Article

• The EMBO Journal (2005) 24, 1352 - 1363
• doi:10.1038/sj.emboj.7600613

Published online: 17 March 2005

• Subject Categories:

  • Structural Biology | Microbiology and Pathogens

Crosslinking renders bacteriophage HK97 capsid maturation irreversible and effects an essential stabilization

Philip D Ross¹, Naiqian Cheng², James F Conway³, Brian A Firek⁴, Roger W Hendrix⁴, Robert L Duda⁴ and Alasdair C Steven²

1. Laboratory of Molecular Biology, National Institute of Diabetes, Digestive and Kidney Diseases, Bethesda, MD, USA
2. Laboratory of Structural Biology Research, National Institute of Arthritis, Musculoskeletal and Skin Diseases, Bethesda, MD, USA
3. Institut de Biologie Structurale CEA-CNRS-UJF, Grenoble, France
4. Department of Biological Sciences, University of Pittsburgh, Pittsburgh, PA, USA

Correspondence to:

Alasdair C Steven, Building 50, Room 1517, 50 South Drive MSC 8025, NIH, Bethesda, MD 20892, USA. Tel.: +1 301 443 7651; Fax: +1 301 496 0132; E-mail: alasdair_steven@nih.gov

Received 1 December 2004; Accepted 10 February 2005

• Keywords:

  • Brownian ratchet,
  • conformational change,
  • cryo-electron microscopy,
  • differential scanning calorimetry,
  • virus assembly

Introduction

For double-stranded (ds) DNA viruses, capsid assembly and packaging of nucleic acid are initially separate events: first a precursor particle (procapsid) is assembled, which then serves as a receptacle for DNA (see, for reviews, Hendrix and Duda, 1998; Cerritelli et al, 2002; Fane and Prevelige, 2003; Leiman et al, 2003). DNA packaging is accompanied by procapsid maturation, which involves major structural and compositional changes (King and Chiu, 1997; Steven et al, 1997). For dsDNA bacteriophages, these changes remodel the surface shell, resulting in an approximate doubling of its internal volume. Such transitions are a widespread phenomenon, also occurring for herpesviruses (Steven and Spear, 1997), retroviruses (Turner and Summers, 1999), dsRNA phages (Butcher et al, 1997), and an insect virus (Canady et al, 2000). In each case, maturation involves a large conformational change, although not necessarily an increase in capsid size.

Certain features are common to most bacteriophage procapsids despite a dearth of sequence similarity among their proteins: in most cases, the procapsids are round in shape and the hexamers of their icosahedral surface lattices exhibit marked departures from six-fold symmetry; in contrast, mature capsids are flat-faceted icosahedra and their hexamers observe six-fold symmetry closely. In general, maturation involves a structural transformation that is large in scale, cooperative, and irreversible. In some systems, for example T4 (Steven et al, 1976; Iwasaki et al, 2000; Olson et al, 2001) and lambda (Imber et al, 1980; Yang et al, 2000), this transformation creates binding sites on the outer surface for 'accessory' proteins. In other cases, for example HK97, capsid maturation is accompanied by the autocatalytic formation of a network of covalent crosslinks (Duda, 1998).

HK97 affords an attractive system to study capsid maturation (Duda et al, 1995a). Its pathway is outlined in Figure 1. The assembly properties of its 385-residue capsid protein, gp5, are well characterized (Hendrix and Duda, 1998), and a crystal structure has been determined for the mature, crosslinked capsid, Head II (Wikoff et al, 2000; Helgstrand et al, 2003). Maturation begins with proteolytic conversion of gp5 to gp5^* by removal of the 102-residue, N-terminal -domains from the earliest procapsid, Prohead-I, producing the metastable Prohead-II, which subsequently undergoes expansion and crosslinking (Conway et al, 1995; Duda et al, 1995a). Cryo-electron microscopy (cryo-EM) analysis of Prohead II revealed that expansion represents rigid-body movements—rotations and translations—of the conserved core of gp5^*, accompanied by refolding of two extended motifs—the N-arm and the E-loop (Conway et al, 2001).

Figure 1.

Acid-induced maturation of wild-type and mutant (K169Y) HK97 capsids. K169 is one of two residues that participates in covalent intermolecular crosslinks (the other is N356); mutating K to Y at this position abrogates crosslinking. Coexpression of the capsid protein gp5 with the protease gp4 leads to the assembly of Prohead I and its proteolytic conversion to Prohead II. If the portal were also present, a dodecamer of it would replace a pentamer of gp5 at one vertex, and the capsids would have 415 copies of gp5 instead of 420. Prohead II maturation may be induced in vitro by acidification to pH around 4. Wild-type capsids progress through the state EI-I (partially expanded) to EI-II (partially expanded with incipient crosslinking) to EI-III (almost completely expanded, that is, balloon morphology, variably crosslinked). If EI-III remains long enough under these conditions, it becomes EI-IV (balloon morphology, all crosslinks except pentonal ones). Upon neutralization, maturation to Head II (Head morphology, all crosslinks formed) is completed. K169Y may be converted to Head I (Head morphology, no crosslinks) by acidification/neutralization. Its expansion intermediates (in parentheses) represented our a priori expectation but were not previously demonstrated. Although slight structural differences appear to exist between EI-I and EI-II (Lata et al, 2000), they have yet to be explored at higher resolution. For the purposes of this paper, we consider EI-I as thermodynamically indistinguishable from non-crosslinked EI-II.

View full figure (18 KB)

Prohead II may be induced to expand in vitro by any of several chemical stimuli (Duda et al, 1995a). Of these, acidification to pH around 4 has been used to study the dynamics of maturation by cryo-EM and small-angle X-ray scattering (Lata et al, 2000; Lee et al, 2004). This approach led to the characterization of the expansion intermediates (EI states) and the creation of maturation movies (Lata et al, 2000; unpublished results, see http://mmtsb.scripps.edu/viper/MOVI
ES/hk97movies.html). At pH 4.15, Prohead II rapidly switches into a semiexpanded state called EI-I, then to the structurally similar EI-II, and finally to a larger, thin-walled spherical particle EI-III (Figure 1). Upon neutralization, the capsids rapidly complete their maturation to the polyhedral Head II state.

Initially, it was thought that the immediate precursor to Head II was Head I, a particle that is fully expanded but not yet crosslinked (Conway et al, 1995; Duda et al, 1995a). However, it has since been established that, on the acid-induced maturation pathway, crosslinking starts at the EI-II stage and EI-III preparations are mixtures of variably crosslinked capsids of which the extreme case, EI-IV, has all crosslinks in place except those involving the penton subunits (Gan et al, 2004). The temporal ordering of expansion and crosslinking on the in vivo pathway, where they are coupled with DNA packaging, is unknown. Because there is not a unique relationship between capsid morphology and crosslinking status, we use the term 'Head' for any particle with the polyhedral mature structure, including Head II (fully crosslinked) and Head I (no crosslinks). Similarly, we use 'balloon' for all capsids with the preceding thin-walled spherical structure, including EI-III, EI-IV, and the putative, completely non-crosslinked K169Y balloon.

In maturation, bacteriophage capsids may be stabilized either by the expansion transformation or by subsequent modifications. In the T4 system, for example, expansion effects the primary stabilization, which is reinforced by binding of the accessory protein Soc (Ross et al, 1985; Steven et al, 1992). On the other hand, the accessory protein Hoc has no effect on T4 capsid stability, and other phages, for example T7 (Serwer, 1976) and P22 (Prevelige et al, 1990), do not have accessory proteins. Even Soc is dispensable. These considerations raise the question: what purpose is served by crosslinking in the HK97 system? While it is plausible that crosslinks might enhance capsid stability, this proposition has not been addressed directly, nor has been it determined whether an ability to crosslink is essential for HK97 to be viable, that is, replication-competent. The goal of this study was to address these questions—in particular, to compare the stabilities of the successive conformational states and to correlate them with the capsid structure and crosslinking status. To this end, we prepared capsids at various stages of maturation for both wild-type gp5^* and the K169Y mutant, and characterized them by differential scanning calorimetry (DSC) and cryo-EM. We also tested the essentiality of crosslinking by complementing a phage with an amber mutant in gene 5 with coexpressed wild-type gp5 or with the K169Y mutant.

Preparations of wild-type and K169Y mutant capsids at various stages of maturation were subjected to calorimetric analysis under standard conditions, either at pH 7.5 or at pH around 4.0. The thermodynamic quantities evaluated in these experiments are reported in Table I. The thermal stability of a specimen is indicated by T_p, the peak temperature of its denaturation endotherm, which in each case is the largest and highest temperature event recorded. The enthalpy change accompanying denaturation, H_m, is given by the area under the peak and may be regarded as reflecting the disruption of interactions between and within protein subunits. The denaturation entropy change, S_m, is calculated from H_m/T_p, and necessarily compensates H_m. The cooperativity of the transition is expressed as the full-width at half-height of this peak, T_FWHH.

Table 1: Thermodynamic parameters from calorimetric scans of HK97 capsids

View full table

Thermal scanning of Prohead II—wild type and K169Y—at pH 7.5

Both particles produced quite similar thermograms (Figure 2). They show a single, fairly broad, denaturation endotherm whose T_p is slightly lower for K169Y (84.5°C) than for wild type (86.3°C). These events were irreversible, as we ascertained by their failure to reappear in a second scan (data not shown). In fact, all denaturation endotherms observed in this study as well as in previous DSC studies of T4 capsid-related particles (Ross et al, 1985; Steven et al, 1992) and P22 capsids (Galisteo and King, 1993; Galisteo et al, 1995) are irreversible. In addition to its major endotherm, Prohead II (K169Y) exhibited a minor endotherm at 73°C that had no counterpart with wild-type Prohead II. This event also was irreversible.

Figure 2.

Thermograms of Prohead II wild type and K169Y at pH 7.5.

View full figure (12 KB)

Thermal scanning of Head II at pH 7.5

Head II denatures in an asymmetric endotherm (Figure 3, middle) that peaks at 96°C and is preceded by smaller events at 80 and 90°C. The 80°C event has 13% of the enthalpy change of the denaturation endotherm and, notably, it is reversible. The 90°C event is much smaller still and also reversible.

Figure 3.

Thermograms of (top) Head I at pH 7.5; (middle) Head II at pH 7.5; (bottom) Head II at pH 4.1.

View full figure (16 KB)

These experiments indicate that Prohead II, which is already rather stable (T_p>85°C), undergoes substantial additional stabilization—by about 11°C—upon conversion to Head II. Maturation also confers a large increase in H_m and S_m, which are 67% higher for Head II than for Prohead II (Table I). At the same time, the denaturation endotherm becomes narrower, that is, more cooperative (T_FWHH decreases from 6.6 to 4.0°C), and it is preceded by three lesser events that have no counterparts with Prohead II. Since we detected no sign of contaminants in these preparations, we attribute the minor events to disruption of some interaction(s) specific to Head II. We investigated the 80°C event by cryo-EM after incubating Head II at 85°C for 10 min and then rapidly cooling it. No evident changes were seen (cf Figure 4A and B). We conclude that, when cooled after this transition, the capsids switch back to their original state too quickly for structural changes to be detected, or such changes are too subtle to be apparent.

Figure 4.

Cryo-electron micrographs showing fields of (A) Head II at pH 7.5; (B) Head II at pH 7.5 after heating at 85°C; after such heating, Head II had a tendency to form close-packed arrays in vitrified films, as here, but dispersed areas were also observed. However, there is no evident change in size and shape. Such variability in appearance as is observed reflects differences in viewing geometry, for example, the hexagonal particles are viewed along a three-fold axis of symmetry or close to it. (C, D) Comparison of fields of K169Y capsids prepared by acidifying aliquots of a Prohead II isolate to pH 3.9 ((C) essentially all EI-II-like capsids) and pH 4.3 ((D) an approximately equal mixture of EI-II and balloons). In (D), a few examples of EI-II capsids are marked with white arrows, and those of balloons with black arrows. (E, F) Comparison of Head I at pH 7.5 (E) and after switching back to pH 4.1 (F). Bar=1000 Å.

View full figure (510 KB)

Thermal scanning of Head I (K169Y) at pH 7.5

To separate effects of expansion from those of crosslinking, we examined Head I (K169Y), which has the mature polyhedral structure but no crosslinks. Head I denatures at 83°C (Figure 3, top), slightly lower than its precursor state, Prohead II (K169Y), and 13°C lower than Head II. We conclude from this observation that no stabilization is conferred by expansion per se.

Like Head II, Head I exhibits minor thermal events, albeit at different temperatures—a reversible event at 60°C, again with 13% of the enthalpy of the major endotherm—and a very small irreversible event at 73°C.

Thermal scanning of Head II at pH 4.1

Acidification causes Prohead II to embark on maturation. At pH around 4, the end point for wild-type capsids is populations that have the same structure—balloons—but variable degrees of crosslinking. To evaluate their thermal stability, these capsids have to be scanned at acidic pH: if neutralized, they rapidly convert to Head II (wild type) or Head I (K169Y). However, the fact that the melting temperature of any given protein tends to drop with reductions in pH in the acidic range (Privalov and Khechinashvili, 1974) complicates comparison with Prohead II. In this context, Head II provides a point of reference, since it is the most stable capsid under study, with the same structure at both pH values. A preparation of Head II was dialyzed to pH 4.1 and then scanned (Figure 3, bottom). Its denaturation endotherm, with a T_p of 88°C, is sharper and more symmetrical than at pH 7.5, and is preceded by a reversible minor event at 72°C. The relative sizes (H_m) and positions of these two events are essentially the same as at pH 7.5, but they are shifted to temperatures that are lower by 8°C.

Thermal scanning of EI-III at pH around 4

To prepare wild-type EI-III, Prohead II was dialyzed at pH around 4 for at least 24 h (Lata et al, 2000). This material was scanned under two closely related conditions: at pH 4.1 and 0.2 M ionic strength, and at pH 4.2 and 0.4 M ionic strength. The adjustment of ionic strength was intended to suppress aggregation. The pH was altered to explore the sensitivity of the results to small changes of pH in this range, because such changes are known to have large effects on the rate at which Prohead II matures (Lata et al, 2000). These thermograms are quite complex. The former scan (Figure 5, top) exhibits a broad and somewhat asymmetric endotherm extending from 75 to 87°C and peaking at 82°C, preceded by a smaller and sharper event at 73°C and followed by a small event at 89°C. The latter scan (Figure 5, bottom) shows essentially the same minor event, but the main endotherm has two peaks—a sharp one at 76°C and a broad one whose maximum is at 80°C. Both peaks are shifted to lower temperatures compared with the scan at pH 4.1 and 0.2 M ionic strength.

Figure 5.

Thermograms of wild-type HK97 capsids produced by incubating wild-type Prohead II under two closely related acidification conditions: (top) pH 4.1 and 0.2 M ionic strength; (bottom) pH 4.2 and 0.4 M ionic strength. The marked differences between the two thermal profiles primarily reflect sensitivity of crosslinking to slight variations in these conditions.

View full figure (12 KB)

Upon acidification, most K169Y Prohead II capsids mature only to the EI-II stage

Our next step was to compare the corresponding preparations obtained by acidifying Prohead II (K169Y), which we expected to consist of non-crosslinked balloons. To test this expectation, cryo-EM was performed on preparations after dialysis to three pH values, 4.43, 4.20 and 3.86, all at 0.4 M ionic strength. The sample at pH 3.86 was found to consist almost exclusively (>98%) of EI-II, together with a few balloons (Figure 4C). The preparation at pH 4.43 had 60% EI-II and 40% balloons (Figure 4D). We went on to quantitate the percentages of EI-II and balloons in four additional experiments at pH between 4.0 and 4.2 and found the fraction of balloons to vary from 40 to <5%. These data establish that the mutant capsids either do not progress to balloons from the EI-II state as readily as wild type, or they are less efficient at retaining the balloon conformation.

The observations described above raised the question of whether Head I would retain its structure if switched to pH 4.1, or whether it would revert to the balloon or an EI-II-like state. (In practice, we prepare Head I by incubating Prohead II (K169Y) at pH 5 in urea for 1 h and then switching to neutral pH, and such preparations are structurally homogeneous; Conway et al, 1995.) To address this question, we dialyzed a sample of Head I to pH 4.1 and examined it by cryo-EM. We found these particles (Figure 4E) to be slightly rounder than the starting material (Figure 4F) but not as regularly spherical as balloons, whence we conclude that the balloon-to-Head transition is not readily reversible, even in the absence of crosslinks.

Thermal scanning of K169Y expansion intermediates at pH around 4

The three samples—at pH 4.43, 4.20, and 3.86—were subjected to thermal scanning (Figure 6). The T_p's of the major endotherms occurred at 77, 71, and 62°C, respectively. The first two thermograms also had minor events at 67 and 64°C. The scan at pH 3.86 did not have a separate minor event but exhibited a low-temperature shoulder, showing that the minor event does not shift to lower temperature with reductions in pH as rapidly as the denaturation endotherm: by pH 3.86, the two events coalesce. It appears to be ruled out that the major and minor events represent the respective denaturations of EI-II-like particles and balloons, since, in that scenario, the two endotherms should be of similar size in the pH 4.43 scan since this preparation contained approximately equal amounts of both capsids (cf Figure 4D), assuming that the two capsids have similar H_m's. In fact, the major and minor events are of very different sizes (Figure 6, top), whence we conclude that the capsids have similar T_p's and both contribute to the same, rather broad, denaturation endotherm (the major event). We further interpret these thermograms (Figures 5 and 6) as follows.

Figure 6.

Thermograms of K169Y mutant HK97 capsids produced by incubating Prohead II at three closely related values of acidic pH. The marked shifts in the endotherms between these thermal profiles reflect an acute sensitivity to small changes of pH in this range.

View full figure (14 KB)

Their complexity reflects heterogeneity. For wild-type EI-III, the minor event at 89°C represents denaturation of the small amount of Head II that is generally present in Prohead II isolates (Lata et al, 2000). We attribute the breadth of the EI-III endotherm centered on 80–82°C to variability in crosslinking: the more a balloon is crosslinked, the higher its T_p. The sharp event at 76°C seen at pH 4.2 in 0.4 M KCl (Figure 5) is tentatively assigned to non-crosslinked capsids. At pH 4.1 in 0.2 M KCl, this event either does not occur or it is diminished to the point that it does not yield a resolved peak. This distinction appears to originate primarily in the change in ionic strength, since Gan et al (2004) observed less crosslinking of wild-type balloons at 2 M ionic strength compared to 0.2 M, and we have also observed less crosslinking at 0.2 M as compared to 0.4 M by SDS–PAGE (data not shown). The T_p for non-crosslinked balloons is lower for K169Y than for wild-type, and we attribute this shift to the K169Y substitution, as with Prohead II (see above). Finally, the minor event at 70–72°C (Figure 5) correlates with a similar event at about the same temperature for Head II (Figure 3, bottom).

The E-loops of EI-II capsids are in an 'up' position

We calculated a three-dimensional density map from cryo-micrographs of EI-II (K169Y) (Figure 7C and D). The resolution achieved, 15 Å, was somewhat lower than anticipated, given the number of particles and a priori assessment of data quality by diffraction. This outcome may be indicative of conformational microheterogeneity despite their uniformity with respect to crosslinking status (none). Nevertheless, it allowed comparison with wild-type Prohead II at 12 Å (Conway et al, 2001) and EI-IV at 14 Å (Gan et al, 2004). On the outer surface, there are rings of small nubbins around the periphery of each capsomer (arrows in Figure 7C and D)—five around each penton and six around each hexon. Although the nubbins are small—only about 10 Å across—they represent genuine features, not residual noise, because the six symmetrically disposed nubbins around each hexon were independently calculated, not generated by icosahedral symmetry. Moreover, it is known from molecular modeling of Prohead II that these nubbins represent E-loops—to be precise, the 'stump', that is, the two-stranded -sheet, of the E-loop, its outer tip being disordered (Figure 8). On EI-II (K169Y), the E-loops still protrude outwards in an 'up' position, although the nubbins are shifted slightly relative to their positions on Prohead II, that is, their center-to-center spacing in local triplets is 40 Å compared to 30 Å.

Figure 7.

Cryo-EM reconstructions at 15 Å resolution of (A, B) EI-II wild type; (C, D) EI-II (K169Y); and (E, F) the K169Y balloon. (A, C, E) Outside views along a two-fold symmetry axis. (B, D, F) The inner surface (right part); a central section (left part); and a blow-up of the outer surface containing a hexon and a penton (inset). Red arrows point to E-loops of penton subunits and blue arrows to E-loops of hexon subunits. Bar=100 Å.

View full figure (689 KB)

Figure 8.

In Prohead II (left; adapted from Conway et al, 2001), the E-loops are in 'up' positions, that is, outward protruding. In the quasiatomic model (upper right, blown up and rotated through 90°), four E-loops on different quasiequivalent subunits are marked with arrows. In Head II, the E-loops are in 'down' positions, that is, lying in the plane of the capsid shell (Wikoff et al, 2000), as shown at bottom right, where E-loops are marked with arrowheads. The E-loop of the green subunit is eclipsed.

View full figure (190 KB)

It seemed to us unlikely that the 'up' position of the E-loops on EI-II (K169Y) would be caused by substituting Lys with Tyr at their tips, since the wild-type residue—Lys—is charged and therefore should readily accept a solvent-exposed position. Nevertheless, to investigate this possibility, we calculated a density map of wild-type EI-II to the same resolution (Figure 7A and B). It shows E-loops in the same positions as on the mutant EI-II capsid.

The peripentonal E-loops of the K169Y balloon remain in an 'up' position

We also reconstructed the K169Y balloon to the same resolution of 15 Å. This particle (Figure 7E and F) is very similar to the maximally crosslinked balloon EI-IV (Gan et al, 2004). Compared with the preceding EI-II state (Figure 7C and D), its shell is thinner, smoother, rounder, and larger. These changes originate in movements of the hexon subunits, the pentons remaining largely unchanged. In particular, the E-loops around the penton are still 'up', as on EI-IV. However, most of the hexon-associated E-loops are no longer visible, although there is weaker density at the sites of the two hexon-associated E-loops that are closest to the penton (Figure 7F, blow-up, blue arrows). We conclude that, on K169Y balloons, the majority of the perihexonal E-loops have swiveled down into the plane of the capsid shell where they are within reach of adjacent N356 residues but are chemically unqualified to crosslink with them.

Crosslinking is essential to produce viable phage

As shown above, all wild-type HK97 capsids from Prohead II on are physically robust in the sense that they have denaturation temperatures in excess of 80°C at neutral pH. This property raises the question of whether the ability to form crosslinks is an essential function. We used genetic complementation to test for the ability of gp5(K169Y) to produce functional capsids: Escherichia coli cells expressing this protein from a plasmid were infected with bacteriophage carrying an amber mutation in gene 5. Complementation efficiency is commonly measured by determining the phage burst size per cell in liquid culture infections, but a comparably sensitive method is to measure the efficiency of plating (EOP) of mutant phage on bacterial strains expressing a candidate mutant gene product, where plaque formation requires production of at least several phages per cell. In our EOP complementation spot tests, sets of serial dilutions of mutant and control phage were spotted on to a plate seeded with bacteria expressing test or control proteins and spots were scored for complete or partial clearing (plaques) after incubation. The data are shown in Table II. When we measured the EOP of gene 5 amber 1 phage on K169Y plasmid-carrying hosts, it was 4 logs lower than that obtained for a wild-type gene 5 or the mutant C362S, both of which produce fully functional gp5. We conclude, therefore, that the crosslinks that are missing from K169Y capsids are essential for phage viability.

Table 2: Plating of amber phages on bacterial strains producing HK97 capsid proteins shows failure of complementation and dominant-negative phenotype for mutant K169Y

View full table

The preceding experiments were carried out using constitutive levels of plasmid gene expression. If we tested under slightly different conditions where expression was enhanced by adding lactose, the K169Y plasmid exhibited a dominant-negative effect: not only did complementation fail, but even a wild-type phage's ability to grow on the strain was reduced 30-fold. The dominant-negative effect of the K169Y protein on wild-type HK97 growth emphasizes the essentiality of crosslinking since it implies that capsids that can only partly crosslink are not viable.

Bacteriophage capsids must be sufficiently robust to withstand the outwards pressure of densely packed DNA as well as other physical and chemical challenges. Moreover, the interactions between their subunits must be strong enough to resist any tendency to dissociate, even at near-zero concentrations. To achieve the requisite stability, two mechanisms may be employed—the expansion transformation, and related effects such as the binding of accessory proteins or the formation of covalent crosslinks. In some systems, for example T7 and P22, expansion—which appears to be a universal feature of capsid maturation—suffices. In others, both factors contribute: for example, T4 expansion confers a primary stabilization that is reinforced by the binding of Soc subunits that serve as molecular clamps. The present observations point to a different situation with HK97.

Crosslinking is an essential function for HK97

Whereas Soc is dispensable to T4 except in extreme conditions (Black et al, 1994), our complementation data indicate that crosslinking is required to produce viable HK97 phage. Since gp5 (K169Y) is competent to assemble and mature (Figure 4C), the defect presumably lies in its inability to crosslink. We surmise that the mutant capsid cannot sustain the pressure of encapsidated DNA. The dominant-negative property of this mutant suggests that, for viability, the network of crosslinks must be complete or nearly so.

Crosslinks stabilize the mature HK97 capsid

At near-neutral pH, Prohead II denatures with a T_p of 86°C, while the K169Y Prohead II denatures at 84°C. If the same temperature differential between wild type and mutant, 2°C, pertains to the hypothetical wild-type Head I (fully expanded but not crosslinked), it should denature at 85°C. On the other hand, Head II has a T_p of 96°C. It follows that crosslinking augments the thermal stability of the Head by +11°C. A similar inference follows from the elevation in T_p that accompanies increasing levels of crosslinking of the expansion intermediates (Figure 5). Taken together, these observations indicate that crosslinking is solely responsible for stabilizing the HK97 capsid. Essentially no change in thermal stability is conferred by expansion per se. Expansion is only indirectly involved in stabilization in that it facilitates crosslink formation. For comparison, in the T4 system, expansion increases the T_p of its Prohead II counterpart (cleaved but unexpanded) by +23°C (Steven et al, 1992) and Soc binding adds a further +6°C (Ross et al, 1985).

Since EI-II and balloon are not stable at pH 7.5, they cannot be analyzed calorimetrically under these conditions. To assign them a T_p for comparison with Prohead II, we refer to their T_p at pH 4.1 and apply an increment of +7.7°C, assuming this correction to be the same as for Head II. The resulting view of stepwise stabilization along the maturation pathway is summarized in Figure 9. At pH 7.5, the first step, to EI-II, is tantamount to a considerable destabilization. In acid-induced maturation in vitro, this deficit is redressed by crosslinking, and individual capsids may vary in the extent to which their crosslinking takes place at the EI-II, balloon, or Head states (Figure 8). In vivo, we envisage that the innate reluctance of Prohead II to switch to EI-II is overcome by pressures imposed by incoming DNA (and deriving ultimately from the hydrolysis of ATP by the packaging translocase), but the transition states of EI-II and balloon are expected to be short-lived under these circumstances.

Figure 9.

Schematic plot of maturation pathways of wild-type capsids (blue circles) and K169Y capsids (red squares) monitored in terms of thermal stability (T_p) at pH 7.5 and 0.1 M ionic strength. Direct measurements are represented by solid symbols and indirect measurements by open symbols. Since the EI-II and balloon states are kinetically inaccessible at pH 7.5, we estimated their T_p for K169Y from the pH 4.1 value of 68.4°C (interpolated from Figure 6), assuming the same difference between pH 4.1 and 7.5 (7.7°C) as was found for Head II. This yields a T_p of 76.1°C. We estimated the T_p of the hypothetical non-crosslinked wild-type EI-II and balloon by adding the 1.8°C difference between mutant and wild-type Prohead II, giving 77.9°C. In fact, wild-type EI-II and balloon produced in vitro by acidification have variable degrees of crosslinking, resulting in ranges of T_p's that are indicated by the gray columns. The lower limits refer to non-crosslinked capsids, and values are deduced as described above. The upper limits cannot be given precisely but values were assigned to accommodate the observations that wild-type EI-II may have up to 50% crosslinking and wild-type balloons up to 86% (Gan et al, 2004). Three maturation pathways are marked by arrows. K169Y at pH 7.5 is depicted in red and non-crosslinked wild type, blue and blue dashes. The first step, from Prohead II to EI-II, corresponds to a substantial destabilization and indeed Prohead II shows no inclination to convert to EI-II under these conditions. After achieving the EI-II state, wild-type capsids may follow different trajectories, one of which is indicated by solid black arrows. This particular capsid gains 3°C in T_p from incipient crosslinking in the EI-II state; then becomes a balloon (no change in T_p); then undergoes additional crosslinking, gaining 2°C; then switches to Head, which we infer to bring a gain of 7°C by analogy with this step for K169Y; and finally completes crosslinking to Head II for an additional gain of 7°C.

View full figure (62 KB)

In previous work comparing the structures of Prohead II and Head II, we observed that the buried surface area at intermolecular interfaces increases by 50% between the two states (Conway et al, 2001), and tentatively correlated this change with stabilization. In retrospect, we would revise this idea by recognizing that the nature of the interactions at these interfaces must also be important. Changes at inter-subunit interfaces may also account for Head I having a T_p that is 1.5°C lower than its precursor state, Prohead II (K169Y).

Evolutionary implications

Despite the basic differences in the mechanism of capsid stabilization employed by HK97, T4 and T7, there is increasing evidence that the major capsid proteins of many, perhaps all, tailed phages share common ancestry. For example, of phages whose capsid protein sequences are similar to that of HK97, about one-third make covalent crosslinks resembling HK97's (Hendrix, 2005). For P22, which has no detectable capsid protein sequence similarity with HK97, there is evidence that it nevertheless has a similar fold (Jiang et al, 2003). We suggest, therefore, that as the amino-acid sequences of these capsid proteins have diverged from their distant common ancestor, there has also been a diversification of thermodynamic strategies of capsid stabilization, resulting in (at least) three distinct, evolutionarily enduring, mechanisms.

Thermodynamics of stabilization

Another measure of a capsid's stability is G°, the difference in free energy between its folded and unfolded forms. This quantity may be calculated, for any temperature T, from the thermodynamic parameters of denaturation, obtained from DSC, and the relationship (Becktel and Schellman, 1987)

For the overall maturation process, Prohead II Head II, the difference in free energy between the two states is the quantity

At the denaturation temperature of Head II, G^o (H-II)=0, thus G^o=-G^o (P-II). Using the thermodynamic quantities for wild-type Prohead II (Table I) and the observed value of C_p (P-II)=539 cal/mol/K (see Materials and methods), we calculate from equation (1) that G^o=4.6 kcal/mol. The positive value of this quantity confirms that wild-type Prohead II is stabilized in the maturation process at pH 7.5. Since the equilibrium constant K (=[native]/[denatured]) is related to the free energy by K=exp(-G^o/RT), the 4.6 kcal/mol difference in free energy represents a 530-fold increase in K at 96°C and more than four times that at 25°C. Since there are 420 gp5^* subunits per capsid (see Figure 1 legend), the corresponding effect per mole of capsids is enormous, that is, 1930 kcal/mol for Head II.

By similar reasoning, we calculate that Prohead II (K169Y) at pH 7.5 is 0.8 kcal/mol less stable than wild-type Prohead II, and its expanded form, Head I, is less stable by an additional 0.6 kcal/mol. These numbers are small, and illustrate that these three capsids are of comparable stability, as already reflected in the similar values of their T_p's.

The minor endotherms represent disruption of state-specific interactions

In addition to the denaturation endotherms, most HK97 capsids register smaller events at lower temperatures, with 2–10% of the enthalpy exchange seen upon denaturation. For example, Head II exhibits such events at 78 and 89°C. We think it very unlikely that they are contributed by protein contaminants since there was no indication of any such contaminants by SDS–PAGE of heavily loaded gels or by EM. Another possibility is that they may represent denaturation of an accompanying small fraction of dissociated capsomers. However, (i) we have no evidence of gp5^*-containing capsids coexisting with a significantly sized pool of unpolymerized protein, and (ii) most of these events are reversible in contrast to the irreversibility of denaturation of gp5 capsomers (unpublished observations). The most likely explanation is that they represent local transitions of the surface lattice. We are exploring this possibility by cryo-EM analysis of appropriately heated capsids (work in progress).

The position of the E-loops is key to crosslinking

The K169 and N356 residues that engage in a given crosslink reside at the ends of two oppositely oriented -hairpins on subunits in adjacent capsomers (Wikoff et al, 2000). K169 is in the E-loop at the end of its -hairpin. In Prohead II, the hairpin diverges from the plane of the capsid shell so that the terminal loop—which is disordered—protrudes externally (Figure 8). In this 'up' conformation, residue K169 is 30 Å from the N356 with which it is destined to crosslink. In maturation, the subunit rotates so that the hairpin lies in the plane of the capsid shell—the 'down' state—with the E-loop assuming a well-defined conformation in Head II (Wikoff et al, 2000).

Our reconstruction of EI-II (K169Y) (Figures 7C and D) shows its E-loops to be in an 'up' position. Since K169 is the only residue mutated, these capsids should, in principle, be as competent as wild-type EI-II to expand. However, K169Y capsids mostly progress only as far as EI-II, whereas essentially all wild-type capsids become balloons, with partial (EI-III) or almost complete (EI-IV) crosslinking, and the E-loops involved necessarily 'down'. We suggest the following explanation. Upon acidification, the energetic barrier between the Prohead II and EI-II states is lowered and the particles rapidly convert to the latter structure with their E-loops still 'up'. With K169Y, there is little free energy difference between the EI-II and balloon states, and EI-II is, in fact, favored. However, the E-loops are mobile and sample different conformations including crosslink-compatible 'down' positions. With wild-type EI-II, some crosslinks form, and as they accrue, tip the balance in favor of the balloon (EI-III) state. In this way, capsids progress from EI-II to EI-III via a Brownian ratchet mechanism (Peskin et al, 1993; Brokaw, 2001) in which the bias is imposed by the crosslinks.

Is such a ratchet required for maturation in vivo, which takes place concurrently with DNA packaging? It is likely (but not proved) that the capsid progresses through the same intermediates but they may be short-lived, as force generated by electrostatic repulsion between encapsidated DNA and the capsid inner surface (Parker and Prevelige, 1998; Conway et al, 2001) should accelerate expansion. In this context, expansion may promote crosslink formation rather than vice versa, depending on the relative kinetics of the two processes under these conditions.

Preparation of capsids

Prohead II was made by expressing proteins from plasmid pV0 (pT7-5Hd2.9, which expresses HK97 genes 4 and 5 under the control of the T7 10 promoter), and purified by a combination of differential centrifugation, polyethylene glycol precipitation, and velocity sedimentation in glycerol gradients, as described previously (Duda et al, 1995a). A plasmid with gene 5 mutation K169Y was used to make K169Y Prohead II (Duda et al, 1995b). Cells were induced for 6 h at 28°C and used immediately. Proheads were further purified by adsorption to Poros HQ20 quaternary amine ion exchange columns (ABI, Foster City, CA) in 20 mM tris(hydroxymethyl)amino-methane hydrochloride (Tris–HCl)-bis-tris-propane buffer at pH 7.5 with 20 mM NaCl and elution with a linear gradient of NaCl to 0.5 M at 13–33 ml/min using a BioCAD chromatography system (ABI, Foster City, CA). Prohead II was matured to Head II and K169Y Prohead II was converted to Head I by a two-step procedure. Prohead II samples (in 20 mM Tris–HCl pH 7.5, 0.2 M KCl, 1 mM 2-mercaptoethanol) were diluted about five-fold into acidic urea (7 M urea (final) in 50 mM sodium acetate pH 5) to induce expansion. After 60 min, the samples were diluted five-fold in 50 mM Tris–HCl pH 7.5 and dialyzed exhaustively to remove urea. Samples were concentrated by ultracentrifugation, resuspended in a small volume, and dialyzed into calorimetry buffer (17.2 mM K₂HPO₄, 2.8 mM KH₂PO₄, 0.1 M KCl, pH 7.5). Purity and homogeneity were assessed by SDS–PAGE and, on many occasions, by EM.

Protein concentration determination

Protein concentrations were determined by ultraviolet absorption spectroscopy in 6 M guanidinium chloride by the method of Edelhoch (1967), modified by using 90% of his values for the extinction coefficients of tyrosine because this resulted in better agreement between the measured OD₂₈₀/OD₂₈₈ ratio and that predicted from the known amino-acid composition of our samples.

Differential scanning calorimetry

The temperature dependence of the difference in heat capacity between capsid solution and buffer was measured using a Micro-Cal (Springfield, MA) MC-2 calorimeter equipped with 0.6 ml cells and operated at a scan rate of 1.98 K/min. Scans of buffer versus buffer were subtracted to minimize systematic differences between the cells. Protein concentrations were about 1 mg/ml. Experiments were carried out in 0.1 M KCl, 0.02 M KPO₄ buffer at pH 7.5 or in 0.2 M KCl, 0.05 M NaAc buffer solutions in the vicinity of pH 4. The endothermic transition region was defined by the points of departure from the linear pretransition and post-transition baselines. Heats of denaturation, H_m, were calculated by numerical integration of the trace of excess heat capacity in the transition region over a normalized sigmoidal progress baseline constructed on the assumption that the fractional heat absorbed up to any temperature relative to the total heat of the transition was proportional to the extent of reaction. The heat capacity change accompanying denaturation, C_p, was evaluated from the difference between the asymptotic linear baselines extrapolated to the temperature of the transition midpoint.

Cryo-EM and image reconstruction

Samples of K169Y capsids were vitrified and then imaged at a nominal magnification of 45 000 on a CM120 electron microscope (FEI, Mahwah, NJ) equipped with a Gatan 626 cryo-holder, as described previously (Cheng et al, 1999). A total of 11 focal pairs were digitized on an SCAI scanner (Z/I Imaging, Huntsville, AL) using a step size of 7 m, corresponding to 1.6 Å at the sample. Image reconstruction, including contrast transfer function (CTF) correction, was performed as described (Conway and Steven, 1999) with several modifications, including automated particle detection and CFT estimation, that will be described elsewhere. For the EI-II reconstruction, the micrographs yielded 1764 particles, of which 883 (50%) were included in the final map. For the EI-III reconstruction, the corresponding numbers were 1472 and 737. Resolution was assessed in terms of the Fourier shell correlation coefficient (Saxton and Baumeister, 1982) with a threshold of 0.3: for both reconstructions, this gave 15 Å. The wild-type EI-II reconstruction was calculated from cryo-electron micrographs already reported (Lata et al, 2000; the 2.5 h sample). Two defocus pairs were redigitized at 7 m (1.8 Å at the sample) and reanalyzed in parallel with the mutant capsid data: 1066 out of 1590 particles were included in the final map, and the resolution was estimated at 14.5 Å, about double that previously obtained. Analyses were performed on Macintosh G5 computers (Apple Computer, Cupertino, CA), and panels in Figure 7 were prepared on Linux workstations (Dell, Austin, TX) with Amira 3.1 software (Mercury Computer Systems/3D Viz group, San Diego, CA and Merignac, France).

EOP complementation tests

Spot tests compared the EOP of an amber mutant phage grown on a host synthesizing mutant gp5 from a plasmid to the EOP when wild-type gp5 is provided from a control plasmid. Spot tests are variants of phage T4 techniques (Benzer, 1955; Doermann and Boehner, 1970). Phages were HK97 (wild type), HK97 5am1 (amber mutant in gene 5(Li, 2000)), and HK97 amC2 (negative control amber not complemented by plasmids used). Overnight cultures of BL21(DE3) containing positive control plasmid pV0 (which carries wild-type gp5 and gp4) or pV0-K169Y (gene 5 mutant) were grown in LB plus 0.4% maltose and 50 g/ml ampicillin. About 0.15 ml of each culture was mixed with 2.5 ml of top agar, spread on to an LB plate with ampicillin, and allowed to solidify. Phages (in 10 mM Tris–HCl pH 7.5, 10 mM MgSO₄, 0.01% gelatin) at 10 PFU/ml were serially diluted in 10-fold steps and 4 l drops applied to the plates, allowed to dry, and incubated overnight at 37°C. The number of plaques was scored by multiplying (10 number of fully cleared spots) by (3 number of partially cleared spots).

We thank Charles Smith for help with sample handling, Frederic Metoz for computational support, David Belnap and Bernard Heymann for provision of software, and Kelly Lee, Lu Gan, and Jack Johnson for discussions. We acknowledge support from an ATIP grant from the CNRS to JFC and NIH grant GM47795 to RH.

Becktel WJ, Schellman JA (1987) Protein stability curves. Biopolymers 26: 1859–1877 | Article | PubMed | ISI | ChemPort |

Benzer S (1955) Fine structure of a genetic region in bacteriophage. Proc Natl Acad Sci USA 41: 344–354 | Article | PubMed | ChemPort |

Black LW, Showe MK, Steven AC (1994) Morphogenesis of the T4 head. In Molecular Biology of Bacteriophage T4, Karam J (ed) pp 218–258. Washington, DC: American Society of Microbiology

Brokaw CJ (2001) Protein–protein ratchets: stochastic simulation and application to processive enzymes. Biophys J 81: 1333–1344 | PubMed | ChemPort |

Butcher SJ, Dokland T, Ojala PM, Bamford DH, Fuller SD (1997) Intermediates in the assembly pathway of the double-stranded RNA virus phi6. EMBO J 16: 4477–4487 | Article | PubMed | ISI | ChemPort |

Canady MA, Tihova M, Hanzlik TN, Johnson JE, Yeager M (2000) Large conformational changes in the maturation of a simple RNA virus, nudaurelia capensis omega virus (NomegaV). J Mol Biol 299: 573–584 | Article | PubMed | ISI | ChemPort |

Cerritelli ME, Conway JF, Cheng N, Trus BL, Steven AC (2002) Molecular mechanisms in bacteriophage T7 assembly, maturation and DNA containment. In Advances in Protein Chemistry: Virus Structure, Chiu W, Johnson JE (eds) San Diego: Academic Press

Cheng N, Conway JF, Watts NR, Hainfeld JF, Joshi V, Powell RD, Stahl SJ, Wingfield PE, Steven AC (1999) Tetrairidium, a four-atom cluster, is readily visible as a density label in three-dimensional cryo-EM maps of proteins at 10–25 Å resolution. J Struct Biol 127: 169–176 | Article | PubMed | ISI | ChemPort |

Conway JF, Duda RL, Cheng N, Hendrix RW, Steven AC (1995) Proteolytic and conformational control of virus capsid maturation: the bacteriophage HK97 system. J Mol Biol 253: 86–99 | Article | PubMed | ISI | ChemPort |

Conway JF, Steven AC (1999) Methods for reconstructing density maps of 'single particles' from cryoelectron micrographs to subnanometer resolution. J Struct Biol 128: 106–118 | Article | PubMed | ISI | ChemPort |

Conway JF, Wikoff WR, Cheng N, Duda RL, Hendrix RW, Johnson JE, Steven AC (2001) Virus maturation involving large subunit rotations and local refolding. Science 292: 744–748 | Article | PubMed | ISI | ChemPort |

Doermann AH, Boehner L (1970) The identification of complex genotypes in bacteriophage T4.I. Methods. Genetics 66: 417–428 | PubMed | ChemPort |

Duda RL (1998) Protein chainmail: catenated protein in viral capsids. Cell 94: 55–60 | Article | PubMed | ISI | ChemPort |

Duda RL, Hempel J, Michel H, Shabanowitz J, Hunt D, Hendrix RW (1995a) Structural transitions during bacteriophage HK97 head assembly. J Mol Biol 247: 618–635 | Article | PubMed | ISI | ChemPort |

Duda RL, Martincic K, Hendrix RW (1995b) Genetic basis of bacteriophage HK97 prohead assembly. J Mol Biol 247: 636–647 | Article | PubMed | ISI | ChemPort |

Edelhoch H (1967) Spectroscopic determination of tryptophane and tyrosine in proteins. Biochemistry 6: 1948–1954 | Article | PubMed | ISI | ChemPort |

Fane BA, Prevelige Jr PE (2003) Mechanism of scaffolding-assisted viral assembly. Adv Protein Chem 64: 259–299 | PubMed | ChemPort |

Galisteo ML, Gordon CL, King J (1995) Stability of wild-type and temperature-sensitive protein subunits of the phage P22 capsid. J Biol Chem 270: 16595–16601 | PubMed | ChemPort |

Galisteo ML, King J (1993) Conformational transformations in the protein lattice of phage P22 procapsids. Biophys J 65: 227–235 | PubMed | ISI | ChemPort |

Gan L, Conway JF, Firek BA, Cheng N, Hendrix RW, Steven AC, Johnson JE, Duda RL (2004) Control of crosslinking by quaternary structure changes during bacteriophage HK97 maturation. Mol Cell 14: 559–569 | Article | PubMed | ChemPort |

Helgstrand C, Wikoff WR, Duda RL, Hendrix RW, Johnson JE, Liljas L (2003) The refined structure of a protein catenane: the HK97 bacteriophage capsid at 3.44 Å resolution. J Mol Biol 334: 885–899 | Article | PubMed | ChemPort |

Hendrix RW (2005) Bacteriophage HK97: assembly of the capsid and evolutionary connections. In Advances in Virus Research, vol. 64, Roy P (ed) in press. San Diego: Academic Press

Hendrix RW, Duda RL (1998) Bacteriophage HK97 head assembly: a protein ballet. Adv Virus Res 50: 235–288 | PubMed | ChemPort |

Imber R, Tsugita A, Wurtz M, Hohn T (1980) Outer surface protein of bacteriophage lambda. J Mol Biol 139: 277–295 | Article | PubMed | ISI | ChemPort |

Iwasaki K, Trus BL, Wingfield PT, Cheng N, Campusano G, Rao VB, Steven AC (2000) Molecular architecture of bacteriophage T4 capsid: vertex structure and bimodal binding of the stabilizing accessory protein, Soc. Virology 271: 321–333 | Article | PubMed | ChemPort |

Jiang W, Li Z, Zhang Z, Baker ML, Prevelige Jr PE, Chiu W (2003) Coat protein fold and maturation transition of bacteriophage P22 seen at subnanometer resolutions. Nat Struct Biol 10: 131–135 | Article | PubMed | ISI | ChemPort |

King J, Chiu W (1997) The procapsid-to-capsid transition in double-stranded DNA bacteriophages. In Structural Biology of Viruses, Chiu W, Burnett RM, Garcea R (eds) pp 288–311. New York: Oxford University Press

Lata R, Conway JF, Cheng N, Duda RL, Hendrix RW, Wikoff WR, Johnson JE, Tsuruta H, Steven AC (2000) Maturation dynamics of a viral capsid: visualization of transitional intermediate states. Cell 100: 253–263 | Article | PubMed | ISI | ChemPort |

Lee KK, Gan L, Tsuruta H, Hendrix RW, Duda RL, Johnson JE (2004) Evidence that a local refolding event triggers maturation of HK97 bacteriophage capsid. J Mol Biol 340: 419–433 | Article | PubMed | ChemPort |

Leiman PG, Kanamaru S, Mesyanzhinov VV, Arisaka F, Rossmann MG (2003) Structure and morphogenesis of bacteriophage T4. Cell Mol Life Sci 60: 2356–2370 | Article | PubMed | ChemPort |

Li Y (2000) Genetic and Functional Analysis of the Major Capsid Protein of Bacteriophage HK97. PhD Thesis, University of Pittsburgh

Olson NH, Gingery M, Eiserling FA, Baker TS (2001) The structure of isometric capsids of bacteriophage T4. Virology 279: 385–391 | Article | PubMed | ChemPort |

Parker MH, Prevelige Jr PE (1998) Electrostatic interactions drive scaffolding/coat protein binding and procapsid maturation in bacteriophage P22. Virology 250: 337–349 | Article | PubMed | ChemPort |

Peskin CS, Odell GM, Oster GF (1993) Cellular motions and thermal fluctuations: the Brownian ratchet. Biophys J 65: 316–324 | Article | PubMed | ISI | ChemPort |

Prevelige PEj, Thomas D, King J, Towse SA, Thomas GJj (1990) Conformational states of the bacteriophage P22 capsid subunit in relation to self-assembly. Biochemistry 29: 5626–5633 | Article | PubMed | ChemPort |

Privalov PL, Khechinashvili NN (1974) A thermodynamic approach to the problem of stabilization of globular protein structure: a calorimetric study. J Mol Biol 86: 665–684 | Article | PubMed | ISI | ChemPort |

Ross PD, Black LW, Bisher ME, Steven AC (1985) Assembly-dependent conformational changes in a viral capsid protein. Calorimetric comparison of successive conformational states of the gp23 surface lattice of bacteriophage T4. J Mol Biol 183: 353–364 | Article | PubMed | ISI | ChemPort |

Saxton WO, Baumeister W (1982) The correlation averaging of a regularly arranged bacterial cell envelope protein. J Microsc (Oxford) 127: 127–138

Serwer P (1976) Internal proteins of bacteriophage T7. J Mol Biol 10: 271–291

Steven AC, Couture E, Aebi U, Showe MK (1976) Structure of T4 polyheads. II. A pathway of polyhead transformation as a model for T4 capsid maturation. J Mol Biol 106: 187–221 | Article | PubMed | ISI | ChemPort |

Steven AC, Greenstone HL, Booy FP, Black LW, Ross PD (1992) Conformational changes of a viral capsid protein. Thermodynamic rationale for proteolytic regulation of bacteriophage T4 capsid expansion, cooperativity, and super-stabilization by soc binding. J Mol Biol 228: 870–884 | Article | PubMed | ISI | ChemPort |

Steven AC, Spear PG (1997) Herpesvirus capsid assembly and envelopment. In Structural Biology of Viruses, Chiu W, Burnett RM, Garcea RL (eds) pp 312–351. New York: Oxford University Press

Steven AC, Trus BL, Booy FP, Cheng N, Zlotnick A, Caston JR, Conway JF (1997) The making and breaking of symmetry in virus capsid assembly: glimpses of capsid biology from cryoelectron microscopy. FASEB l 11: 733–742 | ChemPort |

Studier FW, Rosenberg AH, Dunn JJ, Dubendorff JW (1990) Use of T7 RNA polymerase to direct expression of cloned genes. Methods Enzymol 185: 60–89 | PubMed | ChemPort |

Turner BG, Summers MF (1999) Structural biology of HIV. J Mol Biol 285: 1–32 | Article | PubMed | ISI | ChemPort |

Wikoff WR, Liljas L, Duda RL, Tsuruta H, Hendrix RW, Johnson JE (2000) Topologically linked rings of covalently joined protein subunits form the dsDNS bacteriophage HK97 capsid. Science 289: 2129–2133 | Article | PubMed | ISI | ChemPort |

Yang F, Forrer P, Dauter Z, Conway JF, Cheng N, Cerritelli ME, Steven AC, Pluckthun A, Wlodawer A (2000) Novel fold and capsid-binding properties of the lambda-phage display platform protein gpD. Nat Struct Biol 7: 230–237 | Article | PubMed | ISI | ChemPort |

Main navigation

• Search EMBO Journal

  Search

• Advanced search

• The EMBO Journal

  • Home
  • Aims and scope
  • Current issue
  • Advance online publication
  • Web focuses
  • Archive:
  • Browse by issue
  • Browse by subject
  • Browse by category
  • Free online sample issue
  • Press releases

• Authors and Referees

  • Editorial process
  • Guide for authors
  • Submit an article
  • Guide for referees
  • Editors and Editorial Board
  • Contact editorial office

• Customer Service

  • Subscribe
  • Order sample copy
  • Purchase articles
  • Reprints and permissions
  • Contact NPG
  • Advertising