Fig. 3. The nuclear accumulation of the FER proteins depends on kinase domain integrity. A, wild-type fer-intact p94fer (a), wild-type ferT- intact p51ferT (b), ferT{Delta}1–58 (c), fer{Delta}758–823 (d), fer{Delta}594–634 (e), and ferT{Delta}446–453 (f), all linked to a HA epitope, were exogenously expressed in actively growing COS1 cells. Cells were fixed and stained with {alpha}-HA monoclonal antibody (Babco). Photographs represent stacked confocal laser sections taken 1 µm apart. Scale bar, 20 µm. B, summary of p94fer (fer) and p51ferT (ferT) constructs whose subcellular distribution profiles were determined in growth arrested confluent (conf.) and nonconfluent (nonconf.) COS1 cells. The symbols within the schemes of the tested constructs (structure) are as in Fig. 1A . Enzymatically active (+) and nonactive (-) constructs which were tested in an autophosphorylation assay (kinase act.) are marked. The subcellular distribution profiles are as follows: -, 0–5% of the molecules were found in the nucleus; -/+, 5–20%; +, 20–40%; ++, 40–60%; +++, 60–80%; ++++, 80–95%; and +++++, 95–100%. These values were obtained from at least three independent experiments in which at least 50 transiently transfected cells were obtained for each construct. The expressing cells were then examined by eye for the relative intensity of staining.