Download Nitrogen fixation:

ECOLOGICAL, PHYLOGENETIC AND TAXONOMIC REMARKS ON DIAZOTROPHS AND
RELATED GENERA
E. Martínez-Romero, J. Caballero-Mellado, B. Gándara1, M. A. Rogel, A. Lopez Merino1, E.
T. Wang, L. E. Fuentes-Ramirez, I. Toledo, L. Martinez, I. Hernandez-Lucas, J. MartinezRomero.
Centre de Investigatión sobre Fijación de Nitrógeno, UNAM, Cuernavaca, Morelos, México,
1
Escuela Nacional de Ciencias Biológicas, IPN. Mexico
1.
Introduction
Nitrogen-fixing organisms are very diverse prokaryotes that have been identified scattered in the
phylogenetic trees reconstructed from the comparative analysis of ribosomal RNA gene sequences
(Martinez-Romero, 1985, Young, 1992). Although 16S rRNA gene-based phytogenies have been
criticized and the Universal Tree of Life questioned (Pennisi, 1998), novel approaches from
proteome analysis derived from genomes of 20 organisms provide support to the unrooted Universal
Tree of Life based on 16S rRNA genes (Tekaia et al., 1999). However, the final answer has not yet
been given in this highly controversial field. A critical overview of 16S rRNA gene based
phylogenies compared to other markers has been presented (Martinez-Romero et al., 1999).
Anyhow, it is clear that nitrogen fixers are found in close relationship to non nitrogen fixers.
Nitrogen fixers may constitute clusters restricted in distribution inside groups of non-fixing bacteria
as is the case of Acetobacter diazotrophicus and novel Acetobacter N-fixing species inside
Acetobacter species (Caballero-Mellado et al., 1999; Fuentes-Ramírez et al, this volume) or
Paenibacillus among aerobic endospore-forming Firmicutes (Achouak et al., 1999). In Archea, Nfixation is a general property of the methanogen group and nif genes are found in halophiles, but no
fixers have been found in the sulfur dependent archaeobacteria. The Frankia group in the
actynomycetes includes atypical strains as well as an infective but no nitrogen fixer isolate
(Normand et al., 1996); in Rhizobium there are no nitrogen fixers, nonsymbiotic strains (Segovia et
al., 1991), meaning that nitrogen fixation may be easily lost in some cases. The loss of nitrogen
fixing ability has been proposed to explain non fixers occurring closely related to N-fixing bacteria.
Possessing nitrogen fixing genes would not be advantageous for bacteria if they inhabit N-rich
environments. N-fixers are beneficial in plant-bacteria associations, unfortunately, farmers provide
N-rich media when heavily fertilizing agricultural fields and we supposed that these practices may
negatively affect populations of nitrogen fixers (Martinez-Romero, Caballero-Mellado, 1996). We
have documented evidence of different negative effects of fertilization: lack of recovery (FuentesRamírez et al., 1993) and less genetic diversity of A. diazotrophicus in fertilized sugar cane fields
(Caballero-Mellado et al., 1995) with diminished colonization by A. diazotrophicus of N-fertilized
sugar-cane plants (Fuentes-Ramírez et al., 1999), Fig. 1. We also reported a decrease in the genetic
diversity of Rhizobium recovered from P. vulgaris (bean) nodules in fertilized fields using the
commonly recommended N dose for bean fertilization in Mexico (Caballero-Mellado, MartinezRomero, 1999), Fig. 2. It seems highly probable that not all nitrogen fixers are known at present.
Some may be unculturable. "New findings may fill in some of the blanks on the bacterial map, and
the distribution on N2 fixation may prove to be less patchy than it appears at present" (Young,
1992). However, very few real new N-fixers have been reported in these last years, this may be in
relation to the fact that the ability to fix nitrogen is not a routine test for new isolates and is
performed only in labs specializing in the subject. In addition, pathogenic bacteria are usually
excluded from N-fixing analyses as they are supposedly considered non-fixers. Nevertheless,
155
F.O. Pedrosa et al. (eds.), Nitrogen Fixation: From Molecules to Crop Productivity, 155–160.
© 2000 Kluwer Academic Publishers. Printed in the Netherlands.
156
Agrobacterium tumefaciens strains, a plant tumor inducer, has nif genes and fixes nitrogen
(Kanvinde, Sastry, 1990). Around 30% of klebsiellae, either soil or hospital isolates, are
diazotrophic (Postgate, Eady, 1988) or even 100% of them may fix depending on assayed conditions
(unpublished results). Furthermore, nitrogen fixation may be expressed only under special bacterial
growth conditions. Many free-living organisms require anaerobic conditions for fixation.
Bradyrhizobium requires a burst of glutamate; Rhizobium a differentiation process; Azoarcus,
diazosome formation (Karg, Reinhold-Hurek, 1996) and in Azospirillum nitrogen fixation is
stimulated by plant agglutinins (Karpati et al., 1999).
157
In any case, bioprospection for new fixers should be encouraged and this effort seems highly
justified. We are searching for diazotrophs associated to banana plants.
The identification of new isolates and the phylogenetical classification that looks toward clustering
natural groups are the basis for good science and are clearly related to important changes in family
names of N fixers that would be adopted in the new Bergey's Manual. Rhizobiaceae will no longer
encompass Rhizobium, Bradyrhizobium and Mesorhizobium species. Mesorhizobium will be
included in a new family together with Phyllobacterium and Bradyrhizobium grouped with
Rhodopseudomonas in another new family. Azoarcus will be distributed in several families.
2.
Rhizobia and related genera
From the comparative analysis of 16S rRNA genes, relationships of bacteria to known fixers have
been revealed. For example, strains of sewage water Zooglea ramigera were found to be similar to
the rhizobia group (Rosselló-Mora et al., 1993); Pseudoaminobacter, Chelatobacter, and
Aminobacter with important uses in biotechnology for bioremediation are groups in close
relationship to Mesorhizobium spp (Auling et al., 1993; Kampfer et al., 1999); and the bacterialysing Ensifer isolates are similar to (Sind)rhizobium meliloti (D. Balkwill, personal
communication). We have found that a new cluster of Mesorhizobium represented by SH0172 (Tan
et al., 1999) share with methylotrophic bacteria the capacity to use methanol as sole carbon source.
Brucella constitute a cluster of bacteria in relationship to Rhizobium spp (de Ley, 1987; MartinezRomero et al., 1991; Young et al., 1991). From 16S rRNA gene sequences the position of the nodes
relating this genus to other rhizobia may not be confidently resolved (Ludwig et al., 1998). By
multilocus enzyme electrophoresis we confirmed the close relationship of Brucella and Rhizobium
using more than 100 Brucella isolates and Rhizobium reference strains (Fig. 3) and found that
Brucella was closely related to R. tropici.
All the B. melitensis and B. abortus strains studied have two chromosomes of sizes 2.1 and 1.15 Mb
(Michaux et al., 1993). B. suis biovars 2 and 4 have two chromosomes with different sizes and a
unique chromosome was found in B. suis biovar 3. An ancestor with a single chromosome has been
proposed for all Brucella strains. Rearrangement at the rrn loci led to the existence of two
minichromosomes (Jumas-Bilak et al., 1998). Our data clearly support a monochromosomic
ancestor of Brucella that must have been related to R. tropici as well. It is interesting to remark that
R. tropici is also closely related to Agrobacterium biovar 2 strains and Agrobacterium species have
158
been recovered as clinical isolates in humans (Alnor et al., 1994; Riley, Weaver, 1977). Brucella
chromosomes are in the range of size of the megaplasmids of Rhizobium and megaplasmids have
been observed in R. tropici but these megaplasmids do not harbor rRNA genes (Geniaux et al.,
1995). In addition, R. tropici has smaller plasmids one of them
harboring nod and nif
genes (Martinez et al., 1987). We are pursuing the analysis of the genetic differences of R. tropici
and Brucella towards the elucidation of mechanisms involved in the evolution of a pathogen, bac
genes involved in bacteroid differentiation in R. meliloti (Glazebrook et al., 1996) have homologs in
Brucella required for macrophage invasion thus providing further support to the close relatedness of
both genera.
159
Bioprospection in Rhizobium has yielded fruit. The excellent work in NGR234 led by Broughton
(Freiberg et al., 1997; Perret et al., 1999; Pueppke, Broughton, 1999; Viprey et al., 1998) has been
possible after the pioneer work of bioprospection of legume symbionts by Trinick in New Guinea
in 1965 (Trinick et al., 1980). Similarly, the Rhizobium species described in recent years, such as
Azorhizobium caulinodans, Rhizobium galegae, R. fredii, R. etli, and R. tropici, have enriched our
knowledge in this field. As an example, novel modifications of Nod factors have been recognized in
new species (Bec-Ferté et al., 1994; Poupot et al., 1995; Mergaert et al., 1993) and in isolates as
BR816 (Snoeck et al., this volume). BR816 is a Sinorhizobium isolate from the analysis of 16S
rRNA sequences (Hernández-Lucas et al., 1995). Naming R. tropici as a new species (MartinezRomero et al., 1991), encouraged research in different aspects including studies that led to the
identification of sulfate on Nod factors (Poupot et al., 1993) that was otherwise exclusively
encountered in R. meliloti (Roche et al., 1991). R. tropici strains have been extensively used as bean
(Phaseolus vulgaris) inoculants in South America (Vlassak et al., 1997).
The identification of M. amorpha (Wang et al., 1999a) showed the clear existence of symbiotic
plasmids containing nif and nod genes in Mesorhizobium. M. loti and M. plurifarium have these
genes in the chromosome and in M. loti, symbiotic islands have been described (Sullivan, Ronson,
1998). This finding adds support to Mesorhizobium being intermediate between Rhizobium (having
mainly symbiotic plasmids) and Bradyrhizobium (with symbiotic determinants in chromosome). R.
huautleme (Wang et al., 1998) is the sole Rhizobium species related to R. galegae, that was
otherwise mainly immersed in an Agrobacterium cluster (Young, Haukka, 1996). The R.
huautlense-Sesbania herbacea symbiosis is highly adapted to flooded conditions (Wang, MartinezRomero, submitted) without stem nodules. R. etli represents the main symbiont of Phaseolus in its
site of origin (Segovia et al., 1993) . The proposal of a new biovar (mimosae) in R. etli isolates from
Mimosa affinis (Wang et al., 1999b) allows the identification of a possible ancestral plasmid of
phaseoli and presents clear data on fast plasmidic evolution. It also shows that biovars are a
common feature in R. etli, R. leguminosarum, and R. gallicum.
3.
References
Achouak W et al (1999) Int J Syst Bacteriol 49,961-967
Alnor D et al (1994) Clinical Infectious Diseases 18, 914-920
Auling G et al (1993) Syst Appl Microbiol 16, 104-112
Bec-Ferté M-P (1994) Biochem 33, 11782-11788
Caballero-Mellado J et al (1995) Appl Environ Microbiol 61, 3008-3013
Caballero-Mellado J et al (1999) In Martínez E and Hernandez G, eds. Highlights of Nitrogen
Fixation Research pp 271-274, Kluwer Academic/Plenum Publishers, New York, USA
Caballero-Mellado J, Martinez Romero E (1999) Symbiosis 26, 111-121
De Ley et al (1987) Int J Syst Bacteriol 37, 35-42
Freiberg C et al (1997) Nature 387, 394-401
Fuentes-Ramírez LE et al (1993) Plant Soil 154, 145-150
Fuentes-Ramírez E et al (1999) FEMS Microbiol Ecol 29, 117-128
Geniaux E et al (1995) Int J Syst Bacteriol 45, 392-394
Glazebrook J et al (1996) J Bacteriol 178, 745-752
Hernández-Lucas I et al (1995) Appl Environ Microbiol 61, 2775-2779
Jumas-Bilak E et al (1998) Mol Microbiol 27, 99-106
Kampfer P et al (1999) Int J Syst Bacteriol 49, 887-897
Karg T, Reinhold-Hurek B (1996) J Bacteriol 178, 5748-5754
Kanvinde L, Sastry GRK (1990) Appl Environ Microbiol 56, 2087-2092
160
Karpati E et al (1999) J Bacteriol 181, 3949-3955
Ludwig W et al (1998) Plant Soil 204, 1-19
Martinez E et al., (1987) J Bacteriol 169, 2828-2834
Martinez-Romero E (1985) Ph. D. Thesis, UNAM, Mexico
Martinez-Romero E et al (1991) Int J Syst Bacteriol 41, 417-426
Martinez-Romero E, Caballero-Mellado J (1996) Crit Rev Plant Sci 15, 113-140
Martinez-Romero et al (1999) in de Wit, Bisseling and Stiekema eds. 1999 IC-MPMI Congress
Proceedings "Biology of Plant-Microbe Interactions, vol 2."
Mergaert P et al (1993) Proc Natl Acad Sci USA 90, 1551-1555
Michaux S et al (1993) J Bacteriol 175, 701-705
Normand P et al (1996) Int J Syst Bacteriol 46, 1-9
Pennisi, E (1998) Science 280, 672-674
Perret X et al (1999) Mol Microbiol 32, 415-425
Poupot R et al (1993) Biochem 32, 10430-10435
Poupot R et al (1995) J Biol Chem 270, 6050-6055
Postgate JR, Eady RR (1988) in Bothe H, de Bruijn FJ, Newton WE eds. Nitrogen Fixation:
Hundred Years After. pp 31-40. Gustav Fischer, Stuttgart, Germany
Pueppke SG, Broughton WJ (1999) Mol Plant-Microbe Interact 12, 293-318
Riley PS, Weaver RE (1977) J Clinical Microbiol 5, 172-177
Roche P et al (1991) Cell 67, 1131-1143
Rosselló-Mora R et al (1993) FEMS Microbiol Lett 114, 129-134
Segovia L et al (1991) Appl Environ Microbiol 57, 426-433
Segovia L et al (1993) Int J Syst Bacteriol 43, 374-377
Sullivan JT, Ronson CW (1998) Proc Natl Acad Sci USA. 95, 5145-5149
Tan ZY et al (1999) Int J Syst Bacteriol 49, 1457-1469
Tekaia F et al (1999) Genome Res. 9, 550-557
Trinick MJ (1980) J Appl Bacteriol 49, 39-53
Viprey V et al (1998) Mol Microbiol 28, 1381-1389
Vlassak KM et al (1997) Biol Fertil Soils 24:274-282
Walker G (1999) In Molecular Plant-Microbe Interactions
International Congress, Book of
Abstracts
Wang ET et al (1998) Int J Syst Bacteriol 48, 687-699
Wang ET et al (1999a) Int J Syst Bacteriol 49, 51-65
Wang ET et al (1999b) Int J Syst Bacteriol 49, 1479-1491
Wang ET, Martinez-Romero E (1999) FEMS Microbiol Ecol submitted
Young JPW et al (1991) J Bacteriol 173, 2271-2277
Young JPW (1992) In Stacey G, Burris RH and Evans HJ, eds. Biological Nitrogen Fixation pp 4386, Chapman and Hall, New York, USA
Young JPW, Haukka KE (1996) New Phytol 133, 87-94
4.
Acknowledgments
To DGAPA grant IN202097 and to CONACyT 25075-B. To M. Dunn for reading the manuscript.