Aspartylglycosaminuria: biochemistry and molecular biology

Aspartylglycosaminuria: biochemistry and molecular biology

Biochimica et Biophysica Acta 1455 (1999) 139^154 www.elsevier.com/locate/bba Review Aspartylglycosaminuria: biochemistry and molecular biology Nath...

547KB Sizes 0 Downloads 12 Views

Biochimica et Biophysica Acta 1455 (1999) 139^154 www.elsevier.com/locate/bba

Review

Aspartylglycosaminuria: biochemistry and molecular biology Nathan N. Aronson Jr. * 2146 MSB, Department of Biochemistry and Molecular Biology, University of South Alabama, 307 University Blvd. Mobile, AL 36688-0002 USA Received 19 November 1998; received in revised form 20 May 1999; accepted 20 May 1999

Abstract Aspartylglucosaminuria (AGU, McKusick 208400) is an autosomal recessive lysosomal storage disease caused by defective degradation of Asn-linked glycoproteins. AGU mutations occur in the gene (AGA) for glycosylasparaginase, the enzyme necessary for hydrolysis of the protein^oligosaccharide linkage in Asn-linked glycoprotein substrates undergoing metabolic turnover. Loss of glycosylasparaginase activity leads to accumulation of the linkage unit Asn^GlcNAc in tissue lysosomes. Storage of this fragment affects the pathophysiology of neuronal cells most severely. The patients notably suffer from decreased cognitive abilities, skeletal abnormalities and facial grotesqueness. The progress of the disease is slower than in many other lysosomal storage diseases. The patients appear normal during infancy and generally live from 25 to 45 years. A specific AGU mutation is concentrated in the Finnish population with over 200 patients. The carrier frequency in Finland has been estimated to be in the range of 2.5^3% of the population. So far there are 20 other rare family AGU alleles that have been characterized at the molecular level in the world's population. Recently, two knockout mouse models for AGU have been developed. In addition, the crystal structure of human leukocyte glycosylasparaginase has been determined and the protein has a unique KLLK sandwich fold shared by a newly recognized family of important enzymes called N-terminal nucleophile (Ntn) hydrolases. The nascent single-chain precursor of glycosylase araginase self-cleaves into its mature K- and L-subunits, a reaction required to activate the enzyme. This interesting biochemical feature is also shared by most of the Ntnhydrolase family of proteins. Many of the disease-causing mutations prevent proper folding and subsequent activation of the glycosylasparaginase. ß 1999 Published by Elsevier Science B.V. All rights reserved. Keywords: Aspartylglycosaminuria ; Aspartylglucosaminuria ; Glycosylasparaginase; Glycoprotein catabolism; Lysosomal storage disease; N-Terminal nucleophile amidase

Contents 1.

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

140

2.

Lysosomal degradation of ASN-linked glycoproteins: 2.1. Degradation of oligosaccharide component . . . . 2.2. Degradation of protein and linkage region . . . . . 2.3. Role of glycosylasparaginase . . . . . . . . . . . . . . .

141 142 142 142

general ...... ...... ......

features . ....... ....... .......

. . . .

. . . .

. . . .

. . . .

. . . .

* Fax: +1-334-460-6127; E-mail: [email protected] 0925-4439 / 99 / $ ^ see front matter ß 1999 Published by Elsevier Science B.V. All rights reserved. PII: S 0 9 2 5 - 4 4 3 9 ( 9 9 ) 0 0 0 7 6 - 9

BBADIS 61881 10-9-99

Cyaan Magenta Geel

Zwart

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

140

N.N. Aronson Jr. / Biochimica et Biophysica Acta 1455 (1999) 139^154 3.

Molecular biology . . . . . . . . . . . . . . . . . . . 3.1. cDNA . . . . . . . . . . . . . . . . . . . . . . . . 3.2. Gene structure and AGU mouse model 3.3. 3P-Untranslated region . . . . . . . . . . . .

. . . .

143 143 144 145

4.

Evolution of glycosylasparaginase . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

145

5.

Gene defects causing aspartylglycosaminuria . 5.1. Finnish mutations . . . . . . . . . . . . . . . . . 5.2. Splicing mutations . . . . . . . . . . . . . . . . . 5.3. Missense mutations . . . . . . . . . . . . . . . . 5.4. Nonsense mutations . . . . . . . . . . . . . . . 5.5. Insertions and deletions . . . . . . . . . . . . . 5.6. Genomic rearrangement . . . . . . . . . . . . 5.7. Discussion . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . .

146 146 148 149 150 150 150 150

6.

Glycosylasparaginase as a member of the N-terminal nucleophile (Ntn) hydrolase family of proteins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

151

7.

Future directions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

151

Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

152

References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

152

1. Introduction In 1967, Mahadavan and Tappel [1] puri¢ed and characterized rat liver glycosylasparaginase (EC 3.5.1.26) which is the enzyme responsible for hydrolysis of the glycosylamide bond that joins carbohydrate chains to Asn-linked glycoproteins. They determined the enzyme is an amidase that functions in the lysosomes. In the very same year that the rat enzyme was described, Jenner and Pollitt [3] published a short proceedings report that large quantities of Asn^GlcNAc, the linkage unit of Asn-linked glycoproteins, occurred in the urine of two mentally retarded siblings in England. They then noted that the activity of glycosylasparaginase was absent from the tissues of the two patients [4]. This de¢ciency of the enzyme suggested that they su¡ered from an inborn error of metabolism which they named `aspartylglycosaminuria'. During this same period while working on his Ph.D. thesis in Finland, Jorma Palo observed among a very large number of retarded persons being clinically screened in his country that nine of them excreted an unknown peptide material in their urine [5]. Based on the Jenner and Pollitt publica-

BBADIS 61881 10-9-99

. . . .

. . . .

. . . . . . . .

. . . .

. . . . . . . .

. . . .

. . . . . . . .

. . . .

. . . . . . . .

. . . .

. . . . . . . .

. . . .

. . . . . . . .

. . . .

. . . . . . . .

. . . .

. . . . . . . .

. . . .

. . . . . . . .

. . . .

. . . . . . . .

. . . .

. . . . . . . .

. . . .

. . . . . . . .

. . . .

. . . . . . . .

. . . .

. . . . . . . .

. . . .

. . . . . . . .

. . . .

. . . . . . . .

. . . .

. . . . . . . .

. . . .

. . . . . . . .

. . . .

. . . . . . . .

. . . .

. . . . . . . .

. . . .

. . . . . . . .

. . . .

. . . . . . . .

. . . .

. . . . . . . .

. . . .

. . . . . . . .

. . . .

. . . . . . . .

. . . .

. . . . . . . .

. . . .

. . . . . . . .

. . . .

. . . . . . . .

. . . .

. . . . . . . .

. . . .

. . . . . . . .

. . . .

. . . . . . . .

. . . .

. . . . . . . .

. . . .

. . . . . . . .

tions the unknown compound was identi¢ed to be Asn^GlcNAc [6]. Palo coined the term `AGU' for the disease (probably derived from PKU already being a popularized term for phenylketonuria), and soon it became evident that an AGU gene defect had accumulated in the Finnish population [7]. Aspartylglucosaminuria (AGU, McKusick 208400), although rare outside of the Finnish population, is the most common lysosomal storage disorder that directly involves glycoprotein catabolism. There are over 200 patients in Finland, and approximately one new case in 18 500 live births occurs there [8]. The clinical hallmarks of the disease are: slowly developing, progressive psychomotor retardation; facial grotesqueness; and skeletal and connective tissue abnormalities [9,10]. AGU children appear normal for several years, but as juveniles and throughout adolescence the major traits of mental retardation and facial coarseness progressively worsen. Thus, to a certain extent, AGU behaves as a juvenile, or even adult, form of lysosomal storage disease in which the pathology is not immediately apparent at birth or during infancy [11]. Some common diagnostic changes in the young patients are

Cyaan Magenta Geel

Zwart

N.N. Aronson Jr. / Biochimica et Biophysica Acta 1455 (1999) 139^154

141

aggressive behavior and delayed and diminished communication skills. AGU patients can have relatively long lives compared to patients with other forms of lysosomal storage diseases, surviving up to 45^50 years [12,68]. As in all types of lysosomal storage diseases, AGU tissues are characterized by enlarged lysosomes. The principal storage material as originally determined by Jenner and Pollitt [3] is Asn^GlcNAc, but small amounts of other glycoasparagines with unexplained saccharide structures are also sometimes noted [13,14,75]. Recently the crystal structure of the human leukocyte enzyme was obtained by Oinonen et al. [2], and the mechanism of the reaction determined. This review will therefore focus on our current knowledge about the biochemistry and structural biology of glycosylasparaginase and discuss how this information can now explain the manner by which speci¢c AGU mutations e¡ect the disease. A recent book [87] and other important reviews covering various aspects of AGU and glycosylasparaginase have been published [68,69,79,88^ 92]. 2. Lysosomal degradation of ASN-linked glycoproteins: general features Degradation of Asn-linked glycoproteins completely to amino acids and sugars occurs in lysosomes in a two-part pathway (Fig. 1) [15]: (1) sequential release of non-reducing-end sugars by exoglycosidases that accounts for the bulk of carbohydrate removal; and (2) digestion of the protein and protein-to-carbohydrate linkage region. Serum K1 acid glycoprotein (orosomucoid) was used as a model Asn-linked glycoprotein substrate to determine the lysosomal digestive pathway (Fig. 1A). These studies were done with the rat and human glycoprotein, or their asialo-derivatives (minus terminal sialic acid residues), and in vitro with extracts of puri¢ed lysosomes [16] or in situ using a perfused rat liver. The organ system was especially useful since radioactively labeled asialo-orosomucoid could be metabolically `microinjected' into hepatocyte lysosomes via the well-characterized asialoglycoprotein receptor that is unique to these cells [17^20]. Rat K1 -acid glycoprotein contains 187 amino acids in its polypeptide and six attached N-linked oligosaccharides [21] (Fig. 1A).

BBADIS 61881 10-9-99

Fig. 1. Lysosomal degradation of Asn-linked glycoproteins. (A) Structure of rat serum orosomucoid and its complete hydrolysis into monosaccharides and amino acids. Rat orosomucoid is a 187-amino acid single polypeptide with six biantennary chains, one triantennary chain, and one fucose that account for its total 70 monosaccharides. Linkage positions of the sugars are indicated: O, sialic acid; R, L-D-galactose; E, N-acetyl-LD-glucosamine; b, K-D-mannose; a, L-D-mannose; F, K-L-fucose. (B) Bidirectional hydrolysis pathway in lysosomes. Steps 1^5: monosaccharides are cleaved sequentially from the non-reducing end of the oligosaccharides. Steps I^IV : ordered hydrolysis of protein and linkage region. Hydrolysis of the reducingend GlcNAc can be catalyzed from either of the two reaction pathways (Step IV or 3).

One sugar chain contains a single-branched L-fucose residue attached K-(1^6) to the core reducing-end GlcNAc that is joined to Asn. One chain is a triantennary structure, while the other ¢ve are biantennary. A total of 256 H2 O molecules are required for lysosomal hydrolases to catalyze complete disassembly of this glycoprotein into its 187 amino acids

Cyaan Magenta Geel

Zwart

142

N.N. Aronson Jr. / Biochimica et Biophysica Acta 1455 (1999) 139^154

and 70 monosaccharides. This metabolism only requires a set of six or seven glycosidases, a relatively small group of proteinases (cathepsins A^D, H, L and S) and a speci¢c glycosylasparaginase for the protein-to-carbohydrate linkage, which is the topic of this review (Fig. 1B). In metabolic design, the overall simultaneous bidirectional disassembly is similar to other biochemical pathways, incorporating usual features, such as ordered reaction steps and a requirement for relatively few enzymes [15]. 2.1. Degradation of oligosaccharide component The stepwise degradation of the oligosaccharide component results directly from the substrate specificity of individual lysosomal glycosidases. Most of them are exo-hydrolases that only recognize a nonreducing end glycosyl unit as a substrate. This enzymatic feature determines that lysosomal K-neuraminidase, L-D-galactosidase, N-acetyl-L-D-hexosaminidase and K- and L-D-mannosidases perform their hydrolytic reactions in the same sequential order as the sugars structurally occur along the bi- or triantennary chains (steps 1^5, Fig. 1B). A variation in this reaction path is the possibility for separate K-Dmannosidases to release the 1^3- and 1^6-K-D-mannosyl units joined to the L-D-mannosyl branch point of biantennany chains. A major undigested oligosaccharide fragment in human K-mannosidosis patients contains only the K(1C3)-D-mannose branch, which suggests that a speci¢c K(1C6)-D-mannosidase is still active in their lysosomes [22]. Such an K(1C6)-D-mannosidase has been isolated and characterized [23]. This enzyme appears to require the prior removal of Asn by glycosylasparaginase in order to hydrolyze the 1^6-linked mannose unit [24]. Some Man2 GlcNAc2 Asn fragments having the 1^6mannose were recently found in the liver of an AGU patient [25]. The fact that this fragment is rarely observed in AGU tissues is due to the ability of the major lysosomal K-D-mannosidase to hydrolyze an K-(1^6)-D-mannose [23]. Sequential L-D-mannosidase and N-acetyl-L-D-glucosaminidase attack would then yield the ¢nal Asn^GlcNAc fragment. 2.2. Degradation of protein and linkage region The second part of the bidirectional pathway for

BBADIS 61881 10-9-99

Asn-linked glycoprotein catabolism encompasses digestion of the protein component, the actual linkagestructure Asn^GlcNAc, and the di-N-acetylchitobiose core unit (reaction steps I^IV, Fig. 1B). This portion of degradation is interesting because of the unique enzymes involved, including glycosylasparaginase. The order for degradation of the protein and protein-to-carbohydrate components of Asn-linked glycoproteins is: (I) proteolysis; (II) removal of any fucose; (III) hydrolysis of the L-amide of the Asn^ GlcNAc linkage; and (IV) release of the reducingend GlcNAc from the di-N-acetylchitobiose core structure. The latter step can also be catalyzed by N-acetyl-L-D-glucosaminidase acting from the nonreducing end of the carbohydrate component (glycosidase 3, Fig. 1B). This occurs when the rate of oligosaccharide digestion is fast enough to expose the penultimate GlcNAc prior to the glycosylasparaginase reaction. It is noteworthy that extensive proteolysis must occur ¢rst in order for overall digestion of the linkage region to go to completion. 2.3. Role of glycosylasparaginase Two major biochemical features of the glycosylasparaginase play an important role in de¢ning the order of the linkage-region reactions. A noteworthy speci¢city of this amidase is its requirement for both a free K-amino and K-carboxyl group on the Asn substrate [26,27]. This enzymic property demands that there be prior hydrolysis of both peptide bonds joined to the Asn. Lysosomal cathepsins must therefore release the Asn with its associated oligosaccharide (step I, Fig. 1B) before glycosylasparaginase can react in the pathway. The second property of glycosylasparaginase reactivity that impinges on the degradation process is its inhibition by any K-(1^6)-L-fucose that is bound to the core reducing-end GlcNAc [28,29]. The deoxysugar branch must sterically block the active site of the enzyme from binding the asparaginyl moiety of the substrate. This inhibitory feature explains why the majority of fragments that accumulate in fucosidosis patients are glycopeptides that retain the Asn residue [30], and it determines that removal of any fucose is a prerequisite for glycosylasparaginase to react. Thus, for each Asn-linked glycoprotein that enters the lysosomes for turnover, digestion occurs simulta-

Cyaan Magenta Geel

Zwart

N.N. Aronson Jr. / Biochimica et Biophysica Acta 1455 (1999) 139^154

143

Fig. 2. Human placenta glycosylasparaginase cDNA and AGU mutations. The signal peptide, catalytic nucleophile T206 and exon^intron junctions are in shaded boxes. Two glycosylation sites (N38 and N308) are enclosed by hexagons. Disul¢de bonds C64^C69; C163^C179; C286^C306; and C317^C345 are shown. The sequence G186^D205 (underlined, italicized) is cleaved from the C-terminus of the K-subunit in the lysosomes. Amino acid changes caused by nine missense AGU mutations are shown at arrowheads with the mutated base underlined. There is one nonsense AGU mutation (C64 STOP) and seven AGU-causing insertions or deletions (the latter two classes are indicated by underlined bases). The mutations are further described in Figs. 3 and 6 and in the text.

neously from the outer ends of oligosaccharide chains and along the polypeptide. The common end-point in catabolism involves the linkage region where glycosylasparaginase reacts. The relative rates at which the two parts of the overall degradative system operate will vary depending on the detailed chemical structure of a particular glycoprotein. For example, certain glycosylated asparagines may be in more protected environments than others, and the ability of the lysosomal milieu to denature each glycoprotein would contribute to the disassembly of each component.

BBADIS 61881 10-9-99

3. Molecular biology 3.1. cDNA A full-length cDNA of human glycosylasparaginase was ¢rst cloned in 1990 (Fig. 2) [31]. The 1041-bp open-reading frame1 encoded a protein of

1 The protein sequence is numbered from the start methionine being at position 1 and the nucleotide sequence is numbered from the start ATG codon with A being at position +1.

Cyaan Magenta Geel

Zwart

144

N.N. Aronson Jr. / Biochimica et Biophysica Acta 1455 (1999) 139^154

Fig. 3. Structures of human glycosylasparaginase gene (AGA) and protein with locations of AGU mutations. The glycosylasparaginase gene is 13 kb [36] located on chromosome 4q32^q33 [34,35]. Its nine exons are separated by introns whose sizes in kb are shown in italics. The gene encodes a 346-amino acid single-chain polypeptide [31] that contains a 23-amino acid signal peptide (shaded). Selfcleavage of the nascent single-chain precursor occurs between residues D205 and T206 (at heavy dotted line in exon 5-encoded region) to form the active K/L-glycosylasparaginase [38^40]. The mature enzyme loses 20 amino acids from the C-terminus of the K-subunit in the lysosomes. Relative locations of 21 AGU-causing mutations (a^u) are indicated in both the gene and protein and are described in Table 1 and the text. Exon and intron sequences can be found in GenBank: U21273^U21281.

346 amino acids of which the N-terminal 23 residues form a signal peptide. The signal-free human protein (residues Ser24^Ile346) has a calculated mass of 34.6 kDa. This polypeptide encompasses both the K- and L-subunit of glycosylasparaginase, and it immediately became evident that post-translational cleavage of a single-chain precursor gives rise to the mature two-subunit amidase [31,32]. The calculated mass of the encoded human K-subunit polypeptide (Ser24^ Asp205) is 19.5 kDa and that of the L-subunit (Thr206^Ile346) is 15.1 kDa. The molecular masses of the human K and L-subunits observed by SDSPAGE are 18 and 23 kDa, respectively [31]. Each subunit contains one Asn-linked glycosylation sequon (K, 38AsnAlaThr; L, 308AsnValThr), and each of these two Asn residues is glycosylated in the mature enzyme [33]. These oligosaccharides would be expected to add a mass of about 2^3 kDa to the enzyme. Four disul¢de bonds occur in the protein, two in each subunit: Cys64^Cys69; Cys163^Cys179; Cys286^Cys306; Cys317^Cys345 [48]. 3.2. Gene structure and AGU mouse model The human gene for glycosylasparaginase has been localized to chromosome region 4q32^q33 [34,35]. This 13-kb gene (Fig. 3) has nine exons with all exon^intron junctions ¢tting the gt/ag consensus

BBADIS 61881 10-9-99

splicing rule [36]. Exon 1 encodes 5P-untranslated sequence and the ¢rst 42 amino acids which include the 23 amino acid signal peptide. Exon 9 codes the last 33 amino acids and 939 bases of 3P-untranslated sequence. The major transcriptional start site in the 5P-untranslated region has been experimentally determined to be at 3298 (relative to the ATG start codon) [37]. As observed in genes for many lysosomal hydrolases, the 5P-promoter region of glycosylasparaginase is typical of housekeeping genes and includes a GC-rich island. Promoter studies on the 5P-£anking region of the human glycosylasparaginase gene suggest that Sp1 and Ap-2 binding sites and a possible inhibitory control region may be important for regulating its expression [37]. The only other gene structure for glycosylasparaginase besides that in humans was reported for the mouse [38,39]. The mouse gene spans 11 kb and also contains nine exons. All of the exon^intron boundaries of the mouse and human genes are equivalently positioned. Two mRNA transcripts are produced from both the mouse and human genes. Two experimental mouse models for AGU have now been generated by targeted disruption of the mouse glycosylasparaginase gene [38,39]. The AGU mice exhibit very similar pathophysiology as the human patients [93^96], including widespread hypertrophy of the lysosomes in the central nervous system that leads to a gradual, progressive motor impairment and ¢nally

Cyaan Magenta Geel

Zwart

N.N. Aronson Jr. / Biochimica et Biophysica Acta 1455 (1999) 139^154

145

premature death. Older AGU mice display severe ataxia as seen in dramatic changes in their walking pattern which seems to coincide with a gross loss of Purkinje cells [94]. Promising partial correction of the pathology has been observed in early adenovirus-mediated gene therapy experiments in which both tail vein and intraventricular injections of the virus-encoded human glycosylasparaginase were tested [97]. A normal enzyme activity level returned to the liver of the AGU animals treated by the tailvein injection route. In the brain-treated animals, some transduced ependymal cells lining the ventricles appeared also to replace the enzyme in surrounding areas of neurons. In both types of gene therapy, the corrective e¡ects lasted for at least a month. 3.3. 3P-Untranslated region Northern blot analysis indicated two forms of glycosylasparaginase mRNA were in human ¢broblasts and their sizes were estimated to be 1.4 and 2.2 kb [37,40]. The genomic structure showed that the 3Puntranslated region in exon 9 is 939 bp. Three polyadenylation signals (AATAAA) are located at bases 1152, 1164 and 1949 of the cDNA. Based on the 1.4 and 2.2 kb sizes of the two mRNAs seen in ¢broblasts and the initiation of transcription at 3298, both the repeated poly A signals at 1152/1164 and the one downstream at 1949 appear to be used. This predicted termination of the mRNAs is substantiated in the current databases by an approximately equal number of human glycosylasparaginase ESTs that extend to these two 3P-sites. It is unlikely that an intron is in the 3P-untranslated region since a single cDNA from liver spans this exact genomic sequence. 4. Evolution of glycosylasparaginase Glycosylasparaginase cDNAs have been determined for six species: human [31], rat [41], bovine [41], bacteria (Flavobacterium meningosepticum) [42], insect (Spodoptera frugiperda) [41] and mouse [43]. In addition, a plant asparaginase also appears to have an evolutionary relationship to mammalian glycosylasparaginases [44]. The high homology of the deduced amino acid sequences of these forms of the enzyme was useful for gaining insight into what res-

BBADIS 61881 10-9-99

Fig. 4. Strand diagram showing three-dimensional structure of human leukocyte glycosylasparaginase. The enzyme is a member of a self-processing N-terminal nucleophile amidohydrolase family of proteins [2,78,79]. Autoproteolysis forms the two subunits (K, green; L, orange) from a single-chain precursor [31,64,85]. The unique protein fold of this group consists of an `KLLK sandwich' [2], i.e. K-helices layer I, L-sheets II and III, and K-helices layer IV. The N-terminal nucleophile residue Thr206 that catalyzes hydrolysis of the Asn-GlcNAc substrate and the product Asp are shown in the active-site cleft formed between the two L-sheets. This is a RasMol presentation of ¢le 1APZ [2] from the Brookhaven Protein Data Bank.

idues are important for its function. With the exception of the bacterial and insect enzymes, all cysteines involved in disul¢des are conserved as well as both glycosylation sites [41]. The glycosylation site on the L-subunit is missing from the insect hydrolase, and one of the two potential disul¢de bonds is absent from the same insect subunit due to lack of appropriate cysteines. However, the most C-terminal potential disul¢de that is important in the overall folding/structure of the human enzymes is maintained by insect glycosylasparaginase. There are no disul¢de

Cyaan Magenta Geel

Zwart

146

N.N. Aronson Jr. / Biochimica et Biophysica Acta 1455 (1999) 139^154

Fig. 5. Mechanism of reaction catalyzed by human glycosylasparaginase. Important active site residues [2] are noted in rectangles with their important side chains indicated. (The protein sequence is numbered from the start methionine being at position 1.) The catalytic cycle beginning at the upper left consists of: (a) activation of the hydroxyl of the N-terminal T206 of the L-subunit by its own K-amino group; and nucleophilic attack at the amide carbon of Asn-GlcNAc; (b) stabilization of the tetrahedral intermediate in an oxyanion hole formed by T257 and G258; (c) formation of the L-aspartyl ester through general acid catalysis; (d) deacylation of the enzyme; and (e,f) release of aspartate by this same general base/acid mechanism. Requirement for free K-amino and K-carboxyl groups on the substrates are due to speci¢c binding of these substrate moieties by D237 and R234, respectively. The initially released 1-amino N-acetylglucosamine product (c) is non-enzymatically hydrolyzed to ammonia and GlcNAc.

bonds in the bacterial enzyme [45,46]. Essentially all amino acids that are observed at the active site in the crystal of the human enzyme (Figs. 4 and 5) are conserved in all known glycosylasparaginases [2,45,46]. Another major di¡erence in the bacterial amidase is the absence of a stretch of about 30 amino acids from the C-terminus of its K-subunit [45,46]. However, most of this region is cleaved from human glycosylasparaginase by proteases in the lysosomes [32], and the important His204^Asp205 unit at the self-cleavage site between K- and L-subunits is conserved in the bacterial enzyme. There are several clusters of highly conserved residues in both the Ksubunit (Ala53^Glu58; Val73^Asp80; Thr86^Asp93) and L-subunit (Gly233^Gly244; Gly252^Asp261), whose functions in the protein are not yet understood. 5. Gene defects causing aspartylglycosaminuria Besides the common mutation that has caused AGU in Finland [40,65,68], a second minor Finnish allele [52] and 19 other individual family AGU cases have been noted throughout the world's population

BBADIS 61881 10-9-99

(Table 1). Most of the non-Finnish patients are homozygotes due to having consanguineously related parents. However, recent discovery of four unrelated Arab families with the same AGU-causing mutation [47] suggests that a founder e¡ect may be responsible for the storage disease in a second region of the world besides Finland. The molecular basis of causes and consequences of some of these other mutations are now understood based on both the AGA gene structure and crystal structure of the enzyme (Figs. 3 and 6). 5.1. Finnish mutations The ¢rst AGU mutations to be characterized at the DNA level were the Finnish G482A and G488C base changes which resulted in two amino acid substitutions in the glycosylasparaginase, Arg161Gln and Cys163Ser (Figs. 2, 3 and 6, mutation n) [40,65,66]. The transition G482A (Arg161Gln) was shown to be a neutral polymorphism that does not a¡ect enzyme activity [40]. The G488C transition (Cys163Ser) will by itself cause loss of glycosylasparaginase [40] and therefore is responsible for the major form of Finnish AGU disease. Cys163 forms a

Cyaan Magenta Geel

Zwart

N.N. Aronson Jr. / Biochimica et Biophysica Acta 1455 (1999) 139^154

147

Table 1 AGU mutations Mutation (Type)

Gene alteration

Consequencesa

a (deletion)

37 bp: 99(TTGGCCC)105 or 100(TGGCCCCT)106 or 101(GGCCCTT)107 (exon 1)

Frameshift and premature STOP (23 aa); unstable mRNA

b (splicing)

gt to at (intron 1 splice donor); alternative splicing 6 bp downstream from exon 1; +6 bp in mRNA; 127(+ATGCGG)128

Insertion (+2 aa): 42(AspAla)43 (AH1)b

c (missense)

G179A (exon 2)

Gly60Asp (AH2)b

d (missense)

T183G (exon 2)

Cys61Trp (AH2)b

e (nonsense)

T192A (exon 2)

Cys64(STOP): trace amount of single chain form

f (deletion)

32 bp: 193GA194 or 195GA196 or 197GA198 or 199GA200

Frameshift and premature STOP (45 aa)

g (missense)

T214C (exon 2)

Ser72Pro (active site)

h (missense)

G299A (exon 3)

Gly100Glu(AS3)b

i (missense)

C302T (exon 3)

Ala101Val: (AS3)b single-chain; failed transport to lysosomes

j (deletion)

31 bp: T335 (exon 3)

Frameshift and premature STOP (103 aa)

k (deletion)

35 bp: 367(ACACA)371 (exon 3)

Frameshift and premature STOP (118 aa); unstable mRNA

l (splicing)

a to g (intron 3; 7 bp upstream from exon 4); new favorable ag splice acceptor site; +7 bp in mRNA

Frameshift and premature STOP (146 aa)

m (missense)

T404C (exon 4)

Phe135Ser (AH4)b

n (missense)

G488C (exon 4)

Cys163Ser; unable to form Cys163^Cys179 disful¢de; single-chain; failed transport to lysosomes

o (deletion)

31 bp: G567 (exon 5)

Frameshift

p (deletion)

31 bp: T788 or T789 (exon 7)

Frameshift and premature STOP (239 aa)

q (insertion)

+1 bp: 800(+T)801 (exon 7)

Frameshift and premature STOP (295 aa)

r (missense)

G904A (exon 8)

Gly302Arg (BS5)b

s (missense)

T916C (exon 8)

Cys306Arg: (BS5)b unable to form Cys286^Cys306 disul¢de

t (splicing)

gt to tt (intron 8 splice donor site; exon skipping): -exon 8

Single-chain; shortened and not transported to lysosomes

u (rearrangement)

32 kb: deletion starting 1268 bp from the 5P-end of intron 8 and ending 1830 bp downstream from the STOP codon; triplication of T19 at the junction

Chimeric mRNA with a novel exon 9; normal level of mRNA, polyadenylated; single-chain form, rapid degradation of mutant protein

a b

Sites of mutations (a^r) in the gene and enzyme are noted in Figs. 2, 3 and 6. References are in the text. Location of mutation in secondary structure of glycosylasparaginase (see Fig. 6).

BBADIS 61881 10-9-99

Cyaan Magenta Geel

Zwart

148

N.N. Aronson Jr. / Biochimica et Biophysica Acta 1455 (1999) 139^154

Fig. 6. Secondary structural elements of glycosylasparaginase and locations of AGU-causing missense mutations. The crystal structure of human leukocyte glycosylasparaginase [2] is shown schematically according to the major folding of the protein, i.e. two layers of L-sheet (II, III), each £anked on one side by a layer of K-helices (I, IV, respectively). Sheet II consists of eight strands (AS1^AS4; BS1, BS2 and BS7, BS8) and is £anked by ¢ve K-helices (AH1^AH5); sheet III consists of 4 strands (BS3^ BS6) and is £anked by three K-helices (BH1^BH3). Designation of these secondary structures are A or B = K- or L-subunit; H or S = K-helix or L-strand. Locations of AGU missense mutations are indicated by diamonds (mutation b is a splice site mutation resulting in the insertion of two amino acids, see Table 1).

critical disul¢de bond with Cys179 [48,49] and its genetic substitution by Ser results in a misfolding of the protein. The mutant polypeptide remains as a single-chain form rather than undergoing normal post-translational cleavage and is degraded rapidly in the ER. Syva«nen et al. [50] found that 98% of 115 Finnish AGU patients were homozygous for G488C and G482A, indicating that the major AGUFIN mutation is a result of a founder e¡ect. Seven AGU patients living in northern Norway were also homozygous for the major Finnish mutation [51] and had ancestors who immigrated as early as 1700 from the southeastern parts of Finland, or partly from Russia. A more precise understanding of the historical origin of the major Finnish mutation is yet to be determined [52]. The second AGU-causing allele in heterozygous combination with the Cys163Ser mutation has been found in just seven Finnish patients. This minor Finnish allele is a 2-bp deletion in exon 2 of the gene [52]. One of four GA repeats starting at position 193

BBADIS 61881 10-9-99

of the cDNA is missing in these AGU patients (Figs. 2 and 3, mutation f) and results in a premature termination of the translation of glycosylasparaginase mRNA [52]. This region of the glycosylasparaginase gene has several repeating or symmetrical sequences that may contribute to the accumulation of ¢ve di¡erent AGU mutations clustered in a 50base sequence (nucleotides 175^225) within exon 2. The mRNA level of this -(GA) allele in these compound heterozygous patients was greatly reduced to 3^11% of the normal glycosylasparaginase message. This second mutation contributes to approximately 1.5% of Finnish AGU cases, with a deduced carrier frequency of 1:2600. Historically, it appears to be a relatively new mutation that migrated to Finland from southern and eastern directions according to the restricted geographical distribution of the patient families along the southern coast of the country. 5.2. Splicing mutations The ¢rst non-Finnish AGU mutation to be characterized at the molecular level was in a 12-year-old African-American male [53^55,67]. The 3P-terminal portion of the glycosylasparaginase mRNA from this patient lacked exon 8 (Table 1, Fig. 3, mutation t). The exon skipping was due to mutation of gt to tt at the 5P-splice donor site of intron 8 (Table 1), which must result in the ag acceptor site of intron 8 dominating that of intron 7. Deletion of exon 8 causes a frameshift and premature termination of translation. The inactive mutant enzyme was not processed into two subunits, similar to the consequences of the Finnish AGU major mutation. This same type of mutation, i.e. loss of exon 8, appears in one of the two AGU knockout mouse models produced experimentally by targeted gene disruption [39]. A second human AGU splicing defect has caused a seven-base insertion at the junction of exons 3 and 4 in the cDNA of two Japanese siblings [12,56]. The insertion was due to an a to g transition at the eighth nucleotide preceding the 3P-end of intron 3, which creates a new consensus splicing acceptor site ag (Table 1). Exclusive alternative splicing at the new ag site produced abnormal mRNA with seven nucleotides of previous intron 3 added to the 5P-end of exon 4. A frameshift results from the seven-

Cyaan Magenta Geel

Zwart

N.N. Aronson Jr. / Biochimica et Biophysica Acta 1455 (1999) 139^154

base insertion that causes translation to be terminated prematurely (Fig. 3, mutation l). A third splicing mutation occurred in a Tunisian AGU patient [55] whose cDNA showed a 6-bp insertion at the junction of exons 1 and 2 (Table 1, Fig. 3, mutation b). This insertion can be explained if the gt splice donor site of intron 1 were mutated to at. If this g to a transition then leads to splicing from the next gt in intron 1, the predicted change is: GCA G gtgcgggtCGCA GAT GCG G gt (Table 1). This putative mechanism would account for the observed cDNA sequence of the patient with its insertion of the hexanucleotide ATGCGG. This in-frame two-codon insertion would result in the addition of two amino acids (AspAla). These two extra amino acids in the Tunisian patient are in a helix region (AH1) of the enzyme [2] (Figs. 3 and 6, mutation b), and they would likely disturb the normal three-dimensional structure. 5.3. Missense mutations Two AGU brothers in a British family were characterized as compound heterozygotes [57]. One paternally transmitted mutation was a C302T transition that changed alanine at position 101 to valine (Figs. 2, 3 and 6, mutation i). Transient expression of the mutant enzyme in transfected COS-1 cells showed a similar phenotype to the enzyme formed in the major Finnish AGU patients; the protein remained a single-chain form without enzymatic activity, and it was not targeted to lysosomes. Ala101 is located in a L-strand of the K-subunit (AS3) inside of the core of the enzyme (Fig. 6, mutation i), and change to the larger Val should disturb the packing of the polypeptide. This same paternal C302T mutation occurred in an Italian patient [55]. Missense mutations in two Canadian compound heterozygous AGU patients were recently characterized. The observed G299A transition in exon 3 of one patient causes a Gly100Glu substitution (mutation h, Figs. 2, 3 and 6), and a T404C transition in exon 4 of the second patient causes a F135S substitution (mutation m, Figs. 2, 3 and 6). These substitutions lead to major changes in the side chain chemistry at these two positions. Disruption of overall protein folding likely occurs, since the new amino acids are in important secondary structures of the K-subunit: AS3 of L-

BBADIS 61881 10-9-99

149

sheet II and AH4 of K-helices layer I, respectively (Fig. 6). A German AGU patient was found to have a point mutation G179A in exon 2 which changes Gly60 to Asp (Figs. 2, 3 and 6, mutation c) [55]. This glycine is located in a helix of the K-subunit (AH2) facing the interior of the enzyme [2] (Fig. 6, mutation c) and the substitution by aspartic acid again probably disturbs the packing of the protein. A Turkish AGU patient had a G904A transition in exon 8 of the gene [55]. The resulting Gly302Arg mutation (Figs. 2, 3 and 6, mutation r) occurs in the core of the enzyme in a L-strand of the L-subunit (BS5) (Fig. 6, mutation r). A T916C AGU mutation in exon 8 was found in a caucasian-American patient to cause Cys306 to be changed to Arg (Figs. 2, 3 and 6, mutation s) [55]. Cys306 forms an important disul¢de bond with Cys286. Although the Arg replacement disturbs folding of the protein, experimental mutation of Cys306 to Ser allows activation to the two-subunit form of the enzyme [48]. The latter mutant protein, however, was less stable than normal glycosylasparaginase. The paternal allele responsible for AGU in a Dutch heterozygote patient results from a missense transition T183G in exon 2, which substitutes Trp for Cys61 (Figs. 2, 3 and 6, mutation d) (Y. Liu and N.N. Aronson, Jr., unpublished result). This Cys is the only one in human glycosylasparaginase that is not involved in a disul¢de bond and its experimental substitution by Ser does not a¡ect enzyme activation [48]. The large size and hydrophobic nature of Trp probably disrupts the normal tertiary structure in this helix region (AH2) of the protein (Figs. 2, 3 and 6, mutation d). A T214C transition also in exon 2 was characterized as the cause of AGU in members of four unrelated Palestinian Arab families whose origins are in the region of Jerusalem (Figs. 2, 3 and 6, mutation g) [47,58]. This mutation results in substitution of Ser72 by Pro, and this is the only known disease-causing mutation that involves an active site residue [58]. The hydroxyl group of Ser72 is hydrogen bonded to the free Kamino group of the critical N-terminal nucleophile Thr206 of the L-subunit. This mutant enzyme form is transported to lysosomes, but it is not activated into subunits in the ER. Cleavage into subunits did occur when the Pro72 substituted protein was over-

Cyaan Magenta Geel

Zwart

150

N.N. Aronson Jr. / Biochimica et Biophysica Acta 1455 (1999) 139^154

expressed and secreted into the medium by transfected cells in culture. 5.4. Nonsense mutations In a Puerto Rican AGU patient with parents who were ¢rst cousins [61] a mutation T192A changed the wild-type codon TGT encoding Cys64 to the stop codon TGA (Figs. 2 and 3, mutation e) [60]. A theoretical truncated polypeptide with 63 amino acids (40 amino acids after signal peptide cleavage) would be synthesized from the mutant mRNA. 5.5. Insertions and deletions A single-base insertion of T in exon 7 was observed between positions 800 and 801 in a Spanish^ American patient (Figs. 2 and 3, mutation q) [55]. As a result of frameshifting and premature termination of translation, the expected mutant glycosylasparaginase would be shortened by 28 amino acids. The maternal allele of the British AGU siblings described above with missense mutation A101V [57] is a 7-bp deletion from exon 1 (Figs. 2 and 3, mutation a). The seven consecutive nucleotides could be deleted from the sequence between 99 and 107 (TTGGCCCTT) in three possible ways to account for the change observed in the cDNA [57]. This mutation results in a frameshift that is followed by premature termination of translation. Another deletion of 5 bp located in exon 3 was found in the glycosylasparaginase gene of an American patient [59] (Figs. 2 and 3, mutation k). The deletion of ACACA (367^371) resulted in similar consequences as the 7-bp deletion in the two British AGU patients, including an apparent decreased stability of the mutant mRNA. An AGU patient of Mauritian origin was found to have a single-base deletion of either T787 or T788 from exon 7 (Figs. 2 and 3, mutation p) [61]. As a result of the frameshift and premature translational termination the mutant enzyme will become shortened by 84 amino acids. This truncated polypeptide contains complete sequence of the K-subunit and half of that of the L-subunit. When it was expressed in COS-1 cells the truncated protein remained as a single-chain form which suggests that conversion into two subunits does not take place co-translationally during folding, but instead requires a de¢nite three

BBADIS 61881 10-9-99

dimensional structure for the cleavage between Asp205 and Thr206 to occur. Two Dutch AGU alleles are deletions of single bases, both of which result in frameshifting and premature termination of translation. In one Dutch patient, deletion of a single T occurred at position 336 in exon 3 (Figs. 2 and 3, mutation j) [55], and in the second patient, G567 is deleted from exon 5 (Figs. 2 and 3, mutation o, L. Liu and N.N. Aronson, Jr., unpublished result). 5.6. Genomic rearrangement A genomic rearrangement in the glycosylasparaginase gene was characterized in the DNA from the ¢rst AGU case clinically described in the United States [62]. A 2-kb deletion accompanied by the addition of T19 repeats was found. The 5P-deletion breakpoint starts 1258 bp downstream from the 5Pend of intron 8, while the 3P-breakpoint is located near the 5P-end of an inverted Alu repeat that is 1830 bp downstream from the normal reading frame termination codon in exon 9 (Fig. 3, mutation u). A T19 sequence located at the 5P-end of the inverted Alu repeat was triplicated and inserted between the two ends of the deletion breakpoints. After completely losing the normal exon 9 by this genomic rearrangement, a new chimeric intron was formed with a new splice acceptor site located 4326 bp downstream from the normal termination codon. The mutant chimeric mRNA contains a new exon 9 which also provides a stop codon and a polyadenylation site. This exchange of exon 9 yields a mutant glycosylasparaginase that has a hydrophilic region at its normal C-terminus replaced by a hydrophobic group of 16 amino acids. When the mutant glycosylasparaginase was expressed in COS1 cells, an equal amount of mRNA was produced from the mutant construct compared to the normal one. This AGU-type polypeptide, however, remained as a single-chain precursor form and was quickly degraded within 24 h. One of the knockout mouse models appears to have one of its two AGU alleles very similar to this unusual human mutation [39]. 5.7. Discussion A few patterns can be discerned at the molecular level within this relatively small number of AGU

Cyaan Magenta Geel

Zwart

N.N. Aronson Jr. / Biochimica et Biophysica Acta 1455 (1999) 139^154

mutant alleles described above. Isoniemi et al. [52] pointed out a mutational `hotspot' in exon 2 of the gene (Figs. 2 and 3, mutations c^f) that can be explained by the symmetrical and repeating sequences in this tract of the DNA (Fig. 2). The only portion of the glycosylasparaginase gene lacking an AGU mutation is in its middle region between exons 5 and 6 and most of exon 7. Evolutionary pressure to maintain these genetic sequences unchanged could be explained by the critical importance of this part of the encoded polypeptide for the structure and function of glycosylasparaginase. Indeed, all putative activesite residues (Thr206, Thr224, Arg234, Asp237, Thr257, and Gly258), except Ser72, are in the L-subunit [2,58] and are encoded by these three exons. Similarly, the self-cleavage site (204HisAspThr206) which is an essential part of the novel Ntn-amidase structure and activation mechanism [63,64] is at the 3P-end of exon 5 in this segment of the gene. All of these critical amino acids near the N-terminus of the L-subunit are conserved in the sequences of glycosylasparaginases from bacteria to insects to humans [41]. On the other hand, the C-terminal portion of the K-subunit is also encoded by exon 5, but this segment of the enzyme is the most varied in its sequence among these species. One could therefore predict that gene mutations that alter individual amino acids in this portion of the protein might tend to be polymorphic and not have detrimental e¡ects on the enzyme. Indeed, this segment of the human K-subunit is removed when glycosylasparaginase reaches the lysosomes [32]. 6. Glycosylasparaginase as a member of the N-terminal nucleophile (Ntn) hydrolase family of proteins Many of the human AGU mutations can be explained by their direct e¡ects on the structural biology of glycosylasparaginase (Fig. 6). Recently, this enzyme was noted to be a member of a unique protein group termed N-terminal nucleophile amidases (hydrolases) [2,78,79]. Most of these proteins undergo self-cleavage to create an active hydrolase. The process has been studied in both the bacterial [64] and human [85] glycosylasparaginases. The polypeptide at the cleavage site appears to adopt a confor-

BBADIS 61881 10-9-99

151

mation that maintains a substrate pocket into which the cleaved Asp205 (D151, bacterial) side chain ¢ts and thereby stabilizes. An active His^Thr^Thr triad may then be involved: Thr206 is activated upon deprotonation by His205 that requires interaction with Thr224. This mechanism closely follows that of the active enzyme hydrolyzing Asn^GlcNAc. Mature glycosylasparaginase is folded as an KLLK-sandwich structure which is characteristic of the Ntn-amidases (Fig. 4). These domains are schematically depicted in Fig. 6 as helices I, L-sheet II, L-sheet III and helices IV, respectively. The majority (c, d, h, i, m, r and s) of 10 known missense AGU mutations in the protein coding region occur in either an K-helix or L-sheet secondary structure motif of the protein. Only one mutation, Ser72Pro (mutation g), changes an activesite residue. The major Finnish AGU mutation (mutation n) is detrimental because it destroys the important K-subunit Cys163^Cys179 disul¢de that is critical for folding, which in turn is necessary for selfcleavage into subunits and activation to occur. Activation results because the free K-amino group of the new N-terminal nucleophile, Thr206, is a principal catalytic residue in the amidase reaction [63]. Thus, the unique structural biology requirements of glycosylasparaginase as a member of the Ntn-class of enzymes largely determines which gene mutations yield amino acid substitutions that will cause AGU. This explains why so many of the patient mutations produce altered proteins that exhibit the same aberrant phenotype as the major AGUFIN glycosylasparaginase: failure to fold correctly, failure to self-cleave into the active K/L heterodimer, and failure to be transported to lysosomes. Another unexplained repetition among the 21 AGU mutations is that four of the nine codons for Cys residues have been a¡ected. Because of their importance to the folding and stability of glycosylasparaginase [48], substitutions for Cys are expected to be exceptionally detrimental. Bacterial glycosylasparaginase, however, lacks disul¢des [98]. 7. Future directions The study of glycosylasparaginase has been exceptionally rewarding from many scienti¢c directions. From a genetic perspective, we have come to under-

Cyaan Magenta Geel

Zwart

152

N.N. Aronson Jr. / Biochimica et Biophysica Acta 1455 (1999) 139^154

stand the cause of one of approximately 30 genetic diseases that are almost exclusive to the Finnish heritage [69]. Concentration of this genetic disease in a single human population provides hope that combined current and future scienti¢c knowledge can be used to decrease the number of AGU births in Finland and to alleviate the su¡ering of AGU patients and their families. Medical attempts to accomplish this are underway including genetic testing [73] and counseling [72] and bone marrow transplants [70,71]. Even more sophisticated therapeutic approaches, such as enzyme replacement and gene therapy [97], may become possible due to development of the two knockout-mouse models for AGU [38,39] and successful correction of the AGU phenotype in cultured cells by enzyme replacement [74,76,77]. Another major scienti¢c discovery made during the study of glycosylasparaginase is the unique biochemical nature and structural biology of the protein that is based on its overall folding pattern and on the selfactivation of its nascent single-chain precursor by autoproteolysis. The newly recognized enzyme family of which glycosylasparaginase is a member is called N-terminal nucleophile (Ntn) hydrolyses (amidases) [78]. This group of enzymes encompasses very important catalytic reactions in biology [79,80]: the controlling metabolic steps in biosynthesis of purine nucleotides [81]; aminosugars [82]; and amino acids (in plants) [83] and the proteasome whose proteolytic activity is widely utilized in cellular physiology [84]. Experimental studies on the self-activation of both human [85] and bacterial [46,64] glycosylasparaginases have already revealed important aspects of this biochemical process which is now known to be a unique feature of the N-terminal nucleophile family of proteins. Many of the other Ntn-hydrolase family members are complex proteins with multisubunits or domains, and the results from study of glycosylasparaginase and its mutants are likely to become the model for understanding general features of the activation process. Most of the Ntn-hydrolases are amidases that cleave the L- or Q-side chain amide of asparagine or glutamine, or their analogs [79,80]. Glycosylasparaginase has been found also to hydrolyze L-amide peptides of Asn [86].

BBADIS 61881 10-9-99

Acknowledgements I wish to recognize the contributions of students and colleagues who did most of the research on glycosylasparaginase in my laboratory: Sigran Hofmann, Kris Fisher, Ole Tollersrud, Hyejeong Park, Michelle Vittese and Yuan Liu. My research on glycosylasparaginase was supported by United States Public Health Service Grant DK-33314 from the National Institute of Diabetes and Digestive and Kidney Diseases. My thanks go to Mrs. Connie Smith for preparing the manuscript. References [1] S. Mahadevan, A.L. Tappel, J. Biol. Chem. 242 (1967) 4568^ 4576. [2] C. Oinonen, R. Tikkanen, J. Rouvinen, L. Peltonen, Nat. Struct. Biol. 2 (1995) 1102^1108. [3] F.A. Jenner, R.J. Pollitt, Biochem. J. 103 (1967) 48P^49P. [4] R.J. Pollitt, F.A. Jenner, H. Merskey, Lancet 2 (1968) 253^ 255. [5] J. Palo, Doctoral thesis, University of Helsinki, 1966. [6] J. Palo, K. Mattsson, J. Ment. De¢c. Res. 14 (1970) 168^ 173. [7] J. Palo, R.J. Pollitt, K.M. Pretty, H. Savolainen, Clin. Chim. Acta 47 (1973) 69^74. [8] M. Arvio, S. Autio, P. Louhiala, Acta Paediatr. 6-7 (1993) 587^589. [9] S. Autio, J. Visakorpi, H. Ja«rvinen, Ann. Clin. Res. 5 (1973) 149^155. [10] A. Maatta, H.T. Jarvelianen, L.O. Nelimarkka, R.P. Penttinen, Biochim. Biophys. Acta 1225 (1994) 264^270. [11] J. Rapola, Pathol. Res. Pract. 190 (1994) 759^766. [12] K. Yoshida, S.-I. Ikeda, N. Yanagisawa, T. Yamauchi, S. Tsuji, Y. Hirabayashi, Clin. Genet. 40 (1991) 318^325. [13] J. Palo, R.J. Pollitt, K.M. Pretty, H. Savolainen, Clin. Chim. Acta 47 (1973) 69^74. [14] I. Mononen, V. Kaartinen, T. Mononen, J. Inherit. Metab. Dis. 11 (1988) 194^198. [15] N.N. Aronson Jr., M.J. Kuranda, FASEB J. 3 (1989) 2615^ 2622. [16] N.N. Aronson Jr., C. de Duve, J. Biol. Chem. 243 (1968) 4564^4573. [17] N.N. Aronson Jr., P.A. Docherty, J. Biol. Chem. 258 (1983) 4266^4271. [18] M.H. Kuranda, N.N. Aronson Jr., J. Biol. Chem. 260 (1985) 1858^1866. [19] P.A. Docherty, M.J. Kuranda, N.N. Aronson Jr., J. Biol. Chem. 261 (1986) 3457^3463.

Cyaan Magenta Geel

Zwart

N.N. Aronson Jr. / Biochimica et Biophysica Acta 1455 (1999) 139^154 [20] M.J. Kuranda, N.N. Aronson Jr., J. Biol. Chem. 261 (1986) 5802^5809. [21] G.A. Ricca, R.W. Hamilton, J.W. McLean, A. Conn, J.E. Kalinyak, J.M. Taylor, J. Biol. Chem. 256 (1981) 10362^ 10368. [22] P.F. Daniel, B. Winchester, C.D. Warren, Glycobiology 4 (1994) 551^566. [23] P.F. Daniel, J.E. Evans, R. De Gasperi, B. Winchester, C.D. Warren, Glycobiology 2 (1992) 327^336. [24] J.-F. Haeuw, T. Grard, C. Alonso, G. Strecker, J.C. Michalski, Biochem. J. 297 (1994) 463^466. [25] B.A. Gordon, C.A. Rupar, J.W. Rip, M.D. Haust, M.B. Coulter-Mackie, E. Scott, G.G. Hinton, Clin. Invest. Med. 21 (1998) 114^123. [26] M. Tanaka, M. Kohno, I. Yamashina, J. Biochem. 73 (1973) 1285^1289. [27] V. Kaartinen, T. Mononen, R. Laatikainen, I. Mononen, J. Biol. Chem. 267 (1992) 6855^6858. [28] A.L. Tarentino, T.H. Plummer Jr., F. Maley, Biochemistry 14 (1975) 5516^5523. [29] T. Noronkoski, I. Mononen, Glycobiology 7 (1997) 217^220. [30] J.-C. Michalkski, J.-M. Wieruszeski, C. Alonso, P. Cache, J. Montreuil, G. Strecker, Eur. J. Biochem. 201 (1991) 439^ 458. [31] K.J. Fisher, O.K. Tollersrud, N.N. Aronson Jr., FEBS Lett. 269 (1990) 440^444. [32] E. Ikonen, I. Julkunen, O.K. Tollersrud, N. Kalkkinen, L. Peltonen, EMBO J. 12 (1993) 295^302. [33] O.K. Tollersrud, N.N. Aronson Jr., Biochem. J. 260 (1989) 101^108. [34] K. Gro«n, P. Aula, L. Peltonen, Hum. Genet. 85 (1990) 233^ 236. [35] C. Morris, N. Heisterkamp, J. Gro¡en, J.C. Williams, I. Mononen, Hum. Genet. 88 (1992) 295^297. [36] H. Park, K.J. Fisher, N.N. Aronson Jr., FEBS Lett. 288 (1991) 168^172. [37] A. Uusitalo, K. Tenhunen, J. Tenhunen, S. Matikainen, L. Peltonen, A. Jalanko, J. Biol. Chem. 272 (1997) 9524^9530. [38] V. Kaartinen, I. Mononen, J.W. Voncken, T. Noronkoski, I. Gonzalez-Gomez, N. Heisterkamp, J. Gro¡en, Nat. Med. 2 (1996) 1375^1378. [39] A. Jalanko, K. Tenhunen, C.E. McKinney, M.E. LaMarca, J. Rapola, T. Autti, R. Joesnsuu, T. Manninen, I. Sipila«, S. Ikonen, P. Riekkinen, E.I. Ginns, L. Peltonen, Hum. Mol. Genet. 7 (1998) 265^272. [40] K.J. Fisher, N.N. Aronson Jr., J. Biol. Chem. 266 (1991) 12105^12113. [41] Y. Liu, G.S. Dunn, N.N. Aronson Jr., Glycobiology 6 (1996) 527^536. [42] A.L. Tarentino, G. Quinones, C.R. Hauer, L.M. Changchien, T.J. Plasmmer Jr., Arch. Biochem. Biophys. 316 (1994) 399^406. [43] K. Tenhunen, M. Laan, T. Manninen, A. Palotie, L. Peltonen, A. Jalanko, Genomics 30 (1995) 244^250. [44] T.J. Lough, B.D. Roddington, M.R. Grant, D.F. Hill, P.H. Reynolds, K.J. Farnden, Plant Mol. Biol. 19 (1992) 391^399.

BBADIS 61881 10-9-99

153

[45] J. Xuan, A.L. Tarentino, B.G. Grimwood, T.H. Plummer Jr., T. Cui, C. Guan, P. Van Roey, Protein Sci. 7 (1998) 774^781. [46] H.C. Guo, Q. Xu, D. Buckley, C. Guan, J. Biol. Chem. 273 (1998) 20205^20212. [47] J. Zlotogora, Z. Ben-Neriah, B.Y. Abu-Libdeh, V. Sury, M. Zeigler, J. Inherit. Metab. Dis. 20 (1997) 799^802. [48] A. Riikonen, J. Rouvinen, R. Tikkanen, I. Julkunen, L. Peltonen, A. Jalanko, J. Biol. Chem. 271 (1996) 21340^ 21344. [49] A.L. McCormack, I. Mononen, V. Kaartinen, J.R. Yates III, J. Biol. Chem. 270 (1995) 3212^3215. [50] A.-C. Syva«nen, E. Ikonen, T. Manninen, M. Bengtstrom, H. Soderlund, P. Aula, L. Peltonen, Genomics 12 (1992) 590^ 595. [51] O.K. Tollersrud, O. Nilssen, L. Tranebjaerg, O. Borud, J. Med. Genet. 31 (1994) 360^363. [52] A. Isoniemi, M. Hietala, P. Aula, A. Jalanko, L. Peltonen, Hum. Mutat. 5 (1995) 318^326. [53] S. Hreidarsson, G.H. Thomas, D.L. Valle, R.E. Stevenson, H. Taylor, J. McCarty, S.B. Coker, W.R. Green, Clin. Genet. 23 (1983) 427^435. [54] K.J. Fisher, N.N. Aronson Jr., FEBS Lett. 288 (1991) 173^ 178. [55] E. Ikonen, P. Aula, K. Gron, O. Tollersrud, R. Halila, T. Manninen, A.C. Syva«nen, L. Peltonen, Proc. Natl. Sci. USA 88 (1991) 11222^11226. [56] K. Yoshida, N. Yanagisawa, A. Oshima, H. Sakuraba, Y. Iida, Y. Suzuki, Hum. Genet. 90 (1992) 179^180. [57] H. Park, M.B. Vettese, A.H. Fensom, K.J. Fisher, N.N. Aronson Jr., Biochem. J. 290 (1993) 735^741. [58] R. Tikkanen, A. Riikonen, C. Oinonen, R. Rouvinen, L. Peltonen, EMBO J. 15 (1996) 2954^2960. [59] D. Chitayat, S. Nakagawa, R.W. Marion, G.S. Sachs, S.Y. Hahm, H.S. Goldman, Am. J. Med. Genet. 31 (1988) 527^ 532. [60] M. Peltola, D. Chiatayat, L. Peltonen, A. Jalanko, Hum. Mol. Genet. 3 (1994) 2237^2242. [61] H. Park, M. Rossiter, A.H. Fensom, B. Winchester, N.N. Aronson Jr., J. Inherit. Metab. Dis. 19 (1996) 76^83. [62] A. Jalanko, T. Manninen, L. Peltonen, Hum. Mol. Genet. 4 (1995) 435^441. [63] V. Kaartinen, J.C. Williams, J. Tomich, J.R. Yates III, L.E. Hood, I. Mononen, J. Biol. Chem. 255 (1991) 5860^5869. [64] C. Guan, Y. Liu, Y. Shao, T. Cui, W. Liao, A. Ewel, R. Whitaker, H. Paulus, J. Biol. Chem. 372 (1998) 9695^ 9702. [65] I. Mononen, N. Heisterkamp, V. Kaartinen, J.C. Williams, J.R. Yates III, P.R. Gri¡en, L.E. Hood, J. Gro¡en, Proc. Natl. Acad. Sci. USA 88 (1991) 2941^2945. [66] E. Ikonen, M. Baumann, K. Gro«n, A.-C. Syva«nen, N. Enomaa, R. Halila, P. Aula, L. Peltonen, EMBO J. 10 (1991) 51^58. [67] I. Mononen, N. Heisterkamp, V. Kaartinen, T. Mononen, J.C. Wlliams, J. Gro¡en, J. Biol. Chem. 267 (1992) 3196^ 3199.

Cyaan Magenta Geel

Zwart

154

N.N. Aronson Jr. / Biochimica et Biophysica Acta 1455 (1999) 139^154

[68] P. Aula, S. Autio, K.O. Raivio, J. Rapola, in: P. Durand, J.S. O'Brien (Eds.), Genetic Errors of Glycoprotein Metabolism, Berlin, 1982, pp. 123^152. [69] L. Peltonen, Ann. Med. 29 (1997) 553^556. [70] T. Autti, P. Santavuori, R. Raininko, M. Renkend, J. Rapola, U. Saarinen-Pihkala, Lancet 349 (1997) 1366^1367. [71] A. Laitinen, M. Hietala, J.C. Haworth, M.L. Schroeder, L.E. Seargeant, R. Greenberg, P. Aula, Clin. Genet. 51 (1977) 174^178. [72] M. Hietala, A. Hakonen, A.R. Aro, P. Niemala, L. Peltonen, P. Aula, Am. J. Hum. Genet. 56 (1995) 1493^1500. [73] C.M. Delahunty, W. Ankener, S. Brainerd, D.A. Nickerson, I.T. Mononen, Clin. Chem. 41 (1995) 59^61. [74] N. Enomaa, O. Danos, L. Peltonen, A. Jalanko, Hum. Gene Ther. 6 (1995) 723^731. [75] F. Irie, H. Murakoshi, T. Suzuki, Y. Suzuki, K. Kon, S. Ando, K. Yoshida, Y. Hirabayashi, Glycoconj. J. 12 (1995) 290^297. [76] I. Mononen, N. Heisterkamp, U. Dunder, E.L. Romppanen, T. Noronkoski, I. Kuronen, J. Gro¡en, FASEB J. 9 (1995) 428^433. [77] A. Kyttala, O. Heinonen, L. Peltonen, A. Jalanko, J. Neurosci. 18 (1998) 7750^7756. [78] J.A. Brannigan, G. Dodson, H.J. Duggleby, P.C.E. Moody, J.L. Smith, D.R. Tomchick, A.G. Murzin, Nature 378 (1995) 416^419. [79] N.N. Aronson Jr., Glycobiology 6 (1996) 669^675. [80] N.N. Aronson, Jr. in: A.J. Barrett, N.D. Rawlings, J.F. Woessner (Eds.), Handbook of Proteolytic Enzymes, 1998, Academic Press, London, pp. 482^486. [81] J.L. Smith, E.J. Zaluzec, J.-P. Wery, L. Nia, R.L. Switzer, H. Zalkin, Y. Satow, Science 264 (1994) 1427^1433. [82] M.J. Isupov, G. Obmolova, S. Butterworth, M.-A. BadetDenisot, B. Badet, I. Polikarpov, J.A. Littlechild, A. Teplyakov, Structure 4 (1996) 801^810. [83] T.J. Lough, B.D. Reddington, M.R. Grant, D.F. Hill, P.H.S. Reynolds, K.J.F. Farnden, Plant Mol. Biol. 19 (1992) 391^399.

BBADIS 61881 10-9-99

[84] J. Lo«we, D. Stock, B. Jap, P. Zwickl, W. Baumeister, R. Huber, Science 268 (1995) 533^539. [85] J. Saarela, M. Laine, R. Tikkanen, C. Oinonen, A. Jalanko, J. Rouvinen, L. Peltonen, J. Biol. Chem. 273 (1998) 25320^ 25328. [86] T. Noronkoski, I.B. Stoineva, I.P. Ivanov, D.D. Petkov, I. Mononen, J. Biol. Chem. 273 (1998) 26295^26297. [87] I. Mononen, N.N. Aronson, Jr. (Eds), Lysosomal Storage Disease: Aspartylglycosaminuria, R.G. Landes/SpringerVerlag, Austin, TX/Heidelberg, 1997, pp. 1^197. [88] J. Rapola, Pathol. Res. Pract. 190 (1994) 759^766. [89] I. Mononen, K.J. Fisher, V. Kaartinen, N.N. Aronson Jr., FASEB J. 13 (1993) 1247^1256. [90] R. Halia, E. Ikonen, O. Tollsersud, A.-C. Syva«nen, N. Enomaa, L. Peltonen, Biochem. Med. Metab. Biol. 50 (1993) 1^ 8. [91] E. Ikonen, A.-C. Syva«nen, L. Peltonen, Scand. J. Clin. Lab. Invest. Suppl. 213 (1993) 19^27. [92] E. Ikonen, L. Peltonen, Hum. Mutat. 1 (1992) 361^365. [93] V. Kaartinen, I. Mononen, I. Gonzalez-Gomez, T. Noronkoski, N. Heisterkamp, J. Gro¡en, J. Inherit. Metab. Dis. 21 (1998) 207^209. [94] I. Gonzales-Gomez, I. Mononen, N. Heisterkamp, J. Groffen, V. Kaartinen, Am. J. Pathol. 153 (1998) 1293^1300. [95] K. Tenhunen, A. Uusitalo, T. Autti, R. Joensuu, M. Kettunen, R.A. Kauppinen, S. Ikonen, M.E. LaMarca, M. Haltia, E.I. Gins, A. Jalanko, L. Peltonen, J. Neuropathol. Exp. Neurol. 57 (1998) 1154^1163. [96] A. Jalanko, K. Tenhunen, C.E. McKinney, M.E. LaMarca, J. Rapola, T. Autti, R. Joensuu, T. Manninen, I. Sipila, S. Ikonen, P. Riekkinen Jr., E.I. Ginns, L. Peltonen, Hum. Mol. Genet. 7 (1998) 265^272. [97] M. Peltola, A. Kyttala, O. Heinonen, J. Rapola, T. Paunio, F. Revah, L. Peltonen, A. Jalanko, Gene Ther. 5 (1998) 1314^1321. [98] J. Xuan, A.L. Tarentino, B.G. Grimwood, T.H. Plummer Jr., T. Cui, C. Guan, P. Van Roey, Protein Sci. 7 (1998) 774^781.

Cyaan Magenta Geel

Zwart