Differential effects of inhibitors on the gamma-secretase complex. Mechanistic implications

Gamma-secretase is a protease complex of four integral membrane proteins, with presenilin (PS) as the apparent catalytic component, and this enzyme processes the transmembrane domains of a variety of substrates, including the amyloid beta-protein precursor and the Notch receptor. Here we explore the mechanisms of structurally diverse gamma-secretase inhibitors by examining their ability to displace an active site-directed photoprobe from PS heterodimers. Most gamma-secretase inhibitors, including a potent inhibitor of the PS-like signal peptide peptidase, blocked the photoprobe from binding to PS1, indicating that these compounds either bind directly to the active site or alter it through an allosteric interaction. Conversely, some reported inhibitors failed to displace this interaction, demonstrating that these compounds do not interfere with the protease by affecting its active site. Differential effects of the inhibitors with respect to photoprobe displacement and in cell-based and cell-free assays suggest that these compounds are important mechanistic tools for deciphering the workings of this intramembrane-cleaving protease complex and its similarity to other polytopic aspartyl proteases.     

Formation of an intramolecular cystine disulfide during the reaction of 8-azidoguanosine 5'-triphosphate with cytosolic phosphoenolpyruvate carboxykinase (GTP) causes inactivation without photolabeling

Phosphoenolpyruvate carboxykinase (GTP) (PEPCK) specifically utilizes a guanosine or inosine nucleotide as a substrate, yet it does not share extended sequence homology with other GTP-binding proteins, and the molecular basis for its nucleotide specificity is not understood. In an effort to locate the enzyme's nucleotide-binding site, we have studied the interaction of cytosolic PEPCK from rat liver with the photoprobe 8-azidoGTP, which fulfills the criteria of a specific photoaffinity label for PEPCK. The photoprobe binds reversibly to the enzyme prior to modification and at low concentrations causes greater than 60% inactivation (Ki = 1.2 microM). GTP provides nearly complete protection against inactivation by 8-azidoGTP, whereas phosphoenolpyruvate and metal ions provide partial protection. In addition, the photoprobe is a substrate for the enzyme and has a Km similar to that for GTP. However, the extent of covalent modification by [32P]8-azidoGTP as measured by three independent techniques is significantly lower than the extent of enzyme inactivation. Further investigation of this anomaly has revealed that the loss in enzymatic activity is caused by modification of a critical cysteine residue in a reaction that does not terminate with covalent attachment of the photolabel. Quantitation of the total free thiols of modified PEPCK shows that 2 mol of cysteine is lost per mole of inactivated enzyme. These results indicate that the photoinactivation of PEPCK by 8-azidoGTP is caused by the formation of an intramolecular cystine disulfide bridge, thus providing evidence for the existence of a pair of proximal cysteine residues within the GTP-binding site. The interaction of cysteine residues with the reactive photogenerated derivatives of 8-azidopurines is discussed.     

Transbilayer migration (flip-flop) of 12-(4-azido-2-nitrophenoxy)stearoyl glucosamine in large unilamellar phospholipid vesicles

The ability of the glycolipid photoprobe, 12-(4-azido-2-nitrophenoxy)-stearoyl[1-14C]glucosamine (12-APS-GlcN), to undergo transbilayer flip-flop and intermembrane transfer between liposomes was examined. It was found that probe which was incorporated into membranes during the preparation of large unilamellar vesicles (LUVs) could be rapidly and completely extracted by incubation of these donor vesicles (in the liquid-crystalline state) with probe-free acceptor vesicles.     

Insulin receptors are bivalent as demonstrated by photoaffinity labeling.

Insulin receptors in human placental membranes were photoaffinity-labeled with a radioactive human insulin-like growth factor I (hIGF-I) photoprobe N epsilon B28-monoazidobenzoyl 125I-hIGF-I either alone or together with a non-radioactive insulin photoprobe N epsilon B29-monoazidobenzoyl insulin. Precipitation of the solubilized receptors with anti-insulin antibody showed that receptors labeled with the radioactive hIGF-I photoprobe were detected in the immunoprecipitate only when photolabeling was carried out in the presence of the non-radioactive insulin photoprobe. Comparable results were obtained in converse experiments using a radioactive insulin photoprobe N epsilon B29-monoazidobenzoyl 125I-insulin, a non-radioactive hIGF-I photoprobe N epsilon B28-monoazidobenzoyl hIGF-I, and an antibody to hIGF-I. The amount of radioactive receptors precipitated by either the anti-insulin antibody or the anti-hIGHF-I antibody was close to the expected amount. These observations demonstrate that the insulin receptor is bivalent being capable of binding two molecules of ligand.     

分支位点在真核基因受体位点识别中的作用

真核基因受体位点识别是剪接位点识别的一部分,也是基因识别中的重要环节,一直受到研究人员的关注。已有的研究结果显示受体位点的识别与分支位点有关,然而关于分支位点和受体位点识别的关系问题,目前还无人将其作为专门的问题予以深入研究。从受体位点识别出发,选取不同的受体位点序列长度,以神经网络为识别工具,对分支位点在受体位点识别中的作用做了深入研究和分析。实验结果表明,受体位点序列的特征信息集中在分支位点一例,因此分支位点在受体位点识别中具有重要作用。研究结果为受体位点识别问题中序列特征提取提供了依据。     

Mapping of nucleotide-depleted mitochondrial F1-ATPase with 2-azido-[alpha-32P]adenosine diphosphate. Evidence for two nucleotide binding sites in the beta subunit

Photolabeling of nucleotide binding sites in nucleotide-depleted mitochondrial F1 has been explored with 2-azido [alpha-32P]adenosine diphosphate (2-N3[alpha-32P] ADP). Control experiments carried out in the absence of photoirradiation in a Mg2+-supplemented medium indicated the presence of one high affinity binding site and five lower affinity binding sites per F1. Similar titration curves were obtained with [3H]ADP and the photoprobe 3'-arylazido-[3H]butyryl ADP [( 3H]NAP4-ADP). Photolabeling of nucleotide-depleted F1 with 2-N3[alpha-32P]ADP resulted in ATPase inactivation, half inactivation corresponding to 0.6-0.7 mol of photoprobe covalently bound per mol F1. Only the beta subunit was photolabeled, even under conditions of high loading with 2-N3[alpha-32P]ADP. The identification of the sequences labeled with the photoprobe was achieved by chemical cleavage with cyanogen bromide and enzymatic cleavage by trypsin. Under conditions of low loading with 2-N3[alpha-32P]ADP, resulting in photolabeling of only one vacant site in F1, covalently bound radioactivity was located in a peptide fragment of the beta subunit spanning Pro-320-Met-358 identical to the fragment photolabeled in native F1 (Garin, J., Boulay, F., Issartel, J.-P., Lunardi, J., and Vignais, P. V. (1986) Biochemistry 25, 4431-4437). With a heavier load of photoprobe, leading to nearly 4 mol of photoprobe covalently bound per mol F1, an additional region of the beta subunit was specifically labeled, corresponding to a sequence extending from Gly-72 to Arg-83. The isolated beta subunit also displayed two binding sites for 2-N3-[alpha-32P]ADP. When F1 was first photolabeled with a low concentration of NAP4-ADP, leading to the covalent binding of 1.5 mol of NAP4-ADP/mol F1, with the bound NAP4-ADP distributed equally between the alpha and beta subunits, a subsequent photoirradiation in the presence of 2-N3[alpha-32P]ADP resulted in covalent binding of the 2-N3[alpha-32P]ADP to both alpha and beta subunits. It is concluded that each beta subunit in mitochondrial F1 contains two nucleotide binding regions, one of which belongs to the beta subunit per se, and the other to a subsite shared with a subsite located on a juxtaposed alpha subunit. Depending on the experimental conditions, the subsite located on the alpha subunit is either accessible or masked. Unmasking of the subsite in the three alpha subunits of mitochondrial F1 appears to proceed by a concerted mechanism.     

A cysteine-linkable,short cleavable photoprobe with dual functionality to explore protein-protein interfaces

We developed a bifunctional photoprobe with dual functionality, that can be specifically tethered to cysteinyl residues of peptides and proteins through a short cleavable disulfide bond. Thus, an aryldiazonium moiety is positioned at approximately 8.5 A from the modified cysteinyl alpha-carbon, leading to one of the shortest cleavable linkages. In a sodium azide-containing buffer, the aryldiazonium moiety is transformed into an aryl azide. Therefore, with one bifunctional photoprobe two types of photogenerated species can be obtained: a hydrophilic and positively charged arylcation or a hydrophobic nitrene. We coupled the aryldiazonium probe, in a site-directed manner, to a nicotinic acetylcholine receptor competitive antagonist, obtained by chemical engineering of an analogue of a snake alpha-neurotoxin. In this molecule, Arg33, which is known to interact with the receptor, was replaced by a cysteine residue, where the photoprobe could be attached. Under inactinic light, this novel photosensitive snake toxin behaved as a reversible ligand on the Torpedo acetylcholine receptor. However, when irradiated at 391 nm, it generated a highly reactive arylcation which labeled mostly the receptor alpha-subunit, confirming the location of the tip of the second toxic loop near this receptor subunit. Finally, we showed that reduction of the disulfide bond, linking the ligand to the photocoupled receptor, allowed introduction of radioactivity on the labeled residue(s), opening the way to further characterization and avoiding the synthesis of a radioactive bifunctional photoprobe.     

Novel γ-secretase enzyme modulators directly target presenilin protein

γ-Secretase is essential for the generation of the neurotoxic 42-amino acid amyloid β-peptide (Aβ(42)). The aggregation-prone hydrophobic peptide, which is deposited in Alzheimer disease (AD) patient brain, is generated from a C-terminal fragment of the β-amyloid precursor protein by an intramembrane cleavage of γ-secretase. Because Aβ(42) is widely believed to trigger AD pathogenesis, γ-secretase is a key AD drug target. Unlike inhibitors of the enzyme, γ-secretase modulators (GSMs) selectively lower Aβ(42) without interfering with the physiological function of γ-secretase. The molecular target(s) of GSMs and hence the mechanism of GSM action are not established. Here we demonstrate by using a biotinylated photocross-linkable derivative of highly potent novel second generation GSMs that γ-secretase is a direct target of GSMs. The GSM photoprobe specifically bound to the N-terminal fragment of presenilin, the catalytic subunit of γ-secretase, but not to other γ-secretase subunits. Binding was differentially competed by GSMs of diverse structural classes, indicating the existence of overlapping/multiple GSM binding sites or allosteric alteration of the photoprobe binding site. The β-amyloid precursor protein C-terminal fragment previously implicated as the GSM binding site was not targeted by the compound. The identification of presenilin as the molecular target of GSMs directly establishes allosteric modulation of enzyme activity as a mechanism of GSM action and may contribute to the development of therapeutically active GSMs for the treatment of AD.     

Synthesis and properties of azidonitrophenyl pyrophosphate, a photoaffinity probe of the nucleotide binding sites of mitochondrial F1-ATPase

4-Azido-2-nitrophenyl pyrophosphate (azido-PPi) labeled with 32P in the alpha position was prepared and used to photolabel beef heart mitochondrial F1. Azido-PPi was hydrolyzed by yeast inorganic pyrophosphatase, but not by mitochondrial F1-ATPase. Incubation of F1 with [alpha-32P]azido-PPi in the dark under conditions of saturation resulted in the binding of the photoprobe to three sites, two of which exhibited a high affinity (Kd = 2 microM), the third one having a lower affinity (Kd = 300 microM). Mg2+ was required for binding. As with PPi [Issartel et al. (1987) J. Biol. Chem. 262, 13538-13544], the binding of 3 mol of azido-PPi/mol of F1 resulted in the release of one tightly bound nucleotide. ADP, AMP-PNP, and PPi competed with azido-PPi for binding to F1, but Pi and the phosphate analogue azidonitrophenyl phosphate did not. The binding of [32P]Pi to F1 was enhanced at low concentrations of azido-PPi, as it was in the presence of low concentrations of PPi. Sulfite, which is thought to bind to an anion-binding site on F1, inhibited competitively the binding of both ADP and azido-PPi, suggesting that the postulated anion-binding site of F1 is related to the exchangeable nucleotide-binding sites. Upon photoirradiation of F1 in the presence of [alpha-32P]azido-PPi, the photoprobe became covalently bound with concomitant inactivation of F1. The plots relating the inactivation of F1 to the covalent binding of the probe were rectilinear up to 50% inactivation.(ABSTRACT TRUNCATED AT 250 WORDS)     

Photoaffinity labeling of brain glutamate dehydrogenase isoproteins with an azido-ADP.

The ADP binding site within two types of bovine brain glutamate dehydrogenase isoproteins (GDH I and GDH II) was identified using photoaffinity labeling with [alpha-32P]8-azidoadenosine 5'-diphosphate (8N3ADP). 8N3ADP, without photolysis, mimicked the activatory properties of ADP on GDH I and GDH II activities, although maximal activity with 8N3ADP was about 75% of maximal ADP-stimulated activity. Saturation of photoinsertion with [alpha-32P]8N3ADP occurred at around 40 approximately 50 microM photoprobe with apparent Kd values near 25 and 40 microM for GDH I and GDH II, respectively. Photoinsertion of [alpha-32P]8N3ADP was decreased best by ADP in comparison with other nucleotides. With the combination of immobilized aluminum affinity chromatography and reversed-phase high performance liquid chromatography, photolabel-containing peptides generated by tryptic digestion were isolated. This identified a portion of the adenine ring binding domain of GDH isoproteins as in the region containing the sequence, EMSWIADTYASTIGHYDIN. Photolabeling of the peptide was prevented over 90% by the presence of 1 mM ADP during photolysis, while other nucleotides could not reduce the amount of photoinsertion as effectively as ADP. These results demonstrate selectivity of the photoprobe for the ADP binding site and suggest that the photolabeled peptide with the residues Glu179-Asn197 is within the ADP binding domain of the brain GDH isoproteins.     

A fluorescent photoprobe for the imaging of endothelin receptors

A novel fluorescent photoprobe for the imaging of endothelin A receptors (ET(A)R) was developed. Based on the nonpeptidyl, high-affinity, and selective ET(A)R antagonist 3-benzo[1,3]dioxol-5-yl-5-hydroxy-5-(4-methoxyphenyl)-4-(3,4,5-trimethoxybenzyl)-5H-furan-2-one (PD 156707), a modification of the lead structure with a PEG-spacer containing an amino moiety was performed. Labeling of this precursor with the fluorescent marker Cy 5.5 NHS-ester was accomplished by adaption of common peptide labeling procedures. The affinity of the Cy 5.5-labeled receptor antagonist was evaluated using human carcinoma cell lines with different degrees of ET(A)R expression. Fluorescence microscopy revealed that ET(A)R-positive MCF-7 human breast adenocarcinoma and HT-1080 human fibrosarcoma cells effectively bind the photoprobe at very low doses (nM), while ET(A)R-negative MDA-MB-435 human breast cancer cells showed no fluorescence signal. Binding specificity of the probe could be demonstrated by predosing with a specific ET(A)R antibody or the parent antagonist PD 156707 as a competing inhibitor. The results suggest that the modified photoprobe tightly binds to ET(A) receptors and thus may be a possible candidate for the imaging of ET(A)R-overexpressing tissues in vivo.     

Crystal structure of beta1,4-galactosyltransferase complex with UDP-Gal reveals an oligosaccharide acceptor binding site

The crystal structure of the catalytic domain of bovine beta1,4-galactosyltransferase (Gal-T1) co-crystallized with UDP-Gal and MnCl(2) has been solved at 2.8 A resolution. The structure not only identifies galactose, the donor sugar binding site in Gal-T1, but also reveals an oligosaccharide acceptor binding site. The galactose moiety of UDP-Gal is found deep inside the catalytic pocket, interacting with Asp252, Gly292, Gly315, Glu317 and Asp318 residues. Compared to the native crystal structure reported earlier, the present UDP-Gal bound structure exhibits a large conformational change in residues 345-365 and a change in the side-chain orientation of Trp314. Thus, the binding of UDP-Gal induces a conformational change in Gal-T1, which not only creates the acceptor binding pocket for N-acetylglucosamine (GlcNAc) but also establishes the binding site for an extended sugar acceptor. The presence of a binding site that accommodates an extended sugar offers an explanation for the observation that an oligosaccharide with GlcNAc at the non-reducing end serves as a better acceptor than the monosaccharide, GlcNAc. Modeling studies using oligosaccharide acceptors indicate that a pentasaccharide, such as N-glycans with GlcNAc at their non-reducing ends, fits the site best. A sequence comparison of the human Gal-T family members indicates that although the binding site for the GlcNAc residue is highly conserved, the site that binds the extended sugar exhibits large variations. This is an indication that different Gal-T family members prefer different types of glycan acceptors with GlcNAc at their non-reducing ends.     

Interaction of antigenic peptides with MHC class I molecules on living cells studied by photoaffinity labeling.

Using a direct binding assay based on photoaffinity labeling, we have studied the interaction of antigenic peptides with murine MHC class I molecules on living cells. Photoreactive derivatives were prepared by N-terminal amidation with iodo, 4-azido salicylic acid of the Kd restricted Plasmodium berghei circumsporozoite (P.b. CS) peptide 253-260 (YIPSAEKI) and the Db-restricted Adenovirus 5 early region 1A (Ad5 E1A) peptide 234-243 (SGPSNTPPEI). As assessed in functional competition experiments, both peptide derivatives retained the specific binding activity of the parental peptides for Kd or Dd, respectively. The P.b. CS photoprobe specifically labeled Kd molecules on P815 (H-2d) cells, but failed to label RMA (H-2b) cells. Conversely, the Ad5 E1A photoprobe specifically labeled Db molecules on RMA cells, but failed to label P815 cells. When the two photoprobes were tested on a panel of Con A-activated spleen cells expressing 10 different H-2 haplotypes, significant photoaffinity labeling was observed only on H-2d cells with the P.b. CS photoprobe and on H-2b cells with the Ad5 E1A photoprobe. Labeling of cell-associated Kd or Db molecules with the photoprobes was specifically inhibited by antigenic peptides known to be presented by the same class I molecule. Photoaffinity labeling of Kd with the P.b. CS photoprobe was used to study the dynamics of peptide binding on living P815 cells. Binding increased steadily with the incubation period (up to 8 h) at 37 degrees C and at ambient temperature, but was greatly reduced (greater than 95%) at 0 to 4 degrees C or in the presence of ATP synthesis inhibitors. The magnitude of the labeling was twofold higher at room temperature than at 37 degrees C. In contrast, binding to isolated Kd molecules in solution rapidly reached maximal binding, particularly at 37 degrees C. Dissociation of the photoprobe from either cell-associated or soluble Kd molecules was similar, with a half time of approximately 1 h at 37 degrees C, whereas the complexes were long-lived at 4 degrees C in both instances.     

Recognition of gene acceptor site based on multi-objective optimization

A new method for predicting the gene acceptor site based on multi-objective optimization is introduced in this paper. The models for the acceptor, branch and distance between acceptor site and branch site were constructed according to the characteristics of the sequences from the exon-intron database and using common biological knowledge. The acceptor function, branch function and distance function were defined respectively, and the multi-objective optimization model was constructed to recognize the splice site. The test results show that the algorithm used in this study performs better than the SplicePredictor,which is one of the leading acceptor site detectors.     

Probing the donor and acceptor substrate specificity of the γ-glutamyl transpeptidase

γ-Glutamyl transpeptidase (GGT) is a two-substrate enzyme that plays a central role in glutathione metabolism and is a potential target for drug design. GGT catalyzes the cleavage of γ-glutamyl donor substrates and the transfer of the γ-glutamyl moiety to an amine of an acceptor substrate or water. Although structures of bacterial GGT have revealed details of the protein-ligand interactions at the donor site, the acceptor substrate site is relatively undefined. The recent identification of a species-specific acceptor site inhibitor, OU749, suggests that these inhibitors may be less toxic than glutamine analogues. Here we investigated the donor and acceptor substrate preferences of Bacillus anthracis GGT (CapD) and applied computational approaches in combination with kinetics to probe the structural basis of the enzyme's substrate and inhibitor binding specificities and compare them with human GGT. Site-directed mutagenesis studies showed that the R432A and R520S variants exhibited 6- and 95-fold decreases in hydrolase activity, respectively, and that their activity was not stimulated by the addition of the l-Cys acceptor substrate, suggesting an additional role in acceptor binding and/or catalysis of transpeptidation. Rat GGT (and presumably HuGGT) has strict stereospecificity for L-amino acid acceptor substrates, while CapD can utilize both L- and D-acceptor substrates comparably. Modeling and kinetic analysis suggest that R520 and R432 allow two alternate acceptor substrate binding modes for L- and D-acceptors. R432 is conserved in Francisella tularensis, Yersinia pestis, Burkholderia mallei, Helicobacter pylori and Escherichia coli, but not in human GGT. Docking and MD simulations point toward key residues that contribute to inhibitor and acceptor substrate binding, providing a guide to designing novel and specific GGT inhibitors.     

The mechanism of action of ethanolamine ammonia-lyase, an adenosylcobalamin-dependent enzyme. Evidence that the hydrogen transfer mechanism involves a second intermediate hydrogen carrier in addition to the cofactor

Ethanolamine ammonia-lyase catalyzes the adenosylcobalamin (AdoCbl)-dependent conversion of ethanolamine to acetaldehyde and ammonia. During this reaction, a hydrogen atom migrates from the carbinol carbon of ethanolamine to the methyl carbon of acetaldehyde. Previous studies have shown that this migrating hydrogen equilibrates with the hydrogens on the 5'-(cobalt-linked) carbon of the cofactor. On the basis of those studies, a two-step mechanism for hydrogen transfer has been postulated in which the migrating hydrogen is first transferred from the substrate to the cofactor, then in a subsequent step is returned from the cofactor to the product. We now show that this migrating hydrogen is transferred not only to the cofactor, but also to a second acceptor at the active site. Hydrogens on this acceptor do not exchange with water during the course of the reaction, but are released to water when the enzyme is denatured. The catalytic significance of this second hydrogen acceptor was demonstrated by the findings that the transfer of hydrogen to this acceptor required both AdoCbl and active enzyme and that hydrogen at the second acceptor site could be washed out by unlabeled ethanolamine. On the basis of these results, we propose an expanded hydrogen transfer mechanism in which AdoCbl and the second acceptor site serve as alternative intermediate hydrogen carriers during the course of ethanolamine deamination.     

Identification of two tamoxifen target proteins by photolabeling with 4-(2-morpholinoethoxy)benzophenone

Our quest to identify target proteins involved in the activity of tamoxifen led to the design of photoaffinity ligand analogues of tamoxifen able to cross-link such proteins. A new tritiated photoprobe, 4-(2-morpholinoethoxy)benzophenone (MBoPE), was synthesized and used to identify proteins involved in tamoxifen binding in rat liver. MBoPE, which has structural features in common with the potential antagonist of the intracellular histamine receptor (N,N-diethyl-2-[(4-phenylmethyl)phenoxy]ethanamine HCl: DPPE) is unable to bind the estrogen receptor although it does compete with tamoxifen for an antiestrogen binding site (AEBS). This tritiated benzophenone derivative was obtained by metal-catalyzed halogen-tritium replacement reaction. Because of its high specific activity, four target proteins could be photolabeled, three of which were identified with M(r) of 60,000, 49,500, and 14,000, while the fourth at 27,500 was in too low an amount and could not be sequenced. The 49.5 kDa protein corresponded by mass spectrometry to the microsomal epoxide hydrolase already identified with an aryl azide photoprobe [Mesange, F., et al. (1998) Biochem. J. 334, 107-112]. The 60 and 14 kDa proteins were identified as the carboxylesterase (ES10) and the liver fatty acid binding protein (L-FABP), respectively. The inhibitory effect of tamoxifen on carboxylesterase activity and the competitive efficacy of oleic acid on [(3)H]tamoxifen binding suggest that both proteins are AEBS subunits. Moreover, treatment of hepatocytes with antisense mRNA directed against ES10 or L-FABP abolished both tamoxifen and MBoPE binding. On the basis of previous pharmacological arguments, the 27.5 kDa protein might correspond to the sigma I receptor. Altogether, these results confirm that the microsomal epoxide hydrolase is a target for tamoxifen and provide evidence of two new target proteins implicated in cell lipid metabolism.