ABSTRACT Bacterial phytochromes are attractive molecular templates for engineering fluorescent proteins (FPs) because their near-infrared (NIR) emission significantly extends the spectral

coverage of GFP-like FPs. Existing phytochrome-based FPs covalently bind heme-derived tetrapyrrole chromophores and exhibit constitutive fluorescence. Here we introduce Rep-miRFP, an NIR 

imaging probe derived from bacterial phytochrome, which interacts non-covalently and reversibly with biliverdin chromophore. In Rep-miRFP, the photobleached non-covalent adduct can be

replenished with fresh biliverdin, restoring fluorescence. By exploiting this chromophore renewal capability, we demonstrate NIR PAINT nanoscopy in mammalian cells using Rep-miRFP. SIMILAR

SUPER-RESOLUTION IMAGING Article Open access 11 May 2021 INTRODUCTION Super-resolution microscopy allows the observation of subcellular architectures beyond the diffraction limit of light.

Among nanoscopy techniques, single-molecule localization microscopy (SMLM) offers high spatial resolution with simple optics and low hardware cost1. In SMLM, fluorophores are stochastically

activated and sampled on single-molecule basis. Photoconvertible and photoswitchable fluorescent proteins (FPs) are the workhorses of SMLM tags, but typically require a total of 2–3 lasers

for excitation, activation, and depletion of the fluorescent species2,3. The high spectral occupancy complicates the optics and limits multiplexing potential. We previously developed SPOON,

a green FP that spontaneously blinks under single-laser illumination4. SPOON is simultaneously excited and switched off by 488 nm light at high irradiance. Off-state SPOON molecules rapidly

and thermally relax back to the fluorescent state, therefore bypasses photoswitching and streamlines SMLM. Alternatively, single-laser SMLM has been achieved by transient labeling using

exchangeable probes5,6. A prominent example is Point Accumulation for Imaging in Nanoscale Topography (PAINT) and its variants, which rely on stochastic fluorophore binding to generate

localizable single molecules7,8,9. Recently, FP-based PAINT was demonstrated using the bilirubin-activating protein UnaG10 and its brightness-enhanced mutant eUnaG11. UnaG is derived from

Japanese eel muscle and fluoresces when non-covalently bound to the bilirubin chromophore12. UnaG is switched off by photobleaching of the bilirubin adduct and its release from the binding

pocket, then switched on by rebinding to fresh bilirubin. Here, we extend the spectrum of FP-based PAINT probes to the NIR by reporting a phytochrome-derived FP capable of chromophore

renewal. This FP non-covalently binds the biliverdin IXα (BV). Under red light irradiation, the photobleached chromophore can be displaced by fresh BV. This reversible phytochrome-BV

interaction leads to fluorescence turnover and single emitters compatible with the principle of PAINT localization microscopy. RESULTS PHYTOCHROME-DERIVED NEAR-INFRARED FLUORESCENT PROTEIN

WITH NON-COVALENT BILIVERDIN ADDUCT We developed the chromophore-renewable imaging tag by using the NIR FP miRFP720 as a molecular template13. miRFP720 was previously derived from _Rp_BphP2,

a soluble bacterial phytochrome photoreceptor found in _Rhodopseudomonas palustris_14. Among phytochrome-derived FPs15, miRFP720 has the longest excitation (702 nm) and emission (720 nm)

wavelengths. It covalently binds the BV chromophore via a single thioether bond at cysteine 15 residue, whereas other phytochrome-derived FPs, e.g., miRFP670, attach to BV by two thioether

bonds. When expressed in mammalian cells, apo-miRFP720 took up endogenous BV produced through heme metabolism (Fig. 1a). The stable covalent bond between miRFP720 and BV resulted in

constitutive NIR fluorescence when expressed in mammalian cells (Supplementary Fig. 1a). The fluorescence was not diminished by the efflux of unbound BV after permeabilization of the plasma

membrane (Supplementary Fig. 1b). It was previously reported that a mutant of the _Agrobacterium_ phytochrome Agp1, which non-covalently interacts with BV, can recover its absorption

spectrum from the photobleached state upon the addition of fresh BV16. Interestingly, photodamage primarily targeted the BV cofactor, leaving the phytochrome intact. Fresh BV displaced its

photobleached counterpart from the binding pocket and restored a functional holoprotein. However, the fluorescence quantum yield of wildtype Agp1 is as low as 0.0002 (<1/300 of

miRFP720)17, making it impractical as an bioimaging tag. To develop a functional NIR FP that allows chromophore exchange, we disrupted the strong phytochrome-BV covalent bond by introducing

the C15S mutation in miRFP720 (Fig. 1b, c). Using superfolder GFP (sfGFP) as a transfection marker (Supplementary Fig. 2a), we examined the fluorescence emission of miRFP720(C15S) in

mammalian cells. In BV-free basal medium, the fluorescence of miRFP720(C15S) was barely observable in confocal microscopy (Supplementary Fig. 2b, c). This suggests that either the modified

binding pocket failed to accommodate BV, or endogenous BV in cells was insufficient for the onset of ensemble fluorescence. We next increased the intracellular BV concentration by externally

supplement 0.5 µM BV in imaging buffer after cell fixation. In addition, we localized miRFP720(C15S) to the endoplasmic reticulum (ER) using the Sec61β sequence (Supplementary Fig. 3a), to

distinguish miRFP720(C15S) signal from BV-induced autofluorescence by localization pattern. ER-specific fluorescence was induced after adding BV (Supplementary Fig. 3b), indicating

successful incorporation of BV by miRFP720(C15S). Using miRFP720(C15S) as a fusion tag, we acquired confocal microscopy images of ER, focal adhesions, and clathrin-coated vesicles in fixed

cells supplemented with 0.5 µM BV (Fig. 1e). In live cells, the addition of 0.5 µM BV failed to immediately activate Rep-miRFP in minutes (Supplementary Fig. 3c), suggesting limited

permeability of BV. Next, we investigated whether the chromophore could be delivered in live cells by long-term incubation with BV. To this end, untransfected HeLa cells were incubated for

24 h with up to 122 μM of BV in culture medium. Notably, BV solution can be weakly fluorescent and has an excitation peak at around 603 nm18. This allows the estimation of intracellular and

extracellular BV levels by measuring fluorescence intensity using a sensitive EMCCD camera (Supplementary Fig. 4a). Intracellular BV increased monotonically with extracellular BV

concentrations. Nevertheless, BV levels inside the cells were consistently lower than in the medium (Supplementary Fig. 4b). In addition, high fluorescence was observed on cell edges, which

again confirmed that intact plasma membrane is a barrier for BV delivery (Supplementary Fig. 4a). We obtained time-lapse confocal image of ER-localized miRFP720(C15S) in live HeLa cells

after incubation with 30 μM BV for 24 h (Fig. 1d), though non-specific fluorescence was induced by BV and observed on cell edges. We recently reported cell-level fluorescence of a

non-covalent BV-binding serpin derived from _Boana punctata_ (_Bp_BBS)19. To the best of our knowledge, miRFP720(C15S) is the first non-covalent BV-binding protein that functions as a fusion

tag for the imaging of organelles and subcellular compartments. REVERSIBLE INTERACTION BETWEEN PHYTOCHROME AND BILIVERDIN ENABLES FLUORESCENCE TURNOVER The β-barrel green FP UnaG

incorporates its bilirubin chromophore through a non-covalent bond. It was recently found that the bilirubin adduct can detach from the UnaG protein backbone after photobleaching, freeing

the binding cavity for de novo incorporate of fresh chromophore10. Bilirubin and BV are related bilins found in the heme metabolism pathway. To date, fluorescence turnover by the renewal of

BV chromophore has not been reported. In fixed HeLa cells, the ensemble fluorescence of miRFP720(C15S) could be rapidly depleted by 640 nm laser irradiation (Fig. 2a). The decay curve was

fitted with a bi-exponential function (magenta segment in Fig. 2b), suggesting the presence of two fluorescence states likely related to BV isomerization20,21. The on-to-off speed increased

with the power density of 640 nm irradiation (Fig. 2c). Importantly, photobleached miRFP720(C15S) spontaneously recovered its fluorescence in the dark when fresh BV was available (Fig. 2a).

The off-to-on curve was well fitted with a single-exponential function (blue segment in Fig. 2b). Increasing BV concentration linearly accelerated the fluorescence recovery process (Fig. 

2d). The fluorescence turnover behavior of miRFP720(C15S) is similar to that of UnaG despite the drastic difference in their protein structures. It suggests that non-covalent protein-ligand

interaction could be a general avenue that facilitates chromophore renewal. Switchable FPs enable stochastic fluorescence activation which is the cornerstone of many super-resolution

microscopy techniques. Compared to other switching mechanisms such as photoisomerization and protonation/deprotonation, chromophore renewal has the potential advantage of delaying the

irreversible photodamage in time-lapse imaging. To this end, we switched miRFP720(C15S) for 500 cycles in fixed HeLa cells supplemented with 1 μM BV, then measured the on/off contrast during

the 10-hour duration (Fig. 2e). Notably, the ensemble fluorescence remained detectable at the end of the experiment (Fig. 2f, g). We named this chromophore replenishable miRFP720 mutant as

Rep-miRFP, because of the fluorescence turnover property. NEAR-INFRARED PAINT LOCALIZATION MICROSCOPY BASED ON THE RENEWAL OF BILIVERDIN CHROMOPHORE We next examined whether the chromophore

renewal of Rep-miRFP could be exploited for PAINT localization microscopy to super-resolve a subcellular structure. We applied highly inclined and laminated optical sheet (HILO) illumination

using a 640 nm laser, and imaged fixed COS-7 cells expressing Rep-miRFP localized to the ER. COS-7 was chosen because its flat morphology well complements HILO illumination for capturing

the ER structure. In imaging buffer supplied with 0.5 µM BV and under 640 nm laser illumination at 28.9 W cm−2, the ensemble fluorescence of Rep-miRFP was rapidly reduced to single emitters

in seconds (Fig. 3a and Supplementary Movie 1). Single emitters were sustained in an equilibrium between photodepletion and de novo formation of Rep-miRFP holoprotein. Interestingly, we

noticed a background fluorescence prominent in the intracellular compartment, which was not observed in the cell-free regions (Supplementary Fig. 5). This background was also detected in

untransfected live cells after plasma membrane permeabilization and the influx of BV (Supplementary Movie 2), indicating that BV was the source of this background fluorescence. Upon a closer

examination of confocal images under high magnification, we also found that the non-specific fluorescence preferentially highlighted the mitochondria (blue arrow in Supplementary Fig. 6a).

These results are consistent with reports of BV accumulation at mitochondria22 and binding with proteins partners such as albumin23. The non-specific fluorescence negatively impacts the

fluorescence contrast of Rep-miRFP. In the absence of BV-induced background, the on/off contrast of Rep-miRFP was twice as high as that of Dronpa (Supplementary Fig. 7), a well-documented

photoswitchable green FP commonly used for optical highlighting and super-resolution microscopy24. In the presence of BV-induced background, the Rep-miRFP signal was obscured and the on/off

contrast was compromised (Supplementary Fig. 7). The BV-induced background has relatively little impact on confocal microscopy, because the ensemble fluorescence of the holoprotein is

usually dominant. In SMLM, fixed pattern background could reduce the accuracy of single-molecule localization and the fidelity of super-resolution images25. We sought to eliminate the

background in single-molecule image data prior to emitter localization. The background cannot be removed by an emission filter due to spectral overlap with Rep-miRFP. However, we observed in

time-lapse images that while Rep-miRFP was rapidly switched off by the 640 nm laser, fluorescence intensity of the fix pattern background was stable on the time scale of typical image

acquisition and inert to the irradiation (Supplementary Fig. 6b). This difference in temporal behavior allowed the extraction of single-emitter foreground, by using advanced background

subtraction algorithms such as SMALL-LABS26 and WindSTORM27. Although their detailed implementation differs, both SMALL-LABS and WindSTORM algorithms are designed to separate fast-changing

foreground from slow-changing background. We processed the single-molecule image data of Rep-miRFP using SMALL-LABS in MATLAB. Compared to the raw images, the single emitters in the output

were no longer obscured by the fix pattern background (Supplementary Fig. 5 and Supplementary Movie 3). We reconstructed super-resolution image of the ER using the background-subtracted data

(Fig. 3b). Line profile analysis revealed a 168 nm peak-to-peak distance between separated structures (Fig. 3c, d), which surpassed the diffraction limit. The lateral resolution was also

estimated by Fourier ring correlation (FRC)28,29, which showed a FRC resolution of 90 nm (Fig. 3g). Next, we compared the single-emitter performance of Rep-miRFP with β-barrel FPs. Each

localized Rep-miRFP molecule emitted an average number of 436 photons (Fig. 3e). In comparison, the photon budget of Dronpa, an early generation photoswitchable FP, was estimated to be 304

photon/localization. On the other hand, the very bright SPOON emitted 983 photon/localization. The photon budget of SPOON is comparable to other state-of-the-art SMLM tags such as mEos3.2

and mMaple2,3, which led to a high localization precision of 9.9 nm. The localization precision of Rep-miRFP was 17.6 nm (Fig. 3f), worse than that of SPOON but superior to Dronpa (24.5 nm).

Notably, due to chromophore renewal, the number of detectable Rep-miRFP single emitters remained largely unchanged after 60 min of continuous 640 nm laser irradiation (Fig. 3h). In

comparison, the single emitters of SPOON were almost completely exhausted after 30 min, due to photobleaching by the 488 nm laser (Fig. 3i). In contrast to fixed cells that have unblocked

access to free BV, live-cell SMLM imaging was limited by the inability of medium BV to rapidly diffuse across the plasma membrane. This was demonstrated by the quick exhaustion of diffusive

intracellular BV and localizable molecules, even after incubating live cells with 30 µM BV for 24 h (Supplementary Fig. 8). Rep-miRFP emits in the NIR spectrum which is not occupied by

existing photoconvertible and photoswitchable FPs, making it flexible for multiplexed super-resolution imaging. We performed two-color SMLM using the combination of Rep-miRFP and SPOON.

Rep-miRFP-Sec61β and Lifeact-SPOON were co-expressed in COS-7 cells by co-transfection. Images of the two fluorescence channels were acquired sequentially in fixed cells. The SPOON channel

was acquired first under 488 nm illumination. After adding BV at a final concentration of 0.5 μM, the Rep-miRFP channel was acquired next under 640 nm illumination (Fig. 3j). Axial chromatic

aberration was measured and compensated (Supplementary Fig. 9). As a result, multiplexed SMLM of the ER and F-actin was obtained (Fig. 3k). DISCUSSION Proteins using BV as a fluorescent

cofactor have become attractive templates for the development of FPs, because it is challenging to cover the NIR spectrum by using β-barrel FPs. Extensive mutagenesis has led to some of the

most redshifted β-barrel FPs including mKelly1 (maximum λem = 656 nm)30 and mGarnet2 (maximum λem = 671 nm)31. In comparison, wildtype BV-binding proteins naturally have longer emission

wavelength, e.g., Sandercyanin (maximum λem = 675 nm)32 and _Bp_BBS (maximum λem = 691 nm)19. Here, we have described a phytochrome-derived FP that non-covalently and reversibly interacts

with BV (Fig. 3l). Previously, UnaG was reported for PAINT localization microscopy in green channel. Rep-miRFP expands the color palette of FP-based PAINT probes. Our finding suggests that

chromophore renewal might be a generalizable approach for developing genetically encoded PAINT probes from a larger variety of templates in the future, including BV-binding

phytochromes13,15,33, cyanobacteriochromes34,35, phycobiliproteins36, lipocalin32 and serpins19. To the best of our knowledge, Rep-miRFP is the first non-covalent BV-binding protein that

functions as a fusion tag for fluorescence microscopy. Two other promising candidates in this class are: Sandercyanin, a tetrameric lipocalin found in walleye fish32; _Bp_BBS, a serpin found

in polka-dot tree frog19. Although the fluorescence of these two were documented, their functionality as bioimaging tags has not been well explored. Currently, Rep-miRFP has two major

shortcomings. First, the negatively charged carboxylates of BV limits its membrane permeability37 and the utility of Rep-miRFP for live-cell imaging. BV variants with higher membrane

permeability have been synthesized by modifying the carboxylates to create more neutrally charged molecules. The resulting biliverdin dimethyl ester (BVMe2)36 and BV-PEG-FAM37 can be

incorporated by the phycobiliprotein-derived FP smURFP in live cells and tissues. However, bacterial phytochromes use the carboxylic acids of BV as a recognition motif, thus cannot

incorporate BVMe2. It remains to be seen whether chromophore-renewable FPs could be derived from templates that naturally tolerate BVMe2, and whether molecular evaluation of phytochromes

could lead to a modified chromophore binding pocket that accommodates BVMe2. Second, Rep-miRFP is less bright than the best-in-class photoswitchable FPs in SMLM. Typically, the brightness of

BV-binding proteins is improved by reducing chromophore flexibility and increasing the overall protein rigidity to suppress nonradiative transition38. By extensive mutagenesis and

screening, the fluorescence quantum yield of a monomeric cyanobacteriochrome, miRFP670nano3, recently reached a record high of 0.18535. Stabilizing the BV binding pocket while preserving the

chromophore exchange ability of Rep-miRFP could be a challenging but exciting avenue to go down in the future. METHODS FLUORESCENT PROTEINS The gene encoding miRFP720 was cloned into the

pcDNA3.0 vector (Invitrogen/Thermo-Fisher Scientific) between the BamHI and EcoRI restriction sites. The C15S point mutation was introduced in miRFP720 by inverse PCR using the KOD FX Neo

DNA polymerase (Toyobo Life Sciences). The PCR product was digested by DpnI at 37 °C overnight to remove the template DNA, followed by T4 ligation (Promega). SPOON was previously developed

in our lab by mutagenesis of Dreiklang 39. Dronpa was a gift from Atsushi Miyawaki at RIKEN Center for Brain Science. CONSTRUCTION OF MAMMALIAN EXPRESSION PLASMIDS For the fluorescent

labeling of organelles and subcellular compartments, ER localization was achieved by fusing Sec61β to the C-terminus of FPs and insertion into pcDNA3.0 between BamHI and EcoRI restriction

sites using In-Fusion cloning (Takara Bio). Sec61β fragment was obtained from pAc-GFPC1-Sec61β (Addgene #15108). For labeling focal adhesion sites, emiRFP703 between AgeI and NotI

restriction sites in pzyxin-emiRFP703 (Addgene #136568) was swapped with Rep-miRFP. For labeling clathrin-coated vesicles, Kohinoor between NheI and BglII restriction sites in

Kohinoor-Clathrin (Addgene #67773) was swapped with Rep-miRFP. For labeling the nucleus, Phamret between BamHI and EcoRI restriction sites in Phamret-H2B (Addgene #51955) was swapped with

SPOON or miRFP720. For bicistronic expression of sfGFP and NIR FPs, the two domains were linked by a self-cleaving P2A peptide and inserted into pcDNA3.0 between BamHI and EcoRI restriction

sites. CELL CULTURE AND TRANSFECTION HeLa and COS-7 cells were cultured in low-glucose Dulbecco’s modified Eagle medium (DMEM; Sigma-Aldrich) supplemented with 10% fetal bovine serum (FBS;

Biowest) at 37 °C in a 5% CO2 incubator. Cells were seeded on home-made 35 mm glass-bottom dishes and grown to 50–60% confluence. For transfection, 2 μg of plasmid (or 1 μg of each plasmid

for co-expression) was mixed with 5.0 μg of polyethylenimine MAX (Polysciences) in 200 μL of Opti-MEM medium (Thermo Fisher Scientific). The mixture was incubated for 20 min at room

temperature and then added dropwise to the cell culture. The medium was refreshed after 6 h, and the cells were cultured for 24–48 h before observation. CONFOCAL MICROSCOPY For live-cell

confocal microscopy, cells were immersed in DMEM/F-12 without phenol red (Thermo Fisher Scientific). For fixed-cell confocal microscopy, the cells were treated with 4% paraformaldehyde for

20 min at room temperature, then washed three times with phosphate buffered saline. Cells were imaged with a Dragonfly 200 spinning disk confocal microscope (Andor) equipped with an iXon

Ultra EMCCD camera (Andor), quad-band 405/488/561/640 dichroic mirror, Plan Apo Lambda 60×/1.40 oil immersion or Plan Apo Lambda S 100×/1.35 silicone oil immersion objective lens (Nikon).

The Dragonfly 200 spinning disk confocal microscope and iXon Ultra EMCCD camera were controlled by Fusion 2.3.0.45 (Andor). NIR FPs were excited with 640 nm laser and the fluorescence

collected through a 725/40 emission filter (Semrock). Switching kinetics of Rep-miRFP was measured in time-lapse using the same setup. Green FPs including sfGFP and Dronpa were excited with

488 nm laser and their fluorescence was collected through a 521/33 emission filter (Semrock). PAINT LOCALIZATION MICROSCOPY COS-7 cells were transfected with

sfGFP-P2A-Rep-miRFP-Sec61β/pcDNA3.0. sfGFP was used as a transfection marker to identify transfected cells and expression level. The single-molecule imaging system was based on an inverted

microscope (ECLIPSE Ti, Nikon) equipped with an sCMOS camera (ORCA-Flash4.0, Hamamatsu Photonics), a quad-band 405/488/561/640 dichromic mirror and a 732/68 and 525/45 emission filter

(Semrock). The ORCA-Flash4.0 sCMOS camera was controlled by HCImageLive 4.3.1.33 (Hamamatsu Photonics). Objective-based HILO illumination was achieved using a 640 nm laser (OBIS 640 nm LX

100 mW, Coherent) and an Apo TIRF 100×/1.49 oil-immersion objective lens (Nikon). Single-molecule image track was acquired with an exposure time of 200 ms at a frame rate of 5 Hz. The

irradiance of 640 nm laser at the sample plane was 28.9 W cm−2. Laser power was measured with a laser power meter (power meter console, PM400, Thorlabs; photodiode sensor, S120VC, Thorlabs).

Size of the illumination area was measured with an objective micrometer (OB-M#, 1/100, Olympus). RECONSTRUCTION OF SUPER-RESOLUTION IMAGES The raw single-molecule image track of Rep-miRFP

was processed by the SMALL-LAB algorithm26 in MATLAB 9.8 to remove the heterogeneous background found inside fixed cells after the addition of BV. The workflow of background subtraction

followed the user guide (https://github.com/BiteenMatlab/SMALL-LABS). First, an initial round of crude background subtraction was performed before emitter detection. In each frame, the

low-frequency background that changed slowly over time was globally removed by calculating and subtracting a temporal mean image over a sliding window of 300 frames (1 min). Second,

approximated detection of emitters was performed after the last step. Third, for each detected emitter, its unique local background was calculated by temporal searching for neighboring

frames in which this emitter was in off state and not obscured by other on-state molecules in the diffraction-limited region. Fourth, the true local background specific to each detected

molecule was subtracted. Fifth, the background-subtracted single molecules were localized with diffraction-unlimited precision by gaussian fitting. Lastly, SMLM images were reconstructed

from 10,000 frames using the coordinates of localized molecules obtained in the last step. IMAGE ANALYSES AND FOURIER RING CORRELATION Image data were analyzed using ImageJ 1.54 f and Fiji

software (https://fiji.sc). Fourier ring correlation (FRC) analysis was performed with an ImageJ plugin (https://github.com/BIOP/ijp-frc). The single-molecule track was split into two stacks

containing odd or even number of frames, reconstructed into two statistically independent SMLM images, and then fed into the plugin for FRC analysis. The spatial resolution was estimated by

the inverse of the spatial frequency at the intercept of the FRC curve and the threshold of 1/7 correlation. MULTIPLEXED SMLM COS-7 cells were co-transfected with Rep-miRFP-Sec61β/pcDNA3.0

and Lifeact-SPOON/pcDNA3.0, then fixed. The same microscope setup for single-color SMLM was used, except for an additional light path for 488 nm laser. Sequential image acquisition was

performed in the order of green channel followed by NIR channel. Objective-based HILO illumination of SPOON was achieved using a 488 nm laser (OBIS 488 nm LS 150 mW, Coherent). First, the

green channel single-molecule image track was acquired without adding BV, with an exposure time of 200 ms at a frame rate of 5 Hz using a 525/45 emission filter (Semrock). The power density

of 488 nm laser at the sample plane was 54.5 W cm−2. Subsequently, BV was added at a final concentration of 0.5 µM and the NIR channel single-molecule image track was acquired under 640 nm

laser irradiation. SMLM image of SPOON was reconstructed from 10,000 frames using ThunderSTORM v1.340, using ‘difference of Gaussians’ filter for particle detection, ‘integrated Gaussian’

for PSF model and ‘maximum likelihood estimation’. Protein PAINT image of Rep-miRFP was reconstructed as described in the previous section. STATISTICS AND REPRODUCIBILITY Curve fitting was

performed using GraphPad Prism 9.3.1. Data obtained from each experiment were expressed as the mean ± SD. The sample sizes were described in each figure legend. REPORTING SUMMARY Further

information on research design is available in the Nature Portfolio Reporting Summary linked to this article. DATA AVAILABILITY The source data for the main graphs are provided in

Supplementary Data. All other supplementary data in this study are available from the corresponding author on reasonable request. REFERENCES * Schermelleh, L. et al. Super-resolution

microscopy demystified. _Nat. Cell Biol._ 21, 72 (2019). Article  CAS  PubMed  Google Scholar  * Zhang, M. et al. Rational design of true monomeric and bright photoactivatable fluorescent

proteins. _Nat. Methods_ 9, 727–729 (2012). Article  CAS  PubMed  Google Scholar  * Wang, S., Moffitt, J. R., Dempsey, G. T., Xie, X. S. & Zhuang, X. Characterization and development of

photoactivatable fluorescent proteins for single-molecule–based superresolution imaging. _PNAS_ 111, 8452–8457 (2014). Article  CAS  PubMed  PubMed Central  Google Scholar  * Arai, Y. et al.

Spontaneously blinking fluorescent protein for simple single laser super-resolution live cell imaging. _ACS Chem. Biol._ 13, 1938–1943 (2018). Article  CAS  PubMed  Google Scholar  * Kwon,

J., Elgawish, M. S. & Shim, S.-H. Bleaching-resistant super-resolution fluorescence microscopy. _Adv. Sci._ 9, 2101817 (2022). Article  Google Scholar  * Perfilov, M. M., Gavrikov, A.

S., Lukyanov, K. A. & Mishin, A. S. Transient fluorescence labeling: low affinity—high benefits. _Int. J. Mol. Sci._ 22, 11799 (2021). Article  CAS  PubMed  PubMed Central  Google

Scholar  * Sharonov, A. & Hochstrasser, R. M. Wide-field subdiffraction imaging by accumulated binding of diffusing probes. _Proc. Natl Acad. Sci. USA_ 103, 18911–18916 (2006). Article 

CAS  PubMed  PubMed Central  Google Scholar  * Jungmann, R. et al. Single-molecule kinetics and super-resolution microscopy by fluorescence imaging of transient binding on DNA Origami. _Nano

Lett._ 10, 4756–4761 (2010). Article  CAS  PubMed  Google Scholar  * Jungmann, R. et al. Multiplexed 3D cellular super-resolution imaging with DNA-PAINT and Exchange-PAINT. _Nat. Methods_

11, 313–318 (2014). Article  CAS  PubMed  PubMed Central  Google Scholar  * Kwon, J. et al. Bright ligand-activatable fluorescent protein for high-quality multicolor live-cell

super-resolution microscopy. _Nat. Commun._ 11, 1–11 (2020). Article  Google Scholar  * Ko, S., Kwon, J. & Shim, S.-H. Enhanced UnaG with minimal labeling artifact for single-molecule

localization microscopy. _Front. Mol. Biosci_. 8, 647590 (2021). * Kumagai, A. et al. A bilirubin-inducible fluorescent protein from Eel Muscle. _Cell_ 153, 1602–1611 (2013). Article  CAS 

PubMed  Google Scholar  * Shcherbakova, D. M., Cox Cammer, N., Huisman, T. M., Verkhusha, V. V. & Hodgson, L. Direct multiplex imaging and optogenetics of Rho GTPases enabled by

near-infrared FRET. _Nat. Chem. Biol._ https://doi.org/10.1038/s41589-018-0044-1 (2018). Article  PubMed  PubMed Central  Google Scholar  * Yang, X. et al. Light signaling mechanism of two

tandem bacteriophytochromes. _Structure_ 23, 1179–1189 (2015). Article  CAS  PubMed  PubMed Central  Google Scholar  * Matlashov, M. E. et al. A set of monomeric near-infrared fluorescent

proteins for multicolor imaging across scales. _Nat. Commun._ 11, 1–12 (2020). Article  Google Scholar  * Borucki, B., Seibeck, S., Heyn, M. P. & Lamparter, T. Characterization of the

covalent and noncovalent adducts of Agp1 phytochrome assembled with biliverdin and phycocyanobilin by circular dichroism and flash photolysis. _Biochemistry_ 48, 6305–6317 (2009). Article 

CAS  PubMed  Google Scholar  * Zienicke, B. et al. Fluorescence of phytochrome adducts with synthetic locked chromophores. _J. Biol. Chem._ 286, 1103–1113 (2011). Article  CAS  PubMed 

Google Scholar  * Leem, J. W. et al. Visible light biophotosensors using biliverdin from Antheraea yamamai. _Opt. Express_ 26, 31817–31828 (2018). Article  CAS  PubMed  Google Scholar  *

Manoilov, K. Y., Ghosh, A., Almo, S. C. & Verkhusha, V. V. Structural and functional characterization of a biliverdin-binding near-infrared fluorescent protein from the serpin

superfamily. _J. Mol. Biol._ 434, 167359 (2022). Article  CAS  PubMed  Google Scholar  * Holt, R. E., Farrens, D. L., Song, P. S. & Cotton, T. M. Surface-enhanced resonance Raman

scattering (SERRS) spectroscopy applied to phytochrome and its model compounds. I. Biliverdin photoisomers. _J. Am. Chem. Soc._ 111, 9156–9162 (1989). Article  CAS  Google Scholar  *

Hontani, Y. et al. Real-time observation of tetrapyrrole binding to an engineered bacterial phytochrome. _Commun. Chem._ 4, 11 (2021). Article  Google Scholar  * Shum, M. et al. ABCB10

exports mitochondrial biliverdin, driving metabolic maladaptation in obesity. _Sci. Transl. Med._ 13, eabd1869 (2021). Article  CAS  PubMed  PubMed Central  Google Scholar  * Wei, Y. et al.

Investigation of the association behaviors between biliverdin and bovine serum albumin by fluorescence spectroscopy. _Talanta_ 70, 377–382 (2006). Article  CAS  PubMed  Google Scholar  *

Ando, R., Mizuno, H. & Miyawaki, A. Regulated fast nucleocytoplasmic shuttling observed by reversible protein highlighting. _Science_ 306, 1370–1373 (2004). Article  CAS  PubMed  Google

Scholar  * Hoogendoorn, E. et al. The fidelity of stochastic single-molecule super-resolution reconstructions critically depends upon robust background estimation. _Sci. Rep._ 4, 3854

(2014). Article  PubMed  PubMed Central  Google Scholar  * Isaacoff, B. P., Li, Y., Lee, S. A. & Biteen, J. S. SMALL-LABS: measuring single-molecule intensity and position in obscuring

backgrounds. _Biophys. J._ 116, 975–982 (2019). Article  CAS  PubMed  PubMed Central  Google Scholar  * Ma, H., Xu, J. & Liu, Y. WindSTORM: robust online image processing for

high-throughput nanoscopy. _Sci. Adv._ 5, eaaw0683 (2019). Article  PubMed  PubMed Central  Google Scholar  * Nieuwenhuizen, R. P. J. et al. Measuring image resolution in optical nanoscopy.

_Nat. Methods_ 10, 557–562 (2013). Article  CAS  PubMed  PubMed Central  Google Scholar  * Culley, S. et al. Quantitative mapping and minimization of super-resolution optical imaging

artifacts. _Nat. Methods_ 15, 263–266 (2018). Article  CAS  PubMed  PubMed Central  Google Scholar  * Wannier, T. M. et al. Monomerization of far-red fluorescent proteins. _Proc. Natl Acad.

Sci._ 115, E11294–E11301 (2018). Article  CAS  PubMed  PubMed Central  Google Scholar  * Matela, G. et al. A far-red emitting fluorescent marker protein, mGarnet2, for microscopy and STED

nanoscopy. _Chem. Commun._ 53, 979–982 (2017). Article  CAS  Google Scholar  * Ghosh, S. et al. Blue protein with red fluorescence. _Proc. Natl Acad. Sci._ 113, 11513–11518 (2016). Article 

CAS  PubMed  PubMed Central  Google Scholar  * Shcherbakova, D. M. et al. Bright monomeric near-infrared fluorescent proteins as tags and biosensors for multiscale imaging. _Nat. Commun._ 7,

12405 (2016). Article  CAS  PubMed  PubMed Central  Google Scholar  * Oliinyk, O. S., Shemetov, A. A., Pletnev, S., Shcherbakova, D. M. & Verkhusha, V. V. Smallest near-infrared

fluorescent protein evolved from cyanobacteriochrome as versatile tag for spectral multiplexing. _Nat. Commun._ 10, 279 (2019). Article  PubMed  PubMed Central  Google Scholar  * Oliinyk, O.

S. et al. Single-domain near-infrared protein provides a scaffold for antigen-dependent fluorescent nanobodies. _Nat. Methods_ 19, 740–750 (2022). Article  CAS  PubMed  PubMed Central 

Google Scholar  * Rodriguez, E. A. et al. A far-red fluorescent protein evolved from a cyanobacterial phycobiliprotein. _Nat. Methods_ 13, 763–769 (2016). Article  CAS  PubMed  PubMed

Central  Google Scholar  * Machado, J.-H., Ting, R., Lin, J. Y. & Rodriguez, E. A. A self-labeling protein based on the small ultra-red fluorescent protein, smURFP. _RSC Chem. Biol._ 2,

1221–1226 (2021). Article  CAS  PubMed  PubMed Central  Google Scholar  * Shemetov, A. A., Oliinyk, O. S. & Verkhusha, V. V. How to increase brightness of near-infrared fluorescent

proteins in mammalian cells. _Cell Chem. Biol._ 24, 758–766.e3 (2017). Article  CAS  PubMed  PubMed Central  Google Scholar  * Brakemann, T. et al. A reversibly photoswitchable GFP-like

protein with fluorescence excitation decoupled from switching. _Nat. Biotechnol._ 29, 942–947 (2011). Article  CAS  PubMed  Google Scholar  * Ovesný, M., Křížek, P., Borkovec, J., Švindrych,

Z. & Hagen, G. M. ThunderSTORM: a comprehensive ImageJ plug-in for PALM and STORM data analysis and super-resolution imaging. _Bioinformatics_ 30, 2389–2390 (2014). Article  PubMed 

PubMed Central  Google Scholar  Download references ACKNOWLEDGEMENTS This work was supported by grants from Core Research for Evolutionary Science and Technology, Japan Science and

Technology Agency (JPMJCR15N3 to T.N.; JPMJCR20E4 to T.M.), grant-in-aid from Japan Society for the Promotion of Science (18H03987, 18H05410, and 21K19225 to T.N.; 22K04891 to T.W.),

GM122567 from the US National Institutes of Health and 220011 from the Jane and Aatos Erkko Foundation (both to V.V.V.). AUTHOR INFORMATION AUTHORS AND AFFILIATIONS * SANKEN (The Institute

of Scientific and Industrial Research), Osaka University, 8-1 Mihogaoka, Ibaraki, Osaka, 567-0047, Japan Kai Lu, Tetsuichi Wazawa, Tomoki Matsuda & Takeharu Nagai * Department of

Genetics and Gruss-Lipper Biophotonics Center, Albert Einstein College of Medicine, Bronx, NY, 10461, USA Daria M. Shcherbakova & Vladislav V. Verkhusha * Medicum, Faculty of Medicine,

University of Helsinki, Helsinki, 00290, Finland Vladislav V. Verkhusha Authors * Kai Lu View author publications You can also search for this author inPubMed Google Scholar * Tetsuichi

Wazawa View author publications You can also search for this author inPubMed Google Scholar * Tomoki Matsuda View author publications You can also search for this author inPubMed Google

Scholar * Daria M. Shcherbakova View author publications You can also search for this author inPubMed Google Scholar * Vladislav V. Verkhusha View author publications You can also search for

this author inPubMed Google Scholar * Takeharu Nagai View author publications You can also search for this author inPubMed Google Scholar CONTRIBUTIONS T.N. conceived the project. K.L.,

V.V.V. and T.N. designed the experiments. K.L. performed the experiments. K.L. and T.W. constructed the microscope for SMLM imaging. K.L., T.W. and T.M. analyzed the data. T.N. supervised

the project. D.M.S. and V.V.V. designed the miRFP720 protein and discussed the results. K.L., V.V.V. and T.N. wrote the manuscript. All authors reviewed and revised the manuscript.

CORRESPONDING AUTHOR Correspondence to Takeharu Nagai. ETHICS DECLARATIONS COMPETING INTERESTS The authors declare no competing interests. PEER REVIEW PEER REVIEW INFORMATION _Communications

Biology_ thanks the anonymous reviewers for their contribution to the peer review of this work. Primary Handling Editors: Periklis Pantazis and Christina Karlsson Rosenthal. ADDITIONAL

INFORMATION PUBLISHER’S NOTE Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. SUPPLEMENTARY INFORMATION SUPPLEMENTARY

INFORMATION DESCRIPTION OF ADDITIONAL SUPPLEMENTARY FILES SUPPLEMENTARY DATA SUPPLEMENTARY MOVIE 1 SUPPLEMENTARY MOVIE 2 SUPPLEMENTARY MOVIE 3 REPORTING SUMMARY RIGHTS AND PERMISSIONS OPEN

ACCESS This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format,

as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third

party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the

article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright

holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. Reprints and permissions ABOUT THIS ARTICLE CITE THIS ARTICLE Lu, K., Wazawa, T., Matsuda, T. _et

al._ Near-infrared PAINT localization microscopy via chromophore replenishment of phytochrome-derived fluorescent tag. _Commun Biol_ 7, 473 (2024). https://doi.org/10.1038/s42003-024-06169-7

Download citation * Received: 29 August 2023 * Accepted: 09 April 2024 * Published: 18 April 2024 * DOI: https://doi.org/10.1038/s42003-024-06169-7 SHARE THIS ARTICLE Anyone you share the

following link with will be able to read this content: Get shareable link Sorry, a shareable link is not currently available for this article. Copy to clipboard Provided by the Springer

Nature SharedIt content-sharing initiative