Article
Extended Data Fig. 2 | Substrate specif icity and structure-guided active
site engineering of UGT84A27 in engineered yeast. a, Phenylpropanoids
tested as glucose (Glc) acceptors for UGT84A27 in engineered yeast.
To p, (d)-3-phenyllactic acid (PLA); middle, trans-cinnamic acid (CA); bottom,
trans-ferulic acid (FA). b, Heat map of the percentage conversion of fed
phenylpropanoids to glucosides by yeast engineered for UGT84A27
expression. UGT84A27 orthologues or a BFP negative control were expressed
from low-copy plasmids in CSY1251. Transformed cells were cultured in
selective medium supplemented with 500 μM PLA, CA or FA for 72 h before
LC–MS/MS analysis of culture supernatant. Data represent the mean of n = 3
biologically independent samples ± s.d. c, Representative LC–MS/MS traces
showing conversion of PLA, CA and FA to cognate glucosides by AbUGT in
CSY1251 cultured as in b for 120 h to enable more complete glucosylation. For
PLA, acid and glucoside were distinguished by different NH 4 + adduct parent
masses as well as different retention times. For CA and FA, rapid fragmentation
necessitated detection of the glucosides based on the lower-retention peaks
produced by their phenylpropanoid fragments. d, Homology model of
AbUGT84A27 constructed based on the crystal structure of Arabidopsis
thaliana salicylate UDP-glucosyltransferase UGT74F2 with bound UDP (PDB:
5V2K). PLA (orange) is shown in the preferred binding pose with UDP-glucose
(pink) based on docking simulations. e, Zoomed view of AbUGT active site with
docked d-PLA and UDP-glucose. Potential mutations identified to improve PLA
selectivity (F130Y, L205F, I292Q) are shown; dashed lines indicate putative
polar/hydrogen bond interactions. f, Heat map of the percentage conversion of
fed phenylpropanoids to glucosides by yeast engineered for expression of
AbUGT mutants. AbUGT wild-type, active site mutants, or a BFP negative
control were expressed from low-copy plasmids in CSY1251. Transformed cells
were cultured in selective media supplemented with 500 μM PLA, CA or FA for
72 h before LC–MS/MS analysis of culture supernatant. Data represent the
mean of n = 3 biologically independent samples ± s.d.