Biology Reference
In-Depth Information
a CD or 'read only memory' (ROM). Interestingly, epige-
netic information (genes which are still active, DNA
methylation, chromatin modification states and so on) from
state n to state n
1 may be considered as temporary
memory or, by analogy with computers, 'random access
memory
þ
RAM'. The epigenetic memory provides means
for self-renewal, where a given number of active network
components fluctuates between different semi-stable states.
Finding cell-specific markers or tissue-specific tran-
scriptional regulators is, in fact, tracing the emergence of
new information in differentiating cells. Separation of the
new, specific information from epigenetic memory may
help focus on the right candidate genes and reduce the
complexity of gene networks and the network-based
quantitative models.
e
FIGURE 22.7
Information flow in pluripotency and differentiation.
(A) Epigenetic memory is a set of active genes and gene products in a cell.
This information is analogous to random access memory of computers
(RAM). Upon differentiation, new information is added to the system with
newly activated genes, this new information is retrieved from genome. The
discrete genome information has analogy to read only memory of
computer devices (ROM). Gene activation is shown by the green
/
red
color gradient; shutdown of genes is shown by red
/
green gradient.
(B) Pluripotent and primed pluripotent states, attractors (shown by the
green and the red diamonds correspondingly) utilize epigenetic memory
(RAM), though the already present genes may by changing their expres-
sion levels. Exit to differentiating state (transitions between states are
shown by arrows) typically results in massive activation of new genes,
addition of new information (ROM) to the system. The diagram on (B)
may be considered as a projection of multidimensional space (such as that
shown in Figure 22.5) to a surface corresponding to concentrations of just
two core pluripotency factors.
Waddington Landscapes and Attractor States
Stem cells are capable of self-renewal and of differentiating
into terminal cell types. Waddington
[131]
described the
cells in a metaphorical way as marbles rolling down an
epigenetic landscape containing 'hills' and 'valleys' during
the development process: the 'valleys' represent cell types
separated by the 'hills'. From the network biology
perspective, cells settled in the 'valleys' are in the stable
states, i.e., fixed-point attractors. Alternatively, cells
oscillating around multiple 'valleys' are unstable. Different
stable or transitive states are defined by gene expression
pattern and epigenetic status. Huang et al. provided the first
experimental evidence that cell types can be represented by
high-dimensional attractor states
[132]
. Specifically, they
differentiated human leukemia cells (HL60) into neutro-
phils by two different stimuli. Using gene expression
microarray, they showed that HL60 adopted two different
trajectories of differentiation which finally converged into
the same 'end program' attractor state under the two
conditions. Random Boolean networks (RBNs) were
applied as a simple but generic model to mimic cell fate
commitment back in the 1960s
[84]
. When each node in
a random network had fixed number, K, of input nodes
(parents), the network displayed interesting dynamics. For
K
travel from one semi-stable state to another by means of
molecular noise in gene expression (see
Figure 22.6
B), or
expose cyclic behavior, described via limit cycles, damping
oscillations or even strange attractors.
REFERENCES
[1] Takahashi K, Yamanaka S. Induction of pluripotent stem cells
from mouse embryonic and adult fibroblast cultures by defined
factors. Cell 2006;126(4):663
e
76.
[2] Carvajal-Vergara X, et al. Patient-specific induced pluripotent
stem-cell-derived models of LEOPARD syndrome. Nature
2010;465(7299):808
e
12.
[3] Marikawa Y, et al. Dual roles of Oct4 in the maintenance of
mouse P19 embryonal carcinoma cells: as negative regulator of
Wnt/beta-catenin signaling and competence provider for Bra-
chyury induction. Stem 2011;20(4):621
e
33.
[4] Hayden EC. Stem cells: the growing pains of pluripotency. Nature
2011;473(7347):272
e
4.
[5] Kim HD, et al. Transcriptional regulatory circuits: predicting
numbers from alphabets. Science 2009;325(5939):429
e
32.
[6] Chiang PM, Wong PC. Differentiation of an embryonic stem cell
to hemogenic endothelium by defined factors: essential role of
bone morphogenetic protein 4. Development. 2011;138(13):
2833
2, the network exhibited 'critical' dynamics, which is
the borderline between chaotic and ordered states. Appar-
ently, the model captured the generic properties of bio-
logical gene regulatory networks corresponding to 'critical'
dynamics, which seems to reflect an equilibrium between
adaptiveness and robustness. Therefore, instead of stable
attractors, stem cells would be expected to reside in semi-
stable states in order to prepare themselves for environ-
mental changes, such as external differentiation signals.
Sometimes, such semi-stable states on the borderline of
chaos are called 'primed' states, poised for differentiation
[42,133,134]
(see
Figure 22.7
B). Apparently, cells may
¼
43.
[7] Son EY, et al. Conversion of mouse and human fibroblasts into
functional spinal motor neurons. Cell Stem Cell 2011;
9(3):205
e
18.
e
