FEMS Microbiology Letters 100 (1992) 101-106 © 1992 Federation of European Microbiological Societies 0378-1097/92/$05.00 Published by Elsevier
101
FEMSLE 80051
The new approaches to whole genome analysis of bacteria B.W. Holloway, M.D. E s c u a d r a , A.F. M o r g a n , R. Saffery and V. Krishnapillai Department of Genetics and Developmental Biology, Monash University, Clayton, Victoria, Australia Received 2 June 1992 Accepted 21 July 1992
Key words: Bacterial genome; Pulsed field gel electrophoresis; Combined physical/genetic maps; Pseudomonas aeruginosa ; Pseudomonas putida ; Pseudomonas solanacearum ; Ribosomal genes 1. S U M M A R Y A range of recombinant DNA techniques now enables whole genome analysis of any bacterium to be carried out without recourse to the classical means of bacterial genetic exchange. Using enzymes which cut infrequently, such as SpeI, combined with pulsed field gel electrophoresis, a physical map of ordered fragments can be constructed. By means of cloned fragments of known genes or oligonucleotides synthesized using data from DNA or protein sequence banks, the location of individual genes on this map can be determined. We have used these techniques to study whole genome structure in three species of Pseudomonas: P. aeruginosa, P. putida and P. solanacearum.
2. I N T R O D U C T I O N A comprehensive review of chromosome organization in bacteria has recently been published
Correspondence to: B.W. Holloway, Department of Genetics and Developmental Biology, Monash University, Clayton, Victoria 3168, Australia.
[1]. The best known bacterium from a genetic viewpoint is undoubtedly Escherichia coli K12 with the positions of some 1403 loci known [2], almost all mapped using conjugation and transduction. The introduction of pulsed field gel electrophoresis (PFGE) in its various forms has enabled the separation of linear fragments of double stranded DNA up to several megabases. The identification of enzymes which cut bacterial chromosomes infrequently has permitted the construction of physical maps in which the location of restriction enzyme sites can be mapped relative to each other. For E. coli, there are 27 NotI and 22 Sill sites giving a genome size of 4700 kb [31. Other techniques can be combined to extend the sophistication of any combined genetic and physical map. These include DNA probing, which could involve use of either cloned material, synthetic oligonucleotides or polymerase chain reaction (PCR) generated sequences combined with the use of data retained in the increasingly comprehensive DNA and protein data banks. One of the most attractive features of a combined approach with this range of techniques is that it is equally applicable to all bacteria and independent of the existence of conjugation, transduction or transformation systems. To date,
102 combined genetic and physical maps have been published for a range of bacteria [1] including E. coli [3], Pseudomonas aeruginosa [4], Rhodobacter sphaeroides [5] and Caulobacter crescentus [6]. T h e p s e u d o m o n a d s are a h e t e r o g e n e o u s group of bacteria, with important genetic differences having b e e n d e m o n s t r a t e d between them and those other bacteria which have b e e n genetically analysed. These differences include c h r o m o s o m a l gene arrangement, and the role of plasmids in phenotypic characterization [7]. The need for genetic systems of analysis in a wide variety of species and strains of Pseudomonas, in contrast to the experimental concentration on one strain of one species in Escherichia, created a need for systems promoting genetic exchange which were effective over a range of isolates. Plasmids such as R68.45 and R91-5 [7] solved this problem to some extent, but extensive work with different species and strains d e m o n s t r a t e d that these plasmids diffe'red in their effectiveness with different strains. For example, R91-5 was excellent for c h r o m o s o m e m a p p i n g in P. putida P P N [8] but less effective for P. syringae [9] and in other strains of P. putida (M.I. Sinclair, personal c o m m u n i c a t i o n ) and neither R91-5 nor R68.45 have been of any use with P. solanacearum (M.D. Escuadra and B.W. Holioway, unpublished data). T h e use of P F G E combined with probing and interspecific c o m p l e m e n t a t i o n using clones from a genomic library has resulted in the genetic analysis of a range of p s e u d o m o n a d s which has enabled a m o r e precise comparison of their g e n o m e structure. Even so, the strategy a d o p t e d for the physical analysis of the P. aeruginosa genome, the species first selected, has not been always readily adaptable to other species of this genus. Ratnaningsih et al. [4] have constructed a combined genetic and physical map of P. aeruginosa PAO. In an i n d e p e n d e n t analysis, R6mling and Tiimmler [10] have also constructed a physical map of P. aeruginosa P A O using two-dimensional P F G E , and, while the physical map so constructed is largely similar to that of Ratnaningsih et al. [4], differences do exist. These have b e e n largely identified and a c o m m o n system of nomenclature a d o p t e d [11].
3. R E S U L T S The range of techniques which have proved to be successful for P. aeruginosa g e n o m e analysis is now being extended in P. aeruginosa and applied to two other species of Pseudomonas: P. putida, an important organism in biotechnology; and P. solanacearum, a plant pathogen.
3.1. P. aeruginosa By a variety of techniques, a n u m b e r of genes have now been located on SpeI fragments, the combined physical and genetic map being ext e n d e d and refined. In addition to these reported in Holloway et al. [12], other genes have been
Table 1 Location of genes on the physical map of P. aeruginosa PAO by probing with cloned DNA. The nomenclature of Tiimmler et al. [13] is used for Spel fragment identification Gene Character Symbol
Spel
oriC3m origin of chromosome K replication oriC3o I origin of chromosome replication pilT twitchingmotility (fimbriae) nid/ccf nitrate reductase/ cytochrome C551 pt,d pyoverdinsynthesis lpp membrane lipoprotein braB branched amino acid transport braC branched amino acid transport braZ branched amino acid transport apr alkalineprotease fim A fimbrialsubunit
Reference
fragmentation location 10, 22
K
22
H
23
H
24, 25 ~'
H, F V Q or R
26, 27 b.c,~ 28 c 29, 30 f
U
29, 31 f
D
32 1
M E
33 ~ 23, 34
a Brunori, M. and Silvestrini, M.C. (personal communication). b Meyer, J.M. (personal communication). c H6fte, M. (personal communication). J Visca, P. (personal communication). c van de Lelie, D. (personal communication). f Hoshino, T. (personal communication). g Lazdunski, A. (personal communication).
103
located on SpeI fragments by probing with cloned material as shown in Table 1.
3.2. P. putida PPN A circular chromosomal map has been constructed using Hfr donor strains generated by integration of Tn501 and T n 5 derivatives of the narrow host range P. aeruginosa plasmid R91-5, which acts as a natural suicide vector in P. putida [8,13,14]. The genome size has been shown to be 5960 kb, some 400 kb larger than P. aeruginosa PAO and considerably larger than the 4400 kb reported by Grothues and Tfimmler [15] for P. putida DSM50291. While circularity of the physical map has not yet been established, it is clear that most if not all of the genomic DNA is chromosomal. Interestingly, consideration of genetic (88 min) and physical size (5960 kb) of the P. putida PPN chromosome yields a conjugal transfer rate of 68 kb/min, compared to that for E. coli of 47 kb/min, even though the optimum growth temperature for this strain of P. putida is 28°C. It has proved to be much more difficult to determine the linkage relationships of SpeI fragments in P. putida PPN than was the case in P. aeruginosa. In the first place there are rather more fragments compared to P. aeruginosa (44 compared tO 37) and the complexities of the system increase geometrically with increasing fragment number. Secondly, attempts to construct a Spel junction fragment library using methods developed by Ratnaningsih et al. [4] for P. aeruginosa PAO have led to an extremly biased library. More than 90% of such clones undergo considerable rearrangement immediately following construction. This is possibly due to the different methylation patterns of the two species, revealed by the observation that while DpnI (which recognizes only methylated DNA) cuts P. aeruginosa PAO DNA, it does not cut P. putida PPN genomic DNA. To overcome this a K m r / SpeI cassette with SpeI cohesive ends has been constructed which allows any plasmid vector to be used for positive selection of SpeI junction fragments. The physical data map available has been sufficient for an analysis of ribosomal genes in P.
putida. By probing fragments of the P. putida genome separated by SpeI digestion and P F G E with cloned 5S, 16S or 23S rDNA genes of P. aeruginosa, seven distinct operon-like clusters of ribosomal genes have been identified on the P. putida chromosome. It has not yet been shown that all such clusters are functional. Mapping of these genes has been facilitated by the fact that all seven P. putida PPN 23S genes contain a conserved SpeI site.
3.3. P. solanacearum Little is known about the genome structure of
P. solanacearum, the cause of bacterial wilt in a wide range of plant genera. Boucher et al. [16] have constructed a circular chromosome map of strain GMI1000 containing 19 markers using conjugation promoted by RP4-prime plasmids. Hence this species is well suited to a combined physical approach to genome mapping. P F G E of P. solanacearum has so far identified 30 SpeI fragments, giving a genome size of 5548 kb. It is known that the strain studied contains large plasmid(s) and as their actual size has not been determined, the bacterial chromosome size will be less than this. As with P. putida, the construction of a SpeI linking clone library has not proved as straightforward as was the case with P. aeruginosa, and only 15 of the potential 30 fragment linkages have been identified. To construct a genetic map in P. solanacearum, a cosmid library of this bacterium was made using the wide host range cosmid vector pLA2917 [17]. Auxotrophic markers located on individual cosmids were identified in two ways. Firstly, by complementation of a range of auxotrophic mutants isolated in P. solanacearum, although in most cases these mutants have not been characterized any further than identifying the amino acid requirement. Secondly, this cosmid library was used to complement a range of P. aeruginosa auxotrophs many of which have been characterized with respect to the enzyme deficiency [18]. The interspecific and intergeneric identification of markers by this approach has been used extensively for P. putida [13,19] and has been readily adaptable to P. solanacearum. Once markers have been identified on individ-
104
ual cosmids, then such cosmids can be used as probes to relate the location of markers to individual SpeI fragments and to date, 38 markers have been mapped in this way. Combined PFGE and DNA probing techniques have been used to locate genes affecting extracellular polysaccharide and polygalacturonate synthesis.
4. DISCUSSION These are early days in the whole genomic analysis of bacteria, particularly in those organisms for which there are no characterized systems of genetic exchange or even collections of mutants [20]. Over 100 different bacterial isolates representing different genera, species and strains have been analyzed by PFGE indicating the broad applicability of this technology. The ever increasing size and availability of DNA and protein data bases combined with PCR to generate probe DNA, will permit the physical mapping of any DNA sequence in the species of origin. Depending upon the degree of sequence conservation between related and unrelated bacteria, this will allow gene mapping in a wider range of organisms of interest. While construction of a physical and genetic map is an achievement in itself, the importance of that achievement will be enhanced by the use of the data made available. An example of such use was provided by Sinclair and Holloway [21] who demonstrated highly specific integration sites of the TOL transposons Tn4651 and Tn4653 into the chromosome of P. aeruginosa PAO. As an example of the use of these techniques for mapping genes which could not be mapped by conventional means, Smith et al. [22] have described the precise mapping of replication origin sites in P. aeruginosa. Given the labour-intensive nature of finding, characterizing and using classical systems of genetic exchange in bacteria, these combined techniques should result in an explosion of genome data which will be of general application to microbial genetics, evolutionary studies, applied microbiology and biotechnology.
ACKNOWLEDGEMENTS Work in the authors' laboratory is supported by the Australian Research Council, the Australian Centre for International Agricultural Research and the Celgene Corporation.
REFERENCES [1] Krawiec, S. and Riley, M. (1990) Microbiol. Rev. 54, 502-539. [2] Bachman, B.J. (1990) Microbiol. Rev. 54, 130-197. [3] Smith, C.L., Econome, J.G., Schutt, A., Klco, S. and Cantor, C.R. (1987) Science 236, 1448-1453. [4] Ratnaningsih, E., Dharmsthiti, S., Krishnapillai, V., Morgan, A., Sinclair, M. and Holloway, B.W. (1990) J. Gen. Microbiol. 136, 2351-2357. [5] Suwanto, A. and Kaplan, S. (1989) J. Bacteriol. 171, 5850-5859. [6l Ely, B., Ely, T.W., Gerardot, C.J. and Dingwall, A. (1990) J. Bacteriol. 172, 1262-1266. [7] Holloway, B.W. and Morgan, A.F. (1986) Microbiol. Rev. 40, 79-105. [8] Dean, H.F. and Morgan, A.F. (1983) J. Bacteriol. 153, 485 -497. [9] Nordeen, R.O. and Holloway, B.W. (1990) J. Gen. Microbiol. 136, 1231-1239. [10] R6mling, U. and Tiimmler, B. (1991) Nucleic Acids Res. 19, 3199-3206. [11] Tfimmler, B., R6mling, U., Ratnaningsih, E., Morgan, A.F., Krishnapillai, V. and Holloway, B.W. In: Pseudomonas: Molecular Biology and Biotechnology. (Galli, E., Silver, S. and Witholt, B., Eds.), Am. Soc. Microbiol. Washington, DC. in press. [12] Holloway, B.W., Bowen, A., Dharmsthiti, S., Krishnapillai, V., Morgan, A., Ratnaningsih, E. and Sinclair, M. (1991) In: Proc. 6th Int. Symp. Genetics, Industr. Microorg. (Heslot, H., Davies, J., Florent, J., Bobichon, L., Durand, G. and Penasse, L., Eds.), 1, 227-238. Soc. Francaise de Microbiol., Paris. [13] Morgan, A.F. and Dean, H.F. (1985) J. Gen. Microbiol. 131,885-896. [14] Strom, D., Hirst, R., Petering, J. and Morgan, A. (1990) Genetics 126, 497-503. [15] Grothues, D. and Tfimmler, B. (1991) Mol. Microbiol. 5, 2763-2776. [16] Boucher, C., Arlat, M., Zischek, C. and Boistard, P. (1988) In: Physiology and Biochemistry of Plant-Microbial Interactions. (Keen, N.T., Kosuge, T. and Walling, L.L., Eds.), pp. 83-95. The American Society of Plant Pathologists, Rockville. [17] Allen, L.N. and Hanson, R.S. (1985) J. Bacteriol. 161, 955-963.
105 [18] Holloway, B.W. and Zhang, C. (1990) Genetic Maps 5, 2.71-2.78. [19] Morgan, A.F. (1986) J. Bacteriol. 149, 654-661. [20] Smith, C.L. and Condemine, G. (1990) J. Bacteriol. 172, 1167-1172. [21] Sinclair, M.I. and Holloway, B.W. (1991) J. Gen. Microbiol. 137, 1111-1120. [22] Smith, D.W., Yee, T.W., Baird, C. and Krishnapillai, V. (1991) Mol. Microbiol. 5, 2581-2587. [23] Whitchurch, C.B., Hobbs, M., Livingston, S.P., Krishnapillai, V. and Mattick, J.S. (1990) Gene 101, 33-44. [24] Silvestrini, M.S., Galeotti, C.L., Gervais, M., Schininia, E., Barra, D., Bossa, F. and Brunon, M. (1989) FEMS Microbiol. Lett. 254, 33-38. [25] Nordling, M., Young, S., Karlsson, B.G. and Lundberg, L.G. (1990) FEMS Microbiol. Lett. 259, 230-232. [26] Hohnadel, D., Haas, D. and Meyer, J.M. (1986) FEMS Microbiol. Lett. 36, 195-199.
[27] Ankenbauer, R., Hanne, L.F. and Cox, C.D. (1986) J. Bacteriol. 167, 7-11. [28] Cornelis, P., Boma, A., Belarbi, A., Guyonvarch, A., Kammerer, B., Hannaert, V. and Hubert, J.C. (1989) Mol. Microbiol. 3, 421-428. [29] Hoshino, T. and Kose, K. (1991) J. Bacteriol. 172, 55315539. [30] Hoshino, T., Kose, K. and Uratam, Y. (1990) Mol. Gen. Genet. 220, 461-467. [31] Hoshino, T. and Kose, K. (1989) J. Bacteriol. 171, 63006306. [32] Hoshino, T., Kose-Terai, K. and Urantani, Y. (1991) J. Bacteriol. 173, 1855-1861. [33] Guzzo, J., Murgier, M., Filloux, A. and Lazdunski, A. (1990) J. Bacteriol. 172, 942-948. [34] Hobbs, M., Dalrymple, B., Delaney, S. and Mattick, J.S. (1988) Gene 62, 219-227.
101
FEMSLE 80051
The new approaches to whole genome analysis of bacteria B.W. Holloway, M.D. E s c u a d r a , A.F. M o r g a n , R. Saffery and V. Krishnapillai Department of Genetics and Developmental Biology, Monash University, Clayton, Victoria, Australia Received 2 June 1992 Accepted 21 July 1992
Key words: Bacterial genome; Pulsed field gel electrophoresis; Combined physical/genetic maps; Pseudomonas aeruginosa ; Pseudomonas putida ; Pseudomonas solanacearum ; Ribosomal genes 1. S U M M A R Y A range of recombinant DNA techniques now enables whole genome analysis of any bacterium to be carried out without recourse to the classical means of bacterial genetic exchange. Using enzymes which cut infrequently, such as SpeI, combined with pulsed field gel electrophoresis, a physical map of ordered fragments can be constructed. By means of cloned fragments of known genes or oligonucleotides synthesized using data from DNA or protein sequence banks, the location of individual genes on this map can be determined. We have used these techniques to study whole genome structure in three species of Pseudomonas: P. aeruginosa, P. putida and P. solanacearum.
2. I N T R O D U C T I O N A comprehensive review of chromosome organization in bacteria has recently been published
Correspondence to: B.W. Holloway, Department of Genetics and Developmental Biology, Monash University, Clayton, Victoria 3168, Australia.
[1]. The best known bacterium from a genetic viewpoint is undoubtedly Escherichia coli K12 with the positions of some 1403 loci known [2], almost all mapped using conjugation and transduction. The introduction of pulsed field gel electrophoresis (PFGE) in its various forms has enabled the separation of linear fragments of double stranded DNA up to several megabases. The identification of enzymes which cut bacterial chromosomes infrequently has permitted the construction of physical maps in which the location of restriction enzyme sites can be mapped relative to each other. For E. coli, there are 27 NotI and 22 Sill sites giving a genome size of 4700 kb [31. Other techniques can be combined to extend the sophistication of any combined genetic and physical map. These include DNA probing, which could involve use of either cloned material, synthetic oligonucleotides or polymerase chain reaction (PCR) generated sequences combined with the use of data retained in the increasingly comprehensive DNA and protein data banks. One of the most attractive features of a combined approach with this range of techniques is that it is equally applicable to all bacteria and independent of the existence of conjugation, transduction or transformation systems. To date,
102 combined genetic and physical maps have been published for a range of bacteria [1] including E. coli [3], Pseudomonas aeruginosa [4], Rhodobacter sphaeroides [5] and Caulobacter crescentus [6]. T h e p s e u d o m o n a d s are a h e t e r o g e n e o u s group of bacteria, with important genetic differences having b e e n d e m o n s t r a t e d between them and those other bacteria which have b e e n genetically analysed. These differences include c h r o m o s o m a l gene arrangement, and the role of plasmids in phenotypic characterization [7]. The need for genetic systems of analysis in a wide variety of species and strains of Pseudomonas, in contrast to the experimental concentration on one strain of one species in Escherichia, created a need for systems promoting genetic exchange which were effective over a range of isolates. Plasmids such as R68.45 and R91-5 [7] solved this problem to some extent, but extensive work with different species and strains d e m o n s t r a t e d that these plasmids diffe'red in their effectiveness with different strains. For example, R91-5 was excellent for c h r o m o s o m e m a p p i n g in P. putida P P N [8] but less effective for P. syringae [9] and in other strains of P. putida (M.I. Sinclair, personal c o m m u n i c a t i o n ) and neither R91-5 nor R68.45 have been of any use with P. solanacearum (M.D. Escuadra and B.W. Holioway, unpublished data). T h e use of P F G E combined with probing and interspecific c o m p l e m e n t a t i o n using clones from a genomic library has resulted in the genetic analysis of a range of p s e u d o m o n a d s which has enabled a m o r e precise comparison of their g e n o m e structure. Even so, the strategy a d o p t e d for the physical analysis of the P. aeruginosa genome, the species first selected, has not been always readily adaptable to other species of this genus. Ratnaningsih et al. [4] have constructed a combined genetic and physical map of P. aeruginosa PAO. In an i n d e p e n d e n t analysis, R6mling and Tiimmler [10] have also constructed a physical map of P. aeruginosa P A O using two-dimensional P F G E , and, while the physical map so constructed is largely similar to that of Ratnaningsih et al. [4], differences do exist. These have b e e n largely identified and a c o m m o n system of nomenclature a d o p t e d [11].
3. R E S U L T S The range of techniques which have proved to be successful for P. aeruginosa g e n o m e analysis is now being extended in P. aeruginosa and applied to two other species of Pseudomonas: P. putida, an important organism in biotechnology; and P. solanacearum, a plant pathogen.
3.1. P. aeruginosa By a variety of techniques, a n u m b e r of genes have now been located on SpeI fragments, the combined physical and genetic map being ext e n d e d and refined. In addition to these reported in Holloway et al. [12], other genes have been
Table 1 Location of genes on the physical map of P. aeruginosa PAO by probing with cloned DNA. The nomenclature of Tiimmler et al. [13] is used for Spel fragment identification Gene Character Symbol
Spel
oriC3m origin of chromosome K replication oriC3o I origin of chromosome replication pilT twitchingmotility (fimbriae) nid/ccf nitrate reductase/ cytochrome C551 pt,d pyoverdinsynthesis lpp membrane lipoprotein braB branched amino acid transport braC branched amino acid transport braZ branched amino acid transport apr alkalineprotease fim A fimbrialsubunit
Reference
fragmentation location 10, 22
K
22
H
23
H
24, 25 ~'
H, F V Q or R
26, 27 b.c,~ 28 c 29, 30 f
U
29, 31 f
D
32 1
M E
33 ~ 23, 34
a Brunori, M. and Silvestrini, M.C. (personal communication). b Meyer, J.M. (personal communication). c H6fte, M. (personal communication). J Visca, P. (personal communication). c van de Lelie, D. (personal communication). f Hoshino, T. (personal communication). g Lazdunski, A. (personal communication).
103
located on SpeI fragments by probing with cloned material as shown in Table 1.
3.2. P. putida PPN A circular chromosomal map has been constructed using Hfr donor strains generated by integration of Tn501 and T n 5 derivatives of the narrow host range P. aeruginosa plasmid R91-5, which acts as a natural suicide vector in P. putida [8,13,14]. The genome size has been shown to be 5960 kb, some 400 kb larger than P. aeruginosa PAO and considerably larger than the 4400 kb reported by Grothues and Tfimmler [15] for P. putida DSM50291. While circularity of the physical map has not yet been established, it is clear that most if not all of the genomic DNA is chromosomal. Interestingly, consideration of genetic (88 min) and physical size (5960 kb) of the P. putida PPN chromosome yields a conjugal transfer rate of 68 kb/min, compared to that for E. coli of 47 kb/min, even though the optimum growth temperature for this strain of P. putida is 28°C. It has proved to be much more difficult to determine the linkage relationships of SpeI fragments in P. putida PPN than was the case in P. aeruginosa. In the first place there are rather more fragments compared to P. aeruginosa (44 compared tO 37) and the complexities of the system increase geometrically with increasing fragment number. Secondly, attempts to construct a Spel junction fragment library using methods developed by Ratnaningsih et al. [4] for P. aeruginosa PAO have led to an extremly biased library. More than 90% of such clones undergo considerable rearrangement immediately following construction. This is possibly due to the different methylation patterns of the two species, revealed by the observation that while DpnI (which recognizes only methylated DNA) cuts P. aeruginosa PAO DNA, it does not cut P. putida PPN genomic DNA. To overcome this a K m r / SpeI cassette with SpeI cohesive ends has been constructed which allows any plasmid vector to be used for positive selection of SpeI junction fragments. The physical data map available has been sufficient for an analysis of ribosomal genes in P.
putida. By probing fragments of the P. putida genome separated by SpeI digestion and P F G E with cloned 5S, 16S or 23S rDNA genes of P. aeruginosa, seven distinct operon-like clusters of ribosomal genes have been identified on the P. putida chromosome. It has not yet been shown that all such clusters are functional. Mapping of these genes has been facilitated by the fact that all seven P. putida PPN 23S genes contain a conserved SpeI site.
3.3. P. solanacearum Little is known about the genome structure of
P. solanacearum, the cause of bacterial wilt in a wide range of plant genera. Boucher et al. [16] have constructed a circular chromosome map of strain GMI1000 containing 19 markers using conjugation promoted by RP4-prime plasmids. Hence this species is well suited to a combined physical approach to genome mapping. P F G E of P. solanacearum has so far identified 30 SpeI fragments, giving a genome size of 5548 kb. It is known that the strain studied contains large plasmid(s) and as their actual size has not been determined, the bacterial chromosome size will be less than this. As with P. putida, the construction of a SpeI linking clone library has not proved as straightforward as was the case with P. aeruginosa, and only 15 of the potential 30 fragment linkages have been identified. To construct a genetic map in P. solanacearum, a cosmid library of this bacterium was made using the wide host range cosmid vector pLA2917 [17]. Auxotrophic markers located on individual cosmids were identified in two ways. Firstly, by complementation of a range of auxotrophic mutants isolated in P. solanacearum, although in most cases these mutants have not been characterized any further than identifying the amino acid requirement. Secondly, this cosmid library was used to complement a range of P. aeruginosa auxotrophs many of which have been characterized with respect to the enzyme deficiency [18]. The interspecific and intergeneric identification of markers by this approach has been used extensively for P. putida [13,19] and has been readily adaptable to P. solanacearum. Once markers have been identified on individ-
104
ual cosmids, then such cosmids can be used as probes to relate the location of markers to individual SpeI fragments and to date, 38 markers have been mapped in this way. Combined PFGE and DNA probing techniques have been used to locate genes affecting extracellular polysaccharide and polygalacturonate synthesis.
4. DISCUSSION These are early days in the whole genomic analysis of bacteria, particularly in those organisms for which there are no characterized systems of genetic exchange or even collections of mutants [20]. Over 100 different bacterial isolates representing different genera, species and strains have been analyzed by PFGE indicating the broad applicability of this technology. The ever increasing size and availability of DNA and protein data bases combined with PCR to generate probe DNA, will permit the physical mapping of any DNA sequence in the species of origin. Depending upon the degree of sequence conservation between related and unrelated bacteria, this will allow gene mapping in a wider range of organisms of interest. While construction of a physical and genetic map is an achievement in itself, the importance of that achievement will be enhanced by the use of the data made available. An example of such use was provided by Sinclair and Holloway [21] who demonstrated highly specific integration sites of the TOL transposons Tn4651 and Tn4653 into the chromosome of P. aeruginosa PAO. As an example of the use of these techniques for mapping genes which could not be mapped by conventional means, Smith et al. [22] have described the precise mapping of replication origin sites in P. aeruginosa. Given the labour-intensive nature of finding, characterizing and using classical systems of genetic exchange in bacteria, these combined techniques should result in an explosion of genome data which will be of general application to microbial genetics, evolutionary studies, applied microbiology and biotechnology.
ACKNOWLEDGEMENTS Work in the authors' laboratory is supported by the Australian Research Council, the Australian Centre for International Agricultural Research and the Celgene Corporation.
REFERENCES [1] Krawiec, S. and Riley, M. (1990) Microbiol. Rev. 54, 502-539. [2] Bachman, B.J. (1990) Microbiol. Rev. 54, 130-197. [3] Smith, C.L., Econome, J.G., Schutt, A., Klco, S. and Cantor, C.R. (1987) Science 236, 1448-1453. [4] Ratnaningsih, E., Dharmsthiti, S., Krishnapillai, V., Morgan, A., Sinclair, M. and Holloway, B.W. (1990) J. Gen. Microbiol. 136, 2351-2357. [5] Suwanto, A. and Kaplan, S. (1989) J. Bacteriol. 171, 5850-5859. [6l Ely, B., Ely, T.W., Gerardot, C.J. and Dingwall, A. (1990) J. Bacteriol. 172, 1262-1266. [7] Holloway, B.W. and Morgan, A.F. (1986) Microbiol. Rev. 40, 79-105. [8] Dean, H.F. and Morgan, A.F. (1983) J. Bacteriol. 153, 485 -497. [9] Nordeen, R.O. and Holloway, B.W. (1990) J. Gen. Microbiol. 136, 1231-1239. [10] R6mling, U. and Tiimmler, B. (1991) Nucleic Acids Res. 19, 3199-3206. [11] Tfimmler, B., R6mling, U., Ratnaningsih, E., Morgan, A.F., Krishnapillai, V. and Holloway, B.W. In: Pseudomonas: Molecular Biology and Biotechnology. (Galli, E., Silver, S. and Witholt, B., Eds.), Am. Soc. Microbiol. Washington, DC. in press. [12] Holloway, B.W., Bowen, A., Dharmsthiti, S., Krishnapillai, V., Morgan, A., Ratnaningsih, E. and Sinclair, M. (1991) In: Proc. 6th Int. Symp. Genetics, Industr. Microorg. (Heslot, H., Davies, J., Florent, J., Bobichon, L., Durand, G. and Penasse, L., Eds.), 1, 227-238. Soc. Francaise de Microbiol., Paris. [13] Morgan, A.F. and Dean, H.F. (1985) J. Gen. Microbiol. 131,885-896. [14] Strom, D., Hirst, R., Petering, J. and Morgan, A. (1990) Genetics 126, 497-503. [15] Grothues, D. and Tfimmler, B. (1991) Mol. Microbiol. 5, 2763-2776. [16] Boucher, C., Arlat, M., Zischek, C. and Boistard, P. (1988) In: Physiology and Biochemistry of Plant-Microbial Interactions. (Keen, N.T., Kosuge, T. and Walling, L.L., Eds.), pp. 83-95. The American Society of Plant Pathologists, Rockville. [17] Allen, L.N. and Hanson, R.S. (1985) J. Bacteriol. 161, 955-963.
105 [18] Holloway, B.W. and Zhang, C. (1990) Genetic Maps 5, 2.71-2.78. [19] Morgan, A.F. (1986) J. Bacteriol. 149, 654-661. [20] Smith, C.L. and Condemine, G. (1990) J. Bacteriol. 172, 1167-1172. [21] Sinclair, M.I. and Holloway, B.W. (1991) J. Gen. Microbiol. 137, 1111-1120. [22] Smith, D.W., Yee, T.W., Baird, C. and Krishnapillai, V. (1991) Mol. Microbiol. 5, 2581-2587. [23] Whitchurch, C.B., Hobbs, M., Livingston, S.P., Krishnapillai, V. and Mattick, J.S. (1990) Gene 101, 33-44. [24] Silvestrini, M.S., Galeotti, C.L., Gervais, M., Schininia, E., Barra, D., Bossa, F. and Brunon, M. (1989) FEMS Microbiol. Lett. 254, 33-38. [25] Nordling, M., Young, S., Karlsson, B.G. and Lundberg, L.G. (1990) FEMS Microbiol. Lett. 259, 230-232. [26] Hohnadel, D., Haas, D. and Meyer, J.M. (1986) FEMS Microbiol. Lett. 36, 195-199.
[27] Ankenbauer, R., Hanne, L.F. and Cox, C.D. (1986) J. Bacteriol. 167, 7-11. [28] Cornelis, P., Boma, A., Belarbi, A., Guyonvarch, A., Kammerer, B., Hannaert, V. and Hubert, J.C. (1989) Mol. Microbiol. 3, 421-428. [29] Hoshino, T. and Kose, K. (1991) J. Bacteriol. 172, 55315539. [30] Hoshino, T., Kose, K. and Uratam, Y. (1990) Mol. Gen. Genet. 220, 461-467. [31] Hoshino, T. and Kose, K. (1989) J. Bacteriol. 171, 63006306. [32] Hoshino, T., Kose-Terai, K. and Urantani, Y. (1991) J. Bacteriol. 173, 1855-1861. [33] Guzzo, J., Murgier, M., Filloux, A. and Lazdunski, A. (1990) J. Bacteriol. 172, 942-948. [34] Hobbs, M., Dalrymple, B., Delaney, S. and Mattick, J.S. (1988) Gene 62, 219-227.
