Protein Function

CHAPTER 5 PROTEIN FUNCTION It may seem counterintuitive that a protein’s interaction with another molecule could be important if it does not alter the associated molecule. Yet, transient interactions of this type are at the heart of many complex physiological processes, such as oxygen transport, transmission of nerve impulses, and immune function. Defining which molecules interact and quantifying such interactions are common and illuminating tasks in every biochemical subdiscipline. The study of proteins that function through reversible interactions can be organized around six key principles of protein function, some of which will be familiar from Chapter 4: The functions of many proteins involve the reversible binding of other molecules. A molecule bound reversibly by a protein is called a ligand. A ligand may be any kind of molecule, including another protein. The transient nature of protein-ligand interactions is critical to life, allowing an organism to respond rapidly and reversibly to changing environmental and metabolic circumstances. A ligand binds a protein at a binding site that is complementary to the ligand in size, shape, charge, and hydrophobic or hydrophilic character. The interaction is specific: the protein can discriminate among the thousands of different molecules in its environment and selectively bind only one or a few types. A given protein may have separate binding sites for several different ligands. These specific molecular interactions are crucial in maintaining the high degree of order in a living system. Proteins are flexible. Changes in conformation may be subtle, reflecting molecular vibrations and small movements of amino acid residues throughout the protein. Changes in conformation may also be more dramatic, with major segments of the protein structure moving as much as several nanometers. Specific conformational changes are frequently essential to a protein’s function. The binding of a protein and a ligand is oen coupled to a conformational change in the protein that makes the binding site more complementary to the ligand, permitting tighter binding. The structural adaptation that occurs between protein and ligand is called induced fit. In a multisubunit protein, a conformational change in one subunit oen affects the conformation of other subunits. Interactions between ligands and proteins may be regulated. The themes in our discussion of noncatalytic functions of proteins in this chapter — binding, specificity, and conformational change — are continued in Chapter 6, with the added element of proteins participating in chemical transformations. This discussion excludes the binding of water, which may interact weakly and nonspecifically with many parts of a protein. 5.1 Reversible Binding of a Protein to a Ligand: Oxygen- Binding Proteins Myoglobin and hemoglobin may be the most-studied and best- understood proteins. They were the first proteins for which three- dimensional structures were determined, and these two molecules illustrate almost every aspect of that critical biochemical process: the reversible binding of a ligand to a protein. This classic model of protein function tells us a great deal about how proteins work. Oxygen Can Bind to a Heme Prosthetic Group Oxygen is poorly soluble in aqueous solutions (see Table 2-2) and cannot be carried to tissues in sufficient quantity if it is simply dissolved in blood serum. Also, diffusion of oxygen through tissues is ineffective over distances greater than a few millimeters. The evolution of larger, multicellular animals depended on the evolution of proteins that could transport and store oxygen. However, none of the amino acid side chains in proteins are suited for the reversible binding of oxygen molecules. This role is filled by certain transition metals, among them iron and copper, that have a strong tendency to bind oxygen. Multicellular organisms exploit the properties of metals, most commonly iron, for oxygen transport. However, free iron promotes the formation of highly reactive oxygen species such as hydroxyl radicals that can damage DNA and other macromolecules. Iron used in cells is therefore bound in forms that sequester it and/or make it less reactive. In multicellular organisms, iron is oen incorporated into a protein-bound prosthetic group called heme (or haem). (Recall from Chapter 3 that a prosthetic group is a compound permanently associated with a protein that contributes to the protein’s function.) Heme is found in many oxygen-transporting proteins, as well as in some proteins, such as the cytochromes, that participate in oxidation- reduction (electron-transfer) reactions (Chapter 19). Heme consists of a complex organic ring structure, protoporphyrin, to which is bound a single iron atom in its ferrous (Fe2+) state (Fig. 5-1). The iron atom has six coordination bonds: four to nitrogen atoms that are part of the flat porphyrin ring system, and two perpendicular to the porphyrin. FIGURE 5-1 Heme. The heme group is present in myoglobin, hemoglobin, and many other proteins, designated heme proteins. Heme consists of a complex organic ring structure, protoporphyrin IX, with a bound iron atom in its ferrous (Fe2+) state. (a) Porphyrins, of which protoporphyrin IX is just one example, consist of four pyrrole rings linked by methene bridges, with substitutions at one or more of the positions denoted X. (b, c) Two representations of heme. The iron atom of heme has six coordination bonds: four in the plane of, and bonded to, the flat porphyrin ring system, and (d) two perpendicular to it. [(c) Data from PDB ID 1CCR, H. Ochi et al., J. Mol. Biol. 166:407, 1983.] Iron in the Fe2+ state binds oxygen reversibly; in the Fe3+ state it does not bind oxygen. The structure of heme and the globins to which it is bound represents an evolutionary adaptation to prevent Fe2+ oxidation. Free heme molecules (heme not bound to protein) leave Fe2+ with two “open” coordination bonds. Simultaneous reaction of one O2 molecule with two free heme molecules (or two free Fe2+) can result in irreversible conversion of Fe2+ to Fe3+. The coordinated nitrogen atoms (which have an electron-donating character) help prevent conversion of the heme iron to the ferric (Fe3+) state. In heme- containing proteins, this reaction is further prevented by sequestering each heme deep within the protein structure. Thus, access to the two open coordination bonds is restricted. In the globins that will shortly become the focus of our narrative, one of these two coordination bonds is occupied by a side-chain nitrogen of a highly conserved His residue referred to as the proximal His. The other is the binding site for molecular oxygen (O2) (Fig. 5-2). When oxygen binds, the electronic properties of heme iron change; this accounts for the change in color from the dark purple of oxygen-depleted venous blood to the bright red of oxygen-rich arterial blood. Some small molecules, such as carbon monoxide (CO) and nitric oxide (NO), coordinate to heme iron with greater affinity than does O2. When a molecule of CO is bound to heme, O2 is excluded, which is why CO is highly toxic to aerobic organisms (a topic explored in Box 5- 1). By surrounding and sequestering heme, oxygen-binding proteins regulate the access of small molecules to the heme iron. FIGURE 5-2 The heme group viewed from the side. This view shows the two coordination bonds to Fe2+ that are perpendicular to the porphyrin ring system. One is occupied by a His residue called the proximal His, His93 in myoglobin, also designated His F8 (the eighth residue in α helix F; see Fig. 5-3); the other is the binding site for oxygen. The remaining four coordination bonds are in the plane of, and bonded to, the flat porphyrin ring system. Globins Are a Family of Oxygen- Binding Proteins The globins are a widespread family of proteins. Globins are commonly found in eukaryotes of all classes, as well as in archaea and bacteria. All evolved from a common ancestral protein. Their tertiary structure is highly conserved, made up of eight α -helical segments connected by bends. This folding pattern, illustrated by myoglobin (Fig. 5-3), constitutes a structural motif known as the globin fold. The primary amino acid sequence of globins is less conserved, well reflecting both the ancient origins of this protein family and the evolutionary relatedness of species from which globin comparisons might be made. FIGURE 5-3 Myoglobin. The eight α -helical segments (shown here as cylinders) are labeled A through H. Nonhelical residues in the bends that connect them are labeled AB, CD, EF, and so forth, indicating the segments they interconnect. A few bends, including BC and DE, are abrupt and do not contain any residues; these are sometimes not labeled. The heme is bound in a pocket made up largely of the E and F helices, although amino acid residues from other segments of the protein also participate. [Data from PDB ID 1MBO, S. E. Phillips, J. Mol. Biol. 142:531, 1980.] Most globins function in oxygen transport or storage, although some play a role in the sensing of oxygen, nitric oxide, or carbon monoxide. The simple nematode worm Caenorhabditis elegans has genes encoding 33 different globins. In humans and other mammals, there are at least four kinds of globins. The monomeric myoglobin facilitates oxygen diffusion in muscle tissue. Myoglobin is particularly abundant in the muscles of diving marine mammals such as seals and whales, where it also has an oxygen-storage function for prolonged excursions undersea. The tetrameric hemoglobin is responsible for oxygen transport in the bloodstream. The monomeric neuroglobin is expressed largely in neurons and helps to protect the brain from hypoxia (low oxygen) or ischemia (restricted blood supply). Cytoglobin, another monomeric globin, is found at high concentrations in the walls of blood vessels, where it functions to regulate levels of nitric oxide, a localized signal for muscle relaxation (see Box 12-2). Myoglobin Has a Single Binding Site for Oxygen Any detailed discussion of protein function inevitably involves protein structure. Myoglobin (M r 16,700; abbreviated Mb) is a single polypeptide of 153 amino acid residues with one molecule of heme (Fig. 5-3). About 78% of the amino acid residues in the protein are found in the eight α helices typical of the globin fold, named A through H. An individual amino acid residue is designated either by its position in the amino acid sequence or by its location in the sequence of a particular α -helical segment. For example, the His residue coordinated to the heme in myoglobin — the proximal His — is His93 (the ninety-third residue from the amino-terminal end of the myoglobin polypeptide sequence) and is also called His F8 (the eighth residue in α helix F). The bends in the structure are designated AB, CD, EF, FG, and GH, reflecting the α -helical segments they connect. Protein-Ligand Interactions Can Be Described Quantitatively The function of myoglobin depends on the protein’s ability not only to bind oxygen but also to release it when and where it is needed. Function in biochemistry oen revolves around a reversible protein-ligand interaction of this type. A quantitative description of this interaction is a central part of many biochemical investigations. In general, the reversible binding of a protein (P) to a ligand (L) can be described by a simple equilibrium expression: P + L⇌ PL (5-1) The reaction is characterized by an equilibrium constant, Ka, such that Ka = = (5-2) [PL] [P][L] ka kd where ka and kd are rate constants (more on these below). The term Ka is an association constant (not to be confused with the Ka that denotes an acid dissociation constant; p. 57) that describes the equilibrium between the complex and the unbound components of the complex. The association constant provides a measure of the affinity of the ligand L for the protein. Ka has units of M −1; a higher value of Ka corresponds to a higher affinity of the ligand for the protein. The equilibrium term Ka is also equivalent to the ratio of the rates of the forward (association) and reverse (dissociation) reactions that form the PL complex. The association rate is described by the rate constant ka, and dissociation is described by the rate constant kd. As discussed further in the next chapter, rate constants are proportionality constants, describing the fraction of a pool of reactant that reacts in a given amount of time. When the reaction involves one molecule, such as the dissociation reaction PL → P + L, the reaction is first order and the rate constant (kd) has units of reciprocal time (s−1). When the reaction involves two molecules, such as the association reaction P + L → PL, it is called second order, and its rate constant (ka) has units of M −1 s−1. KEY CONVENTION Equilibrium constants are denoted with a capital K and rate constants are denoted with a lowercase k.

A rearrangement of the first part of Equation 5-2 shows that the ratio of bound to free protein is directly proportional to the concentration of free ligand: Ka[L]= (5-3) When the concentration of the ligand is much greater than the concentration of ligand-binding sites, the binding of the ligand by the protein does not appreciably change the concentration of free (unbound) ligand — that is, [L] remains constant. This condition is broadly applicable to most ligands that bind to proteins in cells, and it simplifies our description of the binding equilibrium. We can now consider the binding equilibrium from the standpoint of the fraction, Y, of ligand-binding sites on the protein that are occupied by ligand: Y = = (5-4) Substituting Ka[L][P] for [PL] (see Eqn 5-3) and rearranging terms gives [PL] [P] binding sites occupied total binding sites [PL] [PL]+ [P] K [L][P] K [L] [L] Y = = = (5-5) The value of Ka can be determined from a plot of Y versus the concentration of free ligand, [L] (Fig. 5-4a). Any equation of the form x = y/(y + z) describes a hyperbola, and Y is thus found to be a hyperbolic function of [L]. The fraction of ligand-binding sites occupied approaches saturation asymptotically as [L] increases. The [L] at which half of the available ligand-binding sites are occupied (that is, Y = 0.5) corresponds to 1/Ka. Ka[L][P] Ka[L][P]+ [P] Ka[L] Ka[L]+ 1 [L] [L]+ 1 Ka FIGURE 5-4 Graphical representations of ligand binding. The fraction of ligand-binding sites occupied, Y, is plotted against the concentration of free ligand. Both curves are rectangular hyperbolas. (a) A hypothetical binding curve for a ligand L. The [L] at which half of the available ligand-binding sites are occupied is equivalent to 1/Ka, or Kd. The curve has a horizontal asymptote at Y = 1 and a vertical asymptote (not shown) at [L]=− 1/Ka. (b) A curve describing the binding of oxygen to myoglobin. The partial pressure of O2 in the air above the solution is expressed in kilopascals (kPa). Oxygen binds tightly to myoglobin, with a P50 of only 0.26 kPa. It is more common (and intuitively simpler), however, to consider the dissociation constant, Kd, which is the reciprocal of Ka (Kd = 1/Ka) and has units of molar concentration (M). Kd is the equilibrium constant for the release of ligand. The relevant expressions change to Kd = = (5-6) [PL]= (5-7) Y = (5-8) When [L] equals Kd, half of the ligand-binding sites are occupied. As [L] falls below Kd, progressively less of the protein has ligand bound to it. For 90% of the available ligand-binding sites to be occupied, [L] must be nine times greater than Kd. In practice, Kd is used much more oen than Ka to express the affinity of a protein for a ligand. Note that a lower [P][L] [PL] kd ka [P][L] Kd [L] [L]+ Kd value of Kd corresponds to a higher affinity of ligand for the protein. The mathematics can be reduced to simple statements: Kd is equivalent to the molar concentration of ligand at which half of the available ligand-binding sites are occupied. At this point, the protein is said to have reached half-saturation with respect to ligand binding. The more tightly a protein binds a ligand, the lower the concentration of ligand required for half the binding sites to be occupied, and thus the lower the value of Kd. Some representative dissociation constants are given in Table 5-1; the scale shows typical ranges for dissociation constants found in biological systems. TABLE 5-1 Protein Dissociation Constants: Some Examples and Range Protein Ligand Kd (M) Avidin (egg white) Biotin 1× 10−15 Insulin receptor (human) Insulin 1× 10−10 Anti-HIV immunoglobulin (human) gp41 (HIV-1 surface protein) 4× 10−10 Nickel-binding protein (E. coli) Ni2+ 1× 10−7 Calmodulin (rat) Ca2+ 3× 10−6 2× 10−5 a b c Color bars indicate the range of dissociation constants typical of various classes of interactions in biological systems. A few interactions, such as that between the protein avidin and the enzyme cofactor biotin, fall outside the normal ranges. The avidin-biotin interaction is so tight it may be considered irreversible. Sequence-specific protein-DNA interactions reflect proteins that bind to a particular sequence of nucleotides in DNA, as opposed to general binding to any DNA site. A reported dissociation constant is valid only for the particular solution conditions under which it was measured. Kd values for a protein-ligand interaction can be altered, sometimes by several orders of magnitude, by changes in the solution’s salt concentration, pH, or other variables. This immunoglobulin was isolated as part of an effort to develop a vaccine against HIV. Immunoglobulins (described later in the chapter) are highly variable, and the Kd reported here should not be considered characteristic of all immunoglobulins. Calmodulin has four binding sites for calcium. The values shown reflect the highest- and lowest-affinity binding sites observed in one set of measurements. WORKED EXAMPLE 5-1 Receptor- Ligand Dissociation Constants Two proteins, A and B, bind to the same ligand, L, with the binding curves shown below. What is the dissociation constant, Kd, for each protein? Which protein (A or B) has a greater affinity for ligand L? a b c SOLUTION: We can determine the dissociation constants by inspecting the graph. Because Y represents the fraction of binding sites occupied by ligand, the concentration of ligand at which half the binding sites are occupied — that is, the point where the binding curve crosses the line where Y = 0.5 — is the dissociation constant. For A, Kd = 2μM ; for B, Kd = 6μM . Because A is half-saturated at a lower [L], it has a higher affinity for the ligand. WORKED EXAMPLE 5-2 Protein- Ligand Binding A protein has a Kd of 2.0 μ M. What is the [L] where Y = 0.6? SOLUTION: Solving Equation 5-8 for [L], first we obtain [L]= Y([L]+ Kd) Distributing Y on the right side gives [L]= Y[L]+ YKd Subtracting Y[L] from both sides [L]− Y[L]= YKd then factoring out [L] [L](1− Y)= YKd and dividing both sides by 1 − Y gives the final expression [L]= Substituting the values for Y and Kd given in the problem gives [L] = 0.6(2.0 μM )/(1− 0.6) = 3.0 μM The binding of oxygen to myoglobin follows the patterns discussed above. However, because oxygen is a gas, we must make some minor adjustments to the equations so that laboratory experiments can be carried out more conveniently. We first substitute the concentration of dissolved oxygen for [L] in Equation 5-8 to give Y = (5-9) As for any ligand, Kd equals the [O2] at which half of the available ligand-binding sites are occupied, or [O2]0.5. Equation 5-9 thus becomes Y = YKd (1− Y) [O2] [O2]+ Kd [O2] [O2]+ [O2]0.5 (5-10) In experiments using oxygen as a ligand, it is the partial pressure of oxygen (pO2) in the gas phase above the solution that is varied, because this is easier to measure than the concentration of oxygen dissolved in the solution. The concentration of a volatile substance in solution is always proportional to the local partial pressure of the gas. So, if we define the partial pressure of oxygen at [O2]0.5 as P50, substitution in Equation 5-10 gives Y = (5-11) A binding curve for myoglobin that relates Y to pO2 is shown in Figure 5-4b. Protein Structure Affects How Ligands Bind The binding of a ligand to a protein is rarely as simple as the above equations would suggest. The interaction is greatly affected by protein structure and is oen accompanied by conformational changes. For example, the specificity with which heme binds its various ligands is altered when the heme is a component of myoglobin. For free heme molecules, carbon pO2 pO2+ P50 monoxide binds more than 20,000 times better than does O2 (that is, the Kd or P50 for CO binding to free heme is more than 20,000 times lower than that for O2), but it binds only about 40 times better than O2 when the heme is bound in myoglobin. For free heme, the tighter binding by CO reflects differences in the way the orbital structures of CO and O2 interact with Fe2+. Those same orbital structures lead to different binding geometries for CO and O2 when they are bound to heme (Fig. 5-5a, b). The change in relative affinity of CO and O2 for heme when the heme is bound to a globin is mediated by the globin structure. FIGURE 5-5 Steric effects caused by ligand binding to the heme of myoglobin. (a) Oxygen binds to heme with the O2 axis at an angle, a binding conformation readily accommodated by myoglobin. (b) Carbon monoxide binds to free heme with the CO axis perpendicular to the plane of the porphyrin ring. (c) Another view of the heme of myoglobin, showing the arrangement of key amino acid residues around the heme. The bound O2 is hydrogen-bonded to the distal His, His E7 (His64), facilitating the binding of O2 compared with its binding to free heme. [(c) Data from PDB ID 1MBO, S. E. Phillips, J. Mol. Biol. 142:531, 1980.] When heme is bound to myoglobin, its affinity for O2 is selectively increased by the presence of the distal His (His64, or His E7 in myoglobin). The Fe-O2 complex is much more polar than the Fe-CO complex. A partial negative charge is distributed across the oxygen atoms in the bound O2 due to partial oxidation of the interacting iron atom. A hydrogen bond between the imidazole side chain of His E7 and the bound O2 stabilizes this polar complex electrostatically (Fig. 5-5c). The affinity of myoglobin for O2 is thus selectively increased by a factor of about 500; there is no such effect for Fe-CO binding in myoglobin. Consequently, the 20,000-fold stronger binding affinity of free heme for CO compared with O2 declines to approximately 40-fold for heme embedded in myoglobin. This favorable electrostatic effect on O2 binding is even more dramatic in some invertebrate hemoglobins, where two groups in the binding pocket can form strong hydrogen bonds with O2, causing the heme group to bind O2 with greater affinity than CO. This selective enhancement of O2 affinity in globins is physiologically important: it helps prevent poisoning by the CO that is generated in small amounts from metabolism and that is sometimes generated in larger amounts by our industrial-age environment. The binding of O2 to the heme in myoglobin also depends on molecular motions, or “breathing,” in the protein structure. The heme molecule is deeply buried in the folded polypeptide, with limited direct paths for oxygen to move from the surrounding solution to the ligand-binding site. If the protein were rigid, O2 could not readily enter or leave the heme pocket. However, rapid molecular flexing of the amino acid side chains produces transient cavities in the protein structure, and O2 makes its way in and out by moving through these cavities. Computer simulations of rapid structural fluctuations in myoglobin suggest that there are many such pathways. The distal His acts as a gate to control access to one major pocket near the heme iron. Rotation of that His residue to open and close the pocket occurs on a nanosecond (10−9 s) time scale. Thus, subtle conformational changes can be critical for protein activity. Structural changes take on more complexity in the multisubunit hemoglobin, which we consider next. Hemoglobin Transports Oxygen in Blood Nearly all the oxygen carried by whole blood in animals is bound and transported by hemoglobin in erythrocytes (red blood cells). Normal human erythrocytes are small (6 to 9 μ m in diameter), biconcave disks. They are formed from precursor stem cells called hemocytoblasts. In the maturation process, the stem cell produces daughter cells that form large amounts of hemoglobin and then lose their organelles — nucleus, mitochondria, and endoplasmic reticulum. Erythrocytes are thus incomplete, vestigial cells, unable to reproduce and, in humans, destined to survive for only about 120 days. Their main function is to carry hemoglobin, which is dissolved in the cytosol at a very high concentration (~34% by weight). In arterial blood passing from the lungs through the heart to the peripheral tissues, hemoglobin is about 96% saturated with oxygen. In the venous blood returning to the heart, hemoglobin is only about 64% saturated. Thus, each 100 mL of blood passing through a tissue releases about one-third of the oxygen it carries, or 6.5 mL of O2 gas at atmospheric pressure and body temperature. Myoglobin, with its hyperbolic binding curve for oxygen (Fig. 5- 4b), is relatively insensitive to small changes in the concentration of dissolved oxygen, so it functions well as an oxygen-storage protein. Hemoglobin, with its multiple subunits and O2-binding sites, is better suited to oxygen transport. As we shall see, interactions between the subunits of a multimeric protein can permit a highly sensitive response to small changes in ligand concentration. Interactions among the subunits in hemoglobin cause conformational changes that alter the affinity of the protein for oxygen. The modulation of oxygen binding allows the O2-transport protein to respond to changes in oxygen demand by tissues. Hemoglobin Subunits Are Structurally Similar to Myoglobin Hemoglobin (M r 64,500; abbreviated Hb) is roughly spherical, with a diameter of nearly 5.5 nm. It is a tetrameric protein containing four heme prosthetic groups, one associated with each polypeptide chain. Adult hemoglobin contains two types of globin: two α chains (141 residues each) and two β chains (146 residues each). The three-dimensional structures of the two types of subunits are very similar to each other and to myoglobin (Fig. 5-6), reflecting their evolution within the larger globin superfamily. However, fewer than half of the amino acid residues are identical in the polypeptide sequences of the α and β subunits, and only 27 are identical in the three polypeptides (Fig. 5-7). The helix-naming convention described for myoglobin is also applied to the hemoglobin polypeptides, except that the α subunit lacks the short D helix. The heme-binding pocket is made up largely of the E and F helices in each of the subunits. FIGURE 5-6 Comparison of the structures of myoglobin and the β subunit of hemoglobin. [Data from (le ) PDB ID 1MBO, S. E. Phillips, J. Mol. Biol. 142:531, 1980; (right) PDB ID 1HGA, R. Liddington et al., J. Mol. Biol. 228:551, 1992.] FIGURE 5-7 Comparison of whale myoglobin with the α and β chains of human hemoglobin. (a) Sequence alignment with dashed lines marking helix boundaries. For optimal alignment, short gaps are introduced into both Hb sequences where a few amino acids are present in the other sequences. With the exception of the missing D helix in the Hb α chain (Hbα ), this alignment permits the use of the helix lettering convention that emphasizes the common positioning of amino acid residues. Residues conserved in all known globins are shaded black. Residues conserved in the three globins shown here are shaded dark blue. Positions where the amino acid class (hydrophobic, charged, etc.) is conserved are shaded light blue. The common helix- letter-and-number designation for amino acids does not necessarily correspond to a common position in the linear sequence. For example, the distal His residue is His E7 in all three structures, but corresponds to His64, His58, and His63 in the linear sequences of Mb, Hbα , and Hbβ , respectively. Nonhelical residues at the amino and carboxyl termini are labeled NA and HC, respectively. (b) The location of shaded residues is shown in a ribbon representation of myoglobin structure. (c) Ribbon representations of the three globin structures are overlaid to show structural conservation. [Data from PDB ID 1MBO, S. E. Phillips, J. Mol. Biol. 142:531, 1980; PDB ID 1HGA, R. Liddington et al., J. Mol. Biol. 228:551, 1992.] The quaternary structure of hemoglobin features strong interactions between unlike subunits. The α1β1 interface (and its α2β2 counterpart) involves more than 30 residues, and its interaction is sufficiently strong that although mild treatment of hemoglobin with urea tends to disassemble the tetramer into αβ dimers, these dimers remain intact. The α1β2 (and α2β1) interface involves 19 residues (Fig. 5-8). The hydrophobic effect plays the major role in stabilizing these interfaces, but there are also many hydrogen bonds and a few ion pairs (or salt bridges), whose importance is discussed below. FIGURE 5-8 Dominant interactions between hemoglobin subunits. In this representation, α subunits are light and β subunits are dark. The strongest subunit interactions (highlighted) occur between unlike subunits. When oxygen binds, the α1β1 contact changes little, but there is a large change at the α1β2 contact, with several ion pairs broken. [Data from PDB ID 1HGA, R. Liddington et al., J. Mol. Biol. 228:551, 1992.] Hemoglobin Undergoes a Structural Change on Binding Oxygen X-ray analysis has revealed two major conformations of hemoglobin: the R state and the T state. Although oxygen binds to hemoglobin in either state, it has a significantly higher affinity for hemoglobin in the R state. Oxygen binding stabilizes the R state. When oxygen is absent experimentally, the T state is more stable and is thus the predominant conformation of deoxyhemoglobin. T and R originally denoted “tense” and “relaxed,” respectively, because the T state is stabilized by a greater number of ion pairs, many of which lie at the α1β2 (and α2β1) interface (Fig. 5-9). The binding of O2 to a hemoglobin subunit in the T state triggers a change in conformation to the R state. When the entire protein undergoes this transition, the structures of the individual subunits change little, but the αβ subunit pairs slide past each other and rotate, narrowing the pocket between the β subunits (Fig. 5-10). In this process, some of the ion pairs that stabilize the T state are broken and some new ones are formed. FIGURE 5-9 Some ion pairs that stabilize the T state of deoxyhemoglobin. Many of the ion pairs stabilizing the T state lie at the α1β2 (and α2β1) interface. Close-up view of a portion of a deoxyhemoglobin molecule in the T state. Interactions between the ion pairs His HC3 and Asp FG1 of the β subunit (blue) and between Lys C5 of the α subunit (gray) and His HC3 (its α -carboxyl group) of the β subunit are shown with dashed lines. (Recall that HC3 is the carboxyl-terminal residue of the β subunit.) These examples do not represent the entire network of ion pairs that stabilize the structure. [Data from PDB ID 1HGA, R. Liddington et al., J. Mol. Biol. 228:551, 1992.] FIGURE 5-10 The T → R transition. In these depictions of deoxyhemoglobin, as in Figure 5-9, the β subunits are blue and the α subunits are gray. Positively charged side chains and chain termini involved in ion pairs are shown in blue, and their negatively charged partners are shown in red. The Lys C5 of each α subunit and the Asp FG1 of each β subunit are visible but are not labeled (compare Fig. 5-9). Note that the molecule is oriented slightly differently than in Figure 5-9. The transition from the T state to the R state shi s the subunit pairs substantially, affecting certain ion pairs. Most noticeably, the His HC3 residues at the carboxyl termini of the β subunits, which are involved in ion pairs in the T state, rotate in the R state toward the center of the molecule, where they are no longer in ion pairs. Another dramatic result of the T → R transition is a narrowing of the pocket between the β subunits. [T state: Data from PDB ID 1HGA, R. Liddington et al., J. Mol. Biol. 228:551, 1992. R state: Data from PDB ID 1BBB, M. M. Silva et al., J. Biol. Chem. 267:17,248, 1992.] Max Perutz proposed that the T → R transition is triggered by changes in the positions of key amino acid side chains surrounding the heme. In the T state, the porphyrin is slightly puckered, causing the heme iron to protrude somewhat on the proximal His (His F8) side. The binding of O2 causes the heme to assume a more planar conformation, shiing the position of the proximal His and the attached F helix (Fig. 5-11). These changes lead to adjustments in the ion pairs at the α1β2 interface. FIGURE 5-11 Changes in conformation near heme on O2 binding to deoxyhemoglobin. The shi in the position of helix F when heme binds O2 is thought to be one of the adjustments that triggers the T → R transition. [T state: Data from PDB ID 1HGA, R. Liddington et al., J. Mol. Biol. 228:551, 1992. R state: Data from PDB ID 1BBB, M. M. Silva et al., J. Biol. Chem. 267:17,248, 1992; R state modified to represent O2 instead of CO.] Hemoglobin Binds Oxygen Cooperatively Hemoglobin must bind oxygen efficiently in the lungs, where the pO2 is about 13.3 kPa, and it must release oxygen in the tissues, where the pO2 is about 4 kPa. Myoglobin, or any protein that binds oxygen with a hyperbolic binding curve, would be ill-suited to this function, for the reason illustrated in Figure 5-12. A protein that bound O2 with high affinity would bind it efficiently in the lungs but would not release much of it in the tissues. If the protein bound oxygen with a sufficiently low affinity to release it in the tissues, it would not pick up much oxygen in the lungs. FIGURE 5-12 A sigmoid (cooperative) binding curve. A sigmoid binding curve can be viewed as a hybrid curve reflecting a transition from a low- affinity state to a high-affinity state. Because of its cooperative binding, as manifested by a sigmoid binding curve, hemoglobin is more sensitive to the small differences in O2 concentration between the tissues and the lungs, allowing it to bind oxygen in the lungs (where pO2 is high) and release it in the tissues (where pO2 is low). Hemoglobin solves the problem by undergoing a transition from a low-affinity state (the T state) to a high-affinity state (the R state) as more O2 molecules are bound. As a result, hemoglobin has a hybrid S-shaped, or sigmoid, binding curve for oxygen (Fig. 5-12). A single-subunit protein with a single ligand-binding site cannot produce a sigmoid binding curve — even if binding elicits a conformational change — because each molecule of ligand binds independently and cannot affect ligand binding to another molecule. In contrast, O2 binding to individual subunits of hemoglobin can alter the affinity for O2 in adjacent subunits. The first molecule of O2 that interacts with deoxyhemoglobin binds weakly, because it binds to a subunit in the T state. Its binding, however, leads to conformational changes that are communicated to adjacent subunits, making it easier for additional molecules of O2 to bind. In effect, the T → R transition occurs more readily in the second subunit once O2 is bound to the first subunit. The last (fourth) O2 molecule binds to a heme in a subunit that is already in the R state, and hence it binds with much higher affinity than the first molecule. An allosteric protein is one in which the binding of a ligand to one site affects the binding properties of another site on the same protein. The term “allosteric” derives from the Greek allos, “other,” and stereos, “solid” or “shape.” Allosteric proteins are those having “other shapes,” or conformations, induced by the binding of ligands referred to as modulators. The conformational changes induced by the modulator(s) interconvert more-active and less-active forms of the protein. The modulators for allosteric proteins may be either inhibitors or activators. When the normal ligand and modulator are identical, the interaction is termed homotropic. When the modulator is a molecule other than the normal ligand, the interaction is heterotropic. Some proteins have two or more modulators and therefore can have both homotropic and heterotropic interactions. Cooperative binding of a ligand to a multimeric protein, such as we observe with the binding of O2 to hemoglobin, is a form of allosteric binding. The binding of one ligand affects the affinities of any remaining unfilled binding sites, and O2 can be considered as both a ligand and an activating homotropic modulator. There is only one binding site for O2 on each subunit, so the allosteric effects giving rise to cooperativity are mediated by conformational changes transmitted from one subunit to another by subunit-subunit interactions. A sigmoid binding curve is diagnostic of cooperative binding. It permits a much more sensitive response to ligand concentration and is important to the function of many multisubunit proteins. The principle of allostery extends readily to regulatory enzymes, as we shall see in Chapter 6. As is the case with myoglobin, ligands other than oxygen can bind to hemoglobin. An important example is carbon monoxide, which binds to hemoglobin about 250 times better than does oxygen (the critical hydrogen bond between O2 and the distal His is not quite as strong in human hemoglobin as it is in most mammalian myoglobins, so the binding of O2 relative to CO is not augmented quite as much). Human exposure to CO can have tragic consequences (Box 5-1). BOX 5-1 MEDICINE Carbon Monoxide: A Stealthy Killer Summer, 2017. A Florida man drove his late-model SUV into his garage, took his wireless key fob with him inside the house, and went to bed. He was found, days later, dead from carbon monoxide poisoning. The man had assumed that removing the fob was enough to shut down the automobile. Instead, the car had idled in the garage until it ran out of gas, ultimately filling the attached house with CO. Since 2006, several dozen deaths and scores of injuries have resulted from similar incidents. Measures to address the safety issues posed by keyless ignitions, now standard in most new cars, have lagged. Meanwhile, the deaths are a vivid reminder that life depends on the reversibility of protein- ligand interactions. Carbon monoxide (CO), a colorless, odorless gas, is one of the most common known causes of death due to poisoning (in countries where such records are kept). CO has an approximately 250-fold greater affinity for hemoglobin than does oxygen. Consequently, relatively low levels of CO can have substantial and tragic effects. When CO combines with hemoglobin, the complex is referred to as carboxyhemoglobin, or COHb. Some CO is produced by natural processes, but locally high levels generally result only from human activities. Engine and furnace exhausts are important sources, as CO is a byproduct of the incomplete combustion of fossil fuels. In the United States alone, nearly 4,000 people succumb to CO poisoning each year, both accidentally and intentionally. Many of the accidental deaths involve undetected CO buildup in enclosed spaces. This might occur when a household furnace malfunctions, or a gas generator is operated indoors during a power outage or shutoff, venting CO into a home. CO poisoning can also occur in open spaces, as unsuspecting people at work or play inhale the exhaust from generators, outboard motors, tractor engines, recreational vehicles, or lawn mowers. Carbon monoxide levels in the atmosphere are rarely dangerous, ranging from less than 0.05 part per million (ppm) in remote and uninhabited areas to 3 to 4 ppm in some cities of the Northern Hemisphere. In the United States, the government-mandated (Occupational Safety and Health Administration, OSHA) limit for CO at worksites is 35 ppm for people working an eight-hour shi . The tight binding of CO to hemoglobin means that COHb can accumulate over time as people are exposed to a constant low-level source of CO. In an average, healthy individual, 1% or less of the total hemoglobin is complexed as COHb. Since CO is a product of tobacco smoke, many smokers have COHb levels in the range of 3% to 8% of total hemoglobin, and the levels can rise to 15% for chain-smokers. COHb levels equilibrate at 50% in people who breathe air containing 570 ppm of CO for several hours. Reliable methods have been developed that relate CO content in the atmosphere to COHb levels in the blood (Fig. 1). CO levels in the home of the Florida victim exceeded OSHA-mandated limits by more than 30-fold. FIGURE 1 Relationship between levels of COHb in blood and concentration of CO in the surrounding air. Four different conditions of exposure are shown, comparing the effects of short exposure versus extended exposure, and exposure at rest versus exposure during light exercise. [Data from R. F. Coburn et al., J. Clin. Invest. 44:1899, 1965.] How is a human affected by COHb? At levels of less than 10% of total hemoglobin, symptoms are rarely observed. At 15%, the individual experiences mild headaches. At 20% to 30%, the headache is severe and is generally accompanied by nausea, dizziness, confusion, disorientation, and some visual disturbances; these symptoms are generally reversed if the individual is treated with oxygen. At COHb levels of 30% to 50%, the neurological symptoms become more severe, and at levels near 50%, the individual loses consciousness and can sink into coma. Respiratory failure may follow. With prolonged exposure, some damage becomes permanent. Death normally occurs when COHb levels rise above 60%. Binding of CO to hemoglobin is affected by many factors, including exercise (Fig. 1) and changes in air pressure related to altitude. Because of their higher base levels of COHb, smokers exposed to a source of CO o en develop symptoms faster than nonsmokers. Individuals with heart, lung, or blood diseases that reduce the availability of oxygen to tissues may also experience symptoms at lower levels of CO exposure. Fetuses are at particular risk for CO poisoning. Fetal hemoglobin, with γ subunits replacing β subunits, has a somewhat higher affinity for CO than adult hemoglobin (see p. 148). Cases of CO exposure have been recorded in which the fetus died but the pregnant woman recovered. It may seem surprising that the loss of half of one’s hemoglobin to COHb can prove fatal — we know that people with any of several anemic conditions manage to function reasonably well with half the usual complement of active hemoglobin. However, the binding of CO to hemoglobin does more than remove protein from the pool available to bind oxygen. It also affects the affinity of the remaining hemoglobin subunits for oxygen. As CO binds to one or two subunits of a hemoglobin tetramer, the affinity for O2 is increased substantially in the remaining subunits (Fig. 2). Thus, a hemoglobin tetramer with two bound CO molecules can efficiently bind O2 in the lungs — but it releases very little of it in the tissues. Oxygen deprivation in the tissues rapidly becomes severe. To add to the problem, the effects of CO are not limited to interference with hemoglobin function. CO binds to other heme proteins and a variety of metalloproteins. The effects of these interactions are not yet well understood, but they may be responsible for some of the longer-term effects of acute but nonfatal CO poisoning. FIGURE 2 Several oxygen-binding curves: for normal hemoglobin, for hemoglobin from an anemic individual with only 50% of her hemoglobin functional, and for hemoglobin from an individual with 50% of his hemoglobin subunits complexed with CO. The pO2 in human lungs and tissues is indicated. [Data from F. J. W. Roughton and R. C. Darling, Am. J. Physiol. 141:17, 1944.] When CO poisoning is suspected, rapid removal of the person from the CO source is essential, but this does not always result in rapid recovery. When an individual is moved from the CO-polluted site to a normal, outdoor atmosphere, O2 begins to replace the CO in hemoglobin — but the COHb level drops only slowly. The half-time is 2 to 6.5 hours, depending on individual and environmental factors. If 100% oxygen is administered with a mask, the rate of exchange can be increased about fourfold; the half-time for O2–CO exchange can be reduced to tens of minutes if 100% oxygen at a pressure of 3 atm (303 kPa) is supplied. Thus, rapid treatment by a properly equipped medical team is critical. A home carbon monoxide detector is a simple and inexpensive measure to avoid possible tragedy. A er completing the research for this box, we immediately purchased several new CO detectors for our homes. Cooperative Ligand Binding Can Be Described Quantitatively Cooperative binding of oxygen by hemoglobin was first analyzed by Archibald Hill in 1910. From this work came a general approach to the study of cooperative ligand binding to multisubunit proteins. For a protein with n binding sites, the equilibrium of Equation 5-1 becomes P + nL⇌ PLn (5-12) and the expression for the association constant becomes Ka = (5-13) [PLn] [P][L]n The expression for Y (see Eqn 5-8) is Y = (5-14) Rearranging, then taking the log of both sides, yields = (5-15) log( )= nlog[L]− logKd (5-16) where Kd = [L]n0.5 Equation 5-16 is the Hill equation, and a plot of log [Y/(1 − Y)] versus log [L] is called a Hill plot. Based on the equation, the Hill plot should have a slope of n. However, the experimentally determined slope actually reflects not the number of binding sites but the degree of interaction between them. The slope of a Hill plot is therefore denoted by nH, the Hill coefficient, which is a measure of the degree of cooperativity. If nH equals 1, ligand [L]n [L]n+ Kd Y 1− Y [L]n Kd Y 1− Y binding is not cooperative, a situation that can arise even in a multisubunit protein if the subunits do not communicate. An nH of greater than 1 indicates positive cooperativity in ligand binding. This is the situation observed in hemoglobin, in which the binding of one molecule of ligand facilitates the binding of others. The theoretical upper limit for nH is reached when nH = n. In this case the binding would be completely cooperative: all binding sites on the protein would bind ligand simultaneously, and no protein molecules partially saturated with ligand would be present under any conditions. This limit is never reached in practice, and the measured value of nH is always less than the actual number of ligand-binding sites in the protein. An nH of less than 1 indicates negative cooperativity, in which the binding of one molecule of ligand impedes the binding of others. Well-documented cases of negative cooperativity are rare. To adapt the Hill equation to the binding of oxygen to hemoglobin, we must again substitute pO2 for [L] and Pn50 for Kd: log( )= nlogpO2− nlogP50 (5-17) Hill plots for myoglobin and hemoglobin are given in Figure 5-13. Y 1− Y FIGURE 5-13 Hill plots for oxygen binding to myoglobin and hemoglobin. When nH = 1, there is no evident cooperativity. The maximum degree of cooperativity observed for hemoglobin corresponds approximately to nH = 3. Note that while this indicates a high level of cooperativity, nH is less than n, the number of O2-binding sites in hemoglobin. This is normal for a protein that exhibits allosteric binding behavior. Two Models Suggest Mechanisms for Cooperative Binding In a multisubunit ligand-binding protein like hemoglobin, how does the T → R transition occur? Two models for the cooperative binding of ligands to proteins with multiple binding sites have greatly influenced thinking about this problem, defining the two major possibilities. The first model was proposed by Jacques Monod, Jeffries Wyman, and Jean-Pierre Changeux in 1965, and is called the MWC model or the concerted model (Fig. 5-14a). The concerted model assumes that the subunits of a cooperatively binding protein are functionally identical, that each subunit can exist in (at least) two conformations, and that all subunits undergo the transition from one conformation to the other simultaneously. In this model, no protein has individual subunits in different conformations. The two conformations are in equilibrium. The ligand can bind to either conformation, but it binds much more tightly to the R state. Successive binding of ligand molecules to the low-affinity T state (which is more stable in the absence of ligand) gradually shis the equilibrium to favor the R state. FIGURE 5-14 Two general models for the interconversion of inactive and active forms of a protein during cooperative ligand binding. Although the models may be applied to any protein — including any enzyme (Chapter 6) — that exhibits cooperative binding, here we show four subunits because the model was originally proposed for hemoglobin. (a) In the concerted, or all-or-none, model (MWC model), all subunits are postulated to be in the same conformation, either all ◯ (low affinity or inactive) or all (high affinity or active). Depending on the equilibrium, Keq, between ◯ and forms, the binding of one or more ligand molecules (L) will pull the equilibrium toward the form. Subunits with bound L are shaded. (b) In the sequential model, each individual subunit can be in either the ◯ form or the form. The equilibrium is altered as additional ligands are bound, progressively favoring the R state. In the second model, the sequential model (Fig. 5-14b), proposed in 1966 by Daniel Koshland and colleagues, ligand binding can induce a change of conformation in an individual subunit. A conformational change in one subunit makes a similar change in an adjacent subunit more likely, and makes the binding of a second ligand molecule more likely as well. There are more potential intermediate states in this model than in the concerted model. The two models are not mutually exclusive; the concerted model may be viewed as the “all-or-none” limiting case of the sequential model. Unfortunately, the two models have proven very difficult to distinguish experimentally. In Chapter 6 we use these models to investigate allosteric enzymes. Hemoglobin Also Transports H+ and CO2 In addition to carrying nearly all the oxygen required by cells from the lungs to the tissues, hemoglobin carries two end products of cellular respiration — H+ and CO2 — from the tissues to the lungs and the kidneys, where they are excreted. The CO2, produced by oxidation of organic fuels in mitochondria, is hydrated to form bicarbonate: CO2+ H2O ⇌ H+ + HCO−3 This reaction is catalyzed by carbonic anhydrase, an enzyme particularly abundant in erythrocytes. Carbon dioxide is not very soluble in aqueous solution, and bubbles of CO2 would form in the tissues and blood if it were not converted to bicarbonate. As you can see from the reaction catalyzed by carbonic anhydrase, the hydration of CO2 results in an increase in the H+ concentration (a decrease in pH) in the tissues. The binding of oxygen by hemoglobin is profoundly influenced by pH and CO2 concentration, so the interconversion of CO2 and bicarbonate is of great importance to the regulation of oxygen binding and release in the blood. Hemoglobin transports about 40% of the total H+ and 15% to 20% of the CO2 formed in the tissues to the lungs and kidneys. (The remainder of the H+ is absorbed by the plasma’s bicarbonate buffer; the remainder of the CO2 is transported as dissolved HCO− 3 and CO2.) The structural effects of H+ and CO2 binding to hemoglobin favor the T state. Thus, the binding of H+ and CO2 is inversely related to the binding of oxygen. At the relatively low pH and high CO2 concentration of peripheral tissues, the affinity of hemoglobin for oxygen decreases as H+ and CO2 are bound, and O2 is released to the tissues. Conversely, in the capillaries of the lung, as CO2 is excreted and the blood pH consequently rises, the affinity of hemoglobin for oxygen increases and the protein binds more O2 for transport to the peripheral tissues. This effect of pH and CO2 concentration on the binding and release of oxygen by hemoglobin is called the Bohr effect, aer Christian Bohr, the Danish physiologist (and father of physicist Niels Bohr) who discovered it in 1904. A complete statement of the binding equilibrium for hemoglobin and one molecule of oxygen can be designated by the reaction HHb+ + O2⇌ HbO2+ H+ where HHb+ denotes a protonated form of hemoglobin. This equation tells us that the O2-saturation curve of hemoglobin is influenced by the H+ concentration (Fig. 5-15). Both O2 and H+ are bound by hemoglobin, but with inverse affinity. When the oxygen concentration is high, as in the lungs, hemoglobin binds O2 and releases protons. When the oxygen concentration is low, as in the peripheral tissues, H+ is bound and O2 is released. FIGURE 5-15 Effect of pH on oxygen binding to hemoglobin. The pH of blood is 7.6 in the lungs and 7.2 in the tissues. Experimental measurements on hemoglobin binding are o en performed at pH 7.4. Oxygen and H+ are not bound at the same sites in hemoglobin. Oxygen binds to the iron atoms of the hemes, whereas H+ binds to any of several amino acid residues in the protein. A major contribution to the Bohr effect is made by His146 (His HC3) of the β subunits. When protonated, this residue forms one of the ion pairs — to Asp94 (Asp FG1) — that helps stabilize deoxyhemoglobin in the T state (Fig. 5-9). Protonation of the amino-terminal residues of the α subunits, certain other His residues, and perhaps other groups has a similar effect. Hemoglobin also binds CO2, again in a manner inversely related to the binding of oxygen. Carbon dioxide binds as a carbamate group to the α -amino group at the amino-terminal end of each globin chain, forming carbaminohemoglobin: This reaction produces H+, contributing to the Bohr effect. The bound carbamates also form additional salt bridges (not shown in Fig. 5-9) that help to stabilize the T state and promote the release of oxygen. When the concentration of carbon dioxide is high, as in peripheral tissues, some CO2 binds to hemoglobin and the affinity for O2 decreases, causing its release. Conversely, when hemoglobin reaches the lungs, the high oxygen concentration promotes binding of O2 and release of CO2. It is the capacity to communicate ligand-binding information from one polypeptide subunit to the others that makes the hemoglobin molecule so beautifully adapted to integrating the transport of O2, CO2, and H+ by erythrocytes. Oxygen Binding to Hemoglobin Is Regulated by 2,3- Bisphosphoglycerate The interaction of 2,3-bisphosphoglycerate (BPG) with hemoglobin molecules further refines the function of hemoglobin, and provides an example of heterotropic allosteric modulation. BPG binds at a site distant from the oxygen-binding site and regulates the O2-binding affinity of hemoglobin in relation to the pO2 in the lungs. BPG is present in relatively high concentrations in erythrocytes. When hemoglobin is isolated, it contains substantial amounts of bound BPG, which can be difficult to remove completely. In fact, the O2-binding curves for hemoglobin that we have examined up to this point were obtained in the presence of bound BPG. 2,3-Bisphosphoglycerate greatly reduces the affinity of hemoglobin for oxygen — there is an inverse relationship between the binding of O2 and the binding of BPG. We can therefore describe another binding process for hemoglobin: HbBPG + O2⇌ HbO2+ BPG BPG is important in the physiological adaptation to the lower pO2 at high altitudes. For a healthy human at sea level, the binding of O2 to hemoglobin is regulated such that the amount of O2 delivered to the tissues is nearly 40% of the maximum that could be carried by the blood (Fig. 5-16). Imagine that this person is suddenly transported from sea level to an altitude of 4,500 meters, where the pO2 is considerably lower. The delivery of O2 to the tissues is now reduced. However, aer just a few hours at the higher altitude, the BPG concentration in the blood has begun to rise, leading to a decrease in the affinity of hemoglobin for oxygen. This adjustment in the BPG level has only a small effect on the binding of O2 in the lungs, but it has a considerable effect on the release of O2 in the tissues. As a result, the delivery of oxygen to the tissues is restored to nearly 40% of the O2 that can be transported by the blood. The situation is reversed when the person returns to sea level. The BPG concentration in erythrocytes also increases in people suffering from hypoxia, lowered oxygenation of peripheral tissues due to inadequate functioning of the lungs or circulatory system. FIGURE 5-16 Effect of 2,3-bisphosphoglycerate on oxygen binding to hemoglobin. The BPG concentration in normal human blood is about 5 m at sea level and about 8 m at high altitudes. Note that hemoglobin binds to oxygen quite tightly when BPG is entirely absent, and the binding curve seems to be hyperbolic. In reality, the measured Hill coefficient for O2- binding cooperativity decreases only slightly (from 3 to about 2.5) when BPG is removed from hemoglobin, but the rising part of the sigmoid binding curve is confined to a very small region close to the origin. At sea level, hemoglobin is nearly saturated with O2 in the lungs, but is just over 60% saturated in the tissues, so the amount of O2 released in the tissues is about 38% of the maximum that can be carried in the blood. At high altitudes, O2 delivery declines by about one-fourth, to 30% of maximum. An increase in BPG concentration, however, decreases the affinity of hemoglobin for O2, so approximately 37% of what can be carried is again delivered to the tissues. The site of BPG binding to hemoglobin is the cavity between the β subunits in the T state (Fig. 5-17). This cavity is lined with positively charged amino acid residues that interact with the negatively charged groups of BPG. Unlike O2, only one molecule of BPG is bound to each hemoglobin tetramer. BPG lowers hemoglobin’s affinity for oxygen by stabilizing the T state. In the absence of BPG, hemoglobin is primarily present in the R state, where it binds O2 efficiently in the lungs but fails to release it in the tissues.

FIGURE 5-17 Binding of 2,3-bisphosphoglycerate to deoxyhemoglobin. (a) BPG binding stabilizes the T state of deoxyhemoglobin. The negative charges of BPG interact with several positively charged groups (shown in blue in this surface contour image) that surround the pocket between the β subunits on the surface of deoxyhemoglobin in the T state. (b) The binding pocket for BPG disappears on oxygenation, following transition to the R state. (Compare with Fig. 5-10.) [Data from (a) PDB ID 1B86, V. Richard et al., J. Mol. Biol. 233:270, 1993; (b) PDB ID 1BBB, M. M. Silva et al., J. Biol. Chem. 267:17,248, 1992.] Regulation of oxygen binding to hemoglobin by BPG has an important role in fetal development. Because a fetus must extract oxygen from its mother’s blood, fetal hemoglobin must have greater affinity than the maternal hemoglobin for O2. The fetus synthesizes γ subunits rather than β subunits, forming α2γ2 hemoglobin. This tetramer has a much lower affinity for BPG than normal adult hemoglobin, and a correspondingly higher affinity for O2. Sickle Cell Anemia Is a Molecular Disease of Hemoglobin The hereditary human disease sickle cell anemia demonstrates strikingly the importance of amino acid sequence in determining the secondary, tertiary, and quaternary structures of globular proteins, and thus their biological functions. Almost 500 genetic variants of hemoglobin are known to occur in the human population; all but a few are quite rare. Most variations consist of differences in a single amino acid residue. The effects on hemoglobin structure and function are oen minor but can sometimes be extraordinary. Each hemoglobin variation is the product of an altered gene. Variant genes are called alleles. Because humans generally have two copies of each gene, an individual may have two copies of one allele (thus being homozygous for that gene) or one copy of each of two different alleles (thus heterozygous). Sickle cell anemia occurs in individuals who inherit the allele for sickle cell hemoglobin from both parents. The erythrocytes of these individuals are fewer and also abnormal. In addition to an unusually large number of immature cells, the blood contains many long, thin, sickle-shaped erythrocytes (Fig. 5-18). When hemoglobin from sickle cells (called hemoglobin S, or HbS) is deoxygenated, it becomes insoluble and forms polymers that aggregate into tubular fibers (Fig. 5-19). Normal hemoglobin (hemoglobin A, or HbA) remains soluble upon deoxygenation. The insoluble fibers of deoxygenated HbS cause the deformed, sickle shape of the erythrocytes, and the proportion of sickled cells increases greatly as blood is deoxygenated. FIGURE 5-18 A comparison of (a) uniform, cup-shaped, normal erythrocytes and (b) the variably shaped erythrocytes seen in sickle cell anemia, which range from normal to spiny or sickle-shaped.

FIGURE 5-19 Normal and sickle cell hemoglobin. (a) Subtle differences between the conformations of HbA and HbS result from a single amino acid change in the β chains. (b) As a result of this change, deoxyHbS has a hydrophobic patch on its surface, which causes the molecules to aggregate into strands that align into insoluble fibers. The altered properties of HbS result from a single amino acid substitution, a Val instead of a Glu residue at position 6 in the two β chains. Replacement of the Glu residue by Val creates a “sticky” hydrophobic contact point at position 6 of the β chain, which is on the outer surface of the molecule. These sticky spots cause deoxyHbS molecules to associate abnormally with each other, forming the long, fibrous aggregates characteristic of this disorder. Sickle cell anemia is life-threatening and painful. People with this disease suffer repeated crises brought on by physical exertion. They become weak, dizzy, and short of breath, and they also experience heart murmurs and an increased pulse rate. The hemoglobin content of their blood is only about half the normal value of 15 to 16 g/100 mL, because sickled cells are very fragile and rupture easily; this results in anemia (“lack of blood”). An even more serious consequence is that capillaries become blocked by the long, abnormally shaped cells, causing severe pain and interfering with normal organ function — a major factor in the early death of many people with the disease. Without medical treatment, people with sickle cell anemia usually die in childhood. Curiously, the frequency of the sickle cell allele in populations is unusually high in certain parts of Africa. Investigation into this matter led to the finding that when heterozygous, the allele confers a small but significant resistance to lethal forms of malaria. The heterozygous individuals experience a milder condition called sickle cell trait; only about 1% of their erythrocytes become sickled on deoxygenation. These individuals may live completely normal lives if they avoid vigorous exercise and other stresses on the circulatory system. Natural selection has resulted in an allele population that balances the deleterious effects of the homozygous condition against the resistance to malaria afforded by the heterozygous condition. SUMMARY 5.1 Reversible Binding of a Protein to a Ligand: Oxygen-Binding Proteins Protein function oen entails interactions with other molecules. A protein binds a molecule, known as a ligand, at its binding site. Myoglobin contains a heme prosthetic group, which binds oxygen. Heme consists of a single atom of Fe2+ coordinated within a porphyrin. Globins are a specialized family of transport proteins containing heme; most globins store oxygen. Oxygen binds to myoglobin reversibly. Reversible binding of a ligand to a protein can be described quantitatively by a dissociation constant Kd. For a monomeric protein such as myoglobin, the fraction of binding sites occupied by a ligand is a hyperbolic function of ligand concentration. Proteins may undergo conformational changes when a ligand binds, a process called induced fit. In a multisubunit protein, the binding of a ligand to one subunit may affect ligand binding to other subunits. Ligand binding can be regulated. Hemoglobin transports oxygen in blood. Normal adult hemoglobin has four heme-containing subunits, two α and two β , similar in structure to each other and to myoglobin. Hemoglobin exists in two interchangeable structural states, T and R. The T state is most stable when oxygen is not bound. Oxygen binding promotes transition to the R state. Oxygen binding to hemoglobin is both allosteric and cooperative. As O2 binds to one binding site, the hemoglobin undergoes conformational changes that affect the other binding sites — an example of allosteric behavior. Conformational changes between the T and R states, mediated by subunit-subunit interactions, result in cooperative binding; this is described by a sigmoid binding curve and can be analyzed by a Hill plot. Two major models have been proposed to explain the cooperative binding of ligands to multisubunit proteins: the concerted model and the sequential model. Hemoglobin also binds H+ and CO2, resulting in the formation of ion pairs that stabilize the T state and lessen the protein’s affinity for O2 (the Bohr effect). Oxygen binding to hemoglobin is also modulated by 2,3- bisphosphoglycerate, which binds to and stabilizes the T state. Sickle cell anemia is a genetic disease caused by a single amino acid substitution (Glu6 to Val6) in each β chain of hemoglobin. The change produces a hydrophobic patch on the surface of the hemoglobin that causes the molecules to aggregate into bundles of fibers. This homozygous condition results in serious medical complications. 5.2 Complementary Interactions between Proteins and Ligands: The Immune System and Immunoglobulins We have seen how the conformations of oxygen-binding proteins affect and are affected by the binding of small ligands (O2 or CO) to the heme group. However, most protein-ligand interactions do not involve a prosthetic group. Instead, the binding site for a ligand is more oen like the hemoglobin binding site for BPG — a cle in the protein lined with amino acid residues, arranged to make the binding interaction highly specific. Effective discrimination between ligands is the norm at binding sites, even when the ligands have only minor structural differences. Almost all organisms have an immune system of some type that allows them to respond to challenges presented by environmental pathogens. The emergence of vertebrates about 500 million years ago was accompanied by the evolution of an adaptive immune system based on the generation of large numbers of distinct cell clones, each expressing a protein variant that could recognize and bind to a particular type of chemical structure. The system distinguishes molecular “self” from “nonself” and then destroys what is identified as nonself. In this way, the immune system eliminates viruses, bacteria, and other pathogens and molecules that may pose a threat to the organism. On a physiological level, the immune response is an intricate and coordinated set of interactions among many classes of proteins, molecules, and cell types. At the level of individual proteins, the immune response demonstrates how an acutely sensitive and specific biochemical system is built upon the reversible binding of ligands to proteins. The Immune Response Includes a Specialized Array of Cells and Proteins Immunity is brought about by a variety of leukocytes (white blood cells), including macrophages and lymphocytes, all of which develop from undifferentiated stem cells in the bone marrow. Leukocytes can leave the bloodstream and patrol the tissues, each cell producing one or more proteins capable of recognizing and binding to molecules that might signal an infection. The immune response consists of two complementary systems, the humoral and cellular immune systems. The humoral immune system (Latin humor, “fluid”) is directed at bacterial infections and extracellular viruses (those found in the body fluids), but it can also respond to individual foreign proteins. The cellular immune system destroys host cells infected by viruses and also destroys some parasites and foreign tissues. At the heart of the humoral immune response are soluble proteins called antibodies or immunoglobulins, oen abbreviated Ig. Immunoglobulins bind bacteria, viruses, or large molecules identified as foreign and target them for destruction. Making up 20% of blood protein, the immunoglobulins are produced by B lymphocytes, or B cells, so named because they complete their development in the bone marrow. The agents at the heart of the cellular immune response are a class of T lymphocytes, or T cells (so called because the latter stages of their development occur in the thymus), known as cytotoxic T cells (TC cells, also called killer T cells). Recognition of infected cells or parasites involves proteins called T-cell receptors on the surface of TC cells. Receptors are proteins, usually found on the outer surface of cells and extending through the plasma membrane, that recognize and bind extracellular ligands, thus triggering changes inside the cell. In addition to cytotoxic T cells, there are helper T cells (TH cells), whose function it is to produce soluble signaling proteins called cytokines, which include the interleukins. TH cells interact with macrophages. The TH cells participate only indirectly in the destruction of infected cells and pathogens, stimulating the selective proliferation of those TC and B cells that can bind to a particular antigen. This process, called clonal selection, increases the number of immune system cells that can respond to a particular pathogen. A host organism needs time, oen days, to mount an immune response against a new antigen, but memory cells permit a rapid response to pathogens previously encountered. A vaccine to protect against a particular viral infection oen consists of weakened or killed virus or isolated proteins from a viral or bacterial protein coat. When injected into a person, the vaccine generally does not cause an infection and illness, but it effectively “teaches” the immune system what the viral particles look like, stimulating the production of memory cells. On subsequent infection, these cells can bind to the virus and trigger a rapid immune response. The importance of TH cells is dramatically illustrated by the epidemic produced by HIV (human immunodeficiency virus), the virus that causes AIDS (acquired immune deficiency syndrome). TH cells are the primary targets of HIV infection; elimination of these cells progressively incapacitates the entire immune system. Each recognition protein of the immune system, either a T-cell receptor or an antibody produced by a B cell, specifically binds some particular chemical structure, distinguishing it from virtually all others. Humans are capable of producing more than 108 different antibodies with distinct binding specificities. Given this extraordinary diversity, any chemical structure on the surface of a virus or an invading cell will most likely be recognized and bound by one or more antibodies. Antibody diversity is derived from random reassembly of a set of immunoglobulin gene segments through genetic recombination mechanisms that are discussed in Chapter 25 (see Fig. 25-42). A specialized lexicon is used to describe the unique interactions between antibodies or T-cell receptors and the molecules they bind. Any molecule or pathogen capable of eliciting an immune response is called an antigen. An antigen may be a virus, a bacterial cell wall, or an individual protein or other macromolecule. A complex antigen may be bound by several different antibodies. An individual antibody or T-cell receptor binds only a particular molecular structure within the antigen, called its antigenic determinant or epitope. It would be unproductive for the immune system to respond to small molecules that are common intermediates and products of cellular metabolism. Molecules of Mr< 5,000 are generally not antigenic. However, when small molecules are covalently attached to large proteins in the laboratory, they can be used to elicit an immune response. These small molecules are called haptens. The antibodies produced in response to protein-linked haptens will then bind to the same small molecules in their free form. Such antibodies are sometimes used in the development of analytical tests described later in this chapter or as a specific ligand in affinity chromatography (see Fig. 3-17c). We now turn to a more detailed description of antibodies and their binding properties. Antibodies Have Two Identical Antigen-Binding Sites Immunoglobulin G (IgG) is the major class of antibody molecule and one of the most abundant proteins in the blood serum. IgG has four polypeptide chains: two large ones, called heavy chains, and two smaller ones, called light chains, linked by noncovalent and disulfide bonds into a complex of Mr 150,000. The heavy chains of an IgG molecule interact at one end, then branch to interact separately with the light chains, forming a Y-shaped molecule (Fig. 5-20). At the “hinges” separating the base of an IgG molecule from its branches, the immunoglobulin can be cleaved with proteases. Cleavage with the protease papain liberates the basal fragment, called Fc because it usually crystallizes readily, and the two branches, called Fab, the antigen-binding fragments. Each branch has a single antigen-binding site. FIGURE 5-20 Immunoglobulin G. (a) Pairs of heavy and light chains combine to form a Y- shaped molecule. Two antigen-binding sites are formed by the combination of variable domains from one light (VL) chain and one heavy (VH) chain. Cleavage with papain separates the Fab and Fc portions of the protein in the hinge region. The Fc portion also contains bound carbohydrate (shown in (b)). (b) A ribbon model of an IgG molecule. Although the molecule has two identical heavy chains (two shades of blue) and two identical light chains (two shades of red), it crystallized in the asymmetric conformation shown here. Conformational flexibility is important to the function of immunoglobulins. [(b) Data from PDB ID 1IGT, L. J. Harris et al., Biochemistry 36:1581, 1997.] The fundamental structure of immunoglobulins was first established by Gerald Edelman and Rodney Porter in the 1960s. Each chain is made up of identifiable domains. Some are constant in sequence and structure from one IgG to the next; others are variable. The constant domains have a characteristic structure known as the immunoglobulin fold, a well-conserved structural motif in the all-β class of proteins (Chapter 4). There are three of these constant domains in each heavy chain and one in each light chain. The heavy and light chains also have one variable domain each, in which most of the variability in amino acid sequence is found. The variable domains associate to create the antigen- binding site (Fig. 5-20), allowing formation of an antigen-antibody complex (Fig. 5-21). FIGURE 5-21 Binding of IgG to an antigen. To generate an optimal fit for the antigen, the binding sites of IgG o en undergo slight conformational changes. Such induced fit is common to many protein-ligand interactions. In many vertebrates, IgG is but one of five classes of immunoglobulins. Each class has a characteristic type of heavy chain, denoted α , δ , ε , γ , and μ for IgA, IgD, IgE, IgG, and IgM, respectively. Two types of light chain, κ and λ , occur in all classes of immunoglobulins. The overall structures of IgD and IgE are similar to that of IgG. IgM occurs either in a monomeric, membrane-bound form or in a secreted form that is a cross- linked pentamer of this basic structure (Fig. 5-22). IgA, found principally in secretions such as saliva, tears, and milk, can be a monomer, a dimer, or a trimer. IgM is the first antibody to be made by B lymphocytes and the major antibody in the early stages of a primary immune response. Some B cells soon begin to produce IgD (with the same antigen-binding site as the IgM produced by the same cell), but the particular function of IgD is less clear. FIGURE 5-22 IgM pentamer of immunoglobulin units. The pentamer is cross-linked with disulfide bonds (yellow). The J chain is a polypeptide of Mr 20,000 found in both IgA and IgM. The IgG described above is the major antibody in secondary immune responses, which are initiated by a class of B cells called memory B cells. As part of the organism’s ongoing immunity to antigens already encountered and dealt with, IgG is the most abundant immunoglobulin in the blood. When IgG binds to an invading bacterium or virus, it activates certain leukocytes such as macrophages to engulf and destroy the invader, and it also activates some other parts of the immune response. Receptors on the macrophage surface recognize and bind the Fc region of IgG. When these Fc receptors bind an antibody-pathogen complex, the macrophage engulfs the complex by phagocytosis (Fig. 5-23). FIGURE 5-23 Phagocytosis of an antibody-bound virus by a macrophage. The Fc regions of antibodies bound to the virus now bind to Fc receptors on the surface of a macrophage, triggering the macrophage to engulf and destroy the virus. IgE plays an important role in the allergic response, interacting with basophils (phagocytic leukocytes) in the blood and with histamine-secreting cells called mast cells, which are widely distributed in tissues. This immunoglobulin binds, through its Fc region, to special Fc receptors on the basophils or mast cells. In this form, IgE serves as a receptor for antigen. If antigen is bound, the cells are induced to secrete histamine and other biologically active amines that cause the dilation and increased permeability of blood vessels. These effects on the blood vessels are thought to facilitate the movement of immune system cells and proteins to sites of inflammation. They also produce the symptoms normally associated with allergies. Pollen and other allergens are recognized as foreign, triggering an immune response normally reserved for pathogens. Antibodies Bind Tightly and Specifically to Antigen The binding specificity of an antibody is determined by the amino acid residues in the variable domains of its heavy and light chains. Some of those residues, particularly those lining the antigen-binding site, are hypervariable — especially likely to differ. Specificity is conferred by chemical complementarity between the antigen and its specific binding site. For example, a binding site with a negatively charged group may bind an antigen with a positive charge in the complementary position. In many instances, complementarity is achieved interactively as the structures of antigen and binding site influence each other as they come closer together. Conformational changes in the antibody and/or the antigen then allow the complementary groups to interact fully. This is an example of induced fit. The complex of a peptide derived from HIV (a model antigen) and an Fab molecule, shown in Figure 5-24, illustrates some of these properties. The changes in structure observed on antigen binding are particularly striking in this example. FIGURE 5-24 Induced fit in the binding of an antigen to IgG. The Fab fragment of an IgG molecule is shown here with the surface contour colored to represent hydrophobicity. Hydrophobic surfaces are yellow and hydrophilic surfaces are blue, with shades of blue to green to yellow in between. (a) View of the Fab fragment in the absence of antigen (a small peptide derived from HIV), looking down on the antigen binding site. (b) The same view, but with the Fab fragment in the “bound” conformation with the antigen omitted to provide an unobstructed view of the altered binding site. Note how the hydrophobic binding cavity has enlarged and several groups have shi ed position. (c) The same view as (b) but with the antigen (red) in the binding site. [Data from (a) PDB ID 1GGC, R. L. Stanfield et al., Structure 1:83, 1993; (b, c) PDB ID 1GGI, J. M. Rini et al., Proc. Natl. Acad. Sci. USA 90:6325, 1993.] A typical antibody-antigen interaction is quite strong, characterized by Kd values as low as 10−10 M (recall that a lower Kd corresponds to a stronger binding interaction; see Table 5-1). The Kd reflects the energy derived from the hydrophobic effect and the various ionic, hydrogen-bonding, and van der Waals interactions that stabilize the binding. The binding energy required to produce a Kd of 10−10 M is about 65 kJ/mol. The Antibody-Antigen Interaction Is the Basis for a Variety of Important Analytical Procedures The extraordinary binding affinity and specificity of antibodies make them valuable analytical reagents. Two types of antibody preparations are in use: polyclonal and monoclonal. Polyclonal antibodies are those produced by many different B lymphocytes responding to one antigen, such as a protein injected into an animal. Cells in the population of B lymphocytes produce antibodies that bind specific, different epitopes within the antigen. Thus, polyclonal preparations contain a mixture of antibodies that recognize different parts of the protein. Monoclonal antibodies, in contrast, are synthesized by a population of identical B cells (a clone) grown in cell culture. These antibodies are homogeneous, all recognizing the same epitope. The specificity of antibodies has practical uses. In a versatile analytical technique, an antibody is attached to a reagent that makes it easy to detect (Fig. 5-25). When the antibody binds the target protein, the label reveals the presence of the protein in a solution or its location in a gel, or even in a living cell (Fig. 5-25a). In one application of this technique, an immunoblot or Western blot assay (Fig. 5-26b), proteins that have been separated by gel electrophoresis are transferred electrophoretically to a nitrocellulose membrane. Aer washing, the membrane is treated successively with primary antibody, secondary antibody linked to enzyme, and substrate. A colored precipitate forms only along the band containing the protein of interest. Immunoblotting allows the detection of a minor component in a sample and provides an approximation of its molecular weight. FIGURE 5-25 Antibodies as analytical reagents. (a) A schematic representation of the use of the specific reaction of an antibody with its antigen to identify and quantify a specific protein in a complex sample. (b) One common application is the immunoblot. Lanes 1 to 3 are from an SDS gel; samples from successive stages in the purification of a protein kinase were separated and stained with Coomassie blue. Lanes 4 to 6 show the same samples, but these were electrophoretically transferred to a nitrocellulose membrane a er separation on an SDS gel. The membrane was then “probed” with antibody against the protein kinase. The numbers between the SDS gel and the immunoblot indicate Mr in thousands. We will encounter other aspects of antibodies in later chapters. They are extremely important in medicine and can tell us much about the structure of proteins and the action of genes. SUMMARY 5.2 Complementary Interactions between Proteins and Ligands: The Immune System and Immunoglobulins The immune response is mediated by interactions among an array of specialized leukocytes and their associated proteins. T lymphocytes produce T-cell receptors. B lymphocytes produce immunoglobulins. Humans have five classes of immunoglobulins, each with different biological functions. The most abundant class is IgG, a Y-shaped protein with two heavy chains and two light chains. The domains near the upper ends of the Y are hypervariable within the broad population of IgGs and form two antigen-binding sites. A given immunoglobulin generally binds to only a part, called the epitope, of a large antigen. Binding oen involves a conformational change in the IgG, an induced fit to the antigen. The exquisite binding specificity of immunoglobulins is exploited in analytical techniques such as immunoblotting. 5.3 Protein Interactions Modulated by Chemical Energy: Actin, Myosin, and Molecular Motors Organisms move. Cells move. Organelles and macromolecules within cells move. Most of these movements arise from the activity of a fascinating class of protein-based molecular motors. Fueled by chemical energy, usually derived from ATP, large aggregates of motor proteins undergo cyclic conformational changes that accumulate into a unified, directional force — the tiny force that pulls apart chromosomes in a dividing cell, and the immense force that levers a pouncing, quarter-ton jungle cat into the air. The interactions among motor proteins, as you might predict, feature complementary arrangements of ionic, hydrogen- bonding, and hydrophobic groups at protein binding sites. In motor proteins, however, the resulting interactions achieve exceptionally high levels of spatial and temporal organization. Motor proteins underlie the migration of organelles along microtubules, the motion of eukaryotic and bacterial flagella, and the movement of some proteins along DNA. Here, we focus on the well-studied example of the contractile proteins of vertebrate skeletal muscle as a paradigm for how proteins translate chemical energy into motion. The Major Proteins of Muscle Are Myosin and Actin The contractile force of muscle is generated by the interaction of two proteins, myosin and actin. These proteins are arranged in filaments that undergo transient interactions and slide past each other to bring about contraction. Together, actin and myosin make up more than 80% of the protein mass of muscle. Myosin (Mr 520,000) has six subunits: two heavy chains (each of Mr 220,000) and four light chains (each of Mr 20,000). The heavy chains account for much of the overall structure. At their carboxyl termini, they are arranged as extended α helices, wrapped around each other in a fibrous, le-handed coiled coil similar to that of α -keratin (Fig. 5-26). At its amino terminus, each heavy chain has a large globular domain containing a site where ATP is hydrolyzed. The light chains are associated with the globular domains. FIGURE 5-26 Myosin. (a) Myosin has two heavy chains: the carboxyl termini forming an extended coiled coil tail and the amino termini having clublike globular head domains. Two myosin light chains are associated with each myosin head. (b) A three-dimensional representation of the myosin head, showing the binding sites for nucleotide (ATP) and actin. [(a) Research from Takeshi Katayama, et al. “Stimulatory effects of arachidonic acid on myosin ATPase activity and contraction of smooth muscle via myosin motor domain,” Am. J. Physiol. Heart Circ. Physiol. Vol 298, Issue 2, pp. H505–H514, February 2010, Fig. 6b. (b) Data from PDB ID 2MYS, I. Rayment et al., Science 261:50, 1993.] In muscle cells, molecules of myosin aggregate to form structures called thick filaments (Fig. 5-27a). These rodlike structures are the core of the contractile unit. Within a thick filament, several hundred myosin molecules are arranged with their fibrous “tails” associated to form a long, bipolar structure. The globular domains project from either end of this structure, in regular stacked arrays. FIGURE 5-27 The major components of muscle. (a) Myosin aggregates to form a bipolar structure called a thick filament. (b) F-actin is a filamentous assemblage of G-actin monomers that polymerize two by two, giving the appearance of two filaments spiraling about one another in a right-handed fashion. (c) Space-filling model of an actin filament with one myosin head bound to an actin monomer within the filament. [(c) Data from PDB ID 2MYS, I. Rayment et al., Science 261:50, 1993; PDB ID 6BNQ, P. S. Gurel et al., eLife 6, 2017, doi 10.7554/eLife.31125] The second major muscle protein, actin, is abundant in almost all eukaryotic cells. In muscle, molecules of monomeric actin, called G-actin (globular actin; Mr 42,000), associate to form a long polymer called F-actin (filamentous actin). The thin filament consists of F-actin (Fig. 5-27b), along with the proteins troponin and tropomyosin (discussed below). The filamentous parts of thin filaments assemble as successive monomeric actin molecules add to one end. Upon addition, each monomer binds ATP, then hydrolyzes it to ADP, so every actin molecule in the filament is complexed to ADP. This ATP hydrolysis by actin functions only in the assembly of the filaments; it does not contribute directly to the energy expended in muscle contraction. Each actin monomer in the thin filament can bind tightly and specifically to one myosin head group (Fig. 5-27c). Additional Proteins Organize the Thin and Thick Filaments into Ordered Structures Skeletal muscle consists of parallel bundles of muscle fibers, each fiber a single, very large, multinucleated cell, 20 to 100 μ m in diameter, formed from many cells fused together; a single fiber oen spans the length of the muscle. Each fiber contains about 1,000 myofibrils, 2 μ m in diameter, each consisting of a vast number of regularly arrayed thick and thin filaments complexed to other proteins (Fig. 5-28). A system of flat membranous vesicles called the sarcoplasmic reticulum surrounds each myofibril. Examined under the electron microscope, muscle fibers reveal alternating regions of high electron density and low electron density, called the A bands and I bands (Fig. 5-28b, c). The A and I bands arise from the arrangement of thick and thin filaments, which are aligned and partially overlapping. The I band is the region of the bundle that in cross section would contain only thin filaments. The darker A band stretches the length of the thick filament and includes the region where parallel thick and thin filaments overlap. Bisecting the I band is a thin structure called the Z disk, perpendicular to the thin filaments and serving as an anchor to which the thin filaments are attached. The A band, too, is bisected by a thin line, the M line or M disk, a region of high electron density in the middle of the thick filaments. The entire contractile unit, consisting of bundles of thick filaments interleaved at either end with bundles of thin filaments, is called the sarcomere. The arrangement of interleaved bundles allows the thick and thin filaments to slide past each other (by a mechanism discussed below), causing a progressive shortening of each sarcomere (Fig. 5-28d). FIGURE 5-28 Skeletal muscle. (a) Muscle fibers consist of single, elongated, multinucleated cells that arise from the fusion of many precursor cells. The fibers are made up of many myofibrils (only six are shown here for simplicity) surrounded by the membranous sarcoplasmic reticulum. The organization of thick and thin filaments in a myofibril gives it a striated appearance. When muscle contracts, the I bands narrow and the Z disks move closer together, as seen in electron micrographs of (b) relaxed muscle and (c) contracted muscle. (d) Thick filaments are bipolar structures created by the association of many myosin molecules. Muscle contraction occurs by the sliding of the thick and thin filaments past each other so that the Z disks in neighboring I bands draw closer together. The thick and thin filaments are interleaved such that each thick filament is surrounded by six thin filaments. Myosin Thick Filaments Slide along Actin Thin Filaments The interaction between actin and myosin, like that between all proteins and ligands, involves weak bonds. When ATP is not bound to myosin, a face on the myosin head group binds tightly to actin (Fig. 5-29). When ATP binds to myosin and is hydrolyzed to ADP and phosphate, a coordinated and cyclic series of conformational changes occurs in which myosin releases the F- actin subunit and binds another subunit farther along the thin filament.

FIGURE 5-29 Molecular mechanism of muscle contraction. Conformational changes in the myosin head that are coupled to stages in the ATP hydrolytic cycle cause myosin to successively dissociate from one actin subunit, then associate with another farther along the actin filament. In this way, the myosin heads slide along the thin filaments, drawing the thick filament array into the thin filament array (see Fig. 5-28). The cycle has four major steps (Fig. 5-29). In step , ATP binds to myosin and a cle in the myosin molecule opens, disrupting the actin-myosin interaction so that the bound actin is released. ATP is then hydrolyzed in step , causing a conformational change in the protein to a “high-energy” state that moves the myosin head and changes its orientation in relation to the actin thin filament. Myosin then binds weakly to an F-actin subunit closer to the Z disk than the one just released. As the phosphate product of ATP hydrolysis is released from myosin in step , another conformational change occurs in which the myosin cle closes, strengthening the myosin-actin binding. This is followed quickly by step , a “power stroke” during which the conformation of the myosin head returns to the original resting state, its orientation relative to the bound actin changing so as to pull the tail of the myosin toward the Z disk. ADP is then released to complete the cycle. Each cycle generates about 3 to 4 pN (piconewtons) of force and moves the thick filament 5 to 10 nm relative to the thin filament. Because there are many myosin heads in a thick filament, at any given moment some (probably 1% to 3%) are bound to thin filaments. This prevents thick filaments from slipping backward when an individual myosin head releases the actin subunit to which it was bound. The thick filament thus actively slides forward past the adjacent thin filaments. This process, coordinated among the many sarcomeres in a muscle fiber, brings about muscle contraction. The interaction between actin and myosin must be regulated so that contraction occurs only in response to appropriate signals from the nervous system. The regulation is mediated by a complex of two proteins, tropomyosin and troponin (Fig. 5-30). Tropomyosin binds to the thin filament, blocking the attachment sites for the myosin head groups. Troponin is a Ca2+-binding protein. A nerve impulse causes release of Ca2+ ions from the sarcoplasmic reticulum. The released Ca2+ binds to troponin (another protein-ligand interaction) and causes a conformational change in the tropomyosin-troponin complexes, exposing the myosin-binding sites on the thin filaments. Contraction follows. FIGURE 5-30 Regulation of muscle contraction by tropomyosin and troponin. Tropomyosin and troponin are bound to F-actin in the thin filaments. In the relaxed muscle, these two proteins are arranged around the actin filaments so as to block the binding sites for myosin. Tropomyosin is a two-stranded coiled coil of α helices, the same structural motif as in α -keratin (see Fig. 4-11). It forms head-to-tail polymers twisting around the two actin chains. Troponin is attached to the actin-tropomyosin complex at regular intervals of 38.5 nm. Troponin consists of three different subunits: I, C, and T. Troponin I prevents binding of the myosin head to actin; troponin C has a binding site for Ca2+; and troponin T links the entire troponin complex to tropomyosin. When the muscle receives a neural signal to initiate contraction, Ca2+ is released from the sarcoplasmic reticulum (see Fig. 5-28a) and binds to troponin C. This causes a conformational change in troponin C, which alters the positions of troponin I and tropomyosin so as to relieve the inhibition by troponin I and allow muscle contraction. Working skeletal muscle requires two types of molecular functions that are common in proteins: binding and catalysis. The actin-myosin interaction, a protein-ligand interaction like that of immunoglobulins with antigens, is reversible and leaves the participants unchanged. This interaction illustrates why reversibility is important; a permanent actin-myosin interaction defines the state of rigor mortis, a state we all want to avoid as long as possible. When ATP binds myosin, however, it is hydrolyzed to ADP and Pi (inorganic phosphate). Myosin is not only an actin-binding protein, but also an ATPase — an enzyme. The function of enzymes in catalyzing chemical transformations is the topic of the next chapter. SUMMARY 5.3 Protein Interactions Modulated by Chemical Energy: Actin, Myosin, and Molecular Motors Protein-ligand interactions achieve a special degree of spatial and temporal organization in motor proteins. Muscle contraction results from choreographed interactions between myosin and actin, coupled to the hydrolysis of ATP by myosin. Myosin consists of two heavy chains and four light chains, forming a fibrous coiled coil (tail) domain and a globular (head) domain. Myosin molecules are organized into thick filaments, which slide past thin filaments composed largely of actin. ATP hydrolysis in myosin is coupled to a series of conformational changes in the myosin head, leading to dissociation of myosin from one F-actin subunit and its eventual reassociation with another, farther along the thin filament. The myosin thus slides along the actin filaments. Muscle contraction is stimulated by the release of Ca2+ from the sarcoplasmic reticulum. The Ca2+ binds to the protein troponin, leading to a conformational change in a troponin- tropomyosin complex that triggers the cycle of actin-myosin interactions. Chapter Review KEY TERMS Terms in bold are defined in the glossary. ligand binding site induced fit hemoglobin heme porphyrin globins globin fold equilibrium expression association constant, Ka dissociation constant, Kd allosteric protein modulator homotropic heterotropic Hill equation Hill coefficient Bohr effect immune response lymphocytes antibody immunoglobulin (Ig) B lymphocyte (B cell) T lymphocyte (T cell) antigen epitope hapten immunoglobulin fold polyclonal antibodies monoclonal antibodies immunoblotting Western blotting myosin actin myofibril sarcomere PROBLEMS 1. Relationship between Affinity and Dissociation Constant Protein A has a binding site for ligand X with a Kd of 3.0× 10–7M. Protein B has a binding site for ligand X with a Kd of 4.0× 10–8M. Calculate the Ka for each protein. Which protein has a higher affinity for ligand X? Explain your reasoning. 2. Modeling Apparent Negative Cooperativity Which of these situations would produce a Hill plot with nH < 1.0? Explain your reasoning in each case. a. The protein has multiple subunits, each with a single ligand-binding site. Ligand binding to one site decreases the binding affinity of other sites for the ligand. b. The protein is a single polypeptide with two ligand- binding sites, each having a different affinity for the ligand. c. The protein is a single polypeptide with a single ligand-binding site. As purified, the protein preparation is heterogeneous, containing some protein molecules that are partially denatured and thus have a lower binding affinity for the ligand. d. The protein has multiple subunits, each with a single ligand-binding site. Ligands bind independently to each site, do not affect the binding affinity of other sites, and bind with identical affinities. 3. Reversible Ligand Binding I The protein calcineurin binds to the protein calmodulin with an association rate of 8.9× 103M –1s–1 and an overall dissociation constant, Kd, of 10 nM. Calculate the dissociation rate, kd, including appropriate units. 4. Reversible Ligand Binding II The E. coli nickel-binding protein binds to its ligand, Ni2+, with a Kd of 100 nM. Calculate the Ni2+ concentration when the fraction of binding sites occupied by the ligand (Y) is (a) 0.25, (b) 0.6, (c) 0.95. 5. Reversible Ligand Binding III You are a technician in a biochemistry lab running receptor binding experiments. The target membrane-bound receptor has been partially purified from mouse, rat, and human cell lines. Using various concentrations of the same radioactive ligand for each receptor in a saturation binding assay, you generate the binding data shown in the table. The dependent variable, Y, is the fraction of binding sites occupied by the ligand. Ligand concentration (nM) Y Mouse receptor Rat receptor Human receptor       0.2       0.048 0.29 0.17       0.5       0.11 0.50 0.33       1.0       0.20 0.67 0.50       4.0       0.50 0.89 0.80 10       0.71 0.95 0.91 20       0.83 0.97 0.95 50       0.93 0.99 0.98 a. Determine the mouse receptor Kd in this binding experiment. b. Which receptor binds most tightly to this ligand? 6. Reversible Ligand Binding IV Exposure to carbon monoxide can lead to unconsciousness and ultimately death. Suffocation occurs when hemoglobin is half-saturated with CO — that is, when only two of every four oxygen-binding sites are occupied with CO. Explain why death may occur at this point, even though half of the oxygen-binding sites are still available to transport O2. (Hint: See Box 5-1.) 7. Cooperativity in Hemoglobin Under appropriate conditions, hemoglobin dissociates into its four subunits. The isolated α subunit binds oxygen, but the O2-saturation curve is hyperbolic rather than sigmoid. In addition, the binding of oxygen to the isolated α subunit is not affected by the presence of H+, CO2, or BPG. What do these observations indicate about the source of the cooperativity in hemoglobin? 8. Oxygen Binding to Hemoglobin The solid curve in the plot shown is an O2-binding curve for human hemoglobin. For each condition, indicate whether the stated physiological change would shi the curve to the le (dashed curve), produce no change (black curve), or shi the curve to the right (dashed curve). a. an increase in the concentration of CO2 b. an increase in the proton concentration (decrease in pH) c. an increase in the concentration of 2,3- bisphosphoglycerate (BPG) 9. Comparison of Fetal and Maternal Hemoglobins Studies of oxygen transport in pregnant mammals show that the O2- saturation curves of fetal and maternal blood are markedly different when measured under the same conditions. Fetal erythrocytes contain a structural variant of hemoglobin, HbF, consisting of two α and two γ subunits (α2γ2), whereas maternal erythrocytes contain HbA (α2β2). a. Which hemoglobin has a higher affinity for oxygen under physiologic conditions? b. What is the physiological significance of the different O2 affinities? When all the BPG is carefully removed from samples of HbA and HbF, the measured O2-saturation curves (and consequently the O2 affinities) are displaced to the le. However, HbA now has a greater affinity for oxygen than does HbF. When BPG is reintroduced, the O2-saturation curves return to normal, as shown in the graph. c. What is the effect of BPG on the O2 affinity of hemoglobin? How can this information be used to explain the different O2 affinities of fetal and maternal hemoglobin? 10. Hemoglobin Variants There are almost 500 naturally occurring variants of hemoglobin. Most are the result of a single amino acid substitution in a globin polypeptide chain. Some variants produce clinical illness, though not all variants have deleterious effects. A brief sample of hemoglobin variants is shown here. HbS (sickle cell Hb): substitutes a Val for a Glu on the surface Hb Cowtown: eliminates an ion pair involved in T-state stabilization Hb Memphis: substitutes one uncharged polar residue for another of similar size on the surface Hb Bibba: substitutes a Pro for a Leu involved in an α helix Hb Milwaukee: substitutes a Glu for a Val Hb Providence: substitutes an Asn for a Lys that normally projects into the central cavity of the tetramer Hb Philly: substitutes a Phe for a Tyr, disrupting hydrogen bonding at the α1β1 interface Select the hemoglobin variants that are described by each statement. a. The Hb variant least likely to cause pathological symptoms b. The variant(s) most likely to show pI values different from that of HbA on an isoelectric focusing gel c. The variant(s) most likely to show a decrease in BPG binding and an increase in the overall affinity of the hemoglobin for oxygen 11. Oxygen Binding and Hemoglobin Structure A team of biochemists uses genetic engineering to modify the interface region between hemoglobin subunits. The resulting hemoglobin variants exist in solution primarily as αβ dimers (few, if any, α2β2 tetramers form). Are these variants likely to bind oxygen more weakly or more tightly? Explain your answer. 12. Reversible (and Tight) Binding to an Antibody An antibody with high affinity for its antigen has a Kd in the low nanomolar range. Assume an antibody binds an antigen with a Kd of 5× 10–8 M. Calculate the antigen concentration when Y, the fraction of binding sites occupied by the ligand, is a. 0.4, b. 0.5, c. 0.8, d. 0.9. 13. Using Antibodies to Probe Structure-Function Relationships in Proteins A monoclonal antibody binds to G- actin but not to F-actin. What does this tell you about the epitope recognized by the antibody? 14. The Immune System and Vaccines Some pathogens have developed mechanisms to evade the immune system, making it difficult or impossible to develop effective vaccines against them. a. African sleeping sickness is caused by a protozoan called Trypanosoma brucei, carried by the tsetse fly. The trypanosome surface is dominated by one coat protein, the variable surface glycoprotein (VSG). The trypanosome genome encodes over 1,000 different versions of VSG. All of the cells in an initial infection feature the same VSG coat on their surfaces, and this is readily recognized as foreign by the immune system. However, an individual trypanosome in the broader population will switch and randomly begin expressing a different variant of the VSG coat. All the descendants of that cell will have the new and different protein on their surface. As the population with the second VSG coat increases, an individual cell will then switch to a third VSG protein coat, and so on. b. The human immunodeficiency virus (HIV) has an error-prone system for replicating its genome, effectively introducing mutations at an unusually high rate. Many of the mutations affect the viral protein coat. Describe how each pathogen can survive the immune response of its host. 15. How We Become a “Stiff” When a vertebrate dies, its muscles stiffen as they are deprived of ATP, a state called rigor mortis. Using your knowledge of the catalytic cycle of myosin in muscle contraction, explain the molecular basis of the rigor state. 16. Sarcomeres from Another Point of View The symmetry of thick and thin filaments in a sarcomere is such that six thin filaments ordinarily surround each thick filament in a hexagonal array. Match each cross section (transverse cut) image of a sarcomere with the correct viewpoint. a. at the M line b. through the I band c. through the dense region of the A band d. through the less dense region of the A band, adjacent to the M line (see Fig. 5-28b, c) BIOCHEMISTRY ONLINE 17. IgG and Lysozyme Crystal Structure To fully appreciate how proteins function in a cell, it is helpful to have a three- dimensional view of how proteins interact with other cellular components. Fortunately, this is possible using online protein databases and three-dimensional molecular viewing utilities such as JSmol, a free and user-friendly molecular viewer that is compatible with most browsers and operating systems. In this exercise, examine the interactions between the enzyme lysozyme and the Fab portion of the antilysozyme antibody. Use the PDB identifier 1FDL to explore the structure of the IgG1 Fab fragment–lysozyme complex (antibody-antigen complex). To answer the questions, use the information on the Structure Summary page at the Protein Data Bank (www.rcsb.org), and view the structure using JSmol or a similar viewer. a. Which chains in the three-dimensional model correspond to the antibody fragment, and which correspond to the antigen, lysozyme? b. What type of secondary structure predominates in this Fab fragment? c. How many amino acid residues are in the heavy and light chains of the Fab fragment? In lysozyme? Estimate the percentage of the lysozyme that interacts with the antigen-binding site of the antibody fragment. d. Identify the specific amino acid residues in lysozyme and in the variable regions of the Fab heavy and light chains that are situated at the antigen-antibody interface. Are the residues contiguous in the primary sequence of the polypeptide chains? 18. Exploring Antibodies in the Protein Data Bank The PDB- 101 Molecule of the Month article on “Antibodies” (http://pdb101.rcsb.org/motm/21) summarizes what you have read in this chapter regarding antibody structure and function. To paraphrase the article, a variety of antibodies, on the order of one hundred million different types, are always circulating in our bloodstream, searching for foreign invaders to attack. Once an invader is discovered, the antibody binds the invader with its flexible arms, containing the Fab region. Thin, flexible chains connect these flexible arms to the antibody base, called the Fc region. This base determines which class the antibody belongs to, as some antibodies have four or ten binding sites due to their structural formation. a. How many specific antigen-binding sites are there on the first immunoglobulin image (derived from PDB ID 1IGT) in the article? b. When a virus enters your lungs, how long does it take for you to produce one or more antibodies that bind to it? c. Approximately how many types of different antibodies are present in your blood? d. Explore the structure of the immunoglobulin molecule (PDB ID 1IGT) by clicking the link in the article or by using a search engine to find the structure summary for PDB ID 1IGT. Use one of the 3D viewers on the PDB site to view a ribbon structure for this immunoglobulin. Identify the two light chains and two heavy chains (use the viewer controls to distinguish them by color). DATA ANALYSIS PROBLEM 19. Protein Function During the 1980s, the structures of actin and myosin were known only at the resolution shown in Figure 5-26a. Although researchers knew that the globular head portion of myosin bound to actin and hydrolyzed ATP, there was a substantial debate about where in the myosin molecule the contractile force was generated. At the time, two competing models were proposed for the mechanism of force generation in myosin. In the “hinge” model, the head bound to actin, but the pulling force was generated by contraction of the “hinge region” in the myosin tail. The hinge region is in the heavy meromyosin portion of the myosin molecule; this is roughly the point labeled “Two supercoiled α helices” in Figure 5-26. In the “S1” model (S1 being a name used to describe the head), the pulling force was generated in the S1 “head” itself and the tail was just for structural support. Many experiments were performed but provided no conclusive evidence. Then, in 1987, James Spudich and his colleagues at Stanford University published a study that, although not conclusive, went a long way toward resolving this controversy. Recombinant DNA techniques were not sufficiently developed to address this issue in vivo, so Spudich and colleagues used an interesting in vitro motility assay. The alga Nitella has extremely long cells, oen several centimeters long and about 1 mm in diameter. These cells have actin fibers that run along their long axes, and the cells can be cut open along their length to expose the actin fibers. Spudich and his group had observed that plastic beads coated with myosin would “walk” along these fibers in the presence of ATP, just as myosin would do in contracting muscle. For these experiments, the researchers used a more well- defined method for attaching the myosin to the beads. The “beads” were clumps of killed bacterial (Staphylococcus aureus) cells. These cells have a protein on their surface that binds to the Fc region of antibody molecules (Fig. 5-20a). The antibodies, in turn, bind to several (unknown) places along the tail of the myosin molecule. When bead-antibody-myosin complexes were prepared with intact myosin molecules, they would move along Nitella actin fibers in the presence of ATP. a. Sketch a diagram showing what a bead-antibody- myosin complex might look like at the molecular level. b. Why was ATP required for the beads to move along the actin fibers? c. Spudich and coworkers used antibodies that bound to the myosin tail. Why would this experiment have failed if they had used an antibody that bound to the part of S1 that normally bound to actin? Why would this experiment have failed if they had used an antibody that bound to actin? To help focus on the part of myosin responsible for force production, Spudich and colleagues used trypsin to produce two partial myosin molecules: heavy meromyosin (HMM) and light meromyosin (LMM), by cleavage of a single specific peptide bond in the myosin tail. Additional incubation with trypsin produced an additional cleavage, eliminating more of the tail and the hinge region to generate short heavy meromyosin (SHMM). d. Why might trypsin attack this peptide bond first, rather than other peptide bonds in myosin? Spudich and colleagues prepared bead-antibody- myosin complexes with varying amounts of myosin, HMM, or SHMM and measured their speed of movement along Nitella actin fibers in the presence of ATP. The graph below sketches their results.

e. Which model (“S1” or “hinge”) is consistent with these results? Explain your reasoning. f. Provide a plausible explanation for the increased speed of the beads with increasing myosin density. g. Provide a plausible explanation for the plateauing of the speed of the beads at high myosin density. Reference Hynes, T.R., S.M. Block, B.T. White, and J.A. Spudich. 1987. Movement of myosin fragments in vitro: domains involved in force production. Cell 48:953–963.