If the tertiary structures of many proteins are examined, it soon becomes evident that they can often be divided into folded domains. Although the precise definition of a domain is still a matter of contention, the common opinion is that they are units of protein structure that could fold autonomously, that is, independently of other domains in the protein. Domains are often quite compact and are connected by relatively unstructured segments of the polypeptide chain. Their connectivity may be enhanced by metal ion binding or occasionally by covalent disulfide bonds formed by oxidation of sulfhydryl side chains. Domains range from 25 to ~500 residues in length. It seems likely that many domains have evolved independently and have been connected to form specifically functional proteins later in evolution. Domain structure is common; about 90% of all known human proteins contain distinguish able domains.
Protein evolution seems to have proceeded from a relatively small number of domain types, which then further evolved and connected into the multidomain proteins we see today. This may account for the fact that features of protein tertiary structure seem to be even more conserved than sequence and can be used as clues to recognizing distant protein relatives, whose relatedness would not be evident from their sequences alone. A gallery of some common domain types is shown in Figure1. Note that domains can be largely classified by their combinations of α and β secondary structure elements.
Fig1. Representative examples of individually folded protein domains and their functions. (A) β-Catenin is made up of a number of repeats of the armadillo domain. Each armadillo repeat contains approximately 42 residues, forming three helices shown in yellow, blue, and green, arranged in a shape that has a triangular cross-section. The 12 contiguous repeats of the armadillo repeat domain form a superhelix that features a long, positively charged groove. Many different protein partners bind to portions of the repeat domain. β-Catenin is involved in cell adhesions, which link the adhesion protein cadherin to the cytoskeleton, or in signaling pathways that change gene expression: in the presence of appropriate signals, β-catenin moves to the nucleus and interacts with transcription factors to activate target genes. (B) Cadherin repeats are named after cadherin proteins that play a fun damental role in calcium-dependent cell–cell adhesion. Adjacent cells adhere to each other via homophilic binding between cadherin molecules on their surfaces. Cadherins are typically formed by a single-pass transmembrane domain and multiple extracellular repeats. Each cad herin repeat consists of about 110 amino acid residues forming seven antiparallel β-strands, which are tightly linked by hydrogen bonds, and a highly conserved calcium-binding motif. (C) Immunoglobulin-like domains were originally described in immunoglobulins but have since been found widely distributed among diverse proteins. Two β-sheets of antiparal lel β-strands form a sandwich that is stabilized by interaction between hydrophobic amino acids on the inner side of the sandwich and highly conserved disulfide bonds. The immuno globulin superfamily includes proteins with roles in the immune system, such as cell surface antigen receptors, antibodies, and major histocompatibility complex; in muscle contraction, such as titin; and in cell adhesion. The immunoglobulin superfamily is the most populous known family of proteins, with more than 765 members in humans. (D) The SH2 domain contains a large β-sheet flanked by two α-helices, shown in orange and blue. A total of 120 SH2 domains, contained within 115 proteins, are encoded in the human genome. The SH2 domain binds to phosphorylated tyrosine and is found in signal transduction proteins. Extracellular signals sensed by receptors in membranes are converted to the chemical signal phosphorylated tyrosine; this initiates a series of events that eventually results in altered patterns of gene expression. (E) The death effector domain, DED, is a protein interaction domain that allows protein–protein binding by DED–DED interactions. DED is composed of a bundle of six α-helices. The dimerization of DED domains is mediated primarily by electro static interactions. DED proteins are involved in apoptosis (programmed cell death) which occurs through the activation of proteolytic cascades. The major players in these cascades are caspase proteases, whose inactive precursors are activated by DED–DED interactions with specific adaptor molecules. The structure shown is that of the DED domain of FADD. (F) The EF-hand motif consists of two perpendicular 10- to 12-residue α-helices with a 12-residue loop region between, forming a single calcium-binding site as a helix–loop–helix. Calcium ions interact with residues contained within the loop region. EF-hand domains are often found in single or multiple pairs; two pairs are shown here. A typical representative of EF-hand proteins is calmodulin, which can bind to and regulate a number of different protein targets in response to calcium binding.
Very often, the combination of several domain types within a protein molecule allows it to carry out multitasking, with each domain having a specific function. An example is pyruvate kinase (Figure 2A), a key enzyme in the utilization of glucose as an energy source. This protein has three domains: an all-β regulatory domain and two α/β domains. One of these α/β domains catalyzes the dephosphorylation of phospho-enolpyruvate to pyruvate, and the other accepts the adenosine diphosphate, ADP, that is converted to adenosine triphosphate, ATP, in the reaction. The regulation in this case is allosteric, mediated by a number of metabolites that bind to the regulatory domain and modify the catalysis. See Box 3.8 for a detailed discussion of allosteric regulation.
Fig2. Two examples of multidomain proteins containing either distinct domains or closely related repetitive domains. (A) Pyruvate kinase, a three-domain protein. The central α/β barrel domain is present in numerous different enzymes that catalyze completely unrelated reactions. It is formed by different noncontiguous segments of the polypeptide chain coming in close, spatial proximity, hence the name discontinuous domain. The β-barrel regulatory domain is continuous, formed by a single stretch of the polypeptide chain. (B) A portion of the giant muscle protein titin which consists of repeating immunoglobulin-like domains. These domains are believed to confer elasticity to the muscle fibers. [B, adapted, courtesy of Wikimedia.]
There are also proteins in which one domain type is repeated. Often, this is an economical way to build a very large protein, such as the giant muscle protein titin (Figure 2B), which consists of hundreds of repetitions of immunoglobulin-like (Ig-like) domains (see Figure 1). This also demonstrates the versatility of domains; the Ig domains were originally noted in the structure of antibodies, which have no function in common with titin. Evolution makes do with whatever useful building blocks it can find.
