Turning a PAGE on Protein Dynamics
By Jason Socrates
Studying protein structure and dynamics is not what it used
In the last several decades, primary, secondary, tertiary,
and quaternary structure determination have all become routine
due to advances in the techniques and technologies of molecular
biology and spectroscopy. Molecular Biology Professor Jack
Johnson at The Scripps Research Institute, for instance, has
been using x-ray crystallography to solve the structures of
intact complex virus particles for over a decade.
So what about observing the motions of intact complex virus
"Usually, you don't think about being able to characterize
the motions [of a molecular machine] biochemically," says
Johnson. "You need spectroscopy or something that is going
to be time-sensitive."
However, in the latest issue of Molecular Cell, Scripps
Research graduate student Lu Gan, Johnson, and their colleagues
report the individual steps of the maturation of a virus called
bacteriophage HK97, which they elucidated with a highly unusual
application of a routine laboratory technique called SDS PAGE
(sodium dodecyl sulfate polyacrylamide gel electrophoresis).
A Virus that Plagues Bacteria
The bacteriophage HK97 is a double stranded DNA virus that
infects Escherichia coli cells. It was first isolated from
pig dung by a rice geneticist in Hong Kong.
The mature virus particles have a hard protein head, the
"capsid," and a long protein tail. In nature, these viruses
infect the bacteria by attaching their tails to bacterial
cell walls and injecting in their DNA. Once inside, the DNA
will circularize, use the bacterial enzymes to make copies
of its DNA, replicate its proteins, assemble new virus particles,
and eventually lyse the cell, spilling out new viral particles.
When the viral subunits assemble, they first come together
to make an immature, round procapsid called "prohead." Then,
when all the proteins are in place and the DNA is ready to
be packaged into the prohead, a viral protein complex sitting
at a vertex (called the portal) of the virus pumps in the
DNA. While this is happening, the prohead matures into its
Maturation is an important part of the viral lifecycle.
When the viral capsid matures, which takes about five minutes
in nature, it goes from being a round particle to an icosahedral
particle with angular vertices and distinct triangular faces.
It also expands from a diameter of about 560 Angstroms to
a diameter of about 660 Angstroms (an Angstrom, one billionth
of a centimeter, is a common yardstick for molecular-scale
objects) While 100 Ang isn't considered a large change at
the macroscopic scale, at the length scale of the entire virus,
the average diameter has increased nearly 20 percent, and
the internal volume has nearly doubled.
At the same time, the capsid forms covalent cross-links
between particular lysine and asparagines amino acids within
420 different protein subunits that make up this complex protein
shell. These cross-links serve to permanently bind the subunits
of the capsid to each other.
The mature and immature forms of the virus are identical
in composition, but completely different in size and stability.
In the final form, almost every subunit has changed its position,
and all the cross-linking serves to tighten the structure
of the viral head.
"The proteins are totally reorganizing," says Johnson. "When
they link together, they form what we call chain mail."
The chain mail is extremely durable and makes the virus
more chemically and mechanically stable. It helps protect
the DNA during transport from one host to another. Once the
DNA is inside and all the linkages are in place, the tail
is attached and the virus is mature and ready to infect another
E. coli cell. As a testament to this stability, the
only known method to proteolyze the mature capsid is to heat
it up to greater than 65 degrees Celsius and then hit it with
a thermophilic protease.
"We think of this [viral capsid] as a molecular machine,"
says Gan. "Its goal is to expand into the mature state."
Using SDS PAGE, Gan observed the individual steps of this
A New Page for PAGE
SDS PAGE is not something that one normally regards as a
powerful technique for analyzing details of protein structure.
It is more a routine assay used to determine crude measurements
of rough molecular weights or purity of a protein solution.
Watching the dynamics of a system of interacting molecules
is normally a job reserved for sensitive techniques like nuclear
magnetic resonance or time-resolved electron microscopy. Divining
fine details about protein quaternary structure with SDS PAGE
should be like smashing a music box with a hammer and spreading
all the pieces out on a table to figure out what songs it
playssomething too crude to work.
The SDS PAGE technique begins with mixing the protein solution
with a strong detergent that completely denatures (unfolds)
the proteins. This solution is then run through a sieve-like
gel, which separates all the proteins in the mixture by virtue
of size or weight. Then the gel is stained and dried, and
finally it shows "bands" of distinct proteins or protein fragments
separated by weight.
But since the subunits of the HK97 virus link covalently
as the viral head expands into its final form, SDS PAGE gave
Gan a perfect way to resolve the individual pieces of the
capsid as the reaction proceeded and to read out how many
of the linkages were in place at various points in the reaction.
Each time a cross link was made within the viral capsid,
a new band would appear on the gel, and this allowed them
to delineate all the individual steps of viral maturation.
Thus, they were able to keep track of the motions of this
molecular machine using simple biochemistry.
"As far as we know," says Johnson, "there are not really
any [other studies] like this."
Part of a Larger Collaboration
The paper is a collaboration between Johnson's group and
several other scientists around the world. Johnson's long-time
collaborator Robert L. Duda, of the University of Pittsburgh
and the Pittsburgh Bacteriophage Institute, is the corresponding
author on the paper and one of the world's leading experts
on these types of viruses. A few years ago, Duda discovered
a way to express the proteins that make the head of the virus
in such a way that they can assemble and mature without the
DNA being present.
Duda and his colleague Brian Firek, who is also an author
on the paper, refined an experimental setup that was used
in the current study and carefully quantified results from
gels with the data from an older paper. Doing so, they were
able to elucidate the "point of no return" for the maturing
virusthe maturation state that can only proceed in a
forward direction and cannot revert to prohead. Usually, it
is impossible to study maturing states of a virus biochemically
because the intermediate states have such a fleeting existence.
But HK97 intermediates biochemically trapped at low pH so
that they have lifetimes on the order of hours to days, which
allows them to be scrutinized with "slow" techniques like
Gan developed the conditions to drive it out of the prohead
state into a nearly mature state, and he characterized the
chemistry that takes place as this maturation occurs. Duda
provided the guidance for the study and is joined on the paper
by his colleague Roger Hendrix, also of the University of
Pittsburgh and the Pittsburgh Bacteriophage Institute.
Johnson and Duda have been collaborating since 1995, and
they originally solved the high-resolution structure of the
mature head state of the HK97 viral capsid. In a later paper
reported a low-resolution cryo-EM structure of the prohead,
into which was fitted the atomic coordinates of the mature
Head. The fit was so good that it became clear the virus matures
by some form of rigid body rotations and local refolding.
The results of these two studies gave the scientists the idea
that parts of the capsid subunits must move a relatively long
distanceup to 35 Angstromsin order to achieve
"Lu came to this laboratory to do a rotation," says Johnson.
This paper is the result of over a year's worth of work that
was a tangent to this original project, and it has grown to
become somewhat larger than the original project from which
At the time, Johnson and his colleagues thought that the
process of cross-linking within the virus capsid was much
simpler and something that happened after the capsid had expanded
into its final form.
"Lu showed that there was this entire complex process [whereby
the virus expanded and cross-linked]," says Johnson. "And
he teased out all the molecular details and has been able
to use chemistry to monitor the transitions."
Gan showed that the amino acids in the subunits of the virus
head cross-linked to each other throughout the transition
from immature to mature virus (instead of at the end, as originally
assumed), and was able to show which subunit linked to which
and the order in which the linking occurred.
Also collaborating on the paper was Alasdair C. Steven of
the Laboratory of Structural Biology National Institute of
Arthitis and Muskuloskeletal and Skin Diseases at the National
Institutes of Health and James F. Conway of the Institut de
Biologie Structurale in Genoble, France.
Conway and Steven applied a powerful structural biology
technique called cryo-electron microscopy to a maturation
intermediate state of the maturing viral capsid and verified
that what Gan had discovered was correct.
In addition to shedding light on basic questions such as
how protein assemblies change their shape, interact with other
proteins, and assemble themselves, the work is important because
the HK97 virus has properties similar to some animal viruses,
particularly herpes viruses. In fact, the vast majority of
complex viruses change their morphology and shape as they
mature. Like HK97, the herpes procapsid morphology is round,
while the capsid is angular. Herpes also packages its DNA
This work may lead to better understanding of the maturation
mechanisms of herpesvirus and other human viruses and may
lead to ways to address the diseases they cause.
Next: High-Resolution Structures
At the moment, Gan and Johnson are working on making crystals
of the nearly mature form of the HK97 virus head so that they
can solve the high-resolution structure, compare it to the
immature prohead and mature head structures, and really investigate
the mechanism of the virus capsid's maturation on the atomic
Johnson quips that, even though Gan is halfway towards his
thesis by now, he would still like to see Gan complete the
original aim of his rotation to crystallize this expansion
intermediate form of the capsida process that involves
subjecting purified virus particles to a myriad of different
salt and buffer concentrations until the exact combination
can be found that allows the virus to array symmetrically
in solution to form a crystal. Only then will a beam of x-rays
diffract off the crystal in a pattern that can be collected,
refined, analyzed, and resolved into a three-dimensional picture
of the viral structure.
"We'd like to study the mechanism by crystallography," says
Johnson, joking that his laboratory focuses on crystallography
and he would like Lu to come out of his lab with some crystallography
All kidding aside, Johnson says, Lu has done a wonderful
job. "This is what quality investigation is all about," he
To read the article, "Control of Crosslinking by Quaternary
Structure Changes during Bacteriophage HK97 Maturation" by
Lu Gan, James F. Conway, Brian A. Firek, Naiqian Cheng, Roger
W. Hendrix, Alasdair C. Steven, John E. Johnson, and Robert
L. Duda, see the June 4, 2004 issue of the journal Molecular
Cell or go to http://www.molecule.org.
Send comments to: email@example.com