(352b) Coordination of Quorum Sensing with Cell Motion

                                                                           
  Coordination of Quorum
Sensing and Cell Motion

                                                                                     
   David N. Quan,
William E. Bentley

Through computer aided modeling, emergent phenomenon may be
elucidated?how the interaction of a billion simple answers (forged through
reductionist approaches) create a cohesive cellular/organismal response can be
understood.  A first step, of course, is to investigate the isolated
interaction of two simple answers.  Here, we investigate phenomenon arising
from the overlap of quorum sensing (QS) and cell motion. 

QS is a concentration dependent phenomenon used by bacteria
to sense local diffusion conditions or local cell concentration.  Among the few
known and first identified QS agents are the autoinducers (AI) AI-1 and AI-2. 
AI-1 is an umbrella term for species specific AHL's secreted from Vibrio
bacteria which, after achieving a sufficient concentration, trigger a positive
feedback mechanism.  AI-2 is much more broadly distributed QS agent working
either through LuxPQ, a two component regulatory system, or the more elaborate LuxS-regulated
(Lsr) system that recompartmentalizes AI-2 to achieve quorum activation. 
Because these QS agents are small molecules, the effective concentration that
the cells are exposed to is highly dependent upon local convection and diffusion
dynamics, which are influenced?especially stochastically?by cell transport
(motility).  Here, we predict nontrivial phenomenon arising from the confluence
of these transports (QS agent and cell) based on in silico experimental
results. 

Cells were treated as autonomous single agents that either
swam or glided within a finite element environment through which AI-2
diffused.  At each time point, effective QS agent concentration was integrated
across local grids, QS state (whether the genes were switched on) was
evaluated, QS agent exchange between the cell and the media took place, cells
divided and moved stochastically, and QS agent diffused between grids. 

Firstly, we successfully recapitulated the supernova-like
activity pattern of a surface attached growing colony under the influence of
AI-1 QS.  Next, we used an ODE model of AI-2 activity to inform the stochastics
governing Lsr behavior.  Applying the forms of the ODEs to colony growth in our
finite element/single agent scheme resulted in a stable, abrupt checkered
activation.  We extended this particular model by adding AI-2 synthase (LuxS)
controlled by the lsr promoter, and predict that the proportion of QS+
cells is enhanced depending on the plasmid copy number and where in parameter
space the system operates.  Lastly, we modeled the overlap of Lsr based QS
dynamics with swimming cells that chemotax according to AI-2 gradient.  In
these simulations, AI-2 and cell transport were limited by virtual boundaries. 
Over the course of the experiment, cells/AI-2 coalesced in random clusters
along these boundaries, especially in the corners, trapped by a positive
feedback loop.  This behavior persisted until QS activation, at which point the
driver of chemotaxis was swept up and bacteria within these clusters
redispersed.