We identified two homozygous missense mutations in the human type II 3beta-hydroxysteroid dehydrogenase (3/betaHSD) gene, the first in codon 6 of exon II [CTT (Leu) to TTT (Phe)] in a male infant with hyperpigmented scrotum and hypospadias, raised as a male and no apparent salt-wasting since neonatal age, and the second in codon 259 of exon IV [ACG (Thr) to ATG (Met)] in a male pseudohermaphrodite with labial scrotal folds, microphallus, chordee, and fourth degree hypospadias, raised as a female and with salt-wasting disorder since neonatal age. In vitro transient expression of mutant type II 3betaHSD complementary DNAs of L6F, T259M, as well as T259R for comparison was examined by a site-directed mutagenesis and transfection of construct into COS-1 and COS-7 cells. Northern blot analysis revealed expression of similar amounts of type II 3betaHSD messenger ribonucleic acid from the COS-1 cells transfected by L6F, T259M, T259R, and wild-type (WT) complementary DNAs. Western immunoblot analysis revealed a similar amount of L6F mutant protein compared to WT enzyme from COS-1 cells, but neither L6F from COS-7 cells nor T259M or T259R mutant protein in COS-1 or COS-7 cells was detectable. Enzyme activity in intact COS-1 cells using 1 micromol/L pregnenolone as substrate in the medium after 6 h revealed relative conversion rates of pregnenolone to progesterone of 46% by WT enzyme, 22% by L6F enzyme, and 8% by T259M enzyme and less than 4% activity by T259R enzyme. Using 1 micromol/L dehydroepiandrosterone as substrate, the relative conversion rate of dehydroepiandrosterone to androstenedione after 6 was 89% by WT enzyme, 35% by L6F enzyme, 5.1% by T259M enzyme and no activity by T259R enzyme. However, the L6F mutant 3betaHSD activity, despite its demonstration in the intact cells, was not detected in homogenates of COS-1 cells or in immunoblots of COS-7 cells, suggestive of the relatively unstable nature of this protein in vitro, possibly attributable to the decreased 3betaHSD activity. In the case of T259M and T259R mutations, consistently undetectable proteins in both COS cells despite detectable messenger ribonucleic acids indicate severely labile proteins resulting in either no or very little enzyme activity, and these data further substantiate the deleterious effect of a structural change in this predicted putative steroid-binding domain of the gene. In conclusion, the findings of the in vitro study of mutant type II 3betaHSD enzyme activities correlated with a less severe clinical phenotype of nonsalt-wasting and a lesser degree of genital ambiguity in the patient with homozygous L6F mutation compared to a more severe clinical phenotype of salt-wasting and severe degree of genital ambiguity in the patient with homozygous T259M mutation in the gene.