Green to red photoconversion of GFP for protein tracking in vivo

Abstract

A variety of fluorescent proteins have been identified that undergo shifts in spectral emission properties over time or once they are irradiated by ultraviolet or blue light. Such proteins are finding application in following the dynamics of particular proteins or labelled organelles within the cell. However, before genes encoding these fluorescent proteins were available, many proteins have already been labelled with GFP in transgenic cells; a number of model organisms feature collections of GFP-tagged lines and organisms. Here we describe a fast, localized and non-invasive method for GFP photoconversion from green to red. We demonstrate its use in transgenic plant, Drosophila and mammalian cells in vivo. While genes encoding fluorescent proteins specifically designed for photoconversion will usually be advantageous when creating new transgenic lines, our method for photoconversion of GFP allows the use of existing GFP-tagged transgenic lines for studies of dynamic processes in living cells.

Figure 1. GFP variants used for photoconversion.

(a) Alignment of the coding region of the GFP variants. The alignment was created using GeneDoc (http://www.nrbsc.org/gfx/genedoc/). (b) Schematic representation of GFP transgenes used in photoconversion experiments. Pr1b-SP, secretory signal peptide from the tobacco pathogenesis-related protein 1; EGFP, enhanced green fluorescent protein; L, linker (GGGS)₃; KDEL, endoplasmic reticulum retrieval signal peptide; Htt^Q103, exon 1 of human Htt^Q103 gene; Cytosolic mGFP4-T in tobacco cell culture and cytosolic S65T-GFP in Drosophila gut cells. In cytosolic mGFP4-T, V (valine) was replaced by A (alanine), and Q (glutamine) was replaced by R (arginine) relative to EGFP. In cytosolic S65T-GFP, L (leucine) was replaced by F (phenylalanine) relative to EGFP.

Figure 2. Photoconversion of purified EGFP in vitro.

(a) Typical pre- and post-photoconversion images of a thin (~10 μm) layer of 30 μM EGFP using 405 nm excitation delivered through a 1.2 NA objective to the region within the circle using the Zen software bleaching mode (Green channel: 505–525 nm; Red channel: 580–670 nm; scale bar: 50 μm). See also Supplementary Movie S1. 10 iterations of 257 ms each were performed. (b) Emission spectra of photoconverted red state EGFP acquired using the Zeiss 710 spectral detector (5 nm bandwidth intervals). (c) Absorption spectra of EGFP (~2 μM) at pH 5.0 and pH 8.0. (d) Time course of the increase in red fluorescence with successive 405 nm irradiations within the region of interest (ROI). EGFP (30 mM) was subjected to a sequence of 10 “bleach” iterations over the ROI (1.8 seconds) followed by image acquisition using 488 and 561 nm illumination. The 405 nm power at the sample was at 2.5 mW, corresponding to 6.8 MW/cm at the focus of the 1.2 NA objective lens. Plotted data points are the average pixel values across the entire field of view which estimates the total photo-product produced. Error bars represent the SEM of three trials at pH 5.0 and two at pH 8.0. Black lines are fits to exponential increase for the red channel (1-exp(-αt)) and decrease (exp(-αt)) for the green. Returned values of α were ~0.14 s⁻¹ for all data sets except the decay of the pH 8.0 green channel (which was 0.24 s⁻¹).

Figure 3. Correlation between EGFP concentration and red fluorescence intensity following irradiation.

EGFP was diluted in PBS buffer at the following concentrations: 0.875, 1.7, 3.4, 6.8, 13.75, 27.5, 55 μM. All photoconversion and imaging parameters were kept similar for all tested EGFP concentrations. Data are from 3 independent experiments.

Figure 4. Photoconversion of GFP from green to red state in vivo.

Single snapshots from the pre-photoconversion and post-photoconversion are shown. (a) Tobacco suspension cell culture in which GFP is targeted to cytosol. Photoconverted GFP spreads through the cytosol via cytoplasmic strands. Photoconversion was performed at 50% laser power and 30 iterations of 111 milliseconds (ms). (b) Schematic representation of a tobacco suspension culture cell. The cell wall (brown) surrounds the plasma membrane (red). A large vacuole (blue) occupies the majority of the space inside the cell. The endoplasmic reticulum network (gray) surrounds the vacuole and the nucleus (white) and extends through the rest of the cytosol (white). Cytoplasmic strands exist as extensions of the cytosol between organelles. (c) Changes in fluorescence intensity in the irradiated area in (a) over time indicate increasing red and decreasing green fluorescence. (d) Drosophila gut cells from 3rd instar larvae expressing S65T-GFP. (e) Rat PC12 cells expressing EGFP fusion to exon 1 of human Htt^Q103 gene. White arrowheads highlight cytosolic inclusions. Images acquired by 30 iterations with the duration of 190 ms each. Laser power was adjusted at 70%. Bars 10 μm.

Figure 5. EGFP photoconversion occurs in N. benthamiana cells transiently expressing EGFP.

(a) ER-targeted EGFP photoconverts from green to red upon excitation with 405-nm laser at 60% of laser power and 100 iterations with duration of 145 milliseconds each. The photoconverted red protein travels within the ER network of the cell. (b) Cytosolic EGFP photoconverts and spreads within the cytosolic space of the cell. Photoconversion was performed at 70% of laser power and 30 iterations with duration of 190 milliseconds each. White circles represent the irradiated area. Bar 20 μm. (c) Schematic representation of a Nicotiana benthamiana epidermal leaf cell. Puzzle-shaped epidermal cells are surrounded with the cell wall (brown). The space between the cell walls of two neighboring cells, apoplast, is shown in green. The plasma membrane (red) surrounds the cell. A large vacuole (blue) occupies the majority of the inner space of the cell. Endoplasmic reticulum (black) is localized around the nucleus and spreads throughout the cytosol (white), which is pushed against the plasma membrane by the vacuole. Chloroplasts are drawn in black as oval structures in the cytosol.