Emerging European-like PRRSV in the U.S. Implications for diagnostic and control strategies

However, sera did not neutralize the North American Type 2 PRRSV prototype isolate VR-2332 and demonstrated intermediate levels ofneutralization against other European-like isolates and

Title: Emerging European-like PRRSV in the U.S.: Implications for diagnostic

and control strategies - NPB# 04-186

Investigator: Eric A Nelson

Institution: South Dakota State University

Co-Investigator: Raymond R.R Rowland (Kansas State University) and Jane

Challenged pigs were monitored daily for clinical signs and blood samples werecollected twice weekly for the first two weeks and weekly thereafter Three pigs fromeach group were euthanized at 14 dpi to evaluate acute lesions and the remaining pigswere retained for 12 weeks to monitor antibody responses, viral loads and persistence.Clinical signs and pathology were variable between groups but generally remained mild.Viremia was monitored using semi-quantitative real-time PCR and viral loads variedsubstantially within and between groups Challenged animals seroconverted, asdetected by the IDEXX ELISA, by approximately 14 dpi Neutralizing antibody

responses against the homologous challenge isolates reached detectable levels asearly as 21 dpi and in some cases, were particularly robust, reaching neutralizationtiters as high as 1:512 However, sera did not neutralize the North American (Type 2) PRRSV prototype isolate (VR-2332) and demonstrated intermediate levels ofneutralization against other European-like isolates and the European prototypic strain,Lelystad virus This study represents the first detailed analysis of the relative virulence,persistence potential and pathogenesis of diverse U.S Type 1 PRRSV isolates It alsoprovides a resource of well-characterized serum samples to serve as reference andcontrol materials for other researchers and diagnostic laboratories.

II Introduction:

Porcine reproductive and respiratory syndrome (PRRS) viruses can be divided intoType 1 or Type 2 isolates Historically, Type 1 isolates, such as the Lelystad virus, wererestricted to the European continent; whereas Type 2 isolates, such as VR-2332, werefound in North America Comparisons at the gene level show only 55-70% nucleotideidentity between Type 1 and Type 2 isolates The first reported European-like PRRSV

in North America appeared in Ontario, Canada in May 1999 (Dewey et al 2000) Thesource was traced to a single group of pigs imported from the Netherlands and theaffected pigs were isolated with no further documented spread of the virus Soon afterthis time, European-like PRRSV isolates were reported by diagnostic laboratories atMinnesota (Rossow et al 2000; Mahlum et al 2000), South Dakota (Ropp et al 2002)and Missouri (Schommer et al 2002) However, unlike the Canadian outbreak, theseviruses could not be traced to a specific European country of origin

European-like (Type 1) PRRSV isolates have now been identified throughout the U.S.and have complicated diagnostic strategies for the detection of PRRSV due to theirsubstantial genetic and antigenic differences from North American (Type 2) PRRSV.Many diagnostic laboratories have modified their PCR assays to also allow the accuratedetection of European-like PRRSV Monoclonal antibody-based reagents are availablefor the detection or confirmation of both PRRSV genotypes by immunohistochemistry(IHC) and virus isolation, although care is required in the selection of appropriatereagents The commercially available IDEXX HerdChek® PRRS ELISA will accuratelydetect an antibody response to both North American and European-like PRRSV, butdoes not differentiate between them However, results obtained using the commonfollow-up indirect fluorescent antibody (IFA) assay for PRRSV serology can beinfluenced by antigenic differences between virus genotypes Virus neutralizationresults are also highly dependant on the virus isolate or genotype used in theneutralization assay Therefore, it is essential for diagnostic laboratories to have access

to well-characterized European-like PRRS virus isolates and control serums to ensureaccurate serodiagnosis of European-like PRRSV These reference materials,developed during the current project, are also important for ongoing research studiesand the development of improved diagnostic assays

Our previous studies suggested that European-like (Type 1) isolates in the U.S were aheterogeneous group, and that these isolates were distinct from Type 1 isolatescurrently circulating on the European continent (Ropp et al 2004; Fang et al 2004).These viruses have also shown substantial differences in clinical presentation in thefield; ranging from completely asymptomatic to quite virulent Elimination strategiesmust take into account the unique genetic, antigenic and biological properties of U.S.Type 1 PRRSV The overall goal of this proposed work was to gain the first insights intothe diagnostic, antigenic and molecular properties of European-like Type 1 PRRSV now

circulating in the U.S These properties will help explain their successful emergenceand evolution in the field, as well as lead to the design of strategies to control theirspread The results of this initial study will lead to tools that will be useful to monitor thespread of Type 1 isolates, rapidly and accurately detect their presence by diagnosticlaboratories and the development of new information for vaccine and control strategies

to prevent their spread

Objective 2 Development of a proficiency panel containing serum samples for the accurate detection of U.S Type 1 PRRSV isolates The IDEXX HerdChek®

PRRSV ELISA is reported to detect antibodies in response to infection by Type 1 andType 2 PRRSV However, many follow-up assays are somewhat strain-specific andhave not been validated with respect to Type 1 field isolates, primarily because we lackknown serum standards Some diagnostic laboratories have already incorporated PCRassays designed to detect European Type 1 isolates but proficiency panels containingknown U.S Type 1 isolates are not readily available A key component of this project is

to develop panels of sequential serum standards obtained from pigs at different stagesafter infection with antigenically distinct U.S Type 1 PRRSV isolates These sampleswill be distributed and made available to diagnostic laboratories to assess sensitivityand specificity of PCR and serological assays This objective will provide producerswith more confidence in reported diagnostic results, reduced risk of transmissionbetween farms, and the validation of new diagnostic assays will be accelerated

Objective 3 Evaluate the virus neutralizing properties of sera from pigs experimentally infected with representative U.S Type 1 isolates Current efforts by

several investigators are directed at developing a typing system for PRRSV Thesesystems can be based on the cross-neutralization properties of mono-specific sera fromexperimentally infected pigs In collaboration with Dr Fernando Osorio and others, thegoal of this objective is to incorporate U.S Type 1 isolates into this classificationsystem A large body of preliminary work performed in our laboratories demonstratesthat the European-like PRRSV isolates in the U.S represent a heterogeneous group

(Fang et al 2004; Ropp et al 2004) Nucleotide heterogeneity would be expected tolead to antigenic variation of the proteins, including potential changes in the virusneutralization determinants in glycoprotein (GP) 5 and other proteins Extensiveevaluation of neutralizing antibody titers against homologous and heterologous PRRSVisolates has shown substantial differences in cross-neutralization titers which maycorrelate with cross-protection between isolates The approach is to prepareneutralizing antisera against four different U.S Type 1 isolates and compare cross-neutralizing activities within the Type 1 group and between Type 1 and North AmericanType 2 isolates Benefits to the producer include; 1) an analysis of the cross-neutralizing capacity among Type 1 viruses (inc Lelystad virus), and between Type 1and Type 2 isolates; and 2) the eventual development of a practical classificationscheme for PRRSV involving cross-neutralization as an indicator of potential cross-protection to aid in the design of future vaccines and controlled exposure strategies.

Objective 4 Evaluate pathogenic properties of U.S Type 1 isolates The

pathogenic potential of European-like PRRSV remains unknown Observation ofinfected herds suggests that Type 1 isolates may exhibit a broad range of pathogenicity.Therefore, the final objective is to perform the first controlled pathogenic studies withthese viruses Benefits include gaining a better understanding of factors such asextended persistence or differences in antibody responses and viral loads that maycomplicate diagnostics Linking virulence with sequence data is critical to new vaccinedesign Also, the identification of possible “naturally attenuated” Type 1 isolates mayprovide a starting point for more traditional modified-live vaccines

IV Materials & Methods:

Genomic sequencing and analysis Nucleotide sequence information for Type 1

viruses was obtained directly from sera or after isolation on porcine pulmonary alveolarmacrophages (PAMs) RNA was extracted from the pellet using a QIAamp viral RNA kit(Qiagen) and cDNA prepared by reverse transcription using Superscript II (GibcoBRL)and random hexamer oligonucleotides according to manufacturer’s recommendations.The forward and reverse oligonucleotides for the PCR amplification of ORF1 weredesigned based on the sequence of Lelystad virus (GenBank accession numberM96262) See Table 1 for a complete listing of PCR primers PCR was performedusing pfu polymerase (Stratagene) in a reaction volume of 100 ul containing Mg-free 1xPCR buffer, 2 mM MgCl2, 200 uM of each dNTP, 1uM of each primer and 2.5 units ofpolymerase The PCR consisted of 1 cycle of 950 C for 1 minute, 30 cycles at 940 C for

30 seconds, 55-600 C for 30 seconds, 680 C for 1 min per kb, then followed by 25 cycles

at 940 C for 20 seconds, 55-600 C for 30 seconds, 680 C for 1 min per kb increasing by

10 seconds per cycle The final extension was at 700C for 20 min Sequencingreactions were prepared using a Big Dye Dideoxy cycle sequencing kit (Perkin-Elmer)and sequencing performed at the Iowa State University DNA Sequencing Facility PCRproducts were sequenced in both directions at least twice Sequences were assembledand analyzed using Gene Tool and Pep Tool software (BioTools) Multiple sequencealignments were performed using Vector NT software (InforMax,), which is based on theClustal W algorithm Maximum-parsimony analysis on the Clustal W sequences wasperformed using PAUP 4.0 (Swofford, 2000) All characters were equally weighted andheuristic search with TBR, stepwise addition, and MULTIPARS options were used in thesearch for most-parsimonious trees A strict consensus tree was generated to evaluatethe differences among equal most parsimonious trees, if more than a single mostparsimonious tree was recovered Bootstrap analyses with 50% majority-rule

consensus from 1000 replicates (PAUP) were conducted to evaluate the support ofindividual clades Neighbor-joining distance analysis was performed on the resultingdistance matrix using MEGA with Lelystad virus as the outgroup The bootstrap optionwas carried out with NJBOOT from 5000 replicates to assess the robustness of interiorbranches of the tree Table 2 shows a complete listing of virus isolates evaluated withstate of origin, date of isolation, and associated GenBank accession numbers.

Table 1 PCR primers used for amplification and sequencing

7100R TTGGGAGGCGGTTTGCAGACCCTCATAATCC 15062R TTTTTTTTTTTTTAATTTCGGTCACATGGTTCCTC 7420R CGCTGTTTCAGTTACAACTAAGC

Table 2 Type 1 PRRSV isolates used for phylogenetic analysis Asterisks identifythose isolates for which full genomes were prepared from this study or are availablefrom GenBank.

Location within

U.S

Isolate Isolation date Accession No.

Sequenced at SDSU

obtained from GenBank)

AY749393 (ORF5)

*9 North Carolina-

Farm 1

of this genome obtained from GenBank, remainder from this study

*10 North Carolina-

Farm 2

AY395074 (ORF5) Complete genome obtained from sequencing performed in this study.

AY395076 (ORF5) Complete genome obtained from sequencing performed in this study.

04-42 02-24-04 Complete genome obtained from sequencing

performed in this study.

*17 North Carolina-

Farm 5

04-43 03-18-04 Complete genome obtained from sequencing

performed in this study.

Animals / challenge groups Forty 5 week-old, PRRSV nạve pigs were obtained from

a certified PRRSV-negative herd and randomly divided among 5 groups housedseparately in isolation facilities in the SDSU Animal Resource Building All animal careand handling procedures were approved and in compliance with Institutional AnimalCare and Use Committee guidelines After a 4-day acclimation period, pigs from eachgroup (n=8) were inoculated intranasally (1 ml/naris) with each U.S Type 1 isolate at1X104 TCID50 The four isolates were chosen based on their genetic heterogeneity,geographic diversity, year of isolation and clinical presentation in the herd Selectedisolates and their phylogenetic relationships are identified in bold in Figure 1 Isolate01-07 was obtained from a 7-10 week old nursery pig in swine herd in Hawaii showing

no morbidity or mortality in this age group Isolate 01-08 was from a group of 8 weekold pigs in North Carolina showing no significant clinical signs Isolate 02-11 wasobtained from a nursery in Nebraska showing 6-30% nursery mortality while 25% of thisage group were visibly ill or unthrifty throughout the nursery and grower phases Isolate03-15, from North Carolina, was associated with a herd experiencing a preweaningmortality of 80-90% for 3 weeks time The subsequent performance in the nursery waspoor and continued through the finisher stage In the adult sow population, there wasonly a mild abortion storm compared to previous U.S PRRSV outbreaks The negativecontrol group consisted of 8 pigs mock-challenged with MARC-145 host-cell culturesupernatant Virus isolates were propagated on MARC-145 cells and used at thelowest passage possible Virus inoculums were also plaque purified to ensurehomogeneity and screened by PCR for adventitious agents, such as PCV-2, prior tochallenge Throughout the study, the clinical condition of animals was assessed, aswell as virus load, antibody responses, virus neutralization activity and gross andmicroscopic pathology Three pigs from each group were euthanized at 14 dpi toevaluate acute lesions and the remaining pigs were retained for 12 weeks to monitorlater term antibody responses, viral loads and persistence, prior to being euthanized at

85 dpi for necropsy evaluation

Sampling and storage Pigs were observed daily for clinical signs and serum samples

were obtained from all pigs twice weekly for the first two weeks and weekly thereafter.The study was terminated at 85 dpi and samples were cataloged by number, dispersedinto appropriate aliquots for analysis and distribution and stored at -80 C

Histopathology At days 14 and 85 dpi, samples of ventral and dorsal lung, liver,

kidney, heart, tonsil, spleen, lymph node, ileum, brain, and nasal turbinate werecollected from pigs following euthanasia and fixed in 10% phosphate buffered formalin.These tissues were then processed and sectioned by standard methods, and stainedwith hematoxylin and eosin for examination

Virus isolation / Serology For the isolation of PRRSV, serum and tissue samples

were evaluated in porcine pulmonary alveolar macrophages and MARC-145 cells aspreviously described (Benfield et al., 1992) Briefly, serum samples were centrifugedand diluted ten-fold in MEM + 2% horse serum; then added to confluent monolayers ofMARC-145 cells in 24 well plates Plates were incubated for 48 hours at 37 C in ahumidified incubator Cells were then washed, fixed in 80% acetone and stained withPRRSV-specific, SDOW-17 FITC-labeled monoclonal antibody Virus titration resultswere reported as TCID50/ml or fluorescent focus units (FFU) /ml Serological responseswere measured using the IDEXX HerdChek® PRRS 2XR ELISA according to themanufacturer’s instructions and results were reported as a sample to positive ratio (S/P)with an S/P value of 0.4 or greater considered to be positive

Figure 1 Phylogenetic placement of Type 1 PRRSV isolates used for animal

challenge Location and year of isolation are in parentheses and isolates selected forchallenge are shown in bold The figure shows the neighbor-joining tree from p-distanceestimation of nucleotide substitutions per site for ORF5 sequence

Group A

Lelystad-like viruses vivvviruses

03-15

(NC-2003 03-16 (NC-2003))

03-17 (NC-2003) 03-01 (MN- 2003) 03-08 (MO-2003) 03-05 (OH-2003) 04-09 (NE -2004) 02-10 (IA-2002) 04-06 (NE- 2004 03-12 (NC-2003)

01-08 (NC-2001)

04-48 (NC-2004) 99-01 (IA-1999)

02-11 (NE-2002)

03-10 (TN-2003)

01-07 (HI-01)

B1 GB1

Lelystad Virus

5A Porcilis ES1 ES5 ES6

PL15 PL6 D2 ES2

PyrSvac 111-92 PL9

.011 025 100

.006 100

.036

.021 100 005 87

.029 055 047

.026 040 024 84

.052 035 046 015

.046 100

100

.016 88

.005 001 003

.013 96 001 008 006

.001 53

82

.002 007 .006

010 002 005

.045

U.S Type 1 Isolates Group B

Remaining European Type 1 Isolates

North American Type 2 Isolates

European Type 1 PRRSV

Virus neutralization Isolate-specific virus neutralization titers were determined using

a fluorescent focus neutralization (FFN) assay as previously described (Wu et al.,2001) The four challenge isolates from this study, as well as selected North Americanreference isolates were evaluated against all serum samples Results were reported asthe greatest serum dilution resulting in a >90% reduction in viral infectivity or FFU.Briefly, each test serum was heat inactivated for 30 minutes at 56 C Serum was diluted

in 2-fold dilution series in MEM + 2% horse serum and an equal volume of selectedvirus suspensions containing 104 FFU/ml were added to 96-well tissue culture plates.The mixture was incubated one hour at 37 C then transferred to a confluent monolayer

of MARC-145 cells in a 96 well format with further incubation at 37 C, in a humidifiedatmosphere of 5% CO2 After 24 hours, the cells were fixed with 80% acetone andstained with SDOW-17 FITC-labeled PRRSV monoclonal antibody for 1 hour at 37 C.Neutralization of the virus was determined by using a 90% reduction of fluorescent foci

as the endpoint

Real-time, quantitative PCR For the detection of viral RNA and determination of viral

load, serum and tissue samples were examined using a real-time, semi-quantitativePCR; Tetracore, Inc VetAlert™ PRRSV assay (Wasilk et al 2004) To determine viral

RNA quantity, a standard curve was generated from an in-vitro transcript dilution

LV occupied a basal position A similar topology was obtained when trees wereconstructed using ORF3 sequence (data not shown)

Figure 2 Phylogenetic analysis of ORF5 showing the location of Group A and Group

B viruses

The proteolytic cleavage sites that produce nsp2 and the role of nsp2 in virus replicationare largely based on information from the related equine arteritis virus (EAV) Cleavagebetween 386-Tyr-Gly by the nsp1-β protein, followed by autocleavage between 1247-Gly-Gly of PP1a and PP1ab yields nsp2, the single largest protein produced duringvirus replication We and others were the first to describe peptide sequencehypervariability in nsp2 of U.S Type 1 viruses, including amino acid insertions anddeletions (Ropp et al 2004; Fang et al 2004) When compared to LV, 18 of the 20Type 1 viruses possessed a 51 nt deletion The potential advantages, if any, thatdeletions and hypervariability offer is unclear One possibility is that hypervariabilitymay eliminate important B and T cell epitopes, which may aid in the persistence of thevirus Two isolates, 02-11 (Nebraska) and 01-07 (Hawaii), lacked the 51 nt deletion butpossessed deletions of 3 and 25 nt (02-11), and a single 3 nt deletion (01-07) Theseresults indicate that the 51 nt deletion cannot be used as a marker for the identification

of all U.S Type 1 viruses The only isolate with an insertion was 03-10, whichpossessed a 24 nt insertion

We took advantage of the 20 U.S Type 1 sequences to better demarcate thehypervariability regions in nsp2 Nsp2 sequences were aligned using ClustalX and thenumber of unique amino acid substitutions counted for every 10 amino acids along thensp2 protein For the purpose of comparison, we also included nsp2 peptidesequences from 16 North American isolates The nsp2 proteins of Type 1 and Type 2viruses are only about 33% identical at the amino acid level The results in Figure 3

04 42- North Carolina-Farm 5

04 43- North Carolina- Farm 5

04 48- North Carolina- Farm 2

01 08- North Carolina -Farm 1

03 12- North Carolina- Farm 2

10 0

2 2

3 5

10 0

6 9 6 9 9

4 6

3

6

8 5

10 0

4 6 3 6

Group B

show that both U.S Type 1 and U.S Type 2 viruses possess at least two hypervariableregions, designated in Figure 3 as HV-1 and HV-2 All insertions and deletions werelocated within the hypervariable domains The conserved and hypervariable domainsare important for the design of Type 1-specific diagnostic tests, as well as assays thatcan be used to identify virus-specific markers.

Figure 3 Hypervariability in nsp2 of U.S Type 1 and Type 2 viruses Solid arrows

show deletions and dotted arrows show insertions The numbers above each arrowindicate the size in amino acids for the insertion/deletion

Phylogenetic analysis of U.S Type 1 nsp2 sequences using neighbor-joining andparsimony analysis showed a similar topology to the ORF5 tree, including branchinginto two major subgroups with the Hawaiian isolate, 01-07, occupying a basal position.Viruses were placed into the same Group A and Group B clades as determined by theORF5 tree Also similar to the ORF5 tree, isolates 02-11, 99-01, and 03-10 remainedloosely associated with the Group A and Group B viruses Isolate 04-40, which wasloosely associated with Group A in the ORF5 tree was relocated into Group B in thensp2 tree (Figure 4)

HV-1 HV-2

US Type 1

US Type 2

Figure 4 Phylogenetic analysis using nsp2 Viruses could be placed into similar

groups according to ORF5 The one important exception is isolate 04-41, which waslocated in ORF5 Group A The significance of this virus in genetic recombination isdiscussed in more detail

One interesting difference between the nsp2 and ORF5 trees was the relocation ofisolate 40-41; from Group A in the ORF5 tree to Group B in the nsp2 tree Arecombination event between a Group A and Group B virus could resolve thisdiscrepancy To look at recombination, we sequenced all ORFs in a subset of eightrepresentative U.S Type 1 viruses We were able to include the genome sequence ofvirus 99-01 or EuroPRRS (Ropp et al 2004) A search for recombination was initiatedusing SimPlot, a computer program originally developed to identify chimeric HIVviruses SimPlot is designed to identify chimeras that result from recombinationbetween known clades or groups of viruses Genomic sequences from isolates 01-08(Group A), 03-15 (Group B), 04-41 (putative A/B chimera) and LV (non-A/B) were

8 9

5 9 3 10 0

10 0

9 8

9 9 8

9 9

10 0 4 5 7 9 9 66 0

01

Group B

Group A