Department of Biology, Virginia Polytechnic
Institute and State University, Blacksburg, Virginia 24061
A major structural element of bacterial endospores is a
peptidoglycan (PG) wall. This wall is produced between the two opposed membranes surrounding the developing forespore and is composed of two
layers. The inner layer is the germ cell wall, which appears to have a
structure similar to that of the vegetative cell wall and which serves
as the initial cell wall following spore germination. The outer layer,
the cortex, has a modified structure, is required for maintenance of
spore dehydration, and is degraded during spore germination. Theories
suggest that the spore PG may also play a mechanical role in the
attainment of spore dehydration. Inherent in one of these models is the
production of a gradient of cross-linking across the span of the spore
PG. We report analyses of the structure of PG found within immature,
developing Bacillus subtilis forespores. The germ cell wall
PG is synthesized first, followed by the cortex PG. The germ cell wall
is relatively highly cross-linked. The degree of PG cross-linking drops
rapidly during synthesis of the first layers of cortex PG and then
increases two- to eightfold across the span of the outer 70% of the
cortex. Analyses of forespore PG synthesis in mutant strains reveal
that some strains that lack this gradient of cross-linking are able to
achieve normal spore core dehydration. We conclude that spore PG with
cross-linking within a broad range is able to maintain, and possibly to
participate in, spore core dehydration. Our data indicate that the
degree of spore PG cross-linking may have a more direct impact on the rate of spore germination and outgrowth.
 |
INTRODUCTION |
Sporulation by certain gram-positive
bacteria, such as Bacillus and Clostridium spp.,
results in the formation of a metabolically dormant cell known as an
endospore, which is resistant to severe physical and chemical
conditions. Sporulation involves an asymmetric septation to produce the
mother cell, which will contribute several components of the mature
spore, and the forespore, which will develop into the mature spore.
Engulfment of the smaller forespore by the larger mother cell
leaves the forespore surrounded by two membranes. Spore
peptidoglycan (PG) synthesis occurs within this intermembrane space.
Spore PG is clearly required for the maintenance of spore core
dehydration, the major factor determining spore heat resistance
(4, 19, 26). It has also been suggested that spore PG may
play a direct role in the attainment of spore core dehydration
(17, 23, 34).
Electron microscopy suggests that spore PG consists of two layers: a
thin layer adjacent to the inner forespore membrane, called the germ
cell wall, and a thicker outer layer, termed the cortex. The structure
of the germ cell wall is believed to be similar to that of vegetative
PG (33). Functionally, the germ cell wall is defined as the
initial cell wall following spore germination and is possibly a
template for the synthesis of vegetative PG. It has been suggested that
germ cell wall PG is synthesized from precursors made in the forespore,
but this has not been clearly demonstrated (33). Cortex PG
is synthesized from precursors that are made in the mother cell and
transported across the outer forespore membrane into the intermembrane
space (8, 33).
Warth and Strominger first determined the structures of both vegetative
cell and spore PG in Bacillus subtilis (35-37).
In both structures the glycan strands contain alternating
N-acetylglucosamine (NAG) and N-acetylmuramic
acid (NAM) residues. Each NAM residue initially has a pentapeptide side
chain that can be utilized in the cross-linking of the glycan strands.
The cross-linking of these peptides involves removal of a terminal
D-alanine by a transpeptidase and the formation of a new
peptide bond between the second D-alanine residue of
the peptide side chain and the diaminopimelic acid (Dpm) residue
of a peptide side chain on another glycan strand. The transglycosylase
and transpeptidase activities involved in PG synthesis are
contained within the high-molecular-weight (high-MW) penicillin-binding proteins (PBPs) (reviewed in references
10 and 13). The low-MW PBPs are
most frequently DD-carboxypeptidases that remove terminal
D-alanine residues. This activity can limit the number of
peptides available to participate in cross-linking and can thereby
decrease the overall percentage of cross-linking (23, 24,
32). More recently, spore PG structure has been analyzed in
greater detail and in a number of mutant strains (3, 24). In
spore PG the peptide side chains are completely removed from 50% of
the NAM residues and these sugars are converted to muramic-
-lactam.
These muramic--lactam residues are produced in a regular pattern at
every second NAM position in the glycan strands. The side chains on
24% of the NAM residues have been cleaved to single
L-alanine residues. A result of this removal of the peptide
side chains is that only 4% of NAM residues are cross-linked whereas
40% are cross-linked in vegetative PG (3, 24, 35-37).
Muramic--lactam appears to be found in cortex PG and not in germ
cell wall PG (3, 25). Spores produced by B. subtilis
cwlD mutant strains do not contain muramic--lactam, are unable
to degrade cortex PG, and cannot complete germination (3, 25,
31). These findings and in vitro evidence (6, 7)
indicate that muramic--lactam functions as a specificity determinant
for spore germination lytic enzymes.
A mutant strain lacking the dacB gene product, a
DD-carboxypeptidase, produces spores that show a fivefold
increase in PG cross-linking and that are slightly delayed in spore
outgrowth in comparison to wild-type spores (26). These
spores exhibit no significant change in spore core dehydration, as
measured on wet density gradients (23, 26). Spore PG
produced by cwlD mutant (3, 25, 31) and
cwlD dacB double-mutant (27) strains have two-
and eightfold increases in cross-linking, respectively. However, these
spores exhibit normal spore core dehydration (25, 27).
A dacF mutant which lacks another sporulation-specific DD-carboxypeptidase produces spores with normal PG
structure, whereas a dacB dacF double-mutant strain produces
spores that have approximately a 10-fold increase in PG cross-linking,
do not appear to achieve normal spore core dehydration, and are even more delayed in spore outgrowth than the dacB single mutant
(23).
Theories suggesting that a mechanical activity of spore PG may
contribute to the attainment of spore core dehydration are dependent on
the low cross-linking of spore PG (17, 23, 34). Loosely
cross-linked PG is highly flexible and is capable of expanding and
contracting in response to ionic changes (21). Theories suggest that spore PG may contract (17) or expand
anisotropically (34) in response to ionic changes in the
environment and could thereby exert pressure on the spore core to
achieve dehydration. A recent theory proposed a gradient of
cross-linking across the layers of B. subtilis spore PG in
which the innermost layers of spore PG are most loosely cross-linked
and the outermost layers are most highly cross-linked (23).
This type of structure could potentially carry out the anisotropic
expansion proposed by Warth (34). The more loosely
cross-linked inner layers would expand more than the highly
cross-linked outer layers. The rigidity of the outer layers would
direct the expansion inwards and could create pressure on the spore
core, thus contributing to its dehydration (23).
The observation that mutant spores with significantly increased PG
cross-linking are able to achieve normal spore core dehydration appears
to be inconsistent with models that suggest that low cross-linking of
the spore cortex allows this structure to actively participate in the
process of dehydration. However, previous studies on the structure of
spore PG only examined the PG structure in dormant spores and thus
could not observe possible temporal changes in PG structure during
sporulation or the spatial distribution of PG structural modifications.
Small changes in structure that might have a significant effect on
spore properties, if concentrated in one area of spore PG, could not be
observed. The dramatic changes in spore PG structure produced by
dacB and cwlD mutations might have no effect on
spore core dehydration if the theorized gradient of cross-linking was
not altered.
We have adapted the spore PG analytical method (3, 24) to
examine the PG structure in developing forespores of B. subtilis at each stage during its synthesis. We have determined
the temporal correlation of changes in spore PG structure with other
well-defined sporulation events. We have analyzed forespore PG
synthesis in the wild type as well as eight mutant strains. We report
that a gradient of cross-linking does span the spore cortex PG.
However, the loss of this gradient in some mutant strains indicates
that it is not required for spore core dehydration.
| |
MATERIALS AND METHODS |
Bacterial strains and growth media.
All strains were derived
from B. subtilis strain PS832, a prototrophic derivative of
strain 168. Cultures were induced to sporulate by the nutrient
exhaustion method (20). Growth and sporulation were in 2×
SG medium (16) in which the concentration of glucose was
reduced twofold from the published value in order to achieve the most
synchronously sporulating culture.
Phenotypic and biochemical assays.
Glucose dehydrogenase
(GDH) activity and dipicolinic acid (DPA) accumulation were assayed as
previously described (20). Resistance of spores to heating
at 80°C for 10 min was measured as described previously
(26). The water content of protoplasted cells was determined
on density gradients (Nycodenz; Sigma) as described previously
(18, 26). To measure total muramic acid and glucosamine
content, 0.5-ml culture samples were centrifuged (15,800 × g for 45 s), the supernatant was discarded, and the sample
was washed in 0.5 ml of cold 1 mM MgCl2. The samples were then resuspended in 0.5 ml of cold 6 N HCl, and 20 µl was transferred to a 1.5-ml tube. The 20-µl samples were hydrolyzed at 95°C for 4 h and subjected to amino acid and amino sugar analysis
(11).
Preparation and analysis of immature forespore PG.
Culture
samples (30 ml) were centrifuged at 8,000 × g for 5 min at 20°C. The supernatant was discarded, and the pelleted cells were resuspended in 5 ml of SMM protoplast solution (5). To degrade the mother cell wall, 25 mg of lysozyme (Sigma) was added and
samples were then incubated at 37°C for 15 min. The protoplasted cells were then added to 45 ml of boiling 4% sodium dodecyl sulfate (SDS) (Sigma)-50 mM dithiothreitol (Labscientific, Inc.) solution and
boiled for 20 min.
Samples were cooled to room temperature and then transferred to a 50-ml
centrifuge tube and centrifuged at 21,000 × g for 30 min at 25°C. The supernatant was discarded, and the samples were
resuspended in 1 ml of warm (60°C) sterile water. The suspension was
boiled for 5 min to solubilize SDS and then centrifuged at 21,000 × g for 20 min. The supernatant was discarded,
and the washes were repeated until no SDS was detected (12).
Each sample was treated with DNase I (10 µg) (Sigma) and RNase A (50 µg) (Sigma) at 37°C for 2 h in a total volume of 1.0 ml of 100 mM Tris-HCl (pH 7.0)-20 mM MgSO4. Trypsin (100 µg)
(Worthington TPCK [tolylsulfonyl phenylalanyl chloromethyl ketone])
and CaCl2 (10 mM final concentration) were added, and
incubation at 37°C was continued for 16 h. Samples were
centrifuged (21,000 × g for 20 min), the supernatant
was discarded, and the samples were resuspended in 1.0 ml of 1% SDS. The samples were then boiled for 20 min to inactivate the trypsin. Samples were washed with H2O as described above until no
SDS could be detected. The isolated spore PG was digested with 125 U of Mutanolysin (Sigma) in a total volume of 250 µl of 12.5 mM
NaPO4 (pH 5.5) for 16 h at 37°C. The solubilized
muropeptides were reduced and separated by reverse-phase high-pressure
liquid chromatography (HPLC) as previously described (24).
| |
RESULTS |
Timing of forespore PG synthesis.
Forespore PG synthesis and
several sporulation phenotypic and biochemical markers were analyzed in
cultures of wild-type B. subtilis. All analyses were
performed on three separate cultures. The data presented are from
one culture; however, the other two analyses produced similar
results. Sporulation was induced in cultures of B. subtilis by nutrient exhaustion as previously described (16,
20). Figure 1 shows the timing of
the different events that occurred during sporulation. GDH
activity reached its maximum during the fifth hour of sporulation
(t5) and then dropped to almost zero (data not
shown) as the forespores became resistant to the lysozyme treatment
used in the GDH assay. Levels of DPA began to increase around
t4.5 to t5 and continued
to increase throughout the latter stages of sporulation. The
appearance of heat-resistant spores followed both of these
events. This order of events has been observed in previous
studies (38). The water content of the developing
forespore core was determined by measuring the wet density
(18) of the protoplasted sporangia. The appearance of
higher-density cells roughly paralleled the appearance of
heat-resistant CFU (Fig. 1).
View larger version (27K):
[in this window]
[in a new window]
|
FIG. 1.
Appearance of biochemical and phenotypic markers during
sporulation. Wild-type cells were grown in 2× SG medium, and samples
were collected for assay as described in Materials and Methods. Results
for GDH activity, DPA accumulation, spore PG produced, and heat
resistant CFU are expressed as percentages of the maximum values
reached during the course of the experiment. Spore dehydration is
expressed as the percentage of protoplasted cell material that
banded on a density gradient at a density greater than that of
vegetative cells.
|
|
The total muramic acid contents of culture samples from each time point
were used to estimate the amounts of forespore PG synthesized. It was
assumed that there was no significant forespore PG synthesis before
t3 since no forespore PG was recovered from samples taken at t3 (see results below) and
since no engulfed forespores were visible microscopically prior to this
time. As seen previously (38), the culture muramic acid
content reached a plateau between t2 and
t3. This was interpreted as the total contribution of mother cell wall PG (and nonsporulating cell wall PG)
to the culture, and this value was not expected to change significantly
during the remainder of the sporulation process. Previous studies
revealed little or no cell wall turnover during starvation of B. subtilis under conditions that did not induce lysis
(9). Values for muramic acid content for culture samples taken at t2.5 and t3 were
averaged to give the mother cell wall PG contribution. This value was
subtracted from the muramic acid content for culture samples taken
between t3.5 and t8 to
give the forespore PG muramic acid content. This value should
correspond to the PG produced between the two membranes surrounding the
engulfed forespore. This is the forespore PG that was purified,
following lysozyme treatment of the culture samples, for structural
analysis. The highest forespore PG muramic acid content was set to
100%, and the percentage of forespore PG made at each time point was calculated. These values were graphed against time in sporulation, and
a best-fitting line was drawn. The reported percentages of spore PG
made were taken from the interpolated line.
Extraction, purification, and analysis of wild-type forespore
PG.
To extract developing forespore PG, samples from sporulating
cultures were centrifuged and the pellets were resuspended in an
isotonic solution containing lysozyme to degrade the mother cell wall.
Microscopic observation revealed that 99% of the cells had become
round after 15 min and that >90% of the developing forespores were
still within protective protoplast membranes. The protoplasted cells
were added to a boiling solution of SDS and DTT to remove the mother
cell material and to prevent autolytic activity on the spore PG. The
samples were then washed extensively to remove the SDS and were treated
with nucleases and proteases to digest the remaining cell
components. The purified spore PG was then treated with a muramidase
that specifically cleaves the PG between the NAM and NAG residues but
not between muramic--lactam and NAG residues. The resulting
muropeptides were reduced and analyzed with reverse-phase HPLC using a
methanol gradient (Fig. 2)
(24). The muropeptide peaks were tentatively
identified based on cochromatography with previously identified peaks
(Table 1) (3, 24).
Muropeptides were collected from the methanol gradient and were run on
a secondary acetonitrile gradient system to verify that each peak
represented a single muropeptide (24). The identities of
muropeptides collected from the acetonitrile gradient were verified
using amino acid analysis (11, 24). Based on elution times
and amino acid analyses of muropeptide peaks derived from developing
forespore PG, there was no significant contamination with the
predominant muropeptides derived from vegetative cell PG: (disaccharide
[DS]-tripeptide [TriP] and DS-TriP-DS-tetrapeptide [TP]) with
amidations of the Dpm residues (2) (data not shown).
View larger version (21K):
[in this window]
[in a new window]
|
FIG. 2.
HPLC analysis of immature forespore PG-derived
muropeptides. Forespore PG was purified from culture samples taken
every 15 min between t4 and
t7, but only data for samples taken every 30 min
are shown. PG was digested with muramidase, reduced, and separated
using a methanol gradient as previously described (24).
Peaks are numbered as previously described (24), and
muropeptide structures are given in Table 1. Peaks labeled B are buffer
components. Peaks labeled X are produced by reduction of
muramic--lactam, and corrections for this reduction were
made during calculation of PG structural parameters (24).
|
|
The muropeptide peaks labeled 3Am and 7Am (Fig. 2) represent unique
peaks that are seen only in developing forespore PG and not in
dormant-spore PG. Amino acid analyses tentatively identified muropeptide 3Am as DS-TriP-TP and muropeptide 7Am as a tetrasaccharide (TS) with no peptide side chain (Table
2). These predicted identities of
muropeptides 3Am and 7Am were confirmed by mass spectrometry (Table 2).
These two novel peaks are the result of muramoyl-L-alanine amidase activity, an enzymatic activity normally associated with both
vegetative and spore PG metabolism. A cwlB (lytC)
mutation which eliminates the most abundant
muramoyl-L-alanine amidase in vegetative cells (14,
15) had no effect on the production of muropeptides 3Am and 7Am
(data not shown). These two muropeptides were not found in forespore PG
purified from a cwlD mutant, suggesting that the amidase
encoded by this gene (3, 25, 31) is responsible for their
production. For two reasons we believe that much of the production of
these amidase products is an artifact of unregulated CwlD activity
during the forespore PG purification procedure. First, amounts of these
muropeptides varied dramatically not only between experiments but also
within each experiment. For example, amidase products were sometimes
seen in samples taken at t4 and t4.5 but not at t4.25.
Second, in muropeptide 7Am, peptides were cleaved from two adjacent
muramic acid residues. If this muropeptide were a normal component of
forespore PG, it would be on the pathway for production of a
hexasaccharide which contains two adjacent muramic--lactam residues.
However, in several instances, production of muropeptide 7Am was in
excess of the eventual amount of hexasaccharide muropeptides
produced (data not shown).
A correction for the production of muropeptides 3Am and 7Am was
included in the calculation of spore PG structural parameters. Since
muropeptide 8 is the only muropeptide ever observed in spore PG that
has a TriP-TP cross-link, it was assumed that muropeptide 3Am is
derived from muropeptide 8 via amidase activity. In calculating structural parameters, the molar values of muropeptides 3Am and 8 were
combined to produce the total amount of TriP-TP cross-links. Muropeptide 7Am can presumably be derived by amidase activity on any of
the other muropeptides that contain TS, such as muropeptides 10, 13, 14, 17, and 20. We have no evidence to suggest that muropeptide 7Am is
derived preferentially from any subset of these TS-containing muropeptides. To correct for this amidase activity, we calculated the
amount of each TS-containing muropeptide as a percentage of total TS.
We then divided the molar amount of muropeptide 7Am according to these
percentages and added these values to the molar amounts of the other
TS-containing muropeptides. The validity of these corrections was
determined by comparing two sets of structural parameters. The first
set of structural parameters was calculated, with the correction, for a
sample that contained significant amounts of muropeptides 3Am and 7Am.
The second set of structural parameters was calculated, without the
correction, for a similar sample that, by chance, had extremely low
levels of these muropeptides. There was no significant difference
between the two sets of structural parameters.
HPLC analysis of the purified forespore PG shows increasing amounts of
PG from the beginning to the end of spore PG synthesis during
sporulation (Fig. 2). We believe that there is little or no degradation
and turnover of forespore PG. Turnover of vegetative-cell PG is
necessitated by the inside-to-outside growth process of the cell wall,
which involves PG synthesis on the inside of the wall with concurrent
shedding of PG on the outside (1). Forespore PG is
synthesized predominantly from the outside (33), so PG degradation is not required for expansion of this wall. Difficulty in
the precise definition of the contribution of mother cell wall PG to
the muramic acid content of the culture resulted in poor accuracy in
the reported percent spore PG made, especially in samples taken early
during forespore PG synthesis. Due to this problem we limit our
conclusions about early forespore PG synthesis to observations of
relatively large changes in PG structure.
Structural parameters (Table 3) for the
forespore PG recovered at each time point revealed that a relatively
low level of muramic--lactam observed in the first 5 to 10% of the
spore PG made was followed by a rapid increase. A high percentage of
TriP side chains was also observed in the first 10% of spore PG made. These two facts suggest that the germ cell wall is synthesized first,
adjacent to the inner forespore membrane. The germ cell wall has a high
level of cross-linking relative to the spore PG as a whole. A
calculation of the total amount of TriP-containing muropeptides (Table
3) and the number of peptide cross-links involving TriPs (data not
shown) revealed that these values do not change significantly following
synthesis of the first 10 to 15% of spore PG. We interpret this to
mean that the TriPs and TriP cross-links made in the germ cell wall are
stable during synthesis of the cortex PG.
There was a short decreasing gradient of cross-linking during synthesis
of the first 30% of spore PG, followed by a slightly increasing
gradient across the outer 70% of the spore PG. The cross-linking
values in Table 3 are for the total spore PG present in each sample and
do not indicate the cross-linking within the spore PG made between
samples. For example, a cross-linking value of 6.0% was observed for
the first 11% of the spore PG made, and a cross-linking value of 3.5%
was observed for the entire first 22% of the spore PG made (Table 3).
The cross-linking in the PG made between 11 and 22% must be lower than
3.5% in order to bring the average value over the total 22% of spore
PG down to 3.5%. These raw values for cross-linking were graphed
against percent spore PG made, and a line was drawn through the points. Cross-linking values were determined from the interpolated line for
each 10% of spore PG made. These new cross-linking values were used to
calculate the amount of cross-linking in each 10% "slice" of spore
PG made (Fig. 3). These calculations
indicate that cross-linking is relatively high in the first 10% of
spore PG synthesized (in the region of the germ cell wall) and rapidly decreases during synthesis of the next 20% of the spore PG,
correlating with the production of larger amounts of
muramic--lactam. An assumption inherent in this conclusion is that
cross-links formed in the first PG synthesized are not removed during
subsequent synthesis. We believe this is true due to the apparent
stability of the TriP-containing cross-links found in the first 10 to
15% of the spore PG made. Cross-linking increases progressively during synthesis of the final 70% of the spore PG. This increase in
cross-linking must be due to higher cross-linking in the new PG
synthesized because we never detected significant amounts of the
pentapeptide side chains required for cross-link formation. Repetition
of this analysis with forespore PG derived from three different
wild-type cultures indicated that the decrease in cross-linking between the 0- to 10% and 20- to 30% slices of spore PG ranged between three-
and sevenfold. The increases in cross-linking between the 20- to 30%
and 90- to 100% slices ranged between two- and sixfold. This
relatively large range in our reporting of the degree of the
cross-linking gradient is a result of the difficulties in determining
precise values for the amount of forespore PG synthesized at early time
points.
View larger version (33K):
[in this window]
[in a new window]
|
FIG. 3.
Cross-linking across the span of the spore PG in
wild-type and mutant B. subtilis strains. The percentage of
muramic acid residues with cross-linked peptides was calculated for
each 10% slice of spore PG isolated from cultures of eight different
strains. (A) Data for the wild-type ( ), dacA ( ),
dacF ( ), dacA dacC dacF ( ), ponA
( ), and cwlD ( ) strains. (B) Same data as in panel A
plus those for the dacB ( ) and cwlD dacB ( )
strains on an expanded vertical axis.
|
|
Forespore PG synthesis in mutant strains.
The method for the
analysis of developing forespore PG structure was applied to eight
B. subtilis mutant strains lacking various enzymes involved
in PG synthesis. Analyses were performed on two independent cultures of
each strain. Results from one analysis of each strain are shown, and
results from the duplicate analyses were similar. The timing of
appearance of GDH activity, DPA accumulation, heat resistance, an
increase in spore core wet density, and spore PG synthesis in all eight
mutant strains was similar to that observed for wild-type spores (data
not shown). Wet density analysis was not done for dacA and
dacF strains, but normal spore core dehydration is expected
since normal heat resistance was observed. Heat resistance for strains
with mutations in cwlD was not measured because the spores
of these strains cannot complete germination and form colonies. However, a normal increase in spore wet density was observed (data not
shown), and normal spore heat resistance was previously demonstrated for these strains (25, 27). A complete set of forespore PG structural parameters was derived for each of the strains (Table 4), except for the
dacB dacF strain, which will be discussed below. For each
strain, samples were collected every 15 min between t4 and t7, as was done
for the wild type. For brevity, only data for samples taken every 30 min are presented in Table 4.
Structural parameters of developing forespore PG from a dacF
mutant (strain PS1901, lacking DacF, a forespore-specific putative DD-carboxypeptidase [24, 39]) and a
ponA mutant (strain PS2062, lacking PBP1, the major class A
high-MW PBP [28, 30]) were similar to those determined
for the wild-type strain (Table 4). This finding is consistent with
previous structural analyses on dormant-spore PG from these two mutant
strains (24, 27). Previous studies of strains with mutations
in dacA (strain PS1900 dacA and strain DPVB17
dacA dacC dacF [24, 27]) demonstrated that dormant-spore PG produced by these mutants had a twofold reduction in
the amount of TriP side chains (24, 27). The dacC
mutation alone has no effect on spore PG structure, and the
dacF and dacC mutations had no effect in the
dacA mutant background (22, 23) (data not shown).
Structural analyses of forespore PG isolated from developing spores of
dacA strains showed that a gradual decrease in the amount of
TriP side chains occurred during spore PG synthesis relative to amounts
observed during the synthesis of wild-type spore PG (Table 4). This
analysis also revealed that spore PG from these strains contained a
significant number of pentapeptide side chains. Pentapeptide-containing
muropeptides were identified based on amino acid analysis data
(11) and cochromatography with pentapeptide-containing
muropeptides observed in cwlD mutant spore PG
(27). Pentapeptide side chains have also been found in
vegetative cell wall PG isolated from a dacA mutant strain (2). As observed previously (23, 24), these
changes in spore PG structure had no effect on spore dehydration or
heat resistance.
Synthesis of forespore PG in strains with mutations in dacB
(strain PS2066 [26]), dacB and
dacF (strain PS2421 [27]), cwlD
(strain PS2307 [25, 31]), and cwlD and
dacB (strain PS2422 [27]) differed
substantially from that in the wild type. Relative to that from the
wild type, developing forespore PG from the dacB strain had
an approximately twofold reduction in NAM residues with single
L-alanine side chains and a corresponding increase in TP
side chains throughout spore PG synthesis (Table 4). A slight decrease
in the amount of muramic--lactam and the large increase in peptide
side chains may be the direct cause of a fivefold increase in total
spore PG cross-linking seen in this strain (Table 4). In addition,
there was no decreasing or increasing gradient of cross-linking across
the span of the spore PG (Fig. 3).
Complete analysis of spore PG synthesis in the dacB dacF
double mutant was prevented by the appearance of many novel peaks that
were eluted later in the gradient (23), especially in
samples taken during the later stages of sporulation (data not shown). The structures of these new peaks were not determined. However, we were
able to identify peaks corresponding to approximately 90% of the
muropeptide material in the earliest samples, and we were able to
calculate structural parameters for the first 20% of the spore PG made
in two independent cultures. Structural changes, relative to the wild
type, seen in the first PG synthesized include (i) a threefold
reduction in the amount of single L-alanine side chains,
(ii) a greater-than-twofold increase in the amount of TP side chains,
and (iii) an approximately threefold increase in cross-linking. These
structural parameters are similar to those for early forespore PG from
the dacB single mutant. We used an alternate method for
determining the degree of PG cross-linking to analyze later samples
from dacB dacF double-mutant cultures (data not shown).
These and previous results (23) indicated that although the
levels of cross-linking in the first spore PG produced by
dacB and dacB dacF mutant strains were similar,
the later PG produced by the double mutant was more highly cross-linked than that of the single mutant.
Previous studies on the structure of PG from cwlD mutant
spores demonstrated that it contained no muramic--lactam and
contained small amounts of pentapeptide side chains and peptide side
chains that contained glycine (3, 25, 27). Analysis of a
cwlD mutant revealed the presence of small amounts of
pentapeptide side chains throughout forespore PG synthesis (Table 4).
Levels of cross-linking in cwlD mutant and wild-type
forespore PG are similar during synthesis of the first 10% of
forespore PG (Table 4). This similarity is due to the low level of
muramic--lactam in the first layers of forespore PG produced in the
wild type. Differences between the two strains become more apparent
during synthesis of the cortex PG. The absence of muramic--lactam in the last 90% of the spore PG synthesized in the cwlD mutant
is tied to the following structural differences relative to the wild type: (i) a slow progression to 60% more single-L-alanine
side chains, (ii) an increase to approximately threefold more TriP side
chains; (iii) a twofold increase in the amount of TP side chains, (iv)
a twofold increase in PG cross-linking (Table 4), and (v) the absence
of any increasing or decreasing gradient of cross-linking across the
span of the spore PG (Fig. 3).
Dormant-spore PG from a cwlD dacB double-mutant strain
demonstrated a combination of the structures observed in the two single mutants (25, 27). Analyses of developing forespore PG in the cwlD dacB double mutant revealed a similar pattern
(Table 4). A twofold decrease in the amount of
single-L-alanine side chains, relative to the
wild type, throughout the period of cwlD dacB mutant
forespore PG synthesis was also observed in the dacB single mutant. The first 20% of forespore PG synthesized in the double mutant
had amounts of TriP side chains similar to those observed during early
wild-type and cwlD mutant forespore PG synthesis. However,
the amount of TriP side chains increased twofold over wild-type values
during synthesis of the last 80% of the spore PG in the cwlD
dacB mutant, just as observed in the cwlD strain. Cross-linking in the first layers of spore PG produced was higher than
that found in the wild type, similar to what was found for the
dacB mutant. Cross-linking increased a further twofold
in the remainder of the spore PG, and no gradient of cross-linking was
observed (Fig. 3).
| |
DISCUSSION |
A method for analyzing the structure of dormant-spore PG using
reverse-phase HPLC (24) was adapted for examination of the structure of PG at each stage of its synthesis in developing forespores of B. subtilis. Structural parameters derived for wild-type
forespore PG suggest that the structure of the first 10% of spore PG
made has a high percentage of TriP side chains and a small amount of muramic--lactam relative to later samples. These two facts lead to
the conclusion that the structure found in the first 10% of spore PG
made represents the germ cell wall, which is predicted to have a
structure similar to that of vegetative PG (2) and also
suggests that the germ cell wall is synthesized first, adjacent to the
inner forespore membrane. Our data do not address whether the
substrates for synthesis of these first layers of spore PG are derived
from the forespore or from the mother cell. Previous studies have
demonstrated that muramic--lactam is a specificity determinant for
cortex-specific lytic enzymes (3, 6, 7, 25), which is
consistent with the conclusion that the germ cell wall has little, if
any, muramic--lactam. Structural analyses of germinating
dacA mutant spores demonstrated that the amount of DS
muropeptides containing TriP side chains remained constant throughout
the process of spore PG degradation during germination (2),
providing further support for the conclusion that the germ cell wall
has a high percentage of TriP side chains. There appears to be very few
TriP side chains produced in the remaining spore PG, the cortex; and
the TriP and TriP-containing cross-links within the germ cell wall
appear to be stable during formation of the remainder of the spore.
A recently proposed model suggested that a gradient of cross-linking,
in which the innermost layers of spore PG are most loosely cross-linked
and the outermost layers are most highly cross-linked, was formed
during spore PG synthesis and that this gradient could be involved in
the attainment of spore core dehydration (27). Cross-linking
values calculated for each section of spore PG suggest that the region
of the germ cell wall, or the first 10% of spore PG made, is highly
cross-linked in comparison to the remainder of the spore PG
(cross-linking of the germ cell wall is still very low relative to that
found in vegetative-cell PG [2, 29, 36]). The level of
cross-linking within the germ cell wall appears to be partly determined
by the activity of the dacB product, PBP5*, a
DD-carboxypeptidase. Loss of PBP5* resulted in a
greater-than-twofold increase in germ cell wall cross-linking. The
further loss of dacF or cwlD in the
dacB background did not raise germ cell wall cross-linking
above 16 to 17%, even though cross-linking in later spore PG synthesis
was increased. Germ cell wall cross-linking may be limited by a
transpeptidase activity specifically involved in synthesis of this PG
layer. It was surprising that the level of germ cell wall cross-linking
was not affected by the product of dacF, a putative
DD-carboxypeptidase that is expressed in the forespore and
that was presumed to cross the inner forespore membrane and remain
associated with the outer surface of that membrane (39).
A rapid decrease in cross-linking was observed within the 20% of the
spore PG synthesized after the germ cell wall. A major factor in this
decrease appears to be the cwlD product, a
muramoyl-L-alanine amidase involved in formation of
muramic--lactam (3, 25, 31). Loss of CwlD specifically
prevents the decrease in cross-linking following germ cell wall
synthesis but only slightly increases the overall degree of
cross-linking. PBP5* also appears to play a role in generating this
decrease in cross-linking; the decrease was not present in the
dacB mutant. The large increase in cross-linking in the
dacB mutant has only a slight effect on
muramic--lactam formation by CwlD, suggesting that either CwlD
activity is not substrate limited in the dacB mutant or CwlD
is able to remove cross-linked peptides. The facts that both CwlD and
PBP5* can affect the degree of cross-linking and that their
effects are additive suggest that they function independently.
A two- to eightfold increase in cross-linking was observed during
synthesis of the final 70% of the wild-type spore PG (Fig. 3). The
process driving this increase is unclear. It could be affected by a
decrease in the activity of a presumed endopeptidase involved in the
production of L-Ala side chains, with a corresponding increase in peptide side chains available for cross-linking or a
decrease in removal of cross-linked peptides. Alternatively, a decrease
in the specific activity of DD-carboxypeptidases would allow more of the remaining peptides to be cross-linked. Interestingly, the loss of dacF in the dacB background led to an
increase in cross-linking outside of the germ cell wall. One
explanation might be that DacF is not anchored to the membrane as
previously suggested (39). This could also explain the
failure of repeated efforts to detect this putative PBP in membrane
preparations (data not shown).
Unresolved is the issue of whether spore PG is actively involved in
spore core dehydration. Our data can rule out some possible models.
While a gradient of cross-linking is produced across the span of the
spore PG, the absence of this gradient in forespore PG from
dacB, cwlD, and cwlD dacB strains,
which produce spores with normal dehydration (25, 27),
indicates that the gradient is not required. Formation of a spore that
can achieve and maintain core dehydration may simply require PG
cross-linking within a broad range. The upper limit of the range
required for achieving dehydration may have been surpassed in the
dacB dacF double mutant. While the dacB mutant
spores achieve normal core dehydration, they are unable to maintain it
upon heating (26). These spores may define a spore PG
cross-linking limit required for maintenance of dehydration. Another
factor affecting the evolution of the specific level of spore PG
cross-linking may be the rate of spore outgrowth. The higher degrees of
spore PG cross-linking in the dacB and dacB dacF
mutants correlate with slowed outgrowth of their spores. Difficulty in
degrading the highly cross-linked spore PG may present a physical
barrier to growth of these germinated cells. PG cross-linking during
sporulation may thus be a balancing act between achieving the wall
stability required for maintaining spore core dehydration and the wall
flexibility required for rapid outgrowth and attainment of spore core dehydration.
This work was supported by a Sigma Xi Grant-in-Aid of Research
(J.M.-P.) and by grant GM56695 (D.L.P.) from the National Institutes of
Health. Mass spectrometry was provided by the Washington University Mass Spectrometry Resource with support from the NIH National Center
for Research Resources (Grant P41RR0954).
We thank Megan Manfredi for technical assistance and Marita
Seppanen Popham for editing the manuscript.
| 1.
|
Archibald, A. R.,
I. C. Hancock, and C. R. Harwood.
1993.
Cell wall structure, synthesis, and turnover, p. 381-410.
In
A. L. Sonenshein, J. A. Hoch, and R. Losick (ed.), Bacillus subtilis and other gram-positive bacteria. American Society for Microbiology, Washington, D.C.
|
| 2.
|
Atrih, A.,
G. Bacher,
G. Allmaier,
M. P. Williamson, and S. J. Foster.
1999.
Analysis of peptidoglycan structure from vegetative cells of Bacillus subtilis 168 and role of PBP 5 in peptidoglycan maturation.
J. Bacteriol.
181:3956-3966[Abstract/Free Full Text].
|
| 3.
|
Atrih, A.,
P. Zollner,
G. Allmaier, and S. J. Foster.
1996.
Structural analysis of Bacillus subtilis 168 endospore peptidoglycan and its role during differentiation.
J. Bacteriol.
178:6173-6183[Abstract/Free Full Text].
|
| 4.
|
Beaman, T. C.,
J. T. Greenamyre,
T. R. Corner,
H. S. Pankratz, and P. Gerhardt.
1982.
Bacterial spore heat resistance correlated with water content, wet density, and protoplast/sporoplast volume ratio.
J. Bacteriol.
150:870-877[Abstract/Free Full Text].
|
| 5.
|
Bron, S.
1990.
Plasmids, p. 75-174.
In
C. R. Harwood, and S. M. Cutting (ed.), Molecular biological methods for Bacillus. John Wiley & Sons Ltd., Chichester, England.
|
| 6.
|
Chen, Y.,
S. Fukuoka, and S. Makino.
2000.
A novel spore peptidoglycan hydrolase of Bacillus cereus: biochemical characterization and nucleotide sequence of the corresponding gene, sleL.
J. Bacteriol.
182:1499-1506[Abstract/Free Full Text].
|
| 7.
|
Chen, Y.,
S. Miyata,
S. Makino, and R. Moriyama.
1997.
Molecular characterization of a germination-specific muramidase from Clostridium perfringens S40 spores and nucleotide sequence of the corresponding gene.
J. Bacteriol.
179:3181-3187[Abstract/Free Full Text].
|
| 8.
|
Daniel, R. A.,
S. Drake,
C. E. Buchanan,
R. Scholle, and J. Errington.
1994.
The Bacillus subtilis spoVD gene encodes a mother-cell-specific penicillin-binding protein required for spore morphogenesis.
J. Mol. Biol.
235:209-220[CrossRef][Medline].
|
| 9.
|
de Boer, W. R.,
P. D. Meyer,
C. G. Jordens,
F. J. Kruyssen, and J. T. Wouters.
1982.
Cell wall turnover in growing and nongrowing cultures of Bacillus subtilis.
J. Bacteriol.
149:977-984[Abstract/Free Full Text].
|
| 10.
|
Ghuysen, J.-M.
1991.
Serine  -lactamases and penicillin-binding proteins.
Annu. Rev. Microbiol.
45:37-67[CrossRef][Medline].
|
| 11.
|
González-Castro, M. J.,
J. López-Hernández,
J. Simal-Lozano, and M. J. Oruña-Concha.
1997.
Determination of amino acids in green beans by derivitization with phenylisothiocyanate and high-performance liquid chromatography with ultraviolet detection.
J. Chromatogr. Sci.
35:181-185.
|
| 12.
|
Hayashi, K.
1975.
A rapid determination of sodium dodecyl sulfate with methylene blue.
Anal. Biochem.
67:503-506[CrossRef][Medline].
|
| 13.
|
Holtje, J. V.
1998.
Growth of the stress-bearing and shape-maintaining murein sacculus of Escherichia coli.
Microbiol. Mol. Biol. Rev.
62:181-203[Abstract/Free Full Text].
|
| 14.
|
Kuroda, A., and J. Sekiguchi.
1991.
Molecular cloning and sequencing of a major Bacillus subtilis autolysin gene.
J. Bacteriol.
173:7304-7312[Abstract/Free Full Text].
|
| 15.
|
Lazarevic, V.,
P. Margot,
B. Soldo, and D. Karamata.
1992.
Sequencing and analysis of the Bacillus subtilis lytRABC divergon: a regulatory unit encompassing the structural genes of the N-acetylmuramoyl-L-alanine amidase and its modifier.
J. Gen. Microbiol.
138:1949-1961[Medline].
|
| 16.
|
Leighton, T. J., and R. H. Doi.
1971.
The stability of messenger ribonucleic acid during sporulation in Bacillus subtilis.
J. Biol. Chem.
254:3189-3195.
|
| 17.
|
Lewis, J. C.,
N. S. Snell, and H. K. Burr.
1960.
Water permeability of bacterial spores and the concept of a contractile cortex.
Science
132:544-545[Abstract/Free Full Text].
|
| 18.
|
Lindsay, J. A.,
T. C. Beaman, and P. Gerhardt.
1985.
Protoplast water content of bacterial spores determined by buoyant density sedimentation.
J. Bacteriol.
163:735-737[Abstract/Free Full Text].
|
| 19.
|
Nakashio, S., and P. Gerhardt.
1985.
Protoplast dehydration correlated with heat resistance of bacterial spores.
J. Bacteriol.
162:571-578[Abstract/Free Full Text].
|
| 20.
|
Nicholson, W. L., and P. Setlow.
1990.
Sporulation, germination, and outgrowth, p. 391-450.
In
C. R. Harwood, and S. M. Cutting (ed.), Molecular biological methods for Bacillus. John Wiley & Sons Ltd., Chichester, England.
|
| 21.
|
Ou, L.-T., and R. E. Marquis.
1970.
Electromechanical interactions in cell walls of gram-positive cocci.
J. Bacteriol.
101:92-101[Abstract/Free Full Text].
|
| 22.
|
Pedersen, L. B.,
T. Murray,
D. L. Popham, and P. Setlow.
1998.
Characterization of dacC, which encodes a new low-molecular-weight penicillin-binding protein in Bacillus subtilis.
J. Bacteriol.
180:4967-4973[Abstract/Free Full Text].
|
| 23.
|
Popham, D. L.,
M. E. Gilmore, and P. Setlow.
1999.
Roles of low-molecular-weight penicillin-binding proteins in Bacillus subtilis spore peptidoglycan synthesis and spore properties.
J. Bacteriol.
181:126-132[Abstract/Free Full Text].
|
| 24.
|
Popham, D. L.,
J. Helin,
C. E. Costello, and P. Setlow.
1996.
Analysis of the peptidoglycan structure of Bacillus subtilis endospores.
J. Bacteriol.
178:6451-6458[Abstract/Free Full Text].
|
| 25.
|
Popham, D. L.,
J. Helin,
C. E. Costello, and P. Setlow.
1996.
Muramic lactam in peptidoglycan of Bacillus subtilis spores is required for spore outgrowth but not for spore dehydration or heat resistance.
Proc. Natl. Acad. Sci. USA
93:15405-15410[Abstract/Free Full Text].
|
| 26.
|
Popham, D. L.,
B. Illades-Aguiar, and P. Setlow.
1995.
The Bacillus subtilis dacB gene, encoding penicillin-binding protein 5*, is part of a three-gene operon required for proper spore cortex synthesis and spore core dehydration.
J. Bacteriol.
177:4721-4729[Abstract/Free Full Text].
|
| 27.
|
Popham, D. L.,
J. Meador-Parton,
C. E. Costello, and P. Setlow.
1999.
Spore peptidoglycan structure in a cwlD dacB double mutant of Bacillus subtilis.
J. Bacteriol.
181:6205-6209[Abstract/Free Full Text].
|
| 28.
|
Popham, D. L., and P. Setlow.
1995.
Cloning, nucleotide sequence, and mutagenesis of the Bacillus subtilis ponA operon, which codes for penicillin-binding protein (PBP) 1 and a PBP-related factor.
J. Bacteriol.
177:326-335[Abstract/Free Full Text].
|
| 29.
|
Popham, D. L., and P. Setlow.
1993.
The cortical peptidoglycan from spores of Bacillus megaterium and Bacillus subtilis is not highly cross-linked.
J. Bacteriol.
175:2767-2769[Abstract/Free Full Text].
|
| 30.
|
Popham, D. L., and P. Setlow.
1996.
Phenotypes of Bacillus subtilis mutants lacking multiple class A high-molecular-weight penicillin-binding proteins.
J. Bacteriol.
178:2079-2085[Abstract/Free Full Text].
|
| 31.
|
Sekiguchi, J.,
K. Akeo,
H. Yamamoto,
F. K. Khasanov,
J. C. Alonso, and A. Kuroda.
1995.
Nucleotide sequence and regulation of a new putative cell wall hydrolase gene, cwlD, which affects germination in Bacillus subtilis.
J. Bacteriol.
177:5582-5589[Abstract/Free Full Text].
|
| 32.
|
Severin, A.,
C. Schuster,
R. Hakenbeck, and A. Tomasz.
1992.
Altered murein composition in a DD-carboxypeptidase mutant of Streptococcus pneumoniae.
J. Bacteriol.
174:5152-5155[Abstract/Free Full Text].
|
| 33.
|
Tipper, D. J., and P. E. Linnet.
1976.
Distribution of peptidoglycan synthetase activities between sporangia and forespores in sporulating cells of Bacillus sphaericus.
J. Bacteriol.
126:213-221[Abstract/Free Full Text].
|
| 34.
|
Warth, A. D.
1985.
Mechanisms of heat resistance, p. 209-225.
In
G. J. Dring, D. J. Ellar, and G. W. Gould (ed.), Fundamental and applied aspects of bacterial spores. Academic Press, Inc., London, United Kingdom.
|
| 35.
|
Warth, A. D., and J. L. Strominger.
1972.
Structure of the peptidoglycan from spores of Bacillus subtilis.
Biochemistry
11:1389-1396[CrossRef][Medline].
|
| 36.
|
Warth, A. D., and J. L. Strominger.
1971.
Structure of the peptidoglycan from vegetative cell walls of Bacillus subtilis.
Biochemistry
10:4349-4358[CrossRef][Medline].
|
| 37.
|
Warth, A. D., and J. L. Strominger.
1969.
Structure of the peptidoglycan of bacterial spores: occurrence of the lactam of muramic acid.
Proc. Natl. Acad. Sci. USA
64:528-535[Abstract/Free Full Text].
|
| 38.
|
Wickus, G. G.,
A. D. Warth, and J. L. Strominger.
1972.
Appearance of muramic lactam during cortex synthesis in sporulating cultures of Bacillus cereus and Bacillus megaterium.
J. Bacteriol.
111:625-627[Abstract/Free Full Text].
|
| 39.
|
Wu, J.-J.,
R. Schuch, and P. J. Piggot.
1992.
Characterization of a Bacillus subtilis operon that includes genes for an RNA polymerase  factor and for a putative DD-carboxypeptidase.
J. Bacteriol.
174:4885-4892[Abstract/Free Full Text].
|
![[Feedback]](/icons/banner/feedbackACT.gif){kind=icon}
![[For Subscribers]](/icons/banner/subscriberACT.gif){kind=icon}
![[Archive]](/icons/banner/archiveACT.gif){kind=icon}
![[Search]](/icons/banner/searchACT.gif){kind=icon}
![[Contents]](/icons/banner/tocACT.gif){kind=icon}
Top
Abstract
Introduction
