Journal of Structural Biology 154 (2006) 255259

www.elsevier.com/locate/yjsbi

Cryo-EM structure of a bacteriophage T4 gp24 bypass mutant:

The evolution of pentameric vertex proteins in icosahedral viruses
Andrei Fokine a, Anthony J. Battisti a, Victor A. Kostyuchenko a, Lindsay W. Black b,
Michael G. Rossmann a,
a
Department of Biological Sciences, Purdue University, 915 W. State Street, West Lafayette, IN 47907-2054, USA
b
Department of Biochemistry and Molecular Biology, University of Maryland School of Medicine, Baltimore, MD 21201-1503, USA

Received 18 November 2005; received in revised form 13 January 2006; accepted 17 January 2006
Available online 21 February 2006

Abstract

Many large viral capsids require special pentameric proteins at their Wvefold vertices. Nevertheless, deletion of the special vertex pro-
tein gene product 24 (gp24) in bacteriophage T4 can be compensated by mutations in the homologous major capsid protein gp23. The
structure of such a mutant virus, determined by cryo-electron microscopy to 26 , shows that the gp24 pentamers are replaced by mutant
major capsid protein (gp23) pentamers at the vertices, thus re-creating a viral capsid prior to the evolution of specialized major capsid
proteins and vertex proteins. The mutant gp23 pentamer is structurally similar to the wild-type gp24 pentamer but the insertion domain
is slightly more distant from the gp23 pentamer center. There are additional SOC molecules around the gp23 pentamers in the mutant
virus that were not present around the gp24 pentamers in the wild-type virus.
2006 Elsevier Inc. All rights reserved.

Keywords: Bacteriophage T4; Capsid vertex protein; Evolution; Virus

1. Introduction evolved a specialized vertex protein subsequent to gene

Most large icosahedral or prolate-icosahedral viral caps-
ids are assembled from roughly planar hexagonal arrays of
capsomers. For example, the heads of many tailed bacterio-
phages are built of hexagonal capsomers that utilize the
HK97 protein fold (WikoV et al., 2000) in their capsid pro-
tein structure. Similarly, a large group of eukaryotic viruses
utilize the jelly-roll fold to assemble arrays of pseudo-
hexagonal capsomers (Benson et al., 2004). There is, how-
ever, an alignment problem at the Wvefold vertices where
the roughly planar hexagonal arrays of capsomers intersect
with each other. The procapsids of some viruses, such as
HK97, have roughly spherical shells, not planar arrays, for
which the assembly problem is solved by having asymmet-
ric hexamers that subsequently become symmetric in matu-
ration (WikoV et al., 2000). Many large viruses have
*
Corresponding author. Fax: +1 765 496 1189.
E-mail address: mr@purdue.edu (M.G. Rossmann).
duplication of a primordial, less specialized capsid protein.
For instance, the lipid-containing phage PRD1 has a penta-
mer at each Wvefold vertex composed of Wve, single, jelly-
roll P31 monomers, whereas the major capsid protein, P3,
consists of two sequential jelly-roll folds that assemble into
pseudo-hexagonal capsomers (Abrescia et al., 2004). Simi-
larly for bacteriophage T4, the major capsid protein, gene
product 23 (gp23), forms hexamers, and the special vertex
protein, gp24, forms pentamers, with both proteins having
most probably an HK97-like fold (Fokine et al., 2005). In
very large double-stranded DNA (dsDNA) icosahedral
viruses, such as Paramecium bursaris Chlorella virus (Yan
et al., 2000), the Wvefold vertices have even more specialized
pentasymmetrons that are themselves arrays of capsom-
ers. These capsomers diVer slightly from the major capsid
protein in conformation (Simpson et al., 2003) and possibly
in their amino acid sequence.
Bacteriophage T4 is a giant, tailed, and dsDNA virus
that belongs to the Myoviridae family and infects

1047-8477/$ - see front matter 2006 Elsevier Inc. All rights reserved.
doi:10.1016/j.jsb.2006.01.008
256 A. Fokine et al. / Journal of Structural Biology 154 (2006) 255259

Escherichia coli. The mature bacteriophage T4 virion con-
sists of a capsid and a tail with a contractile sheath termi-
nating in a baseplate to which are attached six long tail
Wbers (Eiserling and Black, 1994). The T4 capsid is a pro-
late-icosahedron (Fig. 1A) with a length of 1195 and a
width of 860 (Fokine et al., 2004). The major capsid
protein gp23 (48.7 kDa) (where the represents the
cleaved form of the protein when the prohead matures to
infectious virus) forms a hexagonal lattice characterized
by the triangulation numbers (Caspar and Klug, 1962)
Tend D 13 (laevo) for the icosahedral caps and Tmid D 20
for the cylindrical midsection (Fokine et al., 2004). Eleven
vertices of the capsid are occupied by pentamers of the
special vertex protein gp24 (47.2 kDa; Iwasaki et al.,
2000), whereas the 12th vertex is the portal for DNA
entry, tail attachment, and DNA ejection through the tail
(Driedonks et al., 1981). The crystal structure of gp24
(Fokine et al., 2005) was used to build a homology model
of gp23, given the 21% amino acid sequence identity
between gp24 and gp23. Gp24 and most probably gp23
have an HK97-like fold and an additional insertion
domain that forms characteristic protrusions on the T4
capsid surface. Two accessory proteins, highly antigenic
outer capsid protein (HOC, 39.1 kDa) and small outer
capsid protein (SOC, 9.7 kDa), decorate the outside of the
capsid shell (Ishii and Yanagida, 1977). Whereas the SOC
protein helps to stabilize the capsid against extremes of
pH and temperature, HOC has only a marginal eVect on
capsid stability (Steven et al., 1992).
The assembly of a T4 head is initiated with the forma-
tion of a head precursor, called a prohead, consisting of a
shell of gp23 hexamers surrounding a core composed of
various scaVolding proteins (Black et al., 1994). The pro-
head formation is completed by binding gp24 to the penta-
meric vertices. Mutants lacking gp24 produce only
proheads and protein tubes. The attachment of gp24 initi-
ates the proteolytic cleavage of the prohead proteins. This
cleavage is performed by the prohead proteinase, T4PPase
(Showe et al., 1976). The T4PPase cleaves oV a 65-residue-
long amino-terminal peptide from each gp23 molecule,
yielding gp23, and a 10-residue amino-terminal region of
gp24, giving rise to gp24 (Black et al., 1994). When DNA
packaging has started, the gp23 hexagonal lattice expands,
increasing the distance between the centers of the hexa-
meric capsomers from 117 to 140 , decreasing the thick-
ness of the capsid walls and creating binding sites for the
HOC and SOC proteins (Steven et al., 1991; Steven and
Carrascosa, 1979).
Gp24 bypass mutants (gp24byp) map onto gene 23
and have amino acid substitutions that permit head
assembly to escape the need for gp24 (Johnson et al.,
1992; McNicol et al., 1977). Most of the gp24byp muta-
tions are located in the axial (A) domain of gp23, close to
the capsomer center (Fokine et al., 2005). Presumably, the
mutations in the axial domain modify interactions
between monomers within capsomers and allow gp23 to
form not only hexamers but also pentamers. Other
gp24byp mutations map into the amino-terminal region

Fig. 1. Comparison of mutant and wild-type T4 capsid structure. (A) Cryo-EM reconstruction of the native bacteriophage T4 capsid. (B) Cryo-EM recon-
struction of the T4 gp24byp mutant. The gene products 23, 24, HOC, and SOC are colored blue, magenta, yellow, and pink, respectively. A midsection
facet triangle is outlined in black. Red ellipses surround the additional SOC molecules around the Wvefold vertices that are visible in this projection. The
reconstructions used Wvefold averaging about the long axis of the head. The features of the tail (green) appear blurred because the tail has sixfold symme-
try. The bars correspond to 100 . (A is adapted from Fokine et al. (2004).)
 A. Fokine et al. / Journal of Structural Biology 154 (2006) 255259 257

of gp23, which is cleaved during maturation (Johnson 2. Results

et al., 1992). The mutants that derive their bypass pheno-
type from mutations in the amino-terminal segment of
gp23 are more diYcult to explain as the Wnal structure of
assembled gp23 has the same amino acid sequence as in
the wild-type virus.
Here, we present a cryo-electron microscopy (cryo-
EM) structure of a T4 gp24byp mutant that has one
mutation in the amino-terminal section of gp23 and is
missing gp24. The gp23 assembles as hexamers with an
HOC molecule at their centers and into pentamers
located around the Wvefold vertices that alter the
distribution of associated SOC molecules. Thus, the viral
capsid would approximate a primordial virus in which
the major capsid protein and the vertex protein have not
specialized to better serve their speciWc functions. This
study, therefore, explores, in reverse, the evolution of
capsid assembly in large icosahedral viruses.
Electron micrographs of a gp24byp mutant phage
(Fig. 2) contained only particles with a normal head length
and were used to produce a three-dimensional reconstruc-
tion with 26 resolution (see Section 4). The overall archi-
tecture of the mutant capsid (Fig. 1B) was found to be
closely similar to that of the wild-type virus (Fokine et al.,
2004). However, the pentameric vertices of the T4 mutant
capsid were occupied by pentamers of gp23 (Figs. 3 and 4).
Therefore, this structure represents a primordial virus in
which the vertex protein and the major capsid protein have
not yet had a chance for separate evolution of their special-
ized functions.
The distances between the centers of gp23 pentamers
and the centers of the protrusions corresponding to the
insertion domain in the individual monomers that make up
the pentamer were 5 larger than the corresponding dis-
tances for the gp24 pentamers in wild-type virus. Protru-
sions corresponding to the HOC molecules were at the

Fig. 2. Cryo-EM micrograph of the bacteriophage T4 gp24byp mutant. Fig. 4. Comparison of the wild-type and mutant capsid vertex structures.

Fig. 3. Structure of the pentameric vertex, viewed down the Wvefold axis of the capsid, of (A) wild-type, and (B) T4 gp24byp mutant. The SOC molecules
are indicated by white lines. The additional SOC molecules bound around the gp23 pentamers are marked with red lines.
258 A. Fokine et al. / Journal of Structural Biology 154 (2006) 255259
centers of gp23 hexamers as in wild-type virus, but were
missing from the center of the gp23 pentamers. The densi-
ties of the HOC molecules were about one-third or less of
the density of the gp23 monomers, as compared to being
about the same in wild-type virus when computed at the
same resolution or when calculated with the same number
of particles, suggesting that the occupancy of the HOC
molecules is only partial. Possibly, the HOC molecules dis-
sociated because the mutant phage sample had been kept
for about 10 years before being examined by electron
microscopy.
The SOC molecules bind between two adjacent gp23
molecules. Consistent with this observation, SOC does not
bind between gp23 hexamers and gp24 pentamers in the
wild-type, but does bind between gp23 hexamers and
gp23 pentamers in the mutant (Figs. 3 and 4). In the
mutant virus, there are Wve additional SOC molecules
around each of the Wvefold vertices in the distal icosahedral
cap, but only four, not fully occupied, SOC molecules
around each of the midsection vertices. The absence of SOC
molecules around the midsection pentamers is presumably
the result of the variation of orientation of the gp23 mono-
mers in the pentameric capsomer relative to the gp23
monomers in the surrounding hexamers. Similarly, SOC
molecules do not bind between gp23 hexamers and the
portal protein, conWrming that one SOC molecule interacts
with two suitably oriented gp23 molecules.
The nucleotide sequence of gene 23 of the bypass mutant
phage showed that there are two amino acid diVerences rel-
ative to the wild-type protein sequence (GenBank No.
AAD42428). The Wrst mutation is Gln 57 to His located in
the N-terminal part of gp23, which is cleaved oV during
capsid maturation. The second mutation is Ser 405 to Asn,
which is probably irrelevant as some strains of T4 and some
T4 close relatives, such as Aeh1, KVP40, RB49, and
4RR2.8t, have an Asn residue at this site. Moreover, the
bypass mutation was also genetically mapped close to resi-
due 57 near the location of two other bypass mutations in
the amino-terminal region of gp23 (Johnson et al., 1992;
McNicol et al., 1977).
3. Discussion
As the bypass mutation was found to be in the amino-
terminal segment that is cleaved from gp23 during prohead
maturation, the gp23 monomers have wild-type sequence
in the mutant virus. However, whereas the gp23 monomers
assemble only into hexamers in the wild-type virus, they
assemble into hexamers and pentamers in the mutant virus.
Presumably, the mutated amino acid in the amino-terminal
segment encourages the assembly of pentamers without los-
ing the ability to form hexamers prior to the cleavage that
initiates maturation of the prohead. This is consistent with
the movement of residues 4857 from being internal in the
prohead to being externalized during maturation (Steven
et al., 1991). Thus, a mutation in this region is likely to alter
the conformational changes that occur during this process.
Although the head capsid has walls that are only about
20 thick, it has to withstand the considerable pressure of
the concentrated negative charge of the packaged DNA.
Because all the gp23 hexamers have quasi-similar environ-
ments, the capsid is stabilized by making all the contacts
between hexamers of about the same strength and thereby
eliminating weak contacts where capsid disruption might
start. However, the vertices have Wvefold, rather than six-
fold environments and thus require special reinforcement
to avoid creating weak regions in the capsid. The bypass
mutant studied here is cold sensitive, but has an osmotic
shock-resistant phenotype, suggesting increased capsid sta-
bility compared to wild-type (McNicol et al., 1977). This
could be due to the extra SOC molecules around the penta-
mers or to altered permeability of the slightly altered gp23
pentamers and hexamers. However, increased stability may
not necessarily be an advantage as viruses have carefully
balanced properties between stability for transport between
hosts and labile qualities required for virus disintegration
on infection. Thus, the special vertex protein might gain the
advantage of capsid adaptability to diVerent stages of the
virus life cycle.
The major capsid protein gp23 and the vertex protein
gp24 of phage T4 have barely recognizable sequence simi-
larity, but by inference from the structures of other large ico-
sahedral viruses that have special vertex proteins whose fold
is similar to that of the major capsid protein (Abrescia et al.,
2004), it is highly likely that gp23 and gp24 have similar poly-
peptide folds and that these proteins evolved from a common
primordial HK97-like capsid protein after gene duplication.
Such divergence subsequent to gene duplication or multipli-
cation is a common occurrence in the evolution of viral cap-
sid structures. For example, the viral proteins 1, 2, and 3 of
picornaviruses (Rossmann et al., 1985) have probably
evolved in each case from a common ancestor.
Sequence alignment of gp23 with gp24 indicates that
these proteins have accumulated a lot of diVerences while
evolving their specialized functions. Nonetheless, the exis-
tence of the bypass mutants shows that single amino acid
changes in gp23 allow the mutant gp23 to fulWll the func-
tions of both gp23 and of gp24 as must have been the case
for the primordial T4 capsid protein. These observations
might argue for a relatively recent gene duplication of a pri-
mordial capsid gene. However, as all Myoviridae (tailed
phages with contractile tails) examined to date contain a
protein homologous to the vertex protein gp24, the gene
duplication must have occurred prior to the divergence of
Myoviridae from other tailed phages and, furthermore,
suggests that the vertex protein provides a signiWcant
advantage. It is also apparent that the specialization of ver-
tex proteins is not essential for the assembly of infectious
virions, but presumably is advantageous in providing a
more stable and perhaps more versatile structure. Thus,
large dsDNA viruses in which the capsids are constructed
of hexagonal or pseudo-hexagonal arrays have had greater
evolutionary success once a suYciently specialized vertex
assembly had been developed.
 A. Fokine et al. / Journal of Structural Biology 154 (2006) 255259
4. Experimental procedures
The mutant T4 phage 24amNG433-byp241 was grown in
E. coli BE cells and then puriWed by diVerential centrifuga-
tion followed by step-gradient CsCl equilibrium centrifuga-
tion as described by McNicol et al. (1977).
Images of the frozen-hydrated sample were recorded on
Kodak Wlms using an FEI CM300 FEG microscope at a
magniWcation of 47,000, an irradiation dose of 29 elec-
trons/2, and a defocus distance of 13 m. The micro-
graphs were digitized with a ZI Imaging PhotoScan 2001
scanner and used to select 710 phage particles with the help
of the program ROBEM (http://bilbo.bio.purdue.edu/
~workshop/help_robem).
The three-dimensional image reconstructions were
performed with the program SPIDER (Frank et al., 1996)
using the projection-matching technique to orient the
particles. The reconstruction of the native bacteriophage
T4 capsid (Fokine et al., 2004), represented on a 6 grid,
was used as the initial model. Several cycles of image recon-
struction were performed imposing Wve- and twofold sym-
metry (point group symmetry 52 or D5) averaging. The
iterative reconstruction procedure was completed by using
Wvefold averaging only. The resolution was estimated using
a Fourier shell correlation cutoV of 0.5. The Wnal resolution
of the reconstruction obtained by imposing Wve- and two-
fold symmetry was 26 , whereas the resolution of the map
obtained by imposing only Wvefold symmetry was 28 .
The nucleotide sequencing of gene 23 in the mutant
phage was determined with standard PCR-based tech-
niques. Four primers complementary to negative DNA
strand and one primer complementary to positive DNA
strand were used to sequence gene 23 of the mutant
phage.
Acknowledgments
We thank Petr Leiman, Sergei Boudko, and Vadim Mes-
yanzhinov for helpful discussions and Sheryl Kelly and
Sharon Wilder for checking the Wnal manuscript. The work
was supported by a National Science Foundation Grant to
M.G.R. (MCB-9986266) and a Human Frontiers Science
Program Grant to M.G.R., Vadim Mesyanzhinov, and
Fumio Arisaka (RGP 28/2003).
References
Abrescia, N.G., Cockburn, J.J., Grimes, J.M., Sutton, G.C., Diprose,
J.M., Butcher, S.J., Fuller, S.D., San Martin, C., Burnett, R.M., Stu-
art, D.I., Bamford, D.H., Bamford, J.K., 2004. Insights into assem-
bly from structural analysis of bacteriophage PRD1. Nature
(London) 432, 6874.
Benson, S.D., Bamford, J.K., Bamford, D.H., Burnett, R.M., 2004. Does
common architecture reveal a viral lineage spanning all three domains
of life? Mol. Cell 16, 673685.
259
Black, L.W., Showe, M.K., Steven, A.C., 1994. Morphogenesis of the T4
head. In: Karam, J.D. (Ed.), Molecular Biology of Bacteriophage T4.
American Society for Microbiology, Washington, DC, pp. 218258.
Caspar, D.L.D., Klug, A., 1962. Physical principles in the construction of
regular viruses. Cold Spring Harbor Symp. Quant. Biol. 27, 124.
Driedonks, R.A., Engel, A., ten Heggeler, B., van Driel, R., 1981. Gene 20
product of bacteriophage T4. Its puriWcation and structure. J. Mol.
Biol. 152, 641662.
Eiserling, F.A., Black, L.W., 1994. Pathways in T4 morphogenesis. In:
Karam, J.D. (Ed.), Molecular Biology of Bacteriophage T4. American
Society for Microbiology, Washington, DC, pp. 209212.
Fokine, A., Chipman, P.R., Leiman, P.G., Mesyanzhinov, V.V., Rao, V.B.,
Rossmann, M.G., 2004. Molecular architecture of the prolate head of
bacteriophage T4. Proc. Natl. Acad. Sci. USA 101, 60036008.
Fokine, A., Leiman, P.G., Shneider, M.M., Ahvazi, B., Boeshans, K.M.,
Steven, A.C., Black, L.W., Mesyanzhinov, V.V., Rossmann, M.G., 2005.
Structural and functional similarities between the capsid proteins of
bacteriophages T4 and HK97 point to a common ancestry. Proc. Natl.
Acad. Sci. USA 102, 71637168.
Frank, J., Radermacher, M., Penczek, P., Zhu, J., Li, Y., Ladjadj, M., Leith,
A., 1996. SPIDER and WEB: processing and visualization of images in
3D electron microscopy and related Welds. J. Struct. Biol. 116, 190199.
Ishii, T., Yanagida, M., 1977. The two dispensable structural proteins (soc
and hoc) of the T4 phage capsid: their properties, isolation and charac-
terization of defective mutants, and their binding with the defective
heads in vitro. J. Mol. Biol. 109, 487514.
Iwasaki, K., Trus, B.L., WingWeld, P.T., Cheng, N., Campusano, G., Rao,
V.B., Steven, A.C., 2000. Molecular architecture of bacteriophage T4
capsid: vertex structure and bimodal binding of the stabilizing acces-
sory protein, soc. Virology 271, 321333.
Johnson, K., Condie, B., Mooney, D.T., Doermann, A.H., 1992. Mutations
that eliminate the requirement for the vertex protein in bacteriophage
T4 capsid assembly. J. Mol. Biol. 224, 601611.
McNicol, L.A., Simon, L.E., Black, L.W., 1977. A mutation which bypasses
the requirement for p24 in bacteriophage T4 capsid morphogenesis. J.
Mol. Biol. 116, 261283.
Rossmann, M.G., Arnold, E., Erickson, J.W., Frankenberger, E.A.,
GriYth, J.P., Hecht, H.J., Johnson, J.E., Kamer, G., Luo, M., Mosser,
A.G., Rueckert, R.R., Sherry, B., Vriend, G., 1985. Structure of a
human common cold virus and functional relationship to other picor-
naviruses. Nature (London) 317, 145153.
Showe, M.K., Isobe, E., Onorato, L., 1976. Bacteriophage T4 prehead pro-
teinase. I. PuriWcation and properties of a bacteriophage enzyme which
cleaves the capsid precursor proteins. J. Mol. Biol. 107, 3554.
Simpson, A.A., Nandhagopal, N., van Etten, J.L., Rossmann, M.G., 2003.
Structural analyses of Phycodnaviridae and Iridoviridae. Acta Crystal-
logr. D 59, 20532059.
Steven, A.C., Bauer, A.C., Bisher, M.E., Robey, F.A., Black, L.W., 1991.
The maturation-dependent conformational change of phage T4 capsid
involves the translocation of speciWc epitopes between the inner and
the outer capsid surfaces. J. Struct. Biol. 106, 221236.
Steven, A.C., Carrascosa, J.L., 1979. Proteolytic cleavage and structural
transformation: their relationship in bacteriophage T4 capsid matura-
tion. J. Supramol. Struct. 10, 111.
Steven, A.C., Greenstone, H.L., Booy, F.P., Black, L.W., Ross, P.D., 1992.
Conformational changes of a viral capsid protein. Thermodynamic
rationale for proteolytic regulation of bacteriophage T4 capsid expan-
sion, cooperativity, and super-stabilization by soc binding. J. Mol. Biol.
228, 870884.
WikoV, W.R., Liljas, L., Duda, R.L., Tsuruta, H., Hendrix, R.W., Johnson,
J.E., 2000. Topologically linked protein rings in the bacteriophage
HK97 capsid. Science 289, 21292133.
Yan, X., Olson, N.H., Van Etten, J.L., Bergoin, M., Rossmann, M.G.,
Baker, T.S., 2000. Structure and assembly of large lipid-containing
dsDNA viruses. Nat. Struct. Biol. 7, 101103.