Complex Systems Biology of Networks: The Riddle and the Challenge - Systems Biology of Metabolic and Signaling Networks - page 27

• Network organization
Mass-energy/information/signaling networks exhibit an overall loop topol-
ogy. They comprise reaction and transport processes, and some nodes in these
networks represent hubs since they exhibit multiple inputs and outputs while
most nodes only possess a few of them. This feature confers these networks the
trait of “scale-free.” The topologically circular connectivity present in these
networks bestows them with self-making and -maintaining properties that com-
bined with continuous energy and matter exchanges allow them to self-organize
in space and time (Aon and Cortassa 2009 ; Aon et al. 2012b ). Although
constructed with a high degree of redundancy that confers these networks
resilience to attack (Barabasi 2009 ), under stress they may reach critical
conditions that make them collapse, especially if hubs fail (see Chap. 5 ) .
• Top—down and bottom—up interrelationships (heterarchies)
Cells, tissues, and organs can be viewed as networks within networks. One of
the most distinctive features of these networks is that all components interact one
way or another, constituting a heterarchy (Aon and Cortassa 2012 ; Aon
et al. 2012b ; Lloyd and Rossi 2008 ; Yates 1993 ). In a heterarchy, but unlike in
a strict hierarchy, interactions between network components and relevant func-
tional interrelationships (including regulatory ones) flow between levels in both
directions, top-down and bottom-up. This has important consequences for con-
trol and regulation of integrated metabolic and transport networks where every
reaction, metabolite, ion, and process, may contribute, although to differing
extents, to the overall control and regulation of the network (Aon and Cortassa
2012 ; Cortassa et al. 2012 ).
• Control is distributed, and operates through “diffuse loops”
Systemic analysis of extensive networks as given by Metabolic Control
Analysis shows that every process (edge, e.g., enzymatic reaction, channel)
controls and is controlled by every other process in the network. However, the
strength of control exerted by different processes may vary significantly, a trait
that relates the fact that control is distributed (see Chaps. 3 , 9 and 13 ) . In the case
of nodes (e.g., metabolites, second messengers, cofactors), every node can
regulate other processes and in turn can be controlled (e.g., its concentration)
by a process. The character of “distributed control” relates the fact that different
processes (edges) exert control, and can be “diffuse” as well as direct. A diffuse
control was first described in an integrated computational model of the
cardiomyocyte and corresponds to the control exerted by seemingly indirectly
related processes through shared ubiquitous cofactors such as ATP, ADP, and
Ca 2+ (Cortassa et al. 2009a , b ).
In networks involving various compartments, not all the control of the flux,
e.g., in an organelle, resides within the organelle itself. In the heart, the control of
mitochondrial respiration is exerted by cytoplasmic and sarcolemmal
membrane-linked processes, e.g., the myofibrillar and Na/K ATPases, in addi-
tion to processes residing in the mitochondrion (Aon and Cortassa 2012 ). This is
especially true under working conditions, when the interaction between cyto-
plasmic and mitochondrial processes is quantitatively more important.