Temporal Partitioning of the Yeast Cellular Network - Systems Biology of Metabolic and Signaling Networks - page 318

Klevecz 2006 ). A multidimensional clustering approach identified seven temporal
clusters that formed a coherent growth programme in these data. When compared to
a large compendium of independent datasets, it has been shown that the formation
of these clusters is both transcription factor and chromatin state dependent. This
leads to the conclusion that a simple feedback between energetics and chromatin
state may be one of the primary regulatory loops involved in defining the transcrip-
tional landscape (Machn ´ and Murray 2012 ). Furthermore, the DNA synthesis is
intimately timed during each respiratory cycle (Klevecz et al. 2004 ), although the
exact link between the cell division cycle and the respiratory oscillation is unclear,
as DNA synthesis is not limited to one phase of the oscillation. Moreover, respira-
tory oscillations occurred under conditions where changes in DNA synthesis were
not detectable (Slavov et al. 2011 ).
Periods can range from 35 min to several hours depending on environmental
conditions (Finn and Wilson 1954 ; von Meyenburg 1969 ; Sohn and Kuriyama
2001a ; Murray and Lloyd 2006 ; Slavov and Botstein 2011 ; Machn´ and Murray
2012 ). However, the common modes of oscillation appear to be between 40 min
and 5 h. The period is temperature compensated (Murray et al. 2001 ) and thus its
timekeeping characteristics (Edwards and Lloyd 1978 , 1980 ; Lloyd et al. 1982a ;
Marques et al. 1987 ) place it in the ultradian time domain (as defined by its
characteristic of cycling many times during a day). That similar clock control can
be demonstrated when ethanol medium is used further distinguishes these
oscillations from glycolytic and cell-cycle associated oscillations, both of which
are characterised as having highly temperature-dependent periods (Lloyd 2006a ).
Moreover, several other properties of the respiratory oscillation are shared with the
circadian clock, e.g. period sensitivity to Li + and type-A monoamine oxidase

inhibitors (Salgado et al. 2002 ).

12.3 Phase Definitions Guided by Real-Time Monitoring of
Redox State

Here, we define the phase of the oscillation by the dissolved O 2 trace (Fig. 12.1 ), as
this is one of the most stable measurements and responds rapidly to changes in
culture concentrations (Murray et al. 1998 ; Murray 2006 ). We use the minimum
first derivative of this trace to define the reference start point; the difference

Fig. 12.1 (continued) measured using an online flurorimeter. Carbon dioxide excretion rate
(qCO 2 ; d) was calculated from the partial pressure of CO 2 in the fermentor off-gas using an IR
sensor. Hydrogen sulphide excretion rate (qH 2 S; e) was calculated from the partial pressure of H 2 S
in the fermentor off-gas using a silver nitrate sensor. Heat transfer (f) was calculated by Fourier's
equation from the reactor temperature and the controlling bath temperature. The grey line
represents the minimum and maximum first derivatives, i.e. the demarcation between the oxidative
( Ox ) and reductive ( Red ) phases (Sasidharan et al. 2012 )