Oene, 86 (1990) 99-102 Elsevier

99

GENE 03381

Improved plasmid shuttle vectors for Haemophilus i n f l K n z a e and Escherichia coli (Recombinant DNA; ColEI plasmid; replication; plasmid RSF0885 rep locus)

Vuong N. Tfieu and David McCarthy Department of Botany and Microbiology, Universityof Oklahoma, Norman, OK 73019-0245 (U.S.A.) Received by R.E. Yasbin: ! I May 1989 Revised: 20 September 1989 Accepted: 4 October 1989

SUMMARY

In vitro deletion and transposon mutagenesis experiments were performed to localize the region essential for plasmid RSF0885 replication to a 1.7-kb sequence downstream from the p-lactamase gene. This locus was named rep, for replication. Plasmid RSF0885 can replicate in both Haemophilus influenzae and Escherichia coll. Replication in E. coli depended on transcription run-off from the p-lactamase promoter into the rep locus. Insufficient transcription into the rep locus could account for the instability of this plasmid in an E. colibackground, a property which reduced its usefulness as a shuttle vector. Therefore, four improved shuttle vectors for H. influenzae and E. coil were constructed. They possess the ColEl replication origin for maintenance in E. coil and the plasmid RSF0885 rep locus for maintenance in H. influenzae. Together, they provide twelve unique restriction sites for cloning by insertional inactivation of drug-resistance genes,

INTRODUCTION

The commonly used ColE I replicon for £scherlchia coil does not replicate in Haemophllus influenzae. However, a family of related p-lactamase encoding, multicopy plasmids from Haemophgus sp. and Neisserla sp. could replicate in both E. coli and H. influenzae. One member of this family, RSF0885 (De Graaff et al., 1976), was used to construct a shuttle vector called pDM2 (McCarthy et al., 1982). pDM2 has been used to isolate H. influenzae genes by complementation ofmutations in E. coil(McCarthy et al., 1984; Setlow Correspondence to: V.N. Trieu, Ph.D., Lipoprotein/Atherosclerosis Research Program, Oklahoma Medical Research Foundation, 825 N.E. 13th Street, Oklahoma City, OK 73104 (U.S.A.) Tel. (405)271-7448; Fax (405) 27 !-3980.

et al., 1988) and to mutagenize H. influenzae with the miniTnlOKm (McCarthy, 1989). There are some limitations to the use of pDM2. First, in our hands, pDM2 is extremely unstable in E. coli making it difficult to recover the plasmid from this background. Second, many H. influenzae chromosomal fragments that are cloned into the Apa gene of pDM2 are tolerated in H. influenzae but are unable to transform E. coil because they block transcription from the ApR promoter into the rep locus of pDM2. Third, pDM2 does not have many unique restriction sites for cloning. To help clone H. influenzae genes, we have successfully constructed a series of pDM2-based shuttle vectors for H. influenzae and E. coli which has none of the problems associated with pDM2.

EXPERIMENTAL AND DISCUSSION Abbreviations: Ap, ampicillin; BHi, brain heart infusion; bp, base pair(s); Cm, chloramphenicol; H., Haemop~llus; kb, kilobase(s) or 1000 bp; Kin, kanamycin; LB, Luria-Bertani (medium); NAD, p-nicotinamide adenine dinucleotide; ~:r/, origin of DNA replication; n, resistant/resistance; s, sensitive/sensitivity; Sm, streptomycin; Sp, spectinomycin; [ ], denotes plasmid-carrier state; t2, insertional mutation. 0378-1119/90/$03.50 © 1990Elsevier Science Publishers B.V.(Biomedical Division)

(a) Media and strains

E. coTiwas grown in LB broth and agar. Antibiotics were used at the concentration of 50/~g Ap/ml, 20 ~g Cm/ml,

100 Orll I

8mal

pVTSe

I

I 8ram I

Sop I

pVTSS

I BamH I

BimH I

pVT54

I Sins I

Xmn I

I

Ikb

pVTS3

I

pill

Pill

pVTS2

I 8roll I

Xnml

pVTSI

I PvuI

Pvull Pvull

I

I o.i

Kmfl

~ - ~ 1

OimHl Xmnl PMI

I

I Ik~i'~

I

aunHl

I

ApR

~pIOitl

I "

i~ul

: :

pVTSO

XmmI

Fig. !. Deletion derivatives ofplasmid pVTS0. Various deletions ofpVTS0 were generated by in vitro DNA manipulation. The deletion derivatives were recognized by their Ap s phenotype. Six bars indicate the DNA regions deleted. The restriction sites used to generate the deletions are indicated at the ends of each bar. The thick line on the pVTS0 map indicates DNA from pDM2.

50/~g Km/ml and 50/AgSp/ml. Strain JM 109 (recA 1 endA 1 gyrA 96 thi hsdR 17 supE44 relA 1 X- Jlac-proAB [F', traD36 pmAB + lacl q lacZAMl5] (Yanisch-Perron et al., 1985) was used in all cloning experiments. Strain JM 110 (GM48, rpsL thr leu thi iac Y gaIK gaIT ara tonA tsx dam dcm supE 44 Jlac-proAB IF', traD36 proAB + lacl q lacZAM15]) was used to generate methylation free plasmid DNA. JM 107 (thi ~rA96 endA 1 I~dR 17 re/A 1 supE44 X- Lllac.proAB IF', traD36 proAB ÷ lacl q lacZAMI5]) was used in all mutagenesis experiments. H. it~fluenzae strain BC200 (Barnhart and Cox, 1968) was grown in BHI broth and agar supplemented with 10/~g hemin/ml and 2/Ag NAD/ml. Antibiotics were used at the concentrations of 5 #g Ap/ml, 2/~g Cm/ml, 20/~g Km/ml and 15/~g Sp/ml. Transformation of H. influenzae strain BC200 by the MIV medium method was performed as described (Herriott et al., 1970).

(Spratt et al., 1986) to give pVT50. A series of deletion derivatives of pVT50 was generated (Fig. 1). Only two, pVT51 and pVT52, were able to transform H. in./luenzae to Km resistance. The deletion in pVT51 extends from the Smal site in the polylinker to the Xmnl site in the Ap R gene. The deletion in pVT52 extends from the Pstl site in the polylinker to the Pstl site in the Ap R gene of pDM2. Deletions extending beyond the Pstl site of the Ap R gene (pVT53, pVT54, pVT55 and pVT56) were not capable of transforming 14. influenzae to Km resistance. Therefore, an essential element was located downstream from the Ap R gene. This locus was named nep for replication. To delineate the rep locus, pVT52 was mutagenized with transposon TnlT$TCm (Ubben and Schmitt, 1987). The different insertional mutations in pVT52 are depicted in Fig. 2, Only mutation 9 (rep9: :Tn1737¢m) prevented the plasmid from transforming H. influen:ae to Km resistance. Mutation 9 was located close to the BamHl site downstream from the Ap R gene. This observation supported our conclusion about the location ofthe rep locus. Furthermore, we concluded that the rep locus is located within a 1.7-kb DNA fragment bordered by mutation 16 at one end and mutation 15 at the other end. pDM2 replication in H. influenzae requires only the 1.7-kb segment containing the intact nee locus because pVT57 with a deletion extending from mutation 10 to mutation 16 did not affect plasmid

(b) Localization of the plasmld replication (rep) locus of RSF0885

To improve pDM2 stability in E. coil, we analyzed pDM2 in detail to define the locus required for maintenance in £. coil and H. influenzae. The 6.2-kb PmlI fragment of pDM2, which was capable of self replication, was cloned into a P u l l site of pBGSg, a Km R analogue of pUC8 I kb |

,

I

I

Pvull

F_

. ori

KmR

Pvull ~tl

?

la, 12

aU a 13 IS

lkmH I

I 8ep I D m I Pvu I

mp Fig. 2. Insertion mutations ofpVT$2. Plasmid pVT52 was mutagenized with the transposon Tn1737Cm. The open triangles indicate the positions of the mutations. The deletion extending from mutation 16 to the mutation 10 to give rise to pVT$7 is indicated by a bar above the map. The thick bar below

the map indicatesthe rep locus.The thicker line on the map indicates DNA frompDM2.

101 Ikb

i.

t

". Pvuli

Sp R

Pet I

I I

OamH I

II

Hind Ul

l i k b P v u i l fragment lnmt pVT S:l.I

Pwu ii

I

Hind III

I pVTS9

Pvul

Pstl

I

~

BamH I

I~11

I

I

Ap A pVT58 PvuS

BamH I

I

Pat I

I

BamH I

I

Pvull

I

'J

rqJ

Fig. 3. pDM2 replication in E. coli is dependent on the ApH promoter, pv'r$8 which can replicate in E. ¢o/iindependentlyof the ColE! o~ is shown. Replacement of the Pstl.l~ull fragmentfrom pVT$8 whichcarries the Apa promoter with the Pstl-Pmll frasmentfrom pBR325 which also carries the Apa promoter give rise to pV'r$9, pVT$9 also replicated independentlyof the ColEI ori in £. co//. However,when the Apa promoter was deleted as in the 6-kb Pmli fragment from pv'r52.1, replication in £. ¢o/iwas not observed. Other designations as in Fig.2. replication in H. influenzae. During the isolation of these plasmids from H. influenzae, we found that the yield for pVT52fl15 was only one-tenth of the yields for pVT52QT, pVT52t210 and pv'r52Q16. This suggested that mutation 15 reduced the plasmid copy number in H. influenzae. The reason for this reduction remains to be investigated. (c) pDM2 replication in E. ¢oli is dependent on the Ap R promoter We attempted to construct a plasmid derived from pVT$2 which lacks the ColEI ori with the expectation that the rep locus would promote plasmid maintenance in E. coll. We placed a 1.95-kb Hindlll fragment carrying the Sp R marker flanked on both sides by Rho-independent terminators from pRU876 (Prentki oral., 1984; Ubben and Schmitt, 1987) into the Hindlll site on pVT52 to give

_

| ort

pv're4

B

1

I

' ori

H

KmH

Ps

I

I

B

E

I

|

"

PvlO

PvlD ~p

I

:

epn

APR

pVTU pvm I

~ orl

~

~.

PvlH

B

I

or"'~--

H

I

APfl I I

:1 i

•I

,I p,,,x I

,:I

,

APR

pvTee I 0 I

i

APR

PvlD

,

pVT52.1. Self-ligation of the large P m i l fragment of pVT52.1, which carries the Sp R marker and the rep locus (Fig. 3), should have yielded a self-replicating plasmid that conferred Sp resistance to its E. coli host. However, repeated attempts did not yield such a construct, in a parallel experiment, the 6.2-kb P m l l fragment of pv'rso was successfully self-ligated to yield a free replicating plasmid which confers Ap resistance to its E. coli host. The plasmid was named pVT58 (Fig. 3). Replication in E. coil was also observed when the Pstl-Pmll fragment containing the Ap R promoter ofpVT$8 was replaced by the Pstl.Pmll fragment from pBR325 which also carries the Ap R promoter. The construct was named pVT$9 (Fig. 3). We concluded that pDM2 replication in E. coil required an intact Ap R promoter upstream from the rep locus. It seems that, in E. coil, an essential c/s-acting element in the rep locus relied on transcriptional run-off from the Ap n promoter

e

~,m I

CmR

Pa

B

I

E

I

I

PvlH

I

-

8m

PvlH

I

:

CmR 3 I

4 i

S

I

6 i

Tkb i

Fig. 4. Restriction maps ofpVT57 derivatives. A series ofpiasmid derivatives from pVT57 was constructed. Piasmids pVT63, pv'r64, pVT6S, and pVT(~ are suitable as shuttle vectors for Haemophllusinfluenzaeand E. coll.The thicker lines indicate DNA from pDM2. The restriction sites are abbreviated as follows:B, BamHl; D, Dral; E, EcoRl; H, Hincll; Ps, Pstl; Pc, Pmll; S, Sml; Sp, Sphl; X, Xmnl. Two abbreviationsseparated by a slashcorrespond to a hybrid of two restrictionsites.The hybrid site is not recognizedby either of the two restrictionendonucleasesthat were used to create it. The arrows indicate the direction of transcription or replication; when arrowhead is absent, the direction of transcription is not known.

102

TABLE ! Availablecloningsites of pVT63,pv'r64, pVT6$,and pVT66 Plusmid"

Cloning sitesb

pVT63

Ape (Hbecll,Pstl, Seal) KmR (Clal, H/ndlll, Nrul, Smal, Xkol)

pVT64

Apa (Hbscll,Psd, Seal, Sspl) Spa (not determined)

pVT6$

Ape (H/ncll, Psd, Pml, $ml, Sspl) Cma (Ball, Ncol)

pVT~

Ape (H/ncll, Pstl, Pml, Sml) Cma (Sinai)

" See Fig.4 for the structure of these plusmids. b Uniquerestrictionsitesfor insertionalinactivationofeachdrugmarker appear in the parantheses. Pml sites are not very useful for cloning, because Pml ends are not readilyligated.

Therefore, we constructed a series of cloning vectors derived From pDM2. These vectors carry the pUCderivative ofthe ColE 1 or/and therefore are stable in E. coli. When grown in the presence of Cm, 100~o of the pVT66transformed cells carried pVT66. In contrast, only 0.3-3 of the pDM2-transformed cells carried pDM2. The percentage of plasmid-bearing cells From a broth culture derived From a single colony isolate grown to log phase in presence of Cm was determined by plating on plates with and without Cm. Because Cm is bacteriostatic, the cured cells, which become Cm sensitive, were not killed. Therefore, the percentage of Cm e cells in the culture reflects the stability of the plasmid. Insertion of the ColEl ori allows insertional inactivation ofthe Ap e gene without any adverse effect on plasmid maintenance.

ACKNOWLEDGEMENTS because strains JM 109[pVT58] and JMI09[pVT59] failed to support the establishment of the 6-kb P m l l fragment from pVT52.1. Insufficient transcription into the nep locus in £. ¢ol1 would account for the low level of pDM2 maintenance in E. coll.

We thank Dr. Brian G. Spratt for the vector pBGS8, Dr. Rudiger Schmitt for pRU868 and pRU876, and Dr. Nancy KJeckner for pNK862. This work was supported the grant HRC-RRP-B-026 from the Oklahoma Center for the Advancement of Science and Technology, and grant N00014-86-K-0222 from the Office of Naval Research.

(d) Construction of pDM2 derivatives suitable as shuttle vectors REFERENCES Because pDM2 replication in E, cell was dependent on the Ap e promoter, a recombinant, with insertional inactivation of the Ap e gene, will replicate in E. coil only if transcription can proceed through the insert into the mp locus. Passage through E. coil of any recombinant library, which inactivates the Ap e Bane, presumably places selective pressures for inserts containing promoters or lacking transcription terminators. In order to bypass these selective pressures, we constructed a series of cloning vectors containing the or/pBGS8 and the rap locus of pDM2. The parent plasmid used in the construction was pVT57. The resulting constructs are shown in Fig. 4. Their cloning sites are listed in Table I. The yields for these plasmids were approximately 500#g culture per liter of £. coil and 50-100 pg per liter of H. influenme. The fact that transcription from the Km ~ gene is directed toward the rep locus in pVT63 did not have any effect on the yield of this plasmid. All drug markers on these plasmids were found to express effectively in H. influen:ae strain BC200,

(e) Cemelmlmm p D M 2 is an excellent cloning vector in H. b~./lue~.ae. However, in E. coil, it is extremely unstable, possibly because its replication is dependent on the Ap e promoter.

Barohart, B.J,and Cox, S.H,: Radiation.sensitiveand radiation.ruistant mutants of Haemophtlw Olfluenzae.J. Bacteriol. 96 (1968) 280-282. De Graaff, J., Elwell, L.P. and Falkow, S.: Molecular nature of two beta-lactamase-sp~il~in8 plasmld8 isolated from Haemopkilus in. j~enzae type b, J. Bacteriol. 126 (1976)439-446. Herriott, R.M.,Meyer,E,M.and VeSt,M,: Definednonlprowthmediafor staBe il development of competence in Haemopkllw ~enzae. J. Bacteriol. 101 (1970) 517-524. McCarthy, D.: Cloningof the rec-2locusofHaemopkilmblfluenzae.GeM 75 (1989) 175-143. McCarthy, D., (,iayton, N.-L. and Sallow,J.K.: A plasmidcloningvehicle for Haemop~,.,lus~quenzae and £sckerickla co//. J. Bacteriol. 151 (1982) 1605-~607. McCarthy, D., Gr~Mn,K. and Sallow, J.K.: Piasmidcontaininga DNA liaasc 8ene from Haemopkilu; b~enzae. J. Bacteriol. 158 (1984) 730-732. Prentki, P. and Krisch, H.M.: In vitro insmional mutesenesiswith a

selectable DNA frNpmant.GeM 29 (1984) 303-313. Sedow, J,K,, Spikes, D. and GdMn, K.: Characterization of the rec-I gene of Hae#,ophYw tq/luen:ae and behaviour of the sane in F,,vckertck~teA,J. BacterioL 170 (1988)3876-3881. Spratt, B.G.,Hedae,PJ., te Heesen, S., Edehnan,A. and Broome-Smith, J.K.: Kanamycin-rosistantvectors that are analogues of plasmids puca, pucg, pEMBLa,and pEMBL9.GeM 41 (1986) 337-342. Ubben, D. and Seniti, R.: A transposablepromoter and transposable promoter probes derived from Tn1721. Gene 53 (1987) 127-134. Yanisch-Perron, C., Vseira, J. and Messing,J.: Improved MI3 phage cloning vectors and host strains: nucleotide sequences of the Ml3mpl8 and pUCI9 vectors. Geoc 33 (1985) 103-119.