Hybridization kinetics is different inside cells

Abstract

It is generally expected that the kinetics of reactions inside living cells differs from the situation in bulk solutions. Macromolecular crowding and specific binding interactions could change the diffusion properties and the availability of free molecules. Their impact on reaction kinetics in the relevant context of living cells is still elusive, mainly because the difficulty of capturing fast kinetics in vivo. This article shows spatially resolved measurements of DNA hybridization kinetics in single living cells. HeLa cells were transfected with a FRET-labeled dsDNA probe by lipofection. We characterized the hybridization reaction kinetics with a kinetic range of 10 micros to 1 s by a combination of laser-driven temperature oscillations and stroboscopic fluorescence imaging. The time constant of the hybridization depended on DNA concentration within individual cells and between cells. A quantitative analysis of the concentration dependence revealed several-fold accelerated kinetics as compared with free solution for a 16-bp probe and decelerated kinetics for a 12-bp probe. We did not find significant effects of crowding agents on the hybridization kinetics in vitro. Our results suggest that the reaction rates in vivo are specifically modulated by binding interactions for the two probes, possibly triggered by their different lengths. In general, the presented imaging modality of temperature oscillation optical lock-in microscopy allows to probe biomolecular interactions in different cell compartments in living cells for systems biology.

Fig. 1.

TOOL microscopy. An IR laser (wavelength 1,455 nm) is heating the bottom of a cell culture chamber. Fast heat retraction is accomplished by a silicon substrate and thin chamber dimensions. Both heating and epi-illumination are modulated with a tunable phase shift θ and imaged with a standard CCD camera. AOM, acousto-optical modulator; L, lens; LED, light-emitting diode; F, filter; BS, beam splitter.

Fig. 2.

Hybridization kinetics inside a single HeLa cell. (A and B) dsDNA probe design: (A) Complementary strands were labeled with the FRET pair RhG (donor) and ROX (acceptor) and left-handed chimeric cytosines at 3′ and 5′ ends to suppress degradation. Melting curves for the donor (green) and the FRET (red) signals were anticorrelated and fitted by melting temperatures of ≈31 °C for the 12-mer (squares) and ≈35 °C for the 16-mer (circles), respectively. (C and D) Transfer function and best fit (red line) at frequencies 1–200 Hz for a single pixel (arrow in E) for the 16-bp probe. The reaction amplitudes of the donor (C) and the FRET signal (D) show equal magnitude but opposite signs. (E and F) Cellular maps of the hybridization time constant show highly similar kinetics in both the donor and the FRET channel. (Scale bars: 10 μm.) (G and H) Histograms from nuclear (N) and cytoplasmic (C) regions (ellipses in E). Within the error bars, donor and FRET signals yielded identical results. Values are mean ± SD.

Fig. 3.

Reaction speed in cellular compartments. (A) Shown is the 16-bp DNA probe. Maps of the reaction time constant (color-coded) for three individual cells and respective histograms from nuclear (N) and cytoplasmic (C) regions are shown. (B) As in A for a 12-bp DNA probe. Hybridization kinetics were always faster in the nucleus. (Scale bars: 10 μm.)

Fig. 4.

Different kinetics in vivo compared with in vitro for the 16-bp DNA (A and B) and the 12-bp DNA (C and D). The inverse time constant τ⁻¹ is plotted versus dsDNA concentration. Data are given as mean ± SD together with the best fit of Eq. 1 (solid line) and its 95% confidence interval (gray). In vivo experiments comprise data from nuclear and cytosolic regions of >10 individual cells. They overlapped and were treated as a single dataset during fitting. The in vivo data of the 16-mer (A) were well described by the parabolic fit, whereas the shape of the 12-mer data (C) was better described by a buffered kinetics (dashed line; see SI Text). As compared with in vivo, the experiments in vitro showed slower kinetics of the 16-bp DNA (B) whereas the kinetics of 12-bp DNA (D) was distinctly faster.

Fig. 5.

Effect of divalent ions and crowding agents on the hybridization kinetics in vitro for the 16-bp probe (A) or the 12-bp probe (B). Magnesium chloride speeded up the reaction, whereas dextrans and Ficoll had only a minor impact on the kinetics. DNA concentration was 20 μM; crowding agents had a final concentration of 20% (wt/vol). Error bars represent the SEM of four independent experiments. Asterisks indicate a significant difference between the sample against the reference in pure PBS (orange) according to Student's t test to the level P < 0.01.