Structure of the protease domain of memapsin 2 (beta-secretase) complexed with inhibitor.

Memapsin 2 (beta-secretase) is a membrane-associated aspartic protease involved in the production of beta-amyloid peptide in Alzheimer's disease and is a major target for drug design. We determined the crystal structure of the protease domain of human memapsin 2 complexed to an eight-residue inhibitor at 1.9 angstrom resolution. The active site of memapsin 2 is more open and less hydrophobic than that of other human aspartic proteases. The subsite locations from S4 to S2' are well defined. A kink of the inhibitor chain at P2' and the change of chain direction of P3' and P4' may be mimicked to provide inhibitor selectivity.

Differential requirements for basic amino acids in transcription factor IIIA-5S gene interaction.

Basic amino acids Arg, Lys, and His in the Cys2His2 zinc fingers of transcription factor IIIA (TFIIIA) potentially have important roles in factor binding to the extended internal control region (ICR) of the 5S ribosomal gene. Conserved and non-conserved basic residues in the N-terminal fingers I, II, III and the more C-terminal fingers V and IX were analyzed by site-directed mutagenesis and DNase I protection in order to assess their individual requirement in the DNA-binding mechanism. In the DNA recognition helix of finger II, the conserved Arg at position 62 (N-terminal side of the first zinc-coordinating histidine) was changed to a Leu or Gln. Both the R62L and R62Q mutations inhibited Xenopus TFIIIA-dependent DNase I footprinting along the entire 5S gene ICR. When His-58 (non-conserved basic residue with DNA-binding potential in the same helical region) was changed to a Gln, the mutated protein was able to protect the ICR from DNase I digestion. Therefore, Arg-62 is individually required for TFIIIA binding over the entire ICR whereas His-58 is not. Fingers V and IX have conserved Arg residues in positions identical to Arg-62 in finger II (Arg-154 in finger V and Arg-271 in finger IX). When these residues were changed to Leu and Ile respectively, TFIIIA-dependent DNase I protection was observed along the entire 5S gene ICR. These results indicate differing DNA-binding mechanisms by the N-terminal fingers versus the C-terminal fingers at the level of individual amino acid-nucleotide interactions. In the N-terminal finger I, the conserved Lys at position 11 outside the recognition helix and a conserved hydrophobic Trp at position 28 within the helix were changed to an Ala and Ser respectively. The K11A change inhibited TFIIIA-dependent DNase I protection to a much greater extent than the W28S change.

Enzymic characteristics of secreted aspartic proteases of Candida albicans.

Candida yeasts are rarely infectious, but frequently cause life-threatening systemic infections in patients immunocompromised by AIDS or by immunosuppressive therapeutics. The secreted aspartic proteases (Saps) are known virulence factors of pernicious Candida species. The most virulent, Candida albicans, possesses at least nine SAP genes, some of which are specifically expressed from cells with morphologies associated with virulence. Only one of these proteases, Sap2, has been previously purified from yeast in sufficient quantities for enzymic studies. The other enzymes are present in low amounts in yeast culture and are difficult to purify. As a consequence, enzyme properties, including the substrate specificities, of all Saps are poorly studied. Therefore, four Saps that are known to be expressed in C. albicans, Sap1, Sap2, Sap3 and Sap6, were produced in Escherichia coli as recombinant zymogens and purified in large quantities. These proenzymes were autoactivated and purified as active proteases. The enzymic properties including the substrate specificities at the P(1) and P(1)' sites were determined using a competitive hydrolysis method employing synthetic substrate mixtures. All four Saps cleave peptide bonds between larger hydrophobic amino acids, but these somewhat broad specificities differ in detail among the four enzymes at both sites. At the P(1) site, Sap1, Sap2 and Sap6 prefer Phe while Sap3 prefers Leu. Positively charged amino acids are also accommodated, especially by Sap2 and Sap3. The specificities at P(1)' are broader than at P(1) for all four enzymes. Sap6 prefers Ala, whereas other Saps prefer Tyr. Acidic side chains are also accommodated at this site. Analysis of substrates with a hydrophobic amino acid in P(1)' reveals that all the Saps possess a unique preference for Ala at this site. The observed differences of residue preferences among Saps may be utilized for the design of specific substrates and inhibitors.

Rearranging the domains of pepsinogen.

Most eukaryotic aspartic protease zymogens are synthesized as a single polypeptide chain that contains two distinct homologous lobes and a pro peptide, which is removed upon activation. In pepsinogen, the pro peptide precedes the N-terminal lobe (designated pep) and the C-terminal lobe (designated sin). Based on the three-dimensional structure of pepsinogen, we have designed a pepsinogen polypeptide with the internal rearrangement of domains from pro-pep-sin (native pepsinogen) to sin-pro-pep. The domain-rearranged zymogen also contains a 10-residue linker designed to connect sin and pro domains. Recombinant sin-pro-pep was synthesized in Escherichia coli, refolded from 8 M urea, and purified. Upon acidification, sin-pro-pep autoactivates to a two-chain enzyme. However, the emergence of activity is much slower than the conversion of the single-chain zymogen to a two-chain intermediate. In the activation of native pepsinogen and sin-pro-pep, the pro region is cleaved at two sites between residues 16P and 17P and 44P and 1 successively, and complete activation of sin-pro-pep requires an additional cleavage at a third site between residues 1P and 2P. In pepsinogen activation, the cleavage of the first site is rate limiting because the second site is cleaved more rapidly to generate activity. In the activation of sin-pro-pep, however, the second site is cleaved slower than the first, and cleavage of the third site is the rate limiting step.(ABSTRACT TRUNCATED AT 250 WORDS)

Multiple functions of pro-parts of aspartic proteinase zymogens.

The importance of aspartic proteinases in human pathophysiology continues to initiate extensive research. With burgeoning information on their biological functions and structures, the traditional view of the role of activation peptides of aspartic proteinases solely as inhibitors of the active site is changing. These peptide segments, or pro-parts, are deemed important for correct folding, targeting, and control of the activation of aspartic proteinase zymogens. Consequently, the primary structures of pro-parts reflect these functions. We discuss guidelines for formation of hypotheses derived from comparing the physiological function of aspartic proteinases and sequences of their pro-parts.

Structure-based design: potent inhibitors of human brain memapsin 2 (beta-secretase).

Memapsin 2 (beta-secretase) is one of two proteases that cleave the beta-amyloid precursor protein (APP) to produce the 40-42 residue amyloid-beta peptide (Abeta) in the human brain, a key event in the progression of Alzheimer's disease. On the basis of the X-ray crystal structure of our lead inhibitor (2, OM99-2 with eight residues) bound to memapsin, we have reduced the molecular weight and designed potent memapsin inhibitors. Structure-based design and preliminary structure-activity studies have been presented.

Detection of procathepsin D in rat milk.

The presence of procathepsin D, a zymogen of the soluble lysosomal aspartic proteinase cathepsin D, was detected in rat milk using Western blot analysis and assay of proteolytic activity in acidic buffers. No other forms of cathepsin D were found. Two different polyclonal anti-procathepsin D antibodies were used for immunochemical detection of procathepsin D. Both antibodies we found to recognize rat procathepsin D. Proteolytic activity in acidic buffers was detected using a fluorogenic substrate specific for cathepsin D and was abolished by pepstatin A, a specific inhibitor of aspartic proteinases. This study represents third demonstration of presence of procathepsin D in mammal breast milk. Potential sources and physiological functions are discussed.

Human procathepsin D: three-dimensional model and isolation.

Human procathepsin D was isolated from medium of human breast cancer cell line ZR-75-1 potentiated with estrogen. The isolation involved both immunoaffinity chromatography and ion-exchange chromatography. The affinity chromatography employed polyclonal antibodies raised against a synthetic activation peptide of human cathepsin D. We have started preliminary crystallization trials using the isolated material. A model of human procathepsin D was also built using coordinates of human cathepsin D and pig pepsinogen. The model aids understanding of multiple roles played by activation peptides of aspartic proteinases and will be used as a starting model for molecular replacement.

The high-resolution crystal structure of porcine pepsinogen.

The structure of porcine pepsinogen at pH 6.1 has been refined to an R-factor of 0.173 for data extending to 1.65 A. The final model contains 180 solvent molecules and lacks density for residues 157-161. The structure of this aspartic proteinase zymogen possesses many of the characteristics of pepsin, the mature enzyme. The secondary structure of the zymogen consists predominantly of beta-sheet, with an approximate 2-fold axis of symmetry. The activation peptide packs into the active site cleft, and the N-terminus (1P-9P) occupies the position of the mature N-terminus (1-9). Thus changes upon activation include excision of the activation peptide and proper relocation of the mature N-terminus. The activation peptide or residues of the displaced mature N-terminus make specific interactions with the substrate binding subsites. The active site of pepsinogen is intact; thus the lack of activity of pepsinogen is not due to a deformation of the active site. Nine ion pairs in pepsinogen may be important in the advent of activation and involve the activation peptide or regions of the mature N-terminus which are relocated in the mature enzyme. The activation peptide-pepsin junction, 44P-1, is characterized by high thermal parameters and weak density, indicating a flexible structure which would be accessible to cleavage. Pepsinogen is an appropriate model for the structures of other zymogens in the aspartic proteinase family.

Proteolytic activation of recombinant pro-memapsin 2 (pro-beta-secretase) studied with new fluorogenic substrates.

Memapsin 2 (beta-secretase), a membrane-anchored aspartic protease, is involved in the cleavage of beta-amyloid precursor protein to form beta-amyloid peptide. The primary structure of memapsin 2 suggests that it is synthesized in vivo as pro-memapsin 2 and converted to memapsin 2 by an activating protease [Lin et al. (2000) Proc. Natl. Acad. Sci. U.S.A. 97, 1456-1460]. To simulate this activation mechanism and to produce stable mature memapsin 2 for kinetic/specificity studies, we have investigated the activation of recombinant pro-memapsin 2 by several proteases with trypsin-like specificity. Clostripain, kallikrein, and trypsin increased the activity of pro-memapsin 2. Clostripain activation was accompanied by the cleavage of the pro region to form mainly two activation products, Leu(30p)- and Gly(45p)-memapsin 2. Another activation product, Leu(28p)-memapsin 2, was also purified. Kinetics of the activated memapsin 2 were compared with pro-memapsin 2 using two new fluorogenic substrates, Arg-Glu(5-[(2-aminoethyl)amino]naphthalene-1-sulfonic acid (EDANS))-Glu-Val-Asn-Leu-Asp-Ala-Glu-Phe-Lys(4-(4-dimethylaminophe nyl azo)benzoic acid (DABCYL))-Arg and (7-methoxycoumarin-4-yl)acetyl (MCA))-Ser-Glu-Val-Asn-Leu-Asp-Ala-Glu-Phe-Lys(2,4-dinitrophenyl (DNP)). These results establish that the activity of pro-memapsin 2 stems from a part-time and reversible uncovering of its active site by its pro region. Proteolytic removal of part of the pro-peptide at Leu(28p) or Gly(45p), which diminishes the affinity of the shortened pro-peptide to the active site, results in activated memapsin 2. These results also suggest that Glu(33p)-memapsin 2 observed in the cells expressing this enzyme [Vassar et al. (1999) Science 286, 735-741; Yan et al. (1999) Nature 402, 533-537] is an active intermediate of in vivo activation, or that the peptide Glu(33p)-Arg(44p) may serve a regulatory role.

Enzymic activities of two-chain pepsinogen, two-chain pepsin, and the amino-terminal lobe of pepsinogen.

In order to study the relationships of aspartic proteases, we have modified pepsin, a single-chain eukaryotic enzyme, to a two-chain heterodimer, which resembles aspartic proteases from retrovirus, including human immunodeficiency virus. Two fragments of pepsinogen, residues 1P-172 and 173-326, were expressed separately in Escherichia coli. Mixtures of chains were refolded from urea solutions to generate an active two-chain pepsinogen, which was converted to two-chain pepsin in acid solutions. The intramolecular and bimolecular activation constants (k1 and k2) of two-chain pepsinogen are about 1.5-fold and one-sixth, respectively, of those for pepsinogen. Structural evidence suggests that the faster k1 of two-chain pepsinogen is due to decreased interaction of the propeptide with the pepsin moiety, implying that the rate-limiting step in the intramolecular activation of pepsinogen is the "conformational dissociation" of its propeptide. Two-chain pepsin has the same Km but only one-sixth of the kcat of pepsin. Both pepsinogen chains are capable of independent refolding. The refolding of the NH2-terminal chain, which contains the propeptide and the NH2-terminal lobe, generated a small amount of proteolytic activity which is likely derived from the homodimer of the NH2-terminal lobe. It has been postulated that mammalian aspartic proteases, which contain two structurally homologous lobes, are derived in evolution from a homodimer enzyme by gene duplication and fusion (Tang, J., James, M. N. G., Hsu, I.-N., Jenkins, J. A., and Blundell, T. L. (1978) Nature 271, 618-621). The observation of the homodimer activity of the NH2-terminal lobe of pepsinogen suggests that the interface of the lobes is conservative in evolution.

Proteolytic processing mechanisms of a miniprecursor of the aspartic protease of human immunodeficiency virus type 1.

The infectivity of the human immunodeficiency virus (HIV) depends upon correct proteolytic processing of viral polyprotein precursors, the Pr55gag and Pr160gag-pol polyproteins. The processing is mediated spontaneously by the viral protease unit (PR) contained within the Pr160gag-pol precursor. However, little is known about the mechanism of this process. The expression in Escherichia coli and the isolation of a 14-kDa HIV-1 PR "miniprecursor" with Ala28 mutated to serine has permitted study of the mechanism for cleavage at the N-terminus of the protease. The miniprecursor is active against a synthetic peptide substrate, and its specific activity is near that of the mutant mature protease. The rate of conversion of radiolabeled precursor to mature protease is quantitated by measuring the amounts of the two radiolabeled proteins separated by SDS-PAGE. The apparent first-order conversion rate constant, kapp, is dependent on miniprecursor concentration indicating a second-order reaction and suggesting an interdimeric processing mechanism. A significant first-order rate constant is observed when the plot of kapp versus initial precursor concentration is extrapolated to zero. This observation suggests the presence of an alternative processing mechanism involving a single active precursor dimer. The presence of both mechanisms is an advantage for the virus to ensure processing under various conditions.

Mapping and molecular modeling of a recognition domain for lysosomal enzyme targeting.

Lysosomal enzymes contain a common protein determinant that is recognized by UDP-GlcNAc:lysosomal enzyme N-acetylglucosamine-1-phosphotransferase, the initial enzyme in the biosynthesis of mannose-6-P residues. Previously, we generated a lysosomal enzyme recognition domain by substituting two regions (lysine 203 and amino acids 265-292) of the lysosomal hydrolase cathepsin D into a related secretory protein glycopepsinogen. When expressed in Xenopus oocytes, the oligosaccharides of the chimeric protein were efficiently phosphorylated (Baranski, T. J., Faust, P. L., and Kornfeld, S. (1990) Cell 63, 281-291). In the current study, incremental substitutions of cathepsin D residues into glycopepsinogen and alanine-scanning mutagenesis were utilized to define the recognition domain more precisely. A computer-generated model of the cathepsin D/pepsinogen chimeric molecule served as a guide for mutagenesis and for the interpretation of results. These studies indicate that the recognition domain is a surface patch that contains multiple interacting sites. There is a strict positional requirement for the lysine residue at position 203.

Recombinant canditropsin, an extracellular aspartic protease from yeast Candida tropicalis. Escherichia coli expression, purification, zymogen activation, and enzymic properties.

A cDNA fragment which encodes the zymogen of canditropsin, the extracellular aspartic protease from the yeast Candida tropicalis (Togni,G., Sanglard, D., Falchetto, R., and Monod, M. (1991) FEBS Lett. 286, 181-185) was cloned into a T7 expression vector for the synthesis of the recombinant zymogen in Escherichia coli. Recombinant canditropsinogen (Ctg), which was expressed as inclusion bodies in the cytosol of E. coli, was refolded by dialysis from an 8 M urea solution and purified to homogeneity using chromatographies on Sephacryl S-300 and on MonoQ columns. The purified Ctg was converted into canditropsin by either acid activation or trypsin conversion. The specificity of the resulting recombinant canditropsin toward polypeptide substrates is significantly different from other aspartic proteases. Canditropsin hydrolyzes oxidized insulin B chain between Ala-Leu and many other minor cleavage sites. Canditropsin also hydrolyzes keratin and collagen, which are components of connective tissues known to be hydrolyzed by canditropsin during Candida infections. Canditropsin was strongly inhibited by the universal aspartic protease inhibitor pepstatin (Ki = 1.75 x 10(-8) M) and inactivated by two aspartic protease inactivators, DAN and EPNP. Canditropsin is weakly inhibited by leupeptin and antipain, with an apparent Ki of 1.74 x 10(-4)M and 1.5 x 10(-5) M, respectively.