Expression studies with cassettes 0, 2, 3, or 234 integrated at RL1 in the absence of MCNeo. (A) FACS-GAL analyses. All analyses were performed in the presence of 2 mmol/L propidium iodide. Debris and dead cells were excluded from the analyses by gating out cells with low SSC (Side-SCatter) and high (propidium iodide–induced) fluorescence in the red channel (FL2). Two thousand cells are shown in each graph, 10,000 cells were counted. The boxes in the first column are MEL cells with PGK-HYG at the RL1 locus and were used to defined the limit between the β-gal–negative and β-gal–positive populations. β-gal–positive cells are in the upper window (x axis = FSC [Forward Scatter]; y axis = green fluorescence channel [530-nm wavelength with a 30-nm bandwidth]). The lower limit of the window was set at the position at which between 0.3% to 0.4% of control cells appeared positive. The next four columns represent cells with cassettes 0, 2, 3, or 234 at RL1 before and after induction of differentiation by DMSO or hemin. Striking differences in the proportions of expressing cells can be observed. (B) Comparative analysis of cassettes 0, 2, 3, and 234 integrated at RL1. Levels of LacZ expression in cell population with cassettes 0, 2, 3, 234, and 234o2 integrated at RL1 were measured on cell lysates by chemiluminescence. The percentage of cells expressing β-GlobZ within each cell population was determined by FACS-Gal analyses (illustrated above). The level of expression of LacZ per positive cells was estimated by multiplying the level of LacZ expression in cell population by the proportion of positive cells after normalization (see text). For each cassette, the histograms represent the results obtained for two independent pools of several clones (pool A and B, see text). The means and standard deviation of the results obtained for pools A and B combined is also indicated on the right side of the histograms. The smallest level of β-gal expression represented in the histogram (0.8 × 10⁻¹⁵ g of β-gal/cell for cassette 0 in uninduced cells) was about 15 times higher than the background level observed with untransfected cells (data not shown) and was therefore well within the range of detection of our luminometer. This analysis clearly shows that both the probability of expression and the rate of transcription are enhancer and trans-acting factor dependent. (C) Structure of the locus. The structure of the RL1 locus after integration of the cassettes and excision of the MCNeo gene is depicted. LCR fragments were, respectively, HS2, HS3, and HS234 in cassettes 2, 3, and 234 (with HS234 in reverse orientation in HS234o2). In cassette 0, no LCR fragment was present.