Figure 6. A comparison of different pathways for alcohol
biosynthesis. (A-D) The range of average chain lengths afforded by
alternative alcohol pathways: FAS supplemented with (A) TesA, CAR, and
AHR, (B) acyl-ACP reductase (ATR), (C) acyl-ACP thioesterase specific
for medium chains (TesA), a fatty-acid-CoA ligase (FadD), and an
acyl-CoA reductase (ACR2), and (D) TesA, FadD, a differentacyl-CoA
reductase (ACR1), and an aldehyde reductase (AHR). At each production
level, the shaded region denotes the average chain lengths achievable by
changing enzyme concentrations; the range of lengths narrows as
production increases. Dataset S1 contains the source data and enzyme
compositions for plots A-D.
length and production. We explored this effect by repeating the ATR
analysis with a much higher enzyme concentration
(~ninefold). This adjustment broadened the range of
accessible chain lengths at intermediate production levels. We note,
however, that our model does not account for the potential for ATR to
bind holo-ACP, an interaction that can inhibit fatty acid synthesis and
may limit the usefulness of ATR overexpression.
A sensitivity analysis of all pathways examined in this work suggests
that, in general, product profiles are most sensitive to concentrations
of acyl-ACP thioesterases and β-ketoacyl-ACP synthases (Fig. S11). This
sensitivity, however, does not imply that all pathways are compatible
with concentration-based control strategies; the narrow substrate
specificities of ATR and ACR1, for example, are limiting. Rather, our
focused analysis of the alcohol pathways suggest that coordinated
changes in enzyme concentration provide a versatile means of tuning the
product profiles of pathways with promiscuous enzymes downstream of the
FAS.