Comparative Pathogenic Potential of Avian Influenza A/H9N2 Viruses Isolated from Commercial, Backyard and Fancy Birds

Muhammad Furqan Shahid1*, Tahir Yaqub1, Muhammad Yasin Tipu2, Asim Aslam2, Saima Yaqub1, Aziz-ul-Rahman3, Muzaffar Ali1

1Department of Microbiology, University of Veterinary and Animal Sciences, Lahore, Pakistan

2Department of Pathology, University of Veterinary and Animal Sciences, Lahore, Pakistan

3Quality operations laboratory, University of Veterinary and Animal Sciences, Lahore, Pakistan

Abstract | Poultry industry is confronting numerous viral infections particularly low pathogenic avian influenza virus (LPAIV) H9N2, which can bring exceptional abatement in trade of poultry products in endemic countries such as Pakistan. The current study was conducted to evaluate the infectious and antigenic potential of three different AIV (A/H9N2) variants isolated from commercial, backyard poultry and fancy birds (partridge). Commercial broiler birds were infected experimentally at the dose of 106 EID50 by using these three isolates. Results revealed that onset of clinical signs was started at 2nd day post infection (DPI) followed by severity in redness and swelling of eyes, sneezing and lacrimation at 3rd DPI that lasts up to 5th DPI. Cross reactivity of antiserum response toH9N2 was compared among reference isolates when challenged in broilers, Partridge origin anti-H9N2 antibodies showed varied hemagglutination inhibition results (4log2 to 15log2) when cross reacted with different antigens, suggesting H9N2 viruses are circulating with different antigenic characteristics. Sero-biochemical analysis of blood collected at 7th DPI revealed an increase in aspartate amino transaminase (246-308 U/L), alanine amino transferase (31-58 U/L) and blood urea nitrogen (11.21-22.58 mg/dl).

Article History

Received: August 09, 2018

Revised: October 09, 2019

Accepted: October 15, 2019

Published: December 05, 2019

Authors’ Contributions

TY and MYT conceived and designed the experiment. MFS and MA did experimental work. MFS and AA performed statistical analysis. MFS and AR wrote the manuscript. SY did bioinformatic analysis.

Keywords

Antigenic variants, H9N2, Histopathology, Pathogenicity, Sero-biochemical

Corresponding author: Muhammad Furqan Shahid, furqan.shahid@uvas.edu.pk

To cite this article: Shahid, M.F., Yaqub, T., Tipu, M.Y., Aslam, A., Yaqub, S., Rahman, A. and Ali, M. 2019. Comparative pathogenic potential of avian influenza A/H9N2 viruses isolated from commercial, backyard and fancy birds. Punjab Univ. J. Zool., 34(2): 143-148. https://dx.doi.org/10.17582/journal.pujz/2019.34.2.143.148

Introduction

Avian Influenza Virus (AIV) belongs to the family Orthomyxoviridae consists of three different groups (A, B, C) classified on the basis of nucleocapsid (NP) and matrix (M) proteins. Spreading of AIV H5N1 is the most significant problem facing by poultry industry in the world but beside that there is also underestimated challenge, which is the spread of AIV subtype H9N2 in Asia (Gharaibeh, 2008). In South Asian countries H9N2 was found circulating with varying pathogenic characteristics (Nili and Asasi, 2002). Similarly, in Pakistan, it was reported earlier that H9N2 subtype of AIV produce marked pathogenicity with high mortality rate in chickens (Naeem et al., 1999) but now H9N2virus has low pathogenic nature with low mortality rate (Sarwar et al., 2013). So, the infection with H9N2 does not bring about high mortality in poultry but rather it can possibly do only in case of immune-compromised birds subsequent of vaccine stress or any environment effect (Bano et al., 2003).

The H9N2 pathogenicity is variable and it also depends upon the presence of other pathogens circulating in the field (Banet-Noach et al., 2007). As H9N2 has vast range of avian hosts including terrestrial poultry birds worldwide, so there is chance of continuous evolution of virus as the virus circulate among different hosts and thus it has become severe panzootic threat in Eurasia during last two decades (Choi et al., 2004). It has also been clear from the previous study that change in biological diversity of H9N2 plays a critical role in potentiating the antigenic potential of virus and thus it can cause severe pandemics in future (Choi et al., 2004). H9N2 virus is under continuously evolution process to better adapt in multiple host which led to change in pathogenic potential in chicken (Naeem et al., 1999; Nili and Asasi, 2002; Gharaibeh, 2008). Paucity of data regarding pathogenic potential in different avian species neglected for H9N2 vaccination, however, kept in close contact to human. Therefore, current study was planned to investigate the pathogenicity and tissue tropism of H9N2 virus isolated from different hosts and localities.

 

Materials and Methods

Viruses

Three Avian Influenza, affirmed H9N2 viruses, isolated from different hosts were secured from Influenza laboratory, Department of Microbiology, University of Veterinary and Animal Sciences, Lahore, Pakistan. The hemagglutinin (HA) and neuraminidase (NA) gene sequences of these three H9N2 viruses are available under accession numbers: MG720814 (A/Chicken/Pakistan/660BYP/2015/H9N2), MG720815 (A/Chicken/Pakistan/871CF/2015/H9N2) and MG720816 (A/Partridge/Pakistan/260/2015/H9N2) for HA gene. While the accession numbers for the sequences of NA gene are: MF280168 (A/Chicken/Pakistan/660BYP/2015/H9N2), MF280169 (A/Chicken/Pakistan/871CF/2015/H9N2) and MF280170 (A/Partridge/Pakistan/260/2015/H9N2) (Ali et al., 2017).

Experimental infection

After achieving the approval letter from the institution having IERB number 645/Bact on October 2015, the present challenge study was conducted. To investigate the infection potential of aforementioned LP AIV isolates, a challenge or experimental infection in commercial poultry was conducted during November 2015 to June 2016. For this purpose, n=40, day old chicks were acquired from a commercial hatchery and raised under controlled environmental conditions in chicken isolators. The feed and water were supplemented to birds ad-libitum during the experiment period. Prior to two days challenge, all birds were screened out for any subclinical infection of AIV H9N2, Newcastle disease virus (NDV) and Infectious bronchitis virus (IBV). For the screening of these viruses, tracheal swabs were collected from the birds and polymerase chain reaction (PCR) was performed via earlier reported primers (Rashid et al., 2009) after extracting RNA using QIAmp® viral RNA mini kit (cat: 52906). After confirming the depletion of maternal antibodies against H9N2 by using hemagglutination inhibition (HI) test (Alexander and Chettle, 1977) at age of 5 week, birds were divided into different four groups (A, B, C and D) having 10 birds each in a group and placed in different chicken isolators. The group A, B, and C were challenged with A/Chicken/Pakistan/660BYP/2015/H9N2, A/Chicken/Pakistan/871CF/2015/H9N2 and A/Partridge/Pakistan/260/2015/H9N2, respectively via oronasal route with the calculated dose of 106 EID50. The calculation of EID50 of all three studied isolates was done according to the earlier reported method (Reed and Muench, 1938). The D group was kept as control and sterile phosphate buffer saline (PBS) was given via oronasal route.

Clinical signs, histopathological examination

After infection, all birds were inspected daily for any clinical signs. At 5th and 9th DPI, n=3 random birds were slaughtered from each experimental group and all the gross lesions were recorded. Tissues samples including trachea, lungs, kidney and liver were fixed in 10% neutral buffered formalin and were processed in standard techniques for fixation, dehydration, clearing, embedding, sectioning and staining as described by Drury and Wallington (Drury and Wallington, 1967).

Serological analysis

Blood samples were collected at 0, 3, 7 and 14th DPI from the trial birds to collect serum. Immune response (titer) of serum collected from each trial group was assessed by HI test utilizing 4 hemagglutinin units (4HAU) of that particular virus with whom challenge group was infected. Along with, the cross-HI of all studied H9N2 strains was performed against specific antisera collected from recovered birds. HI test of sera was performed by making 2 fold serial dilution in 96 well plate using 4HAU of studied H9N2 strains with 1% chicken RBC’s (Alexander and Chettle, 1977).

Serum chemistry analysis

Serum aspartate amino transaminase (AST), Serum alanine amino transferase (ALT) and Blood urea nitrogen (BUN) of collected serum samples of trial birds were determined by Reitman and Frankel method (Reitman and Frankel, 1957).

Statistical analysis

The obtained data was analyzed using GraphPad Prism version 6.01 (Swift, 1997). To know any significant link of enzyme status within and between different groups at different time interval, two way-ANOVA and multiple comparison were employed. The differences between enzyme levels of different challenged groups were separated by Tukey’s test. P values < 0.5 were considered statistically significant.

 

Results

Results revealed no mortality in all groups when challenged with A/H9N2 isolates. However, most of the effected birds revealed slight depression, low feed and water intake along with mild clinical signs like lethargic, sneezing, gasping, facial swelling, nasal discharge, redness of eyes and lacrimation on 2nd day post infection (DPI). These clinical signs were observed up to 5th DPI with gradually decrease in severity. Diarrhea was only observed in group C and birds were slightly depressed with ruffled feathers till 9th DPI. The most recurrent gross lesions in infected birds were congestion of lungs and trachea along with swollen kidney in all challenged groups. There were no apparent gross lesions observed in any other organs of infected birds. Apparently, birds were started to recover naturally on 6th DPI with signs and symptoms disappeared till 10th DPI. No clinical signs were observed in control group.

Histopathological examination of tracheal tissue showed inflammation and necrosis. Sloughing of epithelial lining of trachea with lymphocytic infiltration was also observed in trachea. Lungs showed congestion and perivascular hemorrhages with neutrophilic infiltration. There was congestion of portal vein in liver with aggregation of leukocytes and macrophages. Hepatocytic swelling and necrosis was also observed in liver. Severe congestion, glomerulus inflammation, vacuolar degeneration in tubular epithelium along with leukocytic infiltration and peri-glomerular fibrosis was observed in kidney. However, neutrophilic infiltration in all tested organs were observed higher at 9th DPI when compared with 5th DPI in all challenged groups (Figure 1).

Serological investigations revealed that antisera against H9N2 of partridge host (group C) produced varying cross reactivity with backyard (group A; A/Chicken/Pakistan/660BYP/2015/H9N2) and commercial (group B; A/Chicken/Pakistan/871CF/2015/H9N2) poultry origin H9N2 virus (Table 1).

No amino acid mutation was observed at receptor binding residues 191, 198, 234-235 (H9 numbering) of the three studied isolates. Furthermore, motif KSSR/G at HA cleavage site confirmed the low pathogenic nature of all three isolates. Off note, NA stalk length showed mutation at residue I62T (N2 numbering) in patridge origin isolate (group C; A/Partridge/Pakistan/260/2015/H9N2). However, no mutation was observed in rest of the NA key residues of stalk length (38-39, 46-50, and 63-64), heamadsorbing site (372, 402-403) and heamadsorbing loop (313, 368, 370, 381).

 

Steady increase in liver enzymes (AST, ALT) and kidney function test (BUN) was recorded till post infection to 7th DPI and then sudden fall was observed at 14th DPI. Maximum level of AST (246-308 U/L), ALT (31-58 U/L) and BUN (11.21-22.58 mg/dl) in all challenged groups was seen at 7th DPI (Table 2). Significant (p<0.001) increase in liver enzymes (AST, ALT) and BUN was recorded between experimental groups (Figure 2). The level of ALT increased with high significant in all challenged groups as compared to the control group at 7th DPI. Similarly, the level of ALT also increased with high significance (p<0.001) in all challenged groups but level of ALT was higher in the group C (58±2.34 U/L) than other challenged groups A (31±2.54 U/L) and B (33.8±1.30 U/L) at 7th DPI. The level of BUN was also significantly (p<0.001) increased in all challenged groups but the level was a bit higher in group B (22.58±1.13 mg/dl) than other challenged groups A (12.11±0.91 mg/dl) and C (14.1±0.60 mg/dl) at 7th DPI. The liver enzyme (ALT, AST) and BUN was dropping to normal level at 14th DPI.

 

Discussion

Current study revealed varying clinical signs in chickens challenged with H9N2 viruses of different host origin, may suggest circulation of different H9N2 varients in birds in Pakistan. Although all birds recovered naturally but indicate that fancy birds neglected with H9N2 vaccine may provide opportunity for genetic mingling and genesis of novel influenza viruses. The clinical infection and

 

Table 1: Cross reactivity of raised hyperimmune sera against different isolates of H9N2 virus.

Isolates	HA Activity (log2)	HI of hyperimmune sera (log2)
		Antise-rum Group A	Antis-erum Group B	Antis-erum Group C	Antis-erum Control
Group A (A/Chicken/Pakistan/660BYP/2015 (H9N2))	11	10	16	13	1
Group B (A/Chicken/Pakistan/871CF/2015 (H9N2))	7	9	12	13	2
Group C (A/Partridge/Pakistan/260/2015 (H9N2))	9	4	15	14	2

 

histo-pathological evidences also confirmed that circulating H9N2 viruses are of low pathogenic in nature and clustered in sub-lineage B2 under G1 lineage of Middle East group (Ali et al., 2017). In Pakistan, frequent outbreaks of H9N2 have been observed from last two decades which shows endemic circulation in the country. Birds of group A and B challenged with H9N2 virus showed clinical presentation similar to previous studies conducted in Pakistan in 2009 (Subtain et al., 2011) and 2015 (Aslam et al., 2015) which suggests that our H9N2 viruses isolated from commercial and backyard poultry has same pathogenic potential with the isolates of H9N2 used by Aslam et al. (2015) and Subatain et al. (2011). The same clinical findings were also observed in experimentally infected chicken in other countries like Egypt in 2018 (Awadin et al., 2018). This may be due to the reason that H9N2 isolate used in Egyptian study was of G1 lineage circulating in Middle East is closely related to the H9N2 isolates used in our study as described in our previous study (Ali et al., 2017). While clinical sign like diarrhea was only observed in group C which was infected with the H9N2 virus of Partridge origin but the other clinical signs of respiratory system were same as observed in previous studies (Subatain et al., 2011; Aslam et al., 2015). In our study it is clearly observed that the A/H9N2 viruses isolated from the field can produce clinical infection which is contradicted to the results of a previous study reported in India (Dash et al., 2015) in which the H9N2 virus isolated from field outbreak does not produce clinical infection in the experimental birds. This contradiction might be due to the change in antigenic nature of used isolates in the study. Present study showed severity of lesions highest at 9th day post infection in group C, however, in other studies it was recorded till 6th day post infection (Subatain et al., 2011). Our results suggest that circulating H9N2 viruses has the higher pathogenic potential and is under strong selection pressure.

H9N2 isolates used under current study produced higher HA titer (8 ̶ 11 log2) than reported previously (7 ̶ 9 log2) in Pakistan in year 2011 (Subatain et al., 2011). This contradiction is assumed on basis of antigenic variation in H9 virus isolates with passage of time (Kim et al., 2006). Varying response of cross reactivity of antisera with the different field A/H9N2 isolates (n=3) suggests circulation of H9N2 subtype with different immunogenic properties at the same time in different localities and hosts in poultry of Pakistan. Current results also highlighted the importance of systematic surveillance and further investigation of pathogenic characteristics of endemically circulating H9N2 virus, which may help to select best H9N2 virus as a vaccine candidate.

No mutation was recorded in HA gene at functionally important key residues; however, one mutation was observed in NA stalk length in partridge origin H9N2 virus, might be associated with host adaptation and increased virulence (Iqbal et al., 2009). Results confirmed that NA gene might be under strong selection pressure (Ali et al., 2017). Off note, mutation was only observed in group C, also exhibited severity of clinical signs along with higher level of varying cross reactivity titer to other chicken and backyard poultry isolate.

 

Table 2: Serobiochemical parameters of birds infected with different isolates of H9N2.

Experimental Birds	Day PI	AST (U/L)	ALT (U/L)	BUN (mg/dl)
Group A	0	155.8±4.32	8±1.58	7.98±0.61
	3	233.8±5.11	15.6±1.40	8.22±0.52
	7	291±5.09a	31±2.54bc	12.11±0.91ef
	14	218.6±4.39	20.6±2.70	7.94±0.66
Group B	0	167.4±5.27	13.6±1.40	9.42±0.42
	3	246±5.29	26.2±0.83	9.4±1.14
	7	306.2±3.03a	33.8±1.30bd	22.58±1.13efg
	14	237±5.74	24.8±0.83	18.34±0.40
Group C	0	154±5.83	10.2±1.64	7.2±0.45
	3	232.6±10.66	30.2±0.83	11.8±0.84
	7	308±6.44a	58±2.34bcd	14.1±0.60efg
	14	164.8±6.61	34±2	11.21±1.15
Control	0	147.±8.98	9.6±1.51	8.4±1.14
	3	158.2±7.49	9±2.44	7.23±1.65
	7	152.8±16.25a	8.4±1.14b	7.5±0.47e
	14	155.4±9.37	10±1.58	7.98±0.43

PI: Post Infection; AST: Aspartate amino transaminase, U/L: unit per litter; ALT: Alanine amino transferase; BUN: Blood urea nitrogen, mg/dl: milligram per deci litter; a,b,c,d,e,f,g: Different letters within the same column indicate significant differences among levels according to Tukey’s test (p≤0.05)

 

Increased level of ALT and AST suggest that pathogenic involvement of liver in H9N2 and this was supported by gross lesion and histopathological examination of liver in infected birds. As liver is considered to be the major source of plasma protein like ALT and AST, so damage to this organ leads to change in levels of liver enzymes in blood serum (Egbal et al., 2007). Results of sero-biochemical analytes revealed that studied H9N2 viruses has also strong affiliation to liver and kidney as reported by other researchers (Subatain et al., 2011; Aslam et al., 2015). The higher level of ALT in group C also supports the clinical sign like diarrhea which was only observed in this group. This suggests that H9N2 virus of partridge origin has strong affiliation to liver and viruses isolated from different hosts may have different pathogenic potential which has also been reported previously (Cong et al., 2007; Yu et al., 2008).

 

Conclusion

Our results indicate circulation of multiple H9N2 variants with varying pathogenic potential in different avian species. Observed pathogenic potential and mutation in molecular marker of fancy bird isolate suggest urgent need to devise strategies because of close interaction to human. Therefore, future investigations related to genomic variations due to influence of host adaptation of virus or immune pressure of irrational usage of field vaccine are warrant.

 

Acknowledgement

The authors would like to thanks Influenza lab, Department of Microbiology, University of Veterinary and Animal Sciences, Lahore for providing facilities. We also thank Dr. Ghulam Mustafa for providing facility of histopathology and Clinical pathology Lab, Department of Pathology, University of Veterinary and Animal Sciences, Lahore for serum chemistry analysis.

 

Conflict of interest statement

The authors declare no conflict of interest.

 

References

Alexander, D. and Chettle, N., 1977. Procedures for the haemagglutination and the haemagglutination inhibition tests for avian infectious bronchitis virus. Avian Pathol., 6: 9-17. https://doi.org/10.1080/03079457708418208

Ali, M., Yaqub, T., Mukhtar, N., Imran, M., Ghafoor, A., Shahid, M.F., Yaqub, S., Smith, G.J., Su, Y.C. and Naeem, M., 2017. Prevalence and phylogenetics of H9N2 in backyard and commercial poultry in pakistan. Avian Dis., 62: 416-424. https://doi.org/10.1637/11690-062117-ResNote.1

Aslam, R., Aslam, A., Tipu, Y., Nazir, J., Ghafoor, A. and Fatima, S., 2015. Histopathological and immunohistochemical studies for the pathogenesis of a low pathogenicity H9 avian influenza virus in experimentally infected commercial broilers. J. Anim. Plant Sci., 25: 45-52.

Awadin, W., Said, H., Abdin, S. and El-Sawak, A.A., 2018. Pathological and molecular studies on avian influenza virus (H9N2) in broilers. Asian J. Anim. Vet. Adv., 13: 232-244. https://doi.org/10.3923/ajava.2018.232.244

Banet-Noach, C., Perk, S., Simanov, L., Grebenyuk, N., Rozenblut, E., Pokamunski, S., Pirak, M., Tendler, Y. and Panshin, A., 2007. H9N2 influenza viruses from israeli poultry: A five-year outbreak. Avian Dis., 51: 290-296. https://doi.org/10.1637/7590-040206R1.1

Bano, S., Naeem, K. and Malik, S., 2003. Evaluation of pathogenic potential of avian influenza virus serotype H9N2 in chickens. Avian Dis., 47: 817-822. https://doi.org/10.1637/0005-2086-47.s3.817

Choi, Y., Ozaki, H., Webby, R., Webster, R., Peiris, J., Poon, L., Butt, C., Leung, Y. and Guan, Y., 2004. Continuing evolution of H9N2 influenza viruses in southeastern china. J. Virol., 78: 8609-8614. https://doi.org/10.1128/JVI.78.16.8609-8614.2004

Cong, Y.L., Pu, J., Liu, Q.F., Wang, S., Zhang, G.Z., Zhang, X.L., Fan, W.X., Brown, E.G. and Liu, J.H., 2007. Antigenic and genetic characterization of H9N2 swine influenza viruses in china. J. Gen. Virol., 88: 2035-2041. https://doi.org/10.1099/vir.0.82783-0

Dash, S., Kumar, M., Kataria, J., Tosh, C., Murugkar, H., Rajukumar, K., Kulkarni, D. and Nagarajan, S., 2015. Experimental assessment of pathogenicity and infectivity of H9N2 influenza virus isolated from a natural outbreak. Indian J. Vet. Pathol., 39: 221-225. https://doi.org/10.5958/0973-970X.2015.00053.X

Drury, R.A.B. and Wallington, E., 1967. Carleton’s histological technique.

Egbal, S., Samia, A., Iman, M. and Inaam, E., 2007. Serobiochemical changes associated with highly pathogenic avian influenza in chickens. Res. J. Anim. Vet. Sci., 2: 53-55.

Gharaibeh, S., 2008. Pathogenicity of an avian influenza virus serotype H9N2 in chickens. Avian Dis., 52: 106-110. https://doi.org/10.1637/8108-090907-Reg

Iqbal, M., Yaqub, T., Reddy, K. and McCauley, J.W., 2009. Novel genotypes of H9N2 influenza a viruses isolated from poultry in pakistan containing ns genes similar to highly pathogenic h7n3 and h5n1 viruses. PLoS One, 4: e5788. https://doi.org/10.1371/journal.pone.0005788

Kim, J.A., Cho, S.H., Kim, H.S. and Seo, S.H., 2006. H9N2 influenza viruses isolated from poultry in korean live bird markets continuously evolve and cause the severe clinical signs in layers. Vet. Microbiol., 118: 169-176. https://doi.org/10.1016/j.vetmic.2006.07.007

Naeem, K., Ullah, A., Manvell, R. and Alexander, D., 1999. Avian influenza a subtype H9N2 in poultry in pakistan. Vet. Record, 145: 560-560. https://doi.org/10.1136/vr.145.19.560

Nili, H. and Asasi, K., 2002. Natural cases and an experimental study of H9N2 avian influenza in commercial broiler chickens of iran. Avian Pathol., 31: 247-252. https://doi.org/10.1080/03079450220136567

Rashid, S., Naeem, K., Ahmed, Z., Saddique, N., Abbas, M. and Malik, S., 2009. Multiplex polymerase chain reaction for the detection and differentiation of avian influenza viruses and other poultry respiratory pathogens. Poult. Sci., 88: 2526-2531. https://doi.org/10.3382/ps.2009-00262

Reed, L. and Muench, H., 1938. Asimple method of estimating fifty percent endpoints. Am. J. Hyg., 27: 493-498. https://doi.org/10.1093/oxfordjournals.aje.a118408

Reitman, S. and Frankel, S., 1957. A colorimetric method for the determination of serum glutamic oxalacetic and glutamic pyruvic transaminases. Am. J. Clin. Pathol., 28: 56-63. https://doi.org/10.1093/ajcp/28.1.56

Sarwar, M., Muhammad, K., Rabbani, M., Younus, M., Sarwar, N., Ali, M. and Ahad, A., 2013. Prevalence of avian influenza viruses in live bird markets of lahore. J. Anim. Pl. Sci., 23: 388-392.

Subtain, S., Chaudhry, Z.I., Anjum, A.A., Maqbool, A. and Sadique, U., 2011. Study on pathogenesis of low pathogenic avian influenza virus h9 in broiler chickens. Pakistan J. Zool., 43: 999-1008.

Swift, M.L., 1997. Graphpad prism, data analysis, and scientific graphing. J. Chem. Inf. Comput. Sci., 37: 411-412. https://doi.org/10.1021/ci960402j

Yu, H., Hua, R.H., Wei, T.C., Zhou, Y.J., Tian, Z.J., Li, G.X., Liu, T.Q. and Tong, G.Z., 2008. Isolation and genetic characterization of avian origin H9N2 influenza viruses from pigs in china. Vet. Microbiol., 131: 82-92. https://doi.org/10.1016/j.vetmic.2008.02.024