created: 9th September 1997, last updated: 10 October 1998©1997 ABRF

Abstract

Racemization of amino acids during solid-phase synthesis of peptides leads to the formation of side products which are chirally modified peptides. The chiral specificity of enzymes can be exploited to identify the sites of the modifications in these impurities. One such impurity, designated X5, was isolated from the target peptide, Fel-1, and demonstrated to be an optical isomer of Fel-1 by N-terminal sequencing and mass spectrometry. A chymotryptic digest was done on the isolated X5 and Fel-1. The fragments were separated on reversed phase HPLC. Mass spectral data on the fragment from X5 with a different retention time from the analogous fragment of Fel-1 suggested that the modification was in the N-terminal portion of the peptide. Enzymatic digestion by Asp-N protease followed by HPLC of the fragments and mass spectral analysis provided evidence that an aspartic acid at position 5 was a D-amino acid in X5 because that position was not cleaved. These results contributed to the identification of X5 as an optical isomer of Fel-1, with a D-aspartic acid replacing an L-aspartic acid normally present at position 5.

Introduction

Racemization of amino acids, leading to the formation of chirally modified peptide side products, has long been a problem in solid-phase peptide synthesis (1). Protecting groups and activating agents have been developed which minimized this problem but not eliminated it entirely. Diastereomeric peptides are often separable, and therefore detectable, by reversed phase HPLC and other forms of chromatography (2).

Characterization of these products of side reactions is initiated by using mass spectrometry and amino acid sequencing to demonstrate that they are isomers of the desired product. Often the next step has been hydrolyzing the peptide to its constituent amino acids by treatment with acid or by using amino or carboxypeptidases. This is followed by either derivitization with a chirally pure compound and separation of the diastereomeric products or treatment with L or D-amino acid oxidase followed by standard amino acid analysis (3). Though these approaches identify the amino acid which racemized, they cannot identify its position if it occurs more than once in the sequence.

The use of amino acid oxidases is an example of using the chiral specificity of enzymes to identify chirally modified amino acids. However, little other literature exists on the use of specific enzymes for the characterization of chirally modified peptides.

This report demonstrates the utility of other enzymes, chymotrypsin and endoprotease Asp-N, in identifying the modified residue in a diastereomeric peptide side product of synthesis.

The sequence of the peptide in this study was derived from the sequence of Fel d 1, the major allergen produced by domestic house cats (4). This peptide, termed Fel-1, contains some of the major human T cell epitopes of Fel d I.

Materials and Methods

The synthesis of the Fel-1 peptide was performed by t-Boc chemistry on a 200 mmole scale in a semi-automatic mode using a CSBio 936 synthesizer. Amino acids were activated by the addition of 2-(1H-Benzotriazole-1-yl)-1,1,3,3,-tetramethyluronium tetrafluorborate (TBTU) and two equivalents of base prior to addition to the resin. Fel-1 has the following sequence:

Chromatography was run on a Waters 626 HPLC system with a model 490 detector and a column heater set to 35^oC. The buffers were 6 mM hydrochloric acid in water for A and 80% acetonitrile for B. The method used to generate Figure 1 was a 35-46.2 % B gradient on a YMC-ODS-AQ, 4.6x150mm, 3 micrometer, 200 angstrom column. This method was used for the purification of X5 from the bulk peptide for sequencing, for mass spectrometry, and enzymatic digests.

Chymotryptic digests of the peptides were performed with immobilized chymotrypsin (ENZYGEL, Boehringer Mannheim, Indianapolis, IN). Aliquots of 1 mg each of the gel were hydrated and washed with 3 aliquots of 300 µl of 100 mM potassium phosphate buffer and then suspended in 200 µl of buffer. The peptide solutions (100 µl aliquots of ~0.36 to 0.1 mg/ml) were added to the gel suspensions. The digestions were left to proceed at room temperature for 1.3 to 2 hours and then stopped by filtering out the gel. The peptide fragments were chromatographed on the HPLC system used for the X5 collection except with a gradient of 0-50 % buffer B over 30 minutes.

Endoproteinase Asp-N (Sigma, St. Louis, MO) digestions of target peptide and purified X5 were performed. The peptides (20 µg each) were incubated with enzyme (1 µg) in 116 µl of 100 mM Tris (pH 8.5) at 37^oC for 18 hours. The peptide fragments were chromatographed on the HPLC system used for the X5 collection except with a gradient of 0-70 % buffer B over 30 minutes.

The sequence of X5 was determined by analysis on a Perkin-Elmer ABI 477A sequencer. The mass spectrum of X5 and the Asp-N digestion fragments were acquired by matrix-assisted laser desorption time of flight (MALDI-TOF) mass spectrometry on a Fisons TofSpec. The FAB mass spectral analyses of the chymotryptic fragments were conducted at M-Scan, Inc. (West Chester, PA).

Results

The closely eluting impurity, designated X5, is shown in a fraction of Fel-1 in Figure 1. This impurity was isolated in sufficient quantities to analyze it by sequencing and mass spectrometry. The sequence and molecular weight of X5 matched those of the target peptide (data not shown).

Figure 1. Chromatogram of target peptide fraction containing approximately 13% of the X5 impurity.

Chymotryptic digestion of Fel-1 was predicted to yield four peptides of between five and eight amino acids. Figure 2 shows the chromatograms of the digests of target peptide and X5 with chymotrypsin. The digests show four large fragments for each peptide labeled 1, 2, 3, and 4. The maps for Fel-1 and X5 are similar except for the fragments numbered 2, which have a different relative retention time to the other fragments.

The collected fragment 2 fractions from the two digests were analyzed by FAB mass spectrometry. Both fragments had molecular masses of 892.0 daltons. This mass corresponds to the mass predicted for the N-terminal 7 amino acid chymotryptic fragment of the target peptide.

The N-terminal chymotryptic fragment of Fel-1 contains two aspartic acids at positions 3 and 5. The third aspartic acid in Fel-1 is close to the middle of the sequence. Therefore, endoproteinase Asp-N, which cleaves peptide bonds on the N-terminal side of aspartic acids, was used to digest Fel-1 and X5. The chromatograms in Figure 3 show that each peptide produced two major fragments. These were collected and their masses detected by MALDI-TOF mass spectrometry (data not shown).

Figure 3. Chymotryptic maps of the target peptide (Panel A) and X5 (Panel B).

The first fragments, which eluted with the same retention time for both digests (marked A in Figure 3), had equal masses corresponding to the theoretical C-terminal fragment. The masses of the second fragments, which eluted with different retention times (marked B and C in Figure 3), corresponded to the N-terminal half of the peptide minus the N-terminal 2 and 4 amino acids for X5 and the target peptide, respectively. This indicated that the Asp residue 5 in X5 was modified and therefore could not be cleaved by the enzyme.

Fel-1 containing D-Asp at position 5 was synthesized and mixed with the X5 containing fraction shown in Figure 1. As seen in Figure 4, the synthetic [D-Asp5]-Fel-1 peptide coeluted with X5. The synthetic [D-Asp3]-Fel-1 peptide was also synthesized and found to elute on the hydrophobic side of peptide Fel-1.

 

Figure 4. Coinjection of fraction shown in Figure 1 with the [D-Asp5]-Fel-1 peptide.

Discussion

The X5 impurity was isolated and demonstrated to be an isomer of the target peptide Fel-1. Racemization of amino acids from the L form to the D form is known to occur during solid-phase peptide synthesis and was suspected to have caused this isomer formation. Chymotryptic digests of target peptide and purified X5 cleaved the peptides into four fragments each. A retention time difference between one of the fragments of peptide Fel-1 and X5 would indicate that fragment contains the modification. The digests of X5 and target peptide did give one fragment from each which did not coelute and the mass spectral analyses of these both gave the predicted mass of the amino terminal fragments. This narrowed the location of the modification to one of the first seven amino acids.

The amino terminal sequence of Fel-1 contains two aspartic acid residues. Aspartic acid residues are relatively susceptible to isomerization. It was assumed that the enzyme endoproteinase Asp-N, which cleaves peptide bonds on the N-terminal side of aspartic acids, would be chirally specific for the L form of aspartic acid. This enzyme was used to digest Fel-1 and X5. The dipeptides, which would be produced by cleavages at positions 3 and 5, were not expected to give significant peaks on the HPLC traces, and therefore only two large peaks were expected for each peptide. As shown in Figure 3, two major peaks were produced for each peptide, only one of which had the same retention time for both. The fragment from X5 with the unique retention time was found to have a mass which corresponds to the theoretical mass of the N-terminal fragment minus the first 2 amino acids. These results suggest that the Asp at position 5 in X5 was protease resistant due to a chiral modification at that position. These results, and coelution with the synthetic [D-Asp 5]-Fel-1 peptide, identified X5 as Fel-1 in which the aspartic acid at position 5 had been isomerized.

It was later determined that the Boc-Asp used in position 5 had been activated for an extended period of time before addition to the peptide resin. This had allowed significant racemization of that amino acid to occur before coupling. Minimizing the pre-activation time limited the formation of X5 in subsequent syntheses.

The example of X5 illustrates the utility of peptide mapping coupled with mass spectrometry in determining sites of chiral modifications in synthetic peptides. Knowledge of the amino acid sequence of a peptide allows for selection of enzymes which will yield peptide maps that will implicate the peptide segment or specific amino acid residue containing a chiral modification.

References

1. Principles of Peptide Synthesis, M. Bodansky, ed., Springer-Verlag, Berlin, page 190, (1993).

2. Kirby, D. A., Miller, C. L., and Rivier, J. E. Separation of neuropeptide Y diastereomers by high-performance liquid chromatography and capillary zone electrophoresis.J. Chromatography, 648: 257(1993).

3. Imai, K., Fukushima, T., Santa, T., Homma, H., Hamase, K., Sakai, K., and Kato. Analytical chemistry and biochemistry of D-amino acids.M. Biomedical Chromatography, 10: 303(1996) .

4. Counsell, C. M., Bond, J. F., Ohman, J. L., Greenstein, J. L., and Garman, R. D. Definition of the human T-cell epitopes of Fel d 1, the major allergen of the domestic cat. J. Allergy Clin. Immunol., 98: 884 (1996).