Nucleic Acids Research

Oligonucleotide synthesis

All DNA oligonucleotides were purchased from Operon Inc. and purified by HPLC with a Zorbax XDB-C8 reverse phase column (300 Å pore size) using aqueous triethylammonium acetate (100 mM, pH 7.2)/acetonitrile (7-20%) eluant and UV detection. Appropriate fractions were combined, dried in vacuo, dissolved in 1 ml dH₂O and desalted by passage through an NAP-10 Sephadex cartridge (Pharmacia). The oligonucleotides in the eluant samples were vacuum dried, dissolved in 0.5 ml of Tris-EDTA buffer (pH 7.0) and quantified by absorption at 260 nm. Molar extinction coefficients for the oligomers were calculated using dinucleotide absorbance data (31). All DNA oligomers utilized are shown in Figure 1b.

For gel mobility shift assays, the oligonucleotides were end-labeled with [[gamma]-³²P]ATP (Amersham) using T4 polynucleotide kinase. Each oligonucleotide (10 pmol) was mixed with 5 U of T4 polynucleotide kinase in 1× T4 polynucleotide kinase buffer and each sample mixture was incubated at 37°C for at least 2 h. The unreacted [[gamma]-³²P]ATP was removed by filtering the labeled sample through a QIAquick nucleotide removal column (Qiagen).

Thermal denaturation experiments

Samples characterized by thermal denaturation/renaturation curves monitored by UV absorbance contained a 1:1 molar ratio of an oligonucleotide and either the 29 nt hairpin DNA (H29) or a 12 nt:18 nt partial duplex with dangling end (e.g. DT12/6). Experiments were carried out using 1 µM each of the single strand and either the duplex or hairpin in 1× NAE (100 mM sodium acetate/acetic acid, 1 mM EDTA) at pH 5.0.

Melting curves were taken with a Varian DMS-300 or Cary1E UV-vis spectrophotometer. A temperature probe inserted in a temperature reference cuvette adjacent to the sample was used to monitor the temperature. Samples were heated to 75°C and slowly cooled to room temperature prior to a melting experiment. Absorption at 268 nm was monitored while the temperature was increased at a rate of ~0.5°C/min. Data were taken every 0.1°C. T_m values were reproducible to ±0.5°C.

Thermodynamic analysis of triplex and duplex transitions

Complexes formed between the hairpin DNA and each single strand exhibited two transition steps. The first step was the transition from the triple helix to single strand and hairpin DNA. The second step corresponded to the transition of the hairpin DNA to its random coil form. These equilibrium reactions may be described as

triplex <--> (H29) + single strand	1
(H29) + single strand <--> coil + single strand	2

K₂ and K₃ are defined as the temperature-dependent equilibrium constants for formation of the hairpin and triplex, respectively (the reverse reaction of equations 1 and 2). The experimental transitions were analyzed by comparing them with a theoretical melting curve determined for the coupled equilibria described by equations 1 and 2 (32).

The fraction of single-stranded oligomers that are triplex, [thetas]₃(t), is given by

[thetas]3(t) = K2K3 (Ct/2) [1 - [thetas]3(t)]/{K2K3 (Ct/2) [1 - [thetas]3(t)] + K2 + 1}

3

and the fraction of H29 molecules that are hairpins, [thetas]₂(t), is given by

[thetas]2(t) = {K2K3 (Ct/2) [1 - [thetas]3(t)] + K2}/{K2K3 (Ct/2) [1 - [thetas]3(t)] + K2 + 1}.

4

C_t is the total concentration of the H29 and the single-stranded molecules. Solving equations 3 and 4 requires initial estimates of K₂ and K₃ and thus estimates of T_m2, [Delta]H°₂ and T_m3, [Delta]H°₃.

T_m2 and [Delta]H°₂ were first determined using the normalized melting curve of hairpin H29 alone. The experimental fractionof H29 molecules in the hairpin conformation at temperature t(in the absence of an oligonucleotide) and the experimental derivative melting curve were compared with theoretical values of [thetas]₂[prime](t) = K₂/(1 + K₂) and d[1 - [thetas]₂[prime](t)]/dt. Initial estimates of T_m2 and [Delta]H°₂ were used to calculate K₂(t) values by integration of the van't Hoff equation and adjusted to give the best fit of the theoretical curve to experiment.

Values for [Delta]H°₃ and T_m3 were determined by an iterative analysis in a similar way. Initial estimates were made to the K₃(t) values. K₃(t) and K₂(t) were then used to calculate [thetas]₃(t)from equation 3 and [thetas]₂(t) from equation 4. The term [1 - [thetas]₂(t) + [alpha][1 - [thetas]₃(t)]]/(1 + [alpha]) was determined and its derivative plotted against temperature and compared with the normalizedexperimental curve dAbs/dt. [alpha] is the ratio of the triplex/duplex transition heights. [Delta]H°₃ and T_m3 values were varied to give the best fit.

The complexes formed between single strands and the tailed duplex molecules displayed single step denaturation transitions. Plots of the fraction of molecules denatured versus temperature were determined from the absorbance-temperature profiles after removing the linear portions of the upper and lower baselines. Since all single strand-tailed duplex complexes had irreversible melting transitions, thermodynamic parameters were not evaluated. We defined T_den as the midpoint temperature of the denaturation curve and T_ren as the midpoint temperature of the renaturation curve.

Gel retardation and competition assays

Triplex formation was analyzed with non-denaturing 20% polyacrylamide gels (19:1 acrylamide/bis-acrylamide) containing 1× NAE buffer (pH 5.0). A gel shift assay was employed to assess the binding of ³²P-labeled single strands to the H29 hairpin DNA or the DT12/6 molecules. Samples contained increasing amounts of H29 or DT12/6 molecules with a constant concentration of single strands (300 nM). Complexes were prepared by heating the DNAs to 75°C in 1× NAE buffer and cooling them to 25°C. Electrophoresis was performed for ~18 h (6 V/cm) at 25°C. The amount of radioactivity in the gel bands corresponding to the free single strand and triplex was quantified using an Ambis 4000 system (Scanalytics).

An association constant for PY12 binding to H29 was determined from gel shift data using the relation K₃ = [T]/[PY12][H29] where [PY12], [H29] and [T] represent concentrations of free PY12, free H29 and the triplex. The amount of bound and free ³²P-labeled PY12 was determined as a function of H29 added. Taking the log of both sides of the equation for the equilibrium constant, one obtains at the point where 50% of the labeled PY12 is in the triplex the relation log₁₀K₃ = -log₁₀[H29]_0.5. [H29]_0.5 is the concentration of free H29 when 50% of PY12 is bound. It was determined from the total concentration of H29 added minus the concentration of triplex at the 50% point.

The gel competition assay was carried out by adding varying amounts of a competing unlabeled single-stranded oligomer(e.g. PY17) to fixed concentrations of ³²P-labeled PY12 (300 nM) and H29 (390 nM). The relative free energy of binding each competing single strand to H29 was determined from the plot of the decrease in the fraction of the H29·PY12 triplex ([thetas]_t) as a function of competing single strand (Fig. 4). [thetas]_t was experimentally determined from the band intensities:

[thetas]t = [triplex - background]/[(triplex - background) + (free strand - background)]

5

From the equations relating equilibrium constants and concen-trations and the total concentration of a molecule to its various forms, one obtains the theoretical relationship

[thetas]t = 1/{1 + ([X] Kx)/(K3 [Tx])}

6

K_x/K₃ is the ratio of binding constants of the competing single strand X and PY12 to H29, [X] is the free concentration of X and [T_x] is the concentration of the triplex formed by X and H29. K_x/K₃ is evaluated from equation 6 at [thetas]_t = 0.5. The values of [X]_0.5 and [T_x]_0.5 are determined from the equations relating the concentrations of molecules.

Influence of oligonucleotide extended sequence on hairpin-oligonucleotide triplex transition

UV absorbance-monitored melting curves were obtained for complexes involving H29 and the oligonucleotides, PY12, PY17 and PY18. Figure 2 shows the melting curves of several of the DNA complexes. Denaturation-renaturation experiments indicate that the transitions occur in a reversible manner under the experimental conditions employed (data not shown). The melting curves show two step transitions allowing separate analysis of the triplex melting transition (first step) and the hairpin DNA transition (second step).

Figure 2. Normalized UV monitored melting transitions of hairpin DNA H29 ([open square]), H29·PY12 complex ([closed square]), H29·PY17 complex ([inverted closed triangle]) and the H29·PY18 complex ([inverted open triangle]). Sample concentrations were 1 µM each of H29 and the single strand. The solvent was 1× NAE at pH 5.0.

The T_m value for the triplex to hairpin plus oligonucleotide transition for PY12·H29 was 54.0°C. When 5 nt were added to the 3[prime]-end of PY12 to produce a region complementary to the loop residues of H29, the triplex transition decreased by 4.7°C (PY17). The interaction between the hairpin loop and the 5 nt extension on PY17 destabilized rather than stabilized the triplex region. Inserting a thymine at the junction between the triplex forming bases and the 5 nt extension (PY18) produced a triplex transition that was more stable than PY17 but was still less stable than PY12. T_m values of the above triplex complexes are listed in Table 1. Analysis of the transitions of the oligomer-hairpin complexes (equations 3 and 4) yielded T_m values for the hairpin-coil step of 69.8 ± 0.4°C. This was essentially the same as the hairpin alone. We note that consideration of the triplex strand unwinding step of the complexes yield T_m values for the hairpin-coil step that are higher than suggested by casual inspection of Figure 2.

Results from a model-dependent van't Hoff analysis (33) of the melting curves of the complexes are summarized in Table 1. The [Delta]H° and [Delta]G° value at 25°C for the triplex formed by PY12·H29 agree very well with values obtained by Roberts for a triplex with a similar structure (32) ([Delta]H° = -80 kcal/mol, [Delta]G° = -15.7 kcal/mol). The triplex studied by Roberts was the same as PY12·H29 except that the hairpin loop sequence was TTTT. The van't Hoff analysis indicates that the PY17·H29 and PY18·H29 complexes have a free energy that is 1.5-2.0 kcal/mol less stable than PY12·H29. Melting studies were also carried out with H29 and the PY19 pyrimidine strand. The result confirmed the destabilizing effect of the strands longer than PY12 (data not shown). The juxtaposition of the loop residues and the extended portion of PY17, PY18 or PY19 appear to weaken Hoogsteen base pairing between H29 and the pyrimidine region of the single strands.

Table 1. van't Hoff analysis of melting transitions of H29 and single strands

Type of transition	Tm (°C)a	[Delta]H° (kcal/mol)b	[Delta]G°25°C (kcal/mol)c	[Delta]S° (cal/K.mol)d
Hairpin
H29 only	69.4	-73 ± 5	-9.5 ± 5	213 ± 17
Triplex
H29·PY12	54.0	-76 ± 5	-15.8 ± 5	-202 ± 17
H29·PY17	49.3	-71 ± 5	-13.8 ± 5	-192 ± 17
H29·PY18	50.3	-72.4 ± 5	-14.3 ± 5	-195 ± 17

^aT_m values were averages of two or three measurements and were reproducible to ±0.5°C.
^bValues and uncertainties were estimated from the best fit of equations 1 and 2 to the transition curves.
^cFree energy for the hairpin duplex was determined from [Delta]G° = [Delta]H°(1 - T/T_m); [Delta]G° for the triplex was derived as described in text and from relations described by Marky and Breslauer (33).
^d[Delta]S° was calculated from the equation [Delta]G° = [Delta]H° - T·[Delta]S°.

Gel mobility shift and gel competition assays

The gel mobility shift and competition assays were employed to provide an independent measure of the effect of the extended sequence on oligonucleotide binding to H29. The equilibrium binding of PY12 to H29 was evaluated by varying the amount of H29 in the presence of ³²P-labeled PY12. The amounts of free PY12 and H29·PY12 triplex were quantified on a non-denaturing gel. This information was used with the data from the gel competition experiments to determine the relative free energies between binding PY17 or PY18 and PY12 for H29.

Figure 3 shows a gel competition experiment in which PY17 was competed with PY12 for H29. When no H29 was added, all labeled PY12 migrated as free single strand (Fig. 3, lane C). In the presence of excess H29, but without a competing single strand, most of the PY12 shifted to a single triplex band (Fig. 3, lane 1). As the concentration of the competing PY17 increased, more PY12 migrated as free strands (Fig. 3, lanes 2-5). Figure 4 plots the fraction of PY12 bound to H29 ([thetas]_t) as a function of the competing oligonucleotides PY17 and PY18. Both oligonucleotides gave similar results within the accuracy of the experiments. Analysis of the data using equation 6 gave K_x/K₃ values indicating a difference in free energy of triplex formation with PY17 or PY18 versus PY12 of [Delta][Delta]G°_gel = 1.6 kcal/mol. This value was close to the values derived from van't Hoff assays, [Delta][Delta]G°_{van't Hoff} = 1.5-2.0 kcal/mol (Table 2). Both the gel shift and melting curve results indicate that the interaction of the extended or overhanging part of the pyrimidine-rich strands with the loop destabilizes the adjacent triplex. The results do not support the notion of adjacent, stable triplex and duplex domains for the complexes between the hairpin DNA and the extended pyrimidine-rich strands.

Figure 3. Gel electrophoresis competition assay. Autoradiogram of native polyacrylamide gel showing titration of unlabeled PY17 competing with ³²P-labeled PY12 for binding to hairpin DNA H29. Lane C, 300 nM ³²P-labeled PY12; lanes 1-5, 300 nM PY12 and 390 nM H29 and increasing amounts of PY17 as indicated above each lane.

Figure 4. Data from gel competition assay experiments. The fraction of ³²P-labeled PY12 in the complex with H29 is plotted as a function of the concentration of competing PY17 and PY18. Dashed lines correspond to two experiments with PY17 and solid lines correspond to two experiments with PY18.

Table 2. The relative free energy for triplex formation of hairpin DNA (H29) and single strands evaluated by van't Hoff analysis of melting curves and gel competition assay

Single strand binding to H29	[Delta][Delta]G°gel (kcal/mol)	[Delta][Delta]G°VH (kcal/mol)
PY12	0	0
PY17	1.6	2.0
PY18	1.6	1.5

It is worth noting that the value of K₃ determined from the gel shift assay yielded a [Delta]G° value for the PY12·H29 triplex at 25°C that was 6.5 kcal/mol lower than the value determined from the van't Hoff analysis. Although several potential explanations for this discrepancy are possible, the reason is uncertain. The van't Hoff determination of [Delta]H° from the melting curve may be inaccurate if the assumptions of a two-state transition and/or temperature-independent extrapolation of [Delta]H° are incorrect. However, van't Hoff derived values and calorimetric measurements for similar systems agree well with our value for the pH 5.0 solvent (34,35). The influence of triplex formation kinetics was examined by extending the annealing time to 2 h in the gel experiments. This did not alter the results. Potential reasons for the discrepancy are differences in the single-stranded states at the higher thermal energy of the melting event versus the measurement at room temperature (2) or alteration of the triplex-hairpin equilibrium due to gel electrophoresis.

Complexes of single strands and DT12/6

Thermal denaturation curves of four complexes of DT12/6 and single-stranded oligomers are shown in Figure 5. All melting curves of the three-stranded complexes showed single cooperative transitions indicating that the complexes dissociated into three single strands in one step. The midpoint temperature of the denaturation curves for all complexes examined are presented in Table 3. T_den for the DT12/6 molecule alone was 37.6°C. When DT12/6 formed a complex with PY12 there was a 7.2°C increase in T_den. This increase in thermal stability demonstrates the stabilizing effect of the Hoogsteen base paired PY12 on the duplex (36). A similar increase in stability was noted for the PY12 complex with DX12, the blunt-ended duplex. The T_den value of the PY17·DT12/6 complex was 46°C, 1.2°C higher than the PY12·DT12/6 complex. Unlike the loop sequence of the H29 hairpin DNA, the single-stranded end of DT12/6 interacted favorably with the five Watson-Crick complementary bases on the 3[prime]-end of PY17.

Figure 5. UV-monitored denaturation transitions of DNA complexes involving DT12/6 and DX12 in 1× NAE solvent at pH 5.0. Samples had 1 µM of the duplex or 1 µM each of duplex and single strand. [open square], DT12/6 only;[open diamond], PY12·DX12; [closed square], PY12·DT12/6; [closed diamond], PY17·DT12/6; [open triangle], PY18·DT12/6; [closed triangle], PY20·DT12/6.

The effect of adding one or two unpaired thymines at the junction of the triplex and duplex binding domains of PY17 was examined using PY18 and PY19. These two single strands can form the same 12 Hoogsteen triplex pairs and five Watson-Crick base pairs as PY17 (Fig. 1). The addition of the thymines at the junction enhanced the stability of the complex. T_den values of PY18·DT12/6 and PY19·DT12/6 were 4.2 and 5.7°C higher than the denaturation midpoint of PY17·DT12/6. The effect of adding one additional Watson-Crick complementary base on the 3[prime]-end of PY19 was examined with PY20. The one additional terminal A·T pair did not have a significant effect on denaturation.

The effect of replacing two G·C base pairs in the dangling end region with G·T mismatches was examined with complexes involving DT12/6M with PY18 and PY20. T_den for PY18·DT12/6 M and PY20·DT12/6M were 44.6 and 43.1°C, respectively, and were ~6-8°C lower in stability than the analogous DT12/6 base paired complexes (Table 3). This demonstrated the expected destabilizing effect of the mismatches. The influence of GC composition in the duplex region was examined using the DNA DT12/6G. All 6 nt of the dangling end were either G or C and were complementary to the six bases on the 3[prime]-end of PY20G (Fig. 1). T_den for this complex, PY20G·DT12/6G, was 54.5°C, or 3°C higher than PY20·DT12/6. The above results are consistent with adjoining duplex and triplex base pairing in these complexes.

Table 3. Denaturation and renaturation midpoint temperature of various DNA complexes^a

Type of complex	Tden (°C)b	Tren (°C)b
DX12 only	36.3	36.5
DT12/6 only	37.6	37.6
DX12 + PY12	43.2	42.1
DT12/6 + PY12	44.8	41.5
DT12/6 + PY17	46.0	40.6
DT12/6 + PY18	50.2	41.7
DT12/6 + PY19	51.7	42.8
DT12/6 + PY20	51.4	43.4
DT12/6M + PY18	44.6	41.3
DT12/6M + PY20	43.1	41.5
DT12/6G + PY20G	54.5	45.5

^aEach single strand was mixed with a duplex (DT12/6, DX12, DT12/6M or DT12/6G) to form the complexes indicated.
^bT_den, denaturation; T_ren, renaturation. All T_den and T_ren values were an average of two or three UV transition curves. Repeat experiments were within ±0.5°C of each other.

Thermal denaturation and renaturation curves of the bimolecular complexes of DT12/6 with PY12, PY17 and PY18 are shown in Figure 6. Unlike the melting transitions of H29 and the single strands, the transitions were not reversible. Although the denaturation temperatures of DT12/6 and the single strands increased with the stability of the duplex region, the midpoint temperatures of renaturation curves were essentially the same for the complexes involving PY12, PY17 and PY18 (Table 3). The average T_ren and standard deviation for the four complexes were 41.3 ± 0.4°C. Similar T_ren values were also obtained for the two complexes involving DT12/6M. The 12 Watson-Crick base pairs between R12 and Y18 or Y18M are common to the complexes. This implies that this duplex region is the likely nucleation site for renaturation. The slightly higher T_ren values of PY19·DT12/6, PY20·DT12/6 and PY20G·DT12/6G may correlate with increased stability of the duplex segment formed at the dangling ends. Watson-Crick base pairing at the dangling end appears to also influence renaturation of the three strands.

Figure 6. Denaturation and renaturation curves of (a) PY12·DT12/6, (b) PY17·DT12/6 and (c) PY18·DT12/6. Denaturation curves are indicated by ([open square]) and renaturation curves are indicated by ([closed square]). Experimental conditions as described in Figure 5 and in text.

Gel mobility shift assay of triplex-duplex formation with DT12/6

The gel mobility shift assay was employed to provide a different approach for characterizing the association of the single strands with the DT12/6 molecule. Figure 7 shows the effect of adding increasing amounts of unlabeled DT12/6 duplex on the gel mobility of ³²P-labeled PY18 at 300 nM. Two distinct bands were observed. The lower one corresponds to the free single strand, PY18, and the upper band corresponds to the complex of PY18 and DT12/6. The concentration of free PY18 decreases in an approximately linear fashion with increasing amounts of DT12/6. Essentially all of the PY18 molecules migrated as a complex when the ratio of DT12/6 to PY18 reached 0.8:1. The saturation of the PY18 molecules at a DT12/6 concentration below a 1:1 ratio is probably due to inaccuracies in estimating the concentrations. The single upper band indicates that a 1:1 complex forms between DT12/6 and the single strand. Titration experiments similar to Figure 7 with PY18 replaced by PY12, PY17 or PY20 gave similar results (not shown). It was not possible to distinguish different affinities among the single strands for DT12/6 from these experiments. This result is similar in outcome to the thermal renaturation studies and appears to reflect the high to low temperature complex formation.

Figure 7. The autoradiogram of a gel mobility shift assay in which increasing amounts of unlabeled DT12/6 was added to ³²P-labeled PY18 at 300 nM. Lane C, PY18 alone; lanes 1-5, PY18 mixed with amounts of DT12/6 indicated above each lane. Bands corresponding to the triplex complex and free PY18 are indicated.

The above studies indicate that a DNA oligonucleotide and a duplex molecule containing a dangling end can form a stable duplex-triplex junction. Insertion of one or two thymines at the junction increased the stability of the complexes to denaturation. Alteration of the bases in the dangling end region effects the stability of the complex in a manner expected for Watson-Crick base pairs. The findings are consistent with results obtained by Francois and Hélène (26) who examined a related but different model system. The observation that the single strand overhang decreases rather than increases the triplex melting curve of the hairpin DNA suggests that conformational constraints of the loop sequence do not allow the extended single strand to form a stable duplex-triplex structure.

ACKNOWLEDGEMENT

We thank April Chester for her assistance with some experiments.

REFERENCES

1. Mirkin,S.M. and Frank-Kamenetskii,M.D. (1994) Annu. Rev. Biophys. Biomol. Struct., 23, 541-576. MEDLINE Abstract

2. Plum,G.E., Pilch,D., Singleton,S.F. and Breslauer,K.J. (1995) Annu. Rev. Biophys. Biomol. Struct., 24, 319-350. MEDLINE Abstract

3. Mirkin,S.M., Lyamichev,B.I., Drushlyak,K.N., Dobrynin,V.N., Flippov,S.A. and Frank-Kamenetskii,M.D. (1987) Nature, 330, 495-497. MEDLINE Abstract

4. Htun,H. and Dahlberg,J.E. (1989) Science, 243, 1571-1576. MEDLINE Abstract

5. Schimizu,M., Hanvey,J.C. and Wells,R.D., (1989) J. Biol. Chem., 264, 5944-5949.

6. Ussery,D.W. and Sinden,R.R. (1993) Biochemistry, 32, 6206-6213. MEDLINE Abstract

7. Strobel,S.A., Doucette-Stamm,L.A., Riba,L., Housman,D.E. and Dervan,P.B. (1991) Science, 254, 1639-1642. MEDLINE Abstract

8. Strobel,S.A. and Dervan,P.B. (1992) Methods Enzymol., 216, 309-321. MEDLINE Abstract

9. Cooney,M., Czernuszewicz,G., Postel,E.H., Flint,S.J. and Hogan,M.E. (1988) Science, 241, 456-459. MEDLINE Abstract

10. Alunni-Fabbroni,M., Manfioletti,G., Manzini,G. and Xodo,L.E. (1994) Eur. J. Biochem., 226, 831-839. MEDLINE Abstract

11. Young,S.L., Krawczyk,S.H., Matteucci,M.D. and Toole,J.J. (1991)Proc. Natl Acad. Sci. USA, 88, 10023-10100. MEDLINE Abstract

12. Duval-Valentin,G., Thoung,N.T. and Hélène,C. (1992) Proc. Natl Acad. Sci. USA, 89, 504-508. MEDLINE Abstract

13. Dayn,A., Samadashwily,G.M. and Mirkin,S.M. (1992) Proc. Natl Acad. Sci. USA, 89, 11406-11410. MEDLINE Abstract

14. Mergny,J.I., Duval-Valentin,G., Nguyen,C.H., Perrouault,L., Faucon,B., Rougee,M., Montenay-Garetier,T., Bisagni,E. and Hélène,C. (1992) Science, 256, 1681-1684.

15. Maher,L.J.I., Dervan,P.B. and Wold,B. (1992) Biochemistry, 31, 70-81.

16. Giovannangeli,C., Perrouault,L., Escudé,C., Thuong,N. and Hélène,C. (1996) Biochemistry, 35, 10539-10548. MEDLINE Abstract

17. Ono,A., Tso,P.O.P. and Kan,L.-S. (1991) J. Am. Chem. Soc., 113, 4032-4033.

18. Miller,P.S., Bhan,P., Cushman,C.D. and Trapane,T.L. (1992) Biochemistry, 31, 6788-6793. MEDLINE Abstract

19. Jetter,M.C. and Hobbs,F.W. (1993) Biochemistry, 32, 3249-3254. MEDLINE Abstract

20. Giovannangeli,C., Thuong,N.T. and Hélène,C. (1993) Proc. Natl Acad. Sci. USA, 90, 10013-10017. MEDLINE Abstract

21. Escude,C., Giovannangeli,C., Sun,J.S., Lloyd,D.H., Chen,J.K., Graznov,S.M., Garestier,T. and Hélène,C. (1996) Proc. Natl Acad. Sci. USA, 93, 4365-4369. MEDLINE Abstract

22. Sun,J.S., Giovannangeli,C., Francois,J.C., Kurfurst,R., Montenay-Garestier,T., Asseline,U., Saison-Behmoaras,T., Thuong,N.T. and Hélène,C. (1991) Proc. Natl Acad. Sci. USA, 88, 6023-6027. MEDLINE Abstract

23. Kukreti,S., Sun,J.S., Garestier,T. and Hélène,C. (1997) Nucleic Acids Res., 25, 4265-4270.

24. Roberts,R.W. and Crothers,D.M. (1991) Proc. Natl Acad. Sci. USA, 88, 9397-9401. MEDLINE Abstract

25. Roberts,R.W. and Crothers,D.M. (1996) Proc. Natl Acad. Sci. USA, 93, 4320-4325. MEDLINE Abstract

26. Francois,J.C. and Hélène,C. (1995) Biochemistry, 34, 65-72. MEDLINE Abstract

27. Wang,S., Booher,M.A. and Kool,E.T. (1994) Biochemistry, 33, 4639-4644. MEDLINE Abstract

28. Booher,M.A., Wang,S. and Kool,E.T. (1994) Biochemistry, 33, 4645-4651. MEDLINE Abstract

29. Roberts,R.W. and Crothers,D.M., (1996) J. Mol. Biol., 260, 135-146. MEDLINE Abstract

30. Mundt,A.A., Crouch,G.J. and Eaton,E.B. (1997) Biochemistry, 36, 13004-13009. MEDLINE Abstract

31. Cantor,C.R., Warshaw,M. and Shapiro,H. (1970) Biopolymers, 9, 1059-1077. MEDLINE Abstract

32. Roberts,R.W. (1993) Physical Chemistry of Nucleic Acid Triple Helices, PhD thesis, Yale University, New Haven, CT.

33. Marky,L.A. and Breslauer,K. (1987) Biopolymers, 26, 1601-1620. MEDLINE Abstract

34. Volker,J., Botes,D.P., Lindsey,G.G. and Klump,H.H. (1993) J. Mol. Biol., 230, 1278-1290. MEDLINE Abstract

35. Manzini,G., Xodo,L.E., Gasparotto,D., Quadrifoglio,F., van der Marel,G.A. and van Boom,J.H. (1990) J. Mol. Biol., 213, 833-843. MEDLINE Abstract

36. Kandimalla,E.R. and Agrawal,S. (1994) Gene, 149, 115-121. MEDLINE Abstract

---

*To whom correspondence should be addressed. Tel: +1 404 894 3735; Fax: +1 404 894 0519; Email: roger.wartell@biology.gatech.edu
⁺Present address: Department of Biology, Yonsei University, Seoul 120-749, Korea

---

This page is run by Oxford University Press, Great Clarendon Street, Oxford OX2 6DP, as part of the OUP Journals
Comments and feedback: www-admin{at}oup.co.uk
Last modification: 15 Jan 1999
Copyright©Oxford University Press, 1999.
CiteULike    Connotea    Del.icio.us    What's this?

This Article Abstract Print PDF (126K) Alert me when this article is cited Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in ISI Web of Science Similar articles in PubMed Alert me to new issues of the journal Add to My Personal Archive Download to citation manager Search for citing articles in: ISI Web of Science (2) Request Permissions Commercial Re-use Guidelines for Open Access NAR Content Google Scholar Articles by Nam, K. H. Articles by Wartell, R. M. Search for Related Content PubMed PubMed Citation Articles by Nam, K. H. Articles by Wartell, R. M. Social Bookmarking          What's this?