Expression-Enhanced Fluorescent Proteins Based ... - ACS Publications

Subscriber access provided by Stockholm University Library

Article

Expression-Enhanced Fluorescent Proteins Based on Enhanced Green Fluorescent Protein (EGFP) for Super-Resolution Microscopy Sam Duwé, Elke De Zitter, Vincent Gielen, Benjamien Moeyaert, Wim Vandenberg, Tim Grotjohann, Koen Clays, Stefan Jakobs, Luc Van Meervelt, and Peter Dedecker ACS Nano, Just Accepted Manuscript • DOI: 10.1021/acsnano.5b04129 • Publication Date (Web): 26 Aug 2015 Downloaded from http://pubs.acs.org on August 27, 2015

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

ACS Nano is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 41

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Nano

Table of Contents Figure 40x18mm (300 x 300 DPI)

ACS Paragon Plus Environment

ACS Nano

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Expression-Enhanced Fluorescent Proteins Based on Enhanced Green Fluorescent Protein (EGFP) for Super-Resolution Microscopy Sam Duwé,ǂ Elke De Zitter,ǂ,ǁ Vincent Gielen,ǂ Benjamien Moeyaert,ǂ Wim Vandenberg,ǂ Tim Grotjohann,† Koen Clays,┴ Stefan Jakobs,† Luc Van Meervelt,ǁ Peter Dedecker*,ǂ ǂ

Laboratory for Photochemistry and Spectroscopy , ǁLaboratory for Biomolecular Architecture, ┴

Laboratory for Molecular Electronics and Photonics, Department of Chemistry, KU Leuven,

Celestijnenlaan 200F, 3001 Leuven, Belgium; †Department of NanoBiophotonics, Max Planck Institute for Biophysical Chemistry, Am Fassberg 11, 37077 Goettingen, Germany.

ABSTRACT: “Smart fluorophores”, such as reversibly switchable fluorescent proteins (RSFPs), are crucial for advanced fluorescence imaging. However, only a limited number of such labels is available and many display reduced biological performance compared to more classical variants. We present the development of robustly photoswitchable variants of EGFP, named rsGreens, that display up to 30-fold higher fluorescence in E. coli colonies grown at 37°C and more than 4-fold higher fluorescence when expressed in HEK293T cells compared to their ancestor protein rsEGFP. This enhancement is not due to an intrinsic increase in the fluorescence brightness of the probes, but rather due to enhanced expression levels that allow many more probe molecules to be functional at any given time. We developed rsGreens displaying a range of photoswitching 1 ACS Paragon Plus Environment

Page 2 of 41

Page 3 of 41

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Nano

kinetics and show how these can be used for multi-modal diffraction-unlimited fluorescence imaging such as pcSOFI and RESOLFT, achieving a spatial resolution of ~70 nm. By determining the first ever crystal structures of a negative reversibly switchable FP derived from Aequorea victoria in both the “on”- and “off”-conformation we were able to confirm the presence of a cis-trans isomerization and provide further insights into the mechanisms underlying the photochromism. Our work demonstrates that genetically encoded “smart fluorophores” can be readily optimized for biological performance, and provides a practical strategy for developing maturation- and stability-enhanced photochromic fluorescent proteins.

KEYWORDS: fluorescent proteins, reversible photoswitching, super-resolution fluorescence microscopy, SOFI, RESOLFT, crystal structure determination, rsEGFP, superfolder Fluorescent proteins (FPs) enable the minimally-invasive labeling of intracellular structures in live systems.1 The discovery and development of “smart photoactive FPs”,2,3 with features such as irreversible photoactivation and photoconversion, or reversible photoswitching, allowed the development of diffraction-unlimited imaging techniques such as (f)PALM4,5 ((fluorescence) photoactivated localization microscopy), RESOLFT6 (reversible saturable optical fluorescence transitions) and (pc)SOFI7,8 ((photochromic) stochastic optical fluctuation imaging). These techniques strongly rely on the performance of the fluorophores and considerable efforts have therefore been dedicated to create optimized “smart labels”.9 This is exemplified by the continuous optimization and diversification of the EosFP family,10–15 or the development of Dronpa16 mutants with different or added photophysical properties.17–22 Probes that combine multiple “smart” behaviors have also been engineered.23–25 On the whole, however, the general acceptance of the FP-based “smart labels” has not quite risen up to the high expectations set by the many applications they enable. In some cases this is 2 ACS Paragon Plus Environment

ACS Nano

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

due to concerns surrounding the biological compatibility of the labels, meaning that the label may interfere with the functioning of the system under study. Many end-users have large bodies of work, resources, and efforts invested in using one or just a few different fluorescent proteins, and are justifiably reluctant to re-engineer their samples using the latest FPs out of concern that this major effort may lead to inconsistent results if the new label has a different biological compatibility. To some extent these concerns are warranted: many FPs derive from corals, which display low sequence homology with the widely-used FPs derived from the A. victoria jellyfish. In some cases, “smart FPs” also display a reduced biological performance compared to their ancestor proteins, simply because the evolutionary approach used in the majority of FP development makes it difficult to optimize for multiple target parameters simultaneously, as this converts an essentially one-dimensional search (e.g. optimizing photoswitching) into an ndimensional search (optimizing photoswitching and brightness and …). A recent addition to the list of “smart labels” is rsEGFP,26 which displays very fast photoswitching and a high resistance to photodestruction. A direct descendant of mEGFP, rsEGFP holds great promise for the wide range of applications in which other (Aequorea victoria) avGFP-based proteins are routinely used, driven by several advantages of avGFP-based labels such as a low toxicity and a monomeric character.27–32 Intrigued by the optical properties of rsEGFP,26 we set out to use it for our own imaging. However, in our hands mammalian cells cultured at 37°C and expressing rsEGFP displayed a total fluorescence signal that was much lower than that of cells expressing EGFP, suggesting that many fewer rsEGFP molecules became functional when expressed under identical conditions. We reasoned that this propensity not only limited imaging experiments at this temperature, but was also likely to place considerable stress on the expressing cells as a result of the prevalent protein misfolding. We therefore set out to

3 ACS Paragon Plus Environment

Page 4 of 41

Page 5 of 41

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Nano

generate expression-enhanced versions of rsEGFP, while also preserving or even creating a range of different photochromic behaviors. In this work we demonstrate what we believe to be a general strategy to dramatically enhance the folding and maturation of reversibly photochromic fluorescent proteins, using a combined rational and evolutionary approach. We shed light on the structural aspects of these improvements using X-ray crystallography, and show that the photochromic properties can be directly modulated using targeted mutations, while preserving improvements in the robust formation of the functional label. Given the highly conserved mechanism of photoswitching,33,34 our insights can be applied in the engineering of other photochromic FPs, and may expand the range of “smart labels” available for imaging.35 Since the overall exterior structure is almost identical to EGFP, our rsGreens can ideally be introduced in any construct where EGFP is currently used, with minimal perturbations. In combination with its robust photochromic properties, this should expand the range of experiments amenable to diffraction-unlimited imaging and other advanced fluorescence imaging techniques. During the initial stages of this work, a faster switching mutant of rsEGFP, rsEGFP2, was published and reported to have improved maturation.36 In what follows we include rsEGFP2 for comparison with our labels. RESULTS AND DISCUSSION While investigating the use of rsEGFP for our own imaging, we noticed only very dim fluorescence when the protein was expressed in HeLa cells at 37°C, severely impeding our imaging. Similarly, E. coli colonies transformed with rsEGFP-encoding expression plasmids barely became fluorescent when grown at 37°C. Incubating colonies at 4°C for prolonged periods of time (3-4 days to 2 weeks) resulted in a brighter fluorescence emission, as did growing the E. coli at lower temperatures (~20°C), though it was still below what could be expected given the 4 ACS Paragon Plus Environment

ACS Nano

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 41

spectroscopic properties of rsEGFP. In all cases, the detectable fluorescence was low compared to other known green RSFPs such as Dronpa16 and members of the mGeos14 series (data not shown). We concluded that rsEGFP displays a strongly impaired folding/maturation pathway and/or reduced structural stability. MUTAGENESIS. We set out to improve the rsEGFP formation and stability while preserving the photoswitching properties. For this purpose, the fluorescence formation of rsEGFP was first further disturbed by creating a genetic fusion to an interfering bait peptide, bullfrog red-cell ferritin subunit-H.37,38 This poorly soluble peptide imposes an additional load on the folding of its fusion partner. Expression of this fusion construct (pRSETb::ferritin-rsEGFP) in E. coli at 37°C resulted in an even more reduced fluorescence signal (Figure S1, supporting information). We used this ferritin fusion construct for all mutagenesis and initial screening, and afterwards subcloned the resulting fluorescent proteins to an empty plasmid for expression and all further characterization. Initially, we introduced the folding-enhancing mutations reported during the evolution of cycle-3 GFP39 and superfolder GFP38 (sfGFP), resulting in a mutant with seven amino

acid

replacements,

named

rsGreen0.6

(rsEGFP-

S30R/Y39N/F99S/N105T/Y145F/M153T/I171V). This bright and switchable clone was subjected to further site-directed mutagenesis, this time focusing on mutations reported in other works (Table S1 provides a complete list of the tested mutations, supporting information). This resulted in an additional two mutations (S72A/T105Y) which further increased in-colony fluorescence compared to rsGreen0.6. We named this rationally improved switchable FP rsGreen0.7 (rsEGFP-S30R/Y39N/S72A/F99S/N105Y/Y145F/M153T/I171V). We took care not to mutate any of the residues reported to be important for the photochromicity (L69, A150, S163 and S205).26 The monomerization mutation A206K40, introduced during the development of rsEGFP, was also kept to ensure the monomeric character of the FPs. 5 ACS Paragon Plus Environment

Page 7 of 41

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Nano

DNA shuffling of rsGreen0.7 and rsEGFP did not identify redundant mutations. Although rsGreen0.7 is an efficient photoswitcher, further characterization revealed a slightly slower offswitching and increased susceptibility to photoswitching fatigue compared to rsEGFP. We attributed this to a single mutation, Y145F. The photoswitching behavior of rsEGFP was restored in rsGreen0.7b (rsGreen0.7-F145Y), but at a loss of fluorescence intensity in E. coli colonies compared to rsGreen0.7. We continued the directed evolution with several rounds of random mutagenesis, starting from the clone with the highest in-colony fluorescence, rsGreen0.7. Screening efforts were primarily focused on a more efficient photoswitching performance, this time preserving the fluorescence level. After two rounds of mutagenesis and screening of 50-100 × 103 colonies, we selected a single clone, named rsGreen0.8 (rsGreen0.7-K162R/A227S), and two additional interesting clones. In an attempt to further improve rsGreen0.8, we directly introduced the beneficial mutations from the other two clones, yielding rsGreen0.9 (rsGreen0.8-L44M/K101E/H169L), which showed an improved photoswitching contrast and speed. Two final clones were identified after screening of approximately 35 × 103 colonies from a final random mutagenesis library. rsGreen1 (rsGreen0.9-N149D) provided improved in-colony fluorescence, while rsGreenF (rsGreen0.9-F145L) displayed both improved in-colony fluorescence and faster photoswitching with a higher contrast (or lower residual off-state fluorescence) compared to rsGreen0.9. Figure S2 (supporting information) shows an amino acid sequence alignment of all rsGreens with EGFP, sfGFP, rsEGFP and rsEGFP2. During the mutagenesis of rsEGFP to the different rsGreens, a faster switching mutant of rsEGFP, rsEGFP2, was published and reported to have improved maturation.36 We therefore included both rsEGFP and rsEGFP2 as benchmarks during the characterization of the rsGreens.


ACS Nano

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 41

Table 1: Spectroscopic properties of EGFP, rsEGFP, rsEGFP2, and the rsGreens. EGFP

rsEGFP rsGreen 1 2 0.7 0.7b 0.8 0.9 1 F a λex (nm) 489 491 478 487 489 487 485 486 485 a λem (nm) 509 510 503 511 511 511 509 509 511 a λabs off (nm) NA 396 408 396 396 396 396 396 390 3 -1 -1 b εon (10 M cm ) 54 57 53 62 53 61 66 61 52 52 46 52 εon, overall(103 M-1 cm-1)b 48 44 49 63 58 41 [53]c [47]d [61.3]e εoff (103 M-1 cm-1)f NA 17.9 20.3 20.4 19.7 18.7 20.5 19.5 18.2 61 42 34 QYfluo (%)g 40 43 38 41 42 39 [60]c [36]d [30]e 5.9 6.7 5.7 pKa 6.7 6.7 6.7 5.7 5.9 6.7 h d [5.8] [6.5] [5.8]e molecular brightnessi 100 61 56 61 60 59 81 77 50 j E. coli fluorescence 100 14 100 304 142 374 343 392 405 j HEK293T fluorescence 100 14 24 59 ND 59 36 42 36 -3 -1 k renaturation (10 s ) 5.5 7.2 7.0 9.2 9.6 9.6 12 12 11 -4 -1 l maturation (10 s ) 14 0.002 0.5 1.4 ND ND 1.4 0.9 1.1 m resid. off-state fluo., pure FP (%) 100 0.3 0.7 5.6 0.3 5 0.9 0.9 0.3 m resid. off-state fluo., cell (%) ND 1.1 2.3 4.7 1.0 4.4 1.5 1.6 0.6 n switching fatigue (% resid. fluo.) NA 52 95 32 51 29 46 35 50 o switching halftime (s) NA 9.71 5.78 20.10 10.56 17.04 10.13 10.10 6.66 -3 p QYOFF (10 ) NA 5.5 8.9 2.6 5.3 3.0 3.8 4.2 8.7 -1 p QYON (10 ) NA 1.7 1.2 1.2 1.6 1.5 1.3 1.4 1.8 a Wavelength of maximal excitation (ex) and emission (em) in the on-state and the absorbance (abs off) maximum in the off-state. bExtinction coefficient of the on-state chromophore determined at λex at pH 9 (εon) and corresponding to the total FP content at pH 7.4 (εon, overall). cReported in 31. dReported in 26 e . Reported in 36. fExtinction coefficient of the off-state chromophore determined at 405 nm. g Quantum yield of fluorescence. hReported in 32. iReported as εon overall × QYfluo normalized to 100 for EGFP. jNormalized total detected per-colony/cell fluorescence emission. kAmplitude averaged renaturation rates. lMaturation rates measured at 37°C using in vitro expression. mResidual fluorescence of the off-state ((off/on)*100) for purified proteins and HeLa cells. nResidual fluorescence in the onstate after 500 completed switching cycles. oTime at which the fluorescence is reduced to half its initial value. pQuantum efficiency of OFF-switching and ON-switching. NA = not applicable. ND = not determined. Values between square brackets are taken from the cited literature. The methods section contains further details on the listed parameters.


Page 9 of 41

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Nano

FLUORESCENCE INTENSITY IN PROKARYOTIC AND EUKARYOTIC CELLS. After the mutagenesis and initial screening of the ferritin-FP constructs, all rsGreens were subcloned in pRSETb. Expression of the generated FPs without fusion partners in E. coli at 37°C showed fluorescence intensity increases in line with those of the ferritin-FP constructs used during the mutagenesis. Measured at the colony level, bacterial cells expressing rsGreen0.7 are approximately 20 times brighter than rsEGFP, while bacteria expressing rsGreen0.7b are only 10 times more fluorescent than rsEGFP. Expression of the other members of the rsGreen series in E. coli resulted in almost 30-fold more fluorescence emission than the ancestor protein, rsEGFP, and up to 4 times more fluorescence signal than rsEGFP2 (Figure 1A & Table 1: E. coli fluorescence). The relative increase in fluorescence intensity was less drastic at lower temperatures, mostly because of an increased performance of rsEGFP. At these lower temperatures, rsEGFP performs almost on par with the rsGreen mutants (Figure 1A). To verify that this performance increase was not limited to FPs expressed in prokaryotic systems, we transfected HEK293T cells with pcDNA3::FP constructs encoding cytosolic expression of the FPs, and assayed the overall fluorescence intensity of individual cells using flow cytometry.41 Healthy cells were selected based on morphology and 7-AAD death cell staining. Between 20 and 25% of living cells expressed fluorescent proteins. Once again, all generated mutants performed considerably better than rsEGFP(2) with the increase in cell fluorescence ranging from 2.5 to 4.5 times that of rsEGFP and 1.5 to 3 times that of rsEGFP2 (Figure 1B & Table 1: HEK293T fluorescence). Taken together, our probes thus result in a strong overall increase in fluorescence signal when expressed in E. coli and HEK293T cells, with a stronger relative improvement in E. coli.


ACS Nano

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 41

Figure 1: (A.) Bacterial plates with colonies expressing different FPs and the average colony fluorescence (n = 29-81) in relative fluorescence units (RFU). Detection settings were optimized for each growth temperature. Measurements for a single temperature were performed on a single bacterial plate to ensure identical growth conditions for the different FPs. (B.) Average fluorescence output of live HEK293T cells (n = ~4200 to ~8500) measured using flow cytometry. Error bars in (A.) and (B.) represent the standard error on the mean. SPECTROSCOPIC PROPERTIES AND pH DEPENDENCE. All of the developed proteins are green FPs absorbing maximally between 485-489 nm, with maximal emission between 509-511 nm (Figure 2 & Table 1: λex and λem). All rsGreens display a characteristic absorbance peak with a maximum between 390 and 396 nm when switched to the non-


Page 11 of 41

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Nano

fluorescent state through irradiation with light (Table 1: λabs

off

and Figure S3, supporting

information). We determined three different extinction coefficients: the intrinsic extinction coefficient of the on-state chromophore (εon), the overall extinction coefficient at pH 7.4, (εon, overall),

which describes the absorption at the excitation maximum as a function of the total

(analytical) FP concentration and accounts for the pKa, and the extinction coefficient of the chromophore in the off-state (εoff). The overall molar extinction coefficients for all developed RSFPs were determined to be between 41000 and 63000 M-1 cm-1 (pH 7.4, λmax), corresponding well with the reported and measured extinction coefficients for EGFP, rsEGFP and rsEGFP2 (Table 1: εon, overall). The fluorescence quantum yield values for the developed RSFPs lay between 0.38 and 0.44 and are also highly similar to rsEGFP (Table 1: QYfluo). The pKa of the different FPs could be divided into two groups. The first group had a pKa value of 5.7-5.9, close to the pKa of EGFP (5.9), and comprised rsEGFP2, rsGreen0.9 and rsGreen1. The pKa of the second group, which consists of rsEGFP, rsGreen0.7, rsGreen0.7b, rsGreen0.8 and rsGreenF was higher, with a value of 6.7 (Table 1: pKa and Figures S4 & S5, supporting information). This difference in pKa value is clearly noticeable in the absorbance spectra measured at neutral pH, where a pronounced absorbance peak around 390-400 nm is present due to the neutral, protonated chromophore (Figure 2A & Figure S4, supporting information).


ACS Nano

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 41

Figure 2: Normalized absorbance (A.), excitation (B. dashes) and emission (B. full lines) spectra of rsEGFP and the members of the rsGreen series in the on-state at pH 7.4. BIOLOGICAL PERFORMANCE Because there are no major differences between the spectroscopic properties (e.g. QYfluo, ε, λex ,λem) of the switchable FPs used here, the improved fluorescence signal in cells can only be explained by large changes in the population of fully functional fluorophores.42 To get a feel for the formation efficiency of the FPs used, we calculated the ratio of the measured fluorescence emission in E. coli and HEK293T cells to the molecular brightness (εon, overall × QYfluo), and scaled it relative to the same ratio for EGFP (Table S2, supporting information). Assuming that the spectroscopic properties of the FPs do not change radically when expressed in cells, a higher ratio corresponds to a greater number of successfully formed FPs and thus a more effective formation and/or stability. We find a marked improvement 11 ACS Paragon Plus Environment

Page 13 of 41

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Nano

in our rsGreen mutants in E. coli (ratios between 2.4 and 8.0) and HEK293T cells (ratios between 0.4 and 1.0) compared to rsEGFP (both 0.2) and rsEGFP2 (1.8 in E. coli and 0.4 in HEK293T cells). Interestingly, we found that several of our photochromic labels outperform EGFP in E. coli. We analyzed the folding and maturation behavior of the different FPs to verify if these were indeed the cause of the improved fluorescence emission when expressed at 37°C. Renaturation experiments revealed only minor differences in the folding kinetics of the different probes (Figure 3, Table 1: renaturation and Table S2, supporting information). The amplitude weighted average rate constant of rsEGFP (7.2 × 10-3 s-1) is very similar to that of EGFP (5.5 × 10-3 s-1) and only slightly slower than the folding behavior of the “superfolding” rsGreen variants (9.2 to 12 × 10-3 s-1). However, this experiment only describes the refolding of the protein with an alreadyformed chromophore, and does not capture the complete folding and chromophore formation process.43–45 We used the PURExpress in vitro protein expression kit to mimic the complete formation process of FPs, and followed the increase in fluorescence in time.46 The results are summarized in Figure 4 and Table 1: maturation. While an increase in fluorescence was apparent for all proteins, it was difficult to obtain quantitative information for the rsEGFP formation due to the very low signal. However, the process is clearly slower than that of the rsGreen mutants. The reduced biological performance of rsEGFP was further exemplified by SDS-PAGE analysis of E. coli lysates, showing very little rsEGFP protein content compared to the rsGreens or EGFP under identical growth conditions (Figure S6, supporting information). Likewise, the rsEGFP content of E. coli cultures grown at 37°C was much lower than in rsEGFP cultures grown at 20°C. These results suggest an efficient degradation of the formed rsEGFP polypeptides. While EGFP


ACS Nano

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 41

outperforms the rsGreens in terms of formation rate, the presence of the superfolder mutations overcomes the deteriorating effects of the rsEGFP mutations, leading to comparable formation rates of the rsGreens measured in this work and sfGFP, reported in the study of Iizuka and coworkers.46 To verify that the renaturation and in vitro expression experiments were not influenced by offswitching of the fluorophores due to the probing light, we repeated the experiments using purified, non-denatured RSFPs at similar concentrations using identical illumination schemes. We confirmed that the used settings did not cause detectable off-switching (data not shown). Taken together, our folding/renaturation and in-vitro expression experiments do not fully explain the observed live-cell fluorescence signals shown in Figure 1. For example, EGFP displays the fastest in-vitro expression kinetics (Figure 4), yet displayed lower fluorescence levels compared to our rsGreens when expressed in E. coli grown at 37ºC (Figure 1). In particular, other effects such as rates of protein expression or protein degradation pathways, possibly activated by protein misfolding or aggregation and influenced by the folding assisting machinery, are likely to play an important role in living cells.47–50 Because all our rsGreens are based on mEGFP (EGFP-A206K), we expect that they are monomeric at physiologically relevant concentrations40, which was confirmed for rsGreen0.7 and rsGreen0.7b by size exclusion chromatography (Figure S7, supporting information). As further controls we created constructs in which the FPs were fused to tubulin and expressed in HeLa cells. Tubulin localization is known to be strongly affected by oligomerization of the FP.51 We found that all proteins localized as expected and showed typical monomeric behavior in fusion constructs with α-tubulin (Figure S8, supporting information). We further confirmed this in Nterminal fusions of the FPs to the NH2-terminal residues of the Lyn kinase (Lyn tag), whose


Page 15 of 41

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Nano

targeting to sphingolipid- and cholesterol-enriched microdomains in the plasma membrane is also disturbed by fusions to oligomeric FPs.40,52

Figure 3: (A.) Renaturation kinetics after denaturation in urea and dilution in renaturation buffer. The return of fluorescence was followed in time. (B.) Initial kinetics of the renaturation show only minor differences between the studied FPs. Dots represent measured data and full lines show the calculated double exponential fits.

Figure 4: Formation kinetics measured at 37°C using in vitro protein synthesis (PURExpress kit). Fluorescence emission was detected every 30 s. Dots represent measured data and full lines show the calculated mono-exponential fits. Data was normalized to the value of the fit at t = ∞. The inset shows the slow kinetics of rsEGFP.


ACS Nano

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 41

PHOTOCHROMIC BEHAVIOR. We characterized the photochromic behavior of the proteins both in purified protein solutions and in HeLa cells. Investigating the photoswitching of the purified proteins allowed the identification of their intrinsic properties such as the quantum efficiency for off- and on-switching.25 rsGreen0.7 and rsGreen0.8 show the lowest efficiency in terms of off-switching, approximately two times slower than rsEGFP, rsGreen0.7b, rsGreen0.9 and rsGreen1. rsEGFP2 is the most efficient photoswitcher, switching off almost twice as efficient as rsEGFP in this assay, but is very closely followed by rsGreenF. The quantum efficiencies of off-switching also correspond nicely with the off-switching half times and contrast values (Table 1: resid. off-state fluo., pure FP; switching halftime; QYOFF; Figure 5A and Figures S3 & S9, supporting information). rsGreen0.7, the least efficient rsGreen, displays an off-switching quantum efficiency that is 16 times higher than that of Dronpa, while the off-switching quantum efficiency of rsGreenF is almost 55 times higher. This range illustrates the possibility of tweaking the photoswitching speed to satisfy the needs of the experiment, optimizing for instance photoswitching speed versus the number of photons emitted per cycle.53 In terms of on-switching, the quantum efficiencies are remarkably similar for all RSFPs (Table 1: QYON), including Dronpa,25 indicating that this is a very efficient and fast process in all negative type RSFPs. To assay the resistance to photoswitching fatigue, we performed experiments on transfected HeLa cells illuminated using a widefield microscope. All proteins were subjected to 500 photoswitching cycles using the same illumination settings. After 500 cycles, rsEGFP2 showed the least signs of photoswitching fatigue. From the rsGreen series, the slower switchers, rsGreen0.7 and rsGreen0.8, are the least resistant, while several other members display similar performance as rsEGFP in terms of switching fatigue resistance. The contrast values obtained


Page 17 of 41

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Nano

from these measurements correspond nicely with the data from the purified proteins (Table 1: resid. off-state fluo., cell; switching fatigue; Figure 5B and Figure S10, supporting information). Since we characterized multiple closely related FPs for this study, we also investigated the presence of correlations between the different spectroscopic parameters. We found that the resistance to photoswitching fatigue correlates well with the off-switching rate and efficiency. This correlation suggests that faster off-switchers are more resistant to switching fatigue, a finding that is reasonable considering that these probes undergo fewer excitation cycles before converting to the weakly- or non-absorbing off-state, in which they are protected from photodegradation. In addition, this correlation appeared to be linear (Figure S11, supporting information), which suggests that it is simply the switching efficiency that determines the fatigue resistance of the probes, rather than an intrinsic resistance to the bleaching process. We also observed a negative correlation between the extinction coefficient of the off-state proteins and the quantum yield of on-switching, which might be explained by an increased stabilization of the non-fluorescent state (lower QYON) with consistent increase in extinction coefficient. (Table 1: εoff & QYON and Figure S11, supporting information)


ACS Nano

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 18 of 41

Figure 5 (A.) Photoswitching of purified fluorescent protein solutions. Protein solutions were illuminated with 488 nm (off-switching, ~0.5 W/cm²) or 405 nm (on-switching, ~0.035 w/cm²) laser light and absorbance and emission spectra were recorded every 5 seconds. Displayed traces represent the integrated emission spectra for each time point. (B.) Photoswitching fatigue of the RSFPs measured on HeLa cells transfected with pcDNA3::RSFP. Cells expressing RSFPs were illuminated with violet (on-switching) or cyan (off-switching) LED light (Lumencor). The residual fluorescence after 500 switching cycles was used to assay the resistance to photoswitching fatigue. Displayed traces represent the residual fluorescence after the indicated amount of photoswitching cycles. STRUCTURE DETERMINATION. To gain more insight into the photochromic and biological behavior, we determined the crystal structure of rsGreen0.7 in the fluorescent on-state and the non-fluorescent off-state to a resolution of 1.25 Å and 1.20 Å, respectively (Table S4, supporting information). rsGreen0.7 shows a similar overall structure as other GFP-like FPs, with a typical 11-stranded barrel fold and a central helix holding the chromophore in place (Figure S12, supporting information).38,54 All mutations affecting the photoswitching behavior are located near the chromophore and are oriented inwards. This includes the original rsEGFP mutations (Q69L, V150A, V163S and S205N)26 and the additional Y145F mutation. Mutated residues mainly affecting the formation and/or stability are located further away from the chromophore, often on the outside of the β-can and near the loops (Figure S12, supporting information). The chromophore of rsGreen0.7, which comprises threonine 65, tyrosine 66 and glycine 67, adopts a cis-conformation in the fluorescent state, analogous with the vast majority of fluorescent proteins. The high resolution of the crystallographic data reveals a heterogeneous chromophore conformation, as multiple conformers of the p-hydroxyphenyl group can be fitted to the electron 17 ACS Paragon Plus Environment

Page 19 of 41

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Nano

density, here modelled with two conformations. Anisotropic refinement using a single conformer further demonstrates an increased heterogeneity of the p-hydroxyphenyl in the on-state compared to sfGFP and EGFP, in which the anisotropically refined B-factors have similar sizes over the entire chromophore (Figure 6). This suggests a significant increase in the flexibility of the chromophore in the on-state compared to EGFP and sfGFP, which is a likely explanation for the reduced fluorescence quantum yield. At the same time, we speculate that this increased flexibility is a consequence of destabilizing the interaction between the chromophore and its environment, and ultimately favors the photoswitching process.21,55–58 The cis-conformation of the on-state chromophore isomerizes into a trans-conformation when the protein switches to the non-fluorescent state (tilt and twist angles of the cis-conformation: 10.7° and -11.3°, respectively; tilt and twist angles of the trans-conformation: -161.4° and -60.9°, respectively). In this conformation the original position of the p-hydroxyphenyl moiety is occupied by the imidazole moiety of histidine 148 hydrogen bonded to water molecule 562 (Figure 7). However, due to steric hindrance of phenylalanine 165, the chromophore is unable to isomerize in its original plane, but makes an angle of approximately 62° compared to the plane of the on-state chromophore (Figure S13, supporting information). The large chromophore rearrangement associated with the photochromism is made possible by the rsEGFP mutations at positions 69 and 150. When the off-state chromophore of rsGreen0.7 is modeled in the (non-mutated) EGFP chromophore environment, unfavorably close contacts appear with the glutamine at position 69 (~2.1 Å) and the valine at position 150 (1.4-1.8 Å) which effectively prohibits this from occurring. The Q69L mutation in the rsGreen0.7 increases the distance from this residue to the chromophore to a minimum of 3.3 Å, while the introduction of an alanine at position 150 increases the distance to the p-hydroxyphenyl moiety to 3.0 Å,


ACS Nano

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 20 of 41

providing ample space for the isomerization of the chromophore (Figure S13, supporting information). A striking feature of the off-state is the presence of a water molecule (W510) that allows for a hydrogen bonding network between the chromophore, serine 163 and glutamine 183, stabilizing the off-state chromophore (Figure S14, supporting information), facilitated by the introduction of serine 163, a key mutation in all EGFP based RSFPs. This water molecule is not resolved in the on-state crystal structure nor is it found in the structures of EGFP or sfGFP. The latter FPs also differ on position 163 having a valine and alanine respectively and are thus incapable of forming similar hydrogen bonds. We inspected the published on- and off-state crystal structures of Dronpa16,57, pcDronpa25, IrisFP23, rsTagRFP59,60, Padron19, Dreiklang61, asFP59562,63 and mTFP0.764 and found that in all of these proteins there are stabilizing hydrogen bonds of the phydroxyphenyl oxygen atom of the chromophore with water molecules or surrounding residues in the non-fluorescent state (Table S5, supporting information). Taken together, our results strongly indicate that well-performing FP photochromism requires both the availability of sufficient space to allow cis-trans isomerization, combined with the presence of stabilizing interactions that slow down the spontaneous recovery to the cis-state. The fourth mutation introduced during the development of rsEGFP was an asparagine at position 205 instead of a serine. Our X-ray data revealed an outward backbone shift of 1 Å in this area when comparing the on-state rsGreen0.7 with EGFP and sfGFP (Figure 7 and Figure S15, supporting information). This outward backbone shift is necessary to accommodate the larger amide moiety of the asparagine side chain, while maintaining the hydrogen network with a water molecule (W591) and the chromophore. Interestingly, the backbone shifts back to the original position occupied in EGFP and sfGFP when rsGreen0.7 is switched off. This backbone shifting can act on the photoswitching behavior in different ways. The outward shift in the on-state 19 ACS Paragon Plus Environment

Page 21 of 41

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Nano

potentially creates more room for the chromophore to isomerize, leading to efficient offswitching. When the backbone is shifted back inwards in the off-state, the presence of the larger amide moiety might create an additional barrier that the chromophore has to overcome to switch back on, increasing the stability of the off-state. In rsGreen0.7, we found a slowing effect on the photoswitching caused by the Y145F mutation, which was confirmed by reverting this residue to the original tyrosine (rsGreen0.7b). The importance of residue 145 in the photoswitching process is further exemplified by the fast photoswitching of rsGreenF compared to rsGreen0.9, differing only by the F145L mutation. Because in rsEGFP and rsGreen0.7b, position 145 is occupied by a tyrosine, which can be incorporated into polar interactions with neighboring amino acids, such as His148 or Asn205, we speculate that the phenylalanine present in rsGreen0.7 is unable to make these interactions, while remaining bulky, and may therefore result in a reduced photochromism. In the case of rsGreenF, Leu145 is also unable to make polar interactions, but is more flexible than phenylalanine and may thus allow faster photoswitching. Interestingly, residue 145 also affects the observed pKa of the FPs with a pronounced difference between rsGreen0.9 (fast switcher, Phe145, pKa = 5.7) and rsGreenF (very fast switcher, Leu145, pKa = 6.7). However, the photoswitching rate is not solely affected by chromophore pKa, as illustrated by the comparison of rsGreen0.7 (slow switcher, Phe145, pKa = 6.7) with rsGreen0.7b (fast switcher, Tyr145, pKa = 6.7). The structural mechanisms responsible for the superfolding behavior are difficult to determine since a static crystal structure provides no information on the maturation process. However, multiple mechanisms likely influence both the rate of the formation and the formation efficiency. Most of the mutations that we introduced in our rsGreens are located on the outside of the protein, away from the chromophore, and therefore enhance the biological performance rather than the intrinsic spectroscopic parameters. This may involve the creation of electrostatic 20 ACS Paragon Plus Environment

ACS Nano

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 22 of 41

networks that stabilize the barrel structure or a reduction of the overall hydrophobicity and proneness to aggregation.38 Such a reduction in aggregation could explain why our rsGreens form slower than EGFP, yet achieve a higher relative fluorescence signal (higher functional FP content) when expressed in E. coli. A complete discussion of the superfolder mutations can be found in the supporting information.

Figure 6: (A.) Two modeled conformations of the on-state chromophore in rsGreen0.7 (left) versus one conformation for the EGFP chromophore (PDB ID: 4EUL, right) as observed in the 2Fo-Fc electron density map contoured at 1.0 σ. (B.) Displacement ellipsoids drawn at 50% probability level after anisotropic refinement with a single chromophore conformation for rsGreen0.7 (left) and EGFP (right).


Page 23 of 41

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Nano

Figure 7: Comparison of the green-on (green) and green-off state (grey) of rsGreen0.7. During photoswitching several structural rearrangements occur: 1: cis-trans isomerization of the chromophore; 2: backbone shift in residues 204, 205 and 206; 3 and 4: conformational change of threonine 203 and histidine 148; 5: disruption of a hydrogen bonding network between asparagine 205, water molecule 591 and the chromophore, and the formation of a new hydrogen bonding network between serine 163, water molecule 510 and the chromophore. (Numbers do not imply chronological ordering). SUB-DIFFRACTION MICROSCOPY. We evaluated the use of the developed RSFPs for sub-diffraction microscopy by employing these labels in pcSOFI8,65 and RESOLFT26,61 experiments. pcSOFI experiments were successful in TIRF mode using the Lyn tag described above (Figure 8 and Figure S17, supporting information), and in epi mode using fusions of the RSFPs targeted to the outer membrane of mitochondria (Figure S18, supporting information).66 22 ACS Paragon Plus Environment

ACS Nano

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 24 of 41

We found that a two-fold resolution enhancement was readily achievable both in TIRF and epi illumination mode, yielding a sub-diffraction spatial resolution together with improved contrast and background reduction. These second order SOFI calculations could be performed on small datasets, achieving a temporal resolution of less than 7.5 seconds (Figure 8). Several samples also allowed analysis using a third-order cumulant, further improving the resolution enhancement and providing additional pixels, created by the calculation of cross-cumulants.8 This allows the more detailed visualization of fine structures present in the samples. Because the third-order SOFI analysis is more susceptible to noise, not all samples were suitable for this analysis, and most samples achieving threefold improved resolution were acquired using TIRF-mode excitation.

Figure 8: Sub-diffraction imaging with pcSOFI. (a-f) Averaged widefield, (g-j) second order and (k,l) third order SOFI images of HeLa cells labeled with Lyn-tagged rsGreen0.8 or rsGreen1, imaged using TIRF-mode excitation. (m,n) Intensity line profiles of cross sections indicated by 23 ACS Paragon Plus Environment

Page 25 of 41

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Nano

the arrowheads. Images were calculated on 250 (c,e,g,i) or 1400 frames (a,b,d,f,h,j,k,l) acquired at 30 msec / frame. (c-l) Expansions of the boxed areas in (a) and (b). Scale bars = 10 µm (a,b), 1 µm (c-l). We reasoned that rsGreenF, which almost rivals rsEGFP2 with respect to switching speed, should also facilitate RESOLFT nanoscopy. To test this, we recorded RESOLFT images of HeLa cells transfected with vimentin-rsGreenF (Figure 9). We measured filaments widths of down to ~70 nm in living cells. Although the obtained resolution is less than previously demonstrated with rsEGFP236, which is presumably due to the reduced switching resistance of rsGreenF, we could discern details entirely obscured in the corresponding diffraction-limited confocal image. We conclude that the rsGreen proteins represent a versatile family of expression enhanced RSFPs, which can be used for different super-resolution approaches depending on the question addressed.

Figure 9: RESOLFT nanoscopy of living HeLa cells expressing vimentin-rsGreenF. Comparison of diffraction-limited confocal microscopy (left) with RESOLFT nanoscopy (right). Shown are raw image data, no background subtraction or other image processing was applied. The line


ACS Nano

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 26 of 41

profiles display the normalized fluorescence intensities in the raw confocal and the corresponding raw RESOLFT image at the sites indicated by the arrows. Each profile is an average across three adjacent line profiles. Scale bar: 1µm. CONCLUSION In conclusion, we have developed a strategy to enhance the formation and biological tolerance of photochromic fluorescent proteins while retaining their spectroscopic parameters. We suggest performing two rational and evolutionary optimizations in parallel, one focusing on introducing photochromism while the other focuses on enhancing folding, maturation and stability (biological performance). This reduces a difficult two-dimensional optimization problem into two much simpler one-dimensional optimizations, which can be combined to achieve both, as this work shows. Using this generic approach, we developed a series of new RSFPs, which we named rsGreens. These EGFP-based proteins exhibit a range of photochromic properties and overall improved biological performance compared to the related proteins rsEGFP and rsEGFP2. To gain more knowledge about the requirements for reversible photoswitching, we determined the first ever crystal structures of a negative reversibly switchable avGFP based protein. The crystal structures of the fluorescent and non-fluorescent state of rsGreen0.7 provided insights into the improved formation efficiency and the photoswitching process. We were able to identify several important factors for the introduction of reversible photoswitching in avGFP based FPs, including the creation of space around the chromophore to facilitate its isomerization and the introduction of stabilizing interactions for the non-fluorescent state through hydrogen bonding. We further show that these labels are well suited to sub-diffraction microscopy such as pcSOFI achieving a resolution down to ~80 nm and RESOLFT obtaining a resolution of ~70 nm. rsGreen0.7 and rsGreen0.8 show the potential for efficient but slower switching EGFP-based RSFPs. The final


Page 27 of 41

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Nano

mutants, rsGreen1 and rsGreenF, show fast and efficient reversible photochromism, with rsGreenF coming close to rsEGFP2 in terms of photoswitching speed, but being outmatched in photoswitching fatigue resistance. All rsGreens do exceed rsEGFP and rsEGFP2 in terms of in cell fluorescence indicating an increase in expression and formation of functional FPs. This combination of robust biological performance and photochromism suitable for sub-diffraction microscopy makes the rsGreens ideally suited for use in demanding experiments across a wide range of advanced fluorescence imaging applications. METHODS Cloning, mutagenesis, expression and purification The gene encoding rsEGFP in a pQE31 expression vector was purchased from Abberior GmbH. The ferritin-rsEGFP fusion construct was created by inserting the PCR-amplified rsEGFP with an N-terminal AGSAAGSG peptide linker between KpnI and HindIII sites of a pRSETb vector. An EcoRI restriction site was included in the linker sequence for easy cloning during the mutagenesis. The DNA sequence encoding bullfrog red-cell ferritin subunit-H was ordered from Integrated DNA Technologies (IDT), PCR-amplified and cloned between the BamHI and KpnI restriction sites. Directed mutagenesis of selected residues was performed according to a modified QuikChange protocol.67 All primers used for directed mutagenesis (Table S3, supporting information) were designed using the QuikChange Primer Design tool (Agilent Technologies) and ordered from IDT. For site-saturation purposes, the target codon was replaced with an NNK sequence for optimal coverage of all possible alternative residues. Random mutagenesis was performed using error-prone PCR with a mutagenic buffer.68 The quality of the generated PCR products was assayed using agarose gel electrophoresis and 26 ACS Paragon Plus Environment

ACS Nano

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 28 of 41

successful PCR products, defined by a distinct band at the desired length (+/- 750bp) and a smear of longer products, were pooled and used to replace the rsEGFP sequence in the ferritin-rsEGFP plasmid, using EcoRI and HindIII restriction sites. DNA shuffling of rsGreen0.7 and rsEGFP was carried out according to the Staggered Extension Process (StEP).69 A total of 10 ng combined plasmid DNA was used to generate short PCR fragments in 100 cycles of 30 s at 94°C and 5 s at 55°C. The length of the generated DNA fragments was checked using agarose gel electrophoresis. 2 µL of the generated fragments was treated with DpnI (Thermo Scientific) to reduce background clones and subsequently used as template in a second PCR generating full size, shuffled genes, which were used to replace the rsEGFP sequence in the ferritin-rsEGFP plasmid, using EcoRI and HindIII restriction sites. The screening and selection of interesting clones was carried out as described in Supporting Method 1. (E. coli fluorescence) For recombinant expression purposes all fluorescent protein coding sequences were cloned between BamHI and EcoRI restriction sites of a pRSETb plasmid in frame with a polyhistidine sequence. All plasmid transformations were carried out using a sonoporation protocol. Chemocompetent cells were incubated with plasmid DNA on ice for 5-15 minutes, placed in a Branson 2210 ultrasonic cleaner for 15 seconds, incubated with SOC medium at 37°C for 20 minutes and plated on suitable bacterial agar plates. Expression and purification was carried out as described in Supporting Method 2. Eukaryotic expression vectors were created as follows. PCR-amplified RSFP genes were inserted between BamHI and EcoRI restriction sites of empty pcDNA3, pcDNA3-Lyn and pcDNA3-ᴅ-AKAP1a to generate cytosolic, membrane targeted and mitochondria targeted expression respectively. For the visualization of human α-tubulin, the mVenus gene was replaced by the PCR amplified gene encoding the desired RSFP in a pcDNA3::mVenus-α-tubulin plasmid 27 ACS Paragon Plus Environment

Page 29 of 41

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Nano

using restriction cloning (HindIII and BamHI sites). prsGreenF-vimentin was created by exchanging the mKate2 sequence in the plasmid pmKate2-vimentin with the PCR amplified rsGreenF sequence using AgeI and NotI restriction sites. All primer sequences are listed in Table S3 (supporting information). All generated constructs were validated by sequencing (LGC Genomics). Spectra, molecular brightness & pKa determination Absorbance spectra were recorded with a Shimadzu UV-1650PC spectrophotometer. (λex and λabs off) Fluorescence excitation and emission spectra were determined with a PTI Quanta-Master fluorimeter with slits set to 2 nm. (λem) Spectra were measured in PBS (pH 7.4). Extinction coefficients were measured according to Ward’s method70 using the literature values of EGFP, rsEGFP, rsEGFP2 and Dronpa as a reference. The on-state extinction coefficient is reported at pH 9 to ensure complete deprotonation of the chromophore. (εon) The overall on-state extinction coefficient is reported at pH 7.4. (εon, overall) The off-state extinction coefficient was determined by comparing the absorbance at λex of the on-switched RSFPs with the absorbance at 405 nm for the off-switched RSFPs. (εoff) The quantum yield of fluorescence was determined relative to fluorescein in 0.1 M NaOH (=0.925). (QYfluo) Absorbance and emission spectra were measured quasi-simultaneous on the setup for photoswitching analysis. The molecular brightness is defined as the product of the overall extinction coefficient and the quantum yield, scaled to a value of 100 for EGFP. pKa values of the FPs were determined as the inflection point of the sigmoidal curve fitted to the absorption maxima of either the neutral or anionic species at pH values ranging from 3 to 12. (pKa) Photoswitching efficiency & Photoswitching Fatigue


ACS Nano

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 30 of 41

The photoswitching behavior of purified RSFPs was investigated as previously described.25 In brief, proteins were diluted in 1 mL TN buffer (100 mM Tris, 300 mM NaCl at pH 7.4), to an optical density of less than 0.2 at the absorption maximum. The cuvette (Hellma) was kept at 4°C and continuously stirred. For off-switching, the sample was irradiated with 488 nm laser light (0.48 W/cm²) for 100 seconds in steps of 5 seconds, while on switching used 405 nm light (0.034 W/cm²) and the same time protocol. After every step, an absorbance and emission spectrum were recorded. The quantum yield of photoswitching was calculated as described earlier.25 (resid. offstate fluo., pure FP; switching halftime; QYON and QYOFF) Photoswitching fatigue and high power photoswitching was measured on a setup consisting of a Sola Light Engine (Lumencor) coupled into an inverted microscope (Olympus IX71) equipped with a zt488rdc dichroic mirror (Chroma) and a 10x objective (UplanSApo, Olympus). Fluorescence images were recorded using an EMCCD camera (iXon, Andor). The system was used to image HeLa cells transfected with pcDNA3::RSFP for untargeted expression (cell culture and transfection described in Supporting Method 3). The proteins in the cells were completely switched off in 5 steps of 0.5 seconds with cyan light (20%) and switched back on in 5 steps of 0.5 seconds with violet light (0.5%). This cycle was repeated 500 times to assess the decrease in on-state fluorescence after repeated switching. (switching fatigue) The first switching cycle also provided information about the switching behavior under more intense illumination. For this purpose, the on- and off- state were compared after subtraction of the background signal. (resid. off-state fluo., cell) For the measurement of Dronpa, the illumination scheme was adjusted to 5 steps of 3 seconds cyan light (100%) for off-switching and 5 steps of 0.5 seconds violet light (0.5%) for on-switching. This cycle was repeated 80 times. SDS-PAGE, renaturation and in vitro expression


Page 31 of 41

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Nano

To determine the FP expression levels with SDS-PAGE analysis, a single colony from a fresh culture plate was inoculated in 2 mL LB medium supplemented with ampicillin and grown overnight in a shaking incubator at 37°C. 25 µL of this starting culture was used to inoculate 2 mL of fresh LB medium supplemented with ampicillin and grown for 4 hours at 37°C, after which 500 µL of this culture was used to inoculate 1.5 mL of fresh LB medium supplemented with ampicillin. After overnight growth, the E. coli were collected by centrifugation, washed twice with 500 µL PBS and finally resuspended in 500 µL PBS supplemented with 1 mg/mL lysozyme and 25 U/mL DNAse. Cells were opened by repeated freezing and thawing. Cellular debris was collected by centrifugation and discarded.. 15 µL of supernatant was mixed with 15 µL 2x SDS Loading buffer and placed in boiling water for 5 minutes. 25 µL of this sample was loaded and separated on a SDS-PAGE gel (4% stacking / 12% resolving) according to manufacturer’s instructions. Denaturation - renaturation experiments were carried out as described earlier.71 Briefly, 1 µL of a fluorescent protein solution in PBS was added to 9 µL denaturation solution (8 M urea, 1 mM DTT) and incubated at >95°C for 5 minutes. Subsequently, the denatured protein was added to 990 µL renaturation solution (35 mM KCl, 2 mM MgCl2, 50 mM Tris, 1 mM DTT, pH 7.4). The recovery of fluorescence emission (detection at 510 nm) upon 488 nm excitation was followed for 500 sec in a PTI QuantaMaster fluorimeter. A maximal value was obtained by measuring the fluorescence emission of an identical dilution of non-denatured protein. (renaturation) The maturation speed was measured, based on a previously described method,46 making use of the PURExpress in vitro protein synthesis kit (NEB). 250 ng plasmid DNA was mixed with solution A and B according to manufacturer’s guidelines and was diluted to a total volume of 25 µL. The samples were incubated at 37°C for 15 minutes before being placed on ice to stop the 30 ACS Paragon Plus Environment

ACS Nano

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 32 of 41

reaction. The sample was diluted in 150 µL air-saturated maturation buffer (50 mM Tris-HCl, 2 mM MgCl2, 35 mM KCl, 5 mM b-mercaptoethanol, and 0.3 mg/mL chloramphenicol at pH 7.5) at 37°C and immediately measured in an Edinburgh 920FLS fluorimeter for 10000 seconds set at 490 nm excitation (480 for rsEGFP2). Emission was measured at 510 nm every 30 seconds. (maturation) Flow cytometry HEK293T cells were cultured, transfected and prepared as described in Supporting Methods 3. The brightness of HEK293T cells was measured on a FACSCanto (BD Biosciences) flow cytometry machine ~24 hours post transfection. Viable cells were gated based on morphology (FSC vs SSC), and green and red fluorescence (FP and 7-AAD respectively) was recorded until 10000 ‘living green cells’ were detected. Excitation was performed by a 20 mW 488 nm laser and fluorescence emission was recorded through a 670 LP filter for the PerCP channel (7-AAD detection), and through a 530/30 BP filter for the FITC channel (FP detection). Data Analysis was performed with Flowing Software 2 (http://www.flowingsoftware.com/). Raw data was imported and living cells were gated based both on morphology (SSC vs FSC) and 7-AAD death-cell staining. Green fluorescence was gated above the cell background level measured for the negative control. (HEK293T fluorescence) Microscopy pcSOFI experiments were carried out according to the protocol steps described in

65

on a

commercial cell^tirf microscope (Olympus) described elsewhere.25 For all experiments, the camera acquisition times were varied between the minimum exposure time (14.2 and 31 msec for the ImageEM X2 and the ImageEM camera respectively at full frame) and 100 msec, while the 31 ACS Paragon Plus Environment

Page 33 of 41

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Nano

direct EM gain was adjusted (4-900) for optimal dynamic range. Image series of 300-10000 frames were recorded for pcSOFI analysis, which was done with the Localizer package for IgorPro.72 For comparison, the average of the time series was calculated as the diffraction limited image. Conventional, static widefield images of tubulin labeled cells were acquired using simultaneous 405 nm and 488 nm illumination and an acquisition time of 0.2-1 sec. RESOLFT nanoscopy was performed using a modified “RESOLFT Quad P” microscope (Abberior Instruments, Göttingen, Germany). The microscope provides three co-aligned focal spots: two normally focused beams at wavelengths of 488 nm and of 405 nm for fluorescence excitation and Off-switching, respectively. One ‘doughnut-shaped’ focal intensity distribution with a central minimum (‘zero’) for Off-switching at the focal periphery in the xy-plane. The RESOLFT image was recorded by applying a pulse scheme, pixel by pixel, with three-fold line averaging. The following pulse scheme was used: i) On-switching with 405 nm (30 µs; 2 µW measured at the back focal plane of the objective lens). ii) Off-switching at the focal periphery with 488 nm (450 µs; 18 µW), and, iii, reading out fluorescence with 488 nm (20 µs; 6 µW). The diffraction-limited confocal image was recorded in exactly the same manner, but omitting the offswitching-step. No background subtraction or image processing was applied. Crystallography rsGreen0.7 cultures were grown and purified as described above. Afterwards, the rsGreen0.7 sample for crystallization was further purified from small impurities and imidazole by sizeexclusion chromatography using a HiLoad Superdex 200pg 16/600 column coupled to an Akta Purifier 10 system (both GE Healthcare) equilibrated and run with 0.1x TN buffer, and concentrated using a Vivaspin 10 000 MWCO (Sartorius Stedim).


ACS Nano

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 34 of 41

Several commercial screens (Hampton Research and Molecular Dimensions) were used to find crystallization hits for rsGreen0.7. Therefore, sitting drops consisting of 1 µl of protein solution (7 mg/ml) and 1 µl of precipitant were equilibrated against 100 µl precipitant in the well. The rsGreen0.7 green-on crystal presented here was grown in 6 months in 0.1 M Bis-Tris pH 6.5 and 25 % PEG 3350 at 16°C. Prior to the X-ray diffraction experiment, it was flash frozen in liquid nitrogen without any cryoprotectant. For the off-state, an on-state crystal which was grown in 0.15 M NaBr and 30 % PEG 2000 MME at 16°C for 12 months and was irradiated with 491 nm light before flash freezing. 25 % PEG 400 was used as cryoprotectant. X-ray diffraction data acquisition, analysis and structure refinement are described in Supporting Method 4. AUTHOR INFORMATION Corresponding Author [email protected] ACKNOWLEDGMENTS This work was supported by the Research-Foundation Flanders (FWO-Vlaanderen) via grants 1521915N, 1502314N, 1525113N and grants G0484.12 and G0B49.15, the Flemish government long-term structural funding ‘‘Methusalem’’ grant METH/08/04 CASAS, and the Hercules Foundation (HER/08/021). SD, EDZ, VG, BM, and WV thank the Agency for Innovation by Science and Technology (IWT) Flanders for a doctoral fellowship. SJ thanks the DFG Center for Nanoscale Microscopy and Molecular Physiology of the Brain for support. PD thanks the Research Foundation-Flanders for a postdoctoral fellowship and the KULeuven for a BOF-ZAP research professorship. EDZ and LVM thank the staff of PROXIMA1, Soleil (Paris), particularly


Page 35 of 41

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Nano

Pierre Legrand and of XRD1, Elettra (Trieste), particularly Maurizio Polentarutti for help with Xray data collection. SUPPORTING INFORMATION A discussion of superfolder mutations. Supporting Method 1: Mutagenesis screening. Supporting Method 2: Expression and purification. Supporting Method 3: Eukaryotic cell culture. Supporting Method 4: X-ray data collection, analysis and refinement. Table S1, mutations introduced through directed mutagenesis. Table S2, additional properties of the developed FPs. Table S3, primers used in this work. Table S4, crystallographic data collection and refinement statistics. Table S5, interactions of the on- and the off-state chromophore of known RSFPs. Figure S1, folding disruption by ferritin. Figure S2, AA sequence alignment of the developed rsGreens, EGFP, rsEGFP, rsEGFP2 and sfGFP. Figure S3, Absorbance and emission spectra during photoswitching. Figures S4 and S5, pH-dependence of rsGreens, EGFP, rsEGFP and rsEGFP2. Figure S6, SDS-PAGE results of E. coli culture extracts. Figure S7, size exclusion chromatography results of rsEGFP, rsGreen0.7 and rsGreen0.7b. Figure S8, rsGreen-Tubulin images. Figure S9, photoswitching behavior of rsEGFP, rsEGFP2 and the rsGreens. Figure S10, initial 25 photoswitching cycles of rsEGFP, rsEGFP2 and the rsGreens. Figure S11, correlations. Figures S12-S16, 19 and 20, observations made in the crystal structures. Figures S17 and S18, SOFI images obtained with rsGreen labeled cells. This material is available free of charge via the Internet at http://pubs.acs.org. REFERENCES (1)

Tsien, R. Y. The Green Fluorescent Protein. Annu. Rev. Biochem. 1998, 67, 509–544.

(2)

Dedecker, P.; De Schryver, F. C.; Hofkens, J. Fluorescent Proteins: Shine On, You Crazy Diamond. J. Am. Chem. Soc. 2013, 135, 2387–2402. 34 ACS Paragon Plus Environment

ACS Nano

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 36 of 41

(3)

Tiwari, D. K.; Nagai, T. Smart Fluorescent Proteins: Innovation for Barrier-Free Superresolution Imaging in Living Cells. Dev. Growth Differ. 2013, 55, 491–507.

(4)

Betzig, E.; Patterson, G. H.; Sougrat, R.; Lindwasser, O. W.; Olenych, S.; Bonifacino, J. S.; Davidson, M. W.; Lippincott-Schwartz, J.; Hess, H. F. Imaging Intracellular Fluorescent Proteins at Nanometer Resolution. Science 2006, 313, 1642–1645.

(5)

Hess, S. T.; Girirajan, T. P. K.; Mason, M. D. Ultra-High Resoluion Imaging by Fluorescence Photoactivation Localization Microscopy. Biophys. J. 2006, 91, 4258–4272.

(6)

Hofmann, M.; Eggeling, C.; Jakobs, S.; Hell, S. W. Breaking the Diffraction Barrier in Fluorescence Microscopy at Low Light Intensities by Using Reversibly Photoswitchable Proteins. Proc. Natl. Acad. Sci. U. S. A. 2005, 102, 17565–17569.

(7)

Dertinger, T.; Colyer, R.; Iyer, G.; Weiss, S.; Enderlein, J. Fast, Background-Free, 3D Super-Resolution Optical Fluctuation Imaging (SOFI). Proc. Natl. Acad. Sci. 2009, 106, 22287–22292.

(8)

Dedecker, P.; Mo, G. C. H.; Dertinger, T.; Zhang, J. Widely Accessible Method for Superresolution Fluorescence Imaging of Living Systems. Proc. Natl. Acad. Sci. U. S. A. 2012, 109, 10909–10914.

(9)

Chozinski, T. J.; Gagnon, L. A.; Vaughan, J. C. Twinkle, Twinkle Little Star: Photoswitchable Fluorophores for Super-Resolution Imaging. FEBS Lett. 2014, 588, 3603–3612.

(10)

Wiedenmann, J.; Ivanchenko, S.; Oswald, F.; Schmitt, F.; Röcker, C.; Salih, A.; Spindler, K.-D.; Nienhaus, G. U. EosFP, a Fluorescent Marker Protein with UV-Inducible Green-toRed Fluorescence Conversion. Proc. Natl. Acad. Sci. U. S. A. 2004, 101, 15905–15910.

(11)

McKinney, S. A.; Murphy, C. S.; Hazelwood, K. L.; Davidson, M. W.; Looger, L. L. A Bright and Photostable Photoconvertible Fluorescent Protein. Nat. Methods 2009, 6, 131– 133.

(12)

Zhang, M.; Chang, H.; Zhang, Y.; Yu, J.; Wu, L.; Ji, W.; Chen, J.; Liu, B.; Lu, J.; Liu, Y.; et al. Rational Design of True Monomeric and Bright Photoactivatable Fluorescent Proteins. Nat. Methods 2012, 9, 727–729.

(13)

Paez-Segala, M. G.; Sun, M. G.; Shtengel, G.; Viswanathan, S.; Baird, M. a; Macklin, J. J.; Patel, R.; Allen, J. R.; Howe, E. S.; Piszczek, G.; et al. Fixation-Resistant Photoactivatable Fluorescent Proteins for CLEM. Nat. Methods 2015, advance online.

(14)

Chang, H.; Zhang, M.; Ji, W.; Chen, J.; Zhang, Y.; Liu, B.; Lu, J.; Zhang, J.; Xu, P.; Xu, T. A Unique Series of Reversibly Switchable Fluorescent Proteins with Beneficial Properties for Various Applications. Proc. Natl. Acad. Sci. U. S. A. 2012, 109, 4455–4460.


Page 37 of 41

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Nano

(15)

Zhang, X.; Chen, X.; Zeng, Z.; Zhang, M.; Sun, Y.; Xi, P.; Peng, J.; Xu, P. Development of a Reversibly Switchable Fluorescent Protein for Super-Resolution Optical Fluctuation Imaging (SOFI). ACS Nano 2015, 150223122350008.

(16)

Ando, R.; Mizuno, H.; Miyawaki, A. Regulated Fast Nucleocytoplasmic Shuttling Observed by Reversible Protein Highlighting. Science 2004, 306, 1370–1373.

(17)

Ando, R.; Flors, C.; Mizuno, H.; Hofkens, J.; Miyawaki, A. Highlighted Generation of Fluorescence Signals Using Simultaneous Two-Color Irradiation on Dronpa Mutants. Biophys. J. 2007, 92, L97–L99.

(18)

Mizuno, H.; Dedecker, P.; Ando, R.; Fukano, T.; Hofkens, J.; Miyawaki, A. Higher Resolution in Localization Microscopy by Slower Switching of a Photochromic Protein. Photochem. Photobiol. Sci. 2010, 9, 239–248.

(19)

Andresen, M.; Stiel, A. C.; Fölling, J.; Wenzel, D.; Schönle, A.; Egner, A.; Eggeling, C.; Hell, S. W.; Jakobs, S. Photoswitchable Fluorescent Proteins Enable Monochromatic Multilabel Imaging and Dual Color Fluorescence Nanoscopy. Nat. Biotechnol. 2008, 26, 1035–1040.

(20)

Rosenbloom, A. B.; Lee, S.-H.; To, M.; Lee, A.; Shin, J. Y.; Bustamante, C. Optimized Two-Color Super Resolution Imaging of Drp1 during Mitochondrial Fission with a SlowSwitching Dronpa Variant. Proc. Natl. Acad. Sci. U. S. A. 2014, 111, 13093–13098.

(21)

Stiel, A. C.; Trowitzsch, S.; Weber, G.; Andresen, M.; Eggeling, C.; Hell, S. W.; Jakobs, S.; Wahl, M. C. 1.8 A Bright-State Structure of the Reversibly Switchable Fluorescent Protein Dronpa Guides the Generation of Fast Switching Variants. Biochem. J. 2007, 402, 35–42.

(22)

Flors, C.; Hotta, J.-I.; Uji-i, H.; Dedecker, P.; Ando, R.; Mizuno, H.; Miyawaki, A.; Hofkens, J. A Stroboscopic Approach for Fast Photoactivation-Localization Microscopy with Dronpa Mutants. J. Am. Chem. Soc. 2007, 129, 13970–13977.

(23)

Adam, V.; Lelimousin, M.; Boehme, S.; Desfonds, G.; Nienhaus, K.; Field, M. J.; Wiedenmann, J.; McSweeney, S.; Nienhaus, G. U.; Bourgeois, D. Structural Characterization of IrisFP, an Optical Highlighter Undergoing Multiple Photo-Induced Transformations. Proc. Natl. Acad. Sci. U. S. A. 2008, 105, 18343–18348.

(24)

Adam, V.; Moeyaert, B.; David, C. C.; Mizuno, H.; Lelimousin, M.; Dedecker, P.; Ando, R.; Miyawaki, A.; Michiels, J.; Engelborghs, Y.; et al. Rational Design of Photoconvertible and Biphotochromic Fluorescent Proteins for Advanced Microscopy Applications. Chem. Biol. 2011, 18, 1241–1251.

(25)

Moeyaert, B.; Nguyen Bich, N.; De Zitter, E.; Rocha, S.; Clays, K.; Mizuno, H.; Van Meervelt, L.; Hofkens, J.; Dedecker, P. Green-to-Red Photoconvertible Dronpa Mutant for Multimodal Super-Resolution Fluorescence Microscopy. ACS Nano 2014, 8, 1664–1673. 36 ACS Paragon Plus Environment

ACS Nano

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 38 of 41

(26)

Grotjohann, T.; Testa, I.; Leutenegger, M.; Bock, H.; Urban, N. T.; Lavoie-Cardinal, F.; Willig, K. I.; Eggeling, C.; Jakobs, S.; Hell, S. W. Diffraction-Unlimited All-Optical Imaging and Writing with a Photochromic GFP. Nature 2011, 478, 204–208.

(27)

Craggs, T. D. Green Fluorescent Protein: Structure, Folding and Chromophore Maturation. Chem. Soc. Rev. 2009, 38, 2865–2875.

(28)

Day, R. N.; Davidson, M. W. The Fluorescent Protein Palette: Tools for Cellular Imaging. Chem. Soc. Rev. 2009, 38, 2887–2921.

(29)

Shaner, N. C.; Steinbach, P. A.; Tsien, R. Y. A Guide to Choosing Fluorescent Proteins. Nat. Methods 2005, 2, 905–909.

(30)

Shaner, N. C.; Patterson, G. H.; Davidson, M. W. Advances in Fluorescent Protein Technology. J. Cell Sci. 2007, 120, 4247–4260.

(31)

Patterson, G. H.; Knobel, S. M.; Sharif, W. D.; Kain, S. R.; Piston, D. W. Use of the Green Fluorescent Protein and Its Mutants in Quantitative Fluorescence Microscopy. Biophys. J. 1997, 73, 2782–2790.

(32)

Haupts, U.; Maiti, S.; Schwille, P.; Webb, W. W. Dynamics of Fluorescence Fluctuations in Green Fluorescent Protein Observed by Fluorescence Correlation Spectroscopy. Proc. Natl. Acad. Sci. U. S. A. 1998, 95, 13573–13578.

(33)

Nienhaus, K.; Nienhaus, G. U. Fluorescent Proteins for Live-Cell Imaging with SuperResolution. Chem. Soc. Rev. 2014, 43, 1088–1106.

(34)

Hedde, P. N.; Nienhaus, G. U. Super-Resolution Localization Microscopy with Photoactivatable Fluorescent Marker Proteins. Protoplasma 2014, 251, 349–362.

(35)

Dean, K. M.; Palmer, A. E. Advances in Fluorescence Labeling Strategies for Dynamic Cellular Imaging. Nat. Chem. Biol. 2014, 10, 512–523.

(36)

Grotjohann, T.; Testa, I.; Reuss, M.; Brakemann, T.; Eggeling, C.; Hell, S. W.; Jakobs, S. rsEGFP2 Enables Fast RESOLFT Nanoscopy of Living Cells. eLife 2012, 2012, e00248.

(37)

Waldo, G. S.; Standish, B. M.; Berendzen, J.; Terwilliger, T. C. Rapid Protein-Folding Assay Using Green Fluorescent Protein. Nat. Biotechnol. 1999, 17, 691–695.

(38)

Pédelacq, J.-D.; Cabantous, S.; Tran, T.; Terwilliger, T. C.; Waldo, G. S. Engineering and Characterization of a Superfolder Green Fluorescent Protein. Nat. Biotechnol. 2006, 24, 79–88.

(39)

Crameri, A.; Whitehorn, E. A.; Tate, E.; Stemmer, Wi. P. C. Improved Green Fluorescent Protein by Molecular Evolution Using DNA Shuffling. Nat. Biotechnol. 1996, 14, 315– 319. 37 ACS Paragon Plus Environment

Page 39 of 41

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Nano

(40)

Zacharias, D. A.; Violin, J. D.; Newton, A. C.; Tsien, R. Y. Partitioning of Lipid-Modified Monomeric GFPs into Membrane Microdomains of Live Cells. Science 2002, 296, 913– 916.

(41)

Telford, W. G.; Hawley, T.; Subach, F.; Verkhusha, V.; Hawley, R. G. Flow Cytometry of Fluorescent Proteins. Methods 2012, 57, 318–330.

(42)

McEvoy, A. L.; Hoi, H.; Bates, M.; Platonova, E.; Cranfill, P. J.; Baird, M. a.; Davidson, M. W.; Ewers, H.; Liphardt, J.; Campbell, R. E. mMaple: A Photoconvertible Fluorescent Protein for Use in Multiple Imaging Modalities. PLoS One 2012, 7, e51314.

(43)

Jackson, S. E.; Craggs, T. D.; Huang, J. Understanding the Folding of GFP Using Biophysical Techniques. Expert Rev. Proteomics 2006, 3, 545–559.

(44)

Hsu, S.-T. D.; Blaser, G.; Jackson, S. E. The Folding, Stability and Conformational Dynamics of Beta-Barrel Fluorescent Proteins. Chem. Soc. Rev. 2009, 38, 2951–2965.

(45)

Remington, S. J. Fluorescent Proteins: Maturation, Photochemistry and Photophysics. Curr. Opin. Struct. Biol. 2006, 16, 714–721.

(46)

Iizuka, R.; Yamagishi-Shirasaki, M.; Funatsu, T. Kinetic Study of de Novo Chromophore Maturation of Fluorescent Proteins. Anal. Biochem. 2011, 414, 173–178.

(47)

Kremers, G.-J.; Goedhart, J.; Van Munster, E. B.; Gadella, T. W. J. Cyan and Yellow Super Fluorescent Proteins with Improved Brightness, Protein Folding, and FRET Förster Radius. Biochemistry 2006, 45, 6570–6580.

(48)

Sacchetti, A.; Cappetti, V.; Marra, P.; Dell’Arciprete, R.; El Sewedy, T.; Crescenzi, C.; Alberti, S. Green Fluorescent Protein Variants Fold Differentially in Prokaryotic and Eukaryotic Cells. J. Cell. Biochem. Suppl. 2001, 81, 117–128.

(49)

Chang, H.-C.; Kaiser, C. M.; Hartl, F. U.; Barral, J. M. De Novo Folding of GFP Fusion Proteins: High Efficiency in Eukaryotes but Not in Bacteria. J. Mol. Biol. 2005, 353, 397– 409.

(50)

Kremers, G.-J.; Goedhart, J.; van den Heuvel, D. J.; Gerritsen, H. C.; Gadella, T. W. J. Improved Green and Blue Fluorescent Proteins for Expression in Bacteria and Mammalian Cells. Biochemistry 2007, 46, 3775–3783.

(51)

Ai, H.-W.; Baird, M. A.; Shen, Y.; Davidson, M. W.; Campbell, R. E. Engineering and Characterizing Monomeric Fluorescent Proteins for Live-Cell Imaging Applications. Nat. Protoc. 2014, 9, 910–928.

(52)

Karasawa, S.; Araki, T.; Yamamoto-Hino, M.; Miyawaki, A. A Green-Emitting Fluorescent Protein from Galaxeidae Coral and Its Monomeric Version for Use in Fluorescent Labeling. J. Biol. Chem. 2003, 278, 34167–34171. 38 ACS Paragon Plus Environment

ACS Nano

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 40 of 41

(53)

Chmyrov, A.; Keller, J.; Grotjohann, T.; Ratz, M.; D’Este, E.; Jakobs, S.; Eggeling, C.; Hell, S. W. Nanoscopy with More than 100,000 “Doughnuts”. Nat. Methods 2013, 10, 737–740.

(54)

Arpino, J. A. J.; Rizkallah, P. J.; Jones, D. D. Crystal Structure of Enhanced Green Fluorescent Protein to 1.35 Å Resolution Reveals Alternative Conformations for Glu222. PLoS One 2012, 7, e47132.

(55)

Mizuno, H.; Mal, T. K.; Wälchli, M.; Kikuchi, A.; Fukano, T.; Ando, R.; Jeyakanthan, J.; Taka, J.; Shiro, Y.; Ikura, M.; et al. Light-Dependent Regulation of Structural Flexibility in a Photochromic Fluorescent Protein. Proc. Natl. Acad. Sci. U. S. A. 2008, 105, 9227– 9232.

(56)

Nguyen Bich, N.; Moeyaert, B.; Van Hecke, K.; Dedecker, P.; Mizuno, H.; Hofkens, J.; Van Meervelt, L. Structural Basis for the Influence of a Single Mutation K145N on the Oligomerization and Photoswitching Rate of Dronpa. Acta Crystallogr. Sect. D Biol. Crystallogr. 2012, 68, 1653–1659.

(57)

Andresen, M.; Stiel, A. C.; Trowitzsch, S.; Weber, G.; Eggeling, C.; Wahl, M. C.; Hell, S. W.; Jakobs, S. Structural Basis for Reversible Photoswitching in Dronpa. Proc. Natl. Acad. Sci. U. S. A. 2007, 104, 13005–13009.

(58)

Bourgeois, D.; Adam, V. Reversible Photoswitching in Fluorescent Proteins: A Mechanistic View. IUBMB Life 2012, 64, 482–491.

(59)

Subach, F. V; Zhang, L.; Gadella, T. W. J.; Gurskaya, N. G.; Lukyanov, K. A.; Verkhusha, V. V. Red Fluorescent Protein with Reversibly Photoswitchable Absorbance for Photochromic FRET. Chem. Biol. 2010, 17, 745–755.

(60)

Pletnev, S.; Subach, F. V; Dauter, Z.; Wlodawer, A.; Verkhusha, V. V. A Structural Basis for Reversible Photoswitching of Absorbance Spectra in Red Fluorescent Protein rsTagRFP. J. Mol. Biol. 2012, 417, 144–151.

(61)

Brakemann, T.; Stiel, A. C.; Weber, G.; Andresen, M.; Testa, I.; Grotjohann, T.; Leutenegger, M.; Plessmann, U.; Urlaub, H.; Eggeling, C.; et al. A Reversibly Photoswitchable GFP-like Protein with Fluorescence Excitation Decoupled from Switching. Nat. Biotechnol. 2011, 29, 942–950.

(62)

Lukyanov, K. a.; Fradkov, A. F.; Gurskaya, N. G.; Matz, M. V.; Labas, Y. a.; Savitsky, A. P.; Markelov, M. L.; Zaraisky, A. G.; Zhao, X.; Fang, Y.; et al. Natural Animal Coloration Can Be Determined by a Nonfluorescent Green Fluorescent Protein Homolog. J. Biol. Chem. 2000, 275, 25879–25882.

(63)

Andresen, M.; Wahl, M. C.; Stiel, A. C.; Gräter, F.; Schäfer, L. V; Trowitzsch, S.; Weber, G.; Eggeling, C.; Grubmüller, H.; Hell, S. W.; et al. Structure and Mechanism of the


Page 41 of 41

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Nano

Reversible Photoswitch of a Fluorescent Protein. Proc. Natl. Acad. Sci. 2005, 102, 13070– 13074. (64)

Henderson, J. N.; Ai, H.-W.; Campbell, R. E.; Remington, S. J. Structural Basis for Reversible Photobleaching of a Green Fluorescent Protein Homologue. Proc. Natl. Acad. Sci. U. S. A. 2007, 104, 6672–6677.

(65)

Duwé, S.; Moeyaert, B.; Dedecker, P. Diffraction-Unlimited Fluorescence Microscopy of Living Biological Samples Using pcSOFI. Curr. Protoc. Chem. Biol. 2015, 7, 27–41.

(66)

Yuliang, M.; Taylor, S. A 15-Residue Bifunctional Element in D-AKAP1 Is Required for Both Endoplasmic Reticulum and Mitochondrial Targeting. J. Biol. Chem. 2002, 277, 27328–27336.

(67)

Sawano, A. Directed Evolution of Green Fluorescent Protein by a New Versatile PCR Strategy for Site-Directed and Semi-Random Mutagenesis. Nucleic Acids Res. 2000, 28, 78e.

(68)

Rasila, T. S.; Pajunen, M. I.; Savilahti, H. Critical Evaluation of Random Mutagenesis by Error-Prone Polymerase Chain Reaction Protocols, Escherichia Coli Mutator Strain, and Hydroxylamine Treatment. Anal. Biochem. 2009, 388, 71–80.

(69)

Zhao, H.; Giver, L.; Shao, Z.; Affholter, J. A.; Arnold, F. H. Molecular Evolution by Staggered Extension Process (StEP) in Vitro Recombination. Nat. Biotechnol. 1998, 16, 258–261.

(70)

Ward, W. W. Bioluminescence and Chemiluminescence. In Basic Chemistry and Applications; DeLuca, M. A.; McElroy, W. D., Eds.; Academic Press, New York, 1981; pp. 235–242.

(71)

Reid, B. G.; Flynn, G. C. Chromophore Formation in Green Fluorescent Protein. Biochemistry 1997, 36, 6786–6791.

(72)

Dedecker, P.; Duwé, S.; Neely, R. K.; Zhang, J. Localizer: Fast, Accurate, Open-Source, and Modular Software Package for Superresolution Microscopy. J. Biomed. Opt. 2012, 17, 126008.


Expression-Enhanced Fluorescent Proteins Based ... - ACS Publications

Recommend Documents