Infection, Genetics and Evolution 29 (2015) 118–126

Contents lists available at ScienceDirect

Infection, Genetics and Evolution

journal homepage: www.elsevier .com/locate /meegid

Increased detection of G3P[9] and G6P[9] rotavirus strainsin hospitalized children with acute diarrhea in Bulgaria

http://dx.doi.org/10.1016/j.meegid.2014.11.0111567-1348/� 2014 Elsevier B.V. All rights reserved.

⇑ Corresponding author at (formerly): National Reference Laboratory of Entero-viruses, Department of Virology, National Center of Infectious and ParasiticDiseases, 44A, Stoletov Blvd., Sofia 1233, Bulgaria. Tel.: +359 2 931 23 22x247;fax: +359 2 943 30 75.

E-mail address: zornitsavmbg@yahoo.com (Z. Mladenova).1 The ZM and SN were equally contributed to the investigation reported in the

present study and to the preparation of the draft.

Zornitsa Mladenova a,b,⇑,1, Sameena Nawaz b,1, Balasubramanian Ganesh c, Miren Iturriza-Gomara d

a National Center for Infectious and Parasitic Diseases, Sofia, Bulgariab Public Health of England, London, United Kingdomc National Institute of Cholera and Enteric Diseases, Kolkata, Indiad Institute of Infection and Global Health, University of Liverpool, Liverpool, United Kingdom

a r t i c l e i n f o a b s t r a c t

Article history:Received 11 June 2014Received in revised form 23 September2014Accepted 12 November 2014Available online 20 November 2014

Keywords:RotavirusHuman strainReassortantZoonotic transmissionRotavirus vaccine

Rotavirus severe disease in children is now vaccine-preventable and the roll-out of vaccination programsglobally is expected to make a significant impact in the reduction of morbidity and mortality in children<5 years of age. Rotavirus is also a pathogen of other mammals and birds, and its segmented RNA genomecan lead to the emergence of new or unusual strains in human population via interspecies transmissionand reassortment events. Despite the efficacy and impact of rotavirus vaccine in preventing severe diar-rhea, the correlates of protection remain largely unknown. Therefore, rotavirus strain surveillance before,during and after the introduction of immunization programs remains a crucial for monitoring rotavirusvaccine efficacy and impact. In this context, molecular characterization of 1323 Bulgarian rotavirusstrains collected between June 2010 and May 2013 was performed. A total of 17 strains of interest wereanalyzed by partial sequence analysis. Twelve strains were characterized as G3P[9] and G6P[9] of poten-tial animal origin. Phylogenetic analysis and comparisons with the same specificity strains detected spo-radically between 2006 and 2010 revealed the constant circulation of these unusual human strains inBulgaria, although in low prevalence, and their increased potential for person-to-person spread.

� 2014 Elsevier B.V. All rights reserved.

1. Introduction

Rotaviruses (RVs) are the leading cause of acute viral gastroen-teritis in children under 5 years of age. The total number of RVdiarrhea cases has considerably declined since 2001, but RVs stillaccount for the death of 420–494 thousand children 65 yearsold, as half of the cases occur in 5 developing countries (Tateet al., 2012; Parashar et al., 2009). In 2006 two RV vaccines wereapproved, and since then they have been included in manynational immunization calendars around the globe. The implemen-tation of RV vaccination program in many countries has revealedthe dramatic decrease of the RV hospitalizations and economicbenefits connected with this (Rha et al., 2014; Linhares andJustino, 2014; Nakagomi et al., 2013a; De Oliveira et al., 2013;Raes et al., 2011). However, in countries without RV vaccine

included into the national immunization programs, RVs are stillresponsible for high morbidity and mortality (Yen et al., 2011;Mladenova et al., 2011).

RVs are classified as a genus of the family Reoviridae. The gen-ome of RVs, enclosed in a triple-layered protein capsid, consistsof 11 double-stranded RNA segments which code for a total of 12viral proteins, six structural (VP1–VP4, VP6, VP7) and six nonstruc-tural (NSP1–NSP6) (Estes and Kapikian, 2007). RVs are classifiedinto groups A to G based on epitopes of the VP6 intermediate cap-side protein, as RVs group A (RVA) have the major human and vet-erinary health impact. The two outer capsid proteins of RVA, VP7(glycosylated, or G-type) and VP4 (protease-sensitive, or P-type),elicit neutralizing antibody responses, and may have a role inhomotypic protection. For this reason G and P-types are the basisfor the traditional binary nomenclature of RVA (Hoshino et al.,1985). To date, at least 27 G and 37 P genotypes have been definedin human, mammals and birds according to molecular-geneticdiversity of the VP7 and VP4, respectively (Matthijnssens et al.,2011a). At least, 12 G types (G1–G6, G8–G12 and G20) and 15 Ptypes (P[1]–P[11], P[14], P[19], P[25], and P[28]) have beendetected in RVA-infected humans. Recently, a novel genotypingclassification scheme based on all 11 genome segments has been

Z. Mladenova et al. / Infection, Genetics and Evolution 29 (2015) 118–126 119

proposed providing valuable information on the RV genetic diver-sity and relationships among human and animal strains(Matthijnssens et al., 2008a). However, the traditional classifica-tion of RVAs into G/P genotype combination is still used in broadand long-term surveillance studies, as among human RVAs thesetwo are the proteins displaying the greatest level of diversity andalso the genes encoding these two proteins are the one involvedin the highest rates of reassortment events when compared tothe remaining nine segments (Patton, 2012).

Although many epidemiological studies have revealed that 5 G/Pcombinations, including G1, G3, G4 and G9 in combination withP[8], or G2P[4], are predominant worldwide, human RVAs withother genotypes (G5, G8, G10, G12, P[6], P[9]) have increasinglybeen detected in several countries, predominantly in Africa, Asiaand South America (Mijatovic-Rustempasic et al., 2014; Ahmedet al., 2013; Nakagomi et al., 2013b; da Silva et al., 2011; Wuet al., 2011; Esona et al., 2010; Iturriza-Gómara et al., 2004a). Fur-thermore, there is increasing recognition that many of these lesscommon human RV strains may originate through interspeciestransmission often accompanied by single or multiple reassortmentevents (Ianiro et al., 2013; Papp et al., 2013; Steyer et al., 2010,2013; Mladenova et al., 2012). These RVA stains, in particular thosethat through reassortment have acquired the ability for efficientperson-to-person transmission, have the potential to emerge asepidemic strains. In fact, this is believed to be the mechanism thatlead to the global emergence of G9P[8] strains in the mid-nineties(Page et al., 2010a; Bozdayi et al., 2008; Rahman et al., 2005;Steyer et al., 2005; Iturriza-Gómara et al., 2004b; Ramachandranet al., 2000), and to the widespread circulation of G8P[6] strainsin various African countries (Nakagomi et al., 2013a; Page et al.,2010b; Todd et al., 2010; Matthijnssens et al., 2006).

Rotavirus strain surveillance started in Bulgaria in 2005 withthe aim to study the natural diversity of RVA circulating in the pop-ulation prior to the introduction of rotavirus vaccines. In excess of2,000 RV strains were genotyped in Bulgaria between 2005 and2010 (Mladenova et al., 2010; Mladenova et al., 2012). In addition,1323 RV strains collected between June 2010 and May 2013 weregenotyped, and 17 strains were investigated further as they failedto be G genotyped or were found as having scarce circulation in thecountry. Here we describe the characterization of these strainsthough partial genome sequencing and phylogenetic analysiswhich revealed a potential for sporadic zoonotic transfer of RVAsin Bulgaria.

2. Materials and methods

2.1. Samples

Between June 2010 and May 2013, 4,431 diarrhoeal stool sam-ples from children and adults admitted to hospitals in 5 regionsin Bulgaria were tested for the presence of RV antigen using acommercial immunochromatography test (Rota-Strip, CorisBioConcept, Belgium) or enzyme-linked immunosorbent assay(RIDASCREEN Rotavirus, R-Biopharm, Germany) at Virology Labora-tory of Specialized Hospital for Active Treatment of Infectious andParasitic Diseases and at National Reference Laboratory of DiarrhealViruses, NCIPD, Sofia. None of the patients tested had received RVvaccine. A total of 1323 RV-positive samples were selected for Gand P genotyping on the basis of the age of the patient (childrenbetween 2 months and 8 years old), the location and month of iso-lation, and the availability of the samples. Genotyping was per-formed as previously described (www.eurorota.net/ last accessedApril, 2014).

Seventeen RV strains of interest were further characterized bypartial genome sequencing. In addition, three G3P[9], two

G3P[8], two G6P[9] and two G6P[8] RVA strains detected in Bul-garia between 2006 and 2010, which were characterized by partialgenome sequence analysis in previous investigations, were alsoused for analysis.

2.2. Nucleotide sequencing

Rotavirus VP7 and VP4 amplicons, obtained with the oligonu-cleotide primers VP7-F, VP7-R, VP4-F, and VP4-R (Iturriza-Gomara et al., 2011) were purified using commercial spin columnmethod (Qiagen GmbH, Hilden, Germany) and sequenced directlyusing the same primers and an ABI PRISM 3100 automated DNAsequencer (Applied Biosystems, Inc., Foster city, California, USA).

2.3. Sequence and phylogenetic analysis

The chromatograms were checked using BioEdit V.5.0.9.Comparisons between Bulgarian sequences and other RV VP7sand VP4s available in the GenBank database were made byBLAST server (www.ncbi.nlm.nih.gov/Blast). Multiple VP7 andVP4 sequence alignments and phylogenetic analysis were doneusing Clustal W in the BioEdit software and MEGA v5.10 analyticalpackage (Tamura et al., 2011), respectively. The phylogenetic treeswere constructed using the Neighbor-joining method by boot-strapping with 1000 replicates, and phylogenetic distances weremeasured by the Tajima–Nei model, implemented in the MEGAsoftware. Clusters were defined when bootstrap values were>85%. Nucleotide sequences of BG strains obtained in this studyhas been deposited in GenBank database under accession numbersKM590348–KM590397.

3. Results

A total of 2010 (45.4%) stool samples were positive in RV-screening test, of them 1323 strains were subjected to genotyping.Single G and P types were obtained for 1230 (93.0%) RV isolates,while 57 (4.3%) were mixed infections and 37 (2.7%) were partiallytyped (G or P untypeable). G1P[8], G2P[4], G4P[8] and G9P[8] rep-resented 96.0% (1181 isolates), 26.3% (n = 324), 31.4% (n = 386),23.5% (n = 289) and 14.8% (n = 182), respectively, of the total ofsamples fully genotyped. Seventeen isolates of interest were sub-jected to sequencing. In RT-PCR 6 strains left as G-untypeable,sequence analysis revealed they were with G6P[9] specificity.The G/P genotype combination of the other 11 were confirmed asG3P[9] (6 isolates) or G3P[8] (5 strains) (Table 1). In addition,Bulgarian strains with G6 (n = 4), G3 (n = 5), and P[9] (n = 5)characterized previously were used for the comparative analysis.

3.1. Analysis of VP7 of the G3 strains

A total of 11 G3 RV strains were sequenced. Nucleotidesequencing of the G3 strains revealed that 5 of them were in com-bination with typical human P genotype, P[8] (BG1299/2010,BG1344/2010, BG1548/2010, BG1596/2010, and BG347/2012),while the rest 6 strains were having P[9] genotype (BG1253/2010, BG1262/2010, BG1328/2010, BG1431/2010, BG1271/2012,and BG1511/2012) (Fig. 1). Five additional Bulgarian G3 strainswere included in phylogenetic analysis for comparison. The Bulgar-ian G3s segregated into two different lineages, all but two G3P[8]strains, BG1299/2010 and BG1344/2010, clustered within the typ-ical human G3 lineage, represented by the prototype strain RVA/Hu-tc/USA/P/1974/G3P[8]. The Bulgarian strains within this clustershowed homology of 99.5–99.8% at nt and 100% at the deducedamino acid (aa) levels, and were 97% and 98% similarity at thent and aa levels, respectively to the human prototype strain

Table 1List of the Bulgarian strains with G3, G6 and P[9] specificity used for the comparativeanalysis and construction of the phylogenetic trees.

Strain Year of detection Location Patient Genotype

Age (m) Sex G P

BG502* 2006 Sofia 48 M G6 P[9]BG885* 2008 Pernik 16 F G6 P[9]BG76* 2010 Sofia-region 10 F G6 P[8]BG105* 2010 Sofia-region 9 F G6 P[8]BG1506 2011 Plovdiv 11 F G6 P[9]BG1452 2012 Plovdiv 30 F G6 P[9]BG1458 2012 Sofia 46 F G6 P[9]BG1461 2012 Sofia 47 M G6 P[9]BG1465 2012 Sofia 27 M G6 P[9]BG291 2013 Sofia 23 F G6 P[9]BG1840* 2007 Sofia 24 M G3 P[9]BG932* 2008 Sofia 15 M G3 P[9]BG11* 2009 Sofia 23 M G3 P[9]BG175* 2010 Sofia-region 7 F G3 P[8]BG228* 2010 Plovdiv 36 F G3 P[8]BG1253 2010 Plovdiv 30 M G3 P[9]BG1262 2010 Plovdiv 3 F G3 P[9]BG1299 2010 Sofia 60 F G3 P[8]BG1328 2010 Sofia-region 24 M G3 P[9]BG1344 2010 Sofia 36 F G3 P[8]BG1431 2010 Pernik 12 M G3 P[9]BG1548 2011 Sofia 8 F G3 P[8]BG1596 2011 Sofia 12 M G3 P[8]BG347 2012 Sofia 42 F G3 P[8]BG1271 2012 Plovdiv 21 F G3 P[9]BG1511 2012 Plovdiv 30 M G3 P[9]

* Bulgarian strains characterized previously by sequencing.

120 Z. Mladenova et al. / Infection, Genetics and Evolution 29 (2015) 118–126

RVA/Hu-tc/USA/P/1974/G3P[8]. On the other hand, the BulgarianG3P[9] strains and two of G3P[8] strains (BG1299/2010 andBG1344/2010) clustered in a separate G3 lineage, alongside typicalfeline, canine and human-animal multiple reassortant strains withP[9] specificity. The Bulgarian strains exhibited 93–98% nt and95–98% aa homology with each other and were closely related tothe strain RVA/Fe-wt/ITA/BA222/2005/G3P[9], isolated from anadult cat in Italy in 2005, and were more distantly related to thehuman-feline reassortant RVA/Hu-tc/JPN/AU-1/1982/G3P[9]reported from Japan in 1982 (89–91%, and 92–95% homology atthe nt and aa levels, respectively).

The comparisons of aa residues in VP7 antigenic epitopesbetween G3 Bulgarian and other human and animal strains, exceptthe strain BA222, showed that they possess 2 conservative aa atpositions 142 (M) and 170 (I) (Suppl. 1). Bulgarian G3P[8]s had 2unique aa changes: aa108 (T ? I) and aa265 (S ? P). In addition,Bulgarian G3P[8]s and G3P[9]s differed in five aa in position 72,87, 122, 159 and 218. Additional aa substitution close to (aa78,T ? A) made Bulgarian G3P[9]s divergent from the rest of thehuman and animal G3s. Further, strains BG1328/G3P[9], BG1299/G3P[8] and BG1344/G3P[8], which clustered together with thehuman-feline reassortant PAH136/G3P[9] had unique aa substitu-tion in aa 87 (T ? A), yet possessed the same substitutions in aa116 (I ? V), 147 (T ? N), 179 (D ? N), 200 (T ? A), 213 (N ? S)and 223 (K ? R).

3.2. Analysis of VP7 of the G6 strains

The 6 new G6 sequences (from nt 88 to 760) were aligned with 4Bulgarian G6 strains (RVA/Hu-wt/BUL/BG502/2006/G6P[9], RVA/Hu-wt/BUL/BG885/2008/G6P[9], RVA/Hu-wt/BUL/BG76/2010/G6P[8],and RVA/Hu-wt/BUL/BG105/2010/G6P[8]) which had been previ-ously detected. Phylogenetic tree were constructed and the compar-ison of nt and deduced aa sequences of Bulgarian strains showed thatthey could be clustered into the two main G6 lineages PA151-like and

PA169-like (Fig. 2). Both groups Bulgarian G6 RVs were distantlyrelated and their homology ranged between 91.3% and 92.3%.

The 6 G6P[9] strains detected in the current study together withthe two Bulgarian G6P[8] isolated in 2010, exhibited high nt and aasimilarity (P99%) and were closely related to the human RV reas-sortants RVA/Hu-wt/BEL/B1711/2002/G6P[6], RVA/Hu-tc/USA/Se584/1998/G6P[9] and RVA/Hu-tc/ITA/PA151/1987/G6P[9],detected in Belgium, USA and Italy. Although the Bulgarian G6strains were in combination with P[9] or P[8] genotype, theyshowed the greatest similarity to the G6P[6] Belgium strain(96.9–97.4% nt and 98.4–98.8% aa identity). A total of 3 aa substi-tutions differed the Bulgarian strains in this cluster from theB1711 strain in positions 29 (I ? M), 145 (N ? D), and 225(A ? V) (Suppl. 2).

In contrast, the two of the Bulgarian G6P[9] strains, BG502/G6P[9] and BG885/G6P[9], clustered with the human PA169 strainassociate and closely related to other RVA G6P[14] strains. Thestrains BG502 and BG885 (99% nt and 100% aa homology to eachother), also shared high homology with the strain Hun5, detectedin Hungary in 1997 (96% nt and 100% aa) (Bányai et al., 2003),and differed by a single aa substitution in position 40 (A ? V).

3.3. Analysis of the VP4 (VP8⁄ fragment) of the P[9] strains

Partial VP4 sequences (from nt 195 to 582) of 17 Bulgarian RVstrains with P[9] specificity were analyzed. Twelve RV isolateswere sequenced in the present study, yet for 2 of the strains,BG1253/2010 and BG1262/2010, short nt sequences were obtainedand were excluded of the phylogenetic analysis. In addition, 5strains sequenced previously were also used for the homologousanalysis. The phylogenetic tree constructed showed that BulgarianP[9] strains, either in combination with G3 or G6 VP7 genes, clus-tered together with feline and feline-human reassortant RVs withP[9] specificity (Fig. 3). Moreover, the comparisons showed thatthe Bulgarian P[9] sequences were 92–100% and 94–100% similarto each other at nt and deduced aa level and were closely relatedto the VP4 sequences of the human reassortant isolate RVA/Hu-tc/JPN/AU-1/1982/G3P[9] (95–96% and 97–99%, respectively).Two conserved aa substitutions, in positions 129 (Q ? N) and250 (I ? M) differed the Bulgarian strains (except the isolateBG11/2009 and BG1452/2012, respectively) with the prototypestrain RVA/Hu-tc/JPN/AU-1/1982/G3P[9] (Suppl. 3). Also, 6 BGstrains (BG502/2006, BG1840/2007, BG885/2008, BG932/2008,BG11/2009, and BG1431/2010) had additional aa change in posi-tion 111 (S ? N).

3.4. Analysis of the VP4 (VP8⁄ fragment) of the P[8] strains

In addition, G3 and G6 RV strains associated with P[8] genotypewere detected, and partial VP4 sequences (from nt 190 to 760) wereanalyzed. Nucleotide sequence analysis of these strains showedthey belonged to two different P[8] lineages. BLAST analysis ofP[8] gene segments of the isolates with animal-like G3, RVA/Hu-wt/BUL/BG1299/2010/G3P[8] and RVA/Hu-wt/BG1344/2010/G3P[8], revealed 99% nt similarity to human RV strains with G1 orG4 VP7 genes detected in Cambodia, Thailand, Vietnam, Russiabetween 2005 and 2011 and 88.4% nt homology to the strainRVA/Hu-tc/USA/D/1974/G1P[8] which belong to lineage P[8]-1(www.rotac.regatools.be). The Bulgarian strains differed from theprototype RV strains in this cluster, RVA/Hu-wt/USA/WA/1974/G1P[8] and RVA/Hu-wt/USA/D/1974/G1P[8] in at least 17 aa (Suppl.4). Also, the comparison showed the Bulgarian G6P[8] strains, RVA/Hu-wt/BG76/2010/G6P[8] and RVA/Hu-wt/BG105/2010/G6P[8]and all G3P[8] strains studied clustered into lineage P[8]-2(Fig. 4). The Bulgarian G6P[8]s were closely related to isolatesdetected in Russia and Slovenia between 2005 and 2009 (100% nt

RVA/Hu-wt/BUL/BG11/2009/G3P9

RVA/Hu-wt/BUL/BG1431/2010/G3P9

RVA/Hu-wt/BUL/BG932/2008/G3P9

RVA/Hu-wt/BUL/BG1511/2012/G3P9

RVA/Hu-wt/BUL/BG1271/2012/G3P9

RVA/Hu-wt/BUL/BG1840/2007/G3P9

RVA/Hu-wt/BUL/BG1253/2010/G3P9

RVA/Hu-wt/BUL/BG1262/2010/G3P9

RVA/Cat-wt/ITA/BA222/2005/G3P9

RVA/Hu-wt/ITA/PAH136/1996/G3P9

RVA/Hu-wt/BUL/BG1328/2010/G3P9

RVA/Hu-wt/BUL/BG1299/2010/G3P8

RVA/Hu-wt/BUL/BG1344/2010/G3P8

RVA/Hu-wt/ITA/PAI58/1996/G3P9

RVA/Raccoon_dog/JPN/RAC-DG5/G3bP9

RVA/Cat/AUS/Cat2/1984/G3P9

RVA/Hu-wt/USA/KC814/1998/G3P9

human and animal G3s with P[9] genotype

RVA/Vaccine//USA/RotaTeq-WI78-8/1992/G3P5

RVA/Hu-tc/JPN/AU-1/1982/G3P9

RVA/Hu-tc/JPN/MO/1980/G3P8

RVA/Hu-tc/USA/P/1974/G3P8

RVA/Hu-tc/AUS/RV3/1977/G3P6

RVA/Hu/USA/VU08-09-20/2008/G3P8

RVA/Hu-wt/BUL/BG175/2010/G3P8

RVA/Hu-wt/BUL/BG1548/2011/G3P8

RVA/Hu-wt/BUL/BG1596/2011/G3P8

RVA/Hu-wt/BUL/BG228/2010/G3P8

RVA/Hu-wt/BUL/BG347/2012/G3P8

human G3s

RVA/Cow-lab/GBR/PP-1/1976/G3P7

RVA/Pig-tc/VEN/A131/1988/G3P7

RVA/Horse/IND/Erv105/G3P12

RVA/Si-tc/ZAF/SA11-N5/1958/G3P2

human-animal and animal G3s with P[3] or P[14] genotype

outgroup_G1/Wa

99

100

100

86

100

97

100

100 93

100

100

100

88

89

9788

94

84

0.05

Fig. 1. Phylogenetic tree based on nucleotide sequences of VP7 of G3 RV strains. The Bulgarian strains with P[9] specificity were marked with j and isolates with P[8] weremarked with N. The tree was constructed using Neighbour-joining method implemented in MEGA v5.10 software.

Z. Mladenova et al. / Infection, Genetics and Evolution 29 (2015) 118–126 121

and 99% aa similarity), while Bulgarian G3P[8]s were closely relatedto G3 RVs found in Asia, North and South America. The similarityamong the representatives of the P[8]-2 lineage, RVA/Hu-wt/COD/DRC88/2003/G8P[8], RVA/Hu-wt/BGD/Dhaka16/2003/G1P[8] andRVA/Hu-wt/BGD/Dhaka25/2002/G12P[8] and Bulgarian strains in

the same cluster varied between 96% and 100% at nt and aa level.In addition, the G6P[8] strains studied here had 4 unique aa substi-tutions in positions 85 (N ? T), 173 (I ? V), 191 (A ? T) and 194(N ? D) which differed them from the prototype strain RVA/Hu-wt/JPN/KU/1974/G1P[8] and the other isolates of the P[8]-1 cluster.

RVA/Hu-tc/ITA/PA169/1988/G6P14

RVA/Hu-wt/BEL/B10925/1997/G6P14

RVA/Hu-tc/AUS/MG6/1993/G6P14

RVA/Goat/SAF/Cap455/G6P14

RVA/Hu-wt/ITA/111-05-27/2005/G6P14

RVA/Hu-wt/HUN/BP1879/2003/G6P14

RVA/Hu-wt/HUN/Hun5/1997/G6P14

RVA/Hu-wt/BUL/BG502/2006/G6P9

RVA/Hu-wt/BUL/BG885/2008/G6P9

RVA/Goat-tc/BGD/GO34/1999/G6P1

RVA/Pig-wt/IND/HP113/1987/G6P13

RVA/Pig/ARG/P22/G6P1

RVA/Cow/ARG/B556/G6P5

RVA/Cow/ARG/B180/G6P5

RVA/Cow-tc/USA/NCDV/1967/G6P1

RVA/Cow-tc/USA/WC3/1981/G6P5

RVA/Vaccine/USA/RotaTeq-WI79-4/1992/G6P8

RVA/Cow/USA/C8336/G6P11

RVA/Cow/JPN/KN4/G6P11

RVA/Buffalo/ITA/10733/G6P3

RVA/Cow/ARG/B88/G6P11

RVA/Hu-tc/ITA/PA151/1987/G6P9

RVA/Hu-tc/USA/Se584/1998/G6P9

RVA/Hu-wt/BEL/B1711/2002/G6P6

RVA/Hu-wt/BUL/BG1465/2012/G6P9

RVA/Hu-wt/BUL/BG1452/2012/G6P9

RVA/Hu-wt/BUL/BG105/0110/G6P8

RVA/Hu-wt/BUL/BG1461/2012/G6P9

RVA/Hu-wt/BUL/BG1506/2011/G6P9

RVA/Hu-wt/BUL/BG76/0110/G6P8

RVA/Hu-wt/BUL/BG1458/2012/G6P9

RVA/Hu-wt/BUL/BG291/2013/G6P9

outgroup_Wa/G1

100

97

100

98

96

100

100

99

100

82

100

81

98

100

82

100

100

0.05

Fig. 2. Phylogenetic tree based on nucleotide sequences of VP7 of G6 RV strains. The Bulgarian strains with P[9] specificity were marked with j and isolates with P[8] weremarked with N. The tree was constructed using Neighbor-joining method implemented in MEGA v5.10 software.

122 Z. Mladenova et al. / Infection, Genetics and Evolution 29 (2015) 118–126

4. Discussion

Since the launch of Bulgarian RV surveillance program in 2005,more than 3000 RV strains have been genotyped. Among these,approximately 2.5% has possessed G and/or P types consideredunusual among human RVAs.

Of the five globally considered typical human RVA G types, G3strains have been rarely detected in Bulgaria in more than 8 yearsof continued surveillance. In the period since 2005, only 23 G3strains have been identified either in combination with P[8](n = 11), P[9] (n = 10), mixed P[8] + P[9] (n = 1), or P[3] (n = 1).The G3s had been detected every year since 2005 in the country,even though in low frequency which ranged from 0.37% in 2008to 1.9% in 2010, the highest rate registered.

The genotype G3 RVs have the broadest host range, includinghumans, domestic and farm animals, birds, and monkeys (Estesand Kapikian, 2007). In humans, the majority of G3 strains are asso-ciated with P[8] VP4s and have a Wa-like genomic constellation

(Matthijnssens et al., 2008b), yet recent studies increasinglyreported G3P[9] strains with AU-1 genomic constellation, typicalfor feline RVs (Iizuka et al., 2011; Hwang et al., 2011; Khamrinet al., 2007; Nakagomi and Nakagomi, 1991). Recently, sequenceinvestigations of all or part of the genes of selected G3 strains hasprovided further evidence of interspecies transmission and reas-sortment among strains of human, feline, canine and bovine RVs(Wang et al., 2013; Matthijnssens and Van Ranst, 2012; Grantet al., 2011; Martella et al., 2011; Matthijnssens et al., 2011b; DeGrazia et al., 2010).

In our study, G3P[8] strains clustered into 2 distinct lineages,one typically associated with human rotavirus strains, and pre-sumably in association with a Wa-genomic backbone, and the sec-ond typically associated with animal RVs or RVs isolated fromhuman cases but which are closely related to animal strains andthought to derive through zoonotic transfer with or without reas-sortment. Furthermore, the G3P[9] strains found in Bulgaria alsoclustered in this latter linage, and the VP7 genes were closely

RVA/Hu-wt/BUL/BG1271/2012/G3P[9]

RVA/Hu-wt/BUL/BG1511/2012/G3P[9]

RVA/Hu-wt/BUL/BG1458/2012/G6P[9]

RVA/Hu-wt/BUL/BG291/2013/G6P[9]

RVA/Hu-wt/BUL/BG1506/2011/G6P[9]

RVA/Hu-wt/BUL/BG1461/2012/G6P[9]

RVA/Hu-wt/BUL/BG1328/2010/G3P[9]

RVA/Hu-wt/BUL/BG1452/2012/G6P[9]

RVA/Cat-wt/ITA/BA222/2005/G3P[9]

RVA/Hu-wt/BUL/BG932/2008/G3P[9]

RVA/Hu-wt/BUL/BG885/2008/G6P[9]

RVA/Hu-wt/BUL/BG1840/2007/G3P[9]

RVA/Hu-wt/BUL/BG502/2006/G6P[9]

RVA/Hu-wt/BUL/BG11/2009/G3P[9]

RVA/Hu-wt/BUL/BG1431/2010/G3P[9]

RVA/Hu-wt/BUL/BG1465/2012/G6P[9]

RVA/Cat-tc/AUS/Cat2/1984/G3P[9]

RVA/Hu-wt/ITA/PAI58/1996/G3P[9]

RVA/Hu-wt/ITA/PAH136/1996/G3P[9]

RVA/Hu-wt/CHN/L621/2006/G3P[9]

RVA/Hu-wt/THA/CMH120/2004/G3P[9]

RVA/Hu-tc/JPN/AU-1/1982/G3P3[9]

RVA/Hu-tc/USA/CC425/1997/G3P[9]

RVA/Hu-tc/USA/Se584/1998/G6P[9]

RVA/Hu-wt/PRY/Py1135ASR07/2007/G12P[9]

RVA/Hu/JPN/K12/1999/G12P[9]

RVA/Hu-wt/KOR/Kor588/2002/G12P[9]

RVA/Hu-tc/ARG/ARG720A/1999/G12P[9]

RVA/Hu-tc/JPN/CP727/G12P[9]

RVA/Hu-tc/THA/T152/1998/G12P[9]

outgroup P[14]/strain R-2

100

97

100

100

81

90

8395

96

82

88

0.05

Fig. 3. Phylogenetic tree based on partial nucleotide sequences of VP4 of P[9] RV strains. The Bulgarian strains with G3P[9] specificity were marked with j and strains withG6P[9] genotypes were marked with �. The tree was constructed using Neighbor-joining method implemented in MEGA v5.10 software.

Z. Mladenova et al. / Infection, Genetics and Evolution 29 (2015) 118–126 123

related to the feline strain BA222 and other BA222-like strainsfound in humans in different countries. This is the first report ofhuman G3P[8] RV strains with a VP7 more closely related to VP7genes derived from feline strains, and it is likely that it is a resultof reassortment with human strains containing the typicallyhuman RV P[8] VP4 gene. Further analysis of these stains thoughcomplete genome characterization should reveal the likely originof the genetic backbone, and the weather the association with aWa-like gene constellation could potentially facilitate person-to-person transmission of this G3 feline-human reassortant strain.

The comparison of aa sequences of Bulgarian G3P[9]s to typicalfeline, canine or human-animal reassortant strains showed severalaa substitutions, including in positions 78 (A), 87 (S), 213 (N), and218 (I). In previous studies focused on the VP7 aa sequence, ninevariable regions, VR1–VR9, have been defined, of them VP5 (anti-genic region A, aa 87–101), VP7 (antigenic region B, aa 143–151),VP8 (antigenic region C, aa 208–221), and VP9 (antigenic regionF, aa 235–242) have been confirmed as the major RV neutralizationsites (Dyall-Smith et al., 1986; Taniguchi et al., 1988; Coulson andKirkwood, 1991; Hoshino et al., 1994). Some particular aa in the

antigenic regions, especially positions 94, 96, 147, 148, 190, 208,211, 213, 238, and 291, have been found to be immunodominantepitopes, and substitutions in these positions is believed to changethe antigenicity of the RV strain and to help them to escape hostimmunity (Mackow et al., 1988; Taniguchi et al., 1988; Coulsonand Kirkwood, 1991; Hoshino et al., 1994). Amino acid substitu-tions, in particular these observed in antigenic regions A and C,detected in Bulgarian G3P[9] strains might have changed the strainantigenicity and increased their potential for person-to-personspread due to the poorer homotypic immunity induced.

On the contrary, RVs with G6 VP7 genes are not typically asso-ciated with human infections. Although they have been found spo-radically in human diarrhea cases in several countries, unlike theG3 in which the VP7s can be classified as human or animal-derived, G6 sequences are thought to derive from animal RVs. AG6P[9] strain was first isolated from an Italian child with diarrhea(Gerna et al., 1992), and related strains in association with P[6],P[9] or P[14] have subsequently been reported in the USA, BurkinaFaso, Mali, Hungary, Japan, Italy, Australia and Tunisia (Ianiro et al.,2013; Nordgren et al., 2012; De Grazia et al., 2011; Yamamoto

RVA/Hu-wt/BUL/BG1548/2011/G3P8

RVA/Hu-wt/BUL/BG228/2010/G3P8

RVA/Hu-wt/BUL/BG175/2010/G3P8

RVA/Hu-wt/BUL/BG1596/2011/G3P8

RVA/Hu-wt/BUL/BG347/2012/G3P8

RVA/Hu-wt/CHN/DD8/Liaoning/2007/G3P8

RVA/Hu-wt/PRY/1747SR/2009/G3G1P8

RVA/Hu-wt/CAN/RT025-07/2007/G3P8

RVA/Hu-wt/BEL/B4633/2003/G12P8

RVA/Hu-wt/BGD/Dhaka25/2002/G12P8

RVA/Hu-wt/BGD/Dhaka16/2003/G1P8

RVA/Hu-wt/COD/DRC88/2003/G8P8

RVA/Hu-wt/SVN/SI-R13/2008/G4P8

RVA/Hu-wt/SVN/SI-MB44/2005/G9P8

RVA/Hu-wt/BUL/BG76/2010/G6P8

RVA/Hu-wt/BUL/BG105/2010/G6P8

RVA/Hu-wt/RUS/Nov09-D91/2009/G4P8

RVA/Hu-wt/RUS/Nov09-D221/2009/G4P8

RVA/Hu-wt/JPN/KU/1974/G1P8

RVA/Hu-wt/BRA/IAL28/1992/G5P8

RVA/Hu-wt/AUS/F45//G9P8

RVA/Vaccine/USA/Rotateq-WI79-4/1992/G6P8

RVA/Hu-wt/USA/P/1974/G3P8

RVA/Hu-wt/JPN/ITO/1981/G3P8

RVA/Hu-tc/Vaccine/Rotarix-A41CB052A/1988/G1P8

RVA/Hu-wt/USA/WA/1974/G1P8

RVA/Hu-wt/USA/D/1974/G1P8

RVA/Hu-wt/BUL/BG1299/2010/G3P8

RVA/Hu-wt/BUL/BG1344/2010/G3P8

RVA/Hu-wt/CAM/NP04-142/2005/G?P8

RVA/Hu-wt/VNM/NhaTrang_V20/2006/G1P8

RVA/Hu-wt/THA/CU20/2004-06/G?P8

RVA/Hu-wt/RUS/Nov09-D297/2009/G1P8

outgroup_P14

94

83

100

98

81

89

89

99

96

100

99

0.05

Fig. 4. Phylogenetic tree based on nucleotide sequences of the P8⁄ fragment of VP4 of P[8] RV strains. The Bulgarian strains with G6 specificity were marked with j andisolates with G3 were marked with N. The tree was constructed using Neighbor-joining method implemented in MEGA v5.10 software.

124 Z. Mladenova et al. / Infection, Genetics and Evolution 29 (2015) 118–126

et al., 2011; Bányai et al., 2009; Rahman et al., 2003; Diwakarlaet al., 2002; Cooney et al., 2001). In contrast to G3 RVs, G6 RVsare not associated with a characteristically human cluster in asso-ciation with Wa or DS-1 like genetic backbone or even typicallyhuman RV-derived VP4 genes. Previous reports have describedthe likely origin of such strains in humans through interspeciestransmission having undergone reassortment events betweenbovine, human and/or feline RVs, with some of the G6 strainsfound in humans, possessing a genetic backbone suggestive of afeline origin in all but the gene encoding the VP7, for which previ-ously a potential bovine origin had been speculated (Bányai et al.,2003).

In fact, the P[9] genotype found in association with G3, G6, andG12 has been isolated only from humans, cats and dogs and hence,has been named the feline/canine-like (Wang et al., 2013). Therecent analysis of full genome sequence data of some of P[9]s sug-gested multiple transmissions of genes from animal to human RV

strains, however, the lack of systematic surveillance in animalsand in particularly among the animal species that live in close con-tact with children, such as pets, the direction of interspecies trans-mission and the true origin of unusual strains found in humandisease is difficult to ascertain with sufficient degree of certainty.

From the surveillance in Bulgaria, despite the overall prevalenceof G6 RVs being relative low, an increase in the detection rate ofG6P[9] in 20,012/13 (6 out of the 8 detected since 2005) was seen.This increase was observed particularly in Sofia and Plovdiv, thetwo largest urban populations sampled. Also, G6P[8] reassortantstrains were identified, and whole genome characterization ofthese strains would allow to assess the strain genetic backbonecomposition and origin as well as the potential adaptation of suchreassortant strains for human-to-human transmission and furtherspread.

The data so far strongly suggests the possibility that some of theG3 and all of the G6 strain identified in Bulgaria have a zoonotic

Z. Mladenova et al. / Infection, Genetics and Evolution 29 (2015) 118–126 125

VP7 component, but also importantly there is evidence though theassociation of some of them with P[8] VP4 they have undergonereassortment with human strains. A feline or canine origin for atleast some of the genes is highly plausible, both from the sequencecharacteristics of the strains and the existing literature, and also,due to the fact that in Bulgaria, as in many other countries in urbanpopulation companion pets such cats and dogs, are likely to theanimals in closest and more frequent contact with children. How-ever, in order to fully understand how these strains emerged inBulgaria and the direction of transmission both animal surveillanceand detail epidemiological contact tracing would be required. Fur-thermore, in order to better assess the potential of these reassor-tant strains to spread further and become epidemic or evenendemic in the humans in the future full genome characterizationwould also be useful.

Worldwide, two RV vaccines are available for the prevention ofRVA infection. The RV5 (RotaTeq�, Merck&Sanofi Pasteur),approved in 2006, is a live, pentavalent human-bovine reassortantvaccine, orally administered in 3-doses, while RV1 (Rotarix�,GlaxoSmithKline) was approved in 2008, and it is a live, attenuatedhuman monovalent vaccine, administered in a 2-dose schedule.Only RV1 is available in Bulgaria on the private market, and evenit is recommended for use by the Bulgarian health authorities,the vaccine coverage has still been relatively low (<5%).

Even though the five common RV G/P genotype combinationsrepresent >90% of circulated strains in Bulgaria, the frequency ofdetection of unusual human strains or strains with animal-likecharacteristics has been increasing since the launch of BulgarianRV Surveillance program in 2005.

The mass use of RV vaccines in many countries in Europe hasrevealed the vaccines are highly effective in decreasing the numberof severe and mild RVGE cases (Vesikari et al., 2007a,b), but as theyconfer under 100% protection toward RV strains not included invaccine composition (Leshem et al., 2014), they might act as aselective tool along with immunity pressure itself and to ‘‘boost’’the fitness of some unusual for the human population RV strains.Thus, the increased detection of uncommon strains in humanswith altered characteristics or which are linked with animal RVshas raised concerns for universal RV immunization programs,and still on-going RV surveillance activities in countries withimplemented mass RV vaccination or without, as Bulgaria, willhelp to trace and understand the trends and fluctuations in RV epi-demiology and evolution in both, wild RV strain population andvaccine escape mutants.

To date, the correlates of protection from RVA infection and dis-ease are largely unknown, and the role of homotypic immuneresponses against heterotypic protection remains to be fullyunderstood. Current evidence clearly suggests that both, protectionconferred by natural infection and by vaccination, are to a greatextent heterotypic. However, although the reasons for the strainfluctuations observed between countries regions and from oneyear to another are not fully understood, it is possible that hetero-geneity of immunity and partial protection at the level of the pop-ulation may provide at least in partial explanation. Rotavirusstrains have emerged and spread in recent years in the absenceof vaccination programs, and host adaptation and partial hetero-typic protections may again explain such emergences. Therefore,despite the proven efficacy of rotavirus vaccines, strain monitoringand a better understanding of the drivers for the emergence andsuccessful persistence of RVs in the human and other animal pop-ulation are still needed.

Acknowledgements

The study was funded by European Rotavirus Surveillance Net-work – EuroRotaNet project and partially by NCIPD (for RV screen-

ing of the stool samples). We thank Maria Georgieva (NRLDV,NCIPD), all staff of Specialized Hospital for Active Treatment ofInfectious and Parasitic Diseases (Sofia) and University Hospital‘‘St. George’’ (Plovdiv), and all the colleagues who contributed tothe Bulgarian Rotavirus Surveillance Project.

Appendix A. Supplementary data

Supplementary data associated with this article can be found, inthe online version, at http://dx.doi.org/10.1016/j.meegid.2014.11.011.

References

Ahmed, S., Klena, J., Albana, A., Alhamdani, F., Oskoff, J., Soliman, M., et al., 2013.Characterization of human rotaviruses circulating in Iraq in 2008: atypical G8and high prevalence of P[6] strains. Infect. Genet. Evol. 16, 212–217.

Bányai, K., Gentsch, J.R., Griffin, D.D., Holmes, J.L., Glass, R.I., Szücs, G., 2003. Geneticvariability among serotype G6 human rotaviruses: identification of a novellineage isolated in Hungary. J. Med. Virol. 71 (1), 124–134.

Bányai, K., Bogdán, A., Domonkos, G., Kisfali, P., Molnár, P., Tóth, A., et al., 2009.Genetic diversity and zoonotic potential of human rotavirus strains, 2003–2006,Hungary. J. Med. Virol. 81 (2), 362–370.

Bozdayi, G., Dogan, B., Dalgic, B., Bostanci, I., Sari, S., Battaloglu, N.O., et al., 2008.Diversity of human rotavirus G9 among children in Turkey. J. Med. Virol. 80 (4),733–740.

Cooney, M.A., Gorrellm, R.J., Palombo, E.A., 2001. Characterisation and phylogeneticanalysis of the VP7 proteins of serotype G6 and G8 human rotaviruses. J. Med.Microbiol. 50 (5), 462–467.

Coulson, B.S., Kirkwood, C., 1991. Relation of VP7 amino acid sequence tomonoclonal antibody neutralization of rotavirus and rotavirus monotype. J.Virol. 65 (11), 5968–5974.

da Silva, M.F., Tort, L.F., Gómez, M.M., Assis, R.M., de Mendonça, M.C., Volotão, E.M.,et al., 2011. Phylogenetic analysis of VP1, VP2, and VP3 gene segments ofgenotype G5 group A rotavirus strains circulating in Brazil between 1986 and2005. Virus Res. 160 (1–2), 381–388.

De Grazia, S., Giammanco, G.M., Potgieter, C.A., Matthijnssens, J., Banyai, K., Platia,M.A., et al., 2010. Unusual assortment of segments in 2 rare human rotavirusgenomes. Emerg. Infect. Dis. 16 (5), 859–862.

De Grazia, S., Martella, V., Rotolo, V., Bonura, F., Matthijnssens, J., Bányai, K., et al.,2011. Molecular characterization of genotype G6 human rotavirus strainsdetected in Italy from 1986 to 2009. Infect. Genet. Evol. 11 (6), 1449–1455.

De Oliveira, L.H., Giglio, N., Ciapponi, A., García Martí, S., Kuperman, M., Sanwogou,N.J., et al., 2013. Temporal trends in diarrhea-related hospitalizations anddeaths in children under age 5 before and after the introduction of the rotavirusvaccine in four Latin American countries. Vaccine 31 (Suppl. 3), C99–C108.

Diwakarla, S., Clark, R., Palombo, E.A., 2002. Expanding distribution of humanserotype G6 rotaviruses in Australia. Microbiol. Immunol. 46 (7), 499–502.

Dyall-Smith, M.L., Lazdins, I., Tregear, G.W., Holmes, I.H., 1986. Location of themajor antigenic sites involved in rotavirus serotype-specific neutralization.Proc. Natl. Acad. Sci. 83 (10), 3465–3468.

Esona, M.D., Steele, D., Kerin, T., Armah, G., Peenze, I., Geyer, A., et al., 2010.Determination of the G and P types of previously nontypeable rotavirus strainsfrom the African Rotavirus Network, 1996–2004: identification of unusual Gtypes. J. Infect. Dis. 202 (Suppl.), S49–S54.

Estes, M., Kapikian, A., 2007. Rotaviruses. In: Knipe, D.M., Griffin, D.E. (Eds.), Fields’Virology, 5th ed. Lippincott Williams & Wilkins, Philadelphia, chapter 53, pp.1917–1974.

Gerna, G., Sarasini, A., Parea, M., Arista, S., Miranda, P., Brüssow, H., et al., 1992.Isolation and characterization of two distinct human rotavirus strains with G6specificity. J. Clin. Microbiol. 30 (1), 9–16.

Grant, L., Esona, M., Gentsch, J., Watt, J., Reid, R., Weatherholtz, R., et al., 2011.Detection of G3P[3] and G3P[9] rotavirus strains in American Indian childrenwith evidence of gene reassortment between human and animal rotaviruses. J.Med. Virol. 83 (7), 1288–1299.

Hoshino, Y., Wyatt, R.G., Flores, J., Midthun, K., Kapikian, A.Z., 1985. Serotypiccharacterization of rotaviruses derived from asymptomatic human neonatalinfections. J. Clin. Microbiol. 21 (3), 425–430.

Hoshino, Y., Nishikawa, K., Benfield, D.A., Gorziglia, M., 1994. Mapping of antigenicsites involved in serotype-cross-reactive neutralization on group A rotavirusouter capsid glycoprotein VP7. Virology 199 (1), 233–237.

Hwang, K.P., Huang, Y.C., Bányai, K., Wu, H.S., Chang, F.Y., Yang, D.C., et al., 2011.Severe gastroenteritis associated with G3P[9] rotavirus in Taiwan. Infection 39(3), 271–275.

Ianiro, G., Delogu, R., Camilloni, B., Lorini, C., Ruggeri, F.M., Fiore, L., 2013. Detectionof unusual G6 rotavirus strains in Italian children with diarrhoea during the2011 surveillance season. J. Med. Virol. 85 (10), 1860–1869.

Iizuka, M., Chiba, M., Masamune, O., Kaga, E., Nakagomi, T., Nakagomi, O., 2011. Ahighly conserved genomic RNA constellation of Japanese isolates of humanrotaviruses carrying G serotype 3 and P serotype 9. Infect. Genet. Evol. 11 (6),1449–1455.

126 Z. Mladenova et al. / Infection, Genetics and Evolution 29 (2015) 118–126

Iturriza-Gómara, M., Kang, G., Mammen, A., Jana, A.K., Abraham, M., Desselberger,U., et al., 2004a. Characterization of G10P[11] rotaviruses causing acutegastroenteritis in neonates and infants in Vellore, India. J. Clin. Microbiol. 42(6), 2541–2547.

Iturriza-Gómara, M., Kang, G., Gray, J., 2004b. Rotavirus genotyping: keeping upwith an evolving population of human rotaviruses. J. Clin. Virol. 31 (4), 259–265.

Iturriza-Gomara, M., Dallman, T., Banyai, K., Bottier, B., Buesa, J., Diedrich, S., et al.,2011. Rotavirus genotypes co-circulating in Europe between 2006 and 2009 asdetermined by EuroRotaNet, a pan-European collaborative strain surveillancenetwork. Epidemiol. Infect. 139 (6), 895–909.

Khamrin, P., Maneekarn, N., Peerakome, S., Tonusin, S., Phan, T.G., Okitsu, S., et al.,2007. Molecular characterization of rare G3P[9] rotavirus strains isolated fromchildren hospitalized with acute gastroenteritis. J. Med. Virol. 79 (6), 843–851.

Leshem, E., Lopman, B., Glass, R., Gentsch, J., Bányai, K., Parashar, U., et al., 2014.Distribution of rotavirus strains and strain-specific effectiveness of therotavirus vaccine after its introduction: a systematic review and meta-analysis. Lancet Infect. Dis. 14 (9), 847–856.

Linhares, A.C., Justino, M.C., 2014. Rotavirus vaccination in Brazil: effectiveness andhealth impact seven years post-introduction. Expert Rev. Vacc. 13 (1), 43–57.

Mackow, E.R., Shaw, R.D., Matsui, S.M., Vo, P.T., Benfield, D.A., Greenberg, H.B., 1988.Characterization of homotypic and heterotypic VP7 neutralization sites ofrhesus rotavirus. Virology 165 (2), 511–517.

Martella, V., Potgieter, A.C., Lorusso, E., De Grazia, S., Giammanco, G.M.,Matthijnssens, J., et al., 2011. A feline rotavirus G3P[9] carries traces ofmultiple reassortment events and resembles rare human G3P[9] rotaviruses. J.Gen. Virol. 92 (Pt. 5), 1214–1221.

Matthijnssens, J., Rahman, M., Yang, X., Delbeke, T., Arijs, I., Kabue, J.P., et al., 2006.G8 rotavirus strains isolated in the Democratic Republic of Congo belong to theDS-1-like genogroup. J. Clin. Microbiol. 44 (5), 1801–1809.

Matthijnssens, J., Ciarlet, M., Rahman, M., Attoui, H., Bányai, K., Estes, M.K., et al.,2008a. Recommendations for the classification of group A rotaviruses using all11 genomic RNA segments. Arch. Virol. 153 (8), 1621–1629.

Matthijnssens, J., Ciarlet, M., Heiman, E., Arijs, I., Delbeke, T., McDonald, S.M., et al.,2008b. Full genome-based classification of rotaviruses reveals a common originbetween human Wa-like and porcine rotavirus strains and human DS-1-likeand bovine rotavirus strains. J. Virol. 82 (7), 3204–3219.

Matthijnssens, J., Ciarlet, M., McDonald, S.M., Attoui, H., Bányai, K., Brister, J.R., et al.,2011a. Uniformity of rotavirus strain nomenclature proposed by the RotavirusClassification Working Group (RCWG). Arch. Virol. 156 (8), 1397–1413.

Matthijnssens, J., De Grazia, S., Piessens, J., Heylen, E., Zeller, M., Giammanco, G.M.,et al., 2011b. Multiple reassortment and interspecies transmission eventscontribute to the diversity of feline, canine and feline/canine-like human groupA rotavirus strains. Infect. Genet. Evol. 11 (6), 1396–1406.

Matthijnssens, J., Van Ranst, M., 2012. Genotype constellation and evolution ofgroup A rotaviruses infecting humans. Curr. Opin. Virol. 2 (4), 426–433.

Mijatovic-Rustempasic, S., Teel, E.N., Kerin, T.K., Hull, J.J., Roy, S., Weinberg, G.A.,et al., 2014. Genetic analysis of G12P[8] rotaviruses detected in the largest U.S.G12 genotype outbreak on record. Infect. Genet. Evol. 21, 214–219.

Mladenova, Z., Korsan, N., Geonova, T., Iturriza-Gómara, M., Rotavirus Study Group,2010. Molecular epidemiology of rotaviruses in Bulgaria: annual shift of thepredominant genotype. Eur. J. Clin. Microbiol. Infect. Dis. 29 (5), 555–562.

Mladenova, Z., Iturriza-Gomara, M., Esona, M.D., Gray, J., Korsun, N., 2011. Geneticcharacterization of Bulgarian rotavirus isolates and detection of rotavirusvariants: challenges for the rotavirus vaccine program? J. Med. Virol. 83 (2),348–356..

Mladenova, Z., Papp, H., Lengyel, G., Kisfali, P., Steyer, A., Steyer, A.F., et al., 2012.Detection of rare reassortant G5P[6] rotavirus, Bulgaria. Infect. Genet. Evol. 12(8), 1676–1684.

Nakagomi, O., Iturriza-Gomara, M., Nakagomi, T., Cunliffe, N.A., 2013a.Incorporation of a rotavirus vaccine into the national immunisation schedulein the United Kingdom: a review. Expert Opin. Biol. Ther. 13 (11), 1613–1621.

Nakagomi, O., Nakagomi, T., 1991. Molecular evidence for naturally occurring singleVP7 gene substitution reassortant between human rotaviruses belonging to twodifferent genogroups. Arch. Virol. 119 (1–2), 67–81.

Nakagomi, T., Doan, Y.H., Dove, W., Ngwira, B., Iturriza-Gómara, M., Nakagomi, O.,et al., 2013b. G8 rotaviruses with conserved genotype constellations detected inMalawi over 10 years (1997–2007) display frequent gene reassortment amongstrains co-circulating in humans. J. Gen. Virol. 94 (Pt 6), 1273–1295.

Nordgren, J., Nitiema, L.W., Sharma, S., Ouermi, D., Traore, A.S., Simpore, J., et al.,2012. Emergence of unusual G6P[6] rotaviruses in children, Burkina Faso, 2009–2010. Emerg. Infect. Dis. 18 (4), 589–597.

Page, N., Esona, M., Armah, G., Nyangao, J., Mwenda, J., Sebunya, T., et al., 2010a.Emergence and characterization of serotype G9 rotavirus strains from Africa. J.Infect. Dis. 202 (Suppl.), S55–S63.

Page, N., Esona, M., Seheri, M., Nyangao, J., Bos, P., Mwenda, J., et al., 2010b.Characterization of genotype G8 strains from Malawi, Kenya, and South Africa. J.Med. Virol. 82 (12), 2073–2081.

Papp, H., Borzák, R., Farkas, S., Kisfali, P., Lengyel, G., Molnár, P., et al., 2013. Zoonotictransmission of reassortant porcine G4P[6] rotaviruses in Hungarian pediatricpatients identified sporadically over a 15 years period. Infect. Genet. Evol. 19,71–80.

Parashar, U.D., Burton, A., Lanata, C., Boschi-Pinto, C., Shibuya, K., Steele, D., et al.,2009. Global mortality associated with rotavirus disease among children in2004. J. Infect. Dis. 200 (Suppl. 1), S9–S15.

Patton, J., 2012. Rotavirus diversity and evolution in pros-vaccine world. Discov.Med. 13 (68), 85–97.

Raes, M., Strens, D., Vergison, A., Verghote, M., Standaert, B., 2011. Reduction inpediatric rotavirus-related hospitalizations after universal rotavirus vaccinationin Belgium. Pediatr. Infect. Dis. J. 30, 120–125.

Rahman, M., De Leener, K., Goegebuer, T., Wollants, E., Van der Donck, I., VanHoovels, L., et al., 2003. Genetic characterization of a novel, naturally occurringrecombinant human G6P[6] rotavirus. J. Clin. Microbiol. 41 (5), 2088–2095.

Rahman, M., Matthijnssens, J., Goegebuer, T., De Leener, K., Vanderwegen, L., van derDonck, I., et al., 2005. Predominance of rotavirus G9 genotype in childrenhospitalized for rotavirus gastroenteritis in Belgium during 1999–2003. J. Clin.Virol. 33 (1), 1–6.

Ramachandran, M., Kirkwood, C.D., Unicomb, L., Cunliffe, N.A., Ward, R.L., Bhan,M.K., et al., 2000. Molecular characterization of serotype G9 rotavirus strainsfrom a global collection. Virology 278 (2), 436–444.

Rha, B., Tate, J.E., Payne, D.C., Cortese, M.M., Lopman, B.A., Curns, A.T., et al., 2014.Effectiveness and impact of rotavirus vaccines in the United States – 2006–2012. Expert Rev. Vacc. 13 (3), 365–376.

Steyer, A., Bajzelj, M., Iturriza-Gómara, M., Mladenova, Z., Korsun, N., Poljsak-Prijatelj, M., 2010. Molecular analysis of human group A rotavirus G10P[14]genotype in Slovenia. J. Clin. Virol. 49 (2), 121–125.

Steyer, A., Poljsak-Prijatelj, M., Barlic-Maganja, D., Bufon, T., Marin, J., 2005. Theemergence of rotavirus genotype G9 in hospitalised children in Slovenia. J. Clin.Virol. 33 (1), 7–11.

Steyer, A., Sagadin, M., Kolenc, M., Poljšak-Prijatelj, M., 2013. Whole genomesequence analysis of bovine G6P[11] rotavirus strain found in a child withgastroenteritis. Infect. Genet. Evol. 13, 89–95.

Tamura, K., Peterson, D., Peterson, N., Stecher, G., Nei, M., Kumar, S., 2011. MEGA5:molecular evolutionary genetics analysis using maximum likelihood,evolutionary distance, and maximum parsimony methods. Mol. Biol. Evol. 28(2011), 2731–2739.

Taniguchi, K., Hoshino, Y., Nishikawa, K., Green, K.Y., Maloy, W.L., Morita, Y., et al.,1988. Cross-reactive and serotype-specific neutralization epitopes on VP7 ofhuman rotavirus: nucleotide sequence analysis of antigenic mutants selectedwith monoclonal antibodies. J. Virol. 62 (6), 1870–1874.

Tate, J.E., Burton, A.H., Boschi-Pinto, C., Steele, A.D., Duque, J., Parashar, U.D., 2012.2008 estimate of worldwide rotavirus- associated mortality in children youngerthan 5 years before the introduction of universal rotavirus vaccinationprogrammes: a systematic review and meta-analysis. Lancet Infect. Dis. 12(2), 136–141.

Todd, S., Page, N.A., Duncan Steele, A., Peenze, I., Cunliffe, N.A., 2010. Rotavirusstrain types circulating in Africa: review of studies published during 1997–2006. J. Infect. Dis. 202 (Suppl.), S34–S42.

Vesikari, T., Karvonen, A., Prymula, R., Schuster, V., Tejedor, J.C., Cohen, R., et al.,2007a. Efficacy of human rotavirus vaccine against rotavirus gastroenteritisduring the first 2 years of life in European infants: randomised, double-blindcontrolled study. Lancet 370 (9601), 1757–1763.

Vesikari, T., Itzler, R., Matson, D.O., Santosham, M., Christie, C.D., Coia, M., et al.,2007b. Efficacy of a pentavalent rotavirus vaccine in reducing rotavirus-associated health care utilization across three regions (11 countries). Int. J.Infect. Dis. 11 (Suppl. 2), S29–S35.

Wang, Y.H., Pang, B.B., Zhou, X., Ghosh, S., Tang, W.F., Peng, J.S., et al., 2013. Complexevolutionary patterns of two rare human G3P[9] rotavirus strains possessing afeline/canine-like H6 genotype on an AU-1-like genotype constellation. Infect.Genet. Evol. 16, 103–112.

Wu, F.T., Bányai, K., Huang, J.C., Wu, H.S., Chang, F.Y., Yang, J.Y., et al., 2011. Diverseorigin of P[19] rotaviruses in children with acute diarrhea in Taiwan: detectionof novel lineages of the G3, G5, and G9 VP7 genes. J. Med. Virol. 83 (7), 1279–1287.

Yamamoto, D., Kawaguchiya, M., Ghosh, S., Ichikawa, M., Numazaki, K., Kobayashi,N., 2011. Detection and full genomic analysis of G6P[9] human rotavirus inJapan. Virus Genes 43 (2), 215–223.

Yen, C., Tate, J.E., Patel, M.M., Cortese, M.M., Lopman, B., Fleming, J., et al., 2011.Rotavirus vaccines: update on global impact and future priorities. Hum. Vacc. 7(12), 1282–1290.