4-coumarate:CoA ligase (4CL), the last enzyme of the general phenylpropanoi
d pathway, provides precursors for the biosynthesis of a large variety of p
lant natural products. 4 CL catalyzes the formation of CoA thiol esters of
4-cournarate and other hydroxycinnamates in a two step reaction involving t
he formation of an adenylate intermediate. 4 CL shares conserved peptide mo
tifs with diverse adenylate-forming enzymes such as firefly luciferases, no
n-ribosomal peptide synthetases, and acyl:CoA synthetases. Amino acid resid
ues involved in 4 CL catalytic activities have been identified, but domains
involved in determining substrate specificity remain unknown. To address t
his question, we took advantage of the difference in substrate usage betwee
n the Arabidopsis thaliana 4 CL isoforms At4CL1 and At4CL2. While both enzy
mes convert 4-coumarate, only At4CL1 is also capable of converting ferulate
. Employing a domain swapping approach, we identified two adjacent domains
involved in substrate recognition. Both substrate binding domain I (sbd I)
and sbd II of At4CL1 alone were sufficient to confer ferulate utilization a
bility upon chimeric proteins otherwise consisting of At4CL2 sequences. In
contrast, sbd I and sbd II of At4CL2 together were required to abolish feru
late utilization in the context of At4CL1. Sbd I corresponds to a region pr
eviously identified as the substrate binding domain of the adenylation subu
nit of bacterial peptide synthetases, while sbd II centers on a conserved d
omain of so far unknown function in adenylate-forming enzymes (GEI/LxIxG).
At4CL1 and At4CL2 differ in nine amino acids within sbd I and four within s
bd II, suggesting that these play roles in substrate recognition.