Aberrant Splicing Is the Pathogenicity Mechanism of the p.Glu314Lys Variant in CYP11A1 Gene

Abstract

Context: The cholesterol side chain cleavage enzyme (CYP11A1) catalyzes the conversion of cholesterol to pregnenolone, the first rate-limiting step of steroidogenesis. CYP11A1 mutations are associated with primary adrenal insufficiency (PAI) as well as disorders of sex development (DSD) in 46,XY patients. Objective: To define the pathogenicity mechanism for the p.Glu314Lys variant, previously reported, and found in four additional patients with CYP11A1 deficiency. Subjects and Methods: DNA of four patients presenting with delayed PAI and/or 46,XY DSD were studied by Sanger or Massively Parallel sequencing. Three CYP11A1 mutations were characterized in vitro and in silico, and one by mRNA analysis on testicular tissue. Results: All patients were compound heterozygous for the previously described p.Glu314Lys variant. In silico studies predicted this mutation as benign with no effect on splicing but mRNA analysis found that it led to incomplete exon 5 skipping. This mechanism was confirmed by minigene experiment. The protein carrying this mutation without exon skipping should conserve almost normal activity, according to in vitro studies. Two other mutations found in trans, the p.Arg120Gln and p.Arg465Trp, had similar activity compared to negative control, consistent with the in silico studies. Conclusions: We provide biological proof that the p. Glu314Lys variant is pathogenic due to its impact on splicing and seems responsible for the moderate phenotype of the four patients reported herein. The present study highlights the importance of considering the potential effect of a missense variant on splicing when it is not predicted to be disease causing.

Keywords: CYP11A1; adrenal insufficiency; alternative splicing; congenital lipoid adrenal hyperplasia; disorders of sex development.

Figure 1

Partial chromatograms showing the CYP11A1 mutations in heterozygous state detected in the four patients. DNA sequencing of a healthy control shows the WT sequence underneath. (A) The base change c.940G>A leads to the missense mutation p.Glu314Lys. (B) The base change c.1393C>T leads to the missense mutation p.Arg465Trp. (C) The base change c.359G>A leads to the missense mutation p.Arg120Gln.

Figure 2

Multiple alignment of human CYP11A1 with orthologs. (A) The E314 residue shaded and marked by a triangle is slightly conserved across species. (B) The R465 residue shaded and marked by a triangle is highly conserved across species. (C) The R120 residue shaded and marked by a triangle is highly conserved across species.

Figure 3

In vitro functional CYP11A1 activity assays. (A) Comparison of mutant's residual activity to WT and negative control's activity. Enzymatic activities are reported as % compared to WT activity. Assays were performed after incubation of transfected cells with 3 μmol/L of 22R-hydroxycholesterol in three independant triplicate experiments. Pregnenolone was measured by HPLC-MS/MS. The data are shown as mean ± s.e.m. Enzymatic activities were compared using Anova 1 and Tukey's multiple comparison tests on Graphpad prism software v5.0 (GraphPad, Inc). No significant difference in activity was found between p.Glu314Lys and WT (p ≥ 0.05). The p.Arg120Gln and p.Arg465Trp mutants had residual activity similar to negative controls (p.Arg120Stop and empty vector) (p ≥ 0.05). ns, not significantly different (p ≥ 0.05). ^***Significantly different (p < 0.0001). (B) Michaelis Menten representation of p.Glu314Lys (in gray) and WT's (in black) activity. Assays were performed after incubation of transfected cells with 0.5, 1, 1.5, 2, 3, 5μmol/L of 22R-hydroxycholesterol in three independant triplicate experiments. Pregnenolone was measured by HPLC-MS/MS. The data are shown as mean ± s.e.m.

Figure 4

Analysis of the mutation p.Arg120Gln on three-dimensional model of CYP11A1 (PDB: 3N9Y). This amino acid replacement leads to disruption of three H-bonds (green dotted line) with the heme in green anis. The cholesterol is in violet, the I helix in light blue, L helix in green, cystein pocket in deep blue, and ferrodoxin in orange.

Figure 5

CYP11A1 mRNA analysis from testicular tissue of patient 1 compound heterozygous for the p.Glu314Lys and p.Arg465Trp mutations. (A) Amplification of CYP11A1 exons 3–9 (the reference sequences of exon 5 and 8 are highlighted in blue). Sequences of patient cDNA found the c.940G>A mutation in exon 5 and the c.1393C>T mutation in exon 8 at heterozygous state. WT nucleotide is more important than mutated nucleotide in exon 5 and in contrast, there was more mutated than WT nucleotide in exon 8. (B) Amplification of CYP11A1 exons 3–6. Sequence of patient exon 5 found a low-level second sequence, which matches sequence of exon 6 (the reference sequence of exon 6 is highlighted in blue). This is not the case for the control sample (underneath). (C) Amplification of exon 3 to intron 6. Sequence of patient cDNA is above and shows the skipping of exon 5 at heterozygous state (the reference sequence of exon 6 is highlighted in blue). Sequence of WT, underneath, shows this skipping at a very low level.

Figure 6

Minigene splicing assay for the p.Glu314Lys mutant. (A) Transfection of the mutant minigene (wells 4, 5, and 6) showed a shorter RT-PCR product in comparison with normal minigene (wells 1, 2, and 3). (B) Sequencing of PCR products revealed normal splice product for the normal minigene and a complete exon skipping for the mutant minigene.