salivary gland morphology and virus transmission during ... - CiteSeerX

12 downloads 0 Views 1MB Size Report
examined in a Philips (Eindhoven, The Netherlands) CM-100 electron ...... 3.206B Mary Moody Northern Pavilion, 301 University Boulevard,. Galveston, TX ...
Am. J. Trop. Med. Hyg., 76(1), 2007, pp. 118–128 Copyright © 2007 by The American Society of Tropical Medicine and Hygiene

SALIVARY GLAND MORPHOLOGY AND VIRUS TRANSMISSION DURING LONG-TERM CYTOPATHOLOGIC WEST NILE VIRUS INFECTION IN CULEX MOSQUITOES YVETTE A. GIRARD, BRADLEY S. SCHNEIDER, CHARLES E. MCGEE, JULIE WEN, VIOLET C. HAN, VSEVOLOD POPOV, PETER W. MASON, AND STEPHEN HIGGS* Department of Pathology, University of Texas Medical Branch, Galveston, Texas

Abstract. The effect of long-term West Nile virus (WNV) infection on Culex salivary gland morphology and viability was evaluated by transmission electron microscopy during a four week period post-blood feeding. These studies showed that apoptosis and other cytopathologic changes occurred more frequently in WNV-infected mosquitoes compared with uninfected controls. The effect of long-term infection on WNV transmission was evaluated by titering virus in saliva over the same time period. Although the mean titer of WNV in mosquito saliva did not change significantly over time, the percentage of saliva samples containing WNV decreased. Because of the importance of saliva in blood meal acquisition and virus delivery, salivary gland pathology has the potential to affect mosquito feeding behavior and virus transmission. Results from this study add to a growing body of evidence that arbovirus infections in mosquito vectors can be cytopathic, and offer a potential mechanism for virus-induced cell death in mosquitoes. identified cytopathologic changes including cell death in the salivary glands of WNV-infected Cx. p. quinquefasciatus.8 As described herein, we performed additional ultrastructural studies of the salivary glands to differentiate age-related changes from virus-induced changes, and to elucidate the frequency and mechanism of salivary gland cell death associated with WNV replication. Because mosquito salivary gland cells support persistent WNV infection and are responsible for saliva production and secretion, we hypothesized that cytopathologic changes observed by TEM might impair saliva secretion, virus secretion, and virus maturation. Therefore, we assessed the quantity and infectivity of WNV secreted in saliva at time points corresponding to ultrastructural analysis. Previously, we described WNV-induced proliferated membranes in infected Cx. p. quinquefasciatus salivary glands, including smooth membrane structures (SMS), convoluted membranes (CM), tubular proliferated membranes (TPM), and row-like parallel membrane structures (PMS).8 The detection of flaviviral genomic RNA, the replicative form of viral RNA, and nonstructural proteins involved in RNA replication (NS1, NS2A, NS3, NS4A, and NS5) in association with SMS suggest strongly that these membranes are the site of virus replication.10–12 Proteolytic processing of WNV likely occurs in association with virus-induced proliferated membranes CM and TPM, which are hypothesized to be alternate and reversible structures.11–13 Our earlier ultrastructural analysis of both WNV-infected and uninfected salivary glands showed endoplasmic reticulum (ER) vacuolization, phagolysosomal-like vacuoles (PhV), and dense striated membranes (DSM), but we were unable to correlate the abundance of these structures with WNV infection.8 Dead cells were observed in WNV-infected salivary glands, displaying characteristics of apoptosis including shrunken and electron-dense nuclei with nuclear membrane blebbing and chromatin marginalization. In infected tissues, dying cells typically contained abundant SMS, CM, shrunken, electron-dense mitochondria, and non-compartmentalized WNV particles. Our present study was designed to test the hypothesis that apoptotic cell death and other cytopathologies are spatially and temporally correlated with WNV infection of the salivary glands. Included in our analysis is an investigation of the frequency of newly-identified morphologies such as saliva pool-

INTRODUCTION West Nile virus (WNV) is an emerging mosquito-borne virus in the family Flaviviridae, genus Flavivirus and is the etiologic agent of West Nile fever, West Nile encephalitis, and West Nile paralysis.1 Outbreaks of WNV-associated disease in humans, horses, and birds have become more frequent since the early 1990s, and have occurred in Africa, eastern Europe, the Mediterranean Basin, Russia, Australia, and the United States.2,3 In the United States, approximately 11,000 cases of West Nile fever and more than 8,000 cases of neuroinvasive West Nile disease in humans have been reported to the Centers for Disease Control and Prevention since the introduction of the virus into the Western Hemisphere in 1999.4 Susceptible avian hosts that may serve as virusamplifying reservoirs typically become infected with WNV by the bite of a competent mosquito vector such as Culex pipiens quinquefasciatus. These mosquitoes are abundant in tropical and subtropical regions throughout the world, including the southern and western United States.5 In the United States in 2004, Cx. p. quinquefasciatus was the most commonly reported WNV-positive species during mosquito pool testing.6 Earlier immunohistochemical (IHC) and electron microscopic (EM) studies indicated that WNV infection is persistent in midgut, salivary gland, and nervous tissues of these vectors during a four-week period after an infectious artificial blood meal.7,8 What remains unknown is the nature of this long-term flavivirus infection including the effects of continuous WNV infection and replication on mosquito cell viability, tissue function, and virus transmission efficiency. As with other vectors, Cx. p. quinquefasciatus ingest WNV whilst taking a blood meal from a viremic bird. A requisite for transmission of the virus, which can occur in Cx. pipiens as early as five days after ingestion of an infectious blood meal,9 is replication in salivary gland cells and secretion of the virus into the saliva cavity. WNV is inoculated into a host along with mosquito saliva that is secreted by the insect to facilitate blood feeding. Using transmission EM (TEM), we recently

* Address correspondence to Stephen Higgs, Department of Pathology, University of Texas Medical Branch, 2.104 Keiller, 301 University Boulevard, Galveston, TX 77555-0609. E-mail: [email protected]

118

WNV CYTOPATHOLOGY IN MOSQUITOES

ing, dense irregular saliva (DIS), and filamentous cytoplasmic membranes (FCM), as well as determination of the distribution, abundance, and morphology of proliferated membrane structures and paracrystalline arrays of WNV particles. Results from these studies suggest an increase in the frequency of apoptotic cell death may correspond to a decrease in WNV transmission. We conclude from these results that virusinduced changes in mosquito cells at late stages of infection could play a role in arbovirus transmission dynamics in nature. MATERIALS AND METHODS Virus. The lineage I WNV isolate (strain 114) used in mosquito infections was derived from a blue jay brain/liver homogenate tested during surveillance in Houston, Texas in 2002. By sequence analysis (Genbank accession nos. AY185907 and AY187013), strain 114 clusters with the dominant genetic variant isolates that have emerged in the United States.14 Mouse virulence studies have shown the virus to be similar to WNV NY99 in phenotype.15 The virus was propagated in Vero cells and cell culture supernatant was harvested and stored at −80°C until used in mosquito blood meals. Mosquitoes. Laboratory-reared, Sebring strain Cx. p. quinquefasciatus mosquitoes used for experiments were obtained from the Harris County Mosquito Control District in Houston, Texas and consisted of mosquitoes from F30 or later generations. Mosquitoes were maintained as previously described.16 Sebring strain mosquitoes are not significantly different from field-caught Cx. p. quinquefasciatus from Houston, Texas with respect to infection and dissemination rates of WNV strain 114 (Klingler KL and Higgs S, unpublished data). Mosquito infection. Female mosquitoes were fed on infectious and non-infectious blood meals 5–7 days post-eclosion and were deprived of sucrose for 24 hours prior to feeding. For infectious blood meal preparation, a Vero cell monolayer was inoculated with WNV at a multiplicity of infection of 2, and cells and supernatant were harvested after three days, which coincides with peak virus titer. Blood meals consisted of either WNV-infected Vero cells and supernatant, or uninfected Vero cells and supernatant mixed in a 1:1 ratio with defibrinated sheep blood (Colorado Serum Company, Denver, CO), along with 2 mM adenosine triphosphate, which was used as a phagostimulant. Mosquitoes were exposed to warm (37°C) blood meals using a Hemotek feeding apparatus (Discovery Workshops, Accrington, United Kingdom) for one hour, after which fully engorged mosquitoes were sorted and kept in cartons at 28°C and fed 10% sucrose ad libitum. All experiments including mosquito infections and dissections were performed in Biohazard Safety Level-3 laboratories and insectaries. Mosquito and blood meal titration. Infectious blood meals and three mosquitoes collected immediately after feeding on infectious blood (day 0) were titered in duplicate as serial 10-fold dilutions on Vero cells to determine the log10 50% tissue culture infectious dose (TCID50/mL or mosquito). Mosquitoes were individually triturated in 1 mL of Leibovitz L-15 medium (Mediatech, Inc., Herndon, VA, supplemented with 10% fetal bovine serum, 10% tryptose phosphate broth, 100 U/mL of penicillin, and 100 ␮g/mL of streptomycin) and filtered through a 0.22-␮m syringe filter prior to titration.

119

Virus titers were calculated as tissue culture infectious dose 50% endpoints (TCID50s) using the computer program ID50 version 5.0 (John L. Spouge; National Center for Biotechnology, National Institutes of Health, Bethesda, MD). Transmission electron microscopy. Salivary glands were dissected from both WNV-infected and uninfected control mosquitoes on days 14, 21, and 28 post-blood feeding. Approximately 10–15 mosquitoes were aspirated from cartons and chilled on ice prior to dissection. Salivary glands were removed from the thorax in a drop of phosphate-buffered saline (PBS, pH 7.4) on a clean glass slide and transferred immediately into TEM fixative. Salivary glands were stored at 4°C for no longer than one month in the fixative composed of 2.5% formaldehyde, 0.1% glutaraldehyde, 0.03% trinitrophenol, and 0.03% CaCl2 in 0.05 M cacodylate buffer (pH 7.3). Post-fixation was carried out in 1% OsO4 in 0.1M cacodylate buffer for one hour at room temperature, followed by en bloc staining in 1% uranyl acetate in 0.1 M maleate buffer for 20 minutes at 60°C. Mosquito salivary glands were dehydrated in an ethanol series and embedded in PolyBed 812 (Polysciences, Warrington, PA). Total sample sizes of 5–7 salivary glands were analyzed per time point per treatment group. Semi-thin (5 ␮m) sections were cut on a Reichert-Leica Ultracut S ultramicrotome (Leica Microsystems Inc., Bannockburn, IL) and stained with toluidine blue for observation by light microscopy. When dark-staining cells or morphologies suggestive of cytopathology were observed in 5-␮m sections, ultra-thin sections (70 nm) were cut and placed on 2–3 copper grids, stained with uranyl acetate and lead citrate, and examined in a Philips (Eindhoven, The Netherlands) CM-100 electron microscope at 60 kV. If dark-staining cells were absent by light microscopy, additional 5-␮m sections were prepared, stained, and examined. Depending on salivary gland orientation, up to 12 levels were analyzed by light microscopy and up to 8 levels were analyzed by TEM. A total of 20 morphologic characteristics relating to cytoplasmic membranes and cell viability were evaluated by TEM in both infected and uninfected groups at each collection time point. The frequencies of each characteristic were tabulated and statistically analyzed between time points and between treatment groups using Fischer’s exact test using SPSS for Windows version 11.5.0 (SPSS Inc., Chicago, IL). Saliva collection. In two experiments, saliva was collected from mosquitoes infected with WNV at 14 (n ⳱ 20), 21 (n ⳱ 20), and either 25 (n ⳱ 12) or 28 (n ⳱ 16) days post-infection (dpi) as previously described.16 In one of the experiments, saliva was collected at 7 dpi (n ⳱ 20). Quantitative reverse transcription–polymerase chain reaction (qRT-PCR). To thawed saliva samples, 105 ␮L of Leibovitz L-15 medium containing 1% fetal bovine serum (FBS) was added, and the sample was vortexed and centrifuged for one minute at 8,161 × g. A 50-␮L subsample was added directly to 300 ␮L of RLT buffer from the QIAamp viral RNA kit (Qiagen, Valencia, CA) and RNA was extracted within one hour following the manufacturer’s protocol. The total amount of RNA, expressed as the number of plaque-forming unit (PFU) equivalents of WNV in each saliva sample, was determined using a TaqMan One-Step RT-PCR (Applied Biosystems, Foster City, CA) as previously described.16 A standard curve was generated using three replicates of serial 10-fold dilutions of RNA extracted from virus stock with a known titer of 2 × 108 PFU/mL. Quantification of WNV PFU

120

GIRARD AND OTHERS

TABLE 1 Salivary gland morphologies in West Nile virus (WNV)–infected and uninfected salivary glands Collection day/treatment group 14

21

28

Morphologic characteristics

WNV

Control

WNV

Control

WNV

Control

Smooth membrane structures Convoluted membrane (CM) Tubular proliferated membrane (TPM) With parallel membrane structure TPM/CM combination With disorganized morphology Paracrystalline arrays of WNV in saliva Vacuolization Phagolysosomal vacuoles Dense irregular saliva Filamentous cytoplasmic membranes Saliva pool With apoptotic dead cells Dense striated membranes Electron transparency

6/6* 6/6* 6/6* 1/6 4/6* 0/6 5/7* 3/6 5/6 2/6 0/6 0/6 0/0 3/6 3/6

0/6 0/6 0/6 0/6 0/6 0/6 0/6 0/6 5/6 2/6 1/6 0/6 0/0 3/6 1/6

7/7* 7/7* 7/7* 3/7 4/7 0/7 6/7* 4/7 4/7 2/7 4/7† 4/7† 3/4 2/7 2/7

0/5 0/5 0/5 0/5 0/5 0/5 0/5 0/5 2/5 2/5 0/5 0/5 0/0 1/5 0/5

6/6* 6/6* 6/6* 4/6* 6/6* 4/6*†‡ 4/6* 5/6* 6/6 2/6 5/6*† 0/6‡ 0/0 5/6 4/6*

0/7 0/7 0/7 0/7 0/7 0/7 0.7 0/7 5/7 2/7 0/7 2/7 1/2 5/7 0/7

* Frequency of characteristic is significantly higher than the alternative treatment group from the same time point (P < 0.05). † Frequency of characteristic is significantly higher than day 14 (P < 0.05). ‡ Frequency of characteristic is significantly higher than day 21 (P < 0.05).

was added and plates were kept at −20°C for a minimum of two hours and a maximum of two days. Cell monolayers were stained for the presence of foci of viral antigen by IHC as previously described.17,18 Foci of infection comprising antigen-positive cells were counted in each well and multiplied by dilution factors to determine the number of focus-forming units, i.e., infectious virus particles, per saliva sample.

equivalents in saliva samples was calculated by comparison of the threshold cycle (Ct) values of the samples to the standards using Cepheid software (Cepheid, Sunnyvale, CA). Values are expressed as the average of two wells multiplied by a dilution factor of 2 to represent the whole saliva sample. Data was visualized and statistically analyzed by Student’s unpaired t-test using SigmaPlot version 9.0 (SPSS Inc.). Focus forming assay (FFA). The remaining 50-␮L saliva subsample was added to 250 ␮L of Leibovitz L-15 medium containing 1% FBS, vortexed, and further serially diluted sixfold. A volume of 125 ␮L of each dilution was added in duplicate to a Vero cell monolayer in a 24-well tissue culture plate (Corning Inc., Corning, NY) and was incubated at room temperature for one hour on a rocker. After incubation, an overlay containing 1% FBS in one part 2× L-15 medium mixed with one part 1.2% Tragacanth (Sigma-Adrich, St. Louis, MO) solution was added to each well. Plates were incubated at 37°C for three days, at which time the overlay was aspirated, and wells were washed twice with PBS and allowed to dry in the biosafety cabinet. To each well, approximately 500 ␮L of a 1:1 (v/v) mixture of cold acetone/methanol

RESULTS Salivary gland morphology during WNV infection. In the current study, 10–15 salivary gland pairs were dissected from WNV-infected and uninfected control mosquitoes 14, 21, and 28 days after blood feeding. The infectious blood meal contained 7.9 log10TCID50/mL of WNV, and the mean ± SD titer of three mosquitoes collected immediately after blood feeding was 5.6 ± 0.6 log10TCID50/mosquito. After dissection into PBS, half of a salivary gland pair was fixed and processed for TEM. Tissues were evaluated for morphologic characteristics including cell viability, proliferated membrane morphology, virus distribution, and saliva morphology (Tables 1 and 2).

TABLE 2 Features of apoptotic cell death in West Nile virus (WNV)–infected and uninfected salivary glands* Collection day/treatment group 14 Morphologic characteristic

Number of SG with apoptotic cells (nucleus observed)/ total number of SG analyzed (%) Apoptotic cells (all samples combined) Nucleus observed No nucleus observed Total Apoptotic cell location in salivary gland Number of SG with pre-apoptotic cells/total number of SG analyzed (%) Pre-apoptotic cells (all samples combined)

21

28

WNV

Control

WNV

Control

WNV

Control

1/6 (14)

2/6 (33)

5/7 (71)

1/5 (20)

5/6 (83)†

2/7 (29)

5 1 6 Central

2 0 2 Peripheral

36 7 43 Central

1 1 2 Peripheral

13 26 39 Both

6 4 10 Peripheral

2/6 (33) 2

0/6 (0) 0

4/7 (57) 8

0/5 (0) 0

4/6 (66)‡ 6

0/7 (0) 0

* SG ⳱ salivary glands. † Frequency of characteristic is significantly higher than day 14 (P < 0.05). ‡ Frequency of characteristic is significantly higher than the alternative treatment group, same time point (P < 0.05).

WNV CYTOPATHOLOGY IN MOSQUITOES

For each characteristic, the frequencies of occurrence were compared for WNV-infected and uninfected tissues within and between time points. At all time points, SMS, CM, and TPM were present in infected tissues but were not observed in uninfected salivary gland cells (Table 1 and Figure 1A and B). The combination CM/TPM structures were less frequent at 14 dpi (4 of 6) and 21 dpi (4 of 7) compared with 28 dpi (6 of 6). The level of CM size and distribution did not vary over time; however, the frequency of CM/TPM disorganization increased significantly

121

28 dpi (4 of 6) compared with 14 and 21 dpi, when disorganization was not observed. Membranes within disorganized CM/TPM appeared loose, less compact, and less regular in conformation. The frequency of row-like parallel membrane structures (PMS) associated with TPM increased over time post-infection (1 of 3 at 14 dpi, 3 of 7 at 21 dpi, and 4 of 6 at 28 dpi). Endoplasmic reticulum vacuolization, defined as marked distention of the ER luminal space, was only observed in infected tissues (Table 1 and Figure 1A and C). Vacuolization increased over time in WNV-infected salivary

FIGURE 1. A, West Nile virus (WNV)–associated proliferated membranes in a salivary gland cell, 21 days post-infection (dpi). TPM ⳱ tubular proliferated membrane; CM ⳱ convoluted membrane; SMS ⳱ smooth membrane structures; PMS ⳱ parallel membrane structure; V ⳱ vacuolization. Scale bar ⳱ 0.5 ␮m. B, Salivary gland, day 21 after feeding on a non-infectious blood meal. N ⳱ nucleus; Cy ⳱ cytoplasm; SC ⳱ saliva cavity. Scale bar ⳱ 1 ␮m. C, Vacuolization in a WNV-infected salivary gland lobe 28 dpi. Scale bar ⳱ 1 ␮m. D, Large saliva pool in uninfected salivary gland containing cellular contents (arrowhead) and nucleus, presumably of a dead cell, day 21. Scale bar ⳱ 1 ␮m.

122

GIRARD AND OTHERS

glands (3 of 6 at 14 dpi, 4 of 7 at 21 dpi, and 5 of 6 at 28 dpi). As shown in Table 1, electron transparency of the salivary gland cell cytoplasm was associated with virus infection and the frequency of this characteristic increased over time postinfection. In infected tissues, 3 of 6, 2 of 7, and 4 of 6 salivary glands showed electron transparency at 14, 21, and 28 dpi, respectively, compared with 1 of 6 control salivary glands that displayed electron transparency only at 14 dpi. Large accumulations of saliva (saliva pooling) occurred in the center of the lobe in both WNV-infected and uninfected salivary glands (Figure 1D). Cells surrounding the pool were flattened, suggesting that the acinar cells of the salivary gland had lost the normal concave shape that keeps extracellular saliva confined to the acinar cavity. In WNV-infected tissues, saliva pooling was only observed at 21 dpi (4 of 7) (Table 1). Saliva pooling was only observed in uninfected salivary glands on day 28 (2 of 7). Dead cell nuclei and cellular debris were observed in saliva pools in both infected (3 of 4) and uninfected (1 of 2) salivary glands (Table 1 and Figure 1D). The term DIS describes the appearance of granular, electron-dense, spherical bodies and membranous material (possibly cellular in origin) within the saliva (Figure 2A). Similar frequencies of DIS, DSM and PhV were found in both infected and uninfected tissues throughout the study (Table 1). The frequency of occurrence of FCM, which was characterized by thin, wavy membranes abundant within the cytoplasmic space of some cells, was analyzed (Figure 2B). Cells containing FCW were only identified in 1 of 6 uninfected salivary gland samples on day 14, but were seen at a high frequency in WNV-infected mosquitoes at later time points (4 of 7 at 21 dpi and 5 of 6 at 28 dpi) (Table 1). Paracrystalline arrays of WNV particles in saliva were observed in most WNV-infected salivary glands at all time points; 5 of 7 salivary glands at 14 dpi, 6 of 7 at 21 dpi, and 4 of 6 at 28 dpi. Arrays were sometimes extremely large (Figure

3A). Within salivary acinar cavities, WNV particles associated with an unknown saliva component that itself crystallized (Figure 3A). Paracrystalline arrays were commonly seen in salivary gland ducts as shown in Figure 3B. Frequency of cell death in WNV-infected and uninfected salivary glands. Morphologic features of apototic cells observed in this study included abnormally shaped, electrondense nuclei with chromatin marginalization. These cells were termed apoptotic-like by investigators in a previous report.8 Cells containing these features are heretofore termed apoptotic based on extensive morphologic examination, although biochemical analysis is needed to further confirm this assignment. In infected tissues, apoptotic cells were generally located centrally in the gland (Table 2) and had increased cytoplasmic electron-density and overall distorted cell shape. Shrunken and electron-dense mitochondria, SMS either free or within ER vesicle packets, and non–membrane-bound WNV particles were observed in infected, but not control, tissues (Figure 4A). In some instances, apoptotic cells showed evidence of membrane permeability, or disruption, which is suggestive of secondary necrosis. In these instances, cellular contents, including nuclei, were observed within the saliva cavity (Figure 4B). In uninfected salivary glands, apoptotic cells were generally located peripherally in the gland and were shrunken with an electron-dense cytoplasm, enlarged mitochondria, and an electron-dense and/or distorted nucleus with condensed chromatin (Figure 4C). Compared with apoptotic cells in WNV-infected mosquitoes, the nuclear envelope of dead cells in uninfected control salivary glands was typically well-defined. In WNV-infected mosquitoes, the percentage of salivary glands containing apoptotic cells increased over time from 14% to 71% and finally to 83% at 14, 21, and 28 dpi, respectively (Table 2). This frequency was significantly greater in tissues collected 28 dpi compared with 14 dpi (P < 0.05). In

FIGURE 2. A, Large saliva cavity filled with dense irregular bodies characteristic of dense irregular saliva, day 28 after a non-infectious blood meal. Arrowhead indicates possible cellular contents in the cavity. B, Filamentous cytoplasmic membranes (FCM) are found in two adjacent salivary gland cells, day 14 after a non-infectious blood meal. Mi ⳱ mitochondria. Scale bars ⳱ 1 ␮m.

WNV CYTOPATHOLOGY IN MOSQUITOES

123

FIGURE 3. A, Large paracrystalline array (PCA) of West Nile virus (WNV) particles in the saliva cavity of mosquito, 14 days post-infection (dpi). CM ⳱ convoluted membrane. B, PCA of WNV particles within the salivary gland duct (SD) of a mosquito salivary gland, 28 dpi. WNV particles (VP) are associated closely with an unidentified salivary component (arrowhead). Scale bars ⳱ 1 ␮m.

contrast, the frequency of apoptotic cells never exceeded 33% in uninfected salivary glands. The total number of apoptotic cells with visible nuclei in WNV-infected salivary glands increased from 5 at 14 dpi to 36 at 21 dpi, but decreased to 13 at 28 dpi. In contrast, in infected tissues, apoptotic cells lacking an observable nucleus increased in number from 1 at 14 dpi to 7 at 21 dpi and 26 at 28 dpi. When combined, the overall number of apoptotic cells (with and without a nucleus observed) in infected tissues plateaued between 21 and 28 dpi at approximately 40 cells. In tissues collected at 28 dpi, apoptotic cells were found both centrally and peripherally within the gland (Table 2). In uninfected salivary glands, the total number of apoptotic cells (nucleus and no nucleus observed) was low, but increased slightly over time: 2 on day 14 and day 21, and 10 (6 with nuclei observed) on day 28 (Table 2). Pre-apoptotic cells were defined as cells displaying increased cytoplasmic condensation or cell shrinkage, mild chromatin condensation or nuclear distortion, and/or cytopathology such as vacuolization or abnormal cell shape (Figure 4D). The frequency of pre-apoptotic cells increased over time in WNV-infected salivary glands, from 33% to 57% and 66% at 14, 21, and 28 dpi, respectively (Table 2), although the total number of pre-apoptotic cells in salivary glands was low (2 at 14 dpi, 8 at 21 dpi, and 6 at 28 dpi). Pre-apoptotic cells were not seen in uninfected tissues. Effect of long-term salivary gland infection on WNV transmission. As described above, salivary gland cells infected with WNV undergo progressive cytopathologic changes during long-term infection. Many of these changes increased in frequency over time suggesting they are associated with ongoing WNV replication. We hypothesized that these pathologic changes would impair virus maturation and saliva secretion (and therefore virus secretion). To test this, we collected saliva from WNV-infected mosquitoes and determined both vi-

rus titer and infectivity in saliva over time post-infection. Two groups of Cx. p. quinquefasciatus mosquitoes were orally infected during two separate experiments using blood meals containing 7.5 log10TCID50/mL of WNV. The mean ± SD titer of mosquitoes (n ⳱ 3) collected immediately after bloodfeeding was 5.6 ± 0.6 and 5.0 ± 0.5 log10TCID50/mosquito. Blood meal and day 0 values were not significantly different (P > 0.05). Each saliva sample was divided into two for testing by FFA to determine the amount of infectious virus per sample, and by qRT-PCR to determine the relative levels of WNV RNA per sample (expressed as PFU equivalents). Despite a clear trend of decreasing saliva virus titers over time in both assays, mean virus titers in saliva as determined by FFA and qRT-PCR were not significantly different between time points (P > 0.05) (Figure 5A and 5B). Mean ± SD WNV titers detected by FFA in mosquito saliva were 2.5 ± 1.4 × 103, 7.1 ± 3.2 × 103, 8.1 ± 3.6 × 103, and 1.9 ± 1.3 × 103 focus-forming units at 7, 14, 21, and 25–28 dpi, respectively. Mean WNV titers detected by qRT-PCR in mosquito saliva were 1.1 ± 1.0 × 104, 5.1 ± 1.8 × 103, 4.9 ± 1.6 × 103, and 1.1 ± 0.5 × 103 PFU equivalents at 7, 14, 21, and 25–28 dpi, respectively. When the ratio of focus-forming units to PFU equivalents was taken for each saliva sample, there was an overall decreasing trend in ratios from 14 to 25–28 dpi, although there was no significant difference in ratios over time (P > 0.05) (Figure 5C). The mean ± SD ratios were 7.1 ± 6.7, 2.0 ± 0.5, 0.9 ± 0.2, and 0.5 ± 0.2 on days 7, 14, 21, and 25–28, respectively. The frequency of WNV-positive saliva samples was determined at each time point of collection. We found that the number of WNV-positive saliva samples decreased significantly over time post-infection by both FFA and qRT-PCR methods (Figure 6). By qRT-PCR, WNV RNA was detected in 88% of saliva samples at 14 dpi, 70% at 21 dpi, and 64% at

124

GIRARD AND OTHERS

FIGURE 4. A, Apoptotic cells in West Nile virus (WNV)–infected salivary glands, 21 days post-infection (dpi). Two dead cells are surrounded by filamentous cytoplasmic membranes (FCM) and contain nuclei with chromatin condensation. Endoplasmic reticulum vesicle packets (ERVP) contain smooth membrane structure(s). B, Apoptotic cell contents within saliva cavity. N ⳱ nucleus; Mi ⳱ mitochondria. TPM ⳱ tubular proliferated membrane. C, Apoptotic cell in an uninfected salivary gland, day 21. D, Pre-apoptotic cell in WNV-infected salivary gland 21, dpi. CM ⳱ convoluted membrane; VP ⳱ WNV particles; SC ⳱ saliva cavity. Scale bars ⳱ 1 ␮m.

25–28 dpi (P < 0.05 compared with 14 dpi). By FFA, the frequency of WNV-positive saliva samples decreased significantly from 83% at 14 dpi to 53% (P < 0.05) at 21 dpi and 39% at 25–28 dpi (P < 0.05 compared with 14 dpi). Titration of WNV in mosquito bodies and salivary glands during long-term infection. Whole body titers were determined for 3 mosquitoes at 0 and 3 dpi and 10 mosquitoes at 7, 10, 14, 17, 21, 25, and 28 dpi after a blood meal containing 6.5 log10TCID50/mL of WNV. Mosquito body titers remained constant at approximately 6 log10TCID50/ mosquito between 10 dpi and 28 dpi, and 100% of mosquitoes were WNVpositive at each time point (Figure 7). A qRT-PCR assay was performed on salivary gland pairs dissected from 8 WNVinfected Cx. p. quinquefasciatus mosquitoes collected every other day between 1 and 21 dpi after a blood meal containing 7.5 log10TCID50/mL of WNV. Salivary gland titers ranged

between 1.5 × 106 and 6 × 106 PFU equivalents/mosquito from 11 dpi to 21 dpi, with no significant change over time (P > 0.05) (Figure 7). DISCUSSION This study details the morphologic and cytopathologic changes that occur in mosquito vector salivary glands during WNV infection, and evaluates the effects of long-term salivary gland infection on WNV transmission. Using age- and generation-matched uninfected control mosquitoes in our TEM analysis, we were able to distinguish age-related from infection-related changes in the salivary gland cell cytoplasm and nucleus during a four-week period after blood feeding. Our findings identified a correlation between long-term WNV infection and both increased frequency of salivary

WNV CYTOPATHOLOGY IN MOSQUITOES

125

FIGURE 6. Detection of West Nile virus (WNV) in mosquito saliva by focus-forming assay and quantitative reverse transcription– polymerase chain rection (qRT-PCR). Numbers above bars represent percentages. Number of mosquitoes tested per time point ⳱ 7 days post-infection (dpi), n ⳱ 20; 14 dpi, n ⳱ 40; 21 dpi, n ⳱ 40; 25–28 dpi, n ⳱ 28.

gland apoptosis and decreased WNV transmission. Mosquitoes in these studies were exposed to blood meal titers similar to, and sometimes lower than, those seen in laboratory infections of natural bird reservoirs.19 Future studies that analyze the frequency of salivary gland cytopathology in mosquitoes exposed to a range of blood meal titers, and performed in other important mosquito vectors such as Cx. tarsalis and Cx. p. pipiens, are necessary to better understand the importance of this phenomenon in nature. It is likely that both the dose of virus to which salivary glands are exposed and the susceptibility of the salivary gland cells to virus replication determines the outcome of infection including the timing, the mechanism, and the frequency of cell death. Since the early 1980s, light microscopy and TEM studies,8,20–22 feeding behavior studies,23,24 and survivorship and reproductive effort studies25,26 of arbovirus-vector interactions have challenged the notion that mosquitoes are unharmed by arbovirus infections. This report adds to the growing body of evidence that arboviruses cause pathologic changes in vector mosquitoes and, impor-

FIGURE 5. A, Quantification of West Nile virus (WNV) in saliva by focus forming assay (FFA). B, Relative levels of WNV plaqueforming unit (PFU) equivalents in individual saliva detected by quantitative reverse transcription–polymerase chain reaction. C, Ratio of focus-forming units to viral PFU equivalents of WNV in individual mosquito saliva samples. Dotted horizontal lines represent mean values. Solid horizontal lines within boxes represent median values. Shaded boxes contain values between the 25th and 75th quartiles. Dots represent outliers. Vertical bars represent 95% confidence intervals.

FIGURE 7. Whole body titration of West Nile virus (WNV)– infected Culex pipiens quinquefasciatus mosquitoes over a 28-day course of infection (●). qRT-PCR of individual salivary glands dissected over a 21-day course of infection (䉭). TCID50 ⳱ 50% tissue culture infectious dose; PFU ⳱ plaque-forming units. Error bars show the SD.

126

GIRARD AND OTHERS

tantly, indicates a need for more studies of transmission dynamics during long-term arbovirus infection in mosquitoes. Salivary gland analysis by TEM showed that proliferated membranes associated with WNV infection did not differ dramatically in abundance over time post-infection; however, there were detectable changes in the morphology of these membranes. Disorganization of CM/TPM structures increased over time, particularly after four weeks of infection. Row-like PMS within TPM also increased in frequency over time. It is possible that PMS result from the disruption or loosening of CM/TPM at late time points of infection. Further research is necessary to determine if the morphologic changes in CM/TPM correspond to a decrease in membrane usage for viral genome replication or proteolytic processing. Increasing electron transparency of the salivary gland cell cytoplasm and vacuolization of the ER observed at late time points during WNV infection are possibly because of the intense involvement of ER membranes and ribosomes in WNV replication. Increasing in frequency over time in WNV-infected tissues, FCM morphology and location suggest that these membranes represent degenerating elements of the ER and Golgi apparatus. Elucidation of the role of these membranes in virusassociated cell death and salivary gland cell turnover or senescence requires further study. Saliva pooling in our study was seen in WNV-infected salivary glands one week earlier than in uninfected tissues. In salivary glands from both infected and uninfected mosquitoes, pooling was associated with cell death and the appearance of cytoplasmic contents in saliva. Although saliva pooling appears to be a normal, age-related phenomenon, the process may be accelerated by WNV infection. Within saliva, paracrystalline arrays of WNV were abundant in frequency and distribution throughout the four-week course of analysis. These arrays were found within the acinar cavity, within the salivary gland duct, and within saliva pools. In contrast, arrays of WNV were not found within the cytoplasm of infected cells, suggesting that the phenomenon requires the particular milieu of mosquito saliva for their formation. Paracrystalline arrays of similar size, morphology, and location have been identified in studies of St. Louis encephalitis virus by Whitfield and others in salivary glands of Cx. p. pipiens.27 These investigators identified paracrystalline arrays at later time points than in our study, namely 25 and 32 dpi. Structural similarities between these two Japanese encephalitis serocomplex relatives28 may explain their common array formations. However, the earlier occurrence of paracrystalline arrays in WNV-infected Cx. p. quinquefasciatus may also reflect differences in the replication dynamics of these two viruses. Paracrystalline arrays of the flavivirus Japanese encephalitis virus (JEV) have also been identified in Cx. tritaeniorhynchus, although these arrays were small, not reportedly abundant, and occurred at 26 dpi in the saliva cavity.29 The significance of paracrystalline arrays in transmission has thus far been overlooked. These arrays, presumably maintained in saliva secretions during mosquito feeding, could increase the dose of virus delivered to a host and explain the wide range of titers detected in mosquito saliva in our study and other studies quantifying arboviruses in saliva.16,30,31 Future research into salivary or viral factors that contribute to the formation of paracrystalline arrays, as well as the significance of array formation in host infection, are likely to contribute to our understanding of arbovirus transmission and pathogenesis.

In mammalian tissues, apoptotic cell death is not uncommon during infections with mosquito-borne flaviviruses such as WNV, JEV, Langat virus, and dengue virus (DENV).32–41 In vitro experiments indicate that both apoptotic and necrotic cell death pathways occur during WNV infections, and these outcomes may be dose dependent.32,42–44 In our study, the number of mosquitoes with apoptotic cell death and the total number of dead cells per gland increased over time. In contrast, the percentage and total number of dead cells in uninfected mosquitoes remained relatively constant over a fourweek period. These data suggest that although age-related salivary gland cell death is a natural phenomenon, long-term WNV infection increases cell death. Our TEM analysis showed that dead cells in WNV-infected salivary glands contain morphologies suggestive of apoptosis similar to those identified in apoptotic mammalian cells infected with WNV.32,33 Comparable morphologies indicative of apoptosis include cell shrinkage, extensive vacuolization, chromatin condensation, nuclear blebbing and degeneration, and a loss of plasma membrane integrity.32,33 Interestingly, in our study, the number of dead salivary gland cells containing observable nuclei decreased between 21 and 28 dpi. It is possible that these nucleus-free figures represent apoptotic bodies. Apoptotic nuclei, virus-induced membranes, WNV particles, and other cellular contents were identified at late time points of infection within saliva cavities, which suggested that membrane disruption occurred in some instances after WNVinduced apoptosis. Loss of plasma membrane integrity and cell lysis in Vero cells infected with high doses of WNV was observed by Chu and Ng to correspond to a necrotic mechanism of cell death.32 We hypothesize that either secondary necrosis can occur after apoptosis during long-term salivary gland cell infection with WNV, or infected salivary gland cells may die by a recently described mechanism of cell death called aponecrosis.45,46 During studies of cytopathology associated with Sindbis virus (Togaviridae, Alphavirus, SINV), Karpf and Brown reported that cell death in a vertebrate cell line (baby hamster kidney cells) corresponded clearly to a large induction of apoptosis, but the virus caused only slight induction of apoptosis in cytopathic Aedes albopictus C7-10 cells.47 In addition, cytopathic C7-10 cells infected with SINV were reported to display morphologic characteristics of necrosis by TEM, which suggested to the investigators that both cell death mechanisms may explain the observed mosquito cell death. Molecular-based methods would further contribute to our understanding of the mechanism of salivary gland cell death caused by WNV. Previous IHC and TEM studies from our laboratory have shown that 100% of Cx. p. quinquefasciatus salivary gland cells are infected with WNV 14 dpi, i.e., approximately 6–9 days after initial salivary gland infection.7,8 As shown in our current study, cell death frequently occurred in salivary glands 21 and 28 dpi, equating to approximately 14 days of viral replication in the glands. This late timing of cell death may indicate that the mechanism of apoptosis is unrelated to expression of viral genes such as capsid and NS2B-NS3, as proposed by other investigators studying WNV-induced apoptosis in mammalian cells.33,35 We propose that apoptotic cell death in Culex salivary glands is due to a shift in the relative abundance of pro-apoptotic and anti-apoptotic proteins during continuous virus replication. Apoptosis may be a response to the physiologic disturbances caused by persis-

127

WNV CYTOPATHOLOGY IN MOSQUITOES

tently high levels of WNV replication and the decrease or cessation of host cell gene transcription and protein synthesis. Considering the extensive ER membrane proliferation and vacuolization that occurs in salivary gland cells during WNV replication, ER malfunctioning or stress may induce an unfolded protein response similar to that seen during JEV infection.40 Investigation into the role of viral gene products, host membranes, and host apoptotic and anti-apoptotic enzymes in salivary gland cell death during long-term WNV infection is warranted. Since effective saliva secretion is essential for WNV transmission by mosquitoes, we investigated the transmission of WNV during long-term infection of Cx. p. quinquefasciatus. Although WNV whole body and salivary gland titers did not decrease over time post-infection, fewer mosquitoes secreted saliva with detectable levels of WNV at 21 and 25–28 dpi compared with 14 dpi. One plausible explanation for the temporal decrease in the proportion of mosquitoes with WNVpositive saliva is that the normal mosquito feeding behaviors are impaired during WNV infection of the mosquito nervous system, as suggested by Platt and others studying DENV-3.23 In the laboratory, cerebral and suboesophagial ganglia of Cx. p. quinquefasciataus are infected with WNV between 8 and 10 dpi.7 These ganglia receive signals from sensilla found on cibarial and pharyngeal pumps, and may regulate the dilator muscles of these structures.48 Virus infection of these nervous tissues could disrupt pump and valve function, thereby affecting mosquito feeding and salivation. Virus transmission, virus tropism, and mosquito behavior studies are necessary to elucidate the role of nervous tissue infection in mosquito blood feeding and saliva output. Studies using TEM are particularly useful for detailed investigation of the underlying mechanisms of virus-vector interactions that determine vector competence. In this report, we successfully used TEM to understand morphologic aspects of the Culex mosquito cellular response to long-term WNV infection. Further investigation of Culex feeding behavior and WNV transmission to vertebrate hosts during long-term infection would greatly complement this work. Additionally, research on the inter-relationship between WNV-induced salivary gland cytopathology and saliva production and secretion will improve our understanding of WNV-vector and vector-host interactions, particularly as they relate to transmission dynamics in nature and the risk of human and avian WNV infection. Received June 29, 2006. Accepted for publication August 29, 2006. Acknowledgments: We thank Jing Huang for assistance in mosquito rearing and Dr. Robert B. Tesh for providing WNV-specific mouse hyperimmune ascites fluid. We also thank Dr. Jean Nichols and Dr. Douglas Watts for helpful discussions of this work. Financial support: This study was supported in part by National Institutes of Health grant R01AI47246. Yvette A. Girard, Bradley S. Schneider, and Charles E. McGee are supported by grant T01/ CCT622892 from the Centers for Disease Control and Prevention Fellowship Training Program in Vector Borne Infectious Diseases. Authors’ addresses: Yvette A. Girard, Bradley S. Schneider, Charles E. McGee, and Stephen Higgs, Department of Pathology, University of Texas Medical Branch, 2.104 Keiller, 301 University Boulevard, Galveston, TX 77555-0609, Telephone: 409-747-2426, Fax: 409-7472436, E-mails: [email protected], [email protected], [email protected], and [email protected]. Julie Wen, Violet C. Han, and Vsevolod Popov, Deaprtment of Pathology, University of

Texas Medical Branch, 2.102 Keiller, 301 University Boulevard, Galveston, TX 77555-0609, Telephone: 409-747-2423, Fax: 409-7472437, Emails: [email protected], [email protected], and [email protected]. Peter W. Mason, Department of Pathology, University of Texas Medical Branch, Sealy Center for Vaccine Research, 3.206B Mary Moody Northern Pavilion, 301 University Boulevard, Galveston, TX 77555-0436; Telephone: 409-747-8143, Fax: 409-7478150, E-mail: [email protected].

REFERENCES 1. Hayes EB, Sejvar JJ, Zaki SR, Lanciotti RS, Bode AV, Campbell GL, 2005. Virology, pathology, and clinical manifestations of West Nile virus disease. Emerg Infect Dis 11: 1174–1179. 2. Komar N, 2000. West Nile viral encephalitis. Rev Sci Tech 19: 166–176. 3. Komar N, 2003. West Nile virus: epidemiology and ecology in North America. Adv Virus Res 61: 185–234. 4. Centers for Disease Control and Prevention, 2006. CDC, DVBID West Nile Virus Homepage. Available from www.cdc.gov/ ncidod/dvbid/westnile. 5. Vinogradova EB, 2000. Culex pipiens pipiens Mosquitoes: Taxonomy, Distribution, Ecology, Physiology, Genetics, Applied Importance and Control. Sofia, Bulgaria: Pensoft Publishers, 4–13. 6. Hayes EB, Komar N, Nasci RS, Montgomery SP, O’Leary DR, Campbell GL, 2005. Epidemiology and transmission dynamics of West Nile virus disease. Emerg Infect Dis 11: 1167–1173. 7. Girard YA, Klingler KA, Higgs S, 2004. West Nile virus dissemination and tissue tropisms in orally infected Culex pipiens quinquefasciatus. Vector Borne Zoonotic Dis 4: 109–122. 8. Girard YA, Popov V, Wen J, Han V, Higgs S, 2005. Ultrastructural study of West Nile virus pathogenesis in Culex pipiens quinquefasciatus (Diptera: Culicidae). J Med Entomol 42: 429– 444. 9. Ebel GD, Carricaburu J, Young D, Bernard KA, Kramer LD, 2004. Genetic and phenotypic variation of West Nile virus in New York, 2000-2003. Am J Trop Med Hyg 71: 493–500. 10. Uchil PD, Satchidanandam V, 2003. Architecture of the flaviviral replication complex. Protease, nuclease, and detergents reveal encasement within double-layered membrane compartments. J Biol Chem 278: 24388–24398. 11. Westaway EG, Mackenzie JM, Kenney MT, Jones MK, Khromykh AA, 1997. Ultrastructure of Kunjin virus-infected cells: colocalization of NS1 and NS3 with double-stranded RNA, and of NS2B with NS3, in virus-induced membrane structures. J Virol 71: 6650–6661. 12. Mackenzie JM, Khromykh AA, Jones MK, Westaway EG, 1998. Subcellular localization and some biochemical properties of the flavivirus Kunjin nonstructural proteins NS2A and NS4A. Virology 245: 203–215. 13. Mackenzie JM, Jones MK, Young PR, 1996. Improved membrane preservation of flavivirus-infected cells with cryosectioning. J Virol Methods 56: 67–75. 14. Davis CT, Ebel GD, Lanciotti RS, Brault AC, Guzman H, Siirin M, Lambert A, Parsons RE, Beasley DW, Novak RJ, Elizondo-Quiroga D, Green EN, Young DS, Stark LM, Drebot MA, Artsob H, Tesh RB, Kramer LD, Barrett AD, 2005. Phylogenetic analysis of North American West Nile virus isolates, 2001–2004: evidence for the emergence of a dominant genotype. Virology 342: 252–265. 15. Beasley DW, Davis CT, Guzman H, Vanlandingham DL, Travassos da Rosa AP, Parsons RE, Higgs S, Tesh RB, Barrett AD, 2003. Limited evolution of West Nile virus has occurred during its southwesterly spread in the United States. Virology 309: 190–195. 16. Vanlandingham DL, Schneider BS, Klingler K, Fair J, Beasley D, Huang J, Hamilton P, Higgs S, 2004. Real-time reverse transcriptase-polymerase chain reaction quantification of West Nile virus transmitted by Culex pipiens quinquefasciatus. Am J Trop Med Hyg 71: 120–123. 17. Rossi SL, Zhao Q, O’Donnell VK, Mason PW, 2005. Adaptation of West Nile virus replicons to cells in culture and use of rep-

128

18.

19.

20. 21.

22. 23.

24.

25.

26.

27. 28.

29. 30. 31.

32. 33.

GIRARD AND OTHERS

licon-bearing cells to probe antiviral action. Virology 331: 457– 470. Scholle F, Girard YA, Zhao Q, Higgs S, Mason PW, 2004. Transpackaged West Nile virus-like particles: infectious properties in vitro and in infected mosquito vectors. J Virol 78: 11605– 11614. Komar N, Langevin S, Hinten S, Nemeth N, Edwards E, Hettler D, Davis B, Bowen R, Bunning M, 2003. Experimental infection of North American birds with the New York 1999 strain of West Nile virus. Emerg Infect Dis 9: 311–322. Weaver SC, Scott TW, Lorenz LH, Lerdthusnee K, Romoser WS, 1988. Togavirus-associated pathologic changes in the midgut of a natural mosquito vector. J Virol 62: 2083–2090. Weaver SC, Lorenz LH, Scott TW, 1992. Pathologic changes in the midgut of Culex tarsalis following infection with Western equine encephalomyelitis virus. Am J Trop Med Hyg 47: 691– 701. Bowers DF, Coleman CG, Brown DT, 2003. Sindbis virusassociated pathology in Aedes albopictus (Diptera: Culicidae). J Med Entomol 40: 698–705. Platt KB, Linthicum KJ, Myint KS, Innis BL, Lerdthusnee K, Vaughn DW, 1997. Impact of dengue virus infection on feeding behavior of Aedes aegypti. Am J Trop Med Hyg 57: 119– 125. Grimstad PR, Ross QE, Craig GB Jr, 1980. Aedes triseriatus (Diptera: Culicidae) and La Crosse virus. II. Modification of mosquito feeding behavior by virus infection. J Med Entomol 17: 1–7. Faran ME, Turell MJ, Romoser WS, Routier RG, Gibbs PH, Cannon TL, Bailey CL, 1987. Reduced survival of adult Culex pipiens infected with Rift Valley fever virus. Am J Trop Med Hyg 37: 403–409. Mahmood F, Reisen WK, Chiles RE, Fang Y, 2004. Western equine encephalomyelitis virus infection affects the life table characteristics of Culex tarsalis (Diptera: Culicidae). J Med Entomol 41: 982–986. Whitfield SG, Murphy FA, Sudia WD, 1973. St. Louis encephalitis virus: an ultrastructural study of infection in a mosquito vector. Virology 56: 70–87. Calisher CH, Karabatsos N, Dalrymple JM, Shope RE, Porterfield JS, Westaway EG, Brandt WE, 1989. Antigenic relationships between flaviviruses as determined by crossneutralization tests with polyclonal antisera. J Gen Virol 70: 37–43. Takahashi M, Suzuki K, 1979. Japanese encephalitis virus in mosquito salivary glands. Am J Trop Med Hyg 28: 122–135. Colton L, Biggerstaff BJ, Johnson A, Nasci RS, 2005. Quantification of West Nile virus in vector mosquito saliva. J Am Mosq Control Assoc 21: 49–53. Smith DR, Carrara AS, Aguilar PV, Weaver SC, 2005. Evaluation of methods to assess transmission potential of Venezuelan equine encephalitis virus by mosquitoes and estimation of mosquito saliva titers. Am J Trop Med Hyg 73: 33–39. Chu JJ, Ng ML, 2003. The mechanism of cell death during West Nile virus infection is dependent on initial infectious dose. J Gen Virol 84: 3305–3314. Ramanathan MP, Chambers JA, Pankhong P, Chattergoon M, Attatippaholkun W, Dang K, Shah N, Weiner DB, 2006. Host cell killing by the West Nile Virus NS2B-NS3 proteolytic complex: NS3 alone is sufficient to recruit caspase-8-based apoptotic pathway. Virology 345: 56–72.

34. Parquet MC, Kumatori A, Hasebe F, Morita K, Igarashi A, 2001. West Nile virus-induced bax-dependent apoptosis. FEBS Lett 500: 17–24. 35. Yang JS, Ramanathan MP, Muthumani K, Choo AY, Jin SH, Yu QC, Hwang DS, Choo DK, Lee MD, Dang K, Tang W, Kim JJ, Weiner DB, 2002. Induction of inflammation by West Nile virus capsid through the caspase-9 apoptotic pathway. Emerg Infect Dis 8: 1379–1384. 36. Couvelard A, Marianneau P, Bedel C, Drouet MT, Vachon F, Henin D, Deubel V, 1999. Report of a fatal case of dengue infection with hepatitis: demonstration of dengue antigens in hepatocytes and liver apoptosis. Hum Pathol 30: 1106–1110. 37. Despres P, Frenkiel MP, Ceccaldi PE, Duarte Dos SC, Deubel V, 1998. Apoptosis in the mouse central nervous system in response to infection with mouse-neurovirulent dengue viruses. J Virol 72: 823–829. 38. Liao CL, Lin YL, Wang JJ, Huang YL, Yeh CT, Ma SH, Chen LK, 1997. Effect of enforced expression of human bcl-2 on Japanese encephalitis virus-induced apoptosis in cultured cells. J Virol 71: 5963–5971. 39. Liao CL, Lin YL, Shen SC, Shen JY, Su HL, Huang YL, Ma SH, Sun YC, Chen KP, Chen LK, 1998. Antiapoptotic but not antiviral function of human bcl-2 assists establishment of Japanese encephalitis virus persistence in cultured cells. J Virol 72: 9844–9854. 40. Su HL, Liao CL, Lin YL, 2002. Japanese encephalitis virus infection initiates endoplasmic reticulum stress and an unfolded protein response. J Virol 76: 4162–4171. 41. Prikhod’ko GG, Prikhod’ko EA, Cohen JI, Pletnev AG, 2001. Infection with Langat flavivirus or expression of the envelope protein induces apoptotic cell death. Virology 286: 328–335. 42. Cantile C, Del Piero F, Di Guardo G, Arispici M, 2001. Pathologic and immunohistochemical findings in naturally occuring West Nile virus infection in horses. Vet Pathol 38: 414–421. 43. Guarner J, Shieh WJ, Hunter S, Paddock CD, Morken T, Campbell GL, Marfin AA, Zaki SR, 2004. Clinicopathologic study and laboratory diagnosis of 23 cases with West Nile virus encephalomyelitis. Hum Pathol 35: 983–990. 44. Xiao SY, Guzman H, Zhang H, Travassos da Rosa AP, Tesh RB, 2001. West Nile virus infection in the golden hamster (Mesocricetus auratus): a model for West Nile encephalitis. Emerg Infect Dis 7: 714–721. 45. Formigli L, Papucci L, Tani A, Schiavone N, Tempestini A, Orlandini GE, Capaccioli S, Orlandini SZ, 2000. Aponecrosis: morphological and biochemical exploration of a syncretic process of cell death sharing apoptosis and necrosis. J Cell Physiol 182: 41–49. 46. Papucci L, Formigli L, Schiavone N, Tani A, Donnini M, Lapucci A, Perna F, Tempestini A, Witort E, Morganti M, Nosi D, Orlandini GE, Zecchi OS, Capaccioli S, 2004. Apoptosis shifts to necrosis via intermediate types of cell death by a mechanism depending on c-myc and bcl-2 expression. Cell Tissue Res 316: 197–209. 47. Karpf AR, Brown DT, 1998. Comparison of Sindbis virusinduced pathology in mosquito and vertebrate cell cultures. Virology 240: 193–201. 48. Clemens AN, 2000. Adult food and feeding mechanisms. The Biology of Mosquitoes. Volume 1. New York: CABI Publishing, 220–250.