Study of cholesterol side-chain cleavage (20,22 desmolase) deficiency causing congenital lipoid adrenal hyperplasia using bovine-sequence P450scc oligodeoxyribonucleotide probes.

Conversion of cholesterol to pregnenolone is mediated by the cholesterol side-chain cleavage (SCC) enzyme, P450scc. Deficient SCC activity causes congenital lipoid adrenal hyperplasia (also known as 20,22 desmolase deficiency), a potentially lethal defect in the synthesis of all steroid hormones. To probe for possible genetic defects causing this disease we synthesized four oligodeoxyribonucleotides containing 63 to 72 bases corresponding to portions of the bovine complementary DNA (cDNA) sequence for P450scc. The bovine oligonucleotides were labeled and used directly to probe Southern blots of normal human genomic DNA, revealing a pattern indicating there is a single P450scc gene in the human genome. Hybridization to Northern blots of normal human and bovine adrenal messenger RNA indicates that P450scc messenger RNA is about 2.0 kilobases long in both species. Hybridizations of the oligonucleotides to genomic DNA from three unrelated patients with SCC deficiency did not detect a deletion in the human P450scc gene. The bovine sequence oligonucleotides were then used to isolate a human P450scc cDNA clone. The isolated P450scc cDNA fragment contains 818 bases encoding 239 amino acids of the protein, the translation termination signal, and 98 bases of the 3' untranslated region. The sequence of this carboxy-terminal half of the human P450scc protein is 72% homologous with the bovine sequence and contains an additional amino acid not found in bovine P450scc; the human and bovine nucleotide sequences are 81% homologous. Repetition of the genomic DNA blotting studies with the cDNA probe gave the same results obtained with the bovine-sequence oligonucleotide probes, confirming that SCC deficiency is not due to a deletion in the regions of the P450scc hybridizing with the probes. Long, chemically synthesized heterologous sequence oligonucleotides containing unknown numbers of base mismatches with human sequences may thus be used to study human genes so that access to a cDNA is not necessary for such studies.