JOURNAL OF BACTERIOLOGY, JUlY 1974, p. 250-257 Vol. 119, No.

1
Copyright © 1974 American Society for Microbiology Printed in U.S.A.

Isolation and Characterization of the Outer Membrane of

Neisseria gonorrhoeae
K. H. JOHNSTON AND E. C. GOTSCHLICH
The Rockefeller University, New York, New York 10021
Received for publication 15 March 1974

The cell envelope of Neisseria gonorrhoeae strain 2686, colonial type 4, was

Downloaded from http://jb.asm.org/ on November 25, 2018 by guest

isolated from spheroplasts formed by the action of ethylenediaminetetraacetic
acid and lysozyme. Isopycnic centrifugation of osmotically ruptured spheroplasts
resolved the cell envelope into two main membrane fractions. Chemical and
enzymatic analyses were used to characterize these isolated membranes.
Succinic dehydrogenase, reduced nicotinamide adenine dinucleotide oxidase,
and D-lactate dehydrogenase were localized in the membrane fraction of buoyant
density, p0 = 1.141 g/cm3. Lipopolysaccharide and over half of the cell envelope
protein were associated with the membrane that banded in sucrose at p0 = 1.219
g/cm3. These fractions were consequently designated cytoplasmic and outer or
L-membrane, respectively. Sodium dodecyl sulfate-polyacrylamide electropho-
resis of isolated membranes demonstrated the relative simplicity of the protein
spectrum of the outer membrane. The majority of the protein in this membrane
could be accounted for by proteins of molecular weights 34,500, 22,000, and
11,500. The protein of molecular weight 34,500 accounted for 66% of the total
protein of the L-membrane. Isoelectric precipitation at pH 4.6 with 10% acetic
acid selectively removed this protein from a 150 mM NaCl in 10 mM tris(hydrox-
ymethyl)aminomethane-hydrochloride, pH 7.4, extract of purified outer mem-
brane. At pH 4.0, the other proteins of the L-membrane were precipitated. It was
concluded that the membrane components of the cell envelope of N. gonorrhoeae
were similar to those of other gram-negative bacteria. The cell envelope fractions
described here, in particular the outer membrane, are sufficiently well
defined to provide a valuable tool for future biochemical and immunological
studies on N. gonorrhoeae.

The cell envelope of gram-negative bacteria is ration of relatively pure membrane fractions
an immunochemically complex structure com- would facilitate this approach.
posed of three morphologically distinct layers For this purpose, a technique for the separa-
(7, 20), namely, the cytoplasmic membrane, a tion of outer membrane from cytoplasmic mem-
rigid peptidoglycan layer, and a second mem- brane was necessary. Miura and Mizushima
branous structure, the outer or L-membrane. (17) first described the separation of the cell
The outer membrane contains substantial envelope of Escherichia coli into two fractions,
amounts of protein, phospholipid, and lipopoly- enriched respectively for cytoplasmic and outer
saccharide (LPS), the class of antigens referred membrane by a technique involving isopycnic
to as the 0-antigen. However, very little is sucrose density centrifugation of membranes
known about the location of protein antigens; obtained by spheroplast lysis. More recently,
therefore, the development of methods for the Schnaitman (27) has reported similar results
isolation of these membranes is one important following disruption of E. coli in a pressure cell.
step in the determination of the nature of the Similar procedures with modifications were
antigens present in these structures. found to be applicable to N. gonorrhoeae, per-
The outer surface of certain colonial types of mitting rapid reproducible separations of cell
Neisseria gonorrhoeae possesses pili which have envelope components.
proved to be antigenic in man and other ani-
mals (2, 3). These and other antigenic compo- MATERIALS AND METHODS
nents of the cell envelope are of interest as Bacteria and media. Strain 2686, colonial type 4,
agents of potential immunoprophylaxis. Prepa- an isolate of N. gonorrhoeae obtained from D. S.
250
VOL. 119, 1974 OUTER MEMBRANE OF N. GONORRHOEAE 251
Kellogg, Jr., Center for Disease Control, Atlanta, Ga., of 1 M sodium succinate (pH 7.5), 1.5 ml of 100 mM
was subcultured for 16 h at 35 C in a candle extinction KCN, 20 ml of 10 mM sodium ethylenediaminetetra-
jar on GC agar base supplemented with Isovitalex acetate (EDTA) (pH 7.5), and 85 ml of distilled
(Baltimore Biological Laboratories, Cockeysville, water. After incubation at 22 C for 20 min, 1.5 ml was
Md.). Approximately 7 x 108 bacteria were sus- transferred to a cuvette, and 20 uliters of cytochrome c
pended in 50 ml of modified Frantz' meningococcal (10% wt/vol) was added. To start the assay, 5 gliters
medium (10) prepared at 0.16% L-glutamic acid, of phenazine methosulfate (1% wt/vol) was added.
0.60% NaCl, 0.25% Na,HPO4, 0.125% NH4Cl, 0.009% D-Lactate dehydrogenase (EC 1.1.1.28) activity was
KCl, 0.002% L-cysteine hydrochloride, 2% anhydrous assayed as for succinic dehydrogenase, with 1 ml of 1
dextrose, 0.3% MgSO4 .7H2O, and 0.5% dialyzed M sodium D-lactate instead of succinate in the
yeast extract. When cultures were to be intrinsically reaction mixture.
labeled, approximately 4 gCi of each of the following Reduced nicotinamide adenine dinucleotide
isotopes was added to 50 ml of medium: uniformly (NADH) oxidase (EC 1.6.3.1.) activity was measured

Downloaded from http://jb.asm.org/ on November 25, 2018 by guest

labeled ["Cjleucine (460 mCi/mmol) and ["C]tyro- according to Osborn et al. (23). A sample volume of 50
sine (304 mCi/mmol), L-[4,5-_H]leucine (5 Ci/ ,gliters of membrane suspension containing 20 to 70 ,tg
mmol), and L-[3,5-3Hjtyrosine (48.2 Ci/mmol), all of protein was used. The rate of decrease in absorb-
obtained from New England Nuclear, Boston, Mass. ance at 340 nm was measured at 22 C.
Portions of radioactive samples were added to 5 ml of Glucose-6-phosphate dehydrogenase (EC 1.1.1.49)
Hydromix liquid scintillation fluid (Yorktown Re- was assayed as described by Noltmann et al. (21).
search, New Hyde Park, N.Y.), and radioactivity Isolation of cell envelope components. After 4.5 h
was determined in a refrigerated Packard Tri-Carb of growth, cultures yielded approximately 3 x 10"
liquid scintillation counter. Counting efficiencies were cells/ml. Cells were harvested by centrifugation at
81% for ["CC] and 45% for [3H]. 10,000 x g for 15 min at 4 C. The cell pellets were
Chemical procedures. Protein was determined by washed once with tris(hydroxymethyl)aminomethane
the method of Lowry et al. (15), using bovine plasma (Tris)-hydrochloride buffer, pH 7.8, centrifuged, and
albumin (Metrix Armour Pharmaceutical Co., Chi- then suspended in cold 0.6 M sucrose and 10 mM
cago, Ill.) as standard. Total hexose, pentose, and Tris-hydrochloride buffer, pH 7.8 (20 ml of buffer per
methylpentose was determined by the a-napthol 10'0 cells). Hen egg white lysozyme (EC 3.2.1.17)
reaction as described by Dische (8); uronic acids and (Worthington Biochemical Corp., Freehold, N.J.) dis-
amino sugars gave no color. Hexosamine content was solved in 100 mM Tris-hydrochloride buffer, pH 7.8,
established by the modified Elson-Morgan procedure was added to a final concentration of 160 U/ml (as
as described by Williams and Chase (34). defined by Worthington Biochemical Corp.). Within 1
Phospholipid analysis was carried out as described min, 100 mM EDTA in 10 mM Tris-hydrochloride
by Osborn et al. (23). buffer, pH 7.4, was added to a final concentration of 2
LPS content of membranes was determined by mM. The suspension was gently agitated at 30 C for
measuring the amount of 2-keto-3-deoxyoctonate re- 60 min; spheroplast formation was monitored by
leased from the membranes by hydrolysis with 0.02 N reduction in absorbance at 600 nm of 0.5 ml of
H2SO4 at 100 C for 20 min. Portions were removed suspension diluted 1: 10 with distilled water. Sphero-
and analyzed for 2-keto-3-deoxyoctonate by the thio- plasts were harvested by centrifugation at 25,000 x g
barbituric acid procedure of Weissbach and Hurwitz for 15 min at 4 C and subsequently suspended in 5 ml
(33) as modified by Osborn (22); the absorbance was of 25 mM MgCl2 in 10 mM Tris-hydrochloride buffer,
read at 548 nm. The amount of LPS in the original pH 7.0, per 20 ml of original spheroplast solution.
digest was calculated from a standard curve obtained Spheroplasts were lysed by diluting the suspension
with purified LPS extracted by the procedure of 20-fold in cold 5 mM MgCl2, pH 7.0, and gently
Osborn et al. (24). stirring for 10 min. The lysed preparation was treated
Nucleic acid was determined spectrophotometri- with bovine pancreatic deoxyribonuclease (EC
cally at 260 nm. 3.1.4.5.) and bovine pancreatic ribonuclease (EC
Proteins were iodinated with either [12'I1 or [131I1 by 2.7.7.16), both obtained from Worthington Biochemi-
the chloramine T procedure (12). Before labeling, cal Corp. To 100 ml of lysed suspension, 1,400 U of
sodium azide was removed by overnight dialysis. deoxyribonuclease and 1,500 U of ribonuclease were
Enzyme assays. All enzyme reactions were fol- added, and the suspension was incubated for 30 min
lowed in a Gilford spectrophotometer. Specific activ- at 37 C. Membranes were harvested by centrifugation
ity of enzymes was expressed as micromole of sub- at 75,000 x g for 25 min, suspended in 5 mM MgCl2 in
strate coverted per minute-' per milligram of pro- 10 mM Tris-hydrochloride buffer, pH 7.0, and pel-
tein. leted as above. The membranes were suspended in 25
Succinic dehydrogenase (EC 1.3.99.1) was mea- mM EDTA in 10 mM Tris-hydrochloride buffer, pH
sured by coupling the enzyme via phenazine metho- 7.0, and dialyzed for 18 h against the same buffer.
sulfate to the reduction of cytochrome c (1) and Sodium azide was added to a concentration of 0.2%
following the reaction spectrophotometrically at 550 (wt/vol) as a preservative, after it was ascertained
nm. The reaction mixture was that essentially de- that azide did not interfere with the enzymatic
scribed by MacGregor and Schnaitman (16). A sam- analyses.
ple volume of 100 ,liters (50 to 75 Ag of protein) was Suspensions of membranes (1.0 ml containing 15 to
added to 2.0 ml of the following reaction mixture: 100 20 mg of protein) were layered over 57% (wt/wt)
ml of 100 mM sodium phosphate buffer (pH 7.5), 5 ml sucrose dissolved in 5 mM EDTA in 10 mM Tris-
252 JOHNSTON AND GOTSCHLICH J. BACTERIOL.

hydrochloride buffer, pH 7.0, and centrifuged in a water. Quantities of lysozyme greater than 200
SW-50.1 rotor (Beckman Instruments, Inc., Fullerton, U/ml resulted in macroscopic agglutination of
Calif.) for 7 h at 200,000 x g. The interface was the cells and spheroplasts of poor quality. When
retained and diluted 10-fold with 5 mM EDTA in 10 amounts of lysozyme less than 100 U/ml were
mM Tris-hydrochloride buffer, pH 7.0; membranes used, only 10 to 15% of the cells were converted
were pelleted by centrifugation at 200,000 x g for 2 h.
Pellets were resuspended in minimal volume of 0.8 M to spheroplasts. At the cell density employed, it
sucrose and 5 mM EDTA in 10 mM Tris-hydrochlo- was observed that 160 U of lysozyme/ml were
ride buffer, pH 7.0, to a protein concentration of 10 sufficient to convert 90 to 95% of the cells to
mg/ml; 500 Mliters of sample were applied to a spheroplasts. Addition of EDTA at concentra-
discontinuous gradient prepared by layering sequen- tions in excess of 5 mM resulted in extensive
tially 0.5-ml samples of sucrose solutions ranging in lysis of osmotically balanced cells. Absence of
concentration from 60 to 20% (wt/wt). Samples were EDTA in the spheroplasting medium decreased

Downloaded from http://jb.asm.org/ on November 25, 2018 by guest

spun at 250,000 x g for 18 h to isopycnic conditions. the amount of cells converted to spheroplasts by
Tube contents were fractionated by puncturing the approximately 65%. The optimal EDTA con-
bottom of the centrifuge tube and collecting drops; to
retrieve banded material, fractions were diluted 1:10 centration was between 1 to 3 mM; addition of
with 10 mM Tris-hydrochloride buffer, pH 7.0, and EDTA solution of pH >8.5 resulted in immedi-
centrifuged at 250,000 x g for 90 min. ate lysis of the culture. In summary, the optimal
Extraction of outer membranes. Isolated outer conditions for spheroplast formation of N.
membranes were extracted with 150 mM NaCl and gonorrhoeae entailed incubation of cell suspen-
0.02% (wt/vol) NaN3 in 10 mM Tris-hydrochloride sions 5 x 108 organisms/ml in 0.6 M sucrose and
buffer, pH 7.4 (TSE buffer). Suspensions of mem- 2 mM EDTA in 10 mM Tris-hydrochloride
branes containing 15 mg of protein per ml were buffer, pH 7.8, with 160 U of lysozyme/ml.
incubated with an equal volume of TSE buffer for 2 h The membranes derived from lysed sphero-
at 37 C. After centrifugation at 50,000 x g for 30 min,
the supernatant fluid was decanted and concentrated plasts used in these studies were free from
on a Diaflo PM-10 membrane (exclusion limit, cytoplasmic contamination as judged by the
> 10,000; Amicon, Lexington, Mass.). Two milliliters absence of the soluble cytoplasmic enzyme,
of concentrate was applied to a 1.5 by 30 cm Sepha- glucose-6-phosphate dehydrogenase. Approxi-
rose 6B (exclusion limit, > 4 x 106 daltons for globular mately 90% of the glucose-6-phosphate dehy-
proteins; Pharmacia, Piscataway, N.J.) column equil- drogenase of intact cells was recovered from the
ibrated in TSE buffer. Fractions eluting at the void supematant of lysed spheroplasts. After centri-
volume were pooled and concentrated. fugation of the initial envelope preparation on a
Polyacrylamide electrophoresis. Membranes and cushion of 57% (wt/wt) sucrose, the buoyant
extracts of membranes were analyzed by sodium
dodecyl sulfate (SDS)-polyacrylamide disc gel elec- material contained no glucose-6-phosphate de-
trophoresis by the procedure described by Weber and hydrogenase activity within the 2% detection
Osborn (32). A 1% SDS, 10% acrylamide, 0.27% limit of the assay. The material buoyant at 57%
methylene bisacrylamide gel was used throughout. (wt/wt) sucrose was unfractionated cell enve-
Gels were prepared in glass tubes 6 mm by 150 mm. lope; analyses of this preparation are presented
Samples to be electrophoresed were treated at 100 C in Table 1.
for 2 min in a solution of 10 mM sodium phosphate, Isopycnic centrifugation of total membrane
pH 7.0, 2% (wt/vol) SDS (Biorad Laboratories, Rich- fractions resolved the mixture as three visible
mond, Calif.), and 2% (wt/vol) 2-mercaptoethanol; bands of turbidity: an upper band (A) of
after cooling, one-half volume of 0.05% (wt/vol) bro-
mophenol blue in 80% (wt/vol) sucrose was added. buoyant density, p0 = 1.141 g/cm3, an interme-
After electrophoresis at 5 mA per tube, the gels were diate band (B) of buoyant density, p0 = 1.176
sliced in 1-mm sections by a Gilson Autogel slicer g/cm3, and a pronounced bottom band (C)
(Gilson Medical Electronics, Inc., Middleton, Wisc.) which had a buoyant density of p0 = 1.219
and monitored for radioactivity. g/cm3 (Fig. 1). Band A accounted for 18 to 25%
of the total protein of the cell envelope; band C
RESULTS contained as much as 58% of the envelope
Spheroplasts of N. gonorrhoeae were formed protein (Table 1). Band B (p0 = 1.176 g/Cm3)
by the action of lysozyme on the peptidoglycan consisted of unseparated material usually con-
layer while cells were in the presence of EDTA. taining less than 19% of the total protein. The
It was found that the ratio of lysozyme to cells amount of this band was determined by several
was critical for optimal spheroplast formation. factors, in particular, the ionic strength of the
"Optimal spheroplast formation" was defined spheroplasting medium. When optimal lyso-
as the conditions existing when 90 to 95% of zyme and EDTA concentrations were used, but
osmotically balanced spheroplasts would lyse NaCl was incorporated into the spheroplasting
within 15 s upon 1:10 dilution with distilled medium, wash buffers, and sucrose gradient
VOL. 119, 1974 OUTER MEMBRANE OF N. GONORRHOEAE 253
TABLE 1. Chemical and enzymatic analysesa of fractionated membranes of Neisseria gonorrhoeae strain 2686,
colonial type 4
Membrane fraction from sucrose gradient
Analysis Unfractionatedb Band A Band B Band C
cell envelope p0 = 1.141 g/cm' p0 = 1.176 g/cm' p0 - 1.219 g/cm'
Protein (%) 100 18.7-25.6 10.4-18.8 48.6-58.2
Phospholipid (mg/mg of protein) 0.41-0.47 0.55-0.62 0.34-0.39 0.29-0.31
Lipopolysaccharide (mg/mg of protein) 0.96-1.16 0.04-0.06 n.d.c 1.19-1.21
Carbohydrate (mg/mg of protein) 0.71-0.78 0.12-0.18 0.42-0.49 0.87-0.92
Hexosamine (mg/mg of protein) ND 0.01-0.02 <0.01 0.42-0.47

Downloaded from http://jb.asm.org/ on November 25, 2018 by guest

Specific activityd of:
Succinic dehydrogenase 0.54-0.68 5.23-7.41 1.94-2.86 0.05-0.08
n-Lacate dehydrogenase 0.36-0.59 4.99-5.62 3.04-3.89 0.02-0.12
NADH oxidase 0.61-0.74 4.29-4.93 1.08-1.26 0.08-0.15
'Figures in table represent range of values determined on a number of different membrane preparations.
b Cell envelope buoyant on 57% (wt/wt) sucrose.
ND, Not done.
d Specific activity expressed as micromole of substrate converted per minute per milligram of protein.

C B A Mg2+ was incorporated into the sucrose gradient

solution, most material banded in the region of
density, 1.18 to 1.19 g/cm3, and the recovery of
.22 bands A and C was minimal. The Mg2+ effect
could be reversed by the presence of 5 mM
I EE EDTA in the gradient medium.
Identification of the membranes resolved by
Io
E 1.14 cyl isopycnic centrifugation was achieved by chem-
0 I.0 1.10 .'
*i)~ ical and enzymatic analyses (Table 1). The
CM
c components of the electron transport system
00 .06 D were specifically localized in the A band, p0 =
0.8
C\ 1.141 g/cm8, with some activities detected in the
0.6 B band, p0 = 1.176 g/cm3. The specific activi-
c ties of succinic dehydrogenase and D-lactate
.0 0.4 8 dehydrogenase were found to have increased
0 10-fold over the unfractionated cell envelope;
.n
D0 0.2 similarly, the A band was enriched eightfold in
NADH oxidase activity. The most dense band,
BOTTOM 5 10 15 20 25 TOI
p° = 1.219 g/cm', was practically free of enzyme
activity. The specific activities of succinic dehy-
Fraction number drogenase, D-lactate dehydrogenase, and
NADH oxidase were 1.0%, 1.3%, and 2.5%,
FIG. 1. Isopycnic sucrose centrifugatior
ically lysed lysozyme-induced spheroplasts of N. gon- respectively, of the activities present in the A
orrhoeae, colonial type 4, strain 2686. Ce 41 envelope band. Repeated sucrose gradient centrifugation
preparations were layered on a discontinu ous sucrose did not reduce this contamination of outer
gradient, 60 to 20%o (wt/wt), and centrifugled for 18 h membrane with cytoplasmic membrane.
at 200,000 g. Upper case letters, A, B, an d C, denote
x Carbohydrate, LPS and hexosamine were
membrane fractions described in Table 1. primarily located in the heavy band, p0 = 1.219
g/cm3.
solutions, the increase in material bantding at p0 Both membranes dissolve completely in 1%
= 1.176 g/cm3 was striking. In concenttrations of SDS (wt/vol). To distinguish the two mem-
NaCl in excess of 0.1 M, 35% of total rn adioactiv- branes from each other a double label experi-
ity of intrinsically labeled cell enveloj pe banded ment was performed. Two cultures of 2686,
at this density. Similar results were obtained type 4, were grown in Frantz' medium (10).
with KCl. Optimal resolution of the cell enve- One culture was grown in the presence of
lope into bands A and C dependeA on the [3H]leucine and [3H]tyrosine enriched medium;
absence of divalent cations as well. When 5 mM the other was supplemented with [14C]leucine
254 JOHNSTON AND GOTSCHLICH J. BACTERIOL.

and ["C ]tyrosine. The two cultures were har- TABLE 2. Relative amounts of outer membrane
vested, spheroplasts were produced and osmot- proteinsa
ically lysed, and the envelope fractions were Percent of total
obtained as described above. There was about Peak no. outer membrane
14% incorporation of total radioactivity into the protein
cell envelope. Intrinsically labeled [4C ]-outer
membrane was mixed with [3H]-labeled cyto- 10 10.1
plasmic membrane and electrophoresed. The 12 66.2
14 11.7
disintegrations per minute were normalized 17 7.6
by conversion to percentage of total disintegra-
tions per minute, enabling the amount of pro- aCalculated from data in Fig. 2. The ["C ] content

Downloaded from http://jb.asm.org/ on November 25, 2018 by guest

tein labeled with each isotope to be compared in each peak was summed and divided by the total
directly. Of 22 proteins resolved on the gel, 13 ["C I content in the gel.
were derived principally from the cytoplasmic
membrane, and five from the outer membrane
indicated in Fig. 2 by circled notation above the outer membrane accounted for a much lower
peaks. One protein which was clearly a compo- percentage of the total cell envelope protein. All
nent of the outer membrane was number 12. It the proteins resolved on gels had molecular
was found to have a molecular weight of 34,500 weights of less than 110,000, and the presence or
by comparison with the migration of proteins of absence of 2-mercaptoethanol had no effect on
known molecular weight. This protein was the the electrophoretic profile of the proteins. In
major protein of the cell envelope (Table 2). It most experiments, the amount of protein re-
accounted for 66% of the labeled material of the maining on top of the gel was from 4 to 7% of the
outer membrane. In another experiment, this total protein (radioactivity) applied to the gel.
protein accounted for 48% of the total labeled Several methods for extraction of proteins
cell envelope protein. It appeared as a single from the outer membrane utilizing detergents
symmetrical peak in SDS-polyacrylamide gel (sodium deoxycholate, Triton X-100) were em-
electrophoresis even when the dye marker had ployed. However, it was found that a modifica-
migrated 140 mm. The other proteins of the tion of the simple method described by Frasch
and Chapman (11) was satisfactory. Outer
membranes were incubated in 150 mM NaCl in
10 10 mM Tris-hydrochloride buffer, pH 7.4, for 2
h at 37 C. When the extraction was carried out
12 on intrinsically labeled outer membranes, ap-
proximately 46% of the radioactivity was re-
10 leased into the supematant. All the radioactiv-
z
8
ity was eluted in the void volume of a Sepharose
C)
0
7D
6B column equilibrated with the extraction
2 6i 0D buffer. Addition of acetic acid to pH 4.0 precipi-
C) tated 86% of the radioactivity in the extract.
2 This precipitate could only be redissolved at a
,Z5 pH >8.0; subsequently, acid precipitates were
22
redissolved in 150 mM NaCl in 10 mM Tris-
Top 10 20 30 40 50 60 70
X.,~
80 Bottom
hydrochloride buffer, pH 8.4. Addition of 2 vol
of cold, 90% (vol/vol) acetone precipitated 95%
Migration (mm) of the radioactivity. After repeated acetone
FIG. 2. SDS-polyacrylamide electrophoresis of in- precipitations, the dry material was found to
trinsically labeled membrane proteins of N. gonor- contain 88.7% protein, 3.1% carbohydrate, and
rhoeae, colonial type 4, strain 2686. Membrane frac- 1.7% hexosamine, a composition distinctly dif-
tions isolated after osmotically lysed spheroplasts ferent from that of the outer membrane.
were centrifuged to isopycnic conditions. Bacteria A gel electrophoretic pattern of a [125I] extrin-
were grown in medium supplemented with either sically labeled Tris-saline extract was identical
[3H]-labeled tyrosine and leucine or ['IC]-labeled to the electrophoretic profile of a Tris-saline
tyrosine and leucine. Tritium-labeled cytoplasmic extract of intrinsically labeled outer membrane,
membrane (- ) was mixed with ['IC]-labeled outer
membrane ( ) and resolved by disc gel electropho-
indicating that the radioiodination procedure
resis. Circled notation denotes proteins of the outer did not introduce artifacts. Addition of 10%
membrane. Distribution of radioactivity is expressed (vol/vol) aqueous acetic acid to the iodinated
as a percentage of total radioactivity. extract caused a pronounced precipitation at
VOL. 119, 1974 OUTER MEMBRANE OF N. GONORRHOEAE 255
25 34,500 doltons
pH 4.6; 63% of the total radioactiivity was
precipitated. When acid was added to bring the
supernatant to pH 4.0, a second p]recipitate 20

formed, accounting for 31% of the totalI radioac- 0

tivity. The sum of both precipitations3, namely
94%, approximated the total radioac tivity re- CL
covered (96%) by a one-step precipitat ion at pH 100 22,000 doltons
4.0. When the acid precipitates were analyzed 5
-\ l1,500 doltons
electrophoretically, the specificity of the reac- 5-
tion was notable. The precipitate ob ltained at
pH 4.6 consisted primarily (72.7%) of Ithe major
outer membrane protein (Fig. 3). The superna- Origin 10 20 30 40 50 60 70 80 90

Downloaded from http://jb.asm.org/ on November 25, 2018 by guest

tant fluid at pH 4.6 was enriched ini proteins | Dye front
Migrotion (mm)
having molecular weights of 22,000 a]nd 11,500 FIG. 4. SDS-polyacrylamide electrophoresis of
(Fig. 4). When acetic acid was addedi to bring acetic acid precipitate of 10 mM Tris-saline extracts
the extract to pH 4.0, the precipitate Eso formed of isolated outer membranes. Extracts were extrinsi-
contained the two minor proteins, naimely the cally labeled with either [125I or ["I]. [113 I-labeled
22,000- and 11,500-dalton species pluIs a small protein remaining in the supernatant after a pH 4.6
amount of 34,500-dalton material. acid precipitation (-----) was mixed with [125I]-labeled
protein acid precipitated at pH 4.6 ( ); the mix-
DISCUSSION ture was resolved by disc gel electrophoresis.
The procedure described above offerrs a repro- In the preparation of the spheroplasts, the
ducible and useful method for the fra(,tionation concentration of EDTA was found to be a
of the cell envelope of N. gonorrh oeae into critical variable; EDTA in excess of 5 mM
cytoplasmic and outer membrane. T]he results caused extensive lysis of osmotically balanced
parallel the earlier observations of I liura and cells. This lethal action of EDTA has been
Mizushima (17), Schnaitman (27), anid Osborn attributed to the chelation of divalent metal
et al. (20), and, more recently, those (of Deneke
and Colwell (5) and Stinnett et al. (330) on the ions which are required for the structural integ-
separation of the membrane of other ggrami-neg- rity of the cell envelope (14). Not all gram-nega-
ative bacteria. The chemical and esnzymatic tive bacteria are so readily lysed, but a similar
composition of the isolated membreanes were phenomenon has been observed with Pseudo-
monas aeruginosa (9). For optimal resolution of
similar to those of other gram-negativ e bacteria gonococcal membranes by equilibrium sucrose
(18, 23, 26, 27). density centrifugation, it was imperative to
34,500 doltons
have EDTA present in the gradient medium.
When divalent cations were present, most ma-
terial remained unresolved. High concentra-
tions of Na+ and K+ appeared to have the same
effect. A similar observation on the influence of
Mg2+ on the separation of meningococcal cell
OL envelope membranes has been made by Hill et
Eu 'tons al. (13).
The buoyant densities of the isolated mem-
branes were akin to values found by other
11,500 daltons
t investigators for either E. coli (17, 27) or Salmo-
nella typhimurium (23). The distribution of
components of the electron transport system
Origin
10 20 30 40 50 60 D 80 90
7C (succinic dehydrogenase, NADH oxidase, and
Dye front D-lactate dehydrogenase) indicated that the
Migration (mm) preparations of outer membrane were virtually
FIG. 3. SDS-polyacrylamide electrop ,horesissof free of cytoplasmic membrane contamination.
acetic acid precipitate at pH 4.6 of 10 mM Tris-saline SDS-polyacrylamide electrophoresis of the
extract of isolated outer membrane. Exitracts were. outer membrane indicated a relatively simple
extrinsically labeled with either [12jI] or [13111 Un-
fractionated extract labeled with [12I] (t)Jjwas but unique protein spectrum. Most interesting
mixed with ["3I1]-labeled protein acid prei cipitated at was the presence of a single peak accounting for
pH 4.6 (-----). Samples were mixed priorr to electro- over 60% of outer membrane protein. This
phoretic resolution. protein either represents a single molecular
256 JOHNSTON AND GOTSCHLICH J . B ACTERIOL.

species or a mixture of proteins so similar in gonococcal infection. III. Correlation of gonococcal

subunit molecular size that the separation tech- colony morphology with infectivity for the chick em-
bryo. J. Exp. Med. 137:196-200.
niques used here cannot resolve them. However, 3. Buchanan, T. M., J. Swanson, K. K. Holmes, S. J.
others (19, 28, 29, 35), using somewhat different Kraus, and E. C. Gotschlich. 1973. Quantitative deter-
fractionation and electrophoretic techniques, mination of antibody to gonococcal pili. Changes in
have demonstrated that the major polypeptide antibody levels with gonococcal infection. J. Clin.
Invest. 52:2896-2909.
of the outer membrane of E. coli consists of a 4. Cox, S. T., Jr., and R. G. Eagon. 1968. Action of
complex association of several distinct proteins. ethylenediamine-tetraacetic acid, tris(hydroxyme-
Electrophoretic resolution of the outer mem- thyl)aminomethane and lysozyme on the cell walls of
Pseudomonas aeruginosa. Can. J. Microbiol.
brane did not demonstrate the presence of pili 14:913-922.
which have a subunit molecular weight of 5. Deneke, C. F., and R. R. Colwell. 1973. Lipopolysaccha-
24,000 (3); this is to be expected as the colonial ride and proteins of the cell envelope of Vibrio marinus,

Downloaded from http://jb.asm.org/ on November 25, 2018 by guest

type 4 gonococcus does not carry pili (31). As it a marine bacterium. Can. J. Microbiol. 19:1211-1217.
6. Depamphlis, M. L. 1971. Dissociation and reassembly of
has been demonstrated that EDTA can release Escherichia coli outer membrane and of lipopolysac-
LPS (4, 14) and protein as an LPS-protein charide, and their reassembly into flagellar basal
complex (6, 25) from the outer membrane of bodies. J. Bacteriol. 105:1184-1199.
gram-negative bacteria, it should be noted that 7. DePetris, S. 1967. Ultrastructure of the cell wall of
Escherichia coli and chemical nature of its constituent
the membrane protein spectra described here layers. J. Ultrastruct. Res. 19:45-83.
represents those proteins not solubilized by 8. Dische, Z. 1955. New color reactions for determination
EDTA. of sugars in polysaccharides, p. 313-358. In D. Glick
To prepare materials suitable for immunolog- (ed.), Methods of biochemical analysis, Interscience,
ical analyses, the purified outer membrane was New York.
9. Eagon, R. G., and K. J. Carson. 1965. Lysis of cell walls
extracted with Tris-saline. This extract con- and intact cells of Pseudomonas aeruginosa by
tained three proteins of the outer membrane, ethylenediaminetetraacetic acid and lysozyme. Can. J.
namely the 34,500-, 22,000-, and 11,500-dalton Microbiol. 11:193-201.
species. These proteins still existed in an aggre- 10. Frantz, I. D. 1942. Growth requirements of the meningo-
coccus. J. Bacteriol. 43:757-761.
gated form; they eluted in the void volume of a 11. Frasch, C. E., and S. S. Chapman. 1972. Classification of
Sepharose 6B column. Upon acidification to pH Neisseria meningitidis group B into distinct serotypes.
4.6, a precipitate formed which was markedly IV. Preliminary chemical studies on the nature of the
serotype antigen. Infect. Immunity 6:674-681.
enriched with the 34,500-dalton protein species. 12. Greenwood, F. C., W. M. Hunter, and J. S. Glover. 1963.
The mechanism of this purification is not clear. The preparation of ''I-labelled human growth hor-
There may be two species of aggregates, one mone of high specific radioactivity. Biochem. J.
containing primarily the protein of 34,500 dal- 89:114-123.
13. Hill, J. C., N. R. Peterson, and E. Weiss. 1972. Character-
tons, the other the proteins of 22,000 and 11,500 ization of spheroplast membranes of Neisseria
daltons. Altematively, raising the hydrogen ion meningitidis group B. Infect. Immunity 5:612-621.
concentration may cause dissociation of aggre- 14. Leive, L. 1965. Release of lipopolysaccharide by EDTA
gates containing all three protein species, and, treatment of E. coli. Biochem. Biophys. Res. Commun.
as the pH is lowered further, isoelectric precipi-
21:290-296.
15. Lowry, 0. H., N. J. Rosebrough, A. L. Farr, and R. J.
tation of the disaggregated proteins results. Randall. 1951. Protein measurement with the Folin
On the basis of quantity, location, and its phenol reagent. J. Biol. Chem. 193:265-275.
apparent electrophoretic homogeneity, the 16. MacGregor, C. H., and C. A. Schnaitman. 1971. Altera-
34,500-dalton material most likely has immuno- tions in the cytoplasmic membrane proteins of various
chlorate-resistant mutants of Escherichia coli. J. Bac-
logical significance. The procedures described teriol. 108:564-570.
above were performed to isolate this protein for 17. Miura, T., and S. Mizushima. 1968. Separation by
immunological investigations on the serology of density gradient centrifugation of two types of mem-
N. gonorrhoeae. branes from spheroplast membrane of Escherichia coli
K12. Biochim. Biophys. Acta 150:159-161.
18. Miura, T., and S. Mizushima. 1969. Separation and
ACKNOWLEDGMENTS properties of outer and cytoplasmic membranes in
This investigation was funded by Public Health Service
Escherichia coli. Biochim. Biophys. Acta 193:268-276.
19. Moldow, C., J. Robertson, and L. Rothfield. 1972. Purifi-
grant no. AI-10615 and from the National Institute of Allergy cation of bacterial membrane proteins. The use of
and Infectious Diseases grant no. U-2185 from the Health guanidinium thiocyanate and urea. J. Membrane Biol.
Research Council of New York. 10:137-152.
The authors express appreciation to Leon Parkes for able 20. Murray, R. G. E., P. Steed, and H. E. Elson. 1965. The
technical assistance. location of the mucopeptide in sections of cell wall of
LITERATURE CITED Escherichia coli and other gram-negative bacteria.
Can. J. Microbiol. 11:547-560.
1. Arrigoni, O., and T. P. Singer. 1962. Limitations of the 21. Noltmann, E. A., C. J. Gubler, and S. A. Kuby. 1961.
phenazine methosulfate assay for succinic and related Glucose 6-phosphate dehydrogenase. I. Isolation of the
dehydrogenases. Nature (London) 193:1256-1258. crystalline enzyme from yeast. J. Biol. Chem.
2. Buchanan, T. M., and E. C. Gotschlich. 1973. Studies on 236:1125-1230.
VOL. 119, 1974 OUTER MEMBRANE OF N. GONORRHOEAE 257
22. Osborn, M. J. 1963. Studies on the gram-negative cell Escherichia coli. II. Heterogeneity of major outer
wall. I. Evidence for the role of 2-keto-3-deoxyoctonate membrane polypeptides. Arch. Biochem. Biophys.
in the lipopolysaccharide of Salmonella typhimurium. 157:553-560.
Proc. Nat. Acad. Sci. U.S.A. 50:499-506. 30. Stinnett, J. D., H. E. Gilleland, and R. G. Eagon. 1973.
23. Osborn, M., J. E. Gander, E. Parisi, and J. Carson. 1972. Proteins released from the cell envelopes of Pseudomo-
Mechanism of assembly of the outer membrane of nas aeruginosa on exposure to ethylenediaminetetra-
Salmonella typhimurium. J. Biol. Chem. acetate: comparison with dimethylformamide-
247:3962-3972. extractable proteins. J. Bacteriol. 114:399-407.
24. Osborn, M. J., S. M. Rosen, L. Rothfield, and B. L. 31. Swanson, J., S. J. Kraus, and E. C. Gotschlich. 1971.
Horecker. 1962. Biosynthesis of bacterial lipopolysac- Studies on gonococcus infection. I. Pili and zones of
charide. I. Enzymatic incorporation of galactose in a adhesion: their relation to gonococcal growth patterns.
mutant. Proc. Nat. Acad. Sci. U.S.A. 48:1831-1838. J. Exp. Med. 134:886-906.
25. Rogers, S. W., H. E. Gilleland, Jr., and R. G. Eagon. 32. Weber, K., and M. Osborn. 1969. The reliability of
1969. Characterization of a protein-lipopolysaccharide molecular weight determinations by dodecyl sulfate-
complex released from the cell walls of Pseudomonas polyacrylamide gel electrophoresis. J. Biol. Chem.

Downloaded from http://jb.asm.org/ on November 25, 2018 by guest

aeruginosa by ethylenediaminetetraacetic acid. Can. J. 244:4406-4412.
Microbiol. 15:743-748. 33. Weissbach, A., and J. Hurwitz. 1959. The formation of
26. Schnaitman, C. A. 1970. Examination of the protein 2-keto-3-deoxyheptonic acid in extracts of Escherichia
composition of the cell envelope of Escherichia coli by coli B. J. Biol. Chem. 234:705-709.
polyacrylamide gel electrophoresis. J. Bacteriol. 34. Williams, C. A., and M. W. Chase. 1968. Elson-Morgan
104:882-889. method (modified), p. 296-298. In Curtis A. Williams
27. Schnaitman, C. A. 1970. Protein composition of the cell and Merrill W. Chase (ed.), Methods in immunology
wall and cytoplasmic membrane of Escherichia coli. J. and immunochemistry, vol. II. Academic Press Inc.,
Bacteriol. 104:890-901. New York.
28. Schnaitman, C. A. 1973. Outer membrane proteins of 35. Wu, H. D. 1972. Isolation and characterization of an
Escherichia coli. I. Effect of preparative conditions on Escherichia coli mutant with alteration in the outer
the migration of protein on polyacrylamide gels. Arch. membrane proteins of the cell envelope. Biochim.
Biochem. Biophys. 157:541-542. Biophys. Acta 290:274-289.
29. Schnaitman, C. A. 1973. Outer membrane proteins of