PLACE IN REl'URN BOX to remove this checkout from your record. TO AVOID FINES return on or before date due. MAY BE RECALLED with earlier due date if requested. DATE DUE DATE DUE DATE DUE 1/” m“ Evaluating Diazotrophy, Diversity, and Endophytic Colonization Ability of Bacteria Isolated from Surface- Sterilized Rice By Jon R. Stoltzfus A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Genetics 1999 ABSTRACT EVALUATING DIAZOTROPHY, DIVERSITY, AND ENDOPHYTIC COLONIZATION ABILITY OF BACTERIA ISOLATED FROM SURFACE-STERILIZED RICE By Jon R. Stoltzfus Information about the nitrogen fixing potential, diversity and sites of colonization of endophytic bacteria from rice are needed to expand our understanding of a new area of microbial ecology and plant microbe interactions and lay the foundations for future studies aimed at using biologically fixed nitrogen to replace nitrogen fertilizers. A collection of 142 bacteria isolated from mechanically-abraded, surface-sterilized rice roots was studied. Polymerase Chain Reaction (PCR) mediated gene amplification using degenerate primers derived from highly conserved regions of the nitrogenase nifD gene revealed 20 isolates harboring nifD gene sequences. Southern hybridization analysis confirmed the presence of nif genes in 19 of these isolates. The diazotrophic nature of these 19 isolates was confirmed using Acetylene Reduction Assays (ARA). Examination of genetic diversity using amplified ribosomal DNA restriction analysis (ARDRA) fingerprints and rep-PCR genomic fingerprints with Gelcompar software revealed 56 unique ARDRA fingerprints and 71 unique rep-PCR genomic fingerprints. Clusters of similar combined fingerprints, consisting of 37, 15, 12, and 9 non-diazotrophic bacteria, as well as two clusters each containing 4 diazotrophic bacteria, were found. Analysis of partial Small Subunit (SSU) ribosomal RNA (rRNA) gene sequences revealed the presence of isolates with similarity to strains from the alpha-, beta-, and gamma subdivisions of the Proteobacteria, and to members of the Bacillaceae and Microbacteriaceae. Many of the ARDRA fingerprints and/or SSU rRNA gene sequences of these bacteria were highly similar to those of other bacteria previously isolated from the rhizosphere of rice. Two isolates from the collection and Sinorhizobium meliloti, a control, were tagged with the biomarker gus or gfp, and the colonization of rice tissue was examined. In situ visualization of colonization of three week old inoculated rice seedlings revealed no endophytic colonization of rice tissue by these bacteria. However, clumps of bacteria, as well as individual cells, could be visualized on the surface of the roots. On very rare occasions an isolated epidermal cell filled with bacteria was observed. DEDICATION To all the bacteria and trees that gave their lives to produce this thesis. Iw< ACKNOWLEDGMENTS I would like to thank Professor Frans J. de Bruijn for supporting me as I conducted independent research in his lab. I wish in addition to thank my committee members, Tom Schmidt, Lee Macintosh, and Frank Dazzo for their helpful advice and comments. | wish to thank our collaborator, J.K. Ladha, for inviting me to spend time in his laboratory at IRRI in the Philippines. Thanks are also due past and present members of the de Bruijn lab, especially Silvia Rossbach, Frank Louws, Jan Rademaker, Krzysztof Szczyglowski, Philipp Kapranov , and Mary Ellen Davey, for their helpful advice and camaraderie. I wish to thank Shirley Owens for her help using the LSCM. Finally, I thank my wife for tolerating my foul moods during the writing of this thesis. TABLE OF CONTENTS LIST OF TABLES ........................................................................................... Xi LIST OF FIGURES ......................................................................................... xii CHAPTER 1 INTRODUCTION ......................................................................... 1 IMPORTANCE OF NITROGEN IN RICE PRODUCTION ................................ 2 DIAZOTROPHIC BACTERIA ASSOCIATED WITH RICE .............................. 2 SUPPLYING NITROGEN TO RICE .............................................................. 6 DEFINITION OF ENDOPHYTIC BACTERIA .............................................. 12 ENDOPHYTIC BACTERIA ASSOCIATED WITH PLANTS ......................... 14 ENDOPHYTIC BACTERIA ASSOCIATED WITH RICE ............................... 14 SUMMARY ............................................................................................ 1 8 REFEREI\CES ........................................................................................ 20 CHAPTER 2 ISOLATION OF ENDOPHYTIC BACTERIA FROM RICE AND ASSESSMENT OF THEIR POTENTIAL FOR SUPPLYING RICE WITH BIOLOGICALLY FIXED NITROGEN ................................................................ 28 ABSTRACT ........................................................................................... 28 INTRODUCTION .................................................................................... 30 MATERIAL AND METHODS ................................................................... 37 Isolation of Endophytic Bacteria. ..................................................... 37 Acetylene Reduction Assays ........................................................... 39 PCR Amplification of nifD-Specific DNA Fragments ......................... 40 rep-PCR Genomic Fingerprinting ...................................................... 41 16S rRNA Gene Sequencing and Phylogenetic Studies ................... 41 Introduction of Transposons into Putative Endophytes and Detection of GUS Activity ............................................................... 41 RESULTS AND DISCUSSION ................................................................. 43 Isolation of Rice Endophytes ........................................................... 43 Diversity of Putative Endophytes in Rice Tissues ............................ 43 Nitrogen Fixation by and Presence of nif Genes in the Endophytic Isolates .......................................................................... 45 Composition of a “Test Collection” and Preliminary Characterization of its Members ..................................................... 46 Re-colonization of Rice Tissue by Putative Endophytes .................. 47 vi Use of Marker Genes to Track Endophytic Bacteria in Rice Tissues ............................................................................................. 49 Do the Proper Physiological Conditions Exist in Rice Tissues to De-repress nif Genes Harbored by Endophytic Bacteria? ................ 51 CONCLUSIONS ..................................................................................... 53 ACKNOWLEDGMENTS .......................................................................... 56 REFEREMDES ........................................................................................ 64 CHAPTER 3 DEVELOPMENT AND USE OF POLYMERASE CHAIN REACTION (PCR) PRIMERS FOR THE RAPID DETECTION OF NITROGEN- FlXING BACTERIA ........................................................................................ 70 ABSTRACT ........................................................................................... 70 INTRODUCTION .................................................................................... 71 MATERIAL AND METHODS ................................................................... 76 Bacterial Strains ............................................................................... 76 Media and Growth Conditions. ......................................................... 76 PCR Template Preparation and Amplification .................................. 76 nifD—PCR Primer Design .................................................................... 77 DNA Analyses of nifD-PCR Products ................................................ 78 Southern Hybridization Analysis of Total Genomic DNA ................. 79 Acetylene Reduction Assays ........................................................... 80 Growth, Sterilization and Inoculation of Rice Seedlings. ................. 81 RESULTS .............................................................................................. 83 PCR Amplification of mm Genes ..................................................... 83 Characterization of PCR Fragments Using DNA Sequencing and Southern Hybridization Analysis ............................................... 85 Screening a Collection of Rice Associated Bacteria for the Presence of mm DNA Sequences .................................................... 87 Confirming the Presence of nif Genes by Southern Hybridization of Total Genomic DNA ............................................... 87 Acetylene Reduction Capacity as an Indicator of Nitrogen Fixing Activity .................................................................................. 88 Direct Detection of Rice-Associated Nitrogen-Fixing Bacteria ........ 89 DISCUSSION ......................................................................................... 91 nifD-PCR Primers for Rapid and Specific Identification of Diazotrophic Bacteria. ..................................................................... 91 The nifD Primer Set Rapidly Identifies Diazotrophic Bacteria in Large Collections of Environmental Isolates .................................... 94 NifD-PCR Detects Novel Diazotrophs ............................................... 95 vii (I! nifD-PCR Directly Detects Diazotrophs in Environmental Samples ............................................................................................ 95 CONCLUSIONS ..................................................................................... 96 FEFEFBICES ..................................................................................... 108 CHAPTER 4 DIVERSITY OF A COLLECTION OF BACTERIA ISOLATED FROM SURFACE-STERILIZED RICE TISSUE ................................................ 1 12 ABSTRACT ........................................................................................ 112 INTRODUCTION ................................................................................. 113 MATERIAL AND METHODS ................................................................ 115 Bacterial Strains and Growth Conditions ...................................... 115 PCR Reactions ............................................................................... 115 ARDRA and SSU Sequencing ......................................................... 116 Sesbania rostrata Plant Growth and Inoculation ......................... 117 Southern Hybridization Analysis of Total Genomic DNA .............. 117 Computer Assisted Pattern Analyses ........................................... 118 RESULTS ........................................................................................... 119 Phylogenetic Diversity of Bacteria Isolated from Surface- sterilized Rice Plants ..................................................................... 119 SSU rRNA Gene Sequence Similarities .......................................... 121 Characterization of Clusters of Bacteria with Similar Fingerprints ................................................................................... 1 22 Characterization of Clusters Containing Non-Diazotrophic Bacteria ......................................................................................... 122 Characterization of Clusters Containing Diazotrophic Bacteria... 127 DISCUSSION ...................................................................................... 131 Genetic Diversity of Putative Endophytic Bacteria ....................... 131 Specificity of Rice Plant/Bacterial Interactions ............................ 131 SSU rRNA Gene DNA Sequence Analysis ...................................... 133 Similarities to Endophytic Bacteria ............................................... 133 Similarities to Rice Rhizosphere Bacteria ...................................... 135 Similarities to Legume Symbionts ................................................ 136 Other SSU rRNA Gene Similarities ................................................. 138 Obligate and Facultative Endophytes ........................................... 139 Pathogenic and Endophytic Bacteria ............................................ 141 CONCLUSIONS .................................................................................. 142 PEFEFECES ..................................................................................... 159 CHAPTER 5 USING GFP AS A BIOMARKER FOR MICROBIAL ECOLOGY ..... 166 INTRODUCTION ................................................................................. 166 viii l.— OP! 8.“ JJ (’ GENETIC IMPROVEMENT OF GFP ....................................................... 1 67 Wild-Type GFP (thFP) ................................................................. 167 Disadvantages of thFP ............................................................... 168 IMPROVING GFP FLUORESCENCE ...................................................... 1 69 Background ................................................................................... 1 69 Mutants of GFP ............................................................................. 169 GFP AS A BIOMARKER/BIOREPORTER IN BACTERIA: GENERAL CONSIDERATIONS ............................................................................. 173 Advantages of GFP ....................................................................... 173 Problems with GFP ........................................................................ 175 Detection of GFP ........................................................................... 177 Delivery and Maintenance of the gfp gene ................................... 179 GFP AS A BIOMARKER IN BACTERIA ................................................. 179 Tracking Bacteria in the Environment ........................................... 180 Localizing Bacteria in situ ............................................................. 181 Monitoring Gene Expression in the Environment .......................... 182 GFP AS A BIOREPORTER IN BACTERIA .............................................. 183 Localizing Protein Fusions ............................................................. 183 Monitoring Gene Expression .......................................................... 186 Isolating Environmentally Regulated Genes .................................. 187 CONCLUSIONS .................................................................................. 189 REFEREhDES ..................................................................................... 209 CHAPTER 6 STUDIES ON RICE ROOT COLONIZATION BY SELECTED BACTERIA USING GFP AND GUS AS MOLECULAR BIOMARKERS .............. 220 ABSTRACT ........................................................................................ 220 INTRODUCTION ................................................................................. 222 MATERIAL AND METHODS ................................................................ 225 Bacterial Strains and Growth Conditions ...................................... 225 Generation of Sterile Rice Seedlings, Bacterial Infection and Re-isolation ................................................................................... 226 Generation of Sterile Rice Seedlings and Bacterial Infection for Microscopic Studies ...................................................................... 227 Construction of gfp and gus Biomarker Transposons .................. 228 Introduction of Biomarker Transposons into Bacteria. ................. 229 Histochemical Staining for Gus Activity and Microscopy ............. 230 Fluorescent Microscopy ................................................................ 230 RESULTS ........................................................................................... 232 Re-lsolation of Bacteria from Infected Rice Seedlings .................. 232 Tagging Bacteria with Molecular Biomarkers ................................ 232 In Situ Microscopic Evaluation of the Colonizing Ability of Biomarker Tagged Strains ............................................................. 234 DISCUSSION ...................................................................................... 238 Re-isolation of Bacteria from Infected Rice Seedlings .................. 238 Generation of Sterile Rice Seedlings and Bacterial Infection for Microscopic Studies ...................................................................... 239 In Situ Microscopic studies on Colonizing Ability of the Tagged Stains ............................................................................... 240 Problems with GF P as a Molecular Biomarker ............................... 241 Colonization by S. meli/oti 1021 , R33(120) and R100(64) ........ 244 Spatial Localization Using Three Dimensional Reconstructions.... 246 CONCLUSIONS .................................................................................. 248 FEFEPEADES ..................................................................................... 261 CHAPTER 7 SUMMARY AND CONCLUSIONS ............................................ 265 DIVERSITY OF RICE ISOLATES ........................................................... 265 UTILITY OF nifDPCR ......................................................................... 266 UTILITY OF GUS AND GFP AS COLONIZATION BIOMARKERS ............ 266 COLONIZATION OF SEEDLINGS BY RICE ISOLATES ........................... 266 APPENDIXI .............................................................................................. 268 LIST OF TABLES Table 2-1 Partial SSU rDNA Similarities ..................................................... 57 Table 3-1 Bacterial Strains and PCR Fragments Generated with the nifD Primer Set dB261/dB260. ............................................... 99 Table 3-2 Results of nifD-PCR, Southern Hybridization and ARA experiments for 35 Rice Isolates .......................................... 101 Table 4-1 Clusters of Isolates with Similar Combined Fingerprints ........ 144 Table 4-2 SSU rRNA Gene Sequence Similarities .................................... 149 Table 5-1 GFP Mutants ........................................................................... 191 Table 5-2 GFP Vectors ........................................................................... 202 Table 5-3 GFP Marked Bacteria .............................................................. 208 Table 6-1 Bacterial Strains ..................................................................... 250 Table 6-2 Plasmids .................................................................................. 251 xi LIST OF FIGURES Figure 2-1 Isolation of Putative Rice Endophytes ..................................... 58 Figure 2-2 Isolation of Endophytic Bacteria Using a Re-infection Step ......................................................................................... 59 Figure 2-3 Gel Profiles and Dendrogram of rep-PCR Generated Genomic Fingerprints of Putative Endophytic Bacteria Isolated from Rice Roots .......................................................... 60 Figure 2-4 rep-PCR Genomic Fingerprints and nifD-PCR Analysis of Putative Diazotrophic Endophytes .......................................... 61 Figure 2-5 Re-infection Potential of Putative Endophytes ........................ 62 Figure 2-6 In planta Visualization of Endophyte Colonization in Rice ....... 63 Figure 3-1 Amino Acid Alignment of Mo-Fe Protein Sequences ............ 102 Figure 3-2 nifD-PCR Primers ................................................................... 103 Figure 3-3 nifD-PCR Products Visualized by Agarose Gel Electrophoresis and Identified by Southern Hybridization 104 Figure 3-4 Agarose Gel Electrophoresis of nifD-PCR Products from Rice Isolates .......................................................................... 105 Figure 3-5 Southern Blot of Bacterial DNA Probed with nifHDK. .......... 106 Figure 3-6 Agarose Gel Electrophoresis of nifD-PCR Products from Rice Tissues .......................................................................... 107 Figure 4-1 Dendrogram Showing Genetic Diversity of Rice Isolates ...... 151 Figure 4-2 Combined Fingerprints for Group A isolates ......................... 152 Figure 4-3 Combined Fingerprints for Group B isolates ......................... 153 Figure 4-4 Combined Fingerprints for Group C Isolates ......................... 154 Figure 4-5 Combined Fingerprints for Group D Isolates ......................... 155 xii o a .4 Oi Oi. - co. . o,. D "' 0". c I .. ”'2'”. “"5 a 'A. . I o-"‘l n."" o D I '.’.": - . "Iv‘. n I "u 0‘ IP.. u._"\ I "v. - .P‘P :"v‘\ "IQ Q n - I :"'v ‘ n.- .. I .0- ‘o' .l . 'u‘ I. A. upl" - C Figure 4-6 Combined Fingerprints for Isolates in Group E ...................... 156 Figure 4-7 Isolate Fingerprints Similar to Azorhizobium ........................ 157 Figure 4-8 Southern Blots of Genomic DNA from Isolates Similar to Azorhizobium cau/inodans .................................................... 158 Figure 6-1 Map of anPSGS-T ................................................................ 252 Figure 6-2 Re-isolation of Bacteria from Sterilized Rice Seedlings ......... 254 Figure 6-3 Bacteria Marked with GUS ..................................................... 255 Figure 6-4 Bacteria Marked with GFP ..................................................... 255 Figure 6-5 Growth of Rice Seedlings in the Pillow System ..................... 256 Figure 6-6 Gus Staining of Rice Plants ................................................... 257 Figure 6-7 Microscopic Observation of Colonization .............................. 258 Figure 6-8 Optical Sections of a Rice Cell Colonized by R100(64) Expressing GFP ...................................................................... 259 Figure 6-9 Three Dimensional Reconstruction of Optical Sections ........ 260 xiii |-’ '1' I III u .o-A l I .v u... FF c - ‘.I ~ . an... .‘ v- ‘ 'I. " u. .' ' a a . I_ ‘P ‘O' l - — 0 ~' "I § CHAPTER 1 INTRODUCTION Use of biologically-fixed nitrogen to replace chemical nitrogen fertilizers is an important goal in agriculture since current cropping systems require large inputs of fertilizer nitrogen and therefore are not sustainable. The recent isolation of endophytic nitrogen-fixing bacteria from Brazilian sugar cane varieties capable of high sustainable yields without input of nitrogen fertilizer has suggested that endOphytic diazotrophic bacteria may be an efficient means for supplying graminaceous crops with biologically-fixed nitrogen (Ladha et al., 1997; James and Olivares, 1998). However, it is not clear to what extent endophytic bacteria colonize crop plants such as rice. This work presents data derived from studies aimed at increasing our understanding of the diversity of bacteria that colonize rice, with special emphasis on diazotrophs. The information presented not only expands our basic understanding of microbial ecology and plant microbe interactions, but is also fundamental for further studies designed to use endophytic bacteria to supply rice and other crops with biologically-fixed nitrogen. 0U...‘ .' b-CV MD! Q. II. I l3 - ov- a n w. 0 -n I v. I "‘qu . '“.. .‘... t, .— ‘ID 4 I“ ‘ "In. \- 'h 'O '- ‘v. ' ’ A. I I 'A \ ~‘ I IMPORTANCE OF NITROGEN IN RICE PRODUCTION Increased rice (Oryza sativa L.) production is clearly essential to meet rising global food demands (Hossain and Fischer, 1995; Peoples et al., 1995). To meet this demand in a sustainable manner the amount of industrially produced fertilizer nitrogen used needs to be reduced (Bohlool et al., 1992; Ladha et al., 1997). One method of achieving this is the use of diazotrophic bacteria which would make biologically-fixed nitrogen available to rice for plant growth. DIAZOTROPHIC BACTERIA ASSOCIATED WITH RICE A considerable amount of information is available about the nature of diazotrophic bacteria found closely associated with rice. In the early 19703, acetylene-reduction assays (ARA) were used to demonstrate the occurrence and activity of nitrogen-fixing organisms in flooded rice fields (Yoshida and Ancajas, 1971; 1973a; 1973b; Dommergues et al., 1973). Many of these early studies lacked proper controls, or were performed using imprecise techniques, causing considerable disagreement about the actual amount of nitrogen fixation taking place (Hirota et al., 1978). Nevertheless, the advantages of cereal crops that are able to fix their own nitrogen have been obvious to many scientists (Postgate, 1974). In efforts to achieve this goal, extensive attempts to isolate the bacteria responsible for this nitrogen fixation have revealed the presence of a broad diversity of diazotrophs inhabiting the rhizosphere, most of which remain poorly studied or uncharacterized. The first diazotrophic bacteria isolated from the rhizosphere in the mid 19705 was called “Spirillum” and was found to be common in the rhizosphere of grasses, including rice (Day and Dobereiner, 1976; Dobereiner et al., 1976; Lakshmi Kumari et al., 1976). The ability of this bacterium to supply nitrogen to plants has been of considerable interest. Renamed Azospirillum, the association between this bacterium and plants has been studied intensively (Fendrik et al., 1995; Vande Broek and Vanderleyden, 1995). Continued isolation revealed additional diazotrophic bacteria associated with rice plants. Watanabe et al. (1979) examined different media for isolation of nitrogen-fixing bacteria from rice and found numerous distinct nitrogen-fixing Gram-negative rod-shaped bacteria, suggesting that Azospirillum might not be the most common diazotroph associated with rice. They studied the percentage of diazotrophic bacteria isolated from different plant tissues and suggested the lower stem as a possible site for _.. n a . D i . v'I I ,g.. IQ I — ‘ «vi a .. u ‘0‘... .. ., 3 ll."' ‘ nitrogen fixation (Watanabe and Barraquio, 1979). In their study, 81% of bacterial isolates from roots that had been shaken with glass beads to remove exogenous bacteria were diazotrophic. Isolation experiments using a semi-solid medium made with exudates from germinating rice seedling as the carbon source revealed members of the family Enterobacteriaceae as the most common nitrogen-fixing rhizosphere bacteria, followed by Azospirillum, and Pseudomonas spp. (Bally et al., 1983). Nitrogen-fixing Pseudomonas diazotrophicus, Enterobacter cloacae, and Klebsiella planticola strains were isolated from the rhizosphere of Philippino rice (Ladha et al., 1983; Watanabe et al., 1987). In an extensive study, Oyaizu-Masuchi and Komagata (1988) reported the isolation of >1000 nitrogen-fixing bacteria from rice roots. Seventy-four of the diazotrophs with the highest acetylene-reducing activity were classified using a variety of biochemical and morphological tests. Bacteria isolated included Xanthobacter autotrophicus, three species of Azospirillum, three species from Enterobacteriaceae, Alcaligenes sp., Protomonas-like bacteria, and Azotobacter—like bacteria. However, 52 of the 74 isolates could not be assigned to any known species and 30 were not similar to any of the reference genera. These 52 bacteria were morphologically and phenotypically diverse, falling into 7 distinct groups (Oyaizu-Masuchi and Komagata, 1988). Bacteria isolated using current culture techniques represent only a small portion of the number of bacteria observed by microscopy (Amann et al., 1995). Therefore, in addition to the culturable bacteria, it is likely that uncultured nitrogen-fixing bacteria also contribute to the diversity of diazotrophs in the rice rhizosphere. To determine the diversity of non- culturable diazotrophs, phylogenetic studies of nitrogen-fixing bacteria associated with rice roots using PCR primers designed to amplify fragments from nif genes have been carried out (Ueda et al., 1995a; 1995b). The sequences of 16 mm and 23 nifH PCR products thus generated distinct from sequences of nitrogen-fixing genes from culturable bacteria, and were found to form seperate branches in a phylogenetic tree (Ueda et al., 1995a; 1995b). This provides further evidence for the notion that many of the nitrogen-fixing bacteria from the rice rhizosphere have not yet been characterized. The studies summarized above revealed a diversity of diazotrophic bacteria associated with rice roots but have provided little insight into the diversity of bacteria that colonize internal rice tissues. Watanabe and Barraquio (1979) attempted to remove rhizoplane bacteria by vigorously shaking the rice tissue with glass beads prior to isolation. While enriching for bacteria from the “inner rhizoplane”, this isolation undoubtedly included bacteria which adhere tightly to the surface of the root. Other isolation protocols made no effort to distinguish between bacteria from the interior of the root and bacteria from the rhizoplane and rhizosphere, nor to determine colonization sites. The lack of information about the diversity of bacteria specifically isolated from internal tissues leaves a gap in our understanding of an important niche in microbial ecology. Information on bacteria that occupy this niche is needed, since bacteria from this niche may present an opportunity to increase the supply of biologically-fixed nitrogen to rice. SUPPLYING NITROGEN TO RICE For bacteria to effectively supply crop plants with nitrogen they must have an adequate carbon supply to sustain the energy-intensive process of nitrogen fixation. In addition, the intracellular oxygen tension must be high enough for efficient nitrogen fixation without inactivating the oxygen-sensitive nitrogenase enzyme, and the nitrogen fixed by the bacteria must be transferred from the bacteria to the plant (Kennedy and Tchan, 1992; Kennedy et al., 1997). There must also be enough bacteria to produce the nitrogen needed, which will be affected by the three conditions mentioned above, as well as the total nitrogen demand of the crop plant. Problems with nutrient transfer between rhizosphere bacteria and plants are illustrated by the interactions between Azospirillum and various cereal crops. Colonization of wheat by Azospirillum brasilense is limited to the surface of the root, with occasional intercellular colonization of the cortex (Schank et al., 1979; Levanony et al., 1989; Bashan and Levanony, 1989; Kennedy et al., 1998). Plants gain little nitrogen from bacterial nitrogen fixation when inoculated with Azospirillum (Nayak et al., 1986; Christiansen-Weniger and van Veen, 1991; Kennedy et al., 1998). The addition of a carbon source to the rhizosphere increases the amount of nitrogen fixation (Christiansen-Weniger and van Veen, 1991) . This indicates that rhizosphere bacteria are limited for carbon and therefore not efficient sources for biological nitrogen fixation (BNF). Nitrogen supplied to the plant can be increased by the use of ammonium-excreting mutants of Azospirillum (Christiansen-Weniger and van Veen, 1991; Christiansen-Weniger 1992), however, these bacteria compete poorly in field soil. This indicates that in the rhizosphere the bacteria use most of the nitrogen they fix for their own growth, rather than supplying nitrogen to the plant. The intimate association between plant and endophyte may provide more suitable conditions for nutrient transfer between bacteria and their host than the association between rhizosphere bacteria and plants. Acetobacter diazotrophicus, an endophytic diazotroph isolated from sugar cane, excretes fixed nitrogen (Cojho et al., 1993), does not survive well in soils (James and Olivares, 1998), colonize the interior of the sugar cane and appear to supply the host with nitrogen (James and Olivares, 1998). This provides an example of a plant host providing protection from rhizosphere competition, and simultaneously allowing an ammonium- excreting bacterium to flourish and provide nitrogen to the host. Based on this example, it is conceivable that genetically modified ammonium- excreting endophytic bacteria could effectively supply the host plant with nitrogen. The ambient oxygen tension required for efficient nitrogen fixation varies between diazotrophs with optimal oxygen tensions ranging from 0.3% to 6.5°/o (Vande Broek et al., 1996). For effective nitrogen fixation bacteria Ir. require enough oxygen for efficient respiration, but low concentrations of free oxygen to protect the oxygen-sensitive nitrogenase enzymes. Sites in the interior of plant tissue, especially flooded rice, may provide the environment required for efficient nitrogen fixation. The apparently high rates of nitrogen fixation by endophytic diazotrophs in sugar cane (Urquiaga et al., 1992) and the de-repression of nif genes by Azoarcus (Egener et al., 1998) and Alcalegenes faeca/is (Vermeiren et al., 1998) following internal colonization of rice roots suggest that conditions favorable for nitrogen fixation, including low oxygen tension, must exist inside the plant tissue. The number of bacteria needed to provide enough biologically-fixed nitrogen to reduce or eliminate fertilizer inputs will depend heavily on the efficiency of nitrogen fixation and nutrient transfer. In sugar cane, estimated population sizes for A. diazotrophicus range from 103 to 107 bacteria per gram fresh weight of roots (Boddey, 1995). Studies of BNF in Brazilian sugarcane cropping systems suggest contributions of 39 kg N ha'1 year'1 to over 170 kg N ha'1 year'1 ( Boddey, 1995). It is unclear if this nitrogen is fixed by A. diazotrophicus alone or in conjunction with other endophytic diazotrophs. However, it has been suggested that BNF by endophytes contributes enough nitrogen for maximum yields, while requiring minimal inputs of chemical nitrogen fertilizer (Boddey et al., 1 995). The intimate interactions between endophytic bacteria and their host may provide the potential for significant contributions of biologically-fixed nitrogen to rice growth and yield. Nature provides us with several examples of bacteria living internally in eukaryotic host tissues and supplying the hosts with nitrogen or other nutrients. The Rhizobium/Iegume symbiosis has been the most widely studied example of the former (Long, 1989; van Rhijn and Vanderleyden, 1995). A similar symbiosis between Frankia and actinorhizal plants has also been studied, but to a lesser degree (Benson and Silvester, 1993). The symbiosis between Anabaena—Azolla and Nostoo Gunnera are examples of cyanobacteria supplying eukaryotic hosts with biologically-fixed nitrogen (Peters and Meeks, 1989; Bergman et al., 1992). As mentioned above, endophytic bacteria present in selected sugar cane varieties appear to contribute enough fixed nitrogen to the plant host so that no chemical nitrogen fertilizer is needed for maximum yields (Lima et al., 1987; Boddey et al., 1991; Urquiaga et al., 1992; James et al., 1994). 10 It has been proposed that diazotrophic bacterial endosymbionts of shipworms (teredinid mollusks) allow these marine invertebrates to utilize wood as a principal food source (Distel et al., 1991), suggesting that the bacteria supply some nitrogen to the eukaryotic host. Thioautotrophic and methanotrophic bacterial endosymbionts of other marine invertebrates supply their hosts with the energy needed for life (Cavanaugh 1994; Distel et al., 1995). Chloroplasts and mitochondria are also ancient endosymbiotic bacteria that have evolved into organelles (Gary, 1989). In each of these cases, the bacterial symbiont provides the eukaryotic host with biochemical pathways not readily available in the host cell. As discussed above, bacteria can supply a eukaryotic host with nutrients. However, diazotrophic endophytes of rice currently do not supply enough biologically fixed nitrogen to sustain high yields. Molecular biology allows scientists to manipulate bacteria in very specific ways, making it theoretically possible to engineer a bacterium with enhanced endophytic interactions with its crop host. One example is the use of genetically engineered endophytic bacteria to provide corn with insect resistance. Clavibacter xy/i subsp. cynodontis is capable of endophytic establishment 11 in maize xylem. Inoculation of corn with this bacterium engineered to express the gene encoding CryIA(c) insecticidal protein reduced European corn borer damage by 60% (Lampel et al., 1994). In the future it should also be possible to enhance diazotrophic endophytes to supply their host with more biologically fixed nitrogen. At present our knowledge of the complex interaction between endophytes and rice is limited and much of the information needed to use diazotrophic endophytic bacteria to replace chemical nitrogen fertilizer is still lacking (Quispel, 1992). The diversity of rice endophytes and the niches they colonize have not been determined. Therefore, more studies that focus on determining the types of endophytic bacteria that associate with rice, the nature of the interactions between host and endophyte, and the sites of endophytic colonization are clearly needed. DEFINITION OF ENDOPHYTIC BACTERIA Disagreements exist about the definition of “endophytic” bacteria. Ouispel (1992) restricts his definition of endophytic organisms to non- pathogenic microorganisms that spend a considerable part of their life cycle within the plant host, either in inter- or intracellular locations. Other definitions of endophytes also exclude pathogens (Misaghi and 12 Donndelinger, 1990; Frommel et al., 1991; Fisher et al., 1992). However, Kloepper et al. (1992) does not exclude pathogens, nor will I in this thesis, because of the difficulty distinguishing between pathogenic and non-pathogenic interactions, especially when considering “quiescent” pathogens and variations in plant reactions to an “endophyte” depending on plant genotype (James and Olivares, 1998). Baldani et al. (1997) define obligate endophytes as those that survive only inside the plant host, while facultative endophytes can also survive in the soil. Distinguishing between these two classes in a robust manner is difficult. However, this distinction may be useful for identifying bacteria most suited for future use in crop inoculations, because “obligate” endophytes many form more intimate associations with the plant. Kloepper et al. (1992) does not included bacteria colonizing only the epidermis. I have also chosen to exclude this type of colonization as it is probable that many opportunistic rhizosphere/rhizoplane bacteria occupy this niche. Therefore, in this thesis endophytic bacteria will be defined as bacteria capable of survival and multiplication either inter- or intracellularly, in plant tissues internal to the epidermis. 13 ENDOPHYTIC BACTERIA ASSOCIATED WITH PLANTS Bacterial endophytes have been isolated from healthy plants including onion and potato (Frommel et al., 1991), corn (Fisher et al., 1992; Mclnroy and Kloepper, 1995; Palus et al., 1996), cotton (Misaghi and Donndelinger, 1990; Mclnroy and Kloepper, 1995), Kallar grass (Reinhold- Hurek and Hurek, 1998a; 1998b), sugar cane (James and Olivares, 1998), C-4 grasses (Kirchhof et al., 1997) and rice (Barraquio et al., 1997; see below). A non-fluorescent Pseudomonas sp. has been shown to colonizes the vascular tissues of potato (Frommel et al., 1991), A. diazotrophicus and HerbaspiriI/um spp. colonize the vascular tissue of sugar cane (James and Olivares, 1998) and Azoarcus can colonize the cortex and vascular tissue of rice and Kallar grass (Hurek et al., 1994). Very little is known about the sites of colonization of other natural bacterial endophytes. ENDOPHYTIC BACTERIA ASSOCIATED WITH RICE The presence of endophytic bacteria is suggested by the isolation of bacteria from surface-sterilized rice tissues. Using surface sterilization protocols and most probable number estimates, Barraquio et al. (1997) calculated the population of putative endophytic bacteria in rice to be between 105 to 108 g dry wt'I. The endophytic bacteria population size 14 varied with the rice genotype examined. Less than 10% of the putative endophytic bacteria were diazotrophic. The presence of endophytic bacteria is supported by electron micrographs of field-grown rice that revealed the presence of bacteria with diverse morphologies in the interior of rice roots (Yanni et al., 1997). In addition to the evidence suggesting endophytic colonization of field grown plants, studies carried out in the laboratory suggest that certain bacteria can infect rice tissue endophytically. Barraquio et al. (1997) have used the gus marker gene to visualize subepidermal colonization of rice roots by Herbaspirillum seropedicae, a diazotroph found associated with many graminaceous crops (Baldani et al., 1986). You and Zhou (1988) have described the colonization of cortical rice root cells by Alcaligenes faecalis A15 using light and electron microscopy. However, A. faecalis A15 cells marked with the gus marker gene were only found in epidermal cells of rice roots (Vermeiren et al., 1998). A. faecalis colonizing the epidermal cells of rice express the gus gene under the control of the nifH promoter, suggesting that the bacteria are fixing nitrogen during colonization of roots (Vermeiren et al., 1998). However, no evidence was presented that the rice plants benefited from this 15 nitrogen fixation activity. A diazotrophic bacterium belonging to the genus Azoarcus has been isolated from the roots of Kallar grass and shown to be able to invade the cortex and vascular system of rice roots and occupy dead plant cells (Hurek et al., 1994). Azoarcus in the intercellular spaces of the cortex and in dead epidermal cells expresses the gfp gene under the control of the nifH promoter, suggesting the bacteria are fixing nitrogen (Egener et al., 1998; Reinhold-Hurek and Hurek, 1998b). Inoculation with Azoarcus has also been observed to increase rice plant growth, however, it has been demonstrated that the observed growth increase was not due to nitrogen fixation by the bacteria (Hurek et al., 1994). There is also evidence that some rhizobia can colonize rice tissue. Cooking et al. (1995) have reported the formation of short thick lateral roots, intercellular infection pockets, and nitrogen fixation activity following inoculation of rice or wheat with Azorhizobium caulinodans ORSS71 or Rhizobium ORS310. These bacteria are capable of forming root and stem nodules on Sesbania rosfrata and Aeschynomene indica, respectively. The ability of A. caulinodans and other rhizobia to cause short, thick lateral roots, colonize intercellular spaces between the 16 epidermis and cortex at the point of lateral root emergence, and occupy apparently dead cortical cells has been confirmed (Reddy et al., 1997; Webster et al., 1997). However, the claim that these bacteria supply fixed nitrogen to the rice plant has not been substantiated. Another rhizobial strain, Rhizobium leguminosarum bv. trifolii, has been isolated from field grown rice rotated with Egyptian berseem clover, the legume host of R. leguminosarum bv. trifolii. These bacteria were shown to be able to colonize the upper portion of the rice root (Yanni et al., 1997). Colonization of rice by these bacteria stimulates plant growth. As in the cases described above, this plant growth stimulation is most likely not due to nitrogen fixation (Yanni et al., 1997). These studies show that diazotrophic bacteria do interact intimately with rice roots, and in some cases have endophytic colonization patterns. However, the mechanisms underlying the ability of these bacteria to colonize rice roots in an endophytic manner are poorly understood, and in none of these cases has it been shown that the rice plant benefits from nitrogen fixed by the bacteria, nor has intracellular colonization of living plant cells been demonstrated. Moreover, comprehensive studies on the diversity of endophytic bacteria, a careful analysis of the ability of 17 bacteria to colonize rice roots endophytically, and a detailed characterization of the invasion mechanisms is still lacking. SUMMARY It is important to realize that nitrogen fixation in cereal crops is a long term project, and while potentially great benefits exist for the future, immediate results should not be expected (de Bruijn et al., 1995; Ladha et al., 1997). Progress has been made in understanding the interactions of specific nitrogen-fixing bacteria with cereals and their contribution to crop yield. However, more information is needed about the bacteria capable of invading rice roots, including the diversity of such bacterial endophytes, modes of bacterial invasion, sites of bacterial colonization, competitiveness of endophytes in colonizing the roots, endophytic population sizes and the ability of the bacteria to fix nitrogen and transfer it to the plant. Previous studies have concentrated on identifying diazotrophic bacteria from the rhizosphere of rice roots or looked at the ability of one type of bacterium to colonize rice and supply it with nitrogen. A comprehensive study of the diversity of rice endophytes and the sites of endophytic colonization has not been undertaken. 18 This study begins to address the genetic diversity of endophytic bacteria associated with rice and the sites they colonize. In Chapter 2 putative rice endophytes are isolated from mechanically-abraded, surface-sterilized rice tissue and preliminary characterization of their genetic diversity and diazotrophy is carried out. Chapter 3 describes the development of nifD- specific PCR primers and the use of these primers to identify diazotrophic bacteria in a collection of putative rice endophytes. In Chapter 4, the genetic diversity of 142 putative endophytic bacteria is evaluated using amplified ribosomal DNA analysis (ARDRA) fingerprinting, rep-PCR genomic fingerprinting, and small subunit (SSU) ribosomal RNA (rRNA) gene DNA sequencing. In Chapter 5, the utility of the green fluorescent protein (GFP) as a biomarker/bioreporter is reviewed. Chapter 6 demonstrates the use of two biomarkers, GUS and GFP, to study the sites and levels of rice colonization using two putative rice endophytes and one Rhizobium control strain. These data provide insight into the diversity of bacteria isolated from surface-sterilized rice tissue and the utility of biomarkers for observing colonization of rice tissue by bacteria, and will be useful in the planning and success of future experiments aimed at using endophytic bacteria to supply rice with biologically-fixed nitrogen. 19 REFERENCES Amann RI, Ludwig W, Schleifer K-H (1995) Phylogenetic identification and in situ detection of individual microbial cells without cultivation. Microbiol Rev 59: 143-169 Baldani JI, Baldani VLD, Seldin L, Dobereiner J (1986) Characterization of Herbaspiril/um seropedicae gen. nov., sp. nov., a root associated nitrogen-fixing bacterium. Int J Syst Bacteriol 36: 86-93 Baldani Jl, Caruso L, Baldani VLD, Goi S, Dobereiner J (1997) Recent advances in BNF with non-legume plants. Soil Biol Biochem 29: 911-922 Bally R, Thomas-Bauzon D, Heulin T, Balandreau J (1983) Determination of the most frequent Ng-fixing bacteria in the rice rhizosphere. Can J Microbiol 29: 881 -887 Bashan Y, Levanony H (1989) Factors affecting adsorption of Azospirillum brasilense Cd to root hairs as compared with root surface of wheat. Can J Microbiol 35: 936-944 Barraquio WL, Revilla L, Ladha .K (1997) Isolation of endophytic diazotrophic bacteria from wetland rice. Plant and Soil 194: 15-24 Benson DR, Silvester WB (1993) Biology of Frankia strains, actinomycete symbionts of actinorhizal plants. Microbial Rev 57: 293- 319 Bergman B, Johansson C, deerback E (1992) Tangsley Review No. 42 the Nostoc-Gunnera symbiosis. New Phytologist 122: 379-400 Boddey FM (1995) Biological nitrogen fixation in sugar cane: A key to energetically viable biofuel production. Critical Reviews in Plant Sciences 1 4: 263-279 Boddey RM, Urquiaga S, Reis V, Dobereiner J (1991) Biological nitrogen fixation associated with sugar cane. Plant and Soil 137: 1 1 1— 1 17 2O Boddey RM, de Oliveira OC, Urquiaga S, Reis VM, de Olivares FL, Baldani VLD, Dobereiner J (1995) Biological nitrogen fixation associated with sugar cane and rice: Contributions and prospects for improvement. Plant and Soil 174: 195-209 Bohlool BB, Ladha JK, Garrity DP, George T (1992) Biological nitrogen fixation for sustainable agriculture: A perspective. Plant and Soil 141: 1-11 Cavanaugh (M (1994) Microbial symbiosis: patterns of diversity in the marine environment. Am Zool 34: 79-89 Christiansen-Weniger C (1992) N2-fixation by ammonium-excreting Azospirillum brasilense in auxine-induced root tumors of wheat (Triticum aestivum L.) Biol Fertil Soils 13: 165-172 Christiansen-Weniger C, van Veen JA (1991) Nitrogen fixation by Azospirillium brasilense in soil and the rhizosphere under controlled environmental conditions. Biol Fertil Soils 1 22100-106 Cocking EC, Kothari SI, Batchelor CA, Jain S, Webster G, Jones J, Jotham J, Davey NR (1995) Interaction of rhizobia with non- Iegume crops for symbiotic nitrogen fixation nodulation. In I Fendrik, M del Gallo, J Vanderleyden and M de Zamaroczy, eds, Azospirillum VI and related microorganisms: Genetics, physiology, ecology. Springer-Verlag, New York, pp 197-205 Coiho EH, Reis VM, Schenberg ACG, DobereingerJ (1993) Interactions of Acetobacter diazotrophicus with an amylolytic yeast in nitrogen-free batch culture. FEMS Microbiol Lett 106: 341 -346 Day JM, Dobereiner J (1976) Physiological aspects of N2-fixation by a Spirillum from Digitaria roots. Soil Biol Biochem 8: 45-50 de Bruijn FJ, Jing Y, Dazzo F8 (1995) Potentials and pitfalls of trying to extend symbiotic interactions of nitrogen-fixing organisms to presently non-nodulated plants, such as rice. Plant and Soil 174: 225- 240 21 Distel DL, DeLong EF, Waterbury JB (1991) Phylogenetic characterization and in situ localization of the bacterial symbiont of shipworms (Teredinidae: Bivalvia) by using 16s rRNA sequence analysis and oligodeoxynucleotide probe hybridization. Appl Environ Microbiol 57: 2376-2382 Distel DL, Lee HK-W, Cavanaugh CM (1995) Intracellular coexistence of methano- and thioautotrophic bacteria in a hydrothermal vent mussel. Proc Natl Acad Sci 92: 9598-9602 Dobereiner J, Marriel IE, Nery M (1976) Ecological distribution of Spirillum lipoferum Beijerinck. Can J Microbiol 22:1464-1473 Dommergues Y, Balandreau J, Rinaudo G, Weinhard P (1973) Nonsymbiotic nitrogen fixation in the rhizosphere of rice, maize, and different tropical grasses. Soil Biol Biochem 5: 83-89 Egener T, Hurek T, Reinhold-Hurek B (1998) Use of green fluorescent protein to detect expression of nif genes in Azoarcus sp. BH72, a grass-associated diazotroph, on rice roots. Mol Plant Microbe Interact 1 1: 71-75 Fendrik I, del Gallo M, Vanderleyden J, de Zamaroczy M eds (1995) Azospirillum VI and Related Organisms. Springer, Berlin Fisher PJ, Petrini O, Lappin Scott I-llIl (1992) The distribution of some fungal and bacterial endophytes in maize (Zea mays L.). New Phytologist 122: 299-305 Frommel MI, Nowark J, Lazarovits G (1991) Growth enhancement and developmental modifications of in vitro grown potato (Solanum tuberosum sp. tuberosum) as affected by a nonfluorescent Pseudomonas sp. Plant Physiol 96: 928-936 Gary MW (1989) The evolutionary origins of organelles. Trends Genet 5: 294-299 Hirota Y, Fujii T, Sano Y, Iyama S (1978) Nitrogen Fixation in the Rhizosphere of rice. Nature 276: 416-417 22 Hossain M, Fischer KS (1995) Rice research for food security and sustainable agricultural development in Asia: Achievements and future challenges. GeoJournal 35: 286-295 Hurek T, Reinhold-Hurek B, van Montagu M, Kellenberger E (1994) Root Colonization and systemic spreading of Azoarcus sp strain BH72 in grasses. J Bacteriol 176: 1913-1923 James EK, Reis VM, Olivares FL, Baldani JI, Dobereiner J (1994) Infection of sugar cane by the nitrogen-fixing bacterium Acetobacter diazotrophicus. J Exp Bot 45: 757-766 James EK, Olivares FL (1998) Infection and colonization of sugar cane and other graminaceous plants by endophytic diazotrophs. Critical Reviews in Plant Sciences 17: 77-119 Kennedy IR, Pereg-Gerk LL, Wood C, Deaker R, Gilchrist K, Katupitiya S (1997) Biological nitrogen fixation in non-leguminous field crops: Facilitating the evolution of an effective association between Azospirillum and wheat. Plant and Soil 194: 65-79 Kennedy IR, Katupitiya S, Yu D, Gilchrist K, Deaker R, Pereg- Gerk L, Wood C (1998) Prospects for facilitated evolution of effective N2-fixing associations with cereals: comparative performance of Azospirillum brasilense Sp7-S with various free-living diazotrophs in para- nodulated wheat. In KA Malik, MS Mirza, JK Ladha, eds, Nitrogen Fixation with Non-Legumes. Kluwer Academic Publishers, Dordrecht, pp 109-131 Kennedy IR, Tchan Y-T (1992) Biological nitrogen fixation in non- Ieguminous field crops: Recent advances. Plant and Soil 141:93-1 18 Kirchhof G, Reis VM, Baldani JI, Eckert B, Dobereiner J, Hartmann A (1997). Occurrence, physiological and molecular analysis of endophytic diazotrophic bacteria in gramineous energy plants. Plant and Soil 194: 45-55 Kloepper JW, Schippers B, Bakker PAI-IM (1992) Proposed Elimination of the term endorhizosphere. The American Phyhtopathological Society 82: 726-727 23 315.: upchf no"! . ‘fi. - .513!” I}: ,_ ~-... -‘ ‘I. Ladha JK, Barraquio WL, Watanabe I (1983) Isolation and identification of nitrogen-fixing Enterobacter cloacae and Klebsiella planticola associated with rice plants. Can J Microbiol 29: 1301 -1308 Ladha J(, de Bruijn FJ, Malik KA (1997) Introduction: Assessing opportunities for nitrogen fixation in rice — a frontier project. Plant and Soil 194: 1-10. Lakshmi Kumari M, Kavimandan SK, Subba Rao NS (1976) Occurrence of nitrogen-fixing Spirillum in roots of rice, sorghum, maize, and other plants. Indian J Exp Biol 1 4: 638-639 Lampel JS, Canter GL, Dimock MB, Kelly JL, Anderson JJ, Uratani BB, Foulke JS, Turner JT (1994) Integrative cloning, expression and stability of the cry/A(c) gene from Bacillus thuringiensis subsp. kurstaki in a recombinant strain of Clavibacter xyli subsp. cynodontis. Appl Environ Microbiol 60: 501 -508 Levanony H, Bashan Y, Romano B, Klein E (1989) Ultrastructural localization and identification of Azospirillum brasilense Cd on and within wheat root by immuno-gold labeling. Plant and Soil 117: 207-218 Lima E, Boddey RM, Dobereiner J (1987) Quantification of biological nitrogen fixation associated with sugar can using a 15N aided nitrogen balance. Soil Biol Biochem 19: 165-170 Long S? (1989) Rhizobium-legume nodulation: Life together in the underground. Cell 56: 203-314 Misaghi IJ, Donndelinger CR (1990) Endophytic bacteria in symptom-free cotton plants. Phytopathologhy 80: 808-811 Mclnroy JA, Kloepper JW (1995) Survey of indigenous bacterial endophytes from cotton and sweet corn. Plant and Soil 173: 337-342 Nayak DN, Ladha JK, Watanabe I (1986) The fate of marker Azospirillum lipoferum inoculated into rice and its effect on growth, yield and N2 fixation of plants studied by acetylene reduction, 15N2 feeding and 15N dilution techniques. Biol Fertil Soils 2: 7-14 24 Oyaizu-Masuchi Y, Komagata K ( 1988) Isolation of free-living nitrogen-fixing bacteria from the rhizosphere of rice. J Gen Appl Microbiol 3 4: 127-1 64 Palus JA, Borneman J, Ludden PW, Triplett EW (1996) A diazotrophic bacterial endophyte isolated from stems of Zea mays and Zea luxurians lltis and Doebley. Plant and Soil 186: 135-142 Peters GA, Meeks JC (1989) The AzoIIa-Anabaena symbiosis: Basic biology. Annu Rev Plant Physiol Plant Mol Biol 40: 193-210 Peoples MB, Herridge DF, Ladha .K (1995) Biological nitrogen fixation: An efficient source of nitrogen for sustainable agricultural production? Plant and Soil 174: 3-28 Postgate .B (1974) New advances and future potential in Biological nitrogen fixation. J Appl Bact 37: 185-202. Quispel A (1992) A search for signals in endophytic microorganisms. In DPS Verma, ed, Molecular Signals in Plant-Microbe Communications. CRC Press, London, pp 471-491 Reddy PM, Ladha JK, So RB, Hernandez RJ, Ramos MC, Angeles OR, Dazzo FB, de Bruijn FJ (1997) Rhizobial communication with rice roots: Induction of phenotypic changes, mode of invasion and extent of colonization. Plant and Soil 194: 81-89 Reinhold-Hurek B, Hurek T (19983) Interactions of gramineous plants with Azoarcus spp. and other diazotrophs: Identification, localization, and perspectives to study their function. Critical Reviews in Plant Sciences 17:29-54 Reinhold-Hurek B, Hurek T (1998b) Life in Grasses: Diazotrophic Endophytes. Trends in Microbiol. 6: 139-202 Schank SC, Smith RL, Weiser GC, Zuberer DA, Bouton JH, Guesenberry KH, Tyler ME, Milam JR, Littell II: (1979) Fluorescent antibody technique to identify Azospirillum brasilense associated with roots of grasses. Soil Biol Biochem 11: 287-295 25 Ueda T, Suga Y, Yahiro N, Matsuguchi T (1995a) Remarkable N2- fixing bacterial diversity detected in rice roots by molecular evolutionary analysis of nifH gene sequences. J Bacteriol 177: 1414-1417 Ueda T, Suga Y, Yahiro N, Matsuguchi T (1995b) Genetic diversity of Ng-fixing bacteria associated with rice roots by molecular evolutionary analysis of mm Library. Can J Microbiol 41: 235-240 Urquiaga S, Cruz KI-IS, Boddey FM (1992) Contribution of nitrogen fixation to sugar cane: Nitrogen-15 and nitrogen-balance estimates. Soil Sci Soc Am J 56:105-114 Vande Broek A, Vanderleyden J (1995) Review: Genetics of the Azospirillum-plant root association. Critical Reviews in Plant Science 14: 445-466 Vande Broek A, Keijers V, Vanderleyden J (1996) Effect of oxygen on the free-living nitrogen fixation activity and expression of the Azospirillum brasilense nifH gene in various plant-associated diazotrophs. Symbiosis 21: 25-40 van Rhijn P, Vanderleyden J (1995) The Rhizobium-Plant symbiosis. Microbiol Rev 59: 124-142 Vermeiren H, Vanderleyden J, Hai W (1998) Colonization and nifH expression on rice roots by AIcaIigenes faecalis A15. In KA Malik, MS Mirza, JK Ladha, eds, Nitrogen Fixation with Non-Legumes. Kluwer Academic Publishers, Dordrecht, pp 167-177 Watanabe I, Barraquio WL, De Guzman MR, Cabrera DA (1979) Nitrogen-Fixing (Acetylene Reduction) activity and population of aerobic heterotrophic nitrogen-fixing bacteria associated with wetland rice. Appl Environ Microbiol 37: 813-819 Watanabe I, Barraquio W (1979) Low levels of fixed nitrogen required for isolation of free-living N2-fixing organisms from rice roots. Nature 277: 565-566 26 Watanabe I, So R, Ladha JK, Katayama-Fujimura Y, Kuraishi H (1987) A new nitrogen-fixing species of pseudomonad: Pseudomonas diazotrophicus sp. nov. isolated from the root of wetland rice. Can J Microbiol 33: 670-678 Webster G, Gough C, Vasse J, Batchelor C A, O’Callaghan K J, Kothari S L, Davey MR, Dénarié J, Cocking EC (1997) Interactions of rhizobia with rice and wheat. Plant and Soil 194: 1 15-122 Yanni YG, Rizk RY, Corich V, Squartini A, Ninke K, Philip- Hollingsworth S, Orgambide G, de Bruijn F, Stoltzfus J, Buckley D, Schmidt TM, Mateos PF, Ladha JK, Dazzo F8 (1997) Natural endophytic association between Rhizobium leguminosarum bv. trifolii and rice roots and assessment of its potential to promote rice growth. Plant and Soil 194: 99-114 Yoshida T, Ancajas FR (1971) Nitrogen fixation by bacteria in the root zone of rice. Soil Sci Soc Am Proc 35: 156-158 Yoshida T, Ancajas FR (1973a) Nitrogen fixation activity in upland and flooded rice fields. Proc Soil Sci Soc Am 37: 42-46 Yoshida T, Ancajas FR (1973b) The fixation of atmospheric nitrogen in the rice rhizosphere. Soil Biol Biochem 5: 153-155 You C, Zhou F (1988) Non-nodular endophytic nitrogen fixation in wetland rice. Can J Microbiol 35: 403-408 27 CHAPTER 2 ISOLATION OF ENDOPHYTIC BACTERIA FROM RICE AND ASSESSMENT OF THEIR POTENTIAL FOR SUPPLYING RICE WITH BIOLOGICALLY FIXED NITROGEN J. R. Stoltzfus‘, R. 804, PP. Malarvithi4, J.K. Ladha4 and F.J. de Bruijn‘v2v3' ’MSU-DOE Plant Research Laboratory, 2Department of Microbiology, and 3 NSF Center for Microbial Ecology, Michigan State University, E. Lansing, MI 48824, USA; 4 International Rice Research Institute, Los Banos, The Philippines. Published in Plant and Soil 194: 25-36, 1997 ABSTRACT The extension of nitrogen-fixing symbioses to important crop plants such as the cereals has been a long-standing goal in the field of biological nitrogen fixation. One of the approaches that has been used to try to achieve this goal involves the isolation and characterization of stable endophytic bacteria from a variety of wild and cultivated rice species that either have a natural ability to fix nitrogen or can be engineered to do so. Here we present the results of our first screening effort for rice endophytes and their characterization using acetylene reduction assays (ARA), genomic fingerprinting with primers corresponding to naturally 28 occurring repetitive DNA elements (rep-PCR), partial 168 rDNA sequence analysis and PCR-mediated detection of nitrogen fixation (nif) genes with universal nif primers developed in our laboratory. We also describe our efforts to inoculate rice plants with the isolates obtained from the screening, in order to examine their invasiveness and persistence (stable endophytic maintenance). Lastly, we review our attempts to tag selected isolates with reporter genes/proteins, such as beta-glucuronidase (gus) or green fluorescent protein (gfp), in order to be able to track putative endophytes during colonization of rice tissues. 29 INTRODUCTION Rice (Oryza sativa L.) is the staple in the diet of over 40% of the world’s population making it the most important food crop currently produced (Hossain and Fischer, 1995). Much of this rice is grown in countries where rapidly growing populations, coupled to limited amounts of land and scarce resources, make high yields per hectare with reduced inputs essential to avoid food shortages. Crop productivity is based on numerous variables including weather, soil type, moisture, and nutrients. One of the most important factors in the generation of high yields from modern rice crops is nitrogen fertilizer. Without the addition of (fertilizer) nitrogen the yield of the present varieties is drastically limited. While biological nitrogen fixation in wetland rice fields contributes significantly to the long term fertility of these systems, it is not enough to produce maximum yields. Studies show that biological nitrogen fixation in flooded rice paddies can yield up to 50 kg N ha'1/crop'1 (Roger and Ladha, 1992). This biologically-fixed nitrogen has been sufficient to maintain traditional flooded rice systems for thousands of years. However, the productivity of these sustainable systems is low, with yields less than four tons per hectare. Modern agriculture has increased rice yields to five to eight tons per hectare, but requires the 30 bun V I 'I-P‘ a. Q ’ I '.~¢ luv ”-1- 1“ Ff yuvdv- v ”4.. AAA I A ~'\ I‘-.uud M An \~ " 'v‘uv ., b."-A ‘ . I'M." III...- I . "\ NW; ' ‘1. n 1" u u .0 .‘1 .~ U V '- 'ula. . fiai '1 ' I l I u. - - .Iu‘ I v, A B1 ..‘_ . J 5 A... a i“. 0,..fi u A ”A '1 U. i .lo ab- 1 .X" input of 60 to 100 kg/ha‘1 of fertilizer to supply additional nitrogen. Global population estimates predict the need for a 70% increase in rice production over the next thirty years. The use of current agronomic practices to generate this increase would require even larger inputs of fertilizer nitrogen. The use of high levels of nitrogen fertilizers in crop production has several drawbacks. Most nitrogen fertilizer is produced via the Haber-Bosch process. This process requires large amounts of natural gas, coal, or petroleum, all non-renewable energy source. In addition it produces CO2, a gas implicated in the greenhouse effect. The chemical production of nitrogen fertilizer is also expensive, and in developing countries the additional costs often exceed the means of farmers, limiting the yield potential of their crops. Once chemical fertilizers are applied, additional problems can arise. Roughly one third of the nitrogen applied is used by the crop. The non-assimilated nitrogen from farming systems has been implicated in nitrate contamination of ground water supplies, a potential health hazard. In addition, excess nitrogen can also lead to production of nitrous oxide (N20), a potent “greenhouse” gas. Therefore, crop systems requiring large additions of fertilizer nitrogen are non-sustainable 31 systems, since they require the use of non-renewable natural resources and can contribute to health hazards and environmental pollution. Decreasing the amount of industrially produced fertilizer nitrogen needed in agricultural systems is an important goal of agricultural scientists in general (Bohlool et al., 1992). In the case of sustainable rice production, in particular, one important aim is to replace industrially fixed nitrogen with biologically fixed nitrogen. According to a National Research Council Report (1994), an estimated 100-175 million metric tons of biologically fixed nitrogen is produced annually. Most of this nitrogen is fixed by the legume/Rhizobium symbiosis. This is significantly more than the 10 million metric tons produced by lightning and 80 million tons produced in 1989 by industrial processes. Therefore, it is obvious that (symbiotic) biological nitrogen fixation has great potential for supplying nitrogen to crops. However, the delivery of biologically fixed nitrogen to plants such as rice has been problematic. For example, supplying biologically fixed nitrogen through associative nitrogen-fixing bacteria or green manuring has some of the same drawbacks that have been observed with industrially 32 nn! #7"; r- ‘ 0 vi b d th‘n’g \" \ C 0'.» U U5 In! AA - '43.» P. t I n ”A a. 1"" I I. .| 'hnn...~fl F -' _ .— no... .l., A Al" I .., 5 D d H U "P 'n n A A b 1 ‘ " - it: I ll 5 f ‘1 ..~ I? ‘c v ‘ ;‘ f ‘7 A ~ v.|\' H "‘u: I: Y]! . O. . l «I t ‘I .‘~~.n ‘ Id ‘:‘. O‘.l \ C... '5 p r 1‘” .H ‘- ' n ‘ v F v... 'I ‘- p "4 ~. produced fertilizers. All three methods rely on plant uptake of nitrogen from the soil nitrogen pool. In order to meet the nitrogen requirements of the plant, a large excess nitrogen pool is needed. Green manure systems using Sesbania rostrata, Aeschynomene afraspera, or Azolla produce yields comparable to fertilized control plots (Ladha et al., 1992; Diekmann et al., 1996; Kumarasinghe et al., 1986). However, the percentage of the nitrogen from green manure incorporated by the rice plant is similar to that from fertilizer nitrogen (Clement et al., 1995; Kumarasinghe et al., 1986) and, therefore, nitrogen losses in green manure systems can be similar to those in fertilized systems (Harris et al., 1994). In legume crops this problem is avoided because the endo- symbiotic relationship with Rhizobium supplies biologically fixed nitrogen directly to the plant, without requiring an excess pool of soil nitrogen, and the ammonia produced is rapidly and efficiently assimilated by the host plant. Three approaches to achieve a more direct transfer of biologically fixed nitrogen to rice plants have been proposed: 1. Development of novel symbiotic interactions resulting in the formation of nitrogen-fixing nodules or nodule-like structures on rice roots; 2. Identification of stably 33 n HA I '3' A “fl l " id 'VU \ .I‘Qfl":. ”I u .yyh ‘Iv auplA A f!‘ D . \ 1. vv lb I ‘ I . . -' a a, ll 5. d 'Nfi§0 \ ” «r it ‘ b “F 9'. a a v .1 “Ma... IA. u. :3‘ al\ v V a. “‘4.“ . ‘rt, "~35 : 3 "I I: K r ' U - I O I. . " M .. U. - '§l a: ”A - fl 1 U U“ ‘5 : ~l‘n ’u . ”S C \.‘. ‘x u 'v- \ V 0.. V - U.‘ 4"“ 5' I A ~.‘n ‘ v‘:: . h. I 1‘25 ,. - 1 v. “‘1'.“ .- D "uq-‘p A I‘: A' u ‘\ .v.‘ .I v'-" Q A . I". "I‘. in, I maintained diazotrophic endophytic bacteria in rice tissues; and 3. Direct incorporation/expression of the required complement of nitrogen fixation genes in(to) rice (Kush and Bennet, 1992; Ladha and Peoples, 1995; de Bruijn et al., 1995; Ladha et al., 1997). While each of these methods has distinct merits, much of the basic knowledge needed to successfully develop any of these systems is still lacking. The development of nodules on legumes is a highly developed program involving specialized genes in both the plant and bacteria (Fisher and Long, 1992). While great progress has been made toward characterizing the genes involved in this process and the functions of the proteins they express, there are still large gaps in our knowledge, and extensive hurdles to be overcome (de Bruijn et al., 1995). Exogenously applied hormones or cell wall degrading enzymes cause hypertrophies to form on the roots of several non- leguminous crops (Kennedy and Tchan, 1992). It has been reported that these structures can be invaded by diazotrophs, and that in some cases an increase in nitrogen uptake by the plant can be observed. However, these structures lack the complexity and specificity of nodules. More knowledge about the plant genes involved in nodulation and symbiotic nitrogen fixation and the function of their gene products is needed before the development of true nodulation of cereals can be seriously considered 34 4. [so ——‘ “u... y — v I l‘" 3 ." U o-an a . ."U .l!!‘ C - '5‘! v... . "r 7". u in. “a i» A II.‘ U ‘I (L) (de Bruijn et al., 1995; Reddy et al., 1997; Kennedy et al. 1997; Webster et al., 1997). The development of plants containing nitrogen fixing genes in their own genome may be even more complex (de Bruijn et al., 1995; Dixon et al., 1997). While most of the genes necessary for nitrogen fixation in bacteria are well characterized, the transfer of the genes to the plant genome, along with the appropriate expression of all these genes is beyond our current technical ability. In addition, the creation of the proper environment, in terms of oxygen concentration/supply, energy provision to the bacteria, and efficient ammonia assimilation within plant cells may also present serious problems (de Bruijn et al., 1995; Dixon et al., 1997). Therefore, the employment of endophytic nitrogen-fixing bacteria may involve the fewest technical challenges. Moreover, this approach has already been shown to be successful in the case of sugar cane (Boddey et al., 1995; Kirchhof et al., 1997) and Kallar grass (Reinhold-Hurek and Hurek, 1997). In addition, a beneficial natural endophytic association of rhizobia and rice grown in rotation with clover has recently been documented (Yanni et al., 1997). By using naturally occurring endophytes that colonize a niche in which conditions appropriate for 35 nitrogen fixation exist, we would have ‘allowed evolution to do some of the work for us’. Even if a stably endophytic microbe was to be identified that did not have the capacity to fix nitrogen, the process of introducing and expressing the nif gene complement in such an endophyte would be significantly easier than engineering the rice plant itself to fix nitrogen. However, it is clear that first some basic knowledge about the presence, predominance and stability of endophytic bacteria in different rice tissues must be obtained. Below we report some recent results of our ongoing study on rice-endophyte associations, that has the following objectives: a) isolating putative endophytic bacteria from diverse rice varieties grown in different soil types and assessing their diversity, b) developing molecular probes for the detection of putative N2-fixing endophytes, and c) studying the internal colonization of rice tissue by putative endophytic bacteria. 36 ll'ERilii ration I CID III I x) o... A..\ ’ 'v-v Vv II. P vc‘ D 5 If! H- v J. "'I A l" - " v. ‘0‘. A . a " v.4. . l"‘n H y. \ n I 'v.. 'H .'u MATERIAL AND METHODS Isolation of Endophytic Bacteria. The isolation of bacteria from 8 varieties of rice grown in the greenhouse in pots containing one of five soil types was carried out at IRRI (Malarvithi, 1995; Figure 2-1). The rice tissue was collected at the heading stage as follows. Plants were carefully removed from pots, washed to remove all soil and separated into stems and roots. The outer layer of the stems was removed, the stems were washed with tap water and deionized water, cut into sections 2-3 cm long, and dried on absorbent towels. Roots were cleaned thoroughly with tap water, rinsed with deionized water, and drained on absorbent towels. All tissues were surface- sterilized in the following manner. Ten grams of tissue was shaken for thirty minutes in a 500 ml Erlenmeyer flask containing 250 ml sterile deionized water and 25 grams of glass beads. The tissue was transferred aseptically to a sterile beaker, washed two times with sterile distilled water, and sterilized using 0.2% HgClz (30 seconds for roots, 60 seconds for stems). The tissue was washed 6 times with sterile distilled water, out into small pieces and homogenized in a Warring blender containing 90 ml sterile distilled water. Serial dilutions were prepared and spread on 37 plates containing bacterial growth media or tubes containing semisolid media described by Malarvithi (1995) and Barraquio et al. (1997). The isolation of bacteria using an in planta selection step to reduce the number of non-endophytes (Figure 2-2) was also carried out at IRRI. Initial isolation was carried out on field grown rice plants (variety IR72) as described above with the following modifications. The duration of surface sterilization of tissue was varied from 0 to 120 seconds. One hundred 2 dilution were inoculated microliters of macerate from the 10'1 or 10' onto three day old aseptically grown rice seedlings generated as follows: Rice seeds (variety IR24, |R42, or Lemont) were gently dehulled, placed in 70% ethanol for five minutes, washed with sterile distilled water, sterilized by addition of 0.2% HgCl2 for four minutes or a fresh solution of 30% Clorox (17.5 ml sterile distilled water, 7.5 ml Clorox Bleach, 30 pl tween 20) for 45 minutes, and washed six times with sterile distilled water. Seeds were placed on plates containing TY medium (Beringer, 1974) and incubated at 30°C for two days to allow germination and check for contamination. Seeds showing no contamination were placed in 25 mm X 200 mm tubes containing either 25 ml of semisolid modified Fahraeus medium (Fahraeus, 1957) or sterile sand watered with Fahraeus 38 medium. Seedlings grown in sand were watered with Fahraeus medium on a regular basis during the remainder of the growth period. Bacteria were isolated from the inoculated seedlings as follows. Seedlings were grown from 23 to 25 days before re-isolation of the bacteria and carefully removed from the growth tubes to keep the roots intact. The roots and shoots were separated and the shoots were cut into small sections. The tissues were placed in 125 ml Erlenmeyer flasks containing 10 ml of sterile distilled water and 5 grams of glass beads, shaken for 30 minutes, and washed with 10 ml sterile distilled water. The tissue was sterilized by adding 10 ml of 0.2% HgCI2 for 30 seconds or 30% Clorox solution for 15 minutes and washed six times with sterile distilled water. Tissues were macerated in one milliliter sterile distilled water and 20 pl of macerate was spread onto plates containing Tryptic Soy Agar (Difco, Detroit, MI, USA), 0.1% Tryptic Soy Agar, TY medium or Davis and Mingolini minimal medium (Atlas, 1997). Acetylene Reduction Assays Acetylene reduction assays were performed at IRRI using malate semi- solid nitrogen-free media and standard protocols (Barraquio et al., 1997). 39 PCR Amplification of nifD-Specific DNA Fragments Using sequence data from known diazotrophs (obtained from medline), the following universal nifD primers were derived: FdB261 (5'- TGGGGICCIRTIAARGAYATG-S') and FdB260 (5'-TCRTTIGCIATRTGRTGNCC- 3'). These primers were synthesized at the MSU macromolecular facility and tested using a reference collection of known nitrogen-fixing and non- nitrogen-fixing bacteria (Chapter 3). A DNA fragment 390 bp in length was amplified from all diazotrophic bacteria tested, using the whole cell ERIC-PCR conditions previously described (Louws et al., 1996; Rademaker and de Bruijn, 1997). The DNA sequence of the amplified DNA fragment from Mesorhizobium Ioti NZP2235 was determined by automated fluorescent sequencing of the purified DNA at the MSU-DOE-PRL Plant Biochemistry Facility using the ABI Catalyst 800 for Taq cycle sequencing and the ABI 373A Sequencer and shown to be highly homologous to nifD sequences of a variety of diazotrophs. The identity of the PCR amplified fragments from the reference collection of diazotrophic microbes was confirmed by standard Southern hybridization (Maniatis et al., 1982), using the amplified M. Ioti nifD fragment as a probe. 40 —..‘ I ‘5". rep-PCR Genomic Fingerprinting Whole cell rep-PCR of bacteria and computer analysis of the resulting fingerprints was carried out as described by Louws et al. (1996) and Rademaker and de Bruijn (1997). 168 rRNA Gene Sequencing and Phylogenetic Studies Two highly conserved eubacterial 168 rDNA primers (8F and 1492R) were used to amplify segments of the 168 rRNA genes of the members of the test collection (see Results and Discussion), using rep-PCR conditions. The PCR amplification products were purified using ultra-Free columns (Millipore, Bedford, MA, USA). Partial 16S rDNA sequences were obtained using primers 8F and 519RB using the ABI373A DNA sequencer (see above). The sequences obtained were submitted to the SSU RDP Database at the University of Illinois using Netscape Navigator and analyzed using the Similarity Rank method described by Maidak et al. (1994). Introduction of Transposons into Putative Endophytes and Detection of GUS Activity Transposons containing the beta-glucuronidase (uidA; gus) marker gene were introduced into the genome of the rice isolates by conjugation with Escherichia coli strains harboring narrow host range plasmids carrying the 41 respective transposons. Detection of Gus activity was carried out as described by Wilson et al. (1995). 42 '-.‘J‘l-. ‘l ‘E- I RESULTS AND DISCUSSION Isolation of Rice Endophytes In order to isolate putative endophytes of rice, two distinct approaches were used. The first approach involved the use of root and stem tissues of diverse varieties grown in different soil types, as described in the Material and Methods (Figure 21). One hundred thirty-three of these isolates were randomly selected for further study. In an attempt to enrich for invasive/endophytic bacteria a second approach incorporating an in planta selection step was utilized (see Material and Methods and Figure 2-2). This procedure led to the isolation of ~300 microbial isolates, of which 175 were brought to MSU for further analysis. Two of the isolates from this collection [R061.S1.3. and R032.82.3] were randomly chosen and included in the study presented below (Test Collection). Diversity of Putative Endophytes in Rice Tissues The diversity of a collection of one hundred thirty-three putative endophytic microbes isolated using the approach outlined in Figure 2-1 was assessed using rep-PCR genomic fingerprinting (Versalovic et al., 1994;1Louws et al., 1996; Rademaker and de Bruijn, 1997). Both REP 43 and ERIC primers were employed and the resulting fingerprints were combined and analyzed using the Gelcompar software package, as described by Rademaker and de Bruijn (1997). The dendrogram derived from this analysis is shown In Figure 2-3. The results presented reveal a large degree of diversity in these putative endophytes, although some distinct clusters of closely related strains could be observed (Figure 2-3). One large cluster of thirty-two similar fingerprints stands out in the dendrogram. The origin of these bacterial isolates is interesting from a standpoint of cosmopolitanism verses endemism. The thirty-two isolates in this cluster were isolated from eleven of the rice variety/soil types tested. Bacteria were isolated from plants grown in all five soil types, and from six of the eight rice varieties used. The largest number of similar fingerprints in the cluster that originated from the same variety/soil Combination was eight. The diversity of their origins suggests that the bacteria found in this cluster are cosmopolitan, being found in a wide distribution of geographically removed soil types and being able to Colonize a number of different rice genotypes. Three strains from the large cluster [-R6a(126), R33(120) and R45(42)] were selected for further analysis (Test Collection; see below). 44 ‘m—-— .‘K .- 'H Nitrogen Fixation by and Presence of nif Genes in the Endophytic Isolates The nitrogen fixing ability of the one hundred thirty-three putative endophytic bacteria from the first collection (Figure 2-1) was examined using the acetylene reduction assay (ARA). In addition, highly conserved DNA primers for PCR mediated detection of nifD genes were developed and used to screen the collection for the presence of nif genes. In the case of diazotrophic bacteria, PCR amplification of genomic DNA using these primers generally leads to the generation of a single 390 bp fragment, which has been shown to be nifD-specific by examining a reference collection of thirty-three well characterized bacteria selected from our laboratory strain collection. This collection included twenty-four Species of nitrogen-fixing bacteria and nine non-nitrogen-fixing bacteria. The identity of the amplified fragment was further confirmed by determining the DNA sequence of the amplified fragment from M. Ioti NZP 2235 and by Southern blot analysis (Chapter 3). Seventeen of the one hundred thirty-three strains fingerprinted (Figure 23) displayed a Characteristic amplification fragment when tested using the nifD primers. Of the seventeen strains shown to contain nifD using PCR, thirteen were found to reduce acetylene (Nif plus). Four additional isolates, which were scored as nifD minus in the PCR reactions, were nevertheless found to be 45 I'. "L_."I J I ARA plus, suggesting that they contain highly divergent nifD genes, or alternative nif genes. Figure 2-4 shows fingerprints and nifD PCR results for ten bacteria which were both nifD plus and Nif plus, four bacteria that were nifD plus and Nif minus, and four bacteria that were nifD minus and Nif plus. The remaining one hundred twelve strains were scored as nifD minus and Nif minus. Three of the isolates that were found to be both nifD and Nif plus [T105, R90(8), R100(64)] and one isolate that was nifD plus but Nif minus [R22(88)] were chosen to be included in the Test Collection (see below). Composition of a “Test Collection” and Preliminary Characterization of its Members A “Test Collection” of bacteria was composed for further studies (see Table 2-1). The nine isolates noted above [-R6a(126), R33(120), R45(42), T105, R22(88), R90(8), R100(64), R032.S2.3 and R061.S1.3] Were included, as well as five previously characterized “control” bacteria. Escherichia coli DH5a was chosen as a negative control because it is Unlikely that this bacterium would colonize rice roots endophytically. Azorhizobium caulinodans ORSS71 was chosen because it has been reported to invade the roots of wheat and rice (Cocking et al., 1995; see also Webster et al., 1997; Reddy et al., 1997). Sinorhizobium meliloti 46 1021 was included to compare a Rhizobium with Azorhizobium caulinodans ORSS71 in its ability to colonize rice roots. Azoarcus indigens LMG9092 and Herbaspirillum seropedicae 267 were included since these two bacteria have previously been reported to invade rice tissues (Reinhold-Hurek and Hurek, 1987). In order to further characterize the endophytic isolates from rice, a partial DNA sequence of their 165 rDNA was determined. The DNA sequences obtained were submitted to the SSU RDP database at University of Illinois using the Similarity Rank function (Maidak et al., 1994). The results of this analysis are shown in Table 2-1. The most similar 16s rDNA sequence in the database for each of the “control” bacteria was found to be “itself”, except for the case of Herbaspirillum seropedicae 267, where 1 68 rDNA sequences were not yet available for comparison. Re-colonization of Rice Tissue by Putative Endophytes To study the endophytic nature of the bacteria in the Test Collection, experiments were performed to verify that the same bacteria inoculated to sterile rice seedlings could be re-isolated from these seedlings (to fulfill Koch’s postulate), and to examine their endophytic competence (lnfeCtion and persistence characteristics). Gnotobiotically grown rice 47 seedlings were inoculated with each of the bacteria from the test collection and bacteria were re-isolated as described in Material and Methods (Figure 2-5). The identity of the colonies was verified by comparing the ERIC-PCR fingerprints of these bacteria with the ERIC-PCR fingerprint of the original inocula. In some cases, no bacteria could be re- Isolated from the inoculated seedlings. In other cases a large number of bacteria (>25,000), could be re-isolated from a single seedling. As expected, no bacteria were re-isolated from seedlings inoculated with Escherichia coli. Surprisingly, no bacteria could be re-isolated from seedlings inoculated with either Herbaspirillum seropedicae or Azoarcus indigens. In the case of the rice strains and other control strains large differences in numbers of re-isolatable bacteria were observed. However, large differences in the number of bacteria re-isolated from different Seedlings inoculated with the same bacteria were also observed. In spite Of these variations, certain endophytic isolates [R22(88), R33(120), R45(42), and R061 .813] appeared to be considerably more aggressive colonizers than others. The most aggressive colonizer [R22(88)] was found to cause stunting of seedling growth and therefore may be perceived by the plant as potential pathogen. Presently, the experimental 48 FF.‘ .' parameters of these tests are being further standardized and the experiments are being repeated using sufficient replicas to ensure statistically significant results. The analysis of bacteria re-isolated from seedlings inoculated with Azorhizobium caulinodans ORSS71, R22(88), R100(64), -R6a(126), R33(120), R45(42), R032.S2.3, and R061.S1.3 using rep-PCR genomic fingerprinting (see Figure 2-5 for the experimental approach) revealed that Koch’s postulate was fulfilled, in that bacterialwith the same fingerprint as the original inoculum strain could be re-isolated (Data not shown). Use of Marker Genes to Track Endophytic Bacteria in Rice Tissues To track bacteria during colonization of rice roots and subsequent inter- Or intracellular persistence, specific marker gene constructs were selected (Figure 2-6). A number of different marker genes are available for the tracking of genetically modified organisms (Jansson, 1995). Initially, we focused on the use of the beta-glucuronidase(uidA; gus) gene, since Gus activity can be easily detected using a colorometric Gus enzyme assays or in situ staining methods (Wilson et al., 1995). In order to obtain stably 49 ‘l"..-Im __ 'c- n tagged strains, we used the Tn59us transposon pCAM111 or pCAM121 kindly provided by Kate Wilson (Wilson et al., 1995). One or the other of these transposons was introduced into 8 out of the 14 test collection strains, and Gus activity of the tagged bacteria could be verified. The remaining six strains exhibit multiple intrinsic antibiotic resistances, and modified transposons/suicide vectors are being developed to tag these strains. Initial results with a limited number of tagged strains indicate that global colonization patterns of gus tagged microbes in rice roots can indeed be monitored by in situ Gus staining (see Chapter 6). However, the work required for sectioning tissues in order to localize individual bacteria using Gus activity may be prohibitive to the screening of large sample numbers. Therefore, in collaboration with the laboratory of J. Jansson at Stockholm University, we have developed Tn5 derived transposons carrying the gene encoding the green fluorescent protein (gfp), originally derived from the jellyfish Aequorea (Chalfie et al., 1994). Using the constitutively expressed Tn5gfp marker transposon, we have shown that GFP is stably and highly expressed in single cells of Pseudomonas fluorescence, and that single gfp tagged cells can be detected on the root hairs of plants 50 using laser confocal microscopy (Tombolini et al., 1997; Unge et al., 1998). The use of fluorescent laser scanning confocal microscopy allows optical sectioning reducing the work required to screen large sample numbers. We are presently attempting to tag the rice endophytes and control strains with this marker gene and use the resulting isolates to study rice plant colonization. Do the Proper Physiological Conditions Exist in Rice Tissues to De-repress nif Genes Harbored by Endophytic Bacteria? Even when stably maintained diazotrophic endophytes (endosymbionts) of rice are identified, it still remains to be shown that the proper physiological environment exists at the site(s) of colonization to allow expression of the nif genes and functioning of nitrogenase (de Bruijn et al., 1995; Dixon et al., 1997). As a first step in this analysis we have made use of a plasmid carrying a translational fusion of the 5’ end of the Azospirillum brasilense nifH gene (including its promoter region) to gus (Vande Broek et al., 1992; 1996). Preliminary results show that selected endophytic strains from the test collection harboring the nifH-gus fusion do express the reporter gene in the free living state under low oxygen tensions, as expected (Vande Broek et al., 1996). These strains will be used to infect rice seedlings, and in situ staining of infected tissues for 51 Gus activity will be performed in order to determine if and where the proper physiological conditions exist for nif gene de-repression, and therefore, by inference, for nitrogen fixation per se, although the latter correlation has not always been found to hold true (Vande Broek et al., 1996). 52 CONCLUSIONS In this paper we have described preliminary results from our first screens for and characterization of naturally occurring diazotrophic endophytes of rice. The approach we have used here differs from strategies to assess the potential for nitrogen fixation in rice we have reported previously (de Bruijn et al., 1995; Yanni et al., 1997). The latter strategies are based on the assessment of the potential of known rhizobia to induce (nitrogen- fixing) nodule-like structures or root hypertrophies on rice roots, or the ability of specific naturally-occurring endophytic rhizobia to stimulate rice growth under different nitrogen regimes and display acetylene reduction activity, respectively. As a basis for the screen described here, we reasoned (hypothesized) that rather than, or in addition to, attempting to induce or modify known rhizobia to engage in symbiotic or endophytic interactions with rice plants resulting in biological nitrogen fixation for plant growth enhancement, we would try to isolate and characterize naturally occurring, stable endophytic microbes from a variety of different rTtodern and more primitive varieties of rice, and subsequently determine their nitrogen fixation potential or genetically modify them to fix nitrogen. A major difference between this study and that of Yanni et al. (1997) was the strategy used to isolate diazotrophs from the rice endophytic 53 community. Whereas Yanni et al. used an apprOpriate legume trap host to specifically enumerate and isolate endophytic rhizobia from rice roots, this study used various plating media to isolate a more diverse collection of microbes belonging to the bacterial community that colonizes rice roots. In fact, these studies are complementary and resulted in the isolation of a large collection of rhizobial and non-rhizobial native microbes that are capable of developing endophytic relationships with rice. The results presented here, and those described by Barraquio et al. (1 997) indicate that a large diversity of apparently diazotrophic (~10%) and non-diazotrophic endophytic bacteria can be isolated from rice tissues, some of which are capable of re-colonizing their host when re- inoculated onto sterile rice seedlings. Past isolations of nitrogen-fixing bacteria from rice roots have already revealed a broad diversity of diazotrophs inhabiting the rice rhizosphere (Lakshmi Kumari et al., 1976; Watanabe and Barraquio, 1979; Bally et al., 1983; Ladha, 1986; Oyaizu- Masuchi et al., 1988; Yanni et al., 1997). In addition, recent studies based on the molecular phylogeny of the DNA sequences generated by PCRamplification of nitrogen fixing genes found in the rice rhizosphere 54 also suggest an even broader range of different rhizosphere diazotrophs (Ueda et al., 1995a; 1995b). However, the ability of these bacteria to colonize rice tissue endophytically and de-repress genes needed for nitrogen fixation remains virtually unstudied. Having developed the tools to mark and track some of the promising stable endophytes and monitor the physiological environment in their colonization niches with regard to expressing nif genes, we are now able to further test our hypothesis that natural or genetically modified rice endophytes may be useful to achieve the long term goal of developing rice-microbe interactions capable of supplying biologically fixed nitrogen for higher yields, in the absence of additional chemical fertilizer. In addition, information gained during these studies may yield interesting new insights into basic mechanisms of plant-microbe interactions and microbial colonization of plant tissues in general, and be helpful in designing strategies for the discovery and adaptation of nitrogen-fixing bacteria that could provide nitrogen to other major cereal crops such as maize and wheat. 55 ACKNOWLEDGMENTS We would like to thank Dr. Kate Wilson for her gift of several GusA A transposons for use as molecular markers, Jan Rademaker for his help learning to use Gelcompar 4.0, Dr. Tom Schmidt for his help with the analysis of the SSU rDNA sequence data and Mary Ellen Davey for her contributions to the GFP portion of this project. This project has been supported by grants from the Danish International Development Agency (DANIDA), and the US Department of Energy (DE-FGO2-91 ER20021). 56 @00830 .<.00. 93:0: .3 52> m00 ~30 7:00. m.3:0_. 25 32> M00 m00303.03..0 00:01.00 50303030 00... 0.000 0100303030 00... 0:000. x-.m 0:. .<.Q.mmm 0.000 9303.30.03 3030:. . om. 9303.30.03 x..3....03m030..0 0.000 9303.30.03 3030:. 0.000 033.3033 000300030 0.0mm.) 033.3033 0003300030 0.3m >m:0000~0w 0m..3.~030..0 0.3 m2.-. 0.3. 330800 306030 :30. meow 3304000 330030 0.... 93:03:00 30.020 :0300 .3 .30 :0. 00.:33 .2010 :000 8 .00. <030:0 30.300 .0: 0<0.:0:03 0. 00.03.N0:03. 003.0. mm: «02> 00::03000 .83 3000 000.030 5040 0:03.300 .0 .30 mm: .000 0000000 .<0:0.03 mb. 0. 3200.2 0. ....30.0 :0.30 93:03:. 303: A<040.03 0.0.. 30030 .233 30. 030 000030 300.. 03:04 mm: 62> 00::03000 .3 .30 0.0 03023. ...30 03:01.2 $3: 830:03 00 .30 mmc .000 :300 .30 00.00000 00::03000 .30. 040 300. 03:04 .0 0 00::0300 0:03.300 3 .30 :00: ...30 08903 00:30 .30 3:300: 0* 3.90 w 0000 0.60300 .3 .30 0:03.300 00::0300 .30. 00 003303 .0 0003 00::0300 .3 30 00.0 0000. ...30 M00 <0.:0 .0 30 «00:: 03600000 00 .30 3:300“ 00 03000 .<. 0000 0.603040 02.000 0< 0:39 .30 3:300: 0. 3.05 0.60300 .3 .30 0:03.300 00::0300 04 30 00000000 000.:0300. <<3.030<0_. .0 .9203 3.3.0 .0 30. 0 03.38.0303 030.<0.0_ 0:0 0:20 000000303. 0.. .30 03:03:00 00.200: .30 00::0300. 1.3.0 000020 .0 c.<03 .30 3030 33033.03 .3 .30 moo 0060000. ...0<<0<0n 3 .0 30. 0.000.: «0.0000 .0 0.304 0303333 00000 03 33.000330 .4000 3000 :0.3© mmc «02> 00::03000. 57 Pot Grown Plants Roots Macerate in 8 Varieties Blender 5 Soil Types Surface Sterilize I i i? e HU++' Plate on 0.1%TSA, ~1000 Isolates ~300 Isolates LPIG, or Congo Fled: from Plates from Semi-Solid Semi-Solid Media (Malate or Rice exudate) Figure 2-‘l Isolation of Putative Rice Endophytes Scheme to isolate putative endophytic bacteria from rice tissues. Surface sterilization eliminates/ reduces the presence of non- endophytic bacteria and maceration releases endophytic bacteria from the tissues. Over 1000 bacteria were isolated at IRRI using this procedure . 58 £01- Stems Roots Field Grown Plants Surface Sterilize for 0, 30, 60, Macerate in Variety |R72 Grown or 120 Seconds Blender at IRRI I "2" ‘9 ‘ u u -> _> lnoculate Sterile Surface Sterilize Macerate in Plate on 0.1%TSA. Seedlings Blender TSA, or Davis Min. with Macerate ~300 Bacteria Figure 2-2 Isolation of Endophytic Bacteria Using a Re- infection Step Scheme adds an in planta selection step to eliminate non-endophytic bacteria. Approximately 300 bacteria were isolated at IRRI using this procedure. 59 . _ ,1. . g 3. wins" A-. .. l:-‘Ill:‘-- -r g r u u . - u > E I- - -- I It'- .. _ ..: :::u Ill ----- - ----- .Illl 1-277: -- v -"- '9 . J“: I I- II .f'i__ ___ ;_. _ _ . I." I all -u I A g“ 1. " -- .. 4"; -...- - .. . g '- i I . . _' _. ._ _—-.. - ........ I. |_.._. .. I ." :fi-n :I III-l- I , u. ...... . . —__- L u-o-I ............................... ;- - ~-~ .. .............. _ ; ~ - .- . -------- . .......... . ............. _!g:- -.- ‘:‘. I I u ‘ - __ I ' 9=-' ‘ |. I. ‘ - "3 " In. -- - 0' - ’ _. : - . '1. I .. _. " ans-al. ;.. . ... '. - nu '3‘ 1 ' . ‘ . ‘ 3 ';';:: u... ' 'r -. 'E .--..., _.--.__- .. .. ..... _. . . uJ-i-l n. _ . .. . . . . _ 8! .- , . . '..""”3‘ '..- __ _‘ I‘ n . . n - I . ' ' -..r ‘ gluon... _ :: 0...“. " " I... Figure 2-3 Gel Profiles and Dendrogram of rep-PCR Generated Genomic Fingerprints of Putative Endophytic Bacteria Isolated from Rice Roots The dendrogram of 133 putative endophytes from rice generated from the combined BOX- and ERlC fingerprints using the Gelcompar 4.0.program is shown The fingerprint patterns revealed a high degree of diversity, as well as several clusters of related bacteria. Analysis was performed using Pearson correlation applied to the densitometric curves formed by the fingerprints followed by Clustering analysis using the unweighted pair-grouping method, using arithmetic averages (UPGMA). 60 i++++++++m 6 Br 4 l Iiiii‘ii i” 3|? I H 5 R92t60J - “Hi-iii I H i ”I" T93 + Hiiiilii . I -' Elli. Rsouog) + FL illli if H: l __.' Argo + i!!!“ .I Hi! [31H "'3, we - l? liliill :1 i HI :. R100(64) + iiitilEé-Eli ii i : if! 3 1 1190(8) + “Hill?“ 1 iii-I if Ml .‘ mum) + ‘ iilflilli i 'I il l 5! -R11C(59) + _i:‘: illillll é iii llti 4116:“) + eliililill 1 ll 3 it i Rllct'm + .i iiifiil it i i :5 fl i 4. R22(88) + L if mm 1 2| m mum: — 3 i Ezliii'iiiiiill ililSl El :l llll l ,1 K113 + l ii iii 1| - ll , , nos + 3| “H Hiii-i l I ill 31 .. 1" R99(65) + _{‘—i:: iiillillsi iiflui l llli mm) + lii‘ ultra: lill f illli ' R9410 - Figure 2-4 rep-PCR Genomic Fingerprints and nifD-PCR Analysis of Putative Diazotrophic Endophytes The dendrogram of the rep-PCR genomic fingerprints of 18 putative nitrogen-fixing strains is shown. The dendrogram was generated using the combined BOX and ERIC and REP fingerprint patterns and the Gelcompar 4.0 software program. This dendrogram shows both the diversity of organisms, as well as clustering of related bacteria. A representative gel on which nifD -PCR fragments were electrophoretically separated is shown below the fingerprints. (+) indicates bacteria from which an amplified nifD fragment could be Obtained. (-) indicates bacteria from which no amplified nifD PCR fragment could be obtained. Results from the acetylene reduction assays are also presented. (+) indicates bacteria capable of reducing acetylene in semi-solid malate media. (-) indicates bacteria which did not reduce acetylene in semi-solid malate media. 61 ++++++III Grow Liquid Cultures Inoculate Sterile Surface Sterilize Seedlings of Purified Putative Seedling§ Endophyte With Ligurd Culture and Grow 3 Weeks - Q I I l + = - - -- = - - — — - _ oll Illll i _ _ _ " _ _ = : Macerate in Plate Macerate and Conpare Fingerprints of Blender Grow Colonies Original Strains with Pie-isolated Bacteria Figure 2-5 Re-infection Potential of Putative Endophytes The scheme for re-infecting rice plants with and re—isolation of putative endophytes is shown. This experiment was carried out to fulfill Koch's postulate that true endophytes should be re-isolatable after infection of gnotobiotically grown rice seedlings. After re- isolation the identity of the bacteria was confirmed using rep-PCR genomic fingerprinting. 62 Constitutive gus A Promoter Marker . . GFP o Constitutive Promoter Marker "’7” gusA Promoter Marker .' Transform Putative Grow Liquid Mini Tn5 Molecular Marker Gene Fusions Endophytes with Cultures of Marked Marker Constructs Putative Endophytes l & & Stain for Gus Observe Invasion and nifH Inoculate Sterile Seedlings with Liquid Culture and Grow 3 Weeks Activity or visulalize Promoter Induction using GFP with UV light Microscopy Figure 2-6 In planta Visualization of Endophyte Colonization in Rice The scheme for the utilization of the molecular markers beta- glucuronidase (GUS) and green fluorescent protein (GFP) to characterize endophyte invasion of rice tissues is shown. A Constitutive promoter fused to the gus or GFP reporter gene will be used to visualize colonization. A nifH promoter fused to the gas reporter gene will be used to monitor nif gene expression during Colonization. 63 REFERENCES Atlas FM (1997) Handbook of Microbial Media. 2nd ed, CRC Press, inc. Boca Raton, Florida, pp 401 Bally R, T homas-Bauzon D, Heulin T, Balandreau J (1983) Determination of the most frequent N2-fixing bacteria in the rice rhizosphere. Can J Microbiol 29: 881-887 Barraquio WL, Revilla L, Ladha .K (1997) Isolation of endophytic diazotrophic bacteria from wetland rice. Plant and Soil 194: 15-24 Beringer JE (1974) R factor transfer in Rhizobium leguminosarum. J Gen Microbiol 84: 188-189 Boddey RM, de Oliveira OC, Urquiaga S, Reis VM, de Olivares FL, Baldani VLD, Dobereiner J (1995) Biological nitrogen fixation associated with sugar cane and rice: Contributions and prospects for improvement. Plant and Soil 174: 195-209 Bohlool BB, Ladha JK, Garrity DP, George T (1992) Biological nitrogen fixation for sustainable agriculture: A perspective. Plant and Soil 141:1-11 Chalfie M, Tu Y, Euskirchen G, Ward WW, Prasher DC (1994) Green fluorescent protein as a marker for gene expression. Science 263: 802-805 Clement A, Ladha JK, Chalifour FP (1995) Crop residue effects on nitrogen mineralization, microbial biomass, and rice yield in submerged soils. Soil Sci Soc Am J 59: 1595-1603 Cocking EC, Kothari SI, Batchelor CA, Jain S, Webster G, Jones J, Jotham J, Davey NR (1995) Interaction of rhizobia With non-legume crops for symbiotic nitrogen fixation nodulation. In | Fendrik, M del Gallo, J Vanderleyden and M de Zamaroczy, eds, Azospirillum VI and related microorganisms: Genetics, physiology, ecology. Springer-Verlag, New York, pp 197-205 64 de Bruijn FJ, Jing Y, Dazzo F8 (1995) Potentials and pitfalls of trying to extend symbiotic interactions of nitrogen-fixing organisms to presently non-nodulated plants, such as rice. Plant and Soil 174: 225-240 Diekmann KH, Ottow JCG, De Datta SK (1996) Yield and nitrogen response of lowland rice (Oryza sativa L.) to Sesbania rostrata and Aeschynomene afraspera green manure in different marginally productive soils in the Philippines. Biol Fertil Soils 2 1: 103-1 08 Dixon R, Cheng Q, Shen G-F, Day A, Dowson-Day M (1997) Nif gene transfer and expression in chloroplasts: Prospects and problems. Plant and Soil 194: 193-203 Fahraeus A (1957) The infection of clover root hairs by nodule bacteria studied by a simple glass slide technique. J Gen Microbiol 16: 374-381 Fisher RF, Long SR (1992) Rhizobium-plant signal exchange. Nature 357: 655-660. Harris GH, Hesterman OB, Paul EA, Peters SE, Janke FR (1994) Fate of legume and fertilizer nitrogen-15 in a long-term Cropping systems experiment. Agron J 86: 910-915 Hossain M, Fischer KS (1995) Rice research for food security and sustainable agricultural development in Asia: Achievements and future challenges. GeoJournal 35: 286-295 Jailsson JK (1995) Tracking genetically engineered microorganisms in nature. Curr Opin Biotechnol 6: 275-283 Kannedy IR, Tchan Y (1992) Biological nitrogen fixation in non- \eQUminous field crops: Recent advances. Plant and Soil 141: 93- ‘\ 1 8 65 Kennedy IR, Pereg-Gerk LL, Wood C, Deaker R, Gilchrist K, Katupitiya S (1997) Biological nitrogen fixation in non-leguminous field crops: Facilitating the evolution of an effective association between Azospirillum and wheat. Plant and Soil 194: 65-79 Kirchhof G, Reis VM, Baldani Jl, Eckert B, Dobereiner J, Hartmann A (1997). Occurrence, physiological and molecular analysis of endophytic diazotrophic bacteria in gramineous energy plants. Plant and Soil 194: 45-55 Kumarasinghe KS, Zapata F, Kovacs G, Eskew DL, Danso SKA (1986) Evaluation of the availability of Azolla N and urea N to rice using 15N. Plant and Soil 90: 293-299 Kush GS and Bennet J eds, (1992) Nodulation and nitrogen fixation in rice: Potentials and prospects. IRRI, Manila Ladha ..K (1986) Studies on nitrogen fixation by free-living and rice-plant associated bacteria in wetland rice field. Bionature 6: 47- 58 Ladha J, Pareek R, Becker M (1992) Stem-nodulating legume- Hhizobium symbiosis and its agronomic use in lowland rice. Advan Soil Sci 20: 147-192. Ladha JK, Peoples MB (1995) Management of biological nitrogen f iXation for the development of more productive and sustainable agricultural systems. Plant and Soil 174: 1-286 Ladha JK, de Bruijn FJ, Malik KA (1997) introduction: SSessing opportunities for nitrogen fixation in rice — a frontier Protect. Plant and Soil 194: 1-10. Lakshmi Kumari M, Kavimandan SK, Subba Rao NS (1976) occurrence of nitrogen-fixing Spirillum in roots of rice, sorghum, maize. and other plants. Indian J Exp Biol 14: 638-639 66 q ADA, - 0...“. ml liaiarvi lv ‘1» U liniati l I :F'i': luv . n a l \ "l,“ P I. i v itione ... '-\_:n UV”. iiaizu. "u'fzn ..:. if.” t. F»- V . UV" Louws F J, Schneider M, de Bruijn FJ (1996) Assessing genetic diversity of microbes using repetitive-sequence-based PCR (rep-PCR). In G Toranzos, ed, Nucleic Acid Amplification Methods for the Analysis of Environmental Samples. Technomic Publishing 00., Lancaster, pp 63-93 Maidak BL, Larsen N, McCaughey MJ, Overbeek R, Olsen GJ, Fogel K, Bandy J and Woese CR (1994) The Ribosomal Database Project. Nuc Acids Res 22: 3485-3487 Malarvithi PP (1995) Association of endophytic diazotrophs in rice genotypes from various soil environments and estimation of their BNF potential using 1 5N dilution methods. Ph.D. thesis submitted to the University of Philippines. pp 195. Maniatis T, Fritsch EF, Sambrook J (1982) Molecular cloning: A laboratory manual. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. National Research Council (1994) Biological Nitrogen Fixation: Research Challenges. National Academy Press. Washington DC. Oyaizu-Masuchi Y, Komagata K (1988) Isolation of free-living nitrogen-fixing bacteria from the rhizosphere of rice. J Gen Appl Microbiol 34: 127-164 Rademaker JLW, de Bruijn FJ (1997) Characterization and Classification of microbes by rep-PCR genomic fingerprinting and Computer assisted pattern analysis. In G Caetano-Anolles, PM resshoff eds, DNA markers: Protocols, Applications, and Overviews. J- Wiley & Sons, Inc. New York, pp 151-171 Reddy PM, Ladha JK, So RB, Hernandez RJ, Ramos MC, Angeles OR, Dazzo FB, de Bruijn FJ (1997) Rhizobial communication with rice roots: Induction of phenotypic changes, (“Ode of invasion and extent of colonization. Plant and Soil 194: 31-89 Reinhold-Hurek B, Hurek T (1997) Azoarcus spp. and their interactions with grass roots. Plant and Soil 194: 57-64 67 limbolii .mm a , Juli q A rz: m uvddd I: .J Ueda T, ‘iZ-‘i‘fic - ,. i .(q i J . "v dt'u’i 7134-14 2" ‘n 1.“) 1'- .. a...” A \ f ‘9': 1‘ l -. C K. ,:n. ., t p 'v: .c. 'v an .1 V": ‘o Roger PA and Ladha .K (1992) Biological N2 fixation in wetland rice fields: Estimation and contribution to nitrogen balance. Plant and Soil 141: 41-55 Tombolini R, Unge A, Davey ME, de Bruijn FJ, Jansson .K (1997) Flow cytometric and microscopic analysis of GFP-tagged Pseudomonas fluorescens bacteria. FEMS Microbiol Ecology 22: 17- 28 Ueda T, Suga Y, Yahiro N, Matsuguchi T (1995a) Remarkable N2-fixing bacterial diversity detected in rice roots by molecular evolutionary analysis of nifH gene sequences. J Bacteriol 1 77: 1414-1417 Ueda T, Suga Y, Yahiro N, Matsuguchi T (1995b) Genetic diversity of N2-fixing bacteria associated with rice roots by molecular evolutionary analysis of nifD Library. Can J Microbiol 41: 235-240 Unge A, Tombolini R, Davey ME, de Bruijn FJ, Jansson ..K (1998) GFP as a Marker Gene. In ADL Akkermans, JD van Elsas, FJ de Bruijn, eds, Molecular Microbial Ecology Manual. Kluwer Academic Publishers, Dordrecht, pp 6.1.1321-16 Vande Broek A, Michiels J, de Faria S M, Milcamps A, Vanderleyden J (1992) Transcription of the Azospirillum brasilense nifH gene is positively regulated by mm and mm and is negatively controlled by the cellular nitrogen status. Molecular and General Genetics 232: 279-283 Vande Broek A, Keijers V, Vanderleyden J (1996) Effect of Oxygen on the free—living nitrogen fixation activity and expression of the Azospirillum brasilense nifH gene in various plant-associated diazotrOphs. Symbiosis 21: 25-40 \le'salovic J, Schneider M, de Bruijn FJ and Lupski .R 0994) Genomic fingerprinting of bacteria using repetitive sequence—based polymerase chain reaction. Methods in Molecular and Cellular Biology 5: 25-40 68 Waiana "a. .7 " Cw EL «:05 I IV. I ‘i‘ebste l. Koth I I. 'Frl. .“’A ‘ ' liini Y ”fillings itckle 2"“ n. 'd‘. 'IL U A A H Watanabe I, Barraquio W (1979) Low levels of fixed nitrogen required for isolation of free-living Ng-fixing organisms from rice roots. Nature 277: 565—566 Webster G, Gough C, Vasse J, Batchelor C A, O’Callaghan K J, Kothari S L, Davey MR, Dénarié J, Cocking EC (1997) Interactions of rhizobia with rice and wheat. Plant and Soil 194: 1 15-122 Wilson K, Sessitsch A, Corbo J C, Giller K E, Akkermans A D L, Jefferson R A (1995) B-Glucuronidase (GUS) transposons for ecological and genetic studies of rhizobia and other Gram-negative bacteria. Microbiol 141: 1691 —1 705 Yanni YG, Rizk RY, Corich V, Squartini A, Ninke K, Philip- Hollingsworth S, Orgambide G, de Bruijn F, Stoltzfus J, Buckley D, Schmidt TM, Mateos PF, Ladha JK, Dazzo F8 (1997) Natural endophytic association between Rhizobium leguminosarum bv. trifolii and rice roots and assessment of its potential to promote rice growth. Plant and Soil 194: 99-114 69 CHAPTER 3 DEVELOPMENT AND USE OF POLYMERASE CHAIN REACTION (PCR) PRIMERS FOR THE RAPID DETECTION OF NITROGEN-FIXING BACTERIA ABSTRACT A set of PCR primers corresponding to highly conserved regions of the nitrogen fixation nifD gene, encoding the a-subunit of the nitrogenase MoFe-protein, was developed using amino acid sequence data derived from a variety of different diazotrophs. This primer set was shown to amplify a characteristic 390 bp DNA fragment from diverse genera of diazotrophs. Southern hybridization and DNA sequence analysis confirmed that the 390 bp fragments corresponded to the nifD gene. This primer set was used to screen a collection of 142 bacteria isolated from surface-sterilized rice tissue. Twenty potential nitrogen-fixing bacteria were identified. Genomic Southern hybridization confirmed the presence of nif genes in 19 of these bacteria. The ability of these 19 bacteria to fix nitrogen was verified using acetylene reduction assays. This primer set was also used to detect the presence of diazotrophic bacteria in laboratory and fieldgrown rice plants. 70 PINS -‘ I 94v. "it: n c'u’v ‘3 “Mn— .. ' WU.- v l- 4“ r. «I. t. 'fl’ Suzn . ,'.. 0,.“ "JJ' Pr- c i." Q I‘h‘. i“. r4: K I 'U Q ,., no, ‘V l.- 'vt'na 0‘. A INTRODUCTION Biological nitrogen fixation constitutes an essential part of the global nitrogen cycle and contributes more fixed nitrogen to agricultural production than any other source (Bockman, 1997). However, in the case of wetland rice available levels of natural biologically fixed nitrogen are not sufficient for maximum yield and additional nitrogen must be supplied in the form of industrially produced fertilizer or green manure (Ladha and Kundu, 1997). It is an important goal of agricultural scientists to decrease the amount of industrially- produced fertilizer nitrogen needed by increasing the amount of biologically-fixed nitrogen available to rice crops (Introduction to Chapter 2; Bohlool et al., 1992; Ladha and Kundu, 1997). Recent studies have shown the presence of diazotrophic bacteria colonizing the internal tissues of non-legume plants such as sugar cane and Kallar grass (James and Olivares, 1998; Reinhold-Hurek and Hurek, 1998a; 1998b). In the case of sugar cane these endophytic bacteria have been proposed to be able to supply the plants with enough nitrogen for sustained high yields (Boddey, 1995). 71 Currently, little information is available about diazotrophic bacteria that colonize the cortex and/or vascular tissue of field grown rice plants. Studies have been carried out to determine the numbers and types of diazotrophic bacteria present in the rhizosphere of rice (Lakshmi Kumari et al., 1976; Watanabe and Barraquio, 1979; Thomas-Bauzon et al., 1982; Bally et al., 1983; Ladha et al., 1983; Watanabe et al., 1987; Oyaizu-Masuchi and Komagata, 1988). These studies relied on the use of selective media and acetylene- reduction assays (ARA) to detect rice-associated bacteria with nitrogen-fixing capability (Watanabe and Barraquio, 1979; Bally et al., 1983; Oyaizu-Masuchi and Komagata, 1988). Unfortunatley, the use of nitrogen-free media is prone to error since some bacteria which do grow are nevertheless unable to fix nitrogen and instead grow by scavenging trace amounts of mineral nitrogen from the media. ABAs are time consuming and may not detect diazotrophic bacteria if the proper physiological growth conditions, including proper carbon source, oxygen concentration, and nitrogen levels are not used for the assay. 72 PCR technology (Saiki et al., 1988) has revolutionized screening of large collections of uncharacterized bacteria for DNA sequences of interest (Steffan and Atlas, 1991). For example, universal PCR primers have been developed that allow detection of Listeria monocytogenes by amplification of a DNA fragment from the haemolysin gene (Furrer et al., 1991) and Agrobacterium by amplification of a DNA fragments from vir and ipt genes (Hass et al., 1995; Sawada et al., 1995). Ueda and coworkers assessed the diversity of diazotrophic bacteria associated with rice by amplifying and sequencing nifH and nifD DNA fragments from total DNA isolated from the rhizosphere of rice. (Ueda et al., 1995a; 1995b). However, at the onset of the work described here, PCR primers suitable for screening large collections of bacteria to identify nitrogen-fixing strains had not yet been developed. Therefore we set out to develop a quick, inexpensive, reliable method to screen large bacterial collections for nitrogen-fixing strains. The ability of all known diazotrophic organisms to convert atmospheric N2 into biologically usable forms depends on the activity of the enzyme nitrogenase. The nitrogenase complex contains two 73 components: an iron-containing homodimer and a molybdenum- and iron-rich r1282 tetramer. The genes encoding these proteins are nifH, nifD, and nifK, respectively (Kim and Bees, 1994) Therefore, the presence of DNA sequence corresponding to these genes in a given bacteria would suggest that the microbe may have the potential to fix nitrogen and constitutes a useful and rapid pre-screen for diazotrophs. Some diazotrophic bacteria contian additional nitrogen fixing genes which encode alternate forms of nitrogenase which utilize different metal cofactors. Sequence comparisons indicates that the nifD PCR primers would not amplifiy DNA fragments from these genes. However, all diazotrophs which are known to contain alternate forms of nitrogenase also harbor nifD (Chisnell et al., 1988; Schuddekopf et al., 1993; Thiel et al., 1993; Zinoni et al., 1993). Therefore, the nifD primers would still be useful for detecting these bacteria. However, additional undiscovered methods of nitrogen- fixation which do not rely on nitrogenase may be present in nature. These primers would not detect diazotrophhic organisms which utilize non-nitrogenase systems. 74 Here we describe the generation of a degenerate PCR primer set corresponding to the nifD gene that can be used to amplify a 390 bp DNA fragment from nifD, allowing specific detection of this nitrogen- fixing gene from diverse genera of diazotrophic bacteria. This primer set was used to identify novel nitrogen-fixing strains from a collection of bacteria isolated from surface-sterilized rice. This primer set was also used to detect nifD gene sequences in total DNA isolated from rice tissue. 75 MATERIAL AND METHODS Bacterial Strains The bacterial strains used to examine the efficacy of the nifD-PCR primers are listed in Table 3-1. The 142 stains isolated from surface- sterilized rice tissue and screened with the nifD-PCR primers are described in Chapter 4 of this work and listed in Appendix 1. The 35 strains characterized using ABA and Southern hybridization are listed in Table 3-2. Plasmid pRSZ carries the Azorhizobium caulinodans nifHDK genes (Elmerich et al., 1982). Media and Growth Conditions. E. coli DH5a harboring pRSZ was cultured at 37°C on LB media containing 100mg/l ampicillin. All other bacteria used in this study were routinely cultured at 28°C on TY media (Beringer, 1974). PCR Template Preparation and Amplification Screening of bacteria using nifD-PCB was carried out using ERIC-PCR conditions as described by Louws et al. (1996) and Rademaker and de Bruijn (1997), with the following modification. Template for PCR was generated by filling a 1 pl disposable loop with a bacterial colony 76 grown on a plate. The contents of the loop were resuspended in 100 pl sterile distilled H20 (sdH20) in a 1.5 ml microfuge tube, boiled for 10 min, incubated at 37°C for 1 hour with 5 pg of RNAase A and stored at —20°C. One microliter of this mixture was used as template for PCR reactions. Whole cells or purified DNA were used to test the efficacy of the nifD primers with different template types. Prior to DNA isolation, rice tissue was subjected to vigorous shaking with glass beads to remove rhizoplane bacteria (Watanabe and Barraquio, 1979). Rice DNA was isolated from rice tissue following the protocol of Ueda et al. (1995a; 1995b) with the following modifications: Following treatment with RNase A, samples were phenol/chloroform extracted, isopropanol precipitated, and resuspended in TE. The DNA was quantified using ethidium bromide staining, as described by Sambrook et al. (1989). 200 ng of the rice DNA was used as template for PCR reactions. nifD-PCR Primer Design Degenerate oligonucleotides FdB261 5'-TGGGGICCIRTIAARGAYATG- 3', and FdB260 5'-TCRTT|GC|ATRTGRTGNCC-3 were synthesized by 77 the Macromolecular Structure, Synthesis and Sequencing Facility at Michigan State University (N = any base; I = inosine; Y = T or C; R = A or G). The sequence was derived from an alignment of amino acid sequences from the alpha subunit of the nitrogenase MoFe-protein (NifD; Figure 3-1) using the pile-up function from the GCG software package (Genetics Computer Group, University Research Park, Madison, WI, USA) at Michigan State University. The primer set (FdB261/FdB260) was designed to amplify a fragment of approximately 390 bp in length from the 3’ end of the nifD gene. A schematic of the primer design is presented in Figure 32 DNA Analyses of nifD-PCR Products Visualization of the PCR products was achieved by adding two microliters of loading dye (0.25% w/v bromophenol blue, 0.25% w/v xylene cyanol, 15% ficol type 400 in 10xTAE) to the total reaction mixture and separating one fourth to one half of this mixture on 2.0% agarose gels in TAE buffer. The gels were stained with ethidium bromide, and DNA fragments were visualized on a transilluminator. The size marker used was the 100 bp ladder of Boehringer Mannheim (Indianapolis, IN, USA). 78 DNA sequencing of the 390 bp fragment amplified from M. Ioti NZP2235 DNA was carried out using automated fluorescent sequencing as performed by the MSU DNA Sequencing Facility using an ABI Catalyst 800 for Taq cycle sequencing and an ABI 373A Sequencer for the analysis. Prior to sequencing the PCR reaction was run on a 2.0% agarose gel and the 390 bp fragment was excised and extracted from the agarose using phenol/T E. The DNA was further purified using a Millipore 30,000 ultrafree NC column (Millipore, Bedford, MA, USA). Southern hybridization analysis of the amplified DNA fragments from the strains listed in Tables 3-1 and 3-2 was performed using the DIG DNA labeling and detection (Boehringer Mannheim, Indianapolis, IN, USA). The purified DNA fragment amplified from M. Ioti NZP 2235 was used as hybridization probe. Southern Hybridization Analysis of Total Genomic DNA Bacterial genomic DNA was isolated using the MasterPure Genomic DNA Purification Kit from Epicenter Technologies (Madison, WI, USA) according to the protocol provided by the manufacturer. DNA 79 digestion, separation, and transfer to Hybond-N nylon membranes (Amersham International, Little Chalfont, Buckinghamshire, England) were carried out according to standard protocols (Sambrook et al., 1989). A 4 Kb DNA fragment containing A. caulinodans nifHDK was excised from plasmid pRS2 (Elmerich et al., 1982) by digestion with Bgll and Sell run on an agarose gel and purified using the Qiaex ll Gel Extraction Kit from Qiagen (Madison, WI, USA). Probe preparation was carried out according to standard protocols, using 32P labeled dATP. Prehybridization and hybridization reactions were carried out at 65°C using buffer containing 5x Denhardts solution, 4x SSC, 1% SDS, and 100 ug/ml salmon sperm DNA (Sambrook et al., 1989). The membranes were washed three times at low stringency (65°C in buffer containing 4x SSC and 0.1% SDS). Acetylene Reduction Assays The bacteria listed in Table 3-2 were examined for nitrogenase activity using acetylene reduction assays. Bacteria were grown to late exponential phase in liquid TY media at 28°C with shaking. The cells in one milliliter of culture were pelleted in a microfuge and the supernatant was removed. The cells were washed two times with nitrogen-free LSO (Elmerich et al., 1982) and resuspended in 80 nitrogen-free LSO to a final 00600 of 0.8. Vacutainer tubes containing 3 ml of nitrogen free LSO plus 0.2% agar were inoculated with 30 pl of this suspension. Following incubation at 28°C for 24 hours, the tubes were stoppered, 1 ml of headspace was removed and replaced by 1 ml of acetylene. The tubes were returned to 28°C and 50 pl of headspace was extracted following 24 and 48 hours of incubation. Samples were analyzed using a Varian Model 3700 Gas Chromatograph equipped with a Propack N 80/100 2m x 1/8” Stainless steel column (Varian, Palo Alto, CA, USA) and a flame ionization detector. The injector temperature was set to 190°C, the column temperature was set to 69°C, and the detector temperature was set to 250°C. Output from the gas Chromatograph was measured with a Perkin-Elmer R 100A recorder (Perkin-Elmer, Norwalk, CT, USA). The ethylene peak height was compared to a standard ethylene curve to determine the moles of ethylene produced per tube. Growth, Sterilization and Inoculation of Rice Seedlings. Dehulled rice seeds were sterilized as follows. Dehulled seeds rice (Variety |R42 from International Rice Research Institute stocks, the Philippines) were immersed in 70% EtOH for 4 minutes with gentle 81 agitation followed by washing in SdHQO, immersion in 0.1% HgCI2 for 4 minutes with gentle agitation, and six washes with sdH20. After sterilization, seeds were placed on plates containing Nutrient Agar (Difco) and incubated for 2 days at 30° Celsius to allow germination and check for contamination. Non-contaminated seedlings were transferred to sterile test tubes (2.5 cm x 20 cm) containing ~40 ml sand. Seedlings were watered with 10-15 ml liquid Fahraeus media (Fahraeus, 1957) after planting, and, as necessary, during growth. Seedlings were grown at 28° Celsius with 16 hours light and 8 hours dark. For inoculation, bacterial cultures were grown in liquid TY media to late log phase. Seedlings inoculated with 100 pl of this liquid culture two to three days after transfer to the tubes. 82 RESULTS The objective of this study was to establish a rapid, inexpensive protocol for identifying putative nitrogen-fixing bacteria. The Polymerase Chain Reaction (PCR) was used to rapidly screen phylogenetically diverse bacteria for the presence of the nitrogen fixation nifD gene. Based on the known amino acid sequences of the nitrogenase molybdenum-iron protein alpha chain (Figure 3-1), oligonucleotides corresponding to conserved regions of the nifD gene were designed and synthesized (Figure 3-2). The two opposing oligonucleotides thus generated, dB261 and dB260, correspond to sequences located at the 3’ end of the nifD gene. Amplification of a 390 bp product would be expected when oligonucleotides dB261 and dB260 are used as PCR primers in reactions containing nifD sequences. PCR Amplification of nifD Genes The ability of primers dB261 and dB260 to detect nifD sequences from different template types including whole cells (Versalovic et al., 1994) was tested using purified DNA, single colonies grown on agar plates and resuspended in PCR buffer, and cells grown in liquid 83 culture, washed, and resuspended in SdHQO. Results were similar each type of template tested. All three types of templates prepared from the diazotrophic control bacteria Sinorhizobium meliloti L5-30, Sinorhizobium meliloti 1021, Mesorhizobium Ioti NZP 2037, Mesorhizobium Ioti NZP 2235, and Azorhizobium caulinodans ORS 571 yielded PCR products of the expected size (390 bp; data not shown). All three types of templates prepared from the non- nitrogen-fixing control bacteria Escherichia coli DH5a, Agrobacterium tumefaciens CS8, Xanthomonas campestris pv vesicatoria, and Pseudomonas syringae pv. tomato failed to yield a PCR product of 390 bp. In some cases minor amplification products of other sizes were observed as faint bands in the agarose gel (Data not shown). However, these minor fragments were easily distinguished from the major 390 bp fragment amplified from the nitrogen-fixing strains. The ability of the dB261/dB260 primer set to detect nifD sequences from a wide range of nitrogen-fixing bacteria was further examined using the bacterial strains listed in Table 3-1. Of the 23 nitrogen-fixing bacteria tested, 21 yielded PCR products in the expected size range (390 bp; Figure 3-3A). One of the two 84 nitrogen-fixing bacteria which failed to yield a clear PCR product (Enterobacter cloacae) produced a low abundance PCR product in the expected size range. The other nitrogen-fixing bacterium which failed to yield a PCR product of the expected size was Enterobacter sp Eu F2612 (Figure 3-3A). No PCR product of 390 bp was observed with any of the 9 non-nitrogen-fixing bacteria tested. Pseudomonas fluorescens was originally included as a non-nitrogen- fixing control strain. However, a low abundance fragment of the appropriate size was observed. A strain background check revealed that the lineage of this strain is unclear, and it is not clear if it is diazotrophic. Characterization of PCR Fragments Using DNA Sequencing and Southern Hybridization Analysis The identity of one of the PCR products was confirmed by DNA sequencing. The 390 bp PCR product from Mesorhizobium Ioti NZP2235 DNA was purified and sequenced as described in the Material and Methods. The dB261/dB260 primer set was used to determine the central portion of the DNA sequence. The translation of the derived DNA sequence revealed a polypeptide with high identity to the nitrogenase molybdenum-iron protein a chain (nifD) 85 sequences used to design the primers (Figure 3-1). Blast analysis of both nucleotide- and amino acid sequences confirmed the identity of the nifD fragment (Data not shown). The identity of the 390 bp PCR product from the other strains was confirmed by Southern hybridization. The amplified DNA from the strains listed in Table 3-1 was electrophoresed on an agarose gel and transferred to a membrane, as described in Material and Methods. The purified PCR product from Mezorhizobium Ioti NZP2235 was labeled and used as a probe. The probe hybridized to the 390 bp PCR product amplified from all nitrogen-fixing bacteria tested (Figure 2-3 B) indicating that these fragments contain significant homology to nifD sequences. In addition, the probe hybridized to some PCR products outside the 390 bp range. This was only observed in the case of a few of the nitrogen-fixing bacteria. Very weak hybridization signals were observed in the lanes containing Pseudomonas fluorescens and Enterobacter cloacae DNA products. Longer exposure of the hybridization filters confirmed this result (Data not shown). The probe did not hybridize to nifD-PCR reactions 86 from non-nitrogen-fixing bacteria, nor to PCR products generated using REP-PCR primers (Data not shown). Screening 3 Collection of Rice Associated Bacteria for the Presence of nifD DNA Sequences A collection of 142 unidentified bacteria isolated from surface- sterilized rice tissue was screened for strain harboring nifD gene sequences using primers dB261/dB260. A PCR product of approximately 390 bp was reproducibly generated from twenty of these bacteria (Table 3-2). An example of this analysis is shown in Figure 3-4. The identity of the PCR products was confirmed using Southern blot hybridization, as described in Material and Methods (Data not shown). The nifD probe was found to hybridize to each fragment observed in the 390 bp range, confirming that these PCR products contain nifD sequences. Confirming the Presence of nif Genes by Southern Hybridization of Total Genomic DNA Genomic DNA for Southern analysis was isolated from 35 of the rice isolates. The 35 isolates were selected based on the observation that in initial screening done at IRRI they either produced ethylene in acetylene reduction assays (ARA) or contained nifD gene sequences 87 in PCR experiments (Chapter 2). Southern hybridization experiments were carried out using a probe containing the nifHDK region from Azorhizobium caulinodans ORSS71, as described in Material and Methods. Hybridization signals were detected in 19 of the 35 isolates, as shown in Table 3-2. An example of the hybridization analysis is shown in Figure 3-5. Acetylene Reduction Capacity as an Indicator of Nitrogen Fixing Activity The ability of the above described 35 rice isolates to fix nitrogen was assessed using ARA. The results of this analysis are shown in Table 3-2. Isolates that failed to produce ethylene were scored as negative (-). The 19 isolates that produced, on average, more than 5 nmoles of ethylene and produced ethylene in all replicates tested were scored as positive (+). These 19 isolates were found to correspond to the same 19 shown to harbor nif genes by Southern hybridization. Isolates which produced some ethylene(> 6 nmole) in one or more experiment, but failed to produce ethylene in at least one replicate, were scored as positive/negative (+/-). 88 Direct Detection of Rice-Associated Nitrogen-Fixing Bacteria The utility of primers dB261/dB260 to directly detect nitrogen-fixing bacteria associated with rice plants under laboratory and field conditions was examined. Sterile rice seedlings (lR42) were inoculated with previously identified nitrogen-fixing rice isolates R22(88), R100(64), and R90(8). After four weeks of growth, DNA template was prepared from infected and uninfected rice tissues as described in Material and Methods, and used for PCR with the dB261/dB260 primer set. A PCR product of 390 bp was observed in DNA from inoculated-, but not from uninoculated plants (Figure 3-6), suggesting the presence of bacteria harboring nifD gene sequences. In addition, field grown rice plants (lR72 and IR24) were examined for the presence of bacteria harboring nifD gene sequences. Tissue from roots and aerial portions of field grown plants were harvested and DNA was isolated as described in Material and Methods. A 390 bp PCR product was observed in DNA samples from both roots and aerial portions of both varieties (Figure 3-6). A 390 bp PCR product was also amplified from DNA isolated from field grown plants following vigorous shaking with glass beads, or shaking with glass beads and 89 surface sterilization (Data not shown). This indicates that the rice tissues indeed contain bacteria harboring nifD gene sequences. 90 DISCUSSION nifD-PCR Primers for Rapid and Specific Identification of Diazotrophic Bacteria. We have shown here that the nifD-PCR primer set, developed based on highly conserved regions of NifD, is useful for the rapid identification of diazotrophs in large collections of environmental microbial isolates. Use of the nifD primer set in PCR resulted in the generation of a ~390 bp fragment from the DNA of 21/23 diazotrophic test bacteria examined. DNA sequence and Southern hybridization analysis confirmed that the ~390 bp PCR products observed contain nifD gene sequences. From one of the diazotrophic bacteria tested, E. cloacae, only a minor PCR product of the appropriate size could be generated, showing a weak hybridization signal with the nifD probe. Lowering the annealing temperature of the PCR reaction from 52°C to 40°C increased the intensity of the 390 bp PCR products, including the fragment from E. cloacae, but also increased the number and intensity of background fragments, making it more difficult to distinguish diazotrophic and non- diazotrophic bacteria (Data not shown). The other nitrogen-fixing 91 bacteria, Enterobacter sp Eu F2612, showed no PCR product at either annealing temperature or with Southern hybridization analysis. PCR with the nifD primer set failed to generate distinct ~390 bp fragments from DNA of non-diazotrophic test bacteria, although faint fragments corresponding to minor PCR products of different sizes were observed in some cases. In the case of P. fluorescens, a minor PCR product of the expected size was observed, which shared homology with nifD sequences. This was surprising, since P. fluorescens has not been reported to carry nif genes. It is possible that the isolate used is a contaminate, or that in rare cases our PCR based identification protocol yields a false positive result. When a large collection of bacteria isolated from surface-sterilized rice tissue was screened for diazotrophs using the nifD primer set, ~390 bp DNA fragments were generated from 20 of these isolates. The fragments hybridized to nifD sequences in all cases. Southern hybridization analysis of total DNA confirmed the presence of nif genes in the genome of 19 of these bacterial isolates. The same 19 92 bacterial isolates were able to reduce acetylene, indicating the production of an active nitrogenase enzyme complex. The one isolate, K113, from which a ~390 bp fragment was generated, but in which no nif genes were detected or acetylene reduction observed, is similar to two other isolates, R100(64) and R90(8), both of which are positive in nifD-PCR, Southern hybridization, and ARA experiments. It is possible that K113 is a diazotrophic strain carrying nif genes that are different enough from the nifHDK genes of A. caulinodans used as a probe, that under the conditions employed, no hybridization signals could be observed, and that the culture conditions used for ARA experiments were not appropriate for this particular isolate. Alternatively, it is possible that K113 is not a diazotrophic strain and that the nifD-PCR product is an artifact. However, the fact that 295% of the isolates from which a ~390 bp fragment was generated were shown to be diazotrophs by Southern analysis and ARA clearly shows the utility of the nifD primer based PCR screening protocol. 93 The nifD Primer Set Rapidly Identifies Diazotrophic Bacteria in Large Collections of Environmental Isolates The nifD-PCR primers were able to identify 19 diazotrophic bacteria from a collection of 142 bacteria isolated from surface-sterilized rice plants. Each of the 19 isolates that was identified by the nifDPCR primers showed acetylene reduction activity and hybridization to nif genes in genomic Southern analyses. Ten bacterial isolates which showed acetylene reduction activity in preliminary experiments carried out at IRRI did not reduce acetylene in our tests, and four of the isolates which reduced acetylene during the screens at IRRI showed weak acetylene reduction in some of the replicate experiments at MSU. The characteristic ~ 390 bp fragment was not generated from DNA isolated from these bacteria, nor were nifHDK hybridization signals observed. As pointed out above for the E. cloacae case, it is possible that these bacteria carry nif genes which have diverged to the point where no hybridization with the nifHDK gene from A. caulinodans would be observed under the conditions used and that, therefore, the nifD-PCR primers also do not anneal. It is also possible that these bacteria are able to fix nitrogen using a novel nitrogenase enzyme. Some isolates 94 formed dense pellicles when grown in a nitrogen free semi-solid medium, even in replicate tubes lacking acetylene reduction activity. This suggests that these bacteria are either good scavengers of traces of nitrogen or are nitrogen-fixing. NifD-PCR Detects Novel Diazotrophs The nifD-PCR primers were able to detect diazotrophs from highly diverse genera, include eleven genera of diazotrophic test bacteria and nine genera of bacteria isolated from surface-sterilized rice fissue. nifD-PCR Directly Detects Diazotrophs in Environmental Samples The nifD-PCR primers were also able to detect nitrogen-fixing bacteria directly in environmental samples, such as infected plant tissues. A PCR product of the expected size was detected in the total DNA of plants inoculated with diazotrophic bacteria, but not in DNA from aseptic, uninoculated plants. This suggests that these primers may be useful tools for tracking nitrogen-fixing bacteria during re-inoculation tests under laboratory conditions. 95 CONCLUSIONS On the basis of the results presented above, we conclude that the nifD primers and PCR provide a quick and reliable method for identifying nitrogen-fixing bacteria from a wide range of genera. While other groups have designed PCR primers capable of amplifying nif genes from environmental samples, the utility of these primers for identifying nitrogen-fixing bacteria was not reported (Ueda et al., 1995a; 1995b). The nifD-PCR primers we developed permit the specific amplification of nifD from a variety of template types including purified DNA, colonies from plates, liquid culture and environmental samples, making them suitable for a variety of purposes. Because the nifD-PCR primers are able to amplify a specific fragment from a wide variety of diazotrophic bacteria, they are useful for rapid screening of large collections of isolates cultured from a specific ecological niche. In addition the nifD-PCR primers allow the detection of diazotrophs in environmental samples, without culturing bacteria. An important aspect of increasing the amount biologically fixed nitrogen available to agricultural crops is the isolation and 96 FFA“ {av-s A.“ . ltd"; Mm. abi identification of novel types of nitrogen-fixing bacteria. The ability of the nifD-PCR primers to rapidly and reliably identify nitrogen-fixing bacteria in a collection of uncharacterized isolates has been shown in this study. This will facilitate isolation of diazotrophic bacteria from a variety of environments, including the rhizoplane and internal tissues of crop plants. Isolation of diazotrophs from the internal tissues of crop plants will lead to a better understanding of the ecology of this niche and provide knowledge important for the long term goal of replacing industrial nitrogen fertilizer with biological nitrogen fixation. 97 TABLE 3-1 Strain Nitrogen nifD PCR , Fixing Product Agrobacterium rhizogenes 158341 - - Agrobacterium tumefaciens C581 — - Alca/igenes eutrophus1 - - Clavibacter michiganensis subsp sepedonicus - - SS#432 Cog/nebacterium flaccumfaciens1 - - Escherichia coli DH5a1 — - Pseudomonas fluorescens1 UHKDOWH + / - Pseudomonas syringae pv tomato ATCC - - 108622 Xanthomonas campestris pv campestris2 ' ‘ Xanthomonas campestris pv vesicatoria ATCC - - 359372 Anabaena sp strain PCC 71203 Azoarcus indigens LMG90921 Azorhizobium caulinodans ORSS711 Azospirillum brasilense Sp71 Azotobacter DSM 22854 ++++++ Bradyrhizobium japonicum1 Enterobacter spEu F26.124 + \ I Enterobacter cloacae4 Frankia CESI 55 Herbaspirillum serogedicae 2674 Klebsiella plantico/a KpM5A14 Rhizobium leguminosarum bv trifolii1 Rhizobium leguminosarum bv viciae JIM 14021 Rhizobium leguminosarum bv. viciae AVH11 Mesorhizobium Ioti NZP 20371 Mesorhizobium Ioti NZP 22351 +++++++++++++++++ +++++++++ Sinorhizobium meliloti 10211 98 TABLE 3-1 Continued Sinorhizobium meliloti L5-301 Rhizobium sp USDA311 Rhizobium sp JRG21 Rhizobium sijLasol Rhizobium sp WBM131 ++++++ ++++++ Rhizobium trite/ii1 Table 3-1 Bacterial Strains and PCR Fragments Generated with the nifD Primer Set dB261/d8260. List of bacteria used to test nifD-PCR primers. For Nitrogen Fixing a (+) indicates strain 3 reported to be diazotrophic and a (-) indicates strains not reported to be diazotrophic. For nifD PCR Product a(+) indicates the presence of a ~390 bp PCR product, a (-) indicates no ~390 bp PCR product, and a (+/-) indicates a faint ~390 bp PCR product in reactions containing that bacteria as a template. Source of bacteria: 1Lab Collection, 2 Kind gift from Dr. Frank Louws, 3Kind gift from Dr. Peter Wolk, 4IRRI Collection, 5 Kind gift from Dr. Marcia Murry. 99 Table 3-2 Isolate Southern ARA K46 R29(91) R68(101) R82(128) R94b R96(115) S2 T14 T94 T96(a) R75(50) R92(60) 1295(7) R97(106) -R14(5) K113 -R11c(59) -R16(47) K103 K107 R100(64) R11c(74) R22(88L R48b R62b(117) R720 04) R81 (90) R90(8) R94(17) R99(65) T105 T60 T62 T90 T93 . ++++++++++++++++++++- +++++++++++++++++++. +++++++++++++++++++ 100 Table 3-2 Results of nifD-PCR, Southern Hybridization and ARA experiments for 35 Rice Isolates Table showing results of nifD-PCR, Southern, and ARA analysis for 35 rice isolates. For nifD-PCR (-) indicates no amplification of a 390 bp DNA fragment, (+) indicates the presence of a 390 bp DNA fragment. For Southern analysis (-) indicates lack of hybridization with nifHDK from A. caulinodans, (+) indicates hybridization with nifHDK from A. caulinodans. For ARA (-) indicates no ethylene production, (+/—) indicates low levels of ethylene production in some but not all tubes tested, and (+) indicates higher levels of ethylene production in all tubes tested. 101 KLEBSIELLA PNEUMONIAE ANABAENA SP. (STRAIN PCC 7120) BRADYRHIZOBIUM JAPONICUM RHIZOBIUM SP. COWPEA (STRAIN IRC78) AZOSPIRILLUM BRASILENSE FRANKIA SP. (STRAIN ARL3) FGPIKDMAHI WGPIKDMIHI WGPIKDMVHI WGPIKDMVHI WGPIKDMIHI MEZORHIZOBIUM LOTI NZP2235 SHGPAGCGQY SHGPVGCGYW SHGPVGCGQY SHGPVGCGQY SHGPVGCGYY SHGPVGCGQY SKLIEEMELL TKLIEELDVL DKILDEIQEL VKILDEIQEL HKVIEEINEL EQGLDEIVEL EKIIDEIEDL VRCEGFRGVS LRCEGFRGVS VRCEGFRGVS VRCEGFRGVS VRCEGFRGVS VRCEGFRGVS SRAERRNYYT SWSGRRNYYV SWGSRRNYYV SWGSRRNYYV SWSGRRNYYV SWATRRNYAH NYYV FPLTKGITIQ FPLNRGVSIQ FPLNNGITIQ FPLNNGITIQ FPLVNGISIQ FPLAKGISVQ FPLSGGISVQ QSLGHHIAND QSLGHHIAND QSLGHHIAND QSLGHHIAND QSLGHHIAND QSLGHHIAND GVSGVDSFGT GVTGINSFGT GTTGIDSFVT GTTGIDSFVT GDTGVDKLGT GHLGVDNFTA GTTGIDTFVT SECPVGLIGD SECPIGSIGD SECPVGLIGD SECPIGLIGD SECPIGLNGD SECPIGLIGD S WGPVKDMVTI LNFTSDFQER MHFTSDFQER LQFTSDFQEK LQFTFDFREK MHFTSDFQEK MQITTDFQEK MQFTSDFQEK DISAVANASS DIEAVAKKTS DIEAVSRAKS DIEAVSRAKS DIEGVSKAKS DIEAVARVSS DIVFGGDKKL DIVFGGDKKL DIVFGGDKKL DIVFGGDKKL DIVFGGDKKL DIVFGGDPKL DIVFGGDKKL KAL.DKPVIP KQI.GKPVVP KEYGGKTIVP KEYGGKTIVP EEL.GKPVVP R..LDIPVMR Figure 3-1 Amino Acid Alignment of Mo-Fe Protein Sequences Amino acid alignment of the translation products of nifD genes. The six sequences shown, along with eight others, were obtained from Medline and used to design degenerate oligonucleotide primers for nifD- PCR. The underlined amino acid sequences were used to design the primers. The translation of the DNA sequence of the nifD—PCR product from Mezorhizobium Ioti NZP 2235 is shown in bold. 102 Forward Reverse (dB261) (dBZ60) y regrrgu EEEIEEQ TGG GGT CCT ATT AAA GAT ATG GAA CAT CAT ATT GCT ATT GAT A A G A G C T C C A A C C G G G G C G C C C C C Reverse Compliment S’TGG GGI CCI ATI AAA GAT ATGB’ S'TC ATT IGC IA! ATG ATG TCC3’ G G C G G G A G C Figure 3-2 nifD-PCR Primers Diagram of primer design for the nifD primers. The amino acid sequence used to design the primers is underlined. All possible nucleic acid codons are shown below each amino acid. The final primers are shown in bold with l representing inosine and nucleotides listed below showing degeneracy. 103 M12 3 4 5 6 7 8 9l0ll12l3l4 M1516l718l9202|22232425262728 M29303] 1‘:- “' ‘ ' 3...... .... A...r::';sami .._ .. 1' 7 '; :43 ' I. f .\._._.‘.. .‘. ,. ... '.. .._‘. .5. L __J ‘ a “"4..':'.' '3 - . .. 3‘ ~.‘ '. '4 ‘34-) ‘ ' _,.. ,3 . {E 'g ,., ,.. a a ‘ ,1; ‘fi , _‘ 1‘. ( i , o c‘ c, u' ., '1! g .8. ._ .. ,f x. - . .1 'r . ‘3 . _ . . . .3 .3 . - I. .- . ' - ' . " 2‘ I ' A. .4, 4'. . ‘. . ‘- . If 3 I~ ;. ‘ I_ . ,5 .r ,, - _ ' ‘s . --' _ -. ‘ :.1. t -, -. - ‘ ' “ " - ‘ . » ' ' ' v - ' 1‘ x '4- 3. -.. ~ .. - .. t . .:_ .1. ,7, ~ . . r. “,- 3 , . .2. ~ _‘ _ ’ ‘ .- . .. " ' ' .- . . .‘ ,1, 1 v .2 7 i- ..r ‘ - , . “v . ' ’ d, ' ,"EV a ’ c ._‘ _ t. t: M] 2 3 4 5 6 7 8 91(lll1213l4 M15lbl7l819202122232425262728 M29303] 1,500 600 100 Figure 3-3 nifD-PCR Products Visualized by Agarose Gel Electrophoresis and Identified by Southern Hybridization (A) Polymerase chain reaction (PCR) product generated with oligonucleotide primers corresponding to nifD genes (nifD-Primers) from liquid culture of known bacteria. A photograph of an ethidium bromide stained 2.0% agarose gel is shown. “M” indicates the size marker lanes (100 bp ladder from Boehringer Mannheim) The numbers above the lanes correspond to the following strains: 1, water; 2, Agrobacterium rhizogenes 15834; 3, Agrobacterium tumefaciens CS8; 4, Alcaligenes eutrophus; 5, Corynebacterium flaccumfaciens; 6, C/avibacter michiganensis sub. sp. sepedonicus SS#43; 7, Escherichia coli DH5a; 8, Pseudomonas fluorescens; 9, Pseudomonas syringae pv. tomato ATCC 10862; 10, Xanthomonas campestris pv. vesicatoria ATCC 35937; 11, Azoarcus indigens LMG 9092; 12, Azorhizobium caulinodans ORSS71; '13, Azospirillum brasilense sp7; 14, Azotobacter DSM 2285; 15, Bradyrhizobium japonicum; 16, Enterobacter spEu F2612; 17, Enterobacter cloacae; 18, Frankia CESI 5; 19, Herbaspirillum seropedicae 267; 20, Klebsiella planticola KpM5A1; 21, Rhizobium leguminosarum bv trifolii; 22, Rhizobium leguminosarum bv viciae JIM 1402; 23, Rhizobium leguminosarum bv viciae AVH1; 24, Mesorhizobium Ioti NZP2235; 25, Sinorhizobium meliloti1021; 26, Sinorhizobium meliloti L5-30; 27, Rhizobium sp USDA31; 28, Rhizobium sp JRG2; 29, Rhizobium sp TAL380; 30, Rhizobium sp WBM13; 31, Rhizobium trifolii. (B) Southern blot of gel A with nifD probe verifying the identity of the PCR product. 104 0 13 :D m ' 33 m :0 —& \I —L I) —L I O) :0 I) §8 Samaaafifixqg‘EEE A A -‘ A A A A —‘ A A b § oz 3 3 5' a: 52 o m 4:- \I oo 3 j 8 CD _. ii % pqSQOAégégdégvmmgdgq ‘ I .348Két‘u" m \ “In ”I i a L ‘4 " i 1 ‘4 I v .g . -... .- ' r . ‘ . ‘ f; “a... , It . t . Figure 3-4 Agarose Gel Electrophoresis of nifD-PCR Products from Rice Isolates Products of nifD-PCR performed on rice isolates. A ~390 bp fragment indicates the presence of the nifD gene in the bacterial genome. Products of sizes other than ~390 bp did not hybridize to nifD gene sequences in Southern analysis unless a prominent 390 bp band was also present (see Figure 3-3). 105 .3> m g . >2“) :D 3311 SEE)“ ooIlcoco awe-t xxx—twee kb evtg'saaaaszs’éfi’ia 29mmomowwwo'33933’l 23.1— 9.4— 6.6— 4.4"... «in 2.3— . 2.0— . b“ Figure 3-5 Southern Blot of Bacterial DNA Probed with nifHDK. Low stringency Southern blot of bacterial DNA probed with the nifHDK region from Azorhizobium caulinodans ORS 571. Hybridization indicates the bacteria contain some of the genes necessary for nitrogen fixation. DNA from Sinorhizobium meliloti and Agrobacterium tumefaciens CS8 was included as positive and negative controls. M12345678M 3‘... Figure 3-6 Agarose Gel Electrophoresis of nifD-PCR Products from Rice Tissues Polymerase chain reaction (PCR) products generated with nifD-Primers and DNA isolated from rice tissue. A photograph of an ethidium bromide stained 2.0% agarose gel is shown on which the nifD-PCR products from DNA isolated from rice was separated. “M” indicates the size marker lanes (100 bp ladder from Boehringer Mannheim). The numbers above the lanes correspond to DNA isolated from the following rice tissues: 1, Variety IR42 grown gnotobiotically from sterilized seeds; 2, Variety |R42 grown gnotobiotically from sterilized seeds and inoculated with strain R22(88); 3, Variety |R42 grown gnotobiotically from sterilized seeds and inoculated with strain R00(64); 4, Variety |R42 grown gnotobiotically from sterilized seeds and inoculated with strain R90(8); 5, Roots of field grown variety IR72; 6, Stems and leaves of field grown variety IR72; 7, Roots of field grown variety |R24; 8, Stems of field grown variety IR24. All DNA was prepared as described in Material and Methods. REFERENCES Bally R, Thomas-Bauzon D, Heulin T, Balandreau J (1983) Determination of the most frequent N2-fixing bacteria in the rice rhizosphere. Can J Microbiol 29: 881-887 Beringer JE (1974) R factor transfer in Rhizobium leguminosarum. J Gen Microbiol 84: 188-189 Bockman CC (1997) Fertilizers and biological nitrogen fixation as sources of plant nutrients: Perspectives for future agriculture. Plant and Soil 194:11-14 Boddey R M (1995) Biological nitrogen fixation in sugar cane: A key to energetically viable biofuel production. Critical Reviews in Plant Sciences 1 4: 263-279 Bohlool BB, Ladha JK, Garrity DP, George T (1992) Biological nitrogen fixation for sustainable agriculture: A perspective. Plant and Soil 141: 1-11 Chisnell JR Premakumar R Bishop PE (1988) Purification of a second alternative nitrogenase from a nifHDK deletion strain of Azotobacter vinelandii. J Bacteriol 170: 27-33 Elmerich C, Dreyfus BL, Reysset G, Aubert .P (1982) Genetic analysis of nitrogen fixation in a tropical fast-growing Rhizobium. EMBO 1: 499-503 Fahraeus A (1957) The infection of clover root hairs by nodule bacteria studied by a simple glass slide technique. J Gen Microbiol 16: 374-381 Furrer B, Candrian U, Hoefelein C, Luethy J (1991) Detection and identification of Listeria monocytogenes in cooked sausage products and in milk by in vitro amplification of haemolysin gene fragments. J Appl Bacteriol 70: 372-379 ' 108 Haas JH, Moore LW, Ream W, Manulis S (1995) Universal PCR primers for detection of phytopathogenic Agrobacterium strains. Appl Environ Microbiol 61: 2879-2884 James EK, Olivares FL (1998) Infection and colonization of sugar cane and other graminaceous plants by endophytic diazotrophs. Critical Reviews in Plant Sciences 17: 77-119 Kim J, Rees DC (1994) Nitrogenase and Biological Nitrogen Fixation. Biochemistry 33: 389-397 Ladha JK, Barraquio WL, Watanabe I (1983) Isolation and identification of nitrogen-fixing Enterobacter cloacae and Klebsiella planticola associated with rice plants. Can J Microbiol 29: 1301-1308 Ladha JK, Kundu DK (1997) Legumes for Sustaining Soil Fertility in Lowland Rice. In OP Rupela, C Johansen, DF Herridge, eds, Extending Nitrogen Fixation Research to Farmers’ Fields: Proceedings of an International Workshop on Managing Legume Nitrogen Fixation in the Cropping Systems of Asia. ICRISAT, pp 76-102 Lakshmi Kumari M, Kavimandan SK, Subba Rao NS (1976) Occurrence of nitrogen-fixing Spirillum in roots of rice, sorghum, maize, and other plants. Indian J Exp Biol 1 42638-639 Louws FJ, Schneider M, de Bruijn FJ (1996) Assessing genetic diversity of microbes using repetitive-sequence-based PCR (rep-PCR). InG Toranzos, ed, Nucleic Acid Amplification Methods for the Analysis of Environmental Samples. Technomic Publishing Co, Lancaster, pp 63-93 Oyaizu-Masuchi Y, Komagata K (1988) Isolation of free-living nitrogen-fixing bacteria from the rhizosphere of rice. J Gen Appl Microbiol 34:127-164 Rademaker JLW, de Bruijn FJ (1997) Characterization and classification of microbes by rep-PCR genomic fingerprinting and computer assisted pattern analysis. In G Caetano-Anolles, PM Gresshoft eds, DNA markers: Protocols, Applications, and Overviews. J. Wiley & Sons, Inc. New York, pp 151-171 109 Reinhold-Hurek B, Hurek T (1998a) Interactions of gramineous plants with Azoarcus spp. and other diazotrophs: Identification, localization, and perspectives to study their function. Critical Reviews in Plant Sciences 17:29-54 Reinhold-Hurek B, Hurek T (1998b) Life in Grasses: Diazotmphic Endophytes. Trends in Microbiol. 6: 139-202 Saiki RK, Gelfand DH, Stoffel S, Scharf SJ, Higuchi R, Horn GT, Mullis KB, Ehrlich HA (1988) Primer-directed enzymatic amplification of DNA with a thermostable DNA polymerase. Science 239: 487-491 Sambrook J, Fritsch EF, Maniatis T, eds, (1989) Molecular Cloning: A Laboratory Manual 2nd ed. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory Press Sawada H, Ieki H, Matsuda I (1995) PCR detection of Ti and Ri plasmids from phytopathogenic Agrobacterium stains. Appl Environ Microbiol 61: 828-831 Schuddekopf K Hennecke S Liese U Kutsche M Klipp W (1993) Characterization of anf genes specific for the alternative nitrogenase and identification of nif genes required for both nitrogenases in Rhodobacter capsulatus. Mol Microbiol 8: 673-84 Steffan RJ, Atlas FM (1991) Polymerase chain reaction: applications in environmental microbiology. Annu Rev Microbiol 45: 137-161 Thiel T (1993) Characterization of genes for an alternative nitrogenase in the cyanobacterium Anabaena variabilis. J Bacteriol 175: 6276-86 Thomas-Bauzon D, Weinhard P, Villecourt P, Balandreau J (1982) The Spermosphere Model. I. Its use in growing, counting, and isolating N2-fixing Bacteria from the Rhizosphere of Rice. Can J Microbiol. 28: 922-928 Ueda T, Suga Y, Yahiro N, Matsuguchi T (1995a) Remarkable N2- fixing bacterial diversity detected in rice roots by molecular evolutionary analysis of nifH gene sequences. J Bacteriol 177: 1414-1417 110 Ueda T, Suga Y, Yahiro N, Matsuguchi T (1995b) Genetic diversity of N2-fixing bacteria associated with rice roots by molecular evolutionary analysis of nifD Library. Can J Microbiol 41: 235-240 Versalovic J, Schnieder M, de Bruijn FJ, Lupski .R (1994) Genomic fingerprinting of bacteria using repetitive sequence-based Polymerase Chain Reaction. Meth Mol Cell Biology 5: 25-40 Watanabe I, Barraquio W (1979) Low levels of fixed nitrogen required for isolation of free-living N2-fixing organisms from rice roots. Nature 277: 565-566 Watanabe I, So R, Ladha JK, Katayama-Fujimura Y, Kuraishi H (1987) A new nitrogen-fixing species of pseudomonad: Pseudomonas diazotrophicus sp. nov. isolated from the root of wetland rice. Can J Microbiol 331670-678 Zinoni F Robson RM Robson RL (1993) Organization of potential alternative nitrogenase genes from Clostridium pasteurianum. Biochim Biophys Acta 1174: 83-6 111 CHAPTER 4 DIVERSITY OF A COLLECTION OF BACTERIA ISOLATED FROM SURFACE-STERILIZED RICE TISSUE ABSTRACT A collection of 142 bacteria isolated from surface-sterilized rice tissue was analyzed for its genetic diversity using amplified Ribosomal DNA restriction analysis (ARDRA) fingerprinting, rep-PCR genomic fingerprinting and small subunit (SSU) ribosomal RNA (rRNA) gene sequencing. Analysis of fingerprints using Gelcompar software revealed 71 unique rep-PCR genomic fingerprints and 56 unique ARDRA fingerprints. Clusters of similar combined fingerprints, consisting of 37, 15, 12, and 9 non-diazotrophic bacteria, as well as two clusters each containing 4 diazotrophic bacteria, were found. Analysis of SSU rRNA gene sequences from selected isolates in the collection revealed similarities to bacteria from the alpha-, beta-, and gamma subdivision of the Proteobacteria, and members of the Bacillaceae and Microbacteriaceae. Several of the bacteria analyzed displayed ARDRA fingerprints and/or contained SSU rRNA gene sequences similar to bacteria previously isolated from the rhizosphere of rice. 112 INTRODUCTION The presence and role of endophytic diazotrophs in graminaceous plants has been the subject of several recent studies (James and Olivares, 1998; Reinhold-Hurek and Hurek, 1998a; 1998b). The presence and genetic diversity of diazotrophs in the rhizosphere and rhizoplane of rice has also been investigated in some detail (Lakshmi Kumari et al., 1976; Watanabe and Barraquio, 1979; Thomas-Bauzon et al., 1982; Bally et al., 1983; Ladha et al., 1983; Watanabe et al., 1987; Oyaizu-Masuchi and Komagata et al., 1988; Ueda et al., 1995a; 1995b). However, the presence and genetic diversity of diazotrophic endophytes in field grown rice has remained relatively unevaluated (For a definition of “endophytes” see Chapter 1). In this chapter we extend our previous preliminary analysis of diazotrophic bacteria associated with rice by presenting a detailed molecular phylogenetic characterization of a collection of bacteria we isolated from surface-sterilized rice tissue (Chapter 2). rep-PCR genomic fingerprinting was initially used to assess the genetic diversity of bacteria isolated from mechanically-abraded, surface- sterilized rice plants (Chapter 2). rep-PCR genomic fingerprints have been found to be extremely useful for distinguishing bacteria at the 113 subspecies- and strain- specific level, and phylogenetic relationships derived from such fingerprints are highly correlated to DNA-DNA homology levels (Vinuesa et al, 1998; Rademaker and de Bruijn, 1997; Rademaker et al., 1999). However, ARDRA fingerprints are clearly more suitable for determining the relationships between bacteria at higher taxonomic levels (Moyer et al., 1996; Rademaker and de Bruijn, 1997). Therefore, the genetic diversity of isolates in a previously described, large collection of putative rice endophytes (Stoltzfus et al., 1997; Chapter 2) was examined over a wider phylogenetic range by linearly combining ARDRA and rep-PCR fingerprints as described by Vinuesa et al. (1998). Similarities between selected isolates in this collection and previously characterized bacteria were examined by analyzing partial DNA sequences of the SSU rRNA genes. The combined analysis permitted a direct comparison with bacteria previously isolated from the rhizosphere of rice and endophytic bacteria isolated from other graminaceous plants. 114 MATERIAL AND METHODS Bacterial Strains and Growth Conditions E. coli HB101 cells harboring pUCNC were cultured at 37°C on LB media containing 100pg/l ampicillin. The other bacterial strains used in this study and their mode of isolation are described in Chapter 2. A complete list of the characteristics of the bacteria studied, including soil type, rice variety and rice tissue from which they were isolated, and the type of medium used for their isolation, is presented in Appendix I. These bacteria were cultured for DNA template preparation on TY media at 28°C (Benngen 1974) PCR Reactions Generation of rep-PCR genomic fingerprints and computer analysis of the resulting fragment patterns were carried out as described by Louws et al. (1996), Rademaker and de Bruijn (1997) and Rademaker et al. (1998) with the following modifications: Template for PCR was prepared by filling a 1 pl disposable loop with cells from a bacterial colony grown on a plate. The contents of the loop were resuspended in 100 pl sterile distilled water in a 1.5 ml microfuge tube, boiled for 10 min, incubated at 37°C for 1 hour with 5 pg of RNAase A and stored at —20°C. One microliter of the resulting extract was used as template for PCR reactions. 115 ARDRA and SSU Sequencing Two highly conserved eubacterial SSU rRNA gene primers (8F and 1492R; Weisburg et al., 1991) were used to amplify the SSU rRNA genes from the bacteria tested using ERIC-PCR amplification conditions (Rademaker and de Bruijn, 1997). For ARDRA, 7.5 pl sterile distilled water, 1.5 pl digestion buffer (supplied by Boehringer Mannheim, Indianapolis, IN, USA), and 1 pl of enzyme were added to 5 pl of the PCR reaction and the mixture was incubated for 1.5 hours at 37°C for Real and Mspl and 65°C for Taql. Following digestion 2 pl of loading buffer was added and the digestion products were separated by overnight gel electrophoresis on a 2.0% agarose gel in 0.5x TAE Buffer and visualized on a UV transilluminator. Documentation and fragment pattern analysis were carried out as described by Louws et al. (1996), Rademaker and de Bruijn (1997), and Rademaker et al. (1998). For DNA sequence analysis, the rDNA PCR amplification products were purified using ultra-Free columns (Millipore, Bedford, MA, USA) or Wizard PCR preps DNA purification systems (Promega, Madison, WI, USA). Partial SSU rRNA gene sequences were obtained using primers 8F and 519RB 116 (Weisburg et al., 1991). The DNA sequence was determined by automated fluorescent sequencing at the MSU DNA Sequencing Facility using the ABI Catalyst 800 for Taq cycle sequencing and the ABI 373A or ABI 377 Sequencer. The DNA sequences obtained were subjected to NCBI Blast Analysis (Altschul et al., 1997). Sesbania rostrata Plant Growth and Inoculation Growth and infection of Sesbania rostrata plants were carried out in sterile test tubes, as described by Pawlowski et al. (1987). Southern Hybridization Analysis of Total Genomic DNA Bacterial genomic DNA was isolated using the MasterPure Genomic DNA Purification Kit from Epicenter Technologies (Madison, WI, USA) according to the protocol provided by the manufacturer. DNA digestion, separation, and transfer to Hybond N nylon membranes (Amersham International, Little Chalfont, Buckinghamshire, England) were carried out according to standard protocols (Sambrook et al., 1989). A 390 bp DNA fragment amplified from the A. caulinodans nifD gene was purified using a Qiaex II Gel Extraction Kit from Qiagen (Madison, WI, USA) and used as a probe for nifD. A 395 bp DNA fragment containing A. caulinodans nodC gene sequence was excised from plasmid pUCNC (K. Goethals Ph.D. thesis, 117 Universiteit Gent) by digestion with Hinclll, run on an agarose gel and purified using a Qiaex II Gel Extraction Kit from Qiagen (Madison, WI, USA). Probe preparation was carried out according to standard protocols using 32P labeled dATP. Prehybridization and hybridization reactions were carried out at 65°C using buffer containing 5x Denhardts solution, 4x SSC, 1% SDS, and 100 pg/ml salmon sperm DNA (Sambrook et al., 1989). The membranes were washed three times at low stringency (65°C in buffer containing 4x SSC and 0.1% SDS). Computer Assisted Pattern Analyses Gelcompar 4.1 software (Applied Maths, Kortrijk, Belgium) was used to analyze linear combined data from Fisal, Mspl, and Taql ARDRA digests; ERIC-, REP-, and BOX- PCR reactions; or all six fingerprints, as described by Rademaker et al. (1998). Cluster analysis was performed using a densitometric curve based UPGMA algorithm (Sneath and Sokal, 1973). 118 RESULTS Phylogenetic Diversity of Bacteria Isolated from Surface- sterilized Rice Plants Two collections of bacteria isolated from surface-sterilized rice tissues, as described in Chapter 2 (Figure 2-1 and 2-2), were initially included in this analysis. However, when tested using Gram staining, the bacteria from the second collection (Figure 2-2) were found to include a high percentage of Gram-positive bacteria. The SSU rRNA gene sequence from two of these bacteria was determined and found to be similar to sequences from Bacillus. Because many Gram-positive bacteria form spores which could be resistant to our surface-sterilization protocol, the decision was made not to study these isolates in detail. Rather, we focused on studying the genetic diversity of a collection of 142 bacteria from the first collection (Figure 2-1) using combined ARDRA and rep-PCR fingerprints in conjunction with Gelcompar computer assisted pattern analysis. The results from the analysis of the linearly combined Fisal, Mspl and Taql ARDRA, and ERIC-, REP- and BOX-PCR genomic fingerprints are shown in Figure 4-1. This analysis revealed both the presence of clusters of similar bacteria, as well as a variety of bacteria with unique fingerprints, as discussed below. 119 Since the triple ARDRA and rep-PCR fingerprint samples were generated on multiple independent gels, two internal controls were included in the combined pattern analysis. Azorhizobium caulinodans was included in each ARDRA and rep-PCR experiment and used as a control for variations in conditions between different PCR reactions, digestions, and gel electrophoretic runs. In addition, the DNA size marker lanes from each gel were included as controls to observe the variation between gels. The red line in Figure 4-1 represents the level of variation between A. caulinodans fingerprints derived from different reactions and run on different gels. The variations in a cluster of 23 DNA size marker lanes is indicated by the letter M in Figure 4-1. Interestingly, we found that the combined fingerprint patterns of four of the rice isolates closely matched the A. caulinodans “internal control” strains (“F”; Figure 4-1). Gelcompar analysis, in conjunction with visual inspection of original fingerprints of the 142 rice isolates in this collection, revealed 56 unique combined Rsal, Mspl and Taql ARDRA fingerprints and 71 unique combined ERIC-, REP- and BOX-PCR genomic fingerprints. Unique fingerprints were defined as those exhibiting less similarity than the seven replicate control fingerprints of A. caulinodans. Fingerprint patterns exhibiting more 120 similarity than the replicate A. caulinodans fingerprints in the Gelcompar analysis but which showed clear differences in fragment patterns by visual inspection were also counted as unique. Eight clusters were identified that contained four or more isolates with combined ARDRA and rep-PCR fingerprints more similar to each other than the combined fingerprints of replicates of A. caulinodans (“A-H”; Figure 4— 1 and Table 4-1). These clusters are described in more detail below. SSU rRNA Gene Sequence Similarities A partial SSU rRNA gene DNA sequence was determined for those bacteria in the collection of rice isolates that had previously been shown to be able to reduce acetylene, or to contain nifD DNA sequences by PCR analysis (Chapters 2 and 3) as well as for selected bacteria from the largest fingerprint clusters (“A-H”; Figure 4-1). These DNA sequences were compared to sequences in the GenBank, EMBL, DDBJ, and PDB databases using NCBI Blast analysis (Altschul et al., 1997). The bacteria from this collection were found to have SSU rRNA genes sharing similarities with 15 different genera of bacteria (Table 4-2). 121 Characterization of Clusters of Bacteria with Similar Fingerprints Cluster analysis of combined ARDRA and rep-PCR fingerprints using Gelcompar software revealed the presence of groups of similar bacteria which were isolated from different rice varieties grown in different soil types. The clusters of bacteria with fingerprints with a higher similarity level than the replicate fingerprints of A. caulinodans are shown in Table 4-1, which also summarizes the number of soil types, rice varieties, and soil type/rice variety combinations from which the strains were isolated. Groups A, B, E and G exemplify isolates with similar ARDRA fingerprints that have a single dominant rep-PCR genomic fingerprint pattern (Table 4- 1; Figures 4-1, 4-2, 4-3, 4-6). For bacteria in these groups the genotype of the strain may be important for interactions with rice. Groups C, D, F and H exemplify isolates with similar ARDRA fingerprints but distinct rep- PCR genomic fingerprints (Table 4-1; Figure 4-1, 4-4 , 4-5, 4-7). For these strains the genotype of the strain may not be as important in interactions with rice. Characterization of Clusters Containing Non-Diazotrophic Bacteria The largest group (“A”; Table 4-1; Figures 4-1, 4-2) was found to contain 37 bacteria isolated from five different soil types and six different rice 122 varieties, representing 11 soil type/rice variety combinations. None of the bacteria in this cluster were found to be diazotrophic by either ARA or nifD-PCR analysis (Chapter 3). These bacteria display combined ARDRA fingerprints highly similar to the combined ARDRA fingerprint for Pseudomonas fluorescence. The only difference between the fingerprints is a 100 bp shift in one fragment from the Rsal digest. The DNA sequence for the SSU rRNA genes from bacteria in this cluster are most similar to those from Pseudomonas rhodesiae and P. marginalis (Table 4- 2). Within this group, three distinct combined rep-PCR genomic fingerprints were observed. Thirty two isolates displayed the same rep- PCR genomic fingerprint, four isolates a second rep-PCR genomic fingerprint, and one isolate a unique rep-PCR genomic fingerprint. However, the rep-PCR genomic fingerprint of P. fluorescence is clearly distinct from the fingerprints of all 37 rice isolates. This data indicates that the bacteria in this group are related to the fluorescent pseudomonads. However, more taxonomic data will be required to determine a more precise phylogenetic position of these isolates. Group B (Table 4-1; Figure 4-1, 4-3) was found to contain 15 bacteria isolated from three different soil types and three different rice varieties, 123 representing four soil type/rice variety combinations. None of the bacteria in this cluster were found to be diazotrophic by either ARA or nifD-PCR analysis (Chapter 3). The ARDRA fingerprints for the 15 bacteria in this cluster are highly similar. However, the fingerprints of two of the bacteria display a minor difference from the fingerprints of the other 13. In these two fingerprints a 900 bp fragment is present in the Fisal fingerprints and a of a 400 bp fragment is absent in the Mspl fingerprints. The ARDRA fingerprints of the bacteria in this cluster are very similar to the ARDRA fingerprints of Klebsiella planticola and Enterobacter cloacae. The ARDRA fingerprint of K. planticola differs from the ARDRA fingerprints of the 13 rice isolates only by the absence of a 400 bp fragment in the Mspl fingerprint and differs from the ARDRA fingerprints of the two rice isolates only by the absence of a 900 bp Fisal fragment. The ARDRA fingerprint of E. cloacae differs from the ARDRA fingerprints of the 13 rice isolates only by the absence of a 400 bp Mspl fragment and the presence of a 600 bp fragment in the Taql digest. It differs from the ARDRA fingerprints of the two rice isolates only by the absence of a 900 bp Rsal fragment and the presence of a 600 bp Taql fragment. The partial SSU rRNA gene DNA sequence from R95(7), a bacteria in the group of 13 rice isolates, is most similar to Pantoea spp. 124 and Enterobacter dissolvens, both belonging to the Enterobacteriaceae. Of the isolates in group B that display identical ARDRA fingerprints, ten display one rep-PCR genomic fingerprint pattern and three a second rep- PCR genomic fingerprint pattern. The two isolates with a distinct ARDRA fingerprint also have a distinct rep-PCR genomic fingerprint. The rep-PCR genomic fingerprints for K. planticola and E. cloacae are distinct from each other and all of the isolates in this group. This suggests that the bacteria in this group are related to Enterobacteriaceae. However, more multiphasic taxonomic data will be required to determine a more precise phylogenetic position in this family. Group C (Table 4-1; Figures 4-1, 4-4) was found to contain 12 isolates from four soil types and five rice varieties, representing nine soil type/rice variety combinations (Figure 4-4). Ten of the isolates in this group have identical ARDRA fingerprints, as explained below. None of the bacteria in this cluster were found to be diazotrophic by ARA or nifD-PCR analysis (Chapter 3). When analyzed using Gelcompar, only eight of the isolates in this group displayed combined fingerprints with higher similarity than the replicate fingerprints of A. caulinodans. However, visual inspection of the original gels revealed that the ARDRA fingerprints of strains K9 and 125 K46 were identical to the ARDRA fingerprints of the eight isolates in this cluster. Apparently, uneven staining of the gels caused the loss of a fragment during computer analysis which resulted in less similar fingerprints. Thus strains K9 and K46 are considered as part of the cluster with the eight other isolates. The ARDRA fingerprint of K96 differs from that of these 10 bacteria only by the absence of a 700 bp fragment in the Taql digest. The ARDRA fingerprint for K6 differs from the ARDRA fingerprints of these 10 isolates by the absence of a 300 bp fragment and the presence of a 100 bp fragment in the Mspl fingerprint. Generally, the isolates in this cluster have very diverse rep-PCR genomic fingerprint patterns. Only two of the ten isolates with identical ARDRA fIngerprints were found to have identical rep-PCR genomic fingerprint Patterns (Figure 4-4). The partial sequence of the SSU ribosomal RNA gene from two isolates in this group, T14 and K46, are most similar to the SSU rRNA gene DNA sequence of Aureobacterium kitamiense, a p0'Ysaccharide producing Microbacterium isolated from the wastewater of a beet sugar factory (Table 4-2). Gr<>l.tp D (Table 4-1; Figures 4-1, 3-5) was found to contain nine bacteria IS(,Nated from five soil types and three rice varieties, representing seven 126 soil type/rice variety combinations. None of the bacteria in this cluster were found to be diazotrophic by ARA or nifD-PCR analysis (Chapter 3). Minor variations in the ARDRA fingerprints were observed in this cluster. Five isolates displayed a major 900 bp fragment, while the other four isolates contained minor fragments of both 900 bp and 700 bp in their Taql fingerprints. The isolates in this group display three distinct rep-PCR genomic fingerprints. Visual inspection of the fingerprints revealed two strains, R97(106) and R97b(34), with ARDRA fingerprints highly similar to ARDRA fingerprints of the other isolates in this cluster. However, these strains were not included in this cluster by Gelcompar analysis. The partial SSU rRNA gene DNA sequence for two isolates in this cluster, strains R75(50) and R97(106), were most similar to SSU rRNA gene DNA Secluence of Alcaligenes and Bordete/la, both members of the family AIczslligenaceae. characterization of Clusters Containing Diazotrophic Bacteria Ge Icompar analysis of the combined ARDRA and rep-PCR fingerprints revealed only two groups (containing more than two isolates) of diazotrophic bacteria with identical ARDRA fingerprints. These two grOups are described below. 127 The first group of diazotrophic isolates (Group E) was found to contain four strains isolated from two soil types and two rice varieties, representing two soil type/rice variety combinations (Table 4-1; Figures 4-1 , 4-6). All bacteria in this group were found to reduce acetylene and harbor nif genes using PCR and Southern blot analysis (Chapter 3). The combined ARDRA and rep-PCR fingerprints of these bacteria were identical. The partial DNA sequences of the SSU rRNA gene of isolates in this group were most similar to the SSU rRNA gene DNA sequence of Burkholderia (Table 4-2). The ARDRA fingerprints for these isolates share many fragments with the ARDRA fingerprint for M130, a putative endophytic Burkholderia sp. isolated from rice in Brazil (Boddey et al., 1995; Hartman et al., 1995). This suggests that the bacteria in this cluster belong to the genus Burkholderia. The Second group of diazotrophic bacteria (Group F) was found to Contain four bacteria isolated from three soil types and three rice Varieties, representing four soil type/rice variety combinations (Table 4-1; Figure 4-1, 4-7). All bacteria in this group were found to reduce acet)Ilene and harbor nif genes using PCR and Southern blot analysis (Q haD'ter 3). These bacteria were shown to have ARDRA fingerprints 128 identical to those of A. caulinodans (Figure 4-7). The complete DNA sequence of the R94(17) SSU rRNA gene was determined and found to be more than 99% identical to the DNA sequence of the A. caulinodans SSU rRNA gene (Table 4-2). A. caulinodans forms root and stem nodules on the tropical legume Sesbania rostrata. However, isolates R81(90), R94(1 7) and R99(65) were unable to nodulate S. rostrata when inoculated to plants (Data not shown). The relationship between isolates in Group F and A. caulinodans was fUl'ther investigated using restriction fragment length polymorphism I RF LP) analysis. No polymorphisms were found between R94(17) and A. caulinodans genomic DNA digested with BamHI, Hindlll, or [5le and hYbr'iclized with the A. caulinodans nifD gene. Likewise, genomic DNA from strains R90(8) and R99(65) digested with BamHI and Hindlll and hybr‘iclized with the A. caulinodans nifD gene showed RFLP patterns similar to those of A. caulinodans. However, polymorphisms between these two Stai hs and A. caulinodans were detected when genomic DNA was digested With EcoRI and hybridized with the A. caulinodans nifD gene (Figure 48). NO hybridization signal was detected with the A. caulinodans nodC gene in s Q-'~‘thern hybridization analysis of genomic DNA from R81(90), R94(17) 129 and R99(65) (Figure 4-8). The fourth bacterium in the cluster, R62b(117) was not examined for nodulation properties or for hybridization with nodC. However, hybridization to genomic DNA from R62b(117) was detected using nifHDK from A. caulinodans (Chapter 3). This suggests that the bacterial isolates in this group are very closely related to, but distinct from, A. caulinodans. 130 DISCUSSION Genetic Diversity of Putative Endophytic Bacteria Bacteria isolated from mechanically-abraded, surface-sterilized rice roots displayed diverse ARDRA and rep-PCR fingerprints. 71 unique rep-PCR genomic- and 56 unique ARDRA fingerprints were observed among the 142 isolates studied (Figure 4-1). Both gram-positive and gram-negative bacteria were represented in the collection of rice isolates analyzed (Table 4-2). While it has not been rigorously shown that these bacteria are true endophytes (See definition in Chapter 1; Chapter 6), these bacteria clearly appear to have an intimate interaction with rice plants. The genetic diversity of the isolates analyzed can be explained by several ecI'J‘euly plausible explanations. It is possible that these bacteria are p r eSent in numerous unique ecological or physiological ‘niches’ associated with rice plants and the genetic diversity of bacteria simply reflects niche dive rsity. It is also possible that these bacteria are present in a few uh iClue niches, and the genetic diversity of the bacteria indicates that a V - a" I ety of bacteria can colonize those few niches. S Decificity of Rice Plant/Bacterial Interactions T he importance of plant and bacterial genotypes in plant colonization s i3 . . . . Ves us lnSlght into the specificity of these interactions. While the data 131 from our study does not contain the numbers needed for proof, it does indicates that in many cases the interactions we observed between rice plants and putative endophytes are non-specific. The rice plant genotype does not appear essential for colonization by specific bacteria in our collection. Every cluster containing four or more bacteria harbored isolates from several rice varieties and soil types (Table 4-1), indicating that each type of bacteria isolated is capable of interacting with a broad variety of rice genotypes. Likewise, in some cases the bacterial genotype does not appear essential for colonization of a specific rice genotype, Since bacterial isolates with a similar SSU rDNA gene structure, but diVe rse genomic fingerprints, were found to colonize the same rice Variety. Groups C and D both contain bacteria which were isolated from the Same rice variety grown in the same soil type with identical ARDRA-, but diverse rep-PCR fingerprint patterns. However, in selected cases, the bacterial genotype may be important for i nte factions with rice plants. Groups A, B, E and G all contain bacteria 'th a Single prominent rep-PCR genomic fingerprint. This could indicate t “at for some types of bacteria a specific genotype confers an advantage Vv -hert interacting with the rice plants. 132 SSU rRNA Gene DNA Sequence Analysis The robust characterization of novel bacterial isolates requires combined data from multiple taxonomic approaches (Polyphasic taxonomy; Vandamme et al., 1996). However, the SSU rRNA gene DNA sequence provides a significant amount of phylogenetic information (Woese, 1987) and permits the preliminary identification of bacteria isolated from the environment (Amann et al., 1995). Partial SSU rRNA gene DNA sequences indicate that selected isolates from this collection are related to endophytes previously isolated from rice, bacteria previously isolated from the rhizosphere of rice and legume symbionts. SiFnilarities to Endophytic Bacteria The SSU rRNA genes from several diazotrophic isolates in the collection are very similar to those of putative endophytic bacteria previously i.S‘Olated from rice. Isolates from group E are diazotrophic and have SSU rR NA genes similar to those of Burkholderia vietnamiensis and M130. 8. Vie tnamiensis was isolated from rice tissue and is the only nitrogen-fixing SF>ecies of Burkholderia (Gillis et al., 1995). M130 is a putative e r“(icaphytic Burkholderia isolated from rice in Brazil (Boddey et al., 1995; H artrnan et al., 1995). The diazotrophic strain R48b has SSU rRNA genes 8. i m - . . Ilar to those of Azoarcus sp6a3, a diazotrophic endophyte of Kallar 133 grass (Hurek et al., 1997b). Some strains of Azoarcus are able to colonize rice endophytically (Hurek et al., 1994). While the isolation of Azoarcus bacteria from rice has not been previously reported in the literature, Azoarcus-like strains have been recently isolated from rice plants grown in Nepal (Barbara Reinhold-Hurek, personal communication). In addition a nifH sequence derived from DNA isolated from rice roots in Japan (Ueda et al., 1995b) has been found to be similar to Azoarcus nifH sequences (Hurek et al., 1997a). Our isolation of Burkholderia- and AZoarcus like bacteria in conjunction with information from previous studies clearly suggests that these bacteria colonize rice endophytically and should be targets for further studies in this area of research. Interestingly, no bacteria with SSU rRNA genes similar to Herbaspirillum SDD - were found in this collection. Herbaspirillum has been reported as a Common diazotrophic endophyte of graminaceous plants, including rice (Baldani et al., 1986; James and Olivares, 1998). The absence of Herbaspirillum-like isolates from this collection does not prove that this t ype of bacteria are not present in Philippino rice varieties, but it does s L"QQest that they are not present in high numbers. 134 Similarities to Rice Rhizosphere Bacteria Some diazotrophic bacteria in this collection have SSU rRNA genes similar to those of Xanthomonas f/avus, Pseudomonas and Azospirillum, diazotrophic bacteria commonly isolated from the rhizosphere or rice (Ladha et al., 1982; Bally et al., 1983; Watanabe et al., 1987; Oyaizu- Masuchi and Komagata, 1988; Reding et al., 1991). It is unclear if these bacteria are novel endophytic strains or simply rhizosphere bacteria that “escaped” surface sterilization. Non-diazotrophic isolates in this collection also have SSU rRNA genes similar to those of diazotrophic bacteria isolated from the rhizosphere of rice. Isolates from the largest cluster, Group A, are non-diazotrophic and have SSU rRNA genes similar to those of the fluorescent pseudomonads. Nitrogen-fixing pseudomonads have been reported as common rice rhiZosphere bacteria (Watanabe et al., 1987). Isolates in Group B, have S3U rRNA genes similar to those of the family Enterobacteriaceae. Ho\Ivever, unlike the members of Enterobacteriaceae previously isolated from rice (Ladha et al., 1983), none of these isolates were diazotrophic. ISOIates from group D have SSU rRNA genes similar to those of Alcaligenaceae. On member of Alcaligenaceae, Alcaligenes faecalis A15, 135 nfll nun t" is a nitrogen-fixing strain isolated from surface-sterilized rice capable of colonizing rice plant epidermal cells (You and Zhou, 1988; Vermeiren et al., 1998). The role of these non-diazotrophic bacteria as endophytes and their importance for future studies in this area of research is unclear. Similarities to Legume Symbionts The isolates in group F have SSU rRNA gene sequences very similar to those of A. caulinodans, the unusual stem- and root nodulating symbiont of the tropical legume Sesbania rostrata (Dreyfus and Domergues, 1981). However, they clearly differ from A. caulinodans since they do not harbor the nodC gene and are unable to form nitrogen-fixing nodules on S. rostrata. To our knowledge, this is the first report of the isolation of Azorhizobium-like bacteria from field grown rice. However, previous Studies have shown that A. caulinodans is capable of colonizing rice Under laboratory conditions (Christiansen-Weniger, 1996; Webster et al., 1 997; Khokhar and Qureshi, 1998). Moreover, A. caulinodans is thought to be a promising candidate for supplying rice with biologically fixed hitrogen since, unlike other rhizobia, it is able to fix nitrogen independent of its legume host (Gebhard et al., 1984) and can tolerate relatively high I . . eveIs of 02 during nitrogen fixation (Ratet et al., 1989). If the isolates in group F share such characteristics with A. caulinodans they are excellent 136 candidates for further studies on the use of bacteria to supply biologically fixed nitrogen to rice. In addition to the isolates sharing similarities with Azorhizobium, ten isolates in this collection have SSU rRNA genes similar to those of Rhizobium, Bradyrhizobium , and Agrobacterium species. Five of these isolates are diazotrophic, including two with SSU rRNA genes similar to those from Agrobacterium. Agrobacterium is closely related to Rhizobium, and when nif and nod genes from Rhizobium are introduced into Agrobacterium it can interact with legumes and fix nitrogen (Hirsch et al-, 1985; Martinez et al., 1987; Novikova and Safronova, 1992). As discussed below, the interactions between these bacteria and rice plants deserve further investigation. Several authors have suggested extending nodulation to rice as a method 0f Supplying rice with biologically fixed nitrogen (Ladha et al., 1997; Reddy et al., 1997; Kennedy et al., 1997). When examining this possibility, the ability of Rhizobium and related species to interact with rice is of keen interest. For example, Rhizobium leguminosarum has been i . S‘9'ated from surface-sterilized rice grown in Egypt and shown to colonize 137 I . ’. 71m :l.. .: twin. 9....) lithe rice grown under gnotobiotic conditions (Yanni et al., 1997). However, the production of nitrogen-fixing nodules on legumes is a complex process requiring the development of specialized structures by the plant host (van Rhijn and Vanderleden, 1995; Long, 1989), a process that clearly does not take place in the interaction between rhizobia and rice (de Bruijn et al., 1995; Reddy et al., 1997 Webster et al., 1997;). Therefore, further investigations of the interactions between putative rhizobial endophytes from rice, which may be well adapted for rice colonization, are clearly needed and will provide information useful for future attempts to extend nodulation to rice and other cereal crops. Other SSU rRNA Gene Similarities Two Isolates from Group C have SSU ribosomal DNA gene sequences similar to those of Aureobacterium kitamiense. Other members of the genus Aureobacterium have been redefined as the genus Microbacterium (Takeuchi and Hatano, 1998), However, A. kitamiense has never been formadly renamed. A. kitamiense is a polysaccharide producing bacterium is*OIated from the wastewater of a beet sugar factory (Table 4-2). To our knOvvledge, there are no other reports of Microbacterium isolated from the rhizosphere or tissues of rice. 138 The partial SSU rRNA gene DNA sequences from the other isolates which were analyzed show similarity to Bacillus, Comamonas, Brenvundimonas and Cau/obacter. Since Bacillus spp. form spores, it is quite likely that this isolate was protected from surface sterilization due to spore formation rather than endophytic colonization. These strains are not of lesser interest as candidates for nitrogen-fixing endophytes because they do not fix nitrogen nor were they isolated from different rice varieties and soil types. Obligate and Facultative Endophytes It has been proposed that endophytic bacteria should be divided into obligate and facultative endophytes (Baldani et al., 1997). Facultative endOphytes can survive in the soil and on the plant surface as well as in the interior of the plant. Azospirillum strains capable of colonizing the interior of the plant could be considered facultative endophytes because they also are commonly found in the rhizosphere (Baldani et al., 1997). However bacteria such as Herbaspirillum spp. and Burkholderia spp. seem to be found only inside plant tissues and could be considered to be Obligate endophytes. Assuming that bacteria isolated from mechanically- at“faded, surface-sterilized rice plants are capable of colonizing the 139 interior of the plant to some extent, the bacteria in this collection can be divided into obligate and facultative endophytes. Bacteria in Groups A and B, the largest groups in the collection, as well as isolates with partial SSU rRNA sequences similar to those of Azospirillum spp- and Xanthobacter spp. should be tentatively classified as facultative endophytes, since Pseudomonads, Enterobacteriaceae, Azospirillum spp. and Xanthobacter spp. are commonly isolated from the rhizosphere of rice (Bally et al., 1983; Ladha et al., 1983; Watanabe et al., 1987; Oyaizu-Masuchi and Komagata, 1988; Reding et al., 1991). Bacteria in Groups C, D, E, and those with SSU rRNA genes similar to those of Azoarcus spp. should be tentatively classified as obligate endophytes, since Microbacterium spp., Alcaligenes spp., Burkholderia spp. and Azoarcus spp. have not been isolated as common rhizosphere baCteria from rice. However, many of the common nitrogen-fixing iSelates from the rhizosphere of rice have not been classified (Bally et al., 1 983; Oyaizu-Masuchi and Komagata, 1988). Additional evidence that some of these isolates may be obligate endophytes comes from the c)t.)Servations that Azoarcus spp. have never been isolated from soil 140 (Reinhold-Hurek and Hurek, 1998a) and that Burkholderia spp. appear to require a living plant host for survival (Baldani et al., 1997). Investigation of bacteria from these groups may be most fruitful for increasing our understanding of rice-endophyte interactions. Bacteria from Group F and those with partial SSU rRNA gene sequences similar to those of rhizobia and Agrobacterium spp. could be classified either as facultative- or obligate endophytes. These bacteria have not been reported as common isolates from the rhizosphere of rice. However, it is known that rhizobia and Agrobacterium spp. can survive in the soil. Better classification and careful study of colonization patterns are needed to determine what type of endophytic characteristics these bacteria display. Pathogenic and Endophytic Bacteria The difference between pathogen and endophyte is often quite subtle (James and Olivares, 1998). It is possible for two very similar bacteria to h3Ve drastically different interactions with plants. Different subspecies of 0’3 Vibacter xyli, distinct species of Herbaspirillum, and Agrobacterium S p DJRhizobium spp. are examples of bacteria having e1"1dophytic/symbiotic interactions with plants in some cases and 141 pathogenic interactions in others. In some cases the type of interaction may vary depending on the plant genotype or environmental conditions. Because of this, great care must be taken when proposing the use of endophytic bacteria as inoculum for supplying crops with biologically fixed nitrogen. Bacteria such as Azoarcus spp. or Azorhizobium spp., which have no close relatives known to cause plant disease, may be more suitable for these approaches than Burkholderia spp. or Herbaspirillum spp. which include phytopathogenic isolates. CONCLUSIONS This study shows that genetically diverse bacteria can be isolated from mechanically-abraded, surface-sterilized rice tissue. It reveals that many of these bacteria are similar to bacteria commonly isolated from the rhizosphere of rice, but others are not. It also demonstrates that a variety of nitrogen-fixing bacteria can be isolated from this niche. Some Of these nitrogen-fixing bacteria, specifically those similar to Azoarcus spp.. and Azorhizobium spp., are not commonly found in the rhizosphere 0f rice, can colonize rice under laboratory conditions, are not closely related to known plant pathogens, and have been isolated from surface- Sterilized rice tissue at multiple sites. Further study of these bacteria will Inc>I’ease our understanding of plant-microbe interactions and may provide 142 new insights into the possibility of using nitrogen-fixing endophytes to supply rice with biologically fixed nitrogen. 143 Group Strains Soil Rice Comb. ARDRA rep-PCR Types Var. A 37 5 6 11 1 3(32,4,1) B 15 3 3 4 2(13,2) 3(10,3,2) C 12 4 S 9 3(10,1x2)11(2,1x10) D 9 5 3 7 2(S,4) 3(3,3,3) E 4 2 2 2 1 1 F 4 3 3 4 1 3(2,1x2) G 4 2 1 2 1 1 H 4 3 4 4 1 4(1x4) Table 4-1 Clusters of Isolates with Similar Combined Fingerprints This table contains information on the eight clusters containing four or more isolates with combined fingerprints more similar than the replicate fingerprints of A. caulinodans. The columns contain the following information: Groups, the letter of the group shown in Figure 4-1; Strains, the number of strains in each group; Soil Types, the number of soil types in which plants containing these strains where grown; Rice Var., the number of rice varieties from which these strains were isolated; Comb, the number of soil type/rice variety combinations from which these strains were isolated; ARDRA, the number of unique ARDRA fingerprints present the cluster; rep-PCR, the number of unique rep-PCR genomic fingerprints present the cluster. The numbers in parenthesis indicate the number of strains within each cluster displaying unique fingerprints. 144 .320 A-» 00:00 2.00.00 :00. wmmcm 503.8508. o\o 303:2 mama. 9.03:0“ 308030.012 @0330 33.50.03 m. .nq: r . momAd @2000 mu. m AMA wmiwmm mm mimaocmoama 382.6% m A m A m m A \ w m m w m $3810“ 3060002012 0630 0:32.062 328200060 90:? 32320030 x....« Acaocmoemlca Sim m A A u A A w \ A A m m o Amaoomoamlcs Sim m A A w A A E A A m m m m... oBmAv. Amsoomoqmzcs 5.2.0 w AAm AwAAAm mm Amaoomgmzca 5.2m m A A m A w A \ A A m m a 32:3,. Amaocmoamzcs 5.2m m wmm quBmm mm Amaocmoamzca 5.20 m w w m m .\.A\ w m m m m mmmfioi Czamaama 9 908030815: m wmm wmmiwum mA Amsocmoqmzcs 30:2533 m w m m w m m \ m V m m A wagons“ 3080002020“ 063 0:32.082 >30<_0cm08« 90% 3 mA: 3.. ANQSROOEB omiioqmzm w a A A V a A w w: Am m m m AN03N0SEB 8330qu w a A A V a A A“: a AA m m m m w m A m we. Anoiwogca omcszoqmzm ._ w m m w m m \ w m w a o o AwosfiooEs omcsjoqmzm a w m m w m w \ w m w a o o m m._ G 8 . ANOAENocEB emcezoqmzm m A m o A w m \ A A o m m AwosfiocEB 0053033 m A m o A m m \ A m m m m 145 400.0 A-» 003.3000 00.0.: 00:00 0000.00 300. 00000 .00:..00.....08_ o\.. .83.... w00810. 0..0800008...0. 0.000 0000.50.00. >:0<.00008_. 0.000 .3». 0 08800008103 ...-mm m Amm Amo\Amw om Awomacm 00. 000 w Amm AmEAmo So 9.08:2 3200000322 00:. 0000.50.02 00303030000000 30233 :3003500 0 08800090103 4 Aum AVEAVQ mm 003030000 00. A Aum AmmEmm mm 902010“ 2220502 >0=:00m0~0:m“ >0200000A01000“ >02303<00§02 2.8380050»? 3500008100000 1 A AS00000~0203 333030 N A m w A .x. m\ A u m m m S06000~0123 6300:9303 m A m w A m A \ A w w m m XAm AS00000A0123 3.330000 4 A w A A m. w \A w A m w A500000A0123 3.330300 4 A m A A m w \ A w A m .x. 148 :CO.CC..:C::C 3:30:31 C— .Z.C.CC__::$ 33:55:90.2“ 333 pkg. 22's. 301.3337..an 0‘ AQCDSCM: sZ~C~QC~RxQUH 4:00 nDXOSO—aa< AVA AzuC: .flufiabcc fiv- 3:44:31 .mu3.%~:Uuw .70 «(0:0352. mu—UC<0.v -uhu~ afiopshu. 400.0 A-» 003.0000 0.8.0 00000 0000.00 000. 00000 8000000408. .0. .0000... 000820. 0880000820. 00330 0000.<.0.00. 00000030000 $00. 00000030000 000.. 03 00000030000 306.00..0 m ANA ANSAVA 0 0 00000030000 30000.00 m A u A A w m . A V A 0 0 00.033 00000030000 306.0000 0 A00 AOA.A.\m m 0 00000030000 30000.00 0 A u m A u 0 . A u m 0 0 -000... mm. 00000030000 306.0000 0 Au 0 A VA. A00 0 0 00000030000 30000.00 m. A00 A0032. 0 0 -0. AG. 00000030000 30000.00 . m on 00A. 0 0 m 0 0 00000030000 3%.008 . mmm 00A. mom 0 w 400,. 00000030000 00. 0.0 0 A00 A8300 0.\ 00000030000 0.0.000 0 A m w A m u . A m A 0 0 000820. 33.0080. 000...:0.o_00.20.03 E00. 000500000 000.03 0000.00 003.00 . A00 A3300 00 000.000 003000 . A00 A8300 0 0 400.0 A-» 00: «02> 0000 00000000 0.3002000 .008 0005.00 000.000 .83 .00 00.00000 5...: mm: 002) 0000 000000000 300. 03:0. .0 .00 00: 52> 0000 000000000 .83 000 00.080. .200 .008 ..0.0. 0.8.0. .00 000 00.08 00300.. 00000 000 0000.00. .00 00000 000 0000.00 0. 000.020 .83 .00 00.00000. 000.. .00 00300. 0. ..300 .00 00: «02> 0000 0.00 000000000 .0. 0000 00.08. 00000. .00 00300. 0. 000800000 .0 .00 00000000 .0. 0000 00.0.0. ...... .00 00300. 0. 003.00. 000.00.000.00 00300.. 0. 00082000 0030080. 000 o\o. .00 00.003000 0. 003.00. 000000000. ...00 88003.. 0. 0000 0800 0. 0.3.8. 00.0.00 .0 000.20 0008 .00. 0800. 149 8-3:- I I 1 I“ .l .I 0 III! II UUUUU I ------ I I I I I I I I .- I I I OOOOOOOOOOOOOOOOOOOOOOOOOOOOOO I --- . - ”It I. I I |||||| .0 - . o 1.. - . ..... . - . c . .. -- - o .................. U . I.-. I I II I I I i I , an I s Q I III II-I,I uuuuuuuuu - Mum .mmru-Oruwmrrflr. 0 . I. ERIC REP BOX - II-¢ . I I II o 0 I 0 MM . 000-000-O.-.o no. .0 u o . I I - , I - - w ' I - I I . I I I I I I II I I I I ..I IV“. IIIIIIIIIIIIIIII I I H ............. -- ‘ -" I u'I II I I I ...................................... | I I I; I o h on oom. a. I 0 . . - I . ..... - ‘I- .-:..III .I.II II I I I I“ I It II...-.IOIII.. In. I.uunIIIn..llev:ic u ..I Io II III III. I II. I ‘ .IIII I. IIIIIIIIIII I.aIIII I, I II I II II IIII . III.’ III .II III-I Iv» "“0 III- 0 CI. II I iII IIIO“CI ICI IOIOIIIIIIIIOIIIII . - -‘ I III I II IIII III I I II. I II II 0 000300 ..IIIIIOIO.oIoo . u. Inn‘l II II II n ..I I - III.IIIIIIIIIIIIIIII a C I C I -O.OI C 0 I III IIIIIII I - IIIIIIIIIII! -I ol-IOIOI 0.. o . I. uII III I. < 0 III -.-o Ivan--uo-..-. o ICIIIIIOI IIIII I III II. III.I.:IIII00¢:IIII.-.o T I} i ‘ II III I IIIIII l IIIIIIII . I... "I” au.l HHIHHJHIII.1.I 150 Figure 4-1 Dendrogram Showing Genetic Diversity of Rice Isolates This dendrogram was created using Gelcompar to show clusters of similar isolates based on linearly combined Fisal, Mspl and Taql ARDRA fingerprints and ERIC-, REP-, and BOX-PCR genomic fingerprints of 142 rice isolates, 12 soil/rhizosphere bacteria, and seven replicates of A. caulinodans (marked with a red asterisk). Also included as a reference are 23 marker lanes (M). The red line shows the level of similarity between the seven A. caulinodans fingerprints. Letters A-H indicate the groups listed in Table 4-1. The position of the letter indicates the branches shown in Figures 4-2 to 4-7, which include the fingerprints from previously isolated soil/rhizosphere bacteria. 151 p: I T ‘ Awe .. J figUrl h flirt Vi. Fisal ———_~_‘I—_’-fl_—-—I~———_————~a—-0-*fi-h——————u.—u————— Msp wmnwr*‘~*flwm--.---- -*-~—m-----‘~wvfi’"‘ 6‘ 4!; ’—-—-I-—--~—-_._—-fl---u— -----*--‘--—_fl Aw~cwfl“~-w‘—_-_--—-—— ”-‘~—D~l-*"-.~M‘o ..fl—u-u ERIC REP BOX -R1 7( 30) -R17(100) -R18(69) -R19(23) R4461) R45(42) R46(9) R4a(118) -R9(33) -R12(43) -R14(99) R84(22) R73c(36) -R19(44) -R6a(126) R55(112) R78(57) R79(108) R7708) R77(132) R5904) R73t4) R79(79) -R14(5) R42(21) R43(41) R8(28) R8(54) R32(95) R32(121) R19(122) R19(77) R33(120) R33(94) R39(129) R4009) R4106) Pl Bicol Bicol Bicol Pangasinan Bulacan Bulacan Bicol Bulacan Bicol Bicol Bicol Maahas Maahas Pangasinan Bulacan Pangasinan Maahas Maahas Maahas Maahas Pangasinan Maahas Maahas Bicol Bulacan Bulacan Bulacan Bulacan Bicol Bicol Pangasinan Pangasinan Banaue Banaue Bulacan Maa has Banaue Figure 4-2 Combined Fingerprints for Group A isolates Oryza minuta Oryza minuta Oryza minuta Pinidua |R42 lR42 Pinidua Bomalasang Oryza minute Oryza minuta Oryza minuta Oking Seroni Oking Seroni Pinidua Bomalasang Oking Seroni Pinikitan Pinikitan Pinikitan Pinikitan Oking Seroni Oking Seroni Pinikitan Oryza minuta IR42 IR42 Bomalasang Bomalasang Pinidua Pinidua Pinidua Pinidua Pinidua Pinidua IR42 Pinidua Pinikitan Shown here are the combined ARDRA and rep-PCR fingerprints of the 37 isolates in Group A, plus the combined fingerprint of Pseudomonas fluorescens (P.f.). The strain designation, soil type and rice variety from which the strain was isolated are indicated in the three vertical columns. 152 Rsal Mspl E Hi! i llll { Hi! i llll ‘ llll 5 llll 1. llll - llll llll llll llll llll ,. ~0’I -* g g...— .w Gnu-m u...- as): am ' m iii-r75 - Taql to. ERIC REP —-—-——-—.—~. BOX le1mz) |R62(15) insemo) {Rea |R950) IR104(113) l R87c(111) IR93b lR87bnio) IR104(75) R92b(34) R65(49) R98M6) I KpMSA‘l -? R89<66> ' i R89(87) E.c. Banaue Banaue Banaue Banaue Maahas Maahas Maahas Maahas Maahas Maahas Maahas Banaue Pangasinan Maahas Maahas Figure 4-3 Combined Fingerprints for Group B isolates IR74 IR74 IR74 lR74 Oking Seroni Pinikitan Oking Seroni Oking Seroni Oking Seroni Pinikitan Oking Seroni IR74 Oking Seroni Oking Seroni Oking Seroni Shown here are the combined ARDRA and rep-PCR fingerprints of isolates from group B, plus the combined fingerprints of Klebsiella planticola (KpM5A1) and Enterobacter cloacae (Be). The strain designation, soil type and rice variety from which the strain was isolated are indicated in the three vertical columns. 153 Rsal llll: E I ‘ I ' i I l | n “H llll llll llél lllf |!-: l I l. Taql 1. l 1 ERIC REP BOX l K96 T96a T116 K71 K108 : T14 K95 K37 K9 E K46 K6 Pangaanan Pangadnan Banaue Blcol Pangagnan Banaue Banaue Pangaanan Banaue Banaue Bubcan Banaue Figure 4-4 Combined Fingerprints for Group C Isolates Punknan Pinikitan Phflkfian Oryza minuta Pnudua Oryza minuta Oking Seroni Pinikitan Bonumasang Oking Seroni Bomalasang Oking Seroni Shown here are the combined ARDRA and rep-PCR fingerprints of the isolates in group C. The strain designation, soil type and rice variety from which the strain was isolated are indicated in the three vertical columns. 154 ggn_—_ a, \“gi‘n V HI Rsal II II II II II II II II II Figure Mspl II II II II ll ll . II i II II I II II o u A»... O.- “I . un— utu-II ”fl”- *0“! -r-: Taql II II ll all H l ERIC REP BOX R93(58) R 10a(86) R10a(32) R70b R70b(68) R70b(37) R76b(27) R54(1 0) R75(50) R97b(34) R97(106) Maahas Bicol BICOI Maahas Maahas Maahas Maahas Pangasinan Maahas Pangasinan Pangasinan 4-5 Combined Fingerprints for Group D Isolates Oking Seroni Oryza minuta Oryza minuta Oking Seroni Oking Seroni Oking Seroni Pinikitan Oking Seroni Oking Seroni Oking Seroni Oking Seroni Shown here are the combined ARDRA and rep-PCR fingerprints of the 10 isolates in group D plus two isolates with similar ARDRA fingerprints (strains R97b(34) and R97(106). The strain designation, soil type and rice variety from which the strain was isolated are indicated in the three vertical columns. 155 Rsal Mspl Taql ERIC REP BOX II I I I I I -R16(47) Bicol Oryza minuta II I I I I I i ; , -R11c(59) Bicol Oryza minuta II I I I I I I j 5 ‘ R22(88) Pangasinan Pinidua u I I I I I Z I I - R11c(74) Bicol Oryza minuta II I II I l I ; . mac Figure 4-6 Combined Fingerprints for Isolates in Group E Shown here are the combined ARDRA and rep-PCR fingerprints of the isolates in group E. The strain designation, soil type and rice variety from which the strain was isolated are indicated in the three vertical columns. 156 Rsal Mspl Taql ERIC REP BOX I I II I I I I I R81(90) Maahas Oking Seroni I I II I I I :I l . , g R99(65) Maahas Pinikitan I I II .- I I I . R94(17) Pangasinan Oking Seroni I I II I I I C ORS 517 I I II » I I h I I R62b(117) Banaue IR74 Figure 4-7 Isolate Fingerprints Similar to Azorhizobium Shown here are the combine ARDRA and rep-PCR fingerprints of four isolates with ARDRA fingerprints identical to Azorhizobium caulinodans (ORS 571). The rep-PCR genomic fingerprints for these isolates contain many distinct fragments, indicating diverse genomic arrangements. The strain designation, soil type and rice variety from which the strain was isolated are indicated in the three vertical columns. 157 Eco RI Bam Hl Hind III I } } O :U :D :0 O I) I ID 0 m I :0 13 (D oo (0 :D (D co <0 I] (o oo (o w i‘ : £9 0) f: :4 £9 0, ,4: :: 59 ‘33 :7 8 III 91 3 g 3 9: :3 8 a \I v V V \I v v V V V V V . .. 23.1Kb ‘_._., 9.4 Kb MD .0 ‘ ’ 6.6Kb -- 4.4Kb- 23.1 Kb ”I. nodC . ~ . 9.4Kb' 65be 44Kb. Figure 4-8 Southern Blots of Genomic DNA from Isolates Similar to Azorhizobium caulinodans Shown here are Southern blots of genomic DNA from three isolates similar to Azorhizobium caulinodans (ORS 571). The genomic DNA was digested with the restriction enzyme EcoRl, BamHl, or Hindlll and hybridized with a nifD or nodC gene probe. 158 Ball 388. Imp REFERENCES Amann RI, Ludwig W, Schleifer K (1995) Phylogenetic identification and in situ detection of individual microbial cells without cultivation. Microbiol Rev 59: 143-169 Altschul SF, Madden TL, Sch'a'ffer AA, Zhang J, Zhang 2, Miller W, Lipman DJ (1997) Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res 25: 3389-3402 Baldani JI, Baldani VLD, Seldin L, Dobereiner J (1986) Characterization of Herbaspirillum seropedicae gen. nov., sp. nov., a root associated nitrogen-fixing bacterium. Int J Syst Bacteriol 36: 86-93 Baldani JI, Caruso L, Baldani VLD, Goi S, Dobereiner J (1997) Recent advances in BNF with non-legume plants. Soil Biol Biochem 2 9:911-922 Bally R, Thomas-Bauzon D, Heulin T, Balandreau J (1983) Determination of the most frequent N2-fixing bacteria in the rice rhizosphere. Can J Microbiol 29: 881-887 Beringer JE (1974) R factor transfer in Rhizobium leguminosarum. J Gen Microbiol 84: 188-189 Boddey RM, de Oliveira OC, Urquiaga S, Reis VM, de Olivares FL, Baldani VLD, Dobereiner J (1995) Biological nitrogen fixation associated with sugar cane and rice: Contributions and prospects for improvement. Plant and Soil 174: 195-209 Christiansen-Weniger C (1996) Endophytic establishment of Azorhizobium caulinodans through auxin-induced root tumors of rice(0ryza sativa L.). Biology and Fertility of Soils 21: 293-302 de Bruijn FJ, Jing Y, Dazzo PB (1995) Potentials and pitfalls of trying to extend symbiotic interactions of nitrogen-fixing organisms to presently non-nodulated plants, such as rice. Plant and Soil 174: 225- 240 159 Dreyfus BL, Dommergues YR (1981) Nitrogen-fixing nodules induced by Rhizobium on the stem of the tropical legume Sesbania rostrata. FEMS Microbiol Lett 1 0: 313-317 Gebhard C, Turner GL, Gibson AH, Dreyfus BL, Bergersen FJ (1984) Nitrogen-fixing growth in continuos culture of a strain of Rhizobium sp. isolated from stem nodules on Sesbania rostrata. J Gen Microbiol 130: 843-848 Gillis M, Van TV, Bardin R, Goor M, Hebbar P, Willems A, Segers P, Kersters K, Heulin T, Fernandez NP (1995) Polyphasic taxonomy in the genus Burkholderia leading to an emended description of the genus and proposition of Burkholderia vietnamiensis sp. nov. for N2 —fixing isolates from rice in Vietnam. Int J Syst Bacteriol 45: 274-289 Hartmann A, Baldani Jl, Kirchhof G, Assmas B, Hutzler P, Springer N, Ludwig W, Baldani VLD, Dobereiner J (1995) Taxonomic and ecologic studies of diazotrophic rhizosphere bacteria using phylogenetic probes. In I Fendrik, M del Gallo, J Vanderleyden, M de Zamaroczy, eds., Azospirillum VI and Related Microorganisms. Springer- Verlag, Berlin, pp 415-427 Hirsch AM, Drake D, Jacobs TW, Long SR (1985) Nodules are induced on alfalfa roots by Agrobacterium tumefaciens and Rhizobium trifolii containing small segments of Rhizobium meliloti nodulation region. J Bacteriol 161: 223-230 Hurek T, Reinhold-Hurek B, van Montagu M, Kellenberger E (1994) Root Colonization and systemic spreading of Azoarcus sp strain BH72 in grasses. J Bacteriol 176: 1913-1923 Hurek T, Egener T, Reinhold-Hurek B (19973) Divergence in nitrogenases of Azoarcus spp., Proteobacteria of the B-subclass. J Bacteriol 179: 4172-4178 Hurek T, Wagner B, Reinhold-Hurek B (1997b) Identification of N2- fixing plant- and fungus-associated Azoarcus species by PCR-based genomic fingerprints. Appl Environ Microbiol 63: 4331-4339 160 James EK, Olivares FL (1998) Infection and colonization of sugar cane and other graminaceous plants by endophytic diazotrophs. Critical Reviews in Plant Sciences 17: 77-119 Kennedy IR, Pereg-Gerk LL, Wood C, Deaker R, Gilchrist K, Katupitiya S (1997) Biological nitrogen fixation in non-leguminous field crops: Facilitating the evolution of an effective association between Azospirillum and wheat. Plant and Soil 194: 65-79 Khokhar SN, Qureshi A (1998) Interactions of Azorhizobium caulinodans with different rice cultivars for increased N2-fixation. In KA Malik, MS Mirza J K Ladha, eds, Nitrogen Fixation with Non-Legumes. Kluwer Academic Publishers, Great Britain, pp 91-93 Ladha JK, Barraquio WL, Watanabe I (1982) Immunological technique for identifying Azospirillum associated with wetland rice. Can J Microbiol 28: 478-485 Ladha JK, Barraquio WL, Watanabe I (1983) Isolation and identification of nitrogen-fixing Enterobacter cloacae and Klebsiella planticola associated with rice plants. Can J Microbiol 29: 1301-1308 Ladha JK, Kundu DK (1997) Legumes for Sustaining Soil Fertility in Lowland Rice. In OP Rupela, C Johansen, DF Herridge, eds, Extending Nitrogen Fixation Research to Farmers’ Fields: Proceedings of an International Workshop on Managing Legume Nitrogen Fixation in the Cropping Systems of Asia. ICRISAT, pp 76-102 Lakshmi Kumari M, Kavimandan SK, Subba Rao NS (1976) Occurrence of nitrogen-fixing Spirillum in roots of rice, sorghum, maize, and other plants. Indian J Exp Biol 1 4: 638-639 Long S? (1989) Rhizobium-legume nodulation: Life together in the underground. Cell 56: 203-314 Louws FJ, Schneider M, de Bruijn FJ (1996) Assessing genetic diversity of microbes using repetitive-sequence-based PCR (rep-PCR). InG Toranzos, ed, Nucleic Acid Amplification Methods for the Analysis of Environmental Samples. Technomic Publishing Co, Lancaster, pp 63-93 161 Novikova N, Safronova V (1992) Transconjugants of Agrobacterium radiobacter harboring sym genes of Rhizobium galegae can form an effective symbiosis with Medicago sativa. FEMS Microbiol Lett 72: 261- 268 Martinez E, Palacios R, Sanchez F (1987) Nitrogen-fixing nodules induced by Agrobacterium tumefaciens harboring Rhizobium phaseo/i plasmids. J Bacteriol 169: 2828-2834 Moyer CL, Tiedje JM, Dobbs FC, Karl DM (1996) A computer- simulated restriction fragment length polymorphism analysis of bacterial small-subunit rRNA genes: efficacy of selected tetrameric restriction enzymes for studies of microbial diversity in nature. Appl Environ Microbiol 62: 2501-2507 Oyaizu-Masuchi Y, Komagata K (1988) Isolation of free-living nitrogen-fixing bacteria from the rhizosphere of rice. J Gen Appl Microbiol 34: 127-164 Pawlowski K, Ratet P, Schell J, de Bruijn FJ (1987) Cloning and characterization of mm and ntrC genes of the stem nodulating bacterium ORSS71, the nitrogen-fixing symbiont of Sesbania rostrata: Regulation of nitrogen fixation (nif) genes in the free living versus symbiotic state. Molecular and General Genetics 206: 207-219 Rademaker JLW, de Bruijn FJ (1997) Characterization and classification of microbes by rep-PCR genomic fingerprinting and computer assisted pattern analysis. In G Caetano-Anolles, PM Gresshoft eds, DNA markers: Protocols, Applications, and Overviews. J. Wiley & Sons, Inc. New York, pp 151-171 Rademaker JLW, Louws FJ, Rossbach U, Vinuesa P, de Bruijn FJ (1998) Computer-assisted pattern analysis of elecrophoretic fingerprints and database construction. In ADL Akkermans, JD van Elsas and FJ de Bruijn, eds, Molecular Microbial Ecology Manual, Kluwer Academic Publishers, Dordrecht, Supplement 4. 162 Ratet P, Pawlowski K, Schell J, de Bruijn FJ (1989) The Azorhizobium caulinodans nitrogen fixing regulatory gene, mm, is controlled by the cellular nitrogen and oxygen status. Molecular Microbiol 3: 825-838 Reddy PM, Ladha JK, So RB, Hernandez RJ, Ramos MC, Angeles OR, Dazzo FB, de Bruijn FJ (1997) Rhizobial communication with rice roots: Induction of phenotypic changes, mode of invasion and extent of colonization. Plant and Soil 194: 81 -89 Reding HK, Hartel PG, Wiegel J (1991) Effect of Xanthobacter, isolated and characterized from rice roots, on growth of wetland rice. Plant and Soil. 138: 221 -229 Reinhold-Hurek B, Hurek T (1998a) Interactions of gramineous plants with Azoarcus spp. and other diazotrophs: Identification, localization, and perspectives to study their function. Critical Reviews in Plant Sciences 1 7:29-54 Reinhold-Hurek B, Hurek T (1998b) Life in Grasses: Diazotrophic Endophytes. Trends in Microbiol 6: 139-202 Sambrook J, Fritsch EF, Maniatis T, eds, (1989) Molecular Cloning: A Laboratory Manual 2nd ed. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory Press Sneath PHA, Sokal FR (1973) Numerical Taxonomy, Freeman, San Francisco. Stoltzfus JR, So R, Malarvithi PP, Ladha JK, de Bruijn FJ (1997) Isolation of endophytic bacteria from rice and assessment of their potential for supplying rice with biologically fixed nitrogen. Plant and Soil 194: 25-36 Takeuchi M, Hatano K (1998) Union of the genera Microbacterium OrIa-Jensen and Aureobacterium Collins et al. in a redefined genus Microbacterium. Int J Syst Bacteriol 48: 739-747 163 Thomas-Bauzon D, Weinhard P, Villecourt P, Balandreau J (1982) The Spermosphere Model. I. Its use in growing, counting, and isolating N2-fixing Bacteria from the Rhizosphere of Rice. Can J Microbiol 28: 922-928 Ueda T, Suga Y, Yahiro N, Matsuguchi T (1995a) Remarkable N2- fixing bacterial diversity detected in rice roots by molecular evolutionary analysis of nifH gene sequences. J Bacteriol 177: 1414-1417 Ueda T, Suga Y, Yahiro N, Matsuguchi T (1995b) Genetic diversity of Ng-fixing bacteria associated with rice roots by molecular evolutionary analysis of nifD Library. Can J Microbiol 41: 235-240 Vandamme P, Pot B, Gillis M, de Vos P, Kersters K, Swings J (1996) Polyphasic taxonomy, a consensus approach to bacterial systematics. Microbiol Rev 60: 407-438 van Rhijn P, Vanderleyden J (1995) The Rhizobium-Plant symbiosis. Microbiol Rev 59: 124-142 Vinuesa P, Rademaker JLW, de Bruijn FJ, Werner D (1998) Genotypic characterization of Bradyrhizobium strains nodulating endemic woody legumes of the canary islands by PCR-restriction fragment length polymorphism analysis of gene encoding 168 rRNA (168 rDNA) and 168- 238 rDNA intergenic spacers, repetitive extragenic palindromic PCR genomic fingerprinting, and partial 168 rDNA Sequencing. Appl Environ Microbiol 64: 2096-2104 Vermeiren H, Vanderleyden J, Hai W (1998) Colonization and nifH expression on rice roots by Alcaligenes faeca/is A15. In KA Malik, MS Mirza, JK Ladha, eds, Nitrogen Fixation with Non-Legumes. Kluwer Academic Publishers, Dordrecht, pp 167-177 Watanabe I, Barraquio W (1979) Low levels of fixed nitrogen required for isolation of free-living N2-fixing organisms from rice roots. Nature 277: 565-566 164 Watanabe I, So R, Ladha JK, Katayama-Fuiimura Y, Kuraishi H (1987) A new nitrogen-fixing species of pseudomonad: Pseudomonas diazotrophicus sp. nov. isolated from the root of wetland rice. Can J Microbiol 33: 670-678 Webster G, Gough C, Vasse J, Batchelor C A, O’Callaghan K J, Kothari S L, Davey MR, Dénarié J, Cocking EC (1997) Interactions of rhizobia with rice and wheat. Plant and Soil 194: 1 15-122 Weisburg WG, Barns SM, Pelletier DA, Lane DJ (1991) 168 ribosomal DNA amplification for phylogenetic study. J Bacteriol 173: 697-703 Woese (R (1987) Bacterial Evolution. Microbiol Rev 51: 221-271 Yanni YG, Rizk RY, Corich V, Squartini A, Ninke K, Philip- Hollingsworth S, Orgambide G, de Bruijn F, Stoltzfus J, Buckley D, Schmidt TM, Mateos PF, Ladha JK, Dazzo FB (1997) Natural endophytic association between Rhizobium leguminosarum bv. trifolii and rice roots and assessment of its potential to promote rice growth. Plant and Soil 194: 99-114 You C, Zhou F (1988) Non-nodular endophytic nitrogen fixation in wetland rice. Can J Microbiol 35: 403-408 165 CHAPTER 5 USING GFP AS A BIOMARKER FOR MICROBIAL ECOLOGY INTRODUCTION Since the cloning of the Green Fluorescent Protein (GFP) gene from the jellyfish Aequorea Victoria (Prasher et al., 1992), and its expression in a variety of other organisms (Chalfie et al., 1994), GFP has rapidly become an important biomarker/bioreporter in a wide variety of eukaryotic and prokaryotic organisms. The utility of GFP as a bioreporter/biomarker lies mainly in the simplicity of the biochemical reaction generating fluorescence. GFP expression and fluorescence does not depend on the addition of co-factors or additional substrates, and only requires oxygen briefly for the auto-oxidation of GFP. When GFP is excited with light of the proper wavelength it autofluoresces. Visualization of this fluorescence can be used to monitor gene expression, to localize proteins, to isolate novel genes, or to track tagged cells (Cubitt et al., 1995; Gerdes and Kaether, 1996; Misteli and Spector, 1997). Because of these advantages, GFP is replacing some of the traditional biomarkers/bioreporters used in molecular microbial ecology (Akkermans 166 et al., 1995, 1996, 1998; Tombolini and Jansson, 1998; Unge et al., 1998; Jansson and de Bruijn, 1999). GENETIC IMPROVEMENT OF GFP Wild-Type GFP (thFP) Wild-type GFP is a 27 kDa monomer consisting of 278 amino acid residues (Prasher et al., 1992). In Aequorea Victoria GFP serves to convert blue chemiluminescence, generated by the protein aequorin, into green light (Prasher, 1995). This conversion is accomplished by means of an internal protein p-hydroxybenzylideneimidazolinone chromophore formed by the cyclization of three amino acid residues of GFP, Ser-Tyr- Gly, followed by the oxidation of the Tyr residue. These essential three residues are located at positions 65-67 of the GFP protein (Cody et al., 1993). The formation and activation of the GFP chromophore does not require additional enzymes or cofactors, and therefore heterologous expression of GFP in numerous organisms has been shown to result in fluorescence (Cubitt et al., 1995). The thFP absorbs light with an excitation maximum of 395 nm, with a second excitation peak at 470 nm. It fluoresces with an emission maximum of 510 nm (Morise et al., 1974; Ward et al., 1980). 167 Disadvantages of thFP The excitation/emission characteristics of GFP have important implications for its use as a biomarker/bioreporter. The 470 nm excitation peak allows observation of thFP fluorescence with a standard fluorescein isothiocyanate (FITC) excitation-emission filter set, or with the 488 nm line of an argon laser, equipped with appropriate filters. This allows the detection of GFP using fluorescence-activated cell sorting (FACS), flow cytometry, epifluorescent microscopy and laser-scanning confocal microscopy (LSCM). However, the fluorescent emission of the thFP when excited with higher wavelength light is less intense than its emission when excited with lower wavelength light. This renders deteCtion of the low levels of fluorescence emitted from thFP expressed in some bacteria problematic. On the other hand, excitation of thFP with higher wavelength light causes less photobleaching than excitation with lower wavelength light, making higher wavelength light a better choice for exciting GFP (Chalfie et al., 1994). One possible reason for the low fluorescence emission levels observed when thFP is expressed in selected bacteria is that much of the thFP produced is found in inclusion bodies in an insoluble non-fluorescent form 168 (Heim et al., 1994). GFP derivations with better excitation/emission spectra and higher fluorescence levels, due in part to better solubility, have increased the utility of GFP as a biomarker/bioreporter. IMPROVING GFP FLUORESCENCE Background GFP from the sea pansy Fieni/Ia reniformis contains the same chromophore as GFP from Aequorea Victoria. However, the GFP from R. reniformis has a single excitation peak at 498 nm. This difference is presumably due to the difference in amino acid residues surrounding the chromophore in the mature protein (Ward, 1980). Neither the isolated chromophore, nor the denatured protein at neutral pH, are fluorescent (Cody et al., 1993). These data suggested that changing the protein environment around the chromophore of GFP could change the excitation/emission characteristics of GFP. Mutants of G? A number of methods to generate GFP mutants with altered excitation- emission spectra and increased emission intensities have been used to increase the utility of GFP as a biomarker/bioreporter (Table 5-1). Heim et al. (1994) used hydroxylamine treatment and error-prone PCR to 169 generate mutant cDNAs encoding new versions of GFP. Colonies of E. coli harboring mutant cDNAs were screened for changes in emission color and brightness when excited with 375 vs 495 nm light. Five mutants were identified in this screen. Three of these mutants displayed significant alterations in excitation maxima, with little or no effect on emission spectra. These three mutants exhibited an increased fluorescence level (1.8 x higher). None of these three mutants contained changes in the chromophore-forming residues. The other two mutants both harbored substitutions at position 66, which forms part of the chromophore, and exhibited blue fluorescence. The excitation maxima of one of these mutants was increased. Site-directed mutagenesis of the serine residue at position 65 yielded four mutants of GFP with single excitation peaks in the 470-490 nm range and an increased level of light emission intensity (_<_6X; Heim et al., 1995). The 865T mutant GFP was found to fold four times faster than thFP, making it useful for applications requiring rapid detection of fluorescence following protein production. The faster folding may also partially explain the increased emission levels from the mutant protein. 170 Additional mutagenesis of primary GFP mutants yielded versions of GFP with multiple mutations and enhanced emission properties. Two variants had lower wavelength emission than thFP, making it possible to obtain three distinct color emissions. Fluorescence resonance energy transfer (FRET) between two version of GFP was also demonstrated (Heim and Tsien, 1996). Delagrave et al. (1995) used optimized combinatorial mutagenesis of amino acid residues 64-69 to create mutations in the GFP chromophore. The mutant proteins were screened for GFP with an increased excitation maximum using Digital Imaging Spectroscopy (DIS). These mutant proteins were named Red-Shifted GFP (RSGFP). No comparison of their fluorescence intensity with thFP was carried out. Ehrig et al. (1995) identified two GFP mutants using a strain of E. coli deficient in DNA polymerase proofreading. One GFP mutant was not fluorescent when excited with 395 nm light; the other mutant was not fluorescent when excited with 470 nm light. Neither of the corresponding mutations mapped to the chromophore. 171 Cormack et al. (1996) used a codon-based mutagenesis scheme to mutate amino acid residues in positions 55-74 and a FACS machine to isolate red shifted GFP mutants with increased emission (_<_30x thFP). They found that 90-100% of the mutant protein was soluble, which probably contributes to the increased emission level. The authors also suggested that the mutation may result in faster chrom0phore formation and increased 488 nm excitation. Using DNA shuffling, Crameri et al. (1996) isolated a non-redshift mutant of GFP that remains soluble and has emission levels 18 times that of thFP. These mutants have greatly increased the utility of GFP as a biomarker/bioreporter in bacterial systems. Variants that change the stability of GFP in bacteria have also been developed. Andersen et al. (1998) added short peptide sequences to GFP that cause bacterial proteases to degrade the protein at different rates, generating GFP variants with short half lives. These mutant GFPs allow temporal gene expression studies. Mutants of GFP designed to optimize expression in non-bacterial systems are also available. Many of these mutants have codon usage optimized 172 for a particular organism. GFP mutants have been developed for yeast (Cromack et al., 1997), plants (Sheen et al., 1995; Chiu et al., 1996; Haseloff et al., 1996; Rouwendal et al., 1997; Davis and Vierstra et al., 1998) and mammals (Zolotukhin et al., 1996; Anderson et al., 1996; Zhang et al., 1996; Yang et al., 1996; Siemering et al., 1996; Kimata et al., 1997; Stauber et al., 1998). Two of these mutants have been optimized for use at 37°C (Siemering et al., 1996; Kimata et al., 1997), which may be useful in bacterial species that grow at higher temperatures. Therefore, it is now possible for researchers to select from a number of GFP variants with different excitation and emission maxima and fluorescence levels (Table 51). Patterson et al. (1997) compared thFP with four mutant versions of GFP and concluded each version has advantages and disadvantages, depending on the application. GFP AS A BIOMARKER/BIOREPORTER IN BACTERIA: GENERAL CONSIDERATIONS Advantages of GFP A variety of systems are now available for use as biomarkers/bioreporters (Jansson and de Bruijn, 1999). GFP is easier to use than most of these 173 monitoring systems and has characteristics that allow it to perform novel bioreporter functions. The advantages described below have lead to the rapid adoption of GFP as a biomarker/bioreporter. GFP production requires expression of a small single gene product independent of additional cofactors (Chalfie et al., 1992). This facilitates DNA manipulations and construction of vectors utilizing GFP as a biomarker/bioreporter. It also allows subcellular localization of a specific protein by fusing the protein with GFP. Because GFP requires no additional substrate for fluorescence (Chalfie et al., 1992), detection is basically non-invasive, requiring no pre-detection processing and allowing rapid and accurate in vivo/in situ analysis. It also allows detection of tagged bacteria regardless of the energy status of the cells (Tombolini et al., 1997). A wide variety of detection methods suitable for analysis of bacteria expressing GFP are available (see below). This allows optimization of experimental design to answer relevant biological questions without the compromises sometimes associated with biomarkers that require special detection protocols. The intense green fluorescence produced by GFP in most bacteria allows robust, accurate detection of the tagged bacteria in environmental samples (Unge et al., 174 1998). In some cases the use of image enhancement software is needed to distinguish the fluorescence of tagged bacteria from soil and root autofluorescence (Gage et al., 1996; Tombolini et al., 1997). Problems with G=P Despite the many advantages of GFP, no biomarker/bioreporter is perfect. The major problems encountered when using GFP in bacteria are variable fluorescence, absence of detectable fluorescence, and/or toxicity of the protein itself. These problems vary from bacteria to bacteria and are often overcome by the use of mutant versions of GFP or different promoters and Ribosomal Binding Sites. However, several attempts may be needed to optimization fluorescence in new species or strains of bacteria. Clonal populations of bacteria expressing GFP sometimes exhibit variable levels of fluorescence. For example, Burlage et al. (1996) observed broad variability in fluorescence of clonal populations of Pseudomonas putida tagged with thFP. Moreover, Matthysse et al. (1996) observed variable fluorescence in clonal populations of E. coli expressing GFPmut2 grown on agar plates. 175 Tombolini et al. (1997) and Olofsson et al. (1998) observed that populations of E. coli cells containing identical plasmids encoding thFP gave rise to a non-fluorescent subpopulation. The reasons for the observed variations in GFP fluorescence have not been elucidated. At times, there is no detectable fluorescence in bacteria containing GFP, even when the gfp gene is controlled by a promoter known to be active in that bacterium. For example, Kremer et al. (1995) observed blue color in all E. coli cells harboring a plasmid with the IacZ gene controlled by a heat shock promoter (hsp60). However, when the IacZ gene in this plasmid was replaced with the gfp gene, no fluorescence was observed. Egener et al. (1998) observed fluorescence in Azoarcus sp BH72 harboring the gene encoding GFPmut2 under the control of the nifH promoter. However, they could not detect fluorescence in bacteria harboring the genes encoding thFP or PM expressed from the same promoter. We have observed similar results using the vector pRL7659fp (Tombolini et al., 1998) to tag uncharacterized soil bacteria. This vector harbors the gfp gene and the kanamycin resistance gene npt, under control of the same promoter, PpsbA. Bacteria isolated from rice and transformed with this vector were resistant to kanamycin, but no fluorescence was 176 observed (Chapter 6). While the exact reasons for lack of detectable fluorescence have not been elucidated, the use of a mutant GFP with higher fluorescence levels may overcome this problem (Egener et al., 1998). Toxicity of GFP can also be a problem when trying to obtain detectable fluorescence levels in bacteria. E. coli cells containing thFP have been reported to grow 2-3 times slower than cells expressing the cycle I” mutant version of GFP (Crameri et al., 1996). Miller and Lindow(1997) have reported that E. coli cells harboring high copy-number plasmids strongly expressing GFPmut1 were prone. to cell lysis. Evidence that GFP can be toxic in mammalian cells has also been presented (Hanazono et al., 1997). Detection of GFP The rapid adoption of GFP as a bioreporter/biomarker has been considerably enhanced by the development of a variety of techniques for detecting GFP. These methods include visualization of bacterial colonies on plates using a transilluminator, visualization of colonies or individual cells using an epifluorescent microscope or laser-scanning confocal microscope, visualization of colonization patterns by tagged bacteria on 177 plant tissue using fluorescence stereomicroscopy, counting fluorescent cells using flow cytometry or FACS, screening bacterial collections using fluorescent microtiter plate readers, and quanitation of fluorescence using spectrofluorimetry. Several reviews detailing these techniques are available (Unge et al., 1997; Khodjakov et al., 1997; Tombolini and Jansson, 1999). A filter set has been developed that optimizes detection of GFP (Zylka and Schnapp, 1996). More specialized techniques for observing GFP-mediated fluorescence have also been described. Scanning near-field optical/atomic force microscopy has been used for spatial localization of tagged bacteria in liquid (Tamiya et al., 1997) and video- endoscopy can be used to detect GFP expression in situ (Flotte et al., 1998). Specialized software can create three-dimensional images from optical sections made from GFP labeled samples using a laser scanning confocal microscope. Because GFP detection is non-invasive, real time and time-lapse videos can be used to track cells (or proteins) labeled with GFP (Gerdes and Kaether, 1996). Studies using different GFP mutants have been carried out using double and even triple labeling allowing simultaneous detection of multiple targets (Heim and Tsien, 1996; Yang et al., 1996; Lybarger et al., 1998; Yang et al., 1998). 178 Delivery and Maintenance of the gfp gene The gfp gene can be delivered to bacteria using the same methods developed for use with other biomarkers/bioreporters, e.g. electroporation, conjugation and transduction. The gfp marker/reporter gene can be maintained in the cells on a plasmid or integrated into the bacterial chromosome, via transposons or homologous recombination. A comprehensive list of GFP vectors used as biomarkers/bioreporters and their delivery vehicles is presented in Table 5-2. A list of the bacterial species studied and the types of studies performed using these vectors is presented in Table 5-3. GFP AS A BIOMARKER IN BACTERIA Use of constitutive promoters or inducible promoters to express GFP in bacterial cells allows tracking of these cells in the environment and visualization of their location (see Unge et al., 1998). Constitutive expression can be obtained by cloning GFP into a vector containing a promoter known to be constitutively expressed in the bacterium of interest (Tombolini et al., 1997). Alternatively, constitutive expression can be obtained by random insertion of a promoterless GFP gene into the bacterial genome and screening for strains that remain fluorescent under a variety of physiological conditions (Burlage et al., 1996). Inducible 179 expression of GFP can be obtained using inducible promoters such as Plac or P4310 (Christensen et al., 1996; Normander et al., 1998). Tracking Bacteria in the Environment Bacteria marked with GFP can be monitored for movement and survival in the environment. For example, Burlage et al. (1996) used a promoterless GFP transposon to obtain fluorescent P. putida strains and monitored their movement through columns of defined medium. They also performed similar experiments with E. coli cells harboring a plasmid containing the gfp gene under control of the inducible Plac promoter. In addition, Left and Leff (1996) used E. coli cells expressing GFP constitutively to monitor survival of the tagged bacteria in stream water. Heuermann and Haas (1998) and Josenhans et al. (1998) used GFP as a fluorescent tag in Helicobacter. Tombolini et al. (1997) used a Tn5 derivative transposon containing the gfp gene expressed from a constitutive promoter to tag P. fluorescens cells, and showed that the tagged cells could be detected by flow cytometry, "black light blue" lamp illumination, epifluorescent microcopy and laser-scanning confocal microscopy. In addition, dual tagging experiments using GFP and Lux allow simultaneous tracking of tagged bacteria cells and monitoring of their energy status (Unge et al., 1999). 180 GFP can also be used to monitor transfer of plasmids between bacteria. Christensen et al. (1996) used a plasmid containing GFP controlled by an inducible promoter, P4910, to monitor plasmid transfer on semi-solid surfaces. The donor strain of P. putida did not contain the T7 RNA polymerase gene therefore the T7 Pd>10 promoter was inactive and the bacteria were not fluorescent. The recipient strain carried the T7 RNA polymerase gene integrated in its chromosome. The T7 RNA polymerase activates the Pc1>10 promoter and the bacteria harboring PCD10-gfp are fluorescent. Thus, fluorescence resulted only when the plasmid was transferred to the recipient strain. Normander et al. (1998) used a similar system with GFP expression regulated by Plac to observe plasmid transfer in the phylloplane of bush bean and on polycarbonate filters. Localizing Bacteria in situ Bacteria marked with GFP can be visualized in situ, allowing identification of the niche they occupy. For example, Kremer et al. (1995) and Luo et al. (1996) followed infection of mice cells using Mycobacterium bovis marked with GFP. The infection of human macrophages and epithelial cells by Mycobacterium avium marked with GFP has also been observed (Parker and Bermudez et al. 1997). Valdivia et al. (1996) used GFP to viSualize infection of live mammalian cells by Salmonella typhimurium, 181 Yersinia pseudotuberculosis and Mycobacterium marinum. Gage et al. (1996) used GFP to monitor early events in rhizobial infection and nodule formation on alfalfa using a Sinorhizobium meliloti isolate carrying a plasmid constitutively expressing GFP. Bloemberg et. al. (1997) used a plasmid-borne GFP to distinguish tagged Pseudomonas aeruginosa from untagged Burkholderia cepacia bacteria attached to an abiotic surface. They also used GFP to monitor P. fluorescens bacteria associated with the roots of tomato seedlings. The biocontrol strain P. chlororaphis MA 342, which controls fungal disease on cereal crops, was chromosomally tagged with two COpieS of the gfp gene to allow visualization of the pattern of colonization of the cells on barley seeds by fluorescence stereomicroscopy. GFP-marked E. coli and Serratia marcescens cells have also been used to study flocculation in activated sludge (Olofsson et al. 1998). We have also used bacteria marked with a transposon containing GFP to monitor the interactions between rice roots and endophytic bacteria (Chapter 6). Monitoring Gene Expression in the Environment Biomarking can be combined with bioreporting to obtain spatial and temporal information on gene expression in situ. For example the urea- inducible promoter from the ureD gene fused to the gfp gene has been 182 used to infer the location of Proteus mirabilis urease gene expression during infection of the mouse urinary tract (Zhao et al., 1998). A nifH promoter-gfp fusion has been used to monitor expression of nitrogen- fixing genes in Azoarcus sp. BH72 in soil and on rice roots (Egener et al., 1998). Moreover, Moller et al. (1998) used P. putida cells containing the gfp gene expressed from promoters controlling the expression of proteins involved in the biodegradation of toluene to visualize gene expression in flow chambers and observe the effects of various community members on gene induction. GFP AS A BIOREPORTER IN BACTERIA Localizing Protein Fusions The fusion of GFP to a protein of interest and monitoring of the expression of the fusion in vivo permits cellular localization of the quion protein (Cubitt et al., 1995; Gerdes and Kaether, 1996; Misteli and Spector, 1997; Jansson and de Bruijn, 1999). Although GFP loses its fluorescence when truncated (Cubitt et al., 1995), the full-length protein can be used successfully in both C-terminal and N-terminal protein fusions (Wang and Hazelrigg, 1994). A flexible linker may be required to maintain the function of both fused proteins (Glaser et al. 1997). In bacteria, GFP 183 has been used extensively as a bioreporter to study cell division in E. coli and Bacillus subti/is. Ma et al. (1996) have created fusions of the cell division proteins FtsZ and FtsA to GFP and reported that both C-terminal and N-terminal fusions of FtsZ with GFP showed midcell localization and polymerization. However, the fusion proteins did not function properly in mediating cell division. C-terminal fusions of the FtsA protein with GFP also showed proper midcell localization. FtsZ and FtsA interactions have also been examined with GFP fusions in Agrobacterium tumefaciens and S. meliloti (Ma et al. 1997). Hale and de Boer (1997) used a C-terminal fusion of the ZipA protein with GFP to localize ZipA in a ring structure at the division site, but this fusion protein also failed to function properly in cell division. Raskin and de Boer (1997) found that C-terminal fusions of MinE with GFP were biologically active and formed a central ring in an FtsZ- independent manner. Yu et al. (1998) used truncations of FtsK fused to GFP to determine that as little as 15% of the FtsK protein's N-terminus is needed for proper targeting. 184 GFP has also been used to study proteins involved in septation during vegetative division and sporulation in Bacillus. Barak et al. (1996) reported that a SpollE-GFP fusion was localized to septum formation sites in Bacillus megaterium. However, one study using thFP in a SpollE-GFP fusion found reduced sporulation in B. subfilis (Aringoni et al., 1995). Edwards and Errington (1997) used a Dile A-GFP fusion to visualize sites of old, nascent, and possibly future vegetative cell division in B. subtilis. Ju et al. (1997) used GFP fused to pro-csE to localize this fusion protein and determine in which compartment of B. subtilis pro-csE processing takes place. Wu et al. (1998) used the GFPmut1 variant to construct a SpollE-GFP fusion. This fusion appeared to be fully functional during sporulation and enabled the localization of SpollE during septation. The localization of proteins involved in chromosome behavior has also been studied using GFP. For example, a SpoOJ-GFP fusion was created that resulted in a functional protein which permitted the visualization of chromosomal partitioning in live B. subtilis cells (Lin et al., 1997; Glaser et al., 1997). Webb et al. (1997) used a GFP-Lacl fusion to localize specific regions of the B. subfilis chromosome in vegetative and sporulating cells by inserting a lacO cassette at specific sites in the chromosome. The 185 same system was used to compare chromosome and plasmid behavior during replication of E. coli (Gordon et al. 1997; Kim and Wang, 1998). Finally, Kohler and Ma (1997) used a HBsu-GFP fusion to visualize the location of this essential histone-like protein in B. subfilis. Monitoring Gene Expression The expression of GFP from a promoter of interest and visualization of fluorescence also permits the monitoring of gene expression in vivo (Cubitt et al., 1995; Misteli and Spector 1997; Jansson and de Bruijn, 1999). For example, the localization of gene expression during sporulation in B. subfilis was investigated using promoters controlled by various sigma factors fused to GFP (Webb et al., 1995; Arigoni et al., 1995). Lewis and Errington (1996) used GFP to localize expression directed by the pre-spore-specific promoter of the dacF gene and the mother-cell-specific promoter of the sap/VA gene in B. subtilis. GFP has been used to study the expression of various promoters from Mycobacterium during infection of macrophages (Dhandayuthapani et al., 1995; Via et al., 1996). Collins et al. (1998) used GFP as a reporter of metabolic activity in Mycobacterium tuberculosis. Finally, Chrast—Balz and van Huijsduijnen (1996) used GFP controlled by the tetA gene promoter to evaluate tetracycline antagonists. 186 Isolating Environmentally Regulated Genes Promoterless GFP constructs can also be used to screen for promoters that are induced under a specific set of physiological conditions. For example, Kremer et al. (1995) inserted fragments of the M. bovis genome in front of the gfp gene and found that 3 to 5% of the chimeric genes showed differing levels of fluorescence on agar plates. In an effort to isolate genes important for S. typhimurium survival during infection of acidic macrophages, Valdivia and Falkow (1996) used a similar strategy to find promoters induced under acidic conditions. Suarez et al. (1997) have developed both plasmid and mini-Tn5—based promoterless GFP vectors. P. putida, Pseudomonas spB13 and Alcaligenes eutrophus were mutated with mini-Tn5::gfp. Up to 5% of the bacteria with chromosomal integration of the mini-Tn5 were fluorescent, displaying varied levels of expression. Moreover, Handfield et al. (1998) developed a promoterless GFP vector that allows rapid selection for active promoters by fusing aspartate B-semialdehyde dehydrogenase with GFP and using it as a selectable marker in Asd' strains. Josenhans et al. (1998) constructed a mini-Tn3 based promoterless gfp transposon for use in Helicobacter pylori and H. mustelae. Other promoterless GFP vectors have also been 187 described, but their utility has not yet been demonstrated (Podbielski et al., 1996; Kalogeraki and Winans, 1997). Another example of the utility of GFP as a bioreporter is the construction of pGreenscript (lnouye et al., 1997). In this vector GFP replaces the B- galactosidase (lacZ gene) as the marker for insertion of DNA in the multiple cloning site of the popular cloning vector pBluescript. 188 CONCLUSIONS The large number of studies employing GFP as a biomarker/bioreporter clearly demonstrate the value and versatility of GFP. While problems with GFP as a biomarker/bioreporter in bacteria have been encountered, many have been overcome by using mutant versions of GFP, adding appropriate translation initiation sites, or adding flexible linkers to protein fusions. The many successful studies using GFP and the continued development of new techniques incorporating GFP will undoubtedly lead to the use of GFP in more species of bacteria to answer diverse questions in molecular microbial ecology. 189 485 m. 0.8 39:0 385:9. 3598.9. 98.9.0: 8.86: 3.29.5 :9. 3B.» E... 5885.0 .50.”... 20:6 ”$9.30 moo . 03:6 9 m... 5?. Io mmomm wwm m: ... V Im.3 m. m_. 3...... to ..mu< swiwom mom ..mm Im.3 m. m... I...» .u... :3... 5.3.8.. mom ..mm Im.3 m. m... 8?. 8.. <8... wmm f8 0.3 Im.3 m. m... .8: << .892 saw ..mo 20 :93 m. m... 5?. 03.4 084 .50 m. o m 16.3 m. m... .mmm Oxam w .380. 2.18.7 <.mA> wmm m. o .m Om39. m. m... 5.... 300.8.— mmbz. 080. 0mm: .3... wow 20 OmeSEm 9 m... .mom 0.835.3013 man... 084 ..mm mo. mo-wo 0938: m. m... 5.... 0.83:.» 08>. <37 03> .2.. mo. mo. cc 0938: m. m... 88 0.83:6 mmm0. mwm> mo. 9. ...o-mo 0938: m. m... 58 088482 Anew. h.oo m. m 20 m:..m m. m... .85 0....umnnn0 mmmm0 am. wow 20 m:...o m. m... .95 0.83:8. 0mm. mmm0.0wm> mo. m... No-3. >:a98: m. w.. 89.. 982.595 0.83:6. s...: 398...... .m: mo. m... ...o-mo >:a98: m. m_. .mmm maama 982.55 0.83:3... 5...: 398...... 8.. mo. m. . .o-mo >:a98: m. m.. .wmm maama m.u..u>>. 0.83:6. 5...: 398...... .9. mo. m. . 3&0 >:a98: 9 m... 53. maama c.8305 0.83:5. 5...: 398...... .m: mo. 3. No-2. >:a98: m. m_. .wom maama 03> wmm> 20 mom m 30% m. m_. .3. 030 080 20 m. w . 505.0 9 m... .8. 190 .785 m. oo:..::ma 0.8 :93 moaacmm 3:5.98 58.5.5: 8.8.0: 322.5 :9. .33.» 33.: 5.95.20 8:... <8... <32U mm. .Km . Im.3 m:a ..m.m: .8: <3 <82. 2.5.. .584. 3.. 3m . Im.3 m:a 3.0: .08 <.mw>. 2.33. <~ <82. :mw<. <3mI. sum ..mo . Im.3 m:a 3.6: 58 13mm. 38...... 53>. 22.... 8a.. 82.. 73mm? xmwmm mos m. s. 20 10.3 m:a ...m.m: .oom 20 - :0 955 28 o. ..5 3.34:0 05.5550 000.. 0003.0. -_.o-.... 03 >:5900: 0. 5.. 50.50. ...>>. . 000 0..0>..m 3.:.-._.:m 05.5550 000.. 00030.0. 5-.: x3 >:5900: 0. 5.. .0.50. b..>>. . 000 0503.0 3.3-30 05.5550 000.. 00030.0. 5.... x3 >:5900: 0. 5.. .0.50. ...<>v . 000 0503.0 3.3-...:0 05.5550 000.. 00030.0. .5... x3 >:5900: 0. 5.. 50.5b .r<>. . 000 0503.0 3.3-30 05.5550 000.. 00030.0. .52: x3 >:5900: 0. 5.. .0.50. P>><. . 000 0003.0 3.34:0 05.5550 000.. 00030.0. 40-... x3 >:a900: 0. 5.. .055 .>><. .000 050.500 3.:.--..:m 05.5550 000.. 000330.. ._.o-._.. 03 >:5900: 0. 5.. .0.50. .>0<. . 000 050).». 3.:.--_.:m 05.5550 000.. 00030.0. ..o-q. x3 >:5900: 0. 5.. 00.50 $05 . 000 0...0>~0 3.3-30 05.5550 000.. 00030.0. 5-... ..3 >:5900: 0. 5.. .250. .000 050.500 3.34:0 05.5550 08.. 00030.0. .52: x3 >:5900: 0. 5.. .0.50. .000 00.204 0.503.200.5090. 000...". 000..m 2.000 20 >0 9.00:. 0. 5.. 59.<5..<0. . 000 00.200 0.503.200.5090. 000...". 000=m 2.000 20 >0. x3 >...00:. 0. 5.. 59.<5..<0 500.0:05 .9 .000 090300035. 3.095203 3.0000... 0_503.5.500.0:05 .9 000...”. 25:25:05. 2.000 20 >0. 03 095.. 0. 5.. .000 090300035. .:.00.5..0:. 80.0: .2. 000=m 192 455.0 0.» 0033005 <00.0_. 4.50 0330.9 45:05.5: 010 49335.90 00.00505 00.90500 2530 3:550: <9.5:. 39.9 0000 0.503.5.0c0.0 0.50 20 0003:.» 20 >0. 40 0.003590 0. 5.. 59.<5..<0. . 004 0000 05035.0me 59.<5..<0. 0.50 20 00035.0 20 >0. 40 0.003590 0. 5.. .000 0030» 05035.0me 59.<5..<0. 0.50 20 00030.0 20 >0 0.003090 0. 5.. .005 4:0000. 4:0 0.50 20 5.000 20 40. 03. 03 09.500 0. 5.. .000 040000 0503.5 44 095 20 2.000 20 >0 0.550 0. 5.. 0330.9 .005 0.05.000 0.503.560.0022 4: 5 .03 ..5:0.5..0:5. 9.0.0 5 45:09.0..05. x3 0.950.055 0. 5.. 59.<5..<0. 9030.9 30.0: 5.. 0.00 .33 .000 .0.> 0>0x00-0..0 003.004 3.34:0 0.030.05000 20 3000 20 x3 0:505:00: 0. 5. .000 003.040 3.34:0 40 05.5 20 2.000 20 03 0:505:00: 0. 5.. 0330.9 .000 40r-0.0 9055 ..00. 5:00 0503.5 44 005 20 5.000 20 x3 0:505:00: 0. 5.. .00. ”.000. .40r 59.<5..<0. 0330.9 .000 0000 0503.5 0.50 20 2.000 20 >0 0.0:.00... 000000.030 0.503.5.03<00. 00000 20 2.000 20 x3 0:5:55..c.:505:. 59.<5..<0. 0. 5.. .000 025000.040 0.503.557.2000 50:0 20 2.000 20 x3 0:5:550. 03 m5£5.50 5:5 0:3300035. .:.00.5..0:. 30.0: s: m...:0.0: .004 0.<.< 0000200 35035 3:90 59.<5..<0 0...... ..5:0_5..0:5_ 00030.. 20 >0 m00:9 0. 5.. 500.0:05 .0. 30.0: s: .000 0:.0300035. .:.W0.5..0:. 2:: 04000.... 0.055 :00. .5:00 0.5035 0.40 ..33 20 0004 20 00. 03 0500 0. 5.. .000 2.03 0950.0..5. 05.30:0..5. 040000 0355 :00. .5:00 0.5035 0.3 ..33 20 2.000 20 00. 03 0500 0. 5.. .000 :33 00.50.0__5V 05.30:0__5. 000.0. 45 0.503.200.3090. 0000.. ..5:0.5..0:5. 0004 20 >0. 03 0.509 0. 5.. 59.<5..<0 500.0:05 .0. 30.0: s: .004 0.3300035. 3.095203 00100,. 00000 0.503.5.00>0-.0 05.50>O 00:35. W00: 20 20 >0 00.50: 0. 5.. 59.<5..<0. 000 . 004 00100 0:500 0.50 ..5:0.5..0:5_ 0004 20 20 15.0 5:5 50 000. 80.0: s: .004 N.0> 194 450.0 0-0 833005 08 <00.9 4<00 0330.9 4.30.5.3: 0.00 49335.90 00.00820 00.9030 2530 335.3: <5..5:. 39.9 0.< M4000 0.503.5.0c.0.50 <00.9 0330.0..000 20 00030.0 20 00 15:5..03 0. 5.. 00040.. 59.<5..<0? . 000 0030:093. 0.503.5.00_c000..0. 0.50 20 000-.. 20 >0 300.0 0. 5.. 59.<5..<0. . 004 00. 03033300.... 59.<5..<0 Q50... ..5:0_5..0:5_ 2.000 20 03 .3 0. 5.. .004 500.0:05 .0. 30.0: 2. 09038035. 3.00.5..03 0.Pm00 05.2005 0.503.5 .00.. 00.0.50 :0:0 20 0004 20 x3 X53095... 5:5 <00.9. <<.:5:0 .004 00000000 0.503.5.0>0_.0m 0.50 ..5:0_5..0:5. 20 20 >0 .03 5:5 <<5:0 59.<5..<0. 30.0: 2. .000 0000 03.000 2835.000 0 59.<5..<0. 40 0330.0. ..5:0_5..0:5_ 2.000 20 >0 .639 5:5 30.0: 2. 35.530. .004 I00: 00.800 0.503.5.0UI00 59.<5..<0 00.050 ..5:0.5..0:5_ 2.000 20 >0. 03 00:9 5:5 500.0:05 .0. 30.0: 2. 35.530. .004 0:.0300035. 3.00%03 I00: 00.000 0.5033 20 ..5:0_5..0:5_ 2.000 20 >0 0039 5:5 30.0: 2. 35.530. .004 10% 0003-.. 0.5033 .2..: :0000 00000 05 2.000 20 x3. 03 0.039 0. 5.. 330050.055. .00..05..0: .000 B 0003..» 0.5033 .2..: :0:0 :0000 200 2.000 20 x3. 03 0.039 0. 5.. 3.805093. 30.053: .000 195 450.0 0-0 00:2:005 <00.9 4.30 0330.9 4.30.5.3: 000 49335.90 0030.503 00.9030 2530 3.2520: <5..5:. 39.9 000..»0 0.503.5.00_c000..0. 20:0 ..0. 0. 20 2.000 20 >0. 03 ..02.0 5:5 59.<5..<0 500.0:05 .0. .9335. m...:0.0: .000 0:.0300035. 3.00.5203 ..5:0_520:5_ 303:0. 000200 0.503.5.00_c000..0. 20:0 ..0. 0. 20 <00: 20 >0. 03 ..02.0 5:5 59.<5..<0 500.0:05 .0. .9335. m...:0.0: .000 0:.0300035. 3.00.5203 ..5:0_520:5. 303:0. 000304 0.503.5.0030003. 20:0 ..0. 0. 20 0004 20 >0. 03 ..02.0 5:5 59.<5..<0 500.0:05 .0. .93.:5. m...:0.0: .000 0:.0300035. 3.00.5203 ..5:0_520:5. 30.9% 000.....5 0.503.5.00_c000..0. 000..... ..5:0.520:5_ 0004 20 >0. 03 ..02.0 5:5 59.<5..<0 500.0:05 .0. 30.0: 2. m...:0.0: .000 0:.0300035. 3.00.5203 M00_<> 000.34 0.503.5.00_:000..0. 5500 ..5:0.520:5_ 0004 20 >0. 03 ..02.0 5:5 59.<5..<0 500.0:05 .0. 30.0: 2. m...:0.0: .000 0:.0300035. 339.5203 0500 000000 0.503.5.00m305. 0000.. 230.5203. 20 20 03 E: 0. 5.. .004 59.<5..<0. 30.0: 2. 0000... 0>05 0.503.5.000 59.<5..<0. 0.50 ..5:0_520:5_ 00030.0 20 >0 35 0. 5.. .000 30.0: 2. 0.0> 000x. 8 0.503.5.000 59.<5..<0. 0.50 20 00030.0 20 >0 35 0. 5.. .000 00M 0.503.5.000 59.<5..<0. 0.50 20 00030.0 20 >0 35 0. 5.. . 000 3005 .0 ..0N 196 450.0 0-0 000.3000 000<50<0. 00500. ..:0. <00.0. 4<00 0.0320. 4.500.500: Q0 40.33530 0030.503 00.0.0000 2530 335.3: <5..5:. 352.0. 0N05 0.5033300 00.35.33 0.50 ..5:0_5.3:5_ 00030.0 20 >0 35 0. 5_. .000 303: 2. 0.0M 00. 00.0 0.050 000. 300+ 0.5033 0.50 4. 0000.0 00030.0 20 23 3500.000 0. 5_. .000 00. 00.0 0.050 000. 300+ 0303.0 000.0 4. 0000.0 0003:.» 20 23 3500.000 0. 5_. .000 04000.0 400 0000 .. 0000.0 00030.0 20 >0 3500.000 0. 5.. .000 0:433.- 3.0.-400 0000 4. 0000.0 0003:.» 20 40. >0 3503.000 0. 5.. 4000.0 .000 0000000 0303.0 .0CO. 0. 0 0.50 .50N 00.00- 2.000 20 >0 3:3. 5:0 0.0002 00.35.33 00.05.00 . 00. 00.00: 03030 8.00.0.0 0.50 .50N 00.00- 00030.. 20 >0 3:3. 500 0.0002 00.35.30. 00.05.00 . 00. 00.0004... 0.50.3.0 6.00.0.0 0.50 00.00- 00030.. 20 >0 3:3. 500 0.0002 00.35.30. 00.05.00 0. .00. 00500. .45. 050.00: .02 000.. 0.050 000. 0. 0.00 .50N 00.00- 00030.. 20 >0 3:3. 500 ..3002 .5000 0.5033 .0<000. 0.0.3.0.. 00.05.00 .00. 000<50<0. 05000 .02 000.. 0.050 000. 0. 0.00 .50N 00.00- 2.000 20 >0 3:3. 500 0.0002 .5000 0.50.3.0 .0<000. 0.030.0.. 00.05.00 .00. 000<50<0. 054.... .02 000. 0.050 000. 0. 0.00 00.00- 00030.. 20 >0 350. 500 C0002 .5000 0303.0 .0<000. 0.030.0.. 00.05.00 0. .00. 197 450.0 0-» 00:00:00 <00.0. .300 0.0320. 4.500300: Q00 40.33533 0030.503 50.0.0000 2530 335.3: <5..5:. 352.0. 0<000.45 .02 000. 0.050 000. 0.50 00.00- 0003:: 20 >0 350. 500 ...0002 .5000 0303.0 .0<000. 00.05.00 0. .00. 00.35.33 00500. .45. 0.5>»0 3.0.400 03 20 0003:.00 4...... 03. 00 30:0. 0. 5. .000 0.5>00 33.400 0: 40 500.. 0003:.00 4.40 53 30:0. 0. 5.. .000 40..." 0.050 000. .5000 0303.0 05.50.00 40 550.. 0003:.00 4. -...0 53 20.35000. 0. 5.. 0530.00 0.0.. 00.35.30. 5.50. .000 0020.0 0.5033 .0..00.00000. 00:0 20 2.000 .3 -.3 00 0000.0.0... 0. 5_. 0:.0.00 <00.0.. .000 005..»5 0303.0 0.50 ..5:0_5.3:5_ 0003:.» 20 ..3 550...: 500 00 30.0: 2. 500. .00. 3.000m 0.5.»05 0.50.3.0 05505 ..5:0_5.3:5_ 0003:.» 20 53 550...: 500 00 30.0: 2. 500. .00. 3.:00m 0>0000 33.400 0000 m.00_. 5.0m 0.0.00 20 500>-0 5:0 0:5.0N 0. 5.. 00:0 0.0 .00. 0>0<» 0303.0 35300000. 000500 0. 500 m.00_. 5.0m 2.000 ..:5 4. >0 0:5.0N 0. 5_. 00.35.33 0. .0 .50003 0000 .00.0.0 .00. 08303.. 0>0<0 0.5033 65300000. 0000 m.00.. 5.0m 2.000 ..05 .3 >0 0:5.0N 0. 5_. 00.35.30. 0000 30.0.0 . 00. 0.030.0b 0>0<0 0303.0 65300000. 0000 m.00_. 5.0.”. 2.000 + ..:w 4. >0 0:50.. 0. 5_. 00.35.33 00:0 02. 30.0.0 . 00. 03303.. 198 450.0 0-0 000..:000 02m <00.0. 4.00 0330.0. 4.500.500: 000 40.33583 00.02520 50.0.0000 2530 3.05000 <5..5:. 35.00. 050.0006 400 0000.. 4. 0000 . o 0.. 20 x3 403023. 0. 5_. .000 0:40.0 33.400 0000.. 4. 0000 .0 0.. 20 ..3 40300.3. 0. 5_. .000 C4055. 33.400 0000.. E000 . o 0.. 20 x3 3&0. 5.. .000 0.20000 0.503.0.0m000o 0000 20 00030.0 20 >0 <5.0.<.5 5:0 00..<5..<0. 05:62 .000 00<000.. 0.503.0.000<00 .00.... 20 00030.0 20 >0 <5.0.<.5 500 00..<5..<0. 05:82 .000 000<. 0.5033 655me 0.50 20 2.000 20 >0 <5.0.<.5 0. 5_. 00..<5..<0. .000 0004u 0.503.0.05_0000..0. 00.0. ..5:0_5.3:5_ 2.000 20 ..:. <<000 0. 5.. .000 00..<5..<0 000.0000 .0. .003: 2. 0:.0300035. 3.055000. 00.m 00.2.0 0.503.0.05.0000..0. 00...H ..500.5..005_ 2.900 20 O3 <<000 0. 5.. .000 00..<5..<0 000.0000 .0. .0030 2. 0:.0300035. 3.095000. 00.m 00.2.0 0.503.205.5023. 00.5 ..5:0_5.3:5_ 2.000 20 ..3 <<000 0. 5_. .000 00..<5..<0 000.0000 .0. .003: 2. 0:.0300035. 3.00.5000. 00.0 00.200 0.503.0.05_0000..0. 000M000 ..5:0_5.3:5_ 2.0.00 20 >0. x... <<000 0. 5_. .mom 00..<5..<0 000.0000 .0. 30.0.. s. . 0:.0300035. 3.00.5200 005.0. 0.503.085.0081... . 002m W00 m. . 00..<5..<0 000.0000 .0. 500 5.605. 2.00.0 20 O... { 0:.0300035. 3.0 .500: .0060 2. <<000 0. 5_. .mmm 199 450.0 0-0 000.3000 <00.0. 4.00 0330.0. 4.500.500: 000 40.3.05.0.0 0200.520 50.0.0000 2530 3.0500: <5..50. 35.00. 00.25 0.503.0.05_0000..0. 0.030.0..000 25 2.000 20 >0 <<000 0. 5_. .000 00..<5..<0. 0>4.. 250.03.050.00 000<0 ..5:0_5.3:5_ .0000 .0 25 .0.-m <<000 0. 5_. .00. 00..<..5<0 .0. 3.00.5..00 .0030 2. .50. 3.0 .00 40.0 .0000. m00<0 05030. 0.503.0.05.0000..0. 20:0 ..0. Q 25 00030.. 20 >0. 03 SE 0. 5_. .000 00..<5..<0 000.0000 .0. .0.335_ 0:.0300035. 3.00.5..00. ..5:0_5.3:5_ 30.000. 000.000 0.503.0.05.0000..0. 000...”. ..5:0_5.3:5_ 00030.. 25 >0. 03 .2: 0. 5_. .005 00..<5..<0 000.0000 .0. .0030 2. 0:.0300035. 3.0E5000. 000..m 0X00 0.503.0.05_0000..0. 0.50 ..5:0_5.3:5_ 00030.0 25 >0 <0 0. 5_. .005 00..<5..<0. .0030 2. 0.0.. 0:5m-5U. 0.503.0.05.0000..0. 0.055 ..5:0_5.3:5_ 00030.0 25 >0 N050 0. 5_. .005 0005 00..<5..<0. .0030 2. :05 080 0_503.0.05_0000..0. 0.030.0..000 25 00030.0 25 00 2.5 0.5.. .00. 00..<5..<0. 0.5033650 00..<5..<0. 0.030.0..000 20 00030.0 20 053 2.5 0. 5_. .00. 0>m0. 0.503.2050 00..<5..<0. 20> ..5:0_5.3:5_ 00030.0 20 053 .55 0.5.. .00. .0030 2. _ 0 0.0> 0Nm0.o 0 50.0. .050 00..<5..<0. ..0N ..500.5..o:5_ 30.8 2. 0.00350 20 05... .05 0. 5_. .00. 0.0N /. 200 450.0 0.» 003.0000 <00.0. 4.00 0330.0. 4.500.500: 000 40.3.05.0.0 00.00.520 00.0.0000 2530 3.05000 <5..50. 2.5.00. 0>m0~0 0.503.0A00_0000..0. 00> 500.363. 0003:.» 20 00 :5 0. 5_. .00. 00..<5..<0. €0.00 2. 0.0> 0Nm0~0 0.503.0A0w.0000..0. ..0N ..500.5..005_ 00030.0 20 00 2.5 0. 5.. .00. 00..<5..<0. €0.00 2. 0.0M 0>m000 0.503.000.050 000. .5000 20> ..500.5..005_ 00030.0 20 0.0. 4. .,..5 0.5.. .00. .000 00..<5..<0. 30.00 2. 0.0> 0Nm000 0.503.03850 000. .5000 ..0N ..500_5..005_ 00030.0 20 00. 4. 2.5 0.5.. . 000 .000 00..<5..<0. 30.00 2. 0.0M 0< 0.503.0.0§ ..03 W00... 0.00. <0.0.00 000.53.00 00> ..03 >0.0050.0..03 830.50.000 500 0303.30.03 30....0... 5.0 5.00 5<5._50_0. 20 - 00 05.5 250 0.<00 .0 .00 0500. 0000 201 450.0 0.0 000 <00.0.0 <00.0.0 000.0000 .0 0....N0 000 50 5 0.035.00..0.0.000..0. .0 050.005. 400 .520 ..0.0 .00 0530 0. .00 <00.0.. .00 200 0. <00.0.. .00 0330.0. 0000 .0 0x0.000 000. 53. ..500.5..00 .0...5..00 000 303.0000 .0 .00 0500.. .00 000 <5..50. 0000. 50.. .0.3.05.0.0 303.0000 .0 .00 0500. 500 .00 00.00.520 350.0. 00 .00 <00.0.. 202 45.0.0 0-0 00.200. 00204 0.05303 4<00 0. 0.00... 4.00 0. <00.0. 2530 000 00.0.0000 <00.0. 30.50. >0.0050.0...:.: 0005.30 0.0.030 3<0.<00 .0 000.03 0.50.3.0 0>m00. 0Nm00. 00030.0 2.5 0. 5.. .00. 330.5203 335.3: 500 0.00.. .00.. 3.0300000 500 3.5.00 >30 0.503.00 305300000 0.05.0 0330.0. 30.000 33.400 0>0>om 0.0.0 0 005.0N 0. 5_. .00. 0530.30 >~053:0 00 000020 >N05.0:0 03.00030 0.. 003300035. 0010200 00030.0 m0000. 0. 5_. .000 0:40 L00000 .0 00.. 500 0: ..00 30.0 .EB30 050330 ..005.30 0.0.0.00 3<0.<00 .0 003300035. 0_.m-000-4 2.000 05.5.. 0. 5_. .000 SEES: 000.0.5..00 000.03 3.3500: 3.095000 050330 0:030 0005.30 03.030 33.30 .0 003300035. 005.00 2.000 >..000. 0. 5_. .000 000.0.5..00 000.03 3.3500: 3.0fi5000 050330 0:030 0005.30 0.0.0.00 33200 3 000.03 003300035. 000.0%. 000-4 m025.00 500 m...:0.0: 3.35000 3.09.5000 .000 050330 0330 0005.30 0.0.030 33200 .0 003300035. 000.03 000-4 0.500. 0. 5.. .00.. 003300035. 05.....003P 3.09.5000 050330 0:030 0005.30 0.0.030 .0 008.330 200.0 003300035. 00. 2.000 00 0. 5_. .004 0300005905000 0.500 .0.00@.00 033:0 0:030 0005.30 1000 003300035. 00.800 2.000 30.0. 500 2.5 . 00. 3.0E30 050.30 033.0 ..005.30 0000 0.0.0003: 5:0 0.0.030 003300035. 000. ..: 000-4 ..02.0 500 m...:0.0: 3<0.<00 .0 0003.500: 3.10%5000 .000 050...:0 0:030 0005.30 0000 0.0.0003: 500 0.0.0.00 003300035. 000. .3 000-4 ..02.0 500 m...:0.0: 3<0.<00 .0 0020.500: .0.L0nP.5..00 .000 9.03:0 0:030 0005.30 0.0.0.00 33200 .0 003300035. 000000 E: 0. 5_. .004 00.0300035. 05.3.0050! 3%.00 033:0 0:030 .00.mm_<5..h.00 www.mmmumfihw 00000 “00....00300035. 0>40. 0>4. .. 2.000 <<000 0. 5_. .00 0 0.5..00 00.2.0. 005:0. 203 450.0 0-0 00000000 ! 0.050.03 4<00 0. 0.00.. 4.30 0. <00.0. 2530 emu 00.0.0000 <00.0. 30.50. 0500.00 003.00 0.000 003300030 05.000030 003300035. 00020. <<000 0. 5_. .00.. 3.00.5000 050.000 0003.0 0005.30 0.0.030 3<0.<00 3 003300035. 000.000 00030.. 5.0 0. 5.. .000 03.0.5000 000.03 3.35000 3.09500: 00000335 00.. 0005.30 0.0.0.00 3<0.<00 .0 000.03 00500 00100 000-4 15.0 500 00 000. .00. 3.35000 000001035 000 0005.30 0.0.030 33200 .0 000.03 0.5033 0>0 00030.0 .<.5 0. 5_. .000 3.3500: 000001035 00.. 0005.30 0.0.0.00 33300 .0 000.03 0.5033 000.300 00030.0 .<.5 0.5.. .000 3.35000 M0000...0...5 00.. 0005.30 0.0.0.00 33300 .0 000.03 0.5033 00N 00030.0 .<.5 0.5.. .000 3.35000 000001035 00.. 0005.30 0.0.030 33300 .0 000.03 0.5033 0N0. 00030.0 .<.5 0.5.. .000 335000 00002335 00.. 0005.30 0.0.0.00 33.30 .0 000.03 0.5033 000. .0 00030.0 050.00 500 00 000. 3.35000 .00. 000001035 00.. 0005.30 0.0.0.00 33300 .0 000.03 0.5033 000.00 00030.0 050.00 500 00 000. 3.35000 .00. 000001035 00.. 0005.30 03.0.00 3<0.<00 3 000.03 0.5030 0X00. 00030.0 <0 0. 5. .000 3.3500: 000001035 00.. 230.3. ..00 0.50.0.0 .0 50.35.00 0.5030 00.00 2.000 0.0.0000 0. 5_. .000 .0 . 0.0000 000300.035 00.. 00.00.30 500 0000000 0. 050.005 3.00400 0C40.0.0x 1.. C80 0. m_. .wmm 0:09 00.00 0.00 500 300.3000 .00.. 000.0.. 0.5.00 00.00 .0090 0.0000103... 00.. 00.3.5.0 0.00 03300.00 .0.... . 0’10... 000000 . 0.05 0.50.0.0 mm-m ..5:000 0C 0.000 0003.3 N000 0. m_. .oom 204 ._.mc_m m-w 833ch 062.53 .23 a «8% .23 2 <38. 2m3o 96 303833 538.. 3:83 M83138 8: roomfim 3.08:5 3528 3 32:3 2335 3me nNmor 033% gm 2 m_. $3 ..:. cm 8338: m3 3:3. :5: 3332.03 mag 6.88 u_mm3am M83302» 8:. 3038. 385338 2 3925 3 2836 no? .3an rm: man rm: Ema LE on 333 £32 5um 3.03803 mmogzoim 8: 9:3 0:83omo3m man 2335 2833 umomo 0030: m" m_. Emu .3» 33.2038 magmzoim no: men? 2835 83203.20 0.335 umouwAwmt mmmA 53 man. Ewan 3mm: .3» :93 manglnim no: fimox 3o-m_u_u 936 w 0:36»me m. 000 03030.2 0:0 3.0800005 00001030000 283.2 003.00. 0.0033 00:0.2 3 0.0033 .0.-30.03200 00030.00 2230:02 0. 0.. .000 00080020 30 03.80.08 0. 0:0: 000: >0.» 00001030000 0080.8: 0:0 000330 0. 000.010 3.34:0 0..-30.0.5. 0.. 300 0. 0.. .000 100800000 0030 0.00 0:0 303.230 .00: 0:29. 00.20... 0.0.00 :03m .0015 00001030000 283.2 0.0033 .8:0.2 0< 0.0033 .5008 0.000 0:108:00: 0. 0.. 00.10 5.0000 003E928: 000.0000. .000 03010000103 <.0:0_.N0 8030.8: 0. 301.0000 0.0033 0..-0000 000-..- 0000 0. 0.. .000 300.80. .00. 0050 0.. 000.010 03010000103 0000.30 08830 38.80 3 000.03 0.0033 0>m00. 0Nm00. 0003:.» 2.0 0. 0.. .00-1 30080 2.000. 230.8: 0:0 0.000 30: 3.200003 0:0 8.0.00 207 400.0 0-0 00:00:00 0600.03 .200 0. 0E0< 0.<00 0. <0.0.00 20.00 96 30.000000 <00.0.. 3503 0000000.? <.0:0_.N0 000.000 3.00200 :00. 00:0 0.00.3.0 0205 2.9.6 <0.0.<.0 000 00:62 $6235.33 000 00:0. 5.00.00 00:0 5.00 :02 .000 38303. 00.30020 .u.00 00.0 30:80 0.030.000 000 .00. 0.00.0.0 0.2000 00.0350 <0.0.<.0 000 00:62 3.23533 .0. 006000.00 0< 000.000 .0 .000 3003000m00 $00020 2.0002 200 0.00:5 .0 00.20.00 0.00.0.0 4070.0 2.00.0 0.0.0000 0. 0.. .000 300000000 0.:mb0 $0. 0003 .0005? 000210. <.0:0_.N0 000.000 3.00000 :00. 00:0 0.00.0.0 00.05 .5020 <0.0.<.0 000 00:62 800005000 000 00:0. 3.00.00 00:0 :0.00 202 .000 3:00.05 ...00.0 0-0 010 2.0160 000.000 :0. 0. 000.000 .0 5.0.00 93v 000 0000 :000 00 0 0.030001060000000 ..00 .00.0 ..0.0 .00 000:0 000 0000.00 0. 000.000 :000 .0 .00 0.:0<. .00 300 0. <00.00 .00 00.00 0. .00 <00.0q 000 .00 $0.0 <0000. 000.00 00 .00 <00.00 208 REFERENCES Akkermans ADL, van Elsas JD and de Bruijn FJ, eds, Molecular Microbial Ecology Manual. Kluwer Academic Publishers Dordrecht, 1995, 1996, 1998 Andersen JB, Sternberg C, Poulsen LK, Bjorn SP, Givskov M, Molin S (1998) New unstable variants of green fluorescent protein for studies of transient gene expression in bacteria. Appl Environ Microbiol 64: 2240-2246 Anderson MT, Tjioe IM, Lorincz MC, Parks DR, Herzenberg LA, Nolan GP, Herzenberg LA (1996) Simultaneous fluorescence- activated cell sorter analysis of two distinct transcriptional elements within a single cell using engineered green fluorescent proteins. Proc Natl Acad Sci 93: 8508-8511 Arigoni F, Pogliano K, Webb CD, Stragier P, Losick R (1995) Localization of protein implicated in establishment of cell type to sites of asymmetric division. Science 270: 637-640 Barak I, Behari J, Olmedo G, Guzman P, Brown DP, Castro E, Walker D, Westpheling J, Youngman P (1996) Structure and function of the Bacillus SpollE protein and its localization to sites of sporulation septum assembly. Mol Microbiol 19: 1047-1060 Bloemberg GV, O’Toole GA, Lugtenberg BJJ, Kolter R (1997) Green fluorescent protein as a marker for Pseudomonas spp. Appl Environ Microbiol 63: 4543-4551. Burlage RS, Yang ZK, Mehlhorn T (1996) A transposon for green fluorescent protein transcriptional fusions: application for bacterial transport experiments. Gene 173: 53-58 Chalfie M, Tu Y, Euskirchen G, Ward WW, Prasher DC (1994) Green fluorescent protein as a marker for gene expression. Science 263: 802-805 209 Chiu W, Niwa Y, Zeng W, Hirano T, Kobayashi H, Sheen J (1996) Engineered GFP as a vital reporter in plants. Curr Biol 6: 325-330 Chrast-Blaz J, van Huijsduijnen RH (1996) Bi-directional gene switching with the tetracycline repressor and a novel tetracycline antagonist. Nucleic Acids Research 24: 2900-2904 Christensen BB, Sternberg C, Molin S (1996) Bacterial plasmid conjugation on semi-solid surfaces monitored with the green fluorescent protein (GFP) from Aequorea Victoria as a marker. Gene 173: 59-65 Cody CW, Prasher DC, Westler WM, Prendergast FG, Ward WW (1993) Chemical structure of the hexapeptide chromophore of the Aequorea green-fluorescent protein. Biochemistry 32: 1212-1218 Collins LA, Torrero MN, Franzblau SG (1998) Green fluorescent protein reporter microplate assay for high-throughput screening of compounds against Mycobacterium tuberculosis. Antimicrob Agents Chemother 42: 344-347 Cormack BP, Bertram G, Egerton M, Gow NA, Falkow S, Brown AJ (1997) Yeast-enhanced green fluorescent protein (yEGFP) a reporter of gene expression in Candida albicans. Microbiol 143: 303-31 1 Cormack BP, Valdivia RH, Falkow S (1996) FACS-optimized mutants of the green fluorescent protein (GFP). Gene 17 3: 33-38 Crameri A, Whitehorn EA, Tate E, Stemmer WP (1996) Improved green fluorescent protein by molecular evolution using DNA shuffling. Nat Biotechnol 14: 315-319 Cubitt AB, Heim R, Adams SR, Boyd AE, Gross LA, Tsien RY (1996) Understanding, improving and using green fluorescent proteins. Trends Biochem Sci 20: 448-455 Davis SJ, Vierstra RD (1998) Soluble, highly fluorescent variants of green fluorescent protein (GFP) for use in higher plants. Plant Mol Biol 36: 521 -528 210 Delagrave S, Hawtin RE, Silva CM, Yang MM, Youvan DC (1995) Red-shifted excitation mutants of the green fluorescent protein. Biotechnology 13: 151-154 Dhandayuthapani S, Via LE, Thomas CA, Horowitz PM, Deretic D, Deretic V (1995) Green fluorescent protein as a marker for gene expression and cell biology of mycobacterial interactions with macrophages. Mol Microbiol 17: 901-912 Edwards DH, Errington J (1997) The Bacillus subti/is DileA protein targets to the division septum and controls the site specificity of cell division. Mol Microbiol 24: 905-915 Egener T, Hurek T, Reinhold-Hurek B (1998) Use of green fluorescent protein to detect expression of nif genes in Azoarcus sp. BH72, a grass-associated diazotroph, on rice roots. Mol Plant Microbe Interact 11: 71-75 Ehrig T, O'Kane DJ, Prendergast FG (1995) Green-fluorescent protein mutants with altered fluorescence excitation spectra. FEBS Lett 367: 163-166 Flotte TR, Beck SE, Chesnut K, Potter M, Poirier A, Zolotukhin S (1998) A fluorescence video-endoscopy technique for detection of gene transfer and expression. Gene Ther 5: 166-177 Gage DJ, Bobo T, Long 92 (1996) Use of green fluorescent protein to visualize the early events of symbiosis between Rhizobium meliloti and alfalfa (Medicago sativa). J Bacteriol 178: 7159-7166 Gerdes HH, Kaether C (1996) Green fluorescent protein: applications in cell biology. FEBS Lett 389: 44-47 Glaser P, Sharpe ME, Raether B, Perego M, Ohlsen K, Errington J (1997) Dynamic, mitotic-like behavior of a bacterial protein required for accurate chromosome partitioning. Genes Dev 11: 1160-1168 211 Gordon GS, Sitnikov D, Webb CD, Teleman A, Straight A, Losick R, Murry AW, Wright A (1997) Chromosome and low copy plasmid segregation in E. coli: visual evidence for distinct mechanisms. Cell 90: 1113-1121 Hale CA, de Boer PA (1997) Direct binding of FtsZ to ZipA, an essential component of the septal ring structure that mediates cell division in E. coli. Cell 88:175-185 fluorescent protein retroviral vectors: low titer and high recombination frequency suggest a selective disadvantage. Hum Gene Ther 8: 1313- 1319 Hanazono Y, Yu JM, Dunbar CE, Emmons RV (1997) Green F Handfield M, Schweizer HP, Mahan MJ, Sanschagrin F, Hoang T, Levesque FD (1998) ASD-GFP vectors for in vivo expression technology in Pseudomonas aeruginosa and other gram-negative bacteria. Biotechniques 24: 261-264 Haseloff J, Siemering KR, Prasher DC, Hodge S (1997) Removal of a cryptic intron and subcellular localization of green fluorescent protein are required to mark transgenic Arabidopsis plant brightly. Proc Natl Acad Sci 94: 2122-2127 Heim R, Prasher DC, Tsien RY (1994) Wavelength mutations and posttranslational autoxidation of the green fluorescent protein. Proc Natl Acad Sci 91: 12501-12504 Heim R, Cubitt AB, Tsien RY (1995) Improved green fluorescence. Nature 373: 663-664 Heim R, Tsien RY (1996) Engineering green fluorescent protein for improved brightness, longer wavelength and fluorescence resonance energy transfer. Curr Biol 6: 178-182 Heuermann D, Haas R (1998) A stable shuttle vector system for efficient genetic complementation of Helicobacter pylori strains by transformation and conjugation. Mol Gen Genet 257: 519-528 212 lnouye S, Ogawa H, Yasuda K, Umesono K, Tsuji Fl (1997) A bacterial cloning vector using a mutated Aequorea green fluorescent protein as an indicator. Gene 189: 159-162 Jansson JK, de Bruijn FJ (1999) Biomarkers and Bioreporters to track microbes and monitor their gene expression. In AL Demain, JE Davies, eds, Manual of Industrial Microbiology and Biotechnology. 2nd ed. ASM Press, Washington, DC pp 651-665 Josenhans C, Friedrich S. Suerbaum S (1998) Green fluorescent if protein as novel marker and reporter system in Helicobacter sp. FEMS Microbiol Lett 161: 263-273 Ju J, Luo T, Haldenwang WG (1997) Bacillus subtilis pro-sigmaE fusion protein localizes to the forespore septum and fails to be processed when synthesized in the forespore. J Bacteriol 1 79: 4888-4893 Kalogeraki VS, Winans sc (1997) Suicide plasmids containing promoterless reporter genes can simultaneously disrupt and create fusions to target genes of diverse bacteria. Gene 188: 69-75 Khodjakov A, Cole RW, Rieder Cl. (1997) A synergy of technologies: combining laser microsurgery with green fluorescent protein tagging. Cell Motil Cytoskeleton 38: 311-317 Kim SK, Wang JC (1998) Localization of F plasmid SopB protein to positions near the poles of Escherichia coli cells. Proc Natl Acad Sci 95: 1523-1527 Kimata Y. Iwaki M, Lim CR, Kohno K (1997) A novel mutation which enhances the fluorescence of green fluorescent protein at high temperatures. Biochem Biophys Res Commun 232: 69-73 Kohler P, Marahiel MA (1997) Association of the histone-like protein HBsu with the nucleoid of Bacillus subtilis. J Bacteriol 1 79: 2060-2064 Kremer L, Baulard A, Estaquier J, Poulain-Godefroy O, Locht C (1995) Green fluorescent protein as a new expression marker in mycobacteria. Mol Microbiol 17: 913-922 213 Leff LG, Leff AA (1996) Use of green fluorescent protein to monitor survival of genetically engineered bacteria in aquatic environments. Appl Environ Microbiol 62: 3486-3488 Lewis PJ, Errington J (1996) Use of green fluorescent protein for detection of cell-specific gene expression and subcellular protein localization during sporulation in Bacillus subtilis. Microbiol 142: 733- 740 Lin DCH, Levin PA, Grossman AD (1997) Bipolar localization of a chromosome partition protein in Bacillus subtilis. Proc Natl Acad Sci 94: 4721 -4726 Luo Y, Szilvasi A, Chen X, DeWolf WC, O’Donnell MA (1996) A novel method for monitoring Mycobacterium bovis BCG trafficking with recombinant BCG expressing green fluorescent protein. Clin Diagn Lab Immunol 3: 761-768 1'" ' .uu' In" Lybarger L, Dempsey D, Patterson GH, Piston DW, Kain SR, Chervenak R (1998) Dual-color flow cytometric detection of fluorescent proteins using single-laser (488-nm) excitation. Cytometery 31: 147-152 Ma X, Ehrhardt DW, Margolin W (1996) Colocalization of cell division proteins FtsZ and FtsA to cytoskeletal structures in living Escherichia coli cells by using green fluorescent protein. Proc Natl Acad Sci 93: 12998-13003 Ma X, Sun 0, Wang R, Singh G, Jonietz EL, Margolin W (1997) Interactions between heterologous FtsA and FtsZ proteins at the FtsZ ring. J Bacteriol 179: 6788-6797 Matthysse AG, Stretton S, Dandie C, McClure NC, Goodman AE (1996) Construction of GFP vectors for use in Gram-negative bacteria other than Escherichia coli. FEMS Microb Lett 145: 87-94 Miller WB, Lindow SE (1997) An improved GFP cloning cassette designed for prokaryotic transcriptional fusions. Gene 191: 149-153 214 Misteli T and Spector DL (1997) Applications of the green fluorescent protein in cell biology and biotechnology. Nature Biotechnology 15: 961 -964 Moller S, Sternberg C, Andersen JB, Christensen BB, Ramos JL, Givskov M, Molin S (1998) In situ gene expression in mixed-culture biofilms: evidence of metabolic interactions between community members. Appl Environ Microbiol 64: 721 -732 Morise JG, Shimomura D, Johnson FH, Winant J (1974) intermolecular energy transfer in the bioluminescent system of Aequorea. Biochemistry 13: 2656-2662 Niwa H, lnouye S, Hirano T, Matsuno T, Kojima S, Kubota M, Ohishi M, Tsuji Fl (1996) Chemical nature of the light emitter of the Aequorea green fluorescent protein. Proc Natl Acad Sci 93: 13617- 13622 Normander B, Christensen BB, Molin S, Kroer N (1998) Effect of bacterial distribution and activity on conjugal gene transfer on the phylloplane of the bush bean. Appl Environ Microbiol 64: 1902-1909 Olofsson AC, Zita A, Hermansson M (1988) Floc stability and adhesion of green-fluorescent-protein-marked bacteria to flocs in activated sludge. Microbiol 144: 519-528 Parker AE, Bermudez LE (1997) Expression of the green fluorescent protein (GFP) in Mycobacterium avium as a tool to study the interaction between Mycobacteria and host cells. Microb Pathog 22: 193-198 Patterson GH, Knobel SM, Sharif WD, Kain SR, Piston DW (1997) Use of the green fluorescent protein and its mutants in quantitative fluorescence microscopy. Biophys J 73: 2782-2790 Podbielski A, Spellerberg B, Woischnik M, Pohl B, Lutticken R (1996) Novel series of plasmid vectors for gene inactivation and expression analysis in group A streptococci (GAS). Gene 177: 137-147 Prasher DC (1995) Using GFP to see the light. Trends Genet 1 1: 320- 323 215 Prasher DC, Eckenrode VK, Ward WW,Prendergast FG, Cormier MJ (1992) Primary structure of the Aequorea Victoria green- fluorescent protein. Gene 111:229-233 Raskin DM, de Boer PA (1997) The MinE ring: an FtsZ-independent cell structure required for selection of the correct division site in E. coli. Cell 91: 685-694 Rouwendal GJ, Mendes O, Wolbert EJ, Douwe de Boer A (1997) Enhanced expression in tobacco of the gene encoding green fluorescent protein by modification of its codon usage. Plant Mol Biol 33: 989-999 Sheen J, Hwang S, Niwa Y, Kobayashi H, Galbraith DW (1995) Green-fluorescent protein as a new vital marker in plant cells. Plant J 8: 777-784 Siemering KR, Golbik R, Sever R, Haseloff J (1996) Mutations that suppress the thermosensitivity of green fluorescent protein. Curr Biol 6: 1653-1663 Stauber RH, Horie K, Carney P, Hudson EA, Tarasova NI, Gaitanaris GA, Pavlakis GM (1998) Development and applications of enhanced green fluorescent protein mutants. Biotechniques 24: 462- 466 Suarez A, Guttler A, Stratz M Staendner LH, Timmis KN, Guzman CA (1997) Green fluorescent protein-based reporter systems for genetic analysis of bacteria including monocopy applications. Gene 196: 69-74 Tamiya E, Iwabuchi S, Nagatani N, Murakami Y, Sakaguchi T, Yokoyama K, Chiba N, Muramatsu H (1997) Simultaneous topographic and fluorescence imagings of recombinant bacterial cells containing a green fluorescent protein gene detected by a scanning near- field optical/atomic force microscope. Anal Chem 69: 3697-3701 Tombolini R, Jansson JK (1999) Monitoring of GFP tagged bacterial cells. In R LaRossa, ed, Bioluminescent Protocols. Methods in Molecular Biology. Humana Press (in press). 216 Tombolini R, Unge A, Davey ME, de Bruijn FJ, Jansson .K (1997) Flow cytometric and microscopic analysis of GFP-tagged Pseudomonas fluorescens bacteria. FEMS Microbiol Ecol 22: 17-28 Unge A, Tombolini R, Davey ME, de Bruijn FJ, Jansson ..K (1998) GFP as a Marker Gene. In ADL Akkermans, JD van Elsas, FJ de Bruijn, eds, Molecular Microbial Ecology Manual. Kluwer Academic Publishers, Dordrecht, pp 6.1.13z1-16 Unge A, Tombolini R, Molbak L, Jansson J (1999) Simultaneous monitoring of cell number and metabolic activity of specific bacterial populations using a dual gfp/quAB marker system. Appl Environ Microbiol 65: 813-821 Valdivia RH, Falkow S (1996) Bacterial genetics by flow cytometry: rapid isolation of Salmonella typhimurium acid-inducible promoters by differential fluorescence induction. Mol Microbiol 22: 367-378 Valdivia RH, Hromockyj AE, Monack D, Ramakrishnan L, Falkow S (1996) Applications for green fluorescent protein (GFP) in the study of host-pathogen interactions. Gene 173: 47-52 Via LE, Curcic R, Mudd MH, Dhandauthapani S, Ulmer RJ, Deretic V (1996) Elements of signal transduction in Mycobacterium tuberculosis: in vitro phosphorylation and in vitro expression of the response regulator MtrA. J Bacteriol 178: 3314-3321 Wang S, Hazelrigg T (1994) Implications for bcd mRNA localization from spatial distribution of exu protein in Drosophila oogenesis. Nature 369: 400-403 Ward WW, Cody CW, Hart RC, Cormier MJ (1980) Spectrophotometric identity of the energy-transfer chemophores in Rani/Ia and Aequorea green-fluorescent protein. Photochem Photobiol 31: 611-615 Webb CD, Decatur A, Teleman A, Losick, R (1995) Use of green fluorescent protein for visualization of cell-specific gene expression and subcellular protein localization during sporulation in Bacillus subtilis. J Bacteriol 177: 5906-5911 217 Webb CD, Teleman A, Gordon S, Straight A, Belmont A, Lin DC, Grossman AD, Wright A, Losick R (1997) Bipolar localization of the replication origin regions of chromosomes in vegetative and sporulating cells of B. subtilis. Cell 88: 667-674 Wu LJ, Feucht A, Errington J (1998) Prespore-specific gene expression in Bacillus subtilis is driven by sequestration of SpollE phosphatase to the prespore side of the asymmetric septum. Genes Dev 12: 1371-1380 Yang TT, Cheng L, Kain SR (1996) Optimized codon usage and chromophore mutations provide enhanced sensitivity with the green fluorescent protein. Nucleic Acids Res 24: 4592-4593 Yang TT, Kain SR, Kitts P, Kondepudi A, Yang MM, Youvan DC (1996) Dual color microscopic imagery of cells expressing the green fluorescent protein and a red-shifted variant. Gene 173: 19-23 Yang TT, Sinai P, Green G, Kitts PA, Chen YT, Lybarger L, Chervenak R, Patterson GH, Piston DW, Kain SR (1998) improved fluorescence and dual color detection with enhanced blue and green variants of the green fluorescent protein. J Biol Chem 273: 8212-8216 Yu XC, Margolin W (1997) Ca2+-mediated GTP-dependent dynamic assembly of bacterial cell division protein FtsZ into asters and polymer networks in vitro. EMBO J 16: 5455-5463 Yu XC, Tran AH, Sun 0, Margolin W (1998) Localization of cell division protein FtsK to the Escherichia coli septum and identification of a potential N-terminal targeting domain. J Bacteriol 180: 1296-1304 Zhang G, Gurtu V, Kain SR (1996) An enhanced green fluorescent protein allows sensitive detection of gene transfer in mammalian cells. Biochem BiOphys Res Commun 227: 707-71 1 Zhao H, Thompson RB, Lockatell V, Johnson DE, Mobley HL (1998) Use of green fluorescent protein to assess urease gene expression by uropathogenic Proteus mirabi/is during experimental ascending urinary tract infection. Infect Immun 66: 330-335 218 Zolotukhin S, Potter M, Hauswirth WW, Guy J, Muzyczka N (1996) A "humanized" green fluorescent protein cDNA adapted for high- level expression in mammalian cells. J Virol 70: 4646-4654 Zylka MJ, Schnapp BJ (1996) Optimized filter set and viewing conditions for 865T mutant of GFP in living cells. Biotechniques 2 1: 220-226 219 CHAPTER 6 STUDIES ON RICE ROOT COLONIZATION BY SELECTED BACTERIA USING GFP AND GUS AS MOLECULAR BIOMARKERS ABSTRACT The endOphytic colonizing ability of nine bacterial stains isolated from mechanically-abraded, surface-sterilized rice tissue was evaluated. These strains were inoculated to sterile rice seedlings and could be re-isolated following three weeks of plant growth. Transposons harboring genes encoding a GUS or GFP biomarker were constructed and used to tag two of the nine bacterial strains previously isolated from rice tissues [R33(120) and R100(64)]. The utility of these biomarkers for in situ visualization of bacterial colonization of plant tissues was evaluated. Three dimensional reconstruction of optical sections generated via laser scanning confocal microscopy was used to visualize spatial relationships between bacteria expressing GFP and plant tissues. Colonization sites for the two tagged strains and Sinorhizobium meliloti, a control strain, were examined in detail using both GUS and GFP as biomarkers. Little internal colonization of rice tissue by these bacteria was observed although both clumps of bacteria and individual cells were observed on the surface of 220 infected rice roots. On very rare occasions an isolated epidermal cell harboring large numbers of R1002:ng bacteria was observed. 221 INTRODUCTION As pointed out in Chapter 1 and 2, the long term goal of the experiments outlined in this thesis is to isolate and characterize stable “endophytic” bacteria from rice plants. Our expectation was that the isolation of bacteria from mechanically-abraded, surface-sterilized rice plant tissue (Chapter 2) would enrich for endophytic bacteria. However, it was also realized that some non-endophytic bacteria would ‘escape’ surface sterilization and be isolated along with the endophytic bacteria. Therefore, the ability of these bacteria to colonize plant tissue endophytically must be verified. To accomplish this we focused on re- infection of gnotobiotically grown rice seedlings with selected bacterial isolates (Chapter 2) and in planta microscopic localization of bacteria harboring biomarkers. The isolation of a bacterial strain following re-infection of gnotobiotically grown rice seedlings demonstrates the ability of that strain to form an intimate association with the plant host, resulting in the protection of these bacteria from the sterilizing agent. However, this approach does not address the possibility that the bacteria reside on the plant surface and escape sterilization by the production of polysaccharide capsules or 222 other means. Nor does this method determine the sites of colonization. Therefore, more direct approaches are required to determine whether a strain is capable of colonizing the rice tissue endophytically, and the sites and extent of colonization. In principle, it should be possible to visualize the colonization of rice by putative endophytes using light microscopy, fluorescence microscopy, and electron microscopy (You and Zhou, 1988; Hurek et al., 1994; Yanni et al., 1997; Reddy et al., 1997). However, unless the location of colonization can be predicted, it is difficult to find the bacteria by random scanning of tissues. Moreover, unless the plants are grown under gnotobiotic conditions or highly specific antibodies are available, the identity of the bacteria observed will remain questionable. Recently different types of biomarkers have become available that facilitate tracking bacteria in the environment (Wilson, 1995; Unge et al., 1998; Jansson and de Bruijn, 1999; Chapter 5). The use of biomarkers facilitates the localization and unambiguous identification of endophytes in infected plant tissues (Hurek et al., 1994; Barraquio et al., 1997; Reddy et al., 1997; Webster et al., 1997; Vermeiren et al., 1998). In the 223 studies cited above, the biomarker B-galactosidase (LacZ) or B- glucuronidase (GUS) was employed. We chose to use the Green Fluorescent Protein (GFP) and GUS biomarkers as they offer distinct advantages such as low background and easy detection (Chapter 5). Both of these markers have been used successfully to detect tagged bacteria in (legume) plant tissues, such as those resulting from infection with rhizobia (Wilson et al., 1995; Gage et al., 1996). Here we will describe the development of transposon borne gfp and gus biomarker constructs, the tagging of selected (putative) endophytes and control strains, and a microscopic analysis of rice seedlings infected with the tagged stains. 224 MATERIAL AND METHODS Bacterial Strains and Growth Conditions The strains and plasmids used in this study are listed in Table 6-1 an 6-2 respectively. Escherichia coli S17-1 Apir strains, harboring plasmid pUT Tn5Km, pRL765gfp and anPSBS-T were grown at 37°C on LB medium (Silhavy et al., 1984) supplemented with kanamycin (20 pg/ml). Eco/i strain HB101, harboring the conjugal transfer helper plasmid pRK2013, was grown at 37°C on LB medium supplemented with kanamycin (20 ug/ml). E. coli strain DHSa harboring pT893F, or pUC 4K and E. coli strain S17-1 kpir harboring plasmid pCAM121 were grown at 37°C on LB medium supplemented with carbenicillin (100 ug/ml), kanamycin (20 ug/ml), or spectinomycin (100 ug/ml), respectively. All other bacterial isolates (Table 6-1) were routinely grown at 28°C on TY medium (Beringer, 1974). Minimal media used for selection of transconjugants following the introduction of the biomarker transposons into bacteria were GTS medium (Kiss et al., 1979) containing 0.2% (NH4)2804 as a nitrogen source, LSO medium (Elmerich et al., 1982) containing 1g/l (NH4)2$O4 as a nitrogen source, and Davis and Mingolini minimal medium (Atlas, 1997), each supplemented with kanamycin (20 pig/ml). 225 Generation of Sterile Rice Seedlings, Bacterial Infection and Re- isolation Seeds of the rice variety Lemont (accession # Pl 475833, National Small Grains Research Facility, Aberdeen, ID, USA) were gently dehulled, incubated in 70% ethanol for five minutes, washed with sterile distilled water, treated with a fresh solution of 30°/o Clorox (17.5 ml sterile distilled water, 7.5 ml Clorox Bleach, 30 1.11 Tween 20) for 45 minutes, and washed six times with sterile distilled water. Seeds were placed on plates containing TY medium and incubated at 28°C for two days to allow germination. Seeds showing no bacterial or fungal contamination were placed in 25 mm X 200 mm tubes containing sterile sand watered with Fahraeus medium (Fahraeus, 1957). After two days the seedlings were inoculated with 100 pl of bacterial cultures grown for 24 hours in liquid TY medium at 28°C. Seedlings were watered with Fahraeus medium to keep the sand moist. Bacteria were re-isolated from the inoculated rice seedlings as follows: Seedlings were grown from 14 to 16 days and carefully removed from the growth tubes to keep the roots intact. Each seedling was placed in a 15 ml conical tube (Sarstedt, lnc., Newton, NC, USA) containing 7 ml of sterile distilled water, shaken for 15 minutes, washed with 7 ml sterile 226 distilled water, treated with 7 ml of 30°/o Clorox solution for 15 minutes and washed six times with sterile distilled water. Each seedling was macerated in 1.5 ml microfuge tubes with sterile sand. Following maceration 1 ml distilled water was added and 20 pl of macerate was spread onto plates containing TY medium. Generation of Sterile Rice Seedlings and Bacterial Infection for Microsc0pic Studies Rice seedlings used for in-situ visualization of colonization by bacteria expressing GFP or GUS were generated and grown as follows: Two seeds of the rice variety Lemont, sterilized and germinated as described above, were placed between two “pillows” in a 10 cm x 10 cm x 15 cm pot. Pillows were constructed by folding a 15 cm x 40 cm piece of very fine spectrum cloth in half and heat sealing two edges, filling the resulting bag with 800 ml of vermiculite and heat sealing the top, as described by Szczyglowski et al. (1998). Each pot was watered with approximately 300 ml of 0.5x Hoaglands solution (Sigma, St. Louis, MO, USA) and placed in a growth chamber under 16 hour light at 28°C and 8 hours dark at 22°C. After two days, each seedling was inoculated with 1 ml of stationary phase culture of gfp or gus tagged bacteria and analyzed two to three weeks after inoculation. 227 Construction of gfp and gus Biomarker Transposons All DNA manipulations were carried out using standard molecular biology protocols (Maniatis et al., 1982). Plasmids used in this study are listed in Table 6-2. A mini-Tn5 transposon containing gusA under the control of the Paph promoter and a kanamycin resistance cassette was constructed as follows: The streptomycin and spectinomycin resiStance cassette was removed from anSSsgusA21 (Wilson et al., 1995) by digestion with the restriction enzyme Smal and replaced by a 1.3 Kb kanamycin resistance cassette from pUC 4K. The resulting vector was named p121TH1. A mini-Tn5 transposon containing a gfp gene encoding the GFPS65-T variant (Chapter 5) under the control of the Ptrp promoter and a kanamycin resistance cassette was constructed as follows: A 1 kb Smal fragment from pTB93F (Gage et al., 1996), containing the gfp gene under control of the Ptrp promoter, was cloned into the Smal site of pBluescript, digested with Kpnl and BamHI, and the resulting fragment inserted into pUC18Not (Herrero et al., 1990). The resulting plasmid was digested with Noll and the Nail fragment inserted into the Notl site of pUT/mini-Tn5 Km (de Lorenzo et al., 1990). The resulting vector was named anPSGS-T (Figure 6-1). 228 Introduction of Biomarker Transposons into Bacteria. The gfp and gus containing Tn5 transposon derivatives were introduced into the target bacteria via conjugal transfer. E. coli strain S17-1 Apir, harboring the appropriate transposon, was grown overnight at 37°C in LB medium supplemented with 20 ug/ml kanamycin. E. coli strain HB101, harboring pRK2013 (Figurski and Helinski, 1979), was grown overnight at 37°C in LB medium supplemented with 20 ug/ml kanamycin. The recipient strains were grown at 28°C to late log phase in TY medium. Bacteria were washed 2x with TY medium, mixed in a 1:5 donorzrecipient ratio, or a 1:1:5 donorzhelperzrecipient ratio, concentrated 5-fold and spotted on TY plates. Donor, helper, and recipient cells were spotted alone as controls. After 24 hr incubation at 28°C the spotted cell mixtures were resuspended in sterile distilled water and several dilutions were spread on plates containing a selective medium (see above). 229 Histochemical Staining for Gus Activity and Microscopy Plants inoculated with bacteria harboring the gus marker gene were examined for B-glucosidase activity using 50ug/ml X-gluc (Gold Biotechnology, Inc., St. Louis, MO, USA) in the assay buffer, as described by Jefferson (1987). Blue color was observable after two hours of incubation. Samples were routinely incubated overnight to allow more intense color development. Samples were examined by eye, with a dissecting microscope (Wild 420, Wild Leitz, Heerbrugg, Switzerland) or with a light microscope (Zeiss Auxiophot, Carl Zeiss, Oberkochen, West Germany, or Leitz Laborlux 8, Wild Leitz, Heerbrugg, Switzerland). Images were recorded using a Kodak Digital Science DC120 digital camera (Eastman Kodak Company, Rochester, NY, USA). Digital images were stored using Kodak DC120 software in conjunction with Photoshop 5.0 (Adobe Systems Incorporated, San Jose, CA, USA). Fluorescent Microscopy Plants inoculated with bacteria harboring transposon TnP865-T were examined using an epifluorescent microscope (Zeiss 10 or Zeiss Axiophot, Carl Zeiss, Oberkochen, West Germany) equipped with a standard FITC filter set. Optical sections were made using laser scanning microscopy (Zeiss 10, Carl Zeiss, Oberkochen, West Germany or Meridian lnSIGHT, 230 Genomis Solutions, Lansing, MI USA). Optical sections were recorded using the software provided by the manufacture. Non-confocal fluorescent images were recorded as described above. Three dimensional reconstruction and analysis of optical sections were carried out using Voxel View software (Virtual Images, Fairfield, IA, USA) on a Silicon Graphics 4D-3O Personal IRIS Workstation. 231 RESULTS Re-isolation of Bacteria from Infected Rice Seedlings Re-isolation experiments were carried out to determine if bacteria previously isolated from surface-sterilized rice tissue (Chapter 2) were able to re-colonize rice seedlings. Fourteen bacterial isolates were selected for this analysis (see Chapter 2, Composition of a “Test Collection”). The results of the re-isolation experiments are shown in Figure 6-2. The number of bacteria isolated from different seedlings re- inoculated with the same isolate showed some variability. No bacteria could be re-isolated from seedlings inoculated with TY medium, E. coli, Azoarcus indigens, Herbaspirillum seropedicae or strain R90(8). In the case of all other bacterial strains, bacteria could be re-isolated from at least one test seedling. Tagging Bacteria with Molecular Biomarkers Multiple attempts were made to obtain green fluorescent derivatives of bacteria by introducing the transposons from pRL765gfp or anPSGS-T into 14 rice isolates and 9 previously described rice rhizosphere/endophytic reference strains. In selected cases E. coli cells, harboring the conjugation helper plasmid pRK2013, were added to the mating mixture to boost the transfer efficiency. Mating mixtures of each 232 strain were spread on plates containing different selective media (Material and Methods). Kanamycin resistant transconjugants were obtained in approximately 40% of the cases. Detectable green fluorescence was observed for only six of the kanamycin resistant transconjugants. Three transconjugant bacteria, Enterobacter cloacae, Klebsiella planticola, and R66(40), showed intense green fluorescence following the introduction of the transposon from pRL765gfp. However, the fluorescence was rapidly lost when the transconjugants were removed from kanamycin selection. Only three bacterial strains, 8. meliloti::gfp, R33::gfp and R100::gfp displayed stable, detectable green fluorescence levels following introduction of transposon TnPSGS-T (Figure 6-4). Transposon Tn121TH1 was introduced into S. meliloti, R133(120) and R100(64) to obtain S. meliloti::gus, R33::gus, and R100::gus marked derivatives, all of which were found to have B-glucuronidase (Gus) activity. 233 “7 In Situ Microscopic Evaluation of the Colonizing Ability of Biomarker Tagged Strains Rice seedlings grown using the pillow system and the inoculated with the tagged strains were examined for colonization. The root systems of rice seedlings inoculated with S. meliloti::gfp, R33::gfp, and R100::gfp where observed using fluorescence microscopy. The root system from twelve seedlings was examined for each strain. Likewise, the root systems of rice seedlings inoculated with S. meliloti::gus, R33::gus, and R100::gus were examined following incubation with X-gluc. The root systems from four rice seedlings were examined for each strain. In addition, the root systems from three inoculated rice seedlings grown in test tubes were examined for each strain tagged with the GUS biomarker. Following staining, colonization patterns were observed by eye and using light microscopy. When bacteria harboring the GUS biomarker were used to infect rice seedlings, blue staining, indicating the presence of large numbers of bacteria, could generally be observed without the aid of a microscope (Figure 6-6). The staining of plants grown in test tubes was more pronounced in all cases. The degree of staining was found to vary between seedlings inoculated with the same tagged bacterial isolate. 234 However, in general, no distinct differences in the degree and distribution of Gus staining was observed in rice seedlings inoculated with different tagged bacterial isolates. Gus staining was heaviest in primary and secondary roots, proximal to the seed. lnfrequently, staining of root tips was also observed (Figure 6-6 A). Seeds were darkly stained in most cases (Figure 6-6 B). These observations suggest that bacteria colonize the root system primarily on the upper portions of the primary and secondary roots. GFP mediated bacterial fluorescence could not be detected on rice roots using a UV transilluminator or a CCD camera equipped with proper filters for GFP excitation and emission detection. Fluorescence could only be detected using a 25x or greater immersion lens and an epifluorescent microscope. Therefore, the search for areas with a high degree of colonization required scanning the entire root system using an epifluorescent microscope, which proved to be a time consuming process. More fluorescent bacteria were observed on the upper portion of the primary and secondary roots of plants inoculated with strains expressing GFP, supporting the observations made using the tagged bacteria expressing GUS. 235 Colonization of rice roots inoculated with bacteria expressing GFP or GUS was further examined microscopically (Figure 6-7, 6-8, 6-9). Clumps of bacteria and individual bacteria cells, tagged with either Gus or GFP, could be observed on the root surface. Roots colonized by bacteria expressing GFP were examined using laser scanning confocal microscopy, which allows optical sectioning (Figure 6-8). A careful determination of spatial relationships between bacteria and the plant tissues was made using software that creates three dimensional images from optical sections (Figure 6-8, 6-9). Observations of the relationship between the plant and bacteria marked with GFP were facilitated by staining the plant with a fluorescent dye such as propidium iodide (Gage et al., 1996; Figure 6-7 E and F, 6-8, 6-9 E-H), which causes the edges of plant cells to emit red fluorescence when exited with blue light. No endophytic colonization by GFP tagged S. meliloti, R33(120) or R100(64) bacteria was observed. Bacteria were found to colonize the rice root surface in clumps, often at the base of root hairs (Figure 6-7 A and F, 6-9 C and D). Individual GFP tagged cells could also be observed to be associated with root hairs and along the junctions of epidermal cells (Figure 6-7 D and E). No qualitative differences in colonization efficiency 236 were observed between the there different bacterial strains. On rare occasions (four cells in the 21 root systems examined) an epidermal cell harboring a large number of R100::gfp bacteria was observed (Figure 6-7 B, 6-8, 6-9 E and F). 237 DISCUSSION Re-isolation of Bacteria from Infected Rice Seedlings isolation of bacteria from mechanically-abraded, surface-sterilized plant tissue has been proposed to lead to the isolation of bacteria which engage in intimate interactions with plants and, in particular, endophytic In- bacteria (See Chapter 1 for a definition). In order to examine this proposal, we selected a number of putative rice endophytes isolated using this approach and observed their ability to re-colonize rice tissues. i? Most of these putative endophytes could be re-isolated following inoculation and growth on gnotobiotic seedlings (Figure 6-2). However, no bacteria (C.F.U.’s) could be isolated from seedlings inoculated with control bacteria, E. coli, suggesting that bacteria not normally found to be endophytes of rice do not survive the surface sterilization protocol used. Interestingly, two putative rice endophytes, Herbaspirillum seropedicae and Azoarcus indigens could not be isolated after inoculation and surface sterilization. This suggests that these bacteria do not colonize 3 week old Lemont rice seedlings endophytically, or that the surface sterilization protocol used was too harsh. However, putative endophytic bacteria from our collection were able to survive this same surface sterilization 238 protocol, suggesting that these bacteria are better protected from the sterilizing agent than H. seropedicae or A. indigens. This protection may be the result of more aggressive endophytic colonization. However, a large variation was observed in the number of re-isolated bacteria from different seedlings inoculated with the same strain, making quantitative interpretation of the data impossible. F Generation of Sterile Rice Seedlings and Bacterial Infection for Microscopic Studies Seedlings grown in test tubes were smaller and less vigorous then those grown in pots, or in the ‘pillow system’. Colonization levels on rice seedlings grown in test tubes and inoculated with bacteria expressing the GUS biomarker were found to be much higher than colonization levels on seedlings grown in the pillow system. This suggests that inoculation experiments carried out in test tubes result in artificially elevated colonization levels due to the fact that the plant is unhealthy and more susceptible to bacterial infection. The pillow system (Szczyglowski et al. 1998) allows rice seedling to grow in a manner that more closely reflects that of plants grown in the field and, therefore, may be more suitable for colonization studies. 239 In Situ Microscopic studies on Colonizing Ability of the Tagged Stains A variety of molecular markers are now available that allow the detection and monitoring of bacteria in the environment (Akkermans et al., 1995; 1997; 1998; de Bruijn and Jansson, 1999). The GUS and GFP molecular markers were selected to facilitate the visualization of bacteria colonizing rice seedlings in our study. Both markers have been used successfully to study plant microbe interactions, for example during nodulation of legumes by rhizobia (Wilson et al., 1995; Gage et al., 1996). Our initial studies revealed that each marker has its advantages. For example, colonization of plant tissue by bacteria expressing Gus could easily be detected at the macroscopic level by an intense blue color (Figure 6-3). This allowed the examination of colonization patterns using entire infected root systems (Figure 6-5). However, it was difficult to determine three dimensional spatial location of Gus tagged bacteria in relation to the surface of the root at the microscopic level. In addition, diffusion of the cleaved GUS enzyme substrate prior to dimerization was found to cause the appearance of blue color in areas not directly colonized by bacteria. 240 Visualization of bacteria expressing GFP was found to overcome both these problems. Since bacteria expressing the gene encoding GFP could be detected using fluorescent microscopy (Figure 6-4), laser scanning confocal microscopy could be used to generate optical sections of infected tissue and to create images that depicted three dimensional spatial relationships. In addition GFP does not diffuse, rendering localization quite precise and reproducible. However, in our experience bacteria expressing GFP in root systems were detectable only with the aid of high power objectives in conjunction with epifluorescent or laser scanning microscopes. This feature rendered GFP less suitable than GUS as a biomarker for examining macroscopic colonization patterns. Problems with GFP as a Molecular Biomarker GFP has been used successfully as a biomarker/bioreporterin a wide variety of bacteria (Chapter 5). However, in our hands repeated attempts using several protocols and two different GFP transposons resulted in the tagging of only three bacteria that displayed stable, detectable levels of green fluorescence. There are several possible explanations for the failure to tag most of the bacterial isolates in the “test collection” with GFP. 241 First, it is possible that the gfp gene is not expressed at high enough levels to allow detection of the GFP protein. For example, in the case of anPS65-T it is possible that the Ptrp promoter is not active in some of the rice isolates. The Ptrp promoter is active in S. meliloti 1021 (Gage et al., 1996; Figure 6-4) However, E. coli cells harboring anP865-T are not fluorescent. It is possible that E. coli and other bacteria do not express genes controlled by this promoter properly, explaining why non- fluorescent kanamycin resistant transconjugants were found. Second, the gfp gene and the npt antibiotic resistance gene contained on pRL765gfp are both under control of the PpsbA promoter. Therefore kanamycin resistant transconjugants would be expected to also express GFP. However, none of the stable kanamycin resistant transconjugants obtained in this study displayed observable levels of green fluorescence, and, at best, showed faint fluorescence that rapidly faded. This would suggest that although the gfp gene is expressed at low levels, the fluorescence emitted from the GFP protein was inefficient. Alternatively, the GFP protein may not be processed correctly and/or sequestered in non-fluorescent inclusion bodies (Heim et al., 1994). 242 Difficulties in obtaining stable green fluorescent transconjugants may also be due to minor toxic effects produced by GFP. This hypothesis is supported by other evidence, including the difficulty constructing plasmids with the gfp gene under the control of strong constitutive promoters, especially using a medium containing high salt concentrations during the transformation process (Stoltzfus and de Bruijn, unpublished data; Herman Spaink, personal communication). Finally, it is also possible that the transposons were not properly integrating into the genomes of these bacteria. Plasmid anPSBS-T carries the origin of replication of R6K and therefore requires the TC protein for replication (de Lorenzo et al., 1990). This plasmid should not replicate in most bacteria. Plasmid pRL7659fp carries the pRK2 origin of replication, which replicates in E. coli and other Enterobacteriaceae. Three bacteria, Enterobacter cloacae, Klebsiella planticola and strain R66(40) showed intense green fluorescence following introduction of plasmid pRL765gfp. R66(40) is related to Enterobacteriaceae (Table 4- 2). However, when removed from kanamycin selection, the green fluorescence was quickly lost. The most logical explanation is that the pRL7659fp plasmid was replicating autonomously in these bacteria and 243 “F7 that the transposon never integrated into the genome. Because of the rapid loss of fluorescence in the absence of a selectable marker these bacteria could not be used for in-situ colonization studies. Colonization by S. meliloti 1021, R33(120) and R100(64) Derivatives of three types of bacteria, S. meliloti 1021, R33(120) and R100(64), were obtained that did express GFP and GUS. No observable differences in colonization patterns were observed between different transconjugants of the same isolates. Nor were major differences in colonization patterns observed when comparing the same isolate expressing GFP or GUS. Bloemberg et al. (1997) observed no difference in growth and motility of Pseudomonas strains harboring the gfp gene on a plasmid, suggesting that GFP activity did not constitute an excessive metabolic burden. Therefore, the integration of only two genes into the genome of these isolates would not be expected to add significantly to metabolic load and/or alter colonization patterns. Two to three weeks after inoculation of R33(120) or R100(64) to Lemont rice seedlings grown in the pillow system no endophytic colonization was observed. Colonization was limited to the surface and, on rare occasions, to isolated epidermal cells (Figure 6-7, 6-8, 6-9). This 244 does not mean that these bacteria are not endophytes. One possible explanation is that they were isolated from Philippino rice varieties at heading stage. It is, therefore, possible that endophytic colonization does not take place until later in rice plant development, It is also possible that Lemont is not amenable to colonization by these isolates or that colonization does not take place due to a shortcoming of the pillow system. For example, colonization may require the mechanical wounding that takes place when a root pushes its way through field soil. Some endophytes may also use fungi as aids for endophytic establishment (Paula et al., 1991). If colonization required the partnership between the tagged bacterial isolate and a fungi or other bacteria, colonization would not be expected to occur in an otherwise gnotobiotic pillow system. Despite the fact that strains R100(64) and R33(120) could be re- isolated following inoculation, endophytic colonization was not observed using variants of these strains harboring the Gus or GFP biomarkers. These re-isolation studies were carried out on plants grown in the test tube system while observations using biomarkers were carried out using plants grown in the pillow system. Therefore, the re-isolation of bacteria experiment can not be directly correlated to the biomarker experiments. 245 Spatial Localization Using Three Dimensional Reconstructions Bacteria expressing GFP could be detected using the laser scanning confocal microscope. This also allowed optical sectioning (Figure 6-8) and three dimensional reconstruction of the images (Figure 6-9). Without endogenous fluorescence of rice tissue/cells it was not possible to gain information on the spatial relationship between the bacteria and the plant tissues (Figure 6-9 A and B). In selected cases, the autofluorescence of plant tissue could be used to orient the viewer (Figure 6-9 C and D). However, in our experience, the yellow fluorescence from rice roots was quite variable and unsuitable for reliable three dimensional reconstruction. Propidium iodide can be used to fluorescently stain plant tissues (Gage et al., 1996). With the proper instrumentation, this adaptation of the protocol allowed real color optical sectioning and three dimensional reconstruction of the images (Figure 6-8, 6-9 E and F). The fluorescence observed using this stain was very bright and tended to mask weak GFP signals, even after short staining times and long periods of destaining. This may not be a problem for digital imaging, since the relative intensities of the different fluorescence could be altered. However, it created problems for scanning infected tissues using the 246 oculars of a standard epifluorescent microscope. In addition the staining of the plant tissue was variable, with intense staining in some areas and virtually no staining in others. Therefore, development of better stains or staining techniques would greatly enhance the use of GFP as a biomarker for bacterial colonization of plant tissues. 247 CONCLUSIONS GFP and GUS are promising tools for visualizing endophytic colonization of rice and other plants. However, better vectors with suitable promoters and new versions of GFP need to be developed for this biomarker to reach its full potential. GFP is useful for observing microscopic colonization by individual bacterial cell, or clumps of cells, and can be used in conjunction with laser scanning confocal microscopy to determine spatial relationships between bacteria and plant tissues. GUS is more suited for visualization of macroscopic colonization patterns. The use GFP and GUS as biomarkers revealed the absence of a significant level of endophytic colonization of rice seedlings by two isolates, R100(64) and R33(120). However, due to the limited scope of this study, definitive conclusions on the endophytic nature of these bacteria can not be made. A more detailed investigation including the putative endophytes described at the end of Chapter 4 needs to be carried out. This investigation should involve the study of endophytic colonization using different rice varieties, specifically Philippino varieties from which these bacteria were originally isolated. It should also examine older plants, as endophytic colonization may not occur in young seedlings. 248 Table 6-1 Strain Relevant Characteristics Reference E. coli DHSa supE44 A lacU (¢8OIac ZAMI 5) Sambrook et al., hdei17 recA1 endA1gyr A96 thi-1 1989 re/A1 E. coli H81 01 supE44 hst20(rB‘mB') recA13 ara- 14 proA2, Iach galK2 rpsL20 xyI-5 mtl-1 Sambrook et al., 1989 E. coli S17-1 Apir thi pro hst hde’ recA RP4 2- TC::Mu-Km::Tn7(Tp’/Sm') A-pir lysogen Victor de Lorenzo, Centro de Investigaciones Biologicas, Madrid Klebsiella Isolated from the rice rhizosphere Ladha et al., 1983 planticola KpM5A1 . Enterobacter Isolated from the rice rhizosphere Ladha et al., 1983 cloaca Azoarcus indigens Endophyte of Kallar Grass Reinhold-Hurek et LMG 9098 al., 1993 Herbaspirillum Endophyte of sugarcane Baldini et al., 1986 seropedicae 26 7 Sinorhizobium Symbiont of Medicago sativa Meade et al., 1982 meliloti 1021 Azorhizobium Symbiont of Sesbania rostrata Dreyfus and caulinodans Dommergues, 1981 ORSS71 T105 Isolated from surface-sterilized rice Stoltzfus et al., 1997 R22(88) Isolated from surface-sterilized rice Stoltzfus et al., 1997 R90(8) Isolated from surface-sterilized rice Stoltzfus et al., 1997 R100(64) Isolated from surface-sterilized rice Stoltzfus et al., 1997 -R6a(126) Isolated from surface-sterilized rice Stoltzfus et al., 1997 R33(120) Isolated from surface-sterilized rice Stoltzfus et al., 1997 R45(42) Isolated from surface-sterilized rice Stoltzfus et al., 1997 R032.S2.3 Isolated from surface-sterilized rice Stoltzfus et al., with an in-planta selection step 1997 R061.S1.3 Isolated from surface-sterilized rice Stoltzfus et al., with an in-planta selection step 1997 R66(40) Isolated from surface-sterilized rice Chapter 3, this work 249 Table 6-1 continued Strain Relevant Characteristics Reference Sinorhizobium Sinorhizobium meliloti 1021 This work meliloti::gfp derivative containing anSBS-T Sinorhizobium Sinorhizobium meliloti 1021 This work meliloti::gus derivative containing an121TH1 R33::gfp R33(120) derivative containing This work an865-T R33::gus R33(120) derivative containing This work an121TH1 R100::gfp R100(64) derivative containing This work anSGS-T R100::gus R100(64) derivative containing This work an121TH1 Table 6-1 Bacterial Strains List of the strains, relevant characteristics and references for the bacteria used in this study. 250 Name Relevant Characteristics Reference/ Source pCAM121 Sm/Sp, Ap;an5SSgusA21(Paph- Wilson et al., gusA-trpA terminator translational 1995 fusion with adjacent unique Spel site) in pUT/mini-Tn5 Sm/Sp pTB93F Cm, Sp; Broad host range vector Gage et al., containing Ptrp-gfpSBS-T 1996 pBluescript Ap; Unique cloning sites Stratagene, La Jolla, CA, USA pUC 4K Km, Ap;Tn903 aminoglycoside Amersham 3'phosphotransferase gene in Pharmacia pUC 4 Biotech, Inc., Piscataway, NJ, USA pUT/mini- Km, Ap;an5 (Kanamycin cassette de Lorenzo et al. Tn5Km with uniqe Noll site) 1990 anPSGS-T Km, Ap; an5P865-T (Ptrp- This work gfpS65-T from pTB93F) in pUT/mini -Tn5 Km p121TH1 Km, Ap; This work an121TH1(an5SSgusA2With the Sm/Sp cassette exchanged with the Km cassette from pUC 4K) pRL7659fp Km; pRL765zgfp oriT(RK2) Tombolini et al., oirV(p15A) 1997 pRK2013 Km; ColE1 orIRK2-mob“ RK2tra” Figurski and Helinski, 1979 Table 6-2 Plasmids List of plasmids, relevant information and references for the constructs used in this study. 251 ‘1 x1 :60 2o; 30:3 mL Emu 0* 3533-4. Immammzmzmmxmoxmwmawmm 38 0: max 922:8 ow E833 0.2.33-4. nosmflqcoga Sq Sm Qm_. wmmNNN—LAA I 22222222: EC ”aaaaaaaa ..EJ' CDCDCDCDQCDOCDCD I CDN-‘ODN-‘CDM—L S.m.t’~l u A.C.u Ital i> Ali! Hai T105”EB R22(88)4 ‘3‘ u r! R90(8) ' R100(64) it" "> '3’ -R6a(126) 9"” R33(120) R45(42) R032.S2.3 R061.S1.3 >- locum a.» 39.3.9.9. 9 $29.» .33 29.59. 3.8 98.5% ..:.m 9%: 952m :5 .5309 o. 399$ $299. .33 9.8 mmma_.:om 500599. .2..: 99m 95:6 0. 3m 8295 .3899. o: Sm x-mx.m. .992: .9 38m Emmxm .: $9.6 5m. Ecmm. ..:Bm 39:59 <<9m 300599. .2..: moo: 95.3. 3:92.30 29.96 £9._.Nm..o:. woo: $92.8 £3 38993 95 a._c..o:m 0. 3m 3mo9m$ 29m 29me 0: 56m ._.< 298. wmo.9.m 29m 8:39. man :6 59:5 0. 358: 8623 2mm <9...ma swim Sumo.» @9538 2899.38. ..c:.:. .-w: 593mg. 38.3% 500598 .2..: 99.5 ._.< 39.5. 254 Figure 6-3 Bacteria Marked with 35 Bacteria grown on a TY plate supplemented with 50 ug/ml X-gluc. ; The bacteria were tagged with the transposon Tn121TH1. Two of the strains, R33(120) (top) and R100(64) (middle), were isolated from surface- sterilized rice. The third strain, , _ Sinorhizobium meliloti 1021(bottom), was used as a control. Parental bacteria are shown on the far left (no color). Three transconjugants for each isolate shown on the right produced a dark blue color, indicative of GUS activity. 511m Figure 6-4 Bacteria Marked with (FF Fluorescent micrographs of three bacterial strains expressing GFP are shown. These strains were used for visualization of colonization of rice seedlings. (A)Sinorhizobium meliloti::gfp, used as a control. (B) R33::gfp and (C) R100::gfp, isolated from surface-sterilized rice. The bacteria were tagged with the transposon TnPS65-T. Bacteria were grown in liquid TY media and placed on glass slides in 0.8% agarose. The micrographs were taken using a Kodak Digital Science DC120 digital camera attached to a Zeiss Axiophot microscope equipped with a standard FITC filter set 255 Figure 6-5 Growth of Rice Seedlings in the Pillow System Pictures of two week old rice seedlings grown in the ‘pillow system’ are shown. (A) Seedlings grow sandwiched between two pillows. (B) Picture showing the root systems of seedlings associated with one of the two pillows. 256 . g, ’ t,-=’ . vi 9 . 4:! '0 I «l .. 3 '. Figure 6-6 Gus Staining of Rice Plants Pictures of rice seedlings inoculated with bacteria expressing the GUS biomarker incubated in buffer containing X-Gluc are shown. The blue color is indicative of bacterial Gus activity. (A) Colonization patterns of R100::gus bacteria on two week old rice roots. (B) Seed colonization pattern of plants inoculated with R33::gus bacteria. 257 20pm __ 10mm 20mm.— Figure 6-7 Microscopic Observation of Colonization Micrographs of plants inoculated with bacteria expressing the biomarker GUS (A and B) or GFP (C-F) are shown. Transmitted micrographs take using a Leitz microscope of (A) R100::gus bacteria (blue color) in a clump on the surface of a rice root and (B) colonizing an epidermal cell. Overlay of a transmitted and a fluorescent micrograph of (C) S. meliloti::gfp bacteria (green color) in a clump on the surface of a rice root and (D) individual bacterial cells at the junction between epidermal cells. Fluorescent micrographs of rice roots stained with propidium iodide (red color) and colonized by (E) R33::gfp bacteria (green color) at the junctions of epidermal cells and (F) R100::gfp bacteria (green color) in clumps on the surface of the root. --- Figure 6-8 Optical Sections of a Rice Cell Colonized by R100(64) Expressing GP Optical sections of an epidermal cell colonized by R100::gfp bacteria (green color). Each section is 0.3 pm deeper into the sample. The rice root tissue was stained with propidium iodide (red color). Section one starts at the surface of the rice root and section 80 is 24pm below the surface of the root. (A-E) Individual optical sections 1 through 5. (F) Overlay of sections 1-5. (G-N) Overlays containing ten optical sections, starting with sections 1 through 10 and ending with sections 71 through 80. (O) Overlay of all 80 optical sections. (P-T) Individual optical sections 76 through 80. (U) Overlay of optical sections 76 through 80. 259 .R I. '1 '! e v I . V i 7". ioc ba 3[ lhl 3? iodide stained root tissue. Figure 6-9 Three Dimensional Reconstruction of Optical Sections Three dimensional computer reconstructions of rice root colonization by bacteria expressing the GFP biomarker. These renderings are constructed from optical sections made using a laser scanning confocal microscope. (A & B) Two views of an epidermal root cell and attached root hair filled with R100::gfp bacteria (green color). (C & D) Two views of R33::gfp bacteria (green color) colonizing the surface of a rice root (red color). Artificial colorization is based on fluorescent intensity. A low level of yellow autofluorescence from the root tissue is depicted in red. The intense green fluorescence from the tagged bacteria is depicted in green. (E-H) Images created using dual channel detection of R100::gfp bacteria colonizing an epidermal cell on a rice root. The red fluorescence is from propidium The green fluorescence is from R100::gfp bacteria. (E) Phi 2 sections along the X and Y axis. (F) Oblique view of the 3D reconstruction showing only the root tissue (red channel), (G) only the bacteria (green channel), and (H) the root tissue and bacteria (red and green channels) combined. 260 REFERENCES Akkermans ADL, van Elsas JD and de Bruijn FJ, eds, Molecular Microbial Ecology Manual. Kluwer Academic Publishers. Dordrecht, 1995, 1996, 1998. Atlas FM (1997) Handbook of Microbial Media. 2nd ed, CRC Press, inc. Boca Raton, Florida, pp 401 Baldani Jl, Baldani VLD, Seldin L, Dobereiner J (1986) Characterization of Herbaspirillum seropedicae gen. nov., sp. nov., a root associated nitrogen-fixing bacterium. Int J Syst Bacteriol 36: 86-93 Barraquio WL, Revilla L, Ladha .K (1997) Isolation of endophytic diazotrophic bacteria from wetland rice. Plant and Soil 194: 15-24 Beringer JE (1974) R factor transfer in Rhizobium leguminosarum. J Gen Microbiol 84: 188-189 Bloemberg GV, O’Toole GA, Lugtenberg BJJ, Kolter R (1997) Green fluorescent protein as a marker for Pseudomonas spp. Appl Environ Microbiol 63: 4543-4551. Dreyfus BL, Dommergues YR (1981) Nitrogen-fixing nodules induced by Rhizobium on the stem of the tropical legume Sesbania rostrata . FEMS Microbial Letters. 1 0: 313-317 Elmerich C, Dreyfus BL, Reysset G, Aubert ..P (1982) Genetic analysis of nitrogen fixation in a tropical fast-growing Rhizobium. EMBO 1: 499-503 Fahraeus A (1957) The infection of clover root hairs by nodule bacteria studied by a simple glass slide technique. J Gen Microbiol 16: 374-381 Figurski D, Helinski D (1979) Replication of an origin-containing derivative of plasmid RK2 dependent on a plasmid function provided in trans. Proc Natl Acad Sci USA 76:1648-1662 261 Gage DJ, Bobo T, Long SR (1996) Use of green fluorescent protein to visualize the early events of symbiosis between Rhizobium meliloti and alfalfa (Medicago sativa). J Bacteriol 178: 7159-7166 Heim R, Prasher DC, Tsien RY (1994) Wavelength mutations and posttranslational autoxidation of the green fluorescent protein. Proc Natl Acad Sci 91:12501-12504 Herrero M, de Lorenzo V, Timmis KN (1990) Transposon vectors containing non-antibiotic resistance selection markers for cloning and stable chromosomal insertion of foreign genes in gram-negative bacteria. J Bacteriol 172: 6557-6567 Hurek T, Reinhold-Hurek B, van Montagu M, Kellenberger E (1994) Root Colonization and systemic spreading of Azoarcus sp strain BH72 in grasses. J Bacteriol 176: 1913-1923 Jansson JK, de Bruijn FJ (1999) Biomarkers and Bioreporters to track microbes and monitor their gene expression. In AL Demain, JE Davies, eds, Manual of Industrial Microbiology and Biotechnology. 2nd ed. ASM Press, Washington, DC pp 651-665 Jefferson RA (1987) Assaying chimeric genes in plants, the gene fusion system. Plant Molecular Biology Reporter. 5: 387-405 Kiss GB, Vincze E, Kalman z, Forrai T, Kondorosi (1979) Genetic and biochemical analysis of mutants affected in nitrate reduction in Rhizobium meliloti. Journal of General Microbiol 1 13: 105-1 18 Ladha JK, Barraquio WL, Watanabe l (1983) Isolation and identification of nitrogen-fixing Enterobacter cloacae and Klebsiella planticola associated with rice plants. Can J Microbiol 29: 1301-1308 de Lorenzo V, Herrero M, Jakubzik U, Timmis KN (1990) Mini- Tn5 transposon derivatives for insertion mutagenesis, promoter probing, and chromosomal insertion of cloned DNA in gram-negative eubacteria. Journal of Bacteriology 172: 6568-6572 262 Maniatis T, Fritsch EF, Sambrook J (1982) Molecular cloning: A laboratory manual. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY Meade HM, Long, SR, Ruvkun GB, Brown SE, Ausubel FM (1982) Physical and genetic characterization of symbiotic and auxotrophic mutants of Rhizobium meliloti induced by transposon Tn5 mutagenesis. Journal of Bacteriology 149: 114-122 Paula MA, Reis VM, Dobereiner J (1991) Interactions of Glomus clarum with Acetobacter diazotrophicus in infections of sweet potato (Ipomoeabatatas), sugarcane (Saccharum spp.), and sweet sorghum (Sorghum vulgare). Biology and Fertility of Soils 1 1: 111-115 Reddy PM, Ladha JK, So RB, Hernandez RJ, Ramos MC, Angeles OR, Dazzo FB, de Bruijn FJ (1997) Rhizobial communication with rice roots: Induction of phenotypic changes, mode of invasion and extent of colonization. Plant and Soil 194: 81-89 Reinhold-Hurek B, Hurek T, Gillis M, Hoste B, Vancanneyt M, Kersters K, De Ley J (1993) Azoarcus gen. nov., nitrogen-fixing proteobacteria associated with roots of Kallar grass (Leptochloa-fusca (L.) Kunth) and description of two species Azoarcus indigens sp. nov. and Azoarcus communis sp. nov. Int J Syst Bacteriol 43: 574-584 Sambrook J, Fritsch EF, Maniatis T, eds, (1989) Molecular Cloning: A Laboratory Manual 2nd ed. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory Press Silhavy TJ, Berman ML, Enquist LW (1984) Experiments with gene fusions. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory Press Stoltzfus JR, So R, Malarvithi PP, Ladha JK, de Bruijn FJ (1997) Isolation of endophytic bacteria from rice and assessment of their potential for supplying rice with biologically fixed nitrogen. Plant and Soil 194: 25-36 263 Szczyglowski K, Shaw RS, Wopereis J, Copeland S, Hamburger D, Kasiborski B, Dazzo FB, de Bruijn FJ (1998) Nodule organogenesis and symbiotic mutants of the model legume Lotus japonicus. MPMI 11: 684—697 Tombolini R, Unge A, Davey ME, de Bruijn FJ, Jansson .K (1997) Flow cytometric and microscopic analysis of GFP-tagged Pseudomonas fluorescens bacteria. FEMS Microbiol Ecol 22: 17-28 Unge A, Tombolini R, Davey ME, de Bruijn FJ, Jansson J( (1998) GFP as a Marker Gene. In ADL Akkermans, JD van Elsas, FJ de Bruijn, eds, Molecular Microbial Ecology Manual. Kluwer Academic Publishers, Dordrecht, pp 6.1.1321-16 Vermeiren H, Vanderleyden J, Hai W (1998) Colonization and nifH expression on rice roots by Alcaligenes faecalis A15. In KA Malik, MS Mirza, JK Ladha, eds, Nitrogen Fixation with Non-Legumes. Kluwer Academic Publishers, Dordrecht, pp 167-177 Webster G, Gough C, Vasse J, Batchelor C A, O’Callaghan K J, Kothari S L, Davey MR, Dénarié J, Cocking EC (1997) Interactions of rhizobia with rice and wheat. Plant and Soil 194: 1 15-122 Wilson KJ (1995) Molecular techniques for the study of rhizobial ecology in the field. Soil Biol Biochem 27: 501-514 Wilson K, Sessitsch A, Corbo J C, Giller K E, Akkermans A D L, Jefferson R A (1995) B-Glucuronidase (GUS) transposons for ecological and genetic studies of rhizobia and other Gram-negative bacteria. Microbiol 141: 1691-1705 Yanni YG, Rizk RY, Corich V, Squartini A, Ninke K, Philip- Hollingsworth S, Orgambide G, de Bruijn F, Stoltzfus J, Buckley D, Schmidt TM, Mateos PF, Ladha JK, Dazzo FB (1997) Natural endophytic association between Rhizobium leguminosarum bv. trifolii and rice roots and assessment of its potential to promote rice growth. Plant and Soil 194: 99-114 You C, Zhou F (1988) Non-nodular endophytic nitrogen fixation in wetland rice. Can J Microbiol 35: 403-408 264 CHAPTER 7 SUMMARY AND CONCLUSIONS DIVERSITY OF RICE ISOLATES Bacteria isolated from mechanically-abraded, surface-sterilized rice tissue are genotypically diverse. This may indicate that many types of bacteria can colonize rice plants endophytically, or it may indicate that the surface sterilization used was inadequate to eliminate rhizoplane and rhizosphere bacteria. Bacteria which could be isolated following inoculation to sterile seedlings did not show endophytic colonization when observed using biomarkers. Without further observation it is impossible to determine the endophytic nature of the isolates in this collection . Many of the bacteria isolated have similar ARDRA fingerprints or SSU rRNA gene sequences to bacteria commonly isolated from the rhizosphere of rice. Some diazotrophic bacteria in this collection have not been commonly isolated from the rhizosphere of rice but show similarities to other putative rice endophytes. These bacteria may represent obligate endophytes and deserve further study as candidates for supplying rice ‘with biologically-fixed nitrogen. 265 UTILITY OF nifD-PCR The nifD-PCR primers presented here are capable of amplifying a specific fragment from most diazotrophic bacteria. These primers identified 20 bacteria from a collection of putative rice endophytes, 19 of which were shown to be authentic diazotrophs. These primers are suitable for identifying nitrogen-fixing bacteria in collections of environmental isolates and for detection of diazotrophs in the environment. UTILITY OF GUS AND GFP AS COLONIZATION BIOMARKERS GUS and GFP are both useful for localizing tagged bacteria following inoculation to plant tissues. GUS is better suited for determining overall colonization patterns. GFP in conjunction with optical sectioning and three dimensional reconstruction allows visualization of spatial relationships between the bacteria and plant tissue. COLONIZATION OF SEEDLINGS BY RICE ISOLATES Bacteria isolated from mechanically-abraded, surface-sterilized rice can be isolated following inoculation to sterile seedlings grown in test tubes. However, since plants grown in test tubes appear less vigorous than seedlings grown in pots the significance of this is unclear. Biomarkers were used to examine the colonization patterns for two of these bacteria, 266 R33(120) and R100(64). No endophytic colonization was found. R100(64) was, on very rare occasions, found to colonize individual epidermal cells. 267 APPENDIX I Strain Soil Type Variety Isolation Medium -R11b(59) Bicol Oryza minuta Culm Extract & Water Agar1 -R11c(59) Bicol Oryza minuta Culm Extract & Water Agar1 -R136(33) Bicol Oj/za minuta Culm Extract & Water Agar1 -R139(43) Bicol Oryza minuta Culm Extract & Water Agar1 -R14(5) Bicol Oryza minuta Culm Extract & Water Agar1 -R141 (99) Bicol Oryza minuta Culm Extract & Water Agar1 -R143(44) Pangasinan Pinidua Culm Extract & Water Agar1 -R144(100) Bicol Oryza minuta Culm Extract & Water Agar1 -R144(30) Bicol Oryza minuta Culm Extract & Water Agar1 -R145(69) Bicol Oryza minuta Culm Extract & Water Agar1 -R146(23) Pangasinan Pinidua Culm Extract & Water Agar1 -R16(47) Bicol Oryza minuta Culm Extract & Water Agar1 -R3 Bicol Oryza minuta Culm Extract & Water Agar1 -R38(26) Pangasinan Pinidua Culm Extract 8. Water Agar1 -R6a(126) Bulacan Bomalasang Culm Extract & Water Agar1 K10 Banaue OkingSeroni Congo Red2 K100 Pangasinan Pinidua Congo Red2 K1 O3 Banaue Oryza minuta Congo Red2 K107 Banaue Oryza minuta Congo Red2 K108 Banaue Oryza minuta Congo Red2 K109 Bulacan Oryza minuta Congo Red2 K1 1 Banaue Oking Seroni Congo Red2 K1 10 Bulacan Oryza minuta Congo Red? K1 13 Bulacan Oryza minuta Congo Red2 K3 Banaue Okim Seroni Congo Red2 K34a Banaue Bomalasang Congo Fied2 K36a Banaue Bomalasang Congo Red2 K360 Banaue Bomalasang Congo Red2 ' K37 Banaue Bomalasang Congo Red? 268 Appendix I continued K45 Bulacan Bomalasang Congo Red2 K46 Bulacan Bomalasang Congo Red2 K5 Banaue Oking Seroni Congo Red2 K6 Banaue Oking Seroni Congo Red2 K71 Pangasinan Pinidua Congo Red? K8 Banaue Oking Seroni Congo Red2 K9 Banaue Oking Seroni Congo Red2 K93 Pangasinan Pinikitan Congo Red:2 K94 Pangasinan Pinikitan Congo Red2 K95 Pangasinan Pinikitan Congo l=ied2 K96 Pangasinan Pinikitan Congo Red2 L 2 1 Banaue Bomalasang LGIP3 R100(64) Maahas Pinikitan Malate Semi Solid Media4 R103(73) Maahas Pinikitan Malate Semi Solid Media4 R103(96) Maahas Pinikitan Malate Semi Solid Media4 R104(113) Maahas Pinikitan Malate Semi Solid Media4 R104(75) Maahas Pinikitan Malate Semi Solid Media4 R1 Oa(32) Bicol Oryza minuta Malate Semi Solid Media4 R1 Oa(86) Bicol Oryza minuta Malate Semi Solid Media4 R11b(103) Bicol Oryza minuta Malate Semi Solid Media4 R1 10(74) Bicol OrLza minuta Malate Semi Solid Media4 R19(122) Pangasinan Pinidua Malate Semi Solid Media4 R19(77) Pangasinan Pinidua Malate Semi Solid Media4 R1 a(35) Bulacan Bomalasang Malate Semi Solid Media4 R22(88) Pangasinan Pinidua Malate Semi Solid Media4 R29(20) Bicol Pinikitan Malate Semi Solid Media4 R29(91) Bicol Pinikitan Malate Semi Solid Media4 R32(121) Bicol Pinidua Malate Semi Solid Media4 R32(95) Bicol Pinidua Malate Semi Solid Media4 R33(120) Banaue Pinidua Malate Semi Solid Media4 R33(94) Banaue Pinidua Malate Semi Solid Media4 269 Appendix I continued R39(129) Bulacan IR42 Malate Semi Solid Media4 R40(39) Maahas Pinidua Malate Semi Solid Media4 R41 (1 e) Banaue Pinikitan Malate Semi Solid Media4 R42(21) Bulacan |R42 Malate Semi Solid Media4 R43(41) Bulacan IR42 Malate Semi Solid Media4 R44(51) Bulacan |R42 Malate Semi Solid Media4 R45(42) Bulacan |R42 Malate Semi Solid Media4 R4649) Bicol Pinidua Malate Semi Solid Media4 R48b Pangasinan Pinikitan Malate Semi Solid Media4 R49(56) Pangasinan Pinikitan Malate Semi Solid Media4 R4a(1 18) Bulacan Bomalasang Malate Semi Solid Media4 R54(10) Pangasinan OkingSeroni Malate Semi Solid Media4 R55(112) Pangasinan Oking Seroni Malate Semi Solid Media4 R58(76) Pangasinan Oking Seroni Malate Semi Solid Media4 R58b(102) Pangasinan Oking Seroni Malate Semi Solid Media4 R59(24) Pangasinan Oking Seroni Malate Semi Solid Media4 R59b(13) Pangasinan Okim Seroni Malate Semi Solid Media4 R61 (12) Banaue lR74 Malate Semi Solid Media4 R62(15) Banaue IR74 Malate Semi Solid Media4 R62b(117) Banaue IR74 Malate Semi Solid Media4 R63b(124) Banaue IR74 Malate Semi Solid Media4 R65(49) Banaue IR74 Malate Semi Solid Media4 R66(40) Banaue IR74 Malate Semi Solid Media4 R68(101) Banaue IR74 Malate Semi Solid Media4 R69(72) Maahas Oking Seroni Malate Semi Solid Media4 R70b Maahas OkilchSeroni Malate Semi Solid Media4 R70b(37L Maahas Oking Seroni Malate Semi Solid Media4 R70b(68) Maahas Oking Seroni Malate Semi Solid Media4 R72(104) Maahas Oking Seroni Malate Semi Solid Media4 R73(4) Maahas Oking Seroni Malate Semi Solid Media4 , R73(6) Maahas Oking seroni Malate Semi Solid Media4 270 Appendix I continued R73c(36) Maahas Oking Seroni Malate Semi Solid Media4 R75(50) Maahas OkingSeroni Malate Semi Solid Media4 R76b(27L Maahas Pinikitan Malate Semi Solid Media4 R77(132) Maahas Pinikitan Malate Semi Solid Media4 R77(18) Maahas Pinikitan Malate Semi Solid Media4 R78453 Maahas Pinikitan Malate Semi Solid Media4 R79Q08) Maahas Pinikitan Malate Semi Solid Media4 R79(79) Maahas Pinikitan Malate Semi Solid Media4 R8428) Bulacan Bomalasang Malate Semi Solid Media4 R8(54) Bulacan Bomalasang Malate Semi Solid Media4 R81(1) Maahas Pinikitan Malate Semi Solid Media4 R81 (90) Maahas Oking Seroni Malate Semi Solid Media4 R82(128) Maahas Oking Seroni Malate Semi Solid Media4 R83(71) Maahas Oking Seroni Malate Semi Solid Media4 R84(22) Maahas Oking Seroni Malate Semi Solid Media4 R85(92) Maahas Oking Seroni Malate Semi Solid Media4 R85b(130) Banaue IR74 Malate Semi Solid Media4 R87b(110) Maahas Oking Seroni Malate Semi Solid Media4 R87c(11 1) Maahas OkinlSeroni Malate Semi Solid Media4 R88 Banaue lR74 Malate Semi Solid Media4 R89(66) Maahas Oking Seroni Malate Semi Solid Media4 R89(87) Maahas Oking Seroni Malate Semi Solid Media4 R90(8) Maahas Oking Seroni Malate Semi Solid Media4 R92(60) Maahas Oking Seroni Malate Semi Solid Media4 R92b(34) Maahas Oking Seroni Malate Semi Solid Media4 R93(58) Maahas Oking Seroni Malate Semi Solid Media4 R93b Maahas Oking Seroni Malate Semi Solid Media4 R94(17) Pangasinan Oking Seroni Malate Semi Solid Media4 R94b Maahas Oking Seroni Malate Semi Solid Media4 R95(7) Maahas Oking Seroni Malate Semi Solid Media4 . R96 (1 1 5) Pangasinan Oking Seroni Malate Semi Solid Media4 271 It; . Appendix I continued R97(106) Pangasinan Oking Seroni Malate Semi Solid Media4 R97b((34) Pangasinan Oking Seroni Malate Semi Solid Media4 R98(2) Pangasinan Oking Seroni Malate Semi Solid Media4 R98(46) Pangasinan Oking Seroni Malate Semi Solid Media4 R99(65) Maahas Pinikitan Malate Semi Solid Media4 S2 Bicol Oryza minuta Malate Semi Solid Media4 S489 Maahas IR74 Malate Semi Solid Media4 T10 Banaue Oking Seroni 0.1% TSA4 T105 Banaue Oking minuta 0.1% TSA4 T116 Bicol Oryza minuta 0.1% TSA4 T12 Banaue Oking Seroni 0.1 % TSA4 T14 Banaue Oking Seroni 0.1% TSA4 T60 Banaue Pinidua 0.1% TSA4 T62 Pangasinan Pinidua 0.1% TSA4 T77 Banaue Pinikitan 0.1% TSA4 T90 Pangasinan Pinikitan 0.1% TSA4 T93 Pangasinan Pinikitan 0.1% TSA4 T94 Pangasinan Pinikitan 0.1% TSA4 T96a Pangasinan Pinikitan 0.1% TSA4 T96b Pangasinan Pinikitan 0.1% TSA4 272 1 Barraquio WL, Revilla L, Ladha JK (1997) Isolation of endophytic diazotrophic bacteria from wetland rice. Plant and Soil 194: 15-24 2 Congro Red medium (g/L) K2HPO4 0.5 MgSO4'7H20 0.2 NaCl 0.1 Yeast Extract 0.5 FeCI3'6H20 0.015 DL-Malic acid 5.0 KOH 4.8 H20 1000 ml Adjusted to pH 7.0 with 0.1N KOH Bacto-Agar 20.00 1:400 ml aqueous solution of congo red 15.0 ml/L was added to media aseptically before plating. (Congro red solution and basal medium were autoclave separately) 3 Reis VM, Olivares FL, Dobereiner J (1994) Improved methodology for isolation of Acetobacter diazotrophicus and confirmation of its endophytic habitat. Wold J Microbiol Biotechnol 10: 401-405 4 0.1% TSA medium (g/l) Tryptic Soy Agar 1.0 H20 1000 ml Adjusted to pH 7.0 with 0.1N KOH Bacto-Agar 20.00 273