Bioorganic & Medicinal Chemistry Letters
Fluorescent photoaffinity probes for mitotic protein kinase Aurora A
Darja Lavogina ⇑, Katariina Kisand, Gerda Raidaru, Asko Uri
Institute of Chemistry, University of Tartu, Ravila 14A, 50411 Tartu, Estonia
a r t i c l e i n f o
Keywords:
Protein kinase Aurora A
Photoaffinity labelling Fluorescent probe
Covalent irreversible inhibitor VX689 (MK-5108)
a b s t r a c t
We combined the advantages of the selective inhibitor VX689, the bisubstrate-analogue conjugate approach, and photoreactive amino acids to develop 8 photoaffinity probes for Aurora A. The most effi- cient compounds possessed one-digit nanomolar KD values in the equilibrium binding assay, inhibited Aurora A at elevated concentrations of ATP in the phosphorylation assay in the presence of TPX2, and formed covalent complexes with the recombinant kinase or Aurora A in HeLa cells upon UV-irradiation. The recognition of the correct target by the probes during formation of the covalent complex in the bio- chemical assay and in situ was demonstrated by competition experiments using the non-labelled inhibi- tors VX689 and MLN8237.
© 2015 Elsevier Ltd. All rights reserved.
Aurora A is a Ser/Thr protein kinase that serves as an important player in mitosis, being involved in intracellular pathways leading to centrosome maturation, separation, and bipolar spindle assem- bly.1–3 The concentration of Aurora A protein peaks in G2/M cells, where the kinase is localized into the cytosol, on the centrosomes, or on the microtubules.4 The latter local pool of Aurora A is formed by the interaction of the kinase with the microtubule-associated protein TPX2,5 which additionally stabilizes Aurora A by protecting the phosphorylated activation loop of the kinase against phos- phatases,6,7 and by affecting recognition of Aurora A by degrading proteases.8 In addition, TPX2 acts as an activator of Aurora A by reducing the size of its ATP-binding pocket and hence lowering the KM value of ATP.9,10
While the normal activity of Aurora A is necessary for the suc-
cessful completion of mitosis and thus sustains the growth of living organisms, the amplification of the aurora a gene as well as the abnormally elevated concentration and catalytic activity of the kinase have been connected to a variety of tumors.11–13 The unveiled value of the kinase as a drug target urged the develop- ment of several inhibitors of Aurora A, with 21 studies currently undergoing clinical trials.14 Still, while most of the Aurora A studies pursue the potential pharmaceutical application of compounds, the recent discovery of Aurora A as a prognostic biomarker opens yet another perspective of importance of this kinase and a need for tools enabling quantification of Aurora A activity in biological samples.15,16
* Corresponding author. Tel.: +372 737 5250.
E-mail address: [email protected] (D. Lavogina).
The affinity-based tagging technique, also termed activity- based protein profiling, is a powerful method for the assessment of the concentration and activity of enzymes, selectivity profiling of enzyme-targeting inhibitors, as well as dissection/monitoring of the enzyme-regulated pathways in cells.17,18 This technology uses chemical probes that contain generally three structural ele- ments: a selective ligand responsible for binding to the enzyme, a reactive group responsible for the development of the covalent linkage to the target (can also be a part of the irreversibly binding ligand), and a reporter group (e.g., fluorescent dye).19,20 By apply- ing photoreactive building blocks or cysteinome targeting, the probes compatible with affinity-based tagging technique have been reported for several classes of enzymes, including protein kinases.21–23 Still, only one such probe (based on the commercially available inhibitor MLN8237) has been designed specifically for native Aurora A,24 and another probe has been used for targeting the genetically engineered kinase.25 Moreover, few reversibly binding fluorescent probes have been developed for Aurora A.26–28 Here, we report the design and synthesis of a series of photoaffinity probes targeting native Aurora A, and the results of the structure– efficacy studies for these novel compounds in studies with recombinant kinase and HeLa cells.
As the starting compound for the design of reactive photoaffin-
ity probes, we chose the well-known inhibitor VX689 that targets the ATP-binding site of Aurora A.29 In our previous study, we demonstrated that a fluorescent dye could be attached to the car- boxylic group of VX689 via a linker without remarkable reduction of affinity of the inhibitor.28 Thus, here we designed probes comprising a photoreactive moiety (4-azido-Phe in Compound I,
2 lM).32 The phosphorylation assay was performed in the presence of 5 mM ATP (a high concentration of ATP was used to mimic the intracellular conditions), 10 mM Mg2+, 30 lM substrate (Kemptide analogue) and 2.5 lM peptide incorporating residues 1–43 of the Aurora A activator TPX2.
The results of the experiments are summarized in Table 1. Importantly, both probes incorporating photoreactive moieties
6 2
O n formed covalent complexes with Aurora A upon irradiation, whereas the azide-containing Compound I was more efficient than
Figure 1. Schematic structure of the probes developed in this study.
or 4-benzoyl-Phe in Compound II) that was attached via a 4- aminobutanoic acid linker to the carboxyl group of VX689 (Fig. 1; the detailed structures and HRMS data of compounds are given in Supplementary Tables S1 and S2). Two photoreactive moi- eties were concurrently introduced into the structure of a probe to increase the probability of the probe for the development of cova- lent bonds with Aurora A upon UV-irradiation of the complex; as a control, Compound A was constructed incorporating Phe residues comprising no photoreactive groups. To the C-terminus of the probes, lysine moiety was added that enabled the attachment of a fluorescent dye to its side-chain.
The ability of the photoaffinity probes to form a covalent irre- versible complex with purified recombinant full-length Aurora A upon irradiation with near-UV light was tested using SDS–PAGE and transfer onto a Western blot membrane by monitoring the presence of a fluorescent band in the molecular weight range slightly exceeding that of Aurora A (theoretical Mw of
46.9 kDa).30 This analysis was followed by the control staining of the membrane with an Aurora A-specific antibody. As the starting conditions of Aurora labelling, final total concentrations of 1 lM
Aurora A and 1.2 lM or 2.4 lM probes and 30 min irradiation time
were used. In parallel, the affinity of the non-irradiated probes towards Aurora A was measured using a binding assay with fluo- rescence polarization/anisotropy readout (final total concentration of probes 1 nM or lower),31 and the inhibitory potency of non-irradiated compounds was established by one concentration- point phosphorylation assay (final total concentration of probes
the benzoyl-containing Compound II (for quantification, the ratio of the signal at 50 kDa to the total fluorescent signal in the given lane was found for each sample; Fig. 2A and B). It was also observed that the percentage of the covalent complex formed between Compound I and Aurora A depended strongly on the pres- ence of DMSO in the reaction mixture, which could be attributed to the relatively hydrophobic nature of the probe (the most efficient complex formation was observed at the highest concentration of DMSO used, 4 vol %).
The variation of UV-irradiation times showed that in case of the azide-containing Compound I, 15 min irradiation was suffi- cient for the covalent complex formation and its yield was not improved upon longer treatment of the sample with UV-light; in case of the benzoyl-containing Compound II, however, 30 min irradiation was required (Fig. 2C and D). The presence of TPX2 (1–43) (final total concentration of 100 lM) did not affect signif- icantly the efficiency of the complex formation between Compound II and Aurora A (Fig. 2A and B). This was an important observation given the fact that the potency of most Aurora A-tar- geting inhibitors (including VX689) is generally remarkably lower in the presence of TPX2 due to the allosterically induced reduc- tion of the size of the ATP-binding site of the kinase. Interestingly, it goes in line with our previous report on the bind- ing of the conjugate comprising MLN8237 and an oligo-arginine
peptide that had shown relatively small changes in affinity and association/dissociation kinetics in the presence versus absence of TPX2 (1–43) as compared to the conjugates lacking the long peptidic fragment.28 Still, none of the Compounds I, II or A showed significant inhibition of the kinase at elevated concentra- tions of ATP in the phosphorylation assay (Table 1).
Table 1
Structures and biochemical characteristics of the probes and control compounds used in this study
Compound Schematic structurea
% Probe in complexb
KDc (nM)
% Aurora A residual activityd
I VX689-Abu-LPhe(N3)-LPhe(N3)-LLys[TAMRA]-NH2 40 ± 5** 1.9 ± 0.6 100 ± 5ns
II VX689-Abu-LPhe(Bz)-LPhe(Bz)-LLys[TAMRA]-NH2 9.0 ± 2.6* ND 100 ± 3ns
III VX689-bAla-DPhe(N3)-NH-PEG3-NH-TAMRA 20 ± 3⁄⁄⁄ 2.2 ± 0.1 73 ± 6⁄⁄⁄
IV VX689-Abu-LPhe(Bz)-LPhe(Bz)-LArg6-DLys[PF555]-NH2 8.5 ± 1.9* ND 100 ± 5ns
V VX689-Abu-LPhe(Bz)-LPhe(Bz)-DArg6-DLys[PF555]-NH2 8.1 ± 0.1* 9.1 ± 2.4 100 ± 5ns
VI VX689-Abu-DPhe(Bz)-DPhe(Bz)-DArg6-DLys[PF555]-NH2 6.9 ± 0.2ns 6.5 ± 0.1 110 ± 4ns
VII VX689-bAla-LPhe(N3)-LPhe(N3)-DArg6-DLys[PF555]-NH2 18 ± 4⁄⁄⁄ 5.7 ± 2.1 71 ± 4⁄⁄⁄
VIII VX689-bAla-DPhe(N3)-DPhe(N3)-LArg6-DLys[PF555]-NH2 15 ± 1⁄⁄⁄ 6.1 ± 2.2 74 ± 6⁄⁄⁄
IX VX689-bAla-DPhe(N3)-DPhe(N3)-DArg6-DLys[PF555]-NH2 11 ± 2⁄⁄ 2.5 ± 0.8 96 ± 7ns
X VX689-Ahx-LPhe(N3)-LArg6-DLys[TAMRA]-NH2 3.6 ± 0.6ns 0.73 ± 0.06 66 ± 6⁄⁄⁄
A VX689-Abu-LPhe-LPhe-LLys[TAMRA]-NH2 1.5 ± 0.7 0.44 ± 0.06 110 ± 2ns
B VX689-Ahx-DArg6-NH2 ND ND 10 ± 4⁄⁄⁄
VX689 — ND ND 67 ± 4***
ND, not determined.
a Abbreviations: Abu, 4-aminobutanoic acid moiety; Ahx, 6-aminohexanoic acid moiety; Bz, benzoyl group; PEG3, 4,7,10-trioxa-1,13-tridecanediamine moiety; PF555, PromoFluor-555 moiety; TAMRA, 5-carboxy-tetramethyl rhodamine moiety.
b Here and below, the percentage of the probe forming the covalent complex with Aurora A was defined as the ratio of the signal at 50 kDa to the total fluorescent signal in the given lane (N P2). A final total concentration of 1.2 lM was used for Compounds II, III and A; for other probes, the final total concentration was 2.4 lM. Samples with Compounds I, II and A were irradiated for 30 min; for samples with other probes, irradiation time was 15 min. Statistical significance of the difference of complex formation efficiency for each probe versus Compound A was calculated by the unpaired t-test; the symbols indicate the following two-tailed P-values (90% confidence level): ***P 60.001,
***P 60.01, *P 60.1, nsP >0.1.
c Equilibrium dissociation constant KD value for the non-irradiated probes (N P2).
d Percentage of catalytic activity of Aurora A in the presence of the non-irradiated probes as compared to the non-inhibited control (N P2). Statistical significance of the difference of catalytic activity for each probe versus non-inhibited control was calculated by the 1-way ANOVA with Dunnett’s multiple comparisons test; the symbols show the following two-tailed P-values (90% confidence level): ***P 60.001, nsP >0.1.
Compound A Compound I Compound II
Figure 2. Assessment of the formation of the covalent complex between Aurora A and photoaffinity probes. (A) Example of a Western blot membrane with samples containing Compounds I, II or A (final total concentration of 2.4 lM was used for Compound I, and 1.2 lM for Compounds II and A), Aurora A (final total concentration of 1 lM) and TPX2 (final total concentration of 100 lM) after 30 min irradiation. (B) Quantification of covalent complex formation efficiency dependent on the presence of 100 lM TPX2 or different concentrations of Aurora A in the samples (N P2). Statistical significance of the difference of complex formation efficiency for the indicated conditions was calculated by the unpaired t-test; the symbols indicate the following two-tailed P-values (90% confidence level): *P 60.1, nsP >0.1. (C) Example of a Western blot membrane with samples containing Compounds 1, 2 or A (final total concentration of 2.4 lM was used for Compound I, and 1.2 lM for Compounds II and A) and Aurora A (final total concentration of 1 lM) after different irradiation times. (D) Quantification of the covalent complex formation efficiency dependent on the irradiation time (N = 2).
The changes in concentration of the kinase and the probe in the sample (final total concentrations of 200 nM Aurora A and 2.4 lM Compound II) also did not significantly influence the efficiency of the covalent complex formation between Compound II and Aurora A. This indicated that the relatively low affinity of the non-irradiated compound towards Aurora A might be the issue: if the amount of the non-covalent complex formed prior to irradi- ation is low, then most of the probe is in solution and cannot form a covalent complex with the kinase upon irradiation. In case of Compound I, on the other hand, the concentration of the kinase proved extremely important for the complex formation. When final total concentrations of 200 nM Aurora A and 4.8 lM Compound I were used, the ratio of the signal at 50 kDa to the total fluorescent signal decreased almost 5-fold as compared to the ini- tial conditions (final total concentrations of 1 lM Aurora A and
2.4 lM Compound I; Fig. 2B). According to the results of the bind-
ing assay, the equilibrium dissociation constant KD values of the non-irradiated Compounds I and A were in the one-digit nanomo- lar range (Table 1). The affinity of the Compound II could not be measured with sufficient accuracy, as the signal of the free unbound probe was relatively high (presumably due to a certain extent of non-specific binding of the hydrophobic probe to the microplate plastic), which overall reduced significantly the mea- surement window.
Next, we decided to incorporate an oligo-arginine sequence into the structure of compounds (bisubstrate-analogue inhibitor approach),33 which was expected to improve the properties of pho- toaffinity probes from the two aspects. First, we presumed that this fragment mimicking the substrate sequence for the basophilic Aurora A would further contribute to the increased affinity of the non-irradiated probes in the presence of TPX2. Second, we aimed at increase of the polarity of the conjugates to overcome the adsorption of the probes to the laboratory plastic and thus to improve handling of those. Therefore, a series of compounds con- taining VX689 as the ATP-site targeting fragment, an oligo-arginine peptide as the presumed substrate binding site-targeting fragment,
and 4-azido-Phe or 4-benzoyl-Phe as the photoreactive moieties was designed and synthesized. To explore the relationship between the structure of the probes and the characteristics of those (such as affinity, inhibitory potency, and efficiency of the covalent complex formation with Aurora A), we also varied the number of the photoreactive moieties, the chirality of the incorpo- rated amino acid residues, and the length of the linkers (Compounds III–X; Table 1). As a control, we synthesized a non- fluorescent Compound B incorporating DArg6 and no Phe residues. According to the results of the biochemical assays with non-irra- diated probes (Table 1), the affinity and the inhibition potency of the probes incorporating azido-Phe (Compounds VII–X) were sys- tematically higher than those of compounds incorporating ben- zoyl-Phe (Compounds IV–VI) irrespective of the chirality of amino acid residues within the structure of compounds. This observation confirmed that the incorporation of large substituents to the para- position of the aromatic ring of Phe is not favored by the kinase. Furthermore, the best inhibitory characteristics were shown by Compound B, which was a more potent inhibitor than VX689 itself in the presence of 5 mM ATP and 2.5 lM TPX2 (1–43). Interestingly, Compound III that contained no oligo-arginine and only one azido- Phe residue demonstrated significantly more potent inhibition of
Aurora A as compared to the similar Compound I.
While the covalent complex formation efficiency was also higher on average for compounds containing azido-Phe than for those con- taining benzoyl-Phe, the effect of the linker length between the VX689 moiety and photoreactive amino acid residues was opposite to the trend observed in the equilibrium binding assay. The probes containing a shorter bAla moiety as a linker (Compounds VII–IX) formed the covalent complex with Aurora A significantly more effi- ciently than the probe with the longer linker derived from 6-amino- hexanoic acid (Compound X). This observation showed that while longer flexible linker structures are preferred for the binding of non-irradiated compounds to the kinase, those lead to a too distant positioning of the photoreactive moieties from the kinase residues that can be attacked by the reactive radicals formed from aromatic
Figure 3. Assessment of the formation of the covalent complex between Aurora A and photoaffinity probes in the presence of competing inhibitors (VX689 or MLN8237) in experiments with recombinant kinase or in HeLa cells. (A) Example of a Western blot membrane with samples containing photoaffinity probes (final total concentration of 1 lM), Aurora A (final total concentration of 1 lM) in the absence or presence of VX689 (final total concentration of 200 lM) after 15 min irradiation. For each probe, the sample in the absence and in the presence of VX689 was prepared on the same day and applied on the same SDS–PAGE gel. (B) Quantification of covalent complex formation efficiency dependent on the presence of VX689 (N = 2). Statistical significance of the difference of complex formation efficiency for the indicated conditions was calculated by
the unpaired t-test; the symbols indicate the following two-tailed P-values (90% confidence level): **P <0.01, *P 60.1, nsP >0.1. (C) Example of a Western blot membrane with samples prepared after 20 h incubation of live HeLa cells with 100 nM nocodazole and 1 lM photoaffinity probes without or with addition of 50 lM MLN8237 prior to 15 min irradiation and lysis of cells. The dotted arrow indicates the bands that were quantified for samples incubated with Compounds I or III, and the solid arrow indicates the bands that were quantified for samples incubated with Compounds VII or VIII. The samples made in the same experiment were applied onto a different SDS–PAGE gel and transferred to Western blot membranes, one of which was subsequently stained with an anti-Aurora A antibody and the other one with an anti-bActin antibody. (D)
Quantification of covalent complex formation efficiency of photoaffinity probes in HeLa cells dependent on the incubation with MLN8237 prior to irradiation and cell lysis (N = 2). The percentage of the probe forming the covalent complex was calculated as above, and each ratio was additionally normalized to that of the sample not incubated with a competing inhibitor (defined as 100%); the additional normalization was required to compensate for slightly different internalization efficiency of probes into the cells in two independent experiments. Statistical significance of the difference of complex formation efficiency for the indicated conditions was calculated by the unpaired t-test; the symbols indicate the following two-tailed P-values (90% confidence level): **P <0.01, *P 60.1, nsP >0.1.
azides upon UV-irradiation. Interestingly, in the absence of TPX2, the probes containing azido-Phe and an oligo-arginine peptide (Compounds VII–IX) were somewhat less efficient labelling reagents than the probe containing only azido-Phe (Compound III), even though the former probes contained two photoreactive moieties and the latter probe only one. This suggests that the non- covalently formed binary complex of Aurora A and the probe could not adopt an optimal conformation for pinpointing both interaction ‘hotspots’, the one important for the binding of the substrate-mim- icking part of the conjugate and the one important for formation of covalent bonds with the kinase upon irradiation.
In order to confirm that the covalent complex between the probes and Aurora A detected by Western blot reflects the specific binding process, we performed competition experiments (Fig. 3). Aurora A was pre-mixed with excess of a non-labelled inhibitor VX689 (final total concentrations of 1 lM and over 100 lM, respectively); after 10 min, a photoaffinity probe (Compound I, III or VII–IX) was added to the sample (final total concentration of 1 lM) and after another 15 min incubation, the irradiation of the sample was performed for 15 min. We expected that the long
residence time of VX689 in complex with Aurora A will enable monitoring of the reduction in covalent complex formation even after relatively long sample incubation and irradiation times.28 Indeed, for all the probes used in the competition experiments, a significantly lower amount of the covalent complex was formed in the presence of VX689 as compared to the samples where the inhibitor competing with the probes for the binding site of Aurora A was absent (Fig. 3A and B).
Finally, we assessed the ability of the most efficient of our novel photoaffinity probes to label Aurora A covalently in the cellular milieu. Compounds I, III, VII and VIII were chosen for this experi- ment as those showed most promising profiles from the aspect of covalent complex formation efficiency with recombinant kinase as well as in the biochemical binding and inhibition assays. HeLa cells were treated simultaneously with 100 nM nocodazol (for enrich- ment of population of mitotic cells) and 1 lM photoaffinity probes dissolved in usual growth medium; after 20 h, Aurora A-targeting inhibitor MLN8237 (50 lM) dissolved in growth medium was added for 30 min to some wells,34 whereas only growth medium was added to other wells. The cells were subsequently transferred
onto ice, the medium was replaced with PBS and 15 min irradiation with UV-light was then performed, followed by cell lysis. We used a 1 lM concentration of the probes, because it was presumed to be sufficient for the penetration of the compounds through the cell plasma membrane (exceeding of a concentration ‘threshold’ is gen- erally required for internalization of conjugates containing oligo- arginines),35,36 but low enough to enable 50 lM MLN8237 to com- pete for binding (higher concentration of MLN8237 could likely result in quick cell death).34 On the other hand, prolonged incuba- tion times in the presence of high concentration of MLN8237 were avoided to minimize the difference in the physiological state of the cells treated with MLN8237 versus non-treated cells. MLN8237 was chosen instead of VX689, as the former has even longer resi- dence time in complex with Aurora A (also in the presence of TPX2).28
The cell lysate samples were applied onto an SDS–PAGE gel, fol- lowed by transfer onto a Western blot membrane, fluorescence imaging and subsequent staining with Aurora A-specific antibody to visualize the presence of the kinase at 50 kDa (Fig. 3C and D).37 In lanes with samples containing Compounds I and III, fluorescent bands were visible at the Mw range slightly above 50 kDa that could correspond to labelling of Aurora A; still, the intensity of these bands did not depend on the addition of MLN8237 to the samples prior to irradiation. In lanes with samples containing Compounds VII and VIII, however, a fluorescent band was visible at the Mw range of 50 kDa, whereas it was significantly weakened in the sam- ples incubated with MLN8237. Additionally, a strong unidentified band was present at 15 kDa in the lanes with samples containing Compound VIII (confirmed in the other independent experiment). We thus concluded that Compound VII was the most suitable pho- toaffinity probe for covalent labelling of Aurora A in HeLa cells in situ, and we will pursue further possibilities for enhancement of its potency (e.g., from the aspect of internalization efficiency) in our future studies.
To sum up, in this study we synthesized 11 fluorescent probes and one non-fluorescent bisubstrate-analogue inhibitor targeting Aurora A. All fluorescent probes showed a one-digit nanomolar equilibrium dissociation constant KD value. Out of 12 novel com- pounds, 5 inhibitors significantly affected the Aurora A-catalyzed
phosphorylation of the Kemptide analogue at 5 mM concentration of ATP in the presence of 2.5 lM TPX2 (1–43), whereas one inhibi- tor was more efficient at these conditions than the commercially available inhibitor VX689 used as the starting compound for the design of novel inhibitors. Upon UV-irradiation, 8 probes could be successfully used for the photoaffinity labelling of the recombi- nant kinase, whereas 2 of these probes enabled labelling of Aurora A in HeLa cells. We expect that the structures of the compounds
can be easily adjusted dependent on the experimental needs (e.g., development of ‘clickable’ probes), and that these probes or their derivatives will find several application possibilities in research of mitotic kinase pathways.
Acknowledgements
This work was supported by Grants from the Estonian Research Council (PUT0007 and IUT20-17). The help of Dr. Katja Gehenn in preparation of the manuscript is gratefully acknowledged.
Supplementary data
Supplementary data (materials and methods, structures of com- pounds, HPLC and HRMS data) associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.bmcl. 2015.05.060.
References and notes
1. Janssen, A.; Medema, R. H. Oncogene 2011, 30, 2799.
2. Hochegger, H.; Hégarat, N.; Pereira-Leal, J. B. Open Biol. 2013, 3, 120185.
3. Salmela, A. L.; Kallio, M. J. Chromosoma 2013, 122, 431.
4. Marumoto, T.; Zhang, D.; Saya, H. Nat. Rev. Cancer 2005, 5, 42.
5. Kufer, T. A.; Silljé, H. H.; Körner, R.; Gruss, O. J.; Meraldi, P.; Nigg, E. A. J. Cell Biol.
2002, 158, 617.
6. Bayliss, R.; Sardon, T.; Vernos, I.; Conti, E. Mol. Cell 2003, 12, 851.
7. Zorba, A.; Buosi, V.; Kutter, S.; Kern, N.; Pontiggia, F.; Cho, Y. J.; Kern, D. eLife
2014, 3, e02667.
8. Giubettini, M.; Asteriti, I. A.; Scrofani, J.; De Luca, M.; Lindon, C.; Lavia, P.; Guarguaglini, G. J. Cell Sci. 2011, 124, 113.
9. Anderson, K.; Yang, J.; Koretke, K.; Nurse, K.; Calamari, A.; Kirkpatrick, R. B.; Patrick, D.; Silva, D.; Tummino, P. J.; Copeland, R. A.; Lai, Z. Biochemistry 2007, 46, 10287.
10. Dodson, C. A.; Bayliss, R. J. Biol. Chem. 2012, 287, 1150.
11. Bischoff, J. R.; Anderson, L.; Zhu, Y.; Mossie, K.; Ng, L.; Souza, B.; Schryver, B.; Flanagan, P.; Clairvoyant, F.; Ginther, C.; Chan, C. S.; Novotny, M.; Slamon, D. J.; Plowman, G. D. EMBO J. 1998, 17, 3052.
12. Warner, S. L.; Bearss, D. J.; Han, H.; Von Hoff, D. D. Mol. Cancer Ther. 2003, 2, 589.
13. Mehra, R.; Serebriiskii, I. G.; Burtness, B.; Astsaturov, I.; Golemis, E. A. Lancet Oncol. 2013, 14, e425.
14. U.S. National Institutes of Health, https://clinicaltrials.gov/ct2/results?term= aurora+a&recr=Open&no_unk=Y, last accessed on April 2, 2015.
15. Chen, J.; Lin, Q.; Wen, J. Y.; Li, X.; Ma, X. K.; Fan, X. J.; Cao, Q. H.; Dong, M.; Wei, L.; Chen, Z. H.; Li, X. Y.; Wang, T. T.; Liu, Q.; Wan, X. B.; Xing, Y. F.; Wu, X. Y. Tumour Biol. 2014, 35, 9361.
16. Umene, K.; Yanokura, M.; Banno, K.; Irie, H.; Adachi, M.; Iida, M.; Nakamura, K.; Nogami, Y.; Masuda, K.; Kobayashi, Y.; Tominaga, E.; Aoki, D. Int. J. Oncol. 2015, 46, 1498.
17. Cohen, M. S.; Hadjivassiliou, H.; Taunton, J. Nat. Chem. Biol. 2007, 3, 156.
18. Cheng, X.; Li, L.; Uttamchandani, M.; Yao, S. Q. Chem. Commun. (Camb.) 2014,
50, 2851.
19. Hagenstein, M. C.; Mussgnug, J. H.; Lotte, K.; Plessow, R.; Brockhinke, A.; Kruse, O.; Sewald, N. Angew. Chem., Int. Ed. 2003, 42, 5635.
20. Niphakis, M. J.; Cravatt, B. F. Annu. Rev. Biochem. 2014, 83, 341.
21. Shi, H.; Zhang, C. J.; Chen, G. Y.; Yao, S. Q. J. Am. Chem. Soc. 2012, 134, 3001.
22. Liu, Q.; Sabnis, Y.; Zhao, Z.; Zhang, T.; Buhrlage, S. J.; Jones, L. H.; Gray, N. S.
Chem. Biol. 2013, 20, 146.
23. Sherratt, A. R.; Nasheri, N.; McKay, C. S.; O’Hara, S.; Hunt, A.; Ning, Z.; Figeys, D.; Goto, N. K.; Pezacki, J. P. ChemBioChem 2014, 15, 1253.
24. Su, Y.; Pan, S.; Li, Z.; Li, L.; Wu, X.; Hao, P.; Sze, S. K.; Yao, S. Q. Sci. Rep. 2015, 5, 7724.
25. Koch, A.; Rode, H. B.; Richters, A.; Rauh, D.; Hauf, S. ACS Chem. Biol. 2012, 7, 723.
26. Yang, K. S.; Budin, G.; Reiner, T.; Vinegoni, C.; Weissleder, R. Angew. Chem., Int. Ed. 2012, 51, 6598.
27. Slatter, A. F.; Campbell, S.; Angell, R. M. J. Biomol. Screening 2013, 18, 219.
28. Lavogina, D.; Enkvist, E.; Viht, K.; Uri, A. ChemBioChem 2014, 15, 443.
29. Shimomura, T.; Hasako, S.; Nakatsuru, Y.; Mita, T.; Ichikawa, K.; Kodera, T.; Sakai, T.; Nambu, T.; Miyamoto, M.; Takahashi, I.; Miki, S.; Kawanishi, N.; Ohkubo, M.; Kotani, H.; Iwasawa, Y. Mol. Cancer Ther. 2010, 9, 157.
30. On membrane, we obtained two bands (Mw of ca. 42 kDa and 48 kDa) that were both recognized by the Aurora A-specific antibodies. Dependent on the batch of protein, the proportion of these bands was variable, whereas the intensity of the lower band increased upon storage of the kinase on ice (indicating that it most likely corresponded to either hyper- autophosphorylated or truncated form of the Aurora A); however, no significant change in the catalytic activity of the kinase was observed. For simplicity, in this Letter we hence further refer to the observed Mw of Aurora A as 50 kDa.
31. Vaasa, A.; Viil, I.; Enkvist, E.; Viht, K.; Raidaru, G.; Lavogina, D.; Uri, A. Anal. Biochem. 2009, 385, 85.
32. Viht, K.; Vaasa, A.; Raidaru, G.; Enkvist, E.; Uri, A. Anal. Biochem. 2005, 340, 165.
33. Lavogina, D.; Enkvist, E.; Uri, A. ChemMedChem 2010, 5, 23.
34. Görgün, G.; Calabrese, E.; Hideshima, T.; Ecsedy, J.; Perrone, G.; Mani, M.; Ikeda, H.; Bianchi, G.; Hu, Y.; Cirstea, D.; Santo, L.; Tai, Y. T.; Nahar, S.; Zheng, M.; Bandi, M.; Carrasco, R. D.; Raje, N.; Munshi, N.; Richardson, P.; Anderson, K. C. Blood 2010, 115, 5202.
35. Vaasa, A.; Lust, M.; Terrin, A.; Uri, A.; Zaccolo, M. Biochem. Biophys. Res. Commun. 2010, 397, 750.
36. Kriisa, M.; Sinijärv, H.; Vaasa, A.; Enkvist, E.; Kostenko, S.; Moens, U.; Uri, A.
ChemBioChem 2015, 16, 312.
37. In lanes with Compound III, the intensity of staining with Aurora A-specific antibody was systematically lower than in the other lanes, although the intensity of staining with the bActin-specific antibody (used as a loading control) was the same. Possibly, a short PEG chain contained in Compound 3 might have interfered with binding of antibodies in the Mw range close to 50 kDa. For MK-5108 quantification in Figure 3D, we pooled the data for both Aurora A and bActin staining in case of both independent experiments.