To study the epidemiology of arbovirus infection, we, an international team of researchers in Japan, Thailand, the Philippines, and Indonesia, conducted arbovirus surveillance in our respective countries during 2016–2018 with the support of our governments. In each country, we collected mosquitoes in and around cattle or pig housing using sweeping nets and aspirators. We collected mosquitoes that had digested blood in their midguts. We identified their species and sorted them into pools, which we used for virus isolation. We also collected serum samples from pigs and wild boars from each country to use for virus isolation.

We passed homogenized mosquito or serum samples through 0.45-μm filters (Corning Inc., https://www.corning.com) and inoculated filtrates onto monolayers of 3 culture cell lines (mammalian cell lines Vero9013 and BHK-21 and mosquito cell line C6/36). We assessed cytopathic effect (CPE) daily and collected supernatants from cells that exhibited CPE. If we observed no CPE, we passaged the cells 5 times for 7 days each, after which, if virus was present, CPE should have become apparent. We extracted RNA from culture supernatants using the QIAamp Viral RNA Mini Kit (QIAGEN, https://www.qiagen.com) and subjected the resulting RNA to reverse transcription PCR (RT-PCR) using the QIAGEN One-Step RT-PCR Kit and 2 universal flavivirus-specific primer sets (MAMD and cFD2 or FU2 and cFD3) (8,9) to screen for flaviviruses. To determine genome sequences, we used the QIAGEN One-Step RT-PCR Kit, TaKaRa LA RT-PCR Kit version 2.1 (Takara Bio, https://www.takarabio.com), and Invitrogen 5′ RACE System for Rapid Amplification of cDNA Ends version 2.0 (https://www.thermofisher.com) as needed in combination with several JEV-specific primers (Appendix Table 1).

Of 945 pig serum samples, we selected 56 candidate samples for virus isolation on the basis of their RT-PCR results with the MAMD and cFD2 primers, and from these samples, we obtained the full or partial genome sequences of 5 JEV isolates (Appendix Table 2). Out of a total of 22,277 mosquitoes comprising >16 species, we obtained the full or partial genome sequences of only 2 JEV isolates (Appendix Table 3). Overall, we obtained the full-genome sequence of 4 of the 7 JEV isolates and the partial genome sequence (envelope gene) of the remaining 3 isolates (DDBJ accession nos. LC461956–62; Table).

A preliminary study we performed showed that many pigs in these countries possessed antibodies against JEV (K. Maeda, unpub. data). We found that JEV was more often isolated from serum samples from JEV antibody–negative pigs in farms where JEV seroprevalence was low. Because JEV isolation seems to be difficult in endemic regions, we suggest selecting younger pigs, which are less likely to be JEV antibody positive, for virus isolation studies to increase the chances of success.

Figure

Figure. Maximum-likelihood phylogeny of JEV isolates, Japan and Southeast Asia, 2016–2018 (circles), and reference isolates constructed on the basis of the envelope gene sequence (1,500 nt). A) Genotype Ia (GIa); B) genotype...

The 2 JEV isolates we recovered in Japan, JEV/MQ/Yamaguchi/803/2016 and JEV/MQ/Yamaguchi/804/2016, were GIa and not GIII. In a phylogenetic analysis, these viruses clustered with JEV isolates recovered from Japan (in 2013 and 2009), China (in 2007), and South Korea (in 2010) (Figure, panel A). The 2 viruses collected in Japan in 2016 were most closely related to viruses recovered in 2013 from the same site that had been sampled in our previous study (10), suggesting that this strain has been maintained in the Yamaguchi area of Japan since at least 2013.

The JEV isolate we obtained from Thailand, JEV/sw/Thailand/185/2017, was GIb and clustered with other JEV isolates detected in Thailand during 1985–2005 (11), as well as isolates from Myanmar in 2010, Cambodia in 2014 and 2015, and Singapore in 2014 (Figure, panel B). Thus, these JEVs have been maintained in Thailand and other parts of the southern peninsula of continental Asia for >30 years.

JEV is distributed extensively throughout the Philippines (12). However, only 3 JEV GIII isolates from the Philippines (which were obtained from pigs during 1984–1986) were available for genetic analysis. Our 2 Philippines-derived JEV isolates, JEV/sw/Mindanao/K3/2018 and JEV/sw/Mindanao/K4/2018, clustered with these isolates (Figure, panel C). Our 2 Indonesia JEV isolates, JEV/sw/Bali/93/2017 and JEV/sw/Bali/94/2017, were obtained from the island of Bali, where a JEV vaccination program began in March 2018 (13). These JEV isolates clustered with other Indonesia isolates obtained in the 1980s (Figure, panel C) (14).

In summary, our phylogenetic analysis revealed that the JEV isolates we obtained from Japan were GIa, the isolate from Thailand was GIb, the isolates from the Philippines were GIII, and the isolates from Indonesia were GIV (Figure, panels A–D; Appendix Figure). These results indicated that JEV GIII and GIV are still active and being maintained in parts of Asia.

Our data demonstrate that a number of the JEV isolates we obtained in select countries of Southeast Asia during 2016–2018 were phylogenetically related to isolates reported in the same country in the 1980s, suggesting that some JEV strains have been maintained in their corresponding regions. Contrary to our expectation, the JEV transmission cycle seems to have been maintained indigenously. JEV strains are presumed to be transferred between JEV-endemic regions by movement of arthropod vectors and bird reservoirs. Nonetheless, we infer that fixation of an invading JEV strain into a new region is difficult unless the new strain possesses properties advantageous for virus growth and expansion (15). However, the genotype shift from GIII to GIa has occurred in East Asia since the 1990s, indicating that GIa must have had some sort of growth advantage over GIII that permitted its spreading to and expansion in these countries. Our findings that JEV strain invasion in Asia is infrequent could assist in public health decision-making regarding vaccine formulation and campaign strategies.