Different dynamics of genome content shuffling among host-specificity groups of the symbiotic actinobacterium Frankia
© Kucho et al.; licensee BioMed Central Ltd. 2014
Received: 7 October 2013
Accepted: 9 July 2014
Published: 19 July 2014
Frankia is a genus of soil actinobacteria forming nitrogen-fixing root-nodule symbiotic relationships with non-leguminous woody plant species, collectively called actinorhizals, from eight dicotyledonous families. Frankia strains are classified into four host-specificity groups (HSGs), each of which exhibits a distinct host range. Genome sizes of representative strains of Alnus, Casuarina, and Elaeagnus HSGs are highly diverged and are positively correlated with the size of their host ranges.
The content and size of 12 Frankia genomes were investigated by in silico comparative genome hybridization and pulsed-field gel electrophoresis, respectively. Data were collected from four query strains of each HSG and compared with those of reference strains possessing completely sequenced genomes. The degree of difference in genome content between query and reference strains varied depending on HSG. Elaeagnus query strains were missing the greatest number (22–32%) of genes compared with the corresponding reference genome; Casuarina query strains lacked the fewest (0–4%), with Alnus query strains intermediate (14–18%). In spite of the remarkable gene loss, genome sizes of Alnus and Elaeagnus query strains were larger than would be expected based on total length of the absent genes. In contrast, Casuarina query strains had smaller genomes than expected.
The positive correlation between genome size and host range held true across all investigated strains, supporting the hypothesis that size and genome content differences are responsible for observed diversity in host plants and host plant biogeography among Frankia strains. In addition, our results suggest that different dynamics of shuffling of genome content have contributed to these symbiotic and biogeographic adaptations. Elaeagnus strains, and to a lesser extent Alnus strains, have gained and lost many genes to adapt to a wide range of environments and host plants. Conversely, rather than acquiring new genes, Casuarina strains have discarded genes to reduce genome size, suggesting an evolutionary orientation towards existence as specialist symbionts.
Frankia is a genus of soil actinobacteria with unique abilities to fix atmospheric dinitrogen (N2) and establish endosymbiotic associations with actinorhizal plants comprising various non-leguminous trees from eight dicotyledonous families [1–3]. This symbiosis, in which Frankia reduces N2 to ammonium and supplies the resulting product to host plants, takes place in root nodules. As a result of the symbiosis, actinorhizal plants grow rapidly, even in nutrient-poor soils, and improve soil fertility. Frankia strains are classified into four host-specificity groups (HSGs) that establish symbiosis with distinct host plant families . “Alnus” strains infect plant species in Myricaceae and the genus Alnus of Betulaceae. “Casuarina” strains infect plant species in the genera Casuarina and Allocasuarina of Casuarinaceae. “Elaeagnus” strains exhibit a broader host range, infecting plant species in five families (including Elaeagnaceae) of the orders Fagales and Rosales. “Rosaceous” strains infect plant species in four families of orders Rosales and Cucurbitales, although no strains have yet been isolated in pure culture. In a phylogenetic tree generated from 16S rDNA sequences, strains belonging to each HSG cluster together in distinct clades .
In 2007, complete genome sequences were determined for representative Frankia strains from Alnus, Casuarina, and Elaeagnus HSGs . A surprising finding was that despite close phylogenetic relationship (>97.8% identity for 16S rDNA), genome sizes were very different among HSGs. The largest genome (Elaeagnus strain EAN1pec) is 9.0 Mbp and contains approximately 7,400 genes, whereas the smallest one (Casuarina strain CcI3) is only 5.4 Mbp and comprises about 4,600 genes. Alnus strain ACN14a possesses an intermediate-sized genome (7.5 Mbp) of approximately 6,800 genes. This size divergence is the largest reported for any such closely related soil bacteria. Genome size of these strains correlates with the breadth of their host ranges. Comparative genome analysis has revealed that the difference in genome size is due to acquisition, loss, and duplication of genes occurring at different rates in different strains .
Two studies have uncovered evidence suggesting how such extensive diversification has occurred in Frankia genomes. Since they are particularly prevalent in Frankia genomes and indeed retain the ability to be excised from chromosomes, actinomycete integrative and conjugative elements (AICEs) may play a role in gene loss and acquisition . Homologous recombination between insertion sequences (IS) could have also caused deletions of chromosomal segments, as genes contained in IS-rich regions of ACN14a and EAN1pec genomes are absent in the smallest genome, that of CcI3 .
In the present study, we analyzed content and size of 12 Frankia genomes using in silico comparative genome hybridization (CGH) and pulsed-field gel electrophoresis (PFGE) to investigate within-HSG diversity of Frankia genomes.
Genome sequencing of Frankiastrains
Frankia strains used in this study
Alnus viridis subsp. crispa
A. hirsuta ssp. microphylla
Petersham, U. S. A.
Ohio, U. S. A.
E. umbellata ssp. rotundifolia
Number of low-coverage-rate (LCR) genes
No. of genesa
No. of LCR genes
Confirmation by PCR
We used PCR to confirm whether the identified LCR segments were structurally missing in the query genomes. We designed primer sets that flanked individual or clustered LCR segments and performed PCR using Asi1, Ceq1, and Ema2 genomic DNA as templates. Most amplification products (69% for Asi1, 100% for Ceq1, and 75% for Ema2) were smaller than the size expected based on reference genome sequences (Additional file 2), indicating that those LCR segments were missing in the query genomes. In contrast, some bands were larger than the expected size, suggesting insertion of DNA segments at these loci (Additional file 2).
LCR gene properties
Dominant functions of the top 10 nonredundant LCR genes identified in each studied Frankia host-specificity group
Hypothetical, conserved hypothetical, unknown function
Nonribosomal peptide and polyketide synthetases
Amino acids metabolism
Hypothetical, conserved hypothetical, unknown function
Restriction and modification system
Putative ATP/GTP-binding protein
AMP-dependent synthetase and ligase
Hypothetical, conserved hypothetical, unknown function
Methyltransferase of unknown function
Clustering of LCR genes
LCR gene clusters associated with potential insertion/deletion elements
AICEs are prevalent in Frankia genomes, and have been experimentally confirmed to retain their ability to be excised from chromosomes . Although three AICEs have been identified in ACN14a and CcI3 genomes, and four in the EAN1pec genome , only one AICE—in ACN14a (Faln2929)—corresponded closely to any LCR gene clusters in our study (data not shown).
DNA segments flanked by homologous sequences (direct repeat sequences) or flanked by ISs from the same family can be excised from chromosomes. LCR gene clusters associated with direct repeat sequences were relatively frequent in ACN14a and EAN1pec, but nevertheless represented only a small fraction of observed clusters (Table 4). In the EAN1pec genome, several clusters were associated with ISs belonging to the same family. To summarize, however, few LCR gene clusters were associated with elements previously known to be involved in insertion and deletion of DNA segments.
In this study, we used an in silico CGH method based on the application of next-generation sequencing technology. Resolution obtained using in silico CGH is higher than that available from DNA array-based CGH. In addition, in silico CGH experiments are less time-consuming, as this technique does not require construction of DNA arrays. On the other hand, this method is inferior to comparative analyses that use assembled genome sequences and homology search programs. When a large number of undetermined genomes need to be compared, however, in silico CGH is useful, because complete genome assembly is very laborious.
Recent comparative genomic studies have revealed that bacterial genome contents vary greatly, even among closely related species and strains [15–19]. As confirmed in our study, this is also true for Alnus and Elaeagnus HSG Frankia strains. An unexpected and novel finding of our study, however, is that this diversity varies depending on the HSG. Alnus strains lacked 14–18% of genes present in a reference genome from the same HSG, whereas more than 20% of genes were absent in Elaeagnus strains, with over 30% lacking in two strains (EP01 and EUr01) (Table 2). These divergences are much greater than that observed between the actinobacterial species Streptomyces coelicolor and Streptomyces lividans (7%) . In the case of Escherichia coli, a comparable level (about 25%) of divergence occurs between pathogenic and non-pathogenic strains . Because Frankia and Streptomyces occupy the same ecological niche, i.e. soil, environmental factors offering differing opportunities for horizontal gene exchange within the bacterial community cannot be responsible for the discrepancy. Inherent properties specific to Frankia, such as domino effects (see below), may allow such dynamic changes.
In spite of these remarkable levels of gene loss, PFGE revealed that actual genome sizes of Alnus and Elaeagnus query strains were not as small as expected based on total LCR gene length (Figure 6). This result indicates that these query strains carry genes that are absent in the reference genomes, thus compensating for the reduced genome size due to gene loss. Insertion of additional DNA segments was indeed observed in query strains (Additional file 2) based on genomic PCR. These strains have thus both lost and acquired significant numbers of genes over the course of evolution; as a consequence, genome contents have diverged greatly, even within the same HSG. Interestingly, such shuffling of genome content appears to have occurred to different extents between the two HSGs. More dynamic shuffling has taken place in genomes of Elaeagnus strains than in Alnus, as evidenced by the greater extent of gene loss (number of LCR genes; Table 2) and higher compensated genome size (Figure 6) in the Elaeagnus HSG. In Alnus strains, gene acquisition and loss seems to have been mostly balanced, because the number of LCR genes and genome size are similar among strains (Table 2 and Figure 6); this balance was not well maintained in Elaeagnus strains.
Unlike the other two HSGs, very few LCR genes were identified among Casuarina strains (Table 2), indicating that genome contents were highly similar within the HSG. In particular, strains CaE03 and CaE04 were missing only two or no LCR genes, respectively (Table 2), revealing that these query strains possessed almost all the genes in the reference genome (CcI3). Genome sizes of the query strains were significantly smaller than that of CcI3, however (Figure 6). These results suggest that some components of multigene families in the CcI3 genome were missing in CaE03 and CaE04. Normand et al.  have pointed out that transposase genes are frequently duplicated in the CcI3 genome, forming large multigene families. Loss of such transposase genes may consequently be responsible for the observed size reductions.
When complete sequences were obtained for three representative Frankia strains, the most surprising finding was their unusual size divergence. To explain the biological significance of this divergence, genome size and content have been proposed to influence host range and biogeographical adaptation of bacterial strains . Casuarina strain CcI3 has the smallest genome, consistent with the narrowest range of hosts and the limited environment of its host plants’ habitat (temperate regions of Australia) . In contrast, Elaeagnus strain EAN1pec has the largest genome, helping it to achieve the broadest host range and to adapt to the wide range of soil types and climates under which its host plants grow . Our PFGE results support this hypothesis, as this correspondence between genome size and HSG held true for the 12 strains analyzed (Figure 6). In addition, our results suggest that the dynamics of genome content shuffling, along with genome size, have contributed to these symbiotic and biogeographic adaptations. Genomes of Elaeagnus strains have likely discarded and acquired a greater number of genes to manage adaptation to a wider range of hosts (spanning five families) and encountered environments. Alnus strains may have also done so, but to a lesser extent, because their host range (spanning two families) is not as broad as that of Elaeagnus strains. Indeed, LCR genes are associated with regulatory, metabolic, and transport functions (Table 3) suggestive of such adaptive roles. In Bradyrhizobium, acquisition of genomic islands is reported to influence symbiotic nitrogen fixation properties . In contrast to Elaeagnus and Alnus, Casuarina strains have not acquired new genes, but have instead discarded them to reduce their genome sizes; this suggests an evolutionary orientation towards existence as specialist symbionts . Casuarina strains infect only a narrow spectrum of hosts, spanning two genera, and show reduced saprophytic activity which is evidenced by the fact that these strains have not been isolated from soils outside the native habitats of their host plants [21, 22]. Such reductive genome evolution is often observed in obligate symbiotic bacteria [23, 24].
Most detected LCR gene clusters were not flanked by elements known to be associated with genomic islands  (Table 4). This is similar to the case of E. coli. Because current cluster structure is a product of multiple DNA rearrangement steps, elements functional in the past may no longer be located at cluster termini. We therefore cannot determine whether the disparate occurrence of such elements explains differences in genome stability.
On the other hand, den Bakker et al.  have proposed a “domino” effect theory to explain why a particular genomic region is subject to active gene acquisition and loss. If a genome has acquired a genomic island that encodes beneficial gene products, the island will be maintained. Most parts of the island will be functionally neutral, however; they may easily accept insertion and deletion of genes without losing the island’s adaptive value, making the region a hot spot for gene exchange. We can use this hypothesis to explain the different dynamics of genome content shuffling observed in Frankia; the more genomic islands (LCR gene clusters) in a genome, the more chances for gene acquisitions and losses.
Our results suggest that two genomic properties have affected diversity in host plant range and biogeography in Frankia strains. The first property, genome size, was previously proposed by Normand et al.  and has been validated by our study. The second property is the dynamics of genome content shuffling. In other words, Elaeagnus strains have both retained and exchanged a large number of accessory genes to adapt to diverse host plant species, soil types, and climates. In contrast, Casuarina strains have discarded rather than acquired genes to limit hosts and inhabited environments, suggestive of an evolutionary preference for specialist symbiosis. Differences in the extent of genome content shuffling can be partially explained by domino effects: if a strain carries more genomic islands, then more neutral regions accompany them, thus enhancing genome flexibility towards gene acquisition and loss.
Frankia strains AH1, AHm01, CaE04, EP01, EU05, and EUr01 were isolated using the differential filtration method  from root nodules collected in the field (Table 1). Lobes of fresh nodules were sterilized with 1% sodium hypochlorite for 5 min, washed with sterilized water, and homogenized in a mortar. The homogenates were passed through filters with 50- and 20-μm nylon mesh screens. Plant residues and Frankia vesicle clusters collected through filtration were mixed in 100-ml flasks with 40 ml of modified BAP medium  lacking ammonium chloride (BAP-). The flasks were placed at 29°C in darkness. Frankia strain T7 was isolated from root nodules of Casuarina cunninghamiana. The fresh nodules were washed and dissected into individual lobes and surface-sterilized as described in . Each lobe was checked for sterility in sterile nutrient-rich medium. Nodules free from contaminant were dissected and transferred to 125-ml flasks containing modified BAP medium  and incubated at 28°C. Frankia filaments were homogenized and diluted 1:100 (v/v) with sterile distilled water; 1 ml of the suspension was then transferred to melted agar DPM medium . The plate was agitated, allowed to solidify, and incubated at 28°C for 3 weeks. A single colony was picked up, homogenized, and cultured in liquid B medium .
Frankia strains were maintained in BAP (ACN14a), BAP- (AH1, AHm01, CaE03, CaE04, T7, EP01, EU05, and EUr01), BAP-T  (CcI3), or Qmod  (Asi1, Mru1, Ceq1, and Ema2) media in tissue culture flasks (TPP, Trasadingen, Switzerland) at 28°C.
Genomic DNA preparation
Frankia cells were suspended in TE buffer (10 mM Tris-HCl [pH 8.0] and 1 mM EDTA) containing 8 mg ml-1 lysozyme, and incubated at 37°C for 1 h. Cells were collected by centrifugation, and genomic DNA was purified using a DNeasy Plant Mini kit (Qiagen, Hilden, Germany) according to the manufacturer’s instructions.
Phylogenetic analysis of 16S rDNA
The full-length 16S rDNA region was amplified by PCR using primers CcI3 16S rRNA f1 (5'-TTGATGGAGAGTTTGATCCTGG-3') and CcI3 16S rRNA r1 (5'-AGAAAGGAGGTGATCCAGC-3'). Residual primers and nucleotides were removed by exonuclease I (Takara Bio, Ohtsu, Japan) and shrimp alkaline phosphatase (Roche, Mannheim, Germany), and PCR products were directly sequenced using BigDye terminator v3.1 (Applied Biosystems, Foster City, CA, USA). A phylogenetic tree was constructed by the neighbor-joining method  using Genetyx (Genetyx, Tokyo, Japan). GenBank accession numbers of generated sequences are as follows: ACN14a, NC_008278.1; AH1, AB849940; AHm01, AB849941; Asi1, AB847121; Mru1, AB848357; CcI3, NC_007777.1; CaE03, AB849939; CaE04, AB849942; Ceq1, AB848358; T7, AB850642; EAN1pec, NC_009921.1; Ema2, AB848359; EP01, AB849943; EU05, AB849944; EUr01, AB849945; and Purshia nodule, AF034776.
Next-generation genomic sequencing
We sequenced genomes of Frankia strains using a SOLiD 4 next-generation sequencing system (Applied Biosystems). Libraries were generated from 1 μg genomic DNA using a SOLiD fragment library construction kit. Templated beads were prepared with a SOLiD ePCR kit v2 and XD beads enrichment kit, and then deposited on a glass slide using a SOLiD XD slide and deposition kit v2. Fifty base pairs at the ends of library fragments were sequenced using a SOLiD ToP fragment BC sequencing kit and a SOLiD ToP instrument buffer kit. All experiments were performed according to the manufacturer’s instructions.
To compare the content of Frankia genomes, we used a strategy named in silico CGH (Additional file 1), which has been used to find missing genes in bacterial genomes . The 50-bp reads (query reads) obtained from Frankia genomes of unknown sequence (query genomes) were mapped to a reference genome whose complete sequence had already been reported  — Frankia strains ACN14a (GenBank: NC_008278.1), CcI3 (NC_007777.1), and EAN1pec (NC_009921.1). Mapping was conducted using Bioscope software (Applied Biosystems). The term “map” means to match an individual 50-bp query read to a region with significant sequence similarity on a reference genome. Regions of a reference genome onto which few or no reads are mapped were deduced to be absent in the query genome. To quantitatively evaluate the mapping results, we dissected reference genome sequences into two types of segments—gene and IGR—and calculated “coverage rate” of each segment. Coverage rate was calculated as the percentage of nucleotides in the segment that were mapped by one or more reads (Additional file 1). A low coverage rate indicated that few query reads were mapped to a segment in a reference genome; in such cases, that segment was likely absent in the query genome. Coverage rates of all the genes in Alnus, Casuarina and Elaeagnus strains are shown in Additional file 4.
Data for GC3 and CAI were obtained from the MaGe database (https://www.genoscope.cns.fr/agc/mage/). A list of ISs annotated for Frankia genomes was taken from Bickhart et al. . We used the ssearch program (http://fasta.bioch.virginia.edu/fasta_www2/fasta_intro.shtml), which implements the Smith-Waterman algorithm , to find repeat sequences at ends of LCR gene clusters.
We conducted PCR using genomic DNAs (10 ng) of strains Asi1, Ceq1, and Ema2 as templates, along with the primers listed in Additional file 5, GC buffer I (Takara Bio), and EX Taq polymerase (Takara Bio).
Pulsed-field gel electrophoresis
Cells of Frankia were harvested from 5–15 ml of culture solution and resuspended in 0.3 ml HE buffer (10 mM 4-[2-hydroxyethyl]-1-piperazineethanesulfonic acid [pH 8.0] and 1 mM EDTA). The cell suspension was mixed with an equal volume of 2% low-melting agarose (Agarose-LM plaque; Nacalai Tesque, Kyoto, Japan) in HE buffer and solidified in plug molds (Bio-Rad, Hercules, CA, USA). The agarose plugs were incubated with 2 mg ml-1 lysozyme in HE buffer at 37°C for 2 h, and then with 1 mg ml-1 proteinase K (Nacalai Tesque) in NDS buffer (0.5 M EDTA, 10 mM Tris-HCl [pH 8.0], and 1% SDS) at 50°C for 24 h. We removed the proteinase K solution and washed the plugs once with 10 ml HE buffer containing 0.1 mM phenylmethylsulfonyl fluoride (Nacalai Tesque) and three times with HE buffer. The plugs were then washed three times with TE buffer and equilibrated with 1× buffer supplied with the restriction enzymes. We digested DNA in 200 μl solution containing 1× restriction enzyme buffer, 0.5 mM dithiothreitol, 1 mg ml-1 bovine serum albumin, and 30 units of DraI (Roche) or 1 μl of FastDigest PsiI (Thermo Scientific, Waltham, MA, USA) at 37°C for 5 h or overnight. Electrophoresis was performed under the conditions described in Additional file 3 using the CHEF-DR III system (Bio-Rad). Chromosomes of Saccharomyces cerevisiae and Schizosaccharomyces pombe were used as size standards (Bio-Rad). Gels were stained with SYBR Gold (Life Technologies, Carlsbad, CA, USA), and electropherograms were obtained under UV irradiation.
Availability of supporting data
All the data supporting the results of this article are included as additional files.
Hideo Sasakawa as emeritus professor.
We thank Dr. Louis S. Tisa (University of New Hampshire) for providing Frankia strain CcI3, Dr. Philippe Normand (CNRS) for providing Frankia strain ACN14a, Ms. Yuu Kucho for performing next-generation sequencing, and Dr. Petar Pujic (CNRS) for sharing the PFGE protocol. This work was supported by the Asahi Glass Foundation.
- Benson DR, Silvester WB: Biology of Frankia strains, actinomycete symbionts of actinorhizal plants. Microbiol Rev. 1993, 57: 293-319.PubMed CentralPubMedGoogle Scholar
- Huss-Danell K: Actinorhizal symbioses and their N2 fixation. New Phytol. 1997, 136: 375-405. 10.1046/j.1469-8137.1997.00755.x.View ArticleGoogle Scholar
- Kucho K, Hay AE, Normand P: The determinants of the actinorhizal symbiosis. Microbes Environ. 2010, 25: 241-252. 10.1264/jsme2.ME10143.PubMedView ArticleGoogle Scholar
- Normand P, Lapierre P, Tisa LS, Gogarten JP, Alloisio N, Bagnarol E, Bassi CA, Berry AM, Bickhart DM, Choisne N, Couloux A, Cournoyer B, Cruveiller S, Daubin V, Demange N, Francino MP, Goltsman E, Huang Y, Kopp OR, Labarre L, Lapidus A, Lavire C, Marechal J, Martinez M, Mastronunzio JE, Mullin BC, Niemann J, Pujic P, Rawnsley T, Rouy Z, et al: Genome characteristics of facultatively symbiotic Frankia sp. strains reflect host range and host plant biogeography. Genome Res. 2007, 17: 7-15.PubMed CentralPubMedView ArticleGoogle Scholar
- Ghinet MG, Bordeleau E, Beaudin J, Brzezinski R, Roy S, Burrus V: Uncovering the prevalence and diversity of integrating conjugative elements in actinobacteria. PLoS One. 2011, 6: e27846-10.1371/journal.pone.0027846.PubMed CentralPubMedView ArticleGoogle Scholar
- Bickhart DM, Gogarten JP, Lapierre P, Tisa LS, Normand P, Benson DR: Insertion sequence content reflects genome plasticity in strains of the root nodule actinobacterium Frankia. BMC Genomics. 2009, 10: 468-10.1186/1471-2164-10-468.PubMed CentralPubMedView ArticleGoogle Scholar
- Normand P, Lalonde M: Evaluation of Frankia strains isolated from provenances of two Alnus species. Can J Microbiol. 1982, 28: 1133-1142. 10.1139/m82-168.View ArticleGoogle Scholar
- Nagashima Y, Tani C, Yamamoto M, Sasakawa H: Host range of Frankia strains isolated from actinorhizal plants growing in Japan and their relatedness based on 16S rDNA. Soli Sci Plant Nutr. 2008, 54: 685-693. 10.1111/j.1747-0765.2008.00320.x.View ArticleGoogle Scholar
- Zhang Z, Lopez MF, Torrey JG: A comparison of cultural characteristics and infectivity of Frankia isolates from root nodules of Casuarina species. Plant Soil. 1984, 78: 79-90. 10.1007/BF02277841.View ArticleGoogle Scholar
- Yamanaka T, Mansour SR: Nodulation of Alnus japonica and Casuarina equisetifolia in liquid culture after inoculation with Frankia. Bull FFPRI. 2013, 12: 97-103.Google Scholar
- Lalonde M, Calvert HE, Pine S: Isolation and use of Frankia strains in actinorhizae formation. Current perspectives in nitrogen fixation. Edited by: Gibson AH, Newton WE. 1981, Canberra: Australian Academy of Sciences, 296-299.Google Scholar
- Fischbach MA, Walsh CT: Assembly-line enzymology for polyketide and nonribosomal peptide antibiotics: logic, machinery, and mechanisms. Chem Rev. 2006, 106: 3468-3496. 10.1021/cr0503097.PubMedView ArticleGoogle Scholar
- Horvath P, Barrangou R: CRISPR/Cas, the immune system of bacteria and archaea. Science. 2010, 327: 167-170. 10.1126/science.1179555.PubMedView ArticleGoogle Scholar
- Hacker J, Kaper JB: Pathogenicity islands and the evolution of microbes. Annu Rev Microbiol. 2000, 54: 641-679. 10.1146/annurev.micro.54.1.641.PubMedView ArticleGoogle Scholar
- Jayapal KP, Lian W, Glod F, Sherman DH, Hu WS: Comparative genomic hybridizations reveal absence of large Streptomyces coelicolor genomic islands in Streptomyces lividans. BMC Genomics. 2007, 8: 229-10.1186/1471-2164-8-229.PubMed CentralPubMedView ArticleGoogle Scholar
- Welch RA, Burland V, Plunkett G, Redford P, Roesch P, Rasko D, Buckles EL, Liou SR, Boutin A, Hackett J, Stroud D, Mayhew GF, Rose DJ, Zhou S, Schwartz DC, Perna NT, Mobley HL, Donnenberg MS, Blattner FR: Extensive mosaic structure revealed by the complete genome sequence of uropathogenic Escherichia coli. Proc Natl Acad Sci U S A. 2002, 99: 17020-17024. 10.1073/pnas.252529799.PubMed CentralPubMedView ArticleGoogle Scholar
- Joyce EA, Chan K, Salama NR, Falkow S: Redefining bacterial populations: a post-genomic reformation. Nat Rev Genet. 2002, 3: 462-473.PubMedGoogle Scholar
- Thompson JR, Pacocha S, Pharino C, Klepac-Ceraj V, Hunt DE, Benoit J, Sarma-Rupavtarm R, Distel DL, Polz MF: Genotypic diversity within a natural coastal bacterioplankton population. Science. 2005, 307: 1311-1313. 10.1126/science.1106028.PubMedView ArticleGoogle Scholar
- Itakura M, Saeki K, Omori H, Yokoyama T, Kaneko T, Tabata S, Ohwada T, Tajima S, Uchiumi T, Honnma K, Fujita K, Iwata H, Saeki Y, Hara Y, Ikeda S, Eda S, Mitsui H, Minamisawa K: Genomic comparison of Bradyrhizobium japonicum strains with different symbiotic nitrogen-fixing capabilities and other Bradyrhizobiaceae members. ISME J. 2009, 3: 326-339. 10.1038/ismej.2008.88.PubMedView ArticleGoogle Scholar
- Benson DR, Dawson JO: Recent advances in the biogeography and genecology of symbiotic Frankia and its host plants. Physiol Plant. 2007, 130: 318-330. 10.1111/j.1399-3054.2007.00934.x.View ArticleGoogle Scholar
- Zimpfer JF, Smyth CA, Dawson JO: The capacity of Jamaican mine spoils, agricultural and forest soils to nodulate Myrica cerifera, Leucaena leucocephala and Casuarina cunninghamiana. Physiol Plant. 1997, 99: 664-672. 10.1111/j.1399-3054.1997.tb05370.x.View ArticleGoogle Scholar
- Simonet P, Navarro E, Rouvier C, Reddell P, Zimpfer J, Dommergues Y, Bardin R, Combarro P, Hamelin J, Domenach AM, Gourbiere F, Prin Y, Dawson JO, Normand P: Co-evolution between Frankia populations and host plants in the family Casuarinaceae and consequent patterns of global dispersal. Environ Microbiol. 1999, 1: 525-533. 10.1046/j.1462-2920.1999.00068.x.PubMedView ArticleGoogle Scholar
- Moran NA: Tracing the evolution of gene loss in obligate bacterial symbionts. Curr Opin Microbiol. 2003, 6: 512-518. 10.1016/j.mib.2003.08.001.PubMedView ArticleGoogle Scholar
- Kuwahara H, Takaki Y, Yoshida T, Shimamura S, Takishita K, Reimer JD, Kato C, Maruyama T: Reductive genome evolution in chemoautotrophic intracellular symbionts of deep-sea Calyptogena clams. Extremophiles. 2008, 12: 365-374. 10.1007/s00792-008-0141-2.PubMedView ArticleGoogle Scholar
- Touchon M, Hoede C, Tenaillon O, Barbe V, Baeriswyl S, Bidet P, Bingen E, Bonacorsi S, Bouchier C, Bouvet O: Organised genome dynamics in the Escherichia coli species results in highly diverse adaptive paths. PLoS Genet. 2009, 5: e1000344-10.1371/journal.pgen.1000344.PubMed CentralPubMedView ArticleGoogle Scholar
- den Bakker HC, Desjardins CA, Griggs AD, Peters JE, Zeng Q, Young SK, Kodira CD, Yandava C, Hepburn TA, Haas BJ, Birren BW, Wiedmann M: Evolutionary dynamics of the accessory genome of Listeria monocytogenes. PLoS One. 2013, 8: e67511-10.1371/journal.pone.0067511.PubMed CentralPubMedView ArticleGoogle Scholar
- Hiyoshi T, Sasakawa H, Yatazawa M: Isolation of Frankia strains from root nodules of Myrica rubra. Soil Sci Plant Nutr. 1988, 34: 107-116. 10.1080/00380768.1988.10415584.View ArticleGoogle Scholar
- Murry MA, Fontaine MS, Torrey JG: Growth kinetics and nitrogenase induction in Frankia sp. HFPArI 3 grown in batch culture. Plant Soil. 1984, 78: 61-78. 10.1007/BF02277840.View ArticleGoogle Scholar
- Mansour SR, Dewedar A, Torrey JG: Isolation, culture, and behavior of Frankia strain HFPCgI4 from root nodules of Casuarina glauca. Bot Gaz. 1990, 151: 490-496. 10.1086/337848.View ArticleGoogle Scholar
- Tzeans S, Torrey JG: Spore germination and the life cycle of Frankia in vitro. Can J Microbiol. 1989, 35: 801-806. 10.1139/m89-133.View ArticleGoogle Scholar
- Baker D, O'Keefe D: A modified sucrose fractionation procedure for the isolation of Frankia from actinorhizal root nodules and soil samples. Plant Soil. 1984, 78: 23-28. 10.1007/BF02277837.View ArticleGoogle Scholar
- Kucho K, Kakoi K, Yamaura M, Higashi S, Uchiumi T, Abe M: Transient transformation of Frankia by fusion marker genes in liquid culture. Microbes Environ. 2009, 24: 231-240. 10.1264/jsme2.ME09115.PubMedView ArticleGoogle Scholar
- Lalonde M, Calvert HE: Production of Frankia hyphae and spores as an infective inoculant for Alnus species. Forest Research Laboratory. Edited by: Gordon JC, Wheeler CT, Perry DA. 1979, Corvallis: Oregon State University, 95-110.Google Scholar
- Saitou N, Nei M: The neighbor-joining method: a new method for reconstructing phylogenetic trees. Mol Biol Evol. 1987, 4: 406-425.PubMedGoogle Scholar
- Gulig PA, de Crécy-Lagard V, Wright AC, Walts B, Telonis-Scott M, McIntyre LM: SOLiD sequencing of four Vibrio vulnificus genomes enables comparative genomic analysis and identification of candidate clade-specific virulence genes. BMC Genomics. 2010, 11: 512-10.1186/1471-2164-11-512.PubMed CentralPubMedView ArticleGoogle Scholar
- Smith TF, Waterman MS: Identification of common molecular subsequences. J Mol Biol. 1981, 147: 195-197. 10.1016/0022-2836(81)90087-5.PubMedView ArticleGoogle Scholar
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly credited. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.