Temporal Partitioning of the Yeast Cellular Network - Systems Biology of Metabolic and Signaling Networks - page 322

signalling during the oscillation. H 2 O 2 (500 M) added at a minimum of dissolved

O 2 both perturbed the respiratory oscillation and elevated H 2 S production in the
subsequent cycle. Menadione causes the generation of superoxide and perturbed the
oscillation in a similar manner as H 2 O 2 when added at 500

μ

M.
Further work on the effects of glutathione perturbation on branched-chain and
sulphur-containing amino acids tends to suggest that the observed effects are more
closely related to amino acid metabolism and H 2 S generation than with cellular
redox state per se (Sohn et al. 2005b ). This interpretation was supported by more
in-depth analyses of the transcriptome and metabolome (Murray et al. 2007 ).

μ

12.6 The Spatio-Temporal Self-Organisation of the
Reactome

With the advent of high-throughput analyses, e.g. microarray-based multiple assays
and mass spectrometric metabolomics, the yeast cellular network is the most fully
characterised among all eukaryotes. During its life cycle and in its response to
environmental challenges (e.g. on exposure to heat), between 20 and 50 % of the
yeast protein coding transcriptome alters (Gasch and Werner-Washburne 2002 ).
This has been confirmed by genome-wide and metabolome-wide elucidation of
properties of the respiratory oscillations in yeast (Klevecz and Murray 2001 ;
Klevecz et al. 2004 ; Li and Klevecz 2006 ; Murray et al. 2007 ; Slavov
et al. 2011 ). The high-throughput methods required new computational approaches
to curate and better understand the implications of vast array of new data.
High-throughput chromatin immuno-precipitates hybridised to DNA microarray
containing the probes for the upstream regulatory sequences or DNA tiling
microarrays (ChIP-chip) revealed that the underlying system structures involved
in global transcription can be profoundly altered in response to environmental
stimuli (Harbison et al. 2004 ; MacIsaac et al. 2006 ). The topology of the
protein-protein interaction network has been approached by series of immuno-
precipitation, 2-hybrid and mass spectrometry analyses (Ito et al. 2001 ; Gavin
et al. 2006 ; Krogan et al. 2006 ). More recently, the focus has turned to global
modes of regulation initiated by distinct protein complexes that dynamically mod-
ify chromatin structure (Basehoar et al. 2004 ; Whitehouse et al. 2007 ; Tsankov
et al. 2010 ). In conjunction with this, a concerted effort by the bioinformatics
community has resulted in a series of advances that have revolutionised the
manipulation and correlation of data obtained (e.g. KEGG, SGD or SwissPROT).
Although these advances have led to a deeper understanding of the structure of
the cell network, they have done little to advance our understanding of cellular
dynamics. Moreover, little has been done to combine the derived networks together,
and even less has been done to model large-scale datasets to produce a coherent
view of the formation of cellular phenotypes. With this in mind, we recently
combined computational and statistical network approaches, with high quality
transcriptional and metabolomic data (Fig. 12.3a ), to analyse the global landscape