54 Original Article

Mild Persistent Isolated Hypermethioninemia Identified through Newborn Screening in Michigan

Kuntal Sen1,2 Michael D. Felice2 Allison Bannick1 Roberto Colombo3 Robert L. Conway1,2

1 Division of Genetic, Genomic and Metabolic Disorders, Children’s Address for correspondence Kuntal Sen, MD, Division of Genetic, Hospital of Michigan, Detroit, Michigan, United States Genomic and Metabolic Disorders, Children’s Hospital of Michigan/ 2 Wayne State University School of Medicine, Detroit, Michigan, Detroit, MI, United States (e-mail: [email protected]). United States 3 Center for the Study of Rare Inherited Diseases (CeSMER), Niguarda Ca’ Granda Metropolitan Hospital, Milan, Italy

J Pediatr Genet 2019;8:54–57.

Abstract Methionine S-adenosyltransferase deficiency, due to mutations in MAT1A,isthemost common cause of persistent isolated hypermethioninemia (PIH). While the recessive Keywords form may cause neurological consequences, the dominant form is typically benign. ► newborn screening This condition may be found in asymptomatic infants through newborn screening ► persistent isolated programs. We describe 16 asymptomatic individuals with PIH. Our data reiterates the hypermethioninemia benign nature of PIH and reports two novel mutations in the gene. There were a ► MAT1A mutation disproportionate number of individuals with African descent in this cohort.

Introduction either homocystinuria (HCU) or MAT deficiency. In addition In 1951, Cantoni described the function of methionine to these two conditions, there are diagnoses such as tyrosi- adenosyltransferase (MAT). This enzyme catalyzes the trans- nemia, citrin deficiency, S-adenosylhomocysteine hydrolase fer of the adenosyl moiety of ATP (adenosine triphosphate) to deficiency, glycine N-methyltransferase deficiency, as well as methionine which forms S-adenosylmethionine (AdoMet).1 liver disease which can also cause hypermethioninemia. We AdoMet is the principal molecule involved in methyl group present the follow-up data of 13 infants identified through transfer in several biological reactions. The gene that codes newborn screening due to elevated methionine and describe for the two isoforms of the enzyme MAT I/III is MAT1A.The the clinical, molecular, and biochemical findings of this

phenotypic spectrum of MAT I/III deficiency is broad ranging cohort. In addition, similar results from two children and Downloaded by: Boston College. Copyrighted material. from a demyelinating disorder of the central nervous system one adult from Italy are also described. and variable neurologic sequelae to persistent isolated hypermethioninemia (PIH) in otherwise asymptomatic indi- Methodology viduals.2 MATIA mutations can be both recessive and domi- nant, with the latter probably being a benign disorder The present study was performed after obtaining institutional causing mild to modest methionine elevations and without review board approval. To characterize the clinical data of risk of neurologic symptoms compared with the recessive patients identified with PIH, we performed a retrospective type, which is associated with severe methionine elevations chart review of abnormal NBS results for increased methionine often > 800 to 1000 µmol/L and variable risk of neurologic identified through the Michigan NBS program over a 10-year symptoms.3 A specific heterozygous MAT1A mutation, period (2005–2015). When reviewing methionine values from c.791G>A (p.Arg264His), has been reported as being parti- our NBS laboratory, it is important to note that the cut-off value cularly associated with the dominant form of PIH. An ele- for abnormal results changed in 2011 from 74 to 56 µmol/L due vated methionine level on newborn screening (NBS) dried to a change in testing method. Infants with elevated methio- blood spot (DBS) analysis typically raises clinical concern for nine levels were referred to our institution’s metabolic clinic

received Copyright © 2019 by Georg Thieme DOI https://doi.org/ December 28, 2018 Verlag KG, Stuttgart · New York 10.1055/s-0039-1683900. accepted ISSN 2146-4596. February 8, 2019 published online March 27, 2019 Mild Persistent Isolated Hypermethioninemia NBS Michigan Sen et al. 55 for further testing and management. Confirmatory testing of ing MAT activity in liver extract. However, this test is not these infants routinely included quantitative plasma amino widely available in the clinical setting. Clinically significant acids, and plasma total homocysteine. S-adenosylmethionine hyperhomocysteinemia was not observed in our cohort (SAM) and S-adenosylhomocysteine (SAH) testing, as well as (highest measured level was 11 µmol/L). parental biochemical testing, was not systematically per- All patients were seen in our clinic for an initial consulta- formed. This is a limitation of this report, as that information tion, and the nature of mildly elevated methionine levels was could help with interpretation of the infant’s elevated methio- discussed with them. No dietary restrictions or use of nine level. A designation of PIH was given to infants who had at AdoMet supplementation was recommended. It was recom- least three elevated methionine levels between the newborn mended that these infants be followed by a health care screen and plasma amino acids. Our center serves as the provider for growth and development surveillance and to metabolic referral center for infants from anywhere in the have periodic check of plasma methionine levels. Families state of Michigan. Therefore, some of the laboratory reference were given the option to continue follow-up with our clinic ranges are different, as the laboratories where confirmatory for this, or to have their primary care provider (PCP) to plasma amino acids were performed could vary. Information complete the monitoring and report to us any health con- regarding demographics, clinical parameters, biochemical, cerns or rising of methionine levels. All families chose to and molecular test results were collected by reviewing avail- follow-up with their PCP. All of our patients were followed up able medical charts. Similar data of two siblings (patients 15 in our clinic for an average period of 5.3 months, and a range and 16) and their mother (patient 14) were contributed from of 1 to 15.2 months. None of them demonstrated any other Italy. neurologic signs reported in the recessive form of MAT I/III deficiency, supporting the observations made by other MAT1A Results authors that PIH due to heterozygosity for mutations represents a benign condition. We have not received any Of the 51 Michigan infants who had an elevated methionine communication from a PCP with a health concern. Time of on the NBS specimen, three cases of HCU and 13 cases of PIH direct follow-up of our cohort is limited, and a follow-up were identified; plasma methionine levels normalized on survey to PCP offices to assess their current growth and confirmatory plasma amino acid testing in the remainder. development status is a strong consideration to add to Among the PIH cohort (►Table 1), the average methionine longitudinal data about individuals with PIH. level on the NBS was 83.1 µmol/L (range, 60–121 µmol/L; Interestingly, 10 of the 13 patients (77%) from Michigan reference < 56 µmol/L prior to 2011 and < 74 in 2011 and were African American, although the African American after). On confirmatory plasma amino acid analysis methio- demographic only represented 13.4% of births in Michigan nine averaged 118.5 µmol/L (range, 76–419 µmol/L; refer- within this same time period.6 All patients for whom we had ence ranges varied but all were 65 µmol/L or less). molecular results were of African American descent. This Methionine levels averaged 104 µmol/L on subsequent fol- suggests the possibility that the prevalence of heterozygous low-up (range, 67–173 µmol/L; reference ranges varied but MAT1A mutations causing PIH may be more frequent among all were 65 µmol/L or less). All patients with multiple this population. measurements of plasma amino acids exhibited a plasma The data from the patients in Italy included two siblings methionine trend characterized by an initial postnatal who had PIH (►Table 1, patients 15 and 16). These two Downloaded by: Boston College. Copyrighted material. increase and subsequent decrease, a phenomenon pre- brothers were of African origin and were 7 and 5 years at the viously reported in another study.4 time of testing. Their plasma methionine levels at diagnosis In all six patients who received molecular testing, a were 68 and 81 µmol/L (reference, 10–50 µmol/L), while their heterozygous mutation in MAT1A was identified. The subsequent methionine levels ranged from 58 to 144 and 75 remainder did not receive genetic testing because they to 159 µmol/L, respectively. Identical to our cohort, neither of were either lost to follow-up or because the insurance denied them had any developmental delays or other neurological coverage of the test, which is a limitation of our report. issues. Their mother, 27 years old and also asymptomatic, However, individuals with autosomal recessive MAT defi- was subsequently ascertained and found to have PIH with a ciency typically have methionine levels elevated much methionine level of 74 µmol/L. They all carried the above- higher (> 200 µmol/L) than those of our cohort.5 Of note, mentioned c.763C>T(p.Pro255Ser)MAT1A variant. one patient did not receive genetic testing though her half fi sibling was identi ed with a mutation, and they had similar Discussion elevations of methionine, implying identical variant geno- types. Three MAT1A sequence alterations were identified Similar to data published from other NBS programs, the among our cohort: c.259C>T (p.Arg177Trp) (a previously Michigan program identifies mild PIH far more frequently reported mutation, n ¼ 3), along with two novel variants, than HCU or MAT deficiency after an increased methionine c.763C>T(p.Pro255Ser)(n ¼ 1) and c.839G>T (p.Gly280Val) on the newborn screening test.7 Our study found that mild (n ¼ 2). The classification of these new variants as patho- PIH identified through NBS in absence of hyperhomocystei- genic is based on biochemical studies showing mild eleva- nemia is most frequently caused by heterozygous sequence tions in methionine. The most definitive functional study to alterations in MAT1A. In the past 15 years, there have been establish the pathogenicity of the variants would be measur- more reports of individuals with dominant PIH compared

Journal of Pediatric Genetics Vol. 8 No. 2/2019 56 Mild Persistent Isolated Hypermethioninemia NBS Michigan Sen et al.

Table 1 Clinical, biochemical, and molecular characterization of 16 patients with PIH

Patient Ethnicity Sex NBS results: Confirmatory Follow-up Molecular Length of follow-up methionine laboratories: laboratories: results with no clinical µmol/L methionine Methionine concerns (days) µmol/L µmol/L 1 Irish, German F 111 97 72–152 Not done 202 (nl < 74) (10–60) 2a African American F 84 90 42–131 Not done 383 (nl < 74) (10–60) 3b African American, F74 96 96–134 Not done 29 Mexican (nl < 74) (10–60) 4b African American, F89 68 68–163 MAT1A: 107 Mexican (nl < 74) (10–60) p.Gly280Val heterozygous 5a African American F 103 182 92–182 Not done 457 (nl < 74) (10–60) 6German,French,F70 66 66–94 Not done 37 Scottish (nl < 56) (< 56) 7 African American M 64 70 67–109 MAT1A: 146 (nl < 56) (10–60) p.Arg177Trp heterozygous 8 Italian,Irish,Polish M 71 78 78–100 Not done 44 (nl < 56) (10–60) 9 African American M 71 124 87–124 MAT1A: 114 (nl < 56) (10–60) p.Gly280Val heterozygous 10 African American F 63 61 61–145 MAT1A: 327 (nl < 56) (11–35) p.Arg177Trp heterozygous 11 African American M 69 98 67–98 Not done 197 (nl < 56) (10–60) 12 African American M 113 74 74–96 MAT1A: 45 (nl < 56) (< 65) p.Arg177Trp heterozygous 13 African American, F60 69 69–122 MAT1A 128 Indian (nl < 56) (10–60) p.Pro255Ser heterozygous Downloaded by: Boston College. Copyrighted material. 14c African F N/A 74 64–117 MAT1A 27 years (10–50) p.Pro255Ser heterozygous 15c African M N/A 68 58–144 MAT1A 7y (10–50) p.Pro255Ser heterozygous 16c African M N/A 81 75–159 MAT1A 5y (10–50) p.Pro255Ser heterozygous

Abbreviations: F, female; M, male; NBS, newborn screening. aFull siblings. bHalf siblings. cPatients from Italy.

– with recessive PIH.4,5,8 10 Literature review showed nearly Pro255Ser) and c.839G>T (p.Gly280Val) identified in our 40 individuals who all carried the heterozygous MAT1A cohort expands the genotypic spectrum of heterozygous mutations leading due to benign PIH without any clinical alterations in this gene associated with mild PIH. One of repercussions.4,9 Most of these individuals have been these novel variants, c.763C>T (p.Pro255Ser), was also found reported to have a heterozygous c.791G>A(p.Arg264His) in African siblings and mother from Italy. The c.259C>T mutation in MAT1A, which was not present in our patients. (p.Arg177Trp) mutation which was present in three of our The two novel MAT1A sequence changes c.763C>T(p. patients, has been suspected as a pathogenic variant in two

Journal of Pediatric Genetics Vol. 8 No. 2/2019 Mild Persistent Isolated Hypermethioninemia NBS Michigan Sen et al. 57 patients in another case series, though that report does not References precisely record the inheritance pattern or the clinical sig- 1 Cantoni GL. Activation of methionine for transmethylation. J Biol nificance of this change within their cohort.10 Although we Chem 1951;189(02):745–754 acknowledge that our genotyping data are limited due to the 2 Chamberlin ME, Ubagai T,MuddSH, WilsonWG, Leonard JV, ChouJY. Demyelination of the brain is associated with methionine adenosyl- fact that not all patients in our cohort received molecular transferase I/III deficiency. J Clin Invest 1996;98(04):1021–1027 MAT1A testing, we observe that mild PIH due to heterozyg- 3 Couce ML, Bóveda MD, Castiñeiras DE, et al. Hypermethioninae- osity appears to be more common among individuals of mia due to methionine adenosyltransferase I/III (MAT I/III) defi- African descent in our state’s newborn screening of popula- ciency: diagnosis in an expanded neonatal screening programme. – tion. The findings of this publication represent the outcomes J Inherit Metab Dis 2008;31(Suppl 2):S233 S239 4 Nagao M, Tanaka T, Furujo M. Spectrum of mutations associated of a public health program and not a research initiative. with methionine adenosyltransferase I/III deficiency among indi- viduals identified during newborn screening in Japan. Mol Genet Conclusion Metab 2013;110(04):460–464 5 Chien YH, Abdenur JE, Baronio F, et al. Mudd’s disease (MAT I/III Our data support the observation that mild PIH associated deficiency): a survey of data for MAT1A homozygotes and com- with MAT1A heterozygosity appears to be a benign disorder pound heterozygotes. Orphanet J Rare Dis 2015;10:99 6 United State Census Bureau. Available at: https://www.census. that does not require treatment. MAT1 heterozygosity, along gov/quickfacts/MI. Accessed November 26, 2018 with other serious disorders that can cause hypermethioni- 7 Iacobazzi V, Infantino V, Castegna A, Andria G. Hyperhomocys- nemia can be found through NBS. Therefore, confirmatory teinemia: related genetic diseases and congenital defects, abnor- testing including molecular analysis should be completed for mal DNA methylation and newborn screening issues. Mol Genet all such cases. If a case is proven to be a result of MAT1A Metab 2014;113(01):(02):27–33 heterozygosity, routine follow-up might be unnecessary and 8 Chien YH, Chiang SC, Huang A, Hwu WL. Spectrum of hyper- methioninemia in neonatal screening. Early Hum Dev 2005;81 can be considered incidental finding. We are also reporting (06):529–533 MAT1A two new variants in gene and postulate that this 9 Couce ML, Bóveda MD, García-Jimémez C, et al. Clinical and biochemical phenotype may be more common among indi- metabolic findings in patients with methionine adenosyltrans- viduals from certain African populations. ferase I/III deficiency detected by newborn screening. Mol Genet Metab 2013;110(03):218–221 10 Chadwick S, Fitzgerald K, Weiss B, Ficicioglu C. Thirteen patients Conflict of Interest with MAT1A mutations detected through newborn screening: 13 None declared. years’ experience. JIMD Rep 2014;14:71–76 Downloaded by: Boston College. Copyrighted material.

Journal of Pediatric Genetics Vol. 8 No. 2/2019