Subsite interactions are considered to define the stringent specificity of proteases for their natural substrates. To probe this issue in the proteolytic pathways leading to apoptosis we have examined the P(4), P(1) and P(1)' subsite preferences of human caspases 1, 3, 6, 7 and 8, using internally quenched fluorescent peptide substrates containing o-aminobenzoyl (also known as anthranilic acid) and 3-nitro-tyrosine. Previous work has demonstrated the importance of the S(4) subsite in directing specificity within the caspase family. Here we demonstrate the influence of the S(1) and S(1)' subsites that flank the scissile peptide bond. The S(1) subsite, the major specificity-determining site of the caspases, demonstrates tremendous selectivity, with a 20000-fold preference for cleaving substrates containing aspartic acid over glutamic acid at this position. Thus caspases are among the most selective of known endopeptidases. We find that the caspases show an unexpected degree of discrimination in the P(1)' position, with a general preference for small amino acid residues such as alanine, glycine and serine, with glycine being the preferred substituent. Large aromatic residues are also surprisingly well-tolerated, but charged residues are prohibited. While this describes the general order of P(1)' subsite preferences within the caspase family, there are some differences in individual profiles, with caspase-3 being particularly promiscuous. Overall, the subsite preferences can be used to predict natural substrates, but in certain cases the cleavage site within a presumed natural substrate cannot be predicted by looking for the preferred peptide cleavage sites. In the latter case we conclude that second-site interactions may overcome otherwise sub-optimal cleavage sequences.

A solid-phase assay for the complete subsite mapping doftheactive site ofeldop eaes has been developed.A lbrary of resin-bound ptase ubstrates was sytsed both on kieselguhr-supported polyamide resin and on a polyethylene glycol-poly-(NN-dlnethylacryhmlde) copolymer type of resin that alows proteas to diffuse into the interior and perform their catalytic activity. An lic acid and 3-it rosine were used a an efclet donor-acceptor p for the resonance energy transfer. The synthesis was performed in a manual library generator that allows simple wet ming of the bead and ll washing p eus. After treatment wit subtiAsin Carlsberg, beads were c and subjected to peptide seq , affording the preferred s quences, their cleavage bond, and a esiqstimation of the turnover. A satitcal dbution of prefed amino acids was obtain for each subsite. The result was compaed with data from kinetic studies In solution.In early studies of the activity of isolated proteolytic enzymes, polypeptides were subjected to digestion with the enzyme (1) and the preferred cleavage site, if any, was determined by isolation of the fragments and subsequent Edman degradation. This often tedious work provided lead amino acid sequences that could be used for determination of the optimal substrates. These were prepared by chemical synthesis and the kinetic constants were measured by determination ofthe rate ofproduct formation by HPLC (2), NMR spectroscopy (3), or spectrophotometric monitoring when chromogenic or fluorogenic substrates could be used (4, 5). More recently the development of internally quenched fluorogenic substrates of the resonance energy transfer type (6, 7) has facilitated the complete subsite mapping (5,(8)(9)(10)(11)(12) for endoproteases. In these substrates, it is important that efficient long-range energy transfer is observed between the donor and the acceptor to span the entire active site of the endopeptidase, thus minimizing the interaction between the enzyme and the chromophoric probes. The donor-acceptor amino acid pair o-aminobenzamide (ABz)-3-nitrotyrosine [Tyr(NO2)] (13) for which excellent quenching of fluorescence is observed has recently been described (5). This donor-acceptor pair is conveniently introduced in parallel multiple-column peptide synthesis (MCPS) (14) of numerous substrates and has been used for subsite mapping of a variety of proteases (6,(15)(16)(17)(18)(19)(20)(21). However, substantial effort is still required to identify the optimal substrates.The use ofcombinatorial peptide libraries (22) and portionmixing libraries (23, 24) is widely accepted as the method of choice for defining binding motifs and unknown biological activities. The portion-mixing library is particularly convenient for the presentation of millions of substrates to a protease with unknown specificity. The problem is, however, to detect, isolate, and characterize the active substances. Due to the high quantum yield of the ABz group, the fluorescence can easily be observed visually in the absence of Tyr(N02), whereas peptides containi...

No abstract

Three types of beaded polyethylene glycol polyacrylamide copolymers (PEGA) with a high content of polyethylene glycol (PEG) were synthesized by inverse suspension polymerization and characterized for peptide synthesis and with respect to their physical properties. Several peptides of high purity have been synthesized on the resin. The properties which were determined were loading of amino groups, swelling, bead size distribution, porosity, flexibility and compatibility with active biomolecules. A loading of 0.35 mmol/g has been obtained and the swelling was excellent in solvents of various polarities ranging from water to dichloromethane. The 13C-NMR T1-relaxation times of a resin containing a peptide were determined in DMSO-d6 and the resin was found to exhibit a behavior similar to the components in free solution.

The binding site of monoclonal antibody Se155-4, which has been the object of successful crystallographic and antibody-engineering studies, is shown by solid-phase immunoassays to be complementary to a branched trisaccharide, alpha-D-Galp(1-->2) [alpha-D-Abep(1-->3)]-alpha-D-Manp(1, rather than to the tetrasaccharide repeating unit alpha-D-Galp(1-->2) [alpha-D-Abep(1-->3)]-alpha-D-Manp(1-->4) alpha-L-Rhap(1- of the bacterial antigen. Specificity for the 3,6-dideoxy-D-xylo-hexose (3,6-dideoxy-D-galactose) epitope present in Salmonella paratyphi B O-antigens was ensured by screening hybridoma experiments with glycoconjugates derived from synthetic oligosaccharides. Detailed epitope mapping of the molecular recognition by modified and monodeoxy oligosaccharide derivatives showed that complementary surfaces and three antibody-saccharide hydrogen bonds are essential for full binding activity. Both hydroxyl groups of the 3,6-dideoxy-D-galactose residue were obligatory for binding and consistent with the directional nature of their involvement in carbohydrate-protein hydrogen bonds; related tetrasaccharides built from the isomeric 3,6-dideoxyhexoses, 3,6-dideoxy-D-glucose, paratose, and 3,6-dideoxy-D-mannose, tyvelose were not bound by the antibody. Titration microcalorimetry measurements were consistent with the hydrogen-bonding map inferred from the crystal structure and suggest that the displacement of water molecules from the binding site accounts for the favorable entropy that accompanies binding of the native trisaccharide determinant. The protein sequences determined for the antibody VL and VH domains reveal somatic mutation of the VL germ line gene, implying that this antibody-binding site results from a mature antibody response.