If you have an ACS member number, please enter it here so we can link this account to your membership. (optional)

ACS values your privacy. By submitting your information, you are gaining access to C&EN and subscribing to our weekly newsletter. We use the information you provide to make your reading experience better, and we will never sell your data to third party members.


Physical Chemistry

New Perspective on Enzyme Action

Hypothesis that nearly all enzymes covalently speed reactions sparks debate

May 16, 2005 | A version of this story appeared in Volume 83, Issue 20

Zhang (left) and Houk have proposed a controversial new basis for the source of most enzymes' power.
Zhang (left) and Houk have proposed a controversial new basis for the source of most enzymes' power.

Let it not be said that professor of chemistry and biochemistry Kendall N. Houk of the University of California, Los Angeles, lacks chutzpah. Case in point: A proposal that highly proficient enzymes work through covalent bond formation.

The hypothesis has proved to be controversial and has sparked considerable debate. A few in the enzyme mechanistic community find it startling and significant, but others are skeptical.

Postdoc Xiyun Zhang and Houk hypothesize that highly proficient enzymes, which accelerate reactions by factors of more than 1011, bond covalently with reacting substrates, or that their coenzymes do so. To lower the activation barriers of such reactions, "enzymes create new chemistry--new bonds and intermediates--that do not form in the corresponding uncatalyzed processes," Houk says.

The hypothesis does not require that a proficient enzyme form an actual covalent intermediate with a substrate. Zhang and Houk's characterization of covalent bonding includes partially covalent or partly ionic bonding--such as when a metal ion associated with an enzyme coordinates with substrate or when a proton transfers between enzyme and substrate.

The hypothesis is surprising, because enzymes have generally been considered to accelerate reactions mainly via noncovalent binding interactions--such as van der Waals and electrostatic forces, hydrogen bonds (H-bonds), and hydrophobic interactions--and not primarily by full or even partial covalent bonding to their substrates. "Our new view of enzyme catalysis changes the world view of chemists about how enzymes work," Houk says.

The concept originated in a paper by Zhang, Houk, and coworkers Andrew G. Leach and Susanna P. Kim on the binding affinities of host-guest, protein-ligand, and protein-transition-state complexes (Angew. Chem. Int. Ed. 2003, 42, 4872). They noted that, with few exceptions, the maximum energy of noncovalent binding is 15 kcal per mol and that this pales in comparison with the binding energies of enzyme- or coenzyme-transition-state complexes, which average 22 kcal per mol and can be as much as 38 kcal per mol.

THEY CONCLUDED that highly proficient enzymes must therefore use covalency to lower the transition-state free energy to the enormous extent that they do. Noncovalent interactions wouldn't be up to the task. According to the criteria they use to define catalytic strength, almost all known enzymes are highly proficient.

Zhang and Houk wrote a follow-up paper describing the concept in greater detail, which they submitted to Nature. "After a lot of discussion and wrangling with the Nature editors, this article did not make it," Houk says. The paper, "Why Enzymes Are Proficient Catalysts: Beyond the Pauling Paradigm," has now appeared in Accounts of Chemical Research (2005, 38, 379).

The title refers to the renowned chemist Linus Pauling because Zhang and Houk consider their paper to be an advance over Pauling's notion of the way enzymes speed reactions along. In a number of publications and talks, Pauling proposed that enzymes accelerate reactions not by actually taking part in them but by binding noncovalently to their substrates, thus stabilizing reaction transition states and making the reactions proceed more easily. This view has been conventional wisdom and textbook gospel for some time.

The Accounts paper is "a worthy successor" to Pauling's writings, which "we have now updated in a profound way," Houk says. In the paper, Zhang and he "overturn what was previously the bedrock upon which understanding of enzyme catalysis rested. In contrast to the venerable Pauling paradigm for enzyme catalysis, we postulate that most enzymes achieve acceleration by partial covalent bond formation to the reacting substrate in the transition state, involving a change in chemical mechanism from that occurring in the uncatalyzed reaction in aqueous solution," he says. Pauling would have arrived at the same hypothesis "had he known the magnitude of acceleration that enzymes achieve."

Pauling proposed that "the same types of noncovalent molecular recognition involved in antibody-antigen binding also operate in enzymes," Houk says. Zhang and he now propose that "this cannot be the case and that fundamentally different mechanisms operate in antibodies and enzymes."

In their study of enzyme mechanisms, Zhang and Houk surveyed the binding or bonding strength (Ka = association constant) of hundreds of different antibody-antigen (blue), enzyme-inhibitor (yellow), and enzyme-transition-state (red) complexes. The enzyme-transition-state interactions were strongest by far, suggesting that covalent bonding is occurring in many such complexes.
In their study of enzyme mechanisms, Zhang and Houk surveyed the binding or bonding strength (Ka = association constant) of hundreds of different antibody-antigen (blue), enzyme-inhibitor (yellow), and enzyme-transition-state (red) complexes. The enzyme-transition-state interactions were strongest by far, suggesting that covalent bonding is occurring in many such complexes.

The modest acceleration generally achieved by catalytic antibodies is a good example of the low level of catalysis achievable with noncovalent mechanisms, Houk says. The average acceleration by catalytic antibodies is about 103, compared with about 1012 by enzymes, he notes. The new model "explains why chemists have made little or no progress in mimicking enzyme acceleration of reactions by following the Pauling recipe."

Many researchers "are excited about this concept," such as professor of chemistry and biochemistry Richard L. Schowen of the University of Kansas, Houk says. "Some do not like the idea much."

THE ZHANG-HOUK paper "is likely to be a healthy stimulus to the field," Schowen says. Enzyme accelerations of 22 or more orders of magnitude are known, and the paper suggests that an enzyme can derive a maximum catalytic acceleration of only about 11 orders of magnitude from noncovalent sources, such as H-bonds, weak electrostatic interactions, and hydrophobic interactions, he explains. The paper further proposes that the additional 11 or more orders of magnitude achieved by some enzymes derive from localized strong interactions, which Zhang and Houk call "covalent interactions," he notes. These comprise the forming or breaking of covalent bonds, the hydrogen bridges involved in general acid-base catalysis, and coordination to metal centers. "These are original ideas about enzyme catalysis, and I think they deserve exploration and testing," he says.

"The suggestion is going to be controversial because it is novel," Schowen adds. "One reason is that many of us are attached to our own views, but the objections will not arise merely for that reason. A novel suggestion always encounters a multiplicity of objections, and rightly so. Scientists ought not merrily to hop from one bandwagon to another."

Donald G. Truhlar, professor of chemistry, chemical physics, nanoparticle science and engineering, and scientific computation at the University of Minnesota, Minneapolis, agrees that the hypothesis is "a very thorough and important contribution to the field, although I would state it slightly differently." Zhang and Houk say that "the average enzyme proficiency amounts to a transition-state stabilization of 22 kcal per mol but that nonbonding effects--'noncovalent effects,' in their terminology--can account for at most 15 kcal per mol," Truhlar explains. "One could say, as they do, that all the highly proficient enzymes must involve covalency, or they would be maxed out at 15. Or one could say that for a typical enzyme reaction, at least one-third of the transition-state stabilization must come from bonding interactions, and up to two-thirds can come from nonbonding interactions. So both kinds of interactions are important," he says.

"Does the hypothesis change the world view of chemists?" Truhlar asks rhetorically. "It should change the world view of any chemist who believes that catalysis is entirely due to nonbonding interactions. Furthermore, it will probably change the way that even many experts discuss their work, by bringing the role of bonding interactions into the limelight for more quantitative focus. It may be a bit of a stretch for Zhang and Houk to say they have overturned a bedrock of belief, since there are whole books of mechanisms that show the presence of enzyme and coenzyme bonding to substrates, although this bonding [has not been] stressed as an organizing principle for understanding catalytic power" in the same way the UCLA researchers have now done. "That is where their work is new."

OTHERS DO NOT favor the hypothesis. The proposal "is a valuable exercise, but I fail to see the new paradigm," says chemistry professor Stephen J. Benkovic of Pennsylvania State University. "What new principle is stated, or is it merely a relabeling? The Pauling concept, at least as it is now popularly used, already includes covalent catalysis, although it was not explicitly stated in the original definition. The aspects of covalent catalysis invoked by Zhang and Houk have long been widely appreciated."

Chemistry professor Judith P. Klinman of UC Berkeley agrees that the hypothesis amounts to "rediscovering the wheel. One very obvious origin of enzyme catalysis is the use of multiple functional groups on the enzyme that simply will not be available in solution" in the absence of enzyme, she says. "This is what the present authors call covalent interactions. They are basically introducing a new term for factors that have long been recognized."

Professor of chemistry, biochemistry, and biophysics Richard V. Wolfenden of the University of North Carolina, Chapel Hill, agrees that the proposal is unlikely "to change the world view of how enzymes work." Some enzymes have long been known to form covalent bonds to the substrate during catalysis, he says, but some of the strongest enzymes do not, and it's not obvious "that covalent catalysis offers any advantage in the very limited number of reactions in which it is known to occur." Noncovalent interactions such as "H-bonds and electrostatic interactions, sometimes involving metal ions, appear to be entirely responsible for the activity of the most proficient enzymes," he says.

Covalent interactions do contribute to catalytic proficiency, but the new proposal "certainly does not represent a new paradigm for understanding enzyme catalysis," says professor of biochemistry Gordon G. Hammes of Duke University Medical Center. The hypothesis--that the lowering of the free energy of activation achieved by most enzymes (relative to the uncatalyzed reaction) is greater than the free-energy changes associated with typical noncovalent interactions--is invalid, he says. Noncovalent interactions are known to be associated with very large free-energy changes, he notes. "Covalent interactions also contribute to enzyme catalysis," but this was already well-known, he adds.

According to professor of chemistry and biochemistry Arieh Warshel of the University of Southern California, "There is nothing new in the proposal except the risk that it will confuse the scientific community." Zhang and Houk state that it is difficult to rationalize the catalytic power of enzymes on the basis of noncovalent interactions alone, Warshel says, but they ignore calculations by his group and others that such catalytic effects are completely accounted for by the effects of optimized electrostatic preorganization, a type of noncovalent interaction that doesn't occur in catalytic-antibody and other nonenzymatic binding. Warshel says he and his coworkers can easily attribute substantial catalytic effects to electrostatic interactions of enzymes with substrates, but the UCLA group "simply dismisses such quantitative demonstrations and has not attempted to estimate the preorganization effects."

Professor of chemistry and biochemistry Thomas C. Bruice of UC Santa Barbara comments that "the efficiency of an enzyme, by convention, relates to the ratio of the rate constant for the enzyme reaction divided by the rate constant for the same reaction in water." Water is a poor nucleophile, so enzymes that bond covalently to substrates "have amazingly greater efficiencies," but the formation of such bonds does not improve enzyme rate constants per se, so the proposal has no valid basis, he says.

CHEMISTRY PROFESSOR Jiali Gao, Truhlar's colleague at the University of Minnesota, says the Zhang-Houk study "surely is newsworthy, and its comprehensive analysis of host-guest and enzyme data is very impressive." The hypothesis "can be useful to interpret a number of enzyme reactions, but covalency does not seem to be the only means by which enzymes can achieve remarkable rate accelerations." For example, orotidine decarboxylase works covalently, according to Zhang and Houk, but Gao and coworkers have previously found that it attains its high catalytic power noncovalently.

"Enzyme chemists know there are covalent interactions," Truhlar says, "but too often they have treated these as mechanistic complications and focused on cases where this does not occur. Some of us were unhappy with this situation, but we did not enunciate our thoughts in a way that changed the focus in a large number of minds. The Zhang-Houk analysis shows clearly that ignoring these mechanistic issues may cause one to miss one of the keys to quantitative understanding of the magnitude of transition-state stabilization. This is why it is an important step forward."

Acknowledging a wide divergence of opinion on the hypothesis, Schowen says: "There is nothing wrong and much that is healthy in such a situation. In the end, there will continue to be a fairly large compendium of different factors required to explain enzyme catalysis. And it seems to me quite possible that the Zhang-Houk hypothesis will be among them."


This article has been sent to the following recipient:

Chemistry matters. Join us to get the news you need.