USE OF HYDRATED PHOSPHOLIPID GELS AS MEDIA FOR CRYSTALLIZATION OF MEMBRANE PROTEINS

1
USE OF HYDRATED PHOSPHOLIPID GELS AS MEDIA FOR
CRYSTALLIZATION OF MEMBRANE PROTEINS

• Robert M. Glaeser
• Shahab Rouhani
• Jonathan P. Remis
• Chris Lunde

2
MEMBRANE PROTEINS HAVE BEEN CRYSTALLIZED BY AT
LEAST THREE INDEPENDENT STRATEGIES
3
CRYSTALLIZATION OF DETERGENT-SOLUBILIZED PROTEIN
(MICELLES) IS CURRENTLY THE BEST APPROACH

• Amost 100 unique structures now in the PDB
• All but a few crystallized as detergent-solubilize
  d micelles

4
CRYSTALLIZATION OF PROTEINS IN DETERENT-MICELLES
HAS MAJOR DRAWBACKS, HOWEVER

• gt20 years after the first success (Michel, JMB,
  1982), the failure-rate for getting
  well-diffracting crystals seems much higher for
  membrane proteins than it is for soluble proteins
• In addition, when crystals diffract well enough
  to do a chain-trace, the resolution is often not
  good enough to support high-quality refinement
• Some membrane proteins even appear to crystallize
  in detergent after one or more transmembrane
  helices have been solubilized free of the rest
  of the bilayer-spanning part of the structure
  (vide infra)

5
MEMBRANE-PROTEIN STRUCTURES ARE FAR LESS STABLE
IN DETERGENT THAN WHEN IN LIPID BILAYERS

• Poor protein-stability in micelles of
  short-length detergents is likely to be one of
  the more important factors that make it difficult
  to crystallize membrane proteins
• Detergent solubilization appears to have
  occasionally produced structures that are
  sufficiently homogeneous and stable to be
  crystallized,
• but these structures are sufficiently different
  from the native structure to mis-direct a
  structure-based interpretation of function

6
3-DIMENSIONALLY CONNECTED BILAYER GELS MIGHT BE
AN ALTERNATIVE CRYSTALLIZATION MEDIUM

• Three-dimensional connectivity of the
  lipid-bilayer gel allows proteins to diffuse in
  space until they nucleate and/or join a growing
  crystal
• The lipid bilayer environment is expected to
  stabilize the correct protein conformation

Caffrey (2003) J. Struct. Biol. 142108-132
7
HYDRATED MONOGLYCERIDE GELS HAVE WORKED
SPECTACULARLY WELL FOR SOME HALOBACTERIAL
RHODOPSINS

• Bacteriorhodopsin
• 1.4 A
• Mutants photocycle intermediates 2 2.5 A
• Halorhodopsin
• 1.8 A
• Sensory rhodopsin II
• 2 A
• Sensory rhodopsin II complex with transducer
• lt 2 A

8
SOME MUTANTS OF bR ARE FAR LESS STABLE IN
HYDRATED MONO-OLEIN THAN THEY ARE IN DETERGENT,
HOWEVER
9
EVEN wt bR IS STRUCTURALLY PERTURBED IN
MONO-OLEINAND YET IT CRYSTALLIZES BEAUTIFULLY
10
wt bR ALSO CRYSTALLIZES FROM 3-D CONNECTED
BILAYER GELS FORMED BY DI-ACYL LIPID MOLECULES

• Fusion of proteoliposomes in the presence of
  precipitant
• Takeda, , Kouyama (1998) J. Mol. Biol.
  283463-474
• The resulting gel is presumably a disordered, 3-D
  connected bilayer
• Fusion of DMPC/CHAPSO bicelles
• Faham Bowie (2002) J. Mol. Biol. 335297-305
• The resulting gel appears to form a perforated
  lamellar phase
• Nieh et al. (2001) Langmuir 172629-2638

11
PHOSPHATIDYLETHANOLAMINE (PE) AND SOME
DERIVATIVES OF PE COULD BE ADDITIONAL, PROMISING
CANDIDATES

• Shyamsunder, Gruner et al. (1988) DOPE forms
  cubic phase gels Biochemistry 272332-6
• Keller, Gruner et al. (1996) alamethicin
  catalyzes the formation of cubic phases by PE
  mixtures BBA 1278241-6
• Koynova et al. (1997) low amounts of PEG-lipid
  induce cubic phase formation by hydrated PE BBA
  1326167-70
• Johnsson Edwards (2001) phase behavior of
  mixtures of DOPE and PEG-derivative lipids
  Biophys. J. 80313-23

Hydrated PEG-DMPG gel Rouhani et al. (2002)
Biopolymers 66300-316
12
STABILITY OF A LABILE MUTANT OF bR IS GREATLY
IMPROVED IN HYDRATED PE
The native structure is more fully retained in
PE gels and it is even seen that the protein
is less sensitive to heat-denaturation. Stabili
zation of the native protein conformation is
not perfect, however!
13
wt bR CRYSTALLIZES FROM PE GELS UNDER A WIDE
VARIETY OF BUFFER CONDITIONS
2.9 M Sodium malonate, pH 6
2 M Ammonium sulfate, pH 4.5-8.5
1.7 M Na/K phosphate, pH 5.6 plus 20 PEG4000
1.6 M Lithium sulfate, pH 4.6
14
UNCOLORED MICROCRYSTALS CAN BE IMPOSSIBLE TO
SEE IN SUCH OPTICALLY COMPLEX MEDIA
15
MICROSCOPY WITH 280 nm LIGHT IS AN EFFECTIVE
SOLUTION
16
CAUTION PROTEIN STABILITY MAY STILL BE A TRICKY
THING, EVEN IN 3-D CONNECTED GELS OF DIACYL-LIPIDS
17
FLUORESCENCE SPECTROSCOPY CAN BE USED TO MONITOR
PROTEIN STABILITY IN THE LIPID-GEL MEDIA
But the native protein in PE changes
fluorescence intensity when it is denatured by
heating.
Tryptophan fluorescence shows no further change
when the already denatured membrane protein
in mono-olein is heated
This test can be applied to any membrane protein.
18
A RESEARCH CHALLENGE CAN ONE CONTROL THE
GEL-STRUCTURES NEEDED FOR LIMITED NUCLEATION BUT
ABUNDANT GROWTH?

• CUBIC PHASES HAVE ABUNDANT CONNECTIVITY,
• BUT TRANSITIONAL (LAMELLAR?) STRUCTURES MAY BE
  REQUIRED FOR NUCLEATION AND GROWTH
• EXTENDED LAMELLAR PHASES MAY NOT BE SUITABLE,
  EITHER
• NUCLEATION AND GROWTH IN THE THIRD DIMENSION MAY
  BE PROBLEMATIC

Caffrey (2003) J. Struct. Biol. 142108-132
19
ACKNOWLEDGEMENTS

• M.T. Facciotti
• J. Tittor D. Oesterhelt
• LBNL Laboratory Directors Research and
  Development grant
• UC Discovery Grant/Genentech
• Staff at BL 8.3.1 of the ALS