An improved method of sequential alcian blue and ammoniacal silver staining of chondroitin sulfate proteoglycan in polyacrylamide gels

An improved method of sequential alcian blue and ammoniacal silver staining of chondroitin sulfate proteoglycan in polyacrylamide gels

ANALYTICAL BIOCHEMISTRY 167,295-300 (1987) An Improved Method of Sequential Alcian Blue and Ammoniacal Silver Staining of Chondroitin Sulfate Prote...

1MB Sizes 1 Downloads 60 Views

ANALYTICAL

BIOCHEMISTRY

167,295-300 (1987)

An Improved Method of Sequential Alcian Blue and Ammoniacal Silver Staining of Chondroitin Sulfate Proteoglycan in Polyacrylamide Gels’ RICHARD

C. KRUEGER,

JR* AND NANCY

B. SCHWARTZ

Kennedy Center for Mental Retardation, Departments of Pediatrics and Biochemistry and Molecular Biology, The University of Chicago, Chicago, Illinois 60637 Received May 7, 1987 Partially deglycosylated chondroitin sulfate proteogiycan (CSPG) or peptide fragments obtained from CSPG are not readily detectable in gels by staining with Alcian blue 8GX or ammoniacal silver using the technique of Oakley et al. (B. Oakley, D. Kirsh, and N. Morris (1980) Anal. Biochem. 105,361).Sequential staining with both reagents allows visualization of intact CSPG or peptides derived from proteoglycans in polyacrylamide gels at protein concentrations as low as 2 rig/mm’, or glucuronic acid and galactosamine concentrations of 1 rig/mm* or less. This method is significantly more sensitive and has broader applicability than that described by H. Min and M. Cowman (( 1986) Anal. Biochem. 155,275) for staining glycosaminoglycan fragments in polyacrylamide gels. 0 1987 Academic PFS, Inc.

Oakley et al. (1) simplified the technique of Switzer et al. (2) for staining proteins in polyacrylamide gels with ammoniacal silver after treating the gels with glutaraldehyde. It was shown by Dion and Pomenti (3) that the intensity of staining was directly proportional to the concentration of lysine, leading them to postulate a mechanism for glutaraldehyde enhancement involving lysine residues in proteins. Therefore, while this technique is invaluable for most proteins, it will not work for certain proteins which contain little or no lysine. Scott et al. (4-7) described the mechanism for the formation of insoluble complexes of glycosaminoglycans with Alcian blue dye and its dependence on salt concentration. Newton et al. (8) demonstrated the usefulness of Alcian blue for detecting and quantitating glycosaminoglycans separated by electrophoresis on cellulose acetate. Wardi et al.

used Alcian blue to stain cellulose acetate (9) or polyacrylamide gels ( 10) for glycoproteins which have had the vicinal diols of their neutral sugars oxidized by periodic acid and potassium metabisulfite. Min and Cowman (11) demonstrated that oligomeric and polymeric fragments of glycosaminoglycans could be detected to 50 ng in polyacrylamide gels by combined Alcian blue and the silver stain method of Merril et al. (12) (the BioRad stain kit). However, they made no mention of this method’s usefulness for proteoglycans or proteoglycan fragments. Alcian blue may be used to stain proteoglycans or proteoglycan fragments in gels; however, Alcian blue staining is not as sensitive as the silver stain technique is for most proteins, especially partially deglycosylated proteoglycans. The lack of a sensitive method for staining proteoglycans or their peptide fragments is a hindrance to careful peptide analysis and protein sequencing. In this paper, we describe a highly sensitive method for staining partially deglycosylated proteoglycan or proteoglycan-derived pep-

’ This research was supported by USPHS Grants AM-19622, HD-09402, HD-17332, and HD-04583. * Supported by Training Grant on Growth and Development HD-07009. 295

0003-2697/87 $3.00 Copy~&t 0 1987 by Academic Press, Inc. All rights of reproduction in any form reserved.

296

KRUEGER

tide fragments in polyacrylamide using both techniques sequentially.

AND SCHWARTZ

gels by

EXPERIMENTAL

Materials. All reagents were of the highest quality. Only distilled and deionized water was used. Chondroitin sulfate proteoglycan (CSPG)3 was extracted from 14-day chick embryo epiphyses with 4 M GuHCl, 0.1 M Tris-HCl, pH 7.4, containing protease inhibitors ( 100 mM e-amino caproic acid, 10 mM EDTA, 5 mM benzamide HCl, 5 mM PMSF, and 10 mM N-ethyl maleimide) and purified on isopycnic CsCl gradients (13). Chondroitinase ABC and keratanase (Miles) digestions were performed for 4 h at 37°C in 0.05 M Tris-HCl, pH 7.6, with protease inhibitors ( 14). Proteoglycan fragments from the chondroitin sulfate-rich region (CS-rich peptides) were produced by clostripain (Sigma) digestion of intact chondroitinase ABC-treated CSPG and purified on DEAE-cellulose and gel filtration HPLC (unpublished results and ( 15)). METHODS

Analytical procedures. Amino acid analysis was performed by precolumn derivatization with PITC and separation on reversephase HPLC (16). Carbohydrates were analyzed by gas-liquid chromatography of trifluoroacetate derivatives of the O-methyl glycosides ( 17). General methods. Gels were run according to the procedure of Laemmli ( 18). Gels were run with a 3% stacking gel and either a 4% resolving gel in the case of partially deglycosylated CSPG, or a 5 to 18% gradient gel in the case of peptides derived from chondroitinase ABC-treated CSPG. Samples were heated to 100°C for 5 min in 0.0625 M Tris3 Abbreviations used: CSPG, chondroitin sulfate proteoglycan; CS, chondoitin sulfate; PMSF, phenylmethylsulfonyl fluoride; PITC, phenylisothiocyanate; SDS, sodium dodecyl sulfate; PAGE, polyacrylamide gel electrophoresis; BSA, bovine serum albumin.

HCI, pH 6.8, containing 2% SDS (w/v), 10% glycerol (v/v), 1% DTT (w/v), and 0.1% bromphenol blue as a tracking dye. The dimensions of the resolving gel were 14 X 11 cm with a l-cm stacking gel. The gels were run on a Hoeffer vertical gel electrophoresis apparatus at 200 V; the gels take approximately 3 h to run. As silver staining is highly sensitive, it was essential that the equipment, especially the gel plates and staining trays, were scrupulously clean; the gels were handled only with clean latex gloves, because fingerprints stain. Gels were agitated during all staining and washing procedures. Periodic acid oxidation of glycoproteins. This procedure is adapted from Wardi and Michos (10). Gels were fixed for 2 h in 25% ethanol and 10% acetic acid in water, and rinsed with 5% acetic acid in water. The gel was then treated with 1% periodic acid in 5% acetic acid in water at 23°C for 50 min, washed with water for 30 min, treated with 0.5% potassium metabisulfite in water for 30 min, and finally washed with water for 15 min. Alcian blue staining. Gels were fixed for 4 h or overnight (with several changes) in 25% ethanol and 10% acetic acid in water. Extensive washing was necessary to remove the SDS from the gel, which precipitated the stain. If precipitation occurred, the gel could be washed with fixative and restained. Fixed gels were stained for 4 h to overnight in 0.025% Alcian blue 8GX (Sigma) in 25% ethanol and 10% acetic acid. This method also stains negatively charged glycoproteins as well as glycosaminoglycan (7). The gel was destained with multiple changes of 25% ethanol and 10% acetic acid. Inadequate destaining resulted in high background upon subsequent silver staining. Ammoniacal silver staining. This technique is essentially as described by Oakley et al. (1). Gels stained with Alcian blue or fixed overnight in 25% ethanol and 10% acetic acid, were washed for 2 h to overnight in 10% acetic acid to remove the alcohol. The gel was then treated for 30 min in 10% unbuffered glutar-

ALCIAN

BLUE/SILVER

STAINING

OF CHONDROITIN

aldehyde (made by diluting 50% glutaraldehyde (Kodak)), followed by three water washes over 1 f h (45 min was sufficient for low porosity (~5%) gels). The water was removed and the gel was treated with fresh ammoniacal silver for 15 min. This solution was prepared by adding 1.4 ml NI&OH to 21 ml of 0.36% NaOH, followed by 4 ml of 19.4% AgN03 (made by dissolving 20 g of AgN03 in 100 ml water) while stirring, and diluting to the final volume of 100 ml. The ammoniacal silver solution was removed, and the gel was washed twice quickly (30 s each) with water. Freshly prepared developer containing 0.005% citric acid and 0.019% formaldehyde (made by diluting 38% formaldehyde stock) was added, allowing the visibility of the protein bands. The development was stopped when the background began to get dark by a brief washing with water and treatment for 30 min with 10% acetic acid. Gels were stored in water or dried. Gels stained with silver without pretreatment with glutaraldehyde were

A

B

SULFATE

PROTEOGLYCAN

washed after Alcian blue staining with water, as above. RESULTS AND DISCUSSION

Samples used in this study were digested with chondroitinase ABC, an endoglycosidase which removes the bulk of the CS chains attached to the protein core leaving only the CS linkage region with a single disaccharide. It is most likely to the carboxyl function on the glucuronic acid and the sulfate on the galactosamine that Alcian blue binds in the highly polyanionic molecules (5). Alcian blue can also bind to the repeating unit of keratan sulfate chains (galactose-Nacetylglucosamine 6-sulfate); however, these sugars have been largely removed by keratanase treatment as well. Therefore, Alcian blue most likely is binding only to the anionic disaccharide at the end of CS-linkage region and, as shown in Fig. 1, stains chondroitinase ABC, keratanase-treated CSPG

C

* 200kd

9116kd I -92kd

I -66kd

5

4

3

2

154321

297

5432

I

FIG. 1. Five percent SDS-PAGE of chondroitinase ABC- and keratanase-treated CSPG. Section A is stained only with Alcian blue, Section C is stained only with ammoniacal silver, and Section B is stained with Alcian blue and then with ammoniacal silver. Lanes l-4 are serial dilutions of intact CSPG; concentrations are given in Table 1. Lane 5 represents high-molecular-weight protein standards (Bio-Rad), and include myosin (200 kDa), &alactosidase (116.25 kDa), phosphorylase B (92.5 kDa), and BSA (66.2 kDa), each at approximately 167 ng of protein/mm*.

298

KRUEGER

AND SCHWARTZ TABLE 1

CARBOHYDRATE,

AMINO

ACID,

AND

LYSINE

CONTENT

Chondroitinase- and keratanase-treated CSPG Lane 1 2 3 4

aa Wmm*)

LYS (pmol/mm*)

294.0 73.5 18.4 4.6

30.3 7.6 1.89 0.47

OF STAINED

BANDS

Chondrotinase-treated

ChoWmm*)

aa Wmm*)

Lane

77.1 19.2 4.81 1.20

1 2 3 4

125.4 31.35 7.83 1.96

CS-rich peptides

LYS (pmol/mm*) 2.25 0.56 0.14 0.04

Cho(w/mm*) 32.75 8.06 2.02 0.50

Note. Carbohydrate, amino acid, and lysine concentration per square millimeter of gel band (bands are 1 X 6 mm) of intact chondroitinase ABC- and keratanase-treated intact CSPG (see Fig. l), or CS-rich peptides from clostripain digestion of chondroitinase ABC-treated CSPG (see Fig. 2). Carbohydrate content represents total glucuronic acid and galactosamine content. Concentrations are determined on samples before electrophoresis.

very poorly (Fig. 1A). In contrast to the modest staining with Alcian blue or ammoniacal silver, Alcian blue in combination with silver demonstrably enhanced the stainability of the proteoglycan, down to at least 1 ng of glucuronic acid-N-acetylgalactosamine 4-

C

200kd-

-

116kd92kd-

*-,

sulfate per square milliliter (Figs. 1 and 2B). This level is one to two orders of magnitude lower in sensitivity than the method of Min and Cowman (11). In order to examine the usefulness of this staining procedure, attempts were made to

B

A

66kd46kd-

54321

54

3

2

I54321

FIG. 2. SDS-PAGE (5 to 18% gradient) of CS-rich peptides. Section A is stained only with Alcian blue, Section C is stained only with ammoniacal silver, and Section B is stained first with Alcian blue and then with ammoniacal silver. Lanes l-4 are serial dilutions of CS-rich peptides from clostripain digestion of chondroitinase ABC-treated CSPG; concentrations are given in Table 1. Lane 5 represents high-molecular-weight protein standards (Bio-Rad), and includes myosin (200 kDa), @-galactosidase (116.75 kDa), phosphorylase B (92.5 kDa), BSA (66.2 kDa), and ovalabumin (45 kDa), each at approximately 167 ng protein/mm*.

ALCIAN

BLUE/SILVER

STAINING

OF CHONDROITIN

detect peptide fragments derived from proteoglycan core protein. However, there are far fewer lysine residues in the CS-rich peptides (2 residues/lOOO) than in the intact (i.e., not treated with protease) CSPG (11.3 residues/1000) (Table 1). It was not surprising that silver staining only was not successful for proteoglycan-derived peptides, since it was previously demonstrated that the intensity of glutaraldehyde-ammoniacal silver staining is in direct proportion to lysine content. In fact, while the partially deglycosylated CSPG stained poorly with the ammoniacal silver alone (Fig. 1C) the CS-rich peptides (Fig. 2) did not stain at all, even at 1O-fold higher concentration. However, Al-

A

ZOOkd-

6

SULFATE

B

A

200kd

PROTEGGLYCAN

299

C

-

116kd92kd66kd-

45kd-

123123123

FIG.4. SDS-PAGE (5% to 18% gradient) of fetuin. Section A is stained with ammoniacal silver, section C is stained with Alcian blue and then with ammoniacal silver, and section B is treated with periodic acid-potassium metabisullite, and then stained with Alcian blue followed by ammoniacal silver. Lane 1 represents highmolecular-weight protein standards (see Fig. 2). Lanes 2 and 3 represent fetuin at approximately 550 and 35 rig/mm*.

*IS

116kd92kd66kd-

45kd-

GLUT:

I +

2 +

I -

2 -

FIG. 3. SDS-PAGE (5 to 18% gradient) of CS-rich peptides. Section A is stained with Alcian blue and ammoniacal silver, with glutaraldehyde, and section B is stained with Alcian blue and ammoniacal silver, without glutaraldehyde. Lane 1 represents high-molecularweight protein standards (see Fig. 2). Lane 2 represents the CS-rich peptides (50.14 ng amino acid/mm’, 12.88 ng glucuronic acid-galactosamine/mm2, 0.90 pmol lysine/mm*).

cian blue in combination with ammoniacal silver resulted in increased staining of the core protein-derived peptides to the same levels of detectability as partially deglycosylated CSPG. When gels are stained first with Alcian blue and then with ammoniacal silver, but the glutaraldehyde step is omitted, the CSrich peptides still stain strongly with silver, but the protein standards do not (Fig. 3). Thus, unlike the method of Min and Cowman (1 I), it is possible to stain Alcian binding molecules such as CSPG to the exclusion of other proteins. This is extremely useful in the analysis of mixtures of proteoglycans and proteins, especially as in crude preparations. This modification of the procedure also indicates that Alcian blue does not need to bind

300

KRUEGER

AND SCHWARTZ

glutaraldehyde to reduce silver, and is consistent with the fact that it has no primary amino groups capable of binding. Furthermore, this technique may be adapted to stain polyacrylamide gels for detection of glycoproteins which have had the vicinal diols of their neutral sugars oxidized by periodic acid and potassium metabisulfite. Although fetuin (Sigma) stained with ammoniacal silver alone, and this staining was not enhanced by prior treatment with Alcian blue, the staining was greatly enhanced by first treating with periodic acid/potassium metabisulfite, then staining with Alcian blue, followed by ammoniacal silver (Fig. 4). This technique should greatly facilitate the study of proteoglycans and their peptides, and glycoproteins and their peptides, by providing a tool for detecting small quantities of these poorly detectable molecules in polyacrylamide gels.

teins and proteoglycans. Lastly, a feasible mechanism by which the selective staining is accomplished is presented. REFERENCES 1. Oakley, B., Kirsh, D., and Morris, N. (1980) Anal. Biochem. 105,361-363. Switzer, R., Merril, C., and Shifiin, S. (1979) Anal. Biochem. 98,232-237. Dion, A., and Pomenti, A. (1983) Anal. Biochem. 129,490-496.

5.

6. 7. 8. 9.

SUMMARY

A method for detecting glycosaminoglycans based on sequential treatment with Alcian blue and silver staining has been reported (11). A similar method was independently developed which is excellent for staining proteoglycans, proteoglycan-derived peptides, and neutral glycoproteins whose vicinal diols have been cleaved with periodic acid/sodium metabisulfite. Furthermore, with the modifications, which include using ammoniacal silver instead of potassium dichromate oxidation followed by silver nitrate ( 12), the sensitivity of our method is increased by one to two orders of magnitude. It is also possible to omit the glutaraldehyde fixation step and selectively stain proteoglycans over protein, which is an extremely useful technique in analyzing mixtures of pro-

10.

Scott, J., Quintarelli, G., and Dellovo, M. (1964) Histochemie 4, 73-85. Quintarelli, G., Scott, J., and Dellovo, M. (I 964) Histochemie 4, 89-98. Quintarelli, G., Scott, J., and Dellovo, M. (1964) Histochemie 4, 99- 112. Scott, J., and Doring, J. (1965) Histochemie 5, 221-233. Newton, D., Scott, J., and Whiteman, P. (1974) Anal. Biochem. 62,268-273. Wardi, A., and Allen, W. (I 972) Anal. Biochem. 48, 621-623. Wardi, A., and Michos, G. (1972) Anal. Biochem. 49,607-609.

Il. Mitt, H., and Cowman, M. (1986) Anal. Biochem. 155,275-285.

Merril, C. R., Goldman, D., Sedman, S. A., and Ebert, M. H. (1981) Science 211, 1437-1438. 13. Hascall, V., and Kimura, J. (1982) in Methods of Enzymology (Cunningham, L. W., and Frederickson, D. W., Eds.), Vol. 82, pp. 769-800, Academic Press. 14. Oike, Y., Kimata, K., Shinomura, T., Nakazawa, K., and Suzuki, S. (1980) Biochem. J. 191, 12.

193-207.

Krueger, R., and Schwartz, N. B. (1987) in Proceedings of the IXth International Symposium on Glycoconjugates, B54. 16. Heinrikson, R., and Meridith, S. (1984) Anal. Biothem. 36,65-74. 17. Zanetta, J., Breckenridge, W., and Vincendon, G. (1972) J. Chromatogr. 16,291-304. 18. Laemmli, U. K. (1980) Nature (London) 227, 15.

680-685.