Download Dynamics of PhiX174 protein E-mediated lysis of

Arch Microbiol (1992) 157:381 388
Archives of
Hicrobiology
~@Springer-Verlag1992
Dynamics of PhiX174 protein E-mediated lysis of Escherichia coli
A. Witte 1, G. Wanner 2, M. Sulzner 1, and W. Lubitz ~
1 Institute of Microbiology and Genetics, University of Vienna, Althanstrasse 14, A-1090 Vienna, Austria
z Institute of Botany, University of Munich, Menzingerstrasse 68, W-8000 Munich 19, Federal Republic of Germany
Received August 30, 1991/Accepted December 12, 1991
Abstract. Expression of cloned gene E of bacteriophage
PhiX 174 induces lysis by formation of a transmembrane
tunnel structure in the cell envelope of Escherichia coli.
Ultrastructural studies of the location of the lysis tunnel
indicate that it is preferentially located at the septum or
at polar regions of the cell. Furthermore, the diameter
and shape of individual tunnel structures vary greatly
indicating that its structure is not rigid. Apparently, the
contours of individual lysis tunnels are determined by
enlarged meshes in the peptidoglycan net and the force
produced at its orifice, by the outflow of cytoplasmic
content. Once the tunnel is formed the driving force for
the lysis process is the osmotic pressure difference between cytoplasm and medium. During the lysis process
areas of the cytoplasmic membrane which are not tightly
attached to the envelope are extended inward by the
negative pressure produced during lysis. After cell lysis
external medium can diffuse through the lysis tunnel
filling the inner cell space of the still rigid bacterial ghosts.
Key words: PhiX174 - Bacterial lysis - Escherichia coli
- Electron microscopy - Membranes - Cell envelope
Initial studies of PhiX174 mediated lysis of E. coli gave
no indications of phage encoded murolytic enzymes
(Eigner et al. 1963; Markert and Zilling 1965). Genetic
analysis of phage mutants (Hutchison and Sinsheimer
1966) identified a single lysis gene of the phage, E, and
it was shown that its expression alone was sufficient to
cause bacterial lysis (Henrich et al. 1982; Young and
Young 1982). D N A sequence analysis of gene E suggests
that it codes for a hydrophobic protein of 91 amino acids
(Barrell et al. 1976). Protein E was detected in the inner
and outer membrane fractions where it has the ability
to oligomerize (Blfisi et al. 1983, 1989; Altman et al.
1985). Based on these facts and on the structural analysis
of protein E it was postulated that protein E is a
Offprint requests to." W. Lubitz
membrane protein which causes the formation of a
transmembrane tunnel structure (Witte and Lubitz 1989;
Witte et al. 1990b).
Protein E integrates into the cell envelope of E. eoli
and exerts its lytic effect by a process which is dependent
on the proton-motive-force of the cells (Witte et al. 1987).
Other cellular factors for E-lysis are its dependence on
the growth phase of the cells as well as on the regulation
of the cells autolytic system (Lubitz et al. 1984a, b). We
have recently shown by electron microscopy that E-lysed
cells show discrete holes in their cell envelopes (Witte et
al. 1990a). Here we extend the electron microscopic
studies and show a preferential location of the lysis tunnel
in potential division zones of the cells. Questions which
remain to be answered are the number and arrangement
of protein E subunits within the tunnel, and whether the
lysis tunnel is solely formed by protein E molecules or
in concert with cellular proteins, or is indirectly a
consequence of local membrane disturbance.
In this communication electron microscopic evidence
is provided that suggest functions of the cell division
machinery may play an important role in E-mediated
lysis. The morphological studies indicate that single lysis
pores appear either in the middle or polar regions of the
cell, areas of the envelope actively involved in cell division
(MacAlister et al. 1983, 1987; de Boer et al. 1990). In
addition, the electron microscopic studies presented
provide a better understanding of the driving force for
the lysis process. The present study is in agreement with
biochemical studies of E-mediated lysis which indicated
that energetic and permeability properties of the inner
membrane change simultaneously with the onset of lysis
(Witte and Lubitz 1989).
Materials and methods
Bacteria, plasmids and growth conditions
Escherichia coli PC1363 (E. coli C wild-type; Phabagen Collection,
Utrecht) as well as plasmids pSB12, pSB22 (Blfisiet al. 1985) and
pci857 (Remaut et al. 1983) have previously been described.
Plasmids pSB12 and pSB22 carry gene E under transcriptional
382
control of the lambda pL promoter. The repressor allele ci857 was
plasmid-encoded and provided sufficient repressor activities to keep
gene E expression silent during growth at 28 °C. Expression of gene
E from the lambda promoter was induced by thermal inactivation
of the ci857 repressor molecules at 42 °C. E. eoli strain PC1363 was
grown with aeration in Luria broth containing, 10 g/1 tryptone, 5 g/l
yeast extract and 5 g/l NaC1. For strains harbouring pci857 and
pSB12, or pci857 and pSB22, ampicillin (200 gg/ml) and kanamycin
(50 gg/ml) were added to the medium to maintain selection. Growth
and lysis of the culture samples were monitored spectrophotometrically be measuring the optical density at 600 nm.
Determination of cellular water and sucrose space
Internal water space and uptake of 14C-sucrose were determined
as described by Rottenberg (1979). To remove glucose impurities,
14C-sucrose was preincubated with E. eoli cells. 3H-H20 and
14C-sucrose were added to culture samples of E. eoli 10 rain prior
to induction of E-mediated lysis at final specificactivities of 2 gCi/ml
to achieve full equilibration of intra- and extracellular 3H-H20
concentration and diffusion of 14C-sucrose through the outer
membrane of the cells, respectively. At various time points, l-ml
culture aliquotes were layered onto silicone oil and cells were
sedimented. For calculation of the total waterspace, radioactivity
in cell pellet and supernatant fractions were determined. For the
estimation of the intracellular 3H-H20, 0.3 mg (dry weight) was
taken to be equivalent to an optical density, OD6oo, of 1. The
external water space (sucrose space) was calculated in percent of
the total water space.
Membrane A TPase activity
Membrane bound ATPase activity of E-lysed E. coli PC1363 was
measured as described by Fillingame and Foster (1986). Cells were
disrupted by using a French press and unbroken cells were removed
by centrifugation at 6000g for 20 rain. The membrane fraction was
then collected by ultracentrifugation as described.
Transmission and high resolution scanning electron microscopy
Transmission electron micrographs were taken with an Elmiscop
101 (Siemens Munich, FRG) electron microscope and scanning
electron micrographs were taken with a Hitachi S-800 field emission
scanning electron microscope (Hitachi, Tokyo, Japan). Fixation of
cells and preparation for electron microscopy was essentially the
same as previously described (Witte et al. 1990a).
Results
The observations that in rich medium the rate of Emediated lysis corresponds to the growth rate of Escherichia coli (Lubitz et al. 1984a), and that stationary cells
or minicells display a lysis-negative phenotype (Lubitz
et al. 1984a; Blfisi et al. 1984, 1985) can be interpreted
that cellular activities involved in cell division might also
be essential for lysis. Inspection of approximately five
hundred cells of which the lysis tunnel was clearly visible
by scanning or transmission electron microscopy show
that the location of the transmembrane lysis tunnel of
more than 90% o f all cells is at or near the potential
division zones at the center or poles of the cells (Fig.
1 A - D ) . As illustrated in Fig. 1, the lysis tunnel is
characterized by a small hole through the envelope.
Sealing of the periplasmic space is achieved by fusion of
the inner and outer membranes (Fig. 1 B). The predominant location of the E-mediated lysis tunnel is in the
middle of E. coli cells (Fig. 1 B and 1 D) but in about one
third oflysed cells it is located near 1/4 and 3/4 cell length
(Fig. 1C). In contrast to earlier determinations of the
diameter of the lysis tunnels between 40 and 80 nm (Witte
et al. 1990a), the inspection of several hundreds of cells
indicates that the tunnel diameters fluctuate between
border values of 40 and 200 nm (Fig. 1 C and D). These
variations in the diameter as well as of the shape of
different lysis tunnels indicate that no regular structure,
such as a defined cylinder is formed during lysis.
It seems likely that the size of the lysis tunnel is
influenced by individual differences o f the cells regarding
the extent of local autolysis and consequently o f particular meshes within the peptidoglycan net. Whereas smaller
orifices of the lysis tunnels can be detected at early stages
of cell lysis larger tunnel structures can be seen more
frequently in samples taken at late stages of lysis. This
indicates that secondary effects occurring in the envelope
of individual cells after lysis also influence the shape of
the tunnel. One candidate activity contributing to this
effect might be the induction of phospholipases in the
envelope complex by protein E (Lubitz and Pugsley
1985). It should, however, be emphasized that E-lysis
does not change or destroy the overall structure of the
envelope except within a small area in nanometer range
dimensions (Fig. 1).
After formation of the transmembrane tunnel structure cytoplasmic material is released very rapidly from
the cells. Evidence for the suggestion that the lysis process
itself is extremly fast comes from the finding that liberated
chromosomal D N A is sheared almost completely to a
uniform size class of 40 to 50 kb whereas plasmid D N A
is preserved in its supercoiled form (data not shown).
This suggests that the driving force for the release of
cytoplasmic material is the osmotic pressure difference
between the cytoplasm and the medium created by the
opening of the tunnel structure. Osmotic protection
experiments with PhiX 174 infected E. coli (Markert and
Zillig 1965) or gene E expression from plasmids (Pocta
and Lubitz, unpublished) showed that addition of 20%
sucrose to the growth medium inhibits phage release and
E-mediated lysis.
Electron microscopic inspection of more than one
thousand lysed cells prepared as ultrathin sections provides additional insight into the dynamics of E-mediated
lysis. Three representative types of ghosts including more
than 90% of the cells inspected by serial sections and
transmission electron microscopy, are summarized in the
reconstruction drawings given in Fig. 2. As the figure
illustrates, the loss of cytoplasm appears to produce a
negative pressure within the lysing cell. In most cases the
inner membrane is pulled inwards where it is not firmly
associated with the envelope complex. Evidence suggests
there are numerous areas where the cytoplasmic membrane is fixed to the envelope complex such that the inner
membrane does not detach completely from the rigid
383
Fig. 1 A - D . E-mediated lysis of E. coli PC1363 (pci857, pSB12). A
and B Transmission electron micrographs of ultrathin sections of
lysed ceils. A The inner (ira) and outer (ore) membranes are
indicated by arrows. Efftux of cytoplasmic material is indicated by
the open arrow. B The location of the lysis tunnel in the division
zone is indicated by an arrow. The inner and outer membranes are
continuous at the borders of the lysis tunnel orifice (arrow). Other
lysed cells visible in this picture typically show detachment of the
inner membrane from the poles. C and D High resolution field
emission scanning electron micrographs of lysed cells. C E-specific
lysis tunnel near the pole region of E. coli. D E-specific lysis tunnel
at the central division zone
384
m
PP
im
P
Fig. 2. Time course of E-mediated lysis of E. coll. Dynamic reconstructions of different types of lysis from serial ultrathin sections of
lysed E. coli cells inspected by transmission electron microscopy.
The cytoplasma (cp), inner membrane (irn), periplasmic space
(pp) and outer membrane (ore) of schematic E. coli cell are
indicated. Formation of the E-specificlysistunnel either in the middle
of the cell or at the polar region is indicated (short dark arrow,
second row). Expulsion of cytoplasm through the E-specific lysis
tunnel (open arrow, third row) and inward bending of the inner
membrane during the course of cell lysis is depicted by thin dark
arrows (first and second panel, third row). Left panel." This figure
shows expansion of the inner membrane starting at a small area in
the middle of the cell to compensate for the negative pressure
produced by the outstreaming cytoplasma. Middle panel: detachment of the inner membrane from the polar sites of the cell to the
central area. Right panel. minor inner membrane detachment from
the envelope complex
peptidog!ycan/outer membrane complex (Fig. 2). In those cells where the lysis tunnel occurs at the midpoint of
the cells, detachment of the inner membrane is often seen
at both polar sites of the cell (Fig. 2, middle pannel).
However, inward bending of cytoplasmic membrane
areas can occur at various places starting apparently at
areas of weak attachment to the envelope complex
(Fig. 2, left pannel). In about 15% of lysed cells minor
inward bendings of the inner membrane were also
detected (Fig. 2, right pannel). In cells where the inner
membrane has detached from the poles, very often a
triple layer membrane complex can be seen in cross
sections of bacterial cylinders (Fig. 3). The well-preserved
layers seen in Fig. 3, made it possible to determine the
dimensions of the outer and inner membranes to be
approximately 6 nm each. They border the periplasmic
space which has a width of about 10 nm with stained
peptidoglycan of approximately 3 nm (Fig. 3). These
dimensions indicate that E-mediated lysis does not alter
the structures of the envelope complex as the measurements of the different envelope components reported here
correspond well to other determinations (Costerton et al.
1974; Kellenberger 1990).
Measurements of membrane-bound ATPase activities
of E. coli ghosts after E-mediated lysis indicate very high
specific activities of the enzyme compared to membrane
fragments produced by French press disrupted cells
(Tab. 1). The high specific activity of the membrane
bound ATPase of E-lysed cells (bacterial ghosts) indicates that the inner membrane structure is well preserved after E-lysis and that substrates are accessible to
membrane enzymes through the E-lysis tunnel. This
result also indicates that the envelope structures, including the inner membrane of E-lysed cells, are well
conserved in their native structure. Bacterial ghosts
obtained after E-mediated lysis may represent a new tool
for studies of membrane-bound enzyme activities which
previously have been performed with inside-out vesicles
(Mfiller and Blobel 1984). Other investigations have also
shown that larger particles such as antibodies, can diffuse
into the interior cell space and bind to specific receptor
sites on the inside of the inner membrane (Szostak et al.
1990). Thus, E-lysis might be an alternative method for
cell disruption which carefully preserves the integrity of
the cytoplasmic membrane.
The augmented specific activity of the membrane
ATPase of E-lysed cells as compared to French press
disrupted envelope fragments (Table 1), most probably
reflects the intact nature of the inner membrane in
bacterial ghosts. If the inner membrane is indeed unim-
385
Fig. 3. Transmission electron micrograph of an
ultrathin cross section through the cylindricalpart
of lysed E. coli PC1363 (pci857, pSB12) with
detached inner membrane from the pole region.
The outer (ore) and inner (ira) membranes
(~)
and the peptidoglycan layer ( . . . . . . )
are indicated. The innermost membrane represents an inverted inner membrane due to the
negative pressure produced by the lysis process
(for illustration see Fig. 1B and Fig. 2, middle
panel). The inset at the right top corner givesthe
entire cross section at lower magnification
Table 1. Specific activities of membrane bound ATPase of E-lysed
E. coli PC1363 and envelope fragments produced by French press.
The values given are the average of three independent determinations
Sample
Specific activity
~tM POJmg protein/min
Bacterial ghosts
Envelope fragments
Whole cells
0,38 +_ 0,06
0,16 _+ 0,01
0,04 __ 0,02
paired, the lysis tunnel itself should be the only route of
passage for solutes which cannot cross the inner membrane. To test this assumption, the influx of sucrose into
E-lysed cells as well as the internal and external water
space of the cells before and after lysis were determined
monitoring the distribution of 14C-sucrose and 3H-H20
in intact and lysed cells (Fig. 4). In intact cells tritiated
water should be uniformly distributed within the cell and
in the growth medium whereas 14C-sucrose should be
excluded from the interior cell volume. With the release
of cytoplasmic material the internal water space should
expand and sucrose should be able to diffuse into the
bacterial ghosts. The difference in the kinetics of 3H-H/O
and 14C-sucrose influx (Fig. 4) is interpretated to indicate
that sucrose can only diffuse through the tunnel structure
into the ghosts, whereas water can also cross the inner
membrane barrier. The onset of lysis of E. coli PC1363
harbouring pSB12 started approximately 17 rain after
induction of gene E expression (Fig. 4). Under the same
conditions, mild growth retardation occurred in the control due to the dilution effect caused by the addition of
14C-sucrose and 3H-water (Fig. 4A). Influx of measurable
amounts of labelled sucrose into the lysed bacteria could
be observed with a delay of a five minutes after lysis
(Fig. 4B and C). The experimental conditions were such
that 10 rain prior to induction of gene E expression (time
0 min, Fig. 4C) 3H-H20 and ~4C-sucrose were added to
achieve full equilibration of intra- and extracellular
3H-H20 and to provide sufficient time for 14C-sucrose
to penetrate the outer membrane and equipoise the
periplasmic space. As sucrose cannot pass the inner
membrane of E. coli, sucrose is often used to compensate
the cytoplasmic tonicity (Osborn and Munson 1974).
However, the concentrations of sucrose used for the
determinations of the sucrose space were below that
required for osmotic stabilization and thus had no effect
on E-mediated lysis (Fig. 4). The constant values of the
water and sucrose space seen in Fig. 4 between the time
point of temperature upshift for gene E induction (time
0min) and onset of E-mediated lysis (time 17 min)
indicate that the system was in full equilibrium. Since
3H-H20 was taken up faster than 14C-sucrose by the
ghosts, it can be concluded that part of the water filling
the interior of lysed cells was aquired during the lysis
process from diffusion through the inner membrane. This
indicates that the cytoplasmic membrane not involved in
lysis tunnel formation remains intact during expulsion
of cytoplasmic material. The remaining water inside the
ghosts originates from the influx of water through the
tunnel structure after cell lysis. Because sucrose cannot
diffuse through the inner membrane, all labeled sucrose
within the ghosts is taken up through the tunnel structure.
After cell lysis, membrane movement of the inwardly
bent cytoplasmic membrane back to the rigid envelope
(Fig. 2) would greatly contribute to the filling of the
internal cell space, and uptake of 3H-H20 and 14Csucrose should occur with almost the same kinetics.
However, the extended lag phase of 14C-sucrose influx
versus influx of aH-H20 in the ghosts strongly suggests
that diffusion plays the major part in this process and
that membrane areas detached from the envelope complex do not move back to their original position after
lysis. On the other hand, the time difference seen can
also be interpretated to incidate that during the course
of E-mediated cell lysis the inner membrane remains
intact with the exception of the area of tunnel formation.
386
1,3 - A
1,o
0,8
o
o
¢D
Q
o
0,6
0,5
0,4
0,3
I
i
I
I
I
I
I
I 5
I
I
B
E
\
v
¢1
O..
I--
C
00
o.
80
60
:3
I
I
I
I
I
40
I
I
0
10
'mio
Fig. 4 A - C . Exchange of cytoplasmic material with extracellular
medium during E-mediated lysis of E. coli. A Induction of gene E
expression of strain PC1363 (pci857, pSB12) (o
e), and strain
PC1363 (pci857, pSB22) (©- ©), by temperature upshift of the
cultures from 28 °C to 42 °C (arrow,0 rain). The dotted lines give
the time of lysis onset of strain PC1363 (pci857, pSB12) as
monitored by measurement of culture turbidity. B Water space in
gl/mg dry weight of cells. The water space givencorresponds to the
internal water space of the cellsplus the waterbound at the hydrated
cell surface.The external water bound to the cells (58% of the total
water space) corresponds to the sucrose space of cells which in C
is given as % sucrose space of the total water space. The internal
water space of intact cells corresponds to 2 gl/mg dry weight of
cells. Cells were seperated from the supernatant by sedimentation
through silicon-oil,radioactivitywas determinedfrom the sediment
fraction.
Discussion
Cellular prerequisites of E-mediated lysis and its dynamic
process will be discussed in context with our current
model of this process. Several lines of evidence suggest
that E-mediated lysis depends on the activity of the
cellular autolytic system (Lubitz and Plapp 1980; Lubitz
et al. 1984a, b; B1/isi et al. 1984). Regulatory functions
of the autolytic systems, for example, the lytA gene
product (Lubitz et al. 1984b), seem to be more important
for E-mediated lysis than single impaired functions of
the peptidoglycan metabolism (Halfmann et al. 1984;
Halfmann and Lubitz 1986). E. coli mutants with altered
sensitivities towards penicillin, moenomycin or EDTA
(Halfmann et al. 1984; Halfmann and Lubitz 1986) or
with defects in penicillin-binding proteins (Witte et al.
1990b) show no altered phenotype in E-mediated lysis.
Thus, not every cellular activity which harms the integrity
of the cell envelope or is involved in autolysis affects
E-mediated lysis.
The electron microscopic findings that the E-specific
lysis tunnel is predominantly located in areas of potential
cell division sites indicate that cell division activities could
be key functions in the pathway of E-mediated lysis.
Investigation of this postulated dependence is in progress.
Preliminary results have shown that the cell division genes
f t s Z andftsA (for recent reviews see de Boer et al. 1990;
Luktenhaus 1990) affect E-mediated lysis (Brand E,
Witte A, Lubitz W, unpublished). Further support for
the idea that theftsZ gene product (a regulator of division
processes; Luktenhaus et al. 1980; Robin et al. 1990),
affects E-mediated lysis comes from the observation that
after induction of gene E expression, lysis of an exponentially growing culture occurs only gradually and not
simultaneously, e.g. phage lambda mediated lysis (Garrett et al. 1981). The rate of E-lysis under such conditions
corresponds to the growth rate of the bacteria (Lubitz
etal. 1984a). Addition of chloramphenicol and/or rifampicin to such cultures showed that short and permanent expression of gene E resulted in the same kinetics
of culture lysis (Witte et al. 1987). This indicates that the
supply of a critical concentration of protein E is not the
rate limiting step for lysis, but rather, cellular factors
determine the time point of cell lysis. In PhiX 174 infection
of E. coli, lysis timing is additionally influenced by the
gen K product of the phage (B1/isi et al. 1988).
One further possibility of the dependence of Emediated lysis on functions involved in cell division
should be discussed here. It is possible that specific
proteins of the cell division machinery are targeted by
protein E, either as a nucleus for tunnel formation or as
essential building blocks of its structure. On the other
hand, it is also feasible that protein E interacts with cell
division in such a way that enlarged openings in the
peptidoglycan cannot be sealed quickly enough by the
ingrowing septation process. Cell lysis could then result
from local disturbance of the rigid envelope structure.
Local peptidoglycan hydrolysis preceeds E-specific cell
lysis (Lubnitz and Plapp 1980). However, the degradation
process seen before lysis is very limited and does not
account for more than 8% of the total peptidoglycan
(Witte 1990). As the structure of the sacculus is not
affected by this limited degradation, we interpret this
activity as a local disintegration of peptidoglycan. From
current estimates of the size distribution of meshes in the
sacculus the existence of pores larger than 20 nm is
excluded (Kellenberger 1990). However, as shown by
electron microscopy oflysis tunnel structures, larger holes
in the peptidoglycan net are necessary (Fig. 1D). As
previously mentioned, other lines of evidence suggest
cooperation of protein E with regulatory factors of the
autolytic system. It is therefore highly likely that processes
associated with cell division are responsible for the local
387
and limited degradation of peptidoglycan. The electron
microscopic investigation of E-lysed cells presented here
indicate that more than 90% of the E-specific lysis
structures are associated with potential cell division sites.
The previously mentioned dependence of E-mediated
lysis on autolytic processes can now be more restricted
to activities involved in cell division.
The determinations of the enzymatic activity of membrane bound ATPase (Table 1) and of sucrose influx into
bacterial ghosts (Fig. 4) support the suggestion of a
largely intact inner membrane after E-lysis of E. coli. It
should be further emphasized that the ultrathin sections
of E-lysed cells (Fig. 1 and 2) show that there are
extended areas of inner membrane associations with the
envelope complex. These zones of adhesion (Bayer 1968)
roughly correspond to the regions determined by Cook
etal. (1986) who plasmolyzed E, coli by hypertonic
sucrose solutions. The possibility that inner membrane
detached from the cell envelope by E-lysis moves back
to the envelope complex and thus contributes to an influx
of external medium could not be supported by measurements of the distribution of 3 H - H 2 0 and 14C-sucrose
during E-mediated lysis. The serial ultrathin sections
(Fig. 2) show that at certain sites the cytoplasmic membrane bulges into the interior cell space of E-lysed cell
while remaining attached to the envelope complex by
areas of adhesion. This electron microscopic evidence
seems to be a strong support for the existence o f such
areas, the existence o f which have been recently questioned (Kellenberger 1990).
A direct measurement of the time course of single cell
lysis is still missing. F r o m the observations presented, it
is assumed that single cell lysis is faster than the time
required for water to diffuse into the cytoplasmic space
to fill this area. Using filtration experiments, Hutchison
and Sinsheimer (1963) estimated the time for single cell
lysis by PhiX174 to be less than 30 s. The rate limiting
step in these experiments was the time required to collect
different samples. From the inward bending of inner
membrane material during E-lysis (Fig. 2), it is estimated
that single cell lysis is much faster, However, direct
experimental data are needed to confirm this assumption.
Acknowledgements. We are grateful to Dr. E. P. Bakker to provide
advice, material and laboratory space to experiments defining the
internal water space of the ghosts. We appreciate the excellent
technical assistance of S. Reese and of S. Schoy and the artwork
done by J. Seifert. We are grateful to Drs. D. Dennis, U. Blfisi and
K. Tedin for critical reading of the manuscript. This work was
supported by grants from the Austrian Fonds zur F6rderung der
wissenschaftlichen Forschung (P6861 Bio) and Deutsche Forschungsgemeinschaft (SFB 145).
References
Altman E, Young KD, Garret J, Altman R, Young R (1985)
Subcellular location of lethal lysis proteins of bacteriophages
lambda and PhiX 174. J Virol 53 : 1008-1011
Barrel BG, Air GM, Hutchison III CA (1976) Overlapping genes
in bacteriophage PhiX174. Nature 264:34-41
Bayer ME (1968) Areas of adhesion between wall and membrane
of Escherichia coli. J Gen Microbiol 53 : 395-404
B1/isi U, Geisen R, Lubitz W, Henrich B, Plapp R (1983) Localization of the bacteriophage PhiX174 lysis gene product in the cell
envelope of Escherichia coli. In: Hakenbeck R, H61tje JV,
Labischinski H (eds) Target of penicillin, de Gruyter, Berlin
New York, pp 205-210
B1/isi U, Halfmann G, Lubitz W (1984) Induction of autolysis of
Escherichia coli by PhiX174 gene E product. In: Nombela C
(ed) Microbial cell wall synthesis and autolysis. Elsevier, Amsterdam New York, pp 213-218
B1/isi U, Henrich B, Lubitz W (1985) Lysis of Escherichia coli by
cloned PhiX174 gene E depends on its expression. J Gen
Microbiol 131 : 1007-1114
B1/isi U, Young R, Lubitz W (1988) Evaluation of the interaction
of PhiX174 gene products E and K in E-mediated lysis of
Escherichia coll. J Virol 62:4362-4364
B1/isi U, Linke RP, Lubitz W (1989) Evidence for membrane-bound
oligomerization of bacterophage PhiX 174 lysis protein E. J Biol
Chem 264:4552-4558
Cook WR, MacAlister TJ, Rothfield LI (1986) Compartimentalization of the periplasmic space at division sites in Gram-negative
bacteria. J Bacteriol 168:1430-1438
Costerton JW, Ingrain JM, Cheng KJ (1974) Structure and function
of the cell envelope of Gram-negative bacteria. Bacteriol Rev
38:87 110
de Boer PAJ, Cook WR, Rothfield LI (1990) Bacterial cell division.
Ann Rev Genet 24:249-274
Eigner J, Stouthamer AH, Van der Sluys J, Cohen JA (1963) A
study of the 70S component of bacteriophage PhiX174. J Mol
Biol 6:61-84
Fillingame RH, Foster DL (1986) Purification of F 1Foil +-ATPase
from Escherichia coll. Methods Enzymol 126:545-557
Garrett J, Fusselman R, Hise J, Chio L, Smith-Grillo D, Schulz J,
Young R (1981) Cell lysis by induction of cloned Lambda lysis
genes. Mol Gen Genet 181:326-331
Halfmann G, Leduc M, Lubitz W (1984) Different sensitivity of
autolytic deficient Escheriehia coli mutants to the mode of
induction. FEMS Microbiol Lett 24:205-208
Halfmann G, Lubitz W (1986) Different induction of Escherichia
coli autolysis by penicillin and the bacteriophage PhiX174 gene
E product. J Bacteriol 66:683-685
Henrich B, Lubitz W, Plapp R (1982) Lysis of Escherichia coli by
induction of cloned PhiX174 genes. Mol Gen Genet 185:
493-497
Hutchison III CA, Sinsheimer RL (1963) Kinetics of bacteriophage
release by single cells of PhiX174-infected E. coll. J Mol Biol
7 : 206-208
Hutchison III CA, Sinsheimer RL (1966) The process of infection
with bacteriophage PhiX174. Mutations in a PhiX174 lysis gene.
J Mol Biol 18:429-447
Kellenberger E (1990) The 'Bayer bridges' confrontated with results
from improved electron microscopy methods. Mol Microbiol
4:697-705
Lubitz W, Plapp R (1980) Murein degradation in Eseherichia
coli infected with bacteriophage PhiX174. Curr Microbiol 4:
301-304
Lubitz W, Halfmann G, Plapp R (1984a) Lysis of Escherichia coli
after infection with PhiX174 depends on the regulation of the
cellular autolytic system. J Gen Microbiol 130:1079-1087
Lubitz W, Harkness RE, Ishiguro EE (1984b) Requirement for a
functional host cell autolytic system for lysis of Escherichia coli
by bacteriophage PhiX174. J Bacteriol 159:385-387
Lubitz W, PugsleyAP (1985) Changes in cell phospholipid composition of PhiX174 gene E product. FEMS Microbiol Lett 30:
171-175
Lutkenhaus JF, Wolf-Watz H, Donachie WD (1980) Organization
of genes in the ftsA-envA region of the Eseherichia coli genetic
map and identification of a new fts locus (ftsZ). J Bacteriol 142:
615-620
388
Lutkenhaus J (1990) Regulation of cell division in E. coli. Trends
Genet 6:22-25
MacAlister TJ, MacDonald B, Rothfield LI (1983) The periseptal
annulus: an organelle associated with cell division in Gramnegative bacteria. Proc Natl Acad Sci USA 80:1372-1376
MacAlister TJ, Coo WR, Weigand R, Rothfield LI (1987) Membrane-murein attachment at the leading edge of the division
septum: a second membrane-murein structure associated with
morphogenesis of the Gram-negative bacterial division septum.
J Bacteriol 169:3945-3951
Markert A, Zillig W (1965) Studies on the lysis of Escherichia coli
by bacteriophage PhiX174. Virology 25:88-97
Mfiller M, Blobel G (1984) In vitro translocation of bacterial
proteins across the plasma membrane of Escherichia coli. Proc
Natl Acad Sci USA 81:7421-7425
Osborn M J, Munson R (1974) Separation of the inner (cytoplasmic)
and outer membranes of gram negative bacteria. Methods
Enzymol 31A: 642-653
Remaut E, Tsao H, Fiers W (1983) Improved plasmid vectors with
a thermoinducible expression and temperature regulated runaway replication. Gene 22:103-113
Robin A, Joseleau-Petit D, D'Ari R (1990) Transcription of the
ftsZ gene and cell division in Escherichia coli. J Bacteriol 172:
1392-1399
Rottenberg H (1979) The measurement of membrane potential and
pH in cells, organells and vesicles. Methods Enzymol LV:
547-569
Szostak M, Wanner G, Lubitz W (1990) Recombinant bacterial
ghosts as vaccines. Res Microbiol 141:1005-1007
Witte A (1990) Untersuchungen zur Wirkung von PhiX 174 Protein
E sowie davon abgeleiteter chim~irer Proteine auf den Zellwandkomplex von Escherichia coli. PhD-qhesis, University of Munich, Faculty of Science
Witte A, Lubitz W, Bakker EP (1987) Proton-motive-force-dependent step in the pathway of lysis of Escherichia coli induced
by bacteriophage PhiX174 gene E product. J Bacteriol 169:
1750-1752
Witte A, Lubitz W (1989) Biochemical characterization of PhiX 174protein-E-mediated lysis of Escherichia coll. Eur J Biochem 180:
393-398
Witte A, Wanner G, B1/isiU, Halfmann G, Szostak M, Lubitz W
(1990a) Endogenous transmembrane tunnel formation mediated by PhiX174 protein E. J Bacteriol 172:4109-4114
Witte A, B1/isiU, Halfmann G, Szostak M, Wanner G, Lubitz W
(1990b) PhiX174 protein E-mediated lysis of Escherichia coli.
Biochimie 72:191-200
Young KD, Young R (1982) Lytic action of cloned PhiX174 gene
E. J Virol 44:993-1002