TY - JOUR
AU - Dupret, Daniel
AB - Abstract Sequencing by the recently reported hybridization technique requires the formation of DNA duplexes with similar stabilities. In this paper we describe a new strategy to obtain DNA duplexes with a thermal stability independent of their AT/GC ratio content. Melting data were acquired on 35 natural and 27 modified duplexes of a given length and of varying base compositions. Duplexes built with AT and/or G4EtC base pairs exhibit a thermal stability restrained to a lower range of temperature than that of the corresponding natural compounds (16 instead of 51°C). The 16°C difference in thermal stability observed between the least stable and the most stable duplex built with AT and/or G4EtC base pairs is mainly due to the sequence effect and not to their AT/G4EtC ratio content. Thus N-4-ethyl-2′-deoxycytidine (d4EtC) hybridizes specifically with natural deoxyguanosine leading to a G4EtC base pair whose stability is very close to that of the natural AT base pair. Oligonucleotide probes involving d4EtC can be easily prepared by chemical synthesis with phosphoramidite chemistry. Modified DNA targets were successfully amplified by random priming or PCR techniques using d4EtCTP, dATP, dGTP and dTTP in the presence of DNA polymerase. This new system might be very useful for DNA sequencing by hybridization. Introduction Nucleic acid analysis is essential for the study of structures and the understanding of many aspects of genes functions (1,2) as well as for diagnostics and prognostics in medicine (3,4 and references cited therein, 5). It includes a variety of techniques such as chemical and enzymatic sequencing (6,7) and hybridization (8,9). The first approach requiring electrophoresis was introduced 20 years ago and has been employed routinely ever since. These techniques are slow and expensive. The second approach, particularly the recently reported reverse hybridization, is a powerful tool for DNA analysis and DNA sequencing (10–17). This approach is based on the detection of perfect hybrids between one or several DNA labelled fragments with a large panel of known oligonucleotides (up to several thousands) arrayed on a solid support. A computational approach can then be used to reconstruct the complete sequence. These large arrays of oligonucleotides allow simultaneous and fast analysis of several DNA fragments. In spite of the great potential of this approach, the use of natural oligonucleotides serving as probes presents a severe drawback. In fact, this method requires absolute discrimination of perfect hybrids from the ones containing mismatches. It is well known that there is a significant variation in duplex stabilities according to their base composition. Thus GC pairing with three hydrogen bonds is more stable than AT or AU pairing which has only two hydrogen bonds. Therefore a perfect AT-rich DNA hybrid could have a similar or even lower stability than do GC containing DNA hybrids involving one mismatch. This leads to false positive or false negative signals depending on the hybridization temperature and washing conditions. Thus for the application of DNA sequencing by hybridization, it is imperative to smooth DNA duplex stabilities in order to make them independent of their base content. Although several groups have been working on the problem of abolishing this stability difference, none have as yet succeeded in resolving it completely (18–28). To achieve this objective, we recently designed and selected the N-4-ethyl-2′-deoxycytidine (dC4Et) which forms with a natural 2′-deoxyguanosine dG a G4EtC base pair having a stability very close to that of the natural AT base pair (29). We also synthesized a few modified DNA duplexes involving AT and G4EtC pairing. Modified duplexes built with AT and/or G4EtC base pairs showed a smoothed thermal stability when compared to their natural analogs. With the purpose of verifying the general properties of these modified duplexes and estimating the influence of the sequence on their thermal stability, we describe in this paper synthesis and thermal stability studies of 35 duplexes of a given length and of varying base compositions and sequences. To assess the influence of the ethyl group at position 4 of the cytosine on the physicochemical properties of modified duplexes, hybridization studies were also carried out with natural duplexes having the same sequence as those that were modified. This system involved the replacement of dC with d4EtC in both strands of the duplexes. In order to prepare modified DNA fragments, d4EtCTP was synthesized and the incorporation of d4EtCMP with DNA polymerase was studied. Moreover, we have verified that d4EtC is not mutagenic when incorporated in a DNA sequence. Incorporation of nucleotides opposite to d4EtC using Klenow and Taq polymerase has been checked. We show that both Klenow and Taq polymerase specifically add a dG opposite to d4EtC. Materials and Methods [α-32P]dATP, [γ-32P]ATP, [α-32P]dCTP, [α-32P]dGTP, [α-32P]dTTP were obtained from NEN. The following products are from Appligène-Oncor: Klenow polymerase, T4 polynucleotide Kinase, Taq polymerase, nonaprimer kit, lambda DNA, dATP, dCTP, dGTP, dTTP. Absorption spectra were recorded on a UVIKON 860 (KONTRON). Absorption studies were carried out on a UVIKON 941 cell changer spectrophotometer (KONTRON). Analysis by reversed-phase chromatography was performed on a Waters 626 E (system controller) equipped with a Waters 996 photodiode array detector. Purifications by reversed-phase chromatography were performed with Kontron HPLC pumps (model 422) connected to a Kontron diode array detector (model 440) on 4 × 250 mm RP18 encapped (5 µm) lichrospher 100 lichrocart column (Merck). Analysis and purification by ion exchange chromatography were performed with FPLC apparatus (Pharmacia). Gel electrophoresis was performed in 3 mm thick 2% agarose gel containing running buffer TAE (0.04 M Tris acetate, 0.001 M EDTA) or in 1.5 mm thick 20% polyacrylamide gel, 7 M urea. Agarose gel electrophoresis was conducted at 20°C and visualized with ethidium bromide and UV irradiation. Autoradiograms were obtained with a XOMAT Kodak film. PCR amplifications were carried out in a DNA thermocycler Crocodile III from Appligène-Oncor. Synthesis, analysis and purification of oligonucleotides 1–35 and 8*–35* The chain assembly was carried out on a Expedite synthetizer on a CPG (Controlled Pore Glass) solid support functionalized with a nucleoside using phosphoramidite chemistry (30). Oligonucleotide syntheses were performed, as described in a preceding paper, on a 1 µmol scale using 10 µmol of commercial phosphoramidite or modified phosphoramidite (29), per cycle with a cycle time of 3.5 min and a coupling time of 1.6 min for the phosphoramidite. The oligonucleotides obtained were then deblocked by overnight treatment with concentrated ammonia at 60°C. After deprotection and extraction of the organic impurities, the crude tritylated reaction products were purified by reversed-phase chromatography using a gradient of CH3CN:0% of CH3CN for 1 min and from 0 to 80% of CH3CN over 20 min, in 0.1 M aqueous triethylammonium acetate buffer pH 7, with a flow rate of 1 ml/min. The fractions were monitored by absorption at 254 nm. After detritylation, the purity of all oligonucleotides described here was checked by capillary electrophoresis with a 270 A-HT capillary electrophoresis system (Perkin Elmer) on microgel 100 capillaries. Runs were performed at −15 kV for 15 min at 50°C. Oligonucleotides were visualized by absorption at 260 nm. Synthesis of N-4-ethyl-2′-deoxycytidine triphosphate Synthesis of N-4-ethyl-2′-deoxycytidine triphosphate were performed according to two protocols leading the same product. In the transamination method, methylamine was replaced with ethylamine (31). Using synthesis according to Eckstein's method 2′-deoxycytidine was replaced with N-4-ethyl-2′-deoxycytidine (32). HPLC analysis of the obtained product was performed on a reversed-phase chromatography using a Lichrocart system (125 × 4 mm) packed with 5 µm Lichrospher RP 18 from Merck with a linear gradient of acetonitrile, 0% over 10 min and 0–16% over 20 min in 0.1 M aqueous triethylammonium acetate buffer pH 7, with a flow rate of 1 ml/min: Rt (d4EtCTP) = 22.8 min. d4EtCTP was ascertained by nuclease degradation leading d4EtC. An aliquot of oligonucleotide was digested with snake venom phosphodiesterase (Pharmacia Biotech) and alkaline phosphatase (Boehringer) in 0.1 M Tris-HCl (pH 8.2) buffer for 2 h at 37°C. After inactivation of the enzyme at 90°C for 2 min, the digestion product was analysed by reversed phase chromatography using linear gradient of acetonitrile: 0% over 15 min and then 0–20% acetonitrile over 20 min in 0.1 M aqueous triethylammonium acetate buffer, pH 7. Detection was performed at 260 nm. d4EtCTP was degraded to nucleoside having the same retention time than d4EtC by comparison with d4EtC authentic sample (Rtd4EtC = 10.9 min). Incorporation of d4EtCMP by Klenow polymerase Efficiency of incorporation of d4EtCMP by Klenow polymerase. λDNA (25 ng) was denatured by heating for 10 min at 100°C and chilled on ice for 5 min. Nonaprimer Mix (1×), 50 µM of each triphosphate (dATP, dGTP and dTTP), 0.83 µM of dCTP, 0.0555 µM of [α-32P]dCTP and 2.5 U of Klenow polymerase were added. Then different isotopic dilutions of this mix (N) were made with either dCTP or d4EtCTP, incubated for 30 min at 37°C. It consists in a dilution of the N = 0.88 µM [α-32P]CTP mixture (0.83 µM of dCTP and 0.0555 µM of [α-32P]dCTP) by increasing quantities (N × 2n) µM with n = 0–5 of either dCTP or d4EtCTP. The percentage of labelled fragments was determined by means of DE-81 adsorption (33). Specificity of incorporation of d4EtCMP by Klenow polymerase. [α-32P]dATP, [α-32P]dCTP, [α-32P]dGTP or [α-32P]dTTP (0.0555 µM) and 50 µM of the three other triphosphates were used. Different concentrations (50, 100, 150 or 200 µM) of d4EtCTP were added to the mix. The protocol was performed as described above using the nonaprimer labelling kit ‘random priming’. Incorporation of d4EtCMP by Taq polymerase The protocol for evaluating the efficiency of incorporating d4EtCMP by Taq polymerase was carried out as described above using the random priming reaction by replacing Klenow polymerase by Taq polymerase. Determination of nucleotides incorporated opposite d4EtCTP Primers (10 pmol) were labelled at the 5′-end by T4 polynucleotide kinase (5 U) for 1 h at 37°C, in the presence of [γ-32P]ATP (0.37 MBq) in a total volume of 10 µl. Reactions catalysed by Klenow were carried out for 1 h at 37°C, in a buffer containing 40 mM potassium phosphate pH 7.5, 6.6 mM MgCl2, 1 mM 2-mercaptoethanol, templates annealed with radiolabelled primers (0.05 µM), 10 µM of a single dNTP and enzyme (0.01 U) in a total volume of 10 µl. Experiments with Taq polymerase were conducted at 72°C for 1 h, in a total volume of 10 µl of a buffered reaction mixture containing 10 mM Tris-HCl pH 9, 50 mM KCl, 1.5 mM MgCl2, 0.1% Triton X-100, 0.2 mg/ml BSA, oligonucleotide templates annealed with radiolabelled primers (0.05 µM), 10 µM of a single dNTP and enzyme (0.01 U) in a total volume of 10 µl. Reactions were stopped by addition of 5 µl blue stop solution (95% formamide, 0.1% bromophenol blue, 0.1% xylene cyanol). Samples were heated at 95°C for 3 min and then 3 µl were applied on a 20% polyacrylamide, 7 M urea gel. Electrophoresis was carried out for 4 h at 800 V and then the gel was autoradiographed. PCR amplification involving d4EtCTP PCR amplifications were carried out with 50 µl solutions containing 200 µM each of triphosphate dATP, dGTP, dTTP and 200 µM of dCTP or d4EtCTP, 50 pmol of each primer, 62.5 ng genomic DNA, 10 mM Tris-HCl pH 9, 50 mM KCl, 1.5 mM MgCl2, 0.2 mg/ml BSA, 0.1% Triton X-100 and 0.5 U of Taq polymerase. The solutions were subjected to 31 cycles of either a high temperature regimen [(5 min at 93°C, 1 min at 55°C, 1 min at 70°C) ° 1], [(1 min at 91°C, 1 min at 55°C, 1 min at 70°C) × 30], 10 min at 70°C ], or a low temperature regimen [(5 min at 93°C, 1 min at 55°C, 1 min at 70°C) ° 1], [(1 min at 91°C, 1 min at 55°C, 1 min at 70°C) ° 3], [(1 min at 85°C, 1 min at 48°C, 1 min at 70°C) ° 27], 10 min at 70°C] in a DNA thermocycler Crocodile III from Appligène-Oncor. Melting experiments Changes in absorbance with the temperature of 2 µM duplexes in 10−2 M sodium cacodylate pH 7 buffer containing 1 M NaCl and 2 ° 10−4 M EDTA were measured at λ = 260 nm on a UVIKON 941 cell changer spectrophotometer equipped with a Huber PD 415 temperature programmer connected with a cryothermostat ministat water circulating bath (Huber). The samples and reference were slowly heated at a rate of 0.5°C/min from 0 to 80°C. Melting temperatures (Tm) were taken as the temperature corresponding to the half-dissociation of the complexes. The Tm values were determined using the first and second derivatives. The margin of error on the Tm values is ∼1°C. The molar extinction coefficient of the sequences was determined as described in the literature (34). Circular dichroism (CD) spectroscopy Oligonucleotide and triphosphate solutions were studied in a buffer containing 10−2 M sodium cacodylate pH 7, 1 M NaCl and 2 ° 10−4 M EDTA. The concentration used to calculate the CD amplitude was that of the nucleotide unit. CD measurements were carried out on a Jobin-Yvon Mark IV dichrograph. Data acquisition and analysis were performed on a microcomputer interfaced to the spectrometer. Optical cells with a pathlength of 1.0 cm were used. The temperature of the cell was adjusted with a refrigerated circulating water bath and held constant to ±0.5°C. Each CD spectrum was run at least twice and checked for possible base line shifts. Results and Discussion To obtain DNA duplexes with thermal stabilities independent of their base content in classical buffer solution, we chose to investigate the melting behaviour of a set of duplexes built with either AT, AT and G4EtC, or only with G4EtC base pairs. This study also allowed us to verify that duplexes of a given length built with AT and/or G4EtC base pairs exhibit a thermal stability independent of the base content and to know the influence of the sequence on their thermal stability. Studies were carried out on 35 natural duplexes and 28 modified duplexes involving d4EtC synthesized as described in a previous paper (29) (Table 1). These duplexes were built with 9 bp and formed by a triplet of bases repeated three times. This allowed us to easily obtain measurable Tm variations as a function of the sequence and base content. Table 1 gives the Tm values of seven duplexes made with nine AT base pairs (duplexes 1–7), 12 duplexes built with six AT and three GC base pairs (duplexes 8–19) or six AT and three G4EtC base pairs (duplexes 8*–19*), 12 duplexes composed of three AT and six GC base pairs (duplexes 20–31) or three AT and six G4EtC base pairs (duplexes 20*–31*) and four duplexes composed of nine GC base pairs (duplexes 32–35) or nine G4EtC base pairs (duplexes 32*–35*). To avoid possible concatenation of the duplexes to form long polymers, two dangling TT arms were added at the 3′- and 5′-extremities of one strand of the duplex (35–36). The thermal stability of hybrids built with AT and/or G4EtC base pairs is independent of their base content Half-dissociation temperatures (Tm) of sequences built with nine AT base pairs are generally very low. Thus Tm of duplexes were determined in a high concentration of NaCl (1 M) (Table 1). Duplexes 5′-d-[T2(N1N2N3)3T2]-3′/3′-d-(N1′N2′N3′)3-5′ are symbolized by the Watson strand T2(N1N2N3)3T2 for natural duplexes and T2(N1N2N3)*3T2 for the corresponding modified duplexes containing d4EtC instead of natural dC. The results obtained showed that the thermal stabilities of modified duplexes of a given length do not depend on their base content. In fact, modified duplexes 8*–19*, 20*–31* and 32*–34* built with six AT and three G4EtC base pairs, three AT and six G4EtC base pairs and nine G4EtC base pairs, respectively, have very similar Tm ranging from 15.5 to 31.5°C. The melting range of 16°C was determined between duplex 25* (Tm = 31.5°C), thermally the most stable and the modified duplex 27* (Tm = 15.5°C), thermally the least stable. These two duplexes belong to the same set of modified duplexes built with three AT and six G4EtC base pairs. They do not belong to different sets of oligonucleotides having different base compositions. We can also note that modified duplexes 8*–34* built with AT and G4EtC base pairs or nine G4EtC base pairs have Tm very close to those of natural duplexes 1–7 made with nine AT base pairs (Tm ranging from 15.5 to 30.5°). Consequently these results show that all natural duplexes (1–7) and modified duplexes (8*–34*) have thermal stabilities independent of their base content (Fig. 1). On the other hand, we observed that thermal stabilities of modified duplexes change with the sequence of base pairs as natural sequences do. This problem will be discussed further on. Table 1 View largeDownload slide Melting temperatures of natural and modified duplexes involving C° = C or 4EtC Table 1 View largeDownload slide Melting temperatures of natural and modified duplexes involving C° = C or 4EtC The usual base composition dependent differences in transition temperatures are observed for the various natural oligonucleotides. Thus as expected the thermal stability of natural duplexes in NaCl solution increases regularly with the GC base pair number. The difference value in the stability between the more stable duplex [duplex 34 built with nine GC base pairs of sequence T2(GCG)3T2, Tm = 66.5°C] and the duplex having only AT base pairs [duplex 3 of sequence T2(TAA)3T2Tm = 15.5°C] is ∼51°C. This difference in the Tm value is mainly due to the effect of the base composition of natural duplexes and is clearly higher than the 16°C Tm difference measured with the modified duplexes built with AT and/or G4EtC. Figure 1 View largeDownload slide Tm of natural (Δ) and modified (▪) duplexes involving d4EtC according to the base content. Figure 1 View largeDownload slide Tm of natural (Δ) and modified (▪) duplexes involving d4EtC according to the base content. Sequence effect on the formation and stability of duplexes It is well known that for a given base composition, the thermal stability of natural DNA varies with the nucleic base sequence. This phenomenon is observed as well for modified duplexes (Table 1). Inside each of the two duplex families, for the same base composition, we found almost the same triplets of natural or modified base pairs which can be classified as the most or least thermally stable. (i) For duplexes built with six AT and three GC or G4EtC base pairs, (CTT), (TTC) and (CTT)*, (TTC)* triplets are among the least stable, whereas (GTT), (TGT), (ATC) and (GTT)*, (TGT)*, (ATC)* are the most stable. (ii) For duplexes containing three AT and six GC or G4EtC base pairs, (CCT), (CTC), (TCC) and (CCT)*, (CTC)*, (TCC)* are among the least stable while (TGC) and (TGC)* are the most stable. (iii) For nine GC or G4EtC base pairs, (CGG) and (CGG)* triplets are the least stable. Very similar results were previously reported with a series of octamer DNA duplexes involving various set of triplets at the central position (37). Inside modified duplexes composed of the four nucleic bases, the triplets homopyrimidic-homopuric PyPyPy/PuPuPu are generally thermally less stable than the triplets of base pairs having a pyrimidic ↔ puric transition (Table 2). For duplexes possessing only AT base pairs, the homopyrimidic-homopuric sequences (duplex 1) and homopuric-homopyrimidic sequences (duplex 2) are the most stable. The previously described high stability of duplexes d(A)n-d(T)n would be due to their particular structures (38,39). In a previous study carried out on a limited number of sequences, we observed that the thermal stability of the G4EtC pairing is slightly higher than that of the AT pairing in NaCl solution. This should lead to Tm values higher for G4EtC-rich sequences than those built with AT base pairs. The results in Table 1 seem to confirm this tendency. However, the sequence effect on the thermal stability of duplexes seems to be more important than the difference in stability between AT and G4EtC base pairs. We also observed that the replacement of a GC base pair in duplex 5′-d[T2TCGTCXTCGT2]-3′/3′-d[AGCAGYAGC]- 5′ (X = G, Y = C) by a G4EtC base pair (X = G, Y = 4EtC) led to a Tm decrease of ∼4°C. This lower value for one substitution inside a nine base pair sequence is similar to that obtained with the 27 modified duplexes (3.7–5.7°C decrease in stability by modification, except for duplexes 18 and 18*) (Table 1). Results show that although the thermal stability of modified duplexes does not depend on the base content, it varies with the nucleic base sequences as do natural duplexes. This 16°C difference observed for the Tm values of the thermally most stable and least stable duplexes among those studied (duplexes 1–7 and 8*–34*) is exclusively due to the sequence effect. Besides this, we noted that the sequence effect on thermal stability is slightly different for natural and modified duplexes involving the same base content. The cooperativity of the modified duplex dissociation is similar to that of natural DNA duplexes involving natural base pairs except for natural duplex 35 and modified duplex 35* composed of (CCC/ GGG) and (CCC/GGG)* triplets repeated three times, respectively, which led to a monotonous variation of absorbance when solutions were heated from 0 to 80°C (data not shown). We attempted to dissociate a self-associated complex by heating oligonucleotide solutions to 100°C before hybridization experiments. Despite this procedure, we did not observe any cooperative transition. This suggests that these ‘duplexes’ are not really formed by hybridization between the nonadeoxyguanylate and the natural complementary sequence T2(CCC)3T2 or modified T2(CCC)*3T2. In fact, it is well known that oligodeoxy- and poly-deoxyguanylate sequences tend to form self-associated complexes, particularly G tetrads (40) which are very stable in the presence of Na+ or K+. To check this point the CD spectrum of dT2G9T2 was taken. It exhibits the characteristic pattern of the parallel intermolecular quadruplex (41,42) with a large positive ellipticity at ∼260 nm and a smaller negative peak at ∼240 nm. The self-association property of G-rich sequences represents a serious difficulty for characterizing an oligo- or polyguanylate sequence in solution by hybridization with an oligocytidylate sequence in solution or immobilized onto a solid support. This problem could be resolved by using modified DNA sequences having a slight tendency to form self-associated complexes, for example by replacing dG by d7deazaG (43). The thermal dissociation of duplexes built with AT and/or G4EtC base pairs have the same cooperativity as that of natural duplexes Nucleic probes in a reverse hybridization system should ideally have the following properties: (i) they should form with complementary sequences hybrids having a similar half-transition temperature in order to allow the discrimination of perfect hybrids from those with mismatches; and (ii) dissociation of duplexes into single strands should be in an as narrow as possible temperature range in order to obtain clear and homogeneous signals. The examination of the thermal dissociation curves of natural duplexes 1–34 or modified duplexes 8*–34* containing d4EtC in place of dC shows that they have, without exception, a similar cooperativity. (This property has been observed with others d4EtC containing duplexes described in a preceding paper; 29). Moreover, for the same sequence the dissociation of natural and modified duplexes have a similar hyperchromicity which is higher for AT-rich duplexes than for GC or G4EtC-rich duplexes. The same dissociation cooperativity observed with natural and modified duplexes is an important factor in obtaining equivalent and optimal amounts of different perfect duplexes retained on a matrix composed of several oligonucleotide probes under defined experimental conditions. Table 2 View largeDownload slide Tm variations of duplexes containing AT and/or G4EtC base pairs as a function of pyrimidine ↔ purine transitions Table 2 View largeDownload slide Tm variations of duplexes containing AT and/or G4EtC base pairs as a function of pyrimidine ↔ purine transitions Specificity of the G4EtC base pair Previous studies (29) of duplexes formed with sequences 5′-d-(T2T4EtCGT4EtCXT4EtCGT2)-3′/3′-d(AG4EtC AG4EtC AG4EtC)-5′ (X = G, C, A, T) have shown that the specificity of the G4EtC base pair was maintained (X = G, Tm = 24°C). The C4EtC, A4EtC or T4EtC mismatch led to a decrease in the Tm value which was impossible to determine (Tm < 10°C). The results obtained showed that by substituting one hydrogen atom of the amino group at position 4 of dC by the ethyl group, its matching with dG was retained. The conservation of the phosphodiesterdeoxyribonucleoside skeleton seemed to be a major factor in obtaining specific hybridization and good cooperativity of duplex dissociations. To lower the stability of GC-rich duplexes, modifications could be realized on either the sugar or internucleotidic bonds. However, this approach was rejected because it would mean an important change in the conformation of oligomers which would make the interpretation of their hybridization properties very difficult. CD studies The CD spectra of modified oligonucleotide duplexes were recorded and compared to those of their unmodified analogs. A representative set of CD spectra are depicted in Figures 2–4. In Figure 2 the duplexes contain three C or three 4EtC residues on the same strand (duplexes 9, 9*, 13, 13*, 14 and 14*); in Figure 3A and B the duplexes contain six C or six 4EtC which are on both strands (duplexes 20 and 20* and duplexes 24 and 24*), whereas they are all on the same strand in Figure 3C (duplexes 28 and 28*). In Figure 4 the duplexes contain nine C or nine 4EtC shared by both strands (duplexes 33, 33*, 34 and 34*). The modified oligonucleotides exhibit characteristic CD spectra with differences compared to their natural analogs. In some cases a small negative band occurs at the longer wavelengths. The differences are small for compounds bearing three or six C (Figs 2 and 3) and larger for those containing nine C (Fig. 4). The differences between the spectra of the natural and modified duplexes may result not only from the spectroscopic properties of the monomer d4EtCTP (the various transition moments, electric and magnetic, are different in orientation and/or intensity) but also from their geometries as well. The resemblance between the spectra of modified and unmodified compounds leads us to believe that the major differences may originate from a change in the spectroscopic properties and not from a geometrical change, all the compounds being in the B-DNA form. This is further confirmed by the comparison between the CD spectra of the monomer triphosphate, dCTP and d4EtCTP, which is shown in Figure 5. There is an increase of the intensity by ∼50% and a shift of the maximum toward shorter wavelengths of ∼10 nm. Synthesis of N-4-ethyl-2′-deoxycytidine triphosphate and its use as substrate for DNA polymerases Synthesis. To apply this new system, every natural dC must be replaced by d4EtC in both DNA strands to be amplified and analyzed. This implies the synthesis of modified N-4-ethyl-2′-deoxycytidine-5′-triphosphate (d4EtCTP). The synthesis of d4EtCTP was carried out according to two protocols described. The first synthesis was achieved by means of a bisulfite catalyzed transamination reaction in the presence of ethylamine (31). The second one was carried out according to the Eckstein's method by replacing 2′-deoxycytidine with N-4-ethyl- 2′-deoxycytidine (32). Incorporation studies of d4EtCMP Incorporation studies of d4EtCMP by Klenow polymerase. In order to compare the efficiency of incorporating the modified nucleotide with that of the natural one, experiments were carried polymerase. Competition in variable ratios between either [α-32P]dCTP and d4EtCTP or [α-32P]dCTP and natural dCTP was set up to proceed to isotopic dilution (Materials and Methods). Thus efficiency of incorporation of d4EtCMP into the newly-synthesized strand depends on the amount of radioactivity incorporated. Figure 6A shows that dCMP and d4EtCMP have the same incorporation profiles. Thus incorporation of d4EtCMP is feasible, but is nine times less efficient compared to the incorporation of the natural dCMP in these conditions. Figure 2 View largeDownload slide CD spectra of natural duplexes (—) and modified duplexes involving d4EtC (---) in a buffer similar to that of Table 1. Nucleotide unit concentration was 10−4 M. (A) duplexes 9 and 9*; (B) duplexes 13 and 13*; (C) duplexes 14 and 14*. Figure 2 View largeDownload slide CD spectra of natural duplexes (—) and modified duplexes involving d4EtC (---) in a buffer similar to that of Table 1. Nucleotide unit concentration was 10−4 M. (A) duplexes 9 and 9*; (B) duplexes 13 and 13*; (C) duplexes 14 and 14*. In a second set of experiments, the specificity of the incorporation of the modified nucleotide was tested by a competition experiment of one of the radiolabelled triphosphates (dATP, dTTP and dGTP) with the d4EtCTP. During the polymerization, whatever the radiolabelled triphosphate used, we did not observe a significantly decrease in radioactivity (Fig. 6B). Although the d4EtCTP concentration increased, incorporation of the other triphosphates remained almost constant. These results showed that d4EtCMP is not incorporated in place of other nucleotides and confirmed that its incorporation is specific. The incorporation of the modified nucleotide has been optimized and the best results were obtained using 10 U of Klenow polymerase and 500 µM d4EtCTP (data not shown). Incorporation studies of d4EtCMP with Taq polymerase Production of modified DNA by PCR amplification implies the incorporation of d4EtCMP by Taq polymerase which was also tested using the same protocol as described above, replacing Klenow polymerase by the latter. Results showed that although the incorporation of d4EtCMP is feasible, it is six times less efficient than that obtained with the natural dCMP under the same conditions (data not shown). Figure 3 View largeDownload slide CD spectra of natural duplexes (—) and modified duplexes involving d4EtC (---)in a buffer similar to that of Table 1. Nucleotide unit concentration was 10−4 M. (A) Duplexes 20 and 20*; (B) duplexes 24 and 24*; (C) duplexes 28 and 28*. Figure 3 View largeDownload slide CD spectra of natural duplexes (—) and modified duplexes involving d4EtC (---)in a buffer similar to that of Table 1. Nucleotide unit concentration was 10−4 M. (A) Duplexes 20 and 20*; (B) duplexes 24 and 24*; (C) duplexes 28 and 28*. Figure 4 View largeDownload slide CD spectra of natural duplexes (—) and modified duplexes involving d4EtC (---) in a buffer similar to that of Table 1. Nucleotide unit concentration was 10−4 M. (A) Duplexes 33 and 33*; (B) duplexes 34 and 34*. Figure 4 View largeDownload slide CD spectra of natural duplexes (—) and modified duplexes involving d4EtC (---) in a buffer similar to that of Table 1. Nucleotide unit concentration was 10−4 M. (A) Duplexes 33 and 33*; (B) duplexes 34 and 34*. Figure 5 View largeDownload slide CD spectra of dCTP (—) and d4EtCTP (---) in a buffer similar to that of Table 1. Nucleotide unit concentration was 10−4 M. Figure 5 View largeDownload slide CD spectra of dCTP (—) and d4EtCTP (---) in a buffer similar to that of Table 1. Nucleotide unit concentration was 10−4 M. To assess the potential for using d4EtCTP in PCR amplification, natural dCTP was replaced by d4EtCTP in the PCR reaction mix. The first experiment was performed using the standard protocol with a denaturation temperature of 91°C and a hybridization temperature of 55°C. Using this standard PCR protocol, the desired fragment of 192 bp was obtained with the canonical dNTPs but not with d4EtCTP (results not shown). Considering that modified GC base pairs show a thermal stability close to that of AT base pairs, the use of d4EtCTP must induce a decrease in the Tm of the hybrid. Therefore we tried to optimize the temperatures of hybridization and denaturation. Amplification of DNA could not be obtained by using only a lower hybridization temperature (48 instead of 55°C), in the PCR program. However, the modification of both the denaturation and hybridization temperatures (85 instead of 91°C and 48 instead of 55°C, respectively) allowed the amplification of the 192 bp fragment. The PCR products obtained with dCTP or d4EtCTP showed identical band intensities when visualized in the agarose gel with ethidium bromide (data not shown). Determination of nucleotides incorporated opposite d4EtC by Klenow and Taq polymerases Klenow and Taq polymerases have been challenged to incorporate a dNTP into a growing oligonucleotide chain opposite d4EtC in a template.The nucleotide incorporation by Klenow polymerase was checked by a 1 h incubation with radiolabelled CE299 primer annealed with the template CE301 containing a d4EtC at the 4th position. As a control, same experiment was performed with radiolabelled CE299 primer annealed with the template CE300 containing a natural dC at the same position (Fig. 7A). Since experiments with Taq polymerase require incubation at 72°C, longer sequences with higher Tm were chosen. Radiolabelled primer CE302 was annealed with template CE304 containing a d4EtC at the 6th position and incubated for 1 h at 72°C. As a control, the same experiment was performed with radiolabelled primer CE302 annealed with template CE303 containing a natural dC at the same position (Fig. 7B). Following incubation, products were resolved by gel electrophoresis on a 20% denaturing gel. Untreated primer-templates were used as migration controls. As shown in Figure 7, both Klenow and Taq polymerases specifically incorporate a dGMP opposite d4EtC. So, this modified nucleotide should not be mutagenic during DNA amplification. Figure 6 View largeDownload slide (A) Efficiency of incorporation of d4EtCMP by Klenow polymerase. Random priming was performed as described in Materials and Methods. Isotopic dilution was realized by an increasing dilution (1 + 2n) with n = 0–5 either with dCTP (□) or d4EtCTP (♦). Competition of the incorporation of the added nucleotide is reported in values of radioactivity. (B) Specificity of incorporation of d4EtCMP by Klenow polymerase. Random priming was performed as described in Materials and Methods, with 0.555 µM of one [α-32P]dNTP tested: [α-32P]dATP (•), [α-32P]dTTP (♦) or [α-32P]dGTP (□). Figure 6 View largeDownload slide (A) Efficiency of incorporation of d4EtCMP by Klenow polymerase. Random priming was performed as described in Materials and Methods. Isotopic dilution was realized by an increasing dilution (1 + 2n) with n = 0–5 either with dCTP (□) or d4EtCTP (♦). Competition of the incorporation of the added nucleotide is reported in values of radioactivity. (B) Specificity of incorporation of d4EtCMP by Klenow polymerase. Random priming was performed as described in Materials and Methods, with 0.555 µM of one [α-32P]dNTP tested: [α-32P]dATP (•), [α-32P]dTTP (♦) or [α-32P]dGTP (□). Conclusion Two factors contribute to the stability of DNA duplexes, hydrogen bonding between the complementary bases and stacking between the base pairs. The stacking effect explains that oligonucleotide duplexes with the same base compositions (that is to say with the same number of hydrogen bonds) exhibit differences in stability. In our system of duplexes made with nine base pairs [AT (or TA) and/or GC (or CG) base pairs], the melting temperature range is 15°C for compounds with nine AT or TA base pairs, 11°C for six AT or TA base pairs, 10°C for three AT or TA base pairs and 7°C for compounds which do not contain any AT or TA base pairs. The stability difference between the most stable and the least stable duplexes is 51°C. With oligonucleotide duplexes bearing AT (or TA) and 4EtCG (or G4EtC) base pairs, this overall variation is reduced to 16°C. An interesting point is that the most stable and the least stable duplexes are obtained in the same class of duplexes with three AT (or TA) and six 4EtCG (or G4EtC) base pairs. Substitution of one hydrogen of the exocyclic amino group at position 4 by the ethyl group gives a modified d4EtC which hybridizes with natural dG leading to a G4EtC base pair whose stability is very similar to that of the natural AT base pair. The specificity of recognition of G is maintained and the cooperativity of dissociation of the modified duplexes into single-strands is similar to that with natural duplexes. Figure 7 View largeDownload slide (A) Incorporation of nucleotides by Klenow polymerase. Template CE300 [3′-d(TACTGCCTTATACTAG)-5′] or CE301 [3′-d(TACTGCCTTATA4EtCTAG)-5′] annealed with primer CE299 [5′-d(ATGACGGAATAT)-3′] were treated with Klenow polymerase in the presence of a single dNTP under the conditions described in Materials and Methods. Lanes 1 and 6: template primer complex; lanes 2 and 7: in the presence of dATP; lanes 3 and 8: in the presence of dGTP, lanes 4 and 9: in the presence of dCTP; lanes 5 and 10: in the presence of dTTP. (B) Incorporation of nucleotides by Taq polymerase. Template CE303 [3′-d(AGTGGAATCCCAACGGGTATTCGATCTGAGT)-5′] or CE304 [3′-d(AGTGGAATCCCAACGGGTATTCGAT4EtCTGAGT)-5′] annealed with primer CE302 [5′-d(TCACCTTAGGGTTGCCCATAAGCTA)- 3′] were treated with Taq polymerase in the presence of a single dNTP under the conditions described in Materials and Methods. Lanes 1 and 6: template primer complex; lanes 2 and 7: in the presence of dATP; lanes 3 and 8: in the presence of dTTP; lanes 4 and 9: in the presence of dGTP; lanes 5 and 10: in the presence of dCTP. similar to that with natural duplexes. Figure 7 View largeDownload slide (A) Incorporation of nucleotides by Klenow polymerase. Template CE300 [3′-d(TACTGCCTTATACTAG)-5′] or CE301 [3′-d(TACTGCCTTATA4EtCTAG)-5′] annealed with primer CE299 [5′-d(ATGACGGAATAT)-3′] were treated with Klenow polymerase in the presence of a single dNTP under the conditions described in Materials and Methods. Lanes 1 and 6: template primer complex; lanes 2 and 7: in the presence of dATP; lanes 3 and 8: in the presence of dGTP, lanes 4 and 9: in the presence of dCTP; lanes 5 and 10: in the presence of dTTP. (B) Incorporation of nucleotides by Taq polymerase. Template CE303 [3′-d(AGTGGAATCCCAACGGGTATTCGATCTGAGT)-5′] or CE304 [3′-d(AGTGGAATCCCAACGGGTATTCGAT4EtCTGAGT)-5′] annealed with primer CE302 [5′-d(TCACCTTAGGGTTGCCCATAAGCTA)- 3′] were treated with Taq polymerase in the presence of a single dNTP under the conditions described in Materials and Methods. Lanes 1 and 6: template primer complex; lanes 2 and 7: in the presence of dATP; lanes 3 and 8: in the presence of dTTP; lanes 4 and 9: in the presence of dGTP; lanes 5 and 10: in the presence of dCTP. similar to that with natural duplexes. The decrease of the Tm value ΔTm observed per 4EtC modification is nearly independent of the sequence and of the number of modifications made in the duplex and of their positions. This leads us to believe that the replacement of C by 4EtC weakens the hydrogen bonding in the 4EtCG (or G4EtC) base pair but has less influence on the stacking energy between adjacent base pairs. The hybridization properties of modified oligonucleotides showed that this substitution leads neither to an important change in the duplex structures nor to a notable steric disturbance. This is further supported by CD studies. This system implies the substitution of dC by d4EtC in both strands of the DNA. Preparation of oligonucleotide probes can easily be carried out by chemical synthesis with phosphoramidite chemistry. Experiments showed that the incorporation of d4EtCMP into DNA fragments by Klenow polymerase and dATP, dGTP, dTTP and d4EtCTP according to the nonaprimer technique is efficient and specific. Moreover, under appropriate temperature Taq polymerase can incorporate d4EtCTP and also read modified DNA templates containing d4EtC to produce modified amplified DNA fragments. Finally, we show that both Klenow and Taq polymerase specifically add a dG opposite to d4EtC. These results suggest that d4EtC and dC would have similar biochemical properties which are predictable. In fact it has been shown that its homologous d4MeC is present in the DNA of certain thermophylic bacteria (44). All hybridization properties of modified DNA allow their use in reverse hybridization implying large numbers of oligonucleotides immobilized on a matrix. Using this system it is possible to discriminate between perfect hybrids built with AT-rich sequences from those involving GC-rich sequences and mismatches. Properties of duplexes containing AT and/or G4EtC base pairs in a classical buffer solution as well as the good incorporation of d4EtCMP by DNA polymerase allow their use in biochemical experiments such as Random priming and LCR used in the preparation and labelling of DNA fragments for diagnoses and prognoses. Acknowledgement This work was supported by a BioMed contract (gene-CT 93-0009). References 1 Landegren U.,  Kaiser R.,  Caskey C.T.,  Hood L.. ,  Science ,  1988, vol.  242 (pg.  229- 237) CrossRef Search ADS PubMed  2 Matthews J.,  Kricka L.J.. ,  Anal. Biochem. ,  1988, vol.  169 (pg.  1- 25) CrossRef Search ADS PubMed  3 Botstein D.,  White R.L.,  Skolnick M.,  Davis R.W.. ,  Am. J. Hum. Genet ,  1980, vol.  32 (pg.  314- 331) PubMed  4 Conner B.J.,  Reyes A.A.,  Morin C.,  Itakura K.,  Teplitz R.L.,  Wallace R.B.. ,  Proc. Natl Acad. Sci. USA ,  1983, vol.  80 (pg.  278- 282) CrossRef Search ADS   5 Jeffreys A.J.,  Wilson V.,  Thein S.L.. ,  Nature ,  1985, vol.  314 (pg.  67- 73) CrossRef Search ADS PubMed  6 Maxam A.M.,  Gilbert W.. ,  Proc. Natl Acad. Sci. USA ,  1977, vol.  74 (pg.  560- 564) CrossRef Search ADS   7 Sanger F.,  Coulson A.R.. ,  Proc. Natl Acad. Sci. USA ,  1977, vol.  74 (pg.  5463- 5467) CrossRef Search ADS   8 Southern E.M.. ,  J. Mol. Biol. ,  1975, vol.  98 (pg.  503- 517) CrossRef Search ADS PubMed  9 Kafatos F.C.,  Jones C.W.,  Efstradiadis A.. ,  Nucleic Acids Res. ,  1979, vol.  7 (pg.  1541- 1552) CrossRef Search ADS PubMed  10 Drmanac R.,  Crkvenjakov R.. ,  Yugoslav Patent Application 570,  1987 11 Lysov Y.P.,  Florentiev V.L.,  Khorlin A.A.,  Khrapko K.R.,  Shick V.V.,  Mirzabekov A.D.. ,  Dokl. Akad. Sci. USSR ,  1988, vol.  303 (pg.  1508- 1511) 12 Southern E.M.. ,  International patent PCT/GB89/00460,  1988 13 Drmanac R.,  Labat I.,  Brukner I.,  Crkvenjakov R.. ,  Genomics ,  1989, vol.  4 (pg.  114- 128) CrossRef Search ADS PubMed  14 Khrapko K.R.,  Lysov Y.P.,  Khorlin A.A.,  Schick V.V.,  Florentiev V.L.,  Mirzabekov A.D.. ,  FEBS Lett. ,  1989, vol.  256 (pg.  118- 122) CrossRef Search ADS PubMed  15 Fodor S.P.A.,  Read J.L.,  Pirrung M.C.,  Stryer L.,  Lu A.T.,  Solas D.. ,  Science ,  1991, vol.  251 (pg.  767- 773) CrossRef Search ADS PubMed  16 Téoule R.,  Roget A.,  Livache T.,  Barthet C.,  Bidan G.. ,  International patent PCT/FR94/00354,  1994 17 Matson R.S.,  Rampal J.,  Coassin P.. ,  Anal. Biochem. ,  1994, vol.  217 (pg.  306- 310) CrossRef Search ADS PubMed  18 Melchoir B.W.,  Von Hippel P.H.. ,  Proc. Natl Acad. Sci. USA ,  1973, vol.  70 (pg.  298- 302) CrossRef Search ADS   19 Orosz J.M.,  Wetmur J.G.. ,  Biopolymers ,  1977, vol.  16 (pg.  1183- 1199) CrossRef Search ADS PubMed  20 Wood W.I.,  Gitschier G.,  Lasky L.A.,  Lawn R.. ,  Proc. Natl Acad. Sci. USA ,  1985, vol.  82 (pg.  1585- 1588) CrossRef Search ADS   21 Jacobs K.A.,  Rudersdorf R.,  Neill S.D.,  Dougherty J.P.,  Brown E.L.,  Fritsch E.F.. ,  Nucleic Acids Res. ,  1988, vol.  16 (pg.  4637- 4650) CrossRef Search ADS PubMed  22 Maskos U.,  Southern E.M.. ,  Nucleic Acids Res. ,  1992, vol.  20 (pg.  1675- 1678) CrossRef Search ADS PubMed  23 Maskos U.,  Southern E.M.. ,  Nucleic Acids Res. ,  1993, vol.  21 (pg.  4663- 4669) CrossRef Search ADS PubMed  24 Rees W.A.,  Yager T.D.,  Korte J.,  Von Hippel P.H.. ,  Biochemistry ,  1993, vol.  32 (pg.  137- 144) CrossRef Search ADS PubMed  25 Ricelli P.V.,  Benight A.S.. ,  Nucleic Acids Res. ,  1993, vol.  21 (pg.  3785- 3788) CrossRef Search ADS PubMed  26 Southern E.M.,  Case-Green S.C.,  Elder J.K.,  Johnson,  Mir K.U.M.,  Wang L.,  Williams J.C.. ,  Nucleic Acids Res. ,  1994, vol.  22 (pg.  1368- 1373) CrossRef Search ADS PubMed  27 Hoheisel J.D.,  Lehrach H.. ,  FEBS Lett. ,  1990, vol.  274 (pg.  103- 106) CrossRef Search ADS PubMed  28 Hoheisel J.D.. ,  Nucleic Acids Res. ,  1996, vol.  24 (pg.  430- 432) CrossRef Search ADS PubMed  29 Nguyen H.K.,  Auffray P.,  Asseline U.,  Dupret D.,  Thuong N.T.. ,  Nucleic Acids Res. ,  1997, vol.  25 (pg.  3059- 3065) CrossRef Search ADS PubMed  30 Beaucage S.L.,  Caruthers M.H.. ,  Tetrahedron Lett. ,  1981, vol.  22 (pg.  1859- 1862) CrossRef Search ADS   31 Draper D.E.. ,  Nucleic Acids Res. ,  1984, vol.  12 (pg.  989- 1002) CrossRef Search ADS PubMed  32 Ludwig J.,  Eckstein F.. ,  J. Org. Chem. ,  1989, vol.  54 (pg.  631- 635) CrossRef Search ADS   33 Maniatis T.,  Fritsch E.F.,  Sambrook J.. ,  Molecular Cloning; A Laboratory Manual ,  1982 Cold Spring Harbor, NY Cold Spring Harbor Laboratory Press 34 Fasman G.D.. ,  Handbook of Biochemistry and Molecular Biology ,  1975 3rd Edition Cleveland, OH CRC Presspg.  589  Nucleic Acids, Volume 1 35 Senior M.,  Jones R.A.,  Breslauer K.J.. ,  Biochemistry ,  1988, vol.  27 (pg.  3879- 3885) CrossRef Search ADS PubMed  36 Williams J.C.,  Case-Green S.C.,  Mir K.U.,  Southern E.M.. ,  Nucleic Acids Res. ,  1994, vol.  22 (pg.  1365- 1367) CrossRef Search ADS PubMed  37 Doktycz M.J.,  Morris M.D.,  Dormady S.J.,  Beattie K.L.,  Jacobson K.B.. ,  J. Biol. Chem. ,  1995, vol.  270 (pg.  8439- 8445) CrossRef Search ADS PubMed  38 Aboul-ela F.,  Koh D.,  Tinoco I.Jr. ,  Nucleic Acids Res. ,  1985, vol.  13 (pg.  4811- 4822) CrossRef Search ADS PubMed  39 Arnott S.,  Chandrasekaran R.,  Hall I.H.,  Puigjaner L.C.. ,  Nucleic Acids Res. ,  1988, vol.  11 (pg.  4141- 4155) CrossRef Search ADS   40 Sasisekharan V.,  Zimmerman S.,  Davies D.R.. ,  J. Biol. Chem. ,  1975, vol.  92 (pg.  171- 179) 41 Person A.M.,  Rich A.,  Krieger M.. ,  J. Biol. Chem. ,  1993, vol.  268 (pg.  3546- 3554) PubMed  42 Blume S.W.,  Guarallo V.,  Zacharias W.,  Miller D.M.. ,  Nucleic Acids Res. ,  1997, vol.  25 (pg.  617- 625) CrossRef Search ADS PubMed  43 Raynaud F.,  Asseline U.,  Roig V.,  Thuong N.T.. ,  Tetrahedron ,  1996, vol.  52 (pg.  2047- 2061) CrossRef Search ADS   44 Ehrlich M.,  Norris K.F.,  Wang R.Y.H.,  Kuo K.C.,  Gehrke C.W.. ,  Biosci. Rep. ,  1986, vol.  6 (pg.  387- 393) CrossRef Search ADS PubMed  © 1998 Oxford University Press
TI - The stability of duplexes involving AT and/or G4EtC base pairs is not dependent on their AT/G4EtC ratio content. Implication for DNA sequencing by hybridization
JO - Nucleic Acids Research
DO - 10.1093/nar/26.18.4249
DA - 1998-09-01
UR - https://www.deepdyve.com/lp/oxford-university-press/the-stability-of-duplexes-involving-at-and-or-g4etc-base-pairs-is-not-u3ayJAKp6V
SP - 4249
EP - 4258
VL - 26
IS - 18
DP - DeepDyve
ER -