In 1961, Childs et al.7 described a male infant with episodic metabolic ketoacidosis, protein intolerance, and remarkably elevated plasma glycine concentration. Several hundred children with similar clinical and biochemical findings have since been described. Many of these children were subsequently found to have methylmalonic acidemia173 ; a few had β-ketothiolase deficiency (see Chapter 93). However, the patient described by Childs et al., and many reported subsequently, had propionic acidemia due to a primary and specific deficiency of propionyl CoA carboxylase activity (Fig. 94-2). This conclusion was derived independently from the description of a patient with massive propionate accumulation in blood,174 from another with impaired propionate oxidation in leukocytes9 and defective carboxylase activity in fibroblast extracts,175 and from a third with both propionic acidemia and defective carboxylase activity.176 We now recognize that propionyl CoA carboxylase deficiency also occurs in children with inherited abnormalities in biotin metabolism, leading to the deficiency of multiple biotin-dependent carboxylases (see Chapter 156). Hence, we must now use the term propionic acidemias to refer to this heterogeneous group of related inborn errors. As will be discussed subsequently, a similar heterogeneity exists among the methylmalonic acidemias.
Propionyl CoA Carboxylase Deficiency.
As mentioned above, this disorder was originally referred to as ketotic hyperglycinemia. E.G., the patient described by Childs and Nyhan and their colleagues,7,177,178 presented with dehydration, lethargy, and coma on the first day of life. He was found to be severely ketoacidotic and responded slowly to massive alkali replacement. The clinical course was characterized by recurrent attacks of ketoacidosis, precipitated by infections or protein ingestion, and by developmental retardation, electroencephalographic (EEG) abnormalities, and osteoporosis. The patient had episodic neutropenia and thrombocytopenia prior to death at age 7. A sister (A.G.) also became ketotic and acidotic during the first 4 days of life, but the course of her condition has been modified dramatically because of the extensive experience gained in studying her brother. Although she has had mild attacks of ketoacidosis during intercurrent infections, maintenance on a low-protein diet has resulted in little need for hospital care and normal somatic and mental development up to 15 years of age.179
In 1968, Hommes and his colleagues174 described a male infant with hyperventilation, areflexia, and grunting at 60 h of age. There was a profound metabolic acidosis (arterial pH 6.98), and despite administration of massive amounts of sodium bicarbonate and tris(hydroxymethyl)-aminomethane, the infant died on the fifth day of life. Leukocytes and platelets were normal. Postmortem examination showed only a fatty liver and degeneration of Purkinje cells and the granular layer of the cerebellum.
Subsequent descriptions of patients with propionic acidemia have confirmed that most patients present in the newborn period with severe metabolic acidosis manifested by refusal to feed, vomiting, lethargy, and hypotonia; dehydration, seizures, and hepatomegaly occur less often.180,181 Other patients have presented later, either with acute encephalopathy or episodic ketoacidosis or with developmental retardation apparently uncomplicated by attacks of ketosis or acidosis.182,183 A 5-year-old boy presented with a fatal necrosis of the basal ganglia without either metabolic acidosis or hyperammonemia.184 Propionic acidemia has been identified in a 29-year-old man who presented initially with adult-onset chorea and dementia.185 Still other children, with almost complete deficiency of propionyl CoA carboxylase activity as measured in extracts of cultured fibroblasts, have had no clinical abnormalities whatever and have been identified only during family studies.186,187 No satisfactory explanation for this striking lack of clinical enzymatic correlation exists at present.
Based on a survey of 65 patients with propionic acidemia, Wolf et al.181 reported that the clinical course of symptomatic patients is characterized by repeated relapses, usually precipitated by excessive protein intake, constipation, or intercurrent infection. Treatment of these children has been quite difficult, and neurologic sequelae have been common. Among the neurologic complications often observed, developmental delay, focal and general seizures, cerebral atrophy, and EEG abnormalities have been the most prominent. Surtees et al.183 also have reported a high prevalence of neurologic sequelae, including dystonia, severe chorea, and pyramidal signs, particularly in patients who survive longer. The cranial computer tomographic and magnetic resonance imaging findings in propionic acidemia were reviewed by Bergman et al., who showed spectroscopic abnormalities, specifically an increase in glutamine/glutamate, even when the patients appeared to be stable.188 Walter et al. described 11 newborn patients with elevated blood ammonia levels and neurologic symptoms; only 4 had clinically important acidosis.189 Leukopenia and thrombocytopenia, perhaps due to marrow suppression by one or more of the toxic metabolites produced, is also not uncommon. Parathyroid hormone resistance and B-cell lymphopenia was described in a 7-week-old patient.190
Childs and Nyhan7,177,178,191 studied their index patient extensively. Because of the hyperglycinemia, they focused their attention on the pathways of glycine formation and utilization but found no consistent abnormalities. Normal hemoglobin concentration in the peripheral blood indicated that the pathway from glycine to δ-aminolevulinic acid was not blocked. Slices of the patient's liver incorporated [14C]glycine into protein and carbon dioxide as well as slices of rat liver did. Salicylate and benzoate were normally conjugated with glycine, and the glutathione concentration of whole blood was normal. Although the rate of conversion of tritiated glycine to serine in vivo was slower than in controls, this difference may have reflected the enlarged glycine pool rather than a specific block in the conversion of glycine to serine.191
Moreover, several observations suggested an abnormality in the catabolism of the branched chain amino acids, methionine, and threonine. Plasma concentrations of valine, isoleucine, and leucine were elevated intermittently; administration of leucine, valine, isoleucine, threonine, and methionine each precipitated attacks of ketoacidosis, but no other amino acids were toxic. Menkes192 reported that the urine contained large amounts of butanone (a four-carbon ketone that is a by-product of isoleucine catabolism) and the longer chain ketones, pentanone and hexanone. These long-chain ketones were not detected in the urine of patients with ketosis due to diabetes, starvation, or ketogenic diets. Because isoleucine, valine, threonine, and methionine are all precursors of propionate, a defect in propionate metabolism seemed likely, but patient E.G. died before any other studies of propionate catabolism could be performed. Subsequently, Hsia et al.9 demonstrated a striking defect in propionate catabolism in A.G., the affected sister of E.G. When leukocytes isolated from her peripheral blood were incubated with [3-14C]propionate, negligible quantities of 14CO2 were evolved as compared with values in controls, but her cells oxidized methylmalonate and succinate normally. Identical findings were obtained using fibroblasts grown in tissue culture. These data showed that the primary metabolic defect in E.G. and A.G. was in the conversion of propionyl CoA to D-methylmalonyl CoA, a reaction catalyzed by propionyl CoA carboxylase. This conclusion was confirmed subsequently by direct assay of carboxylase activity in fibroblast extracts.175
In their child with lethal neonatal acidosis, Hommes et al.174 found that the serum propionic acid concentration was 400 mg/dl (5.4 mM), a value more than 100 times that reported in normal infants. The liver contained fatty acids with 15 and 17 carbon atoms in addition to the even-chain fatty acids found in control livers. From these data, Hommes et al. also postulated a defect in propionyl CoA carboxylation in their patient.
Subsequent investigations have confirmed and extended these early findings. Analysis of body fluids in several additional patients92-94,176 showed that propionate accumulation in blood and urine occurs regularly, its magnitude being related to the severity of the clinical course and the time at which sampling is performed. Ando and colleagues92 have stressed that other propionate derivatives also accumulate in urine. These include methylcitrate, which is probably formed from the intramitochondrial condensation of propionyl CoA with oxaloacetate92 ; propionylglycine, which results from the conjugation of propionate with glycine93 ; β-hydroxypropionate, an intermediate in one of the alternative pathways of propionate catabolism94 (Fig. 94-4); and tiglic acid,193 an isoleucine catabolite several steps proximal to the block. Although the exact amounts of these compounds in urine have not always been determined, they appear to account for a small fraction of the propionate pool that accumulates in vivo in this disease. Their presence may be important in mitigating the toxic effects of propionate excess. Wendel et al. showed elevated levels of odd-numbered fatty acids (OLCFA) in the erythrocyte lipids of five patients with propionic acidemia and suggested that OLCFA levels reflect the continuous burden of propionyl CoA toxicity within cells and could serve as a means of evaluating the quality of long-term metabolic control.194
Other compounds, not directly concerned with the propionate pathway, also have been found in significantly increased amounts. In addition to hyperglycinemia and hyperglycinuria, which were discussed earlier, marked hyperammonemia has been documented in several patients,181,195 and a distinct correlation between plasma propionate and blood ammonia has been noted in two patients.196
The molecular pathology of propionic acidemia is both complex and interesting. Cell extracts from a number of affected patients share a common finding, namely, reduction in propionyl CoA carboxylase activity to 1 to 5 percent of that in controls.197-199 Because the enzyme is composed of two independently encoded enzyme subunits, the causative mutations will necessarily occur in one of two genes. This was first illustrated in complementation experiments in which fibroblast heterokaryons formed between pairs of affected cell lines were assayed for recovery of functional propionyl CoA carboxylase by fixation of 14C-propionate.200-202 Two major complementation groups, and pccBC, were identified, the latter group showing intragroup complementation (subgroups pccB and pccC) compatible with the occurrence of interallelic complementation (see Fig. 94-11). It was shown subsequently that patients in the pccA group have a primary defect in the PCCA gene encoding the α subunit of propionyl CoA carboxylase, whereas patients in the pccBC group and subgroups have defects of the PCCB gene encoding the β subunit.203,204
Model for interallelic complementation between and pccC mutations causing propionic acidemia. The schematic shows a pair of β subunits aligned head to toe so that two functional domains are produced, with the carboxybiotin site coming from one subunit and the propionyl CoA binding site coming from the other. In the case in which each β subunit contains complementing mutations—near the N-terminus of one subunit for dupKICK140 and near the C-terminus of the other subunit for R410W—the outcome is that only one of the two functional sites is inactivated. The second functional site retains activity despite the presence of a mutation on each subunit.
There are several unusual features of mutant fibroblast lines from patients belonging to the different complementation groups. First, many individuals in the pccA group lack detectable α subunit protein; when this is the case, they invariably lack detectable β subunits as well.203,205 This has been explained by the inherent instability and consequent degradation of the β subunit in the absence of α subunit with which to assemble to form the native enzyme. Among α-minus/β-minus cell lines, some lack α subunit mRNA but contain β subunit mRNA, confirming the assignment of the pccA complementation group to mutations of the PCCA gene.204 Second, a number of or subgroup fibroblasts lack β subunits but have α subunits. These have unstable β subunit protein due to mutation, although β subunit protein is present in at least some cases.205,206 Importantly, the small amount of residual propionyl CoA carboxylase activity observed in extracts of most mutant fibroblasts appears to be present even in those with absent α or β subunits. This suggests that the “background” activity is due to the minimal activity of other carboxylases acting on propionyl CoA as substrate, not to propionyl CoA carboxylase itself.207 A third unusual feature of propionic acidemia is that many heterozygotes of the pccBC group or subgroups have propionyl CoA carboxylase activity indistinguishable from that in controls, whereas obligate heterozygotes of the pccA group have the expected 50 percent of control activity.199 This has proved to be due to relative differences in the synthesis of the enzyme subunits. It has been shown that β subunits are synthesized in four- to fivefold excess over α subunits, so that heterozygotes for null-type β subunit mutations still have more than enough wild-type β subunits to interact with the limiting amount of α subunits to form normal amounts of carboxylase enzyme.205 Conversely, any reduction in the amount of α subunit, as in pccA heterozygotes, is directly reflected in a proportionate reduction in carboxylase activity.199 All three of these features of mutant cell lines can be used diagnostically to identify the affected gene, but caution is warranted because patients with point mutations in either gene may have both subunits present, despite having a defective holoenzyme.
A defect in the carboxylation of propionate provides a satisfactory explanation for many of the findings reported in this disorder. This defect would be expected to lead to an elevated concentration of propionate in the blood and an inability of leukocytes to catabolize propionate to carbon dioxide. Because isoleucine, valine, threonine, and methionine are precursors of propionate, such a block also should lead to the observed protein and specific amino acid intolerance. The appearance of long, odd-chain fatty acids in the liver suggests that when propionyl CoA carboxylation is blocked, odd-chain fatty acid biosynthesis may be augmented because propionyl CoA is the “primer” for such compounds. Finally, the presence of such compounds as butanone, methylcitrate, β-hydroxypropionate, propionylglycine, and tiglic acid very likely results from reversal of reactions proximal to the primary carboxylase block or from increased utilization of alternative pathways.
It is not at all clear from the foregoing, however, why some patients have a severe and often life-threatening course, and others are only mildly affected clinically. Major differences in dietary protein uptake and quality, in the contributions of gut bacteria to total propionate load, or in the activity of alternative mechanisms for propionate disposal are possible explanations for the wide clinical spectrum, but the prominent intrafamilial differences in severity are not easily explained this way.186 Furthermore, several other features of the disease are not adequately explained by the block in propionate catabolism. The ketosis produced in E.G. by leucine is not understood, because this amino acid is not catabolized to propionate. However, it is ketogenic in normal subjects, suggesting that its effect in E.G. was nonspecific. The cause for the hyperglycinemia seen in many, but not all, of these patients also has not been adequately defined. Because the infant described by Hommes et al.174 with massive propionic acidemia never demonstrated signs of hyperglycinemia, the latter cannot be ascribed simply to the acidosis or ketosis. Numerous theses have been put forth in explanation. For example, one or more products of isoleucine catabolism may interfere with glycine cleavage or glycine–serine interconversion.191,208,209 Ando et al.92 speculated that methylcitrate cleavage in the cytosol may yield propionate and glyoxylate, the latter being used as a substrate for glycine overproduction. Impaired glycine conjugation systems have been suggested, but no data in support of this notion have been forthcoming. Because plasma glycine concentration may increase in sick children with negative nitrogen balance of many causes,210 the hyperglycinemia may be nonspecific. The hyperammonemia often observed in this disorder has been the subject of considerable investigation. It appears likely that this secondary but clinically important finding results from inhibition of the first enzyme of the urea cycle, mitochondrial carbamyl phosphate synthetase (CPS I), by the organic acids and CoA esters that accumulate intramitochondrially behind the block in propionyl CoA carboxylation. This conclusion rests on data from studies with experimental animals and animal tissues. For example, propionate inhibits ureagenesis in rat liver slices when ammonia, but not citrulline or aspartate, is the nitrogen-donating substrate.211 Administration to rats of sufficiently large amounts of propionate or methylmalonate to produce hyperammonemia is associated with a marked decrease in hepatic concentration of N-acetyl glutamate,212 the required allosteric effector of CPS I, probably by competitively inhibiting N-acetyl glutamate synthetase.213 That such CPS I inhibition occurs in vivo as well as in vitro is supported by case reports that describe selective impairment of CPS I activity in the livers of patients with propionic acidemia214 or methylmalonic acidemia.215
As a prelude to mutation identification in propionic acidemia, it is first necessary to identify the affected gene. This is most readily done by conducting complementation tests and determining whether the affected patient belongs in the or pccBC group or a subgroup. Determination of the subgroup (e.g., pccB or pccC vs. pccBC) is not specifically required, but does provide insight into the functional impact of mutations affecting the β subunit (see Fig. 94-11). Alternatively, as described above, demonstration of some of the peculiarities of propionyl CoA carboxylase activity or of mRNA or enzyme subunit expression can also reveal the affected gene. Thus, the demonstration of normal enzyme activity in parents of an affected child would be compatible with mutations in the PCCB gene, although obtaining the converse, 50 percent activity, does not necessarily implicate the PCCA gene. Absence of both α and β subunits by Western blotting does identify the PCCA gene as responsible for the disease. Other more straightforward findings, such as absence or abnormality of one of the two mRNA species or polypeptide, also will identify the affected gene. Performing the required experiments is worthwhile because there is a great diversity of mutations in the PCCA gene, and although the PCCB gene shows bias toward a small number of mutations that account for about 30 to 60 percent of alleles in different populations, there remains a large diversity of mutations that account for the rest.
Mutations in the PCCA Gene.
Nineteen disease-causing mutations have been identified in the PCCA gene, eight of which fail to produce a complete α subunit. There are four splicing mutations, two nonsense mutations and two small deletions causing frameshifts. All four splicing mutations cause exon skipping. Three of them—1771IVS-2del9, 1824IVS+3del4, and 1824IVS+ 3insCT—affect the same exon.216 The nonsense mutations, R288X and S537X, and the small deletions, 700del5 and 1115del4, are expected to produce truncated proteins. However, fibroblasts from a patient homozygous for 1115del4 failed to show mRNA by Northern blot, although it could be detected by reverse transcriptase polymerase chain reaction (RT-PCR). A similar finding was made for a patient heteroallelic for R288X and 700del5, indicating that both mutant mRNA species are unstable.217 These results suggest a propensity for mutations disrupting normal mRNA translation to produce mRNA-minus outcomes, as has been noted for other genes.
An unusual finding in patient cells with mRNA destabilizing mutations at both alleles is the detection of an RT-PCR product showing an 84 bp insertion at nucleotide 1209, containing two in-frame stop codons.217 The insertion is an anomalous exon derived by aberrant splicing within the adjacent intron. The cryptic transcript is part of the background “noise” of abnormal mRNAs occurring at very low level in normal cells. Cell lines in which the 84 bp insertion is detected as a predominant (but low-level) species share at both alleles severely deleterious mutations, which are consistent with rendering the normally structured mRNA species unstable and rapidly degraded. In a study of 12 mutant cell lines, 4 showed the characteristic transcript. The mutations included a splicing mutation (1671IVS + 5G→C), a nonsense mutation (R288X, three occurrences), and two small deletions (700del5, 1115del4) causing frameshifts.217 Screening for the 84 bp insertion by RT-PCR may provide a diagnostic benefit, given the high proportion of such cell lines.
The remaining mutations cause amino acid substitutions, which are expected to produce inactive or unstable α subunits, although most have yet to be evaluated in expression experiments. Two subunits with point mutations, G643R and Cdel687, have been expressed in E. coli and shown to abolish biotin binding.57 Another mutation, M348K, has been shown to reduce the intramitochondrial stability of the mutant α chain without appreciably affecting its import.218 This α chain failed to be biotinylated, as also was found for two other nearby mutations, D343G and G354V, and one near the N-terminus, A50P,57 indicating that all four mutations likely result in unstable α chains in vivo, accounting for the lack of biotinylation, because the biotin binding domain is far removed toward the C-terminus of the protein.
Mutations in the PCCB Gene.
Twenty-eight disease-causing mutations have been identified in the PCCB gene. There are 16 missense mutations, three nonsense, three insertions, and one complex insertion/deletion (ins/del). Five other mutations cause disease through disruption of normal splicing. The most frequent mutation in white individuals is ins/del,219-221 present in 32 percent of alleles of mixed white heritage. A similar value also was obtained in a study involving 29 patients of Spanish or Latin American heritage.61 Two other mutations have been found in significant frequency in the same populations. These are 1170insT and E168K, with combined frequencies of 14 percent and 17 percent, respectively, in Spanish and Latin American patients.61,222 The remaining mutations occur singly or have been found in only two or three patients among whites.
Six mutation have been identified among Japanese patients. Two of them, R410W and T428I, appear to be prevalent, occurring in 25 and 31 percent of alleles, respectively.221,223 There are also two splicing mutations (IVS4+3del4 and IVS12+3del8), both of which result in exon skipping; one nonsense mutation (R499X), and one other point mutation (R165W).54,223,224 Of these, R410W and R165W also have been found in whites.61,225
The ins/del mutation is a complex mutation resulting from a deletion of 14 nucleotides and their replacement by 12 nucleotides unrelated to those that were lost. The outcome is the loss of an Msp I site, which has been used diagnostically, and generation of a frameshift that results in the production of an unstable, truncated protein.220 The origin of the ins/del is not obvious, but it is attractive to speculate that it is derived from duplication of nine nucleotides just upstream of the site, with random filling of three more nucleotides.219 The R410W mutation occurs in the same location and also results in loss of the Msp I site. This leaves ambiguous the results of diagnosis by Msp I digestion of PCR products. Dot blotting can be used to distinguish these possibilities,219 or PCR products that show at least one allele not cleaved by Msp I can be sequenced.
Two amino acid substitutions, L519P and R512C, are unusual in that they are associated with apparently absolute deficiency of β subunits.61 Both are present in patients with null-type second alleles (1170insT and ins/del, respectively) and both patients' cells are negative for the β subunit by Western blot. Other point mutations, including R44P, G131R, R165W, E168K, R410W, and A497V are associated with unstable β subunits, which appear reduced in quantity and may show smaller-sized fragments by Western blot.61,223 The abundance of β > subunit mutations that are CRM negative (in the presence of detectable α subunit) highlights the affected gene (PCCB) and focuses the analysis for mutations.
The identification of mutations in the PCCB gene has made it possible to assess the basis of interallelic complemention that defines the and pccC subgroups. Each complementing cell line would have one, if homozygous, or two candidate mutations that could be responsible for the interallelic complementation observed between mutant subgroups. The identity of the complementing allele in several cell lines was determined by microinjecting β subunit cDNA plasmids containing the candidate mutations into fibroblasts of each subgroup and assaying for recovery of 14C-propionate metabolism by autoradiography.225,226 The results paralleled the original complementation results, and the distribution of complementing alleles shows a pattern suggestive of functional domains within the β subunit. For example, two mutations from the pccB subgroup in the N-terminal half of the protein, dupKICK140 and P228L, are both complementing alleles. These mutations reside in the putative carboxybiotin binding domain by homology with related sequences in the 12S subunit of P. shermanii transcarboxylase. Similarly, two complementing alleles from the pccC subgroup, Idel408 and R410W, are close together in the C-terminal half of the protein. They are located in a region of high homology with the propionyl CoA binding site of the P. shermanii 12S subunit. These findings suggest the importance of β–β interactions in the enzyme and indicate the benefit of determining subgroup complementation within the PCCB gene as a way of assessing functional characteristics of the enzyme.
A particular difficulty for investigating the impact of mutations is the requirement for a two-subunit expression system and successful biotinylation to obtain functional propionyl CoA carboxylase. So far, this has been achieved in only one instance. Kelson et al.227 coexpressed cDNAs encoding the α and β subunits in E. coli, along with the chaperonin proteins GroES and GroEL to facilitate folding and assembly, and obtained fully assembled, biotinylated propionyl CoA carboxylase. Using this system, they evaluated the Japanese T428I mutation in the β subunit and showed that it resulted in complete obliteration of enzyme activity.
Diagnosis and Mutation Analysis.
A defect in propionate carboxylation must be considered in any child in whom ketosis or acidosis develops in the neonatal period. Other inborn errors of metabolism must be ruled out, as must the more common causes of acidosis in the newborn period. Determinations of propionic acid and its metabolites in blood or urine and studies of propionyl CoA carboxylase activity in leukocyte or fibroblast extracts are required for definitive diagnosis. The enzymatic test is, in fact, the only absolutely specific one, because propionate accumulation can occur in patients with defects of methylmalonate metabolism as well as in those with propionyl CoA carboxylase deficiency. Such assays on cord blood leukocytes should allow immediate diagnosis in a high-risk newborn. Prenatal diagnosis has been accomplished reliably by measuring carboxylase activity in cultured amniotic fluid cells228 or chorionic villous biopsies,229 by measuring [14C]propionate fixation in amniotic fluid cells,230 or by measuring methylcitrate in amniotic fluid.231
The identification of the specific gene defect, in PCCA or PCCB, responsible for the enzyme deficiency is important for monitoring future pregnancies and to contribute to genotype–phenotype correlations. The gene assignment can be achieved most unequivocally through complementation analysis. However, this is an esoteric procedure that is not readily available in most laboratories.200-202 Other methods can be used that take advantage of some of the peculiarities of each complementation group, although the techniques do not apply universally. For example, estimation of PCC activity in parents' fibroblasts will indicate PCCB gene mutations if the activity is within the normal range, although this does not apply in all cases of β subunit defects.199 Western blot, or 3H-biotin labeling in the case of the α subunit, will indicate defects of the PCCA gene, if both the α and β subunits are absent, or implicate the PCCB gene if only the β subunit is absent, reduced in quantity, or partially degraded.205 Finally, Northern blot or semiquantitative RT-PCR will identify the defective gene if the level or mobility of one or the other mRNA species is adversely affected.204
After identification of the defective gene, cell lines can be analyzed for mutation by conventional techniques. In the PCCB gene, it is worthwhile to first screen for common mutations. These include ins/del, which appears widespread, and 1170insT and E168K, which are frequent in Spanish and Latin American populations.61,219-221 In Japanese patients, the common mutations include R410W and T428I.221,223 When screening for the ins/del by loss of the Msp I site, it should be recalled that R410W is also associated with loss of the same site. Although ins/del has not been observed in Japanese patients, R410W and R165W do occur in both Japanese and white individuals. It would be important to confirm the mutation by sequencing if it is identified through loss of the Msp I site. The PCCA gene does not have predominant mutations. Because a number of patients have had null mutations on both alleles, however, it is worthwhile to test for the characteristic 84-bp insertion detectable by RT-PCR in patients with absent or unstable mRNA.217
For determining the identity of unknown mutations, genomic DNA can be analyzed in the case of the PCCB gene. With the structure of the PCCB gene completed, it is possible to amplify each of the 15 exons and flanking intronic sequences for analysis of PCR products.61 Alternatively, RT-PCR and sequencing have been used to identify mutations at the mRNA level222,223 and to examine the effect of splicing mutations.54,224 In the case of the PCCA gene, where complete genomic structure is still lacking, analysis of overlapping cDNA segments generated by RT-PCR has been the method of choice.57,217,218
A low-protein diet (0.5–1.5 g/kg per day) or one selectively reduced in the content of propionate precursors appears to be the best treatment for the disorder at this time. Such diets will minimize the number of attacks of ketoacidosis but will not necessarily prevent them or allow normal development in all patients. Because fasting has been shown to increase the excretion of propionate metabolites in patients, frequent feeding has been recommended.27 Attacks of ketoacidosis should be treated vigorously by withdrawing all dietary protein and administering sodium bicarbonate parenterally; glucose is also required to avoid catabolism. Acute attacks, particularly those accompanied by hyperammonemia, have been treated with peritoneal dialysis.232 Total parenteral nutrition also has been used to treat critically ill patients.233 Because propionyl CoA carboxylase requires biotin as a coenzyme and because some patients' cells show a biotin-dependent increase in enzyme activity,234 it is possible that certain patients could improve when given supplementary biotin, but no clear example of a biotin-responsive patient with isolated propionyl CoA carboxylase deficiency has yet been documented. On the other hand, dramatic biotin responsiveness has been described in several children in whom propionyl CoA carboxylase deficiency is part of the constellation now called multiple carboxylase deficiency (see Chapter 156).
Two additional therapeutic adjuncts deserve mention. Roe and Bohan reported marked, transient clinical improvement in a child with propionic acidemia given a single oral dose (100 mg/kg) of L-carnitine.173 Because this child's urinary hippurate concentration increased markedly after carnitine, because free plasma carnitine was reduced in three other patients with propionic acidemia,235 and because urinary propionylcarnitine was present in large amounts in such patients,235 these workers proposed that patients with propionic acidemia have a relative carnitine deficiency. Subsequently, Wolff et al.236 observed that L-carnitine supplements significantly reduced the ketogenic response to fasting in patients with propionic acidemia. Thus, it appears that long-term L-carnitine supplementation in these patients warrants serious consideration. To date, no results of long-term treatment with carnitine have been published.
The recognition that gut bacteria may contribute significantly to propionate production in at least some individuals25 has led to the suggestion that specific antimicrobial therapy may be of clinical benefit to some children with propionic acidemia by reducing the total amount of propionate in their serum and tissues.237 Metronidazole (10 mg/kg) has been reported to reduce fecal propionate substantially, concomitant with a reduction in the anaerobic bacterial count.237 Plasma propionate also decreased by 50 to 60 percent in two patients, whereas urinary excretion of propionate metabolites was reduced by an average of 34 percent in four.237 No comment was made on the effect of this treatment on the clinical status of these patients. Others have reported using a similar treatment, but also without any firm evidence of clinical efficacy183 (see the later subsection on Diagnosis and Treatment under the section on The Methylmalonic Acidemias). Further study is clearly required to establish whether such therapy improves the management of acute episodes of metabolic decompensation or provides any long-term benefit with respect to growth, mental development, or neurologic outcome for this condition.
In a retrospective study of 17 patients (12 early-onset, 5 late-onset) from a single hospital over 20 years, 7 patients died (5 early-onset, 2 late-onset).238 The neurologic outcome of the surviving patients was felt to be satisfactory, and was even better for early-onset patients. The investigators felt that other treatments such as liver transplantation or somatic gene therapy could improve the quality of life of propionic acidemia patients in the future.
Not unexpectedly, successful application of therapy, particularly in mildly affected patients, is leading to the survival of women with propionic acidemia into their child-bearing years. One report documents a relatively uneventful pregnancy and delivery in a woman mildly affected with propionic acidemia treated with protein restriction and carnitine supplementation.239 Other such pregnancies are likely to occur; each will need to be handled individually according to the mother's specific clinical and biochemical status.
Multiple Carboxylase Deficiency.
In 1971, Gompertz et al.240 reported a male infant (J.R.) thought to have specific deficiency of the mitochondrial, biotin-dependent enzyme, β-methylcrotonyl CoA carboxylase (Fig. 94-5). This infant developed a diffuse, erythematous skin rash at 5 weeks of age and was admitted to the hospital at 5 months of age because of a worsening rash, recurrent vomiting, irritability, and a mild metabolic acidosis. His urine, which smelled like “tomcats' urine,” was analyzed for organic acids and was found to contain large excesses of β-methylcrotonylglycine, tiglylglycine, and β-hydroxyisovaleric acid. When he was given 10 mg biotin (about 100 times the estimated human requirement) by mouth daily for several days, the rash, vomiting, irritability, and abnormal urine metabolites all disappeared dramatically. Several years later, it became clear that J.R. had multiple—not specific—carboxylase deficiency. His reanalyzed urine contained metabolites characteristic of propionyl CoA carboxylase deficiency, as well as β-methylcrotonyl CoA carboxylase deficiency241 ; his cultured fibroblast extracts were deficient in pyruvate carboxylase,202 as well as in propionyl CoA and β-methylcrotonyl CoA carboxylase242,243 ; and supplementation of the fibroblast growth medium with biotin led to complete correction of the deficiency of all three biotin-dependent enzymes.202,242,243 Subsequently, more than 50 children with multiple carboxylase deficiency have been described.106 These children are now known to suffer from defects in one of two steps in biotin metabolism: the transfer of biotin to apocarboxylases, catalyzed by holocarboxylase synthase244-247 ; or the hydrolysis of biocytin, the biotin-containing product of degraded holocarboxylases, to release biotin, catalyzed by biotinidase.248-250
As a group, children with holocarboxylase synthase deficiency tend to present in the first days or weeks of life with feeding difficulties, hypotonia, lethargy, and seizures; some have a diffuse skin rash or alopecia.106 The first described patient with multiple carboxylase deficiency, J.R., proved to have a defect in holocarboxylase synthase. Early studies showed that his enzyme had a reduced affinity for biotin when assayed in vitro.251,252 This was confirmed by the identification of a point mutation in the biotin-binding region of the enzyme and the production of a biotin-responsive mutant enzyme when it was expressed in E. coli .253,254 Several additional mutations have been identified in patients, many of them clustering within the biotin-binding region of the enzyme.253-257
A second, larger group of children is deficient in biotinidase activity.248-250 These children usually present later in life (mean age of onset 3 months) with a variety of neurologic problems (seizures, hypotonia, developmental delay, hearing loss, optic atrophy). Although a large number of mutations have been identified in the biotinidase gene, several are highly prevalent and account for the majority of alleles in patients.258-260
Both groups respond dramatically to biotin supplements (10 mg daily) with prompt and sustained clinical improvement. Thus, multiple carboxylase deficiency differs markedly from isolated propionyl CoA carboxylase deficiency in response to biotin and, hence, in long-term prognosis. It should be emphasized, however, that the clinical presentations may be very similar. For this reason, urinary metabolite identification is of important therapeutic significance.
For more information on multiple carboxylase deficiency, the reader is referred to Chapter 156.
The Methylmalonic Acidemias
In 1967, Oberholzer,10 Stokke,11 and their colleagues described critically ill infants with profound metabolic ketoacidosis and developmental retardation who accumulated huge amounts of methylmalonate in their blood and urine. These children had none of the hematologic or neurologic stigmata of Cbl deficiency, failed to respond to Cbl supplements, and excreted much larger amounts of methylmalonate than those observed in patients with pernicious anemia.6,7 They were presumed to have a congenital defect of methylmalonyl CoA racemase or of the methylmalonyl CoA mutase apoenzyme (Fig. 94-2). Shortly thereafter, Rosenberg,261 Lindblad,12,13 and their co-workers reported children with similar clinical presentations whose methylmalonic aciduria responded dramatically to pharmacologic but not physiologic amounts of CN-Cbl or AdoCbl. Such children were found subsequently to have a primary defect of AdoCbl synthesis that resulted in impaired mutase activity.14,15 The array of different biochemical and clinical disturbances of methylmalonate metabolism was broadened still further in 1969 and 1970, when Mudd,262 Goodman,263 and their associates described children with methylmalonic aciduria whose clinical and chemical findings differed from those described above. Ketoacidosis was not present, and the increased methylmalonate excretion was accompanied by homocystinuria, cystathioninuria, and hypomethioninemia. This biochemical constellation was interpreted as evidence for defective synthesis of both Cbl coenzymes, with secondary impairment of AdoCbl-dependent methylmalonyl CoA mutase and MeCbl-dependent methionine synthase (Fig. 94-7). These early descriptions, coupled with a body of data to be discussed below, have demonstrated that there are many different biochemical bases for inherited forms of methylmalonic acidemia: two distinct defects of the mutase apoenzyme, one producing complete mutase deficiency ( 0), the other partial deficiency (mut −); two distinct defects of AdoCbl synthesis, one probably due to deficiency of a mitochondrial Cbl reductase (), the other to deficiency of mitochondrial cob(I)alamin adenosyltransferase (CblB); and three distinct defects of both AdoCbl and MeCbl synthesis due to abnormal cytosolic or lysosomal metabolism of cobalamins (CblC, CblD, and CblF). Patients with lesions producing methylmalonic acidemia only (mut 0, mut −, CblA, CblB) share many clinical features and will be discussed as a group; discussion of the other group of patients whose lesions produce methylmalonic acidemia and homocystinuria (CblC, CblD, and CblF) will follow.
Methylmalonyl CoA Mutase Deficiency.
Clinical and Laboratory Presentation.
More than 100 children with isolated mutase deficiency have been documented. Although, as mentioned above, there are four known etiologies for such deficiency, the clinical findings in affected patients from the four groups are remarkable more for their similarities than for their differences. Matsui et al. surveyed264 the natural history in 45 such patients: 15 were mut 0; 5 were mut −, 14 were , and 11 were CblB. There were approximately equal numbers of males and females in each group. Information was obtained from questionnaires completed by the patients' physicians, published reports, unpublished communications, and personal experience. The most common signs and symptoms at the onset of clinical difficulty were lethargy, failure to thrive, recurrent vomiting, dehydration, respiratory distress, and muscular hypotonia (Table 94-1). Little interclass difference was observed for these major clinical manifestations or for such less common ones as developmental retardation, hepatomegaly, or coma. Patients in the mut 0 class, however, presented earlier than those in the other groups (Fig. 94-9). Whereas 80 percent of children in the mut 0 class became ill during the first week of life, less than half the children in the three other groups presented during this interval. Furthermore, clinical onset occurred in 90 percent of mut 0 patients before the end of the first month, whereas onset beyond the first month was observed in an appreciable fraction of patients in each of the other groups. A survey of 20 methylmalonyl CoA mutase locus (mut) patients has reached similar conclusions.265
Age of clinical onset in 45 patients with methylmalonic acidemia. Inset numbers denote percentages of patients in each group. (Reprinted with permission from Matsui SM, Mahoney MJ, Rosenberg LE; The natural history of the inherited methylmalonic acidemias. New Engl J Med 308:857 1993)
Table 94-1: Clinical Presentation in 45 Patients wth Methylmalonic Acidemia |Favorite Table|Download (.pdf) Table 94-1: Clinical Presentation in 45 Patients wth Methylmalonic Acidemia
| ||Mutant Class |
| || |
|Signs and Symptoms at Onset || CblA || CblB || mut– || mut° ||Total |
|Lethargy ||78 ||83 ||100 ||85 ||84 |
|Failure to thrive ||75 ||86 ||40 ||77 ||73 |
|Recurrent vomiting ||58 ||86 ||80 ||77 ||73 |
|Dehydration ||64 ||86 ||100 ||62 ||71 |
|Respiratory distress ||89 ||67 ||50 ||55 ||67 |
|Muscular hypotonia ||44 ||57 ||33 ||91 ||63 |
|Developmental retardation ||36 ||33 ||25 ||65 ||47 |
|Hepatomegaly ||11 ||67 ||0 ||57 ||41 |
|Coma ||50 ||29 ||40 ||38 ||40 |
The laboratory findings in affected patients at the time that methylmalonic acidemia (with or without aciduria) was first documented are shown in Table 94-2. As expected, serum Cbl concentrations were routinely normal. Metabolic acidosis, with blood pH values as low as 6.9 and serum bicarbonate concentrations as low as 5 mEq/L, was observed in the majority of patients in all four groups. Ketonemia or ketonuria was found in 80 percent of patients, with hyperammonemia being only slightly less common, occurring in 70 percent of affected patients. Leukopenia, thrombocytopenia, and anemia were the only other manifestations that were noted in 50 percent or more of this group of patients. Earlier case reports266 reported that hypoglycemia occurs in about 40 percent of affected patients. Inadvertently, this parameter was not assessed in this survey.
Table 94-2: Laboratory Findings in 45 Patients with Methylmalonic Acidemia |Favorite Table|Download (.pdf) Table 94-2: Laboratory Findings in 45 Patients with Methylmalonic Acidemia
| ||Mutant Class |
| || |
|Findings at Clinical Onset || CblA || CblB || mut – || mut° ||Total |
|Normal serum cobalamin ||100 ||100 ||100 ||100 ||100 |
|Metabolic acidosis ||100 ||88 ||100 ||85 ||92 |
|Ketonemia/ketonuria ||78 ||67 ||100 ||85 ||81 |
|Hyperammonemia ||50 ||83 ||80 ||75 ||71 |
|Hyperglycinemia/glycinuria ||70 ||83 ||40 ||70 ||68 |
|Leukopenia ||70 ||45 ||60 ||62 ||60 |
|Anemia ||10 ||45 ||0 ||58 ||55 |
|Thrombocytopenia ||75 ||45 ||40 ||40 ||50 |
It should be mentioned that mutase deficiency is not always associated with serious clinical consequences. Ledley et al.267 reported eight children, between the ages of 18 months and 13 years, who had methylmalonate accumulation in blood and urine, but had no symptoms. Presumably, these apomutase-deficient patients have an enzyme defect so “leaky” that homeostasis is not compromised. At least some of these individuals continued to be symptom free 7 years after the initial report.268
Another report determined that patients, initially ascertained by a newborn screening program with methylmalonic aciduria urine levels of less than 1400 mmoles/mmole creatinine, had normal somatic and cognitive outcomes.269
Conversely, another group of patients appear to have methylmalonic acidemia without a demonstrable defect in mutase activity, at least as measured in cultured cells. Although the elevations of methylmalonate are relatively mild as compared with mutase-deficient patients, they are chronic and usually are discovered on laboratory workup for failure to thrive or developmental retardation.270 Because Cbl metabolism also appears to be normal in this group, the cause of the disease remains an enigma. Roe et al. have described a patient with psychomotor delay, methylmalonic aciduria without episodes of metabolic acidosis, and methylmalonic semialdehyde dehydrogenase deficiency.271 It is possible that others in this group have the same defect.
Chemical Abnormalities in Vivo.
Large amounts of methylmalonic acid have been found in the urine or blood of all reported patients. Whereas normal children and adults excrete less than 0.04 mmole (5 mg) methylmalonate daily, children with isolated methylmalonic acidemia have excreted from 2.1 to 49 mmoles (240 to 5700 mg) in a 24-h period. Their plasma concentrations of methylmalonate, almost undetectable in normal subjects, have ranged from 0.22 to 2.9 mM (2.6–34 mg/dl). In the few patients in whom it was measured, the cerebrospinal fluid concentration of methylmalonate equaled that of plasma (for references to early case reports, the reader is referred elsewhere266). It is important to note that patients with mild, late-onset, or “benign”267 disease may have much lower levels, particularly when clinically asymptomatic.267,268 No relationship between the quantities of methylmalonate accumulated in body fluids and the etiology of mutase deficiency (i.e., apoenzyme vs. coenzyme deficiency) has been reported. Methylmalonate is surely the major, but not the only, abnormal metabolite found in body fluids of these patients. Because propionyl CoA carboxylation is reversible, propionate and some of its precursors (butanone) or metabolites (β-hydroxypropionate and methylcitrate) also accumulate in blood and urine,8,92,93,272,273 although their amounts are small compared with that of methylmalonate.
Several groups have studied the relationship between protein or amino acid loading and methylmalonate accumulation in these patients. Without exception, administration of protein or amino acids known to be precursors of propionate and methylmalonate, such as methionine, threonine, valine, or isoleucine, has resulted in augmented methylmalonate accumulation and, in some instances, ketosis or acidosis.8,10-12 When Cbl-responsive patients are given supplements of this vitamin, such augmentation by methylmalonate precursors is lessened considerably.274 All these findings suggest that patients with discrete defects at the mutase step have a major block in the utilization of methylmalonyl CoA that is expressed as methylmalonate accumulation.
Localization of Enzymatic Defects.
Because the conversion of propionate to succinate is blocked in each of the methylmalonic acidemias, an early screening test for these disorders measured the ability of intact peripheral blood leukocytes or cultured fibroblasts to oxidize [14C]propionate to 14CO2 and compared this with the oxidation of [14C]succinate to 14CO2.261 By including estimation of [14C]methylmalonate oxidation as well, this test can distinguish between deficiency of propionyl CoA carboxylase and of methylmalonyl CoA mutase. Incorporation of [14C]propionate into trichloroacetic acid–precipitable material by intact cultured cells has replaced the more cumbersome 14CO2 evolution technique.230,275 Further discrimination among the methylmalonic acidemias has depended on studies of Cbl uptake and AdoCbl formation by intact cultured fibroblasts, on assays of mutase activity in cell extracts, and on genetic complementation studies with cultured cell heterokaryons.
Mutase Apoenzyme Deficiency.
Morrow and colleagues276 provided the first evidence in vitro for apoenzyme abnormalities and for biochemical heterogeneity among the methylmalonic acidemias. In four patients who had died, they studied mutase activity in liver homogenates by measuring the conversion of DL-[3H]methylmalonyl CoA to [3H]succinyl CoA (Table 94-3). Activity was barely detectable in three and showed no response when AdoCbl was added at concentrations sufficient to saturate the normal enzyme. In the fourth, mutase activity was restored to control values by AdoCbl. These findings were interpreted as evidence for a mutase apoenzyme defect in the first three patients and for defective AdoCbl synthesis in the fourth. These findings were confirmed subsequently in studies with cultured fibroblasts.277 Cells from the first three patients synthesized AdoCbl normally but had much reduced mutase activity in extracts regardless of the amount of AdoCbl added; cells from the fourth had a distinct defect in AdoCbl synthesis.
Table 94-3: Methylmalonyl CoA Mutase Activity in Liver Homogenates from Patients with Methylmalonic Acidemia |Favorite Table|Download (.pdf) Table 94-3: Methylmalonyl CoA Mutase Activity in Liver Homogenates from Patients with Methylmalonic Acidemia
| ||Enzymatic Activity* |
| || |
|Subjects ||Without Added AdoCbl ||With Added AdoCbl (4 × 10-5 M) |
|Controls (3) ||535–866 ||799–1058 |
|Patients || || |
|1 ||1 ||3 |
|2 ||8 ||33 |
|3 ||3 ||7 |
|4 ||80 ||1368 |
Subsequently, it has become clear that two general types of apomutase defects exist. In one type, designated 0 and constituting about two thirds of the mut complementation group, mutase activity in extracts of cultured fibroblasts is undetectable (<0.1 percent of control), even when assayed in the presence of AdoCbl concentrations greatly in excess of that normally required to saturate the enzyme.277-279 When CRM was sought by radioimmunoassay under steady-state conditions in cell lines from 21 such patients, 12 had no immunologically identifiable mutase protein (CRM−), whereas 9 had reduced amounts of CRM ranging from 1 to 40 percent of that found in control extracts.280 In a follow-up study,281 cells from this group of patients were pulse labeled to determine how amounts of newly synthesized mutase protein, detected by specific immunoprecipitation, compared with the CRM values obtained under steady-state conditions. As expected, all CRM+ mutants had easily detectable newly synthesized mutase. Of 11 CRM− lines, however, 5 had amounts of newly synthesized mutase ranging from barely detectable to nearly half that seen in controls. Thus, some apomutase mutations lead to the synthesis of unstable mutase proteins, which are rapidly degraded intracellularly. One other result of this study bears mention. Using a pulse-chase experimental protocol, mitochondrial import and cleavage of the apomutase precursor were studied in control lines and in 38 lines from mut mutants that synthesized mutase protein. In one of the 38 mutant lines, an N-terminal deletion resulted in failure of the mutant mutase to be taken up by mitochondria.281,282 All the others underwent normal mitochondrial uptake and processing.
The second type, designated −, involves a structurally abnormal mutase apoenzyme. The mutant apoenzymes in extracts of these cells retain maximally 2 to 75 percent of control activity, have a Km for AdoCbl approximately 200 to 5000 times normal, show a normal Km for methylmalonyl CoA, and exhibit increased thermolability relative to control enzyme.278,279,283 By radioimmunoassay, the amount of immunologically reactive mutase protein in these extracts ranges from 20 to 100 percent of control.280 Because pairwise crosses between mut 0 and mut − generally yield noncomplementing heterokaryons and because there are affected individuals who appear to be mut 0/mut − compound heterozygotes, both mutant types reflect abnormalities of the locus coding for the apomutase structural gene.278,279 The identification of nonsense and missense mutations within the mutase gene that lead to absent enzyme or abnormal enzyme kinetics have firmly established this conclusion.
Ledley and his colleagues first described the molecular changes in mut patients. They surveyed a number of patient fibroblasts and found no evidence for gross rearrangements at the genomic level, but reductions in mRNA in some lines.284 A number of point mutations were described, including nonsense changes that lead to a 0 phenotype,282 missense mutations that also generate a mut 0 phenotype,285,286 and a missense mutation that generates a mut − phenotype.287 Each of these changes was confirmed as a mutation causing deficiency of mutase activity by expressing the variant protein, either in a mut 0 fibroblast line285,286 or in Saccharomyces cerevisiae cells.287 Interestingly, one of the mut 0 lines was able to complement a number of the other mut 0 and mut − lines, although not all,286 an example of interallelic complementation. To date, 28 mutations and 2 benign sequence changes have been identified, some of which have been characterized by expression in cultured cells or in E. coli .68,288,289 A common mutation (G717V) was found in five black patients of African and African-American ancestry,290,291 and in a series of patients from Japan, six patients carried the same mutation (E117X).292
Figure 94-10 is a linear representation of the structure of human mutase,86 based on the crystal structure of the P. shermanii homologue.87 On it are indicated the locations of a number of the missense mutations in mutase identified so far. The effects of some of these on mutase activity have been rationalized in terms of the predicted three-dimensional structure of the human enzyme.68,86,289,293 The easiest to explain are the 0 mutations G630E and G703R, affecting residues that line the binding pocket for the DMB side-chain of AdoCbl. Although both change flexible glycine residues to charged ones, the main effect is to introduce bulky side chains into the narrow binding pocket, effectively blocking access to it and preventing AdoCbl binding.293 Some changes in the Cbl-binding domain (G623R, G626C) appear to affect the positioning of His627, whose side-chain provides the essential bottom ligand to the bound AdoCbl,86,293 whereas others likely affect the position or interaction of the βα strands that form the Cbl-binding domain. One of these, G717V, results in a highly unstable protein, in addition to modifying AdoCbl binding.68 Mutations in the N-terminal TIM barrel, particularly those producing a mut − phenotype, are less easily explained. W105R appears to affect the substrate channel in the TIM barrel, whereas A377E, V368D, and R369H are in the likely dimer interface. The interaction between dimerization or dimer stability and AdoCbl binding, revealed by the mut − phenotype of R369H,68 remains unexplained.
Schematic representation of the secondary structure of human methylmalonyl CoA mutase, predicted by homology modeling based on the radiographic structure of the P. shermanii mutase α subunit. Helices are denoted by rectangles, β strands by bold arrows, and coils or turns by a thin line. For each of the two main domains, the βα pairs are labeled. The AdoCbl is represented by the horizontal thick line labeled Cbl (the position of the corrin ring) attached to an almost vertical thick line indicating the DMB side chain as it extends into the Cbl-binding domain. The approximate positions of a number of characterized missense mutations determined in patients are indicated on the structure by small ellipses. Although many mutations affecting AdoCbl binding are located in the Cbl-binding domain, several, including R93H, Y231N, and R369H, are in the N-terminal (βα)8 barrel domain. (Redrawn with permission from Thoma NH, Leadley PF: Homology modelling of human methylmalonyl-CoA mutase: A structural basis for point mutations causing methylmalonic aciduria. Protein Sci 5:1922, 1996.)
Even less apparent is an explanation of the interallelic complementation supported by two mutations in the N-terminus of mutase, R93H and G94V. R93H was identified in homozygous form in a cell line that showed interallelic complementation with a number of other lines,286 and G94V was found in heterozygous form in a cell line that appeared to show interallelic complementation with the G717V mutation in vivo.68 Subsequent in vitro experiments with mutant proteins expressed in E. coli have confirmed its ability to complement a range of other mut cell lines (J. Janata and W. Fenton, unpublished observations). This region of mutase is the least homologous between the human and P. shermanii enzymes. In the bacterial enzyme, this region forms a long element that wraps around the TIM domain of the other subunit, and thus appears to be involved with subunit–subunit interactions. How this relates to AdoCbl binding (G94V is mut −)68 and how these mutations modify the effects of mutations in the distant Cbl-binding domain to produce interallelic complementation are questions that may only be answered by crystal structures of the human enzyme and these mutant versions.
It seems clear from the available structures, however, that complementation is not simply a matter of bringing together unaffected regions from each subunit of a mutase dimer to form one “normal” active site,293 as may be the case for the β subunits of PCC (Fig. 94-11). The active site of mutase is clearly contained in a single subunit, with no possibility for sharing elements of it between subunits.87 Moreover, the combined active site model predicts that the kinetic parameters of the restored site should be the same as wild type. This is not what is observed in vivo for the complementing G94V/G717V pair, where the “complemented,” presumably heterodimeric, enzyme has a Km at least 10-fold lower than either of its homodimeric parents, but 100-fold higher than wild type.68 Similar results have been obtained in vitro by combining the individually expressed mutant enzymes (J. Janata and W. Fenton, unpublished observations), suggesting a more complex mechanism for restoration of activity upon interallelic complementation between mutase mutants.
Finally, it should be mentioned that the only patient thus far reported to have methylmalonyl CoA racemase deficiency294 has been restudied and shown conclusively to be a mutant biochemically and genetically.17
Defective Synthesis of AdoCbl.
A series of observations by Rosenberg,14 Mahoney,15 and their colleagues on the fibroblasts of the index patient with Cbl-responsive methylmalonic acidemia led to the demonstration of a primary defect in AdoCbl synthesis in intact cells, resulting in a deficiency of mutase activity. Although intact cells were unable to convert OH-[57Co]Cbl to Ado[57Co]Cbl, cell-free extracts from this line synthesized AdoCbl normally when incubated with OH-[57Co]Cbl, ATP, and a reducing system designed to bypass the Cbl reductases (Fig. 94-8). Subsequent biochemical167 and genetic complementation295-297 studies have differentiated two mutant classes among patients defective in AdoCbl synthesis. One class, which contains the index Cbl-responsive patient, is designated , and may be due to a deficiency of one of the mitochondrial Cbl reductases, perhaps an NADPH-linked aquacobalamin reductase.298 Complementation has been demonstrated between several CblA lines, raising the possibility of interallelic complementation in this disorder as well.299 The second, designated CblB, has been shown to result from a specific deficiency of cob(I)alamin adenosyltransferase.168
All studies in vivo and in vitro in patients with methylmalonic acidemia due to specific methylmalonyl CoA mutase deficiency indicate that the primary block in the conversion of methylmalonyl CoA to succinyl CoA explains the accumulation of methylmalonate in blood and urine; the augmentation of methylmalonate excretion and the precipitation of ketosis by protein, amino acids, or propionate; and the excretion of long-chain ketones formed in the catabolism of branched chain amino acids. However, the primary block does not explain several important physiologic disturbances: the acidosis, hypoglycemia, hyperglycinemia, and hyperammonemia. Oberholzer et al.10 pointed out that the concentration of methylmalonate in the blood (no more than 3 mM) could not alone explain the acidosis and suggested other possibilities. They proposed that an accumulation of CoA, “trapped” intracellularly as methylmalonyl CoA, could lead to an insufficiency of this widely utilized coenzyme and, secondarily, to impaired carbohydrate metabolism and subsequent acidosis. Alternatively, they suggested that an excess of methylmalonyl CoA, a known inhibitor of pyruvate carboxylase,300 could interfere with gluconeogenesis and lead directly to hypoglycemia and indirectly to excessive catabolism of lipid, with ketosis and acidosis. Halperin et al.301 showed that methylmalonate inhibited the transmitochondrial shuttle of malate and argued that impairment of this key step in gluconeogenesis could lead to hypoglycemia. As discussed earlier for deficiencies of β-ketothiolase and propionyl CoA carboxylase, the mechanism of the hyperglycinemia and hyperammonemia so often observed in children with any one of these disorders probably reflects inhibition of the intramitochondrial glycine cleavage enzyme and of CPS I, respectively, by the accumulated organic acids or their CoA esters.208,209,211-215 Thus, as shown in Fig. 94-12, each of the major secondary biochemical abnormalities in the propionic and methylmalonic acidemias can be explained satisfactorily by inhibition of specific intramitochondrial processes by the accumulated organic acids and esters.
Proposed mechanisms of hypoglycemia, hyperglycinemia, and hyperammonemia in patients with inherited deficiencies of β-ketothiolase (see Chapter 93), propionyl CoA carboxylase, or methylmalonyl CoA mutase. Inhibitory effects of the enlarged intramitochondrial pools of acyl CoA esters (such as propionyl CoA) or their respective free acids on selected mitochondrial functions are shown by the numbered dashed lines corresponding to the following enzymatic or shuttle-mediated reactions: (1) pyruvate carboxylase; (2) the transmitochondrial malate shuttle; (3) the glycine cleavage enzyme; (4) carbamyl phosphate synthetase I; and (5) N-acetylglutamate synthetase.
As a further consideration, about half the reported patients with isolated methylmalonic acidemia also show pancytopenia.264 One report suggests that methylmalonate inhibits growth of marrow stem cells in a concentration-dependent fashion.302
By comparing and contrasting the findings in patients with isolated mutase deficiency with those in patients with Cbl deficiency (as in classic pernicious anemia), it has been possible to shed some light on the mechanism responsible for the hematologic and neurologic abnormalities in the latter disorder. Thus, the absence of megaloblastic anemia in any patient with isolated mutase deficiency militates against any involvement of this enzyme in the typical megaloblastosis seen in Cbl deficiency. Similarly, the cerebellar and posterior column abnormalities so often encountered in Cbl-deficient patients have never been observed in patients with methylmalonic acidemia due to specific mutase dysfunction. Therefore, the notion that neurologic dysfunction in pernicious anemia reflects aberrant incorporation of odd-chain or branched-chain fatty acids into myelin because of a block in the propionate pathway has little to recommend it. It appears likely, then, that abnormalities in the Cbl-dependent methionine synthase account for the hematologic and neurologic abnormalities in Cbl-deficient patients. This matter will be discussed further when we consider that group of patients with methylmalonic acidemia and homocystinuria.
Each of the four etiologic bases for specific methylmalonyl CoA mutase deficiency ( 0, mut −, , and CblB) is almost certainly inherited as an autosomal-recessive trait. This conclusion is based on the following findings. First, approximately equal numbers of affected males and females are encountered in each group.264 Second, no instance of vertical transmission from affected parent to affected child has been reported. Third, interclass heterokaryons formed between cell lines from different etiologic groups (i.e., mut0× CblA) complement each other, whereas intraclass heterokaryons (i.e., mut0× mut0) generally do not (with the exception of interallelic complementation); thus, each mutant class behaves as a recessive in culture.295-297 Fourth, cell lines from heterozygotes for the mut0, mut−, and CblB mutations show partial mutase apoenzyme deficiency279 and partial adenosyltransferase deficiency,168 respectively. And fifth, among a large group of mut mutants studied, some have inherited a genetically different mutant allele from each parent, thereby being compound heterozygotes rather than true homozygotes.279,284,285
It is not possible to define with any precision the prevalence of these disorders in the general population. A survey of newborns in Massachusetts has suggested that methylmalonic acidemia may occur in 1:48,000 infants.303 A similar survey in Quebec yielded 1:61,000 infants.304 Because this study screened urines from infants 3 to 4 weeks of age and because it is known that many children with methylmalonic acidemia die in the first week of life from ketoacidosis or hyperammonemia or both, the true prevalence must be greater.
Diagnosis, Treatment, and Prognosis.
Because simple colorimetric assays for urinary methylmalonate and more complex gas chromatography–mass spectrometry assays for serum and urinary methylmalonate are now available, it should no longer be difficult to make a diagnosis of methylmalonic acidemia once this condition is considered. Other sources of neonatal or infantile ketoacidosis must be ruled out. If excessive amounts of methylmalonate are found in the urine, Cbl deficiency can be excluded by direct measurement of serum Cbl concentration. Confirmation and etiologic designation (i.e., or cobalamin metabolism locus Cbl) depend on studies with cultured cells and extracts therefrom.17,305 Prenatal detection of methylmalonic acidemia has been accomplished in two ways: by measurement of methylmalonate in amniotic fluid and maternal urine at mid-trimester306,307 and by studies of mutase activity and Cbl metabolism in cultured amniotic fluid cells.295,307,308 Assays of [14C]propionate utilization230 in uncultured chorionic villous biopsy specimens have proven unsatisfactory, however.309 Mutase apoenzyme307,308 and AdoCbl synthesis230,306 deficiencies have been identified prenatally.
Two treatment regimens for children with methylmalonic acidemia exist and should be used in tandem. A diet restricted in protein (or a special formula restricted in amino acid precursors of methylmalonate) should be instituted as soon as life-threatening problems such as ketoacidosis, hypoglycemia, or hyperammonemia have been addressed; and supplementary Cbl (1–2 mg CN-Cbl or, preferably, OH-Cbl intramuscularly daily for several days) should be given as soon as the diagnosis of methylmalonic acidemia is made (or even seriously considered). Such measures should decrease the circulating concentrations of methylmalonate and propionate. Even Cbl-unresponsive children with delayed development have been shown to improve markedly when treated with careful dietary protein restriction.310,311 As discussed above for patients with propionyl CoA carboxylase deficiency, Roe and associates235,312,313 have pointed out that L-carnitine supplements may be a useful therapeutic adjunct in patients with methylmalonic acidemia, presumably by repleting intracellular and extracellular stores of free carnitine that are depleted in affected patients because of exchange with excess methylmalonyl CoA and propionyl CoA. Likewise, oral antibiotic therapy may prove useful here as well. Thompson and his colleagues report that three patients showed subjective improvement in alertness and appetite following brief metronidazole therapy237 ; longer treatment periods have resulted in significant improvements in other patients, including decreased number and severity of acidotic episodes, increased appetite, decreased vomiting, growth acceleration, and improved behavior.29,314 Total parenteral nutrition also has been used in at least one reported case.233
The previously mentioned survey264 suggests that both the response to Cbl supplements and the long-term outcome in affected patients depend considerably on the nature of the biochemical lesion causing the methylmalonic acidemia. As shown in Figure 94-13, essentially none of the children designated 0 or mut − responded to Cbl supplements with a distinct decrease in blood or urinary methylmalonate, whereas over 90 percent of the and about 40 percent of the CblB patients showed such a response. Given the complete absence of mutase activity in cells from the mut 0 group, it is not surprising that they were regularly Cbl-unresponsive in vivo. The disappointing absence of response in four mut − patients presumably means that even parenteral Cbl supplements could not drive tissue concentrations of AdoCbl sufficiently high to increase significantly their mutase holoenzyme activity. The fraction (60 percent) of CblB patients unresponsive to Cbl supplements presumably has such complete adenosyltransferase deficiency that AdoCbl synthesis cannot be augmented by Cbl supplements, as it apparently can in the CblB patients with leaky mutations that permit responsiveness in vivo. Patients in the CblA group were uniformly responsive, suggesting either that the responsible mutations are generally leaky, thereby allowing mass action to result in more AdoCbl synthesis, or that alternative pathways of Cbl reduction that require high substrate concentrations exist in cells. It should be emphasized that clinical responsiveness in vivo does not mean complete correction of mutase deficiency. Even in a patient whose clinical improvement is dramatic, Cbl administration only reduces, rather than eliminates, methylmalonate excretion. Studies with cultured cells from a variety of patients with methylmalonic acidemia17,297 suggest that raising holomutase activity to only 10 percent of normal values by supplementing the growth medium with OH-Cbl results in distinct augmentation of propionate pathway activity (or, conversely, in a distinct decrease in the magnitude of the metabolic block). Some patients in the CblB group, unresponsive to CN-Cbl or OH-Cbl in vivo, might be expected to respond to AdoCbl itself, but published reports on two patients suggest that this logical alternative is ineffective.315,316
Biochemical response to cobalamin supplementation in 45 patients with methylmalonic acidemia. MMA refers to the concentration of methylmalonate. The supplementation protocol generally used 1 mg CN-Cbl parenterally daily for 7 to 14 days. Inset numbers denote the percentage of patients in each group. (Reprinted from Matsui SM, Mahoney MJ, Rosenberg LE: The natural history of the inherited methylmalonic acidemias. New Engl J Med 308:857, 1983.)
The long-term outlook for affected patients is revealing. As noted in Fig. 94-14, the 0 group has the poorest prognosis, with 60 percent deceased and 40 percent distinctly impaired developmentally at the time of the survey. Shevell et al. reported similar findings.265 In sharp contrast, the CblA patients (i.e., the group biochemically most responsive to Cbl supplements) had the best outcome: 70 percent were alive and well at ages up to 14 years. The CblB and mut − groups were intermediate, with about equal fractions in each group being found in the alive and well, the alive and impaired, or the deceased category.
Long-term outcome in 45 patients with methylmalonic acidemia. The ages of the patients surveyed ranged from a few weeks to 14 years. (Reprinted with permission from Matsui SM, Mahoney MJ, Rosenberg LE: The natural history of the inherited methylmalonic acidemias. New Engl J Med 308:857, 1983.)
It is interesting, albeit anecdotal, that the index patient in the CblA group (now over 30 years old) discontinued Cbl supplements at age 9 years despite advice to the contrary. In the ensuing years, his development and general health have remained excellent despite the high concentration of methylmalonate in his blood and his continued excretion of very large amounts of methylmalonate.
Perhaps, as in some other inherited metabolic disorders, treatment of methylmalonic acidemia is most critical during the early years of life, making expert clinical management in the early weeks or months of life most important. There have been several reports of “metabolic stroke” in patients following episodes of metabolic decompensation.317-319 Three of the patients317,318 belonged to the Cbl-responsive group, but were not being treated at the time. Extrapyramidal signs, particularly dystonia, were accompanied by bilateral lucencies of the globus pallidus and persisted after the acute crisis had passed. In one case, the dystonia has been gradually progressive over a period of 7 years without visible progression of the neurologic lesions.319
Another complication of long-term survival of some methylmalonic acidemia patients is chronic renal failure.320 One report has indicated that 8 of 12 non–Cbl-responsive patients (1–9 years of age) had a reduced glomerular filtration rate, with five severely affected.321 In one of these, “greatly improved metabolic control” over a period of 18 months led to increased, but still impaired, renal function.321 Significantly, the index CblA patient referred to above returned for treatment of moderate renal dysfunction due to biopsy-proven interstitial nephritis. We do not yet know what impact better metabolic control and Cbl supplementation may have in this and similar cases.
A 7-year-old boy with methylmalonic aciduria developed persistent lactic acidosis and multiorgan failure. He was shown to have glutathionine deficiency and responded to treatment with high-dose ascorbate.322
The feasibility of prenatal therapy with Cbl supplements also has been demonstrated. Ampola et al.307 showed that administration of Cbl supplements to a woman carrying a Cbl-responsive, affected fetus resulted in significant reduction in maternal excretion of methylmalonate. Other cases also have been reported.323,324 However, the value of this regimen over one in which therapy is instituted immediately postnatally remains to be established.
Because of the poor prognosis of early-onset severe methylmalonic aciduria, liver transplantation has been attempted in a limited number of patients.325-328 Although liver transplantation appears to protect against acute metabolic decompensation, biochemical correction is incomplete and it is not certain that there will be complete protection against the renal and neurologic complications. Finally, it must be noted that preliminary steps have been taken toward somatic gene therapy for mutase deficiency.329,330 Besides all the usual questions of safety and long-term stability of response that surround somatic gene therapy, two important issues remain unanswered for mutase: How much activity must be restored in vivo to normalize the biochemical hallmarks of the disease? And, does correction of the defect in the liver, for example, lead to reversal or amelioration of the pathologic changes in other organ systems, such as the kidneys, and overall clinical improvement? Much work remains to be done before we can address these problems.
Combined Deficiency of Methylmalonyl CoA Mutase and Methionine Synthase.
Clinical and Laboratory Presentations.
Many patients with inherited combined methylmalonic acidemia and homocystinuria have been the subject of individual case reports.263,331-347 Cells from these children comprise three biochemically and genetically distinct complementation groups, designated , CblD, and CblF.167,295,296,348
Clinical findings have varied widely among the more than 100 known patients in the CblC group, including some who have been diagnosed only in adult life. In a review of 50 patients, 44 had onset in the first year of life and 6 had onset after 6 years of age.349 The median age of onset was 1 month, and the range was from birth to 14 years. Thirteen of the early-onset patients died, with a mean age of death at 9.6 months and a range of 1 to 47 months. The early-onset patients had feeding difficulties, hypotonia, failure to thrive, seizures, microcephaly, developmental delay, cortical atrophy, hydrocephalus, nystagmus, pigmentary retinopathy, and decreased visual acuity. Blood findings included megaloblastic anemia, thrombocytopenia, leukopenia, and neutropenenia. In some patients there was renal failure, sometimes with a hemolytic-uremic syndrome. The later-onset patients presented in childhood or adolescence with acute neurologic findings, which included decreased cognitive performance, confusion, dementia, delirium, myelopathy, and tremor. Only one late-onset patient in this series had pigmentary retinopathy. Hematologic abnormalities were seen in half the late-onset patients. Significantly, serum Cbl and folate concentrations are generally normal in patients.
Neither of the two brothers in the CblD group263 had any clinically significant problems until later in life. The older brother came to medical attention because of severe behavioral pathology and moderate mental retardation at 14 years of age. He also had a poorly defined neuromuscular problem involving his lower extremities. His then 2-year-old brother was asymptomatic, although biochemically affected. No hematologic abnormalities were noted in either sib.
Five patients have been reported in the CblF group. The first two (both female) were small for gestational age and had methylmalonic aciduria, poor feeding, growth retardation, and persistent stomatitis.344,345 One had minor facial anomalies, dextrocardia, and abnormal Cbl absorption from the gut. The other had a persistent rash, macrocytosis, and elevated homocysteine and died suddenly despite a good biochemical response to Cbl treatment. A male CblF patient had recurrent stomatitis in infancy, arthritis at age 4, and confusion, disorientation, and a pigmentary dermatitis at age 10.347 Another boy had aspiration pneumonia at birth, hypotonia, lethargy, hypoglycemia, thromobocytopenia, and neutropenia. A native Canadian girl was diagnosed at 6 months of age because of anemia, failure to thrive, developmental delay, recurrent infections, low serum Cbl, and Cbl malabsorption.346 As a group, the patients have responded well to treatment with Cbl.
Chemical Abnormalities in Vivo.
In addition to the methylmalonic aciduria and homocystinuria that characterize the CblC, CblD, and CblF groups of patients, some have shown hypomethioninemia and cystathioninuria. This constellation of chemical abnormalities, plus the normal serum Cbl values, led to the proposal262,331 that these children suffered from a defect in cellular metabolism of Cbl such that both Cbl-dependent enzyme activities (mutase and methionine synthase; see Fig. 94-7) were deficient. The methylmalonic aciduria in these children is distinctly less severe than that encountered in children with isolated mutase deficiency. It is important to note that elevations of homocystine may be unremarkable in some CblC patients, even when acutely ill. One of the CblF patients had no detectable homocystinuria despite a cellular deficit in methionine synthase activity,344 although all the others have shown homocystinuria.345-347 Moreover, neither hyperglycinemia nor hyperammonemia has been reported in any of the CblC, CblD, or CblF patients.
Localization of Defective Cellular Metabolism of Cbl.
It has long been clear that patients in the and CblD groups have a defect in cellular metabolism of Cbl. This conclusion is based on the following data: total Cbl content of liver, kidney, and cultured fibroblasts is markedly reduced262,332,350,351 ; the ability of cultured cells to retain 57Co-labeled CN-Cbl 352 or to convert 57Co-labeled CN-Cbl or OH-Cbl to AdoCbl and MeCbl is markedly impaired15,296 ; activity of methylmalonyl CoA mutase and methionine synthase in cultured cells is deficient, such deficiency being improved by supplementation of the growth medium with OH-Cbl296,297,353 ; and the mutase and methionine synthase apoenzymes in cells from affected patients appear to be normal.262,263,296,297,354 The precise nature of the metabolic defect in the CblC and CblD classes remains elusive, but some progress has been made. Because these mutant cells demonstrate normal receptor-mediated adsorptive endocytosis of the TC II–Cbl complex and normal intralysosomal hydrolysis of TC II,17,153,154,296 perusal of Fig. 94-8 makes it clear that the defects in the CblC and CblD cells affect some step or steps subsequent to cellular uptake, common to the synthesis of both coenzymes, and prior to the binding of the Cbl coenzymes to their respective apoproteins. Significantly, CblC (and, to a lesser extent, CblD) cells use CN-Cbl less well than OH-Cbl353,355 and are unable to convert CN-Cbl to OH-Cbl, a step shown in normal cells to be a metabolic prerequisite for the synthesis of both AdoCbl and MeCbl.355 The latter results have been interpreted as evidence for a defect in a cytosolic cob(III)alamin reductase, which is required for reducing the trivalent cobalt of Cbl prior to alkylation.355 In both CblC and CblD fibroblast extracts, partial deficiencies of CN-Cbl β-ligand transferase and microsomal cob(III)alamin reductase have been described by Pezacka.356,357 Watanabe and colleagues have described a partial deficiency of a mitochondrial NADH-linked aquacobalamin reductase in CblC fibroblast extracts.298 The suggestion that glutathionyl Cbl may be an intermediate in the reductive pathway358 provides another potential site for the mutation in one of these groups. Finally, it should be pointed out that the distinction between the CblC and CblD classes is based first and foremost on complementation studies that define the two classes as unique,296 although essentially only one example of CblD exists. The biochemical differences between them appear to be quantitative rather than qualitative, with the CblC group having more severe metabolic derangements (and, at the same time, more severe clinical involvement) than the sibs designated CblD. Thus, the possibility must be considered that CblD is an allele of CblC that shows interallelic complementation.
Studies using cultured fibroblasts from two patients in the CblF group are of particular interest. As with cells from CblC and CblD patients, both mutase and methionine synthase activities were impaired, and AdoCbl and MeCbl contents were reduced. In contrast to the and CblD mutants, however, the CblF cells accumulated unmetabolized, non–protein-bound CN-Cbl in lysosomes.359 These findings indicate that CblF cells are deficient in the mediated process by which Cbl vitamers exit from lysosomes after being taken up by receptor-mediated endocytosis.
The biochemical features of patient fibroblasts that distinguish the various methylmalonic acidemias are summarized in Table 94-4.
Table 94-4: Salient Biochemial Features of Cultured Fibroblasts from Patients with the Various Methylmalonic Acidemias |Favorite Table|Download (.pdf) Table 94-4: Salient Biochemial Features of Cultured Fibroblasts from Patients with the Various Methylmalonic Acidemias
| ||Mutant Class |
| || |
|Biochemical Parameter || mut 0 || mut – || CblA || CblB || CblC || CblD || CblF |
| Studies with intact cells || || || || || || || |
|[14C]propionate oxidation ||– ||– ||– ||– ||– ||– ||– |
|[14C]MeH4F fixation ||+ ||+ ||+ ||+ ||– ||– ||– |
|MeCbl synthesis ||+ ||+ ||+ ||+ ||– ||– ||– |
|AdoCbl synthesis ||+ ||+ ||– ||– ||– ||– ||– |
|Conversion of CN-Cbl to OH-Cbl ||+ ||+ ||+ ||+ ||– ||± ||– |
|Lysosomal efflux of free Cbl ||+ ||+ ||+ ||+ ||+ ||+ ||– |
| Enzyme activities in cell extracts * || || || || || || || |
|Mutase holoenzyme ||– ||– ||– ||– ||– ||– ||nt |
|Mutase total enzyme ||– ||± ||+ ||+ ||+ ||+ ||nt |
|Met synthase holoenzyme ||+ ||+ ||+ ||+ ||– ||– ||– |
|Met synthase total enzyme ||+ ||+ ||+ ||+ ||± ||± ||± |
|Cob(l)alamin adenosyltransferase ||+ ||+ ||+ ||– ||+ ||+ ||+ |
The megaloblastic anemia so commonly observed in the CblC patients almost surely reflects the enzymatic disturbance of methionine synthase. This can be stated with some assurance because patients with isolated methylmalonyl CoA mutase deficiency (mut 0, −, CblA, CblB) more severe than that encountered in the CblC patients exhibit no such hematologic dysfunction. The early and severe central nervous system abnormalities encountered in the CblC group probably reflect the methionine synthase abnormality as well, in that such patients do not have the severe metabolic ketoacidosis that probably accounts for the neurologic problems in patients with mutase deficiency only. Thus, patients with severe, inherited dysfunction in the synthesis of both Cbl coenzymes resemble closely patients with exogenous Cbl deficiency, both groups having prominent hematologic and neurologic manifestations resulting from the blocked methionine synthase system.
Because equal numbers of affected males and affected females exist in the CblC group, because females have been as seriously affected as males, and because cells from affected patients behave as recessives in complementation studies,295 it seems safe to predict that this disorder is inherited as an autosomal-recessive trait. The mode of inheritance of the and the CblF mutations cannot yet be defined with certainty, because of the paucity of known patients. Identification of heterozygotes for CblC, CblD, or CblF has not yet been accomplished. One additional contribution of the somatic cell genetic studies used to characterize these disorders deserves mention. The locus coding for the human methionine synthase structural gene was originally mapped to chromosome 1 using human–hamster hybrids.354 This assignment has been confirmed with the cloning of the gene.360-363
Diagnosis, Treatment, and Prognosis.
The combination of methylmalonic aciduria, homocystinuria, and normal serum Cbl concentrations is the triad needed to distinguish patients in the CblC, CblD, or CblF groups from those with isolated mutase deficiency, with other causes of homocystinuria such as cystathionine synthase deficiency or N 5,10 -methylenetetrahydrofolate reductase deficiency, with Cbl deficiency, or with the CblE and CblG mutations affecting only methionine synthase (see Chapters 88 and 155). Such distinctions, easily confirmed by cell studies, are critical because appropriate therapy depends on them. Whereas exogenous Cbl deficiency will respond dramatically to physiologic amounts of Cbl and certain forms of homocystinuria will respond to supplements of pyridoxine or folate, successful treatment of , CblD, or CblF patients may demand administration of very large amounts (up to 1 mg daily) of OH-Cbl.263,333,335-337,344 Such treatment has resulted in dramatic decreases in urinary methylmalonate (and less dramatic decreases in urinary homocystine) in patients who have received it.364 Supplementation with betaine can reduce homocysteine levels and restore methionine, although the clinical effects of this treatment are unclear.365 Early diagnosis and prompt institution of therapy with Cbl supplements (and betaine) may be the only way to change the outcome of these patients, which, at least in the case of the CblC group, has been dismal thus far.366 Documentation of experience with such treatments will be particularly important in assessing the clinician's ability to modify the natural history of these disorders.