Image collection and region of interest assignment. Four channels were collected per field: (a) Acceptor channel collected at 200 ms. (b) Acceptor channel collected at 50 ms. (c) Donor channel collected at 200 ms. (d) FRET channel collected at 100 ms. ROIs containing any saturated pixels were deleted. The ROIs remaining are shown as white outlines in this example. ROIs were used from the 200-ms acceptor channel except when there were saturated pixels in that channel. In that case, acceptor ROIs were taken from the 50-ms acceptor channel and the associated intensity multiplied by four to obtain accurate acceptor measurements.

Establishment of calibration constants to convert acceptor-bleedthrough-corrected FRET/donor (F/D) measurements into the proportion of donor transferred to acceptor (E). Background-subtracted and acceptor bleed-through corrected intensity measurements in the acceptor (A), donor (D), and FRET (F) channels were collected from cells expressing a dual-labeled CFP-AR-YFP calibration standard. (a) F/D ratios for 30 different CFP-AR-YFP calibration standards (only five shown here) were collected by plotting F against D for a large number of cells. The F/D ratios for those five standards (mean ± SD from six independent studies) are shown in the inset. (b) A/D ratios from the same cells were collected by plotting A against D. (c) As outlined in Sec. 2.4, Eq. (7), plotting F/D against A/D, yields a curve with a y intercept that is the difference of the constants KfD and KfaD. The slope of that same curve represents KfaD divided by KaD (see Sec. 2.4 for definition of the constants). Establishment of the calibration constants enables E to be extrapolated from the easily and accurately measured F/D ratio [Eq. (3)]. (d) Comparison of E determined from the F/D ratios with E determined from the A/D ratios [Eq. (6)] for the 30 calibration standards. (Color online only.)

FRET measurement of the dimerization kinetics of the estrogen receptor (ERα) labeled (a) and (c) with the dimer competent FPs or (b) and (d) with the monomeric A206K FP mutants. (a) and (b) Plotting the amount of complex formed (represented by E) against the concentration of acceptor present (x axis) shows that the data points fit well to a curve indicative of a one-site, bimolecular interaction, when the cells are grown in the presence of estradiol. The curve fitting values (R², P-values from runs test, P-values for normal distribution) are shown for each dataset, along with the number of cells analyzed (n). P-values of <0.05 (shown in gray) are considered to represent significant deviation from the expected number of runs or from a normal distribution of the data points around the best-fitting curve. The data from five separate experiments are shown in five different columns of data to assess data reproducibility. (c) and (d) Similar analysis for cells grown without the addition of estradiol; the curves obtained for the binding in the presence of estradiol are shown as solid lines to compare against the no estradiol control cells (dashed lines). (e) Direct comparison of the dimeric CFP/YFP and monomeric mCFP/mYFP binding curves obtained in parallel experiments.

Bleed-through control values determined by linear regression. (a) The amounts of background-subtracted fluorescence from the donor-labeled ERα (CFP or mCFP) in the FRET or acceptor channel in relationship to the donor channel. The r² values indicate whether the data points fall along a straight line. For a linear relationship, the slope of the line represents the bleed-through contributions of the donor to each channel. (b) Similar analysis for the acceptor-labeled ERα (YFP or mYFP). The bleed-through and r² values for each of the five experiments are shown in Table 2.

Removal of autofluorescent debris. (a) For high throughput analysis with no user intervention, a quality control (q_c) measurement was derived (Sec. 3.3) to identify ROIs (open boxes) that varied substantially from the ROIs representing actual cellular objects (closed boxes). In the current study, all objects with q_c values between −0.8 and 0.8×10⁻³ were retained for further analysis. Note that the corresponding acceptor fluorescence values here (x axis) and in the remaining graphs are all corrected by dividing A by KaD (the constant that describes how well acceptor fluorescence is collected relative to an equimolar amount of donor). Where helpful, this enables acceptor fluorescence values to be compared in relationship to an equal amount of fluorescence from an equimolar amount of donor. (b) Included data points fit well to the one-site binding bimolecular interaction curve (see Sec. 3.4), whereas most of the excluded data points do not.

FP dimerization does not affect FRET measurement of the interaction kinetics of the estradiol-bound estrogen receptor (ERα) with itself. (a) For each of the five independent studies shown in Fig. 5, the amounts of acceptor required to reach half of B max (Kd) were no different for the interactions of ERα tagged with the dimer-competent (closed bars) or monomeric (open bars) FPs. (b) The associated B max measurements also showed no significant differences within most experiments, although there was a marginally significant (p=0.05) trend for the B max to be less for interactions measured with the monomeric FP. The mean ± standard deviation for all five experiments is shown for both graphs as the right-most bars.