created: 25th May 1998, last updated: 18 February 1999© 1999 ABRF

Abstract

We developed a novel chemistry for C-terminal sequencing of proteins based on derivatization with acetylisothiocyanate (AITC) to yield amino acid thiohydantoins (TH-AAs), and it was used for manual sequencing of biopharmaceutical products on a routine basis. This simple chemistry was automated using a ABI 473A N-terminal sequencer. All the reagents (R1 = trimethylsilylisothiocyanate; R3 = alkaline thiocyanate for cleavage) and solvents required for sequencing were accommodated on the sequencer which was modified to deliver liquid R2 (acetyl chloride) to the reaction vessel. The conversion flask was used for preparing the TH-AAs for analysis by "on-line" HPLC using a graphitized carbon (Hypercarb) column. Results obtained with model proteins and recombinant protein drugs suggest that at least three residues from C-terminus can be easily determined. The C-terminal heterogeneity in rIgGs was determined and the differences in Gly/Lys ratio were consistent with changes observed in the IEF profile of these antibodies. Since the chemistry uses only four reagents delivered to the reaction vessel and three to the conversion flask, we believe that the automated protocol can be easily adapted to any existing N-terminal sequencer.

Introduction

The amino-terminal sequence analysis of protein and peptide based on the Edman degradation has been widely used for protein structure determination. As an orthogonal method, C-terminal sequencing can provide structure information on proteins with blocked N-termini by means of natural modification. In addition, it provides information on the post-translational processing at the carboxy-terminus of gene products and facilitates in the production of more specific probes for gene cloning. Although a number of methods for C-terminal sequencing have been reported, the "thiocyanate method" that was first described in 1926 by Schlack (1) has been the subject of many studies (for review see Ref. 2).

The thiocyanate method involves an activation step using acetic anhydride to activate the carboxylic acid at the C-terminus of protein or peptide to yield a oxazolinone, followed by a derivatization step using an isothiocyanate donor reagent to convert the protein or peptide to a peptidylthiohydantoin (2). The derivatized amino acid is then cleaved by an acid or base to yield a shortened peptide or protein and a thiohydantoin amino acid (TH-AA). Although this chemistry is similar to the Edman chemistry, it is difficult to find suitable activation, derivatization and cleaving reagents which do not modify or cleave the protein test article extensively. Although a number of modifications to the thiocyanate method were reported, recently diphenyl phosphoro-isothiocyanatidate (DPP-ITC) was introduced as an efficient reagent for C-terminal sequence analysis (3). In this method, the protein or peptide sample is treated with diisopropylethylamine to convert the C-terminal carboxylic acid into a carboxylate salt, and then the sample is treated with DPP-ITC followed with pyridine to form a peptidyl isothiocyanate. Cleavage of the derivatized amino acid is performed with sodium trimethylsilanolate solution (3).

We developed a simple novel chemistry for C-terminal sequencing of protein or peptides. In this chemistry, acetylisothiocyanate (AITC) is used in acidic condition as a derivatization reagent to convert the C-terminal amino acid to a thiohydantoin, and an alkaline potassium thiocyanate (KSCN) solution is used to cleave the TH-AAs from proteins (4,5). The AITC chemistry has been used in manual sequencing of biopharmaceutical products on a routine basis for several years in our laboratory. Recently we have automated this chemistry using an Applied Biosystems 473A N-terminal sequencer. The TH-AAs cleaved from the proteins were analyzed by on-line HPLC using a graphitized carbon (Hypercarb) column. TH-AA standards from Hewlett Packard were used for identification and quantitation. However, TH-AA standards can be prepared easily from the epsilon-amino acids by treatment with AITC. This report describes the method, operating conditions for ABI 473 A sequencer, and sequencing results from two representative proteins.

Material and Methods

A. Materials

Acetyl chloride (Ac-Cl), triethylamine (TEA) and b-lactoglobulin (b-Lac) were from Sigma Chemical Company. Acetic anhydride (Ac2O), acetic acid, potassium thiocyanate (KSCN), 4-methylmorpholine and trimethylsilyl isothiocyanate (TMS-ITC) were obtained from Aldrich. C-terminal thiohydantoin amino acid standards in acetonitrile (TH-Stds) were obtained from Hewlett Packard. All other chemicals used were either HPLC or reagent grade. Proteins were purchased from Sigma and dissolved in water or 1% TFA/water.

B. Automated C-terminal sequencing with AITC chemistry

Protein samples (1-5 nmol) were coupled onto the Sequelon-DITC membrane disc (Millipore) as per manufacturer's suggestions, and only a portion of the disc was installed in the glass membrane block for sequencing. Sequencing was performed on an ABI model 473A protein sequencer using a newly created sequencing cycle. Briefly, the cycle consisted of the following major steps: 1) The sample was acetylated with R2 (acetyl chloride/acetic anhydride/acetic acid mixture) at 60^oC for 10 min. 2) The C-terminal residues were derivatized to TH-AAs with multiple sequential delivery of R2 and R1 (TMS-ITC) at 60^oC for 3X for 7 min. 3) After washing the membrane with S2 (methanol), the C-terminal TH-AAs were cleaved from the protein with R3 (KSCN in sodium phosphate solution, pH 11) at 60^oC for 2X for 5 min. The cleavage solution containing TH-AAs was transferred to conversion flask (25^oC), neutralized with X1 (3% acetic acid in methanol), dried, and dissolved in R4 (HPLC solvent A). The solution (20 µl, 17%) was injected for on-line HPLC analysis. After washing the membrane with S2 (methanol), the steps described above were repeated during the second sequencing cycle. Table 1 describes the reagents and solvents used in the sequencer.

C. Separation of TH-AAs by HPLC

TH-AAs released from the C-terminal of proteins were analyzed with on-line HPLC using an ABI 473A sequencer. The TH-AAs were separated on a Hypercarb graphitized carbon column (2.1x100 mm, 7 micron, from Shandon Scientific) at 50^oC using a flow rate of 0.2 ml/min. The following gradient was used: 7% B for 2 min, 7%-58% B over 23 min, and then 80% B wash over 9 min. Solvent A was 0.1% TFA with 0.2% 1-hexanesulfonate in water and solvent B was acetonitrile. The column effluent was monitored at 265 nm and 319 nm (for dehydro TH-Ser and TH-Thr) using two UV detectors connected in series. TH-AA standard mixture from Hewlett Packard was dried and reconstituted with buffer A. Acetylated TH-Lys standard was made from Lys amino acid (in excess) treated with a mixture of R1 and R2 at 60^oC for 20 min.

Results

A. C-terminal sequencing chemistry

The automated C-terminal sequencing method described is based on the novel AITC chemistry (4,5). The overall reaction has two stages (see Figure 1):

 

Figure 1. Reaction scheme for derivatization of protein C-terminal amino acid using acetylisothiocyanate and acetylchloride.

In the first stage, protein or peptide is treated with a mixture of acetyl chloride and trimethylsilyl isothiocyanate (TMS-ITC) in acetonitrile. At this stage, the protein C-termini are activated with acetyl chloride in acidic condition to yield a stable protein carboxyl chloride. At the same time the acetyl chloride is also reacted with TMS-ITC to form acetylisothiocyanate (AITC), which in-turn reacts with C-termini to yield TH derivatives. Indeed, AITC (a fresh mixture of acetyl chloride and TMS-ITC) is used directly in the manual method for convenience. Since this reagent mixture is unstable, the components are installed separately on the ABI 473A sequencer. The reagents (R1 and R2) are delivered separately and sequentially in a repeated manner to achieve mixing of R1 and R2 on the protein or peptide sample. In the second stage, the TH-AAs are cleaved from the C-terminal of the proteins by use of an alkaline thiocyanate (KSCN, pH 11). The yield of TH-AAs is nearly quantitative in the first cycle due to the high efficiency of this chemistry. Since the sequencing chemistry is similar to the Edman degradation, the AITC chemistry can be easily adapted to any N-terminal sequencers.

B. HPLC analysis

  

Figure 2. TH-AA standard mixture (25 pmol each) separated on a graphitized carbon column. Detected at 265 and 319 nm. N-acetylated TH-Lys is indicated by the Ac-Lys. Artifact peaks presented in this mixture are indicated by (*). Elution of unmodified Cys derivative is also indicated.

Figure 2 shows the separation of 19 common amino acids using the graphite column. For each amino acid thiohydantoin, except TH-Ile, a single peak with a characteristic retention time was observed. TH-Ile yielded two peaks. Lysine residue yielded an acetylated thiohydantoin derivative in the current sequencing chemistry, and it eluted immediately after the TH-Lys with free e-amino group. Cysteine and serine residues yielded the same dehydro- thiohydantoin derivative and it had an absorption maximum at 319 nm. TH-Thr was also dehydrated under these conditions and detected at 319 nm. TH-Trp was eluted in the wash in the current HPLC.

C. Automated C-terminal sequencing

We have used ABI model 473A Protein Sequencer with a slight modification for the automation of AITC chemistry. The delivery lines of the cartridge holder were exchanged such that the reagent solutions and gas flowed from the bottom to the top. The line of the R2 was also modified such that the R2 was delivered as a liquid. The total time of the cycle was about an hour. To ensure efficiency of the sequencing chemistry, all the steps including load, deliver, transfer and dry functions in the cycle must be timed and optimized. During the sequencing analysis, R1 and R2 (derivatizing reagents) were delivered separately and sequentially in a repeated manner to achieve mixing thoroughly on the protein sample. The time for derivatization and cleavage steps are usually about 20 min and 10 min respectively. A diverse range of proteins have been analyzed using this automated method, and in our experience three residues from the C-terminal of the proteins can be easily determined. This limitation was found to be due to sample wash-out by >50% at each cleavage step.

 

Figure 3. Automated C-terminal sequencing of lysozyme (approximately 0.4 nmol, without reduction and alkylation). Cycle-1(Leu), cycle-2(Arg) and cycle-3 (Cys) are identified. Leucine recovery in the first cycle is nearly quantitative and Arg in the second cycle is 25%. Cysteine is detected at 319 nm. Artifact peaks are indicated by (*).

 

 

 

 

Figure 4. Automated C-terminal sequencing (first cycle) of a rIgG sample (1 nmol) immobilized on Sequelon-DITC membrane disk. The expected C-terminal residues, Lys and Gly from the heavy chain and Cys from the light chain, are identified. Glycine recovery is nearly quantitative. Artifact peak is indicated by (*).

Figure 3 and 4 show the results of sequencing analysis of lysozyme and rIgG. The expected Lys from the C-terminus of rIgGs was found in trace amounts due to extensive C-terminal processing (typical for IgGs). Other model proteins analyzed by this procedure are described in Table 2 below.

All the results were reproducible using ABI 473A sequencer. The sequencing chemistry uses only four reagents (R1, R2, R3 and S2) to be delivered to the reaction cartridge and three reagents (R4, X1 and S4) to the conversion flask to prepare the TH-AAs for the HPLC analysis. Therefore, we believe that this chemistry can be adapted to any existing N-terminal sequencer.

Discussion

In summary, we have developed a sequencing chemistry based on novel acetylisothiocyanate degradation of C-terminal amino acids. An ABI 473A N-terminal sequencer was used to demonstrate the automation of this chemistry. And therefore, the chemistry can be used in other sequencers designed for N-terminal sequencing. A new HPLC method using graphitized carbon (Hypercarb) column has been developed to identify 19 thiohydantoin-amino acids, except tryptophan which eluted in the column wash. Typically three residues from the C-terminal of the proteins are easily identified. Extended sequencing is not possible at this time due to rather heavy (>50%) wash outs during the cleavage steps. Further optimization of the sequencing method should allow identification of more than three residues.

References

1. Schlack, P., and Kumpf. W. Uber eine neue methode zur Ermitt-lung der Konstitution von peptiden. Z. Physiol. Chem. 154: 125-170 (1926) .

2. Inglis, A. S. Chemical procedures for C-terminal sequencing of peptides and proteins. Anal. Biochem. 195: 183-196 (1991).

3. Bailey, J. M., Nikfarjam, F., Shenoy, N. and Shively, J. E. Automated carboxy-terminal sequence analysis of peptides and proteins using diphenyl phosphoroisothiocyanatidate. Protein Science 1: 1622-1633 (1992).

4. Anumula, K.R. Method for Carboxy Terminal Protein or Peptide Sequencing, US Patent # 5,641,685 (1997).

5. Anumula, K.R, and Sheng Tang. Novel chemistry for sequencing of proteins from carboxyl terminus yields a simple method. FASEB J. 9: A1477 (1995).

John W. Crabb

ajpchem@transit.nyser.net