J. Cell Sci. 3, 273-294 (1968)
Printed in Great Britain
273
FINE STRUCTURE AND RADIATION
RESISTANCE IN ACINETOBACTER: STUDIES
ON A RESISTANT STRAIN
MARGARET J. THORNLEY AND AUDREY M. GLAUERT
Sub-department of Chemical Microbiology, Department of Biochemistry,
University of Cambridge, and Strangeways Research Laboratory, Cambridge
SUMMARY
An electron-microscope study of thin sections and negatively stained preparations of intact
cells and isolated cell walls of a bacterium which is moderately resistant to ionizing radiation,
Acinetobacter strain 199 A, showed that it is similar to other Gram-negative bacteria except for
its mode of division and for the fine structure of some of the surface layers. During division the
cells form a fairly thick septum similar to those observed in Gram-positive bacteria. An examination of the appearance and chemical composition of isolated cell walls before and after treatment
with enzymes, detergents and lipid solvents revealed that three layers, each with a characteristic
fine structure, are present in the cell wall: (1) an outer membrane with an array of peg-like
subunits; (2) a layer of wrinkled material which is digested by proteolytic enzymes; and (3) a
smooth, rigid layer, which contains the mucopeptide components of the cell wall. These observations are compared with the results of other workers for various Gram-negative bacteria.
From comparisons with the structure of more radiation-sensitive strains of Acinetobacter, it
appears that layer (2) may be associated with the radiation resistance of the organism.
INTRODUCTION
Acinetobacter strain 199 A was isolated from poultry during experiments on the
use of ionizing radiation as a method of food preservation (Thornley, Ingram & Barnes,
i960). Among isolates classed as Acinetobacter (earlier called Achromobacter) radiation
resistance varied considerably, and in many strains the property appeared to be inversely related to penicillin resistance (Thornley, 1963). This led to speculation as to
whether both properties were connected with some aspect of cell-wall structure, and
after preliminary experiments had shown that differences in fine structure existed
between a few radiation-sensitive and radiation-resistant strains, a more extensive
investigation was made.
Since the fine structure of these organisms differs in several respects from that of
the Gram-negative bacteria previously studied, observations on one strain (MJT/F5/
199A) are reported here in detail; a comparison of different strains will be made in a
separate paper. The strain selected is one of the more radiation-resistant, and has a
sigmoid survival curve with a D10 value of about 60 Krads in the exponential part of
the curve. It is very sensitive to penicillin, with a minimum inhibitory concentration
of less than o-1 i.u./ml (Thornley, 1967). It was isolated from a chicken carcass irradiated
with 500 Krads of y-radiation, but strains similar in resistance to radiation, sensitivity
274
M. J. Thornley and A. M. Glauert
to penicillin and fine structure have been isolated from control carcasses, so this
strain is not thought to be a mutant with unusual properties.
The classification of this group of organisms is a matter of some doubt, since similar
Gram-negative or Gram-variable, non-motile, coccoid rods or cocci have been
classed as Achromobacter, Alcaligenes, or Achromobacter-Alcaligenes when isolated
from foods, and as Moraxella Iwoffi, M. glucidolytica, Mima, Herellea, Bacterium
anitratum, or a wide variety of other names, when isolated from human sources. On
the basis of a computer survey of strains from culture collections representing all the
above names, together with many isolates from poultry, it was suggested (Thornley,
1967) that all these strains should be included in a single genus, for which the name
Acinetobacter (Brisou & Prevot, 1954) would be suitable.
METHODS
Media and growth conditions
Difco Heart Infusion broth and agar were mainly used, with o-oi % (w/v) CaCl2 added in
later experiments. For comparison, 4 other complex media were tested:
(1) M-5 medium (Weeks & Beck, i960).
(2) TGYM medium (Krabbenhoft, Anderson & Elliker, 1967).
(3) Difco Dextrose Starch.
(4) Difco Brain Heart Infusion.
This organism is strongly aerobic, and was grown either on the surface of agar or in aerated
broth by the following methods:
(A) Thin layers of Heart Infusion agar in Roux bottles were inoculated with 1 ml of a suspension of cells from a slope; after incubation for 24 h at 20 °C, the cells were washed from the
agar surface with 10 ml of water.
(B) T-tubes (Kay & Fildes, 1950) containing 10 ml of liquid medium were aerated by
shaking at 25 °C. These were used to compare different media, and to prepare standard inocula
for larger-scale cultures.
(C) Rotating flasks (Mitchell, 1949) containing 1 1. of Heart Infusion broth with CaCl2
were inoculated with 5 ml of a stationary-phase culture from a T-tube and incubated at 25 °C
for 18-24 h.
Preparation of cell walls
The methods used were based on those of Work (1964). Cells grown by method A or C were
washed twice with water, suspended in water, and disrupted by shaking with glass beads
(Ballotini, grade 12). In early experiments, the Mickle disintegrator was used, 10 ml of cell
suspension and 10 g of beads being shaken for periods of 6-20 min, until microscopic examination showed that most of the cells were broken. Later the Nossal disintegrator (Nossal, 1953)
was found to give more reproducible results. A portion of 12 ml of suspension, containing
about 6 g wet weight of cells and 10 g of glass beads, was shaken at 3760 rev/min for i£ min.
Most of the cells were disrupted by this treatment.
The suspension of broken cells, from either disintegrator, was separated from the glass beads
by nitration and then centrifuged at 34000 g for 30 min. The pellet showed two layers: a more
opaque layer of intact cells at the base, and an overlying layer of more translucent cell walls. The
upper layer was resuspended in water and the bottom layer discarded. The suspension was
recentrifuged and the upper portion of the pellet was resuspended in M NaCl solution, a small
basal region being discarded again. This process reduced the proportion of intact cells until
not more than 3 or 4 could be seen in a microscope field containing several hundred cell walls.
In the routine washing procedure, the walls suspended in M NaCl were centrifuged, then
washed once in water, 3 times in o-i M phosphate buffer, pH 7-0, and 3 times in water. This
•procedure was used except where otherwise stated.
Fine structure of an Acinetobacter
275
Electron-microscope examination and enzyme treatments were made on either freshly prepared or freeze-dried cell walls, and chemical tests always on freeze-dried walls.
Enzyme treatments
Cell walls were incubated at 37 °C with various enzymes, under the conditions shown in
Table 1. The enzyme preparations used are listed below.
Papain (Sigma): twice crystallized, suspended in 0-05 M sodium acetate, pH 4-5.
Ficin: a crude powder.
Streptomyces griseus protease (Sigma): purified.
Pronase (Calbiochem): a crude preparation of Streptomyces griseus protease.
Trypsin (Worthington or Sigma): crystalline.
Lysozyme (Sigma): from egg white, 3 times crystallized.
Steapsin (Nutritional Biochemicals Corporation): a crude preparation of pancreatic lipase.
Naja naja venom: a crude preparation, used as a source of phospholipase A.
Hyaluronidase (Light): from bovine testes.
Table 1. Conditions used during treatment of cell walls with enzymes
Enzyme
Concentration
(mg/ml)
Papain
0-13
Ficin
Saturated
solution
diluted 3/10
Streptomyces
griseus protease
Pronase
Trypsin
Lysozyme
Steapsin
(pancreatic lipase)
Phospholipase A
(Naja naja venom)
Hyaluronidase
1
o-i
5-0, o-if
o-i
0-28
o-i
o-i
Other
additives
(mg/ml)
Cysteine, o-6;
EDTA, i-8
Cysteine, o-6;
EDTA, i-8
Times of
incubation
pH
(h)
6-o
1, 3, 18
60
3
—
7-0
3
—
—
—
8-5*
3
8-o
3,i8
OS. 1. 3. 18
Calcium
acetate, i-o
Calcium
acetate, i-o
—
7-0
7-0
3
7-0
3
7-0
2, 18
The temperature of incubation was 37 °C in all experiments. Cell walls were added at
concentrations up to 2 mg/ml.
* Tris buffer 0-02 M, was used with pronase. In all other experiments, o-i M phosphate
buffer was used.
f With both concentrations, some of the added trypsin remained undissolved, and was
removed by low-speed centrifugation after incubation.
Electron microscopy
Thin sections. Intact cells and cell walls were fixed by the Ryter-Kellenberger (1958) method
or by the glutaraldehyde-osmium procedure described by Glauert & Thornley (1966). After
embedding in Araldite, thin sections showing silver-grey interference colours were cut on an
A. F. Huxley or LKB Ultrotome III microtome. Sections were stained with uranyl acetate
and lead citrate.
Negative staining. Equal quantities of a suspension of cells or cell walls in water and 2 %
potassium phosphotungstate were mixed together in a watch glass. A small drop of the mixture
was placed on the collodion-carbon support film of an electron-microscope grid with a fine
pipette, and excess fluid removed with filter paper.
276
M. J. Thornley and A. M. Glauert
Electron microscopy. Electron micrographs were taken with a Siemens Elmiskop I or AEI
EM6B electron microscope, operating at 60 kV with a 50-/* objective aperture, on Ilford
Special Contrasty Lantern Plates, at instrumental magnifications of 5000-40000.
Chemical estimations
Identification of sugars. Acid hydrolysis was carried out by adding a 10 mg portion of freezedried cell walls to 5 ml of N HC1; this suspension was then sealed in a glass ampoule and
heated at 100° for 2 h. After removal of acid, the solution was evaporated to a small volume
twice, and kept in a vacuum desiccator overnight.
Paper chromatograms of the hydrolysate were made, using ethyl acetate-acetic acid-water
(3:1:3) or ethyl acetate-pyridine-water (2:1:2) (Jermyn & Isherwood, 1949).
Duplicate chromatograms were sprayed with ammoniacal silver nitrate (Partridge, 1948) and
with aniline hydrogen phthalate (Partridge, 1949).
Identification of amino acids
The procedure of Salton (1953) was followed. 10 mg samples of freeze-dried cell walls were
hydrolysed with 5 ml of 6 N HC1 for 24 h at 100 °C. After removal of the acid by repeated
evaporation under vacuum, two-dimensional paper chromatograms (Consden, Gordon &
Martin, 1944) were made. The identification of the amino acids was confirmed, for one preparation, by means of the Technicon autoanalyser.
The effects of enzymes on amino acid composition were studied by washing the residue remaining after enzyme treatment several times with water, freeze-drying it, and then testing
10 mg by acid hydrolysis and paper chromatography in the usual way.
RESULTS
Intact cells
Living cells of Acinetobacter strain 199 A examined in the phase-contrast microscope appear as non-motile cocci or coccoid rods, with a length/width ratio between
1 and 2. Daughter cells are often joined in pairs and larger groups of cells are sometimes observed. The bacteria are Gram-negative but appear a darker red than typical
Gram-negative bacteria such as Pseudomonas when counter-stained with safranin; this
property is common in strains of Acinetobacter (Thornley, 1967) and has been described also for the genus Moraxella (Murray & Truant, 1954; Ryter & Pie"chaud,
Preliminary examination with the electron microscope of thin sections of cells grown
in Difco Heart Infusion broth showed apparently empty spaces and vacuoles in some
cells, and unstained areas, possibly corresponding to the vacuoles, were seen in the
light microscope after staining with Loffler's methylene blue or carbol fuchsin. In an
attempt to avoid what was thought to be an abnormality of structure the appearances
of cells grown in 6 different complex media (see Methods) were compared.
Viable counts were made after overnight growth with aeration at 25 °C, a standard
inoculum being used. Counts were low with both M- 5 and TGYM media, and varied
between 2 x io8 and io9/ml with the Difco media: Dextrose Starch, Brain Heart
Infusion, Heart Infusion and Heart Infusion plus o-oi % CaCl2, the last giving the
highest figure of io9/ml.
Cultures in liquid media were examined unstained by phase-contrast microscopy
and also after staining with Loffler's methylene blue, carbol fuchsin and the Gram
Fine structure of an Acinetobacter
277
stain. Slime formation was greatest in Brain Heart Infusion and also occurred in the
other two infusion media. The cells grew in chains (Fig. 2) in Anderson's medium,
while pairs of cells were commonest in the other media; tetrads (Fig. 3) were frequent
in Dextrose Starch.
The average size of the cells varied with the medium; Heart Infusion with CaCl2
gave smaller cells than Heart Infusion alone. The presence of a great variation in cell
size in one culture, or of irregularities of cell shape, or of unstained cells or portions of
cells was taken as a sign of abnormality of growth. Some of these features were
present in all cultures except those grown in Heart Infusion broth plus CaCl2, in
which the cells were slightly uneven in size but had no other 'abnormal' features.
This medium was therefore chosen for future work and used for growth with aeration
at 25 °C.
During the different phases of growth the cells varied in size and arrangement,
being larger during the log phase (Fig. 4), after 10 h growth by method C, than in the
early stationary phase after 24 h (Fig. 5). Pairs of cells were commoner than larger
groups at both stages, but where groups occurred these were short chains in the 10 h
culture and tetrads in the 24 h culture. Presumably successive planes of division were
parallel to each other in the chains and perpendicular in the tetrads (Figs. 4, 5)
(Thornley, 1967). This difference in the plane of division has been regarded as a
character distinguishing the genera Moraxella and Neisseria in earlier work (Murray &
Truant, 1954; Piechaud, 1961), so it is of interest to find both processes occurring at
different stages in the same culture.
Most preparations were made with 24 h cultures which had received a standard
inoculum and were in the early stationary phase.
A typical electron micrograph of a pair of cells negatively stained with potassium
phosphotungstate is shown in Fig. 6. The bacteria are surrounded by a very dark area
of capsular material which retains the phosphotungstate and nearly obscures the few
fimbriae that are present. The surfaces of the cells show a pattern suggesting the
presence of wrinkles. This pattern is much more clearly seen in preparations of cell
walls (see later section); its presence in intact cells establishes that it is not an artefact produced during breakage of the bacteria.
Thin sections of intact cells
Earlier preparations were fixed for sectioning by the Ryter-Kellenberger method
(1958), but this technique gave variable results even for different cells in the same
preparation. Consequently in later work a combination of glutaraldehyde and Kellenberger fixation (Glauert & Thornley, 1966) was used. This procedure gave more
uniform results, although some of the surface layers of the cells were less clearly
visible.
The cell contents are similar in appearance to those of most other Gram-negative
bacteria. Finely fibrillar nuclear material (Fig. 7, n) occupies irregularly shaped
regions of low electron density, and darkly stained ribosomes (Fig. 7, rs) can be distinguished throughout the cytoplasm. Mesosomes are visible in many of the Acinetobacter cells, where they are often associated with the ingrowing septa of dividing
278
M. J. Thornley and A. M. Glauert
cells (Figs. 8, 9, m). Their frequency varied in different preparations for unknown
reasons.
Two membranes, the plasma membrane (pm) and an outer membrane (om),
separated by an intermediate layer can be distinguished in sections of intact cells
(Figs. 10, 11). The plasma membrane appears as a typical unit membrane and can be
seen at the boundary of the cell (Fig. 10) and next to the septa (Fig. 7). The outer
membrane has a similar appearance to the plasma membrane in section, and either
forms a fairly smooth or a wavy, scalloped, outline to the cell (Fig. 10). After glutaraldehyde fixation two regions are visible in the intermediate layer, the inner being darkly
stained (Fig. 11, r). The layers seen in sections are shown diagrammatically in
Fig. 1.
Peg-like s u b u n i t s - — a _ i _ « _ o - - ^ > ^ . ; : ; C 7 ^ \ ^ ^ ; ^ ^ < r r r r ^ \ _ _ ^ ^ ' membrane
Site of wrinkled
structure
Rigid layer
,
•.''••: 1:v-.:v~'.'-':;.-K '.;..•".;.'-i'-'"';/,:,).V "•'•.).;>'•" :':i';''';'.V;..' '•: I Intermediate
J&^S^ftVSSK^SSS^!SSI^^^ISSS^B^^s^i^S&llS0i J
layer
: : • ; • • . • • • _ • • - . • -•.•••:•-.•••••• - - : ' - . • • • • . - • - • • •.••"<•
•'•••.'-••.•'
. ' . • . • • • . • ' . • • • • • . • . • . • • . • • : • : • '
••••:.:-V:^~• . " • ' . '
p l a s m a
m e m b r a n e
Fig. 1. Diagram of structures observed in thin sections of the cell wall. The peg-like
subunits are only seen in negatively stained preparations.
Cell division begins by the ingrowth of a thick septum (Fig. 8, s) which is apparently
continuous with the material of the intermediate layer, and this is accompanied by a
variable amount of constriction of the whole cell. This constriction is sometimes very
slight, so that a continuous septum is formed across the cell (Fig. 7).
More frequently the thickened ring of septum material shows considerable constriction, and in the extreme example illustrated in Fig. 9 the daughter cells have
nearly separated but the septum is not yet closed in the centre and the space is filled
by a plug of membranous material. At a later stage of division the two cells sometimes
remain connected for a time by an irregular mass of septum material.
Cell walls
Cell walls were prepared by mechanical breakage followed by repeated washing
(see Methods). Thin sections show that the entire cell contents are removed, although
small fragments of the plasma membrane remain in a few cell walls (Fig. 12). The
outer membrane (Figs. 12, 13) is clearly visible, together with an inner dense line
(Fig. 12), which is presumably the same as the inner, darkly stained region of the
intermediate layer (Fig. 11). The broken edges of the walls frequently curl inwards
(Fig. 12). The septa of dividing cells are preserved and are clearly seen to be continuous with the material of the intermediate layer (Fig. 13).
Negatively stained preparations of cell walls which have been washed only twice
show that the walls are still surrounded by some capsular material which retains the
dense phosphotungstate (Fig. 14). The wrinkled appearance of the surface is clearly
evident, and fimbriae (Fig. 14, / ) are visible in the less heavily stained regions of the
Fine structure of an Acinetobacter
279
preparation. The cell wall illustrated in Fig. 14 appears to lie with the break uppermost,
and the curled-back parts of the wall can be seen adjacent to the break. An indication
of a periodic structure is noticeable at the periphery of the cell in a zone where the
wrinkles are absent (Fig. 14, arrow).
After one wash in M NaCl solution and 6 washes in water the structure of the surface of the cell wall is much more clearly revealed (Fig. 15). The wrinkles are still
evident but they are now seen to be covered by a patterned layer containing a regular
array of peg-like subunits. These subunits are most clearly visible at the folded edges
of the cell wall (Fig. 15), where they are spaced about 90 A apart. Their arrangement
on the surface of the wall is not clear; it appears that the arrays of subunits on the
upper and lower surfaces of the wall are both contrasted by the negative stain, and
consequently complex moire patterns are observed, resulting from the superposition
of two patterns at an angle to each other.
The same structures are visible, although less clearly, in cell walls that have been
frozen and thawed or freeze-dried before mounting in potassium phosphotungstate.
Table 2. Effect of chemical and enzymic treatments on different layers of the cell
wall of Acinetobacter strain lgg A, as seen by negative staining
Layer of cell wall
Treatment
None: intact walls
Papain: early stages
Papain: later stages
Lysozyme
Lysozyme + papain
Steapsin
SDS (cone, repeated
extraction)
Phenol
Heat, 60 °C for 15 min,
on freshly broken walls
Heat, 60 °C for 15 min,
on washed walls
EDTA in tris buffer
Ficin
Urea, 67 M
Phospholipase A (Naja
naja venom)
Hyaluronidase
18
Outer
membrane
with pegs
Visible
Loses pegs
Breaks down to
tubes and vesicles
(no pegs)
Vesicles with pegs
Vesicles (no pegs)
Vesicles
Wrinkled
structure
Visible
—
—
Rigid
layer
Not visible
Visible
Visible
—
—
—
—
Debris, probably
from this layer
Visible
—
—
Visible
—
—
Visible
Not visible
Broken in places,
Loosened
blebs formed
Broken down
Debris, probably
to fragments
from this layer
Partially detached
Loosened
in blebs
—
Loses pegs
?Some removed
Visible
Very little effect
Visible
Not visible
Visible
Not visible
Some disorganization of structure
Cell Sci. 3
280
M. J. Thornley and A. M. Glauert
Enzymic and chemical treatment of cell walls
To investigate the chemical nature of the various layers of isolated cell walls observed by electron microscopy, attempts were made to fractionate the walls by enzymic
and chemical treatments, many of which were modifications of those used earlier by
others. The results are summarized in Table 2.
Proteolytic enzymes. Several proteolytic enzymes were compared, and all had
broadly similar effects on the cell walls. Papain, in a twice crystallized form, gave the
most clear-cut results and was the most extensively used.
Cell walls treated with papain for 1-3 h lose the peg-like subunits from the surface,
and the wrinkled pattern also disappears (Fig. 16). An outer membrane remains but
this shows signs of disintegration (Fig. 16) and in many walls it has completely
broken down into vesicles and irregular tubes (Fig. 17) which often have a similar
appearance to myelin figures. Treatment with papain also reveals a smooth inner
layer (Fig. 16, r) which is enclosed by the outer membrane and which is not visible
before enzymic digestion. A thick ring of material is seen between incompletely
separated daughter cell walls (Fig. 16, s) and presumably represents the septum which
has not yet closed (compare Figs. 16 and 9). In some preparations the fragments
of the outer membrane have almost completely separated from the inner layer and
then it is clear that the rings and the inner layer are part of the same structure (Fig. 18).
This separation is even cleaner after digestion for 18 h. Presumably the inner layer
corresponds to the dense inner line of the intermediate layer observed in the thin
sections (Fig. 11). The outer edges of the layer are fairly smooth and its surface is
covered with a large number of small holes or pits which retain the negative stain
(Figs. 17, 18). Chemical tests (see below) show that these smooth inner layers, together with the rings, have similar chemical properties to the ' rigid layers' isolated by
Weidel, Frank & Martin (i960) from Escherichia colt, and this term will therefore be
applied to them.
A search was made for the remains of the wrinkled pattern which is such a striking
feature of untreated walls. A few walls appear to be unaffected by the enzymic digestion but intermediates between them and structures lacking all sign of the wrinkles
are rare. In a very few cell walls some material resembling the wrinkles is present
(Fig. 19, arrow) and appears to be on the outside of the rigid layer and partly detached
from it.
Fixation with glutaraldehyde prevents the removal of the wrinkled material by
papain.
Thin sections of the papain-treated walls were also studied and in some sections the
wall appeared as a thin outer layer and a thick inner layer, separated by a space
(Fig. 20). This appearance was thought to correspond with the stage of digestion
illustrated in Fig. 16, with the two layers consisting of the outer membrane and the
rigid layer. More frequently only the thick, inner, rigid layers remained in the sections and were seen to be continuous with associated septa (Fig. 21). Sections also
showed septa at early stages of formation and in the form of rings (Fig. 22). These
observations confirm the identification of the rigid layer seen in negatively stained
Fine structure of an Acinetobacter
281
preparations with the inner dense line of the intermediate layer that is visible in
sections of intact cells.
Two different preparations of Streptomyces griseus protease were used, one being
purified while the other (pronase) was not, and different concentrations and buffer
systems were also used (Table 1).
After 3 h digestion with the purified preparation, the outer membranes are degraded more extensively than by either papain or ficin, and many fragments roll up
into onion-like and other myelinic figures. There are also signs of attack on the rigid
layers, which are covered with dark spots, possibly representing holes or pits, similar
to those seen after papain treatment. Pronase had a very similar effect, but the fragments of the outer membranes are mostly in small rounded vesicles, and the ' holey'
appearance of the rigid layers is even more marked (Fig. 23). Enzymes capable of
degrading bacterial mucopeptide occur in some Streptomyces species (Ghuysen, i960)
and it is therefore possible that some activity of this kind is present in addition to the
protease.
Treatment of freshly prepared walls with the lower concentration of trypsin
(Table 1) for 3 h has little effect on the outer membrane and the peg-like subunits are
still present. The wrinkles appear to be partly removed, since they are less densely
packed and some of the material rounds up to form small circular structures. In some
regions the material of the wrinkled layer is found lying free from the other layers of
the wall in tangled masses (Fig. 24).
Incubation with the higher concentration of trypsin for 18 h causes the partial
fragmentation of the outer membrane into vesicular structures which are mixed with
the remains of the wrinkled layer to give a confused picture. Pretreatment of the walls
with glutaraldehyde entirely prevents the action of trypsin, while pretreatment by
heat (60 °C for 15 min) enhances the effect of trypsin digestion, and only the rigid
layers remain intact among the debris of the wrinkled layer and outer membrane.
It appears that the wrinkled layer is more resistant to the action of trypsin than to
that of the other proteolytic enzymes studied.
Lysozyme. Intact cells incubated under the conditions shown in Table 1 are sensitive to lysozyme without the addition of EDTA or other substances. The turbidity
decreases to 25 % of the initial value during incubation for 30 min at 37 °C, reaches
22% after 60 min, and then remains nearly constant.
Many of the cell walls incubated for 3 h with lysozyme disintegrate into small
rounded fragments 80-150 m/t in diameter (Fig. 25), with a few larger fragments of
similar appearance. All traces of the rigid layers and rings disappear and it is probable
that the fragments originate from the outer membrane. A pattern of light dots is
visible on many fragments (Fig. 25, arrow); they presumably represent the peg-like
subunits on the outer membrane. This interpretation is supported by the fact that
treatment with papain for 1 h after lysozyme for 3 h removes the light dots without
otherwise altering the appearance of the preparation (Fig. 26).
Shorter treatments with lysozyme, including 30 and 60 min, were tried in an
attempt to see earlier stages in the formation of the fragments, but intermediate
structures were not observed.
18-2
282
M. J. Thornley and A. M. Glauert
Pancreatic lipase. Incubation of cell walls with steapsin (pancreatic lipase) for 3 h
at 37 °C removes most of the outer membranes and wrinkled material from the rigid
layers, which retain their shape. Tangled masses of material (Fig. 27) and densely
packed arrays of small vesicular bodies with a granular surface pattern (Fig. 28) are
observed and possibly represent the remains of the wrinkled material and the outer
membrane respectively.
Lipid solvents and detergents. Various extraction procedures using sodium dodecyl
sulphate (SDS) were tested.
(a) Cell walls were incubated with 0-2% SDS for 15 min at 37 °C, washed
thoroughly, and then mounted in potassium phosphotungstate. Removal of the SDS
proved very difficult even at this low concentration, since solid aggregates formed
which centrifuged down with the walls (Fig. 29, SDS). The treated walls have rough
outlines and it appears as if most of the material in the outer layers is removed, leaving
the rigid layers covered with irregular granular subunits about 100-140 A in diameter (Fig. 29). Thin sections confirm that only the rigid layers and septa remain with
some dense granular deposits (Fig. 30). Many of the walls clump together in groups
during preparation for embedding and sectioning.
(b) Similar results were obtained when the cell walls were added to a 1 % solution
of SDS in water, centrifuged immediately, and then mounted in potassium phosphotungstate. The walls have the same appearance if they are treated with papain for
18 h before extraction with SDS, suggesting that the granular subunits on the rigid
layers represent the remains of the outer membrane, since papain appears to remove
the wrinkled material (Fig. 16).
(c) Cell walls were extracted with a saturated solution of SDS in phosphate buffer
at 37 °C for three successive periods of 24 h each, as in the method used by Schocher,
Bayley & Watson (1962) for the preparation of the mucopeptide of an Aerobacter.
Subsequent removal of the SDS again proved very difficult. This treatment removes
the outer membrane and wrinkled material, and clean rigid layers and rings are obtained (Fig. 31) mixed with only very small fragments of other material.
To investigate the action of phenol, cell walls were treated with a 50% (w/w)
phenol-water mixture at 65 °C for 10 min, cooled, collected from the interface,
dialysed overnight against water, and washed. Only rigid layers with rings remain
intact and are often clumped together. Many small granular and membranous fragments are also present in some preparations. After further purification (see chemical
results) by subsequent incubation with papain for 18 h, cleaner rigid layers are observed (Fig. 32). The fibrous structure of the septa is evident. A few fragments of
other material are still visible and presumably represent the remains of the wrinkled
layers and outer membranes. A similar preparation of rigid layers was obtained when
the material was digested with papain before the phenol treatment.
Successive extraction with methanol, ethanol-ether (3:1) and chloroform does
not remove any complete layer from the cell wall, but leaves the surface with an
irregularly pitted appearance.
Fixation to prevent autolytic changes. It is known that autolysis can produce rapid
changes in cell-wall structure (Weidel, Frank & Leutgeb, 1963), and for chemical
Fine structure of an Acinetobacter
283
analysis, some method of fixation is usually adopted to prevent these changes. For the
present work, the prevention of autolysis was desirable, but both methods tested
produced some change in structure.
(a) Sodium dodecyl sulphate at concentrations of 0-4 or 4% was added during cell
breakage by Weidel et al. (i960) and Weidel et al. (1963) for preparation of mucopeptide material. The effect of 0-2% SDS on the Acinetobacter cell walls has already
been described, and this method was clearly inapplicable here.
(b) Fixation by heat (60 °C for 15 min) has been used to inactivate autolytic
enzymes. When applied immediately after mechanical breakage (Work, 1964) and
before separation of intact and broken cells, or washing, the treatment damages the
outer membrane so that many surface blebs are formed, and also alters the wrinkled
material (Fig. 33). Cell walls prepared with the usual sequence of 8 washes and then
given the same heat treatment are more extensively damaged: the outer membranes
are largely removed and little of the wrinkled material is visible, while the rigid layers
are exposed. The appearance is very similar to that produced by prolonged papain
digestion.
Controls to show autolytic and other changes. As autolytic changes were not prevented
during the enzymic treatment of cell walls, some experiments were made to find the
extent to which such changes might take place during prolonged incubation in the
presence of various additives.
(a) A few drops of chloroform were added to the incubation mixtures in nearly all
enzyme treatments of 18 h duration. The effect on cell walls of incubation in o-i M
phosphate buffer at pH 6-o with added chloroform for 18 h at 37 °C was therefore
examined. There is some slight effect on the wrinkled material, which appears less
densely packed, but in general the cell walls are unaffected. The slight effect seen
may have been due to either autolysis or the presence of chloroform.
(b) A preparation of cell walls was incubated for 18 h at 37 CC in tris buffer (pH 7-05)
to which was added i-8 mg/ml of EDTA (ethylenediaminetetra-acetic acid, disodium
salt) and a few drops of chloroform. Like chloroform alone, this treatment also has a
loosening effect on the wrinkled material and, in addition, partially detaches the
outer membrane so that peripheral blebs are produced (Fig. 34). As in many other
preparations, not all the walls are affected to the same extent, and Fig. 34 illustrates
an extreme example of the effect of incubation with tris-EDTA.
The changes just described are considerably less marked than those produced during
incubation for 18 h with either papain, trypsin or lysozyme, so that it seems likely that
the effects of these treatments are due largely to the added enzymes. However, some
contribution from other factors cannot be excluded, especially in the case of papain,
which was always used in the presence of EDTA.
Further incubation at 37 °C causes an increased release of fragments of the outer
membrane (Fig. 35), but much of the membrane still remains after 10 weeks. At the
same time there are indications that the wrinkled material is beginning to be digested.
Similar alterations occur in walls incubated with tris-EDTA and chloroform at o °C,
but much more slowly, and blebs of the outer membrane only begin to appear after
about 10 weeks.
284
M. J. Thornley and A. M. Glauert
Incubation of intact cells with tris-EDTA causes a similar blebbing and removal of
parts of the outer membrane.
(c) Prolonged incubation at 37 °C in water gives structural changes similar to those
observed in the presence of tris-EDTA except that the outer membrane appears to be
more stable, often remaining as an intact sheet, and fewer membranous fragments are
observed (Fig. 36). The main site of action of the autolytic enzymes appears to be the
wrinkled layer.
Little change occurs on incubation in water at 1 °C, again suggesting that the
changes observed are due to the action of autolytic enzymes.
Table 3. Amino acids in cell walls of strain igg A after various treatments
Cell wall treatment
A
None
A
f
Amino acid
Alanine
Arginine
Aspartic acid
Diaminopimelic acid
Glutamic acid
Glycine
Histidine
Isoleucine
Leucine
Lysine
Phenylalanine
Proline
Serine
Threonine
Tyrosine
Valine
Total
A
io-6
29
Lysozyme
Phenol
papain
+
+
+
+
+
+
+
—
+
+
+
—
+
+
+
—
+
+
+
+
—
+
+
+
+
+
+
+
+
+
+
+
—
+
+
+
+
+
—
—
+
+
+
+
—
—
—
±
—
±
±
B
n-6
•
12-4
4-6
t
2-8|
4-2/
2-9
2-4
3*3
4-2
5-7
±
±
—
76-8
A, Results obtained with 'Technicon' autoanalyser, expressed as percentage dry weight of
walls.
B, All other results are from paper chromatograms.
* DAP was not estimated because no standard was available.
f Histidine was present in too small a quantity to estimate.
Chemical results
Sugars. Cell walls prepared by the routine method did not contain ribose, indicating
the absence of cytoplasmic contamination. Glucose and galactose were both present,
as also were compounds which ran more slowly on the paper chromatograms, but
which were not identified.
Amino acids. Results are shown in Table 3. Diaminopimelic acid (DAP) was
present, as in all Gram-negative bacterial walls so far studied. The walls also contained many other amino acids, but not those containing sulphur. Histidine was found
in a very small quantity by the amino acid analyser, but was not detected on paper
Fine structure of an Acinetobacter
285
chromatograms. Three amino acids found in large quantities were alanine, aspartic
acid and glutamic acid; the results quoted by Salton (1964, table 25) for three other
Gram-negative bacteria showed that aspartic acid was found in similar amounts in
two species but both alanine and glutamic acid were much less abundant in all three.
The percentage of the dry weight of the wall accounted for by amino acids other
than DAP was 77%, a much larger amount than the comparable figures of 43-60%
quoted by Salton (1964, table 25).
Digestion with lysozyme for 18 h caused the loss of DAP from the insoluble
residue. The proline spot was also absent, but this is not thought to be significant,
since proline could not always be detected in chromatograms made from untreated
cell walls.
Papain treatment for 18 h gave no change in the amino acids present in the residue.
Phenol extraction caused the reduction or disappearance of several amino acids originally present in small quantity, and when this was followed by papain digestion for
18 h, the composition was greatly changed. Large spots were formed for alanine and
glutamic acid, a definite spot for DAP, and smaller amounts of aspartic acid, glycine,
serine, valine and possibly tyrosine and leucine/isoleucine were indicated. Since this
preparation consisted of rigid layers with a few fragments of other material (Fig. 32),
the main amino acids of the rigid layers were evidently alanine, glutamic acid and
DAP, as for all other Gram-negative bacteria so far studied (Salton, 1964). The minor
components may be located in either the rigid layers or in fragments of other material.
DISCUSSION
This study of the fine structure of Acinetobacter strain 199A has shown that the
appearance of the cells in thin sections is similar to that of other Gram-negative
bacteria except for the structure and mode of formation of the cross-walls. The
negative-staining technique reveals some structures not previously described in the
surfaces of Gram-negative bacteria.
Mode of division
The ingrowth of a thick septum, accompanied by mesosomes, observed in Acinetobacter strain 199A is the typical mode of division for Gram-positive bacteria (FitzJames, i960; Glauert, 1962), but until recently it was thought that Gram-negative
bacteria divide by a gradual constriction of all the layers of the surface leading to an
eventual pinching-off of two daughter cells. Steed & Murray (1966) have now shown
that septum formation can often be demonstrated in Gram-negative bacteria when the
conditions of growth and fixation are right. Thin septa and simple mesosomes are
observed in both Escherichia coli and Spirillum serpens grown and fixed at 45 °C, and
also in cells grown at 30 °C and fixed in the presence of a buffer more dilute than that
normally used. Steed & Murray (1966) concluded that the tonicity of the fixing environment is critical for the preservation of the fine septa.
Some other Gram-negative bacteria show septation without special conditions of
growth or fixation. Among these are Neisseria gonorrhoeae and Veillonella which also
286
M. J. Thornley and A. M. Glauert
resemble Acinetobacter strain 199A in being coccoid in shape. N. gonorrhoeae
appears to divide by ingrowth of a simple mesosome consisting of one fold of membranous material, followed by the formation of a thin septum, accompanied by some
constriction (Fitz-James, 1964). A fairly thick septum is visible in micrographs published by Bladen & Mergenhagen (1964; figs. 2, 8) of Veillonella, although the authors
do not comment on it. Asticcacaulis excentricus also forms a thick septum (Poindexter
& Cohen-Bazire, 1964), while differing markedly from Acinetobacter in other respects,
and septa have been observed in the photosynthetic green bacterium Chlorobium
thiosulfatum (Cohen-Bazire, Pfennig & Kunisawa, 1964).
Cell-wall structure
Thin sections. Considerable advances have recently been made in our knowledge of
the cell walls of Gram-negative bacteria. Earlier workers studying thin sections
(for example, Kellenberger & Ryter, 1958) found only two structures, each with the
appearance of a unit membrane, at the boundary of cells such as E. coli, and these
were interpreted as the cell wall and plasma membrane. Improved methods have
revealed an intermediate layer between these two (Claus & Roth, 1964) and the cell
wall is now usually defined as the outer membrane plus the intermediate layer (see
Fig. 1). The rigid layer (Weidel et al. i960) of mucopeptide-containing material has
been identified in sections of several bacteria as a thin, darkly stained line within this
intermediate layer. It is often adjacent to the plasma membrane and approximately
parallel to it, while the outer membrane frequently has many convolutions. Although
varying in detail, this basic cell-wall structure has been shown in many Gram-negative
bacteria, including E. coli (Murray, Steed & Elson, 1965; de Petris, 1965, 1967),
Spirillum (Murray, 1963; Murray et al. 1965), Veillonella (Bladen & Mergenhagen,
1964), Moraxella (Ryter & Piechaud, 1963), Acetobacter (Claus & Roth, 1964),
Caulobacter (Poindexter & Cohen-Bazire, 1964) and Rhodospirilium (Cohen-Bazire &
Kunisawa, 1963).
The cell wall of Acinetobacter strain 199A also shows this basic structure in sections
of intact cells and of isolated walls, and the thick septa appear to be continuous with
the dense material of the intermediate layer. The effects of various treatments on the
structure of the cell walls of four Gram-negative bacteria are compared in Table 4.
Negative staining. Three distinct layers are visible in negatively stained preparations
of cell walls of Acinetobacter strain 199A: the outer membrane, the wrinkled material
and the rigid layer. These will be discussed in turn.
(a) An array of peg-like subunits is seen on the outer surface of the cell wall after
negative staining, but is not visible in sections. The outer membrane, which in
section has the appearance of a typical unit membrane, is thought to correspond to
the outer layer seen after papain digestion or prolonged autolysis in water. This outer
layer disintegrates to form tubes and vesicles after longer papain treatment, and
similar vesicles remain after treatment with lysozyme and papain. Lysozyme alone
produces vesicles with a pattern of light dots similar to that seen on the outer layer
after autolysis. It is thought, therefore, that the outer surface of the outer membrane
carries the array of peg-like subunits and that these are removed by treatment with
Fine structure of an Acinetobacter
287
papain or other proteolytic enzymes, but not by lysozyme. This suggestion has been
confirmed by recent experiments with spheroplasts (unpublished observations). The
pattern in which the pegs are arranged is being studied further by diffraction techniques.
Regular arrays of subunits have been observed on the surfaces of a number of other
bacteria by means of shadow-casting, replication and negative staining, but the exact
position of the arrays in relation to the various layers of the cell surface has not always
been clear. The hexagonal pattern on the surface of Spirillum serpens is difficult to
detect in sections, but Murray (1963) concluded that it lies outside the outer membrane in a somewhat similar position to the subunits of Acinetobacter.
The positions of the hexagonal patterns in the surfaces of Micrococcus radiodurans
(Thornley, Home & Glauert, 1965) and Escherichia coli (Fischman & Weinbaum,
1967) are not known so exactly, although it appears that they are outside the rigid,
mucopeptide-containing layers of the walls.
The outer membrane of Acinetobacter strain 199 A has not been prepared in a purified form, so that no direct evidence on its composition is available. After lysozyme
treatment, the remaining fragments and vesicles, which are thought to originate from
this layer, contain a full set of amino acids, apart from DAP, indicating some protein
constituents. The fact that the pegs are removed by proteolytic enzymes suggests that
they either consist of protein or are attached to the membrane by protein. The
particles of similar dimensions that are observed in negatively stained preparations of
the microvilli of intestinal epithelial cells are also removed by papain (Oda & Seki,
1966). There is some evidence that these particles contain the invertase and leucine
aminopeptidase activities of the intestinal membranes, and it will be interesting to
discover whether the subunits associated with bacterial membranes have any enzymic
function.
Lipid would be expected to be present in the outer membranes of Acinetobacter
since they appear as unit membranes in section and form myelin figures and vesicles
after various treatments.
In Table 4 the effect of different treatments on the outer membrane of Acinetobacter
strain 199 A are compared with the results of other workers for various bacteria. In
general, the effects are very similar. Lysozyme treatment leaves the outer membrane
intact except in the Acinetobacter strain, where vesicles are formed. Phenol treatment
causes the disappearance of the outer membrane in Acinetobacter, E. coli and Veillonella, and SDS appears to cause rather more extensive breakdown of the outer membrane in the Acinetobacter strain than in E. coli. Papain treatment results in the
separation of outer membranes and rigid layers in both Acinetobacter and E. coli, but
in the Acinetobacter the outer membrane itself is degraded to vesicles by prolonged
treatment.
The effects of EDTA on Gram-negative bacteria vary with the organism and the
conditions of treatment, and include an increase in permeability, extraction of
lipopolysaccharide (Leive, 1965 a, b) and, in more sensitive bacteria, death of the
organism and dissolution of the cell wall (Eagon & Carson, 1965; Wilkinson, 1967).
Most authors attribute these effects to the chelating properties of EDTA, resulting in
SDS (cone, repeated
extraction)
Pancreatin, amylase, pepsin
Pancreatin, amylase, pepsin+
SDS
Papain + SDS
SDS (dilute)
Phenol
Lysozyme + papain
Heat
Papain
Early stage
Later stage
Lysozyme
Treatment
Hexagonal pattern
disappears
Hexagonal pattern
disappears
n.t.
n.t.
n.t.
OM and R separated
Only R remains
Only R remains
Only R remains
* n.t., not tested; OM, outer membrane; R, rigid layer.
n.t.
n.t.
n.t.
OM remains
R disappears
n.t.
n.t.*
OM and R separated
OM remains
R disappears
OM remains
R disappears
OM disappears
R remains
OM usually unchanged
R remains
n.t.
OM and R separated
OM ->-vesicles
OM -> vesicles
R disappears
OM -> vesicles
R disappears
OM disappears
R remains
OM -> granular bits
R remains
OM disappears
R remains
n.t.
n.t.
n.t.
Spirillum serpens
(Murray (1963))
Blebs on OM
(de Petris (1967))
Escherichia coli B
Blebs on OM
Adnetobacter strain 199 A
(present paper)
Organism and reference
n.t.
n.t.
n.t.
n.t.
OM disappears
R remains
n.t.
n.t.
n.t.
OM remains
R disappears
n.t.
n.t.
Veillonella
(Bladen & Mergenhagen (1964))
Table 4. A comparison of the effects of various treatments on the cell walls of different Gram-negativebacteria
0
S|
00
00
a.
nley t
Fine structure of an Acinetobacter
289
the removal of divalent metal ions, which are thought to be necessary for the integrity
of the outer layers of the cell wall. The disorganization of the outer membrane of the
Acinetobacter by EDTA would be consistent with this theory.
It has been shown that the lipopolysaccharide which can be extracted from Gramnegative bacteria, and which has endotoxic and antigenic properties, is located in the
outer membrane in Veillonella (Bladen & Mergenhagen, 1964). Phenol-water extracts
of Veillonella contain particles similar to those observed in negatively stained preparations of Acinetobacter cell walls after treatment with lysozyme and papain and
which appear to be fragments of the outer membrane, while sections of lipopolysaccharide extracted from E. coli show vesicles and filaments with the same triplelayered structure as the membrane (de Petris, 1967). Additional evidence for the
location of the lipopolysaccharide-lipoprotein complex in the outer membrane of
Gram-negative bacteria is provided by the studies of Knox, Vesk & Work (1966) on a
lysine-requiring mutant of E. coli which excretes lipopolysaccharides identical to those
extracted from cells by phenol, when growing under conditions of limited lysine. It is
thought that the lipopolysaccharide-lipoprotein complex found in the growth medium
probably originates from blebs on the outer membrane which break off and are released (Work, Knox & Vesk, 1966).
From these many observations it seems likely that lipid, protein and polysaccharide
are all present in the outer membranes of Gram-negative bacteria. The basic
unit-membrane structure appears to be determined by the organization of the
lipopolysaccharide, since it is unaffected by proteolytic enzymes, and isolated
lipopolysaccharide has the same triple-layered structure (de Petris, 1967).
Chemical and other evidence suggests that both lipopolysaccharide and lipoprotein are
present at the surface, possibly in some form of mosaic structure (Weidel et al. i960;
Knox et al. 1966; de Petris, 1967).
(b) A wrinkled surface pattern has been observed in various negatively stained
Gram-negative bacteria (Thornley & Home, 1962; Ryter & Piechaud, 1963; Bladen &
Mergenhagen, 1964; Nermut, 1964; Zwillenberg, 1964; Bayer & Anderson, 1965;
Ueda & Takagi, 1965), in contrast to the smooth surface commonly seen in Grampositive organisms (Zwillenberg, 1964). The pattern has usually been interpreted as
indicating that the outer surface of the bacterium is wrinkled and grooved. The comparison of preparations of intact bacteria and isolated cell walls of Acinetobacter
strain 199 A has shown that in this organism the wrinkled material lies beneath the
outer membrane. Similar studies have not been made with other bacteria but it
seems likely that the wrinkled pattern may sometimes have a similar origin.
The wrinkled structure in Acinetobacter is readily destroyed by nearly all the treatments used, and so little evidence on its properties is available. It is seen inside the
pattern of pegs on the outer membrane, and outside the rigid layer in partially digested cell walls, and there seems little doubt that this material occupies the position
marked in Fig. 1. As several proteolytic enzymes remove the wrinkled pattern at
stages when the outer membrane and rigid layer are still intact, it seems likely that the
wrinkles are at least partly composed of protein. The protection of the wrinkles against
papain digestion by glutaraldehyde treatment also suggests the presence of protein.
290
M. J. Thornley and A. M. Glauert
Lysozyme would be expected to attack mucopeptide material only, and the fact that
only fragments of the other layers remain after this treatment is rather surprising. It is
possible that the outer membrane is weakened by the mechanical breakage involved
in the preparation of the cell walls, and that the wrinkled material is too fragile to remain
when not supported by the other layers. In another strain (239) of very similar
structure, only 33 % of the cell wall remains insoluble after lysozyme treatment, and
45 % after papain treatment, so that it is clear that one (or possibly both) of these enzymes releases into solution a quantity of material besides its own substrate. This was
also found by Work (1964) for cell walls of Micrococcus radiodurans.
The wrinkled layer of Acinetobacter 199 A occupies the same position in the cell
wall, just outside the rigid layer, as the protein granules described by Weidel et al.
(i960) in E. coli. Treatment of intact, briefly heated, E. coli cells with papain causes a
separation between the outer membrane and rigid layer (de Petris, 1967). The layer
destroyed was therefore in the same position as the wrinkled structure in Acinetobacter ; this site was thought by de Petris to contain the protein granules of Weidel
et al. (i960) and possibly other protein material also.
In Micrococcus radiodurans a ' compartment layer' is observed immediately outside
the rigid layer (Thornley et al. 1965) and this is also sensitive to proteolytic enzymes
(E. Work & H. Griffiths, personal communication). It may therefore be analogous to
the wrinkled layer although its structure appears more complex.
(c) The rigid layers, together with the thickened rings formed by ingrowing septa,
disappear on treatment with lysozyme. They can be isolated by repeated extraction
with concentrated SDS or by phenol extraction followed by papain digestion. The
insoluble residue after lysozyme treatment of cell walls contains no DAP, while the
preparation of rigid layers after phenol-papain treatment contains a greatly increased
proportion of glutamic acid, alanine and DAP, the amino acids characteristic of the
purified mucopeptide of other Gram-negative bacteria (Salton, 1964). There seems
no doubt, therefore, that the rigid layers shown by negative staining contain the
mucopeptide material. These were shown to correspond in sections to the darkly
stained inner region of the intermediate layer, which is immediately outside the
plasma membrane. This position is the same as that demonstrated for the lysozymesensitive material of other Gram-negative bacteria (Bladen & Mergenhagen, 1964;
Murray et al. 1965; de Petris, 1967).
The surfaces of the rigid layers as seen by negative staining appear to have a finely
granular texture with irregularly spaced black dots, up to about 100 A in diameter,
possibly representing pits or holes.
The rigid layers of E. coli have only been studied by metal-shadowing techniques
(Weidel et al. i960) which did not show any fine structure except for the surface array
of protein granules. In Micrococcus radiodurans the rigid layers in negatively stained
preparations appear to contain holes, also about 100 A in diameter. These penetrate
the whole thickness of the rigid layer giving it a striated appearance in sections
(Thornley et al. 1965; E. Work & H. Griffiths, personal communication), very different
from the appearance of the layer in the Acinetobacter strain.
Fine structure of an Acinetobacter
291
Correlation of structure with radiation resistance
The structure of different strains of Acinetobacter will be described and compared
fully in a subsequent paper. In summary, division by septation, as in 199 A, occurs in
the most radiation-resistant strains, while the most sensitive strains show constriction,
but in the intermediate range the correlation is not maintained. The wrinkled pattern
has been observed only in strains of intermediate and high resistance and so may
possibly be connected with resistance in some way. Further studies are in progress to
investigate whether DNA repair enzymes are important in the resistance of these
organisms, since it is now generally accepted that damage to DNA is primarily responsible for the inactivation of bacteria by u.v. and X-rays. Variations in the ability
to repair this damage seem the most likely cause of the wide range in resistance of
Acinetobacter strains, and the ability to repair DNA may well be influenced by the
extent of the changes in permeability induced by radiation and hence by the organization of the surface layers.
We are grateful to the International Atomic Energy Agency for a Research Contract held by
one of us (M.J.T.) at the Low Temperature Research Station, Cambridge, and to the Wellcome Trust for the loan of the AEI EM 6 B electron microscope. One of us (A. M. G.) is a
Sir Halley Stewart Research Fellow. We acknowledge the skilled technical assistance of Miss
K. Moden and Mr R. A. Parker, and are indebted to Mr D. F. Elsden for results obtained
with the Technicon autoanalyser.
REFERENCES
M. E. & ANDERSON, T.F. (1965). Thesuriacestructureoi Escherichiacoli. Proc.natn. Acad.
Sci. U.S.A. 54, 1592-1599.
BLADEN, H. A. & MERGENHAGEN, S. E. (1964). Infrastructure of Veillonella and morphological
correlation of an outer membrane with particles associated with endotoxic activity. J. Bad.
88, 1482-1492.
BRISOU, J. & PREVOT, A. R. (1954). fitude de systimatique bacte>ienne. X. Revision des
especes rdunies dans le genre Achromobacter. Annls Inst. Pasteur, Paris 86, 722-728.
CLAUS, G. W. & ROTH, L. E. (1964). Fine structure of the Gram-negative bacterium Acetobacter suboxydans. J. Cell Biol. 20, 217-233.
COHEN-BAZIRE, G. & KUNISAWA, R. (1963). The fine structure of Rhodospirillum rubrum.J. Cell
Biol. 16, 401-419.
COHEN-BAZIRE, G., PFENNIG, N. & KUNISAWA, R. (1964). The fine structure of green bacteria.
J. Cell Biol. 22, 207-225.
CONSDEN, R., GORDON, A. H. & MARTIN, A. J. P. (1944). Qualitative analysis of proteins: a
partition chromatographic method using paper. Biochem. J. 38, 224-232.
EAGON, R. G. & CARSON, K. J. (1965). Lysis of cell walls and intact cells of Pseudomonas
aeruginosa by ethylenediamine tetraacetic acid and by lysozyme. Can. J. Microbiol. 11,
BAYER,
193-201.
D. A. & WEINBAUM, G. (1967). Hexagonal pattern in cell walls of Escherichia coli B.
Science, N. Y. 155, 472-474.
FITZ-JAMES, P. C. (i960). Participation of the cytoplasmic membrane in the growth and spore
formation of bacilli. J. biophys. biochem. Cytol. 8, 507-528.
FITZ-JAMES, P. C. (1964). Thin sections of dividing Neisseria gonorrhoeae. J. Bad. 87, 14771482.
GHUYSEN, J. M. (i960). Acetylhexosamine compounds enzymically released from Micrococcus
lysodeikticus cell walls. II. Enzymic sensitivity of purified acetyl hexosamine and acetylhexosamine-peptide complexes. Biochim. biophys. Ada 40, 473-480.
GLAUERT, A. M. (1962). The fine structure of bacteria. Br. med. Bull. 18, 245-250.
FISCHMAN,
292
M. J. Thornley and A. M. Glauert
A. M. & THORNLEY, M. J. (1966). Glutaraldehyde fixation of Gram-negative
bacteria. Jl R. microsc. Soc. 85, 449-453.
JERMYN, M. A. & ISHERWOOD, F. A. (1949). Improved separation of sugars on the paper
partition chromatogram. Biochem. Jf. 44, 402-407.
KAY, D. & FILDES, P. (1950). The calcium requirement of a typhoid bacteriophage. Br. Jf. exp.
Path. 31, 338-348.
KELLENBERGER, E. & RYTER, A. (1958). Cell wall and cytoplasmic membrane of Escherichia
coli. J. biophys. biochem. Cytol. 4, 323-326.
KNOX, K. W., VESK, M. & WORK, E. (1966). Relation between excreted lipopolysaccharide
complexes and surface structures of a lysine-limited culture of Escherichia coli. Jf. Bad. 92,
GLAUERT,
1206-1217.
KRABBENHOFT,
K. L., ANDERSON, A. W. & ELLIKER, P. R. (1967). Influence of culture media
on the radiation resistance of Micrococcus radiodurans. Appl. Microbiol. 15, 178-185.
LEIVE, L. (1965 a). A nonspecific increase in permeability in Escherichia coli produced by EDTA.
Proc. natn. Acad. Sci. U.S.A. 53, 745-750.
LEIVE, L. (19656). Release of lipopolysaccharide by EDTA treatment of E. coli. Biochem.
biophys. Res. Commun. 21, 290-296.
MITCHELL, P. (1949). A new technique for stirred aerated culture. Nature, Lond. 164, 846847.
MURRAY, R. G. E. (1963). On the cell wall structure of Spirillum serpens. Can. Jf. Microbiol. 9,
381-392.
MURRAY, R. G. E., STEED, P. & ELSON, H. E. (1965). The location of the mucopeptide in
sections of the cell wall of Escherichia coli and other Gram-negative bacteria. Can. Jf. Microbiol. 11, 547-560.
MURRAY, R. G. E. & TRUANT, J. P. (1954). The morphology, cell structure and taxonomic
affinities of the Moraxella. Jf. Bad. 67, 13-22.
NERMUT, M. V. (1964). Bacterial surfaces as revealed by the negative staining method. Third
Reg. Conf. Electron Microsc. Prague 2, 525-526.
NOSSAL, P. M. (1953). A mechanical cell disintegrator. Aust. J. exp. Biol. vied. Sci. 31, 58359O.
ODA, T. & SEKI, S. (1966). Molecular basis of structure and function of the plasma membrane
of the microvilli of intestinal epithelial cells. Proc. 6th Int. Congr. Electron Microsc, Kyoto
(1966) z, 387-388.
PARTRIDGE, S. M. (1948). Filter-paper partition chromatography of sugars. I. General description and application to the qualitative analysis of sugars in apple juice, egg white and
foetal blood of sheep. Biochem. J. 42, 238-248.
PARTRIDGE, S. M. (1949). Aniline hydrogen phthalate as a spraying reagent for chromatography
of sugars. Nature, Lond. 164, 443.
PETRIS, S. DE (1965). Ultrastructure of the cell wall of Escherichia coli. Jf. Ultrastruct. Res. 12,
247-262.
PETRIS, S. DE (1967). Ultrastructure of the cell wall of Escherichia coli and chemical nature of its
constituent layers. Jf. Ultrastruct. Res. 19, 45-83.
PIECHAUD, M. (1961). Le groupe Moraxella. A propos des B 5W-Bacterium anitratum. Annls
Inst. Pasteur, Paris 100, suppl. to no. 6, 74-85.
POINDEXTER, J. L. S. & COHEN-BAZIRE, G. (1964). The fine structure of stalked bacteria
belonging to the family Caulobacteraceae. J. Cell Biol. 23, 587-607.
RYTER, A. & KELLENBERGER, E. (1958). Etude au microscope electronique de plasma contenant
de l'acide de'soxyribonucl&que I. Les nucltoides des bacte'ries en croissance active. Z.Naturf.
I3B, S97-6O5RYTER, A. & PIECHAUD, M. (1963). Etude au microscope e'lectronique de quelques souches de
Moraxella. Annls Inst. Pasteur, Paris 105, 1071—1079.
SALTON, M. R. J. (1953). Studies of the bacterial cell wall IV. The composition of the cell
walls of some Gram-positive and Gram-negative bacteria. Biochim. biophys. Ada 10, 512523SALTON, M. R. J. (1964). The Bacterial Cell Wall. Amsterdam: Elsevier.
SCHOCHER, A. J., BAYLEY, S. T. & WATSON, R. W. (1962). Composition of purified mucopeptide
from the wall of Aerobader cloacae. Can. Jf. Microbiol. 8, 89-98.
Fine structure of an Acinetobacter
293
STEED, P.
& MURRAY, R. G. E. (1966). The cell wall and cell division of Gram-negative bacteria.
Can. J. Microbiol. 12, 263-270.
THORNLEY, M. J. (1963). Radiation resistance among bacteria. J. appl. Bad. 26, 334-345.
THORNLEY, M. J. (1967). A taxonomic study of Acinetobacter and related genera. J. gen. Microbiol. 49, 211-257.
THORNLEY, M. J. & HORNE, R. W. (1962). Electron microscope observations on the structure of
fimbriae, with particular reference to Klebsiella strains, by the use of the negative staining
technique. X gen. Microbiol. 28, 51-56.
THORNLEY, M. J., HORNE, R. W. & GLAUERT, A. M. (1965). The fine structure of Micrococcus
radiodnrans. Arch. Microbiol. 51, 267-289.
THORNLEY, M. J., INGRAM, M. & BARNES, E. M. (i960). The effects of antibiotics and irradiation on the Pseudomonas-Achromobacterfloraof chilled poultry. J. appl. Bad. 23, 487498.
UEDA, M. & TAKAGI, A. (1965). Surface pattern of Fusobacterium polymorphum. Jap. J. Microbiol. 9, 145-148.
WEEKS, O. B. & BECK, S. M. (i960). Nutrition of Flavobacterium aquatile strain Taylor and a
microbiological assay for thiamine. J. gen. Microbiol. 23, 217-229.
WEIDEL, W., FRANK, H. & LEUTGEB, W. (1963). Autolytic enzymes as a source of error in the
preparation and study of Gram-negative cell walls. J. gen. Microbiol. 30, 127-130.
WEIDEL, W., FRANK, H. & MARTIN, H. H. (i960). The rigid layer of the cell wall of Escherichia
coli strain B. J. gen. Microbiol. 22, 158-166.
WILKINSON, S. G. (1967). The sensitivity of pseudomonads to ethylenediamine-tetra-acetic
acid. J. gen. Microbiol. 47, 67-^76.
WORK, E. (1964). Chemical structure of bacterial cell walls. Nature, Lond. 201, 1107-1109.
WORK, E., KNOX, K. W. & VESK, M. (1966). The chemistry and electron microscopy of an
extracellular lipopolysaccharide from Escherichia coli. Ann. N.Y. Acad. Sci. 133, 438-449.
ZWILLENBERG, L. O. (1964). Electron microscopic features of Gram-negative and Grampositive bacteria embedded in phosphotungstate. Antonie van Leeuwenhoek 30, 154-162.
{Received 3 August 1967)
M
294
- J- Thornley and A. M. Glauert
Fig. 2. Living cells growing in a chain. Phase contrast, x 2000.
Fig. 3. Living cells grouped in a tetrad. Phase contrast, x 2000.
Fig. 4. Thin section of a chain of cells from a culture in the logarithmic phase,
x 20000.
Fig. 5. Thin section of a tetrad from a culture in the stationary phase, x 20000.
Fig. 6. Negatively stained preparation of a pair of intact cells. Capsular material
appears dark due to its retention of phosphotungstate. The cell surfaces show wrinkles,
x 40000.
Fig. 7. Thin section of a dividing cell showing nuclear material («), ribosomes (rs),
septum (s) and plasma membrane (pm). x 60000.
Journal of Cell Science, Vol. 3, No. 2
rs
1
_
w««5fr«*7 ^-SUP*****
M. J. THORNLEY AND A. M. GLAUERT
(Facing p. 294)
Fig. 8. A dividing cell has a mesosome (m) associated with an ingrowing septum (s).
x 60000.
Fig. 9. A dividing cell showing constriction. The daughter cells are still connected by a
channel containing membranous material (w). x 60000.
Fig. 10. High magnification micrograph of the surface of a cell showing the outer
membrane (om) and the plasma membrane (p»i). Ryter—Kellenberger fixation,
x 120000.
Fig. 11. A dense line (r) is visible between the outer membrane (om) and plasma
membrane (pm) after glutaraldehyde-osmium fixation, x 120000.
Journal of Cell Science, Vol. 3, No. 2
s m
M. J. THORNLEY AND A. M. GLAUERT
Fig. 12. Thin section of an isolated cell wall. The outer membrane (ovi) and the intermediate dense line (r) are still associated, while the plasma membrane (pm) is detached,
x iooooo.
Fig. 13. Part of a cell wall with septum (s) and outer membrane (oni). x 60000.
Fig. 14. Negatively stained preparation of an isolated cell wall which has a wrinkled
appearance. There is a break in the upper surface. Where the wrinkles are absent,
there are indications of a periodic structure (arrow). A few fimbriae (/) are present,
x 60000.
Fig. 15. High-magnification micrograph of the surface of an isolated eel' wall. Peg-like
subunits are visible at the folded edge, x 120000.
Journal of Cell Science, Vol. 3, No. 2
15
M. J. THORNLEY AND A. M. GLAUERT
Fig. 16. Cell wall after treatment with papain. The peg-like subunits and wrinkled
pattern have disappeared, leaving an outer membrane and an inner layer (r) with an
associated septum (s). x 60000.
Journal of Cell Science, Vol. 3, No. 2
16
M. J. THORNLEY AND A. M. GLAUERT
Fig. 17. In another cell wall, papain has caused further breakdown of the outer
membrane to irregular tubes and vesicles, (r, inner layer.) x 60000.
Fig. 18. Fragments of the outer membrane of a papain-treated cell wall, almost completely separated from the inner layer (r) and its associated septum (s). x 60000.
Journal of Cell Science, Vol. 3, No. 2
M. J. THORNLEY AND A. M GLAUERT
Fig. 19. A papain-treated cell wall in which some of the wrinkled material (arrow)
remains, x 60000.
Fig. 20. Thin section of a papain-treated cell wall showing a thin outer and a thick
inner layer. Compare with Fig. 16. x 60000.
Figs. 21, 22. In some papain-treated cell walls, only the dense inner layers and their
associated septa (s) remain, x 60000.
Fig. 23. Negatively stained preparation of a pronase-treated cell wall, x 40000.
Journal of Cell Science, Vol. 3, No. 2
22
M. J. THORNLEY AND A. M. GLAUERT
Fig. 24. Some of the wrinkled material appears to be detached from the walls by
trypsin treatment, x 60000.
Fig. 25. Small vesicles with a granular surface pattern (arrow) remain after lysozome
digestion, x 60000.
Fig. 26. The granular pattern of the small vesicles is removed when lysozyme
treatment is followed by papain. x 60000.
Fig. 27. After steapsin treatment, masses of debris, probably originating from the
wrinkled material, are observed, x 40000.
Fig. 28. Clumps of small vesicles also remain after steapsin treatment, x 40000.
Journal of Cell Science. Vol. 3, No. 2
M. J. THORNLEY AND A. M. GLAUERT
Fig. 29. The walls have a rough appearance after treatment with dilute SDS. Aggregates of SDS are present, x 40000.
Fig. 30. A thin section of walls treated with dilute SDS shows that only the rigid
layers and some dense granular deposits remain, x40000.
Fig. 31. Repeated extraction with concentrated SDS leaves clean rigid layers, x 40000.
Fig. 32. Only the rigid layers remain after treatment with phenol followed by papain.
x 50000.
Journal of Cell Science, Vol. 3, No. 2
•\/i
0-1
32
M. J. THORNLEY AND A. M. GLAUERT
Fig. 33. Fixation by heat causes damage to the outer membrane and wrinkled layer,
x 40000.
Fig. 34. Incubation with EDTA in tris buffer appears to loosen the outer membrane.
x 50000.
Fig. 35. Incubation for 10 weeks in tris-EDTA causes further destruction of the outer
membrane, x 50000.
Fig. 36. Incubation for 31 weeks at 37 °C in water appears to remove the wrinkled
material. The outer membrane retains its granular appearance, x 50000.
Journal of Cell Science, Vol. 3, No. 2
36
M.J. THORNLEY AND A. M. GLAUERT