Photodestruction Intermediates Probed by an Adjacent Reporter Molecule

We used a fluorescence resonance energy transfer donor molecule to probe the multiple intermediates in the photoinduced destruction of an acceptor molecule. These intermediates are nonemitting but are still able to quench the fluorescence of the donor at a distance scale shorter than conventional fluorescence resonance energy transfer, suggesting novel biophysical applications.

Figure 1

The DNA used is a partial duplex: a $3^{\prime }$ single stranded DNA tail ($T_{20}$) is attached to an end of 40mer duplex. The donor ($\mathbf{D}$:Cy3) and acceptor ($\mathbf{A}$:Cy5) are attached as shown. A biotin ($\mathbf{B}$) that tightly binds to Streptavidin protein is attached to the DNA for immobilization to a quartz slide coated with Streptavidin. The sequences are $5^{\prime }$-Cy5-GCCCTGCTGCCGACCAACGATGGTTACATTCCCGCTGCTG-$3^{\prime }$biotin and $5^{\prime }$-CAGCAGCGGGAATGTAACCATCTTGGTCGGCAGCAGGGC-Cy3-$(T{)}_{20}$-$3^{\prime }$. Experiments were performed at room temperature in a buffer (20 mM Tris, pH 7.5, 6 mM NaCl, 1.7 mM ${\mathrm{M}\mathrm{g}\mathrm{C}\mathrm{l}}_2$, 10% ($v/v$) glycerol). (a) Fluorescence time records of donor (dotted line) and acceptor (solid line) of a single DNA in an aerated sample (0.1 s bin): For measurements at much lower laser intensities, larger time bins were used. (b) Fluorescence time records in a deaerated sample (0.4 s bin). ${\mathrm{O}}_2$ was enzymatically removed [12, 13, 22]. Note the time scale differences between (a) and (b). A cartoon illustrates stepwise photodestruction of the acceptor that results in the delay between the acceptor signal cessation and donor signal appearance.

Figure 2

(a) Histogram of $t_{\mathrm{d}\mathrm{e}\mathrm{l}\mathrm{a}\mathrm{y}}$ in aerated solution ($I_{\mathrm{e}\mathrm{x}\mathrm{c}}=I_0$). $I_0$ is the maximum intensity used and is $\sim 300\mathrm{W}/{\mathrm{c}\mathrm{m}}^2$. The data minus the first bin were fit by a single exponential decay yield $⟨t_{\mathrm{d}\mathrm{e}\mathrm{l}\mathrm{a}\mathrm{y}}⟩=0.48\mathrm{s}$ decay time and 32% of the molecules belonged to the fast component beyond the resulting exponential distribution. (b) $⟨t_{\mathrm{d}\mathrm{e}\mathrm{l}\mathrm{a}\mathrm{y}}⟩$ vs $I_0/I_{\mathrm{e}\mathrm{x}\mathrm{c}}$ with a linear fit. At the high $I_{\mathrm{e}\mathrm{x}\mathrm{c}}$ limit, (b) shows a nonzero ($\sim 0.5\mathrm{s}$) $⟨t_{\mathrm{d}\mathrm{e}\mathrm{l}\mathrm{a}\mathrm{y}}⟩$ (arrow). (c) Histogram of $t_{\mathrm{d}\mathrm{e}\mathrm{l}\mathrm{a}\mathrm{y}}$ in deaerated solution. The bin size (6 s) is 15 times larger than the time resolution of this measurement (0.4 s) and the rise in the nonexponential distribution occurs over three bins indicating at least two sequential kinetic steps. Also shown is a model fit (see text) with two time constants, ${\tau }_1$and ${\tau }_2$. (d) A model for the photodestruction pathway of Cy5.

Figure 3

(a) Fluorescence time record of a single DNA molecule labeled with two Cy3 dyes in an aerated sample (0.033 s bin). (b) Time record of (average acceptor intensity)$/$(average total intensity) obtained from 26 molecules that started with Cy5 in the $\mathbf{F}$ state and showed FRET decrease (dashed line and open circles), and time record of the donor signal averaged over 17 molecules that started with Cy5 in the $\mathbf{I}$ state (no fluorescence) and showed donor signal appearance (solid line and filled squares). The time bin was 0.1 s. The cartoon below (b) illustrates that Cy5 in the $\mathbf{I}$ state can quench the donor emission only at short distances, leading to donor dequenching at a very early phase of unwinding. Unwinding by E. coli Rep helicase was triggered by ATP addition at $\sim 1.1\mathrm{s}$.