Discrete and Continuum Multiscale Behaviour in Bacterial Communication - Multiscale Computer Modeling in Biomechanics and Biomedical Engineering - page 305

Gram-positive bacteria. First identified in the pathogen S. aureus [ 17 ], homologues
of the agr operon have since been found to also affect virulence factor production
in Listeria monocytogenes [ 24 ], Enterococcus faecalis (the fsr system) [ 14 ] and
Clostridium botulinum [ 2 ] and perfringens [ 1 , 13 ] amongst others [ 29 ]. This
system, however, is not restricted solely to pathogenic bacteria: it has also been
found to regulate granulose formation and sporulation in C. acetobutylicum [ 27 ]
and cell morphology in Lactobacillus plantarum [ 4 ] for instance. The prevalence
of this system and the fact that it controls such a wide range of cell phenotypes in
different species render it particularly important to understand. Furthermore, since
the signal molecules can be secreted from the cells, cross-talk between strains and
species has been found to occur (for examples relating to S. aureus see e.g. [ 11 , 19 ]
or [ 20 ]). Thus modelling of the agr operon traverses a number of scales—from
subcellular gene regulation, through cell phenotype, to population behaviour and
the repercussions upon multiple populations. We present here a model of a single
species, but examples involving cross-talk between species and strains can be
found in [ 8 ] and [ 9 ] and between two quorum-sensing systems within one strain in
[ 12 ]or[ 21 ] for example (though the last two studies consider quorum-sensing in
Gram-negative bacteria, so the systems studied are not agr homologues).
4.2 Model Formulation
Known agr operons have varying degrees of feedback contained within them:
some have every element of the operon up-regulated in response to signal mole-
cules, others only the elements controlling signal synthesis or those governing
signal detection and response. In order to ensure bistability 3 in our system, we
consider an agr system in which every element is induced in response to increased
levels of quorum-sensing signal molecule, i.e. that first identified in S. aureus.
The operon consists of two genes (agrB and agrD) that interact to produce the
signal molecule (termed AIP: autoinducing peptide) and two genes (agrA and
agrC) which produce proteins to detect and respond to the AIP (these form a two-
component system); see Fig. 1 . Ultimately, sufficiently high levels of signal
molecule induce high levels of phosphorylated AgrA (AgrA * P), which increase
transcription of both the agr genes and the downstream target genes (e.g. those
controlling virulence factor production). We have previously derived a spatially
structured model of the agr operon [ 9 ] which, under suitable simplifying
assumptions, could always achieve upregulation should there be no interference
with the operon. In order to investigate bistability, we cannot employ all of these
simplifications and detail must be added back into the model. The full (ordinary
3 Note that we are thus concerned in this section with the interactions between two stable states,
whereas the focus in the previous sections was on how the solution is driven away from an
unstable state (namely the trivial one).