Cell wall components of gut commensal bacteria stimulate peritrophic matrix formation in malaria vector mosquitoes through activation of the IMD pathway

In our previous studies, we demonstrated that Enterobacter hormaechei, isolated from our colony mosquitoes, has the ability to stimulate PM formation in An. stephensi [22]. To investigate which component of commensal bacterium stimulates PM formation, we administrated mosquitoes, in which gut commensal bacteria were reduced by antibiotics (Abx) treatment, with the supernatant and cell pellets of the overnight culture of E. hormaechei (Fig 1A). The efficacy of bacterial clearance by antibiotics was evaluated by qPCR and plate culture methods (S1 Fig). In mosquitoes fed with E. hormaechei pellets, the bacterial load reached a similar level as that in normal mosquitoes, around 10⁷–10⁸ colony-forming units (CFU), at 45 h after blood feeding. However, the bacterial load in mosquitoes treated with the bacterial supernatant and Abx-treated mosquitoes remained at the same level (Fig 1B). We next analyzed the mRNA and protein levels of Per1, a gene involved in PM formation, in these mosquitoes. Administration of E. hormaechei pellets significantly up-regulated the expression of Per1 mRNA at 24 and 45 h after blood feeding, and Per1 protein at 45 h after blood feeding, reaching levels comparable to those found in the normal mosquito midgut. However, treatment with bacterial supernatant failed to stimulate Per1 expression (Fig 1C and 1D). Consistent with these findings, fully formed PMs were observed by immunofluorescent staining against Per1 and calcofluor staining in mosquitoes treated with bacterial pellets and in normal mosquitoes at 45 h after blood meal, while PM formation was impaired in mosquitoes treated with bacterial supernatant or Abx-treated mosquitoes (Fig 1E and 1F). Based on these results, we conclude that bacterial secretion does not play a role in promoting PM formation. Instead, components present in the bacterial cells are responsible for stimulating PM formation in mosquitoes.

Fig 1. Identification of the effective component of E. hormaechei in promoting PM formation.

(A) Schematic overview of the study design. (B) Bacterial loads in the midgut of normal mosquitoes and mosquitoes with different treatments at 45 h post a blood meal. Abx, antibiotic-treated mosquitoes; Pellets, Abx mosquitoes supplemented with cell pellets of E. hormaechei; Supernatant, Abx mosquitoes supplemented with cell supernatant of E. hormaechei. (C, D) The levels of Per1 mRNA (C) and protein (D) in the midgut of the same treated mosquitoes as in (B). The expression level of Per1 was normalized to S7. The relative expression level of Per1 in treated mosquitoes was normalized to gene expression in Normal mosquitoes. Quantification of Per1 intensity was shown on the panel below in (D). Data are presented as mean ± SEM (n = 5 in B, n = 8~10 in C, and n = 5 in D). (E) Immunofluorescent staining of Per1 (green) in the midgut of the same treated mosquitoes as in (B) at ×200 magnification. Nuclei were stained with DAPI (blue). (F) Calcofluor white staining PM structure in the midgut of the same treated mosquitoes as in (B) at ×200 magnification. Scale bars, 50 μm. Significance was determined by one-way ANOVA followed by Dunnett’s multiple comparison test. *, P P P S1 Data. The uncropped blots are included in S1 Raw Images. PM, peritrophic matrix.

https://doi.org/10.1371/journal.pbio.3002967.g001

A bacterial cell typically consists of a cell wall that surrounds an internal matrix called the cytoplasm [25]. To identify the bacterial effectors responsible for promoting PM formation, we isolated the bacterial cell wall and intracellular contents from an overnight culture of E. hormaechei. These isolated components were individually administered to Abx-treated mosquitoes through a blood meal (Fig 2A). Supplementation of the bacterial cell wall increased the protein level of Per1, indicating its role in stimulating PM formation. However, the intracellular contents had no influence on Per1 protein level (Fig 2B).

Fig 2. Identification of the bacterial cell component in PM formation.

(A) Schematic overview of the study design. (B) Western blot of Per1 in the midguts of normal, Abx mosquitoes and Abx mosquitoes treated with cell wall and cell content at 45 h post a blood meal. (C–E) The expression levels of Per1 of mosquitoes orally supplemented with serial concentrations of DAP-PGN (C), Lys-PGN (D), LPS (E), at 24 and 45 h post a blood meal. (F, G) The expression levels of Per1 mRNA (F) and protein (G) in the midgut of Normal, Abx, and Abx mosquitoes treated with the mixture of DAP-PGN, Lys-PGN, and LPS (PGNs+LPS) at 24 and 45 h post a blood meal. The expression level of the target gene was normalized to S7. The relative expression level of Per1 was normalized to gene expression in Abx or Normal mosquitoes. Quantification of Per1 intensity was shown on the right panel in (G). Data are presented as mean ± SEM (n = 8~10 in C, n = 7~10 in D, n = 8~10 in E, and n = 10 in F and n = 4 in G). (H) Immunofluorescent staining of Per1 (green) in midguts of Normal, Abx, and PGNs+LPS treated mosquitoes at 45 h post a blood meal at ×200 magnification. Nuclei were stained with DAPI (blue). (I) Calcofluor white staining PM structure in the midgut of Normal, Abx, and PGNs+LPS treated mosquitoes at 45 h post a blood meal at ×200 magnification. Scale bars, 50 μm. Significance was determined by one-way ANOVA followed by Dunnett’s multiple comparison test. *, P P P P S1 Data. The uncropped blots are included in S1 Raw Images. DAP-PGN, diaminopimelic acid-peptidoglycan; LPS, lipopolysaccharide; Lys-PGN, lysine-peptidoglycan; PM, peritrophic matrix.

https://doi.org/10.1371/journal.pbio.3002967.g002

Peptidoglycan, including DAP-PGN and Lys-PGN, as well as LPS, are major components of the bacterial cell wall [26]. We next fed Abx-treated mosquitoes with blood containing varying concentrations of DAP-PGN, Lys-PGN, or LPS individually, and analyzed Per1 expression levels of at 24 h and 45 h post-blood feeding. All 3 components stimulated Per1 expression at 24 h post-blood feeding, while DAP-PGN continued to elevate Per1 expression at 45 h (Fig 2C–2E), suggesting that DAP-PGN elicits a more sustained induction of Per1 than other cell wall components. Since the induction of Per1 by LPS may be due to the PGN contamination [27,28], we next treated LPS with mutanolysin, a PGN lytic enzyme [27], and then fed mosquitoes with the treated LPS via blood. Again, significant up-regulation of Per1 expression was observed in both mutanolysin-untreated and treated LPS mosquitoes (S2 Fig), suggesting that LPS is involved in the stimulation of PM formation. Considering that the mosquito gut microbiota consists of both gram-negative and gram-positive bacteria [29], we combined all 3 cell wall components and fed the mixture to Abx-treated mosquitoes via blood. As expected, oral administration of the PGNs/LPS mixture restored the mRNA and protein levels of Per1 to similar levels as those observed in normal mosquitoes (Fig 2F and 2G). Additionally, intact PM formation was observed in the midgut of mosquitoes orally supplemented with the mixture (Fig 2H and 2I). Taken together, these results indicate that the components of the bacterial cell wall, including DAP-PGN, Lys-PGN, and LPS, play a crucial role in stimulating PM formation.

Peritrophins are stored in the secretory vesicles of midgut epithelial cells before blood feeding and are released into the gut lumen immediately after mosquitoes ingest a blood meal [30–32]. To investigate whether the elimination of gut commensal bacteria influences Per1 level before a blood meal, we analyzed the expression level of Per1 in mosquitoes that were only fed sugar (Fig 3A). As expected, the mRNA and protein levels of Per1 were reduced significantly in Abx-treated mosquitoes compared to normal mosquitoes (Fig 3B and 3C). However, when we administered the PGNs/LPS mixture to Abx-treated mosquitoes via a sugar meal, we found that the mixture induced Per1 expression at both the mRNA and protein levels even without a blood meal (Fig 3B and 3C). Additionally, 2 other PM-related genes were also regulated by bacterial cell wall components (S3 Fig) [33–35]. These results suggest that the PGNs/LPS mixture stimulates PM genes expression independently of the type of food that the mosquito ingests. Even in the absence of a blood meal, the administration of the PGNs/LPS mixture can trigger PM gene expression, indicating that the bacterial cell wall components play a direct role in stimulating PM formation in the mosquito gut.

Fig 3. The influence of PGNs and LPS on Per1 expression before blood feeding.

(A) Schematic overview of the study design. (B, C) The levels of Per1 mRNA (B) and protein (C) in the midgut of Normal, Abx, Abx mosquitoes treated with DAP-PGN/Lys-PGNs/LPS mixture via sugar meal for 24 h. The expression level of Per1 was normalized to S7. The relative expression levels of Per1 in Abx, Abx supplemented with PGNs+LPS mosquitoes were normalized to gene expression in Normal mosquitoes. Quantification of Per1 intensity was shown on the right panel in (C). Data are presented as mean ± SEM (n = 7~10 in B, n = 4 in C). Significance was determined by one-way ANOVA followed by Dunnett’s multiple comparison test. *, P P P S1 Data. The uncropped blots are included in S1 Raw Images. DAP-PGN, diaminopimelic acid-peptidoglycan; LPS, lipopolysaccharide; Lys-PGN, lysine-peptidoglycan.

https://doi.org/10.1371/journal.pbio.3002967.g003

The cell wall components, particularly PGNs and LPS, are well-known activators of the mosquito innate immune system [36–38]. We hypothesized that the bacterial cell wall might regulate Per1 expression through immune signaling pathways. In Drosophila, the peptidoglycan recognition proteins PGRP-LC and -SA are responsible for the recognition of PGNs and act as receptors for the IMD and Toll pathways, respectively [39–42]. In mosquitoes, PGRP-LC is the key receptor of IMD pathway [43], while PGRP-S1 is a putative receptor of Toll pathway [44–46]. To investigate whether the IMD and Toll pathways regulate Per1 expression through sensing PGN and LPS, we conducted a co-silencing experiment targeting Pgrp-lc and –s1 (dsLC/S1) and examined the expression level of Per1 in mosquitoes before blood feeding. The mRNA and protein levels of Per1 were significantly reduced in mosquitoes with Pgrp-lc and Pgrp-s1 double knockdown compared to those with dsGFP controls (Fig 4A and 4B). As expected, the knockdown of Pgrp-lc and Pgrp-s1 inhibited the activities of the IMD and Toll signaling pathways, as indicated by the significantly reduced expression levels of antimicrobial peptides (AMPs), including Cecropin3, Gambicin, and Defensin (Fig 4A). The simultaneous decrease in the expression of Per1 and AMPs in dsLC/S1 mosquitoes suggests that these genes may be regulated by common transcription factors or signaling components downstream of the IMD and Toll pathways.

Fig 4. The influence of Toll and IMD pathways on Per1 expression.

(A) Relative expression level of Per1 and immune related genes in mosquitoes treated with dsLC/S1 and dsGFP (n = 10). (B) Western blot of Per1 in the midgut of dsGFP and dsLC/S1-treated mosquitoes. Quantification of Per1 intensity was shown on the right panel in (B). (C) The knocking down efficiency of Rel1 and Rel2 of mosquitoes (dsRel1/Rel2). (D, E) The levels of Per1 transcript (D) and protein (E) in dsRel1/Rel2 and dsGFP-treated mosquitoes prior to blood feeding. Quantification of Per1 intensity was shown on the right panel in (E). (F, G) The levels of Per1 transcript (F) and protein (G) in dsRel1/Rel2 and dsGFP-treated mosquitoes at 45 h post blood feeding. The expression levels of the target genes were normalized to S7. The relative expression level of immune genes in dsRel1/Rel2-treated mosquitoes was normalized to gene expression in dsGFP-treated mosquitoes. Quantification of Per1 intensity was shown on the right panel in (G). (H) The total gut microbiota load was measured in dsRel1/Rel2 and dsGFP treated mosquitoes prior to blood feeding. All Data are presented as mean ± SEM (n = 9~10 in A, n = 3 in B, n = 8~9 in C, n = 8~9 in D, n = 4, n = 9~10 in F, n = 5 in G, n = 8~10 in H). Significance was determined by Student’s t test. The microbiota load was determined by Mann–Whitney test. *, P P P P S1 Data. The uncropped blots are included in S1 Raw Images.

https://doi.org/10.1371/journal.pbio.3002967.g004

The NF-κB transcription factors Rel1 and Rel2, which belong to the Toll and IMD pathways, respectively, are the primary regulators of downstream immune effector expression. To investigate whether Rel1 and Rel2 regulate Per1 expression, we conducted a double knockdown experiment targeting Rel1 and Rel2 and analyzed Per1 expression. When Rel1 and Rel2 were simultaneously knocked down, both the mRNA and protein levels of Per1 were significantly reduced compared to the dsGFP control, regardless of whether mosquitoes were fed sugar or blood (Fig 4C–4G). Moreover, the expression levels of Per14 and Fibrinogen, which were induced by treatment with PGNs/LPS mixture in Abx mosquitoes, were also reduced in dsRel1/Rel2 mosquitoes (S4 Fig). Given that the NF-κB signaling pathways are responsible for controlling the abundance of gut microbiota that stimulates PM formation, we investigated whether the down-regulation of Per1 expression was due to alterations in gut bacteria abundance. The bacterial load in the midgut of dsRel1/Rel2 mosquitoes was significantly increased prior to and post blood meal due to the inhibition of mosquito Toll and IMD signaling activities (Figs 4H and S4C). However, despite the increased bacterial load, the expression level of Per1 in these mosquitoes remained down-regulated, further confirming that the inhibiting Rel1 and Rel2 activity prevents the initiation of Per1 expression. These results suggest that Rel1 and Rel2 directly regulate Per1 expression and underscore the importance of these NF-κB transcription factors in coordinating the immune response and PM formation in mosquitoes.

Mosquito Rel1 and Rel2 contain the Rel-homology domain (RHD), which initiates the transcription of target genes. We next co-expressed the RHD domains of Rel1 and Rel2, respectively, and investigated their activity on Per1 transcription using a dual-luciferase reporter assay. The pCMV6-Rel1-RHD and pCMV6-Rel2-RHD plasmids were co-transfected into 293T cells. The pGL3-Basic Reporter plasmids containing promoter fragment 1,589 bp upstream of the Per1 coding sequence was transfected into 293T cells (Fig 5A). The plasmid that constitutively expresses renilla luciferase (pRLTK-renilla) was co-transfected as an internal control. At 48 h post-transfection, Rel1 and Rel2 drove the expression of the firefly luciferase reporter successfully in the presence of the −1,589 bp promoter region (Fig 5B). We next analyzed the Rel1 and Rel2 recognition region by transfecting cells with different lengths of promoter fragments ranging from −1,589 to −229 bp upstream of the coding sequence. We found that the deletion of −289 to −229 bp resulted in the loss of capability to activate the firefly luciferase reporter expression, compared to transfection with the pGL3-Basic Reporter vector (Fig 5B). This result indicates that the binding motif of Rel1 and Rel2 may exist within the −289 bp to −229 bp region.

Fig 5. Identification of the Rel1/Rel2 binding motif in Per1 promoter.

(A) A diagram of Per1 promoter and the truncated constructs used in the transient transfection assays. The truncations of the Per1 promoter, which were sequentially truncated from the 5′-end of promoter region, derived from the −1,589 bp to −229 bp, were inserted into pGL3-Basic vector. The inserted promoters were followed by a firefly luciferase gene (green), thereby enabling the determination of the regulatory activity of the inserted promoters using dual luciferase reporter assay. The pGL3-Basic vector was function as control named as pGL3-NC. (B) Assessment of promoter activity for Per1 promoter truncations together with Rel1-RHD and Rel2-RHD plasmids. (C) Schematic representation of M1, M2, M3, M4 mutants, derived from the −289 bp to −229 bp in the Per1 promoter. M, mutation. (D) Assessment of promoter activity for Per1 promoter mutations together with Rel1-RHD and Rel2-RHD plasmids. The activity of Per1 promoter with different mutations was normalized to the activity of the control luciferase construct. Each experiment was performed at least 2 times independently and in duplicate. Data are presented as mean ± SEM. Significance was determined by one-way ANOVA followed by Dunnett’s multiple comparison test. *, P P P Per1 promoter by EMSA. The arrows indicate specific DNA-protein complex. The data underlying this figure can be found in S1 Data. The uncropped blots are included in S1 Raw Images. EMSA, electrophoretic mobility shift assay; RHD, Rel-homology domain.

https://doi.org/10.1371/journal.pbio.3002967.g005

To identify binding motifs within the minimal 61 bp region, we constructed 4 deletion mutants from −289 bp to −229 bp of the Per1 promoter sequence and designated them as M1 (−289 bp to −275 bp deletion), M2 (−274 bp to −260 bp deletion), M3 (−259 bp to −245 bp deletion), and M4 (−244 bp to −229 bp deletion), respectively (Fig 5C). Only M1 failed to activate firefly luciferase reporter expression, compared to wild-type controls (pGL3-489), indicating that the binding motif is located at the region from −289 bp to −275 bp of Per1 promoter (Fig 5C and 5D). To further validate this, we performed an electrophoretic mobility shift assay (EMSA). The pCMV6-Rel1-RHD and pCMV6-Rel2-RHD plasmids containing His-tag were co-transfected into 293T cells. When biotin-labeled probes were incubated with cell lysates, a shifted band was observed, indicating that Rel1 and Rel2 proteins bind to biotin-labeled probes. The band intensities gradually decreased with the addition of 5-fold, 50-fold, 100-fold, and 200-fold excess of unlabeled competitive oligonucleotide probes, suggesting that Rel1 and Rel2 proteins binding to probes specifically. When a mutant biotin-labeled probe deleted of −289 bp to −275 bp region was added, the shifted band was disappeared (Fig 5E and 5F). However, we observed a band larger than the shift band possibly due to the nonspecific binding of the mutant probe to the cell lysate. Furthermore, a super-shift band that represents the DNA-protein complex was detected when anti-His antibody was added, indicating that there is one binding site on the Per1 promoter for Rel1 and Rel2.

As Toll and IMD pathways play different roles in maintaining gut microbial homeostasis and defending against pathogens [47], we next investigated the individual contributions to PM formation. We knocked down Pgrp-s1, Pgrp-lc, Rel1, and Rel2 individually and analyzed Per1 expression levels (Fig 6). Knockdown of Pgrp-s1 or Rel1, associated with the Toll pathway, had no effect on Per1 expression (Fig 6A and 6B). In contrast, specific knockdown of Pgrp-lc and Rel2, key components of the IMD pathway, significantly down-regulated Per1 expression (Fig 6C and 6D). Consistent with the stronger effect of DAP-PGN on Per1 expression (Fig 2C), these findings strongly suggest that IMD pathway plays a major role in regulating PM formation.

Fig 6. The impact of single gene knockdown in Per1 expression.

The gene knockdown efficiency and Per1 expression levels in Pgrp-s1 (A), Rel1 (B), Pgrp-lc (C), and Rel2 (D) knockdown mosquitoes. Data are presented as mean ± SEM (n = 9~10 in A, n = 10 in B, n = 10 in C, n = 8~10 in D). Significance was determined by Student’s t test. **, P P P S1 Data.

https://doi.org/10.1371/journal.pbio.3002967.g006

An. stephensi (Strain Hor) larvae were reared in water that was supplemented with fish food daily. Adult mosquitoes were kept in a controlled environment at 28°C, 80% relative humidity and at a 12 h light/dark cycle according to standard procedures [70,71]. They were provided with a 10% sugar solution and BALB/c mice for blood feeding. For antibiotic treatment, newly emerged mosquitoes were given fresh filtered 10% sucrose containing 10 U/ml penicillin, 10 μg/ml streptomycin, and 15 μg/ml gentamicin daily for up to 5 days. To determine the efficacy of antibiotic treatment, mosquitoes were surface sterilized with 75% ethanol twice and 0.9% NaCl twice, then homogenized in 0.9% NaCl. The homogenate was plated on the LB agar plates. CFU were counted after incubating the plates at 28°C for 2 days and 5 days. The genomic DNA was extracted using the method of Holmes and Bonner as described [72] and the bacteria load was determined by qPCR using universal 16S rRNA primers (S1 Table). After removing antibiotics, the mosquitoes were reared in sterilized cages and provided with sterilized sucrose solution to maintain a sterile environment [73].

The purification procedure of bacterial cell wall was performed as described [74]. Briefly, the E. hormaechei was grown in LB medium at 28°C overnight. Then, 1 × 10⁷–10⁸/ml bacterial cells were harvested by centrifugation at 5, 000 g for 10 min at 4°C. The pellets were collected, washed with distilled endotoxin-free water, and boiled in 8% sodium dodecyl sulfate (SDS) (Sigma-Aldrich) for 30 min. Then, the cells were incubated at room temperature overnight. Afterward, they were centrifuged at 25,000 g for 20 min at 4°C. The resulting cell pellets were resuspended in 4% SDS and boiled for 15 min. Following this, the cell wall was harvested by centrifugation at 25,000 g for 20 min at 4°C and resuspended in distilled endotoxin-free water. This process was repeated 3 times. Finally, the pellets were resuspended in 2 M NaCl and collected by centrifugation at 25,000 g for 20 min at 4°C. The supernatant was discarded and the remaining pellets were dissolved in 10% sucrose or blood for oral feeding mosquitoes.

Briefly, the E. hormaechei was grown in LB medium at 28°C overnight. Then, 1 × 10⁷–1 × 10⁸/ml bacterial cells were harvested by centrifugation (5,000 g, 10 min, 4°C). The supernatant was discarded, and the cell pellets were washed 3 times with sterile 1× phosphate-buffered saline (1× PBS). The cell pellets were then resuspended in PBS and sonicated on ice in the presence of protease inhibitor (Beyotime). Cell contents were obtained by centrifugation at 143,000 g for 1 h at 4°C and filtered through a 0.22 μm membrane. Then, the filtrate was used for oral administration to mosquitoes via blood meal [75,76].

The oral administration of E. hormaechei was performed as described [77]. Briefly, the overnight culture E. hormaechei was harvested by centrifugation (5,000 g, 10 min, 4°C). The pellet and the supernatant were collected. The pellet was washed 2 times with PBS and dissolved in 1.5% sucrose to get the final concentration 1 × 10⁷–10⁸/ml. The supernatant was filtered through a 0.22 μm membrane before adding to 1.5% sucrose. The commercial Lys-PGN derived from Staphylococcus aureus (Sigma-Aldrich), DAP-PGN derived from Bacillus subtilis (Sigma-Aldrich), and LPS derived from E. coli (Sigma-Aldrich) were dissolved in sterile water and were administrated to antibiotics treated mosquitoes through a blood meal or 10% sucrose at dedicated concentration. To eliminate the PGN contamination in LPS, the LPS (1 mg/ml) (Sigma) were incubated with mutanolysin (Sigma) at 22°C for 12 h with shaking [78]. The mutanolysin-treated LPS was then orally administrated to Abx-treated mosquitoes through an artificial membrane-feeding system (Hemotek). The mosquito midgut was dissected at 24 h post blood feeding, and the gene expression of Per1 was analyzed by qPCR.

The immunofluorescence analysis of Per1 was performed as previously described [14]. Briefly, mosquito abdomens were collected 45 h after a blood meal and fixed in 4% paraformaldehyde at 4°C overnight. Samples were paraffin-embedded, sectioned at thickness of 5 μm, and stained with anti-Per1 (1:100) and Alexa Fluor 546 (1:5,000) (Thermo Fisher). Images were captured using a Nikon ECLIPSE IVi microscope connected to a Nikon DIGITAL SIGHT DS-U3 digital camera.

Total RNA was extracted from individual mosquitoes using the standard TRIzol reagent (Sigma-Aldrich). The cDNA was synthesized using the Hifair III 1st Strand cDNA Synthesis SuperMix for qPCR (Yeasen). The expression level of the target genes were determined by qPCR using a Roche LightCycler 96 Real Time PCR Detection System with SYBR Green qPCR Master Mix (Yeasen). Data were processed and analyzed using LightCycler 96 software. The ribosomal gene S7 of An. stephensi was used as an internal reference [79]. The relative expression of the target gene was normalized to S7 by the 2^-△△Ct method [80]. Seven to 10 mosquitoes were used for qPCR analysis each time. Primers were listed in supplementary S1 Table.

Mosquito midguts were dissected at 45 h post-blood meal and at least 10 midguts were pooled together. Proteins were extracted using lysis buffer (8 M urea, 2% SDS, 5% β-mercaptoethanol, 125 mM Tris-HCl). Immunoblotting was performed using standard procedures with rabbit anti-Per1 polyclonal antibody (1:5,000) [81], anti-Actin monoclonal antibody (1:2,000) (Abbkine), and secondary anti-rabbit-HRP (1:5,000) (Abbkine). Actin was used as a reference control [14,81]. Intensity of the signals was quantified by ImageJ software [82].

The cDNA clones of target genes were obtained using gene-specific primers (S1 Table). PCR amplicons of GFP (as a control) (BD Biosciences), Pgrp-lc (ASTE016447), Pgrp-s1 (ASTE007708), Rel1 (ASTE011378), Rel2 (ASTE010360) tailed with a T7 promoter (TAA TAC GAC TCA CTA TAG GGA GA) were used to synthesize dsRNA using a MEGAscript T7 High Yield Transcription Kit (Invitrogen). Four- to 5-day-old females were received a total of 69 nL dsRNA (3 to 4 μg/μl), which was injected intra-thoracically using a nanoject II microinjector (Drummond). Gene silencing efficiency was determined 2 days after injection. All experiment was replicated twice with separate cohorts of mosquitoes.

The promoter sequence, located 1,589 bp upstream of the Per1 coding sequence, and the different truncations of the An. stephensi Per1 promoter were amplified using the primers listed in S1 Table. The resulting PCR products were then cloned into plasmid pGL3-Basic vectors carrying the firefly luciferase reporter (Promega). The RHD of Rel1 and Rel2 were amplified from an An. stephensi cDNA using the primers listed in S1 Table, and then cloned into pCMV6-AC-HA vectors (OriGene), named as pCMV6-Rel1-RHD, and pCMV6-Rel2-RHD, respectively. The pRL-TK plasmid carrying renilla luciferase were used as an internal control (Promega). The 4 mutations of the Per1 promoter (M1-M4) were constructed by amplifying the 489 bp promoter region (pGL3-489) using the primers listed in S1 Table. Plasmids for transfection were prepared using the EndoFree Maxi Kit Plasmid Kit (TIANGEN).