Glucose transporter GLUT1 influences Plasmodium berghei infection in Anopheles stephensi

Background Sugar-feeding provides energy for mosquitoes. Facilitated glucose transporters (GLUTs) are responsible for the uptake of glucose in animals. However, knowledge of GLUTs function in Anopheles spp. is limited. Methods Phylogenetic analysis of GLUTs in Anopheles stephensi was performed by the maximum likelihood and Bayesian inference methods. The spatial and temporal expression patterns of four Asteglut genes were analyzed by qPCR. The function of Asteglut1 was examined using a dsRNA-mediated RNA interference method. Transcriptome analysis was used to investigate the global influence of Asteglut1 on mosquito physiology. Results We identified 4 glut genes, Asteglut1, Asteglutx, Asteglut3 and Asteglut4 in An. stephensi. Asteglut1, Asteglut3 and Asteglut4 were mainly expressed in the midgut. Plasmodium berghei infection differentially regulated the expression of Asteglut genes with significant downregulation of Asteglut1 and Asteglut4, while upregulation of Asteglutx. Only knocking-down Asteglut1 facilitated Plasmodium berghei infection in An. stephensi. This might be due to the accumulation of glucose prior to blood-feeding in dsAsteglut1-treated mosquitoes. Our transcriptome analysis revealed that knockdown of Asteglut1 differentially regulated expression of genes associated with multiple functional clusters, especially those related to detoxification and immunity. The dysregulation of multiple pathways might contribute to the increased P. berghei infection. Conclusions Our study shows that Asteglut1 participates in defense against P. berghei in An. stephensi. The regulation of Asteglut1 on vector competence might through modulating multiple biological processes, such as detoxification and immunity.


Background
Anopheles mosquitoes are the primary vectors of human malaria that kill over 450,000 people annually [1]. To transmit between mammalian hosts, malaria parasites have to complete multiple development processes in the mosquito including gametogenesis, fertilization, zygoteto-ookinete conversion and oocyst formation [2]. During this process, complicated interactions between Anopheles mosquitoes and Plasmodium parasites occur [3]. The nutrient availability is one of the key factors that determine the infection outcome [4,5]. Sugar is a key energy resource that influences survival and fecundity of mosquitoes [6]. It also affects the vector competence [6][7][8][9][10]. Trehalose, the main circulating sugar, is a non-reducing disaccharide composed of two glucose molecules linked by an α-α-1,1-glycosidic bond. It enters cell metabolism after catabolized into glucose [11,12]. Trehalose transporter AgTreT1 is responsible for the transportation of trehalose from fat body to hemolymph [13]. Knockingdown Tret1 leads to the reduction of hemolymph trehalose and inhibition of Plasmodium falciparum infection [13].
Glucose is the primary source of energy for both mosquitoes and Plasmodium [14,15]. During the blood stage and liver stage of malaria infection, Plasmodium parasites increase the absorption of glucose in host cells by enhancing the translocation of GLUT1 to the cell membrane [16,17]. Then these parasites scavenger host glucose by their facilitative hexose transporter (PfHT) [18]. However, the interactions of glucose metabolism between Anopheles mosquitoes and Plasmodium are still unclear. Only one glucose transporter, AGAP007752, is reported to be involved in facilitating Plasmodium sporozoites infection in Anopheles gambiae, and its knockdown decreased the number of sporozoites in mosquito salivary glands [19][20][21].
In this study, we identified four Asteglut genes in An. stephensi. RNAi-mediated silencing of Asteglut1 specifically increased P. berghei infection and significantly elevated the glucose level in mosquito midgut prior to blood-feeding. The accumulation of midgut glucose might modulate multiple biological processes, including detoxification and immunity, which in turn increased parasite infection.

Mosquito rearing and maintenance
Anopheles stephensi (strain Hor) was reared at 28 °C and a relative humidity of 80%. Adults were maintained on 2% sucrose solution. Adult female mosquitoes were fed on BALB/c mice for a blood meal.

Plasmodium berghei infection
Plasmodium berghei (ANKA strain) parasites expressing GFP constitutively were maintained by passing through BALB/c mice by mosquito biting [22,23]. When parasitemia of P. berghei infected mice rose to 3-6%, mosquitoes which had been starved overnight were allowed to feed on the infected mice for 15 min. Engorged mosquitoes were maintained at 20 °C and un-engorged mosquitoes were removed 24 h post-blood meal. Midguts were dissected and oocyst numbers were counted under a fluorescence microscope at 8 days post-infection. For the melanizaiton assay, midguts were dissected at 8 days post-infection and fixed in 4% formaldehyde for 30 min. Fluorescent oocysts and melanized ookinetes were visualized under a Nikon fluorescence microscope. Pictures were taken using a Nikon confocal microscope (Nikon, Tokyo, Japan).

RNA interference
The Asteglut genes were amplified by the correspond- Primers with the T7 promoter sequence (5'-TAA TAC GAC TCA CTA TAG GG-3') were used to generate templates for double-stranded RNA (dsRNA). The dsRNAs were synthesized using the MEGAscript RNA kit (Ambio, Invitrogen, Shanghai, China). The plasmid eGFP (BD Biosciences, Shanghai, China) was used as a control. Four-day-old mosquitoes were injected with 69 nl dsRNA (4 μg/μl) using a nanoject II microinjector (Drummond, Philadelphia, USA). The dsRNA-treated mosquitoes were collected two days post-treatment and knockdown efficiency was verified by qPCR as previously described [25].

RNA isolation, cDNA synthesis and quantitative PCR (qPCR)
Total RNA was extracted from mosquitoes using TRIzol reagent (Sigma-Aldrich, Shanghai, China) according to the manufacturer's protocol. One µg of total RNA was used to synthesize cDNA using 5× All-in-One Master-Mix (AccuRT Genomic DNA Removal Kit; ABM, Shanghai, China). The qPCR was performed using a Roche LightCycler 96 Real Time PCR Detection System using SYBR Green qPCR Master Mix (Biomake, Shanghai, China) according to a previously described protocol [25]. The data were processed and analyzed using the Roche LightCycler 96 software. Ribosomal gene s7 was used as the internal reference gene.

Sugar measurement
The glucose and trehalose levels in the mosquito hemolymph and midgut were examined as described [13,26]. Briefly, 30 μl hemolymph was collected from 10 mosquitoes. Ten midguts were pooled together and homogenized in 30 µl PBS buffer. Thirty microliters of midgut homogenates and hemolymph, respectively, were used for glucose and trehalose measurement; 10 µl was used to measure the glucose level using a Glucose Kit (K-GLUC; Megazyme Bray, Ireland); another 10 µl was treated with trehalase enzyme (K-TREH; Megazyme, Bray, Ireland), and then examined for glucose concentration. Trehalose concentration was calculated as described [13]. The remaining 10 µl was used for genomic DNA extraction and quantification [27]. The concentration of glucose and trehalose were normalized to the amount of genomic DNA, respectively.

RNA sequencing
Mosquitoes treated with dsAsteglut1 and dsGFP 24 h post-infectious blood meal were collected for RNA sequencing. Four mosquitoes were pooled for one sample and three biological replicates were used from each treatment. Total RNA was extracted using TRIzol ® reagent according the manufacturer's instructions (Sigma-Aldrich) and sent to Majorbio (Shanghai, China) for library construction and sequencing using Illumina HiSeq ×10. Clean data were aligned to the reference genome AsteS1.6 (https ://www.vecto rbase .org/organ isms/anoph eles-steph ensi). To identify DEGs (differential expression genes) between two groups, the expression level of each transcript was calculated according to the fragments per kilobase of exon per million mapped reads (FPKM) [28]. R statistical package edgeR (Empirical analysis of Digital Gene Expression in R; http://www.bioco nduct or.org/ packa ges/2.12/bioc/html/edgeR .html) was used for differential expression analysis [29]. GO functional enrichment were carried out by Goatools (https ://githu b.com/ tangh aibao /Goato ols). Both the heat map and Venn map were generated using TBtools software (v0.66831; https ://githu b.com/CJ-Chen/TBtoo ls-Manua l) [30]. The pie chart, scatter chart and histogram were created using GraphPad Prism version 6 (GraphPad Software, La Jolla, CA, USA). To demonstrate the similarity across individual biological replicates, principal components analysis (PCA) was performed using the Spotfire DecisionSite for Functional Genomics (DSFG) package (http://spotfi re. tibco .com/).

Statistical analysis
All statistical analyses were performed by GraphPad Prism version 8. Gene expression and sugar levels were compared using the Student's t-test. Oocyst data were not normally distributed as determined by the Shapiro-Wilk test. Thus, the Mann-Whitney test was used to determine the significance of oocyst intensity in dsRNAtreated mosquitoes.

Expression of Asteglut genes in An. stephensi
To determine the expression pattern of Asteglut genes in An. stephensi. We analyzed the expression levels of these genes by qPCR in the head, salivary glands, midgut, ovary and carcass 24 h before a blood meal, respectively. Asteglut1, Asteglut3 and Asteglut4 were mainly localized in the midgut tissue of An. stephensi (Fig. 2a, c, d).
In addition to the midgut, Asteglut1 and Asteglut4 were also expressed in the head and salivary glands (Fig. 2a, d).
Asteglutx was distributed in all five tissues (Fig. 2b). We next investigated the influence of parasite infection on the expression in the midgut of the four Asteglut genes. Asteglut genes were differentially regulated by P. berghei 24 h post-infection. (Fig. 2e). Plasmodium berghei infection significantly decreased the expression of Asteglut1 and Asteglut4 (t (14) = 2.585, P = 0.0216; t (14) = 3.001, P = 0.0095), while increased the expression of Asteglutx compared to those in normal blood feeding mosquitoes. No influence on Asteglut3 expression was observed during parasite infection (Fig. 2e, t (14) = 0.343, P = 0.7369).

Knockdown of Asteglut1 significantly elevates the glucose level in the mosquito midgut
We next analyzed the influence of Asteglut1 on sugar transportation in An. stephensi. The glucose and trehalose levels in the midgut and hemolymph of dsRNAtreated mosquitoes were examined. The glucose level of the Asteglut1-knockdown group was significantly higher than that in dsGFP controls 24 h prior to blood-feeding (Fig. 4a, t (8) = 4.374, P = 0.0024). However, its level in hemolymph is comparable to that in the dsGFP control (Fig. 4c). There was no significant difference between sugar levels in the midgut or hemolymph either just before (0 h) or 24 h post-blood-feeding (Fig. 4, for statistics details, see Additional file 4: Text S1). In addition, knockdown of Asteglut1 did not change the level of trehalose in the midgut or in hemolymph (Fig. 4b, t (8

Transcriptional analysis of Asteglut1-knockdown mosquitoes
To explore how Asteglut1 regulated P. berghei infection, we performed a transcriptome analysis of the mosquito's midgut treated with dsAsteglut1 and dsGFP 24 h post-blood meal, respectively. A total of 6 G PE clean sequences were generated by the Illumina HiSeq ×10 (Additional file 1: Table S1). Principal components analysis (PCA) showed a clear separation between dsAsteglut1 and dsGFP treatments (Additional file 2: Figure S1). The Venn diagram shows that the expression of 10,240 genes was overlapped in the two groups (Fig. 5a). A total of 46  Table S2) with 26 upregulated and 20 downregulated genes. These differentially expressed genes belong to multiple functional clusters, including cytoskeletal and structural, immunity, metabolism, proteolysis, redox, transport and those of unknown function (Fig. 5c).
Among the 'redox' functional cluster, five genes encoding cytochrome P450 (CYP450) were upregulated, indicating that detoxification was activated in mosquitoes [35]. The gene encoding peroxiredoxin that controls cytokine-induced peroxide levels in mammalian cells was also significantly upregulated, but the role of this gene in parasite control in mosquitoes is still unknown [36]. We also observed that DUOX (dual oxidase), which is involved in Plasmodium elimination, was significantly downregulated in dsAsteglut1-treated mosquitoes (P < 0.0001) [37]. It is highly possible that the reduction of DUOX expression might render mosquitoes more permissive to P. berghei infection.
The CLIP (class of serine proteases) family are involved in the melanization of P. berghei in An. gambiae [38].
Two CLIP genes, clip2 and clip19, were significantly downregulated in dsAsteglut1 treated mosquitoes, while clipb3 was upregulated compared to dsGFP mosquitoes (P = 0.0087, P = 0.0034) [38,39]. Next, we examined whether the increasing parasite infection could be due to the dysregulation in mosquito melanization. Midguts of mosquitoes treated with dsRNA 8 days post-infection were collected and melanization was visualized microscopically. We found that the number of melanized ookinetes increased with the number of oocysts (Fig. 5d, t (68) = 0.707, P = 0.482). Thus, there was no significant difference in the melanization rate between dsAsteglut1treated and the dsGFP control group.

Discussion
The glucose transporter family, functionally conserved from insects to mammals, is responsible for the transportation of glucose across the cell membrane [44]. In mammals, GLUT1 is one of the earliest cloned membrane transporters and has been extensively investigated in the past half century [45]. It is ubiquitously expressed in the skeletal, muscle, heart, and other tissues, but predominantly functions in erythrocytes and the blood-brain barrier [32,45,46]. In An. stephensi, Plasmodium undergoes a drastic reduction during the early stage of their infection in mosquitoes. In the entire Plasmodium life-cycle (in both human and mosquito hosts), parasite number is the lowest at the oocyst stage and then quickly increases with thousands of sporozoites produced per oocyst [47]. For this reason, we focused on the interactions between the midgut stages of parasites and mosquitoes, aiming to find a possible target for vector control. We identified 4 glut genes, Asteglut1, Asteglut3, Asteglutx and Aste-glut4. They have distinct expression patterns, suggesting their potential different roles in glucose transportation. Knockdown of Asteglut1 increased the glucose level in the midgut, suggesting its role in maintaining the homeostasis of intestinal glucose. However, we did not observe significant changes of glucose and trehalose levels in the hemolymph. It is highly possible that the sugar level in hemolymph is controlled by multiple factors and functional redundancy exists between members of the Asteglut family.
In addition, we also found that knockdown of Aste-glut1 influences the mosquito's susceptibility to P. berghei infection. In agreement with our findings, GLUT1 is involved in the regulation of pathogen infection in mammals. GLUT1 is a natural receptor of T-lymphotropic virus (HTLV) that facilitates the invasion of HTLV in human cells [48]. GLUT1 is also involved in the regulation of CD4 + T cell function in humans. Knocking out GLUT1 in CD4 + T cells reduces glucose uptake and glycolysis, and also impairs the growth, proliferation, survival and differentiation of these cells [49]. In plants, the expression of sugar transporter (SWEET) is induced by bacterial and fungal infection. Knockout of SWEET limits the growth of these pathogens [50]. In An. stephensi, invasion of P. berghei into salivary glands induces the expression of the glucose transporter AGAP007752. Its knockdown decreased the number of sporozoites in mosquito salivary glands [19][20][21].
Asteglut1 help to defense against P. berghei might occur through regulating midgut glucose level. The accumulation of glucose in the midgut when Asteglut1 is knockeddown might change multiple biological processes, which effect synergistically to increase parasite infection. Our Fig. 4 Influence of Asteglut1 on glucose transportation in the mosquito midgut. The relative concentration of glucose (a, c) and trehalose (b, d) in the midgut and hemolymph of GFP and Asteglut1 dsRNA-injected An. stephensi 24 h prior to (− 24 h), just before (0 h) and 24 h post-blood-feeding (24 h), respectively. Sugar concentration was normalized to genomic DNA extracted from midgut or hemolymph cells. Error bars indicate mean ± SEM (n = 5). Significance was determined by Student's t-test (for details, see Additional file 4: Text S1). Abbreviation: NS, non-significant. **P < 0.01 Error bars indicate mean ± SEM (n = 35). Significance was determined by Student's t-test (for details, see Additional file 4: Text S1). Abbreviation: NS, non-significant. Scale-bars: d, 50 μm transcriptome analysis reveals that a considerable number of upregulated genes are cytochrome P450s, which are responsible for detoxification [51]. The upregulation of cytochrome P450 genes in dsAsteglut1-treated mosquitoes indicate that these mosquitoes might suffer more toxicity than that in control [51,52]. In addition to catabolizing xenobiotics, the cytochrome P450s are also involved in the anabolism and catabolism of hormones [53]. For example, cytochrome P450s are involved in the biosynthesis of 20-hydroxyecdysone (20E) from cholesterol [54]. The steroid hormone 20E not only promotes oogenesis in mosquitoes, but also facilitates Plasmodium infection [55]. Thus, the elevated levels of P450 gene expression might be responsible for increased parasite infection.
The CLIP family members function as either activators or suppressors of melanization that are responsible for elimination of P. berghei in An. gambiae [38]. Although three clips genes are differentially regulated in dsAsteglut1-treated mosquitoes, we did not observe any difference in melanization rates between dsGFP and dsAsteglut1 treated mosquitoes. This result suggests that these CLIPs might function differently from classical CLIPS. Further investigation of their function is required.
We also noticed the significant induction of caudal, and reduction of pgrps, pgrp-la, pgrp-lc, pgrp-ld, and the antimicrobial peptides, defensin in dsAsteglut1-treated mosquitoes, suggesting that Asteglut1 might be involved in regulation of immune responses [25,40,56]. PGRP-LA is a receptor of the mosquito immune deficiency pathway (Imd) [41]. It helps to control the homeostasis of gut microbiota and parasite infection in An. stephensi [57]. PGRP-LC is the primary receptor of the Imd pathway. Silencing PGRP-LC blocks the synthesis of downstream immune effectors, which in turn increases parasite infection [42]. Different from PGRP-LA and LC, PGRP-LD is a negative regulator of the immune signaling pathway. However, knockdown of PGRP-LD similarly increased susceptibility of An. stephensi to P. berghei infection through impairing the integrity of the peritrophic matrix. The compromised peritrophic matrix structure results from the reduction of gut microbiota in the absence of protection by PGRP-LD [25]. Thus, the downregulation of pgrp-la, pgrp-lc and pgrp-ld all lead to increasing susceptibility of the Anopheles mosquito to Plasmodium infection [25,42,57]. However, how Asteglut1 regulates the Imd pathway needs to be investigated in the future.

Conclusions
In summary, we identified 4 GLUT members in An. stephensi and found Asteglut1 participates in the defense against P. berghei infection. The regulation of vector competence by Asteglut1 might occur through modulating multiple biological processes, especially detoxification and immunity (Fig. 6). Our findings pave the way for further understanding as to how sugar transporters regulate vector-parasite interactions and will help to explore potential new targets for vector control.