Blocking the synthesis of the AD pathogen beta amyloid peptide (Αβ) through inhibition of APP processing by gamma secretase (GS) is a promising therapeutic strategy that may provide the first disease-modifying anti-Alzheimer's drug (DMAAD). In animal models, GSI-953 is a Notch-sparing gamma secretase inhibitor (GSI) and this compound has subsequently entered clinical trials. Our objective was to characterize the relative affinity of GSI-953 for binding to GS and compare it to the potency of GS inhibition. A cell-free GS binding assay was developed employing membranes isolated from human neuroblastoma SY5Y cells and the tritiated GSI-953 analog TGSI (5-chloro-N-[(1S,2R)–4,4,4-trifluoro-1-(hydroxyl[3H2]methyl)–2-methylbutyl]thiophene-2-sulfonamide). GSI-953 and benchmark GSIs (DAPT, LY411575, LY450139, DuPont E, L-685458, a BMS sulfonamide and an Amgen sulfonamide) were profiled in this assay for their ability to displace the radiolabeled GSI-953 analog. For comparison to GS binding affinity, the GSIs were also profiled for inhibition of Αβ synthesis in a cellular assay (hAPPCHO cells). GSI-953 competitively displaced TGSI from the GS complex in the cell-free binding assay (IC50=8 nM) and its affinity for GS was comparable to its GSI potency in the cellular assay (EC50Αβ42=15 nM). The stereospecific nature of both the binding to and inhibition of GS by GSI-953 was confirmed by the low affinity (IC50>10,000 nM) and GSI activity (EC50Αβ42>30,000 nM) observed for the enantiomer of GSI-953. Benchmark GSIs DAPT (IC50=29 nM), LY411575 (IC50=3 nM), LY450139 (IC50=26 nM), DuPont E (IC50<10 nM), a BMS sulfonamide (IC50=57 nM) and an Amgen sulfonamide (IC50=986 nM) were able to competitively displace TGSI from GS at concentrations comparable to their EC50s for inhibition of Αβ synthesis. The transition state inhibitor L-685458 was only able to partially displace TGSI. GSI-953 and benchmark GSIs (except L-685458) can competitively displace TGSI from GS suggesting that these GSIs may bind to the same site. Among the GSIs profiled, a good correlation was observed between GS binding affinity and GSI potency.

Journal of Labelled Compounds and RadiopharmaceuticalsVolume 50, Issue 5-6 p. 578-579 Short Research Article Carbon-14 labelling of 3-cyanoquinolines† John D. Olszewski, Corresponding Author John D. Olszewski [email protected] Chemical Development—Radiosynthesis, Wyeth Research, 401 N. Middletown Road, Pearl River, New York 10965, USAWyeth Research, Chemical Development—Radiosynthesis, 401 N. Middletown Road, Pearl River, New York 10965, USASearch for more papers by this authorDan M. Berger, Dan M. Berger Chemical and Screening Sciences, Wyeth Research, 401 N. Middletown Road, Pearl River, New York 10965, USASearch for more papers by this authorMichael K. May, Michael K. May Chemical Development—Radiosynthesis, Wyeth Research, 401 N. Middletown Road, Pearl River, New York 10965, USASearch for more papers by this author John D. Olszewski, Corresponding Author John D. Olszewski [email protected] Chemical Development—Radiosynthesis, Wyeth Research, 401 N. Middletown Road, Pearl River, New York 10965, USAWyeth Research, Chemical Development—Radiosynthesis, 401 N. Middletown Road, Pearl River, New York 10965, USASearch for more papers by this authorDan M. Berger, Dan M. Berger Chemical and Screening Sciences, Wyeth Research, 401 N. Middletown Road, Pearl River, New York 10965, USASearch for more papers by this authorMichael K. May, Michael K. May Chemical Development—Radiosynthesis, Wyeth Research, 401 N. Middletown Road, Pearl River, New York 10965, USASearch for more papers by this author First published: 30 July 2007 https://doi.org/10.1002/jlcr.1284Citations: 1 † Proceedings of the Ninth International Symposium on the Synthesis and Applications of Isotopically Labelled Compounds, Edinburgh, 16–20 July 2006. AboutPDF ToolsRequest permissionExport citationAdd to favoritesTrack citation ShareShare Give accessShare full text accessShare full-text accessPlease review our Terms and Conditions of Use and check box below to share full-text version of article.I have read and accept the Wiley Online Library Terms and Conditions of UseShareable LinkUse the link below to share a full-text version of this article with your friends and colleagues. Learn more.Copy URL No abstract is available for this article. REFERENCES 1 Berger D, Dutia M, Powell D, Wu B, Wissner A, Boschelli DH, Floyd MB, Zhang N, Torres N, Levin J, Du X, Wojciehowicz D, Discafani C, Kohler C, Kim SC, Feldberg LR, Collins K, Mallon R. Bioorg Med Chem Lett 2003; 13: 3031– 3034. 2 Boschelli DH. Curr Topics Med Chem 2002; 2: 1051– 1063 and references therein. 3 Carr RM, Sutherland DR. J Label Compd Radiopharm 1994; 34(10): 961– 971. 4 Thurston DE, Varanasi SM, Langley DR, Jones GB. Synthesis 1990; 1: 81– 84. 5 Wang T, Lui AS, Cloudsdale IS. Org Lett 1999; 1(11): 1835– 1837. 6 Koltai E, Zolyomi G, Komaromy P, Banfi D, Szuts T, Takacs K. J Label Compd Radiopharm 1981; 18(8): 1107– 1113. 7 Shelkov R, Nahmany M, Melman A. J Org Chem 2002; 67(25): 8975– 8982. 8 Wolfe JP, Tomori H, Sadighi JP, Yin J, Buchwald SL. J Org Chem 2000; 65(4): 1158– 1174 and references therein. 9 Wu Y-J, Boissard CG, Greco C, Gribkoff VK, Harden DG, He H, L'Heureux A, Kang SH, Kinney CG, Knox RJ, Natale J, Newton AE, Lehtinen-Oboma S, Sinz MW, Sivarao DV, Starrett Jr. JE, Sun L-Q, Tertyshnikova S, Thompson MW, Weaver D, Wong HS, Zhang L, Dworetzky SI. J Med Chem 2003; 46(15): 3197– 3200. Citing Literature Volume50, Issue5-6Special Issue: Proceedings of the Ninth International Symposium on the Synthesis and Applications of Isotopically Labelled Compounds, Edinburgh, 16–20 July 2006.April ‐ May 2007Pages 578-579 ReferencesRelatedInformation

A synthetic analogue of the tripeptide hemiasterlin, designated HTI-286, depolymerizes microtubules, is a poor substrate for P-glycoprotein, and inhibits the growth of paclitaxel-resistant tumors in xenograft models. Two radiolabeled photoaffinity analogues of HTI-286, designated 4-benzoyl-N,beta,beta-trimethyl-l-phenylalanyl-N(1)-[(1S,2E)-3-carboxy-1-isopropylbut-2-enyl]-N(1),3-dimethyl-l-valinamide (probe 1) and N,beta,beta-trimethyl-l-phenylalanyl-4-benzoyl-N-[(1S,2E)-3-carboxy-1-isopropyl-2-butenyl]-N,beta,beta-trimethyl-l-phenylalaninamide (probe 2), were made to help identify HTI-286 binding sites in tubulin. HTI-286, probe 1, and probe 2 had similar affinities for purified tubulin [apparent K(D(app)) = 0.2-1.1 microM], inhibited polymerization of purified tubulin approximately 80%, and were potent inhibitors of cell growth (IC(50) = 1.0-22 nM). Both radiolabeled probes labeled exclusively alpha-tubulin. Labeling by [(3)H]probe 1 was inhibited by probe 1, HTI-286, vinblastine, or dolastatin 10 (another peptide antimitotic agent that depolymerizes microtubules) but was either unaffected or enhanced (at certain temperatures) by colchicine or paclitaxel. [(3)H]Probe 1 also labeled exclusively tubulin in cytosolic extracts of whole cells. The major, if not exclusive, contact site for probe 1 was mapped to residues 314-339 of alpha-tubulin and corresponds to the sheet 8 and helix 10 region. This region is known to (1) have longitudinal interactions with beta-tubulin across the interdimer interface, (2) have lateral interactions with adjacent protofilaments, and (3) contact the N-terminal region of stathmin, a protein that induces depolymerization of tubulin. Binding of probe 1 to this region may alter the conformation of tubulin outside the labeling domain, since enzymatic removal of the C-terminus of only alpha-tubulin by subtilisin after, but not before, photolabeling is blocked by probe 1. These results suggest that hemiasterlin is in close contact with alpha-tubulin and may span the interdimer interface so that it contacts the vinblastine- and dolastatin 10-binding sites believed to be in beta-tubulin. In addition, we speculate that antimitotic peptides mimic the interaction of stathmin with tubulin.

HTI-286, a synthetic analogue of hemiasterlin, depolymerizes microtubules and is proposed to bind at the Vinca peptide site in tubulin. It has excellent in vivo antitumor activity in human xenograft models, including tumors that express P-glycoprotein, and is in phase II clinical evaluation. To identify potential mechanisms of resistance induced by HTI-286, KB-3-1 epidermoid carcinoma cells were exposed to increasing drug concentrations. When maintained in 4.0 nmol/L HTI-286, cells had 12-fold resistance to HTI-286. Cross-resistance was observed to other Vinca peptide-binding agents, including hemiasterlin A, dolastatin-10, and vinblastine (7- to 28-fold), and DNA-damaging drugs, including Adriamycin and mitoxantrone (16- to 57-fold), but minimal resistance was seen to taxanes, epothilones, or colchicine (1- to 4-fold). Resistance to HTI-286 was retained when KB-HTI-resistant cells were grown in athymic mice. Accumulation of [(3)H]HTI-286 was lower in cells selected in intermediate (2.5 nmol/L) and high (4.0 nmol/L) concentrations of HTI-286 compared with parental cells, whereas accumulation of [(14)C]paclitaxel was unchanged. Sodium azide treatment partially reversed low HTI-286 accumulation, suggesting involvement of an ATP-dependent drug pump. KB-HTI-resistant cells did not overexpress P-glycoprotein, breast cancer resistance protein (BCRP/ABCG2/MXR), MRP1, or MRP3. No mutations were found in the major beta-tubulin isoform. However, 4.0 nmol/L HTI-286-selected cells had a point mutation in alpha-tubulin that substitutes Ser for Ala(12) near the nonexchangeable GTP-binding site of alpha-tubulin. KB-HTI-resistant cells removed from drug became less resistant to HTI-286, no longer had low HTI-286 accumulation, and retained the Ala(12) mutation. These data suggest that HTI-286 resistance may be partially mediated by mutation of alpha-tubulin and by an ATP-binding cassette drug pump distinct from P-glycoprotein, ABCG2, MRP1, or MRP3.

Proc Amer Assoc Cancer Res, Volume 45, 2004 5441 HTI-286, a synthetic analog of hemiasterlin, is a novel tripeptide that depolymerizes microtubules. The binding-sites on tubulin of antimitotic peptides such as hemiasterlin and dolastatin-10 are not well characterized. To determine the binding site for HTI-286 two radiolabeled benzophenone photoaffinity probes of HTI-286 were utilized. We have already established that photoprobe 1 binds within residues 314-338 of α-tubulin. Here we describe the binding of photoprobe 2 (N,β,β-trimethyl-L-phenylalanyl-4-benzoyl-N-[(1S,2E)-3-carboxy-1-isopropyl-2-butenyl]-N,β,β-trimethyl-L-phenylalaninamide). The affinity of probe 2 for bovine brain tubulin was determined to be 3.6 μM, 6-fold higher than that for HTI-286. HTI-286 and probe 2 inhibited tubulin polymerization in a cell-free system by 88% and 69%, respectively. Probe 2 was 20-fold less potent in its ability to inhibit cell growth (IC50 = 22nM). Radiolabeled probe 2 exclusively labeled the α-subunit in bovine brain and HeLa cell tubulin preparations. Photolabeling of tubulin was inhibited by unlabeled photoaffinity analog, HTI-286, vinblastine, and dolastatin-10. However, some concentrations of paclitaxel and colchicine enhanced binding of probe 2. Digestion of labeled tubulin with trypsin, formic acid, Lys C, and CNBr suggests that probe 2 binds N-terminal to residue 306 (formic acid cleavage site), and is likely to reside within amino acids 204-280 in α-tubulin. These and other data from our laboratory suggest that the HTI-286 binding site maps to α-tubulin at the interdimer region. This region has not been reported as a binding site for any other known tubulin-binding drug. The C-terminal (2-4 kDa) regions of α- and β-tubulin are susceptible to cleavage by subtilisin under non-denaturing conditions. Subtilisin cleaves β-tubulin more rapidly than α-tubulin. Vinblastine and dolastatin-10 binding induces polymerization of tubulin into non-microtubule structures that pellet at 100,000 g . Vinblastine-induced polymerization can block limited proteolysis (at room temperature) with subtilisin of α-tubulin, but not β-tubulin. We observed that HTI-286 does not cause tubulin to polymerize into structures that pellet at 100,000 g , but nevertheless protects the subtilisin cleavage site in α-tubulin. This suggests that the mechanism of protection of α-tubulin by HTI-286 is distinct from that seen by vinblastine and dolastatin-10. We propose that the protection of subtilisin-cleavage is due to a change in the conformation of the C-terminus upon binding of HTI-286 to α-tubulin. Taken together these data strongly suggest that HTI-286 binds to α-tubulin in the interdimer region.

ABSTRACT The naturally occurring mannopeptimycins (formerly AC98-1 through AC98-5) are a novel class of glycopeptide antibiotics that are active against a wide variety of gram-positive bacteria. The structures of the mannopeptimycins suggested that they might act by targeting cell wall biosynthesis, similar to other known glycopeptide antibiotics; but the fact that the mannopeptimycins retain activity against vancomycin-resistant organisms suggested that they might have a unique mode of action. By using a radioactive mannopeptimycin derivative bearing a photoactivation ligand, it was shown that mannopeptimycins interact with the membrane-bound cell wall precursor lipid II [C55-MurNAc-(peptide)-GlcNAc] and that this interaction is different from the binding of other lipid II-binding antibiotics such as vancomycin and mersacidin. The antimicrobial activities of several mannopeptimycin derivatives correlated with their affinities toward lipid II, suggesting that the inhibition of cell wall biosynthesis was primarily through lipid II binding. In addition, it was shown that mannopeptimycins bind to lipoteichoic acid in a rather nonspecific interaction, which might facilitate the accumulation of antibiotic on the bacterial cell surface.