Ban cao Xuan Hieu co hoi "Xin cho ví dụ thành phần buffer. Sau khi có protein fraction rồi thì hòa SDS sample buffer ntn? Có heat trước khi điện di SDS-PAGE hay k? Nếu có thì ntn? "
Toi xin tra loi rang
Doi voi hydrophobic protein, thi can phai "try and error", u learn it by doing yourself experiments,
Va hinh nhu cai ban hoi khong phai lien quan den van de purify hydrophobic protein ma la chay SDS gel cho no, noi nhu vay thi toi nghi khong co gi dac biet o buoc nay, ma cai kho no nam o cho purify nhom protein nay, nguoi ta se phai thu cac detergent, va nhieu khi ket hop ca mot so dung moi dac biet nua, nhu toi da de cap,ban co the dung aceton nitril (mot dung moi pho biet de chay HPLC) va dung de rua hay hoa tat cac hop chat hydrophobic, con luong cho vao bao nhieu, phan phai tu tim lay, thong thuong co the tu 5 den 60%
ban nen tham khao mot so reference: protein handbooks Humana Press, (toi nghi ban da co trong tay book nay)
hay article duoi day, xin loi ban toi khong biet lam the nao de attach duoc PDF vao post nay, nen neu ban can toi se co the email
chuc ban thanh cong
International Journal of Biological Macromolecules 39 (2006) 83–87
A simple strategy towards membrane protein purification
Damian Niegowski a,b, Marie Hedr´en a, P¨ar Nordlund a, Said Eshaghi a,∗
a Department of Medical Biochemistry and Biophysics, Karolinska Institutet, SE-171 77 Stockholm, Sweden
b Department of Biochemistry and Biophysics, Arrhenius Laboratories for Natural Sciences, Stockholm University, SE-106 91 Stockholm, Sweden
Received 11 November 2005; received in revised form 13 December 2005; accepted 9 February 2006
Available online 23 February 2006
A simple and cost-efficient detergent screening strategy has been developed, by which a number of detergents were screened for their efficiency
to extract and purify the recombinant ammonium/ammonia channel, AmtB, from Escherichia coli, hence selecting the most efficient detergents
prior to large-scale protein production and crystallization. The method requires 1ml cell culture and is a combination of immobilized metal ion
affinity chromatography and filtration steps in 96-well plates. Large-scale protein purification and subsequent crystallization screening resulted
in AmtB crystals diffracting to low resolution with three detergents. This strategy allows exclusion of detergents with the lowest probability in
yielding protein crystals and selecting those with higher probability, hence, reducing the number of detergents to be screened prior to large-scale
membrane protein purification and perhaps also crystallization.
© 2006 Elsevier B.V. All rights reserved.
Keywords: Membrane proteins; Detergent screen; High-throughput
The recombinant expression and subsequent purification of
integral membrane proteins are considered major challenges,
and together with the crystallization step, the major hurdles
towards routine structure determination of membrane proteins.
Consequently, the number of membrane proteins with known
structure has remained negligible as compared to those of soluble
proteins [1,2]. This in turn causes serious lack of information
in the field of drug discovery, since membrane proteins already
cover more than 50% of the current drug targets [3,4]. Therefore,
it is essential to improve the success rate in membrane
proteins structural determination, by dealing with experimental
difficulties in the production of these proteins.
Experimental procedures for handling and isolating integral
membrane proteins are generally more challenging than their
soluble counterparts, since the former requires purification in
detergent. General experiences from workers in the field with
This paper was presented at the “Challenging Proteins Workshop” in Paris,
October 17–18, 2005.
∗ Corresponding author. Tel.: +46 8 524 86863; fax: +46 8 524 868650.
E-mail address: firstname.lastname@example.org
the problematic experimental behaviour of integral membrane
proteins have lead to the expectation that these proteins are dramatically
harder to produce than soluble proteins. One of the
reasons may be the usage of the wrong detergent during extraction
and purification. There are dozens of different detergents
that are commonly used, dozens more that are less characterized
but still probably useful, and many novel detergents under
development. It has also been reported that some compartments
of the cell membrane show resistance towards certain detergents
. Moreover, mixtures of detergents are sometimes used during
purification and crystallization [6,7]. Altogether, the size
of the detergent parameter space becomes very large. Therefore,
it is crucial to choose the right detergent for an efficient
extraction and purification of the membrane protein of interest.
Recently, we reported an efficient strategy, by which we could
screen the expression of a number of membrane proteins in a
high-throughput manner . Here, we report the application of
this strategy to design downstream protocols for large-scale production
of the recombinantly expressed ammonium transporter
AmtB from Escherichia coli, by which 26 detergents, 4 types
of chromatography columns and various buffer conditions have
been screened using a 96-well plate format. The method is very
cost efficient and may easily be applied to other membrane pro-
0141-8130/$ – see front matter © 2006 Elsevier B.V. All rights reserved.
84 D. Niegowski et al. / International Journal of Biological Macromolecules 39 (2006) 83–87
teins. In addition, we have tested the feasibility of using this platform
for choosing the right detergent for crystallization attempts.
We believe that using the presented strategy will minimize the
number of detergent candidates in initial crystallization screens,
by excluding those that will yield low amount of proteins.
2. Materials and methods
All the detergents were purchased from Anatrace (OH, USA).
Coding sequences for the AmtB was obtained from SwissProt
database and used for primer design usingSGDWebprimer
design program. The cloning was performed as described in
. Briefly, touchdown PCR was performed using Platinum
PfxDNApolymerase (Invitrogen, Stockholm, Sweden); primers
containing at least 15 gene-specific nucleotides (TAG Copenhagen,
and the E. coli template K12
isolated with High Pure PCR Template Preparation Kit (Roche,
Another PCRwas performed using the same Pfx polymerase,
primers containing attB sites (Invitrogen), and the PCR products
from the touchdown reactions as templates. The linear fragments
flanked by attB-sequences were subjected to site-specific recombination
with pDONR201 vector (Invitrogen), containing the
ccdB gene, flanked by attP sites and catalyzed by BP Clonase
(see manufacturer’s protocols) yielding entry clones. The entry
clones were subjected to another round of site-specific recombination
catalyzed by the LR Clonase enzyme mix (Invitrogen) in
order to subclone the genes of interest into the destination vector
(AstraZeneca, S¨odert¨alje, Sweden) containing the ccdB gene
flanked by attR sites, as well as the coding sequences for fusion
tags N-terminal FLAG and C-terminal 6-His (see Invitrogen’s
The resulting expression constructs were used to transform
E. coli C41(DE3) (Avidis, Saint-Beauzire, France). Construct
identity was verified by dideoxysequencing.
2.2.2. Protein overexpression and detergent screening
The cells were grown in a shake flask at 37 ◦C until the cultures
reached the OD600 of ∼0.8. The culture was then cooled
down to 20 ◦C and induced overnight with 0.1mMisopropyl--
d-1-thiogalactoside. Aliquots of 1ml were added to a 96-well
plate and harvested by centrifugation at 3000×g for 10 min.
The overall procedure is outlined in Fig. 1. The cell pellets
were resuspended in 50 l 20mM Tris–HCl pH 8.0, 50mM
NaCl, 0.5mM Tris(2-carboxyethyl)phosphine hydrochloride
(TCEP), 1 mg/ml lysozyme, complete protease inhibitor cocktail
EDTA-free (Roche), 10 units/ml benzonase (VWR International),
and the appropriate detergent at concentrations 1–2%
according to Table 1, respectively. Lysis and extraction were
performed for 1 h at 4 ◦C. The suspensionwas filtered using a 96-
well filter plate (Millipore, Stockholm, Sweden). The filtratewas
Fig. 1. An overall scheme of the detergent screen platform using 96-well plates.
added to 25l Ni-NTA agarose resin (Qiagen, Stockholm, Sweden)
pre-equilibrated with purification buffer (P-buffer) containing
20mMTris–HCl pH 8.0, 300mMNaCl, 0.5mMTCEP, and
the appropriate detergent. After 15 min agitation at 4 ◦C, the
unbound materialwas removed by 30 s centrifugation at 100×g.
The resinwas thenwashed with 20 column volumes (CV) 40mM
imidazole in the P-buffer containing the appropriate detergent
at 100×g for 30 s. The bound recombinant membrane proteins
Detergents and their concentrations used in the screen
Detergent name Abbreviation CMC (mM) Extraction (mM)
FOS-CHOLINE-10 FC10 11 63 (2)
FOS-CHOLINE-11 FC11 1.85 30 (1)
FOS-CHOLINE-12 FC12 1.5 32 (1)
HEGA-10 HEGA-10 7 54 (2)
Nonyl maltoside NM 6 43 (2)
Decyl maltoside DM 1.8 21 (1)
Undecyl maltoside UDM 0.59 20 (1)
Dodecyl maltoside DDM 0.17 20 (1)
CHAPS CHAPS 8 32 (2)
CHAPSO CHAPSO 8 32 (2)
Nonyl thiomaltoside NTM 3.2 32 (1.5)
Decyl thiomaltoside DTM 0.9 20 (1)
Undecyl thiomaltoside UDTM 0.21 19 (1)
Dodecyl thiomaltoside DDTM 0.05 1.9 (1)
Cymal-6 Cy6 0.56 20 (1)
Cymal-7 Cy7 0.19 19 (1)
LDAO LDAO 1 43 (1)
TDAO TDAO 0.29 39 (1)
C8E4 C8E4 8 64 (2)
C8E6 C8E6 10 51 (2)
C10E5 C10E5 0.81 26 (1)
C12E8 C12E8 0.09 19 (1)
Octyl glucoside OG 18 68 (2)
Nonyl glucoside NG 6.5 65 (2)
Triton X-100 TX100 0.23 15 (1)
Triton X-114 TX114 0.20 36 (2)
Values in parenthesis indicate percentages.
D. Niegowski et al. / International Journal of Biological Macromolecules 39 (2006) 83–87 85
were finally recovered in 30 l P-buffer containing 500mMimidazole
and the respective detergent, by centrifugation at 100×g
for 1 min. The eluted material was then analyzed by dot blot.
2.2.3. Large-scale protein purification
Cultures were grown in baffled shake flasks as described
above, and finally harvested by centrifugation at 7000×g for
15 min. Cell pellets were resuspended in 20mM Tris–HCl
pH 8.0, 50mM NaCl, 0.5mM TCEP, 10 u/ml benzonase and
1 mg/ml lysozyme, sonicated and centrifuged at 15,000×g for
10 min to remove cell debris. Cell membranes were finally harvested
by 1 h centrifugation at 150,000×g. The membranes
were resuspended and solubilized with 1% FC12 detergent
in the P-buffer supplemented with 20mM imidazole using a
glass homogenizer, followed by centrifugation for 45 min at
200,000×g to remove unsolubilized material. The clear supernatants
containing solubilized membrane proteins were loaded
on TALON resin pre-equilibrated with the P-buffer, including
20mM imidazole and 0.1% FC12. The resin was washed once
with 15 CV P-buffer containing 20mM imidazole and 0.1%
FC12. When detergent exchanged was desired, 5 CV wash with
0.1% FC12 was performed followed by 10 CV wash with the
appropriate detergent. The recombinant proteins were eluted
with 250mM imidazole in the wash buffer. The IMAC purified
AmtBwas further polished by gel filtration using a Superdex 200
column (GE-Healthcare, Uppsala, Sweden) in 20mMTris–HCl
pH 8.0, 150mMNaCl, 0.5mMTCEP and the appropriate detergent.
2.2.4. Dot blot and SDS-PAGE
One microlitre of sample was applied to nitrocellulose and
allowed to dry. The 6-His-tagged proteins were detected using
INDIA HisProbe-HRP Western blotting probe (Pierce, Stockholm,
Sweden) according to the manufacturer’s protocol. The
signals were detected with FluorS-multiImager (BioRad, Stockholm,
Sweden) and quantified using the Quantity One software
(BioRad). The purity of the expressed proteins was monitored
by SDS–PAGE, using Nu-PAGE 4–12% Bis–Tris-gels (Invitrogen).
The purified proteins were concentrated using Amicon Ultra
100-kDa cut-off (Millipore) to a protein concentration of
4–5 mg/ml. Crystallization was performed by sitting-drop vapor
diffusion at 20 ◦C at a ratio of 1:1. Protein crystals were grown
with undecyl maltoside (0.1%), dodecyl maltoside (0.03%) or
Cymal-6 (0.06%), in either 12% PEG 4000 or 30% PEG 400
buffered with 100mM MOPS pH 7.0 (MbClass II screen, Nextal
Biotechnologies, Montreal, Canada).
Prior to large-scale purification and subsequent crystallization,
we applied the previously reported detergent screening
strategy  in order to explore and refine the extraction and
purification of AmtB. Recombinant AmtB was overexpressed
in a shake flask to ensure equality of the starting material prior
Fig. 2. Detergent screen analysis histograms of AmtB. From left to right: (1)
FC10, (2) FC11, (3) FC12, (4) HEGA10, (5) NM, (6) DM, (7) UDM, (8) DDM,
(9) CHAPS, (10) CHAPSO, (11) NTM, (12) DTM, (13) UDTM, (14) DDTM,
(15) LDAO, (16) TDAO, (17) C8E4, (18) C8E6, (19) C10E5, (20) C12E8, (21)
OG, (22) NG, (23) TX100, (24) TX114.
to further processing with various detergents. Twenty six detergents,
each at final concentrations of 1–2% were screened, to
find the best detergent and detergent concentration required for
an efficient protein extraction. These detergents either belong to
families whose members have been successfully used to produce
crystals, such as maltosides and glucosides , or are commonly
used for protein purification in many laboratories, such as Triton
and CHAPS, in addition to some rather new and uncharacterized,
such as cymals. Using a 96-well plate, cell cultures in 1ml
aliquots were lysed and solubilized simultaneously. Solubilized
membrane proteins and the soluble content of the cells were
successfully separated from cell debris and inclusion bodies by
filtration, as previously described [8,10]. Using the small-scale
His-tag affinity purification, we were able to discard any AmtB
that was partially solubilized, i.e. containing membrane fragments
due to incomplete solubilization, or precipitated, and thus,
the yield of actual native and solubilized protein could be estimated
by dot blot analysis (Fig. 2). Interestingly, the yield of
purified AmtB changed with different detergents. In addition,
for some detergents, an increase in the concentration resulted
in more purified protein, whereas with other detergents such
increase would reduce the yield (data not shown). The latter
would suggest the destabilization effect of certain detergents on
specific membrane proteins, due to, e.g. excessive delipidation
The same platform was used to perform purification optimizations.
To this end, we investigated the various imidazole
concentrations during purification steps in combination with
four different types of metal affinity columns, to find the optimum
purification conditions prior to large-scale purifications
(Fig. 3). The dot blot analysis of the multi-parameter screen
indicated that presence of imidazole during binding improves
the yield of AmtB. Moreover, the TALONresin seemed to purify
AmtB more efficiently than others. Although a second elution
step showed the presence of more material in other columns
than TALON, we decided to use TALON resin for our largescale
purifications, since SDS-PAGE analysis of eluted material
showed purer material coming off TALON (data not shown).
The same procedures as small-scale protein purification were
used to produce milligrams of pure AmtB, suitable for crystallization
trials. However, ultracentrifugation steps were intro86
D. Niegowski et al. / International Journal of Biological Macromolecules 39 (2006) 83–87
Fig. 3. AmtB purification optimization. AmtB was extracted with FC12 (F),
DDM (D) and TX100 (T). Imidazole was used during binding at concentrations
0mM (1 and 2) and 20mM (3 and 4), and during washing at concentrations
20mM(1 and 3) and 40mM(2 and 4). Column volumes used were 25 and 50 l
Ni-NTA agarose in A and B, respectively, following in the same way by Ni-NTA
superflow (Qiagen) (C and D), Ni-sepharose fast flow (GE-Healthcare) (E and
F) and TALON (Clontech, Stockholm Sweden) (G and H), respectively.
duced to increase the purity and maintain the high protein yield.
One percent FC12 was used to initiate the purification, as this
detergent gave the highest purification yield of AmtB. Detergent
exchange was successfully performed during IMAC purifica-
Fig. 4. Gel filtration (A) and SDS-PAGE analysis (B) of AmtB after extraction
with FC12 and further purification with DDM. The peak marked with asterisk
was submitted to SDS-PAGE showing presence of pure AmtB. The other peaks
contained either impurities or no protein.
Fig. 5. AmtB crystals grown in the presence of DDM (A) and Cymal-6 (B).
tion followed by gel filtration to further purify AmtB to higher
homogeneity, and finally the purity of the sample was verified
by SDS-PAGE (Fig. 4). Although, some of the detergents, as
indicated in Fig. 2, showed low efficiency in extracting AmtB,
we decided to include these detergents in the large-scale production
processes to further explore their effects during purification
and subsequent crystallization. Thus, AmtB was extracted with
FC12 and subsequently purified with FC12, UDM, DDM, Cy6,
Cy7, OG and LDAO. AmtB crystals appeared in the first screen
after three days in the presence of DDM and Cymal-6 (Fig. 5).
The crystals could diffract poorly to ∼11A° .
AmtB could not be crystallized in the presence of other detergents,
except for UDM in which micro-crystals were grown.
To produce high quantities of pure native membrane proteins
has always been a major obstacle in crystal structure determination
of membrane proteins. This problem has so far been
connected to lowexpression of recombinant membrane proteins.
However, we have shown here that this problem may also be due
to poor understanding of purification procedures, namely the
usage of wrong detergents. If commonly used detergents such
as Triton X-100, octyl glucoside or CHAPS would have been
D. Niegowski et al. / International Journal of Biological Macromolecules 39 (2006) 83–87 87
used in the present study, the results would be no or very poor
yields of recombinantly produced AmtB, and a poor expression
was wrongly assumed. In addition, screening of many detergents
during crystallization has been shown to be highly important for
obtaining high-quality crystals necessary for solving the structure
of membrane proteins [7,14].
Due to the rather high cost of detergents, traditional methods
for screening various detergents and purification optimizations,
i.e. membrane preparation by ultracentrifugation and further
protein purification and analysis by gel filtration, SDS-PAGE
and Western blotting, is both expensive and time consuming.
We believe the use of present screening procedure, allowing
tens of various purification strategies and detergent screening
at microlitre volumes, and subsequent analysis by dot blot, will
minimize the costs, thus, enabling more thorough optimizations
prior to tedious large-scale membrane protein production. It is
important to remember that the small-scale purification platform
has an elevated degree of initial impurities due to the presence of
soluble as well as membrane proteins, as compared with purification
following membrane isolation. Therefore, any progress
with purification during this step may provide important information,
such as required detergent concentration or critical salt
concentration, for large-scale preparative steps, and thus be of
There is an interesting correlation between the yield of
extracted protein using a certain detergent and its ability to crystallize
in that detergent: UDM, DDM and Cy6 were efficient
in protein extraction and they facilitated crystal growth, while
neither OG nor LDAO could extract sufficient protein and no
crystals were grown in their presence. To suggest that detergents
highly efficient in extracting a certain protein would facilitate
crystallization of that protein is currently rather impossible and
many more studies are required to validate that. However, to
exclude detergents that are not useful for high-yield extraction
of a specific protein (e.g. due to destabilizing the protein) may be
very helpful to minimize the number of detergents during crystallization
screening. The fact that no crystals of AmtB were
grown in the presence of OG and LDAO may be such an indication.
The structure of AmtB was recently solved to high resolution
[15,16]. In those studies, AmtB crystals were grown in the presence
of OG and LDAO, respectively. Interestingly, none of these
detergents could help to produce AmtB crystals in our study.
Khademi et al. reported the extraction of AmtB with almost
10% OG (200 mM), with subsequent purification using 1% of
the same detergent. Our attempt to perform the same procedure
for isolating AmtB did not result in any crystals. Since it
is rather unclear exactly how Khademi et al. purified and crystallized
AmtB, the comparison of the procedures are difficult.
There is also no information from these reports whether the used
detergents in the crystallization were also the best detergents in
extracting AmtB. The differences between the constructs used
in all three studies may be the reason for the differences in crystallization
behaviour of AmtB, as our construct contains a FLAG
followed by an attB-site and a TEV-protease cleavage site at the
N-terminus, which, together with the nature of the used detergent,
may play important role for crystal contacts. The other
two constructs lack this long N-terminal tag. Khademi et al.
have crystallized a protein lacking the first transmembrane helix
and containing two mutations.
We have successfully applied this strategy to 10 other membrane
proteins, out of two have so far resulted in diffracting
We believe the presented platform is indeed an efficient tool
for benchmarking and optimizing important purification parameters,
e.g. the type and concentration of detergents, buffer composition
and column material, in a membrane protein production
and crystallization pipe-line. The choice of the right detergent
is indeed one of the important keys for membrane protein stability
and crystallization. However, due to the large number of
detergents available in the market, it is an elaborate and expensive
task to choose the right detergent(s) for purification and
crystallization of the target protein, by traditional methods, if
one determines to screen a large number of detergents. On the
other hand, screening fewer detergents may result in poor yield,
unstable protein and/or no protein crystals, and is not recommended.
Thus, the presented strategy allows the screening of
tens of detergents, for their efficiency of pure protein production
and crystallization, easily and simultaneously, producing
reliable and reproducible results, at very low cost.
The European Membrane Protein Consortium (E-MEP), the
Swedish Research Council and theG¨oran Gustafsson foundation
are acknowledged for financial support.
 S.H. White, Protein Sci. 13 (2004) 1948–1949.
 G. Muller, Curr. Med. Chem. 7 (2000) 861–888.
 K. Lundstrom, Comb. Chem. High Throughput Screen 7 (2004)
 S. Schuck, M. Honsho, K. Ekroos, A. Shevchenko, K. Simons, Proc.
Natl. Acad. Sci. U.S.A. 100 (2003) 5795–5800.
 V. Koronakis, A. Sharff, E. Koronakis, B. Luisi, C. Hughes, Nature 405
 M.J. Lemieux, J. Song, M.J. Kim, Y. Huang, A. Villa, M. Auer, X.-D.
Li, D.N. Wang, Protein Sci. 12 (2003) 2748–2756.
 S. Eshaghi, M. Hedr´en, M.I. Nasser, T. Hammarberg, A. Thornell, P.
Nordlund, Protein Sci. 14 (2005) 676–683.
 F. Reiss-Husson, D. Picot, in: A. Ducruix, R. Gieg´e (Eds.), Crystallization
of Nucleic Acids and Proteins, Oxford University Press, New York,
1999, pp. 245–268.
 R. Knaust, P. Nordlund, Anal. Biochem. 297 (2001) 79–85.
 M. Auer, M.J. Kim, M.J. Lemieux, A. Villa, J. Song, X.-D. Li, D.-N.
Wang, Biochemistry 40 (2001) 6628–6635.
 J.M. Boulter, D.N. Wang, Protein Expr. Purif. 22 (2001) 337–348.
 M.J. Lemieux, R. Reithmeier, D.N. Wang, J. Struct. Biol. 137 (2002)
 G. Chang, C.B. Roth, Science 293 (2001) 1793–1800.
 S. Khademi, J. O’Connell, J. Remis, Y. Robles-Colmenares, L.J. Miercke,
R.M. Stroud, Science 305 (2004) 1587–1594.
 L. Zheng, D. Kostrewa, S. Berneche, F.K. Winkler, X.-D. Li, Proc. Natl.
Acad. Sci. U.S.A. 101 (2004) 17090–17095.