Introduction

Biological nitrogen fixation (BNF) is an environmentally friendly and easily accessible source of N for agricultural crops’  N nutrition [111]. BNF occurs when a group of prokaryotes known as diazotrophs reduce atmospheric N2 to ammonia via a nitrogenase enzyme complex [111, 174]. The legume-rhizobia symbiosis is the most important BNF process in agriculture in terms of quantities of N fixed [174]. It takes place in root nodules harbouring the rhizobia. In cases where the populations of indigenous soil rhizobia are insufficient, commercial rhizobial strains (inoculants) are introduced into the soil [19]. However, the establishment of these introduced rhizobial strains in the soil, their infectivity and symbiotic effectiveness is affected by a number of biotic and abiotic factors [2, 4, 20, 33, 41, 90, 174]. Such factors include, but are not limited to, antibiotics which are introduced into the soil by antibiotic-producing organisms and by faeces and urine [50, 101] because under concentrated animal feeding operations worldwide, antibiotics are commonly used to treat animal disease and promote animal growth. Ben et al. [14] reported that in China over 8000 t of antibiotics are used as feed additives each year. In Turkey, the usage of veterinary antibiotics in feeding operations has been reported to be 33 % of total pharmaceutical consumption [76]. About 16 million kg of antibiotics are used annually even in the USA, and 70 % are estimated to be used for non-therapeutic purposes [134]. So, the majority of used antibiotics are excreted in faeces and urine which persist and accumulate in soils after repeated manure application [74]. The performance of Rhizobium inoculation in agricultural fields and its efficiency are critical due to competition occurring between the introduced rhizobia and the indigenous rhizobia present at high densities. Maximum nodule occupancy by the inoculant strain is achieved using Rhizobium strains having more host cultivar specificity, inoculum size and better competitiveness. Competition for nodulation is usually measured by comparing the ability of introduced Rhizobium strains to form nodules on the chosen host. Antibiotic resistance has been frequently used in distinguishing the introduced inoculant strain from indigenous rhizobia and monitoring their survival and occupancy of legume nodules [22, 143]. Antibiotic resistance is an ancient and naturally occurring phenomenon widespread in the environment [168]. The rhizosphere contains a mixture of metabolically active microbial populations that compete in this environment in relation to size, diversity and biochemical activity [90]. Production of antibiotics by some soil harbouring microorganisms mainly bacteria and fungi has been largely documented [10, 81, 92, 117]. Such organisms include Streptomyces like Streptomyces coelicolor, Micromonospora purpurea and Streptomyces griseus which produce actinorhodin, gentamycin and streptomycin, respectively [87, 156]. Spore-producing, filamentous, gram-positive actinomycetes, specifically Streptomyces species, produce the majority of known antibiotics [81]. The reasons for production of antibiotics by some soil harbouring microbes are not yet clear. Various reports have suggested a number of hypotheses for the production of antibiotics.

Hypotheses for Antibiotic Production by Some Soil Microbes

Competition

This hypothesis suggests that microbial organisms produce antibiotics for competition purposes [10, 92]. The hypothesis is supported by three lines of evidence [117]. First, various studies have indicated that production of antibiotics from bacteria inhibits the growth of pathogenic fungi of plants [71, 151]. Secondly, a survey on tropical rhizobia reported a prevalence of isolates which were highly tolerant to naturally produced antibiotics [126, 130]. This implies that rhizobial strains susceptible to the naturally produced antibiotics succumbed to the antibiotics. Lastly, when populations of bacteria and antibiotic-producing microbes were mixed, the densities of bacterial populations sensitive to antibiotics were significantly reduced [154]. It is assumed that production of antibiotics gives the producing organism an advantage over non-producing microbial populations [10, 92, 167].

Coordination of Developmental Process in the Organism

Dworkin et al. [48] suggested that some organisms produce antibiotics as a measure of coordinating their developmental stages. For instance, the antibiotics may maintain the dormancy of the spores produced by the antibiotic-producing organisms. This hypothesis is supported by the observation of antibiotic production and spore formation in the same organism that occurred at almost the same time [48].

Symbiotic Associations

This hypothesis was suggested by Kumbhar and Watve [87]. An example of the symbiosis between Streptomyces and beewolf digger wasp was given in support of the hypothesis [85]. The Streptomyces produces antibiotics which protect the beewolf digger wasp’s cocoons from fungal infection hence increasing the larvae’s chances of survival in the soil [75]. In return, the Streptomyces acquires nutrients from the beewolf digger wasp.

Homeostasis

It is hypothesised that the antibiotics act as signalling molecules for homeostasis regulation in the organism [92]. It is believed that at low concentrations, antibiotics may positively impact the susceptible bacterial species [87]. For instance, Bader et al. [11] showed that sub-lethal concentrations of antibiotics boost the growth of bacteria on solid media. Enhancement of motility and/or mutation of the bacterial cell by sub-lethal doses of the antibiotics was also reported [92]. Sub-lethal doses of antibiotics have also been reported to aid in exopolysaccharide synthesis. High concentrations of antibiotics, on the other hand, inhibit the growth of susceptible bacterial species [38].

Predation is the most recent suggested hypothesis. Since most antibiotic-producing organisms (e.g. Streptomyces) are predators, it is hypothesised that they produce antibiotics to aid the predation process. The hypothesis is supported by the fact that antibiotics are produced at the end of the growth period (when nutrients have been depleted) or under nutrient stress. The antibiotics lyse live cells of susceptible bacterial species, thereby providing nutrients for the organism [87].

Despite the conflicting hypotheses as to why some soil organisms produce antibiotics [87], the fact still remains that some microorganisms in the soil produce antibiotics. The produced antibiotics then inhibit growth and/or kill susceptible populations of the soil bacteria [10, 92]. Different classes of antibiotics exhibit different modes of action on the bacteria.

Mode of Action of Antibiotics

Interference with Cell Wall Synthesis

This can be through interfering with enzymes required for the production of the peptidoglycan layer [101]. The antibiotic agent acts by forming covalent complexes with enzymes that generate the mature peptidoglycan molecule [99]. This mode of action is common in β-lactam antibiotics like penicillin and vancomycin. As a result, the bacterial cell wall is weakened resulting into bacterial death due to osmotic pressure [6, 83].

Inhibition of Protein Synthesis

The antibiotic agent binds to the 16S and 23S rRNA hence inhibiting protein translation [2]. Tetracycline, actinomycin, aminoglycoside, oxazolidinone and macrolide-lincosamide-streptogramin (MLS) like kanamycin and streptomycin use this mode of action [50]. Streptomycin, for example, interferes with the 30S subunit of the bacteria’s ribosomal RNA thereby affecting translation [164]. Chloramphenicol also inhibits protein synthesis in both gram-positive and gram-negative bacteria. This is due to its affinity for the peptidyl transferase of the 50S ribosomal subunit of 70S ribosomes [136]. The binding of chloramphenicol to peptidyl transferase enzyme prevents elongation of the peptide chain [138]. The target sites and function of some commonly used antibiotics with rhizobial cells are indicated in Table 1.

Table 1 Some commonly used antibiotics with legume symbionts, their target sites and function

Inhibition of Nucleic Acid Synthesis

This results from inhibition of either DNA or RNA synthesis. For example, fluoroquinolones inhibit DNA synthesis and also cause lethal double-strand DNA breaks during replication [46]. Bleomycin also uses this mode of action.

Inhibition of Metabolic Pathway

The antibiotic may block the pathway for folic acid synthesis in the bacteria [119]. For example, trimethoprim antibiotic acts by inhibiting folic acid metabolism in the bacteria [36].

Disruption of Bacterial Membrane Structure

The antibiotic may increase the rhizobial membrane permeability causing leakage of bacterial cell content [24, 148]. Lipopeptides use this mode of action [2]. For instance, daptomycin is reported to insert its lipid tail into the bacterial cell membrane leading to depolarisation of the bacterial membrane and eventual death of the bacterium [24].

Although antibiotic resistance in bacteria is a threat in the health sector [2, 150, 166], it is a desirable trait in both indigenous and introduced (commercial inoculants) rhizobial populations [5, 9, 174]. Resistance to antibiotics increases the rhizobium’s chances of survival in the rhizosphere. Rhizobial strains should be resistant to concentrations of antibiotics that inhibit the growth of other soil bacteria and they should be able to retain their infectivity and symbiotic effectiveness. While the majority of rhizobial strains in the soil are susceptible to antibiotics, others have developed resistance in response to naturally produced antibiotics [16, 164, 167]. The resistance may be developed towards one or multiple antibiotic classes [5]. Unlike resistance to some soil stress, which may be dependent on the origin of the rhizobial strains, antibiotic resistance is not [167].

Rhizobial genome organisation has been widely studied. Usually, these bacteria have one chromosome and several plasmids and/or megaplasmids that may represent 50 % of the genome. Symbiotic genes are typically located in megaplasmids, known as pSyms. Non-symbiotic plasmids may encode locally adaptive traits that confer phenotypic advantages, such as heavy metal or antibiotic resistance genes. In evolutionary terms, plasmid-encoded genes have the advantage of being more easily exchangeable within a certain population than genes located in the chromosome [3].

Therefore, increase in resistance of rhizobia to antibiotics is mainly due to mobile genes on plasmids that can readily spread through rhizobial populations [86]. It is also important to note that the more rhizobial populations are exposed to antibiotic agents, the more they develop resistance due to selective pressure.

Types/Forms of Rhizobial Resistance to Antibiotics

There are two forms of antibiotic resistance in bacteria, namely intrinsic/innate and acquired resistance [2, 150].

Intrinsic/innate resistance occurs naturally in a bacterial species [99]. It results from genes which naturally occur in the bacterium’s chromosome [2]. Such genes include ampC β-lactamase found in gram-negative bacteria. Cole and Elkan [30] reported that R. japonicum (now Bradyrhizobium japonicum) carries extra chromosomal antibiotic resistance genes. Intrinsic resistance may also result from the multiple drug resistance efflux systems within the bacterial cell. Gram-negative bacteria (including rhizobia) are intrinsically resistant to the activity of macrolides because macrolides are slow to traverse the bacterial cell wall [99]. It is also because the macrolides are simultaneously extruded by the activity of efflux pumps [82]. Intrinsic resistance could also result from a lack of the antibiotic target site [99]. Bacterial species with intrinsic resistance could be resistant to a single or multiple classes of antibiotics [150].

Acquired resistance occurs when susceptible rhizobial strains become resistant to the antibiotic agent they were susceptible to [99, 150]. The new resistant strains proliferate and spread under selective pressure of the antibiotic agent [150]. Unlike intrinsic resistance, acquired resistance only occurs to some strains of a species [99]. Acquired resistance is brought in bacteria through mutation in genes targeted by antibiotics and the transfer of resistance determinants born on plasmids, bacteriophages, transposons and other mobile genetic material [2, 66]. The acquired genes may provide resistance to a whole class of antimicrobials [99]. These genes are frequently associated with large transferable extra chromosomal DNA elements called plasmids, on which may be other mobile DNA elements such as transposons and integrons [99]. Gene transfer may be intra- or inter-genera [31]. This implies that a rhizobial strain may acquire or transfer genes for antibiotic resistance from/to members of the same or different genera [30]. This is referred to as horizontal gene transfer [156]. Kuykendall [88], reported the transfer of plasmid genes (R factor) to and between genetically marked strains of R. japonicum. The plasmid gene transfer was from P. aeruginosa and E. coli to R. japonicum. However, the B. japonicum requires more time for mating and for phenotypic expression compared to other bacteria. In cases of mutation, the antibiotic target gene within the bacterium may be altered or expelled leaving no site to which the antibiotic agent can bind [150]. Antibiotic-resistant (AR) phenotypes come about due to chromosomal DNA mutations [147]. The mutations alter existing bacterial proteins, through processes like transformation which can create mosaic proteins [97]. AR phenotypes can also be brought about by transfer and acquisition of new genetic material between bacteria of the same or different species or genera [97, 110, 147]. Alternatively, mutations may cause the bacterium to produce enzymes which inactivate the antibiotic agent before it reaches the target site [150]. Mutations may also lead to the altering of the bacterium’s outer membrane protein channel which the antimicrobial agents require entering the bacterium [150]. This makes it impossible for the antimicrobial agent to access the target sites within the bacterial cell. Lastly, the bacterium can up-regulate efflux pumps which expel the antibiotic agent from the bacterial cell [66, 101]. It should be noted that some rhizobial species/strains lose their symbiotic effectiveness and infectivity once they mutate to AR forms [175, 176]. Other rhizobial species/strains, however, maintain their infectivity and symbiotic effectiveness [63]. Acquired resistance resulting from chromosomal mutations and selection is termed vertical evolution [150]. Resistance resulting from the acquisition of genetic material from other organisms is referred to as horizontal evolution [150].

Processes Through Which Rhizobia Acquire Antibiotic Resistance Genes

Since soil is a reservoir of antibiotic-resistant genes (ARGs) [51], transfer of antibiotic resistance determinant genes among rhizobia is effected by three main parasexual processes, namely conjugation [68, 69], transformation [122] and transduction [84]. Conjugation and transduction often transfer particular parts of the chromosomes that are associated with the integration sites of episomal elements. Transformation is a generalised mechanism that can in principle mediate the transfer of any part of the chromosome [97].

Conjugation involves the transfer of DNA from one rhizobial cell to another rhizobial cell through direct contact [8]. During conjugation, the antibiotic resistance determinant genes are carried on conjugative plasmids and conjugative transposons from the donor bacteria to the recipient bacteria [2, 66, 150]. More than a single plasmid can occur in a host Rhizobium and their multiplication does not depend on the host. Transposons, on the other hand, are a mobile genetic material which may exist on plasmids, integrate into other transposons or in the host’s chromosome. They can also aid in the transfer of endogenous plasmids from one bacterium to another [2]. Beringer [16] showed the ability of R. leguminosarum to donate and receive R factors belonging to the compatibility class P (which are associated with plasmids carrying antibiotic resistance genes) in mixed culture with other bacteria. The same author reported the transfer of R factors RP4, RK2 and R6886 carrying resistance to kanamycin/neomycin, tetracycline and carbenicillin from E. coli to R. leguminosarum. Datta et al. [37] also reported the transfer of R factors (Rp4) carrying high levels of resistance to carbenicillin from E. coli to R. trifolii and R. meliloti by conjugation. Transfer of antibiotic resistance genes from R. japonicum to Agrobacterium tumefaciens through conjugation was also reported [29].

Transformation is a process through which a rhizobium acquires free/naked DNA from the environment and incorporates it into its genome [89, 150] by homologous recombination or transposition [156]. A wide range of DNA fragments can be transferred between bacteria by transformation [97]. The free DNA is normally released to the environment as a result of cell lysis of the donor bacterium. Balassa [12] reported acquisition of resistance to streptomycin by three rhizobial species (R. japonicum, R. meliloti and R. lupini) through transformation. However, the frequency of transformation varied between the three rhizobial species. Acquisition of penicillin resistance genes by R. phaseolus and R. leguminosarum strains through transformation was also reported [53]. There are three requirements for bacterial transformation, namely opportunity (presence of exogenous DNA in the immediate environment of the host Rhizobium), DNA uptake (presence of DNA uptake mechanisms in the host rhizobium) and incorporation of the acquired DNA into the host rhizobium’s chromosome [94, 97]. Since transformation can allow transfer of any part of the chromosome, it can be equated to recombination in sexual organisms with the one exception that a single part of the chromosome is transferred per a single process of transformation [77].

Transduction is a rare process of gene transfer. It occurs when the antibiotic resistance determinant gene is transferred from the donor bacterium to the recipient Rhizobium by a bacterial phage [2, 150]. Instead of the phage DNA, bacterial DNA is packaged into the phage head and injected into the recipient Rhizobium [156]. There are two types of transduction, namely generalised and specialised transduction. In generalised transduction, any element of DNA is packaged into the phage head [156]. In specialised transduction, only DNA adjacent to the phage insertion site is packaged. Kowalski [84], Sik and Orosz [141] reported gene transfer in rhizobia by transduction. Rhizobium phage 16-3 transferred genetic material to R. meliloti strain 41 [141].

Mechanisms of Rhizobial Resistance to Antibiotic

Mechanisms of resistance to antibiotics in rhizobia can be described as active or passive based on how the mechanisms are put in place. Passive resistance is not directly put in place to resist a particular antibiotic or class of antibiotics. Instead, it results from general adaptive processes, for example, the presence of specific barriers on the outer membrane of gram-negative bacteria. Active resistance involves three main mechanisms, namely modification of antibiotic target site, active efflux of the antimicrobial agent from the bacterial cell and enzymatic modification/degradation of the antimicrobial agent. All active mechanisms require changes in their genetic programming in response to the presence of antibiotic agents [165]. Other resistance mechanisms include the acquisition of alternative metabolic pathways, over production of the target enzyme and permeability changes in the bacterial cell wall.

Modification of the Antibiotics Target Site

The commonest targets of antibiotic agents include enzymes involved in major life processes like cell wall synthesis, protein synthesis and nucleic acid synthesis [101, 150]. Such sites include DNA gyrase, topoisomerase IV, dihydrofolate reductase and penicillin-binding protein (PBP) sites [45, 99, 146]. Rhizobia may modify antibiotic target sites through mutation, for example, mutation of the ribosomal RNA [101, 165]. The mutation may be a point mutation or mutation of the entire chromosome [2]. Gram-negative bacteria (including rhizobia) develop resistance to fluoroquinolones and coumarin by a mutation in the DNA gyrase and topoisomerase IV. Some antibiotics like rifamycins arrest transcription by interacting with RpoB, MLS classes of antibiotic target ribosome to inhibit protein translation. Some bacteria acquire resistance to penicillin through mutation in the PBP.

Some bacteria modify the antibiotic target site by expression of a variant of the target site. The variant may have similar or close to similar functions as the antibiotic target site but with less or no affinity to the antibiotic agent [97, 156]. Some bacteria acquire penicillin resistance through the acquisition of entirely new PBP which have reduced affinity for penicillin or do not bind the penicillin at all [156].

The other modification could be reprogramming the biosynthetic pathway which is the case in the resistance to glycopeptide antibiotics [165].

Enzymatic Modification

In this case, the bacteria synthesise enzymes which selectively target the antibiotic agent [165]. The bacterium acquires genes which encode enzymes that destroy or break down the antibiotic agent to a new form before it reaches the target site. The new form is normally non-toxic to the bacterial cell [66, 150]. Such genes are normally introduced into the bacterial cell from the exterior. Rhizobia acquire resistance to β-lactams like penicillin by the production of one or more β-lactamases that inactivate the antibiotic agent [115]. The β-lactamases inactivate the antibiotics by cleaving their chemical bonds by hydrolysis [156, 166]. Five genes have been reported with regard to bacterial resistance to aminoglycosides (e.g. streptomycin and kanamycin), namely armA [54] npmA, rmtA, rmtB, rmtC and rmtD [31, 42]. Other enzymes modify the antibiotic to a form which impairs the target site binding [165]. The modification, however, requires co- substrates like ATP, acetyl-CoA, glutathione and NAD+ for activity [165]. The best examples of such enzymes are the aminoglycoside acetyltransferases and chloramphenicol acetyltransferases (CATs) which confer resistance in some bacteria to aminoglycosides and chloramphenicol, respectively [109, 138, 165]. CATs inactivate most antibiotics in the chloramphenicol group except florfenicol which has structural modifications. CATs are coded by two genes—classical catA determinants and novel catB variants [156]. These two genes differ significantly in their structure. There are also redox enzymes which inactivate the antibiotic agent by the redox reaction [165]. For example, the redox enzyme Tet X oxidises tetracycline [114, 171]. Lyases follow non-hydrolytic or non-oxidative routes to cleavage off carbon–carbon, carbon–oxygen, carbon–nitrogen and carbon–sulphur bonds [108]. The commonest example is the Vgb lyase which is responsible for type B-streptogramin resistance.

Development of Efflux Pumps

Efflux pumps are membrane-associated proteins [165]. The efflux pumps actively extrude antibiotic agents from the bacterial cell before it reaches the target site [66, 150]. This lowers the concentration of the antibiotic agent in the bacterial cell [99]. Because, it is an active process, it implies that energy in the form of ATP is required. The antibiotic agent is excluded from the bacterial system without being altered or degraded. This mechanism may lead to multiple antibiotics resistance. This is because a single efflux pump may extrude multiple antibiotic agents. It should be noted that the efflux pumps not only extrude antibiotic agents from the bacteria’s system but also other chemically and structurally unrelated compounds. There are five families of efflux pumps: the major facilitator super family (MFS), the adenosine triphosphate (ATP)-binding cassette (ABC) super family, the resistance-nodulation-cell division (RND) sub family, the small multi drug resistance (SMR) family and the multidrug and toxic compound extrusion (MATE) family [98, 113, 117, 132]. Some chloramphenicol-specific efflux proteins like CmlA, expression of the resistance determinant is mediated through an inducible mechanism of translational attenuation. In most of the gram-negative bacteria (rhizobia), intrinsic antibiotic resistance is attributed to the expression of the RND efflux system (s) that allows them to transport drugs from the cytoplasm and across the inner and outer membranes of the cell envelope [2]. The outer membrane of gram-negative cell envelope is a barrier to both hydrophobic and hydrophilic compounds. These organisms have evolved porin proteins (OmpF in E. coli) that function as non-specific entry and exit centres for antibiotics [2]. The tetracycline efflux proteins contain 12 (TetA-E) putative transmembrane-spanning segments in gram-negative bacteria. Expression of proteins is controlled by a transcription repressor such as TetR. The antibiotics inactivate the repressor and allow expression of the tetracycline efflux systems [2]. In R. etli, Rhizobium multiresistance genes (rmrA and rmrB) against phytoalexin and salicylic acids were identified [59]. The protein encoded by gene rmrA belongs to membrane fusion protein (MFP), is similar in sequence, length, hydropathy plots and proteins encoded by genes adjacent to multiresistance proteins EmrA and VceA of Escherichia coli and Vibrio cholera, respectively, while RmrB protein encoded by rmrB gene belongs to the drug resistance protein family. An R. etli rmrA mutant was complemented with the antibiotic resistance genes emrAB from E. coli [59].

Alteration/Changes in the Permeability of the Cell Wall

Bacterial strains acquire resistance to antibiotics by changing their cell wall structure. This may be through changes in cell wall porins that limit movement of the antibiotic agent to the target site. The bacteria may also acquire genes for a metabolic pathway which produces altered bacterial cell walls. The altered cell walls may not contain binding sites for the antibiotic agent [150]. As a result, the antibiotic agent is excluded from the bacterial cell [97]. This mechanism of resistance may lead to multiple antibiotic resistance because more than one type of antibiotic agent may use the same cell wall porin [99].

Benefits of Antibiotic Use in Agriculture

Scientists have noted the potential benefits associated with using antibiotics to know the resistance or susceptibility level of rhizobium. The antibiotic-resistant Rhizobium makes itself competitive in soil environment to occupy high number of nodules in legumes [13, 56]. Large differences in degree of tolerance to antibiotics among fast- and slow-growing rhizobia have been reported [30, 52, 60, 112, 116]. Since the first exploitation of natural rhizobial strain-to-strain variations in intrinsic resistance to antibiotics (IAR) for identification and differentiation of nodule isolates from common bean [17] and chickpea [3], IAR has been used extensively in ecological studies to identify inoculant strains and to determine heterogeneity in natural populations [49]. Some of the intrinsic antibiotic resistance properties of Rhizobium and Bradyrhizobium sps. are shown in Table 2. Belachew [13] reported that field pea variety ‘Wayitu’ from Ethiopia showed the best combination with antibiotic resistant of strain EAL 302 for nodule occupancy. This method has proven practical, rapid and reliable with a discriminating ability dependent on the number of antibiotics and the concentrations used. IAR has been used either as a primary criterion or as a complement to other methods to describe diversity in nodule isolates from alfalfa [73, 140], pea [21, 153], clover [58, 64], Phaseolus sp. [7], chickpea [3, 55, 78], soybean [1, 44, 103, 106, 107, 123], cowpea [144, 167] and various tropical legumes [35, 100, 149]. High doses of antibiotics were used to get the mutant of bradyrhizobia to show resistant to soil rhizobiophage [5]. The authors had found that strain MPSR033 Smr Vr antibiotic and phage-resistant mutant of soybean nodulating Bradyrhizobium was highly competitive for nodule occupancy. Singh et al. [145] reported that soybean nodulating isolate ALSR12, a mutant of gentamicin and also phage-resistant, showed ex-planta nitrogenase activity comparable to the parent strain.

Table 2 Antibiotic resistance and susceptibility of different Rhizobium and Bradyrhizobium species

Numerous reports have indicated that agricultural practices, such as direct application of antibiotics or animal manure, provide positive selective pressure for antibiotic-resistant bacteria, often resulting in increased number and resistance level of these bacteria and ARGs in agricultural soils, and thereby expanding the level of native resistance in soil [80, 121]. In addition, a considerable amount of resistant bacteria added into the soil through manure amendment would lead to the spread of resistance to soil bacteria [57, 67]. The result of Anand et al. [5] showed that inoculation of soybean with an antibiotic and phage-resistant mutant of bradyrhizobia that showed the high ability for nitrogen fixation was effective in increasing soybean production in Indian soil.

Conclusion

Antibiotic-resistant genes (ARGs) found in bacterial chromosome and plasmid are increasing in number due to rapid accumulation of antibiotic in the environment [39, 166]. Antibiotic resistance has become one of the most important sustained driving forces for antibiotic discovery. The soil has been regarded as a rich source of ARGs, emanating from natural and anthropogenic processes. Although a high frequency of antibiotic resistance was found in cultured and uncultured soil bacteria which represent a reservoir of new ARGs, use of metagenomic approach [51] can help in the preparation of single or multiple AR mutants of Rhizobium against these new antibiotic-producing microorganisms. Some rhizobial strains exhibit multiple antibiotic resistances [32, 167] but some possess no detectable antibiotic markers [32, 139]. In that situation, antibiotic markers could be introduced either by isolating spontaneous mutants or by transpositions.