The Specific Conformation Of A Protein Results From Its Pattern Of Hydrogen Bonds

Whereas a portion of the energy stabilizing a protein is provided by hydrophobic interactions, the specific conformation of a protein structure is largely determined b) hydrogen bonds. The energy associated with the hydrophobic stabilization of proteins has no directional component, whereas hydrogen bonds require precise distances and angles (see Figure 3-Q and Table 3-3). In general, all hydrogen-bond donors and acceptors within a protein's interior have suitable mates. Failure to make a hydrogen bond in the protein interior is energetically costly, at the rate of a few kiloealories per hydrogen bond. The vitally important role of hydrogen bonds in proteins is to destabilize incorrect structures as much as to stabilize the correct one.

The necessity of satisfying all the hydrogen-bond donors and acceptors on the polypeptide backbone (two per residue) drives formation of the large sections of« helices and 3 sheets found in most proteins. The only way that 3 polypeptide can traverse the nonaqueous interior of a protein, as it must, and satisfy the hydrogen-bonding necessity is through formation of regular secondary structures. Side chains do not have enough donors and acceptors to do the job. Thus, all large proteins contain significant regions of (3 sheets, a helices, or both. Despite the small number of secondary-structure building blocks, the variety of protein structures that can be built from these is vast. Fven proteins that are composed entirely of j3 sheets or a helices adopt structures spanning a wide range (Figure 5-13).

Of course, some polypeptide sections must lie less regular to allow their chains to turn at the ends of« helices and individual strands of [3 sheets (p strands). Turns are loops of amino acids that link a: helices and p strands but do not exhibit, a defined secondary structure themselves. Turns can vary in length from only a few amino acids to extended segments that are substantially longer. They are, however, generally relatively short so as tn minimise the number of unfulfilled

The Specific Conformation of a Protein Results from /it Pattern of Hydrogen Bonds 79

FIGURE 5-13 Polypeptide chain folding, (a) Protons composed of a lielices: myoglobin and the N-teminal dorr^in of \ repressor, {b} Proteins composed ot p sheets: the Green Fluorescent Protein (GFP) and gamma crystalline, (c) Comparison ot the N terminal domain ot A repressor, composed of a helices with the C terminal domain of \ repressor, composed of p sheeis. ((a) Vb]techwsky j; Bcrendzcn J., Cbu K., Schltchting L, and Sweet RM submitted and Beamer U. and Pabo C.O. 1992. J. Mol. Btd. 227: 1/7 (b) Ormo M, Cubit: A.8., Kdllie K, Gross LA, Tsen RY, and Remington SJ. 1996 Science 273 1392 and Oiirgadze Y t\, LJriessen H.P.C., Wright C., Slmgsby C, Hay R E. and Lindley PF 193 b. Acta Crystdbgrophcr D. Biol. Crystdlogr. 52: 712 (c) Beamer I J and Pabo CO. 1992. J. Mol. Biol 111. 177 and Beil C.F, f-rescura P, and Horfischild A. 2000, Cell 101. 801.) All images prepared with MolScnpt, BobScript. and Raster 3D.

FIGURE 5-13 Polypeptide chain folding, (a) Protons composed of a lielices: myoglobin and the N-teminal dorr^in of \ repressor, {b} Proteins composed ot p sheets: the Green Fluorescent Protein (GFP) and gamma crystalline, (c) Comparison ot the N terminal domain ot A repressor, composed of a helices with the C terminal domain of \ repressor, composed of p sheeis. ((a) Vb]techwsky j; Bcrendzcn J., Cbu K., Schltchting L, and Sweet RM submitted and Beamer U. and Pabo C.O. 1992. J. Mol. Btd. 227: 1/7 (b) Ormo M, Cubit: A.8., Kdllie K, Gross LA, Tsen RY, and Remington SJ. 1996 Science 273 1392 and Oiirgadze Y t\, LJriessen H.P.C., Wright C., Slmgsby C, Hay R E. and Lindley PF 193 b. Acta Crystdbgrophcr D. Biol. Crystdlogr. 52: 712 (c) Beamer I J and Pabo CO. 1992. J. Mol. Biol 111. 177 and Beil C.F, f-rescura P, and Horfischild A. 2000, Cell 101. 801.) All images prepared with MolScnpt, BobScript. and Raster 3D.

hydrogen bonds that accompany their formation (for example, see Figure 5-14).

In addition, the less regular structures of these loops are critical for the formation of binding sites for small molecules, the active sites of enzymes, and the surfaces involved in protein-protein interactions. This will become apparent in the three-dimensional protein structures we discuss in the rest of this chapter and the remainder of the text.

o hairpin loop

00 Oo

figure 5-14 Adjacentantiparaltet p strands are joined by hairpin loops.

Schematic showing an example of a two-residue hairpin loop Hie bonds within the hairpin loop (in shaded area at top of structure) are green

FIGURE 5-15 The leucine zipper from the yeast transcription factor Gcn4. The leuant? dipper is an example ol a coited-coi! (see text). Here we show two views of the leuane lipper: fram (tie side (on the left) and from above (on the right). (Fllenberger TL. Brandl Of, Struhl K., and ttamson S.C. 1992. Cell 71 1223.) Images piepared with MolScnpt, BobScnpt, and Raster 3D.

a Helices Come Together to Form Coiled-Coils

Many polypeptides interact with one another through the supercoil-ing of a helices around each other. Typically, this can only occur when the nonpolar side chains along each a helix are arranged so that their side groups contact the other helix. The twisting of the helices around each other reflects the nonintegral (3.6 residues per turn) nature of the a helix, which allows the side groups to pack neatly together only when the « helices interact at an angle of 1HL from parallel. If the a helices remained perfectly rigid, they could stay in contact for only a few residues. But by supercoiling in a left-handed direction, neatly packed, highly stable, coiled-coils are created [Figure 5-15).

One example of a coiled-coil is found in the leucine zipper family of ONA-binding proteins. These DNA-bindmg factors have two subunits that come together to form a dimer through the use of a coiled-coil region. This coiled-coil region is called a leucine zipper due to the repeating appearance of leucine or other amino acids with an aliphatic side group, such as valine or isoleucine. These leucines appear in a regular pattern as follows. If you consider two turns of an a. helix this will represent a segment of approximately seven amino acids. The aliphatic amino acids are located within each seven amino acid stretch at the first and fourth positions. This positioning ensures that one side of the « helix is aliphatic, since the first and fourth positions will be on the same face of the helix. These faces in two adjacent helices are packed against each other, hurying their hydrophobic side chains away from the aqueous environment.

Most Proteins Are Modular, Cantuiniiig Tivo or Three Domains fl

MOST PROTEINS ARE MODULAR, CONTAINING TWO OiriHRLL DOM A INS___

The subunits of soluble proteins vary in size from less than 100 to larger than 2,000 amino acid residues. The smallest polypeptides that form folded proteins have molecular' weights of about 11,000 daltons (approximately 100 residues], but most are between 20,000 and 70,000 daltons for a single subunit.

Proteins larger than about 20,000 daltons are often formed from two or more domains (Figure 5-1 fi; see also Box 5-2, Large Proteins Are Often Constructed of Several Smaller Polypeptide Chains). The term domain is used to describe a part of the structure that appears separate from the rest, as if it would be stable in solution on its own, which is often the case. Typically, a single domain is formed from a continuous amino acid sequence and not portions of sequence scattered throughout the polypeptide. This is an important point when considering how multidomain proteins have evolved,

Proteins Are Composed of a Surprisingly Small Number of Structural Motifs

Determination of the first half-dozen protein structures showed a bewildering variety of protein folding motifs, implying the existence of an infinite number of protein structures. Now that we know the three-dimensional structures of thousands of proteins, however, it appears that a relatively small number of different domains account for most of the large variety of protein structures. Although an accurate estimate is not possible, the number of truly unique domain motifs will be orders of magnitude smaller than the number of unique proteins.

Specific kinds of domain motifs are often associated with particular kinds of activities. One frequently observed motif has been termed the dinucleotide fold because it is frequently found in enzymes thai bind

FIGURE 5-16 Pyruvate kinase is composed of distinct domains. The predominant domains of the enzyme are shown in turquoise, purple, and red. (Allen 5.C. and Murrhead H. 1996. Acta Cry'.tdiogr 0. Btoi. Crystollgr. 52199 ) image prepared with MolScript BobScript, and Raster 3D.

Box 5-2 Large Proteins Are Often Constructed of Several Smaller Polypeptide Chains

Most large proteins are reguiar aggregates of sevefal smaller polypeptide chains, t he relationship among the polypeptide chains making up such a protein is termed its quarternary structure. For example, the macro molecular complexes responsible for the synthesis of RNA (RNA polymerase) and protein (nbosome) are each assemblies of multiple subunits. The complexes are about 500,OCX) and 2,500,000 daftons, respectively, but do not include any individual subunits greater than 200,000 daltans. The nbosome is composed of both protein and RNA subunits. This type of factor is called a ribonuciear protein (RNP).

Why are large protein complexes composed of multiple subunits rather than a single large subunit? The use of multiple subunits to build large protetn complexes reflects a building principle applicable to all complex structures, nonliving as well as Wing This principle states thai it is much easier to reduce the impact of construction mistakes if faulty subunrts can be discarded before they are incorporated into the final product. For example, let us consider two alternative ways cf constructing a molecule with a million atoms. In scheme 1, we build the structure atom by atom, in scheme 2, we first build a thousand smaller units, each with a thousand atoms, but subsequently put the subunits together into the million-atom product. Now consider that our building process randomly makes mistakes, inserting the wrong atom once every 100,000 times. Let us assume that each mistake results in a nonfunctional product

Under scheme 1, each molecule wilt contain, on the average, ten wrong atoms, and so almost no good products will be synthesized. Under scheme 7, however, mistakes will occur in oniy 1% of the subunits If there is a device to reject the bad subunits, then good products can be made easily, and the cell will hardly be bothered by the occurence of the occasional nonfunctional subunit. This is the same construction strategy that forms the basis of the assembly line, in wNch complicated industrial products, such as radios and automobiles, are constructed. At each stage of assembly, there are devices to throw away bad subunits. In industrial assembly lines, mistakes were initially removed by human hands; now, automation often replaces manual control. In cells, mistakes are sometimes controlled by the specificity of enzymes. If a monomeric subunit is wrongly put together, it usually will not be recognized by the polymer-making enzyme and hence will not be incorporated into a macromolecule. In other cases, faulty substances are rejected because they are unable to spontaneously become part of stable molecular aggregates.

ATP (Figure 5-17). This domain binds ATP through a central, parallel {i sheet with a helices on both sides. The nucleotide binding site is on the carboxyl end of the p strands. What varies is the number and detailed arrangement of the ct helices and, to a lesser exient, the order of the (3 strands. Related domains of similar structure serve the same function in many different proteins.

Different Protein Functions Arise from Various Domain Combinations

The various functional properties of proteins appear to arise from their modular construction in much the same way as computers with different specifications can be assembled from the appropriate modular components. Numerous examples can be given. There are, for example, many dehydrogenase enzymes, each working on a specific substrate. Each enzyme consist of two domains, one a common dinucleotlde

Most Proteins Are Modular. Containing Two or Three iiomams 83

Most Proteins Are Modular. Containing Two or Three iiomams 83

FIGURE 5-17 Enzymes that bind ATP. 7 he red arrows point to the ATP molecules bound within each structure (a) RecA (b) DnaA. {(a) Story RM. and Steitz TA 1992. Nature 355; 374. (b) Erzberger IP., PirrucceSo M.M., and Berger JLM. 2002. EMBOJ. 21: 4763 73.) images prepared with MotSoipt, BobScnpr, and Raster 3D.

binding domain that binds the coenzyme NAD~\ the other a domain that binds substrate and has the catalytic site. The structure of the latter domain varies among different dehydrogenases.

The gene regulatory repressor and activator proteins provide another example of modular construction. The Lac repressor and the calabnlite gene activator protein (CAP) of E. coli both contain multiple domains. The crystal structure of CAP shows two domains; A larger domain binds a molecule of cyclic AMP in its interior, while the smaller domain recognizes specific DNA sequences (Figure 5-18). There are significant amino acid sequence similarities between the cAMP-binding domain of CAP and the regulatory subunit of cAMP-dependent protein kinase, suggesting that the cAMP-binding domain of both proteins evolved from the same

FIGURE 5-18 CAP complex with cAMP interacting with bent DMA- The larger domain of CAP, shown in turquoise, binds cydic AMR shown in red and yellow in the Center of that domain. The smaller, UNA-binding domain (shown in purple), recognizes specific DMA sequences (the double heiix is shown in ted and gray). (Schtiil? SC, Shields OC, and Steitz T.A 1991. Science 253: 1001 ) Image prepared with MolSoipt, Botecnpt, and Raster 3D.

FIGURE 5-18 CAP complex with cAMP interacting with bent DMA- The larger domain of CAP, shown in turquoise, binds cydic AMR shown in red and yellow in the Center of that domain. The smaller, UNA-binding domain (shown in purple), recognizes specific DMA sequences (the double heiix is shown in ted and gray). (Schtiil? SC, Shields OC, and Steitz T.A 1991. Science 253: 1001 ) Image prepared with MolSoipt, Botecnpt, and Raster 3D.

B4 Ufeai: and Strong Bernds Determine Macronwiseular Structure precursor. In CAP, this c AMP-bin ding domain is attached to the DNA-binding domain, so that changes in cAMP levels control transcription levels. In the kinase, the cAMP-b hiding domain regulates the artivity of the First enzyme in a cascade of enzymes that result in the breakdown of stored glycogen.

0 0

Post a comment