Separation of peptides dissolved in a sodium dodecyl sulfate solution by reversed-phase liquid chromatography: Removal of sodium dodecyl sulfate from peptides using an lon-exchange precolumn

186,264-268

ANALYTICALBIOCHEMISTRY

(1990)

Separation of Peptides Dissolved in a Sodium Dodecyl Sulfate Solution by Reversed-Phase Liquid Chromatography: Removal of Sodium Dodecyl Sulfate from Peptides Using an Ion-Exchange Precolumn Hiroshi

Kawasaki

and Koichi

Department of Molecular 3-18-22, Honkomagome,

Received

September

Suzuki

Biology, The Tok.yo Metropolitan Bunk&-ku, Tokyo 113, Japan

Institute

Science,

29,1989

Separation of peptides by reversed-phase liquid chromatography is significantly affected by sodium dodecyl sulfate (SDS) in the sample solution. The strongly acidic group of SDS binds to the reversed-phase column where it serves as an ion exchanger and retards the elution of peptides. By using a DEAE precolumn connected in series to a reversed-phase column, the interference of SDS in the separation of peptides by reversed-phase chromatography can be significantly diminished. This simple method is applicable to the separation of peptide mixtures obtained by digestion of proteins extracted from SDS-polyacrylamide gels. Peptide production with some proteases in the presence of SDS was examined using the present method. Lysylendopeptidase was suitable for digestion in the presence of SDS, but VS protease was not. 0 1990 Academic Press, Inc.

Sodium dodecyl sulfate (SDS)l is one of the most useful detergents for biochemical analysis. SDS dissociates protein complexes and solubilizes membrane proteins. SDS is used as a solvent for the separation of proteins by polyacrylamide gel electrophoresis (PAGE) and gel filtration (1,2) and is a powerful tool for resolving proteins according to their molecular weights. Recent technical advances in microscale structural analysis of proteins, e.g., the gas-phase sequencer, require the devel‘Abbreviations used: SDS, sodium dodecyl sulfate; PAGE, polyacrylamide gel electrophoresis; RPLC, reversed-phase liquid chromatography; TFA, trifluoroacetic acid, CANP, calcium-activated neutral protease or calpain; TPCK, L-l-p-tosylamino-2-phenylethyl chloromethyl ketone; BSA, bovine serum albumin. 264

of Medical

opment of small-scale methods for protein and peptide purification. SDS-PAGE is a convenient method for small-scale protein purification and several procedures for this technique have been reported (3-5). Usually, SDS is removed from proteins before subsequent treatments, e.g., peptide production by protease digestion. Although various procedures have been reported for removal of SDS from proteins (6-ll), these are laborious and cause a significant loss of protein. As some proteases, for example lysylendopeptidase, retain their proteolytic activity and can produce peptides from proteins in the presence of SDS (ll-13), digestion of proteins recovered from polyacrylamide gels or hydrophobic membrane proteins can be carried out in the presence of SDS. Most difficulties arise in the separation of peptides dissolved in an SDS solution. Reversed-phase liquid chromatography (RPLC) is a commonly used technique for separating peptides. Peptide separation by RPLC in the presence of SDS, however, gives poor resolution and the separation of the peptides is dependent on the amount of SDS in the sample (14). Recently, Bosserhoff et al. reported a method for the removal of SDS from peptide mixtures by solvent extraction (14). However, this method seems troublesome and difficult to apply to the large volume samples containing large amounts of SDS that usually result from the extraction of proteins from polyacrylamide gels. In this report, we examine the effects of SDS on the separation of peptides by RPLC and describe a method for removal of SDS from peptides using a DEAE precolumn. This method is simple and applicable to dilute peptide mixtures containing large amounts of SDS. Using this method, we also examined the conditions for protein digestion by lysylendopeptidase in the presence of SDS. 0003-2697/90$3.00 Copyright 0 1990 by Academic Press, Inc. All rights of reproduction in any form reserved.

REMOVAL MATERIALS

AND

High~Perfor~~e

OF SODIUM

DODECYL

SULFATE

FROM

PEPTIDES

265

METHODS

Liquid ~hro~togyaphy

A Hitachi 655-15 HPLC system and a JASCO 880 HPLC system equipped with a high-pressure mixing gradient controller were used. Vydac Cl8 and Hitachi gel No. 3063 were packed in a stainless-steel column (4 X 150 mm). Wakosil5C18 (4 X 150 mm) was obtained from Wako. DEAE-Toyopearl65OS, a hydrophilic anion exchanger, was obtained from Toyo Soda and was used in a stainless-steel column (4 X 50 mm). This precolumn was connected in series at a position between the injection port and the RPLC column either directly or using a six-way valve. Peptides were separated by a linear gradient of acetonitrile in the presence of 0.1% trifluoroacetic acid (TFA) at a flow rate 1 or 0.5 ml/min. Materials Lysylendopeptidase (Achromobacter protease I from Aeh~o~obacte~ lytieus M497-1, see Ref. (15)) was obtained from Wako. V8 protease was obtained from Boehringer-Mannheim. Sperm whale myoglobin was obtained from Sigma. Calcium-activated neutral protease (CANP) was purified from chicken skeletal muscle (16). Carboxymethylated chicken CANP was dissolved in 0.1 M NHIHCOB (2.7 mg/ml), digestedwith TPCK-trypsin (1,000, w/w), and freeze-dried. BSA was reduced and carboxymethylated, dissolved in 100 mM Tris-Cl buffer (pH Q.O), and digested with lysylendopeptidase in the presence or absence of 0.1% SDS at 37°C for 16-26 h. Lysylendopeptidase was used at concentrations greater than 1 pg/ml. Apomyoglobin was produced by treatment of whale myoglobin with hot 50% acetic acid and purified by RPLC. Apomyoglobin was dissolved in 100 mM TrisCl (pH 9.0) (5 pg/50 ~1) in the presence of various concentrations of SDS and then digested with lysylendopeptidase at 37°C for 16 h. Digestion with V8 protease was performed in 100 mM NH,HCO, at 37°C for 17 h. RESULTS

AND

DISCUSSION

Effect of SDS on the Separation of Peptides by RPLC The presence of SDS in the peptide solutions subjected to RPLC decreased the resolution of peptide separation. Small amounts of SDS in the sample solutions brought about broadening of the peaks (Figs. 1A and 1B). An increase in the amount of SDS reduced the separation of peptides significantly and peptide elution was retarded to near the elution position of SDS (Figs. 1C and 1D). Dilution of the samples did not improve the elution patterns significantly (data not shown). The effects of SDS depend primarily on the absolute amount of SDS loaded to the column (Fig. 1).

Retention Time the separation of peptides.

FIG. 1. Effect of SDS on A Hitachi gel No. 3063 column (4 X 150 mm) was eluted at a flow rate of 1 ml/min by a 40-min gradient elution from 0 to 80% acetonitrile in the presence of 0.1% TFA. (A) CANP digest (54 Fg) dissolved in 1 ml of 5 mM TrisCl (pH 7.5). (B) CANP digest (54 cg) dissolved in 100 ~1 of 0.1% SDS (100 pg SDS). (C) CANP digest (54 gg) dissolved in 1 ml of 5 mM TrisCl (pH 7.5) containing 0.1% SDS (1 mg SDS). (D) One milliliter of 5 mM Tris-Cl (pH 7.5) containing 0.1% SDS (1 mg SDS). (E) A DEAE precolumn (4 X 50 mm) was connected in series to the Hiatchi gel No. 3063 column. For the elution conditions, see above. A CANP digest (54 pg) was dissolved in 100 ~1 of 5% SDS (5 mg SDS). The chromatogram is similar to that in A. The broad negative peak of SDS was eluted after peptide separation was completed.

SDS was completely washed out of the column by gradient elution to over 50% acetonitrile, appearing as a broad negative peak on the chromatogram (Fig. 1D). After completion of one cycle of injection and gradient elution, the next separation without SDS exhibited normal separation. However, when gradient elution was stopped

266

KAWASAKI

r f

AND

SUZUKI

D b

d $ s

20min b

Retemtion

b

Time

FIG. 2. Effect of SDS on the digestion of apomyoglobin with lysylendopeptidase. Apomyoglobin (5 pg/50 ~1) was digested with various concentrations of lysylendopeptidase in the presence or absence of SDS. Peptides were separated on a Wakosil5ClS column with a DEAE precolumn, except in A-C. Elution was performed at a flow rate of 0.5 ml/min by a 60-min linear gradient from 0 to 60% acetonitrile in the presence of 0.1% TFA. (A-C) 0% SDS; (D-F) 0.05% SDS; (G-I) 0.1% SDS; (A, D, G) 50 ng lysylendopeptidase (1 pg/ml); (B, E, H) 250 ng (5 ra’ml); 8%F, I) 500 ng (10 t.vz/ml).

at under 40% acetonitrile, the next separation without SDS exhibited poor resolution similar to that observed for samples injected with SDS (data not shown). Resolution recovered only gradually by repeated gradient elutions under 40% acetonitrile. These results indicate the modification of the column by SDS, a compound which has a strong cation-exchange group. The negative charge of SDS could not be suppressed under the conditions commonly used for RPLC (0.1% TFA). The elution of peptides that have positive charges under the conditions used was retarded by the ion-exchange effect of SDS bound to the RPLC column. Recently, Bosserhoff et ab. reported a similar observation on microbore RPLC (14) and suggested that the effect of SDS on peptide retention volumes of peptides was related to the number of Lys and Arg residues in the individual sequences. Other detergents, for example, Triton X-100 or sodium cholate, did not affect the separation of peptides because these detergents have no strongly acidic groups (data not shown). Removal of SDS from Peptide Mixtures a DEAE Precolumn

Using

Removal of SDS from proteins by an ion-exchange resin (Bio-Rad AC 2-X10) under acidic and neutral conditions has been reported (6). This method appears in principle to be applicable to various proteins. We applied

a similar ion-exchange method to the removal of SDS from a peptide mixture and directly combined SDS removal with peptide separation. The method described here uses an ion-exchange precolumn connected in series to an RPLC column. By passing a peptide solution containing SDS through a DEAE ion-exchange column, SDS was trapped on the DEAE column completely while peptides were not trapped under the acidic conditions commonly used in RPLC (0.1% TFA). Simply by insertion of a DEAE precolumn in series, satisfactory peptide separation could be achieved in the presence of SDS (Fig. 1E). The separation of peptides in the presence of SDS using a DEAE precolumn is almost identical to the control separation obtained without a DEAE precolumn in the absence of SDS. In Fig. lE, some peaks could not be detected due to the negative peak which was caused by very large amount of SDS (5 mg). The peptide maps of apomyoglobin in the presence of SDS significantly improved (compare Fig. 2C with Fig. 2F and Fig. 3A with Fig. 3B). The capacity of the DEAE column (4 X 50 mm) was sufficient for SDS loads up to 5 mg. Separation was not affected by the volume of sample injected up to 1 ml. SDS was washed out of the DEAE precolumn by gradient elution of acetonitrile in the presence of TFA after the peptides had been separated and recovered (Fig. 1E). The conventional gradient elution by acetonitrile in the presence of TFA was sufficient to regenerate the DEAE

REMOVAL

OF

SODIUM

DODECYL

SULFATE

FROM

267

PEPTIDES

dient elution should be started to the initial level.

after the baseline returns

Production of Peptides by Lysylendopeptidase and V8 Protease in the Presence of SDS

Retention

Time

FIG. 3. Effect of SDS on the digestion of apomyoglob~n with V8 protease. (A) Apomyoglobin (10 pg/lOO ~1) was digested with V8 protease at a concentration of 20 pg/ml in 100 mM NH,HCO,. Digest (50 ~1) was injected and separated on a Wakosil 5C18 column without a DEAE precolumn. (B) Five microliters of 1% SDS was added to 50 pl of the digest used in A. Separation was performed using a DEAE precolumn and Wakosif 5C18. (C) Apomyoglobin (5 pg/lO ~1) was digested with V8 protease at a concentration of 100 @g/ml in the presence of 0.05% SDS. Separation was performed using a DEAE precolumn and Wakosil SC18 column. For the chromatographic conditions, see Fig. 2.

column, and the performance of the DEAE precolumn was not reduced even after 10-20 cycles of injection and elution with an acetonitrile gradient to over 50% in 0.1% TFA. A precolumn switching method was examined in order to eliminate the negative peak of SDS eluted from the DEAE precolumn; however, switching off the precolumn is not recommended. Although most peptides pass through the DEAE precolumn in 0.1% TFA, some hydrophobic and basic peptides bind to the precolumn in the presence of SDS. These peptides are, however, recovered and eluted from the RPLC column at the same positions as those injected in the absence of SDS. Thus, these peptides may be trapped in SDS micelles bound to the DEAE group. These micelles may be broken by acetonitrile at a concentration lower than that required for elution of the peptides from the RPLC column. When a sample contains salts, e.g., Tris-Cl buffer, a broad negative peak appears slightly after the pass through peak (data not shown). This negative peak was supposed to be retarded anions, e.g., Cl-, eluted from the DEAE precolumn by TFA. It is recommended that gra-

Some proteases, such as trypsin, V8 protease, and lysylendopeptidase, retain their proteolytic activities in the presence of SDS (11-13). However, no information is available as to whether SDS binding to the substrates affects proteolytic cleavages. It has been reported that lysylendopeptidase (Achromobacter protease I) retains full proteolytic activity in the presence of up to 0.1% SDS (12). We analyzed the effects of SDS and enzyme concentration on the production of peptides from proteins by digestion with lysylendopeptidase. In the absence of SDS, the same peptide maps were obtained at lysylendopeptidase concentrations above 5 pg/ml. In the presence of SDS, similar peptide maps were obtained by lysylendopeptidase digestion; however, the increase in peak “a” that accompanied the decrease in peak “b” was slower than that for other peaks (Fig. 2). The N-terminal sequences of the peak a and peak b peptides were YLEFISEAII -- and HKIPIKYLEF - - -, respectively. This indicates that the sequence - IK/Y L - is difficult to cleave in the presence of SDS. The binding of SDS to the hy~ophobi~ site may inhibit cleavage by lysylendopeptidase. The same peptide map as that in the absence of SDS was obtained under conditions of 0.05% SDS and 10 @g/ml lysylendopeptidase (compare Figs. 2B, 2C, and 2F). Clearly, higher protease concentrations are required in 0.1% SDS than in the absence of SDS (Fig. 21). Lysylendopeptidase retains its proteolytic activity in 1% SDS, although the cleavage occurs more slowly (data not shown). It has been reported that V8 protease retains full proteolytic activity in the presence of up to 0.2% SDS (13). However, in contrast to lysylendopeptidase, V8 protease showed weaker activity in the presence of SDS. Figure 3 shows that V8 protease cannot cleave apomyoglobin sufficiently in the presence of 0.05% SDS even at concentrations higher than 100 @g/ml. In the absence of SDS, V8 protease cleaves apomyoglobin sufficiently at a concentration of 20 pg/ml. In conclusion, lysylendopeptidase, but not V8 protease, is suitable for the digestion of proteins in the presence of SDS. The method described here offers the simplest way to obtain peptides from proteins eluted from gels or blotting paper in the presence of SDS (3,17). This method also enables the separation and purification of peptides after protein digestion in a stained gel piece in the presence of SDS. The method for proteolytic extraction of proteins from polyacrylamide gels will be described elsewhere.

268

KAWASAKI

AND

9.

REFERENCES 1. Laemmil, 2. Takagi,

Nature

(London)

j. Chromatogr.

227,680-685.

219,123-12’7.

3. Hunkapiller, M. W., and Lujan, Microcharacterization (Shively, Press, Clifton, NJ.

E. (1986) in Methods of Protein J. E., Ed.), pp. 89-101, Humana

4. Aebersold, R. H., Leavitt, J., Saavedra, R. A., Hood, L. E. and Kent, B. H. (1987) Proe. N&l. Acud. Sci. USA S&6970-6974. 5. Prussak, Biochem. 6. Lenard,

C. E., Almazan,

M.

Biockem.

7. Hebderson, L. E., Groszlan, Biochem. 93.153-157. 8. Kapp,

T., and

Tseng,

B. Y. (1989)

Anal.

178,233-238. J. (1971)

0. H., and

230-235.

Tuszynski,

0. P., and Warren,

I*. (1975)

Anal.

Biochem.

67,55-

65.

U. K. (1970) T. (1981)

SUZUKI

Vinogradov,

Biopkys.

Res. Commun.

S. N. (1978)

45,662-668.

W.

S., and Konigsberg, Anal.

(1979)

Rio&em.

AR.&.

91,

10. Wessel, D., and Flugge, U. I. (1984) Anal. Biochem. 138,141-143. 11. Suzuki, H., and Terada, T. (19881 And. Biockem. 172.259-263. 12. Masaki, T., Fujihashi, T., and Soejima, M. (1984) N&on Nogeikcgaku Kaishi M&865-870. 13. Drapeau, G. R. (1977) in Methods in Enzymology (Him, C. H. W., and Timasheff, S. N., Eds.), Vol. 47, pp. 189-191, Academic Press, San Diego. 14. Bosserhoff, A., Wallach, J., and Frank, R. W. (1989) J. Chromatogr. 437,71-77. 15. Masaki, T., Tanabe, M., Nakamura, K., and Soejima, M. (1981) Biochim. Biophys. Aeta 660,44-50. 16. Ishiura, S., Murofushi, H., Suzuki, K., and Imabori, K. (1978) J. &o&em. 84,225-230. 17. Szewczyk, B., and Summers, F. (1988) Anal. Biochem. 168, 48-

53.