Biology Reference
In-Depth Information
landscape
that C. Waddington sensibly proposed in the 1940s can be considered a
pivotal conceptual intermediate for understanding the one-to-many mapping between the
blueprint (genome) and the variety of cell phenotypes. The epigenetic landscape is akin to
a high-dimensional potential-like surface with multiple potential wells that represent
nonequilibrium (meta)stable steady-states and whose topography is encoded in the
genome.
7
'
With regard to devices or systems that exhibit multiple distinct states, in engineered systems,
orderly state transitions follow a predesigned and deterministic process between explicitly
designed a priori discrete states, e.g. the on
off state of a light signal. By contrast biological
systems exist in a continuous space that nevertheless exhibit discretely distinct behaviors
and achieve the same state through many different state transition paths. More concretely,
changes of system states (
phenotypes) in living cells which appear as quasi-discontinuous
switching are driven by noise-induced state transitions that can be biased if not explicitly
steered by biological signals.
8,9
Hence, the various states must be attractor states for the
latter to ensure discreteness and stability in a continuum of possibilities and allow for
multiple entry (converging) paths to each state. Complex multicellular organisms are
thought to have evolved the needed functionality by
5
of new attractors to the
existing system. In multicellular systems this has led to developmental paths from the egg
cell attractor to the attractor states of the stem cells of the various tissues, and from these
stem cells to the mature adult cell types, giving rise to a particular form of the landscape
that is encoded by the genome.
'
apposition
'
Evolution thus acts on the genome to shape the particular landscape topography.
The attractors are located on the landscape to allow the developing cell to access them in
a particular, controlled order
when needed during development and in homeostasis.
Access to an attractor is controlled locally, based on fine-tuned transition probabilities and
susceptibility to regulatory signals, such as hormones. Thus, while nondeterministic
processes, such as molecular noise and bifurcations,
10
are in general avoided in engineered
systems, biological systems exploit noise to drive the unfolding of the genome-encoded
system of constraints into a highly complex structure. They are poised between being robust
to noise (staying in the same attractor despite it) and being sensitive to noise (switching to
an accessible nearby attractor in response to it). In a cartoonish but useful simplification,
we can view development as the noise-driven successive occupation of the attractor states,
which can represent intermediate or terminal states
82
all encoded in the genome.
11
Thus the cellular states are predestined yet subject to randomness.
Here we are concerned with the modification of the developmental paths between attractors
with the ultimate goal to control attractor state transitions
which amounts to changing cell
types. This has become known as
While the switch from one cell type to
another of a closely related lineage has long been practiced in cell biology research,
12
and can
be achieved by manipulating just one
'
cell reprogramming.
'
gene, it is only a recent
demonstration of drastic changes of cell phenotype, the reprogramming of adult cells back to
an embryonic stem-cell-like state, referred to as iPS (induced pluripotent state)
13
that has
convinced the reductionist defenders of orthodoxy that cell-type identities are, after all,
not carved in stone but a dynamic entity. Such reprogramming can be robustly achieved by
manipulating multiple transcription factors although it remains a stochastic process.
5,14
'
fate-determining
'
The complete reversion of a cell
s phenotype, which in a first approximation can be
regarded as defined by the expression pattern of the tens of thousands of genes of the
genome, warrants the shift of the notion of explorative manipulation for analysis to the
realm of constructive, hence biosynthetic reengineering of a cell phenotype for practical
purposes. In fact, the prospect of manipulating cell types at will, starting from easily
available, proliferating cells (such as certain blood cells, liver cells, or fibroblasts) to more
challenging therapeutically desired cells, has fostered the hope that such universal cell
'
