Figure 1.
Spectral characterization of mMaple.
(A) Excitation and emission spectra of the pre-photoconversion green state of mMaple. (B) Excitation and emission spectra of the post-photoconversion red state of mMaple.
Table 1.
Properties of mMaple and related variants.
Figure 2.
Characterization of mClavGR2 and its improved variants.
(A) Absorbance spectra of mMaple (green line), mClavGR2 (red line) and mEos2 (orange line). Absorbance spectra are normalized to absorbance at 280 nm. (B) Photoconversion from the green state (green lines) to the red state (red lines) of mMaple (diamonds), mClavGR2 (triangles), and mEos2 (circles) for pcFP-H2B fusions expressed in HeLa S3 cells. The red-to-green contrast has been calculated at 47 s and 91 s (indicated with vertical dotted lines). At 47 s the contrast is 5.3 for mMaple, 2.1 for mClavGR2, and 1.8 for mEos2. At 91 s the contrast is 10.8 for mMaple, 4.7 for mClavGR2, and 6.9 for mEos2.
Figure 3.
(f-)PALM/STORM comparison of mMaple, mClavGR2 and mEos2.
(A–C) Images of ΔcheW E.coli expressing CheW fusion proteins at L-arabinose concentrations optimal for swarming. Images contain (A) 1086 mMaple-CheW localizations (B) 694 mClavGR2-CheW localizations and (C) 229 mEos2-CheW localizations. (D) The mean number of photons emitted by each construct per photoconversion event (error is standard error, N = 3 independent measurements from distributions consisting of 4,000–32,000 localizations). Scale bars are 500 nm. (E) Distribution of the number of localizations observed for ΔcheW E. coli cells containing CheW fusions to mMaple, mClavGR2, and mEos2. Greater than 96% of cells expressing either mEos2- or mClavGR2-CheW fusions have less than 500 localizations (boxed region), whereas greater than 50% of cells expressing mMaple-CheW fusions have more than 500 localizations.
Figure 4.
(f-)PALM/STORM characterization of the number of observed localizations and proteins per cell.
(A–B) (f-)PALM/STORM images of fixed E. coli expressing cytoplasmic (A) mMaple (N = 1696 localizations), or (B) mEos2 (N = 472 localizations). Localizations are represented as normalized 2D Gaussian peaks with widths given by their theoretical localization precisions (left panels) and plotted as small markers grouped into clusters with adjacent spacing of 30 nm or less (right panels). Individual protein localizations are shown in grey whereas closely spaced localizations (<30 nm) are grouped into clusters of the same color (right panels). The bright field and conventional fluorescence images are shown for comparison (left panels, left and right inset respectively). Scale bars are 500 nm. (C) Average number of localizations per cell for each cytoplasmically expressed pcFP. (D) The distribution of cluster sizes (<30 nm interlocalization spacing) for cytoplasmically expressed pcFPs. (E) Average number of cytoplasmically expressed proteins per cell. Rather than counting each localization as a single molecule, we count each cluster of localizations (localizations spaced <30 nm) as a single protein. The dotted lines in (A–B) denote the E. coli cell boundary. Scale bars are 500 nm and 50 nm (zooms). Zooms in (A–B) show possible reversible photoswitching events of single proteins. Error is the standard deviation (N = 20 cells (mMaple), N = 17 cells (mClavGR2), N = 16 cells (mEos2)). The large error bars are primarily due to variation in protein expression between cells.
Figure 5.
Green state fluorescence comparison between pcFP-CheW fusions.
(A) E. coli was transformed with plasmids encoding CheW fusion proteins and cultures were induced with 0.01% L-arabinose (top row) and 0.2% L-arabinose (bottom row) for 3 h at 30°C and then imaged using 488 nm excitation. Note the brightness of the cells expressing mMaple fusions and the large percentage of cells with the correct polar localization pattern. (B) Fluorescence image of the bands corresponding to pcFP-CheW fusions extracted from E. coli and analyzed by SDS-PAGE gel. Proteins were either purified from the soluble fraction by Ni2+-NTA, solubilized from inclusion bodies with urea (IB), or loaded onto the gel as a whole cell lysate (WCL). Exposure times were increased from 60 ms to 2 sec in order to visualize the fluorescence of GFP-CheW and mEos2-CheW (right). Relative fluorescence intensities after correction for exposure times are: 1.00, 2.56, 0 (mMaple-CheW), 4.79, 0.43 (mClavGR2-CheW), 0, 0.08, 0.14 (EGFP-CheW), 0, 0.03 and 0 (mEos2-CheW). The Coomassie stained version of this gel is shown in Fig. S2C. (C) Cells expressing only cytoplasmic versions of each pcFP were grown in LB and induced with 0.002% L-arabinose for 3 hours. The mean fluorescence intensity for cells expressing mMaple was more than three times that of mEos2 (1489±636 for mMaple in comparison to 463±281 for mEos2) (N = 58 cells for mMaple, N = 57 cells for mClavGR2 and mEos2. (D) Mean fluorescence intensity for cells expressing each pcFP-CheW fusion. Consistent with the images in (A) and (B), at both induction levels, the mean fluorescence intensity level of mMaple-CheW is higher than all other CheW fusions (for 0.01% arabinose conditions, the number of cells analyzed was 69 for mMaple-CheW and 46 for mClavGR2-CheW respectively; for 0.2% conditions, N = 91, 62, 40, 71 cells for mMaple, mClavGR2, mEos2 and GFP fusions respectively).
Figure 6.
mMaple has improved in vivo brightness and enables 3D-SIM reconstructions.
(A) ΔcheW E.coli expressing mMaple-CheW and a zoom (B) of the polar region of the cell denoted by the boxed region in (A). (C) One 125 nm slice of the 3D-SIM reconstruction of a ΔcheW E. coli expressing GFP-CheW. Red represents fluorescence of the membrane-specific dye FM4-64 and green represents FP fluorescence. Scale bars are 500 nm (A, C) and 100 nm (B).
Table 2.
Advantages and limitations of mMaple.
Figure 7.
pcFP fusions expressed in mammalian cells.
(A–C) Representative widefield fluorescence images of HeLa cells transfected with plasmids encoding either mMaple-actin (A), mEos2-actin (B), or mClavGR2-actin (C). (D) Flow cytometric analysis of the green fluorescence for HeLa cells transfected with the same plasmids used in A–C. A total of 7000 fluorescent cells were analyzed in each experiment. Relative median fluorescence intensity after excitation and emission correction is: 0.68 (mMaple), 0.77(mClavGR2) and 1.00 (mEos2). (E–G) Representative images of HeLa cells transfected with actinin fusions of mMaple (E), mEos2 (F), and mClavGR2 (G). (H) Flow cytometric analysis of the green fluorescence from HeLa cells transfected with the plasmids used in E–G. Relative corrected median fluorescence intensity is: 1.03 (mMaple), 1.25 (mClavGR2) and 1.00 (mEos2).