Immune response and insulin signalling alter mosquito feeding behaviour to enhance malaria transmission potential

Malaria parasites alter mosquito feeding behaviour in a way that enhances parasite transmission. This is widely considered a prime example of manipulation of host behaviour to increase onward transmission, but transient immune challenge in the absence of parasites can induce the same behavioural phenotype. Here, we show that alterations in feeding behaviour depend on the timing and dose of immune challenge relative to blood ingestion and that these changes are functionally linked to changes in insulin signalling in the mosquito gut. These results suggest that altered phenotypes derive from insulin signalling-dependent host resource allocation among immunity, blood feeding, and reproduction in a manner that is not specific to malaria parasite infection. We measured large increases in mosquito survival and subsequent transmission potential when feeding patterns are altered. Leveraging these changes in physiology, behaviour and life history could promote effective and sustainable control of female mosquitoes responsible for transmission.

The effect of anti-ILP morpholino treatment on engorgement success. Our previous experiments demonstrated that low ILP gene expression coincides with decreased hostseeking while high ILP expression coincides with increased host-seeking behaviours in both E.coli challenged and P. falciparum infected mosquitoes. Next, we sought to determine whether a functional connection was evident between ILP expression levels and host-seeking. To this end, we treated mosquitoes with anti-ILP3 or anti-ILP4 morpholinos, which reduced ILP protein levels in vitro by 59% and 89%, respectively. We then measured blood feeding rates in control and anti-ILP morpholino-treated groups. Both anti-ILP3 and anti-ILP4 morpholino-treated mosquitoes showed a statistically significant reduction in blood feeding propensity compared to controls (Chi-squared test, P < 0.05). When offered an artificial bloodmeal for 15 min, 67.1% of mosquitoes in the control group engorged. However, only 43.2% of mosquitoes in the anti-ILP3 morpholino-treated group engorged within the same timeframe, representing a greater than 20% reduction relative to controls. Similarly, 56.3% of mosquitoes in the anti-ILP4 morpholino-treated group took a bloodmeal, a reduction of more than 10% relative to controls.

Effects of E.coli challenge on ILP expression.
Our behavioural assays demonstrated that challenge with heat-killed E. coli can lead to dynamic changes in feeding behaviour in a dose-and time-dependent manner. Changes in insulin signalling have been linked to both the immune response and altered food seeking behaviour in Drosophila (14, 15), providing a potential mechanistic explanation for our previous results. To examine potential long-term consequences of immune challenge on insulin signalling that coincide with behavioural phenotypes, we measured the expression of ILP3 and ILP4 in the mosquito midgut and head on days 6 and 14 post-challenge with E. coli (Fig. 2cd). We found that time had a significant effect on the expression of both ILP3 (F= 7.56, d.f.=1 , P=0.02) and ILP4 (F=4.67 , d.f.=1 , P=0.05) in the midgut. Specifically, expression of these genes was low during the period of reduced host-seeking response (6 days) and elevated when host-seeking was enhanced (14 days). There was no effect of dose on ILP expression and no interaction between dose and time. In the head, we found that time also affected ILP3 expression (F=6.35 , d.f.=1 , P=0.03), but the pattern observed in the midgut was reversed (Fig. S1). That is, ILP3 expression was high at 6 days and low at 14 days. No significant changes in ILP4 expression were observed in the head at any of the time points examined (F=0.1 , d.f.=1 , P>0.05).
The effect of timing on host-seeking patterns. Neither immune challenge (heat-killed E. coli only, Wald Chi-Square= 3.48, P=0.06) nor a bloodmeal alone (bloodfed control, Wald Chi-Square=0.47, P=0.49) significantly altered host-seeking propensity compared to unmanipulated controls. As previously reported (7), bloodfed females challenged with heatkilled E. coli on day 0 exhibited a significantly different phenotype from bloodfed control females (Treatment x Test Period, Wald Chi-Square=10.91, P=0.01). Consistent with the 'manipulation' phenotype, females challenged directly after the bloodmeal were less likely to respond to the host on days 6-8 post bloodmeal compared to the response 13-15 days after the bloodmeal. However, this significant change in feeding propensity across the two sample periods was limited to the treatment in which the immune challenge was received on the same day as the bloodmeal (HK-0 between periods, Wald Chi-Square=13.66 , P<0.001). In the other groups, there was no significant difference between feeding propensity in the two time periods tested, nor was there a significant difference between these treatments and the bloodfed controls. There also was no significant difference between the feeding patterns observed for the bloodfed control and the group challenged on day 2 post bloodmeal (Treatment x Stage, Wald Chi-Sqaure=0.17, P=0.68) or the group challenged on day 4 post bloodmeal (Treatment x Stage, Wald Chi-Squared=0.03, P=0.87). See Figure 3a.
Effect of dose on host-seeking patterns. There was no difference between the injury and unmanipulated controls (Dose x Stage, Injury Control/Control Model, Wald Chi-Squared= 0.054, P=0.973) and thus, we combined these treatments for the remainder of analyses. There was a significant effect of replicate on overall response (Wald Chi-Squared=29.82, df=2, P<0.001), but no significant interactions between replicate and other parameters. Controlling for the replicate effect, there was a significant interaction between test period and dose (Wald Chi-squared= 28.14, df=11, P<0.001). See Figure 3b.
In the first test period (6-8 days post-challenge), all challenged mosquitoes were significantly less responsive to host cues than control females (Bonferroni pairwise comparison, P<0.05). The higher the dose of heat-killed E. coli a treatment group received, the lower its response (Wald Chi-Squared= 36.44, df=4, P<0.001). Mosquitoes receiving a low dose of heat killed E. coli were significantly less likely to respond than mosquitoes receiving no immune challenge (Bonferroni pairwise comparison, P=0.015) and significantly more likely to respond than those receiving the high dose of E. coli (Bonferroni pairwise comparison, P=0.032).
When mosquitoes from the same treatment groups were assayed on days 14-16 post bloodmeal, we observed the opposite trend. The response of mosquitoes to host odor increased with the dose of heat-killed E. coli. There was a significant difference between the high and medium dose treatment groups compared to the control (Bonferroni pairwise comparison, P=0.018). There was also a significant difference between the low and high treatment groups, but no significant difference between the medium and the other two treatment groups.
The effect of dose on expression of DEF1.We found significant effects of dose, sampling time point, and the interaction between dose and sampling time point (Table S1, Fig S2) on the expression of DEF1. Overall, immune challenge relative to no challenge significantly increased the expression of DEF1 (Adjusted Bonferroni: unmanipulated vs. all other immune challenge groups; p < 0.0001). Further, there was no significant difference between DEF1 expression resulting from an injury and the DEF1 expression elicited by intermediate doses of E. coli. The expression of DEF1 peaked at 12 h post injection. There was a significant interaction between dose of immune challenge and sampling time point because the effect of dose on magnitude of the immune response was strongest 12 h post-immune challenge, or the peak of DEF1 expression (Adjusted Bonferonni: 6 hr vs. 12 hr, p = 0.001; 12 hr vs. 24 hr, p = 0.018; 12 hr vs. 48 hr, p < 0.0001). Figure S1: Analysis of DEF1 expression in An. stephensi challenged with low, medium, and high doses of heat-killed E. coli. Immune challenge significantly increased expression of DEF1. The effect of dose on expression was most pronounced at peak expression (12 hr).  Figure S3: Expression of ILP3 and ILP4 in the head tissues of mosquitoes challenged with heat-killed E. coli (three doses).Expression of ILP3 was increased on day 6 in females challenged with heat-killed E.coli, while no significant patterns in ILP4 expression were observed. Figure S4. Anti-AsILP morpholinos knockdown AsILP levels in vitro and in vivo. Full-length (A) AsILP3 and (B) AsILP4 were detected at the highest levels in RIN-5F cells transfected with plasmids for overexpression and treated with control morpholino (Lane 2). Peptides were detected at decreased levels in cells transfected with plasmids for overexpression and treated with anti-AsILP morpholinos (Lane 3), but were not detected (ND) in cells transfected with empty plasmid (Lane 1). Values were normalized to Coomassie brilliant blue stain for total protein with proportional levels indicated below the blots.(C) In vivo, AsILP3 and AsILP4 were detected at the highest levels in protein extracts from mosquitoes fed control morpholinos and were detected at decreased levels in mosquitoes fed anti-AsILP morpholinos. Densitometry values from western blots were normalized to Coomassie brilliant blue stain for total protein and data are represented as mean fold reduction relative to control (broken line, set at 1). Table S1. Results from mixed effects model analysis demonstrate significant effects of treatment (dose of heat-killed E. coli) and sampling time. Replicate was included in the model as a random factor. Variation between samples (mRNA extraction and cDNA conversion) was controlled for using centered rpS7 cDNA counts.