A general strategy for developing cell-permeable photo-modulatable organic fluorescent probes for live-cell super-resolution imaging

Abstract

Single-molecule localization microscopy (SMLM) achieves super-resolution imaging beyond the diffraction limit but critically relies on the use of photo-modulatable fluorescent probes. Here we report a general strategy for constructing cell-permeable photo-modulatable organic fluorescent probes for live-cell SMLM by exploiting the remarkable cytosolic delivery ability of a cell-penetrating peptide (rR)₃ R₂ . We develop photo-modulatable organic fluorescent probes consisting of a (rR)₃ R₂ peptide coupled to a cell-impermeable organic fluorophore and a recognition unit. Our results indicate that these organic probes are not only cell permeable but can also specifically and directly label endogenous targeted proteins. Using the probes, we obtain super-resolution images of lysosomes and endogenous F-actin under physiological conditions. We resolve the dynamics of F-actin with 10 s temporal resolution in live cells and discern fine F-actin structures with diameters of ~80 nm. These results open up new avenues in the design of fluorescent probes for live-cell super-resolution imaging.

ARTICLE

Received 29 Apr 2014 | Accepted 14 Oct 2014 | Published 20 Nov 2014

Deng Pan1,2,*, Zhe Hu1,2,*, Fengwu Qiu1,2,*, Zhen-Li Huang1,2, Yilong Ma1,2, Yina Wang1,2, Lingsong Qin1,2, Zhihong Zhang1,2, Shaoqun Zeng1,2 & Yu-Hui Zhang1,2

Single-molecule localization microscopy (SMLM) achieves super-resolution imaging beyond the diffraction limit but critically relies on the use of photo-modulatable uorescent probes. Here we report a general strategy for constructing cell-permeable photo-modulatable organic uorescent probes for live-cell SMLM by exploiting the remarkable cytosolic delivery ability of a cell-penetrating peptide (rR)3R2. We develop photo-modulatable organic uorescent probes consisting of a (rR)3R2 peptide coupled to a cell-impermeable organic uorophore and a recognition unit. Our results indicate that these organic probes are not only cell permeable but can also specically and directly label endogenous targeted proteins. Using the probes, we obtain super-resolution images of lysosomes and endogenous F-actin under physiological conditions. We resolve the dynamics of F-actin with 10 s temporal resolution in live cells and discern ne F-actin structures with diameters of B80 nm. These results open up new avenues in the design of uorescent probes for live-cell super-resolution imaging.

DOI: 10.1038/ncomms6573 OPEN

A general strategy for developing cell-permeable photo-modulatable organic uorescent probes for live-cell super-resolution imaging

1 Britton Chance Center for Biomedical Photonics, Wuhan National Laboratory for Optoelectronics-Huazhong University of Science and Technology, Wuhan 430074, China. 2 Key Laboratory of Biomedical Photonics of Ministry of Education, Department of Biomedical Engineering, Huazhong University of Science and Technology, Wuhan 430074, China. * These authors contributed equally to this work. Correspondence and requests for materials should be addressed toY.-H.Z. (email: mailto:[email protected]

Web End [email protected] ).

NATURE COMMUNICATIONS | 5:5573 | DOI: 10.1038/ncomms6573 | http://www.nature.com/naturecommunications

Web End =www.nature.com/naturecommunications 1

& 2014 Macmillan Publishers Limited. All rights reserved.

ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/ncomms6573

Recent developments in single-molecule localization microscopy (SMLM), such as photo-activated localization microscopy1,2 and stochastic optical reconstruction

microscopy3, have enabled biological structures to be dened with a spatial resolution beyond the diffraction limit, providing powerful techniques to obtain exciting new insight into the nanoscale structures and dynamics of biological samples4,5. In contrast to deterministic far-eld super-resolution imaging methods, such as stimulated emission depletion6 and structured illumination microscopy7, SMLM does not require sophisticated microscopes but instead critically relies on the use of photo-activatable, photo-convertible or photo-switchable (termed photo-modulatable) uorescent probes8. Organic uorophoreconjugated antibodies used in immunouorescence techniques have been well-established. However, these probes are primarily restricted to xed cells5,9. To specically label intracellular proteins in intact cells for live-cell SMLM, several strategies have been developed, each with their own limitations. One strategy exploits uorescent proteins (FPs) because they are live-cell compatible. However, the generally low-uorescence quantum yield and poor photostability of FPs have resulted in only a few suitable photo-modulatable FPs (for example, photo-activatable green uorescent protein (PAGFP), PAmCherry, PAtagRFP and tdEosFP) being used successfully in live-cell SMLM so far10. Moreover, the overexpression of these FPs may lead to artifacts, such as protein aggregation or inappropriate localization due to saturation of the protein targeting machinery. FP (diameter 34 nm) fusions also substantially increase the size of the protein and can interfere with biological activity11. The second strategy relies on the combination of a genetically encoded target protein (or peptide) with a separate synthetic probe consisting of a photo-modulatable organic uorophore and a recognition unit, such as SNAP-tags12, TMP-tags13 or Halo-tags14. However, for this strategy, both the organic uorophore and the recognition unit of the synthetic probe must be cell permeable for live-cell imaging, which severely restricts the number of the synthetic probes. Moreover, most of photo-modulatable organic uorophores (for example, photo-caged uorophores and Alexa 647) with excellent optical properties cannot be used in this strategy due to their poor cell permeability, particularly after conjugation with a recognition unit15. Furthermore, the large protein tags (for example, SNAP-Tag 20 kDa, eDHFR/TMP-Tag 18 kDa and Halo-Tag 35 kDa) used in these methods can sterically interfere with protein function. Overexpressed fusion proteins rather than endogenous proteins are labelled in these procedures, which may also cause overexpression artifacts.

Here we develop a new general strategy for constructing cell-permeable photo-modulatable organic uorescent probes for live-cell super-resolution imaging by utilizing the remarkable cytosolic delivery ability of a cell-penetrating peptide (CPP), (rR)3R2 (r: D-Arg, R: L-Arg). CPPs are short peptides that can penetrate the cell membrane and translocate linked cell-impermeable cargoes into live cells. Although they are able to enter live cells efciently, most CPPs are often trapped within punctate vesicles and have difculty being released to the cytosol to exert their functions inside cells16. We have recently developed a short CPP (rR)3R2 that can efciently deliver small membrane-impermeable molecules into the cytosol rather than punctate vesicles in live cells17. Based on the excellent properties of (rR)3R2, including low cytotoxicity, ease of synthesis, high uptake efciency and efcient cytosolic delivery17, we designed novel photo-modulatable organic uorescent probes consisting of the CPP (rR)3R2, a organic uorophore (cell impermeable) and a recognition unit (cell impermeable). Our results indicate that these organic probes are not only cell permeable but can also

specically and directly label endogenous proteins. Using these probes, we obtain super-resolution images of F-actin and lysosomes in live cells. Remarkably, we monitor the dynamics of F-actin under physiological conditions with 10 s temporal resolution. To the best of our knowledge, the probes presented here are the rst cell-permeable photo-modulatable organic uorescent probes that can directly label intracellular endogenous targeted proteins in live cells for super-resolution imaging. Considering that the organic uorophore and the recognition unit of the probes can be easily replaced with other organic dyes and recognition groups, respectively, which do not need to be cell permeable anymore, thus greatly expanding the number of organic uorophores and recognition groups that can be used inside live cells for SMLM.

ResultsDesign and synthesis of the organic uorescent probes. We designed and synthesized the organic uorescent probes 16 (Fig. 1, Supplementary Fig. 1) using standard solid-phase peptide synthesis1820. The probes 13 have the following four components: (a) Lifeact (MGVADLIKKFESISKEE), a 17-amino-acid peptide, which can specically and reversibly bind to the F-actin cytoskeleton without compromising cellular processes21;(b) an (rR)3R2 peptide to engender cellular permeability of the probe17; (c) six glycine residues as a spacer and (d) a uorophore (uorescein isothiocyanate (FITC), rhodamine B (RhB) or a photo-activatable rhodamine B derivative (PA-RhB) that has a high uorescence quantum yield, high photochemical stability and a large contrast ratio between the uorescent and the dark state22). We also generated probe 4, which does not contain a CPP, as a control of probe 2 (Fig. 1). Compared with probe 3, probes 5 and 6 contain a different recognition unit, an epoxysuccinyl scaffold, which can selectively form covalent bonds with cysteine cathepsins (typically localized in lysosomes)19,20 and a different uorophore (a caged rhodamine 110 derivative (Caged-Rh110) or 5(6)-carboxyuorescein (FAM))23.

Effect of (rR)3R2 on the labelling of the probes for F-actin. Previous studies showed that the intracellular distribution (for example, diffuse cytosolic distribution or entrapment in punctate vesicles) of CPPs was affected by their cargoes24. Therefore, we rst determined whether the short peptide (rR)3R2 can efciently

Recognition unit Cell-penetrating peptide

MGVADLIKKFESISKEEGGGGK(RhB)GG

MGVADLIKKFESISKEEGGGGK(Ahx-FITC)GGrRrRrRRR

Probe 2

Probe 4

MGVADLIKKFESISKEEGGGGK(RhB)GGrRrRrRRR

Probe 1

Probe 3

MGVADLIKKFESISKEEGGGGK(PA-RhB)GGrRrRrRRR

O OH

O

O

H N

N H

N H

O

O

O O

O

Epoxysuccinyl scaffold

HO

R1

HN NH

N H

O 8

R2

Cell-penetrating peptide

Recognition unit

Probe 5 R1=Caged-Rh110, R2=Arg Probe 6 R1=FAM, R2=Arg

Figure 1 | Structures of constructed probes 16. Ahx, e-aminocaproic acid; FITC, uorescein isothiocyanate; RhB, rhodamine B; PA-RhB, photoactivatable rhodamine B derivative; Caged-Rh110, caged rhodamine 110 derivative; FAM, 5(6)-carboxyuorescein; r, D-Arginine.

2 NATURE COMMUNICATIONS | 5:5573 | DOI: 10.1038/ncomms6573 | http://www.nature.com/naturecommunications

Web End =www.nature.com/naturecommunications

& 2014 Macmillan Publishers Limited. All rights reserved.

NATURE COMMUNICATIONS | DOI: 10.1038/ncomms6573 ARTICLE

transport the recognition peptide (Lifeact) and the dye (FITC or RhB) of probes 1 and 2 into live cells and more importantly into the cytosol rather than punctate vesicles to access F-actin. Trypan blue was used to exclude the dead cells and quench the extracellular uorescence from the probes bound to either the cell membrane or the dish surface25. Figure 2a indicates that after incubation with BSC-1 cells, FITC-labelled probe 1 entered the live cells efciently within 5 min and underwent diffuse localization throughout the cytoplasm. Few punctate signals were observed in Fig. 2 during the time course from 5 to 30 min, suggesting that probe 1 did not enter the cells via endocytosis, which is in good agreement with our previous report17. By contrast, no obvious uorescent signal was detected for probe 4, which did not contain a CPP (Fig. 2c, Supplementary Fig. 2). These results suggest that the CPP (rR)3R2 can efciently deliver the cell-impermeable recognition unit and the dye into the cytosol of live cells. Over time, probe 1 was gradually bound to the lamentous structures (Fig. 2a). After 30 min, both FITC-labelled probe 1 and RhB-labelled probe 2 stained the lamentous structures efciently (Fig. 2a,b), suggesting that altering the uorophore attached to the probe does not inuence its labelling.

The staining of PA-RhB-labelled probe 3 was not determined because PA-RhB is difcult to detect with a confocal microscope.

Next, we investigated the specicity of labelling F-actin with the probes by performing co-localization studies employing both GFP-actin26 and Lifeact-tdEosFP27 as the standard actin markers. The results indicate that RhB-labelled probe 2 co-localized well with the GFPactin in BSC-1 cells (Fig. 3a, Pearsons coefcient:0.837), HeLa cells (Pearsons coefcient: 0.878) and NIH/3T3 cells (Pearsons coefcient: 0.853; Supplementary Fig. 3), suggesting that the conjugation with the polycationic CPP (rR)3R2 has little effect on the high labelling specicity of the recognition unit Lifeact towards actin. These results were further conrmed by the excellent co-localization of RhB-labelled probe 2 with Lifeact-tdEosFP (Fig. 3b, Pearsons coefcient: 0.939). There were some off-target punctate signals detected in Fig. 3, possibly arising from the probes bound to the cell membrane because a careful wash before imaging could decrease the punctate signals.

Super-resolution imaging of F-actin in live cells. Based on the excellent cell permeability and high specicity of probes 1 and 2, we investigated the utility of probes 13 in a live-cell super-resolution microscopy experiment. After incubation with live BSC-1 cells, no photoblinking was observed for FITC-labelled probe 1 (Supplementary Movie 1). Therefore, although FITC has been used for SMLM in xed cells28, it is not suitable for live-cell super-resolution imaging without additives most likely because the endogenous thiols in the live cells, which provide triplet state depletion of FITC, were not sufcient for reversible photoswitching. For RhB-labelled probe 2, although the photoblinking was observed (Supplementary Movie 2), it was difcult to construct a super-resolution image due to the low signal-to-noise ratio of RhB.

The non-uorescent PA-RhB uorophore has been reported to be switched on by ultraviolet light, to be thermally reverted to a dark state, and to be excited by 560-nm laser light, emitting optimally at 600 nm in xed cells22. However, it is unknown whether the live-cell environment changes the properties of PA-RhB and whether the PA-RhB uorophore can be used for live-cell super-resolution imaging. After incubation with live BSC-1 cells for 30 min, PA-RhB-labelled probe 3 was observed to be blinking inside of the cells when irradiated with 561-nm laser light (0.0250.33 kW cm 2) without any ultraviolet light activation (Supplementary Movie 3). Moreover, no obvious photobleaching of probe 3 occurred within at least 30 min, even at the maximum intensity (0.33 kW cm 2) of 561-nm laser radiation without additional photoactivation (Supplementary

Fig. 4). By contrast, in BSC-1 cells, obvious photobleaching was observed for Lifeact-tdEosFP after 4 min (B8,000 frames) at an excitation power of 0.165 kW cm 2 (561 nm) with an ultraviolet light activation (405 nm) (Supplementary Fig. 4). The PA-RhB uorophore has been shown to be spontaneously activated without photo-induced activation for densely labelled samples22. Because probe 3 is a small-molecule organic probe, it can bind to the target protein with little steric hindrance, which may result in high density labelling of actin; this high density labelling most likely led to obvious spontaneous activation. By contrast, in xed cells, only the rst 1001,000 frames could be visualized for the PA-RhB-labelled tubulin antibody without photoactivation using the limited spontaneously activated molecules22 possibly because the labelling density29 of the antibody is much lower than that of the small-molecule organic probe 3. Previous reports claimed that the low afnity of the recognition peptide Lifeact for F-actin in live cells allowed for the replenishment of bleached uorophores and led to rapid uorescence recovery of the probe after photobleaching21,30. However, the photobleaching of

Probe 1 5 min

15 min

10 min

Probe 2 Probe 4

30 min

Figure 2 | The intracellular distribution of probes 1, 2 and 4 in live BSC-1 cells. Confocal microscopy images of live BSC-1 cells after incubation with (a) FITC-labelled probe 1 (15 mM) for 5, 10, 15 and 30 min,(b) RhB-labelled probe 2 (15 mM) for 30 min and (c) probe 4 (15 mM)

for 30 min. Scale bars, 10 mm.

NATURE COMMUNICATIONS | 5:5573 | DOI: 10.1038/ncomms6573 | http://www.nature.com/naturecommunications

Web End =www.nature.com/naturecommunications 3

& 2014 Macmillan Publishers Limited. All rights reserved.

ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/ncomms6573

Probe 2

Probe 2

GFP-actin Merge

Merge

Lifeact-tdEos

Figure 3 | Co-localization studies employing GFPactin or Lifeact-tdEosFP as the standard actin markers. Live BSC-1 cells transiently transfected with (a) GFPactin (green) or (b) Lifeact-tdEosFP (green) were stained with RhB-labelled probe 2 (red, 15 mM) for 30 min and imaged by confocal microscopy. Scale bars, 10 mm.

Lifeact-tdEosFP after 4 min was still obvious, whereas for PA-RhB-labelled probe 3, the photobleaching was not observed even at 30 min, suggesting that the anti-photobleaching characteristic of probe 3 is most likely due to both the excellent photostability of the PA-RhB uorophore and the low afnity of Lifeact for F-actin.

To determine the best imaging conditions for probe 3 in live cells, we rst evaluated the effect of the excitation power. The results indicate that increasing the excitation power (561 nm) from 6 mW (0.025 kW cm 2) to 40 mW (0.165 kW cm 2) led to an obvious improvement of signal intensity (Supplementary

Fig. 5). No signicant differences were observed between the images acquired at 40 mW (0.165 kW cm 2) or 80 mW(0.33 kW cm 2; Supplementary Fig. 5). Considering that higher excitation powers cause more photodamage, we performed our following experiments with an excitation power of 40 mW. Second, we acquired images at different frame acquisition rates (100, 83, 66 and 58 Hz). The results show that slower frame rates yielded higher signal densities (Supplementary Fig. 5). When reconstructing the SMLM images with the Parallel Localization of Multiple Emitters via Bayesian information criterion Recommendation (PALMER) algorithm for high density localization31, we found that the signals at 83, 66 and 58 Hz were too dense to be accurately localized. We therefore acquired images consistently at 100 Hz with an excitation power of 0.165 kW cm 2 (561 nm).

Then we reconstructed SMLM images with 200, 500, 1,000 and 2,000 frames. The results indicate that 1,000 frames (10 s temporal resolution) were sufcient to obtain a smooth and continuous super-resolution image of actin laments (Fig. 4, Supplementary Fig. 6). Compared with the diffraction-limited image, a much higher spatial resolution was obtained (Fig. 4a), and remarkably the ne F-actin structures with diameters of B80 nm could be discerned (Fig. 4b). These results indicate that our PA-RhB-labelled probe 3 can be used for super-resolution

imaging of endogenous actin in live cells with high spatiotemporal resolution under physiological conditions without needing any additive or additional photoactivation.

New insight into the dynamic behaviours of F-actin. F-actin plays key roles in many important cellular processes such as regulation, mechanosensation, vesicle and organelle movement, cell motility and cell division32,33. Filaments in the network are relatively unstable, undergoing continuous changes, such as polymerization and depolymerization, bending and twisting, and severing and branching, enabling the network to remodel itself in response to external or internal stimuli34. However, the diameter of F-actin is far o100 nm and therefore difcult to observe using ordinary light microscopy.

Based on the excellent optical properties of PA-RhB-labelled probe 3, we tried to monitor the dynamic behaviours of F-actin in live cells. We acquired 15,000 images at 100 Hz with a total acquisition time of 2.5 min (excitation power: 0.165 kW cm 2 (561 nm)). By rendering images from sub-stacks of 1,000 frames, a series of time-lapse super-resolution images were obtained (Fig. 4c) at a time-series interval of 10 s. Although most actin laments remained still during the acquisition time, remarkably, these time-lapse super-resolution images reveal a dynamic procedure in which an actin bundle rearranged its connection to the network (Fig. 4c). The following is one possible interpretation of the dynamic procedure. The actin bundle was severed (20 s), then formed a loose twist structure made from two laments (40 s, Fig. 4d), and nally extended the severed actin bundle establishing a new connection with another adjacent F-actin (140 s). To the best of our knowledge, this is the rst report of the dynamics of F-actin in live cells with nanoscale spatial resolution and 10 s temporal resolution. We also recorded time-lapse super-resolution images of actin dynamics in lopodia in live cells (Supplementary Movie 4).

4 NATURE COMMUNICATIONS | 5:5573 | DOI: 10.1038/ncomms6573 | http://www.nature.com/naturecommunications

Web End =www.nature.com/naturecommunications

& 2014 Macmillan Publishers Limited. All rights reserved.

NATURE COMMUNICATIONS | DOI: 10.1038/ncomms6573 ARTICLE

1

0.8

Normalized intensity

0.6

0.4

0.2

83 nm

0 0

250 300

50 100 150 200 Distance (nm)

Distance (nm)

Normalized intensity

10.80.60.4

0.20 0 50 100 150 200 250 300

0 s 10 s 20 s 30 s 40 s 50 s 80 s 110 s 140 s

Figure 4 | Reconstructed images of actin laments in a live BSC-1 cell stained with PA-RhB-labelled probe 3. (a) Diffraction-limited TIRF (upper right) and super-resolution (lower left) images. (b) Enlarged super-resolution image of the blue-boxed region in a and cross-sectional proles of the actin laments with full-width at half-maximum (FWHM) of 83 nm. (c) Enlarged super-resolution time-lapse images of the yellow-boxed region in a.

Each image was reconstructed from 1,000 frames at 100 Hz. (d) Enlarged image of the boxed region in c at 40 s. Scale bars, (a) 5 mm, (b) 200 nm and (c) 500 nm.

Super-resolution imaging of lysosomes in live cells. To evaluate the generality and expanded utility of our strategy, we also constructed another organic uorescent probe 5 (Fig. 1), which contains a different recognition unit and a uorophore from probe 3. The recognition unit of probe 5 is an epoxysuccinyl scaffold, which can selectively form covalent bonds with cysteine cathepsins (typically localized in lysosomes)19,20. We used Caged-Rh110 instead of PA-RhB in probe 5 because Caged-Rh110 is stable under acidic conditions23 (pHE5 in lysosomes), while

PA-RhB is pH sensitive22. The non-uorescent Caged-Rh110 can be uncaged with a 405-nm laser light and excited at 473 nm, emitting optimally at 520 nm23.

Previous reports have shown that the probes containing the epoxysuccinyl scaffold can specically label lysosomes19,20,35, which was further conrmed by the excellent co-localization of LysoTracker Red with probe 6 (Pearsons coefcient: 0.850), a control of probe 5, which contains a FAM for observing with an ordinary light microscopy (Supplementary Fig. 7). To obtain the super-resolution images of lysosomes, live HeLa cells stained with probe 5 were rstly exposed to a high level of 473 nm light(0.38 kW cm 2) to photobleach any previously uncaged molecules. Then, images were acquired under continuous illumination at 473 nm (0.38 kW cm 2) at a frame rate of 260 Hz with an ultraviolet light intermittent activation (405 nm).

We reconstructed super-resolution images of lysosomes from 500

frames (a total acquisition time of 3.8 s) using the PALMER algorithm31 and obtained a super-resolution image of lysosomes with a full-width at half-maximum of 64 nm (Fig. 5). A movie of the live-cell imaging of lysosomes stained with probe 5 was shown in Supplementary Movie 5. These results demonstrate that Caged-Rh110-labelled probe 5 can be used to image lysosomes with a resolution beyond optical diffraction limit in live cells.

DiscussionOrganic photo-modulatable uorophores offer several important photophysical advantages over FPs, including smaller sizes, broader spectral coverage, higher uorescence quantum yields and better photostabilities; therefore, use of organic uorophores could potentially improve the spatial and temporal resolution of SMLM15. Although various photo-modulatable organic uorophores with excellent optical properties have been reported9, necessary cell permeability, particularly after conjugation with a recognition unit, severely restricts the application of these organic uorophores in live-cell SMLM. To address these problems, the general strategy in previous reports has involved modifying the structures of organic uorophores or the recognition unit chemically to improve their permeability, which is expensive and time consuming and, in particular, might eliminate the photo-modulatable properties of organic uorophores or decrease the afnity of the recognition unit.

NATURE COMMUNICATIONS | 5:5573 | DOI: 10.1038/ncomms6573 | http://www.nature.com/naturecommunications

Web End =www.nature.com/naturecommunications 5

& 2014 Macmillan Publishers Limited. All rights reserved.

ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/ncomms6573

1

0.9

0.8

0.7

0.6

0.5

0.4

0.3

Normalized intensity

0.2

0.1

0200 300 400 500 600 700 800Distance (nm)

64 nm

900

Figure 5 | Reconstructed images of lysosomes in a live HeLa cell stained with Caged-Rh110-labelled probe 5. (a) Diffraction-limited TIRF.(b) Super-resolution image. (c) Enlarged TIRF image of the boxed region in a. (d) Enlarged super-resolution image of the boxed region in b. Super-resolution image was reconstructed from 500 frames at 260 Hz. (e) Cross-sectional proles of lysosome shown in c,d (diffraction-limited, blue line; super-resolved, red line with a full-width at half-maximum (FWHM) of 64 nm). Scale bars, (b) 2 mm and (d) 200 nm.

Here we develop a new general strategy for constructing cell-permeable photo-modulatable organic uorescent probes for live-cell super-resolution imaging by exploiting the remarkable cytosolic delivery ability of (rR)3R2. We construct two novel cell-permeable photo-modulatable organic uorescent probes, PA-RhB-labelled probe 3 and Caged-Rh110-labed probe 5, which contain different recognition units and uorophores. Our results indicate that both of the probes can enter live cells efciently and directly label endogenous targeted proteins for super-resolution imaging. Using the probes, we obtain super-resolution images of lysosomes (B 60 nm) and endogenous F-actin (B 80 nm) under physiological conditions.

Remarkably, PA-RhB-labelled probe 3 can be excited by a single laser light under physiological conditions without other additives, such as redox chemicals required for most other organic probes28 or additional photoactivation, thereby greatly reducing the complexity of the experiment and the phototoxicity caused by the second ultraviolet light activation. Moreover, PA-RhB-labelled probe 3 has an excellent anti-photobleaching property, making it potentially very valuable for long-term real-time super-resolution imaging.

Resolving the fast dynamics of actin networks requires both high spatial resolution and temporal resolution. Although the photo-modulatable FP Lifeact-tdEosFP has been used successfully for live-cell SMLM, the highest temporal resolution that can be achieved by Lifeact-tdEosFP was 50 s (ref. 30). Applying higher activation laser intensities leads to higher emitter densities, which can improve the temporal resolution; however, this application also results in the quick photobleaching of Lifeact-tdEosFP. Alternatively, using PA-RhB-labelled probe 3, we resolve the dynamics of F-actin with 10 s temporal resolution in live cells,

and the ne F-actin structures with diameters of B80 nm could be discerned, thereby allowing us to follow the dynamics with high spatiotemporal resolution. Our results reveal a connection rearrangement procedure of an actin bundle, which has never been previously reported in living cells, therefore providing new insights into the dynamic behaviours of F-actin in vivo.

Taken together, we anticipate that probes 3 and 5 will be a useful tool for super-resolution imaging and may substantially benet ultrastructural characterizations of live cells. Notably, the easy replacement of the organic uorophore and the recognition unit of probes 3 and 5 with other cell-impermeable organic dyes and recognition groups (for example, peptides and protein substrates) will greatly expand the number of organic dyes and recognition groups that can be used inside live cells for SMLM. The results here provide a new general strategy for developing cell-permeable organic uorescent probes for live-cell super-resolution imaging. Studies using other dyes (for example, Alexa 647 and Cy3B) and recognition units (for example, phalloidin36, docetaxel37, SNAP-tag12 and TMP-tag13) to construct new cell-permeable probes for live-cell SMLM are currently underway.

Methods

Materials. Probes 16 were prepared by solid-phase peptide synthesis, puried by preparative high performance liquid chromatography to reach a purity 495%, and their appropriate masses were conrmed by electrospray ionization mass spec-trometry (Supplementary Fig. 1).

Cell culture. BSC-1 cells (African green monkey (Cercopithecus aethiops) kidney cells), HeLa and NIH/3T3 (Mouse embryonic broblast cell line) cells were purchased from Boster Biological Technology Co. Ltd, Wuhan, China. BSC-1 cells were cultured in Minimum Essential Medium (MEM; Gibco BRL, Grand Island, NY, USA). HeLa and NIH/3T3 cells were cultured in Dulbeccos modied Eagles

6 NATURE COMMUNICATIONS | 5:5573 | DOI: 10.1038/ncomms6573 | http://www.nature.com/naturecommunications

Web End =www.nature.com/naturecommunications

& 2014 Macmillan Publishers Limited. All rights reserved.

NATURE COMMUNICATIONS | DOI: 10.1038/ncomms6573 ARTICLE

medium (DMEM; Gibco BRL). All media were supplemented with 10% foetal bovine serum (FBS; Gibco BRL) and maintained at 37 C in a humidied 5% CO2 environment.

Transfection. BSC-1, HeLa and NIH/3T3 cells were grown overnight in 24-well plates at 37 C in a 5% CO2 atmosphere. After reaching B80% conuence, the GFPactin plasmid or the Lifeact-tdEosFP plasmid was transfected into the cells using Lipofectamine 2000 (Invitrogen, Grand Island, NY, USA) according to the manufacturers instructions. After 24 h, the transfected cells were digested with trypsin (0.25% EDTA) and seeded into cell culture dishes (glass bottom F15 mm,

NEST Biotechnology Co., Ltd., China) at a density of 1.5 104 per well. The cells

were grown for an additional 12 24 h before incubation with the indicated probe.

Labelling live cells with the probes. For F-actin imaging, BSC-1 cells were seeded into cell culture dishes at a density of 1.5 104 per well in growth medium (MEM

supplemented with 10% FBS, 200 ml). After an overnight incubation, the cells were washed with phosphate-buffered saline (PBS, pH 7.4) for three times. A solution of the indicated probe (15 mM, 200 ml) in PBS was then added, and the cells were incubated in a 5% CO2 atmosphere at 37 C for the indicated (5, 10, 15 and 30 min) times as mentioned in the main text. The supernatant was then discarded, and a solution of Trypan blue (200 ml, 1 mg ml 1) in PBS was added to quench the extracellular uorescence of the probes that were bound to either the cell membrane or the dish surface. After 2 min, Trypan blue was removed, the cells were washed gently twice with PBS and immersed in growth medium prior to optical imaging.

For lysosome imaging, HeLa cells were seeded into cell culture dishes at a density of 2.0 104 per well in growth medium (DMEM supplemented with 10%

FBS, 200 ml). The cells were then grown at 37 C in a 5% CO2 atmosphere overnight. After washing with PBS for three times, the cells were incubated with probe 5 (6 mM, 200 ml) or probe 6 (15 mM, 200 ml) in PBS for 30 min in a 5% CO2 atmosphere at 37 C. The supernatant was discarded and the cells were rinsed with PBS for three times and incubated in growth medium (DMEM supplemented with 10% FBS, 200 ml) for 5 h. To stain lysosomes, the cells were incubated with

LysoTracker Red (500 nM, 200 ml) for 30 min before imaging. Then the cells were washed twice gently with PBS and immersed in growth medium prior to optical imaging.

Confocal laser scanning microscopy. The uorescence signals were detected using a Fluoview FV1000 confocal laser scanning microscope (Olympus) equipped with a 60/1.42 numerical aperture oil-immersion objective lens (FITC channel:

excitation (EX) 488 nm, emission (EM) 510530 nm; GFP channel: EX 488 nm, EM 508532 nm; tdEosFP channel: EX 488 nm, EM 500522 nm; RhB channel: EX 543 nm, EM 580630 nm; FAM channel: EX 488 nm, EM 505540 nm and LysoTracker Red channel: excitation (EX) 543 nm, emission (EM) 560580 nm). All uorescence images were analyzed with Image J software.

Single-molecule localization-based super-resolution imaging. The single-molecule imaging experiments were performed with a laboratory-built total internal reection uorescence (TIRF) microscope setup consisting of an Olympus IX 71 inverted optical microscope, a 100/numerical aperture1.49 oil-immersion

TIRF objective (UAPON 100XOTIRF, Olympus), an iXon 897 EMCCD camera (Andor Tech.), a 561-nm diode-pumped solid-state laser (CNILaser, China) for PA-RhB excitation, a 473-nm diode-pumped solid-state laser (CNILaser) for Caged-Rh110 excitation and a 405-nm laser diode (CNILaser) for caged-Rh110 activation. An electronic shutter (UNIBLITZ VS14, Vincent Associates) was used to control the duration of laser irradiance, and a dichroic mirror (Di01-R488/561, Semrock) and long-pass lter (BLP01561R-25, Semrock) were used to separate the collected uorescence from the scattering laser and impurity uorescence31. For live-cell F-actin imaging, a wavelength of 561 nm with an intensity of B0.165 kW cm 2 was used to record movies at a frame rate of 100 Hz. The data were acquired with the software that was provided by the manufacturer of the camera. For live-cell lysosome imaging, live HeLa cells stained with probe 5 were rstly exposed to a 473 nm light (0.38 kW cm 2) to photobleach any previously uncaged molecules. Images were acquired under continuous illumination at473 nm (0.38 kW cm 2) at a frame rate of 260 Hz with an ultraviolet light intermittent activation (405 nm, 5.523 W cm 2).

Data analysis. A fast, high-precision image analysis method for high density localization, termed PALMER, was used to reconstruct the super-resolved images31. Subsequent rendered images of F-actin were generated from 1,000 video sequences, with no overlap between two adjacent reconstructed images, that is, sequence 11,000, 1,0012,000, 2,0013,000 and so on. Super-resolution images of lysosomes were reconstructed from 500 frames.

References

1. Betzig, E. et al. Imaging intracellular uorescent proteins at nanometer resolution. Science 313, 16421645 (2006).

2. Hess, S. T., Girirajan, T. P. K. & Mason, M. D. Ultra-high resolution imaging by uorescence photoactivation localization microscopy. Biophys. J. 91, 42584272 (2006).

3. Rust, M. J., Bates, M. & Zhuang, X. Sub-diffraction-limit imaging by stochastic optical reconstruction microscopy (STORM). Nat. Methods 3, 793796 (2006).

4. Bates, M., Huang, B. & Zhuang, X. Super-resolution microscopy by nanoscale localization of photo-switchable uorescent probes. Curr. Opin. Chem. Biol. 12, 505514 (2008).

5. van de Linde, S., Heilemann, M. & Sauer, M. Live-cell super-resolution imaging with synthetic uorophores. Annu. Rev. Phys. Chem. 63, 519540 (2012).

6. Hell, S. W. & Wichmann, J. Breaking the diffraction resolution limit by stimulated emission: stimulated-emission-depletion uorescence microscopy. Opt. Lett. 19, 780782 (1994).

7. Gustafsson, M. G. L. Surpassing the lateral resolution limit by a factor of two using structured illumination microscopy. J. Microsc. 198, 8287 (2000).8. Henriques, R., Grifths, C., Hesper Rego, E. & Mhlanga, M. M. PALM and STORM: unlocking live-cell super-resolution. Biopolymers 95, 322331 (2011).

9. Patterson, G., Davidson, M., Manley, S. & Lippincott-Schwartz, J. Superresolution imaging using single-molecule localization. Annu. Rev. Phys. Chem. 61, 345367 (2010).

10. van de Linde, S. et al. Investigating cellular structures at the nanoscale with organic uorophores. Chem. Biol. 20, 818 (2013).

11. Giepmans, B. N. G., Adams, S. R., Ellisman, M. H. & Tsien, R. Y. The uorescent toolbox for assessing protein location and function. Science 312, 217224 (2006).

12. Klein, T. et al. Live-cell dSTORM with SNAP-tag fusion proteins. Nat. Methods 8, 79 (2011).

13. Wombacher, R. et al. Live-cell super-resolution imaging with trimethoprim conjugates. Nat. Methods 7, 717719 (2010).

14. Lee, H. L. D. et al. Superresolution imaging of targeted proteins in xed and living cells using photoactivatable organic uorophores. J. Am. Chem. Soc. 132, 1509915101 (2010).

15. Jones, S. A., Shim, S.-H., He, J. & Zhuang, X. Fast, three-dimensional super-resolution imaging of live cells. Nat. Methods 8, 499505 (2011).

16. Mueller, J., Kretzschmar, I., Volkmer, R. & Boisguerin, P. Comparison of cellular uptake using 22 CPPs in 4 different cell lines. Bioconjug. Chem. 19, 23632374 (2008).

17. Ma, Y. et al. Direct cytosolic delivery of cargoes in vivo by a chimera consisting of D- and L-arginine residues. J. Control. Release. 162, 286294 (2012).

18. Keliris, A. et al. Synthesis and characterization of a cell-permeable bimodal contrast agent targeting b-galactosidase. Bioorg. Med. Chem. 19, 25292540 (2011).

19. Fan, F. et al. Labeling lysosomes and tracking lysosome-dependent apoptosis with a cell-permeable activity-based probe. Bioconjug. Chem. 23, 13091317 (2012).

20. Fan, F. et al. Protein proling of active cysteine cathepsins in living cells using an activity-based probe containing a cell-penetrating peptide. J. Proteome Res. 11, 57635772 (2012).

21. Riedl, J. et al. Lifeact: a versatile marker to visualize F-actin. Nat. Methods 5, 605607 (2008).

22. Flling, J. et al. Photochromic rhodamines provide nanoscopy with optical sectioning. Angew. Chem. Int. Ed. 46, 62666270 (2007).

23. Tunnemann, G. et al. Cargo-dependent mode of uptake and bioavailability of TAT-containing proteins and peptides in living cells. ACS Chem. Biol. 7, 289293 (2012).

24. Sambashiva, B., Damien, M., Suliana, M. & Kai, J. A caged, localizable rhodamine derivative for superresolution microscopy. FASEB J. 20, 17751784 (2006).

25. Manceur, A., Wu, A. & Audet, J. Flow cytometric screening of cell-penetrating peptides for their uptake into embryonic and adult stem cells. Anal. Chem. 364, 5159 (2007).

26. Ning, W. et al. Effects of GSM 1800 MHz on dendritic development of cultured hippo-campal neurons. Acta Pharmacol. Sin. 28, 18731880 (2007).

27. Hedde, P. N., Fuchs, J., Oswald, F., Wiedenmann, J. & Nienhaus, G. U. Online image analysis software for photoactivation localization microscopy. Nat. Methods 6, 689690 (2009).

28. Dempsey, G. T., Vaughan, J. C., Chen, K. H., Bates, M. & Zhuang, X. Evaluation of uorophores for optimal performance in localization-based super-resolution imaging. Nat. Methods 8, 10271036 (2011).

29. Sahl, S. J. & Moerner, W. E. Super-resolution uorescence imaging with single molecules. Curr. Opin. Struct. Biol. 23, 778787 (2013).

30. Izeddin, I. et al. Super-resolution dynamic imaging of dendritic spines using a low-afnity photoconvertible actin probe. PLoS ONE 6, e15611 (2011).

31. Wang, Y., Quan, T. W., Zeng, S. Q. & Huang, Z. L. PALMER: a method capable of parallel localization of multiple emitters for high-density localization microscopy. Opt. Express 20, 1603916049 (2012).

NATURE COMMUNICATIONS | 5:5573 | DOI: 10.1038/ncomms6573 | http://www.nature.com/naturecommunications

Web End =www.nature.com/naturecommunications 7

& 2014 Macmillan Publishers Limited. All rights reserved.

ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/ncomms6573

32. Schoenenberger, C. A., Mannherz, H. G. & Jockusch, B. M. Actin: from structural plasticity to functional diversity. Eur. J. Cell Biol. 90, 797804 (2011).

33. Pollard, T. D. & Cooper, J. A. Actin, a central player in cell shape and movement. Science 326, 12081212 (2009).

34. Pollard, T. D. & Borisy, G. G. Cellular motility driven by assembly and disassembly of actin laments. Cell 112, 453465 (2003).

35. Greenbaum, D., Medzihradszky, K., Burlingame, A. & Bogyo, M. Epoxide electrophiles as activity-dependent cysteine protease proling and discovery tools. Chem. Biol. 7, 569581 (2000).

36. Wulf, E., Deboben, A., Bautz, F. A., Faulstich, H. & Wieland, T. Fluorexcent phallotoxin, a tool for the visualization of cellular actin. Proc. Natl Acad. Sci. USA 76, 44984502 (1979).

37. Lukinavicius, G. et al. uorogenic probes for live-cell imaging of the cytoskeleton. Nat. Methods 7, 731733 (2014).

Acknowledgements

We thank Prof Jian-hong Luo from the Zhejiang University School of Medicine for providing plasmid of GFPactin and Prof Gerd Ulrich Nienhaus from the Karlsruhe Institute of Technology for providing plasmid of Lifeact-tdEosFP. We also thank the Optical Bioimaging Core Facility of WNLO-HUST for the support in data acquisition, and the Analytical and Testing Center of HUST for spectral measurements. This work was supported by the National Basic Research Programme of China (Grant No. 2015CB352002), the National Natural Science Foundation of China (Grant No. 31270906, 31470056 and 61121004) and the National Natural Science Foundation of China Director Fund of WNLO.

Author contributions

Experiments were performed primarily by D.P., Z.H., and F.Q. with regular input and guidance from Y.-H.Z. and Z.-L.H. Y.M. and F.Q. contributed to the synthesis of the probes. D.P., Z.H., Y.W., L.Q. Z.Z. S.Z. and Y.-H.Z. analyzed the data. Experimental strategy was designed by Y.-H.Z. The manuscript was written by Y.-H.Z. and D.P. with discussion and improvements from all authors.

Additional information

Supplementary Information accompanies this paper at http://www.nature.com/naturecommunications

Web End =http://www.nature.com/ http://www.nature.com/naturecommunications

Web End =naturecommunications

Competing nancial interests: The authors declare no competing nancial interests.

Reprints and permission information is available online at http://npg.nature.com/reprintsandpermissions/

Web End =http://npg.nature.com/ http://npg.nature.com/reprintsandpermissions/

Web End =reprintsandpermissions/

How to cite this article: Pan, D. et al. A general strategy for developing cell-permeable photo-modulatable organic uorescent probes for live-cell super-resolution imaging. Nat. Commun. 5:5573 doi: 10.1038/ncomms6573 (2014).

This work is licensed under a Creative Commons Attribution 4.0 International License. The images or other third party material in this article are included in the articles Creative Commons license, unless indicated otherwise in the credit line; if the material is not included under the Creative Commons license, users will need to obtain permission from the license holder to reproduce the material. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/

Web End =http://creativecommons.org/licenses/by/4.0/

8 NATURE COMMUNICATIONS | 5:5573 | DOI: 10.1038/ncomms6573 | http://www.nature.com/naturecommunications

Web End =www.nature.com/naturecommunications

& 2014 Macmillan Publishers Limited. All rights reserved.