LN I. 'b'if' ... Lu :9. ”a. - my Ir- 1.. n- . "iv .- ' .0. :fi;\: 1’. l‘.: :f .,. . . Jig; ’ .‘ ’ I 3 0 ,—.~ .m. ,. . 11” ...¢-::. ICHIGAN STA LIBRARlES Ill ll mmlfill‘ulfllu ll ill 3 1293 01555 0613 This is to certify that the dissertation entitled Mutants of the E. coli dnaA gene: Genetic and Biochemical Analysis of its Replication Activity presented by Mark D. Sutton has been accepted towards fulfillment of the requirements for Ph . D . degree in Biochemistry " -1 r ‘ a or profess r Date W MS U is an Affirmative Action/Equal Opportunity Institution 0-12771 LIBRARY Michigan State University PLACE IN RETURN BOX to remove We checkout from your record To AVOID FINES return on or betore dete due. DATE DUE M15 DUE DATE DUE |L__J [3| fiLI—fi] L 6% l fiV—ll—T MSUIeMWWmeWIWW Mutants of the E. coli dnaA Gene: Genetic and Biochemical Analysis of its Replication Activity By Mark David Sutton A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Biochemistry 1996 Abstract Mutants of the E. chi dnaA Gene: Genetic and Biochemical Analysis of its Replication Activity By Mark David Sutton The Escherichia coli DnaA protein is a sequence-specific DNA binding protein required for initiation of chromosomal replication. Numerous plasmids and bacteriophage including pSClOl, R6K, F factor, and P1 also require DnaA protein for replication. Although the dnaA gene was identified three decades ago, only a handful of mutant alleles have been described. Furthermore, very little is known regarding the domain structure of DnaA protein at the functional or structural level. To identify functional domains of DnaA, novel mutants of the dnaA gene were isolated. These were obtained using two independent approaches, both of which relied on the pSC101 origin which is dependent on the dnaA gene for replication. Genetic and biochemical characterization of these novel alleles suggest that the DnaA protein contains four functionally distinct domains. Detailed characterization of one class of missense mutations that mapped to the C-terminal 85 residues indicated that this region is required for DNA binding. To Laune, my wife and best friend Acknowledgements My time here has been made enjoyable thanks to many individuals. First, I would like to thank my thesis advisor, Dr. Jon Kaguni, for his guidance and advice, both scientific and general, over the past five years. The freedom he afforded me enabled me to pursue many different aspects of my project while simultaneously teaching me the importance of self discipline and critical thought. i would like also to thank my “second” thesis advisor, Dr. Laurie Kaguni. Her many suggestions and long standing interest in my thesis project were much appreciated. Also central to my scientific development was the instruction of my guidance committee. Drs. Arnold Revzin, John Wang, Zachary Burton, and Loren Snyder were available always for advice and ensured that l maintained focus in my research. I would like also to thank my many friends and comrades who have helped me through the years. Carol Smith and Julie Oesterle have resolved many administrative dilemmas for me. I am also indebted to Kevin Carr for his thankless teachings in the use of a computer. His blunt manner and strong skepticism have been appreciated. But remember Kevin, “IRON” Mike Ditka is the personification of American Football. i would like also to express my appreciation to Carla Margulies for her many suggestions and constant enthusiasm, Ben Hummel and his family for their friendship, Richie Halberg, one of the few great Bears fans in America, for the Sunday Bears’ game get- togethers, Mark Kadrofski and Patty Voss for their generous donations of the occasionally needed materials and colorful discussions about football, and Cindy Peterson for her unique slant on things. i am also grateful to Wenge Zhang, Joe Lipar, David Lewis, Andrea Williams, Carol Farr, Yuxun Wang, Anthony Lagina, Matias Vicente, and Li Fan. iv This section of my thesis would not be complete without acknowledging my parents. Their tireless and unwavering support has and always will be appreciated. Last but not least, I must thank my wife, Laurie. Without her love, sacrifice, and unconditional support, completion of this thesis would not have been possible. For this and countless other things, I will be forever indebted to her. Table of Contents List of Tables ........................................................ xiii List of Figures ....................................................... xv List of Abbreviations ................................................ xviii Chapter I. Literature Review ........................................... 1 Introduction .................................................... 2 Principles of Replication ......................................... 3 The Bacterial Cell Cycle ......................................... 3 The Escherichia coli dnaA Gene ................................. 6 DnaA Protein ................................................. 10 Functional Characteristics of DnaA Protein ................. 10 Physical Characteristics of DnaA Protein ................... 13 Role of DnaA in Replication of Plasmids and Bacteriophage .. 14 Mutations of dnaA ....................................... 14 DnaA as a Transcription Factor .................................. 19 The Escherichia coli Origin of Replication, oriC .................... 20 Physical Characteristics of oriC ........................... 20 Transcription in and Around oriC .......................... 23 in vitro Replication of oriC Minichromosomes ..................... 27 Regulation of Replication ....................................... 30 Replication in the Absence of dnaA Function ...................... 33 Stable DNA Replication .................................. 33 vi Integrative Suppression .................................. 34 References ................................................... 35 Chapter II. Novel Alleles of the Escherichia coli dnaA gene That Are Defective in Replication of pSC101 but Not of oriC ................ 47 Abstract ...................................................... 48 introduction ................................................... 49 Materials and Methods ......................................... 53 Bacterial strains and plasmids ............................ 53 in vitro mutagenesis of the dnaA gene ..................... 53 Genetic assay to identify mutations of dnaA defective in pSC101 maintenance .............................. 57 Chemical cleavage of mismatched bases .................. 57 DNA sequence analysis .................................. 60 Quantitative immunoblot analysis of DnaA protein ........... 6O Assay of in vitro replication ............................... 61 Results ....................................................... 62 Isolation of five novel dnaA mutations ...................... 62 Nucleotide sequence determination of mutant dnaA alleles . . . 65 Replication activity of the novel dnaA alleles in a AdnaA strain ............................................. 65 An elevated level of DnaA protein is required for oriC replication compared with that required for pSC101 replication 68 Replicon-specific defect of the novel dnaA alleles ........... 7O Suppression of the temperature sensitive phenotype for pSClO1 maintenance by elevated expression ........ 73 vii B-galactosidase assays ................................. 132 Quantitative immunoblot analysis of mutant DnaA proteins .. 132 S1 nuclease protection assays ........................... 132 Determination of relative pCM128 abundance ............. 133 Results ...................................................... 134 Identification of nonsense and deletion mutations of the Escherichia coli dnaA gene ........................ 134 Mutations mapping within or near the P-loop motif exhibit a cold-sensitive heteropolyploid phenotype ........... 138 The C-terminus of DnaA is required for repression of dnaA gene expression ............................ 141 Quantitative immunoblot analysis of mutant DnaA proteins .. 144 Identification of a third promoter directing expression of the dnaA gene in pACYC184 .......................... 148 The N-terminus of DnaA protein is essential for replication activity .......................................... 151 DnaA protein contains two domains that are alternately dispensable for replication from the pSC101 origin 151 Determination of the relative abundance of pCM128 when maintained by different dnaA alleles ................ 157 Truncated dnaA mutations are inactive for replication from oriC ............................................. 1 58 Discussion ................................................... 1 61 References .................................................. 1 70 Chapter V. identification of Essential Residues in the DNA Binding Domain of the Escherichia coli DnaA Protein .................... 174 Abstract ..................................................... 1 75 Introduction .................................................. 1 76 Materials and Methods ........................................ 178 Bacterial strains ........................................ 178 Recombinant DNA techniques ........................... 178 Qverexpression and purification of recombinant proteins . . . . 180 Protein determinations .................................. 182 Southwestern blotting ................................... 183 DNA binding assays .................................... 184 Competitive gel mobility shift assay ....................... 185 ATP binding assays ..................................... 185 Assay of in vitro replication .............................. 186 Results ...................................................... 187 The C-terminal 89 residues of DnaA are required for specific DNA binding .............................. 187 Mutants of DnaA protein with substitutions within the C-terminal 89 residues are defective in DNA binding .......................................... 195 T435M is defective in specific DNA binding activity ......... 199 Mutant proteins are poorly active for in vitro replication of an oriC plasmid ................................ 207 T435M is active for ATP binding .......................... 208 T435M augments limiting levels of wild type DnaA for in vitro oriC replication ............................ 213 Discussion ................................................... 216 The DNA binding domain of DnaA protein is within the C-terminal 89 residues of DnaA protein ............. 216 xi Proposed secondary structure for the DnaA binding domain of the E. coli DnaA protein ......................... 218 References .................................................. 221 Chapter VI. Summary and Perspectives .............................. 224 References .................................................. 229 xii List of Tables Chapter II Table 1. Bacterial strains and plasmids ................................ 54 Table 2. Inability of plasmid-home dnaA mutations to confer temperature resistant growth to E. coli M8206 (pSC101) ...................... 66 Table 3. Nucleotide and deduced amino acid alterations of five dnaA alleles ........................................................ 67 Table 4. Plasmid-bome dnaA mutations are temperature sensitive for pSC101 maintenance ....................................... 69 Table 5. A subclass of dnaA mutations is specifically defective in pSC101 maintenance .................................................. 74 Table 6. Summary of mutant dnaA phenotypes ......................... 80 Chapter III Table 1. Bacterial strains, plasmids, and bacteriophage ................. 93 Table 2. Agt10 derivatives containing the pSC101 replication origin are able to plate on a nonlysogenic and lysogenic host quantitatively . .. 99 Table 3. Nucleotide and deduced amino acid alterations of novel dnaA missense mutations .......................................... 109 Table 4. Summary of replication phenotypes of novel dnaA mutations . . . . 112 xiii Chapter IV Table 1. Bacterial strains and plasmids ............................... 130 Table 2. Summary of nucleotide sequence analysis of nonsense and in-frame dnaA alleles ......................................... 137 Table 3. Substitutions mapping to the P-loop motif exhibit a cold-sensitive heteropolyploid phenotype ....................... 139 Table 4. Mutations mapping to the C-terrninus of DnaA are defective for regulation of expression of the dnaA’-’IacZ translational fusion in viva ................................................. 145 Table 5. Truncated dnaA mutations are unable to affect expression of a dnaA’-’IacZ translational fusion in vivo .......................... 147 Table 6. Quantitative immunoblot analysis of mutant DnaA proteins ...... 149 Table 7. dnaAA237-378 and dnaA-am361 are active for replication of a lambda derivative from the inserted pSC101 origin ............... 155 Table 8. Truncated dnaA mutations are able to maintain pSC101 ........ 156 Table 9. Relative abundance of pCM128 in various dnaA mutants ....... 159 Table 10. Truncated dnaA mutations are inactive for replication from oriC ......................................................... 160 Chapter V Table 1. Plasmid DNAs ............................................. 179 Table 2. Quantitative Southwestern blot analysis of truncated forms of DnaA ................................................ 194 Table 3. Mutant proteins with substitutions in the C-terminal region of DnaA protein are defective in binding to oriC .................. 196 xiv List of Figures Chapter I Figure 1. Schematic representation of the relationship between the bacterial cell cycle and cell division ............................... 5 Figure 2. Schematic representation of the dnaA operon .................. 9 Figure 3. Comparison of the amino acid sequence of fifteen different dnaA homologs ............................................... 12 Figure 4. Summary of dnaA mutations and corresponding phenotypes . . . . 17 Figure 5. Consensus sequence of the minimal chromosomal origin ....... 22 Figure 6. Promoters in and around oriC ................................ 25 Figure 7. Proposed model for initiation of chromosomal replication in E. coli ...................................................... 29 Chapter II Figure 1. Structural organization of oriC and the pSC101 replication origin ........................................................ 51 Figure 2A. Map of the dnaA gene ..................................... 59 Figure 2B. p08596* construction ..................................... 59 Figure 3. Genetic assay to identify E. coli dnaA mutants defective in pSC101 maintenance at 415°C ................................. 64 Figure 4. Quantitative immunoblot analysis of DnaA protein .............. 72 Figure 5. in vitro oriC replication activity of novel DnaA mutant proteins . . . . 78 XV Chapter III Figure 1A. Construction of kgt10 derivatives containing the pSC101 origin ........................................................ 98 Figure 13. Construction of a lambda derivative expressing phage 434 repressor and capable of lysogeny .............................. 98 Figure 2. Lytic growth of kpCM128 is dnaA-dependent .................. 102 Figure 3. Genetic selection for mutations of dnaA defective in pSC101 replication ................................................... 104 Figure 4. Scheme for nucleotide sequence analysis of dnaA alleles ...... 108 Figure 5. Summary of positions of mutations of novel dnaA alleles ....... 119 Chapter IV Figure 1. Western blot analysis of truncated forms of DnaA protein ....... 136 Figure 2. in viva DNA binding assay .................................. 143 Figure 3. Identification of dnaAP1’ using an S1 nuclease protection assay .............................................. 153 Figure 4. DnaA protein has four functionally distinct domains ............ 163 Figure 5. Secondary structure prediction for DnaA protein ............... 168 Chapter V Figure 1. Structures of truncated forms of DnaA protein ................. 189 Figure 2. DnaA protein binds preferentially to DNA fragments containing oriC by Southwestern blot analysis ............................. 191 Figure 3. Southwestern blot analysis of truncated forms of DnaA protein . . 193 Figure 4. Mutant forms of DnaA protein are defective in DNA binding ..... 198 Figure 5. T435M is defective in high affinity DNA binding ................ 201 xvi Figure 6. T435M does not form discrete complexes by gel mobility shift analysis ................................................. 204 Figure 7. Competitive gel mobility shift analysis of DnaA and T435M ...... 206 Figure 8. Mutant forms of DnaA are poorly active for in vitro replication from oriC .................................................... 210 Figure 9. T435M binds ATP with high affinity ........................... 212 Figure 10. T435M augments limiting levels of DnaA for in vitro replication from oriC .................................................... 215 Figure 11. Secondary structure prediction of the DNA binding domain of DnaA protein ................................................. 220 xvii bp BSA Dam DnaA box D‘I‘l' EDTA Fis HEPES lHF kDa RNAP SDS Tris List of Abbreviations base pair(s) bovine serum albumin deoxyadenosine methyltransferase asymmetric 9-mer consensus sequence bound by DnaA dithiothreitol (ethylenedinitrilo)tetraacetic acid Factor for inversion stimulation N-2-Hydroxyethylpiperazine-N’-2-ethanesulfonic acid Integration host factor kflodaflon RNA polymerase sodium dodecyl sulfate Tris(hydroxymethyl)aminomethane xviii Chapter I Literature Review Introduction As accurate duplication of the genetic material is a prerequisite for cell division in all organisms, the study of the mechanisms and regulation of DNA replication is of considerable significance. DNA replication and its regulation have been studied extensively since the discovery that DNA was the genetic material (140) with the process best understood in the enteric bacterium Escherichia coli. The E. coli chromosome is a covalently closed, circular, duplex molecule of 4,700 kilobase pairs. Replication initiates at a single, unique site termed the origin of chromosome replication, or oriC, and continues in a bidirectional fashion until terminating at a site called terC located directly opposite oriC (74). Following duplication of the bacterial chromosome, cell division proceeds wherein each daughter cell receives a complete copy of the chromosome (31). Principles of Replication Less than a decade following the discovery that DNA was the genetic material (10), a model for its structure was proposed (140). Soon after, Jacob, Brenner, and Cuzin proposed a model for its duplication that is still accepted to this day (63). In this model, an initiator, that is encoded by the replicon, acts at a unique site termed the replicator to promote duplication of the replicon. To date, the best candidates for the replicator and initiator are the loci defined as oriC and dnaA, respectively. Soon after Jacob and coworkers proposed their model, it was recognized that DNA replication must be both a periodic and tightly controlled event. The most obvious mode of regulation, that of regulation of initiation, was soon demonstrated (84). Central to any model attempting to explain initiation of replication is the concept of initiation mass (30). The initiation mass is defined as the ratio of cell mass to chromosome number at the time of initiation and is thought to be constant irrespective of growth rate. Any factor that perturbs the initiation mass is expected to affect the time of initiation of replication as well. The Bacterial Cell Cycle That the initiation mass must be an important aspect of DNA replication led to the examination of different periods of the bacterial cell cycle (27, 51). This examination revealed that the time between DNA replication and cell division, termed D, is a constant period of 20 minutes (75) (Figure 1). Furthermore, complete duplication of the bacterial chromosome, termed C, requires a minimum of 40 minutes. Although a lag may occur between Figure 1. Schematic representation of the relationship between the bacterial cell cycle and cell division (135). The bacterial cell cycle consists of two distinct periods, C and D. C, the period between initiation (ini) and termination (ter) of replication, may vary at slow growth rates but is a constant equal to 40 min at rapid growth rates. D, the period between termination of replication and cell division (div), is a constant equal to 20 min and is independent of the growth rate. The time between subsequent initiation events varies as a function of the doubling time (to) of the cell (a tD>C+D, b tD=C+D, c tD;—:) ( =::) "is ter ini if" t.=35 c =40 o = 20 -_—_<: ;.>__———:<: / / / / §—:—< C/E) C; C?) C): 9% ea / div djv “1" an ini I ini I "N ter ter ter duplication of the genome and the next initiation event, it appeared that at least 60 minutes was required for cell division. That E. coli can divide at time intervals as short as 20 minutes appeared to contradict these findings. However, this apparent dilemma was resolved by the finding that initiation events occur prior to the completion of ongoing rounds of replication. A recent method corroborating this observation is flow cytometry that measures DNA mass relative to cell size. This may then be extrapolated into the number of chromosome equivalents per cell. This method indicated that rapidly growing E. coli cells contain more than one chromosome equivalent per cell, consistent with the proposal that initiation of replication may occur prior to the completion of ongoing rounds of replication, thereby permitting E. coli to divide at time intervals shorter than that required for duplication of the bacterial genome (122). The Escherichia coli dnaA Gene The Escherichia coli dnaA gene was identified nearly three decades ago as an essential gene (52, 72). Mutations mapping to the dnaA locus exhibit a temperature sensitive phenotype for growth (reviewed in reference 92). That dnaA(Ts) mutations exhibit a slow-stop phenotype following an upshift to the nonpermissive temperature (48, 72) is consistent with the requirement of the dnaA gene product for initiation of replication. This slow-stop phenotype is due to completion of ongoing rounds of replication without new initiation due to inactivation of DnaA by a temperature upshift. The E. coli dnaA gene was cloned by compiementation of the temperature-sensitive phenotype of a dnaA mutant. It maps 42 kb to the left of the chromosomal origin at 83.5 minutes on the revised map of the E. coli genome (11). It is the first in an operon that also contains dnaN, recF, and gyrB (49, 106) (Figure 2). dnaN encodes the Bsubunit of the replicative DNA polymerase, DNA polymerase III holoenzyme, recF encodes a protein involved in recombination-directed DNA repair following damage by ultraviolet light, and gyrB encodes the B subunit of DNA gyrase, a heterotetrameric type II topoisomerase. Although these genes appear to form an operon, dnaN, recF, and gyrB are also transcribed from promoters located upstream from each gene (3,4,102,104,110) Nucleotide sequence analysis of dnaA revealed an open reading frame of 467 residues with a calculated molecular mass of 52.5 kDa and deduced pl of approximately 9.6 (43). Transcription is from two promoters, dnaAP1 and dnaAP2. S1 nuclease protection analysis indicated that dnaAP2 is the stronger of the two promoters resulting in approximately 70-80% of total transcripts (25, 76). Binding of DnaA protein to a 9-mer sequence termed a DnaA box (see below) in between dnaAP1 and dnaAP2 results in repression of transcription from both promoters (138) (see “DnaA Protein as a Transcription Factor”). A greater than expected number of deoxyadenosine methyltransferase (Dam) sites were observed in the dnaA promoter region (46, 124) (see Figure 2). Immediately following DNA replication, the dnaA promoter region is naturally hemimethylated. Whereas transcription from dnaAP2 is essentially blocked in the absence of Dam methylation, transcription from dnaAP1 appears to be independent of Dam methylation (17, 76). That the dnaA promoter region does not become fully methylated for up to 1/3 of a generation, while other sites on the chromosome become fully methylated after only 1-2 minutes, has led to the proposal that the promoter region is sequestered when hemimethylated by one or more membrane associated proteins (20). This sequestration is thought Figure 2. Schematic representation of the dnaA operon. Genes and promoters are indicated by shaded boxes and filled triangles, respectively. In the blow-up of the dnaA gene (bottom) symbols are: shaded rectangle, dnaA coding sequence; bent arrows, dnaA promoters; thin ticks in the dnaA promoter region, Dam methylation sites; filled boxes, sequences bound by DnaA protein (DnaA box motifs). Q39» Q32 soon Exam .VV .VVV .YV .V 3 pm ................ .=. -._i_. w 0:; cox Dam) cox 10 to inhibit transcription of dnaA from dnaAP2 and has been proposed to serve some regulatory function. DnaA Protein Functional Characteristics of DnaA Protein DnaA protein, the product of the dnaA gene, is essential for the initiation of replication of the bacterial chromosome (38, 52, 72) as well as numerous plasmids and bacteriophage including pSC101 (32, 34, 50), R6K (143), F factor (44, 69), and P1 (44, 142). Shortly after the dnaA gene was cloned from E. coli (49), dnaA homologs from other members of the family Enterobacteriaceae were identified (reviewed in reference 92). In all, 15 homologs have been described. Based on their deduced amino acid sequence, a remarkable degree of conservation is observed (Figure 3). Biochemical characterization of DnaA protein revealed numerous functional activities. DnaA is a sequence-specific DNA binding protein that recognizes an asymmetric 9-mer consensus matching the sequence 5’—TI'ATC/ACAC/AC-3’, or very close matches (36). However, the 50-fold range in KD depending upon the sequences flanking this 9-mer indicates the importance of these flanking sequences on binding affinity (113). DnaA protein is also a nucleotide binding protein. It binds avidly to ATP and ADP (KD of 30 nM and 100 nM, respectively) in a mutually exclusive manner (116). Bound ATP is hydrolyzed slowly to ADP in the presence of DNA in a sequence-nonspecific manner. Although nucleotide binding does not appear to affect the affinity of DnaA for the DnaA box (15), the ATP-bound form is significantly more active than either the ADP-bound or non-nucleotide-bound forms for replication of an oriC minichromosome in vitro (145). However, the 11 Figure 3. Comparison of the amino acid sequences of fifteen different dnaA homologs (92). The deduced amino acid sequence of the E. coli DnaA protein is indicated relative to the consensus sequence obtained by comparison of the deduced amino acid sequences of fifteen (E. coli, S. typhimurium, S. marcescens, P. mirabilus, B. aphidicola, P. putida, B. subtilis, S. coelicolcr, M. Iuteus, C. crescentus, Fl. meliloti, Synechocystis sp., 8. burgdorferi, S. citri, and M. capricolum) different dnaA homologs. For the consensus sequence, a dot (-) represents a residue conserved (either identical residue or conservative replacement) in less than 9, a letter represents a residue conserved in 9 or more, and an underlined letter represents a residue conserved in 12 or more of the 15 homologs. See reference 92 for a description of the method used for the sequence comparison. 12 l 60 E. coli MSLSLWQQCLARLQDELPATEFSMWIRPLQAELSDNTLALYAPNRFVLDWVRDKYLNNIN consensus MSL.Lfl.Q.LA.L..EL....§..flIR.LQ.EL...TL.L.APN.FVLDfly..KYL..I 61 120 E. coli GLLTSFCGRIAPQLRFEVGTKPVTQTPQAAVTSNVAAPAQVAQTQPQRAAPSTRSGWDNV consensus .LL..F ....... L.E.y ...................................... W... 121 180 E. coli PAPAEPTYRSNVNVKHTFDNFVEGKSNQLARAAARQVADNPGGAYNPLFLYGGTGLGKTH consensus ......... S.MN.K.TEDNEME§.SN.LA.AAAR.MADNPG.AXNELELXQG.§L§KIH 181 240 E. coli LLHAVGNGIMARKPNAKVVYMHSERFVQDMVKALQNNAIEEFKRYYRSVDALLIDDIQFF Consensus LLHA!GN..M...PNAKMVXM.SEREV.DMM.ALQ.N.IEEEK.YIBSVD.LLLQDIQEE 241 300 E. coli ANKERSQEEFFHTFNALLEGNQQIILTSDRYPKEINGVEDRLKSRFGWGLTVAIEPPELE consensus A.K§..QEEFTETENALL§...QLLLTfiflfiYBKEI.GMEDELKSBF.WQL.VAIEP2ELE 301 360 E. coli TRVAILMKKADENDIRLPGEVAFFIAKRLRSNVRELEGALNRVIANANFTGRAITIDFVR consensus MIL-K191313- =l=LP-EV.FEIA-RL-SWIA-A-E- - - 43112172- 361 420 E. coli EALRDLLALQEKLVTIDNIQKTVAEYYKIKVADLLSKRRSRSVARPRQMAMALAKELTNH consensus E.LRDLL...E.LyTIDNlQK.yAEYXKle.QLLSKRBSRSEARPBQMAM.L.KELINH 421 467 E. coli SLPEIGDAFGGRDHTTVLHACRKIEQLREESHDIKEDFSNLIRTLSS consensus SLEELQD.EQQEDHIEYLHACRKLE.LR.E..DlK.DE..LIR.L.. 13 ADP-bound form may augment limiting levels of the ATP-bound form, indicating that it is able to function at certain stages of initiation, and is active in vitro at a concentration approximately four-fold higher than the ATP-bound form. It has also been reported that DnaA protein binds cAMP (KD of 1 uM) at a site distinct from the ATP and ADP binding site (54). The binding of cAMP appears to induce the release of bound ADP but not ATP and is postulated to aid in regeneration of the active form of DnaA. Furthermore, the CAMP-bound form had an increased affinity for oriC relative to the ATP-, ADP-, and non-nucleotide-bound forms. Physical Characteristics of DnaA Protein DnaA protein is isolated in two physical states: a monomeric and aggregated form (59). Although the aggregated form is poorly characterized, it is known to be a large complex. The monomer is active in in vitro replication of an oriC minichromosome; the aggregate is not, suggesting that its physical state (i.e. ratio of monomer to aggregate) may in some way determine the net level of DnaA activity. The aggregate may be activated for in vitro replication by treatment with protein denaturants such as guanidine (118) or the heat shock protein DnaK (59). Both treatments appear to activate the aggregate by its conversion to the monomeric form. Limited biochemical characterization of the aggregate indicated that it was associated with acidic phospholipids (146). It was this observation that led to the finding that the aggregated form could be monomerized also by treatment with phospholipase A2 (59), presumably by cleavage of the phospholipids to result in the destabilization of the aggregated complex. 14 That ATP can stabilize the monomeric form of DnaA and prevent its conversion into the aggregate (28) Is consistent with the ATP-bound form of DnaA being the form preferred for replication in vitro. Role of DnaA in Replication of Plasmids and Bacteriophage That a strain containing a dnaA null mutation was unable to maintain F factor (44, 69), R6K (143), pSCiO1 (32, 34, 50), or P1 (44, 142) demonstrated that these replicons are dependent upon dnaA function. Consistent with this observation is the similarity in structural organization of their origins (2, 26, 44, 143). All contain one or more DnaA box motifs in addition to an adjacent AT-rich region. Although the AT-rich region of oriC can be unwound by DnaA protein (15), its role in the unwinding of plasmid origins dependent on it for replication, with the exception of bacteriophage P1, is largely unknown. The AT-rich region of the P1 origin can be unwound by DnaA (95). That this unwinding is facilitated by the phage-encoded initiator, RepA, suggests that both proteins are involved in the unwinding process in viva. DnaA protein is dispensable yet stimulatory for replication of R1 (44, 89, 101) and ColE1-Iike plasmids (1, 34). The in viva role of DnaA protein in replication may be both versatile and widespread. Mutations of dnaA As described above (see “T he Escherichia coli dnaA Gene”), the dnaA gene was first identified by a conditional allele mapping to the dnaA locus (52, 72). This mutant exhibited a slow-stop phenotype (48, 72) indicative of a defect at the point of initiation of replication. Since its discovery, a limited collection of dnaA alleles have been reported (reviewed in reference 92), all of which exhibit 15 a conditional phenotype (Figure 4). Both the small number and the fact that five of the 14 reported alleles share a common substitution of valine for alanine at position 184 indicates that the current collection is neither representative of all functional domains nor sufficiently large for detailed structure-function studies. However, some insight into dnaA function has been gained from the genetic and biochemical characterization of these alleles. First, the thermolabile phenotype of dnaA mutants may be suppressed in an allele-specific fashion by mutations in rpaB (7) (Figure 4). rpaB encodes the B-subunit of RNA polymerase (RNAP), an essential enzyme under certain conditions for replication of oriC minichromosomes in vitro (65, 96). Although this suggests a physical interaction between DnaA protein and RNAP, no evidence for such an interaction has as yet been reported. Arguing against a DnaA-RNAP interaction is the observation that an rpaC mutation results in an increased chromosome copy number (103, 105). rpaC encodes the 8’ subunit of RNAP. That this rpaC mutation results also in an increase in expression of dnaA is presumably related to the observed increase in chromosome copy number (i.e. via a gene dosage effect). Some rpaB mutations appear to result also in an increase in chromosome copy number suggesting that a similar mechanism may account for the allele-specific suppression of dnaA(Ts) alleles by mutant rpaB alleles. Indeed, the thermolabile phenotype of some dnaA(Ts) mutations may be suppressed by an increase in their gene dosage (or level of expression) (47). In addition to rpaB, the thermolabile phenotype of dnaA(Ts) mutants may be suppressed by extragenic suppressors mapping to trxA (55), mhAl (130), squ (83, 133), or tapA (81). Furthermore, mutations of dnaA have been identified which can suppress the thermolabile phenotypes of dnaFi (phosphoribosylpyrophosphosphate [PRPP] synthase) (109) or dnaX (encoding 16 Figure 4. Summary of dnaA mutations and corresponding phenotypes. DnaA protein is represented by the shaded rectangle. The P-Ioop motif encompassing residues 172-179 and proposed to be the high affinity ATP binding site is indicated. The approximate positions of known amino acid substitutions for missense mutations of dnaA are also shown (below). Phenotypes are grouped into three different classes (above) based on allele-specific suppression of their thermolabile phenotypes by the different rpaB alleles (see text). The mutants that are most affected in synchrony of initiation map to the middle region of the protein. >.:u manic >m<=osqo=ocm 6882 : summon scoops 2-83.55... 0.83.55 mom mom 3N m km m9 \mom mofimom mom .3 A m gm mod 50m @858 mowxmos 18 the y and 1' subunits of DNA polymerase ill holoenzyme) (42) mutants. It has been speculated that these second site suppressors represent genes whose products interact physically with DnaA in the replication process. Biochemical characterization of thermolabile DnaA mutants (DnaA5 and DnaA46) has also been performed (56, 57, 58, 60, 61). DnaA5 and DnaA46 protein are activated for in vitro oriC-dependent replication by the combined action of the heat shock proteins DnaK and GrpE. The thermolabile phenotype of dnaA46 may be suppressed in viva by overexpression of graELS (68), suggesting that these heat shock proteins and chaperonins may play a role in initiation of replication of the bacterial chromosome. The requirement of DnaK for oriC-dependent replication in vitro with mutant forms of DnaA is consistent with the finding that dnaK mutants are defective for replication at the point of initiation (108). Replication activity of DnaA5 and DnaA46 protein is temperature sensitive and more labile than wild type DnaA in vitro (58, 61), leading to the speculation that the role of DnaK is to maintain DnaA (particularly mutant forms) in an active conformation. Consistent with this is the fact that DnaK is not required for oriC-dependent replication in vitro using the replication system consisting of reconstituted purified proteins unless mutant forms of DnaA are used (56, 57, 65). Finally, physiological characterization of dnaA mutants indicated that those sharing a common substitution of alanine-to-valine at position 184 (near a P-loop motif proposed to function in nucleotide binding) exhibited an asynchrony in initiation of replication (123). Asynchrony refers to the inability of a cell containing multiple origins to initiate replication at all origins simultaneously. In a wild type bacterium, all origins are initiated simultaneously, or synchronously, resulting in a 2'1 (where n=1, 2, or 3) number of chromosome equivalents (i.e. 2, 4 or 8) (121). Conversely, cells that initiate 19 asynchronously have a variable number of chromosome equivalents (i.e. 2, 3, 4, and 5). That the DnaA5 and DnaA46 proteins, both of which contain the alanine- to-valine substitution, are defective for ATP binding in vitro (58, 60), and are asynchronous for initiation in viva (123) has prompted speculation that the asynchrony is due to the common ATP binding defect. The ATP-bound form is active in in vitro replication from oriC (145) and appears to be important for proper timing of initiation of replication in viva. DnaA Protein as a Transcription Factor The role of DnaA as a transcriptional regulator is well documented (reviewed in reference 92). DnaA protein has been demonstrated either to activate or repress transcription of genes that contain one or more DnaA box motifs in either their promoter region or coding sequence. The best studied of these is the dnaA gene itself. A DnaA box is in the promoter region located in between dnaAP1 and dnaAP2. A second DnaA box is in the coding region (see Figure 2). Binding of DnaA protein to the DnaA box in the promoter region results in transcriptional repression from both dnaA promoters as measured by S1 nuclease protection (76, 138). Furthermore, transcription from dnaAP1 was repressed in viva when dnaAP2 was deleted and the DnaA box was present but not when both dnaAP2 and the DnaA box were deleted (16). This was determined using translational fusions consisting of various portions of the dnaA promoter region joined in-frame to the IacZ structural gene, and measuring B-galactosidase activity as a function of dnaA induction. The significance of the DnaA box in the coding region of the dnaA gene is not yet known, although it does not appear to affect transcription (91). 20 Other genes whose transcription is repressed by DnaA include rpaH (139), uvrB (132), ftsQ/ftsA (90), the gua operon (128), and the trp operon (8). Although DnaA protein binds to a DnaA box within the palA promoter, the effect of binding on palA transcription has not been demonstrated (137). Transcriptional repression is also observed for a number of promoters within or nearby oriC including Pan-L (5), miaC (79, 129), and aan (40, 112) (see “Transcription in and Around oriC”). Transcription of Pan-m, a promoter in oriC, is stimulated by DnaA protein (5). It also stimulates transcription of the nrd operon encoding both subunits for ribonucleotide reductase (9) and glpD (aerobic glyceraI-3-phosphate dehydrogenase) (cited in reference 92). Stimulation of transcription of the nrd operon by DnaA protein also requires Fis (Factor for inversion stimulation) protein. The Escherichia coli Origin of Replication, oriC Physical Characteristics of oriC The E. coli origin of chromosome replication, oriC, was first cloned by virtue of its ability to confer autonomous replication upon a nonreplicating DNA fragment containing a selectable antibiotic resistance cassette (144). The minimal oriC sequence, and element within it were then identified by a combination of both deletion (100) and random mutagenesis (99). These regions are well conserved in homologs of oriC in the family Enterabacteriaceae (Figure 5). Structurally, the E. coli origin of replication contains a number of sequence elements required for origin function, including four DnaA box motifs (R1-R4) (36), three AT-rich 13-mer sequences (L, M, and R) (15), 11 21 Figure 5. Consensus sequence of the minimal chromosomal origin (147). This consensus sequence is derived from six different bacterial organisms and was aligned such that the least number of changes are introduced into the consensus sequence. In the consensus, a large capital letter indicates that that nucleotide is found in all six origins; a small capital letter indicates that it is found in five of the six origins; a lowercase letter indicates that it is found in three or four of the six origins and only two different nucleotides are found at that position; n indicates that three or four of the four possible nucleotides, or that two different nucleotides and a deletion are present at that position. Bold, large capital letters located at positions 149, 242, and 267, indicate where single-base substitutions produce an ariC- phenotype. The minimal origin of E. coli is bracketed and numbers refer to positions in the E. coli sequence. The three AT-rich 13mers repeats in the 5’-end of the origin are indicated by arrows. Darn methylation (GATC) sites are undenined. Positions and orientations of the four DnaA boxes are indicated by arrows and R1, R2, R3, and R4. 22 _ Q§i_ r .4nnn.. .nnnn.. .4n>o.. .4o>a. . ..4no4. . >4444 -.e >nn44>>a>n4> .4 .0. an“: mn>4nnaaoaq>n4>>>>>a>2m>4n4=444>444>a>o 4n n. a. a. a. >. 4| I 096306 I-Ofiflfi mo 0 I 4-4-r4 >e ..mo >32. .n>n4nnnn nonn>nnn >n4n4n4> nn44n4n4 4nn4n44a n>an444n 34434>+4nnm>4nqo+4>44>mm>4noaaaaaaaa+m4mo>4>>m aaua>nanaaa IIIIIIIIIIII-v IIIIIIIIIIIIv. IIIIIIIIIIIiVnIIIIT AMMflM4M44o Q§i. n>>on.4nnonn nnnno.4nn4n4 n4nnn.4nn>nn a4n>n.4nnonn n4a>44>44n>4 n4no>.>>4n>4 an .00 ~00 a.4=>>m>qn>»=aa==4aaaa.nm>4n=an.3nqm4m>-->4m>4nonfin>4naqunannm4>4>>nn4mnm>4n>a>>qoam onaqq>a.n>n>an=n>> 44.4 >>.> >4.4 94.4 >4.) n>>n no. >>>=am a>naaaomdded4odd4mm>d>>fid>n30044::40neaonn4 .n4o)..>>n> .o4n)..>>nn .nn>4..4nna 44n>nn.>>nn .>nnn..4non .>>>n.4n4>> .f .n>>nn .4nnn4 .>nnn4 .nna44 .n>n>> 4nn>4n fifl’)’fl nn>>>. nn>>>. >>n>>. >>n>>. nonn44 4n444. I>44>>. >n4>a. >n>>4. 00444. n44>n. >>>4n4 uflsdd4u44 n .n . .n In 0 on 0" 0 0" so . .n .> o .a .> . >> 000n>0 0006’. ...n). 000n’0 m..44> .>>>o. 4II wean... ”mo n>).n4.nn4a> n>>.n4.4nn.n n>>.n4.44n.a n>>.n4.44>.n )nn»a>>44>.a >n4.n>.n44.n \ In. an). n n.an4.n n.944.n n.nn4.n n.>nn.> 4n>4na> 44a .9 .) .n .> 44 .> ..>nn. n n.na4>n ..>>n.n ..>>a.o 0 0qnn0q .nn>4.n >n4 . >n4 . nn4 . on) . no) . cr>2>61)-> one 4 ab una>3fi>w >344>Hhfi> .n.n 0n04 .999 040“ .4.0 44.0 . . anao>aaaasarobd " 04>..4nnn>ni. >4n..4nnn>nr. o)>..>>a4>4r.. 44n>.n4nnnnr>. n>>..>an4nna.. m n>4o4444nnn4.4 muoreemorme oon mesaoaeswn «mvrwacamgs mxoeuoooooea possesses auovemeuwn paexaosnoe giiogggga .Rv1rurn3§€m nozmmzmcm mmocmznm 23 deoxyadenosine methyltransferase (Darn) sites (148), two promoters (5), and binding sites each for Fis protein (33, 41, 107) and lHF (Integration Host Factor) (62, 107). The observation that mutations mapping to oriC abolished replication activity when assayed in vivo using a minichromosome but exhibited no mutant phenotype when present in the chromosome (53) suggests that their respective mechanisms of replication may differ. Based on this finding, the minimal oriC as currently defined may be incomplete. Consistent with this is the lack of incompatibility observed between the chromosomal oriC and oriC minichromosomes, even when the minichromosomes are present at high copy numbers (144). Indeed, plasmids containing the Bacillus subtilis origin exhibit incompatibility with the B. subtilis chromosome (93, 94). However, that minichromosomes initiate replication in a cell-cycle- specific manner (77), replicate bidirectionally (64), and show a similar dependence in replication on proteins known to be essential for replication of the chromosome (65) argues that much, if not all of the essential sequences are present. In summary, oriC defines a locus known to be essential for initiation of replication of the bacterial chromosome and represents the best studied and most understood of all chromosomal origins. Whether additional sequences are required for regulation or accurate initiation of chromosomal replication is a subject under continued investigation. Transcription in and Around oriC oriC is flanked by a number of promoters, many of which exhibit a striking degree of regulation by DnaA protein (Figure 6). Located to the right of oriC, mioC (modulation of initiation at oriC) encodes a 16 kDa gene product of 24 Figure 6. Promoters in and around oriC. Genes and DnaA box motifs (>; 5’-TTATCCACA—3’) in the vicinity of oriC, the chromosomal origin, are indicated (above). Relevant structural features of on‘C including the promoters within oriC (see text) are shown in the blow-up (below). Symbols are: open circles, AT-rich 13mers (L, M, and R); pentagons, DnaA boxes (Rt-R4). Binding sites for lHF and Fis protein are also indicated. 25 $6 9.9» 03.0 Al Saflllmmao mun) 26 unknown function (127). The mioC promoter contains a single DnaA box and its leftward transcription through oriC is repressed by DnaA at the point of transcriptional initiation (115, 138). The aan gene encodes a transcriptional activator of asnA, one of the two asparagine synthetases of E. coli (73). Extension of aan transcripts past mioC and into oriC is regulated by DnaA at the boundary between aan and mioC by transcriptional termination (40, 112). The gidA gene (glucose inhibition of division) is located to the left of oriC. The gidA promoter directs transcription leftward away from oriC and has a positive effect on oriC function by its ability to elevate the copy number of an oriC plasmid when contained at its proper position (6, 97). In addition to the promoters discussed above that flank oriC, two promoters (Pan-L and Foam) have been identified within it that direct transcription leftward and rightward, respectively (5, 115). An additional promoter, PoriFi’ located at the right boundary but outside of oriC directs transcription rightward, away from oriC. Whereas transcription from Paul is repressed by DnaA, transcription from Pan-R, is activated. However, transcriptional activation of Pam, is dependent upon its location in the bacterial chromosome as similar activation is not observed using oriC minichromosomes. In the process of transcription, RNA polymerase generates positive supercoils ahead of the transcription complex and negative supercoils behind (78). DnaA represses (or terminates) trancription from promoters that enter into oriC and stimulates transcription from promoters that lead away from oriC, suggesting that the role of RNAP in initiation of replication may be through facilitation of unwinding in the oriC region by an increase in its negative superhelical density. To test the effect of transcription on the function of oriC, T7 promoters were inserted at various distances from oriC and origin function was measured 27 as a function of 17 RNAP (12, 119). This analysis indicated that transcription dependent on T7 RNAP was able to enhance oriC function when the promoters were within 1.4 kb of oriC. It was speculated that the observed activation was through alteration of the DNA structure in the vicinity of oriC, thereby facilitating unwinding of the AT-rich region by DnaA protein. in vitro Replication of oriC Minichromosomes The cloning of oriC permitted the development of in vitro replication systems specific for oriC. The first such system relied on use of a crude, soluble protein fraction obtained from a dnaA(Ts) mutant (37). This system was completely dependent upon both exogenously added template DNA containing a functional oriC and DnaA protein (38). Fractionation of the crude, soluble protein fraction permitted the identification of those proteins essential for in vitro replication from oriC These include DnaA, DnaB, DnaC, primase, RNAP, gyrase, topoisomerase I, FlNase HI, single stranded DNA binding protein, and DNA polymerase III Holoenzyme. This led to the development of a replication system reconstituted with these purified proteins (65). Making use of these in vitro replication systems, a model for the biochemical mechanism of initiation of replication from oriC has emerged (Figure 7). DnaA protein binds to oriC in a cooperative manner through specific recognition of the four DnaA boxes present within the origin resulting in the formation of a large and highly ordered structure referred to as the initial complex (117). Characterization of this initial complex by electron michroscopy suggests that it contains 20-30 DnaA monomers per oriC (36). 28 Figure 7. Proposed model for lnitiatlon of chromosomal replication in E. coli (74). The four stages of DnaA-dependent initiation of DNA replication in E. coli are (1) cooperative binding of the ATP-bound form of DnaA protein to four DnaA boxes contained within oriC resulting in a large nucleoprotein structure refered to as the initial complex; (2) DnaA-dependent unwinding of the three, tandemly repeated AT-rich 13mer sequences in the left-most end of oriC resulting in conversion of the initial complex to the open complex; (3) loading of DnaB, the replicative helicase, into the unwound region followed by dissociation of DnaC protein from the DnaB-DnaC complex creating a structure termed the prepriming complex; (4) further unwinding of the DNA strands by DnaB helicase allowing subsequent priming and bidirectional replication. Both HU protein and lHF have been shown to aid DnaA protein in unwinding the AT-rich region. 5 mM ATP and elevated temperature (38°C) is required for conversion of the initial complex to the open complex. 29 INITIAL COMPLEX TEMPLATE SUPERCOILED ( PREPRIMING COMPLEX (I) \O) PRIMING 0. AND : REPLICATION 30 Although the nucleotide-free form of DnaA protein is able to form the initial complex with an efficiency similar to that observed for the ADP- and ATP- bound forms, only the ATP-bound form is capable of converting the initial complex to the open complex (15, 145). In the open complex, the AT-rich left most region of oriC becomes unwound. This unwinding also requires elevated temperature (38°C) and high levels of ATP (5 mM). Furthermore, histone-like proteins such as HU and lHF aid DnaA in the conversion of the initial complex to the open complex (62). Subsequent entry of DnaB, the replicative helicase, proposed to occur through direct physical contact between DnaA protein in the open complex and the DnaB-DnaC complex, followed by the release of DnaC results in formation of the prepriming complex (39, 88). The sole function of DnaC appears to be the inhibition of the ATPase activity of DnaB helicase prior to its loading into the open complex (141). That DnaA is dispensable following the loading of DnaB helicase was confirmed by isolation of the prepriming complex by gel filtration chromatography in the presence of potassium glutamate (145). Under these conditions, the prepriming complex lacked detectable levels of DnaA and was competent for replication. Establishment of a multi-subunit heterodimeric replication fork complex consisting of all remaining necessary replication factors fulfills the final requirement for the priming and bidirectional synthesis of daughter strands (74). Regulation of Replication Although regulation of replication has been demonstrated to occur at the point of initiation (84), the biochemical mechanism of this regulation is as yet unknown. As the initiation mass, the ratio of chromosomal origins to cell mass, 31 is thought to be a constant and independent of the growth rate (30), it has been postulated that the accumulation of a key factor signals the time of initiation. DnaA protein may act to regulate the frequency of initiation. Two separate groups have determined the intracellular concentration of DnaA protein as a function of growth rate. Whereas one reported it to be a constant irrespective of the growth rate (80), the other found that it varied as a function of the growth rate (24). The reason for this discrepancy is unknown. Consequently, two different models have emerged to attempt to explain regulation of replication. In the model proposed by Zyskind and co-workers (86), initiation of replication is regulated by the level of activity of DnaA protein (i.e. the ATP-bound form), not its concentration. This was based on their finding that the intracellular DnaA protein concentration varied as a function of the growth rate, suggesting that its intracellular concentration was not the key factor determining the initiation mass. In the model of Atlung and co-workers (45), the total concentration of free intracellular DnaA protein is proposed to regulate initiation of replication. As they detect no change in the level of DnaA protein as a function of growth rate, they propose that it must be the concentration of DnaA that is important. They suggest that DnaA protein has a lower affinity for sites within oriC relative to sites outside of oriC, and that only at high levels of DnaA are these low affinity sites bound thereby permitting initiation of replication. A number of observations support the latter model. First, in viva footprinting of oriC plasmids indicated that all but one of the four DnaA boxes (R3) of oriC were occupied for the majority of the cell cycle (111), leading to the proposal that binding of DnaA to box R3 is the rate-limiting step for initiation. Second, binding of DnaA protein to the four DnaA boxes within on‘C is highly 32 ordered (87). Consistent with the in vivo footprinting results (111), R3 is the last of the four boxes to be occupied. Finally, binding constants for DnaA protein to various individual DnaA box motifs were determined (113). In this analysis, DnaA was unable to bind specifically to box R3. Other observations support the former model in which the net level of DnaA activity regulates the frequency of initiation. Two potential candidates that may negatively regulate the activity of DnaA protein have been identified. These are Squ (83, 133) and an uncharacterized soluble factor of ~150 kDa (66). The dam gene of E. coli encodes DNA adenine methyltransferase that methylates adenine in GATC sequences. Eleven sites are in oriC. The squ gene, encoding a 22 kDa protein of 181 amino acids (18, 83), was identified by the ability of an E. coli strain containing a dam13:Tn9 allele and a mutation in squ to be transformed by a fully methylated oriC minichromosome (83). In contrast, a squt, dam" mutant is transformed inefficiently because, after one round of replication, the hemimethylated daughter molecules that are formed are not replicated further (20). It has been proposed that Squ is a membrane associated protein that binds to and sequesters hemimethylated oriC to prevent subsequent initiation events (83). Biochemical studies that support this proposal include the finding that Squ preferentially binds to oriC when hemimethylated relative to the fully methylated form (18, 125). A squ null allele suppresses the thermolabile phenotype of dnaA(Ts) mutations and an increase in its gene dosage and suppresses the cold sensitive phenotype of the hyperactive dnaAcos mutation, suggesting that Squ may also regulate the level of activity of DnaA protein (83). dnaAcos is an intragenic suppressor of dnaA46 (35). Although dnaAcos does not exhibit a thermolabile phenotype, it is hyperactive at 30°C resulting in 33 a cold sensitive phenotype. Biochemical characterization of DnaAcos protein revealed that in contrast to wild type DnaA protein that was inactivated by a soluble factor of ~150 kDa, DnaAcos was not (66, 67). This inactivation was dependent upon DNA (in a sequence-independent manner) and hydrolyzable ribonucleoside triphosphates. Based on this, it has been proposed that the ~150 kDa factor inactivates DnaA protein by stimulating its intrinsic DNA-dependent ATPase activity to convert the ATP-bound form to the less active ADP-bound form by a mechanism analogous to inactivation of Ras GTPase by GTPase-activating protein (GAP) (19). Replication in the Absence of dnaA Function Stable DNA Replication Conditions exist under which replication initiates from sites other than oriC in a DnaA-independent fashion. One, referred to as constituitive stable DNA replication, cSDR (130), results from inactivation of mhA that encodes RNase H. RNase H degrades RNA in RNA-DNA hybrids and appears to confer specificity in initiation to oriC by destruction of potential primers for replication at other sites (71, 98). On inactivation of mhA, initiation of replication appears to occur at random sites referred to collectively as oriK (29). cSDR is completely dependent upon recA function. Whereas replication from oriC is dependent upon concomitant transcription and translation (136), cSDR is independent of concomitant translation (134). A second type of SDR has been reported to occur following induction of the SOS responce. This type of SDR, termed inducible stable DNA replication, or iSDR, may drive replication for several hours (70). iSDR initiates from two origins referred to collectively as oriM (85). One is located within oriC and the 34 other near terC. iSDR is dependent upon derepression of the lexA regulon and is independent of concomitant transcription and translation. Integrative Suppression Integrative suppression is the second condition wherein replication of the chromosome is no longer dependent upon oriC. It occurs by integration of an autonomous replicon that serves as a site for initiation of chromosomal replication in the absence of oriC function (14, 21, 82). If replication from this origin is independent of dnaA function, the dnaA gene becomes dispensable as well (13, 44). Although insertion of F plasmid (131), R6K (126), or bacteriophage P1 (22, 23) into the bacterial chromosome is able to suppress the thermolabile phenotype of dnaA(Ts) mutants, only plasmid R1 (13, 44) and bacteriophage P2 (114) have been demonstrated to suppress a dnaA null or amber mutation. This observation suggests that F, R6K, and P1 are dependent on dnaA function but may not require the full complement of activities of DnaA protein as does oriC. In addition to the use of R1 suppressed, dnaA null mutants to identify DnaA-dependent replicons, they have been used to study dnaA function in cell division (13). Cell division proceeds normally in the absence of dnaA function, suggesting that DnaA does not play a significant role in the process of cell division. 10. 11. 12. 13. References Abe, M. 1980. Replication of CoIE1 plasmid deoxyribonucleic acid in a thennosensitive dnaA mutant of Escherichia coli. J. Bacteriol. 141:1024-30. Abeles, A. L., K. M. Snyder and D. K. Chattoraj. 1984. P1 plasmid replication: replicon structure. J. Mol. Biol. 173:307-324. Adachl, T., K. Mizuuchl, R. Menzel and M. Gellert. 1984. DNA sequence and transcription of the region upstream of the E. coli gyrB gene. Nucleic Acids Res. 12:6389-95. Armengod, M. E., M. Garcia-Sago and E. Lambles. 1988. TranscrIptional organization of the dnaN and recF genes of Escherichia coli K-12. J. Biol. Chem. 263:12109-12114. Asai, T., C. P. Chen, T. Nagata, M. Takanaml and M. lmai. 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. Asai, T., M. Takanaml and M. lmai. 1990. The AT richness and id transcription determine the left border of the replication origin of the . coli chromosome. EMBO J. 9:4065-4072. Atlung, T. 1984. Allele-specific suppression of dnaA(Ts) mutations by rpoB mutations in Escherichia coli. Mol. Gen. Genet. 197:125-128. Atlung, T. and F. G. Hansen. 1983. Effect of dnaA and rpaB mutations on attenuation in the trp operon of Escherichia coli. J. Bacteriol. 56:985-92. Augustin, L. 8., B. A. Jacobson and J. A. Fuchs. 1994. Escherichia coli Fis and DnaA proteins bind specifically to the nrd promoter region and affect expression of an nrd-lac fusion. J. Bacteriol. 176:378-387. Avery, O. T., C. M. MacLeod and M. McCarthy. 1944. J. Exptl. Med. 79:137. Bachmann, B. J. 1990. Linkage map of Escherichia coli K-12, edition 8. Microbiol Rev. 54:130-97. Baker, T. A. and A. Kornberg. 1988. Transcriptional activation of initiation of replication from the E. coli chromosomal origin: an RNA-DNA hybrid near oriC. Cell. 55:113-123. Bernander, R., S. Das upta and K. Nordstrém. 1991. The E. coli cell cycle and the plasmi R1 replication cycle in the absence of the DnaA protein. Cell. 64:1145-1153. 35 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 36 Bird, R. E., M. Chandler and L. Caro. 1976. Suppression of an Escherichia coli dnaA mutation by the integrated R factor R.100.1: Change of chromosome replication origin In synchronized cultures. J. Bacteriol. 126:1215-1223. Bramhill, D. and A. Kornberg. 1988. Duplex opening by dnaA protein at novel se%uences in initiation of replication at the origin of the E. coli chromosome. ell. 52:743-755. Braun, R. E., K. O'Day and A. Wright. 1985. Autoregulation of the DNA replication gene dnaA in E. coli K-12. Cell. 40:159-169. Braun, R. E. and A. Wright. 1986. DNA methylation differentially enhances the expression of one of the two E. coli dnaA promoters in vivo and in vitro. Mol. Gen. Genet. 202:246-250. Brendler, T., A. Abeles and S. Austin. 1995. A protein that binds to the P1 origin core and the oriC13mer region in a methylation-specific fashion is the product of the host squ gene. EMBO J. 14:4083-4089. Cales, C., J. F. Hancock, C. J. Marshall and A. Hall. 1988. The cytoplasmic protein GAP is implicated as the target for regulation by the res gene product. Nature. 332:548-51. Campbell, J. L. and N. Kleckner. 1990. E. coli oriC and the dnaA gene promoter are sequestered from darn methyltransferase following the passage of the chromosomal replication fork. Cell. 62:967-979. Chandler, M., L. Silver and L. Caro. 1977. Suppression of an Escherichia coli dnaA mutation by the integrated R factor R100.1: origin of chromosome replication during exponential growth. J. Bacteriol. 31:421-430. Chesney, R. H. and J. R. Scott. 1978. Suppression of a thermosensitive dnaA mutation of Escherichia coli by bacteriophage P1 and P7. Plasmid. 1:145-63. Chesney, R. H., J. R. Scott and D. Vapnek. 1979. lnte ration of the plasmid prophages P1 and P7 into the chromosome of scherichia coli. J Mol Biol. 130:161-73. Chiaramello, A. E. and J. W. Zyskind. 1989. Expression of Escherichia coli dnaA and mioC genes as a function of growth rate. J. Bacteriol. 171 :4272-4280. Chiaramello, A. E. and J. W. Zyskind. 1990. Couplin of DNA replication to growth rate in Escherichia coli: a possible ro e for guanosine tetraphosphate. J. Bacteriol. 172:2013-2019. Cohen, S. N. and A. C. Change. 1977. Revised interpretation of the origin of the pSC101 plasmid. J. Bacteriol. 132:734-737. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 37 Cooper, S. and C. E. Helmstetter. 1968. Chromosome replication and the division cycle of Escherichia coli B/r. J. Mol. Biol. 1:519-540. Crooke, E., C. E. Castuma and A. Kornberg. 1992. The chromosome origin of Escherichia coli stabilizes DnaA protein during rejuvenation by phospholipids. J. Biol. Chem. 267:16779-82. de Massy, B., O. Fayet and T. Kogoma. 1984. Multiple origin usage for DNA replication in sdrA(mh) mutants of Escherichia coli K-12. Initiation in the absence of oriC. J. Mol. Biol. 178:227-236. Donachie, W. D. 1968. Relationship between cell size and time of initiation of DNA replication. Nature. 219:1077-1079. Donachie, W. D. and K. J. Be 9. 1989. Chromosome partition in Escherichia coli requires postrep ication protein synthesis. J. Bacteriol. 171:5405-9. Felton, J. and A. Wright. 1979. Plasmid SC101 replication in integratively suppressed cells requires dn function. Mol. Gen. Genet. 175:231-233. Filutowicz, M., W. Ross, J. Wild and R. L. Course. 1992. Involvement of Fis protein in replication of the Escherichia coli chromosome. J. Bacteriol. 174:398-407. Frey, J., M. Chandler and L. Caro. 1979. The effects of an Escherichia coli dnaAts mutation on the re lication of the plasmids ColE1 pSCtO1, R100.1 and RTF-TC. Mol. Gen. enet. 174:117-26. Frey, J., M. Chandler and L. Caro. 1984. Overinitiation of chromosome and plasmid replication in a dnaAcos mutant of Escherichia coli K12. Evidence for dnaA-dnaB interactions. J. Mol. Biol. 179:171-183. Fuller, R. S., B. E. Funnell and A. Kornberg. 1984. The dnaA protein complex with the E. coli chromosomal replication origin (oriC) and other DNA sites. Cell. 38:889-900. Fuller, R. S., J. M. Ka uni and A. Kornberg. 1981. Enzymatic replication of the origin 0 the Escherichia coli chromosome. Proc. Natl. Acad. Sci., U. S. A. 78:7370-7374. Fuller, R. S. and A. Kornberg. 1983. Purified dnaA protein in initiation of replication at the Escherichia coli chromosomal origin of replication. Proc. Natl. Acad. Sci., U. S. A. 80:5817-5821. Funnell, B. E., T. A. Baker and A. Kornberg. 1987. In vitro assembly of a prepriming complex at the origin of the Escherichia coli chromosome. J. Biol. Chem. 262:10327-10334. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 38 Gielow, A., C. Kucherer, R. Kolling and W. Messer. 1988. Transcription in the region of the replication origin, oriC, of Escherichia coli: termination of aan transcripts. Mol. Gen. Genet. 214:474-481. Gille, H., J. B. Egan, A. Roth and W. Messer. 1991. The FIS protein binds and bends the origin of chromosomal DNA replication, oriC, of Escherichia coli. Nucleic Acids Res. 19:4167-4172. Ginés—Candelaria, E., A. Bllnkova and J. R. Walker. 1995. Mutations in Escherichia coli dnaA which suppress a dnaX(Ts) polymerization mutation and are dominant when located in the chromosomal allele and recessive on plasmids. J. Bacteriol. 1 77:705-715. Hansen, E. B., F. G. Hansen and K. von Meyenburg. 1982. The nucleotide sequence of the dnaA gene and the first part of the dnaN gene of Escherichia coli K-12. Nucleic Acids Res. 10:7373-7385. Hansen, E. B. and M. B. Yarmolinsky. 1986. Host participation in plasmid maintainance: Dependence upon dnaA of replicons derived from P1 and F. Proc. Natl. Acad. Sci., U. S. A. 83:4423-4427. Hansen, F. G., B. B. Christensen and T. Atlung. 1991. The initiator titration model: computer simulation of chromosome and minichromosome control. Res. Microbiol. 142:161-7. Hansen, F. G., E. B. Hansen and T. Atlung. 1982. The nucleotide seguence of the dnaA gene promoter and of the adjacent rpmngne, co ing for the ribosomal protein L34, of Escherichia coli. EMBO . 1:1043-1048. Hansen, F. G., S. Koefoed and T. Atlung. 1992. Cloning and nucleotide sequence determination of twelve mutant dnaA genes of Escherichia coli. Mol. Gen. Genet. 234:14-21. Hansen, F. G. and K. V. Rasmussen. 1977. Regulation of the dnaA product in Escherichia coli. Mol. Gen. Genet. 155:219-225. Hansen, F. G. and K. van Meyenburg. 1979. Characterization of the dnaA, gyrB and other enes in the dnaA region of the Escherichia coli chromsome on specia ized transducing phages Mna. Mol. Gen. Genet. 175:135-144. Hasunuma, K. and M. Sekiguchi. 1977. Replication of plasmid pSC101 in Escherichia coli K12: requirement for dnaA function. Mol. Gen. Genet. 154:225-30. Helmstetter, C. E. 1968. DNA synthesis during the division cycle of rapidly growing Escherichia coli B/r. J. Mol. Biol. 1:507-518. Hirota, Y., A. Ryter and F. Jacob. Thennosensitive mutants of E. coli affected in the processes of DNA 3 nthesis and cellular division. In (eds.), Cold Spring Harbor Symp. Quant. iol., 1968. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 39 Holz, A., C. Schaefer, H. Gille, W. R. Jueterbock and W. Messer. 1992. Mutations in the DnaA binding sites of the replication origin of Escherichia coli. Mol. Gen. Genet. 233:81-88. Hughes, P., A. Landoulsi and M. Kohiyama. 1988. A novel role for cAMP in the control of the activity of the E. coli chromosome replication initiator protein, DnaA. Cell. 55:343-350. Hupp, T. R. and J. M. Kaguni. 1988. Suppression of the Escherichia coli dnaA46 mutation by a mutation in trxA, t e gene for thioredoxin. Mol. Gen. Genet. 213:471-478. Hupp, T. R. and J. M. Kaguni. 1993. Activation of DnaA5 protein by GrpE and DnaK heat shock proteins in initiation of DNA replication in Escherichia coli. J. Biol. Chem. 268:13137-13142. Hupp, T. R. and J. M. Kaguni. 1993. Activation of mutant forms of DnaA protein of Escherichia coli by DnaK and GrpE proteins occurs prior to DNA replication. J. Biol. Chem. 268:13143-13150. Hupp, T. R. and J. M. Kaguni. 1993. DnaA5 protein is thermolabile in initiation of replication from the chromosomal origin of Escherichia coli. J. Biol. Chem. 268:13128-13136. Hwang, D. S., E. Crooke and A. Kornberg. 1990. Aggregated dnaA protein is dissociated and activated for DNA replication by phospholipase or dnaK protein. J. Biol. Chem. 265:19244-19248. Hwang, D. S. and J. M. Kaguni. 1988. Interaction of dnaA46 protein with a stimulatory protein in replication from the Escherichia coli chromosomal origin. J. Biol. Chem. 263:10633-10640. Hwang, D. S. and J. M. Kaguni. 1988. Purification and characterization of the dnaA46 gene product. J. Biol. Chem. 263:10625- 10632. Hwang, D. S. and A. Kornbeaz origin of Escherichia coli by Dn Chem. 267:23083-23086. Jacob, F., S. Brenner and F. Cusln. 1963. On the regulation of DNA replication in bacteria. Cold Spring Harbor Symp. Quant. Biol. 28:329- 348. Ka uni, J. M., R. S. Fuller and A. Kornberg. 1982. Enzymatic rep ication of E. coli chromosomal origin is bidirectional. Nature. 296:623-627. Kaguni, J. M. and A. Kornberg. 1984. Replication initiated at the origin (oricg of the E. coli chromosome reconstituted with purified enzymes. ell. 38:183-190. r . 1992. Opening of the replication protein with protein HU or HF. J. Biol. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 76. 77. 78. 79. 40 Katayama, T. and E. Crooke. 1995. DnaA protein is sensitive to a soluble factor and is specifically inactivated for initiation of in vitro replication of the Escherichia coli minichromosome. J. Biol. Chem. 270:9265-9271. Katayama, T. and A. Kornberg. 1994. Hyperactive initiation of chromosomal replication in vivo and in vitro by a mutant initiator protein, DnaAcos, of Escherichia coli. J. Biol. Chem. 269:12698-12703. Katayama, T. and T. Nagata. 1991. Initiation of chromosomal DNA replication which is stimulated without oversupply of DnaA protein in Escherichia coli. Mol. Gen. Genet. 226:491-502. Kline, B. C., T. Kogoma, J. E. Tam and M. S. Shields. 1986. Requirement of the Escherichia coli dnaA gene product for plasmid F maintenance. J. Bacteriol. 168:440-3. Kogoma, T. and K. G. Lark. 1975. Characterization of the replication of Escherichia coli DNA in the absence of protein synthesis: stable DNA replication. J. Mol. Biol. 94:243-256. Kogoma, T., N. L. Subia and K. von Meyenburg. 1985. Function of nbonuclease H in initiation of DNA replication in Escherichia coli K-12. Mol. Gen. Genet. 200:103-109. Kohiyama, M., D. Cousin, A. Ryter and F. Jacob. 1966. [Thermosensitive mutants of Escherichia coli K 12. l. Isolation and rapid characterization]. Ann Inst Pasteur (Paris). 110:465-486. Kolling, R. and H. Lother. 1985. Aan: an autogenously regulated activator of asparagine synthetase A transcription in Escherichia coli. J. Bacteriol. 164:310-5. Kornberg, A. and T. A. Baker. 1992. DNA Replication. 931. Kubitschek, H. E. 1974. Estimation of the D period from residual division after exposure of exponential phase bacteria to chloramphenicol. Mol. Gen. Genet. 135:123-130. Kucherer, C., H. Lother, R. Kolling, M. A. Schauzu and W. Messer. 1986. Regulation of transcription of the chromosomal dnaA gene of Escherichia coli. Mol. Gen. Genet. 205:115-121. Leonard, A. C. and C. E. Helmstetter. 1986. Cell cycle-specific replication of Escherichia coli minichromosomes. Proc. Natl. Acad. Sci. U. S. A. 83:5101-5. Liu, L. F. and J. C. Wangj 1987. Sugercoiling of the DNA template during transcription. Proc. atl. Acad. ci., U. S. A. 84:7024-7027. Lebner-Olesen, A., T. Atlung and K. V. Rasmussen. 1987. Stability and replication control of Escherichia coli minichromosomes. J. Bacteriol. 169:2835-2842. 80. 81. 82. 83. 84. 85. 86. 87. 88. 89. 90. 91. 92. 41 Lebner-Olesen, A., K. Skarstad, F. G. Hansen, K. von Meyenburg and E. Boys. 1989. The DnaA protein determines the initiation mass of Escherichia coli K- 12. Cell. 57:881-889. Louarn, J., J. P. Bouche, J. Patte and J. M. Louarn. 1984. Genetic inactivation of topoisomerase I suppresses a defect in initiation of chromosome replication in Escherichia coli. Mol. Gen. Genet. 195:170-174. Louarn, J., J. Patte and J. M. Louarn. 1982. Suppression of Escherichia coli dnaA46 mutations by integration of plasmid R100.1. der1ivatives: constraints imposed by the replication terminus. J. Bacteriol. 15 :657-67. Lu, M., J. L. Campbell, E. Boys and N. Kleckner. 1994. Squ: a negative modulator of replication initiation in E. coli. Cell. 77:413-426. Maaloe, O. and N. O. Kieldgaard. 1966. Control of Macromolecular Synthesis. Magee, T. R., T. Asai, D. Malka and T. Ko oma. 1992. DNA damage-inducible origins of DNA replication in scherichia coli. EMBO J. 11 :4219-4225. Mahaffy, J. M. and J. W. Zyskind. 1989. A model for the initiation of replication in Escherichia coli. J. Theor. Biol. 140:453-477. Margulies, C. and J. M. Kaguni. in press. J. Biol. Chem. Marszalek, J. and J. M. Kaguni. 1994. DnaA protein directs the binding of DnaB protein in initiation of DNA replication in Escherichia coli. J. Biol. Chem. 269:4883-4890. Masai, H. and K. Arai. 1987. RepA and DnaA proteins are required for initiation of R1 plasmid replication in vitro and interact with the oriFi sequence. Proc. Natl. Acad. Sci., U. S. A. 84:4781-4785. Masters, M., T. Paterson, A. G. Popplewell, T. Owen-Hughes, J. H. Pringle and K. J. Begg. 1989. The effect of DnaA protein levels and the rate of initiation at oriC on transcription originating in the fisO and ffsA genes: in vivo experiments. Mol. Gen. Genet. 216:475-483. Messer, W., W. Seufert, C. Schaefer, A. Gielow, H. Hartmann and M. Wende. 1988. Functions of the DnaA protein of Escherichia coli in replication and transcription. Biochim. Biophys. Acta. 951:351-358. Messer, W. and C. Weigel. Initiation of chromosome replication. In Neidhardt, F. C., J. L. Ingraham, K. B. Low, B. Magasanik, M. Schaechter and H. E. Umbarger (eds.), Escherichia coli and almonella typhimurium. Cellular and Molecular Biology. 1996. American Society for Microbiology, Washington, D. C. 93. 94. 95. 96. 97. 98. 99. 100. 101. 102. 103. 104. 42 Moriya, S., T. Atlung, F. G. Hansen, H. Yoshikawa and N. Ogasawara. 1992. Cloning of an autonomously replicating sequence (ars) from the Bacillus subtilis chromosome. Mol. Micro. 6:309-315. Morl ya, 8., T. Fukuoka, N. Ogasawara and H. Yoshikawa. 1988. Re ulation of initiation of the chromosomal replication by DnaA- boxes in t e origin region of the Bacillus subtilis chromosome. EMBO J. 7: 291 1-7. Mukhopadhyay, G., K. M. Carr, J. M. Kaguni and D. K. Chattorai. 1993. Open-complex formation by the host initiator, DnaA, at the origin of P1 plasmid replication. EMBO J. 12:4547-4554. Ogawa, T., T. A. Baker, A. van der Ends and A. Kornberg. 1985. Initiation of enzymatic replication at the origin of the Escherichia coli chromosome: contributions of RNA polymerase and primase. Proc. Natl. Acad. Sci., U. S. A. 82:3562-3566. Ogawa, T. and T. Okazaki. 1991. Concurrent transcription from the gid and miaC promoters activates replication of an Escherichia coli minichromosome. Mol. Gen. Genet. 230:193-200. Ogawa,T ., G. G. Pickett, T. Kogoma and A. Kornberg. 1984. RNase H confers specificity In the dnaA- -dependent initiation of replication at the unique origin of the Escherichia coli chromosome in vivo and in vitro. Proc. Natl. Acad. Sci., U. S. A. 81:1040-1044. Oka, A., H. Sasaki, K. Su imoto and M. Takanaml. 1984. Sequence organization of rep ication origin of the Escherichia coli K- 12 chromosome. J. Mol. Biol. 176:443-458. Oka, A., K. Sugimoto, M. Takanaml and Y. Hirota. 1980. Replication origin of the Escherichia coli K-12 chromosome: the size and structure of the minimum DNA segment carrying the information for autonomous replciation. Mol. Gen. Genet. 178:9-20. Ortega, S., E. Lanka and R. Diaz. 1986. The involvement of host replication proteins and of specific origin sequences in the in vitro replication of miniplasmid R1 DNA. Nucleic Acids Res. 14:4865-79. Perez-Roger, |., M. Garcia-Sogo, N.-A. J. P., C. Lopez-Acedo, F. Macian and M. E. Armengod. 1991. Positive and negative regulatory elements in the dnaA-dnaN-recF operon of Escherichia coli. Biochimie. 73:329-334. Petersen, S. K. and F. G. Hansen. 1991. A missense mutation in the rpoC gene affects chromosomal replication control in Escherichia coli. J. Bacteriol. 173:5200-6. Ouifiones, A. and W. Messer. 1988. Discoordinate gene expression in the dnaA-dnaN operon of Escherichia coli. Mol. Gen. Genet. 213:118-124. 105. 106. 107. 108. 109. 110. 111. 112. 113. 114. 115. 116. 117. 118. 43 Rasmussen, K. V., T. Atlung, G. Kerszman, G. E. Hansen and F. G. Hansen. 1983. Conditional change of DNA replication control in an RNA polymerase mutant of Escherichia coli. J. Bacteriol. 154:443-51. Ream, L. W., L. Margosslan, A. J. Clark, F. G. Hansen and M. K. von. 1980. Genetic and physical mapping of recF in Escherichia coli K-12. Mol. Gen. Genet. 180:115-21. Roth, A., B. Urmoneit and W. Messer. 1994. Functions of histone- like proteins in the initiation of DNA replication at oriC of Escherichia coli. Biochimie. 76:917-923. Sakakibara, Y. 1988. The dnaK gene of Escherichia coli functions in initiation of chromosome replication. J. Bacteriol. 170:972-9. Sakakibara, Y. 1993. Cooperation of the Eprs and dnaA gene products for initiation of chromosome replication in scherichia coli. J. Bacteriol. 175:5559-5565. Sakakibara, Y., H. Tsukano and T. Sako. 1981. Organization and t1r231nscription of the dnaA and dnaN genes of Escherichia coli. Gene. :47-55. Samitt, C. E., F. G. Hansen, J. F. Miller and M. Schaechter. 1989. In vivo studies of DnaA binding to the origin of replication of Escherichia coli. EMBO J. 8:989-993. Schaefer, C. and W. Messer. 1988. Termination of the Escherichia coli aan transcript. The DnaAgrotein/dnaA box complex blocks transcribing RNA polymerase. ene. 73:347-354. Schaper, S. and W. Messer. 1995. Interaction of the initiator protein DnaA of Escherichia coli with its DNA target. J. Biol. Chem. 270: 1 7622-1 7626. Schaus, N., K. O'Da , W. Peters and A. erght. 1981. Isolation and characterization o amber mutations in gene dnaA of Escherichia coli K-12. J. Bacteriol. 145:904-913. Schauzu, M. A., C. Kucherer, R. Kolling, W. Messer and H. Lother. 1987. Transcripts within the replication origin, oriC, of Escherichia coli. Nucleic Acids Res. 15:2479-2497. Sekimizu, K., D. Bramhill and A. Kornberg. 1987. ATP activates dnaA protein in initiating replication of plasmids bearing the origin of the E. coli chromosome. Cell. 50:259-265. Sekimizu, K., D. Bramhill and A. Kornberg. 1988. Sequential early sta es in the in vitro initiation of replication at the origin of the Escheric ia coli chromosome. J. Biol. Chem. 263:7124-30. Sekimizu, K., B. Y. Yung and A. Kornberg. 1988. The dnaA protein of Escherichia coli. Abundance, improved purification, and membrane binding. J. Biol. Chem. 263:7136-7140. 119. 120. 121. 122. 123. 124. 125. 126. 127. 128. 129. 130. 44 Skarstad, K., T. A. Baker and A. Kornberg. 1990. Strand separation required for initiation of replication at the chromosomal origin of 500” is facilitated by a distant RNA--DNA hybrid. EMBO J. 9:2341-2348. Skarstad, K. and E. Bo e. 1994. The initiator protein DnaA: evolution, properties and unction. Biochim. Biophys. Acta. 1217:111-130. Skarstad, K., E. Boys and H. B. Steen. 1986. Timin of initiation of chromosome replication in individual Escherichia coli cel s. EMBO J. 5:1711-1717. Skarstad, K., H. B. Steen and E. Boye. 1985. Escherichia coli DNA distributions measured by flow cytometry and compared with theoretical computer simulations. J. Bacteriol. 163:661-668. Skarstad, K., K. von Meyenburg, F. G. Hansen and E. Boye. 1988. Coordination of chromosome replication initiation in Escherichia coli: effects of different dnaA alleles. J. Bacteriol. 170:852-858. Skovgaard, O. and F. G. Hansen. 1987. Comparison of dnaA nucleotide sequences of Escherichia coli, Salmonella typhimurium, and Serratia marcescens. J. Bacteriol. 169:3976-3981. Slater, S., S. Wold, M. Lu, E. Bo e, K. Skarstad and N. Kleckner. 1995. E. coli Squ protein inds oriC in two different methyl- modulated reactions appropriate to its roles in DNA replication initiation and origin sequestration. Cell. 82:927-936. Sotomura, M. and M. Yoshikawa. 1975. Reinitiation of chromosome replication in the presence of chloramphenicol under an integratively suppressed state by R6K. J. Bacteriol. 122:623-8. Stuitje, A. R., N. de Wind, J. C. van der Spek, T. H. Pors and M. Meiier. 1986. Dissection of promoter sequences involved in transcriptional activation of the Escherichia coli replication origin. Nucleic Acids Res. 14:2333-2344. Tesfa-Selase, F. and W. T. Drabble. 1992. Re ulation of the gua ggeron of Escherichia coli by the DnaA protein. Mo. Gen. Genet. :256-264. Theisen, P. W., J. E. Grimwade, A. C. Leonard, J. A. Bogan and C. E. Helmstetter. 1993. Correlation of Gene Transcription with the Time of Initiation of Chromosome Replication in Escherichia coli. MoI. Micro. 10:575-584. Torrey, T. A., T. Atlung and T. Kogoma. 1984. dnaA sup ressor (dasF) mutants of Escherichia coli are stable DNA replication sdrA/mh) mutants. Mol. Gen. Genet. 196:350-355. 131. 132. 133. 134. 135. 136. 137. 138. 139. 140. 141. 142. 45 Tresguerres, E. F., H. G. Nandadasa and R. H. Pritchard. 1975. Suppression of initiation-negative strains of Escherichia coli by integration of the sex factor F. J. Bacteriol. 121 :554-61. van den Ber , E. A., R. H. Geerse, J. Memelink, R. A. Bovenberg, . A. Magnee and P. van de Putte. 1985. Analysis of regulatory sequences upstream of the E. coli uvrB gene; involvement of the DnaA protein. Nucleic Acids Res. 13:1829-1840. von Freiesleben, U., K. V. Rasmussen and M. Schaechter. 1994. Squ limits DnaA activity in replication from oriC in Escherichia coli. Mol. Micro. 14:763-772. von Meyenburg, K., E. Boye, K. Skarstad, L. Koppes and T. Kogoma. 1987. Mode of initiation of constitutive stable DNA replication in RNase H- defective mutants of Escherichia coli K-12. J. Bacteriol. 169:2650-2658. von Meyenburg, K. and F. G. Hansen. Regulation of chromosome replication. In Neidhardt, F. C., J. L. Ingraham, K. B. Low, B. Magasanik, M. Schaechter and H. E. Umbarger (eds.), Escherichia coli and Salmonella typhimurium. Cellular and Molecular Biology, 1987. American Society for Microbiology, Washington, D. C. von Meyenbur , K., F. G. Hansen, E. Rlise, H. E. Bergmans, M. Meijer and . Messer. 1979. Origin of replication, oriC, of the Escherichia coli K12 chromosome: genetic mapping and minichromosome replication. Cold Spring Harbo Symp. Quant. Biol. 121-128. Wang, Q. 1989. Ph. D. Thesis. Wang, Q. P. and J. M. Kaguni. 1987. Transcriptional repression of the dnaA gene of Escherichia coli by dnaA protein. Mol. Gen. Genet. 209:518-525. Wang, Q. P. and J. M. Kaguni. 1989. DnaA protein regulates transcriptions of the rpoH gene of Escherichia coli. J. Biol. Chem. 264:7338-7344. Watson, J. D. and F. H. C. Crick. 1953. Molecular structure of nucleic acids: a structure for deoxyribose nucleic acid. Nature. 1 71:737-738. Wickner, S. and J. Hurwitz. 1975. Interaction of Escherichia coli dnaB and dnaC(D) gene products in vitro. Proc. Natl. Acad. Sci., U. S. A. 72:921-5. Wickner, S. H. and D. K. Chattora]. 1987. Replication of mini-P1 plasmid DNA in vitro requires two initiation proteins, encoded by the repA gene of phage P1 and the dnaA gene of Escherichia coli. Proc. Natl. Acad. Sci., U. S. A. 84:3668-72. 143. 144. 145. 146. 147. 148. 46 Wu, F., I. Goldberg and M. Filutowicz. 1992. Roles of a 106-bp origin enhancer and Escherichia coli DnaA protein in replication of plasmid R6K. Nucleic Acids Res. 20:811-817. Yasuda, S. and Y. Hirota. 1977. Cloning and mapping of the replication origin of Escherichia coli. Proc. Natl. Acad. Sci., U. S. A. 74:5458-5462. Yung, B. Y., E. Crooke and A. Kornberg. 1990. Fate of the DnaA initiator protein in replication at the origin of the Escherichia coli chromosome in vitro. J. Biol. Chem. 265:1282-1285. Yung, B. Y. and A. Kornberg. 1988. Membrane attachment activates dnaA protein, the initiation protein of chromosome replication in Escherichia coli. Proc. Natl. Acad. Sci., U. S. A. 85:7202-5. Zyskind, J. W., J. M. Cleary, W. S. Brusilow, N. E. Harding and D. W. Smith. 1983. Chromosomal re lication origin from the marine bacterium Vibrio harve i functions in scherichia coli. oriC consensus sequence. Proc. Natl. cad. Sci., U. S. A. 80:1164-1168. 2 skind, J. W., N. E. Harding, Y. Takeda, J. M. Cleary and D. . Smith. The DNA replication ori in region of the enterobactiriaceae. In Ray, D. S. (eds.), The Initiation 0 DNA Replication, 1981. Academic Press, Los Angeles. Chapter II Novel Alleles of the Escherichia coli dnaA gene That Are Defective in Replication of pSC101 but Not of oriC 47 Abstract Five novel mutations of the Escherichia coli dnaA gene were isolated that were temperature sensitive in maintenance of pSC101, a plasmid that is dependent on this gene for replication. Nucleotide sequence analysis revealed that four of the five mutations arose from single base substitutions whereas the fifth contained three base substitutions, two of which were silent. Whereas all were temperature sensitive in vivo for pSCtOt maintenance, genetic and biochemical characterization indicated that only two were defective in replication from the chromosomal origin, oriC. As previously characterized mutations are defective in replication for both pSC101 and oriC, the dnaA mutations specifically defective in pSC101 maintenance represent a novel class. We speculate that one or more of these pSC101-specific mutations are defective in interaction with pSC101 RepA protein that is also required for initiation of plasmid DNA replication. 48 Introduction The dnaA gene of Escherichia coli is essential for initiation of replication of the bacterial chromosome (23, 32, 37). A number of plasmids and bacteriophage also require this gene for DNA replication as indicated by the inability of F (28, 36), R6K (55) and P1 (28, 54) to be maintained in a dnaA null mutant. Replication of pSC101 is also dependent on the dnaA gene (18, 20, 31). However, it is not capable of integrative suppression of a temperature sensitive dnaA allele, in contrast to F (43), R6K (46), and P1 (12, 13). These findings suggest that many of the same functions of DnaA protein are required for replication of pSC101 and for the bacterial chromosome. Biochemical characterization of DnaA protein indicates that it is multifunctional. It is a sequence-specific DNA binding protein that recognizes 9-mer sequences (DnaA boxes) present in oriC, the chromosomal origin, as well as in replication origins of plasmids and bacteriophage that require this protein for replication (21). At oriC, DnaA protein, bound to ATP and stimulated by integration host factor (IHF) or HU protein, is proposed to induce a structural alteration resulting in strand separation of the AT-rich region (7). As binding of DnaA protein to oriC is mutually exclusive and independent of lHF binding, stimulation of unwinding by IHF is apparently at a step subsequent to DNA binding by DnaA protein (34). By contrast, at the pSC101 origin, enhancement of DnaA protein binding by either lHF, or pSCIO1-encoded RepA protein is proposed to occur through formation of a looped structure (47). The differences in mechanisms of initiation at oriC relative to the pSC101 origin likely relates to differences in their structural organization (Figure 1). Whereas both strong and weak DnaA protein binding sites in the pSCtO1 origin are similarly oriented (47), the four DnaA boxes in oriC are arranged with 49 50 Figure 1. Structural organization of oriC and the pSC101 replication origin. DnaA protein binding sites (DnaA boxes) are represented by pentagons with the indicated orientations. DnaA boxes containing at most one mismatch from the consensus are crosshatched whereas those with more than 4 or 5 mismatches are not. AT-rich 13-mer repeats (L, M, and R) are represented by open circles. RepA binding sites (50, 52) are represented by shaded boxes (direct repeats RS1-RS3) or arrows (inverted repeats IR1 and IR2). Also indicated are lHF (19, 48) and Fis (25) binding sites. 51 :5 .mm an.» IV an.» 63:56.. _ woo an 52 opposing polarities (21). Each replicon has an lHF binding site between DnaA boxes and 13-mer motifs (6, 19, 48). lterons bound by RepA protein are important for enhancement of DnaA protein binding, leading to the speculation that RepA protein with lHF interact directly with DnaA protein in the initiation process (47). Although the dnaA gene was first identified nearly 3 decades ago (32, 37), the collection of conditionally defective mutations is limited to 16 alleles (8, 17, 26, 30, reviewed in reference 40) (see Figure 2A). In addition, the amino acid sequence has been deduced for 15 homologs of the dnaA gene from bacteria in the family Enterabacteriaceae (reviewed in reference 40). Their comparative analysis, revealing a remarkable degree of amino acid sequence conservation, suggests that conserved residues contribute to the structure or function of DnaA protein. If so, mutations that affect its functions are expected in conserved residues of this essential gene. We embarked on the isolation and characterization of novel mutations of the dnaA gene by a genetic assay that measures pSCtO1 maintenance. As oriC and pSC101 appear to have different mechanisms of initiation, we felt it possible that mutations of dnaA defective in replication of pSC101 but not oriC may be identified. Some of the mutations described here are specifically defective for pSC101 maintenance. The substituted residues may affect a proposed interaction with pSC101 RepA protein. Materials and Methods Bacterial strains and plasmids. The bacterial strains and plasmid DNAs used in this study are described in Table 1. P1 transduction was performed essentially as described by Miller (41) using P1 vir. pACYCdnaA was constructed by ligation of the 1.9 kb Cla l-Xho I fragment (see Figure 2A) containing the dnaA gene and its promoter from pdnaA/dnaN (10) to the 3.6 kb Cla I-Sall pACYC184 vector fragment (11). The dnaA+ gene of pACYCdnaA was replaced by mutant sequences using the unique Eco NI and Fisr lI restriction sites (encompassing the dnaA coding region but without the dnaA promoters) located internal to the Cla I-Xho I restriction sites (see Figure 2A). Bacteriological methods were performed essentially as described by Miller (41). Bacterial transformation was either by the calcium chloride method or electroporation using a Bio-Rad Gene Pulser according to the manufacturer‘s recommendation. Bacterial strains and their plasmid-containing derivatives were routinely grown in LB medium supplemented, as indicated, with 50 uglml ampicillin, 40 ug/ml kanamycin and 10 ug/ml tetracycline. M9 minimal medium supplemented with 0.5% casamino acids, 0.2% glucose, 1 )1ng thiamine and the indicated antibiotic(s) was used as noted. in vitro mutagenesis of the dnaA gene. 30 ug of pD8596 DNA (33) was incubated for 60 min at 65°C in 1 ml of 1 M hydroxylamine-HCI (Sigma) and 0.1 M sodium phosphate pH 6.5 (41). The sample was then dialyzed against 200 volumes of 10 mM Tris-HCI pH 8.0 and 1 mM EDTA for 2 h, ethanol precipitated, washed with 70% ethanol and resuspended in the above buffer. Hydroxylamine treatment reduced the transformation efficiency of the plasmid DNA to 1% of the untreated DNA. An Eco NI-Xhol restriction fragment containing the complete dnaA coding region from 53 54 TABLE 1. Bacterial strains and plasmids8 Strain Genotypeb Source or construction MC1061 araD139 A(ara-Ieu)7696 AlacX 74 Lab stock galU gaIK hstZ (rK-mK+) mch1 rpsL (F-) EH3894 asnB32 reIAf spoTi thi-f fuc-1 M. Yarmolinsky (28) IysA iIv-192 zia::pKN500 (pKN500=mini-R1) AdnaA mad-10 tnaA::Tn 10 (F-) M83895 EH3894 but lacking Tn 10 Selection for To8 of EH3894 on Bochner media (4) M83896 M83895: mad-2° Selection for growth of M83895 at 42°C KL16-99 recA1 HerL16 L. Snyder M83897 M83896: recA1 IysA+ fucf (M83896) X (KL16-99), select Lys+ and screen for RecA- by UV5 TH433 lea-19 pro-19 frp-25 his-47 This lab thyA59 arg-58 met-55 de0823 lac-11 gal-11 strA56 suI-1 hstK12 dnaA204 tnaA::Tn 10 (F-) M8204 M83896: dnaA204 tnaA::Tn10 (M83896) X P1(TH433), select Tcr, screen for Ts pSC101 maintenance M8205 M8204 but lacking Tn 10 Selection for T08 of M8204 on Bochner media (4) M8206 M8205: recA1 IysA+ fuc+ (M8205) X (KL16—99), select Lys+ and screen for RecA- by UV5 55 TABLE 1 (continued). Plasmid Characteristics Source pING1 Ap'; pBR322 origin, expression vector, D. Ray (35) Salmonella typhimurium araB promoter and araC pDStOS dnaA5 cloned into pING1 This lab (33) pDSZtS dnaA46 cloned into plNGt This lab (33) pDS319 dnaA204 cloned into pING1 This lab (33) pD8596 dnaA+ cloned into pING1 This lab (33) pD8596* p08596 containing an hydroxylamine-treated This work Eco NI-Xho I fragment of the dnaA gene pA31T A31T cloned into pING1 This work pV198M V198M cloned into pING1 This work pG287S G287S cloned into pING1 This work pT301| T310I cloned into pING1 This work pL447W L447W cloned into pING1 This work pACYC184 Cm', Tc'; p15A origin, cloning vector Lab stock (11) pACYCdiaA Cm'; dnaA+ cloned into pACY0184 This work pACYCohaA46 Cm'; dnaA46 cloned into pACYC184 This work pACYCA31T Cm'; A31T cloned into pACY0184 This work pACYCGZB78 Cm'; 62878 cloned into pACYC184 This work pACYCT301I Cm'; T301l cloned into pACYC184 This work pACYCL447W Cm'; L447W cloned into pACYC184 This work pSCfOt Tcr; DnaA-dependent Iow-copy-number S. Cohen (14) repficon pKN505 Apr; pBR322 origin, expresses R1 copA K. Nordstrom (24) activity aAbbreviations are Ap, ampicillin; Cm, chloramphenicol; Tc, tetracycline; UVS, ultraviolet sensitive; Ts, temperature sensitive; P1, P1 Iysate; X, genetic cross. 0 See the report of Bachmann (3) for genetic symbols. 0 mad-1 is a mutation that confers cold-resistant growth on minimal medium and temperature-sensitive growth on rich medium to a strain that contains a dnaA null mutation and is integratively suppressed by a mini-R1 plasmid (zia::pKN500) (28). It is unlinked by P1 transduction to both the integrated mini-R1 and dnaA null mutation and is thought to be a large chromosomal inversion (28). For consistency, we have 56 TABLE 1 (continued). assigned the mad-2 designation to derivatives of AdnaA, zia::pKN500, mad-1 strains that spontaneously acquired temperature resistant growth on both rich and minimal medium. We speculate that mad-2 is also a chromosomal inversion resulting in a more optimal placement of zia::pKN500 relative to terC (39). Replication of both the mad-1 and mad-2 mutations remains dependent on the integrated mini-R1 plasmid (zia::pKN500) as demonstrated by the cold- sensitive phenotype of the strains following their transformation to ampicillin resistance with pOU420, a plasmid encoding an R1 copA(Ts) allele (28, data not shown). 57 hydroxylamine-treated pD8596 was gel-purified (Qiaex DNA extraction kit, Qiagen), and ligated to gel-purified pING1 vector fragment from nonmutagenized p08596 to form pD8596* (Figure 2B). This construction places the dnaA gene without most of its promoter region distal to the araB promoter that is regulated by the plasmid-encoded AraC protein. Genetic assay to identify mutations of dnaA defective In pSC101 maintenance. A genetic screen was devised to identify plasmid-home mutations of dnaA unable to complement the conditional phenotype of the dnaA204(Ts) allele for pSC101 maintenance. M8206 containing pSC101 was transformed with the mutagenized dnaA-containing plasmid, pD8596*, by selection for ampicillin resistance at 30°C on LB medium containing kanamycin and tetracycline (Figure 3). Transforrnants were then grown in LB medium containing the above antibiotics at 30°C in 96-well microtiter plates and diluted to a cell density of approximately 104 colony forming units/ml. Aliquots (1-2 pl) were replica plated in duplicate onto supplemented M9 medium with or without tetracycline and incubated at 30°C or 41 .5°C to screen for the inability of transfonnants to maintain pSC101 at the elevated temperature. The phenotype of plasmid-bome mutations was confirmed by retransformation into M8206 (pSC101) followed by incubation at 30°C and 41 .5°C on supplemented M9 medium. The vector only (pING1), or it containing the dnaA5(Ts) (pDS105), dnaA46(Ts) (pDS215), dnaA204(Ts) (pDS319) and dnaA+ alleles (p08596) served as controls. Chemical cleavage of mismatched bases. The chemical mismatch cleavage procedure was performed on heteroduplex DNA essentially as described (15). As hydroxylamine modifies cytosine mismatches, and osmium tetroxide modifies thymine mismatches (16, 42), analysis of both strands provides mapping information for all possible substitutions. The sizes of 58 Figure 2. (A) Map of the dnaA gene. Relevant restriction enzyme sites, promoters (pentagons), the DnaA box, positions of mutations for existing dnaA alleles (listed below) (8, 17, 26, 30, reviewed in reference 40) as well as those described here (listed above) are indicated. Extragenic suppressors of dnaX2016(Ts) that reside in the dnaA gene are denoted by “Cs, 8x” (26). Positions are not indicated for intragenic suppressor mutations (dnaA46cos, an508005) of the dnaA46 and dnaA508 alleles. Numbers refer to the first and last amino acid residues of DnaA protein. (B) p08596" construction. p08596* was constructed (see Materials and Methods) by insertion into the vector pING1 of an Eco Nl-Xho I restriction fragment containing the hydroxylamine-treated dnaA coding region. The recombinant plasmid carries the mutagenized dnaA gene without most of the promoter region downstream from the araB promoter that is regulated by AraC protein, also encoded by the plasmid vector. 59 > 9m 2 3 *3 95> cox N8 2‘ >94 58: 83m 48: ..th Ems: X30. duo-Mal. 4.1% ...WIHIII I! .I.. . . .. .. . . .... ...!» E .....I. ...... . 4.2.... ...: .. . , . . as; ...... . . Iq‘rci .r . .. . . .. J. .I . 1 ,. ...». . I . 4 I . c . I u... . ...... F.w........n..l.h.uafi.:n.l. «4.. Hui-n. aux...“ J.t.t.....1__..... .. .r.. ...... i. «I... hit... r... . .. 2...»..- .t It . .61... ......2 r5 m... min... .. I?» II. .lLr ...... [In-la ..h .i.. ......npfri .1... ......Do. at H... (r. .... u . mom mom 3.x m «209me 3 moimom motmom mam N: m 3 69me 8 moimom mimom NOSE ~36?me EQHMMNWBSQ m8 Z- 0:40.— 4 Sam) .. ll .3 02> Comma 00. can? mam), #832: 56. 0o. 25? 508143353 m8 2. IIV v.29 use) ll § 60 cleavage products detected by autoradiography were determined relative to molecular weight markers. By chemical cleavage, the nature of putative base changes was deduced by the base preference of hydroxylamine (C-specific) or osmium tetroxide (T-specific), and the template or coding strand that was cleaved. By taking into account the position of the labelled nucleotide relative to the coding sequence of the dnaA gene, the approximate location of the mispaired nucleotides was determined. DNA sequence analysls. Based on the results from chemical cleavage, relevant regions for each of the five mutations were sequenced by the chain-terminating enzymatic method. Oligonucleotide primers JK—7, 5'—GTGGAGTCCCATATGTCACTTTCGCTTTGG—3'; JK—21, 5'-ATAACCCGTTGTI'CC-3'; JK—22, 5'—GACCACCTAACGGACCGCTC—3': JK—26, 5’—GAGCGCTTTG'ITCAGGACAT—3’, synthesized by an in-house facility with an Applied Biosystems model 394, were used with Sequenase (U.S. Biochemicals) essentially as described by the manufacturer with plasmid DNAs purified with the Qiagen mini- or midi-column method. Samples were electrophoresed in a 6% DNA sequencing gel and autoradiographed. The Oligonucleotide primers used for sequencing are homologous to the following nucleotide positions: JK—7, residues -12 to 18 of the coding strand and contains mismatches at positions -3, -2, -1 and 1; JK—21, 494 to 508 of the coding strand; JK—22, 1,459 to 1,478 of the template strand; JK—26, 610 to 629 of the coding strand. Quantitative immunoblot analysis of DnaA protein. Cultures were grown in supplemented M9 medium (41), unless otherwise indicated, containing the appropriate antibiotics at 30°C. At mid-log growth (OD595nm of 0507), cells from 1.0 ODsgsnm of culture were collected by centrifugation and resuspended in 50 mM Tris pH 7.6 and 20% sucrose (w/v). An equal volume of 61 50% glycerol, 0.2 M Tris-HCI pH 7.6, 1% SDS, 0.1% bromophenol blue, and 40 mM DTT was then added. The final concentration of each sample was equivalent to 10'2 ODSQSnm/ul. Samples were frozen in liquid nitrogen and stored at -70°C prior to electrophoresis in 10% SDS polyacrylamide gels, then subjected to immunoblot analysis (51). Detection was with a monoclonal antibody (M43 that recognizes residues 133-141) to DnaA protein incubated in 10 mM Tris pH 7.4, 154 mM NaCl, 0.05% Tween 20, and 2% nonfat milk. 125I-labelled Goat anti-Mouse lgG (ICN) was used as the second antibody. Quantitation of immunoreactive species was with a Packard Instant lmager, or with a Kodak Bio Image Visage 110, as indicated. Assay of in vitro oriC replication. Reactions (25 ul) were performed with a crude protein fraction deficient in DnaA protein activity prepared from WM433 (22). Purified DnaA+ protein was used as a control. Lysate supematants (Fraction I) were prepared using a gentle Iysis procedure as described (33) of cell pellets obtained from 1 L cultures of M83897 harboring the mutant alleles in pING1. nlduced expression of respective gene products was by addition of L-arabinose (Sigma) to 0.7%, followed by growth for 3 hr at 30°C. Resuns Isolation of five novel dnaA mutations. To obtain novel mutations of the dnaA gene, a genetic assay was developed (Figure 3). This method measured the inability of a plasmid-home mutation to complement the conditional defect of the dnaA204(Ts) allele for pSCIO1 maintenance. To measure pSCtO1 maintenance at an elevated temperature, the host strain (M8206) was also integratively suppressed by a mini-R1 plasmid. Whereas this strain was viable up to 41 .5°C due to integrative suppression, an isogenic strain lacking the mini-R1 was viable only up to 38°C (data not shown). When the integratively suppressed strain was transformed by pSC101, this plasmid was not maintained at the elevated temperature unless it was co-resident with another containing the dnaA+ gene (pD8596) (Table 2) whose expression is under control of the araB promoter but was not induced (see Figure 28). Sufficient levels of DnaA protein were expressed from this construct in the absence of arabinose induction to permit the establishment and maintenance of PSC101. Other controls were performed with this strain containing the latter Plasmid but bearing other alleles (dnaA5, dnaA46, and dnaA204) temperature sensitive for replication from oriC (reviewed in reference 40). These dnaA alleles were also temperature sensitive for pSC101 maintenance (Table 2), Confirming that pSC101 maintenance is dnaA-dependent (18, 20, 31). By this method in combination with hydroxylamine mutagenesis of the dnaA gene (Figure 2 and 3), five mutations defective for pSC101 maintenance at high temperature (41.5 °C) were identified from a collection of 1,790 transfon'nants (Table 2). At the elevated temperature, G287S (with a G-to-S mUtation at position 287), T301l and L447W displayed small or minute colonies I’elative to 30°C, or when compared to the dnaA+ transfonnant (p08596). A31T 62 63 Figure 3. Genetic assay to Identify E. coli dnaA mutants defective in pSC101 maintenance at 41 .500. See Materials and Methods for a description. Abbreviations are Ap, ampicillin; Km, kanamycin; Tc, tetracycline. M8206 (pSC1 01 ) (and controls) V Select Ap, Km and To resistant transformants at 30°C on LB media I Pick colonies and grow in LB media + Ap, Km and To at 30°C in microtiter dishes I Dilute to 10-20 cfu/ul and spot onto supplemented M9 media +/- To | I I 30°C 41 .5°C I (+) To H To dnaA mutant 65 and V198M each showed a modest defect, having a reduction in plating efficiency to 0.11 and 0.09, respectively, at the elevated temperature (Table 2). Nucleotide sequence determination of mutant dnaA alleles. The respective mutation(s) of each allele was mapped by a chemical modification method (data not shown) (15), and the appropriate regions for each sequenced (Table 3). A single nucleotide substitution for four of the five alleles was identified. L447W was found to contain 3 nucleotide substitutions, two of which are silent. In all, each mutant allele was found to contain a single deduced amino acid substitution. AII mutations were consistent with hydroxylamine mutagenesis (C—>T transitions) (16, 42) except for the transversion (T->G) of L447W that we presume arose spontaneously. Compared to the dnaA alleles previously described (reviewed in reference 40), these are novel (Figure 2A). The amino acids, V-198 (valine at position 198), T-301, and L-447 are highly conserved among 15 dnaA homologs (Table 3). A-31 and G-287 are less well conserved among all homologs but are fairly well conserved among Gram negative bacteria. As hydroxylamine treatment should produce nonsense mutations (9), the failure to obtain these or to reisolate previously identified mutations may be due to the small number of transformants screened (1,790). Indeed, dnaA5, dnaA46 and dnaA204, were temperature sensitive for pSC101 maintenance as measured by this genetic screen (see Table 2). Furthermore, the residual activity of the dnaA204 allele at 42°C (29) may have restricted the types of mutations detected as those able to function in concert with the dnaA204 gene product would not have been obtained. Replication activity of the novel dnaA alleles In a AdnaA strain. To assess the phenotypes of the novel mutations without the possible contribution of a chromosomal dnaA allele, plasmids bearing the mutant alleles 66 TABLE 2. Inability of plasmid-bome dnaA mutations to confer temperature resistant growth to E. coli M8206 (pSC101) M8206 (950101) dnaA allele Efficiency of plating” transfonnanta pING1 None 8.1 X 10'6 p08596 dnaA+ 1.4 pDS105 dnaA5 5.1 X 10’5 pDSZ15 dnaA46 2.1 X 10'5 pDS319 dnaA204 5.6 X 10'5 pA31T A31T 0.11 pV198M V198M 9.1 X 10'2 pG287S G287S 1.3c pT301l T301 I 0.826 pL447W L447W 1 .3d 8 M8206 (pSC101) transfonnants contain the indicated dnaA alleles in plNGt and expressed from the araB promoter. 0 Transforrnants were grown at 30°C in LB medium containing kanamycin, ampicillin and tetracycline before plating. Plating efficiency is expressed as the number of colony forming units observed at 41 .5°C relative to 30°C on supplemented M9 medium containing the above antibiotics. Efficiency of plating was determined in the absence of arabinose induction. On medium without tetracycline, the colony morphology of M8206 (pSCtOt) with the different plasmid-home mutations was similar to the dnaA+ transfonnant. By contrast to these results with synthetic medium, pSC101 maintenance was not temperature- sensitive on LB medium with these plasmid-encoded dnaA mutations, or with temperature sensitive alleles (dnaA5, dnaA46 and dnaA204). 0 Small cfu observed at 41.500. 0' Tiny cfu observed at 41.500. 67 TABLE 3. Nucleotide and deduced amino acid alterations of five dnaA alleles‘ll Deduced amino acid dnaA Nucleotrde NucleotIde Ammo acrd substitution (sequence allele positionb substitutionc position conservation)d A31T 91 G—>A 31 Ala—>Thr (6/15) V198M 592 G—>A 198 Val—>Met (13/15) G287S 859 G->A 287 GIy—rSer (6/15) T301l 902 C—>T 301 Thr—+Ile (12/15) L447W 1335 G—>A 445 GIu—rGlu 1338 G—>A 446 Gln->Gln 1340 T—>G 447 Leu->Trp (12/15) 8 Chemical modification of the template and coding strand to map the approximate positions of mutations (data not shown) was performed as described in Materials and Methods. As controls, the dnaA+ and dnaA46 alleles were analyzed. The (cytosine) substitutions of the dnaA46 allele at residues 551 and 754 (30) were confirmed by this method. No mispairs in either strand of the dnaA+ gene were detected. Appropriate regions of each mutant allele based on the chemical modification mapping results were sequenced by the chain-terminating enzymatic method. b Numbers refer to those in dnaA coding region with position 1 corresponding to the first nucleotide of the coding sequence (27, 44). c Nucleotide substitutions are expressed as those in the coding strand. d Sequence conservation is indicated by the frequency of residues conserved at respective positions among the 15 dnaA homologs (reviewed in reference 40). 68 were transformed into an integratively-suppressed strain lacking the dnaA gene (M83897). Except for the dnaA null mutation, this strain is isogenic to that used to isolate the novel mutations. Plasmids bearing each of the novel alleles were taken up by this strain as efficiently as controls of the vector (pING1), or it bearing the wild type (pDS596) or dnaA46(Ts) allele (pD8215) (data not shown). The dnaA null strain carrying each of the novel mutations was then transformed by pSC101. At 30°C, their transformation efficiencies were comparable to controls of the dnaA+ or dnaA46 alleles in pING1 (data not shown). At 41 .5°C, the transformation efficiency with plasmids carrying the novel mutations or the dnaA46 allele was greatly reduced relative to dnaA+ (data not shown). The temperature sensitive phenotype for pSC101 maintenance was confirmed by plating representative transfonnants at 30°C and 41 .5°C on supplemented M9 medium (Table 4). At the elevated temperature, an increased plating efficiency was observed with these plasmid-home mutations co-resident with the chromosomal dnaA204 allele relative to the dnaA null allele in the isogenic strain (Table 2 compared to Table 4). This may be due to contribution of the dnaA204 allele that maintains residual activity at 42°C (29). An elevated level of DnaA protein is required for oriC replication compared with that required for pSC101 replication. We discovered that the dnaA+-containing plasmid, pD8596, was unable to maintain an oriC plasmid (data not shown), whereas pSC101 was maintained under comparable conditions (Table 4, line 1). This was in the absence of arabinose induction in the dnaA null strain (M83897). By contrast, the dnaA+ gene expressed from its own promoter contained in pACYC184 was sufficient to maintain an oriC plasmid (data not shown), or to support chromosomal replication from oriC upon inhibition of integrative 69 TABLE 4. Plasmid-home dnaA mutations are temperature sensitive for pSC101 maintenance trahrcggggznta dnaA allele Efficiency 0f platingb [308596 an+ 0.91 PD8215 dnaA46 9.0 X 10-5 pA31T A31T 2.3 x 10-3 PV198M V198M 1.2 x 10-3 PG287S G287S 1.9 x 1o-2 pT301l T301I 8.1 x 10-4 pL447W L447W 1.1 x 104 8 M83897 transformants contain the indicated dnaA alleles in pING1 and expressed from the araB promoter. b Transforrnants were grown at 30°C in LB medium containing kanamycin, ampicillin and tetracycline before plating. Efficiency of plating is expressed as the ratio of the number of colony forming units observed at 41.500 relative to 30°C on supplemented M9 medium containing the above antibiotics. Plating efficiency was determined in the absence of arabinose induction. 7O suppression (see Table 5, line 1 and below). These results suggest that expression of DnaA protein, when under araB promoter control and not induced, is limiting for oriC replication. Quantitative immunoblot analysis supported this interpretation (Figure 4). The level of DnaA+ protein expressed from pACYCdnaA contained in the null strain was about 40-fold higher than the uninduced level expressed from p08596 (Figure 4, inset). In the latter plasmid, the level of expression was not dramatically affected by growth medium. The level of DnaA+ protein expressed from p08596, barely detectable by autoradiography, concurs with the observation of Abeles ef al. (1). Together, these results indicate that the level of DnaA protein required for pSC101 replication in vivo is less than that required for replication of an oriC plasmid or the bacterial chromosome. In the dnaA null strain, DnaA+ and DnaA46 protein encoded in pACYC184 were expressed about 2.1-fold and 2.9-fold higher, respectively, relative to the chromosomally-encoded level of M01061 (Figure 4). The calculated cellular abundance of chromosomally encoded DnaA protein (1.9 x 10‘3 molecules per 10'9 O.D.595nm, see Figure 4 legend) is slightly greater than published estimates (1-1.5 x 103 molecules per 10'9 O.D.595nm) (45). In this experiment, DnaA protein was not detected in the null strain lacking a plasmid or harboring pACYC184, confirming its genotype. Replicon-specific defect of the novel dnaA alleles. The method to obtain novel mutations screened for the inability of a plasmid-encoded allele to maintain pSC101 in a host strain harboring the dnaA204(Ts) allele at the nonpermissive temperature (Figure 3). Consequently, it is possible that the mutations obtained were not defective in replication from oriC. To address this issue, chromosomal replication from oriC was measured by inhibition of replication from the integrated R1 origin in the dnaA null strain (M83897). 71 Figure 4. Quantitative lmmunoblot analysis of DnaA protein. The level of DnaA protein encoded by the various recombinant plasmids in the dnaA null strain (M83897) was determined as described in Materials and Methods. “None” refers to the M83897 (pACYC184) transfonnant. The levels were quantified with a Packard Instant Imager. For each sample, the counts detected during a 100 min exposure minus background, indicated by the right ordinate, were normalized relative to the chromosomally encoded level of the wild type strain MC1061 (indicated by the left ordinate). In the inset, scanning densitometry of an autoradiogram (not preflashed) was required due to the low levels of expression that could not be detected with the Instant Imager. Optical densities of DnaA protein encoded by p08596 grown in either LB medium (0.24 units), or M9 minimal medium (0.15 units) were normalized relative to the level encoded by pACYCdnaA (9.43 units). Relative to known amounts of purified DnaA protein analyzed in parallel, the cellular abundance of DnaA protein encoded by MC1061 was determined. 72 Relative abundance .0 N pACYCdnaA None posssel:MB IIIII omemoo I _s o DnaA+ DnaA46 A31T V198M , G287S T301 I L447W é ' I ' I ' I ' I ' I ' I ' I ' I ' I N ..L _L .5 —L —L O O O O O O O O O O O O O O Counts 73 Inhibition was achieved by introduction of a plasmid containing the R1 copA gene (pKN505) that encodes an antisense RNA to the transcript for R1 RepA protein (24). Formation of the RNA duplex interferes with expression of RepA protein required for R1 replication (49). The level of DnaA protein, when encoded by pACYC184 and expressed from its own promoter, was sufficient to support oriC replication (data not shown, Table 5). To examine the dnaA mutations for oriC replication activity, derivatives of pACYC184 containing these alleles were constructed. They were introduced into the dnaA null strain followed by the R1 copA-containing plasmid (pKN505) to inhibit integrative suppression (Table 5). In contrast to the others, the A31T, G287S and L447W alleles supported replication from oriC as effectively as the wild type control (pACYCdnaA). This observation suggests that the defect of the A31T, G287S and L447W alleles is specific for pSC101 replication. Transformants harboring the copA-containing plasmid (pKN505), and either the plasmid-home dnaA46 or T301I allele were obtained at 30°C but were extremely unstable, preventing an assessment of their replicon specificity. This observation is not understood but is unlikely to be due to insufficient levels of DnaA protein (Figure 4). DnaA46 and T301I proteins were expressed at levels comparable to the other mutant proteins. As a control, pSC101 maintenance was examined in the integratively-suppressed null strain (Table 5). G287S, T301l and L447W were temperature sensitive for maintenance of pSCfOt. By contrast, A31T and V198M were not defective for pSC101 maintenance at the elevated temperature. Suppression of the temperature sensitive phenotype for pSC101 maintenance by elevated expression. All of the mutant dnaA alleles, except G287S which was unchanged, showed an increase in plating 74 TABLE 5. A subclass of dnaA mutations is specifically defective in pSC101 maintenance M83897 transformanta dnaA allele Efficiency of platingc pSC101 pKN505 pACYCdnaA dnaA+ 1.17 1.16 pACYCdnaA46 dnaA46 5.9 X 10'3 NA6 pACYCA31 T A31T 1.01 0.89 pACYCV198M V198M 1.30 6.3 X 10'2 pACYCG287S G287S 1.6 X 10'2 0.91 pACYCT301I T301l 7.8 X 10'2 NA pACYCL447W L447W 4.2 X 10‘2 1.03 8 M83897 transformants contain the indicated dnaA alleles in pACYC184 and expressed from the dnaA promoter. 0 Transformants of pSC101 were grown at 30°C in LB medium containing kanamycin, chloramphenicol and tetracycline before plating. Transformants of pKN505 were grown on supplemented M9 agar medium, picked and diluted in M9 salts before plating. Efficiency of plating is expressed as the number of colony forming units observed at 41.500 relative to 30°C on supplemented M9 medium containing kanamycin, chloramphenicol and either tetracycline (for pSC101) or ampicillin (for pKN505). C Efficiency of plating is not available (“NA”) due to the instability of the transfonnant containing both the plasmid-encoded dnaA allele and pKN505. 75 efficiency of two to three orders of magnitude at the elevated temperature when contained in pACYC184 relative to pING1 (compare Table 4 and 5). The A31T and V198M alleles were not temperature sensitive for pSC101 maintenance when contained in pACYC184. Although we have not measured the cellular abundance of these or the other novel mutants when contained in pING1, we believe that their elevated expression when in pACYC184 suppresses, sometimes partially, the temperature-sensitive phenotype for pSC101 maintenance. This interpretation is based on two observations. First, immunoblot analysis showed that DnaA+ protein was approximately 40-fold elevated in the pACYC184 construct relative to pING1 (Figure 4, inset). Second, the viability of many dnaA(Ts) mutants increases when the alleles were present on both the chromosome and a pBR322 derivative (30), presumably by a gene dosage effect. Also in this analysis (Figure 4), the levels of the mutant proteins encoded in pACYC184 were found to be between 1.9- and 2.7-fold higher than the chromosomally-encoded dnaA+ control (M01061), and comparable or slightly elevated compared to the control of DnaA+ protein encoded by the comparable plasmid (pACYCdnaA). The latter difference may possibly relate to differences in protein stability. Alternatively, the difference in expression levels of the pACYC184 constructs may relate to their relative abilities to autoregulate their own expression. The dnaA promoter region containing a DnaA box is autoregulated by the binding of DnaA protein (2, 28, 38, 53). We do not know whether A31T or V198M, when expressed at the chromosomal level, will be temperature sensitive for pSC101 maintenance nor whether A31T will be temperature sensitive for replication from oriC. This issue may be addressed by substituting the chromosomal copy with each allele, or by modulating their expression from the araB promoter to the chromosomal level. 76 A31T, V198M and T301I are thermolabile In in vitro replication of an oriC plasmid. To corroborate the genetic studies, in vitro replication activity of the mutant DnaA proteins was measured at both 30°C and 42°C (Figure 5). The V198M and T301l mutant proteins were thermolabile in vitro for oriC plasmid replication. Whereas G2878 protein appeared comparable in replication activity at both temperatures to the wild type protein, A31T protein was moderately thermolabile. The replication activity of L447W protein was not thermolabile. However, the- extent of nucleotide incorporation with A31T and L447W protein was lower than that observed for both the wild type and G2878 proteins. The crude protein fractions used in this assay contained comparable levels of mutant or wild type protein as judged by immunoblot analysis (data not shown). Whether a contaminating factor in the samples may have affected the results obtained can be addressed when these proteins are purified. The marginal effect of temperature on the replication activity of A31T, G2878 and L447W proteins is in agreement with the in vivo phenotypes of these mutants for chromosomal replication (Table 5), suggesting that these alleles are specifically defective for pSC101 maintenance. 77 Figure 5. In vitro oriC replication activity of novel DnaA mutant proteins. Replication assays were performed as described in Materials and Methods at 30°C (0) or 42°C (0). Crude lysate supematants containing the indicated mutant DnaA proteins used in these in vitro replication assays were comparable in total protein concentration determined by the dye-binding method (5), and DnaA protein abundance as judged by immunoblot analysis (data not shown). DnaA protein is expressed either in ng (purified DnaA+ protein) or pl of Fraction I (mutants). By immunoblot analysis,1 ul of the respective Fraction I corresponds to ~150 ng of DnaA protein. DNA Synthesis (pmol) 78 600 400 200 600 400 200 600 r- 400 - 200 — 0 _ _ 4 0.0 0.3 0.6 0.9 1.2 600 400 200 600 400 200 600 400 200 0 A31T 0.0 0.3 0.6 0.9 1.2 L447W 0.0 0.3 0.6 0.9 DnaA Protein Discussion dnaA mutations specifically defective in pSC101 maintenance. The method to identify mutations of the E. coli dnaA gene relied on use of an integratively suppressed strain (M8206) containing the dnaA204(Ts) allele. Due to the integrated mini-R1, the dnaA gene and oriC are not essential for bacterial growth (28). Incubation at a temperature nonpermissive for the dnaA204 allele provided a means to assess the phenotype of the plasmid-encoded dnaA gene for maintenance of pSC101 (Figure 3). Whereas bacterial growth in the absence of tetracycline was observed at 30°C and 41 .5°C (unpublished results), pSC101 was not maintained at the elevated temperature in the absence of a functional dnaA gene product (Table 2). With this approach, five mutations unable to complement the conditional phenotype of the dnaA204 allele for pSC101 maintenance were obtained. Additionally, these mutations were temperature sensitive for pSC101 maintenance in a strain lacking the dnaA gene (Table 4). Compared to the 16 other missense mutations of the dnaA gene (8, 17, 26, 30, reviewed in reference 40), these are novel (see Figure 2A). Furthermore, V198M, T301I and L447W have substitutions in amino acid residues that are highly conserved in dnaA homologs of the enterobacteriaceae family (Table 3). Of the five mutations, A31T, G2878, and L447W were not temperature sensitive for oriC replication under conditions in which replication from the integrated R1 origin was inhibited (Table 5). Biochemical studies corroborated these genetic results (Figure 5). Assays using Iysate supematants enriched with the respective mutant proteins indicated that G2878 and L447W, and to a lesser degree A31T, were also temperature resistant in in vitro replication of an oriC plasmid. Collectively (Table 6), these results suggest that the G2878 and 79 80 TABLE 6. Summary of mutant dnaA phenotypes8 pSC101 maintenance oriC replication dnaA allele Low level of Hi h level in vivo (High in vitro DnaAb o DnaAc level of DnaAC) A31T Ts + + +/-Ts V198M Ts + Ts Ts G2878 Ts Ts + + T301l Ts Ts NA Ts L447W Ts Ts + + 8 This Table is based on results presented in Figures 4 and 5 and Tables 4 and 5. Abbreviations are Ts, temperature sensitive; «I, temperature insensitive; +/—, moderately temperature sensitive; NA, data not available. b “Low level of DnaA” refers to the level of expression observed when the mutant alleles were contained in pING1 and under control of the araB promoter in the noninduoed state (see Figure 4, inset). 6‘ “High level of DnaA” refers to the level of expression observed when the mutant alleles were contained in pACYC184 and expressed from the dnaA promoter (see Figure 4). 81 L447W mutations, and possibly A31T are specifically defective in replication of pSCtO1. We do not understand the instability of the dnaA null strain bearing both the copA plasmid and either plasmid-bome dnaA46 or T301I allele (Table 5). That these alleles were plasmid-bome and not chromosomally encoded may be of relevance. Suppressors of dnaX2016(Ts) that reside in the dnaA gene are dominant to plasmid-bome dnaA+ when chromosomally located, but recessive in the converse construction (26). A similar difference in activity depending upon mutant allele location (chromosome versus plasmid) has been reported for those dnaA(Ts) allele containing an alanine-to-valine substitution at residue 184 (30). These alleles, expressed from the dnaA promoter, exhibit a cold- sensitive phenotype when plasmid-encoded and in a strain containing a chromosomal dnaA+ allele. However, cold sensitivity was not observed when the thermolabile allele was present on the chromosome with the wild type allele residing in a plasmid. pSC101 maintenance requires a lower level of DnaA protein than the level needed for replication from oriC. With the dnaA+ gene in p08596, our inability to establish an oriC plasmid in a dnaA null mutant (unpublished results) contrasts with the establishment and maintenance of pSC101 under comparable conditions (Table 4). With the dnaA+ gene in pACYC184 in the dnaA null strain, we were able to establish an oriC plasmid (unpublished results), as well as to sustain chromosomal replication from oriC on inhibition of replication from the integrated R1 origin (Table 5). This result, suggesting that the level of DnaA protein required for pSC101 replication is less than that required for replication of an oriC plasmid or the bacterial chromosome, was confirmed by quantitative immunoblotting (Figure 4). 82 DnaA protein may interact with RepA protein of pSC101. Among the plasmids of E. coli, many require DnaA protein for replication. Of these, F and P1 are capable of integrative suppression of strains containing the dnaA46(Ts) allele (12, 13, 43), but not a dnaA null allele (28, 36). By comparison, the strict dependence of pSC101 on DnaA protein for maintenance (18, 20, 31) was interpreted to suggest that it requires many of the same functions of dnaA that are required for initiation from oriC. The mutations G2878, L447W, and possibly A31T, obtained in this study that appear to have a specific defect in pSC101 replication are particularly interesting for they suggest a unique role of DnaA protein in pSC101 replication. One or more may be defective in the proposed interaction of DnaA protein with pSC101-encoded RepA protein that facilitates the binding of DnaA protein to the pSC101 replication origin (47). Further examination of these mutant proteins in binding to the pSC101 replication origin, and the development of an in vitro system to study pSCtO1 replication should provide a greater understanding of the interplay between these proteins, and if the amino acids that have been substituted normally interact with pSC101 RepA protein. 10. 11. 12. 13. References Abeles, A., T. Brendler and 8. Austin. 1993. Evidence of two levels of control of P1 oriFI and host oriC replication origins by DNA adenine methylation. J. Bacteriol. 175:7801-7807. Atlung, T., E. S. Clausen and F. G. Hansen. 1985. Autoregulation of the dnaA gene of Escherichia coli K12. Mol. Gen. Genet. 200:442-450. Bachmann, B. J. 1990. Linkage map of Escherichia coli K-12, edition 8. Microbiol. Rev. 54:130-197. Bochner, B. R., H. C. Huang, G. L. Schieven and B. N. Ames. 1980. Positive selection for loss of tetracycline resistance. J. Bacteriol. 143:926-933. Bradford, M. 1976. A rapid and sensitive method for the quantitation of microgBram quantities of protein utilizing the principle of protein-dye binding. Anal. iochem. 72:248-254. Bramhill, D. and A. Kornberg. 1988. Duplex opening by dnaA protein at novel sequences in initiation of replication at the origin of the E. coli chromosome. Cell 52:743-755. Bramhill, D. and A. Kornberg. 1988. A model for initiation at origins of DNA replication. Cell 54:915-918. Braun, R. E., K. O'Day and A. Wright. 1987. Clonin and characterization of dnaA(Cs), a mutation which leads to overinitiation of NA replication in Escherichia coli K-12. J. Bacteriol. 169:3898-3903. Brenner, 8., A. O. Stretton and 8. Kaplan. 1965. Genetic code: the 'nonsense' triplets for chain termination and their suppression. Nature 206:994-998. Burgers, P. M., A. Kornber and Y. Sakakibara. 1981. The dnaN ene codes for the beta subunit of D A polymerase III holoenzyme of Escheric ia coli. Proc. Natl. Acad. Sci. USA. 78:5391-5395. Chang, A. C. and S. N. Cohen. 1978. Construction and characterization of amplifrable multicopy DNA cloning vehicles derived from the P15A cryptic miniplasmid. J. Bacteriol. 134:1141-1156. Chesney, R. H. and J. R. Scott. 1978. Suppression of a thermosensitive dnaA mutation of Escherichia coli by bacteriophage P1 and P7. Plasmid 1:145- 163. Chesney, R. H., J. R. Scott and D. Vapnek. 1979. Integration of the plasmid prophages P1 and P7 into the chromosome of Escherichia coli. J. Mol. Biol.130:161-173. 83 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 84 Cohen, S. N. and A. C. Chang. 1977. Revised interpretation of the origin of the pSCtO1 plasmid. J. Bacteriol. 132:734-737. Cotton, R. G. H., N. R. Rodrigues and R. D. Campbell. 1988. Reactivity of cytosine and thymine in single-base-pair mismatches with hydroxylamine and osmium tetroxide and its application to the study of mutations. Proc. Natl. Acad. Sci., U. S. A. 85:4397-4401. Davis, R. W., D. Botsteln and J. R. Roth. 1980. A manual for genetic engineering: advanced bacterial genetics. Eberle, H., Van, de, Merwe, W, K. Madden, G. Kampo, L. Wright and K. Donlon. 1989. The nature of an intragenic suppressor of the Escherichia coli dnaA508 temperature-sensitive mutation. Gene. 84:237-245. Felton, J. and A. Wright. 1979. Plasmid pSC101 replication in integratively suppressed cells requires dnaA function. Mol. Gen. Genet. 175:231-233. Filutowicz, M. and J. Roll. 1990. The requirement of IHF protein for extrachromosomal replication of the Escherichia coli oriC in a mutant deficient in DNA polymerase I activity. New Biol. 2:818-827. Frey, J., M. Chandler and L. Caro. 1979. The effects of an Escherichia coli dnaAts mutation on the replication of the plasmids ColE1 p80101, R100.1 and RTF-TC. Mol. Gen. Genet. 174:117-126. Fuller, R. 8., B. E. Funnell and A. Kornberg. 1984. The dnaA protein complex with the E. coli chromosomal replication origin (oriC) and other DNA sites. Cell 38:889-900. Fuller, R. 8., J. M. Kaguni and A. Kornberg. 1981. Enz matic replication of the origin of the Escherichia coli chromosome. Proc. Natl. cad. Sci. USA. 78:7370-7374. Fuller, R. S. and A. Kornberg. 1983. Purified dnaA protein in initiation of replication at the Escherichia coli chromosomal origin of replication. Proc. Natl. Acad. Sci. USA. 80:5817-5821. Gerhart, E., H. Kagner, J. von Heljne and K. Nordstrém. 1987. Control of replication of plasmid R1: translation of the 7k reading frame in the Re A mRNA leader counteracts the interaction between CopA RNA and CopT NA. EMBO J. 6:515-522. Gille, H., J. B. Egan, A. Roth and W. Messer. 1991. The FIS protein binds and bends the origin of chromosomal DNA replication, oriC, of Escherichia coli. Nucleic Acids Res. 19:4167-4172. Ginés-Candelaria, E., A. Bllnkova and J. R. Walker. 1995. Mutations in Escherichia coli dnaA which suppress a dnaX(Ts) polymerization mutation and are dominant when located in the chromosomal allele and recessive on plasmids. J. Bacteriol. 177:705-715. 27. 28. 29. 30. 31. 32. 33. 35. 36. 37. 38. 39. 85 Hansen, E. B., F. G. Hansen and K. van Meyenburg. 1982. The nucleotide sequence of the dnaA gene and the first part of the dnaN gene of Escherichia coli K-12. Nucleic Acids Res. 10:7373-7385. Hansen, E. B. and M. B. Yarmollns . 1986. Host participation in plasmid maintenance: dependence upon dnaA o replicons derived from P1 and F. Proc. Natl. Acad. Sci. USA. 83:4423-4427. Hansen, F. G. and T. Atlung. 1995. Reversibility of DnaA protein activity in the 'irreversible' dnaA204 mutant of Escherichia coli. Mol. Microbiol. 15:141- 148. Hansen, F. G., S. Koefoed and T. Atlung. 1992. Cloning and nucleotide sequence determination of twelve mutant dnaA genes of Escherichia coli. Mol. Gen. Genet. 234:14-21. Hasunuma, K. and M. Sekiguchi. 1977. Replication of plasmid pSCtO1 in Escherichia coli K12: requirement for dnaA function. Mol. Gen. Genet. 154:225- 230. Hirota, Y., A. Ryter and F. Jacob. 1968. Therrnosensitive mutants of E. coli affected in the processes of DNA synthesis and cellular division. Cold Spring Harb. Symp. Quant. Biol. 677-693. Hwang, D. S. and J. M. Ka uni. 1988. Purification and characterization of the dnaA46 gene product. J. iol. Chem. 263:10625-10632. . Hwang, D. S. and A. Kornberg. 1992. Opening of the replication origin of Escherichia coli by DnaA protein with protein HU or IHF. J. Biol. Chem. 267:23083-23086. Johnston, 8., J. H. Lee and D. 8. Ray 1985. High-level expression of $13 gene ll protein from an inducible po ycistronic messenger RNA. Gene :137-145. Kline, B. C., T. Ko oma, J. E. Tam and M. S. Shields. 1986. Requirement of the scherichia coli dnaA gene product for plasmid F maintenance. J. Bacteriol. 168:440-443. Kohiyama, M., D. Cousin, A. Ryter and F. Jacob. 1966. [Thennosensitive mutants of Escherichia coli K 12. I. Isolation and rapid characterization]. Ann. Inst. Pasteur (Paris). 110:465-486. Kucherer, C., H. Lother, R. Kolling, M. A. Schauzu and W. Messer. 1986. Regulation of transcription of the chromosomal dnaA gene of Escherichia coli. Mol. Gen. Genet. 205:115-121. Louarn, J., J. Patte and J. M. Louarn. 1982. Suppression of Escherichia coli dnaA46 mutations by integration of Iasmid R100.1 derivatives: constraints imposed by the replication terminus. J. acteriol. 151:657-667. 40. 41. 42. 43. 45. 46. 47. 48. 49. 50. 51. 52. 53. 86 Messer, W. and C. Weigel. Initiation of chromosome replication. In Neidhardt, F. C., J. L. Ingraham, K. B. Low, B. Magasanik, M. Schaechter and H. E. Umbarger (eds.), Escherichia coli and Salmonella typhimurium. Cellular and Molecular Biology, 1996. American Society for Microbiology. Washington, D. C. Miller, J. H. 1992. A short course in bacterial genetics. Myers, R. M., L. S. Lerman and T. Maniatis. 1985. A general method for saturation mutagenesis of cloned DNA fragments. Science 229:242-247. Nishimura, Y., L. Caro, C. M. Berg and Y. Hirota. 1971. Chromosome replication in Escherichia coli. IV. Control of chromosome replication and cell division by an integrated episome. J. Mol. Biol.55:441-456. .Ohmori, H., M. Kimura, T. Nagata and Y. Sakakibara. 1984. Structural analysis of the dnaA and dnaN genes of Escherichia coli. Gene 28:159-170. Sekimizu, K., B. Y. Yung and A. Kornberg. 1988. The dnaA protein of Escherichia coli. Abundance, improved purification, and membrane binding. J. Biol. Chem. 263:7136-7140. Sotomura, M. and M. Yoshikawa. 1975. Reinitiation of chromosome replication in the presence of chloramphenicol under an integratively suppressed state by R6K. J. Bacteriol. 122:623-628. Stenzel, T. T., T. MacAlIister and D. Bastia. 1991. Cooperativity at a distance promoted by the combined action of two replication initiator proteins and a DNA bending protein at the replication origin of pSC101. Genes Dev. 5:1453-1463. Stenzel, T. T., P. Patel and D. Bastia. 1987. The integration host factor of Escherichia coli binds to bent DNA at the origin of replicatIon of the plasmid pSC101. Cell 49:709-717. Stougaard, P., S. Molin and K. Nordstrém. 1981. RNAs involved in co y- number control and incompatibility of plasmid R1. Proc. Natl. Acad. Sci. US. . 78:6008-6012. Sugiura, 8., M. Tanaka, Y. Masamune and K. Yamaguchl. 1990. DNA binding properties of purified replication initiator protein (Rep) encoded by plasmid pSC101. J. Biochem. (Tokyo). 107:369-376. Towbin, H., T. Staehelin and J. Gordon. 1979. Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: procedure and some applications. Proc. Natl. Acad. Sci. USA. 76:4350-4354. Vocke, C. and D. Bastia. 1983. DNA-protein interaction at the origin of DNA replication of the plasmid pSC101. Cell 35:495-502. Wang, Q. P. and J. M. Kaguni. 1987. Transcri tional re ression of the dnaA gene of Escherichia coli by dnaA protein. ol. Gen. enet. 209:518-525. 87 54. Wickner, S. H. and D. K. Chattoraj. 1987. Replication of mini-P1 plasmid DNA in vitro requires two initiation proteins, encoded by the repA gene of phage P1 and the dnaA gene of Escherichia coli. Proc. Natl. Acad. Sci. USA. 84:3668-3672. 55. Wu, F., I. Goldberg and M. Filutowicz. 1992. Roles of a 106-bp ori in enhancer and Escherichia coli DnaA protein in replication of plasmid R6 . Nucleic Acids Res. 20:811-817. Chapter III Novel Alleles of the Escherichia coli dnaA Gene 88 chr sell gro the the clap. slral colle obta that Villel Ihe( Abstract The Escherichia coli dnaA gene is required for replication of the bacterial chromosome. In order to identify residues critical for its replication activity, a selection to obtain novel mutations was developed. The method relied on lytic growth of lambda from an inserted pSC101 replication origin. Replication from the lambda origin was inhibited by cl repressor protein encoded by a lysogen of the host. Replication from the pSC101 origin that resulted in lytic growth was dependent on active DnaA protein that was provided by a plasmid. The host strain lacked the chromosomal dnaA gene. With this approach, a large collection of missense, nonsense, and a few internal deletion mutations were obtained. Nucleotide sequence analysis of the missense mutations indicated that 28 of 50 were unique. Of these, one is identical to the dnaA205 allele whereas the remainder represent novel mutations. Examination of the locations of mutation and replication activity indicate that only those mutations mapping to the C-terrninus were inactive for pSCtOt replication. 89 ac se pr: pII the pit pre lorl haI bio, bio: nur blnl moi IOUI I25, lesi Introduction The Escherichia coli dnaA gene was identified nearly three decades ago (24, 30). It is required for initiation of replication of the bacterial chromosome (18, 24, 30) as well as a number of plasmids and bacteriophage including F (21, 29), pSC101 (15, 16,23), and P1 (10, 11). This requirement correlates with its activity as a sequence-specific DNA binding protein and the observation that sequences termed DnaA boxes that are in these replicons are bound by DnaA protein (17). At oriC and aided by HU protein or integration host factor, DnaA protein unwinds an AT-rich region within oriC (4, 5, 28). As DnaA protein and the replicative helicase, DnaB, interact physically (31), a model has been proposed whereby DnaA protein than loads DnaB in the vicinity of oriC, presumably in the single stranded AT-rich region, to establish both replication forks. In an effort to identify functional domains of DnaA protein, mutant alleles have been characterized genetically (reviewed in references 32 and 38) and biochemically (25, 27, reviewed in references 32 and 38). Although the biochemical characterization of DnaA protein has revealed that it contains numerous activities including sequence-specific DNA binding (17) and ATP binding (37), little has been revealed regarding its domain structure. A P-loop motif (GX4GKT, where X is any amino acid) mapping to positions 172-179 and found in many nucleotide binding proteins has been implicated in ATP binding (25, 27). The DNA binding activity appears to reside in the C-terrninal 94 residues (36). Characterization of a large collection of dnaA alleles has the potential of identifying functional domains of DnaA protein. However, the current collection of mutant alleles is small and their mutations not widely distributed (6, 14, 19, 90 91 reviewed in reference 32), rendering them inadequate for in-depth structure- function studies. To identify a large number of residues of DnaA protein critical for its replication activity, a novel selection for dnaA alleles defective in pSC101 replication was developed. E. coli strains harboring lambda as a lysogen express cl repressor to maintain it as a prophage (35). The activity of the prophage-encoded cl repressor also inhibits subsequent infection by lambda by repressing lytic functions, including the expression of phage-encoded genes required for lambda replication. However, if the infecting lambda contains in its genome an altemative replication origin, lytic growth ensues because of the inability of cl repressor to inhibit replication from this origin (7). We relied on these features to develop a method to select novel mutants of the E. coli dnaA gene. The alternative replicon was the pSC101 origin that depends on DnaA protein for its replication (15, 16, 23). The inability of mutant dnaA alleles to replicate the lambda derivative containing the pSC101 origin provided a powerful method for their selection. Characterization of these novel mutant dnaA alleles revealed three regions that affect the replication activity of DnaA protein. b; Materials and Methods Bacteriological methods. Bacterial strains, plasmids and bacteriophage used in this study are described in Table 1. Bacteriological methods were performed essentially as described by Miller (33). Bacterial transformation was either by the calcium chloride method or by electroporation using a Bio-Rad Gene Pulser according to the manufacturer’s recommendation. Bacterial strains and their plasmid-containing derivatives were routinely grown in LB medium supplemented, where indicated, with 100 pg/ml ampicillin, 40 pg/ml chloramphenicol, 40 pg/ml kanamycin, and 10 pg/ml tetracycline. M9 minimal medium supplemented with the indicated antibiotics and either 7 mM M9804, 0.5% casamino acids, 0.2% maltose, and 1 pg/ml thiamine (for work with phage lambda) or 0.5% casamino acids, 0.2% glucose, and 1 pg/ml thiamine (to measure the phenotypes of dnaA alleles) was used as noted. Lambda lysogens and cross-streaking of bacteria and phage lambda were done essentially as described by Davis et al. (13). Recombinant DNA methods. kpCM128, kpCMt 29, and kimm434 were constructed as described in the legend to Figure 1. in vitro chemical mutagenesis of plasmid DNA. pACYCdnaA DNA (30 pg) was incubated for 60 min at 65°C in 1 ml of 1 M hydroxylamine- HCI (Sigma) and 0.1 M sodium phosphate pH 6.5 (33) then ethanol precipitated 2 times to remove hydroxylamine. Hydroxylamine treatment reduced the transformation efficiency of the plasmid DNA to 10% that of the untreated DNA. Genetic selection for mutations of dnaA defective for replication from the pSC101 origin. A selection method was developed to obtain mutations of the Escherichia coli dnaA gene defective for pSC101 replication (Figure 3). After transformation of M83898 with hydroxylamine- mutagenized pACYCdnaA by selection for chloramphenicol resistant growth at 92 93 TABLE 1. Bacterial strains, plasmids, and bacteriophage8 Strain, plasmid, or Genotypeb or characteristics cgggtrfuecgn bacterIophage Strains HM81 74 recA I'K-12‘ mK,12+) Rifr (F‘) Novagen, Inc. M8175 HM8174 but Iysogenized with Mmm‘m This work M83897 asn832 reIA1 spoT1 thi-1 iIv-192 zia::pKN500 This lab (pKN500=mini-R1) AdnaA mad-2 recA1 (F -) M83898 M83897 but Iysogenized with ltimm“34 This work Plasmids pACYC184 Cm', Tc'; p15A origin, cloning vector Lab stock (9) pACYCdnaA dnaA+ cloned into pACYC184 This lab pACYCdnaA46 dnaA46 cloned into pACYC184 This lab pSCt 01 Tcr; DnaA-dependent low-copy-number S. Cohen (12) replicon pCMf 28 Ap'; copy-number-up mutant of pSC101 S. Cohen (41) Phagec kcI85 7 cit5857 imm" L. Snyder Mlts8578am7 clt5857 Sam7 imm" Lab stock kgtt 0 phage lambda cloning vector, er/13° erM° David Dewitt (26) srl2.5° imm434 M527 (Att -) Aimm’m kgt10 but without Asrl3° and M527 (Att +) This work kpCMt 28 Apr; pCM128 cloned into the Eco RI site of This work kgt10, pSC101 origin replicates leftward lpCM129 Apr; pCM128 cloned into the Eco RI site of This work kgt10, pSC101 origin replicates rightward 8 Abbreviations are Ap, ampicillin; Cm, chloramphenicol; Tc, tetracycline. b See the report of Bachmann (1) for genetic symbols. 0 See Figure 1 for a description of the construction of Mmm434, kpCM128, and ApCM129. he Int he cu ch co Ch. Nu en; IDL Unit Ihe 94 37°C, colonies were replica plated in duplicate using sterile velveteen onto supplemented M9 medium (as described above) seeded with 1011 XpCM128 infective particles. After overnight incubation at either 30° and 42°C, isolates were colony purified on the above medium at the same temperature that they were originally selected for. Immunity to kpCM128 infection was confirmed by cross-streaking. That immunity was due to a specific plasmid-home mutation was confirmed by retransformation of M83898 with the isolated plasmid followed by cross-streaking with kpCM128. Ito/857 and Agt10 were used as controls to confirm that all clones were sensitive to lambda infection (kc/857 sensitive) but expressed phage 434 immunity (kgt10 immune). M83898 harboring pACYC184 (see Table 1), pACYCdnaA, or pACYCdnaA46 were included in parallel as controls. Colonies (5) from the control plate of M83898 harboring pACYCdnaA that were immune to XpCM128 infection were assumed to be spontaneous mutations of dnaA and were pursued in parallel. Immunoblot analysis of DnaA protein. Preparation of whole cell lysates and immunoblot analysis was performed as described (40) except that cultures were grown in LB medium supplemented with kanamycin and chloramphenicol to late log phase. Detection of horseradish peroxidase coupled to goat-anti mouse lmmunoglobulin G (Bio-Rad) was by chemiluminescence (ECL, Amersham). Nucleotide sequence analysis of mutant dnaA alleles. Nucleotide sequence determination was either by the chain-terminating enzymatic method with Sequenase (U. S. Biochemicals) and [a35S]-dATP (Dupont, NEN) or an on-campus sequencing facility equipped with an Applied Biosystems Model 343A DNA Sequencing System. For the latter, at least seven unique primers were used that provided overlapping sequence information for the majority of the gene (Figure 4). Sequence compilations were performed 95 using either Seq-Ed (Applied Biosystems) or Sequencher (version 2.1.1, Gene Codes Corporation). Mutations were identified by comparison of the compiled sequences to the dnaA+ gene sequence in GeneBank (accession no A24944). Template DNAs were prepared either by the Qiagen mini- or midi-column procedure or by PCR amplification of the dnaA gene using the Oligonucleotide primers JK—22 and JK—23 and Deep Vent DNA polymerase (New England Biolabs) with a Perkin Elmer Cetus DNA Thermal Cycler. Oligonucleotide primers were synthesized by an in-house facility with an Applied Biosystems Model 394 and are complementary to the indicated positions relative to the coding region (with position 1 corresponding to the first nucleotide of the coding region; see Figure 4): JK—7, 5’—GTGGAGTCCCATATGTCACTTI'CGCTTTGG—3’ (-12 to 18 of the template strand and contains mismatches at positions -3, -2, -1 and 1); JK—22, 5’-GACCACCTAACGGACCGCTC-3’ (1 ,459 to 1,478 of the coding strand); JK—23, 5’-AAGCCAATITI‘I’GTCTATGG—3’ (-320 to -301 of the template strand); JK—25, 5’—CACGT|'TGATAACTTCGTI'G—3’ (408 to 427 of the template strand); JK—26, 5’—GAGCGCT'I'I'G'I‘|'CAGGACAT—3’ (610 to 629 of the template strand); JK—27, 5’—CGTTGAGGATCGTTTGAAAT—3' (831 to 850 of the template strand); JK—28, 5’—CATTGCCAATGCCAACTTI'A—3’ (1,029 to 1,048 of the template strand); JK—31, 5’—CG'ITATCCCAACCTGAGCGC—3’ (339 to 358 of the coding strand); JK—32, 5’—GGTGAGTTTTACCCAGACCC—3’ (519 to 538 of the coding strand); JK—33, 5’—GAGATCGTTC'I'I'TATTAGCA-3’ (720 to 739 of the coding strand); JK—34, 5’—CGT'ITTCGTCGGCCTTTTTC—3’ (921 to 940 of the coding strand); JK—35, 5’—TCCGCCACCGTCTTCTGAAT—3’ (1,135 to 1,154 of the coding strand). Results Agt10 recombinants containing an alternative replication origin form plaques on a host strain immune to infection by kgt10. Infection by bacteriophage lambda of an Escherichia coli host containing lambda as a prophage is repressed by virtue of the prophage-encoded cl repressor (35). CI repressor inhibits lytic propagation of lambda by repressing expression of genes required for lambda replication. However, lytic growth is observed if the infecting lambda genome contains an alternative replication origin (7). Recombinants of Agt10 were constructed that contained the pSCtO1 origin in both orientations inserted at the unique Eco RI site (Figure 1A). The ability of these bacteriophage to plate on a control E. coli strain, or its isogenic partner was measured (Table 2). Agt10 containing the pSCtO1 origin in either orientation plated with an efficiency of approximately 0.8 on the lysogen expressing phage 434 repressor relative to the nonlysogen. By comparison, nonrecombinant Agt10 was unable to form plaques on the lysogenic strain expressing phage 434 repressor, but did on the nonlysogen. The Agt10 derivatives that were active in lytic growth on a lysogen expressing phage 434 repressor utilized a copy-number-up mutant of pSC101 (kpCM128 and kpCM129; see Figure 1A). As suggested by others (7), we assume that lytic growth of the lambda derivative requires titration of the prophage-encoded cl repressor to a level where it is no longer inhibitory to lambda propagation. The wild type pSC101 origin or oriC inserted in either orientation into a lambda vector were unable to promote cell lysis (data not shown). This observation correlated with the reduced copy numbers of these plasmids relative to pCM128 (41, data not shown). 96 97 Figure 1. (A) Construction of Agt10 derivatives containing the pSC101 origin. kpCM128 and kpCM129 are Agt10 derivatives that were constructed by insertion of Eco RI linearized pCM128, a pSC101 derivative with a copy-number-up repA mutation (41), into the unique Eco RI site of Agt10. For this construction, pCM128 was treated with calf intestinal alkaline phosphatase prior to ligation in vitro with T4 DNA ligase. As insertion into the Eco RI site of Ilgt10 inactivates the cl434 gene, kgt10 recombinants were identified as clear plaques (26), then confirmed by DNA restriction analysis. (B) Construction of a lambda derivative expressing phage 434 repressor and capable of lysogeny. I.imm434 was constructed in vitro by joining of the left arm of kclt88578am7 and the right arm of 719110 at the unique Xhol site. Recombinant phage that were 8+ and imm434 were selected by transformation of E. coli HMS174 (relevant marker, Sup°). The phenotype of recombinants was confirmed by screening for immunity to Agt10 and sensitivity to Ito/857 by cross-streaking colony purified lysogens originally picked from the center of plaques. 98 Eco RI bla lgt10 Ecol RI F M J pSC101 Ab527 imm434 origin Eco RI restriction Eco RI restriction Ligate <—-— Dephosphorylate 5'-ends with CIAP lpCM128 r “7527 pSC101 bla origin kpCM129 r '51-... Ab527 p80101 bla origin kclt3857Sam7 ”'01 r % I imma' Agt10 X hol r = _’///2— 10527 imm4~34 Xhol restriction > Mix ‘ Ligate Mmm434 I X ’10] F E J 99 TABLE 2. Agt10 derivatives containing the pSCtOt replication origin are able to plate on a nonlysogenic and lysogenic host quantitatively . Avera 6 Relative HOSt straIn8 Phageb phage aterc phage titerd HM8174 kgt10 2.4 X 109 M8175 1.9110 NDe —' HM8174 kpCM128 1.6 X109 M8175 kpCM128 1,3 x 109 0'81 HMS174 lpCM129 8.6 X 109 0 78 M8175 kpCM129 6.7 X 109 ' 8 HM8174 and M8175 are isogenic except for the Mmm434 prophage contained in M8175 that expresses the same immunity region as kgtt 0, kpCM128, and IpCM129 (see Table 1). b Lambda derivatives are described in Table 1 and their structures are shown in Figure 1. C Expressed as the number of infective particles per ml of phage Iysate. 0' Expressed as the phage titer calculated on M8175 relative to that calculated on HM8174. 8 ND, none detected. f Relative phage titer could not be calculated due to the inability of Agtto to form plaques on M8175. 100 Replication of lambda bearing the pSC101 origin in a host expressing phage 434 repressor is dnaA-dependent. pSC101 replication is dependent on the dnaA gene (15, 16, 23). To demonstrate that replication of the lambda derivative containing the replication origin of pSC101 was dependent on the dnaA gene, growth of the bacteriophage was tested by spotting 107 infective particles onto a dnaA null host expressing phage 434 repressor and harboring different plasmid-encoded dnaA alleles (Figure 2). Lytic growth was observed on this strain harboring the dnaA+-containing plasmid at 30° or 42°C. Lytic growth was also observed on the strain bearing the plasmid-encoded dnaA46 allele at 30°C, but not at 42°C. dnaA46 is inactivated for pSCtO1 replication at 42°C (16, 23). The negative control of this strain containing the vector only did not support lytic growth at either temperature. Other controls were performed as well. Acl857 propagated on this host at both temperatures and independent of dnaA function. In contrast, growth was not observed with kgt10 (that contains the immunity region of phage 434). These results indicate that the host strain was sensitive to lambda infection and expressed phage 434 repressor. Furthermore, they indicate that growth of the kgt10 derivative containing the pSC101 origin on this host strain is dependent on the dnaA gene and the inserted pSC101 origin. A selection method to obtain novel mutants of the E. coli dnaA gene. To identify residues of DnaA essential for its replication activity, a genetic selection for mutant alleles defective for replication from the pSCtO1 origin was developed (Figure 3). As described above, a lambda derivative containing the pSC101 replication origin with a copy-number—up repA mutation permitted the identification of dnaA mutations either inactive or conditionally defective for pSC101 replication (see Figure 3). Replica plating transfonnants 101 Figure 2. Lytic growth of ApCM128 ls dnaA-dependent. The dnaA null strain expressing phage 434 repressor (M83898) and containing the indicated plasmid-bome dnaA allele was streaked then spot tested with 107 infective particles of each indicated lambda derivative (see Table 1). Plates were incubated at 30° (A) or 42°C (B) for 16 h. SM refers to lambda diluent that was used for dilution of the phage and included as a negative control. 102 M S Ito/85 7 ApCM128 719110 5.5065 3.85053 5663 M S 1.0/85 7 kpCM128 kgt1 0 $5055 $55063 .8653 < 103 Figure 3. Genetic selectlon for mutations of dnaA defective ln pSC101 replication. See Materials and Methods for a description. 00 M53898 cat / dnaA Y Hydroxylamine treated pACYC dnaA Select transfonnants at 37°C on LB media containing kanamycin and chloramphenicol I Replica plate onto supplemented M9 agar seeded with 1011 kpCM128 infective particles and incubate at 30°C and 42°C e I I l 30°C 42°C Cold sensitive or inactive Temperature sensitive or dnaA mutants inactive dnaA mutants pr pl pr: ITIL lire 105 of a dnaA null strain expressing phage 434 repressor and harboring a chemically mutagenized, plasmid-home dnaA gene provided a powerful selection for mutations of dnaA. Using this approach, 289 mutants immune to the lambda derivative were obtained from a collection of 8,921 transfonnants. Five of these 289 were from the control plate of M83898 bearing the nonmutagenized dnaA+-containing plasmid. These were presumed to be spontaneous mutations of dnaA. To confirm the mutant phenotype of these 289 isolates, plasmid DNA corresponding to each putative dnaA mutant was isolated and retransforrned into the dnaA null strain used for their original isolation. By cross-streaking, 266 of the 289 were either temperature sensitive or inactive for lytic growth of the lambda derivative, confirming the plasmid-dependence of the mutant phenotype. Qualitative immunoblot analysis of mutant dnaA alleles. lmmunoblot analysis of whole cell Iysates was performed to identify and discard those plasmid-home mutations that resulted in a reduced level of DnaA protein due to a reduced copy number of the plasmid, production of an unstable polypeptide, or insufficient expression, perhaps due to a mutation in the dnaA promoter region (data not shown). As controls, transfonnants of the vector, pACY0184, or it containing the dnaA+ or dnaA46 allele were included. No product was detected for the pACYC184 transfonnant, confirming the dnaA null mutation of the host strain. Immunoreactive polypeptides that comigrated at the position of purified DnaA protein were observed for the dnaA+ and dnaA46 transfonnants. Immunoreactive polypeptides were detected for 129 of the 266 plasmid-bome mutations. Of these, 50 appeared identical in size to DnaA protein and were presumed to be missense mutants. This interpretation was confirmed by DNA 106 sequence analysis (see below). 76 encoded proteins with greater electrophoretic mobilities than DnaA protein. We presumed that these encoded nonsense mutations. DNA sequence analysis of representative isolates identified seven unique nonsense mutations (data not shown). In addition, a few mutations containing internal, in-frame deletions of coding sequence were identified (data not shown). The nonsense mutations and those containing an internal, in-frame deletions will be described elsewhere (see Chapter IV). As no immunoreactive species was detected for the remaining 137 pIasmid-bome mutations, we assume that they contained mutations either in the promoter region to diminish expression or elsewhere that resulted in unstable gene products. A third possibility is that they contain a mutation that reduced the copy number of the plasmid to lower the amount of DnaA protein to undetectable levels. These mutants were not characterized further. Nucleotide sequence analysis of mutant dnaA alleles. The group of 50 missense mutations were sequenced as described in Materials and Methods (see Figure 4). 28 of these 50 were unique (Table 3). 27 of the 28 were novel with one of the 28 being identical to the dnaA205 allele identified previously (3). The five mutations identified on the control plate of the dnaA null strain harboring the non-mutagenized dnaA+-containing plasmid were confirmed and found to represent two different alleles (A412P and 412[QMA]413; see footnote a to Table 3). In all, the 28 novel alleles identified 31 different residues that when altered affect the replication activity of DnaA protein. Many result in substitutions of residues that are extremely well conserved among dnaA homologs (reviewed in reference 32) (Table 3). Four alleles contain mutations that result in substitutions at the same positions as in alleles previously identified (dnaA508, P-28 (proline at position 28); dnaA5, 46, 601/602, 604/606, A-184; dnaA205, V-383; dnaA5, G-426) (22). However, 107 Figure 4. Scheme for nucleotide sequence analysis of dnaA alleles. The top portion represents the Cla I-Xho I dnaA gene fragment contained in pACYC184 (40). The bold line represents pACYC184 sequence. Nucleotide sequence determinations were performed as described in Materials and Methods. At least seven of the 14 indicated primers were used to sequence each allele. The seven primers used in most cases are indicated within dashed boxes that correspond to the portion of the gene sequenced for each allele. Others were used to resolve ambiguities or to confirm mutations. Arrows represent the approximate location and extent of sequence information obtained, on average, for each primer. In some cases, if a clone with a phenotype similar to an allele already sequenced in its entirety was found to contain mutation(s) identical to the latter allele after partial nucleotide sequence analysis, the remaining portion was not sequenced and the alleles were assumed to be identical. Upon completion of the nucleotide sequence analysis (and consistent with their almost wild-type phenotype with respect to lytic growth of ApCM128), no mutations were detected in three of the presumed 50 missense mutations (data not shown). It is possible that these three may have contained mutations that, for some reason, were unstable, thereby preventing their detection. A similar phenomenon has been reported for a mutation residing in dnaA that suppresses the thermolabile phenotype of the dnaX2016(Ts) allele (19). 108 gm _ X36 :9»: 8883 833 _ _ _ OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII ooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooo 109 TABLE 3. Nucleotide and deduced amino acid alterations of novel dnaA missense mutations Mutant dnaA- . . Amino Deduced amino acid Number of containing N13233:? s:::I:3I:gre1c acid substitution (degree of times plasmid8 p position conservation)d identified8 pACHA-9 68 G—>A 23 Ser—rAsn (5/15) 1 976 A—+T 326 Ala—>Thr (14/15) 1 .397 C->T 466 Ser—>Leu (5/15) pACHA-32 77 T—rA 26 lle->Lys (9/15) 1 1 .354 C-vT 452 His-+Tyr (5/15) pACHA-39 26 G—-A 9 Cys—var (6/15) 1 191 C—vT 64 Thr—>Ile (1/15) 1 ,1 10 G—rA 370 Silent (Gln) 1 ,1 49 G—vA 383 Silent (Val) pACHA-40 3 G-vA 1 Met->Ala (1 5/15) 3 pACHA-43 1 .318 G—>A 440 Ala-vThr (12/15) 2 pACHA-45 26 G—>A 9 Cys-+Tyr (6/15) 5 pACHA-49 1 .304 C-+T 435 Thr—rMet (14/15) 3 pACHA-50 74 G-+C 25 Trp-+Ser (12/15) 1 pACHA-54 1 .220 G—>A 407 Arg—vHis (15/15) 1 pACHA-70 1 .309 G-vA 437 Val—vMet (15/15) 3 pACHA-72 61 G—-A 21 Glu—>Lys (6/15) 2 1.099 C->T 367 Silent (Leu) 1,362 C—>T 454 Silent (Leu) pACHA-79 1 .277 G—+A 426 Gly—+Asp (15/15) 2 pACHA-83 1 .31 9 C—vT 440 Ala—>Val (12/15) pACHA-94 530 G->A 1 77 Gly—>Asp (15/15) 742 G—>A 248 Glu-+Lys (14/15) 745 G->A 249 GIu—>Lys (13/15) pACHA-101 524 G—rA 175 Gly—>Asp (15/15) 1 pACHA-105 550 G—>A 184 Ala-Thr (14/15) 2 1 ,1 13 A->G 371 Silent (Glu) 1,224 G->A 408 Silent (Gln) pACHA-1 06 2 6 G—+A 9 Cys-+Tyr (6/1 5) 1 878 C->T 293 Ala—Nat (7/15) pACHA-110 1.147 G—>A 383 Val—+Met (11/15) 1 pACHA-117 128 C—>T 43 Pro-+Leu (11/15) 1 529 G—>A 177 Gly—vSer (15/15) [.11 DA DA 110 TABLE 3 (continued). Mutant dnaA- . . Amino Deduced amino acid Number containg N13333:? s 3513:1336: 0 acid substitution (degree of of times plasmid8 p position conservation)d identifiede pACHA-130 1,282 G—+A 428 Ala—+Thr (5/15) 1 1,318 G—>A 440 Ala-+Thr (12/15) pACHA-143 28 C—iT 10 Leu-vPhe (9/15) 1 1 54 C-T 52 Arg—rTrp (5/15) pACHA-146 530 G—rA 177 Gly-rAsp (15/15) 1 1 .149 G—+A 383 Silent (Val) 1 ,1 64 G—>A 488 Silent (Lys) pACHA-152 82 C-rT 28 Pro—-Leu (6/15) 1 83 C—rT pACHA-162 625 G—>A 209 Asp—>Asn (8/15) 1 1.304 C—>T 435 Thr—-Met (14/15) pACHA-170 1.304 C—-T 435 Thr—-Met (14/15) 1 1,307 C—»T 436 Thr—>Met (15/15) pACHA-200 48 G—rA 16 Silent (Glu) 1 61 G-+A 21 Glu—aLys (6/15) 430 G—+A 144 GIy—+Asn (14/15) 431 G->A 526 C—-T 176 Silent (Leu) pACM-2 1 .234 G—-C 41 2 Ala—*Pro (6/15) 1 pACM-7 1.237- (unease)f 413- 412(0 Met-Ala)413 4 1.242 414 8 Some M83898 (pACYCdnaA) transfonnants were immune to kpCM128 infection (data not shown). These were pursued under the assumption that they may represent spontaneous mutations of dnaA. pACHA refers to mutations of dnaA identified by hydroxylamine treatment of pACYCdnaA. pACM refers to mutations of dnaA identified from non-hydroxylamine treated pACYCdnaA control transformants. 8 Numbers refer to the dnaA coding region, with position one corresponding to the first nucleotide of the coding sequence (20). 8 Nucleotide substitutions are expressed as those in the coding strand. 8 Sequence conservation is indicated by the frequency of residues conserved among the 15 dnaA homologs (reviewed in reference 32). 8 The number of times that each allele was identified is indicated. No mutations were detected in three of the presumed 50 missense mutations (data not shown). ’ Q(ATGGCG) indicated an insertion of the noted nucleotides at the indicated location. 0t ne 3U 111 substitutions at positions A-184 and G-426 are with different amino acid residues than in previously isolated alleles. For dnaA508. a secondary mutation not present in the allele described here substitutes threonine at position 80 with isoleucine. Among the 50 alleles characterized at the nucleotide level, all but six mutations were consistent with hydroxylamine treatment (see Table 3) that produces C-+T transitions (34. 13). We assume that these six mutations arose spontaneously. Indeed. two of the six (A412P and 412[QMA]413) were identified on the plate corresponding to the control of the host strain harboring the dnaA+-containing plasmid that was not mutagenized (see footnote a to Table 3). Replication phenotypes of novel dnaA alleles are elther inactive or temperature sensitive. We were interested in determining if the replication phenotypes among the missense mutations might be distinguishable. To do this with the number of mutations in hand. about 105 cells of the dnaA null strain expressing phage 434 repressor and containing the different plasmid-borne dnaA alleles were spotted with 102, 103. and 104 infective particles of kpCM128 to test the efficiency whereby each allele supported lytic growth of the lambda derivative from the pSC101 origin (Table 4). Lytic growth was observed with all phage dilutions spotted onto the transformant harboring the plasmid-borne dnaA+ allele at both 30° and 42°C. By comparison. lytic growth on the transformant bearing the dnaA46 allele was observed at 30°C, the permissive temperature for dnaA46. but not at 42°C. The negative control of the transformant harboring the vector pACYC184 did not support lytic growth at either temperature. Of the 28 missense mutations, two were almost wild-type for dnaA- function. 21 were temperature sensitive. and five were inactive. The latter class 112 TABLE 4. Summary of replication phenotypes of novel dnaA mutations kpCM128 spottingc 1'“ H 104 103 10?- 104 103 pACYCdnaA None + + + + + Wild type pACYCdnaA46 A1 84V-H252Y + + + +/— — Ts pACYC1 84 — - - — — Inactive pACHA-9 823N-A326T-S466L + + + + + ~Wild type pACHA-32 l26K-H452Y + + + + + ~Wild type pACHA-39 CQY-T64I + - — - - Ts pACHA-40 M1 V + + - + — Ts pACHA-43 A440T + + + + + T8 pACHA-45 C9Y + + - — — Ts pACHA-49 T435M + - — - - Ts pACHA-50 W258 + - - — — Ts pACHA-54 R407H - - - - - Inactive pACHA-70 V437M + + + - - Ts pACHA-72 E21 K + + - — — Ts pACHA-79 G426D — — - - - Inactive pACHA-83 A440V — - - - Ts pACHA-94 G1770-E248K- + + + + - Ts E249K pACHA-101 61750 -I- + + + + T8 pACHA-105 A1 84T + + + + + T8 pACHA-1 06 CQY-A293V + + — - - Ts pACHA-1 10 V383M + + + — — Ts pACHA-1 17 P43L-G177S + + - — - Ts pACHA-1 30 A428T-A440T + — - — - Ts pACHA-143 L10F-R52W -I- + + —- — Ts pACHA-146 G177D + + + + — Ts pACHA-152 P28l. + + + + + T8 pACHA-1 62 D209N-T435M + — - - - Ts pA pA pA alp was MSG Iran. (MI W 113 TABLE 4 (continued). ApCM128 spottingc dnaA-containing Deduced amino acid 30°C 42°C Replication plasmid8 substitution(s)b phenotyped 104 103 102 104 103 102 pACHA-1 70 T435M-T436M — — - — - - Inactive pACHA-200 E21 K-Gt 44N + + - - — - Ts pACM-2 A41 2P — — - — — — Inactive pACM-7 412(QM-A)4138 - — — - — - Inactive 8 See footnote a to Table 3. b Deduced from results presented in Table 3 and expressed as the amino acid present in wild type DnaA, its position, and the deduced substitution (i.e. A184V represents an A-to-V mutation at position 184). 8 The ability of the individual dnaA mutations to propagate LpCM128 from the pSC101 origin was measured by spotting dilutions of LpCM128 (104-102 infective particles) onto ~105 cells of M83898 bearing the respective plasmid-home allele. Growth of hpCMt 28 on the above M83898 transfonnants at 30° and 42°C on supplemented M9 medium was compared with M83898 harboring the vector only (pACYC184). or it containing the dnaA+ (pACYCdnaA) or dnaA46 (pACYCdnaA46) alleles. Abbreviations are +, bacterial clearing (kpCM128 sensitivity); -. bacterial growth (kpCM128 immunity); +/— partial clearing (moderate ApCM128 sensitivity). 0' Deduced from the kpCM128 spotting. Ts, temperature sensitive. 8 412(QMA)413 is a Met. Ala insertion between residues 412 and 413. 114 contained substitutions in the C-terrninal region within residues 407 through 436 (see Table 4). The inability of these five mutant alleles to support replication of pSC101 was confirmed by a transformation assay (data not shown). Whereas transfonnants of the dnaA null strain harboring each of these alleles were transformed efficiently with pBR322 at either 30° or 42°C, they could not be transformed with pSC101. As a control. this strain bearing the plasmid-encoded dnaA+ allele was transformed by pSC101. If this strain harbored a plasmid encoding the dnaA46 allele instead. pSCIO1 transfonnants were obtained at 30°C. but not at 42°C. E1 E5 rel r») Discussion Identification and nucleotide sequence of novel mutant Escherichia coli dnaA alleles. To obtain a large collection of novel Escherichia coli dnaA alleles to identify residues critical for its activity in replication, we developed a method to select mutants defective for pSC101 replication. Relying on the ability of lambda lysogens to exclude superinfection (35), we constructed a lambda derivative containing the immunity region of phage 434 and a pSC101 origin with a copy-number—up mutation. Lytic growth of this lambda derivative on a lysogen expressing phage 434 repressor was dependent on the pSC101 origin and the dnaA gene. With this lambda derivative to infect a dnaA null strain expressing phage 434 repressor, and a plasmid bearing the dnaA gene mutagenized with hydroxylamine, dnaA mutants were selected that were inactive or conditionally defective. These failed to support lytic growth and remained viable in the presence of the lambda derivative. By this approach, we identified 28 unique missense mutations of the E. coli dnaA gene. Twenty seven of the 28 are novel. In all, the 28 alleles identified 31 different residues of DnaA that when substituted affect its ability to support pSC101 replication. Many result in substitutions at positions that are highly conserved among dnaA homologs (T able 3). That no cold sensitive dnaA mutations were identified may be due to the fact that the host strain used for selection of dnaA mutations contained an intact oriC locus. It is possible that a cold sensitive mutant would over initiate replication from oriC, thereby resulting in cell death as is the case with the dnaAcos mutant (6). Alternatively, the possibility that DnaA protein may facilitate replication from the integrated R1 origin (8, 2, 21) may also have 115 116 affected the types of mutations obtained. A third possibility is that cold sensitive dnaA alleles may require multiple mutations as in the case of dnaAcos (6) and that the in vitro mutagenesis method used was insufficient for generation of multiple mutations. Missense mutations residing in the C-terminus of DnaA abolish replication activity. The ability of the various plasmid-home mutations to support lytic growth of the lambda derivative from the pSC101 replication origin was measured by a spot test. Of the 28 missense mutations, 21 were thermolabile, five were inactive. and two were similar to wild type dnaA for lytic growth of the lambda derivative. Interestingly, all of the mutations that abolished replication activity mapped to residues 407 through 436 (Table 4). That these alleles were inactive for replication of pSC101 was substantiated by our inability to transform a dnaA null strain harboring the respective plasmid- bome alleles to tetracycline resistance with the wild-type pSC101 plasmid (data not shown). As the C-terminal region of DnaA is involved in DNA binding (36. Chapter V), we speculate that these mutations affect the DNA binding activity of DnaA protein. The 21 alleles that were thermolabile for pSC101 replication varied both in their degree of temperature sensitivity and overall activity (T able 4). Furthermore. some of the mutants encoding more than one amino acid substitution were comparable in phenotype to single mutants containing an identical mutation. This finding suggests that the substitutions of Qt44N. D209N. E248K-E249K, and A293V that are found with other substitutions that are unique for some alleles may have little effect on replication activity. Finally, it is of particular interest that two alleles with different substitutions at position 440 exhibit such diverse replication phenotypes. Whereas A440V had only minimal replication activity, A440T was nearly wild-type in its ability to support 117 lytic growth of the lambda derivative (Table 4). DnaA contains three distinct domains. The missense mutations were clustered in three discrete regions of the dnaA gene (Figure 5). suggesting the presence of three distinct domains of DnaA protein essential for replication of pSC101. The C-terrninal 94 residues have been determined to be both essential and sufficient for specific DNA binding (36). Although a function for the N-terminus of DnaA is unknown, the large number of mutations mapping to this region indicates its essential role. The central region of DnaA protein contains a P-Ioop motif (GX4GKT, where X represents any amino acid) mapping to positions 172-179 and is thought to comprise a portion of the high affinity ATP binding site. Consistent with this assumption, an alanine-to-valine substitution at position 184 near this P-loop motif has been shown to affect the ability of DnaA to bind ATP (27. 25). Furthermore, mutant alleles containing an alanine-to-valine mutation at position 184 initiate replication from oriC asynchronously whereas those containing mutations mapping elsewhere in the gene do not (39). Although each of the alleles containing the A184V mutation contains a second mutation as well. it has been speculated that the A184V mutation is responsible for the asynchrony in initiation and that the ATP-bound form of DnaA is essential for synchronous initiation of multiple origins in rapidly growing cells (38). As the ability of the A184V mutation, without additional secondary mutations. to initiate replication synchronously from multiple origins is not yet known, it will be interesting to see if the novel alleles described here containing mutations mapping within or near the P-loop motif exhibit asynchrony in initiation. Whether the 118 Figure 5. Summary of positions of mutations of novel dnaA alleles. Relative positions of mutations resulting in the novel dnaA alleles described here are indicated. Each blackened triangle (A) represents the approximate location of a single mutation. The mutations indicated between residues 200 and 350 are not unique and contain one or more additional mutations that may account for its mutant phenotype (see Table 3). (2 indicates the position of the Met, Ala insertion mutation between residues 412 and 413. Amino acid residues and proposed functional domains of DnaA protein (reviewed in references 32 and 39) are indicated. 119 momacm A ._oo Proou Boa moo _ woo watt! b F 02> c.3650 BBo emu 120 mutations mapping to this region of the protein affect ATP binding activity may be addressed following their purification. Finally, the genetic and biochemical characterization of the mutant alleles described here will provide significant insight into functional domains of DnaA and aid in the development of a more refined model for the function of DnaA in initiation of DNA replication in Escherichia coli. 10. 11. 12. 13. References Bachmann, B. J. 1990. Linkage map of Escherichia coli K-12, edition 8. Microbiol. Rev. 54:130-97. Bernander, R., S. Dasgupta and K. Nordstréim. 1991. The E. coli cell cycle and the plasmid R1 replication cycle in the absence of the DnaA protein. Cell. 64: 1145- 1153. Beyersmann, D., W. Messer and M. Schlicht. 1974. Mutants of Escherichia coli B/r Defective in Deoxyribonucleic Acid Initiation: dnal, a New Gone for Replication. J. Bacteriol. 118:783-789. Bramhill, D. and A. Kornberg. 1988. Duplex opening b dnaA protein at novel sequences in initiation of replication at the origin 0 the E. coli chromosome. Cell. 52:743-755. Bramhill, D. and A. Kornberg. 1988. A model for initiation at origins of DNA replication. Cell. 54:915-918. Braun, R. E, K. OD a4, and A. Wright. 1987. Cloning and characterization of dn (Cs), a mutation which leads to overinitiation of DNA replication in Escherichia coli K- 12. J. Bacteriol. 169: 3898- 3903. Brenner, 8., G. Cesareni and J. Karn. 1982. Phasmids: hybrids between ColE1 plasmids and E. coli bacteriophage lambda. Gene. 17:27-44. Chandler, M., L. Silver and L. Caro. 1977. Suppression of an Escherichia coli dnaA mutation by the integrated R factor R100.1: origin of ggr'omosome replication during exponential growth. J. Bacteriol. :421-430. Chang, A. C. Y. and S. N. Cohen. 1978. Construction and Characterization of Amalifyable Multicopy DNA Cloning Vehicles Derived from the P15A Cryptic iniplasmid. J. Bacteriol. 134:1141-1156. Chesney, R. H. and J. R. Scott. 1978. Suppression of a thermosensitive dnaA mutation of Escherichia coli by bacteriophage P1 and P7. Plasmid. 1:145-63. Chesney, R. H., J. R. Scott and D. Vapnek. 1979. Integration of the {Inasmid prophages P1 and P7 into the chromosome of Escherichia coli. J. ol Biol. 130: 161-73. Cohen, S. N. and A. C. Change. 1977. Revised interpretation of the origin of the pSC101 plasmid. J. Bacteriol. 132:734-737. Davis, R. W., D. Botsteln and J. R. Roth. 1980. A Manual for Genetic Engineering: Advanced Bacterial Genetics. 12] 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 122 Eberle, H., W. Van de Merwe, K. Madden, G. Kampo, L. Wright and K. Donlon. 1989. The nature of an intragenic suppressor of the Escherichia coli dnaA508 temperature-sensitive mutation. Gene. 84:237-245. Felton, J. and A. Wright. 1979. Plasmid pSC101 replication in integratively suppressed cells requires dnaA function. Mol. Gen. Genet. 175:231-233. Frey, J., M. Chandler and L. Caro. 1979. The effects of an Escherichia coli dnaAts mutation on the re lication of the plasmids ColE1 pSC101. R100.1 and RTF-TC. Mol. Gen. enet. 174:117-26. Fuller, R. 8., B. E. Funnell and A. Kornberg. 1984. The dnaA protein complex with the E. coli chromosomal replication origin (oriC) and other DNA sites. Cell. 38:889-900. Fuller, R. S. and A. Kornberg. 1983. Purified dnaA protein in initiation of replication at the Escherichia coli chromosomal origin of replication. Proc. Natl. Acad. Sci., U. S. A. 80:5817-5821. Ginés-Candelaria, E., A. Bllnkova and J. R. Walker. 1995. Mutations in Escherichia coli dnaA which suppress a dnaX(Ts) polymerization mutation and are dominant when located in the chromosomal allele and recessive on plasmids. J. Bacteriol. 177:705-715. Hansen, E. B., F. G. Hansen and K. von Meyenburg. 1982. The nucleotide sequence of the dnaA gene and the first part of the dnaN gene of Escherichia coli K-12. Nucleic Acids Res. 10:7373-7385. Hansen. E. B. and M. B. Yarmolinsky. 1986. Host participation in glasmid maintainance: Dependence upon dnaA of replicons derived from 1 and F. Proc. Natl. Acad. Sci., U. 8. A. 83:4423-4427. Hansen, F. G., 8. Koefoed and T. Atlung. 1992. Cloning and nucleotide sequence determination of twelve mutant dnaA genes of Escherichia coli. Mol. Gen. Genet. 234:14-21. Hasunuma, K. and M. Sekiguchi. 1977. Replication of plasmid pSC101 in Escherichia coli K12: requirement for dnaA functron. Mol. Gen. Genet. 154:225-30. Hirota, Y., A. Ryter and F. Jacob. Thermosensitive mutants of E. coli affected in the processes of DNA 8 nthesis and cellular division. In (eds.). Cold Spring Harbor Symp. Quant. iol., 1968. Hupp, T. R. and J. M. Kaguni. 1993. DnaA5 protein is thermolabile in initiation of replication from the chromosomal origin of Escherichia coli. J. Biol. Chem. 268:13128-13136. Huynh, T. V., R. A. Young and R. W. Davis. In Glover, D. M. (eds), DNA Cloning: A Practical Approach. 1985. IRL Press, Oxford. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 123 Hwang, D. S. and J. M. Kaguni. 1988. Interaction of dnaA46 protein with.a stimulatory protein in replication from the Escherichia coli chromosomal orIgin. J. Biol. Chem. 263:10633-10640. Hwang, D. 8. and A. Kornbear'g. 1992. Opening of the replication origin of Escherichia coli by Dn protein with protein HU or HF. J. Biol. Chem. 267:23083-23086. Kline, B. C., T. Kogoma, J. E. Tarn and M. S. Shields. 1986. Requirement of the Escherichia coli dnaA gene product for plasmid F maintenance. J. Bacteriol. 168:440-3. Kohiyama, M., D. Cousin, A. Ryter and F. Jacob. 1966. [Thermosensitive mutants of Escherichia coli K 12. I. Isolation and rapid characterization]. Ann Inst Pasteur (Paris). 110:465-486. Marszalek, J. and J. M. Kaguni. 1994. DnaA protein directs the binding of DnaB protein in initIation of DNA replication in Escherichia coli. J. Biol. Chem. 269:4883-4890. Messer, W. and C. Weigel. Initiation of chromosome replication. In Neidhardt, F. C., J. l_. Ingraham, K. B. Low, B. Magasanik, M. Schaechter and H. E. Umbarger (eds.), Escherichia coli and almonella typhimurium. Cellular and Molecular Biology, 1996. American Society for MIcrobiology, Washington. D. C. Miller, J. H. 1992. A short course in bacterial genetics. Myers, R. M., L. S. Lerman and T. Maniatis. 1985. A general method for saturation mutagenesis of cloned DNA fragments. Science. 229:242-7. Ptashne, M., K. Backman, M. Z. Humayun, A. Jeffrey, R. Maurer, B. Meyer and R. T. Sauer. 1976. Autoregulation and function of a repressor in bacteriophage lambda. Science. 194:156-61. Roth, A. and W. Messer. 1995. The DNA binding domain of the initiator protein DnaA. EMBO J. 14:2106-2111. Sekimizu, K., D. Bramhill and A. Kornberg. 1987. ATP activates dnaA protein in initiating replication of plasmids bearing the origin of the E. coli chromosome. Cell. 50:259-265. Skarstad, K. and E. Bo e. 1994. The initiator protein DnaA: evolution. properties and function. iochim. Biophys. Acta. 1217:111-130. Skarstad, K., K. von Meyenburg, F. G. Hansen and E. Boye. 1988. Coordination of chromosome replication initiation in Escherichia coli effects of different dnaA alleles. J. Bacteriol. 170:852-858. 40. 41. 124 Sutton, M. D. Ph. D. Thesis, Chapter II. Tucker, W. T., C. A. Miller and S. N. Cohen. 1984. Structural and functional analysis of the par region of the pSClO1 plasmid. Cell. 38:191-201. Chapter IV Genetic and Biochemical Characterization of Novel Alleles of the Escherichia coli dnaA Gene: Identification of Four Functionally Distinct Domains of DnaA Protein 125 Abstract The Escherichia coli DnaA protein is a sequence-specific DNA binding protein that promotes the initiation of replication of the bacterial chromosome. pSC101, and other plasmids. We recently described 28 missense mutations of the E. coli dnaA gene, identified by their inability to replicate a recombinant phage lambda derivative from the inserted pSC101 origin (see Chapter III). Here, we describe the genetic and biochemical characterization of these missense dnaA alleles as well as the identification and characterization of seven nonsense mutations and one in-frame deletion mutation. Our findings suggest that the E. coli DnaA protein contains four functionally distinct domains. Mutations that affect residues in the P-loop motif, proposed to function in ATP binding, identify one domain. The second domain maps to the C-terrninus and is involved in DNA binding. The function of the third domain that maps near the N-terminus is unknown. Two mutant alleles encoding different truncated gene products retained the ability to promote replication from the pSC101 origin but not oriC. identifying a fourth domain dispensable for replication of pSC101 but essential for oriC. 126 Introduction DnaA protein is a sequence-specific DNA binding protein that is essential for initiation of replication from the Escherichia coli chromosomal origin, oriC (14). DnaA recognizes and binds to a 9-mer sequence, termed the DnaA box, that is found four times in the E. coli replication origin (28). as well as in the replication origins of plasmids (12, 13. 19. 17, 24) and bacteriophage (17, 44) which require this protein for replication. Once bound at oriC, DnaA is proposed to induce a structural alteration in the AT-rich region of oriC. consisting of three tandem repeats of a 13-mer AT-rich sequence. leading to their unwinding and formation of the prepriming complex (3. 38). Little is known to correlate the structure of DnaA protein to its various biochemical activities. The isolation of extragenic suppressors of conditional dnaA mutations has revealed a number of gene products proposed to interact either functionally or physically with DnaA protein (1, 15, 20, 33). However, with the exception of DnaB helicase (26). biochemical evidence in support of these physical interactions is lacking. . DnaA protein binds ATP and ADP with high affinity (KD of 0.03 pM and 0.10 pM. respectively) (37). Examination of the amino acid sequence identifies a P-Ioop motif (GX4GKT from residues 172-179, where X is any amino acid) found in many nucleotide binding proteins (35). An alanine-to-valine substitution at residue 184 near this P-loop motif has been shown to reduce ATP binding affinity (21, 22), supporting the model that this region functions in ATP binding. DnaA protein interacts physically with DnaB protein (26). This interaction is inhibited by a monoclonal antibody that binds to an epitope within the N-terminal 147 residues of DnaA protein. These observations suggest that 127 128 residues within this region are involved in binding to DnaB. In another study, the region of DnaA protein involved in DNA binding was localized to the C-temIinal 94 residues. This was determined by fusing portions of DnaA protein to B-galactosidase then measuring DNA binding activity (32). Here, we report the genetic and biochemical characterization of the mutations described in Chapter III as well as the identification and characterization of seven nonsense mutations and one in-frame deletion mutation. Examination of the mutant phenotypes relative to the positions of the mutations suggests the presence of four functionally distinct domains of DnaA protein. One class resides near a P-loop motif that appears to be involved in ATP binding. The second affects DNA binding activity. The third group is postulated to affect the interaction between DnaA protein and pSC101 RepA protein. The fourth class maps to a region near the N-terrninus. The function of this region is unknown. Comparison of these findings with the predicted secondary structure of DnaA protein obtained using the PHD method (31) revealed that many substituted amino acids fall in regions of predicted secondary structure, and affect positions that are highly conserved. Materials and Methods Bacteriological methods. The bacterial strains, plasmid DNAs, and phage lambda used in this study are described in Table 1. Bacteriological methods were performed essentially as described by Miller (27). Bacterial transformation was by the calcium chloride method. Lambda lysogens were constructed as described by Davis et al. (11). Bacterial strains and their plasmid-containing derivatives were routinely grown in LB medium supplemented, as indicated, with 100 pg /ml ampicillin, 40 pg/ml chloramphenicol. 40 pg/ml kanamycin, and 10 pg/ml tetracycline. M9 minimal medium supplemented with 0.5% casamino acids, 0.2% dextrose, 1 pg/ml thiamine, and the indicated antibiotics was used as noted. Recombinant DNA methods. pACYCdnaA and pACYCdnaA46 are pACYC184 recombinants (41) that contain the dnaA+ and dnaA46(Ts) alleles. respectively. pMD8421 is a chloramphenicol sensitive derivative of pOU420 (17) constructed by Eco RI restriction of pOU420, end-filling of the linear DNA with the large fragment of DNA polymerase I and dATP and dTTP, and its religation with T4 DNA ligase (Boehringer Mannheim) to obtain a four base pair insertion at the Eco RI site within the cat gene resulting in its inactivation. Selection for mutations of dnaA defective for replication from the pSC101 origin. Mutations of dnaA encoding truncated gene products were identified. and their respective replication phenotypes measured. utilizing the phasemid assay described previously (see Chapter III). Briefly, this assay measures the ability of a plasmid-encoded dnaA allele to direct replication of a phage lambda vector that contains a pSC101 origin (ApCM128). Such a recombinant phage lambda containing a secondary origin of replication is referred to as a lambda phasemid (8). As the host strain contains a prophage 129 130 TABLE 1. Bacterial strains and plasmidsa . . b . . Source or Strain or plasmId Genotype or characterrstrcs con stru on on Strains MC1061 ara0139 A(ara-Ieu)7696 AIacX74 gaIU gaIK Laboratory stock hSdFIZ (ix-12‘ mK-12+) ("0’3 1 ’PSL (F ') M81062 MC1061 but zia::pKN500 This work M83896 asn832 relAf spoT1 fhi-1 IysA fuc-1 iIv-192 This laboratory zia::pKN500 (pKN500=mini-R1) AdnaA mad-2 (F ‘) M83898 M83896: recA1 IysA+ fucf M34 This laboratory M8207 M83898 but dnaA204(Ts) This laboratory SH210 fhuA22 garBiO A(argF-IacU) 169 zai-736‘.:Tn 10 CGSC # 7096 (36) phoA8 ompF627 fadL 701 rel/I1 glpFI2 glpD3 pit-10 spoT1 phoM510 mch Hfr (POZA of Cavalll l-lfr) M83899 M83896 but A(argF-IacU)169 zai::Tn 10 (M83896) X (SH210). select Tc'. screen for Lac - M83900 M83899 but ARB1c This work M83907 M83898 but dnaA+ This work Plasmids pACYC184 Cm', Tc'; p15A origin. cloning vector LaboratOIy stock (9) pACYCdnaA dnaA+ cloned into pACYC184 This laboratory (41) pACYCdnaA46 dnaA46(Ts) cloned into pACYC184 This laboratory (41) pOU420 Apr. Cm'; pBR325 origin, contains an R1 M. Yarmolinsky (17) copA(Ts) allele pMDS421 Ap'; Cm3 derivative of pOU420 This work pSC101 Tcr; DnaA-dependent low-copy-number S. Cohen (10) replicon pCM128 Ap'; copy-number-up mutant of pSC101 S. Cohen (42) 8 Abbreviations are Ap, ampicillin; Cm. chloramphenicol; Tc, tetracycline; Ts. thermolabile; X, bacterial mating; CGSC, E. coli genetic stock center. 8 See the report of Bachmann (2) for genetic symbols. 8 ARB1 is a ANN 955 derivative that contains a dnaA ’-’IacZ translational fusion (7). 131 that expresses the same immunity region as kpCM128. normal lambda propagation of the phasemid is repressed by virtue of the prophage-encoded cl repressor protein (29). However, replication from the pSC101 origin. which is DnaA-dependent, has the ability to produce a kpCM128 copy number sufficiently high to titrate out the prophage encoded repressor protein, thereby promoting cell lysis. Nucleotide sequence analysis of mutant dnaA alleles. The approximate size of each mutant protein was determined by immunoblot analysis of whole cell Iysates of a dnaA null strain harboring the respective plasmid-encoded dnaA mutations. Relevant regions for each of the seven presumed nonsense alleles were then sequenced by the chain-terminating enzymatic method using Sequenase (U.S. Biochemicals) essentially as described by the manufacturer. Plasmid DNAs were purified using the Qiagen mini- or midi-column method. The approximate location of the coding sequence deletion from the pACHA-4 mutation was determined by endonuclease restriction analysis of the plasmid DNA relative to pACYCdnaA and markers of known size. The appropriate region was then sequenced as described above. Oligonucleotide primers JK—21. 5’—ATAACCCGTTG'ITCC—3’; JK—22, 5'—GACCACCTAACGGACCGCTC—3'; JK—26. 5’-GAGCGCTTI'GTTCAGGACAT—3’; JK—27. 5'—CG'I'I'GAGGATCGT'I'I'GAAAT—3’; JK—32. 5’-GGTGAGTTTI'ACCCAGACCC—3’. synthesized by an in-house facility with an Applied Biosystems model 394. were used for sequencing and are homologous to the following nucleotide positions of dnaA (with position 1 corresponding to the first nucleotide of the coding sequence (16)): JK—21, 494 to 508 of the coding strand; JK—22, 1,459 to 1,478 of the template strand; JK—26. 610 to 629 of the coding strand; JK—27. 831 to 850 of the coding strand; 132 JK—32, 522 to 541 of the template strand. B-galactosidase assays. Cultures were grown in supplemented M9 medium at 30° and 42°C. When growth approached mid-log phase, as determined by ODeoo, cells were lysed by sonication on ice using a Heat Systems-Ultrasonics, Inc. Sonicator Model W-225 equipped with a microtip. Debris was removed from the lysates by centrifugation in a microcentrifuge at 4°C prior to performing B-galactosidase assays. 0.5 ml aliquots of each cleared Iysate were assayed for B-galactosidase activity essentially as described by Miller (27). Aliquots were also assayed for protein concentration by the dye binding method of Bradford (6). B-galactosidase activity was calculated using the following formula: ([213] X [A420] / ([protein concentration] X [volume] X [time]), where 213 is the extinction coefficient for 2-Nitrophenyl-B-D-galactopyranoside (ONPG), the protein concentration is in mg/ml, the volume assayed is in ml, and the time of incubation is in minutes. Units of B—galactosidase activity are then expressed as nmol O-nitrophenyl (ONP)/min—mg protein. Quantitative immunoblot analysis of mutant DnaA proteins. Whole cell lysates and quantitative immunoblot analysis were performed as described previously (41), except that cultures were grown in LB medium at 30°C to late log phase and M83898 was used as the host strain in place of M83897 (see Table 1). S1 nuclease protection assays. S1 nuclease protection assays were performed as described previously (43). As a probe. the 1.4 kb Pvu lI-Eco RI fragment from pACYCdnaA was used. This DNA fragment contains pACYC184 sequence (coordinates 515 to 1,517 (see reference 9)). the complete dnaA promoter region, and the first 61 nucleotides of the dnaA coding region. Preparation of this fragment was as follows: The 3.6 kb Eco R1 133 fragment from pACYCdnaA was dephosphorylated with calf intestinal alkaline phosphatase (Boehringer Mannheim), 5’ end-labelled with [732P]ATP (Dupont, NEN) and T4 polynucleotide kinase (New England BioLabs). and digested with Pvu II. The 1.4 kb fragment was used without further purification. Determination of relative pCM128 abundance. The abundance of pCM128 relative to the respective dnaA-containing pACYC184 derivatives was determined using a modification of the method described by Biek and Cohen (5). Briefly, double transfonnants of M83898 harboring both the pACYC184 dnaA-containing derivative (as noted in Table 9) and pCM128 were grown at 30°C in LB medium containing kanamycin, chloramphenicol, and ampicillin. These were subcultured into supplemented M9 medium containing kanamycin and chloramphenicol, but lacking ampicillin which is required for selection of pCM128, followed by growth at 30° or 42°C for 4 h. Plasmid DNAs were prepared by the alkaline lysis miniprep method (34) and the Ava I (New England IioLabs) linearized plasmid DNAs were electrophoresed in a 0.7% agarose gel. Both DNAs contain a single Ava I site and differ in size by approximately 2 kb. The relative abundance of each DNA was determined by scanning densitometry relative to a standard curve using a Bio-Rad Gel Doc 1000 equipped with the Molecular Analyst software package. Results Identification of nonsense and deletion mutations of the Escherichia coli dnaA gene. Using a selection method for mutations of the E. coli dnaA gene. we identified nonsense and deletion mutations. These were originally identified as encoding smaller polypeptides by immunoblot analysis of whole cell Iysates of a dnaA null strain harboring the plasmid-encoded dnaA alleles (data not shown). For the nonsense mutations, electrophoretic analysis of plasmid DNAs revealed these to be similar in both size and structure to the wild type pACYCdnaA parent whereas the deletion mutants appeared slightly smaller (data not shown). In addition. the amber suppressors supE and supF permitted expression of full length polypeptides for the presumed nonsense mutations (data not shown). As many of the nonsense mutations appeared to encode polypeptides judged to be similar in size (data not shown), a representative group was selected. None of the presumed nonsense mutations expressed a detectable level of the full length polypeptide (Figure 1). The region of each mutant allele expected to contain the nonsense mutation based on the estimated size of the truncated polypeptide was sequenced by the chain-terminating enzymatic method as described in Materials and Methods. Nucleotide sequence analysis confirmed the presence of a nonsense mutation near the expected position for each allele sequenced (Table 2). The regions encompassing the deletion of sequence for those mutant plasmids that were smaller in size than the pACYCdnaA parent plasmid were determined by DNA restriction endonuclease analysis (data not shown). This analysis indicated that of the three deletion mutations that expressed a detectable gene product, one appeared similar to a nonsense mutation 134 135 Figure 1. Western blot analysis of truncated forms of DnaA protein. Western blot analysis of whole cell Iysates of a dnaA null strain harboring the respective dnaA alleles under control of the dnaA promoter region was performed as described in Materials and Methods. 136 838-55 @8868 88335 88855 88958 .8558 03.8358 88-8268 +6.5 3 D k 5 W _ 6 5 1 6 4 3 _ _ _ 5 1. 2 _ O- 5" — 14.5 TABLE 2. Summary of nucleotide sequence analysis of nonsense and in-frame dnaA alleles dnaA-containing Position of Nucleotide Deduced plasmid mutation8 alterationb mutation(s)c pACHA-4 709-1 .134 ln-frame deletion A(237-378) pACHA-6 988 CGA-+TGA 330—+opal pACHA-22 442 CAA—>TAA 1 48—>och re pACHA-86 787 CAG —»TAG 263-+amber pACHA-97 739 CAG —>TAG 247—»amber pACHA-1 19 1 .336 CAG—vTAG 446-+amber pACHA-121 1 .081 GAG-*TAG 361 ->amber pACHA-178 733 CGA—>TGA 245-+opal 8 The nucleotide position of each mutation is indicated with position one corresponding to the first nucleotide of the coding sequence (18). 8 Altered nucleotides resulting in the nonsense codon are in bold. 0 Mutations are expressed as the amino acid position in the wild type sequence and the type of nonsense mutation (amber.am; opal, op; ocher, 00) resulting from the base substitution. The allele containing an in-Irame deletion has been named dnaAA237-378. Nonsense alleles have been named with the designation of the nonsense codon and its position in the coding sequence(dnaA-op330). 138 described above (based on polypeptide size and the finding that C-terrninal sequences were deleted; data not shown) and was not pursued further. The remaining two appeared similar to each other (data not shown). The region encompassing the deletion of coding sequence for these two was sequenced as described in Materials. Both contained an identical in-frame deletion of coding sequence corresponding to amino acid residues 237-378 (Table 2). Mutations mapping within or near the P-Ioop motif exhibit a cold-sensitive heteropolyploid phenotype. In an effort to identify mutations of dnaA that exhibited a dominant negative phenotype. all 36 plasmid-encoded alleles (28 missense. seven nonsense. and one deletion) were transformed into a prototrophic strain (W3110). Chloramphenicol resistant transfonnants were selected at both 30° and 42°C. By this analysis, three of the 36 were cold sensitive for transformation (data not shown). None were thermolabile for transformation. Based on nucleotide sequence analysis, all three contained substitutions of highly conserved residues mapping within or near the P-Ioop motif thought to function in nucleotide binding (37). A similar cold-sensitive heteropolyploid phenotype has been described for other dnaA alleles containing the alanine-to-valine substitution at position 184 (18). In light of this, a more careful analysis of the transformation efficiency of all plasmid-bome dnaA mutations containing an amino acid substitution mapping within or near the P-Ioop motif was performed (Table 3). This analysis used two isogenic strains; M01061 and M81062 (see Table 1). The latter strain contains a mini-R1 plasmid inserted into its chromosome that serves as a secondary site for initiation of replication in the absence of dnaA or oriC function (17). In this case, it serves as a control strain inasmuch as it may suppress the othenrvise lethal phenotype of a dominant negative dnaA mutation (data not shown; 23). 139 TABLE 3. Substitutions mapping to the P-Ioop motif exhibit a cold-sensitive heteropolyploid phenotype Relative transformation efficiency“ dnaA- Deduced containing amino acid 24 hours 48 hours p'asm'd SUbsmmmMs) M01061 M81062 M01061 M81062 pACYCdnaA None 1.0 0.9 1.0 0.9 pACYCdnaA46 A184V 2.0 X 103 b 0.7 0 1 .0 0.7 H252Y pACYC184 —— 0.8 1.0 0.8 1.0 pACHA-94 Gt77D 3.0 X 10‘3 b 1.0 d 3.9 X 10'3 b 1.0 E248K E249K pACHA-101 G175D 1.0 1.0 1.0 1.0 pACHA-105 A184T 1.7 X 10'2 1.3 X 10'2 1.7 X 10'2 b 3.9 X 10'1 pACHA-117 P43L 0.9 0.9 0.9 0.9 61778 pACHA-146 G177D 2.8 X 10‘2 b 1.0 d 0.9 1.0 8 Expressed as the transformation efficiency observed at 300 relative to 42°C after 16 h of growth at 42°C and either 24 h or 48 h of growth at 30°C, as noted. Transformation efficiencies at 42°C were on the order of 107-108 per pg of plasmid DNA for each of the eight plasmid DNAs. 8 These values do not include micro colony forming units. 8 Small colony forming units. 8 Tiny colony forming units. 140 The plasmids containing the dnaA+, G175D (with a G-to-D mutation at position 175), or P43L-G177S allele each transformed M01061 with similar efficiencies at 30° relative to 42°C. However. the transformation efficiency of the plasmids containing the dnaA46, G177D, G1770-E248K-E249K. or A184T allele was reduced at 30°C relative to both that observed for the dnaA+ allele at the same temperature, and to their respective transformation efficiencies at the elevated temperature. At 42°C. all transformation efficiencies were similar to that obtained with pACYC184, indicating that none of the plasmid-bome alleles exhibited an inhibitory phenotype at the elevated temperature. We do not understand why the G17SD and P43L-G177S alleles did not result in a cold sensitive phenotype in this analysis. The cold sensitive transformation phenotype of the plasmids containing the dnaA46, G177D. or G1770-E248K-E249K allele was suppressed by chromosomal insertion of a mini-R1 replicon. This was demonstrated by the ability of the respective plasmids to transform the R1 integratively suppressed strain (M81062) at both 30° and 42°C with comparable efficiency. By contrast, the cold sensitive transformation phenotype of the A184T allele was only partially suppressed as indicated by its reduced transformation efficiency at 30° relative to 42°C. Further investigation of the phenotypes of the plasmids containing the G1770, G177D-E248K-E249K. or A184T allele using MC1061 harboring pBSoriC (3), a pBluescript derivative containing E. coli oriC sequence. indicated that their cold-sensitive heteropolyploid phenotype was also suppressed by the presence of a multicopy oriC plasmid (data not shown). Consistent with the results discussed above regarding the R1 integratively suppressed strain (M81062), the A184T allele was least efficiently suppressed by the presence of pBSoriC while G177D was most efficiently suppressed. This 141 was judged both by growth rate and colony size of the double transformant (data not shown). Together, these observations suggest that the cold-sensitive heteropolyploid phenotype exhibited by these mutations mapping within or near the P-loop motif is oriC-specific. The C-terminus of DnaA is required for repression of dnaA gene expression. DnaA protein has been shown to regulate its own expression (7, 25). Autorepression is a function of the binding of DnaA protein to a DnaA box located between the two dnaA promoters (43). Although the mechanism of dnaA autoregulation is not yet understood at the molecular level. the DNA binding activity of DnaA protein is a major component. With this in mind, we developed an in vivo assay to measure the DNA binding activity of DnaA protein (Figure 2). Using a specialized E. coli strain (M83900; see Table 1), we were able to measure the abilities of the plasmid-encoded dnaA alleles to repress expression of the dnaA’-’IacZ fusion by quantitation of B-galactosidase activity. Repression was attributed in part to binding of DnaA protein to the DnaA box within the dnaA promoter region (43). M83900 was transformed to chloramphenicol resistance with all 36 plasmid-encoded mutant dnaA alleles (28 missense. one deletion and seven nonsense). Representative isolates were cultured in supplemented M9 medium and B-galactosidase assays were performed as described in Materials and Methods. As the cell size and/or extent of filamentation for the different classes of dnaA mutations varied (data not shown), B-galactosidase activity was normalized to total soluble protein. In this assay. missense mutants mapping to the N-terminus or P-Ioop motif had only a marginal effect on B-galactosidase activity relative to the controls (vector only and it containing the dnaA+ or the thermolabile dnaA46 allele). By contrast, many of the mutants mapping to the C-terrninus were 142 Figure 2. in viva DNA binding assay. The DNA binding activities of the plasmid-bome dnaA alleles was measured indirectly using the dnaA promoter-‘Iacz fusion contained in ARB1 (7). Mutants of DnaA able to bind to the DnaA box in the dnaA promoter region repress expression of the dnaA'-’IacZ fusion whereas those unable to bind to the DnaA box have no effect. Expression of the dnaA’-’IacZ fusion was measured by quantitation of total B-galactosidase activity as described in Materials and Methods. 143 2.52: 03> 666.: V rmox o_. o. 6983 69685: 0.. new) ...amoN 86635: rm: 63. Q59». .amoN amoex :63 ma: 069820: 2 ene.»...LmQN . ox _. m. : >o= 206.: u mm o 144 unable to repress the dnaA’-’IacZ fusion (Table 4). In addition, none of the dnaA nonsense mutations analyzed were able to represses expression of the dnaA’-’IacZ fusion in comparison to pACY0184 (Table 5). Together. these results indicate that the C-terminus of DnaA protein is required for autorepression of dnaA expression and suggest that it is required for DNA binding. The inability of the mutation containing an in-frame deletion of residues 237-378 to repress expression of the dnaA’-’IacZ fusion suggests that it is either defective in DNA binding. or that the region absent is in some way required for autoregulation of dnaA gene expression. Experiments similar to those described above were performed using a plasmid (pD8596) that contained the dnaA gene under control of the araB promoter (data not shown). Based on quantitative immunoblot analysis, in the absence of araB promoter induction, about 2-3% as much DnaA+ protein was observed relative to when the dnaA gene was expressed from its own promoter (pACYCdnaA; see Chapter II). Under these conditions. only subtle repression of B-galactosidase activity was observed, suggesting that the level of expression from araB in the noninduoed state represented the near minimal level of DnaA+ protein required to observe partial repression of dnaA’-’IacZ expression. Therefore, a likely explanation for the lack of repression observed for the MN mutation in this assay relates to its low level of abundance (see below and Table 6). We do not yet know why 09Y, P43L-G1778, and G177D were less efficient for repression of expression of the dnaA’-’IacZ fusion, although it most likely does not relate to their respective levels of abundance as they were present at levels similar to other proteins analyzed (see below and Table 6). Quantitative immunoblot analysis of mutant DnaA proteins. To corroborate the B-galactosidase assays, quantitative immunoblots of whole cell 145 TABLE 4. Mutations mapping to the C-terrninus of DnaA are defective for regulation of expression of the dnaA’-’IacZ translational fusion in vivo8 Relative units M83900 Deduced amino acid I5‘93"1'5“?“951‘185" transformantb substitution(s)c actrvrtyd 30°C 42°C pACYCdnaA None 5 50 pACYCdnaA46 A184V-H252Y 23 46 pACYC184 — E100 E100 pACHA-9 $23N-A326T-S466L 7 -16 pACHA-32 l26K-H452Y 1 -16 pACHA-39 C9Y-T64I 36 52 pACHA-40 MW 79 68 pACHA-45 C9Y 5 3 70 pACHA-50 W253 31 42 pACHA-72 E21 K 7 31 pACHA-1 06 C9Y-A293V 4 2 71 pACHA-143 L10F-R52W 20 59 pACHA-152 P28L 1 7 32 pACHA-200 E21K-Qt44N 12 42 pACHA-94 G177D-EZ48K-E249K 39 24 pACHA-101 G17SD 25 33 pACHA-105 A184T 9 32 pACHA-117 P43L-G177S 59 73 pACHA-146 G1770 49 40 pACHA-43 A440T 60 60 pACHA-49 T435M 107 106 pACHA-54a R407H 1 1 2 106 pACHA-70 V437M 31 39 pACHA-79° G42GD 1 1 2 104 pACHA-83 A440V 97 37 pACHA-110 V383M 51 91 pACHA-1 30 A428T-A440T 86 1 06 146 TABLE 4 (continued). Relative units M83900 Deduced amino acid B-galactosidase transfonnantb substitution(s)c actrvrtyd 30°C 42°C pACHA-162 D209N-T435M 1 1 3 1 13 pACHA-1709 T435M-T436M 1 O3 1 04 pACM-2° A412? 93 122 pACM-79 412(oM-A)413' 104 124 8 B-galactosidase assays were performed as described in Materials and Methods. 8 M83900 contains a dnaA null mutation, a lac deletion, and itRBt as a prophage. ARBt contains a translational fusion of the dnaA promoter and IacZ structural gene (7). See Table 1 for the complete genotype of M83900. 0 Deduced amino acid substitutions are based on the nucleotide sequence of these mutations that were presented previously (see Chapter III). The deduced amino acid substitutions in dnaA46(Ts) are from Hansen efal. (18). 0' B-galactosidase activity was determined as described in Materials and Methods and is presented as the percentage of the activity observed relative to the pACYCdnaA and pACYCf 84 controls with 100% being equal to that for no DnaA present (pACY0184 transformant) and 0% being equal to that for the level of DnaA present when the dnaA+ allele was plasmid-encoded (pACYCdnaA transformant). To normalize the data, units observed with pACYCdnaA were subtracted from all samples and the resulting values were divided by the units observed with pACY0184 (minus the units observed with pACYCdnaA). Actual units of B-galactosidase activity (nmol ONP/min-mg protein) observed for the controls at 30° and 42°C. respectively, were as follows: pACYCdnaA, 82 and 61; pACYCdnaA46, 144 and 89; pACYC184. 350 and 122. 8 These alleles are inactive for replication from the p80101 origin (see Chapter III) and oriC (data not shown). ' 412(QM-A)413 is 3 Met. Ala insertion between residues 412 and 413. 147 TABLE 5. Truncated dnaA mutations are unable to affect expression of a dnaA’-'IacZ translational fusion in vivo8 Relative units M83900 Deduced B-galactosidase transformant mutation(s)b activityc 30°C 42°C pACYC dnaA None so 20 pACYC dnaA46 A184V-H252Y 24 47 pACY0184 —— 2100 2100 pACHA-4 A(237-378) 98 93 pACHA-6 330-»amber 121 104 pACHA-22 148—>ocher 84 128 pACHA-86 2638amber 120 1 12 pACHA-97 247—»amber 1 17 1 15 pACHA-1 19 446—>amber 124 1 17 pACHA-121 361—amber 94 95 pACHA-178 245—»amber 1 15 1 18 8 See the footnotes to Table 4. 8 Based on the nucleotide sequence information presented in Table 2. 6 Relative B-galactosidase activity is expressed as described in the footnote to Table 4. Actual units of B-galactosidase activity (nmol ONP/min-mg protein) for the controls at 30° and 42°C, respectively. were as follows: pACYCdnaA. 53 and 56; pACYCdnaA46. 82 and 93; pACY0184,177 and 134. 148 lysates of a dnaA null strain harboring the plasmid-bome dnaA alleles was performed. Assuming the various polypeptides were similarly resistant to proteolysis, those defective in DNA binding would be expected to be present at levels higher than those unaffected as they would be unable to autoregulate their own expression. With the exception of the G1770 and G177D-E248K-E249K mutants. only those mapping C-temIinal to residue 406 were detected at levels elevated relative to the plasmid-encoded DnaA+ and DnaA46 proteins (Table 6). Previously, we reported that DnaA46 protein was present at an abundance approximately 1.4-fold elevated relative to the level observed for the wild type protein when both alleles were plasmid-encoded (see Chapter IV). Although this analysis was done using the same plasmids and an isogenic strain, the variation in results is likely due to the different growth conditions used (see Materials and Methods). Some of the mutations mapping to the C-terrninus were unable to repress expression of the dnaA’-’IacZ fusion but the mutant proteins were present at levels lower than that observed for DnaA+ or DnaA46. Presumably. this is due to proteolytic instability of the mutant proteins. Consistent with this interpretation, these same mutants were proteolytically unstable following their overexpression from an inducible promoter (data not shown; see Chapter V). In general. the results of the quantitative immunoblot analysis are consistent with those of the B-galactosidase assay. Together. they indicate that the C-tenninus of DnaA is required for repression of dnaA gene expression and suggest that it is required for DNA binding. Identification of a third promoter directing expression of the dnaA gene in pACYC184. In an effort to correlate the results of the B-galactosidase assay with transcription from the two dnaA promoters, S1 149 TABLE 6. Quantitative immunoblot analysis of mutant DnaA proteins M83898 Deduced amino acid Relative protein transformant substitution(s) level8 pACYCdnaA None 21 .00 pACYC dnaA46 A1 84V-H252Y 0.97 pACYCt 84 — 50.00 pACHA-9 823N-A326T-S466L 0.81 pACHA-32 l26K-H452Y 0. 22 pACHA-39 09Y-T64l 0.1 9 pACHA-40 M1 V 0.04 pACHA-45 09Y 0. 24 pACHA-50 W258 0.30 pACHA-72 E21 K 0.24 pACHA-106 C9Y-A293V 0.24 pACHA-143 L10F-R52W 0.57 pACHA-152 P28L 0.45 pACHA-200 E21 K-G144N 0.20 pACHA-94 G177D-E248K-E249K 1 .99 pACHA-101 01750 0.69 pACHA-105 A184T 0.33 pACHA-1 17 P43L-G1778 0.51 pACHA-146 G177D 1.60 150 TABLE 6 (continued). pACHA-43 A440T 1 .44 pACHA-49 T435M 3.34 pACHA-54 R407H 3.50 pACHA-70 V437M 0.70 pACHA-79 G426D 1 .97 pACHA-83 A440V 1 .62 pACHA-110 V383M 0.21 pACHA-130 A428T-A440T 0.64 pACHA-162 D209N-T435M 2.54 pACHA-170 T435M-T436M 2.31 pACM-2 A412P 0.30 pACM-7 412(oM-A)413 2.59 8 Relative protein levels were determined by quantitative immunoblot analysis using an 12til-labelled antibody as described in Materials and Methods. lmmunoblots were quantitated with a Packard Instant Imager. No DnaA was detected in the pACYC184 transformant, confirming its genotype. The counts detected in this sample were therefore defined as background (normalized to 0.00). Protein levels are corrected for cell density (based on ODsgs)and are expressed relative to that observed for the wild type transformant (pACYCdnaA that was normalized to 1.00) minus background. 151 nuclease protection assays were performed (Figure 3). This analysis revealed that the plasmid-encoded dnaA alleles were expressed from not only dnaAP1 and dnaAP2, but a third promoter as well. We have designated this third promoter dnaAPi’. S1 nuclease mapping of this transcript positioned its 5’ and near the start site for the tetracycline resistance gene. Restriction at the Clal site located between the -35 and -10 regions of the promoter for the tetracycline resistance gene in pACY0184 was expected to inactivate transcription from it. Activity from this promoter is presumably due to its reconstruction by the insertion of the dnaA gene at the Cla I site. A similarity in nucleotide sequence near the -10 region of the tet promoter region to that of the hybrid promoter formed is observed (see Figure 3 legend). Quantitative immunoblot analysis indicated a two-fold higher level of DnaA protein encoded by pACY0184 relative to the chromosomally encoded allele (41). This elevated level of expression may be due to a copy number effect or due to elevated transcription by virtue of the third promoter. The N-terminus of DnaA protein is essential for replication activity. Although mutations mapping to the N-terminus of DnaA protein neither exhibit a cold-sensitive heteropolyploid phenotype nor were defective for autoregulation, their thermolabile phenotype for pSC101 replication indicates the importance of this region of the protein. Biochemical characterization of select mutants from this class will help to elucidate the function(s) of this domain in replication. DnaA protein contains two domains that are alternately dispensable for replication from the p80101 origin. The replication phenotypes of the nonsense and in-frame deletion mutations were measured using the phasemid assay (see Materials and Methods). Two of the eight alleles retained the ability to replicate the lambda derivative from the inserted 152 Figure 3. Identification of dnaAP1 ' using an S1 nuclease protection assay. (A) Structure of the 1.4 kb DNA fragment used for 81 nuclease protection assays. Assays were performed as described in Materials and Methods. Approximate positions and sizes of fragments identified by S1 nuclease protection shown in part B below are indicated. (B) Total RNA was isolated from either the dnaA+ strain, M83907. (lane 1) or the dnaA null strain, M83898. harboring pACYC184 (lane 2) or pACYCdnaA (lane 3). DNA size standards are indicated (lane 4), as are the positions of protected fragments corresponding to dnaAP1 (~300 bp). dnaAP2 (~215 bp). and the hybrid promoter, dnaAP1’ (~400 bp). The -10 region of the fat promoter and the upstream region of the dnaA gene are both are AT rich: dnaA CGATTAAGCCAATTfIEIffIfGTCTAT tet TGIEIQAQAGCTTATCATQQATAAGCTTTAATGCGGTAGTT ~35 Cla I ~10 The -35 and -10 positions in the tat promoter are indicated. as are the Cla I site and putative -10 region of the hybrid promoter. Conserved bases are indicated by the blackened dots. 153 Pvu 11 ClaI Eco RI ; / I I l l - > b l j/ (e! -35 region dnaAP1 dnaAP2 e (K > J) 1.4 kb <—> 215 bp A 300 bp _ ‘ 400 bp 7 Lane — 507 bp dnaAP1’— — 399 — 347 dnaAP1 — — 301 — 223 dnaAP2 — 154 pSC101 origin; both were thermolabile (Table 7). One was the in-frame deletion of residues 237-378 (dnaAA237-378) and the other an amber mutation at position 361 (dnaA-am361). Their abilities to maintain pSC101 were investigated further by use of a transformation assay. dnaA-op245 was included as a control, as were pACYC184 and it containing the dnaA+ or temperature sensitive dnaA46 alleles. Isolates of a dnaA null strain harboring the plasmid-encoded dnaA alleles were transformed with the wild type pSC101 plasmid. the copy-number-up pSC101 mutant. pCM128, or pBR322. The transformation efficiency at 30°C for each plasmid was determined (Table 8). Consistent with the results of the phasemid assay. the dnaA null strain bearing dnaAA237-378 or dnaA-am361 was transformed with pSC101 or pCM128 at 30°C as efficiently as were the dnaA+ and dnaA46(Ts) controls. By contrast. we were unable to transform the dnaA-op245 or pACYC184 controls with either pSC101 or pCM128, but were with pBR322. These observations suggest that the regions defined by the in-frame deletion and the amber mutation at position 361 represent domains of DnaA that are alternately dispensable for p80101 replication. The abilities of the various alleles to maintain pSC101 or pCM128 at 42°C relative to 30°C was measured by plating representative isolates of a dnaA null strain harboring the dnaA-containing plasmid and either p80101 or pCM128 (Table 8). Whereas the plasmid-encoded dnaA+ allele was able to maintain p80101 at both 30° and 42°C with similar efficiency, the dnaA46, dnaAA237-378, and dnaA-am361 alleles were temperature sensitive. The dnaA+. dnaA46 and dnaAA237-378 alleles were temperature insensitive for pCM128 maintenance while the dnaA-am361 allele remained thermolabile (Table 8), indicating that the repA7 mutation contained in pCM128 is able to suppress certain dnaA alleles. The inability to transform dnaA null aIl Illl III "III 155 TABLE 7. dnaAA237-378 and dnaA-am361 are active for replication of a lambda derivative from the inserted pSC101 origin - b dnaA- kpCM128 spottIng 3:11:12). .... 32:21:12: “am", 104 103 102 104 103 102 pACYCdnaA None + + + + + + Wild type pACYCdnaA46 A184V-H252Y + + + +/— - — Ts pACY0184 — - - - - - - Inactive pACHA-4 A(237-378) + + - - — - Ts pACHA-6 330—»opal - — - - - - Inactive pACHA-22 1 48—rochre - - — — - — Inactive pACHA-86 263-2amber — - - - - - Inactive pACHA-97 247—ramber - — - — — - Inactive pACHA-1 19 446—ramber — — - — — - Inactive pACHA-121 361—amber + + - — - - Ts pACHA-178 245—»opal - — - - - - Inactive 8 See Table 2 for nucleotide sequence information. The nucleotide sequence of the dnaA46 allele is from Hansen ef al. (18). 1’ xpCM128 spotting consisted of spotting ~105 M83898 transfonnants of each of the Indicated plasmid-encoded dnaA transfonnants onto supplemented M9 medium in combination with the indicated number of itpCM128 infective particles (102-104) as described in Materials and Methods. kp0M128 spotting was done in duplicate and plates were incubated at 30° and 42°C. Abbreviations are +. bacterial clearing (kp0M128 sensitivity); -. bacterial growth (kpCM128 immunity); +/-. partial clearing (moderate kpCM128 sensitivity). 8 Deduced from kpCM128 spotting. Ts, thermolabile. 156 TABLE 8. Truncated dnaA mutations are able to maintain pSC101 Transformation p80101 pCM128 pSC101 pCM128 pACYCdnaA None 21.0 21.0 1.0 1.0 pACYCdnaA46 A184V-H252Y 1.1 1.4 6.9 X 10'4 1.1 pACYCl 84 —— NTO <0.001 NT NT pACHA-4 A(237-378) 0.8 1.0 1.3 X 10‘3 0.81 pACHA-121 361->amber 1.2 1.3 2.8 X 10'3 2.2 X 10'2 pACHA-178 245—»opal $0.038 $0.038 NT ND8 8 M83898 harboring each of the indicated plasmid-encoded dnaA alleles were transformed with p80101, pCM128. or pBR322, and kanamycin. chloramphenicol. and tetracycline (for p80101 and pBR322) or ampicillin (for pCM128 and pBR322) resistant colonies were selected for at 30°C. The transformation efficiency for p80101 and pCM128 for each dnaA allele was normalized relative to the efficiency observed for pBR322, and is expressed relative to the efficiency observed for the dnaA+ transformant (normalized to 1.0). The transformation efficiency (colony forming units/pg plasmid DNA) of M83898 (pACYCdnaA) for each plasmid DNA was as follows: p80101, 3.6 X 104; pCM128. 9.3 X 104; pBR322. 2.1 X 105. Transformation efficiencies of the other transfonnants with pBR322 was on the order of 105 colony forming units/pg plasmid DNA. 8 Representative isolates of M83898 harboring both the dnaA-containing plasmid and p80101 or pCM128 were grown under selection for both plasmids in LB medium at 30°C, then diluted serially in 1X M9 salts, plated onto supplemented M9 minimal plates containing the appropriate antibiotics, and incubated at 30° and 42°C. Efficiency of plating is expressed as the number of colony forming units observed at 42° relative to the number observed at 30°C. 8 NT. not tested. In the case of the transformation efficiency of the pACY0184 transformant for pSC101. this was not tested because both plasmids confer tetracycline resistance. 8 Double transformants of M83898 containing pACHA-178 and p80101 or pCM128 were tiny. Those containing pSC101 would not grow in liquid culture containing kanamycin. chloramphenicol. and tetracycline. Although those containing pCM128 would grow in liquid culture containing kanamycin. chloramphenicol. and ampicillin. no pCM128 was detected by miniprep analysis (see footnote 3). Reinvestigation of the ability of the R245-ropal mutation to support replication of p80101 or pCM128 using pACHA-213 (not described in this thesis) that contained both an R245—>opal and GZ47—>amber mutation was performed (data not shown). We were unable to transform M83898 (pACHA-213) with either p80101 (transformation efficiency 550 colony forming units/pg DNA) or pCM128 (transformation efficiency _<_50 colony forming units/pg DNA). thus suggesting that the ability of the pACHA-178 mutation to maintain p80101 and pCM128 was due to read through of the opal mutation. No full length polypeptide was detected for the pACHA-178 mutation. This is not inconsistent with the above conclusion as the level of DnaA protein necessary to maintain such a low level of pCM128 may be undetectable by immunoblot analysis (data and theoretical calculation not shown). 8 Efficiency of plating for M83898 (pACHA-178)(pCM128) was not determined (ND) because following growth in liquid culture supplemented with kanamycin. chloramphenicol, and with or without ampicillin, less than 1 in 1,000 colony forming units were resistant to ampicillin. 157 strain harboring pACYC184 with pCM128 is consistent with its dnaA-dependence (42). The differences observed between the replication phenotypes of the mutant alleles in the phasemid assay relative to the transformation assay are not unexpected (compare Tables 7 and 8). Although both measure replication from the pCM128 origin. activity observed in the phasemid assay appears to be proportional to the copy number of the phasemid (8). Titration of the prophage-encoded cl repressor is presumed to be more efficient when the phasemid copy number is high. Therefore, we assume the phasemid assay measures an elevated copy number of the ApCM128 phasemid. However, the transformation assay has few constraints on copy number. A single copy of the bla gene is sufficient for resistance to ampicillin (27). Determination of the relative abundance of pCM128 when maintained by different dnaA alleles. The dnaA. dnaA46. dnaAA237- 378. and dnaA-am361 alleles were able to maintain pCM128 for at least 10 generations at 30°C in the absence of its selection (data not shown). However. the plating efficiencies of these alleles for pSC101 and pCM128 maintenance suggested that dnaA-am361 was less efficient for pSC101 replication than was dnaAA237-378 (Table 8). To investigate this possibility. we determined the relative abundance of pCM128 in strains harboring each of these plasmid-encoded dnaA alleles (Table 9). This analysis indicated that maintenance of pCM128 at 30°C by the dnaAA237-378 or dnaA-am361 allele was less efficient than the wild type allele. Furthermore, the level of pCM128 when maintained by dnaAA237-378 appeared 40% elevated relative to dnaA-am361. These observations are consistent with the dnaA-am361 allele being less efficient than dnaAA237-378 for pSC101 replication. The dnaA46 allele appeared to be less efficient than the wild type allele 158 for p80101 replication as well. This was based on the reduced abundance of pCM128 when maintained by dnaA46 relative to the wild type allele. This finding is consistent with previous reports indicating that dnaA(Ts) mutants maintained pSC101 at a reduced copy number relative to an isogenic dnaA+ host (13). Both the dnaA46 and dnaAA237-378 alleles showed a modest reduction in abundance of pCM128 at 42° relative to 30°C (see Table 9), consistent with the finding that these alleles were able to maintain pCM128 at both 30° and 42°C with similar efficiency (Table 8). By contrast, the dnaA-am361 allele showed a significant reduction in pCM128 abundance at 42° relative to 30°C. This is consistent with the thermolabile phenotype of this allele for pCM128 maintenance (see Table 8). Truncated dnaA mutations are Inactive for replication from oriC. As dnaAA237-378 and dnaA-am361 were active for pSC101 replication, we wanted to test their abilities to support replication from oriC. The ability of these plasmid-encoded dnaA alleles to initiate replication from oriC was measured after a temperature downshift from 42° to 30°C of a dnaA null strain harboring both the plasmid-encoded dnaA alleles and the copA(Ts)-containing plasmid. pMDS421. pMDS421 bears an R1 copA(Ts) allele that encodes an antisense RNA which inhibits translation of the R1 repA mRNA required for R1 replication. The R1 copA(Ts) allele is active at 30° but inactive at 42°C. In contrast to the results obtained with p80101 and pCM128. neither the dnaAAZ37-378 nor dnaA-am361 alleles were active for replication from oriC (Table 10). As controls, the plasmid-encoded dnaA+ or dnaA46 alleles were able to complement the chromosomal dnaA null allele at 30°C for replication from oriC, pACYC184 was not. 159 TABLE 9. Relative abundance of pCM128 in various dnaA mutants Relative abundance of dnaA allele8 pCM1ZBb 30°C 42°C dnaA+ 51.00 51.00 dnaA46 0.66 0.43 dnaAA237-378 0.36 0.33 dnaA-am361 0.26 0.08 8 dnaA alleles were plasmid-encoded and harbored by the dnaA null strain M83898. 8 The abundance of pCM128 was determined relative to the indicated dnaA-containing plasmid following growth at 30°C to log phase maintaining selection for pCM128, then for 4 h in the absence of selection for pCM128 in supplemented M9 medium at 30° and 42°C as described in Materials and Methods. In each case, values are normalized relative to the dnaA+ control (pACYCdnaA that was normalized to 1.00) at that temperature. 160 TABLE 10. Truncated dnaA mutations are inactive for replication from oriC "223933.324 dnaAaHe'e pfigfifé‘iiriLZISmifiisa pACYCdnaA dnaA+ 0.54 pACYCdnaA46 dnaA46 0.12 pACY0184 None 5.4 X 10’6 pACHA-4 dnaAA237-378 5.5 X 10‘6 pACHA-121 dnaA361-am 3.6 X 10'4 8 M83898 containing the indicated plasmid-encoded dnaA alleles (or pACYC184) were transformed with pMDS421 that encodes an R1 copA(Ts) allele. Colonies were picked from plates immediately following their transformation. resuspended in 1X M9 salts. diluted serially, and plated onto supplemented M9 medium containing kanamycin. chloramphenicol. and ampicillin. Efficiency of plating is expressed as the number of colony forming units observed at 30°C relative to 42°C. Discussion The method described in Chapter III provided a large collection of missense and nonsense mutations that were defective in replication of a lambda derivative containing an inserted pSC101 origin. DNA sequence analysis indicated that the mutations were distributed throughout the coding region of the dnaA gene with groupings near the N-terminus, the C-tenninus. and the P-loop motif. These mutations were further characterized to correlate a specific functional defect with each individual mutation (Figure 4). Mutations mapping to the well conserved P-Ioop motif, proposed to function in nucleotide binding, exhibited a cold sensitive phenotype when plasmid-encoded and harbored by a strain dependent upon the chromosomal dnaA+ allele and oriC for growth (Table 3). DnaA5 and DnaA46 exhibit a similar cold sensitive phenotype (18) and are known to be defective for ATP binding in vitro (21, 22). Together, these observations suggest that the cold sensitive phenotype is due to an ATP-binding defect. Biochemical characterization of the novel mutants described here will reveal whether they are affected in ATP binding. The three alleles described here that exhibit the cold-sensitive heteropolyploid phenotype were active for pSC101 maintenance at 30°. but less so at 42°C. Of the three, A184T appeared most active at the elevated temperature in the phasemid assay (see Chapter I"). By contrast, the A184T mutant was less abundant than either DnaA+, G177D, or G1770-E248K-E249K as measured by quantitative immunoblot analysis (Table 6). Despite this, it was least tolerated by MC1061 and M81062 in the transformation assay. Together, these observations suggest that there is a correlation between the level of activity and/or protein abundance and the cold-sensitive heteropolyploid 161 162 Figure 4. DnaA protein has four functionally distinct domalns. (A) The approximate location of each of the deduced substitutions contained in the 28 missense mutations are indicated by a blackened triangle (A). Numbers refer to amino acid residues of DnaA protein. (8) Examination of the positions of mutation and their mutant phenotypes suggested the presence of four functionally distinct domains of DnaA protein. Delineations of the separate domains are indicated and are based on the genetic characterization and nucleotide sequence of the mutant alleles. Specific and and start points for Domains l and II. respectively, were refined based on the PHD secondary structure prediction (see Figure 5). The proposed function for Domains II, III and IV are indicated. The function of Domain l is unknown as is the role of Domain III in oriC replication. 163 .. oo woo woo 0036.: H a... I :....I..tx4s.- (.7... 51in}... I . I . i ... ..- .....I I... ... . 4...J.a‘«\ \»l_‘r4.uui In. .5 1, utx§1.m~,h I ...141. .. .. ..I. . .. ...p All. .hhs‘ . I a I In. .... h r I L . b. . A. ...19' L111 III 09:05 a 00365 E 00865 2 § ....... .... ....... 7///////////<\\\\\\\\\\\\\\\\\1% d 4.» .3 EN mum mwu 000 mum emu >....u 0:m>-.u.oc> . . 0.3.8 58826: 02> 059.6 r . 263302 26603320 .2 .0005. cc. $602.6. .2 03.0 162 Figure 4. DnaA protein has four functionally distinct domains. (A) The approximate location of each of the deduced substitutions contained in the 28 missense mutations are indicated by a blackened triangle (A). Numbers refer to amino acid residues of DnaA protein. (8) Examination of the positions of mutation and their mutant phenotypes suggested the presence of four functionally distinct domains of DnaA protein. Delineations of the separate domains are indicated and are based on the genetic characterization and nucleotide sequence of the mutant alleles. Specific and and start points for Domains I and II, respectively, were refined based on the PHD secondary structure prediction (see Figure 5). The proposed function for Domains II, III and IV are indicated. The function of Domain I is unknown as is the role of Domain III in oriC replication. 163 A So m8 moo s8 23 § ....... .... ....... //////////<\\\\\\\\\\\\\\\\\:% my mama mufiwmwflw 02> 235m. _ _ 23:58? Emoozmmzo so. omouod GS ammo—.3. 3.. 03.0 164 phenotype. The P43L-G177S mutant was less abundant that DnaA+ and appeared less active in the phasemid assay than either G177D or G177D-E248K-EZ49K. These observations may account for the lack of cold sensitivity of this allele when harbored by MC1061. The specific substitution at residue 177 (G177S as opposed to 61770), or the secondary mutation (P43L) may also be responsible for the lack of cold sensitivity. Likewise, the G175D mutant may be expressed at too low a level (see Table 6). Whether an elevated level of P43L-Gl77S or G175D will result in a cold-sensitive phenotype may be addressed by cloning these alleles under control of an inducible promoter and modulating their expression levels accordingly. The C-terminus of DnaA protein was essential for transcriptional repression from the dnaA promoter region (see Tables 4 and 5). Quantitative immunoblot analysis of whole cell lysates of a dnaA null strain harboring these alleles on a plasmid confirmed that their inability to repress expression of the dnaA’-’IacZ fusion was not due to their proteolytic instability or lack of abundance (Table 6). These results, suggesting that the C-terrninus of DnaA is required for DNA binding, substantiate those of Roth and Messer. They found that the C-terminal 94 residues of DnaA were sufficient and required for specific DNA binding activity (32). V383M is identical to dnaA205 that was identified previously (4, 18). Consistent with its thermolabile phenotype, it was able to repress expression of the dnaA’-’IacZ fusion at 30° but not at 42°C. That V383M could repress at 30°C suggests that it is able to bind DNA. Its reduced ability to repress, however, indicates that its DNA binding activity may be affected. This finding contrasts with that of Roth and Messer (32) who stated that they could not detect a DNA binding activity for a fusion protein between B-galactosidase and the 165 C-terminal 94 residues of DnaA205. A similar fusion protein consisting of the C-terminal 94 residues of wild type DnaA protein was active for DNA binding. This discrepancy may be due to either a folding deficiency of the fusion protein, or inactivation on purification. Consistent with the latter proposal, we have noticed that DnaA205 is proteolytically unstable when overexpressed from an inducible promoter (data not shown, see Chapter V). The viability of a dnaA205(Ts) mutant at 30°C argues strongly that DnaA205 is able to bind DNA at its permissive temperature, as our results suggest. None of the mutations mapping to the N-terminus exhibited a phenotype in the assays described. However, their temperature sensitive phenotype for pSC101 replication (see Chapter III) indicates the importance of the N-terminus in replication. Two truncated forms of DnaA were identified that retained replication activity for pSC101 (Tables 7 and 8) but not for on’C (Table 10). One was an in-frame deletion of residues 237-378 and the other an amber mutation at position 361. All other nonsense mutations, including an opal mutation that codes for a stop codon at position 245, were inactive for pSC101 replication. These observations suggest that the domains defined by the in-frame deletion and amber mutation at position 361 identify regions of DnaA protein that are alternately dispensable for pSC101 replication. Furthermore, it suggests that these two domains periorrn redundant functions. However, that DnaAA237-378 retained DNA binding activity whereas DnaA-am361 did not (data not shown, see Chapter V) indicates that these domains perform dissimilar functions. That dnaA-am361 could maintain pSC101 is consistent with a physical interaction between DnaA protein and the pSC101-encoded initiator, RepA. Such an interaction was proposed based on the ability of RepA protein to facilitate the binding of DnaA protein to the p80101 origin (40). This proposed interaction 166 may recruit DnaA protein to the pSC101 origin in the absence of a DNA binding activity. Examination of the phenotypes of dnaAA237-378 and dnaA-am361 suggests that residues 237 through 360-to-378 function in a DnaA-RepA interaction and that residues 360-to-378 through 467 constitute the DNA binding domain (see Figure 4). Although dnaA-am361 was active for pSC101 replication, the C-terminus of DnaA is not entirely dispensable for replication of pSClOl. Five missense mutations mapping in between residues 361-467 were found to be inactive for pSC101 replication (see Chapter III). We propose that these mutants were inactive for pSC101 replication because they were unable to bind specifically to DNA (see Chapter V). Their nonspecific DNA binding activity may be sufficient to sequester them away from the pSC101 origin. Furthermore, we propose that since DnaA-am361 lacks the DNA binding domain, it would be more readily recruited to the pSC101 origin by RepA protein. Although the missense mutations were widely distributed throughout dnaA, those that mapped within residues 237-378 contained one or more additional mutations that may account for the observed mutant phenotype, consistent with this region of DnaA being dispensable for replication of p80101. In further support of this, two mutations of dnaA (62878 and T301 I) thermolabile for p80101 replication that were identified using an alternative approach (see Chapter II) were indistinguishable from the wild type allele in the phasemid assay (data not shown). To complement the results of the genetic and biochemical characterization of the mutant dnaA alleles, a secondary structure prediction of DnaA protein using the PHD method was obtained (31) (Figure 5). Essentially all of the missense mutations were found to map within either an a—helix or B-structure. The most striking result of the secondary structure prediction, 167 Figure 5. Secondary structure prediction for DnaA protein. The deduced amino acid sequence of the E. coli DnaA protein is indicated relative to the consensus sequence obtained by comparison of the deduced amino acid sequences of fifteen (E. coli, S. typhimurium, S. marcescens, P. mirabilus, B. aphidicola, P. putida, B. subtilis, S. coelicolcr, M. Iuteus, C. crescentus, Ff. meliloti, Synechocystis sp., 8. burgdorferi, S. citn', and M. capricolum) different dnaA homologs (27a). For the consensus sequence, a dot (-) represents a residue conserved (either identical residue or conservative replacement) in less than 9, a letter represents a residue conserved in 9 or more, and an underlined letter represents a residue conserved in 12 or more of the 15 homologs. The position of each mutation identified in the 28 dnaA alleles and the deduced substitution are also presented. PHD refers to the PHD secondary structure prediction for the E. coli DnaA protein. Symbols are: -, an unstructured residue, such as one contained in a loop; H, a helical structure; E, [3 structure. 168 l 60 E. coli MSLSLWQQCLARLQDELPATEFSMWIRPLQAELSDNTLALYAPNRFVLDWVRDKYLNNIN consensus MSL.LH.Q.LA.L..EL....E..flIR.LQ.EL...TL.L.APN.FVLDWM..KYL..l. mutations V YF K N SK L L W PHD ---HHHHHHHHHHHHH—--HHHHHHHHHHHH ------ EEEE-—-HHHHHHHHHHHHHHHH 61 120 E. coli GLLTSFCGRIAPQLRFEVGTKPVTQTPQAAVTSNVAAPAQVAQTQPQRAAPSTRSGWDNV consensus .LL..F ....... L.£.y ...................................... W... mutations I PHD HHHHHH ------ EEEEEE ------------------------------------------ 121 180 E. coli PAPAEPTYRSNVNVKHTFDNFVEGKSNQLARAAARQVADNPGGAYNPLFLYGGTGLGKTH consensus ......... S.yN.K.TEDNEyE§.SH.LA.AAAR.yADNPG.AXH£L£LX§G.QLQKIE mutations N D D S PHD -------------------- EE----HHHHHHHHHHHHHH ------- EEEE----HHHHH 181 240 E. coli LLHAVGNGIMARKPNAKVVYMHSERFVQDMVKALQNNAIEEFKRYYRSVDALLIDDIQFF consensus LLHAMGN..M...PNAKEVXM.SEREV.DM¥.ALQ.N.IEEEK.YXBSVD.LLLDQLQEE mutations T N PHD HHHHHHHHHHHH ----- EEEEEHHHHHHHHHHHHHHHHHHHHHHHH----EEEEE-HHHH 241 300 E. coli ANKERSQEEFFHTFNALLEGNQQIILTSDRYPKEINGVEDRLKSRFGWGLTVAIEPPELE consensus A.K§..QEEEIHTPNALLE...QllLTSQBYBKEl.G15D3LK§BE.W§L.VA1EPEELE mutations KK V PHD H ----- HHHHHHHHHHHHH--—EEEEE ------- HHHHHHHHHHHHH---EEE—---HHH 301 360 E. coli TRVAILMKKADENDIRLPGEVAFFIAKRLRSNVRELEGALNRVIANANFTGRAITIDFVR consensus IBXALL.KKADE..I.LP.EV.FEIA.RL.SNEBELBQALNB!IA.A.E....ITLDFM. mutations T PHD HHHHHHHHHHHH ------ HHHHHHHHHHHHHHHHHHHHHHHHHHHHH ------ EEEHHHH 361 420 E. coli EALRDLLALQEKLVTIDNIQKTVAEYYKIKVADLLSKRRSRSVARPRQMAMALAKELTNH consensus E.LRDLL...E.LyTIDNlQK.MAEYXKIKy.DLLSKRBSRSEARPBQMAM.L.KELINH mutations M H P!) PHD HHHHHHHHH ----- EHHHHHHHHHHHH-EEEEHHH ----- HHHHHHHHHHHHHHHHH--- 421 467 E. coli SLPEIGDAFGGRDHTTVLHACRKIEQLREESHDIKEDFSNLIRTLSS consensus fiLEElQD.EQQBQEIIMLHACRKLE.LR.E..DLK.DE..LIR.L.. mutations D T MMM V Y L T PHD --HHHHHH ----- HHHHHHHHHHHHHHHHH--HHHHHHHHHHHHH-- 169 however, was the presence of a canonical nucleotide binding domain, or Rossmann fold (30), mapping to residues 168-235. This domain includes the P-loop motif thought to function in nucleotide binding. Finally, the genetic characterization of these dnaA alleles suggests that DnaA protein contains four functionally distinct domains: an N-terrninal domain of unknown function, a domain containing the P-loop motif involved in nucleotide binding, a region proposed to interact with pSC101 RepA protein, and a C-terminal domain required for DNA binding (see Figure 4). More detailed biochemical characterization of select mutant proteins, as well as the design and characterization of more refined deletion mutations will help to more precisely define the function of each of the four proposed domains and will help to provide a more detailed understanding of the mechanism of DnaA-dependent DNA replication. 10. 11. 12. 13. 14. References Atlung, T. 1984. Allele-specific suppression of dnaA(Ts) mutations by rpoB mutations in Escherichia coli. Mol. Gen. Genet. 197:125-128. Bachmann, B. J. 1990. Linkage map of Escherichia coli K-12, edition 8 . Microbiol. Rev. 54:130-97. Baker, T. A., K. Sekimizu, B. E. Funnell and A. Kornberg. 1986. Extensive unwinding of the plasmid template during staged enzymatic initiation of DNA replication from the origin of the Escherichia coli chromosome. Cell. 45:53-64. Beyersmann, D., W. Messer and M. Schlicht. 1974. Mutants of Escherichia coli B/r Defective in Deoxyribonucleic Acid Initiation: dnal, a New Gene for Replication. J. Bacteriol. 118:783-789. Biek, D. P. and S. N. Cohen. 1992. Propagation of pSC101 plasmids defective in binding of integration host factor. J. Bacteriol. 174:785-92. Bradford, M. 1976. A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem. 72:248-254. Braun, R. E., K. O'Da and A. Wri ht. 1985. Autoregulation of the DNA replication gene naA in E. coli -12. Cell. 40:159-169. Brenner, 8., G. Cesareni and J. Karn. 1982. Phasmids: hybrids between ColE1 plasmids and E. coli bacteriophage lambda. Gene. 7:27-44. Chang, A. C. Y. and S. N. Cohen. 1978. Construction and Characterization of Amplifyable Multicopy DNA Cloning Vehicles Derived from the P15A Cryptic Miniplasmid. J. Bacteriol. 134:1141-1156. Cohen, S. N. and A. C. Chang. 1977. Revised interpretation of the origin of the pSC101 plasmid. J. Bacteriol. 132:734-737. Davis, R. W., D. Botsteln and J. R. Roth. 1980. A Manual for Genetic Engineering: Advanced Bacterial Genetics. Felton, J. and A. Wright. 1979. Plasmid £80101 replication in iIntegratively suppressed cells requires dna function. Mol. Gen. Genet. 75:231-233. Frey, J., M. Chandler and L. Caro. 1979. The effects of an Escherichia coli dnaAts mutation on the re lication of the plasmids ColE1 pSC101, R100.1 and RTF-TC. Mol. Gen. enet. 174:117-26. Fuller, R. S., B. E. Funnell and A. Kornberg. 1984. The dnaA protein complex with the E. coli chromosomal replication origin (oriC) and other DNA sites. Cell. 38:889-900. 170 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 171 Ginés-Candelaria, E., A. Blinkova and J. R. Walker. 1995. Mutations in Escherichia coli dnaA which suppress a dnaX(Ts) polymerization mutation and are dominant when located in the chromosomal allele and recessive on plasmids. J. Bacteriol. 177:705-715. Hansen, E. B., F. G. Hansen and K. von Meyenburg. 1982. The nucleotide sequence of the dnaA gene and the first part of the dnaN gene of Escherichia coli K-12. Nucleic Acids Res. 10:7373-7385. Hansen, E. B. and M. B. Yarmolinsky. 1986. Host participation in plasmid maintainance: De endence upon dnaA of replicons derived from P1 and F. Proc. Natl. Aca . Sci., U. S. A. 83:4423-4427. Hansen, F. G., S. Koefoed and T. Atlung. 1992. Cloning and nucleotide sequence determination of twelve mutant dnaA genes of Escherichia coli. Mol. Gen. Genet. 234:14-21. Hasunuma, K. and M. Sekiguchi. 1977. Replication of plasmid p80101 in Escherichia coli K12: requirement for dnaA function. Mol. Gen. Genet. 154:225-30. Hupp, T. R. and J. M. Kaguni. 1988. Suppression of the Escherichia coli dnaA46 mutation by a mutation in trxA, the gene for thioredoxin. Mol. Gen. Genet. 213:471-478. Hupp, T. R. and J. M. Kaguni. 1993. DnaA5 protein is thermolabile in initiation of replication from the chromosomal origin of Escherichia coli. J. Biol. Chem. 268:13128-13136. Hwang, D. S. and J. M. Kaguni. 1988. Interaction of dnaA46 protein with a stimulatory protein in replication from the Escherichia coli chromosomal origin. J. Biol. Chem. 263:10633-10640. Katayama, T. and A. Kornberg. 1994. Hyperactive initiation of chromosomal replication in vivo and in vitro by a mutant initiator protein, DnaAcos, of Escherichia coli. J. Biol. Chem. 269:12698-12703. Kline, B. C., T. Ko oma, J. E. Tam and M. S. Shields. 1986. Requirement of the scherichia coli dnaA gene product for plasmid F maintenance. J. Bacteriol. 168:440-3. Kucherer, C., H. Lother, R. Kolling, M. A. Schauzu and w. Messer. 1986. Regulation of transcription of the chromosomal dnaA gene of Escherichia coli. Mol. Gen. Genet. 205:115-121. Marszalek, J. and J. M. Kaguni. 1994. DnaA protein directs the binding of DnaB protein in initiation of DNA replication in Escherichia coli. J. Biol. Chem. 269:4883-4890. 172 27a. Messer, W. and C. Weigel. initiation of chromosome replication. In 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. Neidhardt, F. C., J. L. Ingraham, K. B. Low, B. Ma asanik, M. Schaechter and H. E. Umbarger (eds.), Escherichia coli and almonella typhimurium. Cellular and Molecular Biology, 1996. American Society for Microbiology, Washington, D. C. Miller, J. H. 1992. A short course in bacterial genetics. Oka, A., K. Sugimoto, M. Takanami and Y. Hirota. 1980. Replication origin of the Escherichia coli K-12 chromosome: the size and structure of the minimum DNA segment carrying the information for autonomous replciation. Mol. Gen. Genet. 178:9-20. Ptashne, M., K. Backman, M. Z. Humayun, A. Jeffrey, R. Maurer, B. Meyer and R. T. Sauer. 1976. Autoregulation and function of a repressor in bacteriophage lambda. Science. 194:156-61. Rossmann, M. G., D. Moras and K. W. Olsen. 1974. Chemical and biological evolution of nucleotide-binding protein. Nature. 250:194-9. Rost, B., C. Sander and R. Schneider. 1994. PHD--an automatic mail server for protein secondary structure prediction. Comput. Appl. Biosci. 10:53-60. Roth, A. and W. Messer. 1995. The DNA binding domain of the initiator protein DnaA. EMBO J. 14:2106-2111. Sakakibara, Y. 1993. Cooperation of the prs and dnaA gene products for initiation of chromosome replication in Escherichia coli. J. Bacteriol. 175:5559-5565. Sambrook, J., E. F. Fritsch and T. Maniatis. 1989. Molecular Cloning; A Laboratory Manual. Saraste, M., P. R. Sibbald and A. Wittinghofer. 1990. The P-loop-- a common motif in ATP- and GTP-binding proteins. Trends Biochem. Sci. 5:430-434. Schweizer, H. and W. Boos. 1983. Transfer of the A(argF- Iac)U169 mutation between Escherichia coli strains by selection for a closely linked Tn10 insertion. Mol. Gen. Genet. 192:293-4. Sekimizu, K., D. Bramhill and A. Kornberg. 1987. ATP activates dnaA protein in initiating replication of plasmids bearing the origin of the E. coli chromosome. Cell. 50:259-265. Sekimizu, K., D. Bramhill and A. Kornberg. 1988. Sequential early stages in the in vitro initiation of replication at the origin of the Escherichia coir chromosome. J. Biol. Chem. 263:7124-30. Skarstad, K., K. von Meyenburg, F. G. Hansen and E. Boye. 1988. Coordination of chromosome replication initiation in Escherichia coli. effects of different dnaA alleles. J. Bacteriol. 170:852-858. 40. 41. 42. 43. 173 Stenzel, T. T., T. MacAlIister and D. Bastia. 1991. Cooperativity at a distance promoted by the combined action of two replication initiator proteins and a DNA bending protein at the replication origin of pSC101. Genes & Dev. 5:1453-1463. Sutton, M. D. Ph. D. Thesis, Chapter II. Tucker, W. T., C. A. Miller and S. N. Cohen. 1984. Structural and functional analysis of the par region of the pSC101 plasmid. Cell. 38:191-201. Wang, Q. P. and J. M. Kaguni. 1987. Transcriptional repression of the gnaA gene of Escherichia coli by dnaA protein. Mol. Gen. Genet. 09:518-525. Wickner, S. H. and D. K. Chattoraj. 1987. Replication of mini-P1 plasmid DNA in vitro requires two initiation proteins, encoded by the repA ene of ghage P1 and the dnaA gene of Escherichia coli. Proc. Natl. Acad. ci., U. . A. 84:3668-72. . Chapter V Identification of Essential Residues in the DNA Binding Domain of the Escherichia coli DnaA Protein 174 Abstract The Escherichia coli DnaA protein, as a sequence-specific DNA binding protein, promotes the initiation of chromosomal replication by binding to asymmetric 9-mer sequences termed DnaA boxes that are present four times in oriC. Characterization of N-terrninal, internal, and C-tenninal deletion mutants indicated that residues near the C-terrninus of DnaA protein are required for DNA binding. Furthermore, genetic and biochemical characterization of eleven missense mutations mapping within the C-terrninal 89 residues indicated that they are defective in DNA binding. Detailed biochemical characterization of one mutant protein that contains a threonine-to-methionine substitution at position 435 (T435M) indicated that it retained only nonspecific DNA binding activity, suggesting that the residue involved imparts specificity in binding. Finally, T435M was inactive on its own for in vitro replication of an oriC plasmid but able to augment limiting levels of wild type DnaA, consistent with the proposal that not all of the DnaA monomers in the initial complex are bound specifically to oriC. 175 Introduction The E. coli origin of chromosome replication, oriC, is bound by about 20-30 monomers of DnaA protein to form a large nucleoprotein structure (10). Subsequent unwinding of the AT-rich region at the left end of oriC by DnaA that is aided by HU or integration host factor (lHF) (17) allows entry of DnaB helicase resulting in formation of the prepriming complex (13). Biochemical characterization of DnaA protein has revealed that it is multifunctional possessing activities of binding to adenine nucleotides (29), to acidic phospholipids (30, 36), to DnaB protein (20), and to itself as a self- aggregate (12). Finally, it is a sequence-specific DNA binding protein that binds to an asymmetric 9-mer sequence (5’—TTATC/ACAC/AA—3’) referred to as the DnaA box (10). its activity as a DNA binding protein appears to be critical not only for its role in initiation of chromosome replication, but also for the expression of genes containing DnaA box motifs in promoters or within coding regions. Its binding to these DnaA boxes results in repression (18, 24, 32, 34, 35), activation (2), or termination of transcription (1, 28), respectively (reviewed in reference 21). The best studied of these genes whose transcription is affected by DnaA protein is the dnaA gene itself (18, 34). A single DnaA box located in the dnaA promoter region is bound by DnaA protein to repress transcription from both dnaA promoters. Recently, we described the identification of a large collection of E. coli dnaA alleles (see Chapter ill). The genetic and biochemical characterization of these alleles suggested that DnaA protein contains four functionally distinct domains (see Chapter M. An in vivo assay to measure DNA binding activity was established that relied on transcriptional repression of a dnaA promoter’ 176 177 7302 fusion by binding of DnaA protein to the DnaA box in the dnaA promoter region. Mutations encoding amino acid substitutions within the C-terrninal 89 residues or lacking C-tenninal coding sequence were unable to repress expression, suggesting that the C-terrninus of DnaA was involved in DNA binding. Consistent with this conclusion was the recent report by Roth and Messer (27) describing that the C-terminal 94 residues of DnaA when fused to B-galactosidase were sufficient for specific binding to a DNA fragment containing oriC. Here, we report that the C-terminal 89 residues of DnaA are required for specific DNA binding. Furthermore, genetic and biochemical characterization of a large set of missense mutations mapping within this region identify residues critical for DNA binding activity. A structure for the DNA binding domain of DnaA is proposed. Materials and Methods Bacterial strains. E. coli strains were WM433 (4), lea-19 pro-19 trp-25 his-47 thyA59 arg-58 met-55 deoBZ3 lac-11 gal-11 strA56 suI-1 hstK"2 dnaA204(Ts); XL1-Blue (7), recA1 endA1 gyrA96 (Nal') thi hdei17 (rK_12‘mK_12+) supE44 reIA1 lac (F’::Tn 10 proA“B+ Iacl‘l A(IacZ) M15); HMS174(DE3)(pLysS) (9), recA1 hdei (rK,12‘mK_12+) Rifr F‘; BLAL21(DE3)(pLysS), ompT hst (rB’mB‘) recA56 er-300.:Tn 10 F‘. BLAL21(DE3) is a recA56 sd-300:Tn 10 derivative of BL21(DE3) (9) kindly provided by Dr. Laurie S. Kaguni. Recombinant DNA techniques. Plasmid DNAs used in this study are described in Table 1. The deletion mutants containing N-tenninal truncations (dnaAA62, dnaAA129, and dnaAA219) were constructed by polymerase chain reaction (PCR)-mediated site-specific mutagenesis using the following oligonucleotide primers (complementary to the indicated positions with position 1 corresponding to the first nucleotide of the coding sequence (15)) synthesized by an in-house facility with an Applied Biosystems Model 394: JK—36, 5’—AATATCAATQATATQCTAACCAGTTTCTGC—3’ (178-201 of the template strand); JK—37, 5’—CAGAACCGACCQATALGTCTAACGTAAACG—3’ (371-402 of the template strand); JK—38, 5’-CAAAACAACQATA] GGAAGAGT'ITAAACGC—3’ (643-672 of the template strand); JK—6, 5’—CGGAGCGTACCAGGA‘EQQGTTCACCTTCCA-E (1 .683- 1,712 of the coding strand). The oligonucleotides JK—36, JK—37, and JK—38 contained nucleotide sequence that was altered relative to the coding sequence of the dnaA gene that generated an Nde I restriction endonuclease site (5’—CATATG-3’; indicated above by underlining) and the oligonucleotide JK—6 introduced a Bam HI site (5’-GGATCC-3’; indicated above by underlining) at the appropriate locations to allow cloning of the PCR product 178 179 TABLE 1. Plasmid DNAs Plasmid Characteristicsa 03:61:;ng b pET11a Apr, pBR322 origin, expression vector Novagen, Inc. (9) that places gene of interest under 17 promoter control pK0596 dnaA+ cloned into pET11a This laboratory (8) pKCA62 dnaAA62 cloned into pET11a This work pKCA129 dnaAA129 cloned into pET11a This work pKCA219 dnaAA219 cloned into pET11a This work pKCA220-294 dnaA4220-294 cloned into pET11a This laboratory pKCA237-378 dnaAA237-378 cloned into pET11a This work pKC43 A440T cloned into pET11a This work pKC49 T435M cloned into pET11a This work pK054 R407H cloned into pET11a This work pKC70 V437M cloned into pET11a This work pKC79 G426D cloned into pET11a This work pKC83 A440V cloned into pET11a This work pKC110 V383M cloned into pET11a This work pKC130 A428T-A440T cloned into pET11a This work pKC170 T435M-T436M cloned into pET11a This work pKCM2 A412P cloned into pET11a This work pKCM7 412(QMA)413 cloned into pET11a This work a Nomenclature is as follows: A, deletion of amino acid residues; A440T (and like), alanine- to-threonine substitution at position 440; A428T-A440T, dash indicates that both mutations are present in the same mutant (i.e. double mutant); 412(QMA)413, Met, Ala insertion between residues 412 and 413. b See Materials and Methods for a description of plasmid constructions. 180 into the 8am Hl-Ndel linearized pET11a vector fragment. PCR reactions contained Vent DNA polymerase (New England Biolabs) as per the manufacturer's recommendation, pKC596 (see Table 1) as the DNA template, and a Perkin Elmer Cetus DNA Thermal Cycler. pKCA220-294 is a pKC596 derivative that lacks a 225 bp fragment corresponding to residues 220-294 of DnaA that was liberated by Pvul endonuclease restriction and was generated by standard in vitro molecular biological techniques (J. Lipar, M. D. Sutton, and J. M. Kaguni, data not shown). dnaAA237-378 and the dnaA missense mutant alleles that were described previously (see Chapters Ill and IV) were subcloned from pACY0184 into pET11a by Eco Ri-Rsr ll restriction of the pACYC184 recombinant containing the mutant dnaA allele, purification of the 1,402 bp fragment corresponding to the dnaA coding sequence (residues 21-467), and its subsequent ligation to the gel purified 5.9 kb pKC596 vector fragment that was obtained by partial Eco RI restriction of the Fisr ll linearized pKC596 plasmid. Alternatively, some of the mutant dnaA alleles were amplified by PCR using the oligonucleotide primers JK—6 and JK-7, (8) restricted with Ndel and Rsr ii, and ligated to the Nde l-Flsr ll pET11a vector fragment. Qverexpression and purification of recombinant proteins. Overexpression of target genes contained in pET11a recombinants was performed by addition of isopropyl-B-D-thiogalactopyranoside (iPTG) (U. S. Biochemicals) to 0.05 mM (final concentration) when cultures reached an OD595 of 0.6-0.8 units. Purified forms of monomeric DnaA and T435M were obtained as described (22). Briefly, cells (HMSi74(DE3)(pLysS)) induced to express DnaA protein (or T435M) were lysed by a gentle freeze-thaw lysis procedure (11). The soluble fraction was then diluted to a conductivity equivalent to Buffer A (25 mM HEPES-KOH pH 7.6, 15% glycerol, 0.1 mM EDTA, and 0.2 mM D'l'l') 181 containing 50 mM KCl and applied to Heparin Sepharose at a ratio of 20 mg protein to 1 ml of packed resin. After loading, the column was washed with 50 column volumes of Buffer A containing 100 mM KCI. Bound proteins were eluted with a linear gradient of 0.1 M to 1 M KCl in Buffer A. Fractions containing DnaA+ (or T435M) were identified by SDS-PAGE analysis and pooled. DnaA+ (or T435M) was then precipitated by dialysis against Buffer C (50 mM HEPES-KOH pH 7.6, 20% glycerol, 1 mM EDTA, and 2 mM DTT) overnight at 4°C. Precipitated proteins were collected by centrifugation, washed with Buffer C containing 0.6 M (NH4)2SO4, and solubilized in Buffer C containing 0.6 M (NH4)2SO4, 10 mM MgOAc, and 4 M guanidinium hydrochloride at a final protein concentration of approximately 10 mg/ml. Gel permeation chromatography using a Superose 12 column equilibrated with Buffer D (50 mM HEPES-KOH pH 7.6, 15% glycerol, 0.1 mM EDTA, 10 mM MgOAc, 0.2 mM (NH4)ZSO4, and 2 mM DTT) was used to remove the guanidinium hydrochloride and obtain the active monomer of DnaA protein (or T435M). Other mutant forms of DnaA containing amino acid substitutions were purified partially from 100 ml cultures grown at 30°C of BLAL21(DE3)(pLysS) containing the respective mutant dnaA alleles in pET11a. As these mutants were more sensitive to proteolysis then were DnaA+ or T435M (data not shown), cell lysis was performed in the presence of 4 mM phenylmethyl-sulphonyl flouride (PMSF). Soluble fractions were chromatographed batchwise with Blue Dextran Agarose. After extensive washing of the bound material with Buffer A containing 100 mM KCl, bound proteins were eluted stepwise with Buffer A containing 1 M KCI. V383M, A412P, and A428T-A440T were found to be extremely unstable. Full length protein for these mutants was obtained only by rapid lysis on ice by sonication in the presence of PMSF (4 mM) followed by 182 addition of SDS (0.25%) to the soluble fraction. DnaAA62 and DnaAA129 were purified partially by batchwise chromatography essentially as described above for the mutant forms of DnaA containing amino acid substitutions except that Blue Dextran Agarose chromatography was replaced by Heparin Sepharose. Whereas DnaAA62 and DnaAA129 were present almost exclusively in the soluble fraction (data not shown), DnaAA219 partitioned exclusively to the insoluble fraction (data not shown). DnaAA219 was purified from a 200 ml culture of HMS174(DE3)(pLysS)(pKCA219) that was induced by addition of lPTG (0.05 mM). The insoluble fraction that was obtained following cell lysis as described above was washed once with Buffer A containing 100 mM KCI and 2% Triton X- 100, then two times with Buffer A containing 100 mM KCI. The pellet was then solubilized in Buffer A containing 100 mM KCl and 6 M guanidinium hydrochloride and dialyzed against 1,000 volumes of Buffer A containing 0.5 M KCI, 10 mM M9804, and 0.5% Triton X-100 for 16 hours at 4°C. The dialyzed material was then combined with 0.5 volumes of Heparin Sepharose and the mixture was diluted with Buffer A to a conductivity equivalent to Buffer A containing 50 mM KCI. After packing the slurry into a column (0.5 cm diameter) it was washed with 50 column volumes of Buffer A containing 0.5% Triton X-100 and 100 mM KCI, then eluted stepwise with Buffer A containing 0.5% Triton X- 100 and 1 M KCI. Fractions containing DnaAA219 were identified by SDS- PAGE analysis. Protein determinations. Protein concentrations were determined by scanning densitometry of Coomassie-stained polyacrylamide gels using a Bio- Rad Gel Doc 1000 equipped with the Molecular Analyst software package. Alternatively, protein concentrations were determined by quantitative immunoblot analysis using an 125i-labelled antibody specific to DnaA protein as 183 described previously (see Chapter II). Southwestern blotting. Southwestern blotting was performed based on the method of Bowen et al. (5). Samples were electrophoresed and blotted onto Westran membrane (Schleicher and Schuell) as described (31). Membranes were then incubated in binding buffer (20 mM HEPES-KOH pH 7.6, 2.5 mM MgOAc, 15% glycerol, 0.4% Triton X-100, 2 mM EDTA, 4 mM DTT, 5 mg/ml BSA, and 0.5 uM ATP) at 4°C with gentle rocking for 1 h to allow for protein renaturation. Radiolabelled DNA fragment (either the 464 bp Sma l-Xho l oriC fragment from pBSoriC (3) or a 466 bp Bam Hl restricted PCR-generated fragment corresponding to the C-terrninus of DnaA that did not contain a DnaA box motif) was then added at a ratio of 1 to 30 (pmol proteinzfmol DNA) and incubated at 4°C for an additional 5 h. 3’ end-labelling was performed with the large fragment of DNA polymerase I and the appropriate [a32P]dNTP (Dupont, NEN). Membranes were then washed in binding buffer at 4°C to remove unbound fragment. Quantitation of the amount of radiolabelled DNA fragment retained by each protein sample was by use of a Packard instant lmager relative to a standard curve of the respective radiolabelled DNA fragment. Bound DNA was stripped from the membrane by incubation in 154 mM NaCl, 10 mM Tris-HCI pH 7.4, and 0.25% SDS. Membranes were then processed as western blots as described previously (31) using a monoclonal antibody specific to DnaA (M43 that recognizes residues 133 to 141) followed by a secondary antibody that was labelled with 125l (ICN Biochemicals). The amount of DNA retained by each protein sample was then normalized for the amount of that protein present on the membrane based on 125I counts relative to a standard curve. DnaAA62, DnaAA129, and DnaAA219 were retained on Westran membrane as efficiently as was DnaA under the conditions used for 184 electrotransfer (data not shown). Consequently, the amount of protein present for these samples was assumed to be equivalent to that loaded onto the SDS- PAGE as was the case with wild type DnaA. DNA binding assays. Fragment retention assays (20 pl) were performed as described (10) and contained 20 fmol of the indicated 3’ end-labelled DNA fragment and 50 ng (17.5 fmol as nucleotide) Hinfl digested pBR322 as competitor (unless otherwise stated) in buffer containing 40 mM HEPES-KOH pH 7.8, 5 mM MgCIZ, 50 mM KCI, 2 mM DTT, and 0.5 uM ATP. DNA fragments used for fragment retention assays included the 464 bp Sma I-Xhol oriC fragment from pBSoriC (3), the 1,097 bp Ban l ColE1 origin fragment from pBluescript, the 686 bp Eco Ri-Spe l pSC101 origin fragment from pCM128 (33), a 417 bp fragment containing the dnaA promoter region obtained by PCR using the oligonucleotide primers JK-23 (5’-AAGCCAATTTTTGTCTATGG—3’ that is complementary to positions -302 to -321 of the template strand) and the M13 —40 universal primer (U. S. Biochemicals) and pRB1A (6) as the DNA template, and a 466 bp fragment corresponding to the C-terrninus of the dnaA gene that was used as a nonspecific DNA fragment that was also generated by PCR using the oligonucleotide primers JK—6 (see above for the sequence of JK—6) and JK-29 (5’—GATGGCGCTGGCGAAAGAGC—3’ that is complementary to positions 1,230-1,249 of the template strand) and pACYCdnaA as the DNA template. Following incubation at 30°C for 10 min with the indicated amount of DnaA or T435M, reactions were filtered through nitrocellulose filters (Millipore HWAP, 0.22 nM, 13 mm), washed with 250 pl of the above buffer equilibrated to 30°C, and dried. Bound DNA was quantitated by liquid scintillation counting. Alternatively, binding to the same 3’ end-labelled 464 bp Sma l-Xho I oriC fragment, or to the following 22 bp duplex oligonucleotide corresponding to 185 box R4 of oriC and flaking sequences was measured by gel mobility shift assays (25) (the DnaA box is indicated by bold type): 5 ’ -GACAGAGTTATCCACAGTAGAT-3 ’ 3 ’ -TGTCTCMTAGGTGTCATCTAG-5 ’ Reactions (10 ul) contained the indicated amount of wild type or mutant DnaA protein, 25 fmol of the indicated 3’ end-labelled DNA substrate, and 50 ng (17.5 frnol as nucleotide) of Hinfl digested pBR322 as competitor in buffer containing 20 mM HEPES-KOH pH 7.6, 2.5 mM MgOAc, 15% glycerol, 0.4% Triton X-100, 2 mM EDTA, 4 mM DTT, 5 mg/ml BSA, and 0.5 uM ATP. Reactions were electrophoresed through 4% (for reactions containing oriC) or 10% (for reactions containing the oligonucleotide) native polyacrylamide gels (60 acrylamide:1 bis-acrylamide) in 90 mM Tris-borate pH 8.3 and 1 mM EDTA, dried, and exposed to Hyperfilm MS (Amersham) at -70°C using a Cronex Quanta Ill intensifying screen. Competitive gel mobility shift assay. Competitive gel mobility shift assays were performed as described above using the same oriC fragment and the indicated amounts of DnaA or T435M except that in place of Hinfl digested pBR322 as competitor, either the same unlabelled oriC fragment or the 1,031 bp Afl ill-Ava l| fragment from pBluescript were used as competitor at the indicated concentrations. DNAs were added prior to addition of proteins and reactions were incubated for 10 min at room temperature prior to electrophoresis through 4% native gels. Quantitation of the bound and free radiolabelled oriC fragment was by use of a Packard Instant Imager. ATP binding assays. Reactions (25 ul) were performed as described (29) and contained 1 pmol (52 ng) of DnaA or T435M and the indicated amount of [P2P]ATP (Dupont, NEN) in buffer containing 50 mM Tris-HCI pH 8.0, 0.5 mM MgOAc, 15% glycerol, 0.01% Triton X-100, and 5 mM DTT. incubations were at 186 0°C for 10 min followed by filtration through nitrocellulose filters (Millipore HWAP, 0.22 nM, 13 mm) that were equilibrated in ATP binding buffer. Filters were washed with 0.5 ml of the same buffer and dried. Bound [yazP]ATP was quantitated by liquid scintillation counting. Assay of in vitro replication. Assays of in vitro replication were performed as described previously using a crude protein fraction deficient in DnaA protein (11). Resufls The C-terminal 89 residues of DnaA are required for specific DNA binding. DnaA protein binds to four DnaA boxes within oriC to promote initiation of replication, as well as to DnaA boxes within origins of plasmids and bacteriophage that require this protein for their replication (10). To determine the region of DnaA protein required for DNA binding activity, quantitative Southwestern blot analysis was performed with deletion mutants that lacked N-tenninal, C-terminal, or internal coding sequences (Figure 1). The respective ability of DnaA protein to retain a DNA fragment containing oriC (464 bp) was compared to a DNA fragment from the C-terrninus of the dnaA gene (466 bp) lacking a DnaA box motif. This analysis indicated that the oriC fragment was retained approximately 1.5-fold more efficiently (Figure 2). Mutant proteins lacking either N-terminai or internal sequences retained the oriC fragment with an efficiency similar to DnaA protein, indicating that the N-terminal 378 residues of DnaA are not required for specific DNA binding (Figure 3 and Table 2). Deletion mutants lacking C-terminal sequences were defective in retaining the oriC fragment, suggesting that this region is required for DNA binding activity. To confirm that residues N-terrninal to position 379 were dispensable for DNA binding activity, gel mobility shift assays were performed with an oriC- containing fragment or a duplex oligonucleotide containing a DnaA box (box R4 of oriC and natural flanking sequences) (data not shown). Consistent with the results of Southwestern blotting, the deletion mutants, DnaAA219 (lacking the N-terminal 219 residues), DnaAA220-294, and DnaAA237-378 (see Figure 1), were able to bind to either DNA in a manner comparable to that observed with the wild type protein. Attempts to overexpress the C-terrninal 88 residues 187 188 1 Figure 1. Structures of truncated forms of DnaA protein. Structures of the truncated forms of DnaA protein used in this study are indicated. Numbers refer to amino acid residues. DnaAA62, DnaAA129, and DnaAA219 code for a methionine at positions 62, 129, and 219, respectively. DnaAA220-294 and DnaAA237-378 are in-frame deletions. All were constructed in vitro as described in Materials and Methods except dnaAA237-378 that was identified using a novel selection for mutations of dnaA (see Chapter IV). 189 Dam>+ ... .. .. x f a t i. ‘ ... , we...“ . . t a s . .. v . ... .t. ...u: ... ~ .A‘ LII)... . ....u.1 l . W. “at... v. ... Many: ... Q I .. Akin 0333mm ..\. .9... r... ..., ..Au....ta ...1na. . f.mv:. 5,.v . 9>.. .. ,. A.»....\,:: {:.. .. a t .331... ff V6... 2 “:..: .. .. . ., . . ... L ..s ....N s cm... ......ru . . put... ... ..1... a, 4 :...r. .4 . ,‘91‘ . UV...“ rm. .2). ...4. 4 n .1 . a, .5 . u.!. slaw. - t\l. ..‘t 58 . .v t a. . :3”.- ..u: v, n. .12....»4. . .. l - .... u. . . . ,n..5.1.... \Losc.x><.b~. . u _. it . no ~ .~ .- s .. s .u u .u u .N. (3.99.: ...! N .1. . ..r.l.L ..."...l iv! Dzm>>mmo$ms A .. . . ._ .u.rfl..-il.lal 1,1 .., . ‘nqt..1..< ,wwnyr..... , p . x u .....5. . . . .‘1....> .45.... (134...... mil-II .- Oam>>mm$wum Gags—3&3 .q ~.. ...I in. ... N a m. . em... .v.u.n§u....£w.a.. ......(rownuu‘flakRuJNHNNK . . 1... U:m>-m3wmn _ _ . g 190 Figure 2. DnaA protein binds preferentially to DNA fragments containing oriC by Southwestern blot analysis. Southwestern blot analysis was performed in duplicate using the indicated amounts of DnaA protein as described in Materials and Methods. Samples were run on the same 10% SDS-PAGE. After electrotransfer, Westran membrane was cut in half. One half was probed with a 464 bp radiolabelled DNA fragment containing oriC (O), and the other with a 466 bp radiolabelled DNA fragment corresponding to the C-terrninus of the dnaA gene that lacked a DnaA box motif (El). Both halves were incubated in identical volumes of binding buffer containing equivalent amounts (fmol) of the respective radiolabelled DNA fragments. Quantitation of the amount (fmol) of each radiolabelled DNA fragment retained by DnaA+ was with a Packard instant Imager. To calculate fmol of DNA retained, the specific activity of each DNA fragment was determined simultaneously with quantitation of the Southwestern blot by counting known amounts of each fragment. 191 2.00- 1.50— 1.00— Fragment Retained (fmol) a»? l 0.00 0 'r'r'r'r'r'r 50 100150 200 250 300 DnaA (ng) 192 Figure 3. Southwestern blot analysis of truncated forms of DnaA protein. A representative Southwestern blot probed with a radiolabelled DNA fragment containing oriC is shown. Truncated forms of DnaA protein used in this analysis were greater than 50% pure as determined by SDS-PAGE (data not shown). Prior to Southwestern blotting, the membrane was stained with Ponceau S to confirm that proteins were transferred to the Westran membrane (data not shown). Similar experiments were performed to obtain the data presented in Tables 2 and 3 except that Ponceau S staining was replaced by quantitative immunoblot analysis to normalize fmol DNA retained to protein concentration as described in Materials and Methods. 193 194 TABLE 2. Quantitative Southwestern blot analysis of truncated forms of DnaAa Protein Amino acids deleted Relative retention of MC" DnaA+ None 21.00 DnaA-am446 446-467 NDc DnaA-am361 361-467 ND DnaAA62 2-62 1.34 (10.47) DnaAA129 2-129 1.33 (:l:0.41) DnaAA219 2-219 1.19 (10.22) DnaAA220-294 220-294 1.04 (i029) DnaAA237-378 237-378 0.92 (:l:0.14) 8 Binding to oriC was measured by Southwestern blot analysis using 200 ng of each protein as described in Materials and Methods. b Retention of oriC is expressed relative to that observed by DnaA+ that was normalized to 1.00. Depending on the experiment, 200 ng of DnaA+ retained between 0.9 and 1.8 fmol of the on'C fragment. Each value is normalized and represents the average of at least three experiments i the standard deviation (calculated after normalization). 6‘ ND, none detected. Binding to oriC was not detected by either autoradiography or scanning with a Packard instant imager. 195 of DnaA protein to measure its DNA binding activity were unsuccessful. The reason may relate to proteolytic instability of the recombinant protein. Mutants of DnaA protein with substitutions within the C-terminal 89 residues are defective in DNA binding. The above results suggested that residues from position 379 to the C-terrninus are sufficient for DNA binding activity. To substantiate this conclusion, eleven mutants containing amino acid substitutions in this region were examined by quantitative Southwestern blotting (Table 3). All were much less efficient at retention of the oriC fragment compared to DnaA. DNA binding activity was also measured in a gel mobility shift assay with a duplex oligonucleotide containing a DnaA box motif (box R4 of oriC and natural flanking sequences) and a constant amount of wild type or mutant DnaA protein (Figure 48). Whereas none of the mutant proteins formed a complex with the oligonucleotide at the level tested, wild type DnaA protein did at levels as low as 2 ng (Figure 4A). Due to proteolysis of the mutants V383M (with a V-to-M mutation at position 383), A412P, and A428T-A440T on induced expression, we were not able to purify them for biochemical study to corroborate the results of Southwestern analysis. Proteolysis was indicated by the observation that they were overexpressed efficiently then rapidly converted, over time, from a species that migrated to a position similar to purified DnaA protein in an SDS-PAGE (as monitored by lmmunoblotting) to a mixture of smaller species. For these mutants, full length protein was recovered only if induced expression was for 45 minutes instead of for 3 hours that was used for the other missense mutants, followed by cell lysis on ice by sonication in the presence of phenylmethyl- sulfonyl flouride (PMSF) and addition of SDS (0.25%) to the soluble fraction. With these exceptions, results from these two methods indicate that mutants 196 TABLE 3. Mutant proteins with substitutions in the C-terrninal region of DnaA protein are defective in binding to oriCa Protein Relative retention of oriC” DnaA“ a1 .00 V383M 0.1 0 R407H 0.1 8 A412P 0.06 412(QMA)413 0.07 G426D 0.1 2 T435M 0.46 T435M-T436M 0.28 V437M 0.09 A440V 0.08 A440T 0.1 2 A428T-A440T 0.1 2 «9 Binding to oriC was measured by Southwestern blot analysis using 200 ng of each protein as described in Materials and Methods . b Retention of oriC is expressed relative to that observed by DnaA protein that was normalized to 1.00 (200 ng of DnaA retained 1.8 fmoi of the on'C fragment). 197 Figure 4. Mutant forms of DnaA protein are defective ln DNA binding. (A) Binding of DnaA+ protein to a duplex oligonucleotide (25 fmoi) containing nucleotide sequence corresponding to that of box R4 of oriC (and natural flanking sequences) by gel mobility shift analysis. The amounts of DnaA protein used are indicated. assays were performed as described in Materials and Methods. (B) Mutant forms of DnaA protein (25 ng) were assayed for their ability to bind the same duplex oligonucleotide as above. The amino acid, the substitution, and its position are indicated for each of the mutants (i.e. R407H, arginine-to-histidine substitution at position 407; 412(QMA)413, Met, Ala insertion between residues 412 and 413). The positions of wells, complexes, and free oligonucleotide are indicated. 198 g E tn (3“) .Vvua —- xerdwog 259181720 x91dm03 91PM on None ., ' DnaA’ R407H 4 l 2(QMA)41 3 G426D T435M T435M-T436M A440V A440T V437M 199 containing substitutions near the C-terrninus of DnaA protein are defective in DNA binding. T435M is defective in specific DNA binding activity. As T435M retained the greatest affinity for oriC as measured by Southwestern blotting, its DNA binding activity was investigated in more detail by nitrocellulose fragment retention assays using a variety of DNA substrates (Figure 5). In comparison to DnaA+ protein, T435M was inefficient in retention of DNA fragments containing one or more DnaA box motifs (in the presence or absence of competitor DNA; data not shown for experiments without competitor DNA except in the case of the “nonspecific” fragment), including oriC, the ColE1 origin, the dnaA promoter region, and the pSC101 origin. Both DnaA+ and T435M were similarly ineffective in binding to a DNA fragment lacking a DnaA box (this DNA fragment termed “nonspecific” was from the C-terrninal coding region of the dnaA gene). The extent of retention by either protein to this DNA was not affected by the inclusion of Hinfl digested pBR322 (as a competitor) and was similar in extent of binding to that observed with T435M to the other DNAs examined. The relative affinity of DnaA for each of the DNA substrates investigated was similar to previously published reports (10). Comparable fragment retention assays were performed with a restriction digest of a pSC101 plasmid (pCM128). Fragments bound and not bound by DnaA+ or T435M were visualized by autoradiography after electrophoresis (data not shown). Whereas DnaA+ protein bound to the origin-containing fragment preferentially, T435M showed poor preference for the origin- containing fragment. Both T435M and DnaA+ bound to DNA nonspecificaliy at protein concentrations of 100 to 200 ng. Collectively, these observations suggest that the DNA binding activity of T435M is largely nonspecific. To demonstrate further that T435M was defective in the ability to bind 200 Figure 5. T435M is defective in high affinity DNA binding. Fragment retention assays with wild type DnaA protein (0) or T435M (O) and 20 fmoi of each DNA fragment were performed as described in Materials and Methods. All reactions, except that indicated with the nonspecific DNA substrate, contained Hinfl digested pBR322 (50 ng) as competitor DNA. Fragment Retained (fmoi) 20 Doric 15 10 r'r'r'r‘ 050100150200 201 20 _ 00151 origin 'l'l'l‘i' 050100150200 20.5 dnaA promoter region 20.: pSC101 origin 15% 15% 1o; 10% 5% 5% 01-..,.,-,.T.o‘ ,.,.,.,. O 50 100 150 200 0 50 100 150 200 20 Nonspecific O&; i I I [#1 I 20 _‘ Nonspecific (no competitor) 155 105 5.‘ T 0 50 100 150 200 OW' l'l'i 050100150200 Protein (ng) 202 specifically to DNA, the effect of competitor DNAs on its binding to a fragment containing oriC was examined. In a gel mobility shift assay, it was able to bind to a DNA fragment containing oriC (Figure 6). DnaA+ protein formed several discrete complexes. These have been shown to result from the ordered binding of DnaA protein to the four DnaA boxes in oriC (19), in contrast, T435M formed a smear from the position of free DNA to the wells, suggesting nonspecific binding. Furthermore, at least 8-fold more T435M was required relative to DnaA to shift all of the oriC fragment from the “free" position. With this DNA binding assay, the effect of competitor DNAs on the DNA binding activities of DnaA+ and T435M was examined. Binding of fixed amounts of DnaA“ or T435M to a radiolabelled oriC fragment was measured in the presence of increasing amounts of either the same oriC fragment (unlabelled) or a 1,031 bp fragment (unlabelled) from pBluescript (Figure 7). The 1,031 bp DNA fragment lacks a DnaA box motif. Inasmuch as the 1,031 bp fragment was about 2-fold larger in size than the oriC fragment (464 bp), it was a poor competitor for binding of wild type DnaA to oriC. However, the 1,031 bp fragment was comparable to the unlabelled on'C fragment in reducing binding of T435M to labelled oriC (a 4-to-4.5-fold excess of unlabelled oriC was required relative to a 7-to-7.5-fold excess of the 1,031 bp fragment to compete away 50% of bound radiolabelled oriC from T435M). These results indicate that the affinity of T435M for a nonspecific fragment was similar (within 2-fold) to that for oriC. A 4-to-4.5-fold molar excess of unlabelled oriC was required to compete away 50% of the radiolabelled oriC fragment bound to DnaA protein (or T435M). This was presumably due to the fact that ~95-98% of the radiolabelled fragment was shifted from the “free” position. As complexes II through Vl appear to contain more than one DnaA monomer per oriC fragment (19), stoichiometric 203 Figure 6. T435M does not form discrete complexes by gel mobility shift analysis. The DNA binding activity of T435M was measured by use of a gel mobility shift assay as described in Materials and Methods with the indicated amounts of wild type DnaA protein or T435M. 204 Dlafit T435M I ll 1 protein(ng) 0 1 2 4 8 16 3264 4 8 16 3264128256 Wells — 205 Figure 7. Competitive gel mobility shift analysis of DnaA and T435M. Competitive gel mobility shift assays were performed as described in Materials and Methods. Binding of wild type DnaA protein (1.9 ng) or T435M (18 ng) to 25 fmoi of a radiolabelled DNA fragment containing oriC (464 bp) was competed by addition of increasing molar equivalents (from 0 to 25X) of either the same unlabelled oriC fragment (A), or a DNA fragment from pBluescript (1 ,031 bp) lacking a DnaA box motif (B). Quantitation of the percent of radiolabelled oriC fragment bound (i.e. counts shifted from the “free” position) relative to the total oriC fragment present in that reaction (i.e. total counts in that lane) (C) was determined with a Packard Instant Imager. Symbols are: 0, DnaA+ competed with cold oriC fragment; 0, DnaA+ competed with cold nonspecific fragment; El, T435M competed with cold orinragment; I,T435M competed with cold nonspecific fragment. T435M 206 5 T435M 0 DnaA+ ...]. m a :32 .0 § 258 9.5 02m 10'1'5'25'2'5'35'3'5'4335 Competitor (molar equivalent) I 207 (i.e. 1:1 molar ratio of radiolabelled oriC to unlabelled oriC) competition was not observed. The requirement for 4-to-4.5-foid more unlabelled oriC fragment relative to radiolabelled fragment corresponds to a two-fold molar excess of total oriC fragment to total protein. This is consistent with the observed reduction of 50% in binding to the labelled oriC fragment. Two duplex oligonucleotides were also used to compete for binding of DnaA+ or T435M to oriC (data not shown). One contained a DnaA box motif (similar to the oligonucleotide containing box R4 sequence but with different flanking sequences) whereas the other did not. DnaA+ protein formed a single complex with the DnaA box-containing oligonucleotide as measured by gel mobility shift assays (data not shown). Whereas binding of DnaA+ protein to radiolabelled oriC was reduced by 55-60% by a 50-fold molar excess of the DnaA box-containing oligonucleotide (unlabelled), binding of T435M to oriC was unaffected by as much as a 1,000-fold molar excess of the same oligonucleotide. Binding of either protein to on‘C was unaffected by addition of the oligonucleotide that lacked a DnaA box motif, even at a 1,000-fold molar excess. This result, suggesting that T435M was unable to bind specifically to a DnaA box motif was consistent with the inability of T435M to bind a similar oligonucleotide containing sequence corresponding to box R4 of oriC (see Figure 4). Mutant proteins are poorly active for in vitro replication of an oriC plasmid. Of the eleven mutant proteins containing C-tenninal substitutions, five (R407H, A412P, 412(QMA)413, G426D, and T435M-T436M) are inactive for replication from both oriC (data not shown) and the pSC101 origin in vivo (see Chapter llI). Three of the remainder (T435M, A440V, and A428T-A440T) were poorly active for pSC101 replication whereas the other three (V383M, V437M, and A440T) retained replication activity at 30°C. 208 However, the ability of the latter six mutants to sustain replication of the bacterial chromosome from oriC is unknown. As the mutant proteins characterized in this study were defective for DNA binding (Figure 4), we expected that they would also be defective for in vitro replication of an oriC plasmid. Of the eleven mutants, eight were assayed biochemically for their ability to initiate replication from oriC in vitro. We were unable to purify the others (V383M, A412P, and A428T-A440T) because of proteolytic degradation after induced expression (data not shown). Whereas DnaA protein sustained replication of an oriC plasmid in vitro, T435M (see Figure 10 below) and A440V were inactive, even at concentrations as high as 200 ng (Figure 8A). A440T was partially active at 30°C but less so at 40°C, consistent with its temperature sensitive phenotype in vivo for pSC101 replication (see Chapter III). T435M was poorly active even when assayed at high concentrations for a prolonged period of time (Figure BB). Although V437M was fairly active for pSC101 replication in vivo, it was inactive for replication from oriC in vitro. The inactivity of the mutant proteins in replication of an oriC plasmid in vitro is presumably due to their defect in DNA binding. T435M is active for ATP binding. DnaA protein binds ATP and ADP with high affinity (KD of 0.03 and 0.1 nM, respectively) (29). Whereas the ATP-bound form is active for in vitro replication of an oriC plasmid, the ADP-bound form is much less active. To confirm that the defect in DNA binding activity of T435M is specific, the ability of T435M to bind ATP was measured and found to be comparable to DnaA protein (Figure 9). The calculated dissociation constants and estimated stoichiometries (n values) for both T435M (KD 15 nM and n=0.42, respectively) and wild type DnaA (KD 17 nM and n=0.14, respectively) were consistent with those reported previously for the wild type protein (KD 0.03 uM and n=0.48, respectively) (29). 209 Figure 8. Mutant forms of DnaA are poorly active for In vitro replication from oriC. Mutant proteins were assayed for in vitro oriC replication activity at 30° (El, 0) and 40°C (I, 0) using the indicated amounts of protein as described in Materials and Methods. (A) A440V (inset) was representative of all the mutant proteins except A440T (D) that retained partial replication activity. Results for R407H, G426D, 412(QMA)413, T435M (see inset to Figure 10), T435M-T436M, and V437M are not shown. None of the mutant proteins (except A440T) were active for in vitro replication at concentrations as high as 200 ng. Wild type DnaA (O) was saturating for in vitro replication of oriC at levels where A440T was just beginning to show replication activity. (B) Time course of wild type DnaA and T435M for in vitro oriC replication. 50 ng of wild type DnaA protein (0) or 200 ng T435M (El) were assayed for replication activity. Reactions were incubated at 30° (Cl, 0) or 40°C (I, O) and aliquots were removed at the indicated time points and quenched as described in Materials and Methods. 210 0 5 0 2 O 2 r mw 9. 1 r .9, 1mm. 1 _. n O .nlu 0 t 1P 0 5 “...... ... _ _ _ . _ oo o o o m m m m M mm mm 8 6 4 2 A See «485:3