16.3 What Are the Molecular Motors That Orchestrate the Mechanochemistry of Microtubules? 493 Kinesin 1 (KHC) (a) (b) N type N-terminal domain Motor domain Coiled coil domain C-terminal tail domain Light chain Kinesin 13 (MCAK) Conventional kinesin (kinesin I) Myosin V Cytoplasmic dynein M type Kinesin 14 (Ncd) C type Myosin V Dynein Tail Calmodulin Light chain Kinesin light chain N C Motor Neck D1 D2 D3 D4 MT AAA+ Light chains D5 D6 3636 3409 3016 2767 2687 2422 2316 2095 2035 1814 4072 4466 3846 Stalk C-term Light chain FIGURE 16.14 (a) Domain structure of kinesins, myosin V, and cytoplasmic dynein. (b) Molecular models of kinesin 1, myosin V, and cytoplasmic dynein. (Adapted from Vale, R., 2003.The molecular motor toolbox for intracellular transport. Cell 112:467–480.) 494 Chapter 16 Molecular Motors domain is located in different places in the sequence, depending on the function of the specific family. The first dyneins to be discovered were axonemal dyneins, which cause sliding of microtubules in cilia and flagella. Cytoplasmic dyneins were first identified in Caenorhabditis elegans, a nematode worm. Cytoplasmic dynein consists of a dimer of two heavy chains (500 kD) with several other tightly associated light chains (Figure 16.14). Each heavy chain contains a large motor domain (380 kD) encompassing six AAAϩ domains (see Section 16.4) arranged as a hexamer. A 10-nm stalk composed of a coiled coil projects from the head, between the fourth and the fifth AAAϩ do- mains. The stalk is the microtubule-binding domain. Myosin V is a multimeric protein that consists of 16 polypeptide chains. The structure is built around a dimer of heavy chains, each of which includes head, neck, and tail domains. The heavy chain head domain is virtually indistinguish- HUMAN BIOCHEMISTRY Effectors of Microtubule Polymerization as Therapeutic Agents Microtubules in eukaryotic cells are important for the mainte- nance and modulation of cell shape and the disposition of intra- cellular elements during the growth cycle and mitosis. It may thus come as no surprise that the inhibition of microtubule polymerization can block many normal cellular processes. The alkaloid colchicine (see accompanying figure), a constituent of the swollen, underground stems of the autumn crocus (Colchi- cum autumnale) and meadow saffron, inhibits the polymerization of tubulin into microtubules. This effect blocks the mitotic cycle of plants and animals. Colchicine also inhibits cell motility and intracellular transport of vesicles and organelles (which in turn blocks secretory processes of cells). Colchicine has been used for hundreds of years to alleviate some of the acute pain of gout and rheumatism. In gout, white cell lysosomes surround and engulf small crystals of uric acid. The subsequent rupture of the lysosomes and the attendant lysis of the white cells initiate an inflammatory response that causes intense pain. The mecha- nism of pain alleviation by colchicine is not known for certain, but appears to involve inhibition of white cell movement in tis- sues. Interestingly, colchicine’s ability to inhibit mitosis has given it an important role in the commercial development of new varieties of agricultural and ornamental plants. When mito- sis is blocked by colchicine, the treated cells may be left with an extra set of chromosomes. Plants with extra sets of chromo- somes are typically larger and more vigorous than normal plants. Flowers developed in this way may grow with double the normal number of petals, and fruits may produce much larger amounts of sugar. Another class of alkaloids, the vinca alkaloids from Vinca rosea, the Madagascar periwinkle, can also bind to tubulin and inhibit microtubule polymerization. Vinblastine and vincristine are used as potent agents for cancer chemotherapy because of their ability to inhibit the growth of fast-growing tumor cells. For reasons that are not well understood, colchicine is not an effective chemother- apeutic agent, although it appears to act similarly to the vinca alkaloids in inhibiting tubulin polymerization. The antitumor drug taxol was originally isolated from the bark of Taxus brevifolia, the Pacific yew tree. Like vinblastine and col- chicine, taxol inhibits cell replication by acting on microtubules. Unlike these other antimitotic drug s, however, taxol stimulates microtubule polymerization and stabilizes microtubules. The re- markable success of taxol in the treatment of breast and ovarian cancers stimulated research efforts to synthesize taxol directly and to identify new antimitotic agents that, like taxol, stimulate microtubule polymerization. NH H H OH OH O O O O O O O OO O O OH O H 3 C CH 3 H 3 C O H 3 C O O O O C CH 3 CH 3 O O CH 3 NH C O CH 3 C OCH 3 O C H 3 CO H 3 CO CH 2 CH 3 CH 2 CH 3 N HO OH R N N N Taxol Colchicine Vinblastine: R = CH 3 Vincristine: R = CHO ᮡ The structures of vinblastine, vincristine, colchicine, and taxol. 16.3 What Are the Molecular Motors That Orchestrate the Mechanochemistry of Microtubules? 495 able from the head domain of myosin II from skeletal muscle (see Figure 16.5), but the neck domain is three times longer than the myosin II neck helix and it contains six repeats of a calmodulin-binding domain. Myosin V is normally associ- ated with an essential light chain (similar to that of myosin II), together with sev- eral calmodulins. Adjoining the neck is a 30-nm-long coiled-coil domain. The tail domain of myosin V also binds a light chain and is adapted to bind specific or- ganelles and other cargoes. Dyneins Move Organelles in a Plus-to-Minus Direction; Kinesins, in a Minus-to-Plus Direction—Mostly The mechanisms of intracellular, microtubule-based transport of organelles and vesi- cles were first elucidated in studies of axons, the long projections of neurons that ex- tend geat distances from the body of the cell. In these cells, it was found that sub- cellular organelles and vesicles could travel at surprisingly fast rates—as great as 1000 to 2000 nm/sec—in either direction. Cytosolic dyneins specifically move organelles and vesicles from the plus end of a microtubule to the minus end. Thus, dyneins move vesicles and organelles from the Nucleus Mitochondrion Golgi apparatus Synaptic terminal Multivesicular body Microtubule Lysosome Cell body Rough endoplasmic reticulum (a) Vesicles (b) ؊ ؉ Kinesin Organelle Vesicle FIGURE 16.15 (a) Rapid axonal transport along micro- tubules permits the exchange of material between the synaptic terminal and the body of the nerve cell. (b) Vesi- cles, multivesicular bodies, and mitochondria are carried through the axon by this mechanism. (Adapted from a drawing by Ronald Vale.) 496 Chapter 16 Molecular Motors cell periphery toward the centrosome (or, in an axon, from the synaptic termini to- ward the cell body). Most kinesins, on the other hand, assist the movement of or- ganelles and vesicles from the minus end to the plus end of microtubules, resulting in outward movement of organelles and vesicles (Figure 16.15). Certain unconven- tional kinesins move in the opposite direction, transporting cargo in the plus-to-mi- nus direction on microtubules. These kinesins have their motor domain located at the C-terminus of the polypeptide (see Figure 16.14). Cytoskeletal Motors Are Highly Processive The motors that move organelles and other cellular cargo on microtubules and actin filaments must be processive, meaning that they must make many steps along their cellular journey without letting go of their filamentous highway. Dyneins, nonskeletal myosins, and most kinesins are processive motors (Table 16.2). Motors in all these classes can carry cargo over roughly similar distances (700 to 2100 nm) before dissociating. 1 The step size of kinesin 1 is smaller than those of myosin V and cytoplasmic dynein; thus, its overall processivity (the average number of steps made before dissociating) is necessarily higher. Moving at rates of 600 to 1000 nm/sec, these motors can carry their cargoes for a second or more before dissociating from their filaments. ATP Binding and Hydrolysis Drive Hand-over-Hand Movement of Kinesin Kinesin movement along microtubules is driven by the cycle of ATP binding and hy- drolysis. The molecular details are similar in some ways to those of the skeletal mus- cle myosin–actin motor but are quite different in other ways. Kinesins, like skeletal muscle myosin, contain switch 1 and switch 2 domains that open and close in re- sponse to ATP binding and hydrolysis. Together these switches act as a “␥-phosphate sensor,” which can detect the presence or absence of the ␥-phosphate of an adenine nucleotide in the active site. The switch movements between the ATP-bound and the ADP-bound states thus induce conformation changes that are propagated through a relay helix to a neck linker that rotates, in ways similar to skeletal muscle myosin (Figure 16.16). Thus, just as in skeletal myosin, small movements of the ␥-phosphate sensor at the ATP site are translated into piston-like movement of a relay helix and then into rotations of the neck linker that result in motor movement. Here the kinesin and myosin models diverge, however, because the dimer of kinesin heavy chains translates these ATP-induced conformation changes into a hand-over-hand movement of its motor domain heads along the microtubule fila- ment. Ronald Vale and Ronald Milligan have likened this movement of kinesin heads along a microtubule to a judo expert throwing an opponent with a forward swing of the arm. A model of kinesin motor movement is shown in Figure 16.17. Kinesin heads in solution (that is, not attached to a microtubule) contain tightly bound ADP. Bind- 1 Compare these distances with the dimensions of typical cells in Table 1.2. Distance Processivity Traveled (average Rate of Before number of % Chance of Movement Step Size Dissociating steps before Dissociating Motor (nm/sec) (nm) (nm) dissociating) in One Step Kinesin 1 600 8 800–1200 100–120 ϳ1 Myosin V 1000 36 700–2100 20–60 ϳ2–5 Dynein 600 24–32 1000 30–40 ϳ2–3 (cytoplasmic) TABLE 16.2 Processivity of Motor Proteins 16.3 What Are the Molecular Motors That Orchestrate the Mechanochemistry of Microtubules? 497 ing of one head of a kinesin multimer to a microtubule causes dissociation of ADP from that head. ATP binds rapidly, triggering the neck linker to rotate or ratchet forward, throwing the second head forward as well and bringing it near the next binding site on the microtubule, 8 nm farther along the filament. The trailing head then hydrolyzes ATP and releases inorganic phosphate (but retains ADP), inducing its neck linker to return to its original orientation relative to the head. Exchange of ADP for ATP on the forward head begins the cycle again. The structure of the kinesin–microtubule complex (Figure 16.18) shows the switch 2 helix of kinesin in intimate contact with the microtubule at the junction of the ␣- and -subunits of a tubulin dimer. The Conformation Change That Leads to Movement Is Different in Myosins and Dyneins The movement of myosin motors on cytoskeletal actin filaments is presumed to be similar to the myosin–actin interaction in skeletal muscle. Clearly, however, the dif- ferent structure of the dynein hexameric motor domain and its associated coiled- coil stalk (see Figure 16.14) must represent a different motor mechanism. ATP- dependent conformation changes in the ring of AAAϩ modules must be translated Converter Relay helix Relay helix Myosin Kinesin Neck linker FIGURE 16.16 Ribbon structures of the myosin and kinesin motor domains and the conformational changes triggered by the relay helix.The upper panels represent the motor domains of myosin and kinesin, respectively, in the ATP- or ADP-P i –like state. Similar structural elements in the catalytic cores of the two domains are shown in blue, the relay helices are dark green, and the mechanical elements (neck linker for kinesin, lever arm domains for myosin) are yellow.The nucleotide is shown as a white space-filling model.The similarity of the conformation changes caused by the relay helix in going from the ATP/ADP-P i –bound state to the ADP-bound or nucleotide- free state is shown in the lower panels.In both cases, the mechanical elements of the protein shift their positions in response to relay helix motion. Note that the direction of mechanical element motion is nearly perpendicular to the relay helix motion. (Adapted from Vale, R. D., and Milligan, R. A., 2000.The way things move: Looking under the hood of molecular motor proteins. Science 288:88–95.) Kinesin ADP ADP ADP ATP ATP ADP-P i ATP 1 2 3 4 FIGURE 16.17 A model for the motility cycle of kinesin. The two heads of the kinesin dimer work together to move processively along a microtubule.Frame 1: Each kinesin head is bound to the tubulin surface.The heads are connected to the coiled coil by “neck linker” segments (orange and red). Frame 2: Conformation changes in the neck linkers flip the trailing head by 160°, over and be- yond the leading head and toward the next tubulin bind- ing site.Frame 3: The new leading head binds to a new site on the tubulin surface (with ADP dissociation), com- pleting an 80 Å movement of the coiled coil and the kinesin’s cargo.During this time, the trailing head hydro- lyzes ATP to ADP and P i .Frame 4: ATP binds to the leading head,and P i dissociates from the trailing head,completing the cycle. (Adapted from Vale, R.,and Milligan, R.,2000.The way things move:Looking under the hood of molecular motor proteins. Science 288:88–95.) 498 Chapter 16 Molecular Motors into movements of the tip of the coiled-coil stalk along a microtubule. A proposed mechanism for dynein movement (Figure 16.19) suggests that the events of ATP binding and hydrolysis and ADP and P i release at an AAAϩ module swing a linker that joins the AAAϩ domain and the dynein tail. 16.4 How Do Molecular Motors Unwind DNA? The ability of proteins to move in controlled ways along nucleic acid chains is im- portant to many biological processes. For example, when DNA is to be replicated, the strands of the double helix must be unwound and separated to expose single- stranded DNA templates. Similarly, histone octamers (Figure 11.26) slide along DNA strands in chromatin remodeling, Holliday junctions (see Figure 28.22) move, and nucleic acids move in and out of viral capsids. The motor proteins that move directionally along nucleic acid strands and accomplish these many functions are called translocases. The translocases that unwind DNA or RNA duplex substrates are termed helicases. Thus, all helicases are translocases, but not vice versa. FIGURE 16.18 In the kinesin–microtubule complex, the switch 2 helix (yellow) of kinesin (left) lies in contact with the microtubule (right) at the subunit interface of a tubulin dimer (pdb id ϭ 2HXH). (a) Motor domain Microtubule(b) Cargo, such as vesicles ADP + P i Power stroke Hydrolysis = ADP bound = ATP bound Stalk Head Linker Tail ATP FIGURE 16.19 A mechanism for the dynein power stroke involves conformation changes in the head domain (a) that facilitate movement of the stalk along a microtubule (b). ATP binding to the motor domain promotes dissociation of dynein from the microtubule. Hydrolysis of ATP causes a conformation change that primes the structure for a power stroke. Microtubule movement is initiated by tight binding to the tip of the stalk, which promotes a conformation change in the head ring (the power stroke). Release of ADP and P i from the catalytic site causes tilting of the stalk at the end of the cycle. (Adapted from Oiwa, K.,and Sakakibara, H., 2005. Recent progress in dynein structure and mechanism. Current Opinion in Cell Biology 17:98–103.) 16.4 How Do Molecular Motors Unwind DNA? 499 All translocases and helicases are members of six protein “superfamilies” (Table 16.3 and Figure 16.20), all of them related evolutionarily to RecA, a DNA-binding protein (pages 881–882). Motors of superfamily 1 (SF1) and superfamily 2 (SF2) consist of two RecA domains in a tandem repeat. Motors of SF3 through SF6 are built from single RecA domain peptides that associate to form hexamers and dode- camers. Each superfamily possesses characteristic conserved residues and sequence elements (Table 16.3), most of which are shared between several superfamilies. All members of a given superfamily move in the same direction on a DNA or RNA tem- plate (either 5Ј to 3Ј or 3Ј to 5Ј). The hexameric motor proteins of the SF3 and SF6 superfamilies are members of the ancient AAA؉ ATPase family. (AAA stands for “ATPases associated with various cellular activities,” and the “ϩ” sign refers to an ex- panded definition of the family characteristics.) Translocases and helicases, like other molecular motors, require energy for their function. The energy for movement along a nucleic acid strand, as well as for sepa- ration of the strands of a duplex (DNA or RNA), is provided by hydrolysis of ATP. Translocases and helicases move on nucleic acid strands at rates of a few base pairs to several thousand base pairs per second. These movements are carefully regulated by accessory proteins in nearly all cases. Translocases and helicases typically move Quanterary Direction of Representative Family Subunit Structure Structure Movement Motor SF1A Monomeric A* Rep SF1B B RecD SF2 Monomeric A or B NS3 SF3 Hexameric A BPV E1 SF4 Hexameric B T7 gp4 SF5 Hexameric B Rho SF6 Hexameric A or B FtsK MCM Zn binding Superfamily 6 (MCM) Origin-binding domain Superfamily 5 Primase Superfamily 4 Origin binding Superfamily 3 (BPV E1) Protein:protein?Protease Superfamily 2 DNA binding UnknownDNA binding Core domains Accessory domains Superfamily 1 TABLE 16.3 Helicase Superfamilies *A: helicase moves 3Ј→5Ј on nucleic acid. B: helicase moves 5Ј→3Ј. 500 Chapter 16 Molecular Motors along the DNA or RNA lattice for long distances without dissociating. This is termed processive movement, and helicases are said to have a high processivity. For example, the E. coli BCD helicase, which is involved in recombination processes, can unwind 33,000 base pairs before it dissociates from the DNA lattice. Processive movement is es- sential for helicases involved in DNA replication, where millions of base pairs must be replicated rapidly. Helicases have evolved at least two structural and functional strategies for achiev- ing high processivity. The hexameric helicases (of the SF3 through SF6 superfamilies) form ringlike structures that completely encircle at least one of the strands of a DNA duplex. The SF1 and SF2 helicases, notably Rep helicase from E. coli, are monomeric or homodimeric and move processively along the DNA helix by means of a “hand- over-hand” movement that is remarkably similar to that of kinesin’s movement along microtubules. A key feature of hand-over-hand movement of a dimeric motor protein along a polymer is that at least one of the motor subunits must be bound to the poly- mer at any moment. Negative Cooperativity Facilitates Hand-over-Hand Movement How does hand-over-hand movement of a motor protein along a polymer occur? Clues have come from the structures of Rep helicase and its complexes with DNA. The Rep helicase from E. coli is a 76-kD protein that is monomeric in the absence of DNA. Binding of Rep helicase to either single-stranded or double-stranded DNA induces dimerization, and the Rep dimer is the active species in unwinding DNA. Each subunit of the Rep dimer can bind either single-stranded (ss) or double- stranded (ds) DNA. However, the binding of Rep dimer subunits to DNA is negatively co- operative (see Chapter 15). Once the first Rep subunit is bound, the affinity of DNA for the second subunit is at least 10,000 times weaker than that for the first! This negative cooperativity provides an obvious advantage for hand-over-hand walking. When one “hand” has bound the polymer substrate, the other “hand” releases. A conformation change could then move the unbound “hand” one step farther along the polymer where it can bind again. But what would provide the energy for such a conformation change? ATP hydrolysis is the driving force for Rep helicase movement along DNA, and the neg- ative cooperativity of Rep binding to DNA is regulated by nucleotide binding. In the absence of nucleotide, a Rep dimer is favored, in which only one subunit is bound to ssDNA. In Figure 16.21a, this state is represented as P 2 S [a Rep dimer (P 2 ) bound AMP-PNP C core N core (a) (b) FIGURE 16.20 (a) Translocase and helicase motors of SF1 and SF2 are monomers that consist of two RecA domains in a tandem repeat (pdb id ϭ 1QHG). (b) Motor peptides of SF3 through SF6 associate to form hexamers (as shown) or dodecamers (pdb id ϭ 1CR0). (b) 2B 2B 2A 1A 1B Translocation ATP Active unwinding (a) ADP + P i P 2 SP 2 SD P 2 S 2 P 2 S 3Ј (I) (IЈ)(II) (III) 3Ј 3Ј 3Ј FIGURE 16.21 (a) A hand-over-hand model for movement along (and unwinding of ) DNA by E. coli Rep helicase.The P 2 S state consists of a Rep dimer bound to ssDNA.The P 2 SD state in- volves one Rep monomer bound to ssDNA and the other bound to dsDNA.The P 2 S 2 state has ssDNA bound to each Rep monomer. ATP binding and hydrolysis control the interconversion of these states and walking along the DNA substrate. (b) Crystal structure of the E. coli Rep helicase monomer with bound ssDNA (dark blue, ball and stick) and ADP (red).The monomer consists of four domains designated 1A (residues 1–84 and 196–276), 1B (residues 85–195), 2A (residues 277–373 and 543–670), and 2B (residues 374–542).The open (purple) and closed (green) confor- mations of the 2B domain are superimposed in this figure (pdb id ϭ 1UAA). (From Korolev, S., Hsieh, J., Gauss, G., Lohman,T. L., and Waksman, G., 1997. Major domain swiveling revealed by the crystal struc- tures of complexes of E.coli Rep helicase bound to single-stranded DNA and ADP. Cell 90:635–647. Reprinted by permission of Cell Press.) 16.4 How Do Molecular Motors Unwind DNA? 501 to ssDNA (S)]. Timothy Lohman and his colleagues at Washington University in St. Louis have shown that binding of ATP analogs induces formation of a complex of the Rep dimer with both ssDNA and dsDNA, one to each Rep subunit (shown as P 2 SD in Figure 16.21a). In their model, unwinding of the dsDNA and ATP hydroly- sis occur at this point, leaving a P 2 S 2 state in which both Rep subunits are bound to ssDNA. Dissociation of ADP and P i leave the P 2 S state again (Figure 16.21a). Work by Lohman and his colleagues has shown that coupling of ATP hydroly- sis and hand-over-hand movement of Rep over the DNA involves the existence of the Rep dimer in an asymmetric state. A crystal structure of the Rep dimer in com- plex with ssDNA and ADP shows that the two Rep monomers are in different con- formations (Figure 16.21b). The two conformations differ by a 130° rotation about a hinge region between two subdomains within the monomer subunit. The hand- over-hand walking of the Rep dimer along the DNA surface may involve alterna- tion of each subunit between these two conformations, with coordination of the movements by nucleotide binding and hydrolysis. Papillomavirus E1 Helicase Moves along DNA on a Spiral Staircase Papillomaviruses are tumor viruses that cause both cancerous and benign lesions in a host. Replication of papillomaviral DNA within a host cell requires the multifunc- tional 605-residue viral E1 protein. Monomers of E1 assemble at a replication ori- gin on DNA and form a pair of hexamers that wrap around a single strand of DNA. These assemblies are helicases that operate bidirectionally in the replication of viral DNA. The N-terminal half of the E1 protein includes a regulatory domain and a sequence-specific DNA-binding domain, whereas the helicase activity is located in the C-terminal half of the protein. The C-terminal helicase domain (Figure 16.22) includes a segment involved in oligomer formation (residues 300 to 378) and an AAAϩ domain (residues 379 to 605). AAAϩ domains are found in proteins of many functions, including motor activ- ity by dyneins (see Section 16.3) and helicases, protein degradation by proteasomes (see Chapter 31), and disassembly of SNARE complexes (see Chapter 9). This ubiq- uitous module consists of two subdomains: an N-terminal segment known as an ␣/ Rossman fold, and a C-terminal ␣-helical domain (Figure 16.23). The Rossman fold is wedge-shaped and has a -sheet of parallel strands in a 5-1-4-3-2 pat- tern. Key features of this fold include a Lys residue in the Walker A motif, an Asp–Asp or Asp–Glu pair in the Walker B motif, and a crucial Arg residue in a struc- ture called an arginine finger. These three motifs are essential for ATP binding and (a) 100 Å (b) F E D C B A FIGURE 16.22 (a) The papillomavirus E1 protein is a 605- residue monomer that forms hexameric assemblies at specific sites on single-stranded DNA (pdb id ϭ 2V9P). (b) The C-terminal helicase domain shown here includes an oligomerization domain (magenta) and the AAAϩ domain (blue). FIGURE 16.23 The AAAϩ domain is composed of an N-terminal, wedge-shaped Rossman fold and a C-terminal ␣-helical domain (upper left). The P-loop (red), the Walker A motif (purple), the Walker B motif (yellow), and the arginine finger (blue) are shown.The ATP-binding sites lie between subunits of the hexamer. Each ATP site includes the arginine finger of one sub- unit and the Walker A and Walker B motifs of the adja- cent subunit (pdb id ϭ 1D2N). P-loop Walker A Walker B 502 Chapter 16 Molecular Motors hydrolysis. In an AAAϩ hexamer, the ATP-binding sites lie at the interface between any two subunits, involving the arginine finger of any given subunit and the Walker A and Walker B motifs of the adjacent subunit. The structure of a large fragment (residues 306 to 577) of the papillomavirus E1 protein bound to a segment of ssDNA (Figure 16.24) reveals the remarkable mech- anism by which this helicase traverses a DNA chain. The oligomerization domains form a symmetric hexamer, but the six AAAϩ domains each display a unique con- formation. The DNA strand is bound in the center pore of the AAAϩ hexamer, with six nucleotides of the DNA chain each bound to residues from each of the protein subunits. The crucial nucleotide-binding residues include Lys 506 and His 507 on a hairpin loop and Phe 464 , all of which face the center pore of the protein (Figure 16.25). Lys 506 interacts with one DNA phosphate oxygen, and His 507 forms a hydro- gen bond with the phosphate of an adjacent nucleotide in the DNA chain. The aliphatic portion of Lys 506 and the aromatic groups of Phe 464 and His 507 share van der Waals interactions with the DNA sugar moiety linking these two phosphates. The F E D C B A FIGURE 16.24 A view of the E1 helicase along the ssDNA axis, showing the DNA-binding hairpin loops from each monomer (colored) interacting with the phosphates of DNA. DNA is shown as a ball-and-stick model in the center of the structure (pdb id ϭ 2GXA). FIGURE 16.25 The hairpin loops of each subunit in the E1 helicase interact with two adjacent nucleotides in the DNA chain. Interactions include an ionic bond between Lys 506 (yellow) and a DNA phosphate oxygen, a hydrogen bond between His 507 and the phosphate of an adjacent nucleotide, and van der Waals interactions between the aromatic rings of Phe 464 (purple) and His 507 (olive) and the aliphatic chain of Lys 506 , with the sugar linking the two phosphates (pdb id ϭ 2GXA).