10.1 Aspartate Transcarbamoylase Is Allosterically Inhibited by the End Product of Its Pathway

Aspartate transcarbamoylase catalyzes the first step in the biosynthesis of pyrimidines: the condensation of aspartate and carbamoyl phosphate to form N-carbamoylaspartate and orthophosphate (Figure 10.1). This reaction is the committed step in the pathway, which consists of 10 reactions, that will ultimately yield the pyrimidine nucleotides uridine triphosphate (UTP) and cytidine triphosphate (CTP). How is this enzyme regulated to generate precisely the amount of pyrimidines needed by the cell?

Figure 10.1: ATCase reaction. Aspartate transcarbamoylase catalyzes the committed step, the condensation of aspartate and carbamoyl phosphate to form N-carbamoylaspartate, in pyrimidine synthesis.

287

     ATCase is inhibited by CTP, the final product of the ATCase-initiated pathway. The rate of the reaction catalyzed by ATCase is fast at low concentrations of CTP but slows as CTP concentration increases (Figure 10.2). Thus, the pathway continues to make new pyrimidines until sufficient quantities of CTP have accumulated. The inhibition of ATCase by CTP is an example of feedback inhibition, the inhibition of an enzyme by the end product of the pathway. Feedback inhibition by CTP ensures that N-carbamoylaspartate and subsequent intermediates in the pathway are not needlessly formed when pyrimidines are abundant.

Figure 10.2: CTP inhibits ATCase. Cytidine triphosphate, an end product of the pyrimidine-synthesis pathway, inhibits aspartate transcarbamoylase despite having little structural similarity to reactants or products.

     The inhibitory ability of CTP is remarkable because CTP is structurally quite different from the substrates of the reaction (Figure 10.1). Thus CTP must bind to a site distinct from the active site at which substrate binds. Such sites are called allosteric or regulatory sites. CTP is an example of an allosteric inhibitor. In ATCase (but not all allosterically regulated enzymes), the catalytic sites and the regulatory sites are on separate polypeptide chains.

Allosterically regulated enzymes do not follow Michaelis–Menten kinetics

Allosteric enzymes are distinguished by their response to changes in substrate concentration in addition to their susceptibility to regulation by other molecules. Let us examine the rate of product formation as a function of substrate concentration for ATCase (Figure 10.3). The curve differs from that expected for an enzyme that follows Michaelis–Menten kinetics. The observed curve is referred to as sigmoidal because it resembles the letter “S.” The vast majority of allosteric enzymes display sigmoidal kinetics. Recall from the discussion of hemoglobin that sigmoidal curves result from cooperation between subunits: the binding of substrate to one active site in a molecule increases the likelihood that substrate will bind to other active sites. To understand the basis of sigmoidal enzyme kinetics and inhibition by CTP, we need to examine the structure of ATCase.

Figure 10.3: ATCase displays sigmoidal kinetics. A plot of product formation as a function of substrate concentration produces a sigmoidal curve because the binding of substrate to one active site increases the activity at the other active sites. Thus, the enzyme shows cooperativity.

ATCase consists of separable catalytic and regulatory subunits

What is the evidence that ATCase has distinct regulatory and catalytic sites? ATCase can be literally separated into regulatory (r) and catalytic (c) subunits by treatment with a mercurial compound such as p-hydroxymercuribenzoate, which reacts with sulfhydryl groups (Figure 10.4). Ultracentrifugation following treatment with mercurials revealed that ATCase is composed of two kinds of subunits (Figure 10.5). The subunits can be readily separated by ion-exchange chromatography because they differ markedly in charge or by centrifugation in a sucrose density gradient because they differ in size. These size differences are manifested in the sedimentation coefficients: that of the native enzyme is 11.6S, whereas those of the dissociated subunits are 2.8S and 5.8S. The attached p-mercuribenzoate groups can be removed from the separated subunits by adding an excess of mercaptoethanol, providing isolated subunits for study.

Figure 10.4: Modification of cysteine residues. p-Hydroxymercuribenzoate reacts with crucial cysteine residues in aspartate transcarbamoylase.
Figure 10.5: Ultracentrifugation studies of ATCase. Sedimentation velocity patterns of (A) native ATCase and (B) the enzyme after treatment with p-hydroxymercuribenzoate show that the enzyme can be dissociated into regulatory (r) and catalytic (c) subunits.
[Data from J. C. Gerhart and H. K. Schachman, Biochemistry 4:1054–1062, 1965.]

     The larger subunit is the catalytic subunit. This subunit has catalytic activity but displays the hyperbolic kinetics of Michaelis-Menten enzymes rather than sigmoidal kinetics. Furthermore, the isolated catalytic subunit is unresponsive to CTP. The isolated smaller subunit can bind CTP, but has no catalytic activity. Hence, that subunit is the regulatory subunit. The catalytic subunit (c3) consists of three chains (34 kDa each), and the regulatory subunit (r2) consists of two chains (17 kDa each). The catalytic and regulatory subunits combine rapidly when they are mixed. The resulting complex has the same structure, c6r6, as the native enzyme: two catalytic trimers and three regulatory dimers.

288

Most strikingly, the reconstituted enzyme has the same allosteric and kinetic properties as those of the native enzyme. Thus, ATCase is composed of discrete catalytic and regulatory subunits, and the interaction of the subunits in the native enzyme produces its regulatory and catalytic properties. The fact that the enzyme can be separated into isolated catalytic and regulatory subunits, which can be reconstituted back to the functional enzyme, allows for a variety of experiments to characterize the allosteric properties of the enzyme (Problems 33 and 34).

Allosteric interactions in ATCase are mediated by large changes in quaternary structure

What are the subunit interactions that account for the properties of ATCase? Significant clues have been provided by the three-dimensional structure of ATCase in various forms. Two catalytic trimers are stacked one on top of the other, linked by three dimers of the regulatory chains (Figure 10.6). There are significant contacts between the catalytic and the regulatory subunits: each r chain within a regulatory dimer interacts with a c chain within a catalytic trimer. The c chain makes contact with a structural domain in the r chain that is stabilized by a zinc ion bound to four cysteine residues. The zinc ion is critical for the interaction of the r chain with the c chain. The mercurial compound p-hydroxymercuribenzoate is able to dissociate the catalytic and regulatory subunits because mercury binds strongly to the cysteine residues, displacing the zinc and preventing interaction with the c chain.

Figure 10.6: Structure of ATCase. (A) The quaternary structure of aspartate transcarbamoylase as viewed from the top. The drawing in the center is a simplified representation of the relations between subunits. A single catalytic trimer [catalytic (c) chains, shown in yellow] is visible; in this view, the second trimer is hidden below the one visible. Notice that each r chain interacts with a c chain through the zinc domain. (B) A side view of the complex.
[Drawn from 1RAI.pdb.]

289

     To locate the active sites, the enzyme is crystallized in the presence of N-(phosphonacetyl)-l-aspartate (PALA), a bisubstrate analog (an analog of the two substrates) that resembles an intermediate along the pathway of catalysis (Figure 10.7). PALA is a potent competitive inhibitor of ATCase that binds to and blocks the active sites. The structure of the ATCase–PALA complex reveals that PALA binds at sites lying at the boundaries between pairs of c chains within a catalytic trimer (Figure 10.8). Each catalytic trimer contributes three active sites to the complete enzyme. Further examination of the ATCase–PALA complex reveals a remarkable change in quaternary structure on binding of PALA. The two catalytic trimers move 12 Å farther apart and rotate approximately 10 degrees about their common threefold axis of symmetry. Moreover, the regulatory dimers rotate approximately 15 degrees to accommodate this motion (Figure 10.9). The enzyme literally expands on PALA binding. In essence, ATCase has two distinct quaternary forms: one that predominates in the absence of substrate or substrate analogs and another that predominates when substrates or analogs are bound. We call these forms the T (for tense) state and the R (for relaxed) state, respectively, as we did for the two quaternary states of hemoglobin.

Figure 10.7: PALA, a bisubstrate analog. (Top) Nucleophilic attack by the amino group of aspartate on the carbonyl carbon atom of carbamoyl phosphate generates an intermediate on the pathway to the formation of N-carbamoylaspartate. (Bottom) N-(Phosphonacetyl)-l-aspartate (PALA) is an analog of the reaction intermediate and a potent competitive inhibitor of aspartate transcarbamoylase.
Figure 10.8: The active site of ATCase. Some of the crucial active-site residues are shown binding to the inhibitor PALA (shaded gray). Notice that the active site is composed mainly of residues from one c chain, but an adjacent c chain also contributes important residues (boxed in green).
[Drawn from 8ATC.pdb.]
Figure 10.9: The T-to-R state transition in ATCase. Aspartate transcarbamoylase exists in two conformations: a compact, relatively inactive form called the tense (T) state and an expanded form called the relaxed (R) state. Notice that the structure of ATCase changes dramatically in the transition from the T state to the R State. PALA binding stabilizes the R state.

290

     How can we explain the enzyme’s sigmoidal kinetics in light of the structural observations? Like hemoglobin, the enzyme exists in an equilibrium between the T state and the R state.

In the absence of substrate, almost all the enzyme molecules are in the T state because the T state is more stable than the R state. The ratio of the concentration of enzyme in the T state to that in the R state is called the allosteric contstant (L). For most allosteric enzymes, L is on the order of 102 to 103.

The T state has a low affinity for substrate and hence shows a low catalytic activity. The occasional binding of a substrate molecule to one active site in an enzyme increases the likelihood that the entire enzyme shifts to the R state with its higher binding affinity. The addition of more substrate has two effects. First, it increases the probability that each enzyme molecule will bind at least one substrate molecule. Second, it increases the average number of substrate molecules bound to each enzyme. The presence of additional substrate will increase the fraction of enzyme molecules in the more active R state because the position of the equilibrium depends on the number of active sites that are occupied by substrate. We considered this property, called cooperativity because the subunits cooperate with one another, when we discussed the sigmoidal oxygen-binding curve of hemoglobin. The effects of substrates on allosteric enzymes are referred to as homotropic effects (from the Greek homós, “same”).

This mechanism for allosteric regulation is referred to as the concerted model because the change in the enzyme is “all or none”; the entire enzyme is converted from T into R, affecting all of the catalytic sites equally. In contrast, the sequential model assumes that the binding of ligand to one site on the complex can affect neighboring sites without causing all subunits to undergo the T-to-R transition. Although the concerted model explains the behavior of ATCase well, most other allosteric enzymes have features of both models.

291

The sigmoidal curve for ATCase can be pictured as a composite of two Michaelis–Menten curves, one corresponding to the less-active T state and the other to the more-active R state. At low concentrations of substrate, the curve closely resembles that of the T state enzyme. As the substrate concentration is increased, the curve progressively shifts to resemble that of the R state enzyme (Figure 10.10).

Figure 10.10: Basis for the sigmoidal curve. The generation of the sigmoidal curve by the property of cooperativity can be understood by imagining an allosteric enzyme as a mixture of two Michaelis–Menten enzymes, one with a high value of KM that corresponds to the T state and another with a low value of KM that corresponds to the R state. As the concentration of substrate is increased, the equilibrium shifts from the T state to the R state, which results in a steep rise in activity with respect to substrate concentration.

What is the biochemical advantage of sigmoidal kinetcs? Allosteric enzymes transition from a less active state to a more active state within a narrow range of substrate concentration. The benefit of this behavior is illustrated in Figure 10.11, which compares the kinetics of a Michaelis-Menten enzyme (blue curve) to that of an allosteric enzyme (red curve). In this example, the Michaelis-Menten enzyme requires an approximately 27-fold increase in substrate concentration to increase Vo from 0.1 Vmax to 0.8 Vmax.. In contrast, the allosteric enzyme requires only about a 4-fold increase in substrate concentration to attain the same increase in velocity. The activity of allosteric enzymes is more sensitive to changes in substrate concentration near KM than are Michaelis–Menten enzymes with the same Vmax. This sensitivity is called a threshold effect: below a certain substrate concentration, there is little enzyme activity. However, after the threshold has been reached, enzyme activity increases rapidly. In other words, much like an “on or off” switch, cooperativity ensures that most of the enzyme is either on (R state) or off (T state). The vast majority of allosteric enzymes display sigmoidal kinetics.

Figure 10.11: Allosteric enzymes display threshold effects. As the T-to-R transition occurs, the velocity increases over a narrower range of substrate concentration for an allosteric enzyme (red curve) than for a Michaelis–Menten enzyme (blue curve).

Allosteric regulators modulate the T-to-R equilibrium

We now turn our attention to the effects of pyrimidine nucleotides. As noted earlier, CTP inhibits the action of ATCase. X-ray studies of ATCase in the presence of CTP reveal (1) that the enzyme is in the T state when bound to CTP and (2) that a binding site for this nucleotide exists in each regulatory chain in a domain that does not interact with the catalytic subunit (Figure 10.12). Each active site is more than 50 Å from the nearest CTP-binding site. The question naturally arises, How can CTP inhibit the catalytic activity of the enzyme when it does not interact with the catalytic chain?

Figure 10.12: CTP stabilizes the T state. The binding of CTP to the regulatory subunit of aspartate transcarbamoylase stabilizes the T state.

The quaternary structural changes observed on substrate-analog binding suggest a mechanism for inhibition by CTP (Figure 10.13). The binding of the inhibitor CTP to the T state shifts the T-to-R equilibrium in favor of the T state, decreasing net enzyme activity. CTP increases the allosteric coefficient from 200 in its absence to 1250 when all of the regulatory sites are occupied by CTP. The binding of CTP makes it more difficult for substrate binding to convert the enzyme into the R state. Consequently, CTP increases the initial phase of the sigmoidal curve (Figure 10.14). More substrate is required to attain a given reaction rate. UTP, the immediate precursor to CTP, also regulates ATCase. While unable to inhibit the enzyme alone, UTP synergistically inhibits ATCase in the presence of CTP.

Figure 10.13: The R state and the T state are in equilibrium. Even in the absence of any substrate or regulators, aspartate transcarbamoylase exists in equilibrium between the R and the T states. Under these conditions, the T state is favored by a factor of approximately 200.

292

Interestingly, ATP, too, is an allosteric effector of ATCase, binding to the same site as CTP. However, ATP binding stabilizes the R state, lowering the allosteric coefficient from 200 to 70 and thus increasing the reaction rate at a given aspartate concentration (Figure 10.14). At high concen trations of ATP, the kinetic profile shows a less-pronounced sigmoidal behavior. Because ATP and CTP bind at the same site, high levels of ATP prevent CTP from inhibiting the enzyme. The effects of nonsubstrate molecules on allosteric enzymes (such as those of CTP and ATP on ATCase) are referred to as heterotropic effects (from the Greek héteros, “different”). Substrates generate the sigmoidal curve (homotropic effects), whereas regulators shift the KM (heterotropic effects). Note, however, that both types of effect are generated by altering the T/R ratio.

Figure 10.14: Effect of CTP and ATP on ATCase kinetics. CTP stabilizes the T state of aspartate transcarbamoylase, making it more difficult for substrate binding to convert the enzyme into the R state. As a result, the curve is shifted to the right, as shown in red. ATP is an allosteric activator of aspartate transcarbamoylase because it stabilizes the R state, making it easier for substrate to bind. As a result, the curve is shifted to the left, as shown in blue.

The increase in ATCase activity in response to increased ATP concentration has two potential physiological ramifications. First, high ATP concentration signals a high concentration of purine nucleotides in the cell; the increase in ATCase activity will tend to balance the purine and pyrimidine pools. Second, a high concentration of ATP indicates that energy is available for mRNA synthesis and DNA replication and leads to the synthesis of pyrimidines needed for these processes.