Functional characterization of PCCA mutations causing propionic acidemia

Propionic acidemia (PA, MIM 232000 and 232050) is caused by a deficiency of mitochondrial biotin-dependent propionyl-CoA carboxylase (PCC, EC 6.4.1.3), a heteropolymeric enzyme composed of alpha and beta subunits, which are encoded by the PCCA and PCCB genes, respectively. The PCCA protein (alpha subunit) is responsible for the formation of carboxybiotin upon hydrolysis of ATP and contains a C-terminal biotin-binding domain and a biotin carboxylase domain, defined by homology with other biotin-dependent carboxylases, some of them characterized structurally. More than 24 mutations have been found in the PCCA gene in patients with PA, among them 14 missense mutations and one in-frame deletion, for which the precise molecular effect is unknown. In this study, we have established the pathogenicity of 11 PCCA mutations (10 missense and an in-frame deletion) by expression studies in deficient fibroblasts and in a cell-free in vitro system, and analyzed the effect of each mutation on PCC activity, protein stability and domain structure. The results show that most mutant proteins show an increased turnover and are functionally deficient, suggesting that the structural alterations they cause are incompatible with normal assembly to produce a stable, functional PCC oligomer. These results are discussed in the context of the genotype-phenotype correlations in PCCA-deficient PA patients.     

Human propionyl-CoA carboxylase beta subunit gene: exon-intron definition and mutation spectrum in Spanish and Latin American propionic acidemia patients.

Propionyl-CoA carboxylase (PCC) is a mitochondrial biotin-dependent enzyme composed of an equal number of alpha and beta subunits. Mutations in the PCCA (alpha subunit) or PCCB (beta subunit) gene can cause the inherited metabolic disease propionic acidemia (PA), which can be life threatening in the neonatal period. Lack of data on the genomic structure of PCCB has been a significant impediment to full characterization of PCCB mutant chromosomes. In this study, we describe the genomic organization of the coding sequence of the human PCCB gene and the characterization of mutations causing PA in a total of 29 unrelated patients-21 from Spain and 8 from Latin America. The implementation of long-distance PCR has allowed us to amplify the regions encompassing the exon/intron boundaries and all the exons. The gene consists of 15 exons of 57-183 bp in size. All splice sites are consistent with the gt/ag rule. The availability of the intron sequences flanking each exon has provided the basis for implementation of screening for mutations in the PCCB gene. A total of 56/58 mutant chromosomes studied have been defined, with a total of 16 different mutations detected. The mutation spectrum includes one insertion/deletion, two insertions, 10 missense mutations, one nonsense mutation, and two splicing defects. Thirteen of these mutations correspond to those not described yet in other populations. The mutation profile found in the chromosomes from the Latin American patients basically resembles that of the Spanish patients.     

Cloning of functional alpha propionyl CoA carboxylase and correction of enzyme deficiency in pccA fibroblasts.

Propionyl CoA carboxylase (PPC) is a heteromeric enzyme composed of alpha subunits (PCCA) and beta (PCCB) subunits. We describe cDNA clones expressing human PCCA and complementation of the genetic defect in pccA fibroblasts by DNA-mediated gene transfer. Two cDNA clones were constructed. The first corresponds to the previously reported, putatively full-length, open reading frame. The second encodes a chimera composed of the mitochondrial leader sequence of human methylmalonyl CoA mutase and the mature PCCA protein. Both clones reconstitute propionate flux to normal levels in fibroblasts from patients genetically deficient in PCCA (pccA). The maximal level of propionate flux approached, but never exceeded, the levels seen in control plates of normal cells. In contrast, the maximal level of PPC holoenzyme activity reached only 10%-20% that of normal controls, which corresponded roughly to the fraction of cells actually transformed with the recombinant gene. These data suggest that the level of PCCA expression in fibroblasts does not normally limit PCC holoenzyme activity or propionate flux. The fact that a small fraction of cells reconstitutes propionate flux to normal levels suggests that metabolic cooperation between cells is capable of increasing the metabolic capacity of recombinant enzyme in a subpopulation of cells. These factors may have important implications for the rational design of somatic gene therapy for PCCA deficiency.     

Propionicacidemia: absence of alpha-chain mRNA in fibroblasts from patients of the pccA complementation group.

Propionicacidemia is an autosomal recessive metabolic disease resulting from a deficiency of propionyl-CoA carboxylase (PCC) activity. The enzyme has the structure alpha 4 beta 4, with the alpha chain containing a covalently bound biotin prosthetic group. Patients have been placed into two major complementation groups, pccA and pccBC, that may correspond to the genes encoding the alpha and beta chains of PCC. The pccBC group is further divided into two subgroups, pccB and pccC, apparently owing to intragenic complementation. We previously reported combined alpha- and beta-chain deficiency in pccA mutants and absence of beta chain in pccC and pccBC mutants after isotope-tracer labeling and immunoprecipitation of cultured-fibroblast extracts. Using cDNA clones coding for the alpha and beta chains as probes, we found absence of alpha mRNA in four of six pccA strains and presence of beta mRNA in all pccA mutants studied. We also found presence of both alpha and beta mRNAs in three pccBC, two pccB, and three pccC mutants. From these data, we confirm the gene assignments of the complementation groups (PCCA gene = pccA complementation group; PCCB gene = pccBC and subgroups) and support the view that pccA patients synthesize a normal beta chain that is rapidly degraded in the absence of complexing with alpha chains.     

A novel point mutation in an acceptor splice site of intron 32 (IVS32 A–12→G) but no exon 3 mutations in the glycogen debranching enzyme gene in a homozygous patient with glycogen storage disease type IIIb

Genetic deficiency of the glycogen-debranching enzyme (debrancher) causes glycogen storage disease type III (GSD III), which is divided into two subtypes: IIIa and IIIb. In GSD IIIb, glycogen accumulates only in the liver, whereas both liver and muscles are involved in GSD IIIa. The molecular basis for the differences between the two subtypes has not been fully elucidated. Recently, mutations in exon 3 of the debrancher gene were reported to be specifically associated with GSD IIIb. However, we describe a homozygous GSD IIIb patient without mutations in exon 3. Analysis of the patient’s debrancher cDNA revealed an 11-bp insertion in the normal sequence. An A to G transition at position –12 upstream of the 3′ splice site of intron 32 (IVS 32 A^–12→G) was identified in the patient’s debrancher gene. No mutations were found in exon 3. Mutational analysis of the family showed the patient to be homozygous for this novel mutation as well as three polymorphic markers. Furthermore, the mother was heterozygous and the parents were first cousins. The acceptor splice site mutation created a new 3′ splice site and resulted in insertion of an 11-bp intron sequence between exon 32 and exon 33 in the patient’s debrancher mRNA. The predicted mutant enzyme was truncated by 112 amino acids as a result of premature termination. These findings suggested that a novel IVS 32 A^–12→G mutation caused GSD IIIb in this patient. Received: 1 August 1997 / Accepted: 22 September 1997     

Mutations participating in interallelic complementation in propionic acidemia.

Deficiency of propionyl-CoA carboxylase (PCC; alpha 4 beta 4) results in the rare, autosomal recessive disease propionic acidemia. Cell fusion experiments have revealed two complementation groups, pccA and pccB, corresponding to defects of the PCCA (alpha-subunit) and PCCB (beta-subunit) genes, respectively. The pccBCC group includes subgroups, pccB and pccC, which are thought to reflect interallelic complementation between certain mutations of the PCCB gene. In this study, we have identified the mutations in two pccB, one pccC, and two pccBC cell lines and have deduced those alleles participating in interallelic complementation. One pccB line was a compound heterozygote of Pro228Leu and Asn536Asp. The latter mutation was also detected in a noncomplementing pccBC line. This leaves Pro228Leu responsible for complementation in the pccB cells. The second pccB line contained an insertional duplication, dupKICK140-143, and a splice mutation IVS + 1 G-->T, located after Lys466. We suggest that the dupKICK mutation is the complementing allele, since the second allele is incompatible with normal splicing. The pccC line studied was homozygous for Arg410Trp, which is necessarily the complementing allele in that line. For a second pccC line, we previously had proposed that delta Ile408 was the complementing allele. We now show that its second allele, "Ins.Del," a 14-bp deletion replaced by a 12-bp insertion beginning at codon 407, fails to complement in homozygous form. We conclude that the interallelic complementation results from mutations in domains that can interact between beta-subunits in the PCC heteromer to restore enzymatic function. On the basis of sequence homology with the Propionibacterium shermanii transcarboxylase 12S subunit, we suggest that the pccC domain, defined by Ile408 and Arg410, may involve the propionyl-CoA binding site.     

Biochemical and genetic studies of four patients with pyruvate dehydrogenase E1α deficiency

We report studies of four patients with pyruvate dehydrogenase complex (PDH) deficiency caused by mutations in the E1α subunit. Two unrelated male patients presented with Leigh syndrome and a R263G missense mutation in exon 8. This mutation has previously been described in males with the same phenotype. The two other patients had different novel mutations: (1) an 8-bp deletion at the C-terminus (exon 11) was found in one allele of a young girl suffering from microcephaly and (2) a C88S missense mutation (exon 3) in a boy who only presented with motor neuropathy. These mutations were not found in the mothers of any of the four cases. Immunoblot analysis revealed decreased immunoreactivity for the E1α and E1β subunits in three out of the four patients. These findings confirm that: (1) PDH deficiencies are genetically heterogeneous, (2) the R263G mutation is more frequent in male cases than are other mutations and this amino acid is a hot spot for gene mutations, (3) the last eight amino acids may be important for the conformation of the tetrameric E1-PDH enzyme, and (4) the amino acids at positions 88, 263 and 382–387 are essential for the linking of the α subunit with the β subunit and for the activity of the holoenzyme. Received: 28 October 1996 / Revised: 13 January 1997     

Identification of a novel splicing mutation and 1-bp deletion in the 17α-hydroxylase gene of Japanese patients with 17α-hydroxylase deficiency

We report studies of two unrelated Japanese patients with 17α-hydroxylase deficiency caused by mutations of the 17α-hydroxylase (CYP17) gene. We amplified all eight exons of the CYP17 gene, including the exon-intron boundaries, by the polymerase chain reaction and determined their nucleotide sequences. Patient 1 had novel, compound heterozygous mutations of the CYP17 gene. One mutant allele had a guanine to thymine transversion at position +5 in the splice donor site of intron 2. This splice-site mutation caused exon 2 skipping, as shown by in vitro minigene expression analysis of an allelic construct, resulting in a frameshift and introducing a premature stop codon (TAG) 60 bp downstream from the exon 1-3 boundary. The other allele had a missense mutation of His (CAC) to Leu (CTC) at codon 373 in exon 6. These two mutations abolished the 17α-hydroxylase and 17,20-lyase activities. Restriction fragment length polymorphism (RFLP) analysis with a mismatch oligonucleotide showed that the patient’s mother and brother carried the splice-site mutation, but not the missense mutation. Patient 2 was homozygous for a novel 1-bp deletion (cytosine) at codon 131 in exon 2. This 1-bp deletion produces a frameshift in translation and introduces a premature stop codon (TAG) proximal to the highly conserved heme iron-binding cysteine at codon 442 in microsomal cytochrome P450 steroid 17α-hydroxylase (P450c17). RFLP analysis showed that the mother was heterozygous for the mutation. Received: 15 November 1997 / Accepted: 15 March 1998     

Functional analysis of PCCB mutations causing propionic acidemia based on expression studies in deficient human skin fibroblasts

Propionic acidemia (PA) is a recessive disorder caused by a deficiency of propionyl-CoA carboxylase (PCC), a dodecameric enzyme composed of two different proteins alpha-PCC and beta-PCC, nuclear encoded by the PCCA and PCCB genes, respectively. Mutations in either gene cause PA and to date, up to 47 different allelic variations in the PCCB gene have been identified in different populations. In this work, we describe the expression studies of 18 PCCB sequence changes in order to elucidate their functional consequences. We have used a PCCB-deficient transformed fibroblast cell line to target the wild-type and mutant proteins to their physiological situation, analysing the effect of the mutations on PCC activity and protein stability. Of the 18 mutant proteins tested for activity, those carrying the L17M and A497V substitutions showed an activity similar to the wild-type one, which proves that these changes do not have any effect on protein activity. The other 16 mutant proteins exhibited two different functional behaviours, 3 retained substantial activity (K218R, R410W and N536D), and the remaining 13 proteins showed null or very low activity. Western blot analysis demonstrated instability only for the L519P, R512C and G112D mutant proteins. We have proved the pathogenicity of R67S, R165Q and G112D mutation in PCCB gene, expressed for the first time in this work. The information derived from the expression analysis is discussed in the phenotype and genotype context in order to improve the knowledge of this complex disease.     

Variability of biochemical and clinical phenotype in X-linked liver glycogenosis with mutations in the phosphorylase kinase PHKA2 gene

X-linked liver glycogenosis (XLG) resulting from phosphorylase kinase (Phk) deficiency is one of the most common forms of glycogen storage disease. It is caused by mutations in the gene encoding the liver isoform of the Phk α subunit (PHKA2). In the present study, we address the issue of phenotypic and allelic heterogeneity in XLG. We have identified mutations in seven male patients. One of these patients represents the variant biochemical phenotype, XLG subtype 2 (XLG2), where Phk activity is low in liver but normal or even elevated in erythrocytes. He carries a K189E missense mutation, which adds to the emerging evidence that XLG2 is associated with missense mutations clustering at a few sites. Two patients display clinical phenotypes unusual for liver Phk deficiency, with dysfunction of the kidneys (proximal renal tubular acidosis) or of the nervous system (seizures, delayed cognitive and speech abilities, peripheral sensory neuropathy), respectively, in addition to liver glycogenosis. In the patient with kidney involvement, we have identified a missense mutation (P399S) and a trinucleotide deletion (2858del3) leading to the replacement of two amino acids by one new residue (N953/L954I), and a missense mutation has also been found in the patient with neurological symptoms (G1207W). These two cases demonstrate that PHKA2 mutations can also be associated with uncommon clinical phenotypes. Finally, in four typical XLG cases, we have identified three truncating mutations (70insT, R352X, 567del22) and an in-frame deletion of eight well-conserved amino acids (2452del24). Together, this study adds eight new mutations to the previously known complement of sixteen PHKA2 mutations. All known PHKA2 mutations but one are distinct, indicating pronounced allelic heterogeneity of X-linked liver glycogenosis with mutations in the PHKA2 gene. Received: 17 October 1997 / Accepted: 23 December 1997     

Large-scale deletion and point mutations of the nuclear NDUFV1 and NDUFS1 genes in mitochondrial complex I deficiency

Reduced nicotinamide adenine dinucleotide (NADH):ubiquinone oxidoreductase (complex I) is the largest complex of the mitochondrial respiratory chain and complex I deficiency accounts for approximately 30% cases of respiratory-chain deficiency in humans. Only seven mitochondrial DNA genes, but >35 nuclear genes encode complex I subunits. In an attempt to elucidate the molecular bases of complex I deficiency, we studied the six most-conserved complex I nuclear genes (NDUFV1, NDUFS8, NDUFS7, NDUFS1, NDUFA8, and NDUFB6) in a series of 36 patients with isolated complex I deficiency by denaturing high-performance liquid chromatography and by direct sequencing of the corresponding cDNA from cultured skin fibroblasts. In 3/36 patients, we identified, for the first time, five point mutations (del222, D252G, M707V, R241W, and R557X) and one large-scale deletion in the NDUFS1 gene. In addition, we found six novel NDUFV1 mutations (Y204C, C206G, E214K, IVS 8+41, A432P, and del nt 989-990) in three other patients. The six unrelated patients presented with hypotonia, ataxia, psychomotor retardation, or Leigh syndrome. These results suggest that screening for complex I nuclear gene mutations is of particular interest in patients with complex I deficiency, even when normal respiratory-chain-enzyme activities in cultured fibroblasts are observed.     

Characterization of four variant forms of human propionyl-CoA carboxylase expressed in Escherichia coli

Propionyl-CoA carboxylase (PCC) is a biotin-dependent mitochondrial enzyme that catalyzes the conversion of propionyl-CoA to D-methylmalonyl-CoA. PCC consists of two heterologous subunits, alpha PCC and beta PCC, which are encoded by the nuclear PCCA and PCCB genes, respectively. Deficiency of PCC results in a metabolic disorder, propionic acidemia, which is sufficiently severe to cause neonatal death. We have purified three PCCs containing pathogenic mutations in the beta subunit (R165W, E168K, and R410W) and one PCCB polymorphism (A497V) to homogeneity to elucidate the potential structural and functional effects of these substitutions. We observed no significant difference in Km values for propionyl-CoA between wild-type and the variant enzymes, which indicated that these substitutions had no effect on the affinity of the enzyme for this substrate. Furthermore, the kinetic studies indicated that mutation R410W was not involved in propionyl-CoA binding in contrast to a previous report. The three mutant PCCs had half the catalytic efficiency of wild-type PCC as judged by the kcat/Km ratios. No significant differences have been observed in molecular mass or secondary structure among these enzymes. However, the variant PCCs were less thermostable than the wild-type. Following incubation at 47 degrees C, blue native-PAGE revealed a lower oligomeric form (alpha2beta2) in the three mutants not detectable in wild-type and the polymorphism. Interestingly, the lower oligomeric form was also observed in the corresponding crude Escherichia coli extracts. Our biochemical data and the structural analysis using a beta PCC homology model indicate that the pathogenic nature of these mutations is more likely to be due to a lack of assembly rather than disruption of catalysis. The strong favorable effect of the co-expressed chaperone proteins on PCC folding, assembly, and activity suggest that propionic acidemia may be amenable to chaperone therapy.     

Structure of the human biotinidase gene

Biotinidase cleaves biotin from biocytin, thereby recycling the vitamin. We have determined the structure of the human biotinidase gene. A genomic clone, containing three exons that code for the mature enzyme, was obtained by screening a human genomic bacteriophage library with the biotinidase cDNA by plaque hybridization. To obtain a clone containing the most 5′ exon of the biotinidase cDNA, a human PAC library by PCR was screened. The human biotinidase gene is organized into four exons and spans at least 23 kb. The 5′-flanking region of exon 1 contains a CCAAT element, three initiator sequences, an octamer sequence, three methylation consensus sites, two GC boxes, and one HNF-5 site, but has no TATA element. The region from nt −600 to +400 has features of a CpG island and resembles a housekeeping gene promoter. The structure and sequence of this gene are useful for identifying and characterizing mutations that cause biotinidase deficiency. Received: 30 September 1997 / Accepted: 5 December 1997