N-aryl pyrido cyanine derivatives are nuclear and organelle DNA markers for two-photon and super-resolution imaging

Abstract

Live cell imaging using fluorescent DNA markers are an indispensable molecular tool in various biological and biomedical fields. It is a challenge to develop DNA probes that avoid UV light photo-excitation, have high specificity for DNA, are cell-permeable and are compatible with cutting-edge imaging techniques such as super-resolution microscopy. Herein, we present N-aryl pyrido cyanine (N-aryl-PC) derivatives as a class of long absorption DNA markers with absorption in the wide range of visible light. The high DNA specificity and membrane permeability allow the staining of both organelle DNA as well as nuclear DNA, in various cell types, including plant tissues, without the need for washing post-staining. N-aryl-PC dyes are also highly compatible with a separation of photon by lifetime tuning method in stimulated emission depletion microscopy (SPLIT-STED) for super-resolution imaging as well as two-photon microscopy for deep tissue imaging, making it a powerful tool in the life sciences.

Fig. 1. Molecular design and in vitro characterization of N-aryl PC derivatives.

a The structure and structural components of PC1. b The comparison of DNA/RNA selectivity of PC1, Hoechst 33342, and PG. c The titration curve of 100 nM PC1 with various concentration of hairpin oligonucleotides. Fluorescence intensity is expressed in arbitrary units (arb. u.). The error bars indicate means ± s.d. of three independent replicates. Source Data is available as a source data file. d The general structure of PC dyes and their substituent patters with corresponding compound names; the unit of methylene length and N-aryl groups are represented as “n” and “R” respectively. The normalized absorption (e) and fluorescence spectra (f) of all PC dyes when complexed with calf thymus double stranded DNA (dsDNA) in tris-EDTA buffer solution. (pH = 8.0); see details in Figs. S1, S2.

Fig. 2. Live cell imaging with PC1 and PC3.

a Live cell fluorescent microscopy with PC1, PC3, and commercialized red fluorescent DNA dyes. HeLa cells were stained with each dye at 100 nM. Images are maximum z-projections of total planes (1 µm intervals). b Quantification of cell proliferation rate of cells stained with DNA labeling dyes. HeLa cells were stained with each dye at 30 nM and observed every 5 min with z-sectioning (six frames at 3 µm steps) for 24 h. The proliferation rate was quantified as fold changes based on the number of cells between the first frame (0 h) and the last frame (24 h) of bright-field images. Gray bars indicate the results from only bright-field time-lapse imaging without fluorescent time-lapse imaging and white bars indicate the results from both bright-field and fluorescent time-lapse imaging. Error bar shows mean ± s.d. from three independent biological replicates (>25 cells per replicate). Individual data points were also indicated in black cross. Statistical significance (^*P < 0.05, ^**P < 0.01) of difference from control condition was examined by two-sided student t-test using Microsoft Excel. Source Data is available as a source data file. c Live cell fluorescent images of different culture cell types with 500 nM PC1 and PC3. The images are maximum z-projections of total planes (1 µm intervals) The representative images are shown from at two to three similar images. d Live cell fluorescent images of Arabidopsis leaf and root cells with 1 µM PC1 and PC3. The images are maximum z-projections of total planes (1.1 µm intervals) The representative images are shown from five to six similar images. e Comparison of imaging penetration for single and two-photon excitation microscopy in Arabidopsis root tip stained with 1 µM PC1. Images were shown every 10 µm steps from z-sectioning images at 1 µm interval. The representative images are shown from four similar images. f Time-lapse observation by two-photon microscopy excited with 1000 nm in Arabidopsis root stained with 5 µM PC1. Root tip was observed every 2 min with z-sectioning (50 frames at 2 µm steps). The representative images are shown from four similar images.

Fig. 3. Discrimination between nuclear DNA and mt-DNA with fluorescence lifetime of PC1.

a–c Concentration dependence of staining pattern with PC1. a 10 nM, b 1 nM, c 100 pM. The images are maximum z-projections of total planes (0.3 µm intervals). The representative images are shown from three to four similar images in each concentration. d–l Fluorescent intensity images (d, g, j) and FLIM based separation images of nuclear DNA and mitochondrial DNA (e, h, k) by phasor plot analysis (f, i, l). The pseudo colors of (e, h, k) correspond to the colors of circles in (f, i, l). The nuclear DNA, mt-DNA, and ch-DNA are shown in red, cyan, and yellow, respectively. HeLa cells (d–f) and NIH3T3 (g–i) were stained with 1 nM and 10 nM PC1, respectively and the fluorescent spectrum were collected between 540–650 nm excited at 532 nm. The representative images are shown from 5 and 13 similar images, respectively. Stomata in Arabidopsis leaf cells was stained with 300 nM PC1 and excited at 532 nm. The fluorescent spectrum of PC1 and chlorophyll autofluorescence were collected between 540–620 nm and 680–700 nm shown in green and magenta in (k), respectively. The representative images are shown from similar images of ten guard cells.

Fig. 4. Comparison of confocal and SPLIT-STED imaging in living HeLa cells stained with PC3.

a Confocal and super-resolution images of mt-DNA stained with 10 nM PC3 in living HeLa cells. Super-resolution images were obtained by two component separations using an n-exponential reconvolution model. b Enlarged images of the white dotted square region of a. The representative images are shown from five similar images. c An example of normalized fluorescence intensity profiles obtained from the region between arrows in b. Line profiles in STED and confocal image are shown in black and gray, respectively. FWHM values estimated by fitting with a Gaussian function are also indicated in the black line profile. d Box-whisker plots of FWHM value as a function of delay time in the minor axis of mt-nucleoids (n = 15 from three independent cells). Center line, median; box limits, upper and lower quartiles; whiskers, 1.5x interquartile range. Individual data points and the mean points are also indicated in white circle and black cross, respectively. Source Data is available as a source data file.