SEARCH

SEARCH BY CITATION

Keywords:

  • E. coli;
  • Cell division;
  • Murein

Abstract

  1. Top of page
  2. Abstract
  3. 1Introduction
  4. 2Materials and methods
  5. 3Results and discussion
  6. Acknowledgements
  7. References

The chain length distribution of murein glycan strands was analyzed in wild-type cells and in cells in which preseptal and/or septal murein synthesis was prevented in ftsZ84 and ftsI36 mutants of E. coli. This revealed a significant change in glycan chain lengths in newly synthesized murein associated with inactivation of the ftsZ gene product but not with inactivation of the ftsI gene product. This is the first reported abnormality in murein biosynthesis associated with mutation of an essential cell division gene.


1Introduction

  1. Top of page
  2. Abstract
  3. 1Introduction
  4. 2Materials and methods
  5. 3Results and discussion
  6. Acknowledgements
  7. References

Cell division in Escherichia coli and most other rod-shaped bacteria occurs by formation of a division septum at midcell. The septum is formed by the coordinate invagination of the three cell envelope layers – inner membrane, murein (peptidoglycan) and outer membrane. Septation requires a number of essential division proteins but the mechanism by which these proteins facilitate septal ingrowth is not known [1].

The murein layer of the cell envelope is a cross-linked saccular structure in which linear glycan strands containing an average of 33 disaccharide units are crosslinked to each other via peptide side-chains [2]. Septal invagination involves a local change in the general orientation of the murein sacculus, leading to murein ingrowth at right angles to the long axis of the cylinder. Although this must involve a local change in peptidoglycan organization, the details are unknown.

Autoradiographic studies of the pattern of incorporation of [3H]diaminopimelate ([3H]Dap) into the sacculus have shown that the topologic pattern of peptidoglycan synthesis changes during the division cycle [3, 4]. Prior to the onset of septation, [3H]Dap is inserted randomly along the entire length of the cylinder. In contrast, when septal ingrowth begins, a new distribution pattern appears. Incorporation along the body of the cylinder diminishes significantly and new synthesis is predominantly restricted to a narrow zone at midcell. These results have suggested that there is a switch in the mode of murein synthesis during the cell cycle, from an ‘elongation’ mode of incorporation in which murein is synthesized along the entire length of the cell cylinder to a ‘septation’ mode in which synthesis is largely restricted to the new division site [5]. The switch-off of ‘elongation’ synthesis may be incomplete since some cell elongation continues during septation [4].

Using another experimental approach, de Pedro and coworkers have shown that new murein synthesis at division sites occurs within a sharply defined zone [6]. It was pointed out that the size and distribution of these zones were similar to those of the periseptal annuli that have been suggested to represent a preseptation stage of differentiation of the future division site [7, 8]. Interestingly, the localized zones of murein synthesis at division sites were not restricted to cells that formed septal crosswalls but were also observed in cells that were unable to form septal crosswalls because of ftsA, ftsQ, or ftsI mutations, or because of inhibition of synthesis of septal murein by β-lactams that block the activity of the ftsI gene product (PBP3). The appearance of localized zones of murein synthesis at potential division sites in these cells defines a preseptal stage of peptidoglycan synthesis that occurs after the initial identification of the future division site but before invagination of the septal crosswall.

The stage of preseptal murein synthesis is associated with the appearance of ‘blunt constrictions’ at regular intervals along the nonseptate filaments of most ftsA, ftsQ and ftsI mutants [9]. Preseptal murein synthesis does not require penicillin binding protein 3 (PBP3), the product of the ftsI gene, which is required for synthesis of septal murein [5, 6]. Preseptal murein synthesis also appears to be insensitive to penicillins that inhibit PBP2 and PBP1a and 1b, proteins that are involved in synthesis of ‘elongation’ murein [5, 6]. This suggested that preseptal murein synthesis is catalyzed by a thus-far unidentified penicillin-insensitive peptidoglycan synthesizing activity (PIPS) [5]. All of these observations indicate that preseptal murein synthesis at division sites represents a distinct stage in the differentiation of the E. coli division site.

The local zones of peptidoglycan synthesis at potential division sites that are characteristic of ftsA, ftsQ and ftsI filaments were not observed in filaments caused by inactivation of FtsZ in a ftsZ84ts strain grown at its non-permissive temperature [6]. This implies that FtsZ is required for the switch-on of preseptal murein synthesis at the correct time in the division cycle whereas ftsA, ftsQ and ftsI are required for the subsequent stages of PBP3-dependent septal murein ingrowth. It is not known whether FtsZ function is also required for the second part of the normal switch, the ‘switch-off’ of elongation synthesis.

The elongation, preseptal and septal stages of the division cycle are associated with differences in cell morphology. The outline of the cylinder at midcell is undisturbed during elongation synthesis, shows ‘blunt constrictions’ when preseptal synthesis occurs in the absence of septal synthesis, and is characterized by constrictive ingrowth of the cell envelope at right angles to the long axis of the cylinder during septation. Because the shape of the cell is dependent on the murein sacculus, this suggests that the three stages of the division cycle might be associated with changes in murein organization.

Previous studies of muropeptides of septal as compared to elongation murein failed to reveal significant changes in muropeptide composition [10–12]. In the present study we have examined another parameter of murein structure by analyzing the length distribution of the glycan chains that represent the backbone structure of the murein sacculus. The studies were performed during inactivation of the ftsZ and ftsI gene products in an attempt to identify characteristics that might correlate with the switch from lateral wall synthesis to preseptal and septal murein synthesis.

2Materials and methods

  1. Top of page
  2. Abstract
  3. 1Introduction
  4. 2Materials and methods
  5. 3Results and discussion
  6. Acknowledgements
  7. References

2.1Strains and growth conditions

Strain KN126 is the wild-type parent of SP63 [ftsI63Ts], kindly provided by B. Spratt, and KF1000 [ftsZ84Ts]. Strain KF1000 was isolated by P1-mediated cotransduction of leu:::Tn10 and ftsZ84 from KF999 to KN126. Cells were grown in L-broth at 30°C to mid-exponential growth phase (A600 approximately 0.33) and a portion of the culture was shifted to 42°C. When the culture reached an A600 of 0.39, [3H]N-acetylglucosamine (GlcNAc) (40 μl, 0.5 mCi ml−1, 36 Ci mmol−1) was added to 25 ml of each culture and growth was continued for 0.4 generations (10 min and 17 min for the 42°C and 30°C cultures, respectively) as estimated by increase in A600. Growth rates, as indicated by change in A600, were similar in all the 42°C cultures and also were similar in all of the 30°C cultures. At that time the cultures were rapidly chilled by pouring them onto 12.5 ml of frozen L-broth. The cell pellets were suspended in chilled M9 minimal medium and washed several times by centrifugation and resuspension. The final pellets were frozen in an acetone-dry ice bath.

2.2Isolation of murein sacculi

Sacculi were prepared by adding the resuspended cell pellet to an equal volume of hot 8% sodium dodecylsulfate (SDS) and heating the suspension in a boiling water bath for 30 min. After removal of SDS by repeated centrifugation and resuspension in sterile water, the sacculi were collected by centrifugation and treated with alpha-amylase and pronase to remove polysaccharides and murein-lipoprotein as previously described [13].

2.3Determination of length distribution of glycan strands

The labeled sacculi were digested with human serum amidase and fractionated by reversed-phase HPLC according to their degree of glycan strand polymerization, as previously described [13].

3Results and discussion

  1. Top of page
  2. Abstract
  3. 1Introduction
  4. 2Materials and methods
  5. 3Results and discussion
  6. Acknowledgements
  7. References

3.1Murein structure

The structure of newly synthesized murein was compared in sacculi of isogenic wild-type cells and cells of the temperature-sensitive mutant strains KF1000 [ftsZ84] and SP63 [ftsI63] grown at permissive and non-permissive temperatures. The sacculi were labeled by growth in the presence of [3H]GlcNAc for 0.4 generation at 30°C or during the initial 0.4 generation after a shift to 42°C. HPLC was used to analyze the distribution of the glycan chains released by amidase treatment of the labeled murein.

3.2Glycan chain length distribution

Inactivation of FtsZ was associated with a significant change in the length distribution of glycan strands in newly synthesized murein, with a shift from longer to shorter chain lengths when ftsZ84 cells grown at 42°C were compared with the isogenic wild-type strain grown under the same conditions (Fig. 1b, and Table 1). The relative abundance of glycan strands synthesized in the ftsZ strain relative to the wild-type strain (the ftsZ/wild type ratio) was greater than 1 for each of the 13 shortest glycan size classes (Fig. 1b). The probability that this could be due to chance alone is 1.2×10−4 and the results are therefore statistically highly significant. When glycan chains were grouped into four equal size classes (Table 1), the abundance of chains in the shortest length class was more than two-fold higher in the ftsZ84 samples than in the wild-type sacculi. Conversely, the longer chain length populations showed a progressive decrease in abundance in the ftsZ84 samples and a corresponding increase in the wild-type samples.

image

Figure 1. Glycan chain length distribution in wild-type and ftsZ84 cells. Cells were grown at 30°C (panel a) or 42°C (panel b). For each length class the bars indicate the concentration expressed as percent of total glycan strands in the entire population. White bars indicate ftsZ84 samples; filled bars indicate wild-type samples.

Download figure to PowerPoint

Table 1.  Glycan chain length classes
Glycan chain lengthStrains
 ftsZ84/WTftsI63/WT
 30°C42°C30°C42°C
  1. Glycan chain lengths were analyzed as described in Fig. 1. The values for each of the indicated ranges of glycan chain lengths were pooled. For each group, the ratio of the value for the ftsZ84 or ftsI63 strain was divided by the value for the wild-type strain to give the ftsZ84/WT and ftsI63/WT ratios.

1–71.412.231.030.96
8–151.281.441.051.04
16–230.811.551.061.62
14–300.60.711.011.04
>300.890.610.930.9

There was also a shift toward shorter glycan chain lengths in labeled sacculi from ftsZ84 cells grown at 30°C but the magnitude of the shift was much lower than in samples from cells grown at 42°C (compare Fig. 1a and 1b, and Table 1). Thus, the ftsZ84/wild type ratios were 1.41 and 0.89 in the shortest and longest length classes, respectively, in sacculi from cells grown at 30°C. The ftsZ/ wild type ratio was greater than 1 for 12 of the 13 smallest glycan size classes (Fig. 1a). The probability that this distribution could be due to chance alone is 1.6×10−3 and the results are therefore statistically highly significant. We assume that the moderate shift in glycan chain lengths in cells grown at permissive temperature reflects the fact that thermosensitive proteins frequently have some alteration of function in cells grown at permissive temperature although the cellular phenotypic effects may be slight.

In contrast to the changes in glycan chain distribution that were observed when FtsZ was inactivated, there was no significant change in glycan strand lengths associated with inactivation of PBP3, the ftsI gene product (Fig. 2, and Table 1).

image

Figure 2. Glycan chain length distribution in wild-type and ftsI63 cells. Cells were grown at 30°C (panel a) or 42°C (panel b). For each length class the bars indicate the concentration expressed as percent of total glycan strands in the entire population. White bars indicate ftsI63 samples; filled bars indicate wild-type samples.

Download figure to PowerPoint

The shift to shorter glycan chain lengths in ftsZ84 cells is the first described abnormality in murein synthesis to be associated with a cell division mutant. In these unsynchronized cells, the murein biosynthetic pattern reflected murein synthesized in cells from all stages of the cell cycle, including cells that were undergoing elongation murein synthesis, preseptal murein synthesis or septal murein synthesis. Because there was no change in glycan pattern associated with inactivation of ftsI, which leads to a division block in septal murein synthesis, after the stage of preseptal murein synthesis, the observed changes are unlikely to reflect changes in septal synthesis. Instead, the glycan changes in the ftsZ84 mutant correlate with the absence of preseptal murein synthesis that was shown to occur when FtsZ is inactivated [6]. This suggests that preseptal murein at division sites is composed of relatively long glycan strands when compared to the cylindrical murein that is synthesized during cell elongation. Further work will be needed to determine whether this is the only biochemical alteration that occurs at this stage of the division cycle and what relation it might have to the later formation of the division septum.

Acknowledgements

  1. Top of page
  2. Abstract
  3. 1Introduction
  4. 2Materials and methods
  5. 3Results and discussion
  6. Acknowledgements
  7. References

This study was supported by grants to L.R. from the US National Institutes of Health (GM53276) and the Human Frontiers in Science Program (RG-386/95).

References

  1. Top of page
  2. Abstract
  3. 1Introduction
  4. 2Materials and methods
  5. 3Results and discussion
  6. Acknowledgements
  7. References
  • 1
    Rothfield, L.I. and Justice, S. (1997) Bacterial Cell Division: The Cycle of the Ring. Cell 88, 581584.
  • 2
    Park, J.T. (1996) The murein sacculus. In: Escherichia coli and Salmonella (Neidhardt, F., Ed.), pp. 48–57. ASM Press, Washington, DC.
  • 3
    Woldringh, C.L., Huls, P., Pas, E., Brakenhoff, G.J. and Nanninga, N. (1987) Topography of peptidoglycan synthesis during elongation and polar cap formation in a cell division mutant of Escherichia coli MC4100. J. Gen. Microbiol. 133, 575586.
  • 4
    Wientjes, F.B. and Nanninga, N. (1989) Rate and topography of peptidoglycan synthesis during cell division in Escherichia coli: Concept of a leading edge. J. Bacteriol. 171, 34123419.
  • 5
    Nanninga, N. (1991) Cell division and peptidoglycan assembly in Escherichia coli. Mol. Microbiol. 5, 791795.
  • 6
    de Pedro, M.A., Quintela, J.C., Holtje, J.V. and Schwarz, H. (1997) Murein segregation in Escherichia coli. J. Bacteriol. 179, 28232834.
  • 7
    Cook, W. and Rothfield, L.I. (1994) Development of potential division sites in FtsA filaments of E. coli. Mol. Microbiol. 14, 497503.
  • 8
    Cook, W. and Rothfield, L. (1994) Early stages in development of the E. coli division site. Mol. Microbiol. 14, 485495.
  • 9
    Taschner, P.E.M., Huls, P.G., Pas, E. and Woldringh, C.L. (1988) Division behavior and shape changes in isogenic ftsZ, ftsQ, ftsA, pbpB, and ftsE cell division mutants of Escherichia coli during temperature shift experiments. J. Bacteriol. 170, 15331540.
  • 10
    de Jonge, B., Wientjes, F., Jurida, I., Driehuis, F., Wouters, J. and Nanninga, N. (1989) Peptidoglycan synthesis during the cell cycle of E. coli: composition and mode of insertion. J. Bacteriol. 171, 57835794.
  • 11
    Obermann, W. and Holtje, J.V. (1994) Alterations of murein structure and of penicillin–binding proteins in minicells from Escherichia coli. Microbiology 140, 7987.
  • 12
    Driehuis, F. and Wouters, J. (1987) Effect of growth rate and cell shape on the peptidoglycan composition in E. coli. J. Bacteriol. 169, 97101.
  • 13
    Harz, H., Burgdorf, K. and Höltje, J.-V. (1990) Isolation and separation of the glycan strands from the murein of Escherichia coli by reversed-phase high-performance liquid chromatography. Anal. Biochem. 190, 120128.