Figure 1.
Fluorescent decay curves of H2B-GFP (A) and H2B-Halo-TMR (B) during time lapse imaging of entire nuclei in live cells.
The GFP tag exhibited a bi-phasic decay with three different combinations of the laser power and delay time between images (A). Colored curves indicate fits to the data with a bi-exponential decay model. According to the model (C), fluorescent molecules can convert at a rate
into a photoswitched dark state
and then revert to the fluorescent state at a rate
. Fluorescent molecules can also bleach irreversibly to a dark state
at a rate
. In contrast to the GFP tag, the TMR-Halo tag exhibited a monophasic decay that was well fit by a single exponential (colored curves) (B). Fitting parameters are shown in Table 1.
Table 1.
Photoconversion rates of GFP obtained from fits to fluorescent decay curves (estimates±95% confidence intervals).
Table 2.
Photobleaching rates of TMR obtained from fits to fluorescent decay curves (estimates±95% confidence intervals).
Figure 2.
Artifacts in FRAP arising from the strength of the intentional photobleach.
The entire nucleus containing either H2B-GFP or H2B-Halo-TMR was bleached to eliminate any conventional fluorescence recovery by influx of unbleached fluorescence, and then whole nuclear intensity was measured. The data were normalized with the pre-bleach intensity set to 1 and the post bleach intensity set to 0. Artifactual recoveries were observed with the GFP tag (A) but not the TMR-Halo tag (B). Consistent with previous studies [9], the relative size of the recovery with the GFP-tag increased with decreasing strength of the photobleach (colored curves in A). No such dependence was seen with the TMR-Halo tag (colored curves in B). Comparable effects were seen when performing a conventional spot FRAP (0.5 µm radius) with H2B exhibiting a fast component of ∼15% with the GFP-tag and ∼1% with the TMR-Halo tag (C). The TMR-HaloTag fast component of 1% for H2B is consistent with a previous study which predicted a 1% fast component after mathematical correction for photoswitching in FRAP of H2B-GFP [9]. Note that the fast component was estimated by finding the point where the change in slope is minimized in the measured curve.
Figure 3.
Artifacts in FRAP arising from the timing of the intentional photobleach.
Spot FRAPs of H2B tagged with either GFP or TMR-Halo were performed varying only the number of pre-bleach images collected (1, 2, 30, 60). The intensity of the photobleach was kept constant at the level used in Fig. 2C. This yielded FRAP curves that were equivalent to those in Fig. 2C, as long as the number of pre-bleach images was 30 or 60 (60 were used in Fig. 2C). However, when only one or two pre-bleach images were acquired the size of the artifactual fast component increased (A). In contrast, with the TMR-HaloTag no dependence on the timing of the intentional photobleach was seen, with a fast component size of ∼1% in all cases (B).
Figure 4.
Artifacts in FRAP arising from alteration in the temporal sampling rate during the recovery phase.
H2B spot FRAP was performed as in Fig. 2C but now the temporal sampling rate was changed such that data after the photobleach were collected with no delay between images for the first 3 s after the photobleach and then with an interval of 2 s between images at the 3 s time point after the photobleach. With a constant sampling rate, the artifactual fast fraction of ∼15% for the GFP tag is equivalent to that in Fig. 2C (green curve, A), whereas with a change in temporal sampling rate the FRAP curve jumps upward leading to an even larger artifactual fast fraction (black curve, A). In contrast, the TMR-HaloTag shows no such effect with the red and black curves overlapping (B).
Figure 5.
Artifacts in FRAP arising from alteration in the temporal sampling rate at the intentional photobleach.
H2B spot FRAP was performed as in Fig. 2C, but now the photobleach profile immediately after the bleach was measured. With a GFP tag, the photobleach profile appears to extend a distance of up to 6 µm for a 0.5 µm radius bleach spot and could not be fitted with the conventional model for a photobleach profile [3] (A). This is an artifact due to an alteration in temporal sampling rates at the time of the photobleach imposed by our microscope software [9]. This effect is not observed with a TMR-HaloTag, where instead the photobleach profile returns to the normalized intensity of one within 3 um from the center of the photobleach and could be well fitted with the conventional model for a photobleach profile (B).
Figure 6.
The dynamics of H2B-Halo-TMR with and without UV irradiation.
Consistent with a previous study [19], the slow recovery phase of H2B after UV irradiation is faster than the control without UV irradiation. However, we could detect no significant difference in the size of the fast component between with or without UV irradiation suggesting that the free pool of H2B is not altered by UV irradiation.