Lack of the Flavin-Containing Monooxygenase FmoA Partially Impairs the Symbiotic Interaction of Mesorhizobium Huakuii with Astragalus Sinicus

Background: Flavin-containing monooxygenases (FMOs) catalyze the NADPH-dependent N- or S- oxygenation of numerous foreign chemicals, and thus may mediate interactions between microorganisms and their chemical environment. The aim of this study was to investigate the role of FMO in symbiotic nitrogen fixation of Mesorhizobium huakuii and its host plant Astragalus sinicus. Results: A mutation in M. huakuii fmoA gene was generated by homologous recombination. The fmoA mutant grew more slowly than its parental strain, and displayed decreased antioxidative capacity under higher concentration of H 2 O 2 and cumene hydroperoxide (CUOOH), indicating that FmoA plays an important role in response to the peroxides. The fmoA mutant strain displayed no difference of peroxidase activity and glutathione reductase activity, but significantly lower level of glutathione and hydrogen peroxide content than the wild type. Real-time quantitative PCR showed that the fmoA gene expression is significantly up-regulated in three different stages of nodule development. The fmoA mutant was severely impaired in its rhizosphere colonization, and its symbiotic properties in Astragalus sinicus were drastically affected. Transcriptomes in root-nodule bacteroids were analyzed and compared. A total of 1233 genes were differentially expressed, of which 560 were up-regulated and 673 were down-regulated in HKfmoA bacteroids compared with that in 7653R bacteroids. The transcriptomic data allowed us to determine additional target genes, whose differential expression was able to explain the observed the changes of symbiotic phenotype in the mutant-infected nodules. Conclusions: The fmoA gene is essential for antioxidant capacity and symbiotic nitrogen fixation. Furthermore, RNA-Seq based global transcriptomic analysis provided a comprehensive view of M. huakuii fmoA gene involved in nodule senescence and symbiotic nitrogen fixation.

FMOs in eukaryotes have been exploited as diverse biocatalysts, and have been well studied in detail [16]. The best known example of eukaryotic FMO is probably from mammals, where FMO3 is an important hepatic enzyme for the detoxification of xenobiotics including trimethylamine (TMA) [17,18]. Plant FMOs are essential for the initiate systemic accumulation of salicylic acid and systemic defense responses provoked by the virulent strains [19]. In yeast, FMOs are required for cell growth under reductive stress and involved in the maintenance or regulation of intracellular reducing potential [20]. FMOs are also present in E. coli, and the presence of FMO were detected in many bacterial genomes. However, the physiological role of bacterial FMOs has thus far not been known [18].
The functions of the flavin-containing monooxygenases are of interest not only because they oxidatively metabolizes a wide variety of nitrogen, sulfur-, and phosphorous-containing xenobiotics, but also because they appear to share various functions in protecting organisms against ROS and detoxifying xenobiotics. Here, we identified a FMO gene fmoA in M. huakuii 7653R, and the roles of M. huakuii FmoA in free-living bacteria and during N-fixing symbiosis with A. sinicus were investigated by analyzing the phenotypes of a fmoA mutant strain. A transcriptome analysis was also carried out to discover clues that might explain the differences in the nodules induced by the fmoA mutant and the wild type strain.

Construction of M. huakii fmoA mutant
The gene MCHK_8187 in M. huakuii, encoding a flavin-dependent oxidoreductase (FmoA), is expressed at high levels in the nodule during plant symbiosis. fmoA gene is predicted to encode a 350-amino acid polypeptide with an expected molecular mass of 39.4 kDa and a pI value of 5.85. To confirm the function of the fmoA gene in growth, antioxidation capacity and symbiotic nitrogen fixation, we constructed its mutant HKfmoA by a single crossover homologous recombination. In liquid AMS minimal medium with NH 4 Cl as nitrogen source and glucose as carbon source, HKfmoA grew more slowly than the parent strain 7653R (Fig. 1). The growth of HKfmoA in AMS media with glucose as sole carbon source was restored by complementation with fmoA carried on a plasmid (pBBRfmoA) (Fig. 1).

Antioxidation analysis of M. huakii fmoA mutant
To investigate the role of FmoA in protection against oxidative stress, survival assays were carried out to determine the capacity of the mutant HKfmoA to oxide hydrogen peroxide (H 2 O 2 ) and organic oxide cumene hydroperoxide (CUOOH). The survival rates of HKfmoA were not significantly affected by the presence of the lower concentrations of 1, 5 mmol/L H 2 O 2 and 0.1 mmol/L CUOOH, whereas the survival ability of mutant HKfmoA was affected by the treatment with the higher concentrations of 10 mmol/L H 2 O 2 and 1, 5 mmol/L CUOOH (Table 1). All data are averages (± SEM) from three independent experiments. a, b Values in each column followed by the same letter are not significantly different (P ≤ 0.05).
It has been reported that the yeast flavin-dependent monooxygenase (yFMO) devotes to the cellular pools of oxidized thiols that, together with GSH from glutathione reductase, create the optimum redox environment of cellular systems [20]. The role of the M. huakuii FmoA in the function of regulating the cellular redox environment was investigated by quantifying the peroxidase activity, glutathione reductase activity, hydrogen peroxide content and GSH content in 5 mM H 2 O 2 -induced oxidative stress conditions. The results showed that the peroxidase activity and glutathione reductase activity in the mutant HKfmoA were not different from that of parent strain 7653R, but its GSH content and hydrogen peroxide content were significantly lower (Table 2). Therefore, FmoA may play important roles in oxidative stress resistance and regulating the cellular redox environment in M. huakuii.

Rhizosphere colonization by M. huakuii strains
Competition between the fmoA mutant HKfmoA and parent strain 7653R for growth in the plant rhizosphere was measured by inoculating a low number of M. huakuii strains into the A. sinicus rhizosphere (10 3 to 10 4 CFU per seedling), and determining the total amount of bacteria after 7 days. When the mutant HKfmoA and the parent 7653R were inoculated alone into short-term colonization of sterile plant rhizosphere, the ratio between mutant HKfmoA and wild-type7653R was 34.4% (Fig. 2). When both strains were inoculated together in equal proportion, mutant HKfmoA was at a significant disadvantage (23.20% of bacteria recovered) compared to the wild type. In the same case, even when mutant HKfmoA was co-inoculated at a 10-fold excess over the wild type 7653R, it still accounted for only 40.33% of bacteria recovered (Fig. 2). Fmos can mediate interactions between microorganisms and their environment, and the present of Fmos in Rhizobium can provide a competitive advantage in competition for survival in the rhizosphere soil. The decreased ability of the fmoA mutant to grow in a sterile host plant rhizosphere shows that FmoA is essential for colonization of the host plant rhizosphere by M. huakuii.

Effect of mutation of fmoA on nodulation
To observe the nodulation characteristics and measure nitrogenase activity of the fmoA mutant, A. sinicus seedlings were inoculated with M. huakuii strains, and,the number, structure, and acetylene reduction activity (ARA) values of the nodules were analyzed 28 days post inoculation ( Figure S1). The results showed that there was no statistically significant difference in the number of nodules per plant between plants inoculated strain HKfmoA and plants inoculated with wild-type 7653R (Table 3). A notable feature of our study was that the nitrogen fixation capacity was severely affected in the fmoA mutant HKfmoA, with a reduction from 23.42 nmol of C 2 H 4 plant − 1 h − 1 in 7653R to 7.40 nmol of C 2 H 4 plant − 1 h − 1 in the mutant (Table 3). Uninoculated controls showed exhibited severe symptoms of nitrogen deficiency. When recombinant plasmid pBBRfmoA was introduced into mutant HKfmoA, plants inoculated with the resulting strain HKfmoA (pBBRfmoA) formed normal nodules (Fig. 3), and there was no significant difference in nitrogen-fixing ability between 7653R-inoculated plants and HKfmoA-inoculated plants ( Table 3). The structural organization of mature root nodules of A. sinicus is studied by thin-sectioning and scanning electron microscopy techniques. Microscopic analysis of HKfmoA nodules showed that they were spherical rather than elongated (Fig. 3). Moreover, bacterial size, structure and membrane incrassation, as observed by electronic microscopy of nodule sections, indicated that the fmoA mutant bacteroids had undergone premature senescence (Fig. 3). In addition, poly-b-hydroxibutyrate granules were only detected in the fmoA mutant bacteroids".

Expression level of the fmoA gene in 7653R-inoculated nodules
Expression of the fmoA gene in root nodules collected at 14 days, 28 days and 42 days post inoculation was detected by qRT-PCR (Fig. 4). The fmoA gene expression is significantly up-regulated in the early stage of nodule formation (14 d), the nodule maturation stage (28 d) and the late stage (42 d) of nodule development and senescence, and the fmoA gene has the highest expression level (more than 7-fold) in nodules at 14 days post inoculation. Therefor, fmoA gene expression was induced during the symbiotic interaction when compared with free-living cells growing in synthetic medium. and FmoA may play an important role in persistence of nodule bacteroids and prevention of premature nodule senescence.

RNA-seq analyses of gene expression in the nodule bacteroids
To investigate the global effects of FmoA on the gene transcription pattern of M. huakuii in the nodules of A. sinicus, the root nodules infected with a M. huakuii wild-type or an fmoA gene mutant strain were processed at 28 dpi and analyzed by transcriptomics. cDNA samples from the nodule formed by M. huakuii HKfmoA and 7653R were sequenced using Illumina paired-end sequencing. A total of 53-million clean sequencing reads was obtained from the RNA-seq transcriptomics analysis of two samples, with average reads of 26.5-million reads per sample. In total, 6721 expressed genes were detected in both the M. huakuii strains during symbiosis through RNA-SEq. Comparative analysis of gene expression levels found that 1233 genes were differentially expressed (p-value ≤ 0.01, with log 2 (FC) ≥ 1.5 and ≤ − 1.5), of which 560 were up-regulated and 673 were down-regulated in M. huakuii HKfmoA bacteroids compared to in M. huakuii 7653R bacteroids. Among these differentially expressed genes, 1164 (94.4%) were located on the chromosome, 32 (2.6%) were located on the plasmid pMHa, and 37 (3.0%) were located on the symbiotic plasmid pMHb. Symbiotic genes such as nif, fix and nod genes, are located on symbiotic plasmid pMHb. However, there was no difference in expression levels of these genes between fmoA mutant bacteroids and wild-type 7653R bacteroids.
Among the top 500 regulated genes, 242 were up-regulated in nodules induced by the fmoA mutant. All the cell motility genes were found to be significantly up-expressed (Fig. 5), the category includes the cell motility flagella structural genes (fliIKLMPQ and flgABCDEFGHI), a flagellar biosynthesis repressor gene (flbT), a flagellar motor stator protein gene motA and a chemotaxis protein gene motC (MCHK_4458) ( Table 4). And the two categories "carbohydrate transport and metabolism", and "coenzyme transport and metabolism" were also found to be significantly over-represented among these genes (Fig. 5). 35 carbohydrate transport and metabolism genes were found among the top 500 genes showing increased expression in mutant nodules: ten coding for ABC transporters; three coding for tripartite tricarboxylate transporters. The coenzyme transport and metabolism were up-regulated including five biotin mechanism genes, four cobalt mechanism genes, two nicotinatenucleotide genes, and a thiamine biosynthesis gene thiS. Interestingly, these coenzymes all play an indispensable role in rhizobial growth or nitrogenase activity (Table 4). In contrast, among the 258 genes down-regulated in fmoA nodule bacteroids, the four categories "transcription", "defense mechanisms", "signal transduction mechanisms", and "posttranslational modification, protein turnover, chaperones" were notably over-represented (Fig. 5). All the defense mechanism genes were found to be significantly down-expressed. The defense mechanism category includes seven genes encoding stress response and virulence proteins (Table 5). And the category "posttranslational modification, protein turnover, chaperones" also includes stress response genes encoding heat-shock proteins and antioxidant proteins ( Table 5). The signal transduction mechanism category includes two histidine kinases, a sensory box protein, a circadian clock protein KaiC and a phage-shock protein (

Discussion
Flavin-containing monooxygenase (FMO), which can metabolize numerous foreign chemicals, is a monooxygenase that uses the reducing equivalents of NADPH to reduce one atom of molecular oxygen to water, while the other atom is used to oxidize the substrate [21]. An important issue is that M. huakuii fmoA expression is elevated in nitrogen-fixing bacteroids of the A. sinicus root nodules, but no studies have been published regarding the connection between the legumes-root nodule bacteria nitrogen fixing system and the function of FmoA. In this study, we focus on a fmoA mutant strain of M. huakuii that is affected with regard to its symbiotic capacity and oxidative stress response.
Flavin-containing monooxygenases (FMOs) have been reported to be important for the disposition of many therapeutics, environmental toxicants, and nutrients and, thus, mediate interactions between organisms and their chemical environment [22,23]  such as peroxidase and glutathione reductase, and the content of GSH were further investigated. The antioxidant enzyme activities of mutant HKfmoA were no difference compared to that of wild type strain 7653R, but its GSH content was significantly lower ( Table 2), suggesting that M. huakuii FmoA-deficiency-mediated decrease in glutathione increases the sensitivity of mutant cells to peroxides.
Since M. huakuii fmoA gene expression is significantly up-regulated on the whole nodulation process, and it's the highest expression level occurred at 14 days after inoculation (Fig. 4). The roles of FmoA in symbiotic nitrogen fixation and the colonization of the plant rhizosphere were further studied by plant experiments. Although the expression levels of symbiotic genes such as nif, fix and nod, were not significantly different in bacteroids from the fmoA mutant and the wild-type strain, A. sinicus plants inoculated with the fmoA mutant exhibited a large decrease in the nitrogen-fixing activity of root nodules (reduced by more than 65%) ( Table 3). Further investigation revealed that fmoA mutant HKfmoA-induced nodules were spherical rather than elongated, underwent premature senescence, and PHB granules could be detected in the fmoA mutant bacteroids but not in those of the wild-type strain (Fig. 3). Moreover, the M. huakuii fmoA mutant was unable to compete efficiently in the rhizosphere with its wild-type 7653R. Flavin-containing monooxygenases metabolize a vast array of foreign chemicals including antioxidants, phytochemicals and dietary components, and, thus, mediate interactions between bacteria and their chemical environment [19]. The results showed that bacterial FmoA was important for adaptation to the microenvironment of the plant host (Fig. 2). Overall, considering the poor nitrogen-fixing ability of its nodules, the mutant in fmoA gene has a profound influence on the whole nodulation process.
The RNA-seq experiments were performed to provide a foundation for assessing the influence of FmoA on the symbiotic nitrogen fixation. In this study, the carbohydrate transport and metabolism, coenzyme transport and metabolism, and flagellar genes were found to be significantly up-expressed in fmoA mutant induced bacteroids (Fig. 5). Flagellar motility is a critical environmental adaptation for the plant-associated bacteria such as Rhizobium that allows bacteria to escape adverse conditions and populate new environments [27]. The upregulation of flagellar genes in this case might be due to the hostile environment in which the bacteroids with low nitrogenase activity are embedded [28]. In the fmoA mutant-induced nodules, many symbiosomes were aberrant and the bacteroid membrane showed incrassation (Fig. 3). The peribacteroid membrane may be relatively impermeable to sugars and so dictate the carbon source(s) available to the bacteroids [29]. This might also be the reason for the over-represented up-regulation category "coenzyme transport and metabolism". The up-expression of carbohydrate transport and metabolism category containing 10 ABC transporter genes might be due to a delicate balance control between sufficient acquisition and overload (Table 4). In this study, PHB was occured in the fmoA mutant bacteroids (Fig. 4). During the formation of bacteroids in indeterminate-type nodules such as M. huakuii, the PHB granules are broken down. PHB can be used as a carbon and energy source for bacteroid formation, but most rhizobial species such as M. huakuii do not accumulate it during symbiosis with legumes. Biochemically, PHB synthesis directly competes with N 2 fixation for reductant. PHB synthesis is apparently a concomitant reduction in protein synthesis, a process coupled to ATP formation and utilization [30]. PHB granules occurred in undergoing senescence bacteroids infected by fmoA mutant implied that the energy and carbon metabolism was shifted, NAD(P)H was channeled into other biosynthesis reactions, such as PHB synthesis [31].
Among the the down-exprssion genes, the four categories "Transcription", "Defense mechanisms", "Signal transduction mechanisms", and "Posttranslational modification, protein turnover, chaperones" were significantly over-represented (Fig. 5). One of the remarkable findings of the RNA-seq analysis was that nearly all the genes associated with stress response and virulence were significantly differentially down-expressed ( Table 5). The symbiotic nodule is prone to high levels of ROS due to the high rate of respiration necessary to supply energy required for nitrogen reduction by nitrogenase [32], but increased level of ROS causes oxidative damage to important cellular macromolecules [33]. Thiol-containing molecules, such as glutathione, glutathione Stransferases, glutaredoxins, and peroxiredoxins play an important role in maintaining redox homeostasis and redox regulation [34]. Nodules induced by fmoA mutant bacteria presented the lower expression of the thiolcontaining molecules, which was associated with increased levels of superoxide accumulation. It has been reported that heat shock proteins play important roles in innate immune responses [35], and in Agrobacterium, the virulence genes are essential for attachment to plant cells [36]. The expression of three heat-shock protein and several virulence genes was also decreased in the fmoA mutant nodules (Table 5). Therefor, the results suggested that the stress response function of M. huakuii is influenced by deletion of the fmoA gene. In addition, it has been reported that the sensor histidine kinase mediated the pathogenesis by the bacterium Rhizobium radiobacter [37], and mutation of a R. leguminosarum histidine kinase gene chvG destabilized the outer membrane of R. leguminosarum, resulting in increased sensitivity to membrane stressors, and caused symbiotic defects on peas, lentils, and vetch [38]. In early senescent nodules induced by the fmoA mutant, the expression of two histidine kinases and a sensory box protein was significantly decreased, suggesting that FmoA could influence the symbiotic interaction between M. huakuii and A. sinicus by decreasing the expression of sensor molecules.
Moreover, 42 transcriptional regulator genes were found among the top 500 genes showing reduced expression (Table 5). MCHK_1343, MCHK_4626, and MCHK_5064 coding for the Crp-Fnr family transcriptional regulators, which is an important transcriptional regulator that controls the expression of a large regulon of more than 100 genes in response to changes in oxygen availability [49]; MCHK_4665 and MCHK_2899 coding for the AraC family transcriptional regulators, which play a critical role in regulating bacterial virulence factors in response to environmental stress [40]; MCHK_5909 coding for Lrp/AsnC family transcriptional regulator, which is known as feast/famine regulatory protein (FFRPs) [41]; MCHK_2407 coding for a DeoR/GlpR-type protein, which serves as transcriptional repressor or activator of either sugar or nucleoside metabolism [42,43]; MCHK_4182 coding for a PadR family transcriptional regulator that functioned as environmental sensor [44]; MCHK_1555 coding for a MarR family transcriptional regulator, which is involved in the regulation of many cellular processes, including pathogenesis [45]; MCHK_5264 coding for a ArsR family transcriptional regulator involved in symbiosis and virulence [46]; MCHK_5463 coding for transcriptional regulator GcvA, which is required for both glycine-mediated activation and purine-mediated repression of the gcvTHP operon [47]. Taken together, nodules induced by the fmoA mutant are different from those induced by the wild-type strain, and there is also a clear difference in bacteroid aspect, revealing that the fmoA mutant is negatively affected in symbiosis. The reduced nitrogen fixation ability exhibited by the fmoA mutant could be a consequence of a defect in nodule development

Conclusions
Flavin-containing monooxygenases (FMOs) catalyze the oxidation of heteroatom centers and, thus, mediate interactions between microorganisms and their chemical environment. The contribution of FmoA to symbiosis and anti-oxidative damage was investigated using the M. huakuii fmoA mutant. A quantitative RNA-Seq based transcriptomics approach was applied to reveal the global transcriptomic responses to FmoA defect in M. huakuii bacteroids from A. sinicus root nodules. The results showed a total of 1233 genes were differentially expressed, of which 560 were up-regulated and 673 were down-regulated in HKfmoA bacteroids compared to that in 7653R bacteroids. This study provided majority of these differentially expressed genes were grouped into 19 categories and a valuable insight into FmoA-mediated mechanisms during M. huakuii-A. sinicus symbiosis. Furthermore, this study has generated an abundant list of transcript from M. huakuii which will provide a fundamental basis for future functional genomic research in M. huakuii and other closely related species.

Bacterial growth and media
The strains, plasmids and primers used in this study are listed in Table S2. M. huakuii strains were grown at 28 °C in either Tryptone Yeast extract (TY) [48] or Acid Minimal Salts medium (AMS) [49]  To monitor culture growth, strains were grown at 28 °C with shaking (200 rpm) in liquid media, and culture optical density at 600 nm (OD 600 ) was measured during the culture period.

Construction and complementation of fmoA gene mutant strain of M. huakuii 7653R
A single-crossover integration mutation in fmoA was made in 7653R. Primers fmoAUP and fmoALW were used to PCR amplify the fmoA region from 7653R genomic DNA, and the 620-bp internal fragment of the fmoA gene was cloned into the Pst I and Xba I sites of pK19mob, giving plasmid pKfmoA. Plasmid pKfmoA was transferred from E. coli to 7653R and recombined into the genomic fmoA region via single crossover to give strain HKfmoA [49]. Insertions into the fmoA gene of strain 7653R were confirmed by PCR using the fmoAmap primer and a pK19mob-specific primer pK19A or pK19B [50].
To complement the fmoA mutant, primers cfmoAF and cfmoAR were used to amplify the complete fmoA gene from M. huakuii 7653R genomic DNA. The PCR product was digested with Kpn I and Xba I and cloned into the broad-host-range vector pBBR1MCS-5, resulting in plasmid pBBRfmoA. Plasmid pBBRfmoA was mated into the mutant strain HKfmoA using the triparental mating method as previously described [50].

Mutant resistance and catalase activity in relation to oxidative stress
The

Plant experiment and cytological study of nodules
Astragalus sinicus L. cultivar XY202 (Xinyang Company, China) was used as a host plant to test nodulation of the M. huakuii strains. Seeds were surface-sterilized, placed in 500 mL pots filled with sterile vermiculite containing nitrogen-free Fahraeus solution. Inoculation with M. huakuii stains was performed on 7-day-old seedlings. The cultivation was carried out in a controlled environment chamber with 16 h light/8 h dark period. Acetylene reduction activity was determined at 28 day postinoculation (dpi) as previously described [52]. The experiment consisted of two independent experiments, each of which had five repeats, and statistical differences were analyzed with one-way ANOVA (P < 0.05).
Nodules at 28 dpi were fixed for 12 h at 4 °C with 2.5% glutaraldehyde, rinsed, and post-fixed in 1.5% phosphate-buffered osmium tetroxide. Ultra-thin sections stained with lead citrate were examined using a Hitachi H-7100 transmission electron microscope [53]. Sections were cut with a microtome and stained with toluidine blue for light microscopy.

Rhizosphere colonization
Rhizosphere colonization was performed as described previously [54]. Astragalus sinicus seedlings were germinated and grown for 7 days as described above for Acetylene reduction activity, and inoculated with M. huakuii 7653R and HKfmoA in the cfu ratios 1000:0, 0:1000, 1000:1000 and 1000:10000. Shoots were cut-off after 7 days (14 days after plant), and 20 mL of sterile phosphate-buffered saline (PBS) buffer (pH 7.4) was added to the roots and vortexed for 15 min [55]. The samples were further serially diluted and plated on TY agar plates containing either streptomycin or streptomycin and neomycin, giving the total number of viable rhizosphere-and root-associated bacteria. Each treatment consisted of 10 replications, and statistical differences were analyzed with one-way ANOVA (P < 0.05).

RNA isolation and quantitative RT-PCR analysis
Quantitative real-time reverse transcription PCR (qRT-PCR) was used to determine the fmoA gene expression level, with gene-specific primers QfmoAUP and QfmoALW. The total RNA was isolated using TRIzol reagent from free-living M. huakuii 7653R cultivated in AMS liquid medium, or root nodules which were harvested from A. sinicus inoculated with strain 7653R after 28 days postinoculation. RNA were reverse transcribed into cDNA using the SuperScript II reverse transcriptase and random hexamers. Quantitative real-time PCR analysis was performed using a SYBR Premix ExTaq kit following the manufacturer's instructions on the BIO-RAD CFX96 Real-Time PCR Detection System. The 16S rRNA gene of M. huakuii 7653R was used as a calibrator gene, and the data were obtained and analysed as previously described [56] .

RNA-seq library preparation and sequencing using the illumina genome analyzer
At 28 days post inoculation, the nodules of plants inoculated with HKfmoA or 7653R were harvested, immediately frozen in liquid nitrogen and stored at − 80 °C. Total cellular RNA was isolated from frozen nodule tissues using TRIzol Reagent (Invitrogen) and RNeasy Mini Kit (Qiagen). Total RNA of each nodule sample was assessed using Agilent 2100 Bioanalyzer (Agilent Technologies, Palo Alto, CA, USA), and NanoDrop (Thermo Fisher Scientific Inc). 1 µg total RNA with RNA integrity number (RIN) value above 6.5 was used for following library preparation. The rRNA was depleted from total RNA using rRNA removal Kit. The ribosomal depleted RNA was then fragmented and reverse-transcribed into cDNA with random primers. The purified double-stranded cDNA by beads was then treated with End Prep Enzyme Mix to repair both ends and add a dA-tailing in one reaction, followed by a T-A ligation to add adaptors to both ends. Next generation sequencing library preparations were constructed according to the manufacturer's protocol. The Qsep100 (Bioptic, Taiwan, China) and Qubit 3.0 Fluorometer was used to determine the quality of the libraries.
The libraries with different indices were multiplexed and sequenced on an Illumina HiSeq instrument according to manufacturer's instructions (Illumina, San Diego, CA, USA). Sequencing was carried out by Illumina paired-end configuration. The sequencing image processing and base calling were conducted following to Illumina's protocol on the HiSeq instrument. Three independent biological replicates per sample were processed and sequenced.

Data analysis
Differences between the average of gene expression for the control and experimental groups were analyzed by the Student's t-test using SPSS software, version 18 (SPSS, Inc., Chicago, IL). For the RNA-seq study, the unique reads mapping to the M. huakuii genome were used for a differential gene expression analysis using the DESeq2 [57]. The P-values with false discovery rate are adjusted for multiple testing. The false discovery rate P-value < 0.01 and the absolute value of log 2 (FC) ≥ 1.5 and ≤ − 1.5 were used to identify statistically significant changes in gene expression. For quantitative RT-PCR analysis, p < 0.05 was considered to be statistically significant.

Consent for publication
Not applicable.

Figure 2
Bacteria recovered (7 dpi) from the rhizosphere of Astragalus sinicus plants following inoculation with wild-type 7653R and HKfmoA, both individually and together. Inoculation ratios are given on the x axis, with 1 corresponding to 103 CFU. Number of bacteria (per plant) recovered from at least ten independent plants (mean ± SEM) are shown.   Functional categories of the top 500 differentially expressed genes in A. sinicus nodules infected by the fmoA mutant versus the wild type 7653R. Bars represent the number of up-expression (black) and down-regulated (orange) genes in fmoA mutant bacteroids compared with wild type bacteroids.The number in each bar represents its up/down-expression percentage (%). C: Energy production and conversion; D: Cell cycle control, cell division, chromosome partitioning; E: Amino acid transport and metabolism; F: Nucleotide transport and metabolism; G: Carbohydrate transport and metabolism; H: Coenzyme transport and metabolism; I: Lipid transport and metabolism; J: Translation, ribosomal structure and biogenesis; K: Transcription; L: Replication, recombination and repair; M: Cell wall/membrane/envelope biogenesis; N: Cell motility; O: Posttranslational modification, protein turnover, chaperones; P: Inorganic ion transport and metabolism; R: General function prediction only; S: Function unknown; T: Signal transduction mechanisms; U: Intracellular trafficking, secretion, and vesicular transport; V: Defense mechanisms. The asterisks (*) indicate statistical significance ((>3-fold, pvalue < 0.01).

Supplementary Files
This is a list of supplementary files associated with this preprint. Click to download.