Figure 1.
Construction and utility of the FPNLSFP nuclear marker.
A, Amino acids inserted (black font, SV40 NLS underlined) between two mCherry FPs (orange font) within the mCherryNLSmCherry nuclear marker. The insertion sequence and location is similar for the other FPNLSFP nuclear markers created in this study. B, Nuclear fluorescence of the mCherryNLSmCherry marker stably expressed in a HeLa cell line in relationship to nuclei stained with Hoechst 33342. The locations of a YFP-tagged Androgen Receptor (AR) co-expressed in this cell line also are shown. The cells were grown in media containing testosterone, which translocates the AR into the cell nuclei. Images were captured with a 10x objective. *, mitotic cell. #, dying cell.
Figure 2.
Segmentation of cell nuclei marked with FPNLSFP.
The boundaries of objects with contiguous FPNLSFP expression established by a commercial analysis software (yellow circles) marked the boundaries of nuclei stained with the DNA binding dye Hoechst 33342. *, colony of cells in which FPNLSFP expression is lost sporadically.
Table 1.
Growth properties of a FPNLSFP-tagged LNCaP-C4-2 prostate cancer cell line.
Figure 3.
Improved well-to-well reproducibility in cell growth measurement enabled by the FPNLSFP live cell nuclear marker.
Variations in cell numbers plated in each well (Day 0) obscured the ability to reliably detect an increase in cell number after four days of slow growth by LNCaP-C4-2 prostate cancer cells treated with vehicle or 0.2 nM DHT. Each symbol represents the numbers of A, Hoechst 33342-stained nuclei or B, FPNLSFP-marked nuclei segmented in each well. C, Dividing the number of FPNLSFP-marked cells on Day 4 by the baseline (Day 0) number of FPNLSFP-marked cells in the same well improved the reproducibility of growth measurement. Dotted lines, three standard deviations (3sd) above the mean Day 0 (black dotted line) or Day 4 vehicle-treated (gray dotted line) measurements are shown. The 3sd cut-offs were used to determine the number of, respectively, vehicle-treated and DHT-treated wells that were scored falsely in the Day 0 and vehicle-treated wells.
Table 2.
Longitudinal comparison of FPNLSFP-marked cell numbers in the same well on Day 4 and Day 0 minimizes incorrect growth measurements of FPNLSFP-tagged LNCaP-C4-2 cell lines.
Figure 4.
A, Excitation and emission properties of four FPs used to create the bar-code. All FPs were excited by light of 560–590 nm (orange box) but emitted different relative amounts in two emission channels (em1: 635–675 nm; em2: 608–648 nm). For example, the area under the curve collected for mPlum in em1 would be slightly more than that in em2 whereas, for mCherry, em2 emissions would be much higher than em1 emissions. These differences were seen in practice (Table 3). B, The measured em1/em2 ratios for sixteen FPNLSFP markers using all possible combinations of the four FPs (Table 4) were similar to those predicted if one assumes no FRET amongst the FPs. The theoretical em1/em2 ratios were calculated from their relative abilities to be excited by 560–590 nm light (Fig. 4A), their relative brightness once excited (Table 3) and the em1 and em2 emissions detected by our instrument for the four homogeneous FPNLSFPs (mPlumNLSmPlum, mCherryNLSmCherry etc).
Table 3.
em1 (635–675 nm) fluorescence channel relative to em2 (608–648 nm) of the indicated red FPs.
Table 4.
Bar-Coded FPNLSFP Nuclear Markers.
Figure 5.
Application of bar-code to cell counting studies.
A, Concept of bar-code for mixing differentially marked FPNLSFP expressing cells. B, Differential response of two LNCaP-C4-2 cell subclones to an inhibitor of cell growth (actinomycin D). C, em1/em2 ratios of all cells within a representative well (x-axis) compared to the intensities of each cell in the em1 channel. D, LNCaP-C4-2 cells mixed, co-plated, treated exactly as in figure 5B then separated according to the bar-code showed similar treatment responses to the individually plated cells. Growth measurements are shown as the mean +/− sd from 8 (Fig. 5B) or 16 (Fig. 5D) wells for each treatment condition. *, statistically significant (p<0.01) increases or decrease in cell number relative to vehicle-treated cells; #, statistically significant (p<0.01) increase in cell number of DHT/actinomycin D treated wells relative to actinomycin D-treated wells.
Figure 6.
Bar-code separation of different, co-cultured assays.
Representative images of A, the em1/em2 ratio and B, YFP-tagged AR showed the FPNLSFP bar code to accurately discriminate between two differentially marked HeLa cell lines in which a wild-type (wt) or mutant (T877S) AR have different nuclear distributions when grown with 10−8 M estradiol. C, Quantification of nuclear AR levels in the two different cell lines after incubation with 10−7 M of the indicated steroids demonstrated that the differential responses between the wt and T877A ARs observed when plated separately were retained when the cell lines were mixed in a well and sorted according to the bar-code. Nuclear AR measurements are shown as the mean +/− sd from 48 wells for each treatment condition. The distinct responses of the two assays to different hormones are indicated by *, # (statistically significant increases, p<0.01, that were at least double the nuclear AR-YFP intensities in vehicle-treated mPlumNLSmPlum and mCherryNLSmCherry cells, respectively).
Figure 7.
Characterization of three differentially bar-coded LNCaP-C4-2 cell lines, each expressing a different YFP-based reporter assay.
A–C, YFP fluorescence intensities within the FPNLSFP-marked cell nuclei of each separately-cultured reporter line are shown in response to fifteen different steroids (10−8 M each). Nuclear YFP measurements are shown as the mean +/− sd from 10 fields for each treatment condition. Gray bars, measurements for indicated assay plated independently into separated wells. Black bars, measurements from wells in which the three assays are co-cultured and separated by the bar-code. The dotted gray line represents the measurements obtained upon treatment of the independent assays with vehicle only. Steroids: 1: pregnenolone, 2: progesterone, 3: 11-deoxycorticosterone, 4: aldosterone, 5: 17-hydroxypregnenolone, 6: 17-hydroxyprogesterone, 7: 11-deoxycortisol, 8: cortisol, 9: dehydroepiandrosterone, 10: androstendione, 11: estrone, 12:4-androstenediol, 13: testosterone, 14: estradiol, 15: dihydrotestosterone. veh, wells treated with vehicle only. *, steroids that increase an assay (p<0.05).
Table 5.
Effective discrimination of three bar-coded cell-based assays.
Figure 8.
Distinct em1/em2 ratios characteristic for each of the three bar-coded LNCaP-C4-2 cell lines.
A–C, Data from a representative field for each cell line. Each ‘x’ represents a YFP (left panel) or em1 (next panel) fluorescent intensity measurement from a single cell in relationship to the em1/em2 ratio measured for that cell. The images from which the measurements were obtained are shown. The representative fields for A and C were from cells treated with an androgen while the fields shown for B were vehicle treated because the MMTV-YFP assay intensity is much higher upon androgen treatment. The em1/em2 ratios used to assign a specific cell to a specific bar-code are shown as colored bars on the y-axis. The margins for those characteristic em1/em2 ratios were established as 3 sd away for all cellular measurements for the independently grown assays.