Traffic 2001 2: 99–104
Munksgaard International Publishers

Toolbox

Strategies for Prokaryotic Expression of Eukaryotic
Membrane Proteins

Rico Laage

1

and Dieter Langosch*

Universita¨t Heidelberg, Neurobiology Department, Im
Neuenheimer Feld 364, 69120 Heidelberg, Germany
* Corresponding author: D. Langosch,
[e-mail]

High-level heterologous expression of integral mem-
brane proteins at full-length is a useful tool for their
structural and functional characterization. Here, systems
that have previously been used for efficient bacterial
expression of eukaryotic membrane proteins are re-
viewed and novel vectors consisting of a modular fusion
moiety based on nuclease A from

Staphylococcus aureus

are presented.

Key words: Expression, synaptobrevin, syntaxin, SNARE,
transmembrane segment,

E. coli

Received 6 June 2000, revised and accepted for publica-
tion 20 October 2000

Heterologous expression of recombinant proteins or protein
fragments is an extremely powerful tool in the analysis of
structure – function relationships of membrane proteins, e.g.,
of components mediating intracellular traffic. Whereas the
use of eukaryotic host systems based on SF9 cells, yeast, or
mammalian cell lines involves cumbersome procedures and
frequently results in low protein yields (1,2), overexpression
in

Escherichia coli is straightforward and has the potential to

produce large quantities of recombinant protein. High-level
bacterial expression has become a routine procedure for
soluble proteins and has been discussed in several excellent
reviews (3 – 6).

The general strategy for bacterial expression is to genetically
fuse the protein of interest to the N- or C-termini of certain
bacterial proteins and/or short sequence tags. These fusions
are designed to allow for efficient expression, detection, and
affinity purification of the expressed product. The most popu-
lar fusion domains include glutathione-S-transferase (GST,
pGEX vectors, Pharmacia Biotechnology, Freiburg, Germany),
maltose binding protein (malE, pMAL vectors, New England
Biolabs), thioredoxin (ThioFusion system, Invitrogen, Carls-
bad, CA, USA), and chitin-binding domain/intein (IMPACT

vectors, New England Biolabs, Beverly, MA, USA). These
larger fusion domains facilitate efficient expression and allow
for subsequent purification by using appropriate affinity
resins. In a number of vector systems, cleavage sites for
site-specific proteases (e.g., thrombin, factor Xa, etc.) have
been introduced to allow for the removal of the fusion do-
main. In addition, several low-molecular weight sequence
tags that facilitate detection, specific precipitation, and
affinity purification have been developed. These include
calmodulin-binding peptide (pCAL vectors, Stratagene, La
Jolla, CA, USA), the strep-tag (pASK-IBA vectors, IBA, Institut
fu¨r Bioanalytik, Go¨ttingen, Germany), a hexa-histidine se-
quence (pET vectors, Novagen; pQE vectors, Qiagen, Valen-
cia, CA, USA), and the S-tag (pET vectors, Novagen). Further,
epitope tags recognized by monoclonal antibodies (e.g. MYC,
hemaglutinin (HA), FLAG, and T7 tags) allow detection and
immuno-precipitation. Finally, N-terminal fusions to bacterial
ompT or pelB leader sequences can be used to target the
expressed proteins into the periplasmic space, where they
are separated from the bulk of cytosolic proteins and pro-
tected from intracellular proteases.

Efficient expression requires use of an efficient promoter
driving expression of the plasmid-borne protein and the
choice of an appropriate host cell system. The combination
of the bacteriophage T7 promoter with BL21 (DE3) host cells
appears to be the most frequently used expression system
to date. There, the gene encoding T7 RNA polymerase is
produced from the chromosomally integrated

l-lysogen DE3;

its expression is under lacUV5 control and is induced by the
addition of isopropyl-

b-

D

-thiogalactopyranoside (IPTG). Alter-

natively, expression can be induced by infection of the cells
with recombinant

l phages encoding T7 polymerase, thus

minimizing premature expression of toxic recombinant
protein. Other promoters that have been used include the

l

bacteriophage P

L

, Tac, trx, arabinose, phoA, trpE promoter’s,

etc. (3 – 6).

Systems for High-level Expression of
Eukaryotic Membrane Proteins

With integral membrane proteins, high-level expression in

E.

coli remains a difficult task due to toxic effects exerted by
hydrophobic protein domains on the host cells. Therefore,
bacterial expression of membrane proteins has frequently
been restricted to their soluble domains. On the other hand,
over-expression of full-length membrane proteins is highly
desirable since their transmembrane segments (TMS) often
contain important structural information directing their fold-

1

Present address: Biochemie-Zentrum Heidelberg, Universita¨t

Heidelberg, Im Neuenheimer Feld 328, 69120 Heidelberg,
Germany.

99

Laage and Langosch

ing, oligomerization, or subcellular sorting (7 – 13). Indeed,
different prokaryotic membrane proteins, including their
membrane-spanning domains, have been over-expressed for
functional and structural studies. Examples include bacterio-
rhodopsin fragments (14), sensory rhodopsin I (15), the Mscl
stretch-sensitive channel (16), the Kscal potassium channel
(17), the outer membrane channel ompA (18), and lactose
permease (19). Furthermore, a growing number of full-length
eukaryotic membrane proteins have been successfully ex-
pressed in different laboratories employing various expres-
sion systems. Prominent examples are briefly discussed
below and summarized in Table 1.

In general, low-level bacterial expression of membrane
proteins may lead to their integration into the inner mem-
brane, where they display biological activity. For example,
different eukaryotic TMS sequences were expressed in the
context of a tripartite fusion protein consisting of the bacte-
rial ToxR transcription activator domain, a TMS of choice, and
malE. These chimera were inserted into the inner

E. coli

membrane and were allowed to determine TMS – TMS inter-
actions based on transcription activation of a reporter gene
(8,9,20). Inner membrane integration was also achieved with
several G-protein-coupled receptors (GPCRs), which consti-

tute a large superfamily of proteins characterized by seven
TMSs. As GPCRs are major drug targets, their heterologous
expression for ligand-binding and structural studies is pur-
sued with great emphasis. Grisshammer and colleagues
(21 – 23) aim at integrating functional GPCRs into the

E. coli

inner membrane. As the integration step appears to be the
rate-limiting factor, a combination of low copy number plas-
mid, weak promoter, and low temperature during expression
proved to be most beneficial. Thus, they achieved membrane
localization of neurotensin and neurokinin-2 receptors upon
fusion to the C-terminus of malE. The correct conformations
of the recombinant receptors were ascertained by specific
ligand binding. Neurotensin receptor expression was signifi-
cantly improved upon addition of an N-terminal signal peptide
(21,22). Functional GPCR expression appears to strongly de-
pend on the nature of C-terminal tags. When fusions of
truncated rat neurotensin receptor with malE were ex-
pressed with a variety of C-terminal affinity tags, functional
expression was highest (800 receptors/cell) when thiore-
doxin was placed between the receptor C-terminus and the
tag. The fusion protein with the

in vivo-biotinylated Bio-tag

was purified with the greatest efficiency and compared to
receptors with the Strep tag or a hexa-histidine tail (23).

Table 1: Eukaryotic integral membrane proteins expressed in bacterial hosts

Expressed protein

Fusion partner

Reference

TMS*

loc***

Promoter (host cell)**

IB

(7,8)

T7 (BL21(DE3))

His

6

/MYC/nuclease A/HA

1

Synaptobrevin II, syntaxin 1A
Synaptobrevin II, Sed5, Bos1,

GST

T7 (BL21(DE3), AB1899)

1

(31 – 35)

Bet2, Gos1, Sft1, Nyv1,
Vam3

Synaptobrevin II, syntaxin 1A,

His

6

, His

8

1

T7 (BL21(DE3))

IB

(33 – 36)

Sec22, VtiI, Snc1, Snc2,
Sso1, Tlg1

Nuclease A

T7 (BL21(DE3)

Glycophorin A

IB

1

(10,38)

phoA (MM294)

Nuclease A

phoA (DH5

a)

IB

(12)

Phospholamban

1

(27)

IB

Neu TMS fragments

TrpE (JM101, JM109

Anthranilate synthase

1

BL21(DE3), DH5

a))

None

F-ATPase subunit b

T7 (BL21(DE3), C41(DE3),

2

IB

(28)

C43(DE3))

6

(28)

IB

Oxoglutarate–malate trans-

T7 (BL21(DE3), C41(DE3),

None

porter

C43(DE3))

6

IB

Phosphate carrier

(28)

T7 (BL21(DE3), C41(DE3),
C43(DE3))

ADP/ATP carrier

T7 (BL21(DE3), C41(DE3),

6

IB

(28)

C43(DE3))

Neurotensin receptor

7

MalE, His

5

, MYC, strep tag,

(21,23)

T7 (DH5

a), lac (DH5a)

M

Bio tag, thioredoxin

Neurokinin receptor-2

7

MalE

(22)

M

tac (DH5

a)

M

Bacterioopsin fragments, His

6

,

7

bacterioopsin-promoter (

H

.

(26)

Muscarinic receptor M1,

halobium

)

serotonin receptor 5HT2c,

HA

alpha mating factor receptor
Ste2

Various catecholamine,

(24,25)

7

GST, His

6

T7 (BL21(DE3))

IB

peptide and olfactory
receptors, opsins

*Number of transmembrane segments; **promoter (host cell system) used; ***location of recombinant protein in the cells.
M, membrane; IB, inclusion bodies.

100

Traffic 2001: 2: 99 – 104

Prokaryotic Expression of Membrane Proteins

Turner et al. (26) developed a set of expression vectors
where other GPCRs, the muscarinic receptor M1, the sero-
tonin receptor 5HT2c, and the yeast alpha mating factor
receptor Ste2 were fused to upstream regulatory sequences
and various coding regions of bacterio-opsin. These fusions
were expressed at medium yields in

Halobacterium halobium

and targeted to the membrane. In agreement with Gris-
shamer’s results (23), the addition of various C-terminal
affinity tags reduced expression yield, which was attributed
to transcriptional effects. In contrast to low-level expression,
recombinant protein yields may be increased about 100-fold
using strong (e.g., T7) promoters and appropriate host cells.
However, high levels of recombinant protein are either toxic
to the cells or lead to the formation of insoluble inclusion
bodies, which may account for up to about 40% of total
bacterial protein. In inclusion bodies, the recombinant
proteins are usually non-toxic to the cell, protected from
proteolytic degradation, and can be easily separated from the
bulk of soluble host proteins upon cell lysis. Therefore, ex-
pression as inclusion bodies has mostly been the method of
choice to obtain large quantities of recombinant protein. It
has to be borne in mind, however, that proteins forced into
inclusion bodies usually adopt non-native conformations.
Therefore, appropriate strategies for solubilization and refold-
ing have to be devised in order to obtain functionally active
protein.

Eleven G-protein-coupled receptors were expressed in

E. coli

as fusion proteins with an N-terminal GST domain and a
C-terminal hexa-histidine tag by Kiefer and colleagues. Ex-
pression levels varied between 0.1 and 10% of the total
cellular protein. Low expression levels coincided with a toxic
effect of the recombinant proteins, which could be largely
attributed to the positive charge content in the loop regions
between the TMSs. Interestingly, the introduction of posi-
tively charged amino acids into the loop regions avoided
insertion into the bacterial membrane and thus favored inclu-
sion body formation (24). The GPCRs could be refolded by
detergent solubilization into conformations displaying specific
ligand binding (25).

To express hydrophobic fragments of the

Neu proto onco-

gene, a single-span receptor tyrosine kinase, several alterna-
tive

approaches

were

tried.

Whereas

expression

as

GST-fusion protein, or expression directly under control of
the T7 promoter, failed to yield sufficient product, fusion to
the C-terminus of anthranilate synthase of the hexa-histidine-
tagged fragment gave satisfactory results. The inclusion
body isolate was cleaved by cyanogen bromide treatment in
70% formic acid and the thus liberated Neu TMS domains
purified by metal – chelate affinity chromatography (27).

An interesting strategy to optimize the host cell system was
described by Miroux et al. (28) after finding that different
membrane proteins killed the majority of expressing BL21
(DE3) host cells. From the few survivors they isolated mutant
derivatives, termed C41 (DE3) and C43 (DE3). These mutant
host cells were found to favor over-expression of oxoglu-

tarate – malate transport protein, a bovine phosphate carrier, a
bovine ADP/ATP carrier, bovine F-ATPase subunit b, as well
as several bacterial membrane proteins and soluble proteins,
about 4 – 10-fold when compared with expression levels ob-
tained with the parental BL21 (DE3) cells. Although these
mutant hosts did not improve expression of some other
membrane proteins (our unpublished observations) it is evi-
dent that selection of host cell derivatives that survive high-
level expression may be useful for obtaining efficient
expression

hosts

tolerating

certain

toxic

recombinant

proteins.

Several reports describe the expression of soluble NSF (N-
ethylmaleimide-sensitive factor) attachment protein receptor
(SNARE) proteins that are essential for fusion of cargo vesi-
cles with cognate intracellular organelles or the plasma mem-
brane during intracellular traffic. In some studies, bacterial
expression was limited to their cytoplasmic domains lacking
the C-terminal TMSs in order to perform biochemical and
structural studies (29,30). More recently, expression of full-
length SNAREs has been achieved using different fusion
partners. In the majority of reports, coding regions corre-
sponding to rat synaptobrevin II or several yeast SNAREs
(Sed5, Bos1, Bet2, Gos1, Sft1, Nyv1, Vam3) have been fused
to the C-terminus of GST using pGEX vectors. The proteins
were solubilized from inclusion bodies using different deter-
gents and affinity purified on glutathione agarose beads.
They were used for biochemical studies as well as for recon-
stitution of SNARE function upon integration into liposomal
membranes with or without prior proteolytic removal of the
GST portion by thrombin (31 – 35).

Some SNAREs, including synaptobrevin II, syntaxin 1A (36),
as well as yeast SNARES (Sec22 (33); Vti1 (35); Snc1, Snc2,
Sso1, Tlg1 (34)), could even be expressed with sole N- or
C-terminal hexa- or octa-histidine tags using pET vectors.
These tags were used for affinity purification by metal –
chelate affinity chromatography, using columns with immobi-
lized Zn

2 +

or Ni

2 +

ions chelating the histidine sequence.

With some of these constructs, induction with decreased
IPTG concentrations, induction for shorter periods, expres-
sion at 16 or 25°C instead of 37°C, and/or using a codon-op-
timized host strain, proved to increase expression yield
(33,34). This highlights the benefit of optimizing expression
conditions for each individual case of membrane protein.

Taken together, it appears that some SNARE proteins can be
expressed efficiently without a large fusion partner, whereas
others require it. A disadvantage of the most frequently
employed fusion partner, i.e., GST, is its dimeric structure
(37). Therefore, GST fusions are impractical for the analysis
of membrane protein oligomerization or for determining the
stoichiometry of large multi-subunit complexes. Clearly, re-
moval of the GST portion by site-specific proteolysis is
preferable before such types of analyses are performed. On
the other hand, removal of fusion domains is associated with
a significant loss of recombinant protein.

101

Traffic 2001: 2: 99 – 104

Laage and Langosch

An alternative to GST was developed by Engelman and
colleagues (12,13) when they described nuclease A, an 18
kD soluble protein from

Staphylococcus aureus, as a novel

fusion partner. Nuclease A is monomeric and therefore does
not perturb the quaternary structure of the recombinant
protein under study. Upon fusion to the nuclease A C-termi-
nus, the single-span membrane proteins, glycophorin A and
phospholamban, have been successfully expressed and sub-
jected to extensive quaternary structure analysis (10,12).
Proteolytic removal of the nuclease A moiety by trypsin
digestion was used to prepare a TMS peptide of glycophorin
A for subsequent structural analysis by solution nuclear mag-
netic resonance (NMR) spectroscopy (38).

In order to obtain vectors based on nuclease A that are more
versatile than the originally (10) described one, we combined
several useful features of other expression systems. In the
following, we describe this system and outline its usefulness
in

studying

protein – protein

interactions

of

membrane

proteins. Our pSNiR vectors are based on pET21 and contain
different modular expression cassettes under transcriptional
control of the T7

lac promoter. In the basic vector, pSNiR, the

cassette consists of a MYC marker epitope, nuclease A, and
an alanine spacer region linking this fusion moiety to the
N-terminus of the protein to be expressed. In pSNiR2, an
N-terminal hexa-histidine tag allows for affinity purification by
metal – chelate chromatography. In pSNiR3, an optimized
thrombin cleavage site (39) and HA epitope tag between the
nuclease moiety and the spacer region allows for site-spe-
cific proteolytic cleavage of the nuclease moiety from the
desired protein (Figure 1). Since the latter contains an HA
epitope, it can be individually detected by HA antibodies.

Characteristics of Nuclease A Fusion Proteins

Synaptobrevin II or syntaxin 1A expressed in the context of
any of the pSNiR vectors yielded ca. 50 mg of recombinant
protein per liter of culture medium. The proteins formed
inclusion bodies, which could be enriched by centrifugation
and solubilized with a variety of detergents, including 3-[(3-
cholamidopropyl)

dimethyl-ammonio]-1-propane

sulfonate,

polyoxyethylene-9-lauryl ether (CHAPS), Triton X-100, sodium
cholate, dodecyl maltoside, or sodium deoxycholate. Up to
50% of the expressed proteins was solubilized depending on
the type of detergent used and on salt concentration. In
quaternary

structure

analysis,

homo-

and

heterotypic

protein – protein interactions between both proteins could be
demonstrated using a variety of experimental approaches,
including chemical cross-linking, mild sodium dodecylsulfate-
polyacrylamide gel electrophoresis (SDS-PAGE), radioactive
overlay procedures, and co-immunoprecipitation (7,8). Perfor-
mance of pSNiR3 is exemplified in Figure 2 for synaptobrevin
II expression. While homodimerization of the uncleaved fu-
sion protein is detectable by SDS-PAGE only under mild
conditions (7,8), removal of nuclease A by thrombin cleavage
preserves significant homodimerization of free synaptobrevin
II, even upon sample boiling (compare lanes 3 and 4 in Figure
2B). Therefore, removal of the fusion domain may be advan-
tageous when low-affinity interactions are studied. Removal
appears to be, however, not required for the analysis of
high-affinity interaction domains (10,12). Further, the nucle-
ase A domain should not be removed from short protein
fragments containing membrane-spanning domains, which
may otherwise precipitate from detergent solution due to
their hydrophobicity (9). Nuclease A is a highly basic protein
whose three-dimensional structure reveals numerous lysine
residues exposed to solvent (40). Therefore, it reacts effi-
ciently with lysine-reactive cross-linking reagents, fluorescent
dyes, etc. (Reference (7) and our unpublished results). In
contrast to GST, nuclease A is devoid of cysteine residues;
thus it does not interfere with disulfide-mapping (41) of
protein – protein interaction domains. Due to its globular
structure (40), it migrates according to its predicted molecu-

Figure 1: Organization of pSNiR3 vector. A, Plasmid map: T7
p/o designates phage T7

lac

promoter/operator region; his

6

, myc,

nuc A, thrombin, and HA designate sequence segments encod-
ing the hexahistidine tag, MYC marker epitope, nuclease A,
thrombin cleavage site, and HA epitope, respectively; mcs,
multiple cloning site. Coding regions inserted into the multiple
cloning site may carry their original stop codons; alternatively,
the vector encoded stop codon 3

% of the BamHI site may be

used; f1, intergenic region of filamentous phage f1; Ap, ampi-
cillin resistance gene; ori, pBR322 origin of replication; lacI,
promoter and coding sequence of the

lac

repressor. Translational

start and stop sites are indicated. The whole expression cassette
may be excised by XbaI and XhoI. B, Nucleotide sequence of
that part of the expression cassette harboring the thrombin site
optimized by a subsequent gly-spacer, the HA epitope and the
unique restriction sites. The encoded amino acid sequence (in
single letter code) is shown up to the NheI site.

102

Traffic 2001: 2: 99 – 104

Prokaryotic Expression of Membrane Proteins

Figure 2: Expression, purification, and cleavage of the synaptobrevin II fusion protein derived from pSNiR3.Syb. The fusion
protein was expressed in BL21(DE3)pLysS cells (Novagen, Madison, WI, USA) by induction of T7 polymerase with 1 mM IPTG at a
cell density of A

600

:0.3 (7). Purification and cleavage of the fusion protein was achieved in a one-step procedure. The solubilized

protein was immobilized on a Zn

2 +

-affinity column (1 ml HiTrap©, Pharmacia Biotechnology) and washed with buffer (20 mM NaP

i

,

pH 7.2, 0.5 M NaCl, 1% (w/v) CHAPS). The C-terminal 18 kD synaptobrevin was separated from the nuclease domain by site-specific
proteolytic cleavage with thrombin (Pharmacia, 25

m/ml for 16 h at 20°C) and subsequently eluted in \90% pure form. The 21 kD

fusion domain was eluted using high imidazole concentrations (

\200 mM). A, Coomassie-stained 15% SDS-polyacrylamide gel. B, C,

Western blots with MYC or HA antibodies. Lanes: 1, uninduced cells transformed with pSNiR3.Syb; 2, lysate of IPTG-induced cells
containing the 36 kD fusion protein; 3, fusion protein solubilized with CHAPS; 4, HA-tagged synaptobrevin II proteolytically cleaved
from immobilized fusion protein and separated from the nuclease A moiety by washing; 5, remaining MYC-tagged nuclease A moiety.
Whereas the proteins in lanes 2 – 5 originated from 100

ml of bacterial culture each, lane 1 corresponds to only 25 ml of culture since

uninduced bacteria grow to higher cell densities. Note that both antibodies stain the intact fusion protein while the synaptobrevin part
is only recognized by the HA antibody (B) and the nuclease A moiety only by the MYC antibody (C). The small amounts of
immunoreactive material migrating below the fusion proteins are attributed to partial degradation of the recombinant protein. The
arrowhead in B denotes the 34 kD synaptobrevin dimer that was observed only upon removal of the nuclease A moiety under these
conditions.

lar mass in size exclusion chromatography and thus permits
analysis of multimeric structures in the presence of non-de-
naturing detergents (9). We expect, therefore, that our ver-
satile pSNiR vectors will become useful for membrane
protein research.

Acknowledgments

We thank Markus Seiler and Bettina Brosig for expert help with plasmid

construction and Dr. W. B. Huttner for continuous support. This work was

supported by the Deutsche Forschungsgemeinschaft (SFB 317 and

Heisenberg Programm) and the Fonds der Chemischen Industrie.

References

1.

Grisshammer R, Tate CG. Overexpression of integral membrane
proteins for structural studies. Q Rev Biophys 1995;28: 315 – 422.

2.

Tate CG, Grisshammer R. Heterologous expression of G-protein-
coupled receptors. TIBTECH 1996;14: 426 – 430.

3.

Makrides SC. Strategies for achieving high-level expression of
genes in

Escherichia coli. Microbiol Rev 1996;60: 512 – 538.

4.

Uhlen M, Moks T. Gene fusions for purpose of expression: an
introduction. Meth Enzymol 1990;185: 129 – 144.

5.

Murby M, Uhlen M, Stahl S. Upstream strategies to minimize prote-
olytic degradation upon recombinant production in

Escherichia coli.

Prot Express Purif 1996;7: 129 – 136.

6.

Sheibani N. Prokaryotic gene fusion expression systems and their
use in structural and functional studies of proteins. Prep Biochem
Biotechnol 1999;29: 77 – 99.

7.

Laage R, Langosch D. Dimerization of the synaptic vesicle protein
synaptobrevin

/VAMP II depends on specific residues within the

transmembrane segment. Eur J Biochem 1997;249: 540 – 546.

8.

Laage R, Rohde J, Brosig B, Langosch D. A conserved membrane-
spanning amino acid motif drives homomeric and supports het-
eromeric assembly of presynaptic SNARE proteins. J Biol Chem
2000;23: 17481 – 17487.

9.

Gurezka R, Laage R, Brosig B, Langosch D. A heptad motif of
leucine residues found in membrane proteins can drive self-assem-
bly of artificial transmembrane segments. J Biol Chem 1999;274:
9265 – 9270.

10.

Lemmon MA, Flanagan JM, Hunt JF, Adair BD, Bormann B-J,
Dempsey CE, Engelman DM. Glycophorin A dimerization is driven
by specific interactions between transmembrane alpha-helices. J
Biol Chem 1992;267: 7683 – 7689.

11.

Cosson P, Bonifacino JS. Role of transmembrane domain interac-
tions in the assembly of class II MHC molecules. Science 1992;258:
659 – 662.

12.

Arkin IT, Adams PD, MacKenzie KR, Lemmon MA, Bru¨nger AT,
Engelman DM. Structural organization of the pentameric transmem-
brane alpha-helices of phospholamban, a cardiac ion channel.
EMBO J 1994;13: 4757 – 4764.

13.

Lemmon MA, Engelman DM. Specificity and promiscuity in mem-
brane helix interactions. Q Rev Biophys 1994;27: 157 – 218.

14.

Marti T. Refolding of bacteriorhodopsin from expressed polypeptide
fragments. J Biol Chem 1998;273: 9312 – 9322.

15.

Schmies G, Chizhov I, Engelhard M. Functional expression of His-
tagged sensory rhodopsin I in

Escherichia coli. FEBS Lett 2000;466:

67 – 69.

16.

Chang G, Spencer RH, Lee AT, Barclay MT, Rees DC. Structure of
the MscL homolog from mycobacterium tuberculosis: a gated
mechanosensitive ion channel. Science 1998;282: 2220 – 2226.

17.

Doyle DA, Cabral JM, Pfuetzner RA, Kuo A, Gulbis JM, Cohen SL,
Chait BT, MacKinnon R. The structure of the potassium channel:
molecular

basis

of

K

+

conduction

and

selectivity.

Science

1998;280: 69 – 77.

18.

Guiliani N, Jerez CA. Molecular cloning, sequencing, and expression
of omp-40, the gene coding for the major outer membrane protein
from the acidophilic bacterium

Thiobacillus ferrooxidans. Appl Envi-

ron Microbiol 2000;66: 2318 – 2324.

103

Traffic 2001: 2: 99 – 104

Laage and Langosch

19.

Kaback HR, Voss J, Wu J. Helix packing in polytopic membrane
proteins: the lactose permease of

Escherichia coli. Curr Opin Struct

Biol 1997;7: 537 – 542.

20.

Langosch DL, Brosig B, Kolmar H, Fritz H-J. Dimerisation of the
glycophorin A transmembrane segment in membranes probed with
the ToxR transcription activator. J Mol Biol 1996;263: 525 – 530.

21.

Grisshammer R, Duckworth R, Henderson R. Expression of a rat
neurotensin receptor in

Escherichia coli. Biochem J 1993;295: 571 –

576.

22.

Grisshammer R, Little J, Aharony D. Expression of rat NK-2 (neu-
rokinin A) receptor in

E. coli. Receptors Channels 1994;2: 295 – 302.

23.

Tucker J, Grisshammer R. Purification of a rat neurotensin receptor
expressed in

Escherichia coli. Biochem J 1996;317: 891 – 899.

24.

Kiefer H, Vogel R, Maier K. Bacterial expression of G-protein-cou-
pled receptors: prediction of expression levels from sequence. Re-
ceptors Channels 2000;7: 109 – 119.

25.

Kiefer H, Krieger J, Olszewski JD, Von Heijne G, Prestwich GD,
Breer H. Expression of an olfactory receptor in

Escherichia coli:

purification,

reconstitution,

and

ligand

binding.

Biochemistry

1996;35: 16077 – 16084.

26.

Turner GJ, Reusch R, Winter Vann AM, Martinez L, Betlach MC.
Heterologous gene expression in a membrane-protein-specific sys-
tem. Protein Expr Purif 1999;17: 312 – 323.

27.

Jones DH, Ball EH, Sharpe S, Barber KR, Grant CWM. Expression
and membrane assembly of a transmembrane region from Neu.
Biochemistry 2000;39: 1870 – 1878.

28.

Miroux B, Walker JE. Over-production of proteins in

Escherichia

coli: mutant hosts that allow synthesis of some membrane proteins
and globular proteins at high levels. J Mol Biol 1996;260: 289 – 298.

29.

Sutton RB, Fasshauer D, Jahn R, Bru¨nger AT. Crystal structure of a
SNARE complex involved in synaptic exocytosis at 2.4 A

, resolution.

Nature 1998;395: 347 – 353.

30.

Hayashi T, Yamasaki S, Nauenburg S, Binz T, Niemann H. Disas-
sembly of the reconstituted synaptic vesicle membrane fusion com-
plex

in vitro. EMBO J 1995;14: 2317 – 2325.

31.

Weber T, Zemelman BV, McNew JA, Westermann B, Gmachl M,
Parlati F, So¨llner TH, Rothman JE. SNAREpins: minimal machinery
for membrane fusion. Cell 1998;92: 759 – 772.

32.

Poirier MA, Hao JC, Malkus PN, Chan C, Moore MF, King DS,
Bennett MK. Protease resistance of syntaxin-SNAP-25-VAMP com-
plexes — implications for assembly and structure. J Biol Chem
1998;273: 11370 – 11377.

33.

Parlati F, McNew JA, Fukuda R, Miller R, So¨llner TH, Rothman JE.
Topological restriction of SNARE-dependent membrane fusion. Na-
ture 2000;407: 194 – 198.

34.

McNew JA, Parlati F, Fukuda R, Johnston RJ, Paz K, Paumet F,
So¨llner TH, Rothman JE. Compartmental specificity of cellular mem-
brane fusion encoded in SNARE proteins. Nature 2000;407: 153 –
159.

35.

Fukuda R, McNew JA, Weber T, Parlati F, Engel T, Nickel W,
Rothman JE, So¨llner TH. Functional architecture of an intracellular
membrane t-SNARE. Nature 2000;407: 198 – 202.

36.

Margittai M, Otto H, Jahn R. A stable interaction between syntaxin
1a and synaptobrevin 2 mediated by their transmembrane domains.
FEBS Lett 1999;446: 40 – 44.

37.

Tudyka T, Skerra A. Glutathione S-transferase can be used as a
C-terminal, enzymatically active dimerization module for a recombi-
nant protease inhibitor, and functionally secreted into the periplasm
of

Escherichia coli. Protein Sci 1997;6: 2180 – 2187.

38.

MacKenzie KR, Prestegard JH, Engelman DM. A transmembrane
helix dimer: structure and implications. Science 1997;276: 131 – 133.

39.

Guan K, Dixon JE. Eukaryotic proteins expressed in

Escherichia

coli: an improved thrombin cleavage and purification procedure of
fusion proteins with glutathione S-transferase. Anal Biochem
1991;192: 262 – 267.

40.

Heynes TH, Fox RO. The crystal structure of

Staphylococcus nucle-

ase refined at 1.7 Angstroms resolution. Proteins 1991;5: 183 – 201.

41.

Pakula AA, Simon MI. Determination of transmembrane protein
structure by disulfide cross-linking: the

Escherichia coli Tar recep-

tor. Proc Natl Acad Sci USA 1992;89: 4144 – 4148.

104

Traffic 2001: 2: 99 – 104