Biological characterization of porcine pegivirus

University of Veterinary Medicine Hannover

Institute of Virology

Department of Infectious Diseases

Biological characterization of porcine pegivirus

THESIS

Submitted in partial fulfilment of the requirements for the degree

DOCTOR OF PHILOSOPHY (PhD)

awarded by the University of Veterinary Medicine Hannover

by

Johanna Kennedy

Koblenz

Hannover, Germany 2020

Supervisor Prof. Dr. Paul Becher

Supervision Group Prof. Dr. Paul Becher

Prof. Dr. Karl-Heinz Waldmann Prof. Dr. Eike Steinmann

1^st Evaluation Prof. Dr. Paul Becher

Institute of Virology, University of Veterinary Medicine Hannover, Germany

Dr. Imke Steffen

Institute for Physiological Chemistry, University of Veterinary Medicine Hannover, Germany

Prof. Dr. Eike Steinmann

Faculty of Medicine, Department of Molecular and Medical Virology, Ruhr-University Bochum, Germany

2^nd Evaluation Prof. Benedikt Kaufer, PhD

Institute of Virology, Free University of Berlin, Germany

Date of final exam 30^th March 2020

Once again… for science, obviously.

Instant classic.

Parts of the thesis have been published previously in:

Research article

Kennedy J, Pfankuche VM, Hoeltig D, Postel A, Keuling O, Ciurkiewicz M, Baumgärtner W, Becher P, Baechlein C. Genetic variability of porcine pegivirus in pigs from Europe and China and insights into tissue tropism. Sci Rep. 2019 Jun 3;9(1):8174.

doi: 10.1038/s41598-019-44642-0

Poster and oral presentations

Kennedy J, Baechlein C, Hoeltig D, Becher P. Presence of porcine pegivirus in domestic pigs and phylogenetic analysis of pegivirus strains from different parts of the world. 10^th Graduate school days, 2017, Bad Salzdetfurth, Germany.

Kennedy J, Baechlein C, Hoeltig D, Becher P. Presence of porcine pegivirus in domestic pigs and phylogenetic analysis of pegivirus strains from different countries in Europe and Asia. 28^th Annual Meeting of the Society for Virology (GfV), 2018, Würzburg, Germany.

Kennedy J, Pfankuche VM, Hoeltig D, Postel A, Keuling O, Ciurkiewicz M, Baumgärtner W, Becher P, Baechlein C. Porcine pegivirus: genetic variability in pigs from Europe and China, insights into tissue tropism and establishment of antibody ELISA. 11^th Graduate school days, 2018, Hannover, Germany.

Kennedy J, Pfankuche VM, Hoeltig D, Postel A, Ciurkiewicz M, Baumgärtner W, Becher P, Baechlein C. Characterization of persistent pegivirus infection: serology, transmission and replication in PBMCs. 29^th Annual Meeting of the Society for Virology (GfV), 2019, Düsseldorf, Germany.

Kennedy J, Pfankuche VM, Hoeltig D, Postel A, Ciurkiewicz M, Baumgärtner W, Becher P, Baechlein C. Insights into porcine pegivirus infection: global distribution, tissue tropism, and transmission. Keystone Symposia Conference on Positive-Strand RNA Viruses, 2019, Killarney, Ireland.

I

Table of contents __________________________________________________________ I List of abbreviations ______________________________________________________ III List of figures ____________________________________________________________ VI List of tables _____________________________________________________________VII

Table of contents

1 Introduction ___________________________________________________________ 1 1.1 Genus Pegivirus ____________________________________________________ 1 1.1.1 Discovery of pegiviruses _________________________________________ 1 1.1.2 Taxonomy _____________________________________________________ 2 1.1.3 Morphology and genome organization ____________________________ 4 1.1.4 Pegivirus protein functions _______________________________________ 5 1.2 Biology of pegivirus infection in pigs and other hosts ____________________ 6 1.2.1 Prevalence and seroprevalence ___________________________________ 6 1.2.2 Transmission __________________________________________________ 10 1.2.3 Persistence ____________________________________________________ 10 1.2.4 Tissue tropism _________________________________________________ 11 1.2.5 Co-infection with other pathogens and clinical relevance ____________ 12 1.3 Aims of the study __________________________________________________ 13 2 Genetic variability of porcine pegivirus in pigs from Europe and China and insights into tissue tropism _________________________________________________ 15

II

3 Dissecting antibody reactivity and possible transmission routes in porcine

pegivirus infection ________________________________________________________ 35 4 Overall Discussion ____________________________________________________ 59 4.1 PPgV RNA detection in domestic pig serum samples from Europe and Asia _

_________________________________________________________________ 59 4.2 Phylogenetic analyses of PPgV ______________________________________ 60 4.3 No detection of PPgV RNA in wild boar ______________________________ 61 4.4 Persistent and transient PPgV infections ______________________________ 62 4.5 Investigation of PPgV tissue tropism _________________________________ 65 4.6 Insights into PPgV transmission routes _______________________________ 66 4.7 Antibody reactivity in Western blot and ELISA ________________________ 67 5 Summary ____________________________________________________________ 71 6 Zusammenfassung ____________________________________________________ 73 7 References ____________________________________________________________ 75

III

List of abbreviations

°C degrees Celcius

× g gravitational acceleration

µg microgram

µl microliter

µm micrometer

aa amino acid

Ab antibody

BPgV bat pegivirus

bp base pair

cDNA complementary deoxyribonucleic acid Cq cycle quantification

CSFV classical swine fever virus C-terminally carboxyl-terminally

E envelope protein

E2t carboxyl-terminally truncated envelope protein 2 E. coli Escherichia coli

ELISA enzyme-linked immunosorbent assay ELISA100 ELISA coated with 100 ng protein per well ELISA²⁵⁰ ELISA coated with 250 ng protein per well EPgV equine pegivirus

FISH fluorescence in situ hybridization FPLC fast protein liquid chromatography

GBV GB virus

h hour

HCV hepatitis C virus HGV hepatitis G virus HHPgV human hepegivirus

HIV human immunodeficiency virus

HPgV human pegivirus

HRP horseradish peroxidase

IgA immunoglobulin A

IV

IgG immunoglobulin G

IgM immunoglobulin M

IMAC immobilized metal ion chromatography IPTG Isopropyl β-d-1-thiogalactopyranoside IRES internal ribosome entry site

kb kilo base

kDa kilo Dalton

LB lysogeny broth

M molar

min minute

ml milliliter

mM millimolar

nm nanometer

no. number

NS non-structural protein

NS3h non-structural protein 3 helicase domain

nt nucleotide

N-terminal amino-terminal

NW New World

OD optical density

OW Old World

ORF open reading frame

PAGE polyacrylamide gel electrophoresis PBS phosphate buffered saline

PBS-Tw phosphate buffered saline containing 0.05% Tween20 PCR polymerase chain reaction

PPgV porcine pegivirus

PVDF polyvinylidene difluoride

px protein x

qRT-PCR quantitative reverse transcription polymerase chain reaction RdRp RNA-dependent RNA polymerase

RNA ribonucleic acid RPgV rodent pegivirus

V rpm revolutions per minute

RT room temperature

RT-PCR reverse transcription polymerase chain reaction SDS sodium dodecyl sulfate

SPgV simian pegivirus

SPgVcpz simian pegivirus (chimpanzee) ssRNA single-stranded RNA

TBS Tris-buffered saline

TDAV Theiler’s disease-associated virus TierSchVersV Tierschutz-Versuchstierverordnung

TM TaqMan

TMB tetramethylbenzidine UTR untranslated region

WB Western blot

VI

List of figures

Chapter 1

Figure 1-1. Phylogenetic relationship of pegivirus species A-K. __________________ 3 Figure 1-2. Predicted genome organization of porcine pegivirus. _________________ 5

Chapter 2

Figure 2-1. Phylogenetic analysis of porcine pegiviruses from different countries and other mammalian pegiviruses. ______________________________________________ 21 Figure 2-2. Fluorescence in situ hybridization of porcine pegivirus (PPgV) positive and negative pigs using a PPgV specific probe; overlay phase contrast and

immunofluorescence; bar = 100 µm. _________________________________________ 24

Chapter 3

Figure 3-1. Coomassie gel of NS3h protein before and after purification by IMAC. _ 45 Figure 3-2. Western blots of purified NS3h protein incubated with serum samples as first antibody (Ab). ________________________________________________________ 46 Figure 3-3. Western blot of crude E2t protein incubated with serum samples that showed NS3h-specific antibody (Ab) reactivity in Western blot and ELISA as first Ab.

_________________________________________________________________________ 48 Figure 3-4. PPgV viral genome quantity in serum (RNA positive results only) during the course of infection in domestic pigs. ______________________________________ 50

VII

List of tables

Chapter 1

Table 1-1. Pegivirus species nomenclature and their respective hosts (Smith et al., 2016). _____________________________________________________________________ 4

Chapter 2

Table 2-1. Porcine pegivirus genome detection rates and viral genome load in serum samples from individual animals and herds from different countries in Europe and Asia. ____________________________________________________________________ 19 Table 2-2. Number of pegivirus positive pigs of different age groups from Europe and China. _______________________________________________________________ 19 Table 2-3. Porcine pegivirus RNA quantities and fluorescence in situ hybridization results in blood and different tissues from two domestic pigs from Germany. _____ 22

Chapter 3

Table 3-1. Characterization of selected serum samples._________________________ 47

1

1 Introduction

1.1 Genus Pegivirus

1.1.1 Discovery of pegiviruses

The genus Pegivirus owes its name to the history of the discovery of its first members in 1995 (Simons et al., 1995b, Simons et al., 1995a, Linnen et al., 1996). The name was first proposed by Stapleton and colleagues in 2011, consisting of two parts: “pe”

represents the characteristic of the viruses in frequently causing persistent infection in their hosts, and “g” acknowledges the previous names “GB virus” and “hepatitis G virus” (Stapleton et al., 2011). The first virus belonging to this group was discovered in 1995 in tamarins and was identified as a primate virus with a flavivirus-like genome closely related to the species Hepatitis C virus (HCV) (Simons et al., 1995b). Following this, similar viruses were found in human sera by two independent working groups who tentatively named them GB virus C (GBV-C) and hepatitis G virus (HGV) (Simons et al., 1995a, Linnen et al., 1996). Though the newly discovered viruses were speculated to be the causative agent of hepatitis for some time, studies failed to show a clear association between virus infection and disease, rendering the name “hepatitis G virus” misleading (Alter et al., 1997, Alter, 1997, Simons et al., 1995a, Mohr and Stapleton, 2009, Theodore and Lemon, 1997). Further pegivirus species were identified in the following years in a variety of mammalian hosts, including further primates, bats, horses and rodents (Epstein et al., 2010, Chandriani et al., 2013, Kapoor et al., 2013a, Quan et al., 2013, Firth et al., 2014, Kapoor et al., 2013b). In 2016, pegivirus sequences were discovered in serum samples of domestic pigs from Germany, and the newly discovered virus species was designated as Porcine pegivirus (PPgV) (Baechlein et al., 2016).

2 1.1.2 Taxonomy

The taxonomy within the Flaviviridae family was recently updated based on phylogenetic relationships and virus characteristics to include newly discovered virus species, which lead to the addition of the genus Pegivirus (Smith et al., 2016, Simmonds et al., 2017, Adams et al., 2013). Highly significant human pathogens belong to this family, including HCV in the Hepacivirus genus, and Yellow fever virus, dengue virus and West Nile virus in the Flavivirus genus. Economically important animal pathogens like bovine viral diarrhea virus, classical swine fever virus (CSFV) and border disease virus of sheep belong to the Pestivirus genus (Simmonds et al., 2017). In contrast to the major pathogens found within the first three genera of the Flaviviridae family, none of the pegivirus species have been clearly associated with causing disease in their hosts (Smith et al., 2016).

The genus Pegivirus currently contains eleven species, Pegivirus A-K, that infect a variety of mammalian hosts, as shown in Figure 1-1 and Table 1-1 (Simons et al., 1995b, Simons et al., 1995a, Linnen et al., 1996, Epstein et al., 2010, Quan et al., 2013, Chandriani et al., 2013, Kapoor et al., 2013a, Kapoor et al., 2013b, Firth et al., 2014, Baechlein et al., 2016). The type species of Pegivirus K is represented by “PPgV_903”, the full-length viral genome sequence of which was isolated from serum of a domestic pig from Germany (Baechlein et al., 2016, Smith et al., 2016).

3

Figure 1-1. Phylogenetic relationship of pegivirus species A-K. The amino acid multiple sequence alignment of the complete coding region of the respective viruses was performed with ClustalW in BioEdit 7.0 (Hall, 1999). The Maximum Likelihood phylogenetic tree was calculated with MEGA X (Kumar et al., 2018) using the Le and Gascuel model (Le and Gascuel, 2008) with frequencies and a gamma distribution of variation with invariant sites. Analysis was performed with 100 bootstrap replicates (Felsenstein, 1985) and numbers along branches represent the percentage bootstrap values. The scale bar indicates substitutions per site. GenBank accession numbers are in parentheses.

4

Table 1-1. Pegivirus species nomenclature and their respective hosts (Smith et al., 2016).

Species Hosts Previous / other names Abbreviation Pegivirus A NW primate

& OW bat GB virus A SPgV

BPgV

Pegivirus B OW bat GB virus D BPgV

Pegivirus C human

& OW primate

GB virus C hepatitis G virus

HPgV SPgVcpz

Pegivirus D horse Theiler’s disease-associated

virus TDAV

Pegivirus E horse equine pegivirus EPgV

Pegivirus F NW bat bat pegivirus BPgV

Pegivirus G OW bat bat pegivirus BPgV

Pegivirus H human human hepegivirus

human pegivirus 2 HHPgV

Pegivirus I NW bat bat pegivirus BPgV

Pegivirus J NW rodent rodent pegivirus RPgV

Pegivirus K pig porcine pegivirus PPgV

NW, New World; OW, Old World

1.1.3 Morphology and genome organization

Pegiviruses are enveloped viruses with a spherical virion of 60-70 nm in size (Xiang et al., 1999). In contrast to other members of the family Flaviviridae, pegiviruses appear to have a truncated core-coding region or absence thereof, though biochemical characterization and electron microscopy of human pegivirus (HPgV) suggest it has a capsid of uncertain origin (Xiang et al., 1998, Xiang et al., 1999).

The positive-sense, single-stranded RNA genome of pegiviruses is non-segmented and ranges in size from 8.9-11.3 kilo bases (kb). It contains a single open reading frame (ORF), which is flanked by 5’ and 3’ untranslated regions (UTRs) (Simmonds et al., 2017). In contrast to the flavivirus 5’ UTR, which contains a type I cap, the 5’ UTRs of hepaciviruses, pestiviruses and pegiviruses possess an internal ribosome entry site (IRES) for translation initiation (Simmonds et al., 2017, Simons et al., 1996). Pegivirus IRES elements are structurally similar to the type I IRES of picornaviruses, or to the

5

type IV IRES elements (Pegivirus H, J and F) seen in hepaci- and pestiviruses, though in both cases sequence identity is limited (Quan et al., 2013, Kapoor et al., 2015). As is common in hepaciviruses, BPgVs belonging to Pegivirus F contain a micro RNA-122 binding site in their 5’ UTR, while such sites are lacking in other pegiviruses (Smith et al., 2016). Comparison of PPgV sequences evidences high amino acid (aa) identities indicative of conserved genome regions within the putative non-structural protein 3 (NS3) and non-structural protein (NS) 5B coding regions, which coincides with findings when comparing HPgV and HCV (Baechlein et al., 2016, Leary et al., 1996b).

1.1.4 Pegivirus protein functions

The ORF of members of the Flaviviridae family is translated into a large polyprotein that is co- and post-translationally cleaved by cellular- and viral proteases (Figure 1-2).

As mentioned above, the origin of the pegivirus capsid has not been determined, as a core-coding region like that seen in other Flaviviridae members appears to be lacking in most pegiviruses, including PPgV (Mohr and Stapleton, 2009, Simons et al., 1996, Baechlein et al., 2016). Most pegivirus protein functions have not been studied in detail and predicted functions are mostly inferred from sequence comparison with homologous proteins within the hepaciviruses, mainly HCV (Simmonds et al., 2017, Mohr and Stapleton, 2009).

Figure 1-2. Predicted genome organization of porcine pegivirus. Schema was newly constructed for this thesis modified from (Baechlein et al., 2016). Amino acid (aa) sizes of the individual predicted mature proteins are indicated below. The predicted cleavage sites are shown by grey (cellular signal peptidases) and black triangles (viral proteases).

6

Like hepaciviruses, pegiviruses have two predicted envelope glycoproteins (E), E1 and E2, which appear to be inserted into the viral envelope in the form of heterodimers (Mohr and Stapleton, 2009). E1 and E2 of PPgV have four N-X-S/T glycosylation sites each, while the number of such sites varies from three to eleven in other pegivirus species (Smith et al., 2016). Protein x (px), the homologue to p7 in HCV, differs in size within the pegivirus species (Mohr and Stapleton, 2009).

Of the predicted six non-structural proteins of pegiviruses, NS2 is thought to be involved in the cleavage of NS2-NS3 as an autoprotease, as it is seen in HCV. The predicted function of HPgV NS3 is that of a viral helicase and of a chymotrypsin-like serine protease, which is responsible for the cleavage of the remaining NS proteins and uses NS4A as a co-factor, a function that is also thought to occur in other pegivirus species (Epstein et al., 2010, Major and Feinstone, 1997, Mohr and Stapleton, 2009, Moradpour et al., 2007, Robertson, 2001, Leary et al., 1996b, Stapleton, 2003). NS5A appears to be an interferone sensitivity-determining region, and the predicted function of NS5B is that of an RNA-dependent RNA polymerase (RdRp) (Leary et al., 1996b, Stapleton, 2003, Linnen et al., 1996, Simons et al., 2000).

1.2 Biology of pegivirus infection in pigs and other hosts

1.2.1 Prevalence and seroprevalence

Porcine pegivirus RNA detection methods and rates

PPgV was first discovered in serum of domestic pigs from Germany using high- throughput sequencing methods. It was subsequently detected in 10 of 455 (2.2%) serum samples from 10 of 37 (27.0%) pig holdings by SYBR-Green-based quantitative reverse transcription polymerase chain reaction (qRT-PCR) targeting NS3 (Baechlein et al., 2016). Following this, Yang and colleagues used conventional RT-PCR, which targets a conserved region in NS5B, to screen 159 porcine serum samples from 15 US

7

states, of which 24 (15.1%, from 10 US states) were PPgV RNA positive, evidencing a much higher detection rate than that previously found in Germany (Yang et al., 2018).

A study from China reported the detection of PPgV genome in 34 of 469 (7.3%) porcine sera using nested RT-PCR targeting NS3, and observed an ascending detection rate from suckling piglets (1.6%) to nursing piglets (1.9%), finishing pigs (6.6%), and sows (11.3%) (Lei et al., 2019).

Human pegivirus RNA and antibody prevalence

HPgV (Pegivirus C) is distributed globally and an estimated 750 million people are viremic, while another 750 million to 2 billion people have evidence of prior HPgV infection. It is thus possibly the most prevalent human RNA virus causing persistent infection, and a major contributor to the human virome (Stapleton et al., 2014, Stapleton, 2003, Chivero and Stapleton, 2015).

Antibodies (Abs) against HPgV are usually detected after clearance of viremia, thus exposure rates are calculated as the sum of RNA positive and Ab positive rates (Tacke et al., 1997, Gutierrez et al., 1997, Thomas et al., 1998). The prevalence of HPgV viremia in healthy blood donors from developed countries is 1-5%, and another 5-20% of individuals have anti-E2 Abs, leading to a total exposure rate between 6 and 25%

(Mohr and Stapleton, 2009, Stapleton et al., 2011, Blair et al., 1998, Gutierrez et al., 1997, Pilot-Matias et al., 1996a, Tacke et al., 1997). Rates are higher in blood donors from developing countries, reaching close to 20% RNA detection rate in some regions (Mohr and Stapleton, 2009, Polgreen et al., 2003). However, prevalence of viremia is significantly higher in high-risk groups, namely individuals with coexistent blood- borne or sexually transmitted infections, and nearly universal exposure is demonstrated in some populations, such as intravenous drug users and human immunodeficiency virus (HIV)-positive men who have sex with men (Alter, 1997, Theodore and Lemon, 1997, Stapleton, 2003, Scallan et al., 1998, Stapleton et al., 2011,

8

Williams et al., 2004, Dawson et al., 1996, Gutierrez et al., 1997, Schlauder et al., 1995, Xiang et al., 2001).

Recently, Pegivirus H, a further pegivirus species infecting humans, was identified (Berg et al., 2015, Kapoor et al., 2015). Prevalence of human hepegivirus (HHPgV, also previously named human pegivirus 2) infection is much lower than that of HPgV, with evidence of exposure to HHPgV (RNA and Ab) detected in 0.45-1.33% of cases (Berg et al., 2015, Coller et al., 2016, Kapoor et al., 2015). In contrast to HPgV, Abs against HHPgV were frequently detected during viremia (Berg et al., 2015, Coller et al., 2016).

Characteristics and detection methods of HPgV-specific antibodies

While various HPgV proteins have been used for Ab detection by expression in Escherichia coli (E. coli), mammalian expressed, C-terminally truncated E2 was identified as a useful antigen for studying HPgV exposure (Dawson et al., 1996, Pilot- Matias et al., 1996b, Pilot-Matias et al., 1996a, Dille et al., 1997).

Anti-HPgV Ab development is usually restricted to conformation-dependent anti-E2 Abs that develop after clearance of viremia, as described above (Gutierrez et al., 1997, Tacke et al., 1997, Tanaka et al., 1998, Thomas et al., 1998). Anti-E2 antibodies are long- lived and provide a certain degree of protection from reinfection, indicating neutralizing activity (Tillmann et al., 1998, Elkayam et al., 1999, Gutierrez et al., 1997).

Additionally, some studies have described the detection of anti-HPgV peptide reactivity during viremia; however, Ab development is restricted to E2 in most cases, suggesting that the E2 antigenic site is immunodominant in humans (McLinden et al., 2006, Pilot-Matias et al., 1996b, Fernandez-Vidal et al., 2007, Gomara et al., 2010, Schwarze-Zander et al., 2006, Tan et al., 1999, Van der Bij et al., 2005, Xiang et al., 1998).

Moreover, HHPgV-specific Abs have been detected using mammalian expressed E2, as well as bacterially expressed NS4A/B and additional peptides located to NS3, NS4A/B and NS5B (Berg et al., 2015, Coller et al., 2016).

9

Pegivirus RNA and antibody detection rates in other mammals

Equine pegivirus (EPgV, Pegivirus E) RNA has been detected in horse serum samples from the United States (9.5%), Brazil (0.8-14.2%), China (1.1%), Germany (13.4%) and England, Scotland and France (3.6%) (Kapoor et al., 2013a, de Souza et al., 2015, Figueiredo et al., 2019, Lu et al., 2018, Lyons et al., 2014, Postel et al., 2016). Theiler’s disease-associated virus (TDAV, Pegivirus D) was first detected in 16 horses from the United States: one was the donor horse of TDAV-containing antitoxin serum and the other 15 horses had been exposed to this serum (Chandriani et al., 2013). Since then TDAV has only been found in horses from Brazil in one study (1.6%), while two other studies failed to find TDAV RNA in 114 and 177 samples from Brazil and China, respectively (Figueiredo et al., 2019, de Souza et al., 2015, Lu et al., 2016).

For detection of anti-EPgV Abs, the NS3 helicase domain (NS3h) was expressed in E.

coli and crude NS3h-containing bacterial lysates were used in indirect ELISA. Of 328 horses from Scotland, England and France, 218 (66.5%) were positive for Ab in ELISA and 88.5% of those were confirmed by Western blot with the same protein. Among these, of the 12 horses with active EPgV infection (3.7%), 10 were also Ab positive (Lyons et al., 2014).

BPgV was initially detected in 5% (5 of 98) of free ranging bats from Bangladesh (Epstein et al., 2010), and further BPgV species were identified in 4% of 1,615 animals belonging to 21 species of New World and Old World bats (Quan et al., 2013). Simian pegiviruses (SPgV) have been found in various Sanguinus, Callithrix, and Aotus species (Bukh and Apgar, 1997, Leary et al., 1996a, Muerhoff et al., 1995, Simons et al., 1995b).

In addition, viruses belonging to Pegivirus C naturally infect chimpanzees (SPgV^cpz), sequences of which form a separate phylogenetic group to HPgVs (see also Figure 1-1) (Adams et al., 1998, Birkenmeyer et al., 1998).

10 1.2.2 Transmission

HPgV has been shown to be transmitted by exposure to infected blood, sexually, and vertically from mother to child (Bhanich Supapol et al., 2009, Hino et al., 1998, Kleinman, 2001, Lin et al., 1998, Ohto et al., 2000, Stapleton, 2003). These transmission routes are largely comparable with those found in HIV, explaining common HIV- HPgV co-infection, and high-risk groups (i.e. intravenous drug users), as mentioned above (Stapleton, 2003). Sexual transmission of HPgV is much more efficient than that of HCV (Lauer and Walker, 2001, Sarrazin et al., 2000, Scallan et al., 1998, Nerurkar et al., 1998, Bourlet et al., 1999, Hollingsworth et al., 1998, Xiang et al., 2001).

TDAV was shown to be transmitted to healthy horses by experimental intravenous inoculation of antitoxin horse serum containing TDAV (Chandriani et al., 2013). EPgV and TDAV were detected in commercially available equine serum pools from various countries, indicating that transmission by products containing equine serum may be possible (Postel et al., 2016).

The transmission of bat pegiviruses has not been studied in detail, however, BPgV genome was detectable in saliva and rectal swabs, but not in urine of bats, indicating that horizontal transmission by shedding in excreta or during fighting, grooming, or sharing of food may be possible (Epstein et al., 2010, Quan et al., 2013). Studies on the transmission of PPgV between pigs have not been reported to date.

1.2.3 Persistence

As RNA viruses, HPgV and HCV are unusual in frequently causing persistent infections in immunocompetent hosts (Chivero and Stapleton, 2015). Though HPgV persistence is not as frequent as that seen in HCV, it is found in roughly 25% of cases, while the other 75% of infections are cleared within two years (Gutierrez et al., 1997, Tanaka et al., 1998). HPgV infection can be long-lived – it was documented for 16 years in one individual (Masuko et al., 1996) – and during persistence the viral load usually

11

remains constant (Lefrere et al., 1999). Many aspects of immune evasion by RNA viruses, including hepaciviruses and pegiviruses, remain poorly understood (Chivero and Stapleton, 2015). Hypervariable regions of the E2 protein, mutation of immunodominant T-cell epitopes, and chronic stimulation of T-cells leading to T-cell exhaustion are mechanisms that have been proposed for HCV immune evasion, but apparently do not apply to HPgV (Keck et al., 2009, Burke and Cox, 2010, Mohr and Stapleton, 2009, Lemon, 2010, Stapleton et al., 2012). Persistence was also observed in HHPgV infection (Kapoor et al., 2015, Berg et al., 2015), as was life-long SPgV infection (Simons et al., 1995b). Persistent PPgV infection was observed in three pigs from Germany, in which viral RNA was detected for 7, 16 and 22 months (Baechlein, 2016), but has not been studied in greater depth.

1.2.4 Tissue tropism

The tissue tropism of most pegivirus species, including PPgV, has not been assessed in detail. Though HPgV was speculated to be a causative agent of hepatitis for some years, clear evidence of a causal association with hepatitis and evidence of replication in liver tissue, such as viral negative-strand RNA as a replication intermediate, were lacking or inconclusive (Chivero and Stapleton, 2015, Fan et al., 1999, Pessoa et al., 1998, Tucker et al., 2000, Berg et al., 1999, Laskus et al., 1997, Laras et al., 1999).

Compared with HCV, a hepatotropic virus in which viral RNA levels are higher in liver than serum, the opposite is the case for HPgV (Pessoa et al., 1998, Chivero and Stapleton, 2015, Manns et al., 2017). Rather, HPgV negative-strand RNA has been detected in bone marrow and spleen of infected individuals (Radkowski et al., 2000, Tucker et al., 2000) and further evidence suggested that HPgV is lymphotropic, which has also been demonstrated to be the case for SPgV (Fogeda et al., 1999, George et al., 2006, Xiang et al., 2000, Kobayashi et al., 1999, Laskus et al., 1998, Stapleton et al., 2011).

HPgV replicates ex vivo in peripheral blood mononuclear cells (PBMCs) isolated from HPgV RNA positive individuals (Fogeda et al., 1999, George et al., 2003, Rydze et al.,

12

2012). PBMCs can also be infected in vitro using serum-derived HPgV (Chivero et al., 2014, Xiang et al., 2000). Viral RNA was detected in natural killer cells, monocytes and diverse subsets of T- and B-lymphocytes, including naïve, central memory and effector memory T-cells, leading to the suggestion that HPgV may infect hematopoietic precursor cells, which maintain infection during differentiation (Chivero and Stapleton, 2015, Chivero et al., 2014). However, the primary permissive cell type(s) for pegiviruses remain unknown (Chivero and Stapleton, 2015).

1.2.5 Co-infection with other pathogens and clinical relevance Co-infection of HPgV and HIV in humans

As mentioned above, due to high similarity in transmission routes, HPgV and HIV co- infection is common (Stapleton, 2003). Several studies have shown beneficial effects on the outcome of HIV-infection attributed to HPgV, including longer survival in co- infected individuals (Heringlake et al., 1998, Tillmann et al., 2001, Nunnari et al., 2003, Toyoda et al., 1998, Williams et al., 2004, Xiang et al., 2001, Zhang et al., 2006, Vahidnia et al., 2012). Such effects caused by HPgV can be explained by various mechanisms, including alterations of cytokine profile, modification of HIV co-receptor expression, direct inhibition of HIV entry through HPgV E2, and modulation of T-cell activation, among others (Schwarze-Zander et al., 2012, Nunnari et al., 2003, Capobianchi et al., 2006, Chang et al., 2007, Haro et al., 2011, Herrera et al., 2009, Herrera et al., 2010, Jung et al., 2007, Koedel et al., 2011, Mohr et al., 2010).

Clinical relevance of pegivirus infection in humans

Several studies have suggested an association between HPgV infection and an increased risk of non-Hodgkin lymphoma (NHL), attributing a possible etiologic role in the development of NHL to chronic immune stimulation or impaired immunosurveillance, to HPgV (Chang et al., 2014, Civardi et al., 1998, Collier et al.,

13

1999, De Renzo et al., 2002, Giannoulis et al., 2004, Kaya et al., 2002, Minton et al., 1998, Zignego et al., 1997, Ellenrieder et al., 1998, Keresztes et al., 2003, Michaelis et al., 2003, Nakamura et al., 1997, Krajden et al., 2010). Furthermore, HPgV RNA has been detected in brain tissue and in cerebrospinal fluid of individuals with encephalitis of unknown cause, although no causal association has been shown (Kriesel et al., 2012, Fridholm et al., 2016, Balcom et al., 2018, Tuddenham et al., 2019).

Disease association in animals

TDAV was suggested as the causative agent of a Theiler’s disease outbreak in horses in the United States, but recent studies indicate that a newly discovered member of the copiparvoviruses, namely equine parvovirus hepatitis, is responsible for acute serum hepatitis in horses (Chandriani et al., 2013, Divers et al., 2018).

To date, infection with PPgV has not been shown to cause any disease in pigs, and viral RNA can be detected in apparently healthy animals (Baechlein et al., 2016). One study detected PPgV in a serum sample from a farm with pigs exhibiting lameness and vesicles in the United States, but porcine parvovirus and astrovirus were also detected. In the same study, nine vesicular swab samples that were additionally tested for PPgV presence were found negative (Yang et al., 2018).

Disease association is similarly unknown for other pegiviruses and they are considered as apathogenic to date. Further research is necessary to better understand clinical implications of pegivirus infection in animal hosts and humans.

1.3 Aims of the study

Since the first description of PPgV in domestic pigs from Germany in 2016, only a handful of studies have been published, which described mainly PPgV genome detection in pigs from the United States and China (Baechlein et al., 2016, Chen et al., 2019, Lei et al., 2019, Li et al., 2019, Yang et al., 2018). Many aspects of pegivirus

14

infection, even in human hosts, remain elusive, not least due to poor in vitro replication and the lack of an animal model in the case of HPgV (Chivero and Stapleton, 2015).

The overall objective of this project was to further biologically characterize porcine pegivirus, specifically including method development to allow insights into RNA and antibody prevalence, tissue tropism and sites of viral replication, host range, transmission routes, and antibody response. In addition to method development, the acquisition of blood, tissue and excretion samples from PPgV-infected pigs was an essential step in the furtherance of this project.

The first aim of the project was the establishment and validation of a TaqMan-based qRT-PCR and the development of an in vitro-transcribed RNA copy standard to allow accurate quantification of PPgV RNA isolated from serum, tissues and excretion samples. The second objective was the establishment of a nested PCR for the amplification of a genome region suitable for sequencing that permitted subsequent phylogenetic analyses. A further goal was the expression of PPgV proteins, namely of the NS3 helicase domain (NS3h), and of truncated E2 (E2t), to evaluate these as possible markers of past or present PPgV infection and, as such, as possible candidates in antibody detection assays.

The development of these methods allows the investigation of various aspects of PPgV biology, including virus distribution and spread, tissue tropism, transmission routes, and infection dynamics, among others. The understanding of such aspects is essential for examining not only how PPgV may influence porcine health, but also in determining whether PPgV infection in the porcine host may be a valuable asset in the study of HPgV infection in humans.

15

2 Genetic variability of porcine pegivirus in pigs from Europe and China and insights into tissue tropism

Kennedy J, Pfankuche VM, Hoeltig D, Postel A, Keuling O, Ciurkiewicz M, Baumgärtner W, Becher P, Baechlein C.

This chapter was published in Scientific Reports journal

Kennedy J, Pfankuche VM, Hoeltig D, Postel A, Keuling O, Ciurkiewicz M, Baumgärtner W, Becher P, Baechlein C. Genetic variability of porcine pegivirus in pigs from Europe and China and insights into tissue tropism. Sci Rep. 2019 Jun 3;9(1):8174.

doi: 10.1038/s41598-019-44642-0

Contribution as first author:

Experimental work: Establishment and optimization of TM qRT-PCR for screening of the presence of PPgV RNA in serum and tissue samples, establishment and optimization of sequencing RT-PCR for phylogenetic analyses, sample preparation, genome amplification, submission for sequencing. Evaluation and scientific presentation of the results: Analyses and graphical presentation of qRT-PCR and sequencing data, performing phylogenetic analyses. Scientific writing: preparation of the manuscript, tables and figure (phylogenetic tree).

16

Abstract

Pegiviruses belong to the family Flaviviridae and have been found in humans and other mammalian species. To date eleven different pegivirus species (Pegivirus A-K) have been described. However, little is known about the tissue tropism and replication of pegiviruses. In 2016, a so far unknown porcine pegivirus (PPgV, Pegivirus K) was described and persistent infection in the host, similar to human pegivirus, was reported. In this study, qRT-PCR, phylogenetic analyses and fluorescence in situ hybridization (FISH) were implemented to detect and quantify PPgV genome content in serum samples from domestic pigs from Europe and Asia, in tissue and peripheral blood mononuclear cell (PBMC) samples and wild boar serum samples from Germany.

PPgV was detectable in 2.7% of investigated domestic pigs from Europe and China (viral genome load 2.4 × 10² to 2.0 × 10⁶ PPgV copies/ml), while all wild boar samples were tested negative. Phylogenetic analyses revealed pairwise nucleotide identities

>90% among PPgVs. Finally, PPgV was detected in liver, thymus and PBMCs by qRT- PCR and FISH, suggesting liver- and lymphotropism. Taken together, this study provides first insights into the tissue tropism of PPgV and shows its distribution and genetic variability in Europe and China.

17

Introduction

Pegiviruses comprise a group of positive-sense, single-stranded RNA viruses, with a genome size of 9–13 kb, that were recently classified into eleven species (Pegivirus A- K) within the genus Pegivirus in the Flaviviridae family¹. They can infect humans as well as a range of mammalian species, including primates, bats, rodents, horses and ^pigs2–9. While pegiviruses are known to cause persistent infections in humans and horses, their pathogenicity remains largely unknown^1,4,10–12. Though a pegivirus was identified in horses with Theiler’s Disease in the USA⁵, recent studies imply that viruses of the copiparvovirus group are associated with serum hepatitis in horses^13,14. Human pegiviruses (HPgV) are distributed globally and viral RNA is present in roughly 750 million people, making it one of the most prevalent human RNA viruses¹⁵.

Though HPgV was initially thought to be hepatotropic and a possible agent of Non- A-E hepatitis, evidence of viral replication in the liver of infected patients is missing or inconclusive^16–18. Rather, as HPgV replication has been shown in peripheral blood mononuclear cells (PBMCs) ex vivo for several weeks, the virus appears to be lymphotropic^19–21. Additionally, HPgV RNA has been found in serum microvesicles, which have successfully delivered viral RNA to uninfected PBMCs that then supported HPgV replication ex vivo²². Interestingly, pegivirus infection in humans may have a beneficial effect on the outcome of human immunodeficiency virus type 1 (HIV-1) infections in individuals co-infected with both viruses, including reduced retroviral loads, slower progression to AIDS and improved survival rates. These benefits are attributed to immune-modulating effects as well as direct and indirect antagonistic mechanisms of HPgV on HIV-1 infection²³.

Porcine pegiviruses (PPgV) were first described in domestic pigs from Germany in 2016⁹. The study reported a PPgV detection rate of 2.2% (10 of 455) in porcine serum samples and described persistent infection for up to 22 months in three pigs that did

18

not display any clinical signs of disease. Apart from Germany, presence of PPgV has been investigated in North America, where a recent study revealed a PPgV detection rate of 15.1% (24 of 159 samples) in the USA²⁴. Additionally, a recent study investigated 469 porcine serum samples from China, 34 (7.25%) of which were found PPgV positive.

Samples originated from different age groups and proved an ascending trend in the PPgV positive rate from suckling piglets (1.61%) and nursing piglets (1.85%) to finishing pigs (6.56%) and sows (11.34%)²⁵.

In this study we analyzed the presence of PPgV genome in pigs from Europe and Asia.

To clarify whether wild boar might play a role in the epidemiology of PPgV, as seen in infections with, for example, classical swine fever virus^26,27, African swine fever virus²⁸ and atypical porcine pestivirus (APPV)²⁹, we also investigated the presence of PPgV genome in wild boar serum samples from Germany. To date the primary permissive cell type(s) of HPgV and other pegiviruses remain unknown. For this reason, we analyzed the tissue and cell tropism of PPgV through detection and quantification of viral RNA in tissues and PBMCs from PPgV positive pigs using qRT- PCR and fluorescence in situ hybridization (FISH).

Results

PPgV RNA in serum samples from Europe and Asia. The in vitro transcribed RNA copy standard evidenced a highly efficient qRT-PCR assay that was able to detect ten viral genome copies per reaction at Cq values around 36. PPgV genome was detectable in 47 of 1,736 (2.7%) serum samples from domestic pigs corresponding to 20 out of 132 herds (15.2%) (Table 2-1). Highest detection rates were found in individual animals from Great Britain (10.3%) and in herds from China (58.3%). In the different age groups investigated here, the PPgV positive rates were 1.9% in animals under 4 weeks of age, 1.2% in fattening pigs over 4 weeks of age, 3.4% in sows and boars, and 10.1% in pigs of unknown age and use (Table 2-2). Viral loads varied between 2.4 × 10² and 2.0 × 10⁶

19

PPgV RNA copies/ml serum, with an overall average of 3.8 × 10⁵ copies/ml. For individual countries on average, lowest genome loads were detected in Poland (1.9 × 10⁴ copies/ml) and highest in Italy (7.1 × 10⁵ copies/ml). All 800 wild boar samples were negative for PPgV RNA.

Table 2-1. Porcine pegivirus genome detection rates and viral genome load in serum samples from individual animals and herds from different countries in Europe and Asia ¹. ¹ PPgV, porcine pegivirus; pos., positive.

Table 2-2. Number of pegivirus positive pigs of different age groups from Europe and China.

20

Phylogenetic analyses. Altogether 31 PPgV partial NS3 sequences were obtained from domestic pigs, of which nine were identical to one or more other sequences. In total, ten sequences from Germany, three sequences from Italy, four sequences from Poland, nine sequences from Great Britain and five sequences from China were acquired.

Sequences GER/SA/13, GER/SA/91, and PL/159 were identical to one additional sequence each, while IT/77, GB/16, and GB/23 were identical to two further sequences each. In Germany, Poland, and Italy, all identical sequences originated from samples from the same farms, while herd affiliation was unknown for samples from Great Britain.

Twenty-two distinct sequences shown here (Figure 2-1) were submitted to GenBank.

They displayed nucleotide sequence identities of >90%. According to phylogenetic analysis, PPgV formed a separate branch in the tree of pegiviruses and viral sequences segregated into two main clusters, one of which contained only sequences from Europe (Germany, Great Britain and Poland). Within the second main cluster, some branches contained sequences recovered from animals in Europe (GER/NDS/T72 and IT/77) in close proximity to variants from China (i.e. CN/6/5) and USA (i.e.

33/ND/2017)²⁴.

Overall, the most closely related pegivirus sequence found in other species when compared to PPgV was bat pegivirus sequence PDB-1715 (GenBank KC796088), which had a nucleotide sequence identity of 58.1% with PPgV/GB/30. A human pegivirus type 2 sequence, ABT0070P.US (GenBank KT427411), had the lowest nucleotide identity (47.1%) compared to PPgV sequences. When comparing PPgV sequences with pegivirus sequences originating from horses, nucleotide identities ranged from 53.7%

to 55.7%. The sequence identities between PPgV and rodent pegivirus were around 54.0%, while the identities with simian pegiviruses ranged from 50.0% to 55.8%.

21

Figure 2-1. Phylogenetic analysis of porcine pegiviruses from different countries and other mammalian pegiviruses. Numbers along branches represent percentage bootstrap values (bootstrap values < 80 % are not given). GenBank accession numbers are in parentheses. Scale bar indicates nucleotide substitutions per site. PPgV sequences are marked with a circle and the circle color indicates the country of origin.

Pegivirus species A-K are indicated on the right.

22

PPgV RNA in tissue samples. In tissue samples of PPgV positive pigs, PPgV RNA was most abundant in the liver (Table 2-3). Liver samples of animals A and B contained 343.9 and 142.5 viral RNA copies/mg tissue, respectively, while 119.3 copies/mg tissue were found in the liver of animal C using qRT-PCR. Serum samples of these animals contained 2,051.1 copies/µl (animal A), 388.6 copies/µl (animal B) and 157.0 copies/µl (animal C). PBMCs were only available from animal A and contained 46 copies/µl whole blood used for isolation (Table 2-3).

Table 2-3. Porcine pegivirus RNA quantities and fluorescence in situ hybridization results in blood and different tissues from two domestic pigs from Germany ¹. ¹ PPgV, porcine pegivirus; FISH, fluorescence in situ hybridization; GE, genome equivalents;

n.d., not determined; boldface indicates positive FISH results; *fresh blood was not available.

23

FISH was used to investigate the liver, thymus, PBMCs and different lymph nodes of animal A, as well as the liver and thymus of animal B, and respective tissues of negative control pigs. PPgV specific signals were detected in the liver of both PPgV positive pigs (Table 2-3; Figure 2-2). Furthermore, several cells of the medullary and cortical region of the thymus of animals A and B were observed to be virus positive using the PPgV specific probe. Additionally, PBMCs of animal A were found to be virus positive in FISH, while lymph nodes, spleen, tonsils, bone marrow and pancreas of animal A tested virus negative. The non-probe incubation as well as the PPgV PCR- negative pigs showed no detectable positive area in the same tested organs, respectively. During necropsy of animal A, multifocal, mild, subendocardial hemorrhages were present. Histopathology showed a mild, portally accentuated, lymphohistiocytic hepatitis, a mild, diffuse infiltration of eosinophils within the thymus, tonsils and lymph nodes and single multinucleated giant cells within the medullary part of the thymus. Furthermore, lymph nodes revealed a mild, diffuse sinus histiocytosis. A moderate, focal, perivascular, lymphoplasmahistiocytic, partially eosinophilic dermatitis was present at the pinna. Additionally, a mild endocardiosis, a mild, lymphohistiocytic epicarditis and a mild to moderate, focal, follicular, lymphocytic conjunctivitis were observed.

24

Figure 2-2. Fluorescence in situ hybridization of porcine pegivirus (PPgV) positive and negative pigs using a PPgV specific probe; overlay phase contrast and immunofluorescence; bar = 100 µm. (A) Single hepatocytes of the liver of a PPgV positive pig showed an intracytoplasmic positive signal for PPgV using a PPgV specific probe, also shown at higher magnification in the insert; arrows: nuclei of hepatocytes surrounded by intracytoplasmic, red, positive signals. (B) The liver of a PPgV negative pig lacked a PPgV specific signal. (C) Within the thymus of a PPgV positive pig, scattered cells showed an intracytoplasmic red positive signal for PPgV, also shown at higher magnification in the insert. (D) Within the thymus of a PPgV negative pig, all cells were negative for PPgV using a PPgV specific probe. (E) Several PBMCs from a PPgV positive pig showed a red positive signal using a PPgV specific probe, also shown at higher magnification in the insert. (F) PBMCs from a PPgV negative pig were negative for PPgV.

25

Discussion

The genus Pegivirus has grown in recent years, as new viruses were identified in different hosts. Yet little is known about their pathogenicity and the impact on the host’s immune response. In this study, our aim was to gain detailed insights into the distribution of PPgV in different parts of the world and the genetic diversity of PPgV.

Viral RNA was detected in serum samples from domestic pigs from various European countries and China, with an overall individual detection rate of 2.7%.

Investigation of three different age groups from Europe and Asia showed a lower PPgV positive rate in younger animals such as piglets (1.9%) and fattening pigs (1.2%) than in adult animals (3.4%). This observation is concordant with the results from a recently published study from China; however, the increase in PPgV positive rate was more prominent there (1.6–11.3%). Focusing on samples from China, we found similar results: 1.0% detection rate in piglets and 9.7% detection rate in sows and boar²⁵. The PPgV positive rates found in this study differ between countries. While no samples were PPgV positive from Switzerland, Serbia and Taiwan, samples from Germany, Poland and Italy have a positive rate similar to the one described previously for German domestic pigs (2.2%)⁹. High detection rates in China (7.8%) and Great Britain (10.3%) found here are nonetheless lower than the positive rate observed in the USA (15.1%) in a previous study²⁴. In humans, HPgV prevalence ranges from 0.5 to 5%

in healthy blood donors from developed countries, but is higher in blood donors from developing countries (5–18.9%), and in individuals co-infected with blood borne or sexually transmitted diseases, like hepatits C virus or HIV-1^30–32. Equine pegivirus (EPgV) has been found in 12 of 328 horses (3.7%) from Europe and 7 of 74 horses (9.5%) from USA^4,12, thus showing similar detection rates as PPgV. The divergence in PPgV detection rates suggests uneven distribution of virus infection and local spread of PPgV. This may be caused by the occurrence of other infectious diseases in pig

26

populations, similar to observations in humans with co-infection, and needs to be studied further. The viral loads determined here (2.4 × 10² to 2 × 10⁶ copies/ml) are similar to EPgV RNA loads described in one study, which ranged from 3.2 × 10⁴ to 3.2

× 10⁶ RNA copies/m^l4. However, another study found higher EPgV viral loads (4.1 × 10⁵ to 2.0 × 10⁹ RNA copies/ml)¹², and the mean RNA load of HPgV in human plasma typically reaches >1 × 10⁷ copies/ml. This may suggest lower replication of PPgV in vivo compared to HPgV and EPgV.

Although HPgV does not appear to be hepatotropic, high amounts of PPgV RNA in the porcine liver shown by qRT-PCR and in situ techniques suggest that viral RNA may accumulate in the liver or even that PPgV has the ability to replicate in hepatocytes. However, this hypothesis will have to be investigated in future studies, as well as whether PPgV infections might be the cause of histopathological changes in the liver, as seen in animal A. Moreover, presence of PPgV RNA in PBMCs and in the thymus supports lymphotropism analogical to HPgV²². Positive FISH results in primary but not secondary lymphoid organs, such as spleen or lymph nodes, imply that the virus might replicate in the thymus and spread to other tissues (e.g. the liver) via PBMCs, but successfully evades recognition by the immune system, which could lead to a persistent infection in the host. Despite significant amounts of viral RNA detected in cells and tissues, highest viral loads were present in the serum of infected animals. With regard to this, low amounts of PPgV RNA in further organs and tissues can most probably be attributed to blood residues. Possible presence of viral RNA in serum microvesicles and associated virus uptake by PBMCs, as seen for HPgV, remain to be determined²².

Phylogenetic analyses showed close genetic relationships among PPgV sequences from different countries, like sequences GER/NDS/T72 and CN/6/5. This could suggest virus spread by international trade with pigs or pig products, such as feed. While all wild boar samples were tested negative for PPgV RNA in this study, other porcine

27

viruses from the family Flaviviridae, such as APPV, have been shown to be present at a higher rate in wild boar (19%) than in domestic pigs from Germany (6.2%)^29,33. For APPV, virus transmission between wild boar and domestic pigs appears likely, as strains originating from wild and domestic animals show genetic distance of as little as 6.6%²⁹. However, due to the comparatively low prevalence of PPgV in Germany, transmission of the virus from domestic pigs to wild boar and vice versa may be limited. Only samples from wild boar hunted in northern Germany entered the present study. Future studies with extended sampling will reveal whether PPgV is also absent in wild boar from other geographical regions. As genome detection alone may result in underestimation of virus dissemination, upcoming investigations of samples from domestic pigs, wild boar and other species will also address serological reactions upon infection with PPgV.

These results manifest that PPgV, like other pegiviruses, is distributed over several continents. It can be hypothesized that putative immune modulatory effects of PPgV infections are implicated in pig health worldwide. Detection of PPgV RNA in lymphoid cells suggests that the virus has the potential to affect the immune system of pigs. First insights into the cell- and organ tropism of PPgV suggest that the virus may be hepatotropic and/or lymphotropic. Future studies will clarify the pathogenic potential and immune modulatory effects of this newly discovered, widely distributed virus.

Methods

1,736 serum samples from domestic pigs from different countries in Europe (Germany, Great Britain, Poland, Switzerland, Italy and Serbia) and Asia (mainland China and Taiwan) originating from 132 different herds were analyzed in this study. For samples collected in Great Britain, the number of herds was unknown. Samples included 108 piglets up to four weeks old, 923 fattening pigs over four weeks old, 557 sows and

28

boar, and 148 pigs of unknown age and use. Samples were taken between 2014 and 2018, originated from apparently healthy domestic pigs and were taken within the framework of national veterinary health management in concordance with national legal and ethical regulations. Residual volumes of these samples were provided for use in the current study, therefore no ethical approval was required for use of these samples. In addition, 800 serum samples from hunted wild boar from Lower Saxony, Germany, were included. 456 of these wild boar samples were collected during the hunting seasons of 2015/2016 and 2016/2017 and were used in a previous study investigating APPV prevalence²⁹. 344 additional wild boar samples were collected during the hunting season of 2017/2018. Furthermore, blood and post-mortem tissue samples originated from apparently healthy PPgV positive pigs (n = 3, animals A, B, and C) from the Clinic for Swine, Small Ruminants, Forensic Medicine and Ambulatory Services (University of Veterinary Medicine, Hannover) and PPgV negative control pigs (n = 2). To rule out presence of co-infections with APPV and porcine reproductive and respiratory syndrome virus (PRRSV), PPgV positive animals were also tested using RT-PCR and found negative for both viruses (data not shown).

One pig (animal A) was submitted to the Department of Pathology, University of Veterinary Medicine Hannover. A full necropsy was performed and samples of 40 different tissues were collected and stored at −80 °C or fixed in 10% neutral buffered formalin and embedded in paraffin wax. For histopathological examination, 3 µm thick sections were stained with hematoxylin and eosin. Different organ and tissue samples and a liver sample originated from two further PPgV positive pigs, animal B and animal C, respectively. Control samples for FISH were taken from PPgV negative pigs. PBMCs from animal A and one negative control animal were isolated from ~1 ml blood by density gradient centrifugation with Histopaque (Merck, Darmstadt, Germany). Euthanasia and sampling were approved by Lower Saxony’s official authorities (LAVES AZ 15A602 and 17A195) and were carried out in accordance with German legislation (TierSchVersV).

29

RNA was isolated from 140 µl of serum using the QIAmp Viral RNA Mini Kit (QIAGEN, Hilden, Germany) according to the manufacturer’s instructions. Isolation of RNA from preweighed tissue samples was achieved using the NucleoSpin RNA kit (Macherey-Nagel, Düren, Germany) or the RNeasy Mini Kit (QIAGEN) according to the manufacturer’s instructions and RNA samples were stored at −80 °C until testing.

For PPgV genome quantification, a TaqMan based qRT-PCR targeting the highly conserved NS3 encoding region with primers PPgV/ fwd/7

(5′-GTCTATGCTGGTCACGGA-3′), PPgV/rev/8

(5′-CACTCATCGCAAATGACCAC-3′) and probe PPgV/pro/11

(5′-[6FAM]-CCATTTCGCGAACCACTGATTCCA-[BHQ1]-3′) was developed and verified using samples that were PPgV positive in a SYBR Green qRT-PCR (QIAGEN) described in an earlier study⁹. For the new PCR assay, an in vitro transcribed RNA copy standard was developed using MEGAscript Kit (ThermoFisher Scientific, Germany) to allow for absolute quantification of genome copies. Real-time RT-PCR was performed using the Mx3005P QPCR System (Agilent Technologies, Santa Clara, USA) and the QuantiTect Probe RT-PCR Kit (QIAGEN) according to the manufacturer’s instructions on samples and RNA standard dilutions. Briefly, 12.5 µl RT-PCR master mix, 0.25 µl reverse transcriptase, 0.8 pmol of each primer and 0.2 pmol of the probe, 5.25 µl water and 5 µl sample RNA were used in each reaction of 25 µl with the following temperature profile: 50 °C for 30 minutes, 95 °C for 15 minutes and 40 cycles of 94 °C for 15 seconds and 60 °C for 1 minute. Serum samples were initially screened in pools containing three to ten individual samples; subsequently samples from positive pools were tested individually.

For phylogenetic analysis, amplicons corresponding to a partial NS3 coding sequence were generated by one of the following two methods: a) RT-PCR with SuperScript III One-Step RT-PCR System with Platinum TaqDNA Polymerase (Life technologies, Germany) with primers PPgV/fwd/G1 (5′-CACCGGGCTGTTTCTGCTA-3′) and PPgV/rev/G4 (5′-TTCCTTCCACACCAACCCAT-3′), or b) cDNA synthesis with

30

SuperScript II Reverse Transcriptase (Invitrogen, Germany) using random hexamers followed by nested PCR with outer primers PPgV/fwd/G1 and PPgV/rev/G4, and inner primers PPgV/fwd/G3 (5′-CGGGCTGTTTCTGCTAGGT-3′) and PPgV/rev/G2 (5′-CACCAACCCATCGAGGATCA-3′) using Taq polymerase included in the Maxima Hot Start Green PCR Master Mix (2X) (ThermoFisher Scientific) and the following cycling parameters: 95 °C for 4 min, 40 cycles of 95 °C for 30 s, 52 °C for 30 s, 72 °C for 75 s, and 72 °C for 10 min. PCR products with an expected length of 1,290 (method a) and 1,278 (method b) were purified using the GeneJET PCR Purification Kit (ThermoFisher Scientific) according to the manufacturer’s instructions and submitted to Sanger sequencing (FlexiRun, LGC Genomics, Germany) with primers PPgV/fwd/G3 and PPgV/rev/G2. Sequences were trimmed to a final length of 1041 base pairs and a multiple sequence alignment was performed with ClustalW implemented in BioEdit 7.0³⁴. Phylogenetic trees were calculated in MEGA7 using the Maximum-likelihood method and the Kimura 2-parameter substitution model³⁵ with 500 replicates for statistical evaluation.

FISH was performed on formalin-fixed, paraffin-embedded organ sections of two qRT- PCR positive pigs (animal A and B) and on the PBMC pellet of one pig (animal A) using a PPgV specific RNA probe covering parts of the PPgV NS3. The probe set (ViewRNA TYPE 1 Probe Set, ThermoFisher Scientific) covered positions 2–816 of a target sequence with 1,172 nucleotides that overlapped with the partial PPgV sequence of animal A (GenBank MH979651). The procedure was carried out according to the manufacturer´s protocol with minor variations as previously described (ViewRNA TYPE 1 Probe Set; ViewRNA™ ISH Tissue Assay Kit (1-plex) and ViewRNA Chromogenic Signal Amplification Kit; ThermoFisher Scientific;)³⁶. Briefly, sections were deparaffinized, boiled in pretreatment solution^® at 90 °C for 10 minutes, digested by a protease QF^® at 40 °C for 10 minutes and afterwards fixed. Hybridization to the specific probe was performed for 6 hours. Following preamplification and amplification steps, sections were stained with Fast Red Substrate and counterstained

31

with Mayer´s hemalum^® (Carl Roth GmbH, Karlsruhe, Germany). Images were acquired with an inverted fluorescence microscope (Olympus IX-70; Olympus Life Science Europe GmbH, Hamburg, Deutschland). The specificity of the probe was confirmed by including a non-probe incubation which served as system negative control and organ sections and cells of PPgV RT-PCR-negative pigs, respectively.

Accession codes. The obtained DNA sequences were deposited in GenBank (accession numbers: MH979651-MH979672).

References

1. Smith, D. B. et al. Proposed update to the taxonomy of the genera Hepacivirus and Pegivirus within the Flaviviridae family. J Gen Virol 97, 2894–2907, https://doi.org/10.1099/jgv.0.000612 (2016).

2. Simons, J. N. et al. Identification of two flavivirus-like genomes in the GB hepatitis agent. Proc Natl Acad Sci USA 92, 3401–3405 (1995).

3. Quan, P. L. et al. Bats are a major natural reservoir for hepaciviruses and pegiviruses. Proc Natl Acad Sci USA 110, 8194–8199, https://doi.org/10.1073/pnas.1303037110 (2013).

4. Kapoor, A. et al. Identification of a pegivirus (GB virus-like virus) that infects horses. J Virol 87, 7185–7190, https://doi.org/10.1128/ JVI.00324-13 (2013).

5. Chandriani, S. et al. Identification of a previously undescribed divergent virus from the Flaviviridae family in an outbreak of equine serum hepatitis. Proc Natl Acad Sci USA 110, E1407–1415, https://doi.org/10.1073/pnas.1219217110 (2013).

6. Epstein, J. H. et al. Identification of GBV-D, a novel GB-like flavivirus from old world frugivorous bats (Pteropus giganteus) in Bangladesh. PLoS Pathog 6, e1000972, https://doi.org/10.1371/journal.ppat.1000972 (2010).

7. Firth, C. et al. Detection of zoonotic pathogens and characterization of novel viruses carried by commensal Rattus norvegicus in New York City. MBio 5, e01933–01914,

https://doi.org/10.1128/mBio.01933-14 (2014).

8. Kapoor, A. et al. Identification of rodent homologs of hepatitis C virus and pegiviruses. MBio 4, e00216–00213, https://doi. org/10.1128/mBio.00216-13 (2013).

9. Baechlein, C. et al. Pegivirus Infection in Domestic Pigs, Germany. Emerg Infect Dis 22, 1312–

1314, https://doi.org/10.3201/ eid2207.160024 (2016).