Mutation Research, 264 (1991) 119-125 119

1991 Elsevier Science Publishers B.V. All rights reserved. 0165-7992/91/$03.50

ADONIS 0165799291000923

MUTLET 0541

Decreased resistance to antibiotics and plasmid loss in plasmid-carrying

strains of Staphylococcus aureus treated with ascorbic acid

Carlos F. Amfibile-Cuevas 1, Rosa Maria Pifia-Zentella 2 and

Maria Elena Wah-Laborde a

1Departamento de Microbiologla, LUSARA, 08930 Mexico, D.F. and 2Departamento de Ciencias de la Nutrici6n y de los Alimentos,
Universidad lberoamericana, 01210 Mexico, D.F. (Mexico)

(Accepted 5 July 1991)

Keywords." Ascorbic acid; Plasmid loss; Antibiotic resistance

Summary

The effect of ascorbic acid on plasmid-coded antibiotic resistance in Staphylococcus aureus was in-
vestigated. Several strains of S. aureus were cultured in the presence of 1 mM ascorbate for 6 h. This treat-
ment induced an increased loss of resistance markers in 4 o f 6 strains tested, and agarose gel electrophoresis
showed this disappearance of plasmid DNA in ascorbate-induced susceptible colonies. The presence of
ascorbate induced a 50-75% decrease in minimal inhibitory concentrations of different antibiotics for resis-
tant strains. When ascorbate is added, formerly subinhibitory concentrations of penicillin or tetracycline
have an increased inhibitory effect on resistant strains and even induced the death of 25-93% of the intial
population. These results suggest that ascorbate can induce the loss of several plasmids of S. aureus, and that
the levels of antibiotic resistance are also affected by the presence of this compound.

L-Ascorbic acid (vitamin C) has been found to be
capable of modifying the properties of DNA
(McCarty, 1945). Ascorbate can inactivate several
phages by induction of single-strand breaks in their
D N A (Murata and Kitagawa, 1973; Murata et al.,
1971, 1986), and can induce multiple breaks in the
chromosomes of repair-deficient mutants of
Escherichia coli (Van Sluys et al., 1986a). This ef-
Correspondence: Dr. C.F. Am~ibile-Cuevas, (present address)
Laboratory of Toxicology, Harvard School of Public Health,
665 Huntington Ave., Boston, MA 02115 (U.S.A.), Tel: (617)
432-3490; Fax: (617) 432-1780.
fect was initially attributed to the generation of
hydrogen peroxide (H202) a n d / o r hydroxyl free
radicals (HO - ), but this does not seem to be clear.
E. coli x t h A , an H202-hypersensitive mutant
(Demple et al., 1983), is not sensitive to ascorbate
treatment (Van Sluys et al., 1986b). A sequence-
specific breakage following treatment with ascor-
bate, unlikely to be randomly generated by oxygen
radicals, has been observed in phage DNA
(Kobayashi et al., 1988).
Most in vivo tests fail to detect mutagenic or car-
cinogenic activity for ascorbic acid (Douglas et al.,
1984; Norkus et al., 1983). It seems that ascorbate
120
TABLE 1
S. aureus STRAINS
Strain Plasmid
DIM 152 N.C.
DIM162 N.C.
DIM260 N.C.
RN11 p1258
RN794 pI55cl
RN2424 pTl81
Amk, amikacin; Era, erythromycin; Pc, penicillin; Sin, streptomycin; Spc, spectinomycin; Tc, tetracycline; Tin, tobramycin. N.C., not
characterized.
* From Universidad A u t 6 n o m a de Puebla (Mexico). DIM strains were clinically isolated, and aminoglycoside resistance was found to
be plasmid-mediated.
can damage only small and unprotected D N A
molecules, especially circular covalently closed
forms (cccDNA) which depend for their efficient
replication and transcription on the maintenance
of their supercoiled state.
This laboratory has previously reported the
elimination of penicillinase plasmids of S. aureus
after a 6-h treatment with 1 mM ascorbate
(Amfibile-Cuevas, 1988). The possibility that the
elimination of bacterial resistance determinants
could be of some help in controlling the incidence
of extrachromosomal drug resistance, as has been
suggested (Bouanchaud and Chabbert, 1971;
Hahn, 1976), has induced us to examine the range
of this loss of resistance caused by ascorbic acid.
We have tested the effects of this compound upon
several plasmid-bearing resistant strains of S.
aureus, including clinical isolates and laboratory-
constructed strains, resistance to different an-
tibiotics, and in the case of penicillinase plasmids,
different 'families' of extrachromosomal elements
(Shalita et al., 1980). We think that an agent
capable of eliminating resistance plasmids could be
useful in the therapy of infectious diseases with
multi-resistant etiology, if administered together
with a c o m m o n antibiotic. In order to determine if
an antibiotic-resistant organism can be inhibited by
the presence of ascorbic acid and an antimicrobial
drug, we have exposed these resistant strains to the
joint action of ascorbate and antibiotics. Although
ascorbate cannot induce plasmid loss in all of the
Relevant phenotype Reference
(Sm,Tm,Spc,Amk) r B. Baca, unpublished*
(Tm,Spc,Amk) r B. Baca, unpublished
(Tm,Spc,Amk) r B. Baca, unpublished
Pcr,Em r Peyru et al., 1969
Pc r Peyru et al., 1969
Tc r lordanescu and Surdeanu, 1980
strains tested, we found an increase in the in-
hibitory action of antibiotics on resistant strains
when ascorbate is present.
Material and methods
Strains and chemicals
The strains of S. aureus used are listed in Table
1. Liquid cultures were grown in Brain Heart Infu-
sion broth (BHI) and solid ones on Muller-Hinton
agar (both from Difco). L-Ascorbic acid (Sigma)
was used as the sodium salt, adjusting 100 mM
stock solutions to a p H of 7.2 by addition of
NaHCO3. Amikacin, penicillin, spectinomycin,
tetracycline, and tobramycin were obtained from
Laboratorios Bristol de M6xico, Farmac6uticos
Lakeside, Upjohn, Rorer de M6xico, and Upjohn,
respectively.
Determination o f plasmid loss
Ascorbate treatment was performed as follows:
small inocula (106 cells/ml) of each strain were
grown in BHI with or without 1 mM ascorbate, for
6 h at 37C with constant air bubbling. Samples of
these cultures were streaked on agar plates to ob-
tain isolated colonies. Plasmids were scored by the
presence or absence of their resistance markers as
determined by toothpicking to selective and non-
selective media. Selective concentrations were: 50
t~g/ml of amikacin, spectinomycin, and tobramy-
cin, used separately for DIM strains; 50 /~g/ml
 121

TABLE 2

LOSS OF RESIST ANCE MARKERS BY 6-h, l-mM A S C O R B A T E T R E A T M E N T

Strain Treatment Total colonies Sensitive colonies" % loss

DIMI52 None 160 11 6.9

Ascorbate 180 28 15.6
DIM162 None 120 0 < 1.0
Ascorbate 120 12 10.0
DIM260 None 250 2 < 1.0
Ascorbate 250 35 14.0
RNI1 None 180 0 < 1.0
Ascorbate 180 0 < 1.0
RN794 None 300 105 35.0
Ascorbate 300 144 48.0
RN2424 None 120 0 < 1.0
Ascorbate 120 0 < 1.0

" The values given for DIM strains are the number of colonies that lost the resistance to the 3 antibiotics tested. There was no segregation
of resistance.

tetracycline, for strain RN2424; 0.1 #g/ml penicil- Cultures were incubated at 37C with constant air
lin for strains RN11 and RN794. Plasmid loss was bubbling. Colony-forming units (cfu/ml) were
confirmed by agarose electrophoretic screening of determined by diluting cells appropriately, plating
extrachromosomal DNA, performed by the
method of Nahaie et al. (1984) using 0.8% agarose
gels on a horizontal slab. DIM RN
Minimal inhibitory concentrations 152 162 260 7 9 4
Determinations of MICs in the presence or RS RS R S RS
absence of ascorbic acid were done by the tube-
dilution technique, in BHI with or without 1 mM
ascorbate. Preliminary determinations were done
....i 28.2
by doubling dilutions of the antibiotics, and then
ranges of activity were divided with 10-#g in-
10.1
creases. The concentrations tested were: 40-100
# g / m l tetracycline and 40-250 ~g/ml penicillin.
Tubes were incubated at 37C for 12 h, and growth 4.4
inhibition was confirmed 'by turbidimetric
measurement at 620 nm in a Corning 252 col-
orimeter. kb
"ccDNA)
Cultures on antibiotics and~or ascorbate
The effect of simultaneous treatment with ascor-
bate and subinhibitory concentrations of an-
tibiotics was determined by growing the cells in
Fig. 1. Electrophoretic screening of plasmid DNA from resis-
BHI containing 1 m M ascorbate and 50 #g/ml tant (R) and ascorbate-induced susceptible (S) colonies of
tetracycline for strain RN2424, 40 #g/ml penicillin strains DIM152, DIM162, DIM260, and RNl I (RN794). Migra-
for strain RN794, and 15 /~/ml for strain R N l l . tion of cccDNA markers in indicated on the right.
122

TABLE 3
MICs IN THE PRESENCE OR ABSENCE OF ASCORBATE
Strain Antibiotic MIC (#g/ml) M1C (#g/ml)
/
RN 11
RN794
RN2424
Penicillin
Penicillin
Tetracycline
without
ascorbate
250
200
100
with
ascorbate
120
50
50
e /"
il
t h e m , a n d i n c u b a t i n g at 37C for 18 h, a n d then
c o u n t i n g colonies.

Results
//
L o s s o f resistance p l a s m i d s due to ascorbate
A s c o r b a t e increased the o c c u r r e n c e o f resistance
loss in a m i n o g l y c o s i d e - r e s i s t a n t strains, a n d in one
o f the penicillin-resistant strains, pi258 (pen r) a n d

0 2 4 6 8
Time (h)
Fig. 3. Growth curves of S. aureus RN794. Symbols as in Fig. 2.
kd

p T 181 (tet r) c a n n o t be e l i m i n a t e d by t r e a t m e n t with

7 1 mM ascorbate. A comparison of spontaneous
a n d a s c o r b a t e - i n d u c e d loss o f resistance m a r k e r s is
s h o w n in T a b l e 2. T h e sum o f results o f 3
s i m u l t a n e o u s e x p e r i m e n t s is shown; these values
are typical for 4 sets o f i n d e p e n d e n t experiments.
T h e d i s a p p e a r a n c e o f p l a s m i d D N A in the elec-
t r o p h o r e t i c screening o f crude lysates o f a s c o r b a t e -
i n d u c e d d r u g susceptible colonies c o n f i r m s that the
loss o f resistance m a r k e r s is the result o f the
e l i m i n a t i o n o f resistance p l a s m i d s (Fig. 1). 10-20

~U~o/O/ susceptible colonies f r o m each e x p e r i m e n t a n d

each strain were a n a l y z e d a n d all d i s p l a y the loss o f
5 plasmid bands.
J ,

o 2 4 i
Time (h)
Fig. 2. Growth curves of S. aureus RNll, in BHI (e), with
ascorbate (), with penicillin (u), or with ascorbate and
penicillin (D). Symbols represent the mean values of at least 3
independent experiments. Deviations are within the symbols.
h
R e d u c t i o n in M I C s by the p r e s e n c e o f ascorbate
T h e a d d i t i o n o f 1 m M a s c o r b a t e reduces the
M I C s o f penicillin by 5270 for strain RN11, a n d by
75% for RN794. T e t r a c y c l i n e M I C was r e d u c e d by
50% for RN2424. Values in T a b l e 3 are f r o m one

typical experiment, but have been reproduced in at
least 3 independent experiments.
Growth inhibition and cell death by antibiotics if
ascorbate is present
1 mM ascorbate did not itself modify the growth
of strains R N l l , RN794, and RN2424. But the
association of ascorbate and subinhibitory concen-
trations of antibiotics can inhibit the growth of
bacterial populations, and even induce the death of
25, 93, and 89o70 of the initial inocula of R N l l ,
RN794, and RN2424 respectively, whereas the an-
tibiotics themselves just induce a time lag
(Figs. 2-4).
Discussion
L-Ascorbic acid is involved in a wide range of
biological activities, especially against nucleic acids
O
9~
E pression of resistant determinants by interference
u_
o ocDNA, previously reported as an effect of ascor-
4,
3'
\o/
0 2 4 6 8 appear to correlate with the effect of the associa-
Time (h)
Fig. 4. Growth curves of S. a u r e u s RN2424, in BHI (o), with
ascorbate (O), with tetracycline ( l ) , or with ascorbate and
tetracycline ( [] ).
123
(Shamberger, 1984). The present results suggest
that a variety of S. aureus plasmids are sensitive to
the 'curing' effect o f ascorbate, which was
previously reported by this laboratory (Amfibile-
Cuevas, 1988). 4 of the 6 strains tested showed a
loss of the resistant phenotype when treated with
ascorbate that can be attributed to the elimination
of the coding plasmid, as can be proved by elec-
trophoretic screening. In the case of RN2424, 1
mM ascorbate did not induce the loss of resistance
from any cell. This could be the consequence of the
high copy number of plasmid pT181 (20-25), so we
should not discard the possibility that ascorbate
could induce the loss of some of the copies, but not
of all.
The bactericidal effect of the association of
ascorbate and antibiotic on resistant strains, which
can be observed in the determinations of MICs as
well as in the growth curves, could be the conse-
quence of a variety of possible events: (a) the
reduction in the number of plasmids within the
bacterial population; (b) the inhibition of the ex-
with plasmid transcription or by damage to
plasmid DNA (e.g., the conversion of cccDNA to
bate on circular DNA (Morgan et al., 1976)); or (c)
a synergistic effect of ascorbate, which merely
potentiates the antimicrobial activity of the drugs.
We think the last possibility is unlikely, considering
that the 2 antibiotics involved, penicillin and
tetracycline, have very different cellular targets and
mechanisms of action. Furthermore, the effect of
ascorbate differed in magnitude in different RN
strains that differ only in their resistance plasmids.
Therefore, it appears that ascorbate is acting
through the plasmid and not on the host cell.
Moreover, MICs for the host without plasmids did
not change because of the presence of ascorbate
(not shown).
However, the loss of plasmids (Table 2) does not
tion of ascorbate and antibiotics, especially for
strain RN2424. Ascorbate did not eliminate the
plasmids from any cell, but did reduce the
tetracycline MIC to 507o and can induce the death
124
of 89% of an original population exposed to a
subinhibitory concentration of the antibiotic. This
could be the result of a reduction in the plasmid
copy number because there is a gene-dosage
dependence of resistance for this plasmid (Carleton
et al., 1984; Thomas, 1986). Although there is no
evidence that penicillin resistance has the same
behavior, this or similar mechanisms can be pro-
posed, together with an interference with the ex-
pression of plasmid genes, as mentioned earlier.
None of these results give any indication about
the mechanism of action of ascorbate on plasmids.
But it seems unlikely that this could be the result of
the extracellular generation of free radicals that
enter the cell and damage DNA or other structures
involved in plasmid replication or stability. Free
radicals, if any, should be generated very close to
DNA, perhaps even from a DNA-ascorbate com-
plex, as has been reported previously (Jamaluddin
et al., 1981). DNA damage induced by ascorbate
seems to have 2 different intermediates: (a)
hydrogen peroxide, produced during ascorbate
autoxidation, which generates hydroxyl radicals,
the most reactive radical species involved in the
damage by ascorbate of phage D N A in vitro
(Morgan et al., 1976; Murata et al., 1986) and E.
coli chromosome in vivo (Van Sluys et al., 1986a);
and (b) other products of ascorbate oxidation,
probably peroxy or alkoxyl radicals of ascorbate,
which may be involved in the sequence-specific
damage of phage DNA, different from the uniform
cleavage caused by H O - (Kobayashi et al., 1988).
Extracellular catalase or thiourea prevents the kill-
ing of u v r A r e c A mutants of E. coli by ascorbate,
suggesting the participation of H202 and HO"
(Van Sluys et al., 1986a). But the concomitant
DNA single-strand breakage induced by treatment
with ascorbate, which can be, at least in part,
responsible for cell death, may not be caused by
H202 or its products, since the x t h A mutant, which
is hypersensitive to hydrogen peroxide, is not sen-
sitive to ascorbate/copper treatment (Van Sluys et
al., 1986b). ngr-374 mutants of S. aureus (Tam and
Pattee, 1986), which also are hypersensitive to
H202, are not sensitive to ascorbate (Am~ibile-
Cuevas, unpublished results). Moreover, H202
itself did not induce plasmid loss in any of the S.
aureus strains used in this work (Amfibile-Cuevas,
unpublished results). These results suggest that
H202 is not involved as a direct intermediate in
DNA damage or in plasmid loss by ascorbic acid in
S. aureus.
There is also the possibility that genotoxic prod-
ucts of lipid peroxidation of membrane com-
ponents can act as reactive intermediates of the
plasmid elimination by ascorbate treatment, as has
been proposed for radiation mutagenesis (Petkau,
1980). Ascorbate can promote extensive lipid
peroxidation in some biological preparations, such
as in vitro assays of membrane enzymes (Schaefer
et al., 1975), or binding assays of neurotransmit-
ters to membrane receptors (Heikkila, 1984). In-
teraction of ascorbate with membranes may also
result in the distortion of membrane architecture,
either as a consequence of lipid peroxidation, or by
other mechanisms that do not involve free radical
generation or peroxidation (Matsuda et al., 1990).
Other membranes-acting compounds, such as SDS
(Sonstein and Baldwin, 1972), and phenothiazines
and imipramine (Molnfir et al., 1982) are also
capable of eliminating bacterial plasmids. How-
ever, plasmids pi258 and pT181, which can be
eliminated by protoplast formation-regeneration
(Novick et aI., 1980), are not susceptible to ascot-
bate. Plasmid instability during protoplast regener-
ation has been attributed to the involvement of the
cell envelope in plasmid maintenance (Novick et
al., 1980; Gruss and Novick, 1986). Therefore, we
would initially expect these plasmids to be
eliminated by ascorbate treatment, if ascorbic acid
alters the cell membrane; this is not the case. Fur-
ther work is needed to determine the mechanism by
which ascorbic acid acts.
Acknowledgements
We would like to express our gratitude to Bruce
Demple and Mauricio Ludgar for their criticism
and valuable comments, R.P. Novick, P.A. Pat-
tee, and B. Baca for their kind sending of strains,
and to Isabel Niv6n for her excellent technical
assistance.

References
Amabile-Cuevas, C.F. (1988) Loss of penicillinase plasmids of
Staphylococcus aureus after treatment with L-ascorbic acid,
Mutation Res., 207, 107-109.
Bouanchaud, D.H., and Y.A. Chabbert (1971) Practical effec-
tiveness of agents curing R factors and plasmids, Ann. NY
Acad. Sci., 182, 305-311.
Carleton, S., S.J. Projan, S.H. Highlander, S. Moghazeh and
R.P. Novick (1984) Control of pT181 replication, I1. Muta-
tional analysis, EMBO J., 3, 2407-2414.
Demple, B., J. Halbrook and S. Linn (1983) Escherichia coli
xthA mutants are hypersensitive to hydrogen peroxide, J.
Bacteriol., 153, 1079-1082.
Douglas, J.F., J. Huff and A.C. Peters (1984) No evidence of
carcinogenicity for L-ascorbic acid (vitamin C) in rodents, J.
Toxicol. Environ. Health, 14, 605-609.
Gruss, A., and R. Novick (1986) Plasmid instability in
regenerating protoplasts of Staphylococcus aureus is caused
by aberrant cell division, J. Bacteriol., 165, 878-883.
Hahn, F.E. (1976) Experimental elimination of R factors, An-
tibiotics Chemother., 20, 196-226.
Heikkila, R.E. (1984) Ascorbic acid, lipid peroxidation and the
interaction of drugs with their receptors in rat brain tissue
preparations, in: W. Bors, M. Saran and D. Tait (Eds.), Ox-
ygen Radicals in Chemistry and Biology, De Gruyter, Berlin,
pp. 371-375.
lordanescu, S., and M. Surdeanu (1980) Complementation of a
replication defect by autonomous incompatible plasmids in
Staphylococcus aureus, Plasmid, 4, 1-7.
Jamaluddin, M., S. Rajalakshmi and D.S.R. Sarma (1981)
Evidence for a direct interaction of L-ascorbate with rat liver
DNA, Biochim. Biophys. Acta, 653, 423-427.
Kobayashi, S., K. Ueda, J. Morita, H. Sakai and T. Komano
(1988) DNA damage induced by ascorbate in the presence of
Cu z+ , Biochim. Biophys. Acta, 949, 143-147.
Matsuda, T., T. Gemba, A. Baba and H. lwata (1990) Decrease
of Na+-Ca z + exchange activity by ascorbate in rat brain
membrane vesicles, Brain Res., 532, 13-18.
McCarty, M. (1945) Reversible inactivation of the substance in-
ducing transformation of pneumococcal types, J. Exp. M ed.,
81, 507-514.
Molndr, J., Y. Mandi and S. F61dedk (1982) Drug-receptor in-
teraction on plasmid elimination by phenothiazines and imip-
ramine in Escherichia coil, Acta Microbiol. Acad. Sci.
Hungar., 29, 2-25.
Morgan, A.R., R.L. Cone and T.M. Elgert (1976) The
mechanism of DNA strand breakage by vitamin C and
superoxide and the protective roles of catalase and superoxide
dismutase, Nucleic Acids Res., 3, 1139-1149.
Murata, A., and K. Kitagawa (1973) Mechanism of inactivation
of bacteriophage Jl by ascorbic acid, Agr. Biol. Chem., 37,
1145-1151.
125
Murata, A., K. Kitagawa and R. Saruno (1971) Inactivation of
bacteriophages by ascorbic acid, Agr. Biol. Chem., 35,
294-296.
Murata, A., H. Suenaga, S. Hideshima, Y. Tanaka and F. Kato
(1986) Hydroxyl radical as the reactive species in the inactiva-
tion of phages by ascorbic acid, Agr. Biol. Chem., 50,
1481-1487.
Nahaie, M.R., M. Goodfellow and C.R. Harwood (1984) A
rapid screening procedure for staphylococcal plasmids, J.
Microb. Meth., 2, 73-81.
Norkus, E.P., W. Kuenzig and A.H. Conney (1983) Studies on
the mutagenic activity of ascorbic acid in vitro and in vivo,
Mutation Res., 117, 183-191.
Novick, R., C. Sanchez-Rivas, A. Gruss and 1. Edelman (1980)
Involvement of the cell envelope in plasmid maintenance:
plasmid curing during regeneration of protoplasts, Plasmid,
3, 348-358.
Petkau, A. (1980) Radiation carcinogenesis from a membrane
perspective, Acta Physiol. Scand., Suppl. 492, 81-90.
Peyru, G., L.F. Wexler and R.P. Novick (1969) Naturally-
occurring penicillinase plasmids in Staphylococcus aureus, J.
Bacteriol., 98,215-221.
Schaefer, A., M. Koml6s and A. Seregi (1975) Lipid peroxida-
tion as the cause of the ascorbic acid induced decrease of
adenosine triphosphatase activities of rat brain microsomes
and its inhibition by biogenic amines and psychotropic drugs,
Biochem. Pharmacol., 24, 1781-1786.
Shalita, Z., E. Murphy and R.P. Novick (1980) Penicillinase
plasmids of Staphylococcus aureus: structural and evolu-
tionary relationships, Plasmid, 3,291-311.
Shamberger, R.J. (1984) Genetic toxicology of ascorbic acid,
Mutation Res., 133, 708-714.
Sonstein, S.A., and J.N. Baldwin (1972) Loss of the
penicillinase plasmid after treatment of Staphylococcus
aureus with sodium dodecyl sulfate, J. Bacteriol., 109,
262-265.
Tam, J.E., and P.A. Pattee (1986) Characterization and genetic
mapping of a mutation affecting apurinic endonuclease ac-
tivity in Staphylococcus aureus, J. Bacteriol., 168, 708-714.
Thomas, C.M. (1986) Plasmid replication, in: K.G. Hardy
(Ed.), Plasmids, A Practical Approach, IRL, Oxford, pp.
7-36.
Van Sluys, M.A., R. Costa de O1iveira, 1. Felzenszwalb, C.F.
Martins Menck and R. Alcantara-Gomes (1986a) Ascorbate-
copper induced DNA lesions and repair in Escherichia coli
K12 cells, Carcinogenesis, 7, 197-200.
Van Sluys, M.A., R. Alcantara-Gomes and C.F. Martins
Menck (1986b) Escherichia coli xthA mutant is not hypersen-
sitive to ascorbic acid/copper treatment - an HzOz generating
reaction, Mutation Res., 174, 265-269.
Communicated by D.M. Shankel