Multiple Enzyme Purifications from Muscle Extracts by Using Affinity-Elution- Chromatographic Procedures

Biochem. J. (1977) 161, 265-277 265 Printed in Great Britain

Multiple Enzyme Purifications from Muscle Extracts by using Affinity-Elution- Chromatographic Procedures

By ROBERT K. SCOPES Department ofBiochemistry, La Trobe University, Bundoora, Vic. 3083, Australia (Received 20 July 1976)

1. Starting with (NH4)2SO4 fractions of muscle extracts, procedures for purifying four to six separate enzymes from each fraction by using affinity-elution-chromatographic techniques are described. 2. Schemes for purifying 12 separate enzymes from rabbit muscle, and eight from chicken muscle extracts, are included. In nearly all cases the overall procedure involves three steps: the initial (NH4)2SO4 fractionation, the ion- exchange chromatography with affinity elution of the enzyme, and gel filtration. The specific activities of the enzymes so purified are comparable with the highest values in the literature. 3. The five schemes described include illustrations of affinity elution of the separate enzymes at different pH values, at different ionic strengths and in combination with conventional gradient elution. They also include stepwise adsorption on columns at different pH values. 4. Separation of two electrophoretically differing forms of phosphoglycerate kinase was achieved by gradient affinity elution from CM-cellulose. The lower-pI form was eluted by a lower concentration of substrate than the higher-pI form.

Methods for purifying individual glycolytic and of extract (approx. 350g of muscle), and the columns related enzymes from rabbit muscle by using affinity- used were always 16cm2 cross-section and close to elution techniques were described in the preceding 5cm tall unless otherwise indicated. The buffers used paper (Scopes, 1977). Although most investigators in chromatography were 10mM of base (Tris, KOH, are interested in only one or perhaps two different NaOH) adjusted to the required pH with a-picolinic enzymes, it may be desirable to purify others to use, acid, Mes,* Mops or Tricine as appropriate. The for example, in assaying the enzymes or the substrates buffers also contained 0.2mM-EDTA. being investigated. In this laboratory all glycolytic enzymes are required in large quantities both for Results assaying purposes and for metabolic reconstitution studies (Scopes, 1973). Use ofthe affinity-elution pro- Five schemes are presented here (three for rabbit cedurehas beenwidely adopted by us (Scopes & Fifis, enzymes, two for chicken) which cover 12 different 1975; Stewart & Scopes, 1975; Chappel et al., 1976; enzymes. Phosphorylase (EC2.4.1.1), phosphofructo- Scopes, 1977), and the present paper describes kinase (EC 2.7.1.11), glycerol phosphate dehydrogen- procedures for purifying several different enzymes ase (EC 1.1.1.8) and glyceraldehyde phosphate simultaneously from extracts of rabbit muscle. dehydrogenase (EC 1.2.1.12) are not included; In addition, methods for purifying some of the alternative techniques are more convenient for these corresponding chicken muscle enzymes by affinity enzymes (Scopes, 1973). In particular, phospho- elution are described. fructokinase is mostly destroyed by (NH4)2SO4 fractionation at pH6.0; phosphorylase and glyceraldehyde phosphate dehydrogenase form the major Materials and Methods portions of fractions A and E respectively, and each Materials and basic methods were as described in can be crystallized directly after suitable treatments. the preceding paper (Scopes, 1977). In addition, chicken muscle from freshly killed hens was obtained Scheme (a): purification of pyruvate kinase (EC and extracted exactly as therabbit muscle. (NH4)2SO4 2.7.1.40), lactate dehydrogenase (EC 1.1.1.27), fractions obtained as described were dissolved in a aldolase (EC 4.1.2.13) and phosphoglucomutase (EC small volume of Tris/HCl buffer, pH 8.0, and the pH 2.7.5.1)from the 45-60%-satd. (NH4)2S04fraction was adjusted to 8.0 with 1M-Tris before desalting. Fraction BC contains between 70 and 80 % of each The fractions were often stored as a slurry at 4°C of the aldolase, lactate dehydrogenase and pyruvate before desalting, since it was not possible to work up * Abbreviations: Mes, 2-(N-morpholino)ethanesul- more than one fraction at a time. phonic acid; Mops, 3-(N-morpholino)propanesulphonic In all cases procedures relate to material from 1 litre acid; Tricine, N-tris(hydroxymethyl)methylglycine. Vol. 161 266 R. K. SCOPES

kinase present in the extract, these enzymes being the major protein components of the fraction. Proteins in smaller amounts include creatine kinase (EC 00 2.7.3.2), phosphoglucomutase, phosphoglycerate e4 - mutase( EC 2.7.5.3), phosphoglucose isomerase C ._ UE (EC 5.3.1.9) and enolase (EC 4.2.1.11); some of these (A are more concentrated in other fractions. The procedure described here gives four of the main components, together with some others isolated in a) lesser overall yield. cd Fraction BC from 1 litre of extract was first desalted, then adjusted to pH6.5 with Tris/Mes buffer, and diluted to a protein concentration of between 10 200 250 300 350 and 15mg/ml (approx. 400ml). A CM-cellulose Eluate volume (ml) column (of dimensions 16cm2x 10cm) in Tris/Mes buffer was used first, and the sample run on under Fig. 2. Separation ofimpuritiesfrom pyravate kinase bki gel gravity (300-400ml/h). The sample was washed in filtration with buffer, and KCI was added to the buffer to a The fraction applied to the Sephadex G-150 column (8cmnx90cm) was the phosphoenolpyruvate-eluted concentration of 20mM. Approx. 300ml of this was fraction in Fig. 3, after concentration by ultrafiltra- pumped in (300ml/h), and a protein peak, which tion to 25ml. , Pyruvate kinase. - , Percent- included most of the phosphoglucose isomerase in age transmission at 280nm. the fraction, was obtained. This isomerase could be further purified by affinity elution from phospho-

cellulose as described in the preceding paper (Scopes, 1977). 1 2 3 4 5 6 7 8 9 After the column had been equilibrated with Tris/ Mes/KCl buffer, 0.5mM-phosphoenolpyruvate was 100 e added to the buffer [a 'dummy-substrate' wash 80 (Scopes, 1977) was not necessary at this higher ionic - PGM __ strength], which specifically eluted pyruvate kinase, 60 m - -_ _/ PK \ PGI -9 somewhat contaminated with enolase, phospho- C PGK glycerate mutase and phosphoglycerate kinase. These ,EN u --CK 40 ALD contaminants are clearly identifiable by their subunit 0 - LDH sizes on sodium dodecyl sulphate/polyacrylamide-gel x 30 _~~~~~~~ = - PGAM electrophoresis (Fig. 1). They were present because

0 during the affinity-elution step the enolase and mutase combined to convert the added phosphoenol- 20 pyruvate through 2-phosphoglycerate into 3-phosphoglycerate, so causing affinity elution of all four enzymes; however, pyruvate kinase was the major Fig. 1. Diagram of a sodiwn dodecyl sulphate/polyacryl- component, and being a much larger molecule than amide-gel-electrophoresis slab containing samples relevant the other enzymes was subsequently completely to Schemes (a) and (b) separated from the others by gel filtration on Sepha- Sample 1: fraction BC complete; 2: fraction BC not dex G-150 (Fig. 2). adsorbed on CM-cellulose at pH 6.5; 3: phosphoenolpyruvate-eluted pyruvate kinase, Scheme (a); The column now contained only lactate dehydro- 4: pyruvate kinase after gel filtration (Fig. 2); 5: genase and aldolase in significant amounts. The buffer lactate dehydrogenase eluted by NADH, Schemes was changed to KOH/Mops, pH7.2, containing (a) and (b); 6: aldolase eluted by fructose 1,6- 10mM-KCl, and after 200ml of this had been passed bisphosphate, Schemes (a) and (b); 7: phosphogluco- into the column 0.2mM-NADH was added to it mutase eluted by glucose phosphates, Scheme (a); (150ml), specifically eluting lactate dehydrogenase. 8: phosphoglucose isomerase eluted by glucose 6- The column was next washed with KOH/Mops phosphate, Scheme (b); 9: phosphoglycerate mutase buffer, pH7.2, containing 20mM-KCI, followed by Scheme eluted by 2,3-bisphosphoglycerate, (b). the same 0.2mM-fructose Abbreviations: PGM, phosphoglucomutase; PK, buffer containing 1,6- pyruvate kinase; PGT, phosphoglucose isomerase; bisphosphate (200ml), which eluted the aldolase. PGK, phosphoglycerate kinase; EN, enolase; CK, The purities of these preparations can bejudged from creatine kinase; ALD, aldolase; LDH, lactate de- Fig. 1. The complete elution diagram from this hydrogenase; PGAM, phosphoglycerate mutase. column is shown in Fig. 3. All elution diagrams in this 1977 MULTIPLE ENZYME PURIFICATIONS BY AFFINITY ELUTION 267

pH 6.5 pH 7.2 Buffer pH: D. 20 mM-KCI 10 mM-KCI Extra ionic _ strength: I la nn Pbospbhenol- 0.2me- 0.2 Imm- I pyruvate NAOH fructose 1,6-bisphosph-ate I I I I~~~~~ 13 wC 2?.a

\J t - .1 \,! 0 400 600 800 1000 1200 Eluate volume (ml) Fig. 3. Elution diagram for rabbit muscle fraction BC adsorbed to a CM-celildose column atpH6.5 The enzymes obtained by affinity elution were pyruvate kinase (---- ), lactate dehydrogenase ( ) and aldolase

(- .-.). - , Percentage transmission at 280nm in 3mm-path-length cell. For details of procedure, see the text.

pH 6.0 Buffer pH: 0.5 mm-Glucose I-phosphate 0.5 rmm-EDTA + 0.I mm-glucose .6bisphosphate Added ligand:

a 25

Ce aS 1500 -;- 'a

4 Ce .-f, C2 1000 P. §75 4) Oo 500

500 600 Eluate volume (ml) Fig. 4. Affinity elutdon ofphosphoglucomutase from a CM-cellulose column atpH6.0

- The fraction applied was that protein not adsorbed at pH6.5 (see Fig. 3). , Phosphoglucomutase. , Percentage transmission at 280nm. For details of procedure, see the text. Vol. 161 268 R. K. SCOPES paper are typical profiles selected from several through the two columns, which were then washed separate experiments. with a little buffer. Of the five enzymes noted, only The final step in this scheme was to take the protein phosphoglucomutase was not adsorbed on either that had not adsorbed on the CM-cellulose column, column, and could subsequently be purified on CM- lower its pH to 6.0 with 1 M-Mes, and run it on to an- cellulose at pH6.0 as described in Scheme (a). The other CM-cellulose column (16cm2 x 5cm), equilibra- CM-cellulose column was worked up for lactate ted at pH6.0. The enzymes adsorbed included phos- dehydrogenase and aldolase exactly as described in phoglucomutase, creatine kinase and phosphoglycer- Scheme (a). The phosphocellulose column retained ate mutase. Thelatter was denatured and could not be the phosphoglucose isomerase and phosphoglycerate recovered from the column. After a short wash ofthe mutase. The isomerase was eluted in the same pH7.0 column with Tris/Mes buffer, pH 6.0, containing an buffer, by using 0.5mM-glucose 6-phosphate. The additional 0.5mM-EDTA, the wash was changed to mutase was eluted, still at pH7.0, by using 0.2mM-2,3- Tris/Mes, pH6.0, containing both 0.5mM-glucose bisphosphoglycerate. Pyruvate kinase was also 1-phosphate and 0.1mM-glucose 1,6-bisphosphate adsorbed on the phosphocellulose column, but elution (lOOml). Both phospho and dephospho forms of of this enzyme was not attempted. The diagram of phosphoglucomutase were eluted from the column, elution from the phosphocellulose column is illus- the latter being phosphorylated in the process. This trated in Fig. 5, and the activities of the enzymes preparation was essentially pure; it has been crystal- eluted from it and from the CM-cellulose columns lized from (NH4)2SO4 in the form of large needles in are listed in Table 2. The purities of the enzymes as the presence of various bivalent metal ions. obtained by affinity elution are illustrated in Fig. 4; Creatine kinasecould be obtained from this column as with the other examples, the impurities could by the same procedure as described below in Scheme mostly be removed by gel filtration. (c); however, less than one-third of the creatine kinase from the original extract is present in fraction Scheme (c): purification of enolase, phosphoglycerate BC, and so the overall yield is not high. A diagram of kinase (EC 2.7.2.3), triose phosphate isomerase (EC the elution pattern from the second column is shown 5.3.1.1), creatine kinase and myokinase (EC 2.7.4.3) in Fig. 4. The activities ofthe enzymes and the specific from the 60-75 %-satd.-(NH4)2S04fraction activities ofthe purified preparations are summarized Fraction CD contains more than 50% of the in Table 1. enolase, creatine kinase, myokinase and phosphoglycerate kinase from the extract. Triose phosphate Scheme (b): purification ofphosphoglucose isomerase, isomerase is only partially precipitated at 75 %-satd. phosphoglycerate mutase, lactate dehydrogenase, (NH4)2SO4, and there are approximately equal aldolase and phosphoglucomutase from the 45-60%- amounts of this enzyme in each of fractions CD, E satd.-(NH4)2S04 fraction and the supernatant from E. This scheme is an alternative to Scheme (a) as far After desalting or dialysis, fraction CD was made as lactate dehydrogenase, aldolase and phospho- to pH6.6 with Tris/Mes buffer; the buffer included glucomutase are concerned; it has been developed for 1.0mM-magnesium acetate to stabilize enolase, and isolation of the other two enzymes. Fraction BC was lOmM-2-mercaptoethanol as a general protective desalted, then taken to pH7.0 with Tris/Mops buffer. measure for all of these enzymes. The sample, in a Two columns in tandem were used, the first being volume of approx. 300ml, was applied to a CM- CM-cellulose, the second phosphocellulose. The cellulose column in pH6.6 buffer. Nearly all of the sample was diluted to about 400ml, and passed creatine kinase, and all of the triose phosphate

Table 1. Summary ofpurifications in Scheme (a) The 'highest literature specific activities' are extrapolated to assay conditions used in the present experiments (Scopes, 1977). References: a, Bondar & Pon (1969); b, Pesce et al. (1967); c, Penhoet & Rutter (1975); d, Hashimoto & Handler (1966). Pyruvate Lactate Phospho- kinase dehydrogenase Aldolase glucomutase Activity in 1 litre of original extract (k-units) 235 480 42.5 135 Activity in fraction BC (from 1 litre) (k-units) 162 356 32.5 69 Activity not adsorbed at pH6.5 (k-units) 0 7 0 69 Activity affinity-eluted (k-units) 122 228 27.5 56 Specific activity affinity-eluted (units/mg) 260 490 16.5 650 Specific activity after gel filtration (units/mg) 340 600 17.0 750 Highest literature specific activity (units/mg) 340 610b 18.Oc 600' Final yield of enzyme from 1 litre (mg) 320 380 1350 68 1977 MULTIPLE ENZYME PURIFICATIONS BY AFFINITY ELUTION 269

pH 7.0 Buffer pH: 0.2mM-2,3- I 0.5 mm-EDTA I 0.5 mM-Glucose 6-phosphate Bisphosphoglycerate Added ligand: I 'l 'lI

00cd 514 .)uz I

00 e 7!

400 500 600 700 800 Eluate volume (ml) Fig. 5. Elution diagram for the phosphocellulose column at pH7.0 in Scheme (b)

The fraction applied was BC after it had been passed through a CM-cellulose column at pH7.0.. , Phosphoglucose

isomerase; phosphoglycerate mutase. , Percentage transmission at 280nm. For details of procedure, see the text.

Table 2. Summary ofpurifications in Scheme (b) The 'highest literature specific activities' are extrapolated to assay conditions used in the present experiments (Scopes, 1977). References: a, Dyson & Noltmann (1968); b, Grisolia & Cleland (1968); c, Pesce et al. (1967); d, Penhoet & Rutter (1975); e, Hashimoto & Handler (1966). Phospho- Phospho- Lactate Phospho- glucose glycerate dehydro- gluco- isomerase mutase genase Aldolase mutase

Activity in 1 litre of original extract (k-units) 142 440 435 45 145 Activity in fraction BC (from 1 litre) (k-units) 88 260 320 33 72 Activity not adsorbed on CM-cellulose (k-units) 88 260 10 0 72 Activity affinity-eluted (k-units) 62 245 210 27 66 Specific activity affinity-eluted (units/mg) 540 840 500 17.0 620 Specific activity after gel filtration (units/mg) 740 1000 * * 710 Highest literature specific activity (units/mg) 760a 1050b 610c 1890d 600e Final yield of enzyme from 1 litre (mg) 74 205 420 1590 79 * Enzymes were crystallized without gel-filtration step.

isomerase, passed through the column. (At pH6.5 or of the protein. Separation of the two kinases was lower, substantial amounts of creatine kinase were readily achieved by a subsequent gel filtration on retarded on the column.) The sample was washed in Sephadex G-100 or G-75. with 200ml ofbuffer containing an additional 0.5 mM- The column pH was raised slightly, to pH6.8 EDTA, and the myokinase was eluted by replacing (Tris/Mes buffer, including 0.5 mM-EDTA as the additional EDTA with 0.5 mM-ADP. The protein dummy substrate), by using 200ml of buffer. eluted contained, not only all the myokinase, but also The EDTA was next replaced by phosphoenol- the small amount ofcreatine kinase that had been held pyruvate, causing elution of enolase, somewhat con- back on the column, which in fact accounted for most taminated with pyruvate kinase, for the same Vol. 161 270 R. K. SCOPES

Buffer pH: - pH 6.6 pH 6.8 pH 8.0 0.5 mM- 0.9 mm. 0.5 mm- Phosphoenol, 0.5 mM- Added lgand: EDTA 10.5 mM-ADP EDTA j pyruvatee 0.5 mm-EDTA 13-Phosphoglycerate *1I 0r

0 00

Cq *a 0 cis 'a >5 4U 6A) .4 a 4)i

Eluate volume (ml) Fig. 6. Elution diagramfor rabbit musclefraction CD adsorbed on a CM-cellulose column atpH6.6 The enzymes obtained by affinity elution were myokina (-u *), enolase (--) and phosphoglycerate kinase (----). , Percentage transmission at 280nm. For details of procedure, see the text.

pH 6.2 Buffer pH: pH 6.0 0.5 mm-EDTA 0.5 mM-ATP | Added ligand;

CO 0000 a 25 4-aCO _, ._00 a

a)0 50 U CA f;is 75

Eluate volume (ml) Fig. 7. Affinity elutfon ofcreatine kinasefrom CM-cellulose after adsorption atpH6.0 The fraction applied was that protein not adsorbed on CM-cellulose at pH6.6 (see Fig. 6). -..., Creatine kinas

-, Percentage transmission at 280nm. For details of procedure, see the text.

reason as the converse contamination was obtained Tricine containing 0.5 Sm-M A as dummy sub- in Scheme (a). Afte all the enolase had been obtained, strate). As the p1* rose, protein was eluted, including the column was washed with pH 8.0 buffer (NaOH/ some haem pigments. After about 250ml of buffer 1977 MULTIPLE ENZYME PURIFICATIONS BY AFFINITY ELUTION 271 bad been washed into the column, the EDTA was adsorbed, but triose phosphate isomerase was not. replaced by 0.5mM-3-phosphoglycerate. Phospho- The column was washed with 200ml ofcold Tris/Mes glycerate kinase was eluted from the column by this buffer, pH6.2, containing 0.5mM-EDTA and no procedure, and it contained only a little contaminant, magnesium acetate, and the creatine kinase was which was completely removed by gel filtration. The eluted by replacing the EDTA with 0.5mM-ATP elution diagram for this column is illustrated in Fig. (Fig. 7). 6. Finally, triose phosphate isomerase could be ad- The protein that had not adsorbed on the column at sorbed by further lowering the pH of the fraction pH6.6 was adjusted to pH6.0 with 1 M-Mes, cooled which did not adsorb at pH6.0 to 5.5 with picolinic to 0°C, then run on to another ice-cold CM-cellulose acid, and applying it to a third CM-cellulose column column of the same dimensions, equilibrated with equilibrated with Tris/picolinate, pH5.5. The sample Tris/Mes buffer, pH6.0. Creatine kinase was totally was washed in with Tris/Mes buffer, pH6.0. As the

6.0 Buffer pH:' pH 5.5 pH

Ce4 ._A IE 0 dI :3 i) ?

Ce I4

400 500 Eluate volume (ml) Fig. 8. Elution of triose phosphate isomerase from CM-cellulose after adsorption atpH5.5 The fraction applied was that protein not adsorbed on CM-cellulose at pH6.0 (see Fig. 7). ...., Triose phosphate

isomerase. , Percentage transmission at 280nm. For details of procedure, see the text.

Table 3. Summary ofpurifications in Scheme (c) The 'highest literature specific activities' are extrapolated to assay conditions used in the present experiments (Scopes, 1977). References: a, Kress et al. (1966); b, Wold (1971) (assumes A2aogIml= 0.90); C, Scopes (1969); I, Mahowald et al. (1962); e, Krietsch et al. (1970). Phospho- Triose glycerate Creatine phosphate Myokinase Enolase kinase kinase isomerase Activity in 1 litre of original extract (k-units) 78 80 225 116 3250 Activity in fraction CD (from 1 litre) (k-units) 45 47 144 64 990 Activity not adsorbed at pH6.5 (k-units) 0 0 0 48 990 Activity affinity-eluted (k-units) 32 27 117 38.5 700* Specific activity affinity-eluted (units/mg) 210 36 550 95 4600* Specific activity after gel filtration (units/mg) 1410 66t 720 98 6000 Highest literature specific activity (units/mg) 1500" 90" 720c lood 6000e Final yield of enzyme from 1 litre (mg) 20 355 145 345 105 * Not 'affinity' procedure, see the text. t With A4ko"''0 = 0.63; specific activity woud be 92 with Allngul = 0.90 (Wold, 1971). Vol. 161 272 R. K. SCOPES pH slowly rose, the isomerase was eluted, in a fairly muscle triose phosphate isomerase is not successful, pure state (Fig. 8). As indicated in the preceding since at the suitable pH (approx. 5.8) the enzyme has paper (Scopes, 1977), affinity elution of rabbit low affinity for its substrates. A summary of the procedure is given in Table 3. The purity oftheenzymes obtained in this Scheme can be judged by comparison of the specific activities 0 1 2 3 4 5 6 7 8 9 with the highest recorded values in the literature (Table 3), and from the diagram of the sodium 100 dodecyl sulphate/polyacrylamide-gel electrophoreto- 80 gram (Fig. 9). 60 -PK The impurities present were in most cases enzymes which bind the sameligand asthe one beingused in the I0 _PGK - - - m m _EN elution procedure. Thus myokinase was contamin- 40 - -CK 0 ated with creatine kinase, and enolase with pyruvate kinase. Impurities in each case were removed by gel x 30 I -0 X6 - TIM filtration, resulting in virtually pure enzyme, as To-0 indicated in Fig. 9 and by the specific activities in 20 -MK Table 3. Scheme (d):purification ofmyokinase, triosephosphate Fig. 9. Diagram ofa dodecyl sulphate/polyacrylamide-gel- isomerase, enolase and twoforms electrophoresis slab containing samples relevant to ofphosphoglycerate Scheme (c) kinase from the 60-70Y%-satd.-(NH4)2S04fraction of Sample 1: fraction CD complete; 2: fraction CD not chicken muscle adsorbed on CM-cellulose at pH6.6; 3: ADP-eluted This procedure for chicken muscle is included to myokinase (+creatine kinase); 4: myokinase after illustrate how the methods can be adapted to purify gelfiltration; 5: phosphoenolpyruvate-elutedenolase; the enzymes from other species. The same main 6: enolase after gel filtration; 7: phosphoglycerate kinase eluted with 3-phosphoglycerate; 8: creatine enzymes are present in chicken muscle fraction CD kinaseelutedwithATP; 9: triose phosphate isomerase as in that from rabbit muscle, and except for creatine eluted by raising pH. Abbreviations: see legend to kinase, which is too acidic in chicken muscle to adsorb Fig. 1; and TIM, triose phosphate isomerase; MK, on a cation-exchanger in a stable pH range, all have myokinase. been isolated by very similar steps. Comparison ofthe

pH 6.5 pH 6.9 pH 8.0 Buffer pH: 0.5mM- 0.5mM- 0.5 mm- 0.5 mM- 0.5 mM- Dihydroxyacetone Phosphoenol- 3-Phosphoglycerate EDTA ADP EDTA phosphate pyruvate _ 0.5 mm-EDTA gradient Added ligand: I- I l 0 0.5 1.0 0 (mM) -,20 aE 0 1-N 00 ("I 25 c (Aa 0 t:Y 0

t9 50 0 CU a) 0 a7 8)

800 Eluate volume (ml) Fig. 10. Elution diagram for chicken musclefraction CD adsorbed on CM-cellulose at pH6.5 The enzymes obtained by affinity elution were myokinase ( .... ), triose phosphate isomerase ( - -), enolase ( - -) and phosphoglycerate kinase (----). -, Percentage transmission at 280nm. For details of procedure, see the text. 1977 MULTIPLE ENZYME PURIFICATIONS BY AFFINITY ELUTION 273

Table 4. Summary ofpurifications in Scheme (d) The 'highest literature specific activities' are extrapolated to assay conditions used in the present experiments (Scopes, 1977). References: a, Kress etal. (1966) (for rabbit enzyme);", Putmanet al. (1972); c, Wold (1971) (for rabbit enzyme); d' Scopes (1969) (for rabbit enzyme). I and II refer to the lower- and higher-pI forms of phosphoglycerate kinase respectively. Triose phosphate Phosphoglycerate Myokinase isomerase Enolase kinase Activity in 1 litre of original extract (k-units) 110 2500 155 200 Activity in fraction CD (from 1 litre) (k-units) 93 850 85 145 Activity not adsorbed at pH6.5 (k-units) 0 10 0 0 Activity affinity-eluted (k-units) 70 575 81 115 Specific activity affinity-eluted (units/mg) 950 4600 89 I 500 II 120 Specific activity after gel filtration (units/mg) 1450 6000 I 700 II 650 Highest literature specific activity (units/mg) 150oa 5800b 90C 720d Final yield of enzyme from 1 litre (mg) 43 86 910 I 83 II 59 * Crystallized without gel-filtration step.

12 3 4 56 7 8 9 four enzymes are adsorbed on and eluted from a single CM-ellulose column. Application of the sample is the same as for the 044 rabbit muscle fraction (Scheme c), namely in a .ea) Tris/Mes buffer containing 1 mM-magnesium acetate and 10mM-mercaptoethanol, but the pH used was 6.5. The first elution of myokinase by ADP was also C) carried out under the same conditions. The myo- x kinase was not contaminated with creatine kinase, as 0 noneofthe latter enzymehad adsorbedonthecolumn. 8en Triose phosphate isomerase had adsorbed on the 4 column, and could be eluted after the myokinase by raising the pH to 6.9 (Tris/Mes), and then adding substrate mixture generated enzymically from fructose 1,6-bisphosphate (Scopes, 1977) at pH6.9. Fig. 11. Diagram of a dodecyl sulphate/polyacrylamide- gel-electrophoresis slab containing samples relevant to After this, enolase was eluted with 0.5mM-phospho- Scheme (d) enolpyruvate, also at pH6.9, and the pH was then Sample 1: fraction CD complete; 2: ADP-eluted raised to 8.0 with NaOH/Tricine. Phosphoglycerate myokinase; 3: myokinase after gel filtration; 4: kinase in chicken muscle appears to consist of two dihydroxyacetone phosphate-eluted triose phosphate components (Scopes, 1968), and these can be separ- isomerase; 5: triose phosphate isomerase after gel ated either by successive elutions with different filtration; 6: phosphoenolpyruvate-eluted enolase; concentrations of substrate (0.5mM-3-phospho- 7: enolase after gel filtration; 8: 3-phosphoglycerate- glycerate for the less basic of the two, 1 mM-3-phos- eluted phosphoglycerate kinase I; 9: 3-phospho- phoglycerate for the other), or preferably with a linear glycerate-eluted phosphoglycerate kinase II. substrate After with 250ml of Abbreviations: see legends to Figs. 1 and 9. gradient. washing NaOH/Tricine buffer, pH8.0, containing additional 0.5mM-EDTA, 150ml ofthis buffer was set up, 150ml of NaOH/Tricine buffer containing 1.5mM-3-phosphoglycerate being mixed in to form a linear gradient; the two phosphoglycerate kinase components were electrophoretic mobilities of the rabbit and chicken eluted separately, with a substantial amount of enzymes (Scopes, 1968) suggests that enolase, being another protein contaminating the second form. A more basic in chicken, should adsorb more firmly, summary of this procedure is given in Table 4, and whereas triose phosphate isomerase should adsorb the elution diagram in Fig. 10. Purities of the less readily. In practice, the isomerase adsorbs more enzymes obtained are illustrated diagrammatically strongly, whereas the enolase behaves about the same, by the sodium dodecyl sulphate/polyacrylamide-gel as do myokinase and phosphoglycerate kinase. These electrophoretogram in Fig. 11. Vol. 161 274 R. K. SCOPES

0.3

0a a *a 0.2 i %-O csuCe on .O. CD 4. K Cda) 0. 1

0 400 600 800 Eluate volume (ml) Fig. 12. EIlution diagram for chicken musclefraction BC adsorbed on DEAE-cellulose atpH7.2 The enzymes obtained by gradient elution were phosphoglucomutase ( . ) and phosphoglycerate mutase (----). - -, KCl concentration - , Percentage transmission at 280nm. For details of procedure, see the text.

Table 5. Summary ofpurifications in Scheme (e) The 'highest literature specific activities' are extrapolated to assay conditions used in the present experiments (Scopes, 1977). References: as Hashimoto & Handler (1966) (for rabbit and fish enzyme); b, Torralba & Grisolia (1966); C, Pesce et al. (1967); d, Penhoet & Rutter (1975) (for rabbit enzyme). Phospho- Phospho- glycerate Lactate glucomutase mutase dehydrogenase Aldolase Activity in 1 litre of original extract (k-units) 195 550 540 31.0 Activity in fraction BC (from 1 litre) (k-units) 110 345 355 20.5 Activity not adsorbed on DEAE-cellulose (k-units) 1 0 330 20.1 Activity affinity-eluted (k-units) 84* 230* 186 17.7 Specific activity affinity-eluted (units/mg) 500* 1000* 560 18.3 Specific activity after gel filtration (units/mg) 660 -t -1t -t Highest literature specific activity (units/mg) 600" lOOOb 660c 18.0d Final yield of enzyme from 1 litre (mg) 65 230 335 970 * Not 'affinity' procedure. t Crystallized without gel-filtration step.

Scheme (e): purification of the two glycolytic mutases acid, then heated at 45°C for 10 min, cooled to 20°C, from chicken muscle, together with lactate dehydrogen- thepH adjustedto 7.5 with I M-Tris, and thedenatured ase and aldolase protein removed by centrifugation for 10min at This procedure has been included, although part of 10000g. It was then desalted on a Sephadex G-25 it does not involve affinity elution. Fraction BC from column, and run on to a DEAE-cellulose column in chicken muscle, after heat-treatment to remove Tris/Mops buffer, pH7.2. The column was washed creatine kinase, contains only two major components with 200ml of Tris/Mops buffer, pH7.2, then a acidic enoughtoadsorb onDEAE-eUulose atpH7.2. linear gradient in KCI was applied by mixing in400ml Thesearephosphoglucomutaseandphosphoglycerate of 0.15M-KCI in Tris/Mops buffer to 400ml of the mutase. A simple ionic-strength gradient elutes the buffer alone. First phosphoglucomutase was eluted, two enzymes separately, each sufficiently pure to be and then phosphoglycerate mutase. The elution able to be crystallized. Proteins not adsorbed on the diagram for this column is shown in Fig. 12, and the DEAE-cellulose can then be adsorbed on CM- activity details in Table 5. cellulose for affinty elution of lactate dehydrogenase The protein that did not adsorb on the DEAE- and aldolase. cellulose column was run straight on to a CM- Fraction BC was taken to pH5.5 with Im-acetic cellulose column, equilibrated in the same Tris/Mops 1977 MULTIPLE ENZYME PURIFICATIONS BY AFFINITY ELUTION 275

pH 7.2 Buffer pH: 30 mm-KCI Extra ionic strength: 1 5 mM-KCI I 0.2 mm- NADH 0.2 Added ligand: I mM-Fructose 1.6-bisphosphate I

i 0 Ca .a .4 .i:I ni

00, co 4-

400 600 800 1000 1200 Eluate volume (ml) Fig. 13. Elution diagramfor chicken musclefraction BC adsorbed on CM-cellulose atpH7.2 The fraction applied was that protein not adsorbed to DEAE-cellulose (see Fig. 12). Enzymes obtained by affinity elution were lactate dehydrogenase (.... ) and aldolase (----). - , Percentage transmission at 280nm. For details of procedure, see the text.

2 4 5 6 buffer; the two columns were in fact coupled together 1 3 in tandem during sample application. This column 0 was washed with Tris/Mops buffer containing 5mM- 100 KCI, and NADH was then added to the buffer to 80 elute lactate dehydrogenase, giving enzyme of specific activity 560 units (pmol/min) per mg of - - PGM 60 protein. Then further KCI was added before elution of aldolase. It was found necessary to use 30mM-KCl .5ha04 for the pre-elution wash for chicken muscle aldolase, as it adsorbs more strongly on the column than does 40 ------ALD the rabbit enzyme. As in the rabbit muscle example, - - - LDM the enzyme was specifically eluted with 0.2mM- x 30 PGAM fructose 1,6-bisphosphate, giving an enzyme which 0 "- was essentially pure (specific activity 17). The elution diagram for these enzymes is shown in Fig. 13. The procedure is summarized in Table 5, and the electro- 20 phoretogram is shown in Fig. 14. 0D Discussion Fig. 14. Diagram of a dodecyl sulphate/polyacrylamide- The procedures outlined in this paper demonstrate gel-electrophoresis slab containing samples relevant to that affinity-elution chromatography can be used for Scheme (e) relatively large-scale purification of several different BC not Sample 1: fraction BC complete; 2: fraction enzymes from one extract. For multiple purifications adsorbed on DEAE-cellulose; 3: phosphoglucomutase obtained by gradient elution; 4: phospho- there are several possible stages where separation can glycerate mutase obtained by gradient elution; be achieved. Ideally, all enzymes would be adsorbed 5: NADH-eluted lactate dehydrogenase; 6: fructose on one column, then each eluted successively by step- Ifi-bisphosphate-eluted aldolase. Abbreviations: wise changes of pH and ionic strength, and inclusion see legend to Fig. 1. of ligands. In practice this is rarely satisfactory; the Vol. 161 276 R. K. SCOPES pH needed to adsorb the lowest pl protein (on a enzymes were close to (and in one case in excess of) cation-exchanger) is likely to be so low as to denature the highest reported specific activities (extrapolated other proteins. Even ifthe denatured proteins are not to the assay conditions used in the present experi- among the enzymes being purified, this would be ments). Sodium dodecyl sulphate/polyacylamide-gel undesirable, as the coagulated protein clogs the electrophoresis confirmed the homogeneity of the columns. Also, a protein present in relatively small preparations. The electrophoretic technique was also amounts still needs a column large enough to adsorb valuable in indicating the likely identity of impurities it together with other proteins present in larger eluted with the enzyme from the column, from known quantity, leading to high dilutions and possible loss subunit sizes of the major protein components of activity of the small component. Because of these (Scopes & Penny, 1970). Certain results have been factors two or more columns have been used in most anomalous, in particular the properties of enolases cases, a system of stepwise adsorption as well as isolated by this technique. The subunit sizes for the stepwise elution. By using the zwitterionic buffers as chicken and rabbit enzymes were estimated to be described it was possible to lower the pH after passage 50000±2000 and 47000±2000 daltons respectively. through one column without changing the ionic The latter is somewhat larger than reported, but the strength, so that the adsorption characteristics of former is similar to the size ofthe fish enzymes (Wold, each enzyme were unaltered. Each column was 1971), being further evidence for a larger polypeptide separately worked up with stepwise affinity elution, chain in non-mammalian enolases. The extinction for each of the major enzymes present. coefficients at 280nm of both enolases, based on Some of the schemes described took only two or protein determination by u.v. methods (Scopes, three trials before arriving at a satisfactory procedure; 1974) were each 0.63 (for 1 mg/ml), compared with others took considerably longer. Certain enzymes values around 0.90 in the literature. have very narrow ranges of conditions where affinity The principles of affinity-elution chromatography elution can be carried out. Differences of only 0.2 pH as applied to rabbit muscle glycolytic enzymes are unit or changed column dimensions can result in described in the preceding paper (Scopes, 1977). The enzymes either beingeluted too soon or not appearing successful adaptation of the methods to chicken when expected. However, the conditions for other muscle enzymes as described here demonstrates the enzymes, particularly those eluted at pH values above general usefulness ofthe procedure. Since adsorption 7, were less critical. This probably reflects a greater on ion-exchange materials is dependent on net charge, degree of change of adsorptive properties on binding some knowledge of the relative isoelectric points of of ligand, but also relevant is the fact that pH is less the species' enzymes is useful; this can be obtained critical because proteins have few titratable groups by gel-electrophoresis techniques (Scopes, 1968). It in the pH range 7-8. has been possible to purify phosphoglycerate kinase In cases where more than one form ofenzyme was from a wide variety ofspecies by using this technique, present owing to genetic or epigenetic factors, it has and several other enzymes have also been purified been possible to separate the enzymes by a substrate from kangaroo muscle extracts (R. K. Scopes & gradient. Separation ofthe five lactate dehydrogenase T. Fifis, unpublished work). isoenzymes has been reported from an affinity For multiple enzyme purification, as a 'half-way adsorption column, by using an NADH gradient house' between conventional ion-exchange chromato- (Brodelius & Mosbach, 1973). In the present example graphy and affinity-adsorption chromatography, two forms of phosphoglycerate kinase from chicken affinity-elution chromatography combines beneficial muscle (one of which is probably an epigenetic factors of each, with many of the disadvantages modification of the other, as peptide 'maps' are removed. indistinguishable; T. Fifis, unpublished work) were nearly completely separated by a 3-phosphoglycerate This work was carried out with the expert technical gradient. In another example (Stewart & Scopes, assistance of Miss A. Dennis, and was supported by grants 1975), the genetically distinct A and B isoenzymes of from the Australian Research Grants Council. phosphoglycerate kinase (VandeBerg et al., 1975) were completely separated by using a substrate gradient. References Although few of the affinity-eluted enzymes could be as contaminants were Bondar, R. J. L. & Pon, N. G. (1969) Biochim. Biophys. described 'pure', usually Acta 191, 743-747 not more than 20 % of the protein, so they could be Brodelius, P. & Mosbach, K. (1973) FEBS Lett. 35, 223- removed either by direct crystallization ofthe enzyme, 226 or in most cases by gel filtration. It is fortuitous that Chappel, A., Scopes, R. K. & Holmes, R. S. (1976) FEBS contaminating proteins for the most part had molecu- Lett. 64, 59-61 larsizesquite different fromthemajorenzymepresent. Dyson, J. E. D. & Noltmann, E. A. (1968) J. Biol. Chem. After gel filtration, the specific acitvities of the 243, 1401-1414 1977 MULTIPLE ENZYME PURIFICATIONS BY AFFINITY ELUTION 277

Grisolia, S. & Cleland, W. W. (1968) Biochemistry 7, Scopes, R. K. (1968) Biochem. J. 107, 139-150 1115-1121 Scopes, R. K. (1969) Biochem. J. 113, 551-554 Hashimoto, T. & Handler, P. (1966) J. Biol. Chem. 241, Scopes, R. K. (1973) Biochem. J. 134, 197-208 3940-3948 Kress, L. F., Bono, V. H. & Noda, L. (1966) J. Biol. Scopes, R. K. (1974) Anal. Biochem. 59,277-282 Chem. 241, 2293-2300 Scopes, R. K. (1977) Biochem. J. 161, 253-263 Krietsch, W. K. G., Pentchev, P. G., Klingenburg, H., Scopes, R. K. & Fifis, T. (1975) Proc. Aust. Biochem. Soc. Hofstatter, T. & Bucher, T. (1970) Eur. J. Biochem. 14, 8,17 289-300 Scopes, R. K. & Penny, I. F. (1970) Biochim. Biophys. Mahowald, T. A., Noltmann, E. A. & Kuby, S. A. (1962) Acta 236, 409-415 J. Biol. Chem. 237, 1535-1548 Penhoet, E. E. & Rutter, W. J. (1975) Methods Enzymol. Stewart, A. A. & Scopes, R. K. (1975)Proc. Aust. Biochem. 42,240-249 Soc. 8,15 Pesce, A., Fondy, T. P., Stolzenbach, F., Castillo, F. & Torralba, A. & Grisolia, S. (1966) J. Biol. Chem. 241, Kaplan, N. 0. (1967) J. Biol. Chem. 242, 2151-2167 1713-1718 Putman, S. J., Coulson, A. F. W., Farley, I. R. T., Riddle- VandeBerg, J. L., Cooper, D. W. & Close, P. J. (1973) ston, B. & Knowles, J. R. (1972) Biochem. J. 129, Nature (London) New Biol. 243, 48-50 301-310 Wold, F. (1971) Enzymes 3rdEd. 5,499-538

VQI. 161