“sham..." .s. . . . v....:. Ty.» if: . it: .c ill: :2 1:4 .lofi . i L L I :13: ...T. 1.5.. ; I .5. i 1.-4J0:III I. '4', .Drvfiofnl. : v4lv4§ ’01.. . iv v. as. . ‘aLv I. 1...?! I... r. r ‘ . 1 . . . “5.5.5.5. «:5. . r 1. ..,.x!. 4... .22... . . “HES"! \\\\\\\\\\\\\\\\\\\\\3\\\\ \l‘illlllllll 3 1293 00793 573 This is to certify that the dissertation entitled A MOLECULAR APPROACH TO UNDERSTANDING THE GENETIC RESPONSE OF 2,4-D DEGRADING MICROBIAL POPULATIONS IN SOIL UNDER SELECTION presented by J ONG-OK KA has been accepted towards fulfillment of the requirements for _Bh..JL_ degree inW Major ofessor Date NW , [(fl'r Mfiz/ MS U i: an Affirmative Action/Equal Opportunity Institution 0- 12771 I w LIBRARY Michigan State University I i ll PLACE IN RETURN BOX to remove this checkout from your record. TO AVOID FINES return on or before date due. DATE DUE DATE DUE DATE DUE ¥ Li: H WW I i f». I“), 4‘ (- MSU Is An Affirmative ActIoNEquaI Opportunity Institution cmmma-pJ A MOLECULAR APPROACH TO UNDERSTANDING THE GENETIC RESPONSE OF 2,4-D DEGRADING MICROBIAL POPULATIONS IN SOIL UNDER SELECTION By J ong-Ok Ka A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Microbiology and Public Health 1992 ABSTRACT A MOLECULAR APPROACH TO UNDERSTANDING THE GENETIC RESPONSE OF 2,4-D DEGRADING MICROBIAL POPULATIONS IN SOIL UNDER SELECTION By J ong-Ok Ka The diversity of 2,4-dichlorophenoxyacetic acid (2,4-D) degrading bacteria, their population ecology and genetic changes under selective pressure were analyzed at the DNA level. A total of 47 predominant 2,4-D degrading bacteria were isolated from Kellogg Biological Station (KBS) soils. The isolates were diverse in fatty acid methyl ester profiles, REP fingerprint patterns, substrate utilization patterns and hybridization patterns using DNA probes. The tfdA gene and a 6.5 kb probe (cloned from a 2,4-D+ Pseudomonas paucimobilis) hybridized to the DNA of 7 0 % of the isolates which predominated in soil repeatedly treated with 2,4-D. These two DNA probes were useful in studying the population dynamics and genetic alterations in dominant 2,4-D degrading bacteria in soil under 2,4—D selection. Bacterial populations with homology to one or the other of these probes rapidly increased in soil treated with 2,4-D and changes in banding patterns (the position of hybridization signal) were observed in Southern blot analyses with total soil bacterial DNA. These data suggested that population changes or possibly genetic rearrangement in 2,4-D degrading microbial populations occurred in this soil. Strains responsible for the hybridization signals observed in total soil bacterial DNA were identified by comparing DNA fingerprints of the isolates to the pattern observed in total soil bacterial DNA. Four different 2,4-D degrading strains (Pseudomonas cepacia DBO 1/pJ P4 ; and three KBS soil isolates identified as Pseudomonas pseudomallei, Pseudomonas paucimobilis, and Pseudomonas pickettii) were inoculated into a native Kansas prairie soil to study competition among and between the indigenous and the inoculated 2,4-D degrading bacteria under 2,4-D selection. The hybridization patterns fiom total soil bacterial DNA revealed that P. cepacia DBO 1/pJ P4 outcompeted the other added strains and indigenous 2,4-D degrading bacteria. P. paucimobilis was also detected as a secondary dominant, but P. pseudomallei and P. pickettii were not. Broth culture competition experiments were also performed. The results suggested that different factors affected competitive ability in soil vs broth and that the 2,4-D degradative plasmids were important determinants of competitiveness. Among the 47 isolates, strain 2811P, identified as Alcaligenes paradoxus, was observed to have a genetically flexible plasmid. The plasmid appeared to integrate in its entirety into the host chromosome and was excised either precisely or imprecisely. ACKNOWLEDGMENTS I wish to express my deep gratitude and appreciation to my major advisor Dr. James M. Tiedie for providing the opportunity for me to persue my graduate education under his guidance. His enthusiastic attitude, dedicated guidance and constant patient were responsible for much of the success of this study. I would like to thank Prof. C. A. Reddy, Prof. Guy L. Bush, Prof. Loren R. Snyder, and Prof. Patrick J. Oriel for their kind service on my doctoral program committee. I would also like to thank Dr. William E. Holben for his assistance and constructive criticism throughout the course of this study, Dr. James R. Cole for his valuable discussion, and Mr. James Bronson for his help during the field experiment. Sincere appreciation is expressed to the lab members who provided me with help, useful suggestions, and encouragement: Arturo Massol-Deya, Bernard Schroeter, Cathy McGowan, David Harris, Joanne Ghee-Sanford, John Dunbar, Jorge Santo Domingo, Joyce Wildenthal, Klaus Nuesslein, Marcos Fries, Rob Sanford, Serguei Bezborodnikov, Suzanne Thiem, Tamara Tsoi, and Yuichi Suwa. Finally, I wish to thank to my parents for their moral and financial support, and to my wife, Myoung-Ok and my son, Chi-Woo for their unselfish sacrifices, constant assistance and encouragement throughout the years. TABLE OF CONTENTS List of tables ...................................................................................................... viii List of figures ..................................................................................................... .1: Chapter One: Literature review ...................................................................... 1 Introduction. .................................................................................................. 2 Microbial adaptation to 2,4-D ....................................................................... 2 Lag period ................................................................................................ 3 Enrichment of soil ................................................................................... 3 Isolation of 2,4-D degrading bacteria ........................................................... 4 Identification of 2,4-D degrading bacteria ................................................... 5 Enumeration of 2,4-D degrading bacteria .................................................. 5 2,4-D degradative plasmid ........................................................................... 7 2,4-D degradative pathway .......................................................................... 7 Induction of 2,4-D enzymes .................................................................... 8 Metabolic pathway of Arthrobacter sp. and Alcaligenes eutrophus JMP134 .................................................................................................... 8 Metabolic pathway of Pseudomonas sp .................................................. 9 2,4—D degradative genes .......................................................................... 9 Environmental factors afi‘ecting 2,4«D degradation ................................... 10 Soil moisture and temperature .............................................................. 10 pH ............................................................................................................ 10 Oxygen ........ ......... ................. 10 Nutrient .................................................................................................... 10 Soil type and sorption. ............................................................................. 11 Rate of application ................................................................................... 1 1 2,4-D formulations ................................................................................... 11 Conclusion ...................................................................................................... 12 List of references ............................................................................................ 13 Chapter Two: Studies on diversity of 2,4-D degrading bacteria isolated from Kellogg Biological Station soils ...................................... 20 Abstract .......................................................................................................... 2 1 Introduction ................................................................................................... 22 Materials and Methods ................................................................................. 23 Results ........................................................................................................... 28 Discussion ...................................................................................................... 34 List of references ........................................................................................... 57 Chapter Three: Use of gene probes to detect 2,4-D degrading populations in soil maintained under selective pressure ........................ 60 Abstract .......................................................................................................... 61 Introduction ................................................................................................... 62 Materials and Methods ................................................................................. 64 Results ........................................................................................................... 67 Discussion ...................................................................................................... 7 0 List of references ........................................................................................... 81 vi Chapter Four: Gene probe analysis of competition among 2,4-D degrading bacteria in soil under selective conditions ........... 85 Abstract .......................................................................................................... 86 Introduction ................................................................................................... 87 Materials and Methods ................................................................................. 88 Results ........................................................................................................... 92 Discussion ...................................................................................................... 98 List of references ......................................................................................... 117 Chapter Five: Integration and excision of a 2,4-D degradative plasmid in Alcaligenes paradoxus and evidence of its natural intergeneric transfer .............................................................. 1 19 Abstract ........................................................................................................ 120 Introduction ................................................................................................. 121 Materials and Methods ............................................................................... 122 Results ......................................................................................................... 125 Discussion .................................................................................................... 130 List of references ......................................................................................... 147 vii LIST OF TABLES Table Page Chapter Two 2,4-D degrading bacteria and their source ................................................. 39 Distribution of isolates among FAME (fatty acid methyl ester) groups ........................................................................................................... 41 Identification of isolates to species level using FAME (fatty acid methyl ester) analyses ............................................................... 42 Hybridization patterns of the isolates with DNA probes .......................... 43 Patterns of utilization of 2,4-D related compounds by selected isolates..44 Comparison of similarities or difi‘erences among isolates of difi‘erent groups using difi'erent methods within groups and among methods for characterizing structure of the isolates .......................................................................................................... 45 viii Chapter Four Bacterial strains and plasmids ................................................................. 103 Comparison of total viable counts, 2,4-D degrading populations and numbers of P. cepacia DBO 1/pJP4 in Konza Prairie soil ................. 104 Growth characteristics of 2,4-D degrading bacteria in axenic broth culture .............................................................................................. 105 Summary of key features of 2,4-D degrading strains and the outcome of competition in difi‘erent environments ................................................. 106 Chapter Five Sequence homology of plasmid and chromosome with 2,4-D metabolic genes .......................................................................................... 134 Transferability of 2,4-D'l' phenotype from donor to P. cepacia DBOl....135 LIST OF FIGURES Figure Page Chapter Two 1 Dendrogram of fatty acid profiles of the isolates ....................................... 47 2 REP PCR patterns of the isolates ............................................................... 48 3 Autoradiogram of the gel after hybridization with the tfdA gene, tde gene, and tde gene and the 6.5 kb probe .......................................... 51 4 Distribution of hybridization groups of 2,4-D degraders in field .............. 56 Chapter Three 1 Degradation of 2,4-D in soil ........................................................................ 76 2 Detection of indigenous 2,4-D degrading populations in soil by Southern blot analysis ................................................................................. 77 3 Comparison by DNA probe hybridization of soil bacterial DNA and plasmid DNA digested with three difi'erent restriction enzymes ............. 78 Detection of indigenous 2,4-D degrading soil bacteria not detected with the tfdA gene probe ............................................................................. 79 Comparison by DNA probe hybridization of soil bacterial DNA and P. paucimobilis DNA digested with threr different restriction enzymes ........................................................................................................ 80 Chapter Four Degradation of 2,4-D in soil during 10 repeated additions in microcosms ............................................................................................. 109 Change in number of soil microorganisms in response to repeated additions of 2,4-D ....................................................................................... 110 Hybridization of labeled tfdA gene probe and 6.5 kb fragment with total soil bacterial DNA fiom three replicate inoculated soils ....... 1 11 Comparison by DNA probe hybridization of total soil bacterial DNA and plasmid DNA digested with three different restriction enzymes ...................................................................................................... 112 Hybridization of labeled tfdA gene probe with total soil bacterial DNA from three replicate uninoculated soil microcosms ........................ 113 Growth patterns of 2,4-D degrading bacteria in axenic culture ............. 114 xi Competition for 2,4-D among 2,4-D degrading bacteria in mixed liquid culture .................................................................................. 115 Effect of culture conditions on enchancing the growth of P. pickettii 7 12 ........................................................................................... 116 Chapter Five Resolution of plasmids of 2,4—D degrading strains .................................. 138 Southern blot hybridization with the 23S rRNA gene to the total DNA of strains 2811P and 28110 .................................................... 139 Growth characteristics of 2,4-D degrading strains in 2,4-D mineral medium under uninduced condition ........................................... 140 Resolution of plasmids in donors, transconjugants, and recipient ......... 141 Line drawing and autoradiogram of a Southern blot hybridization with labeled plasmid pKA2 of restriction enzyme digests of pKA2, total DNA of strain 2811P, and total DNA of strain 28110 .................... 142 Excision of plasmid DNA in strain 28110 during continuous growth on 2,4-D medium ........................................................................... 144 Negative photograph of agarose gel of restriction enzyme-digested xii plasmid DNA. ............................................................................................. 145 8 Negative photograph and autoradiogram of agarose gel of plasmids digested with restriction endonucleases ................................................... 146 xiii Chapter One Literature Review Introduction Large quantities of synthetic organic chemicals have been introduced to the environment. Among them, 2,4-dichlorophenoxyacetic acid (2,4-D) has received widespread use as a herbicide during the past several decades, because it is selectively toxic to broadleaf plants but not to monocotyledonous plants including cereal crops (14, 41). However, apart from its desired toxic effects on the weeds, a herbicide could have harmful effects on nontarget organisms. First, it could influence soil microflora and microbial processes that play a vital role in maintaining soil fertility. Second, the persistent herbicide could damage the sensitive crops, thus reducing the yield. Third, the runoff and leaching of herbicide applied to field could contaminate rivers, lakes, and groundwater, which is a potential danger to animals and to man. Therefore, it is important to study the behaviour of herbicides in soil, their degradation by microorganisms, adaptation behaviour and population ecology of the responsive microorganisms, and the degradative pathways and genes. The information obtained from these studies may be also useful in understanding the degradation mechanisms of various other haloaromatic compounds currently of interest. Microbial adaptation to 2,4-D 2,4-D applied to soil could affect the indigenous microbial populations in two ways : either it is toxic to them or it serves as additional growth substrate for microorganisms able to degrade it. It has been reported that while this chemical significantly stimulated fungal populations at high 3 application rate (30 kg/ha), it could significantly reduce the bacterial populations (14, 49). However, 2,4-D had no significant effect on soil microflora at agronomic rates (0.3 - 5 kg/ha) (61). The breakdown of 2,4-D in soil was demonstrated by radioactive release as 14C02 from 1‘iC-labelled 2,4- , D (36, 48, 76). Microbial involvement in 2,4-D degradation was observed fi'om the comparison of 2,4-D persistence in autoclaved and normal soils (15) and also by using metabolic poison such as sodium azide that inhibits cytochrome oxidase (7). Lag period. Substantial 2,4-D degradation usually occurs after a lag time of 0 to 4 weeks (6) which was shown to be dependent on the concentration of 2,4-D added (27, 43). The lag period appears to be the time required for the development of substantial soil microflora able to degrade 2,4-D (5). Two mechanisms, induction and mutation, were proposed to explain the adaptation phenomenon during the lag period. The induction mechanism suggests that the responsive soil microbial populations which already have the degradative genes require time to induce enzymes before degradation begins, while the mutation mechanism postulates that the ability to degarde 2,4-D arises through a mutation during the lag period (5, 6, 8, 71). Alternatively, it was proposed that the adaptation period in soil is dependent on the population level of the 2,4-D degrading bacteria, the concentration of 2,4-D (2, 43), and the transfer rates of 2,4-D degradative plasmids (73). Although induction was suggested to be the most siginificant mechanism (41), the origin of the effective population is not yet clear. Enrichment of soil. Once the active microbial populations build up to a level where appreciable degradation occurs, degradation of 2,4-D 4 subsequently added occurs more rapidly without a noticeable lag period (7, 46, 56, 74). It has been reported that the 2,4-D degradation rate of the second treatment was three- to five-fold higher in pretreated soils than in unadapted soils and repeated annual applications of 2,4-D resulted in a reduction in degradation time from 10 to 4 weeks (27). However, this enrichment effect was not observed in soil treated with 2,4-D at low application rates (27 , 71), suggesting that low amount of 2,4-D may be metabolized by pre-existing microbial enzymes before the sufiicient development of the active populations (29). It was also reported that 2,4—D was degraded faster in pretreated soil than in unadapted soil one year after the first application, suggesting that the enrichment effect persists for a long period in the absence of the herbicide (5, 6, 27, 45). Since most of the soil microflora live under starvation conditions (9), the loss of their adaptation caused by alternative metabolism would not be extreme (9, 27). Alternatively, a trace of 2,4-D (58) or naturally occurring, structurally related compound was suggested to maintain the active population for a long time (35). Isolation of 2,4-D degrading bacteria It is necessary to amplify the population level of 2,4-D degrading bacteria before attempting any isolations. For this purpose, a high amount of 2,4-D was added to the original environmental samples and whenever the 2,4-D disappears, successive 2,4-D was added (10, 34, 74) or an aliquot of samples was inoculated into 2,4-D mineral medium (3) for further enrichment. Then 2,4-D degrading bacteria were isolated fi'om the enriched samples by streaking on 2,4-D minimum medium. With this enrichment culture technique, some colonies able to degrade 2,4-D were readily obtained (34), 5 while others required 17 days before they produced enough biomass on plates (7 4). Identification of 2,4-D degrading bacteria A variety of different species of bacteria able to degrade 2,4-D have been isolated and identified. The 2,4—D degrading bacteria isolated from soils in 19508 and 1960s were identified mainly by their colony characteristics, cell morphology, and biochemical characteristics. On the other hand, the bacteria isolated from soils and water samples in 19803 were identified by their growth property, pigment formation, flagella arrangement, and G+C content of' chromosomal DNA. 2,4-D degrading microorganisms include Achromobacter sp. (10), Alcaligenes eutrophus JMP134 (18), Alcaligenes paradoxus J MP116 (25), Arthrobacter sp. (42), Arthrobacter globiformis (4), Aspergillus niger (24), Corynebacterium sp. (56), Flavobacten'um aquatile (34), Flavobacterium peregrinum (65), Mycoplana sp. (74), Nocardia coeliaca (33), Nocardia sp. (57), Pseudomonas sp. (22), Pseudomonas sp. HV3 (35), Pseudomonas sp. NCIB 9340 (22), and Streptomyces viridochromogenes (13). Although the identification of the 1950s and 19608 exhibited more diverse and uncommon genera of 2,4-D degrading bacteria than those of the 19808, this could be attributed to the relative imprecision of the earlier identification methods. Enumeration of 2,4-D degrading bacteria The estimation of 2,4-D degrading populations has been reported by several researchers. A bromocresol purple indicator liquid medium and an eosin- 6 methylene blue agar medium were developed to detect and isolate microorganisms able to degrade 2,4-D (39). The production of hydrochloric acid resulting from the oxidation of 2,4-D by organisms either changes the color of the medium from purple to yellow or forms red colonies on the eosin- methylene blue agar. This technique was further improved to estimate 2,4-D degrading bacteria by a most probable number (MPN) method (38). The MPN determination using indicator is more convenient and rapid than the MPN method monitoring the disappearance of the ultraviolet absorption spectrum of 2,4-D (57). It was also reported that the color change resulting from the reaction between chlorine ions released from 2,4-D degradation and added reagents and the release of 14C02 from MPN medium containing labelled 2,4-D could be utilized in enumerating the 2,4-D degrading microorganisms by MPN methods (26). Soils not previously treated with 2,4-D usually contained fewer 2,4-D degrading microorganisms than those in the corresponding soils treated with 2,4-D, the latter population densities range from 1.2 (38) to 8.8 x 106/g soil (36). Numbers of 2,4-D-co-metabolizing microorganisms in soils not amended with 2,4-D ranged from 1.7 x 104 (26) to 3.0 x 105 (27), but their populations in the corresponding treated soils were slightly to markedly increased. The population of 2,4-D degrading microorganisms in soil was maintained at higher levels than that of the 2,4-D-co-metabolizing organisms with relatively high concentrations of 2,4-D (> 5 mg/l), but the former population was not much higher than that of the latter when the concentration was 1.2 ug/l in the MPN medium (26). 2,4-D degradative plasmids It was suggested that 2,4-D degradative gene was evolved from plasmid home phenoxyacetate degradative gene through gene duplication and subsequent mutation (53). In spite of the extensive use of 2,4-D in agriculture, it seldom persists in the environments because of the widespread occurrance of a variety of soil microflora capable of degrading it. The involvement of a conjugative plasmid in the 2,4-D degradation was first reported in Pseudomonas sp. (52) and in Alcaligenes paradoxus (25), providing a mechanism for the widespread distribution of those genes among microbial populations. Since then, six other 2,4—D degradative plasmids were described from Alcaligenes species isolated from soil (18). Among them, one plasmid from Alcaligenes paradoxus, pJ P4, has been well characterized physically and genetically by using cloning, transposon mutagenesis, and restriction endonuclease analysis (16). The plasmid belongs to the P1 incompatibility group, has broad host range, encodes resistance to mercuric chloride, and degradation of 2,4-D, 2- methyl-4-chlorophenoxyacetic acid (MCPA), and 3-chlorobenzoate (16). 2,4-D degradative pathway Most of the research on the 2,4-D degradative pathway has been performed with aerobic cultures and not much attention has been given to possible anaerobic degradation, since ring cleavage following aerobic hydroxylation seems to be the major mechanism of 2,4-D dgradation. Several different 2,4-D degradative pathways were suggested in studies with Pseudomonas species (23), Achromobacter species (65), and Nocardia 8 species (5) prior to 1967, but the pathways were based solely on indirect evidences and were not well characterized. Since then, details of the pathway was revealed by the extensive studies with growing or resting whole cells, enzyme preparations, transposon mutagenesis, and gene cloning in Arthrobacter sp. (11, 12, 21, 40, 42, 70), Pseudomonas sp. (22), and Alcaligenes eutrophus JMP 134 (17). Induction of 2,4-D enzymes. 2,4-D degrading enzymes of Flavobacterium peregrinum were induced when grown on 2,4-D, MCPA, or 2-chloro-4- methylphenoxyacetic acid, while MCPA was not degraded by this 2,4-D degrading enzyme system (65). In the Achromobacter, the 2,4-D-degrading enzyme system was induced by 2,4-D and 2,4-dichlorophenol (65). In Alcaligenes eutrophus JMP 134, the 2,4-D degradative pathway was induced in the presence of 2,4-D or 3-chlorobenzoate plus Casamino acids, but Casamino acids alone could not induce the 2,4-D enzymes (31). Metabolic pathway of Arthrobacter sp. and Alcaligenes eutrophus JMP134. The pathway of 2,4-D degradation observed fi'om these bacteria involves the removal of the acetic acid side chain to yield 2,4—dichlorophenol, followed by ortho hydroxylation of the phenol to produce 3,5-dichlorocatechol. The catechol is then converted to a muconic acid by ortho cleavage of the aromatic ring. The muconic acid is lactonized to form 2-chlorodiene-lactone, during which one of the chlorines is lost from the aromatic ring. The chlorodiene-lactone is delactonized to produce maleylacetic acid which is subsequently metabolized to succinic acid in Arthrobacter sp. 9 Metabolic pathway of Pseudomonas sp. The pathway observed from two Pseudomonas sp. by Evans et. al. (22) was similar to that for the Arthrobacter sp. However, other related compounds, such as 2-chlorophenol and 2,4-dichloro-6-hydroxyphenoxyacetate, were detected during growth on 2,4-D medium. Since these organisms were capable of' oxidizing a variety of mono- and dichlorophenols and phenoxyacetates, different 2,4-D degradative pathways were also suggested to operate in these bacteria. 2,4-D degradative genes. 2,4-D degradative genes, such as tfdA, tde, tde, tde, tde, and tde were identified, localized, and cloned inAlcaligenes eutrophus J MP 134 (17, 68). Four genes, tde to tde, encode 2,4- dichlorophenol hydroxylase, dichlorocatechol 1,2 -dioxygenase, chloromuconate cycloisomerase, and chlorodienelactone hydrolase, respectively, but the function of tde gene is not yet known (17 ). These five genes were observed to constitute a cluster on plasmid pJ P4, while the tfdA gene, encoding 2,4-D monooxygenase, has been localized to a region 13 kb away from the cluster. Recently a second 2,4-D monooxygenase, designated as tfdAII, has been observed on a different region of pJ P4 (54). The tfdAII gene was clustered with the five genes tdeCDEF on plasmid pJP4 and substantial difference was observed between the tfdA gene and the tfdAII gene from DNA-DNA hybridization experiments. It was also suggested that two different genes encoding chlorocatechol 1,2-dioxygenase were present on plasmid pJ P4 (28). 10 Environmental factors affecting 2,4-D degradation Soil moisture and temperature. Soil moisture and temperature were observed to affect significantly the microbial activity. The optimum moisture tension and temperature was 0.1 bar and 27°C, respectively, in a sandy loam soil (51). At moisture above the optimum (37) and at temperatures above the optimum (51), 2,4-D was degraded less rapidly. In broth cultures, 2,4-D degradation was faster at 25°C than at warmer temperatures (72) and in river water studies, the optimum temperature range was 21 - 25°C (44). pH. The optimum 'pH range for 2,4-D degradation in broth culture was 6.2 - 6.9 (72). The 2,4-D degradation in soil was slower in the acid range (pH 3.8 - 5.6) than in the neutral range (pH 6.6 to 7.4) (67). In the neutral pH range, more 2,4-D is thought to be present as its dissociated form which could be less toxic. 2,4-D is degraded mainly by fungi in acid soils (41), while its degradation in neutral soils is catalyzed mainly by bacteria and actinomycetes (1). Oxygen. Since 2,4-D degradation is mainly by an aerobic process, oxygen concentration is suggested to be an important factor. 2,4-D degradation rate was much lower in flooded soils than that in non-flooded soils (57, 77) and an other study shows that the degradation rate generally increased with increasing oxygen supplies (59). Nutrient. While the addition of glucose plus urea to soil (48) was found to have no significant effect on the 2,4-D degradation rate, most of the studies (19, 20, 44, 48) suggested that the organic amendment or the addition of 11 fertilizer could stimulate the degradation rate. It was suggested that the added compounds could increase the corresponding populations, thus enhancing the degradation rate (55). Soil type and adsorption. The adsorption of 2,4-D and microorganisms to soil particles was found to be controlled by the soil organic content and the sorbed 2,4-D was completely protected from biological degradation (47). Clay was observed to have low adsorptive capacity for 2,4-D (30). It was reported that 2,4-D was degraded more slowly in a fine sand, which had a lower pH (5.2) and a higher organic content, than in a sandy clay loam (63, 69). The mobility of 2,4-D is relatively high (32) and it leaches readily in sandy and low organic soils (75), thus being not metabolized by microorganisms. Rate of application. A high concentration of 2,4-D (500 ug/g soil) was observed to be toxic to the soil microflora in a loam and a sandy loam, causing a long lag prior to substantial degradation (67 ). Lag time was proportionally increased with increasing 2,4-D concentration (50) and 2,4-D degradation rate was reduced due to a high 2,4-D concentration (49). However, it was suggested that a different 2,4-D application rate (0.33 to 330 ug/g) did not influence co-metabolizing soil microflora as long as the 2,4-D concentration was not toxic and the co-substrate was not depleted (27 ). 2,4-D formulations. The amine or ester formulations of 2,4-D were shown to rapidly dissociate to the anion in soils (62, 64, 66). The formulated dimethylamine salt was degraded more rapidly, probably due to the presence of additional nutrients in this salt, than that for technical grade at high application rates (49). 12 Conclusion The herbicide 2,4—D is readily degraded by soil microorganisms. The lag time is usually observed in the first exposure, but it is not noticeable from subsequent doses. The adaptation effect is maintained for a long time in soil after disappearance of the herbicide. The number of 2,4-D degrading microorganisms in soil previously treated with 2,4-D is generally higher than that in soil not amended with 2,4-D. A variety of different bacterial species involve in the 2,4-D degradation and the degradative plasmids and genes have been isolated and described. However, there are still many areas to be investigated regarding 2,4-D degradation. Although various different bacterial species able to degrade 2,4- D have been independently isolated and characterized, few studies have contributed to the systematic study on their diversity and not much information is available on the ecological significance and dynamics of an individual population in an ecosystem. The enumeration of 2,4-D degrading microorganisms by the MPN method has provided information on the population, but it can not furnish information on their competitive interactions and their population and genetic changes in the environments under the 2,4-D selective condition. Although numerous 2,4-D degradative plasmids have been isolated and identified, little attention has been given to their diversity and ecological contributions and further research is required to reveal the possibly diverse pathways of 2,4-D degradation. Finally, there has been little effort to investigate physiologic and genetic aspects of the adaptive power of 2,4-D degrading bacteria to grow more efficiently under 2,4-D selection. LIST OF REFERENCES 13 LIST OF REFERENCES 1. Alexander, M. 1977. Introduction to soil microbiology. 2nd Ed. John Wiley & Sons, New York. 2. Altom, J. D., and J. F. Stritzke. 1973. Degradation of dicamba, picloram, and four phenoxy herbicides in soils. Weed Sci. 21:556-560. 3. Amy, P. S., J. W. Schulke, L. M. Frazier, and R. J. Seidler. 1985. Characterization of aquatic bacteria and cloning of genes specifying partial degradation of 2,4-dichlorophenoxyacetic acid. Appl. Environ. Microbiol. 49:1237-1245. 4. Audus, L. J. 1950. Biological detoxication of 2,4-dichlorophenoxyacetic acid in soils: Isolation of an effective organism. Nature 166:356. 5. Audus, L. J. 1964. Hrebicide behaviour in soil, p. 163-206. In L. J. Audus (ed.), The physiology and biochemistry of herbicides. Academic Press, London. 6. Audus, L. J. 1960. Microbiological breakdown of herbicides in soils, p. 1- 19. In E. K. Woodford, and G. R. Sagar (ed.), Herbicides and the soil. Blackwell, Oxford. 7. Audus, L. J. 1951. The biological detoxication of 2,4-dichlorophenoxy- acetic acid in soil. Plant Soil 3:17 0- 192. 8. Audus, L. J. 1952. The decomposition of 2,4-dichlorophenoxyacetic acid and 2-methyl-4-chlorophenoxyacetic acid in soil. J. Sci. Food Agr. 3:268- 274. 9. Babiuk, L. A., and E. A. Paul. 1970. The use of fluorescein isothiocyanate in the determination of the bacterial biomass of grassland soil. Can. J. Microbiol. 16:57-62. 10. Bell, G. R. 1957. Some morphological and biochemical characteristics of a soil bacterium which decomposes 2,4-dichlorophenoxyacetic acid. Can. J. Microbiol. 3:821-840. 11. Bollag, J. -M., C. S. Helling, and M. Alexander. 1968. Enzymatic hydroxylation of chlorinated phenols. J. Agr. Food Chem. 16:826-828. 14 12. Bollag, J. -M., G. G. Briggs, J. E. Dawson, and M. Alexander. 1968. Enzymatic degradation of' chlorocatechols. J. Agr. Food Chem. 16:829-833. 13. Bounds, H. C., and A. R. Colmer. 1965. Detoxication of some herbicides by Streptomyces. Weeds 13:249-252. 14. Breazeale, F. W., and N. D. Camper. 1970. Bacterial, fungal, and actinomycete populations in soils receiving repeated applications of 2,4- dichlorophenoxyacetic acid and trifluralin. Appl. Microbiol. 19:379-380. 15. Brown, J. W., and J. W. Mitchell. 1948. Inactivation of 2,4-dichloro- phenoxyacetic acid in soil as affected by soil moisture, temperature, the addition of manure and autoclaving. Botan. Gaz. 109:314-323. 16. Don, R. H., and J. M. Pemberton. 1985. Genetic and physical map of the 2,4-dichlorophenoxyacetic acid-degradative plasmid pJ P4. J. Bacteriol. 161:466-468. 17. Don, R. H., A. J. Weightman, H. -J. Knackmuss, and K. N. Timmis. 1985. Transposon mutagenesis and cloning analysis of the pathways for degradation of 2,4-dichlorophenoxyacetic acid and 3-chlorobenzoate in Alcaligenes eutrophus JMP134 (pJ P4). J. Bacteriol. 161:85-90. 18. Don, R. H., and J. M. Pemberton. 1981. Properties of six pesticides degradation plasmids isolated from Alcaligenes paradoxus and Alcaligenes eutrophus. J. Bacteriol. 145:681-686. 19. Doyle, R. C., D. D. Kaufman, and G. W. Burt. 1978. Effect of dairy manure and sewage sludge on 1‘fC-pesticide degradation in soil. J. Agric. Food Chem. 26:987-989. 20. Duah-Yentumi, S., and S. Kuwatsuka. 1982. Microbial degradation of benthiocarb, MCPA and 2,4-D herbicides in perfused soils amended with organic matter and chemical fertilizers. Soil Sci. Plant N utr. 28:19-26. 21. Duxbury, J. M., J. M. Tiedie, M. Alexander, and J. E. Dawson. 1970. 2,4-D metabolism: Enzymatic conversion of chloromaleyiacetic acid to succinic acid. J. Agr. Food Chem. 18:199-201. 22. Evans, W. C., B. S. W. Smith, H. N. Fernley, and J. I. Davies. 1971. Bacterial metabolism of 2,4-dichlorophenoxyacetate. Biochem. J. 122:543- 551. 23. Evans, W. C., and B. S. W. Smith. 1954. The photochemical inactivation and microbial metabolism of the chlorophenoxyacetic acid herbicides. Biochem. J. 57, XXX. 15 24. Faulkner, J. K., and D. J. Woodcock. 1964. Metabolism of 2,4- dichlomphenoxyacetic acid (2,4-D) by Aspergillus niger van Tiegh. Nature 203:865. 25. Fisher, P. R., J. Appleton, and J. M. Pemberton. 1978. Isolation and characterization of the pesticides-degrading plasmid pJ P1 fi'om Alcaligenes paradoxus. J. Bacteriol. 135:798-804. 26. Fournier, J. C. 1980. Enumeration of the soil micro-organisms able to degrade 2,4-D by metabolism or co-metabolism. Chemosphere 9: 169-174. 27. Fournier, J. C., P. Codaccioni, and G. Soulas. 1981. Soil adaptation to 2,4-D degradation in relation to the application rates and the metabolic behaviour of the degrading microflora. 10:977-984. 28. Ghosal, D., and I. -S. You. 1988. Nucleotide homology and organization of chlorocatechol oxidation genes of plasmids pJ P4 and pACZ7. Mol. Gen. Genet. 211:113—120. 29. Greaves, M. P., H. A. Davies, J. A. P. Marsh, G. I. Wingfield. 1976. Herbicides and soil microorganisms. CRC Crit. Rev. Microbiol. 5:1—38. 30. Grover, R. 197 3. The adsorptive behaviour of acid and ester forms of 2,4- D on soils. Weed Res. 13:51-58. 31. Harker, A. R., R. H. Olsen, and R. J. Seidler. 1989. Phenoxyacetic acid degradation by the 2,4-dichlorophenoxyacetic acid (TFD) pathway of plasmid pJ P4: Mapping and characterization of the TFD regulatory gene, tde. J. Bacteriol. 171:314-320. 32. Helling, C. S. 1971. Pesticide mobility in soils II. Applications of soil thin-layer chromatography. Soil Sci. Soc. Am. Proc. 35:737-743. 33. Jensen, H. L., and H. 1. Peterson. 1952. Decomposition of hormone herbicides by bacteria. Acta Agr. Scand. 2:215-231. 34. Jensen, H. L., and H. I. Peterson. 1952. Detoxication of hormone herbicides by soil bacteria. Nature 170:39-40. 35. Kilpi, S., V. Backstrom, and M. Korhola. 1980. Degradation of 2- methyl-4-chlorophenoxyacetic acid (MCPA), 2,4-dichlorophenoxyacetic acid (2,4-D), benzoic acid and salicylic acid by Pseudomonas HV3. FEMS Microbiol. Letters 8:17 7- 182. 36. Kunc, F., and J. Rybarova. 1983. Mineralization of carbon atoms of 14C-2,4-D side chain and degradation ability of bacteria in soil. Soil Biol. Biochem. 15:141-144. 16 37. Lavy, T. L., F. W. Booth, and C. R. Fenster. 197 3. Degradation of 2,4- D and atrazine at three soil depths in the field. J. Environ. Qual. 2:132- 137. 38. Loos, M. A., I. F. Schlosser, and W. R. Mapham. 1979. Phenoxy herbicide degradation in soilszQuantitative studies of 2,4-D and MCPA- degrading microbial populations. Soil Biol. Biochem. 1:37 7 -385. 39. Loos, M. A. 1975. Indicator media for microorganisms degrading chlorinated pesticides. Can. J. Microbiol. 21:104-107. 40. Loos, M. A., J. -M. Bollag, and M. Alexander. 1967. Phenoxyacetate herbicide detoxication by bacterial enzymes. J. Agr. Food Chem. 15:858- 860. 41. L008, M. A. 1975. Phenoxyalkanoic acids, p. 1-56. In P. C. Kearney, and D. D. Kaufinan (ed), Herbicides: Chemistry, Degradation and Mode of Action. Marcel Dekker Inc., New York. 42. Loos, M. A., R. N. Roberts, and M. Alexander. 1967 . Phenols as intermediates in the decomposition of phenoxyacetates by an Arthrobacter species. Can. J. Microbiol. 13:679-689. 43. Miwa, N., and S. Kuwatsuka. 1990. Enrichment process of 2,4-D degraders in different soils under upland conditions. Soil Sci. Plant Nutr. 36:261-266. 44. Nesbitt, H. J ., and J. R. Watson. 1980. Degradation of the herbicide 2,4-D in river water-II. The role of suspended sediment, nutrients and water temperature. Water Res. 14:1689- 1694. 45. Newman, A. S., and J. R. Thomas. 1949. Decomposition of 2,4- dichlorophenoxyacetic acid in soil and liquid media. Proc. Soil Sci. Soc. Am. 14:160-164. 46. Newman, A. S., J. R. Thomas, and R. L. Walker. 1952. Disappearance of 2,4-dichlorophenoxyacetic acid from soil. Proc. Soil Sci. Soc. Am. 6:21- 24. 47. Ogram, A. V., R. E. Jessup, L. -T. Ou, and P. S. C. Rao. 1985. Effects of (2,4-dichlorophenoxy)acetic acid in soils. Appl. Environ. Microbiol. 49:582-587. 48. On, L-T, D. F. Rothwell, W. B. Wheeler, and J. M. Davidson. 197 8. The efi'ect of high 2,4-D concentrations on degradation and carbon dioxide evolution in soils. J. Environ. Qual. 7 :241-246. 49. Ou, L-T., J. M. Davidson, and D. F. Rothwell. 197 8. Response of soil microflora to high 2,4-D applications. Soil Biol. Biochem. 10:443-445. 17 50. Parker, L. W., and K. G. Doxtader. 1982. Kinetics of microbial decomposition of 2,4-D in soil: Efi‘ects of herbicide concentration. J. Environ. Qual. 11:679-684. 51. Parker, L. W., and K. G. Doxtader. 1982. Kinetics of the microbial degradation of 2,4-D in soil: Effects of temperature and moisture. J. Environ. Qual. 12:553-558. 52. Pemberton. J. M., and P. R. Fisher. 1977 . 2,4-D plasmids and persistence. Nature 268:732-733. 53. Pemberton. J. M., B. Corney, and R. H. Don. 197 9. Evolution and spread of pesticide degrading ability among soil micro-organisms, p. 287- 299. In K. N. Timmis, and A. Puhler (ed.), Plasmids of medical, environmental and commercial importance. Elsevier/North-Holland Biomedical Press. 54. Perkins, E. J., and P. F. Lurquin. 1988. Duplication of a 2,4- dichlorophenoxyacetic acid monooxygenase gene in Alcaligenes eutrophus JMP134 (pJP4). J. Bacteriol. 170:5669-5672. 55. Pfaender, F. K., and M. Alexander. 1973. Effect of nutrient additions on the apparent cometabolism of DDT. J. Agr. Food Chem. 21:397-401. 56. Rogoff, M. H., and J. J. Reid. 1956. Bacterial decomposition of 2,4- dichlorophenoxyacetic acid. J. Bacteriol. 71:303-307. 57. Sandman, E. R. I. C. 1984. Enumeration of 2,4-D-degrading microorganisms in soils and crop plant rhizospheres using indicator media ; high populations associated with sugarcane (Saccharum officinarum). Chemosphere 13:1073-1084. 58. Sattar, M. A., and J. Paasivirta. 1980. Fate of chlorophenoxyacetic acids in acid soil. Chemosphere 9:745-752. 59. Shaler, T. A., and G. M. Klecka. 1986. Efi‘ects of dissolved oxygen concentration on biodegradation of 2,4-dichlorophenoxyacetic acid. Appl. Environ. Microbiol. 51:950-955. 60. Simon-Sylvestre, G., and J. C. Fournier. 197 9. Effects of pesticides on soil microflora. Adv. Agron. 31:1—92. 61. Singh. K. 1971. Effect of 2,4-D and simazine on total bacteria, fungi, Azotobacter, ammonification, and nitrification under field conditions. Pesticides. 5:14-17. 18 62. Smith, A. E., and B. J. Hayden. 1976. Field persistence studies with eight herbicides commonly used in Saskatchewan. Can. J. Plant Sci. 56:7 69-7 7 l. 63. Smith, A. E., and D. C. G. Muir. 1980. Determination of extractable and non-extractable radioactivity from prairie soils treated with carboxyl- and ring-labelled (14C)-2,4-D. Weed Res. 20:123- 129. 64. Smith, A. E. 197 6. Use of acetonitrile for the extraction of herbicide residues from soils. J. Chromatogr. 129:309-314. 65. Steenson, T. 1., and N. Walker. 1957 . The pathway of breakdown of 2,4-dichloro- and 4-chloro-2-methyl-phenoxyacetic acid by bacteria. J. Gen. Microbiol. 16:146-155. 66. Stewart, D. K. R., and S. O. Gaul. 1977 . Persistence of 2,4-D, 2,4,5-T and dicamba in a dykeland soil. Bull Environ. Contam. Toxicol. 18:210- 218. 67. Stott, D. E., J. P. Martin, D. D. Focht, and K. Haider. 1983. Biodegradation, stabilization in humus, and incorporation into soil biomass of 2,4-D and chlorocatechol carbons. Soil Sci. Soc. Am. Proc. 47 :66-70. 68. Streber, W. R., K. N. Timmis, and M. H. Zenk. 1987. Analysis, cloning, and high-level expression of 2,4-dichlorophenoxyacetate monooxygenase gene tfdA of Alcaligenes eutrophus JMP 134. J. Bacteriol. 169:2950-2955. 69. Thompson, D. G., G. R. Stephenson, K. R. Solomon, and A. V. Skepasts. 1984. Persistence of (2,4-dichlorophenoxy)acetic acid and 2- (2,4-dichlorophenoxy)propionic acid in agricultural and forest soils of northern and sourthern Ontario. J. Agr. Food Chem. 32:57 8-581. 70. Tiedje, J. M., J. M. Duxbury, M. Alexander, and J. E. Dawson. 1969. 2,4-D metabolism: Pathway of degradation of chlorocatechols by Arthrobacter sp. J. Agr. Food Chem. 17:1021-1026. 71. Torstensson, N. T. L., J. Stark, and B. Goransson. 1975. The effect of repeated applications of 2,4-D and MCPA on their breakdown in soil. Weed Research 15:159- 164. 72. Tyler, J. E., and R. K. Finn. 1974. Growth rates of a Pseudomonad on 2,4-dichlorophenoxyaceticacid and 2,4-dichlorophenol. Appl. Microbiol. 28:181-184. 73. Waid, J. S. 1972. The possible importance of transfer factors in the bacterial degradation of herbicides in natural ecosystems. Residue Review 44:65-71. 19 74. Walker, R. L., and A. S. Newman. 1956. Microbial decomposition of 2,4-dichlorophenoxyacetic acid. Appl. Microbiol. 4:201—206. 75. Wiese, A. F., and R. G. Davis. 1964. Herbicide movement in soil with various amounts of water. Weeds 12:101-103. 76. Wilson, R. G., and H. H. Jr. Chang. 1978. Fate of 2,4-D in a Naif silt loam soil. J. Environ. Qual. 7 :281-286. 77. Yochida, T, and T. F. Castro. 197 5. Degradation of 2,4-D, 2,4,5-T and picloram in two Philippine soils. Soil Sci. Plant Nutr. 21:397-404. Chapter Two Studies on Diversity of 2,4-D Degrading Bacteria Isolated from Kellogg Biological Station Soils. J. 0. KA, W. E. HOLBEN and J. M. TIEDJE*. Center for Microbial Ecology and Department of Microbiology and Public Health, Michigan State University, East Lansing, MI 48824 20 21 ABSTRACT A total of 47 predominant 2,4-D degrading bacteria were isolated at different times from 1989 to 1992 from agricultural soils either not treated with 2,4-D or treated with 2,4-D at different rates. The isolates were analyzed by fatty acid methyl ester (FAME) profiles, patterns of the polymerase chain reaction (PCR) amplified repetitive extragenic palindromic (REP) sequences, and hybridization patterns with tfd genes from the plasmid pJP4 and a 6.5 kb probe subcloned from a 2,4-D degrading isolate (Pseudomonas paucimobilis). Substrate utilization abilities of selected isolates among the strains identified as the same species by FAME analysis were evaluated by using 2,4-D related compounds as sole carbon sources. FAME analysis revealed 12 FAME groups that cluster with Euclidian Distance of less than 10. 55.3 % of the isolates could be identified to the species level ; the remainder could not be identified due to their poor match to any profile within the MIDI (Microbial ID, Inc.) library and to lack of growth on laboratory medium. The REP PCR patterns of the isolates, identified as different species or genera by FAME analysis, were quite distinct from one another. Hybridization analysis revealed four hybridization groups. Group I which showed sequence homology with the four tfd genes clustered within a Euclidian Distance of 31.5 by FAME analysis, group II which had homology only with the tfdA gene clustered at 11.8, group III which had homology only with the 6.5 kb probe clustered at 6.4, and group IV (all others) which showed no homology to any of the available probes clustered at 31.5. Group I, II, and IV strains exhibited fairly diverse REP PCR patterns within groups, while group III strains yielded similar patterns. Strains of the same hybridization groups exhibited general similarity in substrate utilization abilities. Strains 22 of groups I and III were preferably detected from soils treated with higher rates of 2,4-D and were encountered more fi'equently with an increase in number of 2,4-D treatments. INTRODUCTION Bacterial diversity in a natural ecosystem reveals the structure of a microbial community arising fiom the complex interactions between microorganisms and their biotic and abiotic environments. It has been of general interest to microbial ecologists, in part because the composition of the microbial community plays an important role in the effective bioremediation and recyling of a variety of organic and inorganic contaminants. 2,4-D seldom persists in the environment due to the widespread occurrance of a variety of soil microflora capable of its degradation (2, 32). The degradation of 2,4-D by microorganisms has received substantial attention not only because it has been extensively used as a herbicide, but also because it could serve as a model compound in understanding the biodegradation mechanism of other haloaromatic compounds in general use. Numerous different bacterial genera, such as Achromobacter, Alcaligenes, Arthrobacter, Corynebacterium, Flavobacterium, Pseudomonas, and Streptomyces have been reported to be involved in 2,4-D degradation (3, 6, 12, 13, 18, 25, 30). Populations of microorganisms able to degrade 2,4-D were estimated by the most probable number method in soils (14, 22) and the 2,4-D degradative plasmids were isolated and biophysically and genetically characterized (8, 10). However, little is known about the diversity of 2,4-D degrading bacteria isolated from environmental samples. Many researchers have isolated 23 different species of 2,4-D degrading bacteria in 19508 and 19608, and the bacteria were identified based on their morphological and cultural characteristics. Recently, aquatic microorganisms able to degrade 2,4-D were isolated from a variety of water samples and were phenotypically and genetically characterized (1). But this research provided only fragmentary information on the diversity of 2,4-D degrading bacteria and the isolation procedure using original environmental samples as inoculum for an enrichment culture with high amount of 2,4-D could select for only a certain type of 2,4-D degrading bacteria. Moreover, the inconsistency and heterogeneity of strain sources and identification make it difficult to get consistent results on diversity and on ecological significance of the isolates in an ecosystem. In this work we studied the diversity of 2,4-D degrading bacteria isolated at different times from 1989 to 1992 from agriculture soils not treated or treated with 2,4-D at different rates by analyzing FAME profiles, REP PCR patterns obtained from the colonies, hybridization patterns with DNA probes, and capabilities to metabolize 2,4-D related compounds. The results were compared for analysis of relationships between the different grouping methods. Finally, we investigated the influences of different 2,4-D application rates and repeated treatments on the distribution of the hybridization groups of the isolates in field. MATERIALS AND METHODS Media and culture conditions. All isolates were maintained on MMO mineral medium (33) plus 500 ppm (pg/ml) of 2,4-D. Peptone-tryptone-yeast extract-glucose (PTYG) medium containing 0.25 g of peptone (Difco 24 Laboratories, Detroit, MI), 0.25 g of tryptone (Difco), 0.5 g of yeast extract (Difco), 0.5 g of glucose, 0.03 g of magnesium sulfate, and 0.003 g of calcium chloride was used for strain purification and colony production for REP PCR. Isolation of bacterial strains. Agricultural soil samples from duplicate plots that have not been treated with 2,4-D or have been regularly treated with 2,4-D at different rates (1, 10, and 100 ug/g soil) from 1988 to 1992 (Table 1) were taken from the Long-Term Ecological Research (LTER) site of Kellogg Biological Station (KBS, Hickory Corners, MI), sifted through a 2-mm sieve, and kept at 4°C prior to use. A 10-g soil sample of each plot was homogenized with 95 ml of sterilized, cold 0.85 % saline by shaking at 200 rpm for 20 min in a New Brunswick G24 environmental incubator shaker (New Brunswick Scientific Co., New Brunswick, NJ). Samples (0.1 ml) of appropriate 10-fold dilutions were inoculated into MPN (most probable number) tubes containing 3 ml of 2,4-D mineral medium (MMO plus 500 ppm of 2,4-D). The tubes were incubated at 30°C for 3 weeks and then the degradation of 2,4-D was analyzed by using an Hewlett Packard series 1050 HPLC equipped with Lichrosorb RP-18 column (Anspec Co., Ann Arbor, MI). The culture of the highest dilution showing 2,4-D degradation was enriched through two additional transfers into fresh medium. The enriched culture was streaked onto 2,4-D agar medium (MMO + 500 ppm of 2,4-D + 0.1 % of Casamino acids + 1.5 % agar) and then single colonies were tested for 2,4-D degradation in a flesh 2,4-D mineral medium before it was further purified by streaking onto PTYG plates. Some cultures failed to produce single colonies able to degrade 2,4-D. Altogether 47 strains were isolated from the eight plots (duplicate control plots not treated with 2,4-D and duplicate plots treated with 1, 10, or 100 ppm 2,4-D) from 1989 to 1992 (Table 1). These isolates were preserved by freezing (-7 0°C) in sterile 15 % glycerol. 25 Chemicals. Analytical grade 2,4-D, phenoxyacetic acid (PA), and 2-methyl- 4-chlorophenoxyacetic acid (MCPA) were obtained from Sigma Chemical Co. (St. Louis, MO) and 2-chlorophenoxyacetic acid (2-CPA), 4- chlorophenoxyacetic acid (4-CPA), 3-chlorobenzoic acid (3-CB), and 4- chlorobenzoic acid (4-CB) were obtained from Aldrich Chemical Co. (Milwaukee, WI). FAME analysis. The isolates were cultured on tryptic soy agar medium (15) at 28°C for 48-7 2 h and then cells were harvested from the plate by scraping with a sterile glass loop and used for FAME analysis. Saponification, methylation and extraction were performed using the procedure described in MIDI (Microbial Identification, Inc.) manual (31). The extracts were assayed by using a Perkin Elmer 3920 gas-liquid chromatography equipped with a flame ionization detector. Cluster analysis was carried out using an inhouse cluster program and the MIDI software. Colony REP PCR. This method was performed as described by deBruiiin (7). Each isolate was grown on PTYG plate for 24-48 h and then small amount of cells was resuspended in 25 pl PCR mixture containing 50 pmol each of primers REP 1 and REP 2 (7), 1.25 mM deoxynucleoside triphosphates, 0.04 % bovine serum albumin, 10 % dimethylsulfoxide, and 2 U of AmpliTaq polymerase (Perkin-Elmer Cetus). The amplications were carried out with a DNA Thermal Cycler (Perkin-Elmer Cetus). After the reactions, 8 pl of REP PCR products were separated by electrophoresis on horizontal 1.5 % agarose gels. Genetic diversity analysis. Individual isolates were cultured in 2,4-D broth medium, harvested, and lysed as described by Kado and Liu (21) with some modifications (19). DNA samples obtained from the cell lysate were subjected to gel electrophoresis to detect and separate plasmid DNA (if any) 26 from linear chromosomal DNA, Southern transferred to nitrocellulose hybridization membrane, and hybridized with 32P-labelled DNA probes. As preliminary experiments for the genetic diversity study, in 1989 we isolated two 2,4-D degrading bacteria (K11 and 1443) from the plot treated with 2,4-D at 100 ppm and one strain (2811P) from the control plot not treated with 2,4-D. Strain K11 was observed to have sequence homology with the tfd genes (tfdA, tde, tde, and tde) of the 2,4-D metabolic pathway of the well-known plasmid pJ P4. However, strain 2811P had sequence homology only with the tfdA gene but not with the tde, tde, and, tde genes and moreover, strain 1443 did not have sequence homology with any of the tfd genes. Hence, the tfd genes (tfdA, tde, tde, and tde) (16) were selected as probes to examine the possible genetic diversification of the 2,4-D genes of other 2,4-D degrading isolates and we developed a new DNA probe, a 6.5 kb fragment, to distinguish strain 1443 types. The 6.5 kb fragment was subcloned a8 a BamHI fragment into the multiple cloning site of plasmid pUC 19 (27) fi'om the large plasmid of strain 1443 based on weak cross- hybridization under low stringency condition between the DNA of strain 1443 and the plasmid DNA of another 2,4-D degrading isolate, strain 712, which showed the same hybridization patterns with the tfd genes as strain 2811P. The basic idea for subcloning this fragment was that the 2,4-D degradative, transferable plasmid of strain 712, which does not have sequence homology with the tde, tde, and tde genes, might have some homology with the DNA of strain 1443, which does not have homology with all the tfd genes. The 6.5 kb fragment was found to be specific for strain 1443 types and did not show any detectable homology to plasmid DNA such as pJ P4 and the plasmid of strain 712 under high stringency condition. The tfd genes were labelled with 32P by using the Random Primed DNA Labeling kit (Boehringer 27 Mannheim, Indianapolis, IN) and the 6.5 kb fragment was labeled by using the Nick Translation kit (Boehringer Mannheim). Prehybridization, hybridization and post-hybridization washes were performed as described by Holben et. al. (17) with some modifications. The membranes were prehybridized for at least 24 h at 42°C and after hybridization, three washes were carried out at room temperature, followed by one wash at 65°C for 45 min. Hybridization signals were detected by autoradiography using Kodak X- omat AR film (Kodak, Rochester, NY) exposed at -7 0°C with a Quanta III (Sigma, St. Louis, MO) intensifying screen. If necessary, bound probe was stripped fi'om the membranes prior to rehybridization by washing for 15 min, 3 times with boiled distilled water containing 0.1 % sodium dodesyl sulfate (w/v). Individual isolates were grouped based on their hybridization patterns with probes. Degradation phenotype analysis. A representative strain selected from the isolates identified as the same species by FAME analysis was cultured in 2,4-D mineral medium. The same medium containing 250 ppm of sodium acetate instead of 2,4-D was used to produce cells not adapted to metabolize 2,4-D. Cultures were grown at 30°C and aerated by shaking at 200 rpm in an incubator shaker. Cells in the late log phase were harvested by centrifugation (10,000 x g, 10 min, 4°C), washed twice with an equal volume of 0.02 M phosphate buffer (pH 7.0), and resuspended in the same buffer. An aliquot of suspended cells was inoculated into culture tubes containing MMO mineral medium plus 250 ppm of one of the following seven compounds : 2,4- D, 2-chlorophenoxyacetic acid (2-CPA), 4-chlorophenoxyacetic acid (4-CPA), phenoxyacetic acid (PA), 2-methyl-4-chlorophenoxyacetic acid (MCPA), 3- chlorobenzoic acid (3-CB), and 4-chlorobenzoic acid (4-CB). The cultures were shaken at 150 rpm and 30°C for 2 weeks, after which optical density was read 28 at 550 mp with a Hewlett Packard 8452A Diode Array spectrophotometer to measure the cell growth. To determine the degradation of phenoxyacetates, the cultures were centrifuged to remove the cellular material and the absorption of ultraviolet light at the wavelength of maximum absorption was determined by scanning the sample in Hewlett Packard spectrophotometer. RESULTS FAME analysis. FAME analysis of the isolates (Figure 1) revealed 12 FAME groups of strains showing a Euclidian distance of less than 10 and all isolates clustered at a Euclidian distance of 48.8. One dominant FAME group contained 34 % of the isolates (i. e., 16 strains, Table 2). Four FAME groups contained three to seven strains, and the rest (8 strains) were distributed among the remaining 7 FAME groups, each consisting of only one or two isolates. Among the isolates, only 55.3 % of the isolates was reasonably identified to the species level and the rest could not be identified due to their poor match to any profile within the MIDI (Microbial ID, Inc.) library and to lack of growth on laboratory medium (Table 3). The dominant species, P. saccharophyla, accounted for 25.5 % of the isolates (i. e., 12 strains, Table 3) and comprised the dominating FAME group with P. paucimobilis. Colony REP PCR analysis. Colony REP PCR was performed to distinguish strains of the same genera and species determined by FAME analysis as well as to analyze the divergence of strains classified as different genera and species. Representative gel photographs are shown in Figure 2. REP PCR analysis of the isolates revealed 30 different DNA fingerprint patterns. Among the 16 strains belonging to the dominating FAME group, two isolates (1443 and 556) identified as P. paucimobilis by FAME analysis exhibited 29 identical REP PCR patterns, thus indicating that they are the same strains in a single species. Identical PCR patterns were also observed among strains 765, 1124, and 1136, among 9 isolates (936, 947, 957, 1146, 1165, 91462, 9174, 9247, and 9256) classified as P. saccharophyla, and between strains 1173 and 9157 identified as AIcaligenes eutrophus. Among the unidentifiable isolates, strains 583 and 773, strains 9136, 91461, and 9166, and strains 9236 and 9266 produced identical REP PCR patterns, respectively. Since these strains were isolated either at different times or from different plots, their detection frequencies reflect the extent of their dominance in the examined plots. From this point of view, the strain of P. saccharophyla which was detected from 1990 to 1992 and from five different plots seemed to be most widely distributed on the plots during this research. Among Pseudomonas species, the isolates of P. paucimobilis produced very similar PCR patterns to those of P. saccharophyla(Figures 2-A, lanes 1 - 6, 8 - 14, 2-B, lanes 18, 19, 2- C, lanes 40, 44, 45, 47), while both of their PCR patterns were quite distinct from other strains (Figures 2-A, lane 7, 2-B, lanes 22 - 24, 2-C, lane 46) classified as Pseudomonas species by FAME analysis. This observation is well correlated to FAME data in which the isolates of P. paucimobilis and P. saccharophyla were grouped together into one FAME group at a Euclidian distance of 6.4, whereas the other Pseudomonas strains were grouped into several more diverse FAME groups clustering at a Euclidian distance of 31.5. The latter groups, together with Alcaligenes species (Figures 2-B, lanes 21, 25, 26, 30, 2-C, lane 42) and most of the unidentifiable strains (Figures 2-A, lanes 15, 16, 2-B, lanes 20, 28, 29, 31, 2-C, lanes 33 - 39, 41, 43) exhibited patterns distinct from each other, which is also reflected in the FAME dendrogram (Figure 1). 30 Genetic diversity analysis. To investigate the sequence diversity of the genes involved in 2,4-D degradation and to examine the presence of 2,4-D degradative plasmid DNA, the isolates were subjected to the modified Kado's plasmid detection method. Chromosomal DNA and plasmid DNA (if any), after gel separation and Southern transfer, were hybridized to the tfd genes of the plasmid pJ P4 and also to a 6.5 kb DNA probe which is specific for a certain group of 2,4-D degrading bacteria such as P. paucimobilis 1443 undetectable with the known tfd genes. Representative hybridization blots are shown in Figure 3. Based on their hybridization patterns, the 47 isolates were grouped into four groups (Table 4) : hybridization group I included 12 isolates having sequence homology with all of the four tfd genes used ; group 11 contained 3 isolates having homology only with the tfdA gene among the four genes ; group III contained 18 isolates showing homology only with the 6.5 kb probe ; group IV (all others) contained 14 isolates which did not hybridize with any of the probes used. All of the isolates classified into the group I exhibited clear plasmid bands on the gel and most of the detected plasmids (75 %) had sequence homology with the tfd genes (Figures 3-A, lane 5, 3-B, C &D, lanes 9, 10). Although plasmids ranging fiom 2.6 to 350 Mdal were reported to be easily detected by the Kado's method (21), shearing of large plasmids seemed to occur during the procedure as shown by minor hybridization signals observed on the chromosomal and linear DNA band in addition to main signals on plasmid DNA. In some isolates such as 9136, 91461, and 9166, the plasmid DNA was clearly detected on the gel, but it did not hybridize to any of the tfd genes used ; instead the chromosomal DNA gave hybridization signals (Figure 3-A, lanes 2, 3, 6). It seemed that these isolates have 2,4-D degradative genes on the chromosome instead of on the plasmid, although it is in contrast to most 31 of the earlier reports (10, 13, 25) that all or most of the 2,4-D genes are contained on the independent plasmid DNA. However, this observation is not surprising since it has been recently observed that an entire 2,4-D degradative plasmid could integrate into (strain 28110, Figure 3-B, lane 8) and excise fi'om the chromosome in Alcaligenes sp. (19). Alternatively, they might have huge plasmids that behave like chromosomal DNA during the procedure. Another interesting strain in group I is 9112 (Figures 3-B, C & D, lane 14). Its plasmid DNA detected on the gel had sequence homology with the tfdA and tde genes but, with the tde and tde probes, the major hybridization signals were shifted to the chromosomal DNA band. It is likely that the genes corresponding to the tfdA and tde probes are contained on one plasmid, while the genes for the tde and tde are contained either on another large plasmid or on the chromosome. The 12 isolates belonging to group I clustered at a Euclidian distance of 31.5 in FAME analysis and most of their REP PCR patterns (Fig. 2-A, lane 7, 2-B, lanes 20, 23, 25 - 31) were quite different from each other, suggesting that they are not closely related. The isolates classified into group II accounted for only 6.4 % of the total isolates, indicating that they are not the dominant species in the examined plots. Their hybridization patterns were more unique than that of isolate 91 12 of group I in that they do not have sequence homology with the tde, tde, and tde genes, suggesting the diversification of the corresponding enzymes or pathways. These isolates were observed to have 2,4-D degradative, transferable plasmids (19). Although isolates 712 and 782 were identified as the same species, P. pickettii, by FAME analysis (Table 3) and observed to have the same plasmid (data not shown), their REP PCR patterns were quite distinct from each other (Figure 2-B, lanes 22, 24). This observation indicates that their 2,4-D degradative plasmids were obtained 32 through natural conjugation among the indigenous soil microorganisms. Isolate 2811P, identified as Alcaligenes paradoxus, was also observed to have nearly identical plasmid to those of isolates 712 and 782. The possible intergeneric gene transfer between isolates 281 IP and 712 has been discussed in the previous paper (19). Strain 28110 (a derivative strain from isolate 2811P) having the entire plasmid integrated into the chromosome (19) is included in Fig. 3-B (lane 8). The isolates classified to group III did not hybridize with any of the tfd genes but hybridized with the 6.5 kb probe (Figure 3-A & E, lanes 1, 4, 7), suggesting that the degradative enzymes (or pathways) of this group are different fi'om those of groups I and II. Although plasmid DNA band was not detected from most of the isolates (83.3 %) of group III with the Kado's method, they seem to have huge plasmids which behave like chromosomal DNA since one isolate (strain 1443) was observed to have a huge plasmid (data not shown) with the direct lysis method (29). Hence, the hybridization signals detected on the chromosomal area from these isolates may have been derived from their broken large plasmids as was observed from the isolate 1443. The isolates of group III clustered at a Euclidian distance of 6.4 in FAME analysis and their REP PCR patterns were very similar or identical to each other (Figures 2-A, lanes 1 - 6, 8 - 16, 2-B, lanes 18, 19, 2-0, lanes 40, 44, 45, 47 ), suggesting that they are closely related to each other. FAME analysis identified them as either P. paucimobilis or P. saccharophyla, but REP PCR analysis revealed six different patterns among them. The 14 isolates of group IV (all others) did not have sequence homology with the DNA probes used (Figures 3-B, C, D & E, lanes 11, 12, 13) and clustered at a Euclidian distance of 31.5 in FAME analysis (Figure 1). Only two isolates, 912 and 983, among them were identified as Alcaligenes faecalis 33 and Pseudomonas pickettii, respectively, but the rest could not be identified by FAME analysis. These isolates appeared not to be closely related to each other as indicated by their quite distinctive REP PCR patterns (Figure 2-0, lanes 33 - 39, 41 - 43, 46). Degradative diversity analysis. Representative strains of the isolates identified as the same species by FAME analysis were grown on 2,4-D or acetate as the sole carbon source and then examined for their degradation capabilities of other 2,4-D related compounds (Table 5). The isolates selected from the hybridization group I were more versatile in substrate utilization ability than those from the other groups. Especially, isolate 965, identified as P. solanacearum by FAME analysis, was (Table 3) the most versatile strain among the isolates tested. Whether it was grown on 2,4-D or acetate, it vigorously utilized 2,4-D related compounds such as PA, 4-0PA, 3-CB, and MCPA as the sole carbon sources as indicated by the complete disappearance of the substrate absorption spectrum and by the substantial cell growth (Table 5). Isolates 1173 and 745 of the hybridization group I could degrade MCPA or 3-CB, respectively. On the other hand, the selected isolates belonging to the groups II, III, and IV were very stringent in substrate utilization ability. None of them could degrade any of the compounds examined when they were grown on acetate (Table 5). The degradation of 4- CPA by isolates 912, 1443, 1156, 91462, and 1173 only under 2,4-D adapted condition indicates that this compound was metabolized probably due to its structural similarity to 2,4-D or 2,4-D pathway intermediates. None of the selected isolates could degrade 2-0PA and 4-CB. Distribution of 2,4-D degraders. The distribution of hybridization groups (group I, II, III, and IV) was influenced by 2,4-D application rates and repeated treatments of 2,4-D with time (Figure 4). The isolates of groups II 34 and IV were encountered more frequently from soils not treated with 2,4-D or treated at a low rate (1 ppm), whereas those of groups I and III preferred soils treated at high rates (10 and 100 ppm) (Figure 4). This is well indicated by the decreasing ratios of the populations of the former to the latter with the increasing rates of 2,4-D application : i. e., the ratios over the experiments were 6, 1, 0.33, and 0.15 in control soil and soils treated with 1, 10, and 100 ppm of 2,4-D, respectively. At the early phase of the field experiments, the isolates belonging to the groups 11 and IV were detected as predominant 2,4-D degraders in most of the plots. For example, these isolates accounted for 87.5 % of the predominant species isolated from the soil samples of May, 1990. However, the frequency of their occurrence was decreased to 50 % and 37.5 % in July and September of 1990, respectively. Thereafter, with further repeated treatments of 2,4-D, they were rarely detected. In contrast, those of the groups I and III occurred at lower frequency at the early phase but were encountered more frequently with the increasing number of 2,4-D treatments and eventually they accounted for most or all of the predominant species detected in 1991 and 1992. DISCUSSION The diversity of predominant 2,4-D degrading bacteria isolated at different times from an agricultural soil not treated or treated with 2,4-D at difi‘erent rates was analyzed by using different classification systems. The relationships between different grouping methods are summarized in Table 6. The 47 isolates of 2,4-D degrading bacteria exhibited high diversity in their FAME profiles (Figure 1), REP PCR patterns (Figure 2), hybridization 35 patterns (Figure 3), and phenotypic traits (Table 5). While a variety of 2,4-D degrading bacteria belonging to multiple genera were reported in 19508 and 19608, species identification by FAME analysis placed 55.3 % (26 strains) of the isolates in the genus Pseudomonas or Alcaligenes in this study. The rest (21 strains) could not be identified by FAME analysis mainly due to their poor match to any profole of the MIDI library and to lack of growth on ordinary laboratory medium. The REP PCR analysis showed these unidentifiable isolates to be heterogenous, giving 16 distinctive patterns, 4 of which occurred more than once. It is also interesting that most of the isolates (85.7 %) belonging to the hybridization group IV, which included all isolates not showing any detectable homology to the DNA probes used, could not be identified. The isolates of the groups II and IV degrade 2,4-D very slowly and are not versatile in capabilities to utilize 2,4-D related compounds as sole carbon sources (Table 5). They appeared not to be competitive in soils treated with a high rate of 2,4-D or repeatedly treated with 2,4-D over a year period, since their proportion among the isolates was low or decreased according to the respective conditions. This could influence the type of 2,4-D degrading bacteria to be isolated from environmental samples by using the enrichment culture technique. Since, in most of studies, 2,4-D degrading bacteria have been isolated by inoculating undiluted environmental samples to broth culture with a high concentration of 2,4-D and through repeated transfers to fresh medium, the type of obtained isolates could be biased. In this situation, generally, strains belonging to the hybridization groups I are preferably isolated due to their quick growth property, which has already been demonstrated and discussed in soil competition studies (20). This may be the main reason why most of the 2,4-D degradative plasmids independently 36 isolated from numerous AIcaligenes species showed a high degree of similarity to plasmid pJ P4 in biophysical and genetic properties (10) and also why most of the 2,4-D degrading bacteria isolated from various natural water samples exhibited suprising uniformity in their phenotypes, metabolism, and cell structure (1). In this research, the possible bias during enrichment procedure has been avoided by using the highest dilution showing 2,4-D degradation as inoculum for enrichment culture. With this procedure, only the naturally predominant species would be enriched and isolated from the plots without depending on their growth property in laboratory 2,4-D medium. Some of the cultures failed to produce single colonies able to degrade 2,4-D, suggesting that either they were not culturable on laboratory plates or cometabolism was involved in the 2,4-D degradation. 2,4-D is known to be easily utilized by microorganisms able to degrade this compound (24). The presence of 2,4-D degradative genes on plasmids and the extensive use of this compound as a herbicide since the late 19408 have been assumed to stimulate the rapid dissemination of the 2,4-D genes among the indigenous microbial populations (28). Additionally, the studies on 2,4-D degrading bacteria such as Arthrobacter (4, 5, 11, 23, 25, 34) and Alcaligenes (9) species known to be among the predominant species revealed the details of pathways and genes involved in 2,4-D degradation. However, it is worth noticing that still 75 % of the 2,4-D degraders isolated as dominant species during this research do not have sequence homology with all or most of the known tfd genes. It is not likely that gene transfers have extensively occurred among the isolates in the environment, which is further supported by the observation that the isolates of each hybridization group showed a tendency to form a separate cluster in FAME dendrogram instead of being mixed with one another. Moreover, considering that the natural environment 37 usually would not be fi'equently exposed to high doses of 2,4-D such as 10 and 100 ppm, the commonly encountered dominant species in the environment would be the types of the hybridization groups 11 and IV. The ability to metabolize 2,4-D related compounds was quite diverse among the isolates. The selected isolates of group I exhibited relatively versatile property in substrate utilization ability, but those of group II appeared unable to vigorously attack any of the compounds examined and those of group III and IV tended to incidentally attack 4-CPA. While it has been suggested that 2,4-D degrading bacteria isolated from environmental samples commonly degraded 2-0PA, 4-CPA, and MCPA (24), none of the selected isolates could degrade 2-CPA and only 30 % and 60 % of the selected isolates could degrade MCPA or 4-CPA, respectively. It has been reported that soil microflora adapted to 2,4-D could rapidly degrade MCPA and vice versa (2). It was suggested that 2,4-D and MCPA each induced their own specific enzyme systems, either in two different microfloras or in the same microorganisms, and that each enzyme system could incidentally attack the other molecule with less efficiency (2). Torstensson et. al. concluded that two different microfloras were involved in this cross-adaptation, while Audus (2) attributed this to the same organism. However, our results revealed substantial diversification of MCPA metabolism among the isolates. While the 2,4-D-degrading enzyme system of isolate 712 appeared able to incidentally degrade MCPA with less efficiency, which is in agreement with the result of Audus, isolates 965 and 117 3 seemed able to induce their own MCPA-degrading enzyme systems without depending on the previous growth conditions. Moreover, the rest of the selected isolates could not attack MCPA even when they were adapted to metabolize 2,4-D, a finding also observed by MacRae & Alexander (26). 38 The hybridization grouping of the isolates not only reveals the substantial genetic heterogeniety among them but also has usefulness in the interpretation of the ecological evaluation of highly diverse groups of 2,4-D degraders. The observed gradual shifting of the predominant 2,4-D degrading populations from the groups II and IV to the groups I and III in field studies demonstrates how the indigenous microbial populations respond to the environmental changes. Among the four hybridization groups, group III was the most homogenous in population structure, as indicated by the high similarity in FAME profiles, REP PCR patterns, and degradation phenotypes (Table 6). The data presented in this study illustrate the ability of different classification methods to group the 2,4-D degrading isolates and provide new insights into their diversity and ecological dynamics. ACKNOWLEDGMENTS This work was supported by National Science Foundation grants from Long-Term Ecological Research program and the Center for Microbial Ecology. We thank Dr. Peter Burauel for arranging the field plots, Mr. James Bronson for helping with the field experiment, and Miss Helen Garchow for conducting and providing information on the FAME analysis. 39 Table l. 2,4-D degrading bacteria and their source. Rate of 2,4-D Date of soil sample Site Isolate applicationa collectionb Plot 1 2811P control Aug, 1989 Plot4 ' K11, 1443 high Plot 1 & 8 512, 583 control May, 1990 Plot 2 & 7 524, 573 low Plot 3 & 6 535, 565 intermediate Plot 4 & 5 546, 556 high Plot 1 & 8 712, 782 control July, 1990 Plot 2 & 7 723, 773 low Plot 3 & 6 736, 765 intermediate Plot 4 & 5 745, 756 high Plot 1 & 8 912, 983 control Sept, 1990 Plot 2 & 7 924, 974 low Plot 3 & 6 936, 965 intermediate Plot 4 & 5 947, 957 high Plot 2 & 7 1124, 1173 low Nov., 1990 Plot 3 & 6 1136, 1165 intermediate 40 Table 1 (con’d). Plot4 & 5 1146,1156 high Plot 1 & 8 9112, 9182 control Sept, 1991 Plot 7 9174 low Plot 3 & 6 9136, 9166 intermediate 91461, 91462, Plot 4 & 5 9157 high Plot 1 9212 control May, 1992 Plot 2 9224 low Plot 3 & 6 9236, 9266 intermediate Plot 4 & 5 9247, 9256 high a Control : no 2,4-D treatment ; low, intermediate, and high : 2,4—D was applied at the rate of 1 (0.6), 10 (6), 100 (60) jig/g soil (kg/ha), respectively, at the following dates : one application each on October of 1988 and on May, August, October, and December of 1989 ; two applications on May, July, September, and November of 1990 ; on May, July, September, and November of 1991 ; and three applications on May of 1992. b Soil samples for strain isolation were taken in one week following the last application. 41 Table 2. Distribution of isolates among FAME (fatty acid methyl ester) groups. Number of isolates Number of FAME groupsa l6 1 7 1 5 1 4 1 3 1 2 1 l 6 a FAME groups include isolates with < 10 Euclidian distance. 42 Table 3. Identification of isolates to species level using FAME (fatty acid methyl ester) analyses. Isolate FAME identification 1443, 556, 9174, 9224 Pseudomonas paucimobilis 736, 756, 765, 936, 947, 957, 974, 1124, 1136, 1146, l 165, 91462 Pseudomonas saccharophyla 712, 782, 983, 9112 Pseudomonas pickettii 745 Pseudomonas pseudomallei 965 Pseudomonas solanacearum 1173, 9157 Alcaligenes eutrophus 912 Alcaligenesfaecalis 2811P Alcaligenes paradoxus KIl, 512, 524, 535, 546, 565, 573, 583, 723, 773, 924, 1156, 9136, Unidentifiablea 91461, 9166, 9182, 9212, 9236, 9247, 9256, 9266 a These isolates could not be identified due to their poor match to profiles of MIDI library and lack of growth on ordinary laboratory medium. 43 Table 4. Hybridization patterns of the isolates with DNA probes. DNA Probesa-b Hybridization Isolate rfdA tde :de yap 6.5 kb Group K11, 745, 965, 1173. 9112, 9136,91461,9157,9166,9212, + + + + - I 9236,9266 2811P,712,782 + - - - - II 1443, 556, 736, 756, 765, 936 947, 957, 974, 1124, 1136. 1146, 1165, 91462, 9174, - - - - + 111 9224, 9247, 9256 512, 524, 535, 546, 565, 573 583, 723, 773, 912, 924, 983 - - - - - IV 1156. 9182 (all others) a tfdA, Ide, tde, q’dD gene probes were subcloned from plasmid pJP4 ; 6.5 kb probe was subcloned from the large plasmid of strain 1443. b +, shows detectable hybridization signal ; -, does not show detectable hybridization signal. 44 Camumnoe 3. a8: $2886 H a E 2 98388.6 Em com :qu mm: :6 a S 33 03mm mom 3» 2% > C > C > C > c > C > C > C > C > c > C Nab ++++++++++++++++++++ 1+.I..I.++ ++++++++++++ w> - - ++ ++ - - - - - - - - - - - - - - - - who.» - - - - - - - - - - - - - - - - - - - - to.» - - ++ ++ ++ - - - - - ++ - ++ - - - .I. - ++ - uhw + + ++++ - - - - - - - - - - - - - - - - row - - Zoe... s v? era—838838 SE " may? Voaonogoeostoaao some I??? Aéaonogoeoxwwnoao 83 “ ”row. u- 0:583:83 some u how. 5.9888888 83 u 396? VBoEELioEonousgoxwwoono some. a .35 8698 ”858 o». 96 351888: «8:? H. a. E. Ba 2 £68 «8%: o: Psi—u 25 on o: 8888 at can 9o: 88a 8n 8388 3:88: omuchnom. ++. Vmo S 8308: m: Sew saw! ”.88 G< momsasm 25 23:59: «8.5—. 8.38 v “:8 u +. 3 8 8 a. Become: E San ea Bonanza «8&9 AOUmmo vcbwv “ -. A G a» 8230: E For use — ' —_.:.:.-—_..-.-—— H. Fig. 2-B. 50 C 3233343515338394041 4243 4445 4647 _. fl. ,3 4...; Va. .5.“ “—’!;; 3‘"...— Fig. 2-C. Figure 3-A 8 9 1O 11 12 13 52 B Figure 3-B. 14 53 C 8 9 1O 11 12 13 14 . . W” p u chr Figure 3-C. 910111213 14 54 D ‘I Figure 3-D. 55 Figure 3-E. 2,4-D Application Rate (ppm) 100" 10" 56 (GroupI-A, II-CJ, III-l, IV-A) IA IA IA II IA IA II AA II IA II AA AA AA AA IA IA I I U AA CE] AA AA A 8978 9075 7 9' 1'1 91I/7 92'/s TIME Figure 4. LIST OF REFERENCES 57 LIST OF REFERENCES 1. Amy, P. 8., J. W. Schulke, L. M. Frazier, and R. J. Seidler. 1985. Characterization of aquatic bacteria and cloning of genes specifying partial degradation of 2,4-dichlarophenaxyaaetic acid. Appl. Environ. Microbial. 49:1237- 1245. 2. Audus, L. J. 1964. Hrebicide behaviour in soil, p. 163-206. In L. J. Audus (ed. ), The physiology and biochemistry of herbicides. Academic Press, London. 3. Bell, G. R. 1957. Some morphological and biochemical characteristics of a sail bacterium which decomposes 2,4-dichlaraphenaxyacetic acid. Can. J. Microbial. 3:821-840. 4. Bollag, J. -M., C. S. Helling, and M. Alexander. 1968. Enzymatic hydroxylation of chlorinated phenols. J. Agr. Food Chem. 16:826-828. 5. Bollag, J. -M., G. G. Briggs, J. E. Dawson, and M. Alexander. 1968. Enzymatic degradation of chlorocatechols. J. Agr. Food Chem. 16:829-833. 6. Bounds, H. C., and A. R. Colmar. 1965. Detoxication of some herbicides by Streptomyces. Weeds 13:249-252. 17 . do Bruiin, F. J. 1992. Use of repetitive (repetitive extragenic palindromic and entrabacterial repetitive intergeneric consensus) sequences and the polymerase chain reaction to fingerprint the genomes of Rhizobium meliloti isolates and other soil bacteria. Appl. Environ. Microbial. 58:2180-2187. 8. Don, R. H., and J. M. Pemberton. 1985. Genetic and physical map of the 2,4-dichlaraphenaxyacetic acid-degradative plasmid pJ P4. J. Bacterial. 161:466-468. 9. Don, R. H., A. J. Weightman, H. -J. Knackmuss, and K. N. Timmis. 1985. Transposon mutagenesis and cloning analysis of the pathways for degradation of 2,4-dichlarophenaxyacetic acid and 3-chlarobenzoate in Alcaligenes eutraphus JMP134 (pJP4). J. Bacterial. 161:85-90. 10. Dan, R. H., and J. M. Pemberton. 1981. Properties of six pesticides degradation plasmids isolated from Alcaligenes paradoxus and Alcaligenes eutraphus. J. Bacterial. 145:681-686. 58 11. Duxbury, J. M., J. M. Tiedie, M. Alexander, and J. E. Dawson. 1970. 2,4-D metabolism: Enzymatic conversion of chloromaleyiacetic acid to succinic acid. J. Agr. Food Chem. 18:199-201. 12. Evans, W. C., B. S. W. Smith, H. N. Fernley, and J. 1. Davies. 1971. Bacterial metabolism of 2,4-dichlaraphenaxyacetate. Biochem. J. 122:543- 551. 13. Fisher, P. R., J. Appleton, and J. M. Pemberton. 197 8. Isolation and characterization of the pesticides-degrading plasmid pJ P1 from Alcaligenes paradoxus. J. Bacterial. 135:798-804. 14. Fournier, J. C. 1980. Enumeration of the sail micro-organisms able to degrade 2,4-D by metabolism or ca-metabalism. Chemosphere 9: 169-174. 15. Gerhardt, P., R. G. E. Murray, R. N. Costilow, E. W. Nestor, W. A. Wood, N. R. Krieg, and G. B. Phillips. 1981. Manual of methods for general bacteriology. American Society for Microbiology, Washington, DC. 16. Holben, W. E., B. M. Schroeter, V. G. Mathesan, R. H. Olsen, J. K. Kukor, V. O. Biederbeck, A. E. Smith, and J. M. Tiedie. 1992. Analysis by gene probes of soil microbial populations selected by treatment with 2,4-dichloraphenaxyacetic acid (2,4-D). Submitted for publication. 17 . Holben, W. E., J. K. Jansson, B. K. Chelm, and J. M. Tiedie. 1988. DNA probe method for the detection of specific microorganisms in the soil bacterial community. Appl. Environ. Microbiol. 54:703-711. 18. Jensen, H. L., and H. 1. Peterson. 1952. Detoxication of hormone herbicides by soil bacteria. Nature 17 0:39-40. 19. Kn, J. 0., and J. M. ’l‘iedie. 1992. Integration and excision of a 2,4-D degradative plasmid in Alcaligenes paradoxus and evidence of its natural intergeneric transfer. Submitted for publication. 20. Ka, J. 0., W. E. Holben, and J. M. 'l‘iedie. 1992. Gene probe analysis of competition among 2,4-D degrading bacteria in soil under selective conditions. Submitted for publication. 21. Kado, C. 1., and S. -T. Liu. 1981. Rapid procedure for detection and isolation of large and small plasmids. J. Bacterial. 145:1365- 1373. 22. Laos, M. A., I. F. Schlasser, and W. R. Mapham. 197 9. Phenoxy herbicide degradation in sailszQuantitative studies of 2,4-D and MCPA- degrading microbial populations. Sail Biol. Biochem. 1:37 7-385. 59 23. Laos, M. A., J. -M. Bollag, and M. Alexander. 1967 . Phenoxyacetate herbicide detoxication by bacterial enzymes. J. Agr. Food Chem. 15:858- 860. 24. Laos, M. A. 1975. Phenoxyalkanoic acids, p. 1-56. In P. C. Kearney, and D. D. Kaufman (ed.), Herbicides: Chemistry, Degradation and Made of Action. Marcel Dekker Inc., New York. 25. Laos, M. A., R. N. Roberts, and M. Alexander. 1967. Phenols as intermediates in the decomposition of phenoxyacetates by an Arthrobacter species. Can. J. Microbial. 26. MacRae, I. C., and M. Alexander. 1965. Microbial degradation of selected herbicides in soil. J. Agric. Food Chem. 13:72-76. 27 . Maniastis, T., E. F. Fritsch, and J. Sambrook. 1982. Molecular cloning: a laboratory manual. Cold Spring Harbor Laboratory, Cold Spring Habar, NY. 28. Pemberton. J. M., B. Carney, and R. H. Dan. 1979. Evolution and spread of pesticide degrading ability among sail micro-organisms, p. 287- 299. In K. N. Timmis, and A. Puhler (ed.), Plasmids of medical, environmental and commercial importance. Elsevier/Narth-Halland Biomedical Press. 29. Plazinski, J., Y. H. Gen, and B. G. Rolfe. 1985. General method for the identification of plasmid species in fast-growing sail microorganisms. Appl. Environ. Microbial. 48:100 1- 1003. 30. Rogofl, M. H., and J. J. Reid. 1956. Bacterial decomposition of 2,4- dichlarophenaxyacetic acid. J. Bacterial. 71:303-307. 31. Sasser, M. 1990. Technical Note #102: Tracking a strain using the Microbial Identification System. 32. Sinton, G. L., L. T. Fan, L. E. Erickson. and S. M. Lee. 1986. Biodegradatian 0f 2,4—D and related xenabiatic compounds. Enzyme Microb. Technal. 8:395-403. 33. Stanier, R. Y., N. J. Palleroni, and M. Doudorofi. 1966. The aerobic pseudamanads: a taxonomic study. J. Gen. Microbial. 43:159-271. 34. 'l‘iedie, J. M., J. M. Duxbury, M. Alexander, and J. E. Dawson. 1969. 2,4-D metabolism: Pathway of degradation of chlorocatechols by Arthrobacter sp. J. Agric. Food Chem. 17 :1021-1026. Chapter Three Use of Gene Probes to Detect 2,4-D Degrading Populations in Soil Maintained under Selective Pressure. J. 0. KA, W. E. HOLBEN and J. M. TIEDJE*. Center for Microbial Ecology and Department of Microbiology and Public Health, Michigan State University, East Lansing, MI 48824 60 61 ABSTRACT 2,4-D was applied to soils in microcosms and degradation was monitored for each of 5 repeated additions. Total DNA was isolated from the soil bacterial communities after each 2,4-D treatment. The DNA samples were digested with restriction enzymes, and analyzed on Southern blots, using a tfdA gene probe subcloned from plasmid pJP4 and also a 6.5 kb probe derived from the large plasmid of a new 2,4-D degrading isolate (Pseudomonas paucimobilis) which has no detectable cross-homology with the WA gene probe. 2,4-D applied to soil with no prior history of 2,4-D treatment was quickly degraded by indigenous soil microbial populations. There was good correlation between 2,4-D degradation and banding patterns in hybridization analyses performed after each 2,4-D treatment using both the tfdA gene probe and the 6.5 kb probe. In one microcosm, where soil water content was adiusted to 10% initially, changes in banding patterns over time were observed with the tfdA gene probe in Southern blot analyses, suggesting that population changes or possibly genetic rearrangement in 2,4-D degrading microbial populations occurred in this soil. In another microcosm adjusted to 25% water content, 2,4-D degrading populations were maintained stably throughout the 5 additions. Isolates responsible for hybridization patterns observed in the total community DNA were identified. The data demonstrate that gene probe analyses were capable of detecting and discriminating indigenous 2,4-D degrading populations in soil amended with 2,4-D. 62 INTRODUCTION Large amounts of man-made chlorinated organic chemicals have been used extensively in agriculture as herbicides and pesticides. Among these, 2,4- dichlorophenoxyacetic acid (2,4-D) has received widespread use as a herbicide for more than 40 years. Unlike many of the recalcitrant synthetic compounds that have been released into the environment over the past several decades, 2,4—D is rapidly degraded by soil bacteria (4, 13, 29, 39, 40) and its fate in soil is critically affected by microbial degradation (13, 25). Populations of microorganisms able to degrade 2,4-D have been estimated by a most probable number (MPN) method in soils (7, 19, 26, 29, 35). Organisms capable of degrading 2,4-D have been identified in a number of genera such as :Achromobacter, Alcaligenes, Arthrobacter, Corynebacterum, Flavobacten'um, Pseudomonas, and Streptomyces (2, 8, 11, 23, 28, 34, 41). Furthermore, the pathway for 2,4-D degradation on plasmid pJP4 originally identified inAlcaligenes eutrophus was studied and described (8, 9, 38). However, past research describing herbicide biodegradation by microbial populations has provided virtually no information regarding population or genetic changes of microbial communities in natural environments in response to the herbicide. Traditionally, the isolation of pure cultures (6) and the fluorescent antibody techniques (5,12) have been used to confirm the presence of specific microorganisms in environmental samples, each of which is useful but limited in some aspects. The cultural method is selective for certain microorganisms which can be cultivated on or in particular laboratory media. The fluorescent antibody technique has high specificity even to the strain level, but is labor-intensive, relatively expensive and is not well-suited to routine screening for the presence of organisms. As a new methodological 63 approach, gene probes were developed to detect specific microorganisms in environmental samples by using colony hybridization (18, 36). But this method requires that microorganisms be cultivated in laboratory media prior to analysis. Since more than 90% of soil microorganisms are non-culturable in laboratory media, the colony hybridization method would have limited utility for detecting these non-culturable microorganisms in soils. It has been suggested that a DNA probe method for analyzing total soil bacterial DNA might make it possible to detect multiple organisms and populations or genetic changes in natural environmental samples (21, 22 ). However, in those reports, particular gene probes were used to detect microorganisms inoculated into soils in large numbers under laboratory conditions rather than the indigenous soil microfloras. Moreover, little attention has been given to the population dynamics and genetic changes of the soil microfloras in response to environmental changes. It is likely that when the involved genes of the indigenous microbial populations are diversified, it is dificult to detect and assess all of the responsive, diverse populations with a single gene probe. The diversity of 2,4-D degradative genes was described in the previous study (24) where numerous indigenous 2,4-D degrading microorganisms did not show any cross-homology with the tfdA gene which is specific in the 2,4—D degradative pathway. Among the four hybridization groups described in the study, the dominating group III accounted for about 38 % of the total 47 2,4-D degrading bacteria isolated from Kellogg Biological Station (KBS) soils. Since the dominating populations did not have any sequence homology with the tfdA gene (as well as with the tdeCD genes), a new probe, a 6.5 kb fragment, has already been developed to discriminate them from others undetectable with the tfdA probe (24). These two DNA probes, the tfdA gene probe and the 6.5 64 kb probe, are able to account for about 70 % of the indigenous 2,4-D degrading populations in KBS soil and the isolates of the hybridization groups I and III, which can be detected with the tfdA gene probe or the 6.5 kb probe, respectively, dominated in KBS soil repeatedly treated with high amount of 2,4-D (24). Hence, the two DNA probes appear to make it possible to monitor the population dynamics and genetic changes of the indigenous, dominant 2,4-D degrading populations in KBS soil under 2,4-D selective condition. In this research, the objectives were to detect indigenous 2,4-D degrading microbial populations in soil using DNA probes to understand the genetic response of the microbial community to 2,4-D selective conditions, to identify specific 2,4-D degrading isolates responsible for hybridization patterns observed in the total soil community DNA, and to demonstrate that a single gene is not enough to monitor all of the 2,4-D degrading microbial papulations in soil due to the diversity of the 2,4-D degradative genes. MATERIALS AND METHODS Bacterial strains and media. The strains of Pseudomonas pickettii and Pseudomonas paucimobilis capable of utilizing 2,4-D as the sole source of carbon were isolated from agricultural soil samples in Long-Term Ecological Research (LTER) site of Kellogg Biological Station (KBS, Hickory Corners, MI) (24). Escherichia coli JM83 is described elsewhere (3). Strains P. paucimobilis and P. pickettii were cultivated in MO mineral media (37) containing 500 ppm of 2,4-D at 30°C. E. coli JM83 was cultivated in Luria broth (30) at 37°C. Plasmid DNA was isolated by using the method of Hirsch et. al. for large plasmids (19) or Maniatis et. al. for small plasmids (30). 65 2,4-D amendment of soil. Soil (loam) with no history of 2,4-D treatment was taken from the LTER site of KBS, on which a plat has been subjected to microbial ecological research in response to 2,4-D treatment as one of the LTER projects. Soil was stored at field moisture levels at 4°C until used. Soil samples (500g) which had been sifted through a 2-mm sieve were transferred to sterile polyethylene wide-mouth bottles. Soil water content was adjusted to 10% for two of four bottles and 25% for the other two by adding sterile, distilled water. 2,4-D was added to one of the 10% moisture soils and one of the 25% moisture soils to a concentration of 250 ppm and thoroughly mixed. Each control sail received only phosphate buffer. The disappearance of 2,4-D in soil was monitored as described above and the soil respiked for each of 5 cycles of degradation. At the end of each cycle of degradation, a 50 g subsamples were taken from both the 2,4-D-treated and control soils and the total soil bacterial DNA was extracted with cell extraction method which involves the separation of the bacterial cells from the soil particles by differential centrifugation followed by lysis of the recovered cells (2 1). Analysis of 2,4-D concentration in soil. Stock solutions (20 mg/ml) of analytical grade 2,4-D (Sigma, St. Louis, M0) were prepared by dissolving in 0.1M NaH2P04 buffer (pH 7.0) and stored at 4°C until use. For analysis of the concentration of 2,4-D in soil, 1 g soil subsamples were combined with lml of sterilized distilled water in an eppendorf tube and vortexed for 1 min. The soil was pelleted by centrifugation in a microfuge (14,000 rpm, 5 min) and the supernatant was transferred into a clean eppendorf tube. This sample was filtered through a Millipore Millex-GS syringe filter and analyzed for 2,4-D with an Hewlett Packard series 1050 HPLC equipped with 66 Lichrosorb RP—18 column (Anspec Co., Ann Arbor, MI) and a UV detector set at 230nm, using methanol/0. 1% phosphoric acid (60:40) as eluant. Probe construction. To detect 2,4-D degrading microbial populations, the tfdA gene was chosen as a probe among the six structural genes (tfdA, B, C, D, E and F) for 2,4-D metabolism of the well-known plasmid pJP4 ( 1, 10), because it would be apparently more specific to the 2,4-D degradation pathway of plasmid pJP4 than the other genes (17 , 19, 24). The tfdA gene probe was cloned as a 556 bp EcaR 1-BamH1 fragment into the multiple cloning site of plasmid pUC 19 (30). A 6.5 kb fragment, which was cloned as a BamHI fragment from the large plasmid of P. paucimobilis into plasmid pUC 19, was used as a new probe to detect P. paucimobilis and similar 2,4-D degrading bacteria that do not hybridize to the tfdA gene probe due to the diversity of 2,4-D degradative genes (24). All restriction enzymes and T4 DNA ligase were purchased commercially and used according to the manufacturer‘s specifications. The cloned tfdA gene probe and the 6.5 kb probe were isolated as restricted fragment from the corresponding cloned plasmid pUC19 on 0.7% agarose gel, purified with Geneclean Kit (Bio 101 Inc, La J olla, CA), and labeled with 32P by using the Random Primed DNA Labeling Kit (Boehringer Mannheim, Indianapolis, IN) or the Nick Translation Kit (Boehringer Mannheim) according to manufacturer's specifications. Labeled probes were separated from unincorporated nucleotides prior to use with spun column (30). Restriction analysis and Southern hybridization. Total soil bacterial DNA was digested with appropriate restriction endonucleases according to manufacturer’s specifications. DigeMd DNA was size fractionated by 67 electrophoresis through horizontal 0.7 % agarose gels and transfered to nitrocellulose hybridization membrane. Prehybridization, hybridization and post-hybridization washes were performed as described by Holben et. al. (21) with some modifications. The membranes were prehybridized for at least 24 h at 42°C and after hybridization, three washes were carried at room temperature, followed by one wash at 65°C. Hybridization signals were detected using the Betascope radioactive blot analyzer (Betagen Corp., Waltham, MA) or by autoradiography using Kodak X-Omat AR film (Kodak, Rochester, NY) exposed at -70°C with a Quanta III (Sigma, St. Louis, MO) intensifying screen. Exposure times were 1 to 7 days depending on the intensity of the radioactive signal. RESULTS Degradation of 2,4-D in soil. Exposure of 2,4—D degrading populations of soil to 2,4-D for the first time resulted in slow degradation for first treatment of 2,4-D in both 10% and 25% moisture microcosm soils (Figure. 1), requiring about 3 weeks for the added 2,4-D to be degraded completely. The rate of degradation for subsequent additions of 2,4-D was greatly increased, requiring 1 week or less for the complete degradation for each of four more . 2,4-D additions. Gene probe hybridization results. The results of Southern hybridization of the tfdA gene probe are presented in Figure 2. For both the 10% and the 25% moisture soils, a single band of hybridization was detected from the second through the fifth addition. No detectable hybridization was observed after the first treatment. In the 10% soil moisture microcosm, the location of 68 the band of hybridization shifted between the second and third treatments and thereaflzer remained constant. A possible interpretation of this data is that the dominant 2,4-D degrading population present after two treatments was displaced by another 2,4-D degrading population that then stably dominated throughout the course of the experiment. Alternatively, it is possible that the sequence encoding the tfdA homology was somehow rearranged (e.g. by deletion) resulting in a shift in the band of hybridization and that this derivative strain displaced the original population possibly because of its increased fitness. In the 25% soil-moisture sample the band of hybridization was first detected after 2 treatments with 2,4-D and its location remained constant throughout the experiment. Note that the apparent size (8.8kb, Figure 2-B) of this band is similar to the size of the band observed after the shift observed in the 10% soil-moisture samples. A total of 47 2,4-D degrading microorganisms were isolated from KBS soil by serial dilution of soil suspensions and selection for growth on 2,4-D (24). All isolates were screened for homology with the tfdA gene probe. Among them, fifteen isolates including strain P. pickettii hybridized to this probe and, through another prescreening hybridization experiment, P. pickettii was expected to correspond to the hybridization bands of microcosm soil bacterial DNA. Plasmid DNA obtained from this strain was compared to total soil bacterial DNA of the fifth treatment of 2,4-D in Southern analyses using the tfdA gene probe (Figure 3). Matching band patterns were obtained between these two DNA samples when digested with three different restriction enzymes, suggesting that the dominant 2,4-D degrading organism responsible for the hybridization signals observed in soil bacterial DNA from the third to the filth treatment of this microcosm experiment could be P. pickettii. Moreover, its population level was increasing with repeated additions of 2,4- 69 D through the fifth treatment, as indicated by increasing band intensifies on Southern blot (Figure 2-A). Since each lane contained the same amount of total DNA (1.5ug), the intensity of the hybridization for each time point is proportional to the amount of target DNA in each DNA sample. Quantitative hybridization analysis using the Betascope radioactive blot analyzer indicated that the hybridization signal exhibited a 5-fold increase when the third 2,4-D treatment is compared to the fifth treatment, suggesting a five fold increase of the target sequence in the degrading population detected (presumably strain P. pickettii). In the 25% moisture soil the same strain P. pickettii population seemed to have been dominant from first detection (after 2 treatments with 2,4-D) through the fifth treatment, since a single band of hybridization that is same size as in the 10% moisture sample persists (Figure 2-B). For both the 10% and the 25% moisture control soils not treated with 2,4-D, the total bacterial DNA had no detectable hybridization to the tfdA gene probe (Figures 2-A & B). ' Strain P. paucimobilis is one of the 2,4-D degrading isolates from KBS soil. The 6.5 kb fragment was developed as a new probe to detect P. paucimobilis in KBS soil amended with 2,4-D. This was necessary because this strain, which was one of the dominant 2,4-D degrading bacteria in both these experiments and in field experiments, has no detectable homology to the tfdA gene probe. The 6.5 kb fragment was also observed to hybridize to other similar 2,4-D degrading populations that are not detectable with the tfdA probe (24). Total soil bacterial DNA from the 25% moisture soil was digested with HindIII, then hybridized with the 6.5 kb fi'agment (Figure 4). Throughout the 5 treatments of soil with 2,4-D, two bands (5.0 kb and 11.3 kb in size) of hybridization were detected with the 6.5 kb probe. The position of the hybridization bands remained constant throughout the experiment, 70 indicating that they represent a single, stably maintained population. When total soil bacterial DNA from the fifth treatment with 2,4-D was compared to DNA isolated from P. paucimobilis, the observed hybridization band patterns between these two DNA samples were identical in Southern analyses using the 6.5 kb probe (Figure 5). This suggests that P. paucimobilis corresponded to the detected hybridization bands from soil bacterial DNA and that this strain was one of the dominant 2,4-D degraders throughout this experiment. Quantitative hybridization analysis indicates that this population was increased about two-fold in size throughout the course of 5 treatments with 2,4-D. Again, there was no hybridization signal to the total DNA of control soil not treated with 2,4-D. DISCUSSION 2,4-D applied to soil with no prior history of 2,4-D treatment was quickly degraded by indigenous 2,4-D degrading microbial populations. Lag phase was observed in the first treatment of 2,4-D, indicating that during which 2,4- D degrading microbial populations of soil were adapting to a new selective conditions and inducing their 2,4-D degradative enzymes as suggested by Loos(27). The hybridization results presented in this report show that both the tfdA gene probe and the 6.5 kb probe were useful for selective detection of 2,4-D degrading microorganisms in natural soil amended with 2,4-D. Good correlation was observed between 2,4-D degradation activities and hybridization signals with these two probes and no hybridization signal was observed in control soil not treated with 2,4-D. It is known that the fate of 2,4-D applied to soil is critically affected by microbial degradation (13, 25, 27) 71 and that 2,4-D treatment of soil significantly increases the number of 2,4-D degrading organisms (14,15). The enrichment effect of 2,4-D an the number of degrading microorganisms made it possible to detect these organisms using DNA probe method which has a lower sensitivity limit of about 104 cells per g (21) without requiring amplification of target sequences or probe signal. Traditionally, numbers of 2,4-D degrading microorganisms in soils have been estimated by the most probable number enumeration method (MPN) (20, 26, 29, 31). This MPN method has restricted usage because it requires successful cultivation of indigenous 2,4-D degraders and only indicates the presence or absence of 2,4—D degraders which can be cultivated. On the other hand, the DNA probe method allows us to study which species is predominant and whether or not population and genetic changes occur among 2,4-D degrading microorganisms, in addition to the relative population levels, in relation to their environment without cultivation of the recovered microorganisms. The use of this DNA probe method in conjuntion with Southern transfer is analytically powerful, being able to detect multiple organisms, population changes, deletion of sequences, and genetic rearrangements (21, 22). The observed changes in band patterns with tfdA gene probe (Figure 2-A) and the co-existence of the apparently dissimilar 2,4- D degrading strains P. pickettii and P. paucimobilis in the same microbial community under 2,4-D selective conditions (Figures 2 & 4) indicates that there may be population fluctuations or dynamic interactions between populations that would be undetectable by the MPN method. Strain P. paucimabilis, which appears to be one of the dominant 2,4-D degraders in both experiments, has no detectable homology to the tfdA gene probe, suggesting that the 2,4-D degradative genes in this strain are substantially difl‘erent at the DNA sequence level and possibly at the 72 functional level. In contrast to the tfdA gene probe, #210 and tde gene probes subcloned from the plasmid pJP4 did not hybridize to the plasmid DNA of P. pickettii(24). In fact, in both microcosms experiments, no hybridization signal was observed with the tde and tde gene probes on Southern analyses. This observed diversity of the 2,4-D degradative genes and the results presented here demonstrate that all of the indigenous 2,4-D degrading microbial populations can not be monitored with a single probe in soil. Among the 2,4-D degrading isolates, some strains (P. pickettii and P. paucimabilis) were found to account for bands of hybridization obtained on Southern blots probed with the tfdA gene probe and the 6.5 kb probe (Figures 3 & 5). Digestion with three different restriction enzymes and hybridization to specific probes gave good evidence that hybridization bands detected from total bacterial community DNA correspond to a particular soil isolate. It is desirable to relate environmental soil DNA samples to laboratory pure DNA of specific isolate in microbial ecological studies and thus it is a good demonstration that DNA probe method can discriminate a strain among a variety of indigenous microorganisms in soil. While it is possible that there were additional 2,4-D degrading populations present that were not detected by either probe, these data demonstrate that DNA probe method with total soil bacterial DNA was capable of detecting and discriminating the indigenous 2,4-D degrading microbial populations in soil under selective conditions and that it could represent a means to study microbial ecology both autecologically and synecologically at the DNA level. 73 ACKNOWLEDGMENTS This work was supported by National Science Foundation grants from Long-Term Ecological Research program and the Center for Microbial Ecology. 74 Figure legends Fig. 1. Degradation of 2,4-D in soil. 2,4-D was applied to microcosm sails at a concentration of 250 ppm for each of the 5 repeated treatments. Degradation of 2,4-D was monitored by HPLC. Initial soil water content was adiusted to (A) 10% and (B) 25%. Fig. 2. Detection of indigenous 2,4-D degrading populations in soil by Southern blot analysis. Total soil bacterial DNA was isolated at time zero and after 1, 2, 3, 4 and 5 cycles of 2,4-D treatment (+ : 2,4-D treated, — : no 2,4-D controls). 1.5ug DNA samples were digested with BamHI, transferred to nitrocellulose hybridization membrane, and hybridized with the tfdA gene probe. Initial soil water content was (A) 10% and (B) 25%. Fig. 3. Comparison by DNA probe hybridization of soil bacterial DNA and plasmid DNA digested with three different restriction enzymes. Soil bacterial DNA isolated after 5 cycles of 2,4-D treatment at 10% moisture soil microcosm was used. Plasmid DNA was extracted from P. pickettii which was isolated from the same agricultural soil. The tfdA gene was used as probe. 75 Fig. 4. Detection of indigenous 2,4-D degrading soil bacteria not detected with the WA gene probe. Soil DNA samples (1.5ug) from 25% moisture soil microcosm were digested with H indIII, size-fractionated by agarose gel electrophoresis, then transferred to nitrocellulose hybridization membrane. A 6.5kb fragment derived from the large plasmid of P. paucimobilis was used as probe. 0 : soil DNA sample of time zero. 1, 2, 3, 4 and 5 : soil DNA samples after 1, 2, 3, 4 and 5 treatments of 2,4-D (+ : 2,4-D treated, — : no 2,4- D controls), respectively. Fig. 5. Comparison by DNA probe hybridization of soil bacterial DNA and P. paucimobilis DNA digested with three different restriction enzymes. Soil bacterial DNA isolated after 5 cycles of 2,4-D treatment at 25% moisture soil microcosm was used. A 6.5kb fragment derived from the large plasmid of P. paucimabilis was used as the probe. 76 300 300 ' 253 :5. E can as .280 30 40 50 60 Days 20 10 Figure 1. 77 xv. ..... t... 49m mango m. “e amfirs 8'3"? 78 sail DNA p712 soi l DNA p71 2 ‘ soil DNA . p712 8 17» 9‘99 17 69 CD1 I 8093 IHWBH III58 79 4.0.5 '9 mafia 8'17 -> 99" i‘ S9 a. 80 ml soi l DNA KI3 soi l DNA Kl3 soi l DNA KI3 18°33 III P“ !H I 'lsd LIST OF REFERENCES 81 LIST OF REFERENCES 1. Amy, P. 8., J. W. Schnlke, L. M. Frazier, R. J. Seidler. 1985. Characterization of aquatic bacteria and cloning of genes specifying partial degradation of 2,4-dichlorophenoxyacetic acid. Appl. Environ. Microbiol. 49:1237-1245. 2. Bell, G. R. 1957 . Some morphological and biochemical characteristics of a soil bacterium which decomposes 2,4-dichlorophenaxyacetic acid. Can. J. Microbiol. 3:821-840. 3. Bethesda Research Laboratories. 1985. BRL M13 cloning/dideoxy sequencing instruction manual. Bethesda Research Laboratories, Gaithersburg, Md. 4. Biederbeck, V. 0., C. A. Campbell, and A. E. Smith. 1987 . Effects of long-term 2,4-D field applications on soil biochemical processes. J. Environ. Qual. 16:257-262. 5. Bohlool, B. B., and E. L. Schmidt. 1980. The immunofluorescence approach in microbial ecology. Advances in Microbial Ecology 4:203-242. 6. Bonde, G. J. 1977 . Bacterial indicators of water pollution. Advances in Aquatic Microbiology 1:273-364. 7. Cnllimore, D. R. 1981. The enumeration of 2,4-D degraders in Saskatchewan soils. Weed Sci. 29:440-443. 8. Don, R. H., and J. M. Pemberton. 1981. Properties of six pesticide degradation plasmids isolated fiom Alcaligenes paradoxus and Alcaligenes eutrophus. J. Bacteriol. 145:681-686. 9. Don, R. H., A. J. Weightman, H. J. Knackmuss, and K. N. Timmis. 1985. Transposon mutagenesis and cloning analysis of the pathways for degradation of 2,4-dichlorophenoxyacetic acid and 3-chlorobenzoate in Alcaligenes eutrophus JMP 134(pJP4). J. Bacteriol. 161:85-90. 10. Don, R. H. and J. M. Pemberton. 1985. Genetic and physical map of the 2,4-dichlorophenoxyacetic acid-degradative plasmid pJP4. J. Bacteriol. 161:466-468. 82 11. Evans, W. C., B. S. W. Smith, H. N. Fernley, and J. 1. Davies. 1971. Bacterial metabolism of 2,4-dichlorophenoxyacetate. Biochem. J. 122:543-551. 12. Fliermans, C. B., and E. L. Schmidt. 1975. Autoradiography and immunofluorescence combined for autecological study of single cell activity with Nitrobacter as a model system. Appl. Microbiol. 30:676-684. 13. Foster, R. K. and R. B. Mckercher. 197 3. Laboratory incubation studies of chlorophenoxyacetic acids in Chernozemic soils. Soil Biol. Biochem. 5:333-337. 14. Fournier, J. C. 1980. Enumeration of the soil micro-organisms able to degrade 2,4-D by metabolism or ca-metabolism. Chemosphere 9:169- 174. 15. Fournier, J. C., P. Codaccioni, and G. Soulas. 1981. Soil adaptation to 2,4-D degradation in relation to the application rates and the metabolic behaviour of the degrading microflora. Chemosphere 8:97 7-984. 16. Guth, J. A. The study of transformations, p. 123-157. In R, J, Hance (ed.), Interaction between herbicides and the soil. Academic Press, London. 17 . Barker, A. R., R. H. Olsen and R. J. Seidler. 1989. Phenoxyacetic acid degradation by the 2,4-dichlorophenoxyacetic acid (TFD) pathway of plasmid pJP4 : Mapping and characterization of the TFD regulatory gene, tde. 171:314-320. 18. Hill, W. E., W. L. Payne, and C. C. G. Aulisio. 1983. Detection and enumeration of virulent Yersinia enterocolitica in food by DNA colony hybridization. Appl. Environ. Microbiol. 46:636-641. 19. Hirsch, P. R., M. Van Montagu, A. W. B. Johnston, N. J. Brewin and J. Schell. 1980. Physical identification of bacteriocinogenic, nodulation and other plasmids in strains of Rhizobium leguminasarum. J. Gen. Microbiol. 120:403-412. 20. Holben, W. E., B. M. Schroeter, V. G. Mathesan, R. H. Olsen, J. K. Kukor, V. O. Biederbeck, A. E. Smith and J. M. 'I‘iedie. 1992. Analysis by gene probes of soil microbial populations selected by treatment with 2,4-dichlorophenoxyacetic acid (2,4-D). Submitted for publication. 21. Holben, W. E., J. K. Jansson, B. K. Chelm, and J. M. Tiedje. 1988. DNA probe method for the detection of specific microorganisms in the soil bacterial community. Appl. Environ. Microbiol. 54:703-711. 22. Jansson, J. K., W. E. Holben, and J. M. 'I‘iedie. 1989. Detection in soil of a deletion in an engineered DNA sequence by using DNA probes. Appl. Environ. Microbiol. 55:3022-3025. 83 23. Jensen, H. L., and H. 1. Peterson. 1952. Detoxication of hormone herbicides by soil bacteria. Nature 170:39-40. 24. Ka, J. O., W. E. Holben, and J. M. Tiedie. 1992. Studies on diversity of 2,4-D degrading bacteria isolated from Kellogg Biological Station soils. Submitted for publication. 25. Kaufmann, D. D., and P. C. Kearney. 1976. Microbial transformations in the soil, p. 29-64. In L. J. Audus, Herbicides : physiology, biochemistry, ecology, vol. 2. Academic Press, London, U. K. 26. Knnc, F., and J. Rybarova. 1983. Mineralization of carbon atoms of 14C-2,4-D side chain and degradation ability of bacteria in soil. Soil Biol. Biochem. 15:141-144. 27 . Loos, M. A. 1975. Phenoxyalkanoic acids, p. 1-128. In P. C. Kearney, and D. D. Kaufman (ed.), Herbicides : Chemistry, degradation and made of action, vol. 1. 2nd edn, Marcel Dekker Inc., New York. 28. Laos, M. A., R. N. Roberts, and M. Alexander. 1967. Phenols as intermediates in the decomposition of phenoxyacetates by an Arthrobacter species. Can. J. Microbiol. 13:679-690. 29. Laos, M. A., I. F. Schlasser, and W. R. Mapham. 1979. Phenoxy herbicide degradation in soils : Quantitative studies of 2,4-D- and MCPA- degrading microbial populations. Soil Biol. Biochem. 11:377-385. 30. Maniastis, T., E. F. Fritsch, and J. Sambrook. 1982. Molecular cloning : a laboratory manual. Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y. 31. On, L. T. 1984. 2,4-D degradation and 2,4-D degrading microorganisms in soils. Soil Sci. 137:100-107. 32. Parker, L. W., and K. G. Doxtader. 1983. Kinetics of the microbial degradation of 2,4-D in soil : Effects of temperature and moisture. J. Environ. Qual. 12:553-558. 33. Pemberton, J. M. 1979. Pesticide degradation plasmids : a biological gagwer to environmental pollution by phenoxy herbicides. Ambio 8:202- 34. Rogofl, M. H., and J. J. Reid. 1956. Bacterial decomposition of 2,4- dichlorophenoxyacetic acid. J. Bacteriol. 71:303-307. 35. Sandmann, E. R. I. C., and M. A. Loos. 1984. Enumeration of 2,4-D- degradung microorganisms in soils and crop plant rhizospheres using 84 indicator media ; High populations associated with sugarcane(Saccharum afl‘icinarum). Chemosphere 13:1073-1084. 36. Sayer, G. 8., M. S. Shields, E. T. Tedford, A. Breen, S. W. Hooper, K. M. Sirotkin, and J. W. Davis. 1985. Application of DNA-DNA colony hybridization to the detection of catabolic genotypes in environmental samples. Appl. Environ. Microbiol. 49: 1295-1303. 37 . Stanier, R. Y., N. J. Palleroni, and M. Dondorofi. 1966. The aerobic pseudomonads : a taxonomic study. J. Gen. Microbiol. 43:159-271. 38. Streber, W. R., K. N. Timmis, and M. H. Zenk. 1987. Analysis, cloning, and high-level expression of 2,4-dichlorophenoxyacetate monooxygenase gene tfii A of Alcaligenes eutrophus JMP 134. J. Bacteriol. 169:2950-2955. 39. Torstensson, L. 1980. Role of microorganisms in decomposition, p. 159- 178. In R. J. Hance (ed.), Interactions between herbicides and the soil. Academic Press, London, U. K. 40. Torstensson, N. T. L., J. Stark, and B. Goransson. 1975. The effect of repeated applications of 2,4-D and MCPA on their breakdown in soil. Weed Research 15:159-164. 41. Walker, R. L., and A. S. Newman. 1956. Microbial decomposition of 2,4- dichlorophenoxyacetic acid. Appl. Microbiol. 4:201-206. Chapter Four Gene Probe Analysis of Competition among 2,4-D-degrading Bacteria in Soil under Selective Conditions. J. 0. KA, W. E. HOLBEN and J. M. TIEDJE“. Center for Microbial Ecology and Department of Microbiology and Public Health, Michigan State University, East Lansing, MI 48824. 85 86 ABSTRACT Competition among indigenous and inoculated 2,4-D degrading bacteria was studied in a native Kansas prairie soil following 2,4-D additions. The inoculated soil received four different 2,4-D degrading stains at densities of ~103 cells/g soil :Pseudomonas cepacia strain DBO 1/pJ P4 ; and three Michigan soil isolates identified as Pseudomonas pseudomallei, Pseudomonas paucimobilis and Pseudomonas pickettii. The soil community DNA samples were analyzed on Southern blots using a tfdA gene probe and also a 6.5 kb probe derived from P. paucimobilis since this strain has no homology to tfd genes. P. cepacia DBO 1/pJP4, a constructed stain, outcompeted both of the other added strains and the indigenous 2,4-D degrading bacteria. P. paucimobilis was detected as a secondary dominant and P. pseudamallei and P. pickettii were not detected. In contast to the outcome of competition experiments in soil, relative fitness coemcients determined in axenic broth cultures indicated that P. pseudomallei along with P. cepacia DBO 1/pJ P4 should have been the best competitors. These parameters predicted the outcome of competition in soil for some but not all stains. Growth in mixed broth culture enhanced the growth of the slower growing stains principally by reducing the lag time over that found in axenic culture. Plasmids containing the 2,4-D pathway were important determinants of competitiveness since pKA4 in P. cepacia DBOl had the slower growth characteristics of its original host, P. pickettii, rather than the more rapid growth found when this stain harbors pJ P4. 87 INTRODUCTION Microorganisms seldom exist in complete segregation from others but interact both positively and negatively in their habitats. One of the major goals of microbial ecology is to tmderstand how interactions between microbial species influence their survival and abundance in nature. Among the various types of interactions between microbial populations (18), competition for carbon is often the major determinant of the relative abundance of the indigenous organisms in the soil environment (3, 20). Pesticides and other xenobiotic compounds are new carbon compounds that have been intoduced into the environment during past several decades. Some of these compounds are good carbon sources for soil microorganisms capable of their degradation. 2,4—D, for example, is known to be a growth substate for a number of difl‘erent soil microorganisms (4, 14, 16, 17). Since these compounds are usually not present in nature and their supply is under human contol, they make good models to study microbial resource competition in soil. Batch liquid culture and chemostats have provided the basic background for understanding the principles of microbial competition for common carbon sources (9, 10), but it is not clear that we can easily extapolate this information to predict the outcome of competition in the poorly mixed soil habitat. Furthermore, competition studies in soil have been difficult to perform due to limited techniques for enumeration of difl‘erent campetiting organisms. DNA probe methods offer the potential advantage of being able to distinguish and identify a number of competiting organisms in the inoculated as well as native communities without the requirement for prior selective culture. 88 In this study, DNA gene probes were used to study competition for the 2,4-D growth substate among inoculated and indigenous 2,4-D degrading bacteria in a native prairie soil which has no history of cultivation or agricultural chemical use. In addition, batch liquid cultures were used to study the growth patterns of 2,4-D degrading bacteria in axenic and mixed cultures and the result was compared with that from the soil competition experiment. Finally, we evaluated the effects of different 2,4-D degradative plasmids on both the growth patterns and the competitive advantage of the host bacteria. MATERIALS AND METHODS Bacteria and soil. The bacterial stains used or isolated in this work and any sequence homology to tfd genes from plasmid pJP4 are described in Table 1. P. cepacia DBO 1/pKA4 was obtained through filter mating between P. cepacia D801 and P. pickettii 712/pKA4. All tfdA hybridizing stains were distinguishable by unique restriction fragments. The soil used in this study was from the Konza Prairie near Manhattan, Kansas. This is the largest native prairie area remaining in the U. S. (3,400 hectare tact), and has never been cultivated or tested with agricultural chemicals. MPN experiments indicated a low background of native 2,4-D degraders in this soil, ~30 cells/g. This soil was also selected because it should provide a microbial environment dissimilar to the one from which the inoculated stains were isolated, and therefore competition between indigenous stains and invaders (perhaps less fit) could be evaluated. The inoculated stains are from the KBS soil which was formed from glacial till in a humid region and under cultivated agriculture, while the Konza soil is 89 derived from layered limestone and shale in a semi-arid region and under native grasses. Media and culture conditions. All stains were maintained an MMO mineral medium (19) plus 500 ppm of 2,4-D. Stains to be inoculated in soils and flasks were cultured to full growth at 30°C in Luria broth and then reinoculated in fresh Luria broth (1:50) for an overnight culture. Enumeration of total viable cells was accomplished by plating appropriate dilutions of soil suspensions onto PTY G agar, which contained 0.25 g of Bacto-peptone, 0.25 g of Bacta-typtone, 0.5 g of yeast extact, 0.5 g of glucose, 0.03 g of MgSO4, 0.003 g of CaClz and 15 g of agar per liter. Stains for plasmid isolation were cultured in MO mineral medium with 500 ppm of 2,4-D. Plasmids were isolated by the method of Hirsch et. al.(5). Inoculation and sampling of bacteria in soil. The above overnight cultures of four difi'erent 2,4-D degrading bacteria, P. cepacia st DBO l/pJP4, P. pseudomallei 7 45, P. paucimobilis 1443 and P. pickettii 712, grown at 30°C were harvested by centrifugation (10,000 x g, 10 min., 4°C), washed twice with an equal volume of sodium phosphate buffer (15 mM, pH 7 .0), and again collected by centrifugation. The cells were resuspended in 1110 volume of sodium phosphate buffer and left at 4°C for 5 h. Konza Prairie soil was sifted through a 2-mm (square aperture) sieve, adjusted to a water content of ~35%, and inoculated with each of four different species to a final concentation of ~1.5 x 103 cells/g soil. The soil was thoroughly mixed and 500 g was tansferred to each of three replicate polyethylene wide-mouth battles (microcosms I - 111). Three additional replicates were not inoculated with 2,4-D degrading bacteria. Inoculated and uninoculated soils were teated with 2,4-D dissolved in 0.1 M NaH2P04 bufl‘er (pH 7.0) to a concentation of 250 ppm (mg/g) and thoroughly mixed. 90 The disappearance of 2,4-D in soil was monitored by HPLC (8) and the soils were respiked with 2,4-D (to 250 ppm) after its removal for each of 10 cycles of degradation. Stesses of cold temperature and of starvation were imposed on two sets of microcosms at a later stage of teatment to evaluate the robustness of the community to other conditions. One inoculated and uninoculated soil microcosm was incubated at 4°C for 14 days immediately after the eighth addition of 2,4-D and then returned back to room temperature (microcosm II). In another microcosm from each group, the 2,4-D addition was delayed for 14 days after the end of the seventh degradation cycle before these two microcosms were re-subjected to the normal 2,4-D addition (microcosm III). At time zero and at the end of the second, the fifth and the tenth teatment of 2,4-D, a 10 g soil subsample from each microcosm was used for bath MPN determination (6) and total viable counts on PTYG plates, and a 50 g soil subsample was used for isolation of total bacterial DNA by using the cell extaction method (7 ). Batch culture experiments. The stains were cultured, harvested and prepared in sodium phosphate buffer as described above. Liquid culture studies were conducted in 250 ml-flasks containing 100 ml of 2,4-D minimal medium (500 ppm). Axenic culture studies were in duplicate and mixed culture studies in triplicate. Mixed cultures were inoculated at a ratio of 1 : 1 : 1 among these stains ; the initial population densities on PTYG plates were ~7 x105 cells/ml and ~1 x 106 cells/ml in the two experiments. Individual stains were discriminated and counted based on distinctive colony morphology. All cultures were incubated at 30°C and aerated by shaking at 200 rpm in a New Bnmswick G24 environmental incubator shaker (New Brunswick Scientific Co., New Brunswick, NJ). 91 Probe preparation. The tfdA gene and a 6.5 kb fragment of P. paucimobilis were used as probes to detect and discriminate the inoculated and the indigenous 2,4-D degrading bacteria by Southern blot. The tfdA gene probe was cloned as a 556 bp EcoRI-BamHI fragment from the well- known 2,4-D degradative plasmid pJP4 into the multiple cloning site of plasmid pUCl9 ( 15). The 6.5 kb probe, which is specific for P. paucimobilis 1443, was cloned into pUC 19 as a BamHI restriction fragment from the large plasmid of this stain. Cloned plasmids were amplified and isolated by the method of Maniatis et. al. (15). These probes were isolated as restricted fragments from the corresponding cloned plasmid pUC 19 on 0.7 % agarose gel, purified with Geneclean kit (Bio 101 Inc, La Jolla, CA), and labeled with 32F by using the Random Primed DNA Labeling kit (Boehringer Mannheim, Indianapolis, IN) or the Nick Translation kit (Boehringer Mannheim, Indianapolis, IN) according to the manufacturer’s specifications. Labeled probes were separated from uninoculated nucleotides by 3pm column prior to use (15). The probe was used at approximately 107 cme 10 ml of hybridization fluid. Southern hybridization. Total soil bacterial DNA was digested with appropriate restriction endonucleases according to the manufacturer’s specifications. Digested DNA was size-fractionated by electophoresis through horizontal 0.7 % agarose gel and tansferred to nitocellulose hybridization membranes (15). Prehybridization, hybridization and post- hybridization washes were performed as described by Holben et. al. (7) with some modifications. The membranes were prehybridized for at least 24 h at 42°C and after hybridization, three washes were carried out at room temperature, followed by one wash at 65°C. Hybridization signals were detected by autoradiography using Kodak X-omat AR film (Kodak, 92 Rochester, NY) exposed at -70°C with a Quanta III (Sigma, St. Louis, MO) intensifying screen. Exposure times were 1 to 4 days depending on the intensity of the radioactive signal. RESULTS Degradation of 2,4-D in soil. The degradation patterns of 2,4-D in Konza Prairie soil with and without the added 2,4-D degrading bacteria are shown in Figure 1. 2,4-D was quickly degraded without a lag period in inoculated soil maintained at a soil water content of ~35% and incubated at room temperature(Figure l-A). Under these conditions, it took about 1 week for each addition of 250 ppm 2,4-D to be degraded completely throughout the 10 teatments. In microcosm II kept at 4°C for two weeks during the eighth teatment, the 2,4-D degradation rate was significantly retarded during this period, but it was restored to the preceding rate when microcosm was tansferred to room temperature (data not shown). In microcosm IH, where the addition of 2,4-D was delayed for two weeks after the seventh cycle, the 2,4-D degradation rate in subsequent teatments was not affected (data not shown). Without added 2,4-D degrading bacteria, the exposure of indigenous 2,4-D degrading populations of Konza Prairie soil to 2,4-D for the first time resulted in a lag period of about two weeks prior to complete degradation in 3 weeks (Figure 1-B). The lag period was not observed following the second addition and the 2,4-D degradation rate was similar to that found for inoculated soil. In uninoculated microcosm soils incubated at 4°C for two weeks (or delayed in 2,4-D teatment), the 2,4-D degradation patterns in subsequent teatments were similar to those found for inoculated soils. 93 In soil inoculated with 2,4-D degrading bacteria, the initial population of 2,4-D degraders was 7.5 x 103/g soil (Figure 2). After the second teatment of 2,4-D, the population markedly increased to 1.8 x 108/g soil which was stably maintained throughout the subsequent 2,4-D teatments. In the uninoculated soil, the initial population detected by MPN was 3.1 x 101/g soil which increased to 3.1 x 107/g soil with two additions of 2,4-D and then was maintained at this level through subsequent additions of 2,4-D (Figure 2). Throughout the 10 teatments of 2,4-D, the population of 2,4-D degraders detected by MPN was higher by a factor of 6 in inoculated soil than in uninoculated soil. Relationship of 2,4-D degrading populations to total viable counts. Addition of 2,4-D as a specific carbon resource had a marked effect on the community stucture (Table 2). In Konza Prairie soil, where the initial total viable counts were 7.8 x 107 and 4.8 x 107 cells/g soil in inoculated soil and uninoculated soil, respectively, the total viable count increased three to five- fold with repeated 2,4-D teatments in both soils (Table 2). By contast, 2,4- D degrading populations increased 2.4 x 104-fold and 106-fold with repeated 2,4-D teatments in inoculated soil and uninoculated soil, respectively (Figure 2). Under this selective condition, the rapidly growing populations of 2,4-D degrading bacteria became dominant members of the total bacterial community, as indicated by the decreasing ratio of total viable count to 2,4- D degrading bacteria and the absolute increase in viable counts (Table 2). In inoculated soil, the typical colonies of P. cepacia DBO 1/pJP4 (irregular, white and large colonies)were always detected in greatest numbers on PTYG plates (Figure 2), suggesting that this stain was predominant throughout this experiment. The ratio of 2,4-D degrading bacteria to P. cepacia DBO 1/pJP4 decreased from 5.0 to 0.9-1.4 indicating that this stain 94 dominawd the 2.4-D degrading community as well as the total community (Table 2). I Probing total soil bacterial DNA. Total bacterial DNA isolated from the three replicate microcosm soils at the indicated times was hybridized to the 32P-labeled tfdA gene probe after restriction digestion, gel separation and Southern tansfer. In microcosm soils inoculated with the four 2,4—D degrading bacteria, a single DNA hybridization band was detected in the DNA isolated after the second, the fifth and the tenth teatment of 2,4-D, and no band was observed in the DNA sample from time zero (Figure 3-A). The size of the hybridized band obtained from total soil bacterial DNA digested with 15le was about 8.3 kb, which corresponds to a fragment containing the tfdA gene in plasmid pJP4 digested with the same enzyme. To confirm the origin of the band fi'om soil, plasmid DNA of pJP4 and total soil bacterial DNA of the second treatment were digested separately with three different restriction enzymes and the size of bands that hybridized to the tfdA probe were compared (Figure 4-A). Matching band patterns were obtained between these two DNA samples, indicating that these hybridized bands corresponded to P. cepacia DBO llpJP4 and that this stain predominated in this experiment. This result from hybridization experiments is consistent with that from plate counts on PTYG which indicated that P. cepacia DBO 1/pJP4 was the predominant colony type among the added and indigenous 2,4-D degraders throughout the experiment. When total soil bacterial DNA was digested with HindIII, subjected to Southern tansfer and hybridized to the 6.5 kb probe, two hybridized DNA bands (5.0 kb and'11.3 kb in size) were obtained from DNA samples isolated after the second, the fifth and the tenth teatment of 2,4-D, and no band 95 was observed in the DNA sample from time zero (Figure 3-B). These two hybridized bands were also detecwd on Southern blots when P. paucimobilis 1443 DNA was digested with the same enzyme and hybridized to the same probe (8), suggesting that this stain produced the hybridization bands in the microcosm DNA. Since each lane contained the same amount of total DNA (1.5 ug), the weaker intensifies (1/180 - 1/35) of these latter bands indicate that P. paucimobilis 1443 was maintained at a lower density than P. cepacia st DBO 1/pJP4. This interpretation is further supported by the fact that the number of P. paucimobilis 1443 colonies on PTYG plates were not obvious. In uninoculated soils, bands hybridizing to thetfdA probe began to appear only after five teatments of 2,4-D (Figure 5). This population was perhaps becoming prominant at this time since the bands were weak from two of the three microcosms. The intensities of bands from the microcosm with weak hybridization increased and became similar to that of the stong band of uninoculated microcosm I after the tenth teatment of 2,4-D. Total soil bacterial DNA from the tenth teatment and plasmid DNA of a new 2,4-D degrading bacterium, isolated from the same soil and designated stain K17, were digested separately with three different restriction enzymes, and compared (Figure 4-B). Matching band patterns were observed between these two DNA samples, suggesting that one of the indigenous 2,4-D degrading bacteria, K17 , had become predominant in the uninoculated sail. No band was detected when the soil bacterial DNA was hybridized to the 6.5 kb probe. Axenic growth characteristics of 2,4-D degrading bacteria in broth culture. To understand axenic growth patterns of 2,4-D degraders, each stain was inoculated into 2,4-D minimal medium under uninduced 96 conditions. The lag period, specific growth rate (2) and relative fitness coeficients (1 1) were measured (Table 3). P. pseudomallei 7 45, P. cepacia DBO 1/pJP4 and K17 showed short lag periods (<15 h) and began to grow exponentially at about 20 h of incubation (Figure 6). By contast, P. paucimobilis 1443 showed a longer lag period (~35 h) and P. pickettii 712 showed the longest lag period (>60 h). The specific growth rates of these bacteria during exponential growth were similar except for P. pickettii which was markedly lower. The role of the host cell versus the plasmid in determining growth characteristics was evaluated by comparing growth of the same host with two different plasmids, and the same plasmid in two different hosts. P. cepacia DBOl carrying plasmid pKA4 instead of plasmid pJP4, showed the lower growth rate of the plasmid in its original host (Table 3) and a lag time intermediate between the original host, P. pickettii, and the new host, P. cepacia (Figure 6). P. pseudomallei 7 45 showed the largest number of doublings during the first 30 h of incubation and had the highest fitness coefficient among these 2,4-D degraders, suggesting that it might be the best competitor. Competition experiments in broth culture. The behavior of each 2,4-D degrading species in three member broth cultures was monitored by plate counts. When LB-grown P. pseudomallei 745, P. cepacia st DBO 1/pJP4 and K17 were inoculated together into 2,4-D minimal medium at a ratio of 1 : 1 : 1, these 2,4-D degraders exhibited similar growth patterns (Figure 7-A). This result is consistent with that of the axenic growth experiment in which all of these stains showed a short lag period and rapid growth. In mixed cultures of P. pseudomallei 7 45, P. paucimobilis 1443 and P. pickettii 712 at a ratio of 1 : 1 : 1 (Figure 7 -B), P. pseudomallei 745 multiplied quickly and 97 slightly outgrew the other two stains in the initial phase, but after 33 h of incubation its growth rate began to decline even though 65 % of the substate remained. By contast, P. paucimobilis 1443 initiated growth earlier in mixed culture than in axenic culture, thereby overcoming P. pseudomallei 745 in the last phase. This result was unexpected because, from the axenic growth experiment, P. pseudomollei 7 45 should have reached stationary phase and depleted the substate before P. paucimobilis 1443 and P. pickettii 712 would have begun to grow. P. pickettii 712 also showed significant growth dming the first 40 h of incubation in mixed culture, whereas it did not show any growth dming 80 h of incubation in axenic culture. To further analyze the stimulatory efl'ect of co-culture on the stains that grew slowly in axenic culture, we grew the slowest growing stain, P. pickettii 712, under several cultural conditions (Figure 8). Since the above experiments included a nutritional shiftdown fi'om LB medium to 2,4-D mineral medium, we evaluated whether stain 712 could shift to a 2,4-D pathway intermediate, succinate, in the same mineral medium. This occurred without significant lag (Figure 8) and produced a rapid doubling time (1.4 h). Growth in 2,4-D and tansfer to 2,4-D resulted in continuous growth. Growth of inoculum on succinate and LB produced significant lag periods when shifted to 2,4-D, suggesting that induction of the 2,4-D pathway may be delayed. When culture filtate from 2,4-D grown P. pseudomallei 745, a member of the stimulatory triculture, was added to the 2,4-D mineral medium, the lag time of P. pickettii 712 was reduced by one- third. While this does not achieve the shorter lag time found for the triculture, it does suggest that products of P. pseudomallei 745 do benefit this slow growing and otherwise non-competitive strain. 98 DISCUSSION The study evaluated competition in three difl'erent environments : non- sterile soil, mixed broth culture and axenic culture, and among stains with difl'erent habitats of origin ; a laboratory host stain containing a plasmid isolated in Anstalia (1), three stains from Michigan agricultural soils, and one stain indigenous to the native Kansas prairie soil used in the study. The key features of these stains and their competitive performance are summarized in Table 4. Kinetic parameters such as llmax and Ks, that underlie growth rate, are well-documented factors important to competition ; this has recently been illustated for 2,4-D degrading stains (12). But, other factors such as induction period, adaptation to the environment, and crossfeeding of growth factors can also contribute to competitive outcome and were the focus of this study. When four 2,4-D degrading organisms were added at equal densities to a soil from a difl'erent climatic region than their origin, primary and secondary dominant stains emerged, an outcome which was not entirely predicted from their fitness coefficients measured in axenic broth culture. P. pseudamallei 745, P. cepacia DB01/pJP4 and K17 belong to the fast- growing group and thus were anticipated to outgrow both of the slow- growing bacteria, P. pickettii 712, and P. paucimobilis 1443 in soil. Although P. pseudomallei 7 45 had the highest fitness coefficient among these stains, it was not detected with the DNA probing method but instead the constructed stain, P. cepacia DBO 1/pJP4, was observed as the predominant stain in this experiment based on both the plate counts and the hybridization results. K17 , the indigenous 2,4-D degrading bacterium in Konza Prairie soil, was not detected on Southern blot throughout 10 99 teatments of 2,4-D in inoculated soil, whereas this stain was predominant in the uninoculated soil from the fifth teatment of 2,4-D. Considering that this stain has never been exposed to 2,4-D in the past and that its initial population density was lower by a factor of 50, it apparently had no chance for detectable growth because P. cepacia DBO 1/pJP4 and P. paucimobilis 1443 likely consumed most of the 2,4-D. This outcome was not necessarily expected since it did have the highest specific growth rate on 2,4-D and should have been well adapted to the other conditions of its native habitat. P. paucimobilis 1443 was detected as a secondary dominant on Southern blots throughout this experiment in inoculated soil. This strain showed an intermediate lag time in broth culture which was followed by a specific growth rate similar to those of P. pseudomallei 745, P. cepacia DBOllpJP4 and K17 . This suggests that P. paucimobilis 1443 may grow slowly initially but has the potential to catch up with other fast-growing bacteria. This potential was demonstated in mixed broth culture where P. paucimabilis 1443 outgrew P. pseudomallei 745 in the later phase, and may have also led P. paucimobilis 1443 to a secondary dominance in the soil competition experiment. P. pickettii 712, which showed a long lag time followed by slow growth in axenic culture, was not detected on Southern blot throughout this experiment in inoculated soil. It is also noteworthy that P. paucimabilis 1443 shares no DNA sequence homology with the tfd genes that were presumably important to the competitive growth of P. cepacia DBO 1, as well as the lesser growth of the other stains. In uninoculated soil after the second 2,4-D teatment the population density of indigenous 2,4-D degrading bacteria reached levels which should be detected by DNA probes. However, no hybridization bands were detected on Southern blot after the second teatment, and after the fifth teatment 100 bands of weak intensity were seen in two of three microcosms while a stong band was seen in the third (Figure 5). This suggests that another 2,4-D degrading microbial population not detected with the available gene probes dominated in the initial phase and was replaced (or co-established) with K17 after the fifth teatment of 2,4-D. The importance of interactions among competiting organisms on the outcome of competition is apparent when comparing patterns from the axenic broth culture with the results tom the mixed broth cultures (Table 4). P. paucimobilis 1443 and P. pickettii 712 began to grow much earlier in mixed culture with P. pseudomallei 745 than in axenic culture, suggesting that some intermediates produced from P. pseudomallei 745 stimulated the growth of these two stains in mixed culture. One explanation is that pathway intermediates induced the 2,4-D pathway of these two less fit stains, thereby substantially reducing the long lag times. Another is that nutritional factors provided by crossfeeding aided growth on this less favorable substate. Lag periods before 2,4-D degradation in soil are typical ; Laos (13) suggested this could be due to enzyme induction while Miwa and Kuwatsuka (16) felt it was related to the initial populations of 2,4-D degraders. An interesting result in these studies is how dominant lag time can be in determining fitness, but also how this can be compensated for by population interactions. The difference in outcome of competition in the broth vs soil experiments may be due to the fact that the rapid exchange of cell products in the mixed broth environment reduced the fitness difference while this did not occur in the unmixed, physically isolated niches in the soil environment. Although coldness and starvation stesses were applied to replicate inoculated and uninoculated microcosms, these stesses were not enough to 101 change the composition of the 2,4-D degradative microbial communities. Soil microbial communities may be rather stable to typical environmental perturbations. The impact of a difl‘erent 2,4-D degradative plasmids on 2,4-D growth pattern of the host microorganism was noticeable. For example, P. cepacia DBOl carrying pJP4 exhibited a short lag period followed by rapid growth, but the same stain carrying pKA4 instead of pJP4 showed a relatively long lag period followed by slow growth (Figure 6). The specific growth rate of P. cepacia DBO 1/pKA4 (0.077) was similar to that of P. pickettii 712/pKA4 (0.078) and not that of P. cepacia DBO 1/pJP4 (0.179). However, the lag time of P. cepacia DB01/pJP4 was intermediate between that of P. cepacia DBOl with pJP4 and P. pickettii with pKA4. Thus plasmid and plasmid- host interactions determined growth rate and lag time, respectively, key determinants of competitive outcome. This research demonstates that a xenobiotic compound is a useful model to study competition in soil as well as provide population information important to understanding biodegradation. The most interesting findings were that species interactions appear to affect competition to a difi‘erent degree in soil vs broth perhaps making data derived from broth culture less predictive for the soil habitat, that superior competitiveness can be determined by a plasmid, and that non-native soil strains, and in fact a constucted stain not even originating from soil was the most successful competitor in soil. 1 02 ACKNOWLEDGMENTS This work was supported by National Science Foundation grants fi'om Long-Term Ecological Research program and the Center for Microbial Ecoloy. We thank Dr. Charles Rice for providing Konza Prairie soil and information on the site and thank Dr. Larry Forney for his valuable comments on this manuscript. 103 Table 1. Bacterial strains and plasmids. tfd genes which Strain Plasmid hybridize to Original plasmid habitat of isolate Pseudomonas cepacia pJP4 A. B . C . D Laboratorya DBOl/pJP4 Pseudomonas cepacia pKA4 A Laboratory DBOl/pKA4 Pseudomonas pseudomallei 745 pBSS A, B, C, D KBS soil, Mlb Konza Prairie K17 pKOSl A, B, C, D soil, KS Pseudomonas paucimobilis pBS3 None KBS soil, Mb 1443 Pseudomonas 712/pKA4 a Provided by R. Olsen, Univ. of Michigan ; plasmid was isolated in Australia from an Alcaligenes strain (1). b From LTER gene flow plot at the Kellogg Biological Station, Hickory Corners, MI. 104 Table 2. Comparison of total viable counts, 2,4-D degrading populations and numbers of P. cepacia DBOl/pJP4 in Konza Prairie soil. Total viable 2.4—D degrading Number of Total viable counts per 2.4—D bacteria per P. Soil 2.4—D counts degrading cepacia DBOll treatments (x107lg) bacterium pl P4 Inoculated 0 7.8 1.0 x 104 5.0 soil 2 41.4 2.3 0.9 5 37.9 1.1 1.0 10 21.2 1.0 1.4 Uninoculated 0 4.8 1.5 x 105 - soil 2 25.6 8.8 - 5 17.9 3.0 - 10 12.3 2.7 - 105 Table 3. Growth characteristics of 2,4-D degrading bacteria in axenic broth culturea. Specific growth Relative fitness Strain Lag period rate coefficientb P. pseudomallei 745 <15h 0.173 1.0 P. cepacia DBOl/pJP4 <15h 0.179 0.95 K17 <15h 0.185 0.74 P. paucimobilis 1443 25 - 40h 0.181 0.03 P. cepacia DBOl/pKA4 25 - 40h 0.077 0.05 P. pickettii 712/pKA4 >60h 0.078 ~0 3 All values represent means from two independent broth cultures. b Relative fitness coefficient determined in axenic broth culture according to reference (8), i.e., ratio of the number of doublings of each strain to that of P. pseudomallei under same conditions. The first 30 h of incubation was chosen at the time to evaluate relative fitness since the most rapidly growing strain, P. pseudomallei, stopped growing at this time due to depletion of substrate. 106 Table 4. Summary of key features of 2,4-D degrading strains and the outcome of competition in different environments. Hybridization Cometitive ranka Strain Source to tfd genes Axenic broth Mixed broth Soil P. cepacia DBOllpJP4 Constructed A - D l - 1 P. pseudonrallei 745/pBSS Mich. soil A - D 1 2 0 Strain K17 lpKOSI Native soil A - D 1 - 0 P. paucimobilis 1443/pBSB Mich. soil none 2 1 2 P. pickettii 712/pKA4 Mich. soil A 3 3 0 a 0 = present but not detected ; - = not included in this triculture competition. 107 Figure legends Fig. 1. Degradation of 2,4-D in soil during 10 repeated additions in microcosms inoculated with 2.4—D degrading bacteria (A) and with only indigenous 2.4-D degrading bacteria (B). Fig. 2. Change in number of soil microorganisms in response to repeated additions of 2,4-D. MPNs of 2.4-D degraders in inoculated soil (_A_) and in uninoculated soil (-—A-—). Viable counts of P. cepacia DBOl/pl P4 on PTYG (mat-u). Total viable counts on PTYG in inoculated soil {-0—) and in uninoculated soil (—0—). Fig. 3. Hybridization of labeled JdA gene probe (A) and 6.5 kb fragment (B) with total soil bacterial DNA from three replicate inoculated soils. lanes : Total soil bacterial DNA of time zero (0) and after the second (2), the fifth (5), and the tenth (10) treatment of 2,4-D. Each lane contains 1.5 pg of total soil bacterial DNA digested with EcoRI (A) and HindIII (B). Fig. 4. Comparison by DNA probe hybridization of total soil bacterial DNA and plasmid DNA digested with three different restriction enzymes. Lanes : Total soil bacterial DNA of the second treatment of inoculated microcosm 1 (1, 3 and 5) and plasmid pl P4 DNA of P. cepacia DBOllpI P4 (2, 4 and 6), total soil bacterial DNA of the tenth treatment of uninoculated microcosm 1 (7, 9 and 11) and plasmid DNA of strain K17 (8, 10 and 12). DNA was digested with EcoRI ( 1, 2, 7 and 8), HindIII (3, 4. 9 108 and 10), and BgIII (S, 6, 11 and 12). Fig. 5. Hybridization of labeled tfdA gene probe with total soil bacterial DNA from three replicate uninoculated soil microcosms. Lanes : Total soil bacterial DNA of time zero (0) and after the second (2), the fifth (5), and the tenth (10) treatment of 2,4-D. Each lane contains 1.5 pg of total soil bacterial DNA digested with HindIII. Fig. 6. Growth patterns of 2,4-D degrading bacteria in axenic culture. P. pseudomallei 745 (+ ), P. cepacia DBOl/pJP4 (.H, K17 (+), P. paucimobilis 1443 (—o—). P. cepacia DBOl/pKA4 (+), and P. pickettii 712/pKA4 {—0—}. Each point represents an average value of replicate flask cultures. Fig. 7. Competition for 2.4-D (A) among P. pseudomallei 745 (_.I_), P. cepacia DBOl/pIP4 (—u.—) and K17 (+), and (B) among P. pseudomallei 745 (—I—), P. paucimobilis 1443 (—o—) and P. pickettii 712 (_‘_) in mixed liquid culture. Each point represents an average value of three replicate broth cultures. Fig. 8. Effect of culture conditions on enhancing the growth of P. pickettii 712. Lines are labeled with medium on which the inoculum was grown followed by the growth medium. P. cepacia DBOl/pJP4 is shown for comparison. 109 80 \ 20 300 200 ' .0 0 1 Eae :8 E can no .280 100' 60 80 100 40 Figure 1. Days 110 Log cells/g soil .3 m 5 2.552. on N.A-U AnemnBaam . 59:6 u. 111 5986 w. 112 ._Nwhmm Ammaofinn mam—Hum A. Eu Ana: A 0.5 Ath 113 '9 911135,; smoH O 09 00I OQI >\o -I'0 114 OD at 550nm o fa - 8’0 V0 115 50 . - z m m -.. -0 4 .. -0 3S B m . o H 1. IO 2 - -0 1 1 . J.- .. . . 0 0. 0. 0. 0. ow ow 0. fl 8 7 6 5 8 7 6 5 35:8. won 35:8 won 116 OD at 55011111 Pa Pu - PM- PH - Modem 6855... gun-U > v fiancee gen . chuh-U runs-c ES}: I I I O I . 3.5 9:38 I 358 D I D 1‘! \D I I .. ..Iu I . . mo :5 59:6 m. LIST OF REFERENCES 117 LIST OF REFERENCES 1. Don, R. H., and J. M. Pemberton. 1981. Properties of six pesticide degradation plasmids from Alcaligenes paradaxus and Alcaligenes eutrophus. J. Bacteriol. 145:681-686. 2. Drew, 8. W. 1981. Liquid culture, p. 152-178. In P. Gerhardt, R. G. E. Murray, R. N. Costilow, E. W. Nester, W. A. Wood, N. R. Krieg, and G. B. Phillips (ed.), Manual of methods for general bacteriology. American Society for Microbiology, Washington, DC. 3. Finstein, M. 8., and M. Alexander. 1962. Competition for carbon and nitogen between Fusarium and bacteria. Soil Sci. 94:334-339. 4. Fournier, J. C. 1980. Enumeration of the soil micro-organisms able to degrade 2,4-D by metabolism or co-metabolism. Chemosphere 9: 169-174. 5. Hirsch, P. R, M. Van Montagu, A. W. B. Johnston, N. J. Brewin, and J. Schell. 1980. Physical identification of bacteriocinogenic, nodulation and other plasmids in stains of Rhizobium leguminosarum. J. Gen. Microbiol. 120:403-412. 6. Holben, W. E., B. M. Schroeter, V. G. M. Calabrese, R. H. Olsen, J. K. Kukor, V. O. Biederbeck, A. E. Smith and J. M. Tiedje. 1992. Analysis by gene probes of soil microbial populations selected by teatment with 2,4-dichlorophenoxyacetic acid (2, 4-D). Submitted for publication. 7. Holben, W. E., J. K. Jansson, B. K. Chelm, and J. M. Tiedje. 1988. DNA probe method for the detection of specific microorganisms in the soil bacterial community. Appl. Environ. Microbiol. 54:7 03-7 11. 8. Ka, J. 0., W. E. Holben, and J. M. Tiedie. 1992. Use of gene probes to detect 2,4-D degrading populations in soil maintained under selective pressure. Submitted for publication. 9. Kuenen, J. G., J. Boonsta, H. G. J. Schroder, and H. Veldkamp. 197 7. Competition for inorganic-substates among chemoorganotropbic and chemolithotophic bacteria. Microb. Ecol. 3: 119- 130. 10. Laanbroek, H. J ., A. J. Smit, G. Klein Nulend, and H. Veldkamp. 1979. Competition for L-glutamate between specialised and versitily 118 Clostridium species. Arch. Microbiol. 120:61-66. 11. Lenski, R. E., M. R. Rose, 8. C. Simpson, and S. C. Tadler. 1991. Long-term experimental evolution in Escherichia coli. I. Adaptation and divergence during 2,000 generations. Am. Nat. 138: 1315-1341. 12. Greer, L. E., J. A. Robinson, and D. R. Shelton. 1992. Kinetic comparison of seven stains of 2,4-dichlorophenoxyacetic acid-degrading bacteria. Appl. Environ. Microbiol. 58:1027- 1030. 13. Laos, M. A. 1975. Phenoxyalkanoic acids, p. 1-128. In P. C. Kearney and D. D. Kaufman (ed.), Herbicides : Chemistry, degradation, and mode of action, vol. 1. Marcel Dekker, New York. 14. Laos, M. A., I. F. Schlasser, and W. R. Mapham. 1979. Phenoxy herbicide degradation in soils : Quantitative studies of 2,4-D- and MCPA- degrading microbial populations. Soil Biol. Biochem. 11:37 7-385. 15. Maniastis, T., E. F. Fritsch, and J. Sambrook. 1982. Molecular cloning : a laboratory manual. Cold Spring Harbor Laboratory, Cold Spring Harbor, N. Y. 16. Miwa, N., and S. Kuwatsuka. 1990. Enrichment process of 2,4-D degraders in difl‘erent soils under upland conditions. Soil Sci. Plant Nut. 36:261-266. 17 . On, L. T. 1984. 2,4-D degradation and 2,4-D degrading microorganisms in soils. Soil Sci. 137:100-107. 18. Slater, J. H., and A. T. Bull. 1978. Interactions between microbial populations, p. 181. In A. T. Bull and P. M. Meadow (ed.), Companion to Microbiology. Langmans, London. 19. Stanier, R. Y., N. J. Palleroni, and M. Dondoroff. 1966. The aerobic pseudomonads : a taxonomic study. J. Gen. Microbiol. 43:159-271. 20. Stotzlry, G., and J. L. Mortensen. 1957. Efl'ect of crop residues and nitogen additions on decomposition of an Ohio muck soil. Soil Sci. 83: 165- 174. Chapter Five Integration and Excision of a 2,4-D Degradative Plasmid in Alcaligenes paradoxus and Evidence of Its Natural Intergeneric Transfer J. O. KA, and J. M. TIEDJE*. Center for Microbial Ecology and Department of Microbiology and Public Health, Michigan State University, East Lansing, MI 48824. 119 120 ASTRACT A self-tansmissible 2,4-D degradative plasmid, pKA2, has been identified in a new 2,4-D degrading stain, Alcaligenes paradoncus 2811P, isolamd fi'om agricultural soil. pKA2 occurred as a 42.9 kb plasmid in stain 2811P. However, a derivative stain 2811C was isolated from glycerol stock culture of 2811P kept at -20°C for two years. InA. paradoxus 28110, the whole plasmid pKA2 was integrand into the host chromosome without loss of 2,4-D+ phenotype, as suggested by disappearance of a free plasmid DNA band, shifting of the hybridization signal with the tfdA gene probe to the chromosome, nontansmissibility, and Southern hybridization results with plasmid and whole cell DNA. Furthermore, we demonstated that the integrated plasmid could be excised either precisely or in a difl'erent way. Another new 2,4-D degrading stain, Pseudomonas picheuii 712, which was isolated from the same plot but at a different time, was found to carry a nearly identical plasmid to pKA2. The plasmid of this stain, pKA4, is 40.9 kb and shares common features with pKA2 such as high self-tansmissibility and hybridization only to the tfdA gene among 2,4-D metabolic genes of the 2,4-D degradative plasmid pJP4. The similar profiles of restriction endonuclease-generated fragments and the genetic homology shown by Southern hybridization between two plasmids indicate that these two 2,4- D degradative plasmids are closely related to each other, thus suggesting the occurrence of intergeneric gene tansfer in natural environments. 121 INTRODUCTION A number of catabolic plasmids have been isolated and described which enable host microorganisms to utilize xenobiotic compounds as their carbon and energy source. Most catabolic plasmids exist as extachromosomal genetic elements that replicate autonomously. However, there have been some reports that catabolic gene recombination occurs between the chromosome and plasmids under selective conditions. This is well documented with chloroaromatic degradative genes on the Pseudomonas TOL plasmid ; these are excised occasionally by reciprocal recombination between the direct repeats (15) and integrated into the host chromosome (2, 10, 21). The integrated genes are shown to be rescued by plasmids R2 and pMG18 to form novel recombinants (10). For the herbicide 2,4-D, all or most of the genes coding for the degradative enzymes have been reported to be contained on plasmids (3, 4, 17, 18). Unlike the aromatic hydrocarbon degradative genes on TOL (2, 10, 21, 24) and in Alcaligenes BR60 (27), 2,4-D degradative genes have not been shown to move either between chromosomal DNA and plasmid DNA or between plasmids in microorganisms which degrade 2,4-D. Furthermore, there have been no reports that the entire catabolic plasmid can be integrated into the host chromosome and then be excised to form a tee plasmid depending on the culture conditions. This phenomenon could play an important role in persistence and efficient gene expression of plasmid DNA in environments where resources fluctuate as well as in the evolution of plasmids and bacterial chromosomes. The increasing use of xenobiotic compounds may be stimulating a more widespread dissemination of the corresponding degradative plasmids among 122 microbial populations in nature. The tansfer of catabolic plasmids, such as the 2,4-D degradative plasmid pJP4 (5), on agar media is well known. However, naturally occurring horizontal gene tansfer between bacterial populations is dificult to detect due to the absence of the appropriate means for investigating such events in open environments. The horizontal tansfer of genes responsibile for 3-chlarobenzoate degradation in a natural aquatic community was suggested in a microcosm study, where the conditions in a shallow bay were simulated (6). Physically and genetically indistinguishable 2,4-D degradative plasmids were detected in different species of Alcaligenes isolated independently from soil (3), which provides indirect evidence of natural gene tansfer. Nonetheless, there have been few reports that supported degradative gene exchange between difl‘erent genera in open environments. In this paper we present the evidence of chromosomal integration of an entire 2,4-D degradative plasmid and demonstate that it can be excised either precisely or in a difl‘erent way. In addition, we present evidence of natural intergeneric tansfer of this plasmid between A. paradoxus and P. pickettii strains that were isolated from the same soil at two different times. MATERIALS AND METHODS Bacterial strains. Stains 2811P and 712 were isolated from the Long- Term Ecological Research (LTER) site of Kellogg Biological Station (KBS, Hickory Corners, MI) by enrichment for growth on 2,4-D as the sole source of carbon and identified as Alcaligenes paradaacus and Pseudomonas pickettii, respectively, by FAME analysis (11). Stain 28110 was isolated from a glycerol stock culture of stain 2811P kept at -20°C for two years and was 123 reconfirmed as Alcaligenes paradoxus by FAME. Pseudomonas cepacia DBO 1, which was obtained from Dr. R. Olsen (Univ. of Michigan), has Tn 5 tansposon and is resistant to kanamycin (75 rag/ml), bacitacin (50 mg/ml) and carbenicillin (50 mg/ml). Media and growth conditions. Peptone-typtone-yeast extact-glucose (PTYG) broth consisted of 0.25 g of peptone (Difco Laboratories, Detroit, MI), 0.25 g of typtone (Difco), 0.5 g of yeast-extact (Difco), 0.5 g of glucose, 0.03 g of magnesium sulfate, and 0.003 g of calcium chloride. Solid medium containing 1.5 % (wt/vol) agar was used for stain purification. MMO mineral medium (22) plus 500 ppm of 2,4-D was used to cultivate 2,4-D degrading stains to detect and isolate plasmid DNA. All cultures were incubated at 30°C. Broth cultures were aerated by shaking at 200 rpm in a New Brunswick G24 environmental incubator shaker (New Brunswick Scientific Co., New Brunswick, NJ). DNA isolation. Plasmid DNA was isolated by using the procedure of Hirsch et. al. (7 ). For the detection of plasmid DNA, cells were lysed as described by Kado and Lin (12) with some modifications. 2,4-D degrading bacteria were cultivated in 5 ml of MO mineral broth containing 500 ppm of 2,4-D at 30°C up to full growth. Cells were harvested by microcentifugation (14,000 rpm for 1 min), washed with 1 ml of sterilized distilled water and repelleted in a 1.5 ml Eppendorf microcentifuge tube. The pellet was resuspended in 30 ul of distilled water and lysed by adding 120 ul of lysing solution (50 mM Tris base, 3 % sodium dodecyl sulfate, pH 12.6). The solution was incubated at room temperature for 15 min, then heated at 80°C for 1 min in a water bath, and extacted with one volume of phenol-chloroform solution (1:1 vol/vol) saturated with TE buffer (10 mM Tris base, 1 mM EDTA, pH 8.0). The lysate was incubated overnight at room temperature. After microcentrifugation " I" "‘1' li- 124 (14,000 for 15 min), the aqueous phase was tansferred into a new Eppendorf tube and mixed with 1/5 volume of 5x loading dye (14). Samples of 110 ul were subjected to electophoresis in 0.7 % (wt/vol) agarose in Tris-acetate buffer (40mM Tris-acetate, 1mM Nag-EDTA). Gels were photographed on a model Foto/Prep I DNA Transilluminator (Fotodyne Inc.) with type 55 positive-negative film (Polaroid Corp.) Chromosomal DNA was isolated by the procedure of Watson et. al. (20) and subsequently purified by cesium chloride-ethidium bromide ultacentrifugation (14). Southern hybridization. To study hybridization patterns with 2,4-D metabolic genes (tfdA, tde, tde, and tde), plasmid and chromosomal DNA were prepared in crude lysate as described above, separated on horizontal 0.7 % agarose gel, tansferred to nitocellulose hybridization membranes (14), and hybridized with tfd gene probes. To analyse the restriction fi-agment profile, purified plasmid DNA and chromosomal DNA were digested with appropriate restriction endonucleases according to the manufacturer’s specifications. Digested DNA was size fractionated by electophoresis through horizontal 0.7 % (wt/vol) agarose gel and tansferred to nitocellulose hybridization membranes (14). As probes, internal segments of the tfd metabolic genes (8) and 23S rRNA gene (19) were labeled with 32P by using the Random Primed DNA Labeling kit (Boehringer Mannheim, Indianapolis, IN). The plasmid DNA probe was labeled with 32P by using the Nick Translation kit (Boehringer Mannheim). Prehybridization, hybridization and post-hybridization washes have been previously described (9). Hybridization signals were detected by autoradiography using Kodak X-omat AR film (Kodak, Rochester, NY) exposed at -70°C with a Quanta III (Sigma, St. Louis, MO) intensifying screen. If necessary, bound probe was stripped from the 125 membranes prior to rehybridization by washing for 15 min, 3 times with boiled distilled water containing 0. 1 % sodium dodesyl sulfate (w/v). Conjugation. Matings were performed on membrane filters as described by Willetts (26). Cells were grown in PTYG broth. An exponential broth culture (0.5 ml) of the donor stain was mixed with 1.0 ml of the recipient culture (P. cepacia DBO 1, exponential phase) in an Eppendorf tube. Cells were pelleted, resuspended in 50 ml of PTY G media and spread onto a filter membrane (0.45 um pore-size) which was plamd on PTYG agar. The filter was incubated overnight at 30°C and immersed in 2 ml of saline (0.85 % N aCI). Cells were resuspended by vortexing, diluted appropriately, and plated on 2,4-D minimum selective agar containing MMO mineral medum, 500 ppm of 2,4-D, kanamycin (75 mg/ml), bacitacin (50 mg/ml), carbenicillin (50 mg/ml), and 1.5 % Noble agar. Transconjugants were purified and their plasmid content was analyzed by agarose gel electophoresis of crude lysates. The donor viable count was measured by plating dilutions on PTYG plates and counting the distinctive colonies of the donor. The frequency of tansfer was calculated as the number of exconjugants per donor cell. RESULTS Properties of 2811P, 28110 and 712. Stains 2811P and 712 were isolated from the same plot at different times ; they contained a 2,4-D degradative plasmid pKA2 and pKA4 (Figure 1-A and Table 1), respectively, and grew with 2,4-D as the sole carbon source. Stain 28110, a derivative of stain 2811P, appeared not to have any tee plasmid (lane 3, Figure l-A), but had the whole pKA2 plasmid DNA integrated into chromosome. The presence of a free plasmid in 28110 could not be demonstated even by pulse field 126 electophoresis (23). Southern blot hybridization with 23S rRNA gene probe to the total DNA of stains 2811P and 28110 confirmed that these were identical stains as indicated by matching band patterns (Figure 2) except for the different location of a 2,4-D degradative plasmid. Southern blot hybridization with 2,4-D metabolic genes (tfdA, tde, tde, and tde) to the plasmid DNA and chromosomal DNA of these three stains revealed that these stains had sequence homoloy only with the tfdA gene (Table 1) and that the homologous sequence was located on plasmid DNA in stains 2811P and 712, whereas it was located on the chromosomal DNA in strain 28110 (Figure l-B). This indicated that the plasmid DNA not detected in the agarose gel was integrated into the host chromosome in stain 28110. These three stains showed a relatively long lag time (>50 h) and slow growth in 2,4-D minimal medium under uninduced condition (Figure 3). Stain 28110 had relatively longer lag time than stain 2811P in 2,4-D medium, presumably due to the chromosomal location (i. e., lower copy number) of a 2,4-D degradative plasmid. On the other hand, stains 2811P and 7 12, which were identified as difl'erent genera but had a nearly identical 2,4-D degradative plasmid, showed similar growth patterns in 2,4-D medium. Plasmid transfers. The possibility of tansfer of the 2,4-D degradative genes was tested by filter mating. In three independent experiments, tansfer of pKA2 from strain 2811P to P. cepacia DBOl was obtained at an average frequency of 7.7 x 10-3 per donor colony formed after the mating period (Table 2). Under similar conditions, the 2,4-D+ phenotype was not tansferred at a detectable frequency (<10‘9) with stain 28110, supporting the interpretation that the 2,4-D genes were not contained on a free conjugative plasmid in this stain. Plasmid pKA4 of stain 712 was tansferred at an average frequency of 1.0 x 10'2 per donor cell (Table 2). 127 Plasmid bands exhibiting identical electophoretic mobilities were observed in agarose gels of the tansconjugants and their respective donors (arrowheads, Figure 4). A cryptic plasmid DNA band of recipient (P. cepacia DBO 1) was also observed in agarose gels of the tansconiugants and mcipients. Physical evidence for the integration of plasmid into chromosomal DNA. To investigate the fate of pKA2 in stain 28110, plasmid pKA2 and total DNA of stain 2811P and stain 28110 were digested with two different restriction enzymes separately, and analyzed on Southern blots hybridized with 32P-labeled pKA2 plasmid DNA (Figure 5). As expected, all the seven hybridized fi'agments of pKA2 (Figure 5) were also observed in the EcoRI- digested total DNA of stain 2811P that contained the same plasmid, in addition to an unexpected novel fragment (0A, Figure 5-A). The novel fragment was not observed in the digested plasmid DNA but observed in the digested total DNA of both 2811P and 28110. For this reason, it was believed to be derived from the digested host chromosomal DNA. This was further supported by the fact that its radioactive signal was much weaker than the signal with the plasmid DNA fragment of similar size due to its lower copy number in total DNA. On the other hand, in EcoRI-digested total DNA of stain 281 10, one fragment (ED) disappeared with concomitant appearance of three extra hybridized fragments (arrowheads, Figure 5-A). This observation indicates that the whole pKA2 plasmid was integrated into the host chromosomal DNA without the loss in 2,4-D+ phenotype, although normally the appearance of only two additional bands is expected with the loss of one band with chromosome integrations. The loss of one hybridized fragment (HB) with concomitant appearance of three novel fragments (arrowheads) was also observed in HindIII digests (Figure 5). 128 Excision of pKA2 plasmid. The existence of pKA2 plasmid as both an independent plasmid in 2811P and a chromosomally integrated plasmid in 28110, together with the observed difl'erence in their growth patterns on 2,4- D medium, suggested that the population containing a free plasmid might be derived from stain 2811C and that it could be preferably amplified under selection imposed by growth on 2,4-D. To investigate this hypothesis, stain 28110 was cultivated in 2,4-D broth with repeated tansfers into fresh medium. After the seventh tansfer, a weak plasmid band was detected on the agarose gel. The intensity of this band was increased with increasing number of tansfers. Southern hybridization of the gel with tfdA exhibited a clear hybridized plasmid band from the seventh tansfer and the radioactive intensity was much increased after the nineth and the eleventh tansfer (Figure 6). After longer film exposure, weak hybridized bands were observed in chromosomal DNA band on Southern blots of the original culture and in cultures after the seventh, the nineth, and the eleventh tansfer (data not shown). With further tansfers on 2,4-D medium, two difl‘erent plasmid bands were detected on agarose gel from the twentieth tansfer culture, suggesting that the integrated plasmid could excise in several different ways during the continuous growth on 2,4-D. These results indicated that initial pure culture of stain 28110 became a mixed culture of 28110 and at least two other plasmid-containing stains that are probably more fit for 2,4-D growth. To estimate the ratio of population 28110 to the population of the excised plasmid-containing stain, colonies from cultures after the eighth and the fourteenth tansfer were randomly selected, purified on PTY G plates, and subjected to Kado’s procedure. Whereas no plasmid DNA band was detected from the 30 colonies examined from the eighth tansfer, clear plasmid DNA 129 bands were observed in five colonies out of the 28 colonies examined from the fourteenth tansfer. The result suggeswd that the mixed population was mainly composed of stain 28110 even at the eighth tansfer, while plasmid- containing stains have gradually become more predominant and comprised approximately 18 % of the population after the fourteenth tansfer. Plasmid DNA was isolated from three colonies selected randomly from the five colonies to analyze the fragment profiles of restriction endonucleases digests. Two of three plasmids showed identical restriction profiles to plasmid pKA2 in 15le and HindIII digests (Figure 7 ), while the third plasmid revealed deletion of four fragments as well as appearance of an exta novel fragment (arrowhead) in EcoRI digests (Figure 7). The derivative plasmid could result from either an alternative excision event of the integrated pKA2 in stain 28110 or rearrangement of the precisely excised plasmid pKA2 during continuous selection on 2,4-D. Comparison of pKA2 of A. pat-adorns with pKA4 of P. pickettii. As stated above, plasmid pKA2 of A. paradoxus has similar properties to plasmid pKA4 of P. pickettii in regard to size, high tansmissibility and hybridization pattern with tfd probes. In addition to that, restriction enzyme analysis with EcoRI and HindIII revealed the restriction pattern similarity between these two plasmids (Figure 8-A). In EcoRI digests, only two (EB and E0) of the seven fragments showed altered mobilities in the agarose gel electophoresis, while one (HB) of the three fragments revealed different mobility in HindIII digests (Figure 8-A). When fragments EA, ED and EE of both plasmids (pKA2 and pKA4) were isolated from the agarose gel and digested with BglII, the new fragments were identical for each plasmid (data not shown). To determine whether all the DNA fragments of plasmids pKA2 and pKA4 generated by 13le restriction endonuclease are homologous to one another, 130 the fragments were analyzed on Southern blots hybridized with pKA4 DNA as the probe (Figure 8-B). All the fragments of pKA2 were shown to be hybridizable with pKA4. The identical restriction profiles and hybridization patterns demonstate that pKA2 of A. pal-adorns and pKA4 of P. pickettii are genetically homologous, suggesting that intergeneric gene tansfer between these two stains has occurred in natures. DISCUSSION The data presented here demonstated that the 2,4-D degradative plasmid pKA2 can exist as both an integrated plasmid into the host chromosome and a free plasmid in a natural isolate, A. paradoaus. Since, for the herbicide 2,4- D, all or most of the degradative genes have been reported to be carried on a free plasmid in a number of independent natural isolates (3, 4, 17 , 18), the observation of its chromosomal location was surprising. The integrated plasmid could be precisely excised, thus forming the same plasmid pKA2. The selective advantage of strain 2811P over stain 28110 in 2,4-D minimal medium (Figure 3) might make it possible to detect the 2811P population which may normally arise from stain 28110 at a low frequency. In any case, the detection of stain 28110 in glycerol stock culture of 2811P and the reproduction of 2811P tom 28110 on 2,4-D medium stongly suggest that the entire plasmid pKA2 can move into and out of the host chromosome depending on its environment. However, we have been unable to demonstate the reproducibility of plasmid integration into chromosome due to the absence of any selection for this variant. It has been reported that TOL genes could be integrated into and rescued from the bacterial chromosome (2, 10, 21). In those cases, part of the TOL 131 plasmid was located in the chromosome and the integrated DNA was rescued only in the presence of other plasmids to form novel recombinants, suggesting that the TOL genes may be carried on a tansposon-like element( 1, 10, 24). It has also been postulated that repeated sequences were present and involved in gene expression in several 2,4-D degradative plasmids (28) and that repeated sequences were involved in the complex interaction between chromosomal DNA and plasmid DNA in Acetobacter xylinum (25). It is important to note that part of the chromosomal DNA of stains 281 1P and 28110 has sequence homology with plasmid pKA2 (Figure 5). Although it is not known whether the process is dependent of a host recA product, the hybridization results and the occurrence of the precise excision of pKA2 suggest that homologous repeated sequences may be involved in the interaction between host chromosome and plasmid pKA2. While chromosomally located plasmid pKA2 has lost the fi'eedom of conjugal tansfer possessed by the tee plasmid, there may be some compensations which come from the chromosomal location. First, it would allow the stable maintenance of plasmid DNA in any circumstance, while the stability of a free plasmid is influenced by cultural conditions, presence of other incompatible plasmids, the efficiency of partitioning into daughter cells, and the environment. Second, the chromosomally located plasmid pKA2 could be available to other populations of bacteria in good times because it still retains a means of tansferability by virtue of its ability to excise followed by conjugation. Third, it may play an important role in the exchange of bacterial genome because an alternative excision could allow for conjugal tansfer of chromosomal DNA as well as plasmid DNA to other microbial populations. In fact, the observation that some exta fragments of the digested chromosomal DNA were hybridizable with pKA2 plasmid DNA 132 would indicate that DNA exchange between chromosome and plasmid occurred in the past. It is also interesting to see how chromosomally located pKA2 excises during continuous growth on 2,4-D medium. In the smaller derivative plasmid, a 19.6-kb region corresponding to four fragments of £2le digests has been deleted with cancomittant appearance of a novel fragment (Figure 7 ). The identical profile of the other fragments generated by EcoRI digestion between pKA2 and the smaller derivative affords a striking illustation of two different plasmids with the same origin and demonstate that such rearrangement of DNA plays an important role in the evolution of plasmids. We isolated two different 2,4-D degrading bacteria belonging to the genera of Alcaligenes and Pseudomonas from the same soil at two different times. The observation that plasmid pKA2 of A. paradoxus isolated in 1989 showed nearly identical restriction profile to that of pKA4 of P. pickettii isolated in 1990, together with their similar physical properties and genetic homology, stongly suggests that intergeneric gene tansfer followed by plasmid modification, or vice versa, has occurred between these two species in nature. While it is not clear why and how pKA2 is integrated into and excised from the chromosome, our results demonstrated that a conjugative plasmid can integrate in its entirety into the host chromosome, and that the chromosomally located plasmid can be excised either precisely or in a different way. Additionally, we presented the evidence that intergeneric gene tansfer occurred between Alcaligenes and Pseudomonas species in nature. Future research on the different physiology of stains 2811P and 28110 on 2,4-D medium, the integration site on chromosome of pKA2, and the mechanism of integration and excision would reveal the mechanism and strategy for this genetic flexibility. 133 ACKNOWLEDGMENTS This work was supported by National Science Foundation grants from Long-Term Ecological Research program and the Center for Microbial Ecology. We thank Dr. William E. Holben for his valuable suggestions on this study and Miss Cathy McGown for her information on conjugation experiment. 134 Table 1. Sequence homology of plasmid and chromosome with 2.4-D metabolic genes. 2.4-D degradative aHomology to Location of Strain plasmid (size) g‘dA rde y‘dC lde hybridization signal P. pickettii 712 pKA4 (40.9 kb) + - - - plasmid A. paradoxus 2811P pKA2 (42.9 kb) + - - - plasmid A. paradoxus 281 1C none + - - - chromosome a +, has sequence homology : -, does not have sequence homology. 135 Table 2. Transferability of 2,4-D+ phenotype from donor to P. cepacia DBO 1. Plasmid Antibiotic Transfer Donor transferred selectiona (ug/ml) frequencyb Km, 75 ; Ch, 50 ; P. pickettii 712 pKA4 Be, 50 1.0 x 10‘2 A. paradoxus Km, 75 ; Ch, 50 ; 2811P pKA2 Bc, 50 7.7 x 10'3 A. paradoxus Km, 75 ; Ch, 50 ; 2811C - Bc. 50 < 10'9 a Km, kanamycin ; Cb, carbenicillin ; Bc, bacitacin. 1’ Values are means of three independent matings. 136 Figure legends Fig. 1. Resolution of plasmids of 2,4-D degrading strains. (A) Negative photograph of agarose gel and (B) autoradiogram of the gel after hybridization with tfdA gene as a probe. Lanes : P. pickettii 712 containing plasmid pKA2 (1), A. paradoxus 2811P containing plasmid pKA4 (2), A. paradoxus 2811C not containing plasmid DNA (3). Positions of plasmid DNA (P) and chromosomal and linear DNA (Chr) are shown. Films were exposed for (a) 20 min (b) 2 days. Fig. 2. Southern blot hybridization with the 16S rRNA gene to the total DNA of strains 2811P and 2811C. Lanes : Total DNA of 281 1C (1 and 3) and 2811P (2 and 4) digested with Hindlll (l and 2) and EcoRI (3 and 4). Fig. 3. Growth characteristics of 2,4-D degrading strains in 2.4-D mineral medium under uninduced condition. P. pickettii 712 ( —A—), A. paradoxus 2811P (—CI—). and A. paradoxus 2811C (—l—). Fig. 4. Resolution of plasmids in donors, transconjugants, and recipient. Lanes : recipient P. cepacia DBOl (l), transconjugant from 712 x DBOl (2), donor P. pickettii 712 (3), transconjugant from 2811P x DBOl (4), donor A. paradoxus 2811P (5). Fig. 5. (A) Line drawing and (B) autoradiogram of a Southern blot hybridization with labeled plasmid pKA2 of EcoRI (left) and Hindlll (right) restriction enzyme digests of pKA2 plasmid DNA (1), total DNA of strain 2811P (2), and total DNA of strain 2811C (3). Novel extra bands presumably resulting from intergation process are indicated by arrowheads in panel (A). 137 Fig. 6. Excision of plasmid DNA in strain 2811C during continuous growth on 2.4—D medium Each 5 ml of culture grown initially (1) and after the seventh (2), nineth (3), and eleventh (4) transfer was subjected to Kado's procedure and then plasmid and chromosomal DNA in lysate were gel-separated, and analyzed on Southern blots hybridized with labeled rfdA probe. Film was exposed for 4 h. Fig. 7. Negative photograph of agarose gel of restriction enzyme-digested plasmid DNA. Lanes : Lambda DNA digested with HindIII as size marker (1), plasmid pKA2 . (2), plasmids isolated from three independent colonies recovered on 2.4-D selection from strain 2811C (3, 4, and 5). Fig. 8. (A) Negative photograph of agarose gel of plasmids digested with restriction endonucleases. Lanes : Plasmid pKA2 (1) and plasmid pKA4 (2). (B) Autoradiogram of a Southern blot hybridized with labeled plasmid pKA4 of EcoRI-digested pKA2 (1) and pKA4 (2). C r Figure 1. 138 23.1 > a... 65> «- 43> 139 .‘t s. . .- 1.. , ‘.‘..‘:?D.c-‘ 3;“. -' 5 Figure 2. 140 OD at 550nm Pm PH T ch 141 gure4 Fi kb 23.1 ... 9.4 ... 6.5 .— 4.3 — 2.3 '- 2.0— 0.5 — EA.— EB— EC— ED— EE — EF— EG— CA— 142 I A HB .__ _ :— cs— =3 —< HC — — — Figure 5-A. 143 2 312 3 1 ~“ Figure 5-B. 144 Figure 6. 145 Hindlll 2345 EcoRl 2 3 4 5 1 1 Figure 7. 146 “'1' . F" . 3....‘3 I303! .. on_--.—.m-- -- -. O ‘6‘ I" _ . (4 a .’ (‘o- 5"“; .’f.-'-.‘§ :: f ‘?\ :2~,"_ 3: . .... 5.1.1:: . -‘ '. . . w . . fifths 5:, iirtzz J.,, o . O" ', , ‘ '. , . . _ - '_~ . , ”dz. o‘v . I. . .U r '_ .‘ ‘ , * ' ‘ “T. ' 3...... 1;” h .. (‘3‘ ‘.,‘.. . ‘ I . ‘ Q. . - ' . .1‘ . I". -"L ‘..f ‘ m' 4'r . 715.1):W @ r ‘ "‘7'" "*va " ‘ *Vlr' 15.3313? 3': ‘ 1.?!- ~ P. e . . ‘r: 7'{ » if? “(’51, x "‘ ‘ ”ff-”m” Jk - : ‘ .- 3:1??3‘11} .. . it. ’I. .- .0345“) ‘ v- 1.” “1» 71?? l *1 . -‘ K‘ “ ' 4.3.1.3. N3. '. o . ' .':' ’ ' ' b' n‘i . a. . ' p- x o ‘7 X“ " ... . I “ .§. ' «‘7‘ '3 9111333 LIST OF REFERENCES 147 LIST OF REFERENCES 1. Cane, P. A., and P. A. Williams. 1986. A restriction map of naphthalene catabch plasmid pWW60-1 and the location of some of its catabch genes. J. Gen. Microbial. 132:2919-2929. 2. Carney, B. F., and J. V. Leary. 1989. Novel alterations in plasmid DNA associated with aromatic hydrocarbon utilization by Pseudomonas putida R5—3. Appl. Environ. Microbiol. 55: 1523-1530. 3. Don, R. H., and J. M. Pemberton. 1981. Properties of six pesticide degradation plasmids isolated from Alcaligenes paradaxus and Alcaligenes eutraphus. J. Bacterial. 145:681-686. 4. Fisher, P. B., J. Appleton, and J. M. Pemberton. 197 8. Isolation and characterization of the pesticide-degrading plasmid pJPl from Alcalzgenes paradaxus. J. Bacteriol. 135:798-804. 5. Friedrich, B., M. Meyer, and H. G. Schlegel. 1983. Transfer and expression of the herbicide-degrading plasmid pJP4 in aerobic autatraphic bacteria. Arch. Microbiol. 134:92-97. 6. Fultharpe, R. B., and R. C. Wyndham. 1989. Survival and activity of . 3—chlorobenzaate-catabalic genotype in a natural system. Appl. Environ. Microbial. 55:1584-1590. 7. Hirsch, P. B., M. VanMontagu, A. W. B. Johnston, N. J. Brewin, and J. Schell. 1980. Physical identification of bacteriocinogenic, nodulation and other plasmids in strains of Rhizobium leguminasarum. J. Gen. Microbial. 120:404-412. 8. Holben, W. E., B. M. Schroeter, V. G. M. Calabrese, R. H. Olsen, J. K. Kukor, V. 0. Biederbeck, A. E. Smith, and J. M. Tiedie. 1992 Analysis by gene probes of soil microbial populations selected by treatment with 2,4—dichloraphenoxyacetic acid (2,4-D). Submitted for publication. 9. Holben, W. E., J. K. Jansson, B. K. Chelm, and J. M. Tiedie. 1988. DNA probe method for the detection of specific microorganisms in the sail bacterial community. Appl. Environ. Microbiol. 54:703-711. 10. Jeenes, D. J., and P. A. Williams. 1982. Excision and integration of 148 degradative pathway genes from TOL plasmid pWWO. J. Bacteriol. 150: 188- 194. 11. Ka, J. 0., W. E. Holben, and J. M. Tiedie. 1992. Studies on diversity of 2,4-D degrading bacteria isolated from Kellogg Biological Station Sails. Submitted for publication. 12. Kado, C. L, and S. -T. Liu. 1981. Rapid procedure for detection and isolation of large and small plasmids. J. Bacterial. 145:1365-137 3. 13. Latarre, J., W. Reineke, and H. -J. Knackmuss. 1984. Microbial metabolism of chloroanilines : enchanced evolution by natural genetic exchange. Arch. Microbiol. 140:159-165. 14. Maniatis, T., E. F. Fritsch, and J. Sambroak. 1982. Molecular cloning : a laboratory manual. Cold Spring Harbor Laboratory, Cold Spring Harbor, N. Y. 15. Meulien, P., and P. Broda. 1982. Identification of chromosomally integrated TOL DNA in cured derivatives of Pseudomonas putida PAWl. J. Bacterial. 152:911-914. 16. Meulien, P., R. G. Downing, and P. Broda. 1981. Excision of the 40 kb segment of the TOL plasmid from Pseudomonas putida mt-2 involves direct repeats. Mol. Gen. Genet. 184:97-101. 17 . Pemberton, J. M., and P. R. Fisher. 1977. 2,4-D plasmids and persistence. Nature (London) 268:732-733. 18. Pemberton, J. M., B. Carney, and R. H. Dan. 1979. Evolution and spread of pesticide degrading ability among soil microorganisms, p. 287- 299. In K. N. 'I‘immis and A. Puhler (ed.), Plasmids of medical, environmental and commercial importance. Elsevier/Narth-Holland Biomedical Press, Amsterdam. 19. Regensburger, A., W. Ludwig, and K. H. Schleifer. 1988. DNA probes with different specificities from a cloned 23S rRN A gene of Micracaccus luteus. J. Gen. Microbiol. 134:1197-1204. 20. Schafield, P. B., M. A. Djardievic, B. G. Rolfe, J. Shine, and J. M. Watson. 1983. A molecular linkage map of nitrogenase and nodulation genes in Rhizabium trifalii. Mol. Gen. Genet. 192:459-465. 21. Sinclair, M. 1., P. C. Maxwell, B. R. Lyon, and B. W. Holloway. 1986. Chromosomal location of TOL plasmid DNA in Pseudomonas putida. J. Bacterial. 168:1302-1308. 22. Stanier, R. Y., N. J. Palleroni, and M. Doudorofl. 1966. The aerobic pseudomonads : a taxonomic study. J. Gen. Microbial. 43: 159-271. 149 23. Kuritz, T., E. Anneliese, A. B. Todd, and C. P. Walk. High resolution mapping of the chromosome of Anabaena sp. strain PCC 7120. Submitted for publication. 24. Tsuda, M., and T. Lino. 1987 . Genetic analysis of a transposon carrying toluene degrading genes on a TOL plasmid pWWO. Mol. Gen. Genet. 210:270-276. 25. Valla, 8., D. H. Caucheron, and J. Kiosbakken. 1987 . The plasmids of Acetobacter xylinum and their interaction with the host chromosome. Mol. Gen. Genet. 208:76-83. 26. Willetts, N. S. 1988. Conjugation In Methods in Microbiology. vol. 21, Academic Press, London. 27 . Wyndham, R. C., R. K. Singh, and N. A. Strauss. 1988. Catabolic instability, plasmid gene deletion and recombination in Alcaligenes sp. BR60. Arch. Microbiol. 150:237-243. 28. You, I. S. 1992. Regulation of plasmid-borne 2,4-dichlorophenoxyacetate catabolic genes involves multiple regulatory genes. 92nd Annual meeting of SR? American Society for Microbiology 1992, New Orleans, Louisiana, U "7'1? llllllfljll'lljlll