Genetic approaches have been traditionally used to understand
microbial
metabolism, but this process can be slow in nonmodel organisms due
to limited genetic tools. An alternative approach is to study metabolism
directly in the cell lysate. This avoids the need for genetic tools
and is routinely used to study individual enzymatic reactions but
is not generally used to study systems-level properties of metabolism.
Here we demonstrate a new approach that we call “cell-free
systems biology”, where we use well-characterized enzymes and
multienzyme cascades to serve as sources or sinks of intermediate
metabolites. This allows us to isolate subnetworks within metabolism
and study their systems-level properties. To demonstrate this, we
worked with a three-enzyme cascade reaction that converts pyruvate
to 2,3-butanediol. Although it has been previously used in cell-free
systems, its pH dependence was not well characterized, limiting its
utility as a sink for pyruvate. We showed that improved proton accounting
allowed better prediction of pH changes and that active pH control
allowed 2,3-butanediol titers of up to 2.1 M (189 g/L) from acetoin
and 1.6 M (144 g/L) from pyruvate. The improved proton accounting
provided a crucial insight that preventing the escape of CO2 from the system largely eliminated the need for active pH control,
dramatically simplifying our experimental setup. We then used this
cascade reaction to understand limits to product formation in Clostridium thermocellum, an organism with potential
applications for cellulosic biofuel production. We showed that the
fate of pyruvate is largely controlled by electron availability and
that reactions upstream of pyruvate limit overall product formation.