3.4 | A comparison of competing objectives in
alternative alcohol pathways.
Our prior work on fatty acid biosynthesis indicates that coordinated
shifts in enzyme concentration can tune average chain length—a
critical design objective of engineered systems—without altering total
production (Mains et al., 2022). We sought to evaluate this engineering
strategy in four pathways for alcohol biosynthesis, where control over
length remains difficult. We began by supplementing our first alcohol
pathway, which relies on FadD, a CAR, and an aldehyde reductases (AHR),
with three additional pathways that extend the core FAS with 1-4
enzymes: (i) an acyl-ACP reductase (ATR), (ii) a TesA, FadD, and an
acyl-CoA reductase (ACR2), or (iii) TesA, FadD, a different acyl-CoA
reductase (ACR1), and an aldehyde reductase (AHR). As before, we added
new enzymes to our base model and optimized poorly characterized kinetic
parameters with fits to experimental data (Fig. S5). With our four
alternative models in hand, we examined the range of average chain
lengths accessible with simple shifts in enzyme concentration.
Intriguingly, for all pathways, the breadth of chain lengths decreases
as total production increases, and the highest production levels
culminate in a single average chain length (Fig. 6). This tradeoff is
sharpest for the ATR and ACR1 pathways (Figs. 6B and 6D), where critical
enzymes—namely ATR and ACR1—have very narrow substrate
specificities. The CAR and ACR2 pathways, by contrast, enable broad
control until very high production levels, with the CAR pathway
providing the greatest flexibility (Fig. 6A and 6C). The control
afforded by this pathway reflects the broad substrate specificities of
all enzymes downstream of the FAS.
An analysis of enzyme compositions at high and low average chain lengths
reveals the mechanisms of control (Dataset S1). For the CAR, ACR1, and
ACR2 pathways, high concentrations of TesA and low concentrations of
FabF and FabB facilitate the production of short-chain alcohols, while
the opposite shifts encourage long-chain products. This trend is
consistent with influence of TesA, FabF, and FabB on fatty acid
biosynthesis (Mains et al., 2022). Results also suggest that a reduction
in FabA concentration and increase in FabZ can promote the production of
short-chain alcohols—perhaps, the result of the preference of FabZ for
short substrates—but this effect is less pronounced (i.e., changes in
enzyme concentrations are smaller than those of TesA, FabF, and FabB).
For the ATR pathway, which has no thioesterase, chain length appeared
most sensitive to the concentrations of FabB and FabA (with trends
consistent with those exhibited by the other pathways).
For each pathway, we used a total enzyme concentration informed by the
overexpression strategy described in an experimental study. The ATR and
ACR1 enzymes, however, are not particularly active, and high enzyme
concentrations could blunt the tradeoff between chain