Human Molecular Genetics, 2000, Vol. 9, No. 10 1501-1513
© 2000 Oxford University Press
Human adenylosuccinate lyase (ADSL), cloning and characterization of full-length cDNA and its isoform, gene structure and molecular basis for ADSL deficiency in six patients
ebesta1,21Institute for Inherited Metabolic Disorders, Ke Karlovu 2, 120 00 Prague 2, and 2Department of Clinical Biochemistry, Charles University 1st School of Medicine and General Faculty Hospital, 120 00 Prague, Czech Republic
Received 11 February 2000; Revised and Accepted 4 April 2000.
DDBJ/EMBL/GenBank accession nos. AF067853, AF067854, AF106656.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
Adenylosuccinate lyase (ADSL) is a bifunctional enzyme acting in de novo purine synthesis and purine nucleotide recycling. ADSL deficiency is a selectively neuronopathic disorder with psychomotor retardation and epilepsy as leading traits. Both dephosphorylated enzyme substrates, succinylaminoimidazole-carboxamide riboside (SAICAr) and succinyladenosine (S-Ado), accumulate in the cerebrospinal fluid (CSF) of affected individuals with S-Ado/SAICAr concentration ratios proportional to the phenotype severity. We studied the disorder at various levels in a group of six patients with ADSL deficiency. We identified the complete ADSL cDNA and its alternatively spliced isoform resulting from exon 12 skipping. Both mRNA isoforms were expressed in all the tissues studied with the non-spliced form 10-fold more abundant. Both cDNAs were expressed in Escherichia coli and functionally characterized at the protein level. The results showed only the unspliced ADSL to be active. The gene consists of 13 exons spanning 23 kb. The promotor region shows typical features of the housekeeping gene. Eight mutations were identified in a group of six patients. The expression studies of the mutant proteins carried out in an attempt to study genotype–phenotype correlation showed that the level of residual enzyme activity correlates with the severity of the clinical phenotype. All the mutant enzymes studied in vitro displayed a proportional decrease in activity against both of their substrates. However, this was not concordant with strikingly different concentration ratios in the CSF of individual patients. This suggests either different in vivo enzyme activities against each of the substrates and/or their different turnover across the CSF–blood barrier, which may be decisive in determining disease severity.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
Adenylosuccinate lyase [adenylosuccinase (ADSL); EC 4.3.2.2.] is an enzyme acting in two pathways of purine nucleotide metabolism. It catalyses the conversion of succinylaminoimidazole carboxamide ribotide (SAICAR) into aminoimidazole carboxamide ribotide (AICAR) in the purine de novo synthesis pathway, and the formation of adenosine monophosphate (AMP) from adenylosuccinate (S-AMP) in the purine nucleotide cycle. ADSL deficiency (McKusick 103050) was first recognized by Jaeken and van den Berghe (1) in 1984 in patients with hereditary psychomotor retardation and autism and remains so far the only inherited deficiency of de novo purine synthesis recognized in humans. To date, about 50 patients have been diagnosed world-wide and reports on about half of them have been published (1–8). The disease usually appears within the first months of life with neurological involvement. Affected individuals later show various degrees of psychomotor retardation often accompanied by various combinations of epilepsy, hypotonia, growth retardation and muscular wasting. A whole spectrum of behavioural changes such as autism, aggressiveness and self-mutilation have been also described (1,3). The non-specific nature of the neurological symptomatology results in the diagnosis relying on biochemical detection of dephosphorylated substrates of ADSL: SAICA riboside (SAICAr) and succinyladenosine (S-Ado) in body fluids. For their detection, several simple screening procedures have been developed (9–12). To better understand the ADSL function, enzymic assays and metabolic functional studies were performed in different tissues of ADSL-deficient patients with the following results. First, ADSL activity has been repeatedly proven to be decreased with both ADSL substrates, SAICAR and S-AMP, which confirms the bifunctional role of the enzyme (13,14). Secondly, residual enzyme activities varied between different tissues and the enzyme deficiency was not always generalized, suggesting the existence of either hitherto unrecognized tissue-specific ADSL isoforms or tissue-specific gene expression regulation (15). Thirdly, functional studies in fibroblasts showed neither a significant reduction in the rate of de novo purine synthesis nor any abnormality in purine nucleotide concentrations (16). Neither of these biochemical observations could be correlated with the severity of the clinical phenotype nor provide any explanation for the pathogenetic mechanism of selective neuronopathic features of the disease. The only biochemical finding that seems to correlate with the severity of the symptoms is the proportion of the accumulating S-Ado and SAICAr in body fluids, particularly in the cerebrospinal fluid (8,14).
Molecular studies of the ADSL gene were started by the work of Van Keuren (17) who assigned the ADSL gene to chromosome 22 using a somatic cell hybrid panel. The human liver ADSL cDNA (GenBank accession no. S60710) was cloned in 1992 (18) and the gene was mapped to chromosome 22q13.1--q13.2 (19). The nucleotide sequence showed 94% and 85% identity to the mouse (20) and avian ADSL cDNAs (21), respectively, and encoded a protein 459 amino acids long. The enzyme proved to have a homotetrameric structure with a relative molecular mass of ~52 kDa per subunit (18,22,23). However, several indications suggested discrepancy in the cDNA sequence. The C
A exchange in the third nucleotide of the cDNA sequence, accession no. X65867, deposited in the GenBank database, creates an alternative initiation codon located in-frame, 75 nucleotides upstream from the one reported by Stone (18). If translated from this first initiation codon the protein would have an additional 25 amino acids at the N-terminus. The question of the correct position of the initiation codon was also stressed by expression of a recombinant protein transcribed from the latter initiation codon. This resulted, even under standard experimental conditions, in an insoluble and inactive enzyme (22). Despite this discrepancy, cDNA mutation analysis was performed successfully in a dozen ADSL-deficient patients (4,18,24,25). These studies showed broad allelic heterogeneity with 14 mutations identified. The allelic heterogeneity suggests correlation with phenotypic variability; however, no convincing data providing a particular link between them have been presented so far.
Therefore, the present study had the following objectives of (i) identification, cloning, expression and characterization of the correct, full-length ADSL cDNA sequence and of its possible isoform(s), (ii) elucidation and description of the ADSL gene structure and its complete genomic sequence, (iii) identification of ADSL mutations in a series of six patients and (iv) expression and kinetic characterization of the mutant enzyme proteins, including the S-Ado/SAICAr ratio variance in an attempt to elucidate the genotype–phenotype correlations.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
Identification of a complete ADSL cDNA sequence
To identify the complete ADSL cDNA sequence, muscle and lymphocyte full-length cDNAs were prepared, using the CapFinder technology (Clontech, Palo Alto, CA). The full-length cDNAs were ligated to GenomeWalker Adaptors (Clontech) and overlapping 5' and 3' ADSL cDNA ends were amplified using the ADSL cDNA and GenomeWalker Adaptor specific primers. Sequence analysis of the resulting fragments showed a novel 52 bp sequence at the 5'-end of the ADSL cDNA. (Fig. 1, GenBank accession no. AF067853). The sequence contained both initiation codons identical to those described previously (GenBank accession no. X65867).
|
Alternative splicing of ADSL mRNA
Two PCR products of different lengths were consistently generated by the RT–PCR amplification of the full-length ADSL cDNA. To reveal their identities the PCR products were cloned and sequenced. Sequence analysis showed the absence of 177 bp corresponding to exon 12, in the shorter fragment. This mRNA isoform (GenBank accession no. AF067854) can be translated in the same reading frame as the non-spliced ADSL mRNA but the synthesized protein would lack 59 amino acids (amino acids 397–456). The tissue-specific patterns of expression of both ADSL isoforms were studied using PCR analysis of the Clontech Human Multiple Tissue cDNA (MTC) Panel I. The results showed both isoforms to be expressed in all tissues studied in approximately the same relative amount with the unspliced form being ~10-fold more abundant (Figure 2).
|
Expression of ADSL isoforms in Escherichia coli
To characterize functionally the identified sequences, both cDNAs were cloned and expressed. The resulting fusion proteins were purified, cleaved with factor Xa, analysed by SDS–PAGE and the activities of both forms were measured using both ADSL substrates, S-AMP and SAICAR. The results showed that the full-length cDNA encodes an active protein. The product of the alternatively spliced isoform, although stable, was entirely inactive (Fig. 3).
|
Structure and complete genomic sequence of the human ADSL gene
To define the ADSL genomic structure, the known genomic organization of the murine ADSL gene (20) was used to predict conserved splicing sites within the human ADSL cDNA. Primers located within the predicted ADSL exons were designed and used in both a genome walking strategy (Universal GenomeWalker, Clontech) and for direct amplification of ADSL fragments from genomic DNA. The PCR fragments generated by genome walking were gel purified, cloned and sequenced. The fragments generated by genomic PCR were either cloned and sequenced as described above, or filter purified and sequenced using the ADSL cDNA-specific primers.
With this approach, genomic fragments covering the ADSL promoter and all of the exons including flanking intronic sequences were isolated and sequenced. The human ADSL gene was found to consist of 13 exons. Exon lengths, their positions within the cDNA and the sequences of all of the exon–intron boundaries are shown in Table 1. All the intron–exon splice junctions conform to the GT/AG consensus sequence. The genomic organization of the human ADSL gene showed identity to that of the murine gene.
|
To complete the whole gene sequence, the identified sequences were analysed using the BLASTN homology search in the Chromosome 22-specific Blast server database (www.sanger.ac.uk/HGP/Chrom_blast_server.shtml). Several sequences showing significant homologies were identified and allowed the reconstruction of the whole human ADSL genomic sequence which spans over 23 kb (GenBank accession no. AF106656).
The lengths of the introns range from 146 to 4536 bp (Table 1, Fig. 4). Using the databases searches, we identified several typical promoter elements such as Oct 1, AP1, HSF2, Pbx-1a, E2F and Sp1 sites up to 500 bp upstream of the 5'-end of the cDNA sequence reported here. Canonical TATA and CAAT boxes were apparently absent (Fig. 4).
|
Mutation analysis of patients with ADSL deficiency
Full-length cDNA from both patient and control lymphocytes or fibroblasts were prepared and sequenced, with the following mutations identified: 63C
T (A3V), 395TC (Y114H), 624GA (R190Q), 635CT (R194C), 857GA (D268N), 1332GA (R426H) and 1343GA (D430N). The individual genotypes together with clinical and biochemical findings are listed in Table 2. All the mutations identified by cDNA sequencing, except for 63CT, were confirmed by individually designed PCR–RFLP tests carried out on both cDNA and gDNA (Table 3). The 63CT mutation, found in patient 2, appeared to be homozygous according to cDNA analyses (sequencing and PCR–RFLP). However, the PCR–RFLP analysis of the gDNA ruled out the homozygosity and showed the patient to be heterozygous for it. To isolate and analyse the second allele, cDNA from the patient was cloned and an allele-specific PCR assay was used to discriminate between 63C and 63T alleles. In total, 280 clones were analysed and only mutated 63T alleles were found. To identify the missing mutation in patient 2, a panel of intronic primer pairs was designed in order to amplify and sequence the ADSL promoter and all the coding regions (Table 4). Using this approach a heterozygous CT transition in exon 9 (1064CT on cDNA) was found. This mutation creates a premature stop codon in the corresponding mRNA (337RX) which leads to its nonsense-mediated decay.
|
|
|
Characterization of identified mutations
To test the functional consequences of the individual mutations, the corresponding cDNAs were cloned and the mutated recombinant proteins expressed, isolated and characterized. Both the fusion proteins and proteins without the affinity tag were prepared and analysed on SDS–PAGE. Except for the 114H, all the proteins were stable (Fig. 3A and B). The catalytic activities of both protein forms were assayed using both ADSL substrates. Activities of the mutated enzyme proteins ranged from 0 to 140% of the wtADSL activity (Fig. 3C and D). The mutants had activities similar to or even higher than the wild type and compared with wild-type enzyme displayed modest changes in thermal stability and temperature dependence of ADSL activity. Significant effect was observed for the 430N mutant at 45°C and 55°C (t-test) (Fig. 5A and B). All of the mutants also showed changes in some of the kinetic parameters (Table 5). Mutant 190Q showed difference in KM for S-AMP and Vlim for both substrates and mutants 430N and 3V showed lowered KM for SAICAR (z-test).
|
|
Genotype–phenotype analysis
Clinical, biochemical and molecular data of individual patients are shown in Table 2. The patients were ordered according to the severity of their symptoms. The S-Ado/SAICAr concentration ratios in cerebrospinal fluid (CSF) or urine correlate with the severity of symptoms. To explain the variations in S-Ado/SAICAr concentration ratios, the ADSL activities were measured in patients’ skin fibroblasts. The residual activities ranged from 20 to 60% of controls and were proportionally decreased with both substrates. No correlation between the residual activities and severity of symptoms was observed. The S-AMP/SAICAR activity ratios in individual patients were 0.8–1.2 against 1.44 ± 0.16 in controls and did not reflect the S-Ado/SAICAr concentration ratios found in CSF or urine. The same applied to the activities of the expressed mutated recombinant proteins. The residual activities were again decreased proportionately, with the S-AMP/SAICAR activity ratios ranging from 0.6 to 1.0 (wtADSL 1.3 ± 0.2) (Fig. 6B).
|
A trend suggesting a relationship between the phenotype severity and the predicted residual activities of individual genotypes was calculated as a mean of the homoallelic in vitro activities and is shown in Fig. 6A.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
The main goal of the study was to characterize basic molecular aspects of ADSL pathobiology with special emphasis on the clinical and biochemical heterogeneity observed in this disease.
Identification of the full-length cDNA
To isolate and characterize the full-length ADSL cDNA we set up an original approach combining the advantages of the previously described methods for a full-length cDNA preparation (CapFinder technology, Clontech, Palo Alto, CA) with amplification of unknown flanking DNA sequence (vectorette PCR). Using this approach we identified a novel 5'-end cDNA sequence absent among more than 70 ADSL expressed sequence tags (ESTs) deposited currently in the human dbEST database, where the longest deposited EST sequence starts at nucleotide 35 of the novel sequence. This result illustrates the utility of our method for full-length cDNA preparation and its superiority over protocols currently used for full-length cDNA synthesis.
The enzyme protein
The 5'-end cDNA sequence identified contained the alternative initiation codon (M1) previously identified by Fon et al. (GenBank accession no. X65867). To characterize the full-length ADSL translated from the M1, we cloned and expressed the corresponding cDNA sequence in E.coli. The expressed enzyme protein was soluble, active and stable in contrast to the truncated protein translated from the M2, which was insoluble and inactive under standard experimental conditions (22). The kinetic parameters of the full-length enzyme (Table 5) were identical to those of the native enzyme isolated from human erythrocytes (26), rat muscle (27) and even to those of the truncated enzyme kept soluble under specific conditions (22). This implies that the identified 25 amino acid N-terminal ADSL sequence is most probably not essential for the enzyme catalytic competency but may be important for its structural stability. This is further corroborated by the nearly identical kinetic properties but altered temperature activation profiles (Table 5, Fig. 5B) of the A3V mutated protein present in the severely affected patient 2.
All the above corroborated by the identification of two mutations within the said N-terminal sequence (ref. 25 and this study) and a high degree of sequence identity within this part of the sequence with the murine ADSL cDNA (20) provides a strong argument that the native human protein starts with the M1 and is composed of 484 amino acids, the same as the murine ADSL.
ADSL isoforms
Two ADSL mRNA isoforms produced by an alternative splicing of exon 12 have been identified. The presence and the relative abundance of each transcript were studied in a normalized brain, heart, kidney, liver, lung, pancreas, placenta and skeletal muscle cDNA, using RT–PCR. The results showed that both transcripts are expressed in all tissues studied, in the same relative amounts with the unspliced form being ~10-fold more abundant. The exon 12 deletion preserves an open reading frame of the unspliced wtADSL mRNA. If translated, the enzyme would miss 59 amino acids (position 397–456 of the ADSL). To test the kinetic properties of the alternatively spliced isoform (
ADSL ) we cloned the ex12 ADSL cDNA and expressed it as a recombinant protein. The ADSL was produced in an amount comparable to the wild-type ADSL. However, the protein was completely inactive and its kinetic parameters therefore could not be determined. The functional relevance of the ADSL isoform remains to be elucidated. Its mRNA is stable, is expressed in all major tissues and is most probably translated. Apparent is the question of whether the ADSL isoform could form mixed tetramers with full-sized subunits. If so, an array of enzymes with different activities would be possible, dependent on the composition of the tetramer. This may also impact relative expression in those individuals harbouring a mutant subunit, in which case an even larger number of subunit combinations would be possible, which may contribute to quantitative differences in ADSL activities across individuals. We intend to study the presence of the predicted protein isoform using a set of polyclonal anti-ADSL antibodies.
ADSL gene structure and genomic sequence
Complete genomic sequence and gene structure of ADSL were described here for the first time. To initiate analysis of the potential gene expression regulating mechanisms, the 500 bp upstream sequence from the 5'-end of the cDNA sequence reported here was computer-analysed using database searches. The promoter region, the promotor activity of which was established recently (28), has features typical of a housekeeping gene: a high G + C content, a high frequency of CpG nucleotides, and no TATA and CAAT boxes. No motifs suggesting tissue-specific expression regulation were found. Furthermore, knowledge of the genomic ADSL sequence allowed design of a panel of intronic primer pairs for PCR amplification and subsequent sequence analysis of the ADSL promotor and coding sequences. These data provide a basis for reliable postnatal and prenatal DNA-based diagnosis of ADSL deficiency on gDNA.
Mutation analysis
We identified eight mutations in a series of six patients, increasing the total amount of mutations identified within the ADSL gene to 21 [(4,18,24,25,28) and ADSL database (http://www.icp.ucl.ac.be/adsldb/mutations.html )]. All the mutations represent single base pair substitutions. These include 18 missense mutations, one nonsense mutation, one splice site mutation and one mutation in the 5'-UTR (28) of the mRNA. The most frequent mutation, R426H, has been detected in 10 homozygotes and six compound heterozygotes (ADSL database, refs 4,25,28 and this work) and thus accounts thus for more than 50% of the mutated alleles analysed so far. Only two other mutations, R190Q, found in the Czech and Belgian patients, and K246E, found in two unrelated Belgian patients (25), were present on more than one allele. The majority of the mutations, 14, occurred within exons 1, 2, 5 and 12. It is worth mentioning that except for the R426H mutation the other four mutations, located within the alternatively spliced exon 12, were found in the less severely affected patients. This emphasizes further the importance of studying the biological relevance of the ADSL isoform.
The allelic heterogeneity and different ethnic origin of the ADSL-deficient patients (Slavic, Romany, Moroccan, Turkish, Spanish, Italian, Dutch, Belgian, German, Australian, US and Brazilian) suggests that ADSL deficiency is panethnic and is probably more common than previously thought. Thanks to the simple and inexpensive screening procedures available (9–12), ADSL deficiency should be tested in all children with congenital psychomotor retardation and neurological involvement. This approach succeeded in diagnosing five patients out of 2000 children screened for the disorder by our laboratory (3).
Expression studies of the mutated proteins
To see the functional effect of individual mutations we expressed seven relevant amino acid sequence-altering alleles in E.coli and measured enzyme activities of the affinity-purified enzyme proteins with each ADSL substrate. According to the residual activities, the mutant enzymes were divided into three groups: null mutations free of detectable enzymic activity (Y114H and D268N), severe mutations with enzyme activities substantially compromised (R194C and R426H) and mild mutations with activities comparable to that of the wild-type enzyme (A3V, R190Q and D430N). We propose the following mechanisms to explain the causative effect in the category of null mutations. In the case of the Y114H mutation it may be the pronounced protein instability (see Fig. 3). The wild-type tyrosine-114 resides in a highly conserved region (TSCYVGDN), common to all members of the fumarase enzyme superfamily. Based on molecular modelling, this region was predicted to be essential for the Bacillus subtilis ADSL subunit assembly (29). Furthermore, the active site of the B.subtilis ADSL is formed by histidine–histidine residues pairing (29). One of these histidines, His68, is spatially close to this conserved region. Based on the homology with the B.subtilis enzyme it could be that the mutated histidine-114 residue in this region interferes with a proper active site formation. Both mechanisms, the hampered monomer association and improper active site formation, or even a combination of both can thus lead to the observed enzyme instability and inactivity. The diminished activity of the D268N mutant is easily explainable, as the wild-type aspartic acid residue resides in the active enzyme site and is directly involved in the catalytic mechanism where it serves as an electron donor of the acidobasic cleavage (22). The exchange of this acidic residue for the uncharged asparagine leads to active centre disruption but does not seem to interfere with the enzyme protein stability. The effect of the severe and mild mutations is very hard to explain using the experimental data. There is no doubt that the mutations compromising severely the enzymic activity are disease causing. To be able to make such a statement for the group of the mild mutations, temperature activation profiles and kinetic parameters of the expressed proteins were studied. All three mutants, compared with wild-type enzyme, showed modest changes in thermal stability and temperature dependence of ADSL activity suggesting the protein susceptibility to accelerated intracellular degradation.
Genotype–phenotype correlation
As in other metabolic disorders affecting the central nervous system, the relation between the enzyme defect and clinical symptoms is not well understood. Two main hypotheses of the pathophysiological mechanism in ADSL deficiency have been proposed. Impaired synthesis of purine nucleotides leading to purine and particularly adenine nucleotide deficiency was ruled out by the work of Van den Berghe et al. (16). On the other hand, there is evidence for interference of the accumulating succinylpurines with brain glucose metabolism in human patients (30), and of toxic effect of SAICAr on neurons in specific regions of rat hippocampus (31). An inverse relationship between the degree of clinical involvement and the excess of S-Ado over SAICAr has been proposed (8,14), providing further argument in favour of the toxic effect of succinylpurines.
Our attempts at correlating the severity of the clinical course with biochemistry can be summarized as follows. No strict correlation of genotype with phenotype could be drawn. With no exception all the patients carried at least one allele showing some residual activity, supporting the assumption that complete loss of ADSL activity is probably not compatible with life. The tendency in patients carrying a combination of milder mutations to be less affected suggests the pivotal role of the residual enzyme activity in determining severity of symptoms. However, we could not prove any difference in activities using either one of the two substrates with all the expressed enzymes as well as with the patients’ fibroblast cell lines. Such a finding would represent an important link between the modulated enzyme activity and the varied body fluid S-Ado/SAICAr ratio, suggesting an inverse correlation with disease severity (8,14).
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
Patients
Six patients from five unrelated families (four of Czech and one of American origin) were studied. A brief summary of the clinical and biochemical data is given in Table 2. More detailed case histories together with clinical and biochemical data of the Czech patients have already been published (3). The US case was published recently (5).
5' and 3' random amplification of cDNA ends (RACE)
To prepare the ‘full-length’ ADSL cDNA we used the CapFinder PCR cDNA Library Construction Kit (Clontech). Briefly, muscle and lymphocyte RNAs were reverse transcribed by MMLV reverse transcriptase in the presence of Clontech CDS/3'oligo(dT)30 N–1N; (N–1 = A, C or G, N = A, C, G or T) and CapSwitch (template for reverse transcriptase after the switch at the 7-methylguanosine cap structure) oligonucleotides. Prepared cDNAs were amplified by LD-PCR using primers specific for CDS/3' and CapSwitch oligonucleotides, so that only cDNA containing CapSwitch sequence on the 5'-end could be amplified. The cDNA was treated with T4 DNA polymerase and ligated into GenomeWalker Adaptor (Universal GenomeWalker kit, Clontech). 5'- and 3'-ends of ADSL cDNA were prepared by nested PCR using the ADSL and GenomeWalker Adaptor specific primers. The PCR products were purified and sequenced.
RT–PCR
Total RNA was isolated from peripheral blood lymphocytes, fibroblasts or skeletal muscle by the standard procedure (32). Poly(A)+ RNA was isolated using the Oligotex Direct mRNA kit (Qiagen, Hilden, Germany). cDNA was reverse transcribed from total or poly(A)+ RNA using the 1st-strand cDNA Synthesis kit (Clontech, Palo Alto, CA) and oligo(dT) primer. ADSL cDNA was amplified in PTC-200 DNA Engine (MJ Research, Waltham, MA) in a reaction volume of 25 µl containing 2.5U KlenTaq1 polymerase (Ab Peptides, St Louis, MO), 0.1 U DeepVent polymerase (NEB, Beverly, MA), 200 µM dNTPs, 0.15 µM primers and 1.5 mM MgCl2 under the following conditions: the initial denaturation at 94°C for 2 min followed by 30 cycles of denaturation at 94°C for 10 s, primer annealing at 59°C for 10 s and extension at 72°C for 1 min. Final chain elongation was performed at 72°C for 10 min.
The cDNAs used for cloning were amplified in a 50 µl reaction containing 1.2 U of cloned Pfu DNA polymerase (Stratagene, La Jolla, CA), 200 µM dNTPs, 0.10 µM primers and 2.5 mM MgCl2 under the following conditions : the initial denaturation at 94°C for 1 min was followed by 30 cycles of denaturation (94°C for 10 s), primer annealing (55°C for 20 s) and extension (72°C for 4 min). Final chain elongation was carried out at 72°C for 10 min.
DNA sequencing
cDNAs prepared by RT–PCR were purified and concentrated on Microcon 100 microconcentrators (Amicon, Beverly, MA), while the plasmid DNA samples were prepared using the miniprep kits SNAP (Invitrogen, NV Leek, The Netherlands). Dideoxy cycle sequencing was performed in 7 µl containing 150 fmol of template, 1 pmol of the fluorescently labelled primer, 100 µM dNTP (dNTP:ddNTP ratio = 120), 5 mM MgCl2 and AmpliTaq FS polymerase (PE Biosystems, Foster City, CA). Sequencing parameters were: initial denaturation (94°C for 2 min), 40 cycles of denaturation (94°C for 15 s), annealing of primers at specific temperature for 10 s and extension (72°C for 30 s), followed by 10 cycles of denaturation (94°C for 15 s) and extension at 72°C for 30 s. The samples were subsequently denatured in loading buffer (94°C for 3 min). Sequences were read either on the ALF sequencer—EMBL prototype (EMBL, Heidelberg, Germany) (fluorescein primers) or on the ALFExpress sequencer (Pharmacia, Uppsala, Sweden) (Cy5 primers). Sequences were analysed using the GeneSkipper software (EMBL). Both strands of cDNA were always sequenced. All the primers (Generi Biotech, Hradec Králové, Czech Republic) used are listed in Table 3.
Mutation analysis
To verify the nature of mutations and to be able to screen for individual mutations in affected families, we designed PCR–RFLP-based assays on genomic DNA. These assays were all based on the PCR-based introduction of either one or both, diagnostic and control, restriction enzyme recognition sequence sites. The primers and restriction enzymes used are listed in Table 3. The genomic DNA for PCR analysis was isolated either from blood samples (Qiagen, Hilden, Germany) or extracted from dried blood spots (33).
Genomic DNA sequencing
The PCRs were carried out according to standard procedure. Genomic DNA (100 ng) was amplified in 25 µl containing 2.5 U KlenTaq1 polymerase, 0.1 U DeepVent polymerase, 200 µM dNTPs and 0.15 µM primers. Amplification products were gel purified using the phenol-based method (34). Purified fragments were sequenced using the T7 and M13 primers. The dideoxy cycle sequencing was performed in a 7 µl reaction containing 150 fmol of template, 1 pmol of the Cy5 labelled primers, 100 µM dNTP (dNTP:ddNTP ratio = 120), 5 mM MgCl2 and AmpliTaq FS polymerase. The sequences were read on ALFExpress sequencer. Sequences were analysed using the GeneSkipper software.
Alternative splicing
To reveal the identity of the shorter fragment we prepared the RT–PCR product as described above. After 30 cycles, 1 U Taq polymerase was added and the mixture was incubated for 15 min at 72°C. DNA was phenol–chloroform purified and ligated into pCRII-TOPO vector (Invitrogen). The OneShot cells (Invitrogen) were transformed according to the manufacturer’s protocol and grown on Luria–Bertani (LB) plates containing 50 µg/ml ampicillin, X-gal and isopropyl 1-thio-ß-D-galactoside (IPTG). The lithium minipreps were prepared from positive clones. The inserts were cleaved out of the vector by EcoRI and clones containing the fragment of interest were sequenced as described above. Sequence alignment was performed with the CLUSTAL W V1.5 program (35).
Tissue-specific expression patterns of ADSL isoforms
The tissue-specific expression pattern of ADSL mRNA isoforms was investigated by PCR, using the Human Multiple Tissue cDNA (MTC) Panel I (Clontech) as a template. ADSL cDNA was amplified as described above. A control fragment of glyceraldehyde-3-phosphate dehydrogenase (G3PDH) was amplified according to the manufacturer’s protocol. The PCR products were analysed on 1% SeaKem agarose (FMC, Rockland, ME).
Expression of ADSL isoforms in E.coli
The ADSL cDNA was prepared and cloned into pCRII-TOPO vector as described above. The constructs of the correct sequence were used as templates for subsequent PCR amplification of ADSL cDNA using the 56S primer starting at the first ATG codon. PCR products were blunt end ligated into the pMAL-c2 vector (NEB). The constructs with the correct sequences were introduced into the E.coli strain DH5
F'IQ (Gibco, Paisley, UK) for fusion protein production. Briefly, a 600 ml culture of transformed bacteria was grown at 37°C in rich medium, 100 µg/ml ampicillin and 0.2% glucose to a density A600 of 0.5. IPTG was added to a final concentration of 0.3 mM and incubation continued for an additional 3 h. Bacteria were harvested by centrifugation at 3000 g for 10 min. The pellets were resuspended in 20 mM Tris–HCl pH 7.4, 200 mM NaCl and 1 mM EDTA buffer and sonicated four times for 15 s at 40 W. Crude lysates were obtained by centrifugation at 8000 g for 30 min. Fusion proteins were isolated from crude lysate on amylose affinity columns (NEB). The fusion proteins were cleaved overnight at 4°C using factor Xa (NEB). The ADSL activities were measured in crude lysate, isolated fusion protein and in cleaved fusion using the HPLC analysis of AMP and AICAR formed from both ADSL substrates, S-AMP (Sigma) and SAICAR (own synthesis, unpublished data). The reactions were run for 20 min at 37°C in 150 µl containing 10 mM Tris pH 7.5, 2 mM EDTA, 10 mM KCl, 1 mM DTT, 4% glycerol and 15 µg protein. Substrate concentrations were 0.14 and 0.09 mM for S-AMP and SAICAR, respectively. SDS–PAGE analyses were performed according to standard procedure: protein concentrations were determined using the Bradford assay (Sigma).
Expression of mutated ADSL in E.coli
All mutated cDNAs were first cloned into pCRII-TOPO vector as described above. Individual mutated alleles were sequenced and subsequently subcloned into the pMAL-c2wtADSL construct, using the appropriate set of restriction enzymes: for alleles 395C, 624A and 635T, Eco72I–NheI; for alleles 857A, 1332A and 1343A, NheI–BseRI. Allele 63T was prepared using the mismatched 56S primer (ATGGCGGTTGGAGGCGAT). PCR product was blunt end cloned into the pmal-c2 vector as described above. Resulting clones were sequenced and introduced into E.coli strain DH5F'IQ. The protein expression experiments and enzyme activity measurements were performed in 20 ml cultures essentially as described above. The expression experiments were performed five to eight times with the enzyme activity measurements run in duplicate.
Thermal stability experiments and kinetic characterization of the expressed proteins
Thermal stability is defined as an enzyme activity measured at 37°C following the incubation at given temperatures for 35 min. Temperature dependence of ADSL activity is defined as an enzyme activity measured at given temperature after a 35 min incubation. The starting activity represents the protein activity under standard reaction condition (20 min reaction at 37°C). The thermal stability experiments were performed in three expression experiments in duplicate for each temperature indicated. The temperature dependence of ADSL activity was studied in single experiment. The Michaelis–Menten kinetics were established for both substrates at 10 different concentrations ranging from 0.5 to 130 µM. The reaction conditions for enzyme activity measurements were the same as described above.
ADSL activities in cultured fibroblasts
ADSL activities were measured in five controls and five patients’ cell lines essentially as described above. Measurements for each substrate were run in duplicate using a 100 µl reaction volume and 15 µg of protein extract.
Accession numbers
The following sequences have been deposited with the GenBank database: complete ADSL cDNA sequence under accession no. AF067853; the alternatively spliced (ex12) ADSL cDNA isoform under accession no. AF067854; the ADSL genomic sequence under accession no. AF106656.
|
ACKNOWLEDGEMENTS |
|---|
We thank Dr Zumrová, Dr vehláková, Dr Valík and Dr Jones who referred their patients or cell lines for investigation; to M. Elleder for help in manuscript preparation, and V. Koich and M. Høebíèek for helpful discussions. This work was supported by grant 3608-3 from the Grant Agency of Ministry of Health of the Czech Republic and partly by grants VS96-127 from the Ministry of Education and 301/00/0689 from the Grant Agency of the Czech Republic.
|
FOOTNOTES |
|---|
+ To whom correspondence should be addressed. Tel: +420 2 24920294; Fax: +420 2 24919392; Email: skmoch@lf1.cuni.cz
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
1 Jaeken, J. and Van den Berghe, G. (1984) An infantile autistic syndrome characterised by the presence of succinylpurines in body fluids. Lancet, ii, 1058–1061.
2 Van den Bergh, F.A., Bosschaart, A.N., Hageman, G., Duran, M. and Tien Poll-The, B. (1998) Adenylosuccinase deficiency with neonatal onset severe epileptic seizures and sudden death. Neuropediatrics, 29, 51–53.[Web of Science][Medline]
3 ebesta, I., Krijt, J., Kmoch, S., Hartmannová, H., Wojda, M. and Zeman, J. (1997) Adenylosuccinase deficiency: clinical and biochemical findings in 5 Czech patients. J. Inherit. Metab. Dis., 20, 343–344.[Web of Science][Medline]
4 Maaswinkel-Mooij, P.D., Laan, L.A., Onkenhout, W., Brouwer, O.F., Jaeken, J. and Poorthuis, B.J. (1997) Adenylosuccinase deficiency presenting with epilepsy in early infancy. J. Inherit. Metab. Dis., 20, 606–607.[Web of Science][Medline]
5 Valik, D., Miner, P.T. and Jones, J.D. (1997) First U.S. case of adenylosuccinate lyase deficiency with severe hypotonia. Pediatr. Neurol., 16, 252–255.[Web of Science][Medline]
6 Salerno, C., Crifo, C. and Giardini, O. (1995) Adenylosuccinase deficiency: a patient with impaired erythrocyte activity and anomalous response to intravenous fructose. J. Inherit. Metab. Dis., 18, 602–608.[Web of Science][Medline]
7 Jaeken. J., Wadman, S.K., Duran, M., van Sprang, F.J., Beemer, F.A., Holl, R.A., Theunissen, P.M., de Cock, P., van denBergh, F., Vincent, M.F. and Van denBerghe, G. (1988) Adenylosuccinase deficiency: an inborn error of purine nucleotide synthesis. Eur. J. Pediatr., 148, 126–131.
8 Van den Berghe, G., Vincent, M.F. and Jaeken, J. (1997) Inborn errors of the purine nucleotide cycle: adenylosuccinase deficiency. J. Inherit. Metab. Dis., 20, 193–202.[Web of Science][Medline]
9 ebesta, I., Shobowale, M., Krijt, J. and Simmonds, H.A. (1995) Screening tests for adenylosuccinase deficiency. Screening, 4, 117–124.
10 Maddocks, J. and Reed, T. (1989) Urine test for adenylosuccinase deficiency in autistic children. Lancet, i, 158–159.
11 Laikind, P.K., Seegmiller, J.E. and Gruber, H.E. (1986) Detection of 5'-phosphoribosyl-4-(N-succinylcarboxamide)-5-aminoimidazole in urine by use of the Bratton-Marshall reaction: identification of patients deficient in adenylosuccinate lyase activity. Anal. Biochem., 156, 81–90.[Web of Science][Medline]
12 Wadman, S.K., Duran, M. and Fabery de Jonge, H. (1986) Diagnosis of inherited adenylosuccinase deficiency by thin-layer chromatography of urinary imidazoles and by automated cation exchange column chromatography of purines. Clin. Chim. Acta, 156, 279–287.[Web of Science][Medline]
13 Van den Bergh, F., Vincent, M.F., Jaeken, J. and Van den Berghe, G. (1991) Radiochemical assay of adenylosuccinase: demonstration of parallel loss of activity toward both adenylosuccinate and succinylaminoimidazole corboxamide ribotide in liver of patients with the enzyme defect. Anal. Biochem., 193, 287–291.[Web of Science][Medline]
14 Van den Bergh, F., Vincent, M.F., Jaeken, J. and Van den Berghe, G. (1993) Residual adenylosuccinase activities in fibroblasts of adenylosuccinase-deficient children: parallel deficiency with adenylosuccinate and succinyl-AICAR in profoundly retarded patients and non-parallel deficiency in a mildly retarded girl. J. Inherit. Metab. Dis., 16, 415–424.[Web of Science][Medline]
15 Van den Berghe, G., Van den Bergh, F., Vincent, M.F. and Jaeken, J. (1993) The biochemical aspects of Asase deficiency. In Gresser, U. (ed.), Molecular Genetics, Biochemistry and Clinical Aspects of Inherited Disorders of Purine and Pyrimidine Metabolism. Springer Verlag, Berlin, pp. 140–143.
16 Van den Bergh, F., Vincent, M.F., Jaeken, J. and Van den Berghe, G. (1993) Functional studies in fibroblasts of adenylosuccinase-deficient children. J. Inherit. Metab. Dis., 16, 425–434.[Web of Science][Medline]
17 Van Keuren, M.L., Hart, I.M., Kao, F.T., Neve, R.L., Bruns. G, A., Kurnit, D.M. and Patterson, D. (1987) A somatic cell hybrid with a single human chromosome 22 corrects the defect in the CHO mutant (Ade-I) lacking adenylosuccinase activity. Cytogenet. Cell Genet., 44, 142–147.[Web of Science][Medline]
18 Stone, R.L., Aimi, J., Barshop, B.A., Jaeken, J., Van den Berghe, G., Zalkin, H. and Dixon, J.E. (1992) A mutation in adenylosuccinate lyase associated with mental retardation and autistic features. Nature Genet., 1, 59–63.[Web of Science][Medline]
19 Fon, E.A., Demczuk, S., Delattre, O., Thomas, G. and Rouleau, G.A. (1993) Mapping of the human adenylosuccinate lyase (ADSL) gene to chromosome 22q13.1q13.2. Cytogenet. Cell Genet., 64, 201–203[Web of Science][Medline]
20 Wong, L.J., and O’Brien, W.E. (1995) Characterization of the cDNA and the gene encoding murine adenylosuccinate lyase. Genomics, 20, 341–343.
21 Aimi, J., Badylak, J., Williams, J., Chen, Z.D., Zalkin, H. and Dixon, J.E. (1990) Cloning of a cDNA encoding adenylosuccinate lyase by functional complementation in Escherichia coli. J. Biol. Chem., 5, 9011–9014.
22 Stone, R.L., Zalkin, H. and Dixon, J.E. (1993) Expression, purification, and kinetic characterization of recombinant human adenylosuccinate lyase. J. Biol. Chem., 268, 19710–19716.
23 Casey, P.J. and Lowenstein, J.M. (1987) Purification of adenylosuccinate lyase from rat skeletal muscle by a novel affinity column. Stabilization of the enzyme, and effects of anions and fluoro analogues of the substrate. Biochem. J., 246, 263–269.[Web of Science][Medline]
24 Verginelli, D., Luckow, B., Crifo, C., Salerno, C. and Gross, M. (1998) Identification of new mutations in the adenylosuccinate lyase gene associated with impaired enzyme activity in lymphocytes and red blood cells. Biochim. Biophys. Acta, 1406, 81–84.[Medline]
25 Marie, S., Cuppens, H., Heuterspreute, M., Jaspers, M., Tola, E.Z., Gu, X.X., Legius, E., Vincent, M.F., Jaeken, J., Cassiman, J.J. and Van den Berghe, G. (1999) Mutation analysis in adenylosuccinate lyase deficiency, eight novel mutations in the re-evaluated full ADSL coding sequence. Hum. Mutat., 13, 197–202.[Web of Science][Medline]
26 Barnes, L.B. and Bishop, S.H. (1975) Adenylosuccinate lyase from human erythrocytes. Int. J. Biochem., 6, 497–503.
27 Casey, P.J. and Lowenstein J.M. (1987) Purification of adenylosuccinate lyase from rat skeletal muscle by a novel affinity collumn. Stabilization of the enzyme, and effects of anions and fluoro analogues of the substrate. Biochem. J., 246, 263–269.
28 Marie, S., Race, V., Nassogne, M.C., Vincent, M.F. and Van den Berghe G. (1999) A mutation in the 5'UTR of the ADSL gene in a patient with adenylosuccinate lyase deficiency. Cell. Mol. Biol. Lett., 3, 9.
29 Lee, T.T., Worby, C., Bao, Z.Q., Dixon, J.E. and Colman, R.F. (1999) His68 and His141 are critical contributors to the intersubunit catalytic site of adenylosuccinate lyase of Bacillus subtilis. Biochemistry, 38, 22–32.[Medline]
30 De Volder, A.G., Jaeken, J., Van den Berghe, G., Bol, A., Michel, C., Cogneau, M. and Goffinet, A.M. (1988) Regional brain glucose utilization in adenylosuccinase-deficient patients measured by positron emission tomography. Pediatr. Res., 24, 238–242.[Web of Science][Medline]
31 Stone, T.W., Roberts, L.A., Morris, B.J., Duley, J.A. and Ogilvy, H.V. (1997) Are succinylpurines neurotoxic? Clin. Biochem., 3, P41, 169.
32 Chomczynski, P. and Sacchi, N. (1987) Single-step method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroform extraction. Anal. Biochem., 162, 156–159.[Web of Science][Medline]
33 Gregersen, N., Blakemore, A.I., Winter, V., Andresen, B., Kolvraa, S., Bolund, L., Curtis, D. and Engel, P.C. (1991) Specific diagnosis of medium-chain acyl-CoA dehydrogenase (MCAD) deficiency in dried blood spots by a polymerase chain reaction (PCR) assay detecting a point-mutation (G985) in the MCAD gene. Clin. Chim. Acta, 203, 23–34.[Web of Science][Medline]
34 Favre, D. (1992) Improved phenol-based method for the isolation of DNA fragments from low melting temperature agarose gels. Biotechniques, 13, 25–26.
35 Thompson, J.D., Higgins, D.G. and Gibson, T.J. (1994) CLUSTAL W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-specific gap penalties and weight matrix choice. Nucleic Acids Res., 22, 4673–4680.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  S. Sivendran, D. Patterson, E. Spiegel, I. McGown, D. Cowley, and R. F. Colman Two Novel Mutant Human Adenylosuccinate Lyases (ASLs) Associated with Autism and Characterization of the Equivalent Mutant Bacillus subtilis ASL J. Biol. Chem., December 17, 2004; 279(51): 53789 - 53797. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 V. Race, S. Marie, M.-F. Vincent, and G. Van den Berghe Clinical, biochemical and molecular genetic correlations in adenylosuccinate lyase deficiency Hum. Mol. Genet., September 1, 2000; 9(14): 2159 - 2165. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||