Phylogenetic Investigation of Dengue Virus Type 2 Isolated in Malaysia Hui-Yee Chee and Sazaly AbuBakar# Department of Medical Microbiology, Faculty of Medicine, University of Malaya, Lembah Pantai, 50604 Kuala Lumpur, Malaysia by
Abstract Dengue virus type-2 (DEN-2) has been isolated in Malaysia for more than three decades. The virus caused two major outbreaks in the early 1990s and late 1990s. Phylogenetic analyses performed using available E/NS1 junction sequences identified two DEN-2 genotypes: DEN-2 Asian 1 and DEN-2 Cosmopolitan. DEN-2 Cosmopolitan/Malaysia is the predominant genotype comprising more than 80% of the total isolates. Two different clades of DEN-2 Cosmopolitan/Malaysia genotype were identified. Clade I consisted of mainly the older isolates, whereas Clade II consisted of the more recent isolates, including that responsible for both the major DEN-2 outbreaks. Two different strains of DEN2 Cosmopolitan/Malaysia genotype were involved in the outbreaks, yet both strains appeared to share a common ancestral lineage with isolates from the early 1970s. Isolates from the late 1990s showed higher sequence similarities to the late-1960s isolates than the early-1990s isolates. These findings raised the possibility that the different DEN-2 strains noted over the last three decades in Malaysia might have evolved from a pre existing DEN-2 gene pool. Keywords: Dengue, genotypes, outbreak, phylogenetic, Malaysia.
Introduction The existence of dengue fever (DF) in Malaysia was first described following an outbreak in Penang, a northern state in the Malaysian peninsula, in 1902(1). In 1954, the first-ever isolate of dengue virus, a DEN-1 was obtained during an outbreak in Kuala Lumpur(2). Eighteen years later (1962), the first case of dengue haemorrhagic fever (DHF) was reported(3). Dengue has since become endemic in Malaysia, with an average incidence of about 400 to 7,000 #
cases of DF/DHF annually. All four serotypes are found co-circulating in the country with DEN-2 and DEN-3 being the most commonly isolated serotypes. A recent analysis of the trend and pattern of dengue in Malaysia over the last few decades showed that major DF/DHF outbreaks occurred in a cyclical pattern of approximately eight years and it involved mainly DEN-1, DEN-2 and DEN-3(4). The cycle is thought to be associated with the switching of the predominant dengue virus serotypes in the population. For example, it
For correspondence: sazaly@um.edu.my
100
Dengue Bulletin – Vol 27, 2003
Phylogenetic Investigation of Dengue Virus Type 2 Isolated in Malaysia
was reported that a major outbreak of DF/DHF involving predominantly DEN-1 that occurred in Malaysia between 1997 and 1998 was preceded by high levels of DEN-3 in the population. Similarly, the outbreak in 1990-1991 involving mainly DEN-2 was preceded by a high occurrence of DEN-1(4). The outbreaks involving DEN-4, however, occurred only in the 1960s and it has not been reported ever since. The disappearance of outbreaks involving DEN-4 coincided with its replacement by other dengue virus serotypes, especially DEN-2 as the dominant serotype, in Malaysia. Though DEN-2 emerged as a major serotype in the late 1960s until the early 1970s, it has since persisted in the country and has been responsible for the major outbreaks that occurred in late 1980s to early 1990s and the subsequent outbreak in late 1990s to early 2000s. It is not presently known, however, if a similar DEN-2 strain was responsible for these outbreaks and whether that the strain has also persisted in the population over the last three decades. In an effort to address these issues, we sequenced the E/NS1 gene junction of twelve DEN-2 isolates from the two outbreaks and examined the phylogenetic relationships of these isolates against those of the earlier isolates.
initially serotyped using specific monoclonal antibody and subsequently confirmed using multiplex RT-PCR(5) and DNA sequencing. The E/NS1 sequences were amplified using two pairs of oligonucleotide primers that amplify DEN-2 sequence from nucleotide positions 1777-2512 and 2422-3477. The amplification conditions, parameters and DNA sequencing were performed as previously described(6). Contigs and consensus sequences were assembled using Sequencher version 4.0.5 (Gene Codes Corporation, USA) and the nucleotide sequences generated after removing the extra nucleotides were used. Phylogenetic analyses were performed as previously described(6) with additional analysis using the maximum likelihood method as implemented in PHYLIP 3.6(7). The E/NS1 junction sequence was used in the present study as unlike other genes, this gene region is not under the immune selection pressure; hence, is most suitable as a marker for longterm evolutionary study(8). Furthermore, several studies, apart from our own (data not shown), have shown that phylogenetic trees constructed using this genome region are robust and correspond well with trees constructed using other genes such as E(9), NS1(10) and the full-length genome(11).
Materials and methods Three sets of DEN-2 isolates obtained from DF/DHF patients treated at the University of Malaya Medical Center (UMMC) during 1989-1991, 1995-1996 and 1998-2000 were used in the present study. The five isolates each from 1989-1991 (early 1990s) and 1998-2000 (late 1990s) were from the two outbreaks involving DEN-2, whereas the 1995-1996 isolates represented the inbetween outbreak strains. The isolates were Dengue Bulletin – Vol 27, 2003
Results A total of 63 DEN-2 E/NS1 sequences were used in the present study. These included 27 sequences obtained from the GenBank which included a sylvatic DEN-2 isolate (PM33974), a representative DEN-2 American genotype (Genotype I; Ven2), a representative DEN-2 Asian 1 genotype (K0010), a representative DEN-2 Asian 2 genotype (Genotype II; NGC) and a representative DEN-2 Cosmopolitan genotype (Genotype IV; 975). The remaining 101
Phylogenetic Investigation of Dengue Virus Type 2 Isolated in Malaysia
sequences were from the Malaysian isolates, which included 24 sequences obtained from Fong et al,(12) one sequence from the GenBank (LF5-99) and 12 sequences obtained from the present study (Table). A neighbour-joining (NJ) tree depicting the phylogenetic relationships of the isolates was obtained using DEN-2 sylvatic isolate (PM33974) as the outward group (Figure 1). The NJ-generated trees were very similar to that generated with the maximum-likelihood (ML) method, with minor branch swapping at the terminal nodes (data not shown). From the phylogenetic tree, it was noted that the Malaysian DEN-2 isolates were grouped into two different genotypes, Asian 1 and Cosmopolitan, with approximately 83.3% (30/36) of the isolates grouped into the Cosmopolitan genotype(13). The Cosmopolitan genotype included isolates from as early as 1968 to as recently as 2000. The Cosmopolitan genotype, DEN-2, also included two isolates from Sarawak (East Malaysia), KS30 and LF5-99 and three isolates from outside the Klang Valley.
Isolates belonging to the Asian 1 genotype included those obtained in 1982, 1987, 1988, 1991 and 1995. The average percentage of nucleotide divergence between the two DEN-2 genotypes, Asian 1 and Cosmopolitan, was 7.1% (range 4% to 11%) with the average percentage of nucleotide divergence within the Cosmopolitan and Asian 1 genotypes at 3.5% (range 0% to 7%) and 2.8% (range 0% to 7%), respectively. Incidentally, the DEN-2 Asian 1 genotype consisted mainly of isolates from Thailand and Myanmar, whereas the Cosmopolitan genotype included isolates from other parts of the world. Among the 12 isolates sequenced in the present study, 11 were of Cosmopolitan genotype, of which nine were from DHF patients and two were from DF patients (isolates MY91-99133 and MY00-21986). The relationships between genotypes and clinical manifestations, however, could not be ascertained as no additional information on the disease severity was available for most of the isolates.
Table. DEN-2 isolates used in the study *Isolate MY89-83995 MY89-88549 MY91-94129 MY91-95062 MY91-99133 MY95-2303 MY96-5176 MY98-15996 MY98-17933 MY99-19575 MY99-19586 MY00-21986 Year of isolation 1989 1989 1991 1991 1991 1995 1996 1998 1998 1999 1999 2000 Location of isolation Klang Valley Klang Valley Klang Valley Klang Valley Klang Valley Klang Valley Klang Valley Klang Valley Klang Valley Klang Valley Klang Valley Klang Valley Diagnosis DHF DHF DHF DHF DF DF DHF DHF DHF DHF DHF DF GenBank accession no. AJ556803 AJ556804 AJ556805 AJ556806 AJ556807 AJ556808 AJ556809 AJ556810 AJ556811 AJ556812 AJ556813 AJ556814
* Other DEN-2 E/NS1 sequences used in the present study are available from GenBank and Fong et al, 1998(12).
102
Dengue Bulletin – Vol 27, 2003
Phylogenetic Investigation of Dengue Virus Type 2 Isolated in Malaysia
Figure 1. Phylogenetic tree of DEN-2 derived from E/NS1 sequence comparisons†
The tree was constructed using the neighbour-joining (NJ) method with DEN-2 sylvatic isolate, PM33974, as the outgroup. Significant bootstrap value is shown as percentages derived from 1,000 resamplings and the scale reflects the number of nucleotide substitution per site along the branches. The twelve newly-sequenced Malaysian DEN-2 isolates are shown in boxes. The reference DEN-2 sequences used for the construction of the phylogenetic tree were obtained from the GenBank and Fong et al, 1998(12). Ch-China; Gn-Guinea; In-Indonesia; Mm-Myanmar; MyMalaysia; NG-New Guinea; SL-Sri Lanka; Th-Thailand; Vz-Venezuela; ?-year of isolation not known.
†
Dengue Bulletin – Vol 27, 2003
103
Phylogenetic Investigation of Dengue Virus Type 2 Isolated in Malaysia
Figure 2. Phylogenetic relationships amongst DEN-2 Cosmopolitan genotype isolated in Malaysia‡
‡
The phylogenetic tree was derived from E/NS1 sequence comparisons and constructed using the neighbour-joining (NJ) method. The tree is shown with DEN-2 sylvatic isolate, PM33974 as the outgroup. Significant bootstrap value is shown as percentages derived from 1,000 re samplings and the scale reflects the number of nucleotide substitution per site along the branches.
104
Dengue Bulletin – Vol 27, 2003
Phylogenetic Investigation of Dengue Virus Type 2 Isolated in Malaysia
A phylogenetic tree drawn using only the Malaysian Cosmopolitan genotype isolates (Figure 2) showed that the isolates were grouped into at least two clades (Clade I and Clade II), with an average percentage of nucleotide divergence of 2.2% (range 1% to 3%). Clade I showed two clusterings (Ia & Ib) of the isolates with isolates from the late 1960s (P7-845 and P9-573) and 1970s (P70104, P73-987, P75-559, P78-182, P78-612 and V32491) grouped into cluster Ib. Within this cluster all the isolates from late 1970s (1976–1980) were grouped together away from the 1968-1969 isolates. The early1970s isolates (1970-1974) formed the outer group. The bootstrap values supporting these groupings, however, were low. The other cluster (Ia) within Clade I comprised isolates from 1980s (V48171 and V89821) and also 1990s (TG2, V03857 and MY98-15996). Similar to Clade I, two main clusterings of isolates were also noted in Clade II (IIa & IIb) with the main cluster (IIb) comprising isolates from 1986–1996 (early 1990s) and the other cluster consisted of isolates mainly from 1999 to 2000 (late 1990s). The separation of these two later clusters received a strong bootstrap support (>75%), suggesting that the two isolates were different. Hence, this supports the involvement of two different strains of DEN-2 Cosmopolitan genotype in the two major DEN-2 outbreaks in Malaysia. Comparisons of the representative consensus sequences of each cluster showed only four nucleotide differences at nucleotide positions 75, 85, 99 and 234 that differentiated the early- and late-1990s DEN-2 isolates. These substitutions, however, did not result in any amino acid changes. On the other hand, comparisons made between the representative consensus
sequences of isolates from 1968-1969 with those of the early- and late-1990s isolates revealed nine and five nucleotide differences, respectively. The nine nucleotides substitution noted in the sequence of the early-1990s isolates occurred at positions 9, 15, 75, 85, 99, 183, 186, 228 and 234 whereas the five nucleotides substitution for the late-1990s isolates occurred at positions 9, 15, 183, 186 and 228. It was apparent from these results that isolates from the late 1990s shared higher-sequence similarities to the much older DEN-2 isolates than that of the early1990s isolates.
Discussion The findings from the present study revealed the presence of at least two main DEN-2 genotypes in Malaysia. The Cosmopolitan genotype(13), also known as genotype IV(8), is the dominant genotype comprising more than 80% of all DEN-2 isolates of Malaysia. This genotype includes two of the oldest DEN-2 isolates, P7-845 and P9-573, which were isolated in 1968 and 1969, respectively. The other minor genotype present in Malaysia belongs to the Asian 1 genotype, also known as genotype II. This later genotype comprised mainly of isolates from Thailand and Myanmar; hence, raised the possibility that the Malaysian Asian 1 isolates may not be native to Malaysia, but perhaps were associated with the MalaysiaThailand cross-border economic activities, as Malaysia shares it northern land border with southern Thailand. It is interesting to note that despite the long borders and vigorous cross-border economic activities and travels, the DEN-2 genotypes dominant in the two 105
Dengue Bulletin – Vol 27, 2003
Phylogenetic Investigation of Dengue Virus Type 2 Isolated in Malaysia
countries are different. While the Asian 1 genotype appeared localized mainly in Thailand, the Cosmopolitan isolates dominant in Malaysia have been isolated in as far as Burkina Faso in West Africa. The Cosmopolitan genotype has also been isolated in Singapore, India, Sri Lanka, Seychelles, Somalia, Uganda and Australia(13), all of which once used to be part of the British colonial empire. Whether the spread or the lack of it between the two genotypes was related to historical reasons involving movement of people between different countries and regions, Malaysia, unlike Thailand which has never been under the British rule, was a British colony until 1957 or involved different mosquito vectors remains to be investigated. The phylogenetic analyses of the dominant Cosmopolitan/Malaysian genotype DEN-2 isolated in Malaysia over the last three decades did not suggest a clear evolution of the DEN-2 virus from a single common ancestral lineage. Instead, two clades with two sub-clusters each were noted with the 1968-1969 isolates belonging to Clade I, which, incidentally, has an isolate from 1983 as the apparent ancestor, and Clade II which appeared to have arisen from a 1973 isolate. While the bootstrap support for the tree branches was weak, there was a clear trend suggesting a clustering of the isolates by the year of isolation; with the bulk of the isolates from 1970s found in Clade I and the isolates mainly from the early 1990s and late 1990s well segregated in Clade II. Using the calculated estimate of the rate of nucleotide substitution for DEN-2 (7.19 X 10-4 subs/site/year)(14), and with the assumption that the late-1960s isolates consensus sequence as the most recent 106
common ancestor, it was noted that the isolates from the early 1990s had approximately three times more nucleotide substitutions (9 nucleotides) within this short sequence over 20 years when compared to the expected values of 3.4 nucleotides. This is in contrast to the isolates from the late 1990s, which had only five nucleotide substitutions over 30 years, which is within the predicted substitution rate of 5.1 nucleotides. From this simple estimate, it appeared that the sequence of isolates from the early 1990s underwent a much faster rate of nucleotide changes than expected. As the lineage of these isolates did not prevail a decade later, one wonders if this high rate of nucleotide changes within the E/NS1 junction is mirrored also in other genes and hence affected the ability of the virus to replicate efficiently. Furthermore, the appearance in the late-1990s outbreak of isolates with higher sequence similarity to the late-1960s isolates also raised the possibility that the different DEN-2 strains noted over the last three decades in Malaysia may have evolved from the pre-existing DEN-2 gene pool. As intra- and inter-typic recombination amongst dengue viruses are more common than once thought(6,15), it will be interesting to monitor whether DEN-2 genotype Asian 1 sequence will eventually end up in the Malaysian DEN-2 gene pool.
Acknowledgement This study received financial support from the Ministry of Science, Technology, and the Environment, Malaysia, IRPA Grants #0602-03-0303/0304, and the University of Malaya Research Grants.
Dengue Bulletin – Vol 27, 2003
Phylogenetic Investigation of Dengue Virus Type 2 Isolated in Malaysia
References 1. 2. Skae FMT. Dengue fever in Penang. British Medical Journal, 1902, 2: 1581-1582. Smith CEG. Isolation of three strains of type 1 dengue virus from a local outbreak of the disease in Malaysia. Journal of Hygiene, 1956, 54: 569580. Rudnick A, Tan EE, Lucas JK and Omar M. Mosquito-borne haemorrhagic fever in Malaya. British Medical Journal, 1965, 1: 1269-1272. AbuBakar S and Shafee N. Outlook of dengue in Malaysia: A century later. Malaysian Journal of Pathology, 2002, 24: 23-27. Seah CL, Chow VT, Tan HC and Chan YC. Rapid, single-step RT-PCR typing of dengue viruses using five NS3 gene primers. Journal of Virological Methods, 1995, 51: 193-200. AbuBakar S, Wong P-F and Chan Y-F. Emergence of dengue virus type 4 genotype IIA in Malaysia. Journal of General Virology, 2002, 83: 2437-2442. Felsenstein J. PHYLIP-Phylogeny Inference Package (Version 3.2). Cladistics, 1989, 5: 164-166. Rico-Hesse R. Molecular evolution and distribution of dengue viruses type 1 and 2 in nature. Virology, 1990, 174: 479-493. Lewis JA, Chang GJ, Lanciotti RS, Kinney RM, Mayer LW and Trent DW. Phylogenetic relationships of dengue-2 viruses. Virology, 1993, 197: 216-224. 10. Blok J, Gibbs AJ, McWilliam SM and Vitarana UT. NS 1 gene sequences from eight dengue-2 viruses and their evolutionary relationships with other dengue-2 viruses. Archives of Virology, 1991, 118: 209-223. 11. Leitmeyer KC, Vaughn DW, Watts DM, Salas R, de Chacon VI, Ramos C and Rico-Hesse R. Dengue virus structural differences that correlate with pathogenesis. Journal of Virology, 1999, 73: 47384747. 12. Fong MY, Koh C-L and Lam S-K. Molecular epidemiology of Malaysian DEN-2 viruses isolated over twenty-five years (1968-1993). Research in Virology, 1998, 149: 457-464. 13. Twiddy SS, Farrar JJ, Chau NV, Wills B, Gould EA, Gritsun T, Lloyd G and Holmes EC. Phylogenetic relationships and differential selection pressures among genotypes of dengue-2 virus. Virology, 2002, 298: 63-72. 14. Twiddy SS, Holmes EC and Rambaut A. Inferring the rate and time-scale of dengue virus evolution. Molecular Biology and Evolution, 2003, 20: 122129. 15. Worobey M, Rambaut A and Holmes EC. Widespread intra-serotype recombination in natural populations of dengue virus. Proceedings of the National Academic of Sciences USA, 1999, 96: 7352-7357.
3.
4.
5.
6.
7. 8.
9.
Dengue Bulletin – Vol 27, 2003
107