Membrane Protein Crystallization From

1
Membrane Protein Crystallization From Cubic
Lipid Matrices
Marcus D. Collins and Sol M. Gruner, Cornell
University LASSP
Membrane proteins at a glance. Protein
crystallography. What makes membrane proteins
hard to crystallize? What has succeeded? The open
questions and our project.
The x-ray determined structure of the light
harvesting protein bacteriorhodopsin,
from Halobacterium halobium, (5)
2
Membrane proteins at a glance
Membrane proteins are simply those that exist in
cell membranes. They can serve as structural
supports, as both passive and active channels for
ions and chemicals, or serve more specialized
functions such as light reception. Fully one
quarter of the human genome encodes membrane
protein sequences.
Proteins in a bilayer membrane patch, (1)
One of the defining features of membrane proteins
is that both hydrophobic and hydrophilic regions
exist on their surfaces. This allows the
proteins to blend in to the hydrophobic region
created by the lipid bilayer which makes up most
of the membrane, and still to have a stable
interface with the aqueous material on either
side of the membrane. Because of these
environmental restrictions placed on membrane
proteins, it seems likely that their structrural
possibilities are more limited than aqueous
proteins, though this is not fully explored.
Common features include transmembrane a-helices
arranged like the staves of a barrel, sheet
structures called b-barrels, and sometimes large
extramembraneous regions of hydrophilic amino
acid residues.
Various membrane proteins shown from views
parallel (above) and perpendicular to the
membrane, (6)
3
Protein crystallography
Just like minerals and salts, proteins can form
crystals. And, just like other crystals,
proteins will diffract X-rays. From the X-rays
diffraction patterns, the protein structures can
be determined.
bR crystals surrounded by lipids, (5)
However, proteins are much harder to grow than
salts, and by 1975 only 37 structures had been
placed in the Protein Data Bank. These crystals
tend to be extremely fragile and sensitive to
conditions such as pH, specific ion
concentrations, and other factors. They are also
easily damaged by the X-rays used to probe their
structure.
X-ray diffraction images of bacteriorhodopsin
crystals, (3)
Now, as techniques of growing crystals and
inverting the X-ray scattering data have
improved, more than 12,000 structures have been
posted to the PDB.
4
What makes a membrane protein hard to crystallize?
Remember that membrane proteins have both
hydrophilic and hydrophobic regions. Until very
recently all protein crystallization techniques
used an aqueous solvent for crystallization.
Membrane proteins easily denature (that is, lose
their structure) in this environment. In 1984
the first membrane protein, a photosynthetic
reaction center, was crystallized and its
structure determined, earning a German trio the
Nobel Prize.
Detergent surrounds membrane proteins in
solution, (1)
These early efforts centered around using a new
class of synthetic, highly contrived detergents
which surrounded the proteins and protected
them-if just barely-from the nearby water.
Despite this advance, only a small number of
membrane proteins have been crystallized to date,
largely because no general procedure has been
found which can crystallize a variety of membrane
proteins. The process of finding the right
detergent and the correct conditions for
crystallization is extremely laborious it often
involves several scientists entire careers.
b Octylglucopyranoside, a detergent used in
membrane protein research (from the 1997 Aldrich
Chemical Catalog)
5
And then... Membrane protein crystallization in
cubo
In late 1996, a Swiss group led by E. M. Landau
dropped a bombshell on the world of membrane
protein crystallography. They had succeeded in
crystallizing a bacterial light harvesting
protein out of a lipid cubic liquid crystalline
matrix.
bR crystals, (3)
The principle is quite simple if you want to
make a membrane protein stable, why not put it in
a membrane? But the task of crystallization is
more difficult than that. Purity of the protein
in the crystal is paramount, and in any case, one
must deliver the protein to the artificial
membrane somehow. Once the protein is in the
membrane, it then must come out and crystallize.
These are not trivial matters. What interests our
group is that, though the Landau group has
succeeded in extending their technique to a
handful of other proteins in a mere few years,
no one yet knows how the technique actually
works. One idea, pictured at left, depends
directly on the chemical precipitants Landaus
group used.
A hypothetical sketch of protein crystals forming
from a lipidic cubic phase, (1)
6
Our project
There are two intriguing questions that are
raised by the Landau groups experiments. The
first is whether the cubic structure they used is
important, or whether simply being in a membrane
allows the membrane protein to crystallize.
Second, and more fundamental, is how the protein
actually forms crystals. There are several clues
to how this might work, but there are no hard
answers.
(1)
It may be that the precipitant is not directly
responsible for the change in solubility that
leads to crystallization. There is indirect
evidence that the crystals form due to structural
changes in the surrounding lipid matrix. We are
exploring whether changes in lipid crystal phase
and lattice parameters lead to protein crystal
formation.
Two important lipid phases, (1)
One of the challenges we face is that the
complicated detergent-salt system which permits
us to solubilize the protein initially can
interfere with the structural behavior of the
lipids. Indeed, the designer detergents used in
membrane protein experiments are designed to have
properties quite similar to those of lipids. It
is now well known that these detergents do alter
the phase behavior of common lipids
significantly. However, it may be possible to
avoid the use of these detergents entirely. This
is a question which remains unanswered.
7
References

• 1. Caffrey, M, Current Opinion in Structural
  Biology, 2000, 10486-497
• 2. Bowie, JU, Current Opinion in Structual
  Biology, 2000, 10435-437
• 3. Landau, EM and JP Rosenbusch, PNAS, 1996,
  9314532-14535
• 4. Pebay-Peyroula, E et al, Biochimica et
  Biophysica Acta, 2000, 119-132
• 5. Rummel, G et al, Journal of Structural
  Biology, 1998, 12182-91
• 6. Chiu, ML et al, Acta Crystallographica, 2000,
  D56781-784