Schroeter EH, Kisslinger JA, and Kopan R: Notch 1 signalling requires ligand-induced proteolytic release of intracellular domain. Nature 1998; 393:382-386.

Appendix A

Transfections:

3T3 were cultured in DMEM supplemented with 10% bovine calf serum. Transfections were performed as described [Kopan, et al. (1996) "Signal transduction by activated mNotch: Importance of proteolytic processing and its regulation by the extracellular domain. "Proceedings of the National Academy of Sciences of the United States of America, 93: 1683-1687] except that the volumes were scaled down to be appropriate for 2 ml of medium in a 6-well dish. Each well was transfected with a total of 2µg of DNA consisting of: 400 ng HES-1-ßGal, 40 ng pCS2+Luciferase, 0.4 ­ 400 ng of pCS2+mNotch with the remaining DNA made up to 2 µg with empty pCS2+ vector. 16 hours post-transfection cells were rinsed and then fed with fresh medium. 48 hours post transfection cells were lysed in 300µl of 100 mM potassium phosphate pH7.8, 0.2% Triton X-100, 1 mM DTT, 0.2 mM PMSF, 0.5 µg/ml Leupeptin, 1.0 µg/ml Aprotinin. 100 µl of lysate was used to determine Luciferase activity as described in Kopan et al. 1996. 5.0µl of lysate was used to determine ßGal concentration and equalize results for transfection efficiency. ßGal assays were performed using the chemiluminescent substrate Galacton (Tropix) as described by the manufacturer (see also Fig. A). Each data point in Fig. 5 was generated by transfecting three plates with the same DNA mixture. One plate was used to measure HES-ßGal activity in the population relative to the CMV Luciferase transfection controls. The same plot shape was obtained with the HES-ßGal as it was with HES-Luciferase (figure 4 in the paper). The other two plates were stained either with the 9E10 anti myc or with a monoclonal antibody to Notch Ank repeats. The results were similar with the exception that the AN-1 monoclonal was less sensitive to Notch (about ten fold difference). We thus showed the data collected with the more sensitive 9E10. The transfection efficiency, based on the Luciferase control, was within our normal parameters and was consistent throughout the entire experiment and does not explain the changes in the number of HES-ßGal expressing cells.

Antibody staining and data collection:

Plates used for staining were fixed with 4% paraformaldehyde (5 min.), rinsed once with 0.1% Triton X100 in PBS (1 min.) and rinsed (3x5 min.) with PBS. 5prime3prime anti ßGal polyclonal antibody was diluted 1:500 in 9E10 Mouse hybridoma supernatant (containing 10% Fetal Calf Serum) against the Myc epitope. 500 µl of the antibody solution was incubated on the cells for one hour at room temperature. Plates were placed on a nutator for the incubation period. Following the antibody incubation, plates were washed (3x15 min.) in PBS. Goat Secondary antibodies (BODIPY conjugated anti-rabbit and Cy3 conjugated anti mouse, Molecular Probes, Oregon) were placed on the cells over night at 40c. The next day the plates were washed again (3x15 min.) in PBS. Hoechst stain was added to the penultimate wash. Cells were viewed under PBS with a Zeiss 16x water-immersion objective. Cy3 was chosen for the detection of Notch since it gave a brighter signal.

The data presented in Fig 4 are collected by scanning one diameter in a 60 mm plate and counting every labeled cell within that area. This strategy was chosen because in our experience, the distribution of transfected cells (judged by protein expression) is not uniform. Patterns such as concentric circles are often observed and are generated by standing waves during cell plating and/or transfection with CaPO4 (Patterns can be visualized by staining with X-gal after transfection with CMV-ßGal). The area of the plate is 1960 mm2 and the scanned area is 45 mm2 . The same area is covered in all plates. A representative diameter was chosen for cell counting, and all the fluorescent cells along this diameter were counted with a triple filter setup that permits visualization of the three markers (green, red and blue) simultaneously. Representative fields were photographed with a Quantex CCD camera (Photometrix) through the same 16x water immersion objective capturing a 0.22 mm2 field. We estimated the number of total cells that are seen within the region that was scored to be at most 20,500 [high average of 100 cells in photographed field (0.22 mm2 ) out of 45 mm2 scanned]. These numbers predict a final density of 8x105 cells on the average plate at the time of analysis and are consistent with the expected growth of the initial ~3x105 cells which we seeded per plate. At saturation of Notch activity (100 ng DNA/ml) we see >200 positive cells out of ~20,000 or about 1% of cells scored. This transfection efficiency is routinely obtained in the lab with 3T3 cells.

Estimate of how many molecules of NICD per cell are needed for Hes1 activation. To estimate cellular concentration of Notch we titrated input Notch DNA. We find that in populations of transfected cells the maximal rate of change in response to pCS2+NIC occurs at extremely low levels of input Notch DNA (Fig. 4 in our paper, at the range 0.2 to 2 ng/ml; Fig. A). To estimate how much Notch protein is produced within this range from the sCMV expression vector, we titrated pCS2+ßGal. ßGal is stable protein which makes it very suitable as a reporter and estimates of protein concentraion based on it will be conservative since it accumlates to high levels. At concentrations between 0.2 and 2 ng/ml sCMV-ßGal DNA we detect 0.2-2 pg ßGal protein/µl extract. These values are just above background in the assay and are at the limit of detection in our cell extracts. A 60 mm plate is extracted in 300 µl,of lysis buffer. Thus there is approximately 60pg of ßGal protein/plate (or ~3x108 molecules) at the lower limits of detection. The molecular weight of of the ßGal protein subunit is 135,000.

If Notch is produced at similar levels and is as stable as ßGal, we would expect ~30,000 molecules per cell in our experiments (out of ~106 cells per assay at 1% transfection efficiency ~10,000 cells are transfected, giving a concentration of ~6 fg per cell or 3x104 ßGal molecules). The chemiluminescent substrate Galacton (Tropix) can detect as low as 4,000 molecules per assay (Jain, V.K. and I.T. Magrath. 1991. A chemiluminescent assay for quantitation of b-galactosidase in the femtogram range: Application to quantitation of b-galactosidase in lacZ-transfected cells. Anal. Biochem. 199:119-124). However, in our system, background from cell extract prevents us from detecting less than 1 pg ßGal or less than 5x106 molecules per assay. In our cells Notch has a half-life of 3 hours and ßGal grater then 24. We estimate the response to Notch initiates at less then 1000 molecules/cell. By the time an individual cells express 100,000 notch molecules, the response of the population saturates since all cells receiving DNA made at least threshold amounts of Notch.