key: cord-1041114-ijy0canp
authors: Cozza, G.; Fortuna, M.; Meggio, F.; Sarno, S.; Kubbutat, M. H. G.; Totzke, F.; Schaechtele, C.; Pinna, L. A.; Olsufyeva, E. N.; Preobrazhenskaya, M. N.
title: Hydrophobic Derivatives of Glycopeptide Antibiotics as Inhibitors of Protein Kinases
date: 2018-10-15
journal: Biochemistry (Mosc)
DOI: 10.1134/s0006297918100073
sha: ca2ac866aba0297472a5bf3eb9e5796b38d57db0
doc_id: 1041114
cord_uid: ijy0canp

As key regulators of cell signaling, protein kinases (PKs) are attractive targets for therapeutic intervention in a variety of diseases. Herein, we report for the first time the inhibitory activity of polycyclic peptides, particularly, derivatives of glycopeptide antibiotics teicoplanin and eremomycin, against a panel of 12 recombinant human protein kinases and two protein kinases (CK1 and CK2) isolated from rat liver. Several of the investigated compounds inhibited various PKs with IC(50) values below 10 μM and caused >90% suppression of the enzyme activity at 10 μM concentration. Kinetic analysis of the protein kinase CK2α inhibition by the teicoplanin aglycon analogue (7) demonstrated the non-competitive mechanism of inhibition (with regard to ATP). Interestingly, the inhibitory activity of some investigated compounds correlated with the earlier described antiviral activity against HIV, HCV, and other corona- and flaviviruses.

BIOCHEMISTRY (Moscow) Vol. 83 No. 10 2018 1) are able to inhibit several types of viruses, such as HIV, HCV, influenza viruses (A/H1N1, A/H3N2 and B), coro naviruses, and flaviviruses in vitro [10 18 ]. Although natu ral glycopeptide antibiotics do not exhibit pronounced antiviral activity in such tests, chemical removal of sugars and introduction of hydrophobic residues into resulting aglycons or pseudo aglycons generates hydrophobic derivatives that are active against some viruses.

Together with vancomycin and teicoplanin, the gly copeptide antibiotic eremomycin ( Fig. 1) can be used as original compounds for developing a new generation of antiviral agents [11, 12, 16] that represent non nucleoside inhibitors of RNA replicase. Preliminary findings showed that glycopeptide aglycon derivatives act at the entry [11, 17] and post entry [14] stages of the virus infection cycle. The processes of virus penetration into the host cell and its replication provided by specific interactions between viral structural or non structural (co)receptors. The mechanisms of the antiviral activity of glycopeptide agly con derivatives remain unclear. Virus entry to the cell and its life cycle depend on the use of intracellular signal transduction pathways. On these grounds, we screened several hydrophobic glycopeptide aglycon derivatives for their inhibitory activity against a panel of PKs.

Chemicals from Sigma Aldrich (USA) and ProQinase (Germany) were used.

Compound library. Eremomycin was obtained from the pilot plant of Gause Institute of New Antibiotics (Moscow, Russia). Teicoplanin and teicoplanin aglycon (1) were kindly provided by Dr. R. Ciabatti and Dr. A. Malabarba (both from Biosearch S.p.A. Gerenzano, Italy). Teicoplanin aglycon derivatives n decylamide (2), 6 amino n heptyl 1 amide (3), 10 amino n decylamide (4), p (N,N di n butyl)benzylamide (5), p n butylben zyl n heptylamide (6), N 1 adamantyl oxycarbonyl derivative (7), N 1 adamantyl oxycarbonyl cyclopropyl amide (8), N t butyl oxycarbonyl adamantyl 2 amide (9), and N t butyl oxycarbonyl adamantyl 1 methyl [11, 14, 16, 19] . The purity (>95%) and identity of used compounds were assessed by HPLC and spectral analysis.

In vitro PK activity assay with human recombinant PKs. The activity of recombinant human PKs was assessed by the radiometric method ( Table 1 ). The reaction mixture contained 60 mM HEPES NaOH, pH 7.5, 3 mM MgCl 2 , 3 mM MnCl 2 , 3 μM sodium orthovanadate, 1.2 mM dithiothreitol (DTT), 50 μg/ml PEG20000, and 1 μM [γ 33 P]ATP (~5·10 5 Ci; PerkinElmer, USA) in a final volume of 50 μl. The enzyme concentrations in the reaction mix ture were the following: Aurora A, 2 ng/μl; Aurora B, 2 ng/μl; AXL, 2 ng/μl; B RAF V600E, 0.4 ng/μl; CDK2/Cyclin A, 2 ng/μl; CK2α 1, 4 ng/μl; IGF1R, 0.4 ng/μl; MET, 2 ng/μl; PLK1, 1 ng/μl; PRK1, 1 ng/μl; SRC, 0.2 ng/μl; and VEGFR2, 0.5 ng/μl. Incorporation of radioactive phosphate into the following substrates was measured: RBER CHKtide (PLK1); tetra(LRRWSLG) (Aurora A, Aurora B, PRK); casein (CK2 α 1); MEK1 KM (BRAF V600E); histone H1 (CDK2/Cyclin A); poly(Ala, Glu, Lys, Tyr) 6 : 2 : 5 : 1 (MET); poly(Glu, Tyr) 4 : 1 (AXL, IGF1R, SRC, VEGFR2).

In vitro PK activity was assayed with a Beckman Coulter/Sagian Multipette Automated Liquid Handler robotic system (USA) at the ATP concentration of 1 μM. The assay conditions have been optimized for each kinase, as well as for each lot of the recombinant proteins.

To verify the quality of the recombinant proteins, we determined apparent K m (ATP) values under the same conditions that were used for compound in vitro testing.

For each PK lot used, the IC 50 values with three ref erence inhibitors (10 concentrations each) were deter mined at 1 μM ATP to confirm the reproducibility of the obtained results, as well as their correspondence to the earlier published data. Z′ factors have been determined in each assay; all of them were at least 0.41 and, in most cases, exceeded 0.6, thereby indicating good to excellent assay reproducibility.

The reactions were performed at 30°C for 80 min and then stopped by adding 50 μl of 2% H 3 PO 4 (v/v). After washing off unreacted γ 33 P, label incorporation was measured with a Wallac Microbeta Trilux microplate scintillation counter (PerkinElmer).

In vitro PK activity assay with tissue derived CK1 and CK2. Native CK2 (α2β2) and CK1 were purified from rat liver as previously described [20] . Human recombinant CK2 α subunits were expressed in Escherichia coli and used in kinetic analysis after purification (see below) [20] . CK2 and CK1 phosphorylation was carried out at 37°C in the presence of increasing amounts of each inhibitor in a final volume of 25 μl containing 50 mM Tris HCl, pH 7.5, 100 mM NaCl, 12 mM MgCl 2 , and 0.02 mM [γ 33 P]ATP (500 1000 cpm/pmol), unless indicated other wise. The phosphorylation substrates were synthetic pep tides RRRADDSDDDDD (100 μM) and RRKHAAIGDDDDAYSITA (200 μM) for CK2 and CK1, respectively. The reaction was started by adding kinase to the reaction mixture and stopped after 10 min by adding 50 μl of 0.5 M H 3 PO 4 . Aliquots of the reaction mixture were spotted onto phosphocellulose filters (MANN Filter, Germany); the filters were washed four times with 5 10 ml of 75 mM H 3 PO 4 and once with methanol (20 ml) and then dried at 25°C before counting. Each measurement was performed in triplicate; the results were calculated as mean ± SEM (SEM < 15%).

Kinetic analysis. Kinetic data were obtained using CK2α as described above. Initial reaction rates were determined at different substrate and ATP concentrations as described in [21, 22] . K m values were calculated in the absence and presence of increasing inhibitor concentra tions using Lineweaver-Burk double reciprocal plots. 

The inhibitory activity of nine teicoplanin derivatives (2) (10), eremomycin aglycon (11), and its derivatives (12) (15) against CK2α and 11 human PKs (Aurora A, Aurora B, AXL, BRAF V600E, CDK2/CycA, IGF1 R, MET, PLK1, PRK1, SRC, VEGFR2) were evaluated in vitro by either determining IC 50 values (Table 1) or meas uring enzyme residual activity (%) incubation with the tested compounds (10 and 1 μM) ( Table 2) .

The inhibitory activity of teicoplanin aglycon deriv atives (2) (7), (9), and (10), teicoplanin aglycon (1), teicoplanin, and eremomycin was evaluated in vitro against native CK2 and CK1 isolated from rat liver ( Table  3) . We also studied the kinetics of CK2α inhibition by the teicoplanin analogue (7) that exhibited the highest inhibitory activity in our tests (Fig. 3, a and b) .

We compared the inhibitory activity of studied gly copeptide derivatives against PKs and discussed the rela tionship between activity, structure, and antiviral activi ties of the tested compounds.

Teicoplanin aglycon derivatives. Hydrophobic teicoplanin aglycon carboxamides (2) (4) containing var ious hydrophobic residues inhibited all tested human PKs. The IC 50 values for these compounds were in low micromolar range (Table 1) , except for CK2α inhibition by (4) (IC 50 , 18 μM). However, this compound was also inefficient against rat liver protein kinases CK1 and CK2 (IC 50 , 34.76 and >40 μM, respectively) ( Table 3) .

Human CK2α was also less sensitive to the teicoplanin aglycon derivatives (6) and (9) (100 and 83% residual activity at 10 μM concentration, respectively) ( Table 2 ). Compound (6) showed no or weak inhibitory activity (>57% residual activity at 10 μM) toward all investigated human PKs.

Teicoplanin derivatives (2) and (7) (10) demonstrat ed significant inhibitory activity against tested PKs (resid ual activities, <10% at 10 μM concentrations). Among them, compounds (2) (IGF1R, SRC), (7) (Aurora B, IGF1R, SRC), and (9) (IGF1R) were strong inhibitors that suppressed enzymatic activity by more than 50% when used at 1 μM concentration ( Table 2) .

Compounds (2), (5), (7), (9), and (10) efficiently inhibited isolated rat liver CK2 (IC 50 , ~ 2.70 4.00 μM), while compounds (1), (4), and (6) were almost inactive against this enzyme (IC 50 ≥ 40 μM) ( Table 3) .

Unlike native CK2, CK1 was less sensitive to most of the investigated compounds. The IC 50 values for CK1 inhibition were <10 μM only for (5) and (7), while for compounds (1), (2), (4), (6), (9), and (10) they were much higher (~20.29 40.00 μM).

It is important to mention that the inhibitory activi ties of (2), (4), (6), (7), (9), and (10) against CK2 were evaluated using both recombinant and native proteins (Tables 1 3) and correlated with the activities of these compounds in other tests. This holds true for the low inhibitory activity of (6), as well as for higher inhibitory activities of compounds (2), (7), (9), and (10).

Kinetic analysis of CK2α inhibition by the teicoplanin aglycon analogue (7) showed that this com pound suppressed the enzyme activity by the non com petitive mechanism (with respect to ATP) (Fig. 3a) . To further investigate the action mechanism of compound (7), CK2α inhibition was studied at constant ATP con centration in the presence of varying peptide substrate concentrations. As shown in Fig. 3b, compound (7) also showed a non competitive inhibition with respect to the peptide substrate. This is a remarkable result because the majority of kinase inhibitors compete with ATP for binding in the same active site [23] . CK2 inhibitors that are non com petitive with respect to ATP have been rarely reported.

Eremomycin aglycon derivatives. Eremomycin agly con (11) and hexapeptide (12) showed moderate or low inhibitory activities (IC 50 , ~12 and >100 μM, respective ly) against the tested human PKs ( CDK2/CycA, CK2α 1, VEGFR2) were moderate or low (IC 50 , 11.6 75.4 μM). Compound (15) showed remark ably high inhibitory activity (IC 50 , ≤ 10.1 μM) against only four PKs (Aurora A, Aurora B, IGF1R, PRK1). Human CK2α was most resistant to all eremomycin derivatives ((11) (15)) (IC 50 , 14.5 to >100 μM) and teicoplanin aglycon derivative (6) (IC 50 , 18 μM).

The relationship between structure, inhibitory activity against PKs, and antiviral activity. Natural antibiotics eremomycin and teicoplanin (Fig. 1) did not inhibit rat liver CK1 and CK2 in our experiments (IC 50 , ≥ 40 μM) .  Aglycon derivatives (1), (11), and (12) showed moderate to low inhibitory activities against the tested PKs (IC 50 , 12 to >40 μM) (Tables 1 and 3) . However, aglycon derivatives bearing one or two hydrophobic residues (Fig. 2, a and b) inhibited different PKs in micromolar concentrations (Tables 1 3) . In all the cases, hydrophobic substituent in teicoplanin aglycon amides (2) (6), N acyl derivative (7), amides of N acyl derivative (8) (10), as well as in eremomycin aglycon or eremomycin hexapeptide amides (13) (15) was essential for the inhibitory activity of these compounds against PKs.

Comparison of inhibitory efficiencies of teicoplanin and eremomycin aglycon carboxamides bearing the same residue X = NH(CH 2 ) 10 NH 2 showed that teicoplanin derivative (4) was significantly more active against the tested human PKs (Aurora A, Aurora B, AXL, BRAFV600E, CDK2/Cyclin A, IGF1R, MET, PLK1, PRK1, SRC, VEGFR2) than eremomycin derivative (15) (IC 50 , ~ 0.78 4.00 μM vs. 8.3 33.6 μM, respectively) due to the differences in their structures. Teicoplanin, in com parison with eremomycin, has an aromatic hydrophobic amino acid at position 3 instead of asparagine (Asn) and additional chlorine atom in the amino acid 6 of the pep tide core.

It have been found earlier that several compounds ((2) (5), (7) (10)) inhibiting different PKs in our experi ments suppress some coronaviruses, such as HIV 1, HIV 2, feline infectious peritonitis virus (FIPV), severe acute respiratory syndrome associated coronavirus (SARS CoV), and flaviviruses, including hepatitis C virus (HCV), Dengue virus (DENV), yellow fever virus (YFV), tick borne encephalitis virus (TBEV), and Japanese encephalitis virus (JEV) in micromolar concentrations in in vitro tests [11, 14, 16 18] . These compounds exhibit no cytotoxicity against host cells (Vero B cells for SARS CoV, DENV 2, YFV 17D, JEV, and TBEV; CRFK cells for FIPV; CEM cells for HIV) with the selectivity index CC 50 /EC 50 > 10 (where CC 50 is compound concentration required to inhibit host cell proliferation by 50%).

There are several publications about known specific PKs inhibitors suppressing virus replication. For exam ple, 2 (2 chlorophenyl) 5,7 dihydroxy 8 [(3S,4R) 3 hydroxy 1 methylpiperidin 4 yl]chromen 4 one (flavopiridol) is the first CDK inhibitor in human clinical trials that blocks HIV replication [5] . Also, the CK2 inhibitor 2 dimethylamino 4,5,6,7 tetrabromo 1H benz imidazole (DMAT) prevents HCV production in the host cell [24] .

The relationship between inhibitory and antiviral activities of glycopeptide derivatives may be of great inter est for the development of antiviral drugs, since it has been shown that human cyclin dependent kinases (CDKs) and CK2 are required for HIV 1 replication [5, 6, 25, 26] , whereas CK1, CK2, PLK1, and SRC are involved in HCV replication [9, 27 29] .

Eremomycin aglycon derivatives (13) (15) contain ing hydrophobic residues have been investigated as anti HIV 1 agents [16] . The antiviral activity was high for (13) (EC 50 , 1.6 μM) and moderate for (14) and (15) (EC 50 , 12 and 15 μM, respectively) ( Table 1) . Hydrophobic teicoplanin aglycon derivatives (amides (2) (5), (7)) effi ciently inhibited human PKs and CK2 (IC 50 , 10 μM) (Tables 1 3) , as well as exhibited anti HIV 1 activity with EC 50 values between 2.6 and 15 μM [11] . This demon strates that the inhibitory activity of several investigated compounds against PKs correlates with their antiviral activity displayed at micromolar concentrations.

Hydrophobic teicoplanin aglycon derivatives (7) (10) strongly inhibiting human PKs (Aurora B, AXL, IGF1R, MET, PLK1, SRC, VEGFR2) (22% control when used at 10 μM concentration) and CK2 (IC 50 , 10 μM) (Tables 2 and 3 ) selectively suppressed HCV replication in model subgenomic HCV replicon systems and cultured cells (EC 50 , 5.7, 7.0, 4.9, and 3.6 μM, respectively, in Huh 5 2 host cells) [14] .

The most interesting data have been obtained for the teicoplanin aglycon derivative N t butyl oxycarbonyl Table 3 . Inhibitory activity (IC 50 , μM) of teicoplanin, eremomycin, teicoplanin aglycon (1) and its derivatives (2), (4) (7), (9), and (10) against serine/threonine kinas es CK1 and CK2 from rat liver (9) . It was found that this compound targets the early stage (binding/entry) of the DENV 2 infection [17] . The fact that (9) prevents DENV 2 entry into the host cell and HCV replication does not imply the same mechanism of action. Indeed, it was demonstrated that (9) inhibits HCV replication at the post entry stage (in Huh 9 13 cells) but its target has not been established yet [14] . However, considering that DENV 2 and HCV belong to the same family (Flaviviridae), it is remarkable that the same molecule inhibits replication of such close ly related viruses by affecting two entirely different stages of the replication cycle.

In this manuscript, we have for the first time evaluat ed the inhibitory activities of 15 polycyclic peptide deriv atives (teicoplanin and eremomycin aglycons and their derivatives) against a panel of 12 human protein kinases and rat liver protein kinases CK1 and CK2.

Teicoplanin aglycon analogue (7) was a rare example of non competitive inhibitor of protein kinase activity with respect to both ATP and peptide substrate, as demonstrated by kinetic analysis of its inhibitory action on recombinant CK2α.

Several investigated hydrophobic analogues of poly cyclic glycopeptide antibiotics ((2) (5), (8) (10), (12)) that have been earlier found to suppress replication of HIV, HCV, and the other corona and flaviviruses [11, 14, 16 18] inhibited different PKs in the micromolar concen tration range (Tables 1 3) .

The patterns common for PK inhibition and antivi ral activity of the investigated compounds were estab lished: (i) both inhibitory and antiviral activities are determined by the base aglycon structure; (ii) hydropho bic residue in aglycon is essential for the inhibitory activ ity; (iii) the position of this hydrophobic residue in agly con is not important; (iv) the number of hydrophobic residues (one or two) is not important.

Glycopeptide aglycon derivatives are promiscuous inhibitors of PKs, and their inhibitory activity depends on structure. The fact that they inhibit different types of PKs may explain their activity against unrelated viruses (HIV, HCV, etc.) Moreover, the levels and activities of PKs in virus infected cells, as well as the inhibitory potential of glycopeptide aglycon derivatives and their ability to per meate the host cell membrane, can modulate the antivi ral activity of these compounds.

The role of PK inhibition in the antiviral activity of hydrophobic glycopeptide aglycon derivatives remains to be investigated; the ability of these compounds to affect cellular protein kinases in vivo still has to be verified. Since ATP concentrations used in the in vitro assays are significantly below average intracellular ATP concentra tions, the potency of kinase inhibitors in vitro does not necessarily reflect their potency in cells. Therefore, fur ther optimization of glycopeptide derivatives will aim to improve their potency, as well to demonstrate the ability of these compounds to inhibit intracellular protein kinas es.

The search for antiviral compounds among protein kinase inhibitors seems to be very promising. These novel approaches may pave the way to new prospects in drug discovery, while simultaneously providing valuable tools for studying signal transduction and viral replication.

Moreover, such studies may be useful for under standing the mechanism of the antiviral activity of hydrophobic glycopeptide aglycons and promote the development of a novel class of PKs inhibitors based on glycopeptides antibiotics and related compounds.

Ten things you should know about protein kinases

Protein kinase inhibitor

Viral protein kinases and protein phos phatases

Inhibitors of cellular kinases with broad spec trum antiviral activity for hemorrhagic fever viruses

Flavopiridol (L86 8275, HMR1275) is a cyclin dependent kinase (Cdk) inhibitor in clinical trials as a cancer therapy that has been recently shown to block human immunodeficiency virus

Advances on cyclin dependent kinases (CDKs) as novel targets for antiviral drugs

Frontiers in Protein Research

Active cAMP dependent protein kinase incorporated within highly puri fied HIV 1 particles is required for viral infectivity and interacts with viral capsid protein

Involvement of hepatitis C virus NS5A hyperphosphorylation mediated by casein kinase I alpha in infectious virus production

HIV chemotherapy and prophylax is: new drugs, leads and approaches

Antiretroviral activity of semisynthetic derivatives of glycopeptide antibi otics

Polycyclic peptide and glycopeptide antibiotics and their derivatives as inhibitors of HIV Entry

Anti influenza virus activity and structure activity relationship of agly coristocetin derivatives with cyclobutenedione carrying hydrophobic chains

Inhibition of hepatitis C virus replication by semi synthetic derivatives of glycopeptide antibiotics

Synthesis and biolog ical evaluation of lipophilic teicoplanin pseudoaglycon derivatives containing a substituted triazole function

Structure activity relationship studies of a series of antiviral and antibacterial aglycon derivatives of the glycopeptide antibiotics van comycin, eremomycin, and dechloroeremomycin

An ana logue of the antibiotic teicoplanin prevents flavivirus entry in vitro

Inhibition of feline (Fipv) and human (SARS) coronavirus by semisynthetic derivatives of glycopeptide antibiotics

Role of the glycopeptide framework in the antibacterial activity of hydrophobic derivatives of glycopeptide antibiotics

Endogenous phosphate acceptor proteins for rat liver cytosolic casein kinases

Protein kinase CK2 mutants defective in substrate recognition: purification and kinetic analysis

Non ATP com petitive protein kinase inhibitors as anti tumor therapeu tics

Seeding collaborations to advance kinase science with the GSK Published Kinase Inhibitor Set (PKIS)

Regulation of hepatitis C virion production via phosphory lation of the NS5A protein

Casein kinase II is a selective target of HIV 1 transcriptional inhibitors

Unique features of HIV 1 Rev protein phos phorylation by protein kinase CK2

Casein kinase II controls TBK1/IRF3 activation in IFN response against viral infection

Phosphorylation of hepatitis C virus NS5A non structural protein: a new paradigm for phosphorylation dependent viral RNA replication? Virology

Polo like kinase 1 is involved in hepatitis C virus replication by hyperphosphorylating NS5A

We are grateful to Svetlana S. Printsevskaya for fruit ful discussion and Natalia M. Malyutina for HPLC analysis (Gause Institute of New Antibiotics, Moscow).

This work was supported by the European Council 

The authors declare no conflict of interests in finan cial or any other area.

All procedures with animals were conducted in accordance with the legal requirements adopted by the Russian Federation and international organizations (EU Directive 2010/63/EU from September 22, 2010; Article 27).