Blocking of bacteriophages phi W and phi 5 with lipopolysaccharides from Escherichia coli K-12 mutants. - PDF Download Free (2024)

Vol. 121, No.2

JouRNAL oF BAcrTERoLOGY, Feb. 1975, p. 465-470 Copyright X) 1975 American Society for Microbiology

Printed in LT.S.A.

Blocking of Bacteriophages qW and 45 with Lipopolysaccharides from Escherichia coli'K-12 Mutants HANS G. BOMAN* AND DAVID A. MONNERI Department of Microbiology, University of Umed, S 901 87 Umed 6, Sweden Received for publication 1 October 1974

In the preceding paper we presented a formula for the composition of lipopolysaccharides (LPS) from Escherichia coli K-12. This formula contains four regions defined from analyses of LPS from four key strains, the parent and mutants which had lost one, two, or three regions of their carbohydrates. Support for the formula was derived from the susceptibility of the key mutants to several bacteriophages. One of these, phage OW, was found specific for strains which had lost region 4. In this paper we described inactivation in vitro of phage 4W and its host-range mutant 45, using LPS devoid of regions 2 to 4. The blocking of 4W was found to require about 0.15 M concentrations of monovalent cations and to be inhibited by low concentrations of calcium and magnesium ions. One particle of phage 4W required 2 x 10-1' g of LPS devoid of region 4 for stoichiometric inactivation. Phage 45 was blocked by both heptose-less LPS (devoid of regions 2 to 4) and glucose-less LPS (devoid of regions 3 to 4) but was unaffected by LPS devoid of region 4. LPS from a heptose-less mutant of Salmonella minnesota showed the same inactivation ability as did LPS from heptose-less strains of E. coli K-12. Lipid A was prepared from LPS containing all four regions. Such lipid A was found to inactivate 05, whereas both the polysaccharide moiety as well as the intact LPS were without effect. It is suggested that lipid A is part of the receptor site for phage 45.

Bacteriophages often exhibit a marked specificity for their host strains, a fact which has found an increasing use for the identification of bacteria by phage typing. The first level at which this species specificity can be demonstrated is in the adsorption of the bacteriophage to the receptor on the cell surface of the host. That this reaction could be studied in vitro was demonstrated 40 years ago by Burnet (3). Further progress had to await the chemical fractionation of the cell envelope carried out by Weidel (17). This work made it possible to demonstrate that the lipopolysaccharide (LPS) was the receptor for some phages, whereas for others, attachment occurred to a lipoprotein. That proteins can also function as receptors was recently shown for phage lambda (13). The wealth of chemical information concerning LPS from certain strains of Salmonella has also been utilized for the study of the phage receptor function and the subject was recently reviewed (5, 14). Using phage blocking in vitro, we previously demonstrated that Escherichia coli K-12 LPS is at least part of the receptor for 'Present address: Dept. of Physiology, South-Western Medical School, 5323 Harry Hines, Dallas, Tex. 75235.

the Wollman phage, 4W (12). We have now tried to optimize the reaction by studying the influence of mono- and divalent cations. In addition to 4W, we have used a host-range mutant, 45, isolated on a strain with heptoseless LPS and LPS from a series of strains which have lost increasing parts of their core polysaccharide. By phage blocking with intact and hydrolyzed LPS, evidence was obtained that indicates that the lipid A part is the receptor for 45. The chemical composition of LPS from the different strains as well as the phage response was studied in the companion paper (2). MATERIALS AND METHODS Bacterial strains, properties of LPS, and growth

media. The strains used, their phage pattern, and the composition of their respective LPS are summarized in Table 1. Details of the strains, the media, the growth conditions, and the preparation of LPS, as well as back references, are given in the preceding paper (2). Salmonella minnesota, strain R595, was grown in closed 2-liter cultures. Before harvest the bacteria were killed by the addition of phenol to a final concentration of 1% (15). Bacteriophages. The Wollman phage, OW, is the strain originally isolated and propagated on E. coli B (10, 11, 18). The two host-range mutants, 43 and 05,

465

466i

BOMAN AND MONNER

TABLE 1. Phage response of E. coli and Salmonella strains useda Strain

D21 D31 D21e7 D215 D21fl D31m3 D31f2

D31m4 R595

LPS regions

1-4 1-3 (4) 1-3 1-3 1-2 1-2 1 1 1

Response to: 3 05

lw

R

S S S S S S S

R R

S R

R

S S S R R

PIA,0 (Asg/ml) pW 05

R > 100 R 0.03 R 0.04 R 0.05 S > 50 S S > 50 S > 50 R

J. BACTERIOL.

tion tubes were diluted the necessary amount in DM (2) before plating. The effectiveness of a given LPS preparation was determined according to Adams (1) using a series of different concentrations of LPS and expressed as PIA50, the concentration giving 50% inactivation.

> 50 > 100 1.5 1.5 1.5 0.8 0.9

a Details about these strains are given in the companion paper (2). The LPS regions refer to the formula given in Fig. 1 of the companion paper. R, Phage resistance; S, phage sensitivity. PIA5O values, defined as in Materials and Methods, were extrapolated from Fig. 4 and 6 and similar experiments. The PIA5O value for strain D31 refers to LPS from log-phase cells.

were isolated and propagated as described elsewhere (2, 11). Stoichiometric studies of OW inactivation by LPS required high titers of phage. These stocks were made at 30 C from 1 liter of lysate in medium E (2) with 0.2% glucose. After shaking with a small amount of chloroform and chilling, debris were removed by centrifugation at 10,000 x g for 10 min. The phages were pelleted by centrifugation at 75,000 x g for 55 mm using the SW25 rotor in a Spinco L2-65B ultracentrifuge. The pellets were resuspended in 0.1 volume of 0.05 M N-2-hydroxyethylpiperazine-N'-2ethanesulfonic acid buffer (HEPES buffer, Sigma Chemical Co.), pH 7.4, with 2 x 10-3 M ethylenediaminetetraacetic acid (EDTA). The soft agar overlay technique of Adams (1) was used for plaque assays. Plates were generally read after 4 to 6 h at 37 C, before the development of the halo. Assays for inactivation of phages by LPS. Stock solutions of LPS (1 mg/ml in distilled water) were stored frozen. For an assay, the solutions were thawed, warmed to 55 C for 5 min for solubilization, then diluted serially in the desired buffer to twice the concentrations to be used in the reactions. The phage to be tested was diluted in the same buffer to about 4,000 plaque-forming units per ml. After prewarming to assay temperature, 0.5 ml of each solution was mixed in a tube and incubated without shaking for 60 min. For phage 05 room temperature (21 to 23 C) was used, for other phages incubations were always at 37 C. Phage mixed with buffer alone was used as a control. The reaction was stopped by placing the tubes in an ice bath, after which 0.1-ml portions from each tube were placed directly in soft agar with indicator bacteria. Duplicate or triplicate samples were assayed from each tube and all reactions were run in duplicate. With higher phage titers (used for stoichiometric studies), 0.1-ml portions from the reac-

RESULTS Three LPS-specific phages and their properties. In Fig. 1 of the preceding paper (2), we suggested the schematic LPS structure for E. coli K-12. The formula contains four regions and is based on the carbohydrate compositions of LPS from a set of key mutants. A comparison of the growth of several phages on these different mutants showed that the Wollman phage OW was quite specific for strains with an LPS devoid of region 4. From /W we isolated two host-range mutants: 03, which is adjusted for growth on K-12 strains containing region 4; and 05, which can grow on heptose-less LPS containing only region 1. We have shown earlier that LPS from strain D31 (partly devoid of region 4) inactivated OW, whereas 100-times larger quantities of LPS from the parental strain D21 was without any effect (12). We have now extended this work and studied phage blocking in vitro with LPS devoid of regions 2 to 4. We concentrated the work on OW and 05 which in vivo show a more pronounced specificity than 03 (2). Both phage stability and the phage blocking by LPS were found dependent on the ionic conditions used. It was necessary therefore to first investigate the phage stability. Phage OW turned out to be sensitive to low ionic strength and to dilution. However, the addition of either 0.15 M NaCl or 2 x 10-3 M EDTA provided a reasonable stabilization. Phage 05 showed stability requirements contrary to those of OW. As illustrated in Fig. 1, the withdrawal of CaCl2 by the addition of EDTA decreased the stability of 05. To minimize phage instability further, experiments with 05 were performed at room temperature. Phage 03 was not very sensitive to mono- or divalent cations, but during dilutions we always observed a rapid initial 30 to 50% drop after which the phage seemed to stabilize itself. This latent instability may be related to the fact that 03 always produced noninfectious tailless particles (11). Influence of mono- and divalent cations on phage blocking by LPS. Figure 2 shows that phages OW (circles, left part) and 03 (squares, right part) both required NaCl for an effective reaction with LPS at a concentration of 1 ,g/ml. For phage OW there were marked differences

BLOCKING BACTERIOPH'AGES WITH E. COLI LPS

VOL. 121, 1975

tions of 10-3 and 2.5 x 10-2 M, respectively (Fig. 3). In further experiments with kW we therefore added EDTA at a concentration of 2 x 10-3 M. The blocking of phage 05 by heptose-less LPS from strain D31m4 was hardly affected by low concentrations of NaCl, but above 0.3 M blocking was prevented as was the case with the other

100 80 60 -1

40

20 is

n

phages. Blocking of phage qW by LPS devoid of region 4. The inactivation of phage kW versus increasing concentrations of LPS from three

10 0 to

4'

467

5

different strains believed to have mutations

0~ 3

affecting the IpsA gene is shown in Fig. 4. The same PIA60 of 0.04 to 0.06 ,g/ml (see Materials and Methods) was obtained for LPS from logphase cells of strains D21e7, D215, and D31.

'U a,

2

30

60

90

120

Incubation time (min)

FIG. 1. The stability of phage 45 during incubation at different temperatures and in the presence of 10-' M CaCl,. After 15 min (as indicated by the arrow), EDTA was added to the 37 C samples giving a final concentration of 2 x 10-3 M.

This value is about 20 times lower than the value found for LPS from stationary-phase cells of strain D31. Strain D31 produces two types of LPS (2), one seemingly like that produced by IpsA strains and one like wild-type LPS. In log-phase cells the first type dominated, in stationary-phase cells more wild-type LPS was 100 -

- 10

D215-LPS 80 l

c s

60

w

40

3

or

-iL

20

J%O .001

D31-D215-LPS3-

O

L7 60

19 f

.40 20

D31 LDa. b 0.10

0.01 0.10 Molar concentration of NaCI 1.00

CD -

a._ -

.

0.01

80 3 60

100

.80

1.00

FiG. 2. Effects of NaCI concentration on the blocking of phages OW (circles) and 03 (squares) by LPS (0.5 ug/ml) from strains D215 (open symbols) and D31 (filled symbols).

ui

MgCI2 CaC12

40

20 0

S

0110-5

.

w

1 0-4

10o-3

Molar concentration of CaC12 and MgCI2

FIG. 3. Effects of CaCI, (filled circles) and MgCl, (open circles) on the blocking of phage W by LPS (0.5 jg/ml) from strain D215.

between strains D31 and D215. Maximal inactivation occurred at the lowest salt concentration in the case of D31, whereas with D215 the 100 optimal phage blocking was between 0.1 and 3 0.15 M NaCl. Maximal inactivation of phage 03 80 occurred at 0.15 M NaCl with both LPS preparations. A dependence on salt for inactivations c 60 was also found with phages $4 (10) and T4, the .2_ latter having a slightly higher optimal salt to'5 40 > concentration (0.25 M). Similar results were obtained using KCI and potassium phosphate '5 20 buffer, implicating ionic strength or possibly 0.01 0.03 0.1 0.3 specific effects of Na+ and K+ ions as the 1.0 3.0 causative factor. Concentration of LPS (,g/mi) In contrast, divalent cations seem to "poison" FIG. 4. Inactivation of phage 6W by increasing the ability of LPS to block 4W. Complete concentration of LPS from log-phase cells of strains destruction of the phage blocking of the LPS D21e7 (O), D215 (O), and D31 (0) LPS and from was obtained with Ca2+ or Mg2+ at concentra- stationary-phase cells of strain D31 (0). -0

10.0

AAR

BOMAN AND MONNER

made (2). The latter form of LPS is known to be inactive with OW (12). Determinations of PIA,0 measure the effectiveness of a given LPS sample under a given time period but, because of the very low concentrations of the reactants, yield no information about the stoichiometry of the reaction. If phage and LPS concentrations are adjusted to give saturation of the phage with LPS within a very short time after mixing, it becomes possible to determine the amount of LPS needed to inactivate a given number of phages. Figure 5 shows such an end-point titration of OW by LPS from strain D215. The value obtained, 2 x 10-16 g of LPS per phage, is slightly lower than that determined by Jesaitis and Goebel for T4 inactivation with LPS from Shigella (4). Blocking of phage 05 with LPS devoid of regions 2 to 4 or 3 to 4. Such strains are commonly referred to as glucose-less, devoid of regions 3 to 4, and heptose-less, containing only region 1. The blocking of 05 (triangles) was nearly the same for LPS from the Salmonella mutant R595 and our heptose-less and glucoseless mutants of E. coli K-12 (Fig. 6). Also, the small amount of glucose present in LPS from strain D31m3 (2) was without any effect on the PIA6O value (Table 1). With OW (circles) there was no blocking with LPS from strains D21fl and D21f2 (Fig. 7 and Table 1). The results in Fig. 6 suggest that lipid A is

-0

100

e

80

J. BACTERIOL.

c

60 04

40

02111 LnR595

_

(Ti

V

20

0W a D21f2 I

0.5

1.5

5

(I

p

15 50 Concent rat i on of LPS (p.g/ml) FIG. 6. Inactivation of phage 05 (triangles) and OW (circles) with increasing concentrations of heptoseless LPS (containing only region 1) prepared from E. coli K-12 strain D21f2 (A), and S. minnesota strain R595 (A), and of glucose-less LPS from E. coli K-12 strain D21fl (A).

LA

100

80 60 to

m

40

c '0

20 0-

10

o d-

0.

_

/ /

1

6 1

m

2.5 5

20

40 60 Incubalion time(min) FIG. 7. Time curve for the inactivation of phage cp5 with (50 Ag/ml) lipid A (A), polysaccharide (PS) (X) prepared by mild acid hydrolysis of LPS obtained from strain D21, controls with intact LPS from strain D21 (A), and phage incubated in buffer (Ax). The lipid A, the polysaccharide, and the LPS samples all contained bovine serum albumin (0.1 mg/ml) and 2 x 10-3 M EDTA.

10

25

D215 LPS conc.(,ug/ml)

FIG. 5. Stoichiometric relationship between inactivated particles of phage kW and added amount of LPS from strain D215. A high titer phage suspension (4.7 x 101 plaque-forming units per ml) was incubated for 60 min at 37 C with the concentrations of LPS indicated. The number of plaque-forming particles was determined afterwards. The endpoint at about 10 Ag corresponds to 2 x 10-16 g of LPS per phage particle.

part of the phage receptor. To examine this possibility, LPS from strain D21 was subjected to mild acid hydrolysis. The lipid A liberated was removed from the water-soluble polysaccharide fraction by centrifugation (2). After lyophilization, the degradation products were tested for their phage-blocking activity at a concentration of 50 jg/ml. Intact LPS from strain D21 and phage alone were included as controls. The results in Fig. 7 clearly show that lipid A from strain D21 could inactivate phage

VOL. 121, 1975

BLOCKING BACTERIOPHAGES WITH E. COLI LPS

469

45, whereas no such effect was obtained with soluble during the reaction conditions emeither intact D21 LPS or with the polysaccha- ployed. Differences in solubility may explain why our PIA50 values for heptose-less and gluride fraction. cose-less LPS were higher than for LPS devoid DISCUSSION of region 4 (Table 1) or the values reported for Our preliminary results showed that the ionic carbohydrate-rich Salmonella LPS (5). It is also composition and the ionic strength were of possible that the active receptor on the bacterial importance both for the phage stability and the cell surface is made up from many lipid A phage blocking reaction. However, since the moieties arranged in a topologically defined optimal conditions were different, the phage way. Poorly solubilized lipid A may contain such a native blocking reaction had to be performed with only very few aggregates with specific conforassuming Thus, arrangement. somewhat unstable phage particles, thereby likely increasing the experimental errors. De- mational as well as topological requirements, spite this limitation in accuracy there was a lipid A could be part of the receptor for both good agreement between the specificity found phage 4)5 and its parental phage 4W. in vivo for phages 4W and 45 (2) and the LPS ACKNOWLEDGMENTS blocking in vitro shown in Fig. 4 to 6. Blocking We would like to thank Kerstin Helander for technical of these phages can therefore be used as a bioassay in future studies of some types of LPS. assistance. The work was supported by grants from the Swedish Similar blocking experiments could not be ob- Natural Science Research Council (Dnr 2453) and the Swedtained with phages C21, C21-3, or FP3. ish Cancer Society (Project 157). For most other biological activities of LPS it LITERATURE CITED is now well established that it is the lipid A part which is the active moiety (8). It seems clear 1. Adams, M. H. 1966. Bacteriophages. Interscience Publishers, London. also that the active forms of LPS and lipid A H. G., and D. A. Monner. 1975. Characterization often are polymeric aggregates (16) and that 2. Boman, of lipopolysaccharides from Escherichia coli K-12 mulipid disrupting agents, like sodium cholate or tants. J. Bacteriol. 121:455-464. polymyxin B, can destroy the toxic activity of 3. Burnet, F. M. 1934. The phage-inactivating agent of bacterial extracts. J. Pathol. Bacteriol. 38:285-299. LPS (7, 9) as well as the phage-blocking capacM. A., and W. F. Goebel. 1955. Lysis of T4 phage ity (12). So far the phage receptor function of 4. Jesaitis, by the specific lipocarbohydrate of phase II Shigella LPS has generally been attributed to the carbosonnei. J. Exp. Med. 102:733-752. hydrate part of the molecule although Lindberg 5. Lindberg, A. A. 1973. Bacteriophage receptors. Annu. Rev. Microbiol. 27:205-241. in a recent review (6) considers the evidence as 6. Lindberg, A. A., and C. G. Hellerqvist. 1971. Bacterioless than convincing. phage attachment sites, serological specificity, and Is lipid A also an essential part of the receptor chemical composition of the lipopolysaccharides of for 05? Blocking experiments with glucose-less semirough and rough mutants of Salmonella typhimurium. J. Bacteriol. 105:57-64. and heptose-less LPS (Fig. 6) as well as with J., and W. E. Inniss. 1969. Electron microscopy of lipid A (Fig. 7) indicate that this could be the 7. Lopes, of polymyxin on Escherichia coli lipoplysacchaeffect to case. However, the parental phage 4W seems ride. J. Bacteriol. 100:1128-1130. have a clear requirement for the carbohydrates 8. Liideritz, O., C. Galanos, V. Lehmann, M. Nurminen, E. T. Rietschel, G. Rosenfelder, M. Simon, and 0. Westof regions 2 and 3 of the LPS molecule (Table phal. 1973. Lipid A: chemical structure and biological 1). It seems unlikely that one or two mutational activity. J. Infect. Dis. 128(Suppl):17-29. events could change the specificity of adsorp- 9. McIntire, F. C., G. H. Barlow, H. W. Sievert, R. A. tion from a hydrophilic polysaccharide to the Finley, and A. L. Yoo. 1969. Studies on a lipopolysaccharide from Escherichia coli. Heterogeneity and hydrophobic lipid A. We therefore suggest that mechanism of reversible inactivation by sodium deoxyfor 45 require 4W adsorption and phages 8:4063-4067. cholate. strictly defined but somewhat different second- 10. Monner, D.Biochemistry A., and H. G. Boman. 1970. Female strains of that and for structures lipid A, Escherichia coli K12 as selective hosts for the isolation ary and tertiary of female specific mutants of phage OH. Biochem. these conformations are dependent on the presRes. Commun. 39:1017-1020. Biophys. in of the carbohydrates regions ence or absence D. A., and H. G. Boman. 1974. Host dependent 2 and 3 of the LPS. This assumption is consist- 11. Monner, properties of coliphage OW and its female specific ent with the facts that the polysaccharide host-range mutants, 43 and 44. J. Gen. Virol. 24:67-75. H. G. Boman. 1971. moiety contains phosphorus (Fig. 4; 2), and that 12. Monner, D. A., S. Jonsson, andof Escherichia coli K-12 Ampicillin-resistant mutants the LPS blocking of OW was "poisoned" by low alterations affecting mating with lipopolysaccharide concentrations of divalent cations (Fig. 3). It ability and susceptibility to sex-specific bacteriomust also be emphasized that lipid A as well as phages. J. Bacteriol. 107:420-432. heptose-less and glucose-less LPS are poorly 13. Randall-Hazelbauer, L., and M. Schwartz. 1973. Isola-

470

BOMAN AND MONNER

tion of the bacteriophage lambda receptor from Escherichia coli. J. Bacteriol. 116:1436-1446. 14. Rapin, A. M. C., and H. M. Kalckar. 1971. The relation of bacteriophage attachment to lipopolysaccharide structure, p. 267-307. In G. Weinbaum, S. Kadis, and S. J. Ajl (ed.), Microbial toxins, vol. IV. Academic Press Inc., New York. 15. Schmidt, G., I. Fromme, and H. Mayer. 1970. Immunochemical studies on core lipoplysaccharides of enterobacteriaceae of different genera. Eur. J. Biochem.

14:357-366.

J. BACTERIOL.

16. Shands, J. W., Jr. 1971. The physical structure of bacterial lipopolysaccharides, p. 127-144. In G. Weinbaum, S. Kadis, and S. J. Ajl (ed.), Microbial toxins, vol. IV. Academic Press Inc., New York. 17. Weidel, W. 1958. Bacterial virus (with particular reference to adsorption/penetration). Annu. Rev. Microbiol. 12:27-48. 18. Wollman, E. 1947. Rklations entre le pouvoir de synthetiser la proline et la resistance au bacteriophage chez des mutants d'Escherichia coli. Ann. Inst. Pasteur Paris 73:348-363.

Blocking of bacteriophages phi W and phi 5 with lipopolysaccharides from Escherichia coli K-12 mutants. - PDF Download Free (2024)
Top Articles
Latest Posts
Recommended Articles
Article information

Author: Tish Haag

Last Updated:

Views: 5279

Rating: 4.7 / 5 (67 voted)

Reviews: 90% of readers found this page helpful

Author information

Name: Tish Haag

Birthday: 1999-11-18

Address: 30256 Tara Expressway, Kutchburgh, VT 92892-0078

Phone: +4215847628708

Job: Internal Consulting Engineer

Hobby: Roller skating, Roller skating, Kayaking, Flying, Graffiti, Ghost hunting, scrapbook

Introduction: My name is Tish Haag, I am a excited, delightful, curious, beautiful, agreeable, enchanting, fancy person who loves writing and wants to share my knowledge and understanding with you.