Biology Reference
In-Depth Information
maintain the lysogenic state to which it committed when
favored by the prevailing environmental conditions. This
state is maintained for a period, even in the face of dete-
riorating environmental conditions
values for these parameters range from a low of 1.6-fold to
a high of 83-fold, with an overall average of 22-fold.
These results suggest that the system design of the cI
gene circuit is not only robust to local (small) changes in
parameters, but it is also remarkably tolerant to global
(large) changes in parameters and environmental stimuli.
This is consistent with thework of John Little and colleagues
[73
presumably an
advantage if the host is capable of recovering from suble-
thal damage. However, once committed to induction, there
is a penalty for reversing the cell-fate decision, presumably
because such vacillation would reduce the titer of phages
ultimately produced to initiate a new round of infection if
the host is damaged beyond repair. The determinants of the
size of this buffer zone and its tolerance to large changes in
the structural parameters may not be obvious, but they are
readily quantified by means of the generic concept of
phenotype and its manifestation in the system design space.
Near the hysteretic region fluctuations will result in
some phages undergoing induction while others will not.
This produces heterogeneity in the population that hedges
bets in an unpredictable environment.
e
75]
, inwhich they have constructed variants withmajor
changes in the binding constants, and indeed the identity of
the circuit elements, and yet many of their constructs retain
the essential qualitative features of the wild-type phage.
The relationships among these various elements of
design, and their tolerances to large changes in parameter
values, are not at all obvious, but again they are readily
quantified by means of the generic concept of phenotype
and its manifestation in system design space.
There are obvious analogies to other cell-fate decisions
that are well known in bacteria as well as higher organisms.
However, the detailed mechanisms involved can be quite
different. For example, the decision in the case of lambda
induction involves feedback regulation of a transcription
initiation mechanism. That, for the case of spore formation
in Bacillus subtilis, involves feedback regulation in a post-
translational partner-switching mechanism
[34]
. The deci-
sion to switch from a proliferating eukaryotic cell to
apoptosis
[76]
has clear similarities to the decision to switch
from a stable lysogen to induction. Although the mecha-
nisms that govern these systems are very different and most
are not fully characterized, once an appropriate model has
been formulated the procedure to develop the corresponding
system design space is conceptually straightforward.
e
Maintaining the Temperate Lifestyle
Perhaps even more important are the system design prin-
ciples that ensure long-term survival of the temperate
lifecycle itself. If the constellation of structural parameters
that determine the upper bound on K
D
is violated, the phage
becomes locked into the lytic mode of growth. The same
result is obtained if the parameters experience variation that
would shift the lower boundary of the hysteretic region to
values lower than the basal value of the environmental
stress. On the other hand, if the constellation of structural
parameters that determine the lower bound on K
D
is
violated, the phage will reside permanently in the host's
chromosome as a lysogen. Again, a similar result is
obtained if the parameters experience variation that would
shift the upper boundary of the hysteretic region to values
higher than the maximal value of the environmental stress.
Maintenance of the phenotype (large tolerances to the
adjacent phenotypes, e.g., Case 9 and Case 17 relative to
Case 11), proper positioning of a wide hysteretic buffer (left
margin greater than the basal value and right margin less
than the maximal value of environmental stress), and proper
positioning of K
M
within its acceptable band of values (less
than the upper bound and greater than the lower bound) are
best ensured by having large global tolerances for all the
parameters; thus, one can consider a minimum net tolerance
for each of the parameters to be most relevant, as shown in
Table 15.6
. It should be noted that the tolerances mentioned
above for the hysteretic buffer are all larger than these
minimum net tolerances. The most restrictive net tolerances
are for changes in the exponents a and p ([1.39,50.0] and
[1.34,1.07], respectively, although most of the other toler-
ances for change in these exponents are essentially infinite).
Most of the tolerances for change in the other parameters are
essentially infinite; but if these are excluded, the resulting
CONCLUSIONS AND FUTURE
CHALLENGES
While the phenotypes and design principles in the case of
very simple systems, such as the reversible pathway of the
earlier example, may be obvious (or at least may become
obvious in hindsight), this is seldom the case with more
complex systems, as the example of lambda illustrates. We
should have no illusions about the extreme challenges of
relating genotype and environment to phenotype for
complex organisms. However, without a generic concept of
phenotype there can be no deep, predictive understanding
of these important relationships. What then are the major
challenges to developing this type of understanding?
The path forward will require major advances in
moving from the digital information in the genome
sequence to an analog model that captures the essential
mechanistic underpinnings of system behavior in a given
environmental context. Many of the approaches in this
book, and others that have yet
to be developed, will
undoubtedly play a critical
role in overcoming this
