Structural Biology from Zero · 5: Enzymes — Structure as Catalysis

There is a reaction in your cells, catalyzed by an enzyme called orotidine-5′-phosphate decarboxylase, whose uncatalyzed version is so slow that a single molecule would, on average, wait about 78 million years to react. The enzyme performs it in milliseconds. That is a rate enhancement of roughly $10^{17}$ — a hundred quadrillion — achieved with no added heat, no harsh reagents, at body temperature and neutral pH, using nothing but a folded chain of amino acids. Enzymes are the most spectacular catalysts known, routinely outperforming the best industrial chemistry by many orders of magnitude, and they do it through structure: a precise three-dimensional arrangement of a handful of chemical groups. This article is about how a fold becomes a rate, and it is where the abstract architecture of Part 4 turns into function.

Let me start with the one idea that makes all of enzyme catalysis intelligible, because without it the mechanisms are just a zoo.

Catalysis and the transition state

Every reaction passes over an energy barrier. Reactants must distort — stretch bonds, gather charge, contort geometry — into a high-energy, fleeting arrangement called the transition state before they can slide down to products. The height of that barrier, the activation free energy $\Delta G^{\ddagger}$ , controls the rate. Transition-state theory makes the relationship quantitative:

$k = \frac{k_{\mathrm B} T}{h}\,\exp\!\left(-\frac{\Delta G^{\ddagger}}{RT}\right),$

where $k_{\mathrm B}$ is Boltzmann's constant, $h$ is Planck's constant, and the prefactor $k_{\mathrm B}T/h \approx 6\times10^{12}\ \mathrm{s^{-1}}$ at body temperature sets the fundamental attempt frequency. The barrier sits in an exponential, which is why catalysis is so leveraged: the rate depends exponentially on $\Delta G^{\ddagger}$ , so a small reduction in the barrier is an enormous increase in rate.

How small is "small"? A catalyst speeds a reaction by the ratio of the two rate constants, which depends only on how much it lowers the barrier:

$\frac{k_{\text{cat}}}{k_{\text{uncat}}} = \exp\!\left(\frac{\Delta\Delta G^{\ddagger}}{RT}\right).$

At $T = 310\ \mathrm{K}$ , $RT \approx 0.62\ \mathrm{kcal/mol}$ , and since $\ln 10 \approx 2.303$ , each factor of ten in rate costs just $2.303\,RT \approx 1.4\ \mathrm{kcal/mol}$ of barrier reduction. So the $10^{17}$ enhancement above corresponds to lowering the barrier by only

$\Delta\Delta G^{\ddagger} = 17 \times 1.4\ \mathrm{kcal/mol} \approx 24\ \mathrm{kcal/mol}.$

The estimate is about $24\ \mathrm{kcal/mol}$ . An enzyme can produce an enormous rate enhancement by lowering the activation barrier by this comparatively modest amount, using several coupled interactions rather than a simple sum of independent hydrogen bonds.

Linus Pauling emphasized in 1948 that catalysis requires preferential stabilization of the transition state relative to the substrate ground state. Tight ground-state binding alone can deepen the substrate free-energy well and need not lower the activation barrier. An active site that complements the transition state's geometry and charge distribution can use binding energy to reduce $\Delta G^{\ddagger}$ . Molecules designed to mimic important transition-state features can therefore be potent inhibitors, although their affinity depends on how accurately they reproduce the relevant interactions.

The toolkit: how active sites lower barriers

Enzymes stabilize transition states using a small, recurring set of strategies, usually several at once. Each is a way to supply part of that two-dozen kcal/mol.

Proximity and orientation. A bimolecular reaction in solution is slow partly because the two reactants must find each other and collide in exactly the right orientation. An enzyme binds both substrates in its active site, adjacent and correctly aligned, converting a costly search into a near-certainty. In free-energy terms it pre-pays the entropic cost of bringing and orienting the reactants, so that cost is not charged again at the transition state.
General acid–base catalysis. Many reactions need a proton removed or delivered at a precise moment. Side chains that can gain or lose a proton near neutral pH — histidine above all, with its $\mathrm{p}K_{\mathrm a} \approx 6$ — sit in the active site poised to shuttle protons to and from the substrate, stabilizing developing charge in the transition state.
Covalent catalysis. The enzyme forms a transient covalent bond to the substrate, routing the reaction through a different, lower set of barriers than the uncatalyzed path. A nucleophilic side chain (a serine, cysteine, or lysine) attacks the substrate to make a covalent intermediate that is later resolved.
Metal-ion catalysis. A bound metal ion (zinc, magnesium, iron) can polarize bonds, stabilize negative charge, or generate a potent nucleophile — a zinc-bound hydroxide in carbonic anhydrase, for instance, is a far better nucleophile than free water.
Electrostatic preorganization. Arieh Warshel and others have argued that this is a major contribution for many enzymes. In water, the solvent must reorganize to stabilize a developing charge. An enzyme active site can be preorganized, with dipoles and charged groups already oriented to stabilize the transition state's charge distribution. The importance of this contribution depends on the enzyme and reaction mechanism.

Notice that most of these deliberately place polar and charged groups in a specific geometry — the exact opposite of the "bury the hydrophobic, expose the polar" rule that governs the rest of the fold (Part 4). The active site is a purpose-built local exception, and maintaining that reactive, often charged pocket is part of what folding pays for.

The active site and specificity

An enzyme's catalytic machinery occupies a small active site — usually a cleft or pocket, made of residues that may be far apart in sequence but are brought together by the fold. Two things happen there: catalysis (above) and recognition. The active site's shape and chemistry select which substrate can bind, and it can be exquisitely picky.

Two classic pictures describe the fit. Emil Fischer's 1894 lock and key imagined a rigid active site precisely complementary to the substrate. Daniel Koshland's 1958 induced fit corrected this: the enzyme is not rigid but adjusts its conformation as the substrate binds, closing around it like a hand around a ball. Induced fit is not a detail — it is often part of the catalysis, because the conformational change can be what aligns the catalytic groups or excludes water from the site. Specificity itself is frequently read off a specificity pocket: in the digestive protease trypsin, a buried aspartate at the bottom of the substrate pocket attracts the positively charged side chains of lysine and arginine, so trypsin cuts after those residues; its close relative chymotrypsin has a hydrophobic pocket instead and so cuts after bulky aromatic residues. Same fold, one pocket residue swapped, entirely different specificity — a compact illustration of how structure encodes function.

A worked example: the serine proteases

The best way to make all of this concrete is to walk through one mechanism completely, and the serine proteases — chymotrypsin, trypsin, elastase, and thousands of relatives — are the ideal choice, because they combine three of the strategies above into one of the most-studied active sites in biology. Their job is to hydrolyze a peptide bond, which is a genuinely hard reaction: the peptide bond is stabilized by the resonance we met in Part 2, and left alone it can survive for hundreds of years in water.

At the heart of the enzyme is the catalytic triad: three residues — a serine, a histidine, and an aspartate — held by the fold in an exact hydrogen-bonded arrangement. In chymotrypsin these are Ser195, His57, and Asp102. They are distant in sequence but the fold clamps them together, and the geometry does the work:

Activation of the nucleophile. On its own, a serine hydroxyl is a poor nucleophile. But His57, positioned as a general base, pulls the proton off the Ser195 $\mathrm{O\gamma}$ , turning it into a potent alkoxide nucleophile. Asp102, buried behind the histidine, orients it and electrostatically stabilizes the resulting positive charge on the protonated histidine — the "charge relay" identified by Blow and colleagues in 1969.
Nucleophilic attack and the oxyanion hole. The activated serine oxygen attacks the carbonyl carbon of the substrate's scissile bond, forming a tetrahedral intermediate in which the former carbonyl oxygen now carries a full negative charge. This intermediate closely resembles the reaction's transition state, and here the enzyme's preorganization shows: two backbone $\mathrm{N-H}$ groups form a pocket called the oxyanion hole, precisely placed to hydrogen-bond to that developing negative oxygen and stabilize it. This is transition-state stabilization made visible.
Collapse to the acyl-enzyme. The tetrahedral intermediate collapses, breaking the peptide bond; His57 donates its proton to the departing amine, which leaves as the first product. What remains is an acyl-enzyme, with the other half of the substrate covalently attached to Ser195 — the covalent-catalysis step.
Deacylation. A water molecule enters, His57 deprotonates it (acting as a base again), and the resulting hydroxide attacks the acyl-enzyme through a second tetrahedral intermediate, again stabilized by the oxyanion hole. This releases the second product and regenerates the free enzyme.

Every strategy from the toolkit appears: general acid–base (His57), covalent catalysis (the acyl-enzyme), proximity and orientation (the triad geometry), and electrostatic transition-state stabilization (the oxyanion hole). The viewer below lets you look at a real serine protease active site and see the triad's two hydrogen bonds measured in ångströms — the physical hardware behind this mechanism.

Interactive. Tab 1: the catalytic triad of α-chymotrypsin (PDB 1ACB), with the Ser–His and His–Asp hydrogen bonds computed live from the coordinates, plus the bound inhibitor standing in for a substrate. Tab 2: the Michaelis–Menten rate law — drag the substrate concentration and tune $K_M$ and $k_{\text{cat}}$ to see how $V_{\max}$ , $K_M$ , and catalytic efficiency emerge. Opens full-screen in a new tab.

The rate law: Michaelis–Menten kinetics

Mechanism tells you how; kinetics tells you how fast, and the kinetics of nearly every enzyme are captured by one equation worth deriving from scratch, because the derivation is short and the resulting parameters are the language every biochemist speaks. Model the simplest case: an enzyme $E$ binds substrate $S$ reversibly to form a complex $ES$ , which then turns over to product $P$ .

$E + S \underset{k_{-1}}{\overset{k_{1}}{\rightleftharpoons}} ES \overset{k_{2}}{\longrightarrow} E + P.$

The reaction velocity is the rate of product formation, $v = k_2[ES]$ . To find $[ES]$ , use the steady-state approximation of Briggs and Haldane (1925): shortly after mixing, the intermediate $[ES]$ is consumed as fast as it forms, so its concentration is effectively constant,

$\frac{d[ES]}{dt} = k_1[E][S] - (k_{-1} + k_2)[ES] = 0.$

The free enzyme obeys conservation, $[E] = [E]_0 - [ES]$ , where $[E]_0$ is the total enzyme. Substituting and solving for $[ES]$ :

$k_1\big([E]_0 - [ES]\big)[S] = (k_{-1}+k_2)[ES].$

Collect the $[ES]$ terms and define the Michaelis constant $K_M \equiv (k_{-1}+k_2)/k_1$ :

$[ES] = \frac{[E]_0[S]}{K_M + [S]}.$

Multiply by $k_2$ to get the velocity, and write $V_{\max} \equiv k_2[E]_0$ :

$\boxed{\,v = \frac{V_{\max}[S]}{K_M + [S]}\,}$

This is the Michaelis–Menten equation, and its two parameters have clean physical meanings you can read straight off the curve:

$V_{\max} = k_{\text{cat}}[E]_0$ is the maximum velocity, reached when substrate saturates the enzyme ( $[S] \gg K_M$ , so every enzyme molecule is working). The turnover number $k_{\text{cat}}$ (here equal to $k_2$ ) is how many substrate molecules one enzyme converts per second at saturation.
$K_M$ is the substrate concentration at which $v = V_{\max}/2$ — substitute $[S] = K_M$ into the equation and the fraction becomes $1/2$ . It is a rough measure of how tightly the enzyme grabs its substrate: a low $K_M$ means half-maximal rate at low substrate, i.e. high apparent affinity.

Consider the limiting cases. When $[S] \ll K_M$ , the equation reduces to $v \approx (V_{\max}/K_M)[S] = (k_{\text{cat}}/K_M)[E]_0[S]$ , which is first-order in both enzyme and substrate. Cellular enzymes occupy a range of saturation regimes depending on their substrates, regulation, and compartment. When $[S] \gg K_M$ , $v \to V_{\max}$ and becomes independent of substrate concentration.

That low-substrate regime hands us the single most useful number in enzymology. The ratio

$\frac{k_{\text{cat}}}{K_M}$

is the specificity constant or catalytic efficiency — the effective second-order rate constant for substrate meeting free enzyme and going on to product. It has an absolute ceiling: an enzyme cannot process substrate faster than the substrate can diffuse into the active site, and that diffusion limit is about $10^{8}\text{–}10^{9}\ \mathrm{M^{-1}\,s^{-1}}$ . A handful of enzymes have evolved right up against it — they are called catalytically perfect, because every encounter with a substrate molecule leads to reaction. One of them is triosephosphate isomerase, the very α/β barrel we used to illustrate the TIM fold in Part 4: its structure is not just a pretty architecture but a machine tuned to the physical speed limit.

Cofactors: chemistry the twenty can't do

The twenty amino acids are a limited chemical palette — they are good at acid–base chemistry and nucleophilic attack, but they cannot, for example, carry out most redox reactions or transfer certain groups on their own. Enzymes extend their reach with cofactors: metal ions (the zinc of carbonic anhydrase, the iron of the heme we saw in hemoglobin) and small organic coenzymes, many of them derived from vitamins — NAD⁺ and FAD for shuttling electrons, coenzyme A for carrying acyl groups, pyridoxal phosphate for amino-acid chemistry. The protein provides the specificity, the positioning, and the transition-state stabilization; the cofactor provides a reactive center the protein's own side chains could never supply. It is a recurring division of labor worth remembering: the fold organizes, the cofactor does the exotic chemistry.

Where this leaves us

An enzyme uses its fold to position active-site residues and cofactors in a geometry that favors a particular reaction pathway and stabilizes high-energy states along it. The resulting barrier reduction can produce a large rate enhancement through proximity, acid–base and covalent chemistry, metal ions, and electrostatic organization. The Michaelis–Menten framework summarizes the observed kinetics through $k_{\text{cat}}$ and $K_M$ , while their ratio approaches the diffusion limit for some enzymes. The structure supplies the chemical environment, dynamics, and geometric organization required for catalysis.

We have now seen structure build machines (Part 4) and perform chemistry (Part 5). In Part 6 we turn to the structural basis of recognition pushed to its evolutionary extreme — antibodies, where a single fold, decorated with a few hypervariable loops, is tuned by evolution-in-fast-forward to bind essentially any molecule the immune system has never seen before.