Schematic diagram (drawn to scale) of a phosphatidylcholine bilayer and segments of a hypothetical intrinsic membrane protein. The bilayer is fluid and what is shown is a two‐dimensional cross section of it. The apparently empty spaces within it must be imagined as filled with portions of lipid chains that protrude into the pictured cross‐section from adjacent cross‐sections. The hypothetical protein segments are drawn as a bundle of helices, imagined as parts of a single protein molecule, joined together outside the bilayer domain by other parts of the protein polypeptide chain. The small shaded particles are H₂O molecules.

Schematic diagrams showing the topologies of three families of nucleotide binding proteins: (a) for ATP, (b) for GTP, (c) for NAD. All of them contain alternating α‐helices and β‐strands, shown respectively as cylinders and arrows, but the specified arrangements are different. All six β‐strands in (c) are parallel, for example, whereas five are parallel and one is antiparallel in (b).

Schematic ribbon diagram for the three‐dimensional structure of actin, which contains an ATP binding site similar to (a) in Figure 2. The letters and numbers in this diagram identify the amino acids in the sequence at which α‐helices begin and terminate. The ATP binding site is in the cleft between the domains in the lower center of the picture, each domain containing an arrangement of α‐helices and β‐sheets such as those in (a) of Figure 2.

Two functionally unrelated proteins which nevertheless have identical structural topology (and even a similar disulfide bond) over a major part of their three‐dimensional structures. The enzyme superoxide dismutase with Cu and Zn atoms at its active center is on the left. The variable domain of the heavy chain of immunoglobulin G is on the right. The principal structure in both cases is so‐called β‐barrel, composed of seven strands of β‐sheet.

Overall chain trace of bacteriorhodopsin showing helices as solid rods. The N terminus (nt) is at the bottom at the extracellular surface; the C terminus (ct) is at the top at the cytoplasmic surface. The helices range from 30 to 40 Å in length, sufficient to span the hydrophobic core of the lipid bilayer. The retinal is attached by a Schiff base linkage to Lys 216 (helix G) and is tilted roughly 20° out of the plane of the membrane.

Ribbon diagram of matrix porin from E. coli. Arrows represent β‐strands and are labelled 1–16 starting from the strand after the first short turn. The short β‐strand at the N terminus continues the C‐terminal strand β 16. The long loops are denoted L1‐L8, the short turns at the other end T1‐T8. Loop L2 protrudes towards the viewer. Loop L3 folds inside the barrel.

Proposed arrangements of the polypeptide chains of the principal subunits of three transmembrane ion channels: Na⁺ channel from rat brain, Ca²⁺ channel from rabbit skeletal muscle, and A‐current K⁺ channel from Drosophila. Na⁺ channel polypeptides from various sources range in size from 1832 to 2012 amino acid residues. Similarly, the Ca⁺ channel polypeptide consists of 2005 residues. The K⁺ channel protein, however, is only 616 residues long. Putative trans‐membrane helices are shown as numbered cylinders.

Sliding helix model of voltage‐dependent gating. Movement of the S4 helix of domain IV of the Na⁺ channel in response to membrane depolarization. The proposed helix is illustrated as a cylinder with a spiral ribbon of positive charge. At the resting membrane potential (left) all positively charged residues are paired with fixed negative charges on other transmembrane segments of the channel and the transmembrane segment is held in that position by the negative internal membrane potential. Depolarization reduces the force holding the positive charges in their inward position. The S4 helix is then proposed to undergo a spiral motion through a rotation of approximately 60° and an outward displacement of approximately 5 Å. This movement leaves an unpaired negative charge on the inward surface of the membrane and reveals an unpaired positive charge on the outward surface to give a net charge transfer of +1.

Artistic impression showing the relationship between the key residues Asp 85, Asp 96, Asp 212, Lys 216, and Arg 82 and the retinal binding site, the proton channel and overall molecular boundary for the ground state of bacteriorhodopsin. The cytoplasm is at the top of the diagram.

A diagram to describe the 4 proton movements in the photocycle of bacteriorhodopsin. (a) Structure before isomerization of the retinal. (b) Proton transfer from Schiff base to Asp 85. (c) Proton transfers from Asp 96 to Schiff base and from Asp 85 to the extracellular space. (d) Proton transfer from the cytoplasmic surface to Asp 96.

Pictorial representation of the successive stages of the Ca⁺² binding domain of the sarcoplasmic reticulum calcium pump. The binding cavity is assumed to extend physically into the cytoplasm and its inner surface is assumed to possess a net negative charge. E₁ and E₂ are the names given to the two principal conformational states with access to opposite sides of the membrane. E_x, E_y, and E_z are variants of E₁ which had to be assumed to account for details of the experimental data for Ca²⁺ binding from the cytoplasmic side of the membrane.

Speculative model for free energy coupling in the sarcoplasmic reticulum calcium pump. It is suggested that the binding of Ca⁺² (Fig. 11) brings the ATP binding region and the phosphorylation site into close proximity, permitting the transfer of the terminal phosphate group of ATP from one site to the other.