Dual-probe RNA FRET-FISH in Yeast

Abstract

mRNA Fluorescence In Situ Hybridization (FISH) is a technique commonly used to profile the distribution of transcripts in cells. When combined with the common single molecule technique Fluorescence Resonance Energy Transfer (FRET), FISH can also be used to profile the co-expression of nearby sequences in the transcript to measure processes such as alternate initiation or splicing variation of the transcript. Unlike in a conventional FISH method using multiple probes to target a single transcript, FRET is limited to the use of two probes labeled with matched dyes and requires the use of sensitized emission. Any widefield microscope capable of sensitive single molecule detection of Cy3 and Cy5 should be able to measure FRET in yeast cells. Alternatively, a FRET-FISH method can be used to unambiguously ascertain identity of the transcript without the use of a guide probe set used in other FISH techniques.

Keywords: Fluorescence In Situ Hybridization; Budding yeast; RNA FISH; Saccharomyces cerevisiae; Single molecule; Transcription.

Figure 1.. Comparing Epi and inclined illumination.

A. An epi-fluorescence microscope has the illumination incident on the sample through the objective and the entire volume of the sample is illuminated. This leads to poor signal to noise for single fluorophores since the widefield microscope collects out of focus light. B. An inclined illumination geometry has the illumination displaced radially in the back focal plane of the objective so that the light becomes a thin laminated sheet in the sample volume ( Tokunaga et al., 2008 ). This enables z-sectioning and increases signal to noise significantly.

Figure 2.. Emission pathway of a microscope configured for two channel FRET.

The emission is transmitted through a dichroic with two transmission bands compatible with the FRET pair. An adjustible mechanical slit is placed at the image plane. A dichroic is placed near the image plane formed by the tube lens to split the two bands of emission. Filters are placed in the path between the first dichroic and the second dichroic, which recombines the emissions in the same direction. The broadband mirrors are adjusted so that the light from each path covers half of the sensor. If aligned properly, there should be no difference in orientation or magnification between the two images.

Figure 3.. Example images of dual-probe FISH-FRET.

A. Three channels of a FRET acquisition are shown on a single cell expressing yEvenus mRNA where every target should have both probes. The left and right images are during excitation using 532 nm light. These images represent a single z-slice of Cy5 emission (left) and Cy3 emission (right). Cy5 emission during direct excitation with 640nm light is shown in the center. This image was taken second so that the spots that are visible in the FRET (left) image but not the Cy5 (center) image are primarily due to photobleaching. The spots visible in the Cy3 (right) image are due to a combination of non-specific binding (lower melting temperature of the probe) and lower detection efficiency (inactive fluorophores). B. The 256x512 image of Cy5 under direct (Red) and FRET (Green) emission is shown. Where fluorophores were detected in both channels, the image is yellow. The direct excitation was measured second. Approximately 80% of fluorophores were detected in both under this excitation condition with the difference due to photobleaching during excitation by FRET. Shown above is an example where transcripts are countable (~40 spots per cell). A more extreme example (~1,000 transcripts expected) of FRET detection is shown in Wadsworth et al., 2017 . In this example, the control shows that crosstalk between channels is minimal.

Figure 4.. Controls for dual-probe FISH-FRET.

Each image shown is at the same scale and contrast. A. Cells that have only been treated with a Cy3 labeled probe are shown in dual-view where the emission on the top and bottom panels of a slice were acquired simultaneously. Some bleedthrough is expected in FRET experiments. Much of the intensity in the bottom channel during 532 nm excitation is due to cellular auto-fluorescence, however, there are some peaks in intensity due to Cy3 emission in the Cy5 channel. B. Cells with only Cy5 labeled probes are shown. Under 532 nm excitation there is very little fluorescence observed. Under 640 nm excitation, Cy5 is seen in the bottom panel and virtually no emission is observed in the top panel. C. In a TIR setup, Cy3 and Cy5 are observed when labeled on the same DNA oligo bound to the surface with BSA-biotin. These are designed to calibrate the affine transformation that maps the top panel to the bottom panel. They can also be used to calibrate sensitized emission to determine FRET efficiency since these fluorophores are separated by a known number of nucleotides and double stranded DNA is essentially rigid below its persistence length. Variations in spot intensity here are due to non-uniform illumination.