Development of Novel Peptide-Based Michael Acceptors Targeting

Subscriber access provided by Queen Mary, University of London

Article

Development of Novel Peptide-based Michael Acceptors Targeting Rhodesain and Falcipain-2 for the Treatment of Neglected Tropical Diseases (NTDs) Santo Preveti, Roberta Ettari, Sandro Cosconati, Giorgio Amendola, Khawla Chouchene, Annika Wagner, Ute A. Hellmich, Kathrin Ulrich, R. Luise Krauth-Siegel, Peter R. Wich, Ira Schmid, Tanja Schirmeister, Jiri Gut, Philip J. Rosenthal, Silvana Grasso, and Maria Zappalà J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.7b00405 • Publication Date (Web): 01 Aug 2017 Downloaded from http://pubs.acs.org on August 1, 2017

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Journal of Medicinal Chemistry is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 48

Journal of Medicinal Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 48

Development of Novel Peptide-based Michael Acceptors Targeting Rhodesain and Falcipain-2 for the Treatment of Neglected Tropical Diseases (NTDs) Santo Previti,† Roberta Ettari,*,† Sandro Cosconati,‡ Giorgio Amendola,‡ Khawla Chouchene,§ Annika Wagner,ǁ,ϴ Ute A. Hellmich,ǁ,ϴ Kathrin Ulrich,# R. Luise Krauth-Siegel,# Peter R. Wich,⊥ Ira Schmid,⊥ Tanja Schirmeister,⊥ Jiri Gut,∇ Philip J. Rosenthal,∇ Silvana Grasso† and Maria Zappalà† †

Department of Chemical, Biological, Pharmaceutical and Environmental Sciences, University of Messina, Italy, Viale Annunziata, 98168 Messina, Italy ‡

DiSTABiF, University of Campania Luigi Vanvitelli, Via Vivaldi 43, 81100 Caserta, Italy

§

Laboratoire de Chimie des Substances Naturelles UR/11-ES-74, Faculté des Sciences de Sfax, Université de Sfax, Route de l'aeroport, 3000 Sfax, Tunisia,

ǁ

Institute of Pharmacy and Biochemistry, University of Mainz, Johann-Joachim-Becherweg 30, Mainz, DE 55128 Germany,

ϴ

Centre for Biomolecular Magnetic Resonance (BMRZ), Goethe-University Frankfurt, Max-vonLaue-Str. 9, DE 60438 Frankfurt am Main, Germany, #

Biochemistry Center, Im Neuenheimer Feld 328, 69120 Heidelberg, Heidelberg University, Germany,

⊥ Institute

of Pharmacy and Biochemistry, University of Mainz, Staudingerweg 5 Mainz, DE 55128 Germany,

∇Department

of Medicine, San Francisco General Hospital, University of California, 1001 Potrero Ave, San Francisco CA 94110, California, USA 1

ACS Paragon Plus Environment

Page 3 of 48

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


ABSTRACT This paper describes the development of a class of peptide-based inhibitors as novel antitrypanosomal and antimalarial agents. The inhibitors are based on a characteristic peptidesequence for the inhibition of the cysteine proteases rhodesain of T. b. rhodesiense and falcipain-2 of P. falciparum. We exploited the reactivity of novel unsaturated electrophilic functions such as vinyl-sulfones, -ketones, -esters and –nitriles. The Michael acceptors inhibited both rhodesain and falcipain-2, at nanomolar and micromolar level, respectively. In particular, the vinyl ketone 3b has emerged as a potent rhodesain inhibitor (k2nd= 67·106 M-1 min-1), endowed with a picomolar binding affinity (Ki = 38 pM), coupled with a single-digit micromolar activity against T. b. brucei (EC50 = 2.97 µM), thus being considered as a novel lead compound for the discovery of novel effective antitrypanosomal agents.

INTRODUCTION Neglected tropical diseases (NTDs) are a group of disabling infections particularly endemic in developing regions of Africa, Asia, and the Americas. Over one billion people suffer from one or more NTDs. Two of the most important diseases are Human African Trypanosomiasis (HAT) and malaria.1 HAT (also known as sleeping sickness) is an endemic parasitic disease which occurs in 36 countries in sub-Saharan Africa, with around 10 000 new cases reported each year.2 It is caused by two subspecies of Trypanosoma: T. b. gambiense, particularly widespread in central and western Africa and responsible for a chronic form of the disease, and T. b. rhodesiense, the most common

2



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 48

subspecies in southern and eastern Africa, which causes a rapid-onset acute form of HAT, with a higher mortality rate.3 Malaria is the most widespread and severe tropical infectious disease; in humans, it is caused by several species of the Plasmodium genus, with P. falciparum being the most dangerous and most prevalent species.4 Although a number of antitrypanosomal and antimalarial agents are available, these suffer from problems with dosing schedules, toxicity, and increasing drug resistance.5-7 Thus, there is an urgent need to identify new effective drugs, ideally directed against novel targets. To address this need, we focused our attention on parasite cathepsin L-like cysteine proteases, notably rhodesain and falcipain-2 (FP-2), which have been recognized as novel promising targets for the treatment of HAT and malaria because of their key roles for parasite survival.8-11 Rhodesain is a Clan CA, family C1 (papain-family) cathepsin L-like cysteine protease that is essential for T. brucei survival. Its importance stems from several functions, such as its role in crossing the blood brain barrier, thus inducing the neurological stage of HAT;12 other functions include also turnover of variant surface glycoproteins (VSGs) that coat trypanosomes,13 degradation of host immunoglobulins to reduce the host immune response,14 and degradation of parasite and imported host proteins within lysosomes.15 FP-2, a Clan CA, family C1 cysteine protease of P. falciparum, hydrolyzes hemoglobin to provide amino acids that are essential to the parasite for protein synthesis. FP-2 may also be responsible for the cleavage of the cytoskeletal proteins ankyrin and band-4.1 to facilitate rupture of the red-cell membrane.16-17 Both proteases are characterized by a left (L) and a right (R) domain, with the catalytic triad (Cys/His/Asn) located in a cleft between the two domains. In recent years, our group has been actively involved in the development of novel rhodesain and FP-2 inhibitors for the treatment of HAT18-25 and malaria26-33. Peptidyl vinyl sulfones were previously identified as potent inhibitors of rhodesain8-9and FP-2.10 In the present work, we 3


Page 5 of 48

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


designed a new series of peptide-based inhibitors 1-4a-d (Chart 1) as potential rhodesain and FP-2 inhibitors, starting from the vinyl sulfones K11777 (5) and K11002 (6) (Chart 2) as lead compounds.34-37 R

O O

N H

R'

H N

EWG

O

1-4a-d

EWG = COOCH3, COCH3, SO2Ph, CN R = CH(CH3)2, Ph R' = H, NO2

Chart 1. Structure of the Michael acceptors 1-4a-d.

N

H N

N O

O N H

S O

O

O

H N

N O

5

O N H

S O

O

6

Chart 2. Structure of model compounds 5 and 6. Considering that rhodesain and FP-2 can tolerate a range of bulky hydrophobic residues within the S2 pocket and exhibit a strong preference for P2 Phe or Leu, respectively, we introduced these residues at the P2 site, with a hPhe inserted at the P1 site, analogous to the structures of both 5 and 6. The N-terminal amino group was protected with a carbobenzyloxy (Cbz) or p-nitrobenzyloxycarbonyl (pNZ) group, the latter in agreement with the structure of inhibitor 7 (Chart 3), synthesized by our group,25 with a Ki value of 6.81 µM towards rhodesain and an EC50 value of 14.8 µM against cultured T. brucei.

4



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Br

Ph O O

N H

O2N

Page 6 of 48

N

N O

O

7

Chart 3. Structure of model compound 7. Michael acceptors were of interest as inhibitor warheads, due to their ability to covalently trap the active site thiol function of cysteine proteases. Herein we report the synthesis and biological investigation of compounds 1-4a-d against rhodesain and FP-2 and against cultured T. b. brucei and P. falciparum. In silico studies were also performed to elucidate at the molecular level the binding mode of our compounds and rationalize the activity of the newly discovered anti-parasitic agents.

RESULTS AND DISCUSSION Chemistry. The synthesis of Michael acceptors 1-4a-d (Scheme 1) was accomplished starting from the acids 8-11 and the amine 12, synthesized according to literature procedure.38-40 The carboxylic acids 8-11 were coupled to the amine 12 in the presence of HOBt and EDCI, as coupling reagents, to get the dipeptides 13-16 spanning the P1-P3 region and bearing the required terminal olefin for further functionalization via cross-metathesis methodology. These intermediates 13-16 were reacted with the cross-metathesis (CM) partners 17a-d containing the appropriate electron-withdrawing (EWG) group. This synthetic approach allowed us to recover the CM products in excellent yields, particularly for the highly reactive vinyl esters and ketones. The reaction was carried out by employing the Hoveyda-Grubbs 2nd generation catalyst, a homogeneous ruthenium carbene complex easy to handle in air and tolerant towards a large variety of functional groups.41 Furthermore, the use of microwave irradiation to promote the CM reaction resulted in reduced reaction times (from 24 h to 2-4 h) and by-product formation. 5


Page 7 of 48

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Scheme 1a O O

N H

R'

R

O

R OH

O

a

N H

R'

O

H N O

H2N 8-11

12

R 8, 13, 1a-d = CH(CH3)2 9, 14, 2a-d = CH(CH3)2 10, 15, 3a-d = Ph 11, 16, 4a-d = Ph

13-16

R' H NO2 H NO2

b

EWG 17a-d

R

O O

EWG O

O O a

b

O O S Ph c

N

N H

R'

H N

EWG

O

d 1-4a-d

a

Reagents and conditions: (a) HOBt, EDCI, dry DMF/CH2Cl2 (1:1), 0 °C, N2, then 10 min DIPEA,

12, rt, 12 h; (b) Hoveyda-Grubbs 2nd catalyst, dry CH2Cl2, 100-150 °C, MW, 2-4 h. Biological activity. All compounds 1-4a-d were tested against recombinant rhodesain by using Cbz-Phe-Arg-AMC as a fluorogenic substrate.42 First, a preliminary screening at a fixed inhibitor concentration of 50 µM was performed. An equivalent volume of DMSO was used as negative control, and E-64 (18),43 the irreversible standard inhibitor of clan CA family C1 cysteine proteases (papain family), was used as positive control. All compounds except 2d inhibited the enzyme activity by 85-100 %. Continuous assays were then performed at seven different concentrations ranging from minimally inhibitory to fully inhibitory concentrations, to determine the first-order rate constants of inhibition kinac (min-1), the dissociation constants Ki (nM), and the second-order rate constants of inhibition k2nd (M-1 min-1), as k2nd= kinac/Ki (Table 1). All compounds were shown to inhibit rhodesain in an irreversible manner, as demonstrated by analysis of the progress curves at seven different concentrations (see e.g. 1b, Figure 1) and in agreement with literature data.42 6



6000

Fluorescence units

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 48

4000

2000

0 0

200 400 600 800 1000 1200 1400 1600 Time [sec]

Figure 1. Progress curves of substrate hydrolysis in the presence of the inhibitor 1b. Inhibitor concentrations (from top to bottom): 0, 0.005, 0.01, 0.025, 0.05, 0.1, 0.5, 1 µM. With exception of nitriles 1d, 3d and 4d, which inhibited rhodesain with poor k2nd values, (< 19·x 103 M-1 min-1), all the other Michael acceptors exhibited k2nd values in the range 152·x 103 67000·x 103 M-1 min-1 (Table 1). Relative inhibitory activity against rhodesain followed the pattern ketones >> esters > sulfones > nitriles. Regarding the P2 site, generally, inhibitors with a Phe residue were more active than the corresponding Leu derivatives (e.g. 4b vs 2b, 3b vs 1b, 3a vs 1a). Concerning the N-protecting group, the Cbz-derivatives were more potent than the corresponding inhibitors with a pNZ group at the N-terminal moiety (e.g. 3b vs 4b, 3c vs 4c, 3a vs 4a). Overall, the most potent compounds towards rhodesain were the vinyl ketones 3b and 4b which showed k2nd values of 67000·x 103 M-1 min-1 (3b) and 24354 x 103 M-1 min-1 (4b), coupled with the highest binding affinity expressed by Ki values of 38 and 74 pM, respectively. Selectivity assays were also performed by testing the active inhibitors against a papain-family human cysteine protease, i.e. cathepsin L, on the basis of the high structural homology between cathepsin L and rhodesain. An equivalent volume of DMSO was used as negative control, and 18, also in this case, was used as positive control. 7


Page 9 of 48

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


The results of the evaluation indicated that all tested compounds inhibited cathepsin L to a lower extent with respect to rhodesain (Table 2). The best selectivity was generally shown by vinyl ketones 1b, 3b and 4b and by vinyl sulfones 2c and 3c; in these cases, in fact, the second-order rate constants of inhibition for cathepsin L are 1–2 orders of magnitude lower than those observed for rhodesain inhibition.

8



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 48

Table 1. Activity of the Michael acceptors 1-4a-d towards rhodesain

R

O O R'

N H

H N

EWG

O

Rhodesain Comp

R

R’

EWG

H 1a CH(CH3)2 NO2 2a COOMe H 3a Ph NO2 4a H 1b CH(CH3)2 NO2 2b COMe H 3b Ph NO2 4b H 1c CH(CH3)2 NO2 2c SO2Ph H 3c Ph NO2 4c H 1d CH(CH3)2 NO2 2d CN H 3d Ph NO2 4d 18 a Compound 2d did not pass the initial screening

kinac (min-1)

Ki (nM)

k2nd (x 103 M-1 min-1)

0.0038±0.0002 0.0023±0.00004 0.0018±0.0001 0.0012±0.0001 0.0016±0.0004 0.0027±0.0009 0.00255±0.00075 0.0017±0.0006 0.0028±0.0003 0.0043±0.0005 0.0033±0.0004 0.0024±0.0005 0.001±0.0001 n.d.a 0.0005±0.0002 0.0004±0.0005 0.009±0.0004

7.5±0.4 17±6 2.0±0.2 3.6±0.02 0.5±0.2 0.9±0.1 0.038±0.011 0.074±0.032 9.1±2.6 14±3 5.0±0.6 7.2±0.03 53±1 n.d.a 46±3 1743±3 35±5

509 ±2 152±20 919±51 347±44 3795±834 2969±1152 67000±432 24354±1873 342±129 304±20 663±15 334±71 19±2 n.d.a 12±6 0.26±0.03 261± 27

9


Page 11 of 48

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Table 2. Activity of the Michael acceptors 1-4a-d towards human cathepsin-L

Comp 1a 2a 3a 4a 1b 2b 3b 4b 1c 2c 3c 4c 1d 2d 3d 4d 18

kinac (min-1) 0.0032±0.0002 0.0014±0.00005 0.0040±0.0004 0.0016±0.00005 0.0013±0.0001 0.0026±0.00005 0.002±0.0004 0.0016±0.0002 0.0009±0.00005 0.005±0.00005 0.0014±0.0009 0.003±0.0002 0.0012±0.00005 n.d. 0.0022±0.00005 n.d. 0.0032±0.0001

Human cathepsin L Ki (nM) 40±7 41±3 27±7 25±0.2 110±40 3.5±0.1 0.4±0.1 2.2±0.2 0.61±0.05 250±20 290±87 45±5 330±67 n.d. 240±11 n.d. 30±3

k2nd (x 103 M-1 min-1) 81±10 35±2 153±23 64±1 13±4 739±2 5754±523 737±14 1571±213 22±1 4.4±1.9 65±2 3.8±0.6 n.d. 9.3±0.2 n.d. 110±6

The most potent rhodesain inhibitors were tested for their antiparasitic activity against cultured T. b. brucei and, with the exception of compounds 4a and 3c, all derivatives displayed single-digit micromolar activity (Table 3). The most active compounds against cultured trypanosomes were the vinyl ketones 2b and 3b, with IC50 values of 2.48 and 2.97 µM, respectively.

10



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 48

Table 3. Antitrypanosomal activity of the most active Michael acceptors

T. b. brucei EC50 µM

Compounds 1a 3a 4a 1b 2b 3b 4b 1c 3c 4c 19 (Chlorhexidine) #

24 h 6.08±1.72 12.9±7.24 97.2±60.32 3.85±2.61 3.23±1.98 3.18±2.16 6.05±1.36 2.50±0.46 23.7±9.45 5.48±0.49 0.67±0.13

48 h 5.02±2.67 8.28±6.41 #

3.20±1.77 2.48±1.33 2.97±1.19 4.98±1.45 2.62±0.76 18.5±9.81 4.07±0.85 0.53±0.16

the compound was unstable after 24 h.

Compounds 1-4a-d were also tested against recombinant FP-2, the obtained IC50 values (Table 3) revealed a weaker interaction compared to that with rhodesain. The trend of reactivity was similar to that against rhodesain: ketones > sulfones > esters > nitriles. The active compounds displayed IC50 values in the range of 0.11->50 µM, with the strongest inhibitors the vinyl ketones 1b and 2b (IC50 values of 0.25 and 0.11 µM, respectively). Generally, all compounds active against FP-2 showed activity against cultured P. falciparum in the low micromolar range, indicating correlation between protease inhibitory and antiparasitic properties (Table 4). Further, inhibition of FP-2 and cultured parasites was consistently better with a leucine rather than phenylalanine moiety at P2, consistent with the known inhibitory specificity of the protease.44-45

11


Page 13 of 48

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Table 4. Activity of compounds 1-4a-d against FP-2 and P. falciparum

Comp 1a 2a 3a 4a 1b 2b 3b 4b 1c 2c 3c 4c 1d 2d 3d 4d 18

FP-2 IC50 µM 9.53±0.095 9.63±7.15 >50 >50 0.25±0.03 0.11±0.04 3.01±0.22 0.78±0.17 0.78±0.069 1.32±0.22 7.33±1.02 >50 >50 >50

P. falciparum EC50 µM 4.40±1.62 3.97±0.72 >10 >10 3.16±1.14 2.74±1.02 7.03±1.65 7.18±0.81 1.53±0.091 2.31±0.01 >10 >10 >10 >10

>50 >50 0.25±0.06

>10 >10

20 (Artemisinin)

0.023±0.002

21 (Chloroquine)

0.057±0.009

12



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 48

Effect of compound 3b on the thiol level of T. brucei. In trypanosomes, trypanothione [T(SH)2] is the predominant cellular thiol. Under culture conditions, bloodstream (BS) T. brucei contain about 350 µM T(SH)2 and 100 µM glutathione (GSH).46 To assess the effect of the Michael acceptors on the low molecular weight thiols, parasites were treated for 24 h with compound 3b. The cells showed a concentration-dependent growth retardation (Figure 2a) corresponding to the EC50 value of 3 µM (Table 3). Cellular GSH and T(SH)2 concentrations were measured by HPLC analysis of the monobromobimane (mBBr)-labeled thiols. The untreated control contained about 340 µM T(SH)2 which dropped to 160 µM upon treating the cells with 3 µM of compound 3b. The GSH level was only marginally affected. The higher reactivity of T(SH)2 may be explained by its thiol pK value of 7.4 which is much lower than that of GSH.47 In all samples, both thiols were almost exclusively present in reduced form indicating that compound 3b did not cause oxidation of the free thiols. The level of protein-bound GSH, which is ~ 5 µM in BS T. brucei,46 was not affected (data not shown). The drop of free thiols – mainly T(SH)2 – does, however, not explain the EC50 value of 3 µM observed for compound 3b (Table 3). Depletion of trypanothione synthetase in BS T. brucei results after 24 h in a decrease of T(SH)2 to about 50 µM without affecting cell growth.46 This is supported by previous work showing that a severe depletion of T(SH)2 is accompanied by only minimal growth retardation.48 Clearly, the parasites can survive with very low levels of T(SH)2, although under these conditions, they are more sensitive towards oxidative stress.49

13


Page 15 of 48

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 2. Thiol levels in BS T. brucei treated with compound 3b for 24 h. Cells were seeded to 1 x105 cells/ml. a) Cell density after 24 h cultivation. b) Cellular content of free thiols (GSH, T(SH)2) and free thiols plus disulfides (GSSG, TS2) quantified by HPLC analyses of mBBr-derivatized samples. All values are the mean ± SD of three independent experiments.

Finally, cytotoxicity of the most active compounds at cellular level 1b-3b and 1c was evaluated in HeLa cell lines (Table 5): these data demonstrated that the tested compounds have a selectivity index ranging from 2.6 to 4.6, in agreement with the profile of irreversible inhibitors.42

Table 5. Cytotoxicity of 1b-3b and 1c towards HeLa Cell Lines and selectivity index (SI) Comp. CC50 [µM] SI 9.58 ± 1.16 3 1b 8.20 ± 0.08 3.3 2b 7.90 ± 0.26 2.6 3b 12.17 ± 0.90 4.6 1c SI: EC50 (T. b. b.)/CC50 (HeLa cells) 14



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 48

Molecular Modeling. To rationalize the structure-activity relationship data (SARs) of the tested compounds, we performed a series of in silico experiments, consisting of docking calculations and Molecular Dynamics (MD) simulations. To this end, the crystal structure of rhodesain bound to the inhibitor 6 (PDB code 2P86)35 was selected. As this set of compounds covalently binds to the protease, a covalent docking protocol devised by Olson and colleagues was employed.50 In particular, AutoDock4.2 software51-52 was used by exploiting the possibility to include a flexible side chain during the docking simulation. This method requires the intended ligand to be joined in an arbitrary conformation with the residue which participates to covalent bond, in this case Cys25, and then the calculations to be run treating the modified residue as flexible. To probe the predictive power of the protocol, the re-docking of the co-crystal 6 ligand was performed. The software correctly managed to recapitulate the conformation of the crystal structure with a root-meansquared deviation (rmsd) of 2.37 Å compared to the crystal structure; notably, the sole significant discrepancy, between the docked binding pose and the experimental one, was in the position of the hPhe in P1 (data not shown), which, by pointing outwards, does not establish any relevant contact with the nearby residues in the crystal or in the docked pose. Therefore, we were encouraged to proceed with the in silico tests of 1-4a-d. Analysis of the results achieved for these docking calculations revealed that the binding modes of our compounds closely resemble that of 6, interacting with the same set of key residues. As representative of the obtained binding poses, the docked solution achieved for the most active ligand 3b is represented in Figure 3a and 3b.

15


Page 17 of 48

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 3. a) Superimposition of co-crystal ligand 6 (green) and 3b docked conformation (pink) in rhodesain (blue). b) Rhodesain/3b complex. The enzyme is depicted as blue ribbons and sticks and the ligand as pink sticks. Important residues are labelled. H-bonds are shown in green dashed lines. Notably, the binding of 3b, as well as of the other compounds, seems to be governed by the occurrence of several H-bond interactions with the residues lining the enzyme binding cavity. In particular, the backbone atoms of the Phe and hPhe residues of compound 3b are involved in Hbonds with the backbone atoms of rhodesain Gly66 and Asp161, respectively. In addition, most of the warhead moieties are favorably positioned to accept two H-bonds from the Trp184 and Gln19 side-chains. Indeed, the rigid nitrile warhead (1-4d) would not adopt a pose which is conducive to the formation of these two H-bonds, likely explaining the considerably lower activity of the compounds featuring nitriles (compare 1-4b vs 1-4d). As for the P2 moiety (Leu or Phe residues of the inhibitors), our docking studies confirmed that the bulky Phe residue provides a better fit into the hydrophobic S2 binding site which comprises Ala208, Leu160, Ala138, Met68 and Leu67 residues.53 This likely explains the increased activity of the Phe-containing derivatives against rhodesain and cultured trypanosomes, compared to those featuring a Leu in the P2 site (see 1b vs 3b). Of note, consistent with data for FP-2 bound to peptidyl inhibitors45 Leu was preferred at P2 16



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 18 of 48

for inhibition of FP-2 and cultured malaria parasites. The carbobenzyloxy group is lodged in the S3 pocket,53 establishing a T-shaped charge-transfer interaction with the rhodesain Phe61 side-chain. Interestingly, in compounds 2,4a-d the p-nitro substituent of the carbobenzyloxy group seems to have two opposite effects on the stability of the ligand/protein interactions. Specifically, while enhancing the above-described charge-transfer contacts, the same partially negative group is unfavorably pointing towards a negatively charged region of the S3 pocket (Asp60), which likely explains why the nitro group does not substantially contribute to ligand potency. As seen in the cocrystal ligand and for 3b and its structural congeners, the hPhe residue in P1 does not have a critical role in stabilizing the ligand/protein interactions, since it points towards the outer part of the protein active site. When analyzing the Ki constants (Table 1) it is clear that, at least for our set of compounds, the different potencies cannot be entirely ascribed to the different reactivities of the employed warheads. Instead, we suggest that the contacts established by the warheads (e.g. H-bonds) are instrumental in stabilizing the ligand in a pose which is conducive for the formation of the covalent bond with the reactive cysteine residue (Cys25). In this regard, the rigid nitrile-containing analogs were the least active ones, since they cannot adopt the proper orientation within the enzyme cleft. On the other hand, the ketone moiety seems to be well positioned to fulfill steric and electrostatic demands, which likely explains the higher potencies of these compounds 1-4b. Additionally, to overcome the hurdle of protein rigidity in docking and to probe the stability of the predicted ligand/enzyme interaction, the covalent 3b/rhodesain complex was subjected to an 80 ns long MD simulation. Table 6 reports the data regarding the stability of key H-bonds that 3b forms with the enzyme residues. Out of five H-bonds, four of them were maintained for more than 50% of the frames of the simulation, further supporting the predicted contacts.

17


Page 19 of 48

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Table 6. H-bonds statistics from the MD simulation.

a

0.8554

AvgDistb (Å) 2.82

AvgAngc (°) 163.64

Phe NH

0.7246

2.84

156.88

Gln19 – Sidechain NH2

Warhead C=O

0.6410

2.84

160.35

Trp184 – Sidechain NH

Warhead C=O

0.5489

2.86

158.56

Gly66 – Backbone NH

Phe C=O

0.4391

2.88

154.55

Enzyme Residue Atom

Ligand Atoms

Focca

Asp161 – Backbone C=O

hPhe NH

Gly66 – Backbone C=O

Frequency of occurrence of the H-Bond in the frames of the simulation. bMean distance (in

Ångstoms) of the H-bond along the simulation. cMean angle (in degrees) H-bond recorded along the simulation. To rationalize the recorded selectivities of the compounds for rhodesain over FP-2 we also covalently docked 3b into FP-2 using the crystal structure of the enzyme in complex with an inhibitor (PDB code 2OUL).54 The obtained binding pose is presented in Figure 4. Interestingly, while the overall binding pose of 3b with FP-2 is virtually superimposable to that of rhodesain, a number of changes were found in the S2 pocket of FP-2, which is known to play a key role in determining selectivity in this class of proteases.55 Specifically, Ala208, Ala138 and Gly163 of rhodesain are replaced in FP-2 by aspartate, serine, and alanine residues, respectively. These modifications make the cleft narrower and indeed more polar, thereby inducing the bulky and hydrophobic 3b Phe in the P2 position to establish less favorable contacts with the enzyme. Varied interactions with these residues may explain the observed loss of potency against FP-2 compared to that against rhodesain. Interestingly, similar results were also achieved by other authors.44,56,57

18



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 20 of 48

Figure 4. Predicted FP-2/3b complex. The enzyme is depicted as yellow ribbons and sticks and the ligand is depicted as pink sticks. Important residues are labeled. H-bonds are shown as green dashed lines.

To assess the drug-likeness of compound 3b, its physicochemical and pharmacokinetic (PK) features were profiled through the Qikprop software (Schrödinger. LLC New York) employing the same parameters used in a previous work.24 Qikprop predicts a number of PK parameters and compares them with the ranges calculated for the 95% on the marketed drugs. The results, summarized in Table 7, show that 3b would feature a substantial gut/blood barrier permeability (see QPPCaco values), and an acceptable probability of passing the blood/brain barrier according to its QPlogBB and QPPMDCK values. Furthermore, 3b is predicted to have a 100% human oral absorption. According to this analysis, our inhibitor is well suited to be a lead compound for the discovery of novel antitrypanosomal agents.

19


Page 21 of 48

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Table 7. Calculated Physicochemical and Pharmacokinetic Properties of Compound 3b.

Parameter

3b

mol_MW

484.594

Range covered by the 95% of Drugs 130.0 – 725.0

donorHB

1.250

0.0 – 6.0

accptHB

6.250

2.0 – 20.0

QPlogPo/w

6.005

–2.0 – 6.5

QPPCaco

659.806

500 great

QPlogBB

-1.387

–3.0 – 1.2

QPPMDCK

473.226

500 great

Percent Human Oral Absorption

100

>80% is high,

Development of Novel Peptide-Based Michael Acceptors Targeting

Recommend Documents