The following email was sent to a number of researchers active in enzyme function on 28 June 2010. I apologise to those not included on the mailing list - I kept it short to avoid spam filters. The purpose was to generate discussion.
"Dear Colleagues,
I was interested to read the correspondence in PNAS in April between Dr Warshel's group and Dr Karplus’ group, as well as recent papers by Drs Klinman (2009), Sutcliffe, and Scrutton (2008 and 2010). I notice that the various lines of thought are actually converging while the individual scientists focus on the differences.
Insofar as Dr Warshel's "preorganization" is that part of the catalytic process outside of the simulated chemical mechanism, it is just the same as the phenomenon that I have described as "constraint" (Vanselow, D. 2002 Biophys J). Because this "preorganization" or "constraint" involves the whole protein and the surrounding solvent, it has in the past been too big for MM or QM or MD simulation for any length of time and Dr Warshel's group has not pursued it. Constraint is, however, amenable to study by a continuum model of time-averaged forces (Vanselow, D. above ref) that is easy to understand and welcomed by a number of physicists.
As a result if this experience I offer the following comments:
1. It is better to avoid any idea that thermal energy can be focused or funneled to the active site, e.g. through large motions of loops, lids etc. One could inadvertently come up with a mechanism for heat to flow from cold (the environment) to hot (the excited active site) in a regular or probable way and hence produce a perpetual motion machine. They don't occur in biology.
2. We should pay attention to forces. They are usually focused to some extent by transmission through a solid. Because they are vector quantities they place demands on the architecture of a catalytic system that lead to very fruitful comparisons with supposedly known structures (http://nativeproteins.blogspot.com ). Dr Hammes (JBC vol 283, p22338, 2008) is right that an enzyme must be a macromolecule.
3. We should accept the possibility that hydrogen nucleus tunneling can be enhanced during catalysis. Dr Klinman (2009) says that deuterium tunneling does occur and requires atoms to be much closer than the sum of their van der Waals radii. I have found that measured compressibilities of solids and liquids indicate that a pressure of several GPa is required to achieve such close approach. This is within the range of compressive stress in the active site during catalysis, according to a semi-quantitative treatment of constraint (Vanselow, D 2002 above). Hay, Johannissen, Sutcliffe and Scrutton (2010) are right to point out that the organism will exploit any catalytic effect accessible to it during barrier compression.
4. It is a good idea to restrict models to equilibrium conditions. This imposes extra discipline on the imagination. All intermediate states can then be seen to have real properties and constant populations. All transitions from one state to the next must proceed equally well in both directions. After catalysis, the enzyme must be left in exactly the same condition as before catalysis. We should avoid ideas of momentum carrying the process forward; at equilibrium it must be a random walk.
5. We must be prepared for contradictions between the architecture demanded by physics and the structures observed in protein crystals (or by other methods), especially the quaternary structure. Bear in mind that force and energy are very different functions. Subunits in a protein complex can have low energy of interaction and yet exert strong compressive forces over a very short range of motion. Therefore we can expect cases where the active protein complex is not easily observed because of the small energy of interaction of its subunits. Also consider that the degree of association of subunits seems to be not well documented these days. The Protein Data Bank asks authors to specify the “biological unit”, which they do without any documentation. The PDB data is then quoted by others and is treated as fact. Examples I have come across are some neuraminidases http://nativeproteins.blogspot.com/#neu , galactose oxidase http://nativeproteins.blogspot.com/#galox and adenylate kinase.
I liked Dr Klinman’s (2009) philosophical re-evaluation of Pauling’s description of enzyme catalysis. Of course one cannot argue with the Arrhenius treatment of rates because it is a purely mathematical transformation of a definition of rate. However, when Pauling attempted to put the activation energy concept of catalysis into words he inadvertently narrowed the concept, especially with the phrase “forces of attraction”. There is also a problem with the definition of “substrate”. Is the substrate a whole molecule, as implied by Pauling, or is it only the group of atoms reacting? Taking ATP, for example, can binding to the adenine part have any role in catalyzing transfer of the terminal phosphate? What Pauling should have said is that catalysis requires that the interaction between the enzyme and the group of atoms undergoing reaction is such that the energy of interaction is temporarily added to the reacting group. Pauling’s words were much more elegant but too limiting. My own mechanical model of catalysis (see the appendix to my 2002 paper) is more general and abstract. It encompasses Pauling’s idea but also allows for enzyme-substrate interactions other than “attraction”. It also avoids defining the substrate(s). In brief, it says that catalysis results from part of the system doing reversible work on the reactants/products/active site as they pass through transition states.
I was interested in the “pathways of energetic connectivity” described by Lockless and Ranganathan (1999). In particular, pathways of aromatic side chains in contact could transmit either whole charges or electrical potential, provided they are held rigidly together for the appropriate period. Such pathways can be seen in reconstructed galactose oxidase (whole charge transmission) http://nativeproteins.blogspot.com/#galox and in reconstructed hemoglobin (transmission of electrical potential) http://nativeproteins.blogspot.com/#hb . The benefits of using knowledge of protein function to reinterpret protein structure are enormous! I have said enough for one email but I would be delighted to engage in further discussion and, of course, receive assistance.
Yours sincerely
Don Vanselow PhD"
