| Human Molecular Genetics | Pages |
Identification of a mutation in liver glycogen phosphorylase in glycogen storage disease type VI
Introduction
Results
Clinical data
Linkage analysis
Mutational analysis
Discussion
Materials And Methods
Genotyping and linkage analysis
Exon-intron structure and mutation analysis
Acknowledgements
References
Identification of a mutation in liver glycogen phosphorylase in glycogen storage disease type VI
Glycogen storage disease type VI (GSD6) defines a group of disorders that cause hepatomegaly and hypoglycemia with reduced liver phosphorylase activity. The course of these disorders is generally mild, but definitive diagnosis requires invasive procedures. We analyzed a Mennonite kindred with an autosomal recessive form of GSD6 to determine the molecular defect and develop a non-invasive diagnostic test. Linkage analysis was performed using genetic markers flanking the liver glycogen phosphorylase gene (PYGL), which was suspected to be the cause of the disorder on biochemical grounds. Mennonite GSD6 was linked to the PYGL locus with a multipoint LOD score of 4.7. The PYGL gene was analyzed for mutations by sequencing genomic DNA. Sequencing of genomic DNA revealed a splice site abnormality of the intron 13 splice donor. Confirmation of the genomic mutation was performed by sequencing RT-PCR products, which showed heterogeneous PYGL mRNA lacking all or part of exon 13 in affected persons. This study is the first to demonstrate that a mutation in the PYGL gene can cause GSD6. This mutation is estimated to be present on 3% of Mennonite chromosomes and the disease affects 0.1% of that population. Determination of this mutation provides a basis for the development of a simple and non-invasive diagnostic test for the disease and the carrier state in this population and confirms biochemical data showing the importance of this gene in glucose homeostasis.
INTRODUCTION
The glycogen storage diseases are a family of disorders caused by defects in the liberation of glucose from glycogen (1). The step of glycogen metabolism that is perturbed and the tissue specificity of the defective enzyme determines the phenotype of the disorders. Biochemical evidence shows that several of these disorders are caused by defects in the phosphorylase enzyme systems. The three characterized glycogen phosphorylase enzymes (PYGB, PYGM and PYGL, E.C. 2.4.1.1) are predominantly expressed in brain, muscle and liver respectively. These enzymes catalyze glycogen breakdown, producing free glucose 1-phosphate to be used for glycolysis and various synthetic functions. Mutations in the best characterized of the phosphorylases, PYGM, cause glycogen storage disease type V, or McArdle disease (OMIM 232600) (2-5). McArdle disease is characterized by a metabolic myopathy with exercise intolerance, myalgia and myoglobinuria. Glycogen storage disease type VI (GSD6), or Hers disease (OMIM 232700), is a clinically and genetically heterogeneous group of disorders that includes hepatomegaly, early fasting hypoglycemia, growth retardation and hyperlipidemia (6,7). Without treatment affected patients develop hepatomegaly and abdominal distension in the first year of life. Other symptoms and signs include irritability, lethargy, weakness, abnormal sweating, difficulty awakening in the morning and poor linear growth. Abnormal laboratory studies include ketosis with mild hypoglycemia, elevated AST, triglycerides and cholesterol. Elevation of serum glucose in response to glucagon is absent. Cognitive development is normal.
Biochemical evidence shows that GSD6 is associated with defects of liver phosphorylase or the multi-enzyme complex that regulates it, the phosphorylase kinase system (1). The X-linked form of GSD6 has been linked to Xp22.1-22.2 (8) and mutations in phosphorylase kinase have been demonstrated (9-13). Although the PYGL gene has been mapped to 14q21-22 (14), no families with GSD6 have yet been shown to be linked to or excluded from linkage to the PYGL locus. Herein we report a Mennonite kindred segregating the autosomal recessive form of GSD6, report linkage to PYGL on chromosome 14 and describe the first mutation in that gene. These results prove that mutations in PYGL are a cause of GSD6 and provide the basis for a rapid diagnostic test for the disorder in Mennonites.
RESULTS
Clinical data
The diagnosis of GSD6 in the extended family was first established in a 22-month-old girl in 1962. The patient had hepatomegaly, fatigue and decelerating linear growth. Liver and muscle biopsies showed enlarged hepatocytes with a granular substance consistent with glycogen. Muscle glycogen was normal (1.7%) but liver glycogen was 20%, approximately four times control values. Although the original enzymological values are not available, records show that the patient had normal hepatic glucose 6-phosphatase but markedly diminished levels of liver phosphorylase. A second Mennonite patient with similar clinical findings underwent liver biopsy in 1971 at the age of 24 months. She had a liver glycogen content of 9% (normal <5%) and a liver phosphorylase activity of 19 and 25% of a control sample on two independent assays. Other patients were determined to be affected by clinical examination, consistent laboratory studies and relationship to the two affected patients who underwent biopsy.
The 17 individuals studied in this report comprise four of the six GSD6 Mennonite nuclear families known to us. Pedigree analysis showed that all families could be traced back to a couple who lived in eastern Pennsylvania in the 1830s. There are several additional known nuclear families with persons affected with GSD6; however, those persons did not elect to participate in this study. The pedigree shown in Figure 1. was constructed from genealogical interviews with the participating families. The diagram shows a simple looping structure for these sibships and for clarity does not show all possible loops. Because these individuals share multiple ancestors, there are many consanguinity loops that could describe transmission of the abnormal allele. We have performed a manual search to identify the most recent common ancestor of the families who elected to participate. The addition of other affected families to this pedigree could suggest that this presumed ancestor is incorrect if that person is not an ancestor of the additional family. For this reason the pedigree is not presumed to be optimal. That individual VIII-4 and her son are affected is an example of pseudodominant inheritance. Pseudodominant inheritance is a description of an inheritance pattern where a parent and child are affected by a phenotype that is normally inherited in a recessive pattern. This inheritance pattern may occur when a phenotype is normally inherited in a recessive pattern, does not reduce reproductive fitness and the carrier rate is high (15).
Figure 1. Pedigree showing four sibships affected with glycogen storage disease type VI. This pedigree shows evidence of autosomal recessive and pseudodominant inheritance of the trait, the latter demonstrated by vertical transmission of the phenotype from VIII-4 to her son. This pedigree shows the estimated simplest loop structure for the four nuclear families and does not include all possible loops.
Linkage analysis
Two-point genetic linkage analyses of the phenotype and markers flanking the PYGL candidate gene (16) showed positive LOD scores at several of the flanking loci (Table 1). Historically a LOD score >3.0 has been considered sufficient to establish autosomal linkage in a genome-wide scan (17), although some have advocated a higher threshold for the genome-wide scan experiment (18). The high threshold accounts for the many tests done in a genome-wide scan. In contrast, in a candidate gene search only a few loci are tested and a LOD score >3.0 is a much stronger result. The two-point analyses in this study show one score >3.0 and several that are near 3.0, which establish linkage to this locus. Multipoint analyses of markers on chromosome 14q using the pedigree shown in Figure 1. gave a LOD score of 4.7 that peaked within 1 cM of the PYGL locus (data not shown). In addition, this genotyping revealed a conserved haplotype of ~1.1 cM or 1.8 Mb (16). This haplotype included markers D14S978, D14S1018 and D14S989, which was found on all affected chromosomes, consistent with a founder effect in a closed population.
Table 1
| Locus | Distancea | [thetas] | |||||
| 0 | 0.01 | 0.05 | 0.1 | 0.2 | 0.4 | ||
| D14S583 | -4 | 1.43 | 1.55 | 1.51 | 1.23 | 0.62 | 0.06 |
| D14S129 | -4 | 1.26 | 1.23 | 1.16 | 0.97 | 0.48 | 0.04 |
| D14S978 | <-1 | 3.62 | 3.48 | 2.91 | 2.24 | 1.07 | 0.08 |
| PYGL | 0 | ||||||
| D14S989 | <1 | 2.86 | 2.74 | 2.26 | 1.70 | 0.76 | 0.05 |
| D14S281 | 2 | 1.17 | 1.13 | 0.97 | 0.76 | 0.40 | 0.04 |
| D14S991 | 2 | 2.41 | 2.32 | 1.92 | 1.43 | 0.65 | 0.05 |
Mutational analysis
The exon-intron boundaries were determined and showed that the coding region of the gene comprised 20 exons (Table 2) (GenBank accession nos AF046787-AF046796). As expected, the position of the PYGL introns was the same as those of PYGM. The intron sizes were determined for 17 of 19 introns and the length of the PYGL introns was substantially different from the PYGM introns. The splice donor and acceptor sequences of the introns corresponded to the consensus sequences, with the exception of the splice donor of intron 8, as described below. A polymorphic (ATT)n microsatellite was identified in intron 12 of the gene (GenBank accession no. AF046797). This marker had a low heterozygosity score (0.41) but should be useful for future candidate mapping studies attempting to link other phenotypes to this locus. The PCR primers for this marker are TCAGCATGTGGTTTGTGCAG and GCAGTCAGCAGAGATTGCAC. The allele size range in 20 CEPH panel parents was 164-176 bp. PCR amplification and sequencing of exons revealed differences in the nucleotide coding sequence when our sequence data were compared with the published sequence (19). All of these sequence variants were found to be common in unrelated persons, unaffected Mennonite individuals or in EST sequences in genetic databases (data not shown; GenBank accession no. AF046798). Sequencing of the flanking intron boundaries in affected persons revealed two unusual sequences, located in the splice donors of introns 8 and 13. Intron 8 had an apparent GT->GC (IVS8+2T>C) sequence at the splice donor and intron 13 had a GT->AT (IVS13+1G>A) sequence. (Fig. 2). The intron 8 splice donor GC sequence was shown to be the normal sequence, as it was present on all chromosomes examined (10 in total) in Mennonite and unrelated controls. In addition, sequences of RT-PCR products that included this region of the mRNA were normal, suggesting that splicing at this exon-intron junction was normal (data not shown). This is consistent with studies of other genes, showing that although this splice donor sequence does not conform to the GT consensus, it is compatible with normal gene splicing (20). In contrast, the intron 13 splice donor alteration was not found in a control sample of 52 unrelated individuals and was associated with abnormal RT-PCR products in Mennonite individuals affected with GSD6. The RT-PCR analysis of affected persons showed two abnormal cDNA products, one missing exon 13 (102 bp) and the other containing an aberrant splice product that uses a cryptic splice donor site to produce a cDNA product deleted for the last three amino acids of exon 13. No clones with the normal sequence were isolated from affected persons. These data predict that both mRNA species retain an open reading frame. The mutation alters a BspMI restriction enzyme site that allowed direct diagnosis by genomic DNA PCR followed by restriction enzyme digestion to discriminate among normal and mutant alleles (Fig. 2B and C). As predicted by the haplotype analysis, all carrier parents were heterozygous for the mutation and all affecteds were homozygous.
Figure 2. (A) Diagram of three exons of the liver glycogen phosphorylase gene showing the normal splice donor sequence at the 5[prime]-end of intron 13 above the line and the mutant sequence below. The thin angled lines at the top of the figure show the normal splicing pattern of exons 12-14. The angled lines on the bottom of the figure show the splicing patterns caused by the intron 13 GT->AT (IVS13+1G>A) splice donor mutation. The two splice products result in mRNA truncation either by skipping exon 13 (splicing of exon 12 to 14, removing all 102 bp of exon 13) or deleting the last three amino acids of exon 13 (use of a cryptic splice site near the 3[prime]-end of exon 13). The cryptic splice site sequence is shown in exon 13; note that the position of the aberrant splice within the exon is not drawn to scale. (B) Separation of BspMI-digested liver glycogen phosphorylase genomic DNA PCR products on an agarose gel from unaffected (U), GSD6 carrier (C) and GSD6 affected (A) patients. Reverse image photograph of an agarose gel stained with ethidium bromide. The right lane (M) shows size standards indicated in base pairs on the figure. (C) Diagram of PCR products predicted from genomic DNA sequencing. The normal splice site includes a sequence recognized by the BspMI restriction enzyme that cuts the 320 bp PCR product into two fragments of 200 and 120 bp. Unaffected non-carriers are predicted to have the two smaller bands, affected patients should have only the larger band and carriers should have all three, as seen in Figure 1. The two arrows labeled `FOR' and `REV' represent the forward and reverse PCR primers used to amplify the genomic DNA, located in exon 13 and intron 13 respectively. Table 2
Exon no.
Exon size
cDNA position
Splice acceptor-EXON-splice donor
Intron no.
Intron size
1
356
1-356
...CCCCAAGgtaaccgggggcgggacggcgggcc
1
NA
2
102
357-458
attatctctcttttctttatttcagAGGGAAT...TTACCAGgtacattgttcctagatttctctca
2
2400
3
79
459-537
aatattgttttcctttcattaccagCTTGGAT...CTTGCTGgtaagtgacattgtgagtgtattat
3
3850
4
104
538-641
tttgcctgtgctctgttgaatgtagCCTGCTT...ATGGCAGgtgtgtgagccatctttttaaattt
4
NA
5
132
642-773
tctaagcaatactgtatccttgcagGTAGAAG...CACTCAAgtattcagagtgctcgtatagccag
5
2550
6
112
774-885
gcagctgcttctgtttaatccacagGTGGTCC...AGAGACTgtgagtacagcacaggccttggttt
6
350
7
83
886-968
tattgtttctctaactttcccctagTTAATGT...TGACAATgtgagtgatcaaggttgctttctta
7
2800
8
144
969-1112
tttattctatctctctggcttgaagTTTTTTG...GGATCAGgcaagtatttagtctgtggatgcag
8
250
9
93
1113-1205
catgttgatgtctttccccacccagGTGGCCA...GTCCAAGgtctggatagtttggcttctttttc
9
700
10
147
1206-1352
cagctgactccttcctttcccccagGCATGGG...TTTAGATgtaagtcattttgagagattttctc
10
325
11
164
1353-1516
aagggcttttttttcctttggctagAGAATTG...CTAAAGTgtgagcctccattactctggcagtg
11
550
12
115
1517-1631
agtgatgtgtctcctccatctccagATTCAAG...AGCAGAGgtaactgggaagtgctgagaggaaa
12
1450
13
102
1632-1733
gtcatcttttcctttgaattgaaagAAAATTG...GAAGCAGgtgagccttccaggtgtgggtcccc
13
700
14
148
1734-1881
ctggcaccatgcactgtcttaacagGAGAATA...TACAACCgtgagtcagccctgtagccaacaag
14
120
15
59
1882-1940
tcatccacttctccatctttttcagGCATTAA...TGGTAAAgtaagttgaatctattcaatgggga
15
100
16
142
1941-2082
aatgcgtttgaccttcttccccaagGCTGCCC...GAAAAAGgtactaccctcattgcaaaatatgc
16
1500
17
208
2083-2290
taattttaacttacctttcttgcagTCATTCC...AGAAAGGgtaagagaatagcttctctttgcta
17
1000
18
135
2291-2425
tcacccacccacccctaccccacagGTACGAG...ATGACAGgtgagttcccataagaagatggtgc
18
500
19
67
2426-2491
tcctcttccactcacaaaataaaagGTTTAAA...GTACATGgtgagtaaattgccaagatttataa
19
2400
20
336
2493-2828
ctgtatatgcatatgttgtttacagAATCCAA...
DISCUSSION
Molecular diagnosis of metabolic diseases can be an important clinical tool for the proper care of affected persons and for genetic counseling. For GSD6, molecular diagnosis is important because the disease is generally mild and definitive diagnostic tests require liver biopsy. Liver biopsy is an invasive procedure with significant morbidity and the risks of that procedure are difficult to justify for the diagnosis of a mild disorder. Although analysis of leukocyte PYGL is technically feasible (21), such an analysis in the presence of significant levels of residual enzyme activity seen in the affected persons (see Materials and Methods) could be problematical. Nevertheless, confirmation of a diagnosis of GSD6 is important because the differential diagnosis of hypoglycemia and hepatomegaly is considerable and includes other more serious disorders that should be excluded.
The prevalence of GSD6 in the Mennonite population is estimated to be near 1 per 1000. Given the homogeneity of the molecular defect in a closed population and the preference for a non-invasive test, molecular diagnosis of a genomic alteration would be clinically useful. To this end we have determined that Mennonite GSD6 is caused by a single base pair change in a splice donor site of intron 13 in the PYGL gene. A compilation of splice donor mutations in 19 genes shows that GT->AT mutations are strongly associated with disruption of normal splicing (20). The Mennonite GSD6 mutation is associated with an alteration of the PYGL mRNA, generating two abnormal mRNA species. These mRNAs predict a PYGL protein that is deleted for either three or 34 amino acids, with both maintaining the reading frame of the protein. These deletions are theoretically consistent with the residual enzyme function shown in several patients, perhaps accounting for the mild nature of the Mennonite GSD6 phenotype. Previous enzymological studies of an affected patient analyzed in this study demonstrated 19-25% enzyme function that is compatible with a mild alteration in the catalytic function of liver glycogenolysis.
Although the PYGL gene was localized to 14q in 1989, no human phenotype has been linked to this gene nor have any mutations in this gene been described in humans. The clinical heterogeneity of GSD6 outside the Mennonite community will require further genetic analyses to clarify the role of PYGL in other persons with GSD6. We predict that mutations resulting in complete loss of function of PYGL would cause a more severe phenotype and cataloguing of other PYGL mutations may provide evidence of important structural and catalytic domains of PYGL. The biochemical data showing that liver glycogenolysis requires multiple proteins predict that other loci may be involved in the pathogenesis of GSD6. Insight into the genetic, catalytic and regulatory components of glycogen physiology should allow a better understanding of glycogen physiology in the normal and pathological states.
Determination of the causative mutation for GSD6 in Mennonites will provide clinicians with a simple and reliable test for this disease. This molecular assay can provide reassurance that the symptoms in an apparently affected child are not attributable to other causes of hepatomegaly and hypoglycemia and that symptomatic care is appropriate. Although genetic tests may be used for prenatal diagnosis, such a use for Mennonite GSD6 would seem unlikely, since the mild nature of the illness and the cultural mores of the community mitigate against such a use. Instead, this test can take its place alongside other simple and reliable assays that are used in the clinic and at the bedside to provide optimal care for ill children.
MATERIALS AND METHODS
Genotyping and linkage analysis
Peripheral blood specimens were collected from 17 patients and relatives in families with one or more individuals with a clinical diagnosis of GSD6. All families were Mennonites from Lancaster County, Pennsylvania, or other Mennonite settlements in Pennsylvania and Ohio. The study was approved by the NIH NCI Institutional Review Board.
Candidate gene linkage analyses with these families were performed by genotyping 10 microsatellite markers that flank the PYGL locus on chromosome 14 (16). Microsatellite markers were assayed by PCR amplification of 20 ng genomic DNA, using oligonucleotide primers that flank the markers in the presence of [[alpha]-32P]dCTP. PCR products were separated on a 6% denaturing polyacrylamide gel (National Diagnostics) and scored manually. The linkage analysis computations were done with FASTLINK 3.0P (22,23) or parallel FASTLINK (24), which are faster versions of LINKAGE (25). Markers were selected and ordered on the basis of their physical position in the Location Database map (16). The disorder was modeled as an autosomal recessive trait with 95% penetrance and a phenocopy rate of 0.1%. Haplotype analysis was performed manually by minimizing recombinants.
Exon-intron structure and mutation analysis
Mutational analysis was performed by PCR amplification of exons from the PYGL gene using primers from the intron sequences of the gene. Exon-intron boundaries were hypothesized to be positioned identically to those of PYGM. Intron sequences were determined by one of two methods, either direct amplification of genomic DNA using exon-specific primers (deduced from PYGM exon boundaries) or by vector-exon PCR of a mini-library of the PYGL gene. Direct amplification was attempted for all introns, as the sizes of most PYGM introns were small. Introns that could not be amplified by this technique were isolated by creating a mini-library of the PYGL gene from a PYGL-containing BAC clone. This clone was isolated by screening the Genome Systems library with PCR primers from the 3[prime]-end of the published cDNA. That this clone contained the full coding region of the gene was confirmed by amplifying the hypothesized 5[prime]-most exon of the gene. The BAC was digested with one of three four-base cutting enzymes and ligated into pre-digested pUC vector (Pharmacia Biotech). PCR was performed using one cDNA-specific primer and either M13 forward or reverse primer. PCR products were sequenced using either the ABI 373 system (Perkin-Elmer) or Sequitherm Excel (Epicentre Technologies) and [[alpha]-35S]dATP incorporation or ThermoSequenase and [33P]ddNTP terminators (Amersham) followed by denaturing gel electrophoresis. Peripheral blood lymphocytes were transformed with EBV to generate lymphoblastoid cell lines that were used to harvest mRNA. mRNA was analyzed by reverse transcription and amplified by PCR to generate products that were directly sequenced or cloned into pUC18 and sequenced (26). The exon 13 splice donor mutation was analyzed by amplification of genomic DNA with two pairs of nested primers: N12F1 (CCCATGTTCTGTGTGATACG) and N13R1 (GTTGCAGTGAGCCGAGATCG) for the first amplification and N12F2 (TACACTGTGGTAGGTGACAG) and N13R2 (CACTCCAGCCTGGGCAACAG) for the second round. The PCR products were digested with BspMI according to the manufacturers instructions (New England Biolabs), separated on a 2% agarose gel and stained with ethidium bromide.
ACKNOWLEDGEMENTS
The authors thank Robert Nussbaum, Lawrence Brody, Barbara Biesecker and Francis Collins for reviewing previous versions of this manuscript. Lillie Rizack collected the specimens analyzed in this study and Richard Kelley assisted with interpretation of biochemical data. Alejandro Schäffer provided assistance with multiprocessor computer analyses and John Powell and Jim Tomlin of the Division of Computer Resources and Technology provided access and assistance with the IBM SP2 analyses. We also thank the families for their kindness, hospitality and their willingness to assist us in this study. The sequences reported in this study are deposited with GenBank, accession nos AF046787-AF046798.
REFERENCES
This page is run by Oxford University Press, Great Clarendon Street, Oxford OX2 6DP, as part of the OUP Journals
Comments and feedback: www-admin{at}oup.co.uk
Last modification: 4 Apr 1998
Copyright© Oxford University Press, 1998.
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||