Construction of a mobilizable cloning vector for site-directed mutagenesis of Gram-negative bacteria: application to Rhizobium leguminosarum

Gene, 98 (1991) 7-13

7

Elsevier

GENE

03909

Construction of a mobilizable cloning vector for site-directed mutagenesis of Gram-negative bacteria: application to Rhizobium leguminosarum (Recombinant

DNA;

insertion

inactivation;

Werner Kokotek and Wolfgang

mob site cassette;

suicide vector;

broad host range;

plasmid

stability)

Lotz

Institutftir Mikrobiologie und Biochemie, Lehrstuhl Mikrobiologie, UniversitiitErlangen-Niirnberg,8520 Erlangen (F.R. G.) Received by J.A. Hoch: 2 August Revised: 30 September 1990 Accepted: 1 October 1990

1990

SUMMARY

A mobilizable cloning vector was constructed from defined fragments to serve as a suicide plasmid for site-directed mutagenesis. The new vector, pKOK4, closely resembles plasmid pBR325. However, the inverted duplication existing in the latter was not introduced. The useful cloning sites of pBR325 (EcoRI, HindIII, EcoRV, BarnHI, MI, PstI and PvuI) were retained and are located in one of the three resistance markers, ApR, CmR or TcR, respectively. Also, in pKOK4 the CmR gene retains its own promoter. The mob site of plasmid RP4 was introduced as a 760-bp fragment at a defined location. The mobilization frequency of pKOK4 within Escherichia colt’ strains is approx. 4 x 10 2 per recipient cell. The size of pKOK4, deduced from the construction, is 6368 bp. We used pKOK4 for site-directed mutagenesis of hup-specific DNA from Rhizobium leguminosarum B 10. Integration of the vector could be distinguished reliably from marker exchange by screening for the antibiotic resistance(s) of the plasmid. This reduced the number of clones to be retested by colony and Southern hybridization to approx. 1% of the original number. Of these, almost 70 “/, contained the desired marker exchange.

INTRODUCTION

One of the methods used for examining genetic organization is site-directed mutagenesis (Ruvkun and Ausubel, 1981). It requires construction and mapping of mutations on plasmid-borne DNAs in E. coli, transfer of the mutated plasmids into the recipient, and screening for marker exchange. Unfortunately, when suicide vectors are used, i.e., vectors unable to replicate in the recipient strain, cointegrate formation occurs much more often than marker Correspondence Lehrstuhl

to: W. Kokotek,

Mikrobiologie,

strasse 5, W-8520 Erlangen Fax (49-09131)85-8082. Abbreviattons:

(F.R.G.)

und Biochemie,

Erlangen-NUmberg,

Staudt-

Tel. (49-9131)85-8089;

LB, Luria-Bertani (medium); MBN, mung bean nuclease: wrob, recognition site for transfer functions; nt, nucleotide(s); oriT. origin of DNA transfer;

oriF’, origin

of DNA

replication;

(large) fragment off?. coli DNA polymerase rifampicin; ‘, sensitivity/sensitive; Sm,

Ap. ampicillin;

EtdBr, ethidium bromide; a-D-thiogalactopyranoside;

Inst. fiir Mikrobiologie

Universitat

exchange (Ditta, 1986). Also, the ratio of marker exchange vs. cointegration varies to a great degree, depending on the species (deBruijn, 1987). Therefore, a fast, simple and reliable screening procedure is desirable. Screening of the recipients for loss of the plasmid-encoded resistance (Ditta, 1986) was not sufficient since sometimes, although the plasmid (a pACYC184 derivative) had integrated, no resistance was seen (unpublished observations). We ascribe this effect to recombination within the integrated plasmid. Therefore, a time-consuming examination of the desired marker ex-

bp, base pair(s);

Cm, chloramphenicol;

hup,hydrogen-uptake gene; IPTG, isopropylkb, kilobase or 1000 bp; Km, kanamycin;

NaCl/0.015 tryptone-yeast

M Na,

p, plasmid;

‘citrate pH 7.6; r, terminator;

extract;

Pollk,

Klenow

I: n, resistance/resistant; Rif. streptomycin: SSC, 0.15 M Tc, tetracycline:

TY,

u, unit(s); wt. wild type; XGal, 5-bromo-4-chloro-

3.indolyl-fl-D-galactopyranoside;

[ 1. denotes plasmid-carrier

state.

/I-

-

cut

wth

Smol

EcoRI.

396

sph i. 441 Il~n dill, 447

Sea I,

“‘..,,I, 1174 a1, 1180 1186 H:n dIi1. 1204

~solote 1223 bpSou96AI frogmen,, recut w,th Bad. ~soiate Y07-bp frugment contoinlng CmR, flil I” ends with P0llk

Ifgate, transform I co/, 517-l select for Apk+CmE determine orientation

EcoRI,

4359

H~ndlll.

EcoRI,

397

of

29

cut

wth

Kpnl

+

Soil,

isolatemob/K& fragment, make ends with MEN

PvulI,

2064 cut

I

wth

make

blunt

Ifgate.

transform Cdl 517-1: select for TcR+CmR. determini orlentotlon of fragment

f

EcoRl;

ends

blunt

with

MBN

AvaI.

148

l f

cut

wth

So/l;

fragment make

ends

isolate

containing blunt

329Gbp TnlO

with

with Xbol: ttli in ends wth PolIk. dephosphor$ate cut

MBN

ligate; transform E. co/i HBlOl; screen for insertion of frogment; determine orientation

AvaI.

148

>

Fig. 1. Steps in the construction ofpKOK4. (a) Methods. Methods not mentioned separately followed standard protocols (Maniatis et al., 1982). Plasmids used for cloning were purified by CsCI-EtdBr gradients (Maniatis et al., 1982) or by QiagenTM -pack ion exchangers (Diagen GmbH, Diisseldorf, F.R.G.). Small-scale plasmid preparations were done according to Holmes and Quigley (1981). DNA fragments were reisolated by electroelution into dialysis

9 change

by colony

or Southern

hybridization

(deBruijn,

RESULTS

AND DISCUSSION

1987) was necessary. To make screening more reliable, the plasmid described here, pKOK4, was designed. It contains no noticeable repeated sequences, which could promote loss of the resistance markers after integration (besides causing plas-

(a) Construction of pKOK4 The mob site from pSUP202 (Simon et al., 1983) was cloned into pUC19 (Yanisch-Perron et al., 1983; Chambers et al., 1988) (For details of the construction, see Fig. 1).

mid instabilities in the first place). Also, two out of three resistance markers (ApR, CmR, TcR) can always be retained

Next to the n?oh site in the resulting

plasmid,

pKOK I, the

CmR gene from pBR325 (Bolivar, 1978) was inserted, yielding plasmid pKOK2. The CmR/mob cassette was recloned from pKOK2 into pBR322 (Bolivar et al., 1977; Sutcliffe, 1979; Watson, 1988), resuiting in plasmid pKOK3. Down-

and screened for, separately. This should reduce the number of false-positive candidates isolated. The mob site of plasmid RP4 was introduced to render the plasmid mobilizable.

stream

For the construction, sequence data available for all fragments except the 760-bp mob site, were used. Here, only a

pWH961 (Schollmeier plasmid pKOK4.

from

the

core region of 112 bp had been sequenced (Guiney and Yakobson, 1983). The size of the resulting vector (6368 bp) and the location of restriction sites can therefore be deduced exactly.

(b) mobilization properties of pKOK4 pKOK4 was transformed into E. coli Sl7-1 and conjugated to a RifR derivative of E. co/i HB 101. The transfer

of Hammond

et al., 1982) and cleaned

and D’Aiessio (19863. All other enzymes

of King and Blakesley BRL, genotype

and Yakobson,

RP4 (Simon

to that described

(Yanisch-Perron

et al.,

Ap + XGal/IPTG-plates.

on it. The 907~bp fragment

Chambers

map for a Hue11 fragment

was determined

1988).

was isolated

contains

blunt using 75 u MBN/ng

ofsites

required

the fragment

resulting

plasmids

were unstable

upon amplification.

with insert size was observed

promoters

(Brosius,

1984; Gentz

and

Messing.

plasmid

was named

recombination

direction

plasmid,

of the mob sites of RP4

1982)

were analyzed

pKOK2,

et al., 1981; Lambert

were observed and Reznikoff,

contains

were

was isolated

50 u MBN/ng

the MBN-treated of the terminator (Schollmeier

from plasmid

DNA. Plasmid

pKOK3

pWH961

(Schollmeier

was cleaved

et al., 1985). The resulting

piasmid,

of E. co/i S 17-l (Simon

the mob site showed

further

severely retarded

by verifying the single occurrence

a CmR/moh

cassette.

by KpnI + Sal1 double digestion

(d) Construction of an and its ends were made

of E. cofi 517-l were selected oriented

1979; Watson,

on LB plates containing

as shown was named pKOK3.

DNA from R. leguminosarum BIO (Tichy et al., 1987), some of the

population

showed

1985; Stueber

deletions

DNA. Plasmid and Bujard,

extending

destabilization

into both insert

and vector.

has been correlated

1982). The CmK gene fragment

et al., 1985) as a 329-bp Sal1 fragment.

The ends of this fragment

with XbaI, the ends tilled in with PolIk, dephosphorylated,

pKOK4,

has a deduced

contained

at its 3’ end. The 14 bp from

was ligated to pBR322 (Bolivar et al., 1977; Sutcliffe,

in non-amplified

and Quiagen-cleaned terminator fragment. Transformants was determined with ClnI. One of the clones was selected

on

(c) Construction of a

part of pBR322

cleaned

No

with strong

did not contain

terminator, and the promoters of the TcR and CmK gene act in the same direction. Therefore, the instability was ascribed to transcription insert DNA andyor the oriV of the plasmids (Brosius, 1984; Gentz et al., 1981; Stueber and Bujard, 1982). Accordingly, a bidirect~onail~ terminator

selected

in E. c&i (Watt et ai., 1985). The fragment

plasmid having the CmR/mob cassette

of @-specific

A part of the plasmid

and no deletions

the mob (Guiney

with Hpall. The fragment

pKOK1.

the duplicated

of the CmR gene towards

as a 16X9-bp fragment

The resulting

for subcloning

contains

X&I, Sac1 and KpnI was verified.

at its 5’ end and two stop codons

for signiftcant

The CmR/mob fragment

enzyme mapping.

(e) Introduction of a terminator. When using pKOK3 correlation

(Vieira

for Ban?HI,EcoRI,

The resulting

in the opposite

from pKOK2

DNA for the mixture offragments. by restriction

the identity

and recleaved

and recut with BanI, to separate

1988). which had been cleaved by EeoRI and made blunt by 50 u MBN~~lg DNA. Transformants Tc + Cm and were confirmed

pSUP202

1,after filling in the ends ofboth fragment and vector with PolIk. Transformants

was isolated

conditions

cells of DH5rTM (Gibco

the mob site of RK2, was described

was isolated

of E. cc& JM83

digestion.

from pBR325

KpnI, XbaI and PstI. The resulting

HindHI,

the buffer and incubation

fi-agment were made blunt by MBN (0.5 u/pg DNA) and ligated to Snzni-restricted

on LB plates with Ap and Cm. Clones allowing transcription

SalI, BarnHI,

from yielding

of E. coii were selected on LB plates (Miller,

et al., 1979). To confirm

This fragment

the CmK gene with its own promoter

two clones carrying

terminator

to the recommendations

1985). Competent

(b) Cloning of the mob site. Plasmid

(Burkardt

Transformants

at the 5’ end, lie well below the size of homology

‘intermediate’ vector. The CmR/moh cassette

ligations

(Jessee,

of 760 bp, containing

760-bp fragment.

+ AvaI double

by BumHI

fragment

thus obtained

on selective plates. Therefore,

of sites for EcoRI,

et al.,

M, was used according

(1985). Transformants

respectively.

RP4 and RK2 has been proposed

and ligated to ~U~~HI-c~ea~~ed pKOK

et al., 1983) were selected growth

by Hanahan

DNA from one white clone was isolated and the single occurrence

of the fragment

still retained

For blunt-end

for RK2. The ends ofthe 7%bp

1985;

CmR/mob cassette. A 1223.bp Sau96AI pBR322,

from various suppliers.

as described

was digested with NueII, verifying the expected

was identical

The orientation

between

FPLCpure”

of 100, 30 and 20 &ml,

a Tnlfl

-.-

tips. MBN (Pharmacia

were diluted fivefold before transformation

et al., 1983). A restriction

1983) and homology

and RK2, pSUP202

was isolated

were obtained

mixtures

Ap, Tc, or Cm at concentrations

site from plasmid

pattern

(1986) were used. Ligation

given in Focus 10 (1988) 11) were prepared

1972) containing

pUCl9

by Qiagen rM ton-exchange

gene

et al., 1985) was inserted,

-.

-.. tubing (Maniatis

CmR

a

through the active Tnlff

were made blunt with

on a Quiagen

tip, and ligated to

of E. coli HBlOl were selected on plates containing Cm. The orientation where termination of transcription from the Cmu gene was most efficient size of 6368 bp. The levels of CmR were tested

for HBlOl[pKOK3]

and

HB 10 1fpKOK4j by comparing the diameter of single colonies. Growth was indistinguishable on plates containing 20 I_cgCm/ml. On plates with 150 ng Cm/ml, growth of HBlOl~pKOK4] was like on plates with 20 b&gCm/ml, while growth of HBIOI [pKOK3] was severely retarded under these conditions. Except for plasmids

pSUP202

and pWH961,

all restriction

sites, genes and other markers

are drawn

to scale within each plasmid

map. The position

‘0’ in plasmids pKOK3 and pKOK4 has been arbitrarily assigned to the first nt of the former EcoRI site of pBR322. The restriction sites for AvaI and Hinfl only indicate the relative orientation of the mob site; this does not imply that these sites arc unique. All other sites indicated are unique; positions given correspond

to the first nt of the recognition

sequence.

I0

frequency

of pKOK4

was 3.9 x IO-‘+

recipient cell (mean average of triplicate ent titer determined after incubation). (c) Use of pKOK4 for site-directed

0.7 x 1OV’ per experiment.

E. coli by the resulting

plasmids therefore were ApKCmK and ApKTcR, respectively. They were randomly mutagenized by linearization with Suu3AI and ligation to a lucZ-KmR cassette (Kokotek and Lotz, 1989) with BrrrnHI

recipi-

ends. Transformants ofE. co/i DHSX-“~ were selected with Km + Ap + Cm or Km + Ap + Tc. The location of the cassette in several hundred clones was mapped, and 51 selected plasmids were conjugated from E. coli to an Sm”

mutagenesis

Three different fragments of hup DNA were cloned in pKOK4, two (1 kb and 8 kb) into the Hind111 site and one (5 kb) into the EcoRI TABLE

site. The phenotypes

mediated

in

1

Comparison

of antibiotic

screening

Number Tc’

vs. colony

of colonies

Number colony

and Southern

screened

tested

hybridizations”

for TcK: 1702”

by”

Number

of marker

according

hybridization

Number exchanges’

Cm’

of colonies

Number

to hybridizations

screened

tested

for CmK: 3962 h

by”

L

Number of marker exchanges

colony hybridization

according

38(set

34 (89.50”,,

to hybridizations

TypeA Number

showing

wt resistance‘

X(sct as 100” 0)

10

41

7 (X7.50”,, of tested,

as IOO”~,)

0.41 “,> of total)

of tested,

0.86”,, of total)

Type B Number showing increased

1 (50.00”,,

2(set as loo”,,)

resistance‘

12

Total

of tested.

28

19(set as loo”,)

2 (lOSO”,

0.06”,, of total)

2

8 (X0.00”,, of tested,

lO(set as 100”“)

69

57(set

as loo”,,)

36 (63.20”,, of tested, 0.9 I O,)of total)

0.47”,, of total) ,I Filter-matings and Helinski (Beringer,

(Guiney

and Helinski,

(1979) was modified

1974) before spotting

’ All clones obtained

1979) employed

by collecting

” For colony

and 28 type B growth. hybridizations,

were replica-plated

the method

was essential.

cultures

strain

S 17-l (Simon

by centrifugation

onto medium

of Zeph and Stotzki

et al., 1983) as a donor.

and resuspending

The procedure

of Figurski

them in a small volume of TY medium

ofthe filters (by rubbing) were doubled.

Proteinase

(sodium

Tc at 6 pgiml, or Cm at 35 pg/ml, depending

growth

showed

(1989) was modified

(type B). For Cm ‘, 3962 clones were tested.

reduced

antibiotic

as follows:

K concentration

salt) to dissolve

soaked

in NaOH,

was reduced

the slime produced

instead

on the plasmid 41 clones showed

sensitivity.

sterile nylon filters (Pall Biodyne

pressed onto the surface with a sterile Drygalski

The filters were placed on 3MM paper,

with 2”” (w/v) N-Lauroylsarcosine

containing

(type A) and two retarded

In total, of 5664 clones only 81 (1.43;)

were placed on TY plates cooled to 4” C, moderately of air bubbles

of washed

them onto the filter.

from the mutagenesis

used. ’ Of 1702 clones tested for Tc”, ten showed no growth type A growth

the E. co/i mobilizator

mixtures

of tested,

0.04”,, of total)

spatula

and incubated

of immersing

them. All incubation

to 0.25 mgjml. The buffer used in proteinase by the Rhizobium colonies.

Incubation

A, 1.2 pm pore size)

for approx.

15 min. Avoidance

times before washing

treatment

was supplemented

in this buffer was performed

for

90-100 min at 37’C with occasional shaking. Probes were labeled with Biotin-7-dATP or Biotin-11-dATP, using a nick translation kit (BRL). Hybridization was performed overnight at 42’ C in 45 ‘,, (w/v) formamide. 5 x SSC, 1 x Denhardt’s solution, 0.2 mg/mt calf thymus DNA, OS”,, SDS. Probe concentrations The stringency were hybridized

were 100-200

against

pKOK4.

of type B were hybridized. pKOK4.

ng/‘mt. Filter washing

and detection

of probes

followed

the instructions

of the BtuGENE”’

detection

kit (BRL).

wash was done for 2 x 15 min in 0.16 x SSC/O. 1 “c, SDS at 64’C. Of the clones tested for Tc’, eight colonies of type A and two of type B Seven clones of type A and one clone of type B showed

Only two of the 19 type B colonies

Of the 67 clones (1.2”,, oftotal)

gave no signal, whereas

tested by colony hybridization,

no signal. Of the clones tested for Cm”, 38 of type A and 1Y

35 of the 38 type A colonies

45 (67%) showed no hybridization

did not hybridize,

to the vector, therefore

indicating

loss of

being likely to contain

the desired double-crossover. Of these again, 42 (93?,>) were of type A and three (7%) of type B. For all 67 clones tested for presence of pKOK4 an additional colony hybridization against the KmK gene from pUC4K was performed. All but one clone, probably a spontaneous KmK-mutant. hybridized. verifying

the presence

of the larZiKmR

cassette.

’ For hybridizations according to Southern (1975). labeling was done as given for colony hybridizations (see footnote d). Hybridization conditions. washing offitters and detection ofprobes followed the instructions ofthe PhotoGenelM detection kit (BRL). The stringency wash was done for 2 x 15 min. in 0.1 X SSC/O.l”,, SDS at 58°C. From those 66 clones indicating marker exchange in colony hybridizations, total DNA was isolated, restricted with So/I, and Southern hybridizations were performed using the respective fragments originally inserted into pKOK4 as probes. The obtained results were consistent desired

Lvith those for the colony hybridizations, marker

exchange

in the Southern

i.e.. clones

hybridizations

showing

no signal in colony hybridizations

and vice versa. An additional

Southern

with pKOK4

hybridization

as a probe demonstrated

experiment,

the

using the KmK gene from

pUC4K as a probe, demonstrated the presence of the Km”:lucZ cassette in all clones examined. Ofthe 44 marker exchanges obtained. five were derived from primary inaertlons within the I-kb hup fragment. eight from insertions within the 5-kb hup fragment and 31 from insertions within the 8-kb hup fragment.

11 derivative

of R. legu~i~o~a~~

I310 (present

work).

Re-

cipients were selected on TY medium containing Sm (200 pg/ml) and Km (50 pg/ml). E. coli donors were additionally counterselected by phage T4D (Fellay et al., 1989). From every conjugation at least 100 clones were semi-purified by picking them onto new selective plates. The reliability of pKOK4 for the screening by antibiotic

mid then remains

within the recipient.

In comparison,

the

use of suicide vectors is less time-consuming and introduces no additional sequences except the marker used for mutagenesis. However, it requires a suicide vector which is structurally stable and whose integration can be reliably detected. To retain the advantage of shorter time required for mutagenesis,

screening

for integration

e.g., by screening the vector.

(fragment cloned into the EcoRI site) was compared to the wt recipient on solid medium. Differentiation with Cm was rather difficult, as the wt strain already showed a low level of CmR. However, the use of 35 yg Cm/ml yielded sufficient difference in growth after two days of incubation at 28°C. As growth of colonies varied even within the same plate, probably due to unequal inoculation, two different types of growth inhibition were noted: no growth (type A) and retarded growth (type B). For Tc, the wt showed no resistance, and 6 pg/ml were appropriate for differentiation of growth. With Ap, no difference in growth could be found at all. To test if this was due to loss of the ApR gene or to lack of its expression, the 44 clones containing pKOK4 were retested by colony hybridization. The probe used was an 807-bp Hinff-SspI fragment from pBR322, spanning most of the ApR gene. All clones hybridized, showing that the ApR gene was present but not expressed in our strain of R. legu~~~~osaru~z. Recently, Leyva et al. (1990) also observed only marginal expression of the Tn3-derived ApR gene in this strain. Therefore, we could not screen for integration by using two different independent antibiotic resistances as intended. However, the confirmed structural stability of pKOK4 made screening by use of only one resistance marker seem reasonable. The results obtained from this screening are summarized in Table I. Antibiotic screening reduced the original number of clones (5664) to 8 1(1.4”4) candidates for marker exchange. Of these 8 1 clones, 67 (l.2Y,, of the original number) were analyzed by hybridizations. Here again, 45 (679;) of the 67 clones showed the desired marker exchange. These results confirmed that the employment of pKOK4

The vector resembling pKOK4 most closely, pBR325 (Bolivar, 1978) contains an inverted duplication (Prentki et al., 1981) which can lead to instability. The same holds true for mobilizable vectors based on this plasmid. Other mobilizable vectors either are considerably larger than pKOK4, or contain fewer useful cloning sites, or fewer sites where cloning can be nlonitored by insertion inactivation. In pBR329 (Covarrubias and Bolivar, 1982) a CmR gene was inserted into pBR322 without duplicating vector parts. However, in pBR329, the CmR gene is not transcribed from its own promoter. Thus, cloning into the Hind111 site of pBR329 causes a TcS and a CmS phenotype simultaneously in 15-200, of the cases (Covarrubias and Bolivar, 1982). To discriminate integration from marker exchange reliably, it is advantageous to retain two independent resistance markers instead of one, irrespective of the cloning site. No other derivatives of the well-characterized plasmid pBR322, containiIlg three resistance genes and suitable for introducing a moth site, have come to our knowledge (Balbas et al., 1986).

for mutagenesis efficiently reduced to be tested by hybridization.

the number

of colonies

(d) Conclusions (I) Reasons for the construction of a new vector When using broad-host-range plasmids for site-directed mutagenesis, a second incompatible plasmid has to be introduced by an additional conjugation (Ditta, 1986). This plas-

for the antibiotic

also has to be fast,

resistance was to be analyzed. Therefore, a colony hybridization of 44 randomly picked Rhizobium recipients was performed using pKOK4 as a probe. All clones contained the vector. The sensitivity of eight of these clones for Ap and Cm (fragment cloned into the ~~~dlI1 site) or Ap and Tc

resistance

mediated

by

While several other mobilizable suicide vectors exist, we believe that pKOK4 combines most of the properties desirable for a cloning vector for site-directed nlutagenesis.

Only well characterized fragments were used and every cloning site has been thoroughly tested after each construction step. It should be noted that no tailing, BAL 31, or partial digestions were used. To avoid (re)generating restriction enzyme sites by tilling in certain fragment ends, such ends were removed by nuclease digestion. Here, MBN was preferred to S 1 because of the greatly reduced endonuclease activity of MBN (Laskowski, 1980). We also found the supplier of the enzyme to be of critical importance. As even the action of MBN is difficult to control, the concentrations of MBN were calculated individually for each set of ends (Hammond and D’Alessio, 1985). Every MBN-treated fragment was checked for the absence of visible degradation before ligation. Further, in most cases the fragment was reisolated after insertion via restriction sites only some bp away from it and compared directly to the untreated fragment. If sites farther away had to be used for reisolation, several independent isolates were tested with a variety of enzymes and showed restriction patterns indistinguishable

12 among each other, which were as predicted struction.

from the con-

REFERENCES

Therefore, we feel sure that no changes have gone undetected which would seriously modify the mentioned prop-

Balbas, P., Soberon, X., Merino, E., Zurita, M., Lomeli, H., Valle, I,.. Flares, N. and Bolivar, F.: Plasmid vector pBR322 and its spccial-

erties of pKOK4.

Bcringer,

purpose

derivatives

Microbial.

(3) Mobilizntiorl properties pKOK4 can be mobilized by use of the strain S 17-l (Simon et al., 1983). The transfer frequencies of pKOK4 tested in E. co/i are comparable to those of other plasmids with the same mob site. While the actual oriT only comprises 112 bp in the 760-bp t~zoh fragment, the adjacent sequences increase the frequency of transfer about lOO-fold (Guiney and Yakobson, 1983). Therefore, no effort was made to reduce the size of this fragment.

Bolivar,

F.: Construction

selection

and characterization

of plasmid

pBR322

of EcoRI generated

Bohvar,

of new cloning vehicles,

carrying

recombinant

F., Rodriguez,

unique EcoRI sites for DNA molecules.

R.L., Green, P.J., Bctlach,

Boyer, H.W.. Crosa. J.H. and Falkow, terization

Gene 4

M.C., Hcyneker,

S.: Construction

ofnew cloning vehicles, II. A multipurpose

Gene 2 (1977) 95-113. Boycr. H.W. and Roulland-Dussoix, the restriction

clonmg system.

D.: A complemcntation

and modification

H.L.,

and charac-

analysis

of

of DNA in Erc,herichifr co/i. .I. Mol.

Biol. 41 (1969) 459-472. J.: Plasmid

vectors

for the selection

of promoters.

Gene

27

(1984) 151-160. Burkardt,

H.J., Riess. G. and Ptihler, A.: Relationship

mids rcvcaled

by heteroduplex

are tdcntical.

J. Gen. Microbial.

Chambers.

experiments:

R.A., Sherwood.

N.P.: Plasmid pMTL.153: a high copy number its use to obtain pcptidasc

high

expression

L. and Bolivar.

new cloning pBR328

vehicles.

lacking

R.F. and Minton,

verston of PAT‘153 and

of the P.seu&m>ntr\

G2 gene. Appl. Microbial.

Covarrubias,

of group PI plas-

RPI, RP4, R6X and RK2

I I4 (1979) 341-348.

S.P.. Prior, S.E., Evans.

Biotechnol.

F.: Construction

VI. Plasmid

and characterization

pBR329.

the 482.base-pair

carboxy-

29 (lY88) 575-57X. of

a new dertvative

inverted

duplication

of

Gene

17

(IYX2) 7Y-8’). dcBruijn.

F.J.: Transposon

Enzymol.

Tn.F mutagenesis

154 (1987)

Enzymol.

to map genes.

Methods

175-196. of Rhi:ohium

Ditta, G.: Tn.5 mapping

nitrogen

fixatton

genes. Methods

I18 (1986) 519-528.

Fellay. R., Krisch, H.M.. Prcntki, P. and Frey, J.: Omegon-Km: a transposable clement designed for m vivo tnsertional mutagenesis and cloning ofgenes Figurski.

derivative in trans. Gcntz,

in Gram-negative

D.H. and Helinski, ofplasmid

Cloning

RK2 dependent

downstream

A.C.Y.. Cohen,

of strong

placement

of an origin-containing

on a plasmid function provided

Sci. USA 76 (1979)

A., Chang,

and analysis

Gene 76 ( IYXY)215-226.

bacteria.

D.R.: Replication

Proc. Natl. Acad.

R.. Langner,

promoters

of a RNA

164X-1652.

S.N. and Bujard,

is made

tcrmmation

of the plasmid the proposed

RK2: location origin oftransfer.

D.G. and Yakobson,

of the site-specific

origin of the broad host range plasmid

mung bean nucleasc 4-6. Hanahan,

D.: Techniques

(Ed.).

DNA

Oxford. Holmes,

using positive selection

Cloning,

for transformation A Practical

Natl.

complex

I76 ( 1979) 183-189.

and nucleotide

Acad. Sci. USA 80 (1983) 3595-3598. Hammond, A.W. and D’Alessio, J.M.: Removal

Proc.

rclaxatton

H.:

by the

nick in the region of

Mol. Gen. Genet.

E.: Location

possible

signal.

Acad. Sci US.4 78 (19X1) 4936-4Y30. Guincy, D.G. and Helinskt, D.R.: The DN A-protem

the transfer

We thank Drs. R. Simon and U.B. Priefer for providing the E. coli strain S17-1 and several pSUP plasmids, Dr. R. Allmannsberger for providing the E. roli strain DHSc( and Ms. D. Gartner and Prof. W. Hillen, for providing the plasmid pWH961. We thank W. Kokotek for performing the experiments and preparing the manuscript. Our special thanks go to Mrs. C. Sizemore for intensive discussions and extensive help with the English. This vvork was supported by grant No. DFG-Lo 149/10-l from the Deutsche Forschungsgemeinschaft.

in Rhizohiun~ /~gun?inorrrr~l,f,. J. Gen.

(lY78) 121-136.

Guiney, ACKNOWLEDGEMENTS

transfer

84 (1974) 188-198.

III. Dcrivativcs

Brosius,

(4) Sumrnur~~ qff’yroperties qf pKOK4 The size of pKOK4, 6368 bp, and the location of restriction sites can be predicted accurately from the construction. This facilitated the mapping of fragments inserted into this plasmid. Cloning into one of the most commonly used restriction sites can be monitored by insertion inactivation of one of three resistance genes, respectively. The terminator introduced increases the level of CmR mediated by pKOK4. The mob site allows for mobilization of pKOK4 into a wide variety of Gram bacteria. The relatively large size of pKOK4 was no drawback in our work, especially in the insertion mutagenesis of cloned fragments: by (additional) selection for the two resistance genes maintained in each of the derived plasmids, insertions within these genes could be counterselected, reducing the number of insertions in the vector part. Although the .4pK gene is not expressed in R. leguminosurum, screening for integration of pKOK4 by antibiotic resistance was efficient even using only one resistance marker. This confirmed the planned structural stability of pKOK4. In species where the ApR is expressed, the efficiency of this screening can be increased further, as candidates showing no clear response with Tc or Cm alone can be probed additionally by their Apn.

- a review. Gene 50 (19X6) 3-40.

J.E.: R factor

sequence

of

RK2. Proc. Natl.

of 5’ cxtrnsions

plasmids.

with

Focus X (lY86)

ofE. co/i. In: Clover, D.M.

Approach,

Vol. 1. IRL

Press.

1985, pp. 109-135.

D.S. and Quigley,

M.: A rapid

boiling method

for the prepa-

ration of bacterial plasmids. Anal. Biochem. 114 (1981)193-197. Jesscc. J.: Use of competent cells for DNA transformation. Focus (1985) 5-6. King, P.V. and Blakesley, R.W.: Optimizing mation. Focus 8 (1986) l-3. Kokotck.

W’. and Lotz. W: Construction

DNA hgations

for transfor-

of a /ircZ-kanamScin-rcsist~lnce

6

13

cassette, probe.

useful

Lambert,

P.F. and Reznikoff,

stabilize

plasmid

Bacterial. Laskowski

mutagenesis

and

W.S.: Use of transcriptional

copy number

Sr., M.: Purification

Maniatis,

Enzymol.

T., Fritsch,

Laboratory Harbor,

of transcriptional

repressors

to

fusion vectors.

J.

and properties

Manual.

of the mung bean nucle-

Sambrook,

Cold Spring

J.: Molecular

Harbor

Laboratory,

Cloning.

A

Cold Spring

NY, 1982.

A., Palacios,

J.M., Murillo,

Miller, J.H.: Experiments

J. and Ruiz-Argtieso,

Laboratory,

in Molecular

Cold Spring

P., Karch,

Harbor,

Genetics.

T.: Genetic

from Rhizobium Harbor

NY, 1972.

a 482 base-pair-long

inverted

cloning vector

duplication,

B.C. and Ausubel,

signal for termination orfl on transposon R., Priefer,

system

Gene 14

F.M.: A general

method

for site-directed

genes. EMBO J. 1 (1982) 1399-1404. nucleotide

Cold

Spring

of the Ewherichia

sequence

Harbor

Symp.

Quant.

can

ofplasmid co/i plas-

Biol. 43 (1979)

77-90. W.:

of hup DNA

Analysis

leguminosurum

BlO.

Genetics

the Third

International

Plant-Microbe

Dordrecht,

versal primers.

host

D.P.S.

Symposium

Associations,

L.M.. Fees, H. and Lotz,

http

of Plant-Microbe

of Rhizobium

range

and

Brisson,

Interactions.

on the Molecular

Montreal,

Quebec,

N. (Eds.),

Proceedings Genetics

Canada.

of of

Martinus

1987, pp. 279-281.

Vieira, J. and Massing,

Watson,

and

In: Verma,

Molecular

J.: The pUC plasmids, mutagenesis

Gene

an M13mp7-derived

and sequencing

sys-

with synthetic

uni-

19 (1982) 259-268.

N.: A new revision

of transcription TnJO. Nucleic

is located

active fetA and

Acids Res. 13 (1985) 4227-4237.

U. and Ptlhler, A.: A broad

for in vivo genetic

between

engineering:

by gel electrophoresis.

Watt,

Ingles,

of the sequence

of plasmid

pBR322.

Gene

mutagenesis

J. Mol. Biol. 98 (1975) 503-517.

M13mp18 in

C.J., Urdea,

M.S. and

Rutter,

W.J.:

Homology

in Escherichiu co/i. Proc. Natl. Acad.

for recombination

Sci. USA 82 (1985) 4768-4772. Yanisch-Perron, cloning

host range mobilization

transposon

V.M.,

requirements

Gram-negative bacteria. Bio/Technology I (1983) 784-791. Southern, E.M.: Detection of specific sequences among DNA fragments separated

and diminish expression

70 (1988) 399-403.

mutagenesis in prokaryotes. Nature 289 (1981) 85-88. Schollmeier, K., Gartner, D. and Hillen, W.: A bidirectionally

Simon,

from efficient promoters

specified

tem for insertion

(1981) 289-299. Ruvkun,

replication

Sutcliffe, J.G.: Complete

Nijhoff,

Cold Spring

F., Iida, S. and Meyer, J.: The plasmid

pBR325 contains

H.: Transcription

with plasmid

Tichy, H.V., Schild, C., Ripke, H.M., Nelson,

65 (1980) 263-276.

E.F. and

D. and Bujard,

interfere

mid pBR322.

organization of the hydrogen uptake (hup) cluster legumirmsarum. J. Bacterial. 172 (1990) 1647-1655.

Prentki,

Stiiber,

as a promoter

162 (1985) 441-444.

ase. Methods

Leyva,

for site-directed

Gene 84 (1989) 467-471.

C., Vieira,

vectors

and

and pUC19

Zeph, L.R. and Stotzki, bacteria Microbial.

transduced

J. and Messing,

host

strains:

vectors.

J.: Improved

nucleotide

Gene 33 (1985)

G.: Use of a biotinylated by bacteriophage

55 (1989) 661-665.

Pl

Ml3

sequences

phage of the

103-119.

DNA probe to detect in soil. Appl.

Environ,