Head maturation pathway of bacteriophages T4 and T2. III. Isolation and characterization of particles produced by mutants in gene 17

We have isolated and characterized two types of particles produced in comparable amounts by mutants in gene 17: the empty large particle and the empty small particle. Dimensions, morphology, stability, and protein composition of the empty large particle are very similar to those of the capsids or empty heads of mature phage. The other type of particle (empty small particle) is very similar in dimensions and stability to the prehead, but differs in that it is composed of processed proteins (gp23, gp24, IpIII). Structural analysis has shown that the protein subunits of the empty small particles are arranged in an unexpanded type of lattice (11.2 to 11.3 nm), whereas the empty large particles have an expanded lattice (13 nm). The characterization of the empty small particle as being composed of cleaved proteins, but still unexpanded, shows that the expansion of the T4 head shell is not necessarily linked to the cleavage of the structural proteins.     

Conformational changes of a viral capsid protein. Thermodynamic rationale for proteolytic regulation of bacteriophage T4 capsid expansion, co-operativity, and super-stabilization by soc binding.

We have used differential scanning calorimetry in conjunction with cryo-electron microscopy to investigate the conformational transitions undergone by the maturing capsid of phage T4. Its precursor shell is composed primarily of gp23 (521 residues): cleavage of gp23 to gp23* (residues 66 to 521) facilitates a concerted conformational change in which the particle expands substantially, and is greatly stabilized. We have now characterized the intermediate states of capsid maturation; namely, the cleaved/unexpanded, state, which denatures at tm = 60 degrees C, and the uncleaved/expanded state, for which tm = 70 degrees C. When compared with the precursor uncleaved/unexpanded state (tm = 65 degrees C), and the mature cleaved/expanded state (tm = 83 degrees C, if complete cleavage precedes expansion), it follows that expansion of the cleaved precursor (delta tm approximately +23 degrees C) is the major stabilizing event in capsid maturation. These observations also suggest an advantage conferred by capsid protein cleavage (some other phage capsids expand without cleavage): if the gp23-delta domains (residues 1 to 65) are not removed by proteolysis, they impede formation of the stablest possible bonding arrangement when expansion occurs, most likely by becoming trapped at the interface between neighboring subunits or capsomers. Icosahedral capsids denature at essentially the same temperatures as tubular polymorphic variants (polyheads) for the same state of the surface lattice. However, the thermal transitions of capsids are considerably sharper, i.e. more co-operative, than those of polyheads, which we attribute to capsids being closed, not open-ended. In both cases, binding of the accessory protein soc around the threefold sites on the outer surface of the expanded surface lattice results in a substantial further stabilization (delta tm = +5 degrees C). The interfaces between capsomers appear to be relatively weak points that are reinforced by clamp-like binding of soc. These results imply that the "triplex" proteins of other viruses (their structural counterparts of soc) are likely also to be involved in capsid stabilization. Cryo-electron microscopy was used to make conclusive interpretations of endotherms in terms of denaturation events. These data also revealed that the cleaved/unexpanded capsid has an angular polyhedral morphology and has a pronounced relief on its outer surface. Moreover, it is 14% smaller in linear dimensions than the cleaved/expanded capsid, and its shell is commensurately thicker.     

Head maturation pathway of bacteriophages T4 and T2. V. Maturable epsilon-particle accumulating an acridine-treated bacteriophage T4-infected cells.

A maturable head-related particle of bacteriophage T4 has been identified and characterized. This epsilon-particle has the same size as the prehead, but its shell is made of the cleaved product of gene 23 (gp23*). It contains internal matter, most likely the processed core proteins, which is lost or modified by experimental manipulations. It accumulates, together with partially filled ("grizzled") heads, in T4 infected cells that are treated with 9-aminoacridine. On sections of "well-preserved" cells the epsilon-particles are not identifiable with certainty; a more or less empty breakdown product of them becomes visible when cytoplasmic leakage is induced. The number of particles per cell is then in agreement with the biochemically and with the number of particles counted in lysates. Morphologically and biochemically, the isolated epsilon-particles closely resemble the empty small particles of 17- -infected cells described in previous papers of this series. Both are composed of gp23* and are still unexpanded, so that they are not yet able to bind the minor head proteins soc and hoc. We discuss the possibility of the epsilon-particle being an intermediate on the normal T4 wild-type head maturation pathway.     

Isolation and characterization of bacteriophage T4 mutant preheads.

To determine the function of individual gene products in the assembly and maturation of the T4 prehead, we have isolated and characterized aberrant preheads produced by mutations in three of the T4 head genes. Mutants in gene 21, which codes for the T4 maturation proteases, produce rather stable preheads whose morphology and protein composition are consistent with a wild-type prehead blocked in the maturation cleavages. Mutants in gene 24 produce similar structures which are unstable because they have gaps at all of their icosahedral vertices except the membrane attachment site. In addition, greatly elongated "giant preheads" are produced, suggesting that in the absence of P24 at the vertices, the distal cap of the prehead is unstable, allowing abnormal elongation of broth the prehead core and its shell. Vertex completion by P24 is required to allow the maturation cleavages to occur, and 24- preheads can be matured to capsids in vitro by the addition of P24. Preheads produced by a temperature-sensitive mutant in gene 23 are deficient in core proteins. We show that the shell of these preheads has the expanded lattice characteristic of the mature capsid as well as the binding sites for the proteins hoc and soc, even though none of the maturation cleavage takes place. We also show that 21- preheads composed of wild-type P23 can be expanded in vitro without cleavage.     

Structure of T4 polyheads: II. A pathway of polyhead transformations as a model for T4 capsid maturation

We have studied the aberrant tubular polyheads of bacteriophages T4D and T2L as a model system for capsid maturation. Six different types of polyhead surface lattice morphology, and the corresponding protein compositions are reported and discussed. Using in vitro systems to induce transformations between particular polyhead types, we have deduced that the structural classes represent successive points in a transitional pathway. In the first step, coarse polyheads (analogous to the prohead τ-particle) are proteolytically cleaved by a phagecoded protease, a fragment of the gene 21 product. This cleavage of P23 to P23¹ induces a co-operative lattice transformation in the protein of the surface shell, to a conformation equivalent to that of T2L giant phage capsids. These polyheads (derived either from T4 or T2L lysates) can accept further T4-coded proteins. In doing so, they pass through intermediate structural states, eventually reaching an end point whose unit cell morphology is indistinguishable from that of the giant T4 capsids. At least one protein (called soc (Ishii & Yanagida, 1975)) is bound stoichiometrically to P23¹ in the end-state conformation. The simulation of several aspects of capsid maturation (cleavage of P23 to P23¹, stabilization, and lattice expansion) in the polyhead pathway suggest that it parallels the major events of phage T-even capsid maturation, decoupled from any involvement of DNA packaging.     

The maturation-dependent conformational change of the major capsid protein of bacteriophage T4 involves a substantial change in secondary structure

We have investigated the conformational basis of the expansion transformation that occurs upon maturation of the bacteriophage T4 prohead, by using laser Raman spectroscopy to determine the secondary structure of the major capsid protein in both the precursor and the mature states of the surface lattice. This transformation involves major changes in the physical, chemical, and immunological properties of the capsid and is preceded in vivo by processing of its major protein, gp23 (56 kDa), to gp23* (49 kDa), by proteolysis of its N-terminal gp23-delta domain. The respective secondary structures of gp23 in the unexpanded state, and of gp23* in the expanded state, were determined from the laser Raman spectra of polyheads, tubular polymorphic variants of the capsid. Similar measurements were also made on uncleaved polyheads that had been expanded in vitro and, for reference, on thermally denatured polyheads. We find that, with or without cleavage of gp23, expansion is accompanied by substantial changes in secondary structure, involving a major reduction in alpha-helix content and an increase in beta-sheet. The beta-sheet contents of gp23* or gp23 in the expanded state of the surface lattice, and even of gp23 in the unexpanded state, are sufficient for a domain with the "jellyroll" fold of antiparallel beta-sheets, previously detected in the capsid proteins of other icosahedral viruses.(ABSTRACT TRUNCATED AT 250 WORDS)     

DNA packaging of bacteriophage T4 proheads in vitro. Evidence that prohead expansion is not coupled to DNA packaging

We developed a system for DNA packaging of isolated bacteriophage T4 proheads in vitro and studied the role of prohead expansion in DNA packaging. Biologically active proheads have been purified from a number of packaging-deficient mutant extracts. The cleaved mature prohead is the active structural precursor for the DNA packaging reaction. Packaging of proheads requires ATP, Mg2+ and spermidine, and is stimulated by polyethylene glycol and dextran. Predominantly expanded proheads (ELPs) are produced at 37 degrees C and predominantly unexpanded proheads (ESPs) are produced at 20 degrees C. Both the expanded and unexpanded proheads are active in DNA packaging in vitro. This is based on the observations that (1) both ESPs and ELPs purified by chromatography on DEAE-Sephacel showed DNA packaging activity; (2) apparently homogeneous ELPs prepared by treatment with sodium dodecyl sulfate (which dissociates ESPs) retained significant biological activity; (3) specific precipitation of ELPs with anti-hoc immunoglobulin G resulted in loss of DNA packaging activity; and (4) ESPs upon expansion in vitro to ELPs retained packaging activity. Therefore, contrary to the models that couple DNA packaging to head expansion, in T4 the expansion and packaging appear to be independent, since the already expanded DNA-free proheads can be packaged in vitro. We therefore propose that the unexpanded to expanded prohead transition has evolved to stabilize the capsid and to reorganize the prohead shell functionally from a core-interacting to a DNA-interacting inner surface.     

Studies related to the head-maturation pathway of bacteriophages T4 and T2: I. Morphology and kinetics of intracellular particles produced by mutants in the maturation genes

Mutants in the genes governing the maturation of the head of bacteriophage T4 and in gene 24 were studied by electron microscopy of thin sections. We define morphologically: black particles, comprising mature, stable heads and immature, fragile heads, which break down upon lysis; grizzled particles, which apparently are partially filled or partially emptied; empty large particles without DNA or core Which are all the same size as normal heads; empty small particles without DNA and without core which are of the size of the τ particle, which is the prehead of phage T4. The study of single and double mutants of the maturation genes demonstrates that the phenotypes are only different by the proportions of the different particles made except for 17^? where only empty small and empty large particles accumulate. The mutants in gene 24 are epistatic on all other mutants. Mutants in gene 17 are epistatic on the remaining ones. The results are consistent with the hypothesis that the products of several of the maturation genes act on DNA to render it competent for packaging while the others act directly on the particle. By this uncoupling, bypasses and abortive pathways can result.     

Time-resolved molecular dynamics of bacteriophage HK97 capsid maturation interpreted by electron cryo-microscopy and X-ray crystallography

The bacteriophage HK97 capsid is a molecular machine that exhibits large-scale conformational rearrangements of its 420 identical protein subunits during capsid maturation. Immature empty capsids, termed Prohead II, assemble in vivo in an Escherichia coli expression system. Maturation of these particles may be induced in vitro, converting them into Head II capsids that are indistinguishable in conformation from the capsid of an infectious phage particle. One method of in vitro maturation requires acidification to drive the reaction through two expansion intermediates (EI-I, EI-II) to its penultimate particle state (EI-III), which has 86% more internal volume than Prohead II. Neutralization of EI-III produces the fully mature capsid, Head II. The three expansion intermediates and the acid expansion pathway were characterized by cryo-EM analysis and 3D reconstruction. We now report that, although large-scale structural changes are involved, the electron density maps for these intermediate states are readily interpreted in terms of quasi-atomic models based on subunit structures determined by prior crystallographic analysis of Head II. Progression through the expansion intermediate states primarily represents rigid-body rotations and translations of the subunits, accompanied by refolding of two small regions, the N-terminal arm and a beta-hairpin called the E-loop. Movies made with these pseudo-atomic coordinates and the Head II X-ray coordinates illuminate various aspects of the maturation pathway in the course of which the pattern of inter-subunit interactions is sequentially transformed while the integrity of the capsid is maintained.     

Development of Coliphage T5: Ultrastructural and Biochemical Studies

Electron microscopic studies of Escherichia coli infected with bacteriophage T5(+) have revealed that host nuclear material disappeared before 9 min after infection. This disappearance seemed to correspond to the breakdown of host deoxyribonucleic acid (DNA) into acid-soluble fragments. Little or no host DNA thymidine was reincorporated into phage DNA, except in the presence of 5-fluorodeoxyuridine (FUdR). Progeny virus particles were observed in the cytoplasm 20 min postinfection. Most of these particles were in the form of hexagonal-shaped heads or capsids, which were filled with electron-dense material (presumably T5 DNA). A small percentage (3 to 4%) of the phage heads appeared empty. On rare occasions, crystalline arrays of empty heads were observed. Nalidixic acid, hydroxyurea, and FUdR substantially inhibited replication of T5 DNA. However, these agents did not prevent virus-induced degradation of E. coli DNA. Most of the phage-specified structures seen in T5(+)-infected cells treated with FUdR or with nalidixic were in the form of empty capsids. Infected cells treated with hydroxyurea did not contain empty capsids. When E. coli F was infected with the DO mutant T5 amH18a (restrictive conditions), there was a small amount of DNA synthesis. Such cells contained only empty capsids, but their numbers were few in comparison to those in cells infected under permissive conditions or infected with T5(+). The cells also failed to lyse. These results confirm other reports which suggest that DNA replication is not required for the synthesis of late proteins. The data also indicate that DNA replication influences the quantity of viral structures being produced.     

Three-dimensional structure of vaccinia virus-produced human papillomavirus type 1 capsids.

The capsid proteins of papillomavirus self-assemble to form empty capsids or virus-like particles that appear quite similar to naturally occurring virions by conventional electron microscopy. To characterize such virus-like particles more fully, cryoelectron microscopy and image analysis techniques were used to generate three-dimensional reconstructions of capsids produced by vaccinia virus recombinants (V capsids) that expressed human papillomavirus type 1 L1 protein only or both L1 and L2 proteins. All V capsids had 72 pentameric capsomers arranged on a T = 7 icosahedral lattice. Each particle (approximately 60 nm in diameter) consisted of an approximately 2-nm-thick shell of protein with a radius of 22 nm with capsomers that extend approximately 6 nm from the shell. At a resolution of 3.5 nm, both V capsid structures appear identical to the capsid structure of native human papillomavirus type 1 (T. S. Baker, W. W. Newcomb, N. H. Olson, L. M. Cowsert, C. Olson, and J. C. Brown, Biophys. J. 60:1445-1456, 1991), thus implying that expressed and native capsids are structurally equivalent.     

Epitope insertion at the N-terminal molecular switch of the rabbit hemorrhagic disease virus T = 3 capsid protein leads to larger T = 4 capsids

Viruses need only one or a few structural capsid proteins to build an infectious particle. This is possible through the extensive use of symmetry and the conformational polymorphism of the structural proteins. Using virus-like particles (VLP) from rabbit hemorrhagic disease virus (RHDV) as a model, we addressed the basis of calicivirus capsid assembly and their application in vaccine design. The RHDV capsid is based on a T=3 lattice containing 180 identical subunits (VP1). We determined the structure of RHDV VLP to 8.0-Å resolution by three-dimensional cryoelectron microscopy; in addition, we used San Miguel sea lion virus (SMSV) and feline calicivirus (FCV) capsid subunit structures to establish the backbone structure of VP1 by homology modeling and flexible docking analysis. Based on the three-domain VP1 model, several insertion mutants were designed to validate the VP1 pseudoatomic model, and foreign epitopes were placed at the N- or C-terminal end, as well as in an exposed loop on the capsid surface. We selected a set of T and B cell epitopes of various lengths derived from viral and eukaryotic origins. Structural analysis of these chimeric capsids further validates the VP1 model to design new chimeras. Whereas most insertions are well tolerated, VP1 with an FCV capsid protein-neutralizing epitope at the N terminus assembled into mixtures of T=3 and larger T=4 capsids. The calicivirus capsid protein, and perhaps that of many other viruses, thus can encode polymorphism modulators that are not anticipated from the plane sequence, with important implications for understanding virus assembly and evolution.     

Proteolytic cleavage and structural transformation: Their relationship in bacteriophage T4 capsid maturation

Giant T4 phage capsoids formed in canavanine-treated cultures infected by phage mutants in genes 21 and 17, respectively, differ with regard to cleavage of the major capsid protein, gp 23, and in the fine structure of their hexagonal surface lattices. Quantitative computer processing of electron micrographs shows that the significant differences in capsomer morphology amount to six symmetrically placed features present in the uncleaved hexamer but absent after cleavage. These features may be related with the N-terminal portions of gp 23 monomers excised by phage-specific proteolysis. Cleaved 17^? giants can be induced to undergo a further structural transformation (expansion). Structural characteristics of partially transformed giant particles give clues about the dynamics of the cleavage and expansion transformations. Both processes appear to be polar, initiating in one cap and propagating along the particle. The transition zone of partial cleavage is diffuse, whereas the transition between unexpanded and expanded areas is confined to a narrow band of some 20 nm width.     

In vitro association of empty adenovirus capsids with double-stranded DNA.

Several lines of evidence suggest that empty adenovirus capsids are preassembled intermediates in the pathway of virion assembly. We have observed that purified empty capsids of subgroup B adenoviruses have a remarkable affinity for DNA in vitro. The products of capsid-DNA association are sufficiently stable, once formed in low-salt solution, to permit purification and characterization in CsCl density gradients. Neither virions nor the DNA-containing incomplete particles of subgroup B adenoviruses can give rise to such in vitro reaction products. The average molecular weight of the empty adenovirus capsids is about 123 X 10(6), consistent with the absence of viral core peptides and a small deficiency of exterior shell polypeptides. Electron microscopy of negatively stained capsids and the capsids bound to DNA reveals a typical adenovirus size and architecture. The particles appear with a surface discontinuity that is presumed to expose the DNA binding site(s). The DNA molecules associated with the empty capsids are susceptible to the actions of DNase and restriction endonucleases. The dependence of rate of capsid-DNA association on DNA length suggests randomly distributed binding sites on the DNA molecules. Although the DNA molecules can successively acquire additional empty capsids, the empty particles themselves are restricted to interactionwith only one DNA molecule. Electron microscopy of the capsid-DNA complexes spread in cytochrome c films shows that the particles are bo-nd along the contour of extended duplex DNA. The amount of DNA within each bound particle appears to be less than 300 base pairs, as estimated by the length of the DNA molecules visible outside of the bound particle. The empty capsid-DNA association product described in this report provides an interesting substrate for further investigation of the DNA packaging process in a defined in vitro system, with extracts or purified components from infected cells.     

The transformation of τ-particles into T4 heads. I. Evidence for the conservative mode of this transformation.

Asynchroneous T4 phage head maturation includes the step of P23 cleavage: P23 of head-related τ-particles is cleaved into P23

1 This paper is part of the thesis of R. K. L. Bijlenga. It is number X of the series: “Studies on the morphopoieses of the head of phage T-even.”

of capsids with a conservative mode of transformation as evidenced by “heavy” labeling in temperature shift-down experiments with mutant 24 (tsL90). Assuming a subunit pool, data indicate in situ cleavage on individual precursor particles. The interpretation becomes less interesting when assuming a compartmentation of the membrane surface; this hypothesis is not ruled out.     

T=1 capsid structures of Sesbania mosaic virus coat protein mutants: determinants of T=3 and T=1 capsid assembly

Sesbania mosaic virus particles consist of 180 coat protein subunits of 29kDa organized on a T=3 icosahedral lattice. N-terminal deletion mutants of coat protein that lack 36 (CP-NDelta36) and 65 (CP-NDelta65) residues from the N terminus, when expressed in Escherichia coli, produced similar T=1 capsids of approximate diameter 20nm. In contrast to the wild-type particles, these contain only 60 copies of the truncated protein subunits (T=1). CP-NDelta65 lacks the "beta-annulus" believed to be responsible for the error-free assembly of T=3 particles. Though the CP-NDelta36 mutant has the beta-annulus segment, it does not form a T=3 capsid, presumably because it lacks an arginine-rich motif found close to the amino terminus. Both CP-NDelta36 and CP-NDelta65 T=1 capsids retain many key features of the T=3 quaternary structure. Calcium binding geometries at the coat protein interfaces in these two particles are also nearly identical. When the conserved aspartate residues that coordinate the calcium, D146 and D149 in the CP-NDelta65, were mutated to asparagine (CP-NDelta65-D146N-D149N), the subunits assembled into T=1 particles but failed to bind calcium ions. The structure of this mutant revealed particles that were slightly expanded. The analysis of the structures of these mutant capsids suggests that although calcium binding contributes substantially to the stability of T=1 particles, it is not mandatory for their assembly. In contrast, the presence of a large fraction of the amino-terminal arm including sequences that precede the beta-annulus and the conserved D149 appear to be indispensable for the error-free assembly of T=3 particles.     

Three-dimensional structure of the HSV1 nucleocapsid

The three-dimensional structures of full and empty capsids of HSV1 were determined by computer analysis of low dose cryo-electron images of ice embedded capsids. The full capsid structure is organized into outer, intermediate, and inner structural layers. The empty capsid structure has only one layer which is indistinguishable from the outer layer of the full capsids. This layer is arranged according to T = 16 icosahedral symmetry. The intermediate layer of full capsids appears to lie on a T = 4 icosahedral lattice. The genomic DNA is located inside the T = 4 shell and is the component of the innermost layer of the full capsids. The outer and intermediate layers interact in such a way that the channels along their icosahedral two-fold axis coincide and form a direct pathway between the DNA and the environment outside the capsid.     

Structural studies on the empty capsids of Physalis mottle virus

The three-dimensional crystal structure of the empty capsid of Physalis mottle tymovirus has been determined to 3.2 A resolution. The empty capsids crystallized in the space group P1, leading to 60-fold non-crystallographic redundancy. The known structure of Physalis mottle virus was used as a phasing model to initiate the structure determination by real-space electron-density averaging. The main differences between the structures of the native and the empty capsids were in residues 10 to 28 of the A-subunit, residues 1 to 9 of the B-subunit and residues 1 to 5 of the C-subunit, which are ordered only in the native virus particles. An analysis of the subunit disposition reveals that the virus has expanded radially outward by approximately 1.8 A in the empty particles. The A-subunits move in a direction that makes 10 degrees to the icosahedral 5-fold axes of symmetry. The B and C-subunits move along vectors making 12 degrees and 15 degrees to the quasi 6-fold axes. The quaternary organization of the pentameric and hexameric capsomeres are not altered significantly. However, the pentamer-hexamer contacts are reduced. Therefore, encapsidation of RNA appears to cause a reduction in the particle radius concomittant with the ordering of the N-terminal arm in the three subunits. These structural changes in Physalis mottle virus appear to be larger than the corresponding changes observed in viruses for which both the empty and full particle structures have been determined.     

Detecting small changes and additional peptides in the canine parvovirus capsid structure

Parvovirus capsids are assembled from multiple forms of a single protein and are quite stable structurally. However, in order to infect cells, conformational plasticity of the capsid is required and this likely involves the exposure of structures that are buried within the structural models. The presence of functional asymmetry in the otherwise icosahedral capsid has also been proposed. Here we examined the protein composition of canine parvovirus capsids and evaluated their structural variation and permeability by protease sensitivity, spectrofluorometry, and negative staining electron microscopy. Additional protein forms identified included an apparent smaller variant of the virus protein 1 (VP1) and a small proportion of a cleaved form of VP2. Only a small percentage of the proteins in intact capsids were cleaved by any of the proteases tested. The capsid susceptibility to proteolysis varied with temperature but new cleavages were not revealed. No global change in the capsid structure was observed by analysis of Trp fluorescence when capsids were heated between 40 degrees C and 60 degrees C. However, increased polarity of empty capsids was indicated by bis-ANS binding, something not seen for DNA-containing capsids. Removal of calcium with EGTA or exposure to pHs as low as 5.0 had little effect on the structure, but at pH 4.0 changes were revealed by proteinase K digestion. Exposure of viral DNA to the external environment started above 50 degrees C. Some negative stains showed increased permeability of empty capsids at higher temperatures, but no effects were seen after EGTA treatment.     

Assembly-dependent conformational changes in a viral capsid protein. Calorimetric comparison of successive conformational states of the gp23 surface lattice of bacteriophage T4

Inter- and intra-subunit bonding within the surface lattice of the capsid of bacteriophage T4 has been investigated by differential scanning calorimetry of polyheads, in conjunction with electron microscopy, limited proteolysis and sodium dodecyl sulfate/polyacrylamide gel electrophoresis. The bonding changes corresponding to successive stages of assembly of the major capsid protein gp23, including its maturation cleavage, were similarly characterized. The uncleaved/unexpanded surface lattice exhibits two endothermic transitions. The minor event, at 46 degrees C, does not visibly affect the surface lattice morphology and probably represents denaturation of the N-terminal domain of gp23. The major endotherm, at 65 degrees C, represents denaturation of the gp23 polymers. Soluble gp23 from dissociated polyheads is extremely unstable and exhibits no endotherm. Cleavage of gp23 to gp23* and the ensuing expansion transformation effects a major stabilization of the surface lattice of polyheads, with single endotherms whose melting temperatures (t*m) range from 73 to 81 degrees C, depending upon the mutant used and the fraction of gp23 that is cleaved to gp23* prior to expansion. Binding of the accessory proteins soc and hoc further modulates the thermograms of cleaved/expanded polyheads, and their effects are additive. hoc binding confers a new minor endotherm at 68 degrees C corresponding to at least partial denaturation of hoc. Denatured hoc nevertheless remains associated with the surface lattice, although in an altered, protease-sensitive state which correlates with delocalization of hoc subunits visualized in filtered images. While hoc binding has little effect on the thermal stability of the gp23* matrix, soc binding further stabilizes the surface lattice (delta Hd approximately +50%; delta t*m = +5.5 degrees C). It is remarkable that in all states of the surface lattice, the inter- and intra-subunit bonding configurations of gp23 appear to be co-ordinated to be of similar thermal stability. Thermodynamically, the expansion transformation is characterized by delta H much less than 0; delta Cp approximately 0, suggesting enhancement of van der Waals' and/or H-bonding interactions, together with an increased exposure to solvent of hydrophobic residues of gp23* in the expanded state. These findings illuminate hypotheses of capsid assembly based on conformational properties of gp23: inter alia, they indicate a role for the N-terminal portion of gp23 in regulating polymerization, and force a reappraisal of models of capsid swelling based on the swivelling of conserved domains.