- Email: [email protected]
Improved conditions for the removal of 2-oxoacyl groups from the N-terminus of proteins
Biochimica et Biophysica Acta 1388 (1998) 45^52
Improved conditions for the removal of 2-oxoacyl groups from the N-terminus of proteins Margaret Sunde 1 , Michael J. Sparkes, Henry B.F. Dixon * Department of Biochemistry, University of Cambridge, Tennis Court Road, Cambridge CB2 1QW, UK Received 7 April 1998; revised 29 June 1998; accepted 2 July 1998
Abstract Proteins with R^CO^CO^NH^ at the N-terminus, rather than the usual R^CH(^NH3 )^CO^NH^, are produced by nonenzymic transamination and also occur in the pyruvoyl enzymes. The oxoacyl group may be specifically removed from a model peptide, in yields of 70^80%, by treating them in 0.1 M phosphate buffer at 37³C for 24 h with 25 mM of the N-phosphonomethyl derivative of phenylene-1,2-diamine. This provides mild conditions for the stepwise removal of N-terminal residues without denaturation. ß 1998 Elsevier Science B.V. All rights reserved. Keywords: Transamination; Scission; N-Terminal residue
1. Introduction The role of N-terminal residues of a protein can be studied by their stepwise removal in non-denaturing conditions. This can be achieved by transamination, which speci¢cally modi¢es terminal amino groups in the reaction R^CH(^NH3 )^CO^NH^RPCR^CO^ CO^NH^RP, followed by scission of the 2-oxoacyl groups formed to yield NH3 ^RP [1]. The scission can be used also to study pyruvoyl enzymes in which the 2-oxoacyl group acts as an electron sink and is essential for catalytic activity [2,3]. We wished to use this scission to remove the pyruvoyl group from human S-adenosylmethionine decarboxylase (AdoMet Abbreviations: Pyr-, pyruvoyl * Corresponding author. Fax: +44-223-333345; E-mail: [email protected] 1 Present address: New Chemistry Laboratory, Oxford Centre for Molecular Sciences, South Parks Road, Oxford OX1 3QT, UK.
DC) [4], so that it could be replaced with other 2-oxoacyl groups that might be stronger electron sinks in the decarboxylation reaction. For this, the amino group liberated by the scission would need to be acylated with a 2-oxo acid. The anions of such acids might be expected to be poor nucleophiles; nevertheless it proved possible [5] to make the esters of pyruvic, 2-oxobutanoic and phenylpyruvic acids with Nhydroxysuccinimide using bis(succinimid-N-yl) sul¢te and pyridine in dimethylformamide. Hence the amino group liberated could be acylated. To make this reaction speci¢c, all other amino groups in the molecule would need either reversible protection or conversion into guanidino groups by amidination. The enzyme, however, proved to be unstable in the standard conditions for scission [1,6], and also in the milder, neutral conditions developed by Stevens and Dixon [7] of phenylene-1,2-diamine in strong phosphate bu¡er. We therefore tested other scission reagents with a model peptide substrate, Pyr^Ala^Tyr, and analysed the progress of the reaction with reverse-
0167-4838 / 98 / $ ^ see front matter ß 1998 Elsevier Science B.V. All rights reserved. PII: S 0 1 6 7 - 4 8 3 8 ( 9 8 ) 0 0 1 6 2 - 9
BBAPRO 35737 2-10-98
46
M. Sunde et al. / Biochimica et Biophysica Acta 1388 (1998) 45^52
phase high-performance liquid chromatography (HPLC). The highest rates of scission were achieved with the N-phosphonomethyl derivative [8,9] of phenylene-1,2-diamine (Fig. 1) in 0.1 M phosphate at pH 7.0 at 37³C; under these conditions AdoMet DC is stable, and they are mild enough to be unlikely to harm transaminated proteins. 2. Materials and methods 2.1. Materials Sodium glyoxylate monohydrate was prepared [10] from glyoxylic acid. Phenylene-1,2-diamine was puri¢ed by recrystallization from water. Peptides were synthesized by coupling N-t-Boc-L-amino acids [11] and L-amino acid methyl esters [12] using EEDQ [13]. All reagents were purchased from Aldrich. 2.2. Methods 2.2.1. Synthesis of scission reagents The new scission reagents (Fig. 1) were derivatives of phenylene-1,2-diamine, the binucleophilic reagent [1] used most commonly until now in the scission reaction. (i) Phosphonomethyl derivative: (N-(2-aminophenyl)aminomethylphosphonic acid). This was synthesized from chloromethylphosphonic acid and phenylene-1,2-diamine, as previously described [8]. After synthesis the material was dissolved in water at pH 8.0. Any remaining phenylene-1,2-diamine was removed by repeated extraction with ethyl acetate. The product crystallized out of solution when the pH was adjusted to 4.1. (ii) Hydroxyethyl derivative: (2-[(2-aminophenyl)amino]ethanol). 1-Fluoro-2-nitrobenzene (2.1 ml, 0.02 mol) was added to 10 ml of ethanolamine (0.1 mol). The mixture was incubated at 37³C and the progress of the reaction was followed by thin-layer chromatography (TLC) on a silica-coated plate, developed in a mixture of toluene and ethyl acetate (9:1). After 3 h the reaction was judged to be complete and 100 ml of water was added. The pH was adjusted to 1.5 with HCl and the product, 2-[(2-nitrophenyl)amino]ethanol, was extracted from the aqueous solution into ethyl acetate (4U50 ml). The
organic extracts were combined and concentrated under vacuum, leaving an orange oil. The 2-[(2-nitrophenyl)amino]ethanol was reduced by re£uxing with tin in 6 M HCl until the liquid was colourless, and residual tin was precipitated by passing H2 S through the solution after adjustment to pH 0.6 with NaOH. The stannous sul¢de was ¢ltered o¡, and the ¢ltrate was dried down. The residue was dissolved in water, adjusted to pH 10.5 with NaOH, and the product was extracted into ethyl acetate and dried down to a brown crystalline solid. Unfortunately, analysis was not performed; the compound is only characterized by its route of synthesis, decolorization during synthesis, homogeneity by TLC, and, as an o-diamine, by giving scission. 2.2.2. Transamination of Ala^Ala^Tyr to produce Pyr^Ala^Tyr Ala^Ala^Tyr was dissolved with sodium glyoxylate monohydrate (1.2:1 ratio) in a solution of 2 M pyridine, 0.8 M acetic acid and 10 mM cupric acetate. The mixture was incubated at 37³C for 1 h and then applied to a sulfonic acid resin in the acid form. The product was washed through with three column volumes of water and the eluate concentrated under vacuum. The ¢nal concentration of the Pyr^Ala^Tyr stock solution was determined by amino-acid analysis in the Protein and Nucleic Acid Facility of the Department of Biochemistry, University of Cambridge. 2.2.3. Conditions for the scission reaction with peptides The ratio of scission reagent to Pyr^Ala^Tyr was 4:1, usually 25 mM reagent and 6.25 mM peptide. Reaction mixtures were saturated with nitrogen and incubated at 37³C. When samples were removed for analysis during the incubation period, nitrogen was bubbled through the solution again. Reverse-phase HPLC was used to follow the course of the scission reaction. The concentration of the stock solution of Pyr^Ala^Tyr was known from amino-acid analysis. The amount of Ala^Tyr released was determined from the peak area of the product on the HPLC trace. A standard curve was produced by injecting known amounts of a solution of Ala^Tyr in phosphate bu¡er prepared from solid Ala^Tyr peptide.
BBAPRO 35737 2-10-98
M. Sunde et al. / Biochimica et Biophysica Acta 1388 (1998) 45^52
2.2.4. Reverse-phase HPLC systems A Varian Model 5000 liquid chromatograph was used with a Kratos SF 769 UV detector and Shimadzu C-R6A Chromatopac data processor. Samples were injected onto the Varian Reverse Phase CH-10 column manually, through a Rheodyne 7125 injector port ¢tted with either a 100- or 500-Wl sample loop. Samples of scission reaction components were diluted 20-fold in 25 mM citrate bu¡er (pH 2.85) before injection onto the column. The elution of active ester reaction components and of peptides and proteins was monitored by absorbance at 280 nm. All HPLC bu¡ers were made up with Milli-Q water and ¢ltered through a 0.22-Wm ¢lter before use (Type GS from Millipore). A bu¡er £ow rate of 1 ml/min was used for all of the chromatography. Two bu¡er systems were used. (i) When the production of Ala^Tyr was being measured, the reaction components were eluted from the C18 column with a 0% to 60% gradient of acetonitrile in 25 mM citrate bu¡er (pH 2.85 in water). (ii) To follow the decrease in concentration of Pyr^Ala^Tyr, a 0% to 40% gradient of acetonitrile in 100 mM phosphate bu¡er of constant ¢nal concentration (pH 6.0 in water) was used [7]. 2.3. High-voltage paper electrophoresis Scission reactions were also analysed by high voltage paper electrophoresis. Samples were applied to Whatman 3 MM chromatography paper and electrophoretic separation of the reaction components was achieved with a potential gradient of 60^100 V/cm, applied for 10^25 min. The bu¡ers used were those of Ambler [14]. All electrophoresis papers were dried in an oven and then stained to detect functional groups. Amino-positive samples were identi¢ed by reaction with ninhydrin, tyrosine residues were detected with the Pauly stain [11], and carbonyl compounds were detected by their formation of a hydrazone with 2,4-dinitrophenylhydrazine [15]. 3. Results The ¢rst conditions used for scission with AdoMet DC were those developed by Stevens and Dixon [7] for scission at neutral pH with phenylene-1,2-diam-
47
Fig. 1. Phenylene-1,2-diamine and its N-substituted derivatives investigated as scission reagents. 1, Phenylene-1,2-diamine; 2, the phosphonomethyl derivative of 1, N-(2-aminophenyl)aminomethylphosphonic acid; 3, the hydroxyethyl derivative of 1,2[(2-aminophenyl)amino]ethanol.
ine. When the enzyme precipitated in the bu¡ers of 1.0 M phosphate (pH 7.0), it became necessary to try di¡erent conditions. The phosphonomethyl and hydroxyethyl derivatives of phenylene-1,2-diamine were synthesized (Fig. 1). The phosphonomethyl derivative was tested because it is soluble in water at pH 7.0 and had given promising results before [9]. The expected reaction is shown in Fig. 2. Reactions with Pyr^Ala^Tyr as the scission substrate were set up in water at pH 7.0, and in neutral phosphate bu¡ers of varying strengths. The results were analysed by electrophoresis. This showed that the concentration of bu¡er in the reaction mixture is an important factor in determining the yield of product Ala^Tyr. A low level of scission occurs in bu¡ers of high ionic strength (greater than 1.0 M phosphate) or in unbuffered solutions that are adjusted to pH 7.0. Some free starting peptide was still present at the end of the reaction time (up to 30 h) and a tyrosine- and phosphate-positive component developed in the reaction mixture during this time. Stevens and Dixon [7] showed that the best bu¡er for scission at neutral pH with phenylene-1,2-diamine is 2.0 M potassium phosphate, pH 7.0. Under these conditions the phosphonomethyl derivative was as good a scission reagent as phenylenediamine itself, but several side products were present after the long incubation and some starting material remained unaltered. In contrast with this, the phosphonomethyl derivative of phenylenediamine gave a high yield of Ala^Tyr, and few side products, when used in bu¡ers of 0.1^0.2 M potassium phosphate at pH 7.0. As paper electrophoresis is not easily quanti¢ed, the system for the analysis of the scission reaction components by HPLC was developed. After incubation at 37³C for 24 h, the reaction mixture was submitted to HPLC (Fig. 3a). Peak I (the largest) was
BBAPRO 35737 2-10-98
48
M. Sunde et al. / Biochimica et Biophysica Acta 1388 (1998) 45^52
Fig. 2. Overall course of the reaction.
the scission reagent itself, the phosphonomethyl derivative of phenylenediamine. Peak II was found to be a breakdown product of this; it formed when the reagent was incubated in the reaction bu¡er without substrate, and there was more of it when the reaction mixture was not £ushed with nitrogen, so it is probably a quinonoid oxidation product. Peak IV was identi¢ed as the desired product, Ala^Tyr, co-running with the known peptide. Peak III appeared in parallel with Ala^Tyr. It represents the quinoxaline derivative that is the ¢nal product formed when Ala^ Tyr is released, 3-methyl-2-oxo-1-(phosphonomethyl)-1,2-dihydroquinoxaline (Fig. 2). It should therefore be formed with this reagent from any pyruvoyl peptide or from pyruvate itself. To con¢rm this identi¢cation, a scission reaction mixture with Pyr^Arg was analysed by HPLC and this peak was indeed present. For further con¢rmation, solid sodium pyruvate was dissolved in 0.2 M phosphate (pH 7.0) with the scission reagent, and after 2 h the mixture was analysed by HPLC and paper electrophoresis. On paper electrophoresis, two phosphate-positive compounds were present, with the quinoxaline slightly less positively charged than the scission reagent at pH 2.0 as would be expected. The HPLC pro¢le showed that after 2 h very little of the oxidation breakdown peak was present but almost all the quinoxaline had formed (Fig. 3b). Peak V probably contains Pyr^Ala^Tyr and some UV-absorbing bufTable 1 Summary of the results of scission reactions with 25 mM phenylene-1,2-diamine and its N-substituted derivatives, in 0.1 and 0.2 M phosphate, pH 7.0, at 37³C Derivative
Conditions
Time (h)
% Yield
Phosphonomethyl Phosphonomethyl Phosphonomethyl Phosphonomethyl Phosphonomethyl Unsubstituted 2-Hydroxyethyl
0.1 0.1 0.1 0.2 0.2 0.1 0.1
24 30 72 24 30 24 24
72 79 89 78 79 30 27
M M M M M M M
phosphate phosphate phosphate phosphate phosphate phosphate phosphate
Fig. 3. HPLC analysis of the scission reaction mixtures. (a) Four major peaks are separated from the reaction mixture of Pyr^Ala^Tyr with the phosphonomethyl derivative of phenylenediamine (Sections 2.2.3 and 2.2.4). Identi¢cations : I, the phosphonomethyl scission reagent (2 in Fig. 1); II, a breakdown product of the scission reagent; III, the quinoxaline derivative formed when Ala^Tyr is released from its reaction complex with the scission reagent (Fig. 2); IV, the desired reaction product, Ala^Tyr. Peak V contains unused Pyr^Ala^Tyr and some UV-absorbing materials. (b) The scission reagent and sodium pyruvate were incubated in 0.2 M potassium phosphate (pH 7.0) at room temperature for 2 h before analysis. The peaks seen correspond to peaks I, III and V in a. After this short incubation, little of the oxidation breakdown peak (peak II) has formed. This experiment con¢rms that peak III is the quinoxaline product of the scission reaction.
fer components (identi¢ed by control sample and bu¡er chromatographic runs). High concentrations of acetonitrile were required to elute Pyr^Ala^Tyr
BBAPRO 35737 2-10-98
M. Sunde et al. / Biochimica et Biophysica Acta 1388 (1998) 45^52
49
Fig. 4. Possible pathways in the scission of 2-oxoacyl peptides with phenylenediamines.
from the column in this bu¡er and UV-absorbing materials present in the acetonitrile prevented the quanti¢cation of the peptide under these conditions. The scission reactions were therefore also analysed with the HPLC system developed by Stevens and Dixon [7], which separates Pyr^Ala^Tyr from a peak that contains both the phosphonomethyl deriv-
ative of phenylene-1,2-diamine and Ala^Tyr. This system con¢rmed that the appearance of product Ala^Tyr and the disappearance of Pyr^Ala^Tyr occur in parallel. It was important to con¢rm that scission, i.e., the release of free Ala^Tyr, took place at pH 7.0, since it was conceivable that a complex between the pyruvo-
BBAPRO 35737 2-10-98
50
M. Sunde et al. / Biochimica et Biophysica Acta 1388 (1998) 45^52
Fig. 5. The maleyl group and its possible analogue. 1, an amide of maleic acid, whose carboxy group catalyses breakage of the amide bond; 2, a possible analogue of 1, likely to be formed by interaction of an oxoacyl group with ammonium carbonate.
yl peptide and reagent formed at pH 7.0 and only broke down to release peptide when exposed to the citrate bu¡er at pH 2.85 of the chromatographic system. He and Tsou [16] reported that the quinoxaline derivative of the N-terminal residue of insulin is released in acid. The scission reaction mixture was analysed by paper electrophoresis at pH 2.0, 3.5 and 6.5. All samples showed the same amount of scission, as judged by the presence of free Ala^Tyr, con¢rming that scission was indeed taking place at pH 7.0. Since the phosphonomethyl derivative of phenylene-1,2-diamine might have given better scission than the unsubstituted compound merely because alkylation made it a secondary amine, the N-hydroxyethyl derivative was also synthesized. The results of scission reactions with phenylene-1,2-diamine and its derivatives are summarized in Table 1. The phosphonomethyl derivative gives the highest levels of scission at neutral pH. With this reagent in 0.1 M phosphate at pH 7.0, 72% of the product was released after 24 h. The reaction continued slowly, with 79% released after 30 h and 89% after 72 h. When the reaction was performed in 0.2 M phosphate at pH 7.0, the release of product was slightly faster but tailed o¡ sooner. After 24 h 78% had been released, and after 30 h this had only increased to 79%. The yield of free peptide product from scission reactions with phenylene-1,2-diamine and the two derivatives was compared after 24 h. The phospho-
Fig. 6. A conceivable mechanism for the scission observed with ammonium or methylammonium carbonate, phosphate or arsenate.
nomethyl derivative was much more e¡ective than the other two: phenylene-1,2-diamine gave 30% and its hydroxyethyl derivative 27% of the yield given by the phosphonomethyl derivative. This last result is only tentative in view of the absence of characterization of the hydroxyethyl compound. 4. Discussion 4.1. Comparison of scission reagents and conditions The results presented here show that the phosphonomethyl derivative of phenylene-1,2-diamine, used in a bu¡er of 0.1^0.2 M phosphate, pH 7.0, is the best reagent so far developed for releasing NH2 ^RP from R^CO^CO^NH^RP at neutral pH. It released up to 78% of product in 24 h at 37³C. van Heyningen [17] reported that phenylenediamine gave 76% scission in 2 M acetic acid, 2 M sodium acetate (pH 4.76), but as Stevens and Dixon [7] found only 58% yield from Pyr^Ala^Tyr and 63% from Pyr^Tyr in these conditions, there seems to be no advantage in using such harshly acidic conditions. The reverse-phase HPLC system run in the citrate bu¡er system has been useful for analysing the scission reaction with model peptides, provided that they contain a chromophore to allow the detection of products. High-voltage paper electrophoresis is more di¤cult to quantify and is also limited by the need for the peptide to contain a residue that is easily detected on paper, whether a free N-terminus or a reactive side chain. 4.2. Possible mechanisms The proposed mechanism of the scission with phenylenediamine is illustrated in Fig. 4. This also illustrates the side-reactions that may slow the rate of productive scission. The reversible formation of the trans imine (2b) would slow scission because only the cis form (2a) of the initial imine has the second nucleophilic group in the correct position for attack on the amide bond. From the product of a 2-oxoacylpeptide with bis(2-aminophenyl)amine [18], it is clear that a ¢ve-membered ring may be formed by attack of the second nucleophilic group at the same carbonyl group as the ¢rst (5). Produc-
BBAPRO 35737 2-10-98
M. Sunde et al. / Biochimica et Biophysica Acta 1388 (1998) 45^52
tive scission must occur through formation of a 2-amino-2-hydroxy-1,2-dihydroquinoxaline (3). The N-quinoxalin-2-yl derivatives (6) of peptides and proteins are reported to be stable in a basic environment [16], so their formation (3C6) might make the conversion of 6 into 4 a slow step in the overall reaction. This, however, only applies if RQNH, and as we ¢nd approximately equal scission with RQNhydroxyethyl as with RQNH, abortive formation of 6 cannot slow the reaction greatly. The phosphonomethyl derivative of phenylene-1,2diamine was originally synthesized [8] in an attempt to provide intramolecularly the apparent requirement for general acid catalysis, and so remove the need for acidic bu¡ers. In fact, however, the need for a bu¡er is not obviated, so at least one step appears to be assisted by intermolecular acid^base catalysis. In this scission (3C4a) a hydron must move from the 2-hydroxy group of 3 to the amide nitrogen. The phosphonomethyl group, with its ^OH and ^O3 groups might catalyse this, just as phosphate can e¡ect a hydron shift between N and O in a concerted process [19]. The phosphono group may be able to act as an acid^base catalyst even when phosphate bu¡er is also necessary for e¡ective reaction, since these two acid^base catalysts may act in di¡erent steps of the pathway. Another possible role for the phosphono group is suggested by the contrast between the slowness of the conversion of the imine (2a) into 4 and the speed of release of RP^NH3 found by Holley and Holley [20] for the compound that di¡ered from 2a merely in having the ^NNC 6 bond reduced to ^NH^CH 6 . There is free rotation around this bond in their compound; this suggests that the slowness is partly due to a lowering of the steady-state concentration of the cis-imine (2a) by formation of abortive trans-imine (2b). The phosphonomethyl group could favour 2a by hydrogen bonding to the amide ^NH^. Paper electrophoresis of the scission reaction mixture with the phosphonomethyl derivative of phenylenediamine showed a tyrosine- and phosphate-positive compound at intermediate stages of the reaction. HPLC, however, showed no tyrosine-positive compound except the starting Pyr^Ala^Tyr and Ala^ Tyr and the sum of the amounts for these was always between 90% and 100%. This di¡erence may be due to the nature of the analysis, where the conditions
51
for the reverse-phase column might decompose some intermediate that was stable in the electrophoretic system. The electrophoresis results show that it is probably not the formation of an imine that is slow, but the slow step might be formation of the e¡ective cis imine. 4.3. Note on an alternative approach Maleyl groups are removed from amide combination by incubations at acid pH [21], and such removal can be very rapid with 2,3-dialkylmaleyl groups [22^24]. This removal depends on placing the carboxy group close to the amide bond that is split (Fig. 5, 1). We therefore wondered if a rapid scission could be obtained by treating a 2-oxoacylpeptide with ammonium carbonate, by formation of a structure such as 2 of Fig. 5, in which carbon dioxide reacts with the imine of the ketone, to give an analogue of the maleyl group. Some preliminary experiments showed that scission was indeed obtained by incubating the 2-oxoacyl peptide with ammonium carbonate, and more rapidly with ammonium phosphate and ammonium arsenate. The ammonium ion could be replaced with methylammonium, but it was less e¡ective, and traces of scission were found with dimethylammonium. The reactions with phosphate and arsenate indicate that the mechanism of Fig. 5 cannot be responsible; Fig. 6 gives a possible mechanism. Up to 80% scission was obtained in a week at 37³C. The best results were obtained with 2 M (NH4 )2 HAsO4 , adjusted to pH 9.5 with ammonia. 4.4. Conclusion Stevens and Dixon [7] developed conditions for scission at neutral pH. Nevertheless, the high bu¡er strengths they used were not tolerated by AdoMet decarboxylase. In this work we have shown that only 0.1^0.2 M phosphate at pH 7 is su¤cient to achieve good levels of scission if phenylene-1,2-diamine is replaced by its phosphonomethyl derivative. These mild conditions, in which AdoMet decarboxylase is stable, should allow removal of an N-terminal 2-oxoacyl group from many proteins, whether this group is present naturally or produced by transamination.
BBAPRO 35737 2-10-98
52
M. Sunde et al. / Biochimica et Biophysica Acta 1388 (1998) 45^52
References [1] H.B.F. Dixon, R. Fields, Speci¢c modi¢cation of NH2 -terminal residues by transamination, Methods Enzymol. 25 (1972) 409^419. [2] P.A. Recsei, E.E. Snell, Pyruvoyl enzymes, Annu. Rev. Biochem. 53 (1984) 357^387. [3] P.D. Van Poelje, E.E. Snell, Pyruvoyl-dependent enzymes, Annu. Rev. Biochem. 59 (1990) 29^59. [4] A.E. Pegg, R. Wechter, A. Panunen, Increase in S-adenosylmethionine decarboxylase in SV-3T3 cells treated with Smethyl-5P-methylthioadenosine, Biochem. J. 244 (1987) 49^ 54. [5] M. Sunde, N-Terminal Modi¢cation of S-adenosylmethionine Decarboxylase, Ph.D. Thesis, University of Cambridge, 1993. [6] H.B.F. Dixon, N-Terminal modi¢cation of proteins ^ a review, J. Protein Chem. 3 (1984) 99^108. [7] J. Stevens, H.B.F. Dixon, The removal of 2-oxoacyl residues from the N-terminus of proteins and cystatin in non-denaturing conditions, Biochim. Biophys Acta 1252 (1995) 195^ 202. [8] H.B.F. Dixon, The preparation of N-(o-aminophenyl)aminomethanephosphonic acid, Biochem. J. 130 (1972) 317. [9] D. Webster, R.E. O¡ord, The modi¢cation and removal of N-terminal residue of trypsin by transamination, Biochem. J. 130 (1972) 315^317. [10] N.S. Radin, D.E. Metzler, Sodium glyoxylate monohydrate, Biochem. Prep. 4 (1955) 60^62. [11] J.L. Bailey, Techniques in Protein Chemistry, 2nd ed., Elsevier, Amsterdam, 1967, pp. 17^26, 353^394. [12] M. Brenner, W. Huber, Herstellung von K-Aminosa«ureestern durch Alkoholyse der Methylester, Helv. Chim. Acta 36 (1953) 1109^1115.
[13] B. Belleau, G. Malek, A new convenient reagent for peptide synthesis, J. Am. Chem. Soc. 90 (1968) 1651^1652. [14] R.P. Ambler, The amino-acid sequence of Pseudomonas cytochrome c-551, Biochem. J. 89 (1963) 349^378. [15] D.E. Bland, Separation of vanillin and syringaldehyde by paper partition chromatography, Nature 164 (1949) 1093. [16] R. He, C.-L. Tsou, Fluorescence of peptide N-terminal 2-oxoacyl and quinoxaline derivatives, Biochem. J. 287 (1992) 1001^1005. [17] S. van Heyningen, Transamination of Proteins, Ph.D. Thesis, University of Cambridge, 1968. [18] Y. Degani, A. Patchornik, An improved chemical cleavage of cysteinyl peptide bonds, Abstracts of the Seventh International Congress of Biochemistry, Tokyo, 1967, No. A-99, p. 603. [19] W.P. Jencks, Catalysis in Chemistry and Enzymology, McGraw-Hill, New York, 1969, pp. 211^217. [20] R.W. Holley, A.D. Holley, A new stepwise degradation of peptides, J. Am. Chem. Soc. 74 (1952) 5445^5448. [21] P.J.G. Butler, J.I. Harris, B.S. Hartley, R. Leberman, Reversible blocking of peptide amino groups with maleic anhydride, Biochem. J. 103 (1967) 78P. [22] H.B.F. Dixon, R.N. Perham, Reversible blocking of amino groups with citraconic anhydride, Biochem. J. 109 (1968) 312^314. [23] A.J. Kirby, P.W. Lancaster, Structure and e¤ciency in intramolecular and enzymic catalysis. Catalysis of amide hydrolysis by the carboxy-group of substituted maleamic acids, J. Chem. Soc. Perkin II (1972) 1206^1214. [24] A.J. Kirby, R.S. McDonald, C.R. Smith, Intramolecular catalysis of amide hydrolysis by two carboxy-groups, J. Chem. Soc. Perkin II (1974) 1495^1504.
BBAPRO 35737 2-10-98
Improved conditions for the removal of 2-oxoacyl groups from the N-terminus of proteins Margaret Sunde 1 , Michael J. Sparkes, Henry B.F. Dixon * Department of Biochemistry, University of Cambridge, Tennis Court Road, Cambridge CB2 1QW, UK Received 7 April 1998; revised 29 June 1998; accepted 2 July 1998
Abstract Proteins with R^CO^CO^NH^ at the N-terminus, rather than the usual R^CH(^NH3 )^CO^NH^, are produced by nonenzymic transamination and also occur in the pyruvoyl enzymes. The oxoacyl group may be specifically removed from a model peptide, in yields of 70^80%, by treating them in 0.1 M phosphate buffer at 37³C for 24 h with 25 mM of the N-phosphonomethyl derivative of phenylene-1,2-diamine. This provides mild conditions for the stepwise removal of N-terminal residues without denaturation. ß 1998 Elsevier Science B.V. All rights reserved. Keywords: Transamination; Scission; N-Terminal residue
1. Introduction The role of N-terminal residues of a protein can be studied by their stepwise removal in non-denaturing conditions. This can be achieved by transamination, which speci¢cally modi¢es terminal amino groups in the reaction R^CH(^NH3 )^CO^NH^RPCR^CO^ CO^NH^RP, followed by scission of the 2-oxoacyl groups formed to yield NH3 ^RP [1]. The scission can be used also to study pyruvoyl enzymes in which the 2-oxoacyl group acts as an electron sink and is essential for catalytic activity [2,3]. We wished to use this scission to remove the pyruvoyl group from human S-adenosylmethionine decarboxylase (AdoMet Abbreviations: Pyr-, pyruvoyl * Corresponding author. Fax: +44-223-333345; E-mail: [email protected] 1 Present address: New Chemistry Laboratory, Oxford Centre for Molecular Sciences, South Parks Road, Oxford OX1 3QT, UK.
DC) [4], so that it could be replaced with other 2-oxoacyl groups that might be stronger electron sinks in the decarboxylation reaction. For this, the amino group liberated by the scission would need to be acylated with a 2-oxo acid. The anions of such acids might be expected to be poor nucleophiles; nevertheless it proved possible [5] to make the esters of pyruvic, 2-oxobutanoic and phenylpyruvic acids with Nhydroxysuccinimide using bis(succinimid-N-yl) sul¢te and pyridine in dimethylformamide. Hence the amino group liberated could be acylated. To make this reaction speci¢c, all other amino groups in the molecule would need either reversible protection or conversion into guanidino groups by amidination. The enzyme, however, proved to be unstable in the standard conditions for scission [1,6], and also in the milder, neutral conditions developed by Stevens and Dixon [7] of phenylene-1,2-diamine in strong phosphate bu¡er. We therefore tested other scission reagents with a model peptide substrate, Pyr^Ala^Tyr, and analysed the progress of the reaction with reverse-
0167-4838 / 98 / $ ^ see front matter ß 1998 Elsevier Science B.V. All rights reserved. PII: S 0 1 6 7 - 4 8 3 8 ( 9 8 ) 0 0 1 6 2 - 9
BBAPRO 35737 2-10-98
46
M. Sunde et al. / Biochimica et Biophysica Acta 1388 (1998) 45^52
phase high-performance liquid chromatography (HPLC). The highest rates of scission were achieved with the N-phosphonomethyl derivative [8,9] of phenylene-1,2-diamine (Fig. 1) in 0.1 M phosphate at pH 7.0 at 37³C; under these conditions AdoMet DC is stable, and they are mild enough to be unlikely to harm transaminated proteins. 2. Materials and methods 2.1. Materials Sodium glyoxylate monohydrate was prepared [10] from glyoxylic acid. Phenylene-1,2-diamine was puri¢ed by recrystallization from water. Peptides were synthesized by coupling N-t-Boc-L-amino acids [11] and L-amino acid methyl esters [12] using EEDQ [13]. All reagents were purchased from Aldrich. 2.2. Methods 2.2.1. Synthesis of scission reagents The new scission reagents (Fig. 1) were derivatives of phenylene-1,2-diamine, the binucleophilic reagent [1] used most commonly until now in the scission reaction. (i) Phosphonomethyl derivative: (N-(2-aminophenyl)aminomethylphosphonic acid). This was synthesized from chloromethylphosphonic acid and phenylene-1,2-diamine, as previously described [8]. After synthesis the material was dissolved in water at pH 8.0. Any remaining phenylene-1,2-diamine was removed by repeated extraction with ethyl acetate. The product crystallized out of solution when the pH was adjusted to 4.1. (ii) Hydroxyethyl derivative: (2-[(2-aminophenyl)amino]ethanol). 1-Fluoro-2-nitrobenzene (2.1 ml, 0.02 mol) was added to 10 ml of ethanolamine (0.1 mol). The mixture was incubated at 37³C and the progress of the reaction was followed by thin-layer chromatography (TLC) on a silica-coated plate, developed in a mixture of toluene and ethyl acetate (9:1). After 3 h the reaction was judged to be complete and 100 ml of water was added. The pH was adjusted to 1.5 with HCl and the product, 2-[(2-nitrophenyl)amino]ethanol, was extracted from the aqueous solution into ethyl acetate (4U50 ml). The
organic extracts were combined and concentrated under vacuum, leaving an orange oil. The 2-[(2-nitrophenyl)amino]ethanol was reduced by re£uxing with tin in 6 M HCl until the liquid was colourless, and residual tin was precipitated by passing H2 S through the solution after adjustment to pH 0.6 with NaOH. The stannous sul¢de was ¢ltered o¡, and the ¢ltrate was dried down. The residue was dissolved in water, adjusted to pH 10.5 with NaOH, and the product was extracted into ethyl acetate and dried down to a brown crystalline solid. Unfortunately, analysis was not performed; the compound is only characterized by its route of synthesis, decolorization during synthesis, homogeneity by TLC, and, as an o-diamine, by giving scission. 2.2.2. Transamination of Ala^Ala^Tyr to produce Pyr^Ala^Tyr Ala^Ala^Tyr was dissolved with sodium glyoxylate monohydrate (1.2:1 ratio) in a solution of 2 M pyridine, 0.8 M acetic acid and 10 mM cupric acetate. The mixture was incubated at 37³C for 1 h and then applied to a sulfonic acid resin in the acid form. The product was washed through with three column volumes of water and the eluate concentrated under vacuum. The ¢nal concentration of the Pyr^Ala^Tyr stock solution was determined by amino-acid analysis in the Protein and Nucleic Acid Facility of the Department of Biochemistry, University of Cambridge. 2.2.3. Conditions for the scission reaction with peptides The ratio of scission reagent to Pyr^Ala^Tyr was 4:1, usually 25 mM reagent and 6.25 mM peptide. Reaction mixtures were saturated with nitrogen and incubated at 37³C. When samples were removed for analysis during the incubation period, nitrogen was bubbled through the solution again. Reverse-phase HPLC was used to follow the course of the scission reaction. The concentration of the stock solution of Pyr^Ala^Tyr was known from amino-acid analysis. The amount of Ala^Tyr released was determined from the peak area of the product on the HPLC trace. A standard curve was produced by injecting known amounts of a solution of Ala^Tyr in phosphate bu¡er prepared from solid Ala^Tyr peptide.
BBAPRO 35737 2-10-98
M. Sunde et al. / Biochimica et Biophysica Acta 1388 (1998) 45^52
2.2.4. Reverse-phase HPLC systems A Varian Model 5000 liquid chromatograph was used with a Kratos SF 769 UV detector and Shimadzu C-R6A Chromatopac data processor. Samples were injected onto the Varian Reverse Phase CH-10 column manually, through a Rheodyne 7125 injector port ¢tted with either a 100- or 500-Wl sample loop. Samples of scission reaction components were diluted 20-fold in 25 mM citrate bu¡er (pH 2.85) before injection onto the column. The elution of active ester reaction components and of peptides and proteins was monitored by absorbance at 280 nm. All HPLC bu¡ers were made up with Milli-Q water and ¢ltered through a 0.22-Wm ¢lter before use (Type GS from Millipore). A bu¡er £ow rate of 1 ml/min was used for all of the chromatography. Two bu¡er systems were used. (i) When the production of Ala^Tyr was being measured, the reaction components were eluted from the C18 column with a 0% to 60% gradient of acetonitrile in 25 mM citrate bu¡er (pH 2.85 in water). (ii) To follow the decrease in concentration of Pyr^Ala^Tyr, a 0% to 40% gradient of acetonitrile in 100 mM phosphate bu¡er of constant ¢nal concentration (pH 6.0 in water) was used [7]. 2.3. High-voltage paper electrophoresis Scission reactions were also analysed by high voltage paper electrophoresis. Samples were applied to Whatman 3 MM chromatography paper and electrophoretic separation of the reaction components was achieved with a potential gradient of 60^100 V/cm, applied for 10^25 min. The bu¡ers used were those of Ambler [14]. All electrophoresis papers were dried in an oven and then stained to detect functional groups. Amino-positive samples were identi¢ed by reaction with ninhydrin, tyrosine residues were detected with the Pauly stain [11], and carbonyl compounds were detected by their formation of a hydrazone with 2,4-dinitrophenylhydrazine [15]. 3. Results The ¢rst conditions used for scission with AdoMet DC were those developed by Stevens and Dixon [7] for scission at neutral pH with phenylene-1,2-diam-
47
Fig. 1. Phenylene-1,2-diamine and its N-substituted derivatives investigated as scission reagents. 1, Phenylene-1,2-diamine; 2, the phosphonomethyl derivative of 1, N-(2-aminophenyl)aminomethylphosphonic acid; 3, the hydroxyethyl derivative of 1,2[(2-aminophenyl)amino]ethanol.
ine. When the enzyme precipitated in the bu¡ers of 1.0 M phosphate (pH 7.0), it became necessary to try di¡erent conditions. The phosphonomethyl and hydroxyethyl derivatives of phenylene-1,2-diamine were synthesized (Fig. 1). The phosphonomethyl derivative was tested because it is soluble in water at pH 7.0 and had given promising results before [9]. The expected reaction is shown in Fig. 2. Reactions with Pyr^Ala^Tyr as the scission substrate were set up in water at pH 7.0, and in neutral phosphate bu¡ers of varying strengths. The results were analysed by electrophoresis. This showed that the concentration of bu¡er in the reaction mixture is an important factor in determining the yield of product Ala^Tyr. A low level of scission occurs in bu¡ers of high ionic strength (greater than 1.0 M phosphate) or in unbuffered solutions that are adjusted to pH 7.0. Some free starting peptide was still present at the end of the reaction time (up to 30 h) and a tyrosine- and phosphate-positive component developed in the reaction mixture during this time. Stevens and Dixon [7] showed that the best bu¡er for scission at neutral pH with phenylene-1,2-diamine is 2.0 M potassium phosphate, pH 7.0. Under these conditions the phosphonomethyl derivative was as good a scission reagent as phenylenediamine itself, but several side products were present after the long incubation and some starting material remained unaltered. In contrast with this, the phosphonomethyl derivative of phenylenediamine gave a high yield of Ala^Tyr, and few side products, when used in bu¡ers of 0.1^0.2 M potassium phosphate at pH 7.0. As paper electrophoresis is not easily quanti¢ed, the system for the analysis of the scission reaction components by HPLC was developed. After incubation at 37³C for 24 h, the reaction mixture was submitted to HPLC (Fig. 3a). Peak I (the largest) was
BBAPRO 35737 2-10-98
48
M. Sunde et al. / Biochimica et Biophysica Acta 1388 (1998) 45^52
Fig. 2. Overall course of the reaction.
the scission reagent itself, the phosphonomethyl derivative of phenylenediamine. Peak II was found to be a breakdown product of this; it formed when the reagent was incubated in the reaction bu¡er without substrate, and there was more of it when the reaction mixture was not £ushed with nitrogen, so it is probably a quinonoid oxidation product. Peak IV was identi¢ed as the desired product, Ala^Tyr, co-running with the known peptide. Peak III appeared in parallel with Ala^Tyr. It represents the quinoxaline derivative that is the ¢nal product formed when Ala^ Tyr is released, 3-methyl-2-oxo-1-(phosphonomethyl)-1,2-dihydroquinoxaline (Fig. 2). It should therefore be formed with this reagent from any pyruvoyl peptide or from pyruvate itself. To con¢rm this identi¢cation, a scission reaction mixture with Pyr^Arg was analysed by HPLC and this peak was indeed present. For further con¢rmation, solid sodium pyruvate was dissolved in 0.2 M phosphate (pH 7.0) with the scission reagent, and after 2 h the mixture was analysed by HPLC and paper electrophoresis. On paper electrophoresis, two phosphate-positive compounds were present, with the quinoxaline slightly less positively charged than the scission reagent at pH 2.0 as would be expected. The HPLC pro¢le showed that after 2 h very little of the oxidation breakdown peak was present but almost all the quinoxaline had formed (Fig. 3b). Peak V probably contains Pyr^Ala^Tyr and some UV-absorbing bufTable 1 Summary of the results of scission reactions with 25 mM phenylene-1,2-diamine and its N-substituted derivatives, in 0.1 and 0.2 M phosphate, pH 7.0, at 37³C Derivative
Conditions
Time (h)
% Yield
Phosphonomethyl Phosphonomethyl Phosphonomethyl Phosphonomethyl Phosphonomethyl Unsubstituted 2-Hydroxyethyl
0.1 0.1 0.1 0.2 0.2 0.1 0.1
24 30 72 24 30 24 24
72 79 89 78 79 30 27
M M M M M M M
phosphate phosphate phosphate phosphate phosphate phosphate phosphate
Fig. 3. HPLC analysis of the scission reaction mixtures. (a) Four major peaks are separated from the reaction mixture of Pyr^Ala^Tyr with the phosphonomethyl derivative of phenylenediamine (Sections 2.2.3 and 2.2.4). Identi¢cations : I, the phosphonomethyl scission reagent (2 in Fig. 1); II, a breakdown product of the scission reagent; III, the quinoxaline derivative formed when Ala^Tyr is released from its reaction complex with the scission reagent (Fig. 2); IV, the desired reaction product, Ala^Tyr. Peak V contains unused Pyr^Ala^Tyr and some UV-absorbing materials. (b) The scission reagent and sodium pyruvate were incubated in 0.2 M potassium phosphate (pH 7.0) at room temperature for 2 h before analysis. The peaks seen correspond to peaks I, III and V in a. After this short incubation, little of the oxidation breakdown peak (peak II) has formed. This experiment con¢rms that peak III is the quinoxaline product of the scission reaction.
fer components (identi¢ed by control sample and bu¡er chromatographic runs). High concentrations of acetonitrile were required to elute Pyr^Ala^Tyr
BBAPRO 35737 2-10-98
M. Sunde et al. / Biochimica et Biophysica Acta 1388 (1998) 45^52
49
Fig. 4. Possible pathways in the scission of 2-oxoacyl peptides with phenylenediamines.
from the column in this bu¡er and UV-absorbing materials present in the acetonitrile prevented the quanti¢cation of the peptide under these conditions. The scission reactions were therefore also analysed with the HPLC system developed by Stevens and Dixon [7], which separates Pyr^Ala^Tyr from a peak that contains both the phosphonomethyl deriv-
ative of phenylene-1,2-diamine and Ala^Tyr. This system con¢rmed that the appearance of product Ala^Tyr and the disappearance of Pyr^Ala^Tyr occur in parallel. It was important to con¢rm that scission, i.e., the release of free Ala^Tyr, took place at pH 7.0, since it was conceivable that a complex between the pyruvo-
BBAPRO 35737 2-10-98
50
M. Sunde et al. / Biochimica et Biophysica Acta 1388 (1998) 45^52
Fig. 5. The maleyl group and its possible analogue. 1, an amide of maleic acid, whose carboxy group catalyses breakage of the amide bond; 2, a possible analogue of 1, likely to be formed by interaction of an oxoacyl group with ammonium carbonate.
yl peptide and reagent formed at pH 7.0 and only broke down to release peptide when exposed to the citrate bu¡er at pH 2.85 of the chromatographic system. He and Tsou [16] reported that the quinoxaline derivative of the N-terminal residue of insulin is released in acid. The scission reaction mixture was analysed by paper electrophoresis at pH 2.0, 3.5 and 6.5. All samples showed the same amount of scission, as judged by the presence of free Ala^Tyr, con¢rming that scission was indeed taking place at pH 7.0. Since the phosphonomethyl derivative of phenylene-1,2-diamine might have given better scission than the unsubstituted compound merely because alkylation made it a secondary amine, the N-hydroxyethyl derivative was also synthesized. The results of scission reactions with phenylene-1,2-diamine and its derivatives are summarized in Table 1. The phosphonomethyl derivative gives the highest levels of scission at neutral pH. With this reagent in 0.1 M phosphate at pH 7.0, 72% of the product was released after 24 h. The reaction continued slowly, with 79% released after 30 h and 89% after 72 h. When the reaction was performed in 0.2 M phosphate at pH 7.0, the release of product was slightly faster but tailed o¡ sooner. After 24 h 78% had been released, and after 30 h this had only increased to 79%. The yield of free peptide product from scission reactions with phenylene-1,2-diamine and the two derivatives was compared after 24 h. The phospho-
Fig. 6. A conceivable mechanism for the scission observed with ammonium or methylammonium carbonate, phosphate or arsenate.
nomethyl derivative was much more e¡ective than the other two: phenylene-1,2-diamine gave 30% and its hydroxyethyl derivative 27% of the yield given by the phosphonomethyl derivative. This last result is only tentative in view of the absence of characterization of the hydroxyethyl compound. 4. Discussion 4.1. Comparison of scission reagents and conditions The results presented here show that the phosphonomethyl derivative of phenylene-1,2-diamine, used in a bu¡er of 0.1^0.2 M phosphate, pH 7.0, is the best reagent so far developed for releasing NH2 ^RP from R^CO^CO^NH^RP at neutral pH. It released up to 78% of product in 24 h at 37³C. van Heyningen [17] reported that phenylenediamine gave 76% scission in 2 M acetic acid, 2 M sodium acetate (pH 4.76), but as Stevens and Dixon [7] found only 58% yield from Pyr^Ala^Tyr and 63% from Pyr^Tyr in these conditions, there seems to be no advantage in using such harshly acidic conditions. The reverse-phase HPLC system run in the citrate bu¡er system has been useful for analysing the scission reaction with model peptides, provided that they contain a chromophore to allow the detection of products. High-voltage paper electrophoresis is more di¤cult to quantify and is also limited by the need for the peptide to contain a residue that is easily detected on paper, whether a free N-terminus or a reactive side chain. 4.2. Possible mechanisms The proposed mechanism of the scission with phenylenediamine is illustrated in Fig. 4. This also illustrates the side-reactions that may slow the rate of productive scission. The reversible formation of the trans imine (2b) would slow scission because only the cis form (2a) of the initial imine has the second nucleophilic group in the correct position for attack on the amide bond. From the product of a 2-oxoacylpeptide with bis(2-aminophenyl)amine [18], it is clear that a ¢ve-membered ring may be formed by attack of the second nucleophilic group at the same carbonyl group as the ¢rst (5). Produc-
BBAPRO 35737 2-10-98
M. Sunde et al. / Biochimica et Biophysica Acta 1388 (1998) 45^52
tive scission must occur through formation of a 2-amino-2-hydroxy-1,2-dihydroquinoxaline (3). The N-quinoxalin-2-yl derivatives (6) of peptides and proteins are reported to be stable in a basic environment [16], so their formation (3C6) might make the conversion of 6 into 4 a slow step in the overall reaction. This, however, only applies if RQNH, and as we ¢nd approximately equal scission with RQNhydroxyethyl as with RQNH, abortive formation of 6 cannot slow the reaction greatly. The phosphonomethyl derivative of phenylene-1,2diamine was originally synthesized [8] in an attempt to provide intramolecularly the apparent requirement for general acid catalysis, and so remove the need for acidic bu¡ers. In fact, however, the need for a bu¡er is not obviated, so at least one step appears to be assisted by intermolecular acid^base catalysis. In this scission (3C4a) a hydron must move from the 2-hydroxy group of 3 to the amide nitrogen. The phosphonomethyl group, with its ^OH and ^O3 groups might catalyse this, just as phosphate can e¡ect a hydron shift between N and O in a concerted process [19]. The phosphono group may be able to act as an acid^base catalyst even when phosphate bu¡er is also necessary for e¡ective reaction, since these two acid^base catalysts may act in di¡erent steps of the pathway. Another possible role for the phosphono group is suggested by the contrast between the slowness of the conversion of the imine (2a) into 4 and the speed of release of RP^NH3 found by Holley and Holley [20] for the compound that di¡ered from 2a merely in having the ^NNC 6 bond reduced to ^NH^CH 6 . There is free rotation around this bond in their compound; this suggests that the slowness is partly due to a lowering of the steady-state concentration of the cis-imine (2a) by formation of abortive trans-imine (2b). The phosphonomethyl group could favour 2a by hydrogen bonding to the amide ^NH^. Paper electrophoresis of the scission reaction mixture with the phosphonomethyl derivative of phenylenediamine showed a tyrosine- and phosphate-positive compound at intermediate stages of the reaction. HPLC, however, showed no tyrosine-positive compound except the starting Pyr^Ala^Tyr and Ala^ Tyr and the sum of the amounts for these was always between 90% and 100%. This di¡erence may be due to the nature of the analysis, where the conditions
51
for the reverse-phase column might decompose some intermediate that was stable in the electrophoretic system. The electrophoresis results show that it is probably not the formation of an imine that is slow, but the slow step might be formation of the e¡ective cis imine. 4.3. Note on an alternative approach Maleyl groups are removed from amide combination by incubations at acid pH [21], and such removal can be very rapid with 2,3-dialkylmaleyl groups [22^24]. This removal depends on placing the carboxy group close to the amide bond that is split (Fig. 5, 1). We therefore wondered if a rapid scission could be obtained by treating a 2-oxoacylpeptide with ammonium carbonate, by formation of a structure such as 2 of Fig. 5, in which carbon dioxide reacts with the imine of the ketone, to give an analogue of the maleyl group. Some preliminary experiments showed that scission was indeed obtained by incubating the 2-oxoacyl peptide with ammonium carbonate, and more rapidly with ammonium phosphate and ammonium arsenate. The ammonium ion could be replaced with methylammonium, but it was less e¡ective, and traces of scission were found with dimethylammonium. The reactions with phosphate and arsenate indicate that the mechanism of Fig. 5 cannot be responsible; Fig. 6 gives a possible mechanism. Up to 80% scission was obtained in a week at 37³C. The best results were obtained with 2 M (NH4 )2 HAsO4 , adjusted to pH 9.5 with ammonia. 4.4. Conclusion Stevens and Dixon [7] developed conditions for scission at neutral pH. Nevertheless, the high bu¡er strengths they used were not tolerated by AdoMet decarboxylase. In this work we have shown that only 0.1^0.2 M phosphate at pH 7 is su¤cient to achieve good levels of scission if phenylene-1,2-diamine is replaced by its phosphonomethyl derivative. These mild conditions, in which AdoMet decarboxylase is stable, should allow removal of an N-terminal 2-oxoacyl group from many proteins, whether this group is present naturally or produced by transamination.
BBAPRO 35737 2-10-98
52
M. Sunde et al. / Biochimica et Biophysica Acta 1388 (1998) 45^52
References [1] H.B.F. Dixon, R. Fields, Speci¢c modi¢cation of NH2 -terminal residues by transamination, Methods Enzymol. 25 (1972) 409^419. [2] P.A. Recsei, E.E. Snell, Pyruvoyl enzymes, Annu. Rev. Biochem. 53 (1984) 357^387. [3] P.D. Van Poelje, E.E. Snell, Pyruvoyl-dependent enzymes, Annu. Rev. Biochem. 59 (1990) 29^59. [4] A.E. Pegg, R. Wechter, A. Panunen, Increase in S-adenosylmethionine decarboxylase in SV-3T3 cells treated with Smethyl-5P-methylthioadenosine, Biochem. J. 244 (1987) 49^ 54. [5] M. Sunde, N-Terminal Modi¢cation of S-adenosylmethionine Decarboxylase, Ph.D. Thesis, University of Cambridge, 1993. [6] H.B.F. Dixon, N-Terminal modi¢cation of proteins ^ a review, J. Protein Chem. 3 (1984) 99^108. [7] J. Stevens, H.B.F. Dixon, The removal of 2-oxoacyl residues from the N-terminus of proteins and cystatin in non-denaturing conditions, Biochim. Biophys Acta 1252 (1995) 195^ 202. [8] H.B.F. Dixon, The preparation of N-(o-aminophenyl)aminomethanephosphonic acid, Biochem. J. 130 (1972) 317. [9] D. Webster, R.E. O¡ord, The modi¢cation and removal of N-terminal residue of trypsin by transamination, Biochem. J. 130 (1972) 315^317. [10] N.S. Radin, D.E. Metzler, Sodium glyoxylate monohydrate, Biochem. Prep. 4 (1955) 60^62. [11] J.L. Bailey, Techniques in Protein Chemistry, 2nd ed., Elsevier, Amsterdam, 1967, pp. 17^26, 353^394. [12] M. Brenner, W. Huber, Herstellung von K-Aminosa«ureestern durch Alkoholyse der Methylester, Helv. Chim. Acta 36 (1953) 1109^1115.
[13] B. Belleau, G. Malek, A new convenient reagent for peptide synthesis, J. Am. Chem. Soc. 90 (1968) 1651^1652. [14] R.P. Ambler, The amino-acid sequence of Pseudomonas cytochrome c-551, Biochem. J. 89 (1963) 349^378. [15] D.E. Bland, Separation of vanillin and syringaldehyde by paper partition chromatography, Nature 164 (1949) 1093. [16] R. He, C.-L. Tsou, Fluorescence of peptide N-terminal 2-oxoacyl and quinoxaline derivatives, Biochem. J. 287 (1992) 1001^1005. [17] S. van Heyningen, Transamination of Proteins, Ph.D. Thesis, University of Cambridge, 1968. [18] Y. Degani, A. Patchornik, An improved chemical cleavage of cysteinyl peptide bonds, Abstracts of the Seventh International Congress of Biochemistry, Tokyo, 1967, No. A-99, p. 603. [19] W.P. Jencks, Catalysis in Chemistry and Enzymology, McGraw-Hill, New York, 1969, pp. 211^217. [20] R.W. Holley, A.D. Holley, A new stepwise degradation of peptides, J. Am. Chem. Soc. 74 (1952) 5445^5448. [21] P.J.G. Butler, J.I. Harris, B.S. Hartley, R. Leberman, Reversible blocking of peptide amino groups with maleic anhydride, Biochem. J. 103 (1967) 78P. [22] H.B.F. Dixon, R.N. Perham, Reversible blocking of amino groups with citraconic anhydride, Biochem. J. 109 (1968) 312^314. [23] A.J. Kirby, P.W. Lancaster, Structure and e¤ciency in intramolecular and enzymic catalysis. Catalysis of amide hydrolysis by the carboxy-group of substituted maleamic acids, J. Chem. Soc. Perkin II (1972) 1206^1214. [24] A.J. Kirby, R.S. McDonald, C.R. Smith, Intramolecular catalysis of amide hydrolysis by two carboxy-groups, J. Chem. Soc. Perkin II (1974) 1495^1504.
BBAPRO 35737 2-10-98