Eur. J. Biochem. 52. 321 - 330 (1975)

~- .... --·1

EFE 6 ' . j

Nur für persönlichen GeSrauch

Fractionation of Horseradish Peroxidase by Preparative Isoelectric F ocusing, Gel Chromatography. and Ion-Exchange Chromatography t

Henry DELINCEE and Bertold J. RADOLA

Institut für Strahlentechnologie, Bundesforschungsanstalt für Ernährung, Karlsruhe

(Received July 23 / December 2, 1974)

Horseradish peroxidase has been fractionated by preparative isoelectric focusing in a density gradient and in a Jayer of granulated gel using pH-3-10 and narrow-pH-range carrier ampholytes at different total enzyme loads. The resolution of peroxidase isoenzymes in preparative-layer iso­electric focusing was comparable to that obtained by analytical thin-layer isoelectric focusing. Iso­electrically homogeneous isoenzymes could be isolated with good recovery in a single fractionation step. Despite the excellent separation of the individual isoenzymes by isoelectric focusing in gel layers, an effective purification, indicated by the absorbance ratio A403 nm/A278 nm• could not be achiev­ed by focusing applied as a single step. By different fractionation sequences combining gel chromatog­raphy, ion-exchange chromatography, and isoelectric focusing, individual isoenzymes with a high purity and homogeneous with respect to their size and charge properties have been isolated.

The heterogeneity of plant peroxidases, which usually occur as complex isoenzyme systems [1], has been weil established by ion-exchange chromatog­raphy [2 - 11], starch and polyacrylamide gel electro­phoresis [12 - 18], and isoelectric focusing [19 - 21]. For an ·elucidation of the structural basis of the o bserved heterogeneity the individual isoenzymes must be isolated. In an attempt to isolate pure isoenzymes a number of methods have been used for the prepara­tive fractionation of peroxidases. Salt fractionation with ammonium sulfate and repeated ion-exchange chromatography have been employed most frequently, e.g . in the case of fig latex [8, 9], pineapple [10], turnip roots [11 ], rice [6], Japanese-radish [7], and horseradish roots [2 - 5]. Electrophoretic methods, such as moving boundary electrophoresis [22], starch gel electrophoresis [12] and column electrophoresis on Sephadex [23] have been applied occasionally in preparative work with horserac;lish peroxidase, while free-flow electrophoresis has been used for tomato peroxidase [24].

Although extensive purification of horseradish peroxidase isoenzymes can be achieved by repeated

En:yme. Peroxidase or donor: H20 2 oxidoreductase (EC 1..11. 1.7).

Eur. J. Biochern. 52 (19.75)

-------- -„. ------ .- -

ion-exchange chromatography, this procedure has several drawbacks. Strang dilution of the sample, overlapping of components necessitating narrow peak­cutting, and a frequently observed irreversible adsorp­tion (on each column) diminish the yield. Further­more, the method may be very time-consuming. On application of electrophoretic methods the sample load was, in most cases, severely limited, and only electrophoresis on a Sephadex column permitted a somewhat higher load [23]. Preparative isoelectric focusing in granulated gels has recently been shown to be a powerful tool for protein fractionation [25 - 27]. By analytical thin-Iayer isoelectric focusing on Sephadex, commercial horseradish peroxidase has been separated into about 20 distinct isoenzymes differing in their isoelectric points [19]. In view of these results it seemed interesting to explore the poten­tial of preparative isoelectric focusing for the frac­tionation of horseradish peroxidase into individual isoenzymes. Preparative isoelectric focusing has been applied both as a single fractionation step, and in combination with other methods based on different properties of the molecules to be separated, namely gel chromatography and ion-exchange chromatog­raphy.

322

MATERIALS AND METHODS

Materials

Horseradish peroxidase with an absorbance ratio A403 nm/A278 nm of 0.6 was purchased from ·Boehringer (Mannheim, FRG). Sephadex was obtained from Pharmacia (Uppsala, Sweden) and Ampholine carrier ampholytes from LKB (Bromma, Sweden).

Enzyme Assay

Peroxidase act1v1ty was measured by a slightly modified procedure of Chance and Maehly [28]. The

'increase in absorbance at 436 nm by the peroxidation of guaiacol was recorded with a PMQ II spectro­photometer (Zeiss, Oberkochen, FRG) connected with a Servogor recorder (Goerz, Wien, Austria). Initially, the reaction was linear in the interval 0.1 < LlA/min < 1.5.

lsoelectric Focusing

Isoelectric focusing in Sephadex layers was carried out as described previously [19, 27, 29]. Briefly, glass plates (20 x 20 cm or 40 x 20 cm) were coated with a suspension of Sephadex G -75 superfine, containing 1 % of carrier ampholytes. In analytical separations the applied layer was 0. 75 mm, which after gentle drying to give a water loss of about 20 ~~ resulted in a final thickness of 0.6 mm. In the preparative ex­periments the final layer was 1- 2 mm. Focusing was carried out in the Double Chamber (Desaga, Heidel­berg, FRG) with water at 4 - 10 °C circulated through the cooling block. In analytical focusing the sample has been applied as a streak, five toten samples being focused on a 20-cm plate. In preparative runs the sample, with 1 % (w/v) ampholytes added, was mixed with dry gel (60 mg Sephadex G-75 superfine/ml sample), and the resulting slurry was poured into a 15-cm-long slot made in the gel layer with a spatula.

F or detection of enzyme activity in analytical focusing a print was taken with a buffered substrate­impregnated chromatographic paper (SS 2043 b Mg!, Schleicher & Schüll, Dassel, FRG). For peroxidase detection the paper was buffered with a citrate-phos­phate buffer, pH 5.0, urea-peroxide was employed as primary, and o-toluidine as secondary substrate [30]. For protein detection the prints were stained with Coomassie brilliant blue G-250. In preparative focus­ing indicative paper prints were taken with strips (40 cm x 20 cm) ofMN 827 paper (Macherey & Nagel, Düren, FRG). For determination of the isoelectric point (pi) the pH was measured directly in the gel

Fractionation of Peroxidase

layer [19] using a microelectrode equipped with a fia t membrane (diameter 2 mm). After preparative focus­ing was completed, the visible or visualized enzyme zones were removed from the gel plate with a spatula. The removed gel was transferred to small glass columns of variable size closed at the bottom with a cotton plug. For elution approximately twice the gel volume of buffer was layered abov~ the gel.

Density gradient focusing w!~ performed in a 110-ml column (LKB, Bromma, Sweden). The col­umns were eluted by using a peristaltic pump at a fiow rate of 80 ml/h and 80-100 fraction of 20 drops were collected. The absorbance measurements were made in a PMQ II spectrophotometer.

Gel Chromatography

Thin-layer gel chromatography was performed as described previously [31].

Preparative gel chromatography in Sephadex lay­ers was carried out on glass p lates (30 x 20 cm or 20 x 20 cm) coated with a 1 - 2-mm Jayer of a sus­pension of Sephadex G-200 superfine in 0.02 M phos­phate buffer, pH 7.2, containing 0.5 M sodium chloride. Thick paper wicks (MN 827) wcre employed. After equilibration of the plate overnight, the sample in a volume of ~ 1 ml was applied as a 17-cm-long streak. About 50- 200 mg of peroxidase can be sepa­rated within 6 - 8 h. The enzyme zones were scraped from the plate and eluted in small columns by centri­fugation.

Preparative column gel chromatography was per­formed in a column (100 x 5 cm) of Sephadex G-200 or G-150. The eluant was 0.02 M phosphate buffer, pH 7.2, containing 0.5 M sodium chloride. The fiow rate was ~ 2 ml cm- 2 h- 1

; fractions of 200 drops were collected.

Ion-Exchange Chromatography

Fibrous cellulose ion-exchangers (CM-cellulose and DEAE-cellulose) were obtained from Serva (Heidelberg, FRG) and microgranular cellulose ion­exchangers (Whatman CM-52 and DE-52) from Balston Ltd (Maidstone, England). They were condi­tioned as prescribed by the manufacturer.

Protein Concentration

Protein solutions were concentrated with Diafl.o ultrafiltration (UM-10) membranes (Amicon, Ooster-hout, Holland) under nitrogen. ·

Eur. J. Biochem. 52 (1975)

H. Delincee and B. J. Radola

10 A

9

8 7 6

~5 t.

3 2

·1

0 10 20 30 1.0

10

9

8 7

6 ~5

t.

3

2

0

c

cm

9 · , / -/

---

,,,---__ _,,,, -

10 20 30 t.0 cm

!' . .. • - • +,""?. 1 ; _,,, • . - . .l L~ ............ ~ ..:

Fig.1. Preparatil'e-layer isoelec1ric focusing of crude liorseradish peroxidase ( absorbance ratio 0.6) in pH-3- 10 ampholyres. Plate. 40 x 20 cm; thickness of thc Sephadex G-75 superline layer, 2 mm ; focusing. 400 V for 20 h ; 400 mg of the commercial preparation wcre applied, 15 cm from the a node. (A) Densitogram (remission) at 403 nm of a non-stained print. The pH gradient (0---0) was measured in the gel layer at 20 •c. (B) Densitogram (remission) at 600 nm of the same print as in (A). stained with light green SF. (C) Paper strip stained wilh light green SF

RESULTS AND DISCUSSION

Preparative Jsoelectric Focusing

Analytical thin-layer isoelectric focusing revealed that crude commercial peroxidase (absorbance ratio A403 nm/A278 nm = 0.6) contained about 20 isoenzymes and multiple protein zones which did not a lways coin­cide with the enzyme zones (19, 32). For isolation of individual peroxidase isoenzymes preparative iso­electric focusing in Sephadex layers was employed. Amounts of 100- 400 mg of the crude peroxidase were separated in pH-3-10 ampholytes on 40 x 20-cm plates, usually in a 2-mm gel layer. T he resolution of preparative separations is comparable to that of analytical runs (Fig.1). The upper tracing was ob­tained by densitometry of an unstained paper print, representing the visible brown peroxidase zones. The lower densitogram shows the pattern after protein staining with light green SF. The difference in the patterns is quite evident. Different commercial prep-

Eur. J. Biochem. 52 (1975) ·.

323

arations of horseradish peroxidase of comparable purity con tain variable amounts of inactive proteins which results in slightly different patterns (cf (32)). Direct densitometry of peroxidase in the gel layer at 280 nm and 403 nm gave patterns essentially similar to those described earlier [32, 33). Activity measure­ments revealed that about 75 ~6 · of the applied peroxidase activity could be recovered. Total recovery was determined by scraping off lengthways from the gel layer a 2-cm-broad track containing all the enzyme zones. In a d ifferent approach, recovery of enzyme activity was determined for the nine major enzyme zones, scraped off from the plate. The sum of activity of the individual zones gave the much lower recovery of about 50 ~~- The difference in enzyme recovery can be attributed to the loss of some minor enzyme zones, and to zone-cutting of the major zones. The distribution of enzyme activity and the yield for the individual isoenzymes is shown in Table 1. The figures for the yield are based on absorbance measure­ments at 403 nm. assuming a molar absorption coefficient of 105 cm2 x mol - 1 [3, 5). This very con­venient calculation parallels roughly the assay of enzyme activity. The values in Table 1 represent a typical experiment. Deviations from these figures have been observed for different preparations of horse­radish peroxidase, which vary in the content of the acidic and basic isoenzymes.

Analytical refocusing of the iso'enzymes isolated by preparative isoelectric focusing is shown in Fig. 2. An excellent separation of the individual isoenzymes has been achieved. The isolated isoenzymes are basically homogeneous with only slight contamination of adjacent isoenzymes. After extensive dialysis for removal of carrier ampholytes, however, there was only a slight increase in the absorbance ratios (1.0- 1.5)1 of the isolated isoenzymes (Table 1). By refocusing of the main components (the near-neutral isoenzymes with isoelectric points of 6.5 and 7.1, measured under standardized conditions, see below) their absorbance ratio was raised to 1.6- 1.7. The occurrence of enzymatically inactive proteins with isoelectric points close to or identical with those of the isoenzymes offers an explanation for the rather slight increase of the absorbance ratio. For enzyme preparations containing less of the inactive proteins in the pH-6-9 range, the absorbance ratio for the pJ-7. 1 isoenzyme could be raised to 2.2, and that for the pJ-6.5 isoenzyme to 1.6. The absorbance ratios for the acidic isoenzymes in these preparations were lower, probably due to a higher proportion of con­taminating proteins in this range. When the isolation

1 The absorbance ratio was always measured in 0.01 M phos­phate buffer, pH 7.2.

324 Fractionation o f Peroxidase

Ta ble 1. Fractionatinn of horseradish peroxidase br isoe/ec1ric focusing

Step Method Isoelectric range of the fraction pi

Absorbance ratio Totalenzyme Yield activity

A403 nm / A27a nm

0 Crude horseradish peroxidase 3 - 9 0.6

u 18000

0/ I O

100 -----· ···-----------------------··-· ------

Prepara tive-layer 3.7 1.3 75 t 0.7 isoelectric focusing pH 3- 10 4.0 1.5 720 1.7

5.3 0.2 55 0.9 6.0 0.4 330 1.9 6.5 1.l 1335 7 7.1 1.5 4650 20 7.6 0.9 365 2.6 8.2 0.3 270 2.3 8.4 0.5 290 2.0

Preparative-layer 6.5 1.6 7 isoelectric focusing pH 6- 8 7.1 2.3 28

----------~-~----------·-- ~-------

2 Prepara tive-la yer isoelectric focusing pH 5- 9•

6.5 7.1

• Following preparative-layer isoelectric focusing pH 3- 10.

--•

-· ~ ... ---'••·

-­• •

·~

* " -• • Fig. 2. Refocusing of Jractions isolated by preparatil'e focusing ( see Fig. 1) of horseradish peroxidase. Furthest to the left and right is the starting material. Thin-layer isoelectric focusing in pH-3 - 10 ampholytes on a 20 x 20-cm plate. Anode at the top. Focusing at 200 V for 4 h followed by 800 V for 3 h. Enzyme detection with urea - peroxide and o-toluidine

of only a few selected isoenzymes was int~nded, focusing in narrow-range ampholytes was employed. In this case, a decidedly better purification was achieved. Thus, in those preparations containing many inactive proteins in the neutral region, ab­sorbance ratios of 1.6 and 2.3 were obtained for the pl-6.5 and pl-7.1 isoenz.ymes by focusing in pH-6-8 ampholytes on a 40-cm x 20-cm plate. However, by

1.6 1.7

3 12

fractionation ofthe crude enzyme in narrow-pH-range ampholytes, isoenzymes isoelectric outside of the chosen pH range were lost. The results of the focus­ing experiments in Jayers of granulated gel are sum­marized in Table 1.

Density gradient focusing in the 110-ml LKB column of horseradish peroxidase yielded a different pattern than thin-layer isoelectric focusing. A typical separation of the commercial enzyme in a sucrose gradient in pH-3 - 10 ampholytes is shown in Fig.3. The resolution was rather poor, not more than three to four distinct peaks could be separated. Only for the main peak an absorbance ratio higher than 1.0 was achieved. Thin-layer refocusing of the fractions iso­lated by the density gradient technique showed that despite rigorous peak-cutting (only one to three fractions from the peak maximum), substantial a-mounts of adjacent isoenzymes were present (Fig. 4). The discrepancy between the isoelectric points of the peroxidase isoenzymes estimated by the column technique and those determined by focusing in thin layers are difficult to interpret. By several techniques of isoelectric focusing employing gel media, namely thin-layer isoelectric focusing in granulated gels (Sephadex and Bio-Gel), ftat-bed and rod focusing in polyacrylamide gels, horseradish peroxidase did not reach a stationary state even on prolonged focusing, which for a number of other proteins gave constant pi values [29]. lsoelectric points of peroxidase iso­enzymes must be regarded as apparent isoelectric points. However, as long as exactly identical experi-

Eur. J. Bioebern. 52 (1975)

H. J)elincce and B. J. Radola

1.8

1.6 ..

1.4

E l.2 c

0 CO

~1.0 c 0

M 0

. ::: 0.8 0 „ u c _g0.6 ö "' .0

<( 0.4

0.2

0

,.. , 1

1' 1

10

-AL03 ----A200

1

I 1

1 1

1

,-, ,' \ '

20

' 1 1 ·--

' ... -, .... ,

30 Fraction number

12

11

10

9

8

7

5

4

3

2

\

40 so

Fig. 3. De11si1y gradienl focusing of crude lrorseradislr peroxidase in pH-3-10 amplro~r1es wi1/r 0.1 % added argini11e. LKB column, 110 ml; density gradient, 0- 50 % sucrose; anode at the top; focusing, 300 V for 64 h; 50 mg of peroxidase were applied in the middle of the column; isoelectric points ( ..... . ) determined at 20 ·c

Fract1 on no. 42-44

...... ,,.,, ... ·---

-

23-25 18-19 1

-Fig.4. Refocusing offrac1io11s iso/a1ed by densi1y gradie111 focusi11g ( see Fig. 3) of lrorseradislr peroxidase. Furthest to the left is thc starting material. T hin-layer isoelectric focusing in pH-3 - 10 ampholytes. Anode a t the top. Enzyme staining

mental conditions prevail, highly reproducible iso­electric points are obtained, and values referred to in this paper have been determined by thin-layer isoelectric foc using for 24-h focusing at 400 V in pH-3 - 10 ampholytes on a 40-cm plate [19).

Eur. J. Biochem. 52 (1975_)

.. ·r- ~ •. „ ........ „ ----. ·-

325

6

-AL03 s ----Awo

E c

0 ~4 ~

8 :3 0 „ u c

,, '1

' '

_g 2 0 "' .0 <(

1

20

" 1. 1 ' ' \ / ...... ___ ... -

so

' ' 1

'

„ ................ ___ ......

100 Fraction number

„ ..........

JSO

Fig. 5. Co/1111111 gel clrromawgraplry of crude horseradislr peroxidase 011 Seplrodex G-200. Column size, 90 x 5 cm. Elution with 0.02 M phosphate buffer, pH 7.2. containing 0.5 M sodium chloride

Gel Chromatography

Gel chromatography could provide a valuable supplementation of separation procedures according to charge properties. On Sephadex G-200, inactive proteins with a molecular size different from that of the peroxidase have been removed. In a typical experiment with crude peroxidase, some high-molec­ular-weight material migrated with the void volume and showed only traces of activity (Fig. 5). The main peak containing the enzyme corresponded to a molecular weight of about 40000. After pooling, the absorbance ratio for the material of the main peak increased to about 1.3-1.6 without changes in the isoelectric pattern (Table 2). By isoelectric focusing of fractions from different sections of the main peak (leading, medium and trailing parts), minor differences in the distribution of the acidic isoenzymes only were detected , indicating slightly different molecular size. Increases of the absorbance ratio were rarely observed when gel chromatography was applied not as the fi rst step but later in the fractionation procedure, e.g. after focusing.

Cellulose Jon·-Exchange Chromatography

Microgranular cellulose was preferred over con­ventional fibrous cellulose because of significantly sharper zones. The best results on CM-cellulose were achieved with a sodium acetate buffer, pH 4.4 and gradient elution [3). A column ofthe size of 25 x 50 mm gave results comparable to those of the previously used !arger columns, with the added advantage of

326 Fractionation of Peroxidase

Table 2. Fracrionarion of lwrseradish peroxidase hy ge/ chromarography and ion-exchange chmmatography

Step Method Isoelectric range of the fraction p/

Absorbance ratio Yield

% 0 Crude horseradish peroxidase 3 - 9 0.6 100

---------- - ---------- ·-------------- ----·--· --- ·- - -- ·--· -Column gel chromatography on

Sephadex G-200

CM-cellulose

D EA E-cellulose

3 - 9

3 - 5 5 - 7.5 7.5 - 9

5 - .9 3.7 4.0

'j E 2.

-A403 ----A28o (\

c 0 ~ 1.2 -0 c: 01.0

M 0 .... 0 08

II

/1

'1 '1 1 1 1 „ " ~0.6

c

1 1

.J:J

~ 0.4 ' ' I .J:J

<(

0.2

0

2 i

20 40

\

1

\

'

60 Fraction number

3 i

80 100 120

Fig. 6. 1011-exchange chromatogruphy of crude horseradish peroxidase on CM-cellulose (Wharman CM 52). Column size, 50x25mm; 500 mg of the commercial preparation were applied. Buffers for elution : (1) starting buffer 0.005 M sodium acetate, pH 4.4, 150 ml; (2) a linear gradient 0.005-0.05 M sodium acetate, pH 4.4, 400 ml; (3) a linear gradient 0.05-0.2 M sodium acetate, pH 4.4, 200 ml; (4) 0.2 M sodium acetate, pH 4.4, 50 ml. Fraction volume, 100 drops ~ 7ml

shorter separation time, much smaller volumes of eluant and more concentrated fractions (Fig. 6).

Thin-layer isoelectric focusing of the fractions isolated on CM-cellulose revealed a distinct relation­ship between the isoelectric points and the elution sequence. As could be expected, the components with the lowest isoelectric points were e!uted first, whereas the isoenzymes with higher isoelectric points appeared with increasing ionic strength (Fig. 7). Although a substantial improvement of the absorbance ratio of the chromatographic fractions could be achieved by the removal of inactive proteins, all fractions consisted of groups of isoenzymes with close pl values and some

a

• -

1.5

0.8 2.6 0.4

1.8 0.6 2.0

b c . d

r 65

4 45

3

45 0.07 1.l

e f .... .... - - ........

·ertt ......... .... ..... ...,...,,,, .,,,.,_ ......

Fig. 7. Thin-layer isoelecrric /ocusing in pH-3-10 ampholpes o/ peroxidase /ractions, separated by ion-exchange chromatography 011

CM-ce/111/ose (see Fig.6) . From left to right: (a) void volume, (b, c) eluted with the first gradient, (d, e) second gradient, (f) starting material. Anode at the top. Enzyme staining

additional components with largely different plvalues. Further separations of isoenzymes with only small diff erences in their isoelectric points are difficult to achieve by ion-exchange chromatography.

The best separations on DEAE-cellulose have been obtained with a 0.005 M Tris buffer, pH 8.4, and a salt gradient (0-2 M NaCI). Only the acidic isoen­zymes were adsorbed during elution with the equili­bration buffer (0.005 M Tris, pH 8.4), whereas all other isoenzymes came through with the void volume. With increasing ionic strength the acidic isoenzymes were eluted. Higher absorbance ratios were noted for

. the break-through fractions as weil as for the acidic isoenzymes. The results of experiments with the cellulose ion-exchangers are compiled with those of gel chromatography in Table 2.

Eur. J. Biochem. 52 (1975)

1 .,

H . Delincee and B. J . Radola 327

Table 3. Fracrionarion of lwrseradish peroxidase by rhe comhined use of gef chromnrography, chromarography on CM-ce//11fose and preporarive i.rnefecrric focusing in a narrow pH range

Step Method 1 soelectric range of the fraction pi

Absorbance ratio Yield

0

2

3

·3

3

Crude horseradish peroxidase Column gel chromatography on

Sephadex G-200 CM-cellulose

Preparative-layer isoelectric focusing pH 3- 5

Preparative-layer isoelectric focusing pH 6- 8

Preparat ive-layer isoelectric focusing pH 7- 9

3 - 9

3 - 9 3 - 5 5 - 7.5 7.5 - 9 3.7 4.0 6.5 7. l 8.2 8.4

• Freed of carrier ampholytes by gel c hromatography in layers.

0.6

l.5 2. l 3.0 l.6 1.7 2.5 (2.8)' 3.1 3.3 (3.4) b

2.0 2.5 (2.8)'

100

c 65

3 35

2 0.04 0.5 (0.3)' 5

20(l8)b 0.2 0.4 (0.3)•

• Freed of carrier ampholytes by ammonium sulfate precipitation and repeated washing.

Combined Use of Different M ethods

lt follows from the preceding results that different methods must be combined to obtain an efficient separation and purification. lt is important that the first step of the fractionation procedure aff ords a high load. Both column gel chromatography and cel­lulose ion-exchange chromatography satisfy this re­quirement and permit loads exceeding considerably the amounts of peroxidase separated in the experi­m ents described above. H igh total loads, in the order o f 1 - 10 g of protein and even more, can be processed in a single run by preparative isoelectr ic focusing in layers of granulated gels [26,27]. Thus preparative isoelectric focusing could be used even as a first fractionation step, but economical considerations will generally limit the use of the method to a later stage of fractionation. The attempt was made to keep to the number of steps, necessary to isolate individual isoenzymes, to a minimum, with an optimal purifica­tion as indicated by the absorbance ratio.

For a number of reasons the following two frac­tionation sequences appear most attractive : (a) col­umn gel chromatography followed by chromatography on CM-cellulose and preparative-layer isoelectric focusing in a narrow pH range; (b) chromatography on CM-cellulose followed by preparative-layer iso­electric focusing in a narrow pH range and gel chromatography. The results obtained by these se­quences are summarized in Tables 3 and 4.

By fractionation sequence (a), the main component of horseradish peroxidas·e has been isolated with an absorbance rat io of 3.3 - 3.4 and a yield of about

Eur. J. Biochem. 52 (1975)

20 %. which may even be higher when based on activity measurements (cf Table 1). When after the CM-cellulose chromatography the p/-5 - 7.5 fract ion, comprising the ma in activity, was chromatographed in addition on DEAE-cellulose, the absorbance ratio was raised to 3.3. By the subsequent focusing step, however, no further increase in the absorbance ratio could be achieved , and the final purification of the individual isoenzymes was equal to that obtained without D EAE-chromatography. Since the additional chromatographic step lowered the yield, it should be omitted.

The excellent resolution of isoenzymes which can be achieved by preparative isoelectric focusing in Jayers of granulated gels is demonstrated in Fig. 8. The near-neutral isoenzymes have been separated into two major and several minor components in pH-6-8 ampholytes as the third step ofthe fractiona­tion sequence. On analytical refocusing of the iso lated components, it ca n be seen in Fig. 9 that the isolated isoenzymes are essentially homogeneous. A slight contaminat ion with adjacent isoenzymes resulting from an unavoidable spontaneous interconversion of the isoenzymes was observed [34). lmmediately after focusing, the amount of these adjacent com­ponents represented only 0.1-1 % of the isolated component. Their detection was possible only because of the focusing effect by which trace amounts are concentrated in a compact zone, and the high sensitiv­

·ity of the enzyme staining. Similar conversions have a lso been found for the basic and acidic isoenzymes.

Both the basic and acidic isoenzymes have been iso­Jated with relatively high absorbance ratios (Table 3).

--:·~ ----•: ---· --..- -- ·-- --·- - -· ·-

328 Fractionation of Peroxidase

Ta ble 4. Froctionation n( horseradish peroxidose hy the cnmhined use nf chromatography on C M-ce/111/ose, preparati1•e isoe/ectric focusinK in o narro11· pH range wul fiel chromatography

Step

0

2

2

2

3

Method

Crude horseradish peroxidase CM-cellulose

Prepara 1 i ve-la yer i soelect ric focusing pH 3- 5

Preparative-layer isoelectric focusing pH 6-8

Preparative-layer isoelectric focusing pH 7-9

Gel chromatography in layers on Sephadex G-75 Superfine

lsoelectric range of the fraction p/

3 - 9 3 - 5 5 - 7.5 7.5 - 9 3.7 4.0 6.0 6.5 7.1 7.8 8.0 8.2 8.4 4.0 6.5 7.1

8.4

Absorbance ratio

0.6 0.8 2.6 0.4 0.5 1.2 0.6 2.4 2.8 0.6 0.5 1.2 1.7 1.4 2.5 3.1 1.2

pl7.6 pi 7.1

Yield

o; '0

100

44~ 45

3 0.14 0.7 0.8 5

22 0.15 0.09 0.3 0.7 0.3 4

16 0.4

pi 6.5 pi 6.0

·:1

J ~ ijJL----'-' ----'-' -----'''--------'

··-

0 10 20 30 40 cm

Fig. 8. Preparative-layer isoelectric focusing in pH-6-8 ampholytes of the pl-5- 7.5 fraction, isolated by ion-exchange chromatography nn CM-cellu/ose of the gel-chromatographed peroxidase ( see Table 3, step 3). Plate, 40 x 20 cm; Sephadex G-75 superfine layer, 2 mm; focusing, 300 V for 4 h, followed by 600 V for 16 h, and then 800 V for 5 h; about 275 mg peroxidase (p/-5- 7.5 fraction) were applied 15 cm from the anode. Densitogram (remission) at 403 nm of a non-stained print. The pH gradient (0----0) was measured in the gel layer at 20 °C

Analytical refocusing of the basic isoenzymes, iso­lated by preparative focusing in pH-7- 9 ampho­lytes, demonstrates again the good separation ob­tained by this method (Fig. 10). T\1e acidic isoenzymes were recovered with low yield after preparative iso­electric focusing, probably due to denaturation of these isoenzymes, resulting from the prolonged ex-

• 1

Fig. 9. Refocusing of isoen::ymes isolmed by preparative joc11si11g (see Fig.8) of the pl-5-7.5 fraction of lwrseradish peroxidase. Furthest to the Ieft the unfractionated peroxidase. Enzyme staining

posure to the acidic pH of 3.5-4.0 during the last phase of focusing. The risk of denaturation could be at least partially avoided by prefocusing of the plate without applied sample. Later sample application resulted in a much shorter exposure of the enzyme to extreme pH values (Delincee, H. and Radola, B. J., unpublished).

An. additional step is necessary to remove the carrier ampholytes from the isolated isoenzymes, which can be accomplished by a number of methods, e.g. dialysis, repeated ammonium sulfate precipita-

Eur. J. Biochem. 52 (1975)

- ····~

H. Delincee and B. J. Radola

pl7.6 pl8.0 pl&2 pl&L.

-- l ;J:l „ Fig.10. Refocusi11g of isol!n=ymes isolated by preparatil'e .focusi11g of the p-1-7.5-9 fracrion of lwrseradish peroxidase (see Table3. srep 3). Enzyme staining

tion and washing, gel chromatography or ion-exchange chromatography [35]. In Table 3 the effect of the removal of carrier ampholytes either by gel chromatog­raphy (for an acidic and basic isoenzyme) or am­monium sulfate precipitation and repeated washing with 4 M ammonium sulfate (for the main isoenzyme) are shown. A slight increase in the absorbance ratios was observed, accompanied by a decrease in yield.

Although fractionation sequence (b) proved to be eff ective with respect to the simultaneous isolation and purification of several isoenzymes, the yield and absorbance ratios were lower than in combination (a), particularly for the acidic and basic isoenzymes. The advantage of sequence (b) is the removal of carrier ampholytes, originating from the preparative focusing (step 2, see Table 4), by the third fractionation step, namely gel chromatography, by which accompanying proteins with a molecular weight differing from that of the peroxidase, not yet removed by the previous steps, are also separated. This was especially important for the basic isoenzymes, which still displayed size heterogeneity after the focusing step. Only by applying all three steps in sequence (b ), were pure isoenzymes, homogeneous with respect to their charge and size properties, obtained.

Our results underline the advantage of steps based on different fractionation principles to obtain individual isoenzymes with high absorbance ratios, in relatively good yields. The Iargest increase in absorbance ratio has been obtained with ion-ex­change chromatography, but none of the chromato­graphic fractions were homogeneous with respect to charge properties: lt _is difficult to separate iso­enzymes with close isoelectric points by ion-exchange chromatography and, according to our experience,

EurJ Biochcm. 52 (1975)

329

individual isoenzymes can only be isolated by repeated chromatography with rigorous peak-cutting, and at the expense of a substantial decrease in yield. Chroma­tographically homogeneous isoenzymes have been examined for homogeneity by electrophoretic methods [3, 5], but not by the more effective isoelectric focus-ing in gels. ,

Preparative isoelectric focusin,g off ers a high potential for the separation of isdenzymes with dif­ferent charge properties. Isoelectr ic focusing in Jayers of granulated gels affords a high operational flexibility, since the length , width and thickness of the gel layer can be varied within broad limits. This variability combined with a suitable range of carrier ampholytes, permits the separation conditions to be optimized according to the requirements of the separation task. Total Joad and Joad capacity of preparative­layer isoelectric focusing are outstanding [27].

In the experiments described in this paper, 300-400 mg of crude or pre-purified peroxidase have been focused on 40-cm-long gel Jayers in a total gel volume of 150-200 ml, corresponding to a Ioad capacity of ::::: 2.0 mg protein/ml gel suspension. This figure fo r load capacity is well below the maximum value of about 10 mg protein/ml gel suspension found for other proteins [27], and the amount of enzyme applied may easily be raised. By column electrophoresis in Sephadex 120 mg of pre-purified peroxidase have been separated at a load capacity of 0.6 mg/ml [23]. In conventional electrophoresis, however, the optimal buff er must be determined for each isoenzyme, thus limiting the usefulness of this approach for the simul­taneous isolation of several isoenzymes. By thin-layer electrophoresis on Sephadex with different buffers, it was found that an efficient separation of all per­oxidase isoenzymes could not be achieved in a single run. Some isoenzymes could only be separated from other isoenzymes by the use of different buffers. With isoelectric focusing, however, all isoenzymes can be separated efficiently in a single run in pH-3-10 ampholytes, and in addition a ny selected group of isoenzymes can be resolved simply by choosing the appropriate narrow pH range of carrier ampholytes.

While an excellent separation of isoenzymes can be achieved by prepara tive isoelectric focusing in Jayers of granulated gels it is necessary to supplement focusing with other fractionation principles in order to obtain a satisfactory purification. The combined use of gel chromatography, ion-exchange chroma­tography and isoelectric focusing permits the frac­tionation of horseradish peroxidase into individual isoenzymes with a high purity, in three to four steps, in a reasonably short time, in good yield, and displaying homogeneity with respect to charge and size properties.

330

The skilful technical assistance of Mrs A. Sieron and Miss B. Hielscher is gratefully acknowledged.

REFERENCES

l. Shannon, L. M. (1968) Annu. Rev. Plant Physiol. 19, 187-210. 2. Paul. K.-G. (1958) Acta Chem. Scand. 12, 1312-1318. 3. Shannon, L. M., Kay, E. & Lew, J. Y. (1966) J. Bio/. Chem.

241, 2166 -2172. 4. Kasinsky, H. E. & Hackett, D. P. (1968) Phytochemistry, 7,

1147-1150. 5. Paul, K.-G. & Stigbrand, T. (1970) Acta Chem. Scand. 24,

3607-3617. 6. Morita, Y. & !da, S. (1968) Agric. Bio/. Chem. 32, 441 - 447. 7. Morita, Y. , Yashida, C., Kitamura, !. & Ida, S. (1970) Agric.

Bio/. Chem. 34, 1191 - 1197 . . 8. Kon. S. & Whitaker, J . R. (1965) J. Food Sei. 30, 977- 985. 9. El-Fekih, M. & Kertesz, D. (1968) Bull. Soc. Chim. Bio/. 50.

547-568. 10. Beaudreau, C. & Yasunobu, K. T. (1966) Biochemistry, 5,

1405-1412. 11. Mazza, G., Charles, C., Bouchet, M ., Ricard, J. & Raynaud, J.

(1968) Biochim. Biophys. Acta, 167, 89-98. 12. Klapper, M. H. & Hackett, D. P. (1965) Biochim. Biophys.

Acta, 96, 272-282. 13. Siegel, B. Z. & Galston, A. W. (1967) Plant Physiol. 42, 221-

226. 14. Brewbaker,J. L., Upadhya, M. D„ Mäkinen, Y. & MacDonald,

T. (1968) Physiol. Plant. 21, 930 - 940. 15. R acusen, D. & Foote, M. (1968) Anal. Bioclzem. 25, 164-171.

H. Dclincee and B. J. Radola

H. Ddincee and B. J. Radola: Fractionation of Peroxidase

16. Hall, T. C., McCown, B. H .. Desborough, S .. McLeester. R. C. & Beck, G. E. (1969) Phytochemistry, 8. 385- 391.

17. Hamill, D. E. & Brewbaker, J. L. (1969) Physiol. Plant. 22. 945-958.

·. 18. Haard, N. F. & Tobin, C. L. (1971) J . Food Sei. 36, 854-857. 19. Delincee, H. & Radola, B. J. (1970) Biochim. Biophys. Acta,

200, 404- 407. 20. Radola, B. J. & Drawert, F. (1970) Brau"·issenschaft 23, 449-

458. 21. Rücker, W. & Radola, B. J. (1971) Plmlta ( Berl.) 99, 192-198. 22. Theorell , H. (1942) Ark. Kern. Mineral. Geal. 16A, No. 2: 23. Whitehead, J. S., Kay, E., Lew, J. Y. & Shannon, L. M. (1971)

Anal. Biochem. 40, 287- 291. 24. Evans, J. J . (1969) Anal. Biochem . 32, 101-105. 25. Radola, B. J. (1973) Ann. N. Y. Acnd. Sei. 209, 127-143. 26. Radola, B. J . (1974) in lsoelectric Focusing (Arbuthnott, J . P„

ed.) Butterworth, London. 27. Radola, B. J. (1974) Biochim, Biophys. Acta, in press. 28. Chance. B. & Maehly, A. C. (1955) Methods En~ymol. 2, 764-

775. 29. Radola. B. J . (1973) Biochim. Biophys. Acta, 295, 412-428. 30. Delincee, H. & Radola. B. J. (1972) Anal. Biochem. 48, 536-

545. 31. Radola, B. J. (1968) J. Chromatogr. 38, 61- 77. 32. Delincee, H . & Radola. B. J. (1971) in Protides of the Bio/ogical

Fluids (Peeters, H. , ed.) Vol. 18, pp. 493-497, Pergamon Press. Oxford.

33. Radola. B. J. & Delincee, H. (1971) J . Chromatogr. 61 , 365-369.

34. Delincee, H. & Radola, B. J. (1972) Z. Anal. Chem. 259, 223 -224.

35. Vesterberg, 0 . (1 971) Methods Enzymol. 22, 389-412.

Institut für Strahlentechnologie, Bundesforschungsanstalt für Ernährung, D-7500 Karlsruhe. Engesserstraße 20, Federal Republic of Germany

Eur. J. Biochem. 52 (1975)

„