Antibodies to Type A Influenza Virus in Wild Waterbirds from Argentina

Abstract

Limited information exists on avian influenza (AI) virus infection in South American wild birds. As part of a national surveillance program in Argentina, indigenous waterbirds were screened for antibodies to AI virus. From November 2006 to July 2007, serum samples from 540 waterbirds of 12 species were tested for type-specific antibodies to AI virus with the use of a commercially available blocking enzyme-linked immunosorbent assay (bELISA) and the agar-gel immunodiffusion (AGID) test. Thirty-three percent (176/540) of serum samples were positive with the bELISA and 12% (64/540) were positive with the AGID test. The bELISA detected antibodies to AI virus in eight of the 12 species, and the AGID detected positives in only five species. These results provide insight into AI virus circulation in Argentinean waterbirds and preliminary data to guide further surveillance efforts.

Avian influenza (AI) viruses have a global distribution in wild birds, with virus isolated or antibodies to AI detected on all seven continents (Stallknecht et al., 2007). Our understanding of AI virus epidemiology, however, is highly variable among different geographic regions. Extensive data on patterns of AI virus infection in wild birds exist for Europe and North America (Munster et al., 2007; Stallknecht and Brown, 2007), where the majority of surveillance has traditionally been focused, but similar information for many parts of the Southern hemisphere is lacking.

There are limited surveillance data on AI virus infection in South American wild birds. Antibodies to AI virus were not detected in Rockhopper Penguins (Eudyptes chrysocomes) in Argentina (Karesh et al., 1999); Waved Albatross (Poebastria irrorata), or Flightless Cormorants (Phalacrocorax harrisi) in the Galapagos Islands (Padilla et al., 2003; Travis et al., 2006); or Southern Giant Petrels (Macronectes giganteus) in Argentina (Uhart et al., 2003). No AI viruses were detected, with the use of reverse-transcriptase polymerase chain reaction (RT-PCR), in cloacal swabs collected from 156 Red Knots (Calidris canutus rufa) in Argentina (D’Amico et al., 2007). Nine AI viruses were isolated from 2,405 environmental fecal samples collected at waterbird flocks along the central coast of Peru (Ghersi et al., 2009). The nine AI viruses represented multiple hemagglutinin (HA) subtypes (H3, H4, H10, and H13) and were isolated from single-species flocks consisting of Ruddy Turnstones (Arenaria interpres), American Oystercatchers (Haematopus palliatus), Peruvian Pelicans (Pelecanus occidentalis thagus), Whimbrels (Numenius phaeopus), Kelp Gulls (Larus dominicanus), or a mixture of White-cheeked Pintails (Anas bahamensis) and Cinnamon Teal (Anas cyanoptera). At the time of writing, the only published isolations of AI virus directly from wild birds in South America consist of an H7N3 low-pathogenic avian influenza (LPAI) virus from a Cinnamon Teal in Bolivia (Spackman et al., 2006) and an H13N9 LPAI virus from a Kelp Gull in Argentina (Pereda et al., 2008). The later H13N9 LPAI virus was isolated during 2 yr of AI surveillance in Argentina, in which 2,895 cloacal swabs were collected from 56 avian species spanning multiple geographic regions. Eleven additional samples collected during this project were positive for AI virus genetic material via RT-PCR, but virus was not isolated. These RT-PCR–positive/virus-isolation–negative samples were collected from Brazilian Ducks (Amazonetta brasiliensis), Ringed Teal (Callonetta leucophrys), Rosy-billed Pochards (Netta peposaca), and Silver Teal (Anas versicolor) (A. Pereda, pers. comm.).

Virus-detection–based surveillance, including both virus isolation and nucleic acid detection, is sensitive to spatial and temporal variation in AI virus prevalence among wild avian populations. Because antibodies to AI virus can persist after viral shedding is no longer detectable (Spackman et al., 2009), serologic testing may provide an easy method to obtain evidence of previous infection with these viruses. Such information subsequently can be utilized to identify species potentially involved in AI transmission or maintenance that can be targeted for additional surveillance. In this study, we evaluated this surveillance approach by screening multiple waterbird species in Argentina for antibodies to AI virus.

During November 2006–July 2007, 540 serum samples were collected from 12 wild waterbird species in Santa Fe province, Argentina (Table 1). Corresponding virus isolation and RT-PCR results from the birds included in this study have been reported (Pereda et al., 2008; Table 1). The vast majority of the serum samples were collected from species in the order Anseriformes (ducks, geese, and swans); however, two samples were from White-winged Coots (Fulica leucoptera) representing the order Gruiformes.

TABLE 1.

Prevalence of antibodies to avian influenza (AI) viruses in wild waterbird species from Santa Fe province, Argentina. All birds were serologically tested with the agar-gel immunodiffusion (AGID) and a commercial blocking enzyme-linked immunosorbent assay (bELISA), and tested for AI virus genetic material by reverse-transcriptase polymerase chain reaction (RT-PCR).^a

Species	n	Time of sampling	AGID pos/total (%)	bELISA pos/total (%)	RT-PCR pos/totalb (%)
White-faced Tree Duck (Dendrocygna viduata)	9	May–July 2007	0/9 (0)	0/9 (0)	0/9 (0)
Black-bellied Tree Duck (Dendrocygna autumnalis	1	May 2007	0/1 (0)	0/1 (0)	0/1 (0)
Fulvus Whistling Duck (Dendrocygna bicolor)	45		1/45 (2)	12/45 (27)	0/45 (0)
	14	January–February 2007	1/14 (7)	5/14 (36)	0/14 (0)
	31	May–July 2007	0/31 (0)	7/31 (23)	0/31 (0)
Comb Duck (Sarkidiornis melanotos)	1	May 2007	0/1 (0)	0/1 (0)	0/1 (0)
Brazilian Duck (Amazonetta brasiliensis)	5	July 2007	0/5 (0)	1/5 (20)	0/5 (0)
Ringed Teal (Callonetta leucophrys)	6	May 2007	0/6 (0)	3/6 (50)	0/6 (0)
Chilean Wigeon (Anas sibilatrix)	1	May 2007	0/1 (0)	1/1 (100)	0/1 (0)
Speckled Teal (Anas flavirostris)	1	November 2006	0/1 (0)	1/1 (100)	0/1 (0)
Silver Teal (Anas versicolor)	13	May–July 2007	2/13 (15)	7/13 (54)	0/13 (0)
Rosy-billed Pochard (Netta peposaca)	455		59/455 (13)	150/455 (33)	2/455 (<1)c
	97	January–February 2007	28/97 (29)	75/97 (77)	0/97 (0)
	358	May–July 2007	31/358 (9)	75/358 (21)	2/358 (1)
Black-headed Duck (Heteronetta atricapilla)	1	January 2007	1/1 (100)	1/1 (100)	0/1 (0)
White-winged Coot (Fulica leucoptera)	2	January 2007	1/2 (50)	0/2 (0)	0/2 (0)
Total	540		64/540 (12)	176/540 (33)	2/540 (<1)

^apos = positive.

^bRT-PCR testing of cloacal swab samples as reported by Pereda et al. (2008).

^cVirus isolation attempts, in 10-day-old embryonating chicken eggs, on both RT-PCR–positive samples, were negative.

Most serum samples were collected from the carcasses of hunter-killed ducks donated by licensed hunters during the appropriate season (February–July). In addition, a small number of live birds were captured with mist nets (n=21) outside of the designated hunting season (November 2006, January 2007), including White-winged Coots (n=2), a Black-headed Duck (Heteronetta atricapilla; n=1), Rosy-billed Pochards (n=9), a Speckled Teal (Anas flavirostris; n=1), and Fulvous Whistling Ducks (Dendrocygna bicolor; n=8). Hunter-killed ducks were processed within 2–4 hr of death, and blood was collected directly from the heart. Live ducks were sampled by venipuncture of the brachial vein with heparinized 5-ml syringes and 25-G needles. Total blood volume collected was below 0.7% of the host bird body weight. All blood samples were centrifuged at 1,000×G within 4 hr of collection. Serum was transferred to cryovials and stored in liquid nitrogen at the field site. Samples were later transferred to a −20 C freezer and exported to the USA on dry ice. Export permits were obtained from the Argentina veterinary service (Servicio Nacional de Sanidad Animal y Calidad Agroalimentaria), and the wildlife service (Dirección Nacional de Fauna).

Serologic testing for type-specific antibodies to AI virus was performed with the use of the agar-gel immunodiffusion (AGID) test and a commercially available blocking enzyme-linked immunosorbent assay (bELISA) kit (FlockCheck AI MultiS-Screen Antibody Test Kit, IDEXX Laboratories, Westbrook, Maine, USA). The AGID testing was performed using standard procedures (Swayne et al., 2008) and reagents supplied by the National Veterinary Services Laboratories, U.S. Department of Agriculture, Animal and Plant Health Inspection Service, Ames, Iowa, USA. The bELISA testing followed the guidelines provided by the manufacturer.

The bELISA detected antibodies to AI virus in 33% (176/540) of the serum samples, which included samples from eight of the 12 species (Table 1). The AGID test was less sensitive than the bELISA and detected antibodies in 12% (64/540) of samples, representing five of the 12 species. All AGID-positive serum samples were also positive on the bELISA, except for one sample from a White-winged Coot and one from a Rosy-billed Pochard sampled in July 2007; both these samples were AGID-positive and bE-LISA-negative. Two species, Fulvus Whistling Ducks and Rosy-billed Pochards, were sampled at a single geographic location over multiple seasons from January to July 2007 (summer–winter), and antibody-positive ducks were detected at each surveillance effort.

Cloacal swabs from two of the 540 birds sampled in this study previously were found positive for AI virus by RT-PCR (Pereda et al., 2008; Table 1). Both RT-PCR–positive samples were collected from Rosy-billed Pochards in May 2007 from Santa Fe province. Antibodies to AI virus were detected in Rosy-billed Pochards at this time and location (bELISA: 27% [62/231]; AGID: 11% [26/231]), but the two individual RT-PCR–positive birds were antibody-negative with both assays; presumably both these birds were recently infected and had not yet seroconverted.

The potential involvement of migratory waterfowl in the emergence and spread of H5N1 highly pathogenic avian influenza (HPAI) viruses across Eurasia and Africa has highlighted the need to better understand the epidemiology and natural history of AI virus in wild birds. Consequently, an enormous amount of resources recently have been dedicated toward wild bird AI surveillance in order to define global trends of AI infection and search for potential new avian reservoirs. In this current study, antibodies to AI virus were detected in eight wild waterbird species indigenous to Argentina. To our knowledge, AI virus has not reportedly been isolated or detected from wild populations of several of these species, including Fulvous Whistling Ducks, Ringed Teal, Chilean Wigeons (Anas sibilatrix), Speckled Teal, Black-headed Ducks, and White-winged Coots. This largely may be a result of the limited surveillance conducted on the avian taxa included in this study. These positive serology results, however, correspond to the existing general knowledge of AI host range and viral transmission in wild bird populations (Olsen et al., 2006; Stallknecht et al., 2007). AI viruses have been isolated from a wide diversity of mammalian and avian hosts, but aquatic birds in the orders Anseriformes and Charadriiformes currently comprise the natural reservoir for all known viral subtypes (Olsen et al., 2006; Stallknecht et al., 2007). Globally, most reported isolations of AI virus have been from ducks (family Anatidae, subfamily Anatinae), but AI viruses are frequently detected in other sympatric aquatic birds, including other genera of waterfowl, cormorants, herons, egrets, coots, waders, shorebirds, and gulls (Munster et al., 2007; Stallknecht et al., 2007).

Relatively few surveillance studies have determined the prevalence of type-specific antibodies to AI virus in wild avian populations, and this lack of data limits the extent to which we can compare or interpret the results presented herein. The AGID test produces inconsistent results in waterfowl hosts and, consequently, has been considered to be a serologic assay with limited use in wild-bird AI surveillance (Stallknecht et al., 2007). Multiple studies have utilized various indirect or bELISAs to detect type-specific antibodies to AI viruses in wild waterbirds from Spain (Arenas et al., 1990; Astorga et al., 1994), Italy (De Marco et al., 2003), and the USA (Sullivan et al., 2009). In this study, antibodies to AI virus were detected in the majority of Argentinean waterfowl species that were tested with the bELISA (Table 1). Anti-body-prevalence estimates reported here-in approximate those reported for North American dabbling duck species using the same serologic assays (Brown et al., 2010). This suggests that South America wild duck populations routinely are infected with AI viruses and justifies additional surveillance to recover these viruses for further characterization.

The value of the data presented herein is twofold. The presence of antibodies to AI virus in multiple duck species suggests that AI viruses likely are endemic in South American wild waterfowl. This is consistent with trends reported in North America and Eurasia and the existing global knowledge of AI ecology in wild birds. Additional surveillance is needed to further identify important avian host species in different regions of South America, define the spatial and temporal patterns of infection, and determine the subtype diversity maintained in the wild avian reservoir system. The second benefit of this research is as a proof of function for the use of serology in wild bird AI surveillance. Serology is a commonly utilized diagnostic approach for AI detection in domestic poultry and for a variety of wildlife disease surveillance programs. Traditionally, the lack of a valid species-independent type-specific assay has greatly limited the utility of serology in wild-bird AI surveillance. However, with the recent development of commercially available bELISA kits, such as the one used in this study, serology represents an additional diagnostic tool available for wild-bird AI surveillance, which can provide unique and potentially valuable insight into AI virus ecology and epidemiology.

Funding for this work was provided by the National Institute of Allergy and Infectious Diseases, National Institutes of Health (NIH), Department of Health and Human Services, under Contract HHSN266200700007C. This work was additionally supported through the Global Avian Influenza Network for Surveillance (GAINS) program, funded in part by USAID Grant LAG-A-00–99-00047–00. The opinions expressed herein are those of the authors and do not necessarily reflect the views of any of the funding agencies. Students from the Veterinary Schools of Casilda (Universidad Nacional de Rosario), Esperanza (Universidad Nacional del Litoral), Buenos Aires (Universidad Nacional de Buenos Aires), and Tandil (Universidad Nacional del Centro de la Provincia de Buenos Aires) provided valuable assistance in sample collection.