Int. J. Peptide Protein Res. 26, 1985, 362-372 Purification of mammalian arylsulfatase A enzymes by subunit affinity chromatography ABDUL WAHEED and ROBERT L. VAN ETTEN Department of Chemistry, Purdue University, West Lafayette, IN, USA Received 29 January, accepted for publication 4 April 1985 Rabbit liver arylsulfatase A (arylsulfatase sulfohydrolase, EC 3.1.6.1) monomer was immobilized on cyanogen bromide-activated SepharosedMB and on Affi- Gel-10 under various experimental conditions in order to study the effects of variables in sulfatase monomer/oligomer subunit affinity chromatography. First, the number of reactive groups on activated Sepharose-6MB and Affi-Gel-10 was determined by a procedure involving spectrophotometric titration with L- tyrosine. After covalent coupling of sulfatase monomers to the gels, the enzyme binding capacities of the sulfatase subunit affinity gel matrixes were determined at pH 4.5. The maximum binding of free monomers from solution could be achieved when the Affi-Gel-10 protein monomer matrix was prepared at low degrees of covalent loading. The introduction of a batch technique for equili- bration of the protein sample with the monomer affinity matrix also increased the efficiency of the subunit affinity gel in purification procedures. The effect of pH on the stability of the heterodimers formed between monomers of rabbit liver arylsulfatase A immobilized on Affi-Gel-10 and free monomers of aryl- sulfatase A enzymes from different tissues and organisms was studied using the batch technique. For all sulfatase A enzymes tested, the midpoint of the pH transition for subunit association was pH 6.2, suggesting that the amino acid residues involved in the dimerization are similar. The versatility of the Affi-Gel- 10 monomer affinity matrix was further demonstrated by purifying 13 mam- malian arylsulfatase A enzymes to homogeneity, as assessed by Sephacryl chromatography, native and SDS gel electrophoresis. The molecular weights of the homogeneous monomers and their peptide subunits were in the range of 110-1 80KDa and 50-64 KDa, respectively. The amino acid compositions of these enzymes were also determined. Key words: amino acid composition; arylsulfatase A; subunit affinity chromatography Mammalian arylsulfatase A sulfohydrolases diseases in which a deficiency of arylsulfatase A (EC 3.1.6) are acid hydrolases of lysosomal has been observed. Of these, metachromatic origin that catalyze the hydrolysis of synthetic leukodystrophy (MLD) and multiple sulfatase and natural arylsulfate esters (1,2) as well as deficiency (MSD) are the best characterized the presumed natural substrate galactocerebro- (4). Because of its medical and physiological side sulfatide (3). There are several genetic significance, arylsulfatase A has been purified 362
Transcript
Int. J. Peptide Protein Res. 26, 1985, 362-372
Purification of mammalian arylsulfatase A enzymes by subunit affinity chromatography
ABDUL WAHEED and ROBERT L. VAN ETTEN
Department o f Chemistry, Purdue University, West Lafayette, IN, USA
Received 29 January, accepted for publication 4 April 1985
Rabbit liver arylsulfatase A (arylsulfatase sulfohydrolase, EC 3.1.6.1) monomer was immobilized on cyanogen bromide-activated SepharosedMB and on Affi- Gel-10 under various experimental conditions in order to study the effects of variables in sulfatase monomer/oligomer subunit affinity chromatography. First, the number of reactive groups on activated Sepharose-6MB and Affi-Gel-10 was determined by a procedure involving spectrophotometric titration with L- tyrosine. After covalent coupling of sulfatase monomers to the gels, the enzyme binding capacities of the sulfatase subunit affinity gel matrixes were determined at pH 4.5. The maximum binding of free monomers from solution could be achieved when the Affi-Gel-10 protein monomer matrix was prepared at low degrees of covalent loading. The introduction of a batch technique for equili- bration of the protein sample with the monomer affinity matrix also increased the efficiency of the subunit affinity gel in purification procedures. The effect of pH on the stability of the heterodimers formed between monomers of rabbit liver arylsulfatase A immobilized on Affi-Gel-10 and free monomers of aryl- sulfatase A enzymes from different tissues and organisms was studied using the batch technique. For all sulfatase A enzymes tested, the midpoint of the pH transition for subunit association was pH 6.2, suggesting that the amino acid residues involved in the dimerization are similar. The versatility of the Affi-Gel- 10 monomer affinity matrix was further demonstrated by purifying 13 mam- malian arylsulfatase A enzymes to homogeneity, as assessed by Sephacryl chromatography, native and SDS gel electrophoresis. The molecular weights of the homogeneous monomers and their peptide subunits were in the range of 110-1 80KDa and 50-64 KDa, respectively. The amino acid compositions of these enzymes were also determined.
Mammalian arylsulfatase A sulfohydrolases diseases in which a deficiency of arylsulfatase A (EC 3.1.6) are acid hydrolases of lysosomal has been observed. Of these, metachromatic origin that catalyze the hydrolysis of synthetic leukodystrophy (MLD) and multiple sulfatase and natural arylsulfate esters (1,2) as well as deficiency (MSD) are the best characterized the presumed natural substrate galactocerebro- (4). Because of its medical and physiological side sulfatide (3). There are several genetic significance, arylsulfatase A has been purified
362
Arylsulfatase A
300 units per mg of protein (8). Affi-Gel-10, cyanogen bromide-activated SepharosedMB, concanavalin A-Sepharose, Sephacryl S-200 and DEAE-Sepharose were purchased from Pharmacia Fine Chemicals Company. Human liver samples were from autopsy specimens. Human placenta was collected from a local hospital and frozen at - 20". Human urine was collected from individuals and stored at 4" as a suspension in 60% of saturation ammonium sulfate. The media from HeLa and KB cells were collected at the Cell Culture facility of the Purdue Cancer Center and were stored at 4" before use. Bovine testes were obtained from a local slaughter house and frozen at -20". Acetone powders of liver tissue from sheep, pig, horse, dog, cat, whale, seal, together with crude sulfatase preparations from limpet, abalone and Helix pomatia were from Sigma Chemical Co., St. Louis, MO. Opossum liver was obtained from a freshly killed animal. Other reagents used were of analytical grade.
and characterized from several mammalian sources. Most purifications of arylsulfatase A were carried out using tedious and inefficient conventional procedures. In a few cases, the purification of arylsulfatase A from animal tissues has been performed by purification protocols involving affinity chromatography on concanavalin A-Sepharose (5,6) or on p-suc- cinylaminocatechol-Sepharose (7).
Because the enzyme is not very abundant, studies of this protein have been very difficult. Based on our previous studies of the reversible, pH-dependent dissociation-association behavior of rabbit liver arylsulfatase A (8), we developed a new approach for sulfatase purification involving subunit affinity chromatography and demonstrated its potential by purifying the arylsulfatase A from two different mammalian tissues (9). In this procedure, one sulfatase A monomer is covalently bound to a column matrix and used as an affinity ligand for a monomer from solution. However, with our previous subunit affinity column, only 50% of the immobilized monomers of rabbit liver arylsulfatase A were available for the binding of free monomers. Moreover, the attainment of maximal binding of free monomers from solution required relatively long equilibration times (9). Therefore, the objectives of the present work were to develop a new subunit affinity material that would have an optimal, high binding capacity for free monomers from solution, and that would require a minimal amount of time to reach equilibrium in the binding of free monomers to the covalently immobilized monomers. We now describe such experiments. In addition, the considerable potential of the subunit affinity methodology was further established by purifying numerous mammalian arylsulfatase A enzymes to homo- geneity, most of them for the first time. The amino acid compositions of these pure enzymes are also given.
MATERIAL AND METHODS
Materials Initial amounts of rabbit liver arylsulfatase A were purified according to a literature procedure (8). The homogeneous rabbit liver arylsulfatase A that was used for immobilization on the affinity gels had a specific activity of
Methods
Covalent attachment of rabbit liver arylsulfatase A monomer on Affi-Gel-10 and Sepharose- 6MB. Immobilization of arylsulfatase A monomers on Sepharose-6MB and on Affi-Gel- 10 beads was carried out using our previous procedure (9) or according to a manufacturer's procedure (1 0), respectively, using a coupling buffer of 0.1 M HEPES pH 7.5. Under these conditions of pH and ionic strength, rabbit liver arylsulfatase A exists as a monomer (6). At the end of the coupling reaction the gel-enzyme complex was washed with 25 vol. 0.1 M HEPES pH 7.5 + lOmM ethanolamine in order to block any remaining reactive sites. The resulting gels, containing immoblized monomers of arylsulfatase A, were stored in 0.1 M HEPES pH 7.5 + 1OmM glycine at 4".
The coupling of rabbit liver arylsulfatase A monomer to the Affi-Gel-10 or Sepharose- 6MB was also performed using different concentrations of reactive groups on the affinity matrix. The density of the reactive groups was adjusted by first reacting different amounts of ethanolamine with the activated gel. Since lysine residues are involved in the covalent immobilization of the proteins, the number of lysine residues per enzyme monomer
363
A. Waheed and R.L. Van Etten
were estimated from the amino acid compo. sition of rabbit liver arylsulfatase A (1 1).
Determination of reactive groups on Sepharose- 6MB. CNBr-activated Sepharose-6MB (1.5 g) was washed with 250ml 1 mM HCl followed by 400ml cold water and 50ml of 0.1 M HEPES pH 7.5. The gel was suction-dried, 0.25 g portions of the gel were placed in a tube, 3.5-8.8pmol tyrosine in HEPES buffer was mixed in a final volume of 8.01111, and the mixture was incubated at room temperature for 6 h . The supernatant was removed and its absorbance at 280nm was measured. Using a molar extinction €280 = 1200 M - ' cm-' for tyrosine at neutral pH (12), the free and bound tyrosine concentrations were estimated. A graph of amount of the tyrosine added versus the amount bound revealed a sharp break in the titration curve, indicating that the reactive groups on Sepharose-6MB were saturated with tyrosine. From this break point, the amount of reactive groups per gram of SepharosedMB was calculated.
Determination of the immoblized arylsulfatase A concentration and activity. A known amount of the enzyme-Affi-Gel-10 or enzymesephar- rose-6MB complex was hydrolyzed in 1 N NaOH at 98" for 10 min. For controls, identical amounts of Affi-Gel-10 or Sepharose-6MB was also hydrolyzed. The hydrolyzates were then used for protein determination using the Lowry et al. procedure (1 3).
The enzymatic activity of the immobilized arylsulfatase A was determined by incubating 501-11 of a suspension of the immobilized enzyme in 500pl of 5 mM nitrocatechol sulfate in 0.5 M sodium acetate pH 5.5 buffer at 37'. At various times, reaction mixtures were quenched with l m l of 1 N NaOH. The absorbance at 5 15 mm was recorded against an appropriate control. The enzyme activity of soluble and immobilized arylsulfatase A was calculated using a molar extinction coefficient of 12 400 M - ' cm-' for nitrocatechol at 515 nm in 1 M NaOH. One unit of enzyme activity is defined as the amount of enzyme required to hydrolyze 1 pmol of 4-nitrocatechol sulfate in 1 min at 37'C.
Partial purification of arylsulfatase A . Fresh tissues or acetone powders of livers of different animals were homogenized in 3 vol. of 50 mM sodium acetate pH 6.0 buffer. To fresh tissue homogenates at 4", 1 mg protamine sulfate per gram of tissue was added 30min before cen- trifugation. The homogenates were centrifuged at 15 000 r.p.m. for 30 niin using a Beckman JA-20 rotor. The resulting supernatant was saved and the residue was extracted once more with an equal volume of 50 mM sodium acetate pH 6.0 buffer. The culture media obtained from KB and HeLa cells were adjusted to 50mM sodium acetate pH 6.0 by addition of 1 M pH 6.0 sodium acetate buffer. The super- natant from the tissue extract or the cell medium was loaded on a concanavalin A- Sepharose column (50 ml) equilibrated with 50mM sodium acetate pH 6.0 buffer. The column was washed with 5 bed volumes of 50mM sodium acetate pH 6.0 at 4' followed with 5 bed volumes of 50mM sodium acetate and 1 M NaCl at room temperature. The bound protein was eluted with 3 bed volumes of 50mM sodium acetate pH 6.0 + 1 M NaCl + 10% (w/v) D-mannose at room temperature. The post concanavalin A-Sepharose enzyme preparation was concentrated by ultrafiltration using a PM-10 membrane, or it was precipitated at 60% of saturation with ammonium sulfate, at pH4.5. The enzyme preparation was first dialyzed against 50mM sodium acetate, pH 4.5, and then against 50mM HEPES pH 7.5 + 20 mM glycine buffer and stored at 4'.
Subunit-affinity chromatography. Rabbit liver arylsulfatase A monomer - Affi-Gel-10 complex, (3rd) with a binding capacity of 0.5 mg arylsulfatase A protein was suspended in 1.5 ml of the enzyme preparation in 50 mM HEPES pH 7.5 4- 20mM glycine. The mixture was gently stirred at 4' for 60min in order to achieve an equilibrium distribution of solution monomers with gel-bound monomers. (More prolonged stirring did not increase the binding capacity of the affinity complex). In order to initiate the association of the free monomer to the immobilized monomer, the pH of the mixture was changed to 4.5 by addition of 1.5 ml of 0.15 M sodium acetate pH 4.5 buffer
364
plus 0.1 8 M NaCl and the suspension was stirred for another 60min. The mixture was then either packed in a small column or centrifuged at 1200 r.p.m. for 5 min in a Beckman JA-20 rotor in order to separate any non-associated enzyme. The gel in the column or in the centrifuge tube was washed with 20ml of 50mM sodium acetate pH 4.5 buffer. The bound (associated) enzyme was eluted with 20ml of 50 mM HEPES pH 7.5 buffer + 20mM glycine. The enzyme preparation was concentrated by ultrafiltration using a PM-I 0 membrane. The homogeneity of the enzyme was checked by passing through a Sephacryl S-200 column which was equilibrated in 50 mM Tris-HC1 pH 7.5 + 0.1 M NaCl. The fractions containing significant amounts of enzyme activity were pooled and concentrated.
Gel filtration chromatography. A Sephacryl S-200 column (2.5 x 50cm) was packed and equilibrated in 50mM Tris-HC1 pH 7.5 + 0.1 M NaCl buffer. The column was calibrated using standard proteins including immuno- globin (IgG), 160 kDa; bovine serum albumin, 69 kDa and ovalbumin, 46 kDa. The relation- ship between elution volume, Ve and log molecular weight, MW was found to be log MW = 7.44 - 0.0201 Ve. The enzyme preparation (2.5-5ml in the pH 7.5 Tris buffer) was applied to the column. The column was eluted with a flow rate of 40ml/h and fractions of 3.5 ml were collected. Each fraction was monitored for enzyme activity.
Electrophoresis. Polyacrylamide gel electro- phoresis in sodium dodecyl sulfate under re- ducing conditions was performed according to Laemmli (14). Prestained protein standards were used for the determination of the mol- ecular weight of the arylsulfatase A subunits. Polyacrylamide gel electrophoresis of the native arylsulfatase A at pH 7.5 was conducted according to (15). The gels were stained for protein and for enzymatic activity using a- naphthylsulfate as a substrate with fast red TR salt in 0.5 M sodium acetate pH 5.5 buffer at 37'.
Amino acid analyses. Homogeneous preparations of arylsulfatase from different mammalian
Arylsulfatase A
sources (approx. 5Opg) were desalted on PD-I0 (Sephadex G-25 M) column and lyophilized. The lyophilized samples were hydrolyzed in 6 N HCI at 110' for 24h. The hydrolyzed samples were analyzed for amino acid content on a Durrum model 110 amino acid analyzer. The number of residues per subunit were also calculated using an integer fit method (16). Amino acid residue numbers were not corrected for timedependent decomposition or release.
RESULTS AND DISCUSSION
Titration of reactive groups on the affinity matrix Since the concentration of reactive groups can affect the degree of crosslinking of the protein to the affinity matrix, the concentration of the reactive groups on the activated gel was fist determined using L-tyrosine as a reactive probe. A typical titration curve is shown in Fig. 1. The concentration of L-tyrosine at the point where all of the reactive sites on the Sepharose-6MB or Affi-Gel-10 beads were saturated was taken as a measure of the concentration of reactive sites per gram of suction-dry gel. A typical result for activated SepharosedMB was 17.6 k 0.3 pmol/g gel, and similar values were
0 0
D
6 0 I TYROSINE ADDED ip mole1
FIGURE 1 Titration of reactive groups on an affinity matrix. SepharosedMB (0.25 g of suction-dried gel) was incubated at pH 7.5 with amounts of btyrosine ranging between 1 and 8.8pmol for 6 h at room temperature. The amount of bound tyrosine was determined by subtracting the amount of free tyrosine in the supernatant from the total tyrosine added. The break point in the titration curve was used to calculate the reactive group concentration.
365
A. Waheed and R.L. Van Etten
determined for Aff-Gel-l O. From the manufac- turer’s data, the concentration of reactive groups on Affi-Gel-10 was calculated to be 22.5 k 4 . 5 p o l per gram. This number was calculated using the fact that 1 g suction-dry gel occupies a volume of 1.5 ml, and 1 ml Affi- Gel-10 has 15 f 3pmol reactive groups (10). The value determined by reaction with tyrosine is more precise than the one calculated from the manufacturer’s data, and our procedure provides a simple and reliable method for determination of reactive groups on an affinity matrix material immediately before use. While there is a pyridine-barbituric acid spectropho- tometric procedure (1 7) for the determination of cyanate esters on the Sepharose affinity matrix, it is not clear whether the same procedure could be used for affinity matrix system in which reactive groups are not cyanate esters. For example, Affi-Gel-10 has N- hydroxysuccinimide esters as reactive groups. Thus, the present procedure seems to be more versatile and is probably more accurate than estimates made at the time of packaging.
Immobilization conditions The effect of the covalent coupling procedures on the free monomer binding capacity of immobilized monomers of rabbit liver aryl- sulfatase A was investigated. In the present study, we selected two different types of
affinity ligand supports: Sepharose-6MB, which was used earlier (9), and Affi-Gel-10, which has a 1 0 A (six carbon) long spacer arm. Another important parameter for immobilization was the density of loading of the covalently coupled enzyme monomer and the related extent of protein crosslinking. This was systematically changed by controlling the initial concentration of the reactive groups on the affinity matrix. The experimental results are shown in Table 1. The binding capacity of Sepharose-6MB- arylsulfatase A monomer matrix was 47%, a value that is similar to our previous results (9). This value could not be increased even when the density of coupled monomer (and presumably the extent of crosslinking) was decreased by 200-fold. In contrast, the binding capacity of the Affi-Gel-10 arylsulfatase A monomer matrix was significantly increased, from 40-50% to 94-100%. These results show that the sulfatase monomer immobilized on an Affi-Gel-10 matrix under conditions of reduced crosslinking provides an excellent ligand for the association of free monomers from solution. However, the dissociation constant of the free enzyme dimer in buffer solution is about five times smaller than is the dissociation constant determined for the monomer from the immobilized monomer on Affi-Gel-l O, suggest- ing that the conformational accessibility or the structure of the immobilized monomer is
TABLE 1 Association of rabbit liver arylsulfatase subunits to subunit-Affi-Gel-10 affinity matrixes‘
Enzyme applied in solution
(units) Enzyme reversibly bound (units) Gel reactive sites during synthesis
bmol/g)b
22 11 1 .o 0.1
2.5 5.2
10.4 15.6
1.4 2.1 2.5 2.8 4.0 4.8
4.9 5.4 8.4 6.1 6.8 10.7
(2.1P
2.5 5.1 (2.8) 10.1 14.7
‘Rabbit liver arylsulfatase A (22 units) was coupled to 1 g Affi-Gel-10 or SepharosedMB. bEthanolamine was first added to achieve the required concentration of reactive sites on the gels before covalent coupling of the 22 units of enzyme monomer. ‘Numbers in parentheses indicate the association of enzyme from solution onto a subunit Sepharose-6MB affinity matrix; other entries refer to the Affi-Gel column.
366
Arylsulfatase A
altered from that of the free monomer. In any event, Affi-Gel-10 affords a very suitable affinity matrix.
During studies on the rate of binding of the solution monomer to the immobilized monomer, we observed that the binding of the free monomer to the Affi-Gel-10-bound monomer required only 1 h to achieve the maximum binding whereas for the Sepharose- 6MB-bound monomer it was 20h. This result is evidently due to a higher accessibility of the immobilized monomer on the Affi-Gel-10 than on Sepharose-6MB. Presumably this is because in Affi-Gel-10, a six carbon atom spacer arm is connected to the reactive groups, and this maintains the immobilized monomer more accessible to solvent. This conclusion is consistent with an experiment in which the apparent accessibility of the immobilized monomer was measured by the rate of substrate hydrolysis (Fig. 2). The rate of hydrolysis of 4-nitrocatechol sulfate by the Affi-Gel-10 enzyme-bound monomer was 1.12pmol/min, a value that was nearly two-fold greater than the value for the rate of hydrolysis catalyzed by the Sepharose 6MB bound monomer. While we consider that these results are due to the greater accessibility of the immobilized Affi-Gel monomer towards solvent, they may indicate that deleterious structural changes occurred during the covalent coupling process involving cyanogen bromide activation and the Sepharose medium.
Sample application during affiniv chromatography Column techniques and batch techniques represent two possible ways to use the imnio- bilized monomer for the purification of the enzyme from crude protein solutions. Earlier, we used the column technique because of the fragility of Sepharose beads toward mechanical shaking, even though the column technique required a long time to attain the maximum extent of binding of the free enzyme (5 ) . Now, we have developed a batch technique using the Affi-Gel-10-monomer matrix in which the maximum binding can be achieved in 1 h. This step has further improved the efficiency of the present affinity separation method. Importantly, there is no change in the binding capacity of
30 -
0 0 0 -
a
a
0 10 20 30 INCUBATION TIME (min)
FIGURE 2 Hydrolysis of 4-nitrocatechol sulfate by an immobilized monomer of rabbit liver arylsulfatase A. Immobilized enzyme preparations having identical amounts of enzyme protein (120 p g ) coupled either to Affi-Gel-10 (0-0) or to SepharosedMB (0-0) were incubated with 5 mM 4-nitrocatechol sulfate in 0.5 pm sodium acetate pH 5.5 buffer at 37" for different times. Hydrolyzed substrate was determined by measuring the absorbance at 515 nm after quenching the reaction with 1 M NaOH.
the affinity complex after repeated use in the batch technique. Because it is a simple and rapid procedure, we. have used the batch technique to study the dissociation constants of different mammalian arylsulfatase A enzyme monomers from the immobilized monomer of rabbit liver arylsulfatatase A on the Affi-Gel-10 matrix (18).
Effect of pH on the binding of free monomer to the rabbit liver arylsulfatase A monomer immobilized on Affi-Gel-I 0 Using the batch technique, we studied the effect of pH on the stability of the dimer formed between monomers of rabbit liver, bovine testis or human liver arylsulfatase A and
367
A. Waheed and R.L. Van Etten
07
06
05
E n w - In 4
03
02
01
ao-
immobilized monomers of rabbit liver arylsul- fatase A. Results are shown in Fig. 3. The mid- point of the pH transition was pH 6.2 and it was identical for all three different enzymes. This result suggests that the contact interfaces of the two monomers of the different mam- malian enzymes may have similar amino acids. The nature of the curve suggests that there is more than one amino acid residue involved in the interaction of the two monomers, and the pH dependence of the interaction indicates it to be highly cooperative. Although the amino acid residues of the subunit contact region may be highly conserved in the mammalian enzyme, this is evidently not true for other classes of vertebrates. This is suggested by the fact that the arylsulfatase A from opossum, a marsupial, does not bind to the present affinity matrix. In this regard, it was interesting to find that opossum liver enzyme did not exhibit the
-
-
-
-
-
-
- 0
120 fi n o -
g 80- 2 -
6 0 - (0
a 4 0 - $!?
a9
20-
0
PH
FIGURE 3 Effect of pH on the stability of the dimers formed between an immobilized rabbit liver arylsulfatase A monomer and a free monomer of another arylsulfatase A. Arylsulfatase A from rabbit liver (0-o), bovine testis (A-A) or human livcr (0-0) separately, in 0.5 ml 50mM HEPES pH 7.5 containing 33pg enzyme protein was mixed with 1.5 ml AffiGel-10-rabbit liver arylsulfatase A monomer matrix + 1.5 ml of 50mM HEPES pH 7.5 buffer at 4" for 1 h. The pH of thc slurry was changed to 4.5 with 2ml of 140mM sodium acetate pH 4.5 buffer and mixing was continued for another h. The affinity mixture was packed in a small column and the unbound enzyme was removed by washing with 50 mM sodium acetate pH 4.5 buffer. The bound enzyme was eluted with a buffer increasing in pH from pH 4.5 to 7.5.
368
reversible, pH-dependent monomer-dimer interconversion under experimental conditions where mammalian enzymes do undergo such a characteristic dimerization.
Purification of mammalian arylsulsulfatase A by subunit affinity chromatography The rabbit liver subunit column appears useful in purifying most mammalian arylsulfatase A enzymes. Enzyme preparations obtained from concanavalin A-Sepharose chromatography were used for the sulfatase subunit affinity chromatography experiments. Typical chromatographic profiles for the purification of a few enzymes are shown in Fig. 4. The initial peak in the chromatogram corresponds to unbound, excess arylsulfatase A, arylsulfatase B and contaminating proteins. (Data on the protein concentration of each fraction are not shown). As seen in Fig. 4, the arylsulfatase A from Helix pomatia did not bind on the affinity column. However the enzymes from all the mammalian sources tested here were retained on the affinity column. As examples, aryl- sulfatase A from the cell culture medium of K B cells and the enzyme from rabbit liver were retained on the affinity column as indicated by the chromatographic profiles shown in Fig. 4.
4 0 16 20 12 FRACTIONS
FIGURE 4 Chromatographic profiles of different arylsulfatase A enzymes on the Affi-Gel-10 rabbit liver arylsulfatase A monomer affinity column. The enzyme activity of rabbit liver (0-o), K B cells ( 0 - 0 ) and Helix pornaria (0 -0) in each fraction was monitored using 4-nitro- catechol sulfate as a substrate at 5 15 nm.
Arylsulfatase A
purified in this way by subunit affinity chromatography, we utilized gel chromato- graphy, native gel electrophoresis and SDS gel electrophoresis. The enzyme preparations were applied to a calibrated Sephacryl S-200 column at pH 7.5. All the arylsulfatase A enzyme preparations exhibited a single symmetrical peak, consistent with the molecular homo- geneity of the protein. The elution volume of each enzyme was used to estimate the molecular weight of the monomer of each enzyme. The results are given in Table 2. The molecular weights of the sulfatase monomers range between I10 and 182kDa. This wide variation in the molecular weight is due in part to variations in the carbohydrate content. It should be noted that all arylsulfatase A preparations that were used were also retained on a concanavalin A - Sepharose column, indicating that these sulfatases are glyco- proteins. Native gel electrophoresis of the purified arylsulfatase A enzymes showed single, coincident bands when stained for protein with Coomassie blue and for sulfatase activity using a-naphthyl sulfate. The purity of the enzymes were also tested by sodium dodecyl sulfate polyacrylamide gel electrophoresis and the resulting molecular weights of the peptide subunits are given in Table 2. The arylsulfatase A enzymes from all human sources (except that from the medium of cultured HeLa cells)
The rabbit liver monomer column has also been used to purify human urine arylsulfatase A (19). Elution of the bound enzyme by application of a pH 7.5 buffer resulted in nearly homogeneous arylsulfatase A enzymes. Pre- parations from several such column runs were collected and concentrated for further work.
In addition to the arylsulfatase A from Helix pomatia, the enzymes from abalone, limpet and opossum also did not bind to the affinity column. Significantly, the apparent molecular weights of the monomers of these enzymes did not change when the pH was changed from 7.5 to 4.5. These results suggest that the binding of arylsulfatase A to the subunit affinity matrix is a highly specific protein-protein interaction, consistent with our earlier results (5), and with the conclusion that it is a typical characteristic of mammalian arylsulfatase A enzymes that normally undergo a pH dependent monomer- dimer (or related) association. The physiological significance of this association phenomenon is not yet understood.
Characterization of homogeneous aryl sulfatase A enzymes
Molecular weights of the monomers and their peptide subunits from different sulfatase A enzymes. In order to check the homogeneity of the many new arylsulfatase A enzymes
TABLE 2 Molecirlar weights of arylsulfatase A monomers and peptide sicbunits and enzyme specific activities
Enzyme
Monomer Peptide subunit size Specific activity molecular (kDa) (units/mg)
weight I 11 (kDa)
Human liver Human placen Human urine K B cells HeLa cells Bovine testis Porcine liver Dog liver Cat liver Horse liver Sheep liver Whale liver
perhaps even identical. This conclusion is also consistent with immunological studies, where rabbit anti-bovine testis arylsufatase A antisera precipitated identical amounts of the enzyme from bovine liver and testis preparations, and also with the fact that the antiserum showed a single precipitin line for both enzyme prepara- tions (A. Waheed and R.L. Van Etten, unpublished results). However, the reported molecular weights of the subunits of the ‘ox liver enzyme (21) in comparison to the value observed for the bovine testis enzyme (Table 2), merits comment. The earlier report by Roy & Jerfey indicated the presence of 55 and 27 kDa polypeptides, and they concluded that 27 kDa polypeptide may be the “basic unit” of the ox liver enzyme (see Summary in ref. 21). However, our bovine testis enzyme preparation migrated as a single polypeptide of 64kDa (see Table 2) on sodium dodecyl- polyacrylamide gel electrophoresis and we did not observe any low molecular weight poly- peptide. Therefore, the appearance of the low molecular weight polypeptide in earlier studies (21) may have been due to proteolytic nicking of the enzyme molecule during the purification procedure. The present pruification of aryl- sulfatase A from mammalian sources is highly efficient and quite rapid and this rapidity may reduce possible proteolytic nicking of the enzyme molecules.
The present study provides an efficient and simple affinity procedure for the pruification of arylsulfatase A enzymes, even from small amounts of valuable biological fluids, and this should facilitate studies of structure-function relationships involving this protein molecule. During our studies of this procedure, we also developed a useful new method for the determination of reactive groups on typical affinity matrix derivatives, as well as approaches to the immobilization of proteins on affinity matrixes that result in a minimum degree of crosslinking and maximal preservation of the native structure of the immobilized protein molecule. Structure and amino acid compo- sition data for 13 different arylsulfatase A enzymes are also provided. These results should be useful in a variey of comparative studies.
showed two nonidentical subunits. All other mammalian arylsulfatase A enzymes showed only a single type of peptide subunit. Similar nonidentical subunits have been reported from human enzymes such as the urine enzyme (19). The molecular weights of the peptide subunits of all the enzymes tested here were in the range of 50-66 kDa. These results are consistent with the conclusion that the enzyme preparations obtained by sulfatase subunit affinity chroma- tography were homogeneous and that their molecular properties resemble those of other wellcharacterized arylsulfatase A enzymes.
Enzyme specific activities and amino acid compositions. The specific activities of the homogeneous arylsulfatase A enzymes from different animals varied in the range of 3 1-210 units/mg of protein. In cases where the same initial extract has been used to purify an arylsulfatase A by both a conventional and the subunit affinity method, yields by the latter approach are better by more than a factor of two (9,19).
The results of amino acid composition determinations for the different arylsulfatase A enzymes are given in Table 3. (See also ref. 20 for another recent example). Per lOOOOOg protein, the number of amino acid residues (excluding tryptophan and in few cases half- cystine) range between 921 and 1009. The amino acid compositions of all the mammalian arylsulfatase A enzymes are generally similar and show some interesting features that are perhaps characteristic of arylsulfatase A. These include a high content of proline and of hydro- phobic amino acid residues (leucine, isoleucine, tyrosine and phenylalanine). The high proline content feature has been discussed in connection with the secondary structure of the enzyme (11). During amino acid analysis, significant amounts of glucosamine were also observed, which is consistent with the con- clusion that these arylsulfatase A enzymes are glycoproteins of lysosomal origin (1).
When we compare amino acid composition of bovine testis arylsulfatase A with the amino acid composition of the ox liver enzyme (21), we conclude that both enzymes are similar and
371
A. Waheed and R.L. Van Etten
ACKNOWLEDGMENTS
This work was supported by NIH grant GM 22933 from the National Institute of General Medical Sciences. We thank James Cook and Prof. Michael Laskowski for amino acid analyses and Anne Gill and Dr. Linda B. Jambsen for providing the media from KB cells and HeLa cells.
REFERENCES
1. Roy, A.B. (1960) Biochem. .I. 77,380-386 2. Roy, A.B. (1971) in The Enzymes (Boyer, P.,
ed.), Vol. 5, pp. 1-19, Academic Press, New Y ork
3. Mehl, E. & Jatzkewitz, H. (1968) Biochim. Biophys. Acta 151,619-627
4. Moser, H.W. & Dulaney, J.T. (1978) in The Metabolic Basis of Inherited Disease (Stanbury, J.B., Wyngaarden, J.B. & Fredrickson, D.S., eds.), pp. 770-809, McGraw-Hill, New York
5. Balasubramanian, A S . & Bachhawat, B.K. (1975) Biochini. Biophys. Acta 403,113-131
6. Waheed, A. & Van Etten, R.L. (1979) Arch. Biochem. Biophys. 194,215-225
7. Derappe, C. & Dubois, G. (1976) Experientia 32, 1354
