Techniques in Protein Chemistry III

Techniques in Protein Chemistry III compiles papers presented at the Fifth Protein Society Symposium in Baltimore on June 22-26, 1991. This book discusses the protein and peptide recovery from PVDF membranes; high-sensitivity peptide mapping utilizing reversed-phase microbore and microcolumn liquid chromatography; and capillary electrophoresis for preparation of peptides and direct determination of amino acids. The TFMSA/TFA cleavage in t-Boc peptide synthesis; applications of automatic PTC amino acid analysis; and identification of O-glycosylation sites with a gas phase sequencer are also elaborated. This text likewise covers the conformational stability of the molten globule of cytochrome c and role of aqueous solvation in protein folding. This publication is useful to students and researchers interested in methods and research approaches on protein chemistry.

• Language: English
• Publisher: Academic Press
• Release date: Jun 28, 2014
• ISBN: 9781483259468

Book preview

Inc.

Section I

Protein Microsequence Methods

Outline

Chapter 1: Protein and Peptide Recovery from PVDF Membranes

Chapter 2: Automated C-Terminal Sequencing of Peptides

Chapter 3: Elution and Internal Amino Acid Sequencing of PVDF-Blotted Proteins

Chapter 4: Evaluation of Protein Sequencing Core Facilities: Design, Characterization, and Results from a Test Sample (ABRF-91SEQ)

Chapter 5: OPTIMIZATION OF SOLID PHASE PEPTIDE SEQUENCING

Chapter 6: Evaluation of the Blott Cartridge for Enhanced Gas Phase Sequencing at Maximum Sensitivity

Chapter 7: Reusing PVDF Electroblotted Protein Samples After N-Terminal Sequencing To Obtain Unique Internal Amino Acid Sequence

Chapter 8: CHROMATOGRAPHIC CARBON AS AN INERT SAMPLE ADSORBENT FOR PROTEIN SEQUENCING

Chapter 9: Nα-ACYLAMINOACYL-PEPTIDE HYDROLASE: SPECIFICITY AND USE TO UNBLOCK N-ACETYLATED PROTEINS

Protein and Peptide Recovery from PVDF Membranes

Susan Wong, Allan Padua and William J. Henzel,     Department of Protein Chemistry, Genentech, Inc. South San Francisco, CA 94080

Publisher Summary

Polyvinylidene difluoride (PVDF) is the most frequently utilized support for the sequence analysis of electroblotted proteins. This chapter describes a simple method for recovering intact proteins from PVDF membranes. A limitation of this technique is that N-terminal blocked proteins must be chemically cleaved or proteolytically digested to generate internal peptide fragments that can then be sequenced. The chapter describes in situ digestions on PVDF and nitrocellulose membranes. It presents the development of methods that directly extract protein from PVDF. These are based on solutions containing detergents that can limit subsequent proteolytic digestions or interfere with HPLC separations. A simple extraction procedure using the solvent dimethyl sulfoxide (DMSO) was developed that allows the recovery of intact proteins from PVDF membranes. The DMSO can be easily removed, allowing the extract to be analyzed by protein sequencing, amino acid analysis, or by other biochemical analysis. The extracted protein can be cleaved or digested and the resulting fragments can either be separated by HPLC or by tricine SDS-gels followed by a second round of electroblotting. An additional advantage of this technique is that the protein may be directly digested in the extraction solvent after addition of an appropriate buffer.

I Introduction

Electroblotting proteins from gels onto solid supports is a widely utilized technique for the microisolation of proteins. Several electroblotting membranes have been developed that allow direct sequence analysis (1–3). Polyvinylidene difluoride (PVDF) is the most frequently utilized support for sequence analysis of electroblotted proteins (4). A limitation of this technique is that N-terminal blocked proteins must be chemically cleaved or proteolytically digested to generate internal peptide fragments which can then be sequenced. In situ digestions on PVDF (5–8) and nitrocellulose membranes (9) have been described; however, these methods only result in partial digestion of the membrane-bound proteins. Methods that directly extract protein from PVDF have been developed (10,11), but are based on solutions containing detergents which can limit subsequent proteolytic digestions or interfere with HPLC separations. We have developed a simple extraction procedure using the solvent dimethyl sulfoxide (DMSO) that allows recovery of intact proteins from PVDF membranes. The DMSO can be easily removed, allowing the extract to be analyzed by protein sequencing, amino acid analysis, or by other biochemical analysis.

The extracted protein can be cleaved or digested and the resulting fragments can either be separated by HPLC or by tricine SDS-gels (12) followed by a second round of electroblotting. An additional advantage of this technique is that the protein may be directly digested in the extraction solvent after addition of an appropriate buffer.

II Materials and Methods

Materials

Proteins were purchased from Sigma, with the exception of human growth hormone (hGH) which was obtained from Genentech, Inc. Lys-C was from Wako chemicals, and cyanogen bromide was from Pierce. Dimethyl sulfoxide (spectrometric grade) was from Aldrich. PVDF (Immobilon-P) membranes were obtained from Millipore and ProBlott membranes were from Applied Biosystems.

Electroblotting and elution

SDS-PAGE was done as described by Laemmli (13) using a BioRad minigel apparatus. Tricine minigels were obtained from Novex. Electroblotting was done in a BioRad transfer tank using 10 mM 3-[cyclohexylamino]-1-propanesulfonic acid, 20% methanol, pH 11.0 for 1 h at 250 mA (4). Coomassie stained bands from electroblotted proteins were extracted with 200 μl DMSO in a micro Eppendorf tube on a VWR vortexer-2 using a medium setting for 2 h. The DMSO was removed and another 200 μl of DMSO was added. The extraction was continued for an additional 2 h. The extracts were then combined and used for further experiments. Extractions were also performed using a single addition of 200 μl of DMSO, vortexed for 17h.

Fragmentation

Extracted proteins were dried in a Speed-Vac to remove DMSO and were cleaved with cyanogen bromide in 0.1 N HCl at 45°C for 3 h (14). Blots were prewetted with 5 μl of methanol prior to the addition of 0.1 N HCl and cyanogen bromide.

Protease digestions were performed on electroblot extracts after partial removal of DMSO by Speed-Vac evaporation. Ammonium bicarbonate (0.1 M) was added to the residual DMSO (20-50 μl) resulting in a final solution of 200-500 μl of 10% DMSO. Digestion with Lys-C was carried out at 37°C for 17 h. with an enzyme to substrate ratio of 1:20 (w/w).

Electroblots were first treated with a modified methanol-chloroform precipitation (15) prior to DMSO elution in order to remove Coomassie blue and HPLC artifact peaks. Blots (wet) were placed in 100 μl of water in an Eppendorf tube and 400 μl of methanol was added. The solution was vortexed for 2 min. and 100 μl of chloroform was added. The solution was vortexed again for 2 min. and the liquid was removed. The membrane was then extracted with DMSO. This method effectively removed the Coomassie blue and minimized artifact peaks in the HPLC chromatograms.

Peptide separation

Peptides were purified by reversed-phase HPLC using a Synchropak 4000 Å C4 column (2 × 100 mm) on a Hewlett-Packard 1090 M liquid Chromatograph equipped with diode array detector. Peaks were detected at 214 nm after elution with a linear gradient of 0-70% acetonitrile in 0.1% trifluoroacetic acid at a flow rate of 200 μl per minute.

Protein sequencing

Automated Edman degradation was performed with Applied Biosystems models 470A and 473A protein sequencers equipped with on-line PTH analyzers. Sequence interpretation was performed on a VAX 8750 as described (16).

Amino acid analysis

Extracted proteins were dried in a Speed-Vac to completely remove DMSO. Residual DMSO can result in the oxidation of some amino acids including methionine and cysteine. Blots were prewetted with 5 μl of methanol prior to hydrolysis. Proteins were hydrolyzed in the gas phase for 24 h with 300 μl 6N HCl containing 0.1% phenol at 110°C under vacuum using a Millipore Picotag workstation. Hydrolyzed blots were extracted twice with 100 μl of 6 N HCl. The extracts were dried and reconstituted with Beckman sample buffer and hydrolysates were analyzed on a Beckman Model 6300 amino acid analyzer.

III Results and Discussion

In an effort to optimize recovery, a number of proteins were transferred to Immobilon-P or ProBlott membranes, stained with Coomassie and then extracted for various times with DMSO. The optimal extraction time was found to be 4 hours using 2 consecutive 2 hour extractions (of 200 μl DMSO each) or a single extraction of 200 μl for 17 h. Both extraction methods resulted in nearly identical protein recovery. Recovery values obtained from DMSO extractions of proteins on ProBlott and Immobilon-P extractions are shown in Table I. These values were obtained by using the 4 hour extraction procedure. The extraction efficiency from Immobilon-P was better than from Problott membranes. Other extraction solvents were examined including 50% mixtures of DMSO and methanol, and DMSO and water (data not shown). These solvents resulted in lower efficiency of extraction of protein from PVDF membranes compared to using DMSO alone.

TABLE I

Recovery data of proteins eluted from PVDF electroblots. Proteins were separated on 12% SDS gel, electroblotted onto Immobilon-P or ProBlott and stained with Coomassie brilliant blue. Bands were excised and extracted with DMSO, transferred to hydrolysis tubes, and analyzed by amino acid analysis. Extracted blots were hydrolyzed to determine residual protein. Protein on the blot was calculated by adding the amount remaining on the blot after extraction to the extracted amount. All numbers represent the average of triplicate experiments.

Extraction efficiencies can be determined by analyzing an aliquot of the extract by amino acid analysis. Proteins which are found to be intractable to DMSO extractions can be cleaved with cyanogen bromide and the resulting peptides can then be efficiently extracted. Further purification of these peptides can be accomplished by SDS-PAGE or reversed-phase HPLC. Extraction of cyanogen bromide peptides from the membrane may be a preferred alternative to directly extracting intact proteins larger than 70 kDa. No indication of selective retention of hydrophobic peptides was seen when membranes were sequenced after DMSO extractions. The extraction efficiencies of cyanogen bromide cleaved proteins are shown in Table II. Peptide recovery from PVDF and ProBlott membranes were similar.

TABLE II

Recovery of peptides eluted from PVDF electroblots following CNBr cleavage. Proteins bound to PVDF membranes were cleaved with CNBr in 0.1 N HCl at 45°C, dried, then extracted with DMSO. The DMSO extracts were transferred to hydrolysis tubes and analyzed by amino acid analysis. All numbers represent the average of duplicate experiments.

The application of peptide mapping on proteins eluted from electroblots is shown in Figure 1. Human growth hormone was eluted from a ProBlott membrane. An aliquot was removed (25%) and analyzed by amino acid analysis, which indicated a total of 72 pmol was originally present in the extract. The remaining 54 pmol was cleaved with cyanogen bromide, dried in a Speed-Vac, electrophoresed on a 16% tricine SDS gel and electroblotted on a ProBlott membrane. The excised bands were sequenced by automated Edman degradation and revealed initial yields ranging from 2-16 pmol (Fig. 1).

Figure 1 Coomassie blue stained ProBlott membrane of CNBr cleaved hGH after DMSO elution from a ProBlott membrane. (Lane A) Recombinant hGH was electrophoresed on a 12% SDS gel, electroblotted onto ProBlott, stained with Coomassie blue, the band excised and extracted with DMSO. The extracted protein (54 pmol) was cleaved with cyanogen bromide, and the peptides separated on a 16% tricine SDS gel. Bands 1-3 were sequenced (A); (B) control, hGH (100 pmol) + CNBr; (C) control, hGH (100 pmol); (D) LMW marker proteins (BioRad); (E) LMW marker (Diversified Biotech).

Proteins extracted from PVDF are difficult to resolubilize in aqueous buffers if DMSO is completely removed. To circumvent this problem proteins were digested with Lys-C in the presence of DMSO. Figure 2 shows the results of digesting hGH in the presence of increasing amounts of DMSO. Addition of DMSO to the digestion had little effect on proteolysis as shown in lanes C-F.

Figure 2 SDS-PAGE tricine gel (16%) of a Lys-C digest of hGH. Lane A is intact hGH, Lane B is a Lys-C digest of hGH, Lanes C-F are Lys-C digests of hGH with increasing amounts of DMSO added from 10 to 40% in 10% increments, Lane G is a Lys-C digest of hGH with 0.1% SDS. Lane H-K are the same conditions as in C-F except 0.1% SDS was included in the digestion buffer. All digestion were performed in 0.1M NH4HCO3 containing 0.01M DTT.

The addition of detergents such as SDS can be used to solubilize hydrophobic proteins and denatured proteins. The effects of increasing amounts of DMSO on the digestion of hGH by Lys-C in 0.1% SDS, 0.1 M ammonium bicarbonate is shown in lanes H-K in Fig. 2. DMSO significantly enhanced the activity of Lys-C in the presence of SDS (lanes H-K) resulting in greater proteolysis than seen with SDS alone (lane G) or in the absence of DMSO and SDS (lane B). Higher amounts of DMSO (30-40%) (lanes J-K) result in a similar cleavage pattern as observed when hGH is digested in the absence of SDS (lanes B-F). A similar effect with dimethylformamide was recently reported (17).

To demonstrate the utility of this procedure, extracted hGH was digested with Lys-C in 0.1 M ammonium bicarbonate in the presence of DMSO and SDS. The Lys-C digest was separated on a 16% tricine gel and electroblotted onto a ProBlott membrane. The sequence obtained for two of the resultant peptide bands is shown in Table III.

TABLE III

Sequence analysis obtained from an electroblot of a Lys-C digest of hGH eluted from a ProBlott membrane.

Protein digests often generate peptides smaller than 4 kDa and contain peptides having similar molecular weights. These peptides may be better resolved with higher recoveries when separated by reversed phase HPLC. This approach is demonstrated by a Lys-C digestion of DMSO-extracted hGH shown in Figure 3. The control Lys-C digest of hGH (Fig. 3B) appears similar to the digest of extracted hGH (Fig. 3A) except the earlier eluting peaks are less predominate. This is suggestive of more extensive cleavage for the digest of extracted hGH, which may be a result of protein denaturation resulting from the SDS gel electrophoresis. Despite staining the blot initially with Coomassie, no major artifacts are present. The Coomassie artifact peaks are minimized by utilizing a chloroform-methanol precipitation (15) directly on electroblots before extraction. Amino acid analysis of the solvents used for Coomassie extraction on eletroblot membranes containing hGH, lysozyme and BSA resulted in less than 1% protein loss from the membranes. This method can also be used on the extracted protein prior to digestion. A disadvantage of performing precipitation after extraction is that significant losses may occur for proteins smaller than 20 kDa.

Figure 3 HPLC separation of Lys-C digest of hGH. (A) hGH eluted from a ProBlott membrane digested with Lys-C in 0.1 M NH4HCO3 containing 0.01M DTT, 0.1% SDS and 10% DMSO. Peptides were separated on a C4 column using a linear gradient of 0-70% acetonitrile in 0.1% TFA. (B) Control hGH digested with Lys-C using the same condition as in (A).

IV Conclusions

This report describes a simple method for recovering intact proteins from PVDF membranes. Peptides and proteins can be recovered from both Immobilon-P and ProBlott membranes by DMSO extraction. Proteins obtained by DMSO extraction of PVDF membranes can be digested by proteases in solution, allowing more extensive cleavage than obtainable by in situ digestion. Extension of this work may allow additional methods of analysis of intact protein recovered from electroblots.

Acknowledgements

We thank Chris Grimley for providing sequencing analysis.

References

1. Vandekerchove, J., Bauw, G., Puype, M., Van Damme, J., Van Montagu, M. Eur. J. Biochem. 1985; 152(9)

2. Abersold, R.H., Teplow, D.B., Hood, L.E., Kent, S.B.H. J. Biol. Chem. 1986; 261(4229)

3. Eckerskorn, C., Mews, W., Goretzki, H., Lottspeich, F. Eur. J. Biochem. 1988; 176(509)

4. Matsudaira, P. J. Biol. Chem. 1987; 262(10035)

5. Bauw, G.W., Van Damme, S., Puype, M., Vandekerchove, B.G., Ratz, G.P., Lauridsen, J.B., Celis, J.E. Proc. Natl. Acad. Sci. USA. 1989; 86(7701)

6. Simpson, R.J., Ward, L.O., Reid, G., Battenham, M.P., Moritz, R.L. J. Chromatogr. 1989; 476(345)

7. Yuen, S.W., Chui, A.H., Wilson, K.J., Yuan, P.M. Biotechniques. 1989; 7(74)

8. Eckerson, C., Lottspeich, F. Chromatographia. 1989; 28(92)

9. Abersold, R.H., Leavitt, J., Saavedra, R.A., Hood, L.E., Kent, S.B.H. Proc. Natl. Acad. Sci. USA. 1987; 84(6970)

10. Vanfleteren, J.R. Anal. Biochem. 1989; 178(385)

11. Szewcyk, B., Summers, D.F. Anal. Biochem. 1988; 168(48)

12. Schagger, H., VonJagow, G. Anal. Biochem. 1987; 166(368)

13. Laemmli, U.K. Anal. Biochem. 1970; 155:23–27.

14. Zalut, C., Henzel, W.J., Harris, W. J. Biochem. Biophys. Methods. 1980; 3(11)

15. Wessel, D., Flugge, U.I. Anal. Biochem. 1984; 138(141)

16. Henzel, W.J., Rodriguez, H., Watanabe, C. J. Chromatogr.. 1987; 404:41–52.

17. Houen, G., Sando, T. Anal. Biochem. 1991; 193(186)

Automated C-Terminal Sequencing of Peptides

Jerome M. Bailey, Narmada R. Shenoy and John E. Shively,     Division of Immunology, Beckman Research Institute of the City of Hope, Duarte, CA 91010

Publisher Summary

This chapter discusses the development of a chemical method for the sequential degradation of a protein or peptide from the carboxy-terminus. The development of a C-terminal method involves the combination of several diverse areas of research, all of which must be considered concurrently. These include the design of an instrument capable of performing the sequencing chemistry, the choice of a solid support stable to the conditions used for sequencing, the optimization of methodology to covalently couple polypeptide samples to the solid support in high yield, and the development of chemistry to specifically derivatize and hydrolyze the C-terminal amino acid. The chemistry used for C-terminal sequencing consist of three steps: 1) the activation of the C-terminal carboxylic group with acetic anhydride to form an oxazolinone, 2) the derivatization of the C-terminal amino acid to a thiohydantoin with trimethylsilylisothiocyanate, and 3) the specific cleavage of the derivatized amino acid with sodium trimethylsilanolate.

I Introduction

The development of a chemical method for the sequential degradation of a protein or peptide from the carboxy-terminus is a goal of our laboratory. Such a method, in addition to complementing existing N-terminal methods of degradation, would be invaluable for the sequence analysis of proteins with naturally occurring N-terminal blocking groups and for the detection of post-translational processing at the carboxy-terminus of expressed gene products. Although several methods for a sequential C-terminal degradation have been proposed (1,2), the thiocyanate method based on the procedure originally published by Schlack and Kumpf (3) has been the most widely studied. Recent work in our laboratory, introducing new reagents for the derivatization of the C-terminal amino acid (4) and for the specific cleavage of the derivatized amino acid (5,6), has suggested that an automated chemical method for the sequential degradation of polypeptides from the carboxy-terminus, analogous to the Edman method for amino terminal degradation, may be feasible.

The development of a C-terminal method involves the combination of several diverse areas of research, all of which must be considered concurrently. These include the design of an instrument capable of performing the sequencing chemistry, the choice of a solid support stable to the conditions used for sequencing, optimization of methodology to covalently couple polypeptide samples to the solid support in high yield, and the development of chemistry to specifically derivatize and hydrolyze the C-terminal amino acid. As shown in figure 1, the chemistry currently used in our laboratory for C-terminal sequencing can be considered to consist of three steps: 1) activation of the C-terminal carboxylic group with acetic anhydride to form an oxazolinone, 2) derivatization of the C-terminal amino acid to a thiohydantoin with trimethylsilylisothiocyanate (TMS-ITC), and 3) specific cleavage of the derivatized amino acid with sodium trimethylsilanolate.

Fig. 1 Chemical Scheme for C-Terminal Sequencing.

II EXPERIMENTAL

The instrument used for automated C-terminal sequencing was constructed by the Biomedical Instrumentation Services of the City of Hope. This instrument is a modified N-terminal sequencer based on the design described by Hawke et al. (7) and includes the addition of a continous flow reactor (CFR) and an on-line HPLC for detection of the released thiohydantoin amino acids (8). The thiohydantoin amino acids were synthesized as described (5).

Reverse phase separation of the thiohydantoin amino acids was performed as described (6,9). Peptide samples to be sequenced were covalently coupled to a carboxylic acid modified polyethylene film (PE-COOH). This support is commercially available from the Pall Corporation (Long Island, NY). Strips of the polyethylene film (1 × 12.5 mm) were activated by an excess of 1,3-dicyclohexylcarbodiimide in anhydrous DMF (2g/2 ml) for 2 hr. Each strip was then washed with anhydrous DMF and dried in a vacuum centrifuge. The activated PE-COOH strips were then coupled to the peptides overnight (20 hr) at 22°C in a continuous flow reactor (CFR) (8). The microbore tubing on one end of the CFR was sealed by heating and then pinched closed with pliers. Peptide solutions were prepared in water containing 1% hexafluoroacetone. Aliquots of these solutions (2 μl to 5 μl) were transferred to the CFR. Ten microliters of a 1:1 solution of 25 mM HEPES buffer (pH 6.7) and DMF was added to each reaction mixture and the reaction volume was made up to 15 μl with 1% hexafluoroacetone trihydrate in water. The activated membrane was then inserted into each of these peptide solutions within a CFR as described above and the coupling reaction was carried out overnight (20 hr) at 22°C. At the end of the reaction, each membrane was washed with 2 ml of 25 mM Hepes buffer : DMF (1:1), water (2 ml), and methanol (2 ml) and dried in a vacuum centrifuge.

III RESULTS AND DISCUSSION

The Solid Support

The need to apply the chemistry for C-terminal sequencing to the solid phase was recognized early on by a number of groups (4, 10–14). Solution phase methods required laborious gel filtrations for purification of the peptidylthiohydantoin and evaporations to remove reagents (15). By covalently coupling a polypeptide to a solid support and thereby providing the ability to simply rinse away excess solvents without loss of the sample, the time required for a single cycle of degradation could be reduced to hours rather than the several days required by solution methods. Many of the initial studies involving the application of the thiocyanate chemistry for C-terminal sequencing to the solid phase involved the use of glass beads for the covalent immobilization of peptide samples (4, 10–12,14). More recent work has involved the use of DITC-activated amino PVDF (13,16), carboxylic acid modified PVDF (6), and a disuccinimidoyl carbonate polyamide resin (17). The use of the glass and PVDF supports were found by our laboratory to be unsatisfactory for C-terminal sequencing (18). Only one cycle of degradation was obtained when the glass supports were used. Amino acid analysis of the glass support after 3 cycles of automated sequencing showed very little peptide remaining on the support. The loss of the peptide sample is most likely due to hydrolysis of the siloxane bonds formed on derivatization of the support during the basic conditions used for sequencing. This type of linkage is known to be base labile (19,20). Although the derivatized PVDF supports gave better sequencing yields than glass, these supports rapidly turned progressively blacker and more brittle with each cycle of sequencing. This is most likely due to dehydrofluorination of the support during the basic conditions of the cleavage reaction (21,22). This observation is consistent with the results of Shinohara (21) who showed that derivatized PVDF was much more susceptible to degradation (dehydrofluorination) by a basic hydroxide solution than was underivatized PVDF. A more suitable support was found to be a carboxylic acid modified polyethylene film. This novel solid support for C-terminal sequencing offers a number of advantages over the existing supports. These include 1) stability of the support to thec onditions employed for C-terminal sequencing, 2) the ability, due to the hydrophilic nature of the surface groups, to use both aqueous and organic solvents for performing chemistry on covalently coupled polypeptide samples, 3) a high capacity (3.2 nmoles/mm2) to covalently couple polypeptides, 4) the support is easily cut into 1 × 12.5 mm pieces for use in our continuous flow reactor (CFR) which is also used for automated N-terminal sequencing (8).

Covalent Immobilization of Peptides to PE-COOH

A number of reagents were tested for their ability to activate the surface carboxylic acid groups of PE-COOH. The highest yields of covalently attached peptide were obtained when DCC was used as an activating reagent. The optimal activation time was found to be 1 to 2 hours with an overnight coupling reaction with the peptide. Experiments to address the generality of this peptide coupling procedure are shown in Table I. Sixteen different peptides at two concentration levels (1 and 10 nmol) were applied to strips of PE-COOH (1 × 12.5 mm) activated with an excess of DCC in anhydrous DMF and coupled to the peptides overnight (20 hr) in a CFR as described above. A variety of peptides with differing chemical and physical properties were studied, and all of these coupled efficiently to the membrane surface (average yield was approx. 50%). A peptide with proline at the N-terminus (PFAL), which has a secondary amino group, was also found to couple to the membrane surface in relatively good yield. Experiments to determine the capacity of a 1 × 12.5 mm strip of PE-COOH (the size typically used for sequencing) for the covalent coupling of leucine enkephalin, YGGFL, revealed that the capacity of these small strips for short peptides is approximately 40 nmol (3.2 nmol/mm2).

Table I

Effect of Peptide Nature on the Coupling Yields of Carboxyl Activated PE-COOH

Automated Sequencing of Peptides Covalently Immobilized on PE-COOH

The composition of the reagents and solvents used and a description of the program for automated C-terminal sequencing is shown in Tables II and III, respectively. The total time required for each cycle was 120 min. The automated C-terminal sequencing of small peptides (3-10 residues, 3-20 nmol) routinely provided complete sequence determination. Typically the yield of sequencing decreased as the amino acid being sequenced became physically closer to the surface of the solid support. Amino acid composition of the support after several cycles of sequencing revealed that the C-terminal amino acids were completely removed while the N-terminal amino acid was only partially removed (50%). Since this is the amino acid from which the peptide was covalently attached to the solid support, it may be either difficult to derivatize this amino acid or subsequently hydrolyze it due to its proximity to the membrane surface. This surface effect was minimized by the introduction of spacers to the membrane surface prior to the attachment of peptide samples. This will permit the peptide sample being sequenced to be further from the surface of the solid support. Sequencing results, shown in figure 2, using 11-aminoundecanoic acid as a linker arm support this conclusion and resulted in significantly improved yields of sequencing with the pentapeptide, YGGFL. Amino acid analysis revealed only 12% of the tryosine remaining on the support after 6 cycles of sequencing.

Table II

Composition of Reagents and Solvents used in Automated C-Terminal Sequencing

Table III

C-Terminal Sequencer Program Summary

Fig. 2 Automated C-Terminal Sequencing of Leucine Enkephalin (3.7 nmol) Covalently Coupled to PE-COOH.

The generality of our automated C-terminal sequencing methodology was examined by sequencing model peptides containing all twenty of the common amino acids. All of the amino acids were found to sequence in high yield (90% or greater) except for asparagine and aspartate, which could be only partially removed, and proline which was found to not be capable of derivatization. In spite of these current limitations, the methodology should be a valuable new tool for the C-terminal sequence analysis of peptides. Work is continuing to extend this technique to the sequence determination of covalently coupled proteins and toward improvement of the sequencing yields obtained when Asn, Asp, and Pro are encountered.

Acknowledgments

We thank the Tosoh Corporation for funding this research and Dr. Harvey Brandwein of the Pall Corporation for providing us with the carboxyl modified polyethylene films used in this study.

REFERENCES

1. Ward, C. W., Practical Protein Chemistry – A Handbook. Darbre, A. 1986:491–525. [John Wiley and Sons, Ltd.].

2. Rangarajan, M., Protein/Peptide sequence Analysis: Current Methodologies. Brown, A.S. CRC Press, Bora Raton, 1988:135–144. [FA].

3. Schlack, P., Kumpf, W. Z. Physiol. Chem.. 1926; 154:125–170.

4. Hawke, D. H., Lahm, H.-W., Shively, J. E., Todd, C. W. Anal. Biochem.. Neuroradiology. 1987; 166:298–307.

5. Bailey, J.M., Shively, J.E. Biochemistry.. 1990; 29:3145–3456.

6. Bailey, J. M., Shively, J. E., Techniques in Protein Chemistry: II. Villafranca, J.J. Academic Press, Inc., 1991:115–129.

7. Hawke, D.H., Harris, D.C., Shively, J.E. Anal. Biochem.. 1985; 147:315–330.

8. Shively, J.E., Miller, P., Ronk, M. Anal. Biochem.. 1987; 163:517–529.

9. Bailey, J.M., Shenoy, N.R., Ronk, M., Shively, J.E. Submitted 1992;

10. Williams, M.J., Kassell, B. FEBS Lett.. 1975; 54:353–357.

11. Rangarajan, M., Dardre, A. Biochem. J.. 1979; 157:307–316.

12. Meuth, J.L., Harris, D.E., Dwulet, F.E., Crowl-Powers, M.L., Gurd, F.R.N. Biochem.. 1982; 21:3750–3757.

13. Bailey, J. M., Miller, C. G., Kong, C.-T., Shively, J. E., Techniques in Protein Chemistry. Hugli, T.E. Academic Press, Inc., 1989:67–78.

14. Inglis, A. S., Wilshire, J. F.K., Casagranda, F., Laslett, R. L., Methods in Protein Sequence Analysis. Wittmann-Liebold, B. Springer-Verlag, 1989:137–144.

15. Stark, G.R. Biochemistry 1968; 7:1796–1807

16. Inglis, A. S., Moritz, R. L., Begg, G. S., Reid, G. E., Simpson, R. J., Graffunder, H., Matschull, L., Wittman-Liebold, B., Met. Protein Sequence Analysis. Jornvall, Hoog, Gustavsson. Birkhauser-Verlag, Basel, 1991:23–34.

17. Hawke, D., Boyd, V. L., Met. Protein Sequence Analysis. Jornvall, Hoog, Gustavsson. Birkhauser-Verlag, Basel, 1991:35–45.

18. Shenoy, N.R., Bailey, J.M., Shively, J.E. Submitted 1992;

19. IIer, R.K. The Chemistry of Silica, Chap. 6. New York: John Wiley and Sons, Inc., 1979.

20. Taylor, R. F., Protein Immobilization: Fundamentals and Applications. Taylor, R.F. Marcel Dekker, Inc., New York, 1991:139–160.

21. Shinohara, H. J. Polymer Science: Polymer Chem. Ed. 17. 1979:1543–1556.

22. Dias, A.J., McCarthy, T.J. Macromolecules. 1984; 17:2529–2531.

Elution and Internal Amino Acid Sequencing of PVDF-Blotted Proteins

Kathryn L. Stonea, Mary L. LoPrestia and Kenneth R. Williamsa,     aHHMI Biopolymer Facility, New Haven, Ct 06510

Dean E. Mcnulty, J.M. Crawford and Raymond DeAngelis,     Yale University School of Medicine Protein and Nucleic Acid Chemistry Facility, New Haven CT 06510

Publisher Summary

SDS polyacrylamide gel electrophoresis (PAGE) followed by blotting onto polyvinylidene difluoride (PVDF) membranes has rapidly become one of the primary methods of choice for final or the total purification of proteins for microchemical sequence analysis. This chapter discusses the method that is an extension of the procedure of Yuen et al and follows from their finding that there is an indirect relationship between the ease of eluting proteins (with organic solvents) from PVDF membranes and their molecular weights. By first carrying out an in situ cyanogen bromide cleavage, it was found that proteins can be eluted from PVDF membranes in high yield and in a form that is amenable to further enzymatic cleavage. Although the cyanogens bromide peptides could be directly fractionated and then individually sequenced, it is best to carry out a subsequent tryptic digest and then, following reverse phase HPLC, sequence one or more of the resulting cyanogen bromide/tryptic peptides. This approach is used on 26 different unknown proteins and is found to have a 96% or greater success rate.

I Introduction

SDS psolyacrylamide gel electrophoresis (PAGE) followed by blotting onto polyvinylidene difluoride (PVDF) membranes has rapidly become one of the primary methods of choice for final or, in many instances, the total purification of proteins for microchemical sequence analysis. Because of the relatively high specificity with which PVDF membranes bind proteins, compared to most low molecular weight compounds such as salts, amino acids, and detergents, the resulting protein is isolated free of those contaminants that most commonly interfere with subsequent protein chemical analysis. Although PVDF-blotted proteins have been reported to sequence nearly as well as proteins that are applied directly from solution onto automated sequencers [1], many eukaryotic proteins have blocked NH2-termini that prevent direct sequencing. In fact, chemical analysis of soluble proteins [2] and protein database compilations [3] suggest that as many as 80-90% of eukaryotic proteins have blocked NH2-termini. In order to circumvent the problem of in vivo NH2-terminal blocking requires that the electroblotted protein be cleaved into fragments, which then must be individually purified prior to sequencing. This has been accomplished by taking at least two different approaches. Either the remaining protein binding sites on the PVDF membrane have been blocked prior to in situ enzymatic digestion, which is then followed by extraction and subsequent HPLC purification of the resulting peptides [4], or the intact protein has been extracted from the PVDF membrane with a mixture of detergents [5] which have then been removed by inverse gradient reverse phase HPLC prior to enzymatic cleavage [6]. One drawback with both these approaches is a relatively low overall recovery. That is, the overall average recovery (calculated from the ratio of the yield of phenylthiohydantoin amino acid derivative as compared to the amount of protein initially applied to the SDS psolyacrylamide gel) of three peptides isolated by the detergent extraction approach was only 2.3%, even when as much as 1.1 nmol of protein (β-lactoglobulin) was applied to the SDS PAGE gel [6]. Similarly, the overall average recovery of 10 peptides isolated from 400 pmol of bovine serum albumin by the in situ approach was 6.1% [7]. Reasoning that inverse gradient reverse phase HPLC and the relative inaccessibility of PVDF-bound proteins to proteases (leading to the generation of over-lapping peptides in relatively low yield [7]), might significantly contribute to the low recoveries associated with these two procedures, we developed an alternative strategy that avoids these