Antigenic diversity and similarity of Newcastle disease viruses isolated from unvaccinated free-range rural chickens characterised by polyvalent and monoclonal antibodies

M G S Yongolo, A P Muhairwa*, D J Alexander**, K J Handberg***, R. J Manvell** and U M Minga****

Animal Disease Research Institute, Ministry of Agriculture and Co-operatives, PO Box 9254, Dar es Salaam, Tanzania
myongolo@yahoo.co.uk
*Department of Veterinary Medicine and Public Health, PO Box 3021, Morogoro, Tanzania
**Avian virology, VLA, Weybridge, Addlestone, Surrey KT15 3NB, UK 4 Department of Pathobiology
***The Royal Veterinary and Agricultural University, Frederiksberg C
****Open University of Tanzania, PO Box 23409 Dar-es Salaam, Tanzania

Abstract

Antigenic heterogeneity of 33 Newcastle disease virus isolates from healthy and sick free-range rural chickens and ducks was determined by analysis of their binding patterns to a panel of 27 monoclonal antibodies (mAb). The studied viruses had been isolated from apparently healthy and sick unvaccinated village chickens and ducks from four regions and different times of the year.

All isolates were positive to polyvalent APMV-1 antisera and to mAb U85 and negative to the pigeon mAb 161/617. Four isolates had high HI titres to mAb 7D4, which is specific for La Sota vaccine. Five antigenic binding profiles A, B1, EL, EB, and G segregated into 14 antigenic groups. Five antigenic groups had 100% antigenically similar isolates from different regions, indicating a spread from the same source. Isolates with EL and EB binding patterns probably had vaccine strains as progenitor virus. Nine were composed of isolates exclusively from the same region, indicating that isolates had not spread to other regions. Isolation of 100% antigenically similar strains from the same region suggested the persistence of NDV there. Detection of virulent isolates was clear with groups A, B and EB, but failed in group G and one isolate in group EL, possibly due to mixed infection. The use of mAb e, l and H33A for detection of virulent NDVs is recommended.

It is concluded that mAb binding test is a potentially useful tool in poor resource country laboratories, where routine use of in vivo tests and molecular pathotyping are difficult to sustain.

Keywords: Monoclonal antibodies, NDV pathoptyping, Newcastle disease

Introduction

Newcastle disease (ND) caused by an ND virus (NDV) serotype 1 is one of the most serious disease of poultry worldwide (Spradbrow 1999; Alexander 2001). Therefore, efficient and prompt diagnosis, which requires assessment of virulence, is important for timely containment of outbreaks and spread of NDVs. The disease and the virus were first reported in late 1920s (Alexander 2001). From then until the early 1980s using polyvalent antiserum, NDVs were considered homogeneous, evidenced by the use of a single vaccine strain for protection against different field strains (Alexander 1997). However, in the advent of monoclonal antibodies antigenic differentiation of NDVs was made possible and useful for diagnosis, strain differentiation and epidemiology of ND (Russell and Alexander 1983; Alexander et al 1987). Newcastle disease virus [avian paramyxovirus type 1] belongs to the genus Avulavirus, has a non-segmented single stranded RNA genome, contained in an envelope (Mayo 2002). Viruses with such genomes have inexact replication of the RNA, which frequently leads to production of variants with differences, often subtle differences, in phenotype from the parent particle (Spradbrow 1999). Thus, they are considered as quasi-species and the populations that spread in the field, or the populations that make up a vaccine stock, are not clonal. Furthermore, selection pressure can alter the average behaviour of the population (phenotype). Thus, NDVs with similar antigenicity gives indication of an epidemiological link.

In many countries of Africa and Asia, ND outbreaks are common in FRC and the disease can be considered endemic (Spradbrow 1999). These flocks are often composed of birds at different ages, not vaccinated, no biosecurity and mixed with other species (Yongolo et al 2002). It is however not known whether ND viruses circulating in FRC are subjected to selection pressure or not. Antigenic characteristics of NDVs circulating in the field could give an indication on the existence of selection or not. Recently, vaccination against ND has been recommended to protect FRC from ND outbreaks (Spradbrow 1999). However, for any success in the control of ND the background knowledge of its epidemiology is a prerequisite. Ironically, little is documented on the spread and introduction of ND in FRC populations in most African and Asian countries and the little information available is based on observational data. Monoclonal antibodies [mAbs] have been used for virus identification of antigenically variant NDVs (Aldous et al 2003) and have proved useful in epidemiological considerations of the disease (Alexander et al 1984). Furthermore, differentiation of strains with virulent characteristics has been reported (Alexander et al 1992). Thus, data from mAb analysis of NDV can be used in determining the source and explain its spread (Alexander et al 1997, Panshin et al 2000; Panshin et al 2001; Panshin et al 2002).

In FRC populations in Tanzania velogenic, mesogenic and lentogenic NDVs have been isolated from different sources (Yongolo et al 2002). However, there have been no studies on characteristics of NDVs from FRC in Tanzania, which could be of epidemiological importance. The presence and characteristics of NDV can be demonstrated by a number of methods. Currently, genetic characterisation is the method of choice to achieve three needs at one go; pathotyping, strain differentiation and epidemiological links (Aldous and Alexander 2001). Furthermore, in vivo virulence determination needs eggs and chicks from SPF flocks and a high level of laboratory sterility to curb contaminations. However, these techniques are expensive, difficult to maintain and require specialised facilities which are difficult to attain in poor resource countries. Haemagglutination and haemagglutination inhibition tests have traditionally been used to determine presence of NDV (Alexander 2000). They are uncomplicated, require minimum inputs in terms of equipment and reagents, and as a result can be afforded by poor resource laboratories. Despite the fact that monoclonal antibodies have a limited capacity to differentiate viruses that are antigenically similar but genetically distinguishable (Aldous et al 2003), they can still be used in HI tests to rapidly confirm and characterise NDVs and give some information relating to their epidemiology and origins (Aldous et al 2003; King and Seal 1998).

This paper focuses on antigenic variations manifested by NDVs from FRCs in Tanzania, which can be used in virulence assessment and epidemiological inference. Furthermore, to avoid the use of in vivo tests, this paper advocates the use of monoclonal antibodies as a tool for early detection of NDV similarities and pathotyping. Taking in consideration that ND is a trans-boundary problem, achievement of this would enhance early warning and early reaction, which would lead to enabling effective control of Newcastle disease.

Materials and methods

Virus

Allantoic fluid harvests of thirty-three Newcastle diseases virus isolates were used in this study. The isolates were obtained from unvaccinated free-range rural chickens and ducks. Of the 33, two were from apparently healthy ducks reared together with chickens and the remaining 31 were from chickens. Of the 31 from chickens 10 were from sick chickens from an ND suspected outbreak and 21 from apparently healthy chickens from ND free flocks. Characterisation with mAbs was done to all isolates after five passages in 9 - 11 days embryonated chicken eggs. Before characterisation confirmation of NDV was done at the Veterinary Laboratories Agency, Weybridge, Surrey, United Kingdom. Identity of each isolate indicated information on the source of the virus (geographical location, month of isolation and host). The first letters showed the regions in Tanzania where it was isolated, the middle number shows first the month when it was isolate, followed by a dot and the laboratory serial number. The last letters show the host from which it was isolated. Thus, all isolates with MG are from Morogoro, TB are from Tabora, MB from Mbeya and MS from Kilimanjaro. 1 to 12 of the first numbers before the dot indicates months from January to December respectively.

Antibodies

Newcastle disease polyvalent antisera and a panel of 27 monoclonal antibodies were used to characterise the isolates. The ability of each mAb to bind to the NDVs was carried out according to the procedure described by Alexander et al (1999) without modification. Abbreviations used for the identity of mAbs are also according to Alexander et al (1999). Before using the panel of 27 mAbs preliminary NDV identification was done using monoclonal antibody U85 a mAb specific to APMV-1 and not APMV-2 to 9, 7D4 a mAb specific for La Sota vaccine like NDVs, 617/161 a mAb specific for the pigeon variant NDV and polyvalent APMV-1 antiserum were used as a 1:32 dilution for the monoclonal antibodies and 1:16 dilution for the polyclonal antiserum. The haemagglutination inhibition tests procedure was as described by CEC (1992).

Analysis of the data

Serological test results of negative or positive were transformed to two level numerical scale. Where negatives were scored as 1 and positives as 2. The taxonomy programme NYTSYS (Rohlf 1993) commonly used for analysing data from biochemical test reactions of bacteria (Angen et al 1997) was adopted and used to perform the analysis of mAbs reactions of different isolates. Cluster analyses were performed on the distance matrix using the unweighted pair group with arithmetic averages (UPGMA).

Results

Haemagglutination inhibition test results to polyclonal APMV-1 antisera, mAb U85, 7D4 and the mAb 161/617 are shown in Table 1. All 33 isolates reacted positive to APMV-1 antiserum and mAb U85. However, eight isolates had partial inhibition to mAb U85 (Table 1). Results with mAb 7D4 showed that only four isolates were positive and all isolates were HI negative to mAb 617/161 (Table 1). Six isolates MG10.03C, MG6.16C, MG3.35C, TB5.20D, MS8.39C and MG6.7C had the highest titres to mAb U85, and two of these MG10.3C and TB5.20D had the highest HI titres to mAb 7D4. Whereas isolate TB5.22C had the lowest HI titres to all tested antiserum and mAbs (Table 1).

The binding pattern of the study isolates as a results of the ability of the 27 mAbs to bind (+) or not to bind (-) to Vero cells are shown in Table 2. Ten different mAbs (e, f, h, I, j, l, w, x, £, and H339A) had no ability to bind to some isolates as shown by the (-) sign in Table 2. Comparing the binding profiles of the study isolates and that of La Sota and those from previous findings and designation by Alexander et al (1997), our isolate were then judged to fall under five antigenic patterns, eight isolates in A, two in B, five in E/L, three in E/B and 16 in G (Table 2).

Analysis of antigenic similarities according to binding profiles shown in Table 2 revealed presence of two clusters MC1 and MC2 (Figure 1). Where 17 isolates composed MC1 and 16 segregated into MC2. Two sub clusters were observed in each cluster C1 and C2 in the first cluster, which appears at the top of the phenotypic dendogram and C3 and C4 in the second cluster (Figure 1). The C1 isolates were grouped into three groups, three isolates in C1A, four in C1B, one in C1C and two in C1D. All isolates from C1 were chicken isolates from one region except for one isolate TB2.24D, which was isolated from a duck in another region. The C1A isolates were all from chickens, two from the same month (October) and one from February and were 100% antigenically similar. These isolates differed antigenically with C1B, C1C, and C1D by 18, 20 and 28% respectively. The differences in antigenicity were based on binding to the nine mAbs designated as e, f, i, j, l, w, x, £ and H339A. Isolates in C1B were obtained in different months but from the same region. A similar relationship was seen for C1D isolates. The antigenic similarity within C1 isolates ranged from 72 to 100%.

The C2 sub cluster comprised of three groups, where four isolates were in C2A, two in C2B and one in C2C. The antigenic similarity between isolates of C2 ranged from 70 to 100%. The second cluster MC2 comprised of two subclusters C3 and C4. Three groups were segregated into subcluster C3 and three other groups in subcluster C4. The first group C3A had six isolates, the second C3B has two isolates and the third C3C has three isolates. The last subcluster has five isolates segregated into three groups C4A, C4B and C4C (Figure 1). Isolates composing antigenic cluster C3 had a similarity range of 75 to 100%.

Of the duck isolates TB5.20D was grouped together with four other chicken isolates in C2A having the EL monoclonal binding profile. While the other duck isolate TB2.24D was grouped with isolates having A mAb binding profile in C1 subcluster. However, while other chicken isolates were negative to mAb e the duck isolate was positive. In both regions isolates from the same region and month were not at all times grouped in the same group. There was only one group C3B, which had two isolates from the same month and region, grouped together. While, isolates in C2A, C3A, C3C and C4A show that 100% antigenically similar isolates were isolated in the same region from different months January to September.

Correlation of the antigenic groups of the study isolates to virulence (data not shown) showed that isolates in C1 clusters were virulent, their intracerebral pathogenicity index values ranged from 0.87 to 1.67 (Yongolo 2004) and their mAb binding patterns classified them as A (group C1A, C1B, C1C) and B (group C1D). Three mAbs e, l and H339A had negative results to all virulent isolates (isolates in C1) except for the duck isolate. All other isolates in other clusters were positive to these two mAbs. Isolates forming subcluster C2, composed of three groups and having two binding patterns EL and EB, were all low virulent viruses (ICPI 0.0 to 0.5; (Yongolo 2004)) except for isolate TB5.22C, which had an ICPI of 0.86. Different results on correlation of virulence and antigenic patterns were observed in MC2 isolates, i.e. those having a G binding pattern. Where 100% antigenically similar isolates in C3A had manifestly different virulence (ICPI, 0.5 to 1.5; Yongolo 2004). Similarly for isolates in groups C3B, C3C, C4A, C4B and C4C, in each of them viruses showing a range of virulence from low to high were found to have the same antigenic patterns.

Discussion

Monoclonal antibody analysis differentiated the 33 NDV isolates into 13 antigenic sub-clusters. The presence of 13 sub-clusters indicated the existence of antigenic variations, within the field isolates. The existence of antigenic variants has also been reported elsewhere by Alexander et al 1997; Yu et al 2001 and Panshin et al 2002. Determinants that are responsible for antigenic variations were not established in this study and in previous studies. It is known that deletions, insertions, recombination or mutation can cause changes in viral genome or properties, including antigenicity (Murphy et al 1999). Following sequence analysis in previous studies, it was established that there is no evidence of recombination, deletions or insertions among NDVs (Seal et al 2000; Yu et al 2001). However, basing on the accumulation of point mutations that induced amino acid substitutions (Yu et al 2001, Yongolo 2004) it could be speculated that accumulative mutations are responsible for the production of the antigenic variations.

The presence of different isolates with different antigenic properties in FRC in Tanzania implies that NDV circulating in these flocks are not subjected to similar selection pressure. This is clearly shown by results in this study, where isolates with similar binding pattern in each subcluster manifested within cluster antigenic variations. Therefore, there is no single selection factor, which could determine a single dominant strain. The finding of isolates with the same unchanged antigenic pattern over different months indicates an absence of selection pressures that would result in changing. Alternatively, because of the nature of FRC flocks and the environment they are subjected to, this may indicate that in FRC there are multiple NDV infections with different sources or origins. The lack of vaccinations in FRC, the presence of chickens in the same flock or village with different immune status and the free exposure to different NDV hosts means the two explanations could be occurring contemporaneously. It can, therefore, be concluded from this study that the antigenic diversity observed in this study is the result of random mutational events and/or introduction of NDVs from multiple sources of different hosts to chickens.

In the cluster C1 representing virulent isolates, isolate TB2.24D a duck isolate from Tabora region was antigenically distinguishable from other isolates from another region (Morogoro) although they shared the same antigenic pattern A. It is difficult to delineate whether the difference was because it originated from a different host or because it was from a different area. However, it can be concluded that mAbs can differentiate virulent NDVs from different locations. These findings are line with Alexander et al (1997) who reported geographical associations with different mAb groups. Conflicting observations were made for isolates with EL, and G binding profiles (Figure 1). Isolates from different areas were all placed in the same group (C2A, C3A, C3C and C4A) and had 100% antigenically similar mAb binding pattern. Thus, mAbs in this case could not differentiate them according to geographical origin. It is currently known that the weakness of mAb binding tests is their failure to differentiate antigenically similar viruses, which are genetically distinct (Aldous et al 2003). However, the 100% antigenic similarity in some of the groups shows that they probably originated from the same source. This would mean that the progenitor virus spread to the other regions.

The wide spread of these type of viruses and their close similarity raises the suspicion that they may be vaccine derivatives. The finding in this study of isolates having group G mAb binding pattern is of interest. Group G mÁb isolates are distinct from the vaccine strains originating from Australia (Quensland/V4), Ulster2C/67, B1/47 and La Sota (Alexander et al 1997). This suggests that the G group isolated in this study may not have a vaccine origin. Lentogenic mAb group G viruses have been mainly isolated from waterfowl and other wild birds (Alexander et al 1997; Aldous et al 2003). Thus, although mAb group G NDVs were not isolated from ducks in this study, the mere presence of them in FRC, which are constantly in contact with ducks and other wild birds means they could be an exchange of NDVs between them in either way. Furthermore, strain B1/47, LaSota/47 are widely used as live vaccines throughout the world (Aldous et al 2003). These two strains may account for some of the isolates in this study. It is therefore possible that they could have spread to unvaccinated chickens, circulated and perpetuated in FRCs diverging antigenically from the original virus. In our study two groups EL and EB were identified as the subcluster C2. Similar close genetic similarity was reported by Aldous et al (2003), where two clear vaccinal strains were identified. Thus, mAbs used in this study could clearly identify the two vaccine-like NDVs.

The finding that 100% antigenically similar isolates were present in the same region from different months January to September, demonstrated that some strains maintained their antigenicity unchanged during the sampling period. This is evidence that NDVs may be persistent in FRC. Secondly, the presence of similar isolates in the two regions is an indication that there could be an ND epidemiological relationship between the two regions either through the use of a common vaccine or spread of virus from one region to the other. Although there was no direct link for the isolates in this study, it is known that there is one-way marketing of live chickens from western regions to eastern regions (Mlozi et al 2001), and it seems likely NDVs can be introduced by this means. The chronologies of isolations in this study support this. In all groups, NDVs isolated earlier in Morogoro were not isolated in Tabora and conversely (C1A, C1B, C2A, C3A, C3C, C4A). This means that antigenic variations detected by mAbs can be an aid to determining the spread of ND in FRC.

The virulence that was indicted by mAb typing correlated well with ICPI values in those having mAb binding pattern type A and B, but not with the other binding patters. Disparity of mAb typing for virulence and genotypic characteristics have been reported by Collins et al (1998) and Alexander et al (1997). This could be due to the fact that mAb typing is mainly directed to detection of epitopes on different surface antigens on the virus envelopes proteins (Aldous et al 2003). In this study, only four out of the 16 positive mAbs used in this study were specific for fusion protein, the rest were directed to HN, M and NP viral polypeptides (Alexander et al 1997). The second possibility is the presence of mixed NDV in the same culture. Recently Aldous et al (2003) found that a virulent pigeon isolate, which showed mAb group G binding, was actually a mixture of pigeon panzootic-like NDVs and Ulster-like virus. Thus, in our case isolates which showed lentogenic binding patterns and at the same time were virulent could have been mixed infections. It is then recommended that NDV with showing disparity should be cloned to determine their actual characteristic. In general mAbs were able to single out virulent isolates. Therefore, mAb binding tests can be used as a tool to rapidly characterise NDVs. This is primarily important for poor countries that are exporter of wild birds to Europe, America and Asia (Seal et al 1998). Thus, use of simple tests could facilitate early detection and control of ND by screening birds before export. In so doing reduce the risk of spread of NDV by such birds as was the case with the ND panzootic in 1970-1973. Moreover, use of three mAbs in simple tests like HI and ELISA could also minimise the excess use of in vivo tests to determine virulence.

Conclusion

• The study has demonstrated that there is antigenic heterogeneity among NDV isolates circulating in the FRC population in Tanzania.

• Antigenic similarity among isolates from different geographical areas, hosts and with time could be of epidemiological importance, in explaining persistence and spread of NDV.

• The clear delineation of virulent NDVs from other viruses means that mAbs could be used to detect and rapidly assess the virulence of isolates.

• Given that mAbs can be used in a simple test such as haemagglutination inhibition and ELISA makes them potentially useful in poor resource countries laboratories.

Acknowledgements

The authors are indebted to DANIDA and ENRECA of Denmark, who through the 'Improvement of Health and Productivity of Rural Chickens in Africa' project funded this study. We thank Prof. John E. Olsen, Maeda-Machangu A. E. and Prof. Madundo Mtambo for providing the necessary environment to undertake this study. Much gratitude are due to Lisbeth Nielsen of the Danish Institute for Food and Veterinary Research and staff of Avian Virology, VLA Weybridge, UK for the technical assistance that made this study possible.

References

Aldous E W and Alexander D J 2001 Technical review: Detection and differentiation of Newcastle disease virus (avian paramyxovirus type 1). Avian Pathology 30, 117-128.

Aldous E W, Mynn J K, Banks J and Alexander D J 2003 A molecular epidemiological study of avian paramyxovirus type 1 (Newcastle disease virus) isolates by phylogenetic analysis of a partial nucleotide sequence of the fusion protein gene. Avian Pathology 32, 239 - 257

Alexander D J 1997 Newcastle disease and other avian paramyxoviridae infections. In: Calnek B W; Barnes H J; Beard C W; McDougald L R and Y M Saif (Editors) Diseases of Poultry. 10^th edition. Ames, Iowa State University press pp. 5541 - 570

Alexander D J 2000 Newcastle disease and other avian paramyxoviruses. Scientific and technical review office international des épizooties 19, 443-462

Alexander D J 2001 Newcastle disease British Poultry Science 42, 5 - 22.

Alexander D J, Banks J, Collins M S, Manvell R J, Frost K M, Speidel E C and Aldous E W 1999 Antigenic and genetic characterisation of Newcastle disease viruses isolated from outbreaks in domestic fowl and turkeys in Great Britain during 1997. Veterinary Record 145, 417-421.

Alexander D J, Campbell G, Manvell R J, Collins M S, Parsons G and McNulty M S 1992 Characterisation of an antigenically unusual virus responsible for two outbreaks of Newcastle disease in the Republic of Ireland in 1990. Veterinary Record 130, 65 - 68

Alexander D J, Manvell R J, Kemp P A, Parsons G, Collins M S, Brockman S, Russell P H and Lister S A 1987 Use of monoclonal antibodies in the characterisation of avian paramyxovirus (Newcastle disease virus) isolates submitted to an international reference laboratory. Avian Pathology 16, 533 - 565.

Alexander D J, Manvell R J, Lowings J P, Frost K M, Collins M S, Russell P H and Smith J E 1997 Antigenic diversity and similarities detected in avian paramyxovirus type 1 (Newcastle disease virus) isolates using monoclonal antibodies. Avian Pathology 26, 399 - 418

Alexander D J, Wilson G W, Thain J A and Lister S A 1984 Avian paramyxovirus type 1 infections of racing pigeons. Epidemiological considerations. Veterinary Record 115, 213 - 216

Angen Ø, Aalbæk B, Falsen E, Olsen J E andBisgaard M 1997 Relationship among strains classified with the ruminant Pasteurella haemolytica complex using qualitative evaluation of phenotypic data. Zentrablatt fur . Bakteriology 285: 459-479.

CEC (Commission of the European Communities) 1992 Council directive 92/66/EEC of 14^th July 1992 introducing Community measures for the control of Newcastle disease . Official Journal of the European Communities L260, 1 - 20.

Collins M S, Franklin S, Strong I, Meulemans G, Alexander D J 1998 Antigenic and phylogenetic studies on a variant Newcastle disease virus using anti-fusion protein monoclonal antibodies and partial sequencing of the fusion protein gene. Avian Pathology 27, 90-96

King D J and Seal B S 1998 Biological and molecular characterization of Newcastle disease virus (NDV) field isolates with comparisons to reference NDV strains. Avian Diseases 37: 961-969

Mayo M A 2002 A summary of the changes recently approved by ICTV. Archives of Virology 147, 1655 - 1656

Mlozi M R S, Minga U M, Kakengi A M V and Olsen J E 2001 Marketing of free-range local chicken in Morogoro and Kilosa urban markets, Tanzania. In: Proceedings of the 29th Tanzania Society of Animal Production, Volume 29

Murphy F A, Gibbs E P J, Horzinek M C and Studdert M J 1999 Viral replication, genetics and evolution. In: Veterinary Virology 3rd Edition. (pp. 43-80) Academic Press New York USA

Panshin A, Shihmanter E, Weisman Y, Orvell C and Lipkind M 2002 Antigenic heterogeneity among field isolates of Newcastle disease virus (NDV) in relation to the vaccine strain: Studies on viruses isolated from wild birds in Israel. Comparative Immunology and Microbiology of infectious Diseases 25 95 - 108

Panshin A, Shihmanter E, Weisman Y, Orvell C, Kydyrmanov A, Asanov N, Daulbaeva K, Sayatov M and Lipkind M 2000 Antigenic characterisation of the nucleocapsid protein of Newcastle disease virus by means of a new panel of monoclonal antibodies. Comparative Immunology, Mocrobiology and Infectious Diseases 23, 9 - 220

Panshin A, Shihmanter E, Weisman Y, Örvell C, Kydyrmanov A, Sayatov M, Asanov N, Nyaga P N, Kasiiti J L, Macharia M J and Lipkind M 2001 The comparative antigenic characterisation of Newcastle disease virus strains isolated in Kenya and Kazakhstan. Comparative Immunology, Mocrobiology and Infectious Diseases 24, 21-37

Rohlf F J 1993 Numerical taxonomy and multivariate analysis system. NTSYS-pc programme 1.80. New York.

Russell,P H and Alexander D J 1983 Antigenic variation of Newcastle disease virus strain detected by monoclonal antibodies. Archives of Virology 75, 243 - 253

Seal B S, King D J, Locke D P, Senne D A and Jackwood M W 1998 Phylogenetic relationship among highly virulent Newcastle disease virus isolates obtained from exotic birds and poultry from 1989 to 1996. Journal of Clinical Microbiology. 366, 1141 - 1145

Seal B S, Sellers H.S and Meinersmann R J 2000 Fusion protein predicted amino acid sequence of the first US avian pneumovirus isolated and lack of heterogeneity among other US isolates. Virus Research 66, 139-147

Spradbrow P 1999 Epidemiology of Newcastle Disease and the Economics of its Control. In: Proceedings of a Workshop on Poultry as a Tool in Poverty Eradication and Promotion of Gender Equality. March 22-26, 1999 held at Tune Landboskole, Denmark. http://www.husdyr.kvl.dk/htm/php/tune99/16-Spradbrow.htm

Yongolo M G S 2004 Molecular Epidemiology of Newcastle disease virus isolates from Free-range rural chickens of Tanzania. PhD Thesis Submitted to Sokoine University of Agriculture, Morogoro Tanzania pp 362

Yongolo M G S, Maeda Machangu A and Minga U M 2002 Newcastle disease and Infectious bursal disease among free-range village chickens in Tanzania. In: Characteristics and parameters of family poultry production in Africa. IAEA, Vienna, Austria pp107 - 116.

Yu Li, Wang Z, Jiang Y, Chang L and Kwang J 2001 Characterisation of newly emerging Newcastle disease virus isolates from the people's republic of China and Taiwan. Journal of Clinical Microbiology 39, 3512 - 3519

Received 28 July 2006; Accepted 28 August 2006; Published 3 October 2006

Go to top