Structure of a human lysosomal sulfatase.

BACKGROUND: . Sulfatases catalyze the hydrolysis of sulfuric acid esters from a wide variety of substrates including glycosaminoglycans, glycolipids and steroids. There is sufficient common sequence similarity within the class of sulfatase enzymes to indicate that they have a common structure. Deficiencies of specific lysosomal sulfatases that are involved in the degradation of glycosamino-glycans lead to rare inherited clinical disorders termed mucopolysaccharidoses. In sufferers of multiple sulfatase deficiency, all sulfatases are inactive because an essential post-translational modification of a specific active-site cysteine residue to oxo-alanine does not occur. Studies of this disorder have contributed to location and characterization of the sulfatase active site. To understand the catalytic mechanism of sulfatases, and ultimately the determinants of their substrate specificities, we have determined the structure of N-acetylgalactosamine-4-sulfatase. RESULTS: . The crystal structure of the enzyme has been solved and refined at 2.5 resolution using data recorded at both 123K and 273K. The structure has two domains, the larger of which belongs to the alpha/beta class of proteins and contains the active site. The enzyme active site in the crystals contains several hitherto undescribed features. The active-site cysteine residue, Cys91, is found as the sulfate derivative of the aldehyde species, oxo-alanine. The sulfate is bound to a previously undetected metal ion, which we have identified as calcium. The structure of a vanadate-inhibited form of the enzyme has also been solved, and this structure shows that vanadate has replaced sulfate in the active site and that the vanadate is covalently linked to the protein. Preliminary data is presented for crystals soaked in the monosaccharide N-acetylgalactosamine, the structure of which forms a product complex of the enzyme. CONCLUSIONS: . The structure of N-acetylgalactosamine-4-sulfatase reveals that residues conserved amongst the sulfatase family are involved in stabilizing the calcium ion and the sulfate ester in the active site. This suggests an archetypal fold for the family of sulfatases. A catalytic role is proposed for the post-translationally modified highly conserved cysteine residue. Despite a lack of any previously detectable sequence similarity to any protein of known structure, the large sulfatase domain that contains the active site closely resembles that of alkaline phosphatase: the calcium ion in sulfatase superposes on one of the zinc ions in alkaline phosphatase and the sulfate ester of Cys91 superposes on the phosphate ion found in the active site of alkaline phosphatase.

Genotype-phenotype correlations in mucopolysaccharidosis type I using enzyme kinetics, immunoquantification and in vitro turnover studies.

Fibroblasts from 16 patients with known alpha-L-iduronidase gene mutations and different clinical phenotypes of mucopolysaccharidosis type I (MPS I) were investigated in order to establish genotype/phenotype correlations. Enzyme kinetic studies were performed using the specific alpha-L-iduronidase substrate iduronosyl anhydro[1-3H]mannitol-6-sulfate. Specific residual enzyme activities were estimated using the kinetic parameters and an immunoquantification assay which determines levels of alpha-L-iduronidase protein. Cells were cultured in the presence of [35S]sulfate and the in vivo degradation of accumulated labelled glycosaminoglycans measured after different chase times. Residual enzyme activity and different amounts of residual enzyme protein were present in extracts from 9 of 16 cell lines covering a wide spectrum of clinical severity. Catalytic capacity, calculated as the product of kcat/Km and ng iduronidase protein per mg cell protein, was shown in most cases to be directly related to the severity of clinical phenotype, with up to 7% of normal values for patients with the attenuated form of MPS I (Scheie) and less than 0.13% for severely affected patients (Hurler) In vitro turnover studies allowed further refinement of correlations between genotype and phenotype. Scheie disease compared to Hurler disease patients were shown to accumulate smaller amounts of glycosaminoglycans that were also turned over faster. A combination of turnover and residual enzyme data established a correlation between the genotype, the biochemical phenotype and the clinical course of this lysosomal storage disorder.

Identification and molecular characterization of alpha-L-iduronidase mutations present in mucopolysaccharidosis type I patients undergoing enzyme replacement therapy.

Mucopolysaccharidosis type I (MPS I) is an autosomal recessive lysosomal storage disorder caused by a deficiency of alpha-L-iduronidase (IDUA). Mutations in the gene are responsible for the enzyme deficiency, which leads to the intralysosomal storage of the partially degraded glycosaminoglycans dermatan sulfate and heparan sulfate. Molecular characterization of MPS I patients has resulted in the identification of over 70 distinct mutations in the IDUA gene. The high degree of molecular heterogeneity reflects the wide clinical variability observed in MPS I patients. Six novel mutations, c.1087C>T (p.R363C), c.1804T>A (p.F602I), c.793G>C, c.712T>A (p.L238Q), c.1727+2T>A, and c.1269C>G (p.S423R), in a total of 14 different mutations, and 13 different polymorphic changes, including the novel c.246C>G (p.H82Q), were identified in a cohort of 10 MPS I patients enrolled in a clinical trial of enzyme-replacement therapy. Five novel amino acid substitutions and c.236C>T (p.A79V) were engineered into the wild-type IDUA cDNA and expressed. A p.G265R read-through mutation, arising from the c.793G>C splice mutation, was also expressed. Each mutation reduced IDUA protein and activity levels to varying degrees with the processing of many of the mutant forms also affected by IDUA. The varied properties of the expressed mutant forms of IDUA reflect the broad range of biochemical and clinical phenotypes of the 10 patients in this study. IDUA kinetic data derived from each patient's cultured fibroblasts, in combination with genotype data, was used to predict disease severity. Finally, residual IDUA protein concentration in cultured fibroblasts showed a weak correlation to the degree of immune response to enzyme-replacement therapy in each patient.

Mutational analysis of mucopolysaccharidosis type VI patients undergoing a trial of enzyme replacement therapy.

Mucopolysaccharidosis type VI (MPS VI), or Maroteaux-Lamy syndrome, is a lysosomal storage disorder caused by a deficiency of N-acetylgalactosamine-4-sulfatase (ARSB). Seven MPS VI patients were chosen for the initial clinical trial of enzyme replacement therapy. Direct sequencing of genomic DNA from these patients was used to identify ARSB mutations. Each individual exon of the ARSB gene was amplified by PCR and subsequently sequenced. Nine substitutions (c.289C>T [p.Q97X], c.629A>G [p.Y210C], c.707T>C [p.L236P], c.936G>T [p.W312C], c.944G>A [p.R315Q], c.962T>C [p.L321P], c.979C>T [p.R327X], c.1151G>A [p.S384N], and c.1450A>G [p.R484G]), two deletions (c.356_358delTAC [p.Y86del] and c.427delG), and one intronic mutation (c.1336+2T>G) were identified. A total of 7 out of the 12 mutations identified were novel (p.Y86del, p.Q97X, p.W312C, p.R327X, c.427delG, p.R484G, and c.1336+2T>G). Two of these novel mutations (p.Y86del and p.W312C) were expressed in Chinese hamster ovary cells and analyzed for residual ARSB activity and mutant ARSB protein. The two common polymorphisms c.1072G>A [p.V358M] and c.1126G>A [p.V376M] were identified among the patients, along with the silent mutation c.1191A>G. Cultured fibroblast ARSB mutant protein and residual activity were determined for each patient, and, together with genotype information, were used to predict the expected clinical severity of each MPS VI patient.

Psychometric, diurnal, and electrophysiological correlates of activation.

Several studies were performed using Thayer's Activation-Deactivation Adjective Check List (AD-ACL). Factor analysis of college students' scores identified a single activation factor representing a continuum ranging from sleep through wakefulness. Next, changes over the diurnal cycle in activation factor scores were demonstrated. One-week test-retest reliability was also investigated. Finally, activation scores were related to pulse rate, respiration rate, and skin resistance level under passive and active conditions. Correlations between factor scores and electrophysiological measures were higher than were intercorrelations of electrophysiological measures, indicating that scores on this revised activation factor represent a valid measure of phenomenological bodily activation level.

Human liver N-acetylglucosamine-6-sulphate sulphatase. Purification and characterization.

Human N-acetylglucosamine-6-sulphate sulphatase was purified at least 50,000-fold to homogeneity in 78% yield from liver with a simple three-step four-column procedure, which consists of a concanavalin A-Sepharose/Blue A-agarose coupled step, chromatofocusing and Cu2+-chelating Sepharose chromatography. In all, four forms were isolated and partially characterized. Forms A and B, both with a pI greater than 9.5 and representing 30% and 60% respectively of the recovered enzyme activity, were separated by hydroxyapatite chromatography of the enzyme preparation obtained from the Cu2+-chelating Sepharose step. Both forms A and B had native molecular masses of 75 kDa. When analysed by SDS/polyacrylamide-gel electrophoresis, form A consists of a single polypeptide of molecular mass 78 kDa, whereas form B contained 48 kDa and 32 kDa polypeptide subunits. Neither form A nor form B was taken up from the culture medium into cultured human skin fibroblasts. The two other forms (C and D), with pI values of 5.8 and 5.4 respectively, represented approx. 7% and 3% of the total recovered enzyme activity. The native molecular masses of forms C and D were 94 kDa and approx. 75 kDa respectively. Form C contained three polypeptides with molecular masses of 48, 45 and 32 kDa. N-Acetylglucosamine-6-sulphate sulphatase activity was measured with a radiolabelled disaccharide substrate derived from heparin. The development of this substrate enabled the isolation and characterization of N-acetylglucosamine-6-sulphate sulphatase to proceed efficiently. Forms A, B and C had pH optima of 5.0, Km values of 11.7, 14.2 and 11.1 microM respectively and Vmax. values of 105, 60 and 53 nmol/min per mg of protein respectively. The molecular basis of the multiple forms of this sulphatase is not known. It is postulated that the differences in structure and properties of the four enzyme forms are due to differences in the state of processing of a large subunit.

Biochemical characterization of patients and prenatal diagnosis of sialic acid storage disease for three families.

Modifications of the assay method of Aminoff (1961) for the determination of sialic acid levels in urine, amniotic fluid, cultured cell homogenates and tissue homogenates, which reduce the interference from proteins by precipitation and from interfering chromogens by second derivative spectroscopy are described. Biochemical profiles of patients from three families with different clinical forms of sialic acid storage disease were made using data obtained with the new method. A family with two patients with the clinically severe, early-onset form is described, in which a 9-fold elevation of sialic acid can be detected in the skin fibroblasts and a 12-fold elevation in the urine. A patient from the second family is described with very severe clinical features and with 160-fold and 16-fold elevations of sialic acid in the urine and skin fibroblasts respectively. A patient from a third family is described with mild clinical features but with a 160-fold and 6-fold elevation of sialic acid in urine and skin fibroblasts respectively. The data obtained in this study are compared with data obtained in other laboratories on other patients. The method was used to assess the levels of sialic acid present in amniotic cells and chorionic villus cells obtained prenatally from pregnancies in each of the three families. In one case, in which amniotic cells were elevated in sialic acid and were vacuolated, the pregnancy was terminated. Follow-up studies confirmed the diagnosis. Pregnancies from the other two families were assessed to be not affected.

Improved concanavalin A-Sepharose elution by specific readsorption of glycoproteins.

Elution of bound glycoproteins from concanavalin A-Sepharose can be made more efficient by their readsorption to a Blue A agarose column (specific) and Green A agarose column (less-specific) during recycling of the elution buffer. Three lysosomal enzymes were eluted in this way with marked improvement in their specific activities, time and handling and amount of eluting ligand used.

Adaptive Behavior Scale, Part Two: predictive efficiency of severity and frequency scores.

Severity and frequency of occurrence methods of scoring the Adaptive Behavior Scale, Part Two, were compared using ratings for seven groups of mentally retarded persons. Comparisons involved separately correlating the two methods with independently obtained clinical impressions of symptomatology. The severity scoring system predicted approximately 11 percent more of the variance in the criterion than did frequency scores. These results provide considerable construct and predictive validity for the severity scoring system.

Acetyl CoA:alpha-glucosaminide N-acetyl transferase: partial purification from human liver.

The lysosomal enzyme acetyl CoA:alpha-glucosaminide N-acetyltransferase (GNAT) was shown to be an integral membrane protein requiring high concentrations of the detergent Triton X-100 for maximal solubilization. Using a concentration dependent Triton X-100 solubilization procedure and Concanavalin A-Sepharose affinity chromatography, GNAT was purified 50-fold with a yield of 45%. GNAT activity was separated from N-acetyltransferase activity toward glucosamine 6-phosphate, an alternative non-lysosomal pathway for glucosamine metabolism. GNAT was different from other lysosomal enzymes which bound to Concanavalin A-Sepharose in that both alpha-methylmannoside and Triton X-100 were required for elution of enzyme activity. GNAT activity, which bound to Concanavalin A-Sepharose, required at least one other component which did not bind for maximal expression of enzyme activity and for storage stability. Phospholipids and glycolipids, such as phosphatidylethanolamine, phosphatidylcholine, phosphatidylglycerol, sphingomyelin and gangliosides, and bovine serum albumin allowed expression of enzyme activity and storage stability similar to the component(s) which did not bind to Concanavalin A-Sepharose.

Insulin receptor synthesis and turnover in differentiating 3T3-L1 preadipocytes.

3T3-L1 "preadipocytes" can be induced to differentiate in culture into cells having the morphological and biochemical characteristics of adipocytes. The binding of 125I-insulin to the cell-surface of differentiated and undifferentiated 3T3-L1 cells and nondifferentiating 3T3-C2 cells was compared. In the absence of agents which induce adipocyte conversion, ie, insulin or insulin plus methylisobutylxanthine (MIX) and dexamethasone (DEX), 3T3-L1 cells fail to express the adipocyte phenotype and maintain a constant number of insulin binding sites. Induction of adipocyte conversion with 3T3-L1 cells in the presence of insulin causes apparent down-regulation of insulin receptors followed by a 12--15-fold increase in receptor number which parallels differentiation. Approximately 170,000 insulin binding sites per cell are expressed when greater than 75% of the cells have differentiated. The rise of insulin receptor level is differentiation-dependent. 3T3-C2 cells, which do not differentiate in the presence of insulin or insulin plus MIX and DEX, exhibit only insulin-induced down-regulation of insulin receptors. The increase of insulin receptor level in 3T3-L1 cells in receptor-specific since the levels of epidermal growth factor receptor or choleragen receptor, respectively, remain constant or decrease substantially. A heavy isotope, density-shift technique was used to analyze insulin receptor synthesis and turnover in cells labeled with "heavy" (2H, 13C, and 15N) amino acids. Solubilized newly-synthesized "heavy" and old "light" receptors were separated by isopycnic banding on CsCl gradients and quantitated. The size of the soluble receptor isolated after isopycnic banding in CsCl gradients is approximately 400,000 daltons. Mixing of "light" and "heavy" membranes prior to extraction of receptor revealed no change in "light" or "heavy" receptor isopycnic banding densities. Thus, no detectable interchange of subunits occurs between receptor molecules during extraction or equilibrium centrifugation. Insulin receptor synthesis and turnover, studied by the density-shift technique showed that the rise of receptor level during differentiation results primarily from an increased rate of receptor synthesis. The rate of insulin receptor degradation was not significantly altered. The t1/2 for degradation of the insulin receptor in differentiated 3T3-L1 cells in culture was 6--7 hours in the presence of insulin. Removal of insulin from the medium did not materially affect the rate of receptor degradation. Inhibition of protein synthesis with cycloheximide causes a lengthening of the t1/2 for insulin receptor degradation to 26 hours. Thus, the synthesis of a short-lived protein appears to be required for a critical step in the pathway of insulin receptor degradation.

Human alpha-L-iduronidase. 2. Catalytic properties.

The kinetic parameters of human liver alpha-L-iduronidase were determined with three disaccharide substrates: alpha-L-iduronosyl(1----4)2,5-anhydro-D-[1-3H]mannitol 6-sulphate, alpha-L-iduronosyl(1----4)2,5-anhydro-D-[1-3H]mannitol and alpha-L-iduronosyl(1----3)2,5-anhydro-D-[1-3H]talitol 4-sulphate, derived from the natural substrates heparin and dermatan sulphate and one synthetic, fluorogenic substrate, 4-methylumbelliferyl alpha-L-iduronide. The enzyme activity with all four substrates was optimal at about pH 4.5. The Km values derived using the disaccharide substrates were elevated up to 10-fold with up to a 6.5-fold increase in ionic strength whereas that for the synthetic substrate was only increased by 1.7-fold. The V values for all substrates were unaffected. The inhibitory effect of NaCl, Na2SO4, NaH2PO4 or CuCl2 on enzyme activity was more pronounced with the disaccharide substrates than with the synthetic substrate. The moiety which is most important in binding is the idopyranosyl residue. While the aglycone residue adds to the net affinity for the enzyme, it is the substituent groups of both residues which appear to control catalysis. Specifically the carboxyl moiety of the alpha-L-iduronic acid residue is essential for catalysis while the presence of sulphate on the C4 or C6 position of the aglycone residue has a major influence on catalysis rather than binding. alpha-L-Idosyl(1----4)2,5-anhydro-D-[1-3H]mannitol 6-sulphate did not undergo catalysis and was a potent inhibitor of enzyme activity, whereas beta-glucuronosyl(1----4)2,5-anhydro-D-[1-3H]mannitol 6-sulphate, alpha-L-iduronosyl-2-sulphate(1----4)2,5-anhydro-D-[1-3H]-mannitol 6-sulphate and 4-methylumbelliferyl alpha-L-idoside did not undergo catalysis and were not inhibitory. A model of the catalytic requirements of alpha-L-iduronidase is proposed.

Crystallization and preliminary characterization of human recombinant N-acetylgalactosamine-4-sulfatase.

Crystals of human recombinant N-acetylgalactosamine-4-sulfatase have been grown using vapour diffusion. The protein contains approximately 13%(w/w) carbohydrate. The crystals belong to the tetragonal space group P4(1)2(1)2 or its enantiomorph P4(3)2(1)2 with a = b = 108.0 and c = 145.5 A. The crystals diffract to 2.7 A resolution.

Immunopurification and characterization of human alpha-L-iduronidase with the use of monoclonal antibodies.

alpha-L-Iduronidase from human liver was purified by a three-step five-column procedure and by immunoaffinity chromatography with a monoclonal antibody raised against purified enzyme. Seven bands identified by staining with Coomassie Blue had molecular masses of 74, 65, 60, 49, 44, 18 and 13 kDa and were present in both preparations of the liver enzyme. However, relative to the immunopurification procedure, alpha-L-iduronidase purified by the five-column procedure was considerably enriched in the 65 kDa polypeptide band. The seven bands were identified by Western-blot analysis with two different monoclonal antibodies raised against alpha-L-iduronidase. The chromatographic behaviour of alpha-L-iduronidase on the antibody column was dependent upon the quantity of enzyme loaded. Above a particular load concentration a single peak of enzyme activity was eluted, whereas at load concentrations below the critical value alpha-L-iduronidase was eluted in two peaks of activity, designated form I (eluted first) and form II (eluted second). The following properties of the two forms of alpha-L-iduronidase were determined. (1) The two forms from liver were composed of different proportions of the same seven polypeptides. (2) When individually rechromatographed on the antibody column, each form from liver shifted to a more retarded elution position but essentially retained its chromatographic behaviour relative to the other form. (3) Forms I and II of liver alpha-L-iduronidase showed no difference in their activities towards disaccharide substrates derived from two glycosaminoglycan sources, heparan sulphate and dermatan sulphate. (4) The native molecular size of forms I and II of liver alpha-L-iduronidase was 65 kDa as determined by gel-permeation chromatography. (5) Immunoaffinity chromatography of extracts of human lung and kidney resulted in the separation of alpha-L-iduronidase into two forms, each with different proportions of the seven common polypeptide species. (6) Lung forms I and II were taken up readily into cultured skin fibroblasts taken from a patient with alpha-L-iduronidase deficiency. Liver forms I and II were not taken up to any significant extent. Lung form II gave intracellular contents of alpha-L-iduronidase that were more than double those of normal control fibroblasts, whereas lung form I gave contents approximately equal to normal control values. We propose that all seven polypeptides are derived from a single alpha-L-iduronidase gene product, and that different proportions of these polypeptides can function as a single alpha-L-iduronidase entity.(ABSTRACT TRUNCATED AT 400 WORDS)

Cloning and expression of the gene involved in Sanfilippo B syndrome (mucopolysaccharidosis III B).

Sanfilippo B syndrome is caused by a deficiency of alpha-N-acetylglucosaminidase, a lysosomal enzyme involved in the degradation of heparan sulphate. Accumulation of the substrate in lysosomes results in degeneration of the central nervous system with progressive dementia often combined with hyperactivity and aggressive behaviour. In order to clone the deficient gene, we purified the enzyme from human placenta and obtained amino acid sequence information. Alignment of one of the CNBr generated internal peptides to sequence from the database revealed the chromosomal location of the gene in the 5' upstream flanking region of the gene for 17-beta-hydroxysteroid-dehydrogenase at 17q21.1. The available DNA sequence was used to clone the cDNA coding for alpha-N-acetylglucosaminidase and analyse its gene structure. The gene is fully contained in the 5' upstream flanking region of the gene for 17-beta-hydroxysteroid-dehydrogenase and interrupted by five introns. The cDNA clone has a length of 2575 bp and encodes a protein of 743 amino acids. Chinese hamster ovary cells transfected with the cDNA construct show alpha-N-acetylglucosaminidase activity about 17-fold over background. This will allow correction studies with NAG deficient Sanfilippo B cell lines and facilitate the development of enzyme replacement therapy for these patients.

Cellular location of N-acetyltransfer activities toward glucosamine and glucosamine-6-phosphate in cultured human skin fibroblasts.

The intracellular location in normal human cultured skin fibroblasts of the N-acetyltransferase activities that transfer the acetyl group from acetyl-CoA to the 2-amino group of glucosamine and glucosamine-6-phosphate have been investigated. Organelles have been separated using a combination of differential centrifugation and free flow electrophoresis. The intracellular distribution of the enzyme involved in the N-acetyltransfer to glucosamine and an alpha-glucosaminide disaccharide indicated that this enzyme activity concentrates mainly with lysosomal organelles whereas the activity associated with N-acetyltransferase to glucosamine-6-phosphate is non-lysosomal. It is proposed that acetyl-CoA: alpha-glucosaminide N-acetyltransferase may be used as a convenient enzyme marker of lysosomal organelle membranes.

Irreversible inhibition of fatty acid synthase from rat mammary gland with S-(4-bromo-2,3-dioxobutyl)-CoA. Effect on the partial reactions, protection by substrates and stoichiometry studies.

Fatty acid synthase from lactating rat mammary gland is rapidly and irreversibly inhibited by S-(4-bromo-2,3-dioxobutyl)-CoA. Of the seven partial reactions catalysed by the enzyme, the inhibition of the overall catalytic activity is closely paralleled only by inhibition of the beta-oxoacyl synthase (condensing) partial reaction. Three partial reactions. Beta-oxoacyl reductase, beta-hydroxyacyl dehydratase and enoyl reductase, are inhibited to a modest degree. The three partial reactions known to involve an acyl-CoA/CoA-binding site, acetyl acyltransferase, malonyl acyltransferase and palmitoyl thioesterase, are not inhibited by S-(4-bromo-2,3-dioxobutyl)-CoA. The modification process does not cause the enzyme to dissociate into catalytically incompetent monomers. Stoichiometric studies suggest that approx. 6 mol of reagent are incorporated per mol of totally inhibited enzyme (dimer). The formation of acylated enzyme from either acetyl-CoA or malonyl-CoA protects the enzyme equally well against S-(4-bromo-2,3-dioxobutyl)-CoA. Also, pretreatment of the enzyme with 5,5'-dithiobis-(2-nitrobenzoic acid), a thiol-specific reagent reported to block essential thiol groups in the condensing partial reaction, protects against inhibition by the reagent. On the other hand, the presence of up to 770 microM-S-acetonyl-CoA or dethio-CoA does not protect the enzyme from irreversible inhibition. Together, the results suggest that the primary inhibitory process is a bimolecular reaction resulting in alkylation of essential thiol groups in the condensing partial reaction: this process does not require the obligatory formation of a Michaelis-Menten complex of enzyme and reagent before the alkylation reaction.

Adaptive Behavior Scale part two: Relative severity of maladaptive behavior.

Part Two of the Adaptive Behavior Scale contains statements that cover 13 domains of maladaptive behavior. The present system of scoring and profiling reflects the frequency of occurrence of behavior (either occasionally or frequency) and not severity or relative importance. Thirty-three psychologists with experience in mental retardation rated the 452 statements contained in Part Two along a continuum of severity. Median severity scores were computed for all statements. Results indicated that statements prefaced by "frequently" were judged as more serious than those prefaced by "occasionally," and domains varied greatly with regard to relative severity. Correlational analysi revealed that irrespective of absolute differences due to frequency of occurrence, different types of maladaptive behavior were systematically placed along a relative continuum. Substantial reliability was found among judgments. Possible clinical and psychometric advantages of using these severity scores were suggested.