Skip to main content
NCBI home page
Virus Evolution logoLink to Virus Evolution
. 2025 Dec 23;12(1):veaf102. doi: 10.1093/ve/veaf102

Intrahost population dynamics of chikungunya virus in humans and naturally infected Aedes mosquitoes reveal transmission-driven diversity

Cecília Artico Banho 1, Beatriz de Carvalho Marques 2, Olivia Borghi Nascimento 3, Maisa Carla Pereira Parra 4, Maria Vitória Moraes Ferreira 5, Ana Paula Lemos 6, Gabriel Pires Magnani 7, Karine Lima Lourenço 8, Beatriz Cunha de Souza 9, Victor Miranda Hernandes 10, Marini Lino Brancini 11, Cassia Fernanda Estofolete 12,13, João Pessoa Araújo Júnior 14, Nikos Vasilakis 15,16,17, Maurício Lacerda Nogueira 18,19,20,✉
PMCID: PMC12840586  PMID: 41607870

Abstract

RNA viruses exhibit a high mutation rate, contributing to their genetic diversity mainly because their RNA polymerase lacks proofreading ability. Arboviruses, which alternate between vertebrate and invertebrate hosts, are subjected to host-specific selective pressures and population bottlenecks, mainly within mosquito vectors. Although experimental studies have brought insights into their evolutionary dynamics, data from naturally infected vectors remain limited. Here, we investigated the intrahost genetic diversity of chikungunya virus (CHIKV-ECSA lineage) through whole-genome sequencing of 19 human- and 19 mosquito-derived genomes from the 2024 outbreak in São José do Rio Preto, Brazil. Our principal component analysis revealed a greater mutation number in mosquito-derived genomes, predominantly driven by low-frequency and unique variants. Overall, intrahost genetic diversity was significantly higher in mosquito-derived than in human-derived CHIKV genomes, and protein-coding regions showed host-specific patterns. We identified 303 mutations across all CHIKV genomes. Interestingly, shared mutations were predominantly classified as synonymous, whereas unique mutations were mainly nonsynonymous. Gene-wide selection analyses indicated that purifying selection predominates across CHIKV genomes from both humans and mosquitoes, suggesting that most mutations, particularly nonsynonymous ones, are deleterious and subject to purifying selection. However, in mosquito-derived CHIKV genomes, evidence of relaxed purifying selection and neutral evolution, in specific proteins, such as E3 and NSP3, respectively, was observed, in contrast to the stronger purifying selection observed in human-derived CHIKV sequences. Site-specific selection analyses corroborated these results, detecting negatively selected sites in human-derived genomes but not in mosquito-derived genomes for these specific proteins. Together, our results show that these host-specific differences enable mosquitoes to act as reservoirs of genetic diversity by maintaining nonsynonymous variants, likely driven by genetic drift. At the same time, human hosts may impose stronger selective pressures, contributing to preserving the genome stability. This dynamic balance between diversification in vector populations and selective constraints in vertebrate hosts likely drives CHIKV evolution and adaptation.

Keywords: within-host viral diversity, Aedes mosquitoes, arbovirus, evolution

Introduction

RNA viruses are well known for their high mutation rates, which play a key role in the emergence of new variants and, often, the onset of epidemics (Duffy 2018, Domingo and Perales 2019, Villa et al. 2021, Domingo et al. 2021). The high rate of change is largely attributed to the lack of proofreading activity by the viral RNA-dependent RNA polymerase, leading to frequent replication errors during RNA synthesis (Holland 1992, Domingo 2010, Sanjuán et al. 2010, Stapleford et al. 2014, Duffy 2018, Domingo and Perales 2019, Domingo et al. 2021). Despite most emerging mutations being considered deleterious, some may confer adaptive advantages, potentially influencing viral pathogenesis, transmission dynamics, tissue tropism, and immune evasion (Joos et al. 2005, Tsetsarkin et al. 2007, Sanjuán et al. 2010, Sessions et al. 2015, Shi et al. 2016, Villa et al. 2021, Liang 2023). In addition to selective pressures, stochastic processes such as genetic drift also play a role in RNA virus evolution at intra- and interhost levels (Gutiérrez et al. 2012, Lequime et al. 2016, McCrone et al. 2018, Weaver et al. 2021). Together, these forces allow remarkable genome plasticity of RNA viruses, enabling adaptation to changing environments and conditions.

This evolutionary plasticity is particularly relevant for arthropod-borne viruses (arboviruses), which are transmitted between vertebrate and invertebrate hosts. As these hosts impose different selective pressures, alternating between them often exposes viral populations to bottlenecks, driving the fixation or loss of viral variants, contributing to shaping intraspecific genetic diversity over time (Forrester et al. 2014, Coffey and Reisen 2016, Grubaugh et al. 2016, Angleró-Rodríguez et al. 2017, Caldwell et al. 2020, Merwaiss et al. 2021). Within arthropod vectors, anatomical barriers such as infection and escape from the midgut, as well as infection of the salivary glands, impose significant population bottlenecks during viral infection and dissemination (Forrester et al. 2012, Franz et al. 2015). In addition to these physical barriers, mosquitoes present several conserved antiviral pathways, including Jak/STAT (Janus kinase signal transducer and activator of transcription), Toll, and RNA interference (RNAi), which play critical roles in viral diversity (Brackney et al. 2009; Blair 2011; Ahlers and Goodman 2018, Lee et al. 2019, Prince et al. 2023). Despite their conserved nature, these pathways do not act in a universally uniform manner; instead, their effects often depend on the specific virus–host–tissue context (Jupatanakul et al. 2017, Simões et al. 2018, Tikhe and Dimopoulos 2021). Indeed, host-switching events and species-specific immune pressures can influence viral evolution by introducing new selective constraints. In vertebrate hosts, immune evasion mechanisms are typically constrained by purifying selection, particularly when combined with innate and adaptive immune responses, which eliminate less fit viral variants (Ahlers and Goodman 2018, Nelemans and Kikkert 2019). In contrast, hosts with impaired immune function, such as those with B- or T-cell deficiencies, viral immunosuppression, or receiving immunosuppressive therapies, may experience a relaxation of these selective pressures, allowing the persistence of suboptimal or immune-escaping variants and potentially promoting greater intrahost viral diversification (Lustig et al. 2024, Marques et al. 2024, Joseph et al. 2025).

Several studies that have explored the evolutionary dynamics of arboviruses, using in vitro approaches or experimental infections in mosquitoes, have provided important insights into how mutations accumulate and are selected during host switching, replication, and transmission (Coffey et al. 2008, Vasilakis et al. 2009, Coffey and Vignuzzi 2011, Sessions et al. 2015, Coffey and Reisen 2016, Grubaugh et al. 2016, Lequime et al. 2016, Merwaiss et al. 2021, Fitzmeyer et al. 2023). Some in vivo studies have been important in revealing how adaptive mutations shape vector competence and transmission potential (Tsetsarkin et al. 2007, 2014). However, most of these experimental models rely on controlled infections, which may not fully capture the ecological and evolutionary complexity of natural transmission cycles. Thus, studies assessing intrahost viral diversity in naturally infected vectors remain limited. The study of wild-caught mosquitoes, which acquire infection through natural feeding and replicate the virus under field conditions, can provide crucial insights into how stochastic and selective forces act in natural transmission cycles, increasing our understanding of how these evolutionary pressures shape viral populations across hosts and define the constraints and opportunities faced by arboviruses.

Therefore, to address this subject, we conducted intrahost genetic diversity analyses of chikungunya virus (CHIKV) in both humans and naturally infected Aedes aegypti during an outbreak setting. By analysing field-derived viral populations, we aimed to evaluate the evolutionary processes influencing arbovirus diversity in vertebrate and invertebrate hosts and better understand the role of host biology in shaping viral evolutionary patterns.

Material and methods

Mosquito collection

Aedes aegypti were collected in São José do Rio Preto (SJdRP), a city located in the northwestern region of the State of São Paulo, Brazil. Collection points were distributed across a north–south and east–west grid within the municipality and were primarily composed of residential properties, as described by Banho et al. (2025). In total, 43 residences across six neighbourhoods were selected as collection sites, and mosquito sampling was conducted from May to July 2024 (Supplementary Fig. 1). Verbal consent for property access was obtained from residents during each visit. Whenever possible, collection points were spaced 100–300 m apart, corresponding to the average flight range reported for Ae. aegypti (Reiter et al. 1995, Maciel-de-Freitas and Lourenço-de-Oliveira 2009, Moore and Brown 2022). At each collection point, a BG-Sentinel trap (Biogents, Germany) was deployed once per month in peridomestic areas and operated continuously for a 24-h sampling period. Following collection, live specimens were transferred to appropriate plastic containers and transported to the Laboratório de Pesquisas em Virologia (LPV) at the Faculdade de Medicina de São José do Rio Preto (FAMERP). Specimens were cold-ice-anaesthetized for species-level identification using established taxonomic keys (Consoli and de Oliveira 1994, Forattini 1996). Identified specimens were individually stored in 1.5 ml polypropylene tubes at −80°C until further analyses, which were performed within 10 days of collection.

The collection of mosquitoes in the SJdRP area was authorized by SISBIO (Sistema de Autorização e Informação em Biodiversidade) under the protocol number: 88475-4. This study was also approved by the ethics review board of the FAMERP (protocol CAAE 79090324.8.0000.5415, approved on 30 April 2024). All data were analysed anonymously, ensuring total confidentiality for all participants.

Serum samples

To compare the intrahost genetic diversity of CHIKV across hosts, serum samples from symptomatic patients with confirmed chikungunya fever were evaluated. The samples were obtained during January–April 2024 from individuals residing in SJdRP, who were attended at Hospital de Base de São José do Rio Preto, a health reference center that serves patients from 101 municipalities within northwestern São Paulo. The samples were collected for routine diagnosis, and the Ethics Committee of the FAMERP waived informed consent under protocol CAAE 02078812.8.0000.5415, approved on 3 July 2012, and amended on 11 May 2016.

Arbovirus detection

For molecular screening, 400 μl of ice-cold 1× Phosphate Buffered Saline (PBS) supplemented with 2% penicillin and 1% amphotericin B solution was added to each adult mosquito collected monthly. Samples were homogenized with a stainless-steel bead using the L-BEADER mechanical cell disruptor (Loccus, Cotia, Brazil) for three cycles of 30 s at 3000 rpm. The samples were centrifuged at 5340 rcf for 10 min, and the resulting mosquito macerates were used for viral RNA extraction, following the protocol described by MacHado et al. 2012. Total RNA from human samples was extracted using the QIAamp Viral RNA Mini Kit (QIAGEN, Hilden, Germany), according to the manufacturer’s instructions. Molecular analyses were subsequently performed to detect CHIKV RNA. One-step real-time polymerase chain reaction (RT-qPCR) was carried out using the GoTaq Probe 1-Step RT- qPCR System (Promega, Madison, USA) with TaqMan fluorescent primers and probes specific to CHIKV (CHIKV-874: 5′-AAAGGGCAAACTCAGCTTCAC-3′; CHIKV-961: 5′-GCCTGGGCTCATCGTTATTC-3′; CHIKV-899: 5′-[FAM]CGCTGTGATACAGTGGTTTCGTGTG-3′), as described by Lanciotti et al. (2007). Reactions were performed on a QuantStudio 3 Real-Time PCR System (Thermo Fisher Scientific, MA, USA) in linear and multi-component mode, with the baseline set to the positive and negative control cycle thresholds (Ct) to minimize reagent interference. Results were considered positive when the Ct value was below 38.

CHIKV whole-genome sequencing

After molecular screening for CHIKV in the mosquito and human samples, whole-genome sequencing was performed. Library construction, complementary DNA synthesis, genome amplification, and library preparation were carried out according to the instructions provided for the Illumina Microbial Amplicon Prep (Illumina, San Diego, CA, USA), using CHIKV-specific primer pools described by Quick et al. (2017). Library quantification was performed using the Qubit dsDNA HS Assay on a Qubit 2.0 device (Invitrogen, Waltham, MA, USA). The TapeStation 4 150 system and High Sensitivity D1000 ScreenTape kit (Agilent Technologies, Santa Clara, CA, USA) were used to assess library quality control. Sequencing was performed with a MiSeq Reagent kit v2 (2 × 150 cycles) (Illumina, San Diego, CA, USA), and the NextSeq PhiX Control kit was used as a normalization sample with the libraries sequenced on the MiSeq system (Illumina, San Diego, CA, USA).

Genome assembly

The quality of raw reads was assessed using the FastQC v. 0.11.4 program (Andrews 2010), and cutadapt v. 4.6 software (Martin 2011) was used to remove low-quality reads (Phred score > 30) shorter than 50 bp, as well as duplicate sequences, adapters, and primers used during library construction. Clean reads were then mapped against the CHIKV reference genome (KU940225.1) using BWA mem v. 0.7.17-r1188 software (Li and Durbin 2009) and SAMtools v. 1.6 (Li et al. 2009) for read sorting and indexing. After postprocessing steps, the assembled genomes were recovered using iVar v. 1.3.1 (Castellano et al. 2021), and coverage metrics were accessed using SAMtools v. 1.6 (Li et al. 2009).

Phylogenetic analysis

The generated consensus sequences were analyzed using the Genome Detective virus typing tool (Vilsker et al. 2019) to classify genotypes. Phylogenetic analyses were performed to observe similarity patterns of CHIKV sequences from mosquito and human samples. The assembled genomes were aligned with the reference CHIKV genome (KU940225.1) using MAFFT v. 7.520 (Katoh and Standley 2013). AliView v. 1.28 (Larsson 2014) was used to inspect the alignment and manually trim the ends of the aligned sequences. A maximum likelihood (ML) tree was reconstructed using IQ-TREE 2 v. 2.2.2.3 (Minh et al. 2020), using the best nucleotide substitution model inferred according to the Bayesian information criterion in ModelFinder (Kalyaanamoorthy et al. 2017). Branch reliability was tested using ultrafast bootstrap approximation (UFBoot) (Minh et al. 2013) and SH-like approximate likelihood ratio test (SH-aLRT) (Anisimova and Gascuel 2006), each with 10 000 replicates. We then used TreeTime v. 0.9.3 (Sagulenko et al. 2018) to convert the raw ML tree into a tree-scaled tree, as described by Banho et al. (2025). Phylogenetic trees were visualized and edited using R v. 3.6.1 (R Core Team 2021) and the ggtree package (Xu et al. 2022). To investigate the temporal signal from the ML tree, we performed a root-to-tip genetic distance regression against sample collection dates using the TempEst tool v. 1.5.1 (Rambaut et al. 2016), accepting temporal structure if the correlation coefficient was > 0.4, according to Banho et al. (2025).

Intrahost diversity analyses

Intrahost variability analyses were performed using the iVar variants tool (Castellano et al. 2021), applying the following parameters for intrahost single-nucleotide variant (iSNV) identification: minimum coverage of 100 reads per genomic position, base quality score >30, and minimum alternative allele frequency (AAF) ≥10%. The biological effects of the identified iSNVs were annotated using a custom R script, based on the genomic annotation file (GFF3 format) corresponding to the reference genome used for read alignment (KU940225.1).

To assess the genetic diversity of viral populations within individual hosts, we calculated two complementary metrics following Lequime et al. (2016): the nucleotide diversity (π) and the normalized Shannon entropy (Sn). The nucleotide diversity (π) quantifies the average number of pairwise differences per site between coexisting variants and was computed for each genomic position according to:

graphic file with name DmEquation1.gif

where D represents the sequencing depth at the site, and p denotes the minor allele frequency. By definition, π values range from 0 (no polymorphism) to 0.5 (when both alleles occur at equal frequency). For each sample, π values were averaged across all sites that passed quality filters and subsequently summarized per sample to explore overall diversity patterns.

The normalized Shannon entropy (Sn) was calculated for each nucleotide site as:

graphic file with name DmEquation2.gif

where p represents the alternative allele frequency at the given nucleotide site, and ln(4) corresponds to maximum complexity (i.e. four possible nucleotides at each position), as described by Lequime et al. (2016). As in our dataset, only biallelic iSNVs were found; they are expected to have Sn values ranging from 0 (no diversity) to 0.5 (maximum diversity, i.e. when the two alternative nucleotides are present at the same frequency). For each sample, Sn values were averaged across all sites within the same sample, yielding a per-protein estimate of intrahost genetic complexity for each protein-coding gene. Comparisons between Sn and average coverage were performed to assess possible biases in variant detection arising from sequencing depth. Allele frequency visualization was conducted in the R environment v. 3.6.1 (R Core Team 2021) using the ggplot2 package (Wilkinson 2011) for graphical representation, throughout a custom script.

Selection analysis

Subsequently, molecular selection analyses were carried out to infer potential selective pressures acting on specific genomic regions or genes from the consensus sequences. These analyses identified purifying, diversifying, and episodic selection signals across the viral genome. Therefore, two complementary evolutionary inference approaches implemented in the HyPhy software v. 2.5.1 (Kosakovsky Pond et al. 2020) were employed: the Fixed Effects Likelihood (FEL) method (Kosakovsky Pond and Frost 2005), which detects pervasive purifying or diversifying selection at individual codon sites, and the Mixed Effects Model of Evolution (MEME) (Murrell et al. 2012), which identifies both pervasive and episodic diversifying selection. Both methods rely on ML inference to estimate site-by-site nonsynonymous (dN) and synonymous (dS) substitution rates from a codon alignment and the corresponding phylogenetic tree.

Statistical analysis

To compare intrahost diversity between host species (Ae. aegypti and humans), the normality of the variables was first assessed using the Shapiro–Wilk test. Comparisons of π and Sn between Ae. aegypti– and human-derived CHIKV genomes were performed using two-sample independent t-tests, with 95% confidence intervals. To evaluate the potential impact of sequencing depth on diversity estimates, we calculated the Spearman’s rank correlation coefficient (ρ) between mean sequencing coverage and Sn. Next, protein-specific intrahost diversity was compared between hosts using an independent Wilcoxon rank-sum test applied to each viral protein. All statistical analyses were conducted in R version 3.6.1 (R Core Team 2021), using the packages dplyr (Wickham et al. 2025), ggplot2 (Wilkinson 2011), and rstatix (Kassambara 2023).

Results

To better understand the factors shaping intrahost genetic diversity in natural virus populations, we analyzed whole-genome sequences of CHIKV belonging to the East/Central/South African (ECSA) lineage, obtained from 19 human and 19 mosquito samples. All mosquito samples were collected in São José do Rio Preto (SJdRP) (Supplementary Fig. 1), and all human samples were collected from residents of the same city, between January and July 2024. Overall, sequencing coverage was high, with average depths ranging from 1317× to 4482× (Supplementary Table 1).

First, we reconstructed a phylogenetic tree using the Maximum Likelihood (ML) method to examine the genetic relationship among samples and identify clustering patterns (Supplementary Table 2). The analysis revealed that all sequences grouped into two distinct clades corresponding to human- and mosquito-derived samples. However, our root-to-tip regression showed that the clustering observed was not driven by sampling dates, as CHIKV genomes derived from mosquito or human samples presented overlapping evolutionary distances (Fig. 1A). Additionally, it is important to emphasize that, as the phylogenetic analysis was performed using consensus sequences, only nucleotides with allele frequency >50% at each genomic position were considered, as defined by the default settings of the tools employed. Therefore, low-frequency variants were not represented in the tree.

Figure 1.

Figure 1

Open in a new tab

Genetic similarity analysis between human and mosquito samples. (A) Maximum likelihood phylogenetic tree reconstructed using nucleotide consensus sequences obtained from CHIKV-positive human and mosquito samples, highlighting genetic clustering according to host species. (B) PCA based on the mutational profiles of each sequence relative to the reference genome, considering all detected variants regardless of allele frequency.

To address this limitation, we conducted a principal component analysis (PCA) incorporating all mutations detected in each sample, regardless of allele frequency, relative to a reference genome (Supplementary Table 3). The analysis revealed a pattern similar to that of the phylogenetic reconstruction. The PCA also identified two main clusters: one composed exclusively of human-derived samples and another composed of Aedes-derived CHIKV sequences. However, this analysis highlights the role of low-frequency variants in shaping the genomic variability of viral populations. Our results showed that viral genomes obtained from the natural mosquito population exhibited a greater mutational profile, potentially related to selective pressures imposed by specific tissues that act as anatomical barriers, a finding that deserves further investigation (Fig. 1B).

Intrahost viral diversity was assessed using mean nucleotide diversity (π) and the normalized Shannon entropy (Sn). The analysis revealed a statistically significant difference in π (t = 12.37, P < .001, 95% CI: 0.041–0.057), suggesting that CHIKV populations circulating in mosquitoes exhibit greater genetic variability than those infecting humans (Fig. 2A, Supplementary Table 4). Similarly, the Sn was significantly higher in Ae. aegypti-derived CHIKV genomes compared to those found in human samples (t = 13.083, P < .001, 95% CI: 0.04–0.06) (Fig. 2B, Supplementary Table 4).

Figure 2.

Figure 2

Open in a new tab

Intrahost diversity of chikungunya virus. (A) Nucleotide diversity (π) estimates the average number of genetic differences per site between hosts (Ae. aegypti and humans). (B) Shannon entropy index, calculated from the richness and relative frequency distributions of variants within each CHIKV genome derived from Ae. aegypti and human samples. (C) Spearman correlation between mean CHIKV whole-genome sequencing depth and the Shannon entropy index per sample. Colours represent different hosts, and shaded areas represent standard deviation values. (D) Comparison of Shannon entropy index across different CHIKV proteins, according to host species. π and Sn values were averaged across all sites within each sample and each protein-coding gene, generating an individual and per-protein estimate of intrahost genetic complexity.

To investigate whether sequencing depth could influence estimates of intrahost viral diversity, we assessed the correlation between mean sequencing coverage and normalized Sn per sample. The analysis revealed a moderate but statistically significant negative correlation (ρ = − 0.57, P < .001), indicating that higher sequencing depth does not necessarily imply greater detected genetic diversity. These findings suggest that other factors, such as host-specific dynamics or host switching, may play a role in shaping intrahost viral diversity in this dataset (Fig. 2C).

An analysis of intrahost diversity across different CHIKV proteins, stratified by host, showed that Sn differed significantly between Ae. aegypti- and human-derived viruses for all proteins except NSP2 (Fig. 2D, Supplementary Tables 5 and 6). These results indicate that intrahost viral diversity is both protein-specific and host-dependent, underscoring the complexity of CHIKV evolutionary dynamics during natural infections.

Overall, our analysis showed a total of 303 iSNVs across the CHIKV genomes in our sample set. We further investigate whether these mutations were evenly distributed along the viral genome and whether their patterns differed between host types. To explore this, we counted mutations in 500-bp sliding windows. Our analysis revealed that mutations were more frequent in genomic regions encoding nonstructural proteins than in those encoding structural proteins, in human- and mosquito-derived samples (Fig. 3A). The highest concentration of mutations was found between positions 5000 and 5500 bp, which corresponds to the coding region of the NSP3 protein. By comparing the number of mutations per genomic window between Ae. aegypti- and human-derived CHIKV genomes, we found a statistically significant difference (t = 4.287, P = .0002) (Supplementary Table 3). Viral genomes from mosquitoes exhibited, on average, 14 more mutations per sliding window (95% CI, 7.2–20.7) than those from human samples (Supplementary Fig. 2, Supplementary Table 6). The majority of mutations identified in mosquito-derived CHIKV genomes are concentrated between positions 9000 and 10 000 bp and 10 500–11 000 bp, corresponding to the E2, 6 K, and E1 proteins, respectively (Fig. 3A). These results suggest that mutational patterns are not uniformly distributed across the viral genome and may reflect host-specific selective pressures influencing CHIKV evolution.

Figure 3.

Figure 3

Open in a new tab

Intrahost mutational profile of CHIKV in samples derived from Ae. aegypti and humans. (A) Mutation density across the CHIKV genome, estimated using 500-bp sliding windows, comparing genomes from human and mosquito samples. (B) Number of shared mutations (present in two or more samples) and unique mutations (detected in only one sample) across viral proteins, stratified by host and mutation type (synonymous versus nonsynonymous). (C) Allele frequency distribution of shared and unique mutations in human and mosquito samples, categorized by mutation type (synonymous or nonsynonymous). Colours represent either the host species or the mutation classification.

To better characterize the mutational landscape across viral proteins, we classified the 303 observed mutations into two categories: (i) shared mutations, detected in two or more samples, and (ii) unique mutations, detected in only one sample, regardless of host or allele frequency. Of these, 152 were classified as shared and 151 as unique. A clear contrast was observed between the two categories. Among the 152 shared mutations, 75% (n = 114) were synonymous, and 25% (n = 38) were nonsynonymous (X2 = 38, df = 1, P < .001). Conversely, the 151 unique mutations displayed the opposite trend: 65.8% (n = 100) were nonsynonymous, whereas 34.2% (n = 51) were synonymous (X2 = 15.9, df = 1, P < .001). When stratified by host, the distribution of shared mutations was similar between mosquito- and human-derived CHIKV genomes across all viral proteins, with synonymous changes predominating (Fig. 3B). Specifically, we identified 36 and 25 nonsynonymous mutations in mosquito- and human-derived genomes (X2 = 1.98, df = 1, P=.15), respectively, while synonymous mutations accounted for 107 and 86 intrahost single-iSNVs (X2 = 2.28, df = 1, P = .13). In contrast, unique mutations were found in great numbers in Ae. aegypti-derived genomes, and the majority of them were classified as nonsynonymous (Fig. 3B). In total, 93 nonsynonymous mutations were observed in mosquito-derived genomes, compared with only six in human-derived genomes (X2 = 76.45, df = 1, P < .001) (Supplementary Table 3). In mosquito- and human-derived CHIKV genomes, 40 and 11 iSNVs, respectively, classified as synonymous were detected (X2 = 16.49, df = 1, P < .001) (Supplementary Table 3).

When we looked more closely at allele frequency patterns, we found clear, host-specific trends. Among shared mutations, there was a wide range of allele frequencies, with synonymous changes being the most common. CHIKV genomes from mosquito samples exhibited a higher number of shared mutations than human samples, particularly nonsynonymous substitutions within the coding regions of NSP3, NSP4, capsid, and E1 (Fig. 3B). In contrast, shared nonsynonymous mutations in human-derived genomes were mainly restricted to the NSP3 and NSP4 regions (Fig. 3C).

Moreover, in Ae. aegypti-derived CHIKV genomes, unique mutations occurred predominantly at low allele frequencies (<50%), consistent with the idea that mosquito hosts are more permissive to the emergence of new variants. In contrast, human-derived CHIKV genomes showed fewer unique mutations, with a wide allele frequency distribution and a more balanced ratio of synonymous to nonsynonymous changes (Fig. 3C). These distinct patterns suggest different selective pressures or viral population dynamics between the two hosts, potentially driven by differences in immune responses, replication environments, or transmission bottlenecks.

It is noteworthy that the majority of unique mutations arise sporadically during the viral replication cycle, as shown by low allele frequencies (Fig. 3C). In contrast, we observed that shared mutations exhibited a wide range of allele frequencies, reflecting allelic variation patterns that may represent processes of fixation or elimination of variants within the viral population (Fig. 3C).

Thus, to further investigate this pattern, we subdivided the shared mutations into three groups: those restricted to human-derived CHIKV genomes, those limited to mosquito-derived CHIKV genomes, and those detected in CHIKV genomes from both hosts. Among the 152 mutations present in two or more sequences, 9 were exclusive to CHIKV in human samples, 41 were exclusive to CHIKV in mosquito samples, and 102 were common between Ae. aegypti- and human-derived CHIKV genomes (Fig. 4A). Mutations restricted to human hosts were observed at high allele frequencies, predominantly located within nonstructural protein–coding regions, and were mainly synonymous (77.7%, n = 7/9) (Fig. 4B). In contrast, exclusive mutations in mosquito hosts showed low frequency (<50%) and were widely dispersed across nonstructural and structural protein-coding genes, with 68.3% being synonymous (n = 28/41) (Fig. 4B). Among these, one nonsynonymous mutation stands out: K90T, which results in a lysine-to-threonine substitution in the viral capsid (Fig. 4B). This mutation is present in all mosquito-derived samples and exhibits a wide range of allele frequencies, suggesting potential functional relevance that may impact viral replication or vector-mediated transmission.

Figure 4.

Figure 4

Open in a new tab

Allelic distribution of shared CHIKV mutations across viral proteins. (A) Venn diagram showing the number of mutations unique to humans, unique to mosquitoes, and shared between both hosts. (B) Allele frequencies of mutations exclusive to humans (left panel) and mosquitoes (right panel). (C) Allele frequencies of mutations shared between Ae. aegypti- and human-derived CHIKV genomes, shown separately for nonstructural and structural proteins.

Regarding mutations common between human and mosquito-derived CHIKV genomes, the majority were found within nonstructural proteins, of which 78.8% (n = 52/66) were synonymous, and 21.2% (n = 14/66) nonsynonymous (Fig. 4C). Overall, specific patterns in allele frequency shifts were observed across the different groups analysed. As examples, we have the mutations A121E (Alanine → Glutamic acid, nonsynonymous) and V122V (synonymous), both located in NSP1; and the H280H (synonymous) and R374G (Arginine → Glycine, nonsynonymous) located in NSP3, which display high allele frequencies in mosquitoes (>50%) but low frequencies in humans (Fig. 4C). Conversely, mutations such as D277D (synonymous—NSP1), A264A (synonymous—NSP2), A750A (synonymous—NSP2), F129F (synonymous—NSP3), and T204T (synonymous—NSP3) exhibited high allele frequencies in human-derived samples and low frequencies in mosquito-derived samples (Fig. 4C). Similar patterns can be identified in the structural proteins, as mutations such as V222V (synonymous—E2) and C225C (synonymous—E2) showed higher allele frequency in mosquitoes than human-derived CHIKV genomes, whereas the mutations V50V (synonymous—E2) and C68C (synonymous—E3) showed the opposite pattern (Fig. 4C).

In these cases, it is challenging to infer the evolutionary dynamics of such mutations, that is, whether there is a trend toward fixation or loss of these alleles. This uncertainty underscores the need for continuous genomic surveillance of viral variants, particularly during outbreaks, to enable rapid and accurate identification of mutations with potential adaptive effects and to improve our understanding of arbovirus evolutionary patterns and the emergence of novel variants.

Finally, to better understand the selective pressures shaping intrahost CHIKV evolution in different organisms, we applied a combination of two tests to detect signatures of selection across the genome. In human samples, our analyses identified 19 sites under negative selection, of which 68.4% are found in nonstructural proteins, particularly NSP1, NSP2, and NSP3 (Table 1). In mosquito-derived CHIKV genomes, 11 sites under negative selection were detected, of which 90.9% were located in nonstructural proteins (mainly NSP2 and NSP4), as observed in humans (Table 1).

Table 1.

Selection analyses of CHIKV nonstructural proteins in Ae. aegypti- and human-derived genomes

Protein Ae. aegypti Humans
MEME FEL Global
ω (dn/ds estimat es)		 MEME FEL Global ω (dn/ds estimates)
Episodic diversifying positive selection Pervasive positive diversifying selection Negative Selection Episodic diversifying positive selection Pervasive positive diversifying selection Negative Selection
NSP1 0 0 359 0.2138 0 0 429, 430,
446		 0.0881
NSP2 0 0 77, 79,
415, 565,
588		 0.0000 0 0 264, 405,
415, 574,
750		 0.1168
NSP3 0 0 0 1.0674 0 0 209, 295,
346		 0.1092
NSP4 0 0 188, 333,
356, 605		 0.0866 0 0 121, 250 0.1443
Capsid 0 0 0 0.0000 0 0 0 0.3078
E3 0 0 0 0.9251 0 0 29, 37 0.0000
E2 0 0 222 0.0000 0 0 266 0.2441
6 K 0 0 0 0.0000 0 0 38 0.3666
E1 0 0 0 0.1847 0 0 68, 228 0.0000
Open in a new tab

Global ω (dN/dS) estimates are reported for each protein. Episodic diversifying positive selection was assessed using MEME, while pervasive positive diversifying selection and negative (purifying) selection were evaluated using FEL.

Considering the global dN/dS values for each protein-coding gene, most values in both hosts were below 1 (Table 1), indicating purifying selection and suggesting that the majority of mutations, particularly nonsynonymous ones, are deleterious and tend to be eliminated. This trend is especially evident when examining the allele frequencies of unique (sporadic) mutations and exclusive mutations that appeared twice or more times in Ae. aegypti-derived CHIKV genomes, which are overall in low allele frequency (Fig. 3C). However, two proteins in mosquito-derived genomes exhibited dN/dS values equal to or close to 1, indicative of neutral and near-neutral evolution (Table 1). The protein E3 exhibited a dN/dS ratio of 0.92 in mosquitoes, suggesting a balance between nonsynonymous and synonymous substitutions, whereas in humans, the ratio was 0, indicative of strong purifying selection (Table 1). Similarly, NSP3 showed dN/dS values of 1.06 in mosquitoes and 0.10 in humans (Table 1), consistent with neutrality and purifying selection, respectively. Additionally, site-specific selection analysis revealed that NSP3 harboured four negatively selected sites in human-derived sequences but none in mosquito-derived genomes. This pattern suggests that nonsynonymous mutations may persist longer in mosquitoes, potentially due to genetic drift or greater tolerance of structural changes in viral proteins.

Discussion

Arboviruses, unlike most RNA viruses, face unique evolutionary constraints because they must replicate in vertebrate and invertebrate hosts. This dual-host cycle helps maintain genome stability, often through purifying selection, since mutations that improve fitness in one host may be neutral or deleterious in the other (Jerzak et al. 2005, Coffey et al. 2008). Despite these constraints, CHIKV has developed mechanisms for efficient replication in both host types. For example, during the Réunion Island outbreak in 2005–06, a single amino acid substitution changed vector specificity and increased Ae. albopictus’ capacity for viral infection, dissemination, and transmission (Schuffenecker et al. 2006, Tsetsarkin et al. 2007).

Considering the complex factors that shape arbovirus evolution, we applied deep sequencing to investigate the intrahost genetic variability of CHIKV in naturally infected humans and Ae. aegypti mosquitoes during an outbreak in Brazil. Our analyses showed pronounced differences in the frequency and distribution of iSNVs between hosts, with mosquito-derived viral populations exhibiting greater genetic diversity. This pattern was supported by analyses of nucleotide diversity and mean normalized Shannon entropy. On average, CHIKV genomes from mosquitoes showed ~14 more iSNVs per 500-bp genomic sliding window than those from humans. In fact, similar trends have been observed for West Nile virus (WNV), in which viral populations infecting invertebrate hosts exhibit greater intrahost diversity than those infecting vertebrates, reinforcing the role of host biology in shaping viral genetic variability (Jerzak et al. 2005). Additionally, these findings are consistent with previous reports showing that mosquitoes can serve as reservoirs of genetic diversity, mainly due to the accumulation of low-frequency variants driven by stochastic processes such as genetic drift (Forrester et al. 2014, Grubaugh et al. 2017). Corroborating these reports, in our dataset, CHIKV genomes from Ae. aegypti harboured significantly more unique mutations, predominantly nonsynonymous and occurring at low frequencies. Similarly, studies using WNV as a model showed that mosquito-expectorated virus populations were highly diverse and unique to each feeding episode (Grubaugh et al. 2017).

The biological implications of these patterns are closely related to the vector biology. Following a blood meal, arbovirus transmission requires sequential infection of key anatomical compartments. Four major barriers have been described: (i) midgut infection, involving viral entry and replication in epithelial cells; (ii) midgut escape, with dissemination through the basal lamina into the haemolymph; (iii) salivary gland infection, requiring viral access to and replication within salivary tissues; and (iv) salivary gland escape, culminating in viral release into the saliva (Forrester et al. 2014, Franz et al. 2015, Rückert and Ebel 2018, Merwaiss et al. 2021, Weaver et al. 2021). These sequential bottlenecks, together with subsequent increase of diversity via mutation and demographic expansion, can dynamically reshape viral diversity (Forrester et al. 2014, Weaver et al. 2021), allowing nonsynonymous mutations with near-deleterious effects to persist at low frequencies through genetic drift, as observed in our results. In addition, host-specific selective pressures also contribute to this process. Mosquito antiviral immune responses, particularly those mediated by RNA interference (RNAi) (Brackney et al. 2009, Brackney et al. 2015, Jupatanakul et al. 2017, Ahlers and Goodman 2018), together with weak purifying selection (Jerzak et al. 2005, 2008, Grubaugh et al. 2016, Lequime et al. 2016), can modulate viral replication and shape the emergence and persistence of new iSNVs in mosquito hosts.

While these findings highlight general differences in viral population dynamics between hosts, the distribution of genetic diversity across the CHIKV genome provides additional insight into host-specific selective pressures. Detailed analyses revealed protein-specific and host-dependent differences between human- and mosquito-derived CHIKV for all proteins except NSP2. NSP2 is a multifunctional protein with critical roles in the viral life cycle, including evasion of host antiviral responses, as well as helicase, triphosphatase, and protease activities (Rupp et al. 2015, Wong and Chu 2018). Its protease activity, in particular, is responsible for processing the viral polyprotein into the individual nonstructural proteins NSP1–NSP4, a step essential for RNA replication and subgenomic RNA transcription (Akhrymuk et al. 2012). These indispensable functions likely impose strong evolutionary constraints on NSP2, limiting its tolerance to variation in both hosts. Consequently, most nucleotide substitutions, especially nonsynonymous changes, are expected to disrupt its activity, explaining the reduced diversity observed in this region compared with other viral proteins. In fact, some experimental studies have shown that mutations in NSP2 lead to a reduction in CHIKV replication capacity (Utt et al. 2015, Meshram et al. 2019). Such functional constraints can affect replication efficiency and adaptability, which may explain the lower diversity observed in this region compared with other CHIKV proteins in our study.

In contrast, other nonstructural proteins showed more mutations than structural regions in both hosts, as observed for NSP3, despite the majority of the changes being classified as synonymous. These results are similar to those reported in previous studies, which showed that nonstructural proteins, particularly NSP3, exhibit high genetic variability, reflecting their multifunctional roles in replication and host interactions (Mathur et al. 2016, Stapleford et al. 2016, Mohamed Ali et al. 2018, Wong and Chu 2018).

Our results showed significant host-specific differences in structural proteins, with mosquito-derived viruses exhibiting substantially more mutations in the capsid, E2, 6 K, and E1 regions than human-derived viruses. Interestingly, the high density of mutations observed in CHIKV genomes from mosquito samples, particularly nonsynonymous changes, occurred predominantly at allele frequencies below 50%. These likely represent newly generated neutral or slightly deleterious variants, as reported in other studies (Stapleford et al. 2016). In addition, numerous nonsynonymous mutations were detected within the capsid and E1 coding regions in mosquito-derived samples. The capsid protein is essential for assembling the viral nucleocapsid and protecting the viral RNA, whereas the E1 glycoprotein, together with E2, mediates viral entry into host cells (Solignat et al. 2009, Wong and Chu 2018). As these glycoproteins play essential roles in viral entry, the replication cycle, and dissemination, and are key targets for host immune responses, it is expected that they exhibit distinct diversity patterns driven by selective pressures specific to each host. The higher density of nonsynonymous mutations in these regions in mosquito samples suggests an association with genetic bottlenecks imposed by the vector, including anatomical and immunological barriers, which the virus must overcome for successful transmission. Similar findings have been reported in other studies, which identified these proteins as hotspots of both synonymous and nonsynonymous mutations, particularly in mosquito-derived viruses (Solignat et al. 2009, Mohamed Ali et al. 2018).

In contrast, human infections tend to impose stronger purifying selection, which is related to innate and adaptive immune responses (Ahlers and Goodman 2018, Rückert and Ebel 2018), restricting the fixation of nonsynonymous mutations and thereby limiting overall intrahost diversity, as observed for other vertebrate hosts (Jerzak et al. 2008, Deardorff et al. 2011, Grubaugh et al. 2017, Parameswaran et al. 2017). Overall, our results confirm that intrahost CHIKV diversity is shaped by distinct selective pressures in mosquitoes and humans, consistent with the dual-host life cycle of arboviruses (Coffey et al. 2008, Deardorff et al. 2011). The influence of host-specific selective pressures became even more evident when we considered only iSNVs present in more than two sequences (shared mutations). For these mutations, we identified clear patterns reflecting the biological constraints of each host environment.

Mutations exclusive to human-derived CHIKV genomes were rare, often showed high allele frequencies, and were predominantly synonymous, concentrated in nonstructural protein-coding regions. This pattern is consistent with strong purifying selection in vertebrate hosts, where immune pressure and functional constraints limit the fixation of nonsynonymous changes, thereby preserving genome stability and minimizing deleterious effects on viral fitness (Jerzak et al. 2005, Coffey et al. 2008, Deardorff et al. 2011, Ahlers and Goodman 2018). In contrast, mutations exclusive to mosquito-derived CHIKV genomes were, overall, at low allele frequencies and distributed across both structural and nonstructural coding regions. These findings support the idea that replication in mosquito vectors allows greater genetic diversity, which is facilitated by population bottlenecks and relaxed purifying selection (Jerzak et al. 2005, 2008, Forrester et al. 2012, Grubaugh et al. 2016, 2017, Lequime et al. 2016, Weaver et al. 2021). Nonetheless, most shared mutations were present in both vertebrate and invertebrate hosts. Interestingly, we observed a distinct range of allele frequencies, with some mutations common in Ae. aegypti-derived CHIKV but rare in humans, and vice versa, reinforcing that the forces shaping intrahost variability are host-specific.

Selection analyses further supported these findings. Although purifying selection predominated in both hosts, its strength and distribution varied across hosts and proteins. In mosquito-derived genomes, NSP3 and E3 exhibited patterns consistent with neutrality or relaxed purifying selection, suggesting a greater tolerance for amino acid substitutions, as also reported in other studies (Mohamed Ali et al. 2018), whereas in humans, these proteins were under stronger constraint. The contrasting dN/dS values for NSP3 and E3 between mosquitoes and humans suggest that in mosquitoes, viral populations experience weaker purifying selection, allowing persistence or accumulation of nonsynonymous changes, as already observed in other studies (Jerzak et al. 2005, 2008, Deardorff et al. 2011, Lequime et al. 2016). Conversely, stronger negative selection in humans acts to eliminate or prevent fixation of such mutations, as also reported in other arbovirus studies across different hosts (Jerzak et al. 2005, Deardorff et al. 2011). These observations indicate that the strength of purifying selection differs between hosts, likely reflecting differences in viral population dynamics, immune pressures, and transmission bottlenecks in mosquitoes.

In summary, our results from gene-wide and site-specific analyses showed that purifying selection predominates in CHIKV genomes derived from naturally infected humans and mosquitoes, possibly acting to preserve genomic integrity by minimizing the accumulation of deleterious mutations, particularly in functionally important regions. However, for some proteins, such as NSP3 and E3, this pattern varied depending on the host. Specifically, in mosquitoes, NSP3 and E3 showed patterns of neutrality or relaxed purifying selection, indicating increased tolerance to nucleotide substitutions, whereas in humans, these proteins were under stronger constraint. These findings suggest that, despite purifying selection tending to preserve CHIKV genome integrity in both hosts, the replication process in mosquitoes may contribute to greater genetic variability. Thus, for CHIKV, host biology drives divergent evolutionary trajectories, as mosquitoes act as reservoirs of genetic diversity, possibly through genetic drift and relaxed purifying selection, while vertebrate hosts may impose strong immune-mediated constraints that tend to stabilize the virus genome across transmission cycles. This host-dependent balance between diversification and constraint is central to CHIKV evolution and adaptation.

Supplementary Material

Supplementary_figures_1_and_2_(1)_veaf102
supplementary_figures_1_and_2_(1)_veaf102.pdf (560KB, pdf)
Supplementary_tables_Virus_Evolution_(2)_veaf102
supplementary_tables_virus_evolution_(2)_veaf102.xlsx (779.1KB, xlsx)

Acknowledgements

We wish to thank all our colleagues at the Hospital de Base de São José do Rio Preto for their support during CHIKV sample collection. We are also grateful to the Multiuser Laboratory (LMU) at the Faculdade de Medicina de São José do Rio Preto (FAMERP) and the Instituto de Biotecnologia at São Paulo State University (UNESP) in Botucatu, Brazil, for allowing us to use the Illumina MiSeq system. Grammarly AI was used only for language editing: grammar, spelling, and punctuation. The tool did not generate, modify, or interpret any scientific content. All analyses, interpretations, and conclusions are the authors’ own.

Contributor Information

Cecília Artico Banho, Laboratório de Pesquisas em Virologia, Faculdade de Medicina de São José do Rio Preto, Avenida Brigadeiro Faria Lima, 5416, Vila São Pedro, 15090-000, São José do Rio Preto, São Paulo, Brazil.

Beatriz de Carvalho Marques, Laboratório de Pesquisas em Virologia, Faculdade de Medicina de São José do Rio Preto, Avenida Brigadeiro Faria Lima, 5416, Vila São Pedro, 15090-000, São José do Rio Preto, São Paulo, Brazil.

Olivia Borghi Nascimento, Laboratório de Pesquisas em Virologia, Faculdade de Medicina de São José do Rio Preto, Avenida Brigadeiro Faria Lima, 5416, Vila São Pedro, 15090-000, São José do Rio Preto, São Paulo, Brazil.

Maisa Carla Pereira Parra, Laboratório de Pesquisas em Virologia, Faculdade de Medicina de São José do Rio Preto, Avenida Brigadeiro Faria Lima, 5416, Vila São Pedro, 15090-000, São José do Rio Preto, São Paulo, Brazil.

Maria Vitória Moraes Ferreira, Laboratório de Pesquisas em Virologia, Faculdade de Medicina de São José do Rio Preto, Avenida Brigadeiro Faria Lima, 5416, Vila São Pedro, 15090-000, São José do Rio Preto, São Paulo, Brazil.

Ana Paula Lemos, Laboratório de Pesquisas em Virologia, Faculdade de Medicina de São José do Rio Preto, Avenida Brigadeiro Faria Lima, 5416, Vila São Pedro, 15090-000, São José do Rio Preto, São Paulo, Brazil.

Gabriel Pires Magnani, Laboratório de Pesquisas em Virologia, Faculdade de Medicina de São José do Rio Preto, Avenida Brigadeiro Faria Lima, 5416, Vila São Pedro, 15090-000, São José do Rio Preto, São Paulo, Brazil.

Karine Lima Lourenço, Laboratório de Pesquisas em Virologia, Faculdade de Medicina de São José do Rio Preto, Avenida Brigadeiro Faria Lima, 5416, Vila São Pedro, 15090-000, São José do Rio Preto, São Paulo, Brazil.

Beatriz Cunha de Souza, Laboratório de Pesquisas em Virologia, Faculdade de Medicina de São José do Rio Preto, Avenida Brigadeiro Faria Lima, 5416, Vila São Pedro, 15090-000, São José do Rio Preto, São Paulo, Brazil.

Victor Miranda Hernandes, Laboratório de Pesquisas em Virologia, Faculdade de Medicina de São José do Rio Preto, Avenida Brigadeiro Faria Lima, 5416, Vila São Pedro, 15090-000, São José do Rio Preto, São Paulo, Brazil.

Marini Lino Brancini, Laboratório de Pesquisas em Virologia, Faculdade de Medicina de São José do Rio Preto, Avenida Brigadeiro Faria Lima, 5416, Vila São Pedro, 15090-000, São José do Rio Preto, São Paulo, Brazil.

Cassia Fernanda Estofolete, Laboratório de Pesquisas em Virologia, Faculdade de Medicina de São José do Rio Preto, Avenida Brigadeiro Faria Lima, 5416, Vila São Pedro, 15090-000, São José do Rio Preto, São Paulo, Brazil; Hospital de Base de São José do Rio Preto, Avenida Brigadeiro Faria Lima, 5544, Vila São Pedro, 15090-000, São José do Rio Preto, São Paulo, Brazil.

João Pessoa Araújo Júnior, Institute of Biotechnology, São Paulo State University (Unesp), Alameda das Tecomarias s/n, Chácara Capão Bonito, 18607-440, Botucatu, São Paulo, Brazil.

Nikos Vasilakis, Department of Pathology, University of Texas Medical Branch, 301 University Boulevard, Galveston, 77555, Texas, United States of America; Center for Vector-Borne and Zoonotic Diseases, University of Texas Medical Branch, 301 University Boulevard, Galveston, 77555, Texas, United States of America; Institute for Human Infection and Immunity, University of Texas Medical Branch, Galveston, 301 University Boulevard, Galveston, 77555, Texas, United States of America.

Maurício Lacerda Nogueira, Laboratório de Pesquisas em Virologia, Faculdade de Medicina de São José do Rio Preto, Avenida Brigadeiro Faria Lima, 5416, Vila São Pedro, 15090-000, São José do Rio Preto, São Paulo, Brazil; Hospital de Base de São José do Rio Preto, Avenida Brigadeiro Faria Lima, 5544, Vila São Pedro, 15090-000, São José do Rio Preto, São Paulo, Brazil; Department of Pathology, University of Texas Medical Branch, 301 University Boulevard, Galveston, 77555, Texas, United States of America.

Author contributions

C.A.B. and M.L.N. conceived and designed the study. C.A.B., M.C.P.P., O.B.N., G.P.M., M.V.M.F., A.P.L., K.L.L., B.C.S., and V.M.H. collected and identified mosquito samples. C.A.B., O.B.N., M.V.M.F., A.P.L., and B.C.S. processed and screened mosquito samples. M.L.B. collected and screened human samples. C.A.B., O.B.N., M.V.M.F., and A.P. curated the metadata and performed molecular screening in mosquito samples. C.F.E. curated metadata for the human samples. J.P.A.J. provided materials and equipment for the sequencing of CHIKV-positive samples. C.A.B. and B.C.M. performed sequencing. C.A.B. performed data analyses and interpreted the data. C.A.B. and B.C.M. wrote the first draft of the manuscript. C.A.B., N.V., and M.L.N. edited and revised the manuscript. N.V. and M.L.N. provided the resources for the survey. All authors approved the final version of this manuscript.

Conflict of interest: None declared.

Funding

This work received support from the Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP, grant numbers 2022/03645-1 to M.L.N., 2023/14670-0 to C.A.B., and 2022/09229-0 to C.F.E.). This study was partly funded by the by the INCT Viral Genomic Surveillance and One Health grant 405786/2022-0, and partly by the Centers for Research in Emerging Infectious Diseases (CREID), via the ‘Coordinating Research on Emerging Arboviral Threats Encompassing the Neotropics (CREATE-NEO)’ grant U01AI151807 awarded to N.V. by the National Institutes of Health (NIH/USA). M.L.N. is a CNPq Research Fellow. The funders had no role in the study design, data collection, analysis, interpretation, manuscript writing, or the decision to publish the results.

Data availability

All the chikungunya raw data (fastq files) generated and analyzed in this study are available in the SRA-NCBI database (www.ncbi.nlm.nih.gov/sra), under accession number PRJNA1338711. Chikungunya consensus sequences used for phylogenetic analysis are fully available in the GISAID EpiArbo database (https://gisaid.org/), with all accession numbers are listed in Supplementary Table 2. Code and raw data are also available in the GitHub Repository (https://github.com/cab1992/Intra-host-population-dynamics-of-chikungunya-virus-in-humans-and-Aedes-aegypti-), and the Mendeley Data Repository (https://data.mendeley.com/datasets/pr5vsfcf5f/1).

References

  1. Ahlers  LRH, Goodman  AG. The immune responses of the animal hosts of West Nile virus: a comparison of insects, birds, and mammals. Front Cell Infect Microbiol  2018;8:1–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Akhrymuk  I, Kulemzin  SV, Frolova  EI. Evasion of the innate immune response: the old world alphavirus nsP2 protein induces rapid degradation of Rpb1, a catalytic subunit of RNA polymerase II. J Virol  2012;86:7180–91. 10.1128/JVI.00541-12 [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Andrews  S. Babraham bioinformatics - FastQC a quality control tool for high throughput sequence data. Available at: http://www.bioinformatics.babraham.ac.uk/projects/fastqc/. Accessed on January 05, 2010. [Google Scholar]
  4. Angleró-Rodríguez  YI, MacLeod  HJ, Kang  S  et al.  Aedes aegypti molecular responses to zika virus: modulation of infection by the toll and jak/Stat immune pathways and virus host factors. Front Microbiol  2017;8:1–12. 10.3389/fmicb.2017.02050 [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Anisimova  M, Gascuel  O. Approximate likelihood-ratio test for branches: a fast, accurate, and powerful alternative. Syst Biol  2006;55:539–52. 10.1080/10635150600755453 [DOI] [PubMed] [Google Scholar]
  6. Banho  CA, Parra  MCP, Nascimento  OB  et al.  Vector surveillance during a major chikungunya outbreak in northwestern São Paulo state. Brazil npj Viruses  2025;3:78. 10.1038/s44298-025-00161-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Blair  CD. Mosquito RNAi is the major innate immune pathway controlling arbovirus infection and transmission. Future Microbiol  2011;6:265–77. 10.2217/fmb.11.11 [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Brackney  DE, Beane  JE, Ebel  GD. RNAi targeting of West Nile virus in mosquito midguts promotes virus diversification. PLoS Pathog  2009;5:1–9. 10.1371/journal.ppat.1000502 [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Brackney  DE, Schirtzinger  EE, Harrison  TD  et al.  Modulation of flavivirus population diversity by RNA interference. J Virol  2015;89:4035–9. 10.1128/JVI.02612-14 [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Caldwell  HS, Lasek-Nesselquist  E, Follano  P  et al.  Divergent mutational landscapes of consensus and minority genotypes of West Nile virus demonstrate host and gene-specific evolutionary pressures. Genes (Basel)  2020;11:1–18. 10.3390/genes11111299 [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Castellano  S, Cestari  F, Faglioni  G  et al.  Ivar, an interpretation-oriented tool to manage the update and revision of variant annotation and classification. Genes (Basel)  2021;12:1–8. 10.3390/genes12030384 [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Coffey  LL, Reisen  WK. West Nile virus fitness costs in different mosquito species. Trends Microbiol  2016;24:429–30. 10.1016/j.tim.2016.04.005 [DOI] [PubMed] [Google Scholar]
  13. Coffey  LL, Vignuzzi  M. Host alternation of chikungunya virus increases fitness while restricting population diversity and adaptability to novel selective pressures. J Virol  2011;85:1025–35. 10.1128/JVI.01918-10 [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Coffey  LL, Vasilakis  N, Brault  AC  et al.  Arbovirus evolution in vivo is constrained by host alternation. Proc Natl Acad Sci USA  2008;105:6970–5. 10.1073/pnas.0712130105 [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Consoli  RAGB, de  Oliveira  RL. Principais Mosquitos de Importância Sanitária No Brasil Editora FIOCRUZ. Rio de Janeiro: Editora FIOCRUZ, 1994. [Google Scholar]
  16. Deardorff  ER, Fitzpatrick  KA, Jerzak  GVS  et al.  West Nile virus experimental evolution in vivo and the trade-off hypothesis. PLoS Pathog  2011;7:e1002335. 10.1371/journal.ppat.1002335 [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Domingo  E. Mechanisms of viral emergence. Vet Res  2010;41:38. 10.1051/vetres/2010010 [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Domingo  E, Perales  C. Viral quasispecies. PLoS Genet  2019;15:e1008271. 10.1371/journal.pgen.1008271 [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Domingo  E, García-Crespo  C, Lobo-Vega  R  et al.  Mutation rates, mutation frequencies, and proofreading-repair activities in RNA virus genetics. Viruses  2021;13:1882. 10.3390/v13091882 [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Duffy  S. Why are RNA virus mutation rates so damn high?  PLoS Biol  2018;16:e3000003. 10.1371/journal.pbio.3000003 [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Forattini OP. Culicidologia Médica: Princípios Gerais, Morfologia, Glossário Taxonômico. 1st edn. São Paulo: Editora da Universidade de São Paulo (EDUSP), 1996.
  22. Fitzmeyer  EA, Gallichotte  EN, Weger-Lucarelli  J  et al.  Loss of West Nile virus genetic diversity during mosquito infection due to species-dependent population bottlenecks. iScience  2023;26:107711. 10.1016/j.isci.2023.107711 [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Forrester  NL, Guerbois  M, Seymour  RL  et al.  Vector-borne transmission imposes a severe bottleneck on an RNA virus population. PLoS Pathog  2012;8:e1002897. 10.1371/journal.ppat.1002897 [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Forrester  NL, Coffey  LL, Weaver  SC. Arboviral bottlenecks and challenges to maintaining diversity and fitness during mosquito transmission. Viruses  2014;6:3991–4004. 10.3390/v6103991 [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Franz  AWE, Kantor  AM, Passarelli  AL  et al.  Tissue barriers to arbovirus infection in mosquitoes. Viruses  2015;7:3741–67. 10.3390/v7072795 [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Grubaugh  ND, Weger-Lucarelli  J, Murrieta  RA  et al.  Genetic drift during systemic arbovirus infection of mosquito vectors leads to decreased relative fitness during host switching. Cell Host Microbe  2016;19:481–92. 10.1016/j.chom.2016.03.002 [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Grubaugh  ND, Fauver  JR, Rückert  C  et al.  Mosquitoes transmit unique West Nile virus populations during each feeding episode. Cell Rep  2017;19:709–18. 10.1016/j.celrep.2017.03.076 [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Gutiérrez  S, Michalakis  Y, Blanc  S. Virus population bottlenecks during within-host progression and host-to-host transmission. Curr Opin Virol  2012;2:546–55. 10.1016/j.coviro.2012.08.001 [DOI] [PubMed] [Google Scholar]
  29. Holland  JJ (ed.). Genetic diversity of RNA viruses. Current Topics in Microbiology and Immunology, Volume 176. Berlin, Heidelberg: Springer; 1992. [PubMed] [Google Scholar]
  30. Jerzak  G, Bernard  KA, Kramer  LD  et al.  Genetic variation in West Nile virus from naturally infected mosquitoes and birds suggests quasispecies structure and strong purifying selection. J Gen Virol  2005;86:2175–83. 10.1099/vir.0.81015-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Jerzak  GVS, Brown  I, Shi  PY  et al.  Genetic diversity and purifying selection in West Nile virus populations are maintained during host switching. Virology  2008;374:256–60. 10.1016/j.virol.2008.02.032 [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Joos  B, Trkola  A, Fischer  M  et al.  Low human immunodeficiency virus envelope diversity correlates with low In vitro replication capacity and predicts spontaneous control of plasma viremia after treatment interruptions. J Virol  2005;79:9026–37. 10.1128/JVI.79.14.9026-9037.2005 [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Joseph  R, Marais  G, Iranzadeh  I  et al.  Intra-host SARS-CoV-2 diversity in immunocompromised people living with HIV provides insight into the evolutionary trajectory of SARS-CoV-2. J Virol  2025;99:e0078025. 10.1128/jvi.00780-25 [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Jupatanakul  N, Sim  S, Angleró-Rodríguez  YI  et al.  Engineered Aedes aegypti JAK/STAT pathway-mediated immunity to dengue virus. PLoS Negl Trop Dis  2017;11:e0005187. 10.1371/journal.pntd.0005187 [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Kalyaanamoorthy  S, Minh  BQ, Wong  TKF  et al.  ModelFinder: fast model selection for accurate phylogenetic estimates. Nat Methods  2017;14:587–9. 10.1038/nmeth.4285 [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Kassambara  A. Rstatix: Pipe-Friendly Framework for Basic Statistical Tests. R Package Version 0.7.2  2023. Available at: https://rpkgs.datanovia.com/rstatix/. Accessed on January 05, 2026.
  37. Katoh  K, Standley  DM. MAFFT multiple sequence alignment software version 7: improvements in performance and usability. Mol Biol Evol  2013;30:772–80. 10.1093/molbev/mst010 [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Kosakovsky Pond  SL, Frost  SDW. Not so different after all: a comparison of methods for detecting amino acid sites under selection. Mol Biol Evol  2005;22:1208–22. 10.1093/molbev/msi105 [DOI] [PubMed] [Google Scholar]
  39. Kosakovsky Pond  SL, Poon  AFY, Velazquez  R  et al.  HyPhy 2.5 - a customizable platform for evolutionary hypothesis testing using phylogenies. Mol Biol Evol  2020;37:295–9. 10.1093/molbev/msz197 [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Lanciotti  RS, Kosoy  OL, Laven  JJ  et al.  Chikungunya virus in US travelers returning from India, 2006. Emerg Infect Dis  2007;13:764–7. 10.3201/eid1305.070015 [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Larsson  A. AliView: a fast and lightweight alignment viewer and editor for large datasets. Bioinformatics  2014;30:3276–78. 10.1093/bioinformatics/btu531 [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Lee  WS, Webster  JA, Madzokere  ET  et al.  Mosquito antiviral defense mechanisms: a delicate balance between innate immunity and persistent viral infection. Parasit Vectors  2019;12:165. 10.1186/s13071-019-3433-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Lequime  S, Fontaine  A, Ar Gouilh  M  et al.  Genetic drift, purifying selection and vector genotype shape dengue virus intra-host genetic diversity in mosquitoes. PLoS Genet  2016;12:e1006111. 10.1371/journal.pgen.1006111 [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Li  H, Durbin  R. Fast and accurate short read alignment with burrows-wheeler transform. Bioinformatics  2009;25:1754–60. 10.1093/bioinformatics/btp324 [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Li  H, Handsaker  B, Wysoker  A  et al.  The sequence alignment/map format and SAMtools. Bioinformatics  2009;25:2078–79. 10.1093/bioinformatics/btp352 [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Liang  Y. Pathogenicity and virulence of influenza. Virulence  2023;14:2223057. 10.1080/21505594.2023.2223057 [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Lustig  G, Ganga  Y, Rodel  HE  et al.  SARS-CoV-2 infection in immunosuppression evolves sub-lineages which independently accumulate neutralization escape mutations. Virus Evol  2024;10:1–12. 10.1093/ve/vead075 [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. MacHado  DC, Mondini  A, Dos Santos  SV  et al.  First identification of Culex flavivirus (Flaviviridae) in Brazil. Intervirology  2012;55:475–83. 10.1159/000337166 [DOI] [PubMed] [Google Scholar]
  49. Maciel-de-Freitas  R, Lourenço-de-Oliveira  R. Presumed unconstrained dispersal of Aedes aegypti in the city of Rio de Janeiro. Brazil Rev Saude Publica  2009;43:8–12. 10.1590/s0034-89102009000100002 [DOI] [PubMed] [Google Scholar]
  50. Marques  AD, Graham-Wooten  J, Fitzgerald  AS  et al.  SARS-CoV-2 evolution during prolonged infection in immunocompromised patients. MBio  2024;15:1–18. 10.1128/mbio.00110-24 [DOI] [PMC free article] [PubMed] [Google Scholar]
  51. Martin  M. Cutadapt removes adapter sequences from high-throughput sequencing reads. EMBnetjournal  2011;17:10–12. 10.14806/ej.17.1.200 [DOI] [Google Scholar]
  52. Mathur  K, Anand  A, Dubey  SK  et al.  Analysis of chikungunya virus proteins reveals that non-structural proteins nsP2 and nsP3 exhibit RNA interference (RNAi) suppressor activity. Sci Rep  2016;6:1–12. 10.1038/srep38065 [DOI] [PMC free article] [PubMed] [Google Scholar]
  53. McCrone  JT, Woods  RJ, Martin  ET  et al.  Stochastic processes constrain the within and between host evolution of influenza virus. elife  2018;7:1–19. 10.7554/eLife.35962 [DOI] [PMC free article] [PubMed] [Google Scholar]
  54. Merwaiss  F, Filomatori  CV, Susuki  Y  et al.  Chikungunya virus replication rate determines the capacity of crossing tissue barriers in mosquitoes. J Virol  2021;95:1–14. 10.1128/JVI.01956-20 [DOI] [PMC free article] [PubMed] [Google Scholar]
  55. Meshram  CD, Lukash  T, Phillips  AT  et al.  Lack of nsP2-specific nuclear functions attenuates chikungunya virus replication both in vitro and in vivo. Virology  2019;534:14–24. 10.1016/j.virol.2019.05.016 [DOI] [PMC free article] [PubMed] [Google Scholar]
  56. Minh  BQ, Nguyen  MAT, Von Haeseler  A. Ultrafast approximation for phylogenetic bootstrap. Mol Biol Evol  2013;30:1188–95. 10.1093/molbev/mst024 [DOI] [PMC free article] [PubMed] [Google Scholar]
  57. Minh  BQ, Schmidt  HA, Chernomor  O  et al.  IQ-TREE 2: new models and efficient methods for phylogenetic inference in the genomic era. Mol Biol Evol  2020;37:1530–34. 10.1093/molbev/msaa015  Erratum in: Mol Biol Evol. 2020; 37. 10.1093/molbev/msaa131 [DOI] [PMC free article] [PubMed] [Google Scholar]
  58. Mohamed Ali  S, Amroun  A, de  Lamballerie  X  et al.  Evolution of chikungunya virus in mosquito cells. Sci Rep  2018;8:1–12. 10.1038/s41598-018-34561-x [DOI] [PMC free article] [PubMed] [Google Scholar]
  59. Moore  TC, Brown  HE. Estimating Aedes aegypti (Diptera: Culicidae) flight distance: Meta-data analysis. J Med Entomol  2022;59:1164–70. 10.1093/jme/tjac070 [DOI] [PubMed] [Google Scholar]
  60. Murrell  B, Wertheim  JO, Moola  S  et al.  Detecting individual sites subject to episodic diversifying selection. PLoS Genet  2012;8:e1002764. 10.1371/journal.pgen.1002764 [DOI] [PMC free article] [PubMed] [Google Scholar]
  61. Nelemans  T, Kikkert  M. Viral innate immune evasion and the pathogenesis of emerging RNA virus infections. Viruses  2019;11:961. 10.3390/v11100961 [DOI] [PMC free article] [PubMed] [Google Scholar]
  62. Parameswaran  P, Wang  C, Trivedi  SB  et al.  Intrahost selection pressures drive rapid dengue virus microevolution in acute human infections. Cell Host Microbe  2017;22:400–410.e5. 10.1016/j.chom.2017.08.003 [DOI] [PMC free article] [PubMed] [Google Scholar]
  63. Prince  BC, Walsh  E, Torres  TZB  et al.  Recognition of arboviruses by the mosquito immune system. Biomolecules  2023;13:1–19. 10.3390/biom13071159 [DOI] [PMC free article] [PubMed] [Google Scholar]
  64. Quick  J, Grubaugh  ND, Pullan  ST  et al.  Multiplex PCR method for MinION and Illumina sequencing of zika and other virus genomes directly from clinical samples. Nat Protoc  2017;12:1261–76. 10.1038/nprot.2017.066 [DOI] [PMC free article] [PubMed] [Google Scholar]
  65. R core team . R: A Language and Environment for Statistical Computing. R Found Stat Comput Vienna, 2021. Available at: https://cran.r-project.org/mirrors.html. Accessed on January 05, 2026. [Google Scholar]
  66. Rambaut  A, Lam  TT, Max Carvalho  L  et al.  Exploring the temporal structure of heterochronous sequences using TempEst (formerly path-O-gen). Virus Evol  2016;2:1–7. 10.1093/ve/vew007 [DOI] [PMC free article] [PubMed] [Google Scholar]
  67. Reiter  P, Amador  MA, Anderson  RA  et al.  Short report: dispersal of Aedes aegypti in an urban area after blood feeding as demonstrated by rubidium-marked eggs. Am J Trop Med Hyg  1995;52:177–9. 10.4269/ajtmh.1995.52.177 [DOI] [PubMed] [Google Scholar]
  68. Rückert  C, Ebel  GD. How do virus–mosquito interactions lead to viral emergence?  Trends Parasitol  2018;34:310–21. 10.1016/j.pt.2017.12.004 [DOI] [PMC free article] [PubMed] [Google Scholar]
  69. Rupp  JC, Sokoloski  KJ, Gebhart  NN  et al.  Alphavirus RNA synthesis and non-structural protein functions. J Gen Virol  2015;96:2483–500. 10.1099/jgv.0.000249 [DOI] [PMC free article] [PubMed] [Google Scholar]
  70. Sagulenko  P, Puller  V, Neher  RA. TreeTime: maximum-likelihood phylodynamic analysis. Virus Evol  2018;4:1–9. 10.1093/ve/vex042 [DOI] [PMC free article] [PubMed] [Google Scholar]
  71. Sanjuán  R, Nebot  MR, Chirico  N  et al.  Viral mutation rates. J Virol  2010;84:9733–48. 10.1128/JVI.00694-10 [DOI] [PMC free article] [PubMed] [Google Scholar]
  72. Schuffenecker  I, Iteman  I, Michault  A  et al.  Genome microevolution of chikungunya viruses causing the Indian Ocean outbreak. PLoS Med  2006;3:1058–70. 10.1371/journal.pmed.0030263 [DOI] [PMC free article] [PubMed] [Google Scholar]
  73. Sessions  OM, Wilm  A, Kamaraj  US  et al.  Analysis of dengue virus genetic diversity during human and mosquito infection reveals genetic constraints. PLoS Negl Trop Dis  2015;9:1–21. 10.1371/journal.pntd.0004044 [DOI] [PMC free article] [PubMed] [Google Scholar]
  74. Shi  M, Lin  XD, Tian  JH  et al.  Redefining the invertebrate RNA virosphere. Nature  2016;540:539–43. 10.1038/nature20167 [DOI] [PubMed] [Google Scholar]
  75. Simões  ML, Caragata  EP, Dimopoulos  G. Diverse host and restriction factors regulate mosquito–pathogen interactions. Trends Parasitol  2018;34:603–16. 10.1016/j.pt.2018.04.011 [DOI] [PubMed] [Google Scholar]
  76. Solignat  M, Gay  B, Higgs  S  et al.  Replication cycle of chikungunya: a re-emerging arbovirus. Virology  2009;393:183–97. [DOI] [PMC free article] [PubMed] [Google Scholar]
  77. Stapleford  KA, Coffey  LL, Lay  S  et al.  Emergence and transmission of arbovirus evolutionary intermediates with epidemic potential. Cell Host Microbe  2014;15:706–16. 10.1016/j.chom.2014.05.008 [DOI] [PubMed] [Google Scholar]
  78. Stapleford  KA, Moratorio  G, Henningsson  R  et al.  Whole-genome sequencing analysis from the chikungunya virus Caribbean outbreak reveals novel evolutionary genomic elements. PLoS Negl Trop Dis  2016;10:1–18. 10.1371/journal.pntd.0004402 [DOI] [PMC free article] [PubMed] [Google Scholar]
  79. Tikhe  CV, Dimopoulos  G. Mosquito antiviral immune pathways. Dev Comp Immunol  2021;116:103964. 10.1016/j.dci.2020.103964 [DOI] [PubMed] [Google Scholar]
  80. Tsetsarkin  KA, Vanlandingham  DL, McGee  CE  et al.  A single mutation in chikungunya virus affects vector specificity and epidemic potential. PLoS Pathog  2007;3:1895–906. 10.1371/journal.ppat.0030201 [DOI] [PMC free article] [PubMed] [Google Scholar]
  81. Tsetsarkin  KA, Chen  R, Yun  R  et al.  Multi-peaked adaptive landscape for chikungunya virus evolution predicts continued fitness optimization in Aedes albopictus mosquitoes. Nat Commun  2014;5:1–14. 10.1038/ncomms5084 [DOI] [PMC free article] [PubMed] [Google Scholar]
  82. Utt  A, Das  PK, Varjak  M  et al.  Mutations conferring a noncytotoxic phenotype on chikungunya virus replicons compromise enzymatic properties of nonstructural protein 2. J Virol  2015;89:3145–62. 10.1128/JVI.03213-14 [DOI] [PMC free article] [PubMed] [Google Scholar]
  83. Vasilakis  N, Deardorff  ER, Kenney  JL  et al.  Mosquitoes put the brake on arbovirus evolution: experimental evolution reveals slower mutation accumulation in mosquito than vertebrate cells. PLoS Pathog  2009;5:1–18. 10.1371/journal.ppat.1000467 [DOI] [PMC free article] [PubMed] [Google Scholar]
  84. Villa  TG, Abril  AG, Sánchez  S  et al.  Animal and human RNA viruses: genetic variability and ability to overcome vaccines. Arch Microbiol  2021;203:443–64. 10.1007/s00203-020-02040-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
  85. Vilsker  M, Moosa  Y, Nooij  S  et al.  Genome detective: an automated system for virus identification from high-throughput sequencing data. Bioinformatics  2019;35:871–3. 10.1093/bioinformatics/bty695 [DOI] [PMC free article] [PubMed] [Google Scholar]
  86. Weaver  SC, Forrester  NL, Liu  J  et al.  Population bottlenecks and founder effects: implications for mosquito-borne arboviral emergence. Nat Rev Microbiol  2021;19:184–95. 10.1038/s41579-020-00482-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
  87. Wickham  H, François  R, Henry  L  et al.  Dplyr: A Grammar of Data Manipulation. R Package Version 1.1.4, 2025. Available at: https://Dplyr.Tidyverse.Org. Accessed on January 05, 2026.
  88. Wilkinson  L. ggplot2: elegant graphics for data analysis by WICKHAM, H. Biometrics  2011;67:678–9. 10.1111/j.1541-0420.2011.01616.x [DOI] [Google Scholar]
  89. Wong  KZ, Chu  JJH. The interplay of viral and host factors in chikungunya virus infection: targets for antiviral strategies. Viruses  2018;10:2–32. 10.3390/v10060294 [DOI] [PMC free article] [PubMed] [Google Scholar]
  90. Xu  S, Li  L, Luo  X  et al.  Ggtree: a serialized data object for visualization of a phylogenetic tree and annotation data. iMeta  2022;1:1–4. 10.1002/imt2.56 [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary_figures_1_and_2_(1)_veaf102
supplementary_figures_1_and_2_(1)_veaf102.pdf (560KB, pdf)
Supplementary_tables_Virus_Evolution_(2)_veaf102
supplementary_tables_virus_evolution_(2)_veaf102.xlsx (779.1KB, xlsx)

Data Availability Statement

All the chikungunya raw data (fastq files) generated and analyzed in this study are available in the SRA-NCBI database (www.ncbi.nlm.nih.gov/sra), under accession number PRJNA1338711. Chikungunya consensus sequences used for phylogenetic analysis are fully available in the GISAID EpiArbo database (https://gisaid.org/), with all accession numbers are listed in Supplementary Table 2. Code and raw data are also available in the GitHub Repository (https://github.com/cab1992/Intra-host-population-dynamics-of-chikungunya-virus-in-humans-and-Aedes-aegypti-), and the Mendeley Data Repository (https://data.mendeley.com/datasets/pr5vsfcf5f/1).

Articles from Virus Evolution are provided here courtesy of Oxford University Press

RESOURCES