Fatal Malonyl CoA Decarboxylase Deficiency Due to Maternal Uniparental Isodisomy of the Telomeric End of Chromosome 16

Authors


Corresponding author: Enrico Zammarchi, Neurometabolic Unit, Department of Pediatrics, University of Florence, Via L. Giordano 13, Florence Italy. Fax: +39-055-570380. E-mail: enrico.zammarchi@unifi.it; neuromet@meyer.it

Summary

Malonic aciduria is a rare autosomal recessive disorder caused by deficiency of malonyl-CoA decarboxylase, encoded by the MLYCD gene.

We report on a patient with clinical presentation in the neonatal period. Metabolic investigations led to a diagnosis of malonyl-CoA decarboxylase deficiency, confirmed by decreased activity in cultured fibroblasts. High doses of carnitine and a diet low in lipids led to a reduction in malonic acid excretion, and to an improvement in his clinical conditions, but at the age of 4 months he died suddenly and unexpectedly. No autopsy was performed.

Molecular analysis of the MLYCD gene performed on the proband's RNA and genomic DNA identified a previously undescribed mutation (c.772–775delACTG) which was homozygous. This mutation was present in his mother but not in his father; paternity was confirmed by microsatellite analysis. A hypothesis of maternal uniparental disomy (UPD) was investigated using fourteen microsatellite markers on chromosome 16, and the results confirmed maternal UPD. Maternal isodisomy of the 16q24 region led to homozygosity for the MLYCD mutant allele, causing the patient's disease. These findings are relevant for genetic counselling of couples with a previously affected child, since the recurrence risk in future pregnancies is dramatically reduced by the finding of UPD. In addition, since the patient had none of the clinical manifestations previously associated with maternal UPD 16, this case provides no support for the existence of maternally imprinted genes on chromosome 16 with a major effect on phenotype.

Introduction

Malonic aciduria (MIM 248360) is a rare autosomal recessive inborn error of metabolism, due to deficiency of malonyl-CoA decarboxylase (MCD, MIM 606761) that catalyzes the decarboxylation of malonyl-CoA to acetyl-CoA. This deficiency results in an accumulation of malonyl-CoA, which is a powerful inhibitor of carnitine palmitoyltransferase-I (CPT-I), causing impaired β-oxidation of fatty acids in both mitochondria and peroxisomes (McGarry & Brown 1997; Bennet et al. 2001; Wightman et al. 2003). Three different CPT-I enzymes have been described: one of these is expressed in the heart and skeletal muscles, and is much more sensitive to malonyl-CoA inhibition than the liver-specific form. A third enzyme, recently described by Price et al. (2002), is localized in the brain and has a high affinity for malonyl-CoA.

The symptomatology of malonic aciduria is characterized by central nervous system involvement (seizures, psychomotor retardation, muscular tone alterations), cardiomyopathy, muscular weakness, and crisis of metabolic acidosis, partly correlated with the localization and sensitivity to malonyl-CoA of the three CPT-I enzymes.

The diagnosis of malonic aciduria is based on the findings of a high urinary excretion of malonic acid, methylmalonic acid, and a mild increase in dicarboxylic acid; confirmation is usually obtained by the determination of malonyl-CoA decarboxylase activity in cultured fibroblasts, although occasionally normal enzyme activity has been observed in patients with malonic aciduria (Ozand et al. 1994; Gregg et al. 1998).

Acylcarnitine analysis by tandem mass spectrometry shows high blood levels of malonylcarnitine and propionylcarnitine, which can be detected by neonatal screening before the appearance of symptoms (Santer et al. 2003; Ficicioglu et al. 2005; Henderson et al. 1998).

The human MLYCD gene encoding malonyl-CoA decarboxylase has been mapped to chromosome 16q24.3 (FitzPatrick et al. 1999; Gao et al. 1999; Sacksteder et al. 1999). It comprises 5 exons and has 3 potential start codons, although only 2 of these appear to be functional (Wightman et al. 2003). So far, fourteen different mutations have been identified in this gene (FitzPatrick et al. 1999; Gao et al. 1999; Wightman et al. 2003); some of these have been reported to cause subcellular protein mistargeting (Wightman et al. 2003).

Autosomal recessive diseases are characterized by the presence of two defective genes on each complementary DNA-strand (one from each parent). However, with the development and increasingly widespread use of molecular analysis to characterise the genetic bases of inherited diseases, conditions have been found where only one parent is heterozygous for a mutation, albeit very rarely. This can be due to the occurrence of de novo mutations or to UPD (Zlotogora 2004). De novo mutations have been identified in patients with 21-hydroxylase deficiency (Asanuma et al. 1999; Baumgartner et al. 2001), spinal muscular atrophy (Wirth et al. 1997; Lefebvre et al. 1998), and other diseases. UPD, leading to isodisomy and hence exposure of recessive alleles, has been reported for several disorders (Zlotogora 2004) such as mitochondrial diseases (Tiranti et al. 1999), organic acidurias such as methylmalonic aciduria (Abramowicz et al. 1994), trifunctional protein deficiency (Spiekerkoetter et al. 2002), primary hyperoxaluria type 1 (Chevalier-Porst et al. 2005) and fumarase deficiency (Zeng et al. 2006).

Here we describe an infant affected by malonic aciduria with neonatal onset, in whom molecular analysis showed a homozygous MLYCD mutation due to maternal UPD following an error in the first meiosis.

Materials and Methods

Case Report

The male patient was the first child of healthy unrelated Italian parents with an unremarkable family history. He was born at term after an uncomplicated pregnancy and delivery; birth weight was 4000 g, length 54 cm, and head circumference 35 cm. Apgar scores were 9 and 10 at 1 and 5 min, respectively.

Tremors of arms and legs were noted since birth, and on day 2 of life tachypnea, pallor and feeding refusal became evident. Blood glucose concentration was 1.44 mmol/L, and arterial blood gas values indicated a compensated mild metabolic acidosis (pH 7.37, bicarbonate 17.1 mEq/L, base excess −6.1). A chest X-ray and ECG were normal. The echocardiogram showed mild hypertrophy of the right sections, interatrial defect type ostium secundum, tricuspidal insufficiency, and pulmonary hypertension (50 mm Hg). Cerebral ultrasound revealed no abnormalities.

On day 3 of life the patient was transferred to our hospital for diagnostic evaluation and further management. On admission the child was tachypnoic (90 breaths/min) with noisy respiration, hypotonic, and soporous with deep tendon reflexes that were elicitable with difficulty; his heart rate was 140/min. Echocardiographic examination confirmed pulmonary hypertension and hypertrophy of the right atrium, and a hypertrophic and dilated right ventricle was also found. Routine laboratory findings were normal, except for a blood glucose level of 2.27 mmol/L. A few hours after admission he presented with an episode of generalised hypertonia and vomiting. Nasogastric tube feeding was started, and the EEG showed an abnormal background and bilateral spike waves.

Metabolic investigations performed on day 5 of life showed the following results: normal plasma amino acid profile, ammonia 92 μmol/L (normal <62), blood lactate 2.10 mmol/L (normal <2.44). The serum total and free carnitine levels were very low. Urinary organic acid analysis by gas chromatography/mass spectrometry revealed high excretion of malonic, methylmalonic, succinic, adipic and suberic acids, suggesting a malonyl-CoA decarboxylase deficiency. Malonyl-CoA decarboxylase activity in cultured fibroblasts was reduced, confirming the diagnosis.

The patient's treatment included high doses of L-carnitine (250 mg/Kg/day) and a high carbohydrate/low fat diet (30% Energy (Joules) from fat while in the first 3 months of life 50% is recommended). After a few days the child was fed by bottle. After 20 days treatment, urinary malonic acid was decreased (49 mmol/mol creatinine) and the other organic acids normalized.

Echocardiograms performed in the following days showed a progressive reduction of the pulmonar arterial pressure and of the hypertrophy of the right sections, and on day 20 the examination was normal. The brain MRI performed at age 20 days showed diffuse altered signal involving the white matter, which was most evident in the subcortical parietal on the left fronto-occipital portions bilaterally, as well as in the semioval centres. An alteration of neuronal migration with unilateral polymicrogyria affecting the right frontal lobe, and polygyria of left frontoparietal lobes, was present.

The patient gradually improved and began to thrive. During hospitalization a transient increase of alanine aminotransferase (108 IU/L, normal < 50), aspartate aminotransferase (163 IU/L, normal < 50), lactate dehydrogenase (1252 IU/L, normal < 560) and creatine kinase (331 IU/L, normal < 170) was observed.

The patient was discharged from hospital at 1 month of age in good clinical condition, with a low fat diet and L-carnitine (250 mg/Kg/day). At home the child was well but, in spite of a progressive reduction of lipids to 20% of caloric intake, the excretion of malonic acid increased again to 700–800 mmol/mol creatinine, with an increase in the other organic acids found at the earlier diagnosis. At age 4 months he showed a mild psychomotor development delay with normal social behaviour. Moreover, he presented with a mild spasticity with increased deep tendon reflexes and right upper limb dystonia. The VEPs revealed a delayed response bilaterally, while the BAERs were normal. After the age of 4 months cardiokinetic and diuretic drugs were added to L-carnitine therapy, because of the worsening of myocardial function. At age 4 months and 15 days the child died suddenly and unexpectedly. Autopsy was not granted. Written informed consent was obtained from the patient's parents, in accordance with the local institutional review board and medical ethics committee guidelines.

Enzymatic Analysis

Malonyl CoA decarboxylase activity was assayed in cultured fibroblasts from a skin biopsy using the method of Scholte et al. (1987) with some modifications. The potassium phosphate buffer (pH 7.0) concentration was 200 mM, that of [1,3 14C] malonyl-CoA 0.5 mM (675 dpm/nmol), and 4 mM L-carnitine, 2 μl purified carnitine acetyltransferase (0.8 U) and 5 mM iodoacetamide were added. The protein amounts were about 50 μg, volume was 100 μl, incubation time 1 h and reaction temperature 37 °C. The reactions were run in triplicate.

Cell Culture

Human skin fibroblasts from the patient were cultured in Dulbecco's modified Eagles-Hams F10 medium (1:1 vol/vol) with fetal bovine serum (10%) and antibiotics.

RNA Isolation and Synthesis of cDNA

Total RNA was isolated from patient's cultured skin fibroblasts and from parents' lymphocytes using a RNA Isolation Kit purchased from Gentra Systems (Minneapolis, USA). RNA integrity was verified by 0.8% agarose gel electrophoresis, and its concentration determined by the absorbancy at 260 nm. About 500 ng of total RNA and 800 ng oligo-dT, and/or 400 ng of specific oligonucleotide primers mapping to the 3′UTR region of the wild-type MLYCD cDNA, were used for cDNA synthesis. RNA and primers were heated to 68°C for 2 min and chilled on ice. The reaction was carried out in 50 μl total volume by adding of 20 U AMV reverse transcriptase (BM), 10 μl 5X reverse transcriptase buffer (BM), 10 mM each dNTPs and 20 U RNase inhibitor (Boehringer Mannheim Biochemicals, Indianapolis, USA). cDNA synthesis was carried out at 42°C for 1 h, using either specific primers or oligo dT. RT-PCR analysis of MLYCD cDNA was carried out using a set of 5 RT-nested overlapping amplifications encompassing the entire coding region. The sequences of the cDNA primers used in these experiments were as follows:

  • F1. forward 5′ CAGGCGTCTCCTCCCGCTG 3′

  • R1. reverse 5′ CCTTTCAGCACCCCATTCATT 3′

  • F2. forward 5′ CACCACATCAGCAAGCTGGAC 3′

  • R2. reverse 5′ GATTGCCTGGATGTTGCTGGA 3′

  • F3. forward 5′ GTGAGGCTGAGGCTGTGCATC 3′

  • R3. reverse 5′ CGTTTGCGAGTTCAGAAGCCC 3′

  • F4. forward 5′ GTCGTCAAGGAGTTGCAGAGA 3′

  • R4. reverse 5′ GTTCTGCAGGTGGAAGTTGG 3′

  • F5. forward 5′ GTGCGCCTGGTACCTGTATGG 3′

  • R5. reverse 5′ GCATGTCACGCCAGGTAGGAA 3′

Amplification reactions were performed under the following conditions: one microliter of MLYCD-cDNA, corresponding to about 20 ng of total RNA, was amplified with 2.5 U Ampli Taq DNA polymerase (Perkin Elmer Cetus, Branchburgh, NJ), 0.25 mM of dNTPs, 200 ng of each primer and 1XPCR reaction buffer (pH 8.8) in a total volume of 25 μl. Cycling conditions for these primer sets were: denaturation at 94°C for 4 min, then 30 cycles of 94°C for 30 s, 63°C for 30 s, and 72°C for 2 min, with a final extension of 72°C for 10 min.

PCR Amplification and cDNA Sequencing

The 5 fragments of nested RT-PCR products were checked on a 2% agarose gel, excised and purified using a QIAquick Gel Extraction kit (Quiagen). Approximately 100 ng of purified amplifications were used in sequencing reactions. Both strands were sequenced with the same primers used for PCR amplification.

The sequencing reactions were performed using Big Dye Terminator Cycle Sequencing Ready Reaction Kit reagents (PE Biosystems). The reactions were run on an ABI 310 sequencer (PE Biosystems) and analysed using Sequencing Analysis Software, version 3.3.

Analysis of Genomic DNA

To confirm the mutation in the MLYCD gene, genomic DNA from the patient and both parents was prepared from cultured fibroblasts and lymphocytes, respectively. The MLYCD gene region encompassing the mutation was amplified by specific oligonucleotide primers and PCR conditions previously reported by FitzPatrick et al. (1999). The genomic fragments were directly sequenced.

Chromosome 16 microsatellites from the ABI PRISM Linkage Mapping set version 2.5 (Applied Biosystem) were used on DNA samples from the proband and his parents to determine the parental origin of each copy of chromosomes 16. PCR was performed under the conditions provided by the manufacturer. Samples were resolved by capillary electrophoresis on an ABI Prism 310 Genetic Analyzer (Applied Biosystems). Paternity was investigated by microsatellite analysis using a standard commercial kit (Amplifiler, Applied Biosystems).

Results

Urinary organic acid at diagnosis showed an abnormal excretion of malonic acid 731 mmol/mol creatinine (normal <2), methylmalonic acid 54 mmol/mol creatinine (normal <2), succinic acid 343 mmol/mol creatinine (normal <102), fumaric acid 15 mmol/mol creatinine (normal <2), glutaric acid 77 mmol/mol creatinine (normal <2), adipic acid 150 mmol/mol creatinine (normal <12), suberic acid 33 mmol/mol creatinine (normal <2), lactic acid 37 mmol/mol creatinine (normal <25), 3-OH-butyric acid 82 mmol/mol creatinine (normal <3), and acetoacetic acid 21 mmol/mol creatinine (normal <2). Total and free carnitine levels were 9.9 mmol/L (normal 35–60) and 2.4 mmol/L (normal 28–50) respectively. These data and the clinical manifestations suggested a diagnosis of malonyl-CoA decarboxylase deficiency, confirmed by the reduction of enzyme activity in cultured fibroblasts [10 pmol/(min.mg protein)] vs. control cell lines (211, 216); so the residual activity was 4.7%.

A retrospective acylcarnitine analysis by tandem mass spectrometry of a neonatal blood spot from the patient was performed. Although the blood spot had been stored at room temperature for a long time (the sample had been collected 9 years before analysis) malonylcarnitine levels were in the upper normal range (0.25 μmol/L; normal 0.01–0.25, 99th percentile, n= 5000).

Molecular analyses performed on the patient's cDNA showed a homozygous 4 bp deletion (c.772–775delACTG) in the MLYCD gene. The mutation was confirmed by direct sequencing of a PCR-amplified genomic DNA fragment encompassing exon 3. No other mutations were identified in the coding region, including all intron–exon junctions of the gene. Genomic DNA sequencing indicated that the patient's mother was heterozygous for the c.772–775delACTG mutation, while this mutation was not present in the father (Fig. 1). Paternity, supported by parental consent, was confirmed by microsatellite analysis.

Figure 1.

Molecular analysis of family's DNA. (a) Partial nucleotide sequence of patient's MLYCD gene showing the new c.772–775delACTG mutation in the homozygous state. (b) Partial nucleotide sequence of MLYCD gene from father showing both normal alleles. The boxed nucleotides are deleted in the patient. (c) Mother's nucleotide sequence revealing the c.772–775delACTG mutation in the heterozygous state.

Therefore, the possibility of maternal UPD for the chromosomal region encompassing the MLYCD gene was investigated. Segregation analysis of fourteen chromosome 16 polymorphic markers was compatible with maternal UPD (mat UPD-16). The marker order, as well as the position of the MLYCD gene, is indicated in Fig. 2. The patient was heterozygous for 6 out of 7 markers located on the short and proximal long arm of chromosome 16 (from D16S423 to D16S415). For 3 of these markers (D16S415, D16S3068 and D16S3046) he did not share any allele with his father; similarly, the D16S3075 82 bp allele present in a homozygous state in the proband was not found in his father. For the remaining 3 markers located in this region, for which the same allele was present in the proband and his father, the genotype was still compatible with maternal heterodisomy. A complex situation was observed for the 16q21–16qter region. Homozygosity for 1 maternal allele at loci D16S503 and D16S515 in the proband indicated maternal isodisomy for the 16q21 region. Genotypes of the next distal marker, D16S516, were compatible with maternal heterodisomy, while maternal isodisomy was again detected for the most distal markers, D16S520 and D16S3123. D16520 is located approximately 2.6 Mb distally to MLYCD. Overall, the results demonstrated that the patient had maternal UPD due to meiosis I non-disjunction, and that the distal portion of chromosome 16q, including the MLYCD locus, was isodisomic.

Figure 2.

Schematic representation of the segmental maternal UPD16. The different markers used for segregation analysis are shown on the chromosome. The table gives the position for each marker according to the NCBI sequencing map (Mb) and Genethon (cM) map, and the genotypes for both parents and the index patient. WT, wild type allele.

Discussion

The diagnosis of malonic aciduria due to malonyl-CoA decarboxylase deficiency was suggested by clinical manifestations and by excretion of urinary organic acids, and confirmed by decreased malonyl-CoA decarboxylase activity.

To date 23 patients with malonic aciduria, including the one described here, have been reported. In 16 of them a diagnosis of malonyl-CoA decarboxylase deficiency was made by determination of enzyme activity and/or molecular investigation (Brown et al. 1984; Haan et al. 1986; MacPhee et al. 1993; Matalon et al. 1993; Krawinkel et al. 1994; Yano et al. 1997; Gao et al. 1999; Wightman et al. 2003; Ficicioglu et al. 2005; de Wit et al. 2006). In three patients (Henderson et al. 1998, Büyükgebiz et al. 1998), to our knowledge, neither enzymatic nor molecular diagnosis was performed.

Four patients with malonic aciduria and normal malonyl-CoA decarboxylase activity have been reported (Ozand et al. 1994; Gregg et al. 1998), thereby supporting the hypothesis that multiple etiologies could cause malonic aciduria. This evidence has also been confirmed in Maltese dogs with malonic aciduria without enzyme deficiency, reported by O'Brien et al. (1999).

Our patient present with classical symptoms; interestingly brain RMI showed altered white matter signals and disorder of neuronal migration. While white matter abnormalities have been reported in other patients with or without an enzyme defect (Ozand et al. 1994; Wightman et al. 2003; De Wit et al. 2006), only one patient with defects of neuronal migration has been described (De Wit et al. 2006). The pathogenic mechanism is still not completely understood, but some authors have hypothesised that there may be an altered interaction between malonyl-CoA and the brain-specific CPT-I enzyme (Price et al. 2002; De Wit et al. 2006), or between CPT-I and CPT-II, the latter reported to be associated with neuronal migration defect (Taroni et al. 1994; North et al. 1995; Elpeleg et al. 2001).

The flux of long-chain fatty acid oxidation by mitochondria is mainly determined by the activity of CPT-I, which is inhibited by malonyl-CoA (McGarry & Brown 1997; Haan et al. 1986). The increased malonyl-CoA in the patient likely decreased fatty acid oxidation and caused hypoglycemia and other symptoms. In the liver medium-chain fatty acids from medium-chain triglycerides (MCT) do not need carnitine and the related enzyme system for its entry. Maltese dogs with malonic aciduria improved on a high carbohydrate diet and MCT supplementation (O'Brien et al. 1999). MCT treatment was also successful in a patient with malonic aciduria: MCT, a high carbohydrate diet and carnitine, improved the clinical condition and resolved the cardiomyopathy (Ficicioglu et al. 2005).

In spite of data reported by Santer et al. (2003) in which malonylcarnitine from a patient's newborn blood spot was not detectable 4 years after birth, the malonylcarnitine level from our sample, stored for nine years at room temperature, was at the upper limit of the normal range. These data confirm that malonylcarnitine can be used for the detection of MCD deficiency in a retrospective diagnosis or in expanded newborn screening by tandem mass spectrometry.

Molecular analysis showed a homozygous frameshift mutation in exon 3 of the MLYCD gene, resulting in premature truncation of the protein. This genetic lesion causes complete loss of enzyme function, in agreement with the severity of the patient's phenotype.

To date 15 malonic aciduria patients with an identified genetic defect have been described, including the present patient (Wightman et al. 2003; Gao et al. 1999; FitzPatrick et al. 1999) (Fig. 3). Each family had its own unique mutation; ten patients were homozygous, and for five patients, including our own, the parents were not consanguineous. Since the mother of our patient was heterozygous for the mutation but the father was not, we confirmed paternity by microsatellite analysis (data not shown). As a result we tested the hypothesis of a maternal UPD. UPD is defined as the inheritance of both homologues of a pair of chromosomes from only one parent, and includes isodisomy (two copies of the same parental chromosome), heterodisomy (one copy of each homologue from the same parent), or a mixture of both. Depending on the affected chromosome and on the resulting regions of homozygosity, UPD can produce no clinical manifestations or a diversity of abnormal phenotypes; moreover, in isodisomy there is an increased risk of autosomal recessive disorders as a consequence of reduction to homozygosity (Engel 1998; Kotzot 2001).

Figure 3.

Schematic representation of MLYCD genomic organization. The new genetic lesion is boxed. The nucleotide numbers refer to GenBank Accession AF090834.

In our study maternal UPD of chromosome 16 resulted in homozygosity for the c.772–775delACTG MLYCD mutation, which caused malonic aciduria in our patient. The finding of heterozygosity for the short arm and the proximal long arm of chromosome 16 indicates that the double maternal contribution was the result of a non-disjunction event at the first meiotic division, likely favoured by advanced maternal age (42 years), with maintenance of euploidy in the zygote by elimination of the paternal contribution. The presence of two small homozygous regions in the long arm can be explained as the result of two crossing over events in otherwise different maternal chromosomes 16.

To date several UPD-16 fetuses and liveborns have been reported in the literature. They showed clinical abnormalities such as intrauterine growth retardation (IUGR), body stalk anomaly, imperforate anus, and congenital heart disease (Kalousek et al. 1993; Vaughan et al. 1994; O'Riordan et al. 1996; Abu-Amero et al. 1999; Chan et al. 2000) and were frequently found to be associated with trisomy 16-confined placental mosaicism (CPM) (Kalousek et al. 1993; Yong et al. 2003).

Our patient did not show IUGR or congenital anomalies, and his clinical manifestations were clearly related to malonic aciduria. IUGR is a common feature of trisomy 16 CPM, regardless of uni- or biparental origin of the chromosome 16 pair present in the fetus (Benn 1998). The presence of a trisomic line in the placenta is believed to cause placental insufficiency, and consequently IUGR, and the severity of the phenotype is likely related to the proportion of trisomic cells in the placenta. Whether mat UPD-16 contributes to fetal growth restriction in combination with trisomy 16 CPM is still a matter of debate (Benn 1998; Abu-Amero et al. 1999; Yong et al. 2002). Fetal mosaicism is also often observed, and the presence of an overt or occult trisomic line in the fetus/newborn is an additional major phenotypic determinant. In the case reported here, loss of chromosome 16 could have occurred at a very early stage following conception, followed by early loss of the trisomic line or predominant proliferation of disomic cells in the placenta. Although cytogenetic investigations were not performed on the present case, a normal result would not have ruled out the possibility of occult mosaicism.

Acknowledgement

The authors are grateful to the patient's family for their collaboration in this work.

Ancillary