Volume 58, number 1 FEBS LETTERS October 1975

ELECTROPHORETIC RESOLUTION OF THE ‘MAJOR OUTER MEMBRANE PROTEIN’

OF ESCHERICHIA COLZ K12 lNT0 FOUR BANDS

Ben LUGTENBERG, Jos MEIJERS, Roe1 PETERS, Peter van der HOEK
and Loek van ALPHEN
Department of Microbiology and Institute for Molecular Biology, State University, Transitorium 3, Padualaan 8, Utrecht,
The Netherlands

Received 26 June 1975

1. Introduction leading as it excludes the lipoprotein (mol. wt 7000)

described by Braun and Rehn [9], which certainly is
The cell envelope of Enterobacteriaceae consists of a major outer membrane protein too. The designation
two membranes separated by a peptidoglycan layer. ‘major outer membrane protein’ used in the present
Methods have been developed to separate the cyto- paper refers only to the 44 000 dalton protein des-
plasmic membrane from the outer membrane [ 1,2] . cribed by Schnaitman [ 1 ] . Other authors calculated
The outer membrane of Escherichia coli contains 60% mol. wts between 29 000 [6] and 44 000 [l] for the
of the envelope protein [l] . Using polyacrylamide same protein.
gelelectrophoresis, Schnaitman found six protein
bands in the outer membrane fraction, of which one
band (mol. wt 44 000) accounted for 70% of the total 2. Experimental
outer membrane protein [ 11. This protein was referred
to as the ‘major outer membrane protein’. Schnaitman 2.1. Strains and growth conditions
recently reported that this major outer membrane The following E. coli K12 strains were used: strain
protein of E. coli strain 0 111 -B4 could be resolved PC 0008 is HfrH, thi; strain PC 003 1 is HfrR4, gal,
into three distinct bands, designated as protein bands tonA; strain PC 0205 is F-, thr, leu, proA, purA,
1,2 and 3. Only bands 1 and 3 were found in E. coli lacy, gal, xyl, mtl, mal, tsx, strA; strain PC 0221 is
K12 [3]. Henning et al. [4] too found that the major F-, prototrophic; strain PC 0612 is HfrKL16, thi and
protein of outer membrane preparations of E. coli strain PC 1349 is F- , thr, leu, proA, argE, his, thi,
K12 could be resolved into two bands, for which they recB21, recC22, sbcB, lac Y, galK, xyl, mtl, ara, tsx,
reported mol. wts of approx. 40 000. Ames has phx, strA, sup-37 amber. The nomenclature is the one
applied the gelelectrophoresis system of Laemmli [S] used by Taylor and Trotter [lo] . Cells were grown
very successfully for the separation of the major under aeration at 37°C in yeast broth [ 111 .
envelope protein of Salmonella typhimurium. How-
ever, in cell envelopes of E. cofi K12 strain HfrH she 2.2. Membrane fractions
found only two major bands with mol. wts of 29 000 Cell envelopes were isolated at 0-4°C as follows:
and 33 000 [6]. Also the system described by Neville exponentially growing cells were harvested, washed
[7] resolves the major outer membrane protein of with 0.9% NaCl and resuspended in 50 mM Tris-HCl
E. cob K12 into two bands [8]. pH 8.5 containing 2 mM EDTA. After sonic disruption
In this paper we describe a new gelelectrophoresis for 4 times 1.5 set under cooling, unbroken cells and
system which results in a better resolution of the large fragments were removed by centrifugation for
major outer membrane protein of E. coli K12. The 20 min at 1200 g. The supernatant fluid was centrifug-
designation ‘major outer membrane protein’ is mis- ed for 60 min at 225 000 g. The pellet obtained, con-

254 North-Holland Publishing Company - Amsterdam

Volume 58. number 1 FEBS LETTERS October 1975

taining the envelopes, was resuspended and washed Tris, 0.19 M glycine and 0.1% SDS. The pH of this
once in the buffer described above and finally resus- solution was 8.3.
pended in 2 mM Tris-HCl, pH 7.8. Trypsin-treated Protein samples were prepared in 0.0625 M Tris-
cell envelopes were prepared by incubation of cell HCl, pH 6.8, containing SDS (2%) glycerol (10%)
envelopes (10 mg protein per ml) in 20 mM Tris-HCl, bromophenol blue (O.OOl%), 2-mercaptoethanol(5%)
pH 7.2, containing 100 pg trypsin per ml. After and protein (0.5-l .O mg/ml). The samples were boil-
incubation for 1.5 min. at 37°C the suspension was ed for 5 min. If necessary, the samples were stored
chilled at 0°C and centrifuged at 225 000 g. The at -20°C and boiled again before application. Sam-
resulting pellet was washed three times with Tris ples containing a mixture standard proteins of known
buffer and finally resuspended in 2 mM Tris-HCl, molecular weight were prepared as described previously
pH 7.8. Cytoplasmic and outer membranes were and contained 1 pg of each protein per 20 ~1 of
separated according to the procedure of Osborn et sample.
al. [2] with the modification that after the addition Routinely 20 ~1 of a sample were applied per slot.
of EDTA the suspension was incubated for 15 min Electrophoresis was carried out at room temperature
at 37°C to obtain good spheroplasts (> 98%). using a constant current of 25 or 30 mA per gel. The
electrophoresis was stopped when the tracking dye
2.3. Polyacrylamide gelelectrophoresis was about one cm from the bottom of the gel, which
We used the slab gel apparatus described by Studier happened after 3.5 or 4.5 h, depending on the current
[ 121. The slabs had a thickness of 1.5 mm. Routinely applied. Gels were stained overnight, under gentle
experiments were carried out with a running gel of shaking, in a solution of 0.1% Fast Green FCF in
9 cm length; in some experiments longer gels (18 cm) 50% methanol - 10% acetic acid. Gels were destained
were used. The gel was prepared following the ex- in 50% methanol - 10% acetic acid. Scanning of the
perimental details described by Ames [6], except for band pattern was performed at a wave length of 536
the composition of the gel, which is a combination of nm with a Vitatron TLD 100 densitometer at a rate
those described by Laemmli [S] and Neville [7] . For of 0.5 cm/min. In order to store the slabs, they were
the preparation of running and stacking gels the follow- dried after soaking for at least 24 h in a solution of
ing solutions were used: stock solution I contained 50% methanol - 5% glycerol. The gel was spread
44 g acrylamide plus 0.8 g methylene bisacrylamide on Whatman 3 MM chromatography paper and dried
while stock solution II contained 30 g plus 0.8 g for 1.5 h at about 50°C in a glazing dryer. In order
respectively of these components. The volumes of both to release the gel from the metal plate of the glazing
solutions were adjusted to 100 ml with deionized dryer, the plate was cooled for 1 h at 4°C.
water. The solutions were kept at 4°C in the dark and
could be used for at least 2 months. A fresh solution 2.4. Molecular weights
of ammonium persulphate (10 mg/ml) was prepared Protein bands are indicated by their molecular
just before the preparation of the gel. Running gel weights multiplied by 10e3 and followed by the letter
solution contained: stock solution I, 6.25 ml; ammo- K. The molecular weights of the standard proteins are
nium persulphate (10 mg/ml), 0.63 ml; 10% (w/v) SDS, indicated in the figures and refer to the following
0.50 ml; 0.75 M Tris-HCl, pH 8.8, 12.50 ml, and, proteins: bovine serum albumin (67 K), catalase
distilled water, 5.12 ml. Stacking gel solution con- (60 K), ovoalbumin (45 K), lactate dehydrogenase
tained: stock solution II, 0.50 ml; ammonium persul- (36 K), chymotrypsinogen A (25 K) and hen egg
phate (10 mg/ml), 0.12 ml; lO%(w/v) SDS, 0.050 lysozyme (14 K). All standard proteins gave rise to
ml; 0.25 M Tris-HCl, pH 6.8, 2.50 ml, and, distilled one band except catalase which gave two bands very
water, 1.83 ml. Before pouring and polymerizing the close to each other.
gel, the solutions were deaerated. Polymerization was
started by the addition of N,N,N’,AJ’-tetramethyl- 2.5. Chemicals
enediamine in a final concentration of 0.2% (v/v). The Acrylamide, methylenebisacrylamide and Triton-X
gel was stored in a humid atmosphere and used after 100 were obtained from Serva, Heidelberg, G.F.R.;
16-22 h. Both electrode buffer contained 0.025 M Fast Green FCF from Sigma Chemical Co., St. Louis,

255
Volume 58, number 1 FEBS LETTERS October 1975

MO., USA; sodium dodecylsulphate from BDH, Poole, major outer membrane protein bands were visible.
England; glycine and hen egg lysozyme from Fluka, This observation is in agreement with a previous
Buch, Switzerland. All other standard proteins of report in which envelope proteins of the same strain
known molecular weight, used as standard and tryp- were described [6]. With the system described by
sin were obtained from Boehrittger Mannheim, Am- Neville [7] three major outer membrane protein bands
sterdam, The Netherlands. were visible but the pattern was rather diffuse. The
systetn described in this paper, which can be consider-
ed as a combination of the two systems mentioned
3. Results and discussion above, allows the resolution of the major envelope
protein into four sharp bands designated a, b, c and
3.1 sImproved ~~s~ll~ti~~of the mujbr outer mem- 4 (fig.lA). Except for some outer l~lert~brane mutants,
brutze protein ihis result was found for all E. cclli K12 strains tested
When membrane proteins of E CO&K12 strain so far.
PC 0008 were separated with the system according to Comparison of samples of envelopes, cytoplasntic
Laemmli [ 5] the bands were very sharp but only two and outer membranes showed that the polypeptides

A B c 0

67K-
67K-
60K-

45K- -

45K-
- C 36K- amew
7d
36K-
25K-
25K- “Lf’

**p*- d’

12345 12 34 1 2

Fig.1. SIX-polyacrylamide geieiectrophoresis of membrane fractions of E. cdi K12. All samples contained 20 ~18of protein. The
molecular weights of the standard proteins are indicated. A, 1: molecular weight standards; 2 -5: cell envelope proteins of & coli
K12 strains PC 0221, PC 008, PC 0612 and PC 0031 respectively. B, 1: molecular weight standards; 2-4: outer membrane, cell
envelopes and cytoplasmic membrane respectively of strain PC 1349. C: Longer gel (I 8 cm) with a cell envelope preparation of
strain PC’ 1349. Pure SIX was used. I>: Effect of in~Libation of cell envelopes of strain PC 0205 with trypsin. I : cell envelopes;
2: cell envelopes after treatment with trypsin. The degradation product of d. band d’, is indicated.

256
Volume 58, number 1 FEBS LETTERS October 1975

67K $OK represented by these four bands, are all located in

the outer membrane (fig.1 B). Fig.2 shows scans of
A.Outer membrane
the protein patterns of the three membrane prepara-
tions of fig. 1B. From these scans it is clear that b, c
and d are major bands, while band a is rather weak.
The results described so far were obtained with SDS
(BDH, cat. no. 30175) that is contaminated with Cl0
and Cl4 derivatives. Recently it was observed that
with more purified SDS (Serva, cat. no. 20760 or
BDH cat. no. 30176) bands a, b and c are not separat-
ed from each other. However, when pure SDS was
67K 6OK 4yK 36K f5K l?K
used in longer gels, bands a, b and c were also separated
1 1

B.Cell envelope (fig.1 C), showing that they are not artificial bands
derived from one polypeptide. Consequently the new
system resolves four outer membrane protein bands
of E. coli K12 instead of the two bands described up
till now [3,4,8]. From gels ran with pure SDS the
following apparent molecular weights were calculated:
a:40K;b:38.5K;c:38K;andd:36K.
The method described for drying of the slabs is
extremely simple and can easily be used for at least
nine gels at a time. Drying did not significantly
influence the band pattern as was concluded from
autoradiograms of gels containing labelled proteins
(not shown).

3.2. Comparison with other gel systems

In order to compare the bands a, b, c and d with
bands of other gel systems we tested preparations of
a number of membrane fractions in various gel systems.
The preparations included purified proteins b and c
and cell envelopes of a strain that is deficient in poly-
Fig.2. Scans of the three membrane preparations from the gel
peptide d. From the patterns obtained (fig.3) it was
shown in fig.lB. The position of the molecular weights
possible to indicate the positions of bands b, c and d
standard proteins are indicated. Arrows at the left-hand side
indicate the top of the running gel, arrows at the right-hand in other gel systems, except in the one described by
side indicate the position of the tracking dye. It should be Henning et al. [4] as method II. The position of band
noted that the peak, corresponding with the fastest moving a could be established in the systems described by
band is partly artificial, due to a turbid spot in that position
(presumably lipid).

1 2 3 4 5

Fig.3. Schematic representation of patterns of the major outer membrane protein bands in various gel systems. The designation
used by the authors is given at the right hand side. The position of bands a, b, c and d is given at the left hand side of each pattern.
The gel systems are 1, the one described in this paper; 2, Laemmli [5] ; 3, Nevillr [7] ; 4, the Bragg-Hou system as used by
Schnaitman [ 13) ; 5, the system II described by Henning et al. [4] I:or further explanation see text.

257
Volume 58. number 1 FEBS LETTERS October 1975

Laemmli [S] and Neville [7] because band a was Acknowledgements

partly separated from band b. This was not the case
in the system used by Schnaitman [ 131. With method We thank Miss H. S. Felix of the Phabagen Collec-
II described by Henning et al. [4] we were unable to tion for her generous gift of strains. We are very grate-
resolve the major outer membrane protein into more ful to Drs J. Foulds, U. Henning, P. Reeves and J. P.
than one band. However, as both Henning’s protein Rosenbusch for their generous gifts of strains and for
II* and protein d are sensitive to trypsin (fig.lD), communicating results prior to publication.
heat-modifiable [ 141 and absent in tolG mutants
[ 151, they must be identical. Henning’s protein I
consists of two polypeptides (references 16 and 17 References
and U. Henning, personal communication), one of [l] Schnaitman, C. A. (1970) .I. Bacterial. 104, 890-901.
which is the matrix protein isolated by Rosenbusch [2] Osborn, M. J., Gander, J. E., Parisi, E. and Carson, J.
[ 181 from E. coli BE [ 171. We isolated this protein (1972) J. Biol. Chem. 247, 3962-3972.
from ,Y. coli BE and found that it moved to the posi- [3] Schnaitman, C. A. (1974) J. Bacterial. 118, 454-464.
[4] Henning, U., Hahn, B. and Sonntag, I. (1973) Eur. J.
tion of band b. We assume that the other polypeptide
Biochem. 39, 27-36.
in band I is protein c. The position of polypeptide a [5] Laemmli, U. K. (1970) Nature (London) 227, 680-
in Henning’s system remains unknown. 685.
Schnaitman [ 131 reported that band 3 consists of [6] Ames, G. 1;. (1973) J. Biol. Chem. 249, 634-644.
two different polypeptides 3a and 3b. In E. coli K12 [7] Neville, D. M. (1971) J. Biol. Chem. 246, 6328-6334.
[8] Skurray, R. A., Hancock, R. E. W. and Reeves, P. (1974)
strain P 400 the amount of protein 3b is small and
J. Bacterial. 119, 726-735.
its bacteriophage K3 resistant mutant P 460 is missing [9] Braun, V. and Rehn, K. (1969) Eur. J. Biochem. 10,
protein 3a, while protein 3b is present. (P. Reeves, 426-438.
personal communication). Protein d was found to be [lo] Taylor, A. L. and Trotter, C. D. (1972) Bacterial. Rev.
absent from gel profiles of mutant P460, using our 36,504-524.
[ll] Lugtenberg, E. J. J. and de Haan, P. G. (1971) Antonie
gel system. Bands d and II* are therefore most likely
van Leeuwenhoek J. Microbial. Serol. 37, 537-552.
identical to band 3a and might contain small amounts [12] Studier, F. W. (1973) J. Mol. Biol. 79, 237-248.
of protein 3b. [ 131 Schnaitman, C. A. (1974) J. Bacterial. 118, 442-453.
Protein IV described by Henning et al. [4,16] [14] Reithmeier, R. A. F. and Bragg, P. D., FEBS Lett.
corresponds with Braun’s lipoprotein [4]. This is the (1974) 41,195-198.
1151 Chai, T. and Foulds, J. (1974) J. Mol. Biol. 85, 465-
only envelope protein that can be labelled under con-
474.
ditions of starvation of a histidine auxotroph for [ 161 Garten, W. and Henning, U. (1974) Eur. J. Biochem.
histidine [ 191. It is also the only envelope protein 47, 343-352.
that is soluble in 10% trichloroacetic acid after boil- [ 171 Henning, U. and Haller, I., FEBS Lett. in press.
ing of envelopes with SDS [20]. We found that only [ 181 Rosenbusch, J. P. (1974) J. Biol. Chem. 249,
8019-8029.
the fastest moving protein in our system (fig.1) has
[19] Hirashima, A. and Inouye, M. (1973) Nature (London)
these properties (not shown). The amount of dye 242,405-407.
that is bound by this lipoprotein varies from gel to [20] Hirashima, A., Wu, H. C., Venkateswaran, P. S. and
gel. It gives a visible band in fig.lA, B and C. Inouye, M. (1973) J. Biol. Chem. 248,5654-5659.

258