Understanding the expression dynamics of symbiont rhizobial nifH and nitrogen assimilatory NR and GS genes in dry bean (Phaseolus vulgaris L.) genotypes at various growth stages

Selection of dry bean (Phaseolus vulgaris L.) cultivars based on their symbiotic nitrogen fixation (SNF) could reduce the fertilizer N requirement. The nifH gene in the bacterial symbiont facilitates the organic N assimilation by N‐assimilatory enzymes, nitrate reductase (NR), and glutamine synthetase (GS). Fertilizer‐N recommendation of legumes could be adjusted with the relative potential of bacterial nifH gene expression. We hypothesized that quantification of bacterial and plant marker gene expression could predict the amount of total N‐assimilation. We studied the expression dynamics of dry bean NR and GS along with the Rhizobium phaseoli nifH gene in four dry bean commercial cultivars La Paz, Lariat, Windbreaker, and ND‐307 at the third trifoliate (V3) and late flowering (R2) stages using quantitative reverse transcription polymerase chain reaction. Association between gene expression and the amount of SNF using 15N stable isotope dilution was determined. At V3 stage, Lariat had a highest nifH expression, followed by ND‐307, La Paz, and Windbreaker. At R2 stage, ND‐307 had the highest nifH expression followed by Windbreaker, La Paz, and Lariat. At R2 stage, the nifH gene expression was upregulated, whereas NR and GS gene expressions were downregulated compared with the V3 stage, indicating dry bean might be benefitted from late‐season N application. Thus, cultivar dependent nifH, NR, and GS gene expression varied depending on the growth stage. The amount of SNF was significantly correlated (r2 = 0.82) with the relative normalized nifH gene expression in experimental cultivars, validating this molecular assay as a tool to screen SNF potential of dry bean cultivars.


| INTRODUCTION
Symbiotic nitrogen fixation (SNF), a significant component of the global nitrogen (N) cycle, is carried out by a class of prokaryotes (Bacteria and Archaea) known as "diazotrophs" (Burns & Hardy, 1975;Ludden, 2001). About half of the N input in the global ecosystem is contributed by SNF (Vitousek et al., 1997). Although N is abundant in the atmosphere, the dinitrogen (N2) form is relatively inert, thus, is unavailable to most organisms. Only diazotrophs (e.g., Rhizobium and Azotobacter) can fix N2 as free-living organisms or symbiotically in association with higher plants (Howard & Rees, 1996;Phillips, 1980). The percentage of N derived from the atmosphere (%Ndfa) for the dry bean (Phaseolus vulgaris L.) is quite variable depending on complex interactions among genotype, soil conditions, and success of inoculation (Brockwell et al. 1980). Fageria et al. (2014) contributed the low SNF potential of bean cultivars to the following factors and interactions: (i) lack of interactions and specificity between bean cultivar and rhizobia strain, (ii) competition among efficient and inefficient bacterial strains, (iii) suppressions of nodulation by residual soil N, (iv) high demand for phosphorus and photosynthate by rhizobia, and (v) sensitivity of soil moisture stress. Understanding host-symbiont genetic interactions and the resulting genetic regulation of SNF in different cultivars will improve the management of SNF in dry bean production to enhance productivity.
Symbiotic legumes receive N from two sources, soil inorganic N as nitrate (NO 3 − ) or NH 4 + , and NH 3 from SNF. The NO 3 − first must be reduced to NH 3 before it can be assimilated as organic nitrogenous compounds in the plant system (Crawford & Arnst, 1993). Reduction of NO 3 − is a two-step process: NO 3 − is first reduced to nitrite (NO 2 − ) in the cytosol, a reaction catalyzed by the plant enzyme nitrate reductase (encoded by NR gene), and then NO 2 − is reduced to NH 3 with the help of the enzyme nitrite reductase (Lam, Coschigano, Oliveira, Melo-Oliveira, & Coruzzi, 1996). Assimilation of NH 3 into amino acids in stepwise reactions involves glutamine synthetase (GS), glutamate synthase (also known as glutaimine-2-oxoglutarate aminotransferase), and glutamate dehydrogenase present as isoenzymes encoded by distinct genes (Lam et al., 1996;Mengel, Kirkby, Kosegarten, & Appel, 2001). GS (EC 6.3.1.2) is the key enzyme involved in the very first step of NH 3 assimilation, where it catalyzes the condensation of NH 3 and glutamate into glutamine with the help of another enzyme, glutamate synthase (Cullimore & Bennett, 1992;McGrath & Coruzzi, 1991).
Therefore, genes that encode these enzymes involved in Nassimilation in the plant system are the genes that should be considered when studying gene expression to identify N-assimilation activity. About a quarter of the total N 2 fixation in the global ecosystem is contributed by rhizobia-legume symbiosis and the symbiotic compatibility between rhizobia and host legumes varies widely (Wang et al., 2018).
Eukaryotic systems lack the genes to synthesize the nitrogenase enzyme complex that carries out reactions of N 2 fixation (Seefeldt, Dean, & Hoffman, 2017). A class of higher plants in the family Fabaceae can engage in symbiotic relationships with Rhizobia and utilize the fixed N 2 . Rhizobia can colonize legumes by forming nodules and live as N 2 fixing symbionts (Hong et al., 2012;Peter, Young, & Haukka, 1996). During SNF, atmospheric N 2 is reduced to ammonium (NH 4 + ) or ammonia (NH 3 ) through reactions catalyzed by the twocomponent dinitrogenase-dinitrogenase reductase complex that form the nitrogenase enzymes (Ludden, 2001). Nitrogenases are encoded by a set of operons that consists of structural genes (nifH, nifD, and nifK), regulatory genes (nifL and nifA), and a few supplementary genes in diazotrophic bacteria (Q. Cheng, 2008). Nitrogenases have two main subunits: dinitrogenase (molybdenum-iron, Mo-Fe protein), which is the heterotetrameric core encoded by nifD and nifK and dinitrogenase reductase (Fe protein) encoded by nifH (Kim & Rees, 1994). The nifH gene is a universally accepted marker for SNF as it is found to be highly conserved among N 2 fixing organisms in natural environments and thus widely used to study the ecological and evolutionary aspects of N 2 fixing bacteria (Farnelid et al., 2011;Izquierdo & Nüsslein, 2006).
The use of 15 N, a stable isotope of N, is the most reliable and accurate approach to quantify N 2 fixation, and it has been extensively reviewed (Chalk, 1985;Witty, Rennie, & Atkins, 1988). Among several 15 N-isotopic methods, the "isotope dilution technique" has high efficiency in discriminating the N 2 fixed from the atmosphere and the N taken up from the soil, though this technique is most accurate when the "percent N derived from atmosphere" (%Ndfa) exceeds 70% and less suitable when N 2 fixation values are less than 30% (Barrie, 1991;Danso, 1986). A nonfixing reference plant is needed to estimate % Ndfa, and both the fixing and reference plants should uptake N from the soil with the same 15 N isotopic composition (Boddey, Chalk, Victoria, & Matsui, 1984). Moreover, 15 N-enriched fertilizers for media enrichment as well as the analysis of enriched plant tissues in the isotope ratio mass spectroscope is very costly and sensitive to minimal variations.
Interestingly, some studies reported that dry bean cultivars like Aurora, Sangretoro, R-275, Redcloud, and Cargamanto fixed more than 100 kg of N ha −1 (Rennie & Kemp, 1982). Thus, it is evident that there is a large variation in the ability of dry bean genotypes to induce N 2 fixation. In free-living diazotrophic communities, the nifH gene expression was found to be correlated with the amount of N 2 fixed. In soil samples and liquid cultures, the correlation between the amount of inorganic N 2 fixed by free-living Azotobacter vinelandii was found to be significantly higher (Bürgmann, Widmer, Sigler, & Zeyer, 2003); however, symbiotic systems are more complex and still lack enough molecular evidence to establish such correlation.
We hypothesized that correlation between nifH gene expression with host NR and GS gene expression could be related to host N assimilation phenotype, and the trend would possibly correlate with %

| Seed treatment and planting
Nontreated (no fungicide treatment) dry bean seeds were surface sterilized with 70% ethanol for 1 min and rinsed with nonchlorinated sterile water afterward. Seeds were pinched for better germination in the greenhouse. Seeds were treated with a peat-based rhizobia inoculant slurry containing Rhizobium phaseoli following the manufacturer's instructions for the commercial inoculant, N-Dure™ (Verdesian Life Sciences U.S., LLC) and subsequently air dried for 5 min just before sowing. The inoculant was previously tested for its capacity in nodulating all the cultivars under study using germination pouches.
Four seeds were planted in each pot and 1 week after germination, seedlings were thinned to two seedlings per pot, and maintained until sampling. Plants were watered regularly to maintain moisture content at 50% of water holding capacity. homogenized using disposable plastic pestles. Homogenized samples were centrifuged for 7 min for nodule tissues and 4 min for leaf tissues instead of manufacturer-recommended 3 min at the highest speed in a tabletop centrifuge. This extended centrifugation step was required make a pellet of otherwise slimy lysate. After centrifuging, 350 μl of supernatant was carefully transferred to a 2-ml collection tube using a 1,000-ml pipette while being careful to exclude any cellular debris from the pellet; 350 μl of 70% ethanol was added to supernatant and mixed thoroughly. Subsequently,~700 μl of the well-mixed sample was transferred to RNeasy Mini spin column placed in a 2-ml collection tube and centrifuged for 15 s at 8,500 g.

| Fertilization
The flow through was discarded, and~700 μl of buffer RW1 (Qiagen™) was added to the spin column and centrifuged for 15 s at 8,500 g. Thereafter, the flow through was discarded, and~500 μl of buffer RPE (Qiagen™) was added to the spin column and centrifuged for 15 s at 8,500 g. The flow through was discarded, and~500 μl of buffer RPE was added to the spin column and centrifuged for 2 min at 8,500 g. The RNeasy Mini spin column was placed in a new 2-ml collection tube and centrifuged at full speed for 1 min to remove residual RPE buffer. The RNeasy Mini spin column was transferred to a new 1.5-ml collection tube supplied with the kit, and 40 μl of warm (60 C) ultrapure nuclease-free water (Ambion™) was added to the spin column and centrifuged for 1 min at 8,500 g for elution. After the elution step, the eluent was passed through the spin column for a second time and centrifuged for 1 min at 8,500 g. The second elution containing the total RNA was collected in the centrifuge tubes and treated with DNase (Sigma-Aldrich ® ) to avoid genomic DNA contamination. The DNase-treated RNA was stored at −80 C in 1.7-ml nuclease-free collection tubes (Thermo Fisher Scientific™) until further analysis. RNA samples were visualized on 0.8% agarose gels ( Figure 1) stained with gel red (Biotium) to confirm the integrity of samples and quantified before storage using Qubit ® 2.0 Fluorometer (Invitrogen™) with a high-sensitivity RNA detection kit (Thermo Fisher Scientific™).

| cDNA synthesis
Approximately 160 ng of total RNA was used as template to synthesize cDNA using the "GoScript™ Reverse Transcription System" (Promega™) following manufacturers protocol (RevTsc, 2010). Since bacterial (prokaryote) RNA does not contain polyA tails, random hexamer primers were used instead of oligo-dT primers for cDNA synthesis allowing for simultaneous synthesis of both plant and bacterial cDNA from root nodule isolated RNA. To maintain homogeneity in the RT-qPCR reactions, random hexamers were also used for cDNA synthesis from leaf RNA. Synthesized cDNA was diluted 1:5 with ultrapure nuclease-free water (Ambion™) before using as template in qPCR reactions (Solanki, 2017).

| Primer designing and validation
The primer pair reported in a recent nifH gene expression study in dry beans (Akter, Pageni, Lupwayi, & Balasubramanian, 2014) were initially used in our study, but the reported primers did not amplify the specified portion of the nifH gene even after gradient polymerase chain reaction (PCR). Upon sequence analysis, it was determined that mistakenly the 18-s ribosome primers were described as one of the primer pair used to amplify nifH gene transcript. Also, the second pair of nifH primers, redesigned in Akhter et al. (2014) (Bürgmann et al., 2003;Bürgmann et al., 2004).
Thus, we designed five new primers to amplify different portions of the R. phaseoli nifH gene, and the primers producing the most robust amplicons after gradient PCR were selected (Table 2, Figure 3).
F I G U R E 1 Quality of RNA and specificity of primers: (a) Visual representation of RNA extracted from six different root nodule samples on gel electrophoresis. Specificity was checked on the gel electrophoresis after RTPCR with designed primers for and Actin (Actin L-leaf and Actin N -nodule) For plant NR and GS genes, specific qPCR (quantitative PCR) primers were not previously reported in the literature for dry bean.
Thus, we designed specific and robust primers for dry bean NR and GS transcripts (Table 2, Figure 3).
We tested three previously described bacterial reference genes recA, dnaK, and truA for our study and found that dnaK was highly and uniformly expressed across all the samples and timepoint tested. Thus, dnaK was selected as the most consistent and robust gene to normalize the nifH gene expression and was utilized for expression analyses.
For the dry bean NR and GS expression, the actin gene was used for gene expression normalization for the qPCR gene expression analysis.
The use of degenerate primers in qPCR analysis may introduce quantification bias, so degenerate primers were not used in this study (Gaby & Buckley, 2017). Instead, highly efficient and robust gene specific primers were designed for NR, GS, and nifH to reduce the possibility of off-target amplification. Gradient PCRs (Polymerase Chain reaction) were run using GoTaq ® Polymerase (Promega™) to identify annealing temperatures that produced specific amplicons for each primer combination ( Table 2). Amplicons of each primer pair for all the pinto cultivars at two time points were sequenced (GenScript) to confirm the specificity of the amplicons produced ( Figure 2). The amplicons were also separated on 1% agarose gels supplemented with GelRed (Biotium) nucleic acid dye to confirm expected amplicon sizes ( Figure 1).

| Experimental conditions for qPCR
The qPCR was performed on three technical replicates of each of the three biological replicates for root nodules or leaves using Bio-Rad Sso Advanced Universal SYBR ® Green Supermix on a CFX-96 Real-Time PCR detection system (Solanki, Ameen, Richards, & Brueggeman, 2016). A 10-μl qPCR reaction was prepared using 0.5 μl of 500 nM of each gene specific forward and reverse primer ( F I G U R E 2 Representative standards (Actin) used for RT-qPCR: (a) Dilution series (15 × 10 −1 to 15 × 10 −6 picograms per milliliter) used to produce standard curve. (b) Standard curve used to study gene expression and the calculated qPCR efficiency. *RFU, relative fluorescence units; *Cq is the quantification cycle (baseline corrected) at which the amplification curve crosses the threshold value (where the curvature of the amplification curve is maximal) and annealing at primer specific temperature (

| Assay optimization and selection of qPCR reference genes
The primers designed and optimized and for study for both the bacterial and host gene expression analyses were shown to be robust and efficient (Figures 1 and 2). The nifH primer pair standardized for this assay (Table 2)   The "atom % excess 15 N" is the 15 N enrichment above the background (the value is 0.3663%). A nonnodulating dry bean navy mutant line, "R99," was used as the nonfixing reference to quantify SNF (Park & Buttery, 2006). Amount of N 2 fixed was calculated by the following formula: N 2 fixed = Dry matter yield × %Ndfa × N in plant tissue:

| Statistical analysis
Analysis of variance was performed for %Ndfa, total N in plant tissues, plant dry matter, and N 2 fixed using SAS 9.4 (SAS Institute, Cary, NC).
Least significance difference method was used to compare the cultivar means at 95% significance level. Statistical analyses for gene expressions were performed using the CFX manager software statistical tools; t test was performed to compare gene expressions at two different growth stages.

| nifH gene expression in nodule tissue
Normalized nifH gene expression was varied across cultivars with growth stages (Figure 3a). Lariat had a higher relative normalized nifH gene expression level at the V 3 stage, followed by La Paz and Windbreaker. At the R 2 stage, Windbreaker had the highest relative normalized nifH gene expression followed by La Paz and Lariat. The relative normalized nifH gene expression was upregulated 1.50 and 3.10 folds in La Paz and Windbreaker, respectively, but downregulated 1.47 folds in Lariat at the R 2 stage compared with the V 3 stage for each cultivar. Relative normalized nifH gene expression did not change with growth stages for ND-307.

| NR gene expression dynamics in nodule and leaf samples
At both growth stages, Lariat had the highest NR gene expression ( Figure 3b). For the other three cultivars, NR gene expression did not vary with growth stages. The NR gene expression was downregulated in nodule tissues of La Paz (8.2 fold) and Windbreaker (5.8 fold) at the R 2 stage compared with V 3 stage. In inoculated leaf tissue, relative normalized NR gene expression varied among cultivars at the V 3 stage but not at the R 2 stage (Figure 4a). At the V 3 stage, ND-307 had significantly higher NR gene expression than the other three cultivars, whereas NR gene expression of the rest had similar levels (Figure 4a).
Significant variation was found in the relative normalized NR gene expression levels in the inoculated leaves of the cultivars between the V 3 and R 2 growth stages (Figure 4a). The NR gene expression was downregulated 2.9 and 14 folds in La Paz and ND-307, respectively, whereas upregulated in Lariat (4.5 fold) at the R 2 stage as compared with the V 3 stage (Figure 4a).
In the noninoculated leaves, statistically significant relative normalized NR gene expression among cultivars was found at the R 2 stage; the highest NR gene expression was found in Windbreaker,  Relative normalized GS gene expression levels in the inoculated leaves varied significantly among the cultivars at both the V 3 and R 2 growth stages (Figure 6a). ND-307 had the highest relative normalized GS gene expression levels at the V 3 stage, followed by Windbreaker, La Paz, and Lariat; whereas at R 2 stage, Lariat had the highest GS gene expression, followed by ND-307, Windbreaker, and La Paz (Figure 3a).

| GS gene expression dynamics in nodule and leaf samples
The GS gene expression levels in the inoculated leaf tissue of the dry bean cultivars significantly varied between the two growth stages.
The relative normalized GS gene expression was downregulated at the R 2 stage compared with the V 3 stage. The GS gene expression in inoculated leaf tissue was downregulated in La Paz (5.7 fold), Windbreaker (3.5 fold), and ND-307 (5.0 fold) at the R 2 stage compared with the V 3 stage; no statistically significant change in gene expression was found in Lariat between the two growth stages (Figure 3a).
F I G U R E 3 nifH, NR, and GS gene expressions in dry bean root nodules. Relative normalized expressions of (a) nifH, (b) NR, and (c) GS genes in the root nodules of four commercial pinto bean cultivars (La Paz, Lariat, Windbreaker, and ND-307) at two different growth stages (V 3 and R 2 ). nifH gene expression was normalized against the dnaK gene. GS and NR gene expression was normalized against the Actin gene. Values (bars) with overlapped standard error bars are not significantly different (p < .01). Gene expression was normalized against the Actin gene. Mean relative normalized expression values in the bar diagrams with same lowercase letters are not statistically significant (p < .01) Significant variation in relative normalized GS gene expression in the uninoculated control leaves of dry bean cultivars was found at the V 3 growth stage, but no significant variation in gene expression was found among dry bean cultivars at the R 2 stage (Figure 6b). Windbreaker had significantly higher GS gene expression levels in the control leaf tissue than the other three dry bean cultivars at the V 3 stage.
GS gene expression was downregulated in the control leaves of all the cultivars at the R 2 stage compared with V 3 stage (Figure 3b). The GS gene expression in control leaves was downregulated 10.7 folds in Windbreaker, 2.9 folds in Lariat, and 2.4 folds in both Windbreaker and La Paz (Figure 3b).

| %Ndfa, Plant N, and N 2 fixed
Plant N content varied among pinto cultivars, but %Ndfa and amount of N 2 fixed were similar (Table 3). Windbreaker and ND-307 (0.15 g kg −1 ) had significantly higher plant tissue N content compared with In the uninoculated plants, the average N contents were higher or similar compared with the inoculated plants except for Lariat, but mean dry matter was lower compared with inoculated plants due to lack of N sources (Tables 3 and 4).

| Correlation between nifH gene expression and SNF
The quantitative reverse transcription polymerase chain reaction (RT-qPCR) assay was successfully optimized to study the dry bean NR and GS gene expression in association with the symbiotic bacteria, R. phaseoli. To avoid the nonspecificity of degenerate primers, we utilized the publicly available sequence data for R. phaseoli and dry bean to design and validate specific qPCR primers. Previously, different cultivars of Oryza sativa grown under field conditions and without nitrogen fertilizer supplement were reported to show cultivar dependent detectable varied level of nifH gene mRNA in the total root extracts indicating varietal differences in root associated diazotrophic communities (Knauth, Hurek, Brar, & Reinhold-Hurek, 2005 R 2 stage and the amount of N 2 fixed (g N 2 per plant) was observed for all four experimental dry bean cultivars assayed ( Figure 7). Similar findings were reported in earlier studies on N2 fixing diazotrophic cyanobacterial communities but not in a symbiotic system (Vitousek et al., 1997;Warshan et al., 2016;Zehr et al., 2007;Bürgmann et al., 2003;Thaweenut et al., 2011;Turk-Kubo et al. 2012). This outcome strongly supports our hypothesis that cultivars that induce higher nifH gene expression in their rhizobia symbiont would fix higher amounts of atmospheric N 2 than the cultivar showing lower nifH gene expression.

| N assimilation in dry bean plants
Plants absorb the available N from growth media at early growth stages and the reduced supply of N causes the downregulation of Nassimilatory genes. The regulation of NR gene expression by different levels of NO 3 − was reported in earlier literatures (Galangau et al., 1988;Mohr, Neininger, & Seith, 1992). A positive correlation has been shown with supply of NO 3 − and plant response for NR gene expression regulation (de Borne Dorlhac, Vincentz, Chupeau, & Vaucheret, 1994;Galangau et al., 1988;R. Wang, 2000). Studies suggested that plants without any accessible NO 3 − source express very low or undetectable NR mRNA expression (Hoff, Stumman, & Henningsen, 1991;Hoff, Stummann, & Henningsen, 1992). In our gene expression study, plants were intentionally supplied with low amounts of inorganic N to force SNF, which in turn affected NR gene expression. Significant variation was found in the relative normalized NR gene expression levels in the inoculated leaves of the dry bean cultivars between the V 3 and R 2 growth stages ( Figure 4a)

| Relation between N fixing and N assimilatory gene expressions
In general, we found that nifH gene expression was increased among dry bean cultivars in the later R 2 growth stage, particularly in Windbreaker. During the plant-symbiont interaction, it is possible that the bacterial nifH gene expression is modulated in a plant genotype specific manner when other available source of N, NO 3 − are depleted. These data suggest that to meet their physiological N requirement, certain host genotypes are more efficient at inducing SNF activity of their Rhizobia bacteria symbiont. At R 2 stage, plants were also preparing for seed development, which requires the elevated synthesis of amino acids, and thus, more N is required for assimilation. However, at the R 2 stage when SNF is the only F I G U R E 7 Correlation between nifH gene expression and N 2 fixation. Linear regression model between mean amount of N 2 fixed (g N 2 per plant) by four cultivars and corresponding normalized expressions of nifH gene in their root nodules at the R 2 stage; P 1 : La Paz, P 2 : Lariat, P 3 : ND-307, and P 4 : Windbreaker source of N, levels were not adequate to provide the required amount of NH 3 for assimilation into glutamine by the GS enzyme resulting in the suppression of GS gene expression, suggesting that addition of N through other sources is necessary to meet the N requirement at the later stages of growth, that is, late vegetative to early reproductive stage.

| CONCLUSION
A robust RT-qPCR method was developed to study the expression profile of the NR and GS genes in four dry bean cultivars as well as nifH gene expression from the nodule forming symbiont R. phaseoli.
This study utilized molecular analyses to demonstrate that higher N 2 fixing dry bean cultivars also induce higher nifH gene expression with their symbiotic partners. The nifH gene expression was found to be significantly higher at the dry bean R 2 growth stage compared with the V 3 growth stage except for cultivar Lariat. We noticed that at the initial growth stages of dry bean, expression of NR and GS genes were 3 to 7 folds higher than later stages, possibly due to the availability and extractability of N from the growth media. The NR and GS enzyme activities are well covered in literatures, but molecular analysis based on gene expression studies in relation to symbiont bacterial nifH gene expression is not well defined. Our study utilizing molecular assay will help to identify host genotypes that have the capability of inducing nifH gene expression resulting in higher SNF, facilitating the determination of efficient timing of N fertilization. The ability to conduct robust molecular characterization of SNF during plant growth stages will also help fill knowledge gaps concerning the cross talk and signaling occurring during this complex symbiotic interaction.
F I G U R E 8 A conceptual model depicting estimated and predicted expression dynamics of the Rhizobium phaseoli nifH gene and dry bean (Phaseolus vulgaris L.) NR and GS genes at different growth stages (G, germination; V 3 , third trifoliate stage; R 2 , late flowering stage; R 5 , pod-filling stage; M, maturity). The expression levels shown are from two dry bean genotypes in relation to the available nitrogen from the growth media. The R 2 value shows the linear correlation between the relative normalized nifH gene expressions at late flowering stage and the amount of total nitrogen fixed