ANALYTICAL
BIOCHEMISTRY
200,36-41
(1992)
Cytochrome c Aided Resolution of Lupinus a/bus lsoperoxidases in a Cathodal Polyacrylamide Gel Electrophoresis System Philip
Jackson*
and CBndido
P. P. Ricardo**P
*Centro de TecnologiaQuimica e Biolbgica, Apartado 127, 2780 Oeiras,Portugal, and tDepartament0 ok BotSca e &agenhuria Bioldgica, Institute Superior de Agronomia, 1399 Lisboa Codex, Portugal
Received
May
28,199l
In a cathodal polyacrylamide gel electrophoresis system, three distinct groups of isoperoxidases from Lupinus &us were found to achieve retention factors (rf) dependent on the quantity of sample applied onto the gel. The possibility of extract-derived substances weakly associating with peroxidase samples was investigated. Association of the putative agents survived dialysis against electrophoresis buffer with and without 2M CaCl, and freeze-thaw treatments. The addition of polyvinylpolypyrrolidone and polyethylene glycol to the homogenization buffer also proved ineffective in eliminating the variation in isoperoxidase rf although differences in the zymogram profiles of these samples were evident. The addition of spermine and cytochrome c to samples was found to increase the rf of some peroxidase bands. Electrophoresis of samples with cytochrome c resulted in the resolution of peroxidase groups to distinct bands at rf independant of the quantity of peroxidase applied. Control experiments indicate that this treatment did not introduce any detectable artifacts.
0 1992
Academic
Press,
Inc.
Plant peroxidases (EC 1.11.1.7) have been implicated in cell differentiation (l-4) and growth processes (5) and are believed to control important metabolic functions such as lignification (6-7), indole acetic acid oxidation (2), and cell wall cross-linking through the formation of diphenol bridges (8-10). The electrophoretic zymogram technique can be an excellent tool for the study of peroxidases and their variation in relation to physiological processes (1,11,12). When separation of sample proteins with a large range of molecular weights and charges is desired, however, cathodal or anodal electrophoresis of nondenatured enzymes may provide a simple, one-step method with im-
proved separation due to the additional resolution of differing molecular weights. Unexpected zymogram patterns have been observed using this electrophoretic system with at least two enzymes, lactate dehydrogenase (13-14) and RNAse A (15-16). In each case, the observed rf of stained bands was dependent on the amount of sample applied. We report a further example of this behavior in Lupinus albus isoperoxidases and of its correction by the application of cytochrome c to samples prior to electrophoresis. MATERIALS
AND
METHODS
Tissue Source
Roots of L. albus cv. “Rio Maior” were taken from 16-day-old seedlings which were grown in vermiculite under a light/dark cycle of 16/8 h at 24°C and watered with tap water daily. Extraction
of Peroxidases
Tissues were homogenized in 2 ml/g (fwt) of extraction buffer (50 mM Tris-HCl, pH 7.2, containing 5 mM PMSF’) by grinding in a mortar with pestle. Where stated, the homogenate was prepared in the presence of PVPP at 2.5 or 5.0% (w/v) or PEG at 1% (w/v). The homogenate was passed through a 200-pm nylon net and centrifuged at 4500g for 10 min to yield soluble and cell wall preparations. All stages of extraction were performed at 4°C. The soluble fraction was reduced to 0.6 ml/g tissue fwt by vacuum ultrafiltration through Milipore CXlO ultrafilters and was desalted through Sephadex G25 1 Abbreviations used: PAG, polyacrylamide gel; PVPP, polyvinylpolypyrrolidone; PEG, polyethylene glycol; PMFS, phenylmethylsulfonylfluoride; rf, retention factors; IEF, isoelectric focusing.
36 All
0003-2697/92 $3.00 Copyright 0 1992 by Academic Press, Inc. rights of reproduction in any form reserved.
CYTOCHROME
c AIDED
ELECTROPHORESIS
PD-10 columns (Pharmacia) equilibrated with o-alanine (1.75 m&-acetic acid (0.75 InM) buffer (pH 4.5). Cell membranes were removed from the crude cell wall preparation by resuspension in extraction buffer containing 1% v/v Triton X-100 before centrifugation at 4500g for 10 min followed by washing of the pellet 3X in Tris-HCl buffer. The purified cell wall preparation was then extracted with 1 M KC1 for 40 min at 1 ml/g fwt (12,17). When PEG was employed in the preparation of the initial homogenate, PEG at 1.0% (w/v) was also included in the 1 M KC1 extraction solution. Proteins were separated from the cell wall debris by centrifugation at 4500g for 10 min and reduced in volume to 0.6 ml/g fwt using Milipore CX-10 ultrafilters or Amicon PG 20 centriflow. The preparations were then desalted on equilibrated PD-10 columns. Where stated, samples prepared without the use of phenolic scavengers were adjusted to 5.0% PVPP (w/v) in 50 mM Tris-HCl (pH 7.2) buffer and mixed for 30 min prior to centrifugation at 4500g for 20 min. The level of phenolic contamination in samples was assayed by the measurement of the absorbance ratio 253 nm/207 nm (12), where high ratios were taken to indicate high levels of contamination. Measurement
of Peroxidase
Activity
Guaiacol peroxidase activity was measured by continuous assay at 25°C in 0.1 M succinic acid (pH 5.0) with 1 mM guaiacol and 1 mM H,O,. One unit of peroxidase activity was defined as being equivalent to a change of 0.1 absorbance units/min at 450 nm (modified after Sanchez et al. (5)). Polyacrylamide
Gel Electrophoresis
PAG electrophoresis was performed using 18 X 16-cm and 1.5-mm thick 7.5% polyacrylamide gels buffered with electrophoresis buffer (35 KnM P-alanine-14 mM acetic acid, pH 4.5) and polymerized using 17.5 mM ammonium persulfate and 2.7 mM Temed. Samples were loaded in 8-mm wells with glycerol to 10% (v/v) and basic fuchsin was used as front marker. Where indicated in the text, cytochrome c (horse heart, purchased from Sigma Chemical Co.) was added to samples immediately before their application to sample wells. Resolved gels were stained for isoperoxidases by immersion in 0.1 M sodium phosphate buffer (pH 5.0) containing 1 mM guaiacol and 1 mM H,O, over 2 h at room temperature. RESULTS
Relationship
between rf and Amount
of Sample
Applied
In preliminary comparisons of L. albus isoperoxidases of root tissues, PAG electrophoresis of native isoperoxidases was found to show poor band resolution and vari-
OF PEROXIDASES
37
able band rf, hence invalidating any meaningful comparisons. Samples were applied to gels over a range of concentration by dilution of a single stock preparation in sample loading buffer and it was evident that the rf of some stained bands were related to the quantity of peroxidase activity applied (Fig. 1). This relation was examined closer using the ionically bound protein fraction of root cell walls. Figures 1B and 1C show that two distinct staining groups contained variations in band rf in cell wall proteins. The rf of these two groups demonstrated proportionality to the peroxidase activity over the range of lo-40 units guaiacol peroxidase. Above 60 U the rf were independent of the applied peroxidase activity. The higher rf group in Fig. 1C (G2) obtained a relatively stable rf at 40 U, the lower rf group (Gl) at 60 U.
The Role of Tissue-Derived Compounds in the Relation of rf to the Peroxidase Activity Applied Since samples applied to each gel were derived from one single preparation, variation in extraction procedures or states of degradation could not explain the effect observed. Similarly, covalent binding of tissuederived contaminants such as quinones would lead to stable rf. Phenolic compounds are recognized as problematic in the extraction of plant enzymes (18). Hydrogen bonding and ionic interactions between phenolics and proteins are known to occur (19,12) and recently have been implicated as a source of variation in peroxidase rf in nondenaturing electrophoretic systems (20,12). The effect of the ionic strength of sample buffer and dialysis of samples against electrophoresis buffer with or without 2 M CaCl, on Gl and G2 migration was investigated. Additionally, samples were subjected to freezethaw treatments over 72 h in 1 M KC1 (7), a treatment recently shown to alter electrophoretic mobility of L. albus peroxidases by the removal of associated phenols (12). Neither of these treatments was able to alter the electrophoretic mobility of Gl or G2 (data not shown). Differences between samples extracted with or without PVPP or PEG were apparent. The use of these phenol scavengers led to a decrease in the yield of peroxidase activity/g fwt (35-55% decreased yield from control preparations). Differential effects of PVPP added to the homogenization buffer or following sample preparation were found. Samples extracted in the presence of PVPP gave an increased absorbance ratio 253 nm/207 nm, but samples treated with PVPP after preparation produced a slight decrease. Electrophoresis of these extracts showed that PVPP treatment after the isolation of cell wall proteins produced no differences in electrophoretic behavior in control samples. Extracts prepared with PVPP or PEG within the homogeniza-
38
JACKSON
1234567
AND
RICARDO
12345670
FIG. 1.
Continuous nondenaturing cathodic PAG electrophoresis of L. albus root extracts. 40,50,60,70,80,90, and 100 units guaiacol peroxidase, respectively. Ionic cell wall fractions and 50 units guaiacol peroxidase respectively to lanes l-8 (B), and at 10,20,30,40,50,60,80, (0.
tion buffer, however, showed striking differences, as displayed in Fig. 2. Figure 2 shows that in extracts prepared with PVPP or PEG within the homogenization buffer, the change in rf of Gl and G2 per U peroxidase applied was decreased. The control preparation (Fig. 2A) demonstrates that Gl and G2 maintained relatively close positions over a fairly wide peroxidase range. All samples extracted in the presence of phenolic scavengers, however, showed that G2 achieved a stable rf at a low peroxidase activity whereas the mobility of Gl was reduced. The Effect of Cationic Compounds on the Relation of rf to Activity of Peroxidase Applied
The effect of addition of the substances spermine and cytochrome c on Gl and G2 migratory behavior was tested. Spermine and cytochrome c were applied to samples immediately before electrophoresis. The application of spermine to samples at 100 and 200 mM was effective in shifting the rf of both Gl and G2 to higher values, but resolution remained poor (data not shown). Cytochrome c applied at 1.0 and 1.5 mg/ml to cell wall
Soluble fractions (A) were applied to lanes l-7 at (B and C) were applied at 10,15,20,25,30,35,40, and 100 units guaiacol peroxidase to lanes l-8
samples was able to shift both Gl and G2 to higher rf with excellent staining resolution. Similar results were obtained with soluble extracts (Fig. 3). Furthermore, a comparison of the mobilities of samples coelectrophoresed with cytochrome c and control samples in the same gel showed that the rf of cytochrome c loaded samples were independent of the sample quantity applied (Fig. 3). Cytochrome c has peroxidase activity that could produce artifactual staining activity of peroxidase samples. In this electrophoretic system cytochrome c migrated in front of peroxidase bands and contributed no detectable staining activity to the zymogram when electrophoresed alone in concentrations utilized for resolution of peroxidases (Fig. 4A). Evidence that resolution of samples with cytochrome c did not produce artifactual bands was seen in the similarity of band rf in these samples to samples of high peroxidase activity (Fig. 4A). In order to confirm that the peroxidase zymograms obtained with the use of cytochrome c were obtained by the resolution of Gl and G2 peroxidases, cytochrome c was applied to identical peroxidase samples over the range O-O.2 mgl ml sample. Two peroxidase concentrations were chosen:
12345678
FIG. 2. Comparison of the electrophoretic mobility of Gl and G2 guaiacol peroxidases in extracts prepared with or without phenolic scavengers. Samples prepared without phenolic scavengers (A) were applied at 3,6,9,12,15,18,21, and 24 units guaiacol to lanes l-8, respectively. Samples prepared with 1% PEG (w/v) (B) and 2.5% PVPP (w/v) (C) were applied at 5,10,15,20,30,40,50, units guaiacol peroxidase to lanes 1-8, respectively. The dilute peroxidase range shown in (A) was chosen in order to demonstrate separation of Gl and G2 in control preparations.
the use of peroxidase and 60 maximum
CYTOCHROME
c AIDED
ELECTROPHORESIS
OF
39
PEROXIDASES
FIG. 3. The effect of cytochrome c in peroxidase zymograms of ionic cell wall (A) and soluble extracts (B). Ionic cell wall samples with cytochrome c at 1.0 mg/ml were applied to lanes 1,3,5,7, and 9 at 10,20,40,60, and 80 units guaiacol peroxidase, respectively. Samples without cytochrome c were applied to lanes 2,4,6,8, and 10 at 10,20,40,60, and 80 units guaiacol peroxidase, respectively. Cytochrome c was applied to soluble extracts of 140 units guaiacol peroxidase at 0 mg/ml (lanes l-3), 1.0 mg/ml (lanes 4-6), and 1.5 mg/ml (lanes 7-9).
26.5 and 53 U. Figures 4B and 4C show that three stained bands resulted from the resolution of G2 with dilute mixtures of cytochrome c and peroxidase sample (0.033 mg/ml cytochrome c, 26.5 U peroxidase) and that a further three bands at higher rf resulted from the resolution of Gl and could only be resolved with more concentrated mixtures (0.2 mg/ml cytochrome c, 53 U peroxidase) because at intermediate concentrations Gl rf were identical to those of G2. Readily observable was the similarity between the relation of the rf to the amount of cytochrome c and the increase in rf with activity of peroxidase applied in samples without cytochrome c. In both cases a proportionality existed between the rf and the quantities applied until high loading quantities, when a final rf was approached with increased resolution. This suggests that the causative agent in the variation of Gl and G2 rf could be effectively reduced in a similar manner by the
12
34567
6
12
34
addition of greater quantities of sample proteins or cytochrome c. These results indicate that zymograms of isoperoxidases obtained with the use of cytochrome c were obtained by the resolution of Gl and G2 peroxidase groups without the introduction of artifactual staining activity. DISCUSSION
The Relation
of rf to the Activity
of Peroxidase
Applied
The relation of rf to enzymic activity applied has been observed in a cathodal electrophoresis system for at least two other enzymes. RNAse was found to achieve variable rf in /3-alanine-acetate buffer (15,16). The authors suggested that the rf of RNase was dependent on the sample quantity applied, and further suggested that the effect was mediated through buffer anions. Other reports concerning lactate dehydrogenase reported simi-
5676
910
12345676
910
c in peroxidase zymograms (A) and cytochrome c aided resolution of Gl and G2 peroxidase groups (B and FIG. 4. The effect of cytochrome C). The sample content of the lanes was as follows: (A) (l-2) 1.0 mg/ml cytochrome c, (3-4) 30 U peroxidase, (5-6) 30 U peroxidase and 1.0 mg/ml cytochrome c, (7-8) 900 U peroxidase. All lanes in zymogram B contained 26.5 U guaiacol peroxidase and those of C contained 53.0 U guaiacol peroxidase. Cytochrome c was applied to both gels following the order: (l-2) 0 mg/ml; (3-8) 0.033,0.066,0.100,0.133,0.166, and 0.200 mg/ml, respectively; (9-10) 1.0 mg/ml.
40
JACKSON
AND
lar effects with the most basic isozyme (13,14). Here, evidence has been presented for a similar effect in the migration of three groups of isoperoxidases from L. albus. The use of PVPP and PEG to reduce phenolic contamination in plant extracts has been advised (18). These reagents can act in a competitive manner with proteins for phenolic compounds. Consistent with the removal of tissue-derived phenols from peroxidase samples in this work is the reduction of the observed change in rf with activity of peroxidase applied in samples prepared with PVPP and PEG in the extraction buffer. Cationic phenols have been found in association with cell wall isoperoxidases of L. albus and their removal resulted in a change in the p1 of two basic forms to more acid states (12). Here, Gl isoperoxidases were shown to adopt an apparently more acidic state with the use of PVPP or PEG in the extraction buffer. However, this effect could not be seen in samples prepared without the use of phenolic scavengers and subsequently subject to PVPP, suggesting that removal of tissue-derived contaminants cannot account for the electrophoretic behavior of samples extracted in the presence of PVPP. The increased absorbance ratio 253 nm/206 nm of these extracts suggests that the electrophoretic behavior of these samples might be mediated through additional contamination. Interestingly, cytochrome c was able to increase guaiacol peroxidase activity in zymograms. A similar observation was made for the effects of cationic compounds on the observed activity in zymograms of RNAse (15). In our experiments, the activity promoting effect of cytochrome c appears restricted to samples prepared with PVPP or PEG (compare the staining activities of Fig. 3 and Figs. 4B and 4C). This implies that these samples were applied at a supposedly lower activitylmg isoperoxidase due to the presence of inhibitors. This may explain the resolution of G2 peroxidases with lower (15 units) activity when compared to that of samples prepared without PVPP (40 units). Similarly, the apparent charge difference observed in Gl isoperoxidases might be explained by an association of tissuederived contamination. It is feasable therefore that tissue-derived contamination may also interact with Gl and G2 isoperoxidases. However, this interaction seems to be of a different nature from that described for samples prepared without phenolic scavengers, and so difficult to relate to our main observations with cytochrome c. The Role of Buffer Anions in the Relation of rf to Peroxidase Activity Applied The ability of spermine to increase the mobility of peroxidases and the excellent resolution of samples with cytochrome c led us to investigate the role of buffer anions in the relation of peroxidase rf to the quantity of
RICARDO
activity applied. Variation of the anionic content of electrophoresis buffer was achieved by the use of P-alaninebased solutions (35 InM) adjusted to pH 4.5 with either acetate, succinate, or citrate. The results showed that higher buffer anion content could decrease the rf of Gl and G2, suggesting the variation in Gl and G2 rf might be mediated through an interaction of peroxidases with anions (data not shown). The application of suboptimal levels of cytochrome c (O-O.2 mg/ml) to samples with identical quantities of peroxidases could simulate the relation of rf to the activity of peroxidase applied, showing that Gl and G2 rf were increased with increasing cationic residues, whether contributed by cytochrome c or the samples themselves. These results suggest that cytochrome c might aid the resolution of peroxidases by providing cationic residues with which buffer anions may interact. The subsequently reduced interaction of buffer anions with peroxidase enzymes therefore allows for the full expression of protein charge and sample resolution. Because of the widespread occurrence of proteins that are positively charged at acid pH, the same reasoning can be applied to the increased resolution obtained with greater quantities of peroxidase samples alone. The interaction of spermine and cytochrome c with acetate buffer and peroxidase with acetate, succinate, and citrate buffers suggests a certain nonspecificity of these cation-anion interactions. Similar observations were made in the improved resolution of RNAse by a variety of cationic sources (15). Yet comparable results with other protein samples has rarely been described (16). One possible reason for the detection of this effect here is not that L. albus isoperoxidases represent an example of unusual proteins but rather is the sensitivity of the staining reaction (16). Guaiacol peroxidase activity could easily be detected in zymograms when as little as 10 units were applied, allowing for the application of small sample quantities. Less sensitive staining methods would necessitate the application of greater sample quantities in which case the effect would be reduced. Second, preparation of enzyme samples for zymograms does not usually require other than simple extraction methods. Many enzymes would exist at low specific activity in these preparations, wherein the high total protein content of these samples would minimize the interaction of anions with the enzyme. Peroxidase, however, is abundant in plant cells and simple extraction procedures often produce samples of significant specific activity, allowing the application of low quantities of total protein, and consequently, a greater interaction of anions with peroxidase. The use of cytochrome c has now enabled us to continue with our work with L. albus isoperoxidases of different tissues, each of which have shown improved reso-
CYTOCHROME
c AIDED
ELECTROPHORESIS
lution and quantity independent band rf in response to cytochrome c addition. Due to the widespread occurrence of proteins that are positively charged at acidic pH, the advantages offered by the use of cytochrome c in a cathodal electrophoretic system are not necessarily limited to exceptional cases, but might be applicable to other peroxidases, enzymes prepared to high specific activity, or purified proteins studied in conjunction with sensitive staining methods. ACKNOWLEDGMENTS Financial support for this work was received from JNICT Project 87.206lBIO.
REFERENCES 1. Joerbso, M. (1989) Physiol. Plant. 76, 10-16. 2. Mato, M. C., and Vieitez, A. M. (1986) Physiol. Plant. 66, 491494. 3. Mato, M. C., Rua, M. L., and Ferro, E. (1988) Physiol. Plant. 72, 84-88. 4. Rao, K. V., Suprasanna, P., and Reddy, G. M. (1990) Plant Sci. 66,127-130.
OF PEROXIDASES
41
5. Sanchez, 0. J., Pan, A., Nicolas, G., and Labrador, E. (1989) Physiol. Plant. 76,275-279. 6. Verma, D. P. S., and van Huystee, R. B. (1970) Can. J. Bot. 48, 429-431. 7. Mader, M. (1980) Z. Pflanzenphysiol. 96, 283-296. 8. Fry, S. C. (1982) Biochem. J. 203,493-504. 9. Fry, S. C. (1982) Biochem. J. 204,449-455. 10. Fry, S. C. (1979) Planta 146,343-351. 11. Hunter, R. L., and Markert, C. L. (1957) Science 126,1294-1295. 12. Barcelo, A. R., Munoz, R., and Sabater, F. (1987) Physiol. Plant. 71,448-454. 13. Vessell, E. S. (1962) Nature 195,497-498. 14. Ressler, N., Shulz, J., and Joseph, R. R. (1963) J. Lab. C&n. Med. 62.571-578. 15. Thomas, J. M., and Hodes, M. E. (1981) Ekctrophoresis 2, 104112. 16. Thomas, J. M., and Hodes, M. E. (1981) Experientiu 37,457-459. 17. Goldberg, R. (1980) Physiol. Plant. 50, 261-264. 18. Loomis, W. D., and Battaile, J. (1966) Phytochem. 5, 423-438. 19. Swain, T. (1965) in Plant Biochemistry (Bonner, J., and Varner, J. E., Eds.), p. 552, Academic Press, New York/London. 20. Srivistava, 0. P., and van Huystee, R. (1977) Bot. Gaz. 138,457464.
BIOCHEMISTRY
200,36-41
(1992)
Cytochrome c Aided Resolution of Lupinus a/bus lsoperoxidases in a Cathodal Polyacrylamide Gel Electrophoresis System Philip
Jackson*
and CBndido
P. P. Ricardo**P
*Centro de TecnologiaQuimica e Biolbgica, Apartado 127, 2780 Oeiras,Portugal, and tDepartament0 ok BotSca e &agenhuria Bioldgica, Institute Superior de Agronomia, 1399 Lisboa Codex, Portugal
Received
May
28,199l
In a cathodal polyacrylamide gel electrophoresis system, three distinct groups of isoperoxidases from Lupinus &us were found to achieve retention factors (rf) dependent on the quantity of sample applied onto the gel. The possibility of extract-derived substances weakly associating with peroxidase samples was investigated. Association of the putative agents survived dialysis against electrophoresis buffer with and without 2M CaCl, and freeze-thaw treatments. The addition of polyvinylpolypyrrolidone and polyethylene glycol to the homogenization buffer also proved ineffective in eliminating the variation in isoperoxidase rf although differences in the zymogram profiles of these samples were evident. The addition of spermine and cytochrome c to samples was found to increase the rf of some peroxidase bands. Electrophoresis of samples with cytochrome c resulted in the resolution of peroxidase groups to distinct bands at rf independant of the quantity of peroxidase applied. Control experiments indicate that this treatment did not introduce any detectable artifacts.
0 1992
Academic
Press,
Inc.
Plant peroxidases (EC 1.11.1.7) have been implicated in cell differentiation (l-4) and growth processes (5) and are believed to control important metabolic functions such as lignification (6-7), indole acetic acid oxidation (2), and cell wall cross-linking through the formation of diphenol bridges (8-10). The electrophoretic zymogram technique can be an excellent tool for the study of peroxidases and their variation in relation to physiological processes (1,11,12). When separation of sample proteins with a large range of molecular weights and charges is desired, however, cathodal or anodal electrophoresis of nondenatured enzymes may provide a simple, one-step method with im-
proved separation due to the additional resolution of differing molecular weights. Unexpected zymogram patterns have been observed using this electrophoretic system with at least two enzymes, lactate dehydrogenase (13-14) and RNAse A (15-16). In each case, the observed rf of stained bands was dependent on the amount of sample applied. We report a further example of this behavior in Lupinus albus isoperoxidases and of its correction by the application of cytochrome c to samples prior to electrophoresis. MATERIALS
AND
METHODS
Tissue Source
Roots of L. albus cv. “Rio Maior” were taken from 16-day-old seedlings which were grown in vermiculite under a light/dark cycle of 16/8 h at 24°C and watered with tap water daily. Extraction
of Peroxidases
Tissues were homogenized in 2 ml/g (fwt) of extraction buffer (50 mM Tris-HCl, pH 7.2, containing 5 mM PMSF’) by grinding in a mortar with pestle. Where stated, the homogenate was prepared in the presence of PVPP at 2.5 or 5.0% (w/v) or PEG at 1% (w/v). The homogenate was passed through a 200-pm nylon net and centrifuged at 4500g for 10 min to yield soluble and cell wall preparations. All stages of extraction were performed at 4°C. The soluble fraction was reduced to 0.6 ml/g tissue fwt by vacuum ultrafiltration through Milipore CXlO ultrafilters and was desalted through Sephadex G25 1 Abbreviations used: PAG, polyacrylamide gel; PVPP, polyvinylpolypyrrolidone; PEG, polyethylene glycol; PMFS, phenylmethylsulfonylfluoride; rf, retention factors; IEF, isoelectric focusing.
36 All
0003-2697/92 $3.00 Copyright 0 1992 by Academic Press, Inc. rights of reproduction in any form reserved.
CYTOCHROME
c AIDED
ELECTROPHORESIS
PD-10 columns (Pharmacia) equilibrated with o-alanine (1.75 m&-acetic acid (0.75 InM) buffer (pH 4.5). Cell membranes were removed from the crude cell wall preparation by resuspension in extraction buffer containing 1% v/v Triton X-100 before centrifugation at 4500g for 10 min followed by washing of the pellet 3X in Tris-HCl buffer. The purified cell wall preparation was then extracted with 1 M KC1 for 40 min at 1 ml/g fwt (12,17). When PEG was employed in the preparation of the initial homogenate, PEG at 1.0% (w/v) was also included in the 1 M KC1 extraction solution. Proteins were separated from the cell wall debris by centrifugation at 4500g for 10 min and reduced in volume to 0.6 ml/g fwt using Milipore CX-10 ultrafilters or Amicon PG 20 centriflow. The preparations were then desalted on equilibrated PD-10 columns. Where stated, samples prepared without the use of phenolic scavengers were adjusted to 5.0% PVPP (w/v) in 50 mM Tris-HCl (pH 7.2) buffer and mixed for 30 min prior to centrifugation at 4500g for 20 min. The level of phenolic contamination in samples was assayed by the measurement of the absorbance ratio 253 nm/207 nm (12), where high ratios were taken to indicate high levels of contamination. Measurement
of Peroxidase
Activity
Guaiacol peroxidase activity was measured by continuous assay at 25°C in 0.1 M succinic acid (pH 5.0) with 1 mM guaiacol and 1 mM H,O,. One unit of peroxidase activity was defined as being equivalent to a change of 0.1 absorbance units/min at 450 nm (modified after Sanchez et al. (5)). Polyacrylamide
Gel Electrophoresis
PAG electrophoresis was performed using 18 X 16-cm and 1.5-mm thick 7.5% polyacrylamide gels buffered with electrophoresis buffer (35 KnM P-alanine-14 mM acetic acid, pH 4.5) and polymerized using 17.5 mM ammonium persulfate and 2.7 mM Temed. Samples were loaded in 8-mm wells with glycerol to 10% (v/v) and basic fuchsin was used as front marker. Where indicated in the text, cytochrome c (horse heart, purchased from Sigma Chemical Co.) was added to samples immediately before their application to sample wells. Resolved gels were stained for isoperoxidases by immersion in 0.1 M sodium phosphate buffer (pH 5.0) containing 1 mM guaiacol and 1 mM H,O, over 2 h at room temperature. RESULTS
Relationship
between rf and Amount
of Sample
Applied
In preliminary comparisons of L. albus isoperoxidases of root tissues, PAG electrophoresis of native isoperoxidases was found to show poor band resolution and vari-
OF PEROXIDASES
37
able band rf, hence invalidating any meaningful comparisons. Samples were applied to gels over a range of concentration by dilution of a single stock preparation in sample loading buffer and it was evident that the rf of some stained bands were related to the quantity of peroxidase activity applied (Fig. 1). This relation was examined closer using the ionically bound protein fraction of root cell walls. Figures 1B and 1C show that two distinct staining groups contained variations in band rf in cell wall proteins. The rf of these two groups demonstrated proportionality to the peroxidase activity over the range of lo-40 units guaiacol peroxidase. Above 60 U the rf were independent of the applied peroxidase activity. The higher rf group in Fig. 1C (G2) obtained a relatively stable rf at 40 U, the lower rf group (Gl) at 60 U.
The Role of Tissue-Derived Compounds in the Relation of rf to the Peroxidase Activity Applied Since samples applied to each gel were derived from one single preparation, variation in extraction procedures or states of degradation could not explain the effect observed. Similarly, covalent binding of tissuederived contaminants such as quinones would lead to stable rf. Phenolic compounds are recognized as problematic in the extraction of plant enzymes (18). Hydrogen bonding and ionic interactions between phenolics and proteins are known to occur (19,12) and recently have been implicated as a source of variation in peroxidase rf in nondenaturing electrophoretic systems (20,12). The effect of the ionic strength of sample buffer and dialysis of samples against electrophoresis buffer with or without 2 M CaCl, on Gl and G2 migration was investigated. Additionally, samples were subjected to freezethaw treatments over 72 h in 1 M KC1 (7), a treatment recently shown to alter electrophoretic mobility of L. albus peroxidases by the removal of associated phenols (12). Neither of these treatments was able to alter the electrophoretic mobility of Gl or G2 (data not shown). Differences between samples extracted with or without PVPP or PEG were apparent. The use of these phenol scavengers led to a decrease in the yield of peroxidase activity/g fwt (35-55% decreased yield from control preparations). Differential effects of PVPP added to the homogenization buffer or following sample preparation were found. Samples extracted in the presence of PVPP gave an increased absorbance ratio 253 nm/207 nm, but samples treated with PVPP after preparation produced a slight decrease. Electrophoresis of these extracts showed that PVPP treatment after the isolation of cell wall proteins produced no differences in electrophoretic behavior in control samples. Extracts prepared with PVPP or PEG within the homogeniza-
38
JACKSON
1234567
AND
RICARDO
12345670
FIG. 1.
Continuous nondenaturing cathodic PAG electrophoresis of L. albus root extracts. 40,50,60,70,80,90, and 100 units guaiacol peroxidase, respectively. Ionic cell wall fractions and 50 units guaiacol peroxidase respectively to lanes l-8 (B), and at 10,20,30,40,50,60,80, (0.
tion buffer, however, showed striking differences, as displayed in Fig. 2. Figure 2 shows that in extracts prepared with PVPP or PEG within the homogenization buffer, the change in rf of Gl and G2 per U peroxidase applied was decreased. The control preparation (Fig. 2A) demonstrates that Gl and G2 maintained relatively close positions over a fairly wide peroxidase range. All samples extracted in the presence of phenolic scavengers, however, showed that G2 achieved a stable rf at a low peroxidase activity whereas the mobility of Gl was reduced. The Effect of Cationic Compounds on the Relation of rf to Activity of Peroxidase Applied
The effect of addition of the substances spermine and cytochrome c on Gl and G2 migratory behavior was tested. Spermine and cytochrome c were applied to samples immediately before electrophoresis. The application of spermine to samples at 100 and 200 mM was effective in shifting the rf of both Gl and G2 to higher values, but resolution remained poor (data not shown). Cytochrome c applied at 1.0 and 1.5 mg/ml to cell wall
Soluble fractions (A) were applied to lanes l-7 at (B and C) were applied at 10,15,20,25,30,35,40, and 100 units guaiacol peroxidase to lanes l-8
samples was able to shift both Gl and G2 to higher rf with excellent staining resolution. Similar results were obtained with soluble extracts (Fig. 3). Furthermore, a comparison of the mobilities of samples coelectrophoresed with cytochrome c and control samples in the same gel showed that the rf of cytochrome c loaded samples were independent of the sample quantity applied (Fig. 3). Cytochrome c has peroxidase activity that could produce artifactual staining activity of peroxidase samples. In this electrophoretic system cytochrome c migrated in front of peroxidase bands and contributed no detectable staining activity to the zymogram when electrophoresed alone in concentrations utilized for resolution of peroxidases (Fig. 4A). Evidence that resolution of samples with cytochrome c did not produce artifactual bands was seen in the similarity of band rf in these samples to samples of high peroxidase activity (Fig. 4A). In order to confirm that the peroxidase zymograms obtained with the use of cytochrome c were obtained by the resolution of Gl and G2 peroxidases, cytochrome c was applied to identical peroxidase samples over the range O-O.2 mgl ml sample. Two peroxidase concentrations were chosen:
12345678
FIG. 2. Comparison of the electrophoretic mobility of Gl and G2 guaiacol peroxidases in extracts prepared with or without phenolic scavengers. Samples prepared without phenolic scavengers (A) were applied at 3,6,9,12,15,18,21, and 24 units guaiacol to lanes l-8, respectively. Samples prepared with 1% PEG (w/v) (B) and 2.5% PVPP (w/v) (C) were applied at 5,10,15,20,30,40,50, units guaiacol peroxidase to lanes 1-8, respectively. The dilute peroxidase range shown in (A) was chosen in order to demonstrate separation of Gl and G2 in control preparations.
the use of peroxidase and 60 maximum
CYTOCHROME
c AIDED
ELECTROPHORESIS
OF
39
PEROXIDASES
FIG. 3. The effect of cytochrome c in peroxidase zymograms of ionic cell wall (A) and soluble extracts (B). Ionic cell wall samples with cytochrome c at 1.0 mg/ml were applied to lanes 1,3,5,7, and 9 at 10,20,40,60, and 80 units guaiacol peroxidase, respectively. Samples without cytochrome c were applied to lanes 2,4,6,8, and 10 at 10,20,40,60, and 80 units guaiacol peroxidase, respectively. Cytochrome c was applied to soluble extracts of 140 units guaiacol peroxidase at 0 mg/ml (lanes l-3), 1.0 mg/ml (lanes 4-6), and 1.5 mg/ml (lanes 7-9).
26.5 and 53 U. Figures 4B and 4C show that three stained bands resulted from the resolution of G2 with dilute mixtures of cytochrome c and peroxidase sample (0.033 mg/ml cytochrome c, 26.5 U peroxidase) and that a further three bands at higher rf resulted from the resolution of Gl and could only be resolved with more concentrated mixtures (0.2 mg/ml cytochrome c, 53 U peroxidase) because at intermediate concentrations Gl rf were identical to those of G2. Readily observable was the similarity between the relation of the rf to the amount of cytochrome c and the increase in rf with activity of peroxidase applied in samples without cytochrome c. In both cases a proportionality existed between the rf and the quantities applied until high loading quantities, when a final rf was approached with increased resolution. This suggests that the causative agent in the variation of Gl and G2 rf could be effectively reduced in a similar manner by the
12
34567
6
12
34
addition of greater quantities of sample proteins or cytochrome c. These results indicate that zymograms of isoperoxidases obtained with the use of cytochrome c were obtained by the resolution of Gl and G2 peroxidase groups without the introduction of artifactual staining activity. DISCUSSION
The Relation
of rf to the Activity
of Peroxidase
Applied
The relation of rf to enzymic activity applied has been observed in a cathodal electrophoresis system for at least two other enzymes. RNAse was found to achieve variable rf in /3-alanine-acetate buffer (15,16). The authors suggested that the rf of RNase was dependent on the sample quantity applied, and further suggested that the effect was mediated through buffer anions. Other reports concerning lactate dehydrogenase reported simi-
5676
910
12345676
910
c in peroxidase zymograms (A) and cytochrome c aided resolution of Gl and G2 peroxidase groups (B and FIG. 4. The effect of cytochrome C). The sample content of the lanes was as follows: (A) (l-2) 1.0 mg/ml cytochrome c, (3-4) 30 U peroxidase, (5-6) 30 U peroxidase and 1.0 mg/ml cytochrome c, (7-8) 900 U peroxidase. All lanes in zymogram B contained 26.5 U guaiacol peroxidase and those of C contained 53.0 U guaiacol peroxidase. Cytochrome c was applied to both gels following the order: (l-2) 0 mg/ml; (3-8) 0.033,0.066,0.100,0.133,0.166, and 0.200 mg/ml, respectively; (9-10) 1.0 mg/ml.
40
JACKSON
AND
lar effects with the most basic isozyme (13,14). Here, evidence has been presented for a similar effect in the migration of three groups of isoperoxidases from L. albus. The use of PVPP and PEG to reduce phenolic contamination in plant extracts has been advised (18). These reagents can act in a competitive manner with proteins for phenolic compounds. Consistent with the removal of tissue-derived phenols from peroxidase samples in this work is the reduction of the observed change in rf with activity of peroxidase applied in samples prepared with PVPP and PEG in the extraction buffer. Cationic phenols have been found in association with cell wall isoperoxidases of L. albus and their removal resulted in a change in the p1 of two basic forms to more acid states (12). Here, Gl isoperoxidases were shown to adopt an apparently more acidic state with the use of PVPP or PEG in the extraction buffer. However, this effect could not be seen in samples prepared without the use of phenolic scavengers and subsequently subject to PVPP, suggesting that removal of tissue-derived contaminants cannot account for the electrophoretic behavior of samples extracted in the presence of PVPP. The increased absorbance ratio 253 nm/206 nm of these extracts suggests that the electrophoretic behavior of these samples might be mediated through additional contamination. Interestingly, cytochrome c was able to increase guaiacol peroxidase activity in zymograms. A similar observation was made for the effects of cationic compounds on the observed activity in zymograms of RNAse (15). In our experiments, the activity promoting effect of cytochrome c appears restricted to samples prepared with PVPP or PEG (compare the staining activities of Fig. 3 and Figs. 4B and 4C). This implies that these samples were applied at a supposedly lower activitylmg isoperoxidase due to the presence of inhibitors. This may explain the resolution of G2 peroxidases with lower (15 units) activity when compared to that of samples prepared without PVPP (40 units). Similarly, the apparent charge difference observed in Gl isoperoxidases might be explained by an association of tissuederived contamination. It is feasable therefore that tissue-derived contamination may also interact with Gl and G2 isoperoxidases. However, this interaction seems to be of a different nature from that described for samples prepared without phenolic scavengers, and so difficult to relate to our main observations with cytochrome c. The Role of Buffer Anions in the Relation of rf to Peroxidase Activity Applied The ability of spermine to increase the mobility of peroxidases and the excellent resolution of samples with cytochrome c led us to investigate the role of buffer anions in the relation of peroxidase rf to the quantity of
RICARDO
activity applied. Variation of the anionic content of electrophoresis buffer was achieved by the use of P-alaninebased solutions (35 InM) adjusted to pH 4.5 with either acetate, succinate, or citrate. The results showed that higher buffer anion content could decrease the rf of Gl and G2, suggesting the variation in Gl and G2 rf might be mediated through an interaction of peroxidases with anions (data not shown). The application of suboptimal levels of cytochrome c (O-O.2 mg/ml) to samples with identical quantities of peroxidases could simulate the relation of rf to the activity of peroxidase applied, showing that Gl and G2 rf were increased with increasing cationic residues, whether contributed by cytochrome c or the samples themselves. These results suggest that cytochrome c might aid the resolution of peroxidases by providing cationic residues with which buffer anions may interact. The subsequently reduced interaction of buffer anions with peroxidase enzymes therefore allows for the full expression of protein charge and sample resolution. Because of the widespread occurrence of proteins that are positively charged at acid pH, the same reasoning can be applied to the increased resolution obtained with greater quantities of peroxidase samples alone. The interaction of spermine and cytochrome c with acetate buffer and peroxidase with acetate, succinate, and citrate buffers suggests a certain nonspecificity of these cation-anion interactions. Similar observations were made in the improved resolution of RNAse by a variety of cationic sources (15). Yet comparable results with other protein samples has rarely been described (16). One possible reason for the detection of this effect here is not that L. albus isoperoxidases represent an example of unusual proteins but rather is the sensitivity of the staining reaction (16). Guaiacol peroxidase activity could easily be detected in zymograms when as little as 10 units were applied, allowing for the application of small sample quantities. Less sensitive staining methods would necessitate the application of greater sample quantities in which case the effect would be reduced. Second, preparation of enzyme samples for zymograms does not usually require other than simple extraction methods. Many enzymes would exist at low specific activity in these preparations, wherein the high total protein content of these samples would minimize the interaction of anions with the enzyme. Peroxidase, however, is abundant in plant cells and simple extraction procedures often produce samples of significant specific activity, allowing the application of low quantities of total protein, and consequently, a greater interaction of anions with peroxidase. The use of cytochrome c has now enabled us to continue with our work with L. albus isoperoxidases of different tissues, each of which have shown improved reso-
CYTOCHROME
c AIDED
ELECTROPHORESIS
lution and quantity independent band rf in response to cytochrome c addition. Due to the widespread occurrence of proteins that are positively charged at acidic pH, the advantages offered by the use of cytochrome c in a cathodal electrophoretic system are not necessarily limited to exceptional cases, but might be applicable to other peroxidases, enzymes prepared to high specific activity, or purified proteins studied in conjunction with sensitive staining methods. ACKNOWLEDGMENTS Financial support for this work was received from JNICT Project 87.206lBIO.
REFERENCES 1. Joerbso, M. (1989) Physiol. Plant. 76, 10-16. 2. Mato, M. C., and Vieitez, A. M. (1986) Physiol. Plant. 66, 491494. 3. Mato, M. C., Rua, M. L., and Ferro, E. (1988) Physiol. Plant. 72, 84-88. 4. Rao, K. V., Suprasanna, P., and Reddy, G. M. (1990) Plant Sci. 66,127-130.
OF PEROXIDASES
41
5. Sanchez, 0. J., Pan, A., Nicolas, G., and Labrador, E. (1989) Physiol. Plant. 76,275-279. 6. Verma, D. P. S., and van Huystee, R. B. (1970) Can. J. Bot. 48, 429-431. 7. Mader, M. (1980) Z. Pflanzenphysiol. 96, 283-296. 8. Fry, S. C. (1982) Biochem. J. 203,493-504. 9. Fry, S. C. (1982) Biochem. J. 204,449-455. 10. Fry, S. C. (1979) Planta 146,343-351. 11. Hunter, R. L., and Markert, C. L. (1957) Science 126,1294-1295. 12. Barcelo, A. R., Munoz, R., and Sabater, F. (1987) Physiol. Plant. 71,448-454. 13. Vessell, E. S. (1962) Nature 195,497-498. 14. Ressler, N., Shulz, J., and Joseph, R. R. (1963) J. Lab. C&n. Med. 62.571-578. 15. Thomas, J. M., and Hodes, M. E. (1981) Ekctrophoresis 2, 104112. 16. Thomas, J. M., and Hodes, M. E. (1981) Experientiu 37,457-459. 17. Goldberg, R. (1980) Physiol. Plant. 50, 261-264. 18. Loomis, W. D., and Battaile, J. (1966) Phytochem. 5, 423-438. 19. Swain, T. (1965) in Plant Biochemistry (Bonner, J., and Varner, J. E., Eds.), p. 552, Academic Press, New York/London. 20. Srivistava, 0. P., and van Huystee, R. (1977) Bot. Gaz. 138,457464.
