The authors have declared that no competing interests exist.
Conceived and designed the experiments: LBC LL TWS. Performed the experiments: LBC MVA. Analyzed the data: LBC LL. Contributed reagents/materials/analysis tools: TWS. Wrote the paper: LBC LL TWS.
Current address: Nossal Institute for Global Health, The University of Melbourne, Parkville, Victoria, Australia
Environmental factors such as temperature can alter mosquito vector competence for arboviruses. Results from recent studies indicate that daily fluctuations around an intermediate mean temperature (26°C) reduce vector competence of
We tested the effect of temperature fluctuations on
Our results indicate that
Mosquitoes in the wild are exposed to daily fluctuations in temperature, but in the laboratory, the effect of temperature on vector competence is generally assessed using constant temperatures. Recent studies demonstrate that realistic fluctuations in temperature around an intermediate mean (26°C) can alter life-history traits, population dynamics, and the ability of a mosquito to become infected with and transmit dengue virus (DENV). Here we tested how fluctuations around high and low mean temperatures influence vector competence and the extrinsic incubation period. Small fluctuations around a high mean temperature (∼8°C swings around 30°C) had no detectable effect on vector competence. Large fluctuations around a low mean (∼18°C swings around 20°C) demonstrate that only 18.9 days were required for 50% of DENV-exposed mosquitoes to develop a disseminated infection, compared to 29.6 days at constant 20°C. Twenty-eight days post-exposure to the infectious blood meal, 100% of mosquitoes tested had a disseminated infection under fluctuating temperatures, but under a constant temperature this proportion was only 42%. Reduced duration of extrinsic incubation increases the potential for pathogen transmission. Results indicate that the rate of dengue transmission by mosquitoes in temperate regions with natural fluctuations may be underestimated by experiments conducted under constant temperatures.
The ability of
The norms of reaction (i.e., phenotypic variation across environmental variation) of vector competence and EIP have been well documented for a large range of temperatures for
What is much less well-documented is the influence of fluctuations in daily temperature on the norm of reaction of vector competence and EIP. Indeed, environmental temperature under natural conditions does not remain constant, but oscillates between a minimum at night and a maximum during daytime. Results from studies using realistic fluctuating temperature profiles support the notion that fluctuating temperatures may alter estimates of both life history traits and vector competence of mosquitoes
While the effects of realistic temperature fluctuations on
In this study we investigated whether fluctuations at high and low mean temperatures alter adult survival, vector competence and/or EIP of
We determined the effect of constant and fluctuating temperature regimes at both high and low mean temperatures, on the survival and vector competence of
The magnitude and asymmetrical shape of the temperature profiles were based on temperature recordings from Central West Thailand where DENV is endemic
When females were 4–5 days old, access to sucrose was removed for 24–36 hr, after which time females were fed defibrinated sheep blood (QuadFive, Ryegate, MT), mixed with DENV-1 freshly grown in cell culture prior to mosquito exposure, using an artificial feeding system. Virus supernatant was harvested after scraping and then separating all cells by centrifugation. Mosquitoes were fed through a desalted porcine intestinal membrane stretched over the bottom of a warm water-filled jar to maintain a temperature of 37°C. The viral isolate used, SV2951 obtained from Ratchaburi, Thailand, had been passaged at 28°C seven times in
We prepared one blood meal for each experiment. The blood meal for the low temperature experiment was calculated to contain 5.86×105 focus forming units (FFU)/ml of DENV-1. The calculated titer for the high temperature experiment was 7.89×105 FFU/ml. Mosquitoes in both experiments were limited to 35 min feeding, to minimize the effect of virus degradation in the infectious blood meal. Mosquitoes were allowed 2–3 hr to begin digestion after the blood meal. We subsequently sedated them using CO2 and retained only fully engorged females to set up experimental groups. For the low temperature experiment, forty-four replicate 1-pint paper cartons (Science Supplies WLE, NJ) with mesh tops, each containing 20 engorged females were set up. Twelve cartons were placed into each of the experimental temperature regimes, and eight cartons into the control 26°C incubator. For the high temperature experiment, we tested 28 replicate cartons each containing 16 females. Nine cartons were placed into the constant temperature incubators, and 10 into the 30°C plus fluctuation incubator.
We assessed vector competence at 7, 14, 21 and 28 days post exposure (DPE) to the infectious DENV-1 blood meal (i.e., days of EIP) for mosquitoes in the low temperature experiment. At each time point, we sampled three replicate cartons of mosquitoes from each experimental temperature, and two from the control 26°C treatment. At the high temperatures, mosquitoes were sampled at 3, 6 and 9 DPE. Three cartons were randomly removed from each incubator at each time point. The additional carton in the 30°C fluctuation treatment was also tested at 9 DPE. Because the course of DENV infection in the mosquito is faster at higher temperatures than at lower ones
For all surviving mosquitoes in each carton, we measured two components of vector competence, midgut infection and virus dissemination from the midgut in infected females, using a qualitative indirect fluorescence assay (Q-IFA). Virus EIP measurements were based on detection of a disseminated DENV infection in the mosquito.
We separated and tested bodies (comprising of the thorax and abdomen) for midgut infection and heads for disseminated infection, independently. Samples were placed into 1 mL viral transport medium (VTM; 77.2% low glucose DMEM, 18.5% heat-inactivated fetal bovine serum, 3.8% penicillin/streptomycin, and 0.15% gentamycin and nystatin) with approximately ten 2 mm glass beads (Fisher Scientific, Pittsburg, PA) in a screw-top plastic vial. Following collection, all samples were frozen at −80°C for later analysis by Q-IFA. We also collected the whole bodies (without separation of heads) of dead females daily and tested them for infection status. Results from analysis of dead mosquitoes were included in our survival analyses.
All data was analyzed using JMP software, version 10 (SAS Institute Inc., NC). Vector competence was analyzed by nominal logistic regression of the infection or dissemination status as a full-factorial function of temperature and DPE, and carton nested within temperature and DPE. Records of survival for individual females exposed to the infectious blood meals were kept throughout the duration of the both experiments. Female survival was analyzed using a Kaplan-Meier (log-rank) analysis, with females that were sacrificed on scoring days right-censored. We tested for differences in survival curves between different temperature regimes and infection status of recently dead mosquitoes. We corrected for multiple comparisons between treatment groups for our logistic regression and Kaplan-Meier analyses using a Bonferroni correction.
We used an infectious fluorescent focus assay
To test mosquito samples for the presence of infectious DENV, we used a qualitative fluorescence assay. We homogenized the tissue samples for 4 min in a Retsch Mixer Mill 400, at 30 Hz. We filtered 300 µL of the sample through 0.22 µm cellulose acetate centrifuge filters (Costar Spin-X, Corning, Japan) and 50 µL of the filtered supernatant was inoculated in duplicate onto a 1-day old confluent monolayer of Vero cells, seeded at a density of 2.5×105 cells/well in a 96-well culture plate. The inoculum was allowed to infect the cells for 1 hr at 37°C, before a standard maintenance media containing 2% FBS overlay was applied to the cells in each well, and the plate was incubated 37°C for 4 days. Positive and negative controls were used in each plate. We then removed the overlay, washed and fixed the cells in 3.7% formaldehyde for 20 min. The washing and staining steps that followed were exactly the same as for the FFA, except that the volumes used for antibody staining were 50 µL for each of the primary and secondary antibodies. We viewed cells under FITC-fitted fluorescence microscope at 10× to screen for the presence or absence of green fluorescence, which was indicative of a sample being either infected or uninfected by DENV, respectively.
A total of 769 female
Females were held at 16°C, 20°C and 26°C constant, and 20°C with a large DTR and sampled at days 7, 14, 21 and 28 post-exposure to an infectious DENV-1 blood meal. A) Body infection, representing a detectable infection of the midgut tissue. B) Levels of infection in head tissue, representing a detectable disseminated infection.
Dissemination of DENV-1 to head tissue was tested on a total of 145 mosquitoes, 129 of which were sacrificed at the days indicated above. At 16°C, there were no individuals with an infected body that had a detectable disseminated infection and this temperature treatment was, therefore, excluded from subsequent analysis. For the remaining three temperatures, we found a significant effect of temperature at 14, 21 and 28 DPE, but not at 7 DPE (
Time point | Factor | ?2 | df | |
Day 7 | Temperature | <0.0001 | 3 | 1.0000 |
Carton[Temperature] | 1.2749 | 5 | 0.9375 | |
Day 14 | Temperature | 18.4072 | 2 | 0.0001 |
Carton[Temperature] | 5.9605 | 5 | 0.3101 | |
Day 21 | Temperature | 16.8531 | 3 | 0.0008 |
Carton[Temperature] | 5.2339 | 4 | 0.2641 | |
Day 28 | Temperature | 14.2789 | 3 | 0.0025 |
Carton[Temperature] | 6.1889 | 3 | 0.1028 |
At each time point, the effect of temperature, and carton nested within temperature was analyzed for the 26°C and 20°C constant temperatures, and the 20°C cyclic treatment. 16°C constant was not included in tests because only single individuals with a detectable infection at 7 and 21 days post-exposure were identified, neither of which had a disseminated infection.
We tested for pair-wise differences between temperatures at each incubation period; differences are shown in
The time taken for 50% of infected individuals to complete the EIP (EIP50) was estimated using logistic regression of dissemination rates. All three estimates of EIP50 were significantly different from each other (χ2<40.04, df = 2, p<0.0001). The 20°C constant treatment had an estimated EIP50 of 29.6 days (95% confidence interval: 23.9–141.9), for 20°C with fluctuations it was 18.9 days (95% CI: 15.7–20.9), and for 26°C constant it was 11.1 days (95% CI: 5.9–15.1).
We assessed the infection status of a total of 458 female
Females were held at 30°C and 35°C constant, and 30°C with a small DTR and sampled at days 3, 6 and 9 post-exposure to an infectious DENV-1 blood meal. A) Body infection, representing a detectable infection of the midgut tissue. B) Levels of infection in head tissue, representing a detectable disseminated infection.
A total of 251 head tissue samples were tested for dissemination. Three infected mosquitoes from the 30°C treatment that were found dead at 1 DPE tested negative for dissemination. The remaining 248 head tissue samples were tested after their collection at 3, 6 or 9 DPE. Temperature did not influence the proportion of infected females with a disseminated infection (χ2<0.0001, df = 2, p = 0.999). As expected dissemination increased sharply from 3 to 6 DPE (χ2 = 140.16 df = 2, p<0.0001). The earliest dissemination was observed in a single female collected 3 DPE from the 30°C constant treatment. By 6 and 9 DPE, all mosquitoes had 75–100% dissemination to their head tissue (
There were very low mortality rates for female
A) Females held at low temperatures and a 26°C control. Despite the overall effect of temperature (p<0.0001), there were no differences in the survival curves after 28 days, between any of the three low temperature treatments (p>0.7). At 26°C constant, mortality was greater than in each of the low temperature treatments (p<0.001). B) Females held at high temperatures. Temperature influenced overall survival curves (p = 0.006), but only the curves of 30°C plus small DTR and 35°C constant were statistically different from each other (p = 0.001).
Survival of females was followed for nine days under three high temperature treatments (30°C, 35°C and 30°C with small DTR). Females from the 35°C treatment had the highest mortality rates. There was a significant overall temperature effect (χ2 = 10.24, df = 2, p = 0.006,
Compared to a constant temperature, large diurnal temperature fluctuations at a mean of 20°C reduced the EIP50 for
Females exposed to a large DTR around a 20°C mean were more likely to have detectable disseminated DENV-1 after 28 days compared to those reared under a constant, control temperature (100%
We observed a very low proportion of DENV-1 infected females held at 16°C constant. The youngest of the three infected females identified was found dead at 4 DPE, while the remaining two females were collected at 7 and 21 DPE during our weekly sampling. Due to slow digestion at such a low temperature, it is possible that the 4 and 7 DPE mosquitoes retained some infectious blood from the blood meal several days earlier. Although it is possible for a mosquito to become infected with DENV at 16°C, as shown by a single individual with a body infection at 21 DPE, this low temperature sharply reduced vector competence for DENV in
There was no detectable effect of the small fluctuations around a high mean of 30°C in the proportion of females with a midgut infection or disseminated virus, or in the duration of the EIP compared to the constant temperature control. The entire temperature profile (∼27°C to 35°C) falls within limits known to be highly conducive to DENV transmission, therefore, the lack of observable change is possibly due to the magnitude of the DTR not being large enough to produce a detectable response given our sample size. We did not test the large DTR around a mean of 30°C because there are few locations that we are aware of that have such large amplitude fluctuations at high temperatures. We therefore restricted our use of the large DTR to lower temperatures. Our cyclic low temperature treatment was derived from ambient conditions in dengue-endemic northern Thailand between December and January
Results from previous studies indicate that midgut infection levels were lower under fluctuating temperature regimes with a mean of 26°C compared to constant temperatures, leading to reduced transmission potential
Lambrechts et al.
Although we used only a single serotype (DENV-1) to test the hypothesis that fluctuations at high and low mean temperatures would alter mosquito vector competence, the EIP of the virus, and adult survival, cumulative results from our group
Our results indicate that the effect of fluctuations around a low mean temperature markedly reduce EIP, which has important implications for determining DENV transmission risk at the northern and southern edges of DENV's geographic range, areas with a mean temperature that would normally be considered too low for DENV transmission to occur. Additionally, seasonal variation in DENV transmission, which is a common feature of DENV transmission dynamics
Similar responses to temperature changes have been reported for life-history trait estimates of
We observed limited mortality throughout the duration of both experiments, and identified females with a disseminated infection in six of the seven temperature treatments tested (all but 16°C). Mosquitoes were raised under conditions with optimal nutrition and were maintained in an environment with limited risk of death other than intrinsic factors and temperature. We do not know the maximum potential lifespan of mosquitoes exposed to each of these temperature regimes. We planned the experimental duration to be long enough for mosquitoes of each temperature to discern the duration of the EIP under each treatment, but did not attempt to estimate longevity. Epidemiologically, although these estimates represent a conservative estimate of the number of mosquitoes that might survive to such a time in order to transmit DENV, the high survival estimates compared to the duration of the EIP indicate that a relatively large proportion of infected mosquitoes in both experiments were capable under laboratory conditions of surviving to an age where they could transmit DENV to a susceptible host.
Similar to previous studies
Our results indicate that the use of constant temperature experiments to assess
We would like to thank W. K. Reisen for continued support of our use of containment facilities, lab space and fruitful discussions, A. Ponlawat for collecting and sending mosquito eggs, and three anonymous reviewers for constructive comments on an earlier version of the manuscript. This research benefited from discussions with M. B. Thomas, K. P. Paaijmans, and working group members in the Research and Policy for Infectious Disease Dynamics (RAPIDD) program of the Science and Technology Directorate, Department of Homeland Security, and the Fogarty International Center, National Institutes of Health.