The chemical synthesis of small proteins has only recently become feasible with reliability. Progress in this field has depended on the development of the solid phase procedure and the automated synthesizer by Merrifield to reduce the labor and errors associated with repetitive manual procedures. Difficulties in generating pure peptides arose from accumulated byproducts resulting from synthetic difficulties in chain assembly and deprotection of side-chain functionality. Advances in analytical and purification technologies have provided tools for addressing these problems. A variety of strategies have been developed to overcome the synthetic difficulties, and several laboratories have invested sufficiently in the optimization of the technology to produce synthetic small proteins of sufficient purity as to remove any doubts regarding their biological effects (1).
One of the first examples where a small synthetic protein had a major impact on a biological problem of clinical significance was the development of human immunodeficiency virus (HIV) proteinase inhibitors. Although the gene sequence for the 99-residue protein which dimerized to give the active proteinase, was identified when the DNA sequence of the gene for HIV was determined, cloning and expressing of HIV proteinase proved to be exceptionally difficult for pharmaceutical groups, even for those with records of considerable previous success in biotechnology. Synthetic HIV proteinase, a 99-residue homodimer prepared in the laboratory of Stephen Kent (2), was used in the initial determination of HIV proteinase specificity and inhibitor screening. The availability of this synthetic protein certainly expedited the development of HIV proteinase inhibitors used for the treatment of acquired immune deficiency syndrome (AIDS). This example and others would argue that chemical synthesis of small proteins may be more rapid than cloning and expression for some proteins, and it can provide ready access to sufficient quantities (hundreds of milligrams) of protein for screening or biophysical studies. In fact, the first X-ray crystallography structure (3) of HIV proteinase complexed with an inhibitor, MVT-101, used synthetic HIV proteinase and set the record for the largest synthetic molecule whose structure was confirmed by crystallography.
1. Solution Versus Solid Phase
Traditional approaches to peptide synthesis were based on solution methods using convergent synthetic schemes, with isolation and characterization of each intermediate. This provided well-defined products with confidence in their final structures, but at the price of increased labor and losses at each synthetic step in separating product from reagents and byproducts. Until recently, the synthesis of small proteins (60 to 100 residues) has been dominated by fragment condensation in solution with maximal protection of side chains. The desired protein sequence is divided into peptide segments of about 10 residues, which is about the maximum length of peptide readily prepared by stepwise addition in solution. After each reaction, the product is isolated and fully characterized before proceeding with the next reaction. Once the set of fragments has been prepared, they are combined pairwise to generate segments of approximately 20 residues. These are then combined pairwise to generate fragments of approximately 40 residues, and this fragment condensation continues until the desired sequence is obtained. The protecting groups on side chains are then removed to give the fully unprotected peptide chain, which is allowed to fold, resulting in the desired protein. First, the 80-residue protein is divided into eight 10-residue segments that are prepared by stepwise elongation requiring nine coupling and nine deprotection steps, with isolation of each intermediate to give a completely protected fragment with both amino and carboxyl termini protected. Depending on the role of the segment, either its N- or C-terminus is deprotected and reacted with its adjacent fragment. This process is continued until the entire protein is assembled, deprotected, purified, and allowed to fold.
The solid-phase approach utilizes a polymeric protecting group that allows the use of excess reagents to force reactions to near completion by the mass action law and trivial isolation of polymeric product by filtration and washing. Intermediates are not isolated, and purity of the final product depends on complete reaction at each synthetic step and minimization of side reactions during the buildup of the oligomeric peptide and subsequent removal from the polymer with deprotection, to give the desired product. The advent of high-performance liquid chromatography (HPLC) with more sophisticated techniques such as nuclear magnetic resonance (NMR), capillary electrophoresis, and mass spectrometry for purification and characterization of the intermediates and final products allows routine synthesis of peptides in the 50- to 100-residue range. Because of the difficulties in purification of larger peptides and small proteins with only minor differences in structure from byproducts, the unambiguous synthesis of larger peptides and small proteins is best accomplished by assembly of fragments that have been purified and fully characterized. This prevents accumulation of side products with only minor structural differences that can be difficult to remove in the final mixture. Initially, chain assembly was relatively easy to optimize, and the majority of undesirable side products in the final cleavage were due to incomplete deprotection of side chains. Considerable effort over two decades was devoted to understanding the sequence-dependent problems leading to truncated sequences or those missing a residue, but these efforts were hampered by the polymeric support itself, which limited application of the normal methods for characterizing intermediates. This effort has led to the current state of the technology, in which average reaction yields are estimated to be greater than 99.5%. Such yields are essential if multiple sequential chemical reactions are performed without isolation and purification of intermediate products.
Automation of the solid-phase reaction was initiated almost immediately once a viable synthetic scheme for peptides was evolved, and the first automated synthesizer was announced by Merrifield and Stewart in 1965 (4). Continuous development of synthesizers and the associated chemistry allows the automated addition of 75 residues per day to a growing peptide chain (5). In many cases, the repetitive yields are sufficiently high that useful products can be isolated from the synthetic mixture by HPLC when small proteins are prepared. If one desires to be more confident that the observed properties are uniquely determined by the targeted sequence, then a more conservative approach utilizing fragment condensation is still required.
Illustrative are the solid-phase protocols for the two strategies (Boc and Fmoc) commonly used for the synthesis of peptides. The Boc strategy (Fig. 1) is often combined with a 1% to 2% cross-linked polystyrene support and a benzyl ester linkage to the polymer, requiring strong acid such as hydrogen fluoride for deprotection. The procedure favored by most synthetic laboratories uses an acid-labile linkage similar to the p-methoxybenzyl ester linkage of the Wang resin and a base-labile amino protecting group, the fluorenylmethyloxycarbonyl (Fmoc), on the added amino acids (Fig. 2). One can use side-chain protection with similar acid lability to the Wang linkage to give free peptide upon cleavage, or use more stable side-chain protection to give the protected peptide for fragment condensation after purification and characterization. In the latter case, a final deprotection with strong acid such as HF is required.
Figure 1. A common implementation of the BOC strategy for peptide synthesis.
The tydc of chain elongation consists liuc removal with trtfluoroacetic add CTFA), neutralization with abase ttuch as triethylamine (TEA), and coupling of the next Hoc-ami no acul (R2) in the presence of a suitable coupling reagent such as disoprcplycarbodimide <DIC).
![tmp10-122_thumb[1]](http://what-when-how.com/wp-content/uploads/2011/05/tmp10122_thumb1.jpg "tmp10-122_thumb[1]")
Figure 2. The alternative FMOC/Wang support strategy which allows ready synthesis of protected peptide fragments as
The cydc of diain elongation conaata of Fmoc removal with piperidine (Pip), neutrah zatioii with a baa* auch a a t.riet.hylrunine CTBA), and coupling of the next Fmoc-ammo acid (R2) in the presence of a suitable coupling reagent such as dii Bopr o [ilycarbodii ini d e : I)IC >,
3. Coupling Reagents
The choice of one of the numerous reagents to activate the carboxyl group for nucleophilic attack by the amine function, to produce the peptide bond, depends on numerous factors. When coupling peptide fragments, reagents that minimize racemization are preferred; a traditional favorite in solution chemistry for coupling fragments has been the azide group. A common procedure is activation by a combination of a diimide, such as 1,3-diisopropylcarbodiimide (DIC) or 1,3-dicyclohexylcarbodiimide (DCCI), with hydroxybenzotriazole (HOBt), or through the use of an activated derivative of HOBt, such as TBTU, to generate the activated HOBt ester in situ. Incorporation of multiple sequential sterically hindered amino acids (valine, isoleucine, aminoisobutyric acid, etc.) often require the use of special coupling reagents, such as acid fluorides, HATU, etc.) to provide reasonable reaction rates and efficient coupling yields.
4. Minimal Versus Maximal Protection
In order to get selectivity in the reaction, all reactive groups except the two that are desired to react can be protected transiently with chemically stable protecting groups. This is referred to as maximal protection and would be preferred, except for incomplete removal and side reactions associated with deprotection of the fully protected peptide product containing a heterogeneous set of protecting groups. In solid-phase synthesis, assembly of the correct sequence is often quite facile, but incomplete removal of the protecting groups can lead to an intractable mixture. By careful control of reaction conditions (aqueous, pH, etc.), one can eliminate some of the side-chain protection to minimize problems associated with protecting group removal. This is referred to as minimal protection and is preferred if selective acylation of the amino terminus by the carboxyl of the N-terminal peptide can be achieved.
5. Chemical Ligation
Certain groups of chemical functionality have exceptional affinities for selective reaction, particularly in the formation of small rings. Recent developments have led to the synthesis of large peptide fragments by solid-phase synthesis, removal from the polymeric support and of most, if not all, side-chain protecting groups, and then selective coupling of the unprotected fragments based on chemical ligation. The most developed strategy for this approach is the use of an N-terminal cysteine residue with a special reactive C-terminal group on the other peptide. This approach was initially conceived by Kemp in his thiol capture strategy (7) and was reduced to a practically successful general approach by the groups of Kent (8-10) and Tam (11-13). For illustrative purposes, the native chemical ligation procedure of Dawson et al. (9) is described (Fig. 4), because it seems to have had the most practical impact. In this case, solid-phase synthesis is used to prepare two unprotected peptide segments that are combined in aqueous solution. The C-terminal fragment contains an N-terminal cysteine residue, and the C-terminal peptide fragment is prepared as the thioester. The thioester is displaced by the thiolate anion of a Cys residue. If the Cys is N-terminal, an acyl migration through formation of a five-membered ring occurs to generate the desired stable amide bond. If the sulfur atom of a Cys residue within the peptide chain is involved, then the thioether formed is capable of being displaced by other thiols, until the fragment migrates to the N-terminal Cys, when the stable rearrangement can occur. An alternative strategy, using an N-terminal b-bromoalanine of fragment two and the C-terminal thioester of fragment one to give the same covalent thioester intermediate by thioesterification, has been explored by Tam et al. (12)
Figure 4. Native chemical ligation of two unprotected peptide fragment using thiol-thioester exchange with ligation at the N-terminal cysteine residue (9).
While many proteins contain cysteine residues, and some have them optimally spaced for fragment assembly, many do not, and one of the research goals in this area is the extension of these concepts to allow coupling at residues other than the ^-terminus of cysteine residues. Some success has been achieved with ligation at X-Gly and Gly-X sites (14), as well as at X-His sites (15). Another development that bodes well for the chemical synthesis of larger proteins is the adaptation of chemical ligation to solid-phase using agarose as the polymeric support. This should allow repetitive ligation of smaller fragments to generate the larger protein stepwise on the polymer.
6. Applications
One might question why one would want to prepare small proteins when one can easily express the genetic message in a variety of biological systems, including eukaryotic systems that include posttranslational modification. Limitations of normal expression systems to the 20 common amino acids restricts chemical modification of the parent protein for a variety of biophysical and pharmaceutical applications. Incorporation via molecular biology of novel amino acids as spectroscopic probes, or to introduce conformational constraints, is feasible using the suppressor mutation approach pioneered by the Schultz group (16-18). In this case, however, steric limitations of the protein biosynthetic machinery preclude the use of certain unusual amino acids or constrained dipeptides (19), which are readily accessible via organic synthesis. The quantities of protein obtained by this approach are generally small, requiring assay methods (enzymatic or spectroscopic) that are sensitive as well as specific.
6.1. Peptide Libraries
The explosion of combinatorial chemistry in the pharmaceutical industry as a paradigm for drug discovery was foreshadowed in the use of peptide libraries by Geysen (20) to map peptide epitopes or antigenic sites on proteins. Numerous strategies (21-23) to synthesize mixtures of thousands to millions of peptides and allow selection of those with the desired activities (24) have developed over the past 20 years. Combinatorial synthesis of individual compounds for assay has developed rapidly, because it eliminates the problem of deconvolution of the mixture to identify the individual compound responsible for the observed activity. It also simplifies the pharmacological evaluation of compounds, due to concerns over synergy or the inhibition of activity that is potentially possible in the bioassay of mixtures. Nevertheless, the most effective strategy for lead generation and optimization will be determined by the cost and efficiency of the biological screen versus that of synthesis.
