The dynamic origins of phenotypic heterogeneity and genotypic instability a
nd hypermutation have been investigated in simulated tissues comprised of 9
00-25 600 cells each represented by initially identical Rossler attractors
running in the chaotic domain. This attractor, representing the cell cycle
behavior of individual cells in the array, has previously been used to mode
l the dynamic behavior of mammalian cells in culture. In these tissue const
ructs, the behavior of an individual cell is modified by its interactions w
ith its immediate neighbors as a consequence of diffusive coupling through
one of the variables. Differentiation within the initially identical popula
tion of attractors is manifested as a position dependent set of novel stabl
e trajectories in phase space that are revealed through the use of return m
aps. These self-mapping patterns, which we define as the phenotype of the c
ell, are periodic and stable over a considerable period of time. A comparis
on of tissues whose individual cell cycle attractor phases describe an arch
imedean spiral with those that exhibit S-T chaos, or turbulence, suggests t
hat the heterogeneous phenotype of tumor tissues is better modeled by turbu
lence. Instability in the spiral array exists primarily at the boundary bet
ween periodic regions of differing phase and trajectory, and involves infre
quent excursions by these boundary cells away from their stables trajectori
es. Such instabilities are hypothesized to play an important role in the am
plification, hypermutation, and gene conversion events seen in certain norm
al biological tissues and tumors. (C) 1998 Elsevier Science B.V.