Transcriptional antitermination activity of the synthetic nut elements of coliphage lambda I. Assembly of thenutR recognition site from boxA and nut core elements

El25

Gene. 42 (1986) E125-El32 Elsevier

ERRATUM This paper has been publishedin Gene, 39(1985) 121427; however,a numberof importantcorrections bad not been executed. We apologize to the authors for this oversight. GENE

1445

Transcriptional

antitermination

activity of the synthetic nut elements of coliphage lambda

I. Assembiy of the n&R recognition site from box_4 and nut core elements (Recombinant

DNA;

terminator;

promoter;

control

modules;

plasmid;

cloning

vector;

galactokinase)

Alexandra L. Brown* and Waclaw Szybalski** McArdk

Lahormr:~~for

(Received

May 12th, 1985)

(Accepted

August

Cmcer Research, University of Wisconsin, Mcrdison, WI 53706 (U.S.A.)

Tel. (608) 262-1259

I&h, 1985)

SUMMARY

An active nutR antiterminator was reconstructed from two synthetic modules, o?e containing the 8-bp boxA (5’-CGCTCTTA) and the other the 17-bp nutR core (5’-AGCCCTGAAAAAGGGCA) sequence. The modules were synthesized with Hind111 cohesive ends, which upon annealing and ligation created an 8-bp spacer (5’-CAAAGCTT) between the boxA and nutR core. The 8-bp length was the same as in the native nutR (5’-CACATTCC), but the sequence showed less than 387; homology. The antitermination mediated by the synthetic nutR was 68-80% efficient when tested in the p,-nutR-N-t,,-galK expression plasmid, analogous to that used by Drahos and Szybalski [Gene, 16 (1981) 261-2741. The cloned boxA by itself has no activity, while the nutR core alone shows only marginal (5-10%) antiterminator function. Increasing the distance between boxA and the nutR core from 8 bp to 20-28 bp, i.e., by one to two turns of the DNA helix (about 10 bp per turn), has little effect on the antiterminator function, whereas use of spacers with length about halfway between 8 and 20 bp results in reduced antitermination. It appears that both the sequences and spatial arrangement of the boxA and nut elements are important for efficient antiterminator function.

Transcription of an operon is initiated at the promoter and is discontinued at the terminator signal. Bacteriophage 1, has developed specific antitermi-

nation mechanisms designed to overcome the termination signals. When a special sequence, nut, is present between the promoter and terminator, and the trans-acting product of gene N is supplied together with the host factors, transcription can proceed

* Present

nal alkaline

INTRODUCTION

Bethesda,

address:

Bldg.

10, Rm.

MD 20892 (U.S.A.)

** To whom addressed.

correspondence

8D14,

NIADDK,

NIH,

threitol;

Tel. (301) 496-2483. and reprint

requests

should

be

pair(s);

buffers:

MATERIALS

Ap, ampicillin; HSR,

BB, bromphenol

ligase,

AND METHODS.

0378-l 119/X6/$03.50 0

LS,

LSR,

blue; bp, base

MG5-C,

MSR,

see

section a: CIAP, calf intesti-

1986 Elsevier Science Publishers B.V. (Biomedical

ethidium

L broth;

~~ac-Cal-anlp

THODS,

section

tide: Abbreviations:

phosphatase;

EtdBr,

DNA

bromide; plates,

GalK,

polymcrasc

Pollk,

AND

LB, ME-

oligo, oligodeoxynucleo-

Klenow

I: ss. single-stranded:

citrate,

pH 7.6; u, unit(s);

ture; UV, ultraviolet;

XC, xylene cyanol.

Division)

DTT, dithio-

galactokinase;

see MATERIALS

a; nt, nucleotide(s);

PA, polyacrylamide;

0.015 M Na,

Cm. chloramphenicol:

fragment SSC.

T,,

of E. coli

0.15 M NnCI.

melting

tempera-

across the terminators

with efftciencies

(for review see Friedman

up to 100%

and Gottesman,

1983). The

rzzdtL element was geneticaliy defined by point mutations (Salstrom and Szybalski, 1976; 1978) and shown to function placed between

as an autonomous

any promoter

module, when

and terminator

(Dra-

hos and Szybalski, 198 1). Originally, it was believed that the entire nut element corresponded to a 17-bp sequence

AGCCCTGAATAmGCA

shared by

nutL and nufR, the leftward and rightward nators

of /z located

promoters,

downstream

respectively

antitermi-

of the pt. and pK

(see Daniels

et al., 1983).

However, as pointed out by Olson et al. (1982) additional sequences named boxA and boxC are located upstream and downstream of the 17-bp nut core (also called boxB), both in the nutL and nutR elements. Drahos and Szybalski (1981) and Drahos et al. (1982) have shown that the boxA-less, 17-bp IZUIL core, natural or synthetic, could function as an antiterminator module, but at reduced effrciency (especially at elevated temperatures; Peltz et al., 1985). On the other hand, the boxA -containing nutL or n&R fragments show efficient and temperatureresistant antitermination activity (de~rombru~he et al., 1979; Peltz et al., 1985; Luk and Szybalski, 1982). The purpose of our study was to determine the role of both the boxA and 17-bp nut core (boxB) sequences in the N-mediated antitermination process. We have found that the synthetic boxA and boxB elements could be assembled into a functional nutR site, with its activity approaching that of the natural nutR antiterminator. Preliminary results of this study were presented by Brown and Szybalski (1984) at the Bacteriophage Meeting, Cold Spring Harbor, in 1984.

(MSR)

buffer

(a) Enzymes,

AND

50 mM

NaCI,

1 mM DTT;

10 mM

Tris

high salt re-

striction (HSR) buffer was 100 mM NaCI, 50 mM Tris (pH 7.5), 10 mM MgCl,, 1 mM DTT. XC/BB buffer is 10 mM Tris (pH 7.9)

1 mM EDTA,

50”,,

glycerol, 0.002% (w/v) BB and 0.002”/, (w/v) XC. MGS-C buffer was 50 mM Tris (pH 7.6), 5 mM DTT,

100 ;lM spermidine,

100 /irM EDTA.

Elution

buffer was 10 mM Tris pH 7.6, 200 mM NaCt. LS buffer was 10 mM Tris pH 8.0, 1 mM EDTA. Ligase buffer was 20 mM Tris HCl (pH 7.6), 10 mM 10 mM MgClz, 1 mM DTT, 1 mM Na,EDTA, ATP. Mac-gal-amp indicator plates were described by Drahos

and Szybalski

(1981). LB was I :fh Tryp-

tone, 0.5:;) yeast extract, 1;; NaCl (pH 7.5). M9 medium was (per 100 ml) 6 g Na2HP0,, 3g KH,PO,, 0.5 g NaCl, 1 g NH,Cl. M9b medium was M9 containing 1 mM MgSO,, 4 (Ig/ml thiamine, 0.5”b glucose, 0.2(& Casamino acids (Difco). (b) Oligodeoxynucleotide

synthesis

Synthesis of oligos was by the manual phosporamidite method described by Adams et al. (1983), which uses a glass-bead support. Reactions were carried out in a 15-ml sintered glass funnel (medium) set over a one-liter filter flask. After each chemical reaction, reagents and solvents were removed by filtration through the sintered glass, which retained the oligo-bearing glass-bead support. Three-step deblocking of the DNA with thiophenol and N&OH was done as described by Matteuci and Caruthers (198 1). The dried residue from the heattreated NH,OH supernatant was resuspended in 10 mM Tris (pH 7.6). Typical yields of crude 21mers, as determined by absorbance at 260 nm, were 5 10 mg, starting from 64 mg of derivatized support. (c) Purification

MATERIALS

was

(pH 7.5), 10 mM MgCl,,

of synthetic

oligodeoxynucleotides

METHODS

buffers and media

All enzymes and buffers were as described previously by Drahos and Szybalski (1981) or Drahos et al. (1982), except those described below. CIAP was obtained from Boehringer Mannheim. Low salt restriction (LSR) buffer was 10 mM Tris (pH 7.5), 10 mM MgCl,, 1 mM DTT; medium salt restriction

Approx. 1.5 mg (.50- 100 ~1) of crude product was mixed with 30 141 of XCjBB buffer and electrophoresed on a 1.5 mm x 260 mm 20% PA-7 M urea slab gel at 300 V. The gel was stained with 1.5 jig EtdBr/ml for 10 min, and the bands were visualized using a UV lamp. The DNA-~ontaitling band was cut into small pieces using a razor blade, placed in a 1.9-ml eppendorf tube and allowed to sit in elution buffer overnight. The DNA-containing

El27

buffer was recovered wool

by filtration

plug in a l-ml

siliconized

eppendorf

corex centrifuge

through pipet

a glass-

tip into

tube. Centrifugation

(g) Galactokinase

assays

a for

Galactokinase

assays were as described

1 min at 4000 rev./min allowed the DNA-containing buffer to pass through the glass wool while the gel

hos and Szybalski tose was diluted

pieces were retained. The DNA was ethanol-precipitated and taken up in 100 ~1 of 10 mM Tris (pH 7.6).

0.1 mCi/mmol.

Typical

(h) Large-scale

yields for each batch

of purified

ss DNA

by Dra-

(1981), except that [ “C]galacto a final specific activity of

plasmid preparations

were 100-500 pg. E. coli C600 guZK

(d) Kinasing the 5’-hydroxyl and hybridization

termini of the ss DNA

to form oligodeoxynucleotide

du-

The 5’-phosphate groups were added to 5 ng of ss DNA using 20 u of T4 polynucleotide kinase in 20 ~1 of MGS-C buffer and 100 PM ATP with incubation at 37°C for 1 h. The kinased ss DNA was hybridized without intermediate purification by heating at 90’ C for 10 min in 300 ,ul of 6 x SSC and incubating overnight at 5-10” C below its T, . After cooling slowly to room temperature, the DNA was dialyzed against 3 x 250 ml of LS buffer, ethanolprecipitated, and taken up in 20 ~1 of 10 mM Tris (pH 7.6). of plasmids and transformation

HindIII-digested plasmid DNA (5 pg) was heated to 65’ C for 10 min, and phosphatased by diluting to 300 ~1 with 50 mM Tris (pH 8.0) and incubating at 37’ C for 1 h with 0.6 u of CIAP. After ethanol preci-

the plasmids

were grown up to an A575nm of 0.5 at 30°C in M9 medium. Plasmids were amplified by addition of 200 pg Cm/ml with continued

plexes

(e) Construction

cells carrying

shaking

overnight

at

30” C. Each liter of overnight culture was pelleted by centrifugation, and the cells were taken up in 13 ml of lysozyme buffer containing 5 mg/ml lysozyme. After 15 min at room temperature, 6 ml of 0.5 M EDTA (pH 8.1) was added and the mixture was allowed to sit on ice for 30 min. Next, 10 ml of Triton mixture (20/, Triton X-100 in 5 mM Tris, pH 8.1, 50 mM EDTA, pH 8.1) was added to each batch, and again the mixture was allowed to sit for 30 min. Cell debris was removed by centrifugation and the supernatant deproteinized by phenol-chloroform extraction. The plasmid DNA was precipitated using isopropanol, taken up in a small volume of 10 mM Tris (pH 7.6) and further purified by centrifugation in a CsCl-EtdBr density gradient for 2 days. Plasmid DNA was collected, and the EtdBr and salts were removed by dialysis against 10 mM Tris (pH 7.6), 1 mM EDTA

(pH 8.1).

pitation, the DNA was taken up in 10 ~1 of 10 mM Tris (pH 7.6). Ligation of 1 c(g of synthetic DNA

(i) Other procedures

with 2 llg of restricted, phosphatased plasmid DNA was accomplished by incubation with 10 u of T4 DNA ligase in 20 ~11of ligase buffer at 15 “C overnight. Competent Escherichiu coli C600 galK cells were transformed by the procedure used by Drahos and Szybalski (198 1).

PA gel electrophoresis, sequencing of DNA, and enzymatic restriction were as described previously (Drahos and Szybalski, 1981; Drahos et al., 1982) except that enzymatic restrictions were performed in either LSR, MSR or HSR buffers (see MATERIALS AND

(f) Restriction

METHODS,

SeCtiOn a).

analysis (j) Bacterial strains and plasmids

Rapid plasmid analysis was carried out according to the method of Birnboim and Doly (1979) as moditied by Ish-Horowitz and Burke (198 l), except that an ether extraction step was added following the phenol-chloroform extraction.

E. coli C600 gulK -- (Rho +, Nus’, ~~11’) was used as host for all transformations. Plasmids pK03, pD553 and pNPl09 were described by Drahos and Szybalski (1981).

El28

Mismatching permitted creating both wild-type and mutant sequences (to be employed in future studies).

RESULTS

(a) OIig~eoxynucleotide

synthesis and puri~cation

The crude preparations were purified by electrophoresis on 20% PA/7 M urea gels with yields of 40 pg (BP) to 184 pg (BG). Phosphate groups were

The boxA and nutR core (bad) sequences were synthesized as described in MATERIALS AND METHODS, sections

b and c, and sequenced.

were about 5 mg of crude preparation The boxA sequence was synthesized having Hind111 cohesive

added

The yields

THODS,

(b) DNA duplex formation

ends (underlined).

AND

ME-

and plasmid tonstrue-

tions

Eleven

Using the formula:

5’-AGCTTTGTAAGAGCG-3’ T,

The nutR core sequence was synthesized as two 21-mers having Hind111 cohesive ends (underlined). 16 nucleotides were complementary and one in the BG strand was mismatched (overlined): BP strand: 5’-AGC~AGCCCTGAAAAAGGGC-3’ BG strand: 5’-AGCTGCCCATTTTCAGGGCTA-3’

r-

(see MATERIALS

section d).

of each oligo. as two 1%mers

nucleotides were complementary: AG strand: 5’-AGCTCGCTCTTACAA-3’ AS strand:

to the 5’ ends

(Davis AGjAS duplex ry l”/ 1.5”C

boxA

pD553An5 h 1 1

= 16.6 log[Na”] + 0.41[7;(G + C)] + 8 1.5 - SOO/number of bp in short duplex et al., 1980), we calculated the T,‘s for the duplex (11 bp; 54.6” C) and for the BP/BG (4 + 12 bp, one A-A mismatch; 65’ C). Evelowering of the homology lowers the T, by (Davis et al., 1980). The AG + AS and

box6

n

n

4 . --*--____ _m.___^----.___ ___1 I T

1 1

f

t

i

8, 12, XI, 24, 28 bp

z

I

I 1

I i

1 t

cc

boxi

pO5538 I I

t t

1 i

I I

_--_-.--

boxA

-L_._ I_

pU553A

E I

PP

10

Hf 1

tl lid I I 5

282

H tif I 14

64

-%

t

pD553

,,PP R

Lf

ori carrying synthetic Fig. I. Plasmids and Szybalski

(I98l),permits

and the Rho-dependent respcctivcly).

ho.uA and’or ho.vB components

measuring

I,, terminator.

The plasmid

Cloningofsynthetic

with the inverted

20, 24. 28 bp) are shown in Fig. 2. E, EcoRI; sites are shown);

of the rttrrR antiterminator.

the activity ofthe nut antiterminator

P. PsfI. Sizes of several

boxA or boxi3 is described

boxA (pD553Ai)

in RESULTS,

described

by Drahos

the constitutivep~

promoter

section b (see pD553A and pD553I3,

is not shown. The boxA-bo.uB regions of plasmids

H, HuPIII (only the pertinent fragments

The pD553 plasmid.

modules when cloned between

(xx pD553)

sites are shown);

are specified

Hd, NindIII;

in bp. The drawing

pD553AnB

(n = 8, 12.

Hf, Hi&l (only the pertinent

is not to scale.

El29

BP + BG

oligos

were

6 x SSC at 5-10°C dialyzed

against

hybridized

below

overnight

their calculated

LS buffer

to remove

pD553 was enlarged

in

boxB insertion,

T,‘s,

salts,

pD553 g&K expres-

and phosphatased

sion plasmid

(Drahos

and Szybalski,

respectively

(Fig. 1). The orientation

of the inserts was determined by Hi&I + Hind111 digestion, as described in the legend to Fig. 1.

and

ethanol-precipitated. The boxA and boxB duplexes were ligated into the HindIII-cut

to 84 bp or 90 bp by boxA or

(d) Nucleotide

1981), giving

rise to a series of pD553A and pD553B recombinant plasmids, respectively (Fig. 1). Since the plasmids of

sequencing of constructs

the pD553B series carried both wild-type and mutant

For sequencing, the plasmids were prepared as describedin MATERIALS AND METHODS, section h.

n&R core sequences, they were re-transformed to ensure that the cells carried only one kind of plasmid.

fractionated

Construction

of plasmids

They were cut with Hinfl, treated

both boxA and

carrying

plates (see MATERIALS a and e).

AND

and

396-bp fragments were electroeluted, ethanol-precipitated, labeled with [ y-32P]ATP employing polynucleotide kinase, cut with HhaI, and fractionated on 1.5-mm 5% PA gels. The approx. 320-bp fragment was eluted, ethanol-precipitated, and sequenced by

boxB sequences (pD553AB series) is described below in section e. All plasmids were transformed into E. co/i C600 galK - and transformants selected on Mac-gal-amp sections

with CIAP,

on 3 mm 5% PA gels. The 380-bp and

METH-

All recombinant plasmids were screened for the presence of boxA and boxB inserts by restriction

the Maxam and Gilbert (1980) procedure using 8 y0 PA/7 M urea gels. As illustrated in Fig. 1 and Table I, plasmids pD553A and pD553Ai contain the wild-type boxA sequence and plasmid pD553B contains the nutR core (boxB) sequence, all in the predicted orienta-

with Hue111 in MSR buffer and electrophoresis 5% PA gels. The 69-bp (5 + 64) HaeIII fragment

tion, as determined are not shown).

ODS,

(c) Restriction

TABLE

analysis

in of

by restriction

analysis

I

Antitermination

function

of the synthetic

boxA, boxB and nurR (boxA + boxB) elements

Plasmid”

Constructb

GalK

(1)

(2)

(3)

_ __L -. p-bo.rrl-Xbp-boxB-r

156 k 10

I.

pD553A8B

2.

pD553A

3.

activity’

“0 Antiterminationd (4) 68-80

24+

2

0

pD553Ai

p&z-t p-Z”sA-t

25*

3

0

4.

pD553B

p-bo.rB-t

46+

4

5.

pD553

P-1

332

5

6.

pNP109

7.

pK03

P -

J Plasmids

l-5 are described

h Constructions controls: Szybalski.

(the gels

( 100)

200 + 20 6+

-

1

I (for 5-7 see footnote b).

in Fig.

in lines l-4 are ofthep,-nutR-N-t,,-galK

type, derived from pD553 ofDrahos

p,-N-t,,-guK (no nut elements; pD553),p,-N-&K 1981). The sequences

5-10 (0)

of the synthetic

(no terminator;

pNP109)

and Szybalski

and gulK (no promoter;

boxA (lines 2-3). boxB (nutR core) and their combination

(198 I). Lines 5-7 represent pK03)

(see Drahos

and

(line 1) are shown in Fig. 2

(t = t, ,; genes N and galK are not shown). ’ Determined

as described

with numerical conditions.

All assays

‘I Percent

antitermination

pNP109

plasmids

against

were performed was

AND METHODS,

the pNP109

= 75.5””

the range for individual

section

g. The results

are an average

galK assay for each series of experiments

of eleven sets of experiments,

done at the same time under identical

at 30°C.

calculated

separately

are 0 and 100 (in parentheses),

100 x (156-33)/(200-33) Actually.

in MATERIALS

values adjusted

(the

200

for each

respectively. and

O0 antitermination

33

series

of experiments,

Using a numerical

values

are

for

values is given in column

assuming

example

the 4.

pNPlO1

the values

for the

for line 1, the “‘, antitermination and

pD553

controls,

pD553

and

would be respectively).

(a)

38240 . . 5’-GCATAAATAACCCCGCT~TACACATT~~~

nutR

boxA (b)

@553A

core

nutR

.

38285 .

.

GAAAAAGGGCATCAA bo?cB -

Hind111

CGTGCTGAGCCCGGCCAAGCTCGCTCTTA~ACKCCAT boxA

(c)

pDS53B

~~A~TA~CCC~T

CGTGCTGAGCCCGGCC

boxB (d)

pD553A8B

CGTGCTGAGCCCGGCCAAGCT~~T~TA~A~

W=AGCTTACTCCCCAT 8

(e)

bp

---------------------C~CTTAC~GCTA~~AGCC~

pD553A12B

-WGGGCA

12 bp (f)

---------------------CGCTCTTACAAAGCTC!AAGCTTGMXTTAGCCCT

pD553A20B

-WGGGCA

bp

20 (g)

---------------------CGCTCI’TACAAAGCTCAAGCTAGCTTG?&CTTAGCCCT 24

pD553A24B

-WZA

bp

(h)

~

---------------------CGCT~TA-~T~~TT~~~TT~~TTA~~~A

pD553A28B

28 bp Fig. 2. Comparison

ofthe wild-type

1. The nt numbering X bp. The inverted plnsmid pD553A. of the synthetic with distances (indicated

5-bp repeats (c) Sequence

between

by the converging

of)zutR core (ho.*B) (underlined)

in Fig. I). Plasmid

pD553A20B

linker. recutting

w-as constructed

are indicated

pD553A8B.

ho.~A and horB varying

as pD553AnB

by HitId

r~tfR sequences

with HirrdfII

arrows.

(b) Sequence

and neighboring

50-bp sequence

from 12-28 bp. The general was constructed

were constructed

and religation.

of pD553A2UB.

structure

by HirzdIII digestion

Two linkers Were retained

(e) Construction of expression plasmids containing both the boxA and ntitR core regions Plasmids pD553A and pD553B were separately digested with HkdIII + PstI and electrophoresed on 1.5 mm 5% PA gels; the 1200-bp and 3&00-bp fragments, respectively, were electroeluted, ligated, and transformed into E. coli C600 ~uIK . Transformants were selected on Mac-gal-amp plates, and the isolated plasmids were tested for the presence of both the boxA and boxB (nutR core) fragments by digestion of mini preparations with Hue111 and electrophoresis on 5”i, PA gels. Whereas the pD553A and pD553B plasmids yielded the characteristic 84- and 90-bp HaeIII fragments, respectively, these were replaced by a 105-bp fragment in plasmid pD553A8B, which carried both the hoxA and hoxB elements. The proper orientation of these two ele-

ments

sequence

~r
is the same as for pD553AYB

of pD553AXB,

in pD553A28B

filling-in with Pollk, and ligation.

hy

regions in

(d) Analogous

of the extended

of these plasmids of pD5S3A8B.

arc separated

and neighboring

pD553B.

(e-g) Sequences

by HirzdIII digestion

of the /tutR region ofphagc

(underlined)

of boxA (underlined)

regions in plasmid

The Hind111 site is overlined.

pD553A12B

and pDS53AXB digestion

(see Fig. 1). (a) Wild-type

et al. (1083). The 8-bp hovA and 17-bp >zulR core (bo.rB) sequences

UWR region from piasmid

ligation. Plasmids the NitldIII

and synthetic

is as in Daniels

filling-in \+ith Pollk and

filling-in with Pollk, ligation with by chance.

Plasmid

pD553AXB

The H&d111 sites arc overlined.

was confirmed

by digestion

with Hinff

and

HindHI. Nucleotide sequencing provided the final confirmation (results in Fig. 2). Construction of plasmids having boxA separated from boxB by more than 8 bp is described in the legend to Fig. 2. All plasmids were sequenced as in RESUL’I’S, section d. (f) Antitermination

activity of the constructs

As shown in Table I, the synthetic boxA-boxB element was up to SOY/, effective at 30°C in antiterminating the Rho-dependent terminator t, ,, . The distance between boxA and boxB (8 bp) was the same as in the wild-type ntttR sequence, but the nucleotides were different. The boxA alone was ineffective, and the nutR core (boxB) had low activity (5-10% antitermination).

El31

TABLE

II

Effect of the distance

between

ho.uA and the nutR core (bad?) on the synthetic Distance

Plasmid

GalK

(n)

8

pD553A8B

nutR antitermination

functiona “(, Antitermination

activity

6X-80

156 f 10

pD553A I2B

12

14+

13

17-32

pD553A2OB

20

113*

9

33-53

pD553A24B

24

1132

0

38

pD553A28B

28

102k

I

41

,’ Plasmid

pD553A8B

(see column

“Distance”)

is described

in Fig.

I and Table I. The nucleotide

are shown in Fig. 2. For other details

and lengths (n) of the space between ho\-Aand ho.-3

sequences

see Table I.

(g) Effect of the distance between boxA and boxB

the antitermination function, especially for the nutR constructs with spacer lengths of 8-20 bp. For lon-

The highest antitermination activity was obtained for the 8-bp distance between boxA and boxB, the same as in the wild-type nutR sequence. The sequence of the constructs with the longer (12, 20, 24

ger spacers there seems to be less dependence

and 28 bp) spacers is shown in Fig. 2, and the results are presented in Table II. The 4-bp increase in the 8-bp spacer length decreases the antitermination to less than half, whereas there is less reduction of the nutR activities when spacer length is 20-28 bp.

20 bp.

DISCUSSION

The present results show that only two synthetic modules, boxA and the 17-bp nutR core are required to reconstruct an active nutR antiterminator, as long as the distance between them is kept at 8 bp. The sequence of this 8-bp spacer does not appear to be important since we replaced the natural 5’-CACATTCC spacer sequence with the HindIII-sitecontaining sequence 5’-CAAAGCTT, while still retaining the nutR activity (Hind111 site is overlined). In nutL, the 5’-AAAATTA spacer between boxA and the nutL core is only 7 bp long. There seems to be little in common between the sequences of these three spacers, other than the length of 7-8 bp. The n&R constructs with a spacer lengthened by one or two turns of the DNA helix (about 10 bp) still retain largely complete antiterminator activity, whereas the nutR activity is severely reduced when the spacer length is increased by only 5 bp (7 to 12 bp), a half helix turn. It appears that the relative spatial arrangement between boxA and the nut core, as governed by the helix geometry, is important for

proper (within greater

on the

phasing of the boxA and boxB modules the turns of the helix), possibly caused by flexibility of the DNA duplex longer than

ACKNOWLEDGEMENTS

The frequent advice of Dr. David Drahos and the skillful help of Cindy Seelinger Diedrich are greatly appreciated. The studies were supported by the NIH Program Project grant (5-POl-CA-23076) (W.S.), core grant 5-P30-CA-07 175, and NIH training grant CA-09075 to A.L.B.

REFERENCES

Adams.

S.P.,

Galluppi, reagents Sot.

Kavka,

K.S.,

G.R.: Hindered in the synthesis

Wykes,

E.J.,

dialkylamino oftwo

Holder,

S.B. and

nucleoside

phosphite

DNA 51-mers. J. Am. Chem.

105 (1983) 661-663.

Birnboim, H.C. and Daly, J.: A rapid alkaline extraction procedure for screening recombinant plasmid DNA. Nucl. Acids Res. 7 (1979) 1513-1523. Brown, A.L. and Szybalski, on the antitermination coliphage Spring Danicls,

lambda.

Harbor, D.L.,

Co&on,

W.: Effect of site-specific activity of the boxA-nufR

Abstracts,

Bacteriophage

NY, August

Schroeder,

21-26,

J.L.,

A.R., Hong, G.-F.,

Blattner.

F.R.: Complete

Hendrix,

R.W., Roberts,

R.A. (Eds.),

Lambda

Cold Spring

Harbor,

mutations sequence

Meeting,

of

Cold

1984, p. 167.

Szybalski,

W.,

Sanger,

Hill, D.F., Peterson,

annotated J.W., Stahl.

II. Cold Spring

lambda

F.,

G.B. and

sequence,

in

F.W. and Weisberg, Harbor

NY, 1983, pp. 519-676.

Laboratory,

El32

Davis,

R.W., Botstein,

Genetics.

Cold

D. and Roth, J.R.: Advanced

Spring

Harbor

Harbor, NY, 1980. de Crombrugghe, B., Mudryj, M.: Specificity termination

M., DiLauro,

are necessary

R. and Gottesman,

lambda

N gene product

and sufficient for the anti-

G.R., Caruthers,

of the nutL DNA segments

mination

and termination

functions

and analysis

W.:

of antiter-

in coliphage

lambda.

D. and Szybalski,

lambda. Friedman,

W.: Antitermination

and termination

of the cloned n&L, N and t,, modules Gene

development,

Ish-Horowitz,

Gottesman,

in Hendrix,

and Weisberg, Laboratory,

of coliphage

M.: Lytic

R.W., Roberts,

R.A. (Eds.), Lambda Cold Spring Harbor,

mode

of lambda

J.W., Stahl,

F.W.

II. Cold Spring Harbor NY, 1983, pp. 21-51.

and Burke, J.F.: Rapid and efficient cosmid vector

cloning. Nucl. Acids Res. 9 (1981) 2989-2998. Luk, K.-C. and Szybalski, W.: Characterization of the cloned termmators

t,,

site of coliphage

, tL3and t,, and of the IZU~Rantitermination lambda.

Maxam,

on a polymer

A.M. and Gilbert,

with base-specific

M.H.: Synthesis support.

ofdeoxyoligo-

J. .4m. Chcm.

Sot.

103

Gene 20 (1982) 127-134.

W.: Sequencing

chemical

cleavages.

end-labeled

Methods

DNA

Enzymol.

65

(1980) 499-560. a region Peltz,

E.L. and Friedman,

of phage

transcription. S.W.,

lambda

Brown,

A.L.,

Hasan,

W.: Thermosensitivity of a truncated

lambda.

Science

unable

required

D.I.: Analysis

of nufR:

for antitermination

of

Cell 31 (1982) 61-70.

activity Salstrom,

16 (1981) 261-274.

D.I. and

M.D. and Caruthers,

nucleotides

Szybalski,

18 (1982) 343-354.

functions

Matteucci,

Olson, E.R., Flamm,

M. and Szybalski,

Synthesis

Drahos,

Spring

by pN. Cell 18 (1979) 1145-I 151.

D., Galluppi,

Gene

Bacterial

Cold

(1981) 3185-3191.

of the bacteriophage

(pN): nur sequences Drahos,

Laboratory,

nufL

N., Podhajska,

A.J.

of a DNA recognition antiterminator

and site:

of coliphage

228 (1985) 91-93.

J.S. and Szybalski,

W.: Phage lambda

to utilize N product

for leftward

nufL mutants

transcription.

Fed.

Proc. 35 (1976) 1538. Salstrom, class

J.S. and Szybalski, of mutants

utilization

W.: Coliphagc

defective

for antitermination

Mol. Biol. 124 (1978) 195-221. Communicated

inutl,

: :I umquc

in the site of gene :c’ product

by A.J. Podhajska

of leftward

transcription

J.