Books - Non-fiction

A cyclic PNA-based compound targeting domain IV of HCV IRES RNA inhibits in vitro IRES-dependent translation

Description
A cyclic PNA-based compound targeting domain IV of HCV IRES RNA inhibits in vitro IRES-dependent translation
Published
of 10
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
  A cyclic PNA-based compound targeting domain IV of HCVIRES RNA inhibits in vitro IRES-dependent translation Sergio A. Caldarelli, a Mohamed Mehiri, a Audrey Di Giorgio, a Amaury Martin, b Olivier Hantz, b Fabien Zoulim, b Raphael Terreux, c Roger Condom a and Nadia Patino a,* a Laboratoire de Chimie Bioorganique UMR-CNRS 6001, Universite´  de Nice-Sophia Antipolis,Parc Valrose, 06108 Nice Cedex, France b INSERM U271, Virus des he´  patites et pathologies associe´ es, 151, cours Albert Thomas, 69424 Lyon Cedex 3, France c Laboratoire LPCM2, Faculte´  de Pharmacie, Universite´  Claude Bernard Lyon 1, 8 av. Rockefeller, 69373 Lyon, France Received 30 March 2005; revised 3 June 2005; accepted 3 June 2005Available online 2 August 2005 Abstract—  A cyclic molecule  1  constituted by a hepta-peptide nucleic acid sequence complementary to the apical loop of domain IVof hepatitis C virus (HCV) internal ribosome entry site (IRES) RNA has been prepared via a   mixed   liquid-phase strategy, whichrelies on easily available protected PNA and poly(2-aminoethylglycinamide) building blocks. This compound  1  has been elaboratedto mimic   loop–loop   interactions. For comparison, its linear analog has also been investigated. Although preliminary biologicalassays have revealed the ability of   1  to inhibit in vitro the HCV IRES-dependent translation in a dose-dependent manner, the linearanalog has shown a slightly higher activity.   2005 Elsevier Ltd. All rights reserved. 1. Introduction The internal ribosome entry site (IRES) located in the5 0 -untranslated region (UTR) of the hepatitis C virus(HCV) genomic RNA mediates the cap-independentinitiation of viral translation and seems to be involvedin RNA replication. HCV IRES is a phylogeneticallyhighly conserved RNA sequence 1 which adopts anion-dependent tertiary fold including four major con-served secondary structure subdomains. 2 These overallsecondary and tertiary structures were shown to becrucial for specific recognition and binding of severalcellular components such as the eukaryotic initiationfactor eIF3 and the ribosomal subunit 40S, to formthe ribosomal initiation complex that, consequently ini-tiates translation. 3,4 Thus, the subdomain sequences in-volved in such interactions represent targets of interestfor inhibiting viral replication. In this context, thestem–loop of domain IV (Fig. 1) is particularly attrac-tive as it contains in the apical loop the start codonAUG which pairs with the initiator tRNA during thetranslation initiation process. 5,6 Among various approaches used to inhibit IRES func-tion, the antisense strategy using synthetic DNA orRNA analogs complementary to different IRESsequences has been widely studied and seems to be anattractive tool both to understand IRES mechanismsand to develop anti-HCV drugs.Thus, natural oligonucleotides, 7–11 as well as modifiedanalogs such as phosphorothioates, 7,8,12–18 alpha-ano- 0968-0896/$ - see front matter    2005 Elsevier Ltd. All rights reserved.doi:10.1016/j.bmc.2005.06.008 Abbreviations : Alloc, allyloxycarbonyl; Boc,  tert -butyloxycarbonyl;DIPEA, diisopropylethylamine; DMF, dimethylformamide; HATU, O -(7-aza-1-benzotriazolyl)- N  , N  , N  0 , N  0 -tetramethyluronium hexa-fluo-rophosphate; HOAt, 1-hydroxy-7-azabenzotriazole; Mmt, monometh-oxytrityl; PyBop, (benzotriazol-1-yloxy)tripyrrolidinophosphoniumhexafluorophosphate; TFA, trifluoroacetic acid; TFMSA, trifluoro-methane sulfonic acid; THF, tetrahydrofuran. Keywords : Cyclic PNA; HCV IRES; Domain IV; Loop–Loopinteraction.*Corresponding author. Tel.: +33 04 92 07 61 46; fax: +33 04 92 07 6151; e-mail: patino@unice.fr  Figure 1.  Structure of domain IV of HCV IRES RNA.Bioorganic & Medicinal Chemistry 13 (2005) 5700–5709  mer 9 and 2 0 -methoxyethoxy phosphodiesters, 14,18 meth-ylphosphonates, 7 benzylphosphonates, 7,19 morpholi-nos, 20 (phenylalkyl)phosphonates, 21 2 0 - O -methyl, 22 andrecently peptide nucleic acids (PNAs), 23,24 have beendeveloped to target different IRES regions. These studiesshowed that linear natural and modified oligonucleo-tides (containing, in general, 15–25 residues) could bindto domain IV very strongly, thus inhibiting in vitro theinitiation of IRES-dependent translation.Another interesting approach to inhibit stem–loop (hair-pin) structures is to mimic intermolecular loop–loopinteractions that are widely present in RNA–RNA asso-ciations and serve a diverse range of biological func-tions. Due to the increased accessibility of the bases inhairpin loops, intermolecular loop–loop interactionsare particularly well adapted to trigger molecular recog-nition and to induce RNA–RNA annealing. Severalloop–loop complexes are well described in the litera-ture, 25–29 and one can cite as an example the interactionbetween two self-complementary dimerization initiationsignal (DIS) loops responsible for the dimerization of the genomic HIV RNA. 30–33 We previously reported the liquid-phase synthesis aswell as the biological activity of two cyclic PNAmolecules 34,35 and one of them was shown in vitro tointerfere with the dimerization process of the HIV-1 gen-ome. These   loop-like   compounds, constituted by shortPNA fragments (i.e., six residues), should afford severaladvantages over linear analogs, such as a greater selec-tivity, or over classical antisense oligomers, such as alower molecular weight.In this context, we designed compound  1 , which con-tains an antisense heptameric PNA moiety complemen-tary to the seven residues of the domain IV loop and aspacer tethering the C- and N-terminal extremities of the PNA (Fig. 2). A molecular modeling study allowedus (i) to assess the length of the spacer in order to opti-mize the loop–loop interaction and (ii) to introduce aphenylalanine (F) residue at the PNA C-terminalextremity to increase stability of the complex via a  p -stacking interaction between the phenyl moiety andthe adjacent uracil nucleobase (Figs. 2 and 3).In this paper, we report the molecular design and theliquid–phase synthesis of cyclic PNA [UCAUGGU]F c 1 , as well as the ability of   1  and of its linear PNA analog5 0 -UCAUGGU-3 0 (synthesis not shown) to inhibitin vitro IRES-dependent translation. 2. Results and discussion2.1. Molecular modeling studies In the molecular model of interaction between com-pound  1  and domain IV of HCV IRES RNA, theRNA stem and the annealed section constituted by theRNA loop and the PNA fragment, form two stable dou-ble helices (Fig. 3). The axes of these helices are collinearas previously reported for the interaction between a cyc-lic PNA and the DIS RNA of HIV-1. 35 The residueA339 is not strictly aligned with the other RNA resi-dues, but this break ensures the continuity of theRNA sequence without making a kink in the doublehelix structure. The seven nucleobases of the RNA loopform Watson–Crick interactions with the complementa-ry nucleobases of the PNA fragment. The presence of aphenylalanine (F) residue on compound  1  allows tostabilize the last Watson–Crick base pair interaction(A339–U7).In the first step, a complex resulting from the interactionbetween the linear heptameric PNA bound with thephenylalanine residue and the domain IV RNA loopwas built using Insight II molecular modeling packageand the complex was energetically minimized with theCFF forcefield. In a second step, a linker of optimallength was added to close the structure. After an ener-getic minimization, the model was put under rectangularwater box filled with TIP3P water molecules. The globalcharge was neutralized with Mg 2+ counter ions. A 6 nsmolecular dynamics simulation was computed at con- Figure 2.  Structure of cyclic PNA [UCAUGGU]Fc ( 1 ). Figure 3.  Molecular model of interaction between compound  1  anddomain IV of HCV IRES RNA. Compound  1  is in green with its U7residue in blue and its Phe residue in purple. A339 of domain IV is inred. S. A. Caldarelli et al. / Bioorg. Med. Chem. 13 (2005) 5700–5709  5701  stant pressure (1 bar) and at constant temperature(300 K). The simulation was performed on a SGI Origin3800 supercomputer located at the   Centre InformatiqueNational de l  Enseignement Supe´rieur   (CINES) withthe Amber 6 molecular modeling package. The shakealgorithm was selected with 2 fs as integration time.During the first 800 ps, the temperature was linearly in-creased starting from 100 to 300 K. A conformation wassampled every 20 ps, and a close analysis of the trajecto-ry revealed that all Watson–Crick interactions re-mained, and that the length of the linker was optimal.The phenylalanine residue protects the interaction be-tween A339 and U7, since the angle between these twobases is comparable to the other base-pairs, angles inthe RNA/PNA helix. 2.2. Chemistry To synthesize compound  1 , we applied a   mixed   strategyrecently elaborated in our laboratory. 34 It relies on theprotected PNA fragments and protected poly(2-amino-ethylglycinamide) building blocks. This strategy hasbeen shown to be a good alternative to the fully protect-ed backbone (FPB) strategy 35,36 when the PNAs toprepare contain four different nucleobases, because it al-lows to circumvent the difficulty to work with a combi-nation of at least eight orthogonal protecting groups.The retrosynthetic pathway to compound  1  is illustratedin Scheme 1. This compound results from the cyclic pre-cursor  2 , a   mixed   heptameric structure that containstwo PNA units (C Z and A Z ) and a protected penta(2-aminoethylglycinamide) unit. At the C- and N-terminalextremities, a spacer constituted by a 8-aminooctanoicacid, a 6-aminocaproic acid, and a phenylalanine (F),allows to close the molecule. The secondary aminofunctions of the penta(2-aminoethylglycinamide) moietyare protected by allyl and Boc protecting groups whichare subsequently replaced, respectively with guanineand uracil acetic acid units in the last steps of the synthe-sis. Compound  2  is synthesized from spacer  3  and hep-tameric   mixed   compound  4 , which is prepared fromthree key synthons previously described: 37 PNA frag-ment  5  and the protected di- and tetra-(2-aminoethylgly-cinamide) units (respectively,  6  and  7 ).The synthesis of linker  3  is detailed in Scheme 2. It wasprepared in four steps (70% overall yield) from commer-cially available  N  -Boc-aminocaproic acid  8 , methyl8-aminooctanoate  9 , and  N  - a -Fmoc- LL -phenylalanine.The preparation of heptameric   mixed   compound  4  (cf.Scheme 1) required the synthesis of    mixed   compound 16  in four steps from compounds  5  and  6  (Scheme 3).In a first step, condensation of   N  -Z-adenine PNA mono-mer  5  with  6  by means of Bop reagent led to trimer  13 (80%). Then, a selective cleavage of the Alloc protectinggroup with Pd[PPh 3 ] 4 /Et 2 NH afforded  14 , onto whichwas condensed a  N  -Z-cytosine acetic acid unit viaHATU/HOAt activation, to give  15  in 68% overall yield.Finally, acid  16  was obtained by saponification using1 M LiOH.The condensation of   16  with protected tetramer  7  affor-ded   mixed   heptameric compound  4  (Scheme 4). Aftercleavage of the Mmt protecting group by means of asolution of 2% TFA in CH 2 Cl 2 , the resulting compound 17  was coupled with spacer  3  to give compound  18 . Thetwo condensation steps described above were performedvia a PyBop activation, respectively, in 79% and 91%yields. The next steps were consisted removal of Fmocfrom  18  with a Et 2 NH/CH 2 Cl 2  solution, to get  19 Scheme 1.  Retrosynthetic route to compound  1 . Scheme 2.  Synthesis of compound  3 . Reagents: (a) PyBrop, NMM,DMF (89%); (b) TFA, CH 2 Cl 2  (90%); (c) Fmoc-Phe-OH, Bop,DIPEA, DMF (86%); (d) Dioxane, HCl (12 N), (5:1, v/v), reflux(100%). Scheme 3.  Synthesis of compound  16 . Reagents: (a) Bop, DIPEA,DMF (80%); (b) Pd[PPh 3 ] 4 , Et 2 NH, CH 2 Cl 2  (91%); (c) C Z CH 2 CO 2 H,HATU, HOAt, DIPEA, DMF (75%); (d) 1 N LiOH, THF (99%).5702  S. A. Caldarelli et al. / Bioorg. Med. Chem. 13 (2005) 5700–5709  (97%), which was saponified using a 1 M LiOH aqueoussolution (93%) to afford  20 . At last, a head-to-tail cycli-zation of   20  via a HATU/HOAt activation and semi-high dilution conditions (10 mM) yielded cyclic   mixed  precursor  2  in 65% yield.The last stage for the synthesis of   1  consists in the intro-duction of the uracil and guanine acetic acid units togenerate the heptameric PNA moiety (Scheme 4). Toavoid unpredictable side reactions that uracil nucleo-bases are suspected to induce, 31 we introduced firstthe guanine acetic acid units. Thus, a selective cleavageof the two Alloc protecting groups by means of Pd(PPh 3 ) 4 /Et 2 NH, afforded the coupling of two OBn– guanine acetic acid units onto the two free amino func-tions of   21 , via a HATU/HOAt activation, to give deriv-ative  22  in 90% overall yield. Then, treatment of   22  withTFA/CH 2 Cl 2 /TIS led to the removal of the three Bocprotecting groups as well as to the cleavage of the benzylguanine exocyclic protecting groups. HATU-mediatedcondensation of three uracil acetic acid units onto  23 was then successfully achieved (88% yield) despite thepresence of the free exocyclic amine functions onto theguanine residues. Finally, simultaneous Z removal of cytosine and adenine nucleobases of   24  by means of aTFMSA/TFA/thioanisole solution allowed to obtaincompound  1  which was isolated after semi-preparativeHPLC (16% yield). Its purity was determined by HPLCanalyses and its structure was confirmed by MALDI-TOF experiments. 2.3. Inhibitory effect on HCV in vitro translation To investigate the inhibitory action of compound  1  andof its linear counterpart (data not shown) on HCV IRESdependent translation, coupled transcription and trans-lation experiments were performed (Fig. 4). Differentconcentrations of   1  were mixed with linearized plasmidpFl-HCV IRES-hRl leading, after transcription by T7polymerase, to the expression of a single bicistronictranscript and, after translation, to the expression of luciferase proteins. The expression of firefly luciferaseis driven by a cap-dependent mechanism and shouldnot be affected by the addition of   1 . In contrast, theexpression of hRenilla luciferase is under the controlof HCV IRES; its activity should be inhibited uponbinding of   1  to IRES sequence(s), if the interaction isadequate to disrupt the association of proteins.A dose-dependent inhibition of viral translation was ob-served (Fig. 5). Maximal effect was obtained with 30  l M Scheme 4.  Synthesis of compound  1 . Reagents: (a) PyBop, DIPEA, DMF (79%); (b) TFA, CH 2 Cl 2  (0.02:1, v/v) (88%); (c) Fmoc-PheNH(CH 2 ) 5 CONH(CH 2 ) 7 CO 2 H  3 , PyBop, DIPEA, DMF (91%); (d) Et 2 NH, CH 2 Cl 2  (97%); (e) 1 M LiOH, THF (93%); (f) HATU, HOAt,DIPEA, DMF (65%); (g) Pd[PPh 3 ] 4 , Et 2 NH, CH 2 Cl 2  (95%); (h) G OBn CH 2 CO 2 H, HATU, HOAt, DIPEA, DMF (95%); (i) TFA, CH 2 Cl 2 , TIS (93%);(j) UCH 2 CO 2 H, HATU, HOAt, DIPEA, DMF (88%); (k) TFMSA, TFA, thioanisole (100%). Figure 4.  Model used for transcription and translation experiments. Figure 5.  Inhibition assay of HCV-IRES dependent in vitro transla-tion by  1 . Varying concentrations of   1  ( l mol L  1 ) were mixed with arabbit reticulocyte lysate (Quick Master Mix) and 0.5  l g of linearizedplasmid pFl-HCV IRES-hRl. Effect of   1  on the inhibition of IRESdependent translation of hRenilla luciferase was observed. hRenillaactivity was normalized relative to firefly luciferase activity. Percent-ages are calculated relative to luciferase activity in the absence of   1 . Allexperiments were performed at least in triplicate. The percentage meanvalues ± SD are given on the curve. S. A. Caldarelli et al. / Bioorg. Med. Chem. 13 (2005) 5700–5709  5703  of   1 , which corresponds to 45 ± 5% of IRES-mediatedtranslation inhibition. Higher concentrations (up to60  l M) showed increasing inhibition of hRenilla lucifer-ase and also a non-specific inhibition of firefly luciferase.Similar results were obtained when compound  1  was as-sayed directly with RNA transcripts on rabbit reticulo-cyte lysate (decoupling translation–transcription). Thesame experiments on the linear analog also revealed adose-dependent inhibition of translation, of about5  l M (data not shown).The lower activity of cyclic PNA [UCAUGGU]Fc  1 compared with its linear analog could be explained inthe light of the work reported by Toulme´ and co-work-ers. 29,38–40 Actually, they identified by in vitro selectionthe RNA aptamers directed against several RNA hair-pins through loop–loop interactions. In all the cases,the loops of the aptamers showing the highest affinitiesare not strictly complementary to the targeted apicalloops. Thus, the aptamers targeting domain IV of theHCV IRES contain in their loops a consensus sequence5 0 -A UCAUGG -3 0 in which six residues (in bold) interactwith six of the seven residues of the apical loop of domain IV. 29 This result seems to indicate that the struc-ture of cyclic PNA  1  is not well adapted for a high-affin-ity interaction. The molecular modeling as well as thesynthesis of a cyclic PNA containing the consensus se-quence is currently under progress. 3. Conclusions A cyclic PNA-based compound ( 1 ), containing the com-plementary heptameric sequence of the apical loop of domain IV of HCV IRES, has been designed in orderto mimic highly stable loop–loop complexes. For com-parison, its linear counterpart has also been investigat-ed. The cyclic structure, of lower molecular weightthan classical antisense oligomers, should be more selec-tive than the linear analog.Compound  1  has been successfully prepared via a  mixed   liquid-phase strategy which relies on easily avail-able protected PNA and poly(2-aminoethylglycinamide)building blocks. This procedure offers the advantageover the FPB strategy of requiring fewer orthogonalprotecting groups, and enables the preparation of PNA containing the four nucleobases A, C, G, and U.Moreover, to avoid side-reactions due to the uracilmoiety, this nucleobase must be introduced at the laststages of the synthesis.Preliminary biological assays have revealed the ability of compound  1  and of its linear counterpart to interfere invitro with the HCV IRES-dependent translation in adose–dependent manner. This result suggests for thefirst time that a cyclic PNA targeting the HCV5 0 -UTR is able to specifically downregulate HCVIRES-directed translation. The lower activity of thecyclic compound compared with its linear analog couldbe explained by the results obtained by Toulme´ andco-workers. Further experiments are under progress toconfirm the binding site as well as the mode of actionof these compounds, to evaluate their selectivity andto optimize the cyclic PNA structure. 4. Experimental (chemistry)4.1. General Analytical thin-layer chromatography was conducted onMerck precoated silica gel 60F 254  plates and the com-pounds were visualized with ninhydrin test and/or byvisualization under ultraviolet light (254 nm). Chroma-tography was performed on Merck Silica gel 60 (230– 400 mesh ASTM) using the solvent systems (volume ra-tios) indicated below. Analytical HPLC chromatogramswere obtained using a Waters HPLC system (600E sys-tem controller, 996 photodiode array detector or 2487dual wavelength absorbance detector) and a MerckLichrospher 100 (250  ·  4 mm 2 ) RP-18 (5  l m) column.The HPLC flow rate was 1 mL/min, and the elution sol-vents were water (0.1% TFA) as solvent A, and acetoni-trile (0.1% TFA) as solvent B (pH 6).  1 H and  13 C NMRspectroscopies were performed using a Bruker AC 200or AC 500 Fourier Transform spectrometer. Chemicalshifts ( d ) are reported in parts per million (ppm) (in 1 H NMR descriptions s = singlet, d = doublet, t = trip-let, q = quartet, m = multiplet, and bs = broad peak).ESI mass spectra were recorded with an INCOS 500 E FINNIGAN MAT or a TSQ 7000 FINNIGAN MAT.MALDI-TOF-MS spectra were recorded with a MAL-DI-TOF   DE PRO   Applied biosystem. 4.2. Synthesis4.2.1. (UCAUGGU)F c Æ 4TFA (1).  Compound  24  (55 mg,21.7  l mol) was dissolved in 1.5 mL of TFMSA/thioani-sole/TFA (1:0.4:0.3, v/v/v) at 0   C. The mixture was stir-red at room temperature for 1 and the deprotectedproduct was precipitated out by Et 2 O (5 mL). The crudeproduct (79 mg) is obtained after washing with MeOH/Et 2 O (1:1, v/v) and drying, and it was further purifiedon a reverse phase preparative HPLC column (PrepNova-Pack, HR-C18, 100  ·  40 mm 2 ) and eluted with20% acetonitrile in water (0.1% TFA) at a flow rate of 5 mL/min to give  1  (10 mg, 16%) as a white solid afterlyophilization. HPLC (A/B 80:20 to 0:100 over 30 min) t R  = 9.7 min ( k max  = 258.4 nm). MS (ESI+) calcd forC 96 H 123 N 41 O 26  [(M+2H)/2] + : 1134.0; found: 1134.0[(M+2H)/2] + . MALDI-TOF-MS Calcd for C 96 H 123 N 41 O 26  [M+H] + : 2267.9; found: 2267.7 [M+H] + . 4.2.2.(BocC Z A Z BocAllocAllocBoc)F c (2). Toacoldsolu-tion (0   C) of   20  (67 mg, 30.2  l mol) and DIPEA (34  l L,196.2  l mol, 6.5 equiv) in DMF (3 mL), were addedHATU (17 mg, 45.9  l mol, 1.5 equiv) and HOAt (8 mg,60.4  l mol, 2 equiv). The mixture was stirred for 2 h atroomtemperature. The solvent wasevaporated underre-duced pressure. The residue was taken up in CHCl 3  andwashedsuccessivelywitha1 MaqueousKHSO 4 solution,asaturatedaqueousNaHCO 3 solution,brine,andfinallydriedoverNa 2 SO 4 .Thesolventwasthenremovedinvac-uo and the residue was purified by column chromatogra-phy(CHCl 3 /MeOH9:1to8:2)toafford 2 (43 mg,65%)as 5704  S. A. Caldarelli et al. / Bioorg. Med. Chem. 13 (2005) 5700–5709
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