Carboxypeptidase A (PDB ID 1hdu), found as a pancreatic digestive enzyme in Bos Taurus, is a zinc exopeptidase that is the cornerstone protein enzyme for the study of metalloproteases (PDB, 1). It contains 307 residues per subunit, has a molecular weight of 137,580.37 Da, and has an isoelctric point (pI) of 7.47 (Artimo, 1). It is fully crystallized at a pH of 7.5 (PDB, 1). Carboxypeptidase A , when bound a by zinc ion (Zn2+), hydrolyzes the carboxy-terminus (C-terminus) peptide, cleaving it from its polypeptide substrate. Carboxypeptidase A breaks down proteins in the pancreas of a cow into amino acids to later be formed into compulsory proteins for the cow. Carboxypeptidase A also is a key component in the uncovering of the primary structure of all proteins. Due to limitations in substrate binding, it is used in conjunction with carboxypeptidase B and carboxypeptidase Y to cleave the C-terminus of proteins. The rate of cleavage is used to determine which amino acid on the protein substrate was cleaved (Grisham, 115-116). The rate of cleavage of peptides for carboxypeptidase A is near 102 s-1 (Christianson, 62). If a protein is used with only carboxypeptidase A and there is no reaction, the peptide in question is one that carboxypeptidase A cannot cleave, such as proline, arginine, lysine, aspartic acid, and glutamic acid (Grisham, 115-116). At this point, another carboxypeptidase can be used to properly postulate the unknown peptide.
Carboxypeptidase A exists in nature as a homotetramer, four subunits that are all identical. These subunits are separated into two important domains, S’1 and S1, that each have a distinct function in the catalytic cleavage of the C-terminus amino acid (Christianson, 63). Although these domains are not entirely hydrophobic or hydrophilic, each subunit has pertinent residues of both characteristics that carry out a specific function.
The S’1 domain determines the specificity for the substrate. Although carboxypeptidase A is a hydrolytic peptidase, the peptides in S’1 prohibit the cleavage of particular peptides that reside on the C-terminus of a given polypeptide substrate. This is because the S’1 domain uses three main residues, Tyr-248, Arg-145, and Asn-144, and a hydrophobic “pocket” to ensure specificity (Christianson, 63). Specificity of carboxypeptidase A is focused towards the carboxylate moiety of a peptide. When interacting with carboxypeptidase A, the carboxylate of the substrate forms a salt link with the guanidinium moiety of Arg-145. Both Arg-145 and Asn-144 perform in hydrogen bonding with this carboxylate, as well (Christianson, 63). These interactions increase the affinity of the polypeptide substrate for the binding site. Competition in binding at this subunit ensures specificity by tolerating only molecules with such intermolecular interactions to successfully bind.
In accordance with the induced fit model that was postulated by Koshland, when a ligand is bound, the phenolic moiety in Tyr-248 and its associated loop undergo a drastic conformational change to include this residue in S’1 (Cho, 2015). The phenol of this residue “flips down” exposing it to hydrogen bond interactions. This allows Tyr-248 to donate a hydrogen bond to the polypeptide substrate’s C-terminus carboxylate and accept a hydrogen bond from an amide of the preceding peptide on the substrate (Cho 2015). Removal of the ligand, Zn2+, yields an inactive enzyme, showing the importance of the Tyr-248 residue in specificity (Anfinsen, 4). In the absence of a zinc ion ligand, Tyr-248 is unable to change conformation and thus, the reaction will not occur.
Due to the hydrophobic pocket in the S’1 domain, both peptide substrates with aliphatic and aromatic side chains have a higher binding affinity (Christianson, 63). Benzylic or aromatic moieties also exploit this hydrophobic pocket, specifically the tyrosine and phenylalanine peptides, to increase their affinity via “edge to face” aromatic interactions (Jennings, 885).
The S1 domain is the catalytic domain where hydrolysis occurs. Once the polypeptide substrate is successfully bound to the active site, hydrolysis can occur. For hydrolysis to occur, Zn2+ must be present. Because it bears an open coordination sphere, Zn2+ is able to bind to a water molecule and also perform in metallic interactions with the carboxylate in Glu-72 and the nitrogen atoms in His-69 and His-196 (McCall, 1437). Because the zinc ion has a full d orbital, it acts as a Lewis acid and becomes the electrophilic catalyst in this reaction. Due to these properties, it is able to aid in the deprotonation of the water molecule for nucleophilic attack and electrostatically stabilize the negative charges of the carbonyl on the polypeptide substrate.
To begin hydrolysis, the carboxylate anion on Glu-270 deprotonates the water molecule bound to zinc, forming a hydroxide, a more suitable molecule for nucleophilic attack. The hydroxide attacks the carbonyl of the peptide substrate forming a tetrahedral transition state, which is stabilized by ionic interactions from the zinc ion and the hydrogen bonds of the guanidinium moiety of Arg-127 (Cho 2015). Arg-127 also activates the scissile peptide bond via hydrogen bonding. Proton donation by Glu-270 and the desired formation of a more stable neutral carbonyl on the polypeptide substrate causes cleavage of the peptide bond at the C-terminus (Christianson, 68).
The second ligand component for carboxypeptidase A is D-[(amino)carbonyl]phenylalanine. This ligand looks and functions very much like a phenylalanine residue on a polypeptide substrate, however instead of binding to carboxypeptidase A and being cleaved, it binds and inhibits carboxypeptidase A. D-[(amino)carbonyl]phenylalanine utilizes competitive inhibition, having an affinity more than three fold of normal polypeptide substrates, to cease hydrolyzation (PDB, 1). In addition to the hydrogen bonding and ionic interactions that a polypeptide substrate has, the phenyl group on D-[(amino)carbonyl]phenylalanine is fitted into the substrate recognition pocket at the S’1, domain and it also has supplementary hydrogen bonds due to its additional amide moiety, further increasing it affinity for the binding site (Cho 2015).
The secondary structure of each subunit of carboxypeptidase A is 38% alpha helices and 17% beta sheets (PDB, 1). There are, in total, 12 beta sheets strands (PDB) with about an equal amount of parallel and anti-parallel strands (Anfinsen, 9). These beta sheets form a large beta sheet that extends through the center of the molecule and twists by 120 degrees (Anfinsen, 9). This twist is due to the packing of hydrophobic side chains from adjacent strands. On one side of the beta sheet that extends through the molecule there are more alpha helices, while the other side contains the active site, including both S’1 and S1 domains, and the sole disulfide bond between Cys-138 and Cys-161 (Anfinsen, 9). This protein has a comparatively large surface area. 53.4% of all of the side chains, with a preference of aspartic acid and glutamic acid, make contact with non-isolated water molecules, and are considered “outside” (Anfinsen, 10). Many of the key peptide residues are located on random coil. These residues and many others that make up the S’1 and S1 domains mainly project in towards the center of the protein, forming the pocket that is the active site.
Procarboxypeptidase A2, a serine protease found in Homo Sapiens, is very similar to carboxypeptidase A in both structure and function. It contains 401 residues, has a molecular weight of 44,956.60 Da, and has a pI of 5.52 (Artimo, 1). Procarboxypeptidase A2 has an approximate 66% sequence similarity to carboxypeptidase A and has a nearly identical structure (Altschul, 1). PSI-BLAST is a database used to find other proteins with similar primary structure to a given protein. The Dali Server is a database used to find other proteins with similar tertiary structure to a given protein by measuring intramolecular distances given by the Z score. A Z-score above two means the compared proteins have similar tertiary structure. Shown by the results of DALI (Z= 52.0, rmsd= 1.7) (Holm, 1) and protein BLAST (E= 8e-148) (Altschul, 1), these two proteins have an analogous primary and tertiary structure. E-value is calculated by assigning gaps in the total sequence homology. Gaps increase the E-value and homology decreases it. Although these proteins have a similar structure, procarboxypeptidase A2 has only one subunit, whereas carboxypeptidase A has four identical subunits. However, when procarboxypeptidase A2 is superimposed on a subunit of carboxypeptidase A, the differences in their tertiary structures are very subtle (Holm, 1).
Both carboxypeptidase A and procarboxypeptidase A2 are pancreatic zinc C-terminus peptidases, so many of the key residues are conserved to perform the same function under similar conditions. All three residues that interact with the zinc ion, Glu-72, His-69, and His-196, are conserved and reside in the same position. Glu-270 and Arg-127, two residues that are imperative for catalysis, are also conserved. Tyr-248, Arg-145, and Asn-144, three residues essential for substrate specificity, are also all preserved (Catasus, 6651). The first 96 residues on procarboxypeptidase A2 are generally excised from the protein to abdicate its zymogen form, however, additional and different residues in procarboxypeptidase A2 can be explained by the increased specificity of a serine protease in relation to carboxypeptidase A (Garcia-Saez, 6906). Changes in the active site such as a substitution of proline for alanine at position 142 allows for less bulky residues on the substrate, such as the serine, to enter, however, steric hindrance prevents the larger residues, such as phenylalanine, to enter or bind.