Synthesisandbindingstudiesofnoveldi-substitutedphenanthrolinecompoundswithgenomicpromoterandhumantelomericDNAG-quadruplexes.doc

精品论文synthesis and binding studies of novel di-substituted phenanthroline compounds with genomic promoter and human telomeric dna g-quadruplexes5chunying wei, yanbo wang, meiying zhang(key laboratory of chemical biology and molecular engineering of ministry of education,institute of molecular science, shanxi university,taiyuan, 030006, china)abstract: six novel di-substituted phenanthroline derivatives 5a-7a and 3b-5b have been prepared, and their binding interactions with human telomeric (h-telo) and the promoter c-kit2 and c-myc10g-quadruplex dnas were investigated. all the compounds are potent stabilisers of the g-quadruplex structures and the compounds 3b, 4b, and 5b exhibit a high g-quadruplex selectivity over duplex. thebinding affinities of these compounds to g-quadruplex are higher than to duplex dna. cd spectrashow that the compound can induce the formation of anti-parallel structure of h-telo g-quadruplex. each h-telo quadruplex binds two compound molecules by the end-stacking mode. six new compounds15are able to inhibit significantly the telomerase activity at lowm concentration.key words: physiology; binding study; human telomeric; promoter; g-quadruplex; phenanthroline derivatives0introductionguanine-rich dna sequences, which are particularly found in important regions of telomere,20oncogene promoters, ribosomal dna (rdna), and mrna, can fold into g-quadruplex structure.1 the formation of g-quadruplex at the telomeric end can prevent telomere elongation by telomerase that is activated in 80-85% cancer cells, thus leading to inhibition of telomerase activity.2 in addition, g-quadruplex formation may dissociate telomeric dna from its physical association with telomerase and other telomere-binding proteins, which in cells then results in the25triggering of a dna damage response and eventual cell damage.3 g-quadruplexes formed in thepromoter regions of human oncogene are related to the regulation of gene transcriptional level and biological dysfunctions that selectively alter the integrity of cancer cells.4 therefore, guanine-rich sequences have become a very promising target for the development of new antitumoral drugs and attracted a lot of research interest during the last decade.530phenanthroline derivatives have been originally used as duplex dna-intercalating agents.recently, phenanthroline derivatives and its metal complexes have been modified as telomeric g-quadruplex-stabilizing ligands,6 some have exhibited high telomerase inhibitory activity. however, to the best of our knowledge, phenanthroline derivatives targeting the promoter g-quadruplex have been largely unexplored. in order to avoid acute toxic and intolerable side35effects on normal tissues, it is necessary to discriminate between g-quadruplexes by small molecules, apart from controlling the g-quadruplex selectivity over duplex dna.with this aim in mind and based on the previous studies, we designed and synthesized a newclass of phenanthroline derivatives 5a-7a and 3b-5b (scheme 1) that recognize g-quadruplexdna, because the extended aromatic structure of a 1,10-phenanthroline scaffold provides a basis40for stacking interactions with terminal g-tetrads. it have been reported that the length of amine side chains influences the stabilizing potential of compoundsfoundations: the national natural science foundation of china (21041008, 21171108); specialized researchfund for the doctoral program of higher education (20101401110009).brief author introduction:魏春英(1970-),女,博士,博士生导师,教授,设计和合成靶向 g-四螺旋的抗癌药物研究 及生物活性与作用机理研究. e-mail: - 15 -to g-quadruplex,7 herein we synthesized two series of compounds 3a-5a and 3b-5b. due to poor aqueous solubility the compounds 3a and 4a were further methylated to obtain 6a and 7a, respectively. next, we studied the interactions of 5a-7a and 3b-5b with human telomere (h-telo)45and promoter g-quadruplex (c-kit2 and c-myc) dnas by circular dicroism (cd) spectroscopy, uv-vis absorption spectroscopy, fluorescent resonance energy transfer (fret) melting analysis, fluorescent intercalator displacement (fid), polymerase chain reaction (pcr) stop assay, and telomerase repeat amplification protocol (trap) assay. the results show that these compounds can stabilize the structures of these distinct g-quadruplexes, and they are also the potential50telomerase inhibitors.1results and discussion1.1 synthesisthegeneralsyntheticroutetotargetcompoundsisshowninscheme1.1,10-phenanthroline-2,9-dicarbaldehyde and 1,10-phenanthroline-2,9-dicarboxylic acid were55prepared using previously reported procedures.6f the compounds 3a-5a or 3b-5b were synthesized by substitution of either 2a or 2b with piperidine, pyrrolidine, and piperazine, respectively. the nitrogen atoms of piperidine and pyrrolidine rings for 3a and 4a were further methylated by treatment with ch3i because of their poor aqueous solubility to obtain 6a and 7a, respectively. the detailed description of synthesis and characterization of compounds is given in60the experimental.seo2nno hc 60% hno3nnc hon h oocncooh652,6- dia minop yridineo edc i a nd hoatnn o nhhnc lc h2coc lc l(c h2)2co clnnoo nhh nnnnnnh 2h2nnhhn12a n=1 oo702b n=2(c h2)nc l(ch 2) nc lr 2n h75nnoo nhhnnnc h3i,3a,4 annoo nhhnnnnhh nnhh noooo(c h2)nnr2(c h2)nnr2+c h3r2nc h 2+ch 2nr2ch 33a n=1 nr 2= piperidinyl-804a n=1 nr 2= pyrrolidinyl-5a n=1 nr 2= piperazinyl-3b n=2 nr 2= piperidinyl-4b n=2 nr 2= pyrr olidinyl-5b n=2 nr 2= piperazinyl-6a n r2=pip eridinyl-7a n r2=pyrro lidinyl-scheme 1 synthesis of target compounds1.2 fret melting assaythe g-quadruplex-stabilizing properties of target compounds were evaluated by measuring85the compound-induced change in melting temperature (tm) of the labeled fh-telot, fc-kit2t, and fc-myct g-quadruplexes in fret melting experiments. the tm values of three g-quadruplexes in the presence of 1 m compounds are summarized in table 1. at this concentration, all the compounds are able to stabilize the g-quadruplex structures, but with a different extent. in buffer containing kcl, for the one-methyene linker series the compound 7a9095100105110115120exhibits the highest stabilization potential for h-telo g-quadruplex with tm of 20.4 oc, while 6a and 7a give the similar tm values of about 14 oc for c-kit2 g-quadruplex, and 5a and 6a have similar stabilization potential on c-myc g-quadruplex with tm values of 11.2 and 10.4 oc, respectively. in the case of the two-methyene linker series, the compounds 3b and 4b have similar stabilization effect on h-telo g-quadruplex with tm values of about 21.7 and 21.4 oc, respectively, while the compound 5b exhibits slight low tm value of 16.8 oc. however, 5a, 6a, and 7a stabilize c-myc g-quadrupelx with the same extent with tm value of about 14.0 oc, a similar effect is also observed for c-kit2. for h-telo g-quadrupelx, with the exception of 3b and 4b,the tm values in buffer containing nacl are lower than those in buffer containing kcl. because each dna target exhibits a different tm value, the selectivity between two different g-quadruplex targets cannot be directly compared by tm values. in the piperidino terminal substitute series, 3b with two-methyene linker shows the improved h-telo and c-myc stabilization potential over 6a with one-methyene linker, similarly, 4b exhibits higher c-myc stabilization potential compared to7a, both of which possess two pyrrolidino terminal substitutes. however, the compounds 5a and5b possessing piperazino terminal substitues exhibit comparable stabilization temperatures for all the g-quadruplex targets. overallthe stabilization potential of compounds with slight modifications of side chain length and terminal amino alkyl substitutes exhibits a dna target-dependent difference.furthermore, we measured the tm values of target g-quadruplexes in the presence of excess of competitor duplex (ds26) at 1 m compound concentration. as shown in table 1, the conpounds 3b, 4b, and 5b exhibit a high level stabilization on the g-quadruplexes, the tm values was only a little affected at the level of 50-fold excess of ds26 (10.0 m) for a particular g-quadruplex target. in the case of the compound series with one-methyene linker, in the presenceof 50-fold excess of ds26 6a gives a high selectivity for h-telo target, whereas 5a and 7a show a moderate selectivity for h-telo g-quadruplex over duplex with tm value decreases of 5.6 and 4.2 oc, respectively. in buffer containing 100 mm nacl, all the compounds show a selectivestabilization potential for h-telo g-quadruplex over duplex at the level of 10-fold excess of ds26 (2 m), and 6a and 7a can stabilize h-telo g-quadruplex even at the level of 25-fold excess of ds26 (5 m). for c-kit and c-myc targets, 5a, 6a, and 7a show a poor g-quadruplex selectivetycompared to 5b, 3b, and 4b, particularly for 6a and 7a. the competitive fret melting assayshows that the increase of side chain length can improve the selectivity of g-quadruplex over duplex. accordingly, the combined results of these assays demonstrate that the compounds 3b, 4b, and 5b are a new class of potent and selective stabilizers of h-telo, c-kit2, and c-myc g-quadruplexes.table 1 stabilization temperatures (tm) determined by fret melting experiment.tm at 1.0 m compound concentration(oc)compounds fh-telota fh-telotb fc-kit2tb fc-myctcg4d +ds26e +ds26f g4d +ds26e +ds26f +ds26gg4d +ds26e +ds26f +ds26g g4d +ds26e +ds26f +ds26g5a 13.0 10.7 9.3 17.9 16.4 15.4 12.3 12.3 10.4 9.3 7.1 11.2 10.5 9.5 7.86a 10.4 10.1 10.2 15.0 15.0 15.3 15.0 13.7 9.5 - h- 10.4 7.0 - -7a 15.7 14.3 14.3 20.4 19.5 19.5 16.2 13.8 9.7 - - 8.4 6.0 - -3b 21.4 20.5 17.1 21.7 21.1 21.1 21.0 13.5 13.3 13.3 12.4 14.7 14.7 14.2 13.64b 21.1 20.2 16.6 21.4 21.3 19.0 18.3 13.9 12.9 12.4 11.7 14.0 13.4 13.0 11.8 5b 14.2 12.9 9.6 16.8 16.1 15.8 15.1 11.9 11.9 11.7 10.5 13.5 12.4 11.0 11.0 125a100 mm nacl buffer, b100 mm kcl buffer, c40 mm kcl buffer. d g-quadruplex alone at 0.2 m, e with ds26 at 2.0 m, f with ds26 at 5.0 m, g with ds26 at 10.0 m, h no determination. tm values of annealed fh-telot are45.3 and 62.5 c in buffer containing either 100 mm nacl or 100 mm kcl. tm values of annealed c-kit2 andc-myc are 74.3 and 72.0 c in buffer containing 100 and 40 mm kcl, respectively.1301351401451501551601651.3 pcr stop assaythe stabiliz ation potential of the compounds for g-quadruplexes was further evaluated by pcr stop assay, and 6a and 3b were chosen due to their large tm values and good selectivity over duplex.fig. 2 shows that the intensity of bands decreases with increasing concentration of compounds from 0.5 to 5.0 m, suggesting that the template sequences form g-quadruplexes thatcannot hybrid with primer, which results in the decrease of pcr product. for h-telo and c-mycsequences, 3b exhibits stronger inhibitory effect than 6a, and the pcr products are undetectable at3.0 m concentration, whereas both compounds give a comparable inhibitory ability for c-kit2 sequence. these results are consistent with those from fret melting experiment. in contrast, the mutate template sequences (mut h-telo, mut c-kit2, and mut c-myc) that fail to form g-quadruplexstructures do not exhibit a concentration-dependant decrease of pcr product.fig. 1 effect of the compounds 6a and 3b on the hybridization of h-telo, c-myc, and c-kit2 templates by pcr stop analysis.1.4 fluorescence intercalator displacementthe dna binding affinities of compounds were investigated by fid assay. displacement of thiazole orange (to), a known g-quadruplex ligand, from its binding site by another ligand can be employed to assess the relative affinity of the latter toward the g-quadruplex dna.7 the relativebinding affinity of a given ligand is commonly given in form of the g4dc50 value, which indicatesthe concentration of the ligand that is required to displace 50% of to molecules from the binding site of g-quadruplex. as shown in fig. 2 (all see fig. s1, esi) and table 2, the compounds interact strongly with g-quadruplex dnas with micromolar dc50 values for h-telo and c-kit2 and submicromolar ones for c-myc. the low c-mycdc50 values can be tentatively deduced that these compounds bind with a relatively high affinity in the same binding site of to or at least in closevicinity. interestingly, 5a appears to be less to potent displacer than the other compounds, whereas it exhibits a comparable thermal stabilization ability to 5b and 6a. the dc50 values of compounds 6a and 7a are higher than those of 3b and 4b, respectively, suggesting the relative binding affinity of the former is larger than that of latter. these low micromolar dc50 values arerange of good g-quadruplex dna binders.7 8to displacement also allows one to examine whether a ligand binds selectively to g-quadruplex dna compared to duplex dna. for this purpose the dsdc50 values of compounds are determined by the displacement of to from the duplex dna, and the ratio dsdc50/g4dc50 is used to characterize the selectivity of the compound towards duplex and g-quadruplex, that is, a large value of dsdc50/g4dc50 denotes high selectivity. the results show that these compounds exhibit dsdc50/c-mycdc50 values from 10.14 to 13.60, which indicates a high selectivity of c-myc g-quadruplex versus duplex, while they exhibit a moderate selectivity to both h-telo and c-kit2170175180185190g-quadruplexes over duplex with dsdc50/g4dc50 values from 1.60 to 2.54 and 3.05 to 6.86, respectively. for both h-telo and c-kit2 targets, 3b exhibits the largest dsdc50/g4dc50 value suggesting a good selectivity, while the dsdc50/g4dc50 value of 5a is the smallest that corresponds to the lowest selectivity. these results are full consistent with those from the competitive fretmelting experiment. however, in the case of c-myc target, it appears no good correlation withfret melting results.fig. 2 the percentage of to displacement from h-telo g-quadruplex against increasing concentrations of compounds from 0.125 to 26.25 m.table 2 h-telodc50, c-mycdc50, c-kit2dc50, and ds26dc50 values and selectivity determined by fid assay.to displacement (m)selectivitycompounds h-telodc50 c-mycdc50 c-kit2dc50 dsdc50dsdc50/h-telodc50 dsdc50/c-mycdc50 dsdc50/c-kit2dc505a 11.09 1.34 5.82 17.78 1.60 13.27 3.056a 3.31 0.73 1.37 8.05 2.43 11.03 5.887a 2.95 0.54 1.59 5.91 2.00 10.95 3.723b 2.00 0.45 0.82 5.07 2.54 11.27 6.184b 2.24 0.50 1.06 5.07 2.26 10.14 4.785b 3.31 0.52 1.30 7.34 2.22 13.60 5.65avalues are average of three independent measurements.1.5 uv-vis absorption titrationto obtain the binding constants for the interactions between compounds and dnas, uv-vis absorption titrations were carried out. uv-vis absorption spectra of all the compounds shows similar patterns of behaviour, with a strong absorption band about 285 nm that are assigned to the -* transition of phenanthroline.9 addition of increasing amount of h-telo, c-kit2, or c-mycg-quadruplexes to the corresponding compound results in large hypochromicities (36-73%) andnoticeable red shifts ranging from 9 to 16 nm, these spectral features are indicative of an end-stacking binding rather than groove binding. for comparison, we also investigated the interactions of compounds with ctdna. upon addition of increasing amounts of calf thymus dna (ctdna) to 5a-7a, the hypochromicities between 40 and 51% and red shifts from 4 to 7 nm are also observed, whereas the hypochromicities only between 31 and 45% are observed for 3b-5b,suggesting that these compounds are not good duplex dna intercalators.10 fig. 3 gives therepresentative examples of uv-vis absorption titrations of 6a with either g-quadruplexes or ctdna.the intrinsic binding constants (kb) to g-quadruplexes and ctdna were calculated by equation (1) (see experimental),11 and the values were summarized in table 3. the kb values of1952002052102152203b-, 4b- and 5b-dna complexes are larger than those of 6a-, 7a- and 5a- dna ones, respectively, whether for g-quadruplexes or for ctdna, and the binding affinities to g-quadruplexes are higher than to ctdna for all the compounds. these results are in good agreement with those of fid assay.fig. 3 uv-vis absorption titration spectra of 6a (20 m) with increasing amounts of h-telo (a), c-myc (b), c-kit2 (c), and ctdna (d) in pbs