Rapid identification of substrates for novel proteases using a combinatorial peptide library.

Fluorogenic substrates for assaying novel proteolytic enzymes could be rapidly identified using an easy, solid-phase combinatorial assay technology. The methodology was validated with leader peptidase of Escherichia coli using a subset of an intramolecularly quenched fluorogenic peptide library. The technique was extended toward the discovery of substrates for a new aspartic protease of pharmaceutical relevance (human napsin A). We demonstrated for the first time known to us that potent fluorogenic substrates can be discovered using extracts of cells expressing recombinant enzyme to screen the peptide library. The straightforward and rapid optimization of protease substrates greatly facilitates the drug discovery process by speeding up the development of high throughput screening assays and thus helps more effective exploitation of the enormous body of information and chemical structures emerging from genomics and combinatorial chemistry technologies.

The spliceosomal snRNP core complex of Trypanosoma brucei: cloning and functional analysis reveals seven Sm protein constituents.

Each of the trypanosome small nuclear ribonucleoproteins (snRNPs) U2, U4/U6, and U5, as well as the spliced leader (SL) RNP, contains a core of common proteins, which we have previously identified. This core is unusual because it is not recognized by anti-Sm Abs and it associates with an Sm-related sequence in the trypanosome small nuclear RNAs (snRNAs). Using peptide sequences derived from affinity-purified U2 snRNP proteins, we have cloned cDNAs for five common proteins of 8.5, 10, 12.5, 14, and 15 kDa of Trypanosoma brucei and identified them as Sm proteins SmF (8.5 kDa), -E (10 kDa), -D1 (12.5 kDa), -G (14 kDa), and -D2 (15 kDa), respectively. Furthermore, we found the trypanosome SmB (T. brucei) and SmD3 (Trypanosoma cruzi) homologues through database searches, thus completing a set of seven canonical Sm proteins. Sequence comparisons of the trypanosome proteins revealed several deviations in highly conserved positions from the Sm consensus motif. We have identified a network of specific heterodimeric and -trimeric Sm protein interactions in vitro. These results are summarized in a model of the trypanosome Sm core, which argues for a strong conservation of the Sm particle structure. The conservation extends also to the functional level, because at least one trypanosome Sm protein, SmG, was able to specifically complement a corresponding mutation in yeast.

Mass spectrometry: a tool for the identification of proteins separated by gels.

Mass spectrometry (MS) has become the technique of choice to identify proteins. This has been largely accomplished by the combination of high-resolution two-dimensional (2-D) gel separation with robotic sample preparation, automated MS measurement, data analysis, and database query. Developments during the last five years in MS associated with protein gel separation are reviewed.

Purification and characterization of active recombinant human napsin A.

Recombinant human napsin A expressed in human embryonic kidney 293 cells was purified to homogeneity by a single-step procedure using part of napsin A propeptide as affinity ligand. N-Terminal amino-acid sequencing of the purified enzyme identified the mature form of napsin A. Treatment of purified napsin A with endoglycosidases F and H resulted in a decrease in its molecular mass from 39 kDa to approximately 37 kDa, confirming that napsin A is glycosylated. The kinetic properties were analyzed by using two fluorogenic synthetic substrates K(Dabsyl)-TSLLMAAPQ-Lucifer yellow (DS1) and K(Dabsyl)-TSVLMAAPQ-Lucifer yellow (DS3). The Km values obtained were 1.7 microM and 6.2 microM, respectively. A substrate-specificity study using a napsin A-targeted peptide library confirmed the preference of napsin A for hydrophobic residues at positions P1 and P1'. Adjacent positions, P2-P4 and P2'-P4', appeared less restricted in distribution of amino acids. A pH optimum between 4.0 and 5.5 at room temperature was determined. The purified enzyme was fully active for more than 10 h at pH 5.0 and 6.0, while a half-life of 4 h was determined at pH 7.0 and 37 degrees C.

Two-dimensional map of the proteome of Haemophilus influenzae.

We have constructed a two-dimensional database of the proteome of Haemophilus influenzae, a bacterium of medical interest of which the complete genome, comprising about 1742 open reading frames, has been sequenced. The soluble protein fraction of the microorganism was analyzed by two-dimensional electrophoresis, using immobilized pH gradient strips of various pH regions, gels with different acrylamide concentrations and buffers with different trailing ions. In order to visualize low-copy-number gene products, we employed a series of protein extraction and sample application approaches and several chromatographic steps, including heparin chromatography, chromatofocusing and hydrophobic interaction chromatography. We have also analyzed the cell envelope-bound protein fraction using either immobilized pH gradient strips or a two-detergent system with a cationic detergent in the first and an anionic detergent in the second-dimensional separation. Different proteins (502) were identified by matrix-assisted laser desorption/ionization mass spectrometry and amino acid composition analysis. This is at present one of the largest two-dimensional proteome databases.

Biophysical characterization of fungal phytases (myo-inositol hexakisphosphate phosphohydrolases): molecular size, glycosylation pattern, and engineering of proteolytic resistance.

Phytases (myo-inositol hexakisphosphate phosphohydrolases) are found naturally in plants and microorganisms, particularly fungi. Interest in these enzymes has been stimulated by the fact that phytase supplements increase the availability of phosphorus in pig and poultry feed and thereby reduce environmental pollution due to excess phosphate excretion in areas where there is intensive livestock production. The wild-type phytases from six different fungi, Aspergillus niger, Aspergillus terreus, Aspergillus fumigatus, Emericella nidulans, Myceliophthora thermophila, and Talaromyces thermophilus, were overexpressed in either filamentous fungi or yeasts and purified, and their biophysical properties were compared with those of a phytase from Escherichia coli. All of the phytases examined are monomeric proteins. While E. coli phytase is a nonglycosylated enzyme, the glycosylation patterns of the fungal phytases proved to be highly variable, differing for individual phytases, for a given phytase produced in different expression systems, and for individual batches of a given phytase produced in a particular expression system. Whereas the extents of glycosylation were moderate when the fungal phytases were expressed in filamentous fungi, they were excessive when the phytases were expressed in yeasts. However, the different extents of glycosylation had no effect on the specific activity, the thermostability, or the refolding properties of individual phytases. When expressed in A. niger, several fungal phytases were susceptible to limited proteolysis by proteases present in the culture supernatant. N-terminal sequencing of the fragments revealed that cleavage invariably occurred at exposed loops on the surface of the molecule. Site-directed mutagenesis of A. fumigatus and E. nidulans phytases at the cleavage sites yielded mutants that were considerably more resistant to proteolytic attack. Therefore, engineering of exposed surface loops may be a strategy for improving phytase stability during feed processing and in the digestive tract.

Effect of the hydrolysis method on the determination of the amino acid composition of proteins.

Fast and reproducible separation and determination of amino acids serves the economical and reliable characterization and quantification of peptides and proteins as well as the identification of proteins by amino acid composition analysis on a large-scale. A prerequisite of a successful compositional analysis is a complete hydrolysis of the peptides and proteins and a quantitative recovery of the residues in the hydrolyzate. We investigated the effect of different acid-hydrolysis methods on the compositional analysis of known proteins in solution and after blotting onto polyvinylidene difluoride membranes and worked out the conditions for the processing of large numbers of samples. The reliability of each method was studied by introducing the analysis data into the AACompIdent software and deducing the protein identification scores. All acid-hydrolysis methods delivered reliable analysis data. The most accurate data were provided by conventional, thermal hydrolysis of proteins in solution in the presence of methanesulfonic acid, closely followed by hydrolysis with hydrochloric acid and microwave radiation-dependent hydrolysis with hydrochloric or methanesulfonic acid, respectively. For blotted proteins, conventional hydrolysis delivered more accurate analysis data in comparison with the microwave radiation-induced hydrolysis. The extraction of the residues from the membrane hydrolyzate was a critical step for unambiguous protein identification. Microwave radiation-induced hydrolysis was responsible for a higher degree of racemization of the residues.

Hydrolysis and amino acid composition of proteins.

Amino acid composition analysis is a classical protein analysis method, which finds a wide application in medical and food science research and is indispensable for protein quantification. It is a complex technique, comprising two steps, hydrolysis of the substrate and chromatographic separation and detection of the residues. A properly performed hydrolysis is a prerequisite of a successful analysis. The most significant developments of the technology in the last decade consist in the (i) reduction of the hydrolysis time by the use of microwave radiation energy; (ii) improvement in the sensitivity of the residue detection, the quantification of the sensitive residues and separation of the enantiomeric forms of the amino acids; (iii) application of amino acid analysis in the large-scale protein identification by database search; and (iv) gradual replacement of the original ion exchange residue separation by reversed-phase high-performance liquid chromatography. Amino acid analysis is currently facing an enormous competition in the determination of the identity of proteins and amino acid homologs by the essentially faster mass spectrometry techniques. The amino acid analysis technology needs further simplification and automation of the hydrolysis, chromatography and detection steps to withstand the pressure exerted by the other technologies.

Inhibition of bovine nasal cartilage degradation by selective matrix metalloproteinase inhibitors.

N-terminal analysis of aggrecan fragments lost from bovine nasal cartilage cultured in the presence of recombinant human interleukin 1alpha revealed a predominant ARGSVIL sequence with an additional ADLEX sequence. Production of the ARGSVIL-containing fragments has been attributed to the action of a putative proteinase, aggrecanase. The minor sequence (ADLEX) corresponds to a new reported cleavage product; comparison of this sequence with the available partial sequence of bovine aggrecan indicates that this is the product of a cleavage occurring towards the C-terminus of the protein. Matrix metalloproteinase (MMP) inhibitors inhibited aggrecan loss from bovine nasal explants incubated in the presence of recombinant human interleukin 1alpha. A strong correlation between inhibition of aggrecan metabolism and inhibition of stromelysin 1 (MMP 3) (r=0.93) suggests a role for stromelysin or a stromelysin-like enzyme in cartilage aggrecan metabolism. However, the compounds were approx. 1/1000 as potent in inhibiting aggrecan loss from the cartilage explants as they were in inhibiting stromelysin. There was little or no correlation between inhibition of aggrecan metabolism and inhibition of gelatinase B (MMP 9) or inhibition of collagenase 1 (MMP 1). Studies with collagenase inhibitors with a range of potencies showed a correlation between inhibition of collagenase activity and inhibition of collagen degradation in the cartilage explant assay. This indicates that in interleukin 1alpha-driven bovine nasal cartilage destruction, stromelysin (or a closely related enzyme) is involved in aggrecan metabolism, whereas collagenase is principally responsible for collagen degradation.

Isolation and expression of a cDNA clone encoding human kynureninase.

Kynureninase (L-kynurenine hydrolase), a pyridoxal-5'-phosphate-(pyridoxal-P)-dependent enzyme, catalyses the cleavage of L-kynurenine and L-3-hydroxykynurenine into anthranilic and 3-hydroxyanthranilic acids, respectively. In this report, we describe the isolation of a cDNA clone encoding human kynureninase. Degenerate oligonucleotides designed from the amino acid sequences of peptides from rat liver kynureninase, were used as primers for reverse-transcription PCR of rat kidney RNA. The resulting rat cDNA product was then used to screen a human hepatoma cell line (Hep G2) cDNA library. Analysis of a positive cDNA clone showed the presence of an insert of 1651 nucleotides containing an open reading frame coding for a protein of 456 amino acids (theoretical molecular mass = 52357 Da). The predicted amino acid sequence of human kynureninase displayed high similarity to that reported for the rat enzyme and to a Saccharomyces cerevisiae gene product putatively ascribed to kynureninase. Profile analysis of kynureninase primary structure indicated the presence of a pyridoxal-P-binding site consensus sequence assigned to class-V aminotransferases, with Lys276 being the residue binding the cofactor. RNA blot analysis of human tissues, including brain, showed the presence of an approximately 2.0-kb mRNA species in all tissues tested. A second mRNA species (approximately 2.6 kb) was also detected in some tissues. After transfection of HEK-293 cells with the cDNA coding for kynureninase, the K(m) values of L-kynurenine and DL-3-hydroxykynurenine for the recombinant enzyme were 671 +/- 37 microM and 13.2 +/- 2.0 microM, respectively.

Investigation on the structure of the active site of monoamine oxidase-B by affinity labeling with the selective inhibitor lazabemide and by site-directed mutagenesis.

The structural features of the active site of human monoamine oxidase B (MAO-B) were investigated by affinity labeling and site-directed mutagenesis. The pseudosubstrate inhibitor N-[2-aminoethyl]-5-chloro-2-pyridine carboxamide HCl (lazabemide) can be irreversibly linked to MAO-B by reduction of the enzyme-inhibitor complex with NaBH(3)CN. Analysis of the flavin spectrum of [(3)H]lazabemide-labeled human MAO-B indicated that insertion of the inhibitor did not occur into the isoalloxazine ring of FAD. After trypsin digestion and HPLC peptide mapping of the radiolabeled enzyme, two labeled peptides were observed. Sequence analysis showed that both peptides started at Val371 of human MAO-B. These results indicate that [(3)H]lazabemide is incorporated into the MAO-B peptide stretch containing the FAD-modified Cys397. The function of putative active-site residues contained in this region was investigated by site-directed mutagenesis and expression of the mutant proteins in HEK-293 cells. Substitution of His382 of MAO-B with an Arg greatly reduced the enzymic activity, suggesting that this residue may represent a nucleophile relevant for the MAO-B catalytic mechanism. Whereas it has been shown that mutation of Cys389 with a Ser residue does not markedly affect the activity of the enzyme [Wu, H.-F., Chen, K. and Shih, J.C. (1993) Mol. Pharmacol. 43, 888-893] the mutant carrying an Ala at this position was virtually inactive. Conversely, substitution of Lys386 (to Met) and Ser394 (to Ala) did not markedly modify the kinetic properties of the enzyme. We also report that mutation of MAO-B Thr158 (to Ala) resulted in a dramatic loss of enzymic activity.

Quantification of cysteine residues following oxidation to cysteic acid in the presence of sodium azide.

Quantification of cysteines by amino acid composition analysis is inaccurate because of decomposition of these residues during protein hydrolysis. Cysteine (and cystine) residues are oxidized to cysteic acid following hydrochloric acid hydrolysis in the presence of sodium azide. Using selected native and recombinant proteins, containing different numbers of cysteine residues, we investigated the conditions for the quantitative oxidation of cysteines to cysteic acid in the presence of sodium azide. Protein hydrolysis with hydrochloric acid in the presence of 0.20% sodium azide resulted in 87-100% oxidation of the cysteines to cysteic acid which was easily quantified. The results were highly reproducible so that the azide-induced oxidation can be used as a general method to determine cysteine residues in a given protein. The sodium azide-dependent oxidation is superior to oxidation with performic acid because (i) it can be performed in solution not requiring protein lyophilization and in approximately half of the time; (ii) it delivers slightly higher yields of cysteic acid; and (iii) it does not affect tyrosine residues, which can be modified during the performic acid treatment.

Cloning and characterization of a soluble kynurenine aminotransferase from rat brain: identity with kidney cysteine conjugate beta-lyase.

In this study, we describe the cloning and characterization of a soluble form of kynurenine aminotransferase (KAT, EC 2.6.1.7) present in rat brain. Soluble KAT was purified from rat kidney and the amino acid sequences of four tryptic peptides determined. These peptides were found to belong to the amino acid sequence reported for rat kidney soluble cysteine conjugate beta-lyase, indicating that rat kidney KAT and beta-lyase represent the same molecular entity. Oligonucleotide probes derived from the beta-lyase cDNA were then used as primers for PCR of reverse-transcribed rat brain poly(A)+ RNA. After subcloning of the resulting PCR fragment and sequencing of the isolated rat brain clone, its oligonucleotide sequence was found to be identical to that reported for the beta-lyase cDNA. Further evidence that the isolated rat brain clone encoded for KAT was obtained by transfecting HEK-293 cells with a construct containing the coding sequence for the enzyme. The transfected cells exhibited KAT activity and, in the presence of 2 mM pyruvate and 2-oxoglutarate, the Km values for L-kynurenine were 1.2 mM and 86.3 microM, respectively. Northern blot analysis of rat kidney, liver, and brain RNA revealed a single species of KAT/beta-lyase mRNA of approximately 2.1 kb.

Oxidation of cysteine and methionine residues during acid hydrolysis of proteins in the presence of sodium azide.

Sodium azide is widely used as bacteriostatic agent during downstream processing of proteins. Amino acid composition analysis of protein samples, subjected to hydrolysis with hydrochloric acid in a buffer containing sodium azide, revealed the presence of cysteic acid, methionine sulfoxide, and methionine sulfone in addition to the expected reaction products. Hydrolysis with methanesulfonic acid in the presence of sodium azide resulted in detection of only methionine sulfoxide in addition to the expected products. When the proteins were hydrolyzed in a buffer containing no sodium azide or after its removal by dialysis, no oxidation products were detected (except for minor amounts of methionine sulfoxide). The generation of the particular oxidation products was affected by the concentration of sodium azide in the protein solution. Therefore, presence of sodium azide in protein samples intended for amino acid composition analysis may lead to wrong conclusions concerning oxidation of cysteine and methionine residues.

Isolation of a heregulin-like growth factor secreted by estrogen receptor-negative MDA-MB-231 human breast cancer cells that stimulates estrogen receptor-positive cells.

A 52 kDa peptide growth factor secreted by the estrogen receptor (ER)-negative human breast cancer cell line MDA-MB-231 was purified to homogeneity. It induced transient membrane ruffling, lamellipodia formation, cell motility and proliferation exclusively of ER-positive human breast cancer cells. Partial sequencing revealed a high homology to the protein family of heregulins. However, the obtained amino acid sequences of the new factor were not completely identical to any of the members of the heregulin family. This finding together with the observation that the successful purification protocol was significantly different from that used to isolate other members of the heregulin family indicate the isolation of a novel heregulin-like proliferation factor for ER-positive human breast cancer cells.

Mapping of the p56lck-mediated phosphorylation of GAP and analysis of its influence on p21ras-GTPase activity in vitro.

The protein tyrosine kinase p56lck and other members of the src family can transduce signals from activated cell-surface receptors. As we showed earlier the GTPase-activating protein (GAP), a regulator of p21ras, is a substrate of p56lck. Here, tryptic peptides of p56lck-phosphorylated GAP were generated and analyzed by two-dimensional thin layer chromatography and mass spectroscopy. Results revealed that p56lck phosphorylates GAP specifically on Tyr-460 in vitro and in vivo. The effect of tyrosine phosphorylation of GAP on its GTPase-activating activity versus p21ras was then tested using a p21ras-dependent GTPase assay system. Our results demonstrate that p56lck-mediated tyrosine phosphorylation of GAP is not sufficient to change directly its effect on the intrinsic GTPase activity of p21ras.

Mass spectrometric analysis of human soluble catechol O-methyltransferase expressed in Escherichia coli. Identification of a product of ribosomal frameshifting and of reactive cysteines involved in S-adenosyl-L-methionine binding.

Technological advances in the field of mass spectrometry (MS) are providing powerful analytical means for the investigation of proteins and peptides. In the present work we have used pneumatically assisted electrospray (ion-spray) MS for the biochemical characterization of recombinant human catechol O-methyltransferase (rhCOMT). hCOMT could be expressed in Escherichia coli in large quantities but in two forms of different size, both enzymically active. Electrospray MS analysis showed that the smaller rhCOMT protein had a molecular mass of 24352 +/- 2 Da, corresponding to the calculated value for native hCOMT (without the initiating methionine), whereas that mass of the larger protein was of 25775 +/- 4 Da. To investigate the molecular differences between the two proteins, they were digested with trypsin and the peptides produced analysed by electrospray MS. Neither protein apparently contained disulfide bridges and the observed molecular masses of the tryptic peptides corresponded to the calculated values. It was possible to determine, however, that the larger protein contained an extended C-terminus with the correct sequence GPGSEAGP plus an additional stretch, EDLR. This C-terminal extension resulted from ribosomal frameshift at the codon of the last proline (CCC, rare codon in prokaryotes). In fact, rightward frameshifting would produce a hCOMT form with an additional stretch of 11 amino acid (EDLRSHHHHHH) and the calculated molecular mass of this protein (25773.5 Da) is in good agreement with our experimental result. The differential reactivity of the cysteine residues of the correct rhCOMT enzyme, in the presence and in the absence of S-adenosyl-L-methionine (AdoMet) and MgCl2, was also studied. 5-Iodoacetamido fluorescein (5-IAF) was used as thiol-modifying reagent. Under the conditions used, 5-IAF rapidly inactivated rhCOMT but the presence of AdoMet and MgCl2 partially protected it from inactivation. The 5-IAF-labeled tryptic peptides were separated by HPLC and then submitted to electrospray MS and tandem MS. Several cysteine residues appeared to be readily available to chemical modification by 5-IAF. Incorporation of 5-IAF occurred to a larger extent into Cys32, Cys68, Cys94 and Cys172. AdoMet and MgCl2 markedly reduced the label incorporation into Cys68 and Cys94, therefore suggesting that these residues belong to a region at or near the binding site of AdoMet.

Molecular cloning and functional expression of human 3-hydroxyanthranilic-acid dioxygenase.

Increased cerebral levels of the endogenous excitotoxin quinolinic acid (QUIN) have been speculatively linked to neuronal damage following neurological and inflammatory disorders. 3-Hydroxyanthranilic-acid dioxygenase (3-HAO; 3-hydroxyanthranilate 3,4-dioxygenase, EC 1.13.11.6) is the enzyme that catalyzes the synthesis of QUIN from 3-hydroxyanthranilic acid, and evidence suggests that it could play a role in disorders associated with altered tissue levels of QUIN. In this report, we describe the isolation of a full-length cDNA clone encoding human 3-HAO (h3-HAO). Degenerate oligonucleotides were designed from the amino acid sequences of tryptic peptides of rat liver 3-HAO, and they were used as primers for reverse transcription-polymerase chain reaction of rat liver RNA. The resulting rat cDNA product was used to screen a human hepatoma cell line (HepG2) cDNA library and to isolate a human 3-HAO cDNA clone. This clone was found to have an insert of 1276 nucleotides. The deduced primary structure of h3-HAO is composed of 286 amino acid residues with a predicted molecular mass of approximately 32.6 kDa. The human sequence exhibits high similarity (94%) to the rat partial amino acid sequence deduced from the rat reverse transcription-polymerase chain reaction fragment. Insertion of the h3-HAO coding sequence into a eukaryotic expression vector yielded relatively high amounts of the active enzyme in human embryonic kidney HEK-293 cells. The Km value of 3-HANA for recombinant h3-HAO (approximately 2 microM) was in good agreement with that reported for the native enzyme. Immunoblot analysis of recombinant h3-HAO revealed a polypeptide with an apparent molecular mass of 32 kDa, as predicted from the deduced amino acid sequence. RNA blot analysis of human liver and HepG2 cells revealed one major species of h3-HAO mRNA of approximately 1.3 kilobases.

Structural analysis and localization of the carbohydrate moieties of a soluble human interferon gamma receptor produced in baculovirus-infected insect cells.

A soluble form of the human interferon gamma receptor that is required for the identification of interferon gamma antagonists was expressed in baculovirus-infected insect cells. The protein carried N-linked carbohydrate and showed a heterogeneity on denaturing polyacrylamide gels. We investigated the utilization of the potential sites for N-linked glycosylation and the structure of the carbohydrate moieties of this soluble receptor. Amino acid sequence analysis and ion spray mass spectrometry revealed that of the five potential sites for N-linked glycosylation, Asn17 and Asn69 were always utilized, whereas Asn62 and Asn162 were utilized in approximately one-third of the protein population. Asn223 was never found to be glycosylated. The soluble receptor was treated with N-glycosidase F and the oligosaccharides released were analyzed by matrix-assisted laser desorption mass spectrometry, which showed that the protein carried six types of short carbohydrate chains. The predominant species was a hexasaccharide of molecular mass 1,039, containing a fucose subunit linked to the proximal N-acetylglucosamine residue: [formula: see text]