Gene, 2(1977) 75-93 75

0 Elsevier/North-Holland Biomedical Press, Amsterdam - Printed in The Netherlands

CONSTRUCTION AND CHARACTERIZATION OF NEW CLONING

VEHICLES
1. AMPICILLIN-RESISTANT DERIVATIVES OF THE PLASMID pMB9

(Plasmid vectors; molecular cloning; tetracycline resistance transposon; res-

triction enzyme mapping)

FRANCISCO BOLIVAR*, RAYMOND L. RODRIGUEZ# MARY C. BETLACH and

HERBERT W. BOYER ,
Department of Biochemistry and Biophysics, University of California, San Francisco, Calif
94143 (U. S. A. )

(Received January 31st, 1977)

(Accepted June 23rd, 1977)

SUMMARY

In vitro recombination via restriction endonucleases and the in vivo genetic

translocation of the Ap resistance (Apr) gene resulted in the construction of
a new cloning vehicle, the plasmid pBR313. This vector was derived from a
Co1E1-like plasmid and, while it does not produce colicin El, it still retains
colicin El immunity. The Apr and tetracycline resistance (Tcr) markers carried
in pBR313 were derived from the ampicillin transposon (TnA) of pRSF2124
and pSC101 respectively. During the construction of pBR313, the TnA com-
ponent was altered and the Apr gene in pBR313 can no longer be translocated.
This plasmid has a molecular weight of 5.8 Mdalton and has been charac-
terized using thirteen restriction enzymes, six of which (EcoRI, SmaI, Hpal,
HindIIl, BamHI and SaII) cleave the plasmid at unique restriction sites. This
allows the molecular cloning of DNA fragments generated by these six
enzymes. The restriction sites for the latter three enzymes, HindIII, BamHI
and Sal!, are located in the Tcr gene(s). Cloning DNA fragments into these
sites alters the expression of the Tcr mechanisms thus providing a selection,
for cells carrying recombinant plasmid molecules. An enrichment method
for AprTcS cells carrying recombinant plasmid molecules is described.

*Present addresses:{F. B. )Departamento Biologia Molecular, Instituto de Investigaciones

Biomedicas, Universidad Nacional Autonoma de Mexico, Mexico 20, D. F.; (R. L. R. )
Department of Genetics, Briggs Hall, University of California, Davis, CA 95616 (U. S.A. ).
Abbreviations: Ap, ampicillin; Cyc, cycloserine; Tc, tetracycline.
76

INTRODUCTION

There are several critical components which facilitate recombinant DNA

research. The ease with which recombinant DNA research can proceed will in
part depend on the improvement of one of these components, the cloning
vehicle. It is now clear that bacterial plasmids, bacteriophages and animal
viruses can serve as vectors for cloning DNA fragments. Two bacterial plasmids
used in most of the initial cloning experiments as cloning vectors are pSC101
(Cohen et al., 1973) and ColE1 (Hershfield et al., 1974). Other cloning
vectors have been derived from these two plasmids (Hamer and Thomas,
1976; Hershfield et al., 1976; So et al., 1976) which improve on their utility.
We have undertaken the development of a series of plasmids with the goal
of obtaining a set of multipurpose cloning vehicles. In this paper, we describe
the construction of plasmid pBR313 with the Co1E1 mode of replication
and which contains the genetic potential for resistance to the antibiotics
ampicillin (Ap) and tetracycline (Tc). A tentative restriction map has been
determined for this plasmid and its cloning properties have been characterized.
pBR313 permits molecular cloning with DNA fragments derived by endo-
nucleolytic action of the following restriction endonucleases: EcoRI, HindIIl,
BamHI, Sal!, Hpal and SmaI. A procedure for selecting transformants which
contain recombinant plasmids has been developed.

MATERIALS AND METHODS

(a) Bacterial strains

The following derivatives of E. coli K12 were used as recipient cells in
transformation experiments: HB101 F" pro leu thi lacY Strr rkmj EndoI-,
recA- (Boyer and Roulland-Dussoix, 1969); RR1 F- pro leu thi lacY Strr
rkmk; EndoI". The E. coli B strain HB50 (pro leu try his arg met thr gal lacY
Strr rgmg) was used to prepare unmethylated plasmid DNA for EcoRII
digestion (Yoshimori et al., 1972). The bacterial plasmids pRSF2124 and
pSC101 in the E. coli K12 C600 background (F" leu thi thr lacY) were kind-
ly provided by S. Falkow and S. Cohen, respectively. The plasmid ColEl
(JC411 F; leu his arg met gal mal xyl thy lacY Strr) was obtained from
D. Helinski. The plasmids pMB8 and pMB9 were maintained in and prepared
from HB101.

(b) Preparation of plasmid DNA

Preparation of plasmid DNA by amplification in the presence of chlor-
amphenicol (170 jig/ml) was performed according to Clewell (1972). The
DNA was purified by a modification of the cleared lysate technique of
Guerry et al., 1973. The cleared lysate was extracted with an equal volume
of phenol and the aqueous phase precipitated with two volumes of cold
ethanol. The resuspended DNA, in a5 ml volume of A50 agarose buffer
(500 mM NaCl, 50 mM Tris pH 8.0,1 mM EDTA), was applied to a 25 cm X
 77

50 cm Bio-Gel A50 agarose column (Bio-Rad), and the first peak fractions as
determined by A26onm were pooled and ethanol-precipitated. The plasmid
DNA was further purified by dye-buoyant centrifugation in a CsCl-propidium
diiodide gradient, equilibrated for 18 h at 36 000 rpm, 20°C, in a Spinco
SW 50.1 rotor. Small molecular weight RNA contaminates the plasmid DNA
at this stage.
The band of supercoiled DNA as visualized with UV light was collected
and the propidium diiodide extracted by passing the DNA over a1 cm X4 cm
Dowex (AG 50W-X8 Bio-Rad) column. The eluent was dialyzed against 10 mM
Tris-HCI, 1 mM EDTA, pH 7.4, ethanol-precipitated and resuspended in
50 mM Tris-HCI, 10 mM NaCl, 1 mM EDTA pH 7.4. DNA concentrations
were determined spectrophotometrically in the above buffer; A260 of 1.0 =
50 pg DNA/ml (Padmanabhan and Wu, 1972).
Preparation of plasmid DNA for the rapid analysis of the restriction endo-
nuclease digestion pattern was performed according to the procedure described
by Meagher et al. (1977).

(c) Enzymes
EcoRI restriction endonuclease was purified according to the procedure
of Greene et al. (1974). The restriction endonucleases, AIuI (Roberts et al.,
1976), HaeII (Roberts et al., unpublished observations), HaeIII (Roberts et
al., unpublished observations), Bg1I (Wilson and Young, unpublished observa-
tions), BamHI (Wilson and Young, 1975), EcoRII (Yoshimori et al., 1975),
Hincll (Landy et al., 1974), HindIIl (Danna et al., 1973), PstI (Smith et al.,
1976), and SaII (Arrand et al., 1976), were purified according to the proce-
dure described by Heyneker et al. (1976). Hpal (Gromkova and Goodgal,
1972) was obtained from BRL laboratories. T4 polynucleotide ligase, a gift
from H. Heyneker, was purified according to the procedure of Panet et al.
(1973). Colicin El was prepared from the strain JC411 as described by
Schwartz and Helinski (1971). All restriction enzymes were stored at -20°C
in 50% glycerol, 20 mM KH2PO4-K2HPO4 pH 7.0,1 mM EDTA, 1 mM NaN3,
and 100 mM NaCl.

(d) Preparation of various restriction endonucleaseDNA fragments

DNA fragments were generated with various restriction endonuc? eases by
digesting from 0.1 to 5.0 pg of DNA in a 10 to 30 pl volume according to the
following conditions for each endonuclease: BgII, Alul, HaelI, and HaeIII,
6 mM Tris-HCI pH 7.9,6 mM ß-mercaptoethanol, 6 mM MgC12; $pal, 50 mM
Tris-HC1 pH 7.5,5 mM ß-mercaptoethanol, 5 mM MgC12,50 mM NaCl; Smal,
15 mM Tris-HCI pH 9,6 mM MgC12,15 mM KCI; BamHI, 100 mM Tris-HCI,
5 mM MgC12 pH 7.6; EcoRI, 100 mM Tris-HCI, 5 mM MgC12,100 mM NaCl,
0.02% NP40 (Particle Data Laboratories) pH 7.6; EcoRI*, 2 mM Tris-HCI,
2 mM MgC12,20% glycerol pH 8.8; EcoRII, 100 mM Tris-HCI, 5 mM MgC12
pH 7.6; HinclI, 100 mM Tris-HCI, 66 mM MgC12,500 mM NaCl, 60 mM ß-
mercaptoethanol pH 8.0; HindIIl, 6.6 mM Tris-HCI, 6.6 mM MgC12i 50 mM
78

NaCl, 7 mM ß-mercaptoethanol pH 7.5; PstI, 90 mM Tris-HCI, 10 mM MgSO4

pH 7.5; SalI, 8 mM Tris-HCI, 6 mM MgC12,150 mM NaCl, 0.2 mM EDTA,
50 pg/ml bovine serum albumin. Endonuclease reactions were incubated at
37°C and stopped by heating to 63°C for 5 min or by the addition of 10 µl
of 5% sodium dodecyl sulfate, 25% glycerol, 0.025% bromophenol blue (stop
mixture).

(e) Molecular weight standards

For agarosegel electrophoresis, the following markers were used: the six
EcoRI-generated fragments of the X bacteriophage genome (13.7,4.68,3.7,
3.56,3.03, and 2.09 md. ); the EcoRI-generated linear forms of the plasmids
pVH51 (2.2 md. ) (Hershfield et al., 1976) and pMB8 (1.7 md. ) and the six
HindIIl-generated fragments of the SV40 genome (1.13,0.75,0.68,0.35,
0.28,0.13 md. ). For acrylamide gel electrophoresis the seven HindIIl-generated
fragments of the bacteriophage PM2 (3.5,1.34,0.6,0.29,0.26,0.155,0.07
md. ) were used as molecular weight standards (Wes Brown, personal commu-
nication).

(f) Ligation of DNA fragments

The ligation reactions were incubated at 12°C in volumes ranging from 50
to 1000µl containing 66 mM Tris-HC1,6.6 mM MgClz, 10 mM dithiothreitol
(Sigma), 0.2 mM ATP pH 7.6 and varying amounts of DNA. The reaction was
started by the addition of 0.5 to 5.0 µ1 of T4 polynucleotide ligase. -DNA
ligation was monitored by the electrophoresis of small aliquots of the reaction
taken at intervals.

(g) Agarose and acrylamide electrophoresis

Slab gels containing 0.8 to 1.0% agarose (Seakem) were prepared by auto-
claving the agarose for 10 min in Tris-EDTA-borate buffer (90 mM Trizma
Base (Sigma), 2.5 mM Na2 EDTA, 90 mM H3BO3 pH 8.2). After the gels
were poured and allowed to solidify, samples containing from 0.2 to 1 µg of
DNA were loaded onto the gels in a 20 to 40 µl volume per slot. Electro-
phoresis was performed at 130 to 150 V at room temperature for 2 h. Slab
gels of 7.5% acrylamide (1 mm thickness) were prepared by mixing 2.4 ml
of 1OX Tris-EDTA-borate buffer with 15.6 ml of water, 6 ml of 28% acryl-
amide (Bio-Rad), 0.6% bisacrylamide and 0.12 ml of 10% NH4SO3. After de-
gassing, TEMED (12 µml) (Bio-Rad) was added and the mixture poured im-
mediately. Acrylamide electrophoresis was performed at 120 V for 2 h. After
electrophoresis, all gels were soaked for 5 min in 4 jig/ml ethidium bromide
solution and illuminated with a short-wave UV transilluminator (Ultraviolet
Products, San Gabriel, Calif. ). A yellow no. 9 Kodak Wratten gelatin filter
and NP Type 55 Polaroid film were used with a MP-3 Polaroid camera to
photograph the gels.
 79

(h) Transformation of E. colt

Chloroform-sterilized DNA in a 100 µl volume was adjusted to a final con-
centration of 30 mM CaC12 and added to 0.2 ml of cells prepared for trans-
formation by the procedure of Cohen et al. (1972). The transformation mix-
ture was allowed to stand on ice for 1h after which it received a 60 sec, 42°C
heat pulse. The heat pulse was terminated by the addition of 5 ml of Luria
broth. Transformed cells were plated immediately or allowed to proceed into
the logarithmic phase of cell growth before plating.

RESULTS

(a) Construction and characterization of pMB9

The construction and characterization of a series of ColEl-like plasmid
derivatives have been reported (Betlach et al., 1976). From one of these
plasmids we derived a small plasmid which is similar to pVH51 (Hershfield et
al., 1976). pMB8 has a molecular weight of 1.72 X 106 daltons, is immune to
colicin El, exhibits a ColE1 mode of DNA replication in terms of copy num-
ber and replication in the presence of chloramphenicol but does not form a
detectable relaxation complex (Clewell and Helinski, 1969). Although pMB8
offers some advantages as a cloning vector (e. g., small size, good yields of
DNA, low background of protein synthesis in minicell systems), it does not
have a good selective marker or substrate sites for several of the commonly
used restriction endonucleases other than EcoRI. We then constructed a
composite plasmid which affords cloning with different restriction enzymes
by incorporating into pMB8 some components of the pSC101 plasmid which
can confer Tcr to the cell. This plasmid, pMB9, was constructed by ligating
the products of an EcoRI* endonuclease (Polisky et al., 1975) digest of
pSC101 DNA with an EcoRI endonuclease digest of pMB8 (Rodriguez et al.,
1976).
The plasmid pMB9 has a molecular weight of 3.5.106 daltons and one sub-
strate site for each of the following restriction endonucleases: EcoRI, HindIIl,
Sail and BamHI. The relative positions of these sites (Fig. 1) were determined
by acrylamide gel electrophoresis of double and triple digestions of plasmid
DNA with various restriction enzymes. Since pMB8 has no HindIII, BamHI
or Sail sites, these sites should be associated with that pSC101 DNA frag-
ment introduced in pMB9. This assumption is supported by the fact that
these three sites are present in pSC101 in the same relative positions (data
not shown) and that the molecular cloning of DNA into the HindIII, BamHI
and Sail sites alters the Tcr mechanism (Hamer and Thomas, 1976; Rodriguez
et al., 1976). Since DNA inserted into the EcoRI site of pMB9 does not alter
the expression of the Tcr mechanism, we believe that this site lies outside the
Tcr gene. However, cloning into the EcoRI site of pSC101 has been found to
affect the level and inducibility of Tcr (Tait et al., 1977).
Although EcoRI-generated recombinant plasmids of pMB9 can be selected
by virtue of their resistance to tetracycline, transformants with recombinant
80

plasmids constructed by insertion of DNA fragments generated by HindIIl,

BamHI and Sal! digestion and incorporated into the respective restriction
sites can only be selected by colicin El immunity. Since the action of colicin
El is extremely dependent on the physiological state of the cell and is
accompanied by a high frequency of spontaneous mutations to colicin
tolerance (Nagel De Zwaig and Luria, 1967), we chose to introduce another
selective marker into pMB9. This was accomplished by the genetic transloca-
tion of an ampicillin resistance marker (Apr) from pRSF2124 (So et al., 1976)
to pMB9.

(b) Construction of Ap''-TcT derivatives of pMB9

The Apr marker has been shown to be translocated on a 3.2.106 daltons
sequence of DNA which has been termed TnA (Heffron et al., 1975). Trans-
location of the TnA from pRSF2124 to pMB9 was accomplished by cotrans-
forming E. coli strain RRI with a total of 5µg of supercoiled pMB9 and
pRSF2124 DNA at a respective molecular ratio of 2: 1. Apr-Tcr transformants
which occurred at a frequency of 6.10-6 after 5 generations were screened for
plasmid DNA which gave a linear 6.7 mdalton plasmid molecule upon di-
gestions with EcoRI. While the majority of Apr-Tcr transformants contained
varying ratios of pMB9 and pRSF2124, approximately 20% of the total Apr.
Tcr transformants give one linear plasmid DNA molecule upon EcoRI di-
gestion. These EcoRI-generated linear molecules had molecular weights of
6.7.106 daltons which corresponds to the sum of the molecular weights of
the TnA (3.2.106 daltons) and pMB9 (3.5.106 daltons). These plasmid
molecules were shown to confer resistance to Ap and Tc when transformed
back into E. coli RR1.
The presence of a single asymmetrically located Barn HI site about 1.106
daltons from the end of the TnA (Heffron et al., 1977) made it possible to
localize the various TnA insertion sites in pMB9 (Fig. 1). In the case of two
Apr-Tcr derivatives of pMB9, designated pBR312 and pBR26, mapping the
relative positions of the restriction and TnA insertion sites was carried out
by digestion with combinations of various restriction enzymes as shown in
Plate I(a). In relation to the molecular weight standards (slots 4,7 and 10)
linear molecules of pBR312 and pBR26, having molecular weights of 6.7 -10'
daltons, were also generated by digestion with Hindill (data not shown).
Barn HI digests of pBR312 and pBR26 gave two fragments in each case
with molecular weights of 5.1 and 1.6.106 daltons and 5.06 and 1.65 -10"
daltons, respectively. By following a Bam HI digestion of these plasmids with
an EcoRI digest, the smallest band of pBR312 and the largest band of pBR26
were each cleaved by EcoRI to give a third band of 0.4.106 daltons. These
results enable us to localize the TnA insertion counterclockwise to the EcoRI
site in pBR312 and clockwise to the EcoRI site in pBR26. Since the smallest
BamHI fragments of pBR312 and pBR26 are less than 2.2.106 daltons, this
indicates that the orientation of the TnA is such that the one-third portion
which is known to carry the Apr gene (Heffron et al., 1977) is proximal to
the BamHI site in the Tcr gene (Fig. 1).
 81

Ecq RI

(3.5

pMB9 RI
Rl I
x 10° d. Hý
il!
" N1

~Tý- Tcý

sý1,ýP

p6R 312 pBR 26

6.7 x 10° d. 6.7 x 10° d.

N,

Fig. 1. Schematic representation of pMB9 DNA showing the Tcr region with the EcoRI,
HindIIl, Bam HI and SaII sites. The arrows represent the position of the IR from TnA in
two Apr derivatives of pMB9: pBR312 and pBR26.

(c) Construction and characterization of pBR313

Although pBR312 and pBR26 now possess a second strong selective marker
in the form of Apr, their potential usefulness as molecular cloning vectors is
diminished by the presence of an extra BamHI site and their increased molec-
ular weights. Consequently, we decided to remove the BamHI site contributed
by the TnA and simultaneously reduce the size of pBR312 by partial EcoRI*
digestion. After digestion, the DNA was ligated in a 500 µl volume and then
transformed into RR1. Apr-Tcr transformants, which occurred at a frequency
of 3.10"', were screened for plasmid DNA giving linear molecules upon
BamHI digestion. Out of 16 Apr-Tcr clones, 6 gave linear molecules of varying
molecular weights after treatment with BamHI. The three smallest plasmids
(pBR313, pBR315 and pMB316) were selected for further study.
The molecular weight and relative position of restriction sites of pBR313
were determined as described above and the results of the acrylamide gel
electrophoresis are shown in Plate I(b). The data from Plate I(b) have been
summarized in Fig. 2a. As can be seen in Fig. 2b, the EcoRI* digestion has
removed one of the two BamHI sites from pBR313, pBR315 and pBR316
and reduced their molecul weights by 0.8,1.5 and 1.4 -10' daltons respec-
tively. One useful feature o EcoRI* digestions is the ability to lose and
regenerate new EcoRI substrate sites. In the case of pBR316, the EcoRI site
has been lost, whereas it is moved 0.16 -10' daltons nearer to the HindIIl site
in pBR315 and even closer (0.02 -106 daltons) in pBR313 (Fig. 2b). This
suggests that there are at least two EcoRI* sites, 0.16 and 0.02 -10' daltons
from the HindIII site in pMB9 which can be used to recreate new EcoRI sub-
strate sites. It is also apparent from Fig. 2b that the segment of DNA between
the HindIII and the Sall sites of pBR313, pBR315 and pBR316, is unchanged
after EcoRI* digestion as compared to pBR312, and pMB9.
It should be mentioned at this point that the estimated size of the Apr and
Tcr genes represented in Fig. 2 were determined indirectly on the basis of the
reported values for the size of the TEM ß-lactamase (Datta and Richmond,
 83

1966) and the Tcr"associated proteins detected in the minicell system (Levy
and McMurry, 1974; Tait et al., 1977). Positioning the left-hand boundary of
the Tcr gene was based on our knowledge that cloning into the EcoRI site of
pBR313 did not affect Tcr while cloning into the Hindill site did affect the
expression of the Tcr mechanism. The position and size of the Tcr region is
also consistent with the orientation of the TnA in pBR26. This follows from
the consideration of the known position of the inverted repeats to the BamHI
site (Heffron et al., 1977). The 1.65.106 daltons BamHI-generated fragment
of pBR26 allows for 0.65 -10' daltons of DNA between the pMB9 specified
BamHI site and the end of the Tcr gene(s) after accounting for the 1.0 -10"
dalton segment of the TnA.
The restriction endonuclease PstI was used to further characterize pBR313.
As shown in Plate II(a) (slot 5), pBR313 has three PstI sites which give frag-
ments of 0.4,1.25 and 4.15.106 daltons. Since pMB9 was found to have no
PstI sites (Plate II(a), slot 2), it was concluded that the PstI sites were associ-
ated with the TnA. This conclusion is further supported by the presence of
five PstI sites in pRSF2124 (Plate 11(a), slot 3). Two of these five sites are
known to be carried on the ColEl portion of pRSF2124 (data not shown). It
can also be seen that two of the three PstI fragments present in pBR312
(Plate II(a), slot 4) 1.7 and 0.4 -10' daltons, are present in pRSF2124, while
only the smallest fragment is present in pBR313. From these results, we
concluded that the 1.7 . 106 dalton fragment of pBR312 was reduced to
1.25 . 106 daltons in pBR313 by the EcoRI* digestion. The PstI-EcoRI com-
bination digest shown in slots 7,8, and 9 of Plate II(a), corroborate the
placement of the TnA counterclockwise to the EcoRI site in pBR312 (Fig. 1)

Plate I. (a) Agarose slab gel electrophoresis of plasmids pBR312 and pBR26 cleaved by
EcoRI, BamHI and Sail endonucleases. Digested DNA (0.3 to 0.5 µg) was applied to the
sample slots in 40 Al volumes. Agarose gel electrophoresis was carried out as described in
MATERIALS AND METHODS. Molecular weight estimates are based on the 6a fragments
generated by EcoRI, the linear forms of the plasmids pMB8 and pVH51 and the six HindIII-
generated fragments of the SV40 genome (slots 4,7,10) (see MATERIALS AND
METHODS). Slots 1,2, and 5 show the EcoRI, Sail and BamHI digestions respectively of
pBR26 plasmid DNA. Slot 3 shows the double digestion EcoRI-SaII and slot 6 shows the
EcoRI-BamHI double digestions both in pBR26 DNA. Slots 9,12, and 13 show the BamHI,
SaII and EcoRI digestions respectively of pBR312 plasmid DNA. Slot 8 shows the EcoRI-
BathHI and slot 11 the EcoRI-SaII double digestions of pBR312 DNA. For explanation
see RESULTS (construction of Apr-Tcr derivatives of pMB9).
(b) Acrylamide slab gel electrophoresis of plasmid pBR313 DNA fragments obtained by
double and triple digestions using EcoRI, HindIIl, BamHI and SaII restriction endonucleases.
Gel electrophoresis was carried out as described in MATERIALS AND METHODS.
Molecular weight estimates are based on the (seven) HindIIl-(generated) fragments of the
PM2 phage genome. The restriction endonuclease digestion combinations for Fig. 3 are as
follows: (Slot 1) EcoRI-BamHI; (Slot 2) EcoRI-BamHl-SalI; (Slot 3) BamHI-Sall; (Slot 4)
EcoRl-SaII; (Slot 5) EcoRI-HindIII-SalI; (Slot 6) HindIIl-SaII; (Slot 7) HindIII BamHI;
(Slot 8) HindIII-BamHl, SaII; (Slot 9) EcoRI-HindIII. HindIIl-digested PM2 markers are
also present in slots 2 (bands 2,3,4,5,6,8 and 10), slot 5 (bands 2,3,4,6,7,8 and 9),
slot 8 (bands 2,3,4,5,6,8 and 10).
84

a
Hind III
Eco RII B`
Alu y/
ljoe III \e
Eco RI
Hoe 11 N.
Hoe III
Q/
\
NP
j
N` F
N co

5.8 /0

pBR 313
öl 5.8 x 10' d. ý
x ö
-
IN
0.
di Aj
G\

ý",ý'
OT
O `GEN
o
I)\-
11OOH

N UQRI Hind III §gMHIWI

pMB9 3.5md
b.gm HI
gp RI jjjpd III @qm HI WI Tcr
1.15 I 22-4718=15
. pBR312 6.7md
App Eco Al 1fjp¢ III @gm HI $QJ I Tcr

pBR313 5.8md
Apr f&2 RI Kind III n HI W, I Tcr
I16[-. I
18=. 115 pBR315 5.4md
Apr Hind III Bam HI Sal I Tcr
18 15 il I
. 16 5.2md
Apr Tcr
 85

since the Pstl-EcoRI-generated fragment (0.6 -101 daltons) was produced by

cleaving the largest Pstl fragment carrying the Tcr gene. The position of the
EcoRI site in pBR313 is localized at 0.46 mdaltons from one of the Pstl sites.
A derivative of pBR313 containing two Pstl sites (pBR317) was constructed
by transforming RR1 with ligated PstI-generated fragments of pBR313 and
selecting for Apr and Tcr. This plasmid was found to lack the 0.4 -10" dalton
PstI fragment seen in slot 5 of Plate II(a). After screening the plasmid DNA
of a number of Apr-Tcr transformants, it was noticed that the 1.25.106
dalton PstI fragment was present in all Apr-Tcr plasmids. Since this Pstl frag-
ment was also present in some Ap8-Tcr transformants, we believe that one
orientation of this fragment is associated with the Apr phenotype. Conse-
quently, this places one PstI site in the Apr gene. Another ApsTcr plasmid
(pBR318) was found to lack both 0.4 and 1.25 mdalton PstI fragments
present in pBR313 (data not shown).

(d) Characterization of pBR313 by EcoRII digestion

The EcoRII restriction patterns of pMB9, pBR312, pBR313, pBR316 are
shown in Plate II(b). As shown in slot 9 (Plate II(b)), EcoRII produces at
least 9 fragments in pMB9. Some of these fragments are common to pBR312
(slot 8), pBR316 (slot 6), and pBR313 (slot 5). As the largest EcoRII frag-
ment of pMB9 is not present in the EcoRII pattern of pBR312, we assume
that the TnA was inserted into this piece of DNA. As a consequence, four
new fragments appear in the EcoRII pattern of pBR312 which are not seen
in pMB9. Combination PstI-EcoRII digestions reveal that the two largest
fragments of pBR312 and pBR313 contain the three Pstl sites of the TnA
(Fig. 2a). By comparing the EcoRII patterns of pBR312 and pBR313, it can
be seen that the largest fragment of pBR312 (1.9 -101) was reduced by
0.8.106 daltons in pBR313 as a result of EcoRI* activity. As shown in
Fig. 2b, 0.2 -101 daltons of this reduction occurred between the EcoRI and
HindIII sites of pBR312 while 0.14.106 daltons was removed from between
the EcoRI and PstI sites (Plate II(a), slots 9 and 10). The remaining 0.46.106
dalton reduction occurred in that portion of the TnA containing the BamHI
site.
Slots 1,2 and 3 of Plate II(b) show HindlIf, Sail and BamHI digestions of
EcoRII-digested DNA of pBR313 respectively. In the case of the HindIll
digest, the largest EcoRII fragment (1.1.106 daltons) is cleaved to give two

Fig. 2. The circular restriction map of pBR313. (a) The relative position of restriction sites is
drawn to scale on a circular map divided into units of 1.10' daltons. The restriction sites
for AIuI, EcoRII, HaeII and HaeIII represent only those which could be mapped. (b) The
relationship between pMB9 and its Apr derivatives before (pBR312) and after (pBR313,
pBR315 and pBR316) EcoRI* digestion are represented linearly with respect to those
restriction sites located in the Tcs and Apr genes. The origin of replication in this plasmid
has been localized by restriction endonuclease analysis and electron microscopic determina-
tions (unpublished observations).
86

new fragments of 1.05 and 0.05 daltons. Slots 2 and 3 show that Sall
and BamHI digestions cleave the same 0.53 "106 dalton EcoRlr fragment
generating 0.29 and 0.24 -10' dalton fragments in the Sall digest and 0.4 and
0.13 -10' daltons in case of the BamHI digest. Since HindIII is shown to
cleave one EcoRII fragment while both Sall and BamHI cleave a different
fragment and the fact that Hindill and Bam HI sites are adjacent, indicated
 87

that there is one EcoRII site between Hindill and BamHI sites and no EcoRII
site between BamHI and Sail sites. Combination digests with additional
restriction endonucleases have enabled us to localize other EcoRII sites on
pBR313 (Fig. 2a).

(e) Mapping the substrate sites of the restriction endonucleases HincII, Hpal,
Smal, Bgll, Alu, HaeII and HaeIII
The restriction enzyme HinclI recognizes the sequence
GT PyyPu AC (Landy et al., 1974)
CA Pu? y TG

As shown in Fig. 2a, there are four Hincll sites in pBR313, one of which, is
present in the 0.4 -10' dalton PstI fragment missing in pBR317. It should be
noted at this point that the HpaI-HincII and SaII-Hincll double digestion
patterns are identical to the HincII pattern (data not shown). This is due to
the purine-pyrimidine ambiguity present in the Hincll substrate site which
enables this particular enzyme to recognize both the HpaI

GTT'AAC G'1TC GAC

and Sal! substrate
CAAtTTG CAGC TtG

sites (Danna et al., 1973; Bolivar and Shine, 1976, unpublished observation).
As shown in Fig. 2a, the EcoRI site is located 0.29 . 106 daltons from the
Hincll situated in the Apr gene (S. Falkow, personal communication) and
0.35 -101, daltons from the HincII-SaII site. Since a SaII-HpaI double digestion
generates a 1.3 dalton fragment which is also present in a Hincll digest,
this places the HpaI site clockwise of the Sail site as shown in Fig. 2a. When
the 2.29 . 106 dalton fragment produced by a Hincll digest of pBR313 is
cleaved by Smal, this fragment is reduced by 0.12 -10" daltons. The fact that

Plate II. (a) Analysis of PstI and PstI-EcoR1 single and double digestions of pMB9,
pRSF2124, pBR312 and pBR313 using agarose gel electrophoresis. Molecular weights
estimates are based on the 6r fragments generated by EcoRI, the linear forms of the
plasmids pMB8 and pVH51 and the HindIU generated fragments of the SV40 genome
(slots 1,6 and 11). The PstI digestion patterns of the various plasmids are as follows:
(slot 2) pMB9; (slot 3) pSF2124; (slot 4) pBR312 and (slot 5) pBR313. PstI-EcoRI double
digestions of these plasmids are as follows: (slot 7) pMB9; (slot 8) pSF2124; (slot 9)
pBR312 and (slot 10) pBR313.
(b) Acrylamide slab gel electrophoresis of EcoRII cleaved pMB9, pBR312, pBR313 and
pBR316 DNAs. Purified plasmids DNA were cleaved with EcoRII as described in
MATERIALS AND METHODS and the fragments were dialyzed and subjected to acryl-
amide gel electrophoresis. The seven PM2 HindII1-generated fragments were used as
molecular weight markers (slots 4,7 and 10). Slots 5,6,8 and 9 show the EcoRI1 pattern
of plasmids pBR313, pBR316, pBR312 and pMB9 respectively. Double digestions EcoRII-
HindIIl, EcoRIISal1 and EcoRII-BamHI of pBR313 DNA are shown in slots 1,2 and 3
respectively. For explanation see the text (RESULTS Section (d)).
88

the 1.42.106 dalton fragment produced by a SaII-SmaI double digest is also

reduced by 0.12 daltons when digested with HpaI, places the Smal site
0.12 -10" daltons clockwise from the HpaI site (Fig. 2a).
Combination digest of pBR313 and derivative pBR plasmids (Bolivar et al.,
1977) have enabled us to map the five BgII restriction sites shown in Fig. 2a.
The AIuI restriction enzyme cleaves pBR313 into more than eighteen frag-
ments, some of which have been mapped by analysis of double digestion
patterns. At present, we have been able to localize seven AM sites on pBR313.
The AluI site between the EcoRI and HindIIl sites (Fig. 2a) was localized by
the determination of the nucleotide sequence in this region of the DNA
(J. Shine, unpublished observation). Using the strategy of combination digests,
we were also able to map eleven HaeII sites and 4 HaeIII sites present in
pBR313 (Fig. 2a).
(f) Molecular cloning of various restriction endonuclease-generated fragments
in pBR313
DNA fragments from various sources were produced by digestion with
EcoRI, HindIIl, BamHI, SaII and HindIII-BamHI restriction enzymes and
cloned in their respective sites in pBR313 (Table I). EcoRI-recombinant
plasmids of pBR313 gave Apr-Tcr phenotypes while BamHI, Sal! and HindIIl-
BamHI-recombinant plasmids were Apr. Tcs. While some transformant-carrying
HindIIl-recombinant plasmids were AprTcs, others were found to have a low-
level Tcr which was observed when recombinant transformants were incubated
for more than 24 h on Luria agar plates containing 10 jig/ml Tc. As in the
caseof the EcoRI recombinant plasmid, cloning of DNA fragments into the
Hpal or Smal sites of pBR313 does not affect the expression of Tcr (data
not shown).

(g) Tetracycline-cycloserine enrichment for recombinant transformants

Since transformation of E. coli K12 with in vitro ligated recombinant DNA
may yield recombinant transformants at a frequency as low as 10-6 to 10'7/
ml/µg of DNA, a procedure was needed for enriching the number of re-
combinant transformants in the total cell population. Such a technique was
developed by taking advantage of the bacteriostatic nature of Tc and the
bactericidal effect of Ap and Cyc. The rationale behind this procedure is the
temporary inhibition of the growth of Tc$-recombinant transformants by the
addition of Tc to the growth medium. After allowing the Tcr-transformants
a 45 min interval of exposure to Tc, Cyc was added at a concentration which
promoted the exponential lysis of growing cells (Curtiss et al., 1965). The
Tc8-recombinant cells can be recovered after the removal of the Tc and Cyc.
Nontransformed cells can be eliminated from the culture by the addition of
Ap either before or after the Tc-Cyc lytic step. A mixed-culture reconstruc-
tion experiment was conducted to demonstrate the practicality of this
rationale. As shown in Table II, recombinant transformants containing
N. crassa DNA (pBR313-NCS8) initially present at a frequency of 1.10-6
were enriched to 3 . 10' by this procedure.
 89

TABLE I

MOLECULAR CLONING OF VARIOUS DNA FRAGMENTS IN pBR313

The molecular weights (Mr"10-6) under each restriction site represent DNA fragments of
independent clones isolated in this laboratory.

DNA sources Restriction endonuclease substrate site

EcoRI HindIIl BamHI SaII HindIIl-Baml-II

E. coli
EcoRI modification
methylase 1.0
Chloramphenicol
resistance a 3.65

D. melanogaster 3.1 5.8 2.0

4.0 4.82
0.5 4.5
1.2 2.6
0.86

N. crassa 2.5 1.05 3.5 3.2

0.8 1.5 3.05
0.5 8.0 1.8
0.3 4.0
3.1 0.5
4.2

Cauliflower 4.8
Mosaic virus 3.2
(Cabbage B strain)b 1.6

'Derived from an uncharacterized R-factor originally contained in a strain of S. typhimurium

(unpublished observations).
bMeagher et al.

An example of the utility of pBR313 as a cloning vector in conjunction

with this enrichment technique can be illustrated by the following results.
Drosophila melanogaster DNA and pBR313 were cleaved separately with
BamHI and SaII restriction endonucleases. After in vitro ligation, the DNA
was used to transform RR1. When the transformed culture reached satura-
tion, the percentages of Apr cells in cultures from the Sail and BamHI exper-
iments were 0.023% and 0.16% respectively. Samples of the transformed
culture were diluted 1/50 and after logarithmic growth was established, Ap
was added. The stationary phase cultures were found to be approximately
100% Apr. The enriched cultures were diluted 1/100 (one liter total) and
incubated for 1h before Tc was added. After 45 min incubation in the
presence of Tc, Cyc was added to the cultures and incubated for 2.5 h. The
90

ü
CL)
CL) 4ý
Co j2 Co
dMWV di

0rw
40.2 > "3 N. 0
y1 11

0 Mwy
A-5 ßi Oä CD CO CO
WH CD Cd
eD
-0 .NC
'0 OO w
-0 -3 Z- .2

Wq :...
ä`r yYou
MA °°+ý ci pU
Wr ei ;'°,, db3. OM
Z. 0I. ?
qoo J.- how
. _ v2

Ö= ioo
a
ce N"N
Cd mÄ
poe0g
- boo
wOa°p, -3 p0y
iý it
y o,
Z 42
ß+ 2 10 ",
ä CL y V] ü i.+ wÖ
Ow as
cd co

U oý pp ap+ 0
03
aä öw

P. Q'
wNi, Ici ea qöOM
aqä p
p0 v2 ° ÜNw m
Ex.,
ýmN
Z'
'C7 cýö

ýytöa o
;c
z ,da p^, a, dp
p o, rn
o cm ci r+ r+
ti ßi0ä-3--. ww.
%.

GO2w ý1Nqp CO MU
CO-
+' s.
r+
y
r, a
^y ý"ý
14
ßi p4 pý Z
Zd 4) ä
° 22

ci (Lc002
Wpýw 2 )12 A ib'>' FHF
(0
W Z, 'kL
dý üvyää
dFa V j. p, Addd
ra 41

z4d
EZq ýa 14 l
A4)
La üe ü M
CO CO
V PQ
UD m
9Ö äV
ÖÖ üo
ü
UUº. cow p "ý Eý
CA *.... .""4 .ý ."i
H ää HH 2. °-3 ä äz
rn. rn r4
 91

cultures were centrifuged and resuspended in 100 ml of growth medium

without antibiotics and incubated for 10 h. The percentages of cells that
were AprTcs (i. e., cells carrying recombinant plasmids) in the SaII and BamHI
experiments were enriched to 92% and 88% respectively.

DISCUSSION

By means of the EcoRI and EcoRI* reactions and the genetic translocation
of ampicillin resistance translocon (TnA), we have constructed a series of
bacterial plasmid cloning vectors with a ColEl replication mode. These
plasmids have been characterized with thirteen restriction enzymes. One of
these plasmids, pBR313, has some features which make it a more efficient
cloning vehicle than the currently used vectors, for example pSC101 (Cohen
et al., 1973), Co1E1 (Hershfield et al., 1974), pMB9, and phage lambda
(Cameron et al., 1975). The advantages of using pBR313 as a cloning vector
are summarized as follows. (1) The molecular cloning of EcoRI, HindIIl,
BamHI, Sall, HpaI and Smal can now be achieved in a single low molecular
weight, amplifiable plasmid. (2) The substrate sequences for HindIIl, BamHI
and SaII restriction endonuclease are located in Tc' region thus facilitating
the recovery of cells harboring recombinant DNA by virtue of their Apr-Tcs
phenotypes. Although recombinant plasmids generated by EcoRI, HpaI and
Smal do not inactivate the Tcr mechanism in pBR313, double-digested DNA
fragments involving any one of these enzymes and either HindIIl, Bam HI or
SaII will produce Tcs recombinants. We believe that the use of these six
restriction enzymes and the 14 possible combination digests will provide not
only the opportunity for cloning many interesting DNA fragments but also
the further dissection of these DNAs into their component parts. This feature
is of particular importance for the DNA sequencing technique recently
developed by Maxam and Gilbert (Maxam and Gilbert, 1977). (3) Recom-
binant transformants with Apr-Tcs phenotypes are amenable to enrichment
over non-recombinant and non-transformant cells by the use of the Ap-Tc-Cyc
lytic procedure. (4) As a result of the EcoRI* digestion of pBR312, the Apr
gene can no longer be translocated from pBR313 (S. Falkow, personal com-
munication). This eliminates the possibility of translocating cloned DNA
from the plasmid vector to either the chromosome or other resident episomes.
While cloning into the BamHI and SaII sites in pBR313 clearly inactivates
the Tcr mechanism, cloning in the EcoRI site does not. Many DNA fragments
inserted into the HindIIl site also inactivate the Tcr mechanism; however,
other pieces of DNA only reduce the level of Tc resistance. Preliminary data
suggest that the HindIIl site may be localized in a regulatory region, i. e., a
promoter for E. coli DNA polymerase. This notion is supported by the fact
that the HindIII site in pBR313 is protected from digestion in the presence
of RNA polymerase (Rodriguez et al., 1977).
At present, we have mapped more than forty restriction sites in pBR313
using thirteen restriction endonucleases. At least fourteen of these sites are
 92

located in the Tcr gene complex, and we are in the process of mapping addi-
tional substrate sites in this region. This mapping will enable us to sequence
the Tcr gene(s) according to the procedure of Maxam and Gilbert (Maxam
and Gilbert, 1977) which we hope will provide new insight into the Tcr
mechanism and its mode of control. Furthermore, our knowledge of these
restriction sites in pBR313 has enabled us to construct other cloning vectors
which permit the use of PstI and HincIl restriction enzymes for the molecular
cloning of DNA into the Apr gene (Fig. 2a) (Bolivar et al., 1977). Since HincII
is known to produce blunt-ended cleavages in DNA, a cloning vector with one
Hincll site either in the Ap or Tc genes would allow for the cloning of any
blunt-ended DNA fragment by ligation with phage T4 polynucleotide DNA
ligase under the appropriate conditions (Sgaramella et al., 1970; Heyneker et
al., 1976).

ACKNOWLEDGEMENTS

This work was supported by grants to H. W.B. from the National Science
Foundation and the National Institutes of Health. R. L. R. was supported by
"a postdoctoral fellowship from the A. P. Giannini Foundation for Medical
Research. We would also like to acknowledge Dr. Istvan Fodor and Alejandra
Covarrubias for their discussion and technical assistance. We are also grateful
to Patricia L. Clausen for her expert preparation of this manuscript.

REFERENCES

Betlach, M. C., Hershfield, V., Chow, L., Brown, W., Goodman, H. M. and Boyer, H. W.,
Fed. Proc. 35 (1976) 2037-2043.
Bolivar, F., Rodriguez, R. L., Greene, P., Betlach, M. C., Boyer, H. W., Crosa, J. and Falkow,
S., Gene, 2 (1977) 000-000.
Boyer, H. W. and Roulland-Dussoix, D., J. Mol. Biol., 41 (1969) 459-472.
Cameron, J. R., Panasenko, S.M., Lehman, R. R. and David, R., Proc. Natl. Acad. Sci. USA,
72 (1975) 3416-3420.
Clewell, D. B., J. Bacteriol., 110 (1972) 667-676.
Clewell, B. B. and Helinski, D., Proc. Natl. Acad. Sci. USA, 62 (1969) 1159-1166.
Cohen, S.N., Chang, A. C.Y. and Hsu, C.L., Proc. Natl. Acad. Sci. USA, 69 (1972) 2110-
2114.
Cohen, S.N., Chang, A. C.Y., Boyer, H. W. and Helling, R., Proc. Natl. Acad. Sci. USA, 70
(1973) 3240-3244.
Curtiss, R. III, Charamella, L. J., Berg, C.M. and Harris, P., J. Bacteriol., 91 (1965) 1238-
1251.
Danna, K. J., Sack, G.H. and Nathans, D., J. Mol. Biol., 78 (1973) 363-376.
Datta, N. and Richmond, M. D., Biochem. J., 98 (1966) 204.
Greene, P.J., Betlach, M. C., Goodman, H. M. and Boyer, H. W., in R. B. Wickner (Ed. ),
Methods in Molecular Biology Series; DNA Replication and Biosynthesis, vol. 7,
Marcel Dekker, New York, 1974, pp. 87-111.
Gromkova, R. and Goodgal, S.H., J. Bacteriol., 109 (1972) 987.
Guerry, P., LeBlanc, D. J. and Falkow, S., J. Bacteriol., 116 (1973) 1064-1066.
Hamer, D. and Thomas, C., Proc. Natl. Acad. Sci. USA, 78 (1976) 1537-1541.
Heffron, F., Rubens, C. and Falkow, S., Proc. Natl. Acad. Sci. USA, 72 (1975) 3623-3527.
 93

Heffron, R., Bedinger, P., Champoux, J.J. and Falkow, S., Proc. Natl. Acad. Sci. USA, 74
(1977) 702-706.
Hershfield, V., Boyer, H. W., Lovett, M., Yanofsky, C. and Helinski, D., Proc. Natl. Acad.
Sci. USA, 71 (1974) 3455-3461.
Hershfield, V., Boyer, H. W., Chow, L. and Helinski, D., J. Bacteriol., 126 (1976) 447-453.
Heyneker, H. L., Shine, J., Goodman, H. M., Boyer, H. W., Rosenberg, Dickerson, J., Narang,
S.A., Itakura, K., Lin, S. and Riggs, A. D., Nature, 263 (1976) 748-752.
Heyneker, H., Greene, P.J., Betlach, M. C., Bolivar, F., Rodriguez, R., Covarrubias, A.,
Fodor, I. and Boyer, H. W., (1977) Manuscript in preparation.
Landy, A., Ruedisueli, E., Robinson, L., Foeller, C. and Ross, W., Biochemistry, 13 (1974)
2134-2141.
Levy, S. and McMurray, L.,, Biochem. Biophys. Res. Commun., 56 (1974) 1060-1068.
Maxam, A. M. and Gilbert, W., Proc. Natl. Acad. Sci. USA, 74 (1977) 560-564.
Meagher, R. B., Tait, R. C., Betlach, M. and Boyer, H. W., Cell, 10 (1977) 521-536.
Nagel De Zwaig, R. and Luria, S., J. Bacteriol., 94 (1967) 1112-1119.
Padmanabhan, R. and Wu, R., J. Mol. Biol., 65 (1972) 447-464.
Panet, A., Van de Sande, J.H., Loewen, P.C., Khorana, H. G., Rase, A. J., Lillehaug, J. R.
and Kleppe, K., Biochemistry, 12 (1973) 5045-5050.
Polisky, B., Greene, P.J., Garfin, D. E., McCarthy, B. J., Goodman, H. M. and Boyer, H. W.,
Proc. Natl. Acad. Sci. USA, 72 (1975) 3310-3314.
Roberts, R. J., Breitmeyer, J. B., Tabachuik, N. F. and Myers, P.A., J. Mol. Biol., 91 (1974)
121-123.
Roberts, R. J., Myers, P.A., Morrison, A. and Murray, K., J. Mol. Biol., 102 (1976) 157-
165.
Rodriguez, R. L., Bolivar, F., Goodman, H. M., Boyer, H. W. and Betlach, M. C., in D. P.
Nierlich, W.J. Rutter and C.F. Fox (Eds. ), Molecular Mechanisms in the Control of
Gene Expression (ICN-UCLA Symposia on Molecular and Cellular Biology, Vol. V),
Academic Press, New York, 1976, pp. 471-477.
Rodriguez, R. L., Tait, R., Shine, J., Bolivar, F., Heyneker, H. L., Betlach, M. and Boyer,
H. W., Characterization of tetracycline and ampicillin resistant plasmid cloning vehicles,
in Molecular Cloning of Recombinant DNA, Vol. 13 (Ninth Miami Winter Symposia),
1977, in press.
Schwartz, S.A. and Helinski, D., J. Biol. Chem., 216 (1971) 6318-6327.
Sgaramella, V., Van de Sande, J. H. and Khorana, H. G., Proc. Natl. Acad. Sci. USA, 67
(1970) 1468-1475.
So, M., Gill, R. and Falkow, S., Mot. Gen. Genet., 142 (1976) 239-249.
Tait, R. C., Rodriguez, R. L. and Boyer, H. W., Mot. Gen. Genet., 151 (1977) 327-331.
Wilson, G.A. and Young, F. E., J. Mot. Biol., 97 (1975) 123-126.
Yoshimori, R., Roulland-Dussoix, D. and Boyer, H. W., J. Bacteriol., 112 (1972) 1275-
1283.

Communicated by D. R. Helinski.