Drug resistance dependent on allostery: A P-loop rigor Eg5 mutant exhibits resistance to allosteric inhibition by STLC.

The mitotic kinesin Eg5 has emerged as a potential anti-mitotic target for the purposes of cancer chemotherapy. Whether clinical resistance to these inhibitors can arise is unclear. We exploited HCT116 cancer cell line to select resistant clones to S-trityl-L-cysteine (STLC), an extensively studied Eg5 loop-L5 binding inhibitor. The STLC resistant clones differed in their resistance to other loop-L5 binding inhibitors but remained sensitive to the ATP class of competitive Eg5 specific inhibitors. Eg5 is still necessary for bipolar spindle formation in the resistant clones since the cells were sensitive to RNAi mediated depletion of Eg5. One clone expressing Eg5(T107N), a dominant point mutation in the P-loop of the ATP binding domain of the motor, appeared to be not only resistant but also dependent on the presence of STLC. Eg5(T107N) expression was associated also with resistance to the clinical relevant loop-L5 Eg5 inhibitors, Arry-520 and ispinesib. Ectopic expression of the Eg5(T107N) mutant in the absence of STLC was associated with strong non-exchangeable binding to microtubules causing them to bundle. Biochemical assays showed that in contrast to the wild type Eg5-STLC complex, the ATP binding site of the Eg5(T107N) is accessible for nucleotide exchange only when the inhibitor is present. We predict that resistance can be overcome by inhibitors that bind to other than the Eg5 loop-L5 binding site having different chemical scaffolds, and that allostery-dependent resistance to Eg5 inhibitors may also occur in cells and may have positive implications in chemotherapy since once diagnosed may be beneficial following cessation of the chemotherapeutic regimen.

Self protein-protein interactions are involved in TPPP/p25 mediated microtubule bundling.

TPPP/p25 is a microtubule-associated protein, detected in protein inclusions associated with various neurodegenerative diseases. Deletion analysis data show that TPPP/p25 has two microtubule binding sites, both located in intrinsically disordered domains, one at the N-terminal and the other in the C-terminal domain. In copolymerization assays the full-length protein exhibits microtubule stimulation and bundling activity. In contrast, at the same ratio relative to tubulin, truncated forms of TPPP/p25 exhibit either lower or no microtubule stimulation and no bundling activity, suggesting a cooperative phenomenon which is enhanced by the presence of the two binding sites. The binding characteristics of the N- and C-terminally truncated proteins to taxol-stabilized microtubules are similar to the full-length protein. However, the C-terminally truncated TPPP/p25 shows a lower Bmax for microtubule binding, suggesting that it may bind to a site of tubulin that is masked in microtubules. Bimolecular fluorescent complementation assays in cells expressing combinations of various TPPP/p25 fragments, but not that of the central folded domain, resulted in the generation of a fluorescence signal colocalized with perinuclear microtubule bundles insensitive to microtubule inhibitors. The data suggest that the central folded domain of TPPP/p25 following binding to microtubules can drive s homotypic protein-protein interactions leading to bundled microtubules.

IPP51, a chalcone acting as a microtubule inhibitor with in vivo antitumor activity against bladder carcinoma.

We previously identified 1-(2,4-dimethoxyphenyl)-3-(1-methylindolyl) propenone (IPP51), a new chalcone derivative that is capable of inducing prometaphase arrest and subsequent apoptosis of bladder cancer cells. Here, we demonstrate that IPP51 selectively inhibits proliferation of tumor-derived cells versus normal non-tumor cells. IPP51 interfered with spindle formation and mitotic chromosome alignment. Accumulation of cyclin B1 and mitotic checkpoint proteins Bub1 and BubR1 on chromosomes in IPP51 treated cells indicated the activation of spindle-assembly checkpoint, which is consistent with the mitotic arrest. The antimitotic actions of other chalcones are often associated with microtubule disruption. Indeed, IPP51 inhibited tubulin polymerization in an in vitro assay with purified tubulin. In cells, IPP51 induced an increase in soluble tubulin. Furthermore, IPP51 inhibited in vitro capillary-like tube formation by endothelial cells, indicating that it has anti-angiogenic activity. Molecular docking showed that the indol group of IPP51 can be accommodated in the colchicine binding site of tubulin. This characteristic was confirmed by an in vitro competition assay demonstrating that IPP51 can compete for colchicine binding to soluble tubulin. Finally, in a human bladder xenograft mouse model, IPP51 inhibited tumor growth without signs of toxicity. Altogether, these findings suggest that IPP51 is an attractive new microtubule-targeting agent with potential chemotherapeutic value.

STLC-resistant cell lines as tools to classify chemically divergent Eg5 targeting agents according to their mode of action and target specificity.

Determining the mechanism of action of drugs and their target specificity in cells remains a major challenge. Here we describe the use of cell lines expressing two point mutations in the allosteric inhibitor binding pocket of the mitotic kinesin Eg5 (D130A, in the loop L5 region and L214A in helix α3), which following transfection, were selected for their ability to proliferate normally in the presence of STLC, a well known Eg5 inhibitor. The cell lines were used to discriminate the mechanism of action of other chemically distinct small molecule inhibitors of Eg5 that differ in their mode of action. The STLC resistant cells were capable of continuous proliferation in the presence of ATP uncompetitive inhibitors, such as K858 and dimethylenastron, but were still sensitive to ATP competitive inhibitors that are thought to bind to a distinct site on Eg5 than the allosteric binding pocket. The STLC resistant cell lines can therefore be used as a filter to distinguish Eg5 loop L5 binding drugs from drugs binding to other pockets without prior structural information. Additionally, the cells can be used to analyze whether inhibitors of Eg5 are specific to this potential drug target or whether they have additional targets in dividing cells.

Proteome analysis of microtubule-associated proteins and their interacting partners from mammalian brain.

The microtubule (MT) cytoskeleton is essential for a variety of cellular processes. MTs are finely regulated by distinct classes of MT-associated proteins (MAPs), which themselves bind to and are regulated by a large number of additional proteins. We have carried out proteome analyses of tubulin-rich and tubulin-depleted MAPs and their interacting partners isolated from bovine brain. In total, 573 proteins were identified giving us unprecedented access to brain-specific MT-associated proteins from mammalian brain. Most of the standard MAPs were identified and at least 500 proteins have been reported as being associated with MTs. We identified protein complexes with a large number of subunits such as brain-specific motor/adaptor/cargo complexes for kinesins, dynein, and dynactin, and proteins of an RNA-transporting granule. About 25% of the identified proteins were also found in the synaptic vesicle proteome. Analysis of the MS/MS data revealed many posttranslational modifications, amino acid changes, and alternative splice variants, particularly in tau, a key protein implicated in Alzheimer's disease. Bioinformatic analysis of known protein-protein interactions of the identified proteins indicated that the number of MAPs and their associated proteins is larger than previously anticipated and that our database will be a useful resource to identify novel binding partners.

Synthesis and antiproliferative evaluation of pyrazolo[1,5-a]-1,3,5-triazine myoseverin derivatives.

Pyrazolo[1,5-a]-1,3,5-triazine myoseverin derivatives 1a-c were prepared from 4-(N-methyl-N-phenylamino)-2-methylsulfanylpyrazolo[1,5-a]-1,3,5-triazine 2. Their cytotoxic activity, inhibition of tubulin polymerization, and cell cycle effects were evaluated. Compounds 1a and 1c are potent tubulin inhibitors and displayed specific antiproliferative activity in colorectal cancer cell lines at micromolar concentrations.

Structure-activity relationship of S-trityl-L-cysteine analogues as inhibitors of the human mitotic kinesin Eg5.

The human kinesin Eg5 is a potential drug target for cancer chemotherapy. Eg5 specific inhibitors cause cells to block in mitosis with a characteristic monoastral spindle phenotype. Prolonged metaphase block eventually leads to apoptotic cell death. S-trityl-L-cysteine (STLC) is a tight-binding inhibitor of Eg5 that prevents mitotic progression. It has proven antitumor activity as shown in the NCI 60 tumor cell line screen. It is of considerable interest to define the minimum chemical structure that is essential for Eg5 inhibition and to develop more potent STLC analogues. An initial structure-activity relationship study on a series of STLC analogues reveals the minimal skeleton necessary for Eg5 inhibition as well as indications of how to obtain more potent analogues. The most effective compounds investigated with substitutions at the para-position of one phenyl ring have an estimated K i (app) of 100 nM in vitro and induce mitotic arrest with an EC 50 of 200 nM.

Screening for inhibitors of microtubule-associated motor proteins.

The mitotic spindle is an important target for cancer chemotherapy. The main protein target for drugs in clinical use is tubulin, the building block of microtubules. In recent years, other proteins of the mitotic spindle have been identified as potential targets for the development of more specific drugs with the hope that these will have fewer side effects than known antimitotics (taxanes, vinca alkaloids). The human genome contains more than 40 members of the kinesin superfamily, with at least 12 of these involved in mitosis and cytokinesis. HsEg5 (also called KSP, kinesin spindle protein), a member of the kinesin-5 family, involved in the formation of the bipolar spindle, is a very promising target for cancer chemotherapy with specific inhibitors in Phase I and II clinical trails. Several successful approaches exist today to screen Eg5 for inhibitors, including phenotype-based assays and simple in vitro assays that explore the intrinsic enzymatic ATPase activity of Eg5. Here, we describe a robust and straightforward in vitro method to rapidly screen Eg5 for inhibitors. The assay can easily be adapted to other mitotic kinesins that may be identified in the future as potential drug targets, or simply to obtain specific kinesin inhibitors for use in "chemical genetics" to study the function of this important class of proteins.

New chemical tools for investigating human mitotic kinesin Eg5.

We have designed and synthesized a series of monastrol derivatives, an allosteric inhibitor of Eg5, a motor protein responsible for the formation and maintenance of the bipolar spindle in mitotic cells. Sterically demanding structural modifications have been introduced on the skeleton of the parent drug either via a multicomponent Biginelli reaction or a stepwise modification of monastrol. The ability of these compounds to inhibit Eg5 activity has been investigated using two in vitro steady-state ATPase assays (basal and microtubule-stimulated) as well as a cell-based assay. One compound in the series appeared more potent than monastrol by a fivefold factor. Three other compounds that were unable to inhibit Eg5 ATPase activity in vitro proved potent Eg5 inhibitors in the cell-based assay. The results obtained led to the identification of structure-activity relationships further used to design an affinity matrix that can be used for fast and efficient purification of Eg5 from crude lysate of eukaryotic cells.

Structure of human Eg5 in complex with a new monastrol-based inhibitor bound in the R configuration.

Drugs that target mitotic spindle proteins have been proven useful for tackling tumor growth. Eg5, a kinesin-5 family member, represents a potential target, since its inhibition leads to prolonged mitotic arrest through the activation of the mitotic checkpoint and apoptotic cell death. Monastrol, a specific dihydropyrimidine inhibitor of Eg5, shows stereo-specificity, since predominantly the (S)-, but not the (R)-, enantiomer has been shown to be the biologically active compound in vitro and in cell-based assays. Here, we solved the crystal structure (2.7A) of the complex between human Eg5 and a new keto derivative of monastrol (named mon-97), a potent antimitotic inhibitor. Surprisingly, we identified the (R)-enantiomer bound in the active site, and not, as for monastrol, the (S)-enantiomer. The absolute configuration of this more active (R)-enantiomer has been unambiguously determined via chemical correlation and x-ray analysis. Unexpectedly, both the R- and the S-forms inhibit Eg5 ATPase activity with IC(50) values of 110 and 520 nM (basal assays) and 150 nm and 650 nm (microtubule-stimulated assays), respectively. However, the difference was large enough for the protein to select the (R)- over the (S)-enantiomer. Taken together, these results show that in this new monastrol family, both (R)- and (S)-enantiomers can be active as Eg5 inhibitors. This considerably broadens the alternatives for rational drug design.

The marine natural product adociasulfate-2 as a tool to identify the MT-binding region of kinesins.

Kinesins are molecular motors that transport cargo along microtubules (MTs). To move forward the motor must attach to the MT in a defined orientation and detach from it in a process that is driven by ATP hydrolysis. The knowledge of the motor-MT interface is essential for a detailed understanding of how kinesins move along MTs and how they are related to other molecular motors such as myosins or dyneins. We have used the marine natural product adociasulfate-2 (AS-2), previously identified as a MT-competitive inhibitor of conventional kinesin, to infer the secondary structure elements forming the MT interface of two human mitotic kinesins, namely, CENP-E and Eg5. AS-2 inhibits both basal and MT-stimulated ATPase activities of CENP-E (IC50 of 8.6 and 1.3 microM, respectively) and Eg5 (IC50 of 3.5 and 5.3 microM, respectively) and is a MT-competitive inhibitor of CENP-E with a Ki of 0.35 microM. Binding of AS-2 to CENP-E also stimulates the ADP release from the nucleotide-binding pocket. AS-2 is a nonspecific kinesin inhibitor targeting several superfamily members including KHC, MPP1, MKLP1, RabK6, KIFC1, KIFC3, CENP-E, and Eg5. By measuring hydrogen/deuterium exchange with mass spectrometry we have shown that the formation of the CENP-E/AS-2 complex decreases the solvent accessibility of three neighboring peptides on the same face of CENP-E. We deduce that this is the site of MT attachment and conclude that loop L11, helix alpha4, loop L12, helix alpha5, loop L8, and strand beta5 constitute the main MT interface of the CENP-E motor domain. Similarly for Eg5/AS-2, a region of increased solvent accessibility locates the MT interface of Eg5.

Molecular dissection of the inhibitor binding pocket of mitotic kinesin Eg5 reveals mutants that confer resistance to antimitotic agents.

The mitotic kinesin Eg5 plays an essential role in establishing the bipolar spindle. Recently, several antimitotic inhibitors have been shown to share a common binding region on Eg5. Considering the importance of Eg5 as a potential drug target for cancer chemotherapy it is essential to understand the molecular mechanism, by which these agents block Eg5 activity, and to determine the "key residues" crucial for inhibition. Eleven residues in the inhibitor binding pocket were mutated and the effects were monitored by kinetic analysis and mass spectrometry. Mutants R119A, D130A, P131A, I136A, V210A, Y211A and L214A abolish the inhibitory effect of monastrol. Results for W127A and R221A are less striking, but inhibitor constants are still considerably modified compared to wild-type Eg5. Only one residue, Leu214, was found to be essential for inhibition by STLC. W127A, D130A, V210A lead to increased K(i)(app) values, but binding of STLC is still tight. R119A, P131A, Y211A and R221A convert STLC into a classical rather than a tight-binding inhibitor with increased inhibitor constants. These results demonstrate that monastrol and STLC interact with different amino acids within the same binding region, suggesting that this site is highly flexible to accommodate different types of inhibitors. The drug specificity is due to multiple interactions not only with loop L5, but also with residues located in helices alpha2 and alpha3. These results suggest that tumour cells might develop resistance to Eg5 inhibitors, by expressing Eg5 point mutants that retain the enzyme activity, but prevent inhibition, a feature that is observed for certain tubulin inhibitors.

S-trityl-L-cysteine is a reversible, tight binding inhibitor of the human kinesin Eg5 that specifically blocks mitotic progression.

Human Eg5, responsible for the formation of the bipolar mitotic spindle, has been identified recently as one of the targets of S-trityl-L-cysteine, a potent tumor growth inhibitor in the NCI 60 tumor cell line screen. Here we show that in cell-based assays S-trityl-L-cysteine does not prevent cell cycle progression at the S or G(2) phases but inhibits both separation of the duplicated centrosomes and bipolar spindle formation, thereby blocking cells specifically in the M phase of the cell cycle with monoastral spindles. Following removal of S-trityl-L-cysteine, mitotically arrested cells exit mitosis normally. In vitro, S-trityl-L-cysteine targets the catalytic domain of Eg5 and inhibits Eg5 basal and microtubule-activated ATPase activity as well as mant-ADP release. S-trityl-L-cysteine is a tight binding inhibitor (estimation of K(i,app) <150 nm at 300 mm NaCl and 600 nm at 25 mm KCl). S-trityl-L-cysteine binds more tightly than monastrol because it has both an approximately 8-fold faster association rate and approximately 4-fold slower release rate (6.1 microM(-1) s(-1) and 3.6 s(-1) for S-trityl-L-cysteine versus 0.78 microM(-1) s(-1) and 15 s(-1) for monastrol). S-trityl-L-cysteine inhibits Eg5-driven microtubule sliding velocity in a reversible fashion with an IC(50) of 500 nm. The S and D-enantiomers of S-tritylcysteine are nearly equally potent, indicating that there is no significant stereospecificity. Among nine different human kinesins tested, S-trityl-L-cysteine is specific for Eg5. The results presented here together with the proven effect on human tumor cell line growth make S-trityl-L-cysteine a very attractive starting point for the development of more potent mitotic inhibitors.

Use of hydrogen/deuterium exchange mass spectrometry and mutagenesis as a tool to identify the binding region of inhibitors targeting the human mitotic kinesin Eg5.

An experimental procedure associating both hydrogen/deuterium exchange mass spectrometry (H/D-MS) and mutagenesis was developed to identify the protein-binding region of small inhibitors targeting the motor domain of the human mitotic kinesin Eg5. All the tested inhibitors decrease the deuterium incorporation rate of the same peptides corresponding to the following secondary structure elements: loop L5/helix alpha2 (region Tyr125-Glu145) and strand beta5/helix alpha3 (region Ile202-Leu227). Replacement of these two regions by the equivalent ones from N. crassa conventional kinesin heavy chain completely abolishes the modification of the deuterium incorporation rate by the inhibitors as well as their effects on the basal ATPase activity. The six tested inhibitors thus share a common binding site on Eg5. The strategy reported here allows the regions of a protein involved in ligand binding to be rapidly pinpointed and can be applied to other proteins and used as a general in vitro screening procedure to identify compounds targeting specific binding regions.

Essential kinesins: characterization of Caenorhabditis elegans KLP-15.

Kinesins form a superfamily of molecular motors involved in cell division and intracellular transport. Twenty kinesins have been found in the Caenorhabditis elegans genome, and four of these belong to the kinesin-14 subfamily, i.e., kinesins with C-terminal motor domains. Three of these kinesin-14s, KLP-15, KLP-16, and KLP-17, form a distinct subgroup in which KLP-15 and KLP-16 are more than 90% identical and appear to be related by a relatively recent gene duplication. They are essential for meiotic spindle organization and chromosome segregation, and are mostly expressed in the germline. With 587 amino acids each, they are among the smallest kinesins known. Using bacterially expressed KLP-15 constructs with different length extensions preceding the motor domain, we have determined in vitro the following characteristic properties: ATPase activity, microtubule binding, oligomeric state, microtubule gliding activity, and direction of movement. The constructs exhibit a monomer-dimer equilibrium that depends on the length of the predicted alpha-helical coiled-coil region preceding the motor domain. The longest construct with the complete coiled-coil domain is a stable dimer, and the shortest construct with only seven amino acids preceding the motor domain is a monomer. In microtubule gliding assays, the monomer is immobile whereas the fully dimeric KLP-15 construct supports gliding at 2.3 microm/min and moves toward microtubule minus ends, like other members of the kinesin-14 subfamily studied to date.

Identification of the protein binding region of S-trityl-L-cysteine, a new potent inhibitor of the mitotic kinesin Eg5.

Human Eg5, a mitotic motor of the kinesin superfamily, is involved in the formation and maintenance of the mitotic spindle. The recent discovery of small molecules that inhibit HsEg5 by binding to its catalytic motor domain leading to mitotic arrest has attracted more interest in Eg5 as a potential anticancer drug target. We have used hydrogen-deuterium exchange mass spectrometry and directed mutagenesis to identify the secondary structure elements that form the binding sites of new Eg5 inhibitors, in particular for S-trityl-l-cysteine, a potent inhibitor of Eg5 activity in vitro and in cell-based assays. The binding of this inhibitor modifies the deuterium incorporation rate of eight peptides that define two areas within the motor domain: Tyr125-Glu145 and Ile202-Leu227. Replacement of the Tyr125-Glu145 region with the equivalent region in the Neurospora crassa conventional kinesin heavy chain prevents the inhibition of the Eg5 ATPase activity by S-trityl-l-cysteine. We show here that S-trityl-l-cysteine and monastrol both bind to the same region on Eg5 by induced fit in a pocket formed by helix alpha3-strand beta5 and loop L5-helix alpha2, and both inhibitors trigger similar local conformational changes within the interaction site. It is likely that S-trityl-l-cysteine and monastrol inhibit HsEg5 by a similar mechanism. The common inhibitor binding region appears to represent a "hot spot" for HsEg5 that could be exploited for further inhibitor screening.

In vitro screening for inhibitors of the human mitotic kinesin Eg5 with antimitotic and antitumor activities.

Human Eg5, a member of the kinesin superfamily, plays a key role in mitosis, as it is required for the formation of a bipolar spindle. We describe here the first in vitro microtubule-activated ATPase-based assay for the identification of small-molecule inhibitors of Eg5. We screened preselected libraries obtained from the National Cancer Institute and identified S-trityl-L-cysteine as the most effective Eg5 inhibitor with an IC50 of 1.0 micromol/L for the inhibition of basal ATPase activity and 140 nmol/L for the microtubule-activated ATPase activity. Subsequent cell-based assays revealed that S-trityl-L-cysteine induced mitotic arrest in HeLa cells (IC50, 700 nmol/L) with characteristic monoastral spindles. S-trityl-L-cysteine is 36 times more potent for inducing mitotic arrest than the well-studied inhibitor, monastrol. Gossypol, flexeril, and two phenothiazine analogues were also identified as Eg5 inhibitors, and we found that they all result in monoastral spindles in HeLa cells. It is notable that all the Eg5 inhibitors identified here have been shown previously to inhibit tumor cell line growth in the NCI 60 tumor cell line screen, and we conclude that their antitumor activity may at least in part be explained by their ability to inhibit Eg5 activity.

Interaction of the mitotic inhibitor monastrol with human kinesin Eg5.

The microtubule-dependent kinesin-like protein Eg5 from Homo sapiens is involved in the assembly of the mitotic spindle. It shows a three-domain structure with an N-terminal motor domain, a central coiled coil, and a C-terminal tail domain. In vivo HsEg5 is reversibly inhibited by monastrol, a small cell-permeable molecule that causes cells to be arrested in mitosis. Both monomeric and dimeric Eg5 constructs have been examined in order to define the minimal monastrol binding domain on HsEg5. NMR relaxation experiments show that monastrol interacts with all of the Eg5 constructs used in this study. Enzymatic techniques indicate that monastrol partially inhibits Eg5 ATPase activity by binding directly to the motor domain. The binding is noncompetitive with respect to microtubules, indicating that monastrol does not interfere with the formation of the motor-MT complex. The binding is not competitive with respect to ATP. Both enzymology and in vivo assays show that the S enantiomer of monastrol is more active than the R enantiomer and racemic monastrol. Stopped-flow fluorometry indicates that monastrol inhibits ADP release by forming an Eg5-ADP-monastrol ternary complex. Monastrol reversibly inhibits the motility of human Eg5. Monastrol has no inhibitory effect on the following members of the kinesin superfamily: MC5 (Drosophila melanogaster Ncd), HK379 (H. sapiens conventional kinesin), DKH392 (D. melanogaster conventional kinesin), BimC1-428 (Aspergillus nidulans BimC), Klp15 (Caenorhabditis elegans C-terminal motor), or Nkin460GST (Neurospora crassa conventional kinesin).