Preventive Veterinary Medicine 176 (2020) 104905

Contents lists available at ScienceDirect

Preventive Veterinary Medicine

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

Seroprevalence of small ruminant lentivirus (SRLV) infection in wild cervids T

in Poland
Monika Olecha,*, Zbigniew Osińskib, Jacek Kuźmaka
a
Department of Biochemistry, National Veterinary Research Institute, 24-100 Puławy, Poland
b
Department of Hygiene of Animal Feeding Stuffs, National Veterinary Research Institute, 24-100 Puławy, Poland

A R T I C LE I N FO A B S T R A C T

Keywords: Small ruminant lentiviruses (SRLV) are widespread amongst domesticated sheep and goats worldwide. Infection
Small ruminant lentiviruses (SRLV) of wild ruminants in close contact with affected domesticated small ruminants has been proposed as an actor in
Wild ruminants SRLV epidemiology, but studies are limited. The aim of this study was to estimate the apparent (AP) and esti-
ELISA mated prevalence (EP) of exposure to SRLV infection in wild ruminants from Poland. Samples originating from
Cross-species transmission
198 free-living cervids comprising 142 European red deer and 56 roe deer were serologically tested using a
Bayesian approach
multi-epitope recombinant antigen ELISA representing subtypes A1, A13, B1, and B2 of SRLV and a commercial
ELISA test. The estimated prevalence of SRLV infection was estimated using the Bayesian approach with models
that adjusted for the misclassification of animals because of a small population and lack of sampling method, the
imperfect performance of the ELISAs and because sera of different species were tested. The calculated estimated
prevalence ranged from 5.3 % (95 % CI 0.3, 12.5) to 24.6 % (95 % CI 3.3, 38.5) for the ELISA with multi-epitope
antigens while estimated prevalence using the commercial ELISA was 2.5 % (95 % CI 0.2, 6.6). These results may
suggest the existence of a new SRLV reservoir in Poland and highlight the importance of surveilling and con-
trolling SRLV infection in domestic and wild ruminants sharing pasture areas.

1. Introduction de Andrés et al., 2013).

Small ruminant lentiviruses (SRLV) comprise maedi-visna virus
(MVV) and caprine arthritis encephalitis virus (CAEV) initially isolated
from sheep and goats, respectively. SRLV are small enveloped RNA
viruses belonging to the Retroviridae family in the Lentivirus genus,
which show a tropism mainly for monocytes, macrophages, and den-
dritic cells in infected individuals. The main route of transmission is
through ingestion of SRLV-contaminated milk or colostrum or hor-
izontally through respiratory secretions (Blacklaws et al., 2004). These
viruses cause multisystem chronic inflammatory and degenerative dis-
eases clinically characterized by pneumonia, wasting, arthritis, en-
cephalitis, and mastitis. SRLV are widespread throughout the world,
causing damaging economic effects in the small ruminant industry due
to increased mortality and reduced milk production (Martínez-Navalón
et al., 2013).
The genome of SRLV contains structural genes, gag (group-specific
antigen), pol (polymerase) and env (envelope), as well as the accessory
genes tat, vif and rev, which have many regulatory functions. Both
capsid protein (CA) and surface glycoprotein (SU) have diagnostic value
and have been widely used in serological diagnosis (Glaria et al., 2012;
Lentiviruses typically infect hosts in a species-specific manner, and
distinct viral types and subtypes are characteristically associated with
single host species. Since the demonstration of the emergence of human
lentiviruses HIV-1 and HIV-2 from simian lentiviruses following jumps
of the species barrier, it has become evident that lentiviruses have an
ability to effectively infect other species and to adapt to the new hosts
(Apetrei et al., 2004). Other examples of cross-species infection by
lentiviruses with less disastrous consequences than HIV have been re-
ported in the transmission of the feline immunodeficiency virus (FIV)
subtype puma lentivirus (PLVA), where the virus jumped from bobcats
(Lynx rufus) to mountain lions (Puma concolor) (Franklin et al., 2007).
Many studies also reported that MVV and CAEV, originally considered
as species-specific pathogens, are able to cross the species barrier and
infect goats and sheep, respectively (Minardi da Cruz et al., 2013). In
fact, a new classification of SRLV has been proposed in which five major
genetic groups and several subtypes exist. MVV-like and CAEV-like
viruses are classified in groups A and B, respectively, and 18A subtypes
(A1–A18) and 4B subtypes (B1–B4) have been identified to date (Shah
et al., 2004; Olech et al., 2019). Studies conducted on SRLV found in
Poland revealed that goats are infected by strains belonging to subtypes

⁎
Corresponding author at: National Veterinary Research Institute, Department of Biochemistry, Al. Partyzantów 57, 24-100 Puławy, Poland.
E-mail address: monika.olech@piwet.pulawy.pl (M. Olech).

https://doi.org/10.1016/j.prevetmed.2020.104905
Received 19 August 2019; Received in revised form 30 December 2019; Accepted 23 January 2020
0167-5877/ © 2020 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY license
(http://creativecommons.org/licenses/BY/4.0/).
M. Olech, et al.
B1, B2, A1, A12, A16, and A17, while subtypes B2, A1, A12, A13, and
A18 were identified in sheep (Olech et al., 2012, 2018, 2019). The
emergence of new viruses following cross-species infection demon-
strated that these viruses easily adapt to the new host and in con-
sequence can acquire new biological and pathogenic properties
(Deubelbeiss et al., 2014; Fras et al., 2013; Cardinaux et al., 2013).
Undoubtedly, the high ability of SRLV to recombine and its rapid mu-
tation rate during viral replication are the factors that efficiently favour
transgression of the species barrier and their adaptation to the new host
(Ramirez et al., 2011).
While the ability of SRLV to frequently cross the species barrier has
become evident for sheep and goats, the occurrence of natural jumps of
the species barrier to infect wild ruminants has not been proved,
leading to sometimes contradictory results. So far, SRLV infections have
been found in Rocky Mountain goats (Oreamnos americanus) (Patton
et al., 2012), Passirian goats in northern Italy (Gufler et al., 2008), and
recently in red deer (Cervus elaphus) and mouflon (Ovis aries musimon)
but not in fallow deer (Dama dama) during a study conducted with
indigenous free-living ruminants in Spain (Sanjose et al., 2016). Wild
small ruminants, such as Alpine ibexes (Capra ibex) and mouflons (Ovis
aries musimon) were susceptible to experimental and natural infection
with CAEV (Erhouma et al., 2008; Guiguen et al., 2000). However, no
reaction to infection with SRLV was found in wild mouflons in Spain
(López-Olvera et al., 2009), red deer in California (Cutlip et al., 1991;
Chomel et al., 1994) or in several species of wild ruminants in Germany
(Starick et al., 1995).
It can be postulated that contacts between SRLV-infected sheep and
goats and wild ruminants, usually during free grazing or through in-
gestion of milk, may result in cross-species SRLV horizontal transmis-
sion to those wild ruminants and animals thus infected may transmit
the virus to others, including domestic ruminants. Therefore, infected
wildlife and domestic ruminants may represent a pathogen reservoir for
each other (Minardi da Cruz et al., 2013). Furthermore, viruses
adapting to wild ruminants may acquire novel biological properties,
where upon a widening of the spectrum of species tropism and increase
in the virulence of these pathogens could be expected (Morin et al.,
2003; Patton et al., 2012).
The aim of this study was to evaluate serum samples of wild cervids
(red deer and roe deer) for the presence of antibodies to SRLV and
estimate the prevalence of SRLV infection in these ruminants.
2. Material and methods
2.1. Samples from wild ruminants
In total, 198 clotted blood samples were collected from adult wild
cervids including 142 European red deer (Cervus elaphus) and 56 roe
deer (Capreolus capreolus). The samples were collected from 77 male
and 65 female red deer, while roe deer representing only female po-
pulation. The animals came from areas of wilderness which were
mainly forests, and the sites were distributed over 12 Polish voivod-
ships. The samples were collected during the 2010–2014 hunting sea-
sons. All hunters were summoned to take blood samples from the heart
or main vein (aorta or pulmonary artery) as soon as possible after the
shooting the animals. In the majority of cases, there were clotted blood
samples collected from the heart preferably from the ventricles. No
ethical authority approval was required as the samples were collected
post-mortem by licenced hunters. Plasma samples were collected and
stored at −20 °C for serological study.
2.2. Reference sera
50 negative serum samples originating from 36 sheep and 14 goats
and 50 positive serum samples taken from 45 sheep and 5 goats were
used as reference material. The serological status of these sera was
determined by three commercial ELISAs (ELITEST MVV/CAEV (Hyphen
2
Preventive Veterinary Medicine 176 (2020) 104905
Biomed, France), CHEKIT CAEV/MVV Screening (IDEXX, Switzerland),
and LSIVet MAEDI-VISNA/CAEV BLOCKING ELISA (LSI, France)) and
samples were positive or negative when they showed the same positive
or negative result in all three kits. These specimens were used as re-
ference material for evaluation of the diagnostic performance of the
ELISAs developed in this work. Positive and negative control sera for
ELISA tests were prepared by pooling10 different caprine and ovine
sera from animals known to be infected with type A and B viruses.
2.3. Serological study
Plasma samples from 198 wild ruminants and reference sera from
sheep and goats were tested by four home-made indirect ELISAs based
on Gag and Env multi-epitope recombinant antigens representing sub-
types A1, A13, B1 and B2 of SRLV, as previously described (Olech et al.,
2018). These antigens included the Gag domain, containing nearly the
whole matrix (MA) and capsid (CA) proteins, and the SU1 and SU5
antigenic sites. These tests allowed the detection of IgG in plasma of
wild ungulates by replacing the peroxidase-conjugated monoclonal
anti-goat or anti-sheep antibody with the peroxidase-conjugated pro-
tein G. Additionally, samples from wildlife were analyzed by commer-
cial ELISA (ID Screen MVV-CAEV Indirect Screening test, IDvet,
France). The procedure was carried out following the manufacturer’s
protocol with the exception of the second antibody, which was sub-
stituted by peroxidase-conjugated protein G able to react with wild
species IgG. Protein G was used at 1:8000 dilution, where the highest
differences between positive and negative sera were noted. Sera sam-
ples were all tested in duplicate. The manufacturer’s cut-off sample-to-
positive control ratio (S/P) of > 60 % was used.
2.4. Statistical analysis
To estimate the prevalence of SRLV infection in wild cervids we
chose a Bayesian approach that allowed for the incorporation of prior
knowledge by specifying prior distributions for test properties and
prevalence. The apparent (AP) and estimated prevalence (EP) were
estimated according to the overall population of tested wild cervids and
according to species.
To establish cut-off values for wildlife sera tested by ELISAs using
multi-epitope recombinant antigens, the reference sera from sheep and
goats were tested. Due to different reactivities of sheep and goat sera,
the cut-off values and sensitivity (Se) and specificity (Sp) of these
ELISAs were calculated separately for sheep and goat sera and in-
dividually for each recombinant antigen using multiclass receiver op-
erating characteristic (multiclass-ROC) curve analysis based on the
Youden index (Fig. 1). Then the one optimal cut-off value (OD = 0.25)
was estimated with Se 78–100 % and Sp 86–100 %, using multiclass-
ROC analysis in R software (R Foundation for Statistical Computing,
Austria). This function performs multiclass area under curve calcula-
tions as defined by Hand and Till (2001). Sera from wildlife with OD
values equal or higher than the calculated cut-off were regarded as
positive, while those with lower values were regarded as negative.
Using commercial ELISA, the cut-off value recommended by the pro-
ducer was used. These two cut-offs were used to estimate apparent
prevalence rates of wildlife samples together with 95 % confidence
intervals using the Clopper-Pearson exact method (Clopper and
Pearson, 1934).
EP was estimated from apparent prevalence using the Bayesian
approach and Gibbs sampling. For this analysis, two independent
models were used. The first (two populations, one test and no gold
standard) included testing of wildlife and reference sera using ELISAs
with multi-epitope recombinant antigens, representing subtypes A1,
A13, B1, and B2 and the second model (one population, one test, expert
knowledge, no gold standard) included testing of wildlife sera by a
modified commercial ELISA.
An important step in Bayesian analysis is to obtain prior
M. Olech, et al. Preventive Veterinary Medicine 176 (2020) 104905

Fig. 1. Receiver operating characteristic (ROC) curves illustrating sensitivity and specificity values of four ELISAs (corresponding to multi-epitope recombinant
antigens representing subtype A1, A13, B1 and B2 of SRLV and corresponding to goat (panel A) and sheep (panel B) sera.

distributions for all parameters. Therefore, the prior distribution for the inference. For the Se and Sp of the commercial ELISAs, prior informa-
test properties and the prevalence were specified and the prior in- tion was based on expert knowledge (our own unpublished data and
formation for Se and Sp of each test, mode, prevalence and beta prior manufacturer's validation report from the diagnostic kit) and published
distributions, and beta (α, β) parameters were used for Bayesian data (Sanjose et al., 2016; Nowicka et al., 2014). All prior information

3
M. Olech, et al. Preventive Veterinary Medicine 176 (2020) 104905

was transformed into a β distribution model using BetaBuster 1.0 free

(2.3–37.0)

(2.7–52.2)

(3.3–38.5)
software.

21.2

30.1

24.6
A13 The analysis was conducted in a Bayesian approach using WinBUGS
(Lunn et al., 2000). The model was implemented in OpenBugs version
3.2.3 (2014). A total of 25,000 iterations of the Gibbs sampler were
(0.8–23.5)

(1.4–37.9)

(1.0–25.5)
generated in all of the analyses presented in this paper, and the last
16,000 of them were used to estimate the posterior distributions. Esti-
11.6

18.4

13.7
A1

mated prevalence was calculated as the median and 95 % posterior

credibility intervals.
(0.2–11.1)

(0.5–23.5)

(0.3–12.5)

3. Results
4.2

9.0

5.3
Estimated prevalence (%) (95 % CIb)

B2

Plasma of 198 wild ruminants including 142 red deer and 56 roe
deer was tested with multi-epitope ELISAs representative of four sub-
(1.4–31.8)

(1.0–32.6)

(1.4–30.0)

types of SRLV and also tested with a modified commercial ELISA. The
commercial ELISA failed to detect a proportion of animals seropositive
17.8

15.0

17.6
B1

in the multi-epitope ELISAs, since only 10 (5.1 %) out 198 samples were
positive in the former while 43 (21.7 %), 16(8.1 %), 36 (18.2 %) and 57
(28.8 %) samples were positive in the latter, reacting with antigens
(< 0.1–8.4)
(0.6–9.3)

(0.2–6.6)

representing B1, B2, A1 and A13 subtype of SRLV, respectively

IDvet

(Table 1). The proportion of discordant reactions using respective an-

4.0

1.9

2.5

tigens indicate circulation of different genotypes of SRLV in wild ru-

minants. A total of 47 (23.7 %) samples reacted only with one antigen
(20.2–34.6)

(22.9–47.1)

(22.9–35.5)

while 21 (10.6 %) samples reacted with two antigens simultaneously.

Far fewer samples, totalling 9 (4.5 %), reacted with three and again 9
26.8

33.9

28.8
A13

samples (4.5 %) with all four recombinant antigens. Only 3 (1.5 %) out
of 198 samples were positive in all multi-epitope ELISAs and the
commercial ELISA. Out of 10, 4 samples positive in the commercial
(+)

ELISA were positive only in that test while 3 samples reacted ad-
38

19

57

ditionally with one recombinant antigen (B1 or A1).

The overall estimated prevalence ranged from 5.3 % (95 % CI 0.3,
(11.1–23.2)

(14.1–35.8)

(13.4–24.2)

12.5) to 24.6 % (95 % CI 3.3, 38.5) for ELISAs with multi-epitope an-
tigens (Table 1). For roe deer, estimated prevalence ranged from 9.0 %
16.2

23.2

18.2
A1

(95 % CI 0.2, 11.1) to 30.1 % (95 % CI 2.7, 52.2), while for red deer this
values varied from 4.2 % (95 % CI 0.5, 23.5) to 21.2 % (95 % CI 2.3,
37.0). The posterior Bayesian estimate of EP using the commercial
(+)

23

13

36

ELISA was 2.5 % (95 % CI 0.2, 6.6). For roe deer, the EP was 1.9 % (95
% CI < 0.1, 8.4), while for red deer it was 4.0 % (95 % CI 0.6, 9.3). The
(3.4–11.61)

highest prevalence of samples from both species was in the multi-epi-

(6.3–23.7)

(5.1–12.7)

tope ELISA with antigen representing the A13 subtype.

12.5

The four multi-epitope recombinant ELISAs differed from each other

6.3

8.1
B2

in their reactivity patterns in seropositive samples. Distribution of OD

values revealed that the highest reactivity was directed towards the
(+)

antigen representing subtype A1 of SRLV in both roe deer and red deer
16
9

sera (Figs. 2 and 3). The reactivity to remaining antigens was com-
parable among roe and red deer sera, however, higher values were
(16.5–30.1)

(11.4–31.9)

(16.6–28.0)

observed in red deer sera.

22.5

19.6

21.7
B1

4. Discussion
Apparent prevalence (%) (95 % CIa)
Apparent and estimated prevalence of SRLV in wildlife.

Cervids are susceptible to a wide range of pathogens typical for

(+)

32

11

43

other ruminants such as cattle, sheep, and goats and could potentially
CI– Clopper-Pearson confidence interval.

play an important role in spreading these agents (Gortázar et al., 2007;

Daszak et al., 2000). Recovery of SRLV sequences from ibexes (Capra
(3.4–1.6)

(0.4–9.4)

(2.8–9.1)

ibex) that were closely related to CAEV proviral sequences present in

IDvet

CI– Bayesian credibility interval.

6.3

1.8

5.1

local domestic goats, support the hypothesis of cross-species adaptation

(+)- number of positive samples.

of SRLV to new host under natural conditions (Erhouma et al., 2008).

N–number of tested animals.

In Poland, there are no farms where wild ruminants are kept to-
(+)

10
9

gether with sheep and goats, therefore, it has been suggested that
common pasture or breeding areas could be a risk factor for promoting
142

198

transmission of SRLV from infected sheep and goats to wild cervids

56
N

(spill-over) during the grazing season in wilderness areas. A recent

study showed that the infections with SRLV in Poland are quite
Red deer

Roe deer

common and the overall true prevalence in sheep was 9.3 % and 33.3 %
Species
Table 1

Total

at individual animal and flock levels, respectively (Olech et al., 2017).

b
a

In goats, the seroprevalence was 71.9 % at flock level (Kaba et al.,

4
M. Olech, et al. Preventive Veterinary Medicine 176 (2020) 104905

Fig. 2. Optical density (OD 405 nm) distribution of positive and negative roe deer sera obtained by multi-epitope ELISAs with antigens representing A1, A13, B1 and
B2 subtypes. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article).

2013). Interestingly, the highest seroprevalence in sheep was noted in
the Małopolskie and Podkarpackie voivodships, which are regions lo-
cated in the south of Poland in the Carpathian Mountains. These regions
record the highest population of sheep, accounting for close to 30 % of
the total population, and are places where traditional grazing on
mountain pastures is still practised; there is a consequent high prob-
ability of co-grazing with wild cervids. These facts substantiate the risk
of transmission of SRLV from domestic sheep and goats to wild cervids.
Diagnostic testing of wildlife animals is more challenging than do-
mestic ones for a variety of reasons, including difficulties in animal and
sample accessibility, poor sample quality and poor knowledge of pa-
thogenesis/epidemiology of the disease in wildlife. In some already
published studies with wildlife specimens, commercial diagnostic kits
dedicated to domestic animals were used, however, these kits have not
been validated for wildlife. The question remains as to whether there
are any essential differences in diagnostic sensitivity or specificity of
these tests when they are applied to wildlife samples. The availability of
reference material (positive and negative controls) is also a common
problem when samples from wildlife animals are investigated.
Therefore, in the presented study our main effort was directed to esti-
mate diagnostic performances of newly developed ELISA tests and
modified commercially available ELISA for testing wildlife specimen,
using as reference sera from closely related species, goats and sheep.
The serological status of these sera was determined by three commer-
cial ELISAs. The reason to use the sera from both species was the dif-
ferent reactivity of these sera to SRLV antigens as was described pre-
viously by Grego et al., 2002 and Rachid et al., 2013. Thus, the cut-off
values and sensitivity and specificity of ELISAs were calculated sepa-
rately for sheep and goat sera and individually for each recombinant
antigen and then the one optimal cut-off value was estimated. Detection

Fig. 3. Optical density (OD 405 nm) distribution of positive and negative red deer sera obtained by multi-epitope ELISAs with antigens representing A1, A13, B1 and
B2 subtypes. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article).

5
M. Olech, et al.
of the immune response often requires species-specific reagents/test
components which complicates this diagnostic approach in wildlife. In
our study, the limitation entailed by the lack of antisera to IgG specific
for roe and red deer has been overcome, by the use of protein G bound
to enzyme horseradish peroxidase.
In this paper, we showed reactivity to different SRLV antigen pre-
parations in sera of red and roe deer. Sera were tested with ELISAs
based on Gag and Env multi-epitope recombinant antigens representing
subtypes A1, A13, B1, and B2 of SRLV and an adapted commercial
ELISA, and then obtained data on apparent prevalence were in-
corporated in statistical models, using the Bayesian approach. This
model was used because it adjusts for the misclassification of cervid
sera because of factors such as small number of population and lack of
sampling method, as well as the imperfect accuracy of the ELISAs and
their different Se and Sp, use of recombinant antigens representing
different genetic subtypes of SRLV, lack of diagnostic tools considered a
gold standard to correctly define the infectious status of wild animals
exposed to SRLV, and because sera of different species were tested
(Manual of Diagnostic Tests and Vaccines for Terrestrial Animals,
Chapter 3.6.7., 2014). An additional argument for the use the Bayesian
approach was that this analysis has not yet been applied for estimation
of the prevalence of SRLV in wild ruminants, although it has been used
extensively in epidemiologic studies of domestic animals (Lombard
et al., 2013; Verdugo et al., 2015; Mamisashvili et al., 2013).
Serological reactivity mainly detected by multi-epitope ELISAs re-
sulted in estimated prevalence varying from 5.3%–24.6% according to
the particular ELISA, and it was somewhat lower than the respective
apparent prevalence values. This suggests a tendency to overestimate
apparent prevalence data, which may be associated with false-positive
results. Reactivity investigated by multi-epitope ELISAs was also higher
than that noted in the commercial ELISA, which can be explained by the
addition of novel epitopes representing different genetic subtypes of
SRLV. It has been demonstrated that gag-encoded capsid protein (CA)
conveys immunological relatedness between Gag epitopes of different
subtypes, especially among subtypes for a given group, while env-en-
coded surface glycoprotein (SU) generates a type-specific immune re-
sponse (Carroza et al., 2009; Olech et al., 2012, 2019). Therefore, the
discrepancies between multi-epitope ELISAs and the commercial ELISA
may result from the inclusion of novel epitopes based on Gag and Env
multi-epitope recombinant antigens, providing wider cross-reactivity.
Additionally, this study revealed differences in the reactivity of sera to
antigens representing different subtypes of SRLV. The share of SRLV in
toto represented by a circulating subtype is indicated by its percentage
of positive results with the appropriate recombinant antigen utilized in
the multi-epitope ELISAs. The reaction was mainly directed against
antigens belonging to group A of SRLV, especially against subtype A13,
of which circulation has been confirmed only in sheep from Poland
(Olech et al., 2012, 2019). These results additionally confirm the broad
dissemination of subtype A13 in Poland, which was found to be pre-
dominant in sheep from the Malopolska region (Olech et al., 2019).
It is interesting that although the highest estimated prevalence was
noted in sera which reacted to antigen A13, the highest OD values were
noted in serum samples reacted to A1 antigen. This reactivity was noted
in both species, however, sera obtained from roe deer were more re-
active than specimen from red deer. This may be due to the fact that
immunodominants located on GAG and ENV proteins of A1 subtype of
SRLV induced a stronger immune response than immunodominants of
other subtypes, indicating that genotype A-derived antigens seem to be
more suitable for detection of cross-reacting antibodies than genotype B
(Grego et al., 2005; Carrozza et al., 2009; Tolari et al., 2013). Differ-
ences noted for reactivity of roe deer and deer sera may be due to
species-specific differences as was observed for reactivity of sheep and
goats (Rachid et al., 2013). Using multi-epitope antigens representing
subtypes B2, A1, and A13, reactivity was higher in roe deer than in red
deer while with a B1-derived antigen, higher reactivity was noted in red
deer. This may indicate species-specific patterns of antibody reactivity
6
Preventive Veterinary Medicine 176 (2020) 104905
to the immunodominant Gag and SU proteins of SRLV, which probably
reflects different evolution of SRLV in a new host with distinct hos-
t–virus interactions in different species of wild cervids. Probably the
extent of the interaction between the SRLV and the wildlife host is not
limited to a humoral response only but includes other factors, as was
demonstrated by Sanjose et al. (2016) who showed different adaptation
patterns of SRLV in vitro to red deer and fallow deer skin fibroblasts.
In conclusion, this study provides the first data on the prevalence of
SRLV in wild cervids in Poland. Our results strongly indicate that the
prevalence of SRLV in red deer and roe deer may be very high and that
these animals may play an important role in the epidemiology of SRLV
infections and circulation of these viruses between susceptible animals.
Use of multi-epitope ELISA tests with recombinant antigens based on
different SRLV subtypes improved detection of infection, as compared
to the use of a commercially available ELISA. However, further studies
are needed to know whether adaptation of SRLV to free-living hosts
may be associated with the acquisition of new genetic and phenotypic
properties by these viruses and to elucidate the role of intrinsic factors
in the restriction of SRLV in wild animals.
Funding
The work was supported by the National Science Center (NCN)
project no.2016/21/D/NZ7/00626.
Declaration of Competing Interest
The authors declare no conflicts of interests.
Ethical approval
None required.
References
Apetrei, C., Robertson, D.L., Marx, P.A., 2004. The history of SIVs and AIDS: epide-
miology, phylogeny and biology of isolates from naturally SIV infected non-human
primates (NHP) in Africa. Front. Biosci. 1225–1254.
Blacklaws, B.A., Berriatua, E., Torsteinsdottir, S., Watt, N.J., de Andres, D., Klein, D.,
Harkiss, G.D., 2004. Transmission of small ruminant lentiviruses. Vet. Microbiol. 10,
199–208.
Cardinaux, L., Zahno, M.L., Deubelbeiss, M., Zanoni, R., Vogt, H.R., Bertoni, G., 2013.
Virological and phylogenetic characterization of attenuated small ruminant lentivirus
isolates eluding efficient serological detection. Vet. Microbiol. 162, 572–581.
Carrozza, M.L., Mazzei, M., Lacerenza, D., Del Chiaro, L., Giammarioli, M., Marini, C.,
Rutili, D., Rosati, S., Tolari, F., 2009. Seroconversion against SU5 derived synthetic
peptides in sheep experimentally infected with different SRLV genotypes. Vet.
Microbiol. 137, 369–374.
Chomel, B.B., Carniciu, M.L., Kasten, R.W., Castelli, P.M., Work, T.M., Jessup, D.A., 1994.
Antibody prevalence of eight ruminant infectious diseases in California mule and
black-tailed deer (Odocoileus hemionus). J. Wildl. Dis. 30, 51–59.
Clopper, C., Pearson, E.S., 1934. The use of confidence or fiducial limits illustrated in the case
of the binomial. Biometrika 26, 404–413.
Cutlip, R.C., Lehmkuhl, H.D., Brogden, K.A., Schmerr, M.J., 1991. Seroprevalence of
ovine progressive pneumonia virus in various domestic and wild animal species, and
species susceptibility to the virus. Am. J. Vet. Res. 52, 189–191.
Daszak, K., Cunningham, A.A., Hyatt, A.D., 2000. Emerging infectious diseases of wild-
life-Threats to biodiversity and human health. Science 287, 443–449.
de Andrés, X., Ramírez, H., Bertolotti, L., San Román, B., Glaria, I., Crespo, H., Jáuregui,
P., Minguijón, E., Juste, R., Leginagoikoa, I., Pérez, M., Luján, L., Badiola, J.J.,
Polledo, L., García-Marín, J.F., Riezu, J.I., Borrás-Cuesta, F., de Andrés, D., Rosati, S.,
Reina, R., Amorena, B., 2013. An insight into a combination of ELISA strategies to
diagnose small ruminant lentivirus infections. Vet. Immunol. Immunopathol. 152,
277–288.
Deubelbeiss, M., Blatti-Cardinaux, L., Zahno, M.L., Zanoni, R., Vogt, H.R., Posthaus, H.,
Bertoni, G., 2014. Characterization of small ruminant lentivirus A4 subtype isolates
and assessment of their pathogenic potential in naturally infected goats. J. Virol. 11,
65 doi:10.1186/1743-422X-11-65.
Erhouma, E., Guiguen, F., Chebloune, Y., Gauthier, D., Lakhal, L.M., Greenland, T.,
Mornex, J.F., Leroux, C., Alogninouwa, T., 2008. Small ruminant lentivirus proviral
sequences from wild ibexes in contact with domestic goats. J. Gen. Virol. 89,
1478–1484.
Franklin, S.P., Troyer, J.L., Terwee, J.A., Lyren, L.M., Boyce, W.M., Riley, S.P., Roelke,
M.E., Crooks, K.R., Vandewoude, S., 2007. Frequent transmission of im-
munodeficiency viruses among bobcats and pumas. J. Virol. 81, 10961–10969.
M. Olech, et al.
Fras, M., Leboeuf, A., Labrie, F.M., Laurin, M.A., Singh Sohal, J., L’Homme, Y., 2013.
Phylogenetic analysis of small ruminant lentiviruses in mixed flocks: multiple evi-
dence of dual infection and natural transmission of types A2 and B1 between sheep
and goats. Infect. Genet. Evol. 19, 97–104.
Glaria, I., Reina, R., Ramírez, H., de Andrés, X., Crespo, H., Jauregui, P., Salazar, E.,
Luján, L., Pérez, M.M., Benavides, J., Pérez, V., Polledo, L., García-Marín, J.F., Riezu,
J.I., Borrás, F., Amorena, B., de Andrés, D., 2012. Visna/Maedi virus genetic char-
acterization and serological diagnosis of infection in sheep from a neurological out-
break. Vet. Microbiol. 155, 137–146.
Gortázar, C., Ferroglio, E., Höfe, U., Frölich, K., Vincente, J., 2007. Disease shared be-
tween wildlife and livestock: a european perspective. Eur. J. Wild. Res. 53, 241–256.
Grego, E., Profiti, M., Giammarioli, M., Giannino, L., Rutili, D., Woodall, C., Rosati, S.,
2002. Genetic heterogeneity of small ruminant lentiviruses involves im-
munodominant epitope of capsid antigen and affects the sensitivity of single-strain-
based immunoassay. Clin. Diagn. Lab. Immunol. 9, 828–832.
Grego, E., Bertolotti, L., Carrozza, M.L., Profiti, M., Mazzei, M., Tolari, F., Rosati, S., 2005.
Genetic and antigenic characterization of the matrix protein of two genetically dis-
tinct ovine lentiviruses. Vet. Microbiol. 106, 179–185.
Gufler, H., Moroni, P., Casu, S., Pisoni, G., 2008. Seroprevalence, clinical incidence, and
molecular and epidemiological characterisation of small ruminant lentivirus in the
indigenous Passirian goat in northern Italy. Arch. Virol. 153, 1581–1585.
Guiguen, F., Mselli-Lakhal, L., Durand, J., Du, J., Favier, C., Fornazero, C., Grezel, D.,
Balleydier, S., Hausmann, E., Chebloune, Y., 2000. Experimental infection of
Mouflon-domestic sheep hybrids with caprine arthritis-encephalitis virus. Am. J. Vet.
Res. 61, 456–461.
Hand, D.J., Till, R.J., 2001. A simple generalisation of the area under the ROC curve for
multiple class classification problems. Mach. Learn. 45, 171–186.
Kaba, J., Czopowicz, M., Ganter, M., Nowicki, M., Witkowski, L., Nowicka, D., Szaluś-
Jordanow, O., 2013. Risk factors associated with seropositivity to small ruminant
lentiviruses in goat herds. Res. Vet. Sci. 94, 225–227.
Lombard, J., Gardner, I., Jafarzadeh, S., Fossler, C., Harris, B., Capsel, R., Wagner, B.,
Johnson, W., 2013. Herd-level prevalence of Mycobacterium avium subsp. para-
tuberculosis infection in United States dairy herds in 2007. Prev. Vet. Med. 108,
234–238.
López-Olvera, J.R., Vidal, D., Vicente, J., Pérez, M., Luján, L., Gortázar, C., 2009.
Serological survey of selected infectious diseases in mouflon (Ovis aries musimon)
from south-central Spain. Eur. J. Wildl. Res. 55, 75–79.
Lunn, D.J., Thomas, A., Best, N., Spiegelhalter, D., 2000. WinBUGS—a Bayesian model-
ling framework: concepts, structure, and extensibility. Statist. Comput. 10, 325–337.
Mamisashvili, E., Kracalik, I.T., Onashvili, T., Kerdzevadze, L., Goginashvili, K., Tigilauri,
T., Donduashvili, M., Nikolaishvili, M., Beradze, I., Zakareishvili, M., Kokhreidze, M.,
Gelashvili, M., Vepkhvadze, N., Rácz, S.E., Elzer, P.H., Nikolich, M.P., Blackburn,
J.K., 2013. Seroprevalence of brucellosis in livestock within three endemic regions of
the country of Georgia. Prev. Vet. Med. 110, 554–557.
Martínez-Navalón, B., Peris, C., Gómez, E.A., Peris, B., Roche, M.L., Caballero, C., Goyena,
E., Berriatua, E., 2013. Quantitative estimation of the impacts of caprine arthritis
encephalitis virus infection on milk production by dairy goats. J. Vet. 197, 311–317.
Minardi da Cruz, J.C., Singh, D.K., Lamara, A., Chebloune, Y., 2013. Small ruminant
lentiviruses (SRLVs) break the species barrier to acquire new host range. Viruses 5,
Preventive Veterinary Medicine 176 (2020) 104905
1867–1884.
Morin, T., Guiguen, F., Bouzar, B.A., Villet, S., Greenland, T., Grezel, D., Gounel, F.,
Gallay, K., Garnier, C., Durand, J., Alogninouwa, T., Mselli-Lakhal, L., Mornex, J.F.,
Chebloune, Y., 2003. Clearance of a productive lentivirus infection in calves ex-
perimentally inoculated with caprine arthritis-encephalitis virus. J. Virol. 77,
6430–6437.
Nowicka, D., Czopowicz, M., Mickiewicz, M., Szaluś-Jordanow, O., Witkowski, L.,
Bagnicka, E., Kaba, J., 2014. Diagnostic performance of ID Screen® MVV-CAEV
Indirect Screening ELISA in identifying small ruminant lentiviruses-infected goats.
Pol. J. Vet. Sci. 17, 501–506.
Olech, M., Rachid, A., Croise, B., Kuźmak, J., Valas, S., 2012. Genetic and antigenic
characterization of small ruminant lentiviruses circulating in Poland. Virus Res. 163,
528–536.
Olech, M., Osiński, Z., Kuźmak, J., 2017. Bayesian estimation of seroprevalence of small
ruminant lentiviruses in sheep from Poland. Prev. Vet. Med. 147, 66–78.
Olech, M., Valas, S., Kuźmak, J., 2018. Epidemiological survey in single-species flocks
from Poland reveals expanded genetic and antigenic diversity of small ruminant
lentiviruses. PLoS One 13, e193892. https://doi.org/10.1371/journal.pone.0193892.
Olech, M., Murawski, M., Kuźmak, J., 2019. Molecular analysis of small-ruminant lenti-
viruses in Polish flocks reveals the existence of a novel subtype in sheep. Arch. Virol.
164, 1193–1198.
Patton, K.M., Bildfell, R.J., Anderson, M.L., Cebra, C.K., Valentine, B.A., 2012. Fatal
caprine arthritis encephalitis virus-like infection in 4 Rocky Mountain goats
(Oreamnos americanus). J. Vet. Diagn. Invest. 24, 39239–39246.
Rachid, A., Croise, B., Russo, P., Vignoni, M., Lacerenza, D., Rosati, S., Kuźmak, J., Valas,
S., 2013. Diverse host-virus interactions following caprine arthritis-encephalitis virus
infection in sheep and goats. J. Gen. Virol. 94, 634–642.
Ramirez, H., Glaria, I., de Andrés, X., Martínez, H.A., Hernández, M.M., Reina, R., Iráizoz,
E., Crespo, H., Berriatua, E., Vázquez, J., Amorena, B., de Andrés, D., 2011.
Recombinant small ruminant lentivirus subtype B1 in goats and sheep of imported
breeds in Mexico. J. Vet. 190, 169–172.
Sanjose, L., Crespo, H., Blatti-Cardinaux, L., Glaria, I., Martínez-Carrasco, C., Berriatua,
E., Amorena, B., de Andrés, D., Bertoni, G., Reina, R., 2016. Post-entry blockade of
small ruminant lentiviruses by wild ruminants. Vet. Res. 47, 1.
Shah, C., Böni, J., Huder, J.B., Vogt, H.-R., Mühlherr, J., Zanoni, R., Miserez, R., Lutz, H.,
Schüpbach, J., 2004. Phylogenetic analysis and reclassification of caprine and ovine
lentiviruses based on 104 new isolates: evidence for regular sheep-to-goat transmis-
sion and worldwide propagation through livestock trade. Virology 319, 12–26.
Starick, E., Dedek, J., Enke, K.H., Loeppelmann, H., 1995. Serologic investigations on the
occurrence of Maedi-Visna antibodies in-game ruminants Dtsch. Tierarztl.
Wochenschr. 102, 202–203.
Tolari, F., Al-Ramadneh, W., Mazzei, M., Carrozza, M.L., Forzan, M., Bandecchi, P.,
Grego, E., Rosati, S., 2013. Small ruminant lentiviruses in Jordan: evaluation of sheep
and goat serological response using recombinant and peptide antigens. Trop. Anim.
Health Prod. 45, 1335–1340.
Verdugo, C., Toft, N., Nielsen, S.S., 2015. Within-and between-herd prevalence variation
of Mycobacterium avium subsp. paratuberculosis infection among control pro-
gramme herds in Denmark (2011–2013). Prev. Vet. Med. 121, 282–287.