# Planar transition-state of proline racemization facilitated by proline racemase

The racemization of proline can be facilitated by the enzyme proline racemase.

Since the active sites of enzymes complement the conformation of the transition-states of the molecules they act upon, the most effective enzyme-inhibitors often resemble these transition states.

This enzyme can be inhibited by the transition-state (TS) analog pyrrole 2-carboxylic acid.

The strength of the bond between the TS-analog and the enzyme, is approximately 160 times stronger than the proline-enzyme bond.

When I was first reading about this, I did not believe it because of the negative charge on the alpha-carbon. I decided to draw the resonance structure of the transition state, to find a double bond between the alpha and carboxyl carbons.

My guess is that this resonance structure would make the transition state relatively planar (as is pyrrole 2-carboxylic acid).

But my question is, how "planar" is the transition state, and how might one determine the exact bond angles at the alpha-carbon?

TLDR: How "planar" is the transition state of proline isomerization, and how might one determine the exact bond angles at the alpha-carbon?

(P.S. Sorry for the bad scaling of my ChemDraw pictures, I don't know how to fix it.)

• Looks like the picture was transparent. I made it opaque and slightly shrinked it. To shrink it, add either m (for medium) or s (for small) at the end of the file name. For example, if it's https://i.stack.imgur.com/zCuY3.png just make it https://i.stack.imgur.com/zCuY3m.png instead. – Pritt Balagopal May 25 '17 at 5:51
• @PrittBalagopal Use t (as in tiny) instead of s. Because using s will crop the image to a square (even if your original image was a rectangle) while t would leave the proportions untouched. – Berry Holmes May 25 '17 at 6:15
• Unfortunately the only way to be confident that you know what the bond angles are is to have an x-ray structure. As general rule s(a) The effect of the negative charge will almost certainly be negated to a greater or lesser extent by a nearby amino acid and (b) the protein way well also distort the shape of the substrate while it is in the active site. Sorry to be vague but anything more precise would be pure guess work. – porphyrin May 25 '17 at 6:29
• @porphyrin Since the transition-state does not exist for a long period of time, crystallising an ES-complex would be impossible, correct? But, I guess, given evidence of the fact that pyrrole 2-carboxylic acid is able to permanently bind the enzyme would suggest that the transition state of proline racemization probably is (almost) planar. – Bob May 25 '17 at 6:47
• (Note also that the anion you have drawn is probably an intermediate, not a transition state.) – orthocresol May 25 '17 at 12:27

As others have written in the comments, there are large parts of your question that can't be answered by experiment. However, it is possible to come up with a mechanistic hypothesis (and even better, also an alternative hypothesis) and gather evidence that rejects one hypothesis and is consistent with another.

Structural data supports the mechanism you sketched for the PLP-independent racemases. The structure of proline racemase in complex with the transition-state analog (PDB ID 1W61) shows fairly symmetric binding interactions between the enzyme active site and the ligand:

In the figure (a screen grab from https://www.rcsb.org/3d-view/1W61?preset=ligandInteraction&sele=PYC), the inhibitor is labeled [PYC]700, and it has similar interactions on the left and the right (for example two main chain hydrogen bond interactions with the carboxylic acid group each, and one cysteine residue each in the vicinity of the inhibitor.

The binding pocket is fairly symmetric (view from the "backside" with respect to previous image):

In the paper reporting the structure, the authors point out "two Cys residues optimally located to perform acid/base catalysis through a carbanion stabilization mechanism". They also suggest that the sulfur atoms sandwich proline when it binds, "squeezing" it into a conformation closer to planar than it would be in solution.

However, as the proline-bound structure of a different, putative racemase shows, the active site does accommate proline with a non-planar alpha carbon:

When I was first reading about this, I did not believe it because of the negative charge on the alpha-carbon. I decided to draw the resonance structure of the transition state, to find a double bond between the alpha and carboxyl carbons.

As the enzyme is stepping through catalysis, the enzyme active site binds to the molecule in different charge states, and itself changes charges as the cysteine residues accept and donate protons. In the cited paper, it is left open whether the reaction proceed in a concerted manner (one proton abstracted on one side while one is donated on the other) or via a carbanion intermediate. They do point out the the pKa of the alpha carbon is "in the range 21–32", arguing against a long-lived intermediate.

But my question is, how "planar" is the transition state, and how might one determine the exact bond angles at the alpha-carbon?

As you can read from the excellent comments, there is no good experiment to get at the exact structure of the transition state, and it is not a given that the transition state is planar. However, if the carboxylate group is on one side of the ring before the reaction and on the other side of the ring after the reaction (and stays attached to the alpha carbon throughout), at one point on the reaction coordinate the structure around the alpha carbon has to be planar (or some version of trigonal bipyramidal for a concerted mechanism). Therefore, the active site has to be able to accommodate that structure.

It would be possible to study this system using computational methods, to at least get a sense of the energy landscape in the absence of enzyme. The calculation with enzyme is complicated by the fact that the enzyme undergoes fairly large conformational change upon ligand binding.