Abstract

Canine distemper (CD) may pose a serious threat to Alpine wild carnivores and affect their population dynamics. Since 2006, the strain Europe Wildlife 2006–09, a distinct CD virus subgroup within viral lineage Europe 1 (EU1) characterized by increased virulence and host range expansion, has been linked to multiple CD outbreaks in Alpine wild carnivores. The aim of this study was to fill knowledge gaps about ongoing Alpine outbreaks of CD. To do this, we report on the circulation of canine distemper virus (CDV) and outbreaks of CD in Alpine wild carnivores in northwest Italy. A specific diagnostic protocol applied to a sample of 548 wild carnivores collected between January 2013 and December 2015 revealed the circulation of CDV belonging to the EU1 lineage. All isolates were carriers of amino-acid mutations defining the cluster Europe Wildlife 2006–09. A self-maintained multihost pathogen system may have developed in northwest Italy in which interspecies transmission from red foxes (Vulpes vulpes) to other noncanid species enhanced pathogen maintenance in the system.

INTRODUCTION

Canine distemper virus (CDV) is a member of the Morbillivirus genus within the Paramyxoviridae family. It is an enveloped, single-stranded, negative-sense RNA virus prone to frequent genetic and antigenic variation (Deem et al. 2000; Panzera et al. 2015). Canine distemper (CD) is a multisystemic disease; its clinical severity varies with strain pathogenicity and host immune status (Beineke et al. 2009; Ke et al. 2015; Loots et al. 2017). Several families of carnivores are susceptible to CDV (Marsilio et al. 1997; Martinez-Gutierrez and Ruiz-Saenz 2016). Although host range expansion has been reported, the domestic dog (Canis lupus familiaris) is still considered to be the principal viral reservoir (Panzera et al. 2015), and it is not fully understood how wild animals become exposed to CDV. It is thought that the virus circulates in wildlife within a complex system in which several interconnected populations may function as a single reservoir, since its circulation in wildlife is active even in areas where levels of infection in domestic dog are low (Martinez-Gutierrez and Ruiz-Saenz 2016). Although the effects of H protein variability are not entirely clear (Ke et al. 2015; Liao et al. 2015), the amino-acid substitutions at residues 530 and 549 of the H protein are believed to play a role in host shifting and the emergence of CDV in nondog host species through molecular adaptation of receptor-binding domains of the virus (McCarthy et al. 2007). Three of the 14 genetic lineages of CDV recognized to date (Ke et al. 2015; Panzera et al. 2015), Europe 1/South America 1 (EU1/SA1), Europe 2/Europe Wildlife (EU2), and Europe 3/Arctic-like (EU3), have been reported in Italy (Martella et al. 2006). All CDV isolates from wild animals in Europe were included in EU2 or EU3 until a distinct subgroup called Europe Wildlife 2006–09 was identified within the viral lineage EU1 in 2006 (Monne et al. 2011) as being responsible for multiple outbreaks in wild carnivores in the eastern Alpine region, with abnormally high mortality rates (Monne et al. 2011; Sekulin et al. 2011; Origgi et al. 2012). Phylogenetic analysis revealed that all CDV strains were part of a single transnational European CDV epidemic that seems to be moving from east to west through the Alps (Origgi et al. 2012). Moreover, a mutation in the H protein was identified as being responsible for host range expansion and increasing virulence (Monne et al. 2011; Sekulin et al. 2011; Nikolin et al. 2012; Origgi et al. 2012). The aim of the present study was to fill the gaps in the data on the circulation of CDV in wild carnivores in northwest Italy (NWI) and to describe the ongoing Alpine CD outbreak. For these purposes, a diagnostic protocol was developed, and phylogenetic analysis was performed to correlate the detected strains with those circulating in the present outbreaks.

MATERIALS AND METHODS

Study area

The study area encompasses the regions of Piedmont, Liguria, and the Aosta Valley, and it is bordered by Switzerland and France on the north and the west, respectively. It is divided into four topographic zones: alpine in the northwest, Apennine in the south, a hilly foothill zone, and a flat area. Each region is subdivided into administrative provinces: Cuneo (CN), Torino (TO), Alessandria (AL), Asti (AT), Vercelli (VC), Novara (NO), Biella (BI), and Verbano-Cusio-Ossola (VCO) in Piedmont; Imperia (IM), Savona (SV), Genova (GE), and La Spezia (SP) in Liguria; and Aosta (AO) in Aosta Valley.

Common carnivore species that are widely distributed in the study area include red fox (Vulpes vulpes), Eurasian badger (Meles meles), beech marten (Martes foina), and least weasel (Mustela nivalis). Less common are pine marten (Martes martes), western polecat (Mustela putorius), and stoat (Mustela erminea), which inhabit only woodlands or wetlands (Boitani et al. 2003). Since 1992, the western Alps have been recolonized by wolf (Canis lupus italicus; Marucco and Avanzinelli 2017).

Sampling

Between January 2013 and December 2015, the carcasses of 548 wild carnivores were collected and categorized as “found dead,” if the cause of death was unknown, or “hunted,” if they died during the hunting season. Hunting is legal only for red fox. Animals that died in wildlife recovery centers with symptoms of CD were categorized as “suspected infected.” Species, gender, age (young or adult if with deciduous or permanent teeth, respectively), sampling date, location, and clinical signs were recorded. All animals were submitted to necropsy and diagnostic testing at the Istituto Zooprofilattico Sperimentale del Piemonte, Liguria e Valle d'Aosta, to determine the cause of death.

The sample set consisted of red foxes (n=446), badgers (n=78), beech martens (n=18), and wolves (n=6). Adult animals accounted for 63.5% (348/548) and young accounted for 33.6% (184/548) of the sample set; age could not be determined for 2.9% (16/548). The sample set was composed of more males (54.1%, 297/548) than females (40.8%, 224/248); gender could not be determined for 5.1% (27/548) of animals. Over half (52.9%, 290/548) of the subjects were found dead, 31.6% (173/548) were hunted, and 13.1% (72/548) were suspected infected. Carcass type could not be determined for 2.4% (13/548) of animals.

Detection of CD and CDV

We used the fluorescent antibody test (FAT) and PCR as screening tests for CDV on brain and lung tissue. Autolysis prevented collection of some samples. We performed FAT on four sections of frozen tissue, about 4–6 μm thick, which were fixed on a slide with absolute acetone for 30 min and air dried. Sections were then incubated with monoclonal antibodies specific for CDV conjugated to fluorescein isothiocyanate for 30 min at 37 C and observed under a fluorescent microscope. Cell cultures that were infected or not infected with a field strain of CDV were used as positive and negative controls, respectively. For fluorescence specificity, another section of the sample was incubated with specific antiserum for bovine herpesvirus 1. We performed PCR as described in Verna et al. (2017) and Frisk et al. (1999). For each organ sample testing positive by any screening tests, we attempted virus isolation on Vero Dog Slam (Peletto et al. 2018).

We used neuropathologic and histopathologic techniques to examine several tissues (not previously frozen) from a subgroup of animals with anamnesis or necropsy suggestive of CDV infection. Tissues were fixed in 10% buffered formaldehyde solution. We evaluated lesions in coronal sections of brain and other organs stained with H&E. Some samples that showed microscopic changes or that tested positive by PCR were selected and submitted to immunohistochemical (IHC) analyses for the detection of CDV antigen using a mouse monoclonal antibody (VMRD, Pullman, Washington, USA). A dog lung and brain confirmed positive by molecular analysis were used as a positive control.

We detected immunoglobulin G (IgG) titers against CDV in blood using an enzyme-linked immunosorbent assay (Distemper IgG Ab ELISA kit, Agrolabo, Scarmango, Italy). According to the manufacturer's instructions, a semiquantitative antibody titration was also determined, and positive samples were classified as low (equivalent indirect antibody fluorescent test [IFAT] titer of 1/20 or 1/40), medium (equivalent IFAT titer of 1/80 or 1/160), or high (equivalent IFAT titer ≥1/320).

For sequencing and phylogenetic analysis, complementary DNA was used for PCR amplification of the complete H gene of CDV by employing four primer sets (Sekulin et al. 2011). We used the software MEGA7 (Kumar et al. 2016) to calculate p-distance matrices and phylogeny inference according to the maximum likelihood criterion. The nucleotide substitution model was set according to jModelTest2 output (Darriba et al. 2012) and Tamura-Nei 3 parameters with gamma variation (T92+G).

Data analysis

Data were entered in a data set (Excel 2007, Microsoft Corporation, Redmond, Washington, USA), and descriptive analysis was performed using statistical software SAS version 9.4 (SAS Institute Inc., Cary, North Carolina, USA). An animal was defined as a CD case if it tested positive on at least one of two screening tests. We calculated CDV prevalence and 95% confidence interval (CI) using the binomial exact test. We performed χ2 tests and used a multivariate logistic regression model to analyze the association between CDV and study factors (year, age, species, location, and carcass type). We calculated the odds ratio (OR), as a measure of association, for statistically significant factors. To assess the accuracy of the screening tests (FAT and PCR), only the animals for which both target organs were available were considered; a pseudo–gold standard was defined by categorizing the results as CVD positive if at least one of the tests was positive on at least one of the two target organs. The agreement between screening tests was tested, and Cohen's kappa index was calculated.

Two choropleth maps were produced to describe the geographic distribution of the samples and the detected CD cases. For the former, each province was categorized into four sampling classes using quantiles. For the CD cases, a bivariate choropleth map depicting agreement or disagreement between two variables (uncertainty and prevalence) was produced. Population density could not be considered because of lack of updated data.

An open-source geographic information system was used (QGIS version 2.0.1-Dufour, Free Software Foundation, Inc., Boston, Massachusetts, USA), applying World Geodetic System 1984, Universal Transverse Mercator zone 32 (EPSG:32632) as coordinate systems.

RESULTS

Detection of CD and CDV

In seven provinces (BI, GE, SP, VCO, Novara, AT, SV), sample size was 10 or fewer, in two provinces (VC, CN), sample size was between 11 and 34, and for the remaining provinces (AO, TO, AL, IM), sample size was higher than 34 (Fig. 1). At least one screening test was applied on lung and brain tissue of 548 animals. In total, 189 (34.5%, 189/548, 95% CI: 30.5–38.5) samples were defined as a CD case At least one CD case was identified per province; in SV, GE, BI, AT, and VCO, the small sample size (less than 10) led to a high degree of uncertainty about the estimated raw prevalence (Fig. 2). The prevalence of CD was high without significant differences between species (χ2=6.312; P=0.097). Canine distemper prevalence in adults (31.3%, 109/348, 95% CI: 26.5–36.2) was lower than in young animals (39.7%, 73/184, 95% CI: 32.6–46.7), though this difference was not statistically significant (χ2=3.730, P<0.053). The prevalence was higher among the suspected-infected (63%, 45/72, 95% CI: 51–74) and the found-dead animals (42.1%, 122/290, 95% CI: 36.4–47.8) than the hunted animals (9.8%, 17/173, 95% CI: 5.4–14.3). Bivariate analysis showed a statistically significant association between year and CD case (χ2=6.05, P=0.039), with lower seropositive values after 2015 (13/65; 20%) than in 2013 (28/73; 38%) or in 2014 (143/407; 35%); an association was also found with province (χ2=106.366, P<0.0001) and carcass type (χ2=79.052, P<0.0001). However, multivariate analysis assessed by logistic regression modeled only two factors that resulted statistically significance: province (χ2=40.42, P<0.0001) and carcass type (χ2=28.51, P<0.0001). The accuracy of this model was 0.78. The probability of observing a CDV case in Aosta was four times higher than in the other areas (OR=4.1, 95% CI: 2.7–6.4); it was nearly two times higher (OR=1.9, 95% CI: 1.1–3.4) in suspected-infected than in found-dead animals. Finally, CD was more likely to be found among dead than hunted animals (OR=0.3, 95% CI: 0.1–0.5). Among the CD cases, 40.2% (76/189, 95% CI: 39.7–40.7) of the animals showed traumatic lesions referable to a road accident.

Figure 1

Choropleth map describing the variation of the spatial distribution of the collected wild carnivores: red fox (Vulpes vulpes), badger (Meles meles), beech marten (Martes foina), and Italian wolf (Canis lupus italicus) in northwest Italy between January 2013 and December 2015. The provinces are, in Piedmont: Torino (TO), Cuneo (CN), Asti (AT), Alessandria (AL), Vercelli (VC), Biella (BI), Novara (NO), Verbano-Cusio-Ossola (VCO); in Liguria: Genoa (GE), Savona (SV), Imperia (IM), La Spezia (SP); and in the Aosta Valley: Aosta (AO). Provinces were categorized on the basis of the number of collected samples by quantile calculus; four classes by shading of varying intensity were obtained: 1–4, 5–10, 11–34, and 35–233 units.

Figure 1

Choropleth map describing the variation of the spatial distribution of the collected wild carnivores: red fox (Vulpes vulpes), badger (Meles meles), beech marten (Martes foina), and Italian wolf (Canis lupus italicus) in northwest Italy between January 2013 and December 2015. The provinces are, in Piedmont: Torino (TO), Cuneo (CN), Asti (AT), Alessandria (AL), Vercelli (VC), Biella (BI), Novara (NO), Verbano-Cusio-Ossola (VCO); in Liguria: Genoa (GE), Savona (SV), Imperia (IM), La Spezia (SP); and in the Aosta Valley: Aosta (AO). Provinces were categorized on the basis of the number of collected samples by quantile calculus; four classes by shading of varying intensity were obtained: 1–4, 5–10, 11–34, and 35–233 units.

Figure 2

Bivariate choropleth map describing the spatial distribution of canine distemper (CD) cases in wild carnivores: red fox (Vulpes vulpes), badger (Meles meles), beech marten (Martes foina), and Italian wolf (Canis lupus italicus) in northwest Italy between January 2013 and December 2015. An animal was defined as a CD case if it tested positive on at least one of two screening tests (fluorescent antibody test [FAT] and PCR). The provinces are, in Piedmont: Torino (TO), Cuneo (CN), Asti (AT), Alessandria (AL), Vercelli (VC), Biella (BI), Novara (NO), Verbano-Cusio-Ossola (VCO); in Liguria: Genoa (GE), Savona (SV), Imperia (IM), La Spezia (SP); and in the Aosta Valley: Aosta (AO). The abbreviation for each province is indicated in each area. The nine shades shown in the panel describe the agreement between raw prevalence (Prev%) and uncertainty (SE).

Figure 2

Bivariate choropleth map describing the spatial distribution of canine distemper (CD) cases in wild carnivores: red fox (Vulpes vulpes), badger (Meles meles), beech marten (Martes foina), and Italian wolf (Canis lupus italicus) in northwest Italy between January 2013 and December 2015. An animal was defined as a CD case if it tested positive on at least one of two screening tests (fluorescent antibody test [FAT] and PCR). The provinces are, in Piedmont: Torino (TO), Cuneo (CN), Asti (AT), Alessandria (AL), Vercelli (VC), Biella (BI), Novara (NO), Verbano-Cusio-Ossola (VCO); in Liguria: Genoa (GE), Savona (SV), Imperia (IM), La Spezia (SP); and in the Aosta Valley: Aosta (AO). The abbreviation for each province is indicated in each area. The nine shades shown in the panel describe the agreement between raw prevalence (Prev%) and uncertainty (SE).

As depicted in Table 1, sensitivity of PCR was higher (72.93% for brain and 75.71% for lung) than sensitivity of FAT (20% for brain and 24.28% for lung). Agreement between brain and lung tissue was moderate for FAT (kappa=0.52; 95% CI: 0.38–0.67) and good for PCR (kappa=0.65; 95% CI: 0.52–0.72).

Table 1

Diagnostic performances of screening tests (fluorescent antibody test and PCR) in lung and brain of wild carnivores: red fox (Vulpes vulpes), badger (Meles meles), beech marten (Martes foina), and Italian wolf (Canis lupus italicus), sampled in northwest Italy between January 2013 and December 2015.

Diagnostic performances of screening tests (fluorescent antibody test and PCR) in lung and brain of wild carnivores: red fox (Vulpes vulpes), badger (Meles meles), beech marten (Martes foina), and Italian wolf (Canis lupus italicus), sampled in northwest Italy between January 2013 and December 2015.
Diagnostic performances of screening tests (fluorescent antibody test and PCR) in lung and brain of wild carnivores: red fox (Vulpes vulpes), badger (Meles meles), beech marten (Martes foina), and Italian wolf (Canis lupus italicus), sampled in northwest Italy between January 2013 and December 2015.

In total, 169 samples were submitted for viral isolation, and CDV was isolated from 9 animals (5.3%, 9/169), all from the Aosta Valley; in seven of these, the clinical signs or macroscopic lesions were compatible with CD. Gross findings referable to CD were observed at necropsy: conjunctivitis with serous or mucopurulent ocular discharge (Fig. 3a), interstitial pneumonia or purulent bronchopneumonia (Fig. 3b), and splenomegaly (Fig. 3c) with multifocal necrosis. Hard pad disease (Fig. 3d) and vascular disorders of the gastric and bladder mucosa were also observed. Among the animals found dead, 63.1% (183/290) of cases of polytrauma were referable to collision with a motor vehicle.

Figure 3

Gross findings referable to canine distemper observed during necropsy of wild carnivores in northwest Italy between January 2013 and December 2015. (a) Red fox (Vulpes vulpes) with severe mucopurulent conjunctivitis. (b) Badger (Meles meles) with interstitial pneumonia associated with suppurative bronchopneumonia (Streptococcus dysgalactiae spp. equisimilis). (c) Red fox with splenomegaly. (d) Beech marten (Martes foina) with thickening of the foot pads.

Figure 3

Gross findings referable to canine distemper observed during necropsy of wild carnivores in northwest Italy between January 2013 and December 2015. (a) Red fox (Vulpes vulpes) with severe mucopurulent conjunctivitis. (b) Badger (Meles meles) with interstitial pneumonia associated with suppurative bronchopneumonia (Streptococcus dysgalactiae spp. equisimilis). (c) Red fox with splenomegaly. (d) Beech marten (Martes foina) with thickening of the foot pads.

Histopathologic examinations were performed on different organs of 29 animals that were categorized as CD cases according to the screening tests. The most frequently affected organ was the lung, with evidence of an interstitial subacute pneumonia characterized by infiltration of macrophages and lymphocytes, syncytial cells in the alveoli, and intranuclear or intracytoplasmic acidophilic inclusion bodies in bronchial epithelial cells. The urinary system, particularly the bladder, was frequently involved, and viral inclusions without inflammatory reaction were sometimes observed in the urothelium. Histologic lesions were also found in 15 brains, characterized by vasculitis, perivascular cuffs, and inflammatory infiltrate in the meninges composed of mononuclear cells.

We used IHC to examine samples from 23 animals showing microscopic lesions. All showed immunopositivity in the alveolar and bronchiolar epithelial cells and the syncytial cells. The kidney and bladder resulted in positive tests for the transitional epithelial cells in four animals. A positive reaction in lymphoid tissue, liver bile ductal epithelium, and intestinal epithelium was sporadically observed. Three of the 19 brains that tested positive showed no neuropathologic changes.

Serology

Serology was performed on 401 blood samples, and 33.2% (133/401, 95% CI: 32.9–33.4) were positive. Seropositivity was 33.5% (84/251, 95% CI: 33.1–33.8) in adult and 33.1% (46/139, 95% CI: 32.4–33.8) in young animals. Significant differences were observed only between carcass types (χ2=10.828; P=0.02), in which the highest prevalence values were observed in hunted animals (49.2%, 95% CI: 48.3–50), followed by found-dead and suspected-infected animals, in which prevalence was 29.3% (95% CI: 28.9–29.7) and 16.7% (95% CI: 15.4–17.9), respectively (Table 2).

Table 2

Results of serology performed on 401 blood samples from wild carnivores: red fox (Vulpes vulpes), badger (Meles meles), beech marten (Martes foina), and Italian wolf (Canis lupus italicus), sampled in northwest Italy between January 2013 and December 2015. Data are shown with detail of titers of immunoglobulin G antibodies categorized by different age and carcass type classes. Table does not show the results of seropositive animals for which age (n=11) or carcass type (n=8) could not be determined.

Results of serology performed on 401 blood samples from wild carnivores: red fox (Vulpes vulpes), badger (Meles meles), beech marten (Martes foina), and Italian wolf (Canis lupus italicus), sampled in northwest Italy between January 2013 and December 2015. Data are shown with detail of titers of immunoglobulin G antibodies categorized by different age and carcass type classes. Table does not show the results of seropositive animals for which age (n=11) or carcass type (n=8) could not be determined.
Results of serology performed on 401 blood samples from wild carnivores: red fox (Vulpes vulpes), badger (Meles meles), beech marten (Martes foina), and Italian wolf (Canis lupus italicus), sampled in northwest Italy between January 2013 and December 2015. Data are shown with detail of titers of immunoglobulin G antibodies categorized by different age and carcass type classes. Table does not show the results of seropositive animals for which age (n=11) or carcass type (n=8) could not be determined.

Among the seropositive animals, 50.4% (67/133, 95% CI: 49.6–51.1) were in the low-titer category (1/20–1/40), and only 15.8% (21/133, 95% CI: 15–16) of those were categorized as a CD case. Finally, among the CD cases, 18% (34/189, 95% CI: 17.6–18.4) were also seropositive: 21 with low titer, 7 with medium titer, and 6 with high titer. Multi-organ trauma was observed in 47% (16/34) of seropositive CD cases, and 21% (7/34) showed medium or high titers.

Phylogenetic analysis and strain comparison

Good-quality sequences were obtained for 30 samples from foxes (n=20), badgers (n=9), and beech marten (n=1). All animals originated from the Aosta Valley, except for one fox and one badger from the province of Turin. Phylogenetic analysis was performed based on alignment of 1,659 base pairs corresponding to residues from 55 to 605 of the H protein (complete coding sequence=607 amino acids), by aligning the study sequences and 99 homologous sequences representative of the main CDV lineages available in GenBank. The phylogenetic tree (Supplementary Material Fig. S1) showed that all CDV sequences analyzed were classified in the Europe-1 lineage, showing 98.6–100% similarity among these and 96.4–99.6% similarity with other sequences of the same lineage.

Within the Europe-1 lineage, the CDV isolates formed two subclades, defined by three nucleotide changes at positions 171 (G to A), 183 (C to A), and 212 (A to G). The 171-A variant is a synonymous change that seems to be specific for the CDV isolates characterized in our study. The other two nucleotide changes were present also in other sequences of the data set and were nonsynonymous, determining amino-acid changes at positions 61 (H to Q) and 71 (N to S), respectively. The 212-A variant, corresponding to asparagine at position 71, was present only in two other sequences identified in foxes by Origgi et al. (2012) in the course of the distemper epidemics in wild carnivores in Switzerland in 2010. Moreover, all the isolates of the Europe-1 lineage typed in our study carried the coding change Y to H at position 549, as previously reported in isolates from CDV outbreaks described in Germany and northeast Italy (Benetka et al. 2011; Monne et al. 2011; Sekulin et al. 2011). Our strains also shared with these isolates the amino-acid mutations at positions 266 (L to F), 590 (S/T to I), and 597 (R to H), which seem to be peculiar to the Europe Wildlife 2006–09 cluster. A noteworthy finding was that the H protein of two CDV sequences from the foxes originating from the Aosta Valley carried a leucine at position 345; all other sequences in the data set had isoleucine at this site, except for a canine CDV strain (GenBank accession no. HM443710) of the Arctic-like lineage identified in northeast Italy (province of Padua) in 2000 (Monne et al. 2011). Two other CDV isolates from foxes from the Aosta Valley showed coding mutations at positions 354 (I to V) and 371 (Q to E).

DISCUSSION

Our main objective was to determine if CDV circulated in wild carnivore populations of NWI. About a third of the animals were positive by at least one of the two screening tests, indicating the circulation of CDV infection in the study area, as confirmed by isolation of CDV and by IHC. An accurate and updated estimation of the wild carnivore population is lacking, as is adequate knowledge of the epidemiology of the disease in the study area, which precluded calculating the true prevalence of CD. Found-dead animals made up the majority of the carcasses we sampled, the poor quality of which may have limited the diagnostic test performance of some tests (e.g., FAT and virus isolation), although for virus isolation, we used Vero cells modified with a signaling lymphocyte activation molecule receptor to increase sensitivity. The reduction in diagnostic performance was probably due to the low cellular integrity of the matrix, but this did not seem to affect PCR (Table 1). Considering our results, we feel that FAT, which is widely used for the diagnosis of CDV in domestic animals, should be reevaluated or at least refined according to the characteristics of the sample.

Our sample consisted mainly of red foxes; besides being widespread, it is the only huntable carnivore species in the study area. The high raw CD prevalence (Table 3) in the found-dead animals (OR=1.9, 95% CI: 1.1–3.4), and even higher prevalence in the suspected-infected animals (OR=0.3, 95% CI: 0.1–0.5) are expected during epidemic CD outbreaks (Origgi et al. 2012). The statistically significant decrease in prevalence during the sampling period (from 38% in 2013 to 20% in 2015) might have been due to an epidemic peak; the following decrease could have been due to host density reduction and increase in immune subjects.

Table 3

Raw prevalence and association between animal features and positivity in wild carnivores: red fox (Vulpes vulpes), badger (Meles meles), beech marten (Martes foina), and Italian wolf (Canis lupus italicus), sampled in northwest Italy between January 2013 and December 2015 and tested for canine distemper (CD) virus using a fluorescent antibody test and PCR as screening tests. An animal was defined as a case of CD if the tissue sample tested positive by at least one of the two screening tests. Not shown are the results of animals for which age (n=16) or carcass type (n=13) could not be determined. The provinces not listed in Table 3, because sample size was too low, are: Asti (one positive of five animals tested), Biella (two positive of three animals tested), Novara (four positive of 10 animals tested), Verbano-Cusio-Ossola (one positive of six animals tested), Genova (one positive of four animals tested), Savona (two positive of five animals tested), and La Spezia (one negative of one animal tested).

Raw prevalence and association between animal features and positivity in wild carnivores: red fox (Vulpes vulpes), badger (Meles meles), beech marten (Martes foina), and Italian wolf (Canis lupus italicus), sampled in northwest Italy between January 2013 and December 2015 and tested for canine distemper (CD) virus using a fluorescent antibody test and PCR as screening tests. An animal was defined as a case of CD if the tissue sample tested positive by at least one of the two screening tests. Not shown are the results of animals for which age (n=16) or carcass type (n=13) could not be determined. The provinces not listed in Table 3, because sample size was too low, are: Asti (one positive of five animals tested), Biella (two positive of three animals tested), Novara (four positive of 10 animals tested), Verbano-Cusio-Ossola (one positive of six animals tested), Genova (one positive of four animals tested), Savona (two positive of five animals tested), and La Spezia (one negative of one animal tested).
Raw prevalence and association between animal features and positivity in wild carnivores: red fox (Vulpes vulpes), badger (Meles meles), beech marten (Martes foina), and Italian wolf (Canis lupus italicus), sampled in northwest Italy between January 2013 and December 2015 and tested for canine distemper (CD) virus using a fluorescent antibody test and PCR as screening tests. An animal was defined as a case of CD if the tissue sample tested positive by at least one of the two screening tests. Not shown are the results of animals for which age (n=16) or carcass type (n=13) could not be determined. The provinces not listed in Table 3, because sample size was too low, are: Asti (one positive of five animals tested), Biella (two positive of three animals tested), Novara (four positive of 10 animals tested), Verbano-Cusio-Ossola (one positive of six animals tested), Genova (one positive of four animals tested), Savona (two positive of five animals tested), and La Spezia (one negative of one animal tested).

Canine distemper virus appears to be circulating throughout all of NWI (Fig. 2), with a considerable prevalence in the provinces of TO and CN, but especially in the AO (OR=4.1, 95% CI: 2.7–6.4), to which the disease was probably introduced from other Alpine areas. Origgi et al. (2012) hypothesized that the epidemic front seems to be advancing via the territorial continuity of Alpine areas. Considering the geography of the study area and the higher prevalences of disease in AO, TO, and CN, it is plausible that the spread of CD primarily involved the Aosta Valley before entering the provinces of TO and CN. The Aosta Valley and Piedmont are both delimited to the north and west by the Alps. The Aosta Valley is mountainous with an average altitude of 2,100 m; two thirds of Piedmont are Alpine and hilly territories surrounded by a semicircle that sweeps from south to west to north delineating the borders with France and Switzerland. In addition, Liguria is almost entirely mountainous. Some provinces, such as IM, VC, AL, and SP, are generally hilly to flat or are far from the epidemic front, and the lower CD prevalence is likely compatible with a lower spread of infection (Fig. 2 and Table 3). For the remaining provinces, it was not possible to confirm or disprove the hypothesis advanced for the other areas (Fig. 2) because of the high estimated prevalence with an elevated degree of uncertainty. Our data seemed to suggest that the geography of the study area was a determinant factor in the spread of CD; however, considering the high adaptability of these carnivores and the absence of population density data, the reasons for these differences remain hypothetical.

Gross pathology and microscopic features were comparable with those described in the literature (Zhao et al. 2015; Caswell and Williams 2016), with predominant localization of lesions in the lung and in the urinary system. Immunopositivity without lesions was frequently noted in urinary bladder and brain. This finding is consistent with death at the earliest stage of infection, which stops the development of further lesions (Origgi et al. 2012).

Our serologic results suggested a wide CDV circulation in the study area. Although the antibody titers in wild animals that are protective are unknown, the high prevalence of low antibody titers could be related to a low survival rate, resulting in a scarcity of immunized animals. Low titers among CD cases could represent a viremic state or an early stage of infection at the time of sampling, when immunity was still developing (Frölich et al. 2000). Also, for suspected-infected animals, which were less seropositive as compared to the other classes (Table 2), death during the initial stage of disease can be assumed.

The high frequency of multi-organ trauma observed among seropositive animals and in CD cases could be related to a clinical stage of CD that may have predisposed them to accidental trauma. The difficulty of assuring a high quality of samples influenced our choice of serologic techniques. We were unable to apply the serum neutralization test because it is strongly affected by hemolysis and contamination of serum causing cytotoxicity to the cell line. In contrast, commercial ELISA kits allowed us to overcome this limitation and to perform semiquantitative titration also in samples containing hemolytic sera.

Another objective of our study was to define the source of CDV infection in the wild carnivore population in NWI. Phylogenetic analysis showed a close relation between the CDV that we identified and the emergent CDV strain named Europe Wildlife 2006–09, which has been spreading from east to west through the Alps since 2006 (Monne et al. 2011; Sekulin et al. 2011; Origgi et al. 2012; Nouvellet et al. 2013). All Europe 1 isolates characterized in our study were carriers of amino-acid mutations at position 266 (L to F), 590 (S/T to I), and 597 (R to H), which define the Europe Wildlife 2006–09 cluster. The finding of the same strain in different species and the amino-acid mutations Y to H in position 549 responsible for host range expansion (Sekulin et al. 2011) support the hypothesis of multihost epizootic spread.

Here, we highlighted the spread of the Europe Wildlife 2006–09 cluster in NWI. Since no abnormally high mortality was recorded in the past years, we assume that this emergent CDV may have been introduced through the Alps. Therefore, as suggested by Monne et al. (2011), it is possible that a lack of population immunity to the new CDV may have favored the spread of the CD epidemic, aided by the high population density of medium-sized predators (Lucherini and Lovari 1996) through an increase in effective contacts.

Some authors consider CDV a true multihost pathogen that could be endemic in wild and domestic canids with spill-over events to other carnivores (Gortàzar et al. 2007). In order to sustain an epidemic of CD, one or more species has/have to be abundant and susceptible and able to transmit the infection (Loots et al. 2017). Our data did not allow us to establish differences in species susceptibility to CD between mustelids and other carnivore families, perhaps due to the type of sampling, which was based, for some species, on dead animals. However, a high lethality rate of CD has ben reported in mustelids (Nikolin et al. 2012; Origgi et al. 2012; Martinez-Gutierrez and Ruiz-Saenz 2016). Moreover, red foxes seem to be more susceptible to the Europe 1 lineage, probably owing to the closer genetic relationship between dogs and wild canids (Nikolin et al. 2012). The seropositive healthy animals (Martinez-Gutierrez and Ruiz-Saenz 2016) we recorded indicated recovery of red foxes. Although the spatial distribution of the fox population is not accurately described for NWI, it is known that this species has the widest geographic distribution among the canids. It is common in all environments in Italy, with an average density of 1–2.5 foxes/km2 (Boitani et al. 2003). For these reasons, a potential primary role of the red fox in the spread and maintenance of CD in NWI, especially during the reproductive period and juvenile migration, is plausible. A self-maintained multihost pathogen system may have developed in NWI in which interspecies transmission from red foxes to other noncanid species enhances pathogen maintenance in the system (Loots et al. 2017). Some authors (Haydon et al. 2002; Martinez-Gutierrez and Ruiz-Saenz 2016) have also proposed a new concept called metareservoir for contexts like this, in which some interconnected populations may function as a single reservoir.

No spill-over of this new CDV in dogs has been reported (Monne et al. 2011), probably due to the lack of contacts or to the virus strain's strong specificity. However, a fatal transmission to a vaccinated dog during the 2009–10 Swiss epidemic was reported (Origgi et al. 2012). While spill-over cannot be completely excluded, its occurrence is at best uncertain (Frölich et al. 2000; Martella et al. 2002). Although the level of cross-immunity of vaccines in use is not clear, vaccination remains the basic preventive measure for dogs, especially in those areas where there is a spatial overlap between domestic and wild carnivores.

The CDV poses one of the most serious threats to carnivore conservation, especially for small fragmented populations with a density lower than the minimum viable population and low genetic intrapopulation variability in which the introduction of new pathogens can cause a drastic population decline (Fenton and Pedersen 2005; Smith et al. 2009). Our reporting on three of the six wolves as positive for CDV and the severe outbreaks within wolf populations in the Apennines (Di Sabatino et al. 2014) are cause for concern in NWI, which has recently been recolonized by this species. Moreover, CD epidemics in immunologically naïve ecosystems can modify the structure of the infected populations through high lethality rates, with inevitable repercussions on other density factors such as predators, trophic and spatial resources, and other infectious diseases. The full host range and ecology of the Europe lineage and European Wildlife lineage of CDV are currently unknown (Nikolin et al. 2012). Further studies are needed to understand whether this new CDV is well adapted in wildlife and whether it can be maintained in wildlife populations between outbreaks in NWI.

ACKNOWLEDGMENTS

We thank Sabrina Nodari, Daniela Mei, Leonardo Pinto, Marco Perruchon, and Anna Soncin (Istituto Zooprofilattico Sperimentale del Piemonte, Liguria e Valle D'Aosta) for their technical assistance. This research project was funded by the Italian Ministry of Health (project IZS PLV 10/12 RC). None of the authors has a financial or personal relationship with other people or organizations that could inappropriately influence the content of the paper.

SUPPLEMENTARY MATERIAL

Supplementary material for this article is online at http://dx.doi.org/10.7589/2018-09-226.

LITERATURE CITED

LITERATURE CITED
Beineke
A,
Puff
C,
Seehusen
F,
Baumgärtner
W.
2009
.
Pathogenesis and immunopathology of systemic and nervous canine distemper
.
Vet Immunol Immunopathol
127
:
1
18
.
Benetka
V,
Leschnik
M,
Affenzeller
N,
Möstl
K.
2011
.
Phylogenetic analysis of Austrian canine distemper virus strains from clinical samples from dogs and wild carnivores
.
Vet Rec
168
:
377
.
Boitani
L,
Lovari
S,
Vigna Taglianti
A.
2003
.
Fauna d'Italia, Mammalia III: Carnivora–Artiodactyla
.
Calderini
,
Bologna, Italy
,
434
pp
.
Caswell
JL,
Williams
KJ.
2016
.
Infectious respiratory diseases of dogs, viral diseases
.
In
:
Pathology of domestic animals, 6th Ed
.,
Maxie
MM,
editor
.
Elsevier
,
St. Louis, Missouri
,
pp
.
574
577
.
Darriba
D,
Taboada
GL,
Doallo
R,
Posada
D.
2012
.
jModelTest 2: More models, new heuristics and parallel computing
.
Nat Methods
9
:
772
.
Deem
SL,
Spelman
LH,
Yates
RA,
Montali
RJ.
2000
.
Canine distemper in terrestrial carnivores: A review
.
J Zoo Wildl Med
31
:
441
451
.
Di Sabatino
D,
Lorusso
A,
Di Francesco
CE,
Gentile
L,
Di Pirro
V,
Bellacicco
AL,
Giovannini
A,
Di Francesco
G,
Marruchella
G,
Marsilio
F.
2014
.
Arctic lineage–canine distemper virus as a cause of death in Apennine wolves (Canis lupus) in Italy
.
PLoS One
9
:
e82356
.
Fenton
A,
Pedersen
AB.
2005
.
Community epidemiology framework for classifying disease threats
.
Emerg Infect Dis
11
:
1815
1821
.
Frisk
AL,
König
M,
Moritz
A,
Baumgärtner
W.
1999
.
Detection of canine distemper virus nucleoprotein RNA by reverse-transcription-PCR using serum, whole blood, and cerebrospinal fluid from dogs with distemper
.
J Clin Microbiol
37
:
3634
3643
.
Frölich
K,
Czupalla
O,
Haas
L,
Hentschke
J,
Dedek
J,
Fickel
J.
2000
.
Epizootiological investigations of canine distemper virus in free-ranging carnivores from Germany
.
Vet Microbiol
74
:
283
292
.
Gortàzar
C,
Ferroglio
E,
Höfle
U,
Frölich
K,
Vicente
J.
2007
.
Diseases shared between wildlife and livestock: A European perspective
.
Eur J Wildl Res
53
:
241
256
.
Haydon
DT,
Cleaveland
S,
Taylor
LH,
Laurenson
MK.
2002
.
Identifying reservoirs of infection: A conceptual and practical challenge
.
Emerg Infect Dis
8
:
1468
1473
.
Ke
GM,
Ho
CH,
Chiang
MJ,
Sanno-Duanda
B,
Chung
CS,
Lin
MY,
Shi
YY,
Yang
MH,
Tyan
YC,
Liao
PC,
et al
2015
.
Phylodynamic analysis of the canine distemper virus hemagglutinin gene
.
BMC Vet Res
11
:
164
.
Kumar
S,
Stecher
G,
Tamura
K.
2016
.
MEGA7: Molecular evolutionary genetics analysis version 7.0 for bigger datasets
.
Mol Biol Evol
33
:
1870
1874
.
Liao
P,
Guo
L,
Wen
Y,
Yang
Y,
Cheng
S.
2015
.
Phylogenetic features of hemagglutin gene in canine distemper virus strains from different genetic lineages
.
Int J Clin Exp Med
8
:
6607
6612
.
Loots
AK,
Mitchell
E,
Dalton
DL,
Kotzé
A,
Venter
EH.
2017
.
Advances in canine distemper virus (CDV) pathogenesis research: A wildlife perspective
.
J Gen Virol
98
:
311
321
.
Lucherini
M,
Lovari
S.
1996
.
Habitat richness affects home range size in the red fox Vulpes vulpes
.
Behav Proc
36
:
103
105
.
Marsilio
F,
Tiscar
PG,
Gentile
L,
Roth
HU,
Boscagli
G,
Tempesta
M,
Gatti
A.
1997
.
Serologic survey for selected virus pathogens in brown bears from Italy
.
J Wildl Dis
33
:
304
307
.
Martella
V,
Cirone
F,
Elia
G,
Lorusso
E,
Decaro
N,
Campolo
M,
Desario
C,
Lucente
MS,
Bellacicco
AL,
Blixenkrone-Møller
M.
2006
.
Heterogeneity within the hemagglutinin genes of canine distemper virus (CDV) strains detected in Italy
.
Vet Microbiol
116
:
301
309
.
Martella
V,
Pratelli
A,
Cirone
F,
Zizzo
N,
Decaro
N,
Tinelli
A,
Foti
M,
Buonavoglia
C.
2002
.
Detection and genetic characterization of canine distemper virus (CDV) from free-ranging red foxes in Italy
.
Mol Cell Probe
16
:
77
83
.
Martinez-Gutierrez
M,
Ruiz-Saenz
J.
2016
.
Diversity of susceptible hosts in canine distemper virus infection: A systematic review and data synthesis
.
BMC Vet Res
12
:
78
.
Marucco
F,
Avanzinelli
E.
2017
.
Lo status del lupo in Regione Piemonte
.
In
:
Lo Status della popolazione di lupo sulle Alpi Italiane 2014–2016 Relazione tecnica
,
Progetto LIFE 12 NAT/IT/00080 WOLFALPS–Azione A4 (AllegatoA)
. .
McCarthy
AJ,
Shaw
MA,
Goodman
SJ.
2007
.
Pathogen evolution and disease emergence in carnivores
.
Proc Biol Sci
274
:
3165
3174
.
Monne
I,
Fusaro
A,
Valastro
V,
Citterio
C,
Dalla Pozza
M,
Obber
F,
Trevisiol
K,
Cova
M,
De Benedictis
P,
Bregoli
M,
et al.
2011
.
A distinct CDV genotype causing a major epidemic in Alpine wildlife
.
Vet Microbiol
150
:
63
69
.
Nikolin
VM,
Wibbelt
G,
Michler
FU,
Wolf
P,
East
ML.
2012
.
Susceptibility of carnivore hosts to strains of canine distemper virus from distinct genetic lineages
.
Vet Microbiol
156
:
45
53
.
Nouvellet
P,
Donnelly
CA,
De Nardi
M,
Rhodes
CJ,
De Benedictis
P,
Citterio
C,
Obber
F,
Lorenzetto
M,
Dalla Pozza
M,
Cauchemez
S,
et al.
2013
.
Rabies and canine distemper virus epidemics in the red fox population of northern Italy (2006–2010)
.
PLoS One
8
:
e61588
.
Origgi
FC,
Plattet
P,
Sattler
U,
Robert
N,
Casaubon
J,
Mavrot
F,
Pewsner
M,
Wu
N,
Giovannini
S,
Oevermann
A,
et al.
2012
.
Emergence of canine distemper virus strains with modified molecular signature and enhanced neuronal tropism leading to high mortality in wild carnivores
.
Vet Pathol
49
:
913
929
.
Panzera
Y,
Sarute
N,
Iraola
G,
Hernández
M,
Pérez
R.
2015
.
Molecular phylogeography of canine distemper virus: Geographic origin and global spreading
.
Mol Phylogenet Evol
92
:
147
154
.
Peletto
S,
Caruso
C,
Cerutti
F,
Modesto
P,
Biolatti
C,
Pautasso
A,
Grattarola
C,
Giorda
F,
Mazzariol
S,
Mignone
W,
et al.
2018
.
Efficient isolation on Vero.DogSLAMtag cells and full genome characterization of dolphin morbillivirus (DMV) by next generation sequencing
.
Sci Rep
8
:
860
.
Sekulin
K,
Hafner-Marx
A,
Kolodziejek
J,
Janik
D,
Schmidt
P,
Nowotny
N.
2011
.
Emergence of canine distemper in Bavarian wildlife associated with a specific amino-acid exchange in the haemagglutinin protein
.
Vet J
187
:
399
401
.
Smith
KF,
Acevedo-Whitehouse
K,
Pedersen
AB.
2009
.
The role of infectious diseases in biological conservation
.
Anim Conserv
12
:
1
12
.
Verna
F,
Giorda
F,
Miceli
I,
Rizzo
G,
Pautasso
A,
Romano
A,
Iulini
B,
Pintore
MD,
Mignone
W,
Grattarola
C,
et al.
2017
.
Detection of morbillivirus infection by RT-PCR RFLP analysis in cetaceans and carnivores
.
J Virol Methods
247
:
22
27
.
Zhao
J,
Shi
N,
Sun
Y,
Martella
V,
Nikolin
V,
Zhu
C,
Zhang
H,
Hu
B,
Bai
X,
Yan
X.
2015
.
Pathogenesis of canine distemper virus in experimentally infected raccoon dogs, foxes, and minks
.
Antiviral Res
122
:
1
11
.

Supplementary data