A gentle palette of plasma membrane dyes

Abstract

Plasma membrane (PM) stains are important organelle markers for monitoring membrane morphology and dynamics. The state-of-the-art PM stains are bright, specific, fluorogenic, and compatible with superresolution imaging. However, when recording membrane dynamics using advanced fluorescence microscopes, PM is prone to photodynamic damage introduced by dyes due to its phospholipid bilayer nature. Here, we introduce PK Mem dyes tailored for time-lapse fluorescence imaging. By integrating triplet-state quenchers into the MemBright dyes featuring cyanine chromophores and amphiphilic zwitterion anchors, PK Mem dyes exhibited a three-fold reduction in phototoxicity and a more than four-fold improvement in photostability in imaging experiments compared to MemBright prototypes. These dyes enable 2D and 3D imaging of live or fixed cancer cell lines and a wide range of primary cells, at the same time pair well with various fluorescent markers. PK Mem dyes can be applied to neuronal imaging in brain slices and in vivo two-photon imaging. The gentle nature of PK Mem palette enables ultralong-term recording of cell migration, cardiomyocyte beating, spermiogenesis, and axonal growth cone dynamics, which are prohibitively challenging using traditional PM dyes. Notably, PK Mem dyes are optically compatible with STED/SIM imaging, which can handily upgrade the routine of time-lapse neuronal imaging, such as growth cone tracking and mitochondrial transportations, into nanoscopic resolutions.

Keywords: long-term imaging; membrane dyes; neuronal imaging; phototoxicity; super-resolution imaging.

Fig. 1.

PK Mem dyes employ cyclooctatetraenes as triplet-state quenchers to optically strengthen and photodynamically tame the MemGlow dyes. (A) Chemical structures of PK Mem dyes featuring the modular integration of cyclooctatetraenes and amphiphilic anchors to the cyanine palette. (B) Absorption (solid lines) and emission (dashed lines) spectra of PK Mem dyes in MeOH. (C) Schemes for the mechanisms of turn-on, photostability, and phototoxicity of PK Mem dyes. (D) Plot of the maximum fluorescence intensity against the concentration of PK Mem dyes showing their critical micelle concentration (CMC). (E) Photostability comparison among PK Mem dyes, Az-Cy dyes, and MemGlow dyes, labeled on fixed HeLa cells. (F) In vitro detection of reactive oxygen species (ROS) through 1,3-diphenylisobenzofuran decay assay.

Scheme 1.

Synthetic route of PK Mem 590 via a modular derivatization of cyanine dyes with amphiphilic linkers and cyclooctatetraenes.

Fig. 2.

2D and 3D confocal imaging of various cells using PK Mem dyes (A) Cancer and primary cells imaged in this study. (B) Laser scanning confocal microscopy (LSCM) images of live HeLa cells treated with PK Mem dyes (20 nM) for 5 min without washing. (Scale bar, 10 μm.) (C) Left, 3D reconstruction of live KB cells stained with PK Mem dyes. Inset in the Left panel is the top view of intercellular nanotubes indicated by an asterisk. Right, orthogonal projection obtained from z stacks. (Scale bar, 10 μm.) (D) Multicolor images of live mouse embryonic stem cells (mESC) stained with WGA-488 and PK Mem 650. Mitochondria were stained with PK Mito Red. (Scale bar, 10 μm.) (E) Multicolor image of a live rat sperm labeled with PK Mem 555 (200 nM), SPY505-DNA, and SPY650-tubulin. (Scale bar, 10 μm.) (F) Multicolor image of a live adult rat cardiomyocyte labeled with PK Mem 555 (1 μM), Hoechst, PK Mito Deep Red, and Lyso-Tracker Green. (Scale bar, 10 μm.) (G) LSCM of live hippocampal primary neurons stained with PK Mem 555 (20 nM, 10 min) without washing. (Scale bar, 10 μm.)

Fig. 3.

The low phototoxicity of PK Mem 555 enables long-term recording of cell activities. (A) Long-term confocal recordings of HeLa cells labeled with MemGlow 560 and PK Mem 555. Blebbing events are highlighted with arrows, indicating photodamage. (Scale bar, 10 μm.) (B) Scatter plots of the mean frame number in which the blebs emerge (N > 5). Statistical significances were calculated with one-tailed Welch’s t test (or two-tailed for symmetric stimulation), and P values were given for each comparison. (C) Long-term LSCM recordings of L929 cells labeled with MemGlow 560 and PK Mem 555 (Video S4). Mitochondria were stained with PK Mito Deep Red. The rightmost images show magnified views of the dashed white boxes in the second column, illustrating migrasomes and shrunken cells. (Scale bar, 20 μm.) (D) The migration speed of cells 1 to 6 in C as the function of frame number. The blue and red lines represent cells labeled with PK Mem 555 or MemGlow 560, respectively. (E) Long-term spinning-disk confocal microscope recordings of spermatids labeled with MemGlow 560 and PK Mem 555 on the sperm activation process (Video S6). The yellow, green, and purple arrows point to cells undergoing membrane fusion, spermatozoon, and dead cells, respectively. (Scale bar, 10 μm.) (F) Cumulative bar chart of cell status changes over time. Red, yellow, green, and purple bars represent the population of nonactivated cells, cells undergoing activation, activated cells, and dead cells, respectively. The control group was treated with the same concentration of dye and pronase as the experimental group and incubated for an identical duration under light-protected conditions, after which the population of various cell types were quantified.

Fig. 4.

The low phototoxicity of PK Mem 555 enables high-temporal-resolution imaging of neuron growth cone dynamics. (A) Schematic illustrating the experimental workflow for comparing plasma membrane (PM) dyes. Embryonic day 5 (E5) dorsal spinal cord (dSC) explants were dissected from wild-type chick spinal cord, plated, and cultured for 1 to 2 d in vitro (DIV) to allow commissural axons to grow on the substrate. Then, PM dyes were added to the culture medium, and axons were imaged using spinning disk confocal microscopy. (B) Representative examples of a time-lapse recording of commissural axons labeled with either PK Mem 555 or MemGlow 560. One image was taken every 2 s for 20 min. Axons labeled with MemGlow 560 showed rapid growth cone collapse (yellow arrows), whereas those labeled with PK Mem 555 did not show collapse (white arrows; Video S7). (Scale bar, 10 μm.) (C) Quantification of the percentage of growth cone collapse per embryo. N(embryos) = 3 each; n (growth cones) = 26 (PK Mem) and 44 (MemGlow); P = 0.0002; Unpaired two-tailed t test. (D) Single time point (#350) of a time-lapse recording of a growth cone imaged for PK Mem 555 and bright-field, displaying high lamellipodial activity (Video S8). The rectangles indicate the region of interest (ROI) shown in (E). White arrows point to clearly defined filopodial contacts with the substrate that could not be precisely resolved in the bright-field image. (Scale bar, 5 µm.) (E) ROI of the leading edge lamellipodium shown in (D) followed for 20 s at 1 image/s time resolution. PK Mem 555 allowed the dynamic growth of the lamellipodial leading edge to be followed over time. (F) Single time point (#152) of a time-lapse recording of a growth cone imaged for PK Mem 555 exhibiting high filopodial activity (Video S9). The rectangle indicates the ROI containing a filopodium shown in (G). (G) ROI of the filopodium shown in (F) followed for 2.5 s at 1 image/0.25 s time resolution. White arrows point to the highly dynamic lateral motion of the filopodium. Frame numbers are shown in parentheses. (Scale bar, 5 µm.) (H) Four time points were extracted from a 15-min time-lapse recording (1 image/10 s) of a fast-growing axonal growth cone labeled with PK Mem 555 and SPY650-FastAct, revealing actin filaments (Video S10). White arrows indicate filopodia at the growth cone leading edge enriched with actin filaments. (Scale bar, 5 µm.) (I) Four time points were extracted from a 2-min time-lapse recording (1 image/s) of an axonal growth cone labeled with PK Mem 650 and the SPY555-tubulin, revealing microtubules (Video S11). This allowed for concomitant live monitoring of the axonal growth cone PM and microtubule bundle dynamics within the growth cone (white arrows). Frame numbers are shown in parentheses. (Scale bar, 5 µm.)

Fig. 5.

PK Mem dyes are compatible with superresolution imaging. (A) Confocal and STED images of migrasomes from live myocardial fibroblast. (Scale bar, 1 μm.) (B) Intensity profiles corresponding to the white dotted line of the STED and confocal images. (C) Confocal and STED images of axons and dendritic spines from a live hippocampal neuron. (Scale bar, 10 μm.) (D) Magnified view of the blue boxed area from C. (Scale bar, 1 μm.) (E) Intensity profiles corresponding to the white dotted line of the STED and confocal images. (F) Time-lapse SIM images (0.5 Hz) of the axon from the live hippocampal neuron stained with PK Mem 555 (Video S13). Mitochondria were stained with PK Mito Deep Red. Anterograde transport (yellow arrow), retrograde transport (red arrow), and stationary/dynamic pause (green arrow) of mitochondria within the axon were indicated. (Scale bar, 2 μm.)