AU5621394A - Methodology for developing a superior line of domesticated animals - Google Patents

Description

METHODOLOGY FOR DEVELOPING A SUPERIOR

LINE OF DOMESTICATED ANIMALS

FIELD OF THE INVENTION

This invention relates to a methodology for

developing a superior line of domesticated animals, particularly animals which yield food value and more particularly a methodology which provides an animal line of increase productivity and generally healthier through to maturation.

BACKGROUND OF THE INVENTION

Considerable efforts have been expended in

attempting to develop, by way of genetic selection, animal strains which are of a more robust variety and generally yield higher quality meats for food purposes. The following is a list of several references in which investigations have been made in this regard, as well as corresponding methods and diagnosis used in determining various genetic traits.

Appleyard,G. ,B.N.Wilkie,B.W.Kennedy, and

B.A.Mallard. 1992. Antibody avidity in Yorkshire pigs of high and low immune response groups. Vet . Immunol , and Immunopathol . 31: 229-240

Archer,R.K., and L.B.Jeffcott. 1977. Comparative Clinical Haematlogy. Blackwell, Oxford.

Bendixen, P.H. 1981. Reversible detachment of blood-derived bovine macrophages by replacement of culture medium with phosphate buffered saline solution. Am J Vet Res 42:687-688.

Berche, P.A. 1985. Resistance to listeriosis in two lines of mice genetically selected for high and low antibody production. Immunology 56:707-716.

Biozzi, G., Mouton, D., Sant'Anna, O.A., Passos, N.C., Gennari, M., Reis, M.H., Ferreira, V,C.A., Neumann, A.M., Bouthillier, Y., lbanez, O.M.,

Stiffel, C, and Siqueira, M. 1979. Genetics of immunoresporisiveness to natural antigens in the mouse. Curr Top Microbiol Immunol 85:31-98 . Biozzi, G., Mouton, D., Stiffel, C., and

Bouthillier, Y. 1984. A major role of the

macrophage in quantitative genetic regulation and immunoresponsiveness and antiinfectious immunity. Adv Immunol 36:189-233.

Biozzi G., D.Mouton, C. Stiffel, and Y.Bouthiller. 1984. A major role of the macrophage in quantitative genetic regulation of immunoresponsiveness and antiinfectious immunity. Advances in Immunology 36: 189-234.

Biozzi, G., Mouton, D., Siqueira, M., and Stiffel, C. 1985. Effect of genetic modification of immune responsiveness on anti-infection and anti-tumor resistance. In Genetic Control of Host Resistance to Infection and Malignancy, pp 3-18, Alan R. Liss, Inc.

Blackwell, J. 1989. The macrophage resistance gene Lsh/Ity/Bcg. Twenty-seventh forum in Immunology 140:767-828.

Burton, J.L., Kennedy, B.W., Burnside, E.B., Wilkie, B.N., and Burton, J.H. 1989. Variation in serum concentrations of immunoglobulins G, A, and M in Canadian Holstein-Friesian calves. J Dairy Sc! 72:135-149.

Buschmann, H., Krausslich, H., Herrmann, H., Meyer, J., and Kleinschmidt, A. 1985. Quantitative

immunological parameters in pigs - experiences with the evaluation of an immunocompetence profile.

Tierzuchtg. Zuchtgsbiol 102 : 189199.

Cho H.J., H.L.Ruhnke, and E.V.Langford. 1976. The indirect hemagglutination test for detection of antibodies in cattle naturally infected with

mycoplasmas. Can . J. Comp. Med. 40: 20-29.

Corbeil, L.B., Watt, B., Corbeil, R.R., Betzen, T.G., Brownson, R.K., and Morril, J.L. 1984.

Immunoglobulin concentrations in serum and nasal secretions of calves at the onset of pneumonia. Am J Ver Res 45(4):773-778.

Covelli, V., Mouton, D., Di Majo, V., Bouthillier, Y., Bangrazi, C, Mevel, J., Rebessi, S., Doria, G., and Biozzi, G. 1989. J Immunology 142:12241234.

Fumoux, F., Traore-Leroux, T., Queval, R., Pinder, M., and Roelants, G.E. 1985 High and low

responsiveness of bovine lymphocytes to Trypanosome Jbrucei in vitro: lack of correlation with resistance to trypanosomiasis. Immunology 54:195-203.

Covelli V., D.Mouton , V.di Majo, Y.Bouthillier, C.Bangrazi, J.-C.Mevel, S.Rebessi, G.Doria, and G.Biozzi. 1989. Inheritance of immune

responsiveness, life span and disease resistance in interline crosses of mice selected for high or low multispecific antibody production. J. Immunol . 142: 1224-1234.

Devey,M.E., P.J.Major, K.M.Bleasdale-Barr,

G.P.Holland, M.C.Dal Canto and P. Y.Paterson. 1990. Experimental allergic encephalomyelitis (EAE) in mice selectively bred to produce high affinity (HA) or low affinity (LA) antibody responses. Immunology 69: 519-524

Edfors-Lilja,I., M.Bergstrom, U.Gustafsson,

U.Magnusson, and C.Fossum. 1991. Genetic variation in Con A-induced production of interleukin 2 by porcine peripheral blood mononuclear cells. Vet . Immunol . Immunopathol . 27: 351-363.

Edfors-Lilja,I., E.Wattrang, U.Magnusson and

C.Fossum. 1993. Genetic variation in parameters reflecting immune competence of swine. Vet . Immun . and Immunopathol . In press.

Erno,H., and L.Stripkovitz. 1973. Bovine

mycoplasmas: cultural and biochemical studies. II Acta Vet . Scand. 14: 450-463.

Gavora, J.Si 1990. Genetic disease resistance:

mechanisms and strategies for improvement. Proceedings of the 4th World Congress on Genetics Applied to Livestock Production, 427-436.

Gavora, J.S. and Spencer, L. 1983. Breeding for immune responsiveness and disease resistance.

Animal Blood Groups and Biochemical Genetics 14:159- 180.

Gill,H.S., G.D.Gray, D.L.Watson, and A.J.Husband. 1993. Isotype-specific antibody responses to

Haemonchus contortus in genetically resistant sheep. Parasite Immunology 15: 61-67.

Groves,T.C., B.N.Wilkie, B.W.Kennedy and

B.A.Mallard. 1993. The effect of selection of swine for high and low immune-responsiveness on monocyte superoxide anion production and class II MHC antigen expression. In press. Vet . Immunol . and

Immunopathol .

Harmen, B.G., Templeton, J.W., Crawford, R.P., Heck, F.C., Williams, J.D., and Adams, L.G. 1985.

Macrophage function and immune response of naturally resistant and susceptible cattle to Brucella

abortus . In Genetic Control of Host Resistance to

Infection and Malignancy, pp 345-354, Alan R. Liss,

Inc.

Helwig,T.T., and K.A.Council. 1979. SAS users guide. SAS Institute, Raleigh,N.C.

Hopkins, S.J. and A.Meager. 1988. Cytokines in synovial fluid: II The presence of tumour necrosis factor and interferon. Clin . exp. Immunol . 73: 8-92. Ibanez, O.M., Reis, M.H., Gennari, M., Ferreira, V.C.A., Sant'Anna, O.A., Siqueira, M., and Biozzi, G. 1980. Selective breeding of high and low

antibody responder lines of guinea pigs.

Immunogenetics 10:283-293.

Kennedy, B.W. 1990. Use of mixed model in analyses of designed experiments. In: Advances in

statistical methods for genetic improvement of livestock. D. Gionola and K. Hammond (eds.).

Springer-Verlag, pp.77-97.

Kennedy, B.W. and Sorensen, D.A. 1988. Properties of mixed model methods for prediction of genetic merit. In: Proc. 2nd Int. Conf. Quant. Genet. B.S.

Weir, E.J. Eisen, M.H. Goodman and A. Namkoong

(eds.). Sinaur, Sunderland, Mass., pp 91-103.

Lacey, C. , Wilkie, B.N. , Kennedy, B.W., and

Mallard, B.A. 1989. Genetic and other effects on bacterial phagocytosis and killing by cultural peripheral blood monocytes of SLA-defined miniature pigs. Animal Genetics 20:371-382.

LaGrange, P.H. , and Hurtrel, B. 1985. Listeria monocytogenes infection in Biozzi mouse lines with High or low responses to phytohemagglutinin.

Cellular Immunology 96:210-222.

Lai,W.C, M.Bennett, Y.-S.Lu, and S.P.Pakes. 1991.

Vaccination of Lewis rats with temperature-sensitive mutants of Mycoplasma pulmonis . Adoptive transfer of immunity by spleen cells but not sera. Infect.

Immun. 59: 346-350.

Lie, O. 1990. Genetics of disease resistance.

Proceedings of the 4th World Congress on Genetics

Applied to Livestock Production, 421-425.

Lillehoj , H. S. , Ruf f , M. D. , Bacon, L. D. ,

Lamont, S. J. , and Jeffers, T. K. 1989.

Genetic control of immunity to Eimeria tenella . Vet

Immunol and Immunopath 20:135-148.

Mallard, B.A. , Burnside, E.B. , Burton, J.H. , and Wilkie, B.N. 1983. Variation in serum

immunoglobulins in Canadian Holstein-Friesians. J

Dairy Sci 66:862-866.

Mallard, B.A. , Wilkie, B.N. , and Kennedy, B.W.

1989a. Genetic and other effects on antibody and cell mediated immune response in SLA-defined

miniature pigs. Animal Genetics 20:167-178. Mallard, B.A., Wilkie, B.N., and Kennedy, B.W.

1989b. The influence of the swine major

histocompatibility genes on variation in serum immunoglobulin concentration. Vet Immunol and

Isopath 21:139-151.

Mallard, B.A. , Wilkie, B.N. , and Kennedy, B.W.

1989c. Influence of major histocompatibility genes on serum hemolytic complement activity in

miniature swine. Am J Vet Res 50(3) :359-363.

Mallard, B.A., Wilkie, B.N., and Kennedy, B.W.

1989d. Genetic selection for improved immune response in Yorkshire pigs. Immunobiology

(Supplement) Vol 4:104-105.

Mallard, B.A., Wilkie, B.N., and Kennedy, B.W. 1990. Variation in number of circulating leukocytes in lines of pigs selected for high and low immune and innate resistance mechanisms. (Abstract) Seventy- first Conference of Research Workers in Animal

Diseases, Chicago, Nov. 5-6, pp 22.

Mallard,B.A., B.N.Wilkie, S.Rosendal, and S.Keay. 1991. Induction of tumour necrosis factor alpha in SLA-defined mini pigs. Abstract. Proceedings of the 72nd Conference of Research Workers in Animal

Disease, Chicago, November 11-12.

Mallard B.A., B.N.Wilkie, B.W.Kennedy, and

M.Quinton. 1992. Use of estimated breeding values in a selection index to breed Yorkshire pigs for high and low immune and innate resistance factors. Anim. Biotech . 3(2): 257-280.

Mazengara, K.E., Kennedy, B.W., Burnside, E.B., Wilkie, B.N., and Burton, J.H. 1985. Genetic parameters of bovine serum immunoglobulins. J Dairy Sci 68:2309-2314.

Meyer, K. 1988. DFRF-ML - a set of programs to estimate variance components under an individual animal model, J . Dairy Sci . 71 (Suppl.) 2:33-34. Morris, R.S. 1988. The effects of disease on productivity and profitability of livestock: How should it be assessed? Proc New Zealand Soc of An Prodn 48:117-125.

Morse E.E., S.Panek, and R.Menga. 1970-71. Automated fibrinogen determination. Am .J. Clin .Pathol . 55: 671- 676.

Mouton, D., Heumann, A., Bouthillier, Y., Mevel, J., and Biozzi, G. 1979. Interaction of H-2 and Non H-2 linked genes in the regulation of antibody response to a threshold dose of sheep erythrocytes.

Immunogenetics 8:475486.

Mouton, D., Stiffel, C., and Biozzi, G. 1985.

Genetic factors of immunity against infection. Ann Inst Pasteur 1 Immunol 136D: 131-141.

Mulchany G., E.Reid, R.D.Dimarchi, and T.R.Noel.

1992. Maturation of functional antibody affinity in animals immunized with synthetic foot-and-mouth disease virus. Res . in Vet . Sci . 52: 133-140

Nicholas, F.W. 1987. Veterinary Genetics, Oxford Science Publications .

Niewland, M.G.B., Kreuknief, M.B., Hepkema, B.G., Pinard, M.H., and Van der Zijpp, A.J. 1989.

Breeding for high and low antibody production in chickens: Effects on disease resistance, MHC- haplotypes and production traits. lmmunobiology, Supplement 4:106.

Roberts,E.D., W.P.Switzer, and F.K.Ramsey. 1963 a. Pathology of the visceral organs of swine inoculated with Mycoplasma hyorhinis . Am . J. Vet . Res . 24: 9-18. Roberts, E.D., W.P.Switzer, and F.K.Ramsey. 1963 b. The pathology of Mycoplasma hyorhinis arthritis produced experimentally in swine. Am . J. Vet . Res . 24: 19-31.

Rosenstreich, D.L., Weinblatt, A.C., and O'Brien, A. 1982. Genetic control of resistance to infection in mice. Critical Reviews in Immunology 3:263-330.

Ruuth,E. and F.Praz. 1989. Intractions between mycoplasmas and the immune system. lmmunol. Revs . 112: 133-160.

Saklatvala,J. 1986. Tumour necrosis factor α

stimulates resorption and inhibits synthesis of proteoglycan in cartilage. Nature (London) 322: 547- 549.

Salmi,A.A. 1991. Antibody affinity and protection in virus infections. Current Opinion in Immunology 3: 503-506

Sant'Anna, O.A., Ferreira, V.C.A., Reis, M.H.,

Gennari, M., Ibanez, O.M., Esteves, M.B., Mouton, D., and Biozzi, G. 1982. Genetic parameters of the polygenic regulation of antibody responsiveness to flagellar and somatic antigens of salmonellae. J . Immunogenetics 9 :191-205.

Schurr, E., Morgan, K., Skamene, E., and Gros, P. 1989. The search for a human homologue of the mouse Beg host resistance locus. Twenty - seventh fozrum in Immunology 140:767-828.

Sher,T., S.Rottem and R.Gallily. 1990. Mycoplasma capricolum membranes induce tumour necrosis factor α by a mechanism different from that of

lipopolysaccharide. Cancer Immunol . Immunother. 31: 86-92.

Simonsen, M. 1988. The MHC of the chicken, genomic structure, gene products, and resistance to

oncogenic DNA and RNA viruses. In B.N. Wilkie and P.E. Shewen (eds): Veterinary Immunology, pp 243- 253, Elsevier.

Skamene, E., Kongshavn, P.L., and Landi, M. 1980. Genetic control of natural resistance to infection and malignancy. London, New York. Academic Press. Sorensen, D.A. and Kennedy, B.W. 1983. The use of the relationship matrix to account for genetic drift variance in the analyses of genetic experiments. Theor. Appl . Genet . 66:217-220. Sorensen, D.A. and Kennedy, B.W. 1986. Analyses of selection experiments using mixed model methodology. J. Anim . Sci . 63:245-258.

Spencer, J.L. 1988. Marek's Disease and inherent resistance in chickens. In Biological Implications of Pathogenicity. Symposium on Resistance and

Susceptibility to Infection : A Genetic Interplay of Microbe and Host, pp 2.

Standal, N. 1979. Genetic change in the Norwegian Landrace pig population. Acta Agr Scand 29 : 139-144 . Steadham, E.M. , Lamont, S.J. , Kujdych, I. , and Nordskog, A.W. 1987. Association of Marek's Disease with Ea-B and immune response genes in sublime and F2 populations of the Iowa State Si Leghorn line. Poultry Sci 66:571-575.

Steward, M.W. , Reinhardt, M. C. , and Staines, N.A. 1979. The genetic control of antibody affinity . Evidence from breeding studies with mice selectively bred for either high or low affinity antibody production. Immunology 37:697-703.

Stiffel, C., Liacopoulos-Briot, M., Decreusefond, C., and Lambert, F. 1977. Selection genetique de souris d'apres l'intensite de la transformation blastique des lymphocytes induite par la

phytohemagglutinine. Ann. Immunol . Inst .

Pasteur 128C: 303-305.

Stiffel, C., Ibanez, O.M., Ribeiro, C.,

Decreusefond, C., Siqueira, M., Mouton, D., and Biozzi, G. 1987. Genetic regulation of specific and nonspecific components of immunity. Immmol Lett 16:205.

Tempelton J.W., D.M.Estes, R.E.Price, R. Smith, and L.G.Adams. 1990. Immunogenetics of natural

resistance to bovine brucellosis. Proceedings of the 4th World Congress on Genetics Applied to Livestock Production, 396-399. Thirkill, C.E., N.K.Tyler, and A.M.Roth. 1992.

Circulating and localized immune complexes in experimental mycoplasma-induced arthritis-associated ocular inflammation. Infect . Immun . 60: 401-405.

Thomson,B.M., G.R.Mundy, and T.J. Chambers. 1987.

Tumour necrosis factors α and β induce osteoclastic bone resorption. J. Immunol . 138: 775-779.

Van der Zijpp, A.J. , and Leenstra, F.R. 1980.

Genetic analysis of the humoral immune response of White Leghorn chicks. Poultry Sci 59:1363-1369.

Van der Zijpp, A.P. , Scott, T.R. , Glick, B. , and Kreukniet, M.B. 1988. Interference with the humoral immune response in diverse genetic lines of

chickens. Vet Immunol and Immunopath 20:53-60.

Vangen, O. 1979. Studies on a two trait selection experiment in pigs. Acta Agr Scand 29:305-319.

Vespa, L. , and Zwilling, B.S. 1989. Expression of I-A by macrophages from Bcg' and Beg' mice. J

Immunology 143:214-220.

Wakelin, D., and Donachie, A.M. 1983. Genetic control of eosinophilia. Mouse strain variation in response to antigens of parasitic origin. Clinical and Experimental Immunology 51:239-246.

Warner, CM., Meeker, D.L., and Rothschild, M.F.

1987. Genetic control of immune responsiveness: A review of its use as a tool for selection for disease resistance. J Animal Sci 64:394-406.

Wassom, D.L., and Kelly, A.B. 1990. The role of the major histocompatibility complex in resistance to parasite infections. Critical Reviews in Immunology 10 :31-52 .

Windon, R.G., and Dineen, J.K. 1981. The effect of selection of both sire and dam on the response of F, generation lambs to vaccination with irradiated Trichostrongylus colubriformis larvae . Int J

Parasit 11 :11-18 . Windon,R.G. 1993. Selection of sheep for

immunological responsiveness to the intestinal nematode Trichostrongylus colubriformis . In Workshop report: Selection experiments to alter disease resistance traits in domestic animals. In press. Vet . Immunol . and Immunopathol .

Wright, P. 1987. Enzyme Immunoassay : Observations on aspects of quality control. Vet Immunol and Immunopath 17:441-452.

For ease of reference to these references in the remainder of the specification, they will be identified by the first author and the date of the reference.

A variety of techniques have been investigated to provide for genetic selection of better strains of animals. In making such selection, particular attention has been paid to disease resistance or susceptibility thereto. Normally with domestic livestock, selection has been based primarily on production rather than disease traits. There is a continued push, however, to reduce health costs in developing livestock and to minimize chemical residues in the food chain. Such outside pressure has renewed interest in the selection of

beneficial strains of animals for food production. From a health standpoint, selection experiments have been based on a single trait selection and often antagonistic relationships surface between specific and non-specific indicators of such resistance, for example, antibody and macrophage function (Biozzi et al 1979; Harmon et al 1985).

Resistance to infection and disease is the outcome of complex interactions between virulence-related

attributes of the pathogen and a spectrum of innate and immunological mechanisms which have a quantitative basis and are under polygenic control (Lie, 0. 1990).

Population variation in resistance to specific diseases could be exploited in.disease control if heritability of the resistance-related attributes was sufficiently high to permit selective breeding (Gavora, J.S., 1990). There is however an apparent risk in altering population resistance to selected infectious diseases since the underlying resistance mechanisms may represent only some of those needed to enhance overall population resistance to common pathogens. Cattle remaining healthy when naturally infected with the intracellular pathogen

Brucella abortus , resistance to which is correlated with effective mononuclear phagocytic function, may be

relatively susceptible to other infectious diseases

(Templeton, J.W. et al. 1990).

Genetic and phenotypic correlations of this nature are not unusual in mammals and have been described in a number of species. For instance, in a study compiling immune profiles on a variety of German pig breeds a negative association was reported between per cent peripheral blood monocytes and lymphocytes (Buschmann et al. 1985). A similar correlation was seen in the Guelph pigs (r = -0.35, p ≤ 0.002) (Mallard et al. 1990).

Positive associations between lymphocyte blastogenesis using PHA, Con A, LPS and PWM have also been reported in pigs (Buschmann et al. 1985). No association however was apparent between antibody response to tetanus toxoid or ovalbumin and blastogenic response to PHA, Con A or PWM, but a positive correlation existed between antibody response and lymphocyte stimulation using LPS (Buschmann et al. 1985). In lambs selected for responsiveness to T. colubriformis there was again an indication of an inverse relationship between B and T-cell mitogen response. In this case the B-cell response (LPS stimulation) was greater in non-responder lambs, whereas the T-cell response (Con A stimulation) was greater in the responder progeny (Windon and Dineen, 1981). Conversely, in cattle selected for trypanoresistance or trypanosensitivity no correlation was noted between the level of resistance and lymphocyte stimulation by trypanosomes (Fumoux et al. 1985). In the Biozzi High and Low line mice selected for antibody response to SRBC (Selection I), and in cattle determined as resistance or susceptibility to B . abortus , an antagonistic relationship existed between antibody responsiveness and various indicators of macrophage activity (Biozzi et al. 1979, Harmen et al. 1985).

However, in chickens selected over 8 generations for high and low antibody response to SRBCS, innate macrophage activity appeared to be unaffected (van der Zijpp et al. 1988). It becomes obvious that selection for one trait almost always produces alterations in other parameters, but these associations are often species and breed dependent. At present it is therefore difficult to generalize as to the nature and direction of these relationships.

In accordance with this invention, a methodology has been developed for multi-trait selection for several indicators of disease resistance which have proven to yield a superior line of livestock. This methodology has particular application with pigs, but as well cattle, chickens and other valuable livestock.

In accordance with an aspect of this invention, a procedure is provided for ranking immune system

responsiveness in an animal to provide an Estimated

Breeding Value (EBV) indicator of the animal's level of ability to resist disease and ability to pass such disease resistance to offspring, such EBV indicator being useful in selecting animals to be bred in order to produce offspring which inherit said level of ability to resist disease. The procedure comprises:

i) testing an animal's response to at least two tests one of which is a general measure and the other antigen specific which determine heritable humoral immunity traits;

ii) testing the same animal's response to at least two tests one of which is a general measure and the other antigen specific which determine heritable cell-mediated immunity traits; iii) testing the animal's response to the two tests of humoral immunity traits and two tests of cell- mediated immunity traits beginning as soon as possible after the animal has been weaned from its mother and at a time chosen to negate effects of passive immunity;

iv) ranking the animal's EBV indicator relative to other ranked animals based on the animal's level of response to the tests.

According to another aspect of the invention, five traits may be selected on which diagnosis is based are:

1. a measure of antibody - serum IgG;

2. antibody response to hen eggwhite lysozyme

(HEWL);

3. cellular activity - blastogenetic response to concanavalinA (ConA);

4. cellular activity - cutanaceous delay type

hypersensitivity (DHT) to Bacillus Calmette Guerin (BCG) /purified protein derivative (PPD); and

5. indicator of ainnate monocyte function - uptake in killing of S. typhimurium .

In accordance with this invention, diagnosis based on these selected traits provides an estimated breeding value (EBV) which determines the superior line of pigs for providing enhanced productivity and of improved health.

In accordance with the more specific aspect of the invention based on these selected traits, the combined EBV can be calculated for each animal, such as pigs, where each pig is assigned to a high, low or controlled breeding group.

The methodology, according to this invention, may be applied to various livestock to yield a high quality strain of livestock which is disease resistant. It has been found that the traits evaluated in accordance with the above methodology is preserved in generation to generation where, as needed, further EBVs may be

calculated in the third and fourth generations to further direct the continued production of a high quality animal or a low quality animal.

Various aspects of the invention are described with respect to the drawings wherein:

Figure 1 is a bar chart showing the EBVs for a first generation of Yorkshire pigs;

Figure 2 is the combined EBV for first generation Yorkshire pigs; and

Figure 3 is a rate of gain in pigs selected for high-low immune responses.

Figure 4 Mean change in weight of M. hyorhinis challenged and control (Placebo) pigs of High (solid bars) and Low (hatched bars) immune response lines of pigs. Weight gain in control pigs is significantly

(p ≤ 0.01) greater in High than in Low line animals.

Weight loss is equivalent for animals of each line after infection.

Figure 5 Mean clinical scores of arthritis in

M. hyorhinis infected pigs of High (solid bars) and Low (hatched bars) immune response lines. Difference between lines is significant (D7 p≤0.001, D10 p≤0.05, D14

p≤0.05).

Figure 6 Erythrocyte sedimentation rate (ESR,

Figure 6a) and plasma fibrinogen concentration (Figure 6b) in M. hyorhinis infected pigs of High (solid bars) and Low (hatched bars) immune response lines. ESR is

significantly (p≤0.01) higher in Low than in High line pigs on D3 and elevated in both lines on days 7, 10 and 14. Fibrinogen is increased in pigs of both lines from D3 to D14 (p≤0.001) and on D10 the concentration in the Low line exceeds that of the High line (p≤0.05).

Figure 7 Serum antibody (passive haemagglutination) titer (log₂) in M. hyorhinis infected pigs of High (solid bars) and Low (hatched bars) immune response lines. Titer is significantly (p≤0.001) increased on D3 in High line pigs only and in pigs of both lines at subsequent times. High line pigs have significantly (p≤0.05) higher titers than Low line pigs on days 3, 7, 10 and 14.

Figure 8 Postmortem lesion scores in M. hyorhinis infected pigs of High (solid bars) and Low (hatched bars) immune response lines. Lines differ in scores of

peritonitis (p≤0.008), pleuritis (p≤0.001) and arthritis p≤0.002) .

Figure 9 Antibody responses of High-Low line pigs after vaccination with commercial bacterium against

Actinobacillus pleuropneumonia .

In order to demonstrate various aspects of the invention, the ranking procedure of this invention has been applied to pigs; however, it will become apparent that the basis of the ranking procedure may be applied to other types of livestock. The methodology has been applied to the development of a high line of pigs which exhibit increased production, better health where the desirable traits are passed on from one generation to the next. Such increase in productivity and health can result in, for example, one extra pig per litter, moving the pigs to market weight in less time, for example by as much as ten days without reduction in quality of the meat and that such animals respond in a superior way to various vaccination treatments.

The procedure, according to this invention, which provides an Estimated Breeding Value is very useful for ranking amongst breeding animals their immune system responsiveness. The Estimated Breeding Value is an indicator of the animal's responsiveness in resisting disease. The greater the animal's ability, that is level of ability to resist disease, then the higher the EBV. Furthermore the EBV indicator is based on the ability of the animal to pass such disease resistance to offspring. Hence the higher the ability of the animal to pass on this disease resistance correspondingly the higher the EBV for that animal. By way of ranking the immune syseem responsiveness of the animal, animals may be selected for breeding in order to produce offspring which inherit the higher or lower level of ability to resist disease.

Although the ranking procedure is important in respect of development of strains of animals that are disease resistant, it is also an aspect of the invention, however, to develop at the same time a line of animals which has a very low resistance to disease. Such low animals can be very useful in drug screening programs and other tests to determine efficacies of new drugs, vaccines and the like.

The ranking procedure is based on testing the animal's response to several, perhaps related or

unrelated traits, which may or may not relate directly to determining immune system responsiveness. These traits are selected from heritable humoral immunity traits and heritable cell-mediated immunity traits and perhaps one or more other traits which are inheritable, such as stress resistance. The heritable humoral immunity traits may be tested by at least two tests. One of the tests is usually a general measure of the immune response and the other test is usually an antigen-specific immune

response. Similarly with the heritable cell-mediated immunity trait, this can be tested by at least two tests which are correspondingly directed to either the general or antigen-specific types of tests. These tests are conducted as soon as possible on the animal after it has been weaned form its mother and at a time chosen to negate any effects of passive immunity; thereby ensuring the least amount of interference in respect of getting true values for these test results.

The humoral immunity tests more specifically may be selected from tests, such as:

1. serum concentration of immunoglobulin G (IgG) .

This is directed to the general measure of the animal's response in respect of immunity, traits. 2. The specific test may be an antibody response to an antigen which is not expected to be part of the antigens to which the animal and his parents have previously been exposed. For example, these tests may be:

a) peripheral blood lymphocyte blastogenic response to a mitogen not common to the animal and its parents, or

b) a cutaneous delayed type hypersensitivity (DTH) to an antigen purified protein derivative of a bacterium previously used to induce DTH. Specific examples for the antigen may be hen eggwhite lysozyme.

More particularly in respect of the peripheral blood blastogenic response test, the antigen may be concanavalin A. In respect of the antigen used to induce DTH, that may be Bacillus Calmette-Guerin (BCG) or purified protein derived in a human strain of Mycobacterium tuberculosis grown on a protein free synthetic medium. By carrying out these tests, the animals may be grouped into high, low or control. By virtue of this ranking procedure, the indicator divides all ranked animals into these indicated groups of high, low or control. The ranking therefore provides the basis for breeding together only test animals from the same group; that is only animals ranked high are bred with other animals ranked high and correspondingly with respect to the ranked low or control animals. Cross-breeding of a high animal with a low animal would not produce the desired result, because of the resultant degeneracy in the offspring insofar as developing an animal line with the increasing ability to resist disease and other advantages and features of this invention.

By virtue of the ranking procedure of this

invention, several advantages flow therefrom, particularly in respect of this application to the breeding of pigs, cattle, chickens and fish. The ranking procedure when used in selecting breeding animals from the high group provide offspring which achieve market weight consistently faster than offspring bred from control groups or bred from low groups. The offspring also have a higher percentage of live piglets per litter, such as when applied to pigs, also a lower number of litters with less than three piglets, a lower percentage of deformed piglets per litter and a higher production index. Furthermore, the animals ranked in the high and low groups differ in disease manifestations induced by infection. The animals ranked in the high and low groups differ in response to vaccination such that the animals ranked in the high groups respond earlier, produce more antibody against antigens and have a higher percentage of animals that respond to vaccination. The animals ranked in the high and low groups differ in response to

immunization, such that animals ranked in the high group produce more antibody to antigens in addition to those used to derive the original EBVs. This aspect of the invention will be described in more detail in respect of the following Tables. Furthermore the animals ranked in accordance with the procedures of this invention into the high and low groups differ in response to immunization, such that animals ranked in the high group produce antibody of higher binding strengths (avidity) for the antigen administered. This indicates overall a superior immune response by the animals in the high group.

As an example of an embodiment of the invention, approximately 120 first generation piglets (G1) were evaluated in accordance with the ranking procedure.

Based on G₀ plus G₁, heritability estimates were 0.25, 0.23, 0.08, 0.08 and zero for secondary antibody response to HEWL, blastogenic response to Con A, cutaneous DTH to BCG/PPD, serum IgG, and monocyte function, respectively. Least squares means reflected these estimates in that there were significant (p ≤ 0.05) differences between High and Low line G, pigs in antibody, blastogenic, and DTH responses. However, there were no significant line differences in serum IgG or uptake and killing of S .

typhimurium . Response to selection was determined both by differences in least squares means and differences in average EBV between the High and Low lines. After one generation of selection lines were separated by 1.508 (least squares) and 1.205 (EBV) index points, or slightly more than half a standard phenotypic deviation. The following materials and methods and assay procedures for the various traits are provided to illustrate various embodiments of the invention in carrying out the

methodology thereof.

MATERIALS AND METHODS - EMBODIMENT #1

Experimental Animals and Traits Measured

To determine the effectiveness of using a composite selection index for breeding high and low immune response lines, a random bred population of Yorkshire pigs housed in an Specific Pathogen Free (SPF) research facility

(Arkell Station, University of Guelph) was evaluated for various indicators of antibody, cell-mediated and innate resistance-related traits. The initial screening

involved the evaluation of 65 female and 33 male piglets, beginning at approximately 60 days of age. One male and 2 female piglets were sampled from each of 34 litters by 15 sires. Each piglet was evaluated for total serum immunoglobulin G (IgG) and M (I&M), serum antibody response to Hen Egg White Lysozyme (HEWL) , a synthetic peptide (TGAL), and sheep erythrocytes (SRBC). Cellular immune response was assessed by measuring delayed type hypersensitivity (DTH) to a purified protein derivative (PPD) of Bacillus Calmette Guerin (BCG), as well as by contact sensitivity to dinitrochlorobenzene (DNCB). In addition lymphocyte proliferative responses to

concanavalin A (Con A) and PPD were assessed. Serum hemolytic complement activity (CH₅₀), and the ability of peripheral blood monocytes (PBM) to take up and kill Salmonella typhimurium were evaluated as indicators of innate resistance. Immunisation and Sampling Schedule

A standard protocol of immunization and sampling began on day 0 when piglets were about 60 days of age, since at this point interfering maternal antibodies would be minimal, particularly to inert antigens not previously encountered to negate thereby effects of passive

immunity. Piglets were weaned at an average of 21 days of age. To induce antibody production, HEWL (ICN

Biochemicals, Montreal, Quebec) and TGAL (ICN

Immunologicals, Lisle, 111.) were separately dissolved in phosphate buffered saline (PBS, 0.10 M, pH 7.4) such that 10 ug was contained in 1.0 and 0.10 ml respectively.

Solutions were emulsified with an equal volume of

Freund's complete adjuvant (FCA, Difco, Detroit, MI.) and injected intramuscularly (im) to different sites

delivering 10 ug of antigen to each pig. One ml of 40% v/v washed SRBCs in Alsever's solution was injected intraperitoneally (ip). Immunizations were repeated on day 14 using HEWL with FCA and TGAL without FCA and 0.5 ml of the SRBC suspension. Two 10 ml tubes of blood were collected from each pig on day 0 in order to evaluate pre-immunization serum titers, serum IgG and IgM

concentrations, and hemolytic complement activity (one tube kept on ice). Additional sera were collected on days 9, 14, 21, and 30 for titration of primary and secondary antibody responses and evaluation of CH₅₀ at the end of the immunization protocol (day 30).

On day 9, for induction of cell mediated immunity (CMI), the pigs received 0.5 mg of BCG (Connaught,

Willowdale, Ontario) suspended in 0.5 ml of PBS injected intradermally (id). Each pig also received 0.1 ml of 5% w/v DNCB (Sigma, St. Louis, Missouri) in 9.5% ethanol mixed with an equal volume of aqueous 90% dimethyl sulfoxide (DMSO, Fisher Scientific, Ottawa, Ontario) .

This mixture was applied to a designated skin surface area on the outside of the thigh and evaporated to dryness. On day 21 pigs received 0.1 ml (id) of 250 test units of PPD (Connaught, Willowdale, Ontario) and

topically 0.2 ml of 5% w/v DNCB in 4:1 acetone and olive oil. Negative control sites received 0.1 ml of PBS or 0.2 ml of 4:1 acetone and olive oil respectively.

Twenty-four hours later cutaneous responses were measured by double skin fold thickness.

On day 21, 35 mis of whole blood was collected in 50 ml conical tubes containing 5.0 mis of sodium heparin (Sigma, St. Louis, Missouri, 100 units/ml in PBS), and peripheral blood lymphocytes (PBL) isolated by Ficoll- Hypaque separation (specific gravity 1.077, 260

milliosmoles) in order to evaluate blastogenic response to Con A and PPD. On day 14 whole blood was collected in a similar fashion and adherent cells obtained for

assessment of uptake and killing of S . typhimurium .

Antibody Assays

An ELISA assay was performed to quantify primary and secondary antibody responses to HEWL and TGAL on days 0, 9, 14, 21, and 30 of the immunization schedule. Flat bottom polystyrene 96 well microplates (Dynatech,

Alexandria, Virginia) were coated with 100 ul/well of 1 x 10^-4 moles/liter HEWL (ICN Biochemicals, Montreal, Quebec) or 1 x 10^-8 moles/liter TGAL (ICN Immunologicals, Lisle, Illinois) and incubated 2 days at 4°C for HEWL or overnight for TGAL. Plates were washed (EL 403

microplate autowasher, Bio-tek Instruments, Guelph, Ontario) 3 times with 250 ul/well of wash buffer (0.05% Tween-20 in PBS, pH 7.4) then blocked with 200 ul/well of 3% Tween-20 and incubated at room temperature (rt) for 1 hour. Plates were again washed 3 times with the wash buffer and samples added. In addition, a control was provided which included the above reagents without the addition of the samples.

A 1:5 and a 1:125 dilution of each of the

aforementioned test sample was prepared using 0.05%

Tween-20 as diluent and 100 ul dispensed into each of 4 replicate wells using a quadrant system (Wright 1987). Replicates of pooled positive and negative pig sera were also included on each plate. Plates were incubated (2 hours, rt) and washed 3 times with wash buffer. A rabbit anti-swine IgG (H + L) alkaline phosphatase conjugate

(Sigma, St. Louis, Missouri) was diluted 1:1500 for HEWL and 1:1000 for TGAL in 0.05% Tween-20 and 100 ul

dispensed into each well. Plates were incubated (2 hours, rt) and washed 3 times. The substrate (p- nitrophenophosphates, Kirkegaard and Perry, Gaithersburg, Maryland) was dissolved in diethanolamine (10% in dH₂O) to a concentration of 1 mg/ml and 50 ul added to each well. Plates were incubated at 37°C until the absorbance reading (EL 311 Automatic ELISA Scanner, Biotek

Instruments) of the standard positive control is 1.0 at 405 run. This reading was generally achieved in 20 to 30 minutes. The mean, standard deviation and coefficient of variation were calculated for the 4 replicates of each dilution and the mean sum of the 2 dilutions recorded as the antibody titer.

Antibody responses to SRBCs were determined by a haemagglutination assay described previously (Mallard et al. 1989a) and serum IgG and Igm concentrations by single radial immunodiffusion (Mallard et al. 1989b).

Cellular Assays

Lymphocyte blastogenesis was evaluated on day 21 of the immunization schedule by measuring lymphocyte

proliferative response to the T-cell mitogen Con A and the antigen PPD. Blood was collected from each pig and Ficoll-Hypaque separated peripheral blood lymphocytes (PBLS) were suspended in culture medium (RPMI 1640 plus 20% FCS) to a concentration of 5 x 106 cells/ml. For each pig tested 100 ul of this cell suspension was dispensed into an entire row of a 96 well tissue culture plate (Nunclon, Gibco, Grand Island, NY). Wells in columns 1 through 4 then received an additional 100 ul of culture medium providing unstimulated controls (4

replicates/pig), whereas wells in columns 5 through 12 (8 replicates/pig) received 50 ul of culture medium plus 50 ul of either Con A (10 ug/ml in RPMI) or PPD (400 ug/ml in RPMI, M. bovis strain AN5 kindly provided by Dr. B. Stemshorn, Agriculture Canada). Plates were incubated (18 hours, 370°C) and then "pulsed" with 20 ul/well tritiated thymidine (25 pCi/ml in RPMI. Plates were covered with foil, incubated for a further 24 hours

(37°C) and then frozen (-20°C) until a convenient time for harvesting. To harvest cells, plates were thawed (30 minutes, 37°C), and cells deposited onto glass fiber filter discs (PHD Cell Harvester, Cambridge Technology Inc., Watertown, Ma). Filter discs were placed into scintillation vials with 5.0 mls of scintillation fluid and counts per minute (cpm) obtained (Packard, Liquid Scintillation System, Downers Grove, IL). The change in cpm was calculated as the mean of the stimulated wells (with mitogen or antigen) minus the mean of the

unstimulated control wells.

Delayed Type Hypersensitivity (DTH) to PPD and DNCB were determined 24 and 48 hours after challenge by calculating the increase in double skin fold thickness using the method described previously (Mallard et al. 1989a).

Nonspecific Resistance Assays

Total serum hemolytic complement (CH₅₀) activity was determined using day 0 (pre-immunization) and day 30 (post-immunization) samples. Blood was collected, held on ice and sera were harvested and frozen (-700°C) until time of analyses. Sera were then analyzed according to the method described previously (Mallard et al. 1989c). To. assess the ability of peripheral blood monocytes to take up and kill S. typhimurim, mononuclear cells were separated using a Ficol-Hypaque density gradient

(Specific gravity 1.077, 335 milliosmoles) and

resuspended to 5 x 106 cells /ml in culture medium. Ten ml aliquots of suspended cells were dispensed into tissue culture flasks (Nunc, Gibco, Grand Island, NY) and incubated (18 hrs, 37°C). Medium and nonadherent cells were decanted and adherent cells removed by adding 40°C PBS (Bendixen, 1981). Adherent leukocytes were

resuspended at 1 X 106 cells/ml in culture medium and uptake and killing of bacteria determined by a method described previously (Lacey et al. 1989) which is based on the reduction of 3 (4, 5-dimethylthiazoyl-2-yl) 2, 5 diphenyltetrazolium bromide (MTT, Pharmacia, Dorval, Quebec) to a purple formazan by bacterial dehydrogenases.

Statistical Analyses and Selection of Animals

The initial population (G₀) consisted of 65 females and 33 males from 34 litters sired by 15 boars. Because distributional properties of the traits were not known and the data set at G₀ was relatively small (n = 98), observations on all traits were standardized using rank normal scores. Estimates of heritabilities of the standardized records were then obtained by restricted maximum likelihood according to the following model:

Y_ijkl = μ + g_i + S_j + 1_jk+ e_ijkl

where Y_ijkl is a normal rank score on an immune response measure on the ijkl* pig, μ is the population mean, g_i is the fixed effect of the i^th sex of pig (male vs female), S_j is the random effect of the j^th sire (0, Iσ²,), l_jk is the random effect of the jk^th litter ~ (0, Iσ²j) and e_ijkl is the random effect of the ijkl^th pig ~ (0, Iσ² _e). The litter term contains the additive genetic contribution of the dam plus environmental and dominance genetic effects common to littermates. Solutions from the restricted maximum likelihood analyses were used to compute an estimated breeding value of each tested pig for each trait as follows (Kennedy and Sorensen, 1988) :

EBV_ikjl = s_i + kl_jk + h² _w (y_ijkl - g_i - s_j - l_jk)

where k = σ² _s / σ² ₁

and h² _w = 2σ² _s / 2σ² _s + σ² _e

Based on initial estimates of heritabilities and correlations between EBVs of the traits, five traits were chosen as selection criteria to be included in a

composite index. The objective was to include one specific and one general indicator of both antibody and cellular immunity and one indicator of innate resistance. The five traits were serum IgG, antibody response to HEWL, blastogenic response to Con A, cutaneous DTH to BCG/PPD and uptake and killing of S . typhimurium . A total EBV score on the five traits was combined in an index, as well these five traits can be considered independently.

The top, intermediate and bottom ranking 7 young boars, according to combined EBV score, were chosen as foundation breeding stock (G₀) to be sires of High (H), Control (C) and Low (L) line pigs respectively. Although 7 boars were chosen for each line, only 5 were actually used for breeding with 2 held in reserve in case of reproductive problems. similarly, the top, intermediate and bottom ranking gilts were selected and mated to H, C and L line boars respectively. There were 23, 21 and 19 H, C and L gilts. From each litter from these matings, 2 females and 1 male first generation (G₁) piglets were randomly sampled and evaluated according to the same immunization schedule as the parents.

Response to selection was calculated from the data of both generations by least squares and according to an animal model (Kennedy, 1990).

The least squares analysis was according to the following model: Y_{ijkl m} = μ + g_i + t_j + c_l + a_ijklm + e_ijklm

where y_ijklm, μ, and g_i are as defined previously, t_j is the effect of the j^th generation, l_k is the effect of the k^th line, and tl_jk is the line by generation interaction, and e_ijkl the random error. All effects, except the error, were fixed.

Response to selection was measured on the difference between least squares means for the H and L lines at generation 1. Standard errors of the least squares estimates do not account for drift variance and an adjustment for drift was made using the procedure of Sorensen and Kennedy (1983).

Heritabilities of the five traits and breeding values of the pigs of all lines and both generations were estimated under an individual animal model using the derivative free REML programs of Meyer (1988) according to the following model:

y_ijklm = μ + g_i + t_j + c_l + a_ijklm + e_ijklm

where y_ijklm, μ, g_i and t_j are as previously defined, c, is the common environmental effect due to littermates (0,

Iσ² _c), a_ijklm is the additive genetic (breeding) value of the ijklm* pig (0, Aσ² _a) and e_ijklm is the random environmental effect on the ijklm^th pig (0,lσ² _e).

The average genetic value of pigs of the jk^th generation and line was estimated as ∑_i,l,m a_ijklm / N_jk where N_jk is the number of pigs. Response to selection was estimated as differences in average estimated genetic value between the H and L lines in generation 1. Standard errors of response were according to Sorensen and Kennedy (1986).

Both the least squares and animal model analyses were done for the five selection traits on the data on the original scale of measurement because significant departures from normality were not found. After the analyses, the traits were converted to a standardized scale (μ = 0, σ = 1) and combined EBVs for the five traits were again computed and 5 boars and 21 gilts from each of the three lines were selected as breeding stock to produce the next generation.

Data has been gathered from two generation of pigs G₀ and G₁ to develop a production index. The results of these investigations are summarized in Table 9, where it is demonstrated that the high line of pigs reach 90 kilograms at 144 days versus the low line of pigs which reach the necessary weight in 154 days, such pigs having a greater amount of fat than the high line of pigs. This result in a production index of 156 versus 141 for the low line.

Several generations of pigs have been analyzed for reproduction data selected for the high and low immune response traits. As demonstrated in Table 10, there is a significant increase in the number of live piglets per litter, usually in excess of 1, with less poorly

performing sows and less deformed and mummified pigs.

This Table describes the average (mean +\-standard

deviation) number of live piglets born per litter, the number of sows (%) farrowing 3 or less piglets per litter, and the number of deformed and mummified piglets per litter (%) obtained from Guelph Yorkshire sows selectively bred for high and low immune responsiveness. This data was obtained from computerized record sheets kept on all pigs housed at the University of Guelph Arkel Research Center and significant differences between high and low lines are determined based on statistical t-tests and reported at either a 90% (*) or 95% (**) confidence level. Generation of selection is given down th left margin as G0 to G3.

Production data from pigs was also analyzed for high and low immune response with respect to several

generations. In achieving shipping weight of 90

kilograms in carrying out the production, back fat was measured. As summarized in Table 11, the high line of pigs achieved market weight in approximately 10 days less than the control line with less back fat to indicate a healthier, more productive line of pigs with higher quality meat. This Table presents the rate of weight gain on the basis of days required to reach 90K and backfat thickness in millimeters as determined using an ultrasonic probe of Guelph Yorkshire pigs selected for high and low immune responsiveness. The generation of selection is given down the left margin as G1, G2 or G3. The difference in rate of gain between the high and low line pigs (i.e., about 10 days) was reported as

significantly different using a statistical t-test at a 95% confidence level; i.e., p < 0.05 and is depicted using an ** symbol.

RESULTS FOR EMBODIMENT #1 Parental Generation (Go) : Arithmetic Means

Arithmetic means and the standard deviations of innate and immune response traits measured in boars and gilts selected as parents to produce High and Low

breeding lines are presented in Table 1. These provide baseline data on the traits measured.

Analysis of variance

Results from the least squares analysis of data from G₀ are summarized in Table 2. Results of these analyses indicated that the sire significantly contributed to the variation observed in secondary antibody response to HEWL and TGAL, serum IgG, DTH to PPD, and hemolytic complement activity (Day 0 and 30). The litter significantly influenced the secondary antibody response to HEWL, serum IgG and IgM, blastogenic response to PPD and Con A, and complement (day 0). The sex of the pig influenced the primary antibody response to HEWL, serum IgM and

complement activity (day 0 and 30). The analysis of variance for monocyte function showed that sire did not influence this trait, however age and group were

significant factors (Table 2). Correlation and Heritability Estimates

To determine which of the fourteen parameters measured might be most representative of the pigs overall potential to respond to subsequent infection;

correlations, analysis of variance, and heritabilities were taken into account. Based on this information, 5 traits were chosen to be included in the selection index as follows.

Correlations based on EBVs (Table 3) indicated a positive and significant correlation between primary and secondary antibody responses to TGAL and HEWL in both boars and gilts. There was also a slight positive, but nonsignificant correlation between antibody response to HEWL and TGAL. Due to these associations and because heritability estimates (Table 2) of secondary antibody responses were highest, the secondary response to HEWL was chosen as one of the indicators of humoral immunity.

Since serum IgG concentrations have previously been reported to be associated with disease incidence (Mallard et al. 1983, Burton et al. 1989, Corbeil et al. 1984), this was also included as an indicator of humoral immunity. Serum IgG tended to be negatively correlated with antibody responses to both HEWL and TGAL (Table 2), and therefore inclusion in the selection index should prevent IgG concentrations from significantly declining. The heritability of porcine serum IgG concentration was estimated at 0.15 in G₀ (Table 2).

Cutaneous DTH to PPD and DNCB tended to be

negatively but not significantly associated. In the parental generation these traits had heritability

estimates of 0.27 and 0.17 respectively (Table 2). The DTH response to PPD was included in the index as one indicator of CMI because it had the higher heritability, and the response to DNCB was positively correlated with serum IgG which was already marked for inclusion in the index. The lymphocyte proliferative responses to PPD and Con A tended to be positively correlated, as were the proliferative and DTH responses to PPD (Table 3) . The heritability estimates of the blastogenic responses to PPD and Con A in G₀ were 0.15 and 0.37 respectively. For these reasons blastogenic response to the mitogen Con A was chosen as the other indicator of cellular

responsiveness.

In terms of innate immunity, hemolytic complement activity on days 0 and 30 were significantly and

positively correlated. Preimmunization complement activity was also positively correlated with secondary antibody response to HEWL and serum IgG (Table 3). The heritability of complement activity was 0.13 and 0.31 pre (day 0) and post (day 30) immunization respectively. Due to the positive correlation with antibody response to HEWL, which was already included in the index, complement activity was not included in the index.

Monocyte function, measured as anti-bacterial capacity, was positively correlated with hemolytic complement activity (day 0), but negatively correlated with serum IgM (Table 3). The heritability of this trait was 0.18 in G₀ and because of the importance of the monocyte in both antibody and cellular immunity it was included as a parameter in the selection index.

Estimated Breeding Values of the Parents Selected to Produce High, Low, and Control Line Groups

The five EBVs for each of the selected traits were added to give each pig a total composite Estimated

Breeding Value. Boars and gilts were then ranked on this basis. The five top ranking fertile boars were assigned to the high group and had a mean total EBV of 0.24 ± 0.12. The intermediate five boars were assigned to the Control group (x = 0.02 ± 0.03), and the bottom five to the Low group (x = -0.16 ± 0.03) (Table 4). Similarly the high, intermediate, and low one third of the sows were assigned to H (x = 0.21 ± .10), C (x = -0.02 ± .06) and L (x = -0.18 ± .07) line breeding groups. Mean combined EBVs indicated significant differences (p ≤ 0.001) between the breeding groups (Table 4).

First Generation (G₁) : Means, Heritability and

Environmental Correlations

Estimates of means, phenotypic standard deviations, heritabilities, and environmental correlations among litter mates of the five selected immune and innate resistance factors were calculated using the combined data from the parents and first generation piglets (Table 5). The heritability of secondary antibody response to HEWL was estimated at 0.25 and the similarity amongst litter mates due to a common maternal environment (c²) was 0.23 (Table 5). Least squares means (Lsmeans) for antibody response to HEWL of first generation piglets showed a significant (p ≤ 0.04) difference between H and L lines (Table 6). The heritability of lymphocyte blastogenesis to Con A using the combined (G₀ + Gl₁ data was 0.23 and the environmental correlation was estimated at 0.24 (Table 5). Lsmeans indicated significant (p ≤ 0.03) differences between H and L line piglets for this trait (Table 6). The heritability of the DTH response to BCG/PPD was 0.08 and the environmental correlation was 0.28. Again there were significant (p ≤ 0.007)

differences in lsmean responses between H and L line piglets. The heritability estimate of serum IgG

concentrations using the combined data from both

generations had dropped from 0.15 based on G₀ data to 0.08, and the variation amongst littermates was 0.26 (Table 5). Lsmeans revealed no differences between H and L lines in either generation (Table 6). Similarly there were no differences in Lsmeans for monocyte function (Table 6) and the heritability was now estimated to be zero (Table 5). Analysis of Variance

Results from the least squares analysis performed using the combined data from the parental and first generation piglets indicated that the generation factor was significant (p ≤ 0.0001) for each of the five traits and that line, and generation by line also significantly contributed to the variation in each of the traits, except monocyte function (Table 7). There were no significant sex effects on any of these traits.

Estimated Breeding Values and Response to Selection

EBVs calculated on the five traits for each pig within a line, and additively combined EBV, are given in Figures 1 and 2. Mean total EBVs for the High, Control and Low lines were 0.66 ± .36, -0.04 ± .19, and -0.55 ± .31 respectively.

The response to selection was determined using both the least squares and animal model. The differences between H and L lines are reported in terms of original and standardized units (Table 7). The results showed that after one generation of selection the H and L lines were separated by 1.205 as measured by EBVS, and 1.528 as measured by least squares which is a little more than half a standard deviation of the index (Table 7). As expected from the heritability estimates the largest line differences occurred in antibody response to HEWL and blastogenic response to Con A.

Correlated Traits

Differences between the H and L lines were also apparent for several non-selected traits (Table 8) . High line pigs produced more antibody to TGAL (p ≤ 0.02) and SRBC (p ≤ 0.10) as measured on day 21 of the immunization schedule. Low line pigs however, had a substantially greater (p ≤ 0.07) DTH response to DNCB (Table 8).

Genetic selection, in accordance with the

methodology of this invention, enhances inherent disease resistance and increases livestock productivity and profitability without the hazards and costs associated with the traditional use of medications and vaccines. Recent analysis revealed that preventive treatment of animal disease yielded a much higher return on funds invested (500-1500%) than curative methods (Morris, 1988). In addition, High/Low and Resistant/ Susceptible lines provide a powerful tool to investigate mechanisms of host resistance to a variety of infectious agents. Genetically selected lines of pigs are also useful in gene mapping studies.

The multi-trait selection based on predictors of immune and innate resistance demonstrates that immunity in mammals is regulated not by one, but a complex network of factors.

A random bred population of Yorkshire pigs was characterized using various indicators of immune and innate resistance (Table 1) as described. Based on first estimates of heritability and correlations of estimates of breeding values (Tables 2 and 3), serum IgG and secondary antibody response to HEWL were selected as predictors of antibody responsiveness, while lymphocyte stimulation by Con A and DTH to BCG/PPD were chosen as predictors of cellular response, and innate resistance was evaluated in terms of monocyte function. EBVs for each trait were calculated using an animal model that makes use of all known relationships among animals, and pigs were ranked based on combined EBVs and assigned to High, Low or Control breeding groups. Approximately 40 first generation piglets from each line were then

similarly evaluated.

After one generation of selection there were significant differences between the H and L lines. In fact the H and L lines were separated by 1.205 units as measured by EBVs and 1.508 as measured by least squares EBVs (Table 7). Heritability estimates were 0.25, 0.23, 0.08, 0.08 and zero for antibody response to HEWL, lymphocyte stimulation by Con A, DTH response to BCG/PPD, serum IgG, and monocyte function respectively (Table 5). Least squares means (Table 6) and line differences, both in terms of original and standardized measures, (Table 7) reflected these heritability estimates in that the largest differences were apparent in the more highly heritable traits, whereas there were no significant differences between the lines in serum IgG concentration on monocyte function.

Traits not included as original selection criteria were also modified as a result of selection or drift.

Antibody responses to SRBCs and TGAL were significantly higher in the H line and lower in the L line (Table 8). This is similar to the nonspecific effects of selection described in the Biozzi mice (Biozzi et al. 1984).

Consequently the genetic regulation of antibody

responsiveness in the H and L line pigs is not restricted to a single antigen, but may operate at least in part as a more general level such as antigen presentation, rather than B-cell function alone. Conversely, immune based inflammatory response to DNCB was significantly higher in the L line pigs (Table 8). This parameter has also been previously reported to be negatively associated with antibody response in SLA defined miniature pigs (Mallard et al. 1989a).

In accordance with the methodology of this

invention, pigs and other types of livestock can be separated into high and low breeding lines using

aggregate EBVs. Based on the above-identified traits, the methodology of this invention provides a reproducible technique in determining the high line of pigs or other livestocks for breeding purposes to increase the

productivity and improve general health. With reference to Figure 3, this relationship is indicated where the high line of pigs achieves a 90 kilogram weight in first, second and third generations consistently faster than the control and low groups. Antibody avidity can also be investigated in the high and low response groups of animals as ranked by this invention. Antibody avidity is a measure of the

attractions of an antibody for an antigen; i.e., the quality of the antibody. It is possible to measure this value and has been done for a ranked group of pigs as determined by the procedure of Embodiment #1.

Avidity indices of antibody to hen eggwhite lysozyme (HEWL) were measured by chaotropic ion (SCN) elution enzyme-linked immunosorbent assay (ELISA) in pigs grouped as high control of low for various immune and innate resistance-related traits. The avidity index was the molar concentration of SCN- required to reduce by 50% the ELISA optical density value for a given serum. The index was independent of the amount of antibody.

Eight- to ten-week old Yorkshire pigs were immunized with HEWL and serum antibody measured by ELISA as one of five traits used ot assign them to high, low or control response groups. Serum antibody avidity for HEWL was evaluated on day 14 and day 30 after primary (day 0) and secondary (day 14) immunization. The effects of response group, gender, litter, serum IgG concentration and anti- HEWL antibody on avidity were determined using a linear model.

Antibody avidity indices varied amongst individuals. Mean avidity indices for sera collected on days 14 and 30 were 0.61 ± 0.43 and 1.22 ± 0.546, with maximum indices of 2.64 and 2.86 respectively. Avidity of secondary response antibody was significantly higher (p ≤ 0.05). Pigs of the high response group had significantly higher secondary antibody avidity than those of the control (p ≤ 0108) and low groups (p≤ 0.01). Avidity index was positively correlated with antibody to HEWL on days 14 and 30, but not to preimmunization serum IgG

concentration or to other measured traits.

Animals genetically selected as per Embodiment #1 to express high or low immune response or innate resistance- related traits are expected to differ in response to infection and in development of disease. Pigs were tested as selected for high (H) or low (L) expression of serum IgG, serum antibody to hen egg white lysozyme (HEWL), peripheral blood lymphocyte blastogenesis after in vitro stimulation with concanavalin A and cutaneous delayed- type hypersensitivity to PPD after sensitization with BCG. H and L pigs of generation G4 which differed

significantly in traits used in selection, were infected with Mycoplasma hyorhinis , in accordance with the following Embodiment #2, by intraperitoneal injection and compared using a split litter design on the basis of antibody response, nonspecific indicators of infection and disease signs, both antemortem and postmortem. The study period began on day -1 and continued for 14 days after infection (day 1). H line pigs produced

M.hyorhinis-specific serum antibodies earlier (day 3) (p≤0.001) and to higher titer (p≤0.05). Uninfected H line pigs gained more weight between days -1 and 14 (p≤0.001) but within the infected group weight loss was equivalent for pigs of both lines. Erythrocyte sedimentation rate was elevated on day 3 in L and day 7 (p≤0.001) in H and remained elevated in H and L. Fibrinogen was elevated by day 3 (p≤0.001) with L>H on day 10 (p≤0.005). H had more blood lymphocytes than L in the absence of challenge (p≤0.05). Both groups had reduced blood lymphocyte numbers (p≤0.05) by day 7 with H reverting to

prechallenge values by day 10 before L (p≤0.005). Blood neutrophil counts were higher in L than in H (p≤0.001) in the absence of challenge and were elevated in both groups (p≤0.01) after challenge without line-related

differences. The principal antemortem disease sign was arthritis which had onset at day 7 in H and 10 in L

(p≤0.001) with H>L on days 7 (p≤0.001), 10 and 14

(p≤0.005). Postmortem, H displayed less severe

peritonitis (p≤0.008) and pleuritis (p≤0.001) while pericarditis, although less in H was not signnficantly different. Arthritis was more severe in H (p≤0.002) with more M. hyorhinis in synovial fluid (p≤0.005) but not in blood. Synovial fluid antibody did not differ by line. These findings indicate that differences in response to infection have been induced by indirect selection for resistance-related attributes and that animals selected in this way have utility in husbandry as well as in experimentation directed towards vaccine development or investigation of disease pathogenesis.

As demonstrated in Embodiment #1, pigs have been genetically selected for high and low response using an index that combined estimated breeding values for serum IgG concentration, antibody response to hen egg white lysozyme, in vitro blastogenesis of peripheral blood lymphocytes stimulated with the mitogen con A and delayed type hypersensitivity induced by intradermal injection of tuberculin PPD after sensitization with BCG. By the third generation of selection these pigs differed

significantly for the traits incorporated in the

selection index and for certain other traits such as antibody response to unrelated antigens, antibody avidity and weight gain all of which favor the high response line.

To determine whether or not selection for

potentially resistance-mediating traits had altered capacity to resist infection and disease, pigs of the high and low response lines were experimentally infected with Mycoplasma hyorhinis (M. hyorhinis) and their

response to infection was assessed as described in the following Embodiment #2.

MATERIALS AND METHODS - EMBODIMENT #2

Animals. Forty four weaned Yorkshire pigs were 40 to 76 days old at the start of the experiment. The pigs were from the fourth (G4) generation of one of two breeding lines, selected for high (high line, H) and low (low line, L) immune response, respectively (Mallard, B.A. et al. 1992). In brief, the selection was based on an index that combined estimated breeding values for serum

concentration of IgG, antibody response to hen egg white lysozyme (HEWL), cutaneous delayed type hypersensitivity to PPD and lymphocyte blastogenesis to concanavalin A in vitro . Estimated breeding values were determined using all genetic relationships among individuals in the population (Mallard, B.A. et al. 1992).

Experimental design. The study was performed in three sets, comprising 8, 6 and 8 pairs of litter-mates, respectively. In each set, half of the pairs were from line H and the others from line L. Within each pair, one pig was challenged with M. hyorhinis while the other received phosphate buffered saline (PBS) as a placebo. The challenged and non-challenged groups of pigs were housed separately in adjacent rooms of an isolation unit. Jugular blood was collected into appropriate additives in evacuated tubes on the days immediately preceding and following challenge or placebo treatment and on days 3, 7, 10 and 14 thereafter (designated day -1, 0, 3, 7, 10 and 14). The pigs were euthanized on day 14 for

postmortem examination.

Challenge. The challenge strain of M. hyorhinis (497- 14) was originally isolated from joints of a naturally infected pig. The mycoplasmas were cultured in modified Hayflick's broth (Erno, H. et al. 1973), washed by centrifugation, resuspended in PBS and stored at -70°c. Pigs in the challenged group received a single i.p.

injection of 2X10⁹ M. hyorhinis in 2 ml PBS. The non- challenged pigs received PBS only. Antemortem observations. Arthritis was scored as described in Table 12. Pigs given score 3 were euthanized and assigned a score of 3 for each remaining day of the experimental period. Scoring was performed without reference to treatment by the same person throughout the study. Microbiology. At necropsy, 10 μl of blood collected in tubes with EDTA or 10 μl of synovial fluid were used to inoculate modified Hayflick medium agar in Petri plates for incubation at 37 C for 8 days. Colony counts were recorded as follows: no colonies = 0, 1-10 = 1, 11- 100 = 2, > 100 = 3. The maximum score for mycoplasma growth from 4 joints of l pig was thus 12 (4 X 3).

Clinical pathology. White blood cells were counted electronically (Coulter S +4, Coulter Electronics of Canada, Burlington, Ont. Canada) and differential cell counts were made on Wright's stained blood smears. The erythrocyte sedimentation rate (ESR) was determined by the Westergren method (Archer, R.K. et al. 1977). Plasma fibrinogen concentration was measured by coagulation analyzer (Fibrometer, Becton Dickinson, Canada Inc., Missisauga, Ont., Canada) according to Morse et al.

(Morse, E.E. et al. 1970-71).

Antibody titres to Mycoplasma hyorhinis . Indirect haemagglutination (Cho, H.J. et al. 1976), was used to titrate antibody to M. hyorhinis in serum and synovia. For statistical analysis, titers were converted to log₂ :

titre ≤ 1:2 = 0, titre 1:2 = 1, titre 1:4 = 2, titre 1:8 = 3, etc.

Postmortem observations. Pigs were euthanized by intravenous injection of pentobarbital sodium. Synovial fluids were obtained aseptically from the tarsal and carpal joints prior to macroscopic examination. The abdominal and thoracic cavities were examined for signs of serositis. All observations and scoring were done by a single person without knowledge of breeding line. Scores were assigned to reflect relative severity of response in each pig as described in Table 13.

Statistical analysis. Breeding line and other effects on response to mycoplasma-challenge were analyzed by least squares using the SAS general linear model (GLM) procedure (Helwig, T.T. et al. 1979). Normality of distribution of the data was assessed using the

univariate procedure (Helwig, T.T. et al. 1979) and for most traits log_e-transformations were required to obtain normal distributions for statistical analysis. For graphic presentation these data were converted to anti- log_e. For the data describing the course of the response to challenge, the S.E.M. at various days were compared and it was found possible to use pooled S.E.M.

The model used for analyzing the response during the course of the challenge was:

Y_ijklm =u + DAYj + TREATMENT_j + LINE_k + SET_l +

(TREATMENTxLINE)_jk + (DAYxTREATMENTxLINE) _ijk +

LITTER(LINExSET)_klm + ANIMAL(LITTER LINExSET TREATMENT )_jklmm + ERROR_ijklmm

where

Y_ijklmm = an observed value for a trait measuring the response to challenge;

u = population mean for the trait

DAY_i = a fixed effect due to day of observation; TREATMENT_j = a fixed effect due to treatment regime

(challenge/placebo);

LINE_k = a fixed effect due to breeding line;

SET, = a fixed effect due to experimental set;

LITTER (LINExSET)_klm

= a random effect due to litter nested within line and experimental set;

ANIMAL (LITTER LINExSET TREATMENT)_jklmm

= a random effect due to pig nested within litter, line, experimental set and treatment regime; ERROR_ijkmn = a random residual error term.

The animal mean square was used as the denominator to test the effects of treatment, treatment x line

interaction and litter on the various traits. The litter mean square was used as the denominator to test the effects of breeding line and set. The residual mean square was used as the denominator to test the remaining effects.

The model used for analyzing the data obtained at

necropsy was:

Y_ij=u + LINE_i + SET_j + ERROR_ij,

where all terms are as previously defined and y is the value of a trait recorded at necropsy.

RESULTS

Antemortem observations. The effects of M. hyorhinis challenge by line are presented in Table 14. Unless specifically indicated there were no differences between pigs of the H and L lines in the non-challenged group.

Non-challenged pigs of line H gained significantly (p≤O.01) more weight than those of line L from day -1 to day 14 (Figure 4). Weight loss in challenged pigs was approximately 1 kg and did not differ by line (Figure 4).

In the challenged group, two pigs of line H and two of line L were euthanized on day 10 due to severe

disease. The principal clinical sign of disease was arthritis which was observed earlier and with greater severity in H than in L line pigs. Arthritis was first observed on days 7 and 10 respectively in pigs of the H and L response lines resulting in a highly significant (p≤0.001) difference in clinical scores on day 7 (Figure 5). Scores also differed (p≤0.05) on days 10 and 14, the last day of antemortem observation.

The ESR increased significantly (p≤0.001) from day - 1 and 0 to day 3 in pigs of the L line and to day 7 in H line pigs (Figure 6a). The more rapid rate of increase in ESR in pigs of the low line resulted in a significant (p≤0.01) difference between the lines on day 3. The ESR remained elevated in pigs of both lines to the end of the study.

Blood fibrinogen concentration increased (p≤0.001) from day -1 and 0 to day 3 in pigs of each line (Figure 6b) and remained elevated. On day 10 however, the

fibrinogen concentration was significantly (p≤0.05) higher in L than in H line pigs.

The number of circulating lymphocytes decreased (p≤0.05) from day 3 to day 7 in pigs of each line. In those of the H line, lymphocytes reverted earlier to pre- challenge numbers than in L line pigs and there were significantly (p≤0.05) more blood lymphocytes in H than in L line pigs on day 10. In the non-challenged group there were at all times more (p≤0.05) blood lymphocytes in H than in L line pigs.

The numbers of circulating segmented neutrophils (PMN) increased significantly (p≤O.10) from day -1 and 0 to day 3 and remained elevated in pigs of both lines. On days -1 and 0, PMNs were higher (p≤0.05) in the L than in the H line pigs but did not differ by line after

challenge.

Serum antibody titres to M. hyorhinis increased

(p≤0.001) from day -1 and 0 to day 3 in H line pigs and to day 7 in pigs of the L line (Figure 7). Titres thus increased more rapidly in the H line pigs and were significantly (p≤0.05) higher than in pigs of the L line on days 7, 10 and 14.

Postmortem observations. There were no postmortem signs of disease in pigs of the non-challenged group. In pigs infected with M. hyorhinis , signs of both peritonitis and pleuritis were significantly (p≤0.008 and p≤0.001, respectively) more severe in pigs of the L than of the H line (Figure 8). The lines did not differ significantly in scores for pericarditis (Figure 8). Arthritis was significantly (p≤0.002) more severe in H than in L line pigs (Figure 8).

There was no difference between the lines in the numbers of M. hyorhinis cultured from blood collected at necropsy but significantly (p≤0.005) more M .hyorhinis grew in cultured synovial fluid of H than of L line pigs.

There was no difference between the lines in

synovial antibody titres to M. hyorhinis . DISCUSSION

Yorkshire pigs selectively bred to differ in a number of traits that may relate to their ability to resist infectious diseases (Appleyard, G. et al. 1992 Mallard, B.A. et al. 1992 and Mallard, B.A. et al. 1991) have been shown here to differ in a line-related manner in their response to challenge with a single pathogen, Mycoplasma hyorhinis . Infection was associated with antemortem signs of arthritis, pleuritis, pericarditis and peritonitis evident postmortem as expected for infection with this organism (Robert, E.D. et al. 1963a and Roberts, E.D. et al 1963b). Correlates of response suggest that pigs of the H line are more able than those of the L line to resist production of systemic disease signs such as peritonitis, pleuritis and pericarditis but that the L line pigs are less prone to develop arthritis. Protection against infection and disease due to

mycoplasmas is usually associated with production of antibody although in the case of M. pulmonis infection of rats, but not mice, adoptive transfer of resistance could only be achieved using immune spleen cells (Lai, W.C. et al. 1991) . In the present study, H line pigs produced mycoplasma-specific antibody earlier and to higher titers than did the relatively susceptible L line pigs.

The H lines of pigs, as developed from the foregoing Embodiment #1, realized general improvement in disease resistance indirectly by selecting for a range of specific and innate attributes reflecting both humoral and cell-mediated resistance-mediating mechanisms.

Insofar as the H line developed more severe arthritis than pigs of the L line in response to infection with a single pathogen; however, this development may be

overcome and may not be a significant result. For example, the two lines differ in a number of traits including those incorporated in the selection index and correlated traits (Mallard, B.A. et al. 1992), such as antibody production following immunization with other antigens, lytic complement activity and antibody avidity (Appleyard, G.B. et al. 1992), which may have influenced both resistance and development of disease, including arthritis. Antibody avidity is of interest since high avidity antibody has been shown to be most efficacious in mediating protection in virus infections (Mulchany G. et al. 1992 and Salmi, A.A. 1991) and antibody-dependent disease such as allergic encephalitis (Devey, M.E. et al. 1990). Pathogenesis of mycoplasma-associated diseases such as arthritis and uveitis may involve formation of inflammation-inducing antibody-antigen complexes

(Thirkill, C.E. et al. 1992). The relative ability of the H and L line pigs used here to produce toxic immune complexes is not known but favors the H line.

Since cutaneous delayed-type hypersensitivity to PPD of tuberculin was significantly higher in the H line pigs they may have a generally higher ability to produce inflammation based upon antigen-specific cell-mediated immune response which could have resulted in the more severe arthritis observed in the H line animals. Although serum antibody to the challenge agent was produced earlier and to higher titre in the H than L line animals, synovial fluid antibody titres were equivalent while the numbers of M. hyorhinis were greater in the joint fluid of H line pigs. This may suggest that within joints the lines do not differ with regard to antibody production and that antibody quantity may not influence the development of arthritis. However, there was no attempt in the present study to determine the relative amounts, isotype or avidity of M.hyorhinis-specific antibody present in serum or synovial fluid in response to

infection and if the lines differ in these traits this may have influenced the outcome of challenge. Given the fastidious nature of mycoplasmas and their dependence upon intimate association with host cells in order to obtain essential nutrients (Ruuth, E. et al. 1989) it may be that intraarticular growth conditions for mycoplasma were more favorable in H line pigs.

In that the Biozzi (Biozzi, G. et al. 1984 and

Covelli, V. et al. 1989) selection of mice on the basis of immune response resulted in lines divergent in

macrophage function it was surprising that in the present selection of pigs based upon a multi-trait scheme, including aspects of both cell-mediated and antibody response, there was no line-related difference in

macrophage uptake and killing of bacteria, production of oxygen metabolites or expression of MHC II gene products (Groves, T.C. et al. 1993). It is therefore unlikely that macrophage-mediated events such as phagocytosis, killing or cytokine release contributed to the observed

differences between H and L line pigs in response to challenge infection with M. hyorhinis . However, it cannot be excluded that differences do exist between the lines in ability of macrophages to be activated by mycoplasmas with subsequent production of monocyte or lymphocyte- derived inflammatory mediators; events which have been implicated in pathogenesis of mycoplasma-associated arthritis (Hopkins, S.J. et al. 1988; Saklatvala, J.

1986; Shet, T. et al. 1990 and Thomson, B.M. et al.

1987). We and others have confirmed that pigs which differ genetically may vary in their ability to produce cytokines (Edford-Lilja, I. et al. 1991; Edford-Lilja, I. et al. 1993 and Mallard, B.A. et al. 1991). Table 15 describes a number of resistance-related traits of the Guelph Yorkshire pigs (High, Low and

Control Lines) that are not part of the selection

criteria but nonetheless may be relied upon in predicting disease outcome. These traits include antibody response to a small synthetic peptide known as TGAL and sheep erythrocytes (SRBC), skin thickness response to a topical antigen dinitrochlorobenzene (DNCB), and serum haemolytic complement activity (CH50). The averages (means +\- standard deviations) are reported for particular days using specific tests which are described on the left margin under each test. Significant differences between the lines are determined based on statistical t-tests and are reported at either a 90% (*) or 95% (**) confidence level. Generation of selection is indicated as G1 or G3.

Figure 9 depicts least square mean values (i.e, means corrected for unequal sample size, litter effects, and sire effects) of antibody responses of Guelph High Low line pigs before (day 0) and after (days 14 and 21) vaccination with a commercial bacterium against

Actinobacillus pleuropneumonia . Actinobacillus

pleuropneumonia is a bacteria which causes acute and chronic pneumonia in pigs and presently costs the

Canadian pork producers about $4 million annually.

Antibody responses to this vaccine were measured using an Enzyme Immunoassay (ELISA) and units of response are given on the y-axis as optical density (OD) of the test sera at a predetermined optimal dilution of 1\800. The different letters above the bars of the graph indicate that the antibody responses are significantly different as determined using a statistical t-test and are reported at a 95% confidence level; i.e, p≤0.05. The nonresponder status reported at the right side of the graph indicates the percentage of pigs from each line (High, Low and Control pigs of Generation 4) which did not respond in any measurable way to this test vaccine. The experiments reported here confirm that

resistance to infectious disease can be altered

indirectly by breeding using selection for immune

response-related and other traits. There is a general tendency for resistance enhancement in the high responder animals but it is already evident that the H line animals have superior production performance in terms of weight gain (Mallard, B.A. et al. 1992. This may be due to improved resistance to subclinical infection and disease. The relatively low blood neutrophil numbers in the H animals may be an indicator of this while the higher lymphocyte counts of the H animals may reflect immune system fitness. Leucocyte numbers and neutrophil function have been shown to be heritable traits in animals, pigs (Edford-Lilja, I. et al. 1993).

The breeding selection, according to this invention, is predictably applicable to other animals such as cattle, sheep, chickens, fish, horses and other valuable livestock because all of these animals have similar response to the traits used in developing EDVs for ranking the animals for further breeding. Animals so developed can reduce husbandry costs through reduced requirements for health-related inputs such as

antibiotics and vaccines while enhancing product

wholesomeness by reduced use of extraneous materials . Such superior animals also have an apparent role in pathogenetic studies and in vaccine development and efficacy trials in which the L line animals may simulate problematic low responder individuals in outbred

populations. Hence, the selection procedure results in an animal model on which drug screening and the like may be conducted. There are also animal welfare implications if inherent ability to maintain health is improved in production systems and the selected lines allow

experiments to be conducted with fewer animals.

Although preferred embodiments of the invention are described herein in detail, it will be understood by those skilled in the art that variations may be made thereto without departing from the spirit of the invention or the scope of the appended claims.

a Significant differences between boars and gilts: * p≤ 0.05; ** p≤ 0.01. b nt = not tested

c Monocyte function is based on the ability of peripheral blood monocytes to take up and kill at 30 (T₃₀) and 90 T₉₀) minutes respectively

a Variance component estimation was by Maximum Likelihood.

b The significance of sire was tested using the type 111 mean square for litter

(Sire) as the error term.

c^* Traits chosen as actual selection criteria.

d ns = not significant (p > 0.10)

a IgG ; serum IgG, HEWL and TGAL antibody response to hen lysozyme and the synthetic peptide TGAL, PPD and DNCB (24 or 48 hr) = cutaneous DTH measured at 24 or 48 hours post challenge, PPD and Con A (blast) - blastogenic response, CH₅₀ = hemolytic complement activity, M' = monocyte ability to take up and kill S. typhimurium measured at 30 (T₃₀) and 90 (T₉₀) minute respectively.

b^* p≤ 0.01.

c nt = not tested.

a Combined EBV was calculated as the mean EBV of the five traits chosen for selection. The selection index for gilts was similarly determined.

₁

a Standard deviations are within generation, line and sex.

b Monocyte function is based on the ability of peripheral blood monocytes to take up and kill S. typhimurium at 30 and 90 minutes respectively.

₁

^a Monocyte function is based on the ability of PBMs to take up and kill S. typhimurium at 30 and 90 minutes respectively. ^b Probability of significant differences between High and Low lines within a generation.

^a

₁

₁

*High differs from Low line, p 0.10 ** High differs from Low line, p < 0.05 L y g

* **

Claims (14)

1. A procedure for ranking immune system responsiveness in an animal to provide an Estimated Breeding Value (EBV) indicator of the animal's level of ability to resist disease and ability to pass such disease resistance to offspring, such EBV indicator being useful in selecting animals to be bred in order to produce offspring which inherit said level of ability to resist disease, said procedure comprising:

i) testing an animal's response to at least two tests one of which is a general measure and the other antigen specific which determine heritable humoral immunity traits;

ii) testing the same animal's response to at least two tests one of which is a general measure and the other antigen specific which determine heritable cell-mediated immunity traits;

iii) testing the animal's response to said two tests of humoral immunity traits and said two tests of cell-mediated immunity traits beginning as soon as possible after the animal has been weaned from its mother and at a time chosen to negate effects of passive immunity;

iv) ranking said animal's EBV indicator relative to other ranked animals based on said animal's level of response to said tests.

2. The ranking procedure of claim 1, in which the humoral immunity tests are:

i) serum concentration of immunoglobulin G (IgG); ii) antibody response to an antigen not expected to be part of the antigens to which the animal and its parents have been previously exposed, and the cell- mediated immunity tests consisting of:

a) peripheral blood lymphocyte blastogenic response to a mitogen: and b) cutaneous delayed type hypersensitivity (DTH) to an antigen purified protein derivative of a bacterium used to induce DTH.

3. The ranking procedure of claim 2 in which the immunoglobulin measured in said humoral immunity test is IgG.

4. The ranking procedure of claim 2 in which the antigen in said antibody response test is hen eggwhite lysozyme.

5. The ranking procedure of claim 2 in which the mitogen in said peripheral blood blastogenic response test is concanavalin A.

6. The ranking procedure of claim 2 in which the agent used to induce DTH is Bacillus Calmette-Guerin and said purified protein derivative in said delayed type

hypersensitivity test is from a human strain of

Mycobacterium tuberculosis grown on a protein-free synthetic medium.

7. The ranking procedure of claim 2 in which the animal's indicator according to the animal's immune responsiveness is designated high, low or control.

8. The ranking procedure of claim 2, in which the animals are selected from the group consisting of pigs, cattle, fish and chickens.

9. The ranking procedure in claim 7 wherein said indicator divides all ranked animals into indicated groups of high, low or control, such ranking providing the basis for breeding together only test animals from the same group.

10. The ranking procedure of claim 9 wherein said animals are pigs ranked in said high group in accordance with this procedure and bred together to yield offspring which achieve market weight consistently faster than pig offspring from bred control groups or bred low groups, said pigs of said high group have a higher percentage of live piglets per litter, a lower percentage of litters with less than three piglets, a lower percentage of deformed piglets per litter and a higher production index.

11. The ranking procedure of claim 9 wherein said animals ranked in said high and low groups differ in disease manifestations induced by infection.

12. The ranking procedure of claim 9 wherein said animals ranked in said high and low groups differ in response to vaccination such that animals ranked in the high group respond earlier, produce more antibody to antigens and have a higher percentage of animals that respond to vaccination.

13. The ranking procedure of claim 9 wherein said animals in said high and low groups differ in response to immunization such that animals ranked in the high group produce more antibody to antigens in addition to those used to derive the EBV.

14. The ranking procedure of claim 9 wherein said animals ranked in said high and low groups differ in response to immunization such that the animals ranked in the high group produce antibody of higher binding

strength for an antigen.