Funke et al. (2006) explored the molecular basis of this inhibition by looking at the structure of EPSP synthase from the C4 strain of Agrobacterium sp. This is the resistant form of the enzyme that has been genetically engineered into Roundup Ready® plants [Roundup Ready® Transgenic Plants].
Note that the structure of glyphosate resembles one of the substrates of the reaction; namely phosphoenolpyruvate (PEP). It was already known that glyphosate binds tightly to the active site of the enzyme and inhibits the reaction by preventing PEP binding. As it turns out, the site for glyphosate binding is exactly the same as the site for PEP binding and this explains the inhibition.
Funke et al. (2006) looked at the C4 EPSP enzyme with and without one of the other substrates: namely, shikimate-3-phosphate (sometimes called shikimate-5-phosphate). The results reveal the precise location of the active site of the enzyme at the base of a cleft between two domains. This form of the enzyme is called class II EPSP synthase because it is distantly related to the class I enzymes in other bacteria and eukaryotes (30% amino acid sequence identity). This is the first paper to examine the structure of a class II enzyme.
As an aside, notice that the enzyme closes up a little bit when the substrate binds—sort of like a Pacman icon. This mechanism of substrate binding is called induced fit and it's proving to be more common than most people realized.
The glyphosate resistant (Roundup Ready®) mutation in C4 EPSP synthase is a substitution of Alanine (A) for Glycine (G) at amino acid position 100. The glyphosate molecule fits nicely into the wild type G100 form of the enzyme (lower image) and it excludes PEP binding completely. Note that glyphosate (green) is in an extended configuration when it is bound. The dotted lines represent non-covalent interactions between the enzyme and the glyphosate molecule. The blue dots are "frozen" water molecules embedded in the active site.
In the mutant form of the enzyme the extra methyl group on alanine is just big enough to cause glyphosate to distort so it can no longer lie in the optimal extended configuration (top image). This means that glyphosate binds much more weakly and doesn't inhibit enzyme activity.
The important point is that the active site can still accommodate phosphoenolpyruvate because it is smaller than glyphosate. What this means is that the overall activity of the enzyme in the absence of glyphosate is unaffected. There are lots of EPSP synthase mutants that don't bind glyphosate but in almost all cases the rate of the reaction is drastically reduced because PEP binding is also weakened. For example, if you mutate the glycine to alanine at the equivalent position in other bacterial or plant enzymes you abolish PEP binding along with glyphosate binding.
What's special about the class II enzymes in general and the Agrobacterium sp. enzyme in particular, is that the amino acids surrounding the PEP binding pocket are positioned just right so that a slight shift can exclude glyphosate without affecting phosphoenolpyruvate. This is mostly due to the positions of the charged amino acid side chains that form weak interactions with the oxygen atoms and the nitrogen of glyphosate; for example, arginines (R) at 128, 357, and 405; lysine (K) at 28; and glutamate (E) at 354.
The results of this study not only shed light on the mechanism of glyphosate resistance but they also help explain the lack of Roundup® resistant plants. Apparently, the class I enzymes in plants have a binding pocket that is difficult to mutate in a way that excludes glyphosate while still allowing PEP binding. Nevertheless, some examples of Roundup® resistant plants are known. I'll describe them tomorrow.
(Funke et al. had to do a bit of sleuthing and reconstruction in order to solve the structure of the C4 EPSP synthase. The C4 strain of Agrobacterium sp. has, naturally enough, not been given out to scientists outside of Monsanto laboratories. So Funke et al. got the amino acid sequence from US Patent 5633435 and reverse engineered the nucleotide sequence of the gene. They synthesized the nucleotide sequence and amplified the fragments by PCR. They then tacked on a promoter and a transcription termination signal and cloned the articfial gene into an E. coli plasmid. The artificially reconstructed protein was then expressed in E. coli, isolated, purified, and crystallized.)
Funke, T., Han, H., Healy-Fried, M,L., Fischer, M., and Schonbrunn, E. (2006) Molecular basis for the herbicide resistance of Roundup Ready crops. Proc. Natl. Acad. Sci. (USA) 103:13010-13015. [PubMed]