3-Methylglutaconic Aciduria Type I Due to AUH Defect: The Case Report of a Diagnostic Odyssey and a Review of the Literature

3-Methylglutaconic aciduria type I (MGCA1) is an inborn error of the leucine degradation pathway caused by pathogenic variants in the AUH gene, which encodes 3-methylglutaconyl-coenzyme A hydratase (MGH). To date, MGCA1 has been diagnosed in 19 subjects and has been associated with a variable clinical picture, ranging from no symptoms to severe encephalopathy with basal ganglia involvement. We report the case of a 31-month-old female child referred to our center after the detection of increased 3-hydroxyisovalerylcarnitine levels at newborn screening, which were associated with increased urinary excretion of 3-methylglutaconic acid, 3-hydroxyisovaleric acid, and 3-methylglutaric acid. A next-generation sequencing (NGS) panel for 3-methylglutaconic aciduria failed to establish a definitive diagnosis. To further investigate the strong biochemical indication, we measured MGH activity, which was markedly decreased. Finally, single nucleotide polymorphism array analysis disclosed the presence of two microdeletions in compound heterozygosity encompassing the AUH gene, which confirmed the diagnosis. The patient was then supplemented with levocarnitine and protein intake was slowly decreased. At the last examination, the patient showed mild clumsiness and an expressive language disorder. This case exemplifies the importance of the biochemical phenotype in the differential diagnosis of metabolic diseases and the importance of collaboration between clinicians, biochemists, and geneticists for an accurate diagnosis.

Leucine catabolism consists of a six-step process that mainly takes place in the mitochondria ( Figure 1). The first two steps are shared with the other branched-chain amino acids, isoleucine and valine. In humans, any one of these steps (and its corresponding gene) can be impaired, causing a characteristic disease. The most-known diseases associated with the alteration of this catabolic pathway are Maple Syrup urine disease (MSUD), caused by a defect in the branched-chain α-keto acid dehydrogenase complex (BCKDH), and Isovaleric Acidemia (IVA), caused by a defect in the isovaleryl CoA dehydrogenase gene (IVD). Different phenotypes, encompassing severe neonatalonset forms with metabolic decompensation, are recognized, and their management and treatment are well established [4]. A more-benign disease of the leucine degradation pathway is also known, 3-Methylcrotonyl-CoA carboxylase deficiency, due to MCCC1 and MCCC2 gene defects. Phenotypes have been recognized thanks to the implementation of newborn screening that has allowed for the identification of asymptomatic newborn infants, siblings, and mothers [4]. Figure 1. Illustration of the leucine catabolic pathway. BCAT1, branched-chain amino acid aminotransferase 1; BCKDH, branched-chain α-keto acid dehydrogenase complex; IVD, isovaleryl-CoA dehydrogenase; MCCC, 3-methylcrotonyl-CoA carboxylase; 3-MGH, 3-methylglutaconyl-CoA hydratase; HMGCL, 3-hydroxy-3-methylglutaryl-CoA lyase; 3-HIVA, 3-hydroxyisovaleric acid; DBS, dried blood spot [4].
Here, we report the successful diagnostic odyssey of a neonate positive on newborn screening for 3-methylglutaconic aciduria who was diagnosed with MGCA1 through clinical-biochemical-genetic integration.

Case Report
The 31-month-old girl here reported was referred to our center after the detection of increased levels of 3-hydroxyisovalerylcarnitine (C5OH) at newborn screening, which were confirmed in subsequent samples (C5OH between 0.8 and 1.1 µmol/L; cut-off for newborn screening 0.59 at first test and 0.55 afterwards). Biotinidase activity was within the normal range, while urinary organic acid analysis revealed high excretion of 3-methylglutaconic acid associated with a moderate increase in 3-hydroxyvaleric acid and a slight increase in glutaric acid and 3-methylglutaric acid. Her mother's acylcarnitine profile and urinary organic acid analysis were normal, suggesting 3-methylglutaconic aciduria in the patient. The patient was the only child of a non-consanguineous healthy couple. During the pregnancy, her mother experienced symptomatic lumbar disc herniation requiring anti-inflammatory and steroidal therapy and a planned caesarean delivery at term (38 weeks + 2 days). The child's parameters at birth were normal (weight 2.9 kg, 11th percentile; length 49 cm, 29th percentile; head circumference 33 cm, 14th percentile). At the first admission to our center, she showed a normal general and neurological examination, and routine blood test results and heart ultrasound were normal.

Diagnostic Procedures
Due to the wide differential diagnosis in 3-methylglutaconic acidurias, the patient underwent a next-generation sequencing (NGS) panel containing 16 causative genes of 3-methylglutaconic acidurias. The procedure did not identify any genetic alterations, either in coding regions or in intron/exon boundaries, with the exception of a heterozygous pathogenetic variant in the ryanodine receptor 1 (RYR1) gene (MIM ID* 180901) (c.14126C > T; p.Thr4709Met [23]) inherited from her healthy mother. This variant was previously found in a heterozygous state in a child with congenital myopathy and the authors demonstrated that this particular pathogenic variant had a monoallelic expression in muscle when inherited from the father [24]. Our patient did not show signs of myopathy and her mother, who carried the same pathogenic variant, did not show signs of myopathy or 3-methylglutaconic aciduria.
To further investigate the strong biochemical indication, we measured MGH activity. The enzyme assay revealed markedly reduced MGH activity in the lymphocytes of our patient (<0.02 nmol/(min.mg prot), reference values 1.4-4.8 nmol/(min.mg prot)), confirming the biochemical diagnosis of MGCA1.
Single nucleotide polymorphism (SNP) array analysis was then performed in order to verify the presence of AUH intragenic microdeletions. This analysis disclosed two distinct microdeletions of 45 kb and 16 kb encompassing this gene; the former was covered by 43 markers, mapped to chr9:93936592_93981126 (hg19), and partially included the RefSeq gene LINC00484 and exons 8-10 of AUH (NM_001698.3), while the latter was covered by 76 markers, mapped to chr9:94106723_94122848 (hg19), and included exons 2 and 3 of the gene (Figure 2).
The same analysis also disclosed a maternally inherited microduplication at 1p36.33 that included the following genes: FNDC10, LOC105378586, ATAD3C, ATAD3B, SSU72, ATAD3A, and TMEM240, though its clinical significance is currently unknown. Copy number variation (CNV) analysis performed on patient and parental samples confirmed the presence of the two deletions in compound heterozygosity, with the 45 Kb deletion present in the maternal sample and the 16 kb deletion inherited from the father (Figure 3). The same analysis also disclosed a maternally inherited microduplication at 1p36.33 that included the following genes: FNDC10, LOC105378586, ATAD3C, ATAD3B, SSU72, ATAD3A, and TMEM240, though its clinical significance is currently unknown. Copy number variation (CNV) analysis performed on patient and parental samples confirmed the presence of the two deletions in compound heterozygosity, with the 45 Kb deletion present in the maternal sample and the 16 kb deletion inherited from the father (Figure 3).

Clinical Follow-Up
During the follow-up, C5OH tended to increase up to 3.33 µmol/L (r.v. 0.08-0.44), while organic acid excretion remained stable over time.
As soon as the enzyme activity assay result arrived (at 20 months old), the patient started levocarnitine at 80 mg/kg/day and her dietary protein intake was slowly decreased.
At the last examination (31 months of age), her statural and ponderal growth were normal (height 26th percentile, weight 74th percentile, head circumference 52nd percentile), and she showed mild clumsiness and an expressive language disorder.

Discussion
MGH deficiency was first hypothesized on a biochemical basis due to the excessive excretion of 3-methylglutaconic acid (3-MGA), 3-methylglutaric acid, and 3-hydroxyisovaleric acid in two siblings suffering from a language disorder [5]. Reduced MGH activity was then demonstrated by Narisawa and colleagues [6] in patient fibroblasts, which identified the last defect in the leucine catabolic pathway. Approximately 15 years later, the molecular cause of the disease was independently identified in the AUH (*600529) gene by two research groups [1,3]. Many of the gene variants described thus far in compound heterozygosity or homozygosity are predicted to lead to complete loss of protein function [17].
While 3-MGA is found only in traces in healthy individuals, an increased level of this branched-chain organic acid is a relatively common finding in patients suspected of having a metabolic disorder (about 3% in [25]), such as organic acidurias, glycogen storage disorders, fatty acid oxidation disorders, and urea cycle disorders, usually during metabolic decompensation and in association with more disease-specific urinary metabolites [25]. This increase is most likely due to mitochondrial dysfunction, and indeed patients with mitochondrial depletion syndromes and mitochondrial disorders in general (e.g., POLG, SUCLG1, SUCLA2, and TWINKLE) often show 3-methylglutaconic aciduria [26]. However, in some patients increased 3-MGA excretion is a major finding, being the hallmark of the disease and often paramount in the diagnosis. Therefore, the "inborn errors of metabolism with 3-methylglutaconic aciduria as a discriminative feature" label has been proposed for this group of conditions, which consists of two categories: primary 3-methylglutaconic acidurias and secondary 3-methylglutaconic acidurias [25]. Primary 3-methylglutaconic acidurias are disorders of mitochondrial HMG-CoA metabolism and are caused by MGH (AUH) deficiencies, though some authors have also suggested a role of HMG-CoA lyase (HMGCL) defects, an enzyme common to ketogenesis and leucine oxidation pathways [27][28][29]. In the urine of patients with primary 3-methylglutaconic acidurias, there is also excess 3-hydroxyisovaleric acid derived from unmetabolized 3-hydroxyisovaleryl-coenzyme A ester, a leucine oxidation intermediate from just upstream of 3-MG-CoA, and subsequent HMG-CoA formations.
The natural history of MGCA1 is not known and patients diagnosed with this disease may have variable clinical symptoms ( Table 2). The percentage of consanguinity in the parents of affected patients is quite high ( Table 1) and some authors question whether the severe clinical picture of some patients is related to the MGH defect or to another, still undiscovered, genetic defect.
Tandem mass spectrometry now allows newborn screening for more than 40 metabolic disorders simultaneously, but evidence of its effectiveness for some diseases has yet to be established due to the low prevalence and lack of a clinical phenotype. For these reasons, diseases amenable to newborn screening are often listed in primary/core and secondary panels according to whether evidence about the effectiveness of newborn screening and treatment on disease outcome is high or low, respectively [33].
The first patients found to have MGCA1 were diagnosed before the introduction of newborn screening and had a variable phenotype. The introduction of newborn screening allowed for the detection of MGCA1 based on a biochemical phenotype in asymptomatic newborns. MGCA1 belongs to the secondary panel of newborn screening programs due to the low number of patients described and the uncertainty about the clinical phenotype. However, it shares primary metabolic markers with other diseases belonging to the primary panel (e.g., ß-ketothiolase deficiency), which allows for its early diagnosis but leads to questions about follow-up and dietetic treatment indications.
The patient described in this paper is a healthy child that was put on levocarnitine supplementation and started a slow dietary restriction of protein intake after the MGCA1 diagnosis was made as a preventive measure for possible later alterations. At the last examination, when she was 31 months old, she showed mild clumsiness and an expressive language disorder, which is a quite common finding at this age, also in "metabolically healthy" children. Further follow-up is needed to assess the outcome of this condition in this patient diagnosed by means of newborn screening.
This case exemplifies the importance of the biochemical phenotype in guiding the differential diagnosis, at least in metabolic diseases with a biochemical marker. In this case, only the analysis of the biochemical alteration pattern allowed for further investigation after a negative NGS panel. Fortunately, the enzyme activity assay confirmed the suspicion, and SNP array and CNV analyses using real-time polymerase chain reaction (PCR) assays subsequently detected two different deletions in compound heterozygosity, a rare finding in an ultrarare disease, ultimately confirming MGCA1. Thus far, only one intragenic deletion involving exons 1-3 has been reported in a homozygous state in a patient affected by MGCA1 [8], and only two cases involving exon 8 and exons 2-4 are reported in international databases (https://www.ncbi.nlm.nih.gov/clinvar/; accessed on 24 February 2022). The confirmed diagnosis in our patient also highlights the importance of collaboration between clinicians, biochemists, and geneticists.
Finally, this case also describes the challenges of managing patients and parents of children who are diagnosed early as a result of newborn screening. In this case, the challenges included the course of action and communication with the patient's parents and the decision to start dietetic treatment, which needed to be balanced according to the risk of developing symptoms and the burden of dietetic treatment in the absence of sufficient evidence to guide this decision.

Case Report
This paper was written in accordance with the CAse REport (CARE) guidelines (https://www.care-statement.org, accessed on 10 April 2022).

NGS
Genomic DNA of the patient and her parents was extracted from peripheral blood using a QIAsymphony instrument (Qiagen, Hilden, Germany). We performed targeted NGS resequencing of 16 genes associated with 3-methylglutaconic aciduria using a customdesigned panel (Illumina, San Diego, CA, USA). We prepared libraries using the Nextera Rapid Capture Enrichment kit (Illumina, San Diego, CA, USA) according to the manual instructions. Libraries were sequenced by a paired-end 2 × 150 bp protocol on a MiSeq System (Illumina, San Diego, CA, USA) to obtain an average coverage of above 100x, with >95% of target bases covered at least 15x. For data analysis, we used the BWA, Picard, and GATK tools. Variant annotation was performed by the ANNOVAR tool. The reference sequences used for the nomenclature of exons and variants in the AUH and RYR1 genes were NM_001698.3 and NM_000540.3, respectively.

SNP Array Analysis
CNV analysis was carried out using an SNP-array platform (Cytoscan HD, Thermo Fisher Scientific, Inc., Waltham, MA, USA) following the manufacturer's recommendations. Briefly, 250 ng of genomic DNA extracted from lymphocytes (NucleoSpin Blood, Macherey-Nagel, Duren, Germany) was digested with NspI. After digestion, an adaptor was linked to the restricted fragments that were then amplified by PCR. After purification using magnetic beads, PCR products were fragmented and end-labeled using a terminal deoxynucleotidyl transferase, and then hybridized for 16-18 h to the Cytoscan HD chip at 50 • C in a GeneChip Hybridization Oven 640 (Thermo Fisher Scientific, Inc., Waltham, MA, USA). The chips were washed, stained in a GeneChip Fluidics Station 450, and scanned with a GeneChip 3000 7G scanner. Data were analyzed with ChAS software (v4.2.1; Thermo Fisher Scientific, Inc., Waltham, MA, USA). Two hundred and seventy healthy controls belonging to the International HapMap Project were used as a reference sample in data analysis (Thermo Fisher Scientific, Inc., Waltham, MA, USA). Genome-wide microdeletion/microduplication (CNV) analysis was performed only calling losses and gains covered by at least 25 markers and with a minimum length of 75 Kb, and CNVs were classified according to the American College of Medical Genetics and Genomics (ACMG) and ClinGen [34]. CNVs within the candidate gene were analyzed without a filter.
TaqMan ® Copy Number Assays quantify AUH exons normalized on the endogenous reference RNaseP gene. Normalizations were performed by relative quantifications determined by the ∆∆Ct ((FAM Ct-VIC Ct) sample-(FAM Ct-VIC Ct) calibrator) method as previously reported [36]. The gene copy number is two times the relative quantity [36].