Short-Range Spectroscopic Ruler Based on a Single-Molecule Optical Switch

We demonstrate a novel all-optical switch consisting of two molecules: a primary fluorophore (Cy5) that can be switched between a fluorescent and a dark state by light of different wavelengths, and a secondary chromophore (Cy3) that facilitates switching. The interaction between the two molecules exhibits a distance dependence much steeper than that of conventional Förster resonance energy transfer. This enables the switch to act as a ruler with the capability to probe distances difficult to access by other spectroscopic methods, thus presenting a new tool for the study of biomolecules at the single-molecule level.

Figure 1

A single-molecule optical switch. (a) Schematic diagram of the single-molecule switch. (b) Lower panel: alternating red and green laser pulses were used to switch the Cy5 molecule off and on, respectively. A weak, red probe beam ($5\mathrm{W}{\mathrm{c}\mathrm{m}}^{-2}$) was constantly on (not shown). Upper panel: the fluorescence time trace of a single Cy5 shows that the molecule was switched between its two fluorescence states. (c) Lower panel: the red laser was continuously on, and the green laser was periodically turned on and off, effectively gating the fluorescence output of the Cy5. Upper panel: the fluorescence time trace of a single Cy5 shows the molecule being switched on and off by the green laser. (d) Fluorescence images of individual Cy5 molecules taken at times specified in (c). A video showing the switching behavior is available online [14].

Figure 2

Switching kinetics of single molecules. (a) The rate constant $k_{\mathrm{o}\mathrm{f}\mathrm{f}}$ as a function of the red laser intensity. (b) The rate constant $k_{\mathrm{o}\mathrm{n}}$ as a function of the green laser intensity. Solid lines are linear fits to the data. The standard error for each point is indicated.

Figure 3

The switching mechanism. (a) $k_{\mathrm{o}\mathrm{f}\mathrm{f}}$ as a function of the potassium iodide concentration of the imaging buffer. (b) $k_{\mathrm{o}\mathrm{f}\mathrm{f}}$ as a function of the sucrose concentration ($\mathrm{w}/\mathrm{v}$). Concentrations of 0% and 45% correspond to solvent viscosities of 1 and 6 cP, respectively. The refractive index change of the buffer, induced by sucrose, causes a slight increase in the excitation intensity (amounting to 30%) at the maximum sucrose concentration used. The $k_{\mathrm{o}\mathrm{f}\mathrm{f}}$ values corrected for this effect are shown as open symbols. The changes in the potassium iodide and sucrose concentrations did not affect the absorption coefficient of Cy5.

Figure 4

Distance dependence of the Cy3-assisted photo recovery of Cy5. (a) Schematic diagram of the single-molecule switch with Cy5 and Cy3 separated by a distance $R$. (b) The $R$ dependence of the recovery rate constant of Cy5 $(d{k}_{\mathrm{o}\mathrm{n}}/dI)$. The $R$ values are determined according to the structure of B-form double-stranded DNA [14]. We note that the $R$ value reflects the distance between the attachment sites of Cy3 and Cy5 on the DNA but not necessarily the actual distance between the dyes, due to the flexible linkers connecting the dye and the DNA. The $d{k}_{\mathrm{o}\mathrm{n}}/dI$ values were obtained from the dependence of $k_{\mathrm{o}\mathrm{n}}$ on the green laser intensity ($I$), and were normalized against the value obtained at $R=0.8\mathrm{n}\mathrm{m}$. Also shown are the FRET efficiencies between Cy3 and Cy5, defined as $I_A/(I_A+I_D)$, measured at these distances. $I_A$ and $I_D$ are the fluorescence signals from the Cy5 and Cy3 molecules, respectively, when the Cy3 is excited with 532 nm laser light. (c) Schematic diagram of the hairpin ribozyme in the “docked” and “undocked” state. (d) Normalized $d{k}_{\mathrm{o}\mathrm{n}}/dI$ values for Cy5 on docked and undocked ribozymes.