The sequence and conformation of human pancreatic procarboxypeptidase A2. cDNA cloning, sequence analysis, and three-dimensional model

Implementation of Pharmacogenetics to Individualize Treatment Regimens for Children with Acute Lymphoblastic Leukemia

Despite major advances in the management and high cure rates of childhood acute lymphoblastic leukemia (ALL), patients still suffer from many drug-induced toxicities, sometimes necessitating dose reduction, or halting of cytotoxic drugs with a secondary risk of disease relapse. In addition, investigators have noted significant inter-individual variability in drug toxicities and disease outcomes, hence the role of pharmacogenetics (PGx) in elucidating genetic polymorphisms in candidate genes for the optimization of disease management. In this review, we present the PGx data in association with main toxicities seen in children treated for ALL in addition to efficacy, with a focus on the most plausible germline PGx variants. We then follow with a summary of the highest evidence drug-gene annotations with suggestions to move forward in implementing preemptive PGx for the individualization of treatment regimens for children with ALL.

Clinical and Genetic Risk Factors for Acute Pancreatitis in Patients With Acute Lymphoblastic Leukemia

PURPOSE

Acute pancreatitis is one of the common causes of asparaginase intolerance. The mechanism is unknown, and genetic predisposition to asparaginase-induced pancreatitis has not been previously identified.

METHODS

To determine clinical risk factors for asparaginase-induced pancreatitis, we studied a cohort of 5,185 children and young adults with acute lymphoblastic leukemia, including 117 (2.3%) who were diagnosed with at least one episode of acute pancreatitis during therapy. A genome-wide association study was performed in the cohort and in an independent case-control group of 213 patients to identify genetic risk factors.

RESULTS

Risk factors associated with pancreatitis included genetically defined Native American ancestry (P < .001), older age (P < .001), and higher cumulative dose of asparaginase (P < .001). No common variants reached genome-wide significance in the genome-wide association study, but a rare nonsense variant rs199695765 in CPA2, encoding carboxypeptidase A2, was highly associated with pancreatitis (hazard ratio, 587; 95% CI, 66.8 to 5166; P = 9.0 × 10(-9)). A gene-level analysis showed an excess of additional CPA2 variants in patients who did versus those who did not develop pancreatitis (P = .001). Sixteen CPA2 single-nucleotide polymorphisms were associated (P < .05) with pancreatitis, and 13 of 24 patients who carried at least one of these variants developed pancreatitis. Biologic functions that were overrepresented by common variants modestly associated with pancreatitis included purine metabolism and cytoskeleton regulation.

CONCLUSION

Older age, higher exposure to asparaginase, and higher Native American ancestry were independent risk factors for pancreatitis in patients with acute lymphoblastic leukemia. Those who inherit a nonsense rare variant in the CPA2 gene had a markedly increased risk of asparaginase-induced pancreatitis.

Chymotrypsin C is a co-activator of human pancreatic procarboxypeptidases A1 and A2

Human digestive carboxypeptidases CPA1, CPA2, and CPB1 are secreted by the pancreas as inactive proenzymes containing a 94-96-amino acid-long propeptide. Activation of procarboxypeptidases is initiated by proteolytic cleavage at the C-terminal end of the propeptide by trypsin. Here, we demonstrate that subsequent cleavage of the propeptide by chymotrypsin C (CTRC) induces a nearly 10-fold increase in the activity of trypsin-activated CPA1 and CPA2, whereas CPB1 activity is unaffected. Other human pancreatic proteases such as chymotrypsin B1, chymotrypsin B2, chymotrypsin-like enzyme-1, elastase 2A, elastase 3A, or elastase 3B are inactive or markedly less effective at promoting procarboxypeptidase activation. On the basis of these observations, we propose that CTRC is a physiological co-activator of proCPA1 and proCPA2. Furthermore, the results confirm and extend the notion that CTRC is a key regulator of digestive zymogen activation.

The substrate specificity of Metarhizium anisopliae and Bos taurus carboxypeptidases A: insights into their use as tools for the removal of affinity tags

Carboxypeptidases may serve as tools for removal of C-terminal affinity tags. In the present study, we describe the expression and purification of an A-type carboxypeptidase from the fungal pathogen Metarhizium anisopliae (MeCPA) that has been genetically engineered to facilitate the removal of polyhistidine tags from the C-termini of recombinant proteins. A complete, systematic analysis of the specificity of MeCPA in comparison with that of bovine carboxypeptidase A (BoCPA) was carried out. Our results indicate that the specificity of the two enzymes is similar but not identical. Histidine residues are removed more efficiently by MeCPA. The very inefficient digestion of peptides with C-terminal lysine or arginine residues, along with the complete inability of the enzyme to remove a C-terminal proline, suggests a strategy for designing C-terminal affinity tags that can be trimmed by MeCPA (or BoCPA) to produce a digestion product with a homogeneous endpoint.

Post-translational modifications of human thrombin-activatable fibrinolysis inhibitor (TAFI): evidence for a large shift in the isoelectric point and reduced solubility upon activation

Thrombin-activable fibrinolysis inhibitor (TAFI) is distinct from pancreatic procarboxypeptidase B in several ways. The enzymatic activity of TAFIa is unstable and decays with a half-life of a few minutes. During this study, we observed that (i) the isoelectric point (pI) of TAFI shifts dramatically from pH 5 toward pH 8 upon activation and (ii) TAFIa is significantly less soluble than TAFI. The structural bases for these observations were investigated by characterizing all post-translational modifications, including attached glycans and disulfide connectivity. The analyses revealed that all five potential N-glycosylation sites were utilized including Asn22, Asn51, Asn63, Asn86 (located in the activation peptide), and Asn219 (located in the catalytic domain). Asn219 was also found in an unglycosylated variant. Four of the glycans, Asn51, Asn63, Asn86, and Asn219 displayed microheterogeneity, while the glycan attached to Asn22 appeared to be homogeneous. In addition, bisecting GlcNAc attached to the trimannose core was detected, suggesting an origin other than the liver. Monosaccharide composition and LC-MS/MS analyses did not produce evidence for O glycosylation. TAFI contains eight cysteine residues, of which two, Cys69 and Cys383, are not involved in disulfides and contain free sulfhydryl groups. The remaining six cystines form disulfides, including Cys156-Cys169, Cys228-Cys252, and Cys243-Cys257. This pattern is homologous to pancreatic procarboxypeptidase B, and it is therefore unlikely that permutations in the cysteine connectivity are responsible for the enzymatic instability. LC-MS/MS analyses covering more than 90% of the TAFI amino acid sequence revealed no additional modifications. When these results are taken together, they suggest that the inherent instability of TAFIa is not caused by post-translational modifications. However, after activation, TAFIa loses 80% of the attached glycans, generating a large shift in pI and a propensity to precipitate. These changes are likely to significantly affect the properties of TAFIa as compared to TAFI.

Molecular cloning and cDNA sequence analysis of carboxypeptidases A1, A2 and B from the Japanese flounder Paralichthys olivaceus

Although pancreatic serine proteases have been cloned in teleosts, no sequence data are currently available on members of the carboxypeptidase (CP) family. Here, we cloned cDNAs coding for two preproCPAs, corresponding to mammalian preproCPA1 and preproCPA2, and one preproCPB from a pancreatic cDNA library of the Japanese flounder, Paralichthys olivaceus. The activation peptides of flounder proCPs completely retained the sequences for inhibition of enzymatic activity of proCPs just like mammalian proCPs. Of 306-309 amino acids in total, 95 amino acids are completely conserved between bovine CPA1 and CPB and flounder CPs. Notably, amino acid residues for Zn(2+) ligands, catalysis and substrate anchoring are completely conserved between flounder and bovine CPs. Three species of flounder preproCPs are all expressed in the pancreas of first feeding larvae.

Expression of a soluble and activatable form of bovine procarboxypeptidase A in Escherichia coli

Bovine pancreatic procarboxypeptidase A has been overexpressed in a soluble and activatable form in Escherichia coli. When the protein was expressed under the control of bacteriophage T7 promoter in E. coli ADA494 (a thioredoxin reductase deficient bacteria), a thioredoxin fusion protein was produced at relatively high level in the cytoplasm (4 mg/L culture medium). Although the recombinant protein essentially accumulated as inclusion bodies, as much as 30% of the fusion protein was recovered in a soluble form at low growth temperature and could therefore be purified to homogeneity in a single-step procedure by metal-affinity chromatography. The recombinant precursor form of bovine carboxypeptidase A was recognized by a monoclonal antibody directed against purified bovine pancreatic carboxypeptidase A. Moreover, upon tryptic activation it gave rise to an enzyme, the N-terminal sequence, molecular size,and specific activity of which were comparable to those of the enzyme derived from the native precursor purified from bovine pancreas.

Protein secondary structure and stability determined by combining exoproteolysis and matrix-assisted laser desorption/ionization time-of-flight mass spectrometry

In the post-genomic era, several projects focused on the massive experimental resolution of the three-dimensional structures of all the proteins of different organisms have been initiated. Simultaneously, significant progress has been made in the ab initio prediction of protein three-dimensional structure. One of the keys to the success of such a prediction is the use of local information (i.e. secondary structure). Here we describe a new limited proteolysis methodology, based on the use of unspecific exoproteases coupled with matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF MS), to map quickly secondary structure elements of a protein from both ends, the N- and C-termini. We show that the proteolytic patterns (mass spectra series) obtained can be interpreted in the light of the conformation and local stability of the analyzed proteins, a direct correlation being observed between the predicted and the experimentally derived protein secondary structure. Further, this methodology can be easily applied to check rapidly the folding state of a protein and characterize mutational effects on protein conformation and stability. Moreover, given global stability information, this methodology allows one to locate the protein regions of increased or decreased conformational stability. All of this can be done with a small fraction of the amount of protein required by most of the other methods for conformational analysis. Thus limited exoproteolysis, together with MALDI-TOF MS, can be a useful tool to achieve quickly the elucidation of protein structure and stability.

Thermodynamic analysis of helix-engineered forms of the activation domain of human procarboxypeptidase A2

Thermodynamic characterization of the activation domain of human procarboxypeptidase A2, ADA2h, and its helix-engineered mutants was carried out by differential scanning calorimetry. The mutants were engineered by changing residues in the exposed face of the two alpha helices in order to increase their stability. At neutral and alkaline pH the three mutants, alpha-helix 1 (M1), alpha-helix 2 (M2) and alpha-helix 1 and alpha-helix 2 (DM), were more stable than the wild-type domain, in the order DM, M2, M1 and wild-type. Under these conditions the CD and NMR spectra of all the variants are very similar, indicating that this increase in stability is not the result of gross structural changes. Calorimetric analysis shows that the stabilizing effect of mutating the water-exposed surfaces of the helices seems to be mainly entropic, because the mutations do not change the enthalpy or the increase in heat capacity of denaturation. The unfolding behavior of all variants changes under acidic conditions: whereas wild-type and M1 have a strong tendency to aggregate, giving rise to a beta conformation upon unfolding, M2 and DM unfold reversibly, M2 being more stable than DM. CD and NMR experiments at pH 3.0 suggest that a region involving residues of the second and third beta strands as well as part of alpha-helix 1 changes its conformation. It seems that the enhanced stability of the altered conformation of M2 and DM reduces the aggregation tendency of ADA2h at acidic pH.

Protein engineering as a strategy to avoid formation of amyloid fibrils

The activation domain of human procarboxypeptidase A2 (ADA2h) aggregates following thermal or chemical denaturation at acidic pH. The aggregated material contains well-defined ordered structures with all the characteristics of the fibrils associated with amyloidotic diseases. Variants of ADA2h containing a series of mutations designed to increase the local stability of each of the two helical regions of the protein have been found to have a substantially reduced propensity to form fibrils. This arises from a reduced tendency of the denatured species to aggregate rather than from a change in the overall stability of the native state. The reduction in aggregation propensity may result from an increase in the stability of local relative to longer range interactions within the polypeptide chain. These findings show that the intrinsic ability of a protein to form amyloid can be altered substantially by protein engineering methods without perturbing significantly its overall stability or activity. This suggests new strategies for combating diseases associated with the formation of aggregated proteins and for the design of novel protein or peptide therapeutics.

Hydrogen exchange monitored by MALDI-TOF mass spectrometry for rapid characterization of the stability and conformation of proteins

Matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF MS) has been used to monitor hydrogen exchange on entire proteins. Two alternative methods have been used to carry out the hydrogen exchange studies, exchanging deuteron (H to D experiments) or proton (D to H experiments). In the former case, the use of a deuterated matrix has made possible to overcome back-exchange problems and attain reproducible results. The methods presented have been used to determine the slow exchange core of the potato carboxypeptidase inhibitor in different folding states, and to differentially compare the activation domain of human procarboxypeptidase A2 versus three site-directed mutants of different conformational stability. In this work, we show that by using MALDI-TOF MS to monitor hydrogen exchange in entire proteins, it is possible to rapidly check the folding state of a protein and characterize mutational effects on protein conformation and stability, while requiring minimal amounts of sample.

Refinement of modelled structures by knowledge-based energy profiles and secondary structure prediction: application to the human procarboxypeptidase A2

Knowledge-based energy profiles combined with secondary structure prediction have been applied to molecular modelling refinement. To check the procedure, three different models of human procarboxypeptidase A2 (hPCPA2) have been built using the 3D structures of procarboxypeptidase A1 (pPCPA1) and bovine procarboxypeptidase A (bPCPA) as templates. The results of the refinement can be tested against the X-ray structure of hPCPA2 which has been recently determined. Regions miss-modelled in the activation segment of hPCPA2 were detected by means of pseudo-energies using Prosa II and modified afterwards according to the secondary structure prediction. Moreover, models obtained by automated methods as COMPOSER, MODELLER and distance restraints have also been compared, where it was found possible to find out the best model by means of pseudo-energies. Two general conclusions can be elicited from this work: (1) on a given set of putative models it is possible to distinguish among them the one closest to the crystallographic structure, and (2) within a given structure it is possible to find by means of pseudo-energies those regions that have been defectively modelled.

Mapping the pro-region of carboxypeptidase B by protein engineering. Cloning, overexpression, and mutagenesis of the porcine proenzyme

The proteolytic processing of pancreatic procarboxypeptidase B to a mature and functional enzyme is much faster than that of procarboxypeptidase A1. This different behavior has been proposed to depend on specific conformational features at the region that connects the globular domain of the pro-segment to the enzyme and at the contacting surfaces on both moieties. A cDNA coding for porcine procarboxypeptidase B was cloned, sequenced, and expressed at high yield (250 mg/liter) in the methylotrophic yeast Pichia pastoris. To test the previous hypothesis, different mutants of the pro-segment at the putative tryptic targets in its connecting region and at some of the residues contacting the active enzyme were obtained. Moreover, the complete connecting region was replaced by the homologous sequence in procarboxypeptidase A1. The detailed study of the tryptic processing of the mutants shows that limited proteolysis of procarboxypeptidase B is a very specific process, as Arg-95 is the only residue accessible to tryptic attack in the proenzyme. A fast destabilization of the connecting region after the first tryptic cut allows subsequent proteolytic processing and the expression of carboxypeptidase B activity. Although all pancreatic procarboxypeptidases have a preformed active site, only the A forms show intrinsic activity. Mutational substitution of Asp-41 in the globular activation domain, located at the interface with the enzyme moiety, as well as removal of the adjacent 310 helix allow the appearance of residual activity in the mutated procarboxypeptidase B, indicating that the interaction of both structural elements with the enzyme moiety prevents the binding of substrates and promotes enzyme inhibition. In addition, the poor heterologous expression of such mutants indicates that the mutated region is important for the folding of the whole proenzyme.

Cloning, expression, and substrate specificity of MeCPA, a zinc carboxypeptidase that is secreted into infected tissues by the fungal entomopathogen Metarhizium anisopliae

To date zinc carboxypeptidases have only been found in animals and actinomycete bacteria. A cDNA clone (MeCPA) for a novel fungal (Metarhizium anisopliae) carboxypeptidase (MeCPA) was obtained by using reverse transcription differential display polymerase chain reaction to identify pathogenicity genes. MeCPA resembles pancreatic carboxypeptidases in being synthesized as a precursor species (418 amino acids) containing a large amino-terminal fragment (99 amino acids). The mature (secreted) form of MeCPA shows closest amino acid identity to human carboxypeptidases A1 (35%) and A2 (37%). MeCPA was expressed in an insect cell line yielding an enzyme with dual A1 + A2 specificity for branched aliphatic and aromatic COOH-terminal amino acids. However, in contrast to the very broad spectrum A + B-type bacterial enzymes, MeCPA lacks B-type activity against charged amino acids. This is predictable as key catalytic residues determining the specificity of MeCPA are conserved with those of mammalian A-type carboxypeptidases. Thus, in evolutionary terms the fungal enzyme is an intermediate between the divergence of A and B forms and the differentiation of the A form into A1 and A2 isoforms. Ultrastructural immunocytochemistry of infected host (Manduca sexta) cuticle demonstrated that MeCPA participates with the concurrently produced endoproteases in procuring nutrients; an equivalent function to digestive pancreatic enzymes.

Cloning, sequencing and functional expression of a cDNA encoding porcine pancreatic preprocarboxypeptidase A1

A full-length cDNA clone coding for porcine pancreatic preprocarboxypeptidase A1 (prePCPA1) was isolated from a cDNA library. The open reading frame (ORF) of the nucleotide sequence was 1260 nt in length and encoded a protein of 419 amino acids (aa). The cDNA included a short signal peptide of 16 aa and a 94 aa-long activation segment. The calculated molecular mass of the mature proenzyme was 45561 Da, in accordance with that of the purified porcine pancreatic PCPA1. The deduced aa sequence of the corresponding enzyme differed from that predicted by the three-dimensional structure by 40 aa, and showed 85% identity and 55% identity to that of procarboxypeptidases A1 and A2, respectively. Moreover the sequence was identical to that of several independent cDNA clones, suggesting that it is the major transcribed gene. No evidence for a second variant was observed in the cDNA library and PCPA2 is apparently absent from the porcine pancreas. The cDNA was expressed in Saccharomyces cerevisiae under the control of the yeast triose phosphate isomerase promoter. The signal peptide of the PCPA protein efficiently directed its secretion into the culture medium (1.5 mg.L-1) as a protein of the predicted size. The recombinant proenzyme was analyzed by immunological and enzymological methods. Its activation behavior was comparable with that of the native form and led to a 35-kDa active enzyme.

Structure of the transition state in the folding process of human procarboxypeptidase A2 activation domain

The transition state for the folding pathway of the activation domain of human procarboxypeptidase A2 (ADA2h) has been analyzed by the protein engineering approach. Recombinant ADA2h is an 81-residue globular domain with no disulfide bridges or cis-prolyl bonds, which follows a two-state folding transition. Its native fold is arranged in two alpha-helices packing against a four-stranded beta-sheet. Application of the protein engineering analysis for 20 single-point mutants spread throughout the whole sequence indicates that the transition state for this molecule is quite compact, possessing some secondary structure and a hydrophobic core in the process of being consolidated. The core (folding nucleus) is made by the packing of alpha-helix 2 and the two central beta-strands. The other two strands, at the edges of the beta-sheet, and alpha-helix 1 seem to be completely unfolded. These results, together with previous analysis of ADA2h with either of its two alpha-helices stabilized through improved local interactions, suggest that alpha-helix 1 does not contribute to the folding nucleus, even though it is partially folded in the denatured state under native conditions. On the other hand, alpha-helix 2 folds partly in the transition state and is part of the folding nucleus. It is suggested that a good strategy to improve folding speed in proteins would be to stabilize the helices that are not folded in the denatured state but are partly present in the transition state. Comparison with other proteins shows that there is no clear relationship between fold and/or size with folding speed and level of structure in the transition state of proteins.

Overexpression of human procarboxypeptidase A2 in Pichia pastoris and detailed characterization of its activation pathway

The cDNA of human procarboxypeptidase A2 has been overexpressed in the methylotrophic yeast Pichia pastoris and secreted into the culture medium by means of the alpha-mating factor signal sequence, yielding a major protein of identical size and N-terminal sequence as the wild-type form. Two other forms containing the proenzyme have also been overexpressed: one of them resulted from an incomplete processing of the signal peptide, whereas the other was a glycosylated derivative. Recombinant procarboxypeptidase A2 was purified to homogeneity, and it was shown that its mature active form displays functional properties similar to those of the enzyme directly isolated from human pancreas. The overall yield was approximately 250 mg of proenzyme or 180 mg of mature enzyme/liter of cell culture. The proteolysis-promoted activation process of the recombinant proenzyme has been studied in detail. During maturation by trypsin, the increase in activity of the enzyme is a rapid and monotonic event, which reflects the rate of the proteolytic release of the inhibitory pro-segment and the weaker nature of its interactions with the enzyme moiety compared with procarboxypeptidases of the A1 type. Three main forms of the pro-segment (96, 94, and 92 amino acids), with no inhibitory capability in the severed state, and a single mature carboxypeptidase A2 are produced during this process. No further proteolysis of these pro-segments by the generated carboxypeptidase A2 occurs, in contrast with observations made in other procarboxypeptidases (A1 and B). This differential behavior is a result of the extreme specificity of carboxypeptidase A2.

Comparative analysis of the sequences and three-dimensional models of human procarboxypeptidases A1, A2 and B

A full-length cDNA clone for preprocarboxypeptidase B from human pancreas has been isolated and sequenced. The open reading frame is 1254 bp in length, encoding a protein of 417 amino acids that includes a leader signal peptide of 15 amino acids and a 95-amino acid-long pro-segment. It contains two differences when compared to the sequence reported for pancreas-specific protein, a human serum marker for acute pancreatitis identified as procarboxypeptidase B. The main difference is a previously unreported Cys at position 138, which is needed for the formation of one of the three disulphide bridges. Sequence alignments between human procarboxypeptidases A1, A2 and B and other known forms show that the most conserved region is the enzyme moiety followed by the globular domain of the pro-segment. The maximum variability is found in the connecting region between moieties. The known three-dimensional structures of procarboxypeptidases from bovine and porcine species have been used to model all three human procarboxypeptidases and also to estimatethe interaction energies between the different parts of the molecules, in an attempt to gain insight into the structural features responsible for the differences observed in the functionality of the proenzymes, particularly in their proteolytic activation pathways. Taken together, the results obtained confirm that the main determinant for the rate and mode of activation of procarboxypeptidases is the strength of the interaction between the enzyme and the globular domain of the pro-segment, the connecting segment playing a complementary role.

Characterisation and preliminary X-ray diffraction analysis of human pancreatic procarboxypeptidase A2

Human procarboxypeptidase A2 has been expressed in a Pichia pastoris heterologous system and purified by hydrophobic interaction and anion exchange chromatographies. The hydrolytic action of carboxypeptidase A2 on peptide substrates with different lengths and residues at the C-terminus was analysed, and a preference towards long substrates with aromatic amino acids in their C-terminal end, particularly tryptophan, was found; with such substrates its activity is similar or higher than that of bovine carboxypeptidase A1. Procarboxypeptidase A2 has been crystallised using a vapour diffusion approach; the crystals obtained belong to the monoclinic system, spacegroup P2(1), and present one procarboxypeptidase A2 molecule per asymmetric unit. The crystals diffract beyond 1.8 A resolution and are suitable for detailed X-ray analysis.

The three-dimensional structure of human procarboxypeptidase A2. Deciphering the basis of the inhibition, activation and intrinsic activity of the zymogen

The three-dimensional structure of human procarboxypeptidase A2 has been determined using X-ray crystallography at 1.8 A resolution. This is the first detailed structural report of a human pancreatic carboxypeptidase and of its zymogen. Human procarboxypeptidase A2 is formed by a pro-segment of 96 residues, which inhibits the enzyme, and a carboxypeptidase moiety of 305 residues. The pro-enzyme maintains the general fold when compared with other non-human counterparts. The globular part of the pro-segment docks into the enzyme moiety and shields the S2-S4 substrate binding sites, promoting inhibition. Interestingly, important differences are found in the pro-segment which allow the identification of the structural determinants of the diverse activation behaviours of procarboxypeptidases A1, B and A2, particularly of the latter. The benzylsuccinic inhibitor is able to diffuse into the active site of procarboxypeptidase A2 in the crystals. The structure of the zymogen-inhibitor complex has been solved at 2.2 A resolution. The inhibitor enters the active site through a channel formed at the interface between the pro-segment and the enzyme regions and interacts with important elements of the active site. The derived structural features explain the intrinsic activity of A1/A2 pro-enzymes for small substrates.

Crystal structure of an oligomer of proteolytic zymogens: detailed conformational analysis of the bovine ternary complex and implications for their activation

The pancreas of ruminants secretes a 100 kDa non-covalent ternary complex of the zymogen of a metalloexopeptidase, carboxypeptidase A, and the proforms of two serine endopeptidases, chymotrypsin C and proteinase E. The crystal structure of the bovine complex has been solved and refined to an R-factor of 0.192 using synchrotron radiation X-ray data to 2.35 A resolution. In this heterotrimeric complex, the 403 residue procarboxypeptidase A takes a central position, with chymotrypsinogen C and proproteinase E attached to different surface sites of it. The procarboxypeptidase A subunit is composed of the active enzyme part and the 94 residue prodomain, similar to the monomeric porcine homologous form. The 251 residue subunit chymotrypsinogen structure, the first solved of an anionic (acidic pI) chymotrypsinogen, exhibits characteristics of both chymotrypsinogen A and elastases, with a potential specificity pocket of intermediate size (to accommodate apolar medium-sized residues) although not properly folded, as in bovine chymotrypsinogen A; this pocket displays a "zymogen triad" characteristic for zymogens of the chymotrypsinogen family, consisting of three non-catalytic residues (one serine, one histidine, and one aspartate) arranged in a fashion similar to the catalytic residues in the active enzymes. Following the traits of this family, the N terminus is clamped to the main molecular body by a disulphide bond, but the close six residue activation segment is completely disordered. The third zymogen, the 253 residue proproteinase E, bears close conformational resemblance to active porcine pancreatic elastase; its specificity pocket is buried, displaying the second "zymogen triad". Its five N-terminal residues are disordered, although the close activation site is fixed to the molecular surface. The structure of this native zymogen displays large conformational differences when compared with the recently solved crystal structure of bovine subunit III, an N-terminally truncated, non-activatable, proproteinase E variant lacking the first 13 residues of the native proenzyme. Most of the prosegment of procarboxypeptidase A and its activation sites are buried in the centre of the oligomer, whilst the activation sites of chymotrypsinogen C and proproteinase E are surface-located and not involved in intra or inter-trimer contacts. This organization confers a functional role to the oligomeric structure, establishing a sequential proteolytic activation for the different zymogens of the complex. The large surface and number of residues involved in the contacts among subunits, as well as the variety of non-bonded interactions, account for the high stability of the native ternary complex.

Toward antibody-directed enzyme prodrug therapy with the T268G mutant of human carboxypeptidase A1 and novel in vivo stable prodrugs of methotrexate

Antibody-directed enzyme prodrug therapy (ADEPT) has the potential of greatly enhancing antitumor selectivity of cancer therapy by synthesizing chemotherapeutic agents selectively at tumor sites. This therapy is based upon targeting a prodrug-activating enzyme to a tumor by attaching the enzyme to a tumor-selective antibody and dosing the enzyme-antibody conjugate systemically. After the enzyme-antibody conjugate is localized to the tumor, the prodrug is then also dosed systemically, and the previously targeted enzyme converts it to the active drug selectively at the tumor. Unfortunately, most enzymes capable of this specific, tumor site generation of drugs are foreign to the human body and as such are expected to raise an immune response when injected, which will limit their repeated administration. We reasoned that with the power of crystallography, molecular modeling and site-directed mutagenesis, this problem could be addressed through the development of a human enzyme that is capable of catalyzing a reaction that is otherwise not carried out in the human body. This would then allow use of prodrugs that are otherwise stable in vivo but that are substrates for a tumor-targeted mutant human enzyme. We report here the first test of this concept using the human enzyme carboxypeptidase A1 (hCPA1) and prodrugs of methotrexate (MTX). Based upon a computer model of the human enzyme built from the well known crystal structure of bovine carboxypeptidase A, we have designed and synthesized novel bulky phenylalanine- and tyrosine-based prodrugs of MTX that are metabolically stable in vivo and are not substrates for wild type human carboxypeptidases A. Two of these analogs are MTX-alpha-3-cyclobutylphenylalanine and MTX-alpha-3-cyclopentyltyrosine. Also based upon the computer model, we have designed and produced a mutant of human carboxypeptidase A1, changed at position 268 from the wild type threonine to a glycine (hCPA1-T268G). This novel enzyme is capable of using the in vivo stable prodrugs, which are not substrates for the wild type hCPA1, as efficiently as the wild type hCPA1 uses its best substrates (i.e. MTX-alpha-phenylalanine). Thus, the kcat/Km value for the wild type hCPA1 with MTX-alpha-phenylalanine is 0.44 microM-1 s-1, and kcat/Km values for hCPA1-T268G with MTX-alpha-3-cyclobutylphenylalanine and MTX-alpha-3-cyclopentyltyrosine are 1.8 and 0.16 microM-1 s-1, respectively. The cytotoxic efficiency of hCPA1-268G was tested in an in vitro ADEPT model. For this experiment, hCPA1-T268G was chemically conjugated to ING-1, an antibody that binds to the tumor antigen Ep-Cam, or to Campath-1H, an antibody that binds to the T and B cell antigen CDw52. These conjugates were then incubated with HT-29 human colon adenocarcinoma cells (which express Ep-Cam but not the Campath 1H antigen) followed by incubation of the cells with the in vivo stable prodrugs. The results showed that the targeted ING-1:hCPA1-T268G conjugate produced excellent activation of the MTX prodrugs to kill HT-29 cells as efficiently as MTX itself. By contrast, the enzyme-Campath 1H conjugate was without effect. These data strongly support the feasibility of ADEPT using a mutated human enzyme with a single amino acid change.