High titers of antibodies to caprine arthritis-encephalitis virus (CAEV) envelope (Env) glycoproteins are found in infected goats developing a progressive arthritis. In order to identify linear B epitopes of the CAEV Env, which may be involved in the immunopathology of arthritis, we constructed a Agtll Env expression library. By combining library screening with sera from naturally infected Swiss goats with an enzyme immunoassay with overlapping peptides (pepscan), four group-specific epitopes could be precisely defined in the transmembrane envelope proteins: TM1 to TM4, including a conserved structure (TM3) that corresponds to the immunodominant epitope of human immunodeficiency virus type 1 and other lentiviruses. A panel of 190 CAEV naturally infected goat serum samples, obtained from animals with defined clinical status, was tested for reactivity to synthetic peptides corresponding to the TM epitopes in an enzyme-linked immunosorbent assay. Antibody reactivity to two epitopes was highly associated (TM3, P = 0.002, and TM4, P < 0.001) with the presence of clinically detectable arthritis. Such an association is absent for anti-Gag antibody. Antibodies to the immunodominant structures of the TM glycoprotein could thus have an important role in the immunopathogenic process leading to disease.

The distribution of caprine leucocyte antigens (CLA) in goats from four different breeds (n = 546) affected by caprine arthritis-encephalitis virus (CAEV)-induced arthritis were determined and compared breed for breed with those of infected but clinically healthy controls (n = 402). Differences in frequencies of some of the CLA specificities between the affected and control groups were found, but after correction of the ordinary P values for number of observed alleles, only the CLA Be7 specificity in the Saanen breed showed a significant deviation at the 0.05 probability level. Animals of the Saanen breed carrying this specificity are less prone to develop arthritis after CAE virus infection than goats lacking this specificity. Eleven groups (multiple-case families or halfsibling groups with at least two informative diseased offspring/group) were analyzed for manifestation of the disease and segregation of the parental haplotypes. The results of the maximum likelihood test of association (P less than 0.005) and the calculated high lod score value of 5.70 give evidence for linkage between the locus encoding the determined class I CLA alleles and a hypothetical locus (i) coding for genes responsible for arthritis resistance/susceptibility. The particular class I CLA allele associated with the disease susceptibility varied from family to family, however. These data provide the first evidence that CAE virus-induced arthritis in the goat is genetically influenced by the MHC system; they also suggest that susceptibility/resistance genes are not directly associated with the determined class I gene products but rather are in close genetic linkage.

The distribution of CLA class I and II antigens in CAE virus-infected, diseased and healthy Saanen goats has been investigated. Three class I antigens and two class II antigens showed a statistically different frequency in the two groups.

A genetic predisposition to resistance (R)/susceptibility (S) has been demonstrated for cowdriosis in certain goal lines. In order to identify genetic markers of R/S and to follow their transmission to the offspring, 4 groups of sires and dams were crossbred in 1991, 1992, and 1993: Rr, RS, SR, and SS. The offspring were challenged at the age of six months with subsequent challenge of the S parent. From 28 presumed S dams, 7 turned out R and from 2 presumed S sires, 1 was R and the other undefinable. The resistance rate was 76% for 21 RR kids and 45% for 20 RS kids. The R character seems to be better transmitted by dams than the S one. No difference was observed between R and S goats in terms of incubation period and body temperature level after challenge. There was no statistical effect found of sex, litter size, or birth mass on the R/S character; only a slight effect on daily mass gain was observed. Sperm has been collected from 4 S sires (died after challenge) and 5 R sires, and stored in liquid nitrogen in order to conceive additional SR and SS progeny. Several caprine leucocyte antigens seem to be genetic markers of R (Be 1, CLY) and of S (Be 9, Be 22, Be 23). Further powerful tools like microsatellites and probes defining other interesting loci possibly involved in disease pathogenesis are now at hand and will be included in future experiments.

The incidence of subclinical mastitis in Simmental and Simmental-Red Holstein cattle in relation to the bovine major histocompatibility complex (BoLA) was investigated. Quarter milk samples of 166 cows consisting of fifteen halfsib groups of different ages and lactation stages were analysed for somatic cell counts (SCC) and bacteriological infection. From these data the udder health status (UHS) of the animal was determined. Each cow was typed serologically for BoLA class I and class II specificities. The statistical evaluation for UHS was performed using a logistic regression model. The effect of BoLA on SCC was estimated by least square analysis. Animals carrying the BoLA class II haplotype "b" were significantly more affected by subclinical mastitis than those without this haplotype. Breed group showed a significant influence on both UHS and corrected mean logSCC (P < 0.0001 resp. P < 0.05). Lactation stage had a significant effect on SCC but only a weak influence on UHS (P < 0.0001 resp. P < 0.13). ZUSAMMENFASSUNG: Mögliche Assoziation zwischen dem bovinen Haupthistokompatibilitätskomplex und subklinischer Mastitis In dieser Studie wurde die Beziehung zwischen dem bovinen Histokompatibilitätskomplex (BoLA) und der Prävalenz subklinischer Mastitis bei Simmentaler und mit Red Holstein eingekreuzten Simmentaler Kühen untersucht. Bei 166 Kühen, aus 15 Halbgeschwistergruppen, unterschiedlichen Alters und in verschiedenen Laktationsstadien, wurden Milchproben entnommen und auf Zellzahl (SCC) und bakterielle Infektion untersucht. Aus diesen Daten wurde ein Eutergesundheitsstatus (UHS) definiert. Jede Kuh wurde serologisch für BoLA Klasse I und Klasse II Spezifitäten typisiert. Die statistische Auswertung für den UHS erfolgte mit einem logistischen Regressionsmodel. Der Einfluß von BoLA-Haplotypen auf SCC wurde mit der Least Square Analyse ermittelt. Tiere mit dem Klasse II Allel "b" zeigten mehr Euterprobleme als Kühe ohne dieses Allel. Die Rassegruppe übte sowohl auf den UHS wie auch auf den korrigierten Mittelwert der Zellzahlen einen signifikanten Einfluß aus (P < 0.0001 resp. P < 0.05). Der Effekt des Laktationsstadiums auf die Zellzahl war signifikant, aber für den UHS wurde nur ein schwacher Einfluß des Laktationsstadiums festgestellt (P < 0.0001 resp. P < 0.13).