Applicability of the Alkylation Chemistry for Chemical C-Terminal

Protein Engineering, K. L. Ledeganckstraat 35, B-9000 Gent, Belgium, and ... Chemistry, Synthesis and Bio-organic Chemistry, Krijgslaan 281, S.4, B-90...
0 downloads 0 Views 433KB Size
Anal. Chem. 2000, 72, 1389-1399

Applicability of the Alkylation Chemistry for Chemical C-Terminal Protein Sequence Analysis Bart Samyn,† Klaas Hardeman,†,‡ Johan Van der Eycken,‡ and Jozef Van Beeumen*,†

University of Gent, Department of Biochemistry, Physiology, and Microbiology, Laboratory of Protein Biochemistry and Protein Engineering, K. L. Ledeganckstraat 35, B-9000 Gent, Belgium, and University of Gent, Department of Organic Chemistry, Synthesis and Bio-organic Chemistry, Krijgslaan 281, S.4, B-9000 Gent, Belgium

We have evaluated the alkylation chemistry first described some years ago by Boyd et al. which is now routinely applied in a commercial instrument. We have found that the low repetitive yields observed during these analyses are due to the formation of a major side product when alkylating the C-terminal thiohydantoin. This side product, resistant to the chemical cleavage methods currently used, was characterized by NMR experiments in solution. We further demonstrate that chemical C-terminal sequence analysis of proteins using the alkylation chemistry is feasable with low picomole amounts of material. Highsensitivity C-terminal sequencing allows a complementary approach by which a protein is first subjected to Nterminal Edman degradation followed by C-terminal sequence analysis, limiting the amount of material necessary for the characterization of the protein under study. This limited C-terminal sequence information is often sufficient to solve problems that cannot be solved by applying any other analytical method commonly used today. Chemical methods for the C-terminal sequence analysis of proteins have been under investigation for several years (for review, see ref 1). Only one approach, known as the thiocyanate method and first described in 1926,2 has led to the development of automated instruments comparable, in terms of hardware, to the sequencers used for N-terminal sequence analysis.3-4 The C-terminal amino acid of a protein is converted to a thiohydantoin with the use of a carboxyl-group activating reagent and thiocyanate anion. In the so-called alkylation chemistry, Boyd alkylates the C-terminal thiohydantoin to transform it into a better leaving group.4 As both the initial and the repetitive yields of the carboxyterminal sequencing method are much lower than the yields currently obtained for the Edman degradation, a higher sample amount is required for an unambiguous analysis of even only a * Corresponding author. Tel: 0032-9-264-5109. Fax: 0032-9-264-5338. E-mail: [email protected]. † Department of Biochemistry, Physiology, and Microbiology. ‡ Department of Organic Chemistry. (1) Bailey, J. M. J. Chromatogr., A 1995, 705, 47-65. (2) Schlack, P.; Kumpf, W. Hoppe-Seyler’s Z. Physiol. Chem. 1926, 154, 125170. (3) Bailey, J. M.; Shenoy, N. R.; Ronk, M.; Shively, J. E. Protein Sci. 1992, 1, 68-80. (4) Boyd, V. L.; Bozzini, M.; Zon, G.; Noble, R. L.; Mattaliano, R. J. Anal. Biochem. 1992, 206, 344-352. 10.1021/ac991049u CCC: $19.00 Published on Web 03/03/2000

© 2000 American Chemical Society

few steps. Whereas N-terminal sequence analysis is successfully performed on 5-10 pmol samples, the C-terminal instruments still require some 100-fold of this amount.1,5-7 Here we report a critical evaluation of the alkylation chemistry as it is currently applied on the 494 Procise sequencer. Low (20200) pmol amounts of three test proteins were electroblotted to PVDF membranes and sequenced using the alkylation chemistry. The possibility to perform C-terminal sequence analysis at this high sensitivity allows that, prior to C-terminal degradation, N-terminal sequence analysis is performed on the same sample. Sample preparation protocols and several alkylated thiohydantoin (ATH) derivatives are discussed. The different steps, derivatization, alkylation, and cleavage, of the alkylation chemistry are separately studied to investigate their influence on the initial and the repetitive yield. Sequence analysis results, as well as experiments “in solution”, demonstrate that the alkylation of the C-terminal thiohydantoin is actually very incomplete. A major side product, resistant to the currently applied cleavage conditions, has been characterized by NMR and mass spectrometry studies. EXPERIMENTAL SECTION Materials-Instrumentation. Horse heart myoglobin, bovine serum albumin (monomer), fetuin (fetal calf serum), bovine R-lactalbumine (Type III), N-(t-Boc)amino acids, the N-acetylated dipeptides valanylalanine and phenylalanylalanine, Tris, glycine, SDS, boric acid, and acetic acid were obtained from Sigma (St. Louis, Missouri). Acetonitrile, ammonium thiocyanate, ethoxycarbonyl isothiocyanate, 2-bromomethylnaphthalene, ethyl acetate, cyclohexane, sodium tetraborate decahydrate, deuterated chloroform, and deuterated methanol were from Aldrich (Milwaukee, Wisconsin). Acetonitrile was redistilled from phosphorus pentoxide prior to use. Milli-Q water was obtained from Millipore (Bedford, Massachusetts). Pyridine and triethylamine, from Acros (Geel, Belgium), were both redistilled from calcium hydride. N,Ndiisopropylethylamine and TFA were sequencing reagents from Perkin-Elmer (Applied Biosystems Division, Foster City, California). Open-column chromatography was performed using silica (5) Bozzini, M.; Zhao, J.; Yuan, P.-M.; Ciolek; D.; Pan, Y.-C.; Horton, J.; Marshak, D. R.; Boyd, V. L. In Techniques in Protein Chemistry VI; Crabb, J. W., Ed.; Academic Press: San Diego, CA, 1995; pp 229-237. (6) Miller, C. G.; Hawke, D. H.; Tso, J.; Early, S. In Techniques in Protein Chemistry; Crabb, J. W., Ed.; Academic Press: San Diego, CA, 1995; pp 219-227. (7) Burkhart, W. A.; Moyer, M. B.; Bailey, J. M.; Miller, C. G. Anal. Biochem. 1996, 236, 364-367.

Analytical Chemistry, Vol. 72, No. 7, April 1, 2000 1389

gel (230-400 mesh, Merck, Darmstadt, Germany). Thin-layer chromatography was run on glass plates precoated with silica gel (precoated TLC plates SIL G-25 UV245, 0.25 mm, Merck, Darmstadt, Germany). Column chromatography was performed using silica gel (230-400 mesh, Merck, Darmstadt, Germany). 4-20% Tris-glycine gels (50-µL wells) were from Biorad (Hercules, California). SDS-PAGE was performed using the MiniProtean II cell (Biorad, Hercules, California), and electroblotting was performed in a Mini Transblott cell (Biorad, Hercules, California). The power supply used was from Consort (Turnhout, Belgium). Formic acid (98%) and HCl were from Panreac (Barcelona, Spain). Hydrogen peroxide (30 wt %) was obtained from Aldrich (Milwaukee, Wisconsin). Chemical ionization mass spectra of the N-(t-Boc)amino acid derivatives were recorded at 70 eV on a Finnigan 400 mass spectrometer (Hemel Hempstead, United Kingdom). 1H-nuclear magnetic resonance spectra were recorded in deuterated chloroform (when necessary a few drops of deuterated methanol for better solubility were added) at 200 MHz on a Varian Gemini 200 (San Fernando, California). δ-values are expressed in parts per million versus tetramethylsilane as the internal standard. 13Cnuclear magnetic resonance spectra were recorded at 125.72 MHz on a Bruker AN 500 (Silberstreifen, Germany), linked to an Aspect 3000 computer. Electrospray mass spectra of the N-acetylated dipeptide derivatives were obtained on a benchtop HewlettPackard (Waldbronn, Germany) 1100 Series LC/MSD setup. High-performance liquid chromatography was performed on a Hewlett-Packard 1100 Series with diode array detector, equipped with a Phenomenex Bondclone column 10 C18 (300 × 3.9 mm, St. Torrance, California). SDS-PAGE and Electroblotting. Stock solutions of the three proteins containing 50 pmol of protein/µL (made up by weight) were prepared and increasing amounts loaded on the gel. Electrophoresis was performed at 80 V for 1.5 h, unless otherwise stated, using the following running buffer: 200 mM glycine, 50 mM Tris, and 3.5 mM SDS. After electrophoresis, the gel and the ProBlott PVDF sheets (Perkin-Elmer, Applied Biosystems Division, Foster City, California) were equilibrated in the blotting buffer (50 mM Tris, 50 mM boric acid, pH 8.4) and then sandwiched between three filter papers (Whatmann, Maidstone, UK) at each site. Blotting was performed at 20 V for 18-20 h while cooling the buffer tank in an ice bath. After blotting, the PVDF membranes were stained for 5 min in diluted Coomassie brilliant blue R-250 (0.05%, 5% acetic acid/35% methanol), destained for 10 min in 5% acetic acid/35% methanol, and allowed to dry. Hydrolysis and Amino Acid Analysis. Protein bands were excised from the blot and transferred to small pyrrolized Pyrex tubes which were placed in a large hydrolysis tube (Waters, Milford, Massachusetts) containing 200 µL of 6 N HCl. After hydrolysis in the vapor phase under an inert argon atmosphere for 24 h at 106 °C, amino acid analysis was performed directly on the PVDF using the model 420A amino acid analyzer (PerkinElmer, Applied Biosystems Division, Foster City, California). PTC derivatives were on-line analyzed using the following buffers: Solvent A (50 mM sodium acetate buffer, pH 5.4) and Solvent B (32 mM sodium acetate buffer, pH 6.1) in 70% ACN. Thiohydantoin Formation for N-(t-Boc)Amino Acids and N-Acetylated Dipeptides. As an example, N-Ac-VA-OH (0.22 1390

Analytical Chemistry, Vol. 72, No. 7, April 1, 2000

mmol, 51 mg) was weighed in a flask. Ten mL of acetonitrile, ethoxycarbonyl isothiocyanate (1.2 equiv, 32 µL), and pyridine (1.2 equiv, 22 µL) were added, and the mixture was stirred overnight, under argon, at room temperature. The solvent was removed by rotary evaporation yielding a crude powder. Thinlayer chromatography with cyclohexane/ethyl acetate (1:1) revealed one major product, which was purified by column chromatography with the same eluent (Rf ) 0.22). The yield was determined by weight. Alkylation of the Thiohydantoin Derivatives from N-(tBoc)Amino Acids and N-Acetylated Dipeptides. As an example, N-Ac-VA-TH (0.19 mmol, 50 mg) was weighed in a flask. Ten mL of acetonitrile, 2-bromomethylnaphthalene (1.2 equiv, 49 mg), and triethylamine (1.2 equiv, 31 µL) were added, and the mixture was stirred, under argon, at 60 °C for 1 h. The solvent was removed by rotary evaporation yielding a crude powder. Thinlayer chromatography with cyclohexane/ethyl acetate (1:1) revealed two major spots (Rf ) 0.15 and Rf ) 0.25), which were separated and purified with the same eluent. The yield was determined by weight. Sample Preparation. Prior to direct C-terminal sequence analysis, the electroblotted or adsorbed (ProSorb, Perkin-Elmer, Applied Biosystems Division, Foster City, California) proteins were treated with 200 mM phenylisocyanate in ACN under basic conditions (124 mM diisopropyl ethylamine/ACN) in order to derivatize the -amino group of the lysine residues into stable phenylureas.8 Samples for C-terminal sequence analysis already subjected to N-terminal Edman degradation were treated with performic acid. The latter was prepared by adding 5 µL of 30% H2O2 to 95 µL of 99% HCOOH and keeping the Eppendorf for 2 h at room temperature.9-10 The membrane pieces of PVDF were treated twice with 5 µL of the freshly prepared performic acid solution. N- and C-Terminal Sequence Analysis. All reagents and solvents both for N- and C-terminal sequence analysis as well as HPLC solvents were obtained from Perkin-Elmer (Applied Biosystems Division, Foster City, California). N-terminal sequence analysis was performed in the gas-pulsed liquid phase using the cross-flow reaction chamber on the model 476A protein sequencer and an on-line HPLC system for phenylthiohydantoin (PTH) analysis (Perkin-Elmer, Applied Biosystems Division, Foster City, California). Absorbance was monitored at 269 nm. C-terminal sequence analysis was performed on a Procise 494C protein sequencer (Perkin-Elmer, Applied Biosystems Division, Foster City, California) using the C-terminal sequencing chemistry first described in 19924 and most recently reviewed in 1995.5 A detailed description is given in the Results and Discussion section. The liberated alkylated thiohydantoin (ATH) amino acids were on-line analyzed on a thermostated (38 °C) C18 reversed-phase column (2.1 × 220 mm, 5 micro meters) (Brownlee, Perkin-Elmer, Applied Biosystems Division, Foster City, California). A linear gradient with a flowrate of 300 µL/min (8) Guga, P. J.; Bozzini, M.; DeFranco, R. J.; Large, G. B.; Boyd, V. L. C-terminal sequence analysis of the amino acid residues with reactive side-chains: Ser, Thr, Cys, Glu, Asp, His, Lys. Presented at the Seventh Symposium of the Protein Society, San Diego, CA, 1993. (9) Hirs, C. H. W. In Methods in Enzymology; Hirs, C. H. W., Ed.; Academic Press: New York, 1967; pp 197-199. (10) Fontana, A.; Gross, E. In Practical protein Chemistry: a Handbook; Darbre, A., Ed.; John Wiley & Sons Ltd.: New York, 1986; pp 67-120.

Figure 1. General scheme for the alkylation sequence chemistry. Reactions 1-2: Formation of C-terminal oxazolinone through reaction of the C-terminal carboxyl group with acetic anhydride in the presence of 2-6-dimethylpyridine (lutidine). Reaction 3: Derivatization of the oxazolinone with tetrabutylammonium thiocyanate in the presence of TFA vapors resulting in the formation of a proteinyl isothiocyanate which cyclizes to the C-terminal thiohydantoin. Steps 1-3 are performed in the so-called “C-terminal activation 1” cycle. Reaction 4: Alkylation of the C-terminal thiohydantoin with 2-bromomethylnaphthalene under basic conditions (2% DIEA/n-heptane). Reaction 5: ATH cleavage with tetrabutylammonium thiocyanate in the presence of TFA vapors. Steps 4 and 5 are performed in the so-called “sequencing 1” cycle which is the default cycle during sequence analysis. Reaction 5′: Since the cleavage generates the thiohydantoin of the penultimate C-terminal amino acid, the subsequent cycles require only alkylation and cleavage reactions.

was formed using a 140C micro gradient system (Perkin-Elmer, Applied Biosystems Division, Foster City, California) with the following solvents: Solvent A (35 mM sodium acetate buffer/3.5% tetrahydrofuran/MQ-water) and Solvent B (100% ACN). The ATH derivatives were monitored using a 785A absorbance detector set at 254 nm and quantified relative to a 100 pmol ATH amino acid standard. The methyl naphthyl thiohydantoin (mnth) amino acid standards were obtained from the supplier (Perkin-Elmer, Applied Biosystems Division, Foster City, California). The following derivatives were included: acetyl-Tyr mnth, -phenylcarbamoyl Lys-mnth, piperidineamide Asp-mnth, piperidineamide Glu-mnth, and 2-cyanothiomethylnaphthalene (ctmn). RESULTS AND DISCUSSION Description of the Automated C-Terminal Protein Sequencing Chemistry. C-terminal sequence analysis was performed on a specially configured Procise sequencing platform using the C-terminal sequencing chemistry (Figure 1) first described by Boyd4 and most recently reviewed in 1995.5

In the first activation cycle, the C terminus of the immobilized protein is derivatized with acetic anhydride and 2,6-dimethylpyridine to a C-terminal oxazolinone (Figure 1, steps 1 and 2). This oxazolinone is then converted to a thiohydantoin (TH) with tetrabutylammonium thiocyanate in the presence of trifluoroacetic acid (TFA) vapor (Figure 1, step 3). In a second activation cycle the side-chain carboxyl groups of Asp and Glu are selectively amidated using piperidine thiocyanate (Figure 2A). During the first activation cycle, the side-chain carboxyl groups form mixed anhydrides when reacted with acetic anhydride. Delivery of piperidine thiocyanate generates piperidine which, under mildly basic conditions (due to the presence of lutidine), selectively amidates all Asp and Glu side chains. Since the proteinyloxazolinone as well as the proteinyl-thiohydantoin are ionized under basic conditions, these derivatives are less reactive toward the nucleophile.11 To increase the initial yield, the first activation (11) Boyd, V. L.; Bozzini, M.; Guga, P. J.; DeFranco, R. J.; Yuan, P.-M. J. Org. Chem. 1995, 60, 2581-2587.

Analytical Chemistry, Vol. 72, No. 7, April 1, 2000

1391

Table 1. C-terminal Sequence Analysis of Electroblotted Test Proteins horse Myoglobin raw yield (pmol)

amt (pmol) SDS-PAGE

blot AAA

250 200 150 100

98 63 52 28

Leu 5′

Gly 4′

1.0 nd 0.7 0.8

2.3 nd 1.3 1.0

Phe 3′

Gln 2′

Gly 1′a

6.5 ni ni ni

5.5 5.0 3.1 3.2

3.2 4.3 2.4 1.9

bovine Serum Albumin raw yield (pmol)

amt (pmol) SDS-PAGE

blot AAA

Gln 5′

Thr 4′

Ala 3′

Leu 2′

Ala 1′

250 200 150 100

85 65 48 29

nde nd nd nd

nd nd nd nd

8.6 9.7 6.6 3.7

15 15 11 6.7

19 25 20 8.1

amt (pmol) SDS-PAGE 250 200 150 100

fetuin (Fetal Calf Serum) raw yield (pmol)

blot AAA

Arg 5′b

Tyr 4′c

Phe 3′

Lys 2′d

Ile 1′

162 141 87 56

+f

9.8 4.7 2.8 0.9

18 13 7 3.2

16 20 9.1 4.2

30 32 19 10

+ + +

a Partially recovered as acetylated Gly-ATH. The amino acid labeled 1′ indicates the C-terminal amino acid. b Recovered as acetylated ArgATH. c Recovered as acetylated Tyr-ATH. d Recovered as phenylcarbamoyl-Lys-ATH. e nd, not determined; ni, not integrated. f +, present but not quantified.

Table 2. Recovery of ATH Derivatives of All 20 Common Amino Acids ATH Derivatives reliably called Gln, Ala, Val, Met, Ile, Leu, Phe, Trp, Asn, Glya Figure 2. Additional chemical modification cycles performed on the sequencer: (A) In the second activation cycle the carboxyl groups of Asp and Glu are amidated using piperidine thiocyanate; this is the so-called “C-terminal activation 2” cycle. (B) After the alkylation, the hydroxyl groups of serine and threonine are protected by an Nmethylimidazole catalyzed acetylation with acetic anhydride (the socalled “OH-capping & sequencing” cycle). (C) The doubly alkylated Gly adduct observed during the “in-solution” alkylation of N-(t-Boc)Gly-TH. (C′) The oxygen-acetylated Gly adduct formed during the activation cycles if Gly is the C-terminal amino acid.

cycle is repeated once more after the second activation cycle. Next the proteinyl thiohydantoin is S-alkylated with 2-bromomethylnaphthalene under basic conditions (2% N,N-diisopropylethylamine (DIEA)/n-heptane) (Figure 1, step 4). In the same cycle, an acetylation is performed on the hydroxyl groups of Ser and Thr using acetic anhydride and N-methylimadazole as a catalyst (Figure 2B). This acetyl moiety is subject to elimination during sequencing, so the Ser and Thr residues are detected as their dehydrated derivatives. If underivatized, the hydroxyl groups can displace the sulfur from the adjacent alkylated thiohydantoin which will prevent cleavage and result in blocking further sequence analysis.12 The alkylated thiohydantoin (ATH) is finally cleaved with tetrabutylammonium thiocyanate in the presence of TFA vapors 1392 Analytical Chemistry, Vol. 72, No. 7, April 1, 2000

+b

modified dehydro derivatives

Ser Thrc Tyr Arg Asp Glu Lysf Cys Lyse

acetylated piperidine amidated PIC modification performic acid oxidation

ndd + + + + -

problematic Hisg Proh

nd nd

a Oxygen-acetylated if C-terminal amino acid. b The +/- indicates whether the derivative is present or not in the commercially available standard. c Coelutes with ctmn. d nd indicates that this derivative is not detectable. e After Edman degradation. f Only for direct C-terminal sequence analysis. g Unstable derivative. h Brings the sequence analysis to a stop.

(Figure 1, step 5). Since the cleavage generates the thiohydantoin of the penultimate C-terminal amino acid, the subsequent cycles require only alkylation and cleavage reactions (Figure 1, step 5′). After cleavage, the ATH is transferred to the conversion flask where it is dried and redissolved in 20% acetonitrile/water for online HLPC injection.

Figure 3. Separation of the methyl naphthyl thiohydantoin amino acid standards (100 pmol). Included are the propionamide Cys-ATH (pamC), phenylcarbamoyl Lys-ATH (phc-K), acetyl Tyr-ATH (ac-Y), piperidineamide Asp-ATH (pip-D), and piperidineamide Glu-ATH (pip-E). Arrows numbered one to four indicate the retention times of the following (not commercially available) ATH-derivatives: (1) cysteic acid-ATH, (2) acetyl Lys-ATH, (3) acetyl Arg-ATH, and (4) dehydroalanine ATH.

Sample Immobilization for Direct Application of the C-Terminal Alkylation Chemistry. In the course of the development of an automated C-terminal sequencing chemistry, it has often been proposed that proteins and peptides should be covalently attached to a supporting matrix in order to prevent sample washout. Many of the initial studies involved immobilization of peptides on glass beads13-14 or on carboxylic acid modified membranes such as PVDF15 or polyethylene film.16 However, covalent immobilization of proteins has never gained widespread application, mainly because it turned out that quantitative attachment of the sample to the support was only rarely possible and also because special sets of reagents or solvents are required for covalent attachment. Furthermore, gaps are likely to appear in the sequence data at the sites where attachment has been successful. The use of glass-fiber filters coated with Polybrene for retaining noncovalently applied peptides or proteins in modern gas-phase sequencers17 has been superseded by the use of PVDF membranes which were found to give superior yields for electroblotting and sequence analysis.18-20 PVDF membranes have also been reported to resist the C-terminal alkylation chemistry.5 Bailey suggested using Zitex (porous Teflon) to noncovalently apply proteins.21 This membrane is chemically more resistant than (12) Dupont, D.; Woo, S.; Bozzini, M.; Noble, R.; Bergot, J.; Yuan, P.-M.; Boyd, V. Sequencing proteins from the carboxy-terminus. Presented at The 10th Meeting of the Protein Society, San Jose, CA, 1996. (13) Williams, M. J.; Kassell, B. FEBS Lett. 1975, 54, 353-357. (14) Inglis, A. S.; Wilshire, J. F. K.; Casagranda, F.; Laslett, R. L. In Methods in Protein Sequence Analysis; Wittmann-Liebold, B., Ed.; Springer-Verlag: Berlin, 1989; pp 137-144. (15) Bailey, J. M.; Shively, J. E. In Techniques in Protein Chemistry: II.; Villafranca, J. J., Ed.; Academic Press: New York, 1991; pp 115-129. (16) Shenoy, N. R.; Bailey, J. M.; Shively, J. E. Protein Sci. 1992, 1, 58-67. (17) Hewick, R. M.; Hunkapiller, M. W.; Hood, L. E.; Dreyer, W. J. J. Biol. Chem. 1981, 256, 7990-7997. (18) Matsudaira, P. J. Biol. Chem. 1987, 262, 10035-10038. (19) Xu, Q.; Shively, J. E. Anal. Biochem. 1988, 170, 19-30. (20) Speicher, D. W. Methods: A Companion to Methods in Enzymology 1994, 6, 262-273.

PVDF and can also be used for electroblotting from polyacrylamide gels.22 Recently, a novel sample preparation device, the ProSorb Cartridge, was introduced by Perkin-Elmer.23 It uses an absorbent filter to draw the sample containing solution through a PVDFmembrane which is then washed with water to remove salts or other contaminants which may interfere with the reactions of sequencing. C-terminal sequence analysis on a series of ProSorb membranes with increasing amounts of adsorbed protein (200 to 2000 pmol) showed, however, that there is no linear correlation between the amounts of protein applied and the initial yields of the C-terminal amino acids (results not shown). It has been reported that the protein binding capacity of the ProSorb membrane not only varies with the amount of protein applied,23 but also that at higher protein concentrations the material accumulates on the surface of the membrane instead of being adsorbed. Subsequently, protein is lost during the washes with ethyl acetate used to remove salts and side products formed in the activation and sequencing cycles. Therefore, three proteins at increasing quantities of 100, 150, 200, and 250 pmol (see Experimental Section) were electroblotted, after SDS-PAGE, to a PVDF membrane. After staining with Coomassie blue, the bands were excised using a razor blade and subjected to C-terminal sequence analysis. The raw yields of the C-terminal ATHs are given in Table 1. The initial C-terminal sequencing yield is defined as the molar ratio (in %) of amino acid derivative recovered after the first cycle to the total amount of polypeptide loaded. (21) Bailey, J. M.; Rusnak, M.; Shively, J. E. Anal. Biochem. 1993, 212, 366374. (22) Burkhart, W. A.; Moyer, M. B.; Bodnar, W. M.; Everson, A. M.; Valladares, V. G.; Bailey, J. M. In Techniques in Protein Chemistry VI; Crabb, J. W., Ed.; Academic Press: San Diego, 1995; pp 169-176. (23) Werner, W. E.; Hsi, K.-L.; Grimley, C.; Yuan, P.-M. A new sample preparation cartridge for protein and peptide sequencing. Presented at the Ninth Symposium of the Protein Society, Boston, MA, 1995.

Analytical Chemistry, Vol. 72, No. 7, April 1, 2000

1393

Figure 4. (A) Two cycles of Edman degradation on electroblotted Chromatium vinosum cytochrome c-551. (B) C-terminal sequence analysis on the same sample after performic acid oxidation (140 pmol). The C-terminal sequence is -Ile-Leu-Gly-Leu-OH.

As the electrophoretic transfer of proteins during electroblotting is known to vary according to the nature of the protein,20,24 the actual amount of protein electroblotted to the PVDF membrane was determined by amino acid analysis. Table 1 shows that the amounts on the PVDF membrane were nearly all less than 100 pmol. We found that the initial yield for the C-terminal residue of horse myoglobin varied between 3 and 7%. This is much lower than the 20-40% observed for bovine serum albumin and the 20% for fetal calf serum fetuin. Despite the low initial yield, up to 5 C-terminal amino acids could be unambiguously determined even for the lowest amount of myoglobin (28 pmol) (Table 1). (24) Gershoni, J. M.; Palade, G. E. Anal. Biochem. 1982, 124, 396-405.

1394 Analytical Chemistry, Vol. 72, No. 7, April 1, 2000

Chemical Background/Chemical Impurities/ATH Derivatives. As is the case in the Edman degradation chemistry, a number of side products are formed which can interfere with the detection of the separated amino acid derivatives. The excess of cleavage reagent, tetrabutylammonium thiocyanate, elutes just before the Gly-ATH when analyzed on a C18 RPLC column. The major side product formed during sequence analysis is 2-cyanothiomethylnaphthalene (ctmn), the reaction product resulting from the use of excesses of alkylating reagent (2-bromomethylnaphthalene) and the cleavage reagent (tetrabutylammonium thiocyanate). At very low concentration (