Ethics statement

All viral strains, Aedes albopictus populations and shipment used in this study were recorded in the Brazilian System for the Management of Genetic Heritage (SisGen A5D493C and RAAA843). The study protocol was approved by the Research Ethics Committee of the Oswaldo Cruz Institute (CEP FIOCRUZ/IOC) under approval number 75375223.6.0000.5248. Formal consent was not required as the study involved only viral strains and mosquito populations, with no direct participation of human subjects or collection of identifiable personal data.

Mosquito collections and rearing

Brazilian Ae. albopictus populations used in this experiment were obtained through eggs collected in oviposition traps in 2021 and 2022. Aedes albopictus population from the Southeast region (Rio de Janeiro, RJ, 22°52’33.0"S 43°14’51.4"W) originated from the Manguinhos campus, Oswaldo Cruz Institute, Fiocruz. The North region (Pará, PA1°22’28.69"S 48°23’04.04"W) and South region (Paraná, PR, 25°25’49.0"S 49°18’13.8"W) Ae. albopictus populations were collected during routine entomological surveys and provided by the Evandro Chagas Institute, Pará, and Federal University of Paraná, respectively. These sites were selected because they span distinct Brazilian regions and biomes, capturing a wide range of environmental conditions and historical patterns of CHIKV endemicity.

Field-collected eggs were reared in pans containing 1 L of tap water (150 larvae per pan) to adulthood on a diet with approximately 1g of yeast, larvae were fed with the same quantity of yeast, every two days, until pupal development. Upon pupation, pupae were collected daily and placed in cages until eclosion after which adult mosquitoes were identified to species and separated to establish a low-passage lab colony of Ae. albopictus. The mosquitoes were kept in cages (31 cm wide x 32 cm long x 33 cm high) within an insectary maintained at 28 ºC with a 12-hour light/12-hour dark cycle. They were provided with a 10% sucrose solution via cotton wicks and received weekly blood meals from chickens (IACUC protocol 201003892) at UF-FMEL. Aedes albopictus populations from Rio de Janeiro, Belém, and Paraná used in the experiment were from 3rd (F3), 4th (F4), and 4th (F4) generations, respectively.

Virus and mosquito oral infection

For experimental infection epidemic strains of CHIKV representing the genotypes currently circulating in Brazil were used. The PER160/H803609 strain (GenBank accession number KP164571.1), belonging to the Asian genotype, was isolated from the plasma of an infected patient from an epidemic that occurred in Pernambuco in 2014, and the BHI3741/H804705 strain (GenBank accession number KP164569.1), belonging to the ECSA genotype, was isolated from the serum of a patient during the epidemic that occurred in Bahia in 2014[52]. Both were provided by the Arbovirus and Hemorrhagic Fevers Section (SAARB) of the Evandro Chagas Institute, Pará.

For virus suspension preparation, African green monkey (Vero) cell monolayers were inoculated with diluted CHIKV stock at a multiplicity of infection (MOI) of 0.1 and incubated for one hour at 37 ºC in a 5% carbon dioxide atmosphere. These virological procedures were conducted at the UF-FMEL biosafety level 3 (BSL-3) facility, in accordance with established BSL-3 standard operating procedures. After the inoculation procedure, 20 mL of medium (composed of M199, 10% fetal bovine serum, 2% penicillin/streptomycin, and 2% nystatin (Mycostatin) was added to each flask, which were then incubated at 37 ºC under a 5% carbon dioxide atmosphere for 3 days.

Using an artificial membrane feeding system (Hemotek, Lancashire, UK) at 28 ºC for 1 hour, seven-to-ten-day-old adult females from each population were offered defibrinated bovine blood (HemoStat Laboratories, Dixon, CA, USA) containing freshly propagated CHIKV warmed to 37 ºC. The observed blood feeding rate was approximately 30%. The use of an artificial feeding system has the advantage of standardizing the viral dose across all mosquito populations compared to a live host model. Quantitative comparisons of vector competence measurements for two different chikungunya virus genotypes requires that mosquitoes ingest blood with similar viral titers which is much easier to control using the artificial feeding system than mosquitoes feeding on live hosts. Before and after each blood feeding trial, 1 mL aliquots of CHIKV-infected blood were collected and stored in 2 mL cryogenic vials (Millipore Sigma, Burlington, MA, USA) at −80 ºC. These aliquots were used to determine the viral titer of the infected blood, which was found to be 8.0 log₁₀ ± 0.32 plaque-forming unit equivalents per mL (PFUe/mL) in the blood meals. We deliberately used a high-titer blood meal to ensure that enough mosquitoes developed saliva infections for assessing treatment effects. Using lower viral titers would have limited our ability to detect treatment-dependent differences because too few mosquitoes would exhibit saliva infection.

Chikungunya virus dissemination and transmission

Following feeding trials, females were anesthetized with carbon dioxide and kept cool within a metal tray set on ice. Mosquitoes were sorted and thirty to forty fully engorged females were transferred to cages (0.47 L food containers with mesh lids) and held for 3-, 5- and 13-days incubation period until sample collections were tested to determine susceptibility to infection, disseminated infection, and transmission rates. To determine susceptibility to viral infection and disseminated infection, each mosquito body and legs, respectively, were homogenized (TissueLyser II sample disruptor; Qiagen, Germantown, MD, USA) in medium supplemented with fetal bovine serum and centrifuged before viral RNA isolation and detection by quantitative RT-PCR.

For transmission assays, females were deprived of sucrose, but not water, for 24 h and transferred to 37 mL plastic tubes (height x diameter: 8 by 3 cm) with removable mesh lids, containing cationic paper (Q) “Q-paper” (1 cm in diameter) soaked with honey dyed with blue food coloring (McCormick, Hunt Valley, MD, USA). Thus, individually held female mosquitoes were offered a sugary meal to induce female salivation to collect saliva samples. The visual detection of the food coloring through the crop allowed for determination of fed versus unfed individuals, as previously described in Honório et al. 2018 [42].

Viral nucleic acid extraction and quantitative RT-PCR

RNA was extracted from individual samples by processing 140 μL of homogenate with the QIAamp Viral RNA Mini Kit (Qiagen, Valencia, CA, USA). The RNA was eluted in 60 μL of buffer, following manufacturer’s protocol. CHIKV viral RNA was quantified by RT-qPCR, using the SuperScript III Platinum RT-qPCR Kit (Invitrogen, Carlsbad, CA, USA) in a CFX96 thermocycler (Bio-Rad Laboratories, Hercules, CA, USA). Each reaction comprised of a master mix containing 10 μL of 2X Reaction Mix, 2.2 μL of diethyl pyrocarbonate (DEPC) treated water, 1.0 μL of forward primer (10 μM), 1.0 μL of reverse primer (10 μM), 0.4 μL SuperScript™ III RT/Platinum™Taq Mix, 0.4 μL of probe, and 5.0 μL of viral RNA template (25% of reaction volume) for a total volume of 20 μL. Negative controls consisted of a sham viral RNA template of DEPC-treated water or dilute stock virus, each using a volume of 5.0 μL. Each mosquito sample and controls were tested in duplicate. Each sample with a quantification cycle (Cq) value of <35 was determined as positive. The thermocycling conditions were as follows: 50 ºC for 30 min, 94 ºC for 2 min, 39 cycles at 94 ºC for 10 s and 60 ºC for 1 min, and 50 ºC for 30 s. The following sequences represent primers designed to target a nonstructural polyprotein gene (accession ID of transcript, KU365292.1): forward, 5′-GTACGGAAGGTAAACTGGTATGG-3′; reverse, 5′-TCCACCTCCCACTCCTTAAT-3′. The probe sequence was 5′-/56-FAM/TGCAGAACCC ACCGAAAGGAAACT/3BHQ_1/- 3′ (Integrated DNA Technologies, Coralville, IA, USA). These primers and probe have been used successfully in previous studies assessing vector competence of mosquitoes for CHIKV [42]. Viral titer quantification was determined using a standard curve with serial dilutions of CHIKV stock, in parallel with titration by plaque assays of the same virus dilutions, expressed as plaque-forming unit equivalents (pfue)/mL.

Infection rate was determined by the number of females with CHIKV RNA-positive bodies from the total number that fed on the infectious blood meal. Disseminated infection rate was determined by the number of females with infected bodies that had CHIKV RNA-positive legs [42,84]. Figure 1 provides a detailed schematic overview of the experimental design.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 1. Schematic overview of experimental design.

Aedes albopictus adult females from each Brazilian population (Rio de Janeiro, Belém and Paraná) were orally challenged with Brazilian Asian and ECSA CHIKV genotypes infected blood to determine infection rate (IR), dissemination rate (DR) and transmission rate (TR). Figure created by the author (Maria Resck) using Adobe Illustrator and Getty Images (https://www.gettyimages.pt).

https://doi.org/10.1371/journal.pntd.0014522.g001

Statistical analyses

We analyzed the relationship between the presence or absence of CHIKV in bodies, legs, and saliva (dependent variables) of female Ae. albopictus and the following independent variables: CHIKV genotype (Asian and ECSA genotypes), mosquito population origin (Belém, Rio de Janeiro and Paraná) and days post-infection (dpi, 3, 5 and 13). Exploratory analyses were performed by constructing contingency tables, figures, using chi-square tests and univariate logistic models to analyze the overall relationship between each dependent variable and each of the independent variables. Likelihood-ratio Chi-square tests were used to test the significance of the sequential inclusion of the independent variables in binomial GLMs. We modeled this relationship using separate binomial generalized linear models: one focused on the bodies, one focused on the legs, and one focused on the saliva. We tested two-way interactions between genotype x population origin, genotype x days post-infection, and population origin x days post infections in all models, but the inclusion of the interactions did not incur in statistical significance. Models were evaluated via Akaike Information Criterion (AIC) to compare models with and without interactions. AIC provides a measure of the relative quality of statistical models by balancing goodness of fit and model complexity. Likelihood-ratio Chi-square tests were used to test the significance of the sequential inclusion of the independent variables in binomial GLMs. We also analyzed the relationship between viral titers of bodies, legs, and saliva and the main effects using a Gaussian Linear Model. All analyses were done using R [94] and RStudio [95].

Chikungunya virus infection by Ae. albopictus population origin, genotype and days post-infection

A total of 305 and 340 Ae. albopictus were tested following ingestion of ECSA and Asian genotypes of chikungunya virus, respectively, across the geographic populations and three periods during the incubation.

To investigate the dynamics of CHIKV infection in Ae. albopictus, we measured body positivity rates, defined as the proportion of mosquitoes with detectable viral infection at 3-, 5-, and 13-days post-infection (DPI) for both Asian and ECSA genotypes.

Figure 2A illustrates heterogeneity in positivity curves among Brazilian mosquito populations. For the Asian CHIKV genotype, the Ae. albopictus Belém population (red line) shows a relatively low positivity rate at 3 DPI, a substantial increase at 5 DPI, and a slight decrease at 13 DPI (positivity rate and lower and upper 95% Confidence Intervals; respectively 0.865 [0.809, 0.921]; 0.933 [0.888, 0.979]; and 0.92 [0.866, 0.974]). This suggests a slower initial infection establishment in this population, with peak viremia around 5 DPI. In contrast, the Paraná population (green line) exhibits high positivity rates as early as 3 DPI (0.92 [0.866, 0.974]; 1 for 5 and 7 DPI), remaining relatively stable across time points, indicating greater susceptibility or efficient early viral replication. The Rio de Janeiro population (blue line) displays an intermediate pattern, with a moderate initial positivity rate at 3 DPI (0.955 [0.923, 0.986]), an increase at 5 DPI (similar to Belém; Fig 2A), and a subsequent decrease at 13 DPI (0.966 [0.932, 0.999]), approaching initial levels. This late decrease may reflect viral clearance or lower persistence of detectable infection. Error bars indicate greater variability in the Belém and Rio de Janeiro populations, particularly at extreme time points, suggesting a more homogeneous response in the Paraná population.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 2. CHIKV infection rate in Ae. albopictus bodies across three time points (3-, 5-, and 13-days post-infection).

(A) Infection rates separated by virus genotype (Asian and ECSA), comparing the three mosquito populations (Belém, Paraná, and Rio de Janeiro). (B) Infection rates separated by mosquito population (Belém, Paraná, and Rio de Janeiro), comparing the two virus genotypes (Asian and ECSA). Error bars in both panels represent standard error of the mean.

https://doi.org/10.1371/journal.pntd.0014522.g002

Figure 2B presents CHIKV body positivity rates by mosquito population. Specifically, in the Belém population, the ECSA genotype shows heterogeneous positivity rates when compared to the Asian genotype, suggesting a varying susceptibility and replication (0.923 [0.871, 0.975]; 0.767 [0.689, 0.844]; respectively for 3, 5 and 7 DPI). The Paraná population has lower positivity rates for ECSA when compared to Asian genotype at all time points (0.938 [0.903, 0.972]; 0.925 [0.883, 0.967]; 0.933 [0.888, 0.979]). In the Rio de Janeiro population, the ECSA genotype show decreasing positivity from 3 to 5 DPI (respectively 0.86 [0.811, 0.909] and 0.846 [0.788, 0.904]) and a larger decrease at 13 DPI (0.8 [0.72, 0.88]), with the ECSA genotype showing consistently lower positivity rates when compared to the Asian genotype, especially at 3 and 5 DPI. These observations underscore the complex interaction between virus genotype and mosquito population in determining CHIKV infection dynamics.

Despite the observed variability in descriptive patterns (Fig 2A and 2B), the generalized linear model (GLM) analysis revealed no significant effect of time (days post-infection), population, or virus genotype on CHIKV positivity rates (Table 1; ANOVA results: p > 0.05 for all factors). This indicates that, on average, bodies of female Ae. albopictus populations from the tested populations tends to have a high infection rate despite origin, with most of the groups having more than 80% of infected individuals as early as 3 DPI, showing a rapid viral replication in the midgut.

Download:

• PNG
  larger image
• TIFF
  original image

Table 1. ANOVA results for the effect of Time Post-Infection, population, and virus genotype on CHIKV infection (body) and dissemination (legs) in Ae. albopictus.

https://doi.org/10.1371/journal.pntd.0014522.t001

The viral titers in the analyzed bodies significantly associated with time post-infection (p = 0.003) and viral genotype (p = 0.0003). A significant three-way interaction (time_point × population × virus_genotype) was observed (p = 0.027), indicating that the effect of the viral genotype on the titer is not constant and depends on the specific combination of time and population (Table 2).

Download:

• PNG
  larger image
• TIFF
  original image

Table 2. ANOVA results for the effect of time post-infection, population, and virus genotype on CHIKV body viral titers in Ae. albopictus.

https://doi.org/10.1371/journal.pntd.0014522.t002

The Estimated Marginal Means (EMMs) decomposition showed that the Asian – ECSA contrast was positive in most time × population combinations; that is, the Asian genotype exhibited higher viral titers than ECSA in the majority of the evaluated treatments. In Belém, Asian titers were higher than ECSA at 3 DPI (p = 0.0004) and 5 DPI (p = 0.0045), but not at 13 DPI (p = 0.221). In Paraná, there was no detectable difference at 3 DPI (p = 0.121); however, Asian was substantially higher than ECSA at 5 DPI (p < 0.0001) and 13 DPI (p < 0.0001), consistent with the evidence of a time × genotype interaction in this population. In Rio de Janeiro, Asian was higher than ECSA at 3, 5, and 13 DPI (p < 0.0001 for all comparisons), suggesting a strong and persistent genotype effect in this population (S1 Fig).

Chikungunya virus dissemination by Ae. albopictus population origin, genotype and days post-infection

Analysis of CHIKV positivity in Ae. albopictus mosquito legs revealed a distinct pattern compared to that observed in the mosquito bodies (Fig 3A and 3B). Generalized linear modeling demonstrated that time post-infection and mosquito population were significant factors affecting positivity rates in the legs (time point: p < 0.001; population: p = 0.014) (Table 1). There was a significant increase in positivity from 3 DPI to 5 DPI (p = 0.00116) and a further significant increase from 5 DPI to 13 DPI (p < 0.001), indicating a progressive dissemination of the virus in the legs over time, a crucial step for potential transmission. Mosquito population also played a role, with the Paraná population showing a significantly higher positivity rate than the Belém population (p = 0.00342), suggesting potential differences in susceptibility or dissemination efficiency between these mosquito populations. In contrast, the virus genotype (Asian or ECSA) did not significantly influence the overall positivity in the legs (p = 0.74791), suggesting a similar capacity for both genotypes to disseminate to this tissue among different geographic populations of Ae. albopictus.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 3. CHIKV dissemination in Ae. albopictus legs across three time points (3-, 5-, and 13-days post-infection).

(A) Dissemination rates separated by virus genotype (Asian and ECSA), comparing the three mosquito populations (Belém, Paraná, and Rio de Janeiro). (B) Dissemination rates separated by mosquito population (Belém, Paraná, and Rio de Janeiro), comparing the two virus genotypes (Asian and ECSA). Error bars in both panels represent standard error of the mean.

https://doi.org/10.1371/journal.pntd.0014522.g003

The viral titers in the analyzed legs significantly correlated with both time point and viral genotype. Notably, the three-way interaction was significant (p = 0.048), demonstrating a complex dynamic where each population and genotype varies significantly over time (Table 3). According to the Estimated Marginal Means (EMMs) analysis, the Asian–ECSA contrast proved to be strongly context-dependent. In Belém, the Asian genotype exhibited higher titers than ECSA at 5 DPI, though no detectable difference was found by 13 DPI. In Paraná, Asian titers were significantly higher at 3 and 5 DPI but not at 13 DPI, notably without robust evidence of a time x genotype interaction within this specific population. Conversely, in Rio de Janeiro, the time x genotype interaction was significant; the Asian–ECSA contrast was most pronounced at 5 DPI and remained significant at 13 DPI, after showing no significant difference at 3 DPI (S2 Fig). These log₁₀ differences reflect a significant biological fold-change, equivalent to a ten- to one-hundred-fold increase in viral titers.

Download:

• PNG
  larger image
• TIFF
  original image

Table 3. ANOVA results for the effect of time post-infection, population, and virus genotype on CHIKV legs viral titers in Ae. albopictus.

https://doi.org/10.1371/journal.pntd.0014522.t003

Chikungunya virus saliva infection by population origin, genotypes and days post-infection

Figure 4 shows positivity rates in saliva collected from the studied Brazilian Ae. albopictus populations. Overall, positivity rates in saliva were much lower than those observed in bodies and legs (Figs 2–4). For both CHIKV genotypes, the Belém population showed a low positivity rate at all tested DPIs, with mean values never surpassing a rate of 0.2 (Asian: 0.135 [0.0789, 0.191]; 0.0667 [0.0211, 0.112]; and 0.12 [0.0550, 0.185]; ECSA: 0.0385 [0.000747, 0.0762]; 0.0333 [0.000560, 0.0661]; and 0.176 [0.0840, 0.269]). The Paraná population showed higher positivity rates for the Asian genotype (0.318 [0.248, 0.388]; 0.259 [0.200, 0.319]; and 0.179 [0.118, 0.241]) and a mixed profile for the ECSA genotype when comparing the DPIs (0.0833 [0.0434, 0.123]; 0.2 [0.137, 0.263]; and 0.0667 [0.0211, 0.112]). Finally, the Rio de Janeiro population showed a heterogeneous infection pattern for both Asian (0.114 [0.0658, 0.161]; 0.189 [0.125, 0.254]; and 0.172 [0.102, 0.243]) and ECSA (0.12 [0.0740, 0.166]; 0.179 [0.118, 0.241]; and 0.04 [0.000808, 0.0792]) genotypes for DPIs. Model results showed no significant effect of time (days post-infection), population origin and virus genotype, which shows that, in general, all studied populations tended to show a similar saliva infection rate despite origin, virus genotype and how many days elapsed since ingestion of CHIKV infected blood.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 4. CHIKV infection rate in Ae. albopictus saliva across three time points (3-, 5-, and 13-days post-infection).

(A) Infection rates separated by virus genotype (Asian and ECSA), comparing the three mosquito populations (Belém, Paraná, and Rio de Janeiro). (B) Infection rates separated by mosquito population (Belém, Paraná, and Rio de Janeiro), comparing the two virus genotypes (Asian and ECSA). Error bars in both panels represent standard error of the mean.

https://doi.org/10.1371/journal.pntd.0014522.g004

For the saliva viral titers, no significant effects were detected for time (p = 0.725) or population (p = 0.183). However, a robust genotype effect was observed (p < 0.0001): the ECSA genotype presented saliva titers approximately 1.27 log₁₀ units lower than the Asian genotype. This corresponds to a biological difference of approximately 19-fold (fold-change of 10^1.273) in viral titer magnitude, on average, while adjusting for time and population (S3 Fig). These results should be interpreted with caution, as the high rate of missing observations for saliva reduced the statistical power to detect finer variations between treatments.

Taken together, these findings support the hypothesis that genotypic differences are detectable across multiple tissues. However, spatiotemporal heterogeneity (population and time) is more pronounced during the infection and dissemination phases (body and legs). In contrast, at the final stage closest to potential transmission (saliva), the statistical inference is dominated by a CHIKV genotype effect and remains heavily constrained by the limitations of the available sampling design.

Infection

Aedes albopictus, introduced in Brazil in the 1980s, is now well-established and widespread across the country. In this study we compared the vector competence of Brazilian populations of Ae. albopictus for two strains of chikungunya virus (CHIKV), Asian and ECSA, responsible for outbreaks in 2014 in Oiapoque, Amapá and Feira de Santana, Bahia regions, respectively [51,52]. Our investigation revealed distinct patterns across infection, dissemination, and transmission stages among the three Brazilian Ae. albopictus populations but showed similar vector competence for both Brazilian viral strains. Despite observed descriptive heterogeneity in body positivity rates among the three Brazilian Ae. albopictus populations (Belém, Paraná, and Rio de Janeiro) and between Asian and ECSA CHIKV genotypes (Fig 2A and 2B), generalized linear model (GLM) analysis indicated no statistically significant effect of time post-infection, mosquito population, or virus genotype on overall CHIKV infection rates (p > 0.05 for all factors; Table 1).

Our results also align with previous reports in the literature, which have demonstrated a lack of significant genetic and geographic variations in Ae. albopictus susceptibility to infection for CHIKV. Severini et. al. (2018) [79] for instance, demonstrated that all eight Italian populations of Ae. albopictus tested exhibited comparable abilities to be infected by the mutated ECSA CHIKV strain (E1-226V), with no statistically significant differences observed in mean viral titers within the bodies (Kruskal Wallis test, p = 0.825). Fortuna et al (2018) [88] supported that same information with one population of Italian Aedes albopictus for two strains of chikungunya virus (CHIKV), with and without E1:A226V mutation. More broadly, these results underscore the established role of Ae. albopictus as a highly competent vector for CHIKV across diverse genotypes and geographical origins. Several studies, consistently report its efficient infection capabilities, suggesting similar and high permeability of the midgut infection barrier [28,42,49,79,80,85–88,90,91].

The lack of a statistically significant difference suggests that, once exposed to an infectious blood meal, the midgut barrier in these Brazilian Ae. albopictus populations is generally highly permissive to both Asian and ECSA CHIKV genotypes. This rapid and high initial infection establishment, with over 80% of individuals infected as early as 3 days post-infection (DPI) regardless of origin, implies a highly efficient process of viral entry and replication within the midgut cells. It is possible that the descriptive variations observed (e.g., Belém’s slower initial rise for Asian genotype) represent subtle kinetic differences that do not ultimately translate into significant overall differences in the proportion of infected mosquitoes over the 13-day period. The unique kinetic pattern observed in Belém — characterized by a delayed infection peak at 5 — could reflect slower viral replication kinetics within the midgut. This potential ‘saturation effect’ by 3–5 DPI may mask early variations that could be more apparent at lower viral doses or earlier time points. Given the widespread distribution of Ae. albopictus across Brazil, this uniform susceptibility to high-dose infection underscores the species’ significant potential to facilitate the local spread of CHIKV outbreaks, regardless of the circulating genotype.

Dissemination

In contrast to infection, CHIKV dissemination to the legs was significantly influenced by both time post-infection (p < 0.001) and mosquito population (p = 0.014; Table 1). We observed a strong and progressive increase in leg positivity over time, with significant increases for each subsequent period measured following ingestion of CHIKV infected blood. This highlights the time-dependent nature of viral spread within the mosquito. This progressive increase in disseminated infection is consistent with the general understanding of arbovirus replication kinetics within the mosquito vector, where the virus must overcome the midgut barrier, replicate, and then spread systemically to reach secondary tissues like the legs before ultimately infecting the salivary glands. Furthermore, the Paraná Ae. albopictus population exhibited a significantly higher dissemination rate than the Belém population (p = 0.00342), suggesting a potentially greater intrinsic susceptibility or more efficient viral spread within the Paraná mosquitoes. Notably, neither Asian nor ECSA CHIKV genotypes significantly impacted overall dissemination rates (p = 0.748), implying similar capacities for both genotypes to disseminate to peripheral tissues. These observations suggest that populations of Ae. albopictus from Paraná may exhibit a less effective midgut escape barrier or a greater magnitude of viral replication within the midgut — or a combination of both — compared to other Brazilian populations tested, and future studies should investigate the relative contribution of each mechanism.

These observed patterns of dissemination are supported by and expand upon findings from other comparative studies on Aedes vector competence for CHIKV. For instance, Honório et al 2018 [42] conducted studies using the Asian genotype of CHIKV across Brazilian and Florida Ae. aegypti and Ae. albopictus populations, and their results also demonstrated a significant increase in disseminated infection over time, consistent with our findings on the time-dependent nature of viral spread. Honório et al. 2018 [42] primarily focused on the Asian CHIKV genotype, however, in contrast to our observation of a significant population effect on dissemination (as observed with the Paraná vs. Belém populations), these authors also found no significant effect of mosquito population origin on viral dissemination rates in their study. Similarly, Severini et al 2018 [79] found that Italian Ae. albopictus populations showed no statistically significant differences in dissemination rates or mean viral titers in legs and wings (Kruskal Wallis test, p = 0.609), despite high overall dissemination. This discrepancy highlights potential differences in the specific populations tested, experimental conditions, or the statistical power to detect such effects across studies.

Vega-Rúa et al. (2020) [96] further elucidated this complexity, demonstrating that the vector competence of Ae. albopictus populations for CHIKV, including both dissemination and transmission efficiencies, is significantly influenced by their demographic history and genetic ancestry. Their work, which investigated different CHIKV strains (Asian, ECSA, ECSA E1-226V), revealed complex genotype-by-genotype interactions where mosquito genetic lineages and specific viral strains influenced competence, leading to varying dissemination and transmission efficiencies across different Ae. albopictus populations.

This population-specific difference in dissemination, in contrast to the uniform rates of initial infection, points to a potential bottleneck or barrier to viral spread beyond the midgut that varies among mosquito populations. Notably, neither the Asian nor ECSA genotypes significantly impacted overall dissemination rates, a finding that aligns with observations across ten countries in the Americas [28]. This suggests that while these genotypes may exhibit subtle differences in their initial midgut replication kinetics, their ability to overcome systemic barriers and reach peripheral tissues is comparable. Our results further emphasize that once a midgut infection is established, Brazilian Ae. albopictus is highly and equally permissive to the systemic spread of both circulating strains. This time-dependent progression of viral dissemination is a crucial biological determinant of the extrinsic incubation period (EIP), directly impacting the duration of a mosquito’s infectiousness and its overall capacity for onward transmission in the field.

Transmission

Our finding regarding transmission was the consistently low positivity rates in saliva across all Ae. albopictus populations and CHIKV genotypes, which were substantially lower than those observed in bodies and legs (Fig 4). This result highlights a significant barrier in the transmission process occurring at or before the infection of the salivary glands and subsequent release of infectious virus into the saliva. Despite visual variations in the descriptive patterns, a generalized linear model revealed no significant effect of days post-infection, mosquito population, or virus genotype on saliva positivity rates. This indicates that, on average, the probability of infectious virus being present in the saliva, and thus capable of transmission, remains consistently low regardless of these factors.

The mean saliva viral titers observed in positive mosquitoes provide additional context for interpreting transmission potential. For the Asian genotype, mean saliva titers ranged from 2.53 to 4.06 log₁₀ PFUe/mL across populations and time points. Notably, the published minimum infectious dose for CHIKV in Ae. albopictus has been estimated at approximately 3.9 log₁₀ pfu/mL [97], suggesting that only a subset of saliva-positive mosquitoes — those at the higher end of this range — may have shed virus at levels sufficient for effective transmission. In contrast, mean saliva titers for the ECSA genotype were consistently lower across all populations and time points (1.51 to 3.72 log₁₀ PFUe/mL), falling below this threshold in most conditions, further supporting the limited transmission potential of this genotype even among saliva-positive individuals.

We observed a trend for a reduction in the proportion of mosquitoes with a saliva infection at later time points in the Belém and Paraná mosquitoes infected with the Asian genotype of CHIKV. These results are consistent with observations that saliva infection declined with the length of infection in Ae. albopictus from Okeechobee, Florida and Rio de Janeiro, Brazil infected with an Asian genotype of CHIKV [42]. Impaired transmission efficiency among older mosquitoes is likely attributable to virus modulation of infection as observed in other mosquitoes and arboviruses [98–100]

The progression from midgut infection to salivary gland infection involves at least two well-characterized barriers: the midgut escape barrier (MEB), which limits systemic dissemination, and the salivary gland infection barrier (SGIB), which limits viral entry and replication within salivary gland tissue. Our data suggest that the MEB is relatively permissive in all three Brazilian Ae. albopictus populations, given the high and rapid body infection rates (>80% at 3 DPI) and progressive dissemination to the legs over time. However, the substantially lower saliva positivity rates — despite high dissemination — point to the SGIB as the predominant bottleneck limiting transmission. The lower saliva titers observed for the ECSA genotype compared to Asian, even among saliva-positive individuals, may further reflect genotype-specific differences in the ability to replicate within or escape from salivary gland cells, rather than simply differences in systemic viral load. The Belém population showed very low transmission potential, never surpassing a mean 0.2 positivity rate for either genotype across all time points. This might suggest an even more pronounced SGIB or a less conducive environment for viral replication within the salivary glands, although the overall lack of statistical significance for population effect on saliva positivity suggests a generalized challenge for transmission across all tested Ae. albopictus groups. Regardless of similar transmission rates observed among populations, greater absolute numbers of disseminated infections (i.e., from the total number of mosquitoes ingesting infected blood) exhibited by the Paraná population of Ae. albopictus is predicted to yield a higher risk of saliva infection relative to other populations of Ae. albopictus from Brazil.

This overall low and non-variant transmission efficiency, despite high infection and significant dissemination in some populations, contrasts sharply with the known role of Ae. albopictus as an important CHIKV vector in other geographical regions. For instance, Vega-Rúa et al. (2014) [28] reported transmission efficiencies reaching up to 96.7% in American populations, and Honório et al. (2018) [42] found rates as high as 82% in Brazilian Ae. albopictus populations (e.g., from Manguinhos, Rio de Janeiro). Furthermore, Richards et al. (2010) [80] observed the highest infection, dissemination, and transmission rates in Florida Ae. albopictus compared to Ae. aegypti and Culex pipiens quinquefasciatus, concluding that, Ae. albopictus is a competent vector of CHIKV, though they also noted a bottleneck where not all mosquitoes with disseminated infections could transmit the virus in saliva. However, the low numbers of Ae. albopictus females successfully reaching the final stage of the experiment can be considered an important limitation of our study. This reduced sample size in the saliva group may have limited our ability to resolve finer statistical interactions that were more apparent in the body and leg tissues.

The observed epidemiological trends of chikungunya in Brazil, marked by significant case numbers and fatalities from 2023 to 2025, underscore the critical role of vector dynamics in disease transmission. Our findings, revealing higher heterogeneity in body positivity rates among Ae. albopictus populations from Northern, Southeastern, and Southern Brazil exposed to locally circulating Asian and ECSA CHIKV genotypes, raise pertinent questions regarding the intrinsic competence of these specific mosquito populations under varying conditions. Factors such as diverse environmental settings or serial viral passages could influence these rates, suggesting a complex interplay between the vector, the virus, and local ecological factors. A deeper understanding of this variability is crucial for refining predictive models and developing targeted public health interventions, especially given Ae. albopictus’s potential as a “bridge vector” at the urban-forest interface, which could facilitate future arbovirus spill-over events and outbreaks.

While this study provides important insights into CHIKV vector competence, some considerations should be noted. The use of a relatively high viral titer and a standardized artificial feeding system, although necessary for experimental control, may not fully reflect natural infection conditions. In addition, the limited number of mosquitoes reaching the saliva analysis stage may have reduced statistical power. Despite this, the study design enabled robust comparisons across populations and genotypes.