Aberrant segregation in human populations

Abstract

Three cases of aberrant segregation in humans were considered. The first was a possible case of non-random assortment of chromosomes and meiotic drive. Separating children of fathers of given heterozygous ABO genotypes into sperm classes (classified by the ABO gene received from the father), for children of A1B and A2B fathers a change in association of the sperm type with the sex of the child (which would be determined by the father's sex chromosome received) occurred with advancing father's age. Specifically, the B sperm class was associated with the X chromosome when fathers were old and the Y chromosome when they were young. Mother's age and birth order effects were ruled out by multiple regression. This effect was strikingly non-linear: a fifth order polynomial of father's age was very highly significant (for pooled offspring of A1B and A2B fathers). The sex ratio changed within each sperm class in opposite directions. The segregation frequency of the B sperm class for female offspring increased with father's age (but not mother's age or birth order) and again a fifth order polynomial was very highly significant. The progeny of B0, A2O, and A1O plus A1A2 fathers were not affected. There was a non-linear change in segregation frequency of female offspring of AB mothers as mothers grew older which was not highly significant and may have been due to chance. The second study was on the genetic load of the population in Odate, Japan. Prenatal death was examined with respect to several factors and found to fall into at least three classes: early prenatal death (EPRE), up to four months of gestation; late prenatal death (LPRE), up to nine months; stillbirths, more than nine months. LPRE increased significantly (but non-linearly) with inbreeding but EPRE and stillbirths did not; regressions for LPRE and stillbirths differed significantly. Postnatal death was divided into three classes (which differed in several respects): early postnatal death (less than one month old); middle postnatal death (less than two years old); late postnatal death (two or more years old). Only early postnatal deaths increased significantly with inbreeding (.4 lethal equivalents per gamete; B/A = 14.5) and the regression differed significantly from those of the later stages. The last analysis was on ABO incompatibility in the same population. Although a preliminary analysis had not shown significant effects, it was found that the inclusion of three secretor interactions with ABO incompatibility showed very highly significant effects on the stages of mortality considered above. The secretor interactions were the mother's secretor type, the father's secretor type, and the probability that the zygote was See A and B incompatibility on 0 and non-O mothers were analyzed separately for mortality and were found to behave differently. Interactions with parity and (parity)^2 were often highly significant. The interactions tended to balance each other making simpler models non-significant. Some of the regressions suggested viability differences among genotypes of offspring. Significant environmental interactions were also found.