AVIAN BORNAVIRUS IN FREE-RANGING WATERFOWL: PREVALENCE OF ANTIBODIES AND CLOACAL SHEDDING OF VIRAL RNA Author(s): Pauline Delnatte, Éva Nagy, Davor Ojkic, David Leishman, Graham Crawshaw, Kyle Elias, and Dale A. Smith Source: Journal of Wildlife Diseases, 50(3):512-523. Published By: Wildlife Disease Association DOI: http://dx.doi.org/10.7589/2013-08-218 URL: http://www.bioone.org/doi/full/10.7589/2013-08-218

BioOne (www.bioone.org) is a nonprofit, online aggregation of core research in the biological, ecological, and environmental sciences. BioOne provides a sustainable online platform for over 170 journals and books published by nonprofit societies, associations, museums, institutions, and presses. Your use of this PDF, the BioOne Web site, and all posted and associated content indicates your acceptance of BioOne’s Terms of Use, available at www.bioone.org/page/terms_of_use. Usage of BioOne content is strictly limited to personal, educational, and non-commercial use. Commercial inquiries or rights and permissions requests should be directed to the individual publisher as copyright holder.

BioOne sees sustainable scholarly publishing as an inherently collaborative enterprise connecting authors, nonprofit publishers, academic institutions, research libraries, and research funders in the common goal of maximizing access to critical research.

DOI: 10.7589/2013-08-218

Journal of Wildlife Diseases, 50(3), 2014, pp. 512–523 # Wildlife Disease Association 2014

AVIAN BORNAVIRUS IN FREE-RANGING WATERFOWL: PREVALENCE OF ANTIBODIES AND CLOACAL SHEDDING OF VIRAL RNA Pauline Delnatte,1,2 E´va Nagy,1 Davor Ojkic,3 David Leishman,1 Graham Crawshaw,2 Kyle Elias,1 and Dale A. Smith1,4 1 2 3 4

Ontario Veterinary College, University of Guelph, Guelph, Ontario N1G 2W1, Canada Toronto Zoo, Toronto, 361A Old Finch Avenue, Ontario M1B 5K7, Canada Animal Health Laboratory, University of Guelph, Guelph, Ontario N1H 6R8, Canada Corresponding author (email: [email protected])

ABSTRACT: We surveyed free-ranging Canada Geese (Branta canadensis), Trumpeter Swans (Cygnus buccinator), Mute Swans (Cygnus olor), and Mallards (Anas platyrhynchos) to estimate the prevalence of antibodies to avian bornavirus (ABV) and of cloacal shedding of ABV RNA in southern Ontario, Canada. Blood samples and cloacal swabs were collected from 206 free-ranging Canada Geese, 135 Trumpeter Swans, 75 Mute Swans, and 208 Mallards at 10 main capture sites between October 2010 and May 2012. Sera were assessed for antibodies against ABV by enzymelinked immunosorbent assay and swabs were evaluated for ABV RNA using real-time reversetranscription PCR. Serum antibodies were detected in birds from all four species and at each sampling site. Thirteen percent of the geese caught on the Toronto Zoo site shed ABV RNA in feces compared with 0% in geese sampled at three other locations. The proportions of shedders among Mute Swans, Trumpeter Swans, and Mallards were 9%, 0%, and 0%, respectively. Birds that were shedding viral RNA were more likely to have antibodies against ABV and to have higher antibody levels than those that were not, although many birds with antibodies were not shedding. We confirmed that exposure to, or infection with, ABV is widespread in asymptomatic free-ranging waterfowl in Canada; however, the correlation between cloacal shedding, presence of antibodies, and presence of disease is not fully understood. Key words: Avian bornavirus, Canada Goose, Mallard, Mute Swan, serology, shedding, Trumpeter Swan, waterfowl.

waterfowl and gull species across the US (Payne et al. 2011a, 2012; Guo et al. 2012). The bornaviruses circulating among North American waterfowl form a unique genotype (ABV-CG), clustering separately from those described in parrots and canaries, from mammalian Borna disease virus (BDV), and from the BDV sequences isolated from Mallards (Anas platyrhynchos) and Jackdaws (Corvus monedula) in Sweden (Berg et al. 2001; Payne et al. 2011a; Delnatte et al. 2013a). The prevalence and distribution of ABV infection in free-ranging waterfowl in Ontario are not known. In a retrospective study (Delnatte et al. 2013a), the proportion of birds affected by ABV-related neurologic disease differed among species and locations, being higher in birds from the Toronto Zoo than in those found elsewhere. The prevalence of ABV-infected birds in the US ranges from 0% to 50%

INTRODUCTION

The waterfowl genotype of avian bornavirus (ABV) was first identified in freeranging Canada Geese (Branta canadensis) and Trumpeter Swans (Cygnus buccinator) in 2009 in Ontario, Canada, where it was associated with encephalitis and mortality (Delnatte et al. 2011). Retrospective evaluation of 955 necropsy cases from waterfowl species in Ontario confirmed that the presence of ABV in brain is correlated with lymphoplasmacytic inflammation in the central, peripheral, and autonomic nervous tissues (Delnatte et al. 2013a). These lesions are similar to those seen in parrots affected with proventricular dilation disease (PDD), which is caused by various psittacine-associated genotypes of ABV (Gancz et al. 2010). More recently, ABV infection has been shown to be widespread in free-ranging 512

DELNATTE ET AL.—AVIAN BORNAVIRUS IN WATERFOWL

depending on species, sampling site, and tissues examined, also supporting uneven geographic and species distribution of ABV (Guo et al. 2012; Payne et al. 2012). Avian bornavirus infection has been studied best in captive psittacine birds where investigations of PDD outbreaks have shown that ABV is often detectable in cloacal swabs of affected birds and that antibodies can be identified in serum of PDD-affected and asymptomatic birds (de Kloet et al. 2011; Piepenbring et al. 2012). A large-scale survey revealed that ABV infection is widespread among captive psittacines in Europe, with 23% of 1,442 birds considered infected (Heffels-Redmann et al. 2011). Given the paucity of information about bornaviruses in free-ranging birds, we sought to acquire further knowledge about the epidemiology of ABV in waterfowl and, more specifically, to estimate how widespread exposure was to ABV among Anseriformes in Ontario. We hypothesized that the prevalence of antibodies to ABV would be greater than the prevalence of viral shedding in feces, on the basis of the occurrence of intermittent fecal shedding in ABV-infected psittacine birds, and that the proportions of shedders and of antibodypositive birds would vary among locations and among species. To detect exposure to or infection with ABV, we used a combination of serology and reverse-transcription (RT)PCR on cloacal swabs. Our objectives were to survey free-ranging Canada Geese, Mallards, Trumpeter Swans, and Mute Swans (Cygnus olor) to estimate the prevalence of antibodies to ABV and of cloacal shedding of ABV RNA in southern Ontario. MATERIALS AND METHODS Study area and sample collection

Cloacal swabs were collected from 206 Canada Geese, 208 Mallards, 75 Mute Swans, and 135 Trumpeter Swans in southern Ontario between October 2010 and May 2012. We strived to achieve a sample size of 200 birds per species on the basis of a 95% confidence interval (CI) for detection of a disease at an

513

estimated prevalence of 0–4% (Dohoo et al. 2009). Sampling locations are mapped in Figure 1 and global positioning system coordinates of main capture sites are presented in Table 1. Geese and ducks were handled by the Canadian Wildlife Service (CWS) as part of their annual banding exercises, except for the geese at the zoo, which were specifically captured for sampling. Goose and duck capture sites were predefined on the basis of the results of a retrospective necropsy study and on feasibility of bird capture. Canada Geese were caught by herding molting, flightless birds into baited drive traps; Mallards were caught in baited swim-in traps. In contrast, swan samples could not be collected following a predefined schedule. Formerly extinct in Ontario, Trumpeter Swans were reintroduced in the 1980s and a dedicated team of biologists and volunteers from the Trumpeter Swan Reintroduction Program performs a variety of management activities, including wing tagging, leg banding, winter feeding, and daily observations of the wintering flock. Swan samples were thus collected by team members at locations within the swans’ winter range and at times when birds could be baited into catching areas to manually capture them. Capture methods and sample collection were approved by the University of Guelph Animal Care Committee. A Migratory Bird Research Permit (CWS) was obtained. For each bird sampled, species, identification (band/wing tags), sex, age class, capture date, and location were recorded (Table 1). Blood samples were collected from all Mute Swans (n575), from the majority of Trumpeter Swans (n5130) and Canada Geese (n5203), and from 92 Mallards. Blood was collected from the medial metatarsal or ulnar (Mallards) vein using a 3-mL syringe and a 23-G sterile needle. Sera were separated and stored at 220 C. Cloacal swabs were collected by opening the cloaca, inserting a sterile tipped applicator 1–2 cm, and swabbing the mucosa. Any large pieces of feces were shaken off and the swab was placed in a virus transport medium (Multitrans System, Starplex Scientific Inc., Etobicoke, Ontario, Canada). Swabs were frozen at 220 C until shipped and then held at 280 C. Antigen preparation

A pET21a plasmid-based expression cassette for the M24N protein was kindly provided by Dr. Susan Payne (Texas A&M University, College Station, Texas, USA). The isolated plasmid DNA was transformed into

514

JOURNAL OF WILDLIFE DISEASES, VOL. 50, NO. 3, JULY 2014

FIGURE 1. Study area in southern Ontario, Canada showing sampling locations for Canada Geese (Branta canadensis), Mute Swans (Cygnus olor), Trumpeter Swans (Cygnus buccinator), and Mallards (Anas platyrhynchos) from which cloacal swabs and blood samples were analyzed to determine the proportion of birds infected with avian bornavirus, 2010–12.

BL21 cells (DE3) (Novagen 2003). Protein purification was performed with the Qiagen Ni-NTA purification system according to the manufacturer’s instructions (Qiagen, Dusseldorf, Germany). The protein was analyzed by sodium dodecyl sulfate polyacrylamide gel electrophoresis and protein concentration was determined by the Bio-Rad Protein Assay (Bio-Rad, Hercules, California, USA). Serology

An enzyme-linked immunosorbent assay (ELISA) was developed to estimate ABVspecific antibodies in serum samples. ABV-N protein was purified as described above, and diluted in a carbonate buffer (35 mM NaHCO3, 15 mM Na2CO3 [pH 9.6]). Immulon 2HB 96-well microtiter plates (Dynex Technologies Inc., Chantilly, Virginia, USA) were coated with antigen at a concentration of 100 ng/well and incubated at 4 C for 16 hr. The plates were blocked with 3% bovine serum albumin at 37 C for 75 min. The serum samples were heat-treated for 30 min at 56 C. Sera were diluted 1:20 in a wash buffer (0.05%

Tween 20 in phosphate-buffered saline) and 100 mL were added to each well and incubated for 1 hr at 37 C. Each reaction was performed in duplicate wells. The plates were washed four times between each step. One hundred microliters of horseradish peroxidase-labeled goat anti-bird IgG heavy- and light-chain antibody conjugate (Bethyl Laboratories Inc., Montgomery, Texas, USA) at a dilution of 1:5,000 were added and incubated for 75 min at 37 C. The color was developed with ABTS peroxidase substrate system (Kirkegaard & Perry Laboratories, Gaithersburg, Maryland, USA) until the average optical density (OD) of the two positive control wells reached 0.8, and was read in a BIO-Tek ELISA microplate reader (Bio-Tek Instruments, Winooski, Vermont, USA) at 405 nm. When the OD405 between two duplicates differed more than 15%, the sample was retested. The ABV-specific antibody responses were determined by calculating the sample-to-positive (S/P) ratio: (sample mean 2 negative control mean)/(positive control mean 2 negative control mean). Serum from a domestic goose (Anser anser domesticus) (see below) and from a Canada

Two to three locations sampled within a 15-km radius; GPS coordinates indicate the average coordinates.

Only cloacal swabs collected (no blood).

72 3 0 15 90

c

22 12 7 4 45

Winter 2011–12

GPS 5 global positioning system; HY 5 hatching year; AHY 5 after hatching year; U 5 unknown.

43u189N, 79u519W 43u429N, 79u149W 44u009N, 79u299W Various

Winter 2010–11

Winter 2011–12 Winter 2011–12

September 2011 September 2011

2011 2011 2011 (50)/spring 2012 (6) 2011

b

LaSalle Park Bluffers Park Aurora Others

43u189N, 79u519W 43u189N, 79u489W

42u579N, 81u269W ,45u129N, 75u519W

June June June June

Sampling period (n)

a

Total

Trumpeter Swan

Total

Mute Swan

Total

LaSalle Park Hamilton

Komoka Ottawab

Total

Mallard

43u339N, 80u139W ,43u399N, 79u229W 43u499N, 79u119W ,44u299N, 78u299W

GPS coordinates

Sampling location

Guelph Torontob Toronto Zoo Peterboroughb

Area

Canada Goose

Species

94 15 7 19 135

74 1 75

104 104c 208

50 48 56 52 206

Number of birds

26 7 2 7 42

0 0 0

83 20 103

7 12 0 20 39

HY

68 8 5 12 93

74 1 75

21 37 58

43 36 56 32 167

AHY

Age class

0 0 0 0 0

0 0 0

0 47 47

0 0 0 0 0

U

59 7 2 10 78

48 1 49

67 71 138

29 18 17 32 96

Male

35 8 5 7 55

26 0 26

37 23 60

21 27 37 20 105

Female

Sex

0 0 0 2 2

0 0 0

0 10 10

0 3 2 0 5

U

TABLE 1. Capture dates and locations and demographic information for Canada Geese (Branta canadensis), Mallards (Anas platyrhynchos), Mute Swans (Cygnus olor), and Trumpeter Swans (Cygnus buccinator) in southern Ontario, Canada from which blood samples and cloacal swabs were collected between October 2010 and May 2012 to determine the proportion of birds infected with avian bornavirus.a

DELNATTE ET AL.—AVIAN BORNAVIRUS IN WATERFOWL 515

516

JOURNAL OF WILDLIFE DISEASES, VOL. 50, NO. 3, JULY 2014

Goose that was shedding ABV were the negative and positive controls, respectively. Controls and determination of ELISA cutoff values

Setting cutoffs for determination of antibody testing was hindered by lack of a gold standard for determining ABV infection and a lack of known positive and negative birds for each species. Negative serum samples were obtained from 13 domestic geese and 13 domestic ducks (Anas platyrhynchos domesticus) that were part of a larger experimental infection trial (Delnatte et al. 2013b). The goslings and ducklings were obtained at 1 day of age and maintained in isolation. Blood collected at 40, 55, and 70 days (39 samples for each species) was evaluated for antibodies to ABV by ELISA. These groups of birds were determined to be free of ABV infection using histology, immunohistochemistry, and RTPCR on necropsy samples. Field samples were considered positive if the S/P ratio of the sample was higher than the mean S/P ratio for the negative control group plus 2 SD (Ojkic and Nagy 2003). Calculations were also performed using 3 SD to provide a second estimate of the proportion of antibody-positive birds with a lower sensitivity but higher specificity. Negative-control domestic geese were used for the calculation of ELISA cutoff values for the goose and swan samples, and negative-control domestic ducks were used for the Mallard samples. Nucleic acid isolation and RT-PCR

Cloacal swabs were evaluated for the presence of ABV RNA by real-time RT-PCR detection of the ABV-M gene. Total nucleic acids were extracted from 50-mL aliquots of inoculated virus transport medium using MagMAX-96 viral RNA isolation kits in a MagMAX Express-96 magnetic particle processor (Applied Biosystems Inc., Foster City, California, USA) according to manufacturer’s instructions. The RT-PCR assay was a duplex test with two sets of primers and TaqMan probes: 1) ABV_M_120201 – targeting psittacine ABV-M gene sequences; and 2) ABVG_M_111029 – targeting geese ABV-M gene sequences. The amplification was carried out in 25-mL reactions in a LightCycler 480 real-time PCR system (Roche, Laval, Quebec, Canada) using AgPath-ID one-step RT-PCR kits (Applied Biosystems). Primers and cycling parameters were previously described (Delnatte et al. 2013a). A RT-PCR crossing point value (Cp) less than 37.00 was considered positive.

Statistical analysis

Serology results were converted to an ordinal scale with three categories: negative, positive with the 2-SD seropositivity cutoff, and positive with the 3-SD cutoff. To investigate the effects of age, sex, location, shedding status, and species on antibody status, an ordinal logistic model was fitted with the ordinal serology variable as the outcome variable and age class, sex, species, shedding status, and any significant interactions as the explanatory variables. Since location was only available for the Canada Geese, a separate model was implemented with age, sex, and location as explanatory variables. To evaluate the overall difference of a species compared with the average of the other species, separate models with binary species variables were also implemented. The proportional odds assumption was validated by comparing the model with a cumulative categorical model using a likelihood ratio test. The presence of outliers and the fit of the model were evaluated graphically on residual plots. To perform multiple comparisons among species, the reference level of the species categorical variable was changed accordingly. When several models were tested on the same data set, a Bonferroni correction was applied for testing significance (Dohoo et al. 2009). An alpha of 0.05/n (n5number of comparisons) was used for significance. R (R 3.0.1, R Foundation for Statistical Computing, Vienna, Austria) and R-package ‘‘ordinal’’ (Christensen RHB, Ordinal–regression models for ordinal data R package, version 2012.09-11, http:// www.cran.r-project) were used for ordinal logistic regression models. Proportion of shedders was defined as the percentage of birds sampled for which the cloacal swab was positive by RT-PCR. Fisher’s exact test was used for comparisons of proportions of birds shedding ABV among species, among locations for Canada Geese, and between antibody-positive and antibodynegative birds (Graph-Pad QuickCalcs software, http://www.graphpad.com/quickcalcs/). An alpha of 0.05/n (n5number of comparisons) was again used for significance. Odds ratios (ORs) and Bonferroni-adjusted CIs (100% 2 0.05/n, n5number of comparisons) were not calculated because in all comparisons where the P value was significant, the observed probability in one group was 0. Unpaired t tests were used to compare the mean S/P ratios between shedders and nonshedders, and to compare the mean Cp values of shedding birds among species.

DELNATTE ET AL.—AVIAN BORNAVIRUS IN WATERFOWL

517

TABLE 2. Proportion of birds shedding avian bornavirus RNA as determined by real-time reversetranscription PCR on cloacal swabs and proportion of birds with antibodies against avian bornavirus by enzyme-linked immunosorbent assay in 206 Canada Geese (Branta canadensis), 208 Mallards (Anas platyrhynchos), 75 Mute Swans (Cygnus olor), and 135 Trumpeter Swans (Cygnus buccinator) caught in southern Ontario, Canada, October 2010 to May 2012.

Species

Canada Goose

Mallard Mute Swan Trumpeter Swan

Sampling location

Guelph Toronto Toronto Zoo Peterborough Overall Overall Overall Overall

Proportion shedding

Antibody prevalence a

%

No./total

N

0 0 13 0 3.4 0 9.3 0

0/50 0/48 7/56 0/52 7/206 0/208 7/75 0/135

50 48 53 52 203 92 75 130

Mean S/P ratiob

2 SDc (%)

3 SDd (%)

0.370 0.519 0.557 0.632 0.521 2.613 0.636 0.645

50 65 51 65 58 23 71 62

2.0 10 17 27 14 6.5 23 35

a

N 5 number of birds from which a blood sample was available.

b

S/P ratio 5 sample-to-positive ratio calculated as (sample mean of optical density 2 negative control mean)/(positive control mean 2 negative control mean).

c

Sample considered positive if its S/P ratio was higher than the mean S/P ratio of the negative control group plus 2 SD.

d

Sample considered positive if its S/P ratio was higher than the mean S/P ratio of the negative control group plus 3 SD.

RESULTS Presence of antibodies

The ELISA results are summarized in Table 2. We analyzed 500 serum samples. The overall prevalence of antibodies to ABV was 58% (117/203) of Canada Geese, 71% (53/75) of Mute Swans, 62% (81/130) of Trumpeter Swans, and 23% (21/92) of Mallards using the 2-SD cutoff and 14% (29/203) of Canada Geese, 23% (17/75) of Mute Swans, 35% (46/130) of Trumpeter Swans, and 6.5% (6/92) of Mallards using the 3-SD cutoff. The S/P ratios of individual samples by species (Fig. 2A) and by location for Canada Geese (Fig. 2B) are presented with both 2-SD and 3-SD cutoffs. The S/ P ratios ranged from 1.024 to 5.288 with a mean6SD of 0.52160.474 for Canada Geese, 0.63660.478 for Mute Swans, 0.64560.626 for Trumpeter Swans, and 2.61361.09 for Mallards. Antibodies were found in all four species and at each location. Applying the conservative assessments of significance as outlined above, interaction terms were not significant and thus comparisons were made using the entire model, except for location for Canada Geese. The antibody prevalence

for Mallards was significantly lower than for the three other species (Table 3), and the prevalence for Trumpeter Swans was significantly higher than for Canada Geese (Table 3). There was no effect of age or sex in the overall model. Geese at Peterborough had higher odds of being antibody positive than those from Guelph (P50.009; OR52.73 with a 95% CI51.29–5.79). Cloacal shedding

The RT-PCR results are presented in Table 2. The overall prevalence of ABV RNA shedding was 3.4% (7/206) in Canada Geese and 9.3% (7/75) in Mute Swans. None of the cloacal swabs from the 208 Mallards or 135 Trumpeter Swans was positive. Mallards were significantly less likely to shed viral RNA than Canada Geese (P50.0072) or Mute Swans (P,0.0001), and Trumpeter Swans less likely than Mute Swans (P50.0006). The prevalence of shedding was not significantly different between Canada Geese and Mute Swans (P50.06). The proportion of geese caught on the zoo that were shedding ABV RNA was 13% (7/ 56), which was significantly higher than the proportions (0%) of geese sampled at Guelph (P50.0136), Toronto (P50.0144), and Peterborough (P50.0131).

518

JOURNAL OF WILDLIFE DISEASES, VOL. 50, NO. 3, JULY 2014

FIGURE 2. Sample-to-positive (S/P) ratio of serum samples assessed for antibodies against avian bornavirus (ABV) using an enzyme-linked immunosorbent assay (ELISA) for 203 Canada Geese (Branta canadensis), 75 Mute Swans (Cygnus olor), 130 Trumpeter Swans (Cygnus buccinator), and 92 Mallards (Anas platyrhynchos) in southern Ontario, Canada. A sample was considered positive if the S/P ratio ([sample mean of optical density 2 negative control mean]/[positive control mean 2 negative control mean]) from the ELISA was higher than the

DELNATTE ET AL.—AVIAN BORNAVIRUS IN WATERFOWL

519

TABLE 3. Results of post hoc comparisons using an ordinal logistic regression model to investigate the effect of species on antibody status as part of a prospective survey of avian bornavirus infection in free-ranging waterfowl in southern Ontario, Canada. Comparison (among species)

P valuea

CAGO vs. MUSWc CAGO vs. TRUS MALL vs. CAGO MALL vs. MUSW MALL vs. TRUS MUSW vs. TRUS

0.041 0.006* ,0.001* ,0.001* ,0.001* 0.73

a

Odds ratio (99.98% CIb)

1.65 1.81 4.06 6.69 7.32 1.09

(0.92–2.97) (1.08–3.02) (2.08–7.95) (3.11–14.39) (3.57–15.01) (0.57–2.08)

Bonferroni correction was applied for level of significance: results were significant (*) when P,0.0167 (0.05/n, where n 5 the number of comparisons).

b

Bonferroni-corrected confidence interval (CI; CI510020.05/n, where n 5 the number of comparisons).

c

CAGO 5 Canada Goose (Branta canadensis); MALL 5 Mallard (Anas platyrhynchos); MUSW 5 Mute Swan (Cygnus olor); TRUS 5 Trumpeter Swan (Cygnus buccinator).

* Statistically significant.

The Cp values of the 14 PCR-positive birds ranged from 24.61 to 36.89 with a mean and SD of 31.9563.53. The mean Cp for the seven positive goose samples (29.67) was significantly lower than the mean Cp for the seven positive Mute Swan samples (34.23; P50.0087), indicating shedding of greater amounts of viral RNA in the geese. Correlation between cloacal shedding and presence of antibodies

Shedding status had the strongest effect on the OR of antibody status in the overall logistic model, with birds shedding ABV RNA having a greater probability of being antibody positive (P,0.001; OR510.38 with a 95% CI53.15–34.18). The S/P ratio of each serum sample was also plotted against the shedding status of the Canada Goose (Fig. 2C) or Mute Swan (Fig. 2D) from which it was obtained. All Canada Geese that were shedding ABV RNA were antibody positive using both ELISA cutoff values. Six of seven Mute

Swans that were shedding ABV RNA were antibody positive using the 2-SD cutoff; however, only three of seven (43%) were positive using the 3-SD cutoff. Groups of birds that were shedding ABV RNA had significantly higher mean S/P ratios than those that were not shedding: 1.686 vs. 0.480 (P,0.0001) for Canada Geese (all locations considered together); 1.686 vs. 0.385 (P,0.0001) for geese from the zoo only; and 1.058 vs. 0.592 (P50.0013) for Mute Swans. DISCUSSION

We provide the first estimates of the prevalence of cloacal shedding of ABV RNA and prevalence of antibodies to ABV in four species of waterfowl in Ontario. Our results suggest that exposure to ABV is widespread in Ontario in the species tested, but that infection rates and the biology of the disease vary among species and locations. Antibodies were found in all species and at all locations with preva-

r mean S/P ratio for the negative control group plus 2 SD. An alternative cutoff value used 3 SD above the negative control. The +2SD cutoff is indicated by a solid line and the +3SD by a dashed line. (A) Results for all four waterfowl species, showing cutoff values for ducks, and for swans and geese. (B) Results for Canada Geese by sampling location. (C) Results for Canada Geese that were shedding ABV RNA in cloacal swab samples vs. those that were not. (D) Results for Mute Swans that were shedding ABV RNA in cloacal swab samples vs. those that were not.

520

JOURNAL OF WILDLIFE DISEASES, VOL. 50, NO. 3, JULY 2014

lences ranging from 6.5% to 71%, depending on species and location and on how positive samples were determined. Because no similar studies have been conducted for ABV in free-ranging birds, it is difficult to comment on these figures. In a large-scale study conducted in captive psittacine birds in Europe, 17% of the birds had detectable antibodies (HeffelsRedmann et al. 2011). Despite our efforts to conduct an epidemiologic study that would best reflect the prevalence of cloacal shedding and of antibodies at a specific time, several practical constraints (especially related to the choice of sampling sites) were limiting factors in our study design. Widespread persistent asymptomatic infection with less common clinical disease is a consistent feature of bornavirus infections, and may be attributed to the lack of cytopathic effect of bornaviruses and their ability to escape recognition by the innate immune system (Reuter et al. 2010). Although ABV infection has been associated with significant neuropathology in waterfowl (Delnatte et al. 2013a), poor body condition or neurologic deficits were not noted in any birds in this study. Although antibodies to ABV were widespread in all four species of birds, only 9.3% of Mute Swans, 3.4% of Canada Geese, and none of the Trumpeter Swans or Mallards were shedding ABV RNA at the time of sampling. These shedding rates are consistent with those recently reported in American studies, where 2.9% (12/409) of Canada Geese (Payne et al. 2011b) and 6% (14/219) of Mute Swans (Payne et al. 2012) were positive using RTPCR on oropharyngeal/cloacal swabs. However, the difference in shedding rates among species, especially between Trumpeter and Mute Swans, was surprising given that these species commingle and that Trumpeter Swans showing clinical disease associated with ABV infection had previously been identified in the sampled flock. A species-based difference in the biology of ABV infection is suggested.

Further studies are required to verify whether a true difference in shedding frequency exists in natural and, ideally, experimental infections. The use of cloacal shedding of ABV RNA to estimate infection with ABV in waterfowl was based on current recommendations for the investigation of ABV status in psittacine birds (Hoppes et al. 2013), and is supported by work in parrots describing the shedding of ABV RNA in experimentally infected birds (Piepenbring et al. 2012) and the isolation of viable ABV virus from cloacal swabs that were also positive by RT-PCR (Rubbenstroth et al. 2012). Cloacal sampling; however, likely underestimates the prevalence of ABV infection. Rates of viral shedding might be expected to vary with stage of infection, being higher soon after initial infection or in association with viral reactivation and active replication. Intermittent urofecal shedding of ABV is described in both experimentally and naturally-infected parrots (Raghav et al. 2010; Heatley and Villalobos 2012; Piepenbring et al. 2012), and a similar pattern is suspected in waterfowl (Delnatte et al. 2013a). Technical aspects of assessment could also result in underestimation of viral shedding in feces. Despite using a duplex RT-PCR with primers targeting both psittacine and waterfowl ABV sequences, it is still possible that some undiscovered ABV genotypes remained undetected. Fecal matter can also contain RT-PCR inhibitors that would lead to false-negative results. Although the testing of a single cloacal swab is of limited usefulness in determining the ABV infection status of an individual bird, this method remains a useful tool to screen for the presence of active ABV infection on a flock scale. Pooling multiple droppings collected over several days, the current recommendation for psittacine birds (Hoppes et al. 2013), is unrealistic for studies of free-ranging birds. Parallel assessment of viral RNA in other biological samples, such as choanal swabs or

DELNATTE ET AL.—AVIAN BORNAVIRUS IN WATERFOWL

feather calami, would likely increase the sensitivity of detection of infection. Birds that were shedding ABV RNA had higher mean S/P ratios compared with nonshedding birds; however, not all birds shedding viral RNA were antibody positive. Possible explanations for this include the presence of acutely infected birds that had not yet seroconverted, that some individuals simply do not seroconvert, or that, contrary to our underlying assumption, the virus in feces was ingested on contaminated material and was simply passing through the digestive tract. Failure of antibody detection by a serologic test based on a single antigen could also lead to false-negative ELISA results. However, the presence of ABV-shedding, antibodynegative parrots has been described despite using infected cell culture as the source of multiple presenting antigens in an indirect immunofluorescence assay (Herzog et al. 2010). The discrepancy between viral shedding, presence of antibodies, and clinical disease is now well recognized in parrots infected with ABV (Villanueva et al. 2010; Heffels-Redmann et al. 2012; Piepenbring et al. 2012). Viral shedding and presence of antibodies coincided in only one-fifth of the samples in a large study of captive psittacines (Heffels-Redmann et al. 2011). Our research supports previous indications that infection rates vary according to location (Payne et al. 2012; Delnatte et al. 2013a). The proportion of shedding geese caught on the zoo site was 13% compared with 0% of geese sampled at three other locations. Although Canada Geese in Ontario are considered to belong to one migratory metapopulation, local factors might influence the presence and expression of disease. Delnatte et al. (2013a) found that the proportion of birds with a clinical history or pathologic lesions suggestive of ABV infection was significantly higher at the Toronto Zoo (30/132) than elsewhere in Ontario (21/823). This correlates with the higher proportion of shedders at the zoo and supports the

521

premise that environmental or ecologic factors may affect ABV shedding and possibly the onset of disease. The persistently high population density of geese at the zoo, and the fact that most of these birds remain year-round in this attractive geographically restricted area, likely enhance the transmission of ABV leading to a disease cluster. Historically, serologic assays have not been convincing for diagnosing BDV infections of mammals (Staeheli et al. 2010). However, unlike BDV, ABV in affected parrots shows broad tissue tropism and ABV antigens are abundant in affected organs (Gancz et al. 2009; Rinder et al. 2009; Raghav et al. 2010; Piepenbring et al. 2012). This suggests that the immune response to ABV may be stronger than to BDV. Serologic assays described for the identification of antibodies to ABV include indirect immunofluorescence assay (Herzog et al. 2010), Western blot, and indirect ELISAs with a variety of sources of primary antigen (Hoppes et al. 2013). The N-protein appears immunodominant (Lierz et al. 2009; Villanueva et al. 2010), thus our choice to use a purified M24N protein as the source of antigen. Assessment of the specificity and sensitivity of serologic tests is difficult due to differences in methods, the poor understanding of the epidemiology of the disease, the absence of gold standards for identifying infection in live birds, the difficulty in identifying a true control group of uninfected birds, the variability among species, and the relatively poor correlation between viral shedding, pathology, and presence of antibodies. We encountered these same difficulties in developing and interpreting our ELISA, in addition to those related to working with free-ranging, nondomestic species. To design an assay independent of the species tested, we used an anti-bird IgG antibody conjugate designed for diverse avian species and used previously to detect antibodies to ABV in avian sera (de Kloet et al. 2011). The reactivity of this conjugate

522

JOURNAL OF WILDLIFE DISEASES, VOL. 50, NO. 3, JULY 2014

toward immunoglobulins for each species studied was initially verified by dot blots (P.D. unpubl. data). The intensity of OD405 readings in our ELISA varied greatly among species, with values being higher for ducks (including the negativecontrol domestic ducks) than for the geese and swans. This is most likely due to nonspecific binding of the secondary antibodies, the intensity of which appeared to be species specific. This precluded direct comparison of S/P ratios among species and highlighted the need for species-specific secondary antibodies and ELISA cutoff values. In conclusion, this prospective study demonstrated that antibodies directed against ABV are widespread in freeranging Canada Geese, Trumpeter Swans, Mute Swans, and Mallards in Ontario. We also confirmed that viral RNA can be shed in the urofeces of asymptomatic Mute Swans and Canada Geese. Despite their respective limitations, serologic assays and RT-PCR testing of cloacal swabs can be used to estimate the prevalence of ABV infection at a flock level. Because ABV can also be associated with significant pathology, the overall impact of ABV on waterfowl populations remains uncertain. Avian bornavirus infection is likely underrecognized in free-ranging waterfowl and, given the migratory nature of the host species, is unlikely to be restricted to North America. ACKNOWLEDGMENTS

We thank the Ontario Trumpeter Swan Reintroduction Program (Harry Lumsden, Bev and Ray Kingdon, Julie Kee, and Kyna Intini), the Canadian Wildlife Service (James Vanos [CWS London & Guelph] and Christopher Sharp [CWS Ottawa]), Ontario Ministry of Natural Resources (OMNR) (Rob Brook [OMNR Peterborough]), Toronto Region Conservation Authority (Danny Moro), Toronto Zoo staff, Golf Glen Veterinary Clinic (Aurora, Ontario), and many summer students for assistance in collecting samples. We also thank Lenny Shirose (Canadian Cooperative Wildlife Health Center [CCWHC]–Ontario), Sarah Hoyland, Jane Coventry and Elizabeth

Hillyer (Animal Health Laboratory), Nicole Zaranek and Veronica Kay for their help. A special thank you to Betty-Anne McBey (Pathobiology, Virology laboratory) for technical assistance in the early stage of the ELISA design, and Elizabeth Beck and Hugues Beaufre`re for statistical analysis using R. The authors gratefully acknowledge Susan Payne, Texas A&M, for providing the pET21a + M24N clone. Financial support was provided by the Toronto Zoo, the CCWHC-Ontario and the OVC Pet Trust. This work was carried out in partial fulfilment of a DVSc degree by P.D., whose stipend was provided by the Toronto Zoo. LITERATURE CITED Berg M, Johansson M, Montell H, Berg AL. 2001. Wild birds as a possible natural reservoir of Borna disease virus. Epidemiol Infect 127:173– 178. de Kloet AH, Kerski A, de Kloet SR. 2011. Diagnosis of avian bornavirus infection in Psittaciformes by serum antibody detection and reverse transcription polymerase chain reaction assay using feather calami. J Vet Diagn Invest 23:421–429. de Kloet SR, Dorrestein GM. 2009. Presence of avian bornavirus RNA and anti-avian bornavirus antibodies in apparently healthy macaws. Avian Dis 53:568–573. Delnatte P. 2013b. Avian bornavirus infection in waterfowl. DVSc Thesis, University of Guelph, Guelph, Canada, 196 pp. Delnatte P, Berkvens C, Kummrow M, Smith DA, Campbell D, Crawshaw G, Ojkic D, DeLay J. 2011. New genotype of avian bornavirus in wild geese and Trumpeter Swans in Canada. Vet Rec 169:108. Delnatte P, Ojkic D, Delay J, Campbell D, Crawshaw G, Smith DA. 2013a. Pathology and diagnosis of avian bornavirus in Canada Geese (Branta canadensis), Trumpeter Swans (Cygnus buccinator) and Mute Swans (Cygnus olor) in Canada: A retrospective study. Avian Pathol 42:114–128. Dohoo IR, Martin SW, Stryhn H. 2009. Veterinary epidemiologic research, 2nd Ed. AVC Inc., Prince Edward Island, Canada, 706 pp. Gancz AY, Kistler AL, Greninger AL, Farnoushi Y, Mechani S, Perl S, Berkowitz A, Perez N, Clubb S, DeRisi JL. 2009. Experimental induction of proventricular dilatation disease in Cockatiels (Nymphicus hollandicus) inoculated with brain homogenates containing avian bornavirus 4. Virol J 6:100. Gancz AY, Clubb S, Shivaprasad HL. 2010. Advanced diagnostic approaches and current management of proventricular dilatation disease. Vet Clin North Am Exot Anim Pract 13:471–494.

DELNATTE ET AL.—AVIAN BORNAVIRUS IN WATERFOWL

Guo J, Covaleda L, Heatley J, Baroch J, Tizard I, Payne S. 2012. Widespread avian bornavirus infection in Mute Swans in the northeast United States. Vet Med Res Rep 3:49–52. Heatley JJ, Villalobos AR. 2012. Avian bornavirus in the urine of infected birds. Vet Med Res Rep 3:19–23. Heffels-Redmann U, Enderlein D, Herzog S, Herden C, Piepenbring A, Neumann D, Mu¨ller H, Capelli S, Mu¨ller H, Oberha¨user K. 2011. Occurrence of avian bornavirus infection in captive psittacines in various European countries and its association to proventricular dilation disease. Avian Pathol 40:419–426. Heffels-Redmann U, Enderlein D, Herzog S, Piepenbring A, Bu¨rkle M, Neumann D, Mu¨ller H, Capelli S, Mu¨ller H, Oberha¨user K. 2012. Follow-up investigations on different courses of natural avian bornavirus infections in psittacines. Avian Dis 56:153–159. Herzog S, Enderlein D, Heffels-Redmann U, Piepenbring A, Neumann D, Kaleta EF, Muller H, Lierz M, Herden C. 2010. Indirect immunofluorescence assay for intravitam diagnostic of avian bornavirus infection in psittacine birds. J Clin Microbiol 48:2282–2284. Hoppes S, Tizard I, Shivaprasad HL. 2013. Avian bornavirus and proventricular dilatation disease: Diagnostics, pathology, prevalence and control. Vet Clin North Am Exot Anim Pract 16:339–355. Lierz M, Hafez HM, Honkavuori K, Gruber AD, Olias P, Abdelwhab EM, Kohls A, Lipkin IW, Briese T, Hauck R. 2009. Anatomic distribution of avian bornavirus in parrots, its occurrence in clinically healthy birds and ABV-antibody detection. Avian Pathol 38:491–496. Novagen. 2003. Novagen pET system manual, 10th Ed. http://richsingiser.com/4402/Novagen%20pET%20 system%20manual.pdf. Accessed March 2014. ´ . 2003. Antibody response and virus Ojkic D, Nagy E tissue distribution in chickens inoculated with wild-type and recombinant fowl adenoviruses. Vaccine 22:42–48. Payne S, Covaleda L, Jianhua G, Swafford S, Baroch J, Ferro PJ, Lupiani B, Heatley J, Tizard I. 2011a. Detection and characterization of a distinct bornavirus lineage from healthy Canada Geese (Branta canadensis). J Virol 85:12053–12056.

523

Payne S, Shivaprasad HL, Mirhosseini N, Gray P, Hoppes S, Weissenbo¨ck H, Tizard I. 2011b. Unusual and severe lesions of proventricular dilatation disease in Cockatiels (Nymphicus hollandicus) acting as healthy carriers of avian bornavirus (ABV) and subsequently infected with a virulent strain of ABV. Avian Pathol 40:15–22. Payne SL, Delnatte P, Guo J, Heatley JJ, Tizard I, Smith DA. 2012. Birds and bornaviruses. Anim Health Res Rev 13:145–156. Piepenbring AK, Enderlein D, Herzog S, Kaleta EF, Heffels-Redmann U, Ressmeyer S, Herden C, Lierz M. 2012. Pathogenesis of avian bornavirus in experimentally infected cockatiels. Emerg Infect Dis 18:234–241. Raghav R, Taylor M, DeLay J, Ojkic D, Pearl DL, Kistler AL, DeRisi JL, Ganem D, Smith DA. 2010. Avian bornavirus is present in many tissues of psittacine birds with histopathologic evidence of proventricular dilatation disease. J Vet Diagn Invest 22:495–508. Reuter A, Ackermann A, Kothlow S, Rinder M, Kaspers B, Staeheli P. 2010. Avian bornaviruses escape recognition by the innate immune system. Viruses 2:927–938. Rinder M, Ackermann A, Kempf H, Kaspers B, Korbel R, Staeheli P. 2009. Broad tissue and cell tropism of avian bornavirus in parrots with proventricular dilatation disease. J Virol 83:5401–5407. Rubbenstroth D, Rinder M, Kaspers B, Staeheli P. 2012. Efficient isolation of avian bornaviruses (ABV) from naturally infected psittacine birds and identification of a new ABV genotype from a salmon-crested cockatoo (Cacatua moluccensis). Vet Microbiol 161:36–42. Staeheli P, Rinder M, Kaspers B. 2010. Avian bornavirus associated with fatal disease in psittacine birds. J Virol 84:6269–6275. Villanueva I, Gray P, Mirhosseini N, Payne S, Hoppes S, Honkavuori KS, Briese T, Turner D, Tizard I. 2010. The diagnosis of proventricular dilatation disease: Use of a Western blot assay to detect antibodies against avian bornavirus. Vet Microbiol 143:196–201. Submitted for publication 17 August 2013. Accepted 25 February 2014.