Biological characterization of Aedes albopictus (Diptera: Culicidae) in Argentina: implications for arbovirus transmission

Aedes albopictus (Diptera: Culicidae) is an invasive mosquito, native to Asia, that has expanded its range worldwide. It is considered to be a public health threat as it is a competent vector of viruses of medical importance, including dengue, chikungunya, and Zika. Despite its medical importance there is almost no information on biologically important traits of Ae. albopictus in Argentina. We studied life cycle traits, demographic parameters and analyzed the competence of this mosquito as a virus vector. In addition, we determined the prevalence of Wolbachia strains in Ae. albopictus as a first approach to investigate the potential role of this bacteria in modulating vector competence for arboviruses. We observed low hatch rates of eggs, which led to a negative growth rate. We found that Ae. albopictus individuals were infected with Wolbachia in the F1 but while standard superinfection with wAlbA and wAlbB types was found in 66.7% of the females, 16.7% of the females and 62.5% of the males were single-infected with the wAlbB strain. Finally, despite high levels of infection and dissemination, particularly for chikungunya virus, Ae. albopictus from subtropical Argentina were found to be relatively inefficient vectors for transmission of both chikungunya and dengue viruses.


Laboratory life cycle traits and demographic parameters.
A total of 49 adults were obtained (29 males and 20 females) from the 50 F1 larval specimens obtained in the field. Thus, the survival rate from first instar larvae to pupae was 98%. The sex ratio was 0.68 [♀/♂]. Fecundity was 30 eggs per fed female after the first blood feeding and 82 after the second feed. 10% of eggs laid hatched. The average longevity for adults was 16 days for females and 10 days for males. The demographic parameters were as follows: Net Reproductive Rate (R 0 ) was 0.512, Cohort Generation Time (Tg) was 33 days and the Intrinsic Rate of Population Growth (r) was −0.0197.

Wolbachia infection status in adult mosquitoes.
Two out of 20 females of Ae. albopictus were discarded because DNA extraction failed. Therefore, a total of 18 females were analyzed. Overall, 83.3% of females were Wolbachia positive. Of these, 66.7% had double infection with the strains wAlbA and wAlbB, while 16.7% were infected only with wAlbB. From the eight males assayed, six tested positive for Wolbachia, indicating a 75% prevalence. We found that 62.5% had a single infection with the wAlbB strain, while 12.5% had a double infection.
Comparison of the sequences obtained for each locus of the full MLST (by alphabetical order) revealed the following allelic profile for wAlbB: (229-27-210-242-166). Also, sequencing of the wsp gene, which provides complementary information to the standard MLST system, indicated the following genotype for the four hypervariable regions of the gene (HVR1-4): 10-82-10-84. This information is reported for the first time for the wAlbB strain. As expected, wAlbB strain grouped with other strains from the B supergroup (Fig. 2). As it was seen before for other mosquitoes, this strain is not closely related to other Culicidae strains from the same supergroup.
For CHIKV, 69.8% of females (N = 146 fed females) were infected after feeding on blood with 1.7 × 10 7 PFU/ ml. Dissemination of CHIKV began day 5 (40.9%) and reached a peak of 57.8% (N = 102) in infected mosquitoes. Detection in saliva started in females on 10 dpi with a total percentage of 6.2% of individuals with disseminated infections transmitted. The extrinsic incubation period, calculated as the mean day of transmission detection, was 16.5 days for CHIKV, as compared to 21 days for DENV ( Table 1).
The proportion of CHIKV-infected mosquitoes was significantly higher than DENV infection (p-value = 0.005, α = 0.05). Likewise, the proportion of mosquitoes with disseminated virus was significantly higher for CHIKV than DENV (p-value 0.01, α = 0.05). However, there was not a significant difference in the proportions of mosquitoes transmitting the two viruses (p-value = 0.25, α = 0.05).

Discussion
Aedes albopictus has spread from its native range to at least 28 other countries around the globe 1 . Several studies demonstrated that variants of Ae. albopictus from the Northern and Southern hemispheres show a distinct dispersal behavior 33 . However, the geographic expansion of Ae. albopictus across Brazil after the first detection in 1986 from tropical Asia 2 to at least 20 of the 27 Brazilian states 19 , is quite different from what occurs in Argentina. The current distribution of this species in Argentina is limited to four locations in the province of Misiones [21][22][23] . In this study we did not detect the presence of Ae. albopictus in other locations located along the National Route 12, which connects Posada to Puerto Iguazú. But our search was limited to used car tires, thus to confidently rule out other breeding sites it would be necessary to sample in other types of containers, for instance, those belonging to the natural habitat of Aedes albopictus. Field samples of Ae. albopictus were less abundant than other co-existing mosquito species, in agreement with the results of Schweigmann et al. 22 in a nearby area (Eldorado, Misiones province). Furthermore, we obtained differences in some population parameters as compared with other studied populations of Ae. albopictus from America and Asia [34][35][36] . For instance, although larval and pupal survival was high, we found a low hatch rate of eggs, leading to a negative growth rate. Higher rates of egg hatching were recorded in other populations of Ae. albopictus strains as is New Orleans (57%) and Houston (93%) 36 . One explanation for the decreased hatch rates and immature abundance of Ae. albopictus populations from Misiones province is Wolbachia infection status, wich may induce cytoplasmic incompatibility (CI), due to the bacteria is not fixed in this population 37 .
The studied population was polymorphic for Wolbachia infection, containing uninfected individuals, individuals superinfected with both wAlbA and wAlbB strains, or infected only with wAlbB. We isolated the strain from the wAlbB infected individuals to characterize it for the first time by full MLST, which is useful for comparative purposes. In describing the diversity of Wolbachia in this host species, Zhou et al. 38 found the A and B strains using a wsp-based phylogeny. However, Casiraghi et al. and Baldo et al. 39,40 reported the mosaic nature of this gene, which may yield misleading results. Previously, Baldo et al. 41 proved that wAlbA belongs to the A supergroup by full MLST and here we also used it for wAlbB, leading us to confirm its identity as a member of the B supergroup.
Although we recovered the same Wolbachia strains already reported for Ae. albopictus, prevalence in the population from Iguazú National Park was 80.8%, while the prevalence reported for most populations worldwide is 100% [42][43][44] . The latter is generally expected since this bacteria tends to rapidly spread to fixation after it invades a population 45 , yet our result is in agreement with the study of Turelli 46 , who concluded that a CI strain showing no fertility cost to infected females, complete phenotype expression and a transmission rate of 80% will reach a polymorphic equilibrium of 0.72, causing a 20% reduction in population fertility. Experimental crossing assays between superinfected males and wAlbB infected females will be conducted in order to measure the levels of penetrance of CI in this population and thus ascertain if the low hatch rate observed in the present study is ascribable to this phenotype induced by Wolbachia.
Single wAlbB infection in males of Iguazú National Park population can be explained by depletion of wAlbA infection during the life cycle 47 , while the occurrence of negative individuals is more difficult to interpret. Bacteria could have been undetected because of low loads or by absence of infection and further investigation using real time PCR would be helpful to obtain more accurate results.
We also studied the vector competence of this mosquito population to determine its potential importance in DENV and CHIKV transmission in the northeast area of Argentina. Our results indicated Ae. albopictus from subtropical Argentina are competent but relatively inefficient vectors for both CHIKV and DENV. DENV transmission, in particular, was low and delayed, with extrinsic incubation period (EIP) of 21 days identified in a single transmitting individual. The extended EIP and low transmission rates of DENV in Ae. albopictus from Iguazú National Park indicate that this population is unlikely to be a significant contributor to DENV activity in the region. Despite both a shorter EIP and higher rate of dissemination for CHIKV, transmission was still relatively inefficient. At the population level, there was no significant difference between transmission of CHIKV and DENV looking at all mosquitoes exposed to an infectious blood meal. These results are similar to that found by Vega-Rúa et al. 7 and suggest there is either a significant salivary gland infection or transmission barrier preventing highly efficient transmission by this population of Ae. albopictus.
Mousson et al. 31 showed that Wolbachia naturally infecting Ae. albopictus from La Reunion Island limits DENV-2 dissemination and the infection of salivary glands, since removing Wolbachia prevented this inhibition. Our data suggest potential different levels of interference with arboviruses by this endosymbiont. Aedes aegypti is considered the main epidemic vector of DENV in the Americas, while Ae. albopictus is regarded as a secondary vector. Overall, our results with the Argentina population of Ae. albopictus are consistent with this characterization and indicate that, in the absence of both population expansion and viral adaptation, Ae. albopictus are likely to remain a minor contributor to arbovirus transmission in the region.

Methods
Mosquitoes and environmental data collection. A survey of the Ae. albopictus that breeds in containers was performed in order to ascertain their current distribution. The collection sites were located all along the National Route 12, the major roadway that connects Posadas to Puerto Iguazú, both located in the Misiones province, in Subtropical Argentina, in the northern part of the country (Fig. 1). All collections were taken from various artificial containers of standing water in the area using a siphon bottle. For Iguazú National Park, which belongs to Parques Nacionales, Argentina, a NEA 326 permit was obtained. Morphological identification of the specimens (3 rd and 4 th instar larvae) was performed using dichotomous keys 48 , while 1 st and 2 nd instars and pupae were reared either to the 4 th instar or to adult emergence, respectively. Immature stages were transported to the lab forfurther processing. Daily temperature and humidity were recorded by hour using a HOBO data logger during February in Iguazú National Park, where we were able to collect a sample of Ae. albopictus. These data were used to build an average curve of maximum to minimum daily cycle of temperature from this location.

Life table and demographic parameters.
We examined life-history traits of Ae. albopictus using the first generation (F1). F1 eggs (N = 500) were hatched and placed in an incubator with a temperature cycle fluctuating between 21 °C and 34 °C and a photoperiod 14:10 (L:D), simulating the conditions recorded in the field. Fifty first instar larvae were separated in 1 L of dechlorinated water in a plastic flat tray (30 × 18 × 6 cm) with finely ground guinea pig food. Instar stage and mortality were recorded daily, as well as the day of pupation. The pupae were removed to plastic containers (8 × 3.5 cm diameter), and provided with water and raisins in preparation for emergence. Following emergence, adult were sexed and put in a cardboard cage (25 × 22 cm diameter) for 3 to 5 days to allow mating before blood feeding. After having fed on a hamster for 45-60 min, the adults were moved to separate plastic containers (8 × 3.5 cm diameter, 1 per container) with filter paper and wet cotton on the bottom for oviposition (OVI 1). Seven days after the first feeding, adult females were placed together in a cardboard cage for a second blood feeding and then were placed once again in individual containers to oviposit. Adult mortality was checked daily and dead adults were stored at −20 °C. Eggs laid by each blood-fed female were counted, transferred to a Petri dish on filter paper with cotton, and sealed using parafilm to maintain humidity for 7-10 days. Eggs were then recounted to verify numbers and condition, and put in a plastic container with 250 ml of dechlorinated water and 10 mg of yeast for hatching. This procedure was carried out twice and the eggs that did not hatch were not examined to verify embryogenesis. The larvae were counted 48 hours later. Survival was expressed as the percentage of individuals that reached the next instar/stage. Larval and pupal mean development time and sex ratio of emerged adults were also measured. Fecundity was determined as the total number of eggs laid per mosquito (total number of engorged females) and fertility as the proportion of eggs hatched over the total number of eggs per population. Daily mortality records were used to calculate survival from the first day at the first instar as a function of age (lx); the number of eggs laid daily was used to calculate the age-specific fecundity (mx), by dividing total number of eggs laid each day (x) by the number of individuals alive at the end of that day. The lx and mx schedules allowed the estimation of demographic parameters such as the intrinsic rate of natural increase (r), the net reproductive rate (Ro), and the mean generation time (Tg) 49 .

Detection of Wolbachia.
We extracted DNA from adult mosquitoes (20 females and 8 males) stored at −20 °C from life table study using the Wizard Genomic Purification Kit (Promega). Multiplex PCR was carried out using the temperature profile of 95 °C for 1 min, 55 °C for 1.5 min and 72 °C for 2 min for 35 cycles and wsp primers. Primers used were 328 F and 691 R for wAlbA strain and 183 F and 691 R for wAlbB strain, as described by Zhou et al. 38 . When only one strain was detected, and due to the possible competition between both wAlbA and wAlbB DNAs in the multiplex PCR, we conducted independent PCRs to detect either wAlbA or wAlbB in such individuals. Negative results from independent PCRs were repeated twice. The quality of DNA extraction was checked using a primer set that amplifies the mosquito cytochrome oxidase 1 mtDNA locus (COI) under the conditions specified by Rodriguero et al. 50 . Any sample that yielded a negative result for COI was excluded from the data set. Samples that were negative for wsp primers but positive for COI primers were scored as uninfected. All negative results for Wolbachia infection were checked twice with fbpA and coxA primers 40 . Sequencing of Wolbachia. Strain wAlbB was characterized by means of full MLST through amplification and sequencing of the cytochrome oxidase subunit I (coxA), fructose-bisphosphate aldolase (fbpA), cell division protein ftsZ (ftsZ), aspartyl/glutamyl-tRNA amidotransferase subunit B (gatB) and conserved hypothetical protein (hcpA) fragments using the primers and conditions described in Baldo et al. 40 . MLST profile of wAlbA strain can be seen in Baldo et al. 40 . In addition, we provided the wsp sequence, in order to further characterize this strain on the basis of the amino acid motifs of the four hypervariable regions (HVRs) of this sequence.
DNA was purified with a DNA Puriprep-GP (Inbio Highway). DNA fragments were sequenced using a sequencer ABI3730 XL (Macrogen Inc., Korea). Chromatograms were edited using BIOEDIT 51 . Allele number was given to every gene after comparison with the Wolbachia MLST database (http://pubmlst.org/wolbachia). Thus, wAlbB strain was characterized by the combination of the MLST numbers (allelic profile or ST).
We accomplished a phylogenetic analysis including strains from supergroups A, B, C, D, F and H. The coxA, fbpA, ftsZ, gatB and hcpA gene sequences were concatenated and aligned using the CLUSTALW algorithm 52 . The terminal units were 52 Wolbachia strains retrieved from the Wolbachia MLST database and the one obtained in the present work. Our complete data set includes 2079 aligned nucleotide positions. jModelTest 2 53 was used to infer the most appropriate model of molecular evolution. The GTR + G model was selected as the best fit model of nucleotide substitution for the coxA, ftsZ, gatB and hcpA partitions and the GTR + G + I was the best fit for the fbpA partition.
Bayesian phylogenetic analysis of the concatenated MLST sequences was applied through the 'metropolis-coupled Markov chain Monte Carlo' (MC3) algorithm implemented in MrBayes v. 3.2.6 54 using a partitioned algorithm. Two independent analyses were run with a random starting tree over 2,500.000 generations with a sample frequency of 500. The tree space was explored using four chains: one cold and three incrementally heated chains, with temperature (T) set to 0.20. The first 500 trees were discarded as burn-in. We assessed stationarity of the cold Markov chain for all MrBayes analyses in TRACER 55 , in addition to the standard deviation of the split frequencies. All posterior samples of a run prior to the burn-inpoint were discarded. Remaining trees were taken into account to obtain a 50% majority-rule consensus tree and mean branch length estimates. The frequency of all bipartitions was estimated to assess the support of each node 56 .
SCiEnTiFiC REPORTS | (2018) 8:5041 | DOI:10.1038/s41598-018-23401-7 As the root for the overall tree of the Wolbachia genus is still undetermined 39,57 , we did not include any outgroup. However, for analyzing horizontal transfer, it is necessary to examine sister group relationships. Thus, we arbitrarily rooted the tree with supergroup H.

Vector competence assays. Peroral infection.
These studies were conducted in the BSL-3 insectary at the Arbovirus Laboratory, Wadsworth Center, New York State Department of Health, Albany, New York, USA using F1 eggs from obtained from the Iguazú National Park population. Vector competence assays were conducted using DENV2 strain 306 from Nicaragua (GenBank no. SAMN011003699) and CHIK strain 91077 (GenBank no. EF451145). Blood meals containing virus were prepared using freshly harvested virus following 4 (CHIKV) or 6 (DENV) days of growth on Ae. albopictus mosquito cells (C6/36, ATCC). Specifically, 8.5 mL of defibrinated bovine blood and 0.5 mL of 50.0% (w/v) sucrose were used for blood meal preparation, infected cells were scraped and the media with cell suspension was mixed 1:1 with the bloodmeal.
Five to seven day old female adult mosquitoes were allowed to feed on the infected blood meal in sausage casing for 1 hour at room temperature following the protocol of Ebel et al. 58 . Blood suspensions were frozen at −80 °C for subsequent plaque assay to determine the virus titer at the time of mosquito feeding. Fully engorged females were separated and held in 0.5 L cartons in incubators with temperature cycles fluctuating between 21 °C and 34 °C and a 12:12 (light:dark) photoperiod, and provided a 10% sucrose solution ad lib on cotton wicks until mosquitoes were harvested. At 5, 10, 14 and 21 days post-feeding females were anesthetized with Triethylamine (Sigma, MO), and bodies, legs, and salivary secretions were obtained from each individual mosquito and processed as previously described to determine infection, dissemination, and transmission rates, respectively 58 .
The number of mosquitoes with virus in legs (dissemination) and salivary secretion (transmission) was calculated as a proportion of the number of mosquitoes that were infected. Rates were compared between groups using Chi square (χ 2 ) or Fisher's exact tests, as appropriate according to the sample sizes. All analyses were performed using R.