Chromosoma(1992) 102:S1-S6

CHROMOSOMA 9 Springer-Verlag1992

The complex for replication initiation ofEscherichia coli Walter Messer, Heidi Hartmann-Kiihlein, Uwe Langer, Ellen Mahlow, Angelika Roth, Sigrid Schaper, Britta Urmoneit, and Birgit Woelker Max-Planck-InstitutfiirmolekulareGenetik,Ihneslrasse73, W-1000Berlin33,FederalRepublicof Germany ReceivedSeptember12, 1992

Abstract. We probed the complex between oriC and DnaA protein using two types of mutants in oriC. Base changes in the DnaA binding sites, DnaA boxes, had little effect on origin function. Mutations which change the distance between DnaA boxes R3 and R4, on the other hand, inactivated oriC unless the mutation deleted or inserted one complete helical turn. Origins with other 10 base pair insertions in the interval between DnaA boxes R2 and R3 were functional, but not insertions in the R1-R2 interval. FIS protein binds to a bipartite site in oriC between DnaA boxes R2 and R3. A model for the oriC / DnaA complex based on these results suggests an array of DnaA monomers with a 34A spacing upon which oriC is arranged.

Introduction The replication origin, oriC, of Escherichia coli consists of approximately 250 base pairs. A sequence comparison of replication origins of several Gram-negative bacteria shows that highly conserved regions are separated by spacer regions of variable sequence but constant length (Zyskind et al. 1983). This observation, as well as a mutational analysis of oriC ofE. coli (Hirota et al. 1981), suggested that a precise arrangement of conserved protein binding sites in an appropriate distance to each other is of central importance for the formation of a higher order nucleoprotein structure required for the initiation of replication. Prominent among the conserved regions are four (Fuller et al. 1984) or five (Matsui et al. 1985) 9 base pair binding sites, DnaA boxes, for the initiation protein Correspondenceto: W. Messer

DnaA. Binding sites also exist for proteins which modify the DNA conformation: IHF; integration host factor (Polaczek 1990; Filutowicz and Roll 1990), and FIS, factor for inversion stimulation (Gille et al. 1991, Filutowicz et al. 1992). 11 GATC recognition sequences for Dam methyltransferase have to be methylated for replication initiation (Messer and Noyer-Weidner 1988). The interaction of the initiation protein DnaA with the DnaA boxes in oriC, assisted by proteins HU, IHF, and FIS, is the first and crucial step in the initiation process. 10-30 DnaA monomers form a complex with oriC (Funnell et al. 1987) which results in a local distortion of the helical structure in an AT-rich region at the left border of oriC (Bramhill and Kornberg 1988; Gille and Messer 1991). The subsequent formation of at least two DnaA dependent primosomes (Seufert and Messer 1987a; Messer et al. 1988) then leads to priming and bidirectional replication from the right half of oriC (Seufert and Messer 1987b). Here we describe the unique binding site for FIS in oriC, and present a model of the oriC initiation complex based on an analysis of mutants in the oriC spacer regions and in DnaA boxes.

Results and Discussion Mutations which change the distance between DnaA boxes Previous insertion mutagenesis by filling in of restriction sites resulted in inactive origins (Oka et al. 1984), in agreement with the postulate of the recognition frame model of binding sites separated by spacer regions (Hirota et al. 1981). In order to get a more detailed analysis of the function of spacer regions we varied the dis-

$2 tance between the two rightmost DnaA binding sites R3 and R4 systematically over a range of 38 base pairs. A 1541 bp fragment containing the oriC region (oriC coordinates -44 to +1497, Buhk and Messer 1983) was cloned with Xbal linkers between the EeoRI and HindlII sites ofpBR322 (Messer et al. 1991). The resulting joint replicon, pOC161, has a unique HindlII site between DnaA boxes R3 and R4 in which insertions and deletions have been introduced. Insertions ranging from 2 to 24 base pairs were obtained using synthetic SmaI and EcoRI linkers and oligonucleotides of different lengths in combination with Klenow and Mung Bean enzyme treatment. Deletions in a range from -14 to - 4 base pairs were obtained by oligonucleotide-directed mutagenesis (Kunkel 1985) using an oriC fragment subcloned in M13mp8. All mutations were verified by sequence analysis. The origin function of the different oriC mutants was tested by transformation into a polA host, in which oriC but not the pBR322 origin can replicate. The ratios of transformants in polA/polA + hosts were normalized against the wildtype pOC161. Deletions of 12 and 10 base pairs showed moderate, insertions of 10 (AGCTCCCGGG) and 12 (AGCTGCCCGGGC) base pairs good oriC function. All other insertions and deletions resulted in a nonfunctional oriC (Table 1). This shows that mutants which had inserted or deleted one helical turn between the DnaA boxes R3 and R4, at oriC position 244, have a functional origin, and suggests that the DnaA boxes R3 and R4 must be positioned on the same face of the DNA double helix. However, an inser-

tion of two helical tums inactivated oriC. All different 1012 base pair insertion/deletion sequences analyzed so far showed an Ori + phenotype, although with different efficiencies (see below). In order to extend this analysis to other intervals in oriC we inserted 10 base pairs between DnaA boxes R1 and the IHF binding site (position 95), between the IHF site and DnaA box R2 (position 156), into the high-affinity FIS binding site (see below) between DnaA boxes R2 and R3 (position 203), between the FIS site and DnaA box R3 (position 220), and again between R3 and R4 (position 248) (see Figure 1). Defined 10 base pair insertions (GAAGTACTCA), containing a Seal site, were obtained by oligonucleotide-directed mutagenesis (Kunkel et al. 1985). For the analysis oforiC function with thepo/A assay the mutated oriC fragments were subcloned into the joint replicon pOC170. This plasmid contains the oriC region (coordinates -176 to +1497, Buhk and Messer 1983), the pBR322 origin on a NotI cassette, and the bla gene of pBR322 for selection. In the polA test for origin function the insertions in the interval between DnaA boxes R1 and R2 turned out to be nonfunctional (insertions S1 and $2 in Fig. 1 and Table 2). Origins with insertions $3 and $4 between DnaA boxes R2 and R3 showed a relatively high oriC activity, although insertion $3 disrupts the binding site for FIS. 10 base pair insertions in the R3-R4 interval gave ambiguous results. Insertion $5 (GAAGTACTCA) was quite inefficient, whereas the previously analyzed 10 bp insertion mutation S5B (CCCGGGAGCT) gave relatively good

Table 1. In vivo functionality of distance mutations at oriC coordinate 244

Table

insertions/deletions (bp)

relativetransformation frequency (polA/polAD

oriC function

0 (WT) -14

100 0

- 12

39

+++ -++

- 10

17

+

-8 -4 +4 + 10 + 12 + 14 +16 + 22 + 24

0 0 0 67 73 0 0 0 0

---+++ +++ -----

Strains H25 (polA+) and H221 (fadA::TnlO, polA1) were transformedwith pOC161 wildtypeor mutant DNA using the method described(Hanahan 1983).

2. Origin function of 10 bp insertions at differentpositions in

oriC relative transformation frequency (polA/polA4)

oriC function

95 156 203 220 248 248

100 1 1 65 62 4 15

+++ --+++ +++ (+) +

pOC161 (WT) pOC161/S5 248 pOC161/S5B 248

100 1 24

+++ -++

plasmid

pOC170 (WT) pOC170/Sl pOC170/S2 pOCl70/S3 pOC170/$4 pOC170/$5 pOC170/S5B

insertion coordinate

Strains H25 (polA+) and H221 (fadA::TnlO, polA1) were transformed with pOC170 or pOC161 wildtype or mutant DNA using electroporation with the Gene Pulser of Biorad (Richmond, Calif., USA).

$3

Fig. 1. Position and origin function of 10 bp insertion mutations in oriC of E. coll.

origin function which was further enhanced upon subcloning into the joint replicon pOC161 (Table 2). Plasmid pOC161 has an altered gid.4 promoter activity due to the presence of an XbaI linker at oriCposition -44 (K/511ing et al. 1988). It was suggested that transcription from the gidA promoter affects oriC function (Asai et al. 1992). However, introducing the same gid.4 promoter modification into pOC170 gave results which were not significantly different from unmodified pOC170 (not shown). Insertions in the R3-R4 interval thus result in functional origins, but the quantitive aspects of origin function depend on the inserted sequence, the test plasmid, and possibly the method of transformation. pOC170 allows a more stringent test of origin function due to the possibility to remove the pBR322 origin on its cassette, followed by transformation with the resulting minichromosome. The results corroborate those obtained with thepolA test, and demonstrate that the relative transformation efficiencies obtained with minichromosome transformation and in thepo/A test are a reflection ofplasmid copy number (not shown). The results on the fimction of insertion mutations in oriC are summarized in Fig. 1. Contrary to the results obtained with mutations which change the distances within oriC, the introduction of point mutants into the DnaA boxes was without significant effect on oriC function (Holz et al. 1992). A combination of two point mutants in DnaA boxes R1 and R4 gave near wildtype transformation efficiencies in thepo/A test, but a somewhat reduced growth rate of cells carrying such mutant plasmids. We have extended this analysis by

combining mutations, two at a time, in DnaA boxes R1, R2, R3, and R4. As in the previous experiments, all combinations tested showed a relatively good oriC function (data not shown). In vitro analysis of DnaA binding revealed that despite the near-wildtype function point mutations in the different DnaA boxes showed an impaired affinity to DnaA protein in the DnaseI footprinting pattern ( Holz et al. 1992). In contrast, the spacer mutations showed wildtype-like DnaseI protection at all five DnaA boxes (data not shown).

FIS binding in oriC FIS (factor for inversion stimulation) is a small (12 kD), basic, heat stable DNA-binding and bending protein belonging to the histone-like proteins. It was discovered as a factor stimulating the Hin and Gin site-specific recombination processes of Salmonella and phage Mu (Kahmann et al. 1985; Johnson and Simon 1985). It is also involved in the CIN-System of bacteriophage P1, in the phage ~, site-specific recombination and in transcription activation of rRNA and tRNA operons (cf. Finkel and Johnson 1992). Recently we and others showed that FIS binds specifically to a binding site in oriC (Gille et al. 1991; Filutowiez et al. 1992), resulting in a bend of about 50 ~ Transformation offis mutant strains by oriC plasmids is very inefficient (Gille et al. 1991), and the chromosomal replication control is impaired in fis mutants (Boye et al. 1992; yon Freiesleben and Rasmussen 1992).

$4 There are several potential FIS binding sites in oriC. In DnaseI footprinting experiments one site between DnaA binding sites R2 and R3 is protected when low concentrations of FIS are used. Other protected sites appear with higher concentrations of FIS (Gille et al. 1991; Filutowicz et al. 1992). We therefore used a more dis-

criminative technique for the footprinting of oriC in the presence of FIS. MPE footprinting is based on the intercalation of Methidiumpropyl-EDTA-Fe++ (MPE) and subsequent cutting of the phosphate backbone by a hydroxy-radical mechanism (Hertzberg and Dervan 1984). Even very high concentrations of FIS gave only one protected site, between DnaA boxes R2 and R3 (Fig. 2). The precise protection by FIS against MPE action was determined using a parallel G+A reaction. 18 base pairs adjacent to the DnaA binding site R2 were protected (Fig. 3). In comparison, DnaseI footprinting in the presence of FIS resulted in a protected region covering the whole area between DnaA boxes R2 and R3 (Gille et al. 1991; Filutowicz et al. 1992; Fig. 4). Due to the mechanism of action of MPE, this technique is likely to detect high-affinity binding sites. We therefore suggest that there is a unique FIS binding site in oriC between DnaA boxes R2 and R3 with a high-affinity part detected by MPE and DNaseI footprinting, and a low-affinity part seen in DNaseI footprints only. One of the 10 bp insertion mutations described above is an insertion into the high-affmity part of the FIS binding site (insertion $3). The oriC function of this mutation was nearly normal. MPE footprints showed no protection by FIS, demonstrating the destruction of the high-affinity part of the FIS binding site. However, FIS still protected the region from position 213 to 220 in the low-affinity region. Modification of 6 base pairs within the complete 25 bp FIS binding site, however, destroyed the protection against both MPE and DNaseI, and thus completely eliminated FIS binding, oriC activity of this mutant was reduced (data not shown). These results suggest that FIS action on oriC is possibly direct and not or not only due to a pleiotropic effect. A model for the oriC/DnaA initiation complex

Fig. 2. MPE footprintof the complexof FIS with the largeroriCregion. 5 ng ofa MaelI-XhoI fragment(-111 to +417) fromthe oriC region, 32p-end-labelledat the MaelI site, were incubated at room temperature for 15 rain with the indicated mounts of purified FIS protein in 20 p.1buffer (40 mM Hepes/KOH pH 7.6, 100 mM KC1, 10 mM Mg-acetate, 1 mM DTE). 21.tlMPE (Hertzbergand Dervan 1985) were added, and incubationwas continuedfor 10 rain. 2kt110 mM DTr were added and the mixture was incubated at 37~ for 7 rain. The reactionwas stoppedwith 50~10.3 M Na-acetate,and the mixturewas phenolized,EtOH precipitatedand applied to a 8% sequencingpelyacrylamidegel.

Insertion and deletion mutagenesis in the interval between DnaA boxes R3 and R4 demonstrated that variation of the distance by + or - one integral helical turn resulted in functional origins. All other distance variants were inactive. The traditional explanation for such a result is the existence of a DNA loop between two binding sites due to protein-protein contacts between the DNA binding proteins. However, such an interpretation cannot explain that disruption of an individual DnaA box by base changes shows nearly no in vivo phenotype, although DNaseI footprinting in vitro demonstrates a lack of DnaA binding at such a mutated site (Holz et al. 1992). Therefore we want to suggest that the initiation complex is composed of a regular army of DnaA monomers with a fixed spacing of 34A. We assume oriC DNA to be

$5

Fig. 5. Model of the oriC / DnaA initiation complex. Pointed black circles indicate DnaA monomers, black arrows DnaA boxes on the shaded DNA double helix. Binding and bending sites for IHF and FIS are shown.

arranged on the surface of such a structure with the DnaA boxes mediating the binding, but with additional weak interactions and assisting proteins IHF and FIS being responsible for the final shape (Fig. 5). A modification of the distance by one helical turn would then still allow e.g. DnaA box R4 to find a DnaA parmer for binding in the proper orientation. Destruction of the binding capacity of a single DnaA box, or even of two boxes, might be tolerable because of the rigidity of the residual structure. Fig. 5 illustrates this principle. It does not predict a given shape. A spherical arrangement o f DnaA monomers e.g. is as likely as the planar arrangement shown in Fig. 5. Insertions of one helical turn are permissible downstream of DnaA box R2, i.e. within or downstream of the bend imposed by FIS. In the part upstream of R2 no sites for productive 10 base pair inserts were found so far. The net result of this higher order nucleoprotein structure is a helical distortion or partial unwinding in the AT-rich region at the left border of oriC. Such a conformational change mediates the entry of DnaB helicase and the subsequent formation of DnaA primosomes in the right half of oriC. Fig. 3. MPE footprint of the FIS/oriC complex. A BamHI-XhoI fragment (+92 to +417) from oriC was end-labelled at the BamHI site and treated as decribed in the legend to Fig. 2. 20 ng of labelled fragment were used for a Maxam-Gilbert G+A sequencing reaction (Maxam and Gilbert 1980).

Acknowledgements. We are grateful to R. Kahmann and C. Koch for

repeated gifts of purified FIS protein. We thank M. Hearne for the synthesis of oligonucleotides. This work was in part supported by SFB344, project B6, of the Deutsche Forschungsgemeinschafl.

References G

TA A T A C TnnCG..T..CGnnA

FIS consensussequence 200

TTATACACAAC~C

220 TGAACAACAGTTGTTC T TTGGATAA

AATATGTGTTGAGTTTTTGACTTGTTGTCAACAAGAAAC CTATT

DNaseI footprint MPE footprint Fig. 4. FIS binding site between DnaA boxes R2 and R3 in oriC of E. coil

Asai T, Chen C-P, Nagata T, Takanami M, Imai M (1992) Transcription in vivo within the replication origin of the Escherichia coli chromosome: a mechanism for activating initiation of replication. Mol Gen Genet 231:169-178 Boye E, Lyngstadaas A, Loebner-Olesen A, Skarstad K, Wold S (1992) Regulation of DNA replication in Escherichia coll. In: FanningE, KnippersR, WinnackerE-L (eds) DNA Replicationand the Cell Cycle.43. MosbacherKolloquium.Springer,Berlin,New York, in press

$6 Bramhill D, Kornberg A (1988) Duplex opening by dnaA protein at novel sequences in initiation ofreplication at the origin ofthe E. coli chromosome. Cell 52:743-755 Buhk HJ, Messer W (I 983) Replication origin region of Escherichia coli: Nucleotide sequence and functional units. Gene 24:265-279 Filu_towicz M, Roll J (1990) The requirement of IHF protein for exlrachromosomal replication of the Escherichia coli oriC in a mutant deficient in DNA polymerase I activity. New Biologist 2:818-827 Filutowicz M, Ross W, Wild J, Gourse RL (1992) Involvement of Fis protein in replication of the Escherichia coli chromosome. J Bacteriol 174:398-407 Finkel SE, Johnson RC (1992) The Fis protein: It's not just for DNA inversion anymore. Mol Microbiol: in press Fuller RS, Funnell BE, Kornberg A (1984) The dnaA protein complex with the E. coli chromosomal origin (oriC) and other sites. Cell 38:889-900 Funnell BE, Baker TA, Komberg A (1987) In vitro assembly of a prepriming complex at the origin of the Escherichia coli chromosome. J Biol Chem 262:10327-10334 GiUe H, Egan JB, Roth A, Messer W (1991) The FIS protein binds and bends the origin of chromosomal DNA replication, oriC, of Escherichia coll. Nucl Acids Res 19:4167-4 172 Gille H, Messer W (1991) Localized unwinding and structural pertubations in the origin of replication, oriC, ofEscherichia coli in vitro and in vivo. EMBO J 10:1579-1584 Hanahan D (1983) Studies on transformation of Escherichia coli with plasmids. J Mol Biol 166:557-580 Hertzberg RP, Dervan PB (1984) Cleavage of DNA with Methidiumpropyl-EDTA-Iron(I1): Reaction conditions and product analysis. Biochemistry 23:3934-3945 Hirota Y, Oka A, Sugimoto K, Asada K, Sasaki H, Takanami M (1981) Structural organization of the region essential for autonomous replication and the recognition frame model. ICNUCLA Syrup Mol Cell Biol 22:1-12 Holz A, Schaefer C, Gille H, Jueterbock W-R, Messer W (1992) Mutations in the DnaA binding sites of the replication origin of Escherichia coll. Mol Gen Genet 233:81-88 Johnson RC, Simon MI (1985) Hin-mediated site-specific recombination requires two 26 bp recombinational sites and a 60 bp recombinational enhancer. Cell 41:781-791

Kahrnann R, Rudt F, Koch C, Mertens G (1985) G inversion in bacteriophage Mu DNA is stimulated by a site within the invertase gene and a host factor. Cell 41"771-780 KSlling R, Gielow A, Seufert W, Kiicherer C, Messer W (1988) AsnC, a multifunctional regulator ofgenes located around the replication origin ofEscherichia coli, oriC. Mol Gen Genet 212:99-104 Kunkel TA (1985) Rapid and efficient site-specificmutagenesiswithout phenotypic selection. Proc Natl Acad Sci USA 82:488-492 Matsui M, Oka A, Takanami M, Yasuda S, Hirota Y (1985) Sites of dnaA protein-binding in the replication origin of the E. coli K-12 chromosome. J Mol Biol 184:529-533 Maxam AM, Gilbert W (1980) Sequencing end-labeled DNA with base-specific chemical cleavages. Methods Enzymo165:499-560 Messer W, Egan JB, Gille H, Holz A, Schaefer C, Woelker B (1991) The complex of oriC DNA with the DnaA initiator protein. Res Microbiol 142:119-125 Messer W, Noyer-Weidner M (1988) Timing and targeting: The biological functions of Dam methylation in E. coll. Cell 54:735-737 Messer W, Seufert W, Schaefer C, Gielow A, Hartmann H, Wende M (1988) Functions of the DnaA protein ofEscherichia coli in replication and transcription. Biochim Biophys Acta 951:351-358 Oka A, Sasaki H, Sugimoto K, Takanami M (1984) Sequence organization of replication origin of the Escherichia coli K-12 chromosome. J Mol Bio1176:443-458 Polaczek P (1990) Bending of the origin of replication ofE. coli by binding oflHF at a specific site. New Biol 2:265-271 Seufert W, Messer W (1987a) DnaA protein binding to the plasmid origin region can substitute for primosome assembly during replication ofpBR322 in vitro. Cell 48:73-78 Seufert W, Messer W (1987b) Start sites for bidirectional in vitro replication inside the replication origin, oriC, ofEscherichia coli. EMBO J 6:2469-2472 von Freiesleben U, Rasmussen KV (1992) Involvement oflHF and F1S in the control of initiation of DNA replication in Escherichia coli. Res Microbiol: in press Zyskind JW, Cleary JM, Brusilow WSA, Harding NE, Smith DW (1983) Chromosomal replication origin from the marine bacterium Vibrio harveyi functions in Escherichia coli: oriC consensus sequence. Proc Natl Acad Sci USA 80:1164-1168.