English

• The large Stokes shift fluorescent proteins (LSS-FPs) exhibit the characteristic of wide separation between excitation and emission wavelengths, which facilitates channel separation for multicolor imaging to simultaneously visualize multiple biological events in living cells [1,2]. The use of red fluorescent proteins with a large Stokes shift (LSS-RFPs) for in vivo imaging is ideal due to their longer wavelength, which is less toxic, better penetrate biological tissues, lower spontaneous fluorescence signal from tissue surface [3-5]. A lot of research has been devoted to improving the traditional red fluorescent proteins (RFPs) by inducing the blue shift in excitation spectra or red shift in emission spectra, thereby exhibiting large Stokes shift characteristics. LSSmKate, mBeRFP, LSSmOrange, etc. are typical examples of LSS-RFPs that exhibit the blue shift in excitation spectra [6-9]. They introduced the excited-state proton transfer (ESPT) pathway through random and directed mutagenesis, which endowed the residues near the hydroxyl group of the chromophore with the ability to act as proton acceptors [10,11]. As a result, the absorption wavelength of RFP was blue shifted to 420–450 nm, while the fluorescence emission wavelength remained unchanged [12]. However, blue light excitation typically accompanies higher energy, which can lead to cell damage or toxicity [13,14]. Therefore, the development of large Stokes shift far-red fluorescent proteins with red light absorption and far-red emission is more attractive [5]. Representative examples include TagRFP675 and mPlum, developed through directed evolution [15,16]. These variants maintain the excitation wavelength of RFP at ~580 nm while shifting the emission wavelength to a range of 650–670 nm. Nevertheless, the development of such fluorescent proteins is limited by the requirement for directed evolution and the finite pool of available mutant amino acids [17].

  Compared with the complicated protein engineering in directed evolution, the simple design of coupling a fluorophore to FPs provides a novel approach for the development of LSS-FPs [18-20]. Two fluorophores were connected to generate donor-acceptor pair for Förster resonance energy transfer (FRET), resulting in a large Stokes shift through the excitation of the donor chromophore and the emission of the acceptor chromophore. For instance, Kuhn et al. employed unnatural amino acid (UAA) technology to site-specifically incorporate the fluorescent UAA l-(7-hydroxycoumarin-4-yl)ethylglycine into Tyr39 position of eCFP [21]. The donor coumarin fluorophore and the acceptor eCFP fluorophore generated an efficient FRET pair, so that the hybrid FP show a large Stokes shift about 110 nm. Similar work was also implemented in sfGFP [22]. However, the biosynthesis of UAA is extremely difficult, and its low incorporation efficiency and activity greatly limit their utility [23]. In another notable advancement, Hellweg et al. reported the fusion of EGFP with rhodamine-labeled HaloTag to produce a large Stokes shift FRET biosensor, achieving a red shift of more than 150 nm. This work extended the emission wavelength of green fluorescent proteins [24]. Recently, the FRET strategy based on protein fusion was employed to develop LSS-RFPs with red-shifted emission wavelength. Zhao et al. fused mCherry with the far-red fluorescent protein BDFP1.6 [25]. The donor mCherry efficiently transferred excitation energy to the far-red acceptor BDFP1.6 via FRET, producing the far-red fluorescent protein BDFP2.0 with a large Stokes shift of 79 nm. Nonetheless, there are few types of selectable far-red fluorescent proteins, and their poor fluorescence properties (brightness, photostability, etc.) limit the imaging of LSS-RFPs [26]. Therefore, there remains a necessity for the development of novel LSS-RFPs with red-shifted emission wavelength.

  Small molecule fluorophores have simple structures and are easily modifiable [27-31]. Currently, a variety of small molecule fluorophores with excellent performance and far-red emission (620–700 nm) have been developed [32-34], such as the rhodamine fluorophore tetramethyl-Si-rhodamine (TMSiR) [32,35,36], with maximal λ_ex = 640 nm and λ_em = 662 nm. They exhibit good spectral overlap with RFPs to generate efficient FRET pairs. Moreover, small molecule fluorophores can be specifically labeled to target proteins through genetically encoded protein tags such as Halo/SNAP [37-39], which can be used for fluorescence imaging and functional studies of target proteins in living cells [40]. In this work, we coupled RFPs (mCherry and mApple) with TMSiR using the genetically encoded HaloTag. An effective FRET can be carried out between the two fluorophores by using RFP as the donor and TMSiR as the acceptor, resulting in the large Stokes shift red fluorescent proteins with red-shifted emission wavelength (LSS-mApple, Δλ = 97 nm; LSS-mCherry: Δλ = 75 nm). Importantly, based on the excellent fluorescent properties of the small molecule fluorophore TMSiR, LSS-mApple and LSS-mCherry exhibited significantly improved photostability compared to wild-type RFPs, enabling long-term live cell fluorescence imaging. Finally, LSS-mCherry was used for labeling various organelles and simultaneous imaging with multiple fluorescence channels in the living cells.

  To obtain far-red fluorescent proteins with large Stokes shift and red-shifted emission wavelength, we first choose the widely used red fluorescent proteins mApple and mCherry as donors. The successful FRET requires significant overlap between the emission spectrum of donor and the absorption spectrum of acceptor, along with a high molar extinction coefficient for the acceptor. TMSiR is one of the widely used rhodamine-based dyes in live-cell imaging due to its high brightness and excellent photostability [35,41,42]. Notably, TMSiR not only exhibits a high molar extinction coefficient (141,000 L mol^-1 cm^-1 in ethanol with 0.1% trifluoroacetic acid), but also has a significant spectral overlap between its absorption spectrum and the emission spectra of mApple and mCherry (Fig. S1 in Supporting information). Therefore, TMSiR is very suitable as an acceptor for mApple and mCherry. We employed genetic engineering technology to fuse HaloTag to the C-terminus of RFPs. In addition, we deleted a few amino acids at the end of the junction between RFP and HaloTag (Table S1 in Supporting information) to ensure that the distance between the donor and acceptor fluorophores was less than 10 nm for FRET [43]. The RFP-Halo fusion proteins (hereinafter referred to as LSS-mApple and LSS-mCherry) obtained through this modification were expressed and purified in vitro. SDS-PAGE proved that the purified fusion protein can be covalently labeled with Halo-dyes (Figs. S2 and S3 in Supporting information).

  Furthermore, we coupled TMSiR with the chloroalkane ligand to RFPs through a covalent reaction with HaloTag (Fig. 1a). The Stokes shift was determined by the difference in wavelengths between the excitation of donor RFPs and the emission of acceptor TMSiR (Fig. 1b). LSS-mApple maintained an unchanged maximal excitation wavelength (λ_ex = 565 nm) as that of the donor, but generated the red-shifted emission wavelength (maximal λ_em = 662 nm) to the far-red spectrum region. The Stokes shift of LSS-mApple could reach to 97 nm, which was far greater than the 24 nm Stokes shift of wild-type mApple (Fig. 1c). Similarly, the same strategy can also extend the Stokes shift of mCherry to 75 nm (Fig. 1d). We measured the fluorescence emission spectra of LSS-mApple and LSS-mCherry labeled with TMSiR in vitro. There was a decrease in donor fluorescence intensity and an increase in acceptor fluorescence intensity, indicating FRET occurred (Figs. 1e and f). Among them, the FRET efficiency of LSS-mApple was 80.0% and the FRET efficiency of LSS-mCherry is 83.7%. The good FRET effect will facilitate the subsequent applications of LSS-mApple and LSS-mCherry in multi-color biological imaging.

  Figure 1

  

  Figure 1.  (a) The schematic diagram illustrating the use of FRET strategy to construct far-red fluorescent proteins with large Stokes shift. (b) The Jablonski diagram depicting the processes of FRET. Normalized excitation (EX) and emission (EM) spectra of donor red fluorescent proteins mApple (c) and mCherry (d), along with acceptor dye TMSiR. Arrow texts illustrate the comparison of Stokes shift between wild-type RFPs and LSS-RFPs. (e) Fluorescence emission spectra of LSS-mApple with and without the addition of TMSiR. (f) Fluorescence emission spectra of LSS-mCherry with and without the addition of TMSiR.

  DownLoad: Full-Size Img PowerPoint

  Furthermore, we used LSS-mApple and LSS-mCherry for fluorescence imaging in live mammalian cells. For comparison of FRET, we transiently expressed mApple, and LSS-mApple fused to the mitochondrial membrane protein Tom20 (Fig. 2a), as well as mCherry, and LSS-mCherry fused to human histone protein H2B (Fig. 2c) in HeLa cells, followed by confocal fluorescence imaging. Under the excitation of 561 nm, mApple and mCherry proteins showed barely fluorescence in the FRET emission channel (λ_ex = 561 nm, λ_em = 663–703 nm). After LSS-mApple, and LSS-mCherry were labeled with TMSiR, the fluorescence signals of the donor channel were significantly weakened during imaging. In addition, we further quantitatively analyzed the signals in the donor channel and FRET channel of these two cells, and the results showed that the fluorescence intensity ratio of LSS-mApple and LSS-mCherry was significantly reduced compared with that of the parent RFPs (Figs. 2b and d), indicating that the fluorescence of LSS-mApple and LSS-mCherry in the donor channel was weak and did not affect the co-imaging with donor excitation and emission as channels.

  Figure 2

  

  Figure 2.  (a) Confocal fluorescence images of Tom20-mApple, and Tom20-LSS-mApple (labeled with TMSiR) expressed in HeLa cells, respectively, captured separately in the donor channel (excitation: 561 nm; emission: 580–653 nm) and FRET channel (excitation: 561 nm; emission: 663–703 nm). All channels were adjusted with the same contrast. Scale bar: 10 µm. (b) Fluorescence intensity ratio in the donor channel and FRET channel obtained from confocal fluorescence images described in (a). The graph shows the mean ± SD from 8–15 cells in three independent experiments. (c) Confocal fluorescence imaging of H2B-mCherry, and H2B-LSS-mCherry (labeled with TMSiR) with the same settings as in (a). Scale bar: 10 µm. (d) Fluorescence intensity ratio in the donor channel and FRET channel obtained from confocal fluorescence images described in (c). The graph shows the mean ± SD from 8–15 cells in three independent experiments.

  DownLoad: Full-Size Img PowerPoint

  In addition to possessing characteristics of large Stokes shift, LSS-mApple and LSS-mCherry also exhibit enhanced photostability compared to wild-type RFPs. We conducted in vitro photostability tests on LSS-mApple and LSS-mCherry proteins before and after covalent labeling with TMSiR (Figs. 3a and b). Under continuous irradiation at 561 nm laser, the photostability of the FRET emission (λ_em = 662 nm) generated by LSS-mApple was significantly higher than that of the original mApple emission (λ_em = 590 nm). The fluorescence of mApple had decreased to 23% after 100 min. of photobleaching, while LSS-mApple can still retain 48% of the fluorescence (Fig. 3a). Similarly, LSS-mCherry (λ_em = 662 nm) also exhibited more stable emission compared to the parental protein mCherry (λ_em = 610 nm, Fig. 3b).

  Figure 3

  

  Figure 3.  (a) Trends in maximum emission intensity of mApple (561 nm laser, λ_em = 590 nm), TMSiR-labeled LSS-mApple (561 nm laser, λ_em = 660 nm) and HaloTag-connected TMSiR (640 nm laser, λ_em = 660 nm) under continuous laser irradiation over time. (b) Trends in maximum emission intensity of mCherry (561 nm laser, λ_em = 610 nm), TMSiR-labeled LSS-mCherry (561 nm laser, λ_em = 660 nm) and HaloTag-connected TMSiR (640 nm laser, λ_em = 660 nm) under continuous laser irradiation over time. Each curve represents the average of three independent experiments ± SD. [protein]: 2 µmol/L. The comparative photobleaching dynamics of LSS-mApple (c), LSS-mCherry (d) and their respective parental proteins expressed in HeLa cells were captured through confocal fluorescence imaging under continuous illumination with 561 nm laser (33 mW) over time. Each curve represents the average of three independent experiments. (e) Confocal fluorescence images of LSS-mApple and parental mApple fused to Tom20 in HeLa cells at specified photobleaching time points. (f) Confocal fluorescence images of LSS-mCherry and mCherry fused to H2B in HeLa cells at specified photobleaching time points. Scale bar: 10 µm.

  DownLoad: Full-Size Img PowerPoint

  The improved photostability of LSS-mApple and LSS-mCherry mainly came from the excellent photostability of the introduced Si-rhodamine fluorophore. As a control, we tested the photostability of the TMSiR-labeled HaloTag protein under continuous irradiation at 640 nm laser. However, we found that despite the photostability of LSS-mApple and LSS-mCherry was improved compared to the parent RFPs, it was still not as good as the photostability of TMSiR (Figs. 3a and b). We speculate that this is mainly due to the fact that the FRET efficiency of LSS-mApple and LSS-mCherry does not reach 100%. Incomplete FRET results in a portion of the energy from RFPs transitioning to the triplet state via intersystem crossing (ISC), leading to the quenching of RFPs fluorescence [44,45]. Theoretically, the photostability of LSS-mApple and LSS-mCherry could be consistent with that of TMSiR by employing appropriate strategies to achieve complete FRET, which may fully inhibit the photobleaching of RFPs.

  Next, we further evaluated the photostability of LSS-mApple and LSS-mCherry in living mammalian cells. Using laser scanning confocal microscopy, we performed long-term imaging and photobleaching experiments in HeLa cells expressed LSS-mApple (fused to Tom20) and LSS-mCherry (fused to H2B). Under continuous irradiation at 561 nm laser, the fluorescence imaging of mitochondria labeled with LSS-mApple and nucleus labeled with LSS-mCherry remained visualization for almost 15 min. In contrast, the fluorescence of wild-type mApple and mCherry almost completely disappeared at 9 min (Figs. 3e and f). LSS-mApple exhibited significantly improved imaging photostability compared to parental mApple (Fig. 3c), as evidenced by the increase in photobleaching half-life (t_1/2) from 107.9 s (mApple) to 145.4 s (LSS-mApple). Under the same imaging conditions, we further compared the photobleaching kinetics of mCherry and LSS-mCherry both labeled to H2B protein (Fig. 3d). LSS-mCherry also showed better photostability (t_1/2 = 449.3 s) compared to mCherry (t_1/2 = 342.9 s). These results are consistent with previous photostability measurements in vitro.

  In addition to photostability, we also assessed the pH stability and salt sensitivity of LSS-mApple and LSS-mCherry. Regarding pH stability, LSS-mCherry displayed minimal changes in fluorescence intensity within the physiological pH range (from pH 5.5 to pH 9.0) (Fig. S4c in Supporting information). In contrast, LSS-mApple exhibited a higher sensitivity to pH (Fig. S4a in Supporting information), primarily due to the inherent pH sensitivity of mApple [46]. Furthermore, both proteins demonstrated strong tolerance to salt, with fluorescence intensity remaining virtually unchanged at NaCl concentrations up to 500 mmol/L (Figs. S4b and d in Supporting information). These findings highlight the excellent stability and suitability of LSS-mApple and LSS-mCherry for cell imaging in complex biological environments.

  Furthermore, we fused LSS-mCherry with various organelle proteins to observe its localization in different cellular compartments (Fig. 4). By fusing the N-terminus of LSS-mCherry with human histone H2B and expressing it in HeLa cells, we successfully achieved localized labeling of the nucleus. Similarly, we fused LSS-mCherry to the mitochondrial outer membrane protein Tom20, which resulting in a clear mitochondrial network in the cytoplasm, effectively labeling the mitochondria. Additionally, by fusing the C-terminus of LSS-mCherry with GTPase Rab7 and tail-anchored protein Sec61β, we also observed subcellular localization of LSS-mCherry to the lysosome and endoplasmic reticulum, respectively. We further confirmed the correct subcellular labeling and precise localization of these fusion proteins by co-staining with commercially available organelle-specific dyes during confocal fluorescence imaging (Fig. S5 in Supporting information). This indicated that the fusions formed by LSS-mCherry would not interfere with the localization of most subcellular organelles, and LSS-mCherry was suitable for labeling various structural proteins in mammalian cells.

  Figure 4

  

  Figure 4.  Confocal fluorescence images of LSS-mCherry fusion constructs targeting subcellular locations expressed in HeLa cells. H2B and Tom20 were fused to the N-terminus of LSS-mCherry. Rab7 and Sec61β were fused to the C-terminus of LSS-mCherry. Scale bar: 10 µm.

  DownLoad: Full-Size Img PowerPoint

  The large Stokes shift characteristic of LSS-mCherry enables the possibility of multi-color imaging. We previously validated in confocal fluorescence imaging that LSS-mCherry had little interferential fluorescence into the donor channels (Fig. 2, λ_ex = 561 nm, λ_em = 580–653 nm), which benefited from the introduction of the high FRET effect. This also allowed its strong emission in the far-red channel (λ_ex = 561 nm, λ_em = 663–703 nm) to be separated from the intense emission of traditional RFPs. Therefore, we expected that LSS-mCherry could be combined with traditional RFPs for single-excitation and dual-color emission imaging. To verify this hypothesis, we co-expressed Rab7-mCherry and Tom20-LSS-mCherry proteins in HeLa cells (Fig. 5a). Confocal fluorescence imaging was performed by excitation of a 561 nm laser, and the signals of the two channels were read, respectively. We obtained clear dual-color imaging of lysosomes labeled with mCherry and mitochondria labeled with LSS-mCherry. It demonstrated that the slight background of LSS-mCherry in the donor channel had little impact on the dual-color images. Therefore, LSS-mCherry was highly suitable for dual-color imaging in combination with traditional RFP-like proteins.

  Figure 5

  

  Figure 5.  Multicolor imaging with LSS-mCherry protein. (a) Dual-color confocal image of co-expressing Rab7-mCherry and Tomm20-LSS-mCherry (labled with TMSiR) in HeLa cells. The left panel displays the separation channels of lysosomes (green) and mitochondria (red) under 561 nm excitation. The right panel shows the overlap of both channels. (b) Three-color confocal image of co-expressing Sec61β-eGFP, Rab7-mCherry and Tomm20-LSS-mCherry (labled with TMSiR) in HeLa cells. The left panel displays the separation channels of endoplasmic reticulum (green) under 488 nm excitation, lysosomes (orange) and mitochondria (red) under 561 nm excitation. The right panel shows the overlap of three channels. (c) Four-color confocal image of co-expressing SNAP-Ace2 (labled with Alexa Fluor 488 dye), Rab7-mCherry, Tomm20-LSS-mCherry (labled with TMSiR) along with staining Hoechst 33342 and in HeLa cells. The left panel displays the separation channels of nucleus (blue) under 405 nm excitation, cell membrane (green) under 488 nm excitation, lysosomes (orange) and mitochondria (red) under 561 nm excitation. The right panel shows the overlap of three channels. Scale bar: 10 µm.

  DownLoad: Full-Size Img PowerPoint

  Next, we attempted to further expand to three-color and four-color imaging using LSS-mCherry. We co-expressed Sec61β-eGFP, Rab7-mCherry and Tom20-LSS-mCherry proteins in HeLa cells. Confocal fluorescence imaging was collected in green channel (λ_ex = 488 nm, λ_em = 500–550 nm), red channel (λ_ex = 561 nm, λ_em = 580–653 nm), and far-red channel (λ_ex = 561 nm, λ_em = 663–703 nm), respectively. Under 488 nm and 561 nm excitation, we clearly observed three-color confocal imaging displayed the expected localization of three cellular structures: endoplasmic reticulum labeled with eGFP, lysosomes labeled with mCherry and mitochondria labeled with LSS-mCherry (Fig. 5b). And we noticed almost no crossover between the three channels. Furthermore, we co-expressed SNAP-Ace2, Rab7-mCherry and Tom20-LSS-mCherry proteins in HeLa cells. Then we added the commercial dyes Hoechst 33342 dye and SNAP-Surface Alexa Fluor 488 to stain the nucleus and cell membrane, respectively. Based on the settings of three-color channel, we increased a blue channel (λ_ex = 405 nm, λ_em = 417–477 nm) for four-color confocal imaging. We successfully observed co-imaging of four labeled organelles: cell membrane, nucleus, lysosomes, and mitochondria in HeLa cells (Fig. 5c), all of which matched our expected imaging localization and had no crossover between the all channels.

  The large Stokes shift and good photostability of LSS-mCherry make it very suitable for capturing and analyzing the dynamic changes of biomolecules in living cells. Here, we applied LSS-mCherry to label the mitochondrial Tom20 protein and track the dynamic interactions of mitochondria with various organelles. We first tracked the interaction between mitochondria and the endoplasmic reticulum (eGFP labeled the endoplasmic reticulum Sec61β protein) by confocal dual-color imaging (Fig. 6a). In the yellow box of Fig. 6b, we showed the process of the pink arrowhead mitochondria divided (310–440 s) after rapid contact with the endoplasmic reticulum (300 s). In the red box of Fig. 6c, we observed that the mitochondria indicated by the blue arrows elongated and then shrank after contact with the endoplasmic reticulum during 290–310 s. Subsequently, we observed the appearance of vesicle structures (white arrows) at 400 s in the contact site, and they quickly moved out of the field of view after 410 s. We speculate that this may be because the contact between mitochondria and the endoplasmic reticulum promotes the formation and transport of vesicles, which may be a key process involved in regulating Ca²⁺ [47].

  Figure 6

  

  Figure 6.  (a) Dual-color confocal images of HeLa cells co-expressing GFP-Sec61β and Tom20-LSS-mCherry. The endoplasmic reticulum and mitochondria are represented in green and orange channels, respectively. Scale bar: 10 µm. (b) Magnified images of the yellow-boxed region in (a). The pink arrow indicates the location where endoplasmic reticulum-induced mitochondrial fission event occurred. (c) Magnified images of the red-boxed region in (a). The blue arrows indicate the contact between mitochondria and the endoplasmic reticulum, while the white arrows denote vesicle movement induced by this contact event. (d) Dual-color confocal images of HeLa cells co-expressing mCherry-Rab7 and Tom20-LSS-mCherry. The lysosome and mitochondria are represented in green and red channels, respectively. Scale bar: 5 µm. (e, g) Magnified images of the blue-boxed and green-boxed region in (d). The two groups of images represent the two modes of action of lysosomes in the fission of mitochondria. (f) Magnified images of the yellow-boxed region in (a). (g) The image shows the multiple contact events between lysosomes and mitochondria.

  DownLoad: Full-Size Img PowerPoint

  We then used LSS-mCherry to track the different modes of action between mitochondria and lysosomes (mCherry labeled GTPase Rab7). The blue box of Fig. 6e showed the process of lysosomes participating in mitochondrial fission (190 s) at the white arrow site. In the yellow box of Fig. 6f, we observed repeated contact between the lysosomes indicated by the yellow arrow and different mitochondria. The green box of Fig. 6g showed the dynamic process of multiple lysosomes related to mitochondrial fission. The lysosomes indicated by the white arrows induced the adjacent mitochondria to divide (230 s) after contact (150 s), and pulled the fissioned mitochondria to contact other mitochondria (560 s). After the mitochondria completed the fission process, the lysosomes indicated by the yellow arrows at 490 s and the lysosomes indicated by the blue arrows at 560 s quickly moved to the vicinity of the fission site and stayed there for a while, possibly transporting some important molecules, and then the two lysosomes quickly withdrew from the field of vision after 680 s.

  In this paper, LSS-mApple and LSS-mCherry were designed by connecting TMSiR small molecule fluorophore to mApple and mCherry via genetically encoded HaloTag, respectively. They can apply the excitation wavelength of the donor RFPs to emit far-red fluorescence of acceptor TMSiR, resulting the large Stokes shifts of 97 nm and 75 nm, respectively. LSS-mApple and LSS-mCherry can greatly reduce the fluorescence from the donor channel, which makes them very suitable for combining with traditional RFPs like mCherry in single-excitation and dual-color emission imaging. Particularly, we found that LSS-mApple (t_1/2 = 145.4 s) and LSS-mCherry (t_1/2 = 449.3 s) exhibit significantly improved photostability compared to wild-type RFPs mApple (t_1/2 = 107.9 s) and mCherry (t_1/2 = 342.9 s), which made up for the shortcomings of fluorescent proteins themselves. Furthermore, we proved that LSS-mCherry could perform labeling of various organelles in cells and was an excellent protein tag, while we applied LSS-mCherry to perform multi-color imaging of 2–4 channels, and there was no obvious crosstalk between these channels. These favorable characteristics make LSS-mCherry suitable for observing long-term cellular dynamics. Therefore, we applied LSS-mCherry to demonstrate its dynamic tracking imaging of mitochondria, endoplasmic reticulum and lysosomes. We used LSS-mCherry for dynamic imaging of mitochondria, demonstrating its potential for capturing the dynamic behavior of various biomolecules and organelles. This strategy will facilitate the customization of large Stokes shift fluorescent proteins of various colors, providing a reliable tool for protein labeling and stability monitoring in living cells.

  Declaration of competing interest

  The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

  CRediT authorship contribution statement

  Xuelian Zhou: Writing – original draft, Investigation, Data curation. Lu Miao: Writing – review & editing, Writing – original draft, Supervision. Wei Zhou: Data curation. Qinglong Qiao: Supervision, Funding acquisition. Zhaochao Xu: Writing – review & editing, Supervision, Funding acquisition, Conceptualization.

  Acknowledgments

  This work is supported by the National Natural Science Foundation of China (Nos. 22225806, 22078314, 22278394, 22378385) and Dalian Institute of Chemical Physics (Nos. DICPI202142, DICPI202436).

  Supplementary materials

  Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.cclet.2025.110984.

  
  
• 1. [1]

     T.D. Krueger, C. Chen, C. Fang, Chem. Asian J. 18 (2023) e202300668. doi: 10.1002/asia.202300668

  2. [2]

     H. Kim, J. Ju, H.N. Lee, H. Chun, J. Seong, Sensors 21 (2021) 795. doi: 10.3390/s21030795

  3. [3]

     A. Miyawaki, D.M. Shcherbakova, V.V. Verkhusha, Curr. Opin. Struct. Biol. 22 (2012) 679–688. doi: 10.1016/j.sbi.2012.09.002

  4. [4]

     D.M. Shcherbakova, O.M. Subach, V.V. Verkhusha, Angew. Chem. Int. Ed. 51 (2012) 10724–10738. doi: 10.1002/anie.201200408

  5. [5]

     K.D. Piatkevich, E.N. Efremenko, V.V. Verkhusha, S.D. Varfolomeev, Russ. Chem. Rev. 79 (2010) 243–258. doi: 10.1070/RC2010v079n03ABEH004095

  6. [6]

     T. Kogure, S. Karasawa, T. Araki, et al., Nat. Biotechnol. 24 (2006) 577–581. doi: 10.1038/nbt1207

  7. [7]

     K.D. Piatkevich, V.N. Malashkevich, S.C. Almo, V.V. Verkhusha, J. Am. Chem. Soc. 132 (2010) 10762–10770. doi: 10.1021/ja101974k

  8. [8]

     D.M. Shcherbakova, M.A. Hink, L. Joosen, T.W.J. Gadella, V.V. Verkhusha, J. Am. Chem. Soc. 134 (2012) 7913–7923. doi: 10.1021/ja3018972

  9. [9]

     J. Yang, L. Wang, F. Yang, et al., mBeRFP, PLoS One 8 (2013) e64849. doi: 10.1371/journal.pone.0064849

  10. [10]

      Z. Wang, Y. Zhang, C. Chen, et al., Angew. Chem. Int. Ed. 62 (2023) e202212209. doi: 10.1002/anie.202212209

  11. [11]

      P. Kumar, E. Fron, H. Hosoi, et al., J. Phys. Chem. Lett. 12 (2021) 7466–7473. doi: 10.1021/acs.jpclett.1c01653

  12. [12]

      Y. Shen, Y. Chen, J. Wu, N.C. Shaner, R.E. Campbell, PLoS One 12 (2017) e0171257. doi: 10.1371/journal.pone.0171257

  13. [13]

      J.M. Krassovka, C.V. Suschek, M. Prost, et al., J. Photochem. Photobiol. B 209 (2020) 111952. doi: 10.1016/j.jphotobiol.2020.111952

  14. [14]

      S. Zhou, R. Yamada, K. Sakamoto, PLoS One 18 (2023) e0281062. doi: 10.1371/journal.pone.0281062

  15. [15]

      K.D. Piatkevich, V.N. Malashkevich, K.S. Morozova, et al., Sci. Rep. 3 (2013) 1847. doi: 10.1038/srep01847

  16. [16]

      L. Wang, W.C. Jackson, P.A. Steinbach, R.Y. Tsien, Proc. Natl. Acad. Sci. U. S. A. 101 (2004) 16745–16749. doi: 10.1073/pnas.0407752101

  17. [17]

      A. Acharya, A.M. Bogdanov, B.L. Grigorenko, et al., Chem. Rev. 117 (2017) 758–795. doi: 10.1021/acs.chemrev.6b00238

  18. [18]

      W. Lin, L. Yuan, Z. Cao, Y. Feng, J. Song, Angew. Chem. Int. Ed. 49 (2010) 375–379. doi: 10.1002/anie.200904515

  19. [19]

      M. Zhu, Y. Chen, X. Zhang, et al., Soft Matter 14 (2018) 6737–6744. doi: 10.1039/c8sm01183d

  20. [20]

      J. Han, A. Engler, J. Qi, C.H. Tung, Tetrahedron Lett. 54 (2013) 502–505. doi: 10.1016/j.tetlet.2012.11.060

  21. [21]

      S.M. Kuhn, M. Rubini, M.A. Müller, A. Skerra, J. Am. Chem. Soc. 133 (2011) 3708–3711. doi: 10.1021/ja1099787

  22. [22]

      S.C. Reddington, E.M. Tippmann, D.D. Jones, Chem. Commun. 48 (2012) 8419–8421. doi: 10.1039/c2cc31887c

  23. [23]

      M.A. Shandell, Z. Tan, V.W. Cornish, Biochemistry 60 (2021) 3455–3469. doi: 10.1021/acs.biochem.1c00286

  24. [24]

      L. Hellweg, A. Edenhofer, L. Barck, et al., Nat. Chem. Biol. 19 (2023) 1147–1157. doi: 10.1038/s41589-023-01350-1

  25. [25]

      B. Zhao, W. Ding, Z. Tan, Q. Tang, K. Zhao, ChemBioChem 20 (2019) 1167–1173. doi: 10.1002/cbic.201800695

  26. [26]

      D.M. Shcherbakova, O.V. Stepanenko, K.K. Turoverov, V.V. Verkhusha, Trend. Biotechnol. 36 (2018) 1230–1243. doi: 10.1016/j.tibtech.2018.06.011

  27. [27]

      S. Basu, L.M. Needham, D. Lando, et al., Nat. Commun. 9 (2018) 2520. doi: 10.1038/s41467-018-04486-0

  28. [28]

      J. Chen, W. Liu, X. Fang, Q. Qiao, Z. Xu, Chin. Chem. Lett. 33 (2022) 5042–5046. doi: 10.1016/j.cclet.2022.03.120

  29. [29]

      Z. Li, Q. Qiao, N. Xu, et al., Chin. Chem. Lett. 35 (2024) 108824. doi: 10.1016/j.cclet.2023.108824

  30. [30]

      J. Pan, W. Lin, F. Bao, et al., Chin. Chem. Lett. 34 (2023) 107519. doi: 10.1016/j.cclet.2022.05.033

  31. [31]

      Y. Zhang, W. Zhou, N. Xu, et al., Chin. Chem. Lett. 34 (2023) 107472. doi: 10.1016/j.cclet.2022.04.070

  32. [32]

      H. Ohno, E. Sasaki, S. Yamada, K. Hanaoka, Org. Biomol. Chem. 22 (2024) 3099–3108. doi: 10.1039/d4ob00130c

  33. [33]

      C. Ding, T. Ren, Coord. Chem. Rev. 482 (2023) 215080. doi: 10.1016/j.ccr.2023.215080

  34. [34]

      Z. Khan, N. Sekar, Dye. Pigm. 208 (2023) 110735. doi: 10.1016/j.dyepig.2022.110735

  35. [35]

      J.B. Grimm, B.P. English, J. Chen, et al., Nat. Method 12 (2015) 244–250. doi: 10.1038/nmeth.3256

  36. [36]

      H. Lin, Z. Xu, Chin. Chem. Lett. 33 (2022) 573–574. doi: 10.1016/j.cclet.2021.08.093

  37. [37]

      G.V. Los, L.P. Encell, M.G. McDougall, et al., ACS Chem. Biol. 3 (2008) 373–382. doi: 10.1021/cb800025k

  38. [38]

      A. Keppler, S. Gendreizig, T. Gronemeyer, et al., Nat. Biotechnol. 21 (2003) 86–89. doi: 10.1038/nbt765

  39. [39]

      J. Li, Q. Qiao, Y. Ruan, et al., Chin. Chem. Lett. 34 (2023) 108266. doi: 10.1016/j.cclet.2023.108266

  40. [40]

      A. Cook, F. Walterspiel, C. Deo, ChemBioChem 24 (2023) e202300022. doi: 10.1002/cbic.202300022

  41. [41]

      D. Si, Q. Li, Y. Bao, J. Zhang, L. Wang, Angew. Chem. Int. Ed. 135 (2023) e202307641. doi: 10.1002/ange.202307641

  42. [42]

      W. Liu, J. Chen, Q. Qiao, X. Liu, Z. Xu, Chin. Chem. Lett. 33 (2022) 4943–4947. doi: 10.1016/j.cclet.2022.03.121

  43. [43]

      B. Bajar, E. Wang, S. Zhang, M. Lin, J. Chu, Sensors 16 (2016) 1488. doi: 10.3390/s16091488

  44. [44]

      A.V. Mamontova, A.P. Grigoryev, A.S. Tsarkova, K.A. Lukyanov, A.M. Bogdanov, Russ. J. Bioorg. Chem. 43 (2017) 625–633. doi: 10.1134/s1068162017060085

  45. [45]

      R.B. Vegh, K.B. Bravaya, D.A. Bloch, et al., J. Phys. Chem. B 118 (2014) 4527–4534. doi: 10.1021/jp500919a

  46. [46]

      S.A. Li, X.Y. Meng, Y.J. Zhang, et al., Front. Pharmacol. 14 (2024) 1339518. doi: 10.3389/fphar.2023.1339518

  47. [47]

      J.F. Sambrook, Cell 61 (1990) 197–199. doi: 10.1016/0092-8674(90)90798-J

• 

  Figure 1  (a) The schematic diagram illustrating the use of FRET strategy to construct far-red fluorescent proteins with large Stokes shift. (b) The Jablonski diagram depicting the processes of FRET. Normalized excitation (EX) and emission (EM) spectra of donor red fluorescent proteins mApple (c) and mCherry (d), along with acceptor dye TMSiR. Arrow texts illustrate the comparison of Stokes shift between wild-type RFPs and LSS-RFPs. (e) Fluorescence emission spectra of LSS-mApple with and without the addition of TMSiR. (f) Fluorescence emission spectra of LSS-mCherry with and without the addition of TMSiR.

  下载: 全尺寸图片 幻灯片
  

  Figure 2  (a) Confocal fluorescence images of Tom20-mApple, and Tom20-LSS-mApple (labeled with TMSiR) expressed in HeLa cells, respectively, captured separately in the donor channel (excitation: 561 nm; emission: 580–653 nm) and FRET channel (excitation: 561 nm; emission: 663–703 nm). All channels were adjusted with the same contrast. Scale bar: 10 µm. (b) Fluorescence intensity ratio in the donor channel and FRET channel obtained from confocal fluorescence images described in (a). The graph shows the mean ± SD from 8–15 cells in three independent experiments. (c) Confocal fluorescence imaging of H2B-mCherry, and H2B-LSS-mCherry (labeled with TMSiR) with the same settings as in (a). Scale bar: 10 µm. (d) Fluorescence intensity ratio in the donor channel and FRET channel obtained from confocal fluorescence images described in (c). The graph shows the mean ± SD from 8–15 cells in three independent experiments.

  下载: 全尺寸图片 幻灯片
  

  Figure 3  (a) Trends in maximum emission intensity of mApple (561 nm laser, λ_em = 590 nm), TMSiR-labeled LSS-mApple (561 nm laser, λ_em = 660 nm) and HaloTag-connected TMSiR (640 nm laser, λ_em = 660 nm) under continuous laser irradiation over time. (b) Trends in maximum emission intensity of mCherry (561 nm laser, λ_em = 610 nm), TMSiR-labeled LSS-mCherry (561 nm laser, λ_em = 660 nm) and HaloTag-connected TMSiR (640 nm laser, λ_em = 660 nm) under continuous laser irradiation over time. Each curve represents the average of three independent experiments ± SD. [protein]: 2 µmol/L. The comparative photobleaching dynamics of LSS-mApple (c), LSS-mCherry (d) and their respective parental proteins expressed in HeLa cells were captured through confocal fluorescence imaging under continuous illumination with 561 nm laser (33 mW) over time. Each curve represents the average of three independent experiments. (e) Confocal fluorescence images of LSS-mApple and parental mApple fused to Tom20 in HeLa cells at specified photobleaching time points. (f) Confocal fluorescence images of LSS-mCherry and mCherry fused to H2B in HeLa cells at specified photobleaching time points. Scale bar: 10 µm.

  下载: 全尺寸图片 幻灯片
  

  Figure 4  Confocal fluorescence images of LSS-mCherry fusion constructs targeting subcellular locations expressed in HeLa cells. H2B and Tom20 were fused to the N-terminus of LSS-mCherry. Rab7 and Sec61β were fused to the C-terminus of LSS-mCherry. Scale bar: 10 µm.

  下载: 全尺寸图片 幻灯片
  

  Figure 5  Multicolor imaging with LSS-mCherry protein. (a) Dual-color confocal image of co-expressing Rab7-mCherry and Tomm20-LSS-mCherry (labled with TMSiR) in HeLa cells. The left panel displays the separation channels of lysosomes (green) and mitochondria (red) under 561 nm excitation. The right panel shows the overlap of both channels. (b) Three-color confocal image of co-expressing Sec61β-eGFP, Rab7-mCherry and Tomm20-LSS-mCherry (labled with TMSiR) in HeLa cells. The left panel displays the separation channels of endoplasmic reticulum (green) under 488 nm excitation, lysosomes (orange) and mitochondria (red) under 561 nm excitation. The right panel shows the overlap of three channels. (c) Four-color confocal image of co-expressing SNAP-Ace2 (labled with Alexa Fluor 488 dye), Rab7-mCherry, Tomm20-LSS-mCherry (labled with TMSiR) along with staining Hoechst 33342 and in HeLa cells. The left panel displays the separation channels of nucleus (blue) under 405 nm excitation, cell membrane (green) under 488 nm excitation, lysosomes (orange) and mitochondria (red) under 561 nm excitation. The right panel shows the overlap of three channels. Scale bar: 10 µm.

  下载: 全尺寸图片 幻灯片
  

  Figure 6  (a) Dual-color confocal images of HeLa cells co-expressing GFP-Sec61β and Tom20-LSS-mCherry. The endoplasmic reticulum and mitochondria are represented in green and orange channels, respectively. Scale bar: 10 µm. (b) Magnified images of the yellow-boxed region in (a). The pink arrow indicates the location where endoplasmic reticulum-induced mitochondrial fission event occurred. (c) Magnified images of the red-boxed region in (a). The blue arrows indicate the contact between mitochondria and the endoplasmic reticulum, while the white arrows denote vesicle movement induced by this contact event. (d) Dual-color confocal images of HeLa cells co-expressing mCherry-Rab7 and Tom20-LSS-mCherry. The lysosome and mitochondria are represented in green and red channels, respectively. Scale bar: 5 µm. (e, g) Magnified images of the blue-boxed and green-boxed region in (d). The two groups of images represent the two modes of action of lysosomes in the fission of mitochondria. (f) Magnified images of the yellow-boxed region in (a). (g) The image shows the multiple contact events between lysosomes and mitochondria.

  下载: 全尺寸图片 幻灯片
•