ANALYTICALBIOCHEMISTRY 187,147-150 (1990)

Detection of Glycoproteins Separated by Nondenaturing Polyacrylamide Gel Electrophoresis Using the Periodic Acid-Schiff Stain’

Kinchel C. Doerner and Bryan A. White2 Department of Animal Sciences, University of Illinois, 1207 West Gregory Drive, Urbana, Illinois, 61801

Received December 21,1989

We observed that published methods for staining gly- coproteins in sodium dodecyl sulfate-polyacrylamide gels did not stain glycoproteins separated by nondena- turing polyacrylamide gel electrophoresis. Therefore, the periodic acid-Schiff stain for glycoprotein was adapted for use with proteins analyzed by nondenatur- ing polyacrylamide gel electrophoresis. Following non- denaturing polyacrylamide gel electrophoresis pro- teins were denatured in situ by incubation with aqueous 2% sodium dodecyl sulfate, 5% S-mercaptoeth- anol, 40% ethanol, and 5% acetic acid at 90°C for 15 min followed by periodic acid-Schiff staining. This modified procedure will detect at least 0.2 pg protein- associated carbohydrate. Omission of the periodic acid treatment from the protocol was used as a control to de- tect nonspecific staining of some proteins. This modified procedure was validated using both purified glyco- proteins and extracellular culture fluid containing carbohydrate-associated proteins of Ruminococcus flavefaciens. 0 1990 Academic Press, Inc.

The periodic acid-Schiff stain will stain carbohy- drate-containing proteins which have been subjected to sodium dodecyl sulfate-polyacrylamide gel electropho- resis (SDS-PAGE)3 (l-3), but we were unsuccessful in applying this technique using nondenaturing polyacryl- amide gel electrophoresis (ND-PAGE). In studies with Ruminococcus fluuefaciens, an indigenous ruminal bacte- rium, cellulase-enriched preparations showed numerous

i This research was supported by the United States Department of Agriculture Grants 35-0376 and 35-0505 and by the Agriculture Exper- imental Station of the University of Illinois.

’ To whom correspondence should be addressed. 3 Abbreviations used: SDS-PAGE, sodium dodecyl sulfate-poly-

acrylamide gel electrophoresis; ND-PAGE, nondenaturing polyacryl- amide gel electrophoresis; PAS, periodic acid-Schiff.

0003~2697/90 $3.00 Copyright 0 1990 by Academic Press, Inc. All rights of reproduction in any form reserved.

proteins associated with carbohydrates when SDS- PAGE was used, but failed to show any potential glyco- proteins when proteins separated by ND-PAGE were subjected to the periodic acid-Schiff (PAS) treatment (K. C. Doerner M.S. thesis, University of Illinois, Ur- bana, 1989). Therefore we developed a modification of the PAS protocol that could be used to detect carbohy- drate-associated proteins separated by ND-PAGE. This manuscript describes that adaptation of the PAS method.

MATERIALS AND METHODS

Nondenaturing Polyacrylamide Gel Electrophoresis

ND-PAGE was performed with a discontinuous buffer system. The resolving gel was composed of 0.1 M

tris[hydroxymethyl]aminomethane hydrochloride (pH 8.0) mixed with 30%T and 2.7%C acrylamide stock solu- tion to yield a 16% acrylamide solution. The stacking gel was composed of 0.1 M tris[hydroxymethyl]aminometh- ane hydrochloride (pH 7.0) mixed with the a&amide stock solution to yield a 4% acrylamide gel. Monomer solutions were degassed under vacuum and polymeriza- tion was initiated by the addition of ammonium persul- fate (0.5% final concentration) and 0.05 or 0.1% final concentration of N,N,N’,N’-tetramethylethylenedi- amine for the resolving and stacking gels, respectively. These were then cast into 10.2 cm X 5 cm X 0.75 mm slab gels. The upper buffer reservoir was composed of 0.043 M tris[hydroxymethyl]aminomethane hydrochlo- ride and 0.064 M glycine (4) and the lower buffer reser- voir was composed of 0.1 M tris[hydroxymethyl]amino- methane hydrochloride (pH 8.0). Immediately prior to loading, a l/10 volume of 50% (w/v) sucrose and 0.2% (w/v) bromophenol blue solution was added to protein samples previously prepared in 0.1 M tris[hydroxymeth- yllaminomethane hydrochloride (pH 7.0) and electro- phoresed at 200 constant V until the dye front was near

147

148 DOERNER AND WHITE

FIG. 1. (A) Periodic acid-Schiff- and (B) Coomassie brilliant blue-stained gels of bovine a-acid glycoprotein (lanes l-6), bovine a-acid glycoprotein with nonglycosylated proteins (lane 7), and nonglycosylated proteins (lane 8) separated by nondenaturing polyacrylamide gel electrophoresis. Arrows indicate primary location of nonglycosylated proteins. Lanes 1,23 pg; 2, 11 pg; 3,5.6 pg; 4, 2.8 pg; 5, 1.4 pg; 6,0.7 pg; 7, 4.5 fig bovine a-acid glycoprotein with nonglycosylated proteins [5 fig apoferritin (a), 10 pg fl-amylase (b), 5 pg carbonic anhydrase (c), 5 pg bovine serum albumin (d)]; 8, nonglycosylated proteins only.

the bottom of the gel. Nonglycosylated proteins were 5 pg each of carbonic anhydrase, bovine serum albumin, and apoferritin and 10 pg of /3-amylase per gel lane.

Following electrophoresis gels were immediately placed in a denaturing fixative of 40% (v/v) ethanol, 5% (v/v) acetic acid, 2% (w/v) sodium dodecyl sulfate, and 5% (v/v) 2-mercaptoethanol. Glass vessels with fitted caps containing the denaturing fixative were heated in a 90°C water bath for no less than 15 min. ND-PAGE gels were then incubated in the hot solution for 15 min, after which the glass vessels were removed from the water bath and allowed to cool to room temperature. Due to the volatility and toxic nature of 2-mercaptoethanol these

operations were performed in a vacuum hood. To remove excess sodium dodecyl sulfate and 2-mercaptoethanol, gels were transferred into 40% (v/v) ethanol and 5% (v/ v) acetic acid and washed no less than three times for 20 min each.

Schiff Reagent Preparation

The Schiff reagent was prepared by modifications of earlier protocols (1,5). Five grams of basic fuchsin (para- rosaline) was added to 1 liter of water and heated in a microwave oven. Then 25 ml of 2 N HCl and 17 g of so- dium metabisulfite were added and stirred overnight at

FIG. 2. (A) Periodic acid-Schiff and (B) Coomassie brilliant blue-stained gels of chicken ovalbumin (lanes l-5), chicken ovalbumin with nonglycosylated proteins (lane 6), and nonglycosylated proteins (lane 7) separated by nondenaturing polyacrylamide gel electrophoresis. Lanes 1, 90 pg; 2, 45 fig; 3, 23 pg; 4, 11 pg; 5, 5.6 pg; 6, 45 pg chicken ovalbumin with nonglycosylated proteins; 7, nonglycosylated proteins only. Nongiycosylated protein mixture composition and arrows are the same as those in Fig. 1.

ELECTROPHORETIC DETECTION OF NONDENATURED GLYCOPROTEINS 149

4°C in the dark followed by the addition of 25 g of acti- vated charcoal. After 1 h the charcoal and fines were pel- leted by centrifugation (11,3OOg, 30 min, 4°C) and dis- carded. The supernatant was vacuum-filtered through glass wool and lo-15 ml of 2 N HCl was added to the filtrate until a small drop of reagent would dry to a me- tallic yellow by flaming on a glass slide. The Schiff re- agent was stored in the dark at 4°C and used within 2 weeks.

Periodic Acid-Schiff Staining

PAS staining was performed as modified by Dubray and Bezard (6). Briefly, gels were soaked in 7.5% (v/v) acetic acid for 30 min and then with 0.2% (w/v) periodic acid for 1 h at 4°C. The periodic acid solution was re- moved and the Schiff reagent added and incubated for 1 h at 4°C. Reddish-pink bands of stained glycoprotein would then be visible. The PAS reagent was removed and the gels were soaked in 7.5% acetic acid for 1 h and subsequently stored in water. For detection of nonspe- cific staining, periodic acid solution was omitted from the protocol and replaced with water (7).

Coomassie Brilliant Blue Staining

ND-PAGE gels were stained for protein by soaking in 0.1% (w/v) Coomassie brilliant blue R-250 in 40% (v/v) methanol and 10% (v/v) acetic acid. Destaining was achieved by soaking gels in this fixative omitting the stain.

RESULTS AND DISCUSSION

Sodium dodecyl sulfate, 2-mercaptoethanol, and sam- ple heating are used to denature proteins prior to SDS- PAGE, but are not used in ND-PAGE techniques. Since the published PAS protocols utilize SDS-PAGE and are ineffective with ND-PAGE we hypothesized that pro- tein denaturation played an essential role in PAS stain- ing. Therefore, following electrophoresis, ND-PAGE gels were heated in the presence of sodium dodecyl sul- fate and 2-mercaptoethanol to denature proteins in situ. Gels were washed to remove excess reagents and then PAS stained. This modified protocol was found to detect glycoproteins separated by ND-PAGE gels.

The specificity of this technique is demonstrated in Figs. 1 and 2. Figure 1 shows that bovine a-acid glyco- protein (30% by weight carbohydrate (8)) is readily visu- alized by this procedure. A mixture of proteins, none containing carbohydrate, did not stain positive (Fig. lA, lanes 7 and 8). This procedure visualized 0.7 pg bovine a-acid glycoprotein containing 0.2 gg carbohydrate. Pro- tein loads less than 0.7 pg were not tested.

Figure 2A demonstrates the specificity of this tech- nique for another glycoprotein, chicken ovalbumin. This protein contains mannose residues at approximately 4%

FIG. 3. Periodic acid-Schiff stain of Ruminococcus flavefaciens ex- tracellular protein separated by nondenaturing polyacrylamide gel electrophoresis using the described modifications. Horizontal arrows denote bands of protein that stain positive for glycoprotein. Vertical arrow indicates direction of electrophoresis.

by weight (according to the manufacturer). Due to lesser amounts of carbohydrate associated with chicken albu- min, compared to bovine a-acid glycoprotein, more sam- ple is required to detect the carbohydrate. Figure 2A shows sensitivity to at least 5.6 pg protein which is equivalent to approximately 0.2 pg carbohydrate. There- fore, equivalent loads of carbohydrate from two glyco- protein sources result in the same intensity of staining. This demonstrates reproducibility and sensitivity to at least 0.2 pg carbohydrate for this procedure.

Figures 1B and 2B are duplicate gels of the PAS- stained gels, except that, immediately after electropho- resis, these were stained with Coomassie brilliant blue. This shows that the modified PAS stain is specific for carbohydrate-containing proteins.

Periodic acid-Schiff staining of proteins separated by ND-PAGE was validated by a variety of controls. During heat treatment it was shown that sodium dodecyl sulfate and 2-mercaptoethanol caused no significant protein loss from the gel. This was assessed in two ways. First, gels that were protein stained immediately after heat-

150 DOERNER AND WHITE

denaturing treatment showed identical patterns of pro- tein bands with the same degree of intensity as those gels stained immediately after electrophoresis. Second, gels were stained with Coomassie brilliant blue following PAS staining (without periodic acid pretreatment) and again showed that the proteins were present in the same amount.

Omitting the periodic acid treatment preceding the in- cubation with the Schiff reagent precluded staining of known glycoproteins (data not shown). This confirms the specificity of the techniques as the periodic acid treatment is mandatory for PAS staining. This control can also be used to detect nonspecific dye binding. Fig- ures 1A and 2A show that apoferritin stains, although it contains no carbohydrate. Identical gels containing apoferritin, one treated and the other not treated with periodic acid prior to the Schiff reagent, showed the same degree of staining. Furthermore, decreasing amounts of this protein showed decreasing amounts of staining when the periodic acid treatment was omitted. These observations demonstrate the necessity of omit- ting periodic acid from the protocol to ensure accurate determination of carbohydrate-containing proteins.

The validity of this technique was also tested using R. fEauefa&ns extracellular culture fluid (9). This organism has been shown to produce various carbohydrate-con- taining components using SDS-PAGE followed by PAS staining (K. C. Doerner M.S. thesis, University of Illi- nois, Urbana, 1989). Figure 3 shows that only three

bands were detected in culture fluid when the modified technique was used. It is therefore likely multiple carbo- hydrate-associated proteins compose the three bands. These data indicate that the PAS stain will detect carbo- hydrate-associated proteins separated by ND-PAGE us- ing the modifications described.

ACKNOWLEDGMENT

Thanks are extended to Myra L. Coggan for assistance with the fig- ure preparation.

REFERENCES

1.

2.

3.

4.

5.

6.

7.

8.

9.

Segrest, J. P., and Jackson, R. L. (1972) in Methods in Enzymol- ogy (Ginsburg, V., Ed.), Vol. 28, pp. 54-63, Academic Press, San Diego.

Fairbanks, G., Steck, T. L., and Wallach, D. F. H. (1971) Biochem- istry 10,2606-2617.

Zacharius, R. M., Zell, T. E., Morrison, J. H., and Woodlock, J. J. (1969) Anal. Biochem. 30,148-152. Calza, R. E., Irwin, D. C., and Wilson, D. B. (1985) Biochemistry 24,7797-7804.

Andrews, A. T. (1986) Electrophoresis: Theory, Techniques, and Biochemical and Clinical Applications, 2nd ed., pp. 37-38, Oxford Univ. Press, New York.

Dubray, G., and Bezard, G. (1982) And. Biochem. 119,325329.

Matthieu, J.-M., and Quarles, R. H. (1973) And. Biochem. 56, 313-316.

Hao, Y., and Wickerhauser, M. (1973) Biochim. Biophys. Acta 322,99-108.

Gardner, R. M., Doerner, K. C., and White, B. A. (1987) J. Bacte- rid. 169,4581-4588.