Gene, 5(1979) 59-76 59 @ Elsevier/North-Holland Biomedical Press, Amsterdam -- Printed in The Netherlands

CONSTRUCTION OF PLASMID CLONING VEHICLES THAT PROMOTE GENE EXPRESSION FROM THE BACTERIOPHAGE LAMBDApL PROMOTER

(ColE1-), hybrid plasmid; tryptophan synthetase genes; temperature induction of gene expression; X-immunity region)

HANS-ULRICH BERNARD, ERIK REMAUT*, M. VICKERS HERSHFIELD**, HIRENDRA K. DAS*** and DONALD R. HELINSKI

Department of Biology, B-022, University of California, San Diego, La JoUa, CA 92093

and

CHARLES YANOFSKY and NAOMI FRANKLIN

Department of Biological Sciences, Stanford University, Stanford, CA 94305 (U.S.A.)

(Received October 30th, 1978) (Accepted November 8th, 1978)

SUMMARY

Two multiple-copy, ColEl-type, plasmid cloning vehicles, pHUB2 and pHUB4, have been constructed that carry four different single restriction sites down- stream from the phage lambdapromoter pr.. The promoting activity of PL is switched off at low temperature in the presence of a eIts gene that specifies a temperature-sensitive repressor but could be activated by heat induction~ t i t s was located either on the host chromosome, or on a second plasmid pRK248 that is compatible with the cloning vehicle, or on the vehicle itself. Three different restriction fragments, each carrying the gene trpA of Salmonella typhimurlum or Shigella dysenteriae, have been inserted into the EcoRI, BamHI and SalI sites, respectively, of these plasmids andPL dependent expression of the inserted gene in Escherichta coU was determined by measuring the enzymatic activity of the trpA gene product. Heat induction resulted in a level of expression of trpA corresponding to I to 6.6% of the total soluble cell protein as trpA protein. The level of trpA protein production depended on the particular insert and the plasmidused.

*Present address: Laboratorium voor Moleculaire Biologic, Rijksuniversiteit Gent, B-9000 Gent (Belgium). **Department of Microbiolory, Duke University Medical Center, Durham, N.C. 27710 (u.s.~) ***School of Environmental Sciences, Jawaharlal Nehru University, New Delhi-110067 (India).

60

INTRODUCTION

A structural gene on a fragment of DNA that is inserted into an E. coil cloning vector is expressed as a protein product in E. coU in vivo if the fragment includes the promoter and the ribosome binding site of the gene and if these sequences are recognized by the transcription and translation systems of K coK This seems to be the case with genes of prokaryotic origin (Hershfield et al., 1974) and may also explain the expression of certain genes from lower eukaryotes (Struhl and Davis, 1977; Ratzkin and Carbon, 1977; Vapnek et al., 1977). Prokaryofic genes on restriction fragments not carrying their natural promoter may be expressed either by transcription initiation events at random sites or by read through from a piasmid promoter (Selker et al., 1977). It has been observed that restriction fragments from higher eukaryotes produce transcripts o f random size in E. coil, but to date there is no evidence for the expression of these genes as biologically active proteins (Morrow et al., 1974; Chang et al., 1975; Miller et al., 1977).

On the basis of these observations it was considered important to construct a cloning vector that would not only serve to replicate an inserted restriction fragment, but which would also allow transcription of genes on the insert to be controlled from a promoter on the cloning vector. Such a cloning vehicle would be useful for increasing the expression of prokaryotic or eukaryotic genes that would otherwise be only weakly expressed in E. coli.

An effective vehicle for this purpose should have the following features: (a) it should carry a strong promoter; (b) this promoter should be under the control of a regulatory gene so that normally the promoter is switched off but can be activated experimentally; (c) the vehicle should have several unique restriction sites for insertion of DNA downstream from the promoter; and (d) the vehicle oqhould be present in high copy num~er in the cell to provide multiple copies for expression of the DNA insert.

One of the best investigated promoter regions for the construction of such a cloning vehicle is the leftward promoter-operator of phage ),. Transcription of the N operon starts at the promoter PL and is carried out by £. coli RNA polymerase (Taylor et al., 1967; Takeda et al., 1969). The polymerase reads the N gene first and terminates transcription at the transcription termination site tL in the absence of the N gene product (Roberts, 1971; Salstrom and Szybalski, 1978a). The product of t heN gene, together with the nutL recognition site and host factors, modifies transcription in such a way that RNA polymerase reads through tT. and other transcription~ termination signals into subsequent structural genes (Franklin and Yanofsky, 1976; Epp and Pearson, 1976; Adhya et al., 1976; Salstrom and Szybalski, 1978b). Initiation of transcription at pT. canbe alternatively repressed by the products of the genes cI and cro (Kumar et al., 1970; Maniatis et al., 1975; Folkmanis et al., 1976). Thus, a cIts cro- back- ground in a cell offers the possibility of controlling initiation at PL by a temperature shift.

It has been shown with phages made b y illegitimate recombination in ~ivo

61

(Franklin, 1971; Moir and Brarnmar, 1976) and by in vitro recombinant DNA techniques (Hopkins et al., 1976) that genes downstream from PL and N are not essential for lytic infection by X and, for example, can be replaced by genes of the tryptophan (trp) operon of E. coli. Several of these ~trp phages lack the main trp promoter and in such cases expression of the trp genes is dependent on transcription initiation from PL. Recently, by triple ligation of three EcoRI fragments, one of which carried thePL promoter of k, a hybrid plasmid was constructed where the expression of the cI gene of phage 21 was increased ten-fold by transcription from the PL promoter (Hedgepeth et al., 1978).

In this report a DNA restriction fragment obtained from ~trp DNA that contains PL and E. coU trp genes was cloned into different derivatives of the high copy number plasmid ColE1 and the function of PL was monitored by measuring the enzymatic activities of the cloned trp gene products. The trp genes in these plasmids were then deleted in order to construct multiple-copy vehicles that carry four different single restriction sites downstream from the PL promoter. Restriction fragments carrying genes of the trp operon were inserted at several of these sites to determine the effectiveness of these plasmids as vehicles for the promotion of expression of genes on a DNA insert.

MATERIALS AND METHODS

(a ) Materials Lysozyme, kanamycin, tetracycline were obtained from Sigma; the restriction

enzymes PstI, Bali, HpaI, BamHI, HindIII, BglII from New England Biolabs; T4 DNA ligase from Miles Laboratories; and agarose from SeaKem. EcoRI was prepared according to Greene et aL (1978). trpB protein used in the assay for trpA activity was kindly provided by S. Mills.

(b) Bacterial attd phage strains The constr~lction of ~,trp43 and Xtrp48 has been described by Franklin

(1974). Thes~ phages carry the cIAt2 mutant gene that specifies a temperature sensitive repressor protein,

K-12AH1 is the designation for the strain K coil K-12M72F- lac- amSm R ()~cI857N7N53AHlbio') as described by Castellazzi et al. (1972). The cIts gene is present on a ), prophage that has been partially deleted. K-12~ H1 was made AtrpEA2 by transduction with Plc/ear. K coU C600AtrpEB9 or C600()~)-

trp£B9 was used as a host to select transformants for growth on indole conferred by the trpB protein. K coli C600trpA33 or C600(),)trpA33 was used as a host to select for plasmids containing trpA.

Plasmid pCRI, a ColEl-kan It plasmid with a single EcoRI site, was isolated by Covey et al. (1976) and has been described by Armstrong et al. (1977). Plasmid pBR322 was described by Bolivar et al. (1977). pMK2004 is a kana- mycin-resistan~ derivative of pBR322 constructed by M. Kahn.

62

(c) Growth of bacteria Bacteria were either grown in L-broth, L-broth agar, or minimal salts plus

glucose supplemented with 0.2% casamino acids (Difco-certified). The latter medium is designated as minimal medium in this paper. When selection for trpB was attempted, 10 #g/ml indole was added. The concentrations of tryptophan, kanamycin and tetracycline were 50, 50 and 25 #g/ml respectively.

(d) DNA procedures Plasmid DNA was isolated from overnight cultures in L-broth using lysis

with the lysozyme/Triton X-100 procedure of Katz et al. (1973) followed by phenol extraction, dialysis and ethanol precipitation. In some cases DNA was purified by centrifugation in a CsCI ethidium bromide density gradient. The yield of plasmid DNA was approximately 1 #g DNA/ml culture. Plasmids were never -mplified by incubation of the cells in the presence of chloramphenicol since continued replication of the plasmid in the absence of protein synthesis may result in the activation of the p,. promoter. Agarose slab gel electrophoresis was performed as described earlier (Hershfield et al., 1974). ), DNA, digested with £coRI, was used to provide molecular weight markers (Allet et al., 1973). Restriction ligation and gel electrophoresis conditions were described elsewhere (Meyer et al., 1977). Blunt end ligations were done according to S g ~ m e l l a et al. (1977) and transformations according to Cohen et al. (1972). All re- combinant DNA procedures were carried out in accordance with the National Institutes of Health Guidelines for Recombinant DNA Research.

(e) Expression of trpB and trpA genes Cells which contained plasmids carrying PL under the control of a temper-

ature-sensitive repressor were grown in minimal medium supplemented with tryptophan in the absence of antibiotics at 32°C. When the exponentially growing culture reached 70 Klett units, the temperature was shifted for 5 rain to 45 ° C to inactivate the repressor and the culture was then maintained continuously at 41°C. Samples (50 ml) were removed to obtain timepoints of expression of trpB and trpA under the influence of PL and the cells were harvested and frozen. Cells were resuspended in 50 mM potassium-phosphate buffer (pH 7.5) containing 20% glycerol, I mM DDT, 50 mM NaCI, 10 mg/l pyridoxal phosphate, sonicated and centrifuged 15 min at 10 000 g. The enzyme activity of the trpB or trpA gene product was determined by measuring the reaction indole + serine -~ tryptophan as described by Smith and Yanofsky (1962) in the presence of an excess of the other subunit of tryptophan synthe- tase. One unit of TSase catalyses the reaction of 0.1 #mol of indole in 20 min (Smith and Yanofsky, 1962). Determinations of the soluble protein in cell lysates were made by the method of Lowry (1951).

RESULTS

(a) Construction of hybrid ColE1 plasmids carrying PL In phage ktrp48, the genes trpD, C and B ofE. coli are fused by illegitimate

63

recombination in vivo to the N operon of ~ so that expression of these genes is dependent on transcription from the phage promoter PL (Franklin, 1974). One EcoRI fragment of ;~trp48 DNA (9.2 kb in size, Fig. la) extends from an EcoRI site in ~, gene O (Thomas and Davis, 1975) leftward through cro, the PR promoter, the cIts gene, p~., N and trpD, C and B to an EcoRI site within a ~, DNA sequence to the left of or in the b2 region (Thomas and Davis, 1975). This fragment provides a selectible phenotype for transformation of certain trp- bacteria to tryptophan independence, since trpB is weakly and constitutively expressed from the trp operon internal promoter P2 located near the end of trpD (Jackson and Yanofsky, 1972). The trpB protein catalyzes the synthesis of tryptophan from indole (Creighton and Yanofsky, 1970).

DNAs of ;~trp48 and pCRI (Covey et al., 1976) were cleaved with gcoRI, ligated and transformed into C600(~,)A trpEB9. TrpB + trausformants were selected at 32°C on minimal plates supplemented with indole and were then tested for Kan R . Plasmid DNA was isolated from 8 transformants that were TrpB ÷ and Kan R and was shown in each case to be cleaved by EcoRI into two fragments. One Of the fragments was 13 kb in size, corresponding to pCRI, while the other was 9.2 kb, corresponding to the EcoRI fragment of ;~trp48 (data not shown). The relative orientation of the two fragments was determined by double digestion with SalI and PstI making use of the single 8alI cleavage site located in the pCRI component of the hybrid plasmid (Armstrong et al., 1977) and an asymmetric PstI site left of cI at the position of 76.4 %~ wild-type DNA (Smith et al., 1976; J. Salstrom, pers. commun.) in the ~,trp48 component (Fig. le). The hybrid plasmid with PL facing towards the SalI site was designated pER482 (5 out of 8 transformants) while the plasmid with the opposite orientation was designated pER481. Plasmids pER481 and pER482 were transformed into C6005trpEB9 and C600(~)~ trpEB9 and either Karl R or TrpB ÷ was selected. It was observed that therewas no difference in transformation efficiency using either host or either selection condition. C600A trp£B9[PER481 ] and C600A trpEB9[pER482] were grown in minimal medium supplemented with tryptophan at 32°C in the absence of kanamycin to a Kletts4o of rl0. The temperature was shifted for 5 rain to 45°C and then to 41°C, and samples were taken before and at several time points after the heat treatment. The level of trpB enzyme activity was measured in each of the samples after harvesting and disrupting the cells by sonication. Table I shows the level of the trpB protein synthesized after heat induction. The production of 20 units of trpB protein/mg extract protein in the absence of heat induction is presumably due to the activity of the tl~ internal promoter P2. This corr£~ sponds to about 0.8% of the total cell protein. After 60 to 120 rain of heat induction, this level rises to approx. 5% of total cell protein. This increase, presumably the remit of initiation of transcription at PL after heat inactivation of the cIts gene product, indicates that pw. is functional in pER481 and pER482.

d~

TABLE

I

PL

DE

PE

ND

EN

T E

XP

RE

SS

ION

OF

trp

B O

R t

rpA

FR

OM

DIF

FE

RE

NT

PL

AS

MID

VE

HIC

LE

S

Whe

reve

r de

term

ined

in

this

stu

dy,

trpB

was

fuse

d w

ith

PL

and

was

clo

ned

into

the

pla

smid

on

a s

ingl

e D

NA

fra

gmen

t co

ntai

ning

PL

. A

]] t

vpA

exp

ress

ion

data

wer

e ob

tain

ed f

rom

hyb

rid

plas

mid

s co

nstr

ucte

d by

ins

erti

ng t

he tr

pA g

ane

into

a

vehi

cle

carr

ying

PL

. A

das

h in

dica

tes

that

th

e ac

tivi

ty w

as n

ot

mea

sure

d. A

t zQ

ro t

ime

expr

essi

onis

ind

epen

dent

of

PL

an

d w

as p

resu

mab

ly d

ue t

o t

he

trp

op

ero

n i

nter

nal

pro

mo

ter

p,.

Cal

cula

ting

th%

mea

sure

d en

zym

e ac

tivi

ties

rel

ativ

e to

th

e ac

tivi

ty o

f th

e pu

re t

rpB

or

trpA

pro

tein

(C

reig

hton

and

Yan

ofsk

y, 1

970)

all

ows

an e

stim

ate

of

the

leve

l of

gene

ex

pres

sion

pro

mo

ted

by

PL

as

the

perc

enta

ge o

f to

tal

cell

ular

pro

tein

. T

he s

peci

fic

acti

viti

es o

f tr

pB a

nd t

rpA

pro

tein

s ar

e 27

00 a

nd 5

000,

res

pect

ivel

y. E

ach

valu

e re

pres

ents

the

ave

rage

of

two

enzy

me

dete

rmin

atio

ns o

n ea

ch s

ampl

e. O

ne

unit

of

tvpB

or

trpA

spe

cifi

c ae

tivi

ty c

orre

spon

ds t

o t

he r

eact

ion

of

0.1

#too

l in

dole

in 2

0 ra

in a

t 37

°C p

er l

ug e

xtra

ct

prot

ein

Pla

smid

H

ost

trpB

tr

pA

Spe

cifi

c A

ctiv

ity

Spe

cifi

c A

ctiv

ity

0'

60'

120'

18

0'

0'

60'

120'

18

0'

240'

pER

481

pE

R4

82

pM

K20

04;~

trp4

8trp

A

pMK

2004

Atr

p48t

rpA

pH

UB

2trp

A ( E

coR

I)

pHU

B2t

vpC

BA

{Bam

HI)

pH

UB

4tvp

CB

A(B

amH

I)

pHU

B4t

rpD

CB

A (S

ail)

pH

UB

2trp

A(E

coR

I) +

pR

K24

8¢It

s pH

UB

12tr

pA (

Eco

RI )

C60

0trp

AE

B9

21

97

C60

0tro

nEB

9 22

14

S

K-1

2~'I

IAtr

pEA

19

60

K

-12A

HIA

trpE

A

18

40

K-12~HIA

trpEA

--

--

K-12~H1n

tvpEA

--

--

K- 12AH14

tvpEA

--

--

I(-12AHIA t

rpEA

--

--

C60

0trp

A 3

3 C

600t

rpA

33

m

m

127

..

..

..

I13

..

..

..

--

124

12

8

--

13

--

--

98

5

12

--

17

--

--

--

2

20

52

55

114

--

--

3 1

08

--

1

00

--

--

--

29

1

69

--

3

29

--

--

--

15

51

--

68

--

m

m

1 --

94

--

86

1

--

88

--

13

5

65

(b) Construction o f cloning vehicles with a single EcoRI site downstream from the PL promoter

Plasmids pER481 and pER482 were constructed to assess the function of PL cloned into a plasmid. They do not allow any measurement of the effect of the N protein on expression, since the deletion of tL in ktrp48 permits N independent expression of trpB from PL (Franklin, 1974). In addition, they are not suitable vehicles for the insertion of other DNA fragments, since no unique restriction sites were detected downstream from pT.. To overcome these di-oadvantages, another group of plasmids was constructed starting with the ColE1 cloning vector pMK2004 and the DNA of phages ktrp43 and )~trp48 (Fig. 1).

pMK2004 is a derivative cf plasmid pBR322 that carries a kanamycin resistance gene (M. Kahn, unpublished results). )~trp43 has a structure similar to ktrp48 except that it contains an additional 2 kb segment of ;~ DNA between the N gene and the point of fusion with trpD (Fig. lb) (Franklin, 1971). This additional DNA includes tv. and makes trpB expression from PL in the phage dependent on the function of the N protein. The region of DNA between the PstI and EcoRI sites of pMK2004 carries part of an ampicillin resistance gene but is not required for replication of the plasmid (Bolivar et al., 1977; M. Kahn, pers. commun.). Replacement of this DNA with a PstI-EcoRI fragment of either ),trp43 or Xtrp48 that carries pT. and the trp genes, but not cl andPR (Smith et al., 1976 and Fig. la and b), should yield a plasmid that would have a single EcoRI site downstream from pT.. It should then be possible to insert a promoterless DNA fragment carrying a gene specifying a measurable enzyme into this EcoRI site as, for example, the EcoRI fragment carrying the trpA gene of Salmonella typhimurium (Selker et al., 1977) (Fig. lc). These constructions should allow a test of whether or not the N dependent expression of trpB in the Xtrp43 derivatives is as efficient as the N independent expression of trpB in the ),trp48 derivative.

pMK2004 and ~trp43 or ),trp48 DNA were cleaved with Pstl plus EcoRI; the fragments were ligated and the mixture used to transform C600(),)A trpEB9, with selection for Kan R and TrpB ÷. The plasmid DNA of several transformants of each type was analysed and found to possess a single PstI and a single EcoRI site. Double digestion with PstI and EcoRI resulted in two fragments from the Xtrp43 transformants, one of a size of 10.0 kb corresponding to ),trp43 DNA and the other of a size o~ 4.4 kb corresponding to pMK2004 DNA. Similar digestion of the plasmids from the )~trp48 transformants yielded the expected 7.1 kb fragment from )~trp48 and the 4.4 kb fragment of pMK2004 DNA. These two plssmids were designated pMK2004~trp43 (Fig. 2) and pMK2004~trp48.

pMK2004),trp43 and pMK2~J4Xtrp48 were cleaved with EcoRI, ligated with the EcoRI fragment containing the trpA gene of 5. typhimurium, and the mixture transformed into C600(X)trpA 33. Transformants selected for Kan R and Trp ÷ contained plasmids with the trpA EcoRI piece inse.,~ced into the single EcoRI site of ~he two vehicles. This selection relied on the weak unspecific promoting activity of this fr .agment described by Selker et al. (1977). The

66

orientation of the trpA gene relative to p~ could be determined on the basis of the locations of the SalI sites in the tet a gene of the vehicle and in the inserted trpA fragment (Fig. lc and 2). Plasmids havir~ ~he trpA gene oriented in the same direction as transcription from PL were designated pMK2004),trp43trpA and pMK2004),trp48trpA. These two plasmids were transformed into an E. coU strain, K-12~IIA trpEA, that produces a heat sensitive ), repressor from a ), cIts gene on the host chromosome (Castelazzi et al., 1972). The expression of trpB and trpA in these two plasmid containing strains is shown in Table I. The increase in trpB activity in each of these two strains is very similar. These results indicate that RNA polymerase molecules initiating at PL in the pMK2004~trp43trpA plasmid can efficiently overcome the tL termination region. These results suggest that pw. and N are functional on these plasmids and stimulate trpB expression. In contrast, trpA of the insert at the EcoRI restriction site is not expressed with high efficiency in the case of pMK2004~trp48trpA and there is no L~crease with pMK2004~trp43~rpA.

(o) (coRl Hpol Hpo| PstI (coRl - - - I I 5.0 / 1.8 I 2.1 I - - -

0.35 <"t !'~ tr.BtrpC trpD NPL ¢! PR

(b)

(c)

(d)

(coRI HpoI Soil OamHl Hpol Pstl EcoRI _ . ' I f" 6.7 "I i.s'lo.Ti" ,.61 2., i . . _

0.35 4-1 I '-) .pe .pc ,,po T 'L"P" cipR

, ~ ~ e m .e l

EcoRI Soil Hinm EcoR! . . . I 2.e ~o.61 i .g l___

trp A

Soil Hpo! 6910 Sol! _ ._1 1.5 I 2.5 I 1.51___

trpA trpBtrpC trpD ~ u ~ m g ~ m m m o

(coRl

( e ) Psi i

c°RI Psi I ~

Pstl Pst!

67

(c) Construction of cloning vehicles with several single restriction sites down- stream from PL

The site of insertion of an EcoEI fragment into pMK2004),trp43 is located 8.6 kb from the promoter pT.. In the hope that decreasing the distance between p t and an inserted gene would increase its level of expressiov upon heat induction, pMK2004ktrp43 was further modeled. It was found that ),trp43 DNA was cleaved by HpaI at a HincII site located at 72.9 %k molecular length near the mid-point of the N gene (Allet and Bukhari, 1975 and unpublished DNA sequence data of N.F.). To eliminate the 7.9 kb HpaI fragment carrying the trpD, C and B genes (Fig. 2) the plasmid was cleaved with HpaI followed by blunt end ligation and transformation into C600(k)AtrpEB9. Kan R trans- formants were screened for loss of TrpB ÷ on minimal plates containing indole. All Kan R TrpB- clones isolated contained a plasmid that was derived from pMK2004ktrp43 by a deletion of a 7.9 kb HpaI fragment. This plasmid contained single HpaI, EcoRi, BamHI and 8alI sites within a distance of 0.30 to 1.3 kb downstream from pT. and was designated pHUB2 (Fig. 2). After transformation of plasmid pHUB2 into K.12AHIAtrpEA, it could be shown by cross streaking against kimm434NTN53 that no functional N gene product was produced by pHUB2 indicating that the HpaI deletion had removed the segment of DNA carrying the N gene.

By cleavage of pMK2004klrp43 with 8alI and re-ligation, another derivative of pMK2004ktrp43, designated pHUB4, was col~structed, pHUB4 contains the region coding for the N gene, ~ d has a single BamHI site and a single SalI site located 1.2 kb and 2.7 kb respective~.y, downstream from pT..

Fig. 1. Restriction and genetic maps of DNA segments and hybrid plasmi&. Distances are indicated as kilobase p~irs (kb).

(a) EcoRI fragment of ~,trp48. This tragment carries the phage X promoters PI, and PR. Genes trpB, C and D are located dowr~tream of PI~ in the same reading direction as the N gene. The ¢I gene of this fragment carries a mutation rendering the repressor protein heat sensitive. The cro gene is present in mutant form to prevent interference with transcription from PL and PR'

(b) EcoRI fragment of ~frp43 has a similar structure to ~trp48 except that it carries additional genetic information of phage ;~ between PL and trpD. The Pst! site indicated was used for cloning the PL region into plasmid pMK2004.

(c) EcoRI fragment carrying the promoterless frpA gene of Salmonella typhimurium. This fragment was originally isolated by Selkar et aL (1977) from ~80 frpE-A DNA of Salmonella typhimurium. The position and direction of the trpA gene relative to a single HindIII site was determined by Selker et al. For the determination the orientation of this fragment in various plasmids the asymmetric single SalI site was employed.

(d) SalI fragment containing the trpD-A genes of Shigella dysenter[ae. The asymmetric single Hpal site was employed for the determination of the orientation of this fragment in hybrid DNA molecules.

(e) Plasmids pER481 and pEBA82. The thick lines indicate the EeoRI fragment of ~,trp48 that is shown in detail in Fig. la ; the thin lines indicate the pCRI component of the hybrid plasmids. An asymmetric Pstl site in the ~trp48 part and a single SalI site in pCRI define the two possible relative orientations of the two EcoRI fragments.

68

BamHI Soil

Pst I

S o l i d i "" Hpol HpoI / FcoRI ~SolI

BomHI

Hp(] I HpoI

sa,l

Fig. 2. Restriction and genetic maps of plasmid pMK2004~trp43 and its two derivatives pHUB2 and pHUB4. The pMK2004 component carrying an origin of replication (ori) and kanamycin and tetracycline resistance genes is indicated by a thin line. The ~,trp43 component is indicated by a thick line. Cleavage of the plasmid pMK2004ztrp43 at its two HpaI sites followed by blunt end ligation and cleavage at its two SalI sites followed by cohesive end ligation were carried out to construct pHUB2 (7.6 kb), and pHUB4 (6.5 kb), respectively.

(d) Expression of genes inserted into pHUB2 and pHUB4 pHUB2 was cleaved with EcoRI, ligated with the EeoItI fragment carrying

trpA (Fig. lc) and transformed into C600(~)trpA33. Kan n Trp* transformants were shown to ~ontain pHUB2 with the trpA fragment inserted into the single EcoRI site 0.6 kb downstream from PL.

Both possible orientations of this insert occurred and could be distinguished by digestion with SalI (Fig. 3) since there is an asymmetric SalI site within the EcoRI piece (Fig. lc). pHUB2trpA(EcoRI), containing the trpA gene in the same orientation as transcription from p~., was transformed into K-12AHIA trpEA and trpA expression was measured after heat induction of pL. The data show that expression of the trpA gene is almost completely dependent on the heat induction of pt. and that trpA enzyme reaches a level of approx. 2% of total cell protein (Table I and Fig. 4). In contrast to pHUB2, pHUB4 does not contain an EcoRI site but does contain the N gene down- stream from PL- To compare the expression of a gene inserted into these two plasmids, a BglII-SalI fragment of 8higella dysenteriae that contained trpC, B and A (Fig. ld) was inserted into pHUB2 and pHUB4 replacing DNA sequences between the single BamHI and SalI sites of these plasmids. The two plasmids wore designated pHUB2trpCBA(BamHI) and pHUB4trpCBA(BamHI). Table I shows that in K-125H1AtrpEA trpA is expressed from both plasmids after heat inducing PL. The enzyme activity reaches a level of 105 units/rag

69

(o) (e): ( , f ) ( q )

Fig. 3. Agarose gel electrophoresis analysis of pHUB2trpA(EcoRI) DNA cleaved with restriction enzymes. The agarose concentration in the gel was 0.8%.

(a) pVH201~ ltrpA (E. Remaut and V. Hershfield, unpublished) digested with EcoRI. pVH201AtrpA was used as the source of the trpA fragment (lower band) in this experiment.

(b) EcoRI digest of pHUB2. (c) and (e) EcoRI digests of two pHUB2trpA(EcoRI) hybrid plasmids that differ in

orientation of the fragments. A SalI digest of the same two plasmids (d) and (f) reveals the two possible orientatons of the two EcoRI fragments relative to one another. The plasmid DNA analyzed in slots (c) and (d) contains the trpA gene in the sense orientation relative to PL as determined by the location of the SalI sites in the trpA fragment (Fig. lc) and the pHUB2 part (Fig. 2).

(g) EcoRI digested ~ DNA (size markers) with sizes of 20.7, 6.8, 5.4, 4.5 and 3.5 kb (Allet et al., 1973).

protein (2% of total cell protein) in the case of the plasmid pHUB2trpCBA(BamHI) and of 329 units/mg protein (6.6% of total cell protein) with pHUB4trpCBA(BamHI).

To explore ftu~her the effectiveness of PL in promoting the expression of a gene insert, and to test expression after insertion at the Bali site, a SalI fragment of ShigeUa dysenterige containing the genes trpD, C, B and A was inserted into the plasmid pHUB4 in both orientations relative to PL (Fig. l d and Fig. 5). Table I and Fig. 6 show that in the K-12~H1AtrpEA strain trpA is expressed before heat induction to a basic level of 0.3% of the total cell protein presumably as a result of the promoter activiW of p2. This basic level is independent of the orientation of the trpA gene relative to PL (Figs. 5 and 6). The heat induct ion of PL raises the basic level to 1% of tota l cell protein as trpA protein in the case of the plasmid where the trpA gene is oriented in the same direction as transcription from PL (pHUBtrpDCBA(SaII)). When trpA faces toward PL, the

120

basic level (p2~iependent) of transcription appears to be reduced after heat induct ion of PL (Fig. 6),

(e) Cloning o f tl~.~ clts gene Due to the desired r .for a given gene i n - - i n p ~ i d s pHUB2

and pHUB4 to be control led by a v I t s gene these cloning vehicles were main-

I00

• 8 0

", 6 0

< 4 0 Q, I , .

20

0 I l , , , • *

60 120 180 240 300 TiME ( rain )

Fig. 4. Expression of the trpA gene on ~he plasmid pHUB2trpA(EcoRi) in the host K-12~IIA trp£A after heat induction. Conditions for heat induction and the assay of trpA activity are described in MATERIALS AND METHODS.

70

71

rained in the host K-12AHIA trpEA. This strain is deleted for the trp operon and carries the t i t s gene on the chromosome. To overcome the requirement for this host for use of these plasmids, attempts were made to clone the cIts gene either on a separate plasmid that is fully compatible, and therefore, could coexist in the same cell, with pHUB2 and pHUB4, or on the plasrnid cloning vehicle itself.

The BgIII fragment extending from 74--78% of the genetic map of phage contains the ¢I gene as well as Pa with era being cut by BglII (KamF et al., 1977; Schwarz et al., 1978). pRK248 is a mini-plasmid (8 kb) that was derived from the P group plasmid ttK2 (Meyer and Helinski, mai~uscript in preparation) and confers resistance to tetracycline. To clone the BglII fragment containing the t i ts gene into the single BglII site of pttK248, DNA preparations of pRK248 and pER482 were cleaved with BglII, mixed, ligated and transformed into C600trpA33. Tet It transformants were screened for immunity to ).clear at 30°C and loss of immunity at 42°C. A clone that satisfied these criteria was selected and transformed for Kan R with pHUB2trpA(EcottI). Table I shows that heat induction of PL results in a level of expression of trpA in this system similar to that of the expression of trpA in K-12~H1~ trpEA. Analysis of the plasmid

50

• S 4 0 .?

" I¢

- - O l

2

o 3'o 6'0 9'0 ,~o TIME ( min )

Fig. 6. Expression of the trpA gene on the plasmid pHUB4trpDCBA(Sall)I and pHUB4trpDCBA(SaII)2 in the host K-12~IIAtrpEA. Conditions for heat induction and the assay of trpA activity are described in MATERIALS AND METHODS.

Fig. 5. Agarose gel electrophcre~is of pHUB4trpDCBA(SalI) DNA cleaved restriction enzymes. The agarose gels contained 0.8% of agarose.

(a) Soft digest of bacteriophage ~80 trpE-A DNA. An arrow indicates the band containing the trpDCBA region (C. Yanofsky, unpublished).

(b) SalI digest of pHUB4. (c) and (e) SalI digests of two pHUB4trpDCBA(SoII) hybrid plasmids that differ in

orientation of the fragments. The two orientations of the SalI inserts relative to pHUB4 could be distinguished by a HpaI digest (d) and (f). 'A~he plasmid DNA analyzed in slots (e) and (d) contains the trp genes in the same orientation as PL and is desiguated pHUB4trpDCBA(SaII)I. The plasmid with the opposite orientation is designated pHUB4 trpDCBA ( $aII)2.

72

DNA isolated from this strain showed the coexistence of the two plasmids. One of the plasmids exhibited the molecular weight of pHUB2trpA(EcoRI). The other plasmid, which was present at about a ten-fold lower copy number, was cleaved by BglII into two fragments; one of the ~ e n t s showed a molecular weight identical to pRK248 while the other fragment exhibited a molecular weight corresponding to the BglH fragment of pER482 that carries cIts (data not shown),

In addition, the BglII fragment carrying the clts gene was inserted into the BglII Ate of a plasmid that is identical to pHUB2trpA(EcoRl)except that it lacks the kan a gene (pHUB12~A(£coRI), map not shown) and trpA expression was measured after heat induction of C600trpA33 carrying this plasmid. As shown in Table I the level of trpA protein produced by this hybrid plasmid is similar to that of PL promoted expression of the trpA gene in the K-12AHIA trpEA strain.

DISCUSSION

This study was undertaken for the purpose of constructing a cloning vehicle in which the combination of the high copy number of the plasmid and ihe strength and ability to regulate the leftward promoter PL of bacteriophage would provide a controlled high rate of transcription of genes on inserted DNA fragments. Transcription initiation at )~ PL is dependent of several factors. In the lysogenic state the product of gene cI of ~. binds to the leftward operator leaving PL inaccessible to RNA polymerase (Maniatis et al., 1975). In addition, PL is subject to repression by the product of the cro gene (Kumar et al., 1970; Folkmanis et al., 1976). The plasmids constructed in this study do not produce a functional cro gene product either because of inactivating mutations or because the cro gene was deleted. Use of a temperature-sensitive cI gene allows for the controlled activation of transcription at PL by heat induction. Besides the availability of a switch for promoting gene expression, it was considered that the properties of the N gene product of phage ~ would provide an additional advantage to the cloning vehicle. N protein modified RNA polymerase activated at the nutL site, is insensitive to many transcription termination signals (Fravklin and Yanofsky, 1976; Epp and Pearson, 1976; Adhya et al., 1976; Salstrom and Szybalski, 1978a, b) so that transcription starting at p,. could continue into an inserted restriction fragment downstream from p~.

)~trp phages, constructed in vivo by illegitimate recombination (Franklin, 1971; Moir and Brammar, 1976) and in vitro recombinant DNA techniques (Hovkins et al.. 1976), have been shown to express trp genes in vivo by transcription from PL- To determine whether or not PL functions after insertion into a plasmid, restriction fr.agments of two different ),trp phages were cloned into two different plasmids, ~ trpB expre~on was measured as a test for t h e f u n ~ o n of p,. i ~ e following gene~observations were made with the p~mid~ pER481, pEP,4$2 and p ~ 2 0 0 4 ~ 4 3 t r p A

73

pMK2004~trp48trpA. At 32°C trpB is weakly expressed, probably as a result of the activity of the trp operon internal promoter P2. This basic level of expression in independent of the host or the vector, occurs in the derivatives of both ~trp phages, and is also independent of the orientation of the ~,trp EcoRI fragment relative to the vector pCRI. At 41~C trpB expression is enhanced by a factor of 6 presumably due to initiation from pw.. It is perhaps not too surprising that activation of the pT. promoter increased the trpB level only by a factor of 6. It is possible that this level of expression is limited by the overall metabolic capacity of the cell. Thus, the level of trpB gene product alone in these induced cells represents 5% of the total cell protein. Since the N gene, trpD and C are all presumably transcribed and translated at near equal rates after heat induction, as much as 20% of the total cell protein could be the products of transcription initiated at px.. These considerations suggest that PL functions efficiently on the plasmid vehicles.

An important question in the use of the N operon of phage ~ to construct a cloning vehicle is the influence of the N protein on the p~ dependent transcription of inserted genes. To resolve this question, the expression of trpB, which is fused to PL, and the simultaneous expression of trpA on an EcoRI insert downstream of pT., and trpB, were measured. In phage ktrp48 trpB expression is not dependent on the presence of a functional N protein since the fusion of the trp operon with the N operon has deleted ~v.. trpB expression is strongly dependent on the function of the N protein in ~trp43, where tT. terminates transcription from pT. unless transcription is modified by the N functic~r~ (Franklin, 1974). Since trpB expression is similar in pMK2004~trp43trpA ~'ld pMK2004~,trp48trpA, the level of N protein specified by these plasmids must be sufficient to overcome the effect of tL. On the other hand, the trpA gene in the EccRI insert is expressed at only about 10% the rate of trpB expression in strains carrying the pMK2004~trp43trpA or pMK2004~.trp48trpA plasmids. It is likely that some particular sequence between the fused trpB gene and the inserted trpA gene prevents RNA polymerase from reading into the insert with high frequency.

In general the experiments reported here do not provide clear evidence for the usefulness of having the N gene present in these cloning vehicles. The observations with pMK2004~trp48trpA and pMK2004;~trp43trpA indicate that the presence of the N gene in the pMK2004~trp43trpA plasmid did not enhance the expression of the trpA gene insert. In addition, the threefold difference between the trpA activity made from pHUB2trpCBA(BamHI) (which does not produce a functioning N protein) and pHUB4trpCBA(BamHI) (which has retained the N gene) cannot be interpreted simply in terms of N function since the DNA sequence between pT. and the insert differs in these plasmids.

In the hope that a reduction of the large distance between pT. and the site of insertion of a gene would enhance the expression of this gene, plasmids pHUB2 and pHUB4 were constructed, pHUB2 contains four different single zestriction sites available for insertion downstream from PL. pHUB4 has a

74

HpaI, a BamHI and a SoJI site avzilable for cloning with the entire h s e ~ e n t downstream from PL derived from the N operon of phage h. It was found that the expression of gen~ inserted into the EcoRI site of pHUB2, into the BamHl site of pHUB2 and pHUB4, and into the SaiI site of pHUB4, is p,.- dependent, with I to 6.6% of the total cell protein specified by the trpA gene i n~r ted into one of these three sites. The m ~ levels of the trpA gene product were reproducible in several independent expression experiments and seemed to be dependent of the particular plasmid and the particular restriction site used for in~rtion. It was not possible by prolonged incubation at 41°C to raise the lew~l of trpA gene product higher than the levels shown in Table I. One possible explanation for this observation is that transcription from p~ may interfere with replication of the plasmid. Observations providing indirect support for this possibility are the findings that plasmids carrying PL but not cI could not be transformed into non-lysogens, and that cells carrying piasmids with PL under the control of a temperature sensitive repressor were cured of the plasmids when they were incubated at 41°C continuously.

It is reasonable to assume that genes inserted into the HpaI restriction site upstream from the three sites used in this study will be transcribed with similar efficiency. Therefore, these vehicles should be useful for controlled expression of prokaryotic genes that are cloned by blunt end tigation at the HpaI site or by insertion into the EcoRI, BamHI and SalI sites. Although the plasmids described in this report are potentially very useful for enhancing the level of transcription of a gene inserted at any one of these sites, it is likely that for genes from higher eukaryotes additional nucleotide sequence information will have to be built into these vehicles to allow high efficiency translation of the mRNA specified by these foreign DNA inserts.

ACKNOWLEDGEMENTS

This work was supported by grants from the Nati~ : Institutes of Health and the National Science Foundation. H.U.B. is the recipient of a postdoctoral fellowship of the Deutsche Forschungsgemeinschaft. E.R. is the recipient of a European Molecular Biology Organization (EMBO) long-term postdoctoral fellowship, a NATO research fellowship and a Fulbright/Hays Travel Grant. H.K.D. is supported by the Indo-.4merican Fellowship Program. We want to thank Drs. S. Mills and M. Kahn for helpful advice and Virginia Horn and Magda van Cleemput for their expert technical assistance.

REFERENCES

Adhya, S., Gottesman, M., deCrombrugghe, B. and Court, D., Transcription termination regulates gene expression, in Losick, R. and Chamberlin, M. (Eds.), RNA Polymerase, Cold Spring Harbor Laboratory, Cold Spring Harbor, New York, 1976, pp. 719-730.

AIIc, ~, B., Jeppensen, P.G.N., Karag/ri, K.Y. and Delius, H., Mapping the DNA fragments produced by cleavage of ~, DNA with endonuclease RI, Nature, 241 (19"/3)199-123.

Allet, B. and Bukhari, A.L, Analysis of bacteriophage Mu and ~-Mu hybrid DNAs by specific endonucleases, J. MoL Biol., 92 (1975) 59.9-540.

75

Armstrong, K.A., Hershfield, V. and Helinski, D.R., Gene cloning and containment properties of plasmid ColE1 and its derivatives, Science, 196 (1972) 172--174.

Bolivar, F., Rodriguez, R.L., Greene, P.Y., Betlaeh, M.C., Heyneeker, H.L., Boyer, H.W., Crosa, Y.H. and Falkow, S., Construction and characterization of new cloning vehicles, II. A multiple cloning system, Gene, 2 (1977) 95--113.

Castallazzi, M., Brachet, P. and Eisen, H., Isolation and characterization of deletions in bacteriophage k residing as prophage in E. coli K-12, Mol. Gen. Genet., 117 (!972) 211--218.

Chang, A.C.Y., I~n~man, R.A., Clayton, D.A. and Cohen, S.N., Studies of mouse mitochondrial DNA in Escherichia coli: Structure and functio, of the eucaryotic- procaryotie chimeric plasmids, Cell, 6 (1975) 231--244.

Cohen, S.N., Chang, A~C.Y. and Hsu, C.L., Nonchromosomal antibiotic resistance in bacteria: Genetic transformation of Escherichia coli by R-factor DNA, Proe. Natl. Acad. Sci. USA, 69 (1972) 2110--2114.

Covey, C., Richardson, D. and Carbon, J., A method for the deletion of restriction sites in bacterial plasmid DNA, Mol. Gen. Genet., 145 (1976) 155--158.

Creighton, T.E. and Yanofsky, C., Chorismate to tryptophan (E. coU), in Colowick, S.P. and Kaplan, N.O. (EdL), Methods in Enzymology, Col. XVIIA, Academic Press, New York, 1970, pp. 365--380.

El)p, C. and Pearson, M.L., Association of bacteriophage lambda N gene protein with E. coil RNA polymerase, in Losiek, R. and Chamberlin, M. (Eds.), RNA Polymerase, Cold Spring Harbor Laboratory, Cold Spring Harbor, New York, 1976, ~ 667--691.

Folkmanis, A., Takeda, Y., Simuth, J., Gussin, G. and Echols, H., Purification and properties of a DNA-binding protein with characteristics expected for thz Cro protein of bacteriophage X, Proe. Natl. Acad. Sci. USA 73, (1976) 2249--2253.

Franklin, N.C., The N operon of lambda: Extent and regulation as observed in fusions of the tryptophan operon of £scherichia coli, in Hershey, A.D. (Ed.), the Bacteriophage Lambda, Cold Spring Harbor Laboratory, New York, 1971, pp. 621--638.

Franklin, N.C., Altered reading of genetic signals fused to the N-operon of bacteriophage x: Genetic evidence for modification of polymerase by the protein product of the N gene, J. Mol. Biol., 89. (1974) 33--48.

lZmaklin, N. and YanoC-ky, C., The N protein of x: Evidence bearing on transcription termination, polarity and the alteration of E. coil RNA Dolymerase, in Losick, R. and Chamberlin, M. (Eds.), RNA Polymerase, Cold Spring Harbor Laboratory, New York, 1976, pp. 693--706.

Greene, P.J., Heynecker, H.L., Bolivar, F., Rodriguez, R.L., Betlach, M.C., Covarrubias, A.A., Backman, K., Russel, D.J., Talt, R. and Boyer, H.W. A general method for the purification of restriction enzymes, Nucleic Acids Res., 5 (1978) 2373--2376.

Hedgpeth, J., Ballivet, M. and Eisen, H., Lambda phage promoter used to enhance expression of a plasmid-eloned gene, Mol. Gen. Genet., 163 (1978) 197--203.

Hershfield, V., Boyer, H.W., Yanofsky, C., Lovett, M. and Helinski, D.R., Plasmid ColE1 as a molecular vehicle for cloning and amplification of DNA, Proe. Natl. Acad. Sci. USA, 71 (1974) 3455--3459.

Hopkins, A.S., Murray, N.E. and Brammar, W.G., Characterization of x trp-transducing bacteriophages made in vitro. J. Mol. Biol., 107 (1976) 549--569.

Jackson, E.N. and Yanofsky, C., Internal promoter of the tryptophan ope:on of £scherichia coil is located in structural gene, J. Mol. BioL, 69 (1972) 307--313.

Kamp, D., Kahmann, R., Zipser, D. and Roberts, R.J., Mapping of restriction sites in the attachment site region of bacteriophage lambda. Mol. Gen. Genet., 154 (1977) 231--248.

Katz, L., Kingsbury, D.K. and Helinski, D.R., Stimulation by cyclic adenosine mono- phosphate of plasmid deoxyribonucleic acid replication and catabolic repression of the plasmid deoxyribonucleic acid protein relaxation complex, J. Bacteriol., 114 (1973) 577--591.

Kumar, S., Calef, E. and Szybalski, W., Regulation of the transcription of Escherichia coli phage ~. by its early genes N and tof, Cold Spring Harbor Symp. Biol., 35 (1970) 331--340.

76

Lowry, O.H., Rosebrough, N.G., Farr, ,4.1. and Randall, R.G., Protein meseuremenf; with the folin phenol reagent, J. BIoL Chem., 198 (1961) 265--276.

Maniatls, T., Pteshne, M., Backmann, K., Kleid, D., Fleshman, S., Jeffrey, A. and Mauer, R., Recognition sequences of repressor and polymerase in the operators of bacterio- phage Z-mbda, Cell, 5 (1975) 109--113.

Meyer, It., Figurski, D. and Helinski, D.R., Physicad.and genetic studies with restriction endonuclesses on the broad host/range plasmid~tK2, Mol. Gen. Ganet., 152 (197"/) 129-135.

Miner, D.I~, Gubbins, F_~G., Pagg, E.W. and Done]son, G.E., Trenseription and translation of cloned Drosophila DNA fraIpnents in £scherfchia eoli, Biochemistry, 16 (1977) 1031--1038.

Molt, A. and Brammar, W.J., The use of specialized tr~_nnducing phages in the amplification of enzyme production, MoL Gen. Genet., 149 (1976) 87-~99.

Morrow, J.F., Cohen, S.N., Chang, A.C.Y., Boyer, H.W., Goodman, H.M. and Hemng, R.B., Replication and transcription of enkaryotie DNA in £scherichia coli, Proe. Natl. Acad. SoL USA, 71 (1974) 1743--1747.

Morse, D.E. and Yanofsky, C., The internal low efficiency promoter of the tryptophan operon of£. ¢oli, J. MoL BinL, 38 (1968) 447--459.

Ratzkin, B. and Carbon, J., Functional expression of cloned yeast DNA in £scherichia coli, Proc. NatL Aead. SoL USA, 74 (1977) 5041--5045.

Roberts, J.W., The rho factor: Termination and anti-termination in Imnbda, Cold Spring Harbor Syrup. Quant. Biol., 35 (1971) 10.1--126.

Salstrom, J.S. and Szybalski, W., Tranm~iptlon termination sites in the major leftward operon of coliphage lambda, Virology, 88 (1978a) 9.52--9.60.

Salstrom, J.S. and Szybaisk/, W., Coliphage XnutL-: A unique class of mutants defective in the site of N utilization for antitermln#tion of leftward transcription, J. Mol. Biol., 124 (1978b) 195-221.

Schwarz, E., Seherer, G., Hobom, G. and Kocuel, H., Nueleotide sequence of ero, eH and part of the O gene in phage ~, DNA, Nature, 0.72 (1978) 410--414.

Selker, E., Brown, K. and Yanofsky, C., Mitomyein-C-indueed expression of frpA of Salmonella typhimurium inserted into the plasmid ColEI, J. Bacteriol, 10.9 (1977) 388--394.

Sgaramella, V., Bursztyn-Pettegrew, M. and Ehrlieh, S.D., Use of T4 ligase to join flush ended DNA segments, in Beers, R.F, and Bsssett, E.G. (Eds. }, Recombinant Molecules: Impact on Science and Society, 10th Miles Inst. Syrup., Raven Press, New York, 1977, pp. 57--68.

Smith, D.L., Blattner, F.R. and Davies, J., The isolation and partial characterization of a new restriction endonueleese from Providenc~z st~arffi, Nucleic Acids Res., 3 (1976) 343--353.

Smith, O.H. and Yanofsky, C., Enzymes involved in the biosynthesis of tryptophan, in Colowiek, S.P. and Kapl~n~ N.O. (Ed&), Methods in Enzymology, Vol. V, Academic Press, New York, 196o-, pp. 794--806.

Struhl, K. and Davis, R.W., Production of a functional eucaryotic enzyme in £seherichia toll: Cloning and expression of the yeast structural gene for fmidazole-glyeerolphosphate dehydratase (his 3), Proe. Natl. Acad. Sei. USA, 74 (1977) 60.65--50.69.

Takeda, Y., Oyama, Y., NJk~yima, I~ and Yura, R., Role of host RNA polymerese for ~. phage development, Biochem. Biophys. Res. Commun., 36 (1969) 533--538.

Taylor, K., Hradecna, Z. and Szybakki, W., Asymmetric distribution of the transcribing regions on the complementary strands of the eoliphage X DNA, Proe. Natl. Acad. Sei. USA, 57 (1967) 1618--1625.

Thomas, M. and Davis, ILW., Studies on the cleavage of bacteriophage distribution lambda DNA with EcoRI restriction endonucleese, J. MoL Biol., 91 (1975) 315-'378.

Vapnek, D., Hautala, J.A., Jacobson, J.W., Giles, N.H. and Knshner, S.R., Expression in Escherichia coli K-19 of the structural gene for catabolic dehydroquinese of Neurospora crass~ Proc. Natl. Aead. SeL USA, 74 (1977) 3508--3572.

Communicated by W. Szybalski.