Missense mutations in the human glutathione synthetase gene result in severe metabolic acidosis, 5-oxoprolinuria, hemolytic anemia and neurological dysfunction
Missense mutations in the human glutathione synthetase gene result in severe metabolic acidosis, 5-oxoprolinuria, hemolytic anemia and neurological dysfunctionNiklas Dahl1,*, Maritta Pigg1, Ellinor Ristoff2, Rayappa Gali3, Birgit Carlsson1, Bengt Mannervik4, Agne Larsson2 and Philip Board3
1Department of Clinical Genetics, Uppsala University Children's Hospital, S-751 85 Uppsala, Sweden, 2Department of Pediatrics, Karolinska Institute, Huddinge University Hospital, S-141 86 Huddinge, Sweden, 3Molecular Genetics Group, John Curtin School of Medical Research, Australian National University, Canberra, ACT 2601, Australia and 4Department of Biochemistry, Uppsala University Biomedical Centre, S-751 23 Uppsala, Sweden
Received February 19, 1997;Revised and Accepted April 8, 1997
Severe glutathione synthetase (GS) deficiency is a rare genetic disorder with neonatal onset. The enzymatic block of the [gamma]-glutamyl cycle leads to a generalized glutathione deficiency. Clinically affected patients present with severe metabolic acidosis, 5-oxoprolinuria, increased rate of hemolysis and defective function of the central nervous system. The disorder is inherited in an autosomal recessive mode and, until recently, the molecular basis has remained unknown. We have sequenced 18 GS alleles associated with enzyme deficiency and we detected missense mutations by direct sequencing of cDNAs and genomic DNA. In total, 13 different mutations were identified. Four patients were found to be compound heterozygotes and two individuals were apparently homozygous. Reduced enzymatic activities were demonstrated in recombinant protein expressed from cDNAs in four cases with different missense mutations. The results from biochemical analysis of patient specimens, supported by the properties of the expressed mutant proteins, indicate that a residual activity is present in affected individuals. Our results suggest that complete loss of function of both GS alleles is probably lethal. It is postulated that missense mutations will account for the phenotype in the majority of patients with severe GS deficiency.
Glutathione synthetase (EC 6.3.2.3) (GS) catalyzes the last biosynthetic step in the [gamma]-glutamyl cycle and therefore plays a central role in the synthesis of glutathione. Glutathione is present in virtually all mammalian tissues and is important for a variety of biological functions. For instance, glutathione provides reducing capacity and protects cells against reactive electrophiles. In addition, it has been proposed that glutathione plays a role in the membrane transport of different amino acids (1 ,2 ).
Glutathione is synthesized in two consecutive steps catalyzed by [gamma]-glutamylcysteine synthetase and glutathione synthetase (Fig. 1 ). The second step requires [gamma]-glutamylcysteine and glycine in the presence of ATP. Glutathione acts as a feedback inhibitor of the initial step in its biosynthesis (3 ). In patients with hereditary deficiency of glutathione synthetase the lack of glutathione, i.e. the feed-back inhibitor, leads to the formation of increased amounts of [gamma]-glutamylcysteine which is converted into 5-oxoproline by [gamma]-glutamyl cyclotransferase and excreted in massive amounts by the patients. It is notable that many of the patients with defects of the [gamma]-glutamyl cycle are mentally retarded and exhibit other central nervous system defects (4 -6 ). This indicates that glutathione is important in the normal function of the central nervous system. Other clinical features include hemolytic anemia and neonatal jaundice, which are thought to reflect the more rapid decline in the residual enzyme activity during the maturation of erythrocytes as compared to the situation in nucleated cells. A milder form of glutathione synthetase deficiency apparently restricted to erythrocytes, is associated with decreased erythrocyte glutathione levels and hemolytic disease, which is usually well compensated (7 -11 ).
The GS patients studied here present a wide range of the clinical features observed in severe GS deficiency (Table 1 ). The probands are unrelated and of different geographical origins. No absolute correlation was observed between clinical course, neurological manifestations and decrease in enzymatic activity of GS.
Clinical features, enzyme activity and mutations in GS deficient patients
aAll patients present with metabolic acidosis and hemolytic anemia.bGS activity was determined by duplicate analysis of cell free extracts of cultured skin fibroblasts from patients and eight healthy control individuals. the control range (mean +- 2 SD) was found to be 37.6 +-14.8 pkatal/mg protein. the coefficient of variation was 11% intra-assay and 19% inter-assay.cMutations present in a homozygous state.
We obtained GS cDNA from nine individuals with lowered enzyme activity. Direct sequencing and cycle sequencing of GS cDNA was achieved by the use of five sets of oligomers covering the entire coding sequence (Fig. 2 ). Comparison of these GS derived sequences with the previously published sequence of normal GS revealed unique base changes (Table 1 ). All mutations were confirmed on genomic DNA by direct sequencing of PCR products corresponding to exons 1, 5, 6, 7, 8, 9 and 12, respectively.
Thirteen different point mutations were identified in alleles derived from nine patients. The resulting amino acid substitutions were predicted, and in all cases we can infer these to be the likely cause of GS deficiency as they are the only substitutions found in the entire protein coding sequence when compared to the previously published cDNA sequence. Two probands were found to be homozygous for a mutation (cases 3 and 29). In four cases more than one mutation was identified which indicates that they are compound heterozygotes (cases 8, 19, 21 and 22). In three individuals only one missense mutation was found in a heterozygous state (cases 17, 28 and 36). The phase of the mutations could not be ascertained as no samples from the parents were available. Patient 17 was found to be heterozygous for a transition resulting in a Tyr270Cys substitution. The same codon was altered in a heterozygous state in patient 28 resulting in a Tyr270His substitution. Similarly, two different mutations were identified in codon 219 (cases 8 and 19) resulting in either Asp219Gly or Asp219Ala substitutions, respectively. In patient 8 three predicted amino acid substitutions were identified on both cDNA and genomic DNA.
Four mutations were visualized on genomic DNA by restriction enzyme digest of PCR products (Fig. 3 ). The mutation in codon 26 destroys a BanII recognition site, whereas the transition in codon 188 and the two different mutations affecting codon 270 creates recognition sites for MspI, SphI and HaeIII, respectively.
Southern blot analyses of genomic DNA from all nine patients and with the GS cDNA as a probe revealed hybridization patterns that were identical to that of control individuals. Two polymorphisms were detected. An additional 8 kb EcoRI band was detected in 17% of the alleles with the 1/4 probe. Another, rarer, polymorphic band of 6 kb was detected in 5% of alleles with the 3/8 probe after HindIII digest.
Comparison of the human amino acid sequence to GS sequences in Xenopus and rat revealed that 12 out of 13 predicted amino acid substitutions affect positions that are perfectly conserved in the three species (Fig. 4 ).
Identification of mutations that cause the clinical spectrum of symptoms recognised in patients with GS deficiency is of interest not only for the understanding of GS deficiency per se but also for the information that it may provide concerning the function of GS and glutathione in normal individuals. The biochemistry of GS and the clinical manifestations of its deficiency have been studied previously (1 -3 ,6 ), yet little is known about the human enzyme at the molecular level. The primary structure of human GS was predicted through the isolation of the full-length cDNA (24 ) and recently the first four patients were characterised at the molecular level (27 ). We have now identified 13 different point mutations in this coding sequence which cause lowered enzyme activity in severely affected individuals. The two base substitutions 799C -> T and 847C -> T are derived from the same patient who was presented recently (27 ) and, in the same report, the mutation 656A -> G is described in an independent individual.
We demonstrate that different missense mutations are associated with severe clinical phenotypes (Table 1 ) and the nucleotide substitutions provide direct proof of the genetic heterogeneity of GS deficiency. In addition, our data provide an insight into which amino acids are important for the function of GS either by affecting the specific activity or the stability of the protein.
All patients in our study presented with metabolic acidosis, hemolytic anemia and 5-oxoprolinuria. However, involvement of neurological symptoms was heterogeneous. Among the 13 different missense mutations identified, two were found in patients presenting with functional impairment of the central nervous system. Whether the phenotypic variation is derived from different clinical management of the patients, from specific alterations of GS, or both, is not possible to evaluate. The high degree of amino acid conservation of GS between human and rat (88%) and between human and Xenopus (64%) is evenly distributed along the protein. This indicates an important function of the gene product in mammalian cells throughout evolution, and the conservation suggests that residues along the entire polypeptide chain may be equally important in maintaining its structure and catalytic function. All but one of the amino acid substitutions (12 out of 13) affect codons that are conserved in Xenopus (23 ,24 ) and a close examination of highly conserved regions found in the frog, rat and human enzymes may provide additional support for residues involved in function or maintenance of structure. Ten out of 13 different amino acid substitutions associated with GS deficiency are located between residues 188 and 330 (Fig. 2 ). Among the substitutions identified there are three exceptions to a location in the central part of the cDNA. The three mutations involve codons 26, 464 and 469 indicating that certain amino acids at both ends of the protein sequence are critical for either the catalytic activity or for the stability and structure of the enzyme. Notably, the last 37 codons (437-474) are completely conserved between human and rat. The 10 residue segment between amino acids 280 and 289 is also perfectly conserved between the three species. Two different substitutions within this segment (residues 283 and 286) were identified on GS deficient alleles. The Arg283Cys mutation was expressed in vitro and the protein showed a 10-fold reduction in enzymatic activity. Interestingly, two different mutations were identified in each of the two codons 219 and 270. The mutation that results in Asp219Gly was recently found independently in a patient with GS deficiency (27 ). The importance of the codon 270 mutations was substantiated by the heterologous expression of the two different mutant proteins Tyr270Cys and Tyr270His, both demonstrating a 100-fold reduction of GS activity. The 13 mutations presented here add substantial information about how specific amino acid replacements cause a reduction of GS activity. However, a definite assignment of specific functions to different amino acids within the protein awaits elucidation of the three-dimensional structure of human GS.
In three patients (cases 17, 28 and 36) only one mutated allele was found at the cDNA level. One possibility is that the undiscovered mutations are located outside the protein coding region. Such a mutation may influence the level of transcription or the stability of GS mRNA. Further analyses at the genomic level may clarify the molecular basis for GS deficiency in these three alleles.
The enzymatic activity varied markedly in our series of patients. No correlation was found between the level of enzymatic activity and presence of central nervous symptoms. Surprisingly, one patient exhibited severe symptoms with an enzymatic activity of up to 31% of normal.
We have not found evidence for complete loss of GS activity in the patients studied here. This is supported by the enzymatic activity measured from the expressed mutant cDNAs and in specimens from the patients which show the presence of measurable activity. Interestingly, GS deficiency has been observed in patients without 5-oxoprolinuria (7 -11 ). The phenotype of these patients is compensated hemolytic anemia without neurologic manifestations and the deficiency appears to be restricted to the erythrocytes. It has been speculated that the genetic lesion in these patients leads to the synthesis of an unstable GS molecule. The replacement of a degraded enzyme is not possible in mature erythrocytes as they lose their capacity for protein synthesis when they are released into the circulation. The analysis of such patients is underway and may uncover the biochemical basis of the milder expression of the disorder.
Taken together, the findings reported begin to define the pattern of mutations in a housekeeping gene, in which complete loss of function would probably be lethal.
Nine non-related index cases were studied (Table 1 ). Biochemical findings included severe metabolic acidosis and 5-oxoprolinuria (pyroglutamic aciduria). The patients presented with hemolytic anemia in the neonatal period. Two patients died within 3 weeks of birth from infections (case 19) or severe metabolic acidosis (case 22). Neurological symptoms were apparent in two patients out of those who survived the neonatal period.
GS activity was determined as described by Wellner et al. (5 ). Decreased GS activity was demonstrated in extracts of erythrocytes and leukocytes as well as fibroblasts derived from skin biopsies. This study was approved by the ethical committees of the Karolinska Institute and Uppsala University.
Total RNA from patients and normal control subjects was extracted from cultured fibroblasts, either immediately after harvest or after the cells had been stored at -70oC or in liquid nitrogen. Cells were resuspended in acid guanidinium thiocyanate according to Chomczynski and Sacchi (28 ). For reverse transcription (RT), 5 [mu]g of total RNA from patients and controls was annealed to 5 pmol of primer 4 or 8 specific for GS RNA (Table 2 ). The RT reaction was performed in 20 [mu]l containing 10 mM Tris pH 8.3, 5 mM MgCl2, 1 mM of each dNTP and 2.5 U of RT polymerase (Pharmacia-Biotech). Aliquots (5 [mu]l) of the RT reaction mixture were subjected to 30 cycles in a total volume of 100 [mu]l containing all four dNTPs (each at 200 [mu]M), 1.5 mM MgCl2, 2 U of Taq polymerase and oligonucleotides at 0.1 [mu]M. Each cycle consisted of 94oC for 1 min, 60oC for 1 min and 72oC for 2 min. The RT-PCR products were generated with primer pairs 3/8 and 1/4 and followed by 25 cycles of nested PCR with the four primer pairs 1/B3, 9/B1, B2/10 and B4/8 (Fig. 2 ). The four primers designated `B' were biotinylated at the 5' end. In order to verify the sequence between codons 182 and 358 at the cDNA level, an independent RT-PCR product was generated from each of the nine cell lines using the cDNA primer 8 followed by 30 cycles of amplification with primers 12 or 20 under the same conditions as above. Amplicons from genomic DNA corresponding to exons 1, 5, 6, 7, 8, 9 and 12, respectively, were generated as described previously (27 ).
Primers specific for the GS gene used in this studya
aPrimer pairs used for the amplification of exons 1, , 6, 7, 8, 9 and 12 at the genomic level are described elsewhere (27).bThe nucleotides are numbered with respect to the first nucleotide in the start codon.cMutations present in a homozygous state.
The biotinylated amplicons were bound to streptavidin-coated magnetic beads (Dynal, Oslo, Norway), washed and denatured (29 ). Both strands were sequenced using a fluorescent GS specific primer on an ALF Express automated DNA sequencer (Pharmacia-Biotech). The non-biotinylated RT-PCR fragments and amplicons from genomic DNA were sequenced using the cycle sequencing core kit (Applied Biosytem). The products from the cycle sequencing reaction were separated and detected on an Applied Biosystem automated sequencer (model 373A). The resulting sequences were aligned for a comparative study with the human GS cDNA sequence (DDBJ/EMBL/GenBank accession no. L42531). Patients with mutations in codon 26, 188 or 270 were also analyzed at the genomic level by restriction enzyme digestion. Amplification with primer pairs 11/15 (codon 26), 16/18 (codon 188), and 17/19 (codon 270) resulted in specific amplicons from genomic DNA (Table 2 ). For the detection of mutations in codon 26, the PCR product was digested with BanII for codon 188, MspI and, for the two different mutations in codon 270, SphI and HaeIII, respectively. The resulting fragments were resolved in a native 15% polyacrylamide gel stained with ethidium bromide.
Southern blot analyses were performed with 8 [mu]g of genomic DNA from each of the nine patients and control subjects after digestion with the restriction enzymes EcoRI and HindIII, respectively. The transferred DNA was hybridized to cDNA probes generated with primer pairs 3/8 and 1/4 that were radioactively prepared by random oligonucleotide priming (30 ).
A normal GS cDNA was placed in the pQE-31 vector (Qiagen, Australia) (31 ). Four different cDNA constructs containing the mutations Leu188Pro, Tyr270Cys, Tyr270His and Arg283Cys (Table 3 ) were created by mutagenesis using mutant primers and a DNA Sculptor kit (Amersham). Expression vectors were grown in E.coli in 500 ml LB broth cultures containing 100 [mu]g/ml of ampicillin at 37oC until A600 was 0.7. The cultures were then induced with 1 mM IPTG and the cells harvested 6 h later by centrifugation at 5000 g for 20 min. Cells were resuspended in 50 ml of buffer A (50 mM sodium phosphate buffer pH 7.3, 300 mM NaCl) and lysed by passage through a Ribi cell disruptor. The lysed cells were centrifuged at 10 000 g for 30 min. The supernatant was mixed gently on a revolving rotor with 10 ml of Ni-NTA agarose in a 50 ml plastic tube for ~6-8 h. The resin was washed twice with buffer A on a bench top centrifuge (1000 r.p.m. for 3 min at 4oC) and packed onto a column and connected to an FPLC system. The column was further washed with buffer A at a flow rate of 1 ml/min until the A280 was <0.01. Non-specific binding proteins were washed out with buffer B (50 mM imidazole HCl, pH7.5) at a flow rate of 0.5 ml/min and the eluate (2 ml/tube) was screened by 12% SDS-PAGE. Once the non-specific proteins were removed from the column the imidazole concentration was increased to 500 mM and the glutathione synthetase eluted from the column. The rapid purification results in enzymes with higher specific activities than were obtained previously (24 ,31 ).
We are grateful to a number of physicians and scientists in different countries for providing clinical information as well as samples from patients and their relatives. We thank Progene AB for technical assistance. This work was supported by grants from the Swedish Medical Research Council (4792, 5445, 10326 and 10829), Torsten and Ragnar Söderbergs foundation, the foundation Frimurarebarnhuset, Sävstaholm Society and the Swedish Natural Science Research Council.
1 Larsson, A., Orrenius, S, Holmgren, A. Mannervik, B. (eds) (1983) Functions of Gluthatione Biochemical, Physiological, Toxicological and Clinical Aspects. Raven Press, New York.
2 Taniguchi, N., Higashi, T., Sakamoto, Y. and Meister, A. (eds) (1989) Gluthatione Cennential: Molecular Perspectives and Clinical Implications. Academic Press, New York.
3 Meister, A. (1974) In Boyer, P.D. (ed.) The Enzymes. Academic Press, New York, 3rd Ed., Vol. 10, pp. 671-691.
4 Jellum, E., Kluge, T., Borresen, H.C., Stokke, O. and Eldjarn, L. (1970) Scand. J. Clin. Lab. Invest., 26, 327-330.MEDLINE Abstract
5 Wellner, V.P., Sekura, R., Meister, A. and Larsson, A. (1974) Proc. Natl. Acad. Sci. USA, 71, 2505-2509.MEDLINE Abstract
6 Meister, A. and Larsson, A. (1995) In Scriver, C.F., Beaudet, A.L., Sly, W.S. and Valle, D. (eds) Metabolic and Molecular Bases of Inherited Disease. McGraw Hill, New York, 7th Ed. pp. 1461-1477.
7 Prins, H.K., Oort, M., Loos, J.A., Zurcher, C. and Beckers, T. (1966) Blood J. Hematol., 27, 145-166.
8 Mohler, D.N., Majerus P.W, Hess C.E. and Garrick, M.D. (1970) N. Engl. J. Med., 283, 1253-1257.MEDLINE Abstract
9 Boivin, P., Galand, C. and Schaison, G. (1978) Nouv. Presse. Med., 7, 1531-1535.MEDLINE Abstract
10 Prachal, J.T., Crist, W.M., Roper, M. and Wellner, V.P. (1983) Blood, 62, 754-757.
11 Beutler, E., Gelbart, T. and Pegelow, C. (1986) J. Clin. Invest., 77, 38-41.MEDLINE Abstract
12 Majerus, P.W., Brauner, M.J., Smith, M.B. and Minnich, V. (1971) J. Clin. Invest., 50, 1637-1643.MEDLINE Abstract
13 Wendel, A., Schaich, E., Weber, L. and Flohe, L. (1972) Hoppe-Zeyler's Physiol. Chem.,353, S523-S530.
*To whom correspondence should be addressed. Tel: +46 18 662799; Fax: +46 18 554025; Email: niklas.dahl{at}klingen.uu.se
-->
This page is maintained by OUP admin. Last updated Tue Jun 10 19:01:42 BST 1997. Part of the OUP Journals World Wide Web service.
Copyright
Oxford University Press, 1996
Table 1.
Table 2.