MedimlHypothrse.s(1990)32,235-238 0 LMlgnm Group UK Ltd 1990
Major Histocompatibility Antigens and Spontaneous Abortion: An Evolutionary Perspective P. A. VERRELL”
and N. R. MCCABE+
Department of Ecology and Evolution, The University of Chicago, 940 E. 57th St, Chicago, IL 60637, USA and ‘Department of Pediatrics, Biochemistry and Molecular Biology, The University of Chicago, Box 82, 5841 S. Maryland Ave, Chicago, IL 60637, USA (reprint request to PAW
Abstract - Maternal-fetal histocompatibility as a consequence of sharing of MHC antigens between males and females may result in an increased incidence of spontaneous abortion. We propose that such a response may be of evolutionary significance in inbred populations where partners that are similar at MHC loci also are likely to share deleterious recessive alleles due to common descent. In this scenario, spontaneous abortion is regarded as an adaptive response on behalf of the mother, who terminates at an early stage her investment in an offspring which would likely be of low reproductive value if carried to term.
Introduction The mammalian immune system is under the control of a large number of highly variable, linked genes termed the major histocompatibility complex, or MHC (1). In humans, these genes (labelled HLA) occupy a segment of chromosome 6; in mice, analogous genes (labelled H-2) are found on chromosome 17. The function of the MHC in both species is best known through studies of transplantation, in which incompatibility of HLA and H-2 types between recipients and donors of tissues and organs results in the rejection of grafts and transplants. Genes within the MHC determine aspects of the activity of various components of the immune system; these include complement, cell-mediated
responses, intercellular communication in antibody production and T-cell factors. By.enabling the individual to recognize ‘self from ‘non-self’, the products of the MHC result in the direction of adaptive responses towards pathogenic organisms and other disease states, e.g., tumors. It has been suggested that the primary function of MHC antigens during evolution was in the context of development and morphogenesis, their function in defense arising later in vertebrate phylogeny (2, 3). The role of the MHC in reproduction has been the subject of recent empirical investigation. Studies have addressed the role of H-2 antigens in both behavioral and physiological aspects of reproduction in rodents. With respect to the former, studies of inbred mouse lines have shown
235
236 that individuals, especially males, often prefer to mate with females that differ from themselves at H-2 loci (4). Evidence suggests that male mating preferences are acquired experientially through preweaning exposure to parental, MHCcontrolled chemosignals (5). Physiological studies of mice, rats and hamsters have revealed that homozygosity between a mother and her fetus with regard to MHC antigens may result in pregnancy failure. With such MHC compatibility, the mother mounts an immune response towards her fetus, which may result in spontaneous abortion (6-9). MHC compatibility between the mother and her fetus also may be responsible for certain cases of spontaneous abortion in humans. In a study of spontaneously aborting women, those with recurrent episodes of, unknown etiology shared- significantly more HLA antigens with their husbands that did women whose abortions were of known etiology. Spousal MHC compatibility was correlated with hyporeactivity in mixed (wife-husband) lymphocyte culture reactions (10). It has been suggested that the expression of paternal HLA antigens in the fetus triggers the maternal production of blocking factors which protect the fetoplacental unit, but on/y if the paternal HLA antigens are sufficiently different from those of the mother (11). A final influence of HLA compatibility, especially at the DR locus, is on the sex ratio of infants born to homozygous couples, in which sons may outnumber daughters. H-Y antigen, which directs male sexual differentiation, may confer some protection to the feto-placental unit of sons (12). Might the spontaneous abortion of fetuses produced by MHC-compatible males and females be of any advantage? Extrapolating results from laboratory studies of rodents to natural, outbred populations, Palm (9) suggested that the spontaneous abortion of fetuses homozygous with their mothers would preserve heterozygosity in H-2 at the level of the population; a similar explanation has been advanced for the mating preferences of mice (5). It also has been postulated that the abortion of human fetuses homozygous with their mothers might be of evolutionarily importance in maintaining HLA antigen polymorphism in populations (13). Such an advantage might be greatest in inbred human populations. For example, studies of the Mataco Indians. of Argentina have revealed a relatively homogeneous gene pool and restricted polymorphism with respect to HLA antigens Historically, these people have lived in small
MEDICAL HYPOTHESES
communities (50 to 60 individuals) with high levels of inbreeding (14). However, evolutionary explanations of these types are group-selectionist as they posit an advantage to spontaneous abortion at a level higher than that of the individual organism; if we seek a functional explanation, it should posit advantages at the level of the individual (15). Could spontaneous abortion in females sharing MHC antigens with their mates be of any advantage at the level of the individual? Here we develop a plausible scenario in which spontaneous abortion due to MHC compatibility between related mates prevents the deleterious consequences of inbreeding. A new hypothesis Consider a population in which mating occurs between related males and females. In such a population, certain male-female pairs are likely to share MHC antigens by common descent. Due to their relatedness, these individuals also may share deleterious recessive alleles (perhaps linked to MHC genes) which could, in a homozygous state, result in inbreeding depression, (i.e., the production of offspring of low or negligible reproductive value). Each human gamete may carry the equivalent of as many as four or five lethal and three or four deleterious recessive alleles in heterozygous condition (16). It is well known that both the frequency of deleterious recessive alleles and the probability of prereproductive mortality increase with increasing degrees of consanguinity in human populations, demonstrating an increase in genetic load with increased inbreeding (16, 17). In our scenario in which matings are between related individuals, sharing of MHC antigens may result in the spontaneous abortion of a fetus which might otherwise have lived to be an offspring of low reproductive value (due to homozygosity for a deleterious recessive allele). Evidence that the rate of spontaneous abortion is related to the degree of consanguinity in human populations is both sparse and conflicting, although, as might be predicted, the trend appears to be a positive correlation (18, 19). Some of the best data relevant to the issue of HLA compatibility and reproductive performance in inbred human societies come from studies of the Hutterites, a religious sect in which communities consist of isolates of between 100 and 150 individuals. Comparing Hutterite couples in which spouses share HLA antigens at the A, B or DR loci with
MAJOR HlSTOCOMPATlBILITY
ANTIGENS’AND
SPONTANEOUS
those in which alleles are not shared, Ober et al. (13, 20) found evidence of reduced reproductive performance in the former group. Longer interbirth intervals were apparent in homozygous couples (the effect becoming stronger with increasing parity), and spontaneous abortion rates were higher in couples sharing DR alleles. We suggest that an advantage may accrue to the female which aborts under an inbred mating system, because she terminates investment in an offspring which, with inbreeding depression, would likely be of low relative reproductive value (16, 17, 21). We consider spontaneous abortion as potentially advantageous only in inbred populations, where matings between close relatives sharing HLA alleles and also carrying deleterious alleles are most likely. Any advantage should be greatest in populations in which the extent of inbreeding fluctuates temporally, such as many species of rodents in which population sizes boom and crash cyclically. Under constant inbreeding, deleterious consequences will be reduced as natural selection decreases the mutational load within the population (22). The hypothesis that, under certain circumstances, females might manipulate their pregnancies in order to avoid investing in offspring of low or negligible reproductive value is not new. A similar hypothesis was advanced regarding pregnancy failure in female rodents exposed to males other than their mates (a phenomenon known as the Bruce Effect). In many rodents in which such pregnancy failure occurs, males kill neonates to whom they are unrelated. The Bruce Effect can thus be seen as an adaptive device which enables a female to terminate, at an early stage, her investment in an offspring which will likely suffer infanticide after birth (23). Another example in which females apparently manipulate their own pregnancies in an adaptive manner is provided by those mammals in which the sex ratio of offspring at birth deviates from unity as a function of the expected reproductive value of sons versus daughters (24). With reference to clinical medicine, it is not our intention to imply that women who are HLAcompatible with their partners should not attempt to become pregnant using one of the several clinical procedures currently available (25). Most human populations are largely outbred; on average, the inbreeding coefficient of human populations is less than 1 per 1,000. Not withstanding cultural taboos prohibiting incest,
ABORTIONS:
AN EVOLUTIONARY
PERSPECTIVE
237
inbreeding coefficients are usually highest in isolates (16). In outbred populations, sharing of MHC antigens between spouses likely is due to random association. It is interesting to note that in those cases where MHC compatibility is circumvented clinically, no increased incidence of genetic abnormality is apparent among infants carried to term (25). In the scenario developed in the present paper, genetic defects in offspring are most likely to occur only if couples which share MHC antigens also share the same deleterious recessive alleles due to common descent. This situation is most likely to arise in inbred populations due to the high average relatedness between mating partners (this could be tested empirically by assaying for defects due to recessive alleles in abortus material). In more outbred populations, the genetic load will be lower (16). We are sensitive to the fact that we should be careful to construct neither naive adaptationist explanations nor untestable hypotheses (26). In addition, we believe care should be taken to distinguish between the “function” of a character and its “effects”, especially when considering adaptation (15). Nevertheless, we believe the relationships between MHC compatibility, reproductive failure and inbreeding to merit further study. Acknowledgment gratefully acknowledges financial support from the National Science Foundation (grant BSR 85-06766). PAV
References 1. Penagi G S, David C S (eds). Immunogenetics. Butterworths, London, 1984. 2. Bodmer W F. Evolutionary significance of the HL-A system. Nature 237: 139, 1972. 3. Klein J. Many questions (and about no answers) about the phylogenetic origin of the major histocompatibility complex. p. 467 in Immunologic Phylogeny (W H Hildemann, A A Benedict, eds) Plenum Press, New York, 1975. 4. Yamazaki K, Boyse E A. Mike V. Thaler H T, Mathieson B J, Abbott J, Boyse J, Zayas Z A, Thomas L. Control of mating preferences in mice by genes in the major histocompatibility complex. J. Exp. Med. 144: 1324, 1976. 5. Beauchamp G K, Yamazaki K, Bard J, Boyse E A. Preweaning experience in the control of mating preferences by genes in the major histocompatibility complex of the mouse. Behav. Genet. 18: 537, 1988. 6. Billington W D. Influence of immunologic dissimilarity of mother and foetus on size of placenta in mice. Nature 202: 317, 1964.
238 7. Beer A E, Scott J R, Billingham
8. 9. 10.
11. 12.
13.
14.
15.
R. Histocompatibility and maternal immunologic status as determinants of fetoplacental weight and litter size in rodents. J. Exp. Med. 142: 180, 1975. Chaouat G, Kolb J-P, Kiger N, Stanislawski M, Wegmann T G. Immunological consequences of vaccination against abortion in mice. J. Immunol. 134: 1594, 1985. Palm J. Maternal-fetal histoincompatibihty in rats: an escape from adversity. Cancer Res. 34: 2061, 1974. Beer A E, Quebbeman J F, Ayers J W T, Haines R F. Major histocompatibility complex antigens, maternal and paternal immune responses, and chronic habitual abortions in humans. Am. J. Obstet. Gynecol. 141: 987, 1981. Gall S A. Immunologic factors influencing pregnancy. Obstet. Gynecol. Ann. 14: 25, 1985. Radvany R M, Vaisrub N, Ober C, Pate1 K M, Hecht F. The human sex ratio: increase in first-born males to parents with shared HLA-DR antigens. Tissue Antigens 29: 34, 1987. Ober C, Martin A 0, Simpson J L, Hauck W W, Amos D B, Kostyu D D, Fotino M Allen F H. Shared HLA antigens and reproductive performance in the Hutterites. Am. J. Hum. Gen. 35: 994, 1983. Vu110 C M, Celis E M, Serra H M, Riera C M. Study of HLA system in a Mataco population: a geographically isolated American Indian tribe. Tissue Antigens 23: 33, 1984. Williams G C. Adaptation and Natural Selection. Princeton University Press, Princeton, 1966.
MEDICAL HYPOTHESES
16. Cavalli-Sforza L L, Bodmer W F. The Genetics of Human Populations. Freeman. San Francisco. 1971. 17. Schull W j, Neel J V. The effects of Inbreeding on Japanese Children. Harper and Row, New York, 1%5. 18. Goldschmidt E (ed). The Genetics of Migrant and Isolate Populations. Williams and Wilkins, New York, 1963. 19. Morton N E. Morbidity of children from consanauinous marriages. Progr. Med: Genet. 1: 261, 1961. 20. Ober C. Hauck W W. Kostvu D D. O’Brien E. Elias A. Simpson J L, Martin ‘A 0. ‘Adverse effects of HLA-DR sharing on fertility: a cohort study in a human isolate. Fert. Steril. 44: 227, 1985. 21. Stern C. Principles of Human Genetics (Third Edition). Freeman, San Francisco, 1973. 22. Futuyma, D J. Evolutionary Biology. Sinauer, Sunderland, 1986. 23. Schwagmeyer P L. The Bruce effect: an evaluation of male/female advantages. Am. Nat. 114: 932, 1979. 24. Clutton-Brock T H. Birth sex ratios and the reproductive success of sons and daughters. p. 221 in Evolution (P J Greenwood, P H Harvey, M Slatkin, eds) Cambridge Universitv Press. Cambridae. 1985. 25. Beer A E; Immunologic aspects of normal pregnancy and recurrent spontaneous abortion. Semin. Reprod. Endocrinol. 6: 163, 1988. 26. Gould S J, Lewontin R C. The spandrels of San Marco and the Panglossian paradigm: a critique of the adaptationist program. Proc. Roy. Sot. Lond. B 205: 581, 1979.
Major Histocompatibility Antigens and Spontaneous Abortion: An Evolutionary Perspective P. A. VERRELL”
and N. R. MCCABE+
Department of Ecology and Evolution, The University of Chicago, 940 E. 57th St, Chicago, IL 60637, USA and ‘Department of Pediatrics, Biochemistry and Molecular Biology, The University of Chicago, Box 82, 5841 S. Maryland Ave, Chicago, IL 60637, USA (reprint request to PAW
Abstract - Maternal-fetal histocompatibility as a consequence of sharing of MHC antigens between males and females may result in an increased incidence of spontaneous abortion. We propose that such a response may be of evolutionary significance in inbred populations where partners that are similar at MHC loci also are likely to share deleterious recessive alleles due to common descent. In this scenario, spontaneous abortion is regarded as an adaptive response on behalf of the mother, who terminates at an early stage her investment in an offspring which would likely be of low reproductive value if carried to term.
Introduction The mammalian immune system is under the control of a large number of highly variable, linked genes termed the major histocompatibility complex, or MHC (1). In humans, these genes (labelled HLA) occupy a segment of chromosome 6; in mice, analogous genes (labelled H-2) are found on chromosome 17. The function of the MHC in both species is best known through studies of transplantation, in which incompatibility of HLA and H-2 types between recipients and donors of tissues and organs results in the rejection of grafts and transplants. Genes within the MHC determine aspects of the activity of various components of the immune system; these include complement, cell-mediated
responses, intercellular communication in antibody production and T-cell factors. By.enabling the individual to recognize ‘self from ‘non-self’, the products of the MHC result in the direction of adaptive responses towards pathogenic organisms and other disease states, e.g., tumors. It has been suggested that the primary function of MHC antigens during evolution was in the context of development and morphogenesis, their function in defense arising later in vertebrate phylogeny (2, 3). The role of the MHC in reproduction has been the subject of recent empirical investigation. Studies have addressed the role of H-2 antigens in both behavioral and physiological aspects of reproduction in rodents. With respect to the former, studies of inbred mouse lines have shown
235
236 that individuals, especially males, often prefer to mate with females that differ from themselves at H-2 loci (4). Evidence suggests that male mating preferences are acquired experientially through preweaning exposure to parental, MHCcontrolled chemosignals (5). Physiological studies of mice, rats and hamsters have revealed that homozygosity between a mother and her fetus with regard to MHC antigens may result in pregnancy failure. With such MHC compatibility, the mother mounts an immune response towards her fetus, which may result in spontaneous abortion (6-9). MHC compatibility between the mother and her fetus also may be responsible for certain cases of spontaneous abortion in humans. In a study of spontaneously aborting women, those with recurrent episodes of, unknown etiology shared- significantly more HLA antigens with their husbands that did women whose abortions were of known etiology. Spousal MHC compatibility was correlated with hyporeactivity in mixed (wife-husband) lymphocyte culture reactions (10). It has been suggested that the expression of paternal HLA antigens in the fetus triggers the maternal production of blocking factors which protect the fetoplacental unit, but on/y if the paternal HLA antigens are sufficiently different from those of the mother (11). A final influence of HLA compatibility, especially at the DR locus, is on the sex ratio of infants born to homozygous couples, in which sons may outnumber daughters. H-Y antigen, which directs male sexual differentiation, may confer some protection to the feto-placental unit of sons (12). Might the spontaneous abortion of fetuses produced by MHC-compatible males and females be of any advantage? Extrapolating results from laboratory studies of rodents to natural, outbred populations, Palm (9) suggested that the spontaneous abortion of fetuses homozygous with their mothers would preserve heterozygosity in H-2 at the level of the population; a similar explanation has been advanced for the mating preferences of mice (5). It also has been postulated that the abortion of human fetuses homozygous with their mothers might be of evolutionarily importance in maintaining HLA antigen polymorphism in populations (13). Such an advantage might be greatest in inbred human populations. For example, studies of the Mataco Indians. of Argentina have revealed a relatively homogeneous gene pool and restricted polymorphism with respect to HLA antigens Historically, these people have lived in small
MEDICAL HYPOTHESES
communities (50 to 60 individuals) with high levels of inbreeding (14). However, evolutionary explanations of these types are group-selectionist as they posit an advantage to spontaneous abortion at a level higher than that of the individual organism; if we seek a functional explanation, it should posit advantages at the level of the individual (15). Could spontaneous abortion in females sharing MHC antigens with their mates be of any advantage at the level of the individual? Here we develop a plausible scenario in which spontaneous abortion due to MHC compatibility between related mates prevents the deleterious consequences of inbreeding. A new hypothesis Consider a population in which mating occurs between related males and females. In such a population, certain male-female pairs are likely to share MHC antigens by common descent. Due to their relatedness, these individuals also may share deleterious recessive alleles (perhaps linked to MHC genes) which could, in a homozygous state, result in inbreeding depression, (i.e., the production of offspring of low or negligible reproductive value). Each human gamete may carry the equivalent of as many as four or five lethal and three or four deleterious recessive alleles in heterozygous condition (16). It is well known that both the frequency of deleterious recessive alleles and the probability of prereproductive mortality increase with increasing degrees of consanguinity in human populations, demonstrating an increase in genetic load with increased inbreeding (16, 17). In our scenario in which matings are between related individuals, sharing of MHC antigens may result in the spontaneous abortion of a fetus which might otherwise have lived to be an offspring of low reproductive value (due to homozygosity for a deleterious recessive allele). Evidence that the rate of spontaneous abortion is related to the degree of consanguinity in human populations is both sparse and conflicting, although, as might be predicted, the trend appears to be a positive correlation (18, 19). Some of the best data relevant to the issue of HLA compatibility and reproductive performance in inbred human societies come from studies of the Hutterites, a religious sect in which communities consist of isolates of between 100 and 150 individuals. Comparing Hutterite couples in which spouses share HLA antigens at the A, B or DR loci with
MAJOR HlSTOCOMPATlBILITY
ANTIGENS’AND
SPONTANEOUS
those in which alleles are not shared, Ober et al. (13, 20) found evidence of reduced reproductive performance in the former group. Longer interbirth intervals were apparent in homozygous couples (the effect becoming stronger with increasing parity), and spontaneous abortion rates were higher in couples sharing DR alleles. We suggest that an advantage may accrue to the female which aborts under an inbred mating system, because she terminates investment in an offspring which, with inbreeding depression, would likely be of low relative reproductive value (16, 17, 21). We consider spontaneous abortion as potentially advantageous only in inbred populations, where matings between close relatives sharing HLA alleles and also carrying deleterious alleles are most likely. Any advantage should be greatest in populations in which the extent of inbreeding fluctuates temporally, such as many species of rodents in which population sizes boom and crash cyclically. Under constant inbreeding, deleterious consequences will be reduced as natural selection decreases the mutational load within the population (22). The hypothesis that, under certain circumstances, females might manipulate their pregnancies in order to avoid investing in offspring of low or negligible reproductive value is not new. A similar hypothesis was advanced regarding pregnancy failure in female rodents exposed to males other than their mates (a phenomenon known as the Bruce Effect). In many rodents in which such pregnancy failure occurs, males kill neonates to whom they are unrelated. The Bruce Effect can thus be seen as an adaptive device which enables a female to terminate, at an early stage, her investment in an offspring which will likely suffer infanticide after birth (23). Another example in which females apparently manipulate their own pregnancies in an adaptive manner is provided by those mammals in which the sex ratio of offspring at birth deviates from unity as a function of the expected reproductive value of sons versus daughters (24). With reference to clinical medicine, it is not our intention to imply that women who are HLAcompatible with their partners should not attempt to become pregnant using one of the several clinical procedures currently available (25). Most human populations are largely outbred; on average, the inbreeding coefficient of human populations is less than 1 per 1,000. Not withstanding cultural taboos prohibiting incest,
ABORTIONS:
AN EVOLUTIONARY
PERSPECTIVE
237
inbreeding coefficients are usually highest in isolates (16). In outbred populations, sharing of MHC antigens between spouses likely is due to random association. It is interesting to note that in those cases where MHC compatibility is circumvented clinically, no increased incidence of genetic abnormality is apparent among infants carried to term (25). In the scenario developed in the present paper, genetic defects in offspring are most likely to occur only if couples which share MHC antigens also share the same deleterious recessive alleles due to common descent. This situation is most likely to arise in inbred populations due to the high average relatedness between mating partners (this could be tested empirically by assaying for defects due to recessive alleles in abortus material). In more outbred populations, the genetic load will be lower (16). We are sensitive to the fact that we should be careful to construct neither naive adaptationist explanations nor untestable hypotheses (26). In addition, we believe care should be taken to distinguish between the “function” of a character and its “effects”, especially when considering adaptation (15). Nevertheless, we believe the relationships between MHC compatibility, reproductive failure and inbreeding to merit further study. Acknowledgment gratefully acknowledges financial support from the National Science Foundation (grant BSR 85-06766). PAV
References 1. Penagi G S, David C S (eds). Immunogenetics. Butterworths, London, 1984. 2. Bodmer W F. Evolutionary significance of the HL-A system. Nature 237: 139, 1972. 3. Klein J. Many questions (and about no answers) about the phylogenetic origin of the major histocompatibility complex. p. 467 in Immunologic Phylogeny (W H Hildemann, A A Benedict, eds) Plenum Press, New York, 1975. 4. Yamazaki K, Boyse E A. Mike V. Thaler H T, Mathieson B J, Abbott J, Boyse J, Zayas Z A, Thomas L. Control of mating preferences in mice by genes in the major histocompatibility complex. J. Exp. Med. 144: 1324, 1976. 5. Beauchamp G K, Yamazaki K, Bard J, Boyse E A. Preweaning experience in the control of mating preferences by genes in the major histocompatibility complex of the mouse. Behav. Genet. 18: 537, 1988. 6. Billington W D. Influence of immunologic dissimilarity of mother and foetus on size of placenta in mice. Nature 202: 317, 1964.
238 7. Beer A E, Scott J R, Billingham
8. 9. 10.
11. 12.
13.
14.
15.
R. Histocompatibility and maternal immunologic status as determinants of fetoplacental weight and litter size in rodents. J. Exp. Med. 142: 180, 1975. Chaouat G, Kolb J-P, Kiger N, Stanislawski M, Wegmann T G. Immunological consequences of vaccination against abortion in mice. J. Immunol. 134: 1594, 1985. Palm J. Maternal-fetal histoincompatibihty in rats: an escape from adversity. Cancer Res. 34: 2061, 1974. Beer A E, Quebbeman J F, Ayers J W T, Haines R F. Major histocompatibility complex antigens, maternal and paternal immune responses, and chronic habitual abortions in humans. Am. J. Obstet. Gynecol. 141: 987, 1981. Gall S A. Immunologic factors influencing pregnancy. Obstet. Gynecol. Ann. 14: 25, 1985. Radvany R M, Vaisrub N, Ober C, Pate1 K M, Hecht F. The human sex ratio: increase in first-born males to parents with shared HLA-DR antigens. Tissue Antigens 29: 34, 1987. Ober C, Martin A 0, Simpson J L, Hauck W W, Amos D B, Kostyu D D, Fotino M Allen F H. Shared HLA antigens and reproductive performance in the Hutterites. Am. J. Hum. Gen. 35: 994, 1983. Vu110 C M, Celis E M, Serra H M, Riera C M. Study of HLA system in a Mataco population: a geographically isolated American Indian tribe. Tissue Antigens 23: 33, 1984. Williams G C. Adaptation and Natural Selection. Princeton University Press, Princeton, 1966.
MEDICAL HYPOTHESES
16. Cavalli-Sforza L L, Bodmer W F. The Genetics of Human Populations. Freeman. San Francisco. 1971. 17. Schull W j, Neel J V. The effects of Inbreeding on Japanese Children. Harper and Row, New York, 1%5. 18. Goldschmidt E (ed). The Genetics of Migrant and Isolate Populations. Williams and Wilkins, New York, 1963. 19. Morton N E. Morbidity of children from consanauinous marriages. Progr. Med: Genet. 1: 261, 1961. 20. Ober C. Hauck W W. Kostvu D D. O’Brien E. Elias A. Simpson J L, Martin ‘A 0. ‘Adverse effects of HLA-DR sharing on fertility: a cohort study in a human isolate. Fert. Steril. 44: 227, 1985. 21. Stern C. Principles of Human Genetics (Third Edition). Freeman, San Francisco, 1973. 22. Futuyma, D J. Evolutionary Biology. Sinauer, Sunderland, 1986. 23. Schwagmeyer P L. The Bruce effect: an evaluation of male/female advantages. Am. Nat. 114: 932, 1979. 24. Clutton-Brock T H. Birth sex ratios and the reproductive success of sons and daughters. p. 221 in Evolution (P J Greenwood, P H Harvey, M Slatkin, eds) Cambridge Universitv Press. Cambridae. 1985. 25. Beer A E; Immunologic aspects of normal pregnancy and recurrent spontaneous abortion. Semin. Reprod. Endocrinol. 6: 163, 1988. 26. Gould S J, Lewontin R C. The spandrels of San Marco and the Panglossian paradigm: a critique of the adaptationist program. Proc. Roy. Sot. Lond. B 205: 581, 1979.
