A

BOScRgSdgCtr©

K

1

tmc

s

s9T

1d

Journal of Clinical Virology 40 (2007) 135–141

Diversity of group A rotavirus strains circulating in Paraguay from2002 to 2005: Detection of an atypical G1 in South America

Gabriel I. Parra a,b,∗, Emilio E. Espınola a, Alberto A. Amarilla a, Juan Stupka c, MagalyMartinez a, Marta Zunini d, Maria E. Galeano a, Karina Gomes c,

Graciela Russomando a, Juan Arbiza b

a Departamento de Biologıa Molecular, Instituto de Investigaciones en Ciencias de la Salud, Universidad Nacional de Asuncion, Asuncion, Paraguayb Seccion Virologıa, Facultad de Ciencias, Universidad de la Republica, Montevideo, Uruguay

c Laboratorio de Gastroenteritis Virales, Departamento de Virologıa, INEI-ANLIS “Dr. Carlos G. Malbran”, Buenos Aires, Argentinad Hospital Regional de Ciudad del Este, Alto Parana, Paraguay

Received 6 February 2007; received in revised form 17 July 2007; accepted 17 July 2007

bstract

ackground: Group A rotaviruses are the main cause of severe gastroenteritis in children worldwide.bjectives: To survey human rotavirus strains circulating in Paraguay.tudy design: One hundred ninety-six rotavirus-positive fecal samples collected from children up to 5 years old, from 2002 to 2005, wereharacterized.esults: The most common G genotype detected was G9 (36.2%), followed by G1 (34.2%), G2 (11.7%) and G4 (8.7%). Changes in the Genotype frequency were observed from year to year. The G4 genotype was predominant in 2002; G1 in 2003; and G9 from 2004 to 2005.equence and phylogenetic analysis of the VP7 gene from Paraguayan G1 strains suggested that the high frequency of G1 in 2003 could beue to the introduction of an atypical sub-lineage. In addition, there were amino acid changes in the variable/antigenic regions of the VP7ene from G4 and G9 strains detected in different years.

onclusions: This study further indicates that antigenic pressure can drive the evolution of rotaviruses, and also suggests that a vaccine

hat protects against the most prevalent strains and its variants, will be necessary to elicit a protective immune response against the range ofotavirus types currently circulating in Paraguay.

2007 Elsevier B.V. All rights reserved.

wspb8

eywords: Group A rotavirus; Genetic diversity; Phylogenetic analysis

. Introduction

Group A rotaviruses are the main cause of severe gastroen-eritis in infants and young children worldwide, resulting in

ore than 454,000 deaths per year, mainly in developing

ountries (Parashar et al., 2006).

The rotavirus genome contains 11 segments of double-tranded RNA (dsRNA) surrounded by a triple layer capsid,

∗ Corresponding author. Present address: Department of Neuro-ciences/NC30, Lerner Research Institute, The Cleveland Clinic Foundation,500 Euclid Avenue, Cleveland, OH 44195, United States.el.: +1 216 444 6149; fax: +1 216 444 7927.

E-mail address: gabriel parra@hotmail.com (G.I. Parra).

P

(mdba(

386-6532/$ – see front matter © 2007 Elsevier B.V. All rights reserved.oi:10.1016/j.jcv.2007.07.006

hich can be separated by polyacrylamide gel electrophore-is (PAGE), yielding a characteristic profile, or “electro-herotype”. The outermost capsid is composed of VP4 (codedy gene segment 4) and VP7 (coded by gene segment 7, or, or 9 depending on the strain) proteins, which specify theand G serotypes/genotypes, respectively (Estes, 2001).Since rotavirus can reassort its gene segments

Desselberger et al., 2001; Gouvea and Brantly, 1995),any combinations of G and P genotypes are possible. To

ate, 42 different combinations of G and P genotypes haveeen detected, but four of them (G1P[8], G2P[4], G3P[8]nd G4P[8]) are the most prevalent in humans globallyGentsch et al., 2005; Santos and Hoshino, 2005).

136 G.I. Parra et al. / Journal of Clinical

Table 1G and P genotype combinations of Paraguayan rotavirus strains detectedfrom 2002 to 2005

G genotypes P genotypes

P[8] P[4] P[4] + P[8] P untypeablea Total

G1 50 17 67G2 21 1 2 24G4 16 1 17G9 63 1 5 7 76G1 + G4 1 1G1 + G9 1 1G untypeable 3 7 10

Total 134 22 6 34 196

gh22wydGrTbo

aeO1G

arwtR

2

2

twcttcseaao

2

Ciwzggan9wa2u

2

agarose gel and purified using the QIAquick Gel ExtractionKit (Qiagen, Valencia, CA). The sequencing was carried out

TD

Y

2222

T

a For 16 samples the 1st round amplicon from VP4 gene was obtained.

However, seasonal (or annual, depending on the geo-raphic area) changes in the predominant strain/genotypeave been detected in many geographic areas (Berois et al.,003; Bok et al., 2001; Parra et al., 2005; Yoshinaga et al.,006), the G1P[8] strain is by far the most commonly foundorldwide. In a study carried out during 19 consecutiveears in Italy, it was shown that the emergence or intro-uction of new variants (new lineages or sub-lineages) of1 strains could explain the continuous circulation of G1

otavirus in a given geographic area (Arista et al., 2006).hus, intraserotypic heterogenicity is a possible mechanismy which G1 strains have become the most prevalent strainf rotavirus worldwide.

Uncommon rotavirus strains, such as G5, G8, G9 and G12,re epidemiologically important in human disease (Gentscht al., 2005; Samajdar et al., 2006; Santos and Hoshino, 2005).n this regard, G9 strains were rarely detected before the mid-990s, but now is considered to be the fourth most prevalentgenotype (Gentsch et al., 2005; Santos and Hoshino, 2005).To expand our knowledge of what type of rotavirus strains

re circulating in Paraguay, we conducted a surveillance ofotavirus strains from children under 5 years of age in twoidely separated areas of Paraguay. We characterized a selec-

ion of rotavirus-positive samples from 2002 to 2005 byT-PCR and nucleotide sequencing.

bP

able 2istribution of G genotypes of Paraguayan rotavirus strains detected from 2002 to

ear No. rotavirus-positive/No.of samples collected

No. of samplestested

No. of G ge

G1

002 39/121 34 8 (23.5)003 158/469 57 50 (87.7)004 75/404 36 5 (13.9)005 170/594 69 4 (5.8)

otal 442/1588 196 67 (34.2)a Percentage based on tested samples.b Bold numbers indicate predominant type by year.c Includes one P[8]G1 + G4 and one G1 + G9P[8].d Includes one G2P[4 + 8] and five G9P[4 + 8].

Virology 40 (2007) 135–141

. Materials and methods

.1. Patients and samples

Fecal samples (N = 1588) were collected from children upo 5 years old with acute diarrhea who were outpatients orere hospitalized at one of two main private hospitals in the

apital city of Asuncion, or at one of three main hospitals fromhe Alto Parana Department located 330 km from Asuncion,oward the eastern border with Brazil. The samples wereollected between August 2002 and December 2005. Theamples were screened for the presence of rotavirus by RNAxtraction and polyacrylamide gel electrophoresis (PAGE),s previously described (Candia et al., 2003). This study waspproved by the Ethics Committee of the Research Institutef Health Sciences, National University of Asuncion.

.2. G and P typing

Viral RNA was extracted by TRIzol (Invitrogen, Carlsbad,A) following the manufacturer’s recommendations. The sil-

ca powder method was used to eliminate PCR inhibitorshen P and/or G type could not be determined with the TRI-

ol extraction (Boom et al., 1990). For G typing, the VP7ene was reverse-transcribed and amplified using a pair ofeneric primers (Beg9 and End9) (Gouvea et al., 1990). Then,second round of PCR was performed with a pool of inter-al primers (9T1-1, 9T1-2, 9T1-3P, 9T1-4, FT5, 9T-9B andCon1) (Das et al., 1994; Gouvea et al., 1994). The P typesere detected with a similar RT-PCR strategy (Gentsch et

l., 1992). Amplicons were analyzed by electrophoresis in a% agarose gel, stained with ethidium bromide and observednder UV light.

.3. Nucleotide sequencing

Amplicons from the full-length VP7 gene were run in an

y the dideoxynucleotide chain terminator method on an ABIrism 3100 automatic sequencer (Applied Biosystems, Fos-

2005

notypes (%)a

G2 G4 G9 Mixed G untypeable

0 15 (44.1)b 7 (20.6) 2 (5.9) c 2 (5.9)0 0 4 (7) 0 3 (5.3)0 0 29 (80.6) 0 2 (5.6)

23 (33.3) 2 (2.9) 31 (44.9) 6 (8.7)d 3 (4.3)

23 (11.7) 17 (8.7) 71 (36.2) 9 (4.6) 10 (5.1)

G.I. Parra et al. / Journal of Clinical

twE

2

avew(ob

3

3

iPapspspbmcG(

bcGbgpmG

FGNta1rBIR

Virology 40 (2007) 135–141 137

er City, CA). The nucleotide sequences reported in this paperere submitted to the GenBank database under the numbers:F179186–EF179201.

.4. Sequence and phylogenetic analysis

The Paraguayan sequences were aligned with those avail-ble in the GenBank database and compared using Bioedit7.0.1 (Hall, 1999). The phylogenetic relationships werestablished using neighbor-joining or parsimony methodsith the MEGA 3.1 and Phylip v3.65 software packages

Kumar et al., 2004), respectively. The statistical significancef the phylogenetic trees constructed was estimated by theootstrap method (1000 replicates).

. Results

.1. G and P typing

Of the 1588 samples obtained, 442 (27.8%) were pos-tive for rotavirus: 394 from Asuncion and 48 from Altoarana. One hundred and ninety-six were selected for Gnd P typing, based on dsRNA migration pattern (electro-herotype) differences, month, year and place of isolation,o as to cover a wide spectrum of circulating variants. Sam-les showing the same electropherotype and collected in theame month were randomly selected. In 159 (81.1%) sam-les, the P and G genotype could be successfully determined,ut for 3 (1.5%) samples, only the P genotype could be deter-ined and for 27 (13.8%) samples, only the G genotype

ould be determined. In 7 (3.6%) samples, neither the P norgenotype could be determined, despite several attempts

Table 1).There were no differences in the genotype frequencies

etween the two geographic areas under study. The mostommon strain was G9P[8] (32.1%; 63/196), followed by1P[8] (25.5%; 50/196), G2P[4] (10.7%; 21/196), and finallyy G4P[8] (8.2%; 16/196) (Table 1). Rotaviruses bearing theenotype G3 were not detected. It is noteworthy that one sam-

le showed an uncommon combination, G9P[4]. In addition,ixed infections were found: P[8]G1 + G4; P[8]G1 + G9;2P[4 + 8] and G9P[4 + 8].

ig. 1. Phylogenetic analysis of VP7 nucleotide sequences of genotype1 strains. The tree was constructed using the Kimura 2-parameter andeighbor-Joining methods using MEGA 3.1. Bootstrap values are shown at

he branch nodes (values < 65% not shown). The lineages and sub-lineagesre indicated at the right-hand side. The Paraguayan strains isolated in998–1999 and 2002–2003 are marked with a filled square and filled circle,espectively. Abbreviations for locations: Aus: Australia; Ban: Bangladesh;rz: Brazil; Chi: China; Cos: Costa Rica; Egy: Egypt; In: India; Ita: Italy; Isr:

srael; Jap: Japan; Kor: Korea; Mvd: Montevideo (Uruguay); Py: Paraguay;us: Russia.

1 linical

3

i2idgiti2i(

3

olVistliala2

3

(lscPstf122

3

(PGGPaiv

FGs

38 G.I. Parra et al. / Journal of C

.2. Annual distribution

Rotaviruses with the G4 genotype were predominantn 2002, accounting for 44.1% of the samples. From002 to 2003, rotaviruses with the G1 genotype increasedn frequency from 23.5 to 87.7%, and became the pre-ominant genotype in 2003. Rotaviruses with the G9enotype were detected each year, with frequencies rang-ng from 7 to 80.6%. From 2004 to 2005 it washe most common genotype detected. Rotaviruses bear-ng the G2 genotype were not detected from 2002 to004; however, a high frequency (33.3%) was detectedn 2005, when it was the second most common genotypeTable 2).

.3. Sequence analysis from Genotype G1

To determine if there are unique features in the VP7 proteinr any epidemiologic linkage to the previous strains circu-ating in Paraguay or worldwide, the coding region of theP7 gene (nt 49-1029) was sequenced from 10 G1 strains

solated in 2002 and 2003. Phylogenetic analysis, using 212equences downloaded from the GenBank database, groupedhe Paraguayan G1P[8] strains in two clusters, both within theineage II. The majority of the samples (9/10) were groupedn a cluster related to strains from sub-lineage IIb (Arista et

l., 2006). The remaining sample was grouped (with a lowevel bootstrap) with strains from sub-lineage IIc, IId and

Paraguayan strain detected in 1999 (Fig. 1) (Parra et al.,005).

aw2r

ig. 2. Amino acid substitutions from the variable regions 3, 4, 5 (antigenic region4 and G9 strains. Strains from other lineages and sub-lineages isolated worldwide

hown.

Virology 40 (2007) 135–141

.4. Sequence analysis from Genotype G4

The coding region of the VP7 gene was partially sequencednt 58-1003) from two G4 strains isolated in 2002. Phy-ogenetic analysis grouped the two G4 isolates within theub-lineage Ic (data not shown), previously described cir-ulating in Paraguay (Parra et al., 2005). Although thearaguayan G4 strains isolated in 2002 and 1998–1999howed a high degree of amino acid similarity (98.9%),hree amino acid substitutions were present in the strainsrom 2002 that were not seen in the samples from 1998 to999: 147Gly → Ala (variable region 7 [antigenic region B]),18Val → Ile (variable region 8 [antigenic region C]) and45Thr → Ile) (Fig. 2).

.5. Sequence analysis from Genotype G9

The coding region of the VP7 gene was partially sequencednt 67-1004) from five G9 strains, two isolated from Altoarana Department and three from Asuncion. All these9 strains grouped within the lineage III, along with the9P[8] strains detected in neighboring countries (Fig. 3). Thearaguayan strains detected from 2002 to 2004 had 11 aminocid substitutions, compared with strains previously detectedn Paraguay. Five of these substitutions were present in theariable regions. The change 29Ala → Val was found in vari-

ble region 2; the changes 44Ala → Val and 46Pro → Serere found in variable region 3; and 210Thr → Ala and21Ser → Asn were found in variable region 8 (antigenicegion C) (Fig. 2).

A), 7 (antigenic region B) and 8 (antigenic region C) from Paraguayan G1,are used for comparison. For each strain the lineage and/or sub-lineage is

G.I. Parra et al. / Journal of Clinical Virology 40 (2007) 135–141 139

Fig. 3. Phylogenetic analysis of VP7 nucleotide sequences of genotype G9 strains. The tree was constructed using the Kimura 2-parameter and Neighbor-Joiningmethods using MEGA 3.1. Bootstrap values are shown at the branch nodes (values < 60% not shown). The lineages are indicated at the right hand side. TheP squareo ns for loM .

4

t

araguayan strains isolated in 2000 and 2002–2004 are marked with a filledf origin, the associated P-type and year of isolation are shown. Abbreviatiow: Malawi; Py: Paraguay; Thai: Thailand; US: United States; Vn: Vietnam

. Discussion

There are two rotavirus vaccines currently available inhe market, one of them (Rotarix®; G1P[8]) developed

ooPt

and filled circle, respectively. For each strain (where available), the countrycations: Arg: Argentina; Aus: Australia; Brz: Brazil; In: India; Jap: Japan;

n the hypothesis of heterotypic protection and the otherne (Rotateq®; G1-G4P[8]) on homotypic protection (Ruiz-alacios et al., 2006; Vesikari et al., 2006). Therefore, one of

he considerations of policy makers in Paraguay is to under-

1 linical

sc

nGh

GwpGoapr

icfsfidt

sottstAltc(paGr(tekfcstuw

lsatci

4ot2s2c(saGbt

inMasaTtt(bt

A

rMtt

R

A

B

B

B

C

D

40 G.I. Parra et al. / Journal of C

tand the evolution and strain diversity of rotaviruses in thisountry to estimate which vaccine would be most effective.

In contrast with our previous work, in which the predomi-ant strain was G4P[8] (Parra et al., 2005), strains G1P[8] and9P[8] were the most frequently detected during 2002–2005,owever, most frequent strains changed from year to year.

All the strains showed the common combinations, i.e. G1,4 and G9 with P[8] and long electropherotype, and G2P[4]ith short electropherotype, with the exception of one sam-le. This sample was detected in 2005, is a combination of9P[4] with long electropherotype. In 2005, a high incidencef G2P[4] strains were co-circulating with G9P[8] strains,nd mixed infections were detected. Therefore, this couldrovide a good opportunity for the appearance of the rotaviruseassortants described above (Iturriza-Gomara et al., 2001).

As shown previously, mixed infections are rarely detectedn countries when one or a small number of genotypes cir-ulate in a given season (Kirkwood et al., 2003), but arerequently detected when many genotypes circulate at theame season (Unicomb et al., 1999). In accordance with thisnding, 2002 and 2005 were the only 2 years in which weetected mixed infections; these were also the only 2 yearshat no single genotype predominated.

During a survey carried out by our lab in 1998–1999 G1trains from lineage I were mainly detected. In addition, onlyne of the G1 strains sequenced in this work grouped withhe strains previously detected. Therefore, these data suggesthat the introduction of a new variant of G1 could be respon-ible for the high frequency of G1 in 2003. It is noteworthyhat all the Paraguayan strains sequenced had the amino acidsn at position 94, which is linked to the monotype 1a and

ineage II specificity. Moreover, the Paraguayan G1 strainshat grouped in the sub-lineage IIb had unique amino acidhanges; these same changes are also present in the lineage I41Ser → Phe), III (147Asn → Ser [antigenic region B]) andorcine G1 rotavirus (55Leu → Ile). Therefore, the aminocid differences within this sub-lineage could allow these1 strains to infect a population naıve to this variant. These

esults reinforce the recently published work of Arista et al.2006), where it was shown that the emergence or introduc-ion of new variants of G1 strains could be responsible for thepidemic peaks in a given season in Palermo, Italy. To ournowledge, this is the first report of circulation of G1 strainsrom sublineage IIb in South America. Even though that vac-ine trials have shown high levels of protection against G1trains, no sequence analyses of the strains circulating duringhe trials were reported. Therefore, further analyses should bendertaken in order to know if the currently available vaccinesere effective against strains from this atypical sub-lineage.The two G4 strains from 2002 grouped within the sub-

ineage Ic were epidemiologically linked to the previous G4trains circulating in Paraguay. In addition, the two amino

cid changes found in the antigenic sites B and C withinhe G4 strains detected in 2002 (Fig. 2), suggest that thesehanges could allow these G4 rotaviruses to escape from herdmmunity and persist in the population throughout the years.

D

Virology 40 (2007) 135–141

The emergent genotype G9 was present throughout theyears studied, and was responsible for a large number

f rotavirus infections in recent years in the region andhroughout the world (Parra et al., 2005; Santos and Hoshino,005; Santos et al., 2005; Stupka et al., 2007). Our VP7equence analysis of five G9 strains detected from 2002 to004 grouped within the lineage III, considered to be theirculating lineage of this emerging/reemerging genotypeRamachandran et al., 2000). Of note are the amino acid sub-titutions found between the G9 strains circulating in 2000nd 2002–2004. Thus, similar to the G1 strains, the G4 and9 strains could escape from the acquired herd immunityy intraserotypic variability and persist in a given populationhroughout the years.

Since one of the rotavirus vaccines currently availablen South America (Rotarix®; G1P[8]) does not share the Gor the P genotype with the rotavirus G2P[4] (O’Ryan andatson, 2006), it will be important to monitor the prevalence

nd persistence of the G2P[4] strains and/or reassortants thathare one of the DS-1 genes (exemplified here by G9P[4]),fter the introduction of this rotavirus vaccine in Paraguay.his study reinforces the notion that antigenic pressure drives

he evolution of rotaviruses (Arista et al., 2006), and suggestshat a vaccine that protects against the most prevalent strainsG1P[8], G2P[4], G4P[8] and G9P[8]) and its variants, mighte necessary to elicit a protective immune response againsthe rotavirus types currently circulating in Paraguay.

cknowledgments

We are grateful to Dr. Mitzi Baker for the language cor-ections and critical comments to the manuscript and Dr.

aria Elena Penaranda for her encouragement throughouthis study. This study was partially funded by a grant fromhe Sustainable Science Institute to GIP.

eferences

rista S, Giammanco GM, De Grazia S, Ramirez S, Lo Biundo C, ColombaC, et al. Heterogeneity and temporal dynamics of evolution of G1 humanrotaviruses in a settled population. J Virol 2006;80:10724–33.

erois M, Libersou S, Russi J, Arbiza J, Cohen J. Genetic variation in the VP7gene of human rotavirus isolated in Montevideo-Uruguay from 1996 to1999. J Med Virol 2003;71:456–62.

ok K, Castagnaro N, Borsa A, Nates S, Espul C, Fay O, et al. Surveillancefor rotavirus in Argentina. J Med Virol 2001;65:190–8.

oom R, Sol CJ, Salimans MM, Jansen CL, Wertheim-van Dillen PM, vander Noordaa J. Rapid and simple method for purification of nucleic acids.J Clin Microbiol 1990;28:495–503.

andia N, Parra GI, Chirico M, Velazquez G, Farina N, Laspina F, et al.Acute diarrhea in Paraguayan children population: detection of rotaviruselectropherotypes. Acta Virol 2003;47:137–40.

as BK, Gentsch JR, Cicirello HG, Woods PA, Gupta A, Ramachandran M,

et al. Characterization of rotavirus strains from newborns in New Delhi,India. J Clin Microbiol 1994;32:1820–2.

esselberger U, Iturriza-Gomara M, Gray JJ. Rotavirus epidemiologyand surveillance. Novartis Found Symp 2001;238:125–47 [discussion147–152].

linical

E

G

G

G

G

G

H

I

K

K

O

P

P

R

R

S

S

S

S

U

VZ, et al. Safety and efficacy of a pentavalent human-bovine (WC3) reas-

G.I. Parra et al. / Journal of C

stes M. Rotaviruses and their replication. In: Knipe DM, Howley PM,Griffin DE, Lamb RA, Martin MA, Roizman B, Straus SE, editors. Fieldsvirology. 4th ed. Philadelphia, PA: Lippincott/Williams & Wilkins; 2001.p. 1747–85.

entsch JR, Glass RI, Woods P, Gouvea V, Gorziglia M, Flores J, et al.Identification of group A rotavirus gene 4 types by polymerase chainreaction. J Clin Microbiol 1992;30:1365–73.

entsch JR, Laird AR, Bielfelt B, Griffin DD, Banyai K, Ramachandran M,et al. Serotype diversity and reassortment between human and animalrotavirus strains: implications for rotavirus vaccine programs. J InfectDis 2005;192(Suppl. 1):S146–59.

ouvea V, Brantly M. Is rotavirus a population of reassortants? TrendsMicrobiol 1995;3:159–62.

ouvea V, Glass RI, Woods P, Taniguchi K, Clark HF, Forrester B, et al.Polymerase chain reaction amplification and typing of rotavirus nucleicacid from stool specimens. J Clin Microbiol 1990;28:276–82.

ouvea V, Santos N, Timenetsky Mdo C. Identification of bovine and porcinerotavirus G types by PCR. J Clin Microbiol 1994;32:1338–40.

all T. BioEdit: a user-friendly biological sequence alignment editor andanalysis program for Windows 95/98/NT. Nucleic Acids Symp Ser1999;41:95–8.

turriza-Gomara M, Isherwood B, Desselberger U, Gray J. Reassortment invivo: driving force for diversity of human rotavirus strains isolated in theUnited Kingdom between 1995 and 1999. J Virol 2001;75:3696–705.

irkwood C, Bogdanovic-Sakran N, Palombo E, Masendycz P, Bugg H,Barnes G, et al. Genetic and antigenic characterization of rotavirusserotype G9 strains isolated in Australia between 1997 and 2001. J ClinMicrobiol 2003;41:3649–54.

umar S, Tamura K, Nei M. MEGA3: integrated software for molecularevolutionary genetics analysis and sequence alignment. Brief Bioinform

2004;5:150–63.

’Ryan M, Matson DO. New rotavirus vaccines: renewed optimism. J Pedi-atr 2006;149:448–51.

arashar UD, Gibson CJ, Bresee JS, Glass RI. Rotavirus and severe child-hood diarrhea. Emerg Infect Dis 2006;12:304–6.

Y

Virology 40 (2007) 135–141 141

arra GI, Bok K, Martinez V, Russomando G, Gomez J. Molecular charac-terization and genetic variation of the VP7 gene of human rotavirusesisolated in Paraguay. J Med Virol 2005;77:579–86.

amachandran M, Kirkwood CD, Unicomb L, Cunliffe NA, Ward RL, BhanMK, et al. Molecular characterization of serotype G9 rotavirus strainsfrom a global collection. Virology 2000;278:436–44.

uiz-Palacios GM, Perez-Schael I, Velazquez FR, Abate H, Breuer T,Clemens SC, et al. Safety and efficacy of an attenuated vaccine againstsevere rotavirus gastroenteritis. N Engl J Med 2006;354:11–22.

amajdar S, Varghese V, Barman P, Ghosh S, Mitra U, Dutta P, et al.Changing pattern of human group A rotaviruses: emergence of G12 asan important pathogen among children in eastern India. J Clin Virol2006;36:183–8.

antos N, Hoshino Y. Global distribution of rotavirus serotypes/genotypesand its implication for the development and implementation of an effec-tive rotavirus vaccine. Rev Med Virol 2005;15:29–56.

antos N, Volotao EM, Soares CC, Campos GS, Sardi SI, Hoshino Y. Pre-dominance of rotavirus genotype G9 during the 1999, 2000 and 2002seasons among hospitalized children in the city of Salvador, Bahia,Brazil: implications for future vaccine strategies. J Clin Microbiol2005;43:4064–9.

tupka JA, Parra GI, Gomez J, Arbiza J. Detection of human rotavirus G9P[8]strains circulating in Argentina: phylogenetic analysis of VP7 and NSP4genes. J Med Virol 2007;79:838–42.

nicomb LE, Podder G, Gentsch JR, Woods PA, Hasan KZ, Faruque AS,et al. Evidence of high-frequency genomic reassortment of group Arotavirus strains in Bangladesh: emergence of type G9 in 1995. J ClinMicrobiol 1999;37:1885–91.

esikari T, Matson DO, Dennehy P, Van Damme P, Santosham M, Rodriguez

sortant rotavirus vaccine. N Engl J Med 2006;354:23–33.oshinaga M, Phan TG, Nguyen TA, Yan H, Yagyu F, Okitsu S, et al. Chang-

ing distribution of group A rotavirus G-types and genetic analysis of G9circulating in Japan. Arch Virol 2006;151:183–92.