OF S T A P H Y L O C O C C U S A U R E U S

                                            AND I. CHOPRA
                                  R. W. LACEY
                         Department of Bacteriology, The Medical School,
                              University of Bristol, Bristol BS8 ITD

D U R I N G the last 30 years, strains of Staphylococcus aureus isolated from
hospital sources have varied in three major respects: (1) in their resistance to
antibiotics, due chiefly to acquisition and loss of plasmids (Dyke, Parker and
Richmond, 1970; Lacey, Lewis and Grinsted, 1973), (2) in their phage-typing
patterns, as a result of the gain or loss of prophages (Asheshov and Winkler,
1966; Jevons, John and Parker, 1966; Jessen et al., 1969), and probably
(3) in their virulence, which appears to have declined (Williams, 1971). Because
transduction is the only mechanism likely to have effected plasmid transfer in
nature, gain of plasmids has probably followed from the emergence of trans-
ducing phages within the staphylococcal population. But the relationship of
virulence to alterations in either prophage or plasmid carriage is not clear.
Loss of virulence might be expected to have resulted from the acquisition of
certain prophages, because these inhibit the production of extracellular lipase
(Jevons et al., 1966; Jessen et al., 1969), an enzyme that probably facilitates
the invasion of normal skin by staphylococci (Alder, Gillespie and Herdan,
1953; Jessen and Bulow, 1967). Although primary skin lesions are now a rare
consequence of infection by hospital staphylococci, except perhaps in very
debilitated patients, a considerable number of isolates still produce this lipase
(e.g., Bulow, 1971), so that other factors may be responsible for the decline in
    In a few instances, particular plasmids determine the production of specific
" virulence factors " such as enterotoxins in Escherichia coli (e.g., Smith and

Linggood, 197 l), but little is known about the general effects of plasmid carriage
on bacterial virulence. Thus, although isolates of E. coli that contain R factors
survive less well in the gut than do R-factor-free bacilli (Anderson, 1974),
these findings cannot necessarily be extrapolated to predict the virulence of the
    In this paper we describe investigations of the growth kinetics and virulence
for the chick embryo of plasmid-positive and plasmid-negative variants of
staphylococcal strains.
                                        MATERIALS AND METHODS
   Cultures of staphylococci. Strains no. 649 and 6936 have been described previously
(Lacey and Chopra, 1974). Strain 649MR is a derivative of the wild strain 649 that contains

                               Received 2 May 1974; accepted 17 July 1974.
   J. MED. MICROBIOL.-VOL.   8 (1975)              137
 138                             R. W. LACEY AND I. CHOPRA

five plasmids as covalently closed circular (CCC) DNA and also three probable plasmids
not present as CCC DNA (Lacey and Chopra, 1974). Strain 649N is a derivative of strain
649 that has the determinant for pigmentation but no other plasmid and no CCC DNA.
A series of derivatives of no. 649N, each possessing a single additional plasmid, have been
described (Lacey and Chopra, 1974); among these is no. 649 tet-r(P), which harbours a
plasmid coding for tetracycline resistance; no. 649 tet-r(C),on the other hand, has chromo-
somal genes for tetracycline resistance (Chopra, Lacey and Connolly, 1974). Strain
649MR-tN was obtained from no. 649MR by subculture in vitro without exposure to
mutagens; apart from retention of pigmentation, it does not possess any markers attribu-
table to a plasmid locus and any CCC DNA (Chopra et al., 1974). The phage-typing patterns
of strain 649 and its derivatives are shown in table I. The plasmids employed have been
described previously (Lacey and Chopra, 1974).
    No. 13136-wild is a methicillin-resistant strain of S. aureus (Jevons, 1960) and no. 13136
met-s is a spontaneously occurring methicillin-sensitivevariant of it (Lacey, 19726).
    General methods. The media used, and the methods of detecting loss of antibiotic
resistance, recording of pigment, bateriophage typing, and transduction were described
previously (Lacey and Chopra, 1974). Survival on glass was studied by the method of
Lacey (1972a).
    Growth kinetics. To minimise environmental variables, pairs of cultures were grown in
nutrient broth and the growth of the test culture compared with that of a standard strain
(usually 649N). Pairs of overnight cultures that had been incubated statically were mixed
and then diluted 1 in 100 into fresh pre-warmed nutrient broth. These cultures were incu-
bated statically at 37°C and the proportions of each derivative were estimated at 0, 2,4, 6, 8
and 24 h by performing viable counts on nutrient agar and nutrient agar containing the
appropriate antibiotic.
    Some experiments were performed with individual cultures of 649MR and 649N to
exclude the possibility of interactions between the cultures.
   Assay of difiuible toxins and enzymes. Quantitative assay of coagulase, fibrinolysin,

                                            TABLE   I
                  Phage-typing patterns of strain 649-wild and derivatives of it

                       Designation of strain        Phage-typing pattern at
                           or derivative             routine test dilution

                   649-wild (pig+, str-r, mir-r)   6/42E/47/53/54/75/81(88)
                   649N (pig')                     6147153154175( 88)
                   649tet-r(P)                     6/47/53/54/75/8 l(88)
                   649tet-r(C)                     6/47/53/54/75/8 l(88)
                   649neo-r                        6/47/53/75(88)
                   649chm-r                        6/47/53/54/75(88)
                   649pen +                        6/428/47/53/54/75/8 l(88)
                   649mir-r                        6/47/53/54/75/81(88)
                   649str-r                        6/47/53/54/75/81(88)
                   649pig -                        6/47/53/54/75/85/8l(88)
                   649MR                           6/47/53/54(88)
                   649MR+N                         6/47/53/54( 88)

    pig+ = Pigmented;pig- = unpigmented; fet-r(P) = plasmid-mediated resistance to tetracycline;
tet-r(C) = chromsomal resistance to tetracycline; neo-r = resistant to neomycin; chm = to chlor-
amphenicol; pen+ = producer of penicillinase and resistant to erythromycin; mir-r = resistant to
cadmium, arsenate and mercuric ions; sfr-r to streptomycin;MR = the possession of eight plasmids
[respectively coding for pigmentation, metal-ion resistance, streptomycin resistance, tetracycline
resistance, penicillinaseproduction, chloramphenicolresistance, neomycin resistance, and methicillin
resistance (see Lacey and Chopra, 1974)l; MR-tN = a derivative of strain 649MR that has lost all
these plasmids except that coding for pigment; (88) = experimental phage.
            PLASMID CARRIAGE A N D STAPHYLOCOCCAL VIRULENCE                                     139
a-,,6- and &haemolysins, lipase and phosphatase was by the method of Hallander (1963).
DNAase was measured semi-quantitatively by a plate method.
    Virulence for chick embryos. Intra-allantoic injection (McCabe, 1964) was performed,
and the embryos were examined for viability daily for 4 days. Death of the embryo was
indicated by absence of spontaneous movement and was usually associated with haemorrhage,
loss of vascular architecture, production of gas and the disappearance of normal morphology.
    Preparation of cell membranes and analysis of membrane proteins by polyacrylamide-gel
electrophoresis were as described previously (Chopra et al., 1974).

   Efect o single plasmids on growth kinetics o strain 649 in nutrient broth
           f                                     f
   Derivatives of strain 649N that harboured a variety of plasmids singly were
grown in mixed culture with strain 649N to determine their exponential doubling
times relative to that of no. 649N. Strains 649pen+ and 649mir-r, which harbour
plasmids of M.W. respectively 2 0 106 and 35 x 106 daltons, each had a
doubling time c. 20 % longer than that of no. 649N; nos. 649neo-r and 649chm-r
harbour plasmids of M.W. respectively 5.9 x 106 and 2.9 x 106 daltons, and these
had doublingtimes c. 10 % longer than that of no. 649N (table II). These findings
suggest that the property of the plasmid that affects growth rate is its size,
Further evidence for this was obtained by showing that derivatives of no.
649N in which the tetracycline-resistance genes were on a small plasmid
(M.W. 2.9 x 106 daltons) or were chromosomeal had similar exponential
doubling times to that of no. 649N (table 11).
   The effect of methicillin resistance on growth rate was studied with strain

         E f e c t of the presence of genetic determinants on the exponential growth rate
                           of Staphylococcus aureus strains in mixed culture

                   Plasmid or                                   Doubling time* of
                  other genetic         Composition of           cells carrying the
                  determinant            mixed culture          stated determinant
                                                                  or determinants

                Pen'               649pen+, 649N                     121, 123
                mv-r               649mir-r,649N                     114, 118
                neo-r              649neo-r, 649N                    108, 109
                chm-r              649chm-r, 649N                    107, 111
                tet-r(C)           649tet-r(C),649N                  102,106
                tet-r(P)           649tet-r(P), 649N                 loo, 104
                str-r              649str-r, 649N                    103, 97
                Pig +
                                   649pig+, 649pig-                  130, 118
                MR                 649MR, 649N                       128,125
                MR-tN              649MR+N, 649tet-r(P)               96, 100
                met-r              131 36-wild, 131 36met-s          124,126

   For abbreviations see footnote to table I; met-s = sensitive to methicillin; met-r = resistant to
   * Relative to that of the cells without the determinant or determinants (= 100); results of
duplicate experiments.
140                            R. W. LACEY AND I. CHOPRA
 13136-wild and its methicillin-sensitive counterpart (13136met-s; Lacey,
 1972b). The doubling time of no. 13136-wild was about 25 % longer than that
of 13136met-s (table 11),which is consistent with the hypothesis that the genes
for methicillin resistance are carried by a relatively large plasmid (Lacey,
 19723). Although certain features of streptomycin resistance in strain 649
are similar to those of methicillin resistance-the genes cannot be isolated as
CCC DNA and transduction occurs at low frequency if at all (Lacey and
Chopra, 1974)-the presence of genes conferring streptomycin resistance had
no effect on growth rate. A variety of pigmented strains have previously been
found to have slower growth rates than their non-pigmented derivatives
(Grinsted and Lacey, 1973). Similarly, in strain 649, the pigmented cell
(i.e., no. 649N) had a doubling time c. 25% longer than the non-pigmented
(no. 649 pig -).
     In summary, six plasmids (determining production of penicillinase or of
pigment, or resistance to metal ions, neomycin, chloramphenicol, or methi-
cillin) reduced the growth rates of cells possessing them, but two (coding for
resistance to tetracycline or streptomycin) had little effect on generation time.
     However, the effect of plasmid carriage on growth kinetics in vitro was
complicated by variations in the duration of the lag phase that often did not
correspond to the effects that these plasmids had on the exponential growth rate.
Thus the plasmids determining resistance to neomycin and methicillin reduced
the lag phase, whilst the remainder lengthened this period. These variables
are a factor in determining the changes in the proportions of organisms in mixed
cultures at 20 h compared with those at the beginning of the experiment
(table 111).

  Efect of incubation for 20 h in mixed culture* on the proportion of determinant-positive
                                to determinant-negative cells

                         Plasmid or     Change in ratio$ determinant-
                        other genetic   positive: determinant-negative
                        determinant?      cells between 0 h and 20 h

                      Pen'                        0.53, 0.58
                      mir-r                       0.63, 0.56
                      neo-r                       1.28, 1-00
                      chm-r                       0*57,0*77
                      tet-r(C)                    0*89,0*68
                      tet-r(P)                    0-92, 0.73
                      str-r                       0.34, 0-29
                      Pig -t                      0.41, 0.50
                      MR                          0.39, 0.30
                      MR-tN                       1-73, 2.02
                      met-r                       1.33, 1.74

                  * Composition of mixed cultures as in table 11.
                  t For abbreviations see tables I and 11.
                  $ A constant ratio = 1.0; results of duplicate experiments.
          PLASMID CARRIAGE AND STAPHYLOCOCCAL VIRULENCE                                                 141

               Growth kinetics of cultures 649MR and 649MR-+N
    The mean generation time of strain 649MR was c. 25% longer than that of
no. 649N (table 11), and the ratio of 649MR to 649N cells was greatly reduced
after incubation for 24 h in mixed culture (table 111). A derivative of strain
649MR that had apparently lost all plasmid genome except for the determinant
for pigmentation (649MR-+N) might be expected to show a return to growth
characteristics of strain 649N. The growth of strain 649MR-+N was monitored
in mixed culture with no. 649tet-r(P) which has growth kinetics similar to that
of no. 649N (table 11). The mean doubling time of strain 649MR+N was
identical to that of no. 649tet-r(P) and hence by implication to that of no. 649N,
but the lag phase of no. 649MR+N was longer than that of no. 649tet-r(P) so
that the relative proportions of each differed after growth for 24 h (table 111).

               Efect o plasmid carriage on the virulence o strain 649
                      f                                     f
                          for 10-day chick embryos
    In preliminary experiments, the intra-allantoic injection of about 104-105
viable cocci into 10-day chick embryos produced about 50% mortality in
4 days. Accordingly, subsequent inocula were standardised to c. 4 x 104 viable
cocci (range 2.3 x 104-53 x 104).
    After the inoculation, the viability of chick embryos was assessed daily
for 4 days and results were expressed as total numbers of " viable days " or
" non-viable days ".    When cultures differed in their lethality, these differences
occurred to a similar extent on each day. Experimentsperformed with the same
cultures prepared on different occasions gave good agreement. For example,
when strain 649MR+N was injected into a series of embryos on five different
occasions, the proportions of viable days to non-viable days were respectively
27 : 40,28 : 40,25 : 40,29 :48 and 19 : 36.
    Each of the derivatives of strain 649N was less pathogenic than strain 649N
itself (table IV), although the figures for strain 649tet-r(C) were not significantly

                                         TABLE   IV
    Efect of plasmid carriage on the lethality of strain 649 for 10-day-old chick embryos

                                Number      Number (and percentage) of
           Strain*                                                           \
                               inoculated         !
                                                "&$             non-viable

    649N                            52         43 (21)           165 (79)
    649MR                           30         85 (70)            35 (30)        x2 = 80.5;   P<O.OOl
    649MR-tN                        29         75 (65)            41 (35)        x2 = 62.2;   P<O.OOl
    6 4 9 M ~ + N(after
      four passages)                24         58 (60)            38 (40)        x2 = 46.7; P<O.OOl
    649tet-r(P)                     32         49 (38)            79 (62)        x2 = 12.4; P<O*OOl
    649tet-r(C)                     29         35 (30)            81 (70)        x2 =   3.7; P>0*05

                          *   For abbreviations, see table I.
                          -f Of difference from results obtained with strain 649N.
142                              R. W.LACEY AND I. CHOPRA
different at the 5 % level from those for strain 649N. Strain 649MR was the
least virulent derivative and resulted in only 30% non-viable days compared
with 79% for strain 649N. Stain 649MR+N was also avirulent and not
significantly different from strain 649MR 012 = 1-03; P>0.05). Possession
of the plasmid coding for tetracycline resistance was associated with a significant
decrease in non-viable days-from 70% for strain 649N to 62% for no.
649tet-r(P)-although the effect was less marked than that obtained with
no. 649MR. The differences in virulence between organisms with chromosomal
or plasmid-mediated tetracycline resistance were not significant (x2 = 1 -78;
     An unexpected finding was that even after loss of plasmids from strain
 649MR, the resultant derivative (649MR+N) was as avirulent as no. 649MR.
 Attempts to restore virulence to strain 649MR+N by repeated passage were
 unsuccessful. After four passages in chick embryos, the viability figures
 obtained on subsequent inoculation of embryos were not significantly different
from those for the unpassaged culture (x2 = 0.40; P>0.05, table IV). Thus
loss of virulence in strain 649MR-+N seems irreversible.
    To account for the alterations in virulence of strains 649MR and 649MR-+N,
these and strain 649N were assayed quantitatively for the production of
a-, and 6-haemolysin, coagulase, fibrinolysin, DNAase, and Tween 80 lipase,
but the cultures did not differ in these respects. Furthermore, their colonial
morphology, pigmentation, and phage-typing patterns were similar.
    The pathogenicity of culture filtrates of strains 649MR+N and 649N were
also determined to test further the possibility that loss of production of extra-
cellular factors was responsible for avirulence. Filtrates from stationary-phase
cultures of each (equivalent to 5 x 1 7 viable cells) were added to 3 x 104
viable cocci of strain 649MR+N, and each mixture was then inoculated into
chick embryos. No differences in pathogenicity resulted, the viable to non-
viable days being 29 : 19 for each mixture.
    Because differences in diffusible products were unlikely to explain the
variation in virulence, a structural component of the cell itself was probably
responsible. However, there was no signgcant difference in pathogenicity

                                            TABLE  V
           Frequency of transduction' of plasmids coding for resistance to tetracycline,
              neomycin and cadmium ions (pen+>to strains 649N and 649MR-+N

                                         Frequency of transduction of the
                         Plasmidt               plasmid to strain
                                            649N             649MR-tN

                      tet                 6.2 x 10-6          6.3 x 10-6
                      neo                       ~
                                          3 . 7 10-7          5.3 x 10-7
                      pen   +                 x
                                          1-51 10-5           1.5 x 10-7

      *   Transducing lysates obtained by propagaion of phage 88 on derivatives of strain 649.
          For abbreviationssee footnote to table I.
              PLASMID CARRIAGE AND STAPHYLOCOCCAL VIRULENCE                                        143

when UV-killed cells of either no. 649N or no. 649MR+N (equivalent to
108 cells) were added separately to viable cells of no. 649MR-+N (3.6 x 104 cells)
and the mixtures tested independently for their effect on chick embryos. With
the addition of dead cells from strain 649N, the ratio of viable to non-viable
days was 44 : 40; with the addition of dead cells from strain 649MR+N,
it was 49 : 31 ( x 2 = 1-31; P>0*05).

           Establishment o plasmids in strains 649N and 649MR+N
    A possible reason for the loss of virulence in strain 649MR+N might be
that only avirulent cells of strain 649N can act as recipients in transduction
experiments. If so, both strains 649MR and 649MR-tN would be expected
to be avirulent. To test this possibility nos. 649N and 649MR+N were
examined for their capacity to act as recipients when the pen+, neo-r and
tet-r plasmids were transduced to them. No differences in transduction fre-
quency to the recipients were found (table V). This seems to exclude the
possibility that " avirulent " cells are superior to " virulent " in accepting

      Efect o carriage of several plasmids on the survival o Staphylococcus
              f                                             f
                             aureus strain 649 on glass
     The loss or gain of individual plasmids has previously been found to have
little effect on the capacity of staphylococci to survive on glass (Lacey, 1972~).
The effect on survival of a large number of plasmids present simultaneously is
not known. To test this, the survival of strains 649MR and 649N was studied
by comparing their survival on glass relative to a standard strain (no. 6936)
(Lacey, 1972~).At 21 "C, 30°C and 37"C, there was no difference in the relative
survival of the two strains (table VI).

    Analysis o membrane polypeptides from strains 649N and 649MR+N
   Polyacrylamide-gel electrophoresis in the presence of sodium dodecyl
sulphate (SDS) resolved membrane preparations from strains 649N and
649MR+N into approximately 35 polypeptide bands (the figure) with M.W.
between approximately 15,000 and 120,000 daltons (in this system, polypeptide

                                                TABLE  VI
                              Survival of strains 649N and 649MR on glass

                                       Survival* of the indicated strain at a temperature of
                      Strain                                     A                         \
                                            21"C              30°C               37°C

                   649MR           1      119,122
                                          113, 121
                                                            104, 130
                                                            137, 161
                                                                               185, 197

                               *   Expressed relative to that of strain 6936 (%).
  J.   MED. MICROBIOL.-VOL.    8 (1975)                                                        K
144                        R. W. LACEY AND 1. CHOPRA
mobility is inversely related to molecular weight). As previously noted
(Theodore and Panos, 1973 ; Chopra et al., 1974), membranes from a number of
staphylococcal strains are characterised by the presence of a major polypeptide
of M.W. c. 35,000 daltons (indicated by dotted lines in the figure).
    Although there was some quantitative variation between the profiles
obtained from individual preparations of the same culture, there were no
qualitative nor consisent quantitative differences between membrane poly-
peptides prepared from the virulent (649N) or avirulent (649MR-tN) strains
(respectively traces a and b in the figure). In addition, passage of no.
649MR+N had no detectable effect on the mobility or content of membrane
polypeptides (trace c).
    Although the pathogenicity of the staphylococcusfor man and for chick em-
bryos is closely correlated (McCabe, 1963 and 1964), caution must be exercised
in applying our findings in chick embryos to infection in man. Similarly, the
possession of certain plasmids produces marked and reproducible alterations
in the growth kinetics of staphylococci in vitro, but the extent to which similar
variations occur in vivo is uncertain. These findings do suggest, however, that


FIG.1 .-Densitometer tracings of membrane polypeptides separated by SDS polyacrylamide disk-
   gel electrophoresis: a = 649N; b = 649MR+N; c = 649MR-N after four passages in chick
   embryos. The broken lines represent a polypeptide of M.W. c. 35,000 daltons (see text).
          PLASMID CARRIAGE AND STAPHYLOCOCCAL VIRULENCE                         145

 in general plasmid carriage will be deleterious for the organism in vivo; and
  there is no evidence that possession of any of the plasmids tested benefits the
 cell other than by enabling it to withstand the effect of antibiotics. Thus i     n
  vivo, in the absence of antibiotics, plasmid-negative cells will probably be at an
 advantage in comparison with cells possessing plasmids. This may explain the
 decrease in the numbers of resistant staphylococci that occurred after the
 withdrawal of certain antibiotics (e.g. Lepper et al., 1954; Forfar et al., 1966;
 Bulger and Sherris, 1968; Ridley et al., 1970). This decline has probably
 resulted from plasmid loss in addition to overgrowth of plasmid-positive by
 plasmid-negative organisms.
     Our finding that decrease of virulence persists even after plasmids have been
 lost from the cell may have considerable clinical significance. For example,
 it might explain the relatively low invasiveness of the hospital staphylococci
 in recent years (Williams, 1971). Although the genetic and biochemical basis
 of this loss of virulence is not clear, the genes determining virulence ” may

 be extrachromosomal ; that loss of virulence in strain 649MR-+Nis irreversible
 is consistent with this view. Because some plasmids become unstable in the
 multiresistant strain described here (649MR), probably because of the ability
 of the cell to maintain only a finite amount of plasmid DNA (Lacey and
Chopra, 1974), loss of genome for virulence ” could be a further manifestation

 of the genetic instability of strain 649MR.
     The lack of correlation between virulence and the synthesis of extracellular
products suggests that the differences in virulence we have observed may be
related to changes in the cell surface, though the presence of an extremely
 unstable soluble “ virulence factor ” has not been formally excluded. Because
changes in the bacterial membrane may be necessary to maintain plasmid
DNA within the cell, an alternative explanation for the loss of virulence may
be that plasmid carriage brings about permanent alterations in the membrane.
However, neither qualitative nor consistent quantitative differences in membrane
polypeptides could be demonstrated between virulent and avirulent derivatives
that lacked known plasmids. Further biochemical analysis of these strains
may reveal differences in the cell surface, but it will be difficult to establish
that any abnormality ” is responsible for the loss of virulence and is not an

incidental accompaniment of it, particularly because the addition of dead whole
cells did not enhance the virulence of an avirulent derivative.
     Coincidental loss of virulence and of plasmid carriage in strain 649MR
cannot be excluded, but the partial loss of virulence in cells with one plasmid
-e.g., in strain 649tet-r(P)-suggests that a causal relationship between plasmid
carriage and avirulence does exist. It was not possible to test a “ cured”
derivative of strain 649tet-r(P) for virulence because of the stability of the
plasmid in this derivative.
     The possibility that only avirulent cells can acquire plasmids has been
excluded by showing that virulent and avirulent cells are equally competent
recipients of plasmids.
     The presence of several plasmids simultaneouslyhad no effect on the survival
of the organism on glass. The latter probably corresponds to survival on
146                        R. W. LACEY AND I. CHOPRA

exposed skin (Lacey, Alder and Gillespie, 1970), so it is unlikely that any
change in the epidemiology of hospital staphylococci is due to alterations in
its survival in air or on the body surface. There is also little evidence that the
frequency of nasal carriage of staphylococci has declined recently. Thus
decrease in the sepsis rate is more likely to be related to loss of virulence of
the organism than to other factors.
    Even if carriage of plasmids has been responsible for a decline in virulence
of hospital staphylococci, strains may be expected to evolve in the future that
are both fully virulent and resistant to many antibiotics. This could occur by
the selection of strains capable of maintaining stably a large amount of extra-
chromosomal DNA, or by integration into the chromosome of regions of
plasmid DNA specifying antibiotic resistance and virulence. This process has
probably already occurred for the genes coding for tetracycline resistance
(Kayser, Wust and Corrodi, 1972; Chopra et al., 1974).

    The possession of any of eight different plasmids by Staphylococcus aurew
strain 649-either singly or simultaneously (in no. 649MR)-caused changes in
growth kinetics. Six of the plasmids caused an increase in exponential doubling
time (by 8-25%), and most also altered the duration of the lag period. Strain
649MR was significantly less virulent for 10-day chick embryos than the
corresponding plasmid-negative culture (no. 649N). The avirulence persisted
even after loss of the plasmids from no. 649MR. The presence of a single
plasmid specifying tetracycline resistance produced a moderate reduction in
virulence, but chromosomal tetracycline resistance had an insignificant effect
on it. The decrease in virulence could not be attributed to reduced formation
of soluble products. It probably resulted from alterations in the cell surface,
but membrane-polypeptide profiles of virulent and avirulent cells lacking
plasmids were similar. Survival of strains 649MR and 649N on glass was
identical. Therefore, reduction in the incidence of staphylococcal sepsis may
be due in part to loss of virulence that has resulted from plasmid carriage.
    We thank Mrs E. Lewis for excellent technical assistance and D V. G. Alder for phage
typing. Financial support was by a grant from the Medical Research Council to Professor
M. H. Richmond.

ALDER, G., GILLESPIE, A. AND HERDAN, 1953. Production of opacity in egg-yolk
       V.               W.                     G.
   broth by staphylococci from various sources. J. Path. Bact., 66, 205.
ANDERSON,D. 1974. The effect of R-factor carriage on the survival of Escherichia coli
   in the human intestine. J. med. Microbiol., 7 , 85.
ASHESHOV, H. AND WINKLER, C. 1966. Staphylococcus areus of the ‘‘ 52, 5 2 ~ 80,
          E.                    K.                                                        ,
   81 ” complex. Nature, Lond., 209, 638.
        R.                 J.
BULGER, J. AND SHERRIS,C. 1968. Decreased incidence of antibiotic resistance among
   Staphylococcus aureus: a study in a university hospital over a 9-year period. Ann. intern.
   Med., 69, 1099.
          PLASMID CARRIAGE AND STAPHYLOCOCCAL VIRULENCE                                 147
BULOW,P. 1971. Prevalence of extrachromosomal drug resistance. Staphylococci in
    Danish hospitals during the last decade: factors influencing some properties of pre-
    dominant epidemic strains. Ann. N. Y. Acad., 21 Sci., 182.
CHOPRA, LACEY, w. AND CONNOLLY, 1974. The biochemical and genetic basis of
          I.,        R.                         J.
    tetracycline resistance in Staphylococcus aureus. Antibiotics Chemother., 6, 397.
DYKE,K. G. H., PARKER, T. AND RICHMOND, H. 1970. Penicillinase production and
                             M.                       M.
    metal-ion resistance in Staphylococcus aureus cultures isolated from hospital patients.
    J. med. Microbiol., 3, 125.
FORFAR,O., KEAY, J., MACCABE, F., GOULD, C. AND BAIN, D. 1966. Liberal
                        A.               A.              J.             A.
    use of antibiotics and its effect in neonatal staphylococcal infection, with particular
    reference to erythromycin. Lancet, 2, 295.
GRINSTED, AND LACEY, W. 1973. Ecological and genetic implications of pigmentation
            J.              R.
    in Staphylococcus aureus. J. gen. Microbiol., 75, 259.
HALLANDER, 0. 1963. Fractionation of staphylococcal toxins by gel filtration. Acta
   path. microbiol. scand., 59, 543.
JESSEN, AND BULOW,P. 1967. Changes of pathogenicity of Staphylococcus aureus by
    lysogenic conversion influencing lipase production, as evidenced by experimental skin
    infections in pigs. Acta path microbiol. scand., Suppl. 187, p. 48.
        O.,               K.,                       V.
    staphylococci and staphylococcal infections. A ten-year study of bacteria and cases
    of bacteraemia. New Engl. J .Med., 281, 627.
JEVONS, P. 1961. " Celbenin "-resistant staphylococci. Br. rned. J., 1, 124.
JEVONS, P., JOHN, AND PARKER, T.1966. Cultural characters of a newly recognised
         M.            M.
    group of hospital staphylococci. J. clin. Path., 19, 305.
KAYSER, H., WUST,J. AND CORRODI, 1972. Transduction and elimination of resistance
          F.                              P.
    determinants in methicillin-resistant Staphylococcus aureus. Antimicrob. Agents
     Chemother., 2, 217.
LACEY, W. 1972a. Effect of antibiotic resistance on the survival of Staphylococcus aureus.
    J. clin. Path., 25, 713.
LACEY, W. 19726. Genetic control in methicillin-resistant strains of Staphylococcus
     aureus. J. rned. Microbiol., 5, 497.
LACEY, W., ALDER, G. AND GILLESPIE, A. 1970. The survival of Staphylococcus
         R.               V.
     aureus on human skin. An investigation using mixed cultures. Br. J. exp. Path., 51,305.
LACEY, W. AND CHOPRA, 1974. Genetic studies of a multi-resistant strain of Staphylo-
        R.                     I.
     coccus aureus. J. med. Microbiol., 7, 285.
LACEY, W., LEWIS, AND GRINSTED, 1973. Loss of antibiotic resistance in Staphylo-
        R.               E.                  J.
     coccus aureus in vivo probably resulting from cloxacillin therapy. J. med. Microbiol.,
     6, 191.
          M.                   B.,              H.            G.                      S.
     Epidemiology of erythromycin-resistant staphylococci in a hospital population-
    effect on therapeutic activity of erythromycin. Antibiot. A., 1953-54, p. 308.
MCCABE, R. 1963. Staphylococcal infections in chick embryos. J. clin. Invest., 42,953.
 MCCABE, R. 1964. Studies of staphylococcal infections. 1. Virulence of staphylococci
     and characteristics of infections in embryonated eggs. J. clin. Invest., 43, 2146.
RIDLEY, BARRIE, LYNN, AND STEAD, C. 1970. Antibiotic-resistant Staphylo-
          M.,           D.,         R.              K.
     coccus aureus and hospital antibiotic policies. Lancet, 1, 230.
SMITH, W. AND LINGGOOD, A. 1971. The transmissible nature of enterotoxin pro-
        H.                         M.
     duction in a human enteropathogenic strain of Escherichia coli. J. med. Microbiol.,
     4, 301.
THEODORE, S. AND PANOS, 1973. Protein and fatty acid composition of mesosomal
     vesicles and plasma membranes of Staphylococcus aureus. J. Bact., 116, 571.
 WILLIAMS, E. 0. 1971. Changes in the virulence and antibiotic resistance of Staphylo-
     coccus aureus. In Bacterial infections: changes in their causative agents, trends and
     possible basis, edited by M. Finland, W. Marget, and K. Nartmann, Berlin, p. 99.

To top