Temporal Partitioning of the Yeast Cellular Network - Systems Biology of Metabolic and Signaling Networks - page 316

12.2 Continuous Culture of Yeast: An Ideal System
for Study

Temporal compartmentalisation of the progress of energy generation, metabolic
transformations, synthesis and assembly of membranes and organelles, as well as
the organisation of chromosomal dynamics and the cell division cycle, requires
studies either in single cells or in synchronous populations of cells or organisms.
Whilst the former is even now restricted, the latter is usually problematic because
of limitations associated with the preparation of material without perturbation.
The observation that budding yeasts form stable oscillatory dynamics during
continuous culture was first reported only a year after the DNA double helix
model (Watson and Crick 1953 ; Finn and Wilson 1954 ). However it was not until
the early 1990s, when high temporal resolution, computer acquisition became
available that this spontaneous self-synchrony of dense (~5 10 8 organisms/ml)

cultures of S. cerevisiae could readily be monitored (Fig. 12.1 ) to reveal short
period and multi-timescale oscillatory dynamics (Satroutdinov et al. 1992 ; Keulers
et al. 1994 ). The strain primarily employed (IFO-0233, IFO, Institute of Fermenta-
tion, Osaka, Japan) was an acid-tolerant diploid yeast (Naiki and Yamagata 1976 ).
During batch growth glucose is consumed and the culture produces biomass,
CO 2 ,H 2 S and ethanol as well as many other fermentation products. The pH is
controlled at 3.4 and air flow rate is kept constant. The air flow rate is calculated
for each reactor according to its specific oxygen transfer coefficient (Mueller
et al. 2012 ). When ethanol is completely used up, depletion of trehalose and
glycogen in the second stage of this diauxic growth process results in the initiation
of oscillatory respiration, as indicated by 40-min cycles of dissolved O 2 and CO 2
(Murray 2004 ). A steady and continuous supply of growth medium at this stage
provides material adequate both for long-term monitoring of the organisms through
many thousands of generations over a period of many months, and discrete time
samples for biochemical analyses. The rapidly responding probes immersed in the
culture give either continuous outputs (e.g. for NAD(P)H fluorescence) or fre-
quently sampled voltages at 0.1-10 Hz (Keulers et al. 1996a ; Murray et al. 2007 ;
Sasidharan et al. 2012 ). The most convenient readout, dissolved O 2 , can also be
monitored by membrane inlet mass spectrometry (MIMS) (Roussel and Lloyd
2007 ) as can CO 2 ,H 2 S or ethanol. Off-gas measurements for CO 2 ,O 2 and H 2 S
can either be measured by MIMS (Keulers et al. 1996a ) or by an array of dedicated
sensors (Murray et al. 2011 ) to quantify transfer rates. Near-infrared spectroscopy
can be used online for ethanol, glucose, NH 3 , glutamine and biomass (Yeung
et al. 1999 ).
The oscillatory state of this autonomously self-sustained system is observed
when the cultures are supplied with glucose, ethanol and acetaldehyde as the main
carbon source (Keulers et al. 1996b ), implying mechanistic differences between
respiratory and glycolytic oscillations. It exerts control on the production of
hundreds of metabolites and many transcription factors (Murray et al. 2007 ). The
genome-wide oscillation in transcription is pivotal (Klevecz et al. 2004 ; Li and