A photostable fluorescent marker for the superresolution live imaging of the dynamic structure of the mitochondrial cristae

Abstract

Stimulation emission depletion (STED) microscopy enables ultrastructural imaging of organelle dynamics with a high spatiotemporal resolution in living cells. For the visualization of the mitochondrial membrane dynamics in STED microscopy, rationally designed mitochondrial fluorescent markers with enhanced photostability are required. Herein, we report the development of a superphotostable fluorescent labeling reagent with long fluorescence lifetime, whose design is based on a structurally reinforced naphthophosphole fluorophore that is conjugated with an electron-donating diphenylamino group. The combination of long-lived fluorescence and superphotostable features of the fluorophore allowed us to selectively capture the ultrastructures of the mitochondrial cristae with a resolution of ∼60 nm when depleted at 660 nm. This chemical tool provides morphological information of the cristae, which has so far only been observed in fixed cells using electron microscopy. Moreover, this method gives information about the dynamic ultrastructures such as the intermembrane fusion in different mitochondria as well as the intercristae mergence in a single mitochondrion during the apoptosis-like mitochondrial swelling process.

Keywords: STED microscopy; fluorescence probe; live-cell imaging; mitochondrial cristae; superresolution.

Fig. 1.

Properties of phosphole dyes. (A) Chemical structures of MitoPB Yellow and its simplified (1) and water-soluble (2) derivatives. (B) Normalized emission spectra of 1 in organic solvents and 2 in a PBS buffer (pH = 7.4). The dotted line indicates the depletion wavelength used in this study. (C) Comparison of the photostability of live HeLa cells labeled with MitoPB Yellow and cells stained with Rh123 or MTG. Confocal images were recorded under the same acquisition conditions (λ_ex = 488 nm). Signal intensities of each image (I) relative to the initial value (I₀) are plotted as a function of the recorded number. (D) Comparison of the fluorescence lifetime of MitoPB Yellow with Rh123 and MTG in live HeLa cells. Signal intensities (I) detected in each delay time (t_g) are normalized to the initial value (I₀) and plotted as a function of t_g. The fluorescence lifetime (τ) of each dye in mitochondria were determined by a single-exponential-decay fitting.

Fig. 2.

STED imaging of mitochondria of living HeLa cells stained with MitoPB Yellow (λ_ex = 488 nm; λ_STED = 660 nm). (A) Comparison of scanning images of mitochondria for the confocal (Left) and STED (Right; t_g = 3 ns; P_STED = 270 mW) recording. (Scale bar, 2 µm.) (B) STED images of mitochondria before and after deconvolution. (Scale bar, 1 µm.) (C) Intensity profile plot along the white lines in the raw (black) and deconvoluted (blue) STED images. (D) The FWHM resolution as a function of STED power P_STED. The solid line indicates a theoretical fit of the data to the equation for STED microscopy. (E) FWHM as a function of delay time t_g. Solid lines show fits of the data to the equation for gated CW STED with τ = 7.5 ns for P_STED = 108 mW (open square) and P_STED = 270 mW (filled square).

Fig. 3.

Two-color STED image: (A) mitochondrial inner membrane labeled with MitoPB Yellow (λ_ex = 470 nm; green); (B) outer membrane labeled with TOMM20-TMR (λ_ex = 540 nm; magenta); (C) merged image; (D) signal intensity profile across the mitochondrial membranes for the two channels. The images were deconvoluted using Huygens deconvolution software. Zoomed views of boxed regions in white are shown in Insets [Scale bar, 1 µm (white) and 0.5 µm (yellow for Inset).]

Fig. 4.

Morphological changes of the mitochondrial inner membrane captured by STED microscopy (λ_ex = 488 nm; λ_STED = 660 nm). (A) Deconvoluted STED images showing changes in the mitochondrial morphology under concomitant change of the cristae density upon nutrition starvation for 3 and 12 h (HBSS containing Ca²⁺ and Mg²⁺). (Scale bar, 2 µm.) (B) Comparison of the number of cristae per micrometer of mitochondrial length before and after incubation for 3 and 12 h under starvation conditions (n = 20). (C) STED image of cristae in HeLa cells, pretreated with 10 µM mitochondrial DNA replication inhibitor (ddC) for 5 d followed by staining with MitoPB Yellow. (Scale bar, 2 µm.)

Fig. 5.

(A–C) Time-lapse STED imaging of mitochondrial dynamics in MitoPB Yellow-labeled cells; acquisition conditions: λ_ex = 488 nm; λ_STED = 660 nm; P_STED = 108 mW; t_g = 3 ns; line average = 1; scan speed = 100 Hz; frame rate = 0.77 fps. (Scale bar, 2 µm; Huygens deconvolution was applied.)