Switchable DNA interfaces for the highly sensitive detection of label-free DNA targets

Rant et al. 10.1073/pnas.0703974104.

Supporting Information

Files in this Data Supplement:

SI Figure 7
SI Text
SI Figure 8
SI Figure 9
SI Figure 10
SI Table 1
SI Figure 11
SI Figure 12
SI Figure 13

Fig. 7. Hybridization of a 48-nt probe DNA layer with 10 nM perfectly matching complements, monitored by the absolute (DF) and relative (DF/F) fluorescence modulation amplitudes. Switching duty cycles of 100% and 5% are shown.

Fig. 8. Chemical structure of Cy3 and linker.

Fig. 9. Fluorescence spectra of Cy3-labeled 24-mer DNA in aqueous buffer solution as a function of hybridization ratio ( = conc_ds/conc_ss); 0% and 100% correspond to single- and double-stranded samples, respectively. The hybridization reactions were carried out in solutions of high salinity (containing [Tris] = 200 mM and [NaCl] = 1 M) for 2 h previous to the measurement. Solution composition during measurement: [DNA] = 0.5 mM, [Tris] = 10 mM, [NaCl] = 50 mM, which corresponds to the buffer used for the switching experiments. (Inset) Real-time observation of the increase of Cy3 fluorescence upon adding 0.5 mM unlabeled complementary DNA to a solution containing 0.5 mM Cy3-labeled single-stranded 24-mer DNA.

Fig. 10. Fluorescence intensity decay of Cy3 attached to single- and double-stranded 24-mer DNA. Excitation and detection wavelengths were 440 and 565 nm, respectively. Solid lines are fits obtained by a double exponential decay model (FluoFit software, Picoquant), taking into account the instrument response function (IRF, measured by detection scattered laser light). Solution composition: [DNA] = 0.5 mM, [Tris] = 10 mM, [NaCl] = 50 mM.

Fig. 11. Thermal DNA denaturation measured in bulk solution. mm0, 1, 2, 2a correspond to perfect complements, a single GC mismatch, two spatially separated mismatches, and two adjacent mismatched in the DNA sequences, respectively. ss is ssDNA-Cy3. Solution composition: [DNA] = 0.2 mM, [Tris] = 10 mM, [NaCl] = 50 mM.

Fig. 12. T_m determination using FRET. (Left) FRET principle. (Right) Temperature-dependent spectra. The excitation wavelength was 515 nm, selective to excite Cy3. Solution composition: [DNA] = 0.5 mM, [Tris] = 10 mM, [NaCl] = 50 mM.

Fig. 13. Temperature dependence of the switching amplitude of a 24-mer ssDNA layer. (Left) The measured fluorescence has been normalized by the fluorescence modulation of dsDNA (mm0) to allow a comparison with Fig. 4. (Right) Correction of the fluorescence modulation DF with the ssDNA-Cy3 temperature dependence (SI Fig. 11): DF/F_ssDNA-Cy3 (normalized to 25°C).

SI Text

Influence of DNA Switching on Target-Probe Hybridization

To test whether the electrical switching procedure affects the hybridization reaction, we conducted hybridization experiments during which a layer was continuously modulated by ac potentials (100% duty cycle) and compared the binding process to a second hybridization of the same layer after refreshing the sensor by thermal denaturation (see SI Fig. 7). The second hybridization was monitored by switching the layer for ~15 s every 5 min only, corresponding to a 5% duty cycle.

Both experiments yielded comparable binding kinetics and efficiencies, demonstrating that the DNA switching does not interfere with the hybridization reaction.

Photophysical Properties of Cy3 Conjugated to DNA

Cy3-labeled DNA was used as obtained from IBA (Göttingen, Germany). Cy3 was conjugated to the 3' end by a modified cytosine (NHS coupling), shown in SI Fig. 8.

Remarkably, we observed that the fluorescence intensity emitted by the Cy3 labels depends on the DNA conformation, i.e., differs for single- and double-stranded DNA.

SI Fig. 9 shows that the fluorescence intensity increases by a factor of 1.8 (±0.1) when ssCy3-DNA is hybridized with an unlabeled complementary strand and transforms into a double helix.

To investigate this conformation-dependent fluorescence enhancement further, we performed time-resolved fluorescence measurements. Fluorescence was excited by a pulsed laser diode (440-nm, 80-ps pulse width) and detected by time-correlated single photon-counting equipment (Picoquant, Berlin, Germany).

Analysis of the time-resolved data reveals two excited state lifetimes, t₁ » 0.7 ns and t₂ » 2.6 ns for both ssDNA and dsDNA. However, the relative contributions of these distinct decays, represented by the amplitudes A₁ and A₂, differ significantly (see SI Table 1). Whereas the fast decay clearly dominates for ssDNA, the slow decay gains relative weight for dsDNA.

Cyanine dyes are known to exist in different steroisomers, formed by rotation about a bond in the polymethine chain that connects the two head groups (1, 2). The thermodynamically stable conformation is the all-trans configuration, but population of a cis state can be induced, e.g., by photoisomerization (3). The isomerization is strongly influenced by the length of the polymethine chain, substituents along the chain, endgroups, solvent characteristics, temperature, etc. (4-7). The occurrence of two distinct excited-state lifetimes in time-resolved fluorescence measurements has been explained by the existence of trans and cis ensembles. Because of its dominant amplitude, we assign the shorter lifetime t₁ to the thermodynamically favorable trans state of Cy3 (1, 3).

When contrasting single-stranded data with double-stranded data, we find that the excited-state lifetimes are comparable, but their relative amplitudes differ significantly. The enhanced amplitude A₂ for dsDNA-Cy3 indicates that the interaction of the dye with the double helix (for instance, by stacking) results in a higher probability for Cy3 to adopt the cis configuration.

In summary, we find an enhanced fluorescence quantum yield for dsDNA-Cy3 compared with ssDNA-Cy3. At the same time, time-resolved measurements indicate an increased proportion of Cy3 cis isomers in the case of dsDNA-Cy3.

DNA Melting Temperature Measured in Bulk Solution

In addition to the temperature-dependent DNA-switching experiments, we performed reference measurements in bulk solution to assess the temperature dependence of the Cy3 fluorescence and compare the DNA melting temperatures (T_m) determined on gold surfaces (by the switching method) with T_m values determined in solution. The solution measurements were carried by using a temperature-controlled cuvette holder.

Generally, elevated temperatures are expected to have deteriorating effects on the fluorescence emission of dye molecules, as nonradiative deactivation mechanisms arising from, e.g., collisions or internal vibrations, intensify.

Indeed, we observe a gradual decrease of the Cy3 fluorescence with increasing temperatures, but the scaling behavior for dsDNA-Cy3 and ssDNA-Cy3 is markedly different. The logarithmic plot in SI Fig. 11 shows that the ssDNA-Cy3 fluorescence exhibits an exponential decrease (straight line) with increasing temperature, whereas the fluorescence decrease observed for dsDNA-Cy3 is nonexponential (for T <50°C, it is roughly linear).

In the temperature range between 55°C and 70°C, the dsDNA-Cy3 fluorescence suddenly drops. The exact position of this decay depends on the number and location of mismatched base pairs in the DNA sequence. Subsequently, the traces coincide with the ssDNA-Cy3 trace.

Based on the observation that dsDNA-Cy3 exhibits stronger fluorescence than ssDNA-Cy3, we attribute the sudden fluorescence decrease to the melting (denaturation) of dsDNA.

We assume that the midpoint of the transition region approximately corresponds to the duplex T_m and find:

T_m^mm0 » 67°C, T_m^mm1 » 61°C, T_m^mm2 » 59°C, T_m^mm2a » 57°C.

Because the temperature dependence of ssDNA-Cy3 is inherently different from that of dsDNA-Cy3, it is not possible to account for the Cy3 temperature dependence by adjusting the dsDNA-Cy3 trace and using ssDNA-Cy3 data.

To validate that T_m can be determined from the double-stranded/single-stranded fluorescence decay, we carried out T_m measurements with Cy3-labeled 24-mer probes and Cy5-labeled complementary targets (see SI Fig. 12). Here, Cy3/Cy5 acts as donor/acceptor pair for FRET.

T_m was determined by evaluating the midpoint of the Cy3-fluorescence increase in the range of 60°C to 70°C:

T_m^mm0 » 66°C (FRET measurement).

The T_m value determined by the FRET measurement agrees with the T_m value determined by the double-stranded/single-stranded fluorescence decay measurement within experimental accuracy.

Temperature Dependence of ssDNA Switching

Here, we present temperature-dependent switching data for ssDNA as complementary information to the dsDNA data shown in Fig. 4.

SI Fig. 13 shows switching data from a ssDNA-Cy3 layer that was measured before the layer was consecutively hybridized (and thermally denatured) with different target sequences.

The temperature dependence of the fluorescence modulation amplitude, DF, observed for ssDNA and dsDNA is strikingly different. Whereas DF_dsDNA(T) decreases gradually with increasing temperature, we find an increase in DF_ssDNA(T) up to ~50°C, followed by a slight decrease at higher temperatures. When correcting for the intrinsic temperature dependence of ssDNA-Cy3 fluorescence (compare SI Fig. 11), the real switching efficiency of the ssDNA layer is revealed. Data in SI Fig. 13 Right show that the switching efficiency drastically increases with increasing temperatures, until it saturates around 70°C.

We assume that thermal fluctuations cause the DNA to extend further from the surface, which results in an enhanced switching efficiency (fluorescence modulation).

We have observed a similar, yet not as pronounced, trend for dsDNA when correcting for the dsDNA-Cy3 fluorescence temperature dependence within the range of 25°C to 50°C. We note, however, that it is not possible to account for Cy3-fluorescence temperature effects at higher temperatures, because of the double-strand melting. Because the temperature dependence of the Cy3 fluorescence depends on the DNA conformation (single-stranded or double-stranded), a meaningful correction of temperature influences on the observed fluorescence across the melting transition is not straightforward.

1. West W, Pearce S, Grum F (1967) J Phys Chem 71:1316-1326.

2. Scheibe G (1936) Angew Chem 52:631.

3. Widengren J, Schwille P (2000) J Phys Chem A 104:6416-6428.

4. Sahyun MRV, Serpone N (1997) J Phys Chem A 101:9877-9883.

5. Aramendia PF, Negri RM, San Roman E (1994) J Phys Chem 98:3165-3173.

6. Noukakis D, Van der Auweraer M, Toppet S, De Schryver FC (1995) J Phys Chem 99:11860-11866.

7. Sundström V, Gillbro T (1982) J Phys Chem 86:1788-1794.