Tuesday, December 09, 2014

On the specificity of enzymes

Most biochemistry students are taught that enzymes are highly specific. It's certainly true that the stereospecificity of some enzymes is extraordinary but is it true in general? Here's one of the exam questions that the students in my course had to answer ....
All three of the enzymes (trypsin, alcohol oxidase, β-galactosidase) that you assayed in the past three months are active with several different substrates substrates. Is this behaviour typical or are most enzymes highly specific? Aminoacyl tRNA synthetases are the classic examples of enzymes that are highly specific. Why? Do aminoacyl-tRNA synthetases ever make mistakes?


  1. The Big Larry decided he is going to remove my comments... Why...?
    Because he can't handle the truth.. he never did... and he never will... congratulations larry... you won...

    1. I am not (yet) deleting all of your comments, only the obnoxious ones where you use inappropriate language.

  2. Yeah, one of the big problems with the ID people is they make some assumption based on textbook learning or whatever (or, some quote from the literature from someone who is overgeneralizing, or who really isn't an expert on that specific point, or someone who is clueless -- or, they quote some news article which has garbled or oversimplified something, etc.), don't realize it's an oversimplification or mistaken, then build an antievolution argument on it, and then hold to it come hell or high water.

    This is precisely what Ann Gauger is doing with her assumption that all / almost all enzymes are highly specific. The reality is that enzymes with varying degrees of specify, promiscuous enzymes, etc., are a dime a dozen, out there in the diversity of real biology (which can be very different from the small subset of study enzymes and study organisms someone may have been exposed to in their particular graduate training).

    You also see this kind of thing with regular ID claims/assumptions along the lines that all functions proteins must have well-defined folds, that point mutations never change protein structures, that only one structure can perform a particular function, etc. None of this is right, although these are all the kinds of things you might mistakenly assume from an education in some narrow area of biochemistry without any education or self-education in the diversity of what's out there.

    Ann Gauger does similar things in numerous other areas where she has only the most rudimentary understanding of the field -- e.g., phylogenetics, ancestral sequence estimation, studies of adaptational pathways. In each case, explaining what is wrong in her assumptions pretty much would require writing her a tutorial of the particular field, explaining everything from the ground up.

    1. NickM,

      Can you create the cell membrane... without hiding under numerous NickMames...?

    2. I think readers will find that in order to 'prove' Ann wrong, Nick needs to make the exceptions the rule.

      Easy to prove that notion wrong. Nick does not need to write a book for Ann and lurkers to understand.

      Just provide a couple of 'obvious' examples of each below:

      1. Proteins with poorly defined folds.
      2. Point mutations that change protein structure.
      3. Structures that perform more than one function.

      Next, show the percentages of each to show how ubiquitous they are rendering Ann's position weak.

      1. What percent of proteins have poorly defined folds?
      2. What percent of proteins have their structures changed by point mutations?
      3. What percent of structures have more than one function.

      I know it will take Nick only a few minutes to rattle of the above as he has a deep grasp of the subject.

      No nuancing/hedging required for the above.

    3. Hey Quest, have you recreated genesis 30:39 yet?

    4. Steve, where were you in the other thread where Axe and Gauger's work was heavily criticized. Mind taking a stab at the questions I posed there?

    5. Steve -

      I think readers will find that in order to 'prove' Ann wrong, Nick needs to make the exceptions the rule

      Nick may respond if he likes, but why does he have to do this work? Ann's position is weak to start with; its weakness does not depend on the representativeness of the proteins chosen.

      All modern proteins could be absolutely specific, with no possibility of point-mutational conversion of one to another through functional intermediates. This 'fact', accepted arguendo, would be silent on the possibility of their evolution from a common ancestor.

  3. Funny... Ann Gauger shows up and evolutionists going ape...

    1. What is wrong with the criticisms leveled at her work? Can you explain it?

    2. Even funnier ... completely uncritical acceptance of her work from the Creationist community .... she must have given evolutionistas one in the eye. Because ... well, just because, OK?

  4. Getting back to Larry's original question:

    Do aminoacyl-tRNA synthetases ever make mistakes?

    Not only do errors occur - (I sorta knew this all along)


    ...these errors can be edited! (Something I did not know!)


    Thank you - What a great question! I wish I could sit in your class!

    Out of curiosity - do you ever mention amber, ochre and opal suppressor tRNA genes ?

  5. If I were in charge of propaganda for the DI my argument would be this:

    'Yes enzymes are promiscuous, but the alternate substates they can use have only trivial differences: one 4C sugar versus another 4C sugar. This doenst come close to explaining the huge diversity in enzymes we see in the biosphere'

    My counterargument would be this

    1. There are a few cases where an enyzme can have a very different substrate ( though I cant think of any examples) and many enzymes are assayed using subtrates that difer substantially from their normal substrate
    2. Analysis of the teritiary structure of many enzymes suggest that there have been many transitions in catalysis. Axe and Gauger work to attempt to retrace the evolution of an enzyme could have never worked because they ignored a key finding from Thornton's lab: there are 'molecular ratchets' which would prevent selection from evolving previously existing enzymes.
    3. Evidence from many areas: ribozyme evolution, abzymes, de novo designed proteins, IDPs suggest that the big complex streamlined protiens we see now are not necessary for workable catalysis
    4. Ignoring the origin of proteins, the evidence for the recent evolution of enzymes is enough to suggest that a great deal of change can occur in living systems by natural mechanisms. No designer is required.


  6. Do people with a predilection for teleological thinking tend to ascribe extreme specificity to enzymes?

    There does not seem to be a good empirical or theoretical reason to believe that enzymes have almost exact specificity for a given substrate in order to have satisfactory functionality. All an enzyme needs to do is to catalyze a biologically useful reaction for an organism, and there is no obvious reason why this necessarily requires a high degree of substrate specificity or the reason to exclude other possible substrates from its active site, other than the usual substrate. The fitness landscape of most metabolic enzymes is not constrained by a need for extreme substrate specificity nor a substrate specific enzyme should be considered optimal. A notable exception where fidelity is at a premium are the enzymes involved in DNA replication and transcription and the enzymes involved in the translation (such as the aminoacyl tRNA synthetases).

    It is simple thermodynamics and chemical kinetics to understand that most enzymes are not perfectly specific. For a given enzyme, any substrate, especially one with structurally and chemically similar to its typical substrate, has an associated affinity with the enzyme. (The free energy value is inversely related to the reaction rate as expressed in the Arrhenius equation.) For example, polymerase enzymes do not always add the correct nucleotide to a nascent strand to form a Watson-Crick pair with the template as it occasionally incorporate the incorrect nucleotide to the nascent strand. This happens because the transition state of the a reaction catalyzed by a polymerase that incorporates the incorrect nucleotide has some free energy value that is low enough to allow it to occur occasionally. Of course, the free energy of the transition state for the "correct" is reaction is lower than that for the "incorrect" reaction. The relative difference of free energy between the transition states of a new Watson-Crick pair versus one that has non Watson-Crick pairing is what determines the degree of fidelity of polymerases. (This fidelity is much higher than what would be expected between the differences in free energy between Watson-Crick and non-Watson-Crick pairing). Furthermore, many polymerase enzymes have exonucleolytic proofreading that excises, preferentially but not exclusively, a non-Watson-Crick pairs as a non-Watson-Crick pair from the nascent strand is more likely to go into the exonucleolytic site. This step grants another kinetic layer that increase the fidelity of the polymerase.

    Again, it does not seem obvious that the degree of promiscuity would affect the fitness of an enzyme, since the influence of natural selection would only impact an enzyme if affects the catalysis of a physiologically useful reaction (of course, in a given environment context.) But there little biological context with the notion of promiscuity, especially if it the enzyme is promiscuous regarding substrates that would rarely encounter in its environment. In such a case, this affirms the contention that promiscuity, or the lack of it, is not a property influenced by selection. If an enzyme encounters a novel substrate that is introduced to its environment, the ability of the enzyme to catalyze a useful reaction with the novel substrate cannot be determined a priori. Perhaps, some variants of that enzyme within a given population would be able to catalyze a physiologically useful reaction and some would not, maybe because one variant has an amino acid residue that would sterically exclude the novel substrate. Such variants would not affect fitness before the introduction of the novel substrate but would nevertheless be a source of genetic variance and potentially influence fitness-related differences in phenotype (which is the primary ingredient of natural selection).