Social Media

2,4-Diaminopyrimidines as inhibitors of Leishmanial and Trypanosomal dihydrofolate reductase

Description
2,4-Diaminopyrimidines as inhibitors of Leishmanial and Trypanosomal dihydrofolate reductase
Categories
Published
of 19
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
Share
Transcript
  2,4-Diaminopyrimidines as Inhibitors of Leishmanial andTrypanosomal Dihydrofolate Reductase Didier Pez, a Isabel Leal, b Fabio Zuccotto, a Cyrille Boussard, a Reto Brun, c Simon L. Croft, d Vanessa Yardley, d Luis M. Ruiz Perez, b Dolores Gonzalez Pacanowska b and Ian H. Gilbert a, * a Welsh School of Pharmacy, Cardiff University, Redwood Building, King Edward VII Avenue, Cardiff CF10 3XF, UK  b Instituto de Parasitologia y Biomedicina, Consejo Superior de Investigaciones Cientificas, C/ Ventanilla 11, 18001-Granada, Spain c Swiss Tropical Institute, Socinstrasse 57, CH-4002 Basel, Switzerland  d London School of Hygiene and Tropical Medicine, Keppel Street, London WC1E 7HT, UK  Received 13 June 2003; accepted 8 August 2003 Abstract— This paper describes the synthesis of 4 0 -substituted and 3 0 ,4 0 -disubstituted 5-benzyl-2,4-diaminopyrimidines as selectiveinhibitors of leishmanial and trypanosomal dihydrofolate reductase. Compounds were then assayed against the recombinantparasite and human enzymes. Some of the compounds showed good activity. They were also tested against the intact parasites usingin vitro assays. Good activity was found against  Trypanosoma cruzi  , moderate activity against  Trypanosoma brucei   and  Leishmaniadonovani  . Molecular modeling was undertaken to explain the results. The leishmanial enzyme was found to have a more extensivelipophilic binding region in the active site than the human enzyme. Compounds which bound within the pocket showed the highestselectivity. # 2003 Elsevier Ltd. All rights reserved. Introduction Leishmaniasis, African trypanosomiasis and Chagasdisease are major causes of mortality, mainly in thedeveloping world. There is need for new treatments forthese diseases as the current drugs are toxic, expensiveand require long treatment. The situation has beencompounded by increasing treatment failures by currentdrugs. 1 The causative organisms for these diseases arespecies and subspecies of   Leishmania  (in particular,  L.donovani  ,  L. infantum ,  L. mexicana  and  L. amazonensis ), Trypanosoma brucei   and  Trypanosoma cruzi  , respectively.We have been interested in investigating dihydrofolatereductase as a drug target for these diseases. 2  7 Dihy-drofolate reductase (DHFR) is responsible for thereduction of folates within the cell. A reaction of partic-ular importance is the reduction of dihydrofolate totetrahydrofolate. Tetrahydrofolate is then methylated tomethylene tetrahydrofolate which is a vital cofactor forthe methylation of deoxyuridine monophosphate tothymidine monophosphate. Thus inhibition of DHFRprevents biosynthesis of thymidine, leading to celldeath. DHFR has been a successful drug target for anti-malarials (pyrimethamine and cycloguanil), anti-bacter-ials (trimethoprim), and anti-cancer (methotrexate)amongst other diseases. However, these classical DHFRinhibitors show no selectivity for the leishmanial or try-panosome enzymes. 2 We have carried out modellingthat suggests 3 that there are structural differencesbetween the leishmanial and trypanosome DHFR andthe corresponding human enzyme which could beexploited for selective inhibition.Sirawaraporn et al. 8 reported some 5-benzyl-2,4-diami-nopyrimidines as selective inhibitors of leishmanialDHFR. These compounds also showed some in vitroactivity against  L. donovani   amastigotes. The mostactive and selective compound against the parasiteenzyme was the 3 0 -octyloxy derivative  1  ( n =7) (Fig. 1).We then conducted an extensive structure activity studywith the 3 0 -series of compounds and extended the studyto include  T. brucei   and  T. cruzi  . 4 Essentially, com-pounds with a chain length of 2–6 carbon atoms showed 0968-0896/$ - see front matter # 2003 Elsevier Ltd. All rights reserved.doi:10.1016/j.bmc.2003.08.012Bioorganic & Medicinal Chemistry 11 (2003) 4693–4711*Corresponding author. Tel.: +44-29-2087-5800; fax: +44-29-2087-4149; e-mail: gilbertih@cf.ac.uk  maximum enzyme activity and selectivity. We couldrationalize the data using modeling; the alkyl chainsextended towards a phenylalanine residue in the para-site enzyme active site. The closest (and presumablystrongest) interaction occurred for a chain length of about 4–6 carbon atoms. With regards to selectivity, inthe human enzyme, the phenylalanine residue has beenreplaced by an asparagine. The alkyl chain of com-pound  1  does not interact with the asparagine and isprobably folded in a different conformation, explainingthe selectivity of these compounds for the parasiteenzymes. However, the most active compounds againstthe parasite enzymes were not the most active againstthe intact parasites. The most active compounds had achain length of 8–10 carbon atoms and corresponded tothe most lipophilic that were prepared. These longerchain compounds showed toxic effects on mammaliancells, presumably due, at least in part, to inhibition of the mammalian DHFR. The compounds ( n =8–10)showed good activity against  T. brucei   trypomastigotes,but minimal activity against  L. donovani   amastigotes.The lack of activity against  L. donovani   amastigotesmay be due to problems of accessibility of the parasiteto the compounds:  T. brucei   is an extracellular parasite, T. cruzi   is an intracellular parasite in the cytoplasm of host cells, whilst  L. donovani   is also an intracellularparasite found in a vacuole within macrophages.We decided to further investigate this class of com-pounds with the aim of maximizing activity and selec-tivity for the enzyme, whilst improving in vitro activityagainst the parasite. In particular it appeared that forcompounds to have in vitro activity, they should belipophilic. The proposed target compounds are shownin Figure 2. Firstly the substitution on the 4 0 -position(compound  2 ) was investigated. Secondly substituentswere added at both 3 0 - and 4 0 -positions to see if thepresence of two alkyl substituents could maximizeinteraction with the phenylalanine residue and increasethe lipophilicity. This was done in two ways: both sub-stituents on the 3 0 - and 4 0 -positions were kept the same(compound  3 ) or with both substituents being different(compound  4 ).In addition, further modeling studies were undertakento try and understand interaction of the inhibitors withthe active site and the basis of selectivity for the parasiteenzymes. Chemistry Compounds were prepared based on methodologyreported previously. 4 A number of different strategiesare described below. Compounds of series 2 4-Hydroxybenzaldehyde ( 6 ) was alkylated (Scheme 1).The conditions for alkylation were optimized and it wasfound that the best conditions were the alkyl iodide in 2-butanone with potassium carbonate as base. 9 The etherswere then condensed with 3-ethoxypropionitrile usingsodium ethoxide as base. Use of modified Dean–Starkconditions seemed to improve this step, where waterwas removed by molecular sieves. A final cyclisationwith guanidine provided compounds  2c ,  2d ,  2f   and  2h  ingood yields.An alternative strategy to compounds of type  2  con-sisted in protection of 4-hydroxybenzaldehyde as theTHP compound, condensation with 3-ethoxypropioni-trile, cyclisation with guanidine and then removal of theprotecting group (Scheme 2). The resulting 4-hydroxyphenol ( 2a ) was alkylated, using potassium carbonate as Figure 1.  The lead structure. Figure 2.  The target molecules.4694  D. Pez et al./Bioorg. Med. Chem. 11 (2003) 4693–4711  base in ethanol to give other compounds in series  2  ( 2e , 2g ,  2i ). Compounds of series 3 3,4-Dihydroxybenzaldehyde ( 13 ) was alkylated under avariety of conditions to give the di-substituted com-pounds  14  (Scheme 3). Condensation with 3-ethoxy-propionitrile followed by condensation with guanidinegave the compounds of series  3 . Compounds of series 4 These compounds are differentially substituted on the3 0 - and 4 0 -positions, so a different strategy had to beapplied (Scheme 4). 3,4-Dihydroxybenzaldehyde ( 13 )was treated with one equivalent of benzyl bromide usingacetone as solvent and potassium carbonate as base. 10 Only one regioisomer was found in 77% yield. This wasdeduced to be the 4 0 -substituted compound  16 . Thereason for selective alkylation appears to be the greateracidity of the 4 0 -hydroxyl. The regioselectivity of alkyl-ation was established by a long-range NOESY spectrumwhich showed a signal enhancement of the 5 0 -hydrogen,but not of the 2 0 -hydrogen on irradiation of the benzylicCH 2  (Fig. 3). Subsequently alkylation with hexyl iodidegave the differentially substituted compound  17 .Removal of the benzyl protecting group was initiallyaccomplished with trimethylsilyl iodide. However due tothe reactivity of the trimethylsilyl iodide, different con-ditions were investigated for deprotection of the benzylgroup. Hydrogenolysis using 10% palladium on char-coal gave the required mono-alkylated compound  18 .Use of palladium hydroxide as catalyst lead to reduc-tion of the aldehyde group to the methyl. 11,12 Finallyalkylation of the free hydroxyl group was carried outwith propyl iodide and octyl iodide, followed by con-densation with 3-ethoxypropionitrile and then guanidineto give the required final compounds. Biological AssaysEnzyme assays Compounds were assayed against the recombinant  L.major ,  T. cruzi   and human enzymes. The data for thisare shown in Table 1. The presence of a 4 0 -alkoxy deri-vative appeared to increase activity and selectivity forthe parasite enzymes (series  2 ). The most active andselective compounds were the hexyl derivative ( 2d )against the  L. major  enzyme and the methyl ( 2b ) andheptyl ( 2e ) derivatives against the  T. cruzi   enzyme. Inaddition both the benzyl and THP derivatives were Scheme 1.Scheme 2. D. Pez et al./Bioorg. Med. Chem. 11 (2003) 4693–4711  4695  active and selective for both of the parasite enzymes.However no distinct trends could be observed. The dis-ubstituted derivatives (series  3 ) did not seem to make asignificant difference in selectivity or activity comparedto the corresponding mono-substituted derivatives. Seethe section of modeling for further discussion of thisdata. In vitro assays The compounds were also assayed against the clinicallyrelevant stage of the intact parasites (Table 2). For  L.donovani   and  T. cruzi   this was the amastigote stage cul-tured in mammalian cells. The compounds were alsoassayed against the related parasite,  Trypanosoma brucei rhodesiense , which causes African trypanosomiasis(sleeping sickness). For series  2  compounds, none of thecompounds were notably active against  L. donovani  .Against  T. cruzi   and  T. b. rhodesiense , the activityappeared to increase for the longer alkyl chain deriva-tives ( 2c  –  2h ); however the selectivity for the parasiteover the mammalian cells did not seem to be improved.Interestingly, the THP derivative  2j  appeared to havegood activity and selectivity for both of these organ-isms. For bis-substituted compounds (series  3  and  4 ), anumber of compounds showed moderate activity, mostnotably the bis-hexyl derivative  3b . However, the selec-tivity was not high. In vivo assays Compounds  2d ,  2g ,  2h  and  3b  were considered activeenough for assays against a rodent model of Chagas’disease. At a dose of 50 mg/kg these compounds showedno effect on the progression of the disease. Modelling In order to find a correlation between the compoundsstructure and their activity, molecular modelling wasundertaken. The first step was to dock our compoundsinto the active site of the human,  L. major 13 and  T.cruzi  3 enzymes. Two programmes were considered,Genetic Optimisation for Ligand Docking (GOLD) 14 and FlexX 15 part of the Sybyl suite of software. 16 Thesehave scoring functions, which may give rise to somecorrelation between enzyme inhibition and score. Tovalidate the methodology, two known structures wereused: the crystal structure of the human enzyme withfolate bound in the active site (pdb 1DFR), and thecrystal structure of the  L. major  enzyme 13,17 with meth-otrexate bound in the active site. Firstly, folate wasremoved from the active site of the human enzyme andthen docked into the same active site using the pro-grammes FlexX and GOLD. The same operation wascarried out with the active site of the  L. major  enzymeby removal and subsequent redocking of the metho-trexate. With GOLD, the docking of folate and metho-trexate failed as the two ligands did not fit properly intotheir own active sites. It is reported that this programmefailed with very flexible compounds or with a stronglipophilic character, which is the case with folate andmethotrexate. 14 With the progamme FlexX, the folatewas docked in a very similar orientation to that found inthe crystal structure (RMSD=1.43 A ˚) (Fig. 4a). Simi- Scheme 3. 4696  D. Pez et al./Bioorg. Med. Chem. 11 (2003) 4693–4711  larly, methotrexate was docked into the active site of the L. major  enzyme and gave a very similar orientation tothat of methotrexate in the active site in the crystalstructure (RMSD=1.65 A ˚) (Fig. 4b).FlexX was then used to dock inhibitors  2  –  4  into theactive sites of the human and  L. major  enzymes, andalso into the active site of a homology model of the  T.cruzi   enzyme. 3 The compounds with their highest Consensus Score 18  20 appeared to dock in a reasonable conformation.Consensus Score or Cscore takes into account a numberof different scoring functions:   G_Score: computes the hydrogen bonding ener-gies, complex–ligand energies and internalligand–ligand energies. 15   PMF_Score: analyses the free energies of inter-actions for protein–ligand atom-contact pairs. 21   D_Score: computes the charge and Van derWaals interactions between the protein andligand. 22   ChemScore: includes hydrogen bonding, metal– ligand interaction, lipophilic contact and rota-tional energy. 20 There was no clear correlation between any of these scor-ing functions and enzyme inhibition for each compoundseries. Scheme 4.Table 1.  Inhibition constants ( K  i ) of compounds against  L. major ,  T.cruzi   and human dihydrofolate reductaseNo. R  L. major T. cruzi   Human K  i  ( m M)  K  i  ( m M)  K  i  ( m M) 2a  H 2.8 (1.2) 2.3 (1.5) 3.4 2b  Me 2.8 (2.0) 0.034 (163) 5.5 2c  Pentyl 0.45 (3.6) 0.99 (1.6) 1.6 2d  Hexyl 0.007 (331) 0.42 (5.8) 2.4 2e  Heptyl 0.86 (7.1) 0.04 (152) 6.1 2f   Octyl 16.7 (1.2) 0.48 (40.4) 19.4 2g  Nonyl 0.81 (3.1) 1.1 (2.2) 2.5 2h  Decyl 5.6 (1.5) 2.8 (3.0) 8.3 2i  Benzyl 0.22 (11) 0.03 (84) 2.5 2j  THP 0.32 (19) 0.037 (163) 6.0 3a  Propyl 2.1 (2.9) 0.55 (10.8) 5.9 3b  Hexyl 3.2 (0.51) 0.35 (4.6) 1.6 3c  Octyl 17.4 (1.5) 4.4 (5.8) 25.4 3d  Decyl 0.22 (7.4) 2.7 (0.61) 1.7 4a  Propyl 0.68 (4.5) 0.049 (62.5) 3.0 4b  Octyl 3.2 N.D. N.D. 1 a Octyl 0.097 (25) 1.1 (2.1) 2.4Selectivity values [calculated as  K  i  (human)/ K  i  (parasite)] are shown inparentheses. a Data obtained in our laboratories 4 for compound  1 ,  n =7. D. Pez et al./Bioorg. Med. Chem. 11 (2003) 4693–4711  4697
Search
Similar documents
View more...
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks
SAVE OUR EARTH

We need your sign to support Project to invent "SMART AND CONTROLLABLE REFLECTIVE BALLOONS" to cover the Sun and Save Our Earth.

More details...

Sign Now!

We are very appreciated for your Prompt Action!

x