CLONING AND CHARACTERIZATION OF A METHYL-DEPENDENT RESTRICTION ENDONUCLEASE AND A CELL CYCLE REGULATING DNA METHYLTRANSFERASE FROM Zymomonas mobilis SUBSPECIES mobilis CP4 By PRISCILLA LORRAINE PHILLIPS A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2005
Copyright 2005 by Priscilla Lorraine Phillips
iii ACKNOWLEDGMENTS I wish to acknowledge my mentor, Franci s C. Davis, for his encouragement, support, and guidance from my early years as an undergraduate in his lab, inspiring my desire to pursue a Doctoral Degree in Microbiology, and thro ughout my tenure as a graduate student, helping me learn and grow as a scientist. To the members of my committee, Lonnie Ingram, Keelnatham Shan mugam, Julie Maupin, Dean Gabriel, and the former members of my committee Thomas Bobik and James Marunick, I would like to extend my gratitude for their invaluable advice and support. To Angel and Celeste from Thomas Bobikâ€™s lab, I would like to ex tend special thanks for their stimulating conversation, their encouragement, and most of all for their friendshi p. I would also like to thank Louise Munroe for her guidance that taught me to be a be tter teacher and help me discover my enjoyment in instructing st udents. To my husband Alexander Grams and my daughter Veronica, I would like thank for th eir love and support. I would also like to acknowledge my loving sisters, Patricia a nd Petrina, for their encouragement and confidence. To my devoted parents, Du ane A. Phillips and Kyong Hui Phillips, who kindled my love of learning, without whom none of this would be possible, I would like to extend my deepest gratitude and most h eartfelt thanks. To them, I dedicate this dissertation.
iv TABLE OF CONTENTS page ACKNOWLEDGMENTS.................................................................................................iii LIST OF TABLES............................................................................................................vii LIST OF FIGURES.........................................................................................................viii ABSTRACT....................................................................................................................... ..x CHAPTER 1 Zymomonas mobilis ......................................................................................................1 Introduction...................................................................................................................1 Identification and Taxonomy........................................................................................2 General Physiology.......................................................................................................4 Metabolism...................................................................................................................6 Potential Commercial Applications of Zymomonas mobilis ......................................11 An Alternative Source of Fuel Ethanol: Advantages and Disadvantages.................13 Bioengineering............................................................................................................17 R-M Systems of Zymomonas mobilis .........................................................................18 2 DNA RESTRICTION AND MO DIFICATION SYSTEMS......................................20 Discovery and Historical Applications.......................................................................20 General Description of REases and DNA MTases.....................................................22 Classifications and Characterizations.........................................................................25 Nomenclature......................................................................................................25 Type I Systems....................................................................................................27 Type I systems general characteri stics: Physiology and function...............27 Type I systems: Regulation.........................................................................33 Type II Systems...................................................................................................37 Type II systems general characteri stics: Physiology and function.............37 Type II systems: Regulation........................................................................44 Type III Systems..................................................................................................47 Type III systems general characteristics: Physiology and function............47 Type III systems: Regulation......................................................................50 Type IV Systems.................................................................................................50 Type IV systems general characteri stics: Physiology and function............50
v Type IV systems: Regulation......................................................................54 Unclassified Systems...........................................................................................54 Homing Endonucleases.......................................................................................56 Solitary DNA Methyltransferases.......................................................................58 Dam and Dcm of Escherichia coli................................................................58 CcrM: A cell cycle regulating MTase.........................................................64 Solitary antirestriction MTases....................................................................70 Evolution and the Taxonomic Relationships..............................................................71 Type I Systems....................................................................................................73 Type II Systems...................................................................................................74 Type III Systems..................................................................................................77 DNA MTases.......................................................................................................78 Homing Endonucleases.......................................................................................79 R-M Systems: Mobility Aiding Evolution.........................................................80 Maintenance of R-M Systems.............................................................................81 R-M Systems: â€œSelfish genetic unitsâ€.........................................................81 R-M Systems: A promoter of recombination..............................................85 R-M Systems: Cellular defense and maintenance of species identity.........88 3 STUDY RATIONALE AND EXPE RIMENTAL APPROACH...............................90 4 CLONING AND CHARACTERIZATION OF A METHYL-DEPENDENT RESTRICTION ENDONUCLEASE FROM Zymomonas mobilis SUBSPECIES mobilis CP4.................................................................................................................92 Introduction.................................................................................................................92 Materials and Methods...............................................................................................94 Bacterial Strains, Plasmids, Media, and Reagents..............................................94 DNA Isolation.....................................................................................................95 Construction of and Ap propriate Cloning Vector...............................................95 Isolation of Zymomonas mobilis CP4 Genomic DNA........................................96 Preparation of a Zymomonas mobilis CP4 Genomic Library..............................97 Transformation....................................................................................................98 Screening a Zymomonas mobilis CP4 library for REases and DNA MTases.....99 Sequencing and Analysis of Positive Clones......................................................99 Primer Extension Analysis of ZmCP4mrr mRNA............................................100 Recombinant Expression of the ZmCP4 mrr .....................................................102 Protein purification of ZmCP4Mrr....................................................................102 Transformation Studies in Zymomonas mobilis CP4........................................103 BstNI Restriction Analysis of Zymomonas mobilis CP4 DNA.........................104 Isolation and Sequence Analysis of Zymomonas mobilis CP4 Clones..............105 Primer extension and promoter sequence analysis of ZmCP4mrr ....................108 Transformation Studies of Recombinant ZmCP4mrr ........................................108 Partial purification of ZmCP4Mrr.....................................................................109 Discussion.................................................................................................................109
vi 5 CLONING AND CHARACTERIZATION OF A CELL CYCLE REGULATING DNA METHYLTRANSFERASE FROM Zymomonas mobilis SUBSPECIES mobilis CP4...............................................................................................................125 Introduction...............................................................................................................125 Materials and Methods.............................................................................................127 Bacterial Strains, Plasmids, Media, and Reagents............................................127 DNA Preparation and Screening of Zymomonas mobilis CP4 Clones..............127 Primer Extension Analysis of ZmCP4ccrm mRNA..........................................128 Recombinant Expression of the ZmCP4CcrM Protein.....................................128 Purification of a Recombin ant ZmCP4CcrM Protein.......................................129 Analysis of the Methylation Specific ity of ZmCP4CcrM Using the Type IIS Restriction Endonuclease BsaI......................................................................131 Construction of pBROriV ROP/ ZmCP4ccrM and pBBR1MCS/ ZmCP4ccrMPRO for Overexpression of ZmCP4ccrM in Zymomonas mobilis CP4...............................................................................132 Examination of Cell Morphology and DNA Distribution Using Microscopy..134 Results.......................................................................................................................135 Identification of a Z. mobilis CP4 DNA MTase ORF and Sequence Analysis.135 Primer Extension and Promoter Sequence Analysis of ZmCP4ccrM ...............138 In vivo ZmCP4CcrM activity: HinfI digestion of Z. mobilis CP4 DNA...........139 In vivo Activity of Recombinant ZmCP4CcrM.................................................140 Recombinant ZmCP4CcrM Protein Purification...............................................142 In vitro Activity of Recombinant ZmCP4CcrM................................................143 Determination of the Nucleotide Methylated by ZmCP4CcrM........................144 Effect on Growth and Cell Viability of ZmCP4ccrM Overexpression in Z. mobilis CP4....................................................................................................145 Effect of Overexpression of Zmcp4ccrm in Zymomonas mobilis CP4 on Cell Morphology....................................................................................................146 Discussion .................................................................................................................148 6 GENERAL SUMMARY AND DISCUSION..........................................................184 LIST OF REFERENCES.................................................................................................189 BIOGRAPHICAL SKETCH...........................................................................................206
vii LIST OF TABLES Table page 1 Bacterial strain, plasmid, and primer description...................................................116 2 Dependency of transformation efficiency of Z. mobilis CP4 on plasmid source...119 3 Transformation of cloned ZmCP4mrr into E. coli with different MTase backgrounds...........................................................................................................124 4 Bacterial strains, plasmids, and primers used in ZmCP4CcrM study....................160 5 Determination of the methylation site of ZmCP4CcrM using Type IIs BsaI........178 6 Effect of ZmCP4ccrM overexpression on cell viability.........................................180
viii LIST OF FIGURES Figure page 1 Proposed CcrMI autoregulation and regulation of CtrA expression in Caulobacter crescentus ............................................................................................66 2 pP3226 construct....................................................................................................118 3 Restriction analysis of DNA isolated from Zymomonas mobilis CP4...................120 4 Identification of the ORFs of a cloned Z. mobilis CP4 region containing ZmCP4mrr ..............................................................................................................121 5 REase sequence: ZmCP4mrr .................................................................................122 6 Primer extension analysis of ZmCP4mrr ...............................................................123 7 pBROriV Rop construct........................................................................................162 8 DNA sequence and predicte d amino acid sequence of: ZmCP4ccrM ...................163 9 Identification of the ORFs of a cloned Z. mobilis CP4 region containing ZmCP4ccrM ...........................................................................................................165 10 Amino acid sequence alignment of CcrM homologues.........................................166 11 Primer extension and promoter sequence analysis of ZmCP4ccrM .......................168 12 Promoter region of CcrM homologues compared to ZmCP4ccrM ........................169 13 In vivo activity of recombinant ZmCP 4CcrM from pPROTet.E133/ ZmCP4ccrM ...........................................................................................................170 14 In vivo activity of recombinant Zm CP4CcrM in cultured GM4715/ pBBR1MCS/ ZmCP4ccrMPRO ..............................................................................171 15 In vivo activity of recombinant ZmCP4C crM in cultured HMS174(DE3)/pET24b/ ZmCP4ccrM ....................................................................................................172 16 ZmCP4CcrM protein purification and analysis on SDS-PAGE from BL21Pro/ pPROTet.E133/ ZmCP4ccrM transformants...........................................................173
ix 17 ZmCP4CcrM protein purification and analysis on SDS-PAGE from HMS174(DE3)/pET-24b/ ZmCP4ccrM transformantsB-II)...................................174 18 I n vitro activity of partially purified r ecombinant ZmCP4CcrM extracted from cultured BL21Pro/pPROTet.E133/ ZmCP4ccrM ....................................................175 19 In vitro activity of purified recombinant ZmCP4CcrM extracted from cultured HMS174(DE3)/pET-24b/ ZmCP4ccrM ..................................................................176 20 Analysis of the methylation specificity of ZmCP4CcrM.......................................177 21 Growth curves of Zymomonas mobilis CP4 with or without plasmid borne copies of ZmCP4ccrM ............................................................................................179 22 Microscopy of Zymomonas mobilis CP4 overexpressing ZmCP4ccrM at 200X magnification..........................................................................................................181 23 Microscopy of Zymomonas mobilis CP4 overexpression ZmCP4ccrM at 400X magnification..........................................................................................................182 24 Unstained and DAPI stained Zymomonas mobilis CP4 transformants overexpressing ZmCP4ccrM ..................................................................................183
x Abstract of Dissertation Pres ented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy CLONING AND CHARACTERIZATION OF A METHYL-DEPENDENT RESTRICTION ENDONUCLEASE AND A CELL CYCLE REGULATING DNA METHYLTRANSFERASE FROM Zymomonas mobilis SUBSPECIES mobilis CP4 By Priscilla Lorraine Phillips December 2005 Chair: Francis C. Davis Major Department: Micr obiology and Cell Science A Zymomonas mobilis CP4 genomic library was screened using the E. coli indicator strains AP1-200-9 and ER1992 to isolate clones of enzymes that cause DNA damage. Sequence analysis of positive clones identified two open reading frames encoding DNA modification enzymes: (1) a 92 4 base pair open reading frame with sequence similarity to mrr , a methyl-dependant restri ction endonuclease, which was designated ZmCP4mrr , and (2) a 1149 bp open reading fr ame with amino acid sequence similarity to ccrM , a cell cycle regulating DNA methyltr ansferase, which was designated ZmCP4ccrM. Sequence analysis indicates that ZmCP4mrr is a solitary methyl-dependent restriction endonuclease without a cognate DNA methyltransferase. Transformation of Escherichia coli K12 strains with various DNA methyltransferase backgrounds demonstrated that a plasmid borne ZmCP4mrr gene readily transforms E. coli strains that express dcm, hsdM , and EcoKccrM DNA methyltransferases, indicating that ZmCP4Mrr
xi does not recognize and restrict sites methylated by thes e DNA methyltransferases. E. coli strains that express the dam DNA methyltransferase coul d only be transformed if expression of plasmid borne ZmCP4mrr was repressed. Subs equent induction of ZmCP4mr r expression in these cells resulted in inhibition of grow th and cell death, indicating that ZmCP4Mrr spec ifically restricts Dam N6-a denine methylated DNA (5â€™GmATC-3â€™). Plasmid DNA originating from dam deficient E. coli strains did not improve transformation effi ciency, indicating that Z. mobilis CP4 has a restriction system in addition to ZmCP4Mrr contributing to low frequency of gene transfer from foreign DNA. Sequence analysis indicates that ZmCP4ccrM is a solitary DNA methyltransferase with two possible in-frame translation initiation sites. A ribosomal binding site containing a sequence, 5 â€™-AGGA-3â€™, conserved in Z. mobilis promoters of highly expressed genes is located adjacent to the fi rst possible translation initiation site and not the second, suggesting that ZmCP4ccrM is being expressed from the first translational initiation site. The specificity for ZmCP4Ccr M methylation was directly determined to be the N6-adenine of its r ecognition site 5â€™-GANTC-3â€™. Z. mobilis CP4 cells overexpressing ZmCP4ccrM exhibited a subpopulation of filamentous cells, ranging from 10-90 M in length, with multiple chromosomes. Overexpression of ZmCP4ccrM caused disruption of normal cell division and ch romosomal segregation, suggesting that ZmCP4CcrM is involved in cell cycle regulation.
1 CHAPTER 1 ZYMOMONAS MOBILIS Introduction In the early seventies, the first energy cr isis triggered biotechnological research into renewable resources, such as microbial production of fuel ethanol from biomass using recombinant microorganisms such as Saccharomyces cerevisiae, Zymomonas mobilis , and Escherichia coli (Montenecourt, 1985; Crueger and Crueger, 1989; Doelle et al ., 1993; Danner and Braun, 1999; DiPardo, 2000; Willke and Vorlop, 2004). Despite the incredible rate at which mol ecular biotechnology has grown in the past thirty years, the relatively low cost of fossil fu els, as compared to producti on of renewable resources, has curbed research and stymied any significant repl acement of petroleum based products in the U.S. with bio-based fuels (Bothast et al., 1999; Danner and Braun, 1999; DiPardo, 2000; Willke and Vorlop, 2004). Four growing problems have reinvigorated efforts in researching and implementing the replacem ent of petroleum based products with renewable resources in the last decade: (1) th e depletion of fossil fuels coupled with ever increasing demand, (2) the increasing cost of imported oil, (3) the disposal of increasingly large amounts of waste produced each year, and (4) the damage to environmental and human health cau sed by the increas ing rate of CO2 emissions. Although yeast ( Saccharomyces cerevisiae ) strains have traditionally been used as a commercial source of ethanol, Zymomonas mobilis has various advantages over Saccharomyces cerevisiae as a microbial ethanol producer such as (1) a higher specific growth rate, (2) a greater specific ethanol productivity, (3) a lower production of other
2 byproducts, (4) a higher rate of glucose uptake, and (5) a greater to lerance to high sugar concentrations, ethanol, and other waste products that inhibi t continuous ethanol fermentation in yeast (Crueger and Crueger, 1989; Sahm et al. , 1992; Doelle et al ., 1993; Gunasekaran and Raj, 1999). The main disadvantage Zymomonas mobilis has as a biocatalyst for ethanol producti on is the narrow range of subs trates it can metabolize: glucose, fructose, and sucrose (Sahm et al. , 1992; Doelle et al. , 1993; Gunasekaran and Raj, 1999; Aristidou an d Penttila, 2000) Identification and Taxonomy Zymomonas was most likely first identified in 1912 by Baker and Hillier as the cause of â€œcider sicknessâ€ (unnamed Strain A) but the isolate had not been maintained (Swings and De Ley, 1977; Montenecourt, 1985). The identification of the named organism is generally attributed to Linder, who first isolated the organism from agave plant sap in Mexico in 1923 (Swi ngs and De Ley, 1977) and named it Thermobacterium mobile (Montenecourt, 1985). In 1936, Kluyver and van Niel designated the organism the genus name of Zymomonas â€œfor polarly flagellated bacteria causing alcoholic fermentationâ€ (Swings and De Ley, 1977). It was reclassified and renamed many times and was not â€œsettledâ€ into th e current nomenclature of Zymomonas mobilis until the late 1970â€™s (Montenecourt, 1985). Zymomonas belongs to the rRNA superfamily C (Ingram et al. , 1989) and alpha subdivision of Gram-negative bacteria (NCBI Taxonomy Browser). The genus Zymomonas is presently considered to have only one species ( Zymomonas mobilis ) and two subspecies, mobilis and pomaceae (Swings and De Ley, 1977; Gunasekaran and Raj, 1999) . The ability to utilize sucros e as a carbon source primaril y distinguishes the former from the later (Gunasekaran and Raj, 1999). Zymomonas mobilis subspecies pomaceae
3 can also be distinguished by the fact that st rains of this subspeci es cannot grow above 34C; while strains of the subspecies mobilis can grow at 36C (some strains as high as 42C) (Swings and De Ley, 1977). Z. anaerobia can be found listed as another species of Zymomonas in some reports; however, high phenotypic similarity suggests that these are different strains of the subspecies mobilis (Swings and De Ley, 1977) . With cells grown under standard conditions, comparisons of prot ein gel electropherograms of three strains of Z. mobilis pomaceae show a pattern distinct from the other Z. mobilis strains tested (including Z. anaerobia strains), which all had a sim ilar pattern (Swings and De Ley, 1977). Genetically, Zymomonas mobilis subspecies pomaceae (ATCC 29192), when compared to strains of Zymomonas mobilis subspecies mobilis , has a genome sequence similarity of less than 32% D (D=duplex fo rmation) based on the renaturation rate method performed under stringent conditions (Swings and De Ley, 1977). All the other strains of Z. mobilis tested (40, including Z. anaerobia ) had at least 76% D genomic DNA sequence similarity (Swings and De Ley, 1977). A recent study suggests that members of the Zymomonas subspecies mobilis and pomaceae can be quickly distinguished from each other by restric tion digestion analysis of amplified 16S rRNA DNA (Coton et al. , 2005). The genome of Zymomonas is relatively small, less than half the size of the genome of Escherichia coli (Swings and De Ley, 1977). The single circular genome of Zymomonas mobilis subspecies mobilis ZM4 is 2056416 bp with a 46.33% G+C content (Seo et al. , 2005). Z. mobilis subspecies mobilis has two plasmid profiles (Yablonsky et al. , 1988). The first group has the same f our large native plasmids (31.5 kb, 32.5 kb, 33 kb, and 35 kb) and members of this group are designated CP4 strains (Yablonsky et al. ,
4 1988). The second group has the same three large native plasmids (32.5 kb, 34 kb, and 40.5 kb), and members of this group ar e designated ZM4 strains (Yablonsky et al. , 1988). Zymomonas has been identified as being genetically, phenotypically, and ecologically related to the acetic acid bacteria Gluconobacter and Acetobacter (Swings and De Ley, 1977; Sahm et al. , 1992). It seems to more closely resemble Gluconobacter because of its polar flagella, incomplete tric arboxylic acid cycle, and the occurrence of the Entner-Doudoroff pathway, suggesting that Zymomonas and acetic acid bacteria may have evolved from a common aerobic an cestor (Swings and De Ley, 1977; Sahm et al. , 1992). Hybridization studies of rRNA have supported this theory (Sahm et al. , 1992). Recent amplified 16S rRNA DNA restriction analysis (ARDRA) places Spingomonas as a closely related organism to Zymomonas , followed by the more distantly related Gluconobacter and Acetobacter (Coton et al. , 2005). General Physiology Zymomonas is a rod shaped Gram-negative obligatively fermentative bacterium, which lacks spores or a capsul e and produces ethanol 3 to 4 tim es more rapidly than yeast (Doelle et al ., 1993; Glazer and Nikaido, 1995). Anal ysis of 38 different strains of Zymomonas (â€œwild typeâ€ isolates) revealed that the typical size range of cells are 2-6 m long and 1-1.4 m wide and are arranged singly but mo stly in pairs (Swings and De Ley, 1977). Some strains of Zymomonas also produce rosettes, ch ains, and/or filamentous cells (Swings and De Ley, 1977); however, the growth conditions and proportion of variant cells observed for these stra ins have not been described. Two Zymomonas strains were found to also produce U-shaped cel ls (Swings and De Ley, 1977). If motile, Zymomonas has one to four polar flagella (Swi ngs and De Ley, 1977). Motility is not considered essential because most Z. mobilis strains are not motile, and motility can be
5 lost spontaneously (Sahm et al. , 1992; Swings and De Ley, 1977). The optimal pH for growth is generally abou t 7.3, although most strains can grow between pH 3.8-7.5 (Swings and De Ley, 1977). Its optimal growth temperature is generally between 25-31 C, with slow growth below 15 C a nd above 34 C for most strains (Sahm et al. , 1992; Swings and De Ley, 1977). Zymomonas is tolerant to high su gar concentrations where most strains can grow in 30% glucose and more than half the strains tested can grow in 40% glucose (Swings and De Ley, 1977); however, Z. mobilis has a low tolerance to inorganic salts (Vrieselkoop et al. , 2002). When the growth of 38 strains of Z ymomonas in media containing various concentrations of NaCl was examined, 2 strains could not grow at 0.5% NaCl, 39% of the strains could not grow at 1% NaCl, and none could grow at 2% NaCl (Swings and De Ley, 1977). More recent studies showed that chloride ions inhibited growth and resulted in f ilamentous cells (at 0.175 M NaCl or NH4Cl) (Vrieselkoop et al. , 2002). When comparing growth in batch cultures containing NaCl, Na2SO4, or NH4Cl, sodium ions were shown to have a much greater inhibitory effect on growth than chlorine ions, however, the result s indicated that occurr ence of filamentous cells was due the presence of chloride ions, not the sodium ions (Vrieselkoop et al. , 2002). All strains of Zymomonas seem to require pantothenate for growth, most require biotin, and some require a dditional growth factors such as thiamine, vitamin B12, riboflavin, folic acid, lipoic acid, cyanoc obalamine etc. (Swings and De Ley, 1977). Most strains of Zymomonas are resistant to a wide array of antibiotics (Swings and De Ley, 1977; Sahm et al. , 1992). Of 40 strains tested fo r antibiotic resist ance, all were found to be resistant to streptomycin, nalid ixic acid, bacitracin, gentamycin, kanamycin,
6 neomycin, polymyxin, penicillin, and lincomycin (Swings and De Ley, 1977; Sahm et al. , 1992). Various percentages of the Z. mobilis strains tested were resistant to methicillin (>95%), erythromycin (~90%), nitrofur antoin (~75%), vancomycin (~50%), cephaloridine (~30%), ampicillin (~40%), and chloramphenicol (~5%) (Swings and De Ley, 1977; Sahm et al. , 1992). Of the antibiotics and th e 40 strains tested, most strains were found to be sensitive to chloramphenicol, tetracyclin e, rifampicin, fusidic acid, novobiocin, and sulphafurazole (Swings and De Ley, 1977; Sahm et al. , 1992). Spontaneous mutants may be responsible for sensitivity or resistance to one or more of the antibiotics listed (Swi ngs and De Ley, 1977; Sahm et al. , 1992). Metabolism Zymomonas is considered a â€œhomofermentativ eâ€ bacterium that produces ethanol and CO2 as virtually the only ferm entative products when in an anaerobic environment (Neidhardt et al. , 1990 p.164). The viability of Z. mobilis rapidly decreases after all the carbon source is metabolized, primarily due to the rapid degradation of RNA (Swings and De Ley, 1977). The amount of RNA per cell falls to as little as 3-5% of its original level after one day, and is the only cel lular constituent that is si gnificantly broken down during starvation (Swings and De Ley, 1977). After one day of starvation, 90% of the cells rapidly die and the remainder continues to die off slowly (Swings and De Ley, 1977). Cell survival is improved in the presence of 33 mM MgCl2, which reduced cell death to only 40% initially, due to suppression of RN A degradation, but even tually the cells all die off, typically within 7 days (Swings and De Ley, 1977). Zymomonas is an aerotolerant anaerobe that metabolizes carbohydrates by the way of the Entner-Doudoroff pathway followed by pyruvate decarboxylation (Swings and De Ley, 1977; Crueger and Cr ueger, 1989; Sahm et al. , 1992; Gunasekaran and Raj, 1999).
7 In the 1950â€™s, Zymomonas became the first anaerobic organism discovered to use the Entner-Doudoroff pathway, a pathway that occu rs mainly in strict aerobes (Swings and De Ley, 1977). Currently, Zymomonas is the only genus known to exclusively use the Entner-Doudoroff pathway anaerobically (Aris tidou and Penttila, 2000). The main effect oxygen has on the growth of Zymomonas is a reduced ethanol yi eld due to the oxidation of some of the ethanol to byproducts such as acetaldehyde, acetic acid, and acetone (Swings and De Ley, 1977; Doelle et al ., 1993; Gunasekaran and Raj, 1999). Zymomonas contains a number of other enzymes commonly associated with oxidative metabolism. Four enzymes of th e tricarboxylic acid (TCA) cycle have been detected in Z. anaerobia : citrate synthase, aconitase, isocitrate dehydrogenase, and malate dehydrogenase (Swings and De Ley, 1977) . The activities of all four of these enzymes are not increased by aeration (Swings and De Ley, 1977). As is typical for Gram-negative bacteria, Z. mobilis has a large molecular weight citrate synthase; however, unlike the Gram-negative bacter ia and like the Gram-positive bacteria, Z. mobilis citrate synthase is not inhibi ted by NADH (Swings and De Ley, 1977). Zymomonas also contain cytochrome c and cytochrome oxidase 2, constitutively produced under aerobic and an aerobic conditions, and cytochrome b, typically produced only under anaerobic conditions (Swings and De Ley, 1977). A membrane bound glucose dehydrogenase, contai ning a pyrroloquinoline-quinone prosthetic group, has also been detected in Z. mobilis (Strohdeicher et al. , 1988; Sahm et al. , 1992). Quino-protein dehydrogenases are usually found in aerobic or ganisms to feed electrons into the respiratory chain (Duine et al. , 1986). Z. mobilis probably respires through a complete respiratory chain; however, transfer of electro ns via the respiratory chain is not coupled
8 to oxidative phosphorylation (Gunasekaran a nd Raj, 1999). Glucos e-fructose oxidoreductase and glucose dehydrogenase leads to the formation of gluconic acid, which enters the Entner-Doudoroff pathway afte r phosphorylation (Gunasekaran and Raj, 1999). Under aerobic conditions, Z. mobilis oxidizes NADH and NADPH, catalyzed by membrane bound oxidases (Bringer et al. , 1984; Sahm et al. , 1992; Gunasekaran and Raj, 1999). It was observed th at aerated and starved Z. mobilis cells generate a transmembrane pH gradient where ATP s ynthesis is apparently coupled to NADH oxidation (Kalnenieks et al. , 1993). Under aerobic conditio ns, part of the NADH and ethanol generated in the Entner-Doudoroff pathway is oxidized by respiration forming byproducts (Kalnenieks et al. , 1993). The Entner-Doudoroff pathway produces only one ATP per glucose fermented, instead of the two ATPs produced by the Embeden-Meyerhof pathway (Crueger and Crueger, 1989). Consequently, the amount of glucose that is converted to cell mass rather than ethanol and CO2 is reduced (Crueger and Crueger, 1989; Sahm et al. , 1992). When Z. mobilis is grown on glucose, the ke y enzymes for ethanol production, pyruvate decarboxylase (Pdc) and alcohol dehyd rogenase (Adh), compose 5% and 2-5% of the soluble proteins, respectively (Doelle et al ., 1993; Gunasekaran and Raj, 1999). Pdc rarely occurs in bact eria and is found mainly in yeast and fungi (Doelle et al ., 1993). The Z. mobilis Pdc is unique in that it requires thiamine for activity (Gunasekaran and Raj, 1999). Z. mobilis has two alcohol dehydrogenases, AdhI and AdhII, which are unrelated isoenzymes (Gunasekara n and Raj, 1999; Kalnenieks et al. , 2002). AdhI is zinc dependent, as is its is ofunctional counterpart in yeast (Gunasekaran and Raj, 1999). AdhII is unusual in that it is iron dependent rather than zinc dependent (Gunasekaran and
9 Raj, 1999; Kalnenieks et al. , 2002). AdhI decreases in conc entration in stationary phase, unlike AdhII (An et al. , 1991). AdhII protein production was shown to be induced in response to heat and ethanol stress (An et al. , 1991). Consequently, the function of AdhII was originally believed to facilitate the continued fermentation of sugars at high concentrations of ethanol (Gunasekaran and Ra j, 1999). Recent studi es show that AdhI reduces acetaldehyde to ethanol faster th an it can oxidize etha nol to acetaldehyde, reaching its maximal rate at pH 6.5 (~physiological pH of Z. mobilis ), while AdhII oxidizes ethanol to acetaldehyde faster than it can reduce acetaldehyde to ethanol, with an alkaline optimal pH (Kalnenieks et al. , 2002). It appears that A dhII functions to oxidize part of the ethanol produced by AdhI, ther eby supplying NADH to the respiratory chain during aerobic conditions (Kalnenieks et al. , 2002). In summary, Zymomonas appears to have fragments of a TCA cycle, a fully functional electron transport chain, and undergoes changes in metabolism when switched from an anaerobic to an aerobic environmen t, producing organic byproducts other that ethanol and CO2 (Swings and De Ley, 1977; Gunasekaran and Raj, 1999; Kalnenieks et al. , 2002; Doelle et al ., 1993). Consequently, Zymomonas undergoes â€œaerobic fermentationâ€ but not aerobic respirat ion in the presence of oxygen (Doelle et al ., 1993). The occurrence of pathways and enzymes found mainly in strict aerobes, such as the Entner-Doudoroff pathway and the inco mplete TCA cycle, suggest that Zymomonas may have originated from aerobic ance stors (Swings and De Ley, 1977; Sahm et al. , 1992). Zymomonas is also known for a unique characteristic, its ability to uncouple catabolism from anabolic growth to produce le ss cell mass than exp ected per mole of glucose from the Entner-Doudoroff pathway (Sahm et al. , 1992; Doelle et al ., 1993).
10 Glucose is catabolized at a constant rate of one ATP/glucose, appa rently independent of the state of biosynthesis or composition of the media (Swings and De Ley, 1977). Mechanisms to maintain tight coupling of anabolism to catabolism, like those found in most microorganisms such as E. coli and yeast, must function poorly or do not exist in Zymomonas (Swings and De Ley, 1977). Some of th e factors believed to be required for uncoupled growth in Zymomonas are a lack of pantothenate (required for optimal growth) (Swings and De Ley, 1977), and nitrogen, phos phorus, or potassium limitation (Doelle et al ., 1993). Under suitable conditions ( i.e ., available pantothenate , optimal temperature, etc.), the biosynthetic pathways will use the ATP produced by catabolism (Swings and De Ley, 1977). Under unsuitable conditions ( i.e ., in synthetic or minimal media, low pH, in the presence of some antibiotics such a chlo ramphenicol, etc.), growth will decrease or stop, but the cell will continue to make AT P and waste products (like ethanol) until all metabolizable carbon sources are exhaus ted (Swings and De Ley, 1977). When Zymomonas is grown in complex media, such as y east extract, the organic constituents of the media are primary used as a source of carbon in biosynthesis and not used as a source of energy (Swings and De Ley, 1977). Specifica lly, the amount of glucose converted to cell mass can be as low as 2% and the rest is converted primarily to ethanol and CO2 (under optimal growth conditions) (Swings and De Ley, 1977; Crueger and Crueger, 1989; Sahm et al. , 1992). The glucose incorp orated into cell mass in Zymomonas produces only 48% of the cellu lar carbon; the rest is deri ved from the carbon in yeast extract and/or other organic components of the media (Swings and De Ley, 1977). Z. mobilis prefers an inorganic source of nitroge n, such as ammonium sulfate, over the
11 organic nitrogen provided by yeast extract, fo r a maximal growth rate with anabolism coupled to catabolism (Veeramallu and Agrawal, 1988). Potential Commercial Applications of Zymomonas mobilis Z. mobilis has potential to be used as a commer cial producer of levan, a polymer of fructose (Gunasekaran and Raj, 1999). Levan is produced by Z. mobilis when grown on sucrose at pH 5 and temper atures no higher than 34 C (Doelle et al ., 1993). Levan is used in pharmaceutical production, as a thickening/suspending agent, and as a blood plasma substitute (Doelle et al ., 1993; Gunasekaran and Raj, 1999). Microbial acetaldehyde, produced in abundance by Z. mobilis when grown in the presence of oxygen, is also of potential intere st as flavorings in the food industry (Ingram et al. , 1989; Sahm et al. , 1992). Because NADH availabil ity decreases when oxygen is present, less of the acetaldehyde is reduced to ethanol and acetaldehyde accumulates (Bringer et al. , 1984; Sahm et al. , 1992). Under aerobic conditi ons, ethanol is oxidized to acetaldehyde, where acetaldehyde accumulation in culture increases with increasing aeration (from 0.1 to 2.0 g L-1) (Kalnenieks et al. , 2002). Acetaldehyde is currently produced by catalytic dehydration or oxidati on of ethanol, but because the boiling point of acetaldehyde is 20.8 C (making separation of acetaldehyde from the fermentation broth easy and cheap) direct microbial productio n of acetaldehyde is a viable commercial possibility (Danner and Braun, 1999). Gluconic acid and sorbitol, produced by Z. mobilis by way of the glucose-fructose oxidoreductase reaction, are metabolites of commercial interest because they are intermediates to the production of ascorbic acid (vitamin C) (Sahm et al. , 1992; Glick and Pasternak, 1998; Silveira and Jonas, 2002) . Traditionally produced by catalytic
12 hydrogenation of D-glucose, one quarter of th e industrially produced sorbitol is used to synthesize vitamin C (Silveira and Jonas, 2002) . Normally, phosphorylated gluconic acid enters the Entner-Doudoroff path way to produce ethanol (Sahm et al. , 1992; Gunasekaran and Raj, 1999). Z. mobilis cells permeabilized with r eagents such as CTAB are no longer able to convert gluconic acid to ethanol and instead produce high yields of gluconic acid and sorbitol at rates that may be as high as 2.1 and 1.8 g/liter hour, respectively (Silveira and Jonas, 2002). Untreated cells grown under the osmotic pressure of 650g/l of equimolar glucose-fruc tose substrate (which causes loss of cell viability and disrupts normal metabolism), produ ce similar high concentrations and yields of gluconic acid and sorbitol (S ilveira and Jonas, 2002). So rbitol is also industrially important as a sweetener, humectant, soften er, texturizer, and in the production of pharmaceuticals, synthetic plasticizers, alkyd resins, and other products (Silveira and Jonas, 2002). Biotechnological production of sorbitol and gluconic acid is not economically practical in the United States, b ecause the cost of these products is low compared to the cost of the substrates (glu cose and fructose) and the cost of current methods of bioconversion (Silveira and Jonas, 2002). In countries like South America, where the cost of sorbitol and gluconic acid are considerably higher, bioconversion of glucose and fructose to gluconic acid and sorb itol may be an attractive option (Silveira and Jonas, 2002). Zymomonas is naturally found in various pl ant saps and honey (Swings and De Ley, 1977; Ingram et al. , 1989). It is commonly involved in the production of alcoholic beverages in tropical areas, particularly in the production of palm, agave, and sugarcane wines (Swings and De Ley, 1977; Sahm et al. , 1992; Gunasekaran and Raj, 1999).
13 Therefore, Zymomonas is of some commercial intere st for the development of new beverages, and in the indust rialization of trad itional ones (Swings and De Ley, 1977; Sahm et al. , 1992; Gunasekaran and Raj, 1999). Zymomonas is an organism suitable for production of products for human consumption since it has been shown to be harmless to humans; in fact, there are reports of Zymomonas being of therapeutic use in cases of chronic enteric and gynecological infections due to its antagonistic effect against a number of bacteria and filament ous fungi (Swings and De Ley, 1977). Under optimal conditions, Z . mobilis metabolizes car bohydrates with CO2 and ethanol as virtually the only waste products (Gunasekaran and Raj, 1999). Z. mobilis can theoretically produce as much as 2 moles of ethanol and 2 moles of CO2 gas per mole of fermentable sugar (Swings and De Ley, 1977). Of the 40 different Zymomonas strains tested, each strain produced at least 1.5 moles of ethanol per mole of glucose (Swings and De Ley, 1977). Because Zymomonas can produce at least 97% of the theoretically expected ethanol from fermentable sugars, it is a potential bio catalyst for industrial ethanol production (Zhang et al. , 1995; Aristidou and Penttila, 2000) . In the interest of finding alternative sources for the producti on of fuels and other organic chemicals, Z. mobilis has been the subject of numerous gene tic and biochemical studies designed to enhance its suitability as a bi ocatalyst for ethanol production. An Alternative Source of Fuel Etha nol: Advantages and Disadvantages Microbial production of fuel ethanol from biomass utilizes a renewable resource, unlike fossil fuels which are being rapidly de pleted. The main obstacle in lowering the cost of microbial ethanol produc tion is the cost of the car bohydrate substrates that are currently used (Danner and Braun, 1999; Di pardo, 2000). One study estimated that at least 50% of the cost of producing microbial et hanol in the German ma rket is the cost of
14 the substrate (Willke and Vorlop, 2004). A potentially less expensive source of substrate is the complex carbohydrates present in bioma ss. An estimated 2.45 billion metric tons of biomass is available per year in the Un ited States, representing a potential 270 billion gallons of ethanol per year (Aristidou and Penttila, 2000). Various biomass feedstocks include agriculture and wood residues, food manufacturing a nd solid municipal/consumer waste, agricultural surpluses, and crops specifi cally grown to be used as a feedstock for fuel and chemical production (Crueger and Crueger, 1989; Danner and Braun, 1999; Aristidou and Penttila, 2000; Dipardo, 2000). The use of cellulose-based feedstock (wood, straw, and other plant waste) for th e microbial production of ethanol has been getting much attention from the U.S. De partment of Energy (Danner and Braun, 1999; Dipardo, 2000). The current obstacle in reducing the cost of microbial ethanol production from cellulose-based feedstock is the cost of hydrolyz ing cellulose into fermentable sugars (Danner and Braun, 1999; Dipardo, 2000). The U.S. Department of Energyâ€™s goal is to reduce the cost of micr obial ethanol production from cellulose based feedstock by using genetically engineered microorganisms (D ipardo, 2000), in particular, recombinant Zymomonas mobilis , Escherichia coli , and Saccharomyces cerevisiae (Danner and Braun, 1999). Most of the ethanol currently produced in the United States is produced by yeast fermentation (Danner a nd Braun, 1999). In 2004, the United States had the capacity to produce 3.41 billion gallons of bio-ethanol per year, most of which is used as a fuel additive for 30% of the gasoline sold in the United States (Ethanolrfa.org). In Hawaii and Minnesota, the law requires the addition of ethanol to all gasoline sold within the state (Ethanolrfa.org). While so me envision a 90% replacement of fossil fuels with renewable resources by 2090, it has been recently predicted that the United States
15 will most likely replace fossil fuels with renewable resources by at least 25% by 2030 (Willke and Vorlop, 2004). The EU (European Union) directed that a minimum of 5.75% of all transportation fuels must be bio-fuels by 2010 (Willke and Vorlop, 2004). In the long-run, with continued depletion of fossil fuels and ever increasing demand, and the present rate of biotechnol ogical advances in genetic e ngineering of microorganisms (such as Z. mobilis ) to produce clean burning fuel etha nol from more cost effectively from biomass, ethanol will become a major energy resource worldwide. Zymomonas has several advantages as a biocat alyst of industrial ethanol production when compared to the yeast Saccharomyces cerevisiae . Zymomonas has a 2.4 times higher specific growth rate and at least 2.9 ti mes higher ethanol formation rate (g/Gal h) than the yeast Saccharomyces (Crueger and Crueger, 1989). Zymomonas produces about five times less nonethanol byproducts than y east (Gunasekaran and Raj, 1999). Unlike yeast, Zymomonas does not require initial aerobic growth to maximize cell mass nor controlled addition of oxygen dur ing fermentation to avoid yi eld loss due to decrease in cell mass (Crueger and Crueger, 1989; Guna sekaran and Raj, 1999). In yeast, tight coupling of anabolism and catabolism means that high cell concentrations are required for high ethanol yields (Doelle et al ., 1993). In contrast, Z. mobilis has the ability to uncouple catabolism and anabolic growth, a ch aracteristic that may be exploited to maximize ethanol production (Doelle et al ., 1993). Uncoupling catabolism and anabolic growth, counterbalanced with maintaining su fficient cell mass, means increased ethanol yield because less of the carbon source is directed towards biosynthesis (Doelle et al ., 1993). Zymomonas transports sugars by high velocity, low affinity facilitated diffusion, rather than active transport, at a Vmax of 300 nmol/min/mg protein in strain CP4 (Doelle et
16 al ., 1993). Consequently, Z. mobilis has 2.6 times higher glucose uptake rate than yeast and is ideally suited to grow in high suga r concentration environments (Crueger and Crueger, 1989; Doelle et al ., 1993). Zymomonas has greater osmotolerance to high sugar concentrations than yeast (Swings and De Ley, 1977; Crueger and Crueger, 1989; Sahm et al. , 1992). Most Zymomonas strains can grow in media with up to 40% glucose (Swings and De Ley, 1977; Crue ger and Crueger, 1989; Sahm et al. , 1992). Zymomonas has a greater tolerance to et hanol, at least 12% and up to 16.5% wt/vol, compared to the typical 1-2% wt/vol of other bacteria, and the 5-6% wt/vol of yeast (Scopes and GriffithsSmith, 1986; Crueger and Crueger, 1989; Sahm et al. , 1992). Furthermore, Zymomonas has a high tolerance to inhibitors created duri ng hydrolysis of lignoc ellulosic feedstocks (Danner and Braun, 1999). Zymomonas also has a broader pH range (pH 5-7) required for optimal ethanol production, and a â€œh igher optimal growth temperatureâ€ (30 C for wild-type, with close to theore tical ethanol yield even at 37 C) for optimal ethanol production than yeast (Crueg er and Crueger, 1989). The primary disadvantage wild-type Z. mobilis has as a biocatalyst for the production of ethanol is the narrow range of substrates it can metabolize: glucose, fructose, and sucrose (Sahm et al. , 1992; Doelle et al ., 1993; Gunasekaran and Raj, 1999; Aristidou and Penttila, 2000). Wh en cultured in sucrose, or gl ucose plus fructose, rather than glucose alone, ethanol production is signif icantly reduced from the theoretical yield (Lee and Huang, 2000). A second disadvantage Zymomonas has as a biocatalyst for ethanol production is a low salt tolerance, wh ich would be a problem if the substrates used as the carbon source contains a high salt concentration, as occurs in molasses (Gunasekaran and Raj, 1999).
17 Bioengineering Bioengineering Z. mobilis to overcome its disadvantages, specifically, to enable it to use a larger and less expensive range of subs trates, is of particular interest for those looking toward Z. mobilis as an alternative source of ethanol producti on. Mutagenesis has been one way researchers have used to isolate microorganisms with new or improved phenotypes. In the past, various Zymomonas mutants have been isolated that exhibit desirable traits such as salt tolerance, tole rance to higher temperatures (such as stable growth with better ethanol tolerance at 42 C), greater tolerance to ethanol or other toxic waste products, flocculent gr owth, levansucrase mutants wi th higher ethanol yield when grown on sucrose, ability to grow in lower pH, and ability to grow on mannitol as the sole carbon source (Ingram et al. , 1989; Gunasekaran and Raj, 1999). Direct genetic engineering by introducing foreign genes into an organism is the method researchers use to develop recombinant microorganisms with traits that can not be developed by mutagenesis, such as new biosynthetic or catabolic pathways. Transformation and conjugation are the methods used to introduce foreign DNA into Z. mobilis (Sahm et al. , 1992). Transduction has not been used because there are yet no bacteriophages known to infect Zymomonas (Sahm et al. , 1992). In recent years, Z. mobilis has been successfully engineered to use xylose and arabin ose as sole carbon sources (Dien et al. , 2003). However, there are a few probl ems that make bioengineering Zymomonas relatively difficult. First of all, little is known about the regulatory mechanisms Zymomonas uses to control expression, such as what are the transcriptional and translational sequence requirements for expression. Secondly, Zymomonas has relatively few selectable markers and a limited selection of stable expression vectors (Conway et al. , 1987a, Sahm et al. ,
18 1992; Doelle et al ., 1993). Since Zymomonas is typically resistan t to a wide range of antibiotics, the genes for tetracycline resist ance and chloramphenicol resistance have been the primary selectable marker s used in gene transfer (Sahm et al. , 1992). Finally, a major problem that makes bioengineering Zymomonas relatively difficult is that the efficiency of stable transfer of foreign genes into Z. mobilis is relatively low, typically 100-1000 CFU/ g of foreign DNA (Ingram et al. , 1989). R-M Systems of Zymomonas mobilis R-M systems are known to protect the b acteria from transformation by foreign DNA. Bacterial transformation is the transfer of free DNA to recipient cells without direct contribution or contact from the inta ct donor cell (Joset and Guespin-Michel, 1993, p.279). The relatively low efficiency of transfer of foreign DNA into Z. mobilis may be an indicator of the presence of a restriction-modification system. REase systems have been identified in two Z. mobilis strains. Recent release of the genome sequence of Z. mobilis ZM4 revealed the presence of a puta tive Type I restriction system (Seo et al. , 2005). In Zymomonas mobilis subspecies anaerobia (NCI B8227), the REase ZanI was identified by protein purifi cation and activity assays (S un and Yoo, 1988). ZanI is a Type II REase with the molecular weight of 30,000 +/1,000 dalton and is maximally active at 37 C (Sun and Yoo, 1988; Sahm et al. , 1992; Roberts and Macelis, 1996). Based on genomic sequence analysis, a Type II system does not appear to be present in Z. mobilis ZM4. The recognition site of ZanI is 5â€™-CCWGG-3â€™, the same sequence recognized by BstNI and M.EcoKDcm (Sun and Yoo, 1988; Sahm et al. , 1992; Roberts and Macelis, 1996). ZanI and BstNI are not blocked by Dcm methylat ion and will cleave the internally methylated recognition s ite Cm5C(AT)GG (Sun and Yoo, 1988). Since BstNI cleavage is blocked by an N4-methylcyto sine at its recognition site and because
19 plasmid DNA isolated from Z. anaerobia resisted cleavage by BstNI, it was predicted that there is a cognate ZanI DNA MTase, functionally similar to M.BstNI, which produces N4-methycytosines at its recognition site (Sun and Yoo, 1988).
20 CHAPTER 2 DNA RESTRICTION AND MO DIFICATION SYSTEMS Discovery and Historical Applications The existence of restriction systems in b acteria was first detected when certain strains of Escherichia coli were found to possess a sort of â€œimmune systemâ€ against infection by bacteriophage (Luria and Hu man, 1952; Bertani and Weigle, 1953). The first molecular evidence that bacteriophage resistant bacteria possessed an enzyme system (a restriction-modifi cation system) that specifica lly recognizes and destroys foreign DNA, while modifying its own DNA to prevent destruction, was found in 1962 by Werner Arber and Daisy Dussoix (Arber and Dussoix, 1962; Dussoix and Arber, 1962). In 1963, Francois Jacob developed the replicon theory which predicted that any DNA molecule that contained its own origin of replication could be replicated by the host replication machinery (Jacob et al. , 1963). Salvador Luria a nd Werner Arberâ€™s groups observed that a single cycle of bacteriophage growth in a pa rticular host a lters the hostrange of all its progeny (Arber and Linn, 1969). It is a noninheritab le process described as an acquisition of a host controlled â€œrev ersible phenotypic modifi cationâ€ of the phage DNA that altered its host-ra nge specificity, a â€œhost-co ntrolled restriction and modificationâ€ (Arber and Linn, 1969). Math ew Meselson and Robert Yaun, as well as Werner Arberâ€™s group, extracted the firs t known restriction endonucleases (these endonucleases cleaved DNA nonspecifically) (Linn and Arber, 1968; Meselson and Yuan, 1968). In 1969 Urs Kuhnlein, Werner Ar ber, and Stuart Linn isolated the first DNA methyltransferase (Kuhnlein et al. , 1969). In 1970 Hamilton Smith and Kent
21 Wilcox isolated HindII from Haemophilus influenzae (Smith and Wilcox, 1970). It differed from the previously described endonuc leases by having separate restriction and modification enzymes, and by specifically re cognizing and predictably cleaving the DNA within the recognition site (Smith and Wilcox, 1970). In 1971, Kathleen Danna and Daniel Na thans showed the applicability of restriction endonucleases when they used HindII to digest simian virus 40 and polyacrylamide gel electrophoresis analysis of the restriction frag ments to produce the first restriction map (Danna and Nathan s, 1971). With the use of restriction endonucleases and the recognition of replicon th eory (leading to the use of plasmid and viral vectors), Paul Bergâ€™s group and Stan ley Cohenâ€™s group independently developed different methods for creating the first recombinant DNA molecules in 1972 (Jackson et al. , 1972; Cohen et al. , 1973). David Jackson, Robert Sy mons, and Paul Berg developed the â€œtailing methodâ€ of joining DNA molecules, modeled after the â€œs ticky endsâ€ found at the chromosome ends of the bacteriophage lambda (Jackson et al. , 1972). Using a modified transformation prot ocol developed in 1970 by Morton Mandel and Akiko Higa, Stanley Cohen and coworkers showed th at plasmid DNA could be transformed, maintained, and replicated in Escherichia coli cells that were made competent by calcium chloride treatment (Cohen et al. , 1972). Stanley Cohen and coworkers then created a biologically functional recombinant DNA mol ecule containing DNA from two different DNA species of different origin using the restriction endonuclease EcoRI [an enzyme shown by Janet Mertz and Ronald Davis (1972) to create its own â€œstickyâ€ ends] to cleave the DNA, (Cohen et al. , 1973). They created the recombinant plasmid by mixing and ligating ( in vitro ) the two populations of EcoRI digested plasmid DNA, one carrying the
22 gene for tetracycline resistan ce and the other carrying the ge ne for kanamycin resistance, and then selecting for transformed E. coli cells containing recombinant plasmid by its resistance to both tetracy cline and kanamycin (Cohen et al ., 1973). These ground breaking studies revolutionized molecular biology and set the foundation for modern biotechnology. It is little wonder that Wern er Abner, Daniel Nathans, and Hamilton O. Smith were awarded the Nobel prize in physiology or medicine in 1978 â€œfor the discovery of restriction enzymes and their a pplication to problems of molecular geneticsâ€ (Nobelprize.org). General Description of REases and DNA MTases Many restriction-modification (R-M) systems have been discovered in prokaryotes and viruses (Roberts and Macelis, 1996). It is generally believed that restrictionmodification systems function to reduce th e frequency of genetic exchange via transformation, transduction, or conjugation. RM systems originally referred to as host specificity of DNA (Hsd) systems, typically consist of the compleme ntary functions of a restriction endonuclease and a DNA methyltran sferase, in which the corresponding genes of known systems are usually linked (Wils on, 1991; Joset and Guespin-Michel, 1993, p.171). It should be mentioned that current reviewers refer to only the Type I R-M systems as Hsd systems (Roberts et al. , 2003). In eukaryotes, REases (Chevalier and Stoddard, 2001) and MTases (Jeltsch, 2002; Gr omova and Khoroshaev, 2003) have also been identified that are associated with other cellular functions. An excellent illustration of the function of R-M systems in prokaryotes can be found when analyzing the ones located in Escherichia coli strains K12 and B (Sain and Murray, 1980; Birge, 1981). A lthough they are members of th e same species, each strain is capable of degrading DNA orig inating from the other (Sain and Murray, 1980; Birge,
23 1981). Consequently, the R-M systems they pos sess genetically isolate them from each other (Birge, 1981; Sain and Murray 1980). In addition, it has been shown that bacteria naturally competent for conjuga tion poorly restrict the conjug al transport of homologous single stranded chromosomal DNA into cells due to (and dependent on) the efficient formation of heteroduplex DNA, which is inse nsitive to cleavage because of its hemimethylated state (Raleigh and Brooks, 1998). Therefore, the opposing R-M systems of the naturally competent for conjugation E. coli strains K12 and B would only efficiently restrict the otherâ€™s invading nonhomologous single stranded DNA (Raleigh and Brooks, 1998). Despite its perceived importance in cellular function, R-M systems are not essential; the cell is still vi able if the R-M system is knoc ked out (Jost and Saluz, 1993; Joset and Guespin-Michel, 1993, p.170-171). DNA methyltransferases (MTases) transf er a methyl group from S-adenosyl-Lmethionine (AdoMet) to a specific site with in a recognition sequen ce of double stranded DNA. Exceptions such M1.DpnII (Cerritelli et al., 1989) and M1.BcnI (Merkiene et al. , 1998) (Type II R-M systems) can site specifi cally methylate ssDNA (Bujnicki, 2001). MTases of Type I and II R-M systems usua lly methylate both strands while Type III MTases usually methylate only one st rand of dsDNA (Hyone-Myong, 1996; Bujnicki, 2001). DNA MTases only methylate adenine, cyto sine, or their modified equivalents and are grouped into two classe s based on the chemistry of methylation (Timinskas et al. , 1995, Jeltsch, 2002). The first class modifies the exocyclic nitrogen of adenines or cytosines to create N6-methyladenine (m6A) or N4-methylcytosine (m4C), respectively (Timinskas et al. , 1995, Jeltsch, 2002). The second cla ss modifies the 5-carbon of the
24 pyrimidine ring of cytosine to create 5-methylcytosine (m5C) (Timinskas et al. , 1995, Jeltsch, 2002). It has been shown that DNA MTases are not part of the replication complex and that methylation usually occurs after, not during, replication (Joset and Guespin-Michel, 1993, p. 170). Consequently, a region of DNA ma y exist in a hemi-methylated state for a significant period in the life cycle of a cell (0.5-3 min for E. coli at 30C) (Joset and Guespin-Michel, 1993, p.170-171). However, hemi -methylation is usually sufficient to prevent cleavage by most cognate REases of R-M systems (Bujnicki, 2001). Unlike DNA MTases of eukaryotic cells, it has been shown that most prokaryotic DNA MTases have no strong preference for hemi-m ethylated DNA over unmethylated DNA, the exception being the Type I MTases (Rao et al. , 2000) (see the section below on Type I system characteristics). The methyl groups are located in the major groove of dsDNA and do not interfere with base pairing (Bujnicki, 2001). Although DNA MTases have primarily been found to be part of R-M systems in prokaryotes, other solitary DNA MTases have been found where the â€œepigeneticâ€ information imparted by DNA methylation is involved in other cellular functions such as replication, DNA repair, regulating gene expression, and cell cycle regulation (Jost and Saluz, 1993; HyoneMyong, 1996; Bujnicki, 2001; Jeltsch, 2002). A restriction endonuclease (REase) is stimul ated by either a specific unmodified or modified (for methylation or glucosylati on dependent restriction systems) recognition sequence, and cleaves both strands of double stranded DNA one or more times at specific/predictable or at nons pecific/nonpredictable sites. In R-M systems stimulated by unmodified recognition sites, it is the modi fication of the recognition sequence, not the
25 modification of the cleavage site, which bl ocks restriction (Hy one-Myong, 1996). Most REases (all known Type II) are known to pr oduce 5â€™-phosphate and 3â€™-hyroxyl termini, while others have yet to be defined (Roberts et al. , 2003). As mentioned previously, REases are commonly used as molecular tool s for procedures such as DNA analysis or creating recombinant DNA. The actual number of fragments produced would depend on the pattern of DNA methylati on that may block cleavage, th e specificity of the REase used, and the number of restri ction recognition sites on the DNA. In other words, the length of the recognition site a nd the strictness of specificity of a given REase will affect the number of fragments produced by digest ion of a given molecule of DNA. Some REases are â€œstimulatedâ€ by abbreviated ma tches to the normal recognition sequence (Nasri and Thomas, 1986; Hyone-Myong, 1996). These enzymes cleave at these sites under conditions of high enzyme concentration or under â€œnonoptimalâ€ environmental conditions such as at low ionic strengths , high pH (>8.0), >5% glycerol, and in the presence of organic solvents and/or divale nt cations other than magnesium (Nasri and Thomas, 1986; Chirikjian, 1987; Hyone-Myong, 1996). A situation of altered specificity is termed star activity (Nasri and Thomas, 1986; Hyone-Myong, 1996). It should be noted that the terms REase and MTase refer to the functional enzyme, including all polypeptide subunits required for catalytic activ ity of restriction endonuclease cleavage and DNA methyltran sferase modification respectively. Classifications and Characterizations Nomenclature Herbert W. Boyer first proposed the classifi cation of REases as Type I or Type II, which has since been greatly expanded (Bu rrell 1993). In 1973, Smith and Nathans proposed the basis of the currently used nomenclature for individual restriction
26 endonucleases and DNA methyltransferases (Smith and Nathans 1973; Burrell, 1993; Hyone-Myong, 1996, Roberts et al. , 2003). REase and DNA MTase systems are currently named according to the following nomenclature guidelines as described by Roberts et al. (2003): (1) The fi rst letter is the initial letter of the genus from which it was isolated. (2) The second and third letters are usually the initial letters of the species. (3) The fourth letter, if present, indicates the strain of the organism. (4) The Roman numeral usually indicates the order of discovery. (5) The prefixes R, M, C, V, and N (Nt, Nb) are used to designate REases, MTases, controlling proteins, G/T mismatch Vsr-lik e nicking enzyme, and regular nicking enzymes, respectively, where Nt and Nb indicate top and bottom strand cleavage, respectively. Enzymes with the REase and the DNA MTas e activities catalyzed by a single protein have the prefix RM, however, th e prefix R or RM may be excluded. (6) If there are two REase or two MTase genes associ ated with the same R-M system, then the prefix is followed by an Arabic 1 or 2. (7) The single period or a hyphen (for homing endonucleases) is placed between the complete prefix (letters and numbers) and the main part of the name and no other punctuation mark s are used. (8) Italics and spaces are no longer used in the name. (9) The suffix P (for putative) should be added to the end of putative enzymes until biochemical activity is demonstrated. The number of R-M systems found in a singl e prokaryote is variable (Raleigh and Brooks, 1998; Williams, 2003). Most have more than one R-M system (Bujnicki, 2001), but they usually possess no more than two or three (Wilson, 1991); a noteworthy exception is the approximately two dozen found in Helicobacter pylori J99 (Bujnicki, 2001). Classic R-M systems are divided, based on their physical and enzymatic
27 properties, into Type I, Type II, and Type III classes, where these subdivisions are further broken down into subtypes and families. The review by Hyone -Myong (1996) proposed a Type IV class for systems that have bot h the R and M domains on a single polypeptide, but this kind of R-M system has since been reclassified as Type IIG (Roberts et al ., 2003). It has also been proposed that the Type IIB subclass be reclas sified as Type IV, because these R-M systems are unique in thei r ability to cleave on both sides of their recognition site (Williams, 2003). The Type IV classification, as described in this review, consists of systems that are not cla ssical R-M systems. The members of the Type IV class are methyl-dependent REases that do not have a cognate DNA MTase, usually do not have defined recognition nor cleavage sites, and do not produce a defined pattern of DNA fragments (Roberts et al. , 2003). While it may seem more logical to classify the R-M systems as proposed by Williams (2003) , which groups solitary MTases and modification dependent REases separate from the classical R-M systems, the REBASE Database, consisting of current informa tion on REases, DNA MTases, and related proteins (Roberts and Macelis, 1996), coincide s with the classification and nomenclature as described by Roberts et al. (2003). The classification a nd nomenclature as described by Robert et al. (2003), is used by a large majority of the leading reviewers in this field. Type I Systems Type I systems general characteri stics: Physiology and function Type I or Hsd R-M systems are the most complex and consist of three different types of high molecular we ight proteins, HsdR, HsdM, and HsdS, forming a single multifunctional enzyme (Dryden et al. , 2001). Typically, the protein complex consists of two ~140 kDa R (restriction) subunits, two ~50-60 kDa M (modification) subunits and one ~50 kDa S (speci ficity) subunit (Rao et al. , 2000; Bujnicki, 2001; Bourniquel and
28 Bickle, 2002). Currently, all known Type I syst ems are divided into four families: A, B, C, and D (Bickle and Kruger, 1993; Rao et al. , 2000; Dryden et al. , 2001). While nucleic acid and amino acid sequence identity within a family is high, there is typically little sequence similarity between families (Rao et al. , 2000; Dryden et al. , 2001). An exception is that some sequence similarity has been found among some S subunits that belong to different families but have th e same target recognition site (Rao et al. , 2000; Dryden et al. , 2001). The genes of Type IA, IB, and ID systems are allelic and are linked to the serB gene locus on the chromosomes of E s cherichia coli K12 and Salmonella enterica , while Type IC are usually found on pl asmids (Bickle and Kruger, 1993; Rao et al. , 2000). One Type IC exception ( EcoprrI ) found on the chromosome is part of a defective prophage (Rao et al. , 2000; Titheradge et al ., 2001). The genes that code for these systems are typically closely linked, consisting of hsdR , hsdM , and hsdS (not necessarily in that order) (Bickle and Kruger, 1993; Murray, 2000; Rao et al. , 2000). With few exceptions, the DNA methyltransferase and specificity genes are transcribed as a single operon, and the restriction gene is tran scribed from its own promoter, typically in the same direction (Sain and Murray, 1980; Bickle and Kruger, 1993; Murray, 2000; Rao et al. , 2000). One interesting exception is that some plasmids have been found that only contain the hsdS gene (Rao et al. , 2000). It has been shown that when introduced into certain host strains with a Type I system, th e plasmid encoded specifi city gene conferred a new R-M phenotype on the host, indicati ng that the plasmid encoded HsdS subunit could interact with the host HsdR and HsdM subunits (Rao et al. , 2000). The specificity proteins (HsdS) of Ty pe I R-M systems, determine the site specificity for both the restriction endonucl ease and the DNA methyltransferase subunits
29 (Rao et al. , 2000; Dryden et al. , 2001; Bourniquel and Bick le, 2002). Binding analysis has shown that the HsdS subunit binds to the DNA at its target rec ognition sites and the HsdM subunits bind to the HsdS subunit, away from the DNA helix (Powell et al. , 1998). The recognition sequences are asymmetric and bipartite, consisting of a 5â€™ 3-4 bp and a 3â€™ 4-5 bp sequence interrupted by a nonspecific sequence (Rao et al. , 2000; Dryden et al. , 2001). Generally, Type IA recognition site s have an 8 bp interruption, Type IB recognition sites have a 9 bp interruption, Type IC rec ognition sites have a 7-8 bp interruption, and Type ID recognition s ites have a 6 bp interruption (Murray, 2002). DNA cleavage occurs at variable distance fr om the unmodified recognition site (Rao et al. , 2000; Dryden et al. , 2001). Consequently, it has b een reported that no discrete banding pattern is detected on agarose gels, except under special c onditions (Raleigh and Brooks, 1998). Currently, all known Type I MTases form m6A (N-6 methyladenine) methylation (Roberts and Macelis, 1996; Roberts et al. 2003). MTase activity of most Type I R-M system enzymes requires at least the trimeric complex consisting of two DNA methytra nsferases and one specificity subunit (M2S1), and requires the presence of AdoMet and magnesium as cofactors (Rao et al. , 2000; Dryden et al. , 2001; Bourniquel and Bickle, 2002); however, some Type I systems may only require heterodimer formation for activ ity (Williams, 2003). Type I R-M systems modify dsDNA, normally the hemi-methylated DNA that is formed following replication, transferring the methyl group from AdoMet to the N-6 position of a specific adenine in its recognition site, re sulting in specific adenine methylation on both strands (Rao et al. , 2000; Bourniquel and Bickle, 2002). AdoMet bi nds to the HsdM subunits of free enzyme complex and acts as a positive allosteric e ffector, inducing formation of the correct
30 enzyme conformation required for restri ction activity, promoting DNA binding, and possibly playing a role in determ ining DNA methylation status (Rao et al. , 2000; Bourniquel and Bickle, 2002). ATP also acts as an allosteric effector (Bickle et al. , 1978) and enables the enzyme to discriminate the methylation status of the DNA (Rao et al. , 2000; Bourniquel and Bickle, 2002). Unlike MTases found in the other types of R-M systems, Type I MTases strongly prefer hemi-methylated DNA (Rao et al. , 2000; Bujnicki, 2001; Dryden et al. , 2001). ATP reduces enzyme affinity for fully methylated DNA, stimulates methylation of hemi-m ethylated DNA, and is required for DNA translocation (Rao et al. , 2000, Dryden et al. , 2001). More specifically, studies have shown that the Type IA system enzyme complexes (HsdRMS) generally methylate unmethylated DNA very slowly, and methyl ation of unmethylated DNA is further inhibited by ATP (Rao et al. , 2000). The Type IC system enzyme complexes (HsdMSR) methylate unmethylated DNA very slowly as well, but ATP stimulates methylation (Rao et al. , 2000). The Type IB system enzyme complexes (HsdRMS) methylate each substrate type (unmethylated slightly slower than hemi-m ethylated) rapidly but require ATP to efficiently methylate unmethylated DNA (Rao et al. , 2000). The Type ID system enzyme complexes (HsdMSR) are still being characterized. The REase activity of Type I R-M system enzymes requires complete complex formation of all three type s of subunits: typically R2M2S1 (Rao et al. , 2000; Bourniquel and Bickle, 2002; Williams, 2003). Depending on the Type I system, the two R subunits bind to the hsdM2S trimer with varying affinity (e ither two strong, or one strong one weak, or two weak) (Rao et al. , 2000; Dryden et al. , 2001). REase activity requires the presence of ATP, magnesium ions , and AdoMet as cofactors (Rao et al. , 2000;
31 Bourniquel and Bickle, 2002; Williams, 2003). Unlike other types of R-M system, Type I systems are generally consid ered to have an absolute requirement for AdoMet for endonuclease restric tion activity (Rao et al. , 2000; Williams, 2003). The exception is the Type IC enzyme EcoR124I which is reported to be stimulated by but not require AdoMet for DNA cleavage (Rao et al. , 2000). AdoMet acts as a pos itive allosteric effector of REase activity (Rao et al. , 2000; Bourniquel and Bickle, 2002 ). After a single cleavage event, the enzyme comple x losses endonuclease activity and remains bound to the DNA (it is not recycled), but it can continue to hydrolyze ATP in larg e quantities while bound (Dreier and Bickle, 1996; Dryden et al. , 2001; Bourniquel and Bickle, 2002). The current model for Type I endonucleas e cleavage is based mainly on electron and atomic force microscopy studies. Wh ile cleavage of linear DNA requires the presence of two or more unmethylated site s, circular DNA with only one unmethylated site can be cleaved (Rao et al., 2000; Dryden et al. , 2001). In the presence of AdoMet, the multifunctional enzyme tightly binds to the two unmodified specific sequences of the bipartite recognition site and weakly to nonspecific sites on the DNA (Rao et al., 2000; Bourniquel and Bickle, 2002). Tw o DNA bound Type I holoenzymes (R2M2S1) â€œcollideâ€ or dimerize, producing the restric tion endonuclease active form (Dryden et al. , 2001). The binding of ATP causes a conformational ch ange in the enzyme, producing a more compact structure, resulting in further weak ening of its attachment to the nonspecific sites (Rao et al., 2000; Bourniquel and Bickle, 2002). Th e active site utilizes the energy from ATP hydrolysis by the two HsdR subunits to form two graduall y enlarging loops of DNA (translocating in opposite directions) fed though the nonspecific DNA sequence of the bipartite binding sites bound to the multimeric enzyme (Rao et al., 2000; Dryden et
32 al. , 2001; Szczelkun, 2002). The current kinetic m odel for Type I translocation describes it as processive, where it is significantly more likely for the endonuclease to translocate forward than to dissociate from th e DNA (Bourniquel and Bickle, 2002; Szczelkun, 2002). While some agree with the original co llision model that st ates that collision occurs after transloca tion of two converging REase comple xes, halting translocation and triggering cleavage (Rao et al., 2000; Dryden et al. , 2001; Bourniquel and Bickle, 2002; Szczelkun, 2002), other more recent eviden ce, including atomic force microscopy experiments show dimerized Type I holoenz ymes in the absence of ATP (required for translocation) (Dryden et al. , 2001; Bujnicki, 2001; Bourniqu el and Bickle, 2002). This supports the current model that the dimeri zed REase active form is created before translocation, and cleavage is triggered by stalling of transl ocation when the expanding loops pause or can no longer expand (Dryden et al. , 2001; Bujnicki, 2001; Bourniquel and Bickle, 2002). Either way, the REase ac tive complex dimer cleaves the strand, and then both enzymes complexes disengage from each other but remain bound to the DNA, resulting in a double stranded cut nonsp ecifically at about 50-10000 bp from the recognition sequence (Szczelkun, 2002). In some cases, the cleavage sites have been found to be around the midpoint between adjacent recognition sites, in others, cleavage sites are broadly distributed (possibly due to the premat ure release of DNA from one holoenzyme complex) (Dryden et al. , 2001; Szczelkun, 2002). While REase activity requires dimerization of two Type I enzymeDNA complexes in the proper conformation, evidence suggests that they do not have to be bound to neighboring sites nor do the sites need to be in reverse orientation, as described below for Type III systems (Rao et al., 2000; Dryden et al. , 2001; Bourniquel and Bickle , 2002; Szczelkun, 2002). While
33 evidence also suggests that any type of physical barrier that causes stalling of translocation (such as holiday junctions) can trigger cleavage, build-up of supercoils during translocation is not one of them (Rao et al., 2000; Bourniquel and Bickle, 2002). The greater the degree of positive supercoi ling, the greater the difficulty in dimer formation, resulting in slower rates but not abolishment of cleavage, and likewise, increase of negative supercoiling al so decreases REase activity (Rao et al., 2000; Bourniquel and Bickle, 2002). Type I systems: Regulation Restriction alleviation (RA) at the posttran slational level is a general characteristic of Type I R-M systems, and it is probably accomplished by several different mechanisms, two of which are described by Makovets et al. (2004): (1) the ClpXP dependent proteolysis mechanism used when acquiring new Type IA or IB systems, or in response to DNA damage, and (2) the mechanism of disassembling the enzyme complex in Type IC systems in response to DNA damage. Othe r mechanisms of regul ation such as gene arrangement (Rao et al. , 2000) and enzyme localization (Holubova et al. , 2000) have also been described. Early studies showed that during establishm ent of the Type IA system EcoKI in a new unmodified host (one C and two K12 strains of E. coli ), REase activity is delayed for 15 generations post conjugation (full activity occurs after 30 generations), although all the hsdRMS genes are expressed immediately (Prakash-Cheng and Ryu, 1993b). In contrast, methylation activ ity is immediately detected at zero generations post conjugation, which suggests that there is some sort of REase regulation (Prakash-Cheng and Ryu, 1993b). The ClpXP serine protease is essential for establishment of several Type IA (such as EcoKI) and Type IB (such as EcoAI) systems (Makovets et al. , 1999;
34 Rao et al. , 2000; Dryden et al. , 2001). ClpXP alleviates rest riction activity in response to DNA damage by degrading the RE ase HsdR subunit (Makovets et al. , 1999). HsdR degradation is stimulated by mutations th at make the host deficient in relevant methyltransferase activity and by mutations affecting DNA replication fidelity ( dam , topA , and mutD ), as well several external DNA damagi ng agents that can eventually lead to generation of unmodified sites (UV light, nalidixic acid, and 2-aminopurine) (Makovets et al. , 1999). However, HsdR degrada tion is not stimulated by double stranded breaks (Makovets et al. , 1999). This was evident when it was discovered that DNA damage due to restriction by one Type I system does not stimulate degradation of the HsdR subunit of another Type I system produced in the same cell, although both systems are known to be regulated by ClpXP protease degradation of their restriction endonuclease subunit (Makovets et al. , 1999). HsdR is only degraded if there are unmodified restriction site s on the DNA, if all three hsdRMS products are produced, and if the HsdR subunit is func tional in the cell (Makovets et al. , 1999). Any missense mutation of the HsdR that produces a nonfunc tional protein but does not prevent enzyme complex formation, DNA binding or ATP binding ( i.e ., a missense mutation that causes a defect in ATP hydrolysis and consequently ATP-dependent translocation) prevents degradation of the HsdR subunit by ClpXP (Makovets et al. , 1999). In studies on the disassembly and degradation of the Mu tran sposase by the ClpXP protease, the model for ClpXP activity states that the chaperone ClpX is the subunit that recognizes the target (MuA-DNA complex) while the ClpP subunit is responsible for the protease activity (Makovets et al. , 1999). Based on the evidence from studies on ClpXP and the model for ClpXP activity, it is believed that only af ter restriction of th e host DNA is initiated
35 (during the ATP-dependent tr anslocation down the chromosomal DNA) but prior to cleavage, the target for ClpX recognition is exposed on the Type I enzyme complex bound to unmodified sites, allowing subse quent degradation of the HsdR subunit (Makovets et al. , 1999). In other words, ClpXP does not target and degrade free Type IA or Type IB HsdR subunits; ra ther, it targets the bound active HsdR subun its in a specific conformation. The Type I enzyme complex also ap pears to be regulated by an unknown mechanism so that it does not restrict the DNA of progeny that has lost its hsd genes (Kulik and Bickle, 1996; Oâ€™Neill et al. , 1997; Murray, 2002; Makovets et al. , 2004). Even in the absence of ClpXP, E. coli is completely insensitive to the loss of the Type IA system EcoKI, suggesting another mechanis m to protect the host unmethylated progeny DNA from residual REase activity (Murray, 2002). This may be due in part to the fact that Type I restriction endonuc leases lose their activity af ter a single cleavage event (it remains bound to the DNA and is not recycled), resulting in rapid decr ease in restriction activity, which would increase the ch ance for progeny that has lost its hsd genes to sufficiently repair its DNA and su rvive (Kulik and Bickle, 1996). Early studies showed that the Type IC sy stem EcoR124I also controls restriction activity by a mechanism at th e posttranslational level (Kulik and Bickle, 1996). Although restriction activity is immediately detected after transf er of the EcoR124I system into a new naive host (full activity occu rs after six generations ), it is not lethal when this system is transferred into a ne w host that is already expressing an EcoR124I HsdR subunit (Kulik and Bickle, 1996). Unlik e Type IA and Type IB systems, it was shown that the establishment of a Type IC system in a new host does not require ClpXP
36 (Makovets et al. , 1999; Rao et al. , 2000; Dryden et al. , 2001). Perhaps this difference evolved in part by the fact that Type IA a nd IB are located on the chromosome and Type IC systems are generally located on a plasmi d, because Type IC systems are more likely to be transferred to a new host (Kulik and Bickle, 1996). Recent studies showed that both Type IC and Type ID systems alleviate re striction in response to DNA damaging agents, independent of ClpXP (Makovets et al. , 2004). Based on mutationa l analysis studies, it appears that restriction alleviation in Type IC systems may be cont rolled by a mechanism of disassembling the enzyme complex require d for REase activity (dissociation of the HsdR subunits) (Makovets et al. , 2004). Makovets et al. (2004) has also suggested that assembly and disassembly of the active enzyme complex may be a factor in allowing either acquisition or loss of Type IC systems. A third relatively well-characterized T ype I R-M regulation mechanism is found in Mycoplasma pulmonis , an organism that undergoes a high frequency of DNA rearrangements, in part due to its susceptibility to infection by the Mycoplasma virus P1 (Rao et al. , 2000). This pathogenic bacterium has a Type I R-M system, located in a 6.8 kb invertible region (hsd 1 locus), consisting of two hsdS genes ( hsdSA and hsdSB ) in converging orientation, flanking the hsdM and hsdR genes (Rao et al. , 2000). This Type I R-M system was shown to only restrict invading P1 virus if the hsdRM genes are in one of two orientations (in th e same orientation as the hsdSA gene), meaning that expression of the R-M system is regulated by an invertible DNA element (Rao et al. , 2000). Control of REase activity at the transcriptional leve l by repression or activa tion of expression has not yet been shown in Type I systems (Murray, 2002).
37 A fourth possible mechanism for Type I sy stem regulation has been suggested from studies on the localiza tion of the HsdM, HsdR, and HsdS subunits of the EcoKI system (Holubova et al. , 2000). Immunoblotting experiments i ndicate that the HsdR and HsdM subunits are membrane associated when all three hsd genes are expressed, but not when the hsdS gene is not expressed (Holubova et al. , 2000). The HsdS subunits are found in both the cytoplasm and periplasmic space (Holubova et al. , 2000). This suggests that when the enzyme complex is associated with the cytoplasmic membrane, the periplasmic HsdS subunit recognizes and binds to i nvading foreign DNA, and the restriction endonuclease subunit, unlike the methyltransf erase subunit, can cross the cytoplasmic membrane allowing it access to the peri plasmic space, facilitating cleavage but preventing modification of the foreign DNA (Holubova et al. , 2000). This recent finding is particularly relevant in the pursuit of extracting, purif ying, and characterizing Type I R-M system proteins. Type II Systems Type II systems general characteri stics: Physiology and function Type II systems have defined recognition and cleavage sites, where cleavage occurs at a fixed position or at a known yet limited position within or near its recognition site, producing a predicable pattern of DNA fragments (Roberts et al. , 2003). Because of this predictability of cleavage, Type II endonucleases are the type of endonucleases most commonly used as molecular tools. Type II R-M systems are the most common of all R-M systems (Roberts and Macelis, 1996). The size of Type II REases vary widely, with the average molecular weight of 22-37 kDa per subunit (Hyone-Myong, 1996). The size of the MTase averages around 380 amino acids (~42.6 kDa) (Hyone-M yong, 1996). Most Type II systems
38 recognize a specific 4-8 bp sequence (alt hough there are many exceptions) (HyoneMyong, 1996; Bujnicki, 2001). Type II system REases may or may not have a cognate DNA MTase (Roberts et al. , 2003); however, the majority of Type II systems consist of one restriction endonuclease and one DNA met hyltransferase that typically function independently as homodimeric and mono mer enzymes, respectively (Hyone-Myong, 1996; Bujnicki, 2001; Williams, 2003). A survey of Type II systems reveals that there are several exceptions such as R-M system s that have two DNA MTases, restriction systems with methyl-dependant REases (wit hout a cognate MTase), R-M systems with R and M domains found on a single polypeptid e, as well as some monomeric, homotetrameric, heterodimeric, and hetero tetrameric Type II systems (Roberts and Macelis, 1996; Roberts et al. , 2003; Williams, 2003). Finally, some Type II R-M systems, such as BcgI, have a separate s ubunit that contains the specificity (S) domain that recognizes and binds to the target DNA sequence, like in Type I systems, but are biochemically classified as T ype II (Bujnicki, 2001; Roberts et al. , 2003). Structurally, all currently known Type II REases have a cleavage domain and may or may not have their own target recognition domain/eleme nts (most do) (Roberts and Macelis, 1996; Bujnicki, 2001). All Type II DNA MTases have three separate domains, whether or not they are all located on the same polypeptide, consisting of an AdoMet binding domain, a target recognition domain, and a catalytic dom ain (Roberts and Macelis, 1996; Bujnicki, 2001; Jeltsch, 2002). Unlike Type I and III system REases, Type II restriction endonuc lease activity does not require ATP hydrolysis (Hyone-Myong, 1996; Raleigh and Brooks, 1998; Rao et al. , 2000; Bujnicki, 2001; Roberts et al. , 2003). Type II systems usually require magnesium
39 ions as a cofactor (Raleigh and Brooks, 1998; Rao et al. , 2000; Bujnicki, 2001; Roberts et al. , 2003) with the exception of BfiI (Sapranauskas et al. , 2000). A few are stimulated by AdoMet (Bath et al. , 2002; Raghavendra and Rao, 2003; Roberts et al. , 2003). Most Type II REase require the pr esence of only one unmethyl ated recognition site for cleavage with many exceptions, such as EcoR II which requires simultaneous binding to two recognitions sites for cleavage of the substrate DNA (Bickle and Kruger, 1993; Bath et al. , 2002). The DNA MTase only requires AdoMet to act as the methyl donor (HyoneMyong, 1996; Raleigh and Brooks, 1998; Bujnicki, 2001). In R-M systems it is essential that the DNA methyltransferase completely modify the genome at each of its recognition sites to protect it from the cognate endonuclease. This is consistent with the finding that thei r genes are usually linke d in the genome, on native plasmids or on the chromosome, where they may be aligned in tandem or in opposite orientation (Bickle and Kruger 1993; Hyone-Myong, 1996). Type II systems may have one or two DNA methyltransferas es. Type II systems with palindromic recognition sites (Type IIP) usua lly have one MTase that met hylates both strands with the same specificity (Bujnicki, 2001). Type II systems with asymmetric recognition sites (Type IIS) most commonly have two DNA MTas es, each with its own strand specificity, one to methylate one strand and the other to methylate the other strand, as found in the HgaI system, and may even methylate differe nt base types (adeni ne on one strand and a cytosine on the other) as found in the NgoBVIII system (Szybalski et al. , 1991; Bujnicki, 2001; Roberts et al. , 2003). However, the structural form and functional activity of the DNA MTase(s) found among Type II systems w ith asymmetric recognition sites vary widely and the currently known arrangements are described as follows. Some Type II
40 systems with an asymmetric recognition site have only a single MTas e that modifies both strands of its recognition site (Roberts and Macelis, 1996). So me can do so because they recognize a degenerated quasi-palindromic sequence, such as M.BstNBI, which recognizes 5â€™-GASTC-3â€™ (S= G or C) (Bujni cki, 2001); other Type II systems with a single DNA MTase polypeptide may have two independent methyltransferase domains (due to gene fusion) using one to modify each strand of an asymmetric recognition site, as in the case of M.FokI (Szybalski et al. , 1991; Bujnicki, 2001). It is interesting to note that hemi-methylation of the recognition site is sufficient to block re striction in the FokI system (Szybalski et al. , 1991). Similarly, some Type II systems with an asymmetric recognition site may have a single MTase that modifies only one strand in the recognition site yet confers complete protection (Szybalski et al. , 1991). Some Type II systems have a DNA MTase fused to its cognate REase and methylate both strands of an asymmetric recognition site, as in the cas e of HaeIV (Bujnicki, 2001) and AloI systems (Roberts and Macelis, 1996). Other Type II R-M system s may have two DNA MTases where one is fused to the REase and modifies only one stra nd and the other (separate) MTase modifies both strands of an asymmetric recognition si te, as found in the Type IIG Eco57I system (Bujnicki, 2001; Williams, 2003). Type II systems have been categorized into eleven subtypes (A, B, C, E, F, G, H, M, P, S, and T) based on physical and/or RE ase enzymatic properties, where each system may belong to more than one s ubdivision (Bujnicki, 2001; Roberts et al. , 2003). Type IIA is the general term for all Type II systems that have asymmetric (not palindromic) recognition sequences and may cl eave within or away from this sequence (Roberts et al. , 2003).
41 Type IIB systems (such as BcgI) cleave DNA on both sides of its recognition site, resulting in excision of a short dsDNA fr agment containing the recognition site (Szybalski et al. , 1991; Bujnicki, 2001; Roberts et al. , 2003). Type IIC is the general cla ssification of all Type II systems that contain both the REase and MTase catalytic domains within a single polypeptide, such as BcgI (Roberts et al. , 2003). The target recognition site may be asymmetric or symmetric (Roberts et al. , 2003). Type IIE systems, such as EcoRII, have two DNA binding sites on the same REase (Bujnicki, 2001; Bath et al. , 2002; Roberts et al. , 2003). Each REase DNA binding site binds to separate recognition si tes on DNA, where one acts as an allosteric effector of the enzyme and the other is the target recogniti on site and is cleaved (Bujnicki, 2001; Bath et al. , 2002; Roberts et al. , 2003). The kinetics of cleavage depends on the distance separating the target recogn ition sites on the same DNA molecule (Bickle and Kruger, 1993). For EcoRII, the two target recogniti on sites must be within 1kb (Bickle and Kruger, 1993). If the enzyme binds two different DNA molecules, the kinetics of cleavage depends on both the concentration and size of DNA molecules containing the target recognition sites (Bickle and Kruger, 1993). The MTase is generally a single separate monomeric enzyme (Williams, 2003). Type IIF, such as SfiI, also has tw o DNA binding sites on the REase like IIE systems (Bujnicki, 2001; Bath et al. , 2002; Roberts et al. , 2003). The REase DNA binding sites bind to separate recognition sites on DNA and each target recognition site is cleaved coordinately (Bujnicki, 2001; Bath et al. , 2002; Roberts et al. , 2003). All known
42 Type IIF system REases are homotetrameric (B ujnicki, 2001). The MTase is generally a single separate monomeric enzyme (Williams, 2003). Type IIG systems also belong to the Type IIC subclass because they contain both the REase and MTase domains within a single polypeptide (Szybalski et al. , 1991; Bujnicki, 2001; Roberts et al. , 2003). Furthermore, Type IIG systems have the additional criteria of being affected by AdoMet (stimulated or inhibited) (Szybalski et al. , 1991; Bujnicki, 2001; Roberts et al. , 2003). The target recognition site may be asymmetric or symmetric (Roberts et al. , 2003). Type IIH systems, such as BcgI, act biochemically like Type II systems but genetically resemble Type I systems, havi ng two or three genes where one codes for a separate specificity protein (Roberts et al. , 2003). The target recognition site may be asymmetric or symmetric (Roberts et al. , 2003). Type IIM systems are modification depe ndent restriction (MDR) endonucleases, like Type IV systems, but are biochemically cl assified as Type II because they recognize a specific recognition site and cl eave at a fixed site (Roberts et al. , 2003). They do not have a cognate MTase. Type IIP is the general term for all sy stems that have palindromic recognition sequences, typically consisting of 4-8 specifi c bp, and may be interrupted by additional nonspecific bp within the site (Roberts et al. , 2003). The cleavage site is within or immediately adjacent to the unmodified rec ognition sequence at symmetrical points on both strands, producing either bl unt ends, if it cleaves in the middle of the recognition site, or cohesive single strande d ends, if it cleaves to one side or the other of the center of
43 the recognition site (Roberts et al. , 2003). The MTase is generally a single separate monomeric enzyme (Williams, 2003). Type IIS systems, such as FokI, have an uninterrupted asymme tric recognition site and an asymmetric cleavage site (Szybalski et al. , 1991; Bujnicki, 2001; Roberts et al. , 2003). The REase of Type IIS systems cleaves dsDNA, where at leas t one stand of the DNA is cleaved at a defined distance to one side of the recognition site rather than within the recognition site (Szybalski et al. , 1991; Roberts et al. , 2003). Most Type IIS systems cleave both strands at a precise distance outsi de of the recognition si te, generating unique ends and leaving the recogniti on site intact (Szybalski et al. , 1991; Roberts et al. , 2003). Type IIS system REases are generally larger than other Type II systems, ranging from 47-108 kDa, and most are catalytically active in the monomeric form, able to cleave at least one strand, rather than the typical hom odimeric form required for REase activity of other Type II systems (Szybalski et al. , 1991; Hyone-Myong, 1996; Bath et al. , 2002). For dsDNA cleavage, two unmodified recogni tion sites are require d by most Type IIS systems (making them Type IIE as well), where only one of the two recognition sites is cleaved per reaction by mechanisms that ar e currently unclear (B ujnicki, 2001; Bath et al. , 2002). Type IIS REases appear to be clos ely related to â€œnickingâ€ endonucleases that only cleave the top strand of the DNA to whic h it is bound (Bujnicki, 2001). It has been suggested that the inability of â€œnickingâ€ e ndonucleases, such as N.BstNBI, to cleave the bottom strand is due to loss of the ability to dimerize (Bujnicki, 2001). The unique ability of most Type IIS endonucleases to cleave DNA without destr oying the recognition site makes these enzymes particularly interesting for their varied usefulness as tools in molecular science (where many exam ples are described by Szybalski et al. , 1991). The
44 form and activity of the cognate MTas e vary, depending on the specific system (Williams, 2003). Type IIT systems have two R subunits, each encoded by its own gene (Bujnicki, 2001; Roberts et al. , 2003). The functiona l enzyme may be heterodimeric, such as Bpu10I, or heterotetrameric ( 22), such as BslI (Bujnicki, 2001; Roberts et al. , 2003). The target recognition site may be asymmetric or symmetric (Roberts et al. , 2003). The form and activity of the cognate MTas e vary, depending on the specific system (Williams, 2003). Type II systems: Regulation In some Type II systems, there is a third ge ne that is tightly linked to the REase and DNA MTase genes, which codes for a regulatory protein referred to as the C (controller) protein (Bickle and Kruger 1993; Ives et al. , 1995; Rimseliene et al. , 1995; Bujnicki, 2001; Roberts et al. , 2003). Mutation of the C gene eith er knocks out restriction activity, as in the case of pvuIIC (Tao and Blumenthal, 1992), or results in decrea sed restriction and increased methylation ac tivity, as in the case of bamHIC (Ives et al. , 1992). The C protein is believed to be a positive tran scriptional regulator of the restriction endonuclease mediating temporal control of restriction endonuclease expression (Ives et al. , 1992, 1995; Vijesurier et al. , 2000; Bujnicki, 2001; Roberts et al. , 2003). It is involved in maintenance (resistance to loss) of R-M systems via a â€œpostsegregational cell killingâ€-like mechanism (Nakayama and Kobaya shi, 1998). If the C protein specificity of the new R-M system does not already exist in the cell, it allows new R-M systems to be established in a cell by delaying REase activ ity (Nakayama and Kobayashi, 1998) (see the section below on selfish genetic units). If the host already has an R-M system with
45 the same C specificity, but different R-M sp ecificity, the host cell â€œcommits suicideâ€ when the new R-M system is introduced (N akayama and Kobayashi, 1998). Apoptotic mutual exclusion (host cell suic ide) occurs because the C prot ein of the resident system causes premature expression of the incoming REase before its cognate MTase can fully protect the host DNA (Nakayama and Kobayashi, 1998). In the PvuII system, the promoters for the DNA MTase and the C protei n are in opposite orientation, which seems to be the typical orientation found in the Type II systems with identified C genes (Wilson, 1991; Anton et al. , 1997; Vijesurier et al. , 2000). An exception is BglII, where the C protein, REase, and DNA MTase genes are a ll in the same orie ntation (Wilson, 1991; Anton et al., 1997; Vijesurier et al. , 2000). In all C protein regulated R-M systems analyzed to date, the C gene occurs immediat ely upstream of, and in the same orientation as, the REase gene and in some cases even overlaps the REase gene (Vijesurier et al. , 2000; Bujnicki, 2001). Nakayama and Kobayash i (1998) proposed that the genes for the DNA MTase and the C protein are expressed be fore the REase gene , ensuring that the host DNA is completely protected by methylat ion. Although the mechanism of C protein regulation is not completely clear, studies ha ve shown that the C protein in PvuII is a sequence specific DNA binding protein that bind s to identified sites upstream of the C gene, referred to as the â€œC box regionâ€, a nd appears to act as a homodimer (Rimseliene et al. , 1995; Vijesurier et al. , 2000). Evidence from studies of the PvuII and BamHI R-M systems suggests that temporal regulation of the REase expressi on is due at le ast in part to the autoregulatory activa tion of the polycistronic pvuIICR and bamHICR promoters, respectively, by the C protein binding to the C box region (Vijesurier et al. , 2000; Bujnicki, 2001). The sequence, spacing, and polarity of the C boxes seem to be
46 conserved in BcnI, BglII, BamHI, Ec o72I, MunI, SmaI, and PvuII (Rimseliene et al. , 1995; Vijesurier et al. , 2000). Vijesurier et al. (2000) has also suggested that there may be 5â€™-end mRNA hybridization of the C protein and REase transcripts in the PvuII system because they overlap. Hybridization of the 62-101 nt complementary sequences would block the termination of the C protein mR NA while translation of the REase mRNA continues from a second translation initiati on codon, shown to be used in 90% of the initiations of cloned PvuII systems in E. coli (Vijesurier et al. , 2000). Currently, at least 34 Type II R-M systems with C proteins ha ve been identified (Roberts and Macelis, 1996). Type II systems, without a C regulatory gene, must have another means of tightly regulating R-M system gene expression. The C protein appears to be self-regulating and functions as a regulatory activat or of expression of any REase gene to which it is directly upstream (Ives et al. , 1995), if it has the same C specificity (Nakayama and Kobayshi, 1998). In addition, C genes that are closel y related are associat ed with seemingly unrelated R-M system genes. Consequently, Vijesurier et al. (2000) speculated that the C gene may have evolved independently of th e R-M system genes, may be easily moved, and may possibly regulate the expression of othe r genes in addition to R-M system genes. In the PvuII system, there is a f ourth open reading frame, designated pvuIIW, which has been shown to be transcribed (Adams and Blumenthal, 1995; Bujnicki, 2001). This gene is located within, and in opposite orientation to, the PvuII DNA MTase gene (Adams and Blumenthal, 1995; Bujnicki, 2001) . This predicted 28 amino acid protein has amino acid similarity to a region on the restriction endonuclease involved in proteinprotein interaction at the dimer interface a nd is proposed to be a modulator of REase dimer formation (Adams and Blumenthal, 1995; Vijesurier et al. , 2000; Bujnicki, 2001).
47 While this protein has been shown to have no effect on the active REase dimer, it has been shown to inhibit rena turation of denatured restri ction endonuclease (Adams and Blumenthal, 1995). Consequently, pvuIIW may be involved in delaying restriction activity by delaying REase dimer formation, ther eby facilitating the es tablishment of the R-M system in a new host (Adams and Blumenthal, 1995). Another possible mode of regulati ng Type II R-M system activity is compartmentalization. In early studies of PstI and EcoRI, e ndonuclease activity was recovered from the periplasm (Holubova et al. , 2000). In another report, PstI was isolated and purified by extracting the rest riction endonuclease from the periplasmic space using osmotic shock (Smith et al. , 1976). As for EcoRI system, there are conflicting reports on its cellula r location, with some reporting that the REase is located in the periplasmic space and the DNA MTase is located in the cyt oplasm (from electron microscopy studies), while another report stated that almost all re striction activity is recovered from the cytoplasm and that none of the REase was excreted from a tolA excretory mutant (Holubova et al. , 2000). By compartmentalizing the REase in the periplasm, a cell would have a physical ba rrier, in addition to DNA methylation by its cognate MTase, to block REase digestion of its chromosomal DNA. More importantly, invading foreign DNA would be exposed to i mmediate digestion by the REases in the periplasm, before it reaches the cytoplasm where it may be methylated by the cognate DNA MTase. Type III Systems Type III systems general characteri stics: Physiology and function Type III R-M systems were initially thought to be Type I systems, but despite mechanistic similarities, are considerably di fferent in several as pects (Bujnicki, 2001;
48 Dryden et al. , 2001; Bourniquel and Bickle, 2002). The Type III recognition sequences are currently charac terized as asymmetric, uninter rupted, and 4-6 bp long, where cleavage occurs at a defined distance 25-27 bp 3â€™ from the unmodi fied recognition site (Roberts and Macelis, 1996; Rao et al. , 2000; Dryden et al. , 2001; Bourniquel and Bickle, 2002). Type III systems consist of two protein subunits, Mod and Res (Rao et al. , 2000; Bujnicki, 2001; Dryden et al. , 2001; Bourniquel and Bickle, 2002; Roberts et al. , 2003). The first subunit (Mod) is the functional analog of the hsdM and the hsdS gene products of the Type I system , providing both the methyla tion as well as the DNA recognition functions of the complex (Rao et al. , 2000; Bujnicki, 2001). The second subunit (Res) is responsible for the restriction endonucl ease activity (Rao et al. , 2000; Bujnicki, 2001). The Type III modification subunit can f unction independently, methylating single recognition sites, and does not have a requirement for the nu mber or orientation of the target recognition sites (Rao et al. , 2000; Bujnicki, 2001; Dryden et al. , 2001; Roberts et al. , 2003; Williams, 2003). Studies have s hown that M.EcoP15I is a dimer (Mod2) in solution (Raghavendra and Rao, 2003). Type I II MTases require AdoMet as a methyl donor (Dryden et al. , 2001; Bourniquel and Bickle, 2002) and are believed to also use AdoMet as a positive allosteric effector, to increase efficiency of binding to its unmodified recognition sequence (Dryden et al. , 2001; Bourniquel and Bickle, 2002). Type III MTases have an unusual requirement of magnesium ions for activity (unlike other DNA MTases) and may use ATP to in crease discrimination between specific recognition sequences from nons pecific sequences (Dryden et al. , 2001; Bourniquel and Bickle, 2002). Furthermore, Type III system s methylate only one strand of the substrate
49 DNA on the N-6 of adenines of the recognition sites (Roberts et al. , 2003; Williams, 2003). Type III REase activity requires both the Mod and Res subunits to form a complex currently believed to be heterotetrameric (Res2Mod2) (Bujnicki, 2001; Bourniquel and Bickle, 2002). Studies have shown that the Type III system R.EcoP15I was found to be degraded when expressed alone (Ragha vendra and Rao, 2003). REase cleavage of dsDNA requires ATP (which is not hydrolyzed unlike Type I systems), magnesium ions, and two unmodified recognition sites in inve rse orientation (Bujnicki, 2001; Dryden et al. , 2001; Roberts et al. , 2003; Williams, 2003). In some Type III systems, AdoMet is absolutely required as a positive allosteric effector for REase activity, and like Type I systems, AdoMet is likely a general requi rement for most Type III systems (Rao et al. , 2000; Bujnicki, 2001; Bourniquel and Bickle , 2002; Raghavendra and Rao, 2003). When cells with Type III systems are grown in the presence of methionine analogs, newly synthesized DNA is degraded and the cells die because only methyl ation is inhibited while the REase remains active (Lark and Arber, 1970). This differs from Type I systems which show no degradation of DNA when the availability of AdoMet is inhibited (Lark and Arber, 1970). Although the Type III MTase only methylates one strand of the target recognition sites, unmodified recognition sites produced af ter replication are in the same orientation, consequently, the requirement of two unmodifi ed recognition sites in inverse orientation for Type III REase activity is not in conflict with protecti ng self-DNA from restriction (Bujnicki, 2001; Dryden et al. , 2001; Raghavendra and Rao, 2003; Roberts et al. , 2003). Two DNA bound Type III R-M complexes tran slocate in an ATP dependent manner
50 toward each other, collide, and cleave (only one of the two sites) at a relatively short fixed distance from the recognition site (inde pendent of the distance between recognition sites) (Bujnicki, 2001; Dryden et al. , 2001; Bourniquel and Bickle, 2002). Cleavage of the dsDNA is believed to be induced by coll ision of the two complexes, where the top strand is cut by the REase bound to the recognition site nearest to the cleavage site and the bottom strand is cut by the REase bound to the furthest recogniti on site (Bujnicki, 2001; Bourniquel and Bickle, 2002). Unlike Ty pe I systems, Type III system enzymes are released and recycled after DNA cleavage (Dryden et al. , 2001). Type III systems: Regulation There are at least two different mechanis ms which regulate the restriction activity of the different Type III R-M systems (Kulik and Bickle, 1996). One type of mechanism is implicated in the chromosomally encoded StyLTI system which cannot be transferred by conjugation into a new host (another strain of S. typhimurium ) because it results in death of the recipient host (Kulik and Bickle , 1996). In contrast, EcoP1I, encoded on a mobile P1 prophage element, and Eco P15I, encoded on a p15B plasmid, can freely transfer their systems to a new host, sugge sting that their rest riction enzymes are controlled by some other mechanism that dela ys restriction but not modification activity (Kulik and Bickle, 1996; Rao et al. , 2000). Transcriptional c ontrol has not yet been shown in Type III systems (Murray, 2002). Type IV Systems Type IV systems general characteri stics: Physiology and function Type IV systems are not classic R-M syst ems that require modification of selfDNA to protect it from restriction by its cogn ate REase, but rather, they are modificationdependent restriction (MDR) systems that only cleave DNA with one of the following
51 modifications: methylation, hydroxymethyl ation, or glucosyl-hydroxymethylation (Roberts et al. , 2003). These systems consist of one or more proteins and have recognition sites which may or ma y not be well defined (Roberts et al. , 2003). Like all the other REases, they require magnesium ions for activity (Bujnicki, 2001; Williams, 2003). The best studied examples of Type IV systems are McrA, McrBC, and Mrr, of E. coli K12. Many ORFs with sequence similarity have been found in other bacteria and archaea (NCBI BLAST; Roberts and Macelis, 1996; Rao et al. , 2000). EcoKmrr and EcoKmcrBC genes flank the Type I system EcoKhsdRMS genes, located at about 98.5 minutes (Bickle and Kruger, 1993; Joset a nd Guespin-Michel, 1993; Jost and Saluz, 1993; Rao et al. , 2000). The mcrA gene (located within an excisable prophage-like element e 14) maps away from the mcrC-mcrB-hsdS-hsdM-hsdR-mrr restrictionmodification gene cluster, at about 25 minutes on the standard map of E. coli K12 (Bickle and Kruger, 1993; Joset and Guespin-Mich el, 1993; Jost and Saluz, 1993; Rao et al. , 2000). The first restriction systems detected in prokaryotes (Luria and Human, 1952) are the MDR systems EcoKMcrA and EcoKMcrBC of E. coli K12 (Piekarowicz, 1991a; Bickle and Kruger, 1993; Bourniquel and Bickle, 2002). The specific MDR endonucleases were named based on the initial ch aracterization of thei r activity: The Mrr proteins are endonucleases involved in met hylated adenine recogni tion and restriction (Waite-Rees et al. , 1991), and the Mcr proteins are e ndonucleases involved in modified cytosine restriction (Bourni quel and Bickle, 2002). Mcr pr oteins, initially named Rgl (restricts â€œglucoselessâ€ bacteriophage), were identified by their ability to restrict
52 nonglucosylated T-even phage DNA (Piekarowicz et al. , 1991a). Wild-type T-even bacteriophage (T2, T4, and T6), which have glucosylated C5-hydoxymethylated cytosine DNA, resists Mcr cleavage when infecting E. coli K12 (Waite-Rees et al. , 1991; Bickle and Kruger, 1993; Rao et al. , 2000; Bujnicki, 2001; Bourni quel and Bickle, 2002). Mcr proteins function to restrict methylcytosine or hydroxyl-m ethylcytosine residues, where the cytosine ring is methylated on either th e C5 or N4 positions or hydroxymethylated on the C5 position (Bickle and Kruger, 1993; Jost and Saluz, 1993; Rao et al. , 2000; Bujnicki, 2001; Bourniquel and Bickle, 2002). McrA, McrBC, and Mrr are all known to cleave dsDNA methylated at 5'-mCG-3' by M.SssI (Bickle and Kruger, 1993). EcoK McrA (RglA) was found to also restrict M.HpaII methylated DNA (5'-CmCG-3'); however, no consensus recognition sequence has been deduced for EcoKMcrA (Bickle and Kruger, 1993; Roberts a nd Macelis, 1996). EcoKMcrBC (RglB) binds to two RmC sites (R = purine) in tandem in the general form 5'-RmC (N40-3000) RmC-3â€™ and cleave s dsDNA at multiple positions 10-50 bp (preferentially ~30bp) away from one of th e RmC sites, in between the two recognition sites (Bourniquel and Bickle, 2002; Pieper et al. , 2002; Roberts et al. , 2003). The RmC sequence can be on one strand of each site, either on the same strand or on opposite strands, or on both strands of each site (Dryden et al. , 2001). EcoKMrr appears to restrict both adenine and cytosine modified DNA, methylated by a variety of known MTases such as M.AccI, M.HhaII, M.HincII, M.Hi nfI, M.HpaI, MNlaIII, MTaqI, M.CviRI, M.PstI, M.HhaI, and M.SssI; however, a cons ensus sequence has not been deduced for EcoKMrr (Waite-Rees et al. , 1991; Roberts and Macelis, 1996).
53 The EcoKmrr gene appears to code for a si ngle MDR endonuclease generally referred to as a Mrr REase, however, it actual ly has two phenotypes referred to as Mrr (methylated adenine restricting) and McrF (m ethylated cytosine rest ricting) (Bickle and Kruger, 1993; Jost and Saluz, 1993). EcoKmcrA appears to encode a solitary MDR endonuclease (Bickle and Kruger, 1993; Rao et al. , 2000). EcoKmcrBC operon consists of two genes, mcrB and mcrC, encoding three polypeptides McrBL (long/full length), McrBS (short/truncated), and McrC (Rao et al. , 2000; Dryden et al. , 2001; Bourniquel and Bickle, 2002). The full length McrBL contains the sequence specific DNA binding site in the N-terminal half, where stab le DNA binding appears to require GTP and magnesium ion, and the central and C-terminal half appears to c ontain the GTP binding site and the McrC binding determinants (Rao et al. , 2000; Bujnicki, 2001; Dryden et al. , 2001; Bourniquel and Bickle, 2002). McrBS is translated from an internal in-frame initiation site, producing a C-terminal 287 aa polypeptide which has no REase activity alone or in the presence of McrC (Panne et al. , 1998). McrBS is proposed to function to sequester McrC, forming inactive McrBSC complexes, thereby modulating the formation of the active complex McrBLC (Panne et al. , 1998). The precise function of McrC is unknown, but sequence analysis reveals a putative DNA cleavag e catalytic center (Rao et al. , 2000; Dryden et al. , 2001; Bourniquel and Bickle, 2002), which is supported by recent mutational analysis studies of th e McrC â€œPD-(D/E) XKâ€ motif common to a variety of nucleases (Pieper and Pingoud, 2002). In the absence of McrC, McrB binds to a single 5â€™-RmC-3â€™ site with a ratio of least four McrB molecules per site, while in the presence of McrC, two McrB bound 5â€™-RmC-3â€™ sites pull together forming the required large complex of unknown stoichiometry (Pieper et al. , 2002). Electr on microscopy
54 reveals that McrBL forms heptameric rings that stack to form a tertradecamer, stabilized by an unknown number of McrC subunits, to produce the active REase complex (Bujnicki, 2001; Bourniquel and Bickle, 2002). While EcoKMcrBC generally requires two modified recognition sites and requires formation of a multimeric complex for REase activity like Type I and Type III systems, it uniquely requires GTP rather that ATP for cleavage (Bickle and Kruger, 1993; Rao et al. , 2000; Bujnicki, 2001; Bourniquel and Bickle , 2002). It has been shown that EcoKMcrBC can cleave circular DNA that ha s only one modified site while linear DNA requires two sites or a blocka ge on the DNA for cleave, indica ting that translocation of DNA between two bound sites eventually stalls inducing DNA cleavage (Dryden et al. , 2001). Like Type I systems, EcoKMcrBC co mplex binds to its recognition sites and translocates DNA, using the energy of GTP hydrolysis, unti l stalling induces cleavage between the two recognition si tes (Bujnicki, 2001; Dryden et al. , 2001). In contrast, EcoKMrr and EcoKMcrA do not appear to re quire NTP hydrolysis for activity (Dryden et al. , 2001). Type IV systems: Regulation Studies on regulation of McrBC activity indi cate that it does not undergo restriction alleviation after treatment with DNA damagi ng agents, unlike Type I systems (Murphy et al. , 2002). In addition, McrB is unaffected by clpX mutations and moderately affected by clpA or clpP mutations, suggesting that McrB is a substrate fo r ClpAP protease (Murphy et al. , 2002). Unclassified Systems Several restriction-modificat ion systems have been found that are unique and are currently unclassified. For example, Lla I is a R-M system whose properties are an
55 intermediate between Type IV and IIS, consisting of four protein subunits: an independent Type IIS-like DNA MTase, with two methyltransferas e domains due to a gene fusion, and a REase requiring three diffe rent protein subunits, similar to McrBC, where one is required for cleavage and the other two to modify cleavage (Oâ€™Sullivan et al. , 1995; Bujnicki, 2001). This system also e xpresses a regulatory protein (C.LlaI) (with sequence similarity to Type II C proteins) th at has been shown to positively regulate expression of the REase post-transcriptionall y, unlike the C proteins found in many Type II systems that appear to by transc riptional activators (Oâ€™Sullivan et al. , 1995; Bujnicki, 2001). There is a MDR system named PvuRts1I, found on the Rts1 plasmid isolated from Proteus vulgaris, which is so unique that it is referr ed to as a â€œweirdoâ€ in the REBASE Database (Roberts and Macelis, 1996). PvuR ts1I is a hydroxymethylcytosine-specific restriction system and can cleave these site s even if they are glucosylated (unlike EcoKMcrBC); however, its consensus recogni tion sequence has yet to be determined (Janosi et al. , 1994; Roberts and Macelis, 1996). It wa s found to specific ally restrict Teven bacteriophage (T2, T4, and T6) th at contains DNA with glucosylated hydroxymethylcytosine residues (Janosi et al. , 1994). In general, these T-even bacteriophage use beta-glucosyltransfe rase mediated glucosylation of the hydroxymethylcytosine residues to provide protection of the phage DNA from restriction while allowing phage enzymes to discrimina te phage and host DNA so it can selectively degrade host DNA (Morera et al. , 1999; Murray, 2000). It appears that PvuRts1I may have evolved in response to this se lective pressure. Upstream of the PvuRts1I is a gene with amino acid similarity to MTases, but no DNA methyltransferase activity has been
56 detected (Janosi et al. , 1994). Perhaps PvuRts1I evol ved from a classic Type II R-M system. Homing Endonucleases Homing endonucleases have been found in all three biological kingdoms and are encoded by introns, inteins, or are free-s tanding genes (Belfort and Roberts, 1997; Roberts et al. , 2003). While the vast majority of these sequence spec ific endonucleases have been found in eukaryotes (in chromoso mal, extra-chromosoma l, mitochondrial, or chloroplast genes), several have also been found in bacteria, archaea, and bacteriophage (Bujnicki, 2001; Chevalier and Stoddar d, 2001; Williams, 2003). Homing REases are relatively small, typically less than 40 kDa (Chevalier and Stoddard, 2001) and are classified into to four families based on the conserved amino acid sequence motifs of their catalytic domains (Belfort and Roberts, 1997; Roberts et al. , 2003). The current nomenclature for homing REas es, patterned after restriction system enzymes, designates intron encoded homing REas es with the prefix I (for intron), the intein encoded homing endonuclease with the pref ix PI (for protein in sert), and the freestanding homing endonucleases with the prefix F or the prefix Endo (which has been used by some) (Belfort and Roberts, 1997; Roberts et al. , 2003). These REases are called homing endonucleases because of their genera l ability to make double stranded breaks and promote recombination with a homologous allele (homing site) lacking the intron or intein, resulting in duplicati on of the intron or intein (Belfort and Roberts, 1997; Bujnicki, 2001). Studies have shown that a great many introns and inteins, many of which are mobile, contain internal ORFs, suggesting that homing is widespread (Chevalier and Stoddard, 2001). Mobile intron s are genetic elements that disrupt open reading frames (ORFs) and are able to se lf-splice its intron mR NA sequence from the
57 host mRNA sequence after transcription of the host ORF (Chevalier and Stoddard, 2001; Gogarten et al. , 2002). Mobile inteins (internal protei ns) are genetic elements that disrupt ORFs and are able to excise the intein polypeptide from the host polypeptide sequences (exteins) after transl ation (Chevalier and Stoddard, 2001; Gogarten et al. , 2002). The two host exteins are then spliced together to produce the functional host protein (Gogarten et al. , 2002). The recognition sequences of homing REases are asymmetric, between 12-40bp long, and are usually bipartite (interrupted by nonspecific ba ses) (Belfort and Roberts, 1997; Gogarten et al. , 2002). While their target sequence specificity is less stringent than other REases (Chevalier and Stoddard, 2001; Gogarten et al. , 2002), single base changes result in varying loss of efficiency (Willia ms, 2003). Some homing REases have been found to cleave dsDNA as a monomer while ot hers function as a homodimer (Belfort and Roberts, 1997; Williams, 2003). Homing REase are encoded within â€œparasiticâ€ mobile genetic elements (a self-splic ing intron or intein) in the middle of their own recognition site (Chevalier and Stoddard, 2001; Gogarten et al. , 2002; Burt and Koufopanou, 2004). The intron or intein encoded homing endonucle ase specifically binds to and cleaves its recognition site, producing double stranded breaks in the DNA that typically induces the homologous recombination repair mechanisms of the host, resulting in lateral transfer of the intron to the sec ond allele to produce duplicate intr ons or inteins located in both alleles (Chevalier and Stoddard, 2001; Burt and Koufopanou, 2004). Most of the homing REases currently id entified are located in group I introns (Chevalier and Stoddard, 2001), typically with in essential genes (Burt and Koufopanou, 2004). Similarly, the homing REases located in inteins are typically found in conserved
58 genes of the mitochondria, chloroplast, or chromosome (most commonly in genes involved in DNA replication and nu cleic acid metabolism) (Gogarten et al. , 2002). Group II introns, containing homing REases , are usually found on plasmids and transposons (Burt and Koufopanou, 2004). Gr oup II intron encoded homing REases have maturase activity and form complexes with their respective intron mRNA, creating a ribonucleoprotein (RNP) that aids in splic ing the intron mRNA from the host mRNA (Chevalier and Stoddard, 2001; Burt and Koufopanou, 2004). The RNA component of the RNP mediates DNA binding by base pairing with the targ et recognition site and â€œreverse splicesâ€ into one strand of the hom ing REase recognition site, while the protein component mediates cleavage of the othe r strand followed by DNA synthesis using the intron mRNA as template by the reverse tran scriptase activity of the protein component of the RNP (Belfort and Roberts, 1997; Chevalier and Stoddard, 2001). Like most REases, homing endonucleases require diva lent metal ions for activity (typically magnesium) (Chevalier and Stoddard, 2001; Williams, 2003). Solitary DNA Methyltransferases Not all prokaryotic DNA MTases are part of a restriction-modification system pair (Wilson, 1991; Jost and Saluz, 1993; Hyone -Myong, 1996; Roberts and Macelis, 1996; Bujnicki, 2001; Low et al. , 2001; Jeltsch, 2002). Many solitary DNA MTases have been discovered that contribute to the DNA methylation pattern of the organism, such as Dam, Dcm, CcrM, and the â€œantirestrictionâ€ MTases of various bacteriopha ges (Bujnicki, 2001). Dam and Dcm of Escherichia coli The Dam (DNA adenine-methylation) system of E. coli K12 is known to be involved in several cellular pr ocesses, including coordina te control of replication initiation, methyl-directed mismatch repair (M MR), regulation of transposition of certain
59 transposons, regulation of gene expression at the transcripti on level, bacterial virulence (Arraj et al. , 1990; Jost and Saluz, 1993; Hyone-Myong, 1996; Heithoff et al. , 1999; Bujnicki, 2001; Low et al. , 2001), very short patch (VSP) mismatch repair (Bell and Cupples, 2001), chromosome structure (L obner-Olesen, 2003), and conjugation (Camacho and Casadesus, 2002). Currentl y, the Dcm (DNA cytosine-methylation) system has only been found to be necessary in VSP mismatch repair in conjunction with the Vsr nuclease (related to REases but with less stringent specif icity), which cleaves DNA at T/G mismatches in E. coli , producing a single-stranded nick (Jost and Saluz, 1993; Macintyre et al. , 1997; Bujnicki, 2001; Gonzalez-Nicieza et al. , 2001). It is interesting to note that, similar to the geneti c linkage typical of most R-M systems, the dcm MTase gene is located upstream ( vsr overlaps dcm in a +1 reading frame) of its partner vsr endonuclease gene (Macintyre et al. , 1997; Gonzalez-Nicieza et al. , 2001). Early studies have shown that deletion of the dam or dcm gene has no effect on the restriction-modification ability or the viability of the E. coli K12 (Arraj et al. , 1990; Hyone-Myong, 1996). Although dam was shown to be unessential in E. coli K12 alone, it was shown that particular recombination genes ( recA, recB, recC, ruvA, ruvB , ruvC, and possibly recG ) are required for cell viability of dam mutants, explaining the sensitivity dam mutants to DNA damage (Marinus, 2000) . More recent studies have shown that Dam is essential for cell viability of Yersinia pseudotuberculosis and Vibrio cholerae (Julio et al. , 2001). Dam belongs to the alpha stru ctural subclass of N-MTases (Low et al. , 2001; Jeltsch, 2002). In addition, Dam homo logues are primarily found in gamma proteobacteria (Kossykh and Lloyd, 2004). Ec oKDam methylates the N6-adenine of its
60 recognition site 5â€™-GATC-3â€™ sites, while Ec oKDcm methylates the internal C5-cytosine of its recognition site 5â€™-CC(A/T)GG-3 â€™ (Jost and Saluz, 1993; Hyone-Myong, 1996; Low et al. , 2001; Bujnicki, 2001). Dam functions as a monomer and requires AdoMet as a methyl donor (Nwosu, 1992; Hyone-Myong, 1996). Dam methylation is not entire ly random (Hyone-Myong, 1996; Low et al. , 2001). Sites with certain flanking sequences, su ch as 5â€™-gGATCa-3â€™, are preferentially methylated by Dam, compared to 5â€™-aGATCg-3â€™ (Hyone-Myong, 1996). Dam also methylates dsDNA preferentially over ssDNA and hemi-methylated DNA preferentially over unmethylated DNA (Hyone-Myong, 1996). Id eally, unmethylated sites only exist after replication and are Dam methylated after post-replicat ive mismatch repair of the hemi-methylated daughter strands occur (Jos t and Saluz, 1993). Most undermethylated 5â€™-GATC-3â€™ sites are found within noncoding regions, which are more likely involved in regulation (Low et al. , 2001). Methyl-directed mismatch repair (MMR) of unmethylated DNA occurs without strand preference, while repair of hemi-methylated DNA is biased toward the unmethylated strand (Jost and Salu z, 1993). Repair of fully methylated DNA requires repair pathways other than MMR (J ost and Saluz, 1993). Mismatch MutHSL excision (related to REases) distinguishes post-re plicative daughter from parent strands by its Dam methylation (Bujnicki, 2001). In E. coli K12, dam mutants have a 10-100 fold increase in the number of spontane ous mutations, while enhanced methylation interferes with repair and re sults in increased number of s pontaneous mutations (Jost and Saluz, 1993; Heithoff et al. , 1999). Studies in E. coli K12 on the involvement of Dam in MMR and the involvement of Dcm in VSP mi smatch repair originally lead to the supposition that Dam was not involved in V SP mismatch repair; how ever, direct studies
61 later showed that Dam is also required in VSP mismatch re pair (Bell and Cupples, 2001). Another study showed that plasmids cont aining multiple short direct repeats were unstable, resulting in propagation of plasmids with deletions of long stretches of whole repeat units, unless propagated in dam mutants (Troester et al. , 2000). This deletion mechanism appeared to be independent of recA -dependent recombination, MMR, and methyl-dependent restriction, suggesting the existence of an unknown Dam-dependent deletion mechanism (Troester et al. , 2000). Dam has been shown to be required for coordinate replication of E. coli K12, where Dam overproduction or deficien cy results in random or asynchronous initiation of replication (Boye and Lobner-Olesen, 1990; Campbell and Kleckner, 1990; Jost and Saluz, 1993; Lu et al. , 1994; Lobner-Olesen et al. , 2003). Coordination of initiation of chromosomal replication is regulate d by the methylation state of the oriC and dnaA regions (Boye and Lobner-Olesen, 1990; Camp bell and Kleckner, 1990; Jost and Saluz, 1993; Lobner-Olesen et al. , 2003). Only when these regi ons are fully methylated will dnaA be efficiently transcribed to produce enough DnaA to initiate replication at oriC (Campbell and Kleckner, 1990; Jost a nd Saluz, 1993). Hemi-methylated oriC and dnaA binds to outer membrane proteins, with the he lp of SeqA and possibl y other sequestration requiring proteins, preventing complete methylation of oriC and dnaA (Campbell and Kleckner, 1990; Jost and Saluz, 1993; Lu et al. , 1994; Lobner-Olesen et al. , 2003). This state interferes with both rein itiation of replica tion for about 30-40% of the cell cycle, and efficient transcription of dnaA, drastically decreasing the concentration of DnaA (Boye and Lobner-Olesen, 1990; Campbell and Kl eckner, 1990; Jost and Saluz, 1993). Release of o riC and dnaA is due, at least in part, to slow methylation of the 5â€™-GATC-3â€™
62 sites, leading to reduc ed binding affinity between the out er membrane proteins and the dnaA and oriC regions (Campbell and Kleckner, 1990; Jost and Saluz, 1993). When multiple origins of replication are present, initiation occurs simultaneously at each site but only once per cell cycle (Boye and Lobner-Olesen, 1990; Campbell and Kleckner, 1990; Jost and Saluz, 1993; Lu et al. , 1994). Of eleven 5â€™-GAT C-3â€™ sites found in the â€œminimalâ€ oriC region of E. coli K12, eight appear to be conserved in several other related organisms (Nwosu, 1992; Jost and Saluz, 1993; Lu et al. , 1994). Globalmicroarray analysis of E. coli gene expression showed that , while surprisingly few genes were detected that show ed altered expression in dam mutants, dam overexpression affected the overall transcriptional activit y in localized region s of the chromosome (Lobner-Olesen, 2003). Overexpression of dam resulted in increased transcriptional activity in regions of low expr ession and decreased transcript ional activity in regions of high expression, suggesting that Dam methylation is impor tant in forming and/or maintaining chromosome structure (Lobner-Olesen, 2003). Dam methylation also affects replicati on of ColE1-type plasmids and RepI encoding plasmids (Jost and Saluz, 1993). ColE 1plasmids have three 5â€™-GATC-3â€™ sites in or near the RNAII prom oter region (Patnaik et al. , 1990; Jost and Saluz, 1993). Only when they are hemi-methylated do they have an inhibitory effect on replication via a mechanism similar to the one described for chromosomal replicati on coordination where the oriC of the plasmid is likewise bound to the membrane when hemi-methylated (Boye and Lobner-Olesen, 1990; Patnaik et al. , 1990; Jost and Saluz, 1993). RepI encoding plasmids cannot be replicated unless they ar e fully or hemi-methylated at the multiple GATC sites at the ori region (Jost and Saluz, 1993). Dam methylation facilitates both
63 RepI binding and st rand opening at the ori region of RepI encoding plasmids (Jost and Saluz, 1993). Transposition of transposons, such as IS 10/Tn10 and IS50/Tn5, are also regulated by Dam methylation, where hemi or unmethylat ed 5â€™-GATC-3â€™ sites in the transposable elements allows transposition (Campbell and Kleckner, 1990; Jost and Saluz, 1993). A hyperactive mutant Tn5 transposase has been isolated that prefers to transpose Dam methylated Tn5 transposons, rather than be ing inhibited by methyl ation, apparently by altering its specificity (a mutation that alleviat es the problem of steric hindrance normally caused by methylation of its DNA recogniti on site) (Naumann and Reznikoff, 2002). Some bacterial species, such as E . coli and S. typhimurium, code for the pyelonephritis-associated pili ( pap ) operon, an important virulence determinant in urinary tract infections (Heithoff et al. , 1999; Low et al. , 2001). Dam is known to regulate Pap pili gene expression in these organisms (Heithoff et al. , 1999; Weyand and Low, 2000; Low et al. , 2001). Specifically, th e binding of the global regulator Lrp to the pap regulatory sequences depends on the methylation pattern in the pap control region (Heithoff et al. , 1999; Weyand and Low, 2000; Low et al. , 2001). It appears th at several regulatory proteins (such as Lrp, GutR, H-NS, and OxyR ) binding to their regulatory sites depends on or alters the Dam methylati on pattern of those sites (Low et al., 2001). In E. coli , Pap Pili gene expression is regulated by the met hylation pattern of two GATC sites in the control region of the papBA operon (Heithoff et al. , 1999; Weyand and Low, 2000; Low et al. , 2001). The Pap pili are produced when the 5â€™-GATC-3â€™ site proximal to the papBA promoter is methylated and the 5â€™-GATC-3â€™ site di stal to the promoter is unmethylated while the Pap p ili are not produced during the reverse methylation pattern
64 (Weyand and Low, 2000). This discovery le d to studies on the role of DNA adenine methylation in bacterial virulence (Heithoff et al. , 1999). It was discovered that adenine methylation, but not cytosine methylation, is required for Salmonella pathogenesis via direct control of virulenc e gene expression (Heithoff et al. , 1999; Low et al. , 2001). It was later found that the conjugal transfer of the virulence plasmid (pSLT) of Salmonella enterica is activated by Lrp and repr essed by Dam methylation (Low et al. , 2001; Camacho and Casadesus, 2002). Dam is a global regulator and has been s hown to control expression of at least 20 in vivo -induced genes in Salmonella (Heithoff et al. , 1999; Low et al. , 2001). In contrast (rather than requiring Dam for virulence), group B pathogenic Neisseria meningitidis strains are dam mutants (Low et al. , 2001). Genetic analys is indicate that Neisseria meningitidis dam mutants had apparently replaced dam with a Type IIM methyldependant REase gene, bordered by pseudo-tran sposable small repeat elements (SREs), with the same site specificity as Dam (Cantalupo et al. , 2001), Consequently, these dam mutant strains now restrict Dam methylated DNA. Furthermore, various species of Neisseria have been found with R-M systems that recognize and methylate 5â€™-GATC-3â€™ sites (Roberts and Macelis, 1996). CcrM: A cell cycle regulating MTase The solitary MTase CcrMI was first identified in Caulobacter crescentus (Zweiger et al. , 1994), an unusual dimorphic fr eshwater bacterium that as ymmetrically divides into two different cell types, a nonr eplicating swarmer cell and th e replicating stalked cell (Poindexter, 1981). CcrMI (cell cycle regula ting DNA methyltransferase) is a 39-kDa protein with 49% identity to the Haemophilus influenzae Type II adenine MTase M.HinfI, having the same specificity as M.Hinf I, and is reported to methylate the adenine
65 of 5â€™-GANTC-3â€™ sites in C. crescentus (Reisenauer et al. , 1999b). Although somewhat similar in function, the beta subclass N-MTase CcrMI of C. crescentus is unrelated to the alpha subclass N-MTase Dam of E. coli (Reisenauer et al. , 1999b). CcrMI preferentially binds to hemimethylated DNA and appears to processively methylate its target recognition sites (Berdis et al. , 1998). Based on experiments on the physical properties of CcrMI, it appears that CcrMI naturally forms dimers, but binds to DNA and is active as a monomer, suggesting that dimerization may be a mechanism used to prevent premature methylation (Shier et al. , 2001). Early studies in C. crescentus of the promoter region of ccrMI revealed similarity to some of the temporally controlled class II flagellar promoters (Stephens et al. , 1995). CcrMI promoter regi on includes several inverted repeats (IR1, IR2, IR3) and four CcrMI me thylation sites (two found in IR1 and two found in IR3) (Stephens et al. , 1995). Like these particular class II flagellar promoters, it appears that ccrMI transcription is repressed wh en DNA replication is inhibited (Stephens et al. , 1995). Mutational analysis indicates that methylation of the CcrMI sites in IR3 is not essential for ccrMI expression but does appear to affect the temporal repression of ccrM , suggesting possible CcrMI autoregulation (Stephens et al. , 1995). IR1 does not appear to be essential for CcrMI activity while IR2 mutation (without disrupting the sequences conserved with class II flagellar promoters) appears to affect promoter activity (activation) (Stephens et al. , 1995). Of the ~12194 statistically expected 5 â€™-GANTC-3â€™ sites, only ~4495 are found in the C. crescentus genome, with a greater concentrati on found in the intergenic regions (Reisenauer and Shapiro, 2002). Based on evidence of CcrMI autoregulation and the identification of 5â€™-GANTC-3â€™ sites in the pr omoter of the global ce ll cycle transcription
66 regulator (CtrA), studies were done to investigate CcrM involvement in regulating CtrA expression C. crescentus (Reisenauer and Shapiro, 2002). These studies showed that complete CcrMI methylation represses ctrA transcription from its P1 promoter (Figure 1). (Reisenauer and Shapiro, 2002). Temporal c ontrol of the state of methylation of the ctrA P1 promoter depends on its position relative to the origin of replic ation (Reisenauer and Shapiro, 2002). As replication progressively converts the genome from fully methylated to hemimethylated, ctrA is released from the inhibitory effect of methylation (Reisenauer and Shapiro, 2002). CtrA is part of a tw o component phosphorelay signal transduction system and was determined to activate transcription of ccrMI in C. crescentus (Quon et al ., 1996; Reisenauer et al. , 1999a) (Figure 1). Microarray an alysis indicates that CtrA is involved, directly or indirect ly, in regulating the expressi on of 26% of the cell cycle regulated genes of C. crescentus (Brun, 2001). The phosphorylated form of CtrA appears to directly regulate at least 95 genes (such as ccrMI ) in C. crescentus (Laub et al. , 2002). Figure 1. Proposed CcrMI autoregulation and regulation of CtrA expression in Caulobacter crescentus (as reviewed by Marczynski and Shapiro, 2002).1 P1 P2 ctrA ccrMI CtrA CtrA-P CcrMI DNA Methylation ClpXP mediated degradation Repression Activation Repression of Cori chromosomal DNA replication initiation Regulate expression of various genes Lon mediated degradation
67 Although the mechanism(s) by which CcrM methylation regulates the cell cycle has yet to be identified, the following evid ence suggests that the presence and essential activity of CcrM is strictly controlled (Figure 1). CcrMI is only present in C. crescentus during the predivisional stage of the cell cycle (end of S-phase), after replication initiation (Stephens et al. , 1996). The global cell cycle tr anscriptional regulator CtrA binds to the origin of replication ( Cori ), inhibiting initiation of DNA replication (Reisenauer et al. , 1999a). During early predivisional st age, CtrA is cleared from the cell by ClpXP protease mediated degradation (Marczynski and Shapiro, 2002), allowing replication initiation if the DNA is fully methylated (Reisenauer et al. , 1999a, 1999b). Once replication starts, CtrA accumula tes and initiates transcription of ccrMI to methylate the now hemi-methylated DNA (Reisenauer et al. , 1999a). Upon complete chromosomal methylation, transcription of ccrMI is again repressed (Wright et al. , 1997). During cell division, CcrMI is constitutively degraded, mediated by a Lon-like protease, and cleared from the cell (Wright et al. , 1996). Genomic analysis reveals that CcrM sites are found in higher frequency around the origin of replication in C. crescentus ( Cori ) suggesting that CcrM methylation may be i nvolved in timing and c ontrol of re plication initiation (Reisenauer et al. , 1999b; Marczynski, 1999). In early studies of C. crescentus mutants, where CcrMI is pr esent throughout the cell cycle and a fully methylated chromosome is maintained, inspection of cells grown in culture reveal abnormal cell morphology and aberrant cell division (Zweiger et al. , 1994). About 50% of exponentially growing C. crescentus cells display a wide variety of morphological abnormalities such as twisted misshapen cells and elongated cells (up to three times as long as wild-type cells) (Zweiger et al ., 1994). In stationary phase, the majority of the
68 cells (~90%) show abnormal morphology (Zweiger et al ., 1994). Normally, about 7% of wild-type C. crescentus cells are elongated in st ationary phase (Zweiger et al ., 1994). In wild-type C. crescentus , chromosomal replication init iates only once per cell cycle (Marczynski, 1999). Around 29% of the cells in culture have 1 copy of the chromosome, and 71% of the cells have two copi es of the chromosome (Zweiger et al ., 1994). In contrast, most of the ccrMI overexpression mutants appear to contain two (64.5% of the cells) or three (22.6% of the cells) copies of the chromoso me, which suggests a relaxation in control of replicat ion initiation (Zweiger et al. , 1994). It was later shown that CcrMI is essential for cell viability in C. crescentus , making it the first reported essential solitary DNA MTase found in prokaryotes (Stephens et al. , 1996). Since replica tion is restricted to fully methylated chro mosomes and knockout of the ccrM gene results in cell death, it can be presumed that cell death is du e to arrest of replication (Stephens et al. , 1996; Wright et al. , 1997; Marczynski, 1999; Robertson et al. , 2000; Kahng and Shapiro, 2001). Twenty members of the alpha subdivision of proteobacteria, id entified as having possible CcrMI homologues (Stephens et al. , 1996), were shown to methylate the CcrMI recognition site when assayed by chromo somal DNA resistance to HinfI digestion (Reisenauer et al. , 1999b). In addition, the C. crescentus ccrMI gene hybridized to DNA from these twenty bacteria but not to DNA from bacteria of ot her proteobacterial subdivisions (Reisenauer et al. , 1999b). The full-length ccrM genes were cloned from the soil bacterium R. meliloti , the animal pathogen B. abortus (Wright et al. , 1997) and the plant pathogen A. tumefaciens (Kahng and Shapiro, 2001). Like in C. crescentus , CcrM methylation is essential for cell viability and appears to be involved in regulation
69 of the cell cycle of R. meliloti (Wright et al. , 1997), B. abortus (Robertson et al. , 2000), and A. tumefaciens (Kahng and Shapiro, 2001). In a ddition, overexpression of CcrM produces aberrant cell division, increased DNA content, and aberrant cell morphology such as thickening, branching, a nd some degree of elongation in R. meliloti (Wright et al. , 1997), B. abortus (Robertson et al. , 2000), and A. tumefaciens (Kahng and Shapiro, 2001). Similar to the role of Dam methylation in th e virulence of Salmonella typhimurium , CcrM methylation in B. abortus appears to be involved in its ability to survive as an intracellular pathogen (Robertson et al. , 2000). Having found similar characteristics among the C. crescentus CcrMI homologues that have been studied, the putative CcrM homologues found in the members of other alpha subdivision proteobacteria are likely to ha ve a similar involvement in regulating cell cycle functions (Wright et al. , 1997; Reisenauer et al. , 1999b). The identification of CcrM, CtrA, and CtrA targets in many alpha pr oteobacteria (Bellefontaine et al. , 2002; Hallez et al. , 2004), and comparative morphological analysis of C. crescentus, B. abortus, S. meliloti, and A. tumefaciens that indicate that they each have some level of functional and morphological asymmetry (Hallez et al. , 2004), seem to suggest that the lifestyles among alpha proteobacteria of requiring adaptation to unique environm ents to various degrees (depending on the organism) is the common fact or pressuring conservation of a complex CtrA coordinated cell cy cle regulatory network. Genomic sequence analysis of C. crescentus identified four long motifs (>100bp) containing intergenic repeats associated with CcrMI (5â€™-GANTC-3â€™) recognition sites (Chen and Shapiro, 2003). Analysis of the Brucella melitensis abortus genome showed that, like the C. crescentus genome, it contains similar intergenic motifs in which all had
70 two inverted repeats, flanking a conserved sequ ence, with a CcrM site centered in one of the inverted repeats (Chen and Shapiro, 2003). It was suggested that these motifs may be involved in CcrM regulation of the associated genes or may in stead have a role in gene maintenance since they are conserved in C. crescentus and B. abortus (Chen and Shapiro, 2003). Genomic analysis of C. crescentus , A tumefaciens, B. melitensis (abortus), M. loti, S. meliloti, and R. prowazekii (which does not have a CcrM homologue) revealed conserved 15-mers centered on CcrM sites, wh ere almost all were found in intergenic regions (Chen and Shapiro, 2003). More recently, a solitary CcrM DNA MT ase M.EcoKCcrM, originally named the putative yhdJ MTase, was cloned and characterized in Escherichia coli (Kossykh and Lloyd, 2004). It appears to methylate the firs t adenine of its rec ognition site ATGCAT (Kossykh and Lloyd, 2004). Like the CcrMI of C. crescentus , it is cell cycle regulated and essential for cell viab ility (Kossykh and Lloyd, 2004). With overexpression, DNA content increases in the cell, growth slow s, and cell morphology becomes abnormal (with a significant portion of the cell population appearing elongated) (Kossykh and Lloyd, 2004). Sequence analysis reveals putative homologues in other gamma proteobacteria (Kossykh and Lloyd, 2004). Solitary antirestriction MTases Some bacteriophage and conj ugative plasmids have evolved mechanisms to avoid host REase restriction such as (1) a lack or rarity of host restric tion sites via counterselection, (2) m5C hydroxymethylated DNA in T-even phage which is unrecognized by MDR REases other than McrBC, (3) Type I a nd III inhibitor antirestriction proteins Ocr in T7 phage, (4) Type I inhibitor antirestricti on proteins DarA and DarB in P1 phage, (5) the plasmid borne Type I inhibitor antires triction Ard family proteins, (6) Ral in phage
71 that acts as a Type IA MTase stimulator, and (7) several Bacillus subtilis bacteriophage encoded DNA MTases (Bickle and Kruger, 1993). The Bacillus subtilis bacteriophage encoded m5C-MTases typically have multi-specific recognition sites (methylating different, sequence unrelated, recognition sites), protecting the infecting bacteriophage from the host R-M systems (Bickle and Kruger, 1993; Bujnicki, 2001). Genome analysis of the number of short palindromic sequen ces found in bacteria suggest that some bacteria, like some bacteriophage and plasmids , have evolved to avoid REase cleavage by eliminating palindromic restriction site s (Kobayashi, 2001; Fuglsang, 2003). For example, GC rich palindromes are underre presented in the genomically GC rich E. coli and P. aeruginosa , but their Type II R-M systems mo re commonly recognize GC rich palindromes (Fuglsang, 2003). In cont rast, the genomically AT rich C. perfringens (and possibly B. subtilis ) has underrepresented AT rich pa lindromes (Fuglsang, 2003). In some organisms (such as B. subtilis and C. perfringens , but not E. coli and P. aeruginosa ), hexapalindromes appear to be even ly distributed along the coding regions and dramatically drop in number in th e Shine-Dalgarno region (Fuglsang, 2003). Evolution and the Taxonomic Relationships Not all strains of a species possess an R-M system. It was estimated that 1 in 4 bacteria have REases (Williams, 2003). So far, Type II systems have been found in 12 out of 13 phyla of bacteria and archaea, the current exception being Spirochaetes (Roberts and Macelis, 1996). Among the char acterized REases currently listed in the REBASE Database (Roberts and Macelis , 1996), ~97.9% belong to Type II, ~1.8% belong to Type I, and <0.5% belong to T ype III and IV systems. Including putative REases, ~12.8% belong to Type I, ~79.9% be longs to Type II, and ~7.3% belongs to Type III and IV systems (Roberts and Macelis, 1996). It should be noted that some
72 putative REases and DNA MTases genes had la ter been determined to not produce active enzymes (Bujnicki, 2001; Roberts et al ., 2003). The biased intere st in Type II systems, due to their usefulness as molecular tools, has been suggested as a possible reason why the majority of R-M systems identified to date belong to the Type II classification (Joset and Guespin-Michel, 1993). Alternatively, the Type II bias may be due to Type II REases being much easier to identify by scr eening crude cellular extracts for sequence specific REase activity, because they produce defined restriction patterns (Raleigh and Brooks, 1998). Then again, there may be a selective advantage for R-M system to develop more strictly defined DNA recogniti on and cleavage, as found in the Type II systems (Joset and Guespin-Michel, 1993). Great REase diversity is commonly f ound within a species. Among different strains of Escherichia coli for example, there are at least six different R-M classes, including Types I, II, III, and three di fferent MRD types (Raleigh and Brooks, 1998). Despite a wide genetic diversity and a general lack of significant amino acid similarity among most REases, there is a significant limitation to the number of actual unique recognition site specificities th at have been found, when comp ared to all statistically possible ones (Roberts and Macelis, 1996). For example, 3684 characterized Type II REases with a total of 256 dis tinct recognition site specificities (among tens of thousands of possibilities) have been found, where only 17 are distinct four base uninterrupted recognition site specificities (among 256 possi bilities) (Roberts and Macelis, 1996). Although the total number of dist inct specificities found is lim ited, there appears to be a bias toward systems having more specifi c and longer (more exclusive) recognition
73 sequences, suggesting the presen ce of selective pressure during the course of R-M system evolution. Type I Systems Of the Type I R-M systems that have been characterized, the sequence homology between the four different families (A, B, C, and D) is restricted to short amino acid sequence motifs such as (1) two motifs in the HsdM subunit commonly found in adenine methyltransferases, one that binds to AdoMet and one that is involved in catalysis of DNA methylation; (2) two motifs in the HsdR subunit, one for ATP-binding and one in common with helicases; and (3) regions of conserved DNA sequence homology in the hsdS specificity gene theorized to code fo r domains responsible for protein-protein interactions of the HsdS sp ecificity subunit with the Hs dM DNA methyltransferase and HsdR endonuclease subunits (Rao et al. , 2000). The C-terminus of HsdR subunit appears to contain the domain for protein-protein inte ractions required to form the active REase holoenzyme (Bujnicki, 2001). In cont rast, within a Type I family, the hsdR and hsdM genes are largely homologous (Rao et al. , 2000). The putative control protein, encoded by the hsd C gene of E. coli C, has no DNA sequence similarity to the host hsdRMS genes (Prakash-Cheng et al. , 1993a). The hsdS specificity gene has two large re gions that are generally nonhomologous within each family (Rao et al. , 2000). There are also some repeated sequences in the hsdS genes of all the families believed to be remnants of gene duplication events during evolution (Rao et al. , 2000). Each target sequence of the bipartite recognition site of most Type I systems has its own corres ponding target recognition domain on the HsdS subunit (Bujnicki, 2001). These variable regions code for the target recognition DNA binding domains, one for each half of the bipartite rec ognition site (Rao et al. , 2000).
74 The C-terminal domain binds to the 3â€™ tetraor pentanucleotide specific sequences while the Nterminal domain binds to the 5â€™-tri nucleotide specific sequences (Wilson, 1991; Rao et al. , 2000). It should be mentioned that isoschizomers in Type I systems often have some nucleotide sequence similarity in the variable regions of the HsdS subunit that bind the target recognition sequence, irre spective of family classification (Rao et al. , 2000). Recombination within the conserved region betw een the target recognition domains have been shown to cause a change in the length of the nonspecific spacer sequence of the recognition site, while rearra nging the target rec ognition domains have been shown to produce new recognition specificities (Rao et al. , 2000). The hsd genes of families A, B, and D are allelic and are found linked to the serB genes at 98.5 minutes on the standard chromosomal map of E. coli K12 strains (Rao et al. , 2000). Because the insertion site of the conjugative transposon-like IncJ element has been recently mapped to a region located in between the uxuA and the serB loci (98.099.5 minutes) (nearer the serB loci) in E. coli K12 (Murphy and Pembroke, 1999), the region located near serB may be a region with higher fre quency of genetic rearrangement, which would help explain the genetic divers ity as well as the evidence of frequent horizontal transfer of the hsd genes located in this region (Oâ€™Neill et al. , 1997). It is believed that there is likely only one inse rtion site for IncJ (Murphy and Pembroke, 1999). In contrast, the hsd genes of family C are generally found on plasmids (Rao et al. , 2000). Type II Systems Type II R-M system REases and MTases each autonomously recognize their target recognition sites. Of the Type II restric tion-modification systems that have been sequenced, no significant ami no acid homology has been found between cognate pairs of
75 REases and MTases (Wilson, 1991; Bujnicki, 2001). Although they recognize the same target sequence, their proposed target recogn ition domains are diffe rent, suggesting that these cognate enzymes pairs are unrelate d, independently evolved, and that they recognize targets in different manners (Wilson, 1991; Raleigh and Brooks, 1998; Bujnicki, 2001). As stated previously, ther e is no homology between the different Type II REases known to date and no conserve d sequence motifs, with few exceptions (Wilson, 1991; Bujnicki, 2001). The few exceptions to the complete lack of amino acid sequence similarity among Type II REases are between isoschizomers (Wilson, 1991; Bujnicki, 2001). It is important to note that many Type II REase isoschizomers have been found to only be isofunctional a nd not homologues (Wilson, 1991; Roberts and Macelis, 1996). This general lack of sequence similarity had lead to the orig inal supposition that REases evolved independently and not by divergence from a common ancestor by mutations in its target recognition domain (Wilson, 1991; Bujnicki, 2001). However, comparative genomic and protein structure anal ysis suggest that members of structurally related REase families probably evolved from common ancestors and diverged (Kovall and Matthews, 1998; Bujnicki, 2001). Type II REases may either have a single continuous target recognition domain fused to the catalytic domain, or a series of separate target recognition elements that project fr om the protein via elongated loops from the catalytic center, or some vari ation of both (Bujnicki, 2001). Structural analysis revealed that Type II REases have a common structural core and catalytic site referred to as â€œPD(D/E)XKâ€ (for its weakly c onserved catalytic site signatu re) (Bujnicki, 2001). This common structural characteristic is shared by a large variety (a superfamily) of nucleases
76 such as exonuclease, MutH, Vsr, TnsA T7 tran sposase, Hjc resolvase, T7 endonuclease I, Mrr, and McrC (Bujnicki, 2001). St ructurally related Type II REases are not necessarily found in the same system subtype , which explains why some members from different subtypes appear to share greater se quence similarity then they do with members of their same subtype (Pingoud et al. , 2002). This suggests that all REase may have evolved from a common ancestor and diverged functionally (Pingoud et al. , 2002). It should be noted that the sequence similarity described is very weak by most standards and is taken in considerati on only in conjunction with ove rall structural similarity. Because no common sequence motifs have been discerned among Type II REases, they cannot be recognized as endonucleases by their amino acid sequence alone; consequently, their presence must be demonstrated by screening cell extracts for endonuclease activity (Wilson, 199 1; Raleigh and Brooks, 1998). In contrast, there are extensive similarities among a ll DNA methyltransferases of R-M systems identified to date (Wilson, 1991; Bujnicki, 2001). Excep tions are the modification enzymes which inhibit restriction by DNA m odifications other than methylation, such as the tetrahydrofolate dependent C5-cytosin e hydroxymethyltransferase of T-even bacteriophage (see Type IV McrBC restriction) (Bujnicki, 2001). They are apparently unrelated to the DNA MTases (Bujnicki, 2001). Of the cloned Type II systems, five of th em known to have the â€œcontrolâ€ C protein gene were compared, and four (BamHI, Pvu II, EcoRV, and SmaI) were found to have amino acid similarity (even higher than th e sequence similarity found between their corresponding DNA MTases) (Bickle and Krug er, 1993). Consequently, it has been suggested that evolution of the C gene was independent of their corresponding DNA
77 MTase and REase (Bickle and Kruger, 1993). As stated previously, the promoters for the MTase and the C protein are typically in opposite orientation with the exception being BglII where the C protein, REase, and DNA MTas e genes are all in the same orientation (Wilson, 1991; Anton et al. , 1997; Vijesurier et al. , 2000). Interestingly, another open reading frame between the endonuclease and DNA methyltransferase genes of the BglII system of Bacillus globigii was found with significant se quence similarity to the Fis (factor for inversion stimulation) protein from E. coli (Anton et al. , 1997). Fis is a DNA binding protein that stimulates si te-specific recombination (Anton et al. , 1997). Considering the additional fact that the BglII MTase has considerable sequence homology to a DNA methyltransferase downstream of the fis gene in E. coli , it has been suggested that the collinear c onfiguration of all the genes of the BglII R-M cassette may have evolved during horizontal transfer of this system into an ancestor Bacillus globigii cell (Anton et al. , 1997). Horizontal transf er may have activated the fis -like gene, causing Fis stimulated recombination, resulting in the fis -like gene and the adjoining DNA MTase gene to reverse their relative orientation (Anton et al. , 1997). Type III Systems There is little information on the struct ural relationships among Type III R-M systems. The endonuclease genes of the we ll characterized EcoPI and Eco15I systems are largely conserved and their DNA met hyltransferase genes are a â€œmosaicâ€ of conserved and nonconserved regions includ ing conserved ATP-binding and AdoMetbinding sites, as well a motif in comm on with helicases and involved in DNA methyltransferase catalysis (B ickle and Kruger; 1993; Rao et al. , 2000). Like Type I systems, Type III systems have a helicas e-like DNA translocation domain (DEAD-box) (Dryden et al. , 2001) and both types are be lieved to belong in a group of evolutionarily
78 related enzymes that translocate DNA (includi ng the helicase superfamilies I and II and various enzymes involved in DNA recombinati on and repair) (Bujnick i, 2001). Evidence shows that the Type III system EcoP15I is sequestered by the inta ct recognition site on cleaved DNA, decreasing its effective c oncentration (Raghavendra and Rao, 2003). Additional rounds of cleavage can only occu r if EcoP15I is re leased by exonuclease degradation of the intact recognition site s (Raghavendra and Rao, 2003). This suggests that infecting foreign DNA, such as phage, may avoid restriction by â€œoverwhelmingâ€ the bacteria, where cleaved phage sequesters the REases allowing the escaping phage to be methylated (Raghavendra and Rao, 2003). If this is generally true for Type III and Type I systems, which maintain their recognition si tes intact, then they are less efficient at restricting infecting forei gn DNA than the more prevalent Type II R-M systems that cleave within their recognition site, suggesting that this would put e volutionary pressure on R-M systems to evolve the ability to cleave within their recognition sites (Raghavendra and Rao, 2003). DNA MTases Chemically, all DNA MTases are divided into two main groups, C-MTases which form C-C bonds when methyl ating the fifth position carbon on cytosines, and N-MTases that form C-N bonds when methylating the sixth position nitrogen on adenines or the fourth position nitrogen on cytosines (Jelts ch, 2002). All DNA MTases generally have two domains: a small variable sequence domain that contains sequence required for target recognition, and a large structural dom ain that contains both the AdoMet binding site and the catalytic center (Jeltsch, 2002; Gromova and Khoroshaev, 2003). N-MTases are subdivided into three structural subclasse s, depending on the relative position of the small domain within the large domain: , , and (Bujnicki, 2001; Gromova and
79 Khoroshaev, 2003). C-MTases are most similar to the subclass of N-MTases (Bujnicki, 2001; Jeltsch, 2002). In N-MTases, the large do main also contains sequence required for target recognition (Jeltsch, 2002). C-MTases (prokaryotic and eukaryotic) have ten conserved amino acid motifs where the mo st conserved are motif I (involved in magnesium ion cofactor binding along with se veral other motifs) and motif VI (involved in structural rearrangement of the DNA during MTase catalysis) (Gromova and Khoroshaev, 2003). N-MTases of prokaryotes have nine conserved amino acid motifs which generally correspond to the motifs found in C-MTases (except the unique motif IX of C-MTases), but the structural classes ( , , ) differ from each other in the order the motifs are typically found in each class (Gro mova and Khoroshaev, 2003). Motif I and IV of N-MTases are the most conserved (Jel tsch, 2002). In general, the specificity of MTases is not as stringent as REases a nd they have a smaller number of specific interactions with the DNA (Je ltsch, 2002). Analysis of th is overall conservation of structure among DNA MTases, referred to as th e â€œMTase foldâ€, and its resemblance to Rossmann-fold proteins, has lead to the hypot hesis that DNA MTases originated from a common Rossmann-fold ancestor and dive rged extensively (Bujnicki, 2001). Homing Endonucleases Because homing endonucleases are classified into four distinct families, based on the conserved amino acid sequence motifs of their catalytic domains (Belfort and Roberts, 1997), they may have evolved on mu ltiple independent occasions (Chevalier and Stoddard, 2001). In addition, sequence anal ysis suggests that the homing REases evolutionary origins are indepe ndent from the origin of the intervening (intron or intein) sequences (Chevalier and Stoddard, 2001). B ecause of their ability to make double
80 stranded breaks and promote recombinati on resulting in duplica tion of the homing endonuclease gene in the intron or intein, homing endonucleases are considered highly invasive â€œparasiticâ€ elements (Belfort a nd Roberts, 1997; Chevalier and Stoddard, 2001; Gogarten et al. , 2002). Since homing endonucleases appear to persist in diverse organisms and promote DNA rearrangements, they are likely influenci ng the evolution of the organism (Belfort and Roberts, 1997). Their widespread distribution in diverse organisms and in diverse or ganellar genomes suggests freq uent horizontal transfer (Chevalier and Stoddard, 2001; Gogarten et al. , 2002). There is little apparent selective pressure to maintain most homing REases and evidence suggest that they are easily lost by degeneration (loss of functi on) or deletion (Chevalier and Stoddard, 2001; Gogarten et al. , 2002). R-M Systems: Mobility Aiding Evolution A growing number of open reading fram es resembling DNA mobility proteins such as the IncJ transposon like element of E. coli K12, the pseudo-trans posable small repeat elements (SREs) in Neisseria meningitidis , and the fis -like gene in Bacillus globigii , as well as other transposases, resolvases, invertas es and integrases, have been found near the genes of many restri ction systems (Anton et al., 1997; Bujnicki, 2001). McrA is located within an excisable prophage-like element e 14 (Bickle and Kruger, 1993; Joset and Guespin-Michel, 1993; Jost and Saluz, 1993; Rao et al. , 2000). The e 14 element can be excised from the chromosome as a nonreplica tive circular DNA molecu le after UV light treatment and eventually eliminated from the population during subsequent cell division (Bickle and Kruger, 1993). DNA mobility proteins, like the ones men tioned above, have been proposed to aid in the recombination and horizontal transfer of R-M genes, contribu ting to the evolution
81 of these systems (Anton et al. , 1997). Genomic comparisons found that the location of R-M systems are often associated with genomic polymorphisms and that apparently unrelated R-M systems in some related bacter ia are located in corresponding positions in their chromosomes (Bujnicki, 2001). Maintenance of R-M Systems Three main hypotheses on the function and th e basis of R-M system maintenance in an organism have been proposed, as descri bed below, which include: (1) the hypothesis that R-M systems are selfish genetic units, (2) the hypothesis that R-M systems function a promoter of recombination, and (3) the hypothesis that R-M systems function as a cellular defense against invading foreign DNA, thereby maintaining species identity. R-M Systems: â€œSelfish genetic unitsâ€ As mentioned previously, it is generally believed that R-M systems function to reduce the frequency of genetic exchange via transformation, transduction, or conjugation. Based on this, it was originally believed that they evolved due to the perceived positive selective advantage of having an â€œimmune systemâ€ against bacteriophage infection and ot her invading foreign DNA in orde r to increase survivability of the host and ensure continuation of its species existen ce though production of uncorrupted progeny. Recent studies suggest th at there may be an entirely different force, other than cellular defense, that pr essured cells to evolve and maintain R-M systems. More recent evidence supports another persp ective that describes the evolution of some R-M systems as resulting from negativ e selection. This hypot hesis, proposed by Naito et al. (1995), states that some Type II R-M systems are â€œselfish genetic unitsâ€. Several studies showed that th e loss of the R-M genes (of vari ous Type II systems) lead
82 to killing of progeny by the re sidual restriction endonucleases , similar in mechanism to a plasmid maintenance function called postseg regational cell killing (reminiscent of eukaryotic programmed cell death and viral anti-hos t-death) (Kulakauskas et al. , 1995; Naito et al. , 1995; Oâ€™ Neill et al. , 1997; Nakayama and Kobayashi, 1998; Chinen et al. , 2000; Handa, 2000). In bacteria, postsegregati onal cell killing is mediated though pairs of genes that produce a stable toxin and a labile antitoxin (Engelberg-Kulka and Glaser, 1999). They have mainly been found in E. coli on low copy number plasmids where loss of the plasmid results in residual toxin mediated deat h of the plasmid free cell (Engelberg-Kulka and Glaser, 1999). There ar e two classes of these selfish pairs of genes: (1) systems where both the stable toxin and unstable antitoxin are proteins where the antitoxin tightly binds to the toxin, bloc king its activity; and (2) systems where the stable toxin is a protein and the unstable antito xin is a small antisense RNA molecule that binds to the stable toxin-encoded mRNA, blocking translation (Engelberg-Kulka and Glaser, 1999). In one study, the EcoRII RM system was cloned into a temperature sensitive plasmid where loss of plasmid repli cation resulted in reduction of cell viability, filamentous growth, and chromosome degrada tion, consistent with postsegregational cell killing (Chinen et al. , 2000). The postsegregational killing caused by loss of the plasmid borne EcoRII R-M system could be suppresse d by co-expression of a second R-M system with a less specific MTase (SsoII), able to protect all EcoRII r ecognition sites (Chinen et al. , 2000). These observations lead to the proposal that there is selective pressure to decrease specificity of the recognition sequence in the abse nce of invading foreign DNA (Chinen et al. , 2000).
83 Studies of some Type II R-M systems have shown them to be st able selfish genetic units, not easily lost from the cell (Naito et al. , 1995; Nakayama and Kobayashi, 1998; Sadykov et al. , 2003). When the genes of one of th ese Type II R-M systems are located on a plasmid, the plasmid cannot be readily displaced by a second plasmid carrying R-M genes from their incompatibility group, even if the second plasmid is not vulnerable to restriction by the host (Naito et al. , 1995; Nakayama and Kobayashi, 1998). An incompatibility group consists of R-M systems with C proteins of the same specificity (Nakayama and Kobayashi, 1998). Those cells that did pick up the second plasmid also retained the original hos t plasmid and exhibited inhibited growth (Naito et al. , 1995; Nakayama and Kobayashi, 1998). In cells that had lost their plasmid encoded R-M system, the SOS response system was induced (Handa et al. , 2000). When the plasmid encoded R-M system was lost in mutants de fective in RecBCD, RecA, RecN, RecG, and RuvABC, more severe death sy mptoms were observed (Handa et al. , 2000). Handa et al. (2000) hypothesized that it is the recBCD/Chi/RecA machinery that functions to destroy foreign DNA while repairing self-DNA after rest riction. Once a cell ca rries a restrictionmodification system, it is e ssential that the DNA MTase is sufficiently expressed to protect the host genome. A cell with two R-M systems of the same recognition sequence specificity does not have the ability to mainta in both systems because one is easily lost without resulting in suicide (Nakayama and Kobayashi, 1998). This may help explain the sequence diversity and apparently independent evolution of Type II REases, particularly for those found in different strains within the same species (Nakayama and Kobayashi, 1998). This may also help explain the evol ution of extremely specific and exclusive recognition sequences, as well as the evolution of rare cutt ers in Type II R-M systems,
84 which apparently has some selective advantage contrary to what would be expected if the only consequence of REase activity is restriction of foreign DNA (Naito et al. , 1995; Nakayama and Kobayashi, 1998). More recent studies of several Type II R-M systems, located on plasmids or in the chromosome, have been shown to be st able and resist loss (Sadykov et al. , 2003), supporting the selfish genetic unit hypothesis of R-M system maintenance. Despite there being almost no sequence similarity among Type II R-M systems, comparative genomic and protein structure anal ysis suggest that members of structurally related REase families most likely evolved from common ancestors and diverged (Kovall and Matthews, 1998; Bujnicki, 2001). The naturally competent bacteria Bacillus subtilis , naturally able to re lease DNA, amplified their single copy of bamHI in a â€œburst-like fashionâ€ (Sadykov et al. , 2003). This observation, along w ith studies of the phylogenetic relationships of organisms with related REases suggesting ho rizontal transfer, the linking of some mobility elements with some R-M genes, and the selfish genetic unit hypothesis, supports a suggestion made in 2001 by Kobayash i that R-M systems could be classified as a type of selfish mobile element that sp reads, like viruses and transposons (Kobayashi, 2001; Sadykov et al. , 2003). In the case of modification dependent restriction systems, while they may not act as selfish genetic units in the way described for some Type II systems, they do commit â€œsuicideâ€ if (for example) invading forei gn DNA expresses a MTase that methylates the host chromosome at sites recognized by its me thyl-dependent restriction enzyme, thus limiting transfer of MDR sy stems among these organisms (Nakayama and Kobayashi, 1998).
85 R-M Systems: A promoter of recombination Evidence for the selfish behavior hypot hesis has only been shown in Type II systems that have R-M system regulatory C proteins that regu late transcription, explaining the ability of the bacterium to acquire new Type II systems (with unique C proteins) but be sensitive to a Type II system loss (Nakayama and Kobayashi, 1998; Murray, 2002). Oâ€™Neill et al. (1997) showed that the genes of the Type I Hsd R-M system of E. coli K12 are not â€œselfishâ€ (they are not se nsitive to the loss of the Type I RM system) and that the diversity and distribut ion of Type I R-M systems are probably due to selective advantage, as conventionally hypothesized, and/or due to other selective pressures yet to be understood (Oâ€™Neill et al. , 1997). Little has b een suggested about the function and evolution of the relatively uncommon Type III R-M systems. There are many large structural and mechanistic differe nces between the different types of R-M systems and perhaps the evolution of R-M sy stems can not be all encompassing. Because of the different properties of each system T ype, they may have different evolutionary pressures. The Type I system is also comp aratively less efficient at restricting foreign DNA than Type II and Type III R-M systems (Joset and Guespin-Michel, 1993, p. 191). Type I R-M systems protective ability is furt her weakened by the fact that its cleavage site (randomly chosen) is di stant from its recognition s ite (Dreier and Bickle, 1996; Dryden et al. , 2001; Bourniquel and Bickle, 2002). DNA fragments are created that may be potential substrates for recombinati on (Joset and Guespin-Michel, 1994, p.190-191). Consequently, it has been suggested that Type I R-M systems may be important in general recombination, contribu ting a possible selective advant age to the organism (Joset and Guespin-Michel, 1994, p.190-191).
86 While only some R-M systems appear to act as selfish genetic units, evidence suggests that the cellular function of all RM systems may include more than just the restriction of foreign DNA resulting in some degree of genetic isolation. Complete genetic isolation is generally considered de trimental for survival in the long run. As mentioned previously, the function as a stim ulator of recombination is proposed to contribute to the selective adva ntage of maintenance and evol ution of the diverse Type I systems, which may explain the unus ual allelic diversity as found in hsd genes of E. coli (Oâ€™Neill et al. , 1997). All types of R-M systems may in fact facilitate genetic exchange by breaking up foreign DNA into double strande d fragments containing one or more genes that may be used in general recomb ination. Genetic isolation may depend on the combination of factors such as (1) the promiscuity of the ge netic exchange vectors (broad host range, conjugal plasmids, genetic homol ogy/divergence), (2) promiscuity of the chromosomal DNA (promoter recogniti on, conjugal transfer, sequence homology/divergence, mobile genetic units coding for mobility proteins such as transposons, that can be transferred vi a conjugation, or exci sed and invade via transformation), (3) functiona lly incompatible gene produc ts (codon bias resulting in nonfunctional products, products that are detrim ental to cellular function), (4) physical isolation (microhabitat), (5) barriers to tran sformation or transducti on, (6) efficiency of recombination, (7) efficiency of mismatch re pair, as well as (8) restriction-modification systems (Matic et al. , 1996). As mentioned previousl y, the transfer of single stranded homologous DNA via conjugation is generally po orly restricted in cells by most R-M systems, usually due to hemi-methylated he teroduplex formation with the host DNA and homologous recombination (Raleigh and Br ooks, 1998). The pers istence of ssDNA
87 (such as when the DNA is nonhomologous) induces the SOS repair system, resulting in the overproduction of RecA (which me diates recombinant strand exchange), overproduction of RuvAB (which mediates re combinant branch migration of Holliday structures), restriction alle viation, inhibition of RecBCD exonuclease activity, increased transposon mobility, and increased permeability of the outer membrane (increasing ability to transform) (Matic et al. , 1996; Handa et al. , 2000). These observations also lead to the hypothesis that th e induction of the SOS response system not only functions to repair DNA damage in the host, such as damage due to restriction, but also increases the survivability of stressed b acterial populations by prom oting genetic exchange and increasing diversity (Matic et al. , 1996). One interesting and well studied pair of complementary but opposing restrictions systems provides an excellent example of how some R-M systems may function to promote genetic exchange between different strains of the same organism while maintaining its ability to protect the orga nism from potentially detrimental invading DNA, thereby contributing to its survival. Th e Type IIM MDR restriction system DpnI is found in certain strains of Streptococcus pneumoniae , while other strains of S. pneumoniae have DpnII (De la Campa et al. , 1988; Bickle and Kruge r, 1993). DpnII is a complementary Type IIP R-M system consisting of a REase and two MTases, one of which (M1.DpnII ) can methylate ssDNA or dsDNA (De la Campa et al. , 1987; De la Campa et al. , 1988; Cerritelli et al. , 1989; Bickle and Kruger, 1993; Jost and Saluz, 1993). Double stranded DNA, like bact eriophage, originating from the DpnII strains is cleaved when introduced into DpnI strains and visa versa (Cerritelli et al. , 1989; Bickle and Kruger, 1993; Jost and Saluz, 1993) . Conversely, when ssDNA is introduced ( S.
88 pneumoniae is naturally transformable via conj ugation), it is quickly methylated by M1.DpnII, avoiding rest riction (Cerritelli et al. , 1989; Bickle and Kr uger, 1993; Jost and Saluz, 1993). To reiterate, these types of complementary but opposing systems restricts potentially detrimental foreign dsDNA such as naked dsDNA, or dsDNA carried by bacteriophage or conjugative plasmids (pre venting it from being transferred between different strains of an organism with a c onversely restrictive syst em, as well as from other organisms), but allows transfer of a dvantageous mutations and promoting evolution via transfer of ssDNA by conjugation follo wed by recombination, particularly between closely related organisms. R-M Systems: Cellular defense a nd maintenance of species identity The suggestion that R-M systems function to stimulate recombination is believe by some to be a side effect, rather that a pr imary function and force driving evolution and RM system maintenance (Jeltsch, 2003). In fact , other studies show a different relationship between R-M systems and recombination, wher e recombination is an inevitable side effect to maintain cell viability rather than the cell maintaining R-M systems in order to promote recombination. One study showed that a recA E. coli mutant containing a mutant form of the Type I system EcoR124I , unable to respond to mutagenic treatment by alleviating restriction lik e the wild-type, produced 41% anucleated cells (Makovets et al., 2004). Compared to the 10% anuc leation normally expected in recA mutants, this large percentage of anucleation due to DNA de gradation suggests that recombination is essential to maintain chromosome integrity (Makovets et al. , 2004). The recently proposed selfish genetic unit hypothesis is based on evidence that cells of some Type II R-M systems die with loss of the genes, and this hypothesis does not appear to apply to other types of R-M systems. A more recent paper in essence
89 supports the broad cellular defense hypothesis bu t refines it by basically stating that R-M systems function to maintain species ident ity as a competitive mechanism to survival, exerting pressure to evolve unique specificiti es and be maintained in the cell (Jeltsch, 2003). Unique R-M systems pose a genetic barr ier to gene transfer between competing cells, explaining why there is great variety in the R-M systems f ound in a single species (differing between strains) (J eltsch, 2003). One example s upporting this view is that Helicobacter pylori typically encodes more than 20 putative R-M systems (Jeltsch, 2003). Strain specific R-M systems all appear to be active, while most of the shared R-M genes have been found to be inactive (Jeltsch, 2003).
90 CHAPTER 3 STUDY RATIONALE AND EXPE RIMENTAL APPROACH The discovery of the first sequence speci fic R-M system (Smith and Wilcox, 1970) and the first application of RE ases as a molecular tool to create recombinant DNA in the early seventies (Cohen et al. , 1972; Mertz and Davis, 1972; Cohen et al. , 1973) laid the foundation of modern biotechnology. Simultaneous ly, the first energy crisis resulted in avid research into bioengineering recombin ant biocatalysts for industrial production of fuel ethanol from biomass. Z. mobilis has been of particular interest as a potential biocatalyst in the production of fuels and ot her organic chemicals, such as levan and sorbitol (Doelle et al. , 1993). As described in Chap ter 1, the main disadvantage Zymomonas has as a biocatalyst for industrial etha nol production is the narrow range of substrates it can metabolize: gluc ose, fructose, and sucrose (Sahm et al. , 1992; Doelle et al ., 1993; Gunasekaran and Raj, 199 9; Aristidou and Penttila, 2000). Z. mobilis has been successfully engineered to use xylose and arabinose as sole carbon sources (Dien et al. , 2003). Regardless, the efficiency of stable transfer of foreign genes into Z. mobilis is reported to be relatively low (Sahm et al. , 1992), typically 1001000 CFU per g of foreign DNA (Ingram et al. , 1989), making it relatively difficult to bioengineer. Considering RM systems function to reduce the frequency of genetic exchange, it was hypothesized that Zymomonas contains restricti on endonuclease and/or DNA methyltransferase systems. A better understanding of any restriction and/or modification systems of Zymomonas may lead to increasing the transformation efficiency
91 of foreign genes into Z. mobilis, and thus facilitate multiple gene transfer and gene replacement. After reviewing the characteristics that may be attributed to the organism Zymomonas as well as characterist ics of REases and DNA MTases commonly found in prokaryotes, we designed expe rimental approaches to identify REases and DNA MTases that may be found in Z. mobilis CP4. This lead to determining if Z. mobilis CP4 contained the same R-M systems identified in other strains. Early experiments indicated that Z. mobilis subspecies anaerobia has the Type II restrictio n-modification system ZanI (Sun and Yoo, 1988). In this study, experiments showed that Z. mobilis CP4 does not have ZanI. The next step was to identify REase and/or DNA MTase systems that may be in Z. mobilis CP4. In this study, the initial approach was to construct and then screen a genomic library using a method developed by Piekarowicz et al. (1991b) and by Fomenkov et al. (1994) for cloning and rapid identif ication of genes that encode DNA damaging enzymes, such as restriction e ndonucleases and DNA modification enzymes. Subsequent isolation and sequence analysis of positive clones revealed that Z. mobilis CP4 contains an mrr -like endonuclease and a solitary DNA methyltransferase. The following chapters describe the identificati on and characterization of these two cloned enzymes: ZmCP4Mrr and ZmCP4CcrM.
92 CHAPTER 4 CLONING AND CHARACTERIZATION OF A METHYL-DEPENDENT RESTRICTION ENDONUCLEASE FROM ZYMOMONAS MOBILIS SUBSPECIES MOBILIS CP4 Introduction Zymomonas mobilis has been the subject of ge netic and biochemical studies directed at assessing its potenti al as a biocatalyst for the production of fuels and other organic chemicals (Swings and De Ley, 1977; Crueger and Crueger, 1989; Doelle et al ., 1993; Danner and Braun, 1999). The main disadvantage Zymomonas has as a biocatalyst for ethanol production is the na rrow range of substrates it can metabolize: glucose, fructose, and sucrose (Sahm et al. , 1992; Doelle et al ., 1993; Gunasekaran and Raj, 1999; Aristidou and Penttila, 2000) (see Chapter 1). Transformation and conjugation are the methods used to introduce foreign DNA into Zymomonas , however, the efficiency of stable transfer of foreign genes into Z. mobilis is relatively low, typically 100-1000 CFU per g of foreign DNA (Ingram et al. , 1989; Sahm et al. , 1992). RestrictionModification (R-M) systems are widely found in prokaryotes and f unction to reduce the frequency of genetic exchange via transf ormation, transduction, or conjugation. The relatively low transformation efficiency of foreign DNA into Z. mobilis CP4 indicates that this strain contains restriction endonucleases a nd/or DNA methyltransferases. The Type II restriction-modification system ZanI had been previously identified in the Z. mobilis subspecies anaerobia (Sun and Yoo, 1988). In this study, we examined Z. mobilis DNA for methylation of ZanI rec ognition sites to determine if Z. mobilis CP4
93 contains a ZanI-like R-M system. Resu lts of this analysis indicate that Z. mobilis CP4 does not have the Type II ZanI R-M system found in Z. mobilis anaerobia . Based on the absence of a ZanI-like R-M system, we investigated the possible presence of other restriction endonuclease (REase) systems in Z. mobilis CP4. A Z. mobilis CP4 genomic library was constructe d and screened for DNA modification enzymes, and a putative mrr -like endonuclease was identified. Mrr is a Type IV methyldependent restriction system that has been identified and characterized in Escherichia coli K12 (Waite-Rees et al. , 1991; Roberts and Macelis, 1996). Putative EcoKMrr homologues have been identified by amino acid sequence similarity in numerous bacteria (Roberts and Macelis, 1996). The EcoKmrr gene codes for a single methyl-dependant restriction endonuclease with two phenotypes referred to as Mrr (methylated adenine restricting) and McrF (methyl ated cytosine restricting) (Bickle and Kruger, 1993; Jost and Saluz, 1993). EcoKMrr restricts adenin e and cytosine modified dsDNA, methylated by a variety of MTases such as M.SssI, M. HinfI, M.HhaII and M.PstI, but no consensus recognition sequence has b een deduced (Waite-Rees et al. , 1991; Roberts and Macelis, 1996). In this study, transformation efficien cy experiments indicate that the putative mrr like endonuclease, designated ZmCP4Mrr, is a methyl-dependent restriction endonuclease. ZmCP4Mrr, cloned into the cloning vector pP3226 and the expression vector pPROTet.E133, was transformed into E. coli K12 strains with various DNA MTase deficiency backgrounds. Recombinant pP3226/ ZmCP4mrr plasmids could not be transformed into dam + E. coli strains. Recombinant pPROTet.E133/ ZmCP4mr r plasmids could only be transformed into a dam+ E. coli strain under conditions where ZmCP4mrr expression was repressed; however, growth of this strain was arrested upon
94 inducing expression of ZmCP4mrr . These results suggest that ZmCP4Mrr restricts M.EcoKDam N6-adenine methylated DNA (5â€™-GmATC-3â€™) and that ZmCP4Mrr is a methyl-dependant restriction endonucleas e which recognizes methylated sites overlapping Dam sites. Materials and Methods Bacterial Strains, Plasmids, Media, and Reagents Bacterial strains, plasmids, and primers used in this study are described in Table 1. Zymomonas mobilis subspecies mobilis var. recifensis CP4/D strain, from L. O Ingramâ€™s laboratory, is a derivative of a Zymomonas strain originally isolat ed from a sugar cane in Recife, Brazil (via A. Ben-Basset) (Yablonsky et al. , 1988). The Mrr#3 custom designed primers were purchased from SIGMA-GENOSYS (Woodlands, TX.). E. coli was grown in Luria-Bertani (LB) brot h media (10 g/L tryptone, 5 g/L yeast extract, 5 g/L NaCl, 1ml 1N NaOH) (Atlas, 1997) at 37C with vigorous shaking (200300 RPM) for aeration. Z. mobilis was grown in a GY broth media (100 g/L glucose, 20 g/L yeast extract) (Atla s, 1997) with 1 g/L KH2PO4 (GYx) at 30C with slow shaking (80-100 RPM) and minimized aeration unless stated otherwise. Solid media was prepared by adding 15 g/L of agar to the broth media. For selection, Z. mobilis cultures contained the appropriate antibiotics at the following concentrations: 80 mg/L chloramphenicol and 24 mg/L tetracycline. E. coli cultures contained the appropriate antibiotic at the following concentrations: 50 mg ampicillin, 40 mg chloramphenicol, 12 mg tetracycline, 100 mg spectinomycin, and 30 mg kanamycin per liter. Deionized water (ddH2O ) used for reagents was purified using NANOpure II ultrapure water system at 16.7 megohm-cm resistivity (Barnstead Internat ional, Dubuque, IA.). All chemicals used were reagent grade.
95 DNA Isolation Plasmids were isolated using QIAGEN pl asmid extraction kits (QIAGEN, Inc., Valencia, CA. 91355) according to the manufact urerâ€™s direction or by the standard TELT method or cesium chloride gr adient centrifugation (Ausubel et al. , 1998). Standard protocols were used for DNA etha nol precipitation with either 0.25 M sodium chloride or 2.5 M ammonium acetate (Berger and Kimm el, 1987). Restriction endonucleases, alkaline calf intestinal p hosphatase, T4 kinase, T4 DNA lig ase, and Quick Ligation kit used in this study were purchase d from New England Biolabs, Inc. (Beverly, MA.) and used according to the supplierâ€™s protocols. T4 DNA polymerase (New England Biolabs, Inc.) was used to blunt-end DNA fragments. PfuTurbo DNA polymerase (Stratagene, La Jolla, CA.) was used for PCR amplifications according to the supp lierâ€™s protocols. DNA was purified between enzymatic reactions usi ng the following methods as appropriate: (1) standard phenol-chloroform-octanol (50: 48:2) extraction (Berger and Kimmel, 1987), (2) separation by 0.8% agarose (Sigma-A ldrich Co., Saint Louis, MO.) gel electrophoresis and elution of DNA fragmen ts using QIAquick gel extraction kit (QIAGEN, Inc.), and (3) PCR amplification of DNA fragments and purification using the QIAquick PCR purification kit (QIAGEN, In c.) according to the supplierâ€™s protocol. DNA molecular weight standards for agaros e and polyacrylamide gel electrophoresis were /HindII Digest and 100 bp DNA Ladder (New England Biolabs, Inc.). Construction of and Appropriate Cloning Vector The vector used to clone the re striction-modification genes of Z. mobilis CP4 was constructed from pBR322 (Figure 2). An 80 bp fragment with multiple cloning sites from pNEB193 was inserted into the EcoRI-H indIII site of pBR322. Both pBR322 and pNEB193 were digested with EcoRI and Hi ndIII and the DNA fragments were isolated
96 following agarose gel electrophoresis. The appr opriate bands were ge l purified and the DNA was extracted using the QIAquick gel ex traction kit. The 72 bp EcoRI to HindIII polylinker fragment of pNEB193 was ligated into the EcoRI to HindIII site of pBR322. The pBR322/pNEBpolylinker construct, was dige sted with StyI and EcoRV to remove an 1184 bp sequence containing the BamHI site fr om the plasmid, reducing the size of the vector. The resulting vector DNA was bl unt-ended with T4 DNA polymerase and the ends ligated with T4 DNA ligase, produc ing the 3226 bp plasmid vector pP3226 that contains the selectable marker for ampicillin resistance. Isolation of Zymomonas mobilis CP4 Genomic DNA Zymomonas mobilis CP4 genomic DNA was isolated from 100 ml liquid culture grown to OD550 of 0.6-0.9 at 30C with gentle shaking. Chilled cells were centrifuged for 15 min at 5000 x g , wash twice with 10 ml ice-co ld 1x basal salt solution (1 g/L (NH4)2SO4, 1 g/L K2HPO4, 1 g/L KH2PO4, and 0.5 g/L NaCl), and the washed pellet was resuspended in 0.5 ml ice-cold 1X basal sa lt solution. Fresh prot ease-lysis buffer was prepared by mixing 10 ml of freshly prepar ed lysis buffer (10 mM Tris-HCl pH 8.0, 0.1 mM EDTA, 1% wt/vol N-laurolysarcosine), 500 l 10% wt/vol SDS, 200 l, 0.2 M EDTA, and 40 mg of nucleasefree nonspecific protease Pronase E (Sigma-Aldrich, Co.), which is added after SDS is completely di ssolved. The protease-lysis buffer, preincubated for 30 min at 37C, was added to the washed cell slurry, mixed gently, and incubated overnight at 37C in a polypropylen e screw-cap centrifuge tube. The genomic DNA was extracted by the CTAB (hexadecyltrime thyl ammonium bromide) protocol for preparation of bacterial genomic DNA as described by Ausubel at al (1998). The supernatant, from the chloroform-octanol ex traction, was extracted tw o to three time with
97 buffered phenol-chloroform-octa nol pH 6.7-8.0 (25:24:1) unt il a clear aqueous solution was obtained. Two volumes of 95% ethanol wa s carefully layered on top of the extracted aqueous phase containing the Z. mobilis CP4 genomic DNA, and the genomic DNA was spooled with a pasture pipette that had been fl amed in order to seal the end and bent to form a glass hook. The DNA was gently dissolv ed in 4 ml of 10 mM Tris-HCl pH 8.0, 10 mM NaCl, 10 mM Na2EDTA, 100 g/ml DNase-free RNase (Sigma-Aldrich, Co.) and incubated for 3 hrs at 37C. After in cubation the solution was adjusted to 0.1M NaCl, gently mixed, layered with 2 volumes of 95% ethanol, and the genomic DNA was gently spooled and redissolved in 1 ml TE pH 8.0 (10 mM Tris-HCl, 1 mM EDTA) (Ausubel et al. , 1998) and 2 l of 0.2 M Na2EDTA was added. The DNA was stored at 20 C. Preparation of a Zymomonas mobilis CP4 Genomic Library Z. mobilis CP4 genomic DNA was partially dige sted with Sau3AI and the 4-6 kb fragments were isolated afte r 1% agarose gel electrophores is using the QIAquick gel extraction kit (QIAGEN, Inc.). The desire d size range was iden tified by co-resolved molecular weight marker lanes ( /HindIII Digest) that were cut off and stained with EtBr and visualized with UV light. The isolated genomic DNA fr agments were not exposed to EtBr or UV light. To obtain a source of stable clones for repeated screening, an amplified genomic library was prepared in E coli ER1647. The cloning vector pP3226 was digested with BamHI and dephosphorylated with alkaline calf intestinal phosphatase. The Z. mobilis CP4 genomic Sau3AI 4-6 kb gel purified DNA fr agments were ligated into the BamHI site of pP3226, phenol-chloroform-octanol (50:48:2) extracted, transformed via
98 electroporation into electrocompetent ER1647, a nd plated onto LB ampicillin plates. The cells (~6900 CFU total) were washed off the plates, and the plasmids were extracted and dissolved in TE buffer at pH 8.0 and stored at -70 C. Transformation E. coli and Z. mobilis CP4 electrocompetent cells we re prepared from exponentially growing cultures (OD600=0.5-0.6 and OD550= 0.8-1, respectively). Chilled cells were harvested by centrifugation at 4C for 15 min at 5000 x g , washed three times with culture volume of ice cold ddH2O, and washed once with ice cold 10% vol/vol glycerol in ddH2O. The cells were resuspended in an equal pellet volume of chilled (4 C) 10% vol/vol glycerol ( E. coli ) or in 10% vol/vol glycerol to a final concentrat ion of 10-10 cells/ml ( Z. mobilis ), and dispensed in pre-chilled mi crofuge tubes in volumes of 50 or 100 l. The cells were quick frozen in a -75C ethanol bath and stored at -75C. Prior to electroporation, electroc ompetent cells were thawed on i ce, mixed with 10 ng of plasmid DNA in 1-2 l TE pH 8.0, and transferred to a prechilled electroporation cuvette with a 0.2 cm electrode gap (Bio-Rad Laboratories, Inc., Hercules, CA.). Electroporation was performed in a Gene Pulser II (Bio-Rad Labor atories, Inc.) set at 2.5 kV at 200 ohms for E. coli and 400-600 ohms for Z. mobilis . The ohms used fo r electroporation into Z. mobilis were adjusted for a target time constant of 10 msec (Liang and Lee, 1998). The cells were immediately resuspende d in 1 ml of recovery media and incubated at 37C for 1 hr ( E. coli in Soc media described by Ausubel et al. , 1998) or 30C for 3 hrs ( Z. mobilis , in GYx with KH2PO4 concentration increased to 2 g/L). Aliquots of 1-200 l, diluted to a final volume of 200 l, were spread on appropriate selective media.
99 Screening a Zymomonas mobilis CP4 library for REases and DNA MTases E. coli AP1-200-9 (Piekarowicz et al. , 1991b) and ER1992 (Fomenkov et al. , 1994) indicator strains were used to screen the Z. mobilis CP4 amplified genomic library. AP1200-9 selection is based on temper ature sensitive mutations in mcrA, mcrB , and mrr MDR systems and expression of a SOS inducible promoter fused to the lac operon ( dinD1::LacZ+ ) induced by DNA damage . ER1992 is a restriction system deficient mutant with selection based on expression of a SOS inducible promoter fused to the lac operon. Electrocompetent AP1-200-9 and ER 1992 cells were prepared and transformed via electroporation with Z. mobilis CP4 library clones and plated onto LB ampicillin Xgal plates. The plated transformed ER1992 cells were incubated overn ight at 37C. The plated transformed AP1-200-9 cells were incubated overnight at the nonpermissive temperature (42C), transferre d to the permissive temperat ure (30C) for 3 hrs, and transferred back to the nonpermissive temperature for 6 hrs. The plates were screened for DNA damage positive blue colonies. The positiv e clones were isolated and digested with a variety of restriction endonuc leases (NotI, PstI, SalI, Pm eI, BanII, AvaI, PacI, SacI, XmaI, KpnI, SphI, SpeI, HindIII, XbaI, SfiI, AscI). The resulting restriction fragment patterns were analyzed by agarose gel elec trophoresis to identify unique clones for DNA sequence analysis. Sequencing and Analysis of Positive Clones More than forty positive clones were analyzed by restriction endonuclease digestion. Of these, 10 isolated from AP1-200-9 and 8 isolated from ER1992 were partially or completely sequenced. The Z. mobilis CP4 inserts were sequenced, either after they were subcloned, by PmeI/KpnI dige stion and ligation into the PmeI to KpnI site of pNEB193, or directly using custom sequencing primers for pP3226 purchased
100 from LI-COR, Inc. (Table 1). Plasmids borne cloned inserts were sequenced from both directions by the standard dideoxy method using custom fluorescent primers (LI-COR) (Table 1), the Sequitherm Excel II DNA seque ncing kit (Epicentre, Madison, WI.), and the LI-COR automated DNA sequencer (model 4200-L global IR2) (Lai et al. , 1997). The sequencing data was then analyzed. Open reading frames (ORF s) were identified by an ORF finder program on th e Internet (NCBI ORF Finder). The sequence data was translated in all six reading frames into amino acid sequence using a Web based 6-frametranslation program (BCM Search Launc her: Sequence Utilities) (Smith et al ., 1996). All predicted amino acid sequences were compared to sequences in the GenBank database (NCBI BLAST) to identify protei ns with significant amino acid sequence similarity. The predicted amino acid sequen ces were analyzed for conserved sequence motifs and the expected molecular weight of the putative proteins were determined using Web based proteomics and sequence analysis tools (ExPASy). Primer Extension Analysis of ZmCP4mrr mRNA The ZmCP4mrr transcription initiation site wa s identified by primer extension analysis. Total RNA was isolated fro m either 2 ml or 30 ml of cultured Z. mobilis CP4 cells in exponential phase (OD550=0.6-0.8) using QIAGEN RNeasy kit (QIAGEN, Inc.) or a scaled-up method for isolation of R NA from Gram-negative bacteria (Ausubel et al. , 1998) respectively. For large scale total RNA preparation, cells were collected by centrifugation at 4 C at 7,500 x g for 10 min. Harvested cells were resuspended in a sucrose protoplasting buffer (15 mM Tris pH 8.0, 450 mM sucrose, 8 mM EDTA pH 8.0) containing 10 g/ml egg white lysozyme (Sig ma-Aldrich, Co.), incubated on ice for 25 min, harvested by centrifugation, and incubated in a SDS lysis buffer (10 mM Tris pH
101 8.0, 10 mM NaCl, 1 mM sodium citr ate, 1.5% wt/vol SDS) at 37 C for 5 minutes, ethanol precipitated, and extracted using phenol-chloroform-octanol (25:24:1) followed by chloroform-octanol (24:1) extraction as described by Ausubel et al. (1998). Total RNA was treated with RQ1 RNase-free DNase (Promega Corp., Madison, WI.) followed by a 10 min heat inactivation at 65 C. Phenol and chloroform extractions and ethanol precipitation was repeated. Total RNA con centration and purity were estimated by measuring UV absorbance (260nm and 280nm) of serial diluted RNA in DEPC (diethyl pyrocarbonate) treated ddH2O. Absorbance was measured in a 10 mm quartz cuvette (Pyrocell) with a Beckman spectrophotom eter model DU 640 (Beckman Coulter, Inc., Fullerton, CA.). The integrity of the total RNA was analyzed by standard formaldehyde agarose (FA) gel elec trophoresis (Sambrook et al ., 1989). Primer extension of 20-40 g total Z. mobilis CP4 RNA was performed using M-MLV reverse transcriptase (Promega Corp.), as recommended by the supplierâ€™s protocols, using denaturing RNA-DNA hybridization conditions. Custom designe d IRD41-labeled primers (MrrEXT2) to internal ZmCP4mrr sequence, purchased from LI-COR Inc., were used (Table 1). The primer extension products were incubated with 1l of 10g/ml DNase-free RNase A (Sigma-Aldrich), incubated for 30 minutes at 37C, ethanol precipita ted, and dissolved in loading buffer (Epicentre). A Dideoxy sequencing ladder of ZmCP4mrr was generated with the same primer used for the primer extension and the Sequitherm Excel II DNA sequencing kit (Epicentre). The primer extension products and parallel dideoxy sequences were analyzed using the LI-COR model 4200-L global IR2 automated DNA sequencer (Lai et al. , 1997).
102 Recombinant Expression of the ZmCP4 mrr The mrr -like sequence from Z. mobilis was PCR amplified using the Mrr#3 primers (containing KpnI sites) (Table 1), purified using the QIA quick PCR purification kit, digested with KpnI, and cloned into th e KpnI site of the expression vector pPROTet.E133. Clones were verified by rest riction analysis and by sequencing using custom sequencing primers to the expression vector purchased from LI-COR, Inc. (Table 1). Genes cloned into pPROTet.E133 were unde r the control of a te tracycline inducible promoter ( PLtetO-1) to produce a C-terminal polyhistid ine tagged recombinant protein, with a predicted molecular ma ss of 38.61 kDa, to facilitate protein purification. The expression construct pPROTet.E133/ ZmCP4mrr was transformed into the expression E. coli strain BL21PRO containing the Tet repressor plasmid. Protein purification of ZmCP4Mrr BL21PRO transformants containing pPROTet.E133/ ZmCP4mrr were cultured in 200 ml LB to an OD600 of 0.5. The cells were induced with anhydrotetracycline (aTc) at 500 ng/ml for 4 hrs and harv ested by centrifugation (3,000 x g ). Cell pellets from 40 ml of culture were flash frozen and stored at -75 C. Extraction-wash buffer (50 mM Sodium Phosphate, 300 mM NaCl, pH 7.0) was prepar ed from stock buffers provided in the TALON Buffer Kit purchased fr om CLONETECH Laboratories, Inc. (Palo Alto, CA.). Frozen cell pellets were thawed on ice with addition of 2 ml of chilled 1 x extractionwash buffer containing 0.75 mg/ml egg white lysozyme (Sigma-Aldrich, Co.). Samples were sonicated using a Soni fier Cell Disruptor W185 (B ranson Ultrasonics Corp., Danbury, CT.) with 3 x 10 sec bur sts (output level set at 6) with 30 sec pauses between each burst and centrifuged for 20 min at 20,000 x g at 4 C to obtain clarified cell-free
103 extracts. Cell-free extracts were applied to TALON metal affinity resin (Sepharose Cl6B) (CLONETECH Laboratories, Inc.) accordin g to the supplierâ€™s Batch/Gravity-Flow Column Purification protocol. The TALON resin column was prepared with a 2.5 ml bed volume. Extraction-wash and elution buffe rs were prepared from the stock TALON buffers with addition of 10% vol/vol glycerol. Protein ex tracts were desalted with prepacked HiTrap Desalting columns (5 ml) (Sephadex G-25) (Amersham-Pharmacia, Corp.) according to the supplierâ€™s protocol. Protein concentration was determined by the standard Bradford method (Ausubel et al. , 1998) and the recombin ant ZmCP4Mrr protein purity was determined by SDS-PAGE and sta ndard coomassie staining (Laemmli, 1970). Transformation Studies in Zymomonas mobilis CP4 To examine the effect of DNA methylat ion on transformation efficiency of Z. mobilis CP4, plasmids isolated from E. coli strains containing diffe rent combinations of native E. coli K12 DNA MTases, or the same plasmid isolated from Z. mobilis, were used. When Z. mobilis CP4 was transformed by elec troporation with the plasmid pLOI1844 extracted from different strains of E. coli, or from Z. mobilis CP4, the efficiency of transformation was approxi mately 100 fold higher when the plasmid originated from Z. mobilis CP4, as compared to the same plasmid originating from different E. coli strains, regardless of the native DNA MTase profile of each E. coli strain (Table 2). E. coli strain Gm272 was described has having no detectable methylation, suggesting that Z. mobilis may be restricting unmethylated DNA (see Discussion). The dependence of transformation efficiency of Z. mobilis CP4 on the source of the plasmid is consistent with the interpretation that it ha s a restriction system that causes the low efficiency of transfer of foreign DNA.
104 BstNI Restriction Analysis of Zymomonas mobilis CP4 DNA The only restriction endonuclease characterized in a Zymomonas species is ZanI, found in Z. anaerobia , whose cognate DNA methyltransfer ase appears to methylate the internal cytosine w ithin the 5â€™-CCWGG-3â€™ recognition site (Sun and Yoo, 1988). This is the same recognition sequence as BstNI RM system that catalyzes N4-cytosine methylation (5â€™-CmCWGG-3â€™) and M.EcoKDcm that catalyzes C5-cytosine methylation (5â€˜-CmCWGG-3â€™). In order to determine if Z. mobilis CP4 has a ZanI-like Type II R-M system, Z. mobilis CP4 and E. coli ER1647 were transformed with the shuttle vector pLOI1844. Plasmid DNA extracted from Z. mobilis CP4 and E. coli ER1647 was used as the substrate for restriction digestion with BstNI (5â€™-CCWGG-3â€™), HinfI (5â€™-GANTC-3â€™), and TfiI (5â€™-GAWTC-3â€™) (Figur e 3). The expected digest ion pattern of pLOI1844 with BstNI, TfiI, and HinfI is shown in lanes 57, Figure 3. Each enzyme produced two major bands, and a series of small fragments that are not resolved at the bot tom of the gel. The banding pattern in the lanes cont aining plasmid extracted from Z. mobilis CP4 is more complex. Z. mobilis CP4 has four native plasmids, approximately 31.5 bp, 32.5 bp, 33 bp, and 35 bp in length (Yablonsky et al., 1988), which are co-extr acted with the test plasmid pLOI1844. Comparison of the restrict ion fragments produced by digestion of plasmid DNA from Z. mobilis CP4 and E. coli demonstrate that BstNI digests Z. mobilis CP4 extracted DNA, as indicated by the presence of restric tion fragments of equal size (Figure 3, lanes 2 and 5). In contrast, the DNA extracted from Z. mobilis CP4 is resistant to both TfiI and HinfI digestion. Because Z. mobilis CP4 DNA does not resist BstNI digestion, unlike Z. anaerobia DNA (Sun and Yoo, 1988), it appears that Z. mobilis CP4 does not have a ZanI-like Type II R-M system. Z. mobilis DNA resists TfiI and HinfI
105 digestion due to DNA methylation by Zm CP4CcrM (see Chapter 5 on ZmCP4CcrM characterization). Isolation and Sequence Analysis of Zymomonas mobilis CP4 Clones The host indicator systems for detecti on of clones containing DNA MTases or REases contain a SOS inducible dinD1 promoter fused to a lac operon . DNA damage is detected by induction of the SOS promoter and expression of the lac gene. For detecting REases, a relatively low copy number vector is required to allow a detectable level of gene expression without killing the host ce ll. A relatively lo w copy number vector, pP3226, was constructed and used in the construction of a Z. mobilis genomic library (Figure 2). Sau3AI restrict ion fragments between 4-6 kb long, predicted to contain between two and four genes, were ligated into pP3226. The library obtained contained approximately 6900 colonies with an estimated maximum of 3% background of colonies containing religated vector alone. The numbe r of transformants was three fold greater than the number required to have a 99% proba bility of obtaining a clone of each gene, based on calculations using the formula N=ln[1 -P]/ln[1-(I/G)], where I is the size of the Z. mobilis DNA fragments (4x103-6x103 bp), G is the size of the Z. mobilis genome (2.0x106 bp), P is the probability of obtaining th e desired clone, and N is the number of individual clones required to obt ain the desired clone (Ausubel et al. , 1998). Several methods to identify restrictionmodification genes have been developed and published since the discovery of R-M sy stems. One novel method, used to clone DNA MTases, was developed by Piekarowicz et al. (1991b), using an E. coli indicator strain, AP1-200-9. AP1-200-9 wa s engineered based on the observations, described by Heitman and Model (1987, 1991), that the expres sion of site-specific methyltransferases induces the SOS response in E. coli strains that are mcrA +, mcrB +, or mrr + (genes that
106 code for the E coli MDR systems). AP1-200-9 carries mcrA , mcrB and mrr temperature sensitive (ts) mutations and a SOS inducible promoter fused to the lac operon ( dinD1 :: LacZ +). At the nonpermissive (42C) te mperature, AP1-200-9 transformants grow normally. At permissive temperatures (30C), the DNA damage elicited by the host restriction systems on the chromosomal D NA, which is methylated by the cloned MTases, is directly indicated by an increas e in beta-galactosidase expression, and is visualized as colonies on X-gal indicator plates turning blue after transfer from nonpermissive to permissive temperatures. It is also possible to clone restriction enzymes, provided that the level of expre ssion is low enough as not to kill the host but high enough to damage DNA, because such cl ones will induce the SO S response at both the permissive and nonpermissive temperatures. The â€œendo-blue methodâ€ is a method deve loped to directly clone REase genes (Fomenkov et al. , 1994). E. coli indicator strain ER1992 wa s developed based on the observations and work done by Heitman and Mo del (1987, 1991) as well as Piekarowicz et al. (1991b). ER1992 is an E. coli mutant deficient in methyl ase dependent restriction ( hsd, mcrA, mcrBC, mrr ) and has a SOS inducible promoter fused to the lac operon ( dinD1::LacZ+ ). The â€œendo-blue methodâ€ depends on expression of cloned genes, such as REases, producing damage on the unmethylated host DNA and eliciting the SOS response in ER1992 transformants. Conse quently, there will be an increase in galactosidase expression which can be visuali zed as the formation of blue colonies on Xgal indicator plates. DNA sequence analysis of positive clones in the indicator strains identified seven clones from AP1-200-9 and two clones from ER1992 that containe d overlapping inserts
107 from the same region in the Z. mobilis CP4 genome. The cloned and subcloned overlapping sequences were aligned and the consensus sequence deduced (Figure 4). This region contained a 924 bp ORF with seque nce similarity to the Mrr methylation dependent restriction system protein and was designated ZmCP4mrr . ZmCP4Mrr had the highest identity to the 306 amino acid sequence of a putative Mrr protein from Deinococcus radiodurans where 41% (126/306) of the amino acid sequence was identical, and 63% (193/306) of the sequence was identified as positive matches (NCBI BLAST). The amino acid sequence of ZmCP 4Mrr is 37% identical and 63% positive matches to the amino acid sequence of EcoKMrr protein of E. coli (NCBI BLAST). ZmCP4mrr encodes a 307 aa protein with the pr edicted molecular mass of 34.06 kDa. The sequence of ZmCP4mrr was verified in both strands (Figure 5). Analysis of sequences downstream of ZmCP4mrr revealed an ORF with amino acid sequence similarity to a conserved put ative protein with Se l1-like repeats, a subfamily of the TPR (tetratricopeptide re peat) sequence (NCBI BLAST) (Figure 4). The TPR motif is hypothesized to be involved in protein-protein in teractions because they are found in a wide variety of protei ns associated with protein complexes in eukaryotes, bacteria, and arch aea (Blatch and Lassle, 1999). An open reading frame upstream of ZmCP4mrr has significant amino acid sequence similarity to the bifunctiona l phosphoribosylaminoimidazolecarboxamide formyltransferase (AICAR transformylase)/ IMP cyclohydrolase prot ein (ATIC) (Figure 4). ATIC is involved in th e ninth and tenth step in de novo purine biosynthesis (Tibbetts and Appling, 1997).
108 Primer extension and promoter sequence analysis of ZmCP4mrr To determine the transcription initiation site for ZmCP4mrr , Z. mobilis CP4 RNA was extracted, analyzed by primer extensi on, and compared to the genomic nucleotide sequence using the same primer (Figure 6). Primer extension consistently produced a primer extension product terminating at th e thymine 110 bp upstream of the translation initiation site. Sequence analysis of the promoter sequence iden tified a possible RBS (Figure 6), however, the sequence (5â€™AGGA-3â€™) conserved in the RBS, and the sequence (5â€™-ANNNNNCTNG-3â€™) conserved in the -35 region, characteristic of highly expressed genes of Z. mobilis (Ingram et al., 1989) were not found. Transformation Studies of Recombinant ZmCP4mrr To characterize ZmCP4Mrr activity, two recombinant plasmids containing the mrr coding sequence were constructe d. The first construct, p3226/ ZmCP4mrr , contains both the coding sequence and the natural promoter of ZmCP4mrr (a ~1.9 kb subclone). The second construct, pPROTet.E133/ ZmCP4mrr , contains the mrr coding sequence under the control of a heterologous (nonnative) PLtetO-1 promoter. The two constructs were transformed into E. coli strains with different DNA MTase backgrounds: AP1-200-9 (Dam + , Dcm+, HsdM+), BL21PRO (Dam + , Dcm-, HsdM-), GM272 (Dam , Dcm-, HsdM-), GM48 (Dam , Dcm-, HsdM+), Gm161 (Dam , Dcm+, HsdM-) (Table 3). The construct pP3226/ ZmCP4mrr could not be transformed into any dam+ strains. All of the dam mutant strains could be tr ansformed with pPROTet.E133/ ZmCP4mrr ; however, none of the dam + strains could be transformed except for BL21PRO, which expresses the Tet repressor protein that re presses expression of genes cl oned into the pPROTet.E133 expression vector. When expression of the cloned ZmCP4mrr from the construct pPROTet.E133/ ZmCP4mrr was induced with aTc (CLONE TECH, Laboratories, Inc.) in
109 the dam+ host BL21PRO, cell growth was completely arrested and the cells died. Similarly, p3226/ ZmCP4mrr could not be transformed into any dam+ cell line, including BL21PRO. Both pP3226/ ZmCP4mrr and pPROTet.E133/ ZmCP4mrr could be transformed and maintained in all dam mutant strains tested, whether or not these strains had EcoKDcm or EcoKI (HsdM) DNA MTases. Th ese results are interpreted to indicate that ZmCP4Mrr does not recognize EcoKDcm or EcoKI (HsdM) methylation of DNA. In addition, it appears that Zm CP4Mrr specifically restricts sites that overlap EcoKDam sites (5â€™-GmACT-3â€™), and in doing so, it creates enough DNA damage in dam+ hosts that it is lethal to the cell. Partial purification of ZmCP4Mrr Attempts to purify ZmCP 4Mrr from pPROTet.E133/ ZmCP4mrr /BL21PRO cultured transformants were unsuccessful. The purificati on method described in the Material and Methods section produced the best results (~50% purity) compared to purification attempts using ni ckel affinity chromatogra phy (His-Trap kit) and anion exchange (Q Sepharose) (Amersham-Pharmacia, Corp.) (data not shown). Recombinant protein concentrations in TA LON affinity resin column el uted fractions were low, indicating poor expression (d ata not shown). Similar results were seen using the pPROTet.E133 expression vector to express ZmCP4ccrM (see Chapter 5). Discussion In this study, the possible presence of restriction endonucleases and DNA methyltransferases in Z. mobilis CP4 was investigated. R-M systems often function to reduce the frequency of genetic exchange in prokaryotes. Based on the relatively low transformation efficiency of foreign DNA into Z. mobilis CP4, it was hypothesized that Z. mobilis CP4 contains one or more restriction sy stem prohibiting efficient gene transfer.
110 In order to clone any REases or MTases that may be found in Z. mobilis CP4, a genomic library was prepared and screened using the indicator strains AP1-200-9 (Piekarowicz et al. , 1991b) and ER1992 (Fomenkov et al. , 1994). These strains were designed to signal the presence of DNA damage (caused by e xpression of the foreign genes transformed into them), that induces e xpression of the recombinant lac operon from its SOS response promoter, in the presence of X-gal. DNA se quence analysis of isolated positive clones identified a 924 bp ORF, predicted to produ ce a 307 aa protein, with sequence similarity to a putative methyl dependant mrr designated ZmCP4mrr (Figure 4). Since REases and their cognate MTase of R-M systems are typi cally linked (Wilson, 1991) , analysis of the sequences flanking ZmCP4mrr suggest that it is a solita ry methyl-dependant REase. Although primer extension consistently produced a primer extension product terminating at the thymine 110 bp upstream of the translation init iation site, sequence analysis of the -10 region appears more G/ C rich than expected (Figure 6). The -10 regions of Z. mobilis promoters of highly expressed gene s tend to be A/T rich (Ingram et al ., 1989). An A/T rich sequence more consis tent with what was expected for a -10 region was identified ~20 bp upstream of the transcription initiati on site (Figure 6) determined by primer extension. Either ZmCP4mrr has a seemingly atypical -10 region or premature termination may possibly ha ve occurred in the primer extension. Transformation of E. coli strains with various Da m, Dcm, and/or HsdM DNA methyltransferase backgrounds (Table 3) demonstrated that the ZmCP4mrr gene cloned in pP3226 or pPROTet.E133, read ily transform strains when dcm and hsdM are expressed. In addition to dam, dcm , and hsdM, E. coli K12 has recently been shown to contain EcoKccrM (Kossykh and Lloyd, 2004). EcoKCc rM is characterized as an
111 essential cell cycle regulating N6-adenine DNA MTase that methylates the first adenine of ATGCAT sites, contributing to the DNA methylation pattern of E. coli (Kossykh and Lloyd, 2004). Because E. coli strains that contained DNA that was methylated by Dcm (5â€™-CmC(A/T)GG-3â€™), HsdM (5â€™-AmACNNNNNNGTGC-3â€™), and EcoKCcrM (5â€™mATGCAT -3â€™) were not advers ely affected by transformati on with plasmids containing ZmCP4mrr , it can be concluded that the ZmCP4M rr does not recognize methylation at these sites. E. coli strains which are dam+ could not be transformed if plasmid borne ZmCP4mrr was expressed. When the expression of ZmCP4mrr was repressed, dam+ cells were transformed; however, subsequent induced expression of ZmCP4mr r in the dam+ cells resulted in inhibition of grow th and cell death. Dam recognizes and methylates 5â€™-GmATC-3â€™ sites in DNA. Sin ce only those strains with Dam methylation restricted transformation w ith plasmids containing ZmCP4mrr , it appears that ZmCP4Mrr specifically restricts Dam N6-adenine met hylated DNA (5â€™-GmATC-3â€™). This sequence recognition specificity is unlike EcoKMrr. Ec oKMrr is reported to restrict M.HinfI N6adenine methylated DNA (5â€™-GmANTC 3). Because Z. mobilis has a cell cycle regulating DNA MTase ZmCP4CcrM (see Chapter 5) that recognizes and methylates 5â€™GmANTC-3â€™ sites, EcoKMrr would restrict Z. mobilis ZmCP4CcrM methylated DNA. This further supports that observation that ZmCP4Mrr must recognize different sites than EcoKMrr does. The approximately 100 fold higher transformation efficiency of Z. mobilis CP4 by plasmids originating from Z. mobilis , as compared with the same plasmid from E. coli (Table 2), indicates that Z. mobilis CP4 has a restriction system that contributes to low
112 frequency of gene transfer. Plasmi d DNA originating from Dam deficient E. coli strains does not improve transformation efficiency indicating that Z. mobilis CP4 has a restriction system other than ZmCP4M rr. Transformati on efficiency of Z. mobilis remained relatively poor, whether transfor med with methylated DNA or DNA from strains (GM272) reported to have no detectable methylation (recent identification of the essential MTase EcoKCcrM indi cates otherwise). These resu lts suggest that a methyldependent restriction system is not likely the sole transformation barrier into Z. mobilis CP4. In summary, Z. mobilis CP4 likely has another, yet unidentified, methyl-dependant restriction system (perhaps one that recogni zes the EcoKCcrM methylation expressed in all the E. coli strains studied) or a restrictionmodification system that restricts unmethylated sites in DNA. Z. anaerobia has a Type II R-M system, ZanI, that digests unmethylated BstNI sites (Sun and Yoo, 1988). Conse quently, methylation by M.ZanI in vivo results in resistance to BSTNI digestion. Restriction analysis of Z. mobilis CP4 extracted DNA with BstNI shows that Z. mobilis CP4, unlike Z. anaerobia , does not resist BstNI, indicating that Z. mobilis CP4 does not have a ZanI R-M system. There may be other DNA MTases or REases that exist in Z. mobilis CP4 besides ZmCP4mrr that were not isolated in this study. Creating an amplified genomic library as described in the materials and methods involve d initial mass transformation of the library into ER1647. While this st rain was restriction and hsdM deficient (Table 1), so cloned MTases would not have been affected, RE ases that recognize unmethylated DNA, or Dam and Dcm methylated DNA, ma y have been selected against. This would depend on the expression of active cloned REases at levels detrimenta l or lethal to ER1647, thereby
113 diminishing or obliterating the probability of isolating them in subsequent screenings in the indicator strains. It is interesting that multiple clones containing ZmCP4mrr , which restrict Dam methylation, were isolated from the Z. mobilis genomic library, despite amplification of the library by transformation into the dam+ strain ER1647 prior to screening with the indicator strain. It appears that ZmCP4mrr expression from plasmids bearing the large 4-6 kb cloned inserts was low enough for both creation of the genomic library and screening in the indicator strain s as not to be lethal, yet high enough to be detected in the indicator strains due to ZmCp4Mrr cleavage of host genomic DNA at Dam methylated sites. Attempts to transform subcloned ZmCP4mrr , without large flanking sequences, were unsuccessful in dam+ strains of E. coli. The sequences flanking ZmCP4mrr (or the products of the flanking ORFs) may have reduced the expression or affected the activity of Mrr. Expression from subclones of ZmCP4mrr with less flanking sequence is lethal to dam+ hosts when only the coding sequence of ZmCP4mrr was cloned into pROTet.E133, or when the coding and promoter sequence was cloned into pP3226 (Table 3). Transf ormation efficiency of Dam methylated plasmid DNA into E. coli strain GM4715 containing pPROTet.E133/Z mCP4mrr , which expresses mrr from a heterologous promoter, was reduced 100 fold (1 x 105) compared to transformation efficiency of the same plasmid into GM4715 containing pP3226/Z mCP4mrr (1 x 107), which expresses mrr from its natural promoter (data not shown). These resu lts suggest that E. coli does not express ZmCP4mrr well, if at all, from the natural ZmCP4mrr promoter and that expression of ZmCP4mrr in the original clones with larger flanking sequenc es was likely due to read-through.
114 REases of R-M systems that cleave unmodi fied DNA would only be cloned in the indicator strains (AP1-200-9 and ER1992) if th eir expression is poor enough to allow cell growth while still being detectab le to allow isolation of the cloned gene. In addition, if a cloned REaseâ€™s recognition site s are blocked by host methyl ation, they would not be detected in the indicator strain s. Methyl-dependant REases could only have be isolated in these strains if these enzymes recognized host methylation (Dam, Dcm, HsdM, and EcoKCcrM), and restricted the host chromosome at detectable but nonlethal levels. An MTase clone would only be detected if it methylates sites recognized by the hostâ€™s methyl-dependant restriction systems. Finally, R-M systems that require multiple proteins for activity, such as Type I systems, would have less probability of being cloned since they would require multiple genes being cloned and expressed together to produce an active enzyme. Conditions under which partially purifi ed His-tagged recombinant ZmCP4Mrr readily cleaves various p BR322 derived plasmid DNA or DNA were not found, however, prolonged incubation (5 -6 hours in low salt buffers) resulted in degradation of the substrate DNA without produc ing discrete restriction fr agments (data not shown). Although definitive conclusions can not be dr awn from the attempts to determined in vitro activity of recombinant Zm CP4Mrr, the general observati ons deserve mention. The degree of degradation of DNA substrates that had been methylated in vitro by various commercially available DNA MTases (M.D am, M.BamHI, M.SssI, M.HpaII, and M.MspI) and incubated with partially purified recomb inant ZmCP4Mrr varied. Preliminary observations suggest that ZmCP4Mrr recognizes both Dam (5â€™-GmATC-3â€™) and BamHI (5â€™-GGATmCC-3â€™) methylated DNA, but does not recognize M.HpaII
115 methylated DNA (5â€™CmCGG-3â€™) under the reacti on conditions tested. Considering the fact that Dam (5â€™-GmATC-3 â€™) recognition sites overlap M.BamHI (5â€™-GGATmCC-3â€™) recognition sites, perhaps ZmC p4Mrr recognizes and restricts methylated sites containing the sequence 5â€™-GATC-3â€™ but has no preference to the specific type of methylation (N4cyctosine or N6adenine) it recognizes. Although these obs ervations were preliminary, ZmCP4Mrr may recognize multiple methylati on patterns (both adenine and cytosine methylated DNA) as seen in EcoKMrr.
116 Table 1. Bacterial strain, pl asmid, and primer description. Strain, Plasmid, and Primer Genotype or Description Reference or Source Escherichia coli AP1-200-9 FendA1 thi-1 supE44 hsdR17 mcrB251 mrr253 mcr252 lacZ ::Tn10 [ dinD1 :Mu dI1734 (Kanr lacZ+)] [Fâ€™ lacIq lacZ ::Tn5]. An indicator strain that carries mcr A, mcr B and mrr temperature sensitive (ts) mutations Piekarowicz et al. (1991b) BL21PRO ompT, hsdSB( rB -mB -), gal, dcm, F-; containing tetR, Placi q /laci, and ( Spr) on an autonomously replicating plasmid. CLONETECH Laboratories, Inc. DH5 Fâ€™/ endA1 hsdR17 (rk mk) supE44 thi-1 recA1 gyrA (Nalr) relA1 (lacIZYA-argF)U169 deoR ( 80 dlac(lacZ)M15 ) New England Biolabs, Inc. ER1647 F-, fhuA2, supE44, mcrA1272:: Tn 10 (Tetr) , trp-31, his-1, rspL104 (Strr), xyl-7, mtl-2, metB1, ( lacZ ) r1, ( mcrC-mrr ) 114:: IS 10 (Tetr), recD1041 . http://users.uma ssmed.edu/ martin.marinus/ dstraines.html ER1992 F( argF-lac ) U169 supE44 e14dinD1 ::Mu dI1734 (Kanr lacZ+) rfbD1? relA1? endA1 spoT1? Thi-1 ( mcrC-mrr ) 144 ::IS10 Fomenkov et al. (1994) GM48 dam-3, dcm-6, thr-1, leuB6, ara-14, tonA31, lacYI, tsx-78, glnV44, galK2, galT22, thi-1. http://users.uma ssmed.edu/ martin.marinus/ dstraines.html GM161 F-, dam-4, thr-1, leuB6, supE44, lacY1, tonA21, hsdS1 (rkmk). http://users.uma ssmed.edu/ martin.marinus/ dstraines.html GM272 F, dam-3, dcm-6, hsdS21, metB1, mtl-2, galK2, galT22, SupE44, tonA2 or A31, lacYI or Z4, (thi-1)? http://users.uma ssmed.edu/ martin.marinus/ dstraines.html GM4715 Dam-16 ::KanD ( mcrB-hsd-mrr ) 10 mcrA3 trp31 his-1 tonA2 rspL104 D( lacZ )r1supE44 xyl-7 mtl-2 metR1 argG6 http://users.uma ssmed.edu/ martin.marinus/ dstraines.html HB101 F(gpt-proA)62 leuB6 supE44 ara-14 galK2 laY1 (mcrC-mrr) rpsL20 (Strr) xyl-5 mtl-1 recA13 New England Biolabs, Inc Zymomonas mobilis CP4 wild-type ( Zymomonas mobilis subspecies mobilis var. recifensis ) L.O. Ingram
117 Table 1. Continued. Strain, Plasmid, and Primer Genotype or Description Reference or Source Plasmids pLOI1844 (Cmr). Broad host range vector. RSF1010 derivative. L.O. Ingram pPROTet.E133 (Cmr). Expression vector using a PLtetO-1 promoter ( PL promoter of phage with the repressor sites replaced by two copies of operator2 of the Tn 10 tetracycline resistance operon) repressed by the Tet repressor protein. (6xHN: His-tag) CLONETECH Laboratories, Inc. pBR322 (Tetr, Ampr), 4361 bp ColE1 compatibility group cloning vector. Balbas et al. (1988) pP3226 (Ampr ) pBR322 derivative with Tetr deleted and a multiple cloning site from pNEB193 inserted. This study pNEB193 (Ampr) pUC19 derivative. ColE1 compatibility group cloning vector. New England Biolabs, Inc. Primers CACGACGTTGTAAAACGAC Forward for pNEB193(M13/pUC) GGATAACAATTTCACACAGG Reverse for pNEB193(M13/pUC) LI-COR, Inc. TGCCTGACTGCGTTAGC Reverse for pP3226 CGTCTTCAAGAATTCGAGC Forward for pP3226 CACATCAGCAGGACGCACTGA forward for pPROTet.E133 Sequencing Primers CGCTCGCCGCAGCCGAAC Reverse for pPROTet.E133 This study CAAGGTACCAATGGCGGTTCCCAGTTT Forward Primer Mrr#3 (To clone ZmCP4mrr into KpnI site of pROTet.E133) GGCGGTACCCATTCAGGATTAAAGAAATC Reverse Primer This study Primer extension AGGCAGCATCAATTCGGAA MrrEXT2 This study
118 pBR3224361 bps 1000 2000 3000 4000 HindIII EcoRV BamHI StyI AvaI PvuII NdeI SapI BsaI AseI PstI PvuI ScaI AatII EcoRI Tet Rop Ori APpP32263226 bps 500 1000 1500 2000 2500 3000 AvaI DsaI PvuII NdeI SapI BsaI AseI PstI PvuI ScaI AatII ApoI EcoRI BanII SacI Asp718I KpnI AvaI SmaI XmaI AscI BamHI PacI XbaI SalI PmeI BspMI PstI SphI HindIII ORI ROP AP polylinker Figure 2. pP3226 construct. A. Restriction map of pBR322 (Balbas et al. , 1988). B. A restriction map of pP3226, a 3226 bp de rivative of pBR322. The AP gene codes for -lactamase that confers ampicillin resistance. The multiple cloning sites, derived from the pNEB193 polylinker, are highlighted in cyan. The exact positions of the c oding sequences, the multiple cloning sites, and the origin of replication are: Ap 2873-2085, Rop 707-893, multiple cloning sites 3148-5, and ORI starting at 1324.
119 Table 2. Dependency of transformation efficiency of Z. mobilis CP4 on plasmid source. DNA MTase background Source of Plasmid pLOI1844 dam dcm hsdM Transformation efficiency CFU/g DNA E. coli DH5 + + + 2 x 104 E. coli HB101 + + 4 x 104 E. coli GM48 + 5 x 104 E. coli GM272 2 x 104 Z. mobilis CP4 Native 6 x 106 Plasmid pLOI1844 extracted from E. coli strains, with different DNA MTase backgrounds, or Z. mobilis CP4 was electroporated into Z. mobilis CP4 with equal amounts of DNA. Efficiency of tr ansformation is expressed as CFU/g DNA.
120 Figure 3. Restriction analysis of DNA isolated from Zymomonas mobilis CP4. E. coli ER1647 and Z. mobilis were transformed with pLOI1844 plasmid. Lane 1 shows the molecular weight marker /HindIII Digest containing DNA fragments of 23.13 kb, 9.42 kb, 6.56 kb, 4.36 kb, 2.32 kb, and 2.03 kb, from top to bottom. Lanes 2-4 contai ns plasmid DNA extracted from Z. mobilis . DNA extracted from Z. mobilis and digested with BstNI, TfiI, and HinfI, are shown in lanes 2, 3, and 4 respectivel y. Lanes 5-7 contains plasmid DNA extracted from ER1647. The DNA frag ments produced by digestion of pLOI1844 from ER1647 with BstNI, TfiI, and HinfI are shown in lanes 5, 6, and 7, respectively. The two major BstN I digestion products of pLOI1844 are indicated by the red a rrows in lanes 2 and 5. 1 2 3 4 5 6 7
121 Figure 4. Identification of the ORFs of a cloned Z. mobilis CP4 region containing ZmCP4mrr. The boundaries of the genomic region cloned from Z. mobilis CP4 is indicated by the verti cal lines in the diagram. ZmCP4mrr 924 bp ORF is represented by the yellow box. The fl anking ORFs are represented by green boxes. The putative proteins encoded by the flanking ORFs, based on amino acid sequence similarity (NCBI-BLAST), are indicated below each box. The direction of transcrip tion of each ORF is indicated by the open arrows. TonB-dep./CirA/FepA-like outer membrane receptor (ion transporter) (C-terminal) Hypothetical protein TRP repeat Sel-1 family motif (protein/protein binding) Bifunctional AICAR transformylase/ IMP cyclohydrolase (ATIC) (purine biosynthesis) Hypothetical protein (C-terminal) * * glyoxalase/dioxygenase lactoylglutathione lyase ZmCP4mrr 924 bp
122 CCGCCATTTCCGGCAT TAA AAACAGCTAACGGGAGGGCGGTTTATTCGGCC TTCCTGTTTCTAATCTATAAGGCGTGCTTGGAGGTTAGCTTTGGGCGCGCC TTTTTAAAAAATGAAGAGGGCTTTCCTGTCTGTTATTCTGTCGCCGGATAG GCGAGTATAAGTTTTTCTTTCGATAGATTATTTTAAGTAAAAGGCTCGATA 1 ATG GCGGTTCCCAGTTTTTCCGAATTGATGCTGCCTGTTTTGCGTTTTGCC M A V P S F S E L M L P V L R F A 52 GCAACAGGAAAAAAGCGTATTGCGGAAACCGAAACGGCGATTGCCGATGGA A T G K K R I A E T E T A I A D G 102 TTAGGGCTTTCGGCAGAGGATCGCGAAGCGATGCTGCCTAGTGGTCGGGAA L G L S A E D R E A M L P S G R E 153 AGACTGCTTTATAACCGCATCGCATGGGCCAAAAAATATCTCACGCAAGCC R L L Y N R I A W A K K Y L T Q A 204 GGATTACTGGATATTCCGGCAAGAGGGCAATTTATTATCAGTAAGGCTGGA G L L D I P A R G Q F I I S K A G 255 AAAAGGCTTTTACGACGCAATCCTCCTGAAATTACAGTAGCGACCTTAAAG K R L L R R N P P E I T V A T L K 306 CAATATCCGTCTTTTATCGAATATTGTAACCGCAATTCTAAATCCGACACG Q Y P S F I E Y C N R N S K S D T 357 CGCGAAGATGTCGTTCAGATTGTCGCCCAAGAAACGGTGACCGGAACACCA R E D V V Q I V A Q E T V T G T P 408 GAAGAGCGGATTGATGCCGCCCATGAAGAGCTTCAAAATGAATTACGCACT E E R I D A A H E E L Q N E L R T 459 GATATTTTGAACCAGATTTCCGAGAAAAATCCCAGTTTTTTTGAACAGCTG D I L N Q I S E K N P S F F E Q L 510 ATCGTCGATTTGATGGTGGCTATGGGCTATGGCGGCAATCATGAGGATGCG I V D L M V A M G Y G G N H E D A 561 GCTCGTCGGATTGGGGGAACGGGTGATGGTGGTGTCGATGGTGTTATCAAT A R R I G G T G D G G V D G V I N 612 GAAGATCGTTTGGGCATTGATTGCATCTATGTTCAGGCCAAACGTTATGCT E D R L G I D C I Y V Q A K R Y A 663 GCCCATGTCAGTGTCGGACGGCCGGAAATACAGGCTTTTGTCGGTAGTCTG A H V S V G R P E I Q A F V G S L 714 GTCGGGCATCGTGCTACAAAAGGGGTTTTTGTAACGACCTCTTCTTTTAGC V G H R A T K G V F V T T S S F S 765 GCCCCCGCTTTGGAATATGTCGAGCATTTGCCGCAGCGGGTTATTCTGATC A P A L E Y V E H L P Q R V I L I 816 AATGGCGAAAAACTCGCAAATTTAATGATCGAACATAATGTTGGTGTCAGA N G E K L A N L M I E H N V G V R 867 ACCAGTCGCAGAATAGAAATCAAACGGGTTGATTTGGATTTCTTTAATCCT T S R R I E I K R V D L D F F N P 918 GAA TAG GCTTTTTGAAAAAgccgtttcctgtattgttacgattctggcttt E * Figure 5. REase sequence: ZmCP4mrr . The sequence shown includes the ZmCP4mrr DNA sequence in the 5â€™ to 3â€™ direction as well as some of the upstream and downstream sequence. The DNA sequence in uppercase was verified in both strands. The predicted amino acid sequen ce is presented in single letter code below the first nucleotide of each codon. ZmCP4mrr is a 924 bp DNA sequence producing a predicted 307 aa transl ated polypeptide. The translation initiation site (ATG) of ZmCP4mrr is underlined and highlighted in green and the termination codons for ZmCP4mr r and the upstream ORF (TAG and TAA, respectively) are underlined and highlighted in red.
123 Figure 6. Primer extension analysis of ZmCP4mrr . The four ZmCP4mrr sequencing lanes 14-17, read from bottom to top, is the reverse complement to the sequence shown (presented 5â€™ to 3â€™). The position of the primer extension products in lanes 1, 8, 10, and 12, is indicated by the red arrow. The transcription initiation site (T) is highli ghted in red in the promoter sequence. The translation initiation site (ATG) is highlighted in grey. A possible RBS is underlined. AAACAGCTAACGGGAGG GCGGTTTATTCGGCCTTC CTGTTTCTAATCTATAAG GCGTGCTTGGAGGTTAGC TT T GGGCGCGCCTTTTTA AAAAATGAAGAGGGCTT TCCTGTCTGTTATTCTGTC GCCGGATAGGCGAGTAT AAGTTTTTCTTTCGATAG ATTATTTTAAGTAAAAGG CTCGATA ATG GCGGTTCC CAGTT A CGT 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
124 Table 3. Transformation of cloned ZmCP4mrr into E. coli with different MTase backgrounds E. coli strain dam dcm hsdM pPROTet.E133/ ZmCP4mrr transformation pP3226/ ZmCP4mrr transformation AP1-200-9 + + + + repressed BL21PRO + aTc induced GM161 + + + GM48 + + + GM272 + + E. coli strains, with different MTase b ackgrounds were transformed with the p3226 plasmid containing the subcloned ZmCP4mrr or pPROTet.E/ ZmCP4mrr expressing a recombinant His-tagged Zm CP4Mrr. Recombinant His-tagged ZmCP4mrr expression is repressed when pPROTet.E/ ZmCP4mrr is transformed into BL21PRO (a strain that expresses the Tet repressor) but can be induced with aTc. ZmCP4mrr expression is not repressed when pPROTet.E/ ZmCP4mrr is transformed into the other E. coli strains listed.
125 CHAPTER 5 CLONING AND CHARACTERIZATION OF A CELL CYCLE REGULATING DNA METHYLTRANSFERASE FROM ZYMOMONAS MOBILIS SUBSPECIES MOBILIS CP4 Introduction Most prokaryotic DNA MTases are part of a restriction-modifi cation system pair (Wilson, 1991; Jost and Saluz, 1993; Hyone -Myong, 1996; Roberts and Macelis, 1996; Bujnicki, 2001; Low et al. , 2001; Jeltsch, 2002). However, several solitary DNA MTases have been discovered such as Dam, Dcm, CcrM, and the â€œantirestrictionâ€ MTases of various bacteriophages (Bujnicki, 2001). CcrM is solitary DNA methyltr ansferase first iden tified in the dimorphic freshwater alpha-proteobacterium Caulobacter crescentus (Zweiger et al. , 1994). CcrMI is only present in C. crescentus during the predivisiona l stage of the cell cy cle (end of S-phase), after replication initiation (Stephens et al. , 1996). CcrMI was characterized as a N6adenine DNA MTase that methylates 5â€™-GmANTC3â€™ sites, appears to be involved in cell cycle regulation (Stephens et al. , 1995), and is required fo r cell viability (Stephens et al. , 1996). The mechanism(s) by which CcrM methyl ation regulates the ce ll cycle has yet to be identified; however, in studies of C. crescentus mutants, where CcrMI is active throughout the cell cycle and a fully met hylated chromosome is maintained, DNA replication produced multiple chromoso mes per cell, the cell morphology was significantly altered, and cell division was defective (Zweiger et al. , 1994). More recent studies of the CcrMI of C. crescentus suggested that it is autoregulatory, regulates
126 expression of the global cell cycle regulator CtrA (Reise nauer and Shapiro, 2002), and may possibly regulate expression of ot her genes (Chen and Shapiro, 2003). CcrM homologues have been identified and characterized in alpha-proteobacteria Rhizobium meliloti (Wright et al. , 1997), Brucella abortus (Robertson et al. 2000), and Agrobacterium tumefaciens (Kahng and Shapiro, 2001), and are believed to be widespread in alpha-pro teobacteria (Reisenauer et al ., 1999b). More recently, a solitary CcrM DNA MTase M.EcoKCcrM, was cloned and characterized in Escherichia coli K12 (Kossykh and Lloyd, 2004). It appears to met hylate the first adenin e of its recognition site ATGCAT (Kossykh and Lloyd, 2004). Like the CcrMI in C. crescentus , the E. coli CcrM is cell cycle regulated and essen tial for cell viability (Kossykh and Lloyd, 2004). Zymomonas mobilis is a plant sap alpha-prote obacterium (Swings and De Ley, 1977). Little is known about the regulatory mechanisms Z. mobilis uses to control protein expression and no thing is known about Zymomonas cell cycle regulatory mechanisms. We report the cloning and characterization of a putative DNA methyltransferase from a Z. mobilis CP4 genomic library. Amino acid sequence analysis revealed that the putative solitary DNA MTas e has significant sequenc e similarity to the cell cycle regulating DNA MTase CcrMI found in C. crescentus (NCBI BLAST) . Analysis of the activity of purified reco mbinant ZmCP4CcrM demonstrates that it methylates the same site as CcrMI of C. crescentus. DNA sequence analysis of the promoter region of ZmCP4ccrM revealed similarities to ccrM promoter elements found in C. crescentus . Overexpression of plasmid borne recombinant ZmCP4ccrM in Z. mobilis alters the cell cycle a nd causes changes in cell morphology and cell division. Collectively, the results of this study show that ZmCP4CcrM is a N6-adenine specific
127 solitary DNA methyltransferase that is likely involved in cell cycle regulation as reported for the CcrM homologues found in C. crescentus and other alpha-proteobacteria. Materials and Methods Bacterial Strains, Plasmids, Media, and Reagents Bacterial strains and plasmids are describe d in Table 1 of Chapter 4 and Table 4. Growth media used for bacterial culture s, the method used in preparation of electrocompetent cells, transformation t echniques, and general DNA preparation are described in the Materials and Methods section of Chapter 4. Restriction endonucleases, DNA methyltran sferases, alkaline calf intestinal phosphatase, T4 kinase, T4 DNA ligase, and Qu ick Ligation kit used in this study were purchased from New England Biolabs, Inc. a nd were used according to the supplierâ€™s protocols. T4 DNA polymerase from New E ngland Biolabs, Inc. was used in bluntending DNA fragments, according to supplierâ€™s protocols. PfuT urbo DNA polymerase purchased from Stratagene was used for PCR amplifications (unle ss stated otherwise), according to the supplierâ€™s protocols. DNA molecular weight standards /HindII Digest and 100 bp DNA Ladder were purchased from Ne w England Biolabs and were used in agarose and polyacrylamide gel electrophoresis . Low range protein molecular weight standards for SDS-PAGE were purchased from Bio-Rad Laboratories, Inc. All chemicals and reagents used were reagent grade. DNA Preparation and Screening of Zymomonas mobilis CP4 Clones The methods used in preparation of the Z. mobilis CP4 genomic library (amplified in ER1647), its screening in E. coli indicator strains AP1-200-9 (Piekarowicz et al. , 1991b), subsequent subcloning of positive clone s, and DNA sequence analysis to identify and annotate the ORFs found in each clone were described in the Materials and Methods
128 section of Chapter 4. MTase clones and subc lones were maintained in the endonuclease deficient E. coli strains ER1992 or GM4715. Primer Extension Analysis of ZmCP4ccrm mRNA The ccrM transcription initiation site was iden tified by primer extension analysis using the methods for RNA isolation, primer extension, and standa rd dideoxy sequencing as described in the Material and Methods s ection of Chapter 4. Custom designed IRD41labeled primers to internal ZmCP4ccrM sequence (CcrMEXT3) were purchased from LICOR Inc. and used for primer extension (Table 4). A standard dideoxy sequencing ladder of ZmCP4ccrM was generated with the same primer used for the primer extension. Recombinant Expression of the ZmCP4CcrM Protein The ZmCP4ccrM coding sequence was PCR amplif ied using the pair of custom primers CcrM#3 purchased from SIGMA-GENOSY S (Table 4) containing KpnI sites. The PCR product was purified using a QIAqui ck PCR purification kit, digested with KpnI, and cloned into the K pnI site of the expression vector pPROTet.E133, under the control of a heterologous ( nonnative) anhydrotetracycline (a Tc) inducible promoter PLtetO1. The recombinant ZMCP4CcrM protein produced by pPROTet.E133/ ZmCP4ccrM has a C-terminal polyhistidine tag to facilitate protein purification. The host strain BL21PRO, expressing the Tet repressor protein, was transformed via electroporation with pPROTet.E133/ ZmCP4ccrM . The recombinant plasmid sequence was verified by standard dideoxy sequencing using the cu stom sequencing primers designed for pPROTet.E133 (Chapter 4, Table 1). The ZmCP4ccrM sequence was PCR amplified using the pair of CcrM#6 custom primers purchased from XX IDT Integrated DNA Technologies, Inc. (Coralville, IA.) (Table 4), extracted using the QIAquick PCR purification kit, digested with NdeI and
129 XhoI, and cloned into the NdeI to XhoI site of the expression vector pET-24b (Novagen, Inc.), under the control of an IPTG inducible T7 lac promoter. The CcrM#6 primers were designed to PCR amplify the ZmCP4CcrM OR F from the second possible translation initiation site (Figure 8) to the second nucleot ide of the translation stop site, and included sequence to produce a PCR product flanked by Nd eI and XhoI recognition sites. The recombinant ZMCP4CcrM protein produced by pPET-24b/ ZmCP4ccrM has a C-terminal polyhistidine tag to facilitate protein purification. DE3 strains of AP1-200-9, GM272, and GM4715, E. coli strains were created using the DE3 Lysogenization kit (Novagen, Inc) as recommended by the supplierâ€™s protocol. The recombinant plasmid pPET-24b/ ZmCP4ccrM was transformed via electroporation into HMS174(DE3) , AP1-200-9(DE3), GM272(DE3), GM4715(DE3), and Rosetta(DE3). To determine which DE3 E. coli strain produced the highest level of protein for subsequent purification of reco mbinant ZmCP4CcrM, the transformants were cultured (OD550=0.5) and induced for 2 hrs with 1 mM IPTG. Cells were harvested by centrifugation (3,000 x g ) from 1 ml of culture and nativ e proteins were extracted using the TALON metal affinity resin (Sepharose CL-6B) (CLONETECH, Laboratories, Inc.), according to the supplierâ€™s mini-scale protein pur ification protocol. Protein samples were analyzed on an SDS-PAGE gel. Purification of a Recombinant ZmCP4CcrM Protein To isolate recombinant His-tagged Zm CP4CcrM protein, 300ml cultures of pPROTet.E133/ ZmCP4ccrM or pET-24b/ ZmCP4ccrM /HMS174(DE3) were grown to OD550=0.5-0.6 and expression was induced with addition of aTc (100-500 ng/ml) or IPTG (1 mM) respectiv ely. After two hours of induction the cells were harvested by
130 centrifugation, washed in Bu ffer A (20 mM Na-phosphate pH 7.4, 20% glycerol, 0.5 M NaCl, 5 mM -ME) and stored at -75C. Frozen cell pellets were thawed on ice and resuspended in 1.2 ml of break buffer (50 mM Na-phosphate, pH 4, 20% glycerol, 1 mM EDTA, 2.5 mM PMSF). -ME and Egg white lysozyme (1 0mg/ml) (Sigma-Aldrich, Co.) were added to the cell slurry to 5 mM and 0.75 mg/ml, respectively, and incubated at room temperature for 20 min and on ice for 30 min. The cells were lysed by sonication, and cell debris was sediment ed by centrifugation at 10,000-12,000 x g for 20 min at 4C. The cleared lysate was filtered through a 0.45 micron cellulose acetate membrane syringe filter (Corning Life Sciences, Corning, NJ.), a nd 1-1.25 ml of the filtrate was applied to a 1.5 cm diameter by 51-54 cm long Sephadex G-75 (Amersham-Pharmacia, Corp., Piscataway, NJ.) column at 4C, pre-equili brated with Buffer A. The eluted 0.5-3 ml fractions were collected. Fr actionated proteins in the 14-97 kDa range were resolved and analyzed by SDS-PAGE protei n electrophoresis (Laemmli, 19 70). The eluted fractions determined to contain the highest concen trations of recombinant ZmCP4CcrM (43.7 kDa) were combined and adjusted to 10 mM imidazole. The collected Sephadex G-75 fractionated protein was then applied to a 1 ml Ni2+ charged HiTrap Chelating Sepharose HP column (His-Trap Kit, Amersham-Pha rmacia, Corp.) as recommended by the supplierâ€™s protocol. His-tagged recombinan t ZmCP4CcrM protein was eluted with a stepwise imidazole gradient (10 mM , 20 mM, 40 mM, 60 mM, 100 mM, 300 mM, 500 mM) in Buffer A. Protein concentration was determined by the standard Bradford method (Ausubel et al. , 1998) and the protein purity wa s determined by SDS-PAGE and standard coomassie staining (Laemmli, 1970). The protein concentration (mM) was calculated based on the predicted molecular weight of the recombinant ZmCP4CcrM.
131 Analysis of the Methylation Specifici ty of ZmCP4CcrM Using the Type IIS Restriction Endonuclease BsaI To determine the nucleotide methylated by the recombinant ZmCP4CcrM protein, a strategy based on a method described by Posf ai and Szybalski (1988) was used which employs a Type IIs restri ction enzyme, BsaI, and [3H]-AdoMet as the methyl donor. Substrate DNA (pNEB193 PacI/BsaI) was engineered to co ntain a BsaI restriction site whose cleavage site overlap s a CcrM methylation site (Figure 20). The pNEB193 plasmid was digested with XbaI and BamHI, to remove the PacI site, and the linear vector DNA fragment was extracted after agar ose gel electrophoresis as described in the Material and Methods section of chapter 4 on vector construction. Two single stranded custom oligos, 5â€™-GATCCGGTCT-3â€™ and 5â€™-CTAGAGACCG -3â€™ purchased from GenoMechanix, L.L.C. (Gainesville, Fl.), were annealed creating a short DNA fragment containing a BsaI recognition site and stic ky ends complementary to XbaI and BamHI created sticky ends of the digested plasmi d DNA. This BsaI site containing dsDNA fragment was ligated into th e construct pNEB193/XbaI/BamHI/ PacI using T4 DNA ligase, creating the 2708 bp plasmid construct pNEB193 PacI/BsaI. The construct was confirmed by sequencing. The substrate DNA pNEB193 PacI/BsaI (60 ng/ l) was incubated with 0.5-1 mM recombinant ZmCP4CcrM, 6-8.5 M [3H]-AdoMet (Perkin Elmer, Boston, MA.), 20 M AdoMet, 50 mM potassium phosphate buffer pH 7.5, and 0-50 mM potassium acetate, in 50 l reaction volumes at 30 C for 1 hr. The reactions were stopped by incubation at 65 C for 5 min. The reaction volu mes were increased to 100 l with buffer conditions adjusted for subsequent PvuII di gestion (incubated for 2 hrs at 37 C), followed by BsaI or HinfI digestion (inc ubated for 2 hrs at 50 C), according to the supplierâ€™s protocol. The
132 DNA restriction fragments were phenol-chl oroform extracted and ammonium acetateethanol precipitated by standard protoc ols (Berger and Kimmel, 1987). The DNA was dissolved in TE pH 8.0, and 20 l samples were mixed with PAGE loading dye (Laemmli, 1970). The restriction fragme nts were fractionated by nondenaturing-PAGE (Ausuel et al. , 1998) at 50 volts for 1 hour. A 100 bp DNA Ladder molecular weight standard, and the 344 bp PvuII substrate DNA fragment, were fractionated in parallel lanes. The gels were EtBr stained and th e DNA fragments were excised from the gel. DNA was eluted from the gel slices us ing the protocol described by Berdis et al. (1998). Gel slices were incubated in elution bu ffer (0.5 M ammonium acetate pH 8.0 and 1 mM EDTA) overnight at 37 C to elute the DNA. The samp les containing the eluted DNA were spotted onto DE81 DEAE cellulose ion exchange papers (Whatman International Ltd.), washed three times with 0.5 M Na-phos phate buffer pH 7, and washed once with ethanol. The D81 discs were air-dried, 5 ml of Betaplatescint scin tillation fluid (PerkinElmer) was added to each filter, and the amount of [3H]-labeled DNA was determined using a scintillation coun ter (Beckman model LS6500). Construction of pBROriV ROP/ ZmCP4ccrM and pBBR1MCS/ ZmCP4ccrMPRO for Overexpression of ZmCP4ccrM in Zymomonas mobilis CP4 To create a vector expressing ZmCP4ccrM from a heterologous promoter, first the 4096 bp construct pBROriV (Figure 7) was cr eated by replacing th e AhdI/AatII sequence of pBR322, containing the Ap (Ampr) gene, with a PCR amplified 664 bp AatII/AhdI DNA fragment from RSF1010, c ontaining the origin of re plication OriV, using the custom primer pair ORIV purchased from XX IDT, Inc. (Table 4). The vector sequence of pBROriV was PCR amplified using the Expanded Long Template PCR System from Roche Molecular Biochemicals (Mannheim, Ge rmany) as recommended by the supplierâ€™s
133 protocol, with the custom prim er pair PBRORIV purc hased from XX IDT, Inc. (Table 4). The resulting PCR product deleted the sequen ce starting between the RBS and translation initiation site of ROP to the PvuII site, and created a unique AseI re striction site. The PCR product was blunt-ended using T4 DNA pol ymerase and circularized using T4 DNA ligase, producing the 39 59 bp construct pBROriV Rop (Figure 7). The cloned ZmCP4ccrM gene was PCR amplified using CcrM #4 custom primers purchased from XX IDT, Inc. (Table 4) and subcloned into the AseI/PvuII site of pBROriV Rop, creating pBROriV ROP/ ZmCP4ccrM , which expresses ZmCP4ccrM from the Rop promoter. Z. mobilis CP4 containing pLOI1844 was transformed with pBROriV ROP/ ZmCP4ccrM to produce transformants overexpressing ZmCP4ccrM . Co-expression of a helper plasmid (like the broad host range vector pLOI188) that produces Rep proteins required for replication of plasmids with an OriV origin of replication is requi red for replication of pBROriV ROP/ ZmCP4ccrM in Z. mobilis CP4. A second vector, expressing ZmCP4CcrM from its own promoter was created. The ZmCP4ccrM sequence with its promoter region was PCR amplified using CcrMPRO#2 custom primers (purchased from SIGMA-GENO SYS) and cloned into the XhoI/SpeI site of the broad host range vector pBBR1MCS, creating pBBR1MCS/ ZmCP4ccrMPRO which expresses ZmCP4ccrM from its native promoter. CcrMPRO#2 primers were designed to include 260 bp of sequence upstream of the first possible translation initiation site (Figure 8). Z. mobilis CP4 was transformed with pBBR1MCS/ ZmCP4ccrMPRO to produce transformants overexpressing ZmCP4ccrM.
134 Z. mobilis CP4 transformants were verifi ed to contain the pBROriV ROP and pBBR1MCS expression plasmids (with or with out ZmCP4ccrM ) by plasmid extraction and restriction analysis. The transformants were cultured and assa yed for changes in growth rate by monitoring the increase in OD550 over time using a spectrophotometer. Maximal cell density of cultured transformants was examined by determining the OD550 of diluted 3 day old cultures. The cell viability of th e transformants was examined by plating 20-500 l of 3-day cultures, each diluted to the same OD550, onto GYx plates. Examination of Cell Morphology and DNA Distribution Using Microscopy To examine changes in cell morphology, exponentially and stationary Z. mobilis cells containing plasmid borne copies of ZmCP4ccrM were examined using Bright-field and Nomarski microscopy. Z. mobilis CP4/pLOI1844 cells transformed with pBROriV Rop or pBROriV Rop/ ZmCP4ccrM were cultured to an OD550= ~0.7 and examined to determine the relative number of cells showing gross morphological changes using a Petroff-Hausser counting chambe r with a Bright-field microscope. To examine cell morphology and DNA distribution, Z. mobilis CP4/pLOI1844 transformed with pBROriV Rop/ ZmCP4ccrM or pBROriV Rop was cultured and cells from 3 ml was harvested at an OD550=0.5 and from overnight cu ltures. The harvested cells were resuspended in 500 l fixing solution (12.5% fo rmaldehyde, 150 mM Naphosphate buffer pH 7.4), incubated for 20 min at room temperature, washed twice with chilled PBS buffer (137 mM Na Cl, 2.7 mM KCl, 4.3 mM Na2HPO 4, 1.4 mM KH2PO4) (Ausubel et al. , 1998), and resuspended in 1 ml chille d PBS to examine unstained cells. To examine DNA distribution, formaldehyde fixed cells were resuspended in 500 l
135 chilled PBS buffer for subsequent staining w ith DAPI (Sigma-Aldrich). The prepared cells were stored at 4 C overnight. DAPI stock solution was prepared at 1 mg/ml in DMSO (Sigma-Aldrich) and diluted in PBS to create DAPI-PBS solution (900 l PBS buffer and 2 l of DAPI stock) when ready to stai n prepared formaldehyde fixed cells. The formaldehyde fixed cells (300 l) were centrifuged, resuspended in 300 l of DAPIPBS solution, incubated for 20 minutes at room temperature, washed three times with 600 l PBS buffer, and resuspended in 100 l PBS buffer. DAPI stain binds to DNA and fluoresces blue under ultraviolet light at a maximal absorption wavelength of 358 nm and a maximal visible emission wavelength of 461 nm (Microscopyu.com). Cells and reagents, during and after DAPI staining, were protected from light exposure. The slides were prepared by mixing 5 l of the formaldehyde fixed cells, with or without DAPI staining, with 5 l of molten 1% low melt SeaPlaque GTG agarose (FMC Bioproducts, Rockland, ME.), dispensed onto warm glass slides, and spread by pressing a glass cover slip onto the sample. The speci mens were observed using Nomarski optics (differential interference contrast, DIC) microscopy at 200X and 400X magnification. The DAPI stained cells were visualized with UV illumination. Results Identification of a Z. mobilis CP4 DNA MTase ORF and Sequence Analysis A Z. mobilis CP4 library was screened using E. coli AP1-200-9 DNA damage indicator strain and several positive clones were isolate d. DNA sequences identified three overlapping clones that contained an 1149/1122 bp ORF with significant amino sequence similarity to DNA methyltransf erases. The ORF has two potential ATG translation initiation codons producing a pr edicted 382 or 373 amino acid polypeptide, with an estimated average mass of 43.55 kD a or 42.49 kDa respectively (Figure 8). The
136 best match to this ORF had 62% (218/348) identical amino acids and 79% (275/348) positive matches with the previously charac terized 377 amino acid solitary adenine DNA methyltransferase M. BabI of Brucella abortus (NCBI BLAST). M.BabI is a homologue of the solitary adenine DNA methyltransferase M.CcrMI of C. crescentus , as well as the solitary adenine DNA CcrM methyltransferas es found in several other Gram-negative alpha-proteobacteria (Wright et al. , 1997). The ORF immediately upstream of the putative DNA MTase, designated ZmCP4ccrM , has significant amino acid sequence si milarity to ribonucl ease HII (Figure 9). Upstream from this putative RNase H II homologue is an ORF with significant amino acid similarity to a Phosphomethylpyri midine kinase/hydroxymethylpyrimidinephosphate kinase. HMPP-kinase is a biosynthetic enzyme involved in synthesis of thiamine pyrophosphate, an esse ntial cofactor mediating al dehyde transfer (Voet and Voet, 1995; Peterson and Downs, 1997) (Figur e 9). Upstream of this putative HMPPkinase homologue, to the end of the cloned re gion, are sequences th at have significant amino acid similarity to MrsA phosphoma nnomutase (Figure 9). Downstream of ZmCP4ccrM , to the end of the cloned region, are sequences that have significant amino acid similarity to Dihydropteroate synthase (DhpS) (Figure 9). The predicted amino acid sequence of putative DNA MTase ZmCP4CcrM from Z. mobilis was aligned with the amino acid se quence of the CcrM proteins of C. crescentus, B. abortus , R. meliloti, and A. tumefaciens (NCBI ENTREZ) to compare sequence similarity and alignment of the conserved catalytic and AdoMet binding domains found in DNA MTases (Malone et al. , 1995; Timinskas et al. , 1995; Jeltsch, 2002) (Figure 10). The putative primary target DNA recognition do main of DNA MTases is reported to be
137 located between the catalytic and Ad oMet binding domains (Timinskas et al. , 1995; Malone et al. , 1995; Jeltsch, 2002). Based on the arra ngement of the conserved domains (N-IV.V.VI.VII.VIII.TRD.X. I.II.III-C), CcrMI of C. crescentus belongs to the -subclass of N-MTases (Malone et al. , 1995; Reisenauer et al ., 1999b; Jeltsch, 2002; Gromova and Khoroshaev, 2003). The alignment (Figure 10) shows that the amino acid sequence of these homologues share identity beyond these conserved regions. Amino acid sequence analysis of ZmCP4CcrM also identified a putative N6-adenine-specific DNA MTase signature (Timinskas et al. , 1995) and the conserved sequence motif arrangement found in N-MTases (N6-adenine and N4-cyctosine MTases) (Malone et al. , 1995; Timinskas et al. , 1995). Comparison of Z. mobilis CP4 ORFs cloned in this study (Figure 9) with the corresponding ORFs of the recently released Z. mobilis ZM4 genome sequence (NCBI BLAST) shows that the put ative RNaseHII and CcrM homologues have identical predicted amino acid sequences (data not shown). The putative rnaseHII homologues have a single base change in the coding region and the ccrM homologues have five base changes. Each base change is located in the third base of an amino acid codon. The first 66 bp upstream of the putative rnaseHII homologues are identical except for one base change at the nucleotide immedi ately upstream of the putative rnaseHII translation initiation site. They have identical in tergenic DNA sequence between the putative rnaseHII and ccrM homologues. The first 93 bp downstream of the ccrM homologues are also identical. Comparison of all cloned ORFs in this genomic region of Z. mobilis CP4 (Figure 9) revealed that they are homologous in seque nce and gene arrangement to the corresponding ORFs found in Z. mobilis ZM4.
138 Primer Extension and Promoter Sequence Analysis of ZmCP4ccrM To determine the transcription initiation site for ZmCP4ccrM , Z. mobilis CP4 RNA was extracted and analyzed by primer extens ion and compared to the nucleotide sequence created using the same primer (Figure 11). Primer extension analysis identified the transcription initiation site of ZmCP4ccrM to be at a cytosine, located 27 bp upstream of the first (upstream) possible tr anslational initiatio n codon (Figure 11). In a study where Z. mobilis promoters of seven highly expressed ge nes were examined, it was reported that they all have a conserved 5â€™-AGGA-3â€™ seque nce in the RBS (ribosomal binding site) region (Ingram et al. , 1989). Examination of the ZmCP4ccrM promoter region revealed a 5â€™-AGGA-3â€™ sequence in the possible RBS im mediately upstream of the first possible translation initiation site, but not the second (Figure 11). Kanhere and Bansal (2005) hypothesized that the region of least DNA stability in a prokaryotic promoter region (re gions â€œprone to meltingâ€) coin cides with the -10 region. The -10 regions, where low stability tended to peak, can be predicted by calculating the average stability (average G) of a 15 nucleotide moving window along the promoter sequence (Kanhere and Bansal, 2005). A moving window of 15 nucleotides along the promoter region of ZmCP4ccrM was analyzed to determine the regions that have an average change in free energy ( G) with the smallest negativ e value (which corresponds to the lowest temperature of melting) usi ng a web-based program (NetPrimer) (Figure 11). A region of low stability with a G : Tm of -18.05:25.47 C was identified that corresponds to the -10 region ups tream of the transcription in itiation site determined by primer extension (Figure 11). Examination of the -10 region revealed that it is A/T rich and examination of the -35 region identif ied a 5â€™-ANNNNNCTNG-3â€™ sequence (Figure
139 11) conserved in the -35 re gions of highly expressed Z. mobilis genes (Ingram et al. , 1989). Stephens et al. (1995) reported the presence of regulatory elements in the 151 bp promoter region of ccrMI of Caulobacter crescentus including three IRs (inverted repeats) and four CcrM recogniti on sites. CcrMI expression in Caulobacter crescentus is believed to be autoregulated via the sec ond IR element (Figure 12) and possibly by methylation of CcrM recognition site s in its promoter region (Stephens et al. , 1995). CcrMI has also been shown to be a regulator of CtrA expression from its P1 promoter by methylation of CcrM recognition sites (Rei senauer and Shapiro, 20002) (see the Chapter 2 section on CcrM solitary MTase) . The 169 bp promoter region of ZmCP4ccrM was examined for similar potential regulatory elem ents. Two inverted repeats and two CcrM recognition sites were found (Figure 12) . IRs were not found in the promoter region of the ccrM homologues of Brucella abortus (NCBI ENTREZ) (Figure 12), Agrobacterium tumefaciens (NCBI ENTREZ) (data not shown), and Sinorhizobium meliloti (NCBI ENTREZ) (data not shown). One CcrM re cognition site was found in the promoter region of the ccrM homologue of S. meliloti (NCBI ENTREZ) (data not shown). In vivo ZmCP4CcrM activity: HinfI digestion of Z. mobilis CP4 DNA To determine if Z. mobilis CP4 methylates the same site (5â€™-GANTC-3â€™) in vivo that is methylated by CcrMI of C. crescentus (Stephens et al. , 1995), plasmid DNA (pLOI1844 and native plasmids) extracted from Z. mobilis CP4 was digested with HinfI (5â€™-GANTC-3â€™) and TfiI (5â€™-GAWTC-3â€™) a nd resolved on 0.8% agarose gels by electrophoresis. Z. mobilis CP4 has four native plasmids, approximately 31.5 bp, 32.5 bp, 33 bp, and 35 bp in length (Yablonsky et al., 1988). For comparative analysis, unmethylated pLOI1844 was also extracted from E. coli ER1647 and digested with HinfI
140 and TfiI (see Chapter 4, Figure 3). BstN I (5â€™-CCWGG-3â€™) digestion of DNA isolated from both Z. mobilis CP4 and E. coli ER1647 is not blocked by DNA methylation (see Chapter 4). Analysis of the gel lane cont aining BstNI restriction fragments of pLOI1844 isolated from E. coli (Figure 3, lane 5) compared to the BstNI fragments of plasmid DNA isolated from Z. mobilis CP4 (Figure 3, lane 2) indicate the presence of the pLOI1844 restriction fragments among the BstNI restriction fragments produced by digestion of plasmid DNA extracted from Z . mobilis CP4. Both HinfI and TfiI restriction is blocked by N6-andenine methylation of their recognition site (Roberts and Macelis, 1996). The pLOI1844 restriction fr agments produced from TfiI and HinfI digestions of pLOI1844 isolated from E. coli ER1647 (Figure 3, lanes 6 and 7, respectively) are absent in the lanes containing TfiI and HinfI di gestions of plasmid DNA extracted from Z. mobilis CP4 (Figure 3, lanes 3 and 4, respectiv ely). These results indicate that DNA isolated from Z. mobilis CP4 resisted TfiI and HinfI digestion due to in vivo DNA methylation at sites that overlaps Tf iI and HinfI recognition sites. Like C. crescentus (Reisenauer et al ., 1999b) , DNA in Z. mobilis CP4 appears to be methylated at 5â€™GANTC-3â€™ sites. In vivo Activity of Recombinant ZmCP4CcrM To determine if recombinant ZmCP4ccrM is expressed in E. coli, methylating DNA in vivo, the expression vector pPROTet.E1 33 alone or recombined with the ZmCP4ccrM gene was transformed into E. coli , BL21PRO (Figure 13). If recombinant ZmCP4CcrM is active in vivo , DNA extracted from the transformants will resist digestion by HinfI. The 3.3 kbp recombin ant plasmid, after 4 hours of induction at varying levels of aTc (0, 25, 50, 100, and 200 ng/ ml), was extracted and digested with HinfI. Each reaction contained 1 mg of s ubstrate DNA at 20 ng/l, 3 units of REase/g
141 DNA, and was incubated for 7 hours at 37C, to ensure complete digestion of unprotected sites. The rest riction fragments were reso lved in 0.8% agarose gel electrophoresis. Each lane in the top gel wa s loaded with approximately 85 ng of DNA, while each lane of the bottom gel was load ed with 400 ng DNA to visualize restriction products at low concentrations (except lanes 1, 2) (Figure 13). With increasing levels of aTc inducer, there was an incr easing level of resistance to digestion by HinfI, where 100 ng/ml (Lane 10) and 200 ng/ml (Lane 11) of aTc inducer appears to produce DNA completely protected from HinfI digestion, indicating that recomb inant ZmCP4CcrM is active in vivo (in E. coli ) and methylates HinfI recognition sites. There are seven HinfI sites in pPROTet.E133/ZmCP4C crM, producing 8 fragments when digested with BamHI and HinfI (1217, 774, 396, 365, 227, 117, 94, 91 bp). Inspection of the restriction fragment pattern in lanes 7-11 (Figure 13) suggests that some HinfI sites (in particular) may not be protected by recombinant ZmCP4C crM methylation as readily as others at lower levels of induction. To verify in vivo activity of recombinant ZmCP4CcrM from the pBBR1MCS/ ZmCP4ccrMPRO construct, the recombinant plasmid was transformed into E. coli GM4715. Plasmid DNA from transformants grown to log phase were extracted, linearized with SfiI, and assayed for resistance to HinfI digestion (Fi gure 14). Complete digestion of pBBR1MCS/ ZmCP4ccrMPRO with SfiI and HinfI would produce 11 fragments (ranging from 1884 bp to 10 bp). Th e plasmid DNA resisted HinfI digestion (Figure 14, lanes 3 and 4), indicat ing recombinant ZmCP4CcrM from pBBR1MCS/ ZmCP4ccrMPRO methylates HinfI recogniti on sites in this plasmid.
142 To verify in vivo activity of recombinant ZmCP4CcrM from the pET24b/ ZmCP4ccrM construct, it was transformed into E. coli strain HMS174(DE3). Transformants were grown to log phase and ZmCP4ccrM expression was induced 2 hours with 1 mM IPTG. The plasmid DNA was extracted and assayed for resistance to HinfI digestion (Figure 15, lane 4) and compared to ClaI linearized plas mid (Figure 15, lanes 2 and 3). Complete digestion of pET-24b/ ZmCP4ccrM (6356 bp) with HinfI would produce 19 fragments (ranging from 1194 bp to 6 bp). The plasmid DNA from the HMS174(DE3)/ ZmCP4ccrM transformants resisted HinfI digestion (Figure 15, lane 4), indicating recombinant Zm CP4CcrM from pET-24b/ ZmCP4ccrM methylates HinfI recognition sites. Most of the DNA appears undigested, with a minor band consistent in size with linearized reco mbinant plasmid (~6.5 kb). All th e major bands (three above and one below the position of linear ized recombinant plasmid) ar e consistent with undigested recombinant plasmid DNA (data not shown). Recombinant ZmCP4CcrM Pr otein Purification Expression of ZmCP4ccrM in B21Pro transformed with pPROTet.E133/ ZmCP4ccrM was examined. The highest leve l of expression of recombinant ZmCP4CcrM induced at 500 ng/ml aTc and pro duced a protein that migrated in the agarose electrophoresis gel as expected for the predicted 47.45 kDa C-terminal histidine tagged recombinant ZmCP4CcrM protein (Figur e 16). The level of protein expression was relatively low and attempts to purify the recombinant protein were not satisfactory. A second construct, pET-24b/ ZmCP4ccrM , transformed into HMS174(DE3) produced a much higher level of expression of recombinant ZmCP4ccrM when induced at 1 mM IPTG, than was achieved with pPROTet.E133/ ZmCP4ccrM (Figure 17). Purified CcrM maximally eluted at 300 mM imidazole from the HiTrap Chelating HP
143 column. SDS-PAGE analysis revealed a 43-44 kDa protein consistent with the predicted molecular mass of the 43.74 kDa C-terminal histidine tagged recombinant ZmCP4CcrM protein encoded by the plasmid pET-24b/ ZmCP4ccrM in HMS174(DE3) (Figure 17). The purified recombinant protein (Figure 17; ge l A-II lane 11, and gel B-II lane 9) was determined to be >95% pure, based on SD S-PAGE and coomassie staining (Laemmli, 1970). Batch fractionation using anion exch ange (Q sepharose and DEAE sepharose; Amersham-Pharmacia, Corp.) and/or Sephade x G-75 in conjunction with metal affinity chromatography [TALON affinity resin (CL ONETECH Laboratories, Inc.) and His-Trap Kit] were tested in order to obtain purified recombinant CcrM protein (data not shown). Crude cell extracts that were batch fractiona ted with a Sephadex G-75 column prior to affinity purification with a Ni2+ charged HiTrap Chelating Sepharose HP column produced the purest and greates t amount of recombinant CcrM protein extracts (Figure 17). In vitro Activity of Recombinant ZmCP4CcrM The methylation of DNA by partially purif ied recombinant protein, extracted from cultured BL21PRO/pPROTet.E133/ ZmCP4ccrM, was assayed for its relative ability to protect substrate DNA (pP3226) from HinfI digestion under various buffer conditions in vitro (Figure 18). The reaction in buffer 3 had the highest final NaCl concentration at 118 mM and showed no evidence of methylati on protection from Hinf I digestion (Figure 18, lane 4); while the reaction buffer , with the lowest final Na Cl concentration at 18 mM, had the highest level of methylation prot ection from HinfI digestion (Figure 18, lane 7). Comparing activity of ZmCP4CcrM in buffer (150 mM potassium acetate) to buffer 4 (50 mM potassium acetate) indicates that increased concentration of potassium acetate also has a negative effect on ZmCP4Ccr M activity (Figure 18, lanes 6 and 5
144 respectively). Of the buffers tested, ZmCP 4CcrM activity was highest in the potassium phosphate buffer (Figure 18, lane 7). The partially purified recombinant ZmCP4ccrM protein is active in vitro , as shown by the resistance of the substrate DNA to HinfI cleavage after incubation with the extracted protein. The activity of CcrM is salt dependent, as demonstrated by increased CcrM protection from HinfI digestion as the total salt concentration decreases. ZmCP4CcrM activity was also examined fo r purified recombinant protein extracted from cultured and IPTG induced HMS174(DE3)/pET-24b/ ZmCP4ccrM cells (Figure 19). If there was no in vitro methylation protection from HinfI digestion of the 2708 bp substrate DNA (pNEB193 PacinsertBsaI ), cleavage at th e HinfI sites would have been produced 7 restriction fragments ( 1432, 517, 396, 194, 75, 65, and 29 bp). While a fraction of the sample was completely protected from HinfI digestion by in vitro CcrM methylation, the majority of the resolved fragments consist of two bands approximately 2kb and 1.6 kb in size. It appears that some sites in the plasmid may be unmethylated or selectively undermethylated. Determination of the Nucleoti de Methylated by ZmCP4CcrM To determine whether the recombinant CcrM protein methylates the adenine or the cytosine of its recognitio n site, substrate DNA (pNEB193 PacI/BsaI) was engineered, containing a BsaI restriction si te whose cleavage site overla ps a CcrM methylation site (Figure 20), as described in the Materials and Methods section. The substrate DNA was methylated by recombinant ZmCP4CcrM in vitro, using [3H]-AdoMet as the methyl donor, and digested with PvuII or PvuII/BsaI. The restriction fragments were isolated and analyzed by scintillation counting for [3H]-labeling of each DNA fragment. If the cytosines of the CcrM recognition site we re methylated by recombinant ZmCP4CcrM,
145 both the 125 bp and the 219 bp DNA re striction fragments would be [3H]-labeled. If the adenines of the CcrM recognition site we re methylated by recombinant ZmCP4CcrM, only the 219 bp restrictio n fragment would be [3H]-labeled (Figure 20). The CPMs from experimental set 1 shows that CPM of the 125 bp fragment was not significantly different from background, while the CPM for the 219 bp fragment was [3H]-labeled (Table 5). The CPMs from e xperimental set 2 shows that the 125 bp fragmentsâ€™ CPM was sligh tly above background, however , the 219 bp fragmentsâ€™ CPM was ~34 fold higher (~97% of the total CPM). The [3H]-labeling pattern of the BsaI restriction fragments indicate that ZmCP4C crM specifically recognizes and methylates the adenine of its recognition site 5â€™-GANTC3â€™ (Table 5). The experiment was done with the presence of both 6 M labeled and 20 M unlabeled AdoMet to ensure sufficient amounts of AdoMet for complete met hylation of the substr ate. The relatively low level of total DNA labeli ng is due to the specific activity of the AdoMet. Effect on Growth and Cell Viability of ZmCP4ccrM Overexpression in Z. mobilis CP4 To determine if the presence of plasmid borne ZmCP4ccrM has any effect on growth, the growth rate of wild-type Z. mobilis cells or transformants with pLOI1844 (helper plasmid), or pLOI1844/ pBROriV Rop or pLOI1844/pBROriV Rop/ ZmCP4ccrM was determined in GYx media (includi ng appropriate antibiotics). Cultures of transformants containing plasmid borne ZmCP4ccrM were started with exponentially growing cells to minimizing lag. The cells were grown at 30 C with slow shaking and minimal aeration to maintain cells in suspension. The optical density at 600 was monitored every hour for eight hours using a spectrophotometer (Figure 21). The generation time (GT) of wild-type Z. mobilis CP4 was ~1 hour 50 minutes, which
146 increased slightly to ~2 hours when transformed with pLOI1844. Z. mobilis CP4 containing pLOI1844 and pBROriV Rop with or without ZmCP4ccrM had a similar GT of ~2 hours 23 minutes. Growth rates show no significant difference for cultured cells with or without having plasmid borne copies of ZmCP4ccrM . The increase in generation time from wild type Z. mobilis CP4 to cells containing plasmi ds appears to be due to an increase in plasmid load. To determine the effect of extra copies of ccrM in Z. mobilis CP4 in stationary phase cells, three day Z. mobilis cultures, with or without plasmid borne copies of ZmCP4ccrM, were examined for differences in cu ltural cell density and recovery of viable cells (Table 6). The maximal cultural cell density was reduced for Z. mobilis /pLOI1844 transformed with pBROriV Rop ZmCP4ccrM while the maximal cell density of Z. mobilis transformed with pBBR1MCS/ ccrMPRO was not significantly different from wild-type Z. mobilis or cells transformed with vector DNA alone. The greater growth inhibition ma y indicate a higher level of ccrM expression from the plasmid Rop promoter as compared to expres sion from its native promoter. The CFU counts of viable cells compared to the estimated number of ce lls that were plated indicate that almost all the cells (>99.9%) in both cu ltures were no longer viable after 3 days (Table 6), indicating that overexpression of ccrM resulted in decrease of cell viability. Effect of Overexpression of Zmcp4ccrm in Zymomonas mobilis CP4 on Cell Morphology Z. mobilis CP4 carrying the broad host range vector pLOI1844 was transformed with pBROriV Rop or pBROriV Rop/ ZmCP4ccrM , cultured to an OD550=0.5, and examined using Nomarski optics micr oscopy at 200X (Figure 22) and 400X magnification (Figure 23). Cultures of Z. mobilis cells with pLOI1844 and transformed
147 with pBROriV Rop/ ZmCP4ccrM clearly contain of a subpopulat ion of cells with grossly elongated aberrant morphology that are not seen in transfor mants containing the plasmid without the ZmCP4ccrM insert, nor in wild-type cells. Z. mobilis CP4 strain cells are ch aracterized as Gram negative rods 2 to 6 m in length and 1-2 m in width, with cells occurring as singles or in pairs, and rarely forming chains of a few cells (Swings and De Ley, 1977). The cells found as singlets in the control culture with vector alone average a bout 3 m in length and 2 m in width and appear to make up at least half the populat ion (Figure 22-A). Wh ile doublet length cells were commonly observed in the control cultu re (most with an obvious septum forming pairs), it was very rare that ch ains of three or four cells we re seen. No filamentous cells were seen. Examination of the Z. mobilis transformants overexpressing ZmCP4ccrM revealed that the cell arrangement a ppeared fairly unchanged where cells were found as singlets or pairs; however, the morphology of a s ubpopulation of the cells changed drastically (Figure 22-B and Figure 23). Measurement of the dimensions of the cells indicated that these cells varied greatly in length and diam eter. The cells ranged between 3 to 90 m in length and 1.5-7 m in width. To compare the number of grossly elonga ted cells (>10 m in length) to the number of singlets and doublet s, a Petroff-Hausser counting chamber was used to count the pBROriV Rop/ ZmCP4ccrM transformed Z. mobilis CP4 cells observed under Bright-field microscopy. Approximately 2-5% of a cell population in mid-log phase (6 x 108) was made up of these extremely elongate d cells, calculated based on the number of individual cells found in each block of the Petroff-Hausser counting chamber. When the
148 number of grossly elongated cells and their average sizes are considered, compared to single cells and doublets, they represent a pproximately 10-30% of the populationâ€™s cell mass. The cells of singlet length are less common, than in the control cells, and the majority of the cells were the length of a normal doublet or larger, with a great many not having obvious septum formation (Figure 22-B) . The Petroff-Hausser counting chamber was also used to count the pBROriV Rop transformed Z. mobilis CP4 cells observed under Bright-field microscopy. A mid-log pha se control culture th at contained ~7.08 x 108 cells/ml contained no elongated cells. Z. mobilis CP4 with pLOI1844 transformed with pBROriV Rop/CcrM was stained with DAPI and examined using microscopy under UV to visualize DNA distribution (Figure 24). The cells w ith obvious changes in cell morphology exhibited varying patterns of DNA distribution. Some of the elongated filamentous cells that showed points of the possible beginnings of septum formation had correspond ing localized points of greater DNA density (Figure 24-A), indica ting the presence of multiple chromosomes. Other filamentous cells had very few if any possible points of septum formation, yet still showed distinctly separate points of great er DNA density (Figure 24-B), indicating the presence of multiple chromosomes. Some filamentous cells had no visually evident septum formations and DNA stain was fairly uniform along the length of the cell (Figure 24-C). It appears that cell division was disrupted, while chromosomal replication continued, producing elongated filamentous cells with multiple copies of the chromosome. Discussion In this study, DNA sequence analysis a nd alignment of th ree positive clones isolated from the indicator strain AP1-200-9 (Piekarowicz et al. , 1991b) identified an
149 ORF with amino acid sequence similarity to DNA MTases. The best match was to a solitary adenine DNA methyltransferase M.BabI of Brucella abortus (NCBI BLAST), a homologue of the solitary cell cycle regul ating DNA MTase CcrM, first identified in Caulobacter crescentus (Zweiger et al. , 1994). DNA sequence and NCBI BLAST analysis of the region up stream of the putative ccrMlike DNA MTase ORF indicate that the adjacent genes are not REases. A cl one, pZM46, isolated by Lai, Xiaokuang and Ingram, L.O. (unpublished) was identified that overlapped the clones isolated in this study containing the putative ccrMlike MTase ORF. Sequen ce analysis revealed an ORF, immediately downstream of the putative ccrMlike MTase, which appears to code for a dihydropteroate synt hase/dihydropteroate pyrophosphorylase (DhpS) (Lai, Xiaokuang and Ingram, L.O., unpublished; NCBI BLAST). During folate biosynthesis, DhpS is involved in the first and second step in synthesizing dihydrofolate (Williams et al. , 2000). The ORF downstream of the putative DhpS gene codes for a conserved hypothetical protein of unknown function (NCBI BLAST). REases and DNA MTases of R-M systems are typically linked (Wilson, 1991). Because sequence analysis of the DNA immediately flanking the putative ccrM -like ORF had significant amino acid sequence similarity to known proteins and ar e not likely to be ORFs coding for REases, the putative ccrM -like ORF was considered to likely code for a solitary DNA MTase. This open reading frame was designated ZMCP4ccrM . Sequence analysis of ZMCP4ccrM from Z. mobilis CP4 identified two possible inframe translation initiation sites. ZmCP 4CcrM has a predicted 382 or 373 amino acid sequence with an average mass estimated to be 43.55 kDa or 42.49 kDa, depending on which of two possible in-frame ATG translation initiation sites it uses (Figure 8).
150 Examination of the ZmCP4ccrM promoter region revealed two appropriately located possible RBS. A 5â€™-AGGA-3â€™ sequence, wh ich is conserved in the RBS region of several highly expressed Z. mobilis genes (Ingram et al. , 1989) was found upstream of the first possible translation initiation site but not the s econd, suggesting that ZmCP4ccrM is likely being expressed from the first translational initiation site in Z. mobilis CP4 (Figure 11). In contrast, compara tive amino acid sequence analysis revealed that a protein produced from the first possible translation initiation site of ZmCP4ccrM would be longer than th e CcrM homologues from C. crescentus, B. abortus, S. meliloti , and A. tumefaciens , while the second possible translati on initiation site exactly coincides with the relative position of the initiation codons of the ccrM homologues of B. abortus and S. meliloti (Figure 10). The constructs created in this study to obtain purified recombinant proteins (pPROTet.E133/ ZmCP4ccrM and pET-24b/ ZmCP4ccrM ) were designed to produce ZmCP4CcrM from the seco nd translation initiation site, and were shown to be active in E. coli and in vitro . Likewise, the construc t created to overexpress ZmCP4ccrM in Z. mobilis CP4 from a heterologous promoter (pBROriV Rop/ ZmCP4ccrM ) was designed to produce ZmCP4C crM from the second translation initiation site, and was shown to be active in vivo in both E. coli and Z. mobilis CP4. Further studies must be done to directly determine if ZmCP4ccrM is being expressed in Z. mobilis CP4 from one or both possible translation initiation sites. The recognition site of CcrMI of C. crescentus was originally identified by its sequence similarity to M.HinfI as the enzy me responsible for methylating the HinfI recognition sites in the C. crescentus genome (Reisenauer et al ., 1999b). To determine if Z. mobilis methylates its DNA in vivo at HinfI sites like CcrMI of C. crescentus , DNA
151 was extracted and assayed for resistance to HinfI digestion. The results showed that DNA isolated from Z. mobilis Cp4 resists digestion by both HinfI (5â€™-GANTC-3â€™) and TfiI (5â€™-GA(A/T)TC-3â€™) (Chapt er 4, Figure 3). Because th e DNA resisted digestion by both HinfI and TfiI (seemingly e qually as well), it appears that Z. mobilis CP4 methylates DNA at sites overlapping HinfI sites. Subsequent in vivo and in vitro assays of ZmCP4CcrM methylation confirmed that Zm CP4CcrM methylates sites that overlap HinfI recognition sites. To directly determine if ZmCP4 specifically methylates the adenine of 3â€™-GANTC-5â€™ CcrM sites like M.HinfI, the nucleo tide specificity of ZmCP4CcrM methylation was anal yzed by creating a construct, pNEB193 PacI/BsaI, with a BsaI site positioned so that its cleavage site is within the adjacent CcrM site. BsaI cleaves the CcrM site so that both adenosines are in the 219 bp fragment and the two cytosines are separated into 125 bp and 219 bp fr agments (Figure 20 and Table 5). BsaI digestion of CcrM [3H]-methyl-labeled substrate DNA produced fragments in which the [3H]-label was detected in the 219 bp fragme nts containing both adenine and cytosines targets, while there was little or no [3H]-labeling detected in the 125 bp fragments containing only one cytosine of the CcrM site. Since the fragment containing only a cytosine of the CcrM site remained relati vely unlabeled, it indi cates that ZmCP4CcrM specifically recognizes and me thylates the adenine of its recognition site 5â€™-GANTC-3â€™ (Table 5). The results of this experiment, supported by in vivo and in vitro assays of recombinant ZmCp4CcrM activit y, indicate that ZmCP4CcrM is a N6-adenine specific DNA methyltransferase that recognizes and methylates 5â€™-GANTC-3â€™ sites. The restriction pattern produced in the ZmCP4CcrM methylation protection assays ( in vivo and in vitro ) (Figures 13, 14, 18, and 19) were examined and compared to the
152 expected restriction fragments produced fr om complete and partial digestion, to determine which CcrM recognition sites a ppear to be selectively undermethylated. Despite the ability to obtain fu lly methylated plasmid DNA from E. coli transformants expressing recombinant ZmCP4ccrM (when a sufficient amount of inducer is present) (Figure 13, Lane 10 and 11), the discrete re striction fragments produced from HinfI digestion of plasmid DNA isolated from transformants with lower ccrM expression levels indicate that ZmCP4CcrM has a site specificity preference in vivo that is stricter than recognition of all 5â€™-GANTC-3â€™ sites. Based on analysis of the Hinf I restriction patterns of partially methylated DNA, it appear s that 5â€™-GA(A/T)TC-3â€™ sites may be preferentially/more readily methylated that 5â€™-GA(G/C)TC-3â€™ sites. DNA fragments that would produce the restriction pa ttern in Figure 13 (lanes 8 a nd 9) are consistent with cleavage of 5â€™-GA(G/C)TC-3â€™ sites alone. Ho wever, the restriction pattern produced in Figures 14, 18, and 19 are more complex and can not be explain simply by N nucleotide specificity of 5â€™-GANTC-3â€™ si tes. CcrM may be preferen tially methylating certain 5â€™GANTC-3â€™ sites in context with specific fl anking sequences. Direct evidence for any preference for ZmCP4CcrM to methylate 5â€™GA(A/T)TC-3â€™ sites, or the influence of flanking sequence on CcrM methylation of 5â€™-GANT C-3â€™ sites is still to be determined. Analysis of the Z. mobilis ZM4 genome (Seo et al. , 2005) revealed that it has 66% of the statistically expected number of 5â€™-GANTC-3â€™ sites. Analysis of C. crescentus genome revealed that it has 37% of the statistically expected 5â€™-GANTC-3â€™ sites (Reisenauer and Shapiro, 2002). In both genomes, 5â€™-GA(G/C) TC-3â€™ sites are the most statistically underrepresented. It has been reported that about of CcrM recognition sites have been found to be associated with c onserved intergenic motifs in C. crescentus , and it has been
153 suggested that they may possibly be involved in regulation of the associated genes or maintenance (Chen and Shapiro, 2003). Examin ation of the intergenic CcrM sites that were identified in C. crescentus and B. abortus by Chen and Shapiro, (2003) revealed that 5â€™-GA(A/T)TC-3â€™ was the predominate form found in these intergenic motifs. This suggests that a preferential bias to met hylate 5â€™-GA(A/T)TC-3â€™ sites may perhaps be advantageous, optimizing the functional role th ese motifs may have. Further studies must be done to determine if Z. mobilis has similar intergenic motif containing CcrM methylation sites. Overexpression of ZmCP4ccrM in Z. mobilis CP4 appeared to have little effect on the initial growth rate (Figur e 21); however, analysis of cu ltured cell viability revealed that the cell density of cultures in st ationary phase expressing recombinant ZmCP4ccrM from a heterologous promoter was reduced compared to expression from its native promoter (Table 6). This suggests that expression of ZmCp4ccrM from its native promoter may be repressed to some degree. After 3 days, culture OD600 may reflect growth, cell death, and cell lysis, and requires repeat analysis. Because the promoter of ZmCP4ccrM has a putative RBS, a -10 region, and a -35 region which have sequences conserved in highly expressed Z. mobilis proteins (Ingram et al. , 1989), it seems less likely that maximal expression of ZmCP4ccrM is lower from its native promoter than from a Rop promoter within a Z. mobilis cell. Recovery of vi able cells from 3 day old cultures was poor in cell containing plasmid borne ZmCP4ccrM , regardless of the promoter it used, indicati ng that overexpression of ZmCP4ccrM decreases cell viability (Table 6). It was estimated that more than 99.9% of the cells in both cultures were no longer viable after 3 days. Future studi es should include monitoring the effect
154 overexpression of ZmCP4ccrM in Z. mobilis CP4 has on exponential growth through to stationary phase. Overexpression of CcrM in the alpha-proteobacteria C. crescentus (Zweiger et al ., 1994), R. meliloti (Wright et al ., 1997), B. abortus (Robertson et al ., 2000), and A tumefaciens (Kahng and Shapiro, 2001) were show n to cause aberrant cell division, increased DNA content, and a subpopulation of cells with aberrant morphology such as thickening, branching, and some degree of elongation. Microscopic examination of Z. mobilis CP4 transformed with pBROriV Rop/ ZmCP4ccrM revealed a subpopulation of cells with aberrant morphology that were not seen in the control cells (Figure 22). Approximately 2-5% of the cell pop ulation in mid-log phase (6 x 108) was made up of these extremely elongated filamentous cells ( 10-90 m). When their average sizes were considered, it was estimated that they w ould represent approximately 10-30% of the populationâ€™s cell mass. The average size of the remaining cells in the population was larger than those found in th e control. Thus, as in C. crescentus and the other alphaproteobacteria that have b een studied, overexpression of ZmCP4ccrM has a detrimental effect on normal cell morphology and cell division. The DNA distribution in Z. mobilis CP4 cells overexpressing ZmCP4ccrM was revealed by DAPI stain (Figure 24). While visual distribution of DNA in the cells varied, the elongated cells clearly had multiple chromosomes, indicating that overexpression of ZmCP4ccrM causes disruption of normal cell division. Since many of the filamentous cells examined did not show distinct point s in DNA density, chromosomal segregation may also be disrupted. Collectively, the re sults indicate that overe xpression of CcrM in Z. mobilis CP4 disrupts the normal pattern of DNA replication, DNA segregation, cell
155 division, and viability, suggesting that ZmCP 4CcrM is involved in cell cycle regulation, as described for CcrM homologues found in other alpha-proteobacter ia (Marczynski and Shapiro, 2002). CcrMI of C. crescentus was proposed to be autore gulated via the second IR element found in the promoter region of ccrMI (activation) and by me thylation of the two CcrM recognition sites (rep ression) that make-up the IR3 found between the transcriptional and translational initiation sites (Stephens et al. , 1995) (Figure 12). Similarly, DNA sequence analysis of ZmCP4ccrMâ€™s promoter region revealed the presence of two inverted repeat s and two CcrM recognition sites . The second IR element is centered on the transcri ption initiation site of ZmCP4ccrM (Figure 12), which implicates a possible mechanis m of regulating expression of ZmCP4ccrM at the transcriptional level. The similarity of sequences present in the promoter region of ccrMI from C. crescentus and ZmCP4ccrM suggests that ZmCP4ccrM may also be autoregulated in a manner similar to the tr anscriptional re gulation mechanisms proposed to regulate ccrMI expression in C. crescentus . Although analysis of the B. abortus , A. tumefaciens , and S. meliloti genomes revealed conserve d 15-mers centered around CcrM sites in intergenic regions and the presence of long intergenic motifs in B. abortus containing inverted repeats and CcrM sites like those found in C. crescentus (Chen and Shapiro, 2003) (see Chapter 2 on CcrM solitary MTase), analysis of the promoter region of the ccrM homologues of B. abortus (NCBI ENTREZ) (Figure 12) and A. tumefaciens (NCBI ENTREZ) reveal that they do not ha ve CcrM recognition sites or IRs in the promoter region of their CcrM genes. Sequence analysis revealed that S. meliloti (NCBI ENTREZ) also has no IRs but does have one CcrM site located approximately in the
156 center of the intergenic region, between the tr anslation initiation site s of an ORF of an unknown hypothetical protein and ccrM . Reviewing studies on cel l cycle regulation in C. crescentus , and considering the phenotypic effect CcrM has in seve ral alpha-proteobacteria, including Z. mobilis CP4, suggests that CcrMâ€™s proposed involvement in cell cycle regulation may result from CcrM regulation of gene expression, as shown for ctrA and proposed for several other genes. In C. crescentus , the CcrM methylation sites have been found in the promoters of several genes, including ccrMI, ctrA, ftsZ , and the class II flagellar genes fliQ and fliL (Reisenauer et al., 1999b). In addition to the proposed autoregulation of ccrM , repressing its own transcription (Stephens et al. , 1995), CcrM has been shown to represses ctrA transcription from its P1 promoter in C. crescentus (Reisenauer and Shapiro, 2002). CtrA is a global regulator, part of a two com ponent phosphorelay signal transduction system, involved in temporal control of se veral cell cycle events (Reisenauer et al., 1999a; Brun, 2001). CtrA was determined (by microarray analysis) to be involved directly or indirectly in regulating the expression of 26% of the cell cycle regulated genes in C. crescentus (Laub et al. , 2002). The phosphorylated form of CtrA appears to directly regulate at least 95 genes (most identified by CtrA binding assays and/or expression assays) such as the DNA MTase ccrMI , the cell division initiator protein ftsZ , and class II flagellar proteins (Laub et al. , 2002). In C. crescentus, CtrA activates the transcription of ccrMI (Reisenauer et al. , 1999a) (see Chapter 2, Figure 1) . Morphological changes have been observed in C. crescentus under certain growth conditions, or when the expression of certain genes was altered. Ob servation of phosphate starved C. crescentus cells reveal that they all become extremel y elongated filamentous cells (Quon et al., 1996). C.
157 crescentus also becomes filamentous when the Ct rA regulated class II flagellar genes fliQ, fliR (Zhuang and Shapiro, 1995), and fliX (Mohr et al. , 1998) are mutated. These observations, in conjunction with the morphological effect of ccrM overexpression in Z. mobilis CP4, C. crescentus , and other several alpha-prote obacteria, suggest that CcrM may be involved in regulating the ce ll cycle due to its regulation of ctrA expression (Reisenauer and Shapiro, 2002), and the pr oposed involvement in regulating gene expression of other genes require d for normal cell cycle progression. Alternatively, because CcrM sites (5â€™-GANT C-3â€™) are found in higher frequency at the origin of replication ( Cori ) of C. crescentus (6 sites with 5 clustered in a 418 bp region downstream of hemE ) (GenBank accession number U13664; NCBI ENTREZ), it has been suggested that CcrM may play a ro le in regulating initia tion of DNA replication (Reisenauer et al., 1999b), as shown for the E. coli solitary DNA MTase Dam (5â€™-GATC3â€™) (Boye and Lobner-Olesen, 1990; Ca mpbell and Kleckner, 1990). The hemE /origin of replication/RP001 (first ORF) genetic organization had been identified in R. prowazekii and C. crescentus genomes and predicted to be characteristic of other alphaproteobacteria (Brassinga et al., 2001). Sequence analysis of a 1902 bp region of Z. mobilis ZM4 (NCBI ENTREZ), starting at sequence position 2054661, overlapping the putative hemE, and containing the putative origin of replication, revealed nine 5â€™GANTC-3â€™ sites with eight cluste red downstream of the putative hemE ORF. Perhaps Z. mobilis DNA replication is also regulated by CcrM methylation as proposed for C. crescentus . The recently characterized EcoKCcrM of E. coli K12 (5â€™-ATGCAT-3â€™) (Kossykh and Lloyd, 2004) has also been proposed as having an effect on DNA replication. DNA sequence analysis of the region (468 bp) containing E. coli K12â€™s
158 origin of replication oriC (GenBank accession number X02820; NCBI ENTREZ) revealed that there are no EcoKCcrM r ecognition sites and 17 Dam recognition sites, suggesting that DNA replica tion is not directly regul ated by methylation of oriC by EcoKCcrM. CtrA is a global regulator involve d in temporal control of several cell cycle events, including repression of DNA replication in C crescentus (Quon et al., 1998) (see Chapter 2, Figure 1). Overexpression of CtrA blocks DNA replication (Quon et al ., 1998). Because ctrA expression has been shown to be regulated by CcrM methylation in C. crescentus (Reisenauer and Shapiro, 2002), and CcrM expression is regulated by CtrA (Quon et al ., 1996; Laub et al. , 2002), suggests that CcrM is also indirectly involved in regulating initiation of DNA replicat ion by its direct regulation of ctrA expression. Putative CcrM DNA MTase homologues are wi dely found in alpha -proteobacteria (Wright et al. , 1997), and gamma-proteobacteria (with a different sequence specificity for M.EcoKCcrM) (Kossykh and Lloyd, 2004), sugg esting that these ce ll cycle regulatory DNA MTases may have originated from a common ancient ancestor. The gene found upstream of the ccrM homologues found in both C. crescentus (NCBI ENTREZ) and Z. mobilis is the putative RNase HII, a highly cons erved gene found in all kingdoms of life (Rydberg and Game, 2002). In contrast, the RNase HII of A. tumefaciens, B. abortus, and S. meliloti are found elsewhere in their chromosomes (NCBI ENTREZ). The similarity in gene arrangement (RNase HII-CcrM) and putative regulatory elements found in the promoter regions of the ccrM genes of both Zymomonas and Caulobacter suggests a possible common evol utionary origin of their ccrM genes. Alphaproteobacteria are diverse in the ecological niches in which they are found such as: the free-living water bacterium C. crescentus (Poindexter, 1981), th e nitrogen-fixing plant
159 symbionts M. loti (Wright et al ., 1997) and S. meliloti (Bellefontaine et al., 2002), the plant pathogen A. tumefaciens (Kahng and Shapiro, 2001), the facultative animal pathogen B. abortus (Robertson et al ., 2000), and the obligate animal pathogen R. prowazekii , which does not have a CcrM homol ogue (Chen and Shapiro, 2003), and is considered the closest living relative to eukaryotic mitochondria (Brassinga et al ., 2001). Nothing is known about cell cycle regulation in Z. mobilis, or its ecological relationship to other organisms in nature (other than being found in plan t sap), or its evolutionary relationship to other bacteria (being the only member species of its genera) (Swings and De Ley, 1977). The evolutionary relationship of several alpha-prote obacteria has been analyzed by Hallez et al . (2004) centered on the presen ce of the global cell cycle response regulator CtrA (unique to alpha-prote obacteria) and associat ed proteins of the cell cycle regulatory network, as modeled in C. crescentus . Genomic analysis of six alpha-proteobacteria (each bel onging to a different genera) identified the homologues of several genes known to be involved in CtrA controlled cell cycle regulatory networks which, in conjunction with elec tron microscopic examination of four of these organisms, was the basis of the propositi on that, in general, all al pha-proteobacteria may have morphological and functional asymmetry (a char acteristic most obvious in the dimorphic organism C. crescentus ) (Hallez et al. , 2004). With recent release of the Z. mobilis ZM4 genome, future investigati on of the ZmCP4CcrM and its involvement in cell cycle regulation should includ e analysis of the Z. mobilis genome to identify other putative components of a cell cycl e regulatory network.
160 Table 4. Bacterial strains, plasmids , and primers used in ZmCP4CcrM study Strain, Plasmid, and Primers Genotype or Description Reference or source E. coli Strain HMS174(DE3) FrecA1 hsdR(rk12 mk12) (DE3) (Rifr) Novagen, Inc. DH5 Fâ€™/ endA1 hsdR17 ( rk mk) supE44 thi-1 recA1 gyrA (Nalr) relA1 ( lacIZYA-argF ) U169 deoR ( 80 dlac ( lacZ ) M15 ) New England Biolabs, Inc. HB101 F( gpt-proA )62 leuB6 supE44 ara-14 galK2 laY1 ( mcrC-mrr ) rpsL 20(Strr) xyl-5 mtl1 recA 13 New England Biolabs, Inc Rosetta(DE3) (Cmr) (DE3) derivative of Turner (B) BL21: lacYZ lon ompT Novagen, Inc. Plasmid pBBR1MCS (Cmr). Broad host range vector Kovach et al. , 1995 pET-24b(+) (Kanr) 5310 bp expression vector carries Nterminal T7 Tag sequence; C-terminal His-Tag sequence; T7 promoter, transcription initiation site, and terminator. F1 origin. LacI .Multiple cloning site. ColE1 compa tibility group Ori origin of replication. Novagen, Inc. pNEB PacI/ BsaI (Ampr) pNEB193 derivative with PacI site deleted and BsaI inserted overlapping a HinfI site This study pBROriV Rop (Tetr) pBR322 derivative with Ampr deleted, a RSF1010 OriV insert, and delete ROP. 3959 bp This study Primers GATGGTACCCATGAGAA GAAGACATCGG Forward Primer CcrM#3 (To clone ZmCP4crrM into KpnI site of pROTet.E133 TCCAGGTACCCAAGGATTCTCGGCAAGAT AG Reverse Primer This study CGGCGATTAATGAGTGAAGAAGACATC Forward Primer CcrM#4 (To clone ZmCP4ccrM into the AseI/PvuII site of pBROriV Rop) GATCAGCTGTCAAGGATTCTCGGC Reverse Primer This study GTGAGCTGCATATGAGTGAAGAAGACATC GG Forward Primer CcrM#6 (To clone ZmCP4ccrM into the NdeI/XhoI site of pET-24b) AAACTCGAGCCAAGGATTCTCGGCAAGAT AG Reverse Primer This study
161 Table 4. Continued. Strain, Plasmid, and Primers Genotype or Description Reference or source TTCTCGAGACATCGTGAAGCCATTCAA Forward Primer CcrMPRO#2 (To clone ZmCP4ccrM with native promoter into XhoI/SpeI site of pBBR1MCS) GTACTAGTGAAGCGCCATCGGTGC Reverse Primer This study ATGACTCCCCGTCCACGCCAGAAGGATGA G AhdI Forward Primer ORIV (To create pBROriV) GCGACGTCGGTCGGTGGCTCTGGTAAC AatII Reverse Primer This study TTACCGCAGCTGCCTCGC PvuII Forward Primer PBRORIV (To create pBROriV Rop) GGTCATTAATGCCTCCGTGTAAG AseI Reverse Primer This study Primer Extension AACGGTAAGTCGGGCTGTCAGAT CcrMEXT3 This study See Chapter 4, Table 1, for the description of additional bacterial strains, plasmids, and primers.
162 Figure 7. pBROriV Rop construct. A . The pBROriV construct derived from pP3222 with the RSF1010 broad host range vector OriV origin of re plication sequence inserted between the AhdI and AatII site s (highlighted in blue). The genes highlighted in red code for Tet (tet racycline resistance, 89-1276 bp), Ori origin of replication, 3122-2534), and Rop (1915-2106). The Rop RBS (ribosomal binding site, 1905-1909) is marked in blue. B . The pBROriV Rop construct is derived from pB ROriV with deletion of the Rop gene and insertion of an AseI cloning site. pBROriV4096 bps 1000 2000 3000 4000 PvuII AhdI AatII Tet Rop Ori 3959 bps 500 1000 1500 2000 2500 3000 3500 AseI PvuII AhdI AatII Tet Ori pBROriV Rop Rop RBS Rop RBS A . B . OriV OriV
163 1 gggctctgttggttctcccgca 23 acgtactctgtttaccaggtcaggtccggaagaagcagccaaggc 68 agatgacgcgtgtgccgggatgtagctggcagggcctgttatggc 113 cggagccgtttatcttcaccgtgagcatattcccgaagagtaaat 158 gattcaaaaaaactgactgcgcgccgtcgccatcttttatccgac 203 atg ctgcataatcaggctgattatgcaacgggaatggcgaatgtt RNase HII M L H N Q A D Y A T G M A N V 248 catgagattgaccggatcaatatccggcaagccagtcttttggcg H E I D R I N I R Q A S L L A 293 atgaagcgggcagtagaggccttgatacaaaaaataggtagggaa M K R A V E A L I Q K I G R E 338 cctgattgtattttagtggatggccgcgatattcctgattggccg P D C I L V D G R D I P D W P 383 tggccatcattgcctattatcaaaggggatagtctttctttgtcg W P S L P I I K G D S L S L S 428 attgcggcggcctctattgtagccaaagtcgaacgggatgaaata I A A A S I V A K V E R D E I 473 atggtaaaagccagccaagaatatcccggctatggttgggaacac M V K A S Q E Y P G Y G W E H 518 aatatgggatatccgacaaaggaacatcgtgaagccattcaaaga N M G Y P T K E H R E A I Q R 563 ttaaagccaacaaaattccatcgccggagtttttcacctattcgc L K P T K F H R R S F S P I R 608 caattttatgaaaatgtagat tag gatgtctttgacctctttttc Q F Y E N V D * 653 gttactttttcta 666 cattttttgaagctataaaataatcacttcaatcattatccataa 711 ccgcttttgtatttttacaatgaagcctcttgaccgaagagtcgc 756 attgaatcacAGTAAAGTCGGGTTTCTGTGAAAGGATCAGTATTT 801 ATG CGCCTTGCCGAATGTGTATCACGC ATG AGTGAAGAAGACATC ZmCP4CcrM M R L A E C V S R M S E E D I 846 GGAGGCGATACAGAGTTATCTGACAGCCCGACTTTACCGTTAAAC G G D T E L S D S P T L P L N 891 AGTATTTTGGCCGGTAATTGCATCGAGATTTTAAAAACGCTGCCT S I L A G N C I E I L K T L P 936 GATAATTCGGTTGATCTTATTTTTGCTGATCCGCCTTACAATCTT D N S V D L I F A D P P Y N L 981 CAATTAAGTGGTGAGCTTTTTAGACCGGAAGGCAGCCGTGTCGAT Q L S G E L F R P E G S R V D Figure 8. DNA sequence and pred icted amino acid sequence of: ZmCP4ccrM . The DNA sequence in the 5â€™ to 3â€™ direction and the predicted amino acid sequence presented in single letter code belo w the first nucleotide of each codon are shown for the putative RNaseHII and ZmCP4ccrM ORFs. The DNA sequence includes the ZmCP4ccrM promoter region as well as upstream and downstream sequence of both ORFs . The ATG translation initiation sites are highlighted in green and the TAG and T GA stop sites are highlighted in red. ZmCP4ccrM has two possible translation in itiation sites. The nucleotide sequence in uppercase was veri fied in both strands.
164 1026 GCCGTTAATAACGCATGGGATAAATTCGATACTTTTGCCGCCTAT A V N N A W D K F D T F A A Y 1071 GACCATTTTACCCGTTTGTGGCTGAAAGAAGCGCATCGCGTTTTA D H F T R L W L K E A H R V L 1116 AAAGAAGATGGCACGATTTGGGTTATTGGCAGCTATCACAATATT K E D G T I W V I G S Y H N I 1161 TTCAGGGTAGGCACCGCCTTACAGGATCAGGGCTTTTGGATATTA F R V G T A L Q D Q G F W I L 1206 AACGATATCATCTGGCGTAAATCCAATCCGATGCCCAATTTTAAA N D I I W R K S N P M P N F K 1251 GGTCGACGATTTACCAATGCCCATGAAACCCTGATATGGGCTTCA G R R F T N A H E T L I W A S 1296 AAATCAGATAAATCCCGCTATGTCTTTAATTATGCCTCTTTGAAA K S D K S R Y V F N Y A S L K 1341 ACTTTTAACGATGACCTTCAGATGCGGTCTGACTGGTTATTACCG T F N D D L Q M R S D W L L P 1386 ATATGTTCAGGAAATGAGCGTTTAAAAGGAGAAAATGGGCAGAAG I C S G N E R L K G E N G Q K 1431 ATACATCCGACCCAAAAGCCCGAAGCCTTGCTGTATCGGATTATA I H P T Q K P E A L L Y R I I 1476 TTGGCTTCATCGCGTCCCGATGATGTCATTCTTGATCCTTTTTTT L A S S R P D D V I L D P F F 1521 GGCACGGGAACCACGGGCGTTATTGCCCGCCATCTTCGGCGGCAT G T G T T G V I A R H L R R H 1566 TGGATCGGAATTGAACAAGACCCGACTTATATTAAAGCCGCCCAA W I G I E Q D P T Y I K A A Q 1611 GCCAGAATAGATAAAGCCGAAGTATTTGATGAAGCCTTGATGGGG A R I D K A E V F D E A L M G 1656 CAGGCCTCTAATAAACGGAAACAACCTCGCGTTACCTTTGGTTGT Q A S N K R K Q P R V T F G C 1701 TTGATGGAAAACGGTTTTATCCGTCCCGGTCATATTCTGTATGAC L M E N G F I R P G H I L Y D 1746 AGCCGCCGCCGCTTTAAAGCGGTTGTGAATGTTGATGGTGCGTTG S R R R F K A V V N V D G A L 1791 CAATCGGCCGATGGTCGGTCAGGATCTATCCATAAGCTGGGTGCC Q S A D G R S G S I H K L G A 1836 CAATTACAACAGGCCGTTTCCTGTAATGGCTGGATTTTTTGGCAT Q L Q Q A V S C N G W I F W H 1881 TTTGAAGAAAATAATATTTTGTTGCCGTTAGATATCTTACGACAA F E E N N I L L P L D I L R Q 1926 CGCTATCTTGCCGAGAATCCT TGA tgtctttttataaccaaaatg R Y L A E N P * 1971 tttcagtacctgaaatgaccgatatccgtctttatttttgtccga 2016 ccgcctttgtgaatgcaccgatggcgccttcgggtgcaacattcc 2061 gtttggctggaggcttgaaatggtttggcgctttttccgttatca 2106 tggtttcaaaaggcgaaattgttcaggaacgggtgatttctgttg 2151 cggatatcaatagctttttgaacagcctttcaacagaaatgcagt 2196 tggatgcctatagaaccatttcaaggattgtggcaccaagagcgg 2241 ctatccaattaggtaaaagcaa Figure 8. Continued.
165 Figure 9. Identification of the ORFs of a cloned Z. mobilis CP4 region containing ZmCP4ccrM. The boundaries of the genomic region cloned from Z. mobilis CP4 is indicated by the verti cal lines in the diagram. ZmCP4ccrM 1149/1122 bp ORF is represented by the yellow box. The flanking ORFs are represented by green boxes. The putative proteins encoded by the flanking ORFs, based on amino acid sequence similarity (NCB I-BLAST), are indicated below each box. The direction of transcription of each ORF is indicated by the open arrows. The double open arrows in the ZmCP4ccrM box indicate two possible translation initiation sites. ZmCP4ccrM1149/1122 bp RNase HII HMPP kinase (thiamine phosphate biosynthesis) MrsA Phosphomannomutase (glycolysis/ gluconeogenesis) C-terminal Dihydropteroate synthase (DhpS) (folate biosynthesis) N-terminal
166 ZM MRLAECDSRMSEEDIGGDTELSDSPTLPLNS CC MKFGPET BA MSLVRLAHELPIEAPRTAWLDS SM MSSVVSLAEISRAARPLNWLDS AT MAAVFPLADFRRAGFDRAGESDAWKDS MOTIF IV ZM ILAGNCIEILKTLPDNSVDLIFADPPYNLQLSGELFRPEGSRVDAVNNAWDKFD CC IIHGDCIEQMNALPEKSVDLIFADPPYNLQLGGDLLRPDNSKVDAVDDHWDQFE BA IIKGDCVSALERLPDHSVDVIFADPPYNLQLGGDLHRPDQSMVSAVDDHWDQFE SM IIKGDCVAALNALPDHSVDVVFADPPYNLQLGGTLHRPDQSLVDAVDDDWDQFA AT IIKGDCVAALDALPSQSVDAIFADPPYNLQLGGTLHRPDQSLVDAVDDEWDQFA * * * * *** ********** * * ** * * ** ** * MOTIF V MOTIF VI MOTIF VII ZM TFAAYDHFTRLWLKEAHRVLKEDGTIWVIGSYHNIFRVGTALQDQGFWILNDIIWRK CC SFAAYDKFTREWLKAARRVLKDDGAIWVIGSYHNIFRVGVAVQDLGFWILNDIVWRK BA SFQAYDAFTRAWLLACRRVLKPNGTIWVIGSYHNIFRVGTQLQDLGFWLLNDIVWRK SM SFEAYDAFTRAWLLACRRVLKPTGTIWVIGSYHNIFRVGAILQDLHFWVLNDIVWRK AT SFDAYDAFTRAWLLACRRVLKPNGTIWVIGSYHNIFRVGAMLQNLDFWILNDIVWRK * *** ****** **** **************** * ** ******** MOTIF VIII ZM SNPMPNFKGRRFTNAHETLIWASKSDKSRYV-FNYASLKTFNDDLQMRSDWLLPICS CC SNPMPNFKGTRFANAHETLIWASKSQNAKRYTFNYDALKMANDEVQMRSDWTIPLCT BA TNPMPNFRGRRFQNAHETLIWASREQKGKGYTFNYEAMKAANDDVQMRSDWLFPICT SM TQPDAELQGRRFQNAHETLIWATANAKAKGYTFNYEAMKAANDDVQMRSDWLFPICS AT TNPMPNFKGRRFQNAHETMIWASRDPKAKSYTFNYDALKASNDDVQMRSDWLFPICS * * ** ********* *** * *** ****** * * MOTIF X MOTIF I ZM GNERLKGENGQKIHPTQKPEALLYRIILASSRPDDVILDPFFGTGTTGVIARHLR CC GEERIKGADGQKAHPTQKPEALLYRVILSTTKPGDVILDPFFGVGTTGAAAKRLG BA GSERLKDENGDKVHPTQKPEALLARIMMASSKPGDVILDPFFGSGTTGAVAKRLG SM GSERLKGDDGKKVHPTQKPEALLARILMASTKPGDVVLDPFFGSGTTGAVAKRLG AT GHERLKGDDGKKVHPTQKPEALLARIIMASTKPGDIVLDPFFGSGTTGAVAKRLG ** * * ********** * ******************* * Figure 10. Amino acid sequence alignment of CcrM homologues. The predicted amino acid sequence of the CcrM sequence of Z. mobilis was aligned with the amino acid sequence of the CcrM sequence homologues from C. crescentus, B. abortus , S. meliloti, and A. tumefaciens (NCBI ENTREZ) as shown. The identical amino acids are underscored with fuchsia asterisks. The conserved catalytic domains (IV, V, VI, VII, VIII) are highlighted in yellow and AdoMet binding domains (X, I, II, III) ar e highlighted in green (Malone et al. , 1995; Timinskas et al. , 1995; Jeltsch, 2002). The putat ive N6 adenine-specific DNA MTase signature is highli ghted in red (Timinskas et al. , 1995).
167 MOTIF II MOTIF III ZM RHWIGIEQDPTYIKAAQARIDKAEVFDEALMGQASNKRKQPRVTFGCLMENGFI CC RKFIGIEREAEYLEHAKARIAKVVPIAPEDLDVMGSKRAEPRVPFGTIVEAGLL BA RHFVGIEREQPYIDAATARINAVEPLGKAELTVMTGKRAEPRVAFTSVMEAGLL SM RHFVGIEREQPYIDAAAERIAAVEPLGKATLSVMTGKKAEPRVAFNTLVESGIL AT RHFVGIEREQDYIDAASARITAVEPLGKAELTVMTGKKAEPRVAFNTLVESGFV * *** * ** *** * * * ZM RPGHILYDSRRRFKAVVNVDGALQADGR-SGSIHKLGAQLQQAVSCNGWIFWHF CC SPGDTLYCSKGTHVAKVRPDGSITV-GDLSGSIHKIGALVQSAPACNGWTYWHF BA RPGTVLCDERRRFAAIVRADGTLTANGE-AGSIHRIGARVQGFDACNGWTFWHF SM KPGTVLTDAKRRYSAIVRADGTLASGGE-AGSIHRLGAKVQGLDACNGWTFWHF AT RPGQVLTDARRRYSAIIRADGTLASGGT-AGSIHRLGAKVQGLDACNGWTFWHF ** ** **** ** * **** *** GenBank accession number (NCBI ENTREZ) ZM SEENNILLPLDILRQRYLAENP CC -KTDAGLAPIDVLRAQVRAGMN Q45971 BA -EENGVLKPIDALRKIIREQMAAAGA AAB71351 SM -EEGSVLKPIDELRSVIRNDLAKLN AAB71350 AT -EDGDALKPIDDLRTIIRSEMAKAE AAK53552 * *** ** Figure 10. Continued.
168 0 5 10 15 20 25 30 35 1183552698610312013715417115 nt window ZmCP4ccrM promoterFree energy value (kcal/mol) Figure 11. Primer extension and promoter sequence analysis of ZmCP4ccrM. The red arrow points to the primer extensi on product which coincides with the cytosine nucleotide highlighted in red in the sequence. The four ZmCP4ccrm sequencing lanes, read from bottom to the top, is the reverse complement to the sequence shown. The two possibl e translation initiation sites are highlighted in grey. Highlighted in cyan is the -35 region with the nucleotides (5â€™-ANNNNNCTNG-3â€™) conserved in Z. mobilis promoters (Ingram et al. , 1989) in bold. Underlined are possible RBS. Highlighted in yellow is the sequence 5â€™-AGGA-3â€™ conserved in Z. mobilis RBS regions (Ingram et al. , 1989). The sequence highlighted in green was determined to be the region of low thermal stability, which coincides w ith the primer extension, determined 10 region, by plotting the calculated free energy values of a moving 15 nucleotide window along the promoter s hown in the graph, based on a method of -10 region prediction developed by Kanhere and Bansal (2005). The corresponding graphed points ar e pointed out with arrows. ACGTâ€¢CGTTACTTTTTCTACATTTTTTGAAGCTATAA â€¢AATAATCACTTCAATCATTATCCATAACCGC â€¢TTTTGTATTTTTACAATG A AGCCT CT T G ACCG â€¢AAGAGTCGCATTG AATCACAGTAAAGTC GG â€¢GTTTCTGTGAA AGGA TCAGTATTT ATG CGCC â€¢TTGCCGAATG TGTATCACGC ATG AGT First ATG -10 re g ion
169 Promoter region of ZmCP4ccrM of Zymomonas mobilis CP4 GCTTTTGTATTTTTAC --IR1 ----------IR2 -AATGAAGCCTCTTGACCGAAGAGTCGCATTGAATCACAGTAAAGTCGGGTT ----TCTGTGAAAGGATCAGTATTTATGCGCCTTGCCGAATGTGTATCACGCATG Promoter region of ccrMI of Caulobacter crescentus : Accession number: S76857 ----------IR1----------TGAACGTCTTCAACTTTTGAGTCTGATCAGACTCAAAAGCGCCTGAA ---IR2 ----AGGTGAAAGGCCGTGGTTAACGGCCCGCTAACCACGTCTCTCAACACCGGAT ---IR3--TTACCAGGAAGACTCATGATTCCGCTCTCTTTCTTGAGGACGTGGGACCATG Promoter region of ccrM of Brucella abortus : Accession number: AF411571 CTTGCACATGGATCACGTCGTCACGATGACAAGTCGATAATTATCTCTG CCTTATTGGGCGCGCAAAGGCCGCAAAGCCGGGCTTTCCCTGTGATATTAAG AAAAGATTTACGATTTCAAGCACTTGGCGTTAACGGCATATTTACCCTACGC AGTAACCATAGGAACAAGTTTTTTGCGTTCACAGGTAATCGAGTATCCCATG Figure 12. Promoter region of CcrM homologues compared to ZmCP4ccrM . Comparative sequence analysis of ZmCP4ccrM promoter region identified similar putative regulatory elements as found in the ccrMI promoter of C. crescentus (Stephens et al. , 1995). Highlighted in grey are the translation initiation sites. Highlighted in yellow ar e the CcrM recognition sites. The IRs are indicated by fuchsia arrows above the sequence. Highlighted in green are the putative RBS site for ZmCP4ccrM and the RBS sites for the ccrM homologues found in B. abortus (Robertson et al., 2000) and C. crescentus (Robertson et al., 2000). Underlined is th e second possible RBS for ZmCP4ccrM . Highlighted in red are the ccrM transcription initiation sites for B. abortus, C. crescentus (Robertson et al., 2000) and Z. mobilis . The GenBank accession number for the CcrM homologues are as indicated (NCBI ENTREZ).
170 Figure 13. In vivo activity of recombinant ZmCP 4CcrM from pPROTet.E133/ ZmCP4ccrM. To assay in vivo activity of the recombinant ZmCP4CcrM, pPROTet.E133/ ZmCP4ccrM was extracted from E. coli BL21pro after incubation with different amounts of the aTc expression inducer. The concentration of aTc is indicated abov e the lanes 7-11. All of the plasmid extracts were digested with BamHI, to linearize the expression vector. Lanes 1 and 2 contain BLl21PROâ€™s Tet repre ssor plasmid digested with BamHI and BamHI/HinfI respectively. Lane 5 contains the standard DNA molecular weight marker /HindII Digest containing DNA fragments of 23.13 kb, 9.42 kb, 6.56 kb, 4.36 kb, 2.32 kb, 2.03 kb, and 564 bp, from top to bottom. Lanes 3 and 4 contain the 2.2kb expression vector pPROTet.E133 digested with BamHI and BamHI/HinfI respec tively. Lane 6 contains pPROTet.E133/ ZmCP4ccrM and Tet repressor plasmid digested with BamHI. Lanes 7-11 contain pPROTet.E133/ ZmCP4ccrM and Tet repressor plasmid digested with both BamHI and HinfI. Approximately 85ng of DNA was loaded into lanes 3-4 and 6-11 in the top gel, while the corresponding lanes of the bottom gel were load ed with 400ng of DNA. 3 4 5 6 7 8 9 10 11 0 25 50 100 200 ng/ml aTc 1 2
171 Figure 14. In vivo activity of recombinant Zm CP4CcrM in cultured GM4715/ pBBR1MCS/ ZmCP4ccrMPRO . Lane 1 contains th e standard DNA molecular weight marker /HindII Digest containing DNA fra gments of sizes (kb) as indicated next to the gel. Lane 2 contains 160 ng of pBBR1MCS/ ZmCP4ccrMPRO , extracted from E. coli GM 4715, digested with SfiI. Lanes 3 and 4 contains 160 ng and 320 ng respectively of pBBR1MCS/ ZmCP4ccrMPRO digested with SfiI and HinfI. kb 23.13 9.42 6.56 4.36 2.32 2.03 1234
172 Figure 15. In vivo activity of recombinant ZmCP4C crM in cultured HMS174(DE3)/pET24b/ ZmCP4ccrM . Lane 1 contains the st andard DNA molecular weight marker /HindII Digest containing DNA fragments of 23.13 kb, 9.42 kb, 6.56 kb, 4.36 kb, 2.32 kb, 2.03 kb, and 564 bp, from top to bottom. The pET24b/ ZmCP4ccrM recombinant plasmid was extracted from E. coli HMS174(DE3), after 2 hour of 1 mM IPTG induction, and digested with ClaI, shown in lanes 2 and 3, or with both ClaI and HinfI, shown in lane 4. 1 2 3 4
173 Figure 16. ZmCP4CcrM protein purification and analysis on SDS-PAGE from BL21Pro/ pPROTet.E133/ ZmCP4ccrM transformants induced at 500 ng/ml aTc. The protein band appropriately positioned for the expected 47.42 kDa C-terminal histidine tagged recombinant ZmCP4C crM protein is pointed out with a fuchsia arrow. Lane 1 contains th e Low Range protein molecular weight marker at kDa as indicated next to the gel. Lane 2 contains 500 ng bovine serum albumin (BSA) (66.4 kDa). Lane 3 contains whole cell-free lysate. Lane 4 shows the HiTrap Chelating HP column wash fraction. Lane 5 contains the HiTrap Chelating HP column elu tion fraction with 300 mM imidazole. kDa 97.4 66.2 45 31 21.5 14.4 1 2 3 4 5
174 Figure 17. ZmCP4CcrM protein purifica tion and analysis on SDS-PAGE from HMS174(DE3)/pET-24b/ ZmCP4ccrM transformants induced at 1 mM IPTG. The protein bands approp riately positioned for th e expected 43.74 kDa Cterminal histidine tagged recombinan t ZmCP4CcrM protein is pointed out with fuchsia arrows. Two set of SDSPAGE gels are shown. A-I and B-I are gels containing fractions eluted from a Sephadex G-75 column and directly below each of them are their corresp onding gels, A-II and B-II respectively, containing the subsequent fr actions eluted from a Ni2+ charged HiTrap Chelating Sepharose HP column. In gel I, lane 1 contains whole cell-free lysate, lane 4 contains the low range molecular weight standard (97.4, 66.2, 45, 31, 21.5, 14.4 kDa), and lanes 3 and 515 contain Sephadex G-75 elution fractions containing protei ns with decreasing average molecular weight. In gel A-II, lane 1 contains a sample of the combined eluted fractions from the Sephadex G-75 column (gel A-I, lanes 5-11), before applying them to the HiTrap Chelating HP column, lane 9 contains the low range DNA molecular weight standard, and the rest of the lanes contain the HiTrap Chelating HP column stepwise imidazole gradient eluted fractions. In gel B-I, lanes 3 and 8 contain the low range molecular weight standard, and the rest of the lanes contain Sephadex G-75 column elution fractions, where 15 l of every other eluted fraction was loaded. In gel B-II, lanes 1 and 5 contain the low range molecular weight standard, and the re st of the lanes contain the HiTrap Chelating HP column stepwise imidazole gradient eluted fractions, derived from the collected Sephadex G-75 frac tions (Gel B-I, lanes 4-7 and 9-11, including the eluted fractions in between that were not loaded on the gel). The concentration of imidazole in the el ution fractions are as follows: (A-II) Lanes 2-4 at 60 mM, lanes 5-8, 10 at 100 mM, and lanes 11-13 at 300 mM. (B-II) Lanes 2-4 at 60 mM, lanes 6-8 at 100 mM, and lanes 9-10 at 300 mM. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 1 2 3 4 5 6 7 8 9 10 11 12 13 A I B-I A II B-II 1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10 11 12
175 1 2 3 4 Buffer 1 2 3 4 5 6 7 8 Lane Figure 18. I n vitro activity of partially purified reco mbinant ZmCP4CcrM extracted from cultured BL21Pro/pPROTet.E133/ ZmCP4ccrM . Lane contained the DNA molecular weight marker /HindIII Digest cont aining DNA fragments of 23.13 kb, 9.42 kb, 6.56 kb, 4.36 kb, 2.32 kb, and 2.03 kb, from top to bottom. Unmethylated pP3226 (3226 bp)substrat e DNA was incubated for 1 hour at 30C in various buffers with partially purified recombinant protein extracts, then digested with HinfI (lane2-7). In a 20 l reaction, CcrM recombinant protein extract was added to 640 ng of DNA to produce an estimated final buffer composition as follows: Lane 2, buffer 1: 10 mM Bis-Tris Propane-HCl, 3 mM sodium phosphate, 18 mM NaCl, 10 mM MgCl2, 1 mM dithiothreitol, 3% glycerol, pH 7.0, 80 M AdoMet, 5 mM ME. Lane 3, buffer 2: 10 mM Tris-HCl, 3 mM sodium phosphate, 68 mM NaCl, 10 mM MgCl2, 1 mM dithiothreitol, 3% glycerol, pH 7.9, 80 M AdoMet, 5 mM -ME. Lane 4, buffer 3: 50 mM Tris-HCl , 118 mM NaCl, 3 mM sodium phosphate, 10 mM MgCl2, 1mM dithiothreitol, 3% glycerol, pH 7.9, 80 M AdoMet, 5 mM ME. Lane 5, buffer 4: 20 mM Tris-Acetate, 50 mM potassium acetate, 3 mM sodium phosphate, 18 mM NaCl, 10 mM Magnesium acetate, 1 mM dithiothreitol, 3% glycerol, pH 7.9, 80 M AdoMet, 5 mM ME. Lane 6, buffer : 50 mM potassium phosphate, 150 mM potassium acetate, 3 mM sodium phosphate, 18 mM NaCl, 3% glycerol, pH 7.5, 80 M AdoMet, 5 mM ME. Lane 7, buffer : 50 mM potassium phosphate, 3 mM sodium phosphate, 18 mM NaCl, 3% glycerol, pH 7.5, 80 M AdoMet, 5 mM ME. Lane 8: The DNA substrate was also incuba ted for 1 hour at 37C with M.SssI to protect it from digestion by AvaI, to demonstrate in vitro methylase protection from REase digestion. In a 20 l reaction, 1 unit of M.SssI was added to 640 ng of DNA in 1x buffer mix (10 mM Tris-HCl, 50 mM NaCl, 10 mM MgCl2, 1 mM dithiothreitol, pH 7.9, 80 M AdoMet).
176 bp 1 2 3 23130 9416 4361 6557 2322 2027 564 Figure 19. In vitro activity of purified recombinant ZmCP4CcrM extracted from cultured HMS174(DE3)/pET-24b/ ZmCP4ccrM . Lane 1 contains the DNA molecular weight marker /HindII Digest containing DNA fragm ents of the sizes (bp) as indicated next to the ge l. Unmethylated pNEB193 PacinsertBsaI (2708 bp) was incubated for 1 hour at 30C with purified recombinant ZMCP4CcrM extracts at 1mM in 36.7 mM NaCl, 20 M AdoMet, 50 mM potassium phosphate buffer pH 7.5. The sample was subsequently divided. Approximately 180 ng of substrate DNA wa s digested with excess HindIII, to linearize the plasmid (Lane 2). Approximately 360 ng of the substrate DNA was digested with excess HindIII and Hi nfI for 2 hours at 37C (Lane 3).
177 Figure 20. Analysis of the methylation specificity of ZmCP4CcrM. ZmCP4CcrM specificity was analyzed using the Type IIS restriction endonuclease BsaI and the plasmid construct pNEB193 PacI/BsaI as the s ubstrate DNA. After in vitro methylation of the DNA substrate with recombinant ZmCP4CcrM, the methylated substrate DNA was digest ed with PvuII, producing a 344 bp fragment. The BsaI recognition site (5â€™-GGTCT-3â€™) and its downstream cleavage site are highlighted in fuch sia. The CcrM DNA MTase recognition site (5â€™-GANTC-3â€™) is highlighted in orange. BsaI digestion cleaves the CcrM site so that, two PvuII/BsaI DNA fragments are produced: a 125 bp and a 219 bp fragment. The red star represents [3H]-labeling of the 219 bp fragment that would be produced if only the adenines in the CcrM recognition site are methylated by purified recomb inant ZmCP4CcrM. Both adenines of the CcrM recognition site are pr esent in the 219 bp fragment. CGGATCC GGTCT CTA GAGTC GACTGTTTA GCCTAGG CCAGA GAT CTCAG CTGACAAAT PvuII PvuII pNEB193 Pac I/ Bsa I 125bp 219bp CcrM BsaI 344bp Pvu II + ZmCP4CcrM 2708b p PvuII/BsaI
178 Table 5. Determination of the methylation site of ZmCP4CcrM using Type IIs BsaI. DNA fragment 344 bp 125 bp 219 bp Exp set 1 110 CPM 0 CPM 0% 81 CPM 100% Exp set 2 --14 CPM 3% 475 CPM 97% 1.5% 98.5% .5% In vitro methylation reactions of substrate DNA contained 6 M [3H]-AdoMet + 20 M AdoMet and 1 mM of purified recombinant ZmCP4CcrM. The DNA restriction fragments 344 bp (containing the CcrM site), 129 bp (containing the 5â€™ half of the CcrM site including only cy tosine methylation targets), and 219 bp (containing the 3â€™ half of the CcrM si te including both adenine and cytosine methylation target) were fractionated by nondenaturing PAGE and isolated. [3H]labeling of the extracted DNA fragments wa s determined by scintillation counting, and reported as CPM minus background CPM and the percentage of the [3H]-label in the 125 bp and 219 bp fragments. The amount of [3H]-label in the DNA fragments was relatively low due to the specific activity of the AdoMet used to label the DNA.
179 0.01 0.1 1 10 123456789 Time (Hour)OD600 Figure 21. Growth curves of Zymomonas mobilis CP4 with or without plasmid borne copies of ZmCP4ccrM. The wild-type Z. mobilis growth curve is highlighted in dark blue with points marked by diamonds. The growth curve for Z. mobilis transformed with pLOI1844 alone is hi ghlighted in fuchsia with points marked by squares, or transformed with pLOI1844 and pBROriV Rop is highlighted in yellow with points marked by triangles, or transformed with pLOI1844 and pBROriV Rop/ ZmCP4ccrM is highlighted in cyan with points marked by xâ€™s. Growth was mon itored every hour for eight hours by determining the OD600.
180 Table 6. Effect of ZmCP4ccrM overexpression on cell viability. OD600 CFU CFU Z. mobilis 0.445 TNTC TNTC Z.m /pBBR1MCS/ ccrMPRO 0.64 118 201 Z.m /pLOI1844/pBROriV Rop ZmCP4ccrM 0.195 30 14 Z.m /pLOI1844/ pBROriVRop 0.52 TNTC TNTC 200 l of 3 day culture was diluted with 3 ml GYx media and its OD600 was determined with a spectrophotometer. Cell culture was plated onto GYx agar with volumes adjusted in order to plate equa l cell densities based on spectrophotometer readings.
181 A. B. Figure 22. Microscopy of Zymomonas mobilis CP4 overexpressing ZmCP4ccrM at 200X magnification. A . pBROriV Rop in Z. mobilis CP4 with pLOI1844 at 200X magnification using Nomarski microscopy. B . pBROriV Rop/ ZmCP4ccrM in Z. mobilis CP4 with pLOI1844 at 200X ma gnification using Nomarski microscopy. The more grossly elongat ed cells are pointed out by arrows. 10 microns 5 microns
182 Figure 23. Microscopy of Zymomonas mobilis CP4 overexpression ZmCP4ccrM at 400X magnification. Z. mobilis CP4 containing pLOI1844 transformed with pBROriV Rop/ ZmCP4ccrM cells were examined at 400X magnification using Nomarski optics microscopy. Th e dimensions of the cells were measured relative to a micrograph of a micrometer at the same magnification. 20 micron 10micron
183 Figure 24. Unstained and DAPI stained Zymomonas mobilis CP4 transformants overexpressing ZmCP4ccrM . Z. mobilis CP4 containing pLOI1844 transformed with pBROriV Rop/ ZmCP4ccrM cells was stained with DAPI to examine DNA distribution in the cel ls and viewed at 400X magnification. The top two micrographs show th e cells observed using Nomarski microscopy. The bottom two micrographs show the same cells under UV to observe DAPI staining. Arrow A points to an elongated cell w ith what appears to be the beginning of septum formation at multiple point (seen under Nomarski in the top micrograph) whic h corresponds to separate points of greater DNA density (seen under UV, causing DAPI stained nucleic acid to fluoresce blue, in the bottom microgra ph). Arrow B points to an elongated cell with one visually evid ent point of septum forma tion, yet shows distinctly separate multiple points of greater DNA density. Arrow C points to an elongated cell where neither septum formation, nor variation in DNA density, is visually evident. A C B C B A
184 CHAPTER 6 GENERAL SUMMARY AND DISCUSION Zymomonas mobilis CP4 has a relatively low effi ciency of transformation with foreign DNA from several strains of E. coli compared to DNA originating from Z. mobilis CP4. This observation had l ead to the h ypothesis that Z. mobilis CP4 contains a restriction system or systems that reduce th e efficiency of genetic exchange. Based on this hypothesis, the initial focus of this project was to identif y and characterize restriction endonuclease and or DNA methyltransferases present in Z. mobilis CP4 that may act to reduce the transformation efficiency wh en foreign genes are introduced. An understanding of the mechanism(s) that cau se the low transformation efficiency of Z. mobilis CP4 may provide a basis for modifying Z. mobilis CP4 to obtain strains with higher transformation efficiency. In this study, an mrr -like endonuclease, ZmCP4Mrr, was cloned from Z. mobilis CP4. Transformation efficiency experi ments of plasmid borne recombinant ZmCP4mrr into E. coli K12 strains with different MTas e backgrounds indicate that ZmCP4Mrr recognizes and restricts EcoKDam methyl ated DNA (5â€™-GmATC-3â€™), but not EcoKDcm (5â€™-CmC(A/T)GG-3â€™), EcoKHsdM (5â€™-AmACNNNNNNGTGC-3â€™), or EcoKCcrM (5â€™mATGCAT-3â€™) methylated DNA. Consequently, it was concluded that ZmCP4Mrr is a methyl-dependant restriction endonucleas e that specifically recognizes adenine methylated sites that overlap Da m methylated sites. Putative mrr homologues have been widely identified in prokaryotes, however , until now, the only Type IV Mrr REase system that has been characterized is from E. coli K12 (Waite-Rees et al . 1991; Roberts
185 and Macelis, 1996). Based on transformation efficiency experiments, EcoKMrr is characterized as a methyl-dependant restri ction endonuclease that appears to cleave adenine and cytosine methylated DNA, methylated by various known DNA MTases (Waite-Rees et al . 1991). A consensus recognition site has not yet been determined for EcoKMrr (Waite-Rees et al . 1991; Roberts and Macelis, 19 96). If ZmCP4Mrr is like EcoKMrr, it may recognize and cleave me thylated DNA in a similar manner, and may recognize other methylation patterns in addition to Dam methylated DNA. Transformation efficiency experiments of Z. mobilis CP4 with plasmid DNA originating from damstrains of E. coli K12 when compared to plasmids from dam+ strains did not improve transformation efficien cy. Consequently, it was concluded that Z. mobilis contains one or more REase systems in addition to ZmCP4Mrr. A restriction endonuclease ZanI had been pr eviously iden tified in the Z. mobilis strain Z. anaerobia (Sun and Yoo, 1988); however, DNA restriction analysis in this study indicated that Z. mobilis CP4 does not have a ZanI-like REase. Recent release of the genome sequence of a Zymomonas strain closely related to Z. mobilis CP4, Z. mobilis ZM4 revealed the presence of a putative Type I R-M system (Seo et al ., 2005; NCBI ENTREZ). Closely related organisms, different st rains of the same species, often have different R-M systems (Birge, 1981). Consequently, while Z. mobilis CP4 may have a similar Type I R-M system like Z. mobilis ZM4, an entirely different restriction system(s) may be responsible for low transformation efficiency. Most REase systems have a cognate MT ase, however, solitary DNA MTases like Dam, Dcm, and CcrM have been identifie d in prokaryotes that do not function in restriction systems but instead are involve d in other cellular processes (Roberts and
186 Macelis, 1996; Bujnicki, 2001; Low et al., 2001; Jeltsch, 2002). In this study, a solitary CcrM DNA MTase, ZmCP4CcrM, was cloned from Z. mobilis CP4. Sequence analysis indicates that ZmCP4CcrM has high simila rity to the previously identified and characterized CcrM homologues found in C . crescentus and several other alphaproteobacteria. In vivo and in vitro DNA restriction assays indicate that ZmCP4CcrM methylates DNA at sites overl apping HinfI recognition site s (5â€™-GANTC-3â€™) and protects the DNA from cleavage by HinfI. Like ot her characterized CcrM homologues, the protection from HinfI digesti on indicates that ZmCP4CcrM specifically methylates the adenine of 5â€™-GANTC-3â€™ sites. Methylation of the adenosine of a construct containing a 5â€™-GANTC-3â€™ site by ZmCP4CcrM wa s directly confirmed by an in vitro nucleotide specificity assay. Overexpression of CcrM from plasmid borne ZmCP4ccrM in Z. mobilis CP4 showed a phenotypic effect analogous to the effect of overexpression of CcrM homologues characterized in severa l alpha-proteobacteria (Reisenauer et al., 1999b). Microscopic examination of Z. mobilis CP4 cells overexpressing ZmCP4ccrM revealed that most cells are larger in length and diameter than th e control and a subpopulation in the culture are extremely elongated filamentous cells. The filamentous cells contained multiple chromosomes not seen in control cells. It was concluded that overexpression of ZmCP4ccrM in these cells caused disruption of normal chromosomal segregation and cell division. Analysis of cultured cell viability indicated that Z. mobilis CP4 overexpressing ZmCP4ccrM has decreased viability. Overexpression of ZmCP4ccrM has a detrimental effect on normal cell morphology, cell division, and viability, which is consistent with an involvement of ZmCP4CcrM in cell cycl e regulation, as proposed for the CcrM
187 homologues found in C. crescentus and other alpha-proteobacteria (Reisenauer et al. , 1999b). Since all studied CcrM homologues have high sequence similarity, and have all been shown to have similar activity, it is likely that they all also ha ve similar functions in the cell via similar mechanisms that ar e still to be directly determined. In conclusion, two DNA modification enzymes have been identified in Z. mobilis CP4: a solitary methyl-dependant re striction endonuclease, ZmCP4Mrr, which recognizes and cleaves adenine methylat ed DNA (methylated by EcoKDam), and a solitary N6-adenine specific DNA methyltransf erase, ZmCP4CcrM, which contributes to the DNA methylation pattern in Z. mobilis CP4, and is proposed to be involved in cell cycle regulation. Future studies of ZmCP4M rr should include: (1) de termining conditions for purification of ZmCP4M rr, (2) determining conditions required for in vitro activity including possible co-f actor requirements, (3) determining if ZmCP4mrr is essential for cell viability or if ZmCP4mrr mutants are defective in DNA repair, and (4) determining if overexpression of ZmCP4mrr in Z. mobilis CP4 is detrimental for growth or viability. Future studies of ZmCP4CcrM should include: (1) mass spectrophotometry or amino acid sequence analysis to determin e the translation initiation site of ZmCP4ccrM , (2) transformation of Z. mobilis CP4 with M.HinfI (expressed from a strong promoter or from a ZmCP4ccrM promoter) to determine if unregul ated methylation of 5â€™-GANTC-3â€™ sites effect cell morphology, growth, and viability, and determine if ZmCP4ccrM expression is transcriptionally regulated (3) transformation of Z. mobilis CP4 with plasmid borne ZmCP4ccrM (expressed from a strong promoter) to determine if the
188 detrimental effect ZmCP4CcrM overexpr ession has on cell morphology, growth and viability increases. Future studies of Z. mobilis REase systems should include (1) determining if Z. mobilis CP4 contains a R-M system similar to th e putative Type I R-M system identified in the Z. mobilis ZM4 genome (Seo et al. , 2005), and (2) engineering a Z. mobilis ZM4 strain that is ZmCP4mrr and putative Type I REase deficient, which may create a Z. mobilis strain with improved transformation efficiency.
189 LIST OF REFERENCES Adams, G. M., and R. M. Blumenthal. 1995. Gene pvuIIW : A possible modulator of Pvu II endonuclease sub unit association. Gene . 157:193-199. An, H., R. K. Scopes, M. Rodriguez, K. F. Keshav, and L. O. Ingram. 1991. Gel electrophoretic anaysis of Zymomonas mobils glycolytic and fermentative enzymes: Identification of alcohol dehydroge nase II as a stress protein. Journal of Bacteriology . 173:5975-5982. Anton, B. P., D. F. Heiter, J. S. Benner, E. J. Hess, L. Greenough, L. S. Moran, B. E. Slatko, and J. E. Brooks. 1997. Cl oning and characterization of the Bgl II restriction-modification system reveals a possible evolutiona ry footprint. Gene . 187:19-27. Arber, W., and D. Dussoix. 1962. Host specificity of DNA produced by Escherichia coli. 1. Host controlled modification of bacteriophage lambda. Journal Molecular Bioiology 5:18-29. Arber, W., and S. Linn. 1969. DNA modification and restriction. Annual Review of Biochemistry. 38:467-500. Aristidou, A., and M. Penttila. 2000. Metabo lic engineering applications to renewable resource utilization. Current Opinion in Biotechnology . 11:187-198. Arraj, J. A., T. H. We, and M. G. Mari nus. 1990. Expression of a DNA methylation ( dam ) gene in Escherichia coli K-12. Current Microbiology . 20:133-136. Atlas, R. M., (L. C. Parks, ed.). 1997. Handbook of Microbial Media . 2nd edition. CRC Press, Inc. Boca Raton, Fl. Ausubel, F. M., R. Brent, R. E. Kingston, D. D. Moore, J. G. Seidman, J. A. Smith, and K. Struhl. 1998. Current Protocols in Molecular Biology . Greene publishing assocociation and Wiley Interscience. NY. Balbas, P., X. Soberon, F. Bolivar, and R. L. Rodriguez. 1988. The plasmid, pBR322. In R. L. Rodriguez and D.T. Denhardt (eds.). Vectorsâ€”A Survey of Molecular Cloning Vectors and Their Uses . Butterworth, Boston. p. 5-41 Bath, A. J., S. E. Milsom, N. A. Gormle y, and S. E. Halford. 2002. Many type IIs restriction endonucleases inte ract with two recognition sites before cleaving DNA. The Journal of Biological Chemistry . 277:4024-4033.
190 Belfort, M., and R. J. Roberts. 1997. Homing endonucleases: keeping the house in order. Nucleic Acids Research. 25:3379-3388. Bell, D. C., and C. G. Cupples. 2001. Very-short-patch repair in Escherichia coli requires the dam adenine methylase. Journal of Bacteriology . 183:3631-3635. Bellefontaine, A. F., C. E. Pierreux, P. Merten s, J. Vandenhaute, J. J. Letesson, and X. De Bolle. 2002. Plasticity of a transc riptional regulation network among alphaproteobacteria is supported by the identification of CtrA targets in Brucella abortus. Molecular Microbiology . 43:945-960. Berdis, A. J., I. Lee, J. K. Coward, C. Stephens, R. Wright, L. Shapiro, and S. J. Benkovic. 1998. A cell cy cle-regulated adenine DNA methyltransferase from Caulobacter crescentus processively methylates GANTC sites on hemimethylated DNA. Proceedings of the National Academy of Sciences USA. 95:2874-2879. Berger, S. L., and A. R. Kimmel (eds.). 1987. Methods in Enzymology: Guide to Molecular Cloning Techniques. Volume 152. Academic Press, Inc. Harcourt Brace Jovanovich. San Diego, California. Bertani, G., and J. J. Weigle. 1953. Host controlled variation in bacterial viruses . Journal of Bacteriology. 65:113-121. Bickle, T. A., C. Brack, and R. Yaun. 1978. ATP-induced conformational changes in the restriction endonuclease for Escherichia coli K-12. Proceedings of the National Academy of Sciences USA . 75:3099-3103. Bickle, T. A., and D. H. Kruger. 1993. Biology of DNA Restriction. Microbiological Review . 57:434-450. Birge, E. A. 1981. DNA restriction and m odification systems: 14: Gene splicing, the production of artificial DNA constructs. Bacterial and Bacteriophage Genetics . Springer-Verlag New York Inc. p. 314-319. Blatch, G. L., and M. Lassle. 1999. The tetr atricopeptide repeat: A structural motif mediating protein-protein interactions. BioEssays . 21:932-939. Bothast, R., N. N. Nichols, and B. S. Di en. 1999. Fermentations with new recombinant organisms. Biotechnology Progress . 15:867-875. Bourniquel A. A., and T. A. Bickle. 2002. Complex restriction enzymes: NTP-driven molecular motors. Biochimie . 84:1047-1059. Boye, E., and A. Lobner-Olesen. 1990. The role of dam methyltransferase in the control of DNA replication in E. coli. Cell . 62:981-989.
191 Brassinga, A. C., R. Siam, and G. T. Mar czynski. 2001. Conserved gene cluster at replication origins of the -proteobacteria Caulobacter crescentus and Rickettsia prowazekii . Journal of Bacteriology. 183:1824-1829. Bringer, S., R. K. Finn, and Sahm. 1984. Effect of oxygen on the metabolism of Zymomonas mobilis . Archives of Microbiology. 139:376-378. Brun, Y. V. 2001. Global analysis of a bact erial cell cycle: tracking down necessary functions and their regulators. TRENDS in Microbiology . 9:405-407. Bujnicki, J. M. 2001. Understanding the evol ution of restriction-modification systems: Clues from sequence and structure comparisons. Acta Biochimica Polonica. 48:935-967. Burrell, M. M. 1993. Enzymes of Molecular Biology . Humana Press, Inc. New Jersey. p. 107-211. Burt, A., and V. Koufopanou. 2004. Homing e ndonuclease genes: The rise and fall and rise again of a selfish element. Current Opinion in Genetics and Development. 14:609-615. Camacho, E. M., and J. Casadesus. 2002. Conj ugal transfer of the virulence plasmid of Salmonella enterica is regulated by the leucine-re sponsive regulatory protein and DNA adenine methylation. Molecular Microbiology . 44:1589-1598. Campbell, J. L., and N. Kleckner. 1990. E. coli oriC and the dnaA gene promoter are sequestered from dam Methyltransferase following the passage of the chromosomal replication fork. Cell . 2:967-979. Cantalupo, G., C. Bucci, P. Salvatore, C. Paglia rulo, V. Roberti, A. Lavitola, C. B. Bruni, and P. Alifano. 2001. Evolution and function of the Neisserial dam -replacing gene. FEBS Letters . 495:178-183. Cerritelli, S., S. S. Springhorn, and S. A. Lacks. 1989. DnpA, a methylase for singlestranded DNA in the DnpII restriction system, and its biological function. Proceedings of the National Academy of Sciences.USA . 86:9223-9227. Chen, S. L., and L. Shapiro. 2003. Identific ation of long intergenic repeat sequences associated with DNA methylation sites in Caulobacter crescentus and other proteobacteria. Journal of Bacteriology . 185:4997-5002. Chevalier, B. S, and B. L. Stoddard. 2001. Homing endonucleases: Structural and functional insight into the catalysts of intron/intein mobility. Nucleic Acids Research . 29:3757-3774. Chinen, A., Y. Naito, N. Handa, and I. Kobayashi. 2000. Evolution of sequence recognition by restriction-modi fication enzymes: Selective pressure for specificity decrease. Molecular Biology and Evolution . 17:1610-1619.
192 Chirikjian, J. G. 1987. Gene amplification and analysis. Restriction Endonucleases and Methylases. Volume 5. Elsevier Science Publishing Co., Inc., NY. NY. Cohen, S. N., A. C. Y. Chang, and L. Hsu. 1972. Nonchromosomal antibiotic resistance in bacteria: Genetic transformation of Escherichia coli by R-factor DNA. Proceedings of the National Academy of Sciences USA . 69:2110-2114. Cohen, S. N., A. C. Y. Chang, and H. W. Boyer. 1973. Construction of biologically functional bacterial plasmids in vitro. Proceedings of the National Academy of Sciences USA . 70:3240-3244. Conway, T. M. K Byun, and L. O. Ingr am. 1987a. Expression vector for Zymomonas mobilis. Applied and Environonmental Microbiology 53:2:235-241. Coton, M., J. M. Laplace, and E. Coton. 2005. Zymomonas mobilis subspecies identification by amplified ribosom al DNA restriction analysis. Letters in Applied Microbiology . 40:152-157. Crueger, W., and A. Crueger, (Brock, T.D. English, ed.). 1989. Biotechnology: A textbook of industr ial Microbiology . 2nd edition. Sinauer Associates, Inc., Sunderland, MA. p.124-127. Danna, K., and D. Nathans. 1971. Studies of SV40 DNA. 1. Specific cleavage of simian virus 40 DNA by restriction endonucleases of Hemophilus influenzae . Proceeding of the National Academy of Sciences of the United States of America. 68:29132917. Danner, H., and R. Braun. 1999. Biotec hnology for the produc tion of commodity chemicals from biomass. Chemical Society Reviews 28:395-405 De la Campa, A. G., P. Kale, S. S. Springhor n, and S. A. Lacks. 1987. Proteins encoded by the Dpn II restriction gene cassette: Two methylases and an endonuclease. Journal of Molecular Biology . 196:457-469. De la Campa, A. G., S. S. Springhorn, P. Kale , and S. A. Lacks. 1988. Proteins encoded by the Dpn I restriction gene cassette. Journal of Biological Chemistry . 263:14696-14702. Dien, B. S., M. A. Cotta, and T. W. Jeffries. 2003. Bacteria engineered for fuel ethanol production: Current status. Applied Microbiology and Biotechnology. 63:258266. DiPardo, J. 2000. Outlook for biomass ethanol production and demand. U.S. Department of Energy Web site. Retrieved August 20, 2000, from http://www.eia.doe.gov/oiaf/an alysispaper/biomass.html
193 Doelle, H. W., L. Kirk, R. Crittenden, H. Toh, and M. B. Doelle. 1993. Zymomonas mobilis Science and industrial application. Critical Reviews in Biotechnology . 13:57-98. Dreier, J., and T. A. Bickle. 1996. ATPa se activity of the type IC restrictionmodification system EcoR124II. Journal of Molecular Biology. 257:960-969. Dryden D. T. F., Noreen E. Murray, and D. N. Rao. 2001. Nucleoside triphosphatedependent restriction enzymes. Nucleic Acids Research . 29:3728-3741. Duine, J. A., J. Frank, and J. A. Jonge jan. 1986. PQQ and quinoprotein enzymes in microbial oxidations. FEMS Microbiology Letters . 32:165-178. Dussoix, D., and W. Arber. 1962. Host specificity of DNA produced by Escherichia coli. II: Control over acceptance of DNA from infecting lambda phage. Journal of Molecular Biology. 5:37-49. Engelberg-Kulka, H., and G. Glaser. 1999. Addiction modules and programmed cell death and antideath in bacterial cultures . Annual Review of Microbiology . 53:4370. Etanolrfa.org. 2005. Homegrown for the Homeland: Ethanol Industry Outlook 2005. RFA: Renewable fuels association Web site. Retrieved October 29, 2005, from http://www.ethanolrfa.org/objects/pdf/outlook2005.pdf ExPASy. (n.d.). Proteomics and Sequence an alysis tools. Expert Protein Analysis System. Proteomics server of the Swiss In stitue of Bioinformatics (SIB) Web site. Last accessed October, 2005 from http://www.expasy.ch/ Fomenkov, A., J. P. Xiao, D. Dila, E. Rale igh, and S. Y. Xu. 1994. The â€˜endo-blue methodâ€™ for direct cloning of restriction endonuclease genes in E. coli . Nucleic Acids Research . 22:2399-2403. Fuglsang, A. 2003. Distribution of potential type II restrict ion sites (palindromes) in prokaryotes. Biochemical and Biophysica l Research Communications . 310:280285. Glazer, A. N., and H. Nikaido. 1995. Microbial Biotechnology : Fundamentals of Applied Microbiology . W.H. Freeman and Company. New York. p. 69, 359-377. Glick, B. R., and J. J. Pasternak. 1998. Molecular Biotechnology . 2nd edition. American Society for Microbiolo gy. Washington, DC. Gogarten, P. J., A. G. Senejani, O. Zhaxybaye va, L. Olendzenski, and E. Hilario. 2002. Inteins: Structure, function, and evolution. Annual Review of Microbiology . 56:263-287.
194 Gonzalez-Nicieza, R., D. P. Turner, and B. A. Connolly. 2001. DNA binding and cleavage selectivity of the Escherichia coli DNA G:T-mismatch endonuclease (vsr protein). Journal of Molecular Biology . 310:501-508. Gromova, E. S., and A. V. Khoroshaev. 2003. Prokaryotic DNA methyltransferases: The structure and the mechanism of interaction with DNA. Molecular Biology . 37:260-272. Gunasekaran, P., and K. Chandra Raj. 1999. Ethanol fermentation technologyZymomonas mobilis. Current Science . 77:56-68. Hallez, R., A.F. Bellefontaine, J. J. Lete sson, and X. De Bolle. 2004. Morphological and functional asymmetry in -proteobacteria. TRENDS in Microbiology . 12:361-365. Handa, N., A. Ichige, K. Kusano, and I. Kobayashi. 2000. Cellular responses to postsegregational killing by re striction-modification. Genes . Journal of Bacteriology . 182:2218-2229. Heithoff, D. M., R. L. Sinsheimer, D. A. Lo w, and M. J. Mahan. 1999. An essential role for DNA adenine methylation in bacterial virulence. Science . 284:967-970. Heitman, J., and P. Model. 1987. Site speci fic methylases induce the SOS DNA repair response in Escherichia coli. Journal of Bacteriology . 169:3243-3250. Heitman, J., and P. Model. 1991. SOS induction as an in vivo assay of enzyme-DNA interactions. Gene . 103:1-9. Holubova, I., S. Vejsadova, M. Weiserova, a nd K. Firman. 2000. Localization of the type I restriction-modification enzyme Eco KI in the bacterial cell. Biochemical and Biophysical Research Communications . 270:46-51. Hyone-Myong, E. 1996. Chapter 4: Rest riction endonucleases and modification methylases. Enzymology Primer for Recombinant DNA Technology : Academic Press Inc. San Diego, California. p. 233-306. Ingram, L.O., C.K. Eddy, K.F. Mackenzi e, T. Conway, and F. Alterthum. 1989. Genetics of Zymomonas mobilis and ethanol production. Developments in Industrial Microbiology . 30:53-69. Ives, C. L., P. D. Nathan, and J. E. Brooks. 1992. Regulation of the Bam HI restrictionmodification systems by a small intergenic open reading frame, BamHIC, in both Escherichia coli and Bacillus subtilis. Journal of Bacteriology . 174:7194-7201. Ives, C. L., A. Sohail, and J. E. Brooks. 1995. The regulatory C prot eins from different restriction-modification systems can cross-complement. Journal of Bacteriology . 177:6313-6315.
195 Jackson, D., R. Symons, and P. Berg. 1972. Biochemical method for inserting genetic information into DNA of Simian virus 40; circular SV40 DNA molecules contain lambda phage genes and the galactose operon of E. coli . Proceedings of the National Academy of Sciences USA . 69:2904-2909. Jacob, F., F. Cuzin, and S. Brenner. 1963. On the regulation of DNA replication in bacteria. Cold Spring Harbor Symposia on Quantitative Biology . 28:329-348. Janosi, L., H. Yonemitsu, H. Hong, and A. Kaji. 1994. Molecular cloning and expression of a novel hydroxymethylcytosi ne-specific restriction enzyme ( Pvu Rts1I) modulated by glucosylation of DNA. Journal of Molecular Biology . 242:45-61. Jeltsch, A. 2002. Beyond Watson and Cric k: DNA methylation and molecular enzymology of DNA me thyltransferases. ChemBioChem . 3:274-293. Jeltsch, A. 2003. Maintenance of species iden tity and controlling speciation of bacteria: A new function for restriction/modification systems? Gene . 317:13-16. Joset, F., and J. Guespin-Michel. 1993. Prokaryotic Genetics: Genome Organization, Transfer and Plasticity . Blackwell Science Ltd. Oxford. Jost, J.P., and H.P. Saluz (ed.). 1993. DNA methylation: Molecular biology and biological significance . Birkhauser Verlag Basel/Switzerland. Julio, S. M., D. M. Heithoff, D. Provenzano, K. E. Klose, R. L. Sinsheimer, D. A. Low, and Michael J. Mahan. 2001. DNA adenine me thylase is essential for viability and plays a role in the pathogenesis of Yersinia pseudotuberculosis and Vibrio cholerae . Infection and Immunity . 69:7610-7615. Kahng, L. S., and L. Shapiro. 2001. The CcrM DNA methyltransferase of Agrobacterium tumefaciens is essential, and its activity is cell cycle regulated. Journal of Bacteriology. 183:3065-3075. Kalnenieks, U., A.A. De Graaf, S. Br inger-Meyer, H. Sahm. 1993. Oxidative phosphorylation in Zymomonas mobilis . Archives of Microbiology . 160:74-79. Kalnenieks, U., N. Galinina, M. M. Toma, a nd U. Marjutina. 2002. Ethanol cycle in an ethanologenic bacterium. FEBS Letters . 522:6-8. Kanhere, A., and M. Bansal. 2005. A novel me thod for prokaryotic promoter prediction based on DNA stability. BCM Bioinformatics . 6:1-10 Kobayashi, I. 2001. Behavior of restric tion-modification systems as selfish mobile elements and their impact on genome evolution. Nuclic Acids Research . 29:37423756.
196 Kossykh, V. G., and R. S. Lloyd. 2004. A DNA adenine methyltransferase of Escherichia coli that is cell cycle regulated and essential for viability. Journal of Bacteriology. 186:2061-2067. Kovach, M. E., P. H. Elzer, D. S. Hill, G. T. Robertson, M. A Farris, R. M. Roop II, and K. M. Peterson. 1995. Four new deriva tives of the broad-host-range cloning vector pBBR1MCS, carrying different antibiotic-resistance cassettes. Gene . 166:175-176 Kovall, R. A., and B. W. Ma tthews. 1998. Structural, functional, and evolutionary relationships between -exonuclease and the type II restriction endonucleases. Proceedings of the National Academy of Sciences. USA. 95:7893-7897. Kuhnlein, U., S. Linn, and W. Arber. 1969. Host specificity of DNA produced by Escherichia coli. Proceedings of th e National Academy of Sciences USA . 63:556562. Kulakauskas, S., A. Lubys, and S. D. Ehrlich. 1995. DNA restriction â€“modification systems mediate plasmid maintenance. Journal of Bacteriology . 177:3451-3454. Kulik, E. M. and T. A. Bickle. 1996. Regulation of the activity of the type IC Eco R124I restriction enzyme. Journal of Molecular Biology . 264:891-906. Laemmli, U. K. 1970. Cleavage of structural pr oteins during the assembly of the head of bacteriophage T4. Nature. 227:680. Lai, X., F. C. Davis, R. B. Hespell, a nd L. O. Ingram. 1997. Cloning of cellobiose phosphoenolpyruvate-dependent phosphotransfe rase genes: Functional expression in recombinant Escherichia coli and identification of a putative binding region for disaccharides. Applied and Environmental Microbiology . 63:355-363. Lark, C., and W. Arber. 1970. Ho st specificity of DNA produced by Escherichia coli. Journal of Molecular Biology. 52:337-348. Laub, M. T., S. L. Chen, L. Shapiro, and H. H. McAdams. 2002. Genes directly controlled by CtrA, a ma ster regulator of the Caulobacter cell cycle. PNAS . 99:4632-4637. Lee, W. C., and C. T. Huang. 2000. Modeling of ethanol fermentation using Zymomonas mobilis ATCC 10988 grown on the media containing glucose and fructose. Biochemical Engineering Journal. 4:217-227. Liang, C. C., and W. C. Lee. 1998. Characteristics and transformation of Zymomonas mobilis with plasmid pKT230 by electroporation . Bioprocess Engineering. 19:8185. Linn, S., and W. Arber. 1968. In vitro restriction of phage fd replicative form. Proceedings of the National Academy of Sciences USA. 59:1300-1305.
197 Lobner-Olesen, A., M. G. Marinus, and F. G. Hansen. 2003. Role of SeqA and Dam in Escherichia coli gene expression: A gl obal/microarray analysis. PNAS . 100:46724677. Low, D. A., N. J. Weyand, and M. J. Maha n. 2001. Minireview: Roles of DNA adenine methylation in regulating bacteria l gene expression and virulence . Infection and Immunity . 69:7197-7204 . Lu, M.. J. L. Campbell, and E. Boye. 1994. SeqA: A negative modulator of replication initiation in E. coli. Cell. 77:413-426. Luria, S. E., and M. L. Human. 1952. A nonhereditary, host-induced variation of bacterial viruses. Journal of Bacteriology . 64:557-569. Macintyre, G., K. M. J. Doiron, and C. G. Cupples. 1997. The Vsr endonuclease of Escherichia coli : An efficient DNA repair enzyme and a potent mutagen. Journal of Bacteriology . 179:6048-6052. Makovets, S., V. A. Doronina, and N. E. Murray. 1999. Regulation of endonuclease activity by proteolysis prevents breakage of unmodified bacterial chromosomes by type I restriction enzymes. Proceedings of the National Academy of Sciences USA. 96:9757-9762. Makovets, S., L. M. Powell, A. J. B. Tithera dge, G. W. Blakely, and N. E. Murray. 2004. Is modification sufficient to protect a bacterial chromosome from a resident restriction endonuclease? Molecular Microbiology . 51:135-147. Malone, T., R. M. Blumenthal, and X. Che ng. 1995. Structure-gui ded analysis reveals nine sequence motifs conserved among DNA amino-methyltransferases, and suggests a catalytic mechanism for these enzymes. Journal of Molecular Biology . 253:618-632. Marczynski, G. T. 1999. Ch romosome methylation and m easurement of faithful, once and only once per cell cycle ch romosome replication in Caulobacter crescentus . Journal of Bacteriology . 181:1984-1993. Marczynski, G. T., and L. Shapiro. 2002. Control of chromosome replication in Caulobacter crescentus . Annual Review of Microbiology . 56:625-656. Marinus, M. G. 2000. Recombination is essential for viability of an Escherichia coli dam (DNA adenine methyltransferase) mutant. Journal of Bacteriology . 182:463468 Matic, I., F. Taddel, and M. Radman. 1996. Genetic barriers among bacteria. Trends in Microbiology . 4:69-73.
198 Merkiene, E., G. Vilkaitis, and S. Kllimasa uskas. 1998. A pair of single-strand and double strand DNA cytosine-N4-methyltransferases from Bacillus centrosporus. Biological Chemistry. 379:569-571. Mertz, J., and R. W. Davis. 1972. Cl eavage of DNA by R1 endonuclease generates cohesive ends. Proceedings of the National Academy of Sciences USA . 69:33703374. Meselson, M., and R. Yuan. 1968. DNA restriction enzyme from Escherichia coli. Nature . 217:1110-1114. Microscopyu.com. (n.d.). Fluoresence Micr oscopy. Nikon MicroscopyU: The source for microscopy education Web site. Retrieved October 12, 2005, from http://www.microscopyu.com/articles/fluoresce nce/filtercubes/trip le/dapifitctritc/da pifitctritcindex.html Mohr, C. D., J. K. MacKichan, and L. Shap iro. 1998. A membrane -associated protein, FliX, is required for an early step in Caulobacter flagellar assembly . Journal of Bacteriology . 180:2175-2185. Montenecourt, B. S. 1985. Chapter 9: Zymomonas , a unique genus of bacteria. In Arnold L. Demain, and Nadine A. Solomon (eds.). Biology of Industrial Microorganisms . The Benjamin/Cummings Publishing Company, INC. Menlo Park, California. p.261-289. Morera, S., A. Imberty, U. Aschke-Sonnenborn, W. Ruger, and P. S. Freemont. 1999. T4 phage -glucosyltransferase: subs trate binding and proposed catalytic mechanism . Journal of Molecular Biology . 292:717-730. Murphy, D. Brendan, and J. Tony Pembroke. 1999. Monitoring of chromosomal insertions of the IncJ elements R391 and R997 in Escherichia coli K-12. FEMS Microbiology Letters . 174:355-361. Murphy, M., S. Nuoffer, and T. A. Bickle. 2002. Lack of regulation of the modificationdependent restricti on enzyme McrBC in Escherichia coli . Journal of Bacteriology . 184:1794-1795. Murray, N. E. 2000. Type I restriction syst ems: sophisticated molecular machines (a legacy of Bertani and Weigle). Microbiology and Molecu lar Biological Reviews (MMBR) . 64:412-434. Murray, N. E. 2002. 2001 Fred Griffith review lecture. Immigration control of DNA in bacteria: Self versus non-self. Microbiology . 148:3-20. Naito, T., K. Kusano, and I. Kobayashi. 1995. Selfish behavior of restrictionmodification systems. Science . 267:897-899.
199 Nakayama, Y., and I. Kobayashi. 1998. Re striction-modification gene complexes as selfish gene entities: Roles of a regul atory system in their establishment, maintenance, and apoptotic mutual exclusion. Proceedings of the National Academy of Sciences USA May. 95:6442-6447. Nasri, M., and D. Thomas. 1986. Relaxa tion of recognition sequence of specific endonuclease HindIII Nucleic Acid Research . 14:811-821. Naumann, T. A., and W. S. Reznikoff. 2002. Tn5 transposase with al tered specificity for transposon ends. Journal of Bacteriology. 184:233-240. NCBI ENTREZ. (n.d.). Entrez, The life scie nce search engine. United States National library of medicine (NLM). National instit utes of health (NIH). National center of biotechnological information (NCBI) Web site. Last accessed October, 2005, from http://www.ncbi.nlm.nih.gov/gquery/gquery.fcgi NCBI BLAST. (n.d.). Basi c Local Alignment Search T ool. United States National library of medicine (NLM). National instit utes of health (NIH). National center of biotechnological information (NCBI) Web site. Last accessed October, 2005, from http://www.ncbi.nlm.nih.gov/BLAST/ NCBI Taxonomy Browser. (n.d. ). United States National li brary of medicine (NLM). National institutes of he alth (NIH). National center of biotechnological information (NCBI) Web site. Last accessed October, 2005, from http://www.ncbi.nlm.nih.gov/Taxonomy/taxonomyhome.html/ NCBI ORF Finder. (n.d.). NCBI Tools for Data Mining: ORF Finder. United States National library of medicine (NLM). Nati onal institutes of health (NIH). National center of biotechnological information (N CBI) Web site. Last accessed October, 2005, from http://www.ncbi.nlm.nih.gov/Tools/index.html Neidhardt, F. C., J. L. Ingraham, and M. Schaechter. 1990. Physiology of the Bacterial Cell: A Molecular Approach . Sinauer associates, Inc. NetPrimer. (n.d.). NetPrimer: Primer design. Premier Biosoft International Web site. Retrieved September 20, 2005, from http://www.premierbiosoft.com/netprimer/index.html Nobelprize.org. (n.d.). The Nobel Prize web site. Retrieved September 26, 2005, from http://nobelprize.org/index.html Nwosu, V. U. 1992. Overexpression of the wildtype gene coding for Escherichia coli DNA adenine methylase ( dam ). Biochemical Journal . 283:745-750. Oâ€™ Neill, M., A. Chen, and N. E. Murray. 1997. The restriction-modification genes of Escherichia coli K12 may not be selfish: They do not resist loss and are readily replaced by alleles conferring different specificities. Proceedings of the National Academy of Sciences USA . 94:14596-14601.
200 Oâ€™Sullivan, D. J., K. Zagula, and T. R. Klaenhammer. 1995. In vivo restriction by Lla I is encoded by three genes, arranged in a n operon with llaIM , on the conjugative Lactococcus plasmid pTR2030. Journal of Bacteriology . 177:134-143. Panne, D., E. A. Raleigh, and T. A. Bickle. 1998. McrBs, a modulator peptide for McrBC activity. The EMBO Journal . 17:5477-5483. Patnaik, P., S. Merlin, and B. Polisky. 1990. Effect of altering GATC sequences in the plasmid ColE1 primer promoter. Journal of Bacteriology . 172:1762-1768. Peterson, L.A., and D.M. Downs. 1997. Iden tification and characte rization of an operon in Salmonella typhimurium involved in thiamine biosynthesis. Journal of Bacteriology . 179:4894-4900. Piekarowicz, A., R. Yaun, and D. C. Stein. 1991a. Isola tion of temperature-sensitive McrA and McrB mutations and complement ation analysis of the McrBC region of Escherichia coli K12. Journal of Bacteriology . 173:150-155. Piekarowicz, A., R. Yaun, and D. C. Stein. 1991b. A new method for the rapid identification of genes encoding rest riction and modification enzymes. Nucleic Acids Research. 19:1831-1835. Pieper, U., D. H. Groll, S. Wunsch, F. U. Gast, C. Speck, N. Mucke, and A. Pingoud. 2002. The GTP-dependent restriction enzyme McrBC from Escherichia coli forms high-molecular mass complexes with DNA a nd produces a cleavage pattern with a characteristic 10-base pair repeat. Biochemistry . 41:5245-5254. Pieper, U., and A. Pingoud. 2002. A mutati onal analysis of the PDD/EXK motif suggests that McrC harbors the cataly tic center for DNA cleavage by the GTPdependent restriction enzyme McrBC from Escherichia coli . Biochemistry . 41:5236-5244. Pingoud, V., E. Kubareva, G. Stengel, P. Fr iedhoff, J. M. Bujnicki, C. Urbanke, A. Sudina, and A. Pingoud. 2002. Evolutiona ry relationship between different subgroups of restriction endonucleases. The Journal of Biological Chemistry . 277:14306-14314. Poindexter, J. S. 1981. The Caulobacters : Ubiquitous unusual bacteria. Microbiological Reviews . 45:123-179. Posfai, G., and W. Szybalski. 1988. A simp le method for locating methylated bases in DNA using class-IIs rest riction enzymes. Gene . 74:179-181. Powell, L. M., B. A. Connolly, and D. T. F. Dryden. 1998. The DNA binding characteristics of the trimeric EcoKI methyltransferase and its partially assembled dimeric form determined by fluorescen ce polarisation and DNA footprinting. Journal of Molecular Biology . 283:947-961.
201 Prakash-Cheng, A., S. S. Chung, and J. R yu. 1993a. The expression and regulation of hsdk genes after conjugative transfer. Molecular and General Genetics . 241: 491496. Prakash-Cheng, A., and J. Ryu. 1993b. Delayed expression of in vivo restriction activity following conjugal transfer of Escherichia coli hsd k (restriction-modification) genes. Journal of Bacteriology . 175:4905-4906. Quon, K. C., G. T Marczynski, and L. Shapiro. 1996. Cell cycle control by an essential bacterial two-component signal transduction protein. Cell . 84:83-93. Quon, K. C., B. Yang, I. J. Domian, L. Shapir o, and G. T. Marczynski. 1998. Negative control of bacterial DNA repli cation by a cell cycle regulato ry protein that binds at the chromosome origin. Proceedings of the National Academy of Sciences USA. 95:120-125. Raghavendra, N. K., and D. N. Rao. 2003. Functional c ooperation between exonucleases and endonucleases-basis for th e evolution of restriction enzymes. Nucleic Acids Research . 31:1888-1896. Raleigh, E., and J. E. Brooks. 1998. Restric tion modification systems: Where they are and what they do? In Frans J. De Bruijn, J. R. Lupski, and G. M. Weinstock (eds.). Bacterial Genomes: Physical Structure and Analysis . Chapman and Hall. NY. NY. p. 78-91. Rao, D. N., S. Saha, and V. Krishnamurthy. 2000. ATP-dependent restriction enzymes. Progress in Nucleic Acid Research and Molecular Biology . 64: 1-63 Reisenauer, A., K. Quon, and L. Shapiro. 1999a. The CtrA response regulator mediates temporal control of gene expression during the Caulobacter cell cycle. Journal of Bacteriology . 181:2430-2439. Reisenauer, A., L. S. Kahng, S. McCollum, and L. Shapiro. 1999b . Bacterial DNA methylation: a cell cycle regulator? Journal of Bacteriology . 181:5135-5139. Reisenauer, A., and L. Shapiro. 2002. DNA methylation affects the cell cycle transcription of the Ct rA global regulator in Caulobacter . The EMBO Journal . 21:4969-4977. Rimseliene, R., R. Vaisvila, and A. Janulaitis. 1995. The eco72IC gene specifies a transacting factor which influences expres sion of both DNA methyltransferase and endonuclease from the Eco72I rest riction-modification system. Gene . 157:217-219. Roberts, R. J., and D. Macelis. 1996. REBASE-Restriction enzymes and methylases. Nucleic Acids Research 24:1:223-235. REBASE: The restriction enzyme Database Web site. Last accessed October, 2005, from http://rebase.neb.com
202 Roberts, R. J., M. Belfort, T. Bestor, A. S. Bhagwat, T. A. Bickle, J. Bitinaite, R. M. Blumenthal, S. K. Degtyarev, D. T. F. Dryden, K. Dybvig, K. Firman, E. S. Gromova, R. I. Gumport, S. E. Halford, S. Hattmann, J. Heitman, D. P. Hornby, A. Janulaitis, A. Jeltsch, J. Josephsen, A. Kiss, T. R. Klaenhammer, I. Kobayashi, H. Kong, D. H. Krger, S. Lacks, M. G. Mari nus, M. Miyahara, R. D. Morgan, N. E. Murry, V. Nagaraja, A. Piekarowicz, A. Pingoud, E. Raleigh, D. N. Rao, N. Reich, V. E. Repin, E. U. Selker, P. C. Shaw, D. C. Stein, B. L. Stoddard, W. Szybalski, T. A. Trautner, J. L. Van Etten, J. M. B. Vitor, G. G. Wilson, and S. Y. Xu. 2003. A nomenclature for restriction enzy mes, DNA methyltransferases, homing endonucleases and their genes. Nucleic Acids Research . 31:1805-1812. Robertson, G. T., A. Reisenauer, R. Wright, R. B. Jensen, A. Jensen, L. Shapiro, and R. M. Roop II. 2000. The Brucella abortus CcrM DNA methyltransferase is essential for viability, and its overexpression attenua tes intracellular replication in murine macrophages. Journal of Bacteriology . 182:3482-3489. Rydberg, B., and J. Game. 2002. Excision of misincorporated ribonucleotides in DNA by RNase H (type 2) and FEN-1 in cell-free extracts. Proceedings of the National Academy of Sciences USA. 99:16654-16659. Sadykov, M., Y. Asami, H. Niki, N. Handa, M. Itaya, M. Tanokura, and I. Kobayashi. 2003. Multiplication of a restric tion-modification gene complex. Molecular Microbiology . 48:417-428. Sahm, H., S. Bringer-Meyer, and G. Sp renger. 1992. Chapter 112: The genus Zymomonas . In A. Belows, H. G. Truper, M. Dworkin, W. Harder, K. H. Schleifer (eds.). The Prokaryotes. Volume 3, 2nd edition. Springer, New York Berlin Heidelbeg. p. 2287-2301. Sain, B., and N. E. Murray. 1980. The hsd (Host specificity) Genes of E. coli K12. Molecular and General Genetics. 180:35-46. Sambrook, J., E. F. Fritsch, and T. Maniatis. 1989. Molecular cloning--a laboratory manual , 2nd edition. Cold Spring Harbor, NY. Cold Spring Harbor Laboratory Press. Sapranauskas, R., G. Sasnauskas, A. Lagunavicius, G. Vilkaitis, A. Lubys, and V. Siksnys. 2000. Novel subtype of type IIs restriction enzymes: BfiI endonuclease exhibits similarities to the EDTA-resistant nuclease Nuc of Salmonella typhimurium. The Journal of Biological Chemistry . 275:30878-30885. Scopes, R. K., and K. Griffiths-Smit h. 1986. Fermentation capabilities of Zymomonas mobilis glycolytic enzymes. Biotechnology Letters . 8:653-656.
203 Seo, J. S., H. Chong, H. S. Park, K. O. Yoon, C. Jung, J. J. Kim, J. H. Hong, H. Kim, J. H. Kim, J. I Kil, C. J. Park, H. M. Oh, J. S. Lee, S. J. Jin, H. W. Um, H. J. Lee, S. J. Oh, J. Y. Kim, H. L. Kang, S. Y. Lee, K. J. Lee, and H. S. Kang. 2005. The genome sequence of the ethanologenic bacterium Zymomonas mobilis ZM4. Nature Biotechnology . 23:63-68. Shier, V, K., C, J. Hancey, and S, J. Benkovic. 2001. Identification of the active oligomeric state of an essential adenine DNA methyltransferase from Caulobacter crescentus . Journal of Biological Chemistry . 276:14744-14751. Silveira, M. M., and R. Jonas. 2002. Th e biotechnological produc tion of sorbitol. Applied Microbiology and Biotechnology . 59:400-408. Smith, D. I., F. R. Blattner, and J. Davies. 1976. The isolation and partial characterization of a new re striction endonuclease from Providencia stuartii. Nucleic Acids Research. 3:343-353. Smith, H. O., and K. W. Wilcox. 197 0. A restriction enzyme from Haemophilus influenzae . I: Purification and general properties. Journal of Molecular Biology . 51:379-391. Smith, H. O., and D. Nathans. 1973. A s uggested nomenclature for bacteria host modification and restriction systems and their enzymes. Journal of Molecular Biology . 81:419-423. Smith, R. F., B. A. Wiese, M. K. Wojzynski , D. B. Davison, and K. C. Worley. 1996. BCM Search Launcher--An Integrated In terface to Molecular Biology Data Base Search and Analysis Services Av ailable on the World Wide Web. Genome Research. 6:5:454-62. BCM (Baylor College of medicine) Search Launcher: Sequence utilities, Web site. La st accessed October, 2005, from http://dot.imgen.bcm.tmc.edu:9331/seq-util/seq-util.html Stephens, C. M., G. Zweiger, and L. Shapir o. 1995. Coordinate ce ll cycle control of a Caulobacter DNA methyltransferase and the flagellar genetic hierarchy. Journal of Bacteriology . 177:1662-1669. Stephens, C. M., A. Reisenauer, R. Wright, and L. Shapiro. 1996. A cell cycle-regulated bacterial DNA methyltransferase is essential for viability. Proceedings of the National Academy of Sciences USA. 93:1210-1214. Strohdeicher, M., B. Schmitz, S. Bringer-M eyer, and H. Sahm. 1988. Comparison of plasmids in strains of Zymomonas mobilis . Plasmid . 9:138-146. Sun, D. K., and O. J. Yoo. 1988. Purifica tion and characterization of a restriction endonuclease ZanI from Zymomonas anaerobia . Korean Biochem . J. 21:419-422. Swings, J., and J. De Ley. 1977. The biology of Zymomonas. Bacteriological Reviews . 41:1-46.
204 Szczelkun, Mark D. 2002. Kinetic models of translocation, head-on collision, and DNA cleavage by Type I restri ction endonucleases. Biochemistry . 41:2067-2074. Szybalski, W., S. C. Kim, N. Hasan, and A. J. Podhajska. 1991. Class-IIS restriction enzymes â€“ A review. Gene. 100:13-26. Tao, T., and R. M. Blumenthal. 199 2. Sequence and characterization of pvuIIR , the Pvu II endonuclease gene, and the pvuIIC, its regulatory gene. Journal of Bacteriology . 174:3395-3398. Tibbetts, A. S., and D. R. Appling. 1997. Saccharomyces cerevisiae expresses two genes encoding isozymes of 5-aminoi midazole-4-carboxamide ribonucleotide transformylase. Archives of Biochemistry and Biophysics . 15:340:2:195-200. Timinskas, A., V. Butkus, and A. Janulaitus. 1995. Sequence motifs characteristic for DNA [cytosine-N4] and DNA [ade nine-N6] methyltransferas es. Classification of all DNA methyltransferases. Gene . 157:3-11. Titheradge, A. J. B., J. King, J. Ryu, and N. E. Murray. 2001. Families of restriction enzymes: an analysis prompted by molecular and genetic data for Type ID restriction and modification systems. Nucleic Acids Research . 29:4195-4205. Troester, H., S. Bub, A. Hunziker, and M. F. Trendelenburg. 2000. Stability of DNA repeats in Escherichia coli dam mutant strains indica tes a Dam methylationdependent DNA deletion process. Gene. 258:95-108. Veeramallu, U., and P. Agrawal. 1988. Study of fermentation behavior and formulation of a semidefined nutrient medium ba sed on acid-production measurements in Z. mobilis cultures. Biotechnology and Bioengineering . 31:770-782. Vijesurier, R. M., L, Carlock, R, M. Blumen thal, and J, C. Dunbar. 2000. Role and mechanism of action of C. Pvu II, a regulatory protein conserved among restrictionmodification systems. Journal of Bacteriology . 182:477-487. Voet, D., and J, G. Voet. 1995. Biochemistry . 2nd edition. John Wiley & sons, Inc. p. 337, 466-468, 543, 618-620, 740. Waite-Rees, P. A., C. J. Keating, L. S Mora n, BE. Slatko, L. J. Hornstra, and J. S. Benner. 1991. Characterizati on and expression of the Escherichia coli Mrr restriction system. Journal of Bacteriology . 173:5207-5219. Weyand, N. J., and D. A. Low. 2000. Regul ation of Pap phase variation: Lrp is sufficient for the establishm ent of the phase off pap DNA methylation pattern and repression of pap transcription in vitro . Journal of Biological Chemistry . 275:3292-3200.
205 Williams, D. L., L. Spring, E Harris, P. Roch e, and T. P. Gillis. 2000. Dihydropteroate synthase of Mycobacterium leprae and dapsone resistance. Antimicrobial Agents and Chemotherapy . 44:1530-1537. Williams, R, J. 2003. Restriction endonucl eases: Classification, properties, and applications. Molecular Biotechnology . 23:225-243. Willke, T., and K. D. Vorlop. 2004. Industria l bioconversion of renewable resources as an alternative to conventional chemistry. Applied Microbiology and Biotechnology. 66:131-142 Wilson, G, G. 1991. Organization of restriction-modification systems. Nucleic Acids Research. 19:2539-2566 Wright, R., C. Stephens, G. Zweiger, L. Shapiro, and M. R. Alley. 1996. Caulobacter Lon protease has a critical role in ce ll-cycle control of DNA methylation. Genes and Development . 10:1532-1542. Wright, R., C. Stephens, and L. Shapiro. 1997. The CcrM DNA methyltransferase is widespread in the alpha subdivision of pr oteobacteria, and its essential functions are conserved in Rhizobium meliloti and Caulobacter crescentus . Journal of Bacteriology . 179:5869-5877. Yablonsky, M. D., A. E. Goodma n, N. Stevnsborg, O. Goncalves de Lima, J. O. Falcao de Morais, H. G. Lawford, P. L. Rogers, and D. E. Eveleigh. 1988. Zymomonas mobilis CP4: A clarification of st rains via plasmid profiles. Journal of Biotechnology . 9:71-80. Zhang, M., C. Eddy, K. Deanda, M. Finkelste in, and S. Picataggio. 1995. Metabolic engineering of a pentose meta bolism pathway in ethanologenic Zymomonas mobilis. Science . 267:240-243. Zhuang, W. Y., and L. Shapiro. 1995. Caulobacter FliQ and FliR membrane proteins, required for flagellar biogenesis and cell di vision, belong to a family of virulence factor export proteins. Journal of Bacteriology . 177:343-356. Zweiger, G., G. Marczynski, and L. Shapiro. 1994. A Caulobacter DNA methyltransferase that functions only in the predivisional cell . Journal of Molecular Biology . 235:472-485.
206 BIOGRAPHICAL SKETCH I was born on a military base in Fort Knox, Kentucky, in June 1972. Having a father in the military, I spent my childhood tr aveling all over the world attending various schools in the U.S. and abroad. While we we re stationed at Fort Stewart, Georgia, I completed my secondary schooling at Hinesv ille Institute, graduating with honors. While attending the University of Florida as an undergraduate, with the original intention of applying to medical school upon graduating, I enrolled in undergraduate research at the Department of Microbiology a nd Cell Science. My se nior research thesis for graduating with high honors was based on my project goal to crea te a restriction map of the Zymomonas mobilis CP4 genome. My experiences in undergraduat e research in Francis C. Davisâ€™s laboratory nurtured my inte rest in molecular science and prompted a change in my career goal of becoming a Doct or of Medicine to becoming a Doctor of Philosophy in microbiology. Upon graduating Magna Cum laude in 1996 from the University of Florida with a B.S degree in both zoology and microbiology, I applied to graduate school and was accepted to the Depa rtment of Microbiology and Cell Science in the College of Agricultural and Life Sciences. I returned to Franci s C. Davisâ€™s laboratory in order to continue the research project I st arted as an undergraduate. Since then, I faced many challenges both personal and professional. I started a new research project to identify DNA modification and restriction systems in the bacterium Zymomonas mobilis, which was not my primary focus until anot her research group comp leted and published a restriction map of Zymomonas mobilis ZM4 before I completed my original project.
207 After successfully passing the comprehens ive written and oral examinations and becoming a candidate for the Doctor of Philosophy degree in the fall of 2000, my husband was involved in a serious motorcycle accident that required a year of physical rehabilitation, taking a great d eal of my time and energy away from graduate research. After his recovery, there was a period when my research became par ticularly challenging; however, I eventually began to make real progres s in my graduate project. With so much time and effort invested, I was determined to obtain a Doctor of Philosophy degree. Fortunately, I had the encouragement and support of my family and my mentor. I am happily married to Alexander Gram s, my high school sweetheart, who has also been my best friend for 17 years. We have a three year old daughter, Veronica, who gave me a whole new perspective on the important things in life.