Wolbachia infection-responsive immune genes suppress Plasmodium falciparum infection in Anopheles stephensi

Wolbachia, a maternally transmitted symbiotic bacterium of insects, can suppress a variety of human pathogens in mosquitoes, including malaria-causing Plasmodium in the Anopheles vector. However, the mechanistic basis of Wolbachia-mediated Plasmodium suppression in mosquitoes is not well understood. In this study, we compared the midgut and carcass transcriptomes of stably infected Anopheles stephensi with Wolbachia wAlbB to uninfected mosquitoes in order to discover Wolbachia infection-responsive immune genes that may play a role in Wolbachia-mediated anti-Plasmodium activity. We show that wAlbB infection upregulates 10 putative immune genes and downregulates 14 in midguts, while it upregulates 31 putative immune genes and downregulates 15 in carcasses at 24 h after blood-fed feeding, the time at which the Plasmodium ookinetes are traversing the midgut tissue. Only a few of these regulated immune genes were also significantly differentially expressed between Wolbachia-infected and non-infected midguts and carcasses of sugar-fed mosquitoes. Silencing of the Wolbachia infection-responsive immune genes TEP 4, TEP 15, lysozyme C2, CLIPB2, CLIPB4, PGRP-LD and two novel genes (a peritrophin-44-like gene and a macro domain-encoding gene) resulted in a significantly greater permissiveness to P. falciparum infection. These results indicate that Wolbachia infection modulates mosquito immunity and other processes that are likely to decrease Anopheles permissiveness to Plasmodium infection.


Introduction
Wolbachia, a maternally transmitted symbiotic bacterium of certain mosquito species, has already been developed as a tool for the control of arboviral diseases such as dengue.Wolbachia has been shown to modify Aedes aegypti mosquito biology in ways that make the mosquitoes less permissive to arboviruses, rendering the mosquitoes incapable of transmitting these diseases [1][2][3].Given the previous success in using Wolbachia as a virus transmission-blocking strategy in Ae. aegypti, efforts are now being made to expand this strategy to anopheline mosquitoes that are vectors of malaria [4].Previous studies involving transient somatic infection have indicated that Wolbachia may impair Plasmodium transmission in Anopheles mosquitoes, possibly by inducing mosquito antiparasitic immune responses [5,6].These Wolbachia-related effects support the feasibility of Wolbachia-based interventions for malaria vector control, but only if stable Wolbachia infections can be established in Anopheles malaria vectors.
While early attempts to establish stable Wolbachia infection in anopheline mosquitoes were unsuccessful, several recent studies have detected Wolbachia DNA in various Anopheles species [7][8][9][10][11][12][13][14].Bian et al. (2013) generated the first stable Wolbachia (wAlbB)-infected An. stephensi LB1 line [15], leading to successful maternal transmission and cytoplasmic incompatibility (CI).Importantly, the wAlbB-transinfected An. stephensi showed significantly reduced permissiveness to P. falciparum and P. berghei when compared to non-Wolbachia-infected control mosquitoes [15,16].A study by Chen et al. [17] showed that the microbiome of An. stephensi remains unaffected upon Wolbachia infection, suggesting that Wolbachia-mediated Plasmodium suppression does not involve the mosquito microbiome.Two other studies, by Kambris et al. [6] and Hughes et al. [5], have indicated that laboratory-reared An. gambiae that are transiently infected with wMelPop and wAlbB become resistant to P. berghei and P. falciparum infection.However, several studies have shown that the effect of Wolbachia on Plasmodium infection in Anopheles can vary depending on the Wolbachia strain and the parasite and mosquito species involved.For example, two studies have shown that natural Wolbachia infection in An. coluzzii and An.gambiae field mosquitoes is negatively correlated with Plasmodium development [11,18].Other studies have indicated that the Wolbachia wPip strain renders Culex pipiens mosquitoes more permissive to the avian malaria parasite P. relictum [19] and that Wolbachia infection does not influence P. falciparum development in An. moucheti mosquitoes [20].
The molecular mechanisms underlying Wolbachia-mediated anti-Plasmodium activity in mosquitoes are not completely understood.Bian et al. (2013) have shown that wAlbB-induced

PLOS PATHOGENS
Wolbachia-induced anti-Plasmodium immunity reactive oxygen species (ROS) potentially play a role in the mosquitoes' resistance to Plasmodium infection [15].Similarly, Wolbachia induced ROS-dependent activation of the Toll pathway is associated with anti-viral protection in Ae. aegypti [21].Furthermore, Joshi et al. (2017) have shown that mosquito immune genes such as defensin, TEP1, PGRP and LRMs are upregulated upon Wolbachia infection and are possibly involved in regulation of P. berghei infection in wAlbB-infected mosquitoes [16].
To examine the role of the mosquito's immune response in Wolbachia-mediated anti-Plasmodium activity, we have now analyzed the midgut and carcass transcriptomes of sugar-fed versus blood-fed An. stephensi that are stably infected with Wolbachia (wAlbB).Focusing on the immune genes that were significantly regulated by Wolbachia infection, we investigated a potential role for selected upregulated genes in modulating Plasmodium infection in An. stephensi.We demonstrated that silencing of Wolbachia-regulated TEP 4, TEP 15, lysozyme C2, and CLIPB2, CLIPB4, PGRP-LD, peritrophin-44 like, and macrodomain genes leads to significant increases in P. falciparum infection intensity in An. stephensi mosquitoes, indicating that these genes may suppress Plasmodium infection in Wolbachia-infected Anopheles.

Changes in the midgut and carcass transcriptomes of An. stephensi in response to Wolbachia infection
To determine the effect of Wolbachia infection on gene expression in the midguts and carcasses (the mosquito whole body without the midgut) of An. stephensi, we compared midgut and carcass transcriptomes between Wolbachia-infected (LB1) and uninfected (LIS) females at 1-day post-blood feeding (blood-fed) or without blood feeding (sugar-fed).Overall, ~90% of the sequencing data could be aligned with the An.stephensi genome (S1 Table ).Principal component analysis (PCA) showed four distinct clusters for the sugar-fed midguts, blood-fed midguts, sugar-fed carcasses, and blood-fed carcasses (Fig 1A).In each cluster, three Wolbachiainfected or uninfected samples were associated with each other, except in the cluster of sugarfed carcasses, in which LISCA2, LB1CA2, and LB1CA3 transcriptomes were closely related (Fig 1A ).Hundreds of genes were significantly upregulated or downregulated in response to Wolbachia infection in midguts and carcasses, with and without blood feeding (Fig 1B), but the number of Wolbachia-regulated genes was higher in blood-fed mosquitoes than in sugarfed mosquitoes.Five and ten shared genes were significantly upregulated and downregulated, respectively, in Wolbachia-infected samples as compared to their corresponding uninfected samples (Fig 1C ).

Enrichment of Immune genes in the carcasses of Wolbachia-infected An. stephensi
GO analyses of upregulated and downregulated DE genes showed that innate immune response process and defense response were among the significantly regulated biological process (BP) and molecular function (MF) groups that also included metabolic process, singleorganism process, hydrolase activity, and lipid binding groups (Fig 2).These predicted immune and defense response genes include genes involved in innate immune responses, immune system processes, defense responses, antibacterial humoral responses, and oxidoreductase, suggesting Wolbachia infection may modulate the immune and stress responses in mosquitoes.Previous studies have shown upregulation of six immune genes, TEP1, REL1, PGRPLC, DEF1, LRM1, and CAT1, in Wolbachia-infected midguts and carcass tissues [16].Consistently, in carcasses of blood-fed mosquitoes, we found that Wolbachia infection led to

PLOS PATHOGENS
Wolbachia-induced anti-Plasmodium immunity 31 upregulated and 15 downregulated immune genes (Fig 3A and S2 Table), and the upregulated genes included predicted AMP, leucin-rich repeats (LRMs), defensin, lysozyme, C-type lectin-like (CTL), and CLIP domain serine protease and TEP genes (Fig 3B and S2 Table).In carcasses of sugar-fed mosquitoes, 11 and 9 immune genes were upregulated and downregulated, respectively, in Wolbachia-infected mosquitoes as compared to uninfected mosquitoes, and also among the many upregulated immune genes were CTL and CLIP-domain serine protease genes; among the downregulated genes were three AMP genes (Fig 3C).These data suggest that Wolbachia infection induces a more prominent immune response in blood-fed mosquitoes than in sugar-fed mosquitoes.
We also compared the differential expression of immune genes in the midguts of Wolbachia-infected and uninfected blood-fed and sugar-fed mosquitoes.In the midguts of bloodfed mosquitoes, 10 and 14 immune genes were upregulated and downregulated, respectively, by the presence of Wolbachia, and these genes had diverse functions (Fig 3D).However, Wolbachia infection led to the upregulation of only three genes and downregulation of only one gene in the midguts of sugar-fed mosquitoes (Fig 3E).Therefore, transcriptome analysis of both the carcasses and midguts supports the conclusion that Wolbachia-induced immune responses mainly occur in mosquitoes after ingestion of a blood meal, likely as a result of the increase in Wolbachia density that occurs upon blood feeding [22].Our analysis also indicates that Wolbachia infection stimulates a weaker immune response in midguts than in carcasses.

Testing selected Wolbachia-regulated immune genes for anti-Plasmodium activity
Since stable Wolbachia (wAlbB) infection can suppress Plasmodium in An. stephensi and modulate expression of many immune genes, we hypothesized that at least some of these immune genes are likely to play a role in regulating Plasmodium infection in mosquitoes.To test this hypothesis, we selected 18 genes that were significantly regulated by Wolbachia infection (Table 1).Among these genes, three upregulated genes encoding the CAP10 domain (ASTEI01800), macrodomain (ASTEI04078), and DUF4794 domain (ASTEI06123) and three downregulated genes encoding carboxypeptidase A (ASTEI04483), peritrophin-44 like
We used qPCR to validate the differential expression of the selected genes in midgut and carcass tissues of the wAlbB-infected and uninfected An. stephensi.Among the selected genes, lysC2, CLIPB2, 3-glucan binding protein, CLIPB4, PGRP-LD, CLIP-like, TEP4, TEP15, and TEP14 showed significantly higher expression in both blood-fed midguts and carcasses of the wAlbB-infected (LB1) mosquitoes than of the uninfected (LIS) An. stephensi (Fig 4A -4L).Of these genes, CLIPs, TEPs, PGRP-LD, and PPO have previously been linked to inhibitory or antagonistic effects on Plasmodium development [27][28][29][30][31].Some novel genes, namely, peritrophin-44 like, DUF4764 domain-containing, CAP10 domain-containing, Macro domain-containing, and transcription factor MafK, also showed significantly higher expression in both the midguts and carcasses of the wAIbB-infected mosquitoes than in uninfected An. stephensi (Fig 4M -4R).However, the role of these genes in anti-Plasmodium activity has not previously been examined.

Increase in P. falciparum infection after silencing of the selected Wolbachia-regulated genes in An. stephensi
To determine whether a wAlbB infection-response gene can influence mosquito permissiveness to P. falciparum infection, we investigated the effect of gene silencing on P. falciparum infection in non-Wolbachia-infected An. stephensi.The silencing efficiency of the selected genes is shown in S1  in a significantly (p<0.001) higher intensity of infection with P. falciparum (as measured by oocyst numbers in the midgut at 7 days post-feeding on Plasmodium gametocytes) as compared to the GFP control.Similarly, knockdown of two TEP genes, TEP4 and TEP15, also resulted in a significantly higher oocyst load, indicating a likely role for these genes in Wolbachia-mediated Plasmodium suppression (Fig 6C).Furthermore, significantly increased oocyst counts were observed upon silencing the peritrophin-44-like and macro domain genes (Fig 6D).Notably, the infection prevalence was significantly increased in CLIPB2, lysC2, CLIPB4, TEP4, TEP15, and perotrophin-44 like gene-silenced mosquitoes, but not significantly in PGRP-LD-and macro domain-silenced mosquitoes (Fig 6E -6H).

PLOS PATHOGENS
Wolbachia-induced anti-Plasmodium immunity In summary, silencing of 8 of the 18 selected Wolbachia-regulated genes significantly increased the P. falciparum infection intensity in An. stephensi.Our results, for the first time, provide evidence for a significant contribution of Wolbachia-induced upregulation of the CLIPs, TEPs, lysC2, PGRP-LD, peritrophin-44 like and macro domain genes to suppress Plasmodium infection and development in Anopheles.

Orthologs of selected Wolbachia-regulated An. stephensi genes
To gain insight into the possible orthologs and function of the selected Wolbachia-induced An. stephensi genes, we performed a phylogenetic analysis using the protein sequence of the putative orthologs from An. gambiae or other insect species.The sequence alignment and phylogenetic analysis of three predicted CLIPs gene showed that two (ASTEI08922 and ASTEI08923) were clustered together with the An.gambiae CLIPB4 and CLIPB2 branches, Mann-Whitney test between sugar fed versus blood fed samples (A) and P. falciparum uninfected versus infected samples (B); * P < 0.04; ** P < 0.02; *** P < 0.002.https://doi.org/10.1371/journal.ppat.1012145.g005The neighbor-joining tree for all the selected genes was subjected to numerical re-sampling by bootstrapping, and the resulting bootstrap values were depicted at the tree branch nodes.Each value indicates the number of times (out of 1000 replicates) that the identified orthologs occurred in the re-samplings.

Discussion
The intracellular bacterium Wolbachia has an extensively documented ability to block the infection of a number of insect hosts with various viruses and parasites, including several Drosophila viruses, the Aedes-vectored dengue and chikungunya viruses, and filarial nematodes, as well as Plasmodium in Anopheles mosquitoes [32][33][34][35][36].The means by which Wolbachia blocks insect-borne pathogens are likely to rely on multiple mechanisms and to represent a diversity of insect-pathogen infection models [37][38][39][40].Here we have investigated the basis of Wolbachia based anti-Plasmodium defense mechanisms in the An.stephensi mosquito malaria vector.We compared midgut and carcass transcriptomes between Wolbachia-infected and uninfected An. stephensi with and without blood feeding and found that Wolbachia infection significantly increased the expression of numerous immune genes in blood-fed midguts and carcasses and that some of the Wolbachia-modulated genes showed significant anti-Plasmodium activity.Our results provide insights into how Wolbachia-modulated immunity regulates infection with the human malaria parasite in An. stephensi.
Previous studies have demonstrated a major role for mosquito immunity in the defense against Plasmodium infection [41][42][43], and Plasmodium infection has been shown to upregulate the expression of a variety of mosquito immune genes [44]; apparently, anti-Plasmodium defenses are significantly regulated at the transcriptional level.Given the significant upregulation of numerous immune genes in Wolbachia-infected An. stephensi, we selected 18 of them in order to evaluate their potential role in defending against Plasmodium infection.Serine proteases (CLIPB2 and CLIPB4), TEPs, lysozyme C2 (lysC2), peptidoglycan recognition protein 1 like (PGRP-LD) and two novel genes, namely peritrophin-44 like and macro domain, were chosen because they showed significantly higher expression in both Wolbachia-infected An. stephensi and blood-fed midguts; these genes were also linked to mosquito immune defense against parasitic infection and infection with other pathogens [15,45,46].
Two CLIPs (CLIPB2 and CLIPB4) showed significantly higher mRNA abundance in Wolbachia infected blood-fed midgut and carcass tissues as well as in Wolbachia-uninfected P. falciparum infected An. stephensi (Figs 4B and 4F and 5B).Interestingly, the silencing of CLIPB2 and CLIPB4 significantly increased the oocyst loads and infection prevalence in the mosquito midguts, suggesting that these factors are P. falciparum antagonists that mediate Wolbachiabased suppression of Plasmodium.CLIP-domain serine proteases are a diverse group of proteolytic enzymes that are frequently involved in immune response signaling and amplification cascades (i.e., such as the Toll and PPO pathways) as well as developmental pathways (I.e., the Toll pathway) [47][48][49].Studies of An. gambiae CLIPs have shown that the CLIP-B and -C

PLOS PATHOGENS
Wolbachia-induced anti-Plasmodium immunity families, including CLIPB4, CLIPB8, CLIPB9, CLIPB10, CLIPB14, CLIPB17, and CLIPC9, are upregulated during the parasite's traversal of the midgut, and that they participate in the anti-Plasmodium defense by regulating P. berghei ookinete melanization [50][51][52].Wang et al. (2021) have reported that silencing of CLIPB15 leads to a significant decrease in Phenoloxidase (PO) activity, which acts as a catalyst for the formation of active intermediates of quinone for the synthesis of melanin, but silencing of CLIPB22 does not alter the PO activity [48].Silencing of both CLIPB15 and CLIPB22 affects the survival of Ae. aegypti after pathogenic bacterial infection [48].Consistent with these studies, we identified three CLIP domain-containing serine proteases, two of which, CLIPB2 and CLIPB4, were primarily upregulated in Wolbachiainfected blood-fed An. stephensi midguts and showed a significant anti-Plasmodium activity based on gene-silencing assays.Since CLIP domain serine proteases mediate anti-Plasmodium defenses as part of a complex network of regulatory cascades, it is impossible to predict all the effects of Wolbachia-induced altered expression of these genes.However, given the ability of serine proteases to act against Plasmodium infection, these CLIPs have the potential to be important for mosquitoes' anti-Plasmodium immunity.
In addition, we observed upregulated expression of lysozyme C2 in the Wolbachia-infected tissues and Wolbachia-uninfected P. falciparum infected mosquitoes (Figs 4A and 5B).dsRNA-mediated knockdown of lysozyme C2 resulted in a significant increase in P. falciparum infection intensity and prevalence when compared to those of the control GFP dsRNAinjected mosquitoes, suggesting it is a Wolbachia-inducible P. falciparum antagonist.Interestingly, the An.gambiae lysozyme C-1 (lysc1) is a protective Plasmodium agonist [53] and has been shown to inhibit the melanization of non-biologic targets, thereby playing an opposite role to that of the An.stephensi lysozyme C2 [54].Studies have also shown that An. gambiae lysC1 interacts directly with Plasmodium oocysts, and reducing lysC1 lowers the parasite load in the mosquito host [53][54][55].The increased oocyst load that we observed upon silencing of lysozyme C2 (lysC2) suggests a possible interaction with oocysts.However, further investigation of a possible interaction between lysC2 and Plasmodium will be required to understand the complexities of these intricate relationships.Based on the increased abundance of PGRP-LD (ASTEI02158) gene mRNAs in Wolbachia-infected tissues and Wolbachia-uninfected P. falciparum infected mosquitoes (Figs 4G and 5B), we silenced PGRP-LD in Wolbachia uninfected An. stephensi mosquitoes followed by P. falciparum infection.Silencing of the predicted PGRP-LD (ASTEI02158) also led to an increased susceptibility to P. falciparum infection.This result is consistent with a previous study showing that PGRP-LD (ASTEI010245) maintains the homeostasis of the gut microbiota by negatively regulating innate immune responses and protecting the An.stephensi mosquitoes from malaria parasite infection [26].These results further indicate that different isoforms of the PGRP-LD gene may exist in the An.stephensi genome.
Apart from CLIPs, lysC2, and PGRP-LD, we also selected three TEP genes whose expression levels were significantly higher in the Wolbachia-infected midgut tissue compared to the uninfected control.Significantly higher oocyst numbers were observed in TEP4-and TEP15silenced mosquitoes compared to the control GFP dsRNA injected mosquitoes.However, the silencing of TEP 14 (ASTEI0664) did not produce a phenotype similar to that observed for the other two TEP genes.A previous study has shown that transient wMelPop somatic infection in An. gambiae induces TEP1 expression, and Wolbachia-induced TEP1 upregulation contributes to the Plasmodium suppression [6].
From the pool of Wolbachia-regulated An. stephensi genes, we also selected six genes with no predicted immunity or anti-Plasmodium functions; peritrophin-44-like, DUF4794 domain, macro domain, transcription factor MafK, carboxypeptidase A, and CAP10 domain.We assessed their potential role in modulating Plasmodium infection using our standard gene-silencing and infection assays.These genes showed increased expression in Wolbachia LB1-infected blood-fed midgut tissues and in P. falciparum-infected An. stephensi.RNAi-mediated knockdown of the DUF4794 domain, transcription factor MafK, carboxypeptidase A, and CAP10 domain genes had no effect on Plasmodium infection; however, knockdown of peritrophin-44-like and macro domain -encoding gene led to a significant increase in the number of oocysts per midgut.While the oocyst counts were increased after knockdown of either macro domain or peritrophin-44, only silencing of the peritrophin-44-like protein gene had a statistically significant effect on infection prevalence.Macro domain-containing proteins (ASTEI09413) are not well characterized in insects [56,57]; therefore, an association with the Wolbachia-induced anti-Plasmodium defense could be an interesting new field to explore.Similarly, little information is available about the peritrophin-44-like gene in insect-parasite interactions.Elvin et al. (1996) characterized peritrophin-44 in the fly Lucilia cuprina and suggested that it plays important roles in the maintenance of insect gut structure, facilitation of digestion, and protection of digestive epithelial cells from bacterial damage and parasitic invasion [58].Additionally, among the 18 selected genes, those encoding the predicted CLIP-like, 3-glucan binding protein, DUF4794 domain, transcription factor MafK, CAP10 domain, carboxypeptidase A, and TEP14 did not show any influence on P. falciparum infection upon genesilencing (Figs 4 and 6).
Altogether, our study demonstrates that Wolbachia-regulated immune genes such as CLIP domain serine proteases, lysozyme C2, PGRP-LD, and TEPs have a significant inhibitory or antagonistic effect on Plasmodium development and are therefore likely to mediate Wolbachia's suppression of P. falciparum infection.Also, we show here that two novel genes, macro domain containing protein and peritrophin-44, also play a likely role in Wolbachia-mediated Plasmodium suppression.We did not perform gene expression or mRNA silencing assays with co-infections (Wolbachia and Plasmodium) due to difficulties with data interpretation when tripartite (Wolbachia-Anopheles-Plasmodium) interactions are addressed.Silencing of a Wolbachia induced gene in a coinfection assay could for example either affect Plasmodium directly, or indirectly through a change in Wolbachia infection.To our knowledge, this is the first-time links between these genes and the Wolbachia infection-modulated Plasmodium suppression has been established in a stably Wolbachia-transinfected mosquito and therefore warrants further investigation.Furthermore, we assessed effects of Wolbachia infection on the early midgut stages that will be influenced by Wolbachia-regulated gene expression occurring prior to Plasmodium introduction to the mosquito midgut.A greater understanding of these interactions may facilitate the development of future Wolbachia-based malaria control strategies as has been done for dengue control.

Ethics statement
This study was conducted in accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health, the Animal Care and Use Committee (ACUC) of Johns Hopkins University, and the institutional Ethics Committee.The Institutional Animal Care and Use Committee (IACUC) approved the protocol RA21H388.Mice were used for mosquitoes rearing.Anonymous, commercial blood from human donors was used for Plasmodium falciparum gametocyte cultures and infection assays in mosquitoes.

RNA extraction and real-time PCR
Total RNA was extracted from the pool of ten laboratory-reared sugar-and blood-fed An. stephensi mosquitoes using TRIzol (Invitrogen, USA) followed by Turbo Dnase I treatment.Similarly, midguts and carcass tissues were dissected in 1X PBS from Wolbachia-infected (wAlbB or LB1) or uninfected 5-to 6-day-old female LIS mosquitoes and collected in TRIzol reagent for RNA extraction.Complementary DNA (cDNA) was synthesized using 1 μg of total RNA with oligo dT primers and Moloney murine leukemia virus (MMLV) reverse transcriptase (Promega, USA).Real-time PCR was done using SYBR green PCR Master Mix (Applied Biosystem) with a final volume of 15 μl in a StepOne real-time PCR system (Applied Biosystem).For all assays, the expression of selected genes was normalized to the expression of the ribosomal protein S7 gene (Gene ID-MF999156.1).Three replicates were used per gene and, for tissue-specific expression, samples with five females were pooled to make one replicate.The relative expression of selected genes was calculated as a 2 -ΔΔCT value between sugar-fed and blood-fed, P. falciparum-infected and uninfected, and Wolbachia-infected (LB1) and uninfected An. stephensi.Specifically, we normalized the expression reads of blood-fed with sugar fed and P. falciparum infected with uninfected.Similarly, Wolbachia uninfected was used as normalizing control for assessing the expression of each gene in Wolbachia infected tissues.The sequences of all primers are given in S3 Table.

RNA sequencing and bioinformatics analysis
Total RNA was extracted from midguts and carcasses of Wolbachia-infected (LB1) and uninfected (LIS) one-week-old female mosquitoes using TRIzol.The quality of the RNA was assessed by Agilent 2100, and mRNA was enriched using oligo (dT) beads.An Illumina sequencing library was constructed for each RNA sample according to the manufacturer's instructions and sequenced by Novogene Co., LTD (Beijing, China) on the Illumina platform with paired-end 150 bp (PE 150).Data was deposited in NCBI's Sequence Read Archive (SRA: SUB14206326).Raw data in FASTQ format were processed to remove reads containing adapters, reads containing ploy-N, and low-quality reads; clean reads were aligned to the An.stephensi genome (VectorBase-57).Feature Counts was used to quantify transcript abundance in each sample by using the gene annotation obtained from VectorBase.Differentially expressed (DE) genes between LB1 and LIS mosquitoes in midguts and carcasses were identified with DESeq2.Gene ontology (GO) and -fold enrichment of the DE genes was analyzed using the built-in program in Vectorbase or ShinyGO (http://bioinformatics.sdstate.edu/go/).The interaction networks of the DE genes were generated with Revigo (http://revigo.irb.hr/) and modified by Cytoscape [59].

dsRNA synthesis and RNAi-mediated gene silencing
PCR fragments of 600-700 bp were amplified with each selected gene-specific primer, tailed with a short T7 promoter sequence, 5'taatacgactcactataggg'3, using the cDNA from sugar-fed An. stephensi female mosquitoes as template.Each PCR fragment was purified, and specific dsRNA was synthesized using the HiScribeT7 Quick HighYield RNA synthesis kit (New England Biolabs) according to the manufacturer's protocol.The PCR fragment for GFP that served as a control was amplified using a plasmid template containing the GFP gene [60].The concentration and quality of the dsRNA were determined spectrophotometrically by Nano-drop1000 (company) and agarose gel electrophoresis.The gene-specific primers used for

PLOS PATHOGENS
Wolbachia-induced anti-Plasmodium immunity dsRNA synthesis are summarized in S3 Table .For nano-injections, three-day-old An. stephensi Lis female mosquitoes were cold-anesthetized and injected intrathoracically with 69 nl of a 3μg/ ul dsRNA for each gene of interest.A control group of mosquitoes were injected with dsGFP.All injections were repeated 3-5 times using a Nanoject microinjector, and approximately 80-100 mosquitoes were silenced for each gene per experiment.Each biological replicate corresponds to a different mosquito population cage.After dsRNA injection, mosquitoes were left for 3 days under optimal insectary conditions, with 10% sucrose as the sugar source [23,61].

Gene silencing efficiency
The efficiency of gene silencing was determined 3 days post-dsRNA injection by real-time quantitative reverse transcription-PCR (qRT-PCR) for all the selected genes, along with control dsGFP.Total RNA was extracted from a pool of five to six mosquitoes, followed by cDNA synthesis as described in the previous section.qPCR was performed using the respective gene primers in a StepOnePlus real-time PCR machine as mentioned previously.The primer sequences used for silencing validation are given in S3 Table.

Plasmodium falciparum infection
To determine the anti-Plasmodium activity, gene-knockdown female mosquitoes were starved for 4-5 h.In brief, 14-to 16-day-old P. falciparum NF54 cultures were diluted to 0.1-0.3% with fresh red blood cells (RBCs), and 60% human serum was added to the final volume of infected blood.Starved female mosquitoes were allowed to feed on the infected blood through artificial membrane feeders maintained at 37˚C for 30-45 min [23,61].After feeding, the unfed females were removed, and the fed mosquitoes were kept for 8-10 days at 27˚C for oocyst counting.Midguts were dissected on the 8 th day post-feeding in phosphate-buffered saline (PBS) and stained with 0.2% mercurochrome to determine the oocyst load under a light microscope.

Phylogenetic analysis
The full-length sequences of selected genes and their putative orthologs in An. gambiae were retrieved from VectorBase and aligned in the FASTA format using the Mafft multiple sequence alignment tool.Each amino acid residue was aligned pairwise and compared with other residues of the same row, and identical residues between the species were marked with colored residues based on their biochemical properties.Phylogenetic relationships between the An.stephensi selected gene sequences and their respective orthologs from An. gambiae and An.coluzzii were analyzed using MEGA 11.0.13[62].Phylogenetic trees were constructed by neighbor-joining method.Protein sequences of the target genes of An. stephensi were aligned with their corresponding putative orthologs from An. gambiae and An.coluzzii, and phylogenetic analysis was performed.

Statistical analysis
For Plasmodium infection experiments, the dot plots for oocysts/midgut were generated with GraphPad Prism5 software.Statistical differences between three independent biological replicates were assessed, and the data were pooled.The significance of the differences in infection load (the number of oocysts per midgut) and prevalence (the number of infected mosquitoes per total number of mosquitoes examined) between the control dsGFP and the gene-silenced groups were determined through nonparametric Mann-Whitney tests and the Fisher's exact test, respectively.Two-tailed P values are given for all experiments in the figure legends along with information concerning the total number of midguts, median, and infection prevalence.Data from three biological replicates for each experiment were shown in all figures.

Fig 6 .
Fig 6.Effect of silencing of the 18 selected immune genes on Plasmodium falciparum infection in An. stephensi.Female mosquitoes were injected with dsRNA of each of the immune genes; dsRNA of GFP was used as a control.(A-D) P. falciparum intensity in individual midguts of immune gene-silenced mosquitoes.Each dot indicates the number of oocysts in the individual midgut, and horizontal red bars indicate the median value.(E-H) P. falciparum infection prevalence in immune gene-silenced mosquitoes.Two-tailed P values obtained with the Mann-Whitney test (infection intensity) or Fisher exact test (infection prevalence) are shown: *, P < 0.05; **, P < 0.01; ***, P < 0.001.https://doi.org/10.1371/journal.ppat.1012145.g006