106b is positively correlated with a decrease in the immune protection of malaria vaccines. It has been reported that increasing the expression of miR-106b in vitro can enhance the expression of transmembrane activator and CAML interactor (TACI) on the surface of memory B cells (MBCs) []. This indicated that hsa-miR-106b-5p is closely related to malaria, therefore we focused on its analysis.
We used bioinformatic analysis such as target gene prediction, GO and KEGG analyses to understand the biological functions of hsa-miR-106b-5p and its association with malaria comprehensively and systematically. First, in order to make up for the inevitable false positives and false negatives of bioinformatic predictions, four authoritative miRNA databases, TargetScan, DIANA-microT, miRDB, and miRTarbase, were used to predict the target genes of hsa-miR-106b-5p. The intersection of these results was analyzed using Venn, and a more reliable target gene set, containing 78 overlapping predicted target genes was obtained (Fig. 1). Subsequently, Metascape was used to perform GO and KEGG analyses for overlapping target genes.
GO results revealed that the target genes were signi cantly enriched in the related functions of small GTPases (Additional le 2). Small GTPases are key participants in a series of pathophysiological processes, most of which are associated with the regulation of protein secretion, endocytosis and vesicular transport [, ]. Previous studies have shown that the dysfunction and negative regulation of certain small GTPases, such as members of the Ras and Arf subfamilies, are related to the promotion and progression of tumors [, ], and recent studies have con rmed that GO:0007264 ~ small GTPase mediated signal transduction is related to the occurrence of breast cancer []. Studies have shown that miR-106b-5p is the most regulated miRNA in cancer tissues. It is involved in the emergence and development of pancreatic cancer, lung cancer, and breast cancer among others [,,,], which not only veri es the accuracy of our approaches and directions, but also suggests a link between malaria and tumors.
In 2011, antitumor effects were observed in a Lewis lung cancer model mouse after infection with Plasmodium yoelii. The proliferation of tumor cells was signi cantly inhibited and the survival of tumorbearing mice was remarkably prolonged []. An epidemiological survey also showed a signi cantly negative correlation between the incidence of malaria and tumor mortality []. In general, there are three reasons: (1) Periodic high fevers bouts occur in the host in the acute phase of malaria infection, which can kill certain non-heat-resistant cancer cells. Hyperthermia has been clinically used for the treatment of some cancers [, ]. (2) Plasmodium infection can stimulate the innate immunity of host, activating and enhancing antitumor immunity. It mainly occurs through activation of NK and DC cells to promote the production of a large number of cytokines such as IFN-γ and TNF-α, which antagonize the tumor immunosuppressive microenvironment, leading to an antitumor effect. In fact, malaria infection can enhance the host immune response and it has been used as an adjuvant to treat cancer []. GO analysis showed that GO:0002376 ~ immune system process (P < 0.01), GO:0048518 ~ positive regulation of biological process (P < 0.01) and GO:0008283 ~ cell proliferation (P < 0.01) are all related to immune system. (3) Excluding immune responses, malaria infection also acts on VEGFR2 through miRNAs in exosomes and lncRNA F63 in tumor tissues, thereby inhibiting tumor angiogenesis. Angiogenesis is the formation of new blood vessels from pre-existing blood vessels. It plays an important role in tumor development, and it is responsible for the transportation of nutrients needed for tumor growth. Cutting off the nutrient supply from tumor blood vessels will cause the cancer cells to "starved to death" [25, ]. This is strongly associated with GO:0048514 ~ blood vessel morphogenesis (P < 0.01), suggesting that malaria infection inhibits tumor progression. Besides, GO:0051781 ~ positive regulation of cell division and GO:003052 ~ intracellular receptor signaling pathway are also important for further investigation.
KEGG analysis results revealed that the overlapping target gene set was enriched in hsa04144: Endocytosis (P < 0.01), hsa01521: EGFR tyrosine kinase inhibitor resistance (P < 0.01), hsa05212: Pancreatic cancer (P < 0.01) and other pathways ( Table 3). The most signi cant enrichment was observed in endocytosis, which plays an essential role in malaria infection. For Plasmodium, the importance of endocytosis is self-evident []. In the pre-erythrocytic stage of malaria, endocytosis mediates the uptake and degradation of up to 80% of host cell cytoplasm which is mainly composed of hemoglobin by the malaria parasite. However, the molecular mechanism of this process remains unclear []. In general, endocytosis requires core transport factors, including Rab-GTPase [, ], which is similar to the results of GO analysis, and thus, it further provides evidence for the link between hsa-miR-106b-5p and malaria. Interestingly, a recent study found that mutations of Kelch13 (K13) protein in Plasmodium are resistant to artemisinin (ART) and its derivatives due to a reduction of hemoglobin endocytosis []. Artemisinin is a frontline antimalarial drug. However, drug resistance has gradually emerged in Cambodia and the Greater Mekong Subregion due to long-term use and abuse. The main clinical manifestation of artemisinin resistance is that the clearance of malaria parasite is delayed or reduced after treatment with ART [, ]. The mutated K13 gene in Plasmodium is related to a delayed parasite clearance after treatment with Artemisinin-based combination therapy (ACT), as recommended by WHO []. Molecular characteristics of the highly conserved K13 C-terminal region in P.falciparum, kelch13 propeller region (Pfk-13), are closely linked to the Plasmodium phenotype from Cambodia having a delayed ART clearance []. Moreover, K13 is related to hemoglobin internalization and catabolism. Thus, reducing K13 impairs hemoglobin catabolism and reduces artemisinin activation []. Studies on compartment proteins of K13 and their functions revealed the entire pathway of ART resistance, that is, K13 and its compartment proteins play important role in the endocytosis of hemoglobin uptake by malaria parasite. Hemoglobin digestion products are necessary for the activation of ARTs. Therefore, an inactivation of all kinds of K13 proteins will reduce hemoglobin uptake by malaria parasites, thereby reducing the activation and concentration of ART decreasing the susceptibility of malaria parasites to ART, and eventually resulting in the resistance of malaria parasite to ART. Moreover, the main manifestation observed in wildtype anti-ART P.falciparum explains that changing the stability of K13 protein can affect endocytosis and lead to resistance to ARTs from both negative and positive sides. An opportunistic inactivation of K13 gene will cause the malaria parasite to remain in the ring stages and no longer grow and divide. However, the inactivation of K13 only affects the ring stage and does not work for trophozoite stages. This may be the reason for the frequent mutations found in drug-resistant Plasmodium isolates due to K13 gene, out of all proteins that affect endocytosis [36, , ]. Since the main reason for the emergence of ART resistance among Plasmodium is the reduction of endocytosis and the predicted target genes of hsa-miR-106b-5p are signi cantly concentrated in the signaling pathway of endocytosis, hsa-miR-106b-5p is the key miRNA in malaria. Is it possible to enhance endocytosis by targeting hsa-miR-106b-5p and improve ART resistance?
Finally, we constructed a PPI network using STRING, containing 39 predicted target genes of hsa-miR-106b-5p. Additionally, we identi ed ve hub genes, VEGFA, STAT3, RACGAP1, OCRL, and RBBP7, which may play vital roles in the regulation of biological effects in hsa-miR-106b-5p (Fig. 5). The most critical hub gene was VEGFA, a key factor for vascular endothelial activation, which induces the survival, migration, and proliferation of endothelial cells during angiogenesis, and promotes vascular penetration and endothelial activation during in ammation []. It has been proposed as a biomarker for malaria severity, and it is expressed in the plasma and brain tissues of malaria patients [, ]. Endothelial activation is an early feature of malaria. It is probably bene cial for the isolation of infected RBCs and it also leads to thrombocytopenia. The number of platelets was negatively correlated with parasitemia and malaria severity. These results reveal a correlation between hsa-miR-106b-5p and malaria. Furthermore, the expression of VEGFA increased in many cancers. When its production is inhibited, tumor angiogenesis reduces, thus, decreasing the source of nutrition for cancer cells and nally achieving an anti-tumor effect. Monoclonal antibodies against VEGFA have been used to treat cancer. However, they cause some side effects and redundancy of other antigenic factors []. Therefore, using miRNA to target genes is highly speci c and safe. An equally important gene, STAT3, regulates cell growth and apoptosis, whereas RACGAP1, a GTPase-activating protein belonging to the Rho GTPase family, is highly expressed in malignant tumor cells. These hub genes are of great signi cance in malaria and tumors.
In brief, the GO results of our bioinformatic analysis suggest that hsa-miR-106b-5p may mediate malaria to inhibit tumor progression by controlling vascular morphogenesis and other complex functions. The KEGG results indicate that hsa-miR-106b-5p can improve ART resistance and affect the growth and division in Plasmodium through endocytosis. Likewise, the PPI of predicted target genes revealed the relationship between VEGFA and malaria severity, and the link between malaria parasite and tumor growth and metastasis. However, our study had some limitations. First, the number of volunteers was small, and it may not be representative due to large individual differences in humans. However, the study detected 341 miRNAs, obtained serum miRNA expression pro ling of malaria patients, and identi ed certain differentially expressed miRNAs. To reduce errors, we will expand the sample size in future experiments. Second, disease factors of volunteers, which may affect the results were not considered. Confounding factors can be increased in information statistics. Third, false positives or false negatives inevitably exist in bioinformatics analysis approaches. Therefore, a further experimental veri cation for the predicted target genes and their biological functions is needed. The dual-luciferase reporter assay can be used to identify binding sites of miRNA target genes and choose miRNA mimics or inhibitors for in vitro induction and in vivo experiments for functional veri cation.
In summary, this study analyzed the changes in serum miRNA expression pro ling of malaria patients using bioinformatic analysis, identi ed differentially expressed miRNAs, and conducted a comprehensive and systematic analysis of hsa-miR-106b-5p, which shows the maximum statistical difference and has attracted tremendous research interest. Additionally, we constructed a PPI network for the predicted target genes of hsa-miR-106b-5p. The results demonstrate that it plays a role in inhibiting tumor progression and the emergence of ART resistance in Plasmodium, which may be achieved by regulating vascular morphogenesis, endocytosis, and VEGFA. The study ndings will provide a systematic understanding of the biological processes associated with hsa-miR-106b-5p, and they will provide data support and theoretical basis for further studies to verify the link between malaria and tumors, design malaria vaccines, and regulate miRNA functions to achieve disease treatment.