Dengue virus genetic diversity and lineage dynamics in Thailand: implications for surveillance and vaccine implementation

Abstract

Background

Dengue fever, caused by dengue virus (DENV), is a major public health problem in tropical and subtropical regions. Thailand is considered endemic for dengue, and all four DENV serotypes co-circulate in this country. Monitoring the genetic diversity of DENV is important for establishing local control policies.

Methods

This study sequenced the complete envelope gene from 92 DENV isolates collected at the Hospital for Tropical Diseases between October 2023 and September 2024. These sequences were analyzed together with dengue virus sequences circulating in Thailand between 2000 and 2024, retrieved from the National Center for Biotechnology Information (NCBI) virus database. The DENV sequences were evaluated for lineage determination and phylogenetic analysis.

Results

Our findings identify the presence or introduction of dengue genotypes in the endemic area. The results revealed the co-circulation of all four DENV serotypes, and DENV-2 was the dominant serotype (44.57%). The predominant serotype switched in response to dengue outbreaks throughout the study period. Phylogenetic analysis revealed that DENV-1 sequences belonged to genotype I, whereas DENV-2 sequences belonged to genotype II and V. Moreover, the dengue vaccine strain was identified via sequencing (Thailand 2024). DENV-3 genotype III was the major circulating genotype from 2013 to 2024. Meanwhile, DENV-4 isolates belonged to genotype I, lineage A.3. In addition, the substitution of the dominant lineage was associated with dengue epidemics.

Conclusions

The co-circulation of all four DENV serotypes has been exhibited in Thailand. Dengue vaccine implementation is both affected by existing DENV circulation and can influence future DENV circulation within a population. Understanding DENV evolution, genotype replacement, and multiple-virus lineage circulation is vital for enhancing surveillance, responding to DENV epidemics, and implementing vaccination programs.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12985-026-03194-2.

Background

Dengue virus (DENV) is a single-positive-stranded RNA virus of approximately 10.6 kb in length belonging to the Flaviviridae family. DENV can be transmitted to humans by the bites of female Aedes mosquitoes that carry the virus [1]. Dengue fever, caused by DENV, is a major public health problem in tropical and subtropical countries [1, 2]. DENV can be classified into four distinct serotypes (DENV-1, DENV-2, DENV-3, and DENV-4) that share approximately 65% amino acid sequence similarity [3, 4]. In 2024, more than 7.6 million dengue cases were reported to the World Health Organization, including 3.4 million confirmed cases, more than 16,000 severe cases, and more than 3,000 deaths [5]. Dengue is an arboviral disease highly endemic in Bangkok, Thailand, and all four DENV serotypes co-circulate throughout the country [6, 7]. In 2024, Thailand reported a total of 105,250 cases, including 90 deaths (case fatality rate [CFR] = 0.09%), compared with 158,620 cases, including 181 deaths (CFR = 0.11%), in 2023 [8, 9]. Most symptomatic patients have only mild symptoms, called dengue fever (DF), which rarely result in serious consequences, but approximately 5% of patients will develop plasma leakage, classified as severe illness, dengue hemorrhagic fever (DHF) [10]. A minority of DHF patients might develop hypovolemic shock, also termed dengue shock syndrome (DSS), which has a notable mortality rate [11, 12]. Vector control has been implemented for dengue prevention and control; however, dengue outbreaks have not been significantly reduced. Therefore, the development and deployment of dengue vaccines represent a promising approach for reducing the global dengue burden.

The DENV genome consists of an open-reading frame flanked by two untranslated regions (5′-UTR and 3′-UTR) [4]. Six elements exist within the 5′-UTR (95–101 nucleotides in length): stem-loop A, stem-loop B, 5′-upstream AUG region, 5′-downstream AUG region, C-coding region hairpin, and 5′-cyclization sequence [4]. The 3′-UTR, approximately 450 nucleotides long, is important for DENV replication, and it modulates viral growth and RNA synthesis in mammalian cells [4]. The open-reading frame encodes three structural proteins (capsid, premembrane/membrane, and envelope proteins) and seven nonstructural (NS) proteins (NS1, NS2A, NS2B, NS3, NS4A, NS4B, and NS5) [7, 13]. The DENV envelope gene is the most frequently selected target for genotyping [14] and vaccine development [15].

DENV is currently classified into four serotypes, which include varying numbers of genotypes [16]. Globally, DENV-1, DENV-2, DENV-3, and DENV-4 can be divided into six (I, II, III, IV, V, and VI), six (genotype I [American], genotype II [Cosmopolitan], genotype III [American/Asian], genotype IV [Asian II], genotype V [Asian I], and genotype VI [sylvatic]), five (I, II, III, IV, and V) and four (I, II, III, and sylvatic) genotypes, respectively [4, 16, 17]. Previous studies have shown that all four DENV serotypes are consistently transmitted throughout Thailand, with varying dominant serotypes [18, 19]. DENV evolution and surveillance are key components for assessing the risk of future dengue epidemics by identifying emerging genotypes in populations. Therefore, the detection of such genotypes incorporates sequence data into surveillance activities, thereby serving as an early warning of subsequent epidemic threats [17].

The geographical distribution of DENV-1 shows that genotype I is predominant in Southeast Asia, China, and East Africa [4]. DENV-1 genotype IV strains have been detected from the West Pacific Islands and Australia, whereas DENV-1 genotype V is present in the Americas, West Africa, and Asia [4]. DENV-1 genotype II has historically been reported in Thailand in the 1950 s and 1960 s [4]. DENV-1 genotype III, the sylvatic strain, has historically been reported in Malaysia in 2005 [4]. Regarding DENV-2 genotypes, genotypes V (Asian I) and IV (Asian II) circulate across Asia [20, 21]. The extinct genotype I (American) was largely restricted to Central and South America [21]. Genotype III (Asian–American) circulates in Central and South America and Southeast Asia [21]. The genotype II (Cosmopolitan) is the most widespread and genetically heterogeneous [21]. This genotype circulates in Asia, the Middle East, the Pacific Islands, and Africa, and contributes substantially to the global dengue burden [20, 21]. DENV-3 genotypes I and II have formed independent subpopulations. Genotype I is present in Southeast Asia and the South Pacific, and genotype II originated in Thailand [7, 22]. Genotype III was originally isolated in the Indian subcontinent, and genotype IV in the Americas [7]. Genotype V was predominantly detected in both Asia and the Americas [4, 7]. Although sylvatic strains have been identified in DENV-1, DENV-2, and DENV-4, no sylvatic strain has been characterized in DENV-3 [22]. DENV-4 genotype I was first reported in the Philippines in 1956 (as the prototype), and has historically been reported in Southeast Asia. Genotype II is distributed in Southeast Asia and the Americas, whereas genotype III has been reported only in Thailand since 1997. Genotype IV is found exclusively as sylvatic strains in Malaysia [4, 7, 23]. In Thailand and Southeast Asia, phylogenetic analyses based on envelope gene sequences revealed the presence of DENV-1 and DENV-4 genotype I, and the co-circulation of DENV-2 genotypes II (Cosmopolitan) and V (Asian I) [7, 19]. For DENV-3, genotypes I and III co-circulate in Thailand [7]. The extensive diversity in viral spatial distribution may affect disease outcomes, underscoring the importance of DENV surveillance [7]. Therefore, constant DENV surveillance using molecular sequencing is crucial for understanding the real-time status of DENV strains and assessing the risk of a subsequent outbreak in Thailand.

In addition, the development of dengue vaccines has spanned several decades, with current leading candidates based on live-attenuated tetravalent platforms. TAK-003 (Qdenga^®) vaccine is based on an attenuated DENV-2 (PDK-53) backbone with chimeric viruses expressing prM/E of DENV-1, −3, and − 4 [15]. In contrast, CYD-TDV (Dengvaxia^®) uses a yellow fever backbone, while Butantan-DV consists of four attenuated DENV strains with 3′UTR deletions [24, 25]. These vaccines are derived from well-characterized laboratory strains, with established attenuation, genetic stability, and suitability for manufacturing [24, 26, 27].

This study analyzed the genetic diversity of clinical specimens from patients with confirmed DENV infection at the Hospital for Tropical Diseases in 2023–2024 using real-time RT-PCR. A total of 92 DENV sequences from our clinical cohort were analyzed together with sequences representing dengue viruses circulating in Thailand between 2000 and 2024, retrieved from the National Center for Biotechnology Information (NCBI) virus database, for lineage determination and phylogenetic analysis. This study will help assess the current genetic diversity of DENV strains circulating in the region. Our findings will help identify the presence or introduction of dengue genotypes in endemic areas, which is important for establishing local epidemiological control policies. Our comparison of the amino acid sequences of the envelope proteins of wild-type DENV strains with those of three dengue vaccines will provide useful information for vaccine implementation.

Methods

Study site and sample collection

This study retrospectively investigated 92 serum samples collected from patients with RT-PCR-confirmed dengue, diagnosed in the Outpatient Department of the Hospital for Tropical Diseases (Bangkok, Thailand) between October 2023 and September 2024.

Ethics declarations

This study was approved by the Faculty of Tropical Medicine, Mahidol University Ethics Committees (MUTM 2025-016-01) and the Institutional Biosafety Committee, Faculty of Tropical Medicine, Mahidol University (Approval No. TM 2025-006).

Confirmation of dengue virus serotypes using real-time RT-PCR

DENV serotypes were identified using the DENV-1–4 Real-Time RT-PCR Assay [28]. The primers and probes for this test (Table S1) were designed to detect currently circulating DENV-1–4 strains with comparable sensitivity. The 25-µL multiplex RT-PCR mixture consisted of 5 µL of RNA mixed with 2.2 µL of nuclease-free H₂O, 12.5 µL of 2× premix, 0.5 µL each of forward and reverse primers for DENV-1 and DENV-3 (final concentration 1 µM), 0.25 µL each of forward and reverse primers for DENV-2 and DENV-4 (final concentration 500 nM), 0.45 µL of each TaqMan probe (final concentration 180 nM), and 0.5 µL of SuperScript III RT/Platinum Taq mix (Thermo Fisher Scientific, Waltham, MA, USA). The thermocycling parameters were as follows: reverse transcription at 50 °C for 30 min, reverse transcriptase inactivation at 95 °C for 2 min, and fluorescence detection for 45 cycles, with 95 °C for 15 s and 60 °C for 1 min.

DENV envelope gene sequencing

The DENV envelope (E) gene was amplified using specific primers as listed in Table S2. In brief, 4 µL of RNA template were added to a 16-µL master mix of the SuperScript™ III One-Step RT-PCR System with Platinum™ Taq DNA Polymerase (Thermo Fisher Scientific, Waltham, MA, USA). The RT-PCR protocol consisted of reverse transcription at 50 °C for 30 min; reverse transcriptase inactivation at 94 °C for 5 min; 40 cycles of 94 °C for 15 s, 50 °C for 30 s, and 72 °C for 2 min; final extension at 72 °C for 5 min; and incubation at 4 °C. The product was visualized by 1% agarose gel electrophoresis, and the target amplicon size was purified using the NucleoSpin Gel and PCR Clean-up kit (Machery-Nagel, Düren, Germany) following the manufacturer’s protocol. The purified samples were sent for Illumina MiSeq sequencing at Tsinke (Beijing, China).

Dengue lineage identification and phylogenetic tree analysis

The classification of DENV lineages was based on the envelope gene region, which is 1,485 bp for DENV-1, DENV-2, and DENV-4 and 1,479 bp for DENV-3. To elucidate the long-term molecular epidemiology of DENV, a total of 92 sequences obtained from our clinical cohort were analyzed alongside DENV sequences circulating in Thailand between 2000 and 2024, which were retrieved from the NCBI virus database [29] and are presented in Table S3. Sequences were searched using the keyword “Dengue virus,” with filters applied to sequences from Thailand and those containing the complete envelope region, and limited to those available from 2005 to 2025. Dengue lineages were analyzed by Nextclade v3.10.0 [30] using Nextclade dataset for DENV-1 based on the reference “Nauru/NMRI-45AZ5PDK-0/1974,” DENV-2 based on the reference “Thailand/CDC-16681/1964,” DENV-3 based on the reference “D3/H/IMTSSA-SRI/2000,” and DENV-4 based on the reference “rDEN4” [16].

Phylogenetic analyses were performed using Nextclade v3.10.0 [30] with the curated datasets [16], which contained approximately 349 sequences for DENV-1, 390 for DENV-2, 179 for DENV-3, and 117 for DENV-4. The complete envelope gene sequences generated in this study, including 25 sequences for DENV-1, 41 for DENV-2, 4 for DENV-3, and 22 for DENV-4, were submitted to the Nextclade platform for phylogenetic tree analysis. The resulting phylogenetic trees were exported and subsequently visualized using iTOL v7.0 [31]. Sequence similarity to previously published strains was inferred from phylogenetic clustering in the reconstructed trees. The nucleotide and amino acid identity percentages of the wild-type DENV-2 envelope gene from Thailand (2001–2024) versus the DENV vaccine were calculated using the SIAS (sequence identities and similarities) web server [32].

Bayesian phylogeny of DENV virus

The evolutionary trees for DENV were analyzed using 92 sequences from our clinical cohort, combined with data from the NCBI virus database. In total, 257, 202, 118, and 184 complete E gene sequences of DENV-1, DENV-2, DENV-3, and DENV-4, respectively, circulating in Thailand between 2005 and 2024, were included in the analysis. The complete envelope gene sequences were selected from sequences stored in the NCBI virus database (Table S4). Maximum likelihood phylogenies were constructed by IQ-TREE [33] utilizing TN93 + G for DENV-1, DENV-2, and DENV-3 and TN93 + G + I for DENV-4, which were determined as the best-fit models using MEGA X [34]. The temporal signal assessment was performed using TempEst v1.5.3 [35] through root-to-tip genetic distance regression analysis relative to sampling dates (Supplemental Fig. 1). The time to the most recent common ancestor (tMRCA) was analyzed using BEAST v10.5.0 [36] based on Bayesian Markov chain Monte Carlo (MCMC) analysis. To select the best-fitting model, an uncorrelated relaxed log-normal clock (UCLN) and four priors (constant, exponential, Bayesian Skygrid, and Gaussian Markov random field (GMRF)) were considered, as presented in Table S5. MCMC was run for 30 million states, with sampling every 1,000 steps to ensure stationarity and achieve adequate effective sample sizes for all statistical parameters. The maximum clade credibility (MCC) tree was generated and annotated with posterior probabilities by TreeAnnotator v10.5.0 [36], with a 10% burn-in, and visualized in FigTree v1.4.4 [37].

Results

Dengue case distribution in Thailand

The E gene was successfully sequenced for all 92 dengue strains encountered at the Hospital for Tropical Diseases between October 2023 and September 2024. The results demonstrated the co-circulation of all four DENV serotypes. DENV-2 was the predominant serotype throughout the study period (44.57%), followed by DENV-1 (27.17%), DENV-4 (23.91%), and DENV-3 (4.35%). As in the previous report for 2019–2020, DENV-3 was rare [7]. The IDs, GenBank accession numbers, and other details regarding the obtained sequences are listed in Table S6.

To better understand the distribution of dengue in Thailand, all 92 dengue cases from our cohort were analyzed together with national dengue case data from 2011 to 2024, as illustrated in Fig. 1A [8, 38]. The highest number of dengue cases was recorded in 2023 (158,620 cases, Fig. 1A). As expected, the dengue incidence rate (238.15 per 100,000 population) and number of dengue-related deaths (181 cases) also peaked in 2023 (Fig. 1A). The lowest number of dengue cases (9,956 cases), incidence rate (15.03 per 100,000 population), and number of deaths (6 cases) throughout the review period were recorded in 2021 (Fig. 1A), occurring after the COVID pandemic and lockdown in many areas. Because of a lack of dengue serotype reports from Thailand’s Ministry of Public Health, the dengue serotype distribution data in 2000–2024 were retrieved from the NCBI virus database, as presented in Fig. 1B. In total, 1,553 complete DENV envelope gene sequences were retrieved from the NCBI virus database. The findings showed that DENV-1 and DENV-2 were the predominant serotypes in Thailand. Conversely, DENV-3 was rarely detected throughout the study period. Similar to the findings for dengue cases in Thailand, the lowest number of cases was recorded in 2021, and the dengue serotype from the NCBI virus was also the lowest in this year.

Fig. 1.

Epidemiological trends and serotype distribution of DENV in Thailand. (A) The annual number of reported dengue incidences in Thailand between 2011 and 2025, including dengue cases (●), dengue incidence rate per 100,000 population (), and dengue death cases (). (B) The annual number of DENV serotype distribution, including serotypes DENV1 (●), DENV2 (), DENV3 (), DENV4 ()

Phylogenetic analysis of DENV-1 in Thailand

In Thailand, the incidence of DENV-1 peaked in 2006, and three genotype I lineages (F, H, and K) were detected, as shown in Fig. 2B. The envelope gene (1,485 bp) from samples collected in 2023 and 2024 was amplified. Phylogenetic analysis of the sequencing data revealed that all 25 DENV-1 sequences belonged to genotype I lineage K.1.1, which predominantly includes viruses circulating in Southeast Asia (Fig. 2A, Table S6). These sequences shared a high degree of similarity with the Thailand 2019 (OM281593.1), Cambodia 2019 and 2020 (OQ678061.1, OL411495.1), Myanmar 2022 (OP984871.1), and China 2019 strains (MN923088.1). Historically, DENV-1 strains detected in Thailand during the 1960 s belonged to genotype II [39]. Still, this genotype was not detected in these samples (Fig. 2B). Conversely, circulating DENV-1 strains in Thailand have predominantly clustered within genotype I.

Fig. 2.

Phylogenetic analysis of DENV-1 in Thailand. (A) Results of the phylogenetic analysis. The sequences newly determined in the present study are marked with dots. The left panel provides a legend denoting the genotypes. Dengue lineages were analyzed using Nextclade v3.10.0 [30]. The phylogenetic tree was visualized using iTOL v7.0 [31]. (B) Circulation of DENV-1 lineages in Thailand between 2000 and 2024. The data were analyzed according to the complete envelope gene sequence (n = 974). (C) Molecular clock analysis of 257 DENV-1 envelope sequences in Thailand between 2004 and 2024. Sequences reported in the presented study are colored red. The left panel provides a legend denoting the lineages

Using 974 DENV-1 envelope gene sequences collected in Thailand from 2000 to 2024 and retrieved from the NCBI virus database, we found that the majority of DENV-1 strains clustered within genotype I, specifically lineages H (39.6%) and K (37.5%), as illustrated in Fig. 2B and Table S3. Lineage H was the predominant circulating lineage between 2001 and 2006, whereas lineage K was dominant from 2007 to 2024. The time-scaled Bayesian phylogenetic tree based on the envelope sequences of 257 DENV-1 sequences in 2005–2024 (Table S4) was constructed using UCLN and Bayesian Skygrid coalescent, which was the best-fitting model (Table S5). The tMRCAs for lineages H and K were 1999.4 and 2002.5, respectively. Lineage H was not detected after 2020 (Fig. 2C), suggesting a potential decline or replacement by other circulating lineages. The tMRCA for genotype I lineage K.1.1, which represents the current lineage, was estimated at 2015.16 (Fig. 2C; Table 1). These results suggest that lineage K.1.1 likely emerged several years before becoming the dominant circulating lineage. This pattern is consistent with a process of lineage turnover, in which newly introduced or emerging variants gradually replace previously circulating lineages over time.

Table 1.

The time of the most recent common ancestor (tMRCA), 95% highest probability density (HPD)

Virus	Lineage	tMRCA	tMRCA 95% HPD interval
DENV-1	1I_H	1999.40	2001.17, 1997.45
	1I_K	2002.50	2008.79, 2001.09
	1I_K.1.1	2015.16	2016.15, 2014.17
DENV-2	2II_F.1.1	2011.38	2012.90, 2009.78
	2II_F.1.3	2006.74	2011.00, 2001.81
	2 V	1969.84	1984.20, 1952.94
	2V_A.2	2013.34	2015.38, 2011.09
DENV-3	3I_A.1.2	2005.28	2010.33, 1998.65
	3II_A	1992.70	1997.74, 1987.13
	3II_B	2003.21	2005.46, 2000.39
	3III_B.1	2004.12	2006.47, 2001.25
DENV-4	4I_A.2	1997.65	2000.14, 1994.82
	4I_A.3	2001.37	2003.70, 1998.70
	4II_A.1	2008.48	2009.62, 2005.42

Phylogenetic analysis of DENV-2 in Thailand

For DENV-2–positive samples, we determined 41 complete envelope gene sequences (1,485 bp; Fig. 3A, Table S6). Phylogenetic analysis revealed that 17 sequences belonged to genotype II lineage F.1.1 (Fig. 3A, green dot). This clade shared ancestral strains with DENV-2 strains, including Cambodia 2022 (PP411257.1 and PP411252.1), China 2023 (PP564601.1, PP564586.1, and PP564661.1), and India 2021 (OM730078.1). One sample (ID0066) belonged to the genotype II lineage F.1.3 (Fig. 3A, cyan dot) and shared sequence similarity with Vietnam 2022 (PP269908.1 and OP984834.1). Additionally, 22 sequences were classified as genotype V lineage A.2 (Fig. 3A, red dot), and they were similar to Thailand 2019 (MZ636801.1), Myanmar 2019 (MW788987.1), and China 2019 (MN923120.1). One DENV-2 sequence (ID0244) was classified as genotype V (Fig. 3A, blue dot). These findings indicate the co-circulation of two distinct DENV-2 genotypes in this region.

Fig. 3.

Phylogenetic analysis of DENV-2 in Thailand. (A) Results of the phylogenetic analysis. The sequences newly determined in the present study are marked with dots. The left panel provides a legend denoting the genotypes. Dengue lineages were analyzed using Nextclade v3.10.0 [30]. The phylogenetic tree was visualized using iTOL v7.0 [31]. (B) DENV-2 lineage distribution in Thailand between 2000 and 2024 based on the complete envelope gene sequence (n = 767). (C) Molecular clock analysis of 202 DENV-2 envelope sequences in Thailand between 2005 and 2024. Sequences reported in the presented study are colored red. The left panel provides a legend denoting the lineages

Using 767 complete envelope gene sequences from the NCBI virus database collected in Thailand from 2000 to 2024, the majority of DENV-2 strains were clustered into genotypes II (13.56%) and V (86.44%), as shown in Fig. 3B and Table S3. DENV-2 genotype V was the major circulating genotype between 2000 and 2016. DENV-2 genotype II lineage F was first identified in Thailand in 2014 and co-circulated with DENV-2 genotype V during 2016–2024. The co-circulation of two DENV-2 genotypes prompted us to estimate the time of introduction using molecular clock analysis. We prepared datasets consisting of 41 envelope sequences of DENV-2 obtained in the present study, along with other DENV-2 sequences collected in Thailand between 2005 and 2024 (Table S4). The time-scaled Bayesian phylogenetic tree based on envelope sequences was constructed using UCLN and Bayesian Skygrid coalescent, which was the best-fitting model (Table S5). The tMRCAs of DENV-2 lineages II_F.1.1 and II_F.1.3 were 2011.38 and 2006.74, respectively (Fig. 3C; Table 1), suggesting that these lineages likely emerged at different times and subsequently co-circulated in Thailand, reflecting the complex evolutionary dynamics of DENV-2. The tMRCAs of DENV-2 lineage V_A and V_A.2 were 1997.39 and 2013.34, respectively (Fig. 3C; Table 1), suggesting that the more recently emerged V_A.2 lineage may have replaced or become dominant over the earlier V_A lineage.

As presented in Fig. 3B and Table S3, using 767 complete envelope gene sequences from the NCBI virus database collected between 2000 and 2024, we found that the oldest DENV-2 genotype V strain (similar to the dengue vaccine strain) was predominantly detected in Thailand in 2001 (JF812114.1), 2006 (MW946313.1), 2023 (PV628776.1), and 2024 [this study; PV663128.1 (ID0244)]. The TAK-003 (Qdenga^®) vaccine strain and DENV-2 wild-type strains had 100% nucleotide identity with the Thailand 2023 isolate (PV628776.1) and 99.93% (1-bp difference) nucleotide identity with the ID0244 isolate (Thailand 2024) (PV663128.1), as illustrated in Table 2. Notably, the TAK-003 vaccine has been in use in Thailand since 2023. TAK-003 features three attenuation sites at 5′-UTR-c57t, NS1-G53D, and NS3-E250V [15]. Sequencing analysis using primers targeting attenuation sites detected three attenuation sites in the DENV-2 genotype V isolate ID0244 (Thailand, 2024, PV663128.1; Supplemental Fig. 2). However, vaccination information for the patients is not available due to ethical restrictions on the retrospective samples included in this study. Therefore, no definitive conclusion regarding vaccine-associated viremia can be drawn from the available data.

Table 2.

Percentage identity of the wild-type genotype V DENV-2 envelope gene in Thailand (2001–2024) versus the DENV vaccine

Phylogenetic analysis of DENV-3 in Thailand

Although DENV-3 was infrequently detected in this study, four DENV-3 cases were identified. We sequenced the complete envelope gene (1479 bp) of all four isolates. All sequences in this study were classified as genotype III, specifically lineage B.1, and they exhibited high similarity to a strain from China in 2015 (MN018375.1), as presented in Fig. 4A and Table S6.

Fig. 4.

Phylogenetic analysis of DENV-3 in Thailand. (A) Results of the phylogenetic analysis. The sequences newly determined in the present study are marked with dots. The left panel provides a legend denoting the genotypes. Dengue lineages were analyzed using Nextclade v3.10.0 [30]. The phylogenetic tree was visualized using iTOL v7.0 [31]. (B) DENV-3 lineage distribution in Thailand between 2000 and 2024 based on the complete envelope gene sequence (n = 510). (C) Molecular clock analysis of 118 DENV-3 envelope sequences (1,479 bp) in Thailand between 2004 and 2024. Sequences reported in the presented study are colored red. The left panel provides a legend denoting the lineages

Using 510 complete DENV-3 envelope gene sequences from the NCBI virus database collected in Thailand between 2000 and 2024, we found that genotype II was the predominant circulating genotype during 2000–2012. By contrast, genotype III was the major circulating genotype from 2013 to 2024 (Fig. 4B, Table S3). For DENV-3, the time-scaled Bayesian phylogenetic tree was constructed from envelope sequences using the UCLN and constant-size coalescent models, which were the best-fitting models (Table S5). The 118 DENV-3 sequences comprised four DENV-3 envelope sequences obtained in the present study, along with other DENV-3 sequences collected in Thailand during 2004–2024 (Table S4). The time-scaled Bayesian phylogenetic tree revealed that DENV-3 genotype II lineages A and B had tMRCAs of 1992.70 and 2003.21, respectively, after which both clades disappeared (Fig. 4C). Subsequently, DENV-3 genotype III lineage B.1 became the major circulating lineage. It had a tMRCA of 2004.12 (Fig. 4C; Table 1). These findings suggest a lineage turnover event in which genotype III lineage B.1 replaced the previously circulating genotype II lineages in Thailand. DENV-3 lineage I_A.1.2, which has a tMRCA of 2005.28, co-circulated in Thailand in 2020, indicating the persistence of multiple DENV-3 lineages during this period.

Phylogenetic analysis of DENV-4 in Thailand

For DENV-4, we amplified the envelope gene (1,485 bp) from samples collected in 2023 and 2024. Sequencing analysis revealed that 21 of the DENV-4 isolates belonged to genotype I, specifically lineage A.3, and they exhibited high similarity to strains from Thailand in 2017 (LC410203.1) and Myanmar in 2018 (MW788999.1), as presented in Fig. 5A (red dot) and Table S6. Genotype I is currently the predominant genotype circulating in Thailand [39]. Notably, one DENV-4 isolate was classified as genotype II lineage A.1, as presented in Fig. 5A (blue dot) and Table S6. As presented in Fig. 5B and Table S3, the lineage distribution of DENV-4 in Thailand from 2000 to 2024 featured a temporal shift in dominant lineages. Between 2000 and 2008, genotype I lineage A.2 was the predominant circulating strain. From 2009 to 2024, genotype I lineage A.3 became the dominant lineage. For DENV-4, the time-scaled Bayesian phylogenetic tree was constructed from envelope sequences using the UCLN and constant-size coalescent models, which were the best-fitting models (Table S5). The DENV-4 dataset comprised 22 envelope sequences obtained in the present study, along with other DENV-4 sequences collected in Thailand between 2005 and 2024 (Table S4). Co-circulation of DENV-4 genotypes I and II was noted in Thailand. The time-scale tree indicated that genotype I lineage A.2 circulated from 2000 to 2008 and then disappeared, with an estimated tMRCA of 1997.65. In contrast, genotype I lineage A.3 emerged later, with an estimated tMRCA of 2001.37. This lineage expanded and became the major circulating lineage (Fig. 5C; Table 1). These findings demonstrate dynamic lineage turnover of DENV-4 genotype I lineage A.3 replacing the previously circulating A.2 lineage. Our study found that DENV-4 lineage II_A.1 was co-circulating in Thailand in 2023. The tMRCA of this lineage was estimated at 2008.48 (Fig. 5C; Table 1), suggesting it may have been circulating in the region for several years.

Fig. 5.

Phylogenetic analysis of DENV-4 in Thailand. (A) Results of the phylogenetic analysis. The sequences newly determined in the present study are marked with dots. The left panel provides a legend denoting the genotypes. Dengue lineages were analyzed using Nextclade v3.10.0 [30]. The phylogenetic tree was visualized using iTOL v7.0 [31]. (B) DENV-4 lineage distribution in Thailand between 2000 and 2024 based on the complete envelope gene sequence (n = 629). (C) Molecular clock analysis of 184 DENV-4 envelope sequences (1,485 bp) in Thailand between 2005 and 2024. Sequences reported in the presented study are colored red. The left panel provides a legend denoting the lineage

Sequence differences between vaccine strains and circulating wild-type viruses

Genotypic differences between wild-type DENV strains and vaccine strains might influence vaccine efficacy, particularly when amino acid changes occur in epitopes targeted by DENV-neutralizing monoclonal antibodies. In this study, we compared the amino acid sequences of the E proteins of wild-type DENV isolates with those of the vaccine strains CYD-TDV (Dengvaxia^®), TAK-003 (Qdenga^®), and Butantan-DV, as detailed in Tables S7–S10 [40–42]. The genotypes of the vaccine strains were analyzed using Nextclade DENV dataset [16], which showed that DENV-1 was II (TAK-003, AF180818.1), IV (CYD-TDV, KX239894.1), and IV (Butantan-DV, AY145123.1). The genotype of the vaccine strain for DENV-2 was genotype V ((TAK-003, U87412.1), (CYD-TDV, KX239895.1)) and genotype I (Butantan-DV, AY744149.1). The vaccine-strain genotypes of DENV-3 were V (TAK-003, KU725665.1), II (CYD-TDV, KX239896.1), and I (Butantan-DV, AY656170.1). The vaccine strain of DENV-4 was genotype II across all three vaccines. Vaccine efficacy was associated with the degree of amino acid dissimilarity between the vaccine strains and disease-causing DENV strains [41]. Specifically, we analyzed amino acid sequence identity at known B cell epitope regions to assess potential antigenic mismatches between circulating wild-type viruses and vaccine strains. The monoclonal antibodies used in this analysis included 1F4, 14C10, and E103 for DENV-1 [43–45]; EDE.C8, EDE.A11, EDE.B7, EDE.C10, and 2D22 for DENV-2 [46, 47]; and 5J7, 4.8 A, D11C, 1.6D for DENV-3 [48, 49]. 5H2 is a neutralizing human monoclonal antibody targeting DENV-4 [50]. The monoclonal antibodies targeting these epitopes are among the best-defined DENV epitopes recognized by potent human monoclonal antibodies. They are therefore of particular interest in vaccine development and immune-correlate analysis [41]. When comparing the vaccine strains (CYD-TDV, KX239894.1; TAK-003, AF180818.1; and Butantan-DV, AY145123.1) with wild-type DENV-1 sequences obtained in this study, we identified amino acid differences in the envelope protein at positions 52, 155, and 171. These positions are located within epitope regions targeted by well-characterized neutralizing monoclonal antibodies, including 1F4 and 14C10, as presented in Fig. 6A.

Fig. 6.

Sequence conservation between the DENV vaccine component and wild-type DENV at epitope locations targeted by neutralizing human monoclonal antibodies. Amino acid targets for human neutralizing monoclonal antibodies are colored as indicated in the key (top) and compared with the vaccine sequence and wild-type sequences. Sites are indicated at the top of the columns

Two wild-type DENV-2 genotypes, II and V, are currently circulating in Thailand. Comparative analysis of the genotypes of vaccine strains (CYD-TDV, KX239895.1; TAK-003, U87412.1; and Butantan-DV, AY744149.1) with wild-type DENV-2 sequences from this study revealed amino acid substitutions in the envelope protein at positions 71, 83, 149, 226, and 362. These positions overlap with epitope regions targeted by the EDE.C8, EDE.A11, EDE.B7, EDE.C10, and 2D22 monoclonal antibodies, as illustrated in Fig. 6B. As presented in Fig. 6C, the sequence comparison between vaccine sequences (CYD-TDV, KX239896.1; TAK-003, KU725665.1; and Butantan-DV, AY656170.1) and wild-type DENV-3 sequences from this study revealed no amino acid changes in the envelope protein at monoclonal antibodies targeting epitopes. The comparison between envelope amino acid sequences of DENV-4 vaccines (CYD-TDV, KX239897.1; TAK-003, U18429.1; and Butantan-DV, AF326827.1) and wild-type identified amino acid changes in the envelope protein at position 163 of 5H2, as described in Fig. 6D. The amino acid positions, which differed between dengue vaccine strain and the wild-type DENV-1–4 sequences from this study, were dispersed across the envelope protein. They did not cluster at any particular structural domain. Our results identified specific amino acid changes in epitopes targeted by DENV-neutralizing monoclonal antibodies. These observations highlight potential sequence variation; however, they should not be interpreted as evidence of functional effects on vaccine-induced immunity without experimental validation.

Discussion

Our results revealed the genetic diversity of DENV strains collected from patients with dengue treated at the Hospital for Tropical Diseases from 2023 to 2024. We aimed to perform phylogenetic analyses of the four DENV serotypes in Thailand to assess the complexity of dengue epidemiology during active DENV circulation. The introduction of new serotypes, genotypes, and lineages into the region, and the subsequent replacement of endemic viruses, have occurred frequently across all serotypes, and these shifts may be associated with a significant increase in dengue severity [51]. A better understanding of the forces influencing DENV evolution and genotype replacement could perhaps provide an early warning for dengue outbreaks and inform subsequent treatment [17]. A previous study found that DENV-1 was consistently the dominant serotype in Thailand in 2018–2020 [7]. From our report, DENV-1 was the second most common serotype after DENV-2 in 2023 and 2024. DENV-1 genotype I, the common genotype in Southeast Asia, China, and East Africa [4], was the major serotype circulating in Thailand [7, 19]. The tMRCA estimates indicate successive lineage replacement within DENV-1 genotype I in Thailand, with earlier H lineages being replaced by the more recently emerged K.1.1 lineage, which became predominant during 2020–2024. This pattern highlights the dynamic nature of DENV-1 evolution, where newly emerging lineages gradually replace previously circulating ones. In Thailand, DENV-1 lineage replacement was associated with dengue outbreaks and reported to be associated with mosquito transmission [52]. The genotype replacement of DENV-1 was also associated with a change in serotype prevalence, as increased DENV-1 clade diversity was associated with increased abundance of this serotype and a concomitant decrease in DENV-4 [53].

Our study demonstrated that DENV-2 was the dominant serotype in 2023 and 2024, and that DENV-2 genotypes II and V co-circulated in Thailand, consistent with prior reports [7, 19]. DENV-2 genotype II circulates in Australia, East and West Africa, the Pacific and Indian Ocean islands, the Indian subcontinent, and the Middle East, whereas genotype V circulates in Malaysia and Thailand [4]. Multiple DENV-2 genotypes and lineages were observed to co-circulate in Thailand, including genotype II (lineages F.1.1 and F.1.3) and genotype V (lineage A.2). Earlier genotype V lineages were predominant before being co-circulating with emerging genotype II lineages, reflecting ongoing lineage turnover. Lineage replacement was associated with dengue epidemics and may contribute to disease severity, as evidenced by the replacement of DENV-2 genotype III by genotype V in Vietnam in the 1990 s [54]. An investigation of Vietnamese patients revealed higher viral titers in those infected with DENV-2 genotype V strains than in those infected with DENV-2 genotype III strains [54].

For DENV-3, genotype II, which circulates in Thailand, Vietnam, and Bangladesh, has constituted the dominant genotype reported in Thailand since 1973 [4, 55]. DENV-3, a clear genotype shift was observed in Thailand, with genotype II being replaced by genotype III in more recent years [55]. DENV-3 genotype III lineage B.1 was the predominant circulating genotype from 2013 to 2024. Genotype replacement occurred in Bangladesh, specifically, DENV-3 genotype I replaced DENV-3 genotype II, which was predominant between 2000 and 2009 [56].

DENV-4 genotype I strains circulate in Thailand, the Philippines, Sri Lanka, and Japan [4]. Our results revealed that DENV-4 genotype I lineages A.2 and A.3 were the major lineages circulating in Thailand, consistent with previous findings [19, 55]. DENV-4, genotype I, remains the predominant genotype in Thailand, with a clear shift in dominant lineages over time from A.2 to A.3, reflecting ongoing lineage turnover. Additionally, DENV-4 genotype II circulates in Indonesia, Malaysia, Tahiti, the Caribbean, and the Americas [4]. Our study identified the circulation of DENV-4 genotype II, lineage A.1, in Thailand in 2023, indicating co-circulation of multiple genotypes, suggesting continued viral diversification and possible introduction events.

This temporal gap between tMRCA and dominance lineage may reflect undetected or low-level circulation before their recognition, suggesting that newly emerging lineages may circulate at low frequencies before expanding and eventually replacing previously dominant lineages. Similar to the previous study of DENV, the cryptic circulation observed in Colombia suggested a period of cryptic transmission lasting several years before the recent epidemic [57]. These findings imply that major DENV outbreaks may not always result from recent viral introductions. Instead, circulating lineages may undergo phases of regional spread before contributing to local outbreaks.

A similar scenario involving a temporal gap was also reported in Nigeria [58], Brazil [59], and Vietnam [60]. Lineage turnover is a common feature of DENV evolutionary dynamics, and newly emerging lineages often circulate at low prevalence before expanding within the population [53, 61]. Lineage turnover was also reported for DENV in several studies, such as India [61] and Nicaragua [62]. However, our analysis was limited to sequences obtained from Thailand, and the inferred temporal patterns may be influenced by incomplete sampling or the absence of sequences from surrounding countries where cross-border viral movement may occur.

In 2023, TAK-003 was approved in Thailand for the individual prevention of dengue fever. TAK-003 is a live-attenuated tetravalent chimeric vaccine produced using recombinant DNA technology and is recommended for individuals aged 4–60 years [63]. The TAK-003 parent strain for DENV-2 was Thailand 16,681/PDK-53 (genotype V), which was isolated in 1964 [64]. However, there are three attenuation sites in TAK-003 at 5′-UTR-c57t, NS1-G53D, and NS3-E250V [15]. Interestingly, DENV-2 genotype V (similar to the dengue vaccine strain) remained detectable in Thailand in 2023 and 2024, and its existence might have been influenced by vaccine implementation. These three attenuation sites were confirmed in the DENV-2 genotype V ID0244 isolate (Thailand 2024, PV663128.1) by sequencing (Supplemental Fig. 2). Patient vaccination information is unavailable due to ethical restrictions on the retrospective samples included in this study. Therefore, we are unable to confirm whether the patient from whom isolate ID0244 (Thailand, 2024; PV663128.1) was obtained had recently received the TAK-003 vaccine. In addition, as TAK-003 is commercially used in Thailand, the possibility of vaccine-associated viremia should be considered. The case of severe dengue hemorrhagic fever following TAK-003 dengue vaccination has been reported in Thailand [65], although such events appear to be rare. In addition, TAK-003 is a tetravalent live-attenuated vaccine. The sample was screened using a multiplex real-time RT-PCR assay capable of detecting all four DENV serotypes (including TAK-003 vaccine strain). Based on this analysis, the sample was positive only for DENV-2, with no co-detection of DENV-1, DENV-3, or DENV-4. Consistent with findings from the TAK-003 clinical trial (DEN-301), vaccine-associated viremia has been reported predominantly for only DENV-2 [66]. Similar to the previous report from the TAK-003 study, vaccine viremia mainly occurred with TDV-2 (DENV-2) and resolved within 1 month of vaccination [67]. In addition, RT-qPCR detected viremia in 69.5% of participants (66 of 95) who received CYD-TDV, primarily occurring 6–14 days after injection [68]. In addition, previous reports have identified older DENV-2 genotype V lineages in public databases, suggesting that genetically similar strains may still circulate at low levels. These findings highlight the importance of carefully distinguishing vaccine-related sequences from circulating wild-type viruses.

To reassure that the vaccine is effective against currently circulating DENV genotypes, we have provided a snapshot of DENV genetic diversity. This will facilitate the development of more effective dengue vaccines and disease controls. Our results indicated that the mutations were dispersed across the envelope protein and did not cluster within any particular structural domain. Although amino acid substitutions were observed within reported neutralizing antibody epitopes, these findings are based solely on sequence comparisons. In the absence of functional assays, the potential impact of these substitutions on vaccine-induced immunity remains uncertain. Therefore, these observations represent potential antigenic variation rather than demonstrated effects on vaccine performance. Moreover, it is important to determine whether vaccination campaigns exerted selection pressure on the circulation of wild-type DENVs, thereby favoring certain DENV lineages [42]. Understanding DENV genetic variability and its potential impact on dengue vaccine effectiveness is therefore crucial [42].

This study has the limitations inherent to Nextclade-based analysis, as it relies on curated reference datasets for lineage assignment. The reduced dataset may not fully capture global DENV diversity. Moreover, statements describing sequences as “similar to” previously reported strains should be interpreted with caution, as the analysis is based on the curated dataset in Nextclade, which may not include all closely related sequences available in public databases. Notably, most DENV sequences deposited in the NCBI Virus database originate from individual DENV research studies rather than routine national surveillance in Thailand. In addition, there is often a delay between sample collection and sequence release, which in some cases can span several years.

Conclusions

The present study demonstrated the co-circulation of all four DENV serotypes in Bangkok, Thailand, in 2023 and 2024. DENV-1 and DENV-2 were predominantly detected in Thailand. Phylogenetic analysis revealed dynamic genotype distribution and lineage turnover. All DENV-1 sequences from this study belonged to genotype I, lineage K.1.1, which has replaced the earlier lineage H. In addition, DENV-2 genotypes II and V co-circulated in Thailand. Significantly, one DENV-2 strain showed high genetic similarity to the TAK-003 vaccine strain, highlighting the importance of distinguishing vaccine-related sequences from those of circulating wild-type viruses. DENV-3 and DENV-4 also exhibited genotype-specific patterns, with evidence of lineage turnover and multiple genotype co-circulation. Amino acid changes among the important monoclonal antibody-targeted epitopes for DENV-1, DENV-2, and DENV-4. Dengue vaccine implementation is elaborated by existing DENV circulation and can also alter future DENV circulation within a community. Continuous molecular monitoring of DENV in endemic areas is thus warranted to control dengue better and to support the development of effective vaccines in the future.

Abbreviations

CFR

Case fatality rate

DENV

Dengue virus

DF

Dengue fever

DHF

Dengue hemorrhagic fever

DSS

Dengue shock syndrome

GMRF

Gaussian Markov random field

HPD

Highest probability density

iTOL

Interactive Tree of Life

tMRCA

Time to the most recent common ancestor

MCC

Maximum clade credibility

MCMC

Markov chain Monte Carlo

NCBI

National Center for Biotechnology Information

UCLN

Uncorrelated relaxed log-normal clock

UTR

Untranslated region

Author contributions

S.C. and P.Si. (Pichamon Sittikul) wrote the manuscript. P.Si. (Pichamon Sittikul), K.T., and I.W carried out the data analysis using the bioinformatic tools. P.Sr. (Pimolpachr Sriburin) and I.W performed the molecular diagnosis. J.R., W.N., and P.Sr. (Pimolpachr Sriburin) contributed to the clinical sample management. S.L., C.R., C.S., and K.L. contributed to the overall research strategy and revised the manuscript. S.C. finalized the manuscript. All authors have read and agreed to the published version of the manuscript.

Funding

Open access funding provided by Mahidol University. This research paper is supported by Specific League Funds from Mahidol University and the Technical Support Accelerator Award, Mahidol-Oxford Translational Innovation Partnership, and Mahidol-Oxford Tropical Medicine Research Unit (MORU).

Data availability

All 92 DENV sequences were deposited in the NCBI under the accession numbers PV663069.1–PV663160.1 (Table S6).

Declarations

Ethics approval and consent to participate

This study was approved by the Faculty of Tropical Medicine, Mahidol University Ethics Committees (MUTM 2025-016-01) and the Institutional Biosafety Committee, Faculty of Tropical Medicine, Mahidol University (Approval No. TM 2025-006).

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.