Determining Matrilines by Antibody Response to Exotic Antigens

-The antibody responses of female Microtus pennsylvanicus inoculated with a series of antigens not normally encountered under field conditions were examined, and the kinetics of maternally acquired antibody loss in their offspring were determined. The initial antibody response in adults was rapid, peaking in 4-9 weeks, and long-lasting, with a half life of 4-5 months. Antibody levels in females were unaffected by parity, and more than one antigen could be given without affecting circulating-antibody titers. Antibody titers could be enhanced with additional inoculations. Maternal antibody in offspring increased until weaning at 3 weeks then declined exponentially. Minimal detectable titers were reached at 7-11 weeks. In many instances, maternal antibody remained detectable even after offspring reached adult (35 g) size. Examination of uninoculated wild-caught voles showed only one of 130 tests produced a "false" positive response. The exotic-antigen technique may be generally applicable for determining maternity in small, secretive mammals. Many current theories in population biology and genetics have at their foundations the relatedness, and the relative reproductive success of population members (Eomnicki, 1988; Mangel and Clark, 1988). Some of these theories were addressed in detailed studies of large mammals (LeBoeuf and Reiter, 1988; Packer et al., 1988), but the low population densities, long generation time, and vagility of these species make their use expensive, labor intensive, and of extremely long duration. In addition, populations of large mammals rarely are amenable to experimental manipulations. These difficulties are not associated with studies of small mammals, but the young usually are sampled only after they have terminated contact with their mothers. As a consequence, maternal-offspring relationships often are impossible to determine. A variety of techniques have been used to associate females and their offspring in the field; nest boxes (Goundie and Vessey, 1986; Smith and Sloan, 1988), trapping patterns (Beacham, 1979), sampling at burrow entrances (Jones, 1987), and patterns of allozymic variation (Chesser, 1983) have been tried but none is generally applicable. Rongstad (1965), and Wolff and Holleman (1978) showed that female rodents injected with radioactive isotopes passed them to their infants during nursing. Tamarin et al. (1983) identified offspring by the characteristic spectrographic pattern associated with the decay of several different radionuclides. Scott and Tan (1985), Sheridan and Tamarin (1986), and Tamarin et al. (1983) used this technique to identify sucJ. Mamm., 71(2):129-138, 1990 129 130 JOURNAL OF MAMMALOGY Vol. 71, No. 2 cessfully the relative reproductive success of free-ranging Antechinus stuartii and Microtus pennsylvanicus. However, these workers reported that injected animals excrete radionuclides into the environment, raising potential problems with state and federal regulatory statutes concerning the disposal of radioactive materials (R. H. Tamarin, pers. comm.). Also, animals must be placed in a whole-body counter for 1.5-6 min. Time constraints could occur if large numbers of animals must be processed as all new animals must be examined before release. Additionally, females often eliminate most of their radionuclide burden by the end of weaning necessitating booster doses after each litter is born to make offspring identification unequivocal (Tamarin et al., 1983). Sampling-induced trap shyness may make it difficult to provide booster doses. Finally, it is not possible to mark both males and females with isotopes in the same study, as spectrographic patterns may overlap. Consequently, two separate methods are required; one to establish paternity, the other to determine maternity. If an alternative method could be developed to establish maternity then radionuclide marking could be used to establish mating success of males (Scott and Tan, 1985). An alternative method for marking offspring could take advantage of the humoral immune response of mammals whereby individuals exposed to an antigen react by producing antibodies specifically directed against the antigen (Klein, 1982). Immunoglobulins in pregnant and lactating females are transmitted to their offspring across their placentae and during nursing, providing the mechanism for passive immunization of infants by maternally derived antibodies (Brambell, 1970). By this process, we proposed to determine maternal-offspring relationships by immunizing females against antigens not encountered in the natural environment (exotic antigens), and measuring antibody profiles in their offspring. Herein, we outline laboratory studies of antibody responses to exotic antigens in breeding female M. pennsylvanicus, and report on the kinetics of maternal antibody loss in their offspring. Our goals were to examine the duration of antibody responses in mothers, the effects of multiple litters on their antibody responses, the possibility of increasing the antibody responses by providing boosters of antigens, the effect of immunizations with multiple antigens on antibody responses, the ability of the assay systems to detect differential responses to different but similar antigens, and the frequency of uninoculated wild voles with antibodies to these or related antigens. MATERIALS AND METHODS Female M. pennsylvanicus were obtained from a breeding colony at the U.S. Army Medical Research Institute of Infectious Diseases, Frederick, Maryland. Twelveto 24-week-old females were inoculated subcutaneously at two to four sites in their flanks with 200 ul of antigen-adjuvant mixture. Antigens were used at a concentration of 1 mg/ml dissolved in sterile saline and mixed with equal volumes of complete Freund's adjuvant. Females were inoculated with either tetanus toxoid, diphtheria toxoid, bovine thyroglobulin, porcine thyroglobulin, or keyhole limpet (Megathura crenulata) hemocyanin as antigens. Antigens were obtained from commercial sources. Four to eight females were assigned to each group. Three females were inoculated with complete Freund's adjuvant alone, as negative controls. To examine the effect of multiple antigens on the primary immune response, seven females were inoculated with diphtheria and tetanus toxoids, simultaneously. After 3 (diphtheria-tetanus) or 6 (tetanus) months half of the females in the tetanus toxoid and the diphtheria-tetanus groups received a second (booster) immunization with 100 ul of the appropriate antigens in sterile saline. Freund's adjuvant was not used during these inoculations. One week after inoculations, females were paired individually with males. Cages were examined 17 days later and daily, thereafter, for offspring. The date of birth was noted for each litter. Body weights for most offspring were recorded to the nearest 0.5 g beginning at 2 weeks of age, and at biweekly intervals thereafter. The presence of circulating immunoglobulin-G antibodies was determined from serum samples. Females were bled (100 M1) from their infraorbital sinuses, by capillary tube, before inoculation and at biweekly intervals unless newborn offspring were present. Offspring were bled, with their mothers, at 2 weeks of age. Blood samples were centrifuged, the sera collected, and stored at -20"C until tested. Antibody titers were determined by indirect enzyme-linked immunosorbent assays optimized for each antigen (Drummond et al., 1985; Voller and Bidwell, 1986). Antigens (2 Mg/ml) dissolved in a carbonatebicarbonate buffer (pH = 9.6), were coated onto 96-well microtiter plates by incubating overnight at 40C. May 1990 GLASS ET AL.-DETERMINING MATRILINES BY ANTIBODY RESPONSE 131 Plates were stored for 52 weeks under these conditions. Before use they were washed in phosphate buffered saline. Empty binding sites were blocked by incubation with 5% bovine serum albumin in phosphate buffered saline at 37*C for 1 h. Two microliters of vole sera were diluted 1:100 in 1% bovine serum albumin and phosphate buffered saline, added to each plate, and serially diluted in two-fold steps to 1:12,800. Plates were incubated at 37*C for 1 h, washed, and 100 Al of alkaline phosphatase-conjugated goat antimouse immunoglobulin-G (1:250 in phosphate buffered saline) added. Plates were incubated for 2 h at 370C, washed and incubated with p-nitrophenyl phosphate (0.5 mg/ml) dissolved in diethanolamine for 0.5 h at room temperature to visualize the reaction. Reactions were stopped by adding 50 ul of 3 M NaOH. Endpoint titers were determined by measuring optical densities of test reactions in a spectrophotometer, and fitting a polynomial curve to the graph of the optical densities as functions of dilutions. Antibody titers of sera were defined as the inverse of the dilutions >3 SD above the value of the negative control serum. Kinetics of primary antibody responses were described by graphing the average titer ? 95% confidence intervals against time for each experimental group. Dynamics of maternal-antibody loss in offspring of inoculated females were determined by regressing the average logarithm (base 2) of the titer against time. Antibody profiles generally follow an exponential decay over time (Brambell, 1970). Thus, the expected fit was a straight line with a negative slope, where the x-intercept defined the expected time that detectable maternal antibody disappeared from the circulations of the offspring. The specificity of the antibody responses was examined by comparing antibody titers of animals inoculated with porcine or bovine thyroglobulins against the heterologous antigens. Sera from 26 wild-caught voles (Childs et al., 1987) were screened at 1:100 dilutions against each antigen to determine if naive wild animals have antibodies to these or other cross-reacting antigens. Sera with optical densities >3 SD units above the mean of three negative control sera, included on each plate, were titrated to endpoint.