Behavioral Ecology Vol. 13 No. 6: 816-820
© 2002 International Society for Behavioral Ecology
Stage-specific manipulation of a mosquito's host-seeking behavior by the malaria parasite Plasmodium gallinaceum
a Laboratoire de Parasitologie Evolutive, CNRS UMR 7103, Université P. & M. Curie, Paris, France b Unité de Biochimie et Biologie Moléculaire des Insectes, Institut Pasteur, Paris, France
Address correspondence to J.C. Koella, Laboratoire de Parasitologie Evolutive, CC237, Université P. & M. Curie, 7 quai Saint Bernard, 75252 Paris Cedex 05, France. E-mail: jkoella{at}snv.jussieu.fr.
Received 25 July 2001; revised 28 March 2002; accepted 28 March 2002.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
We present experimental evidence that different stages of the malaria parasite Plasmodium gallinaceum differentially affect the host-seeking behavior of its mosquito vector Aedes aegypti. In uninfected mosquitoes, host-seeking behavior is continued if mosquitoes have ingested less than a threshold volume of blood, whereas a larger blood meal inhibits host seeking. We investigated the parasite's effect on this behavior by feeding infected and uninfected mosquitoes for variable amounts of time and assaying 30-45 min later whether they continued their attempts at blood-feeding. Mosquitoes infected with oocysts (which cannot be transmitted) had a smaller threshold volume and were less likely to return for further probing, whereas individuals infected with transmissible sporozoites increased the threshold volume required to inhibit host-seeking behavior. We conclude that the stage-specific effect of the parasite on host-seeking behavior is likely to be an active manipulation by the parasite to increase its transmission success.
Key words: Aedes aegypti, behavioral manipulation, host-seeking behavior, malaria, Plasmodium gallinaceum.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
As a vector's biting rate largely determines the dynamics and epidemiology of parasites that rely on hematophagous insects for transmission (Macdonald, 1957
), the vector
is expected to be the target for behavioral manipulation by the parasite.
Indeed, several parasites appear to increase the biting rate of their insect
vectors, generally by impairing the vector's ability to obtain a full blood
meal and thereby inducing the vector to bite several times before it is fully
engorged. Examples include the protozoan Leishmania in sandflies
(Beach et al., 1985;
Jefferies et al., 1986), the
plague bacterium Yersinia pestis transmitted by the tropical rat flea
Xenopsylla cheopis (Bacot and
Martin, 1914), and trypanosomes in tse-tse flies
(Jenni et al., 1980;
Roberts, 1981).
The most detailed studies of altered feeding behavior mediated by parasites
have been done with malaria parasites and their mosquito vectors. The mosquito
Anopheles gambiae infected with sporozoites of the parasite
Plasmodium falciparum bites more people during a single night than
uninfected mosquitoes (Koella et al.,
1998). The mechanisms underlying this observation are not clear
but could be the result of at least two synergistic processes. First, the
parasite could increase the mosquito's motivation to resume a meal after it
has been interrupted, thus increasing the probability that it bites several
times. Second, the parasite could decrease the amount of blood obtained at
each biting attempt, thereby increasing the number of bites required to obtain
a given amount of blood. Some evidence for the latter mechanism has been
provided by the observation that sporozoites of the parasite Plasmodium
gallinaceum lower the apyrase activity in the salivary glands of infected
Aedes aegypti. As a result, an infected mosquito's ability to locate
blood is impaired, and it probes for a longer time than do uninfected
mosquitoes (Rossignol et al.,
1984). Furthermore, infection by sporozoites reduces the fecundity
of mosquitoes, which may be a consequence of obtaining less blood than
uninfected ones in a restricted amount of time
(Rossignol et al., 1986).
Direct evidence, however, for either mechanism is lacking.
Not all stages of the malaria parasite, however, would benefit from an
increase in the mosquito's biting rate. After infection, malaria oocysts must
develop for several days on the mosquito's midgut wall before they can produce
the sporozoites, the only stage that can be transmitted to the vertebrate
host. During this developmental period, the only way to increase overall
transmission is to increase the mosquito's survival. An important component of
the probability that a mosquito lives long enough to transmit malaria may be
the mortality of the mosquito associated with blood-feeding
(Anderson and Brust, 1996;
Edman et al., 1984). Therefore,
oocysts should be expected to decrease the biting rate of mosquitoes, in
contrast to the sporozoites. In the only experiment investigating differential
manipulation by different stages of a parasite, the feeding behavior of
Anopheles stephensi infected by Plasmodium yoelii
nigeriensis followed this prediction: When unfed mosquitoes were allowed
to probe but not to imbibe any blood, sporozoites increased the persistence to
probe, but oocysts decreased persistence
(Anderson et al., 1999).
The goal of this study was to compare the effects of malaria oocysts and
sporozoites on the two mechanisms that could affect the mosquito's biting
rate: motivation and efficiency. To achieve this goal, we used the yellow
fever mosquito A. aegypti infected by P. gallinaceum. Our
predictions are twofold. First, as sporozoites decrease the anticoagulant
activity of apyrase (Rossignol et al.,
1984), mosquitoes infected with sporozoites should imbibe less
blood in a fixed amount of time than uninfected mosquitoes. As there is no
published report on the effect of oocysts on apyrase activity, we have no
prediction about the effect of oocyst infection on blood-meal size. Second,
the host-seeking behavior of this species is regulated by a neurohormonal
system that allows the mosquito to recognize when it is sufficiently full
(Klowden et al., 1987). Thus,
host-seeking behavior is continued if mosquitoes have ingested less than a
threshold volume of blood, while a larger blood-meal size inhibits host
seeking (Klowden and Lea,
1978,
1979). If the parasite can
manipulate this behavior, the threshold volume of blood that inhibits further
feeding may be expected to be shifted to a lower level by oocysts but to a
higher level by sporozoites.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Species
The experiment was performed with strain 8A of the malaria parasite P. gallinaceum and the Liverpool Blackeye colony of its mosquito vector A. aegypti. To infect the mosquitoes, we used a domestic race of the parasite's natural host, 3- to 4-week-old white leghorn chickens (Gallus gallus domesticus). The chickens were infected by intravenous injection of 0.5 ml of infected blood (15% parasitemia). The experimental animals were maintained according to European Union guidelines.
General design
We fed mosquitoes on an infected or an uninfected chicken and assayed their
blood-feeding behavior (see below) either 6 or 12 days later (i.e., when the
parasites had developed to mature oocysts or sporozoites). To control for the
potentially confounding effect of age on the difference between infection by
oocysts and sporozoites, we infected mosquitoes 5, 10, or 15 days after their
emergence. Each treatment was done in three replicates. Mosquitoes within each
treatment group were maintained in population cages, given the opportunity to
lay eggs, and provided with sugar (cotton soaked with 6% sugar solution) until
1 day before the feeding behavior was assayed.
Feeding behavior
To assay the blood-feeding behavior, we isolated mosquitoes in individual
plastic cups (8 cm diam x 10 cm high) covered with mosquito netting and
allowed them to feed on one of our arms for different amounts of time ranging
from 0 to 3 min. Between 30 and 45 min later, we tested whether they were
still motivated to blood-feed by giving them access to one of our arms for at
most 5 min. For each mosquito we recorded the time when it inserted its
stylets into the arm in an active attempt at probing and then immediately
withdrew our arm to prevent further uptake of blood. Mosquitoes were then
frozen at -20°C for further processing.
Laboratory analyses
We measured the amount of blood imbibed by the mosquitoes by dissecting the
abdomen on a glass slide (which later allowed detection of oocysts), smearing
the blood onto filter paper, eluting the blood into 1 ml of Drabkin's solution
[1.0 g NaHCO3, 0.1 g K2CO3, 0.05 g KCN, and
0.2 g K3Fe(CN)6 in 1 l distilled water], and measuring
the optical density of a 200-µl aliquot in 96-well plates with an ELISA
reader (Briegel et al., 1979).
To standardize the optical density, we used a sample of 4 µl of blood
obtained with a fingerprick. Midguts were stained with mercurochrome and
checked for oocysts at 200x magnification. Sporozoites were detected
visually after dissection of the salivary glands. Both wings were fixed onto
slides and measured to the nearest 0.05 mm from the tip (excluding the fringe)
to the distal end of the allula; we used the mean length of the wings as a
measure of mosquito size. We performed the laboratory procedures without
knowing the treatment group of the mosquito being analyzed (except testing for
parasites, which we only did with mosquitoes that had fed on an infected
chicken).
Statistical analyses
We considered two aspects of blood-feeding behavior. First, we analyzed the
effect of infection on the amount of blood taken up in the limited amount of
time they had access to our arms. We did this with an ANCOVA, where the time
available to the mosquitoes for blood-feeding was the covariate, infection
status and days after infection were the main factors, and age at infection
and wing length were confounding factors. The amount of blood was
log-transformed to obtain a distribution that was close to Gaussian. Second,
we analyzed the effect of infection on the motivation to continue
blood-feeding after having taken up a limited amount of blood. We did this
with a logistic analysis of the proportion of mosquitoes that started probing
after having been blood-fed, where the amount of blood previously taken up,
infection status, and days after infection were the main factors, and where we
controlled for age at infection, wing length, and replicate. A more detailed
survival analysis used the same factors to analyze the time required for the
mosquitoes to start their probing behavior. We present the results of a
parametric survival algorithm using a Weibull distribution, but an exponential
and a log-normal distribution and a proportional hazard model gave similar
results. As the experiment was replicated three times, we blocked all of our
analyses by the three replicate cages. Note that in our tables of results,
interactions are only shown if they are statistically significant. All
analyses were done with the statistical package JMP
(http://www.jmpdiscovery.com).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
We analyzed 454 mosquitoes. Among the mosquitoes fed on an infected chicken, 93% were found to harbor parasites; only these were used in further analyses.
The amount of blood taken up was between 0 and 5.8 µl. The amount of blood increased with the time available for feeding and with wing length of the mosquitoes, but not with age at infection (Table 1). Uninfected mosquitoes took up more blood if they had been starved of blood for 12 days than if they had starved for only 6 days (Figure 1). Infected mosquitoes took up less blood than uninfected ones 12 days after infection (i.e., when the parasites had developed sporozoites), but there was no difference between infected and uninfected mosquitoes 6 days after infection (i.e., at the oocyst stage; Figure 1).
|
|
Overall, 72% of the mosquitoes resumed host-seeking behavior and attempted to probe 30-45 min after having been blood-fed for a limited time. Motivation increased with wing length and decreased with the amount of blood previously obtained but was not affected by the age at infection (Table 2). At 6 days after the initial engorgement, 78% of the uninfected and 67% of the oocyst-infected mosquitoes attempted to blood-feed. At 12 days after the initial engorgement (i.e., when the parasite had developed sporozoites), 63% of the uninfected and 82% of the infected mosquitoes probed. The increased motivation in sporozoite-infected mosquitoes and decreased motivation in oocyst-infected mosquitoes remained apparent when the amount of blood (Figure 2) and the other confounding factors (Table 2) were controlled for.
|
|
Similar results were obtained for the time required for the mosquito to start probing (Table 3). In particular, sporozoite-infected mosquitoes began probing earlier than uninfected ones at the same age (Figure 3). Although mosquitoes infected by oocysts were less likely to probe than uninfected ones (see above), if they did probe, they did so slightly earlier than uninfected ones. Wing length did not influence the time required to start probing, but older mosquitoes took longer to start probing than young ones.
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
At least in the laboratory situation described here, the malaria parasite P. gallinaceum affected the host-seeking behavior of its mosquito vector, A. aegypti, in two ways. First, sporozoites decreased the amount of blood imbibed in a fixed amount of time (Figure 1). This has been suspected as the mechanism leading to lower fecundity in infected mosquitoes (Rossignol et al., 1986
), and
is indeed expected as the malaria sporozoites modify apyrase, an enzyme in the
mosquito's saliva with anti-coagulatory properties, to make the uptake of
blood less efficient (Rossignol et al.,
1984). Therefore, to obtain the same amount of blood, infected
mosquitoes must therefore bite more often than uninfected ones.
The effect of decreasing the efficiency of blood-feeding was enhanced by a
second mechanism of manipulation: altering the host-seeking behavior if the
blood meal is interrupted before the mosquito has obtained a full blood meal.
Moreover, the two stages of the parasite within the mosquito, oocysts and
sporozoites, affected the blood-feeding behavior in opposite ways. Thus the
sporozoites, which can be transmitted to the next vertebrate host bitten by
the mosquito, increased the volume of blood required to inhibit the mosquito's
host-seeking behavior. In contrast, the oocysts, the developmental stage that
cannot be transmitted, decreased this blood volume. Although the physiological
mechanism for this manipulation is unknown, it is likely to be linked to the
neurohormonal regulation of the mosquito's host-seeking behavior (Klowden and
Lea, 1978,
1979).
The evolutionary reasons selecting for stage-specific manipulation are
clear. As the oocysts cannot be transmitted, their main interest is that the
mosquito survives their development period, as survival is the major factor
determining the parasite's transmission success
(Macdonald, 1957). One of the
main determinants of the mosquito's survival is the risk associated with
biting (Anderson and Brust,
1996; Edman et al.,
1984). Thus, the oocysts can increase the mosquito's survival and
thereby their own transmission by decreasing the mosquito's biting rate. One
of the mechanisms for doing so is to inhibit the host-seeking behavior of
mosquitoes that have so little blood in their midgut that they would normally
try to increase their blood meal with further biting attempts. In contrast,
the sporozoites require biting for their transmission. Their interest in
biting rate, however, does not coincide with the mosquito's. Although the
sporozoites are transmitted during probing
(Li et al., 1992;
Ponnudurai et al., 1991), the
mosquito's reproductive success depends not only on biting, but also on
surviving for several days to develop and lay its eggs. This asymmetry leads
the parasite to favor a higher biting rate than does the mosquito vector
(Koella, 1999), and thus leads
the parasite to manipulate the mosquito to bite more frequently. Thus, the
combined manipulation for decreased host-seeking behavior by oocysts but for
increased motivation by sporozoites increases transmission of the
parasite.
Combining these two mechanisms leads to at least two predictions for
mosquitoes in natural populations. First, as sporozoites increase the
threshold volume of blood that inhibits host-seeking behavior, mosquitoes
infected with sporozoites should be found to have more blood than uninfected
ones. Second, as sporozoites decrease the amount of blood taken up in a
limited time, infected mosquitoes should bite more often to obtain their blood
meal and should thus be more likely to bite several people than uninfected
ones. Both of these predictions have been found to hold true in a Tanzanian
population of A. gambiae infected with Plasmodium falciparum
(Koella et al., 1998). Thus,
while 72% of the uninfected mosquitoes had obtained a full blood meal, 82% of
the sporozoite-infected ones had engorged fully, and while 10% of the
uninfected blood-fed mosquitoes contained blood from at least two people, 22%
of the infected blood-fed mosquitoes did. The second prediction is also
supported by data from a field study in the Gambia. Here, 50% of pairs of
children sleeping in the same house harbored identical genotypes (assessed at
three loci) of P. falciparum parasites, whereas only 1% of pairs of
children picked at random were infected by the same malaria genotypes
(Conway and McBride, 1991). The
explanation was that individual infected mosquitoes had bitten (or at least
probed on) several people within a house and had transmitted the parasite
during several of these bite attempts. Unfortunately, field data on
oocyst-infected mosquitoes are not available to test analogous
predictions.
Patterns in our data concerning uninfected mosquitoes are interesting in
their own right. Mosquitoes that had been starved for 12 days imbibed more
blood than those starved for only 6 days, although both groups of mosquitoes
had fully digested their previous blood meal and laid eggs. As age per se has
no effect on blood meal size, one may speculate that the mosquitoes increase
the risk associated with feeding when hosts are only infrequently encountered.
Such a behavior extends the idea that mosquitoes balance the trade-off between
mortality and fecundity in an attempt to maximize fitness during each blood
meal (Anderson and Roitberg,
1999), to an approach considering bet-hedging and lifetime
reproductive success (Stearns and
Crandall, 1981). Furthermore, the fact that neither blood-meal
size nor host-seeking behavior increased with age contrasts with studies on
other insects, for which reproductive behavior becomes more risky as age
increases and as expected life span decreases
(Roitberg et al., 1993).
However, a more constant behavior is reasonable for mosquitoes because their
daily mortality is independent of age
(Gillies and Wilkes, 1965;
Lines et al., 1991); old
mosquitoes can expect a similar duration of remaining life as young ones.
The mechanisms suggested here to manipulate the behavior of the mosquito
are only two among many possible ones. Plasmodium yoelii nigeriensis,
for example, manipulates the feeding persistence of unfed A.
stephensi; while sporozoites increase the time period for which unfed
mosquitoes continue to probe despite being inhibited from blood-feeding,
oocysts decrease this time (Anderson et
al., 1999). Again, this manipulation is expected to increase
transmission for the reasons described above.
In conclusion, malaria parasites appear to have evolved several mechanisms
to manipulate the blood-feeding behavior of their mosquito vectors. Although
similar manipulation is observed in many other host—parasite systems
(Furlow, 1998), the complexity
of the physiological and evolutionary mechanisms often makes it difficult to
disentangle the effects due to adaptive evolution of the parasite, to an
adaptive response by the host, or to accidental byproducts of infection
(Poulin, 1994;
Poulin, 2000). In contrast,
the variability of specific behavioral patterns and, in particular, the
specific (and indeed opposing) patterns of manipulation by different stages
make it likely that the manipulation by malaria parasites has evolved to
increase the parasite's transmission.
|
ACKNOWLEDGEMENTS |
|---|
We thank A. Carmi-Leroy for her help in rearing the mosquitoes, Solveig Schørring for comments, and the members of the Laboratoire d'Ecologie for not complaining (too much) about the all too frequent mosquito bites.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Anderson RA, Brust RA, 1996. Blood feeding of Aedes aegypti and Culex nigripalpus (Diptera: Culicidae) in relation to defensive behaviour of Japanese quail (Coturnis japnocia) in the laboratory. J Vector Ecol 21: 94-104.[ISI]
Anderson RA, Koella JC, Hurd H, 1999. The effect of Plasmodium yoelii nigeriensis infection on the feeding persistence of Anopheles stephensi liston throughout the sporogonic cycle. Proc R Soc Lond B 266: 1729-1733.[Medline]
Anderson RA, Roitberg BD, 1999. Modeling trade-offs between mortality and fitness associated with persistent blood feeding by mosquitoes. Ecol Lett 2: 98-105.
Bacot AM, Martin CJ, 1914. Observations on the mechanism of the transmission of plague by fleas. J Hyg Plague 3(suppl.): 423-439.
Beach R, Kiilu G, Leeuwenburg J, 1985. Modification of
sand fly biting behavior by Leishmania leads to increased parasite
transmission. Am J Trop Med Hyg 34:
278-282.
Briegel H, Lea AO, Klowden MJ, 1979. Hemoglobinometry as a method for measuring blood meal sizes of mosquitoes (Diptera: Culicidae). J Med Entomol 15: 235-238.
Conway DJ, McBride JS, 1991. Genetic evidence for the importance of interrupted feeding by mosquitoes in the transmission of malaria. Trans R Soc Trop Med Hyg 85: 454-456.[ISI][Medline]
Edman JD, Day JF, Walker ED, 1984. Field confirmation of laboratory observations on the differential antimosquito behaviour of herons. Condor 86: 91-92.
Furlow B, 1998. The body snatchers. New Scientist 163: 42-44.
Gillies MT, Wilkes TJ, 1965. A study of the age composition of populations of Anopheles gambiae Giles and Anopheles funestus Giles in North-Eastern Tanzania. Bull Entomol Res 56: 237-262.[ISI][Medline]
Jefferies D, Livesey JL, Molineux DH, 1986. Fluid dynamics of blood-meal uptake by Leishmania-infected sandflies. Acta Trop 43: 43-53.[Medline]
Jenni L, Molyneux DH, Livesey JL, Galun R, 1980. Feeding behaviour of tsetse flies infected with salivarian trypanosomes. Nature 283: 383-385.[Medline]
Klowden MJ, Davis EE, Bowen MF, 1987. Role of the fat body in the regulation of host-seeking behavior in the mosquito Aedes aegypti. J Insect Physiol 33: 643-646.
Klowden MJ, Lea AO, 1978. Blood meal size as a factor
affecting continued host-seeking by Aedes aegypti (L.). Am J
Trop Med Hyg 27:
827-831.
Klowden MJ, Lea AO, 1979. Abdominal distention terminates subsequent host-seeking behavior of Aedes aegypti following a blood meal. J Insect Physiol 25: 583-585.[Medline]
Koella JC, Sørensen FL, Anderson R, 1998. The malaria parasite Plasmodium falciparum increases the frequency of multiple feeding of its mosquito vector Anopheles gambiae. Proc R Soc Lond B 265: 763-768.[Medline]
Koella JJ, 1999. An evolutionary view of the interactions between anopheline mosquitoes and malaria parasites. Microb Infect 1: 303-308.[Medline]
Li X, Sina B, Rossignol PA, 1992. Probing behavior and sporozoite delivery by Anopheles stephensi infected with Plasmodium berghei. Med Vet Entomol 6: 57-61.[Medline]
Lines JD, Wilkes TJ, Lyimo EO, 1991. Human malaria infectiousness measured by age-specific sporozoite rates in Anopheles gambiae in Tanzania. Parasitology 102: 167-177.
Macdonald G, 1957. The epidemiology and control of malaria. Oxford: Oxford University Press.
Ponnudurai T, Lensen AHW, van Gemert GFA, Bolmer JHE, Meuwissen JHET, 1991. Feeding behaviour and sporozoite ejection by infected Anopheles stephensi. Trans R Soc Trop Med Hyg 85: 175-180.[ISI][Medline]
Poulin R, 1994. Meta-analysis of parasite-induced behavioural changes. Anim Behav 48: 137-146.
Poulin R, 2000. Manipulation of host behaviour by parasites: a weakening paradigm. Proc R Soc Lond B 267: 787-792.[Medline]
Roberts LW, 1981. Probing by Glossina morsitans
morsitans and transmission of Trypanosoma (Nannomonas)
Congolense. Am J Trop Med Hyg 30:
948-951.
Roitberg BD, Sircom J, Roitberg CA, van Alphen JJM, Mangel M, 1993. Life expectancy and reproduction. Nature 364: 108.[Medline]
Rossignol PA, Ribeiro JMC, Spielman A, 1984. Increased
intradermal probing time in sporozoite-infected mosquitoes. Am J Trop
Med Hyg 33:
17-20.
Rossignol PA, Ribeiro JMC, Spielman A, 1986. Increased
biting-rate and reduced fertility in sporozoite-infected mosquitoes. Am
J Trop Med Hyg 35:
277-279.
Stearns SC, Crandall RE, 1981. Bet-hedging and persistence as adaptations of colonizers. In: Evolution today (Scudder GGE, Reveal JL, eds). Philadelphia: Hunt Institute; 371-383.
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||