Generation of AsOBP1 knockout mutants and AsOBP1 overexpression transgenic lines
To investigate a potential role for AsOBP1, we generated knockout (KO) mutants using CRISPR/Cas9. Single guide RNA (sgRNA) targeting of the first exon of the AsOBP1 gene was selected, which was confirmed with high on-target activity by CRISPR/Cas9-mediated KO technology (Fig. S2, Table S1). To facilitate the screening and establishment of homozygous lines, we then knocked in a 3×P3-EGFP visible marker construct at the AsOBP1 target site using CRISPR/Cas9-mediated knock-in technology (Fig. 1A). By injecting the mixture of sgRNA, homologous template and Cas9 protein into An. sinensis eggs (Fig. 1B), we obtained 46 fertile G0 females; one female produced fluorescent G1 larvae with a mutation efficiency of 2.17% (Fig. 1C, Table S1), which contained the specific insertion of the eye color marker (Fig. 1D). The AsOBP1 mRNA level in the AsOBP1−/− adults was undetectable in the homozygotes compared with the wild-type (WT) line (Fig. S3; one-way ANOVA, P < 0.001). We did not find any mutations at two potential off-target sites (Table S2) in the AsOBP1−/− line through polymerase chain reaction (PCR) amplification and sequencing.
We established the AsOBP1 overexpression (OE) transgenic lines by transposon-based transgenesis. In total, 1225 pre-blastoderm embryos were injected with the construct and the helper phspBac: 112 embryos hatched into larvae (9.12%), of which 80 (71.43%; 42 female and 38 male) reached adulthood (G0) (Table S1). The G0 females were crossed with WT males and laid eggs separately; three of 13 fertile G0 females produced fluorescent G1 larvae (Table S1). Each line had a single insertion integrated into a different genomic locus (Fig. S4, Table S3). The expression of AsOBP1in the transgenic females was significantly elevated compared with the WT mosquitoes, as determined by quantitative PCR (Fig. S4). These results indicated that the AsOBP1-T2A-EGFP fusion gene was successfully integrated into the genome of An. sinensis, and expression of AsOBP1 was artificially elevated. The OE2 line was preserved, and a homozygous transgenic line was established for follow-up studies.
Tissue-specific expression of AsOBP1 in mosquito antennae
Enhanced green fluorescent protein (EGFP) expression of the transgenic line was robust and allowed direct visualization of EGFP-labelled tissues in larvae and in adult mosquitoes without immunostaining (Fig. 2). The AsOBP1 promoter is in the 4000-bp region upstream of the transcription start site, but further analysis is required to refine the core promotor sequence. GFP fluorescence was specifically observed in the antennae of the larval and adult stages of An. sinensis (Fig. 2A, D, E), whereas no fluorescence was detected in any other tissues, including the proboscis and maxillary palps of both sexes (Fig. 2L–O), consistent with previous results (Qin et al. 2014; He et al. 2022). We did not detect GFP fluorescence in the antennae of embryonic and pupal stages (Fig. 2C). More importantly, expression patterns of AsOBP1 in the adult antennae were sexually dimorphic: AsOBP1 was expressed in the segments 4–13 of the female antennae (Fig. 2E), but only in the last segment of the male antennae (Fig. 2H). The sections of the female antennae showed that AsOBP1 was mainly concentrated in the cuticle (Fig. 2K). Among the four types of sensilla, including sensilla trichodea, sensilla chaetica, sensilla basiconica, and sensilla coeloconica (Zhang et al. 2019), AsOBP1 was only expressed in the sensilla trichodea of the male and female antennae (Fig. 2F, I).
AsOBP1 KO or OE did not affect the fitness of mosquitoes
Neither AsOBP1 KO nor OE homozygotes had obvious defects in external morphology or body color. No difference was observed among the WT, KO and OE lines in the developmental duration of the larval and pupal stages, size, and fecundity of adults (Fig. 3A–C). The deletion or OE of AsOBP1 did not affect the egg production of females or hatching of eggs (Fig. 3D–F). The survival rate of KO and OE homozygotes was not significantly different from that of WT mosquitoes under fasting conditions (log-rank test, P = 0.8746; Geahand–Breslow–Wilcoxon test, P = 1.0000; Fig. 3G). These results suggested that the mutation or overexpression of the AsOBP1 gene did not affect the fitness of mosquito, which indicates that AsOBP1 is not essential for normal development and production of An. sinensis.
AsOBP1 mutation affects blood feeding
To investigate a potential role for AsOBP1 in blood feeding, we compared the feeding success after starvation among WT, KO and OE mosquitoes. When exposed to an anesthetized mouse for 5 min, only 30% of KO homozygotes fed, significantly lower than that of WT mosquitoes (63.33%) and OE homozygotes (73.33%) (one-way ANOVA, P < 0.0001, Fig. 4A). When mice were exposed for 20 min, the percentage of the three genotypes fed on blood increased; the feeding rate of KO mutants was still the lowest, while that of OE homozygotes was the highest (one-way ANOVA, P < 0.0001, Fig. 4B). The difference in the percentage of blood-fed mosquitoes among the KO, OE and WT mosquitoes was not observed when they were exposed to mice for 30 min (one-way ANOVA, P = 0.4911, Fig. 4C). We then assessed the effects of AsOBP1 mutation or overexpression on the size of ingested blood meal. As expected, the amount of blood ingested by KO mutants was significantly lower than that of WT and OE mosquitoes (one-way ANOVA, P5min < 0.0001, P20min < 0.0001; Fig. 4D, E) after either 5 or 20 min exposure to mice. However, no significant difference in blood meal size was found between the three genotypes after feeding on mice for 30 min (one-way ANOVA, PKO vs WT = 0.1983; PKO vs OE = 0.0712; Fig. 4F). These results suggest that AsOBP1 is required for blood feeding in An. sinensis.
AsOBP1 is involved in host seeking
To obtain a blood meal for egg development, the mosquito must first find a suitable host, then alight on the host, and probe a suitable site for insertion of the mouthparts. Therefore, we investigated the effects of AsOBP1 mutation on the three successive behaviors. Firstly, we evaluated the host-seeking abilities of the WT, KO and OE mosquitoes using human host proximity assay and two-port olfactometer assay. In the first assay, 17.78 ± 2.22% KO mosquitoes were attracted to the arm within 1 min, which was significantly lower than that of WT mosquitoes (27.89 ± 4.20%; χ2 test, χ2 = 6.9979, P = 0.0082) and OE mosquitoes (36.22 ± 1.51%; χ2 test, χ2 = 22.2548, P < 0.0001) (Fig. 5A). In the second assay, we observed similar results. The number of attracted KO mutants was significantly less than that of WT mosquitoes (χ2 test, χ2 = 28.4342, P < 0.0001) and overexpression lines (χ2 test, χ2 = 75.0302, P < 0.0001) (Fig. 5B). Both behavioral tests clearly showed that mutation of AsOBP1 gene seriously impaired the host-seeking abilities of mosquitoes, while overexpression of AsOBP1 enhanced the sensitivity of mosquitoes to their hosts. Then, we observed the dynamic landing process of the WT, KO and OE mosquitoes on the human arm in 5 min by arm-in-cage assays. KO mosquitoes landed on human arms more slowly than WT and OE mosquitoes at any time point within 5 min (Fig. 5C). This phenomenon may be explained by the hindered ability to find their hosts in KO mosquitoes compared with WT and OE mosquitoes. Finally, we assessed the impact of AsOBP1 mutation or overexpression on probing behaviors. No difference among the WT, KO and OE lines in the times of the first probe (one-way ANOVA, F = 0.7618, P = 0.4703), cumulative probing times (one-way ANOVA, F = 0.8264, P = 0.4415) and interprobe times (one-way ANOVA, F = 0.0291, P = 0.9713) (Fig. 5, Table S5). The results suggest that AsOBP1 gene is not involved in the probing behaviors of An. sinensis mosquitoes. Thus, disruption of AsOBP1 gene impaired the host-seeking abilities and made the mosquitoes take a longer time to find their hosts, resulting in a significant reduction in feeding success.
AsOBP1 is involved in host seeking by binding specific odorants
To reveal the mechanism by which AsOBP1 affects the host-seeking behaviors of mosquitoes, it is necessary to identify human odorants to which AsOBP1 may bind. The fluorescence competitive binding assay showed that AsOBP1 specifically bound nine of 83 human odorants, including 1-tetradecanol, dodecanal, heptanal, indole, 3-methylvaleric acid, butanal, trans-2-octene, trans-2-octene-1-ol and trans-3-octene (Fig. 6B). Their binding constants were 6.42, 11.30, 16.78, 19.06, 23.96, 29.25, 21.32, 48.04 and 55.48 µmol/L, respectively (Table S5). Among them, 1-tetradecanol showed the strongest affinity to AsOBP1, followed by dodecanal, heptanal and indole, while the others had weak affinity.
To confirm that these odorants acted on mosquitoes, we measured the EAGs of the WT, KO and OE antenna responses to a panel of compounds, including the nine human odorants and 1-octen-3-ol. 1-Octen-3-ol did not bind to AsOBP1 but is known as a mosquito attractant. The EAG responses to 1-tetradecanol and heptanal were significantly weaker in KO females than in WT and OE females (Fig. 7A–C). The EAG response elicited by 1-tetradecanol decreased with the increase in dose; in contrast, the EAG response produced by heptanal increased with dose (Fig. 7B, C). Surprisingly, we found that the EAG responses to butanal and trans-3-octene were significantly higher in KO females than WT or OE females (Fig. 7A), in contrast to the lower response elicited by 1-tetradecanol and heptanal in KO females. EAG studies revealed no difference in the response of KO and OE females to 1-octen-3-ol, trans-2-octene, trans-2-octene-1-ol, and indole.
Subsequently, we assessed the effects of the nine compounds bound with AsOBP1 on the host-seeking activities of An. sinensis mosquitoes. The mice were perfumed with compounds on the abdomen (~ 1 µg). Two compounds with strong affinities to AsOBP1 (1-tetradecanol and heptanal), but not other compounds, attracted more mosquitoes than the solvent did (Fig. 7D). The two-port olfactometer assays showed 1 µg 1-tetradecanol and heptanal on the mouse abdomen had a significant effect on host-seeking behavior of mosquitoes (Fig. 7E, F). When applied to human hands, both compounds demonstrated an enhanced mosquito-attracting effect (Fig. 7E, F). These results indicate 1-tetradecanol and heptanal can mediate the host-seeking behaviors of An. sinensis mosquitoes.