Activated STING in a vascular and pulmonary syndrome.

BACKGROUND: The study of autoinflammatory diseases has uncovered mechanisms underlying cytokine dysregulation and inflammation. METHODS: We analyzed the DNA of an index patient with early-onset systemic inflammation, cutaneous vasculopathy, and pulmonary inflammation. We sequenced a candidate gene, TMEM173, encoding the stimulator of interferon genes (STING), in this patient and in five unrelated children with similar clinical phenotypes. Four children were evaluated clinically and immunologically. With the STING ligand cyclic guanosine monophosphate-adenosine monophosphate (cGAMP), we stimulated peripheral-blood mononuclear cells and fibroblasts from patients and controls, as well as commercially obtained endothelial cells, and then assayed transcription of IFNB1, the gene encoding interferon-ß, in the stimulated cells. We analyzed IFNB1 reporter levels in HEK293T cells cotransfected with mutant or nonmutant STING constructs. Mutant STING leads to increased phosphorylation of signal transducer and activator of transcription 1 (STAT1), so we tested the effect of Janus kinase (JAK) inhibitors on STAT1 phosphorylation in lymphocytes from the affected children and controls. RESULTS: We identified three mutations in exon 5 of TMEM173 in the six patients. Elevated transcription of IFNB1 and other gene targets of STING in peripheral-blood mononuclear cells from the patients indicated constitutive activation of the pathway that cannot be further up-regulated with stimulation. On stimulation with cGAMP, fibroblasts from the patients showed increased transcription of IFNB1 but not of the genes encoding interleukin-1 (IL1), interleukin-6 (IL6), or tumor necrosis factor (TNF). HEK293T cells transfected with mutant constructs show elevated IFNB1 reporter levels. STING is expressed in endothelial cells, and exposure of these cells to cGAMP resulted in endothelial activation and apoptosis. Constitutive up-regulation of phosphorylated STAT1 in patients' lymphocytes was reduced by JAK inhibitors. CONCLUSIONS: STING-associated vasculopathy with onset in infancy (SAVI) is an autoinflammatory disease caused by gain-of-function mutations in TMEM173. (Funded by the Intramural Research Program of the National Institute of Arthritis and Musculoskeletal and Skin Diseases; ClinicalTrials.gov number, NCT00059748.).

Epitope diversity of hepatitis B virus capsids: quasi-equivalent variations in spike epitopes and binding of different antibodies to the same epitope.

To investigate the range of antigenic variation of HBV capsids, we have characterized the epitopes for two anti-capsid antibodies by cryo-electron microscopy and image reconstruction of Fab-labeled capsids to approximately 10A resolution followed by molecular modeling. Both antibodies engage residues on the protruding spikes but their epitopes and binding orientations differ. Steric interference effects limit maximum binding to approximately 50% average occupancy in each case. However, the occupancies of the two copies of a given epitope that are present on a single spike differ, reflecting subtle distinctions in structure and hence, binding affinity, arising from quasi-equivalence. The epitope for mAb88 is conformational but continuous, consisting of a loop-helix motif (residues 77-87) on one of the two polypeptide chains in the spike. In contrast, the epitope for mAb842, like most conformational epitopes, is discontinuous, consisting of a loop on one polypeptide chain (residues 74-78) combined with a loop-helix element (residues 78-83) on the other. The epitope of mAb842 is essentially identical with that previously mapped for mAb F11A4, although the binding orientations of the two monoclonal antibodies (mAbs) differ, as do their affinities measured by surface plasmon resonance. From the number of monoclonals (six) whose binding had to be characterized to give the first duplicate epitope, we estimate the total number of core antigen (cAg) epitopes to be of the order of 20. Given that different antibodies may share the same epitope, the potential number of distinct anti-cAg clones should be considerably higher. The observation that the large majority of cAg epitopes are conformational reflects the relative dimensions of a Fab (large) and the small size and close packing of the motifs that are exposed and accessible on the capsid surface.

Beta-helix model for the filamentous haemagglutinin adhesin of Bordetella pertussis and related bacterial secretory proteins.

Bordetella pertussis establishes infection by attaching to epithelial cells of the respiratory tract. One of its adhesins is filamentous haemagglutinin (FHA), a 500-A-long secreted protein that is rich in beta-structure and contains two regions, R1 and R2, of tandem 19-residue repeats. Two models have been proposed in which the central shaft is (i) a hairpin made up of a pairing of two long antiparallel beta-sheets; or (ii) a beta-helix in which the polypeptide chain is coiled to form three long parallel beta-sheets. We have analysed a truncated variant of FHA by electron microscopy (negative staining, shadowing and scanning transmission electron microscopy of unstained specimens): these observations support the latter model. Further support comes from detailed sequence analysis and molecular modelling studies. We applied a profile search method to the sequences adjacent to and between R1 and R2 and found additional "covert" copies of the same motifs that may be recognized in overt form in the R1 and R2 sequence repeats. Their total number is sufficient to support the tenet of the beta-helix model that the shaft domain--a 350 A rod--should consist of a continuous run of these motifs, apart from loop inserts. The N-terminus, which does not contain such repeats, was found to be weakly homologous to cyclodextrin transferase, a protein of known immunoglobulin-like structure. Drawing on crystal structures of known beta-helical proteins, we developed structural models of the coil motifs putatively formed by the R1 and R2 repeats. Finally, we applied the same profile search method to the sequence database and found several other proteins--all large secreted proteins of bacterial provenance--that have similar repeats and probably also similar structures.

Lysine 188 substitutions convert the pattern of proteasome activation by REGgamma to that of REGs alpha and beta.

11S REGs (PA28s) are multimeric rings that bind proteasomes and stimulate peptide hydrolysis. Whereas REGalpha activates proteasomal hydrolysis of peptides with hydrophobic, acidic or basic residues in the P1 position, REGgamma only activates cleavage after basic residues. We have isolated REGgamma mutants capable of activating the hydrolysis of fluorogenic peptides diagnostic for all three active proteasome beta subunits. The most robust REGgamma specificity mutants involve substitution of Glu or Asp for Lys188. REGgamma(K188E/D) variants are virtually identical to REGalpha in proteasome activation but assemble into less stable heptamers/hexamers. Based on the REGalpha crystal structure, Lys188 of REGgamma faces the aqueous channel through the heptamer, raising the possibility that REG channels function as substrate-selective gates. However, covalent modification of proteasome chymotrypsin-like subunits by 125I-YL3-VS demonstrates that REGgamma(K188E)'s activation of all three proteasome active sites is not due to relaxed gating. We propose that decreased stability of REGgamma(K188E) heptamers allows them to change conformation upon proteasome binding, thus relieving inhibition of the CT and PGPH sites normally imposed by the wild-type REGgamma molecule.

Functional domains of the ClpA and ClpX molecular chaperones identified by limited proteolysis and deletion analysis.

Escherichia coli ClpA and ClpX are ATP-dependent protein unfoldases that each interact with the protease, ClpP, to promote specific protein degradation. We have used limited proteolysis and deletion analysis to probe the conformations of ClpA and ClpX and their interactions with ClpP and substrates. ATP gamma S binding stabilized ClpA and ClpX such that that cleavage by lysylendopeptidase C occurred at only two sites. Both proteins were cleaved within in a loop preceding an alpha-helix-rich C-terminal domain. Although the loop varies in size and composition in Clp ATPases, cleavage occurred within and around a conserved triad, IG(F/L). Binding of ClpP blocked this cleavage, and prior cleavage at this site rendered both ClpA and ClpX defective in binding and activating ClpP, suggesting that this site is involved in interactions with ClpP. ClpA was also cut at a site near the junction of the two ATPase domains, whereas the second cleavage site in ClpX lay between its N-terminal and ATPase domains. ClpP did not block cleavage at these other sites. The N-terminal domain of ClpX dissociated upon cleavage, and the remaining ClpXDeltaN remained as a hexamer, associated with ClpP, and expressed ATPase, chaperone, and proteolytic activity. A truncated mutant of ClpA lacking its N-terminal 153 amino acids also formed a hexamer, associated with ClpP, and expressed these activities. We propose that the N-terminal domains of ClpX and ClpA lie on the outside ring surface of the holoenzyme complexes where they contribute to substrate binding or perform a gating function affecting substrate access to other binding sites and that a loop on the opposite face of the ATPase rings stabilizes interactions with ClpP and is involved in promoting ClpP proteolytic activity.

Translocation pathway of protein substrates in ClpAP protease.

Intracellular protein degradation, which must be tightly controlled to protect normal proteins, is carried out by ATP-dependent proteases. These multicomponent enzymes have chaperone-like ATPases that recognize and unfold protein substrates and deliver them to the proteinase components for digestion. In ClpAP, hexameric rings of the ClpA ATPase stack axially on either face of the ClpP proteinase, which consists of two apposed heptameric rings. We have used cryoelectron microscopy to characterize interactions of ClpAP with the model substrate, bacteriophage P1 protein, RepA. In complexes stabilized by ATPgammaS, which bind but do not process substrate, RepA dimers are seen at near-axial sites on the distal surface of ClpA. On ATP addition, RepA is translocated through approximately 150 A into the digestion chamber inside ClpP. Little change is observed in ClpAP, implying that translocation proceeds without major reorganization of the ClpA hexamer. When translocation is observed in complexes containing a ClpP mutant whose digestion chamber is already occupied by unprocessed propeptides, a small increase in density is observed within ClpP, and RepA-associated density is also seen at other axial sites. These sites appear to represent intermediate points on the translocation pathway, at which segments of unfolded RepA subunits transiently accumulate en route to the digestion chamber.

Capsid structure of Kaposi's sarcoma-associated herpesvirus, a gammaherpesvirus, compared to those of an alphaherpesvirus, herpes simplex virus type 1, and a betaherpesvirus, cytomegalovirus.

The capsid of Kaposi's sarcoma-associated herpesvirus (KSHV) was visualized at 24-A resolution by cryoelectron microscopy. Despite limited sequence similarity between corresponding capsid proteins, KSHV has the same T=16 triangulation number and much the same capsid architecture as herpes simplex virus (HSV) and cytomegalovirus (CMV). Its capsomers are hexamers and pentamers of the major capsid protein, forming a shell with a flat, close-packed, inner surface (the "floor") and chimney-like external protrusions. Overlying the floor at trigonal positions are (alpha beta(2)) heterotrimers called triplexes. The floor structure is well conserved over all three viruses, and the most variable capsid features reside on the outer surface, i.e., in the shapes of the protrusions and triplexes, in which KSHV resembles CMV and differs from HSV. Major capsid protein sequences from the three subfamilies have some similarity, which is closer between KSHV and CMV than between either virus and HSV. The triplex proteins are less highly conserved, but sequence analysis identifies relatively conserved tracts. In alphaherpesviruses, the alpha-subunit (VP19c in HSV) has a 100-residue N-terminal extension and an insertion near the C terminus. The small basic capsid protein sequences are highly divergent: whereas the HSV and CMV proteins bind only to hexons, difference mapping suggests that the KSHV protein, ORF65, binds around the tips of both hexons and pentons.

Transgenic mice expressing a mutant form of loricrin reveal the molecular basis of the skin diseases, Vohwinkel syndrome and progressive symmetric erythrokeratoderma.

Mutations in the cornified cell envelope protein loricrin have been reported recently in some patients with Vohwinkel syndrome (VS) and progressive symmetric erythrokeratoderma (PSEK). To establish a causative relationship between loricrin mutations and these diseases, we have generated transgenic mice expressing a COOH-terminal truncated form of loricrin that is similar to the protein expressed in VS and PSEK patients. At birth, transgenic mice (ML.VS) exhibited erythrokeratoderma with an epidermal barrier dysfunction. 4 d after birth, high-expressing transgenic animals showed a generalized scaling of the skin, as well as a constricting band encircling the tail and, by day 7, a thickening of the footpads. Histologically, ML. VS transgenic mice also showed retention of nuclei in the stratum corneum, a characteristic feature of VS and PSEK. Immunofluorescence and immunoelectron microscopy showed the mutant loricrin protein in the nucleus and cytoplasm of epidermal keratinocytes, but did not detect the protein in the cornified cell envelope. Transfection experiments indicated that the COOH-terminal domain of the mutant loricrin contains a nuclear localization signal. To determine whether the ML.VS phenotype resulted from dominant-negative interference of the transgene with endogenous loricrin, we mated the ML.VS transgenics with loricrin knockout mice. A severe phenotype was observed in mice that lacked expression of wild-type loricrin. Since loricrin knockout mice are largely asymptomatic (Koch, P.K., P. A. de Viragh, E. Scharer, D. Bundman, M.A. Longley, J. Bickenbach, Y. Kawachi, Y. Suga, Z. Zhou, M. Huber, et al., J. Cell Biol. 151:389-400, this issue), this phenotype may be attributed to expression of the mutant form of loricrin. Thus, deposition of the mutant protein in the nucleus appears to interfere with late stages of epidermal differentiation, resulting in a VS-like phenotype.

HIV-1 rev depolymerizes microtubules to form stable bilayered rings.

We describe a novel interaction between HIV-1 Rev and microtubules (MTs) that results in the formation of bilayered rings that are 44-49 nm in external diameter, 3.4-4.2 MD (megadaltons) in mass, and have 28-, 30-, or 32-fold symmetry. Ring formation is not sensitive to taxol, colchicine, or microtubule-associated proteins, but requires Mg(2+) and is inhibited by maytansine. The interaction involves the NH(2)-terminal domain of Rev and the face of tubulin exposed on the exterior of the MTs. The NH(2)-terminal half of Rev has unexpected sequence similarity to the tubulin-binding portion of the catalytic/motor domains of the microtubule-destabilizing Kin I kinesins. We propose a model wherein binding of Rev dimers to MTs at their ends causes segments of two neighboring protofilaments to peel off and close into rings, circumferentially containing 14, 15, or 16 tubulin heterodimers, with Rev bound on the inside. Rev has a strong inhibitory effect on aster formation in Xenopus egg extracts, demonstrating that it can interact with tubulin in the presence of normal levels of cellular constituents. These results suggest that Rev may interact with MTs to induce their destabilization, a proposition consistent with the previously described disruption of MTs after HIV-1 infection.

Maturation dynamics of a viral capsid: visualization of transitional intermediate states.

Typical of DNA bacteriophages and herpesviruses, HK97 assembles in two stages: polymerization and maturation. First, capsid protein polymerizes into closed shells; then, these precursors mature into larger, stabler particles. Maturation is initiated by proteolysis, producing a metastable particle primed for expansion-the major structural transition. We induced expansion in vitro by acidic pH and monitored the resulting changes by time-resolved X-ray diffraction and cryo-electron microscopy. The transition, which is not synchronized over the population, proceeds in a series of stochastically triggered subtransitions. Three distinct intermediates were identified, which are comparable to transitional states in protein folding. The intermediates' structures reveal the molecular events occurring during expansion. Integrated into a movie (see Dynamic Visualization below), they show capsid maturation as a dynamic process.

Visualization of substrate binding and translocation by the ATP-dependent protease, ClpXP.

Binding and internalization of a protein substrate by E. coli ClpXP was investigated by electron microscopy. In sideviews of ATP gamma S-stabilized ClpXP complexes, a narrow axial channel was visible in ClpX, surrounded by protrusions on its distal surface. When substrate lambda O protein was added, extra density attached to this surface. Upon addition of ATP, this density disappeared as lambda O was degraded. When ATP was added to proteolytically inactive ClpXP-lambda O complexes, the extra density transferred to the center of ClpP and remained inside ClpP after separation from ClpX. We propose that substrates of ATP-dependent proteases bind to specific sites on the distal surface of the ATPase, and are subsequently unfolded and translocated into the internal chamber of the protease.

Separation and crystallization of T = 3 and T = 4 icosahedral complexes of the hepatitis B virus core protein.

The icosahedral nucleocapsid of human hepatitis B virus is a homopolymer of the dimeric capsid protein also known as hepatitis B core antigen or HBcAg. Purified capsid protein obtained from an Escherichia coli expression system was reassembled into a mixture of T = 3 and T = 4 icosahedral particles consisting of 90 and 120 dimers, respectively. The two types of capsid were separated on a preparative scale by centrifugation through a sucrose gradient. In addition to this heterogeneity, the capsid protein has three cysteines, one of which has a great propensity for forming disulfide bonds between the two subunits, forming a dimer. To eliminate heterogeneity arising from oxidation, alanines were substituted for the cysteines. T = 3 and T = 4 capsids crystallized under similar conditions. Crystals of T = 3 capsids diffracted to approximately 8 A resolution; crystals of T = 4 capsids diffracted to 4 A resolution.

Structural invariance of constitutively active and inactive mutants of acanthamoeba myosin IC bound to F-actin in the rigor and ADP-bound states.

The three single-headed monomeric myosin I isozymes of Acanthamoeba castellanii (AMIs)-AMIA, AMIB, and AMIC-are among the best-studied of all myosins. We have used AMIC to study structural correlates of myosin's actin-activated ATPase. This activity is normally controlled by phosphorylation of Ser-329, but AMIC may be switched into constitutively active or inactive states by substituting this residue with Glu or Ala, respectively. To determine whether activation status is reflected in structural differences in the mode of attachment of myosin to actin, these mutant myosins were bound to actin filaments in the absence of nucleotide (rigor state) and visualized at 24-A resolution by using cryoelectron microscopy and image reconstruction. No such difference was observed. Consequently, we suggest that regulation may be affected not by altering the static (time-averaged) structure of AMIC but by modulating its dynamic properties, i.e., molecular breathing. The tail domain of vertebrate intestinal brush-border myosin I has been observed to swing through 31 degrees on binding of ADP. However, it was predicted on grounds of differing kinetics that any such effects with AMIC should be small [Jontes, J. D., Ostap, E. M., Pollard, T. D. & Milligan, R. A. (1998) J. Cell Biol. 141, 155-162]. We have confirmed this hypothesis by observing actin-associated AMIC in its ADP-bound state. Finally, we compared AMIC to brush-border myosin I and AMIB, which were previously studied under similar conditions. In each case, the shape and angle of attachment to F-actin of the catalytic domain is largely conserved, but the domain structure and disposition of the tail is distinctively different for each myosin.

Localization of the N terminus of hepatitis B virus capsid protein by peptide-based difference mapping from cryoelectron microscopy.

Recently, cryoelectron microscopy of isolated macromolecular complexes has advanced to resolutions below 10 A, enabling direct visualization of alpha-helical secondary structure. To help correlate such density maps with the amino acid sequences of the component proteins, we advocate peptide-based difference mapping, i. e., insertion of peptides, approximately 10 residues long, at targeted points in the sequence and visualization of these peptides as bulk labels in cryoelectron microscopy-derived difference maps. As proof of principle, we have appended an extraneous octapeptide at the N terminus of hepatitis B virus capsid protein and determined its location on the capsid surface by difference imaging at 11 A resolution. Hepatitis B virus capsids are icosahedral particles, approximately 300 A in diameter, made up of T-shaped dimers (subunit Mr, 16-21 kDa, depending on construct). The stems of the Ts protrude outward as spikes, whereas the crosspieces pack to form the contiguous shell. The two N termini per dimer reside on either side of the spike-stem, at the level at which it enters the shell. This location is consistent with formation of the known intramolecular disulfide bond between the cysteines at positions 61 and -7 (in the residual propeptide) in the "e-antigen" form of the capsid protein and has implications for why this clinically important antigen remains unassembled in vivo.

Shared motifs of the capsid proteins of hepadnaviruses and retroviruses suggest a common evolutionary origin.

The structure of the dimeric C-terminal domain of the HIV-1 capsid protein (CA), recently determined by X-ray crystallography (Gamble et al. (1997)), has a notable resemblance to the structure of the hepatitis B virus (HBV) capsid protein (Cp) dimer, previously determined by cryo-electron microscopy (Conway et al. (1997), Böttcher et al. (1997)). In both proteins, dimerization is effected by formation of a four-helix bundle, whereby each subunit contributes a helix-loop-helix and most of the interaction between subunits is mediated by one pair of helices. These are the first two observations of a motif that is common to the capsid proteins of two enveloped viruses and quite distinct from the eight-stranded anti-parallel beta-barrel found in most other virus capsid proteins solved to date (Harrison et al. (1996)). Motivated by the structural resemblance, we have examined retroviral and HBV capsid protein sequences and found weak but significant similarities between them. These similarities further support an evolutionary relationship between these two virus families of great medical importance -- the hepadnaviruses (e.g. HBV) and retroviruses (e.g. HIV).

Cornified cell envelope assembly: a model based on electron microscopic determinations of thickness and projected density.

In stratifying squamous epithelia, the cornified cell envelope (CE), a peripheral layer of crosslinked protein, is assembled sequentially from precursor proteins initially dispersed in the cytoplasm. Its major component is loricrin (37 kDa in mouse), which contributes from approx. 60% to >80% of the protein mass in different tissues. Despite its importance to the mechanical resilience and impenetrability of these tissues, detailed information has not been obtained on CE structure, even on such basic properties as its thickness or uniformity across a given CE or from tissue to tissue. To address this issue, we have studied CEs isolated from three murine epithelia, namely epidermis, forestomach and footpad, by electron microscopy of metal-shadowed specimens and scanning transmission electron microscopy (STEM) of unstained specimens. The former data reveal that the cytoplasmic surface is smoothly textured whereas the extracellular surface is corrugated, and that the average thickness is 15.3+/-1.2 nm, and strikingly uniform. Measurements of mass-per-unit-area from the STEM images yielded values of approx. 7.0+/-0.8 kDa/nm2, which were remarkably consistent over all three tissues. These data imply that the mature CE has a uniquely defined thickness. To explain its uniformity, we postulate that loricrin forms a molecular monolayer, not a variable number of multiple layers. In this scenario, the packing density is one loricrin monomer per 7 nm2, and loricrin should have an elongated shape, 2.5-3.0 nm wide by approx. 11 nm long. Moreover, we anticipate that any inter-tissue variations in the mechanical properties of CEs should depend more on protein composition and cross-linking pattern than on the thickness of the protein layer deposited.

Enzymatic and structural similarities between the Escherichia coli ATP-dependent proteases, ClpXP and ClpAP.

Escherichia coli ClpX, a member of the Clp family of ATPases, has ATP-dependent chaperone activity and is required for specific ATP-dependent proteolytic activities expressed by ClpP. Gel filtration and electron microscopy showed that ClpX subunits (Mr 46, 000) associate to form a six-membered ring (Mr approximately 280, 000) that is stabilized by binding of ATP or nonhydrolyzable analogs of ATP. ClpP, which is composed of two seven-membered rings stacked face-to-face, interacts with the nucleotide-stabilized hexamer of ClpX to form a complex that could be isolated by gel filtration. Electron micrographs of negatively stained ClpXP preparations showed side views of 1:1 and 2:1 ClpXP complexes in which ClpP was flanked on either one or both sides by a ring of ClpX. Thus, as was seen for ClpAP, a symmetry mismatch exists in the bonding interactions between the seven-membered rings of ClpP and the six-membered rings of ClpX. Competition studies showed that ClpA may have a slightly higher affinity (approximately 2-fold) for binding to ClpP. Mixed complexes of ClpA, ClpX, and ClpP with the two ATPases bound simultaneously to opposite faces of a single ClpP molecule were seen by electron microscopy. In the presence of ATP or nonhydrolyzable analogs of ATP, ClpXP had nearly the same activity as ClpAP against oligopeptide substrates (>10,000 min-1/tetradecamer of ClpP). Thus, ClpX and ClpA interactions with ClpP result in structurally analogous complexes and induce similar conformational changes that affect the accessibility and the catalytic efficiency of ClpP active sites.

At sixes and sevens: characterization of the symmetry mismatch of the ClpAP chaperone-assisted protease.

ClpAP, a typical energy-dependent protease, consists of a proteolytic component (ClpP) and a chaperone-like ATPase (ClpA). ClpP is composed of two apposed heptameric rings, whereas in the presence of ATP or ATPgammaS, ClpA is a single hexameric ring. Formation of ClpAP complexes involves a symmetry mismatch as sixfold ClpA stacks axially on one or both faces of sevenfold ClpP. We have analyzed these structures by cryo-electron microscopy. Our three-dimensional reconstruction of ClpA at 29-A resolution shows the monomer to be composed of two domains of similar size that, in the hexamer, form two tiers enclosing a large cavity. Cylindrical reconstruction of ClpAP reveals three compartments: the digestion chamber inside ClpP; a compartment between ClpP and ClpA; and the cavity inside ClpA. They are connected axially via narrow apertures, implying that substrate proteins should be unfolded to allow translocation into the digestion chamber. The cavity inside ClpA is structurally comparable to the "Anfinsen cage" of other chaperones and may play a role in the unfolding of substrates. A geometrical description of the symmetry mismatch was obtained by using our model of ClpA and the crystal structure of ClpP (Wang et al., 1997, Cell 91, 447-456) to identify the particular side views presented by both molecules in individual complexes. The interaction is characterized by a key pair of subunits, one of each protein. A small turn (8.6(o) = 2pi/42; equivalent to a 4-A shift) would transfer the key interaction to another pair of subunits. We propose that nucleotide hydrolysis results in rotation, facilitating the processive digestion of substrate proteins.