Why Blame the Age of the Mother? Lionel Penrose Did Not

The effects of parental age upon the occurrence of abnormalities in offspring never ceased to intrigue Penrose. It was Weinberg who first noted the phenomenon in 1912, and there were others who suggested that it was the father's not the mother's age that played a significant role in the occurrence of infants with, for example, achondroplasia. However, it was Penrose who, again using the partial correlation method employed by Wright, provided the data to prove the hypothesis. His results were a mirror image of those found in families with Down syndrome. In other words, if the maternal age was kept constant, a significantly positive correlation (+0.273) was found between the paternal age and the incidence of achondroplasia, whereas the maternal age effect disappeared completely if the paternal age was regressed out. Furthermore, Penrose noted that the statistical significance of the paternal age effect increased considerably when nonsurviving infants with achondroplasia were excluded from the calculations (PENROSE 1955 , PENROSE 1957 ). It is possible, although no proof exists from the old data, that the infants with "achondroplasia" who died neonatally or shortly thereafter had different diagnoses of more severe, perhaps lethal, short-limbed dwarfing disorders. The survivors represented true instances of new (paternally derived) mutations of the gene.

Together with Haldane, Penrose also studied the paternal age effect upon mutations of genes located on the X chromosome. Haldane had accurately postulated an increased occurrence of hemophilia in the grandsons of older maternal grandfathers, implying another example of a higher mutation rate in males. Four decades later, it was Crow who finally completed the story by summarizing data showing that, at the molecular level, the rate for base substitutions (the cause of the achondroplasia mutation) is indeed higher in males than in females; he also showed that the phenomenon can be attributed partially to the larger number of cell divisions, estimated to be about 430 at age 30 in the male, as contrasted with the female germ line, in whom there are only about 24 divisions from zygote to egg (CROW 1997A , CROW 1997B ). More specifically, the mutation in achondroplasia, in the fibroblast growth factor receptor 3 gene (FGFR3) on 4p16.3, consists of a single substitution of the normal glycine residue by an arginine residue at codon 380. This same Gly380Arg mutation is, surprisingly, present in almost all of the hundreds of patients with achondroplasia who have had mutational analysis to date, and it is of paternal origin in those in which the parental origin could be determined (HORTON 1997 ). Thus, we owe our understanding of the paternal age factor, its causes as well as underlying mechanisms, to the imaginatively creative minds of three generations of genetics heroes