Reprinted from

Proc. Natl. Acad. Sci. USA

Vol. 74, No. 4, pp. 1680-1682, April 1977
Genetics

Replication and expression of plasmids from Staphylococcus aureus

in Bacillus subtilis

(DNA/genetic transformation/molecular cloning/biohazards)

S. D. EHRLICH*

Department of Genetics, Stanford University Medical School, Stanford, California 94305

Communicated by Joshua Lederberg, January 31, 1977

ABSTRACT One S. aureus plasmid coding for tetracycline
resistance, pT 127, and four plasmids (pC194, pC221, pC223, and
pUB1I12) coding for chloramphenicol resistance have been in-
troduced by transformation into B, subtilis. The plasmids rep-
licate in—and confer antibiotic resistance upon—their new
host. These experiments show that the potential for genetic ex-
change between diverse bacterial species is greater than has
been commonly assumed.

 

Most of the plasmids studied so far have a narrow host range.
Some, however, can replicate in a wider host range. For ex-
ample, plasmid RP4 from Pseudomonas aeruginosa can be
transferred to other Gram-negative bacteria such as Escherichia
coli, Salmonella typhimurium, Klebsiella aerogenes, Rhizo-
bium leguminosarum, and Agrobacterium tumefaciens (1).

Another instance of plasmid-replicon transfer among bac-
terial species, perhaps even more widely separated, is reported
in this work. Five Staphylococcus aureus plasmids, coding for
tetracycline or chloramphenicol resistance, have been intro-
duced by direct DNA transformation into Bacillus subtilis. The
plasmids can replicate and express their genetic information
(antibiotic resistance) in this new host.

MATERIALS AND METHODS

Bacterial Strains. S. aureus strains used were SA231 (pC194)
Cm', RN154 (pC223) Cm*, RN1305 (pC221) Cm", RN1777
(pS177) Sm, RN1801 (pT127) Te", RN1953 (pK545) Km*/
Nm’, and RN2438 (pUB112) Cm' from R. Novick (2). B. sub-
tilis strains SB634 thy~ aroB tyr-1 and SB748 his-2 trypC2
thy~, derived from $B168, are from the Stanford collection.

Media. B. subtilis was grown in L and Penassay liquid
media. Resistant bacteria were selected on L-agar plates sup-
plemented with antibiotics [tetracycline (Tc), 15 ug/ml; chlo-
ramphenical (Cm), 5 ug/ml; streptomycin (Sm), 30 ug/ml;
kanamycin (Km), 3 ug/ml] S. aureus cells were grown in CY
liquid media or on GL-agar plates (3).

DNAs and Enzymes. Plasmid DNAs were prepared from
S. aureus strains essentially as described by Novick (2). Low-salt
lysates of stationary phase cultures were clarified by centrifu-
gation, concentrated with polyethylene glycol (molecular
weight; 6000), and centrifuged to equilibrium in CsCl density
gradients containing ethidium bromide. Cleared lysis treatment
of B. subtilis strains was essentially as described for E. coli (4).
Lysates were then processed similarly to the S. aureus ones (see
above).

EcoRI endonuclease and T4 ligase were purified and used
as described (5, 6). HindIIS was a commercial preparation
(BioLabs).

 

Abbreviations: Tc, tetracycline, Cm, chloramphenicol; Sm, strepto-

mycin; Km, kanamycin; Nm, neomycin.

* Present address: Institut de Biologie Moléculaire, Faculté de Science,
2, pl. Jussieu, Paris 75005, France.

1680

Transformation Procedure. Induction of competence and
transformation of B. subtilis were as described (7).

Electrophoresis and Electron Microscopy. Horizontal
agarose slab gels were used as described (8). Electron micros-

copy was also performed (9).

RESULTS
Tetracycline resistance plasmid pT127

S. aureus strain RN1801 carries the plasmid pT127 that confers
tetracycline resistance on its host (3). The plasmic DNA can
transform B. subtilis strains to tetracycline resistance: 10° cells
of SB634 (competence level 0.2%) exposed to 0.1 ug of pT 127
DNA yielded 10 Tc! colonies. No colonies were observed if ei-
ther the cells or the DNA was omitted. The typical appearance
and phenotypic match of the SB634 parent strain, auxotrophic
thy~aroB~ tyr, to the Tc’ colonies confirmed that the colonies
were B. subtilis.

The level of resistance of the transformants was higher than
25 ug/ml in L-broth, although at concentrations above 15
ug/ml some inhibition of growth could be seen. The parental
B. subtilis strain is inhibited by 5 ug/ml of Tc. Resistance was
found to be a stable trait: growth for some 20 generations in
liquid medium devoid of antibiotic, followed by plating on the
solid medium of the same type, resulted in less than 2% ob-
served colonies sensitive to tetracycline, as revealed by rep-
lica-plating on medium supplemented with the drug.

One of the Tc colonies was chosen for further study. A
profile of the cesium chloride/ethidium bromide density gra-
dient for its cleared lysis supernatant is displayed in Fig. 1. Two
peaks of radioactivity can be seen. Identically treated parental
Tc cells yielded only the lower density peak (Fig. 1) composed
of linear DNA molecules.

The heavier peak, detected in the extract of Tc" transfor-
mants, contains supercoiled circular DNA molecules, as re-
vealed by electron microscopy. These match the pT127 DNA
molecules extracted from S. aureus, by the following criteria:
(i) the two intact circular DNA preparations have identical
electrophoretic mobilities; (ii) they are both resistant to EcoRI
endonuclease and are cleaved by the HindIII nuclease into
three matching segments. These data are displayed in Fig. 2.

Equivalence of the two DNA preparations was verified by
further genetic tests: two B. subtilis strains, SB634 and SB748,
were transformed to tetracycline resistance with the DNA
isolated from the Tc Bacillus colonies, as well as with the pT 127
DNA isolated from S. aureus.

Chloramphenicol resistance plasmids

Four S. aureus Cm" plasmids (pC194, pC221, pC223, and
pUB112) were tested for their ability to transform B. subtilis
to chloramphenicol resistance. The results were similar to those
described for the Tc" plasmid, above.
Genetics: Ehrlich

 

 

 

| T
10,000 |- tt 4
e
E :
° 5,000- , 4
rr 8
e \ e | .
| \f 5
é e.e/ Ne
3b
Leeeestoccocas |
0 10 20

Fraction number

FIG. 1. Cesium chloride/ethidium bromide gradient of B. subtilis
cleared lysates. We used a cleared-lysis procedure (see Materials and
Methods) on 200 ml of a tritium-labeled B. subtilis culture (1 «Ci of
|’H]thymidine per ml). The concentrated supernatant was centri-
fuged for 36 hr at 386,000 rpm in a Spinco 50 rotor. Tubes were punc-
tured at the bottom, and 0.22-ml fractions were collected. A 5-,l ali-
quot of each fraction was assayed for radioactivity. The solid line
corresponds to a Tc’ transformant; dashed line corresponds to the
parental strain. Density increases from right to left. Specific activity
of DNA was 5000 cpm/ug.

(i) B. subtilis transformants resistant to 50 ug/ml of chlor-
amphenicol (parental strains are sensitive to 5 wg/ml) were
obtained at about the same frequency as the Tc" transformants,
upon exposing competent cells to plasmid DNAs. Growth
without antibiotic for 20 generations led to the loss of resistance
in less then 2% of cells transformed with pC194, about 10% in
those transformed with pUB112, and about 20% for the other
two plasmids.

(ii) Cm* transformants contain plasmid DNA indistin-
guishable from that isolated from the corresponding S. aureus
strains as indicated by (a) electrophoretic mobility of intact
DNAs and (b) response toward EcoRI and HindIIl restriction
enzymes. The number of restriction sites detected in various
plasmids is indicated in Table 1, together with their molecular
weights.

(iii) Plasmid DNAs isolated from the transformed B. subtilis
cells carry the information specifying chloramphenicol resis-
tance and can transform other B. subtilis cells for the same
genetic determinant.

Streptomycin and kanamycin/neomycin plasmids

The results obtained with pS177Sm‘ and pK545Km'-Nm‘
plasmids differ from those obtained with the Tc’ and Cm". We

Table 1. Resistance markers, size, and HindIII sites
of S. aureus plasmids*

 

Resistance M,,* No. of
Plasmid marker millions Hindlil sites
pC194 Cm 1.8 1
pC221 Cm 3.0 1
pC223 Cm 3.0 i
pUBI12 Cm 3.0 1
pT127 Te 2.9 3

* None of the plasmids has an EcoRI-sensitive site.

* From Novick (2) and confirmed here by comparing cleaved DNAs
with the B. subtilis phage Phi-3-T EcoRI segments (5). M,, mo-
lecular weight.

} Sizes of the segments are 1.5, 0.9, and 0.4 million, respectively.

Proc. Natl. Acad. Sci. USA 74 (1977) 1681

 

 

Fig. 2.
A, C. and E contain DNA extracted from S. aureus; lanes B, D, and
F contain DNA from B. subtilis. Untreated DNA is in lanes A and B.
EcoRI-treated DNA in lanes C and D, and HindIll-cleaved DNA is
in lanes E and F.

Agarose gel electrophoresis of plasmid pT 127 DNA. Lanes

did not observe B. subtilis transformants resistant to strepto-
mycin or kanamycin. A number of explanations could be put
forward—these plasmids might fail to replicate or to express
their genetic information in the new host. Alternatively, a
particular regimen of selection might be required to detect the
transformants, because in both cases B. subtilis colonies resistant
to antibiotic, due to endogenous mutations, were observed at
a frequency of about 1077 for Km, 1076 for Sm.

Efficiency of transformation of B. subtilis with
plasmid DNAs

B. subtilis strains of competence greater than 0.1% could be
transformed with S. aureus plasmids at an efficiency of about
10-° colonies per genome equivalent, that is, close to 100
colonies per ug of DNA. Colonies per genome equivalent are
expressed on the basis of input DNA. (Under the conditions of
1682 Genetics: Ehrlich

these experiments only a small fraction of the DNA is taken up.)
The efficiency increased some 50 times, approaching the level
of 1077 colonies per genome equivalent (10* colonies per ug of
DNA) with plasmids isolated from transformed B. subtilis cells.
Less than 0.5% of the efficiency remained after cleavage of
plasmid DNAs with the HindIII restriction endonuclease.
Treatment of cleaved DNAs with T4 ligase, resulting in about
50% recircularization (as revealed by electron microscopy in-
spection), restored 30-50% of the original efficiency for Cm"
plasmids, and less than 0.2% for the Tc" one.

DISCUSSION

The experiments reported here indicate that (i) B. stbtilis
strains can be transformed to antibiotic resistance with S, aureus
plasmid DNAs; (ii) the transformants acquire plasmid DNA;
(iii) this DNA is indistinguishable from the S. aureus plasmid
DNA by criteria of size, restriction enzyme pattern, and genetic
information. This evidence shows that S. aureus plasmids can
replicate and be expressed in B. subtilis.

The five S. aureus plasmids introduced in B. subtilis can be
subdivided into three groups: Tc" (pT127), Cm" small (pC194)
and Cm! large (pC221, pC223, and pUB112, Table 1). Data
reported by Novick (3) indicate that the three plasmids of the
last group are not identical: they respond differently to various
treatments which induce relaxation of supercoiled plasmid-
protein complexes. It appears therefore that at least five dif-
ferent S. aureus plasmids can be maintained in B. subtilis.

The efficiency of interspecies transformation described here
is high enough to allow the process to be demonstrated easily
in the laboratory by employing competent cells and plasmid
DNA.

Hind Il restriction endonuclease cleavage of the plasmids
decreases their biological activity to < 0.5%. This might be due
to the possible presence of the restriction site within the gene
coding for the antibiotic resistance and/or to the destruction
of the circular structure of the plasmids necessary for their
replication in the host. Another plasmid, pFT23, a hybrid be-
tween the pSC101 replicon and the thy gene of the B. subtilis
phage Phi-3-T, did not lose any transforming efficiency when

Proc. Natl. Acad. Sci. USA 74 (1977}

made linear by the action of Bam endonuclease (5). In that case,
circular structure was not obligatory because the transforming
thy gene could be integrated into another replicating structure:
the chromosome of the host. Ligation of the cleaved Cm"
plasmids almost fully restored their biological activity, whereas
the Tc’ plasmid was not reactivated, presumably because of a
lower probability of correct reassembly of the three HindIII
segments.

S. aureus plasmids introduced into B. subtilis are promising
vectors for cloning in this new host because of their small size,
easily selectable markers, and a small number of cleavage sites
for certain restriction enzymes.

The demonstration that replicating plasmids are shared
among species of bacteria as widely diverse as Staphylococcus
and Bacillus, or Escherichia and Agrobacterium |P. aeruginosa
plasmid RP4, (1}| makes it likely that plasmid sharing occurs
commonly in nature. This is pertinent to our views of natural
microbial evolution and, in turn, to the uniqueness of con-
structing DNA recombinants in the laboratory, which is a
premise of much policy discussion.

I gratefully acknowledge interest, criticisms and support of Dr. J.
Lederberg, in whose laboratory these investigations were performed.
Dr. R. Novick kindly provided S. aureus strains. The work was sup-
ported by National Institutes of Health Grant 3RO1 CA 16896-18 and
National Aeronautics and Space Administration Grant NGR-05-
020-004S17 to Dr. J. Lederberg.

Chakrabarty, A. M. (1976) Annu. Rev. Genet. 10, 7-30.

Novick, R. (1976) J. Bacteriol. 127, 1177-1187.

Novick, R. P. & Brodsky, R. (1972) J. Mol. Biol. 68, 285-302.

Clewell, D. B. & Helinski, D. R. (1970) Proc. Natl. Acad. Sei. USA,

62, 1159-1166.

5. Ehrlich, $. D., Bursztyn-Pettegrew, H., Stroynowski, I. &
Lederberg, J. (1976) Proc. Natl. Acad. Sci. USA 73, 4145-4149.

6. Sgaramella, V., Bursztyn-Pettegrew, H. & Ehrlich, $. D. (1977)

Proc. X Miles Symposium, in press.

Stewart, C. (1969) J. Bacteriol. 98, 1239-1247.

Harris- Warrick, R., Elkana, Y., Ehrlich, $. D. & Lederberg, J.

(1975) Proc. Natl. Acad. Sci. USA 72, 2207-2211.

9. Sgaramella V., Ehrlich, S. D.. & Lederberg, J. (1976) J. Mol. Biol.

105, 603-609.

BONE

~1

oe