Lipidic cubic phases as matrices for membrane protein crystallization.
- PubMed: 15325652
Abstract
This review provides detailed procedures for the crystallization of membrane proteins via the lipidic cubic phase method. Bacteriorhodopsin-specific, hands-on protocols are given for (i) the preparation of bacteriohordopsin from purple membrane by monomerization in octylglucoside and gel filtration chromatography or by selective extraction after pre-treatment with dodecyl-trimethylammonium bromide, (ii) the incorporation of bacteriorhodopsin into lipidic cubic phases by mixing in vials or within coupled syringes and, (iii) the crystallization of bacteriorhodopsin in the lipidic matrix by adding a solid salt or an overlaying with a solution. References for further useful procedures and materials are listed in order to provide biochemists and crystallographers with all information that is necessary to grow crystals of the membrane protein bacteriorhodopsin.
Author-supplied keywords
Lipidic cubic phases as matrices for membrane protein crystallization.
r
E
M
ti
th
ph
o
pi
b 2004 Elsevier Inc. All rights reserved.
1. Introduction
It is the purpose of this review to familiarize the
experimental scientist with the practicalities involved in
the use of lipidic cubic phases for membrane protein
crystallization purposes. The detailed protocols given
here enable the reader to readily generate crystals of the
membrane protein bacteriorhodopsin. After mastering
the crystallization of bacteriorhodopsin, further
resources may be consulted, e.g., those listed in Table 1.
It lists references that describe useful procedures and
lipidic cubic phase related crystallization reports which
may be instrumental in designing crystallization screens,
handling of non-colored proteins, or the retrieval of
crystals for X-ray diVraction purposes. It is hoped that
the simple protocols in this review serve as a basis for
benchmark crystallization experiments that embolden
the crystallographer to experiment with more diYcult
transmembrane -helices (bacteriorhodopsin, halorho-
dopsin, sensory rhodopsin II, and sensory rhodopsin II
with its transducer domain). It is therefore believed that
heptahelical membrane proteins that are of non-bacte-
rial origin, namely G-protein coupled receptors (GPCR),
may be appropriate targets for this crystallization
method. After its inception and description in 1996 [9],
the crystallization of bacteriorhodopsin has been
repeated in many laboratories (Table 2). The reproduc-
tion of well-diVracting bacteriorhodopsin crystals, the
application of the method to other membrane proteins,
and the continuing development of lipidic cubic phase-
based crystallization technology demonstrate the value
and future promise of this approach. Numerous original
research reports and reviews have been published on the
topic of lipidic cubic phase-based protein crystallization.
For an introduction to this subject, consult Table 1.Methods 34 (2004
Lipidic cubic phases as matrices fo
Peter N
deCODE genetics, BioStructures Group, 7869 N
Accepted 24
Abstract
This review provides detailed procedures for the crystalliza
Bacteriorhodopsin-speciWc, hands-on protocols are given for (i)
monomerization in octylglucoside and gel Wltration chromatogra
methylammonium bromide, (ii) the incorporation of bacteriorhod
syringes and, (iii) the crystallization of bacteriorhodopsin in the li
References for further useful procedures and materials are listed in
mation that is necessary to grow crystals of the membrane protein1046-2023/$ - see front matter 2004 Elsevier Inc. All rights reserved.
doi:10.1016/j.ymeth.2004.03.030
proteins, in particular those that have not yielded crys-
tals employing conventional crystallization methods.
The track record of the lipidic cubic phase crystalliza-
tion method is summarized in Table 2. It has been partic-
ularly successful for membrane proteins with seven
E-mail address: pnollert@decode.com348–353
www.elsevier.com/locate/ymeth
membrane protein crystallization
ollert
Day RdW, Bainbridge Island, WA 98110, USA
arch 2004
on of membrane proteins via the lipidic cubic phase method.
e preparation of bacteriorhodopsin from purple membrane by
y or by selective extraction after pre-treatment with dodecyl-tri-
psin into lipidic cubic phases by mixing in vials or within coupled
dic matrix by adding a solid salt or an overlaying with a solution.
order to provide biochemists and crystallographers with all infor-
acteriorhodopsin.2. Protocols: crystallization of bacteriorhodopsin in
practice
In general, lipidic cubic phase-based crystallizations
are often carried out as a two-step process. At Wrst the
protein solution is mixed with the lipid, and a lipidic
cubic phase forms spontaneously. This Wrst step is
ec
a
ta
n
isp
li
d
r
e
m
f s
la
t
an
nd
s
d
on
e
oP. Nollert / Method
considered complete when the material is transparent,
Table 1
Useful methods and procedures have been developed to aid in various asp
Topic Description
Lipidic cubic phase based crystallization Pioneering bacteriorhodopsin
Monoolein as matrix for crys
Review of the vial-based tech
Micro crystallization Syringe and manual rachet d
instead of solid salt as crystal
Reproduction of above proce
ReWnement of above procedu
Observation of crystallization setups
by microscopy
Observation options and guid
Demonstration that HoVman
are useful for the detection o
Crystal harvesting Detergent mediated release
Enzymatic release
Overview with protocols simi
Cubic phase forming lipids Lists of lipids and lipid mixes
in Appendix A of the review
Crystallization in absence of
Screening conditions A list with compatible salts a
The destabilization eVect of a
Boundaries on protein size an
estimated from a crystallizati
Temperature-dependent phas
Lyo- and thermotropic monobenchmark experiment.34 (2004) 348–353 349
2.1.2. Monomerization in -octylglucoside and gel
ts of the lipidic cubic phase crystallization methodology
Reference
nd lysozyme crystallization report, glass vials are employed [9,10]
llization of Wve diVerent membrane proteins [3]
ique [21]
enser are employed, 200 nl volume; use of liquid
zation inducing agent
[15]
ure [22]
e and further reduction of volume [5]
to interpretation of crystallization results [16]
odulation contrast and polarization microscopy
mall, colorless protein crystals
[17]
[12]
[13]
r to the ones given here [18]
hat support bacteriorhodopsin crystallization are given [11]
y detergent [14]
solutions is given [18]
et of crystallization solutions is investigated [4]
crystallization conditions are
model
[7]
behaviour of monoolein-detergent mixes [24]
lein phase diagram [1]non-birefringent, and very viscous. Then the second step,
crystallization, is started by the addition of a crystallant.
The latter may be a solid salt mix or a solution. Depend-
ing on the precise conditions, crystals typically appear
within days or weeks 1.
Bacteriorhodopsin is the major protein component in
‘purple membrane,' a membrane fraction obtained from
Halobacterium salinarum. Purple membrane prepara-
tions may be obtained from cultivated H. salinarum
strain ET1001 or strain S9 [20]. Alternatively, they are
available from several commercial sources, i.e., as lyoph-
ilized powders from Sigma (Product No. B0184) Table 3.
A total of 10 mg of purple membrane should be suYcient
for initial crystallization experiments. Purple membrane
is fairly stable when exposed to daylight and room tem-
perature. In contrast, detergent solubilized bacteriorho-
dopsin should be kept cold and in the dark whenever
possible.
2.1. Bacteriorhodopsin preparation from purple membrane
2.1.1. Avoiding the detergent solubilization step
Bacteriorhodopsin crystals may be grown directly
from purple membrane which is mixed with monoolein,
without any pretreatment [14]. Though conceptually
simple, the timeframe for this crystallization strategy,
many months, is prohibitively long to be considered as a
Wltration chromatography
Purple membrane may be solubilized directly with -
octylglucoside. To obtain monomeric bacteriorhodop-
sin, a gel Wltration chromatography step is required [6].
1. Resuspend 10 mg lyophilized purple membrane
(Sigma–Aldrich, B0184) in 10 ml of 150 mM KCl and
collect the purple membrane by centrifugation
(30 min, 18,000 rpm JA-20, 4 °C).
2. Resuspend pellet in 6 ml of 25 mM NaH2PO4, pH 6.9,
and add 2 ml of 10% (w/v) -octylglucoside.
3. Sonicate in bath sonicator for 1 min in the dark at
room temperature and incubate over night without
stirring.
4. Adjust pH to 5.5 with 0.1 N HCl.
5. Centrifuge for 45 min at 15 °C at 55,000 rpm (Ti 70
rotor) and discard the pellet.
6. Concentrate with Centriprep YM-50 Wltration device
to less than 2 ml.
7. Apply concentrate on a BioGel A 0.5 m column or a
comparable preparative gel Wltration column (e.g.,
TSK G3000SW, Tosoh Bioscience, Montgomeryville,
PA, USA,) with 1.2% -octylglucoside in 25 mM
NaH2PO4, pH 5.5, as the stationary phase. Discard
the Wrst and collect the second peak.
8. Concentrate the pooled bacteriorhodopsin fraction
with a Centriprep YM-50 Wltration device. Adjust to a
Wnal concentration of 9 mg/ml in 25 mM NaH2PO4,
pH 5.5.
2.1.3. Selective extraction after DTAB pretreatment
Alternatively, bacteriorhodpsin may be extracted
from purple membranes which have been pretreated
with dodecyl-trimethylammonium bromide (DTAB), a
cationic detergent that is frequently used in detergent
solubilization experiments [8].
1. Resuspend 10 mg lyophilized purple membrane
(Sigma–Aldrich, B0184) in 10 ml of 150 mM KCl and
collect the purple membrane by centrifugation
(30 min, 18,000 rpm JA-20, 4 °C).
2. Resuspend pellet in 10 ml of 10 mM DTAB, 150 mM
Na/K acetate. Tip-sonicate for a total of 1 min, chill
on ice between 20 s pulses.
3. Discard supernatant after centrifugation (1 h,
Table 2
Track record of the lipidic cubic phase-assisted protein crystallization
method
Protein Resolution
(Å)
PDB accession
code or
reference
Bacteriorhodopsin 3.7 [9]
Bacteriorhodopsin 2.35 1AP9
Bacteriorhodopsin 2.3 1BRX
Bacteriorhodopsin 1.55 1C3W
Bacteriorhodopsin D96M 2.0 1C8S
Bacteriorhodopsin, K-state 2.5 1DZE
Bacteriorhodopsin, K-state 2.6 1IXF
Bacteriorhodopsin, L-state 2.1 1EOP
Bacteriorhodopsin, L-state 2.3 1R3P
Bacteriorhodopsin E204Q 1.8 1F4Z
Bacteriorhodopsin E204Q 1.7 1F50
Bacteriorhodopsin 2.3 1IW6
Bacteriorhodopsin D85S/F219L 2.0 1JV6
Bacteriorhodopsin early M state 2.0 1KG8
Bacteriorhodopsin D85S O state 2.25 1JV7
Bacteriorhodopsin mock early M 1.81 1KG9
Bacteriorhodopsin 1.65 1KGB
Bacteriorhodopsin, K-state 1.43 1MOK
Bacteriorhodopsin 1.47 1MOL
Bacteriorhodopsin M-1 1.43 1MOM
Bacteriorhodopsin 1.9 1QHJ
Bacteriorhodopsin early state 2.1 1QKO
Bacteriorhodopsin early state 2.1 1QKP
Sensory rhodopsin II 2.4 1JGJ
Sensory rhodopsin II K state 2.27 1GU8
Sensory rhodopsin II K state 2.27 1GUE
Sensory rhodopsin II 2.1 1H68
Sensory rhodopsin II 2.4 1JGJ
SRII with transducer complex 1.93 1H2S
Halorhodopsin 1.8 1E12
Halorhodopsin 3.2 [3]
Photosynthetic reaction centre
Rhodopseudomonas viridis
3.7 [3]
Photosynthetic reaction centre
Rhodobacter Sphaeroides
2.35 1OGV
Photosynthetic reaction centre
Rhodobacter Sphaeroides
ca. 6 [3]
Light harvesting complex 2, Rp.
acidophila
ca. 25 [3]
Nicotinic acetylcholine receptor [19]18,000 rpm JA-20, 4 °C) and resuspend purple pellet34 (2004) 348–353
in 12 ml of 1.2% (w/v) -octylglucoside, 25 mM
NaH2PO4, pH 6.9.
4. Sonicate again as above and adjust pH to 5.5 with
0.1 N HCl.
5. Clear by centrifugation (1 h, 18,000 rpm JA-20, 4 °C)
and discard pellet. The supernatant contains
monomerized, -octylglucoside solubilized bacterio-
rhodopsin.
6. Concentrate bacteriorhodopsin solution with a Cen-
triprep YM-50 Wltration device (Millipore, Billerica,
MA, USA). Adjust the Wnal concentration to 9 mg/ml
in 25 mM NaH2PO4, pH 5.6.
The concentration of bacteriorhodopsin is usually
determined spectrophotometrically. The extinction coeY-
cient of -octylglucoside solubilized bacteriorhodopsin
at 560 nm is ca. 63,000 cm2/mmol (MWbacteriorhodopsin
26.8 kDa). An absorbance ratio A280/A560 of 1.5–2.0 is
acceptable. The concentrations of -octylglucoside and
those of residual purple membrane lipids in bacteriorho-
dopsin preparations are substantial. They depend on the
type of pretreatment and they typically vary from batch
to batch. This variation may be the source of slightly
diVerent outcomes of crystallization experiments with
respect to crystal growth kinetics, crystal size, and X-ray
diVraction quality.
2.2. Crystallization setup: mix protein and lipid, add
crystallant
In the Wrst step of the crystallization experiment, the
bacteriorhodopsin solution is mixed with the lipid to
form a lipidic cubic phase (ca. 3.5 mg/ml bacteriorhodop-
sin, ca. 60% monoolein). This procedure may be carried
out in a glass vial (Fig. 1) or in two coupled syringes [15]
as shown in Fig. 2. Protocols are given below for both
options. For best results the crystallization experiments
should be carried out under dim light at room tempera-
ture. The crystallization materials should be kept in total
darkness after preparation.
2.2.1. Step 1: Incorporation of bacteriorhodopsin into a
lipidic cubic phase
2.2.1.1. Vials as mixing and crystallization containers.
1. Transfer 6 mg of dry monoolein (monooleoyl-rac-
glycerol [C18:1c9]) into a vial (e.g., a PCR tube or a
glass tube with an inner diameter of 3 mm).
2. Add 4l of bacteriorhodopsin solution (9 mg/ml)
from 2.1.2 or 2.1.3 to the lipid and seal the vial.
3. Mix monoolein and bacteriorhodopsin solution by
repeated centrifugation in a benchtop centrifuge.
Rotate by 180° about the long axis between short
runs to provide mixing. Continue until a homoge-
neous purple, transparent, non-birefringent, and gel-
like material has formed.
od
is
l-
r
-
laP. Nollert / Method
Table 3
Materials that are used in the protocols for the preparation of bacteriorh
dopsin in a monoolein lipidic cubic phase
Purpose Item
Microcrystallization 250 l gas-tight syringe
10 l microsyringe
Semi-automatic ratchet d
Terasaki plate
Clear transparent tape
Cubic phase forming lipid Monoolein, 1-monooleoy
Source of bacteriorhodopsin Lyophilized purple memb
Halobacterium salinarum
Detergent to solubilize bacteriorhodopsin -Octylglucoside, n-octyl
Cationic detergent for purple membrane
pretreatment
DTAB, dodecyl-trimethy
Filtration device Centriprep YM-50below) and cover with transparent tape.
Fig. 1. Growth of bacteriorhodopsin crystals within a monoolein-based lipid
ment was set up in a glass vial as lined out in 2.2.1.1. At Wrst the purple ba
The bottom of the 3 mm diameter tube is Wlled with solid Sørensen salt. (B)
close-up of purple bacteriorhodopsin crystals with sizes up to ca. 100 m alo34 (2004) 348–353 351
opsin from purple membrane and for the crystallization of bacteriorho-
Order information
Model 1725, Hamilton, Reno, NV, USA
Model 701-26s, Hamilton, Reno, NV, USA
penser Model PB600, Hamilton, Reno, NV, USA
Nunc, Rochester, NY, USA
Crystal Clear, Manco, USA
rac-glycerol NuCheck, MN, USA or Sigma–Aldrich, USA
ane from S9 Product No. B0184, Sigma–Aldrich, USA
-D-glucopyranoside No. O311 Anatrace, OH, USA
mmonium bromide Aldrich Chem. Metuchen, NJ, USA
Millipore, Billerica, MA, USAic cubic phase matrix in a glass vial. (A) The 10 l crystallization experi-
cteriorhodopsin is distributed homogeneously in the lipidic cubic phase.
Bacteriorhodopsin microcrystals formed within 1 month. (C) Schematic
ng their longest dimension.2.2.1.2. Coupled syringes as mixing devices
1. Transfer 60 mg of dry monoolein (monooleoyl-rac-
glycerol [C18:1c9]) into a 250l Hamilton syringe
(gastight, RN).
2. Add 40l of bacteriorhodopsin solution (9 mg/ml)
from 1.2 or 1.3 into a second 250l Hamilton syringe
(gastight, RN).
3. Join both syringes via a coupler (Fig. 2).
4. Mix monoolein and bacteriorhodopsin solution by
repeated transfer of the material from one syringe
into the other. Continue until a homogeneous purple,
transparent, non-birefringent, and gel-like material
has formed.
5. Transfer all of the material into one syringe and
attach a suitable dispenser, e.g., a manual rachet dis-
penser with a needle (Fig. 2).
6. Dispense portions of 0.2l into a pre-Wlled well (see
2.2.2. Step 2: Crystal nucleation and growth
Bacteriorhodopsin nucleation and crystallization can
be initiated by the addition of a mix of solid salt to the
preformed lipidic cubic phase. Since weighing of very
small amounts of salt is cumbersome, solutions have
been formulated that can likewise trigger the crystalliza-
tion process [18]. Typically such solutions are used for
multi-well experiments, whereas salts are used primarily
for ‘large-scale' crystallizations in vials.
2.2.2.1. Addition of a solid salt.
1. Prepare Sørensen salt mix by grinding 94.8 g KH2PO4
and 5.2 g Na2HPO4 · H2O.
2. Remove seal from vial (containing 10 mg of lipidic
cubic phase, from 2.2.1.1) and place on balance.
3. Add 2–3 mg of Sørensen phosphate. Seal vial.
4. Centrifuge once in benchtop centrifuge until salt crys-tals combine in bottom of vial (Fig. 1A).
1. Prepare the crystallization solution, 27% (w/v)
PEG 2000 in 100 mM Sørensen phosphate buVer,
pH 5.6.
2. Fill 1l of the crystallization solution into a single
well and cover it with transparent tape. To this well-
based setup, a portion of lipidic cubic phase is added
according to Section 2.2.1.2. See Figs. 2 and 3.34 (2004) 348–3532.3. Observation of bacteriorhodopsin crystallization
setups
Bacteriorhodopsin crystallization setups should be
stored in darkness at constant temperature (e.g., room
temperature). Often crystals can be detected within a few
days after setup. It is simple to spot bacteriorhodopsin
crystals in setups because the crystals are colored purple
e
e
s
c
s
o
Fig. 2. Schematic procedure for the rapid preparation of micro crystallization experiments using the cubic phase method. Crystallization setups are pre-
pared using syringes in combination with a semi-automatic dispenser-driven micro syringe and Terasaki plates. The procedure involves four basic
steps: [I] Preparation of the proper lipidic cubic phase using two 250 l gas-tight syringes coupled together as described [2] (1). Typically, the cubic
phase consists of ca. 60% monoolein, ca. 40% water and protein—i.e., bacteriorhodopsin—at a concentration of ca. 3.5 mg/ml in 20 mM Sørensen phos-
phate buVer, pH 5.6. For dispensing into microwells, the material is transferred into a microsyringe with 10 l total volume either directly into the
syringe barrel or via an appropriate adapter (2). This microsyringe is assembled into a semi-automatic dispenser (3) which allows dispensation in 50
steps of ca. 0.2 l each. [II] Prior to dispensation of the lipid material, six microwells in one lane of a 72-well Terasaki plate are Wlled with 1 l of the pre-
cipitant solutions (4) and dried optionally at 60 °C for 30 min. The latter resembles crystallization of bacteriorhodopsin using dry salt as the precipita-
ri
e
g
i
u(Figs. 1 and 3). Although large crystals can be seen by
the naked eye, it is advisable to inspect crystallization
setups with a dissecting stereo microscope. When investi-
gating crystallization setups, care must be taken to keep
the temperature constant, i.e., prevent warming the set-
ups during manipulation and by microscope light. Oth-
erwise liquid bubbles form inside the lipidic cubic phase
[16,17]. These are detrimental to observation and disrupt
the crystallization process.
References
[1] J. Briggs, H. Chung, M. CaVrey, J. Phys. II France 6 (1997) 723–
751.
[2] A.H. Chen, B. Hummel, H. Qiu, M. CaVrey, Chem. Phys. Lip. 95
Fig. 3. Growth of bacteriorhodopsin crystals using the syringe and microw
crystallization setups in a microwell of a Terasaki plate. (A) Overview of w
Well bottom diameter is ca. 1.2 mm. (B) Close-up view of crystallization
neously distributed in the lipidic cubic phase. (C) Bacteriorhodopsin micro
tallization-inducing agent. (D) Purple bacteriorhodopsin crystals with size
of initial monoolein cubic phase containing 3.5 mg/ml bacteriorhodopsin t
tion agent as described previously [9,11,23]. A single well is shown on the
microwell by positioning the shortened syringe needle at an angle over the w
weight of a single shot of 60% (w/v) monoolein containing cubic phase usin
ble in an excess of an overlaying liquid (7). [IV] The microwells are sealed w
temperatures. Inspection of crystals (10) with a microscope can be carried o[10] E.M. Landau, G. Rummel, S.W. Cowan-Jacob, J.P. Rosenbusch, J.
Phys. Chem. B 101 (1997) 1935–1937.
[11] M. Loewen, M.L. Chiu, C. Widmer, E.M. Landau, J.P. Rosen-
busch, P. Nollert, G-Protein coupled receptors, in: T. Haga, G.
Berstein (Eds.), Methods in Signal Transduction, CRC-Press,
London, 2000, pp. 365–388.
[12] H. Luecke, B. Schobert, H.T. Richter, J.P. Cartailler, J.K. Lanyi, J.
Mol. Biol. 291 (1999) 899.
[13] P. Nollert, E.M. Landau, Biochem. Soc. Trans. 26 (4) (1998) 708–
713.
[14] P. Nollert, A. Royant, E. Pebay-Peyroula, E.M. Landau, FEBS
Lett. 457 (1999) 205.
[15] P. Nollert, J. Appl. Cryst. 35 (2002) 637–640.
[16] P. Nollert, in: S. Iwata (Ed.), Methods and Results in Crystalliza-
tion of Membrane Proteins, 2003, pp. 59–73.
[17] P. Nollert, J. Appl. Cryst. 36 (2003) 1295–1296.
[18] P. Nollert, J. Navarro, E.M. Landau, Methods Enzymol. 343
(2002) 183–199.
[19] Y. Paas, J. Cartaud, M. Recouvreur, R. Grailhe, V. Dufresne, E.
ll-based crystallization procedure. Shown are photographs of cubic phase
ll containing 0.2 l cubic phase dispensed with manual syringe dispenser.
etup prior to crystallization. The purple bacteriorhodopsin is homoge-
rystals formed in 0.2 l of cubic phase when solid salt was used as a crys-
up to ca. 75 m along the longest dimension were grown by adding 0.2 l
1 l of 27% PEG 2000 in 100 mM Sørensen phosphate buVer, pH 5.6.
ght (5). [III] Ca. 0.2 l of the preformed lipidic material is added to each
ll bottom and by triggering the semi-automatic dispenser (6). The average
the dispenser was determined to be 222§ 10.3 g. The cubic phase is sta-
th a clear transparent tape (8) and the plates may be stored (9) at diVerent
t without disturbance from both sides simply by Xipping the entire plate.(1998) 11–21.
[3] M.L. Chiu, P. Nollert, M.C. Loewen, H. Belrhali, E. Pebay-Pey-
roula, J.P. Rosenbusch, E.M. Landau, Acta Cryst. D 56 (2000)
781.
[4] V. Cherezov, H. Fersi, M. CaVrey, Biophys. J. 81 (2001) 225–242.
[5] V. Cherezov, M. CaVrey, J. Appl. Cryst. 36 (2003) 1372–1377.
[6] N.A. Dencher, M.P. Heyn, Methods Enzymol. 88 (1982) 5–10.
[7] M. Grabe, J. Neu, G. Oster, P. Nollert, Biophys. J. 84 (2003) 854–
Pebay-Peyroula, E.M. Landau, J.P. Changeux, Proc. Natl. Acad.
Sci. USA 100 (20) (2003) 11309–11314.
[20] L. Packer, Methods Enzymol. 88 (1982) 1–27.
[21] E. Pebay-Peyroula, R. Neutze, E.M. Landau, BBA 1460 (1) (2000)
119–132.
[22] S. Rouhani, M.T. Facciotti, G. Woodcock, V. Cheung, C. Cunn-
ingham, D. Nguyen, B. Rad, C.T. Lin, C.S. Lunde, R.M. Glaeser,
Biopolymers 66 (5) (2002) 300–316.868.
[8] C. Kiweon, P.J. Reeves, H.G. Khorana, Proc. Natl. Acad. Sci. USA
97 (7) (2000) 3016–3021.
[9] E.M. Landau, J.P. Rosenbusch, Proc. Natl. Acad. Sci. USA 93
(1996) 14532.
[23] G. Rummel, A. Hardmeyer, C. Widmer, M. Chiu, P. Nollert, K.
Locher, I. Pedruzzi, E.M. Landau, J. Rosenbusch, J. Struct. Biol.
121 (1998) 1–11.
[24] C. Sennoga, A. Heron, J.M. Seddon, R.H. Templer, B. Hankamer,
Acta Cryst. D 59 (2003) 239–246.
Sign up today - FREE
Mendeley saves you time finding and organizing research. Learn more
- All your research in one place
- Add and import papers easily
- Access it anywhere, anytime


