Restructuring of focal adhesion plaques by PI 3-kinase. Regulation by PtdIns (3,4,5)-p(3) binding to alpha-actinin.

Focal adhesions are an elaborate network of interconnecting proteins linking actin stress fibers to the extracellular matrix substrate. Modulation of the focal adhesion plaque provides a mechanism for the regulation of cellular adhesive strength. Using interference reflection microscopy, we found that activation of phosphoinositide 3-kinase (PI 3-kinase) by PDGF induces the dissipation of focal adhesions. Loss of this close apposition between the cell membrane and the extracellular matrix coincided with a redistribution of alpha-actinin and vinculin from the focal adhesion complex to the Triton X-100-soluble fraction. In contrast, talin and paxillin remained localized to focal adhesions, suggesting that activation of PI 3-kinase induced a restructuring of the plaque rather than complete dispersion. Furthermore, phosphatidylinositol (3,4, 5)-trisphosphate (PtdIns (3,4,5)-P(3)), a lipid product of PI 3-kinase, was sufficient to induce restructuring of the focal adhesion plaque. We also found that PtdIns (3,4,5)-P(3) binds to alpha-actinin in PDGF-treated cells. Further evidence demonstrated that activation of PI 3-kinase by PDGF induced a decrease in the association of alpha-actinin with the integrin beta subunit, and that PtdIns (3,4,5)-P(3) could disrupt this interaction in vitro. Modification of focal adhesion structure by PI 3-kinase and its lipid product, PtdIns (3,4,5)-P(3), has important implications for the regulation of cellular adhesive strength and motility.

Chemistry of pheromone and hormone metabolism in insects.

Chemical evidence is needed in both insect endocrinology and sensory physiology to understand hormone and pheromone action at the molecular level. Radiolabeled pheromones and hormones have been synthesized and used to identify binding and catabolic proteins from insect tissues. Chemically modified analogs, including photoaffinity labels and enzyme inhibitors, are among the tools used to covalently modify the specific acceptor or catalytic sites. Such targeted agents can also provide leads for the design of growth and mating disruptants by allowing manipulation of the physiologically important interactions of the chemical signals with macromolecules.

Chemical self-defense by termite workers: prevention of autotoxication in two rhinotermitids.

Soldiers of the lower termites Prorhinotermes simplex and Schedorhinotermes lamanianus (Isoptera, Rhinotermitidae) have electrophilic contact poisons used in colony defense. Workers of these termites die when exposed to the defense secretion of the other species, but survive when exposed to chemicals from conspecific soldiers. Detoxication occurs by an initial substrate-specific reduction of the electron-deficient double bond of the nitroalkene (Prorhinotermes simplex) or vinyl ketone (Schedorhinotermes lamanianus) followed by complete catabolism to acetate.

Chemical ecology: studies from East Africa.

The International Centre of Insect Physiology and Ecology (ICIPE), in Nairobi, provides a laboratory at which a multinational group of scientists pursues interdisciplinary research. In collaboration with their colleagues in biology, ICIPE chemists have characterized the sex pheromones of the tick which serves as a vector of East Coast fever and have identified a termite queen-cell-building pheromone. The structure of many anthropod defensive chemicals have been determined; most interesting of these are the trinervitenes, structurally novel diterpenoids from nasute termites. Several highly active insect antifeedants were discovered using a simple bioassay to screen selected East African plants. These antifeedants may provide leads for the development of new insect-control techniques.

Lysophosphatidic acid-induced platelet shape change revealed through LPA(1-5) receptor-selective probes and albumin.

Lysophosphatidic acid (LPA), a component of mildly-oxidized LDL and the lipid rich core of atherosclerotic plaques, elicits platelet activation. LPA is the ligand of G protein-coupled receptors (GPCR) of the EDG family (LPA(1-3)) and the newly identified LPA(4-7) subcluster. LPA(4), LPA(5) and LPA(7) increase cellular cAMP levels that would induce platelet inhibition rather than activation. In the present study we quantified the mRNA levels of the LPA(1-7) GPCR in human platelets and found a rank order LPA(4) = LPA(5) > LPA(7) > LPA(6) = LPA(2) >> LPA(1) > LPA(3). We examined platelet shape change using a panel of LPA receptor subtype-selective agonists and antagonists and compared them with their pharmacological profiles obtained in heterologous LPA(1-5) receptor expression systems. Responses to different natural acyl and alkyl species of LPA, and octyl phosphatidic acid analogs, alpha-substituted phosphonate analogs, N-palmitoyl-tyrosine phosphoric acid, N-palmitoyl-serine phosphoric acid were tested. All of these compounds elicited platelet activation and also inhibited LPA-induced platelet shape change after pre-incubation, suggesting that receptor desensitization is likely responsible for the inhibition of this response. Fatty acid free albumin (10 microM) lacking platelet activity completely inhibited platelet shape change induced by LPA with an IC(50) of 1.1 microM but had no effect on the activation of LPA(1,2,3,&5) expressed in endogenously non-LPA-responsive RH7777 cells. However, albumin reduced LPA(4) activation and shifted the dose-response curve to the right. LPA(5) transiently expressed in RH7777 cells showed preference to alkyl-LPA over acyl-LPA that is similar to that in platelets. LPA did not increase cAMP levels in platelets. In conclusion, our results with the pharmacological compounds and albumin demonstrate that LPA does not induce platelet shape change simply through activation of LPA(1-5), and the receptor(s) mediating LPA-induced platelet activation remains elusive.

PDGF induces an early and a late wave of PI 3-kinase activity, and only the late wave is required for progression through G1.

BACKGROUND: Platelet-derived growth factor (PDGF) triggers cytoskeletal rearrangements and chemotaxis within minutes. These events are at least in part due to the activation of phosphoinositide (PI) 3-kinase; there is good temporal correlation between these events and the accumulation of 3-phosphorylated products of the kinase. Prolonged and continuous PDGF exposure results in S-phase entry many hours after the initial burst of activity. Although early signals appear responsible for the early responses, they may not fully account for later responses, such as cell-cycle progression. RESULTS: We assessed when PI 3-kinase products accumulate in PDGF-stimulated cells. In addition to the previously identified early accumulation of products, we detected a second, prolonged wave of accumulation 3-7 hours after stimulation. To determine the relative contribution of each phase to PDGF-dependent DNA synthesis, we first developed an assay in which synthetic 3-phosphorylated lipids were used to rescue DNA synthesis in cells expressing a PDGF-receptor mutant. The lipids rescued DNA synthesis only when added 2-6 hours after PDGF. In addition, PI 3-kinase inhibitors failed to block PDGF-dependent DNA synthesis if added during the first wave of PI 3-kinase activity, but adding them later, in G1 phase, prevented PDGF-dependent cell-cycle progression. CONCLUSIONS: PDGF induces distinct waves of PI 3-kinase activity. The second wave is required for PDGF-dependent DNA synthesis, whereas the initial wave is not. One of the ways in which cells use PI 3-kinase to mediate distinct cellular responses seems to be by regulating when its products accumulate.

GCS1, an Arf guanosine triphosphatase-activating protein in Saccharomyces cerevisiae, is required for normal actin cytoskeletal organization in vivo and stimulates actin polymerization in vitro.

Recent cloning of a rat brain phosphatidylinositol 3,4, 5-trisphosphate binding protein, centaurin alpha, identified a novel gene family based on homology to an amino-terminal zinc-binding domain. In Saccharomyces cerevisiae, the protein with the highest homology to centaurin alpha is Gcs1p, the product of the GCS1 gene. GCS1 was originally identified as a gene conditionally required for the reentry of cells into the cell cycle after stationary phase growth. Gcs1p was previously characterized as a guanosine triphosphatase-activating protein for the small guanosine triphosphatase Arf1, and gcs1 mutants displayed vesicle-trafficking defects. Here, we have shown that similar to centaurin alpha, recombinant Gcs1p bound phosphoinositide-based affinity resins with high affinity and specificity. A novel GCS1 disruption strain (gcs1Delta) exhibited morphological defects, as well as mislocalization of cortical actin patches. gcs1Delta was hypersensitive to the actin monomer-sequestering drug, latrunculin-B. Synthetic lethality was observed between null alleles of GCS1 and SLA2, the gene encoding a protein involved in stabilization of the actin cytoskeleton. In addition, synthetic growth defects were observed between null alleles of GCS1 and SAC6, the gene encoding the yeast fimbrin homologue. Recombinant Gcs1p bound to actin filaments, stimulated actin polymerization, and inhibited actin depolymerization in vitro. These data provide in vivo and in vitro evidence that Gcs1p interacts directly with the actin cytoskeleton in S. cerevisiae.

Fluorescently labeled neomycin as a probe of phosphatidylinositol-4, 5-bisphosphate in membranes.

Phosphatidylinositol-4,5-bisphosphate (PI(4,5)P(2)), a minor component of the plasma membrane, is important in signal transduction, exocytosis, and ion channel activation. Thus fluorescent probes suitable for monitoring the PI(4,5)P(2) distribution in living cells are valuable tools for cell biologists. We report here three experiments that show neomycin labeled with either fluorescein or coumarin can be used to detect PI(4,5)P(2) in model phospholipid membranes. First, addition of physiological concentrations of PI(4,5)P(2) (2%) to lipid vesicles formed from mixtures of phosphatidylcholine (PC) and phosphatidylserine (PS) enhances the binding of labeled neomycin significantly (40-fold for 5:1 PC/PS vesicles). Second, physiological concentrations of inositol-1,4,5-trisphosphate (10 microM I(1,4,5)P(3)) cause little translocation of neomycin from PC/PS/PI(4,5)P(2) membranes to the aqueous phase, whereas the same concentrations of I(1,4,5)P(3) cause significant translocation of the green fluorescent protein/phospholipase C-delta pleckstrin homology (GFP-PH) constructs from membranes (Hirose et al., Science, 284 (1999) 1527). Third, fluorescence microscopy observations confirm that one can distinguish between PC/PS vesicles containing either 0 or 2% PI(4, 5)P(2) by exposing a mixture of the vesicles to labeled neomycin. Thus fluorescently labeled neomycin could complement GFP-PH constructs to investigate the location of PI(4,5)P(2) in cell membranes.

Site-directed mutagenesis of squalene-hopene cyclase: altered substrate specificity and product distribution.

BACKGROUND: Two regions of squalene-hopene cyclase (SHC) were examined to define roles for motifs posited to be responsible for initiation and termination of the enzyme-catalyzed polyolefinic cyclizations. Specifically, we first examined the triple mutant of the DDTAVV motif, a region deeply buried in the catalytic cavity and thought to be responsible for the initiation of squalene cyclization. Next, four mutants were prepared for Glu45, a residue close to the substrate entrance channel proposed to be involved in the termination of the cyclization of squalene. RESULTS: The DDTAVV motif in SHC was changed to DCTAEA, the corresponding conserved region of eukaryotic oxidosqualene cyclase (OSC), by the triple mutation of D377C/V380E/V381A; selected single mutants were also examined. The triple mutant showed no detectable cyclization of squalene, but effectively cyclized 2,3-oxidosqualene to give mono- and pentacyclic triterpene products. Of the Glu45 mutants, E45A and E45D showed reduced activity, E45Q showed slightly increased activity, and E45K was inactive. A normal yield of pentacyclic products was produced, but the ratio of hopene 2 to hopanol 3 was significantly changed in the less active mutants. CONCLUSIONS: Initiation and substrate selectivity may be determined by the interaction of the DDTAVV motif with the isopropylidene of squalene (for SHC) and of the DCTAEA motif with the epoxide of oxidosqualene (for OSC). This is the first report of a substrate switch determined by a central catalytic motif in a triterpenoid cyclase. At the termination of cyclization, the product ratio may be largely controlled by Glu45 at the entrance channel to the active site.

The binding site for an inhibitor of squalene:hopene cyclase determined using photoaffinity labeling and molecular modeling.

BACKGROUND: The squalene:hopene cyclases (SHCs) are bacterial enzymes that convert squalene into hopanoids, a function analogous to the action of oxidosqualene cyclases (OSCs) in eukaryotic steroid and triterpenoid biosynthesis. We have identified the binding site for a selective, potent, photoactivatable inhibitor of an SHC. RESULTS: SHC from Alicyclobacillus acidocaldarius was specifically labeled by [3H]Ro48-8071, a benzophenone-containing hypocholesteremic drug. Edman degradation of a peptide fragment of covalently modified SHC confirmed that Ala44 was specifically modified. Molecular modeling, using X-ray-derived protein coordinates and a single point constraint for the inhibitor, suggested several geometries by which Ro48-8071 could occupy the active site. CONCLUSIONS: A covalent complex of a potent inhibitor with a squalene cyclase has been characterized. The amino acid modification and molecular modeling suggest that Ro48-8071 binds at the junction between the central cavity and substrate entry channel, therefore inhibiting access of the substrate to the active site.

Probing the phosphoinositide 4,5-bisphosphate binding site of human profilin I.

BACKGROUND: Profilin is a widely and highly expressed 14 kDa protein that binds actin monomers, poly(L-proline) and polyphosphoinositol lipids. It participates in regulating actin-filament dynamics that are essential for many types of cell motility. We sought to investigate the site of interaction of profilin with phosphoinositides. RESULTS: Human profilin I was covalently modified using three tritium-labeled 4-benzoyldihydrocinnamoyl (BZDC)-containing photoaffinity analogs of phosphatidylinositol 4,5-bisphosphate (PtdIns(4,5)P2). The P-1-tethered D-myoinositol 1,4,5-trisphosphate (Ins(1,4,5)P3) modified profilin I efficiently and specifically; the covalent labeling could be displaced by co-incubation with an excess of PtdIns(4,5)P2 but not with Ins(1,4,5)P3. The acyl-modified PtdIns(4,5)P2 analog showed little protein labeling even at very low concentrations, whereas the head-group-modified PtdIns(4,5)P2 phosphotriester-labeled monomeric and oligomeric profilin. Mass spectroscopic analyses of CNBr digests of [3H]BZDC-Ins(1,4,5)P3-modified recombinant profilin suggested that modification was in the amino-terminal helical CNBr fragment. Edman degradation confirmed Ala1 of profilin I (residue 4 of the recombinant protein) was modified. Molecular models show a minimum energy conformation in which the hydrophobic region of the ligand contacts the amino-terminal helix whereas the 4,5-bisphosphate interacts with Arg135 and Arg136 of the carboxy-terminal helix. CONCLUSIONS: The PtdIns(4,5)P2-binding site of profilin I includes a bisphosphate interaction with a base-rich motif in the carboxy-terminal helix and contact between the lipid moiety of PtdIns(4,5)P2 and a hydrophobic region of the aminoterminal helix of profilin. This is the first direct evidence for a site of interaction of the lipid moiety of a phosphoinositide bisphosphate analog with profilin.

Peptides that mimic glycosaminoglycans: high-affinity ligands for a hyaluronan binding domain.

BACKGROUND: Hyaluronan (HA) is a non-sulfated glycosaminoglycan (GAG) that promotes motility, adhesion, and proliferation in mammalian cells, as mediated by cell-surface HA receptors. We sought to identify non-carbohydrate ligands that would bind to and activate cell-surface HA receptors. Such analogs could have important therapeutic uses in the treatment of cancer, wound healing, and arthritis, since such ligands would be resistant to degradation by hyaluronidase (HAse). RESULTS: Peptide ligands that bind specifically to the recombinant HA binding domain (BD) of the receptor for hyaluronan-mediated motility (RHAMM) were obtained by screening two peptide libraries: (i) random 8-mers and (ii) biased 8-mers with alternating acidic side chains, i.e. XZXZXZXZ (X=all-L-amino acids except Cys, Lys, or Arg; Z=D-Asp, L-Asp, D-Glu, or L-Glu). Selectivity of the peptide ligands for the HABD was established by (i) detection of binding of biotin- or fluorescein-labeled peptides to immobilized proteins and (ii) fluorescence polarization of FITC-labeled peptides with the HABD in solution. HA competitively displaced binding of peptides to the HABD, while other GAGs were less effective competitors. The stereochemistry of four biased octapeptides was established by synthesis of the 16 stereoisomers of each peptide. Binding assays demonstrated a strong preference for alternating D and L configurations for the acidic residues, consistent with the calculated orientation of glucuronic acid moieties of HA. CONCLUSIONS: Two classes of HAse-resistant peptide mimetics of HA were identified with high affinity, HA-compatible binding to the RHAMM HABD. This demonstrated that non-HA ligands specific to a given HA binding protein could be engineered, permitting receptor-specific targeting.

Cancer-targeted polymeric drugs.

A major challenge in cancer chemotherapy is the selective delivery of small molecule anti cancer agents to tumor cells. Water-soluble polymer-drug conjugates exhibit good water solubility, increased half-life, and potent anti tumor effects. By localizing the drug at the desired site of action, macromolecular therapeutics have improved efficacy and enhanced safety at lower doses. Since small molecule drugs and macromolecular drugs enter cells by different pathways, multi-drug resistance (MDR) can be minimized. Anti-cancer polymer-drug conjugates can be divided into two targeting modalities: passive and active. Tumor tissues have anatomic characteristics that differ from normal tissues. Macromolecules penetrate and accumulate preferentially in tumors relative to normal tissues, leading to extended pharmacological effects. This "enhanced permeability and retention" (EPR) effect is the principal reason for current successes with macromolecular anti-cancer drugs. Both natural and synthetic polymers have been used as drug carriers, and several bioconjugates have been clinically approved or are in human clinical trials. While clinically useful anti-tumor activity has been achieved using passive macromolecular drug delivery systems, further selectivity is possible by active targeting. Attachment of targeting moieties to the polymer backbone can further exploit differences between cancer and normal cells through selective receptor-mediated endocytosis. This strategy would augment the EPR effect, thereby further improving the therapeutic index of the macromolecular drug. This review discusses the development and therapeutic potential of prototype macromolecular drugs for use in cancer chemotherapy. Specific examples are selected to illustrate the basic design principles for soluble polymeric drug delivery systems.

Using photolabile ligands in drug discovery and development.

Photoactivatable ligands are important tools used in drug discovery and drug development. These ligands enable researchers to identify the targets of drugs, to determine the affinity and selectivity of the drug-target interaction, and to identify the binding site on the target. Examples are presented from three fundamentally different approaches: (1) photoaffinity labeling of target macromolecules; (2) photoactivation and release of 'caged ligands'; and (3) photoimmobilization of ligands onto surfaces.

Bacterial expression and photoaffinity labeling of a pheromone binding protein.

The first high-level production of a binding-active odorant binding protein is described. The expression cassette polymerase chain reaction was used to generate a DNA fragment encoding the pheromone binding protein (PBP) of the male moth Antheraea polyphemus. Transformation of Escherichia coli cells with a vector containing this construct generated clones which, when induced with isopropyl beta-D-thiogalactopyranoside, produced the 14-kDa PBP in both the soluble fraction and in inclusion bodies. Purification of the soluble recombinant PBP by preparative isoelectric focusing and gel filtration gave > 95% homogeneous protein, which was immunoreactive with an anti-PBP antiserum and exhibited specific, pheromone-displaceable covalent modification by the photoaffinity label [3H]6E,11Z-hexadecadienyl diazoacetate. Recombinant PBP was indistinguishable from the insect-derived PBP, as determined by both native and denaturing gel electrophoresis, immunoreactivity, and photoaffinity labeling properties. Moreover, the insoluble inclusion body protein could be solubilized, refolded, and purified by the same procedures to give a recombinant PBP indistinguishable from the soluble PBP. Proton NMR spectra of the soluble and refolded protein provide further evidence that they possess the same folded structure.

Protein farnesyltransferase: production in Escherichia coli and immunoaffinity purification of the heterodimer from Saccharomyces cerevisiae.

Protein farnesylation in Saccharomyces cerevisiae is mediated by a heterodimeric enzyme, protein farnesyltransferase (PFTase), encoded by the genes RAM1 and RAM2. A series of plasmids for the expression of RAM1 and RAM2 in Escherichia coli was prepared and evaluated. Maximal production of functional PFTase was seen in strains containing a multicopy plasmid with a synthetic operon in which the RAM1 and RAM2 structural genes were translationally coupled by overlapping TAATG stop-start codons and by locating a ribosome-binding site near the 3' end of the upstream gene. This was accomplished by an insertional mutation at the 3'-end of RAM1 that embedded an AGGAGGAG sequence within codons for the tetrapeptide, QEEF, added to the end of the Ram1 protein. The QEEF C-terminal motif in the Ram1 subunit of PFTase facilitated purification of the enzyme by immunoaffinity chromatography on an anti-alpha-tubulin column prepared using monoclonal antibodies that recognized a tripeptide EEF epitope. Heterodimeric recombinant yeast PFTase::QEEF (re-PFTase::QEEF) constituted approximately 4% of total soluble protein in induced cells and was readily purified 25-fold in two steps by ion exchange and immunoaffinity chromatography in an overall 25% yield. Michaelis constants for farnesyl diphosphate (FPP) and Hras protein (modified to contain a yeast a-mating factor PACVIA sequence at the C terminus) were 5.5 and 15 microM, respectively; the kcat was 0.7 s-1.

Detection and microsequencing of juvenile hormone-binding proteins of an insect by the use of an iodinated juvenile hormone analog.

An [125I]iodinated juvenile hormone (JH) analog can be used as a sensitive and highly selective probe for the visualization of high-affinity, (JH)-specific binding proteins from insect hemolymph samples. The proteins can be detected in their native form using a two-dimensional (isoelectric focusing then native gradipore gel) separation of the crude protein mixture containing the 125I-labeled iodinated JH analog. The proteins can be transferred to activated glass fiber paper by electroblotting, and the location of the bound gamma-emitter can be found by exposure of the dried gel or the electroblot to X-ray film. The radiolabeled protein spot can be excised from the Coomassie-stained glass fiber paper and subjected directly to gas-phase N-terminal amino acid sequencing. This non-destructive, non-denaturing technique may have wide applicability in identifying and sequencing ligand-specific binding proteins in complex mixtures.

Squalene analogues containing isopropylidene mimics as potential inhibitors of pig liver squalene epoxidase and oxidosqualene cyclase.

Several squalene analogues containing 1,1-dihaloalkene, acetylene, allene, diene, and cyclopropane functionalities were synthesized and evaluated as potential inhibitors of pig liver squalene epoxidase and oxidosqualene cyclase. Both monofunctionalized and bisfunctionalized analogues were prepared. Poor inhibition of squalene epoxidase and oxidosqualene cyclase was found for most compounds (IC50 much greater than 400 microM), with the exception of the monofunctionalized alkynol (IC50 = 300 microM). This alkynol showed mixed-function inhibition with KI = 0.95 mM. Oxidation of the alcohol to the alkynone resulted in loss of epoxidase activity, indicating that the hydroxyl group is necessary for inhibition and that the alkynol is not a proinhibitor. Molecular mechanics calculations indicated that a good inhibitor should possess hydrophobic substituents on an unpolarized, unsaturated system; additionally, the presence of a pro-C-3 hydroxyl group can confer inhibitory potency.

Inhibition of vertebrate squalene epoxidase by extended and truncated analogues of trisnorsqualene alcohol.

The epoxidation of squalene to (3S)-2,3-epoxysqualene and subsequent cyclization to lanosterol are keys steps in vertebrate cholesterol biosynthesis. Trisnorsqualene alcohol (TNSA) has previously been reported as a potent inhibitor of vertebrate squalene epoxidase, with IC50 = 4 microM for pig liver (J. Am. Chem. Soc. 1989, 111, 1508-1510). Analogues with extended and truncated carbon skeletons have been prepared and tested for pig liver squalene epoxidase (SE) inhibition. Most of the structural analogues were poor inhibitors of vertebrate SE, with the exception of bisnorsqualene alcohol which had the same activity as TNSA. These results support the theory that an intact trisnorsqualene moiety is required for activity.

Synthesis and inhibition studies of sulfur-substituted squalene oxide analogues as mechanism-based inhibitors of 2,3-oxidosqualene-lanosterol cyclase.

The synthesis and biological evaluation of three new sulfur-substituted oxidosqualene (OS) analogues (1-3) are presented. In these analogues, C-11, C-15, or C-18 in the OS skeleton was replaced by sulfur. The sulfur position in the OS skeleton was chosen to disrupt one or more key processes involved in cyclization: (a) the folding of the B-ring into a boat conformation, (b) the anti-Markovnikov cyclization leading to the C-ring, or (c) the formation of the D-ring during the lanosterol biosynthesis. Enzyme inhibition kinetics using homogeneous mammalian oxidosqualene cyclases (OSC) were also examined for the previously reported S-19 analogue 4. The four analogues were potent inhibitors of mammalian OSCs (IC50 = 0.05-2.3 microM for pig and rat liver OSC) and fungal cell-free Candida albicans OSC (submicromolar IC50 values). In particular, the S-18 analogue 3 showed the most potent inhibition toward the rat liver enzyme (IC50 = 50 nM) and showed potent, selective inhibition against the fungal enzyme (IC50 = 0.22 nM, 10-fold more potent than the S-19 analogue 4). Thus, 3 is the most potent OSC inhibitor known to date. The Ki values ranged from 0.5 to 4.5 microM for pig OSC, with 3 and 4 showing about 10-fold higher potency for rat liver OSC. Interestingly, the S-18 analogue 3 showed time-dependent irreversible inhibition with homogeneous pig liver OSC (kinact = 0.06 min-1) but not with rat OSC.