The Evolutionary History of Nitrogen Fixation, as Assessed by NifD

The evolutionary history of nitrogen fixation has been vigorously debated for almost two decades. Previous phylogenetic analyses of nitrogen fixation genes (nif) have shown support for either evolution by vertical descent or lateral transfer, depending on the specific nif gene examined and the method of analyses used. The debate centers on the placement and monophyly of the cyanobacteria, proteobacteria, and Gram-positive bacteria (actinobacteria and firmicutes). Some analyses place the cyanobacteria and actinobacteria within the proteobacteria, which suggests that the nif genes have been laterally transferred since this topology is incongruent with ribosomal phylogenies, the standard marker for comparison. Other nif analyses resolve and support the monophyly of the cyanobacteria, proteobacteria, and actinobacteria, supporting vertical descent. We have revisited these conflicting scenarios by analyzing nifD from an increased number of cyanobacteria, proteobacteria, and Gram-positive bacteria. Parsimony analyses of amino acid sequences and maximum likelihood analysis of nucleic acid sequences support the monophyly of the cyanobacteria and actinobacteria but not the proteobacteria, lending support for vertical descent. However, distance analysis of nucleic acid sequences placed the actinobacteria within the proteobacteria, supporting lateral transfer. We discuss evidence for both vertical descent and lateral transfer of nitrogen fixation.     

Progress on the genetics of the N2-fixing actinorhizal symbiont Frankia

The actinomycete Frankia is of fundamental and ecological interests for several reasons including its wide distribution, its ability to fix nitrogen, differentiate into sporangium and vesicle (specialized cell for nitrogen-fixation), and to nodulate plants from about 24 genera. Here, we present a review on the genetics performed so far on Frankia. At the end of July 2001, 293 kbp of Frankia DNA sequences were found in the databases. Thirty five percent of these sequences corresponded to full gene or gene cluster sequences. These genes could be divided according to their role into 6 key activities: gene translation (rrnA and tRNA ^pro gene), proteolysis (pcr genes), assimilation of ammonium (glnA and glnII), protection against superoxide ions (sodF), nitrogen fixation (nif cluster), and plasmid replication. We present a review of these genetic islands; their function, expression, localization and particular properties are discussed. A comparative analysis of Frankia nif genes from various strains and species is presented. An improved nomenclature for some of these genes is suggested to avoid conflicts. Frankia plasmids DNA sequences are also presented. The novel trends in Frankia genetics are described.     

Expression of Frankia nif genes in nodules of Alnus glutinosa

Expression of Frankia genes involved in nitrogen fixation was studied in Alnus glutinosa nodules using the in situ hybridization technique. The results show that high level expression of nif genes does not occur immediately upon infection of cortical cells by Frankia. Also, only in the infected cells near the tips of the nodule lobes, nif genes are expressed at high levels. In the majority of infected cells, nif gene expression is rather low.     

Phylogeny of members of the <Emphasis Type="Italic">Frankia</Emphasis> genus based on <Emphasis Type="Italic">gyr</Emphasis>B, <Emphasis Type="Italic">nif</Emphasis>H and <Emphasis Type="Italic">gln</Emphasis>II sequences

To construct an evolutionary hypothesis for the genus Frankia, gyrB (encoding gyrase B), nifH (encoding nitrogenase reductase) and glnII (encoding glutamine synthetase II) gene sequences were considered for 38 strains. The overall clustering pattern among Frankia strains based on the three analyzed sequences varied among themselves and with the previously established 16S rRNA gene phylogeny and they did not reliably reflect clear evolution of the four discerned Frankia clusters (1, 2, 3 and 4). Based on concatenated gyrB, nifH and glnII, robust phylogenetic trees were observed with the three treeing methods (Maximum Likelihood, Parsimony and Neighbor-Joining) and supported by strong bootstrap and posterior probability values (>75%) for overall branching. Cluster 4 (non-infective and/or non-nitrogen-fixing Frankia) was positioned at a deeper branch followed by cluster 3 (Rhamnaceae and Elaeagnaceae infective Frankia), while cluster 2 represents uncultured Frankia microsymbionts of the Coriariaceae, Datiscaceae, Rosaceae and of Ceanothus sp. (Rhamnaceae); Cluster 1 (Betulaceae, Myricaceae and Casuarinaceae infective Frankia) appears to have diverged more recently. The present study demonstrates the utility of phylogenetic analyses based upon concatenated gyrB, nifH and glnII sequences to help resolve previously unresolved or poorly resolved nodes and will aid in describing species among the genus Frankia.     

Creation of novel nitrogen-fixing actinomycetes by protoplast fusion of Frankia with streptomyces

Protoplast fusion was used for the creation of a novel actinomycete capable of fixing atmospheric nitrogen. Protoplasts of Streptomyces griseofuscus, a fast-growing actinomycete, and Frankia, a slow-growing actinomycete which fixes atmospheric nitrogen in culture and in symbiotic association with alders, were allowed to fuse and regenerate on media without supplied nitrogen. Colonies which regenerated acquired the fast-growing characteristic of Streptomyces and the ability to grow on nitrogen-deficient media from Frankia. These colonies resembled Streptomyces in their morphology and fixed atmospheric nitrogen in culture. They contained both the parent Streptomyces DNA sequences and the Frankia DNA sequences homologous to nif structural genes HDK of K. pneumoniae. In addition to in vitro nitrogen-fixing capacity, one out of 20 colonies also formed nitrogen-fixing root nodules on Alnus rubra, the host plant for the Frankia strain. Examination of the root nodules induced by the hybrids showed only the presence of hyphae-like structures. The typical vesicle-like structures present in Frankia were absent.     

Comparative Genomic Analysis of N2-Fixing and Non-N2-Fixing Paenibacillus spp.: Organization,Evolution and Expression of the Nitrogen Fixation Genes

We provide here a comparative genome analysis of 31 strains within the genus Paenibacillus including 11 new genomic sequences of N₂-fixing strains. The heterogeneity of the 31 genomes (15 N₂-fixing and 16 non-N₂-fixing Paenibacillus strains) was reflected in the large size of the shell genome, which makes up approximately 65.2% of the genes in pan genome. Large numbers of transposable elements might be related to the heterogeneity. We discovered that a minimal and compact nif cluster comprising nine genes nifB, nifH, nifD, nifK, nifE, nifN, nifX, hesA and nifV encoding Mo-nitrogenase is conserved in the 15 N₂-fixing strains. The nif cluster is under control of a σ⁷⁰-depedent promoter and possesses a GlnR/TnrA-binding site in the promoter. Suf system encoding [Fe–S] cluster is highly conserved in N₂-fixing and non-N₂-fixing strains. Furthermore, we demonstrate that the nif cluster enabled Escherichia coli JM109 to fix nitrogen. Phylogeny of the concatenated NifHDK sequences indicates that Paenibacillus and Frankia are sister groups. Phylogeny of the concatenated 275 single-copy core genes suggests that the ancestral Paenibacillus did not fix nitrogen. The N₂-fixing Paenibacillus strains were generated by acquiring the nif cluster via horizontal gene transfer (HGT) from a source related to Frankia. During the history of evolution, the nif cluster was lost, producing some non-N₂-fixing strains, and vnf encoding V-nitrogenase or anf encoding Fe-nitrogenase was acquired, causing further diversification of some strains. In addition, some N₂-fixing strains have additional nif and nif-like genes which may result from gene duplications. The evolution of nitrogen fixation in Paenibacillus involves a mix of gain, loss, HGT and duplication of nif/anf/vnf genes. This study not only reveals the organization and distribution of nitrogen fixation genes in Paenibacillus, but also provides insight into the complex evolutionary history of nitrogen fixation.     

Cluster analysis of genes for nitrogen fixation from several diazotrophs

Hierarchical clustering and similarity coefficients of pairwise alignments of the published nucleotide sequences of 27nifH genes suggest thatnif genes are as ancient as the archaebacteria and clostridia. The positions ofnifHl ofMethanococcus thermolithotrophicus, nifH3 ofClostridium pasteurianum, nifH3 ofAzotobacter vinelandii andnifH ofFrankia suggest that a variety of lateral transfers may have occurred during evolution ofnifH gene. The genes for type 3 nitrogenase ofA. vinelandii may have diverged early from methanogens and clostridia. A high similarity coefficient with the derived amino acid sequence of type 3 nitrogenase suggests the presence of a functionally similar enzyme inC. pasteurianum. The type 2 nitrogenase genenifH2 of azotobacters seems to have originated recently from the genenifHl for conventional type I nitrogenase. RhizobialnifH genes comprise two closely related but discrete clusters that are in consonance with the plasmid or chromosomal location ofnif genes. The chromosomal and plasmid locatednifH of rhizobia seem to have evolved independently but contemporaneously.     

Molecular analysis of actinorhizal symbiotic systems: Progress to date

The application of molecular tools to questions related to the genetics, ecology and evolution of actinorhizal symbiotic systems has been especially fruitful during the past two years. Host plant phylogenies based on molecular data have revealed markedly different relationships among host plants than have previously been suspected and have contributed to the development of new hypotheses on the origin and evolution of actinorhizal symbiotic systems. Molecular analyses of host plant gene expression in developing nodules have confirmed the occurrence of nodulin proteins and in situ hybridization techniques have been successfully adapted to permit the study of the spatial and temporal patterns of gene expression within actinorhizal nodules. The use of heterologous probes in combination with nucleotide sequence analysis have allowed a number of nif genes to be mapped on the Frankia chromosome which will ultimately contribute to the development of hypotheses related to nif gene regulation in Frankia. The use of both 16S and 23S rDNA nucleotide sequences has allowed the construction of phylogenetic trees that can be tested for congruence with symbiotic characters. In addition the development of Frankia-specific gene probes and amplification primers have contributed to studies on the genetic diversity and distribution of Frankia in the soil.     

Phylogeny of nitrogenase sequences inFrankia and other nitrogen-fixing microorganisms

Summary The complete nucleotide sequence of a nitrogenase (nifH) gene was determined from a second strain (HRN18a) ofFrankia, an aerobic soil bacterium. The open reading frame is 870 bp long and encodes a polypeptide of 290 amino acids. The amino acid and nucleotide sequences were compared with 21 other published sequences. The twoFrankia strains were 96% similar at the amino acid level and 93% similar at the nucleotide level. A number of methods were used to infer phylogenies of these nitrogen fixers, based onnifH amino acid and nucleotide sequences. The results obtained do not agree completely with other phylogenies for these bacteria and thus make probable occurrences of lateral transfer of thenif genes. The time of divergence of the twoFrankia strains could be estimated at about 100 million years. The vanadium-dependent (Type 2) nitrogenase present inAzotobacter spp. appears to be a recent derivation from the conventional molybdenum-dependent (Type 1) enzyme, whereas the iron-dependent (Type 3) alternative nitrogenase would have a much older origin.     

Host species and habitat affect nodulation by specific Frankia genotypes in two species of Alnus in interior Alaska

Alders (Alnus spp.) are important components of northern ecosystems due to their ability to fix nitrogen (N) in symbiosis with Frankia bacteria. Availability of optimal Frankia may be a contributing factor in limiting the performance and ecological effects of Alnus, but the factors underlying distribution of Alnus-infective Frankia are not well understood. This study examined the genetic structure (nifD–K spacer RFLP haplotypes) of Frankia assemblages symbiotic with two species of Alnus (A. tenuifolia and A. viridis) in four successional habitats in interior Alaska. We used one habitat in which both hosts occurred to observe differences between host species independent of habitat, and we used replicate sites for each habitat and host to assess the consistency of symbiont structure related to both factors. We also measured leaf N content and specific N-fixation rate (SNF) of nodules (¹⁵N uptake) to determine whether either covaried with Frankia structure, and whether Frankia genotypes differed in SNF in situ. Frankia structure differed between sympatric hosts and among habitats, particularly for A. tenuifolia, and was largely consistent among replicate sites representing both factors. Leaf N differed between host species and among habitats for both hosts. SNF did not differ among habitats or host species, and little evidence for differences in SNF among Frankia genotypes was found, due largely to high variation in SNF. Consistency of Frankia structure among replicate sites suggests a consistent relationship between both host species and habitat among these sites. Correlations with specific environmental variables and possible underlying mechanisms are discussed. Nomenclature: Flora of North America ().     

Frankia Populations in Soil and Root Nodules of Sympatrically Grown Alnus Taxa

The genetic diversity of Frankia populations in soil and in root nodules of sympatrically grown Alnus taxa was evaluated by rep-polymerase chain reaction (PCR) and nifH gene sequence analyses. Rep-PCR analyses of uncultured Frankia populations in root nodules of 12 Alnus taxa (n?=?10 nodules each) growing sympatrically in the Morton Arboretum near Chicago revealed identical patterns for nodules from each Alnus taxon, including replicate trees of the same host taxon, and low diversity overall with only three profiles retrieved. One profile was retrieved from all nodules of nine taxa (Alnus incana subsp. incana, Alnus japonica, Alnus glutinosa, Alnus incana subsp. tenuifolia, Alnus incana subsp. rugosa, Alnus rhombifolia, Alnus mandshurica, Alnus maritima, and Alnus serrulata), the second was found in all nodules of two plant taxa (A. incana subsp. hirsuta and A. glutinosa var. pyramidalis), and the third was unique for all Frankia populations in nodules of A. incana subsp. rugosa var. americana. Comparative sequence analyses of nifH gene fragments in nodules representing these three profiles assigned these frankiae to different subgroups within the Alnus host infection group. None of these sequences, however, represented frankiae detectable in soil as determined by sequence analysis of 73 clones from a Frankia-specific nifH gene clone library. Additional analyses of nodule populations from selected alders growing on different soils demonstrated the presence of different Frankia populations in nodules for each soil, with populations showing identical sequences in nodules from the same soil, but differences between plant taxa. These results suggest that soil environmental conditions and host plant genotype both have a role in the selection of Frankia strains by a host plant for root nodule formation, and that this selection is not merely a function of the abundance of a Frankia strain in soil.     

Evaluation of the 23S rRNA gene as target for qPCR based quantification of Frankia in soils

The 23S rRNA gene was evaluated as target for the development of Sybr Green-based quantitative PCR (qPCR) for the analysis of nitrogen-fixing members of the genus Frankia or subgroups of these in soil. A qPCR with a primer combination targeting all nitrogen-fixing frankiae (clusters 1, 2 and 3) resulted in numbers similar to those obtained with a previously developed qPCR using nifH gene sequences, both with respect to introduced and indigenous Frankia populations. Primer combinations more specifically targeting three subgroups of the Alnus host infection group (cluster 1) or members of the Elaeagnus host infection group (cluster 3) were specific for introduced strains of the target group, with numbers corresponding to those obtained by quantification of nitrogen-fixing frankiae with both the 23S rRNA and nifH genes as target. Method verification on indigenous Frankia populations in soils, i.e. in depth profiles from four sites at an Alnus glutinosa stand, revealed declining numbers in the depth profiles, with similar abundance of all nitrogen-fixing frankiae independent of 23S rRNA or nifH gene targets, and corresponding numbers of one group of frankiae of the Alnus host infection only, with no detections of frankiae representing the Elaeagnus, Casuarina, or a second subgroup of the Alnus host infection groups.     

Inferring the Evolutionary History of Mo-Dependent Nitrogen Fixation from Phylogenetic Studies of nifK and nifDK

The ability to fix nitrogen is widely, but sporadically distributed among the Bacteria and Archaea suggesting either a vertically inherited, ancient function with widespread loss across genera or an adaptive feature transferred laterally between co-inhabitants of nitrogen-poor environments. As previous phylogenetic studies of nifH and nifD have not completely resolved the evolutionary history of nitrogenase, sixty nifD, nifK, and combined nifDK genes were analyzed using Bayesian, maximum likelihood, and parsimony algorithms to determine whether the individual and combined datasets could provide additional information. The results show congruence between the 16S and nifDK phylogenies at the phyla level and generally support vertical descent with loss. However, statistically significant differences between tree topographies suggest a complex evolutionary history with the underlying pattern of vertical descent obscured by recurring lateral transfer events and different patterns of evolution between the genes. Results support inheritance from the Last Common ancestor or an ancient lateral transfer of the nif genes between Bacteria and Archaea, ongoing gene transfer between cohabitants of similar biogeographic regions, acquisition of nitrogen-fixing capability via symbiosis islands, possible xenologous displacement of one gene in the operon, and possible retention of ancestral genes in heterocystous cyanobacteria. Analyses support the monophyly of the Cyanobacteria, αβγ-Proteobacteria, and Actinobacteria (Frankia) and provide strong support for the placement of Frankia nif genes at the base of combined the Cyanobacteria/Proteobacteria clades.     

Analysis of codon usage in genes for nitrogen fixation from phylogenetically diverse diazotrophs

Summary A cluster analysis based on codon usage in genes for biological nitrogen fixation (nif genes) grouped diazotrophs into three distinct classes: anaerobes, cyanobacteria, and aerobes. In thenif genes ofKlebsiella pneumoniae there was no evidence for selection pressure in favor of highly translatable codons. However, in the nitrogen regulatory operonglnAntrBntrC of enteric bacteria the stoichiometrically high level of glutamine synthetase may be facilitated by the presence of efficiently translatable codons inglnA. Thenif genes of the cyanobacteriumAnabaena showed codon selection in favor of translational efficiency. Computation of codon adaptation indices for expression in heterologous systems indicated that the reading frames most suitable for expression ofnif genes inEscherichia coli, Bacillus subtilis, andSaccharomyces cerevisiae were present in azotobacters, clostridia, and cyanobacteria, respectively. In codon-usage-based cluster analysis, type 3 nitrogenase genes ofAzotobacter vinelandii grouped along with type 1 and type 2 genes. This is in contrast to the nucleotide sequence-based multiple alignment in which type 3 nitrogenase genes ofA. vinelandii have been reported to cluster with entirely unrelated diazotrophs such as methanogens and clostridia. This may be indicative of lateral transfer ofnif genes among widely divergent taxons. The chromosomal- and plasmid-locatednif genes of rhizobia also cluster separately in nucleotide sequence-based analysis but showed similar codon usage. These analyses suggested that the phylogeny ofnif genes drawn on the basis of nucleotide sequence homology was not masked by the taxon-specific pressure on codon usage.     

Water use and yield of tomatoes under limited water and excess boron

The role of tripartite associations among Frankia, Alpova diplophloeus (an ectomycorrhizal fungus) and Alnus tenuifolia in growth, nitrogen fixation, ectomycorrhizal formation, and mineral acquisition of A. tenuifolia was investigated. Seedlings of A. tenuifolia were planted in pots containing a mixture of ground basalt–perlite, or perlite alone, which served as the control. The seedlings were inoculated with Frankia isolated from root nodules of alder, followed by spores of A. diplophloeus and grown for 5 months in a greenhouse. The seedlings grown in the pots with a mixture of ground basalt–perlite after dual inoculation with Frankia and A. diplophloeus had the heaviest shoots and root nodules in dry weight, and showed the greatest nitrogen-fixing ability measured by acetylene reduction. Ectomycorrhizae formed with A. diplophloeus increased when this fungus was inoculated together with Frankia. The mineral composition (P, K, Ca, Fe, Mg, Mn, Na, Si and Al) in the seedlings was also determined. The results of these experiments showed that the tripartite associations could improve the growth, nitrogen fixation and mineral acquisition (rock solubilization) of A. tenuifolia.     

Concurrent evolution of nitrogenase genes and 16S rRNA in Rhizobium species and other nitrogen fixing bacteria

It was known that nitrogenase genes and proteins are well conserved even though they are present in a large variety of phylogenetically diverse nitrogen fixing bacteria. This has lead to the speculation, among others, that nitrogen fixation (nif) genes were spread by lateral gene transfer relatively late in evolution. Here we report an attempt to test this hypothesis.We had previously established the complete nucleotide sequences of the three nitrogenase genes from Bradyrhizobium japonicum, and have now analyzed their homologies (or the amino acid sequence homologies of their gene products) with corresponding genes (and proteins) from other nitrogen fixing bacteria. There was a considerable sequence conservation which certainly reflects the strict structural requirements of the nitrogenase iron-sulfur proteins for catalytic functioning. Despite this, the sequences were divergent enough to classify them into an evolutionary scheme that was conceptually not different from the phylogenetic positions, based on 16S rRNA homology, of the species or genera harboring these genes. Only the relation of nif genes of slow-growing rhizobia (to which B. japonicum belongs) and fast-growing rhizobia was unexpectedly distant. We have, therefore, performed oligonucleotide cataloguing of their 16S rRNA, and found that there was indeed only a similarity of S _AB=0.53 between fast- and slowgrowing rhizobia.In conclusion, the results suggest that nif genes may have evolved to a large degree in a similar fashion as the bacteria which carry them. This interpretation would speak against the idea of a recent lateral distribution of nif genes among microorganisms.