Effect of FGFR on EphB2 cell segregation and repulsion. Stable HEK293 cell lines were generated that express EphB2, EphB2+GFP, EphB2+iFGFR+GFP, or ephrinB1. Cell behavior was studied after mixing different combinations of these cell lines with GFP-expressing and -nonexpressing cells visualized by fluoresence and relief-contrast microscopy. (a–c) Cell segregation assays in which cells are plated at moderate density and incubated for 48 h, during which time they achieve confluence. (a) EphB2+GFP cells remain intermingled with EphB2 cells. (b) EphB2+GFP cells segregate from ephrinB1 cells. (c) EphB2+iFGFR +GFP cells fail to segregate from ephrinB1 cells. (d–f) Time-lapse videos of typical behaviors of EphB2-expressing cells after interaction with ephrinB1 cells. (d) An EphB2+GFP cell is not repelled and remains in contact with an EphB2 cell. (e) Upon touching an ephrinB1 cell, an EphB2+GFP cell rapidly retracts and rounds up for >35 min (n = 10/10). (f) After interaction with an ephrinB1 cell, an EphB2+iFGFR+GFP cell retracts but does not round up and reestablishes cell processes by 25 min (n = 8/8). (g) Quantitation of the results of cell segregation assays (a–c) was performed by the nearest neighbor method (Mochizuki et al., 1998). For a random distribution of two populations, on average half of the contacts of any cell type are with cells of the same type. The proportion of contacts between like cells increases for segregated populations. Because of clonal growth, cells are not randomly distributed in control assays. Error bars indicate SEM (n = 4).