3.3 | A sensitivity analysis of different models.
Enzyme concentration is an important design consideration for engineered
pathways, where it can be controlled with carefully selected promoters,
ribosome binding sites, and dynamic transcriptional systems (Meyer et
al., 2019; Salis et al., 2009; Zhao et al., 2018). We used a global
sensitivity analysis to determine how changes in enzyme concentrations
influence three important engineering objectives—the total production,
unsaturated fraction, and average length of oleochemical products (Fig.
4). Our results indicate that total production is most sensitive to the
concentrations of acyl-ACP thioesterase and downstream
oleochemical-specific enzymes, while product profile (i.e., unsaturated
fraction and average chain length) is most sensitive to concentrations
of acyl-ACP thioesterase and core FAS enzymes. These sensitivities are
consistent with the experimentally observed improvements in titer (and
constant product profiles) that result from the overexpression of
oleochemical-specific enzymes (Song et al., 2016; Yan et al., 2020).
The specific effects revealed by our analysis are consistent within vivo studies. We will briefly comment on a few examples:
Starting with the alcohol pathway, we found that concentrations of TesA,
CAR, and AHR have a pronounced influence total production, while TesA
and FabF affect average chain length. The original study of this pathway
did not examine each enzyme in isolation—a difficult feat for an
experimental analysis—but it found that the incorporation of a highly
active AHR could increase alcohol production (Kalim Akhtara et al.,
2013); additionally, shifts in fatty acid compositions produced
corresponding shifts in alcohol