Sanfilippo syndrome type D: identification of the first mutation in the N-acetylglucosamine-6-sulphatase gene

Mucopolysaccharidosis type IIID is the least common of the four subtypes of Sanfilippo syndrome. It is caused by a deficiency of N-acetylglucosamine-6-sulphatase, which is one of the enzymes involved in the catabolism of heparan sulphate. We present the clinical, biochemical, and, for the first time, the molecular diagnosis of a patient with Sanfilippo D disease. The patient was found to be homozygous for a single base pair deletion (c1169delA), which will cause a frameshift and premature termination of the protein. Accurate carrier detection is now available for other members of this consanguineous family.

alpha-Glucosidase and N-acetylglucosamine-6-sulphatase are the major mannose-6-phosphate glycoproteins in human urine

Most newly synthesized lysosomal enzymes contain a transient carbohydrate modification, mannose 6-phosphate (Man-6-P), which signals their vesicular transport from the Golgi to the lysosome via Man-6-P receptors (MPRs). We have examined Man-6-P glycoproteins in human urine by using a purified soluble fragment of the soluble cation-independent MPR (sCI-MPR) as a preparative and analytical affinity reagent. In a survey of urine samples from seven healthy subjects, the pattern of Man-6-P glycoproteins detected with iodinated sCI-MPR as a probe in a blotting assay was essentially identical in each, regardless of sex or age. Two bands of approx. 100 and 110 kDa were particularly prominent. Man-6-P glycoproteins in human urine were purified by affinity chromatography on immobilized sCI-MPR. Seven distinct bands revealed by SDS/PAGE and Coomassie Blue staining were subjected to N-terminal sequence analysis. The prominent 100 and 110 kDa Man-6-P glycoproteins were identified as N-acetylglucosamine-6-sulphatase and alpha-glucosidase respectively. This identification was confirmed by molecular mass determinations on the two major bands after deglycosylation. Sequence analysis revealed arylsulphatase A and several previously unidentified proteins as minor species. Man-6-P glycoproteins were also purified on an analytical scale to determine the proportion of a number of lysosomal enzyme activities represented by the mannose-6-phosphorylated forms. The lysosomal enzymes in urine containing the highest proportion of mannose-6-phosphorylated form were beta-mannosidase (82%), hexosaminidase (27%) and alpha-glucosidase (24%). The profiles of Man-6-P glycoproteins detected by blotting in urine and plasma were not similar, suggesting that the urinary species are not derived from the bloodstream.

Some urinary enzymes in bilharziasis

Urinary alkaline phosphatase (ALP), acid phosphatase (ACP), aryl sulphatase (Ar. sulph.), beta-glucuronidase (beta-gluc.) and galactosidase were assayed in a group of Bilharzia haematobium patients and another group of healthy subjects (control group). The results for most of the determined enzymes revealed high activities as compared to the controls. The activity of acid phosphatase in male urine samples increased also, though not significantly. These elevated enzyme activities could be used to establish the diagnosis of schistosomiasis in patients whose urine contains no ova or when it is difficult to detect them. The results are discussed in the light of localization of each enzyme in the urinary tract as well as in other organs like the liver.

Desulphation of dextran sulphate during kidney ultrafiltration

The renal clearance of [3H]dextran sulphate by the isolated perfused rat kidney was associated with desulphation of the molecule, as demonstrated by ion-exchange and affinity chromatography of material resident in both glomeruli and urine samples. This process also occurred in vivo. The molecular size distribution of glomerular dextran sulphate in the perfused kidney was indistinguishable from that in the perfusate, and although urinary material was smaller it remained macromolecular. Sulphatase activity was not detected in urine or in the perfusate of perfused kidneys, but was detected in glomerular and non-glomerular cortex fractions isolated by a sieving procedure. The identification of significant biochemical changes to dextran sulphate demonstrates that it does not function as an inert transport probe, and supports the concept of cellular involvement in the process of renal charge selectivity.

Human glucosamine-6-sulphatase deficiency. Diagnostic enzymology towards heparin-derived trisaccharide substrates

Glucosamine-6-sulphatase (6S) activity towards a series of radiolabelled heparin-derived trisaccharide substrates was determined in cultured human skin fibroblast and leucocyte homogenates, and in urine supernatants of normal individuals and patients affected with 6S deficiency [Sanfilippo D syndrome; mucopolysaccharidosis (MPS) type IIID]. The N-sulphated and N-acetylated derivatives of the trisaccharide substrate O-(alpha-glucosamine 6-sulphate)-(1----4)-L-O-(alpha-iduronic acid 2-sulphate)-(1----4)-D-O-2,5-anhydro[1-3H]mannitol 6-sulphate (GlcNH6S-IdoA2S-anM6S) were prepared by enzymic digestion of a pentasulphated tetrasaccharide isolated following the HNO2 deamination of heparin. Purified lysosomal enzymes and MPS-patient skin fibroblasts were used along with chemical degradation to confirm the structure of each of the substrates that were utilized to study the interaction of the enzyme activities required to degrade the highly sulphated regions of heparan sulphate. Human liver, skin fibroblast and urine 6S activities were separated by chromatofocusing into at least four and possibly up to six individual activities. 6S activities present in each of the tissues generally had similar catalytic properties, including Km values, pH optima and inhibition with NaCl, Na2SO4 and NaH2PO4. Leucocyte and skin fibroblast 6S activities towards GlcNAc6S-IdoA2S-anM6S were maximal at pH 4.1 and 3.9 respectively, with Km values of 2.8 microM and 0.9-1.7 microM respectively. Urine 6S activity towards GlcNAc6S-IdoA2S-anM6S was stimulated 30-fold by BSA at pH 3.9, which shifted the pH optimum from 5.1 to 4.2 and decreased the Km value at pH 4.2 from 4.0 microM to 0.5 microM. Residual 6S activity present in the skin fibroblast homogenates from MPS IIID patients was characterized for activity towards GlcNAc6S-IdoA2S-anM6S and observed to have similar pH optima and Km values to normal skin fibroblast 6S activities, although the residual 6S activity was less than 1% of the normal control range.

The heterogeneity of arylsulphatase-A and arylsulphatase-B in normal human urine and urine from cancer patients

Arylsulphatase-A and arylsulphatase-B heterogeneity in normal and cancer patient urine was investigated using high resolution agarose isoelectricfocusing. Normal urine contained up to nine forms of arylsulphatase-A activity with isoelectric points from 4.45 to 5.43 and at least 5 forms of arylsulphatase-B between 8.58 and 9.15 along with a broad zone of activity between pH 6.5 and 7.6. Although cancer patients had significantly higher levels of arylsulphatase-A and arylsulphatase-B activity, their pattern of activity was essentially the same as for the normals with only minor quantitative differences in some peaks.

Radioimmunoassay for arylsulfatase A in urine

Purified arylsulfatase A (EC 3.1.6.1) from human urine was radioiodinated under conditions that caused no significant loss of antigenic activity. We used this labeled arylsulfatase A (specific radioactivity 4-7.5 Ci/g) together with nonlabeled enzyme and rabbit antiserum produced against homogeneous enzyme to develop a radioimmunoassay for arylsulfatase A in urine. A solid-phase, second-antibody technique (Immunobead Second Antibody; Bio-Rad Laboratories) was used to separate free enzyme from antigen-antibody complexes. The working range of the assay was 0.1-4.0 ng of enzyme; within- and between-assay CVs were around 10%, and the analytical recovery was 105.5% (SD 7.7%). The lower limit of detection was 0.08 ng of arysulfatase A per assay, substantially less than that of typical activity-based assays. Over a wide range of urinary arylsulfatase A activities, results by this method agreed well (r = 0.99) with those obtained by activity assays. We measured the enzyme in urines of 59 healthy volunteers and 92 patients with different diseases, including a group of colorectal cancer cases, to determine whether this could serve as a reliable marker for cancer of the colon; however, urinary excretion of arylsulfatase A by most patients with colon cancer was within normal limits.

Structural and immunochemical characterization of human urine arylsulfatase A purified by affinity chromatography

Arylsulfatase A (aryl-sulfate sulfohydrolase, EC 3.1.6.1) was isolated from an ammonium sulfate precipitate of urinary proteins using two different affinity chromatography methods. One method involved the use of concanavalin A-Sepharose affinity chromatography at an early stage of purification, followed by preparative polyacrylamide gel electrophoresis. The other procedure employed arylsulfatase subunit affinity chromatography as the main step and resulted in a remarkably efficient purification. The enzyme had a specific activity of 63 U/mg. The final preparation of arylsulfatase A was homogeneous on the basis of polyacrylamide gel electrophoresis at pH 7.5, and by immunochemical analysis. However, when an enzyme sample obtained by either method of purification was subjected to sodium dodecyl sulfate polyacrylamide gel electrophoresis under reducing or non-reducing conditions, peptide subunits, of 63.5 and 54.5 kDa, were observed. Immunological tests with 125I-labeled enzyme established the presence of a common protein component in both of the electrophoretically separable peptide subunits of human urine arylsulfatase. The amino acid analysis of homogeneous human urine arylsulfatase A showed only a few differences between it and the human liver enzyme. However, immunological cross-reactivity studies using rabbit anti-human urine arylsulfatase revealed immunological difference between the human urine and liver arylsulfatase A enzymes.

Polyclonal antibodies against iduronate 2-sulphate sulphatase from human urine

Human iduronate 2-sulphate sulphatase (EC 3.1.6.-) from urine has been purified by affinity chromatography on concanavalin A-Sepharose, ammonium sulphate fractionation and DEAE-cellulose chromatography. With ion-exchange chromatography, the enzyme was resolved in two activity peaks. The less anionic of these forms was further purified by polyacrylamide gel electrophoresis under non-denaturing conditions. Anti-iduronate 2-sulphate sulphatase antibodies were obtained from mice immunized with polyacrylamide eluted enzyme. The specificity of the antibodies towards iduronate 2-sulphate sulphatase was demonstrated by immunoprecipitation of the enzyme from partially purified urine protein. The procedure described in this work opens the way to the application of hybridoma technology to iduronate 2-sulphate sulphatase.

Arylsulfatase A activities in urine and tissues taken from bladder cancer patients

Urinary arylsulfatase A activity expressed as units/mg of urinary creatinine was significantly increased in bladder cancer patients, but not in patients with other genitourinary tract disorders, such as cystitis, urethritis and prostatic cancer, nor in patients with non-urological malignant diseases. The urinary enzyme activity was positively correlated with the stage of the bladder cancer, while post surgical follow-up revealed a marked decrease of the activity. Arylsulfatase A activity was also shown to be higher in malignant than in normal bladder tissue, demonstrating the activity to be a function of the grade of the tumor. Furthermore, the isoelectric point (pI 5.2-5.3) of the tissue enzyme in the bladder tumor coincided with that of the urine enzyme from the same cancer patients; the pI of the enzyme in urine from normal subjects was 4.7. These results suggest that most of the urinary arylsulfatase A in bladder cancer originates from tumor tissue.

Urinary glucuronidase and arylsulfatases in identical twins of bladder cancer patients

Studies showing that bladder cancer patients have unusually high levels of urinary beta-glucuronidase and arylsulfatases A and B led to the suggestion that these urinary enzymes may participate in bladder cancer etiology. An alternative explanation of the high levels of these urinary enzymes in bladder cancer patients is that the disease itself causes the elevation. Since the levels of these enzymes are genetically determined, measuring these enzymes in healthy identical twins of bladder cancer patients can test whether high enzyme levels occurred prior to bladder cancer. Five healthy identical cotwins of bladder cancer patients, together with matched controls, were measured for urinary beta-glucuronidase, arylsulfatases A and B, and two other lysosomal enzymes as controls, alpha- and beta-galactosidases. The mean levels of all five enzymes were not very different in the cotwins and controls, suggesting that high levels of urinary enzymes observed in bladder cancer patients are a consequence of disease rather than occurring prior to disease and contributing to its etiology.

Clinical significance of a variant form of urinary arylsulfatase A

A variant form of urinary arylsulfatase A which was not detected in normal voided urine was demonstrated in the urine obtained directly from the renal-pelvis [Ishibashi, T. et al.: Biochim. Biophys. Acta, 616, 218-227 (1980)]. The variant form was not observed in urine collected directly from the ureter during operations for uretero-cutaneostomy or ileal conduit. However, the urine from intubated uretero-cutaneostomy patients collected near the pelvi-ureteric junction showed the presence of the variant form, further suggesting the origin specificity of the variant. This arylsulfatase was not demonstrated in the voided urine from patients with non-urologic malignant disorders such as uterine endometrial, uterine cervical, rectal, pancreas head and gastric carcinomas, in spite of its appearance in high levels in the urine from patients with advanced bladder cancer.

Urinary arylsulfatase in normal children and in patients with pediatric malignant disease

Urinary arylsulfatase (AS) activities were measured in 20 normal children and 10 patients with malignant disease, consisting of leukemia (6), Hodgkin's disease (1), neuroblastoma (2), and malignant teratoma (1), Seven of the patients showed a significantly high activity, and a serial measurement carried out in 4 patients showed a well correlated relationship between AS activity and the activity of disease. Thus the measurement of urinary AS activity could be a laboratory test for monitoring the activity of malignant diseases because of its simple and rapid procedure.

Aryl sulfatase A in acute nonlymphocytic leukemia

To ascertain whether estimation of urinary aryl sulfatase A (ASA) is a reliable noninvasive technique for monitoring disease activity in adult acute nonlymphocytic leukemia, we studied the excretion pattern of this enzyme during active disease, partial remission, and complete remission in ten patients and compared the results with those of ten healthy volunteers. There was no significant difference in enzyme excretion between complete remission cases and the control group. Patients with partial remission and active disease had significantly different urinary enzyme levels from each other and from the control and complete remission groups. Estimation of urinary ASA is a useful noninvasive method of monitoring disease activity in adult acute nonlymphocytic leukemia.

Some observations on the assay of arylsuphatase A in urine

A method for the measurement of arylsulphatase A (ASA) in urine is assessed. Urine samples should be stored refrigerated at 4 degrees C and not deep frozen prior to dialysis. The within-batch coefficient of variation of method, in the normal range, is 6.7%. The major source of imprecision is the dialysis step. The normal range is assessed and results are expressed as units per mole creatinine.

A variant form of arylsulfatase A in human urine derived directly from the renal pelvis: kinetic and immunological characterization

Arylsulfatase A (aryl-sulfate sulfohydrolase, EC 3.1.6.1) was examined in voided and in nephrostomic urine. A variant form of the enzyme was found in nephrostomic urine, in addition to the minor form, which is the sole component of arylsulfatase A in voided urine. The nephrostomic enzyme differed from the voided urine enzyme with respect to the kinetic parameters, the isoelectric point, heat stability and immunological reactivity. The isoelectric points of the voided urine and nephrostomic enzymes were 4.7 and 5.3, respectively. The nephrostomic enzyme was more heat-labile at 62.5 degrees C than the voided urine enzyme. Although the Km values of the two enzymes with nitrocatechol sulfate as substrate were almost the same, the V value of the nephrostomic enzyme was approx. one-hundredth that of the voided urine enzyme. The molecular weight (almost 130 000) did not differ between the voided urine and nephrostomic enzymes. It was demonstrated by various methods, using IgG antibody against the purified voided urine enzyme, that the nephrostomic enzyme was antigenically distinct from the voided urine enzyme.

Arylsulfatase A activity of urine in patients with various genitourinary tract disorders

Arylsulfatase A activity was measured in urine from patients with various genitourinary tract disorders such as bladder tumor and inflammation. No significant difference in enzyme activity was found between normal and affected urine on the basis of either the volume or protein content of urine, a finding which differed from previous results. However, it was demonstrated that urine from affected patients was more labile to heat treatment in comparison with the control. Upon pretreatment of urine arylsulfatase A at 62.5 degrees C for 10 min, an average of 57% of the original activity was lost in samples from patients with bladder tumor, 58% in those with testicular tumor and 62% in cystitis and urethritis, respectively, while the enzyme activity in the control urine lost only 27% with similar heat treatment. These results were statistically significant with p < 0.001. The arylsulfatase A from patients with advanced bladder tumors demonstrated the presence of a variant form (pI 5.3) which was not detected in normal urine. This variant of arylsulfatase A was also demonstrated in nephrostomy urine from patients with congenital obstruction of the pelvi-ureteric junction.

[Metachromatic leukodystrophy: report of 2 cases with histochemistry of nerves and muscles]

Two cases of metachromatic leukodystrophy, of the late infantile form are reported. The patients were a girl and a boy of 2 years 10 months old, with initial normal development, but by the age of 18 months began with gait disturbances, difficulty to speak and developed progressive mental deterioration, with signs of long tract involvement, absence of deep tendon reflexes, spasticity, blindness, muscle atrophy and finished in a vegetative state. The diagnosis was made electromyography (signs of denervation), motor nerve conduction velocity (very decreased), assay of arylsulfatase A in the urine (absence of activity), sural nerve biopsy (demyelination and presence of metachromatic granules by the cresyl-violet and toluidine blue) and muscle biopsy (atrophy of type I fibers and presence of metachromatic material in the intramuscular nerve fibers). A quick revision about diagnostic methods, transmission, pathogenesis and variant forms is made.

[Urinary excretion of arylsulfatases A and B in patients with renal transplants]

Arylsulfatase A and B activities have been determined in urine samples from four patients who received a Kidney allograft from living donors as a treatment for terminal uraemia. The values have been compared with those obtained in the urine of the respective donors. Two patients, with optimal renal functionality, the enzymatic activities were in the order of that observed in the urine of the donor in one case and lower than that of the donor in the other case. A patient showing kidney rejection episodes, reversed by the specific therapy, had the enzymatic activities higher than those shown by the donor. Another patient who suffered for two rejections, the last one irreversible, showed a constant higher value of the two enzymatic activities compared with those of the donor and a further increase during the rejection period. In the light of these preliminary results it seems that the determination of the arylsulfatase activities in the urine of transplanted subjects could contribute to establishing the functional activity of the transplanted kidney and also in establishing the tendency of the kidney to undergo rejection.

Fluorometric assay of the arylsulphatases in human urine

A method is described which differentiates arylsulphatase A and arylsulphatase B in human urine. 4-Methylumbelliferyl sulphate serves as the substrate, and silver ions are used to inhibit arylsulphatase A activity. Using fresh urine it is possible to obtain separate values for arylsulphatase A and B when both are present, thus providing a diagnostic test for metachromatic leucodystrophy.

Urinary excretion of arylsulfatases in malnourished/vitamin A deficient children

Serum vitamin A (retinol) levels were generally low in all malnourished children (6-15 microgram/100 ml) compared with control children (50 microgram/100 ml). A significant increase in vitamin A after appropriate therapy was observed in all malnourished groups. Dietary supplements of proteins and calories even without extra vitamin A supplements increased serum vitamin A levels in cases of kwashiorkor indicating active mobilization of liver vitamin A. Total urinary arylsulfatase A activity excreted in 24-h or within 8-h in the morning (6 a.m. to 2 p.m.) was significantly reduced in cases of malnutrition with or without mild vitamin A deficiency symptoms. The excretion of arylsulfatase B was not altered. In cases of severe vitamin A deficiency coupled with malnutrition increased excretion of both arylsulfatases A and B was evident. These results on urinary arylsulfatases excretory pattern have been obtained either in samples collected for 24-h or specifically for 8-h (morning) and it is suggested that this test on urinary arylsulfatases may prove useful for detection of acute vitamin A deficiency with malnutrition in field studies. A ratio of arylsulfatases A/B of 2.0 or less seems to indicate mild malnutrition, the normal ratio being 3.4. Furthermore a low ratio coupled with increased excretion of both arylsulfatases A and B may be considered specific for acute vitamin A deficiency.

N-Acetylglucosamine-6-sulfate sulfatase from human urine

N-Acetylglucosamine-6-sulfate sulfatase, which liberates sulfate from the N-acetylglucosamine 6-sulfate residue at the nonreducing terminus of a 3H-labeled trisaccharide prepared from heparan sulfate, was purified 136-fold from human urine. The final N-acetylglucosamine-6-sulfate sulfatase preparation was free of all lysosomal sulfatases known to act on sulfated polysaccharides and gave a single band in polyacrylamide gel electrophoresis. The enzyme appears to be a glycoprotein with a molecular weight of around 97,000 and displays considerable charge heterogeneity. Multiple forms with pI values between 5.4 and 8.3 with a maximum at pH 7.7 were detected. The enzyme acts on the 3H-trisaccharide with a pH optimum at 5.5 and is active towards the sulfated monosaccharides N-acetylglucosamine 6-sulfate and glucose 6-sulfate. Although predominantly in exosulfatase, the enzyme catalyzes hydrolysis of sulfate from internal N-acetylglucosamine 6-sulfate moieties at a low rate. The Km for the 3H-trisaccharide, N-acetylglucosamine 6-sulfate, and glucose 6-sulfate were 0.15, 1.5, and 7.7 mM, respectively. The enzyme is inhibited by albumin, Hg2+, PO43-, SO42-, and CN-. Enzyme activity was highest in kidney and cultured fibroblasts but could be demonstrated in all human tissues tested.

The activity of arylsulfatase A and B on tyrosine O-sulfates

L-Tyrosine O-sulfate was hydrolyzed by pure human arylsulfatase A (arylsufate sulfohydrolase, EC 3.1.6.1). The rate of hydrolysis was 1/20 of the rate with nitrocatechol sulfate, but was comparable to the rate with cerebroside sulfate. The reaction was optimal at pH 5.3--5.5 and displayed zero order kinetics with time and enzyme concentration. The Km was about 35 mM. The enzyme showed no stereospecificity and hydrolyzed D-tyrosine O-sulfate with Km and V similar to those for the L-isomer. Arylsulfatase B was less than 5% as effective as arylsulfatase A in catalyzing the hydrolysis of the tyrosine sulfates. The daily urinary excretion of tyrosine sulfate by a patient with metachromatic leukodystrophy (arylsulfatase A deficiency) was comparable to the excretion by control subjects. The biological relevance of the tyrosine sulfatase activity of arylsulfatase A remains uncertain.

The laboratory diagnosis of Sanfilippo disease

The biochemical findings in 29 patients with Sanfilippo disease are reported and a scheme for laboratory diagnosis is outlined. A grossly elevated urinary excretion of heparan sulphate was a consistent and diagnostic finding, even at birth. The excretion of heparan sulphate and chondroitin sulphate was quantitatively similar in types A and B of the condition. Modifications of previously described methods for the determination of heparin sulphamidase in leucocytes or skin fibroblasts and N-acetyl-alpha-D-glucosaminidase in plasma or fibroblasts facilitated the measurement of specific activities. Sanfilippo A disease appeared to be the commonest mucopolysaccharidosis occurring in England and Sanfilippo B disease, one of the rarest forms.

Diurnal variations of urinary enzyme excretion

Variations in the urinary excretion of arylsulphatase A, beta-galactosidase, alpha-glucosidase and beta-glucuronidase throughout a 24-h period were studied in 8 healthy subjects. Urine was collected at 3-h intervals and enzyme activities were assayed after gelfiltration of the urine specimens. Significant intra-individual changes of the excretion of all 4 enzymes during the 24-h period were found. Enzyme output was high between 3 a.m. and 9 a.m. and low during the afternoon and evening hours. The most striking pattern was seen for arylsulphatase A. Diurnal variations of urinary enzyme excretion seemed not to be flow dependent. Both modes of expression of enzyme output (mU/min or U/g creatinine) gave corresponding results. It is concluded that for the measurement of the excretion of these enzymes urine should be collected during a fixed time interval, e.g. from 6 a.m. to 9 a.m.

Defective heparan sulfate metabolism in the Sanfilippo syndrome and assay of this defect in the assessment of the mucopolysaccharidoses patient

The Sanifilippo syndrome is an inherited dementia caused by defective degradation of heparan sulfate. In the course of its catabolism the heparan sulfate polymer must be desulfated. Heparan sulfate sulfatase activity was demonstrated in homogenates of normal tissues and cultured skin fibroblasts, and in normal urine. This activity was found to be grossly depressed or absent in necropsy specimens of liver and spleen from two Sanfilippo patients. The heparan sulfate sulfatase activity was not demonstrable in urine from eleven, or cultured fibroblasts from four Sanfilippo patients. Activities of alpha-N-acetyl-glucosaminidase, the site of the metabolic defect in the Sanfilippo B variant were either normal or slightly elevated in the Sanfilippo tissues and cultured fibroblasts whereas the mean level in the urine of our Sanfilippo patients was about one-third of that encountered in control urines.

Purification and properties of arylsulfatase. A from human urine

Arylsulfatase A (cerebroside sulfate sulfohydrolase) was purified 3500-fold at a 7% yield from human urine. A crude urinary protein concentrate was prepared by treating pooled urine with ammonium sulfate and subsequently drying the precipitate with acetone. The powder thus obtained was extracted with buffer and was subjected to chromatographic and electrophoretic procedures as follows: (a) ammonium sulfate reverse gradient solubilization chromatography; (b) DEAE-cellulose chromatography; (c) Sephadex G-200 gel filtration; (d) preparative polyacrylamide gel electrophoresis; (e) SP-Sephadex chromatography; and (f) antialbumin-Sepharose chromatography. The enzyme was judged to be essentially homogeneous by: (a) a single band on polyacrylamide gel electrophoresis at two pH values; (b) formation of a single precipitin line on immunodiffusion against its antiserum: (c) complete freedom from albumin, the major contaminating protein; and (d) a single band on sodium dodecyl sulfate gel electrophoresis.