Martin Quibell and Tony Johnson
MRC Laboratory of Molecular Biology
Through continued research and refinement of solid phase peptide synthesis techniques, the stepwise synthesis of 50-residue peptides has now become routine. However, for longer sequences, even with near quantitative reactions an accumulation of resin -bound deletion peptides occurs that remain to contaminate the final product. These may prove impossible to separate from the target peptide. An attractive alternative approach towards synthetic proteins is via solid phase segment assembly (1). Here, a large target sequence is synthesized on the solid phase from smaller, purified, fully protected segments. Because deletions will differ by at least one segment (typically 10-20 residues), they should be more readily purified from the target peptide by chromatography.
Two major problems have hampered development of segment assembly techniques: (i) the largely unpredictable and often low solubility of fully protected peptides in solvents necessary for purification and coupling (2) and (ii) the potential epimerization of segments containing chiral carboxyl-terminal residues during activation and coupling (3). Low segment solubility results in slow reaction kinetics, compounding the first-order epimerization of segments containing a chiral carboxyl-terminal residue. Solving the problem of low solubility would have a profound effect on both of these shortcomings.
Usually the problem of low segment solubility is addressed by carefully selecting the protecting groups for the individual segments to improve their solubility, purification, and coupling kinetics. In the most successful cases, relatively short segments are made (less than 15 residues), and segment coupling can take days, even then not reaching completion (4 and references therein). However, a backbone protection strategy removes this hurdle. No optimization is required, and segment solubility is guaranteed. Purification and coupling of backbone-protected segments then becomes a routine process (see below). In addition, the strategy of assembling segments offers, automatically, local site variation within a target sequence through synthesis of segment analogues.
The insolubility of some protected peptide segments also manifests itself during solid phase peptide synthesis as the occurrence of "difficult sequences" (5). Both are characterized by a sudden and largely unpredictable change in conformation from disordered, freely solvated species to ordered and poorly solvated, b-sheet-like, amide-backbone hydrogen bonded structures. Earlier studies concluded that replacing secondary with tertiary amide bonds prevented these ordered structures from forming by removing the potential for hydrogen bonding and led to the development of the TFA-labile N-(2-hydroxy-4 -methoxybenzyl) (Hmb) (6) amide backbone protecting group and its reversibly acid-stable counterpart, N-(2-acetoxy-4 -methoxybenzyl) (AcHmb) (Figure 1) (7). We reasoned that backbone protection would significantly enhance the solubility of protected peptide segments in solution and lead to substantial improvements in ease of purification, characterization, and subsequent use in segment coupling reactions. The following syntheses clearly show that strategically positioning backbone protection effectively eliminates problems associated with poor segment solubility and ensures excellent peptide-resin solvation.
Figure 1: Structure of the N-(2-acetoxy-4-methoxybenzyl) backbone protecting group.
The first example, a 72-residue tat protein from HIV-1 Bru was prepared by the solid phase assembly of 5 segments ( structure I, segments are indicated by numbered arrows, N-backbone protected residues are shown in bold) (8). Where possible, segments were chosen with a carboxyl-terminal glycine residue to eliminate the problem of epimerization during segment coupling. Segments were also chosen to ensure that a freely solvated peptide-resin state (5) was maintained during each on -resin reaction.
structure I
Placement of backbone protection within each segment required careful consideration, and we have devised three simple rules to facilitate this process.
i. Backbone protection need only occur at about every sixth residue, providing a regular spacing of tertiary amide bonds during segment assembly and ensuring good peptide -resin solvation (9).
ii. The position chosen for backbone protection needs to be compatible with Hmb chemistry (6). The only limitation here is that b-branched residues cannot be coupled to Hmb-substituted residues other than glycine. All other backbone-amide bonds can be readily protected.
iii. Backbone protection of susceptible aspartyl amide bonds (e.g., -Asp-Asn-) completely inhibits any deleterious base -catalyzed side reactions, and it may be prudent to protect these where possible (9).
The Fmoc/tBu protecting group strategy was chosen as the basic method for smooth and efficient preparation of fully protected segments using Polyhipe SU 500 support (0.25-0.30 mmol/g) anchored through a 2-chlorotrityl linker, prepared as described recently (4). Backbone-substituted, fully protected segments were cleaved using 0.7% TFA/DCM (dichloromethane), in which all segments were completely soluble. Standard preparative HPLC conditions (Vydac C4 and diphenyl) were used to purify the segments to homogeneity. No solubility problems were encountered, and isolated yields were typically 25-50%. A typical purification is illustrated in Figure 2. The stepwise solid phase assembly of the target protein was performed using two equivalents of each purified segment and equimolar amounts of BOP, HOBt, and DIEA in DMF, the only exception being the coupling of segment 4 (containing the chiral carboxyl -terminal lysine residue) where DIC/HOBt activation in DCM was found to give low epimerization (11) [BOP, (benzotriazol-1 -yloxy)tris(dimethylamino)-phosphonium hexafluorophosphate; HOBt, 1-hydroxybenzotriazole; DIEA, N,N-diisopropyl-ethyl -amine; DMF, N,N-dimethylformamide; DIC, 1,3-diisopropyl -carbodiimide]. All coupling reactions were conducted for 6 hours, and the individual coupling efficiency of each segment was 95-98%. HPLC analysis of the crude and purified target protein is shown in Figure 3; the target protein was isolated at 38.4% overall yield and had the expected mass and amino acid composition.
Figure 2: Analytical HPLC of crude (trace A) and purified (trace B, offset) fully-deprotected segment 2 for HIV-1 tat protein synthesis. Analytical HPLC conditions were column: Vydac phenyl reversed-phase, 4.6 x 250 mm; solvents: 0.1% TFA in water (solvent A), 0.01% TFA, 90% acetonitrile (solvent B); gradient: 80-95% solvent B in 27 min; flow rate: 1.5 ml/min; wavelength: 215 nm. For purification, 1,800 mg of crude segment 2 was dissolved in 20 ml trifluoroethanol:water (3:1) and loaded onto a diphenyl reversed-phase HPLC column (Vydac 219TP1022, 22 x 250 mm). Preparative HPLC conditions were as described above, except the flow rate was 10 ml/min and the gradient was 80-95% solvent B from 2 to 27 min. The sample was injected as fifty 400 ml aliquots, fractions were pooled, neutralized with 0.01 M sodium carbonate, and lyophilized. The purified sample was dissolved in 25 ml dichloromethane, washed with two 500 ml aliquots of saturated NaCl, dried, and re-lyophilized from trifluoroethanol/water. Yield: 820 mg, 166 mmol, 36.5%.
Figure 3: Analytical HPLC of crude (trace A) and purified (trace B, offset) HIV-1 tat protein. This corresponds to structure I prepared by solid-phase assembly of AcHmb backbone -substituted, fully protected segments 1-5 with Acm protecting groups at Cys residues 22, 25, 27, 30, 31, 34, and 37. The HPLC conditions were column: Vydac C4 reversed-phase, 4.6 x 250 mm; solvents: 0.1% TFA in water (solvent A), 0.01% TFA, 90 % acetonitrile (solvent B); gradient: 20-35% solvent B in 27 min; flow rate: 1.5 ml/min; wavelength: 215 nm. Yield: 38.4%.
A second, longer target from the three-repeat region of human tau-2 was prepared following essentially the same procedure as described above (structure II) (4). For this 94 -residue target protein, it was possible to design the segments so that all contained a carboxyl-terminal glycine residue, so BOP /HOBt coupling of segments in DMF was used exclusively. Again, no problems were encountered with segment solubility during synthesis, cleavage, or preparative HPLC purification. After a standard 6-hours coupling, individual segment incorporation ranged from 87-99%. HPLC profiles of the crude and purified target protein are shown in Figure 4; the overall yield for the purified material was 16.7%, and the purified target protein had the expected mass (Figure 5) and amino acid composition.
structure II
Having shown that one of the major limitations of solid phase segment assemblyunpredictable solubility and poor reaction kineticshas been overcome, we investigated the limitations of sequential assembly of segments. Is there an upper limit for the length of assembled, resin-bound segments?
Fully protected, backbone-substituted segment 8, from the human tau-2 synthesis described above, was repetitively coupled to (segment 7-segment 6)-resin to yield a resin-bound, 210 -residue protein (structure III, nine couplings of segment 8).
structure III
After each addition of segment 8, a sample was removed to quantitate coupling by UV absorbance and for MALDI-TOF MS and HPLC analysis of the protein product. After the sixth addition of segment 8, the protein became too insoluble to analyze. However, UV and amino acid composition data indicated good incorporation up to nine consecutive couplings of segment 8. As shown in Table 1 and Figure 6, the crude 126-residue peptide (from five couplings of segment 8) is of excellent quality.
Figure 4: Analytical HPLC of crude (trace A) and purified (trace B, offset) human tau 2 protein. This corresponds to structure II prepared by solid-phase assembly of AcHmb backbone -substituted, fully protected segments 6-13 with Acm protection of Cys 197. The HPLC conditions were column: Vydac C8 reversed-phase, 4.6 x 250 mm; solvents: 0.1% TFA in water (solvent A), 0.01% TFA, 90% acetonitrile (solvent B); gradient: 25-35% solvent B in 27 min; flow rate: 1.5 ml/min; wavelength: 215 nm. Yield: 16.7%.
Figure 5: MALDI-TOF mass analysis of purified human tau 2 protein. This corresponds to structure II with the Acm protecting group still attached to Cys 197. Insulin was used as an internal calibrant. The expected mass for the sample is 10,164.7 Da.
Table 1: MALDITOF mass analysis of intermediates in the preparation of structure III
Mass (Da)
-------------------------------
n Residues Found Expected
-----------------------------------------------------
1 42 4,632 + 2.4 4,631
2 63 7,002 + 5.9 7,005
3 84 9,382 + 4.3 9,378
4 105 11,748 + 17.4 11,750
5 126 14,126 + 16.7 14,123
6 147 *
7 168 *
8 189 *
9 210 *
------------------------------------------------------
n Number of sequential couplings of segment 8
* Too insoluble to analyze
Figure 6: MALDI-TOF mass analysis (panel A) and analytical HPLC (panel B) of crude 126-residue, human tau 2 analogue. This corresponds to structure III with n equal to 5. The expected mass for the sample is 14,123.1 Da. The HPLC conditions were column: Vydac C8 reversed phase, 4.6 x 250 mm; solvents: 0.1% TFA in water (solvent A), 0.01% TFA, 90 % acetonitrile (solvent B); gradient: 10-90% solvent B in 27 min; flow rate: 1.5 ml/min; wavelength: 215 nm.
In conclusion, the use of AcHmb backbone substitution has led to the development of a simple, reliable series of protocols for the synthesis and purification of protected peptide segments. Routine preparative HPLC can be used to obtain backbone -substituted, fully protected segments at high purity. Due to the high solubility of these segments and the freely solvated peptide -resin complex afforded by backbone substitution, coupling efficiency remains high throughout the course of a solid phase protein assembly (4). As yet, no data are available on the size limitations of protein synthesis by this technique, but early indications are that coupling efficiency remains high at more than 120 residues. This, together with recent promising results concerning carboxyl-terminal epimerization (11), opens the way for synthetic peptide chemists to contribute to the fields of protein chemistry and molecular biology.
References
1. Lloyd-Williams, P., Albericio, F. and Giralt, E. Tetrahedron (1993) 49, 11065.
2. Atherton, E., Cameron, L. R., Cammish, L. E., Dryland, A., Goddard, P., Priestly, G. P., Richards, J. D., Sheppard, R. C., Wade, J. D. and Williams, B. J. in "Innovations and Perspectives in Solid Phase Peptide Synthesis, First International Symposium", (Ed., Epton, R.), SPCC (UK) Ltd, Birmingham, England (1990) p. 11.
3. (a) Benoiton, N. L. in "The Peptides: Analysis, Synthesis, Biology, Vol. V", (Eds., E. Gross and J. Meienhofer), (1983) p. 217. and (b) Benoiton, N. L. and Kuroda, K. Int. J. Peptide Protein Res. (1981) 17, 197.
4. Quibell, M., Packman, L. C. and Johnson, T. J. Am. Chem. Soc. (1995) 117, 11656.
5. Kent, S. B. H., Alewood, D., Alewood, P., Baca, M., Jones, A. and Schnollzer, M. in "Innovations and Perspectives in Solid Phase Synthesis, Second International Symposium", (Ed., Epton, R.), Intercept Ltd., Andover UK. (1992) p. 1.
6. Johnson, T., Quibell, M. and Sheppard, R. C. J. Pep. Sci. (1995) 1, 11.
7. Quibell, M., Turnell, W. G. and Johnson, T. Tetrahedron Lett. (1994) 35, 2237.
8. Quibell, M., Packman, L. C. and Johnson, T. J. Chem. Soc. Perkin Trans 1 (1996) in press.
9. Bedford, J., Hyde, C., Johnson, T., Jun, W., Owen, D., Quibell, M and Sheppard, R. C. Int. J. Peptide Protein Res. (1992) 40, 300.
10. Quibell, M., Owen, D., Packman, L. C. and Johnson, T. J. Chem. Soc. Chem. Commun. (1994) 2343.
11. Quibell, M., Packman, L. C. and Johnson, T. J. Chem. Soc. Perkin Trans 1 (1996) in press.
The authors may be contacted at Peptide Therapeutics Ltd., 321 Cambridge Science Park, Milton Road, Cambridge CB4 4WG, United Kingdom, E-mail: tonyjo@peptide.co.uk..
Return to the The ABRF Home Page