Costs and benefits of mosquito refractoriness to malaria parasites: implications for genetic variability of mosquitoes and genetic control of malaria.
In this study, we examined the fitness costs associated with resistance to the avian malaria parasite, Plasmodium gallinaceum, in its natural vector, Aedes aegypti mosquitoes. Following ingestion in a blood meal from an infected vertebrate host, the malaria parasite must complete a series of complicated biological events within the mosquito. Successful development includes exflagellation, fertilization, ookinete formation, and penetration through the peritrophic matrix and midgut epithelium to form oocysts. Oocysts then mature and develop into infective sporozoites, which then migrate to the salivary glands. Failure to complete any of these events inhibits parasite development and prevents transmission to the vertebrate host. Several parasite-inhibiting mechanisms have been recognized in mosquitoes. One mechanism involves the failure of the parasite to exflagellate or to undergo syngamy in the midgut of mosquitoes, probably because mosquitoes lack an exflagellation stimulator (Nijhout 1979). A second is the requirement for activation of parasite chitinase for penetration of the peritrophic matrix (Shahabuddin et al. 1993). A third is the melanotic encapsulation of the developing oocyst, which usually occurs extracellularly between the basolateral epithelial cell membrane and the basement membrane (Collins et al. 1986; Paskewitz et al. 1988). A fourth is salivary-gland barriers that prevent sporozoite penetration or reduce survivorship of sporozoites (Rosenberg 1985). A fifth is an apparent physiological incompatibility between the mosquito and the parasite, commonly referred to as a refractory condition (Kilama and Craig 1969; Thathy et al. 1994; Severson et al. 1995). Although intracellular ookinete killing by midgut epithelial cells has been observed in the Anopheles gambiae/P. gallinaceum system (Vernick et al. 1995), the biochemical and physiological basis for mosquito refractoriness remains unknown.
Mosquito refractoriness to Plasmodium exhibits complex modes of inheritance (Kilama and Craig 1969; Thathy et al. 1994; Severson et al. 1995; Vernick et al. 1995). Quantitative trait locus (QTL) mapping, using restriction fragment length polymorphism (RFLP) markers, demonstrated that the susceptibility of A. aegypti to P. gallinaceum is determined by at least two unlinked loci, and that each locus exhibits a partial dominance effect on refractoriness (Severson et al. 1995). The contribution of each locus was found not to be equal. A locus identified on chromosome 2 has a major effect, while a locus on chromosome 3 a small effect. Ward (1963) observed a significant response to selection for reduced susceptibility, suggesting the heritability of this trait is fairly high. If the reproductive success of mosquitoes is reduced by parasite infection, even at a level that is undetectable in the laboratory, the selection pressure from the parasite could increase the frequency of refractory alleles within a population if there is no genetic trade-off between refractoriness and susceptibility. Therefore, one would expect a high frequency or fixation of refractory alleles in natural mosquito populations.
However, significant genetic variation in susceptibility to P. gallinaceum in natural and laboratory A. aegypti populations has been documented (Ward 1963; Kilama 1973; Thathy et al. 1994). In particular, these studies identified high levels of susceptibility among Asian and New World populations, and relative high levels of refractoriness in African populations. Rausher and Simms (1989) describe several mechanisms concerning plant/herbivore interactions that may apply to this phenomenon. First, genetic variation in host susceptibility may not be due to selection by the parasite, rather it may simply be a result of founder effects. Although several studies have demonstrated reduced fecundity in mosquitoes exposed to vertebrate hosts infected with malaria parasites (Hacker and Kilama 1974; Freier and Friedman 1976), it is unclear to what degree this effect can be attributed to the malaria parasite itself. That is, blood from an infected animal may be of reduced nutritive quality to the mosquito (Bennett 1970). Infected hosts also may be less attractive, so that fewer mosquitoes feed on infected hosts, and those that do imbibe a smaller blood meal (Freier and Friedman 1976). In addition, the prevalence of malaria parasites in natural mosquito populations, even in areas with high vertebrate disease transmission, is low (Paul et al. 1995). Genetic variability for refractoriness may be selectively neutral and not directly associated with fitness, but simply the result of genetic drift or other ecological constraints.
A second hypothesis is that genetic variation in susceptibility is subject to stabilizing selection, resulting in intermediate levels of refractoriness. That is, if infection by a parasite reduces host fitness, but refractoriness is costly for the host to produce and maintain, stabilizing selection for an intermediate level of refractoriness could occur. Fitness costs of refractoriness may include self-inflicted toxicity, disruption of previous functions for biological pathways that are involved in refractoriness, devotion of resources to resistance and away from other fitness-enhancing functions, or other pleiotropic effects of genes involved in refractoriness (Simms and Triplett 1994). For example, the melanotic encapsulation of parasites by mosquitoes requires tyrosine as a precursor, which also is important for other biological events, such as egg-chorion and cuticular tanning and wound healing (Li et al. 1994). Indeed, reduced ovary size and protein content, and delayed oviposition have been observed in mosquitoes actively encapsulating parasites (Ferdig et al. 1993). When refractoriness is associated with reduced fitness, a negative additive genetic correlation exists between refractoriness and mosquito fitness measured in a parasite-free environment. Therefore, this hypothesis predicts that the frequency of refractory alleles should be higher in endemic regions that in nonendemic areas, because selection pressure from the parasite would be stronger in endemic regions. In areas without the parasite, refractory alleles should be rare.
A third hypothesis is that the observed genetic variation in susceptibility is a transient state prior to complete refractoriness or susceptibility. If there are no negative fitness effects associated with refractoriness, parasite-imposed selection could ultimately result in fixation for refractory alleles. The time to fixation may be very long, because alleles for the refractory condition usually have dominant or partially dominant effects (Severson et al. 1995). On the other hand, when the parasite exerts no selection and the fitness costs of refractoriness are high, a polymorphic population could rapidly become completely susceptible.
The aims of the present study were: (1) to determine whether exposure to the malaria parasite, P. gallinaceum, affected fitness in the mosquito, A. aegypti; (2) to determine whether there were fitness costs associated with refractoriness; and (3) to determine whether refractory mosquitoes obtained a fitness advantage in the presence of the parasite. Fitness components examined included body size, blood-meal acquisition and size, female fecundity, larval-to-adult developmental time, time to initial oviposition, and adult survivorship. These traits may be interrelated, and may have important consequences for mosquito population dynamics and disease transmission. Fitness components affected by parasitism have the potential to evolve under parasite-imposed selection. The results are discussed in relation to genetic diversity of natural mosquito populations and to the genetic control of malaria.
MATERIALS AND METHODS
Mosquitoes and Parasites. - Two populations of A. aegypti were used in this study. The MOYO-R and MOYO-S populations were selected for refractoriness to P. gallinaceum, and susceptibility from the Moyo-In-Dry strain (MOYO) by four generations of recurrent sib mating with a founder density of a pair of mosquitoes each (Thathy et al. 1994). Thus, the two populations have an inbreeding coefficient (F) of 0.5 (Crow and Kimura 1970). The MOYO strain was originally collected from Shauri Moyo Village, Mombasa, Kenya, in 1974.
Unless otherwise stated, mosquitoes were reared as previously described (Christensen and Sutherland 1984). Because the traits of interest could be sensitive to environmental parameters, the two populations were reared under identical conditions. The differences observed between them are therefore likely due to their genetic divergence or the interaction between genotype and environment, but not due to the environment itself. In particular, we controlled water, food availability, and larval density during development to adults. Cotton pads soaked in 0.3 M sucrose solution were constantly available to adults and were changed every two days. All mosquitoes were maintained in an environmental chamber regulated at 26.5 [+ or -] 0.5 [degrees] C and 80 [+ or -] 5% RH under a 16-h light: 8-h dark light cycle. Four- to five-day-old female mosquitoes were deprived of sucrose for 24 h before receiving an infected or uninfected blood meal.
Maintenance of the P. gallinaceum parasite was as previously described (Thathy et al. 1994). Female mosquitoes (50 mosquitoes/carton of 0.473 L), were allowed to engorge on restrained white leghorn chicks (Northern Hatchery Inc., Beaver Dam, WI). Parasite-infected chicks were obtained by sporozoite transmission. The parasitemia and percentage of gametocytes were monitored daily beginning one-week post-infection (PI). When the parasitemia (percentage of infected red blood cells) of an infected chick reached at least 10% with greater than 1% gametocyte development, the naive mosquitoes were allowed to bloodfeed on the infective chick for 15 min. Following exposure to chicks, fully engorged mosquitoes were separated and continuously supplied with sucrose pads. Mosquito susceptibility was examined by dissecting individual mosquitoes six to seven days post blood-feeding, and scoring the number of oocysts that developed on the midgut.
Mosquito Body Size. - Wing length is a reliable measurement of adult mosquito size (e.g., Ward 1963; Kelly and Edman 1992). Adult size may influence survivorship, developmental time, and ability to acquire a blood meal. Wing length was measured from the axial incision to the apical margin, not including the fringe of scales (Kelly and Edman 1992). Two independent trials were conducted. For trial 1, we measured 50 male and 50 female mosquitoes of each population. Only females (n = 50 per population) were measured in trial 2.
Blood-Meal Size. - Blood-meal size influences not only mosquito fecundity (Woke et al. 1956), but also the potential for acquiring pathogens (Kershaw et al. 1955). Blood-meal size was estimated indirectly by measuring hemoglobin concentration ingested by a mosquito, using the HiCN method (Briegel et al. 1979) with slight modifications. The bloodfed females were dissected and the midgut containing blood from a single mosquito was placed into a test tube with 0.5 mL of Drabkin's solution (see Briegel et al. 1979). The mixture was ground thoroughly and incubated at room temperature for at least 30 min, then 400 [[micro]liter] of chloroform was added, and the mixture was vortexed and centrifuged at 1500 rpm for 5 min. The supernatant was transferred to a test tube with 800 [[micro]liter] distilled water, and vortexed briefly. The absorbance was read at 415 nm using a Hitachi model U-2000 spectrophotometer.
A calibration curve was prepared for each chick by using known amounts of chick blood (0.8, 1.6, 2.4, 3.2 and 4.0 [[micro]liter]) and measuring the corresponding absorbance. The absorbance of midguts taken from unfed females (n = 10) was also measured, and the average of their reading was subtracted from that of the bloodfed females. Two independent trials were conducted. In trial I, 50 females of each population were exposed to the same uninfected or infected 17-day-old chick for 15 min, Twenty-five mosquitoes from each of the four treatments (2 populations x 2 infections = 4) were randomly selected to determine blood-meal size and wing length. In trial 2, 50 females from each population were exposed to an uninfected seven-day-old chick for 15 min. This experiment tests for a difference in blood-meal size between the MOYO-R and MOYO-S populations, effects of chick infection status and chick age on blood-meal size, and a correlation between mosquito body size and blood-meal volume.
Larval-to-Adult Developmental Time. - Sixty newly hatched larvae of each population were placed individually into glass vials (27.5 [cm.sup.3]) containing 2 mL of double-distilled water. Small amounts of finely ground fish food (Tetramin) were added to each vial every two days. Vials were checked at three-hour intervals beginning on day 6, and the time period during which an adult mosquito emerged and its gender were recorded. Larvae that failed to develop into adults (n = 13 for MOYO-R and n = 15 for MOYO-S) were excluded from the analysis.
Mosquito Adult Survivorship. - The survivorship of males and females of the two populations was examined separately. Male and female pupae were collected from stock population, and placed into 0.473-L paper cartons (50 individuals each). After emergence, male and female mosquitoes of each population were placed in 10 cartons (50 individuals per cartons) with marquisette coverings. Sucrose pads were changed every two days. On day 5 postemergence, female mosquitoes were starved overnight and allowed to feed on uninfected or infected chicks, and then placed into cartons in the same manner as the males. Mosquitoes were examined daily, and dead individuals were counted and removed until all individuals had died. We started with 500 mosquitoes for each treatment and each population except the MOYO-R infected treatment (n = 400). At day 6-7 postinfection, 85-100 mosquitoes of each population were dissected to monitor P. gallinaceum infection.
Ovipositional Studies. - Fifty females of each population were allowed to mate with males (50/carton) for four days prior to bloodfeeding. Fully engorged females then were placed individually in glass vials, each containing a strip of filter paper (10 x 50 [mm.sup.2]) as an oviposition substrate and about 1.5 mL of double-distilled water, and covered with marquisette. Sucrose was provided each day until day 5 post-bloodfeeding. Each female then was transferred to a fresh vial, and monitored for another five days. Female mosquitoes usually start to lay eggs two to four days after a blood meal. Excess water was discarded, and eggs were counted. The filter strips containing eggs were dried for one week, the eggs then hatched, and number of larvae was scored. Because both populations were exposed to the same infected or uninfected chick, any fecundity difference between the two populations within a treatment should not be influenced by differences in the blood-meal source. Parasite susceptibility was monitored by dissecting 16 females of each population at day 6 postinfection. An additional experiment was conducted to determine the time to initial oviposition. The basic protocol and design were as described above. Individual females were checked every three hours, beginning 36 hours after the blood [TABULAR DATA FOR TABLE 1 OMITTED] meal, and the time period required for the first eggs to be deposited was recorded.
Data Analysis. - We tested each dataset for normality using the Shapiro-Wilk test (SAS Institute1994) and for homogeneity using the F-max test (Sokal and Rohlf 1981). The data were transformed when appropriate (Table 1). Analysis of variance (ANOVA) then was used to test for the effect of population on body size and larvae-to-adult developmental [TABULAR DATA FOR TABLE 2 OMITTED] time, and the effect of population and exposure to infected hosts on blood-meal size, fecundity, and time to initial ovi-position. When appropriate, orthogonal contrasts were used to compare the populations and/or infection treatments with controls. Survivorship data were analyzed using the Kalpan-Meier method (SAS Institute 1994), and the log-rank test results are presented.
The mean infection intensity (the number of malaria oocysts in an infected mosquito) and prevalence (the proportion of mosquitoes infected in a population) of the mosquitoes used in the ovipositional and survivorship experiments are presented in Table 2. The two independent trials yielded similar infection levels. The MOYO-S population was significantly more susceptible than MOYO-R (Table 1). The majority of MOYO-R mosquitoes harbored no parasites, and none of these mosquitoes had more than 11 oocysts. Overall, 84.3% of the MOYO-S females were susceptible, and 73.4% of these infected individuals had more than 11 oocysts (oocyst distribution not shown). Our infection data for the two populations were similar to those reported by Thathy et al. (1994).
Mosquito Body Size
Since mosquitoes exhibit sexual dimorphism in body size, and the data combining both sexes did not meet the assumptions of normality and homogeneity required for analysis of variance (ANOVA), we analyzed the male and female body size data separately. MOYO-R females were significantly smaller than MOYO-S females, as revealed by wing length ([ILLUSTRATION FOR FIGURE 1 OMITTED]; trial 1: 3.11 vs. 3.22 mm, ANOVA orthogonal contrast, t = 6.41, P [less than] 0.0001; trial 2:3.06 vs. 3.24 mm, t = 9.87, P [less than] 0.0001). Similar results were found in males (2.54 mm for MOYO-R vs. 2.57 mm for MOYO-S, P = 0.04). There was no significant difference in female wing length between the two trials (Table 1).
Linear regression was used to establish correlations between absorbance at 415 nm and actual blood volume for the three chicks used in this experiment [ILLUSTRATION FOR FIGURE 2 OMITTED]. The regression coefficient for the 17-day-old uninfected chick was not significantly different from the seven-day-old uninfected chicks (ANOVA, F = 2.36; df = 1, 14; P [greater than] 0.05), however, it was significantly different from the 17-day-old infected chick (F = 17.57; df = 1, 14; P [less than] 0.001). This is due to a reduced hemoglobin content in blood from infected chicks ([ILLUSTRATION FOR FIGURE 2 OMITTED]; Bennett 1970).
There was no significant difference in blood-meal size within treatments between the two independent trials (Table 3; ANOVA, F = 3.87; df = 1, 96; P [greater than] 0.05), therefore the data for the two trials were combined for subsequent analysis. The proportion of MOYO-R females that took a blood-meal was smaller, and the average blood-meal size for MOYO-R was significantly less than for MOYO-S when they were exposed to either infected or uninfected chicks (Tables 1, 3, ANOVA orthogonal contrasts, t = 4.09, df = 1, P [less than] 0.0001 for uninfected chicks; t = 4.56, df = 1, P [less than] 0.0001 for infected chicks). Since the chick was restrained while feeding, defensive behaviors by the animal likely were not responsible [TABULAR DATA FOR TABLE 3 OMITTED] for the difference in blood-meal size between the two populations. When exposed to the infected chick, MOYO-R females ingested less blood as compared to those exposed to uninfected chicks (Table 1; t = 2.44, df = 1, P [less than] 0.05), but no significant difference was found in MOYO-S (t = 0.45, df = 1, P [greater than] 0.05).
We further examined whether mosquito body size was, in part, responsible for the observed variations among individuals within a population or between populations. Simple correlation coefficients between wing length and blood-meal size were calculated for the two populations exposed to either infected or uninfected chicks. When all individuals were included in the analysis, significant correlations were found in MOYO-S population [ILLUSTRATION FOR FIGURE 3A OMITTED], but not in MOYO-R [ILLUSTRATION FOR FIGURE 3B OMITTED]. When individuals not bloodfed were excluded, only one correlation was significant, that is, MOYO-S exposed to uninfected chicks ([r.sup.2] = 0.38, df = 20, P [less than] 0.01).
Larva-to-Adult Developmental Time
The MOYO-R population had a significantly shorter larva-to-adult developmental time than did MOYO-S for both males (Tables 1, 4; ANOVA orthogonal contrast, t = 3.66, df = 1, P [less than] 0.001) and females (t = 5.43, df = 1, P [less than] 0.0001). The faster development of MOYO-R may be related to its smaller body size.
TABLE 4. The larva-to-adult developmental time for two Aedes aegypti populations. Mean larval-to-adult developmental time Population Sex N [+ or -] SD (h) Range (h) MOYO-R Female 25 186.98 [+ or -] 13.08 172-222 Male 22 178.75 [+ or -] 11.68 167-214 MOYO-S Female 23 206.24 [+ or -] 10.43 186.5-222 Male 22 191.68 [+ or -] 12.98 172-207.5
Adult Mosquito Survivorship
Survivorship curves of MOYO-R and MOYO-S are illustrated in Figure 4. MOYO-R male mosquitoes had a significantly lower survivorship than did MOYO-S males ([ILLUSTRATION FOR FIGURE 4A OMITTED]; [[Chi].sup.2] = 64.88, df = 1, P [less than] 0.0001). The 50% survival point was day 20-21 postemergence for MOYO-R males (mean longevity = 20.06 d) and day 23-24 postemergence for MOYO-S males (mean longevity = 24.27 d). Compared to MOYO-S females, the survivorship of MOYO-R females was reduced dramatically [ILLUSTRATION FOR FIGURE 4B OMITTED]. The 50% survivorship occurred at day 26-27 postexposure for MOYO-R females and at day 47-48 for MOYO-S females when uninfected, or at day 26 for MOYO-R females and at day 49-50 for MOYO-S females when infected. The mean longevity of MOYO-R females was nearly half that of MOYO-S females, whether they were uninfected (25.11 vs. 45.67 d; [[Chi].sup.2] = 429.19, df = 1, P [less than] 0.0001) or infected (26.69 vs. 45.06 d; [[Chi].sup.2] = 325.57, df = 1, P [less than] 0.0001). Exposure to P. gallinaceum had no detectable effect on female survivorship for either population ([ILLUSTRATION FOR FIGURE 4B OMITTED]; [[Chi].sup.2] = 1.40, df = 1, P [greater than] 0.05 for MOYO-R; [[Chi].sup.2] = 0.01, df = 1, P [greater than] 0.05 for MOYO-S).
MOYO-R females produced a similar number of eggs as MOYO-S females when they were exposed to an infected chick (Table 5; ANOVA orthogonal contrast, t = 0.72, df = 1, P [greater than] 0.05). When they were exposed to an uninfected chick, more eggs were laid by MOYO-S than MOYO-R, and the difference was marginally significant (Tables 1, 5; t = 1.92, df = 1, P = 0.056). MOYO-R mosquitoes exposed to P. gallinaceum laid 14.3% fewer eggs than those exposed to an uninfected blood meal (Tables 1, 5; t = 3.16, P [less than] 0.01), while the reduction was 23.9% for MOYO-S (t = 6.07, df = 1, P [less than] 0.0001). Because MOYO-R population is refractory to P. gallinaceum, the reduced fecundity in this population, exposed to an infected chick, may reveal infected vertebrate host effects, for example, degraded blood quality. The parasite itself may contribute directly to the greater reduction in fecundity of P. gallinaceum-infected MOYO-S females.
Infection status of the blood meal did not affect egg hatch-ability in either population (Tables 1, 5; ANOVA orthogonal contrast, t = 1.55, df = 1, P [greater than] 0,05 for MOYO-R; t = 0.52, df = 1, P [greater than] 0.05 for MOYO-S). However, a significant difference between the two populations was observed. A greater proportion of MOYO-S eggs hatched as compared to MOYO-R eggs when the mosquitoes received either an uninfected (83% vs. 57%, t = 3.91, df = 1, P [less than] 0.001) or an infected blood meal (81% vs. 64%, t = 4.23, df = 1, P [less than] 0.0001).
The time to initial oviposition was significantly delayed [TABULAR DATA FOR TABLE 5 OMITTED] in MOYO-R females as compared to the MOYO-S females, when they each received an uninfected blood meal (Tables 1, 6; ANOVA orthogonal contrast, t = 2.04, df = 1, P [less than] 0.05) or an infected blood meal (t = 4.39, df = 1, P [less than] 0.0001). This trait was not affected by the infection status of the blood meal with MOYO-S mosquitoes (t = 0.26, df = l, P [greater than] 0.05). However, the delay in oviposition in MOYO-R females was significantly greater when they received an infected as compared to an uninfected blood meal (t = 2.34, df = 1, P [less than] 0.05).
TABLE 6. Mean time to the initiation of oviposition for two Aedes aegypti populations. No. of Mosquito mosqui- Mean time to infection toes oviposition Range Population status examined (h) [+ or -] SD(h) MOYO-R Control 35 55.80 [+ or -] 7.32 51-81 Infected 40 59.25 [+ or -] 9.67 51-81 MOYO-S Control 44 52.91 [+ or -] 2.83 51-60 Infected 46 53.28 [+ or -] 4.39 51-78
Our results demonstrate a reduction in fitness in an A. aegypti population (MOYO-R) that is refractory to P. gallinaceum under our experimental conditions, when compared to a susceptible population (MOYO-S) (Table 7). The decreased fitness of MOYO-R was manifested in several fitness components. MOYO-R has a significantly shorter egg-to-adult developmental time and a smaller adult body size, took a smaller blood meal, and subsequently laid fewer eggs than MOYO-S. The mean longevity of MOYO-R was significantly shorter than that of MOYO-S, regardless of the infection status of the blood meal. Exposure to the parasite had little effect on fecundity and no detectable effect on the survivorship of either population. Because the two mosquito populations were derived from the same ancestor population, and because all experiments were conducted in the same controlled environment, the observed fitness differences between the two mosquito populations most likely resulted from their genetic divergence.
If mosquito refractoriness to Plasmodium is associated with fitness costs, the frequency of alleles conferring refractoriness should decrease in polymorphic populations in the absence of selective pressure by the parasite. The rate of decrease may be a function of the initial frequency of refractory alleles, the fitness of homozygotes and heterozygotes, and the mating structure among the genotypes. The pattern of allele frequency change over generations may be complex because of the multigenic nature of refractoriness (Severson et al. 1995), and there is some experimental evidence to support this prediction (Kilama 1973). That is, significant decreases in refractoriness were observed over about eight generations in several artificial populations that were initiated with varying proportions of susceptible and refractory genotypes. Only some minor allelic frequency fluctuations were observed in populations fixed for the susceptible or refractory genotypes over the same time period (Kilama 1973).
TABLE 7. Summary of differences in susceptibility and fitness components between two Aedes aegypti populations refractory (MOYO-R) and susceptible (MOYO-S) to Plasmodium gallinaceum. Comparison between Traits analyzed MOYO-R and MOYO-S A. Susceptibility to P. gallinaceum MOYO-S [greater than] MOYO-R B. Adult body size females MOYO-S [greater than] MOYO-R males MOYO-S [greater than] MOYO-R C. Blood-meal size when exposed to an uninfected MOYO-S [greater than] MOYO-R vertebrate host when exposed to an infected ver- MOYO-S [greater than] MOYO-R tebrate host D. Larva-to-adult developmental time females MOYO-S [greater than] MOYO-R males MOYO-S [greater than] MOYO-R E. Fecundity uninfected mosquitoes MOYO-S [greater than] MOYO-R infected mosquitoes MOYO-S = MOYO-R F. Hatchability MOYO-S [greater than] MOYO-R G. Time to initial oviposition uninfected mosquitoes MOYO-S [less than] MOYO-R infected mosquitoes MOYO-S [less than] MOYO-R H. Mosquito survivorship males MOYO-S [greater than] MOYO-R uninfected females MOYO-S [greater than] MOYO-R infected females MOYO-S [greater than] MOYO-R
Our data suggest that P. gallinaceum likely imparts weak selection pressure on the mosquito host. The parasite exhibits no detectable effect on mosquito survivorship [ILLUSTRATION FOR FIGURE 4 OMITTED], and little effect on fecundity (Table 5). Therefore, these results support the suggestion that pathogens or parasites transmitted by biting arthropod vectors should exhibit low virulence in the motile vector host (Ewald 1983). Successful parasite transmission depends on vector susceptibility to parasites, vector longevity, vigorous biting behavior, and other factors. Severely affected vectors would reduce parasite transmission. Our data also support the hypothesis that genetic variation in mosquito susceptibility is subject to selection for increased frequency of susceptible genes (see hypothesis 3 in the introduction). Therefore, we would predict a high frequency or even fixation of susceptible genes in natural mosquito populations. This is true for Asian and New World A. aegypti populations, but not for some African populations (Ward 1963; Kilama 1973). What causes such a paradox?
Careful examination of our data may provide insight as to why refractory alleles are maintained in certain African mosquito populations. The significantly shorter developmental time found in MOYO-R (Table 4), probably correlated with a smaller body size [ILLUSTRATION FOR FIGURE 1 OMITTED], may provide the mosquitoes with a competitive advantage in environments where hosts, water availability, or other resources are limited. In such environments, early adult emergence may represent an important fitness-enhancing strategy. Although the refractory genotype exhibits a reduced fitness when exposed to our laboratory conditions, it may not be so in a harsh natural environment. For example, in some regions of Africa, particularly regions near the Sahara Desert where the rainy season is very short and water is scarce, mosquito eggs must hatch and develop into adults during a short window of opportunity. Such environmental conditions would favor any characters that reduce mosquito developmental time. If parasite refractory alleles originated in and were maintained by such ecological constraints, genetic correlations between refractoriness, body size, and developmental time are required. Although our data demonstrate phenotypic correlations between these traits, genetic correlations remain to be determined. Therefore, although we did not observe fitness trade-offs in mosquitoes between costs and benefits of refractoriness to the malaria parasite under our laboratory conditions, such fitness trade-offs may exist in some natural environments. Theories about host-parasite coevolution should consider the interactions between environments and host or parasite genotypes. In addition, the contributions of founder effects to the genetic diversity of natural mosquito populations need to be evaluated.
Our results may have implications for the genetic control of malaria. Malaria, one of the most important parasitic diseases today, infects 270 million people and results in 1-2 million deaths per year, primarily in children from tropical Africa (Knudsen and Slooff 1992). The emergence of pesticide resistance in mosquitoes and antimalaria drug resistance in Plasmodium has significantly limited malaria control efforts. Also, efforts toward vaccine development show little promise for disease control in the near future. Novel control strategies based on genetic disruption of mosquito vector competence have been proposed recently (James 1992; Collins and Besansky 1994). However, the success of genetic control strategies critically depends on the introduction and spread of parasite-inhibiting genes in natural vector populations. Understanding the fitness effects of these genes will provide important information toward developing such strategies. Two mechanisms have been proposed for driving parasite-inhibiting genes into mosquito populations (Curtis 1994). The first is for mosquitoes carrying refractory alleles to possess a competitive advantage, particularly in the presence of the parasite. The second is that refractory alleles may be spread by some driving force, such as mobile DNA sequences (e.g., transposons) or rickettsial symbionts (e.g., Wolbachia). If refractoriness is associated with high fitness costs and the parasite causes little damage, refractory alleles will not spread throughout the mosquito population without some driving mechanism. When such driving forces are used, the ultimate fate of refractory alleles that negatively influence host fitness will be determined by numerous factors. Investigation of these possibilities requires extensive theoretical (Ribeiro and Kidwell 1994) and experimental studies. The success of genetic control of malaria therefore requires an understanding of the genetic structure of natural mosquito populations.
Finally, the biochemical or genetic mechanisms underlying the observed fitness differences between MOYO-R and MOYO-S populations remain unknown. Three hypotheses, though not mutually exclusive, could explain the reduced fitness of MOYO-R. First, fitness variation between the two populations may reflect founder effect, inbreeding depression, or both. Since both populations were derived from the original MOYO population through selective inbreeding with a pair of mosquitoes as founders, it is possible that genes responsible for reduced fitness cosegregated with refractory genes. Although the effects of inbreeding are unknown for mosquitoes, genetic variation in inbreeding depression among lineages has been demonstrated in other insects (e.g., Pray and Goodnight 1995). RFLP data for several A. aegypti strains or crosses suggest that they have greater heterozygosity than expected (Severson et al. 1993; Severson and Kassner 1995), suggesting that mosquitoes may possess unknown mechanisms to avoid inbreeding. Second, the reduced fitness of MOYO-R may be caused by uncharacterized genes tightly linked to genes determining refractoriness. Third, genes conferring refractoriness may have direct pleiotropic effects on other factors determining fitness. Our recent population studies seem to refute the first hypothesis. By mixing MOYO-R with another highly susceptible strain (RED) and examining gene frequency changes over generations, we found that frequencies of the major QTL marker alleles specific to MOYO-R were decreased substantially after six generations (G. Yan, unpubl. data), but the frequencies of MOYO-R specific genes on chromosome 1 did not exhibit such changes. Similar results were observed for another independently selected refractory population (Kilama 1973): when the refractory population was mixed with a susceptible population in equal proportion, refractory gene (pls) frequency was reduced significantly in the first few generations. We have developed a linkage map for A. aegypti using RFLP arkers and have used these markers to identify two unlinked QTL affecting P. gallinaceum susceptibility (Severson et al. 1993, 1995). The genetic relationships between refractoriness to Plasmodium and traits associated with fitness can be identified using QTL mapping techniques (see Severson 1994). Such studies would add significantly to our understanding of the relationship between refractoriness and fitness.
We thank L. Christensen for technical assistance, and L. Stevens and D. Zaitlin for helpful comments on the manuscript. This work was supported by a National Research Service Award No. T32 (National Institutes of Health grant AI 07414) to GY, National Institutes of Health grant AI 33127 to DWS, and the Burroughs Wellcome Fund to BMC.
ANDERSON, R. M. 1988. The population biology and genetics of resistance to infection. Pp. 233-263 in D. Wakelin and J. M. Blackwell, eds. Genetics of resistance to bacterial and parasitic infection. Taylor and Francis, London.
BENNETT, G. F. 1970. The influence of the blood meal type on fecundity of Aedes (Stegomyia) aegypti L. (Diptera: Culicidae). Can. J. Zool. 48:539-543.
BREMERMANN, H. J. 1980. Sex and polymorphism and strategies in host-pathogen interactions. Journal of Theoretical Biology 87: 671-702.
BRIEGEL, H., A. O. LEA, AND M. J. KLOWDEN. 1979. Hemoglobinometry as a method for measuring blood meal sizes of mosquitoes. J. Med. Entomol. 15:235-238.
CHRISTENSEN, B. M., AND D. R. SUTHERLAND. 1984. Brugia pahangi: Exsheathment and midgut penetration in Aedes aegypti. Trans. Am. Microsc. Soc. 103:423-433.
COLLINS, F. H., AND N. J. BESANSKY. 1994. Vector biology and the control of malaria in Africa. Science 264:1874-1875.
COLLINS, F. H., R. K. SAKAI, K. D. VERNICK, S. M. PASKEWITZ, D. C. SEELEY, L. H. MILLER, W. E. COLLINS, C. C. CAMPBELL, AND R. W. GWADZ. 1986. Genetic selection of a Plasmodium-refractory strain of the malaria vector Anopheles gambiae. Science 234:607-610.
CROW, J. F., AND M. KIMURA. 1970. An introduction to population genetics theory. Harper and Row, New York.
CURTIS, C. F. 1994. The case for malaria control by genetic manipulation of its vector. Parasitol. Today 10:371-374.
EWALD, P. W. 1983. Host-parasite relations, vectors, and the evolution of disease severity. Annu. Rev. Ecol. Syst. 14:465-485.
-----. 1994. Evolution of infectious disease. Oxford Univ. Press, Oxford.
FERDIG, M. T., B. T. BEERNTSEN, F. J. SPRAY, J. LI., AND B. M. CHRISTENSEN. 1993. Reproductive costs associated with resistance in a mosquito-filarial worm system. Am. J. Trop. Med. Hyg. 49:756-762.
FREIER, J. E., AND S. FRIEDMAN. 1976. Effect of host infection with Plasmodium gallinaceum on the reproductive capacity of Aedes aegypti. J. Invertebr. Pathol. 28:161-166.
GROETERS, F. R., B. E. TABASHNIK, AND N. FINSON. 1994. Fitness costs of resistance to Bacillus thuringiensis in the diamondback moth (Plutella xylostella). Evolution 48:197-201.
HACKER, C. S., AND W. L. KILAMA. 1974. The relationship between Plasmodium gallinaceum density and the fecundity of Aedes aegypti. J. Invertebr. Pathol. 23:101-105.
HALDANE, J. B. S. 1949. Disease and evolution. La Ricerca Scient. Suppl. 19:68-76.
HAMILTON, W. D. 1980. Sex versus non-sex versus parasite. Oikos 35:282-290.
-----. 1982. Pathogens as causes of genetic diversity in their host populations. Pp. 269-296 in R. M. Anderson and R. M. May, eds. Population biology of infectious diseases. Springer, New York.
HAMILTON, W. D., AND M. ZUK. 1982. Heritable true fitness and bright birds: A role for parasites? Science 218:384-387.
JAENIKE, J. 1978. An hypothesis to account for the maintenance of sex within populations. Evol. Theory 3:191-194.
JAMES, A. A. 1992. Mosquito molecular genetics: The hands that feed bite back. Science 257:37-38.
KELLY, R., AND J. D. EDMAN. 1992. Mosquito size and multiple transmission of avian malaria in the laboratory. J. Am. Mosq. Control Assoc. 8:386-388.
KERSHAW, W. E., M. M. J. LAVOIPIERRE, AND W. N. BEESLEY. 1955. Studies on the intake of microfilariae by their insect vectors, their survival, and their effect on the survival of their vectors. VII. Further observations on the intake of the microfilariae of Dirofilaria immitis by Aedes aegypti in laboratory conditions: The pattern of the intake of a group of flies. Ann. Trop. Med. Parasitol. 49:203-211.
KILAMA, W. L. 1973. Distribution of a gene for susceptibility to Plasmodium gallinaceum in populations of Aedes aegypti (L.). J. Parasitol. 59:920-924.
KILAMA, W. L., AND G. B. CRAIG. 1969. Monofactorial inheritance of susceptibility to Plasmodium gallinaceum in Aedes aegypti. Ann. Trop. Med. Parasitol. 63:419-432.
KNUDSEN, A. B., AND R. SLOOFF. 1992. Vector-borne disease problems in rapid urbanization: New approaches to vector control. Bull. WHO 70:1-6.
LENSKI, R. E. 1988. Experimental studies of pleiotropy and epistasis in Escherichia coli. Variation in competitive fitness among mutants resistant to virus T4. Evolution 42:425-432.
LEONARD, K. J. 1977. Selection pressures and plant pathogens. Ann. N.Y. Acad. Sci. 287:207-222.
LEONARD, K. J., AND R. J. CZOCHER. 1980. Theory of genetic interactions among populations of plants and their pathogens. Annu. Rev. Phytopathol. 18:237-258.
LEVIN, B. R., AND R. E. LENSKI. 1983. Coevolution in bacteria and their viruses and plasmids. Pp. 99-127 in D. J. Futuyma and M. Slatkin, eds. Coevolution. Sinauer, Sunderland, MA.
LI, J., X. ZHAO, AND B. M. CHRISTENSEN. 1994. Dopachrome conversion activity in Aedes aegypti: Significance during melanotic encapsulation of parasites and cuticular tanning. Insect Biochem. Mol. Biol. 24:1043-1049.
MAY, R. M., AND R. M. ANDERSON. 1983. Epidemiology and genetics in the coevolution of parasites and hosts. Proc. R. Soc. Lond. B, Biol. Sci. 219:281-313.
MCKENZIE, J. A., AND A. Y. GAME. 1987. Diazinon resistance in Lucilia cuprina: Mapping of a fitness modifier. Heredity 59:371381.
MINKOFF, C., AND T. G. WILSON. 1992. The competitive ability and fitness components of the methoprene-tolerant (met) Drosophila mutant resistant to juvenile hormone analog insecticides. Genetics 131:91-97.
NIJHOUT, M. M. 1979. Plasmodium gallinaceum: Exflagellation stimulated by a mosquito factor. J. Parasitol. 48:75-80.
PASKEWITZ, S. M., M. R. BROWN, A. O. LEA, AND F. H. COLLINS. 1988. Ultrastructure of the encapsulation of Plasmodium cynomolgi (B Strain) on the midgut of a refractory strain of Anopheles gambiae. J. Parasitol. 74:432-439.
PAUL, R. E. L., M. I. PACKER, M. WALMSLEY, M. LAGOG, L. C. RANFORD-CARTWRIGHT, R. PARU, AND K. P. DAY. 1995. Mating patterns in malaria parasite populations of Papua New Guinea. Science 269:1709-1711.
PRAY, L. A., AND C. J. GOODNIGHT. 1995. Genetic variation in inbreeding depression in the red flour beetle Tribolium castaneum. Evolution 49:176-188.
RAUSHER, M. D., AND E. L. SIMMS. 1989. The evolution of resistance to herbivory in Ipomoea purpurea. I. Attempts to detect stabilizing selection. Evolution 43:563-572.
RIBEIRO, J. M. C., AND M. G. KIDWELL. 1994. Transposable elements as population drive mechanisms: Specification of critical parameters. J. Med. Entomol. 31:10-16.
ROSENBERG, R. 1985. Inability of Plasmodium knowlesi sporozoites to invade Anopheles freeborni salivary glands. Am. J. Trop. Med. Hyg. 34:687-691.
SAS INSTITUTE. 1994. JMP[R] user's guide. Cary, NC.
SEVERSON, D. W. 1994. Applications of molecular marker analysis to mosquito vector competence. Parasitol. Today 10:336-340.
SEVERSON, D. W., AND V. A. KASSNER. 1995. Analysis of mosquito genome structure using graphical genotyping. Insect Mol. Biol. 4:279-286.
SEVERSON, D. W., A. MORI, Y. ZHANG, AND B. M. CHRISTENSEN. 1993. Linkage map for Aedes aegypti using restriction fragment length polymorphisms. J. Hered. 84:241-247.
SEVERSON, D. W., V. THATHY, A. MORI, Y. ZHANG, AND B. M. CHRISTENSEN. 1995. Restriction fragment length polymorphism mapping of quantitative trait loci for malaria parasite susceptibility in the mosquito Aedes aegypti. Genetics 139:1711-1717.
SHAHABUDDIN, M., T. TOYOSHIMA, M. AIKAWA, AND D. C. KASLOW. 1993. Transmission-blocking activity of a chitinase inhibitor and activation of malarial parasite chitinase by mosquito protease. PNAS USA 90:4266-4270.
SIMMS, E. L., AND J. TRIPLETT. 1994. Costs and benefits of plants responses to disease: Resistance and tolerance. Evolution 48: 1973-1985.
SOKAL, R. R., AND F. J. ROHLF. 1981. Biometry. Freeman, San Francisco, CA.
THATHY, V., D. W. SEVERSON, AND B. M. CHRISTENSEN. 1994. Reinterpretation of the genetics of susceptibility of Aedes aegypti to Plasmodium gallinaceum. J. Parasitol. 80:705-712.
TOOBY, J. 1982. Pathogens, polymorphism, and the evolution of sex. J. Theor. Biol. 97:557-576.
TURNER, S. J. 1990. The identification and fitness of virulent potato cyst-nematode populations (Globodera pallida) selected on resistant Solanum vernei hybrids for up to eleven generations. Ann. Appl. Biol. 117:385-397.
VERNICK, K. D., H. FUJIOKA, D. C. SEELEY, B. TANDLER, M. AIKAWA, AND L. H. MILLER. 1995. Plasmodium gallinaceum: A refractory mechanism of ookinete killing in the mosquito, Anopheles gambiae. Exp. Parasitol. 80:583-595.
WAKELIN, D., AND J. M. BLACKWELL. 1988. Genetics of resistance to bacterial and parasitic infection. Taylor and Francis, London.
WARD, R. A. 1963. Genetic aspects of the susceptibility of mosquitoes to malaria infection. Exp. Parasitol. 13:328-341.
WOKE, P. A., M. S. ALLY, AND C. R. ROSENBERGER. 1956. The number of eggs developed related to the quantities of human blood ingested in Aedes aegypti (L.). Ann. Entomol. Soc. Am. 49:435-441.
|Printer friendly Cite/link Email Feedback|
|Author:||Yan, Guiyun; Severson, David W.; Christensen, Bruce M.|
|Date:||Apr 1, 1997|
|Previous Article:||Mitochondrial DNA phylogenies for the Drosophila obscura group.|
|Next Article:||Evolution of the genetic architecture underlying fitness in the pitcher-plant mosquito, Wyeomyia smithii.|