An X-linked cobalamin disorder caused by mutations in transcriptional coregulator HCFC1

Abstract

Derivatives of vitamin B12 (cobalamin) are essential cofactors for enzymes required in intermediary metabolism. Defects in cobalamin metabolism lead to disorders characterized by the accumulation of methylmalonic acid and/or homocysteine in blood and urine. The most common inborn error of cobalamin metabolism, combined methylmalonic acidemia and hyperhomocysteinemia, cblC type, is caused by mutations in MMACHC. However, several individuals with presumed cblC based on cellular and biochemical analysis do not have mutations in MMACHC. We used exome sequencing to identify the genetic basis of an X-linked form of combined methylmalonic acidemia and hyperhomocysteinemia, designated cblX. A missense mutation in a global transcriptional coregulator, HCFC1, was identified in the index case. Additional male subjects were ascertained through two international diagnostic laboratories, and 13/17 had one of five distinct missense mutations affecting three highly conserved amino acids within the HCFC1 kelch domain. A common phenotype of severe neurological symptoms including intractable epilepsy and profound neurocognitive impairment, along with variable biochemical manifestations, was observed in all affected subjects compared to individuals with early-onset cblC. The severe reduction in MMACHC mRNA and protein within subject fibroblast lines suggested a role for HCFC1 in transcriptional regulation of MMACHC, which was further supported by the identification of consensus HCFC1 binding sites in MMACHC. Furthermore, siRNA-mediated knockdown of HCFC1 expression resulted in the coordinate downregulation of MMACHC mRNA. This X-linked disorder demonstrates a distinct disease mechanism by which transcriptional dysregulation leads to an inborn error of metabolism with a complex clinical phenotype.

Figure 1

Metabolite Measurements in cblX Subjects Compared to Treated, Early-Onset cblC Subjects Plasma total homocysteine (tHcy) and serum and urine methylmalonic acid (MMA) were measured in cblX subjects (single measurements) and 23 subjects with early-onset cblC deficiency (1–13 measurements per subject). In cases where multiple measurements were available for the same subject, the mean of all readings was used. n = the total number of subjects in each group. (A) There was no statically significant difference in plasma tHcy measurements between cblC subjects and cblX subjects (cblC, 60.63 ± 4.88 μM; cblX, 64.25 ± 18.30 μM [mean ± SEM]). Normal levels of plasma tHcy are <13 μM. Triangles indicate two subjects with reportedly normal tHcy (no concentration was provided to the referring diagnostic laboratory). (B) There was no significant difference in serum MMA measurements between cblC subjects and cblX subjects (cblC, 11.57 ± 2.99 μM; cblX, 15.00 ± 3.43 μM [mean ± SEM]). Normal levels of serum MMA are <0.4 μM. (C) cblX subjects had higher urine MMA (t test p < 0.001) than did cblC subjects (cblC, 96.52 ± 23.89 mmol/mol creatinine; cblX, 354.20 ± 93.70 mmol/mol creatinine [mean ± SEM]). Normal levels of urine MMA are <4 mmol/mol creatinine.

Figure 2

Pathogenic Variants of HCFC1 in cblX (A) The top panel shows the 26 exons of HCFC1 as gray boxes. The bottom panel shows the predicted HCFC1 domains, including the kelch domain (kelch motifs K1–K5), Fn3 (fibronectin type 3), the basic domain, HCF-proteolysis repeats (HCF-pro; represented as triangles), the acidic domain, and NLS (nuclear localization signal) domains (adapted from Wilson et al.²⁴). The HCFC1 mutations are clustered within exon 2 and 3 of the cDNA, corresponding to the first (K1) and second (K2) kelch motifs, respectively, in HCFC1. (B) Comparative analysis of HCFC1 from multiple species demonstrated that Gln68, Ala73, and Ala115 (highlighted in red) are evolutionarily conserved throughout vertebrates.

Figure 3

Expression Analysis (A) qPCR analysis of mRNA expression. MMACHC expression was either completely lost or reduced by ∼76% in fibroblasts derived from subjects 1 and 11, respectively (asterisks indicate statistical significance). Error bars represent the SEM of relative expression levels. The two control samples used were from healthy individuals with no known biochemical or neurological phenotypes. (B) Immunoblot analysis of MMACHC. Fibroblast lysates from a reference (control) sample (human dermal fibroblast C-013-5C, Life Technologies), two individuals with cblX, and one with cblC were analyzed. The cblX (lanes 2 and 3) and cblC (lane 4) lines show only trace amounts of MMACHC (top panel). (C) Immunoblot analysis of HCFC1. Fibroblast lysates from the same control used in (B) and two cblX cell lines were analyzed. HCFC1 levels in all the samples remained unchanged. β-actin was used as a loading control in all immunoblots (in the bottom of B and C).

Figure 4

siRNA Knockdown of HCFC1 HEK293 cells were transfected with HCFC1-specific siRNA or scrambled siRNA (control), and the relative expression of HCFC1 and MMACHC was assayed by qPCR using ACTB (β-actin) as an endogenous control. Compared to control cells (average expression levels were fixed at 1), siRNA-treated cells had significantly reduced HCFC1 (0.275) and MMACHC (0.424) expression levels, but the MMADHC expression level remain unchanged. Error bars represent the SEM of relative expression levels.