Molecular epidemiology and genetic ...

Journal of General Virology (2010), 91, 2286–2301

DOI 10.1099/vir.0.019737-0

Molecular epidemiology and genetic characterization of equine arteritis virus isolates associated with the 2006–2007 multi-state disease occurrence in the USA Jianqiang Zhang, Peter J. Timoney, Kathleen M. Shuck, Gong Seoul, Yun Young Go, Zhengchun Lu, David G. Powell, Barry J. Meade and Udeni B. R. Balasuriya Correspondence Udeni B. R. Balasuriya

Department of Veterinary Science, Maxwell H. Gluck Equine Research Center, University of Kentucky, Lexington, KY 40546, USA

[email protected]

Received 5 January 2010 Accepted 4 May 2010

In 2006–2007, equine viral arteritis (EVA) was confirmed for the first time in Quarter Horses in multiple states in the USA. The entire genome of an equine arteritis virus (EAV) isolate from the index premises in New Mexico was 12 731 nt in length and possessed a previously unrecorded unique 15 nt insertion in the nsp2-coding region in ORF1a and a 12 nt insertion in ORF3. Sequence analysis of additional isolates made during this disease occurrence revealed that all isolates from New Mexico, Utah, Kansas, Oklahoma and Idaho had 98.6–100.0 % (nsp2) and 97.8–100 % (ORF3) nucleotide identity and contained the unique insertions in nsp2 and ORF3, indicating that the EVA outbreaks in these states probably originated from the same strain of EAV. Sequence and phylogenetic analysis of several EAV isolates made following an EVA outbreak on another Quarter Horse farm in New Mexico in 2005 provided evidence that this outbreak may well have been the source of virus for the 2006–2007 occurrence of the disease. A virus isolate from an aborted fetus in Utah was shown to have a distinct neutralization phenotype compared with other isolates associated with the 2006–2007 EVA occurrence. Full-length genomic sequence analysis of 18 sequential isolates of EAV made from eight carrier stallions established that the virus evolved genetically during persistent infection, and the rate of genetic change varied between individual animals and the period of virus shedding.

INTRODUCTION Equine arteritis virus (EAV) is the causative agent of equine viral arteritis (EVA) in horses, a respiratory and reproductive disease that occurs in many equine populations throughout the world (Timoney & McCollum, 1993). EAV is an enveloped virus with a single-stranded, positive-sense RNA genome of approximately 12.7 kb and belongs to the family Arteriviridae in the order Nidovirales (Cavanagh, 1997; Snijder & Meulenberg, 1998). The genomic RNA comprises nine known open reading frames (ORFs) flanked by a 59 untranslated region (59UTR) and a 39UTR (Snijder & Meulenberg, 1998; Snijder et al., 1999). ORF1a and -1b encode two replicase polyproteins (pp1a and pp1ab) that are processed post-translationally by three ORF1a-encoded proteinases (nsp1, -2 and -4) to yield at least 13 nonstructural proteins (nsp1–12, including nsp7a and 7b) The GenBank/EMBL/DDBJ accession numbers of the sequences reported in this paper are GQ903794–GQ903901. Three supplementary tables are available with the online version of this paper.

2286

(Snijder & Meulenberg, 1998; van Aken et al., 2006; Ziebuhr et al., 2000). The other seven ORFs (2a, 2b and 3– 7), respectively, encode the envelope proteins E, GP2, GP3, GP4, GP5 and M, and the nucleocapsid protein, N (Snijder & Spaan, 2006). The EAV GP5 protein has been shown to carry the major neutralization determinants of the virus. Four major neutralization sites (A–D) in the N-terminal ectodomain of the GP5 protein have been identified, including aa 49 (A), 60–61 (B), 67–90 (C) and 98–106 (D) (Balasuriya et al., 1997, 2004a; Yamaguchi et al., 1997; Zhang et al., 2008b). The vast majority of EAV infections are inapparent or subclinical, but occasionally outbreaks of EVA occur that are frequently characterized by an influenza-like illness in adult horses, abortion in pregnant mares and interstitial pneumonia in very young foals (McCollum et al., 1999; Timoney & McCollum, 1993). A variable percentage (from ,10 to 70 %) of stallions acutely infected with EAV can become persistently infected carriers, shedding the virus constantly in semen (Timoney & McCollum, 1993). Carrier stallions transmit EAV venereally to susceptible mares following natural breeding or artificial insemination.

Downloaded from www.microbiologyresearch.org by 019737 G 2010 SGM IP: 178.18.19.222 On: Sat, 17 Oct 2015 18:36:56

Printed in Great Britain

Characterization of the 2006–2007 EVA outbreak isolates

Infected mares can then disseminate the virus to susceptible cohorts with which they have close contact (Timoney & McCollum, 1988, 1993; Timoney et al., 1997). In 2006–2007, a multi-state occurrence of EVA was confirmed for the first time in Quarter Horses in the USA (Powell & Timoney, 2006; Timoney et al., 2006). In the year prior to the 2006–2007 occurrence of EVA, an outbreak of EVA was diagnosed retrospectively on a Quarter Horse breeding farm in New Mexico (NM) in proximity to the index premises of the 2006–2007 occurrence of the disease. To date, the molecular characteristics of the EAV isolates associated with the 2005 outbreak in NM or the 2006–2007 multi-state disease occurrence have not been determined. The primary objective of this study was to undertake molecular characterization of the EAV isolates obtained from the 2005 outbreak and during the 2006–2007 EVA occurrence in an attempt to identify the source of the virus strain initiating the 2006–2007 EVA occurrence. A second objective was to study the genetic evolution and phenotypic variation of EAV over the course of an extended series of outbreaks of the disease and during the course of persistent infection in stallions.

RESULTS Isolation and identification of EAV from the 2006– 2007 multi-state EVA occurrence and the 2005 EVA outbreak An outbreak of EVA was suspected on a Quarter Horse breeding farm in NM in June 2006. Laboratory testing of

sera from mares as well as serum and semen samples from the stallions on the index premises confirmed a diagnosis of EAV infection. Fresh cooled semen collected from one of the breeding stallions on the index premises in the late spring and early summer of 2006 together with donor/ recipient mares that had visited the index premises during the same time frame were traced to premises in 19 states (Fig. 1). Among these 19 states, EAV was isolated from horses on the premises of seven states (NM, UT, KS, OK, ID, CA and TX). In MT and AL, recent EAV infection was confirmed based on seroconversion or a significant (fourfold or higher) rise in neutralizing antibody titre to EAV between paired (acute and convalescent) sera, but no EAV isolates were obtained. In CO, animals had high antibody titres to EAV as well as epidemiological links to the index premises in NM, but no virus isolates were obtained. No virological or serological evidence of recent EAV infection was found in the nine remaining states (FL, IN, KY, LA, MN, MO, MS, SD and WY) that received shipped freshcooled semen and/or had mares visit the index premises in NM. This was the most extensive recorded occurrence of EVA resulting primarily from shipment of virus-infective semen and movement of donor or embryo recipient mares. In the course of this occurrence of EVA, EAV also spread to horses of another ten breeds in the USA (Table 1). Over the duration of the 2006–2007 EVA occurrence, EAV was isolated from 107 samples received from NM, UT, KS, OK, ID, CA and TX. These included 70 sequential isolates from the semen of 14 carrier stallions, 13 isolates from aborted fetuses, 18 from peripheral blood mononuclear cells (PBMCs) and six from sera. Four EAV isolates were obtained from the 2005 outbreak in NM; these comprised

Fig. 1. Multi-state occurrence of EVA in 2006–2007 in the USA. AL, Alabama; CA, California; CO, Colorado; FL, Florida; ID, Idaho; IN, Indiana; KS, Kansas; KY, Kentucky; LA, Louisiana; MN, Minnesota; MO, Missouri; MS, Mississippi; MT, Montana; NM, New Mexico; OK, Oklahoma; SD, South Dakota; TX, Texas; UT, Utah; WY, Wyoming. http://vir.sgmjournals.org

Downloaded from www.microbiologyresearch.org by IP: 178.18.19.222 On: Sat, 17 Oct 2015 18:36:56

2287

J. Zhang and others

Table 1. Background data on EAV isolates used in this study Sample source

Virus isolate ID

Collection date

Source of virus

2006–2007 multi-state EVA occurrence isolates Stallion B S3685* 18 Jun 2006 Semen S3861* 24 Feb 2007 Semen Stallion C S3699* 10 Jul 2006 Semen S3961* 11 Jun 2007 Semen S4417* 17 Jul 2008 Semen Stallion D S3712 1 Aug 2006 Semen S4007 13 Jul 2007 Semen Stallion E S3711 1 Aug 2006 Semen S4227 6 Feb 2008 Semen Stallion F S3854 20 Feb 2007 Semen S4333 29 Apr 2008 Semen Stallion G S3886 17 Mar 2007 Semen S4421 22 Jul 2008 Semen Stallion H S3943 9 May 007 Semen S4445 13 Aug 2008 Semen Stallion I S3686* 19 Jun 2006 Semen Stallion J S3707* 18 Jul 2006 Semen Stallion K S3901 22 Mar 2007 Semen Stallion L S3881 13 Mar 2007 Semen Stallion M S3874 4 Mar 2007 Semen Stallion N S3955 31May 2007 Semen Stallion O S4222 1 Feb 2008 Semen Mare SR7269 27 Jun 2006 Serum Filly SR7336 26 Jun 2006 Serum Mare SR7406 1 Jul 2006 Serum Stallion SR7415 1 Jul 2006 Serum Mare SR7423 1 Jul 2006 Serum Mare SR7478 6 Jul 2006 Serum Stallion M436* 26 Jun 2006 PBMCs Mare M437 27 Jun 2006 PBMCs Aborted fetus M438* 30 Jun 2006 Lung Stallion M440 7 Jul 2006 PBMCs Mare M441 7 Jul 2006 PBMCs Mare M442 10 Jul 2006 PBMCs Mare M455 10 Jul 2006 PBMCs Aborted fetus M456 10 Jul 2006 Lung Aborted fetus M468 20 Jul 2006 Lung and Aborted fetus M470 19 Jul 2006 Lung and Mare M476 29 Jul 2006 PBMCs Aborted fetus M477 29 Jul 2006 Lung and Aborted fetus M479 29 Jul 2006 Lung and Mare M488 14 Aug 2006 PBMCs Colt M495 23 Aug 2006 PBMCs Stallion M498 23 Aug 2006 PBMCs Mare M507 22 Sep 2006 PBMCs Gelding M509 22 Sep 2006 PBMCs Mare M511 22 Sep 2006 PBMCs Gelding M515 27 Sep 2006 PBMCs Mare M517 27 Sep 2006 PBMCs Gelding M519 27 Sep 2006 PBMCs Aborted fetus M527 25 Oct 2006 Lung Mare M528 25 Oct 2006 PBMCs Mare M532 29 Oct 2006 PBMCs Aborted fetus M533 29 Oct 2006 Lung and

2288

placenta placenta placenta placenta

placenta

Horse breed

Quarter Horse Quarter Horse Quarter Horse Quarter Horse Quarter Horse Quarter Horse Quarter Horse Quarter Horse Quarter Horse Arabian Arabian Arabian Arabian Paint Paint Quarter Horse Quarter Horse Swiss Warmblood American Saddlebred Quarter Horse Andalusian Hannovarian Quarter Horse Quarter Horse Quarter Horse Arabian Quarter Horse Quarter Horse Quarter Horse Quarter Horse Quarter Horse Quarter Horse Arabian Quarter Horse Quarter Horse Quarter Horse Quarter Horse Quarter Horse Quarter Horse Quarter Horse Quarter Horse Arabian Arabian Arabian Oldenburg Oldenburg Welsh Thoroughbred Paint Paint Paint Quarter Horse Quarter Horse Quarter Horse Quarter Horse

State

NM NM NM NM NM UT UT UT UT UT UT UT UT ID ID NM NM NM UT ID TX CA NM OK UT UT UT NM NM KS NM UT UT UT NM NM NM NM NM NM NM UT UT UT UT UT UT UT UT UT UT UT UT UT


Sequences determined (GenBank accession nos) Full-length (GQ903794) Full-length (GQ903795) Full-length (GQ903796) Full-length (GQ903797) Full-length (GQ903798) Full-length (GQ903799) Full-length (GQ903800) Full-length (GQ903801) Full-length (GQ903802) Full-length (GQ903803) Full-length (GQ903804) Full-length (GQ903805) Full-length (GQ903806) Full-length (GQ903807) Full-length (GQ903808) nsp2 (GQ903812); ORF2–7 nsp2 (GQ903813); ORF2–7 nsp2 (GQ903814); ORF2–7 nsp2 (GQ903815); ORF2–7 nsp2 (GQ903816); ORF2–7 nsp2 (GQ903817); ORF2–7 nsp2 (GQ903818); ORF2–7 nsp2 (GQ903819); ORF2–7 nsp2 (GQ903820); ORF2–7 nsp2 (GQ903821); ORF2–7 nsp2 (GQ903822); ORF2–7 nsp2 (GQ903823); ORF2–7 nsp2 (GQ903824); ORF2–7 nsp2 (GQ903825); ORF2–7 nsp2 (GQ903826); ORF2–7 nsp2 (GQ903827); ORF2–7 nsp2 (GQ903828); ORF2–7 nsp2 (GQ903829); ORF2–7 nsp2 (GQ903830); ORF2–7 nsp2 (GQ903831); ORF2–7 nsp2 (GQ903832); ORF2–7 nsp2 (GQ903833); ORF2–7 nsp2 (GQ903834); ORF2–7 nsp2 (GQ903835); ORF2–7 nsp2 (GQ903836); ORF2–7 nsp2 (GQ903837); ORF2–7 nsp2 (GQ903838); ORF2–7 nsp2 (GQ903839); ORF2–7 nsp2 (GQ903840); ORF2–7 nsp2 (GQ903841); ORF2–7 nsp2 (GQ903842); ORF2–7 nsp2 (GQ903843); ORF2–7 nsp2 (GQ903844); ORF2–7 nsp2 (GQ903845); ORF2–7 nsp2 (GQ903846); ORF2–7 nsp2 (GQ903847); ORF2–7 nsp2 (GQ903848); ORF2–7 nsp2 (GQ903849); ORF2–7 nsp2 (GQ903850); ORF2–7

(GQ903857) (GQ903858) (GQ903859) (GQ903860) (GQ903861) (GQ903862) (GQ903863) (GQ903864) (GQ903865) (GQ903866) (GQ903867) (GQ903868) (GQ903869) (GQ903870) (GQ903871) (GQ903872) (GQ903873) (GQ903874) (GQ903875) (GQ903876) (GQ903877) (GQ903878) (GQ903879) (GQ903880) (GQ903881) (GQ903882) (GQ903883) (GQ903884) (GQ903885) (GQ903886) (GQ903887) (GQ903888) (GQ903889) (GQ903890) (GQ903891) (GQ903892) (GQ903893) (GQ903894) (GQ903895)

Journal of General Virology 91


Table 1. cont. Sample source

Virus isolate ID

Aborted fetus M535 Aborted fetus M537 Aborted fetus M547 Aborted fetus M558 Aborted fetus M578 2005 EVA outbreak isolates Stallion A S3583 S3817 S4216 Aborted fetus M405

Collection date

Source of virus

Horse breed

State

Sequences determined (GenBank accession nos)

28 Oct 2006 30 Oct 2006 8 Jan 2007 12 Jan 2007 16 Apr 2007

Placenta Paint Lung and placenta Quarter Horse Lung and placenta Quarter Horse Lung and placenta Appaloosa Lung Quarter Horse

UT UT UT UT ID

nsp2 nsp2 nsp2 nsp2 nsp2

5 Nov 2005 12 Jan 2007 29 Jan 2008 11 Dec 2005

Semen Semen Semen Placenta

NM NM NM NM

Full-length (GQ903809) Full-length (GQ903810) Full-length (GQ903811) nsp2 (GQ903856); ORF2–7 (GQ903901)

Quarter Horse Quarter Horse Quarter Horse Thoroughbred

(GQ903851); (GQ903852); (GQ903853); (GQ903854); (GQ903855);

ORF2–7 ORF2–7 ORF2–7 ORF2–7 ORF2–7

(GQ903896) (GQ903897) (GQ903898) (GQ903899) (GQ903900)

*EAV isolates obtained from the index premises in NM.

one isolate from the placenta of an aborted fetus and three sequential isolates from one carrier stallion.

envelope glycoproteins, GP2–5, with the greatest variation observed in the GP3 protein (Supplementary Table S1).

A total of 59 EAV isolates obtained during the 2006–2007 multi-state EVA occurrence [including 22 sequential isolates from the semen of 14 carrier stallions (B–O) and all 37 isolates recovered from aborted fetuses, PBMCs and sera] as well as four isolates from the 2005 EVA outbreak in NM were subjected to sequence and phylogenetic analysis (Table 1).

The genome of isolate S3685 contained a 15 nt insertion in the nsp2-coding region in ORF1a and a 12 nt insertion in ORF3 when compared with the VB and CW96 strains of EAV (Fig. 2). The consecutive in-frame 15 nt insertion (59-GACGCCGTCCACAGT-39) was located between nt 1559 and 1560, resulting in the insertion of five amino acids (DAVHS) between aa 445 and 446 of the nsp2 protein (Fig. 2a). The consecutive in-frame 12 nt insertion (59-AGTGCATTTGGA-39) was located between nt 10668 and 10669, resulting in the insertion of four amino acids (SAFG) between aa 121 and 122 of the GP3 protein (Fig. 2b). Further analysis revealed that these insertions in nsp2 and ORF3 were absent from all previously sequenced EAV strains reported in the scientific literature or deposited in GenBank (data not shown), suggesting that EAV isolate S3685 from stallion B is a previously unrecorded strain of EAV.

EAV isolates associated with the 2006–2007 multi-state EVA occurrence have unique genetic hallmarks in nsp2 and ORF3 Fresh cooled semen from one of the infected stallions (stallion B) on the index premises in NM had been shipped to premises in a significant number of other states prior to 16 June 2006, when such shipments were suspended. The first EAV isolate (S3685) from the semen of this stallion (B) was subjected to full-length genome sequencing. The full-length sequence of S3685 was compared with that of the prototype virulent Bucyrus (VB) strain of EAV (North American phylogeny group, GenBank accession no. DQ846750; Balasuriya et al., 2007) and the CW96 strain of EAV (European phylogeny group, GenBank accession no. AY349167; Balasuriya et al., 2004b). The complete genome of isolate S3685 was 12 731 nt in length and was 27 and 23 nt longer than the VB and CW96 strains of EAV, respectively. It had 1606 (87.3 % identity) and 1955 (84.6 % identity) nucleotide differences from the VB and CW96 strains of EAV, respectively, at the level of the entire genome. Among the replicase proteins, the greatest variation occurred in the nsp2 protein, followed by the nsp12 and nsp11 proteins (see Supplementary Table S1, available in JGV Online). However, all predicted cleavage sites on the replicase polyproteins were conserved compared with the published sequence of the VB strain. Among the structural proteins, most variation occurred in the four http://vir.sgmjournals.org

To examine further the genetic characteristics of EAV isolates associated with the 2006–2007 EVA occurrence, nsp2 and ORF3 sequences were determined for 58 additional isolates, which comprised 21 isolates from the semen of 14 carrier stallions (stallions B–O) and all 37 EAV isolates from aborted fetuses, PBMCs and sera (Table 1). All isolates obtained from NM, UT, KS, OK and ID had 98.6–100 % (based on nsp2) and 97.8–100 % (based on ORF3) nucleotide identity with isolate S3685 and with each other; all contained the unique 15 nt insertion in the nsp2-coding region in ORF1a (Fig. 2a) and the unique 12 nt insertion in ORF3 (Fig. 2b). These findings were strongly indicative that the occurrence of EVA in these states originated from the same virus strain. In contrast, two EAV isolates obtained from TX (isolate S3955) and CA (isolate S4222) had significantly lower nucleotide identity with the outbreak isolates (S3955: 80.3–80.8 % nucleotide identity with the outbreak isolates based on nsp2 and 82.2–83.8 % based on ORF3; S4222: 83.7–84.2 % based on nsp2 and 82.2–83.8 % based on ORF3). Neither strain contained the characteristic 15 nt insertion in the


2289

J. Zhang and others

nsp2-coding region or the 12 nt insertion in ORF3 (Fig. 2), confirming that the EAV infections in TX and CA were epidemiologically unrelated to the major EVA occurrence involving the other states. Tracing the putative source of EAV for the 2006– 2007 multi-state occurrence of EVA In late 2005, an outbreak of EVA was diagnosed retrospectively on another large Quarter Horse breeding farm in NM in proximity to the index premises of the 2006–2007 occurrence of EVA. The nucleotide sequences of nsp2coding region and ORF3 of four EAV isolates obtained in connection with the 2005 outbreak (isolates S3583, S3817 and S4216 from carrier stallion A and isolate M405 from the placenta of an aborted fetus, Table 1) were determined and compared with those of the 2006–2007 isolates. These four isolates had 98.6–99.5 % (based on nsp2) and 97.4– 99.8 % (based on ORF3) nucleotide identity with the isolates from 2006–2007 (from NM, UT, KS, OK and ID) and all four contained the unique 15 nt insertion in the nsp2-coding region and 12 nt insertion in ORF3 (Fig. 2). These findings provide strong circumstantial evidence that the 2005 outbreak may well have served as the source of virus for the 2006–2007 multi-state EVA occurrence. Phylogenetic analysis of EAV isolates from the 2006–2007 occurrence and the 2005 outbreak Based on phylogenetic analysis of the EAV ORF5 sequences (Fig. 3), available global isolates of EAV clustered into two distinct groups: a North American group and a European group. The vast majority of isolates in the European group could be further divided into two subgroups, European subgroup 1 (EU-1) and European subgroup 2 (EU-2). The EAV isolates obtained from NM, UT, KS, OK and ID during the 2006–2007 EVA occurrence and the four isolates from the 2005 EVA outbreak formed a separate cluster within EU-1 and were distinctly different from any previously sequenced strains of EAV (Fig. 3). Not surprisingly, isolate S3955 from TX and isolate S4222 from CA clustered differently from isolates from the 2006–2007 occurrence as well as the 2005 outbreak isolates (Fig. 3). Isolate S3955 from TX clustered with the North American group, whereas isolate S4222 from CA belonged to EU-1 but was distantly related to other virus isolates in this phylogenetic subgroup. Phylogenetic analysis findings reaffirmed the conclusion that the 2006–2007 multi-state EVA occurrence and the 2005 outbreak probably originated from the same unique strain of EAV, whilst the EAV outbreaks in TX and CA were epidemiologically and virologically unrelated to the major occurrence of the disease in 2006–2007. Genetic evolution of EAV during widespread disease occurrence In an attempt to determine the genomic stability of EAV during horizontal and vertical transmission of the virus in 2290

the 2006–2007 occurrence of EVA, sequence analysis of the nsp2-coding region in ORF1a as well as ORF2a, ORF2b and ORF3–7 was carried out for 49 EAV isolates. These comprised the first isolates from semen of each of the carrier stallions B–M (S3685, S3699, S3712, S3711, S3854, S3886, S3943, S3686, S3707, S3901, S3881 and S3874, 12 isolates in total) and 37 isolates from aborted fetuses, PBMCs and sera (Table 1). All 49 EAV isolates had the same length with respect to the nsp2-coding region in ORF1a (1728 nt) and ORF2–7 (2907 nt). The six EAV isolates collected on the index premises in NM from June to July 2006 differed only slightly from each other by 0 to 3 nt (100–99.8 % identity) based on comparison of 1728 nt in the nsp2-coding region, and by 0 to 6 nt (100–99.8 % identity) based on comparison of 2907 nt in ORF2–7 (see Supplementary Table S2, available in JGV Online). With virus spread to other premises in other states (NM, UT, OK and KS) during the period June–October 2006, the genomic variability among EAV isolates increased slightly (Supplementary Table S2). Following spread of the virus to additional premises and states over the period June 2006– April 2007, genomic variability observed among EAV isolates increased (Supplementary Table S2). This suggested that the EAV genome is genetically relatively stable over the course of a short time frame. However, genomic variability tended to increase where horizontal and vertical transmission took place over a more extended time period. Comparative amino acid sequence analysis of the 49 EAV isolates revealed that the potential N-glycosylation sites on the GP2 and GP4 proteins were conserved among all 49 isolates, whereas loss or acquisition of potential Nglycosylation sites occurred on GP3 and GP5 proteins (Fig. 4a). In the case of the GP3 protein, potential Nglycosylation sites Asn-28, Asn-29, Asn-49, Asn-96 and Asn-106 were conserved in all 49 isolates. The amino acid substitution at residue 120 (SerAAsn) resulted in loss of the N-glycosylation site Asn-118 in two isolates (S3854 and S3943), but led to the acquisition of an additional putative N-glycosylation site, Asn-120, in these two isolates. In the case of the GP5 protein, the potential N-glycosylation sites Asn-56 and Asn-81 were conserved in all 49 isolates. A third potential N-glycosylation site at aa 73 (Asn-73) was observed for the first time on the GP5 protein of five EAV isolates obtained from clinical specimens from three different states (NM: M405 and M477 from aborted fetuses; UT: M517 from PBMCs and M547 from an aborted fetus; ID: M578 from an aborted fetus). Genetic evolution of EAV during persistent infection in the stallion The full-length genomic sequences of 18 EAV isolates sequentially collected from eight carrier stallions over an 8– 26 month period (stallions A–H; three isolates from each of stallions A and C and two isolates from each of stallions B and D–H; Table 2) were determined to study the genetic evolution of EAV during persistent infection in the stallion.




Fig. 2. Alignment of (a) partial nucleotide sequences of the nsp2-coding region in ORF1a and (b) ORF3. The nsp2 and ORF3 sequences of 22 EAV isolates from the semen of 14 carrier stallions (B–O) and all 37 EAV isolates from aborted fetuses, PBMCs and sera were compared with each other and with the VB and CW96 strains of EAV. Four EAV isolates from the 2005 EVA outbreak were also included for comparison (S3583, S3817, S4216 and M405), as shown at the bottom of the alignment. The nucleotides were numbered according to their position in the published sequence of the VB strain of EAV (GenBank accession no. DQ846750).

http://vir.sgmjournals.org


2291

J. Zhang and others

The complete genome of all 18 EAV isolates was 12 731 nt in length and each maintained the unique 15 nt insertion in nsp2 and 12 nt insertion in ORF3, although point mutations within the insertion region were observed for

some isolates (Fig. 2). Nucleotide and amino acid substitutions were observed in sequential isolates from all eight carrier stallions, but the evolution rate varied. As shown in Table 2, the nucleotide change rate (per year at

Fig. 3. Phylogenetic analysis of the partial ORF5 nucleotide sequences of 63 EAV isolates included in this study and 287 previously published EAV sequences (350 in total). Bootstrap analysis was carried out on 1000 replicate datasets and values are indicated adjacent to the major branching points. The North American group and European group (including European subgroups 1 and 2) are depicted in the figure. The 63 EAV isolates analysed in this study [2006–2007 EVA occurrence: 22 sequential isolates from the semen of 14 carrier stallions (B–O) and 37 isolates from aborted fetuses, PBMCs and sera; 2005 EVA outbreak: four isolates] are shown in bold. Bar, 0.05 nucleotide substitutions per site. 2292




the entire genome level) of the virus in the reproductive tract could vary from 0.08 to 0.36 % in different carrier stallions. In individual carrier stallions, the longer the shedding period, the more nucleotide substitutions accumulated in the virus genome, although some nucleotide reversions also occurred during this time frame. For example, in stallion A, isolates S3583 and S3817 had 12 nucleotide substitutions at the entire genome level over a 14-month interval and isolates S3583 and S4216 had 42 nucleotide changes over a 26-month interval; in stallion C, isolates S3699 and S3961 had 15 nucleotide changes over an 11-month interval and isolates S3699 and S4417 had 72 nucleotide changes over a 24-month interval. Sequence analysis of 18 EAV isolates from eight carrier stallions revealed that nucleotide mutations observed in the course of persistent infection could span the entire genome. However, the most consistently observed mutations during persistent infection were only identified at a number of positions (Fig. 4c). These included two (10409CAT and 12068CAT), three (10679A«G, 11389AAG

CAT) and three (9828CAT, 10681G«A and TAA) nucleotide changes that occurred during persistent infection in three out of eight, four out of eight and five out of eight carrier stallions, respectively. The nucleotide mutations in ORF2a, -2b and -6 (encoding the E, GP2 and M proteins, respectively) were silent mutations and amino acid changes were only observed in the GP3 and GP5 proteins encoded by ORF3 and -5, respectively (Fig. 4c).

and

12144

11372

To compare selective pressure exerted on individual EAV proteins, the ratios of the number of non-synonymous substitutions per non-synonymous site (dN) and synonymous substitutions per synonymous site (dS) were estimated. In this study, none of the dN : dS ratios estimated by pairwise comparison of all non-structural and structural coding regions of the EAV isolates from each of the eight carrier stallions was significantly greater than 1 (data not shown). Therefore, positive selection did not contribute significantly to nucleotide diversity of the EAV isolates during persistent infection in stallions.

(a) Potential N-glycosylation site changes during horizontal and vertical transmission (49 isolates) 120 29 56 73 49 96 106 118 (amino acid position) 28 GP3

NNTT 49 49

NVT 49

** * * NCS NAS NSSASAFGG 49 49 47 ** * * NSNASAFGG 2 (no. isolates)

81 (amino acid position)

NCS IIT NDT 44 49 49 NIT 5

GP5

(no. isolates)

(b) Potential N-glycosylation site changes during persistent infection in eight carrier stallions (18 isolates) 29 120 106 118 (amino acid position) 49 96 28 * SAFG is 4 aa insertion resulting from the 12 nt insertion in ORF3 * * * * NNTT NVT NCS NAS NSSA/TSAFGG GP3 15 17 18 18 12 **** NNIT NVI NSNA/TSAFGG 3 1 6 (no. isolates) (c) Mutations consistently identified during persistent infection in eight carrier stallions Non-structural protein genes

Structural protein genes E GP2 2a 2b

5' L

ORF1a

3

ORF1b

M

GP4 4 GP3

3'

6 5 GP5

7 N

An

12068 (M-47) C (Tyr)

10409 (GP2-191; GP3-30) C (Leu; Thr)

Mutations observed in three of eight carrier stallions T (Tyr)

T (Leu; Ile)

10679 (GP3-120) 11389 (GP5-73) 12144 (M-73) A (Asn) C (Leu) A (Ile)

Mutations observed in four of eight carrier stallions

G (Ser)

Mutations observed in five of eight carrier stallions

9828 (E-21) C (Ile)

G (Val)

T (Leu)

10681 (GP3-121) 11372 (GP5-67) G (Ala) T (Val)

T (Ile)

A (Thr)

A (Glu)

Fig. 4. Changes in potential N-glycosylation sites in EAV isolates during horizontal and vertical transmission (a) and during persistent infection in the stallion (b), as well as mutations consistently observed during persistent infection in the stallion (c). L, Leader sequence. http://vir.sgmjournals.org


2293

J. Zhang and others

2294

Table 2. Genetic evolution of EAV during persistent infection in eight carrier stallions NA,

Not applicable.

Genome region/ ORF (nt range)*

Protein/region (length in aa)

Carrier stallion and sample IDD Stallion A


59UTR (1–224) NA Non-structural proteins (nsp) ORF1ab (225– 1ab polyprotein 9766) (3180) nsp1: Met-1–Gly260 (260) nsp2: Gly-261–Gly836 (576) nsp3: Gly-837–Glu1069 (233) nsp4: Gly-1070– Glu-1273 (204) nsp5: Ser-1274– Glu-1435 (162) nsp6: Gly-1436– Glu-1457 (22) nsp7: Ser-1458– Glu-1682 (225) nsp8: Gly-1683– Asn-1732 (50) nsp9: Gly-1683– Glu-2375 (693) nsp10: Ser-2376– Gln-2842 (467) nsp11: Ser-2843– Glu-3061 (219) nsp12: Gly-3062– Val-3180 (119)

Stallion B

Stallion C

Stallion D

Stallion E

Stallion F

Stallion G

Stallion H

S3583 and S3817 (14)

S3583 and S4216 (26)

S3685 and S3861 (8)

S3699 and S3961 (11)

S3699 and S4417 (24)

S3712 and S4007 (11.5)

S3711 and S4227 (18)

S3854 and S4333 (14)

S3886 and S4421 (16)

S3943 and S4445 (15)

0 (NA)

1 (NA)

0 (NA)

0 (NA)

1 (NA)

0 (NA)

1 (NA)

0 (NA)

0 (NA)

0 (NA)

0 (0)

2 (1)

0 (0)

0 (0)

6 (0)

2 (1)

0 (0)

0 (0)

0 (0)

3 (1)

0 (0)

1 (0)

3 (1)

1 (0)

14 (6)

0 (0)

0 (0)

0 (0)

7 (1)

5 (2)

0 (0)

2 (0)

1 (0)

0 (0)

2 (1)

2 (0)

0 (0)

0 (0)

1 (0)

0 (0)

1 (1)

2 (1)

4 (2)

1 (0)

5 (1)

0 (0)

0 (0)

1 (1)

2 (0)

1 (0)

1 (0)

2 (0)

2 (1)

0 (0)

2 (0)

0 (0)

0 (0)

0 (0)

1 (0)

0 (0)

0 (0)

0 (0)

0 (0)

0 (0)

0 (0)

0 (0)

0 (0)

0 (0)

0 (0)

0 (0)

1 (1)

3 (1)

2 (0)

0 (0)

3 (0)

0 (0)

1 (1)

0 (0)

2 (0)

1 (0)

0 (0)

0 (0)

0 (0)

0 (0)

0 (0)

1 (0)

0 (0)

0 (0)

0 (0)

0 (0)

1 (1)

3 (2)

5 (1)

5 (5)

13 (4)

5 (2)

1 (1)

0 (0)

3 (0)

2 (1)

1 (1)

7 (2)

3 (0)

1 (0)

4 (0)

2 (0)

0 (0)

2 (1)

3 (0)

5 (0)

0 (0)

2 (0)

2 (1)

0 (0)

4 (1)

2 (1)

0 (0)

0 (0)

3 (1)

2 (0)

0 (0)

2 (0)

1 (0)

0 (0)

2 (1)

2 (0)

0 (0)

0 (0)

1 (0)

1 (0)



Table 2. cont. Genome region/ ORF (nt range)*

Protein/region (length in aa)

Carrier stallion and sample IDD Stallion A

Stallion C

Stallion D

Stallion E

Stallion F

Stallion G

Stallion H

S3583 and S3817 (14)

S3583 and S4216 (26)

S3685 and S3861 (8)

S3699 and S3961 (11)

S3699 and S4417 (24)

S3712 and S4007 (11.5)

S3711 and S4227 (18)

S3854 and S4333 (14)

S3886 and S4421 (16)

S3943 and S4445 (15)

0 (0)

1 (0)

2 (0)

0 (0)

1 (0)

0 (0)

1 (0)

1 (0)

2 (0)

1 (0)

0 (0)

0 (0)

4 (1)

2 (2)

4 (3)

2 (1)

3 (2)

0 (0)

2 (1)

0 (0)

3 (2)

2 (1)

3 (2)

0 (0)

5 (3)

3 (2)

6 (4)

4 (4)

7 (2)

4 (2)

0 (0)

6 (4)

1 (0)

0 (0)

2 (1)

0 (0)

1 (0)

1 (0)

1 (0)

1 (0)

3 (2)

4 (3)

0 (0)

4 (3)

5 (4)

6 (4)

4 (2)

3 (1)

2 (1)

3 (2)

1 (0)

3 (0)

1 (0)

0 (0)

1 (0)

2 (0)

0 (0)

2 (0)

3 (1)

3 (0)

0 (0)

0 (0)

0 (0)

1 (0)

0 (0)

1 (0)

0 (0)

0 (0)

0 (0)

0 (0)

0 (0)

0 (0)

0 (0)

0 (0)

0 (0)

0 (0)

0 (0)

0 (0)

0 (0)

0 (0)

12 (8)

42 (15)

30 (9)

15 (10)

72 (25)

27 (11)

15 (10)

14 (7)

38 (7)

32 (8)

0.30

0.15

0.36

0.13

0.29

0.23

0.08

0.10

0.23

0.20

*Nucleotides are numbered according to the sequence of EAV isolate S3685 (GenBank accession no. GQ903794). DResults are shown as the number of nucleotide (amino acid) differences between each two indicated EAV isolates in the indicated genome regions. The interval between the collection dates of two sequential isolates is given in parentheses (months). dTotal number of nucleotide differences at the entire genome level. Because of the overlapping nature of each two adjacent ORFs, the total number of nucleotide differences is not the simple sum of the nucleotide differences of each ORF.

2295 Downloaded from www.microbiologyresearch.org by IP: 178.18.19.222 On: Sat, 17 Oct 2015 18:36:56


Structural proteins ORF2a (9766– E (67) 9969) ORF2b (9839– GP2 (227) 10522) ORF3 (10321– GP3 (167) 10824) ORF4 (10727– GP4 (152) 11185) ORF5 (11173– GP5 (255) 11940) ORF6 (11928– M (162) 12416) ORF7 (12340– N (110) 12672) NA 39UTR (12673– 12731) Total no. differencesd 0.08 Nucleotide change year21 at entire genome level (%)

Stallion B

J. Zhang and others

Comparative amino acid sequence analysis of 18 EAV isolates from the eight carrier stallions revealed that the potential N-glycosylation sites on the GP2, GP4 and GP5 proteins were conserved during persistent infection in all eight carrier stallions. All EAV isolates from the eight carrier stallions had two potential N-glycosylation sites (Asn-56 and Asn-81) on the GP5 protein and the potential new N-glycosylation site, Asn-73, was not observed. However, loss or acquisition of potential N-glycosylation sites occurred in the GP3 protein during persistent infection in the stallion (Fig. 4b). The potential Nglycosylation sites Asn-29, Asn-96 and Asn-106 were conserved during persistent infection in all eight carrier stallions. The substitution at aa 30 (ThrAIle) resulted in loss of the N-glycosylation site Asn-28 during persistent infection in three carrier stallions (B, D and G). The substitution at aa 51 (ThrAIle) resulted in loss of the Nglycosylation site Asn-49 during persistent infection in carrier stallion E. The substitution at aa 120 (SerAAsn) resulted in loss of the N-glycosylation site Asn-118 in six isolates (S4417, S4007, S3854, S3943, S4445 and S3583), whereas an additional N-glycosylation site, Asn-120, was acquired by these six isolates. Characterization of neutralization phenotypes of EAV isolates A panel of 14 monoclonal antibodies (mAbs) and a panel of nine polyclonal equine antisera were used to characterize the neutralization phenotypes of 14 EAV isolates from this study as well as four previously described EAV strains (EAV ATCC, ARVAC, KY84 and CW96) for comparison. With the exception of the EAV M547 isolate, all of the isolates from this study were neutralized by mAbs 5G11, 6D10, 7E5, 9F2, 10F11, 10H4 and 6A2 (Table 3). However, mAb 6A2 exhibited significant variability in its ability to neutralize different EAV isolates. The neutralizing mAbs 1H7, 1H9, 5E8, 7D4 and 10B4 did not neutralize any of the EAV isolates evaluated. With respect to the polyclonal equine sera, the M547 isolate was not neutralized to a high titre by EAV strain-specific polyclonal equine sera except for the anti-KY77 serum; all other EAV isolates were neutralized by the polyclonal equine sera, but the anti-ARVAC and antiNVSL polyclonal equine sera generally neutralized these EAV isolates to a lower titre compared with the other antisera (Table 3). In summary, the neutralization data suggested that the M547 isolate had a distinct neutralization phenotype. Comparative amino acid sequence analysis of the critical neutralization regions of the GP5 protein of this isolate (Fig. 5) suggested that aa 62 (site B), 73 (site C), and 99 and 101 (site D) on the GP5 protein may contribute to its distinctive neutralization phenotype.

DISCUSSION The seroprevalence of EAV infection can vary among horses of different breeds (Hullinger et al., 2001; 2296

McCollum & Bryans, 1973; McCue et al., 1991; Timoney & McCollum, 1993). The 1998 National Animal Health Monitoring System’s equine survey showed that only 0.6 % of the Quarter Horse population was seropositive to EAV (NAHMS, 2000), indicating that the American Quarter Horse population was essentially totally naı¨ve with respect to prior contact with EAV and, therefore, fully susceptible should future exposure to infection occur. In mid-2006, an EVA outbreak occurred on a large Quarter Horse breeding farm in NM, with EAV infection spreading rapidly in a presumed immunologically naı¨ve horse population (Timoney et al., 2006). The virus subsequently spread to premises in other states, and outbreaks continued to occur up to early 2007. The origin of the virus responsible for this extensive multi-state occurrence of EVA had not been identified. The findings of this study clearly demonstrated that the multi-state disease occurrence in NM, UT, KS, OK and ID was directly related to infection with a previously unrecorded and genetically unique strain of EAV. The virus present in the semen of stallion A and the virus isolated from the placenta of an aborted fetus during the 2005 EVA outbreak were very similar (high nucleotide identity as well as unique insertions in nsp2 and ORF3) to viruses recovered from the 2006–2007 EVA occurrence. Therefore, the 2005 EVA outbreak on another Quarter Horse farm in NM may well have been the source of virus for the 2006–2007 multi-state EVA occurrence. In fact, there is epidemiological evidence to suggest that infection was introduced onto the index premises for the 2006–2007 occurrence through the movement of a mare from the premises involved in the EVA outbreak in 2005. The findings from this study re-emphasize the value of sequence analysis and comparison of virus strains in tracing the origin of disease outbreaks. Molecular characterization of the EAV isolates associated with the 2006–2007 disease occurrence revealed several features that had not been observed previously. Firstly, this new strain of EAV possessed a unique 15 nt insertion in the nsp2-coding region of ORF1a and a unique 12 nt insertion in ORF3. This represents the longest EAV genome (12 731 nt) reported to date. The biological significance (e.g. the effects on virus growth, genome replication, virulence phenotype and abortion potential) of these insertions remains to be determined. Secondly, this study found some previously unrecognized potential N-glycosylation sites on the GP5 and GP3 proteins of EAV. The potential N-glycosylation site Asn-56 on GP5 has been conserved among all strains of EAV examined thus far. The presence of the potential N-glycosylation site Asn-81 on GP5 was variable; it has been observed in some field and laboratory strains of EAV (Balasuriya & MacLachlan, 2004). A third potential N-glycosylation site, Asn-73, on GP5 was observed for the first time in this study. Its occurrence was variable and was identified only on some field isolates of EAV from this occurrence of EVA. The EAV isolates recovered from this disease occurrence had the potential N-glycosylation site Asn-118 or Asn-120, but




Table 3. Neutralization titres of mAbs and polyclonal equine antisera against various EAV isolates Neutralization titres are expressed as the inverse of the antibody dilution providing 50 % protection of RK-13 cell monolayers against 200 TCID50 of virus. EAV strain

Neutralizing mAbs*

5G11 6D10 7E5

Polyclonal equine antiserad

10F11 10H4 1H7 1H9 5E8 6A2 7D4

10B4

3E2

12A4

4096 .4096 512 512 .4096 .4096 ,32 ,32 ,32 256 ,32 ,32 512 32 ,32 2048 2048 ,32 ,32 ,32 ,32 ,32 512 2048 256 256 4096 1024 ,32 ,32 ,32 256 ,32 128 1024 512 64 2048 1024 ,32 ,32 ,32 256 ,32 1024 4096 128 128 2048 1024 ,32 ,32 ,32 256 ,32 512 2048 128 256 1024 1024 ,32 ,32 ,32 64 ,32 2048 2048 512 1024 4096 1024 ,32 ,32 ,32 1024 ,32 2048 4096 128 128 4096 2048 ,32 ,32 ,32 256 ,32 1024 4096 512 1024 4096 1024 ,32 ,32 ,32 2048 ,32 2048 .4096 128 1024 4096 2048 ,32 ,32 ,32 512 ,32 1024 4096 256 512 4096 1024 ,32 ,32 ,32 1024 ,32 1024 4096 256 256 4096 2048 ,32 ,32 ,32 32 ,32 2048 .4096 128 512 4096 4096 ,32 ,32 ,32 1024 ,32 512 4096 64 256 4096 2048 ,32 ,32 ,32 ,32 ,32 1024 2048 512 128 4096 2048 ,32 ,32 ,32 256 ,32 2048 4096 512 1024 4096 2048 ,32 ,32 ,32 1024 ,32 ,32 ,32 ,32 ,32 ,32 ,32 ,32 ,32 ,32 ,32 ,32 512 4096 1024 1024 2048 2048 ,32 ,32 ,32 256 ,32

,32 ,32 ,32 ,32 ,32 ,32 ,32 ,32 ,32 ,32 ,32 ,32 ,32 ,32 ,32 ,32 ,32 ,32

,32 ,32 ,32 ,32 ,32 ,32 ,32 ,32 ,32 ,32 ,32 ,32 ,32 ,32 ,32 ,32 ,32 ,32

,32 .1024 ,32 256 ,32 64 ,32 128 ,32 128 ,32 128 ,32 128 ,32 128 ,32 128 ,32 256 ,32 256 ,32 256 ,32 1024 ,32 64 ,32 128 ,32 128 ,32 8 ,32 128

Control serum

Anti- Anti- Anti- Anti- Anti- Anti- AntiAnti- NVSLNEG§ KY53 Bibuna Vienna KY77 KY84 CW01 ARVAC EAVNVSL 512 64 32 128 128 128 128 128 128 64 128 128 1024 64 128 256 8 256

*Neutralizing mAbs to EAV (Balasuriya et al., 1995a, 1997). DControl mAbs to (3E2) N protein and (12A4) nsp1 of EAV. dPolyclonal equine sera against EAV generated by the late Dr William McCollum (Balasuriya et al., 2004b). §Negative-control equine serum.

2297 Downloaded from www.microbiologyresearch.org by IP: 178.18.19.222 On: Sat, 17 Oct 2015 18:36:56

256 32 32 64 256 64 512 64 512 32 32 64 512 64 32 32 8 128

1024 256 128 512 1024 256 1024 512 512 128 128 1024 1024 512 128 256 64 512

1024 64 64 128 256 64 128 512 128 64 512 128 1024 16 64 256 8 512

512 64 128 512 1024 256 128 128 256 256 256 256 1024 64 1024 512 16 1024

256 256 8 16 16 64 32 32 32 16 8 16 16 16 32 16 8 16

128 128 32 64 64 16 32 32 32 32 32 16 256 32 128 16 8 64

,8 ,8 ,8 ,8 ,8 ,8 ,8 ,8 ,8 ,8 ,8 ,8 ,8 ,8 ,8 ,8 ,8 ,8


ATCC ARVAC KY84 CW96 S3583 S4216 S3685 S3861 S3854 S4333 S3943 S4445 S3955 S4222 M477 M517 M547 M578

9F2

Control mAbsD

J. Zhang and others

Fig. 5. Aligned deduced amino acid sequences of the critical neutralization regions of the GP5 protein of various EAV strains. The conserved (C1) and variable (V1) regions as well as the critical neutralization (Neut.) sites (A–D) in the GP5 protein are indicated above the sequences. Dots indicate the same amino acid as the aligned sequence at the top. The predicted N-glycosylation sites in the GP5 protein are underlined. The M547 isolate sequence is boxed and several critical amino acid changes are shown in bold. 2298




not both, on the GP3 protein. Interestingly, it was the serine residue of the predicted 4 aa SAFG insertion (resulting from the 12 nt insertion in ORF3) that constituted the third residue of the N-glycosylation site Asn-120 (NXS, Fig. 4a, b). Thirdly, this study identified several amino acid substitution sites in the EAV envelope proteins (GP3: aa 30, 120 and 121; GP5: aa 67 and 73) that were present consistently during persistent infection in the stallion. Interestingly, amino acid substitutions in GP3 (aa 30) and GP5 (aa 67 and 73) were also observed during persistent EAV infection in two long-term carrier stallions that had been characterized previously (Hedges et al., 1999). Moreover, the M547 isolate from an aborted fetus in UT was found to have a distinct neutralization phenotype. This isolate was not neutralized by any of the mAbs that neutralized the other isolates. Similarly, this isolate was neutralized to a significantly lower titre by polyclonal equine sera compared with the other isolates. Sequence analysis suggested that amino acid residues in neutralization sites B, C and D may have contributed to the unique neutralization phenotype of this isolate. However, this needs to be confirmed by further reverse genetic studies. A previous report suggested that the EAV genome remained genetically relatively stable during horizontal and vertical transmission in a disease outbreak on a Warmblood breeding farm in Pennsylvania (Balasuriya et al., 1999). The 2006–2007 EVA occurrence spanned a period of more than 9–10 months of active circulation of EAV in horses of several breeds in multiple states. This event provided a unique opportunity to assess the genomic stability of EAV over an extended period of time, where the principal mode of virus transmission on affected premises was horizontal and vertical, and where several transmission cycles may have occurred. Data from this study suggested that the EAV genome remains genetically relatively stable over a short time period. However, genomic variability tended to increase if horizontal and vertical transmission took place over a more extended time period. Several studies have shown previously that genetic and phenotypic variation of EAV occurs during persistent infection in the stallion or in cell culture (Balasuriya et al., 2004b; Hedges et al., 1999; Zhang et al., 2008c). However, such studies were based on a small number of carrier stallions and mostly on partial genomic sequences of the viruses involved. In this study, we determined the full-length genomic sequences of 18 EAV isolates collected sequentially from eight carrier stallions. It was evident that EAV evolved genetically during persistent infection in the stallion, but the evolution rate of the virus varied depending on the individual stallion and period of viral shedding. The longer the shedding period, the more mutations accumulated in the virus genome. An extensive outbreak of EVA in Kentucky thoroughbreds in 1984 generated widespread interest, publicity and concern regarding this disease (Timoney, 1985). The 2006–2007 multi-state EVA occurrence in the USA (Timoney et al., 2006) and the EVA outbreak in France http://vir.sgmjournals.org

in 2007 (Hamon, 2007; Pitel et al., 2007) have again increased awareness among horse owners and breeders of a disease that can have a significant financial impact on the equine industry. Over the years, importation of carrier stallions and virus-infective semen into the USA and certain European countries has unquestionably been responsible for the introduction of new strains of EAV and precipitation of EVA outbreaks (Balasuriya et al., 1998; Cullinane, 1993; Higgins, 1993; McCollum et al., 1999; Timoney et al., 1998). Ironically, notwithstanding the economic significance of EAV infection to the US$102 billion year21 horse industry in the USA (Owens, 2005), to this date the USA stands alone as the only country with no import testing requirements or controls for EVA. Another feature of the 2006–2007 EVA occurrence is that movement of donor/recipient mares also contributed to virus transmission. The widespread practice of embryo transfer in the Quarter Horse breed and proliferation in the number of recipient mare farms in recent years were significant industry-driven factors not previously recognized as playing a role in the epidemiology of EVA (Timoney et al., 2006). Hopefully, the 2006–2007 multistate EVA occurrence will convince the horse industry and the US animal regulatory authorities of the need to address prevention and control of EVA in a more progressive and determined manner.

METHODS Cells. The rabbit kidney cell line RK-13 (ATCC CCL-37) was

maintained in Eagle’s minimum essential medium with 10 % ferritinsupplemented bovine calf serum, penicillin, streptomycin, amphotericin B and sodium bicarbonate. Clinical specimens. During the 2006–2007 disease occurrence, 19

states that received shipments of fresh cooled semen from the index premises in NM and/or were associated with the movement of donor/ recipient mares submitted 2529 serum samples, 490 semen samples, 143 unclotted blood samples (PBMCs) and 47 aborted fetuses to the World Organization for Animal Health (OIE) EVA reference laboratory at the Gluck Equine Research Center (GERC) for serological and virological testing. One aborted fetus and three semen samples from the 2005 EVA outbreak in NM had been submitted earlier to the GERC. Microneutralization assay. The neutralizing antibody titres of the

test sera against EAV were determined in accordance with an OIEapproved test procedure (OIE, 2008). The neutralization phenotypes of EAV isolates were determined by a microneutralization assay using 14 mAbs and nine polyclonal equine antisera as described previously (Balasuriya et al., 1995a, 1997, 2004b). Virus isolation. Isolation of EAV from semen, sera, PBMCs and fetal tissues was attempted in RK-13 cells according to an OIE-approved protocol (OIE, 2008). The infective tissue culture fluids (TCFs) of the EAV isolates, each in its first passage level (P1) in cell culture, were stored at 280 uC. The identity of each EAV isolate was confirmed by real-time RT-PCR as described previously (Lu et al., 2008). Viral RNA extraction, RT-PCR amplification and sequencing.

Viral nucleic acid was extracted from the infective TCF (P1) containing each EAV isolate using a QIAamp Viral RNA Isolation


2299

J. Zhang and others kit (Qiagen) according to the manufacturer’s instructions. When determining the full-length genome sequence, the entire genome was amplified by RT-PCR in five overlapping fragments using Superscript III (Invitrogen) and the high-fidelity proofreading PfuTurbo DNA polymerase (Stratagene) according to previously described procedures (Zhang et al., 2008a; see Supplementary Table S3, available in JGV Online). When determining the entire ORF2–7 sequences, ORF2–7 and the flanking portion of the EAV genome were RT-PCR-amplified using primers 9570P2 and 12687N (Supplementary Table S3). A similar approach was used to amplify the region (partial ORF1a) encoding the nsp2 protein using primers 833P and 3481N (Supplementary Table S3). The PCR products were gel-purified using a QIAquick Gel Extraction kit (Qiagen). Both sense and antisense strands were sequenced. Sequence analysis. Sequence data were analysed using Aligner

Balasuriya, U. B., Evermann, J. F., Hedges, J. F., McKeirnan, A. J., Mitten, J. Q., Beyer, J. C., McCollum, W. H., Timoney, P. J. & MacLachlan, N. J. (1998). Serologic and molecular characterization of

an abortigenic strain of equine arteritis virus isolated from infective frozen semen and an aborted equine fetus. J Am Vet Med Assoc 213, 1586–1589. Balasuriya, U. B., Hedges, J. F., Nadler, S. A., McCollum, W. H., Timoney, P. J. & MacLachlan, N. J. (1999). Genetic stability of equine

arteritis virus during horizontal and vertical transmission in an outbreak of equine viral arteritis. J Gen Virol 80, 1949–1958. Balasuriya, U. B., Dobbe, J. C., Heidner, H. W., Smalley, V. L., Navarrette, A., Snijder, E. J. & MacLachlan, N. J. (2004a).

Characterization of the neutralization determinants of equine arteritis virus using recombinant chimeric viruses and site-specific mutagenesis of an infectious cDNA clone. Virology 321, 235–246.

v3.0.1 (CodonCode), VectorNTI Advance 11 (Invitrogen) and BioEdit v7.0.0 (http://www.mbio.ncsu.edu/BioEdit/bioedit.html). Potential N-glycosylation sites were predicted using NetNGlyc (http://www.cbs.dtu.dk/services). Multiple sequence alignments were performed using CLUSTAL_X v1.83 (Thompson et al., 1997). Estimates of dN and dS were calculated using the MEGA4.1 program (Tamura et al., 2007). The dN : dS ratios were calculated for each nonstructural and structural protein gene using the method of Li et al. (1985).

Balasuriya, U. B., Hedges, J. F., Smalley, V. L., Navarrette, A., McCollum, W. H., Timoney, P. J., Snijder, E. J. & MacLachlan, N. J. (2004b). Genetic characterization of equine arteritis virus during

Phylogenetic analysis. EAV ORF5, especially the most variable

Cavanagh, D. (1997). Nidovirales: a new order comprising

region within ORF5 (518 nt in length from nt 11296 to 11813; numbered according to the sequence of the VB strain of EAV), has been used widely for phylogenetic studies (Balasuriya et al., 1995b; Hornyak et al., 2005; Larsen et al., 2001; Mittelholzer et al., 2006; Stadejek et al., 1999, 2006; Zhang et al., 2007). The sequence of this region of the 63 EAV isolates analysed in this study together with 287 additional sequences available from GenBank was used for phylogenetic analysis in the current study. The GenBank accession numbers and origins of these EAV isolates have either been reported previously by Zhang et al. (2007) or are provided in Table 1. Unrooted neighbour-joining trees were constructed using MEGA4.1 (Tamura et al., 2007).

ACKNOWLEDGEMENTS

persistent infection of stallions. J Gen Virol 85, 379–390. Balasuriya, U. B., Snijder, E. J., Heidner, H. W., Zhang, J., Zevenhoven-Dobbe, J. C., Boone, J. D., McCollum, W. H., Timoney, P. J. & MacLachlan, N. J. (2007). Development and characterization of

an infectious cDNA clone of the virulent Bucyrus strain of equine arteritis virus. J Gen Virol 88, 918–924. Coronaviridae and Arteriviridae. Arch Virol 142, 629–633. Cullinane, A. A. (1993). Equine arteritis virus in an imported stallion.

Vet Rec 132, 395. Hamon, M. (2007). L’e´pide´mie normande d’arte´rite virale de l’e´te´ 2007 vue par le praticien. Pratique Ve´te´rinaire Equine 39, 29–34 (in French). Hedges, J. F., Balasuriya, U. B., Timoney, P. J., McCollum, W. H. & MacLachlan, N. J. (1999). Genetic divergence with emergence of novel

phenotypic variants of equine arteritis virus during persistent infection of stallions. J Virol 73, 3672–3681. Higgins, A. J. (1993). Equine viral arteritis – a challenge for the British horse industry. Br Vet J 149, 305–306. Hornyak, A., Bakonyi, T., Tekes, G., Szeredi, L. & Rusvai, M. (2005). A

This work was supported by the Grayson-Jockey Club Research Foundation, Lexington, KY, USA, and the Frederick Van Lennep Chair Endowment Fund.

novel subgroup among genotypes of equine arteritis virus: genetic comparison of 40 strains. J Vet Med B Infect Dis Vet Public Health 52, 112–118. Hullinger, P. J., Gardner, I. A., Hietala, S. K., Ferraro, G. L. & MacLachlan, N. J. (2001). Seroprevalence of antibodies against equine

REFERENCES

arteritis virus in horses residing in the United States and imported horses. J Am Vet Med Assoc 219, 946–949.

Balasuriya, U. B. & MacLachlan, N. J. (2004). The immune response

Larsen, L. E., Storgaard, T. & Holm, E. (2001). Phylogenetic

to equine arteritis virus: potential lessons for other arteriviruses. Vet Immunol Immunopathol 102, 107–129. Balasuriya, U. B., Maclachlan, N. J., De Vries, A. A., Rossitto, P. V. & Rottier, P. J. (1995a). Identification of a neutralization site in the

major envelope glycoprotein (GL) of equine arteritis virus. Virology 207, 518–527. Balasuriya, U. B., Timoney, P. J., McCollum, W. H. & MacLachlan, N. J. (1995b). Phylogenetic analysis of open reading frame 5 of field isolates of

equine arteritis virus and identification of conserved and nonconserved regions in the GL envelope glycoprotein. Virology 214, 690–697. Balasuriya, U. B., Patton, J. F., Rossitto, P. V., Timoney, P. J., McCollum, W. H. & MacLachlan, N. J. (1997). Neutralization

determinants of laboratory strains and field isolates of equine arteritis virus: identification of four neutralization sites in the amino-terminal ectodomain of the GL envelope glycoprotein. Virology 232, 114–128. 2300

characterisation of the GL sequences of equine arteritis virus isolated from semen of asymptomatic stallions and fatal cases of equine viral arteritis in Denmark. Vet Microbiol 80, 339–346. Li, W. H., Wu, C. I. & Luo, C. C. (1985). A new method for estimating

synonymous and nonsynonymous rates of nucleotide substitution considering the relative likelihood of nucleotide and codon changes. Mol Biol Evol 2, 150–174. Lu, Z., Branscum, A. J., Shuck, K. M., Zhang, J., Dubovi, E. J., Timoney, P. J. & Balasuriya, U. B. (2008). Comparison of two real-

time reverse transcription polymerase chain reaction assays for the detection of Equine arteritis virus nucleic acid in equine semen and tissue culture fluid. J Vet Diagn Invest 20, 147–155. McCollum, W. H. & Bryans, J. T. (1973). Serological identification of

infection by equine arteritis virus in horses of several countries. In Proceedings of the Third International Conference on Equine Infectious



Characterization of the 2006–2007 EVA outbreak isolates Diseases, Paris, 1972, pp. 256–263. Edited by J. T. Bryans & H. Gerber. Basel, Switzerland: S. Karger.

Thompson, J. D., Gibson, T. J., Plewniak, F., Jeanmougin, F. & Higgins, D. G. (1997). The CLUSTAL_X windows interface: flexible

McCollum, W. H., Timoney, P. J., Lee, J. J. W., Habacker, P. L., Balasuriya, U. B. R. & MacLachlan, N. J. (1999). Features of an

strategies for multiple sequence alignment aided by quality analysis tools. Nucleic Acids Res 25, 4876–4882.

outbreak of equine viral arteritis on a breeding farm associated with abortion and fatal interstitial pneumonia in neonatal foals. In Proceedings of the Eighth International Conference on Equine Infectious Diseases, Dubai, United Arab Emirates, 1998, pp. 559– 560. Edited by U. Wernery, J. F. Wade, J. A. Mumford & O. R. Kaaden. Newmarket, UK: R & W Publications.

Timoney, P. J. (1985). Clinical, virological and epidemiological features of the 1984 outbreak of equine viral arteritis in the thoroughbred population in Kentucky, USA. In Proceedings of the Grayson Foundation International Conference of Thoroughbred Breeders Organizations on Equine Viral Arteritis, Dromoland Castle, Ireland, 1984, pp. 24–33. Lexington, KY: Grayson Found Inc.

McCue, P. M., Hietala, S. K., Spensley, M. S., Stillian, M., Mihalyi, J. & Hughes, J. P. (1991). Prevalence of equine viral arteritis in California

Timoney, P. J. & McCollum, W. H. (1988). Equine viral arteritis:

horses. Calif Vet 45, 24–26.

Timoney, P. J. & McCollum, W. H. (1993). Equine viral arteritis. Vet

Mittelholzer, C., Stadejek, T., Johansson, I., Baule, C., Ciabatti, I., Hannant, D., Paton, D., Autorino, G. L., Nowotny, N. & Belak, S. (2006). Extended phylogeny of equine arteritis virus: division into

Timoney, P. J., McCollum, W. H. & Vickers, M. L. (1997). The carrier

new subgroups. J Vet Med B Infect Dis Vet Public Health 53, 55–58. NAHMS (2000). Equine viral arteritis (EVA) and the U.S. horse industry. http://ageconsearch.umn.edu/bitstream/32749/1/eq98eva1. pdf OIE (2008). Equine viral arteritis. In OIE Manual of Diagnostic Tests

and Vaccines for Terrestrial Animals, 6th edn, pp. 904–918. Paris: World Organisation for Animal Health. Owens, A. (2005). American Horsepower: the Economic Impact of the Horse Industry on the United States. Washington, DC: American Horse Council Publications. Pitel, P. H., Legrand, L., Marcillaud-Pitel, C., Miszczak, F., Guix, E., Pronost, S. & Fortier, G. (2007). L’e´pizootie d’arte´rite virale equine de

epidemiology and control. J Equine Vet Sci 8, 54–59. Clin North Am Equine Pract 9, 295–309. stallion as a reservoir of equine arteritis virus. Equine Disease Quarterly 6, 2. Timoney, P. J., McCollum, W. H. & Vickers, M. L. (1998). The rationale

for greater national control of EVA. Equine Disease Quarterly 7, 2–3. Timoney, P. J., Creekmore, L., Meade, B. J., Fly, D., Rogers, E. & King, B. (2006). Multi-state occurrence of EVA. Report of the

Committee on Infectious Diseases of Horses: addressing equine viral arteritis in the United States. In Proceedings of the 110th Annual Meeting of the United States Animal Health Association, pp. 354–362. Minneapolis, MN: Pat Campbell & Associates and Spectrum Press. van Aken, D., Zevenhoven-Dobbe, J., Gorbalenya, A. E. & Snijder, E. J. (2006). Proteolytic maturation of replicase polyprotein pp1a by

2007: diagnostic et epidemiosurveillance. Pratique Veterinaire Equine 39, 19–27 (in French).

the nsp4 main proteinase is essential for equine arteritis virus replication and includes internal cleavage of nsp7. J Gen Virol 87, 3473–3482.

Powell, D. G. & Timoney, P. J. (2006). EVA outbreak in New Mexico.

Yamaguchi, S., Kanno, T., Akashi, H. & Kondo, T. (1997).

Equine Dis Q 15, 2–3.

Identification of two neutralisation antigenic sites in GL protein of equine arteritis virus isolates by means of monoclonal antibodies. J Equine Sci 8, 7–11.

Snijder, E. J. & Meulenberg, J. J. (1998). The molecular biology of

arteriviruses. J Gen Virol 79, 961–979. Snijder, E. J. & Spaan, W. J. (2006). Arteriviruses. In Fields Virology,

5th edn, pp. 1337–1355. Edited by D. M. Knipe and P. M. Howley. Philadelphia, PA: Lippincott Williams & Wilkins. Snijder, E. J., van Tol, H., Pedersen, K. W., Raamsman, M. J. & de Vries, A. A. (1999). Identification of a novel structural protein of

arteriviruses. J Virol 73, 6335–6345. Stadejek, T., Bjorklund, H., Bascunana, C. R., Ciabatti, I. M., Scicluna, M. T., Amaddeo, D., McCollum, W. H., Autorino, G. L., Timoney, P. J. & other authors (1999). Genetic diversity of equine

arteritis virus. J Gen Virol 80, 691–699. Stadejek, T., Mittelholzer, C., Oleksiewicz, M. B., Paweska, J. & Belak, S. (2006). Highly diverse type of equine arteritis virus (EAV)

from the semen of a South African donkey. Acta Vet Hung 54, 263– 270.

Zhang, J., Miszczak, F., Pronost, S., Fortier, C., Balasuriya, U. B., Zientara, S., Fortier, G. & Timoney, P. J. (2007). Genetic variation and

phylogenetic analysis of 22 French isolates of equine arteritis virus. Arch Virol 152, 1977–1994. Zhang, J., Go, Y. Y., Maclachlan, N. J., Meade, B. J., Timoney, P. J. & Balasuriya, U. B. (2008a). Amino acid substitutions in the structural

or nonstructural proteins of a vaccine strain of equine arteritis virus are associated with its attenuation. Virology 378, 355–362. Zhang, J., Timoney, P. J., Maclachlan, N. J. & Balasuriya, U. B. (2008b). Identification of an additional neutralization determinant of

equine arteritis virus. Virus Res 138, 150–153. Zhang, J., Timoney, P. J., MacLachlan, N. J., McCollum, W. H. & Balasuriya, U. B. (2008c). Persistent equine arteritis virus infection in

HeLa cells. J Virol 82, 8456–8464.

Tamura, K., Dudley, J., Nei, M. & Kumar, S. (2007). Molecular

Ziebuhr, J., Snijder, E. J. & Gorbalenya, A. E. (2000). Virus-encoded

evolutionary genetics analysis (MEGA) software version 4.0. Mol Biol Evol 24, 1596–1599.

proteinases and proteolytic processing in the Nidovirales. J Gen Virol 81, 853–879.



2301