Simulation Models Examining the Effect of Brugian ... - CiteSeerX

Am. J. Trop. Med. Hyg., 80(1), 2009, pp. 44–50 Copyright © 2009 by The American Society of Tropical Medicine and Hygiene

Simulation Models Examining the Effect of Brugian Filariasis on Dengue Epidemics Jefferson A. Vaughan,* Dana A. Focks, and Michael J. Turell US Army Medical Research Institute of Infectious Diseases, Fort Detrick, Maryland; Diagnostic Systems Division, Department of Biology, University of North Dakota, Grand Forks, North Dakota; Infectious Disease Analysis, Gainesville, Florida

Abstract. Concurrent ingestion of microfilariae (mf) and arboviruses by mosquitoes can enhance the transmission of virus compared with when virus is ingested alone. We studied the effect of mf enhancement on the extrinsic incubation period (EIP) of dengue 1 virus within Aedes aegypti mosquitoes by feeding mosquitoes on blood that either contained virus plus Brugia malayi mf or virus only. Mosquitoes were sampled over time to determine viral dissemination rates. Co-ingestion of mf and virus reduced viral EIP by over half. We used the computer simulation program, DENSiM, to compare the predicted patterns of dengue incidence that would result from such a shortened EIP versus the EIP derived from the control (i.e., virus only) group of mosquitoes. Results indicated that, over the 14-year simulation period, mf-induced acceleration of the EIP would generate more frequent (but not necessarily more severe) epidemics. Potential interactions between arboviruses and hematozoans deserve closer scrutiny. INTRODUCTION

additional complexity to what may already be rather complex transmission systems for many zoonotic arboviruses (i.e., transmission systems involving multiple species of amplifying hosts and vectors). In this study, we chose a relatively simple model system; dengue-1 virus, Aedes aegypti mosquitoes, and Brugia malayi mf. Transmission of dengue typically involves only one amplifying host species (i.e., humans) and one vector species (i.e., Ae. aegypti). Other reasons for choosing this particular system were 1) all three components are sympatric in nature, 2) dengue and Brugian filariasis have public health relevance, 3) B. malayi mf are readily obtained from the National Institute of Allergy and Infectious Disease Filariasis Research Reagent Repository Center (NIAID FR3), and most importantly, 4) a simulation model capable of calculating the effects of mf enhancement on dengue epidemiology has been developed and is available for use.10 The experimental approach was to feed mosquitoes on blood containing either virus or virus plus mf and to use the experimentally derived values on EIP as input values in computer simulations to determine what effect microfilarial infections in a host population might have on the long-term patterns of arboviral disease incidence within an area where both filarial and arboviral infections co-exist.

It has been shown that the per os infectivity and subsequent transmission of arboviruses by mosquitoes and Culicoides (= biting midges) can be significantly greater when they ingest virus concurrently with microfilariae (mf) than when they ingest the same dose of virus alone.1–6 This phenomenon has been referred to as microfilarial enhancement of arboviral transmission. It is based on the supposition that, in nature, vertebrate hosts of arboviruses (including humans) can harbor pre-existing filarial infections with circulating mf. If these vertebrates become infected with an arbovirus, they will become dually infected (i.e., they will develop a concurrent microfilaremia and viremia). Most mf species penetrate the vector midgut soon after being ingested and migrate to a preferred site of development (e.g., flight muscle, fat body). If a microfilaremic host is also viremic when fed on by a vector, there is the possibility that penetration of the midgut by mf will allow some of the ingested virus to enter directly into the vector hemocoel. The facilitated movement of virus directly into the hemocoel can have two important epidemiologic consequences. First, it can increase vector competence by allowing virus to bypass both midgut infection and midgut escape barriers,7,8 transforming otherwise incompetent vector species into competent vector species1,2,4–6 and thus increasing the number of vector species involved in an arbovirus transmission cycle. Second, mf enhancement can accelerate arboviral development within a vector2,3 and thus shorten the time required for a virus-exposed vector to become infectious (i.e., shorten the extrinsic incubation period [EIP]). This report examines the second of these two potential consequences (i.e., acceleration of the viral EIP). In theory, mf-mediated acceleration of viral EIP could occur in many different species combinations of virus-arthropod-filaria where mf are capable of penetrating the mosquito midgut, including combinations in which the filarial species are not normally associated with or vectored by a particular arthropod species.9 In fact, it is not necessary for the mf to be able to survive in the arthropod for microfilarial enhancement to occur.5,6 This can introduce

MATERIALS AND METHODS Mosquitoes. Aedes aegypti BLACKEYE strain were reared at 27°C using standard methodology. Before use in membrane feeding experiments, cages of 150–200 4- to 7-day-old adult female mosquitoes were maintained overnight without a sugar source to promote blood feeding. Virus. One week before conducting membrane feedings, frozen dengue 1 virus stocks were thawed and inoculated into the thoraces of ~30 Ae. aegypti mosquitoes (inoculum size, ~0.3 µL). Mosquitoes were maintained at 31°C for 1 week to allow the virus to grow. On the day of membrane feeding, the virus-inoculated mosquitoes were triturated in 1 mL of diluent and centrifuged, and the supernatant containing the virus was added to 10 mL of fresh, pre-warmed human blood containing EDTA. Parasites. Tubes containing live B. malayi mf harvested from peritoneally infected gerbils were shipped overnight from the University of Georgia, courtesy of the NIAID FR3. On receipt, mf were centrifuged and washed twice in diluent, checked for viability, held briefly at 32°C, and used for membrane feeding the same day.

* Address correspondence to Jefferson A. Vaughan, Department of Biology, PO Box 9019, Grand Forks, ND 58202-9019. E-mail: [email protected]

44

MICROFILARIAL ENHANCEMENT OF DENGUE TRANSMISSION

Membrane feeding. For each of the two membrane feeding trials, equal volumes of the blood-virus mixture were placed in two labeled tubes. Microfilariae were added to one of the tubes (= mf + virus), and an equivalent volume of diluent was added to the other tube (= control group). The tubes were mixed thoroughly, and the two solutions were dispensed into the chambers of pre-warmed glass membrane feeders covered at one end with a Baudruche membrane and connected to a circulating water bath (32°C). Approximately 0.5 mL of each solution were stored at –70°C for later viral titration. The feeders were lowered on to the screened tops of mosquito cages and mosquitoes were allowed ~1 hour to feed. Unfed mosquitoes were removed. In the first trial, cages were incubated at a constant 31°C. In the second trial, one set of cages was incubated at 28°C and a second set of cages was incubated at 24°C. Immediately after mosquito feedings in both trials, midguts from three to five mosquitoes from the mf + virus groups were carefully excised intact and placed in individual droplets of saline. After several hours of incubation, the excised midguts were examined microscopically, and the numbers of mf penetrating each midgut were recorded. These midguts were stored individually at –70°C for later viral titration. At intervals of 3, 5, 7, 10, 11, 13, and 18 days after feeding, 10–20 live mosquitoes were removed from each cage, chilled, and placed on ice to anesthetize them. The legs were removed from each mosquito, and individual mosquito bodies and legs were triturated in diluent and frozen at –70°C for later viral assay. Mosquito body samples were assayed for virus to determine viral infection rates. If a mosquito body was found to be virus positive, the corresponding leg sample was assayed to determine whether that infected mosquito had a disseminated viral infection (i.e., virus present in the legs) or a non-disseminated infection confined to the midgut (i.e., no virus in the legs).11 Mosquito mortality. At various time points throughout these experiments, dead mosquitoes in each cage were removed and counted. A 21-day observation period was used for the 24 and 28°C groups, and a 13-day observation period was used for the 31°C groups. At the end of each observation period, all mosquitoes remaining alive were counted. Daily survivorship was calculated using the following formula: (proportion of mosquitoes surviving the observation period)(1/duration of the observation period) Mosquitoes taken for viral assays were not included as part of the mortality assessments. Virus assays. Specimens (e.g., blood-virus mixtures, mosquito bodies, and mosquito legs) were tested at serial 10-fold dilutions for infectious virus by plaque assays on Vero cell monolayers, modified from the methods described by Gargan and others.12 Data analyses. χ2 analyses were used to test for significant differences between experimental groups in the rates of viral infection, dissemination, and mosquito mortality. Microfilarial counts were expressed as geometric means. Viral extrinsic incubation period for each experimental group of mosquitoes was defined as the period of time elapsed before a mosquito was detected with a disseminated infection (i.e., virus in the legs). Computer simulations. The dengue transmission model, DENSiM,10 and its entomologic companion program

45

CIMSiM13,14 were used to estimate the theoretical consequences of Brugian filariasis on the long-term transmission of dengue. The entomologic factors generated from CIMSiM were used to create the theoretical mosquito populations, which were used in the DENSiM simulations. The DENSiM simulated a theoretical human population dynamics driven by country- and age-specific birth and death rates. Both CIMSiM and DENSiM are site specific because each country has its own unique demography and climate (which is a key determinant for mosquito development and survival). Therefore, to generate realistic output, the simulations needed accurate databases of human demography and daily weather information for specific countries and regions. For this project, Indonesia was chosen as the most appropriate country in which to run CIMSiM and DENSiM simulations because 1) Indonesia is endemic for both dengue and Brugian filariasis and 2) there are historical demographic, weather, and entomologic data sets available. The CIMSiM simulations were parameterized using actual entomologic field surveys conducted in Indonesia by the US Navy and Gadjah Mada University. Weather data were courtesy of the Indonesian Air Force. Simulations were conducted to cover a 14-year period from 1985 to 1999. The CIMSiM used a default value for daily adult survival (Sa) set at 0.89.15 For each CIMSiM output, DENSiMs were run to simulate dengue transmission, assuming initial seroprevalence to circulating dengue virus at ~70%, independent of age. Four basic scenarios were compared. The “control scenario” (i.e., no mf enhancement) used Sacontrol = 0.89 and a standardized EIP values based on previous observations.16,17 Because the EIP varies according to environmental temperature, an enzyme kinetics model18 was used to correlate EIP with temperature. Another series of simulations were conducted to analyze the outcome of B. malayi mf infections on dengue transmission (i.e., the “filariasis-endemic scenarios”). These simulations used the same enzyme kinetics model for the EIP–temperature relationship but adjusted the standardized EIP input values in proportion to the reduced EIP observed for the mf + virus groups in our membrane-feeding experiments. Three different simulations were performed for filariasis-endemic scenarios. One used the same default value for adult daily survival as in the control scenario (i.e., Samf = 0.89) to represent the scenario where mf penetration of Ae. aegypti midguts did not reduce adult daily mortality. Two additional simulations substituted lower values of adult daily survival— i.e., Samf = 0.87 and Samf = 0.85—in recognition of the fact that some studies have reported that mf ingestion and penetration of the mosquito midgut can reduce mosquito survival, whereas others have found no effect.19–21 The outcomes for each of the four scenarios were the number of new dengue cases generated each successive year throughout the 14-year simulation period. For each of the four scenarios, yearly incidences of dengue were graphed to compare the long-term patterns of dengue incidence. RESULTS Virus concentration and mf midgut penetration during infectious feeding. In the first trial, mosquitoes were incubated at 31°C, and the concentrations of dengue 1 virus were 106.0 plaque-forming units (PFU)/mL for the infectious blood contained within the membrane feeders and 103.1 PFUs for

46

VAUGHAN AND OTHERS

individual mosquito blood meals. In the second trial, mosquitoes were incubated at 28°C and 24°C, and the concentration of virus was considerably lower—105.3 PFU/mL for the infectious blood within the feeders and 102.6 PFU for individual mosquito blood meals. The intensity of mf penetration through mosquito midguts was similar for both trials—with geometric means of 31.6 and 26.4 mf per mosquito for the first and second trials, respectively. Mosquito mortality. There were no significant differences in mosquito mortalities between the treatment groups for mosquitoes held at 24°C (χ2 = 0.5, P = 0.47) or 28°C (χ2 = 3.6, P = 0.16; Table 1). However, under the 31°C temperature, mosquito mortality in the mf + virus group (26%) was significantly higher than in the virus only group (1%; χ2 = 23.7, P < 0.001; Table 1). When the data from all three temperature regimens were combined, there was no significant difference in overall mosquito mortality between the mf + virus treatment group (31%, N = 367) and virus only group (25%, N = 374; χ2 = 3.6, P = 0.060). The daily survivals in all groups (Sa ≥ 0.97) were substantially higher than estimated daily survival of Ae. aegypti in the field (Sa = 0.89).15 Viral infection. The higher virus concentration used in the first trial (i.e., the 31°C trial) resulted in a significantly higher viral infection rate in mosquitoes (68%, N = 180) compared with the viral infection rates observed during the second trial (9%; N = 322; χ2 = 201.3, P < 0.001; Table 2). However, there were no significant differences in overall viral infection rates of mosquitoes between treatment groups when all temperature regimens were combined (χ2 = 0.03, P = 0.85) nor were there significant differences in mosquito infection rates among treatment groups within a temperature regimen (χ2 = 2.7, P ≥ 0.10). Viral dissemination and EIP. As expected, the development of disseminated viral infections occurred more rapidly at warmer temperatures (Tables 3 and 4). When mosquitoes ingested high titer virus and were incubated at 31°C, the viral dissemination rate of the mf + virus group (61%) was significantly higher than that of the virus only group (28%; χ2 = 18.1, P < 0.03). When all temperature regimens were combined, overall viral dissemination rates were significantly higher in the mf + virus group (26%, N = 244) than in the virusonly group (10%, N = 264; χ2 = 21.2, P < 0.001). However, if

mosquitoes were counted beginning on the day that disseminated infection were first detected for each group, overall viral dissemination rates did not differ between the mf + virus group (32%, N = 199) and the virus-only group of mosquitoes (24%, N = 114; χ2 = 2.5, P = 0.11). More importantly, disseminated infections developed much sooner in the mf + virus groups of mosquitoes than in the corresponding virus-only group of mosquitoes. At each of the three temperature regimens, presence of B. malayi mf in the infectious blood meal reduced the development of disseminated viral infections by over half (Table 3). The time it took for disseminated viral infections to develop served as a proxy for EIP in the DENSiM simulations. Because temperature did not affect the magnitude of the reduction (Table 4), we multiplied the standardized EIPs used in DENSiM by the averaged proportional reduction observed in our experiments for all three temperatures (i.e., multiplied by 0.40) and substituted this reduced EIP into the DENSiM for simulations involving filariasis-endemic scenarios. Simulation analysis. The DENSiM simulations showed just how sensitive dengue transmission dynamics are to reductions in adult survival. When adult daily survival for the filariasisfree scenario was set at Sacontrol = 0.89 and compared with a filariasis-endemic scenario wherein ingestion of mf caused high mosquito mortality (Samf = 0.85), the patterns of dengue epidemics between filariasis-free and filariasis-endemic scenarios looked nearly identical (Figure 1A). In other words, because of the trade-off between reduced EIP and reduced survival, mf enhancement accompanied with high mf-induced mosquito mortality had no net effect on dengue transmission. Interestingly, simulations where mf-associated reductions in mosquito survival were either low (Samf = 0.87; Figure 1B) or absent altogether (Samf = 0.89; Figure 1C) yielded no substantial differences in the overall average numbers of dengue cases throughout the 14-year simulation period. The filariasis-free scenario where Sacontrol = 0.89 produced an average of 2,480 cases/yr, the filariasis-endemic scenario with low mf-induced mortality (Samf = 0.87) produced an average of 2,832 cases/yr, and the filariasis-endemic scenario with no mf-induced mortality (Samf = 0.89) produced an average of 3,019 cases/yr. However, mf enhancement did change the long-term pattern of dengue incidence by changing the temporal dynamics of

Table 1 Numbers of Ae. aegypti mosquitoes dying after ingesting either dengue 1 virus + B. malayi mf or dengue 1 virus only and held thereafter at different temperatures 24°C Day

1 2 3 5 7 8 10 11 13 18 21 Mortality rate (N)* Daily survival†

mf + virus

nd 12 11 5 5 nd nd 3 nd 2 0 31% (122) 0.9824

28°C

31°C

Virus only

mf + virus

Virus only

mf + virus

Virus only

nd 6 5 6 7 nd nd 7 nd 4 8 35% (121) 0.9793

nd 22 9 5 1 nd nd 2 nd 0 0 37% (105) 0.9781

nd 18 7 5 7 nd nd 6 nd 2 3 30% (166) 0.9839

9 17 nd 7 1 1 0 nd 1 – – 26% (140) 0.9774

0 0 nd 0 1 0 0 nd 0 – – 1% (87) 0.9991

* Percentage of mosquitoes found dead by Day 21 (number of mosquitoes in original population). † Daily survival = (proportion surviving the observation period)(1/duration of the observation period). nd = not determined.

47

MICROFILARIAL ENHANCEMENT OF DENGUE TRANSMISSION

Table 2 Rates of dengue 1 virus infection in Ae. aegypti fed through membrane feeding on either virus + B. malayi mf or virus only and held thereafter at selected temperatures Holding temperature (infectious dose in PFU/mL) 24°C (105.3)

28°C (105.3)

31°C (106.0)

Day after blood meal

mf + virus

Virus only

mf + virus

Virus only

mf + virus

Virus only

3 5 7 10 11 13 18 Overall virus infection (N)

0/10* 0/20 1/20 – 2/20 2/14 – 6% (84)

– – 1/10 – 3/20 3/20 2/15 14% (65)

1/15 2/20 1/20 – 1/11 – – 8% (66)

1/14 3/20 3/20 – 0/20 1/20 1/23 8% (117)

14/20 13/20 12/20 11/20 – 11/14 – 65% (94)

8/10 14/20 13/20 14/20 – 13/16 – 72% (86)

Viral concentrations fed to mosquitoes were lower (105.3 PFUs) for the 24°C and 28°C groups than for the 31°C groups (106.0 PFUs). * Number of mosquitoes infected/number tested.

infections, so that the epidemic curves in filariasis-endemic scenarios were out-of-phase with the normal course of the epidemiology. Indeed, in the filariasis-endemic scenario where there was no mf-associated reduction in daily mosquito survival (Figure 1C), the periodicity of the inter-epidemic intervals was considerably shortened, resulting in more frequent “mini-epidemics” occurring throughout the 14-year simulation period than would otherwise occur without the filariasis (i.e., five outbreaks instead of the usual three). DISCUSSION Vertebrate hosts, dually infected with mf and an arbovirus, can transform incompetent mosquito species into competent vectors,1,2,5,6 and perhaps more importantly, reduce the EIP of the virus within vector and non-vector species alike.2,3 The existence of chronic filarial infections within a host population subject to periodic arboviral epidemics could potentially lead to a portion of that population being infected with both microfilariae and virus. In the 1980s, Turell and others showed that co-ingestion of B. malayi mf with Rift Valley fever virus2 and with dengue 2 virus3 increased the rapidity of viral dissemination in Ae. taeniorhynchus and Ae. aegypti mosquitoes, respectively. This study corroborates this phenomenon using dengue 1 virus within its primary vector, Ae. aegypti. Regardless of the temperatures at which mosquitoes were maintained, co-ingestion of B. malayi mf and dengue 1 virus by Ae. aegypti mosquitoes reduced the time interval between ingestion of the viremic blood meal and the development of

disseminated dengue 1 virus by over half that of the normal duration. To understand how mf-mediated acceleration in viral dissemination might affect viral epidemiology, we used a comprehensive computer simulation model of dengue virus transmission, DENSiM, and adjusted the key input parameters (e.g., daily temperature, rainfall, human seroprevalence) to represent a region of the world where both dengue and Brugian filariasis co-exist (i.e., Indonesia). We compared the long-term pattern of dengue incidence that resulted from the computer simulations when normal EIP values (derived from the control group of mosquitoes) were used versus when EIP values were multiplied by a factor of 0.40 (derived from the mf + virus group of mosquitoes; Table 4). The simulations showed that the mf-induced reduction in the EIP produced an initial peak in disease incidence during the first year of virus introduction, but after that, it did not necessarily lead to an overall increase in the total number of new human cases over the ensuing years. Rather, simulations suggested that the reduced EIP would generate more frequent (but not necessarily larger) epidemic waves, such that over the 14-year simulation period, there were five rather than the usual three epidemics that resulted from running simulations with normal EIP values (Figure 1). However this enhancing effect disappeared if scenarios were simulated whereby mf penetration of the mosquito midgut lowered adult mosquito daily survival. Because mf-induced reductions in EIP and daily survival are opposing forces, it is important to understand the effect that mf penetration has on mosquito daily survival in nature.

Table 3 Development of viral dissemination in Ae. aegypti after feeding on either virus + B. malayi mf or on virus alone and held at selected temperatures Holding temperature (infectious dose in PFU/mL) 24°C (105.3) Day after infectious blood meal

mf + virus

3 5 7 10 11 13 18 Dissemination rate (N) Dissemination rate ≥ EIP (N)

0/10 0/20 1/20* – 1/20 2/14 – 5% (84) 7% (54)

28°C (105.3) Virus only

– – 0/10 – 0/20 0/20 1/15* 2% (65) 6% (15)

mf + virus

0/15 1/20* 1/20 – 1/11 – – 5% (66) 6% (51)

31°C (106.0)

Virus only

mf + virus

Virus only

0/14 0/15 0/20 – 0/20 1/20* 1/23 2% (112) 5% (43)

11/20* 13/20 11/20 11/20 – 11/14 – 61% (94) 61% (94)

0/10 0/20 8/20* 8/20 – 8/16 – 28% (86) 43% (56)

Viral concentrations fed to mosquitoes were lower (105.3 PFUs) for the 24°C and 28°C groups than for the 31°C groups (106.0 PFUs). Asterisks indicate the first appearance of disseminated infections within each group.

48

VAUGHAN AND OTHERS

Table 4 EIP of dengue 1 virus in Ae. aegypti mosquitoes as a function of temperature and co-ingestion of B. malayi mf Temperature (°C)

Treatment

24

Mf + virus Virus only Mf + virus Virus only Mf + virus Virus only

28 31 Average

EIP (days)

7 18 5 13 3 7

Proportional reduction in EIP caused by co-ingested mf

0.39 – 0.38 – 0.43 – 0.40

In our experimental trials, the overall mosquito mortality caused by ingestion of mf was not statistically significant from that observed in the control groups (Table 1). This contrasts with our earlier findings where significant mortality was observed in Ae. aegypti ingesting B. malayi mf.6 The difference between the two studies probably reflects differences in the

levels of mf densities ingested by the mosquitoes (~5,000 versus 10,000 mf/mL blood in the earlier study). It has been documented that Aedes spp. mosquitoes incur increasing mortality when fed increasing densities of mf, but significant mortality occurs only when mf intensities are very high (> 5,000 mf/ mL, > 20 mf ingested).22,23 In Indonesia, mf intensity of Brugia infections may range from 1 to 6,028 mf/mL.24,25 Unfortunately, there has been little experimental work quantifying Ae. aegypti mortality after feeding on microfilaremic people infected with B. malayi. However, there are reports in the literature about the more widespread human filarid, Wuchereria bancrofti and Ae. aegypti. Although Ae. aegypti is not considered an important vector for Bancroftian filariasis, W. bancrofti mf readily penetrate the midgut of Aedes mosquitoes after ingestion.26–29 Under natural conditions, mosquito mortality resulting from ingestion of W. bancrofti mf is generally nil. For example, Calheiros and others30 reported that the survival rates over 21 days of a Brazilian population of Ae. aegypti fed on microfilaremic people (mf intensity = 906–1,830 mf/mL) were virtually

Figure 1. DENSiM simulations on the effect of B. malayi mf on dengue 1 epidemiology whereby mf reduces the EIP of dengue 1 virus in Ae. aegypti by an average of 0.40 over that of the normal EIP (Table 2) and either reduces daily mosquito survival to 0.85 (high excess mortality; top panel), 0.87 (low excess mortality; middle panel), or 0.89 (no excess mortality; bottom panel).

MICROFILARIAL ENHANCEMENT OF DENGUE TRANSMISSION

identical to the survival rates of Ae. aegypti fed on non-microfilaremic people. Likewise, in Samoa, Samarawickrema and other31 reported that local populations of Ae. polynesiensis and Ae. samoanus did not suffer an increase in mortality when fed on W. bancrofti –infected people (mf intensity = 1–1,300 mf/mL). Failloux and others32 observed that, although mfinduced mortality in Ae. polynesiensis did occur from ingesting W. bancrofti mf, mosquito mortality was less a function of mf intensity (range, 700–8,000 mf/mL) and more a function of the geographic origin of the mosquito strain—that is, mortality was greater in allopatric combinations of parasite and mosquito, suggesting that co-adaptation of local strains of parasite/mosquito may serve to mitigate mf-induced mortality. When a local population of Ae. aegypti in Trinidad fed on people infected with another human filarid, Mansonella ozzardi, there was no excess mosquito mortality even though M. ozzardi mf penetrated the midgut.9 Taken together, the evidence to date indicates that reduced survival of Ae. aegypti as the result of feeding on microfilaremic people in nature may be mild or absent, suggesting that the DENSiM simulations of filariasis-endemic scenarios where there is no decrease in normal daily mosquito survival (Figure 1, bottom panel) is the more realistic scenario over filariasis-endemic scenarios that incorporate a reduced mosquito survival (Figure 1, top and middle panels). The prevalence of mf infections within a vertebrate host population is of paramount importance to the real-life significance of mf enhancement of arborviral transmission. The higher the prevalence of mf infections in a population, the more likely dual infections could occur and the more likely that mf enhancement could happen. Prevalence estimates of Brugian microfilaremia in endemic areas of Malaysia and Indonesia range from 12% to 30%.24,33–35 Although they have not been tested in the laboratory, other microfilarial species found in human blood (e.g., W. bancrofti, M. ozzardi, Loa loa) could also, in theory, contribute to accelerated dengue EIP in Ae. aegypti. Thankfully, the global prevalence of lymphatic filariasis is in decline, in large part because of the efforts of the Global Campaign for the Eradication of Lymphatic Filariasis.36 Lymphatic filariasis may be eliminated by 2020. Therefore, will the elimination of human lymphatic filariasis make the results of these DENSiM simulations irrelevant? For dengue, the answer may be yes. However, the general principles and outcome of these simulations may apply to scenarios involving other arboviruses, particularly zoonotic arboviruses. The prevalence of mf infections in wildlife is often very high. If an arboviral epidemic swept through such populations, there is a good chance that dual infections would occur. For example, as West Nile virus (WNV) spread westward across North America during the early to mid-2000s, the virus encountered susceptible populations of adult American crows in southern Ontario37 and upstate New York (JAV and L. Patrican, unpublished data) with 65–90% prevalence of Chandlerella chitwoodae mf infections, adult common grackles in Indiana38 and Illinois39 with > 90% prevalence of Ch.quiscali mf infections,and adult black-billed magpies in northern Colorado40,41 with 79–90% prevalence of Splendidofilaria picardina mf infections. There can be little doubt that during WNV expansion across the continent, mosquitoes fed on dually infected crows, grackles, and magpies. Preliminary evidence has indicated that at least two of these mf species, Ch. quiscali and S. picardina, can penetrate the midguts of

49

several North American mosquito species.41,42 Therefore, the question becomes is it possible that such dually infected birds could have enhanced amplification of WNV by reducing the viral EIP within the vector? The answer at present is unknown. However, considering the lack of other reasonable hypotheses and/or mathematical models to explain the incredibly rapid spread of WNV through North American bird populations, together with the simulation results presented in this report, we suggest that the potential of mf infections to enhance arboviral transmission in nature warrants closer scrutiny. Received June 19, 2008. Accepted for publication August 4, 2008. Acknowledgments: Microfilarial parasites were provided by the University of Georgia through the National Institute of Allergy and Infectious Disease Filariasis Research Reagent Repository Center. The authors thank K. Kenyon for editorial assistance. Research was conducted in compliance with the Animal Welfare Act and other federal statutes and regulations relating to animals and experiments involving animals and adheres to principles stated in the Guide for the Care and Use of Laboratory Animals, National Research Council, 1996. The facility where this research was conducted is fully accredited by the Association for Assessment and Accreditation of Laboratory Animal Care International. Financial support: This study was supported in part by the National Research Council Senior Fellowship Program and National Institutes of Health Grant AI49477 (JAV). Disclaimer: The views of the authors do not necessarily reflect the position of the Department of Defense or the Department of the Army. Authors’ addresses: Jefferson A. Vaughan, Department of Biology, University of North Dakota, Grand Forks, ND 58202-9019, E-mail: [email protected]. Dana A. Focks, Infectious |Disease Analysis, PO Box 12852, Gainesville, FL 32604, E-mail: [email protected]. Michael J. Turell, Virology Division, USAMRIID, 1425 Porter Street, Fort Detrick, MD 21702-5011, E-mail: [email protected].

REFERENCES 1. Mellor PS, Boorman J, 1980. Multiplication of bluetongue virus in Culicoides nubeculosus (Meigen) simultaneously infected with the virus and the microfilariae of Onchocerca cervicalis (Railliet & Henry). Ann Trop Med Parasitol 74: 463–469. 2. Turell MJ, Rossignol PA, Spielman A, Rossi CA, Bailey CL, 1984. Enhanced arboviral transmission by mosquitoes that concurrently ingested microfilariae. Science 225: 1039–1041. 3. Turell MJ, Mather TN, Spielman A, Bailey CL, 1987. Increased dissemination of dengue 2 virus in Aedes aegypti associated with concurrent ingestion of microfilariae of Brugia malayi. Am J Trop Med Hyg 37: 197–201. 4. Zytoon EM, El-Belbasi HI, Matsumura T, 1993. Mechanism of increased dissemination of chikungunya virus in Aedes albopictus mosquitoes concurrently ingesting microfilariae of Dirofilaria immitis. Am J Trop Med Hyg 49: 201–207. 5. Vaughan JA, Turell MJ, 1996. Dual host infections: enhanced infectivity of eastern equine encephalitis virus to Aedes mosquitoes mediated by Brugia microfilariae. Am J Trop Med Hyg 54: 105–109. 6. Vaughan JA, Trpis M, Turell MJ, 1999. Brugia malayi microfilariae enhance the infectivity of Venezuelan equine encephalitis virus to Aedes mosquitoes. J Med Entomol 36: 758–763. 7. Chamberlain RW, Sudia WD, 1961. Mechanism of transmission of viruses by mosquitoes. Annu Rev Entomol 6: 371–390. 8. Kramer LD, Hardy JL, Presser SB, Houk EJ, 1981. Dissemination barriers for western equine encephalomyelitis virus in Culex tarsalis infected after ingestion of low viral doses. Am J Trop Med Hyg 30: 190–197. 9. Vaughan JA, Bell JA, Turell MJ, Chadee DD, 2007. Passage of ingested Mansonella ozzardi (Spirurida: Onchocercidae)

50

10.

11. 12.

13.

14.

15. 16.

17. 18. 19.

20.

21.

22. 23.

24.

25.

VAUGHAN AND OTHERS

microfilariae through the midgut of Aedes aegypti mosquitoes (Diptera: Culicidae). J Med Entomol 44: 111–116. Focks DA, Daniels E, Haile DH, Keesling JE, 1995. A simulation model of the epidemiology of urban dengue fever: literature analysis, model development, preliminary validation, and samples of simulation results. Am J Trop Med Hyg 53: 489–506. Turell MJ, Gargan TP II, Bailey CL, 1984. Replication and dissemination of Rift Valley fever virus in Culex pipiens. Am J Trop Med Hyg 33: 176–181. Gargan TP, Bailey CL, Higbee GA, Gad A, El Said S, 1983. The effect of laboratory colonization of the vector-pathogen interactions of Egyptian Culex pipiens and Rift Valley fever virus. Am J Trop Med Hyg 32: 1154–1163. Focks DA, Haile DH, Daniels E, Mount GA, 1993. Dynamic life table model of a container-inhabiting mosquito, Aedes aegypti (L.) (Diptera: Culicidae). Part 1. Analysis of the literature and model development. J Med Entomol 30: 1003–1017. Focks DA, Haile DH, Daniels E, Mount GA, 1993. Dynamic life table model of a container-inhabiting mosquito, Aedes aegypti (L.) (Diptera: Culicidae). Part 2: Simulation results and validation. J Med Entomol 30: 1018–1028. Muir LE, Kay BH, 1998. Aedes aegypti survival and dispersal estimated by mark-release-recapture in northern Australia. Am J Trop Med Hyg 58: 277–282. McLean DM, Clarke AM, Coleman JC, Montalbetti CA, Skidmore AG, Walter TE, Wise R, 1974. Vector capability of Aedes aegypti mosquitoes for California encephalitis and dengue viruses at various temperatures. Can J Microbiol 20: 255–262. Watts DM, Burke DS, Harrison BA, Whitmire RE, Nisalak A, 1987. Effect of temperature on the vector efficiency of Aedes aegypti for dengue 2 virus. Am J Trop Med Hyg 36: 143–152. Sharpe PHJ, DeMichele DW, 1977. Reaction kinetics of poikilotherm development. J Theoret Biol 64: 639–670. Krishnamoorthy K, Subramanian S, Van Oortmarssen GJ, Habbema JD, Das PK, 2004. Vector survival and parasite infection: the effect of Wuchereria bancrofti on its vector Culex quinquefasciatus. Parasit 129: 43–50. Brito AC, Fontes G, Williams P, Rocha EM, 1998. Bancroftian filariasis in Maceio, state of Alagoas, Brazil: Observations on Culex quinquefasciatus after blood feeding on individuals with different densities of microfilariae in the peripheral blood stream. Am J Trop Med Hyg 58: 489–494. Russell RC, Geary MJ, 1996. The influence of microfilarial density of dog heartworm Dirofilaria immitis on infection rate and survival of Aedes notoscriptus and Culex annulirostris from Australia. Med Vet Entomol 10: 29–34. Ibrahim MS, Trpis M, 1987. The effect of Brugia pahangi infection on survival of susceptible and refractory species of the Aedes scutellaris complex. Med Vet Entomol 1: 329–337. Klowden MJ, 1981. Infection of Aedes aegypti with Brugia pahangi administered by enema: results of quantitative infection and loss of infective larvae during blood feeding. Trans R Soc Trop Med Hyg 75: 354–358. Melrose W, Rahmah N, 2006. Use of Brugia Rapid dipstick and ICT test to map distribution of lymphatic filariasis in the Democratic Republic of Timor-Leste. Southeast Asian J Trop Med Public Health 37: 22–25. Supali T, Ismid IS, Wibowo H, Djuardi Y, Majanati E, Ginanjar P, Fischer P, 2006. Estimation of the prevalence of lymphatic filariasis by a pool screen PCR assay using blood spots on filter paper. Trans R Soc Trop Med Hyg 100: 753–759.

26. Rosen L, 1955. Observations on the epidemiology of human filariasis in French Oceania. Am J Hyg 61: 219–248. 27. McGreevy PB, Kostrup N, Tao J, McGreevy MM, de Marshall TF, 1982. Ingestion and development of Wuchereria bancrofti in Culex quinquefasciatus, Anopheles gambiae and Aedes aegypti after feeding on humans with varying densities of microfilariae in Tanzania. Trans R Soc Trop Med Hyg 76: 288–296. 28. Lowichik A, Lowrie RC, 1988. Uptake and development of Wuchereria bancrofti in Aedes aegypti and Haitian Culex quinquefasciatus that were fed on a monkey with low-density microfilaremia. Trop Med Parasit 39: 227–229. 29. Zielke E, 1992. On the uptake of Wuchereria bancrofti microfilariae in vector mosquitoes of different susceptibility to filarial infection. Angew Parasitol 33: 91–95. 30. Calheiros ML, Fontes G, Williams P, Rocha EMM, 1998. Experimental infection of Culex (Culex) quinquefasciatus and Aedes (Stegomyia) aegypti with Wuchereria bancrofti. Mem Inst Oswaldo Cruz 93: 855–860. 31. Samarawickrema WA, Spears GFS, Sone F, Ichimori K, Cummings RF, 1985. Filariasis in Samoa. II. Some factors related to the development of microfilariae in the intermediate host. Ann Trop Med Parasitol 79: 101–107. 32. Failloux A-B, Raymond M, Ung A, Glaziou P, Martin PMV, Pasteur N, 1994. Variation in the vector competence of Aedes polynesiensis for Wuchereria bancrofti. Parasitol 112: 19–29. 33. Rahmah N, Ashikin AN, Anuar AK, Ariff RH, Abdullah B, Chan GT, Williams SA, 1998. PCR-ELISA for the detection of Brugia malayi infection using finger-prick blood. Trans R Soc Trop Med Hyg 92: 404–406. 34. Cox-Singh J, Pomrehn AD, Rahman HA, Zakaria R, Miller AO, Singh B, 1999. Simple blood-spot sampling with nested polymerase chain reaction detection for epidemiology studies on Brugia malayi. Int J Parasitol 29: 717–721. 35. Supali T, Wibowo H, Ruckert P, Fischer K, Ismid IS, Purnomo, Djuardi Y, Fischer P, 2002. High prevalence of Brugia timori infection in the highlands of Alor Island, Indonesia. Am J Trop Med Hyg 66: 560–565. 36. Anonymous, 2007. Global programme to eliminate lymphatic filariasis. Wkly Epidemiol Rec 82: 361–380. 37. Bartlett CM, Anderson RC, 1980. Filarioid nematodes (Filarioidea: Onchocercidae) of Corvus brachyrhynchos brachyrhynchos Brehm in southern Ontario, Canada and a consideration of the epizootiology of avian filariasis. Syst Parasitol 2: 77–102. 38. Welker GW, 1962. Helminth parasites of the common grackle, Quiscalis quiscula versicolor Veillot in Indiana. PhD dissertation, The Ohio State University, Columbus, OH. 39. Granath WO, 1980. Fate of the wild avian filarial nematode Chandlerella quiscali (Onchocercidae: Filarioidae) in the domestic chicken. Poult Sci 59: 996–1000. 40. Stabler RM, 1961. Studies of the age and seasonal variations in the blood and bone marrow parasites of a series of black-billed magpies. J Parasitol 47: 413–416. 41. Hibler CP, 1963. Onchocercidae (Nematoda: Filarioidea) of the American Magpie, Pica pica hudsonia (Sabine), in northern Colorado. PhD dissertation, Colorado State University, Fort Collins, CO. 42. Brewer CM, 2006. The potential for microfilarial enhancement of West Nile virus transmission in the Red River Valley of North Dakota and Minnesota to occur. MS thesis, University of North Dakota, Grand Forks, ND.