IHF supresses the inhibitory effect of H-NS on HU function in the hin inversion system

Gene. 141 (1994) 17-23 0 1994 Elsevier Science B.V. All rights

reserved.

17

037X-l 119/94/$07.00

GENE 01130

IHF supresses the inhibitory inversion system (Escherichia

coIi histone-like

Naoki Goshima, Fumio Imamoto Department

of Molecular

I Shichono-cho,

Misasqi,

proteins;

Yasunobu

Genetics, Institute Yamashina-ku.

effect of H-NS on HU function in the hin

hupA-hupB

mutants;

Kano, Hiromitsu

ofMolecularand Kyoto,

Cellular

himA-himD

Tanaka,

Bio/ogy,for

mutants;

Kyoko

Phurmaceuticul

hns mutants;

Kohno,

DNA loop)

Toshio Iwaki and

Sciences, Kyoto

Phurmaceuticul

Uniwrsity.

Japan

Received by J. Wild: 20 July 1993; Revised/Accepted:

20 October/21

October

1993; Received at publishers:

29 November

1993

SUMMARY

In the bin-mediated DNA inversion system, HU facilitates formation of the synaptic complex composed of two recombination sites spaced 996 bp apart and of the enhancer situated between them, by looping the DNA as to promote interaction of Hin invertase with the Fis enhancer factor [Johnson et al., Nature 329 (1987) 462-4651. The HU requirement for the in vivo L-mediated inversion was demonstrated previously [ Wada et al., Gene 76 ( 1989) 345-352; Hillyard et al., J. Bacterial. 172 (1990) 5402-5407; Haykinson and Johnson, EMBO J. 12 (1993) 2503-25121 and in the current experiments. This HU action, however, required IHF when H-NS was present in the cell; i.e., the inversion reaction of the bin-invertible DNA fragment carried by the pKK1202R plasmid proceeded efficiently in host cells either deficient in H-NS or in the presence of both H-NS and IHF, but not in the cells depleted for IHF alone. The level of bin mRNA in mutant cells lacking HU or IHF, in which bin inversion did not occur, was normal or slightly increased. When IHF was absent, the stimulating effect of HU on in vitro DNA circle formation of a 125-bp bin fragment between hixL and the enhancer where Fis binds was inhibited by H-NS. The present study provides an example of a multicomponent interaction between HU, H-NS and IHF on the bin DNA region, which contains three characteristic sites, a d(A/T), stretch and bent DNA site, and two putative IHF-binding sites.

INTRODUCTION

The HU protein constitutes a major fraction of the bacterial histone-like components that form the nucleoid. The intracellular level of HU has been estimated to be 30000 dimeric protein molecules per cell. approx. Recently, HU has been shown not only to compact chro-

to: Dr. F. Imamoto, Department of Molecular Genetics, Institute of Molecular and Cellular Biology for Pharmaceutical Sciences, Kyoto Pharmaceutical University, 1 Shichono-cho, Misasagi, Yamashina-ku, Kyoto 607, Japan. Tel. (81-075) 593-2920; Fax (81-075) 502-1613. Correspondence

Abbreviations: pair(s): BSA, SSDl

Ap, ampicillin; bovine serum

0378-1119(93)E0726-T

hla, gene encoding

albumin;

Cm,

B-lactamase; chloramphenicol:

bp, base DTT,

mosomal DNA within the cell in a form of nucleoid, but also to play important roles in various cellular processes such as DNA replication, site-specific DNA recombination and regulation of genetic transcription possibly through formation of the functional configuration of the genomic DNA (Drlica and Rouviere-Yaniv, 1987; Pettijohn, 1988; Flashner and Gralla, 1988; Imamoto and

dithiothreitol; Fis, factor for inversion stimulation; himA, gene encoding IHFs(; himD, gene encoding IHFB; hin, gene encoding Hin invertase; hns, gene encoding H-NS; HU, see INTRODUCTION: hupA, gene encoding HU-2; hupB, gene encoding HU-1; Hy, hygromycin B: IHF. integration host factor; kb, kilobase or 1000 bp; Km, kanamycin; LB, LuriaBertani (medium); oligo, oligodeoxyribonucleotide; PA, polyacrylamide: PCR, polymerase chain reaction; TBE, 10.8 g Tris base/5.5 g boric acid/O.93 g Na,EDTA per liter; wt, wild type.

18 Kano, 1990). A typical case of this action of HU is the looping of a short DNA region ( 122 bp) between the hixL site and

Fis-binding

site of the bin-mediated

system of Escherichia allowing between

coli (Glasgow

inversion

et al., 1989), thereby

recombination of the invertible hixL and hi.uR. More recently,

DNA segment it has been re-

ported that HU has two- to four-fold higher affinity for DNA duplexes containing 5-bp dA stretches, particularly on curved

DNA

(Tanaka

et al., 1993). HU

was also

gene (Silverman

and Simon, 1980; Zieg and Simon, 1980).

Both in vitro (Johnson et al., 1986) and in vivo (Wada et al., 1989; Hillyard et al., 1990; Haykinson and Johnson, 1993), studies

on the hin system have shown that HU is

required for a high rate of the hin inversion reaction. HU is thought to participate in formation of a loop of the 122-bp DNA segment between hixL and the Fis-binding site (Johnson et al., 1987; Glasgow et al., 1989). We have reported that the requirement of both IHF and HU for

shown to bind with high affinity to sharply bent or kinked

Mu phage development

is altered

in the cruciform DNA structures DNA present (Pontiggia et al., 1993). HU can also form a supercoiled

in H-NS,

is no longer

DNA structure 1991; Yasuzawa

(Broyles

and Pettijohn,

et al., 1992), possibly

1986; Hsieh et al., independently

from

in which IHF

in host cells deficient required

and HU

alone supports Mu development, suggesting that IHF suppresses the inhibitory action of H-NS (Kano et al., 1993). The aim of present

study was to examine

whether

the action of a DNA gyrase (H.T., K. Yasuzawa, K.K., N.G., Y.K., T. Saiki and F.I., submitted), which helps in

the HU function in the hin-mediated inversion can be compensated for or competed with by two other proteins,

formation of the looped DNA and extrusion of cruciformstructured DNA. The H-NS protein is also known to significantly compact DNA (Schmid, 1990) having particularly strong DNA-binding affinity for bent DNA (Yamada et al., 1990; Owen-Hughes et al., 1992). This protein has recently been reported to participate in negative regulation of expression of genes such as proV (Owen-Hughes et al., 1992) and hgl (Lopilato and Wright, 1990) by binding to sites in the upstream region or the region surrounding the promoter of these genes. The intracellular level of H-NS is estimated to be approx. 10000 dimeric protein molecules per cell. It might be anticipated that HU and H-NS can compete and interfere with each other action at a bent DNA site. In addition to these two proteins, IHF has been shown to have a potentially important role in chromosomal DNA organization; i.e., it is required for integration and excision of h phage by organizing DNA-h integration protein interaction which is mediated by IHF-induced DNA bending (Goodman and Nash, 1989). The IHF protein also participates in positive and negative regulation of transcription of several bacterial and phage genes, in DNA replication and in TnlO transposition (see Friedman, 1988; Kur et al., 1989). IHF was also found to be involved in facilitating formation of the NifA-os4 holoenzyme complex necessary for initiation of transcription, by looping the DNA between two binding sites for these proteins (Hoover et al., 1990). In the specific systems that have been examined, the role of IHF is to promote or interfere with access of other proteins, or to facilitate bending or looping the DNA into a functional conformation by binding to specific sites. The site-specific DNA inversion of the hin DNA segment that is mediated by Hin invertase alters two different states of gene expression by changing the orientation of the promoter relative to the flagellin structural

H-NS and IHF, both of which bind preferably to the bent DNA. This paper shows that the HU proteins is always required for the hin-mediated inversion, but IHF is required only when H-NS is present in the cell.

RESULTS AND DISCUSSION

(a) Hin-mediated DNA inversion in cells depleted for HU, IHF and/or H-NS To investigate the effect of single or double deficiency of these proteins in cells, we used E. coli strains carrying a single, double or triple deletion mutations of the hupA, hupB, himA, himD and hns genes. These mutations were: hupA16 hupBl1 (YK4135), himA himD (YK4207), hupA16 hupBl1 hns-2 (YK4137), hupA16 hupBl1 himA (YK4197), himA himD hns-2 (YK4205) and hns-2 (YK4124). The parental wt strain YK4122 was used as a control. The hupA16, hupBl1 and hns-2 alleles are deletion mutations of the genes replaced by the KmR, CmR and HyR, respectively (Wada et al., 1989; Yasuzawa et al., 1992); himA and himD are 82[himA]::TnlO and 3[hip or himD]::cat insertions, respectively (Kano and Imamoto, 1990). The transformants of these strains were constructed using pKK1202R harboring a 996-bp hin invertible DNA segment in one (R) orientation (Fig. la), that was prepared from the hupA16 hupBl1 host. Plasmid DNA prepared from the transformants was digested with HincII and run on agarose gel. The plasmid possesses four HincII sites, one of which lies within the hin (H) DNA segment. If the H segment was inverted, six NincII fragments of distinct sizes (A, B, B’, C, C’ and D) would be formed. If the inversion was blocked, then four HincII fragments (A, B, C and D) would be produced. Fig. lb shows the HincII fragments, that hybridized to 32P-labeled 61-mer synthetic oligo corresponding to the C and B’ part of the H segment. The results of gel analysis of

19 the genes encoding 1234567

inversion

activity

the IHF subunits (Glasgow

these in vitro experiments

still display

the bin

et al., 1989). However, three-times

in

higher amount

of

HU than of Fis was used, so enhancement of the bin inversion by crude extracts containing Fis (Factor II) prepared

++---++ H-NS Fig. I. Structure

of the plasmid

-

-

of the Hin-mediated

DNA inversion (a) and the inversion activities in cells depleted for one or two of the three proteins HU, IHF and H-NS (b). (a) Heavy lines in pKK1202

(Wada et al., 1989) represent

DNA flanking DNA. Arrows

Salmonella

typhimurium

genomic

the H segment and the thin line represents pBR322 indicate the invertible segment H in both orientations

(R and L). pKKl202R is the plasmid in which segment H is in the R orientation. Arrowheads indicate Hid1 sites. The sizes of fragments are (in kb): A (3.26), B (3.04) B’ (2.14). C (1.30) C’ (2.17). D (0.51). (b) For assay of the Hin-mediated DNA inversion, pKK1202R was transformed into cells depleted for HU, IHF and/or H-NS by electroporation (Calvin and Hanawalt, 1988). The transformants were streaked on LB plates containing 50 ug Ap/ml and colonies were inoculated into 5 ml of LB broth containing was purified by the mini-preparation Hind1 and separated The DNA fragments

himA

clude the possibility

-++++

used for assay

from

or hip (himD) deletion

mutants,

would not be affected significantly by H-NS which might be present in the extracts (Johnson et al., 1986). To ex-

50 pg Ap/ml. The plasmid DNA (Birnboim, 1983), digested with

in 1% agarose gel in TBE electrophoresis buffer. C and B’ produced by inversion were detected by

Southern-blotting analysis (Southern, 1975). An ohgo that was part of the invertible region [indicated as dashed lines in Fig. la], 5’-TTTACATCAGATAAGAATTTTAGTCCCTTTTCTCATGGAGGATTGCTTTATCAAAAACCTT, was used as a probe. The host E. coli strains were: YK4124 (hns-2, lane l), YK4137 (hupA16 hupBl1 hns-2, lane 2), YK4205 (himA himD hns-2, lane 3). YK4207 (him.4 himD. lane 4), YK4197 (hupA16 hupBl I himA, lane 5) YK4135 (hupA16 hupB1 I, lane 6) and YK4122 (wt, lane 7). A description of the hup, him and hns mutant alleles is in section a. YK4135 and YK4122 are trp’ derivatives of Y K 1340 and YK 1100, respectively, described previously (Wada et al., 1989). The presence(+) or absence (-) of HU, IHF or H-NS is indicated at the bottom.

HincII fragments indicated that HU was required for bin inversion (lane 6) consistent with our previous observations (Wada et al., 1989). This HU action, however, seemed to require IHF to overcome the inhibitory effect of H-NS; i.e., the inversion proceeded efficiently in the absence of H-NS when HU alone (lane 3) or HU and THF (lane 1) were present in the cells, but not when HU and H-NS were present but IHF was missing (lane 4). H-NS or IHF alone (in the absence of HU) did not sustain the inversion reaction (lanes 2 and 5). Obviously, the negative effect of H-NS on the HU function is nullified by IHF, as is the case for wt cells (lane 7). IHF requirement on bin inversion was also observed with the recipient cells carrying a pin mutation (EKKll, EKKll hupA 16 hupBl1 and EKKll himD157; Wada et al., 1989) upon pKK1202R plasmid transformation (data not shown). These findings seem to be inconsistent with the previous notion that strains carrying mutations only in

that expression

of the bin structural

gene is somehow affected by the mutations employed the present experiments, thereby resulting in reduction the level of the Hin invertase of the inversion

reaction,

levels of bin mRNA formants

leading

we measured

produced

by Northern-blotting

to the inactivation the intracellular

from the pKK1202R analysis

in of

trans-

using 32P-labeled

DNA probe (67-mer) of the part of H segment. As a reference the level of blu mRNA was determined using labeled DNA probe (2%mer) corresponding to part of the bla gene on the plasmid. The results presented in Table I show that the level of bin mRNA was normal or even higher in mutant cells lacking HU or IHF in which hin inversion was inactive, thus supporting the conclusion that HU participates in the hin inversion processes and that in the absence H-NS.

of IHF

its action

is antagonized

by

(b) Effects of HU, IHF and H-NS on circularization of the 125-bp bin DNA fragment The ability of HU to stimulate the rate of T4 DNAligase-mediated circle-formation of DNA has been demonstrated in vitro by using short (99-126 bp) DNA fragments (Hodges-Garcia et al., 1989). We examined whether H-NS interfered with the in vitro action of HU for DNA circularization and whether this interference was relieved by IHF. The 122-bp region (the Hind111 fragment used in this study was 125 bp) between the hixL and enhancer where Fis binds of the hin invertible DNA region was cloned into pUCl19 and amplified. Looping of this DNA region has been proposed to be facilitated by HU (Johnson et al., 1987; Glasgow et al., 1989). The 125-bp fragment containing the 122-bp region was prepared by Hind111 digestion of the plasmid and mixed with various combinations of the three proteins HU, IHF and H-NS, and was ligated with T4 DNA ligase. The amount of monomeric circular DNA formed was determined by electrophoresis in agarose gel to separate it from monomeric, linear and polymeric DNA. The results in Fig. 2 indicate that the monomeric circular DNA is efficiently formed in the presence of HU (lane l), but its formation is significantly reduced by H-NS in the absence of IHF (lane 5). IHF when present, in addition to both HU and H-NS proteins, relieved this H-NS interference (lane 7). H-NS alone did not sustain DNA circularization (lane

20 TABLE

I

Levels of bin mRNA Strain

in cells depleted

YK[pKKl202]

for HU, IHF and/or

H-NS”

Genotype

Depleted

protein(s)

Relative

level of

hin mRNA

h/a mRNA

1.00 0.97

1.oo 0.46

YK4122 YK4135 YK4207

wt himAmhimDm

HU IHF

YK4124

hns-

H-NS

2.57 1.18

3.14 I.15

YK4197 YK4137

hupA_hupB_himA-

HU IHF

1.01

3.90

hupA-/wpB~hns~

Y K4205

hintA_himD-hns-

HU H-NS IHF H-NS

I .oo 0.40

1.00 2.52

“mRNAs

hupA-hupB-

were

isolated

from

YK4122,

YK4135,

YK4207,

YK4124,

YK4197,

YK4137

and

YK4205

harboring

the

pKK1202

plasmid.

The

mRNAs were analyzed by Northern-blotting analysis according to the procedure of Aiba et al. (1981). The synthetic oligo. 5’-ATTTTGGTCAATTGTTGACACCCGAATATACCCAATAGTAGCCATGATTTTCTCCTTTACATCAGAT. at 53 to 120 nt from hixL, labeled with [Y-~‘P]ATP and T4 polynucleotide kinase. the 1600-bp BumHI-Hind111 rrnB promoter fragment (which includes about 80 nt of the 16s rRNA gene) prepared from pKKlO-2 (Brosius, 1984) labeled by nick-translation, and the synthetic oligo, T-GCGGCGACCGAGTTGCTCTTGCCCGGCG at 249 to 476 nt from the h/u coding

region,

labeled

with [y-“‘P]ATP

and T4 polynucleotide

and b/u mRNA, respectively. The levels of hin and blu mRNA are represented as values relative of rRNA for the rrnB gene. The other interpretations are described in the section a.

3). IHF alone could slightly but discernibly promote formation of monomer-circular DNA (lanes 2 and 6). At such high [IHF]/[DNA] ratios, IHF may bind to the DNA fragment side by side as HU does, or it may bind to the putative IHF-binding sites on the hin fragment, as described in section c, and promote bending of the fragment which can be ligated at some frequency into circles. (c) Possible inactivation and activation of the 125-bp /zin DNA fragment The enhancing effect of HU on the bin-promoted DNA inversion is believed to facilitate assembly of the synaptic complex formed by the recombination sites and the enhancer, and/or stabilize this complex (Johnson et al., 1986). The HU protein is probably required to promote or stabilize bending of the 122-bp DNA region between the hixL site and the enhancer, thus allowing formation of a loop to achieve interaction of the Hin recombinase and the Fis protein. The bending of DNA would be attained when HU dimers occupy sites on the linear DNA as proposed by Tanaka et al. (1984). Intracellularly, the level of HU is estimated to be about 30000 dimers per cell, which corresponds to one dimer per 280-420-bp DNA region assuming two to three molecules of chromosomal DNA (4.2 x lo6 bp) per cell. Although the binding abreast of several HU dimers to a linear DNA fragment at intervals of approx. 9-bp binding length was demonstrated in previous studies under in vitro conditions (Bonnefoy and Rouviere-Yaniv, 1991; Tanaka et al., 1993) this state of DNA binding would be possible only when large excess of HU binds to the DNA sequences, which may contain the sites exhibiting the avid HU binding.

kinase

were used as probes

for analysis

to those of wt cells and normalized

of hin, rrnB

against

the level

From the present study, the 122-bp hin DNA region seems likely to contain some characteristic features that are favorable for IHF and H-NS binding. In the 122-bp hin region, there are three characteristic sequences; i.e., 5’-NNTCAANNANTNA at 105 to 117 nt from the hixL site, near the enhancer, and 5’-NATCAANNANNTN at 17 to 5 nt, near the hixL site, both of which are similar to the putative IHF consensus sequence, 5’-AATCAANNANTTA, proposed by Goodrich et al. (1990) and 5’-AAAANNNNNNAAAA located at 35 to 48 nt which possibly cause bending of DNA. Recently, HU has been demonstrated in vitro to bind preferably to DNA duplexes containing d(A/T), stretches, particularly to curved DNA, exhibiting two- to fourfold higher affinity for these DNAs than DNA without such d(A/T), stretches (Tanaka et al., 1993). H-NS could bind preferably to the d(A/T), stretched and bent DNA site by interfering with access of HU to this DNA site in the absence of IHF. Binding of IHF to either one or both putative IHF-binding sites resulting in strong DNA bending may change the bending feature of the d(A/T), stretches of DNA, thus reducing the accessibility of H-NS to this site and thereby enhancing the binding affinity of HU to this d(A/T), stretch of DNA. It is known that a diagnostic feature of a bent DNA fragment is its slower mobility at a relatively low temperature during PA gel electrophoresis (Koo et al., 1986; Diekmann, 1987). The migration rate of the 125-bp bin DNA fragment was determined on PA gels at 4’C and 60°C with reference to 131-bp noncurved N131 DNA and 132-bp curved B132 DNA. As seen in Fig. 3A, the 125-bp hin DNA fragment migrated at slower rates than N131 DNA at both temperatures, but more markedly at a low

A

B 1

2

3 4

5

6

7

8 9

10

60 “C N131

Hin125

r-

0 1.3 2.5

5 10

I( 4)

IHF (PrOteinS

present)

HU

1

Fig. 2. Effects

IHF

H-NS

HU IHF

HU H-NS

2

3

4

6

of HU,

IHF

and

IHF H-NS

HU IHF H-NS

6 H-NS

7 on

NOW

BSA

8

9

circularization

-

C

10

12345

6 7 8

9 10

of the

125-bp bin DNA segment. (A) The 122-bp sequence between the hixL and the enhancer of the hin invertible DNA region was amplified by HindllI-linker (primer 1, PCR using primers containing 5’-CCCAAGCTTAGGTTTTTGATAAAGCAATCCT;

primer

( n9)

2,

S-CCCAAGCTTTTTT-GGTCAATTGTTGACACCCG) and cloned into the Hind111 site of pUCl19. The 125-bp fragment having Hind111 sites (1 ng) was labeled by an exchange-reaction with [Y-~‘P]ATP. The

H-NS

DNA fragment (20 ng) was mixed with various combinations of the three proteins HU, IHF or H-NS. The combinations of the three proteins were: HU (lane 1); IHF (lane 2); H-NS (lane 3); HU and IHF (lane 4); HU and H-NS (lane 5); IHF and H-NS (lane 6); HU, IHF and H-NS (lane 7); no protein (lane 8). A control containing BSA and the 125-bp hin DNA fragment alone (no T4 DNA ligase) was run in lanes 9 and 10, respectively. The total concentration of the protein or proteins added was 50 ng; i.e.. 50 ng in lanes 1, 2. 3 and 9; 25 ng each in lanes 4, 5 and 6; 17 ng each in lane 7. [In the experiments in which 25 ng of each protein was additively combined, essentially similar results were obtained (data not shown).] After incubation for 20 min at 24 C in the buffer containing

50 mM TrisHCl

pH 7.6/ 10 mM

Free

MgClJIO mM

DTT/l mM ATP (Hodges-Garcia et al., 1989), the DNA was ligated with T4 DNA ligase (70 units) for 12 h at 2OY. The reaction was stopped by addition of SDS (final concentration O.l%), treated with phenol and precipitated with ethanol. The DNA recovered was separated in 4% agarose gel in TBE electrophoresis buffer. The gel was dried with gel-drier and autoradiographed. (+) and (-) indicate the monomer-circular and monomer-linear DNA, respectively. (B) The densitometer tracing of (+) and (-) bands are shown below. The other interpretations are described in section b.

temperature, thus showing that the bin fragment is actually bent to a discernible extent. Fig. 3B shows that IHF and H-NS bound more strongly to the 125bp bin fragment than to a control DNA fragment (N131) that did not contain neither a dA stretch nor an IHF consensus

C

11 12 13 14 15 16

Fig. 3. DNA bending

17 18

Free

19 20 21 22

of the 125-bp bin fragment

(A) and affinities

of

IHF and H-NS for this DNA fragment (B). (A) DNA bending was assayed by electrophoresis on 7% PA gel in a TBE system at 4°C (lanes 1, 2, 3 and 4) and 6O’C (lanes 5, 6. 7 and 8). Nl31 (131-bp noncurved DNA: Tanaka et al., 1993) is indicated in lanes 1 and 5, B132 (132-bp curved DNA; Tanaka et al., 1993) in lanes 2 and 6, Hin125 in lanes 3 and 7 and size-markers in lanes 4 and 8. (B) Gel retardation assay of Hinl25 and Nl31 fragments bound with IHF or H-NS. Hin125 and Nl31 fragments were labeled with [y-3ZP]dATP and T4 polynucleotide kinase. 3ZP-labeled Hin125 was incubated with IHF (lanes 2, 3, 4 and 5) or H-NS (lanes 12, 13, 14, 15 and 16) and 32P-labeled Nl31 was incubated with IHF (lanes 7, 8, 9 and 10) or H-NS (lanes 18, 19, 20. 21 and 22). The sample of H-NS binding on Hin125 and N131 or the IHF binding on Hin125 and N131 were electrophoresed in the same gel. Numbers above gels indicate the amount of IHF or H-NS added to the reaction. The reaction mixture (20 ~1) containing 0.5 ng of 3ZPlabeled DNA fragment and indicated amount of IHF or H-NS in 10 mM Tris-HCl pH 8.0/50 mM NaCl/lO mM MgClJ0.1 mM DTT/O.Ol% BSA/O.Ol% Brij-58/8% glycerol was incubated for 20 min at 23 “C. The reaction mixtures were separated by electrophoresis in 6% PA gel in TBE buffer. The gels were fixed in 10% acetic acid/ 10% methanol,

dried and autoradiographed.

22 sequence.

When

was present,

a relatively

more than

high concentration

one or two proteins

bind to these DNA fragments.

The possibility

of IHF seemed

to

that H-NS

and IHF exhibit antagonistic or cooperative effects by interacting with each other and/or with HU without involvement

of DNA sequences

in the gin inversion

is excluded

by the fact that,

system, the inhibitory

action of H-NS

on HU in the absence of IHF has not been observed in spite of strong dependency of this gin inversion activity on HU (data not shown). This supports the idea that the effects Further

of H-NS

and

IHF

studies to confirm

are

exerted

this possibility

on

the

DNA.

are in progress.

(d) Conclusions (I) The in vivo function of HU in stimulating the hinmediated inversion reaction required IHF when H-NS was present in the cell, suggesting that HU action was inhibited by H-NS and IHF suppressed the inhibitory effect of H-NS. (2) The in vitro circularization of a 125bp hin DNA fragment was facilitated by HU. This activity was inhibited by H-NS when IHF was absent. Thus, effects of these three proteins were similar to their actions on the hin inversion system. (3) The 125bp hin DNA fragment was found to be bent significantly in vitro and was bound by H-NS and IHF more strongly than the control DNA without a d(A/T), stretch and an IHF-binding consensus sequence. (4) The present study provides an example of a multicomponent interaction between HU, H-NS and IHF on the hin DNA region, which contains three characteristic sites, a d(A/T), stretch and bent DNA site, and two putative IHF-binding sites.

with DNA: evidence for formation of nucleosome-like structures with altered DNA helical pitch. J. Mol. Biol 187 (1986) 47760. Calvin, N.M. and Hanawalt, P.C.: High-efficiency transformation of bacterial cells by electroporation. J. Bacterial. 170 ( 1988) 2796--2801, Diekmann, S.: Temperature and salt dependence of the gel migration anomaly 2477265. Drlica,

We are grateful of H-NS.

to Dr. Heisaburo

Shindo

for the gift

REFERENCES Aiba, H., Adhya, S. and de Crombrugghe, B.: Evidence for two functional gal promoters in intact Escherichia coli cells. J. Biol. Chem. 256(1981) 11905511910. Birnboim, H.C.: A rapid alkaline extraction method for the isolation of plasmid DNA. Methods Enzymol. 100 (1983) 243-255. Bonnefoy, E. and Rouviere-Yaniv, J.: HU and IHF, two homologous histone-like proteins of Eschrrichia coli. form different protein-DNA complexes with short DNA fragments. EMBO J. 10 (1991) 6X7-696. Brosius. J.: Toxicity of an overproduced foreign gene product in Escherichicl coli and its use in plasmid vectors for the selection of transcription terminators. Gene 27 (1984) 161-172. Broyles, S. and Pettijohn, D.E.: Interaction of the E. co/i HU protein

DNA fragments.

K. and Rouviere-Yaniv,

Nucleic

J.: Histone-like

Microbial. Rev. 51 (1987) 301-319. Flashner, Y. and Gralla, J.D.: DNA dynamic

Acids Res. 15

( 1987)

proteins

of bacteria.

flexibility

and protein

recognition: differential stimulation by bacterial histone-like protein HU. Cell 54 (1988) 713-721. Friedman, D.I.: Integration host factor: a protein for all reasons. Cell 55(1988)545-554. Glasgow, A.C., Hughes.

K. and Simon, M.I.: Bacterial

DNA inversion

systems In: Berg, D.E. and Howe. M.M. (Eds.), Mobile DNA. American Society for Microbiology. Washington, DC. 1989, pp. 6377659. Goodman, S.D. and Nash, H.A.: Functional replacement of a proteininduced bend 251 254.

in a DNA

recombination

site. Nature

341 (1989)

Haykinson, M.J. and Johnson, R.C.: DNA looping and the helical repeat in vitro and in vivo: effect of HU protein and enhancer location on Hin invertasome assembly. EMBO J. 12 (1993) 250332512. Hillyard, D.R., Edlund. M., Hughes, K.T., Marsh, M. and Higgins, N.P.: Subunit-specific phenotypes of S&ton& ryphimurium HU mutants. J. Bacterial. 172 (1990) 540225407. Hodges-Garcia, Y., Hagerman, P.J. and Pettijohn, D.E.: DNA ring closure mediated by protein HU. J. Biol. Chem. 264 (1989) 14621~14623. Hoover, T.R., Santero, E. Porter, factor stimulates interaction transcriptional activator for (1990) 11-22. Hsieh, L., Rouviere-Yaniv, J. and

S. and Kustu, S.: The integration host of RNA polymerase with NIFA, the nitrogen fixation operons. Cell 63 Drlica, K.: Bacterial

DNA supercoil-

ing and [ATP]/[ADP] ratio: changes associated with salt shock. J. Bacterial. 173 (1991) 3914-3917. Imamoto, F. and Kano, Y.: Physiological characterization of deletion mutants of the hupA and hupBgenes in Escherichiu coli. In: Drlica, K. and Riley, M. (Eds.), The Bacterial Chromosome. American Society for Microbiology, Washington. DC, 1990, pp. 259.-266. Johnson, R.C., Bruist, M.F. and Simon, MI.: Host protein requirements for in vitro site-specific DNA inversion. Cell 46 (1986) 531 -539. Johnson,

ACKNOWLEDGEMENT

of curved

R.C., Glasbow,

A.C. and Simon,

M.I.: Spatial

relationship

of

the Fis binding sites for Hin recombinational enhancer activity. Nature 329 (1987) 462.-465. Kano, Y. and Imamoto, F.: Requirement of integration host factor (IHF) for growth of Escherichiu co/i deficient in HU protein, Gene 89 (1990) 133~_137. Kano, Y., Yasuzawa, K., Tanaka, H. and Imamoto, F.: Propagation of phage Mu in IHF-deficient Eschrrichiu co/i in the absence of the H-NS histone-like protein. Gene 126 (1993) 93397. Koo,H.S., Wu, H.M. and Crothers, D.M.: DNA bending at adenine thymine tracts. Nature 320 (1986) 501-506. Kur, J., Hasan, N. and Szybalski, W.: Physical and biological consequences of interactions between the integration host factor (IHF) and coliphage lambda late pk promoter and its mutants. Gene 81 (1989) I-15. Lopilato. J. and Wright, A.: Mechanisms of activation of the cryptic hgI operon of Eschrrichitr co/i. In: Drlica, K. and Riley. M. (Eds.). The Bacterial Chromosome. American Society for Microbiology, Washington, DC, 1990, pp. 4355444. Owen-Hughes. T., Pavitt, G., Santos, D.S., Sidebotham, J.M., H&on, C.S.J., Hinton. J.C.D. and Higgins, CF.: The chromatin-associated protein H-NS interacts with curved DNA to influence DNA topology and gene expression. Cell 71 ( 1992) 2555265.

23 Pettijohn,

D.E.: H&one-like

proteins

ture. J. Biol. Chem. 263 (1988) Pontiggia,

A., Negri Beltrame,

and bacterial

chromosome

struc-

12793312796.

M. and Bianchi,

M.E.: Protein

HU binds

specifically to kinked DNA. Mol. Microbial. 7 (1993) 3433350. Schmid, M.B.: More than just ‘histone-like’ proteins. Cell 63 (1990) 451-453. Silverman, M. and Simon, M. : Phase variation: switching mutants. Cell 19 (1980) 845-854.

genetic

analysis

of

Southern, E.M. : Detection of specific sequences among DNA fragments separated by gel electrophoresis. J. Mol. Biol. 98 (1975) 503-517. Tanaka, H., Goshima, N., Kohno, N., Kano, Y. and Imamoto, F.: Properties of DNA-binding of HU heterotypic and homotypic dimers from Escherichirc coli. J. Biochem. 113 (1993) 5688572. Tanaka, 1, Appelt. K., Dijk, J., White, S. and Wilson, K.: 3-A resolution

structure of a protein with histone-like Nature 310 (1984) 3766381. Wada,

M., Kutsukake,

K., Komano,

properties

T., Imamoto,

in prokaryotes. F. and Kano,

Participation of the hup gene product in site-specific in Escherichiu coli. Gene 76 (1989) 345-352.

Y.:

DNA inversion

Yamada, H., Muramatsu, S. and Mizuno, T.: An Escherichiu coli protein that preferentially binds to sharply curved DNA. J. Biochem. 108 (1990) 420-425. Yasuzawa, K., Hayashi,

N., Goshima,

Kano, Y.: Histone-like constraint of supercoils

N., Kohno.

K.. lmamoto,

F. and

proteins are required for cell growth in DNA. Gene 122 (1992) 9-15.

and

Zieg, J. and Simon, M.: Analysis of the nucleotide sequence of an invertible controlling element. Proc. Natl. Acad. Sci. USA 77 (1980) 4196-4200.