Phylogenetic diversity analysis reveals Bradyrhizobium yuanmingense and Ensifer aridi as major symbionts of mung bean (Vigna radiata L.) in Pakistan

Abstract

The present study was carried out to evaluate the diversity of rhizobia associated with nodules of mung bean in Pakistan, because this information is necessary for inoculum development. Based on sequence analysis of 16S rRNA gene of thirty-one bacteria, 11 were assigned to genus Bradyrhizobium, 17 to Ensifer, and 3 to Rhizobium. Phylogenetic analyses on the basis of 16S-23S ITS region, atpD, recA, nifH, and nodA of representative strains revealed that B. yuanmingense is the predominant species distributed throughout different mung bean–growing areas. Among the fast-growing rhizobia, Ensifer aridi was predominant in Faisalabad, Layyah, and Rawalpindi, while E. meliloti in Thal desert. Sequence variations and phylogeny of nifH and nodA genes suggested that these genes might have been co-evolved with the housekeeping genes and maintained by vertical gene transfer in rhizobia detected in the present study. Host infectivity assay revealed the successful nodulation of host by rhizobia related to genera Bradyrhizobium, Ensifer and Rhizobium. Among all, Bradyrhizobium and Ensifer spp. inoculation exhibited a significantly higher number of nodules (11–34 nodules plant⁻¹) and nitrogenase activity (nodule ARA 60–110 μmol g⁻¹ h⁻¹). Contrary to the previous studies, our data reveal that B. yuanmingense and E. aridi are predominant species forming effective nodules in mung bean in Pakistan. Furthermore, to the best of our knowledge, this is the first report showing the effective symbiosis of E. aridi, E. meliloti, and Rhizobium pusense with mung bean. The diversity of rhizobia in different habitats revealed in the present study will contribute towards designing site-specific inocula for mung bean.

Supplementary Information

The online version contains supplementary material available at 10.1007/s42770-020-00397-9.

Introduction

Symbiotic interaction among legume and soil bacteria, specifically rhizobia, is of critical importance in the nitrogen cycle and agriculture production. The rhizobia interact with the roots of legume plants and lead to the formation of specialized plant organs, known as nodules, wherein they reside and fix atmospheric nitrogen. The interactions exhibit high levels of specificity through the exchange of signal molecules between rhizobia and legumes. The symbiotic dialogue is initiated through the production and secretion of flavonoid compounds through the roots of the plant. Flavonoids interact with the transcriptional regulators which activate the synthesis of Nod-factors in rhizobia [1]. These Nod-factors activate various signaling pathways in plants for nodule formation. Legume-rhizobia symbiotic specificity depends upon the type of flavonoids and Nod-factors [1]. Besides, nodulation and host specificity are also affected by exo-lipo polysaccharides and proteins secreted by the rhizobia [2, 3].

Mung bean is an important legume crop that provides cereal-based diets to many people in China, India, Indonesia, Pakistan, Philippines, and Thailand and is well-known for its detoxification activities and regulation of gastrointestinal distress [4, 5]. Regarding nodulation, it is promiscuous in nature as it is known to be nodulated by rhizobia related to four different genera, i.e., Bradyrhizobium, Ensifer, Mesorhizobium, and Rhizobium [6–9]. However, previous studies from China, India, Nepal, and Thailand indicated that Bradyrhizobium-related species were predominant in the root nodules of mung bean [7–10].

For many years, 16S rRNA gene is used as a molecular marker in defining the taxonomic position of isolated bacterial strains. Due to the high level of conservation, this molecule can be widely used for defining bacteria at class and genus. However, the resolving power of this molecule at the species level is relatively low especially for the genus Bradyrhizobium [11]. To overcome, sequence analysis of internal transcribed spacer (ITS) regions, symbiotic genes, and housekeeping protein-coding genes is usually carried out to complement the 16S rRNA phylogenetic analysis. The ITS region between the 16S and 23S rRNA genes provides information to elucidate inter- and intraspecific relationships as it shows considerable variation in length and sequence among closely related strains [12]. It was previously used for the characterization of rhizobia associated with various legumes [13]. Symbiotic genes known to be horizontally transferred among rhizobia such as nif and nod genes have also been used for phylogenetic analyses due to variation in evolutionary history from core genes [14, 15]. Housekeeping protein-coding genes, on the other hand, have more sequence divergence and sufficiently conserved to retain genetic information for precisely characterizing the prokaryotes at the intra- and interspecific levels. Currently, most commonly used housekeeping genes for the classification of rhizobia at species level include recombinase A (recA), ATP synthase subunit beta (atpD), glutamine synthetase (glnA), chaperone protein (dnaK) and threonine synthase (thrC), DNA gyrase subunit beta (gyrB), and citrate synthase (gltA) [16].

Mung bean–associated symbionts have been identified from various regions of South Asia through polyphasic studies, but from Pakistan, detailed studies are lacking. Few studies reported the phenotypic and biochemical characterization of mung bean–associated symbionts [17, 18] but not at molecular level. Recently, we have used 16S rRNA–based metagenomics approach to assess the diversity of microbes in the root nodules of mung bean collected from different mung bean–growing areas of Pakistan [19, 20]. Rhizobia related to three genera, i.e., Bradyrhizobium, Ensifer, and Rhizobium, were detected in the root nodules, of which Bradyrhizobium and Ensifer were the predominant in nodules.

The aim of the present study was, therefore, the analysis of the diversity of rhizobia isolated from the root nodules collected from different areas of Pakistan using cultivation-based techniques. Secondly, phylogenetic analyses of rhizobia were based on the 16S rRNA, 16S-23S rRNA ITS region, two symbiotic (nifH and nodA), and two additional conserved housekeeping genes (recA and atpD). The data has validated our previous metagenomics data and will be useful in finding inter- and intraspecies variation among strains from different areas and selecting site-specific strains for inoculum production.

Materials and methods

Sample collection and isolation of root nodule endophytes (putative rhizobia)

Soil and root samples were collected from mung bean plants grown at five sites in Pakistan including Bhakkar (31° 43′ 13.2″ N, 71° 05′ 48.4″ E), Faisalabad (31° 23′ 42.5″ N, 73° 01′ 45.5″ E), Layyah (30° 59′ 01.5″ N, 70° 56′ 27.4″ E), Rawalpindi (33° 22′ 45.0″ N, 73° 30′ 16.3″ E), and Thal desert (31° 44′ 31.1″ N to 32° 12′ 11.9″ N, 71° 45′ 05.6″ E, 72° 01′ 44.3″ E) (Fig. S1). Soil samples were analyzed for physical and chemical soil quality parameter as described [19] (Table 1). From each location, five plants were sampled and 5 nodules were collected from each plant. Nodules were surface-sterilized as described previously [20] and crushed in 1 mL of sterilized water. One hundred microliters of nodule suspension was spread on Yeast Mannitol Agar (YMA) plates supplemented with Congo Red (25 μg mL⁻¹). The inoculated YMA plates were incubated at 32 ± 2 °C for 2 days and at 28 ± 2 °C for 2–8 days for the isolation of fast-growing and slow-growing rhizobia, respectively, as indicated by previous study [20]. Nodule isolates were purified by repeatedly streaking on the YMA medium, and colony and cell morphology of pure cultures was recorded.

Table 1.

Physical and chemical properties of soil samples from mung bean–growing areas

Parameter	Bhakkar	Faisalabad	Layyah	Rawalpindi	Thal desert
Latitudea	31° 43′ 13.2″ N	31° 23′ 42.5″ N	30° 59′ 01.5″ N	33° 22′ 45.0″ N	31° 44′ 31.1″ N to 32° 12′ 11.9″ N
Longitudea	71° 05′ 48.4″ E	73° 01′ 45.5″ E	70° 56′ 27.4″ E	73° 30′ 16.3″ E	71° 45′ 05.6″ E to 72° 01′ 44.3″ E
Rainfall during crop seasonb	123 mm	79 mm	12 mm	182 mm	77 mm
Soil texturec/classificationd	Silty clay loam/Inceptisol	Clay loam/Aridisol	Sandy loam/Entisol	Silty clay loam/Inceptisol	Sandy loam/Entisol
Organic matter (%)c	0.73 ± 0.01	0.597 ± 0.024	0.50 ± 0.01	0.629 ± 0.04	0.35 ± 0.01
pHc	7.89 ± 0.10	8.05 ± 0.13	7.85 ± 0.15	7.840 ± 0.15	8.10 ± 0.08
ECc (dSm−1)	0.28 ± 0.02	0.398 ± 0.04	0.43 ± 0.01	0.32 ± 0.03	0.41 ± 0.04
Total Pc (μg g−1)	1054.3 ± 9.10	1148.7 ± 13.9	894.67 ± 9.30	1249.0 ± 14.20	561.7 ± 12.7
Available Pc (μg g−1)	4.09 ± 0.11	6.873 ± 0.19	3.15 ± 0.19	3.17 ± 0.14	2.91 ± 0.03
Available Kc (μg g−1)	115.0 ± 9.64	189.2 ± 13.25	107.3 ± 6.77	72.0 ± 3.61	65.7 ± 4.63
Available Nc (μg g−1)	0.006 ± 0.0004	0.00856 ± 0.001	0.0052 ± 0.0002	0.0066 ± 0.001	0.0041 ± 0.0002

The standard analysis procedures were used for the estimation of soil organic matter (Loss-on-Ignition method) [21], available nitrogen (Kjeldahl method) [22], total and available phosphorus (Bray’s method) [23], and potassium (Knudsen method) [24]

Data represent the mean of three independent replicates; ± standard deviation

^aGoogle Earth

^bPakistan Metrological Department

^cData from this study

^dFAO (Food and Agriculture Organization of the United Nations) 2017, Soil Fertility Atlas of Pakistan: The Punjab Province

Authentication of putative rhizobia and assessment of N₂ fixation

The nodule isolates were evaluated for their ability to form nodules on the host plant. Host inoculation experiment was conducted in the net house during mung bean–growing season under natural light. For surface sterilization, seeds of mung bean variety NM-2011 were immersed for 10 min in ethanol (70%, v/v) and then dipped into sodium hypochlorite (4%, w/v) for 4 min followed by washing 6 times with sterilized water. Twenty seeds were sown (2 seeds per sowing point) in plastic containers (40 L × 30 W × 22 H) containing10-kg sterilized sand. Putative rhizobia purified from root nodules were cultured in 50 mL of YMA medium for 3–4 days. The cells were pelleted, washed, and resuspended in 0.85% saline (100 mL) and then 1 mL aliquot (approximately 1 × 10⁹ cells) applied to each of the germinated seedlings, while non-inoculated control received 1 mL of sterilized water. Plants were grown during mung bean–growing season; inoculated plants were irrigated with nitrogen-free Hoagland solution, while the non-inoculated plants received Hoagland with nitrogen. The plants were uprooted 28 days post-inoculation (28 dpi) to observe nodule formation and different growth parameters. Nitrogenase activity was estimated through acetylene reduction assay (ARA) which is the most commonly used method of indirect quantification of N₂ fixation rate [25]. ARA is based on reduction of acetylene to ethylene by nitrogenase enzyme, instead of reducing N₂ to NH₃ [26]. The nodulated roots were incubated at room temperature in 10 mL air-tight glass vials to which acetylene (10% v/v) was injected. After 24 h, 200 μL of the gas sample was analyzed for ethylene production by gas chromatography (Thermoquest, Trace G.C, Model K, Rodono Milan, Italy) using a Porapak Q column with hydrogen flame ionization detector. Samples of root nodule and plant shoots were dried for 2 days at 60 °C and weighed to record the data of nodule and plant dry weight.

Identification of rhizobia by 16S rRNA gene sequence analysis

Rhizobia were cultured in TY medium at 28 ± 2 °C for 2–4 days, and total genomic DNA was extracted using the CTAB method [27]. PCR amplification of 16S rRNA gene from genomic DNAs was carried out using fD1/rD1 primers (Table 2). The PCR products were purified using QIAquick spin Columns (Qiagen, Germany, Hilden) and sequenced from Macrogen Inc., Seoul, South Korea. Forward and reversed sequences of 16S rRNA gene were aligned using Clustal X [28], and the Basic Local Alignment Search Tool (BLAST) [29] on the EMBL/GenBank databases was used to find out the similar sequences. 16S rRNA gene from isolates and reference sequences retrieved from GenBank were aligned using Clustal W software [30]. Maximum likelihood–based phylogenetic trees were constructed using MEGA 7 software [31]. The bootstrap values (≥ 50%) were shown in the phylogenetic tree.

Table 2.

Primers and PCR conditions used in the present study

Primer	Sequence (5′–3′)	Target gene/bp amplified	PCR conditions	References
fD1	AGAGTTTGATCCTGGCTCAG	16S rRNA (~ 1500)	2 min 95 °C, 30 × (1 min 93 °C, 45 s 55 °C, 2 min 72 °C) and 5 min 72 °C	[32]
rD1	AAGGAGGTGATCCAGCC
ITS-For2	TACACACCGCCCGTCACACC	16S-23S ITS region (~ 1100–1450)	2 min 95 °C, 30 × (1 min 93 °C, 45 s 55 °C, 2 min 72 °C) and 5 min 72 °C	[33]
ITS-Rev2	TGGTCCGCGTTCGCTCGCC
atpD255F	GCTSGGCCGCATCMTSAACGTC	atpD (~ 550)	3 min 95 °C, 30 × (1 min 93.5 °C, 40s 55 °C, 1 min 72 °C) and 5 min 72 °C	[34]
atpD782R	GCCGACACTTCMGAACCNGCCTG
recA41F	TTCGGCAAGGGMTCGRTSATG	recA (~ 500)	3 min 95 °C, 30 × (1 min 93.5 °C, 1 min 58 °C, 1 min 72 °C) and 5 min 72 °C	[34]
recA640R	ACATSACRCCGATCTTCATGC
POL F	TGCGAYCCSAARGCBGACTC	nifH (~ 370)	2 min 95 °C, 30 × (1 min 94 °C, 1 min 55 °C, 2 min 72 °C) and 5 min 72 °C	[35]
POL R	ATSGCCATCATYTCRCCGGA
nodA-1	TGCRGTGGAARNTRNNCTGGGAAA	nodA (~ 666)	2 min 93 °C, 35 X (45 s 93 °C, 1 min 49 °C, 2 min 72 °C) and 5 min 72 °C	[36]
nodA-2	GGNCCGTCRTCRAAWGTCARGTA

Sequence and phylogenetic analyses of 16S-23S ITS, recA, atpD, nifH, and nodA genes

Amplification of internal transcribed spacer region, symbiotic, and housekeeping genes was performed from the representative strains related to different genera of rhizobia from each site. 16S-23S ITS region, atpD, recA, nifH, and nodA region were amplified using primers and PCR conditions described in Table 2. Phylogenetic analyses of gene sequences were performed as described above for 16S rRNA gene, and sequences were submitted to EMBL-EBI and accession numbers were obtained (Fig. 1 and Table S2).

Fig. 1.

Phylogenetic tree of mung bean rhizobia based on 16S rRNA sequences (~ 1300 bp) showing the relationship of rhizobia isolated in this study with already published strains. The tree was constructed using ML method, and bootstrap values (1000 replicates) higher than 50% are shown at the nodes. Bacterial isolates obtained in the present study are shown in bold

Detection of nucleotide polymorphism and statistical analyses

Nucleotide polymorphism among strains related to genera Bradyrhizobium, Ensifer, and Rhizobium was calculated using DNASP v5 software. Sequences related to each gene fragment were aligned using the Clustal X and manually trimmed to keep the equal length. These homologous aligned sequences were used as input for the estimation of nucleotide variation. The sequence types (STs) representing sequence type for each gene and nucleotide diversity (Pi) were detected through the nucleotide polymorphism. Nucleotide diversity at synonymous and non-synonymous sites was estimated by assigning the coding frame of the gene.

The effect of rhizobial inoculation on nodulation and growth parameter of mung bean plants was determined through analysis of variance (ANOVA) at a level of significance (p = 0.05) using Statistix 8.1 software.

Results

Soil analyses indicated that soils sampled from Bhakkar, Faisalabad, and Rawalpindi were silty clay loam with more organic matter, total P, and available N contents as compared to the sandy loam soil from Layyah and Thal desert (Table 1). The soil of Thal desert showed maximum pH, i.e., 8.10, followed by Faisalabad (8.05).

A total of 31 putative rhizobia were isolated from surface-sterilized root nodules in the present study. These isolates included both slow-growing and fast-growing bacteria. Fifteen rhizobial isolates were obtained from the root nodules collected from Faisalabad, 5 from Bhakkar, 4 from Layyah, 5 from Rawalpindi, and 2 isolates from the Thal desert.

Authentication of rhizobia

Host infectivity assay was performed to assess the nodulation ability of isolates. The seedlings were grown in sterilized sand contained in rectangular plastic containers. All the isolates formed nodules on the host plant (Supplementary Table S1). The maximum number of nodules (33.3 nodules plant⁻¹) were formed in mung bean plants inoculated with Ensifer sp. Vr33 followed by Ensifer sp. Vr25 (28.5 nodules plant⁻¹) at 21 days post-inoculation (dpi). Among bradyrhizobia, maximum nodulation was observed in plants inoculated with Bradyrhizobium sp. Vr50 (24.4 nodules plant⁻¹) followed by Bradyrhizobium sp. Vr61 (23.6 nodules plant⁻¹). Rhizobium strains formed very less number of nodules (2–5 nodules plant⁻¹). The maximum plant dry weight (0.46 g) was observed in plants inoculated with Ensifer sp. Vr33 followed by Bradyrhizobium sp. Vr50 (0.40 g). Nitrogen fixation in the root nodules was estimated through the acetylene reduction assay (ARA). The results indicated that all the rhizobia-induced effective nodulation and maximum activity (109.3 μmol⁻¹ g⁻¹ h⁻¹) were observed in nodules produced by the slow-growing isolate Vr50. Among fast-growing isolates, Vr33 showed maximum ARA activity (102.7 μmol⁻¹ g⁻¹ h⁻¹) followed by Vr34 (99.0 μmol⁻¹ g⁻¹ h⁻¹). The maximum number of nodules and the nitrogenase activity was observed in plants inoculated with the isolates Vr33 and Vr50.

Sequence and phylogenetic analysis of 16S rRNA gene

Sequence analysis of 16S rRNA gene sequence indicated that 17 rhizobia were related to genus Ensifer, 11 were assigned to genus Bradyrhizobium, while 3 strains belonged to genus Rhizobium. None of the isolate purified from root nodule belonged to genus Mesorhizobium.

All the Ensifer strains obtained from Faisalabad, Layyah, and Rawalpindi showed identical 16S rRNA gene sequence and clustered in the first clade with Ensifer aridi LEM 451 and Ensifer kostiense LMG 19225^T in phylogenetic analysis (Fig. 1), while two Ensifer strains Vr77 and Vr78 from the Thal region clustered with Ensifer meliloti IAM 12611^T and formed the second clade.

I6S rRNA gene sequences from all the bradyrhizobia were identical and formed single cluster with Bradyrhizobium yuanmingense B071^T (clade 4). Two Rhizobium strains Vr69 and Vr76 purified from root nodules collected from Rawalpindi clustered with Rhizobium pusense NRCPB10^T while one isolate Vr65 from Layyah showed different gene sequence and clustered with Rhizobium sp. UCCB 145 (clade 3).

To further clarify the taxonomic position of these strains at the species level and find the variations at intra- and interspecific levels, 14 representative strains were selected from different genera for further analysis.

Phylogeny of rhizobia based on 16S-23S rRNA ITS region

The 16S-23S rRNA ITS region of 1100 to 1500 bp was amplified from 14 representative strains (Supplementary Table S2) and sequenced. The ITS region amplified from bradyrhizobia has a sequence length of 1100 to 1160 bp while the ITS region of Ensifer- and Rhizobium-related strains have a sequence length of 1300 to 1420 bp. The alignment of the ITS region indicated the deletion of base at various sites in Bradyrhizobium strains. tRNA gene present in the ITS region was scanned through tRNAscan-SE [37]. Two types of tRNA genes for Ile and Ala were present in the 16S-23S rRNA ITS regions amplified from all the representative rhizobia strains. Five Bradyrhizobium strains had sequence variation in the ITS region and clustered with B. yuanmingense B071^T in the phylogenetic tree (Fig. 2). Out of 6 rhizobia related to genus Ensifer, ITS sequence from five strains, i.e., Vr25, Vr28, Vr33, Vr66, and Vr70, from Faisalabad, Layyah, and Rawalpindi have 100% homology and clustered with E. aridi LEM 451. Ensifer sp. Vr77 from Thal region has 21.8% sequence variation from other Ensifer isolates. All the three Rhizobium-related strains have sequence variation in their ITS region. Two Rhizobium strains Vr66 and Vr76 clustered with Rhizobium pusense strain CFEB5875, while Rhizobium sp. Vr65 showed the highest homology with Rhizobium sp. RBG74.

Fig. 2.

Phylogenetic tree of selected rhizobia in this study based on ITS sequences (~ 1100 bp) showing the relationship of rhizobia isolated in this study with already published strains. The tree was constructed using ML method, and bootstrap values (1000 replicates) higher than 50% are shown at the nodes. Bacterial isolates obtained from this study are shown in bold

Analyses of protein-coding housekeeping genes atpD and recA

The atpD gene fragment of approximately 550 bp and recA gene of 500 bp were amplified from representative strains. Ensifer strains Vr25, Vr28, Vr33, Vr66, and Vr70 have identical atpD gene sequence and show maximum similarity with E. aridi LEM451 while Vr77 with E. meliloti IAM 12611^T (Supplementary Table S2). All the Rhizobium strains showed variation in their gene sequence; however, two strains Vr69 and Vr76 showed 99–100% sequence similarity with R. pusense strain NRCPB10^T. Rhizobium sp. Vr65 has 11–12% sequence variation from other two isolates and showed 98.3% homology with Rhizobium sp. GMF8. Bradyrhizobium-related strains Vr50, Vr57, Vr63, and Vr73 showed identical gene sequence, while Vr61 has 1.25% sequence variation from other Bradyrhizobium strains.

Based on recA gene sequence analysis, Ensifer strains Vr25, Vr28, Vr33, Vr66, and Vr70 showed maximum sequence similarity with E. aridi LEM451 while Vr77 with E. meliloti IAM 12611^T (Supplementary Table S2). Similarly, Vr69 and Vr76 showed maximum homology with R. pusense NRCPB10^T, while Vr65 showed maximum sequence homology with Rhizobium sp. NCSU 2642. Interestingly, recA gene amplified from Bradyrhizobium strains showed variations in the gene sequences and divided these strains into three subgroups.

Maximum likelihood phylogenetic tree was constructed using housekeeping genes (recA and atpD) concatenation. Based on the two gene-concatenated analysis, strains Vr25, Vr28, Vr33, Vr66, and Vr70 clustered together with E. aridi LEM451 while Vr77 clustered with E. meliloti IAM 12611^T (Fig. 3). Rhizobium strains Vr69 and Vr76 clustered together with R. pusense NRCPB10^T and Vr65 formed cluster with Rhizobium sp. NCSU 2642. Bradyrhizobium strains Vr50 and Vr57 clustered with B. yuanmingense CCBAU 10071^T, Vr61 Bradyrhizobium sp. JNVU VA7 and Vr63 and Vr73 were placed with B. yuanmingense CCBAU 33112 in phylogenetic tree.

Fig. 3.

Concatenated tree using atpD and recA gene sequences showing the relationship of rhizobia isolated in this study with already published strains. The tree was constructed using ML method, and bootstrap values (1000 replicates) higher than 50% are shown at the nodes. Bacterial isolates obtained from this study are shown in bold

Phylogeny of rhizobia based on nifH and nodA gene sequences

Phylogenetic analyses of nifH gene (370 bp) amplified from 14 rhizobial strains revealed that bradyrhizobia were clustered into three groups, Ensifer strains into two groups, and Rhizobium strains into three groups. Bradyrhizobium sp. Vr50 originating from Faisalabad had nifH gene sequence variation from other strains and clustered with B. yuanmingense CCBAU 10071^T in the phylogenetic tree (Fig. 4). The isolates Vr57, Vr63, and Vr73 sampled from Bhakkar, Layyah, and Rawalpindi have identical nifH sequences and showed homology with B. yuanmingense M7. nifH sequences amplified from Vr61 isolated from Bhakkar was different from all other bradyrhizobia and showed homology with Bradyrhizobium sp. JNVU VA7. The nifH gene sequences amplified from five Ensifer isolates Vr25, Vr28, Vr33, Vr66, and Vr70 were identical and clustered with E. aridi JNVU TP6 in phylogenetic analysis, while sequence from Ensifer sp. Vr77 showed homology with E. meliloti CCBAU 83493. All the isolates related to genus Rhizobium have variation in their nifH gene sequence and formed three clusters in the phylogenetic tree. Ensifer sp. Vr65 showed the highest homology with Rhizobium sp. ORS571, Vr69 was clustered with R. tropici CFN ESH25, and Vr76 placed with R. pusense AmV15 in phylogenetic tree.

Fig. 4.

Phylogenetic tree of selected rhizobia in this study based on nifH gene sequences showing the relationship of rhizobia isolated in this study with already published strains. The tree was constructed using ML method, and bootstrap values (1000 replicates) higher than 50% are shown at the nodes. Bacterial isolates obtained from this study are shown in bold

The nodA gene sequences were amplified from 12 strains while the sequences were not amplified from Vr61 and Vr65 using the primers given in Table 2. Phylogenetic analysis showed the clustering of Bradyrhizobium strains in two groups. Bradyrhizobium strains Vr57, V63, and Vr73 shared one nodA gene sequence, which differed by Bradyrhizobium sp. Vr50 clustered with B. yuanmingense CCBAU 10071^T (Fig. 5). Similarly, Ensifer strains were clustered in two group, of which Vr25, Vr28, Vr33, Vr66, and Vr70 have identical nodA sequence and showed homology with E. aridi JNVU TP6, while Ensifer sp. Vr77 clustered with E. meliloti CCNWSX0020. The nodA gene sequence obtained from two Rhizobium strains Vr69 and Vr76 formed a cluster with Rhizobium pusense YIC4072 in phylogenetic tree.

Fig. 5.

Phylogenetic tree of selected rhizobia in this study based on nodA gene sequences showing the relationship of rhizobia isolated in this study with already published strains. The tree was constructed using ML method, and bootstrap value (1000 replicates) higher than 50% are shown at the nodes. Bacterial isolates obtained from this study are shown in bold

Nucleotide diversity inferred from different genes

Nucleotide polymorphisms among different genera of all isolates for 16S rRNA gene and 14 representative strains for ITS, atpD, recA, and nifH and 12 strains for nodA were calculated (Table 3). Two sequence types (STs) were detected in genus Ensifer on the basis of nucleotide diversity in the targeted genes, i.e., 16S rRNA (0.0144), ITS (0.0398), atpD (0.0294), recA (0.03713), nifH (0.0464), and nodA (0.1064). These sequence types represent two species of this genus, i.e., E. aridi and E. meliloti. No nucleotide polymorphism was detected in any of the genes amplified from five E. aridi strains originating from three different sites, i.e., Faisalabad, Layyah, and Rawalpindi. Rhizobium strains showed high level of nucleotide diversity and represent three STs based on ITS (0.1678), atpD (0.0794), and nifH (0.1824) and two STs based on nodA (0.0058). Bradyrhizobium strains represent single sequence type based on the 16S rRNA gene while showing nucleotide polymorphism in all other genes and represent three STs based on ITS, recA, and nifH genes, while two STs based on atpD gene and nodA.

Table 3.

Nucleotide polymorphism detected among different genes amplified from rhizobia

Genus (no. of strains)	Length	S	Eta	STs	Pi	Pi(s)	Pi(n)
16S rRNA
Ensifer (17)	1323	19	19	2	0.0144	0.0225	0.0120
Rhizobium (3)	1076	16	16	2	0.0149	0.01211	0.0159
Bradyrhizobium (11)	1307	0	0	1	0.00	0.00	0.00
ITS
Ensifer (6)	1282	153	153	2	0.0398	-	-
Rhizobium (3)	1295	322	330	3	0.1678	-	-
Bradyrhizobium (5)	1077	2	2	3	0.0009	-	-
atpD
Ensifer (6)	443	39	39	2	0.0294	0.1004	0.0040
Rhizobium (3)	466	55	56	3	0.0794	0.2355	0.0254
Bradyrhizobium (5)	479	6	6	2	0.0050	0.0206	0.0000
recA
Ensifer (6)	393	44	44	2	0.0373	0.1269	0.0052
Rhizobium (3)	393	45	45	2	0.0763	0.2667	0.0069
Bradyrhizobium (5)	393	14	14	3	0.0182	0.0681	0.0000
nifH
Ensifer (6)	309	43	43	2	0.0464	0.1601	0.0079
Rhizobium (3)	318	87	91	3	0.1824	0.5084	0.0825
Bradyrhizobium (5)	284	7	7	3	0.0099	0.0293	0.0037
nodA
Ensifer (6)	522	167	167	2	0.1064	0.2195	0.0662
Rhizobium (2)	519	3	3	2	0.0058	0.0217	0.0000
Bradyrhizobium (4)	554	55	55	2	0.0496	0.1182	0.0257

S, total number of polymorphic sites; Eta, number of mutations; STs, sequence types; Pi, nucleotide diversity; Pi(s), Pi at synonymous sites; Pi(n), Pi at non-synonymous sites

Discussion

Present study demonstrates the diversity of mung bean nodule rhizobia from different regions of Pakistan at species level using cultivation-based sequence analyses of multiple genes. The study further reveals few novel rhizobial symbionts found in mung bean. Symbiotic bacteria especially rhizobia fix atmospheric nitrogen within root nodules of legumes, and this symbiosis is among the most beneficial process in the agriculture system. Exploration of different biogeographic regions for the isolation and identification of rhizobia can lead to the selection of elite microsymbionts which could be further used for site-specific inoculum production.

Previous 16S rRNA–based metagenomics studies revealed that two genera Bradyrhizobium and Ensifer are co-dominant (> 60% of total rhizobia) in mung bean nodules from Faisalabad [20] while Bradyrhizobium is the single dominant genus in Bhakkar, Layyah, and Rawalpindi regions (82–94% of total rhizobia) and Ensifer is the dominant genus (99.9%) at Thal desert. A very small fraction of sequences were related to the genus Rhizobium [19]. Furthermore, the occurrence of Ensifer in nodules was found to be positively correlated with soil pH. The results of the present study validate the promiscuity of the mung bean plant as revealed previously by the cultivation-independent analysis [19, 20].

During this study, a total of 31 putative rhizobia were isolated from different sites of Pakistan. Fifteen rhizobial strains were isolated from the root nodules collected from the local site in Faisalabad (soil pH = 8.05) in several attempts at different time intervals. Out of these, 13 strains belonged to genus Ensifer and two strains were related to genus Bradyrhizobium. All the isolated strains from Bhakkar (soil pH = 7.89) belonged to genus Bradyrhizobium, while the strains related to three genera, i.e., Bradyrhizobium (50–60%), Ensifer (20–25%), and Rhizobium (25–40%), were isolated from Layyah (soil pH = 7.85) and Rawalpindi (soil pH = 7.84). Isolates originated from Thal desert (pH = 8.10) were related to genus Ensifer. The present study indicated the occurrence of Ensifer in root nodules where the soil pH was relatively higher which validated the finding of previous study based on culture-independent approach [19]. Soil pH has been found to influence rhizobial diversity in legumes [38–40], due to the influence of soil pH on the bioavailability of mineral nutrients for growth. Soybean nodulation by B. japonicum and B. elkanii is favored when pH is below 7.96, while B. yuanmingense when pH is high pH > 8.19 [41]. Similarly, both Bradyrhizobium and Ensifer were symbionts of legumes in Indian Thar desert, but Ensifer was found to be preferred symbiont-nodulating legumes in alkaline soil [42], while Bradyrhizobium nodulate legumes in acidic soil [43]. All the rhizobial strains effectively nodulated the host plant. Among all, Bradyrhizobium and Ensifer formed significantly higher number of nodules (11–34 nodules plant⁻¹) and exhibited maximum ARA activity (60–110 μmol g⁻¹ h⁻¹). Maximum plant dry weight was observed in plants inoculated with Ensifer sp. Vr33 and Bradyrhizobium sp. Vr50, while most of other plants produced less dry weight as compared to the control irrigated with Hoagland with available N. It might be due to higher efficiency of abovementioned strains and vice versa and/or early harvesting of plants. Rhizobium-inoculated plants showed less number of nodules (2–5 nodules plant⁻¹) and ARA activity (13–25 μmol g⁻¹ h⁻¹) which imply them as less effective symbionts of mung bean. Various studies have reported the ability of Bradyrhizobium, Ensifer, and Rhizobium to nodulate mung bean but bradyrhizobia were predominantly isolated from nodules and assumed to be the major symbionts of mung bean [7–9], which might be due to the biasness of the isolation protocols and/or growth conditions. Our current and previous studies [20] show that the optimum temperature for fast-growing Ensifer strains is 32 °C, while for slow-growing Bradyrhizobium isolates, it is 28 °C. Few Ensifer colonies appear at 28 °C but they are dominated by bradrhizobia and vice versa.

The dataset shows that 16S rRNA gene sequences from the same genus show relatively less variation among the strains related to different genera, e.g., all the slow-growing rhizobia from different sites showed identical 16S rRNA gene sequences and showed high sequence similarity (≥ 99%) with reference strains for 25 Bradyrhizobium species. This shows that a single Bradyrhizobium species is prevalent throughout mung bean–growing regions of Pakistan. Similarly, all the Ensifer strains from Faisalabad, Layyah, and Rawalpindi showed identical 16S rRNA gene sequence and sequence similarity (≥ 99%) with reference strains for 12 Ensifer species. It is well-reported that 16S rRNA gene is widely used for determining the phylogenetic relationship and taxonomic position at the genus level but has limited resolving power to differentiate closely related species of bacteria [44]. In such cases, sequence analyses of the 16S-23S ITS region and housekeeping protein genes such as recA, atpD, dnaK, glnII are frequently used to resolve the species status of bacteria, especially rhizobia [16, 45–49]. The sequence analyses of the ITS region and atpD and recA genes of rhizobia further validated the species status as well as detected the polymorphism among the strains.

Sequence analysis of the ITS region and housekeeping genes revealed that these slow-growing rhizobia belong to B. yuanmingense. Nucleotide polymorphism observed in the ITS sequences divided these strains into three sequence types (STs) (Table 3). The ITS region of Bradyrhizobium strains has been widely used as a useful marker to differentiate closely related species as it exhibited high sequence variation [48, 49]. The differentiation of two atpD STs and three recA STs among the isolates related to B. yuanmingense evidenced the existence of different strains in the Bradyrhizobium population obtained in this study. The higher variability of housekeeping genes in rhizobia as compared to 16S rRNA for better delineation of species and differentiation of subspecies and strains is well-reported [16, 50–52]. Furthermore, bradyrhizobia are divided into three and two STs based upon nifH and nodA gene sequences, respectively, and clustered with different strains of B. yuanmingense. The nodA gene was not amplified form Bradyrhizobium sp. Vr61 and Rhizobium sp. Vr65 using primer pairs nodA-1/nodA-2. It is previously documented that nodA gene is under direct positive selection due to involvement in symbiosis process; thus, it is less conserved than housekeeping genes [53]. Sequence analyses collectively revealed that mung bean nodules in different areas of Pakistan host multiple strains of B. yuanmingense, although multiple species of Bradyrhizobium, i.e., B. japonicum, B. elkanii, B. liaoningense, and B. yuanmingense, were reported to nodulate mung bean in China and Nepal [7–9]. B. yuanmingense was identified as a major slow-growing symbiont of mung bean in India [10], while some novel strains of Bradyrhizobium originating from acidic soils of the sub-Himalayan regions of India were found to be compatible with mung bean [54]. So, the cultivation-based study corroborated the findings of previous reports [7–9] and the bradyrhizobial community associated with mung bean may vary in different regions. However, nucleotide polymorphism was observed among these bradyrhizobia on the basis of ITS, atpD, recA, nifH, and nodA genes. So, sequence analysis of multiple genes clarified the species as well as intraspecies polymorphism.

Likewise, two Rhizobium strains Vr69 and Vr76 from Rawalpindi and one isolate Vr65 originating from Layyah showed 98.6–99.7% 16S rRNA gene sequence similarity with R. pusense previously obtained from chickpea rhizosphere in India [55]. The definition of these isolates as R. pusense enlarged the host range of this species, while the division of three STs in analysis of ITS, atpD, and nifH and two STs in nodA gene sequences evidenced the existence of other Rhizobium species as nodule endophytes of mung bean. Symbiotic genes are commonly transferred horizontally as these are carried on mobile genetic elements [56]. However, in present study, the topologies of nifH and nodA genes were congruent to the phylogeny of housekeeping genes suggesting that these strains might have evolved through vertical gene transfer. The coherence of symbiotic and housekeeping gene phylogenies has been reported in several studies [53, 57].

All the fast-growing rhizobia related to genus Ensifer originated from three sites (Faisalabad, Layyah, and Rawalpindi) exhibit 100% sequence identity of 16S rRNA gene and clustered with E. aridi LEM 451, whereas, Ensifer strains from Thal desert show 1.51% sequence variation and belonged to E. meliloti IAM 12611^T. Similar to 16S rRNA gene sequence, representative Ensifer strains (Vr25, Vr28, Vr33, Vr67, and Vr70) from Faisalabad, Layyah, and Rawalpindi contain identical ITS, atpD, recA, nifH, and nodA gene sequences and belonged to E. aridi. These findings suggest that a single strain of Ensifer might be distributed among the mung bean at all sites/regions. E. aridi is newly described species, which has been isolated in Asian, African, and American deserts [58], from root nodules of Tephrosia purpurea in the Thar desert in India [59]; Acacia gummifera in Merzouga desert, Morocco [60]; and Phaseolus filiformis in the Mexican desert [58]. More recently, the strains of E. aridi were isolated from nodules of Retama monosperma in Eastern Morocco [61]. Among Ensifer strains, only E. fredii has been reported as a symbiont of mung bean from China [8, 9]. Rhizobial strain from Thal desert was related to E. meliloti which is dominantly isolated from root nodules of Acacia tortilis, Argyrolobium uniflorum, Cicer arietinum, Genista saharae, Hedysarum carnosum, Hippocrepis bicontorta, Medicago spp., Lotus spp., Ononis natrix, Phaseolus vulgaris, and Retama raetam [62]. Ensifer strains have also been identified as major symbionts of legume in Thar desert of India which are closely related to E. saheli, E. arboris, E. kostiensis, Ensifer terangae [63], and E. aridi [43]. The findings of the present study suggest that E. aridi and E. meliloti are among the major symbionts of mung bean in Pakistan. The occurrence of both of these symbionts has not been reported in mung bean earlier.

In conclusion, the diversity analyses of symbionts of mung bean through the cultivation-based method showed predominance of Bradyrhizobium and Ensifer. Sequence analyses of multiple genes indicated that B. yuanmingense and E. aridi are the two major symbionts of mung bean at different sites (Bhakkar, Faisalabad, Layyah, and Rawalpindi), while E. meliloti seems specific to Thal desert of Pakistan. This study has contributed to the global database on the distribution of rhizobial species and reported the presence of Bradyrhizobium yuanmingense, Ensifer aridi, E. meliloti, and R. pusense, and a novel Rhizobium sp. as nodulating microsymbionts of mung bean. Furthermore, intraspecies polymorphism detected through the sequence analyses of housekeeping genes validated that these genes could be used for the differentiation of isolates at the strain level. Moreover, two representative strains from B. yuanmingense and E. aridi could be used as universal inoculum as they were the dominant groups in almost all the sampling sites. However, Ensifer sp. Vr77 related to Ensifer meliloti could be used as inoculum in Thal desert. These strains could be further evaluated under field conditions to improve the productivity of plants and fertility of soil.

Funding

The study was financially supported by HEC through Ph.D. fellowship to Sughra Hakim.

Compliance with ethical standards

Conflict of interest

The authors declare that they have no conflict of interest.