Infection, Genetics and Evolution 10 (2010) 746–754

Complete genomic analysis of a Bangladeshi G1P[8] rotavirus strain detectedin 2003 reveals a close evolutionary relationship with contemporary humanWa-like strains

Mustafizur Rahman a,*, Jelle Matthijnssens b, Farjana Saiada a, Zahid Hassan a, Elisabeth Heylen b,Tasnim Azim a, Marc Van Ranst b

a Laboratory of Virology, ICDDR,B: International Centre for Diarrhoeal Disease Research, Bangladesh, Mohakhali, Dhaka 1212, Bangladeshb Laboratory of Clinical and Epidemiological Virology, Rega Institute for Medical Research, University of Leuven, Leuven, Belgium

A R T I C L E I N F O

Article history:

Received 19 January 2010

Received in revised form 15 April 2010

Accepted 26 April 2010

Available online 2 May 2010

Keywords:

Rotavirus

Complete genome

Sequence

G1P[8]

Wa-like

Bangladesh

Vaccine

A B S T R A C T

More than 120 variants of rotavirus strains with different VP7 (G type) and VP4 (P type) combinations are

reported thus far. Among them Wa-like G1P[8] rotaviruses are the most common human strains

worldwide. However, characterization of their entire genome complement is limited to a few old

prototype strains, and no complete genome data for any G1P[8] strain isolated in the last decade are

available. Both the currently licensed rotavirus vaccines RotarixTM and RotaTeqTM possess the G1 and

P[8] specificities. Therefore, comprehensive genetic information of the currently circulating G1P[8]

strain is important to assess the impact of rotavirus vaccines on the circulating rotavirus strains. Here we

report the complete genome sequence of a G1P[8] rotavirus strain Dhaka16-03 isolated in 2003 from a

Bangladeshi child hospitalized with severe diarrhea. Based on a full-genome classification system,

Dhaka16-03 was shown to posses the typical Wa-like genotype constellation: G1-P[8]-I1-R1-C1-M1-A1-

N1-T1-E1-E1-H1. The strain was phylogenetically more closely related to contemporary human

rotavirus strains (isolated in the 2000s) with a range of G and P-genotypes than to those of the prototype

G1P[8] strains. Since the vaccine strains are developed based on strains isolated several decades ago, it is

important to know how much the vaccine strains differ from the currently circulating G1P[8] and other

Wa-like strains. Our complete genome characterization of a recent G1P[8] strain will be helpful to assess

the ongoing rotavirus vaccine trials and their implementation programs in the forthcoming years.

� 2010 Elsevier B.V. All rights reserved.

Contents lists available at ScienceDirect

Infection, Genetics and Evolution

journal homepage: www.elsev ier .com/ locate /meegid

1. Introduction

Group A rotaviruses are one of the most common causes ofinfantile diarrhea (Parashar et al., 2006). Rotaviruses are the onlymammalian agents known to contain 11 segments of doublestranded (ds) RNA (Estes and Kapikian, 2007). Each dsRNA segmentencodes one protein with the exception of the 11th gene segmentwhich encodes two. There are six structural proteins (VP1–VP4,VP6 and VP7) and six non-structural proteins (NSP1–NSP6)encoded by the rotavirus genome. The gene segments range insize from 664 (segment 11) to 3302 base pairs (segment 1), withthe total genome containing approximately 18,500 base pairs(Estes and Cohen, 1989).

* Corresponding author at: Virology Laboratory, Laboratory Sciences Division,

ICDDR,B: International Centre for Diarrhoeal Disease Research, Bangladesh, 68,

Shaheed Tajuddin Ahmed Sharani, Dhaka 1212, Bangladesh.

Tel.: +880 2 8811751–60x2409; fax: +880 2 8812529.

E-mail address: mustafizur@icddrb.org (M. Rahman).

1567-1348/$ – see front matter � 2010 Elsevier B.V. All rights reserved.

doi:10.1016/j.meegid.2010.04.011

In 1978, the International Committee on Taxonomy for Virusesofficially adopted Rotavirus as a genus in the family Reoviridae

(Matthews, 1979). Since then, rotaviruses have been classified indifferent ways, such as serotyping, serogrouping, subgrouping,genotyping, genogrouping, electropherotyping etc (Estes andKapikian, 2007). However, a dual classification system, based onthe two outer capsid proteins VP7 (G genotype) and VP4 (Pgenotype), has most widely been used. Recently, a genotypingsystem based on complete ORF nucleotide sequences of all the 11gene segments has been proposed and a rotavirus classificationworking group (RCWG) has been formed (Matthijnssens et al.,2008a, 2008b). In this classification system, for each of therotavirus gene segments a nucleotide cut-off percentage wascalculated to distinguish different genotypes. According to thisclassification system, 23 G (glycosylated, VP7), 31 P (proteasesensitive, VP4), 13 I (intermediate capsid shell, VP6), 6 R (RNApolymerase, VP1), 6 C (core shell, VP2), 7 M (methyltransferase,VP3), 16 A (antagonist of interferon, NSP1), 6 N (NTPase, NSP2), 8 T(translation enhancer, NSP3), 12 E (enterotoxin, NSP4), and 8 H(phosphoprotein, NSP5) genotypes have been published (Mat-

M. Rahman et al. / Infection, Genetics and Evolution 10 (2010) 746–754 747

thijnssens et al., 2009a, 2009b; Schumann et al., 2009; Solberget al., 2009; Trojnar et al., 2009; Ursu et al., 2009).

A considerable diversity among rotaviruses strains circulatingin different geographical locations and time periods has beenobserved (Matthijnssens et al., 2008c). However, five majorgenotype combinations, G1P[8], G2P[4], G3P[8], G4P[8], andG9P[8] are currently the major human rotavirus strains all overthe world. Among them, G1P[8] strains are the most predominantin humans (Castello et al., 2004; Desselberger et al., 2006; Gentschet al., 2005; Santos and Hoshino, 2005).

In Bangladesh, G1P[8] strains have been one of the mostcommon rotavirus strains during the last two decades (Rahmanet al., 2007b; Unicomb et al., 1993, 1999). These studies alsodetected G2P[4], G4P[8] and G9P[8] as common strains with someuncommon G11 and G12 strains (Rahman et al., 2007b). Athorough investigation on G12 rotaviruses revealed that theacquirement of gene segments from human-adapted rotavirusesthrough reassortments might allow G12 strains to better propa-gate in humans, and hence to develop into an important emerginghuman pathogen (Rahman et al., 2007a).

Rotavirus vaccine efficacy trial on RotaTeqTM has beenconducted in Bangladesh and RotaRixTM trial is ongoing. Themonovalent live-attenuated human rotavirus vaccine RotaRixTM

was originally developed by tissue culture passage of a wild-typehuman rotavirus isolate G1P[8] (strain 89-12), and thus representsthe most common of the human rotavirus VP7 (G1) and VP4 (P[8])antigens. The vaccine was first licensed in Mexico and theDominican Republic in 2004 and received a favorable recommen-dation from the U.S. Food and Drug Administration’s (FDA)Vaccines and Related Biological Products Advisory Committee(VRBPAC) in 2008 (Dennehy, 2007; http://www.fda.gov/NewsE-vents/Newsroom/PressAnnouncements/2008/ucm116875.htm).

Although both licensed vaccines contain the G1 and P[8]specificities, no complete genome data are available for any wild-type G1P[8] rotavirus strain isolated in the last decade! Therefore,we sequenced the complete genome (11 gene segments) of aG1P[8] rotavirus strain Dhaka16-03 isolated during a G1P[8]dominated rotavirus season 2002–2003 (Rahman et al., 2007b) andclassified the virus based on the recently proposed rotavirusclassification system and guidelines of the RCWG. A thoroughphylogenetic analysis was also conducted to reveal the relation-ship of Dhaka16-03 with currently circulating rotavirus strains andolder reference strains.

2. Materials and methods

2.1. Detection of rotavirus strain

Rotavirus strain Dhaka16-03 was detected by enzyme immu-noassay in a routine surveillance system on diarrheal patientsattended the hospital of International Centre for DiarrhoealDisease Research, Bangladesh (ICDDR,B) in 2003. In brief, rotavirusantigen (group A rotavirus-specific VP6 protein) was detected inthe stool specimens using a solid-phase sandwich-type enzymeimmunoassay modeled after the Dakopatts commercial kit(Dakopatts, Copenhagen, Denmark), incorporating rabbit hyper-immune antisera produced at ICDDR,B and an anti-humanrotavirus–horseradish peroxidase conjugate (Unicomb et al.,1999).

2.2. Electropherotyping

One-hundred microliters of 2% stool suspensions in phosphate-buffered saline was treated with sodium acetate and extractedwith an equal volume of phenol:chloroform:isoamylalcohol(25:24:1) mixture. The extracted RNA was tested for E-type by

polyacrylamide gel electrophoresis (PAGE). RNA migration patternof the strain was visualized on the silver stained gel as described byHerring et al. (1982).

2.3. RNA extraction

Total viral RNA was extracted from stool specimen of thepatient using the QIAamp Viral RNA minikit (QIAGEN/Westburg,Leusden, The Netherlands) according to the manufacturer’sinstructions.

2.4. RT-PCR

The extracted RNA was denatured at 97 8C for 5 min, andreverse transcriptase PCR (RT-PCR) was carried out using a QIAGENOneStep RT-PCR kit (QIAGEN/Westburg). Primers for the amplifi-cation of the 11 gene segments were used as described byMatthijnssens and colleagues (Matthijnssens et al., 2008a). The RT-PCR was carried out with an initial reverse transcription step of30 min at 45 8C, followed by PCR activation at 95 8C for 15 min, 40cycles of amplification with a final extension of 7 min at 70 8C. Thecycle conditions for the amplification of VP1, VP2, VP3 and VP4genes were 30 sec at 94 8C, 30 sec at 45 8C and 6 min at 70 8C; forthe other gene segments the conditions were 30 sec at 94 8C, 30 secat 45 8C and 2.5 min at 72 8C.

2.5. Nucleotide sequencing

Nucleotide sequencing was performed for the 11 complete genesegments, VP1–VP4, VP6, VP7, NSP1–NSP5. The PCR products werepurified with the QIAquick PCR purification kit (QIAGEN/West-burg) and sequenced with the ABI Prism BigDye terminator cyclesequencing reaction kit (Applied Biosystems Group, Foster City,CA) on an ABI Prism 3100 automated sequencer (AppliedBiosystems Group) as described previously (Matthijnssens et al.,2006a).

2.6. Determination of the 50 and 30 terminal sequences

The 50 and 30 terminal sequences of the 11 gene segments ofstrain Dhaka16-03 were determined using a modified version ofthe single-primer amplification method (Lambden et al., 1992).Briefly, a modified amino-linked oligonucleotide was ligated to the30 end of both strands of the viral dsRNA with T4 RNA Ligase(Promega, Leiden, the Netherlands). RT-PCR with primers com-plementary to linker sequence and appropriate gene specificprimers (based on the known internal sequences of each segment)was carried out (Matthijnssens et al., 2008d). The amplifiedproducts were purified and sequenced as described above.

2.7. Sequence analysis

The chromatogram sequencing files were inspected usingChromas 2.23 (Technelysium, Queensland, Australia) and consensussequences were prepared using SeqMan II (DNASTAR, Madison, WI).Multiple sequence alignments were calculated using ClustalX 1.81(Thompson et al., 1997). Sequences were manually edited in theGeneDoc, version 2.6.002 alignment editor (Nicholas et al., 1997).

2.8. Percentage of identity

All nucleotide sequences of rotavirus strains available inGenBank database were included for the comparison betweendifferent genotypes. Genetic distances were calculated by theKimura-2 correction parameter at the nucleotide level using MEGAsoftware (Tamura et al., 2007).

M. Rahman et al. / Infection, Genetics and Evolution 10 (2010) 746–754748

2.9. Phylogenetic analysis

Initially, all rotavirus nucleotide sequences available inGenBank database (http://www.ncbi.nlm.nih.gov/genbank) wereacquired and aligned. Phylogenetic analyses were conducted at thenucleotide level using the MEGA version 4.1 software (Tamuraet al., 2007). The dendrograms were constructed using theneighbor-joining method. In the final phylogenetic trees, thestrains which belonged to the respective genotypes to our strainDhaka16-03 with an outgroup were shown.

2.10. Assignment of genotypes

The genotypes of each of the 11 genome segments weredetermined according to the genotyping recommendations of theRotavirus Classification Working Group (RCWG) using the onlineavailable RotaC rotavirus genotyping tool (Maes et al., 2009).

2.11. Accession numbers

The nucleotide sequences of Dhaka16-03 were submitted to theGenBank under the accession numbers DQ492669–DQ492679.Other accession numbers used in the trees are provided inSupplement 1.

Fig. 1. Average nucleotide sequence identity percentages of genes encoding viral structur

of rotavirus strains belonging to different genotypes available in GenBank. Identities wer

M1–M7, (d) VP4 genotypes P[1]-P[31], (e) VP6 genotypes I1–I11, and VP7 genotypes G1–

established cut-off identity percentages defining different genotypes.

3. Results

3.1. Case history of the patient

A VP6-specific enzyme-linked immunosorbent assay detectedrotavirus antigen in stool specimen from a 7-month-old childattended the hospital. Polyacrylamide gel electrophoresis (PAGE)of the RNA segments extracted from the stool specimen alsoshowed a typical long electropherotype of group A rotaviruses (4-2-3-2). The specimen was negative for the other pathogensincluded in the hospital surveillance system. The child infected bystrain Dhaka16-03 had severe watery diarrhea and abdominalpain. The frequency of vomiting was more than 10 times per dayand a mild fever was recorded. The stool frequency had been morethan 21 times during the previous 24 h. Only oral rehydrationsolution was given to the patient and he could leave the hospitalafter 2 days.

3.2. Sequence based classification of Dhaka16-03

3.2.1. VP1–VP3

The complete sequence of gene segment 1 (encoding VP1) ofDhaka16-03 was compared with VP1 gene segments available inGenBank. The average of percentage identity between Dhaka16-03

al proteins (VP1–VP4, VP6 and VP7) of Dhaka16-03 with other corresponding genes

e shown for: (a) VP1 genotypes R1–R6, (b) VP2 genotypes C1–C6, (c) VP3 genotypes

G23. The most similar genotype to Dhaka16-03 is shaded. Dashes lines indicate the

M. Rahman et al. / Infection, Genetics and Evolution 10 (2010) 746–754 749

and the established R-genotypes indicated that the VP1 gene ofDhaka16-03 was most similar (average nt identity 95.5%) to strainsbelonging to the R1 genotype with the highest nucleotide (nt)identity (99.2%) with a Bangladeshi Wa-like G12P[8] strainDhaka25-02. The average identities with other genotypes (R2–R6 VP1 genotypes) were less than the VP1 cut-off value of 83%(Fig. 1a). Phylogenetic analysis confirmed that Dhaka16-03clustered with the Wa-like strains (R1 genotype) and was distantlyrelated to other genotypes (Fig. 2a). It was also apparent thatseveral recently identified human G9 and G12 strains (Dhaka25-02, B4633-03, B3458, Matlab13-03 and Dhaka12-03) clusteredclosely together with Dhaka16-03, whereas several Wa-likeprototype strains (Wa, KU, ST3, P, WI61) clustered more distantly.

The VP2 gene segment of a Bangladeshi G12 strain Dhaka12-03was the most similar (97% nt identity) to our G1P[8] strain, and avery high identity (average 95.2%) with strains belonging to the C1genotype was found (Fig. 1b). The identity between Dhaka16-03and other VP2 genotypes (C2–C5) was lower than the established

Fig. 2. Phylogenetic analyses of viral structural protein genes: (a) VP1, (b) VP2, (c) VP3 a

sequences. The numbers adjacent to the nodes represent the percentage of bootstrap su

lower than 70% are not shown. Only the genotype in which the strain under investigatio

prototype and historical strains detected before 1993 are in red and 2000s green. Hu,

84% cut-off value. Phylogenetic analysis clustered Dhaka16-03with other recently isolated rotavirus strains belonging to the C1genotype (Fig. 2b).

The average identity for the VP3 gene of Dhaka16-03 was highestwith strains belonging to the M1 VP3 genotype (93.4%) and less than81% with other M-genotypes (Fig. 1c). The most similar strain toDhaka16-03 was a Bangladeshi G12P[6] strain Matlab13-03 (99% ntidentity). Phylogenetic analysis confirmed that Dhaka16-03belonged to the M1 genotype (Fig. 2c). Although the VP3 genesegment of our strain is phylogenetically more closely related torecently isolated G9 and G12 strains, unexpectedly, a closerrelationship was found with the corresponding genes of prototypeG3 and G4 human strains P and ST3 (96% nt identity) than to theprototype G1 strains Wa and D (92% nt identity) (Fig. 2c).

3.2.2. VP4

Among the 31 currently published P-genotypes, P[8] was themost similar to Dhaka16-03 (average nt identity 93.7%). The other

nd (d) VP6 that include nucleotide sequences of Dhaka16-03 and other respective

pport (of 1000 replicates) for the clusters to the right of the node. Bootstrap values

n clusters is shown completely and one genotype has been added as outgroup. The

human; Po, porcine. Dhaka16-03 is boxed.

Fig. 3. Phylogenetic analyses of two outer capsid protein genes: (a) VP4 and (b) VP7 that include nucleotide sequences of Dhaka16-03 and other respective sequences. Only

the P[8] and P[4] genotypes in the VP4 tree and G1 and G4 (as outgroup) in VP7 tree are shown. The strains detected before 1995 are in red and 2000s green. Hu, human; Po,

porcine. Dhaka16-03 is boxed.

M. Rahman et al. / Infection, Genetics and Evolution 10 (2010) 746–754750

genotypes except for P[4] showed nt identities below the cut offvalue of 80% differentiating the VP4 genotypes (Fig. 1d). GenotypesP[8] and P[4] were initially described as subtypes (P1A and P1B) ofa single serotype and it is therefore not surprising that P[4] strainsalso show a high nt identity (average 87.5%) with Dhaka16-03(Fig. 1d). This high nucleotide identity between strains belongingto the P[8] and P[4] genotypes was reported previously (Mat-thijnssens et al., 2008a). However, the phylogenetic tree (showingonly the P[8] and P[4] genotypes), confirms that strains Dhaka16-03, belongs to the P[8] genotype (Fig. 3a) and is most similar toBangladeshi strain Dhaka25-02 (99% nt identity), Belgian strainB3458 (99% nt identity) and Korean strain CAU202 (98% ntidentity) isolated during 2002-03, but only distantly related to theprototype G1P[8] strain Wa isolated in 1974 (90% nt identity).

3.2.3. VP6

Dhaka16-03 belonged to the I1 genotype as it showed thehighest identity with strains belonging to the I1 VP6 genotype(average nt identity 95.5%) (Fig. 1e). This finding was confirmed inthe VP6 phylogenetic tree where Dhaka16-03 clustered withstrains belonging to the I1 genotype (Fig. 2d). Again the strain wasclosely related to concurrent Bangladeshi G12 strains Matlab13-03and Dhaka12-03, and more distantly related to the prototype G1strains Wa, D and KU.

3.2.4. VP7

Thai and Indian G1 strains (CMH042/04 and ISO-4, 99% ntidentity) were the most similar to our G1P[8] strain based on theVP7 gene. The VP7 of Dhaka16-03 had an average nt identity of 96%with other G1 strains (Fig. 1f). Genotypes G2–G23 showed lower

than 80% average nt identity with our strain. In the phylogenetictree Dhaka16-03 was placed with recently isolated G1 rotavirusstrains from Bangladesh, India and Thailand. The prototype G1strains, KU, D, and Wa and other G1 strain isolated in the 1970s and1980s were found in a different lineages distantly related to the2000s G1 strains (Fig. 3b). Unusual porcine and bovine G1 strainsclustered in a separate distantly related branch far away fromhuman strains.

Antigenic regions (A–C) of the outer capsid protein VP7 ofDhaka16-03 were compared to prototype and recent G1 strains(Fig. 4). Prototype strains Wa and K8 had Asn and Asp at amino acidpositions 94 and 97 respectively; whereas the recent strainsincluding our strains contain Ser and Glu. In antigenic region B onesubstitution (Ser147Asn) was identified in all recent strains as wellas in some historical strains. Amino acid differences Met217Thrand Ile218Val were also detected in the recent strains in antigenicregion C. Also in the non-antigenic regions some mutations atamino acid positions 41, 71, 266, 281 and 291 are present inBangladeshi recent strains compared to prototype strains (data notshown).

3.2.5. NSP1–NSP5

There are 16 NSP1 genotypes (A1–A16) described of which A1was the most similar to Dhaka16-03 (87.5% average nt identity)(Fig. 5a). A Bangladeshi strain Dhaka25-02 (99.1% nt identity) wasthe most similar to our strain based on the NSP1 gene sequence.The average nt identities of Dhaka16-03 with other genotypes (A2–A16) were less than 79%. Phylogenetic analysis also indicated thatour strain clustered with strains belonging to the A1 genotype(Fig. 6a). Dhaka16-03 belongs to the N1 genotype based on the ORF

Fig. 5. Average nucleotide sequence identity percentages of genes encoding viral non-structural proteins (NSP1–NSP5) of Dhaka16-03 with other corresponding genes of

rotavirus strains belonging to different genotypes. Identities were shown for: (a) NSP1 genotypes A1–A16, (b) NSP2 genotypes N1–N6, (c) NSP3 genotypes T1–T8, (d) NSP4

genotypes E1–E12 and (e) NSP5 genotypes H1–H8.

Fig. 4. Comparison of VP7 amino acid sequences of Dhaka16-03 (bold) with prototype (boxed), historical (collected during 1985–1992) and recent (the 2000s) G1 rotaviruses.

Consensus sequence is based on corresponding amino acid sequence of G1P[8] prototype strain Wa.

M. Rahman et al. / Infection, Genetics and Evolution 10 (2010) 746–754 751

of the NSP2 gene as it showed the highest average nt identity of93.9% with strains belonging to the N1 genotype (Fig. 5b). The mostclosely related strains were Bangladeshi strains Dhaka6 (G11),Dhaka-25-02 (G12), and the Belgian strain B4633-03 (G12)(Fig. 6b). Others genotypes (N2–N6) showed lower identities (lessthan cut off value 85%).

Dhaka16-03 belongs to the T1 NSP3 genotype with an averageidentity of 97% to this genotype (Fig. 5c). Although our strainclustered with other human strains, interestingly there were someporcine and bovine strains including prototype porcine strainGottfried and OSU placed in the same genotype (Fig. 6c). Based onthe ORF sequence of NSP4, Dhaka16-03 was classified as belongingto the E1 genotype as it was most similar to an Indian strain J23170isolated in Kolkata in 2004, and again closely related to recent G12strains isolated in Bangladesh (Figs. 5d and 6d). Some porcine andequine strains in the phylogenetic tree also clustered in the same

genotype; however, they were placed in different branchesdistantly related to human branches (Fig. 6d). The average NSP5identity of our G1P[8] strain was highest with strains belonging tothe NSP5 H1 genotype (Fig. 5e). The other genotypes (H2–H8)showed lower identities (65%-89%) with Dhaka16-03. Phylogenet-ically the recently isolated Wa-like G9 and G12 strains were againfound to cluster closely with Dhaka16-03 (Fig. 6e).

4. Discussion

Historically, rotaviruses have been classified in several differentways. However, none of them covered the full range of rotavirusgenetic strain diversity. Since each of the proteins encoded bydifferent gene segments plays a vital role in rotavirus replicationcycle, assembly of viroplasm and pathogenesis of the virus (Ruizet al., 2009), complete characterization of strains is very important.

Fig. 6. Phylogenetic analyses of non-structural protein genes NSP1–NSP5 that include nucleotide sequences of Dhaka16-03 and other respective sequences. Only the genotype

in which the strain under investigation clusters is shown completely and one genotype has been added as outgroup. The protype and historical strains detected before 1993

are in red and 2000s green. Hu, human; Po, porcine. Dhaka16-03 is boxed.

M. Rahman et al. / Infection, Genetics and Evolution 10 (2010) 746–754752

M. Rahman et al. / Infection, Genetics and Evolution 10 (2010) 746–754 753

Therefore, full-genome sequencing and classification is very wellsuited to distinguish one virus from another, especially forsegmented viruses such as rotaviruses and ensures a precisemolecular characterization. It is noteworthy that complete genomesequences are available for a limited number of human rotavirusstrains and most of them are unusual or novel strains such as ahuman lapine-like G3P[14] strain, a human bovine-like G6P[6]strain, human G8P[4], G8P[6] and G8P[8] strains, a humanneonatal G10P[11] strain, G3P[3] human feline-like strains,recently emerging human G11, G12 and P[14] strains and so on(Banyai et al., 2009; Matthijnssens et al., 2006a, 2008d, 2006b;Pietsch et al., 2009; Rahman et al., 2007a; Ramani et al., 2009;Tsugawa and Hoshino, 2008). Very recently the complete genomesof 51 human G3P[8] strains isolated in the United states between1976 and 1991 have been published (McDonald et al., 2009).However, currently circulating common human rotavirus strainssuch as G1P[8], G2P[4], G3P[8], G4P[8], G9P[8] have been largelyignored while the 11 segments of just a few of the old prototypestrains have been reported (Heiman et al., 2008; Matthijnssenset al., 2008a, 2008c). Among the common human rotavirus strainsG1P[8] has remained globally the most common G-P genotypecombination in both developed and developing countries for a longtime. Therefore we sequenced and analysed the complete genomeof such a common strain Dhaka16-03 from Bangladesh isolated in2003.

Several important findings could be deduced from this study.First, complete genetic characterization demonstrates that ourstrain Dhaka16-03 is completely a Wa-like rotavirus strain, similarto reference strains Wa, D, KU (G1), P, 51 strains isolated in USAbetween 1976 and 1991 (G3), ST3 (G4) and WI61 (G9), and nosegments from DS-1-like, Au-1 like or animal rotavirus strainswere present in these strains (Heiman et al., 2008; Matthijnssenset al., 2008a; McDonald et al., 2009). Secondly, the geneconstellation of our G1P[8] strain Dhaka16-03 was shown to bephylogenetically very closely related to currently circulatingrotavirus strains, such as US6668 (G1P[8]), B3458 (G9P[8]),Dhaka22-01 (G11P[8]), and B4633-03, Dhaka25-02, Matlab13-03, Dhaka12-03 and US6588 (G12P[8]/P[6]), suggesting thatunderneath a diverse set of VP7 and VP4 outer capsid proteinswhich indicate several reassortment events between strains, aconserved Wa-like machinery is present (Freeman et al., 2009;Matthijnssens et al., 2009b; Rahman et al., 2007a). G1, G3, G4, G9and G12 rotavirus strain constitute the large majority (more than90%) of human rotaviruses indicating that Wa-like gene constella-tion may have the ability to propagate extremely well in thehuman host and has been most successful backbone of rotavirusesall over the world (Castello et al., 2004; Desselberger et al., 2006;Gentsch et al., 2005; Santos and Hoshino, 2005).

Obviously, many evolutionary changes have occurred in differentgene segments over time and thus the current strains have becomemore distantly related to the prototype strains. This information iscrucial for vaccine evaluation programs since changes in aminoacids, particularly in the antigenic sites of the outer capsid proteins,but also in other structural and non-structural rotavirus proteinsmay theoretically lead over time to reduced vaccine efficacy or evento vaccine failure. Nucleotide changes in the antigenic regions of VP7genes of G1 strains from other parts of the world were foundresponsible for escaping neutralization in several studies (Diwakarlaand Palombo, 1999; Jin et al., 1996; Kirkwood et al., 1993). Weidentified quite a few amino acid changes in the antigenic regions ofVP7 genes of currently circulating G1 strains when compared withprototype strains (Fig. 4). Whether these changes have any effect onthe immunological reactivity in cell culture or in animal model is yetto be determined.

Although, the complete genome sequence of the RotarixTM

vaccine strain 89-12 (Bernstein, 2007) has not yet been released, it

can be expected to be rather closely related to the older referencestrain, and less closely to the currently circulating Wa-like strains,as the original vaccine strain was isolated more than two decadesago. However, several vaccine trials have shown a very goodprotection against a variety of currently circulating (Wa-like) Ggenotypes and have led to a massive reduction in the number ofdeaths and hospitalization due to rotavirus infection (Bernsteinet al., 2002, 1999; Ward et al., 2006). However, the continuingaccumulation of point mutations in the entire rotavirus genomeunder a vaccine induced selective pressure will be important tomonitor as this might influence vaccine efficiency over time(Matthijnssens et al., 2009a). Sequence data of the vaccine strainsin RotaTeqTM will be published soon (personnel communication),but also for these 5 reassortant strains the G1, G2, G3, G4 and P[8]proteins are likely to be related to old reference strains, and less tothe currently circulating rotavirus strains. Therefore, continuousmonitoring of changes in the genetic make-up of circulating strainsis required and investigation on the evolutionary rate for differentgene segments over time would be important to evaluate andupdate the vaccine strains over time if necessary.

Acknowledgement

This research study was funded by ICDDR,B and Rega Institutefor Medical Research, University of Leuven, Belgium. JM wassupported by an FWO (Fonds voor Wetenschappelijk Onderzoek)postdoctoral fellowship. ICDDR,B acknowledges with gratitude thecommitment of Rega Institute for Medical Research to the Centre’sresearch efforts. ICDDR,B also gratefully acknowledges thefollowing donors which provide unrestricted support to theCentre’s research efforts: Australian Agency for InternationalDevelopment (AusAID), Government of the People’s Republic ofBangladesh, Canadian International Development Agency (CIDA),Embassy of the Kingdom of the Netherlands (EKN), SwedishInternational Development Cooperation Agency (Sida), SwissAgency for Development Cooperation (SDC), and the Departmentfor International Development, UK (DFID).

Appendix A. Supplementary data

Supplementary data associated with this article can be found, in

the online version, at doi:10.1016/j.meegid.2010.04.011.

References

Banyai, K., Martella, V., Molnar, P., Mihaly, I., Van Ranst, M., Matthijnssens, J., 2009.Genetic heterogeneity in human G6P[14] rotavirus strains detected in Hungarysuggests independent zoonotic origin. J. Infect. 59, 213–215.

Bernstein, D.I., 2007. RIX4414 (Rotarix): a live attenuated human rotavirus vaccine.J. Pediatria 83, 193–195.

Bernstein, D.I., Sack, D.A., Reisinger, K., Rothstein, E., Ward, R.L., 2002. Second-yearfollow-up evaluation of live, attenuated human rotavirus vaccine 89-12 inhealthy infants. J. Infect. Dis. 186, 1487–1489.

Bernstein, D.I., Sack, D.A., Rothstein, E., Reisinger, K., Smith, V.E., O’Sullivan, D.,Spriggs, D.R., Ward, R.L., 1999. Efficacy of live, attenuated, human rotavirusvaccine 89-12 in infants: a randomised placebo-controlled trial. Lancet 354,287–290.

Castello, A.A., Arvay, M.L., Glass, R.I., Gentsch, J., 2004. Rotavirus strain surveillancein Latin America: a review of the last nine years. Pediatr. Infect. Dis. J. 23, S168–172.

Dennehy, P.H., 2007. Rotavirus vaccines – an update. Vaccine 25, 3137–3141.Desselberger, U., Wolleswinkel-van den Bosch, J., Mrukowicz, J., Rodrigo, C., Gia-

quinto, C., Vesikari, T., 2006. Rotavirus types in Europe and their significance forvaccination. Pediatr. Infect. Dis. J. 25, S30–41.

Diwakarla, C.S., Palombo, E.A., 1999. Genetic and antigenic variation of capsidprotein VP7 of serotype G1 human rotavirus isolates. J. Gen. Virol. 80 (Pt 2),341–344.

Estes, M.K., Cohen, J., 1989. Rotavirus gene structure and function. Microbiol. Rev.53, 410–449.

Estes, M.K., Kapikian, A.Z., 2007. Rotaviruses. In: Knipe, D.M., Howley, P.M., Griffin,D.E., Lamb, R.A., Martin, M.A., Roizman, B., Straus, S.E. (Eds.), Fields Virology.fifth ed. Lippincott Williams & Wilkins, Philadelphia, pp. 1917–1974.

M. Rahman et al. / Infection, Genetics and Evolution 10 (2010) 746–754754

Freeman, M.M., Kerin, T., Hull, J., Teel, E., Esona, M., Parashar, U., Glass, R.I., Gentsch,J.R., 2009. Phylogenetic analysis of novel G12 rotaviruses in the United States: amolecular search for the origin of a new strain. J. Med. Virol. 81, 736–746.

Gentsch, J.R., Laird, A.R., Bielfelt, B., Griffin, D.D., Banyai, K., Ramachandran, M., Jain,V., Cunliffe, N.A., Nakagomi, O., Kirkwood, C.D., Fischer, T.K., Parashar, U.D.,Bresee, J.S., Jiang, B., Glass, R.I., 2005. Serotype diversity and reassortmentbetween human and animal rotavirus strains: implications for rotavirus vaccineprograms. J. Infect. Dis. 192, S146–159.

Heiman, E.M., McDonald, S.M., Barro, M., Taraporewala, Z.F., Bar-Magen, T., Patton,J.T., 2008. Group A human rotavirus genomics: evidence that gene constella-tions are influenced by viral protein interactions. J. Virol..

Herring, A.J., Inglis, N.F., Ojeh, C.K., Snodgrass, D.R., Menzies, J.D., 1982. Rapiddiagnosis of rotavirus infection by direst detection of viral nucleic acid insilver-stained polyacrylamide gels. J Clin Microbiol 16, 473–477.

Jin, Q., Ward, R.L., Knowlton, D.R., Gabbay, Y.B., Linhares, A.C., Rappaport, R., Woods,P.A., Glass, R.I., Gentsch, J.R., 1996. Divergence of VP7 genes of G1 rotavirusesisolated from infants vaccinated with reassortant rhesus rotaviruses. Arch.Virol. 141, 2057–2076.

Kirkwood, C., Masendycz, P.J., Coulson, B.S., 1993. Characteristics and location ofcross-reactive and serotype-specific neutralization sites on VP7 of human Gtype 9 rotaviruses. Virology 196, 79–88.

Lambden, P.R., Cooke, S.J., Caul, E.O., Clarke, I.N., 1992. Cloning of noncultivatablehuman rotavirus by single primer amplification. J. Virol. 66, 1817–1822.

Maes, P., Matthijnssens, J., Rahman, M., Van Ranst, M., 2009. RotaC: a web-basedtool for the complete genome classification of group A rotaviruses. BMCMicrobiol. 9, 238.

Matthews, R.E., 1979. The classification and nomenclature of viruses. Summary ofresults of meetings of the International Committee on Taxonomy of Viruses inThe Hague, September 1978. Intervirology 11, 133–135.

Matthijnssens, J., Bilcke, J., Ciarlet, M., Martella, V., Banyai, K., Rahman, M., Zeller, M.,Beutels, P., Van Damme, P., Van Ranst, M., 2009a. Rotavirus disease andvaccination: impact on genotype diversity. Future Microbiol. 4, 1303–1316.

Matthijnssens, J., Ciarlet, M., Heiman, E., Arijs, I., Delbeke, T., McDonald, S.M.,Palombo, E.A., Iturriza-Gomara, M., Maes, P., Patton, J.T., Rahman, M., Van Ranst,M., 2008a. Full genome-based classification of rotaviruses reveals a commonorigin between human Wa-Like and porcine rotavirus strains and human DS-1-like and bovine rotavirus strains. J. Virol. 82, 3204–3219.

Matthijnssens, J., Ciarlet, M., Rahman, M., Attoui, H., Banyai, K., Estes, M.K., Gentsch,J.R., Iturriza-Gomara, M., Kirkwood, C.D., Martella, V., Mertens, P.P., Nakagomi,O., Patton, J.T., Ruggeri, F.M., Saif, L.J., Santos, N., Steyer, A., Taniguchi, K.,Desselberger, U., Van Ranst, M., 2008b. Recommendations for the classificationof group A rotaviruses using all 11 genomic RNA segments. Arch. Virol. 153,1621–1629.

Matthijnssens, J., Potgieter, C.A., Ciarlet, M., Parreno, V., Martella, V., Banyai, K.,Garaicoechea, L., Palombo, E.A., Novo, L., Zeller, M., Arista, S., Gerna, G., Rahman,M., Van Ranst, M., 2009b. Are human P[14] rotavirus strains the result ofinterspecies transmissions from sheep or other ungulates that belong to themammalian order Artiodactyla? J. Virol. 83, 2917–2929.

Matthijnssens, J., Rahman, M., Ciarlet, M., Van Ranst, M., 2008c. Emerging humanrotavirus genotypes. In: Palombo, A.E. (Ed.), Viruses in the Environment.Research Signpost, Trivandrum, India.

Matthijnssens, J., Rahman, M., Martella, V., Xuelei, Y., De Vos, S., De Leener, K.,Ciarlet, M., Buonavoglia, C., Van Ranst, M., 2006a. Full genomic analysis ofhuman rotavirus strain B4106 and lapine rotavirus strain 30/96 providesevidence for interspecies transmission. J. Virol. 80, 3801–3810.

Matthijnssens, J., Rahman, M., Van Ranst, M., 2008d. Two out of the 11 genes of anunusual human G6P[6] rotavirus isolate are of bovine origin. J. Gen. Virol. 89,2630–2635.

Matthijnssens, J., Rahman, M., Yang, X., Delbeke, T., Arijs, I., Kabue, J.P., Muyembe,J.J., Van Ranst, M., 2006b. G8 rotavirus strains isolated in the Democratic

Republic of Congo belong to the DS-1-like genogroup. J. Clin. Microbiol. 44,1801–1809.

McDonald, S.M., Matthijnssens, J., McAllen, J.K., Hine, E., Overton, L., Wang, S.,Lemey, P., Zeller, M., Van Ranst, M., Spiro, D.J., Patton, J.T., 2009. Evolutionarydynamics of human rotaviruses: balancing reassortment with preferred ge-nome constellations. PLoS Pathogens 5, e1000634.

Nicholas, K.B., Nicholas, H.B., Deerfield, D.W., 1997. GeneDoc: analysis and visuali-zation of genetic variation. Embnet News 4 .

Parashar, U.D., Gibson, C.J., Bresse, J.S., Glass, R.I., 2006. Rotavirus and severechildhood diarrhea. Emerg. Infect. Dis. 12, 304–306.

Pietsch, C., Petersen, L., Patzer, L., Liebert, U.G., 2009. Molecular characteristics ofGerman G8P[4] rotavirus strain GER1H-09 suggest that a genotyping andsubclassification update is required for G8. J. Clin. Microbiol. 47, 3569–3576.

Rahman, M., Matthijnssens, J., Yang, X., Delbeke, T., Arijs, I., Taniguchi, K., Iturriza-Gomara, M., Iftekharuddin, N., Azim, T., Van Ranst, M., 2007a. Evolutionaryhistory and global spread of the emerging G12 human rotaviruses. J. Virol. 81,2382–2390.

Rahman, M., Sultana, R., Ahmed, G., Nahar, S., Hassan, Z.M., Saiada, F., Podder, G.,Faruque, A.S., Siddique, A.K., Sack, D.A., Matthijnssens, J., Van Ranst, M., Azim, T.,2007b. Prevalence of G2P[4] and G12P[6] rotavirus, Bangladesh. Emerg. Infect.Dis. 13, 18–24.

Ramani, S., Iturriza-Gomara, M., Jana, A.K., Kuruvilla, K.A., Gray, J.J., Brown, D.W.,Kang, G., 2009. Whole genome characterization of reassortant G10P[11] strain(N155) from a neonate with symptomatic rotavirus infection: identification ofgenes of human and animal rotavirus origin. J. Clin. Virol..

Ruiz, M.C., Leon, T., Diaz, Y., Michelangeli, F., 2009. Molecular biology of rotavirusentry and replication. Sci. World J. 9, 1476–1497.

Santos, N., Hoshino, Y., 2005. Global distribution of rotavirus serotypes/genotypesand its implication for the development and implementation of an effectiverotavirus vaccine. Rev. Med. Virol. 15, 29–56.

Schumann, T., Hotzel, H., Otto, P., Johne, R., 2009. Evidence of interspecies transmis-sion and reassortment among avian group A rotaviruses. Virology 386, 334–343.

Solberg, O.D., Hasing, M.E., Trueba, G., Eisenberg, J.N., 2009. Characterization ofnovel VP7, VP4, and VP6 genotypes of a previously untypeable group A rotavi-rus. Virology 385, 58–67.

Tamura, K., Dudley, J., Nei, M., Kumar, S., 2007. MEGA4: Molecular EvolutionaryGenetics Analysis (MEGA) software version 4.0. Mol. Biol. Evol. 24, 1596–1599.

Thompson, J.D., Gibson, T.J., Plewniak, F., Jeanmougin, F., Higgins, D.G., 1997. TheCLUSTAL_X windows interface: flexible strategies for multiple sequence align-ment aided by quality analysis tools. Nucleic Acids Res. 25, 4876–4882.

Trojnar, E., Otto, P., Johne, R., 2009. The first complete genome sequence of a chickengroup A rotavirus indicates independent evolution of mammalian and avianstrains. Virology 386, 325–333.

Tsugawa, T., Hoshino, Y., 2008. Whole genome sequence and phylogenetic analysesreveal human rotavirus G3P[3] strains Ro1845 and HCR3A are examples ofdirect virion transmission of canine/feline rotaviruses to humans. Virology.

Unicomb, L.E., Bingnan, F., Rahim, Z., Banu, N.N., Gomes, J.G., Podder, G., Munshi,M.H., Tzipori, S.R., 1993. A one-year survey of rotavirus strains from threelocations in Bangladesh. Arch. Virol. 132, 201–208.

Unicomb, L.E., Podder, G., Gentsch, J.R., Woods, P.A., Hasan, K.Z., Faruque, A.S.,Albert, M.J., Glass, R.I., 1999. Evidence of high-frequency genomic reassortmentof group A rotavirus strains in Bangladesh: emergence of type G9 in 1995. J. Clin.Microbiol. 37, 1885–1891.

Ursu, K., Kisfali, P., Rigo, D., Ivanics, E., Erdelyi, K., Dan, A., Melegh, B., Martella, V.,Banyai, K., 2009. Molecular analysis of the VP7 gene of pheasant rotavirusesidentifies a new genotype, designated G23. Arch. Virol. 154, 1365–1369.

Ward, R.L., Kirkwood, C.D., Sander, D.S., Smith, V.E., Shao, M., Bean, J.A., Sack, D.A.,Bernstein, D.I., 2006. Reductions in cross-neutralizing antibody responses ininfants after attenuation of the human rotavirus vaccine candidate 89-12. J.Infect. Dis. 194, 1729–1736.