The impact of temperature and Wolbachia infection on vector competence of potential dengue vectors Aedes aegypti and Aedes albopictus in the transmission of dengue virus serotype 1 in southern Taiwan

We evaluated the impact of temperature and Wolbachia infection on vector competence of the local Aedes aegypti and Ae. albopictus populations of southern Taiwan in the laboratory. After oral infection with dengue serotype 1 virus (DENV-1), female mosquitoes were incubated at temperatures of 10, 16, 22, 28 and 34 °C. Subsequently, salivary gland, head, and thorax-abdomen samples were analyzed for their virus titer at 0, 5, 10, 15, 20, 25 and 30 days post-infection (dpi) by real-time RT-PCR. The results showed that Ae. aegypti survived significantly longer and that dengue viral genome levels in the thorax-abdomen (103.25 ± 0.53–104.09 ± 0.71 PFU equivalents/ml) and salivary gland samples (102.67 ± 0.33–103.89 ± 0.58 PFU equivalents/ml) were significantly higher at high temperature (28–34 °C). The survival of Ae. albopictus was significantly better at 16 or 28 °C, but the virus titers from thorax-abdomen (100.70–102.39 ± 1.31 PFU equivalents/ml) and salivary gland samples (100.12 ± 0.05–101.51 ± 0.31 PFU equivalents/ml) were significantly higher at 22–28 °C. Within viable temperature ranges, the viruses were detectable after 10 dpi in salivary glands and head tissues in Ae. aegypti and after 5–10 dpi in Ae. albopictus. Vector competence was measured in Ae. albopictus with and without Wolbachia at 28 °C. Wolbachia-infected mosquitoes survived significantly better and carried lower virus titers than Wolbachia-free mosquitoes. Wolbachia coinfections (92.8–97.2%) with wAlbA and wAlbB strains were commonly found in a wild population of Ae. albopictus. In southern Taiwan, Ae. aegypti is the main vector of dengue and Ae. albopictus has a non-significant role in the transmission of dengue virus due to the high prevalence of Wolbachia infection in the local mosquito population of southern Taiwan.


Background
Dengue fever is the most common arboviral disease in tropical and subtropical regions in the world. In recent decades, the numbers of dengue cases and countries with endemic dengue fever have dramatically increased because of global warming [1], rapid and frequent international traveling, ineffective vector control, changing life styles and unplanned urbanization [2]. The disease burden of dengue is estimated at 390 million dengue infections per year, with 96 million cases exhibiting apparent symptoms [3]. Moreover, a total of 3.9 billion people living in 128 countries have been identified as at risk for dengue virus infection [4]. In Taiwan, since 2004, outbreaks have occurred annually with peaks in 2014 and 2015 (15,492 and 43,419 indigenous cases, respectively). These local mosquito-transmitted cases were mainly found in southern Taiwan where Aedes aegypti L. and Ae. albopictus (Skuse) coexist. The former species has limited distributions in southern Taiwan, Taitung City in eastern Taiwan and the Penghu Islands. The latter species is found in islands below the altitudes of 1760 m above sea level [5].
Dengue outbreaks have occurred in late spring or early summer and have peaked in the fall in many countries, including Taiwan. Previous studies have found that temperature is the most important abiotic environmental factor because it affects the vector efficiency [6][7][8]. Viral replication and propagation speed are faster at higher temperatures. The extrinsic incubation period (EIP), the time from the initial acquisition of the dengue virus by mosquito vectors to the time of transmission to a human host ranges from 5 to 33 days at 25°C and from 2 to 15 days at 30°C, with an average of 15 and 6.5 days, respectively [9]. Additionally, the mosquito species and strain [10], virus serotypes and genotypes [11], virus uptake quantity [6], size of adult mosquitoes [12], and Wolbachia infection [13][14][15] also affect dengue viral replication or limit viral transmission. Among these factors, the effect of Wolbachia is worth paying attention to due to the disease control strategy. Although the mechanism of Wolbachia-mediated pathogen interference is unclear, Wolbachia-induced, density-dependent inhibition of dengue virus in Aedes mosquitoes is temperature sensitive [16]. Therefore, some studies showed no Wolbachia effect on chikungunya virus infection [17] and dengue virus [18] in Ae. albopictus mosquitoes, but other studies showed that Wolbachia infection reduced dengue viral transmission but not the viral load [14].
Dengue is not considered an endemic disease in Taiwan. Most dengue outbreaks have originated from imported dengue cases, in which the introduced virus was transmitted to local Aedes mosquitoes in the late spring or the early summer, except for four winter outbreaks in Taiwan [19]. Furthermore, an ovitrap survey from 2010 to 2011 found that Ae. aegypti mosquitoes were able to oviposit all year long in Kaohsiung City [20]. The magnitude of dengue outbreaks can reach hundreds or thousands of people in the distribution areas of Ae. aegypti. However, only clusters (less than 20 dengue cases) of DENV-1 occurred in areas without the presence of Ae. aegypti, where Ae. albopictus mosquitoes served as the transmitting vectors instead. In Taiwan, Wolbachia infection with wAlbA and wAlbB is very common in the local Ae. albopictus population but not the Ae. aegypti population [21]. Therefore, the objective of this study was to compare the vector competence of local Ae. aegypti and Ae. albopictus populations in high-risk dengue areas in southern Taiwan for DENV-1 at different incubation temperatures in the laboratory. In addition, the natural effect of Wolbachia infection on the vector competence of the local Ae. albopictus population was also evaluated.

Mosquito collection and maintenance
Aedes aegypti and Ae. albopictus larvae and pupae were collected from a dengue high-risk area, the Chien-chen district, Kaohsiung City, in 2015 and 2016, respectively. Most collections occurred in the public domain and for any collections referred to the private sector, owners/residents were asked for permission to collect mosquitoes on their land/in their residences. The F1 adults used for the various temperature experiments, for each species, were from a single collection of larvae reared under the same conditions [20-30°C with a photoperiod of 10:14 h (L:D)]. Immature mosquitoes were reared in a plastic pan (21 × 14 × 7 cm) containing 450 ml of deionized water. A sufficient amount of food (yeast extract and pig liver powder; 1:1 by weight) was provided daily. Adult mosquitoes were kept in an acrylic cage (29 × 20 × 20 cm) and were provided with a 10% sucrose solution. The first generation (F1) of these field populations was used in the following experiments. Additionally, to eliminate Wolbachia infection, adult F1 Ae. albopictus mosquitoes were fed a 1 mg/ ml tetracycline solution in sucrose as described previously [22]. Each generation of mosquitoes was fed with tetracycline for at least two weeks and given a blood meal, and then, eggs hatched to produce the next generation (F1 to F3). When the third generation (F3) was shown to have no infection of Wolbachia, the fourth generation (F4) was used as the Wolbachia-free mosquitoes for further experiments. To minimize the impact of antibiotic on mosquito microflora, the eighth generation (F8) was also used as the Wolbachia-free mosquitoes for giving a recovery of 2 additional generations (tetracycline using in F1-F5).

Experimental oral infection
Two-to five-day-old female mosquitoes were deprived of sugar solution for 24 h prior to the oral challenge. The feeding mixture was prepared by mixing equal parts of DENV-1-infected C6/36 supernatant and human blood treated with an anticoagulant (7 ml of human blood:18 mg of K2EDTA) (cat. no. 367525, Becton Dickinson and Company, New Jersey, USA). The virus strains (H1030440) used in this study was the dominant strain of 2014 [19], which was isolated from blood specimens of the indigenous cases. All were harvested at virus titer of 10 7 plaque forming units (PFU) per ml (determined by plaque assay). A high virus titer of 1.63 × 10 7 PFU/ml was used for the following experiments, and this titer was close to the high viral load (ranging from 1.9 × 10 6 to 4.7 × 10 7 PFU/ml) of viremic patients [23]. Ten females were placed in a small paper cup (8 cm diameter × 9.5 cm height) covered with fine nylon mesh. The Hemotek 5 W1 membrane feeding system (Hemotek Ltd., Lancashire, UK) was used and Ae. aegypti and Ae. albopictus mosquitoes were allowed to feed for 1 h. The fed mosquitoes were then held in a growth chamber at 10, 16, 22, 28 or 34°C, with 75% relative humidity (RH). Cotton soaked in a 10% sugar solution was provided on the mesh and changed every 2 days. Mosquitoes were frozen at 0, 5, 10, 15, 20, 25, and 30 day intervals as knockouts. Then, the heads, salivary glands and thorax-abdomens were dissected for samples and stored at −80°C until processing for virus or Wolbachia detection.

Viral detection and titration
Individual samples (salivary glands, head, thoraxabdomen, midgut or fat-body) were homogenized and purified by centrifugation. Viral RNA (70 μl) was extracted from 140 μl of sample suspension using the QIAamp Viral RNA Mini Kit (cat. no. 52,906, Qiagen, Hilden, Germany) according to the manufacturer's instructions. Amplification by RT-PCR was performed using the LightCycler quantitative PCR system (Roche Applied Science, California, USA). Samples were assayed in a 50 μl reaction mixture containing 10 μl of sample RNA and optimal concentrations of the primers using the QuantiTect SYBR Green RT-PCR Kit (cat. no. 204, 243, Qiagen, Hilden, Germany). Type I dengue virusspecific primers DN-F (5′-CAA TAT GCT GAA ACG CGA GAG AAA-3′) and D1-R (5′-CGC TCC ATA CAT CTT GAA TGA G-3′) [24], were used for realtime RT-PCR, and the products were expected to be 193 bp. The thermal profile consisted of a 30 min reverse transcription step at 50°C, followed by 15 min of Taq polymerase activation at 95°C; these steps were then followed by 45 cycles of PCR (94°C for 15 s, annealing temperature 55°C for 30 s, and 72°C for 20 s).
Additionally, serial ten-fold dilutions of dengue virus (strain H1030440) were performed with an initial viral load of 1.63 × 10 7 PFU/ml. Each dilution was added to 140 μl of the salivary glands, heads, thorax-abdomens, midguts and fat-body sample suspension of one noninfected Ae. aegypti or Ae. albopictus female mosquito to estimate the virus titers in the salivary glands, heads, thorax-abdomens, midguts and fat-body of orally infected mosquitoes. The virus titer standard curve was generated by real-time RT-PCR. The linear regressions of the Ct value (Y) against log (viral load, Z) in the salivary glands, heads, thorax-abdomens, midgut and fat-body of

Wolbachia detection
The first generation (F1) of Ae. albopictus adults was used to detect Wolbachia infection. DNA was extracted from the mosquitoes used in the previous experiments by the QIAamp® DNA Mini Kit (cat. no. 51306, Qiagen, Hilden, Germany). The PCR reaction was conducted as described by previous studies [25]. Two primer pairs (328F, 5′-CCA GCA GAT ACT ATT GCG-3′, and 691R, 5′-AAA AAT TAA ACG CTA CTC CA-3′; 183F, 5′-AAG GAA CCG AAG TTC ATG-3′ and 691R) were used to amplify and detect Wolbachia surface protein genes (wsp), group A wAlbA (501 bp) and group B wAlbB (379 bp) in this study. Universal primers (12SRNA-Forward, 5′-AAA CTA GCA TTA GAT ACC CTA TTA T-3′ and 12SRNA-Reverse, 5′-AAG AGC GAC GGG CGA TGT GT-3′) were used to amplify a cDNA fragment of the insect mtDNA (12SRNA gene), as a control to assess the quality of the template DNA extracted from mosquitoes.

Quantification of Wolbachia density
The density of Wolbachia in the first generation (F1) of Ae. albopictus adults reared at 28°C and different temperatures was determined for the wAlbA and wAlbB infections. Real-time PCR was used for the quantification of the Wolbachia gene (wAlbA and wAlbB) as described by [18,26]. The wAblA and wAlbB gene was normalized to the mosquito housekeeping gene mRpS6 to adjust for different tissues. Primer pairs (wAlbA, F-5′-GGG TTG ATG TTG AAG GAG-3′ and R-5′-CAC CAG CTT TTA CTT GAC C-3′; wAlbB, F-5′-CCT TAC CTC CTG CAC AAC AA-3′ and R-5′-GGA TTG TCC AGT GGC CTT A-3′; mRpS6, F-5′-AGT TGA ACG TAT CGT TTC CCG CTA C-3′ and R-5′-GAA GTG ACG CAG CTT GTG GTC GTC C-3′) were used. The qPCR cycling conditions used are as follows: an initial incubation at 90°C for 5 min, followed by 45 cycles of amplification at 95°C for 10 s, 56°C for 20 s and 72°C for 20 s and a melting curve detection and a final cooling step of 40°C for 10 s. The relative quantification fold of Wolbachia was calculated using the following formula 2-(ct of wAlb/ct of mRpS6).

Statistical analysis
Statistical analyses were performed with Statistica 10 software (StatSoft, Inc., Tulsa, OK, USA). Initially, we performed a test to compare the five survival curves against dpi for each mosquito species using the nonparametric Kaplan-Meier analysis [27]. If this test revealed significant differences among all curves for each species, we then ran a paired log-rank test to identify which paired curves were different. Mosquitoes that were killed to detect the virus for each dpi category were classified as censored observations [28]. Additionally, a paired log-rank test was used to compare survival against dpi for Wolbachia-infected and Wolbachia-free mosquitoes of Ae. albopictus. Because of the small number of samples tested for each combination, Fisher's exact test was used to test the differences in thoraxabdomen and salivary gland infection rates. The Kolmogorov-Smirnov test was used to compare the overall differences in the entire growth curve of the virus at different temperatures in the thorax-abdomen (data on viral uptake at 0 dpi were excluded), salivary glands, and head samples. Student's t tests were used to detect the differences in virus titers between mosquito species, incubation temperatures and dpi.
The virus growth curves of infected Ae. aegypti at higher temperatures (28 and 34°C) were not significantly different (Kolmogorov-Smirnov test, P > 0.05) but were significantly different (Kolmogorov-Smirnov test, P < 0.05) from other temperatures (10, 16 and 22°C). When incubated at 34°C, the level of viral genomes was significantly (t-test, t = 2.38-4.12, df = 8-11, P < 0.05) higher than at 10°C and 5-10 dpi, at 16°C and 5-15 dpi or at 22°C and 20 dpi. At 22-34°C, the viral loads showed an increasing trend with a peak at 15 dpi. The peak viral titers at 22, 28 and 34°C were  Infection rates of Aedes albopictus at 10 dpi with incubation at 28°C were significantly higher than those at 34°C, as determined by Fisher's exact test. Significant differences were also observed at 15 dpi between 10°C and 16°C and between 28°C and 16 or 22°C  Table S1). The rate or titer of viral infection of the salivary gland is known as an indicator of transmission potential. After 10 dpi, higher salivary gland infection rates (range 25-100%) at 22-34°C were found compared with the lower incubation temperatures (range 0-20%), although no significant differences were detected (Fisher's exact, P > 0.05) ( Table 1). In the salivary gland samples of Ae. aegypti, no viral genomes were detected at 10°C or at 0-5 dpi and 16-34°C (Fig. 1b, Table 2). The viral load increased with infection time after 10 dpi. The virus growth curves of infected Ae. aegypti at higher temperatures (28 and 34°C) were not significantly different from each other (Kolmogorov-Smirnov test, P > 0.05) but were significantly different (Kolmogorov-Smirnov test, P < 0.05) from that at low temperatures (10, and 16°C). The growth curve at 22°C was not significantly different (Kolmogorov-Smirnov test, P > 0.05) from that at 28°C but was significantly different (Kolmogorov-Smirnov test, P < 0.05) from that of 34°C. Significant differences (t-test, t = 2.94-4.44, df = 5, P < 0.05) in virus titer were detected at 15 dpi between 16°C and higher (28 or 34°C) incubation temperatures. Additionally, the viral loads at 15-20 dpi and 34°C or at 15 dpi and 28°C were significantly higher (t-test, t = 2.43-3.85, df = 6-9, P < 0.05) than those at 22°C. At 16°C, only viral genomes were  Ae. aegypti 0 0 (0/10) 0 (0/10) 0 (0/10) 0 (0/10) 0 (0/10) Infection rates at 10 dpi with incubation at 22 or 28°C were significantly (P < 0.01 or 0.05) higher than those at 10 or 16°C, as determined by Fisher's exact test detected in one female at 10 and 15 dpi with viral loads of 10 1.54 and 10 1.82 PFU equivalents/ml, respectively. At 22°C , viral genomes were detected in the salivary glands of 2 females at 10 dpi with 10 2.09 ± 0.08 PFU equivalents/ml of the viral load. The viral loads increased to 10 2.69 ± 0.43 , 10 2.54 ± 0.31 , and 10 2.38 PFU equivalents/ml at 15, 20 and 25 dpi, respectively, with a salivary gland infection rate range of 83-100%. At 28 and 34°C, viral genomes were first detected in the salivary gland samples at 10 dpi and viral loads of 10 2.67 ± 0.33 and 10 2.96 ± 0.52 PFU equivalents/ ml (salivary gland infection rates = 50-63%). At 15-30 dpi, the viral loads detected were between 10 3.38 ± 0.49 and 10 3.89 ± 0.58 PFU equivalents/ml with a salivary gland infection range of 57-88%. In the head samples, viral genomes were first detected at 16°C and 15 dpi with a quantity of 10 1.95 PFU equivalents/ml (Fig. 1c). At 22-34°C , virus genomes were first detected at 10 dpi and were maintained as constant titers thereafter. Viral titers of viruses grown at low (10 or 16°C) and high (28 or 34°C) incubation temperatures at 10 dpi were significantly different (t-test, t = 3.05-55.33, df = 5-8, P < 0.05).

Impact of Wolbachia on Ae. albopictus
Since lower viral replication was found in Ae. albopictus compared with Ae. aegypti mosquitoes, we analyzed the infection rate of Wolbachia in the tested Ae. albopictus mosquitoes. Interestingly, the results showed that 92.8-97.2% of Ae. albopictus mosquitoes were co-infected with Wolbachia group A (wAlbA) and group B (wAlbB) ( Table 3). Only 1.4-2.9% of females were not infected with Wolbachia. To determine whether Wolbachia in Ae. albopictus affects DENV-1 replication, we generated Wolbachia-free mosquitoes with tetracycline treatment in 3 generations (Additional file 1: Figure S1). We fed Wolbachia-free mosquitoes with a blood-meal with the same DENV-1 and then incubated them at 28°C, which has the best viral load among the tested temperatures. The survival trend of Wolbachia-free Ae. albopictus mosquitoes fed with DENV-1 significantly (log-rank test, P < 0.05) decreased compared with Wolbachia-infected mosquitoes (Fig. 2a). The infection rate in Wolbachiafree Ae. albopictus mosquitoes increased from 5 to 20 dpi and then slightly dropped (Fig. 2b). The infection rate in the Wolbachia-infected mosquitoes was significantly higher at 5 (Fisher's exact test, P < 0.01) and 10 dpi (Fisher's exact test, P < 0.05) than in the Wolbachiafree mosquitoes. We also found the same transmission trend (increasing from 5 to 20 dpi and then dropping slightly) in the Wolbachia-free mosquitoes. The salivary gland infection rate of the Wolbachia-free mosquitoes was significantly (Fisher's exact test, P < 0.05) higher at 15 and 20 dpi compared to Wolbachia-infected mosquitoes at 28°C (Fig. 2c). We also found that the salivary glands of Wolbachia-infected mosquitoes became infected earlier at 5 dpi than those of the Wolbachia-free mosquitoes at the margin of significance (Fisher's exact test, P = 0.0542).
To minimize the impact of the antibiotic on mosquito microflora, the eighth generation (F8, tetracycline was stopped at F6) was also evaluated as the Wolbachia-free mosquitoes at 28°C condition. Furthermore, insect midgut and fat-body is highly related to the immune defense system for pathogens. We also detected the dengue viral genome copy in these two samples. High viral genome copies were detected in midgut samples, followed by fatbody, salivary gland, thorax-abdomen, and head samples in Wolbachia-free (F8) from 10 to 30 dpi (Fig. 3a). Only few viruses could be detected in all samples of Wolbachia-infected (F1) Ae. albopictus from 10 to 20 dpi, and no virus could be detected after 25 dpi (Fig. 3b).
Wolbachia can induce density-dependent inhibition of dengue virus in Aedes mosquitoes, and its density is temperature sensitive. Wolbachia density was also measured in different tissues at different temperatures in Wolbachia-infected F1 Ae. albopictus (Fig. 4). wAlbB (Fig. 4a) had higher densities than wAlbA (Fig. 4a). The relative density of wAlbA and wAlbB (ratio of wAlbB or wAlbA to host rps6 genomes) was commonly found in thorax-abdomen samples (0.1104 ± 0.0179 and 0.6048 ± 0.0617, respectively), followed by fat-body (0.0052 ± 0.0017 and 0.0652 ± 0.0103, respectively) and midgut samples (0.0003 ± 0.0002 and 0.0008 ± 0.0001, respectively). For thorax-abdomen samples, the temperature-dependent effect from 10 to 28°C on Wolbachia density was found in both strains. The results showed that both wAlbA (Fig. 4a) and wAlbB (Fig. 4b) were significantly located in thorax-abdomen compared to other tissues and had the highest density at the 28°C rearing condition. At 16°C and 22°C, Wolbachia density was also detected more than at 10°C and 34°C. Additionally, the density of wAlbB was higher than that of wAlbA in all tissues of Ae. albopictus (2.4-6.3 times in thorax-abdomen samples and over 12.4 times in the fat-body).

Discussion
Our data confirmed the temperature effects on virus replication as in previous studies, but we detected the native Wolbachia effect on Ae. albopictus vector competence by increasing DENV replication in Wolbachiacured Ae. albopictus. Aedes aegypti mosquitoes survived better, and virus titers were significantly higher at high temperature (28 or 34°C). However, Ae. albopictus mosquitoes lived longer at 16 and 28°C, and virus titers were significantly higher at 22 and 28°C (Fig. 1a, b, d, e and Fig. 5). Viruses were first detected at 10 dpi in salivary glands and head tissues in Ae. aegypti and 5 or 10 dpi in Ae. albopictus (Fig. 1b-d, f ). Wolbachia infections were detected in up to 97% of the tested Ae. albopictus mosquitoes (Table 3). We concluded that in southern Taiwan, Ae. aegypti is the main vector of dengue and Ae. albopictus has an insignificant role due to the high native Wolbachia infection in the local mosquito population. Therefore, the elimination or a significant reduction of Ae. aegypti populations is an effective method for dengue prevention in southern Taiwan, where Ae. aegypti and Ae. albopictus coexist.
The EIP is known to be temperature dependent [6,25]. Within viable temperature ranges, a higher temperature is associated with a shorter EIP. In our study, when incubation temperatures were greater than 22°C, the EIP in Ae aegypti for the dengue virus was between 5 and 10 days, regardless of the incubation temperature. These results were confined to our mosquito collection setting at 5 dpi intervals with a fixed virus uptake quantity, which was not sensitive for the detection of the EIP temperaturedependent trend in Ae. aegypti. However, this trend was clearly detected in the Ae. albopictus species, in which EIPs (between 0 and 5 days) were shorter at higher temperatures (28 and 34°C) and longer (between 5 and 10 days) at 22°C (Fig. 1e). The thresholds for dengue transmission in constant temperatures were 13°C [29] and 35°C [6,30], which was consistent with our results in Ae. aegypti (Fig. 1b, c), no virus particles in salivary glands and head samples were found at 10°C). Recent studies have simulated diurnal field temperature fluctuations to show that this factor could shorten the extrinsic incubation period [31][32][33].
In our study, the survival rates of Ae. albopictus mosquitoes infected with Wolbachia were significantly better (Fig. 5). Although the study of Mousson et al. showed the same survival trend, no significant difference was detected [14]. This outcome might be due the different virus strain infected (DENV-1 vs DENV-2) or mosquito strain. This study found that local Ae. albopictus has a high percentage of infection with wAlbA and wAlbB (Table 3). In Taiwan, 51.7% of mosquitoes in 29 species were infected with Wolbachia, and co-infections with wAlbA and wAlbB have also been recorded [21]. However, none of the Ae. aegypti mosquitoes (the first generation of field collected larvae, n = 32) were infected with Wolbachia in this study.
Wolbachia-free (F4 and F8) Ae. albopictus mosquitoes showed sufficient vector competence after orally fed dengue virus compared to original mosquitoes (F1) and Ae. aegypti, indicating that Wolbachia infection may play a role in Ae. albopictus transmission of the dengue virus. In our study, we showed the reduction of virus load in all tissue samples (Figs. 2, 3), which was inconsistent with other natural Wolbachia infection studies. In these studies, either no native Wolbachia effect [18] or only a b Fig. 5 Mosquito survival rates of Aedes aegypti (a) and Aedes albopictus (b) with days post-infection (dpi) at different incubation temperatures (n = 10 for each combination of temperature, dpi, and mosquito species) virus transmission reduction [14] was found in dengue virus in Ae. albopictus mosquitoes.
Altogether, we showed that low temperatures limit the vector competence of local Ae. aegypti and Ae. albopictus populations in the transmission of DENV-1. Furthermore, the highly native Wolbachia infection was able to reduce viral titers and limit the transmission of DENV-1 in the local Ae. albopictus population.

Conclusions
In southern Taiwan, Ae. aegypti is the main vector of dengue and Ae. albopictus has a non-significant role in the transmission of dengue virus due to the high prevalence of Wolbachia infection in the local mosquito population of southern Taiwan.

Additional file
Additional file 1: Table S1. Details of the virus titers of tested F1 mosquitoes. Table S2. Details of the virus titers of tested Wolbachia free F4 mosquitoes. Figure S1. Generation of the Wolbachia-free Ae. albopictus mosquitoes (DOCX 112 kb)