Molecular links between the E2 envelope glycoprotein and nucleocapsid core in Sindbis virus

A three-dimensional reconstruction of Sindbis virus at 7.0 Å resolution presented here provides a detailed view of the virion structure and includes structural evidence for key interactions that occur between the capsid protein (CP) and transmembrane (TM) glycoproteins E1 and E2. Based on crystal structures of component proteins and homology modeling, we constructed a nearly complete, pseudo-atomic model of the virus. Notably, this includes identification of the 33-residue cytoplasmic domain of E2 (cdE2), which follows a path from the E2 TM helix to the CP where it enters and exits the CP hydrophobic pocket and then folds back to contact the viral membrane. Modeling analysis identified three major contact regions between cdE2 and CP, and the roles of specific residues were probed by molecular genetics. This identified R393 and E395 of cdE2 and Y162 and K252 of CP as critical for virus assembly. The N-termini of the CPs form a contiguous network that interconnects 12 pentameric and 30 hexameric CP capsomers. A single glycoprotein spike cross-links three neighboring CP capsomers as might occur during initiation of virus budding.     

Synthesis of dioxane-based antiviral agents and evaluation of their biological activities as inhibitors of Sindbis virus replication

The crystal structure of the Sindbis virus capsid protein contains one or two solvent-derived dioxane molecules in the hydrophobic binding pocket. A bis-dioxane antiviral agent was designed by linking the two dioxane molecules with a three-carbon chain having R,R connecting stereochemistry, and a stereospecific synthesis was performed. This resulted in an effective antiviral agent that inhibited Sindbis virus replication with an EC(50) of 14 microM. The synthesis proceeded through an intermediate (R)-2-hydroxymethyl-[1,4]dioxane, which unexpectedly proved to be a more effecting antiviral agent than the target compound, as evidenced by its EC(50) of 3.4 microM as an inhibitor of Sindbis virus replication. Both compounds were not cytotoxic in uninfected BHK cells at concentrations of 1mM.     

The structure of barmah forest virus as revealed by cryo-electron microscopy at a 6-angstrom resolution has detailed transmembrane protein architecture and interactions

Barmah Forest virus (BFV) is a mosquito-borne alphavirus that infects humans. A 6-Å-resolution cryo-electron microscopy three-dimensional structure of BFV exhibits a typical alphavirus organization, with RNA-containing nucleocapsid surrounded by a bilipid membrane anchored with the surface proteins E1 and E2. The map allows details of the transmembrane regions of E1 and E2 to be seen. The C-terminal end of the E2 transmembrane helix binds to the capsid protein. Following the E2 transmembrane helix, a short α-helical endodomain lies on the inner surface of the lipid envelope. The E2 endodomain interacts with E1 transmembrane helix from a neighboring E1-E2 trimeric spike, thereby acting as a spacer and a linker between spikes. In agreement with previous mutagenesis studies, the endodomain plays an important role in recruiting other E1-E2 spikes to the budding site during virus assembly. The E2 endodomain may thus serve as a target for antiviral drug design.     

Placement of the structural proteins in Sindbis virus

The structure of the lipid-enveloped Sindbis virus has been determined by fitting atomic resolution crystallographic structures of component proteins into an 11-A resolution cryoelectron microscopy map. The virus has T=4 quasisymmetry elements that are accurately maintained between the external glycoproteins, the transmembrane helical region, and the internal nucleocapsid core. The crystal structure of the E1 glycoprotein was fitted into the cryoelectron microscopy density, in part by using the known carbohydrate positions as restraints. A difference map showed that the E2 glycoprotein was shaped similarly to E1, suggesting a possible common evolutionary origin for these two glycoproteins. The structure shows that the E2 glycoprotein would have to move away from the center of the trimeric spike in order to expose enough viral membrane surface to permit fusion with the cellular membrane during the initial stages of host infection. The well-resolved E1-E2 transmembrane regions form alpha-helical coiled coils that were consistent with T=4 symmetry. The known structure of the capsid protein was fitted into the density corresponding to the nucleocapsid, revising the structure published earlier.     

Design, synthesis, and evaluation of dioxane-based antiviral agents targeted against the Sindbis virus capsid protein

Dioxane-based antiviral agents targeted to the hydrophobic binding pocket of Sindbis virus capsid protein were designed by computer graphics molecular modeling and synthesized. Virus production using SIN-IRES-Luc and capsid assembly were monitored to evaluate antiviral activity. A compound with a three-carbon linker chain connecting two dioxane moieties inhibited virus production by 50% at a concentration of 40 microM, while (R)-hydroxymethyldioxane inhibited virus production by 50% at a concentration of 1 microM. Both compounds were not cytotoxic in uninfected BHK cells at concentrations of 1mM.     

Interactions of the cytoplasmic domain of Sindbis virus E2 with nucleocapsid cores promote alphavirus budding

Alphavirus budding from the plasma membrane occurs through the specific interaction of the nucleocapsid core with the cytoplasmic domain of the E2 glycoprotein (cdE2). Structural studies of the Sindbis virus capsid protein (CP) have suggested that these critical interactions are mediated by the binding of cdE2 into a hydrophobic pocket in the CP. Several molecular genetic studies have implicated amino acids Y400 and L402 in cdE2 as important for the budding of alphaviruses. In this study, we characterized the role of cdE2 residues in structural polyprotein processing, glycoprotein transport, and capsid interactions. Along with hydrophobic residues, charged residues in the N terminus of cdE2 were critical for the effective interaction of cores with cdE2, a process required for virus budding. Mutations in the C-terminal signal sequence region of cdE2 affected E2 protein transport to the plasma membrane, while nonbudding mutants that were defective in cdE2-CP interaction accumulated E2 on the plasma membrane. The interaction of cdE2 with cytoplasmic cores purified from infected cells and in vitro-assembled core-like particles suggests that cdE2 interacts with assembled cores to mediate budding. We hypothesize that these cdE2 interactions induce a change in the organization of the nucleocapsid core upon binding leading to particle budding and priming of the nucleocapsid cores for disassembly that is required for virus infection.     

Structural insights into viral determinants of nematode mediated Grapevine fanleaf virus transmission

Many animal and plant viruses rely on vectors for their transmission from host to host. Grapevine fanleaf virus (GFLV), a picorna-like virus from plants, is transmitted specifically by the ectoparasitic nematode Xiphinema index. The icosahedral capsid of GFLV, which consists of 60 identical coat protein subunits (CP), carries the determinants of this specificity. Here, we provide novel insight into GFLV transmission by nematodes through a comparative structural and functional analysis of two GFLV variants. We isolated a mutant GFLV strain (GFLV-TD) poorly transmissible by nematodes, and showed that the transmission defect is due to a glycine to aspartate mutation at position 297 (Gly297Asp) in the CP. We next determined the crystal structures of the wild-type GFLV strain F13 at 3.0 Å and of GFLV-TD at 2.7 Å resolution. The Gly297Asp mutation mapped to an exposed loop at the outer surface of the capsid and did not affect the conformation of the assembled capsid, nor of individual CP molecules. The loop is part of a positively charged pocket that includes a previously identified determinant of transmission. We propose that this pocket is a ligand-binding site with essential function in GFLV transmission by X. index. Our data suggest that perturbation of the electrostatic landscape of this pocket affects the interaction of the virion with specific receptors of the nematode''s feeding apparatus, and thereby severely diminishes its transmission efficiency. These data provide a first structural insight into the interactions between a plant virus and a nematode vector.     

Hepatitis C virus glycoprotein E2 contains a membrane-proximal heptad repeat sequence that is essential for E1E2 glycoprotein heterodimerization and viral entry

The E1 and E2 glycoproteins of hepatitis C virus form a noncovalently associated heterodimer that mediates viral entry. Glycoprotein E2 comprises a receptor-binding domain (residues 384-661) that is connected to the transmembrane domain (residues 716-746) via a highly conserved sequence containing a hydrophobic heptad repeat (residues 675-699). Alanine- and proline-scanning mutagenesis of the E2 heptad repeat revealed that Leu675, Ser678, Leu689, and Leu692 are important for E1E2 heterodimerization. Furthermore, Pro and Ala substitution of all but one heptad repeat residue (Ser678) blocked the entry of E1E2-HIV-1 pseudotypes into Huh7 cells, irrespective of an effect on heterodimerization. Two conserved prolines (Pro676 and Pro683), occupying consecutive b positions of the heptad, were not required for E1E2 heterodimerization; however, Pro683 was critical for viral entry. Thus, disruption of the predicted alpha-helical structure by proline at position 683 is important for E2 function. The inability of mutants to mediate viral entry was not explained by a loss of receptor binding function, because all mutants were able to interact with a recombinant form of the CD81 large extracellular loop. Chimeras formed between the E1 and E2 ectodomains and the transmembrane domains of flavivirus prM and E glycoproteins, respectively, were able to heterodimerize, although with lower efficiency in comparison with wild type E1E2. The heptad repeat of E2 therefore requires the native transmembrane domain for full heterodimerization and viral entry function. Our data indicate that the membraneproximal heptad repeat of E2 is functionally homologous to the stem of flavivirus E glycoproteins. We propose that E2 has mechanistic features in common with class II fusion proteins.     

Mutations in the E1 hydrophobic domain of rubella virus impair virus infectivity but not virus assembly

Rubella virus (RV) virions contain three structural proteins, a capsid protein that interacts with viral genomic RNA to form a nucleocapsid and two membrane glycoproteins, E2 and E1. We found that substitution of either an aspartic acid residue at Gly93 (G93D) or a glycine residue at Pro104 (P104G) in the internal hydrophobic domain of E1 affected virus infectivity but not virus assembly. Viruses carrying G93D and P104G mutations had impaired infectivity, reduced 1,000-fold and 10-fold, respectively. A revertant was isolated from the G93D mutant. Sequencing analysis showed that the substituted aspartic acid residue in G93D mutant had reverted to the original glycine residue, suggesting the involvement of Gly93 in membrane fusion during viral entry.     

The 2.3-angstrom structure of porcine circovirus 2

Porcine circovirus 2 (PCV2) is a T=1 nonenveloped icosahedral virus that has had severe impact on the swine industry. Here we report the crystal structure of an N-terminally truncated PCV2 virus-like particle at 2.3-Å resolution, and the cryo-electron microscopy (cryo-EM) image reconstruction of a full-length PCV2 virus-like particle at 9.6-Å resolution. This is the first atomic structure of a circovirus. The crystal structure revealed that the capsid protein fold is a canonical viral jelly roll. The loops connecting the strands of the jelly roll define the limited features of the surface. Sulfate ions interacting with the surface and electrostatic potential calculations strongly suggest a heparan sulfate binding site that allows PCV2 to gain entry into the cell. The crystal structure also allowed previously determined epitopes of the capsid to be visualized. The cryo-EM image reconstruction showed that the location of the N terminus, absent in the crystal structure, is inside the capsid. As the N terminus was previously shown to be antigenic, it may externalize through viral “breathing.”     

A single deletion in the membrane-proximal region of the Sindbis virus glycoprotein E2 endodomain blocks virus assembly

The envelopment of the Sindbis virus nucleocapsid in the modified cell plasma membrane involves a highly specific interaction between the capsid (C) protein and the endodomain of the E2 glycoprotein. We have previously identified a domain of the Sindbis virus C protein involved in binding to the E2 endodomain (H. Lee and D. T. Brown, Virology 202:390-400, 1994). The C-E2 binding domain resides in a hydrophobic cleft with C Y180 and W247 on opposing sides of the cleft. Structural modeling studies indicate that the E2 domain, which is proposed to bind the C protein (E2 398T, 399P, and 400Y), is located at a sufficient distance from the membrane to occupy the C protein binding cleft (S. Lee, K. E. Owen, H. K. Choi, H. Lee, G. Lu, G. Wengler, D. T. Brown, M. G. Rossmann, and R. J. Kuhn, Structure 4:531-541, 1996). To measure the critical spanning length of the E2 endodomain which positions the TPY domain into the putative C binding cleft, we have constructed a deletion mutant, DeltaK391, in which a nonconserved lysine (E2 K391) at the membrane-cytoplasm junction of the E2 tail has been deleted. This mutant was found to produce very low levels of virus from BHK-21 cells due to a defect in an unidentified step in nucleocapsid binding to the E2 endodomain. In contrast, DeltaK391 produced wild-type levels of virus from tissue-cultured mosquito cells. We propose that the phenotypic differences displayed by this mutant in the two diverse host cells arise from fundamental differences in the lipid composition of the insect cell membranes which affect the physical and structural properties of membranes and thereby virus assembly. The data suggest that these viruses have evolved properties adapted specifically for assembly in the diverse hosts in which they grow.     

Structure of adeno-associated virus serotype 5

Adeno-associated virus serotype 5 (AAV5) requires sialic acid on host cells to bind and infect. Other parvoviruses, including Aleutian mink disease parvovirus (ADV), canine parvovirus (CPV), minute virus of mice, and bovine parvovirus, also bind sialic acid. Hence, structural homology may explain this functional homology. The amino acids required for CPV sialic acid binding map to a site at the icosahedral twofold axes of the capsid. In contrast to AAV5, AAV2 does not bind sialic acid, but rather binds heparan sulfate proteoglycans at its threefold axes of symmetry. To explore the structure-function relationships among parvoviruses with respect to cell receptor attachment, we determined the structure of AAV5 by cryo-electron microscopy (cryo-EM) and image reconstruction at a resolution of 16 A. Surface features common to some parvoviruses, namely depressions encircling the fivefold axes and protrusions at or surrounding the threefold axes, are preserved in the AAV5 capsid. However, even though there were some similarities, a comparison of the AAV5 structure with those of ADV and CPV failed to reveal a feature which could account for the sialic acid binding phenotype common to all three viruses. In contrast, the overall surface topologies of AAV5 and AAV2 are similar. A pseudo-atomic model generated for AAV5 based on the crystal structure of AAV2 and constrained by the AAV5 cryo-EM envelope revealed differences only in surface loop regions. Surprisingly, the surface topologies of AAV5 and AAV2 are remarkably similar to that of ADV despite only exhibiting approximately 20% identity in amino acid sequences. Thus, capsid surface features are shared among parvoviruses and may not be unique to their replication phenotypes, i.e., whether they require a helper or are autonomous. Furthermore, specific surface features alone do not explain the variability in carbohydrate requirements for host cell receptor interactions among parvoviruses.     

Structures of the Procapsid and Mature Virion of Enterovirus 71 Strain 1095

Enterovirus 71 (EV71) is an important emerging human pathogen with a global distribution and presents a disease pattern resembling poliomyelitis with seasonal epidemics that include cases of severe neurological complications, such as acute flaccid paralysis. EV71 is a member of the Picornaviridae family, which consists of icosahedral, nonenveloped, single-stranded RNA viruses. Here we report structures derived from X-ray crystallography and cryoelectron microscopy (cryo-EM) for the 1095 strain of EV71, including a putative precursor in virus assembly, the procapsid, and the mature virus capsid. The cryo-EM map of the procapsid provides new structural information on portions of the capsid proteins VP0 and VP1 that are disordered in the higher-resolution crystal structures. Our structures solved from virus particles in solution are largely in agreement with those from prior X-ray crystallographic studies; however, we observe small but significant structural differences for the 1095 procapsid compared to a structure solved in a previous study (X. Wang, W. Peng, J. Ren, Z. Hu, J. Xu, Z. Lou, X. Li, W. Yin, X. Shen, C. Porta, T. S. Walter, G. Evans, D. Axford, R. Owen, D. J. Rowlands, J. Wang, D. I. Stuart, E. E. Fry, and Z. Rao, Nat. Struct. Mol. Biol. 19:424–429, 2012) for a different strain of EV71. For both EV71 strains, the procapsid is significantly larger in diameter than the mature capsid, unlike in any other picornavirus. Nonetheless, our results demonstrate that picornavirus capsid expansion is possible without RNA encapsidation and that picornavirus assembly may involve an inward radial collapse of the procapsid to yield the native virion.     

Molecular mechanisms of virus spread and virion components as tools of virulence. A review

Despite of differences in replication strategy among virus families, some basic principles have remained similar. Analogous mechanisms govern virus entry into cells and the use of enzymes which direct the replication of the virus genome. The function of many cell surface receptors (such as glycosoaminoglycans, glycoproteins, proteins) which interact with viral capsid proteins or envelope glycoproteins has recently been elucidated. The list of cellular receptors (Table I) is still far from being final. The capsid components, similarly as the envelope glycoproteins, may form specific pocket like sites, which interact with the cell surface receptors. Neutralizing antibodies usually react with antigenic domains adjacent to the receptor binding site(s) and hamper the close contact inevitable for virion attachment. In the case of more complex viruses, such as herpes simplex virus, different viral glycoproteins interact with several cellular receptors. At progressed phase of adsorption the virions are engulfed into endocytic vesicles and the virion fusion domain(s) become(s) activated. The outer capsid components of reoviruses which participate in adsorption and fusion may get activated already in the lumen of digestive tract, i.e. before their engulfment by resorptive epithelium cells. Activation of the hydrophobic fusion domain(s) is a further important step allowing to pass through the lipid bilayer when penetrating the cell membrane in order to reach the cytosol. Activation of the virion fusion domain is accomplished by a conformation change, which occurs at acid pH (influenza virus hemagglutinin, sigma 1 protein of the reovirus particle) and/or after protease treatment. The herpes simplex virus fusion factors (gD and gH) undergo conformation changes by a pH-independent mechanism triggered due to interaction with the cell surface receptor(s) and mediated by mutual interactions with the viral envelope glycoproteins. The virion capsid or envelope components participating in the entry and membrane fusion are not the only tools of virulence. The correct function of virus coded proteins, which participate in replication of the viral genome, and/or in the supply of necessary nucleotides, may be very essential. In the case of enteroviruses, which RNA interacts with ribosomes directly, the correct configuration of the non-coding viral RNA sequence is crucial for initiation of translation occurring in the absence of the classical "cap" structure.     

Novel inhibitor for prolyl aminopeptidase from Serratia marcescens and studies on the mechanism of substrate recognition of the enzyme using the inhibitor

Prolyl aminopeptidase from Serratia marcescens hydrolyzed x-beta-naphthylamides (x=prolyl, alanyl, sarcosinyl, L-alpha-aminobutylyl, and norvalyl), which suggested that the enzyme has a pocket for a five-member ring. Based on the substrate specificity, novel inhibitors of Pro, Ala, and Sar having 2-tert-butyl-[1,3,4]oxadiazole (TBODA) were synthesized. The K(i) value of Pro-TBODA, Ala-TBODA, and Sar-TBODA was 0.5 microM, 1.6 microM, and 12mM, respectively. The crystal structure of enzyme-Pro-TBODA complex was determined. Pro-TBODA was located at the active site. Four electrostatic interactions were located between the enzyme and the amino group of Pro inhibitors (Glu204:0E1-N:Inh, Glu204:0E2-N:Inh, Glu232:0E1-N:Inh, and Gly46:O-N:Inh), and the residue of the inhibitors was inserted into the hydrophobic pocket composed of Phe139, Leu141, Leu146, Tyr149, Tyr150, and Phe236. The roles of Phe139, Tyr149, and Phe236 in the hydrophobic pocket and Glu204 and Glu232 in the electrostatic interactions were confirmed by site-directed mutagenesis, which indicated that the molecular recognition of proline is achieved through four electrostatic interactions and an insertion in the hydrophobic pocket of the enzyme.     

Rapid increase of near atomic resolution virus capsid structures determined by cryo-electron microscopy

The recent technological advances in electron microscopes, detectors, as well as image processing and reconstruction software have brought single particle cryo-electron microscopy (cryo-EM) into prominence for determining structures of bio-molecules at near atomic resolution. This has been particularly true for virus capsids, ribosomes, and other large assemblies, which have been the ideal specimens for structural studies by cryo-EM approaches. An analysis of time series metadata of virus structures on the methods of structure determination, resolution of the structures, and size of the virus particles revealed a rapid increase in the virus structures determined by cryo-EM at near atomic resolution since 2010. In addition, the data highlight the median resolution (～3.0?Å) and size (～310.0?Å in diameter) of the virus particles determined by X-ray crystallography while no such limits exist for cryo-EM structures, which have a median diameter of 508?Å. Notably, cryo-EM virus structures in the last four years have a median resolution of 3.9?Å. Taken together with minimal sample requirements, not needing diffraction quality crystals, and being able to achieve similar resolutions of the crystal structures makes cryo-EM the method of choice for current and future virus capsid structure determinations.     

How does a symmetric dimer recognize an asymmetric substrate? A substrate complex of HIV-1 protease

The crystal structure of an actual HIV-1 protease-substrate complex is presented at 2.0 A resolution (R-value of 19.7 % (R(free) 23.3 %)) between an inactive variant (D25N) of HIV-1 protease and a long substrate peptide, Lys-Ala-Arg-Val-Leu-Ala-Glu-Ala-Met-Ser, which covers a full binding epitope of capsid(CA)-p2, cleavage site. The substrate peptide is asymmetric in both size and charge distribution. To accommodate this asymmetry the two protease monomers adopt different conformations burying a total of 1038 A(2) of surface area at the protease-substrate interface. The specificity for the CA-p2 substrate peptide is mainly hydrophobic, as most of the hydrogen bonds are made with the backbone of the peptide substrate. Two water molecules bridge the two monomers through the loops Gly49-Gly52 (Gly49'-Gly52') and Pro79'-Val82' (Pro79-Val82). When other complexes are compared, the mobility of these loops is correlated with the content of the P1 and P1' sites. Interdependence of the conformational changes allows the protease to exhibit its wide range of substrate specificity.     

The role of subunit hinges and molecular "switches" in the control of viral capsid polymorphism

The coat protein (CP) of cowpea chlorotic mottle virus assembles exclusively into a T=3 capsid in vivo and, under proper conditions, in vitro. The N-terminal domain of CP has been implicated in proper assembly and was viewed as a required switch for mediating hexamer and pentamer formation in T=3 assembly. We observed that a mutant CP lacking most of the N-terminal domain, NDelta34, assembles, in vitro, into statistically predictable numbers of: native-like T=3 capsids of 90 dimers; "T=2" capsids of 60 dimers; T=1 capsids of 30 dimers. We generated cryo-EM image reconstructions of each form and built pseudo-atomic models based on the subunits from the crystal structure of plant-derived T=3 virus allowing a detailed comparison of stabilizing interactions in the three assemblies. The statistical nature of the distribution of assembly products and the observed structures indicates that the N-terminus of CP is not a switch that is required to form the proper ratio of hexamers and pentamers for T=3 assembly; rather, it biases the direction of assembly to T=3 particles from the possibilities available to NDelta34 through flexible dimer hinges and variations in subunit contacts. Our results are consistent with a pentamer of dimers (PODs) nucleating assembly in all cases but subunit dimers can be added with different trajectories that favor specific T=3 or T=1 global particle geometries. Formation of the "T=2" particles appears to be fundamentally different in that they not only nucleate with PODs, but assembly propagates by the addition of mostly, if not exclusively PODs generating an entirely new subunit interface in the process. These results show that capsid geometry is flexible and may readily adapt to new requirements as the virus evolves.     

Structural analysis of lipid complexes of GM2-activator protein

The GM2-activator protein (GM2-AP) is a small lysosomal lipid transfer protein essential for the hydrolytic conversion of ganglioside GM2 to GM3 by beta-hexosaminidase A. The crystal structure of human apo-GM2-AP is known to consist of a novel beta-cup fold with a spacious hydrophobic interior. Here, we present two new structures of GM2-AP with bound lipids, showing two different lipid-binding modes within the apolar pocket. The 1.9A structure with GM2 bound shows the position of the ceramide tail and significant conformational differences among the three molecular copies in the asymmetric unit. The tetrasaccharide head group is not visible and is presumed to be disordered. However, its general position could be established through modeling. The structure of a low-pH crystal, determined at 2.5A resolution, has a significantly enlarged hydrophobic channel that merges with the apolar pocket. Electron density inside the pocket and channel suggests the presence of a trapped phospholipid molecule. Structure alignments among the four crystallographically unique monomers provide information on the potential role for lipid binding of flexible chain segments at the rim of the cavity opening. Two discrete orientations of the S130-T133 loop define an open and a closed configuration of the hydrophobic channel that merges with the apolar pocket. We propose: (i) that the low-pH structure represents an active membrane-binding conformation; (ii) that the mobile S130-T133 loop serves as a gate for passage of ligand into the apolar pocket; and (iii) that this loop and the adjacent apolar V59-W63 loop form a surface patch with two exposed tryptophan residues that could interface with lipid bilayers.     

Bacteriophage lambda stabilization by auxiliary protein gpD: timing, location, and mechanism of attachment determined by cryo-EM

We report the cryo-EM structure of bacteriophage lambda and the mechanism for stabilizing the 20-A-thick capsid containing the dsDNA genome. The crystal structure of the HK97 bacteriophage capsid fits most of the T = 7 lambda particle density with only minor adjustment. A prominent surface feature at the 3-fold axes corresponds to the cementing protein gpD, which is necessary for stabilization of the capsid shell. Its position coincides with the location of the covalent cross-link formed in the docked HK97 crystal structure, suggesting an evolutionary replacement of this gene product in lambda by autocatalytic chemistry in HK97. The crystal structure of the trimeric gpD, in which the 14 N-terminal residues required for capsid binding are disordered, fits precisely into the corresponding EM density. The N-terminal residues of gpD are well ordered in the cryo-EM density, adding a strand to a beta-sheet formed by the capsid proteins and explaining the mechanism of particle stabilization.