Morphological Changes Associated with NovobiocinResistance in Bacillus licheniformis
ROBERT L. ROBSON AND JAMES BADDILEY*Microbiological Chemistry Research Laboratory, University ofNewcastle upon Tyne, Newcastle upon Tyne
NE1 7RU, United KingdomReceived for publication 8 September 1976
Spontaneously occurring novobiocin-resistant (Nov) mutants of Bacillus li-cheniformis ATCC 9945, resistant to low levels of novobiocin (15 ug/ml), wereisolated with a frequency of 3 in 106 organisms. Such isolates grew well, butnearly all exhibited consistent pleiotropic alterations in colonial and cell mor-phologies. One mutant, nov-12, grew as chains of unseparated but clearlydistinct daughter cells in the absence of novobiocin in liquid culture. Whennovobiocin was present, nov-12 grew as very long "filaments" which were,however, septate. Septa formned in the presence of the antibiotic were normal,except that no annular cleavage of the septal wall was observed. Septa were alsoirregularly positioned along the filament. These observations were comparedwith previous findings on the effects of novobiocin and novobiocin resistancedescribed for other organisms. It was concluded that the primary action ofnovobiocin might differ in gram-positive and gram-negative organisms. How-ever, when the low-level novobiocin sensitivity, normally associated with gram-positive organisms, was genetically abolished in Nov strains ofB. licheniformis,they became susceptible to an action of novobiocin more analogous to that foundfor gram-negative organisms. The morphological alterations associated with theNov phenotype in this organism, together with observations in other organisms,indicate that novobiocin resistance might be generally useful in the search formutants of gram-positive organisms with altered cell walls.
Novobiocin is a clinically useful antibioticthat is particularly active against gram-posi-tive bacteria, but is also active against certaingram-negative species. It has been shown tocause filamentation in gram-negative rods (5,30), impair membrane integrity (3, 11, 25, 26,31, 32), cause the accumulation of cell wallnucleotide precursors (35, 36), and inhibit nu-cleic acid synthesis (31, 32). In vitro effects ofnovobiocin include inhibition of respiration andoxidative phosphorylation (M. M. Weber andG. Rosso, Bacteriol. Proc., p. 93, 1963), nitratereductase (7), bacterial luminescence (6), aden-osine triphosphatase (1), deoxyribonucleic acid(DNA) and ribonucleic acid polymerases (31),and teichoic acid synthesis (8, 15, 19). Brock hasproposed (6) that the primary action of novobio-cin is the "sequestration" of Mg2+ ions whichare essential for each of these inhibited proc-esses.
Novobiocin-resistant strains are frequentamong clinical isolates, and such mutants wereused in early studies of genetic transformationin Haemophilus influenzae (16). However, lit-tle is known about the mechanisms of resist-ance in such strains. In Staphylococcus aureus,
novobiocin resistance is often pleiotropically as-sociated with other lesions, such as coagulasedeficiency, altered patterns in the utilization ofsugars, and alterations in the sensitivity toInternational Staphylococcal Typing Phage(21-23). Furthermore, these novobiocin-resist-ant strains have cell walls with altered compo-sition. These strains contained an increasedproportion of phosphorus in the wall, either asincreased amounts of the nornal teichoic acidor as an additional teichoic acid not previouslyreported in this organism. It was suggestedthat the excess of teichoic acid in the wall en-hanced the binding of Mg2+ by the organisms,thereby reducing their sensitivity to the actionof novobiocin. It was concluded that the pleio-tropy exhibited by these strains results primar-ily from a genetic change in the control of thebiosynthesis of teichoic acid in this organism(22).At present, selection methods for mutants in
teichoic acid biosynthesis rely upon bacterio-phage resistance, which is necessarily limitedto those organisms whose teichoic acid serves asthe phage receptor site. The possibility thatnovobiocin could be used to select mutants al-
tered in regulation of teichoic acid was investi-gated in Bacillus licheniformis ATCC 9945, inwhich much is known of the biosynthesis ofthese polymers (4, 8, 9, 17).This report describes the isolation ofnovobio-
cin-resistant strains of this organism that in-variably exhibit an altered cell morphology,suggesting that changes in cell wall composi-tion might also be pleiotropically associatedwith resistance to this antibiotic in B. licheni-formis. The observations indicate that the anti-biotic activity of novobiocin against differentspecies might not be explicable by a singlemechanism.
MATERIALS AND METHODSGrowth and maintenance of organisms. B. Ii-
cheniformis ATCC 9945 wild-type and mutantstrains were maintained as freeze-dried ampoules.Organisms were grown aerobically at 370C in eithera complex medium (NB) containing nutrient broth(25 g/liter of distilled water) or a defined minimalmedium (MM) containing glucose (10 g), (NH4)2SO4(1.0 g), K2HPO4 (10.5 g), KH2PO4 (4.5 g), sodiumcitrate (0.47 g), and MgSO4 (0.05 g) in 1 liter ofdistilled water. When required, MM was supple-mented with L-tryptophan at a final concentration of100 ,ug/ml. Solid media contained, in addition, 1.5%(wt/vol) agar and were termed NA and MA, respec-tively. Novobiocin was added to media at a concen-tration of 15 Ag/ml. Spores were obtained by aseptictransfer of exponentially growing organisms fromNB to the resuspension medium of Sterlini andMandelstam (34). Free spores were obtained after 12to 48 h of incubation, washed in distilled water, andkept at 40C. Spores were germinated by heating at800C for 10 min before inoculation into media.
Mutagenesis. Organisms were treated with N-methyl-N'-nitro-N-nitrosoguanidine (NTG) essen-tially by the method of Adelberg et al. (2). WhenNov strains were mutagenized, only spores weretreated with NTG, as vegetative forms grew only aschains or filaments and each colony-forming unitwould therefore be genetically heterogeneous, partic-ularly after mutagenesis. Log-phase NB culture orspore suspensions (10 ml) were centrifuged at 1,500x g for 15 min, and the cells were resuspended insodium citrate buffer (0.1 M, pH 5.5). NTG wasadded to a final concentration of 50 ,ug/ml, andsuspensions were incubated at 370C for 15 min. Mu-tagenized organisms were washed twice in sodiumcitrate buffer at 30C, and vegetative organisms wereoutgrown in NB for 6 h at 370C. Outgrown orga-nisms or spores were inoculated onto NA after ap-propriate dilution.
Electron microscopy. A sufficient volume of cul-ture was centrifuged to yield a thin pellet of orga-nisms (about 3 mm in diameter). These bacteriawere fixed and prestained for 2 h in acetate-Veronalbuffer, pH 6.1, containing 0.5% (wt/vol) uranyl ace-tate by the method ofKellenberger et al. (20). Dehy-dration and embedding of fixed, stained organismswere carried out by the method of Mackey and Mor-
ris (24). Ultrathin sections were cut on an LKBultratome 3, stained with lead citrate (28), and ex-amined in a Metropolitan Vickers EM6 electron mi-croscope.
Phase-contrast microscopy. Bacterial suspen-sions were embedded in 0.25% (wt/vol) agar, andexamination was carried out with a Zeiss microscopefitted with a Planachromat 100/1.25 phase-contrastoil immersion objective. Photographs were taken onIlford Pan F film.Measurement of growth. Bacterial growth was
routinely measured by following the absorbancy at680 nm. Dry weight determinations of bacterial cul-tures were carried out after collection of organismson preweighed membrane filters (4.5 cm, 0.45-,umpore size; Millipore Corp.) that were washed thor-oughly with distilled water before being dried downto a constant weight at 1100C.
Materials. Novobiocin and osmic acid (vials con-taining 2% [wt/vol] aqueous solution) were obtainedfrom BDH Chemicals Ltd., Poole, Dorset, UnitedKingdom. L-Tryptophan and NTG were obtainedfrom Sigma Chemical Co., London, United King-dom. Nutrient broth (no. 2) and agar (speciallypure) were obtained from Oxoid Ltd., London,United Kingdom. All other chemicals were analyti-cal grade reagents.
RESULTSIsolation of mutants. Spontaneously occur-
ring novobiocin-resistant mutants (Nov) of B.licheniformis ATCC 9945, resistant to low lev-els of novobiocin, were isolated by inoculatingabout 107 organisms onto NA containing 15 ggof novobiocin per ml. Plates were incubatedaerobically for 48 h at 370C, and mutant strainswere purified by transfer to fresh novobiocinNA. The frequency of isolation of Nov strainswas about 3 in 106 organisms. Resistant mu-tants were also isolated from a Try- auxotrophwith a similar frequency, and all isolates re-tained the Try- phenotype. B. licheniformisATCC 9945 gives rise to the characteristic li-chen-like colonial morphology on solid media,but Nov strains usually gave rise to small, circu-lar, entire, tenacious colonies with a centraldepressed zone. However, rare isolates, lessthan 0.1% of all Nov mutants, gave rise tonormal wild-type colonies. Revertants of onemutant, nov-12, were obtained after mutagene-sis and identified on the basis of restoration ofwild-type colony morphology. They were alsofound to have recovered wild-type sensitivity tonovobiocin.Growth and morphology of Nov strains. A
total of 20 Nov isolates were cultured in NB con-taining novobiocin (15 ,ug/ml). All strains grewwell in the presence of the antibiotic, but gaverise to "fibrous" cultures. The growth and mor-phology of one such mutant, nov-12, were exam-ined in greater detail. This strain grew with a
doubling time in NB of 32 min in both thepresence and absence of 15 ,ug of novobiocin perml, equaling the doubling time of the wild-typestrain in this medium. Growth in the absence ofnovobiocin gave an apparently normal turbidculture, but microscopic examination revealedthat the mutant grew as long chains of cells(Fig. la). In the presence of novobiocin theorganisms grew as long filaments which be-came entangled, giving rise to macroscopic"ropelike" conglomerates (Fig. lb). When fila-ments were suspended in 20% (wt/vol) sucrosesolution, plasmolyzed cell units were seenwithin the filaments, demonstrating that theywere not aseptate (Fig. lc). Brief treatmentwith ultrasonic irradiation led to incompletelysis of the filament, leaving some cell-like re-gions intact; this suggests that composite cellunits were entirely bounded by an osmoticallyprotective wall. These observations were con-firmed by electron microscopy. In the absence ofnovobiocin, although chains of cells were ob-served (Fig. le), septal walls and membranesappeared to be normal. In the presence of novo-biocin, cell units were often unusually long andthere was no evidence of cleavage of septalwalls, which was consistent with a total lack ofcell separation capacity in such conditions (Fig.lf). nov-12 sporulated well in a sporulationmedium, giving rise to long chains of spores,but liberation of free spores was much delayed.Sporulation was prevented by the presence oflow levels of novobiocin. Spore germination andoutgrowth were parallel in both wild-type andnov-12 strains. Outgrowth of nov-12 spores inthe presence of novobiocin gave rise to thecharacteristic filaments. It was noteworthythat spore coats remained attached to the end ofthe filaments (Fig. ld), which continued todouble in length with a periodicity equal tothe mass doubling time of the organism.
DISCUSSIONNovobiocin-resistant (Nov) mutants of B. Ii-
cheniformis, selected for resistance to low lev-els of novobiocin, invariably exhibited morpho-logical alterations. In the absence of novobio-cin, mutants grew as long chains of cells, but inthe presence of novobiocin they grew as verylong filaments. These filaments were not asep-tate, but consisted of chains of cell units ofvariable length in which little or no annularconstriction of septal walls was observed. Bysubjecting filaments to brief periods of ultra-sonic irradiation designed to cause partial lysisalong the filament, it was shown that suchsepta were still able to provide structural pro-tection against unfavorable osmotic pressureproduced by the lysis of the adjacent cells.
Thus, Nov mutants grow as chains of un-separated, but clearly distinct, daughter cells,while the presence of novobiocin apparentlyintensifies this phenomenon to the point whereno invagination of the cell wall occurs at septalsites, which are in other respects normal butpositioned irregularly along the length of thefilament.The mechanism of action of novobiocin is not
well understood, and indeed it is possible thatno single mechanism can satisfactorily explainits antibiotic activity, since widely differinglevels of novobiocin are required to producebacteriostasis in different species (5). The ob-servations with Nov strains ofB. licheniformistend to support this conclusion.
In general, gram-negative and gram-positiveorganisms differ in their response to novobio-cin. The former are much less sensitive to theantibiotic and, when grown in its presence,yield filaments that have not been observed ingram-positive organisms (5). In Escherichiacoli such filaments are aseptate (5), and it hasbeen concluded that filamentation is due to afailure in cell septation which is a consequenceof inhibition ofDNA synthesis, known to be theprimary action of novobiocin in this organism(32). However, in gram-positive organisms themode of action has not been fully elucidated,although many effects both in vivo and in vitrohave been attributed to novobiocin (6).
If DNA synthesis were the primary site ofaction of the antibiotic in such organisms, fila-mentation ought to occur in the presence ofsub-inhibitory levels of novobiocin, particularly inrod-shaped species. Since filamentation has notpreviously been observed in gram-positive rods,DNA synthesis might not be affected by concen-trations of novobiocin sufflcient to cause inhibi-tion of growth in these organisms.Thus, novobiocin might have at least two
modes of action. One bacteriostatic action, towhich gram-negative organisms might not besusceptible, occurs at low levels of novobiocinand is not associated with inhibition of DNAsynthesis. Thus, Nov strains of B. lichenifor-mis would no longer be sensitive to this effect.A second bacteriostatic activity, probably inhi-bition of DNA synthesis with concomitant fila-ment formation, is possibly the primary site ofaction in gram-negative organisms, e.g., E.coli, and becomes the primary site of action inmutants resistant to low levels of novobiocin,such as those described in this study.
Novobiocin resistance is almost always asso-ciated with chain formation in B. lichenifor-mis. Chain formation is a characteristic of mu-tants deficient in autolytic activity (Lyt-) (10,12, 13, 27, 33, 37). In B. licheniformis NCTC
FIG. 1. Morphology and ultrastructure of nov-12. (a through d) Phase-contrast photomicrographs; (ethrough g) electron micrographs ofultrathin, lead citrate-stained sections. (a) Morphology ofnov-12 grown inthe absence of novobiocin; (b) morphology of nov-12 grown in the presence of15 Mg of novobiocin per ml; (c)plasmolysis offilaments grown in the presence ofnovobiocin (Cu, cell units); (d) outgrowth ofspores ofnov-12in the presence ofnovobiocin (Sc, spore coat); (e) longitudinal section through chains ofnov-12 cells grown inthe absence of novobiocin; (f) longitudinal section through filament of nov-12 grown in presence of 15 pg ofnovobiocin per ml. Bar markers represent 25 ,um in (a) and (b), 5 gm in (c) and (d), and 0.5 gm in (e) and (t).
6346, such mutants arise either from a defi-ciency in the levels of autolytic enzymes per seor from alterations in the cell wall, leading toresistance to autolytic enzymes and, in particu-lar, changes in constituents of the associatedanionic polymers teichoic acid and teichuronicacid (13, 14). Therefore, it seemed likely thatresistance to novobiocin in B. licheniformisATCC 9945, as in S. aureus (22), may be pleio-tropically associated with alterations in anionicpolymers which, in this organism, give rise tothe Lyt- phenotype. Data presented in an ac-companying paper (29) show that this is indeedthe case as cell walls of nov-12 were shown tolack teichuronic acid. Thus, novobiocin could begenerally useful in the search for mutants ofgram-positive organisms with alterations incell wall composition.
ACKNOWLEDGMENTSWe are grateful for the technical assistance of Xinia
Loveday and Phillip Holroyd.This investigation was supported by a grant from the
Science Research Council.
LITERATURE CITED1. Abrams, A., P. McNamara, and F. B. Johnson. 1960.
Adenosine triphosphatase in isolated bacterial cellmembranes. J. Biol. Chem. 235:3659-3662.
2. Adelberg, E., M. Mandel, and G. C. C. Chen. 1965.Optimal conditions for mutagenesis by N-methyl-N'-nitro-N-nitrosoguanidine in E. coli K12. Biochem.Biophys. Res. Commun. 18:788-795.
3. Allwood, M. C., and A. D. Russell. 1969. Thermallyinduced changes in the physical properties ofStaphy-lococcus aureus. J. Appl. Bacteriol. 32:68-78.
4. Anderson, R. G., L. J. Douglas, H. Hussey, and J.Baddiley. 1973. The control of synthesis of bacterialcell walls. Interaction in the synthesis of nucleotideprecursors. Biochem. J. 136:871-876.
5. Brock, T. D. 1956. Studies on the mode of action ofnovobiocin. J. Bacteriol. 72:320-323.
6. Brock, T. D. 1967. Novobiocin, p. 651-655. In D. Gott-lieb and P. D. Shaw (ed.), Antibiotics. I. Mechanismof action. Springer-Verlag, Berlin.
7. Brock, T. D., and M. L. Brock. 1959. Effect of novobio-cin on permeability of Escherichia coli. Arch. Bio-chem. Biophys. 85:176-185.
8. Burger, M. M., and L. Glaser. 1964. The synthesis ofteichoic acids. I. Polyglycerophosphate. J. Biol.Chem. 239:3178-3186.
9. Burger, M. M., and L. Glaser. 1966. The synthesis ofteichoic acids. V. Polyglucosylglycerol phosphate andpolygalactosylglycerol phosphate. J. Biol. Chem.241:494-506.
10. Chaterjee, A. N., D. Mirelman, H. J. Singer, and J. T.Park. 1969. Properties of a novel pleiotropic bacterio-phage-resistant mutant of Staphylococcus aureus H.J. Bacteriol. 100:846-853.
11. Elliot, H. J., and W. H. Hall. 1957. Studies on syner-gism with twelve antibiotics against thirty hospitalstrains of Staphylococcus aureus. J. Lab. Clin. Med.50:242-249.
12. Fan, D. P., and M. M. Beckman. 1971. Mutant ofBacillus subtilis demonstrating the requirement oflysis for growth. J. Bacteriol. 105:629-636.
13. Forsberg, C. W., and H. J. Rogers. 1971. Autolyticenzymes in growth of bacteria. Nature (London)229:272-273.
14. Forsberg, C. W., and H. J. Rogers. 1974. Characteriza-tion of Bacillus licheniformis 6346 mutants whichhave altered lytic enzyme activities. J. Bacteriol.118:358-368.
15. Glaser, L. 1964. The synthesis of teichoic acids. II.Polyribitol phosphate. J. Biol. Chem. 239:3178-3186.
16. Goodgal, S. H., and R. M. Herriot. 1957. A study oflinked transformations in Haemophilus influenzae.Genetics 42:372.
17. Hancock, I. C., and J. Baddiley. 1972. Biosynthesis ofthe wall teichoic acid in Bacillus licheniformis. Bio-chem. J. 127:27-37.
18. Hughes, R. C. 1966. The cell wall ofBacillus lichenifor-mis NCTC 6346. Biosynthesis of the teichuronic acid.Biochem. J. 101:692-697.
19. Ishimoto, N., and J. L. Strominger. 1966. Polyribitolphosphate synthetase of Staphylococcus aureus. J.Biol. Chem. 241:638-650.
20. Kellenberger, E., A. Ryter, and J. Sechaud. 1958. Elec-tronmicroscopic study of DNA-containing plasms.Vegetative and mature phage DNA as compared withnormal bacterial nucleoids in different physiologicalstates. J. Biophys. Biochem. Cytol. 4:671-678.
21. Korman, R. Z. 1963. Coagulase-negative mutants ofStaphylococcus aureus: genetic studies. J. Bacteriol.86:363-369.
22. Korman, R. Z. 1975. Cell wall composition of novobio-cin-resistant pleiotropic mutant staphylococci. J.Bacteriol. 124:724-730.
23. Korman, R. Z. 1975. Bacteriophage-typable revertantsfrom pleiotropic staphylococcal mutants. J. Bacteriol.124:731-735.
24. Mackey, B. M., and J. G. Morris. 1971. Ultrastructuralchanges during sporulation of Clostridium pasteu-rianum. J. Gen. Microbiol. 66:1-13.
25. Morris, A., and A. D. Russell. 1968. Studies on themode of action of novobiocin. Biochem. Pharmacol.17:1923-1929.
26. Morris, A., and A. D. Russell. 1970. Novobiocin inducedleakage of intracellular substances from Escherichiacoli. Microbios 2:35-41.
27. Pooley, H. M., G. D. Shockman, M. L. Higgins, and J.Porres-Juan. 1972. Some properties of two autolytic-defective mutants of Streptococcus faecalis ATCC9790. J. Bacteriol. 109:423-431.
28. Reynolds, E. S. 1963. The use of lead citrate at high pHas an electron opaque stain in electron microscopy. J.Cell Biol. 17:208-212.
29. Robson, R. L., and J. Baddiley. 1976. Role of teichu-ronic acid in Bacillus licheniformis: defective autoly-sis due to deficiency of teichuronic acid in a novobio-cin-resistant mutant. J. Bacteriol. 129:1051-1058.
30. Smith, C. G., A. Dietz, W. T. Sokolski, and G. M.Savage. 1956. Streptonivicin, a new antibiotic. Anti-biot. Chemother. 6:135-142.
31. Smith, D. H., and B. D. Davis. 1965. Inhibition ofnucleic acid synthesis by novobiocin. Biochem. Bio-phys. Res. Commun. 18:796-800.
32. Smith, D. H., and B. D. Davis. 1967. Mode of action ofnovobiocin in Escherichia coli. J. Bacteriol. 93:71-79.
33. Soper, J. W., and C. G. Winter. 1973. Role of the cellwall autolysin in chain formation by a mutant ofStreptococcus faecalis. Biochim. Biophys. Acta297:333-342.
34. Sterlini, J. M., and J. Mandelstam. 1969. Commitmentto sporulation in Bacillus subtilis and its relationshipto development of actinomycin resistance. Biochem.J. 113:29-37.
35. Strominger, J. L., and R. H. Threnn. 1959. The optical