C-Terminal Peptide and Protein Sequencing

This invention relates to methodology involving novel coupling reagents for the carboxyl-terminal amino acid sequence analysis of peptides and proteins.

Known carboxyl-terminal (C-terminal) sequencing methodologies are enzymatic physical or chemical. The enzymatic approach is basically a time-course carboxypeptidase procedure. It is limited by differential hydrolysis rates of the involved peptide bonds and by potential unaccessibility of the COOH carboxyl terminus in proteins. The results may include the correct amino acids but in the wrong order and may not extend to more than three to five amino acids.

Physical approaches include mass spectrometry and nuclear magnetic resonance (NMR) and are most suitable for small peptides. Fast atom bombardment--Mass Spectrometry (FAB/MS) sensitivity for determining an entire peptide sequence is in the range of 1-10 nmol and is limited to expensive multisector instruments. Micromolar samples are required for NMR analysis.

Four chemical methods of some interest are known. In 1978 Parkam and Loudon reported a method in which the carboxyamido peptide derivative is treated with bis(1,1 trifluoroacetoxy)iodobenzene to yield a derivative of the amino acid. Free COOH groups were treated with bis-p-nitrophenylphosphoryl azide to generate the carboxyamido derivative through a Curtius rearrangement..sup.1

Loudon and coworkers presented another version of the method which entailed reaction of the COOH terminus with pivaloylhydroxyl amine in the presence of carbodimide to effect a Lossen rearrangement. This method failed to degrade aspartic and glutamine residues..sup.2

The method reported by Stark.sup.3 releases the COOH-terminal amino acid as a thiohydantoin. It entails activation of the COOH group with acetic anhydride, followed by reaction with ammonium thiocyanate and cleavage by acid or base hydrolysis to release the thiohydantoin from the peptide chain.

Hawke reported a modification of the Stark chemistry in which trimethylsilylisothiocyanate is utilized as the coupling reagent..sup.4

Notwithstanding these procedures, there is a continuing substantial need for a generally applicable chemical method for C-terminal sequencing. Such a method would have particular value with respect to, among other things, sequencing N-terminal blocked polypeptides and proteins, verifying the primary protein structures predicted from DNA sequences, providing practical detection of post translational processing of gene products from known codon sequences, and as an aid in the design of oligonucleotide cDNA or gene bank probes.

This invention provides practical C-terminal sequencing methodology utilizing novel phosphoryl(thio)amide coupling reagents which yield arylthiohydantoin cleavage products. As compared with the prior art chemical procedures yielding thiohydantoin, aldehyde or iminohydantoin cleavage products, elaboration of the aryl ring imparts enhanced molar absorptivity to the cleavage product molecule and hence greater sensitivity to detection either in the UV or fluorescent spectra. Practical C-terminal sequencing of nanomole and sub-nanomole peptide and protein samples with positive identification of released amino acids is facilitated.

The novel phosphoryl(thio)amide coupling reagents of this invention are represented by the compounds of Formulae A, B, C, and D in which "PhO" is ##STR1##

The coupling agents of the invention are not known to be reported in the literature. They may be synthesized by chemistry related to that used to produce certain known phosphoryl(thio)amide compounds..sup.5 For example a phosphorylating agent such as ##STR2## may be reacted with a mercaptobenzothiazole or mercaptobenzoxazole to produce the compound of Formula A or B. The reaction proceeds in the manner illustrated by Equation 1: ##STR3##

The synthesis is appropriately accomplished at a temperature from about 20.degree. C. to about 100.degree. C. in an inert solvent in the presence of an amine. Appropriate solvents include dimethyl formamide, dichloromethane, acetonitrile, saturated aliphatic hydrocarbons having from about 5 to about 10 carbon atoms, aromatic hydrocarbons including benzene, toluene, xylene and mesitylene. Tertiary amines which do not react with the phosphorylating agent under the reaction conditions may be utilized. Trialkyl amines having from 1 to about 5 carbon atom alkyl groups are appropriate. Triethylamine is preferred. The reaction time may range from about 15 min. to about 60 min. The reaction is normally complete in about 30 min.

1 millimole of 2-mercaptobenzothiazole was dissolved in 0.5 ml of toluene followed by the addition of 1 millimole of triethylamine. This solution was added to a solution of 1 millimole of diphenylchlorophosphate in 1 ml of toluene heated at 70.degree. C. The mixture was reacted for two hours at 70.degree. C., allowed to cool, filtered, and the filtrate evaporated to yield a yellow solid analyzed for the compound of Formula A. Calculated exact means of the Formula A compound is 400.0231 atomic mass units (amu). The average of several exact mass determinations of the compound produced in Example I as determined by FAB-MS was 400.0256 amu. A yield of 40% to 50% of theoretical was obtained.

This example illustrates the synthesis of the compound of Formula B. 1 millimole of 2-mercaptobenzoxazole was dissolved in 0.5 ml toluene followed by the addition of 1 millimole of triethylamine. This solution was added to a solution of 1 millimole of diphenyldichorophosphate in 1 ml of toluene heated to 70.degree. C. and the mixture was reacted for 30 minutes. The reaction mixture was filtered and the filtrate evaporated leaving a white solid. The yield of Formula B compound from a plurality of such reactions of ranged from 20% to 50% of theoretical. The calculated exact mass for the compound of Formula B is 384.0460 amu. The average of several exact mass determinations of the product of Example II as determined by FAB-MS is 384.0415 amu.

Formula C and D compounds are produced in like manner by reacting 2-mercaptobenzothiazole or 2-mercaptobenzoxazole with ##STR4##

The reaction of dansylnorvaline with the compound of Formula A provides a model for carboxylic acid activation as the first coupling step requires in the C-terminal sequencing methodology of this invention.

As illustrated by Equation 2, in the model reaction the piperidine salt of dansylnorvaline is utilized to permit direct reaction of an amino acid carboxylate with the phosphoryl thiourea of Formula A. ##STR5##

Under the conditions shown by Equation 2, the reaction proceeds at 25.degree. C. in an acetonitrile solvent. Molecular ions of 418 and 503 amu are detected by FAB-MS after 2 minutes reaction time and persist after three hours.

FIG. 2 is a FAB mass spectrum evidencing a molecular ion of 418 amu after the Equation 2 reaction had proceeded for two minutes.

FIG. 1 is a FAB mass spectrum evidencing a molecular ion of 503 amu after the Equation 2 reaction had proceeded for three hours.

These two products are consistent with the carboxyl activation of the amino acid by the compound of Formula A to provide the intermediate activated carboxylate: ##STR6## which decomposes or reacts with the initially present nucleophilic amine piperidine.

Generalized coupling reactions involving the Formula A compound and the subsequent cyclization reaction are illustrated by Equations 3a and 3b. A peptide is depicted from the amino-terminal residue (R.sup.1 through R.sup.3) up to the carboxy terminal residue which is labelled as R.sup.3, although peptides or proteins of any length may be considered. ##STR7##

This example illustrates a peptide coupling reaction with the use of the amino-terminally blocked tripeptide N-dansyltriglycine (DG3).

About 20 nanomoles of N-dansyltriglycine, dissolved in 20 .mu.l (microliters) of acetonitrile, was reacted with about 10-20 .mu.moles of compound A in 15 .mu.moles of triethylamine. The reaction was performed several times at about 25.degree. C. over time courses of from one to 24 hours. FAB-MS data show a molecular ion of 673 amu (see FIG. 3) which is consistent with the triethylamine (TEA) salt of the coupled peptide molecular ion formed pursuant to Equation 4: ##STR8##

This structure indicates the cyclization of the intermediate coupled peptide, i.e., ##STR9## The cyclized product exists as its triethylamine salt and gives rise to the 673 amu molecular ion.

A molecular ion of 348 amu (see FIG. 4) is consistent with a fragmention ##STR10## of the cyclized product.

A FAB mass spectral ion detected at 571 amu, consistent with the parent molecular ion calculated for the coupled peptide (not as its triethylamine salt), was found when N-dansyltriglycine was reacted with a 5,000-fold molar excess of compound A and triethylamine for 30 minutes at 50.degree. C. and subsequent brief (1 to 2 minute) exposure to a dilute aqueous sodium hydroxide solution (0.1N).

Coupling utilizing the novel reagents of this invention, is facilitated by immobilization, preferably covalent, of the protein in known manner to a solid support. Preferred solid supports include derivatized polyvinyldifluoride membranes (Milllipore Immobilon) and inert carriers such as polystyrene beads and controlled pore glass (CPG). Immobilization through nucleophilic groups, e.g., protein lysine residues is preferred. In this way the requirement for pretreatment of the protein with an amino-blocking reagent such as acetic anhydride may be avoided.

The coupling reaction may be accomplished in the presence of various bases. Triethyl amine is preferred. Other amines which may be utilized include trialkyl amines in which the alkyl groups have from one to about 5 carbon atoms, aryl amines such as pyridine or the Aldrich proton sponge.

The coupling reaction rate is a function of reaction temperature. A preferred temperature range is from about 25.degree. C. to about 60.degree. C. Any appropriate solvent may be utilized. Acetonitrile is preferred. N,N-dimethylformamide and dimethylsulfoxide are useful solvents.

Cleavage may be accomplished in known manner, for example, utilizing acetohydroxamic acid,.sup.6 a cation exchange resin or concentrated HCL or dilute NaOH as a cleavage reagent.

Acetohydroxamic acid in a basic medium, preferably triethylamine is preferred. The conditions for acetohydroxamate-assisted cleavage of the protein arylthiohydantoin may be varied to include other organic bases, e.g., pyridine, alkyl-, or arylamines, alternate temperatures, solvent systems, and reaction times. The cleavage reagent is utilized in a mixed organic-aqueous solvent system. Other cleavage reagents and solvent systems which may be utilized include thiolates (mercaptide or thiophenoxide) such as thiophenol, 2-mecaptopyridine, N-acylcysteine, or O-acylmercaptoethanol in basic media such as water-acetonitrile with pH greater than 7 adjusted with pyridine, a triakylamine (TEA), or dilute hydroxide.

The released arylthiohydantoins (or specifically, thiophenoxy or phenoxy-thiohydantoins) are identified by separation and detection with reversed-phase HPLC techniques.

Equation 5 illustrates a generalized cleavage reaction applied to the protein-thiophenoxythiohydantoin product generated by Equation 3b: ##STR11##

The hydrolytic lability of the coupled peptide to release the arylthiohydantoin was evidenced by the FAB-MS molecular ions at 366 and 325 amu which are consistent with the dipeptide, N-dansylglycylglycine, and the thiophenylthiohydantoin of the C-terminal glycyl resiude. The generation of these products is attributed to the presence of water, not rigorously excluded, from the basic coupling reaction mixture.

The ion detected at 366 amu (see FIG. 4) is consistent with the dansyldiglycine peptide which is, in effect, the cleavage product that would result from a small amount of water present in the reaction mixture, i.e., ##STR12##

The ion detected at 325 amu (see FIG. 4) (control reactions also show background at 325 amu) may be contributed in part by the cleaved thiophenylthiohydantoin, i.e., ##STR13##

The C-terminal sequencing methodology of this invention is applicable to protein and peptide samples of all free carboxylic acids and without preclusive restraint consequent from the chain length of amino acid residue. Sequencing pursuant to the methodology of the invention provides improved sensitivity, speed and yield on sequential degradation cycles. A single cycle of degradation can be accomplished in from about 1 to about 2 hours.