Why Nobody Has P450scc (20,22 Desmoslase) Deficiencyg

Abstract

Okuyama et al. (1) recently reported a case of congenital lipoid adrenal hyperplasia (lipoid CAH) caused by a splicing mutation in the gene for the steroidogenic acute regulatory protein (StAR). In citing the relevant literature, they pointed out that nearly all patients with the lipoid CAH phenotype have been found to have StAR mutations. In two large series, Bose et al. (2) and Nakae et al. (3) found StAR mutations in 39 of 40 patients. Lipoid CAH was formerly misnamed 20,22 desmolase deficiency and was thought to be caused by mutations in P450scc (4–6), the cholesterol-side chain cleavage enzyme that converts cholesterol to pregnenolone, but these patients were recently shown to lack P450scc mutations (7) and have StAR mutations (8). Nevertheless, three factors have contributed to the persistence of the notion that some of these patients may harbor P450scc mutations: first, the weight of history and the lingering misnomer of 20,22 desmolase deficiency; second, the failure to find a StAR mutation in Bose’s patient #14 (2); third and most important, the description by Yang et al. that rabbits that are homozygous for P450scc gene deletions have a phenotype that very closely corresponds with lipoid CAH (9). While we cannot be certain that all patients with lipoid CAH will have StAR mutations (especially in view of Bose’s patient #14), we can be quite confidant that P450scc mutations will never be found in a liveborn human being. The reasons for this were outlined recently (10) but are presented in more detail below. As shown by numerous biochemical experiments (for review see 11) and proven by the P450scc deficient rabbit of Yang et al. (9), P450scc is the only enzyme that can convert cholesterol to pregnenolone, which is the needed precursor for the production of progesterone. Progesterone is required for the maintenance of mammalian pregnancy (for review see 12); without progesterone, uterine contractions and spontaneous abortion occur. During pregnancy, there are two potential sources of progesterone—the corpus luteum of the mother’s ovary and the placenta. These are fundamentally different: the corpus luteum is maternal tissue, but the steroidogenic syncytiotrophoblasts of the placenta are fetal tissue. In rabbits (and goats), the corpus luteum secretes progesterone throughout pregnancy (13), so that a disorder in placental progesterone secretion will not adversely affect the pregnancy. In these animals, placental choriogonadotropin is needed to maintain the maternal corpus luteum, but the placenta is not needed to produce progesterone (14). By contrast in human beings, the corpus luteum secretes progesterone only during the first trimester; placental progesterone becomes sufficient to maintain pregnancy only after about 6 weeks gestation; the so-called“ luteoplacental shift” (13, 15). Consequently human maternal ovariectomy in the first 6 weeks will cause spontaneous abortion, but maternal ovariectomy thereafter will not (13, 15). Thus in the rabbit, a homozygous P450scc mutation in the fetus will not disrupt progesterone production by the heterozygous mother, and the pregnancy will be carried to term. By contrast a human fetus homozygous for a P450scc mutation will spontaneously abort at about 6–7 weeks, when production of progesterone from the maternal corpus luteum wanes, and the mutant placental system is unable to take over. This same physiology explains why mutations have never been found in the human genes for the proteins that transfer electrons to P450scc, ferredoxin reductase and ferredoxin. Similarly, placental production of progesterone requires 3β-hydroxysteroid dehydrogenase (3βHSD). The placental form of this enzyme (3βHSD-I) is encoded by a different gene that the 3βHSD-II gene expressed in the adrenals and gonads. Thus all patients with 3βHSD deficiency have mutations in the adrenal/gonadal 3βHSD-II but never in 3βHSD-I (16), as 3βHSD-I is required for placental production of progesterone. Fetuses with StAR mutations causing lipoid CAH reach term because their placentas continue to produce progesterone (17). The placenta does not express StAR (18) and produces progesterone through a poorly-characterized process called StAR-independent steroidogenesis (2), which may involve the StAR-like protein MLN-64 (19). However, consideration of the genetics of progesterone production by the maternal-fetal-placental unit tells us that human fetuses with homozygous mutations for any of the genes required for placental production of progesterone will not survive to the time of delivery.