Wall Peptidoglycan Biosynthesis - American Society for Microbiology

JOURNAL OF BACTERIUOLOGY, May 1988, p. 2197-2201

Vol. 170, No. 5

0021-9193/88/052197-05$02.00/0 Copyright © 1988, American Society for Microbiology

murH, a New Genetic Locus in Escherichia coli Involved in Cell Wall Peptidoglycan Biosynthesis DEXI DAI AND EDWARD E. ISHIGURO* Department of Biochemistry and Microbiology, University of Victoria, Victoria, British Columbia, Canada V8W 2 Y2 Received 28 September 1987/Accepted 10 February 1988

A temperature-sensitive mutant of Escherichia coli defective in peptidoglycan synthesis was characterized. The incorporation of radiolabeled'meso-diaminopimelate into peptidoglycan by the mutant was inhibited at the restrictive growth temperature, res'ulting in autolysis. The defective step appeared to be part of the terminal stage in peptidoglycan synthesis involving the incprporation of disaccharide peptide units into the wall peptidoglycan. The mutation was assigned to a new locus, designated murH, at 99.2 min on the E. coli linkage map.

The synthesis of cell wall peptidoglycan in Escherichia coli occurs in three stages (for a review, see reference 4). (i) A group of soluble enzymes catalyzes the synthesis of the soluble precursors, UDP-N-acetylglucosamine (UDP-GlcNAc) and UDP-N-acetylmuramyl-pentapeptide (UDP-MurNAc-pentapeptide). (ii) Membrane-bound enzymes catalyze the sequential translocation of phospho-MurNAc-pentapeptide and GlcNAc to a glycosyl carrier lipid in the membrane, undecaprenyl-phosphate. (iii) The disaccharide-pentapeptide units are then transferred from the lipid carrier to an acceptor site in the nascent wall peptidoglycan. The terminal steps of this process are catalyzed by one or more penicillin-binding proteins, presumably on the outer face of the cytoplasmic membrane. The penicillin-binding proteins involved are bifunctional enzymes which exhibit both transpeptidase and transglycosylase activities. The mechanistic details of the terminal steps in peptidoglycan synthesis in vivo are not well understood. An important consideration in this respect is that newly synthesized peptidoglycan undergoes a maturation process which results in an increased level of cross-linkage and a higher content of covalently bound lipoprotein. In E. coli, the various soluble enzyme activities involved in UDP-MurNAc-pentapeptide synthesis appear to be tightly coupled during normal growth. Thus', the end product, UDP-MurNAc-pentapeptide, accounts for over 90% of the total UDP derivatives, and' accumulations of the other intermediates in the pathway are relatively insignificant (7). Furthermore, the rate of UDP-MurNAc-pentapeptide synthesis is normally tightly coupled to the rate at which this intermediate is utilized by the particulate enzymes involved in lipid intermediate synthesis and peptidoglycan polymerization; i.e., the UDP-MurNAc-pentapeptide pool size remains relatively constant during steady-state growth. Evidence has been presented indicating that UDP-MurNAcpentapeptide regulates its own synthesis by feedback inhibiting an early enzyme in the biosynthetic pathway, and this may be largely responsible for the coupling between the soluble and particulate enzyme systems during normal steady-state growth (7, 10). The UDP-MurNAc-pentapeptide biosynthetic pathway also appears to be closely

coordinated with the UDP-GlcNAc biosynthetic pathway, but the basis for this is still unknown (E. Ishiguro, unpublished data). When E. coli is subjected to amino acid deprivation, the activities of both the soluble and particulate enzyme systems become regulated by the stringent control mechanism (6, 16). Thus, amino acid-deprived relA+ bacteria accumulate guanosine 5'-diphosphate 3'-diphosphate (2, 3), which apparently mediates the concomitant inhibition of peptidoglycan synthesis. Furthermore, if RelA function is interfered with, e.g., by treating the amino acid-deprived relA+ bacteria with stringent control antagonists such as chloramphenicol or by introducing a mutation in relA, the activities of both the soluble and particulate enzymes are relaxed. As noted above, the synthesis of UDP-MurNAc-pentapeptide and its utilization by the particulate' enzyme system are normally tightly coupled. However, this relationship is uncoupled when peptidoglycan synthesis is relaxed in amino acid-deprived bacteria, e.g., by chloramphenicol treatment, and UDP-MurNAc-pentapeptide accumulates (7). We have therefore proposed that guanosine 5'-diphosphate 3'-diphosphate is responsible for preventing the overproduction of the UDP derivatives during the stringent response, i.e., the feedback inhibition control mechanism operating in normal growing cells is apparently ineffective in this regard (7). On the basis of the preceding information, a general method for localizing steps in peptidoglycan biosynthesis which are blocked either through the action of antibiotics or by mutations has been described (5). This procedure involves tracing the incorporation of a suitable radiolabeled peptidoglycanspecific compound by amino acid-deprived bacteria in which peptidoglycan synthesis has been relaxed. This approach is actually a refinement of a technique originally developed through the pioneering efforts of Lugtenberg and his coworkers (9, 10) and is further rationalized below. Mutants of E. 'coli defective in a variety of steps in peptidoglycan synthesis have been isolated and characterized (11-13, 15, 17-19, 20). We report here what appears to be a new genetic locus involved in peptidoglycan synthesis, which we have designated murH. Using the procedure referred to above, we present evidence indicating that the murH mutant was defective in a biosynthetic step occurring after the synthesis of the lipid-linked intermediates, i.e., in the last stage of peptidoglycan synthesis. However, murH

* Corresponding author. 2197

2198

DAI AND ISHIGURO

was mapped at 99.2 min on the E. coli linkage map (1), and this location does not coincide with any of the known loci for penicillin-binding proteins or other peptidoglycan biosynthetic enzymes.

J. BACTERIOL.

was terminated by applying duplicate 0.1-ml amounts from each sample onto Whatman 3MM filter paper disks and immersing these in cold 10% trichloroacetic acid. The filters were then processed and counted as previously described

(6).

MATERIALS AND METHODS

E. coli K-12 strains and culture conditions. Strain VC440 was a spontaneous murH mutant derived from our laboratory strain, VC7 (thi-J lysA23). Strains VC462 (thi-J lysA23 dapD2 murH zaa-J::TnS zaa-J00::TnJO) and VC482 (thi-J lysA23 dapD2 murH zji-J0J::TnJO) were constructed by transduction of the murH allele into strain LD5 (6) with the aid of murH-linked transposon insertions as described below. The strains used in mapping experiments were CP78 (thi-J thr-i leu-6 his-65 arg46), JB1 (metBIl uxuAl), and JEF8 (thr-31 carB8 metBI), and these were kindly provided by B. J. Bachmann from the E. coli Genetic Stock Center. Bacteria were grown in M9 minimal medium with nutritional supplements as previously described (6) or in tryptic soy broth (Difco Laboratories, Detroit, Mich.). Cultures were incubated in water bath shakers at the indicated temperatures, and growth was monitored with a Klett-Summerson colorimeter with either a blue (for M9 minimal medium) or a green (for tryptic soy broth) filter. In genetic experiments, Davis minimal agar (Difco) containing appropriate supplements or tryptic soy agar was used. Genetic techniques. Conjugation and bacteriophage P1 vir-mediated generalized transduction were performed according to the methods of Miller (14). Genetic linkages were calculated as described by Wu (21). The various TnlO insertions linked to murH were obtained by screening random TnJO insertion pools prepared on a protrophic strain, W3110, with the phage X derivative NK55 as described by Kleckner et al. (8). The screening procedure involved the P1 vir-mediated transduction of the random TnJO insertions into the murH mutant strain and the subsequent selection of appropriate transductants on tryptic soy agar containing 20 ,ug of tetracycline per ml at 42TC. Solubilization of radiolabeled peptidoglycan. Peptidoglycan was labeled with [G-3H]diaminopimelic acid ([3H]DAP; Amersham Canada Limited, Oakville, Ontario, Canada). Strain VC462 (murH) was grown in M9 minimal medium at 30°C for two doublings to a density of 2 x 108 cells per ml. [3H]DAP was added to a concentration of 0.3 jig/ml (1.1 After one doubling, labeling was terminated by RCi4iug). adding unlabeled DAP at 25 ,ug/ml. The labeled culture was divided into three portions (see Fig. 1B). One portion was kept at 30°C, and the other two portions were shifted to 420C.'One of the cultures shifted to 42°C was simultaneously subjected to isoleucine deprivation (by adding 500 ,g of L-valine per ml) and chloramphenicol (200 ,ug/ml) treatment to determine the effect of inhibiting protein synthesis on solubilization of radiolabeled peptidoglycan. At indicated times, 0.5-ml samples were pipetted into equal volumes of boiling sodium dodecyl sulfate and boiled for 30 min. The sodium dodecyl sulfate-insoluble fractions were collected on Millipore filters (pore size, 0.45 ,um). The filters were rinsed with distilled water, dried, and counted in a Beckman LS3145T liquid scintillation counter by using a toluenebased scintillation cocktail. Peptidoglycan synthesis. Bacteria were grown in M9 minimal medium at 30 or 42°C. At indicated times, 0.3-ml samples of the cultures were removed and incubated for 5 min at the apprQpriate temperature (i.e., either 30 or 42°C) in the presence of 1 ,uCi (0.2 ,;g) of [3H]DAP per ml. Labeling

In attempts to identify the step in peptidoglycan synthesis which was blocked by the murH mutation, the cellular distribution of [3H]DAP incorporated at the permissive and nonpermissive temperatures was compared. Strain VC462 was grown in M9 at 30°C to a density of 4 x 108 cells per ml. The culture was divided into three portions (see Table 1). One portion was maintained at 30°C without further treatment. Two portions were subjected simultaneously to isoleucine deprivation and chloramphenicol treatment; one of these was incubated at 30°C, and the other was shifted to 420C. [3H]DAP was added to all cultures at 0.3 ,ug/ml (1.1 ptCi/,g). After 30 min of labeling, cell samples were analyzed by paper chromatography as described previously (7, 16). Briefly, cells from 5-ml samples were collected by centrifugation and suspended in 50 ,ul of ice-cold distilled water. The samples were applied to a sheet of Whatman 3MM chromatography paper (57 cm long), and the chromatogram was developed in isobutyric acid-1 M NH40H (5:3) for 16 h. In this system, cell-bound peptidoglycan remains at the origin. On the other hand, the solvent extracts the UDP-MurNAcpeptides and the lipid intermediates which migrate to Rf values of 0.2 and 0.8, respectively. Where necessary, the development time was extended beyond 48 h to separate UDP-MurNAc-tripeptide from UDP-MurNAc-pentapeptide (7). The amounts of radioactivity associated with these areas were determined. The components of the lipid intermediate fractions were quantified after paper chromatography of mild acid hydrolysates as described in detail elsewhere (16). Cultures of VC462 were labeled with [3H]DAP at 30 and 42°C for 30 min as described above. Suspensions of labeled cells were boiled for 4 min to release soluble radiolabeled material. The washed cell suspensions were divided into two equal parts. One part, referred to below as the unhydrolyzed control, was applied directly to a sheet of Whatman 3MM chromatography paper. The other part was hydrolyzed with 10% acetic acid in sealed vials at 1050C for 20 min and then applied to chromatography paper. The chromatogram was developed in isobutyric acid-1 M NH40H (5:3) for 16 h and analyzed as described previously (16). RESULTS Phenotypic properties. The murH mutant was discovered fortuitously during the screening of a collection of spontaneous temperature-sensitive mutants which apparently had defects in peptidoglycan metabolism. The temperature sensitivity was apparently associated with cellular autolysis. Thus, when cultures of strain VC462 growing in either complex (data not shown) or minimal (Fig. 1A) medium at 30°C were shifted to 420C, a decrease in culture turbidity occurred within one doubling time (based on the growth rate of the wild-type parent at 42°C). This loss in turbidity was correlated with the solubilization of peptidoglycan which had been prelabeled with [3H]DAP (Fig. 1B). The lysis of strain VC462 at 42°C was prevented when growth was inhibited by amino acid deprivation, chloramphenicol treatment, or (Fig. 1A aqd B) by a combination of both methods. Since the temperature-dependent lysis exhibited by strain VC462 resembled the phenotypes of mutants defective in peptidoglycan synthesis (11-13, 15, 17-19, 20), the effect of

murH LOCUS OF E. COLI

VOL. 170, 1988

temperature upshift on peptidoglycan synthesis was determined. Peptidoglycan synthesis in VC462 was inhibited by 75% within 10 min after upshift from 30 to 42°C (Fig. 2). This is significantly earlier than the time of onset of lysis, suggesting that lysis was a consequence of the inhibition of peptidoglycan synthesis. Temperature-resistant revertants of VC462 occurred at a frequency of about 3 x 10-6, suggesting that the mutant phenotype was attributable to a single mutational event. Site of mutational block. The step in peptidoglycan synthesis blocked by the murH mutation was localized by using cultures of VC462 which were simultaneously amino acid deprived and chloramphenicol treated (Table 1). The main steps in peptidoglycan synthesis in such nongrowing bacteria can be quantitatively monitored by following the incorporation of [3H]DAP into the various cell fractions (5, 16). Another important consideration here was that the progress of peptidoglycan synthesis in strain VC462 could be followed at the nonpermissive temperature without the complication of lysis by using amino acid-deprived cells treated simultaneously with chloramphenicol, i.e., because lysis at 42°C was inhibited under these conditions (Fig. 1A and B). Nevertheless, these experiments were restricted to a 30-min period (i.e., before lysis became evident in cultures growing at 42°C) as an added precaution. Cultures of VC462 were labeled with [3H]DAP at 30 and 42°C for 30 min, and the amounts of radioactivity incorporated into the various cellular fractions were determined.

(A

c

a4)

-I.

0

co

0

0) E

E

Q

C.)

3 H ours

FIG. 1. Temperature-dependent autolysis of strain VC462 (murH) and its prevention by inhibiting protein synthesis. Cultures were incubated at 30 (0), 42 (0), or 42°C (A) with 500 1Lg of L-valine and 200 jig of chloramphenicol per ml. (A) Lysis was determined turbidimetrically. (B) Cellular peptidoglycan was prelabeled with [3H]DAP, and solubilization of the radiolabeled peptidoglycan was measured as an index of lysis.

2199

L-a

onww...0

O..---.O..---O

0 08

E.4

E a.

b 0

10

20 Minutes

30

FIG. 2. Effect of growth temperature on peptidoglycan synthesis in strain VC462 (murH). Samples from cultures incubated at 30 (0) and 42°C (0) were pulse-labeled with [3H]DAP for 5 min, and amounts incorporated into peptidoglycan fractions were determined. Average values of duplicate samples are plotted as a function of the sampling times.

The following results are from controls incubated at 30°C which confirm that peptidoglycan synthesis in VC462 was under stringent control (Table 1). (i) The amounts of radioactivity incorporated into the peptidoglycan fractions by the untreated control culture and the chloramphenicol-treated, amino acid-deprived culture at 30°C were the same, indicating that chloramphenicol treatment relaxed peptidoglycan synthesis as expected. (ii) A threefold accumulation of labeled UDP derivatives occurred in the treated culture, confirming that RelA function is necessary to prevent the overproduction of the UDP intermediates during amino acid deprivation (7). (iii) The amounts of radioactivity in the lipid intermediate fractions were the same for both of the cultures. To localize the step blocked in the murH mutant, the cellular distribution of radioactivity in amino acid-deprived cells treated with chloramphenicol at 42°C was compared with the data obtained from cells treated similarly at 30°C. The incorporation of [3H]DAP into the peptidoglycan fraction was inhibited at 42°C by about 40% (Table 1). Otherwise the labeling patterns for the two sets of cells were identical. These results suggest that a late step in biosynthesis was mutationally blocked in VC462. To reinforce this conclusion, the labeled peptidoglycan intermediates in the chloramphenicol-treated cells were characterized further. Over 90% of the UDP derivatives accumulated during chloramphenicol treatment at either 30 or 42°C was accounted for as UDP-MurNAc-pentapeptide (data not shown). Similar results have been reported for wild-type E. coli strains (7). This proves that the synthesis of neither UDP-MurNAc-pentapeptide nor UDP-GlcNAc (because UDP-GlcNAc is a precursor for the synthesis of UDPMurNAc) was defective in VC462. The fact that there was no depletion of radiolabel from the lipid intermediate fraction also indicates that the synthesis of UDP derivatives was unaffected by the mutation in murH. The lipid intermediates in cells labeled as described in Table 1 were quantified further. Typical results are shown in Table 2. The intracellular pools of labeled UDP derivatives and DAP were completely removed by boiling and washing the cell preparations. The peptidoglycan and lipid intermediate fractions were the only detectable labeled components remaining under these conditions (the unhydrolyzed samples in Table 2). Portions of the boiled preparations were subjected to mild acid hydrolysis and then analyzed by paper chromatography. All of the labeled peptidoglycan was con-

2200

DAI AND ISHIGURO

J. BACTERIOL.

TABLE 1. Cellular distribution of [3H]DAP incorporated by strain VC462 at permissive and nonpermissive temperatures Amt of [3H]DAP (103 cpm/mg of cell dry wt) Temp (°C)

Treatment

in the following fraction:

PeptidglycanUDP Uderivatives

Peptidoglycan

43.8 44.2 25.5

30 30 42

Untreated control Amino acid-deprived cells plus chloramphenicol cells

3.6 11.4 16.8

Lipid

intermediates

1.1 1.1 1.3

a Mixture of UDP-MurNAc-tripeptide and UDP-MurNAc-pentapeptide.

sistently recovered after mild acid hydrolysis, and this as a convenient internal standard. Hydrolysis of the lipid intermediate fraction was quantitative (i.e., no intact lipid intermediate was detectable following hydrolysis), but only about 60% of the radioactivity from the original starting material could be subsequently accounted for. The major product was identified as disaccharide-pentapeptide, and the only other labeled derivative was unidentifiable. We are certain, however, that the latter was not MurNAc-pentapeptide. In fact, no labeled MurNAc-pentapeptide was detected in these samples. These results suggest that the disaccharide-pentapeptide derivative was the major lipid intermediate. It is noteworthy that the lipid intermediate fractions from both 30°C- and 42°C-treated cells were identical. Therefore, although our analysis of the lipid intermediate fractions was far from ideal in terms of recovery, we feel that lipid intermediate synthesis was probably normal in VC462. If so, the defect must be in some step involving the incorporation of disaccharide-peptide units from lipid carrier derivatives into peptidoglyan. Mapping of murH. To facilitate the mapping of murH, TnlO insertions linked to murH were obtained. These TnJO insertions were transduced into a collection of Hfr strains. The locations of the TnJO insertions were then determined to be in the 0-min region of the linkage map through a series of conjugation experiments involving these Hfr strains and various appropriate recipient strains (data not shown). Figure 3 shows a linkage map determined by phage P1-mediated transduction. The murH locus was assigned to 99.2 min. component served

DISCUSSION Cellular autolysis occurs whenever peptidoglycan synthesis is blocked in growing bacteria. The basis for this is unclear. However, it has been often proposed (4) that peptidoglycan synthesis may involve the coordinated activities of the biosynthetic enzymes and the peptidoglycan hydrolases. One view is that the hydrolase activities become dissociated when a biosynthetic step is blocked, and lysis results. Thus, strain VC462 resembled other previously

described mutants with defects in peptidoglycan synthesis in terms of the lysis phenotype it exhibited at the nonpermissive temperature. This defect was unequivocally established in VC462 by demonstrating the inhibition of [3H]DAP incorporation into peptidoglycan at 42°C. To analyze the function of the murH locus, we traced the incorporation of [3H]DAP in VC462 cultures which were simultaneously amino acid deprived and chloramphenicol treated. As expected, the normal tight coupling between the soluble and particulate enzyme systems referred to above was disrupted under these conditions, and UDP-MurNAcpentapeptide accumulated at the permissive temperature of 30°C. The fact that UDP-MurNAc-pentapeptide also accumulated at the nonpermissive temperature of 42°C clearly indicates that all of the soluble enzyme activities were normal. As already noted, this includes the synthesis of UDP-GlcNAc because this derivative is a precursor of UDP-MurNAc. Further identification of the biosynthetic step blocked by the mutation in murH has posed some difficulties. These are related mainly to the low levels of lipid intermediates normally present in E. coli and to the incomplete recoveries experienced during fractionation of these intermediates. These technical problems will not necessarily be easy to overcome. The maximum level to which the lipid intermediates can accumulate under conditions of relaxed control is fixed by the level of undecaprenylphosphate present, and this appears to be low (unpublished data). Our results suggest that the disaccharide-pentapeptide derivative was the major product synthesized by VC462, and the amounts observed at the permissive and nonpermissive temperatures were identical. We therefore tentatively conclude that the synthesis of lipid intermediates was not affected by the mutation in VC462. The same conclusion was drawn from results (data not shown) obtained with the conventional in vitro assay method using particulate cell envelope preparations. It is also worth noting that the only known mutation affecting lipid intermediate synthesis has been assigned to the mrbB locus at 90 min and is thus unlinked to murH.

TABLE 2. Analysis of lipid intermediate fractions Amt of [3HlDAP (102 cpm/mg of cell dry wt) in the following fraction:

Sample

Unhydrolyzed Hydrolyzed

Growth temp

(OC)

Peptidoglycan

Lipid intermediates

30 42 30 42

492.3 211.4 487.4 201.2

12.1 11.6

Disaccharide-

pentapeptide

Unknown

7.6 6.9

2.2 1.8

VOL. 170, 1988

cuSI

C

5

< x

97.5

murH LOCUS OF E. COLI

7

mliO4

-4---

98 b

1

10/0

99

17

6. 7.

60

40

I

24

68 c

d 46 j 64 10

8.

e

9. 65 f 28

FIG. 3. Linkage map of the murH region. Thie results of the following crosses (recipient x donor) are shown, wiith the number of recombinants scored in each case indicated parenth etically: a, CP78 x VC462 (148); b, JB1 x VC482 (183); c, JEF8 x VC462 (62); d, JEF8 x VC462 (100); e, JEF8 x VC462 (112); f, JEF8 x VC462 (112). For each set of arrows, the vertical line indic ates the position of the selected marker and the arrowheads indicate the positions of the unselected markers. The numbers represent th4 e cotransduction frequencies (%).

The murH allele most likely conferred a defect in some aspect of the terminal step in peptidoglycan s,ynthesis (1). It is possible that the mutation alters a membr*ane-associated function which indirectly affects this aspect of peptidoglycan final steps in synthesis. There are clear indications that th e final peptidoglycan synthesis are complex (4), anId with the assumption that our conclusion is correct, the deztermination of the function of murH should further our under-standing of the e

sptepsgin

10.

11.

12.

Temperature-sensitive

impaired D-alanine:D-alanine ligase. J. Bacteriol. 113:96-104.

14. Miller, J. H. 1972. Experiments in molecular genetics. Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y.

15. Miyakawa, T., H. Matsuzawa, M. Matsuhashi, and Y. Sugino. 1972. Cell wall peptidoglycan mutants of Escherichia coli K-12:

ACKNOWLEDGMENTS

16.

This work was supported by operating grant A67331 to E.E.I. from the Natural Sciences and Engineering Research Coouncil of Canada. D.D. was supported by a University of Victoria Ciraduate Fellow-

17.

ship. We thank Barbara Bachmann for bacterial strainIS.

18.

Genet. 13:393-415. Holtje, J.-V., and U. Schwarz. 1985. BiosyntheEsis and growth of the murein sacculus, p. 77-119. In N. Nanninga (ed.), Molecular cytology of E. coli. Academic Press, Inc., Ne%w York. 5. Ishiguro, E. E. 1982. Inhibition of uridine 5''-diphosphate-N-

acetylmuramyl-L-alanine synthetase by P-chloro-L-alanine in Escherichia coli. Can. J. Microbiol. 28:654-659. Ishiguro, E. E., and W. D. Ramey. 1976. Stringent control of peptidoglycan biosynthesis in Escherichia coli K-12. J. Bacteriol. 127:1119-1126. Ishiguro, E. E., and W. D. Ramey. 1978. Involvement of the relA gene product and feedback inhibition in the regulation of UDPN-acetylmuramyl-peptide synthesis in Escherichia coli. J. Bacteriol. 135:766-774. Kleckner, N., J. Roth, and D. Botstein. 1977. Genetic engineering in vivo using translocatable drug-resistance elements. New methods in bacterial genetics. J. Mol. Biol. 116:125-159. Lugtenberg, E. J. J., and P. G. De Haan. 1971. A simple method for following the fate of alanine-containing components in murein synthesis in Escherichia coli. Antonie van Leeuwenhoek J. Microbiol. Serol. 37:537-552. Lugtenberg, E. J. J., L. De Haas-Menger, and W. H. M. Ruyters. 1972. Murein synthesis and identification of cell wall precursors of temperature-sensitive lysis mutants of Escherichia coli. J. Bacteriol. 109:326-335. Lugtenberg, E. J. J., and A. van Schijndel-van Dam. 1972. Temperature-sensitive mutants of Escherichia coli K-12 with low activities of the L-alanine adding enzyme and the D-alanylD-alanine adding enzyme. J. Bacteriol. 110:35-40. Lugtenberg, E. J. J., and A. van SchiJndel-van Dam. 1972. Temperature-sensitive mutants of Escherichia coli K-12 with

low activity of the diaminopimelic acid adding enzyme. J. Bacteriol. 110:41-46. 13. Lugtenberg, E. J. J., mutant and A.ofvan Schijndel-van Dam. 1973. Escherichia coli K-12 with an

process.

LITERATURE CITED 1. Bachmann, B. J. 1983. Linkage map of Esche. richia coli K-12, edition 7. Microbiol. Rev. 47:180-230. 2. Cashel, M. 1975. Regulation of bacterial ppG] pp and pppGpp. Annu. Rev. Microbiol. 29:301-318. 3. Gallant, J. A. 1979. Stringent control in E. ccoli. Annu. Rev.

2201

19.

20.

4.

21.

existence of two clusters of genes, mra and mrb, for cell wall peptidoglycan biosynthesis. J. Bacteriol. 112:950-958. Ramey, W. D., and E. E. Ishiguro. 1978. Site of inhibition of peptidoglycan biosynthesis during the stringent response in Escherichia coli. J. Bacteriol. 135:71-77. Salmond, G. P. C., J. F. Lutkenhaus, and W. D. Donachie. 1980. Identification of new genes in a cell envelope-cell division gene cluster of Escherichia coli: cell envelope gene murG. J. Bacteriol. 144:438-440. Spratt, B. G. 1975. Distinct penicillin binding proteins involved in the division, elongation, and shape of Escherichia coli K12. Proc. Natl. Acad. Sci. USA 72:2999-3003. Suzuki, H., Y. Nishimura, and Y. Hirota. 1978. On the process of cellular division in Escherichia coli: a series of mutants of E. coli altered in the penicillin-binding proteins. Proc. Natl. Acad. Sci. USA 75:664-668. Wisman, H. J. W. 1972. A genetic map of several mutations affecting the mucopeptide layer of Escherichia coli. Genet. Res. 20:65-74. Wu, T. T. 1966. A model for three point analysis of random general transduction. Genetics 54:405-410.