Biomedical Engineering Reference
In-Depth Information
analysis of biological networks at a higher level of abstraction. Such approaches
could reveal salient features of biological phenomena missed by non-network-
based approaches.
The appearance of hierarchical modularity in biological networks supports
the assumption that evolution acts on many levels. The accumulation of local
changes, affecting the small highly integrated modules, slowly impacts the lar-
ger, less integrated modules as well. Thus, evolution might act in a self-similar
fashion, copying and reusing existing modules to further increase the organism's
complexity. Especially in the face of eukaryotic evolution, this network-based
framework might be suitable to describe the explosion of complexity in the de-
velopment of the single-celled
S. cerevisiae
toward the multicellular
H. sapiens
.
It is widely accepted that different cellular functions—such as information
storage, processing, and execution—are carried out by the genome, transcrip-
tome, proteome, and metabolome. Although the functional distinction between
these organizational levels is not always clear cut since, e.g., the proteome is
crucial for short-term information storage, all cellular functions can be described
by networks of various heterogeneous components. One way to visualize the
complex relationships between these components is to organize them into a sim-
ple complexity pyramid (38) in which various molecular components—genes,
RNAs, proteins and metabolites—organize themselves into recurrent patterns
such as metabolic pathways and genetic regulatory motifs. In turn, motifs and
pathways are seamlessly integrated to form functional modules that are respon-
sible for distinct cellular functions (19). These modules are nested in a hierar-
chical fashion and define the cell's large-scale organization (Figure 8).
Our present knowledge about the architecture of biological networks em-
phasizes two major aspects: (1) discrete cellular functions are mediated with the
aid of distinct albeit often-blurred modules; (2) network integrity is assured by a
handful of highly connected nodes, making networks robust against random
failures but exceedingly vulnerable to targeted attack. These features explain the
observation that many mutations have little or no phenotypic effect (55), which
appears to be consistent with the presence of genes that either cannot propagate
their failure or whose function can be replaced by other components of the net-
work. The presence of genes that integrate multiple signals and can trigger
widespread changes upon their failure proves the crucial role of highly con-
nected genes.
For example, the tumor suppressor gene
p53
has been identified as such a
highly connected and thus crucial node, which, once mutated, severely jeopard-
izes genome stability and integration of signals related to control of the cell-
cycle and apoptosis (30,54). Emphasizing its crucial role, disfunctional p53 pro-
teins are involved in more than half of all human cancer phenotypes. From a
biomedical point of view, highly connected proteins in general and proteins that
maintain the integrity of modules can be perceived as disease factors and thus
potential drug targets. With the increasing ability to identify and collect protein-
