Search

An analysis of heteronuclear-NMR-based screening data is used to derive relationships between the ability of small molecules to bind to a protein and various parameters that describe the protein binding site. It is found that a simple model including terms for polar and apolar surface area, surface complexity, and pocket dimensions accurately predicts the experimental screening hit rates with an R² of 0.72, an adjusted R² of 0.65, and a leave-one-out Q² of 0.56. Application of the model to predict the druggability of protein targets not used in the training set correctly classified 94% of the proteins for which high-affinity, noncovalent, druglike leads have been reported. In addition to understanding the pocket characteristics that contribute to high-affinity binding, the relationships that have been defined allow for quantitative comparative analyses of protein binding sites for use in target assessment and validation, virtual ligand screening, and structure-based drug design.

Reversal of aberrant gene expression that is induced by the proto-oncogene c-myc is likely to be effective for treating a variety of tumors that rely on this pathway for growth. One strategy to down-regulate the c-myc pathway is to target transcription factors that regulate its own expression. A host of proteins act in coordination to regulate c-myc expression and any one of them are theoretical targets for small-molecule therapy. Experimentally, it has been shown that the far upstream element (FUSE) binding protein (FBP) is essential for c-myc expression, and reductions in FBP levels both reduce c-myc expression and correlate with slower cell growth. FBP binds to ssDNA by capturing exposed DNA bases in a hydrophobic pocket. This suggests that a small molecule could be designed to occupy this pocket and inhibit FBP function. Using a variety of screening methodologies, we have identified ligands that bind to the DNA binding pockets of the KH domains of FBP. Gel shift analyses using full length FBP and a related transcription factor confirm that a small-molecule lead compound inhibits DNA binding in a specific manner. The benzoylanthranilic acid compounds described here represent leads in the design of FBP inhibitors that can serve as useful tools in the study of c-myc regulation and in the development of therapeutics that target the c-myc pathway.

Inhibitor of apoptosis (IAP) proteins are overexpressed in many cancers and have been implicated in tumor growth, pathogenesis, and resistance to chemo- or radiotherapy. On the basis of the NMR structure of a SMAC peptide complexed with the BIR3 domain of X-linked IAP (XIAP), a novel series of XIAP antagonists was discovered. The most potent compounds in this series bind to the baculovirus IAP repeat 3 (BIR3) domain of XIAP with single-digit nanomolar affinity and promote cell death in several human cancer cell lines. In a MDA-MB-231 breast cancer mouse xenograft model, these XIAP antagonists inhibited the growth of tumors. Close structural analogues that showed only weak binding to the XIAP−BIR3 domain were inactive in the cellular assays and showed only marginal in vivo activity. Our results are consistent with a mechanism in which ligands for the BIR3 domain of XIAP induce apoptosis by freeing up caspases. The present study validates the BIR3 domain of XIAP as a target and supports the use of small molecule XIAP antagonists as a potential therapy for cancers that overexpress XIAP.

Protein Tyrosine phosphatase 1B (PTP1B) has been implicated as a key negative regulator of both insulin and leptin signaling pathways. Using an NMR-based screening approach with ¹⁵N- and ¹³C-labeled PTP1B, we have identified 2,3-dimethylphenyloxalylaminobenzoic acid (1) as a general, reversible, and competitive PTPase inhibitor. Structure-based approach guided by X-ray crystallography facilitated the development of 1 into a novel series of potent and selective PTP1B inhibitors occupying both the catalytic site and a portion of the noncatalytic, second phosphotyrosine binding site. Interestingly, oral biovailability has been observed in rats for some compounds. Furthermore, we demonstrated in vivo plasma glucose lowering effects with compound 12d in ob/ob mice.

The NMR-based discovery of biaryl hydroxamate inhibitors of the matrix metalloproteinase stromelysin (MMP-3) has been previously described (Hajduk et al. J. Am. Chem. Soc. 1997, 119, 5818−5827). While potent in vitro, these inhibitors exhibited no in vivo activity due, at least in part, to the poor pharmacokinetic properties of the alkylhydroxamate moiety. To circumvent this liability, NMR-based screening was implemented to identify alternative zinc-chelating groups. Using this technique, 1-naphthyl hydroxamate was found to bind tightly to the protein (KD = 50 μM) and was identified as a candidate for incorporation into the lead series. On the basis of NMR-derived structural information, the naphthyl hydroxamate and biaryl fragments were linked together to yield inhibitors of this enzyme that exhibited improved bioavailability. These studies demonstrate that the NMR-based screening of fragments can be effectively applied to improve the physicochemical or pharmacokinetic profile of lead compounds.

The interaction between leukocyte function-associated antigen-1 (LFA-1) and intracellular adhesion molecule-1 (ICAM-1) has been implicated in inflammatory and immune diseases. Recently, a novel series of p-arylthio cinnamides has been described as potent antagonists of the LFA-1/ICAM-1 interaction. These compounds were found to bind to the I domain of LFA-1 using two-dimensional NMR spectroscopy of ¹⁵N-labeled LFA-1 I domain. On the basis of NOE studies between compound 1 and the I domain of LFA-1, a model of the complex was constructed. This model revealed that compound 1 does not directly inhibit ICAM-1 binding by interacting with the metal ion dependent adhesion site (MIDAS). Instead, it binds to the previously proposed I domain allosteric site (IDAS) of LFA-1 and likely modulates the activation of LFA-1 through its interaction with this regulatory site. A fragment-based NMR screening strategy was applied to identify small, more water-soluble ligands that bind to a specific region of the IDAS. When incorporated into the parent cinnamide template, the resulting analogues exhibited increased aqueous solubility and improved pharmacokinetic profiles in rats, demonstrating the power of this NMR-based screening approach for rapidly modifying high-affinity ligands.

A strategy is described for designing high-affinity ligands using information derived from the NMR-based screening of fragments. The method involves the fragmentation of an existing lead molecule, identification of suitable replacements for the fragments, and incorporation of the newly identified fragments into the original scaffold. Using this technique, novel substituents were rapidly identified and incorporated into lead inhibitors of adenosine kinase that exhibited potent in vitro and in vivo activities. This approach is a valuable strategy for modifying existing leads to improve their potency, bioavailability, or toxicity profile and thus represents a useful technique for lead optimization.

Using an NMR-based screen, a novel class of urokinase inhibitors were identified that contain a 2-aminobenzimidazole moiety. The inhibitory potency of this family of inhibitors is similar to that of inhibitors containing a guanidine or amidine group. However, unlike previously described guanidino- or amidino-based inhibitors which have pKa values greater than 9.0, urokinase inhibitors containing a 2-aminobenzimidazole have pKa values of 7.5. Thus, 2-aminobenzimidazoles may have improved pharmacokinetic properties which could increase the bioavailability of inhibitors which contain this moiety. A crystal structure of one of the lead inhibitors, 2-amino-5-hydroxybenzimidazole, complexed with urokinase reveals the electrostatic and hydrophobic interactions that stabilize complex formation and suggests nearby subsites that may be accessed to increase the potency of this new series of urokinase inhibitors.

A statistical analysis of NMR-derived binding data on 11 protein targets was performed to identify molecular motifs that are preferred for protein binding. The analysis indicates that compounds which contain a biphenyl substructure preferentially bind to a wide range of proteins and that high levels of specificity (>250-fold) can be achieved even for these small molecules. These results suggest that high-throughput screening libraries that are enriched with biphenyl-containing compounds can be expected to have increased chances of yielding high-affinity ligands for proteins, and they suggest that the biphenyl can be utilized as a template for the discovery and design of therapeutics with high affinity and specificity for a broad range of protein targets.

bcl-x is a member of the bcl-2 family of genes. The major protein product, Bcl-x(L), is a 233-amino-acid protein which has antiapoptotic properties. In contrast, one of the alternatively spliced transcripts of the bcl-x gene codes for the protein Bcl-x(S), which lacks 63 amino acids present in Bcl-x(L) and has proapoptotic activity. Unlike other proapoptotic Bcl-2 family members, such as Bax and Bak, Bcl-x(S) does not seem to induce cell death in the absence of an additional death signal. However, Bcl-x(S) does interfere with the ability of Bcl-x(L) to antagonize Bax-induced death in transiently transfected 293 cells. Mutational analysis of Bcl-x(S) was conducted to identify the domains necessary to mediate its proapoptotic phenotype. Deletion mutants of Bcl-x(S) which still contained an intact BH3 domain retained the ability to inhibit survival through antagonism of Bcl-x(L). Bcl-x(S) was able to form heterodimers with Bcl-x(L) in mammalian cells, and its ability to inhibit survival correlated with the ability to heterodimerize with Bcl-x(L). Deletion mutants of Bax and Bcl-2, which lacked BH1 and BH2 domains but contained a BH3 domain, were able to antagonize the survival effect conferred by Bcl-x(L). The results suggest that BH3 domains from both pro- and antiapoptotic Bcl-2 family members, while lacking an intrinsic ability to promote programmed cell death, can be potent inhibitors of Bcl-x(L) survival function.

C24-Deoxyascomycin was prepared in a two-step process from ascomycin and evaluated for its immunosuppressant activity relative to ascomycin and FK506. An intermediate in the synthetic pathway, Δ^23,24-dehydroascomycin, was likewise evaluated. Despite lacking the hydrogen-bonding interactions associated with the C24-hydroxyl moiety of ascomycin, C24-deoxyascomycin was found to be equipotent to the parent compound both in its immunosuppressive potency and in its interaction with the immunophilin, FKBP12. Conversely, Δ^23,24-dehydroascomycin which also lacks the same hydrogen-bonding interactions did not exhibit this potency. NMR studies were conducted on the FKBP12/C24-deoxyascomycin complex in an attempt to understand this phenomenon at the molecular level. The NMR structures of the complexes formed between FKBP12 and ascomcyin or C24-deoxyascomcyin were very similar, suggesting that hydrogen-bonding interactions with the C24 hydroxyl moiety are not important for complex formation.

The Erm family of methyltransferases confers resistance to the macrolide−lincosamide−streptogramin type B (MLS) antibiotics through the methylation of 23S ribosomal RNA. Upon the methylation of RNA, the MLS antibiotics lose their ability to bind to the ribosome and exhibit their antibiotic activity. Using an NMR-based screen, we identified a series of triazine-containing compounds that bind weakly to ErmAM. These initial lead compounds were optimized by the parallel synthesis of a large number of analogues, resulting in compounds which inhibit the Erm-mediated methylation of rRNA in the low micromolar range. NMR and X-ray structures of enzyme/inhibitor complexes reveal that the inhibitors bind to the S-adenosylmethionine binding site on the Erm protein. These compounds represent novel methyltransferase inhibitors that serve as new leads for the reversal of Erm-mediated MLS antibiotic resistance.

A new knowledge-based scoring function (PMF-score), implemented into the DOCK4 program, was used to screen a database of 3247 small molecules for binding to the FK506 binding protein (FKBP). The computational ranking of these compounds was compared to the binding affinities measured by NMR. It was demonstrated that small, weakly binding molecules have, on average, higher computational scores than nonbinders and are enriched in the upper ranks of the computational scoring lists. In addition, the results obtained with the PMF scoring function were superior (by 30−120% larger enrichment factors) to those obtained with the standard force field score of DOCK4. The reliable ranking of small, weakly binding molecules offers new ways of designing building blocks in combinatorial libraries as well as SAR by NMR libraries with the increased chance of identifying suitable lead compounds for drug design.

Lipopolysaccharide (LPS) samples isolated from a parent and two antibiotic-hypersusceptible mutant strains of Escherichia coli were analyzed for polycation affinity and level of binding. Purified salts of the LPSs from the parent strain, UB1005, and from one of the mutant strains, DC1, bound similar amounts of sodium and magnesium, but the samples from the second mutant strain, DC2, had significantly greater amounts of counterions bound per phosphate than did the other two isolates. The 31P nuclear magnetic resonance spectra indicated that, compared with LPS from the parental strain, the sample from strain DC1 was similar but the DC2 sample contained fewer diphosphodiester and more diphosphomonoester groups. Motion within the lipid A head group regions of the magnesium salts of the three isolates was dramatically different, as revealed by an electron spin resonance probe. The binding of the cations to the LPS aggregates was measured by the displacement of this cationic spin probe from the LPS samples. The polycations polymyxin, gentamicin, and spermine displaced more probe from samples of the two mutant strains than from that of the parental strain. The sample from the most antibiotic-susceptible strain, DC2, had the highest affinity for all the polyvalent cations tested. The results indicate that antibiotic hypersusceptibility can result from at least two distinct alterations in LPS structure. The decrease in diphosphodiesters and increase in diphosphomonoesters in the LPS of the DC2 sample resulted in more acidic phosphate moieties and a more antibiotic-susceptible cell. In contrast, the alterations in the LPS of DC1 that resulted in antibiotic hypersusceptibility of the cell were not in the phosphate substituents. In both mutants, however, hypersusceptibility resulted in an alteration in LPS structure that increased the affinity of the molecules for polycations.

2H NMR spectra were obtained for dimyristoylphosphatidylcholine multilamellar dispersions, perdeuterated in the fatty acid chains, in the presence and absence of two steroid analogs. The presence of the active anesthetic steroid alphaxalone results in consistently smaller 2H quadrupolar splittings (delta nu Q) for individual C2H2 segments of the fatty acid chains, indicating increased molecular disorder. In contrast, the inactive analog delta 16-alphaxalone causes no significant change in the 2H spectra of the phospholipid. The conformational analysis of alphaxalone and delta 16-alphaxalone in solution was carried out with the help of 1H and 13C high resolution NMR spectroscopy and the results were used to propose a molecular model for the interaction of the two steroids with membrane phospholipids. The model correlates the observed differences in the manner in which the two steroids interact with model membranes with differences in their respective conformations and provides a molecular basis for anesthetic steroid activity.

Avoparcin is a commercially important glycopeptide antibiotic which is active against Gram-positive bacteria. Recently, beta-avoparcin, the major component of the avoparcin mixture, and several other analogues have been structurally characterized and tested for their antibacterial activity. Varying degrees of activity were observed for only minor structural differences. For example, beta-avoparcin and epi-beta-avoparcin, which differ only in the stereochemistry at position 1' of the NH2-terminal phenylsarcosine subunit, exhibit a 10- to 100-fold difference in antibacterial activity. Following the supposition that the conformational properties of these molecules may explain the differences in their antibacterial activity, we have analyzed the conformations of beta-avoparcin, epi-beta-avoparcin, and other avoparcin analogues in aqueous solutions using NMR methods. On the basis of an analysis of the 1H chemical shifts, 1H-1H nuclear Overhauser enhancement data, and pH titration experiments, it was concluded that the conformations of all of the analogues are similar at the COOH terminus. However, for beta-avoparcin and epi-beta-avoparcin, conformational differences were observed in the NH2-terminal region of the molecules. In this paper, we present a detailed description of the over-all conformations of these two glycopeptides as deduced from an in-depth analysis of their corresponding 1H NMR spectra and discuss the possible relationships this may have with their markedly different antibacterial activities.