(474f) Prediction of Weak Acid Toxicity in S. Cerevisiae Using Genome-Scale Models

Introduction

Renewable and economically feasible production of biofuels
and biochemical relies on the efficient use of lignocellulosic biomass as a
feedstock.  However, necessary pretreatment
of the biomass generates an array of growth inhibitors such as weak acids, phenolics and furan derivatives that adversely affect the
fermentation characteristics.  The
concentrations and relative amounts of each inhibitor in the hydrolysate vary
among the sources of lignocellulosic biomass [1].  Acetic acid is a large contributor to
weak acid inhibition in lignocellulosic hydrolysate and ranges in concentration
from 1.6-4.4 g/L [1].

One mechanism by which weak acids inhibit cellular growth,
known as the uncoupling theory, is a result of the diffusion of undissociated weak acids through the cytosolic membrane
into the cytosol. Due to the near neutral pH of the cytosol, the weak acid
dissociates into its conjugate base and proton, thereby decreasing cytosolic pH.  In order to
maintain pH neutrality, protons are pumped from the cytosol against the
concentration gradient by the membrane bound ATP synthase at the cost of ATP,
thus reducing the amount of ATP available for cell growth.  Alternatively, the accumulation of
anions in the cytosol has been suggested to play a role in growth inhibition.

Metabolic modeling has emerged as a powerful tool for
enhancing our understanding of microorganisms and aiding in design of microbes
for specific phenotypes.  Previous
studies have modeled the impact of weak acids on microorganisms using kinetic
and thermodynamic approaches.  The
focus of such studies has largely been the toxic affects of weak acids in
bacteria for food preservation purposes. 
However, the yeast S. cerevisiae is
the most abundant organism used for industrial fermentations, as it is known to
be more robust to inhibitory compounds present in hydrolysate.  It would therefore be of interest to
examine the affects of weak acids on the growth of an organism relevant to the
biotechnology industry such as S.
cerevisiae.

An alternative modeling approach, flux balance analysis
(FBA), is a commonly used linear programming based, steady state framework that
uses a stoichiometric model of the organism and a set of physicochemical and
thermodynamic constraints to establish intracellular fluxes. 

This work is a combined computational and experimental
approach to examine the affects of weak acid toxicity in S. cerevisiae.  We
focused on the toxicity of acetic acid as it is an
abundant inhibitor in lignocellulosic hydrolysate.  It was hypothesized that the energy
usage required to pump protons out of the cytosol via ATP-synthase could be
represented using a genome-scale metabolic model.  Specifically, we extended the iMM904
consensus model of S. cerevisiae to
include the uncoupling mechanism of weak acid toxicity [2].  We implemented the uncoupling mechanism
by adding a series of reactions to the iMM904 model that describe the
inhibitory effect.  Growth rate was
optimized with glucose as the carbon source under varying concentrations of
acetic acid and pH under both aerobic and anaerobic conditions.  Predicted growth rates and product
profiles of S. cerevisiae were
compared to experimental results for aerobic conditions obtained in this study,
as well as anaerobic conditions reported by Tahezadeh
et al [3].

Batch cultivations of S.
cerevisiae CEN.PK MAT A/a were conducted in glucose mineral medium under aerobic
conditions in the presence of acetic acid ranging from 0-4 g/L.  Throughout the cultivations, temperature
and pH were maintained at 30°C
and 5.5, respectively.  Growth and
metabolic concentrations were monitored throughout the batches by optical
density and HPLC measurements.

Results

Growth of microorganisms such as S. cerevisiae in the presence of acetic acid is hindered due to the
consumption of ATP necessary to maintain near-neutral cytosolic pH.  Model
predictions were consistent with the aerobic growth rates in the batch cultures
with varying concentrations of acetic acid at pH 5.5, as shown in Figure
1.  In this figure, the experimental
growth rate is shown to decrease linearly (R² = 0.93) with
increasing acetic acid concentration. 
Growth rates predicted by FBA at pH 5.5 follow
this trend closely and predicted flux of metabolites were also in agreement
with experimental data. 
Interestingly, these results suggest that the toxic affect of acetic
acid may be largely attributed to the uncoupling mechanism, rather than anion
accumulation.

Growth rates and metabolite profiles were also predicted for
varying acetic acid concentration for anaerobic conditions, and compared to the
experimental values reported by [3]. 
Model predictions of growth rate and ethanol yield in this case were
also in good agreement with experimental values.

Figure 1:
Experimental and predicted growth rates of S.
cerevisiae in the presence of acetic acid.  Error bars indicate the standard
deviation of two or more batch cultivations.

Conclusion

It was shown in this work that the genome-scale metabolic
model is able to capture the growth associated energetic burden incurred by the
presence of acetic acid.  The
consensus model of S. cerevisiae was
extended to include the uncoupling mechanism of weak acid toxicity, whereby
intracellular dissociation of acids generates protons that are subsequently
exported at the expense of ATP. 
Model predictions of acetic acid toxicity agreed well with aerobic batch
cultivation data of S. cerevisiae
CEN.PK MAT A/a.  The model was also shown to predict the
toxicity of acetic acid during anaerobic growth.

We expect this work to find application in the optimization
of industrial processes.  By
systematic evaluation of the model, the trade off between product yield,
cellular growth rate and neutralization cost may be established, thereby
improving the bioconversion of lignocellulosic hydrolysate to renewable
chemicals.

References

[1] Almeida JRM, Modig T, Petersson A, Hahn-Hagerdal B, Liden G, Gorwa-Grausland MF. (2007) Increased tolerance and
conversion of inhibitors in lignocellulosic hydrolysates
by Saccharomyces cerevisiae. J Chem Technol Biotechnol, 82:340-349.

[2] Mo ML, Palsson BO, Herrgard MJ. (2009) Connecting
extracellular metabolomic measurements to intracellular flux states in yeast.
BMC Syst Bio,
3:37.

[3] Taherzadeh MJ, Niklasson C, Liden G. (1997) Acetic acid – friend or foe in
anaerobic batch conversion of glucose to ethanol by Saccharomyces cerevisiae? Chem Eng Sci, 52(15),
p.2653-2659.