Lupin, a Unique Legume That Is Nodulated by Multiple Microsymbionts: The Role of Horizontal Gene Transfer

Lupin is a high-protein legume crop that grows in a wide range of edaphoclimatic conditions where other crops are not viable. Its unique seed nutrient profile can promote health benefits, and it has been proposed as a phytoremediation plant. Most rhizobia nodulating Lupinus species belong to the genus Bradyrhizobium, comprising strains that are phylogenetically related to B. cytisi, B. hipponenese, B. rifense, B. iriomotense/B. stylosanthis, B. diazoefficiens, B. japonicum, B. canariense/B. lupini, and B. retamae/B. valentinum. Lupins are also nodulated by fast-growing bacteria within the genera Microvirga, Ochrobactrum, Devosia, Phyllobacterium, Agrobacterium, Rhizobium, and Neorhizobium. Phylogenetic analyses of the nod and nif genes, involved in microbial colonization and symbiotic nitrogen fixation, respectively, suggest that fast-growing lupin-nodulating bacteria have acquired their symbiotic genes from rhizobial genera other than Bradyrhizobium. Horizontal transfer represents a key mechanism allowing lupin to form symbioses with bacteria that were previously considered as non-symbiotic or unable to nodulate lupin, which might favor lupin’s adaptation to specific habitats. The characterization of yet-unstudied Lupinus species, including microsymbiont whole genome analyses, will most likely expand and modify the current lupin microsymbiont taxonomy, and provide additional knowledge that might help to further increase lupin’s adaptability to marginal soils and climates.


Introduction
Fabaceae or Leguminoseae (nom. cons.) constitute a large plant family, widespread all over the world, that includes more than 19,700 species [1,2]. Many legume species, including all the agronomically important legume crops, establish symbiotic associations with soil bacteria that are able to fix atmospheric nitrogen (N 2 ) in specific symbiotic organs called nodules. The legume-nodulating bacteria, known as rhizobia, are gram-negative bacteria that have the ability to induce the formation of a de novo organ in the roots and/or stems of leguminous plants [3]. They are capable of fixing atmospheric N 2 owing to the microaerobic environment that is created inside the nodule. Most rhizobia belong to the alpha subgroup of the phylum Pseudomonadota (formerly Proteobacteria). The class α-Proteobacteria comprises most of the N 2 -fixing genera that nodulate legumes, while some symbiotic genera belong to the class β-Proteobacteria. The great diversity of legume species is reflected in their wide variety of nodulating microsymbionts.
Ribosomal ribonucleic acid (rRNA) is the primary component of ribosomes, and is essential to all cells. rRNA is a ribozyme that facilitates protein synthesis in the ribosomes [4]. The 16S rRNA gene is present in all bacteria, and phylogenetic analyses based on these gene sequences are used to determine genetic relationships among organisms. Including conserved and variable regions, the 16S rRNA gene has a considerable size (around 1500 bp), the transfer of symbiotic genes to bacteria that are well-adapted to local soil conditions allows these microorganisms to become rhizobial symbionts of originally incompatible legumes [29]. The horizontal transfer of symbiotic genes most often leads to phylogenies that differ from those of the core genome of the receiving bacteria [21,24].
There is abundant evidence that horizontal gene transfer occurs commonly in nature, involving all kinds of transposable elements and all classes of living organisms [30,31]. Even phylogenetically distant species may experience the transference of transposable elements to/from their genomes through transfer vectors such as viruses, bacteria, or other parasites [32]. There are many examples in the literature of how horizontal gene transfer has driven significant evolutionary changes, or helped organisms to adapt to abiotic stress [33]. There are also some examples of horizontal gene transfer helping plants to adapt to biotic stress: for example, when symbiotic genes are acquired by bacteria with PGPR traits [34]. It is well known that biofilms facilitate horizontal gene transfer between bacteria and that soil bacteria are often organized in biofilms on root, litter, and soil particles [35]. Therefore, biofilms may play an important role in the movement of transposable elements between soil organisms.
Initially identified within a species, biological variants, commonly named biovars, have been described in diverse bacterial species. A biovar represents a specific group of bacterial strains which are distinguishable from other strains in the same species on the basis of biochemical or physiological characteristics. Biovars were described for the first time in Rhizobium leguminosarum by Jordan [36] in a taxonomical revision of rhizobial species. In rhizobia, biovars and symbiotic variants, or symbiovars, are used to differentiate distinct subgroups, usually within a genus, comprising species that nodulate the same legume [37]. Legume host range specificity is encoded in rhizobial genetic elements that can be disseminated by horizontal transfer; thus, symbiovars can be shared by rhizobial species belonging to different genera [9,37].
Until the present century, lupin was assumed to be a microsymbiont-specific legume. However, owing to the evolution of molecular taxonomy, nowadays, lupin can be considered as a promiscuous legume that can be nodulated by different Bradyrhizobium lineages [38][39][40], and by bacteria within the genera Phyllobacterium [41,42], Ochrobactrum [27], Rhizobium, Neorhizobium, Agrobacterium [43], Microvirga [42,44], and Devosia [12]. The aim of this review was to collect the existing information on the diversity and taxonomy of the rhizobial bacteria associated with Lupinus spp. nodules. Phylogenetic trees, including the published 16S rRNA, nodC, and nifH gene sequences of lupin microsymbionts, will be analyzed. The incongruences between the classifications derived from the 16S rRNA gene and the symbiotic gene sequences analyses, and the symbiovars that can be defined thereafter, will be discussed. The importance of lupin, a unique legume with specific characteristics, as summarized below, substantiates the significance of the present review. Moreover, the fact that different lupin-nodulating bacteria might be able to modulate or even determine some legume agronomic traits, such as its adaptability to different soils, its tolerance to abiotic stress, or the nutrient profile of its grain adds relevance to the current knowledge on lupin microsymbionts, and to the discussion on the role that horizontal gene transfer appears to play.

Lupin, a Unique Legume in So Many Ways
The legume genus Lupinus (Papilionideae: Genisteae) comprises 200-500 species of perennial and annual shrubs and herbs with an amphi-Atlantic distribution [1]. Lupinus species are distributed in Europe; the Near East; North and East Africa; and North, Central, and South America. Less than 13 species are native of the Old World, mainly surrounding the Mediterranean basin, while American species are much more numerous [45,46]. Lupins have been grown since antiquity and represent an important pulse crop due to the highly effective N 2 -fixing symbiosis they establish with rhizobia, as well as their adaptation to different climates and soils.
Lupins are mainly employed as a source of protein in both human and animal nutrition. The high protein content of lupin seeds (up to 44% dry weight) is similar to that of soybean, and lupin has been proposed as a viable alternative to soybean cultivation in Europe [47]. Lupin seeds possess a high nutritional value and unique phytochemical profile and biological activities [48,49]. Among the lupin species, L. albus (white lupin), L. luteus (yellow lupin), L. angustifolius (blue lupin), and L. mutabilis (tarwi) have gained agricultural market importance and have become part of the modern agriculture and food systems for high-protein production [47]. Lupinus species have also been studied as medicinal plants [50] and used in traditional medicine to treat various diseases, including urinary tract infections, heart conditions, or skin disorders [51]. Due to the presence of essential fatty acids, specific proteins, and amino acids, as well as minerals, alkaloids, and dietary fiber in lupin seeds, component fractionation allows for extracts to be obtained that display promising anticarcinogenic, antidiabetic, antihypertensive, anti-inflammatory, antimicrobial, and antioxidant activities [51]. The phenolic compounds and flavonoid constituents in lupin seed extracts have important antioxidant activity [51,52]. Lupin protein hydrolysates present anti-inflammatory properties [53]. The antifungal and antibacterial activities have been demonstrated in different studies that report an inhibitory effect toward various bacterial and fungal pathogens [54,55].
Lupin plants have also been shown to have great potential for the recovery of degraded soils and in heavy metal phytoremediation [56][57][58][59]. Lupinus albus appears to be an especially well-suited species to withstanding heavy metal stress, and white lupin plants inoculated with a mercury-tolerant Bradyrhizobium strain have been reported to present increased mercury tolerance [59]. Additionally, its adaptive mechanisms make white lupin a choice crop for acidic soils affected by aluminum toxicity [60]. Nitrogen and phosphorus are, together with potassium, the two principal nutrients needed by plants, and are essential to ensure crop yield. In phosphate-deficient soils, some lupin species, including white lupin, are able to modify their root architecture to form cluster roots, which are bottlebrush-like structures consisting of hundreds of short rootlets [61]. Cluster roots display enhanced synthesis and secretion of acid phosphatases, organic acids, flavonoids, and proton efflux. Such modifications lead to the mobilization of the soil's unavailable phosphorus. Therefore, lupin arises as a choice crop for nutrient-poor soils, as it reduces the need for both N and P fertilizers [61]. Lupin's ability to adapt to acid, nutrient-poor, or moderately polluted soils, and to arid and cold climates, enables lupin cultivation in affected areas, where other more demanding crops, such as soybeans, are not economically viable [62]. Identifying lupin cultivars tolerant to a variety of abiotic stresses may allow for the expansion of lupin cultivation to a broader range of edaphoclimatic conditions [47].
Lupin nodules are also unique. According to the host legume and independently of the microsymbiont species, nodules can be determinate or indeterminate [21,63]. Determinate nodules present a transient meristem, while indeterminate nodules maintain meristematic activity throughout their lifespan. All genera examined in the Fabaceae subfamilies Mimosoideae and Caesalpinioideae, and most tribes within the subfamily Papilionoideae, form indeterminate nodules, while the papilionoid species within the tribes Desmodieae, Dalbergieae, Psoraleeae, Phaseoleae, and some genera in the Loteae form determinate nodules [63]. The lupinoid nodule is a subtype of indeterminate nodule, first identified in lupin [64], and to our knowledge has only been described in another legume genus, Listia (Papilionideae: Crotalarieae) [65]. The lupinoid nodule is characterized by the presence of lateral meristems that allow the nodule to encircle the root. Another characteristic of the lupinoid nodule is that the infected cells remain mitotically active and the symbiosomes, much like other cell organelles, are equally distributed between the daughter cells [66,67]. The infection process that leads to lupin nodule organogenesis is also uncommon, as rhizobia infection does not occur through an infection thread, as happens in most legumes. Rhizobia enter the root intercellularly and specifically infect a cell beneath a root hair [66]. Remarkably, in the work in which the infection process was first described, and despite the fact that lupin was thought to be a microsymbiont selective legume exclusively nodulated by Bradyrhizobium sp. (Lupinus), a Mesorhizobium loti strain was used that was able to elicit nodules, but unable to form effective nitrogen-fixing nodules [66].
Lupinus mariae-josephi, the last Lupinus species discovered in the Old World, is an endemic species, growing in alkaline soils in Eastern Spain [81]. Remarkably, this species, unlike other lupins, thrives in high-pH lime soils [82]. The phylogenetic analysis of 16S rRNA gene sequences of the rhizobia isolated from L. mariae-josephi nodules evidenced a new species, B. valentinum, belonging to a distinct evolutionary lineage that also includes B. jicamae [39,83], and which is phylogenetically distant from any other lineage of lupin microsymbionts isolated in the Iberian Peninsula. Considering that most New World lupin species microsymbionts have not yet been characterized, the number of new Bradyrhizobium species able to nodulate lupins might substantially increase in the future. The Bradyrhizobium strains isolated in Europe from Lupinus spp. also prevail in the soils of South Africa and Western Australia as a result of their introduction with lupin seeds [74]. Housekeeping gene trees reveal a relatively high level of diversity among Old World lupin-nodulating bradyrhizobia, but most isolates are placed in the B. japonicum lineage, and some are closely related to B. canariense [72,74].
Despite the prevalence of slow-growing Bradyrhizobium species, fast-growing bacteria belonging to the genera Phyllobacterium [41,42,71], Ochrobactrum [27], Microvirga [42,44,71,84], Rhizobium, Neorhizobium, Agrobacterium [43], and Devosia [12] have also been described as lupin microsymbionts. Burkholderia, a genus that nodulates other legume species, has also been isolated from L. perennis nodules [85]; however, its symbiotic effectiveness after re-inoculation has yet to be confirmed. Phylogenetic analyses based on the 23S and 16S rRNA gene sequences of fast-growing bacteria isolated from L. honoratus nodules affiliated them to the genus Ochrobactrum, and classified a new species (O. lupini) [27]. They were able to nodulate L. albus and fix nitrogen. A Phyllobacterium strain, isolated from Trifolium pretense and classified as a new species (P. trifolii), was reported to nodulate L. albus [41]. A Phyllobacterium strain was also isolated from L. micranthus nodules [42]. Two studies have reported that Microvirga sp. strains were able to nodulate L. micranthus [42,71] and L. cosentinii [44]. The 16S rRNA gene sequence of those isolates showed high nucleotide identity (up to 98%) with other species in the genus Microvirga. The isolates were classified as a new species (M. tunisiensis) [9]. Microvirga bacteria have also been isolated from L. texensis nodules and classified as a new species (M. lupini) [84] according to a phylogenetic analysis based on 16S rRNA gene sequences. Msaddak et al. [12] reported that Devosia sp. was able to nodulate L. micranthus. A phylogenetic analysis of the concatenated dnaK, atpD, and recA gene sequences led to the classification of various bacterial strains isolated from L. albus nodules in Tunisia as belonging to the genera Agrobacterium, Rhizobium, and Neorhizobium. Although the nodA and nodC genes could not be amplified, infectivity tests showed nodulation and substantial increases in the dry weight of shoots [43]. Besides nodulating bacteria, legume nodules often harbor other non-nodulating endophytic bacteria whose role in symbiosis is not clearly defined [3]. Some nodule endophytes have been described as plant growth-promoting rhizobacteria (PGPR). To cite a few examples of lupin nodule endophytes, a Paenibacillus glycanilyticus strain isolated from L. luteus nodules has been reported to modulate the lipidic and phenolic profile in white lupin, which points to PGPR inoculation as a possible strategy to enhance the nutritional quality of lupin seeds by increasing the quantity and quality of lipids and enhancing their phenolic profile [86]. Paenibacillus has also been reported as a nodule endophyte in L. albus [87]. Bacteria isolated from L. luteus and L. angustifolius have been reported to cluster with strains in the genera Rahnella, Serratia, Raoultella, and Stenotrophomonas [88]. Most of the endophytes isolated from L. mutabilis nodules belong to the family Bacillaceae [89]. A potentially new species of Micromonospora has been isolated from L. angustifolius nodules, and a nifH-like gene which was amplified from Micromonospora that presented a 99% identity with the nifH gene of Frankia alni, an actinobacterium that induces nitrogen-fixing nodules in plants within the genus Alnus [90]. Micromonospora bacteria have also been identified in nodules formed by A. glutinosa, A. viridis, and other actinorhizal plants [91], where they may have acquired the nifH gene from Frankia. Horizontal transfer of symbiotic genes from symbiotic to non-symbiotic nodule bacteria has also been reported in different legumes [92,93]. The physical proximity of nodule endophytes with symbiotic bacteria appears to make them likely recipients of symbiotic genes by horizontal transfer.

Horizontal Transfer of nod and nif Genes: On How to Become a Lupin Microsymbiont
Depending on the rhizobial genus, symbiotic genes are located in the chromosome (Bradyrhizobium), in "symbiotic islands" within the chromosome (Mesorhizobium), or in plasmids (Rhizobium, Ensifer) [94]. These genes are easily interchangeable among bacteria by lateral or horizontal gene transfer [95], and they can be used to define symbiovars within one or several rhizobial species according to the legume host specificity [37]. Nodulation (nod) genes are involved in nodule organogenesis in plant roots and stems. They are exclusive of symbiotic bacteria, and the nodulation (nodA, nodB, nodC, and nodD) genebased phylogenies are normally similar to each other, but substantially differ from the 16S rRNA gene phylogeny [96].
Previous works have shown that the nod gene phylogenies somewhat correlate with the host legume [9,96,97]. The nodC gene is the most common nodulation gene marker, as this nod gene is essential for nodulation in all rhizobial species. On the contrary, nif genes are present in many bacteria besides rhizobia, and are not unique to bacteria that are able to fix nitrogen. Nevertheless, it is not clear whether nif genes are an evolutionary part of the rhizobial genome or belong to the common bacterial gene pool [98]. It has been reported that the phylogeny of the nifH gene is similar to that of the 16S rRNA gene, which suggests that both genes are likely to share the same evolutionary history [98,99]. However, when horizontal transfer of nif genes has taken place, a phylogenetic discordance appears [100]. As indicated above, there are various bacteria previously considered as non-symbiotic that have been described as effective lupin microsymbionts, which appear to have acquired their nod and nif genes most likely through horizontal gene transfer [9,12,84].

nodC, an Essential Gene for Nodulation and Host Specificity
The symbiotic nodC gene encodes for N-acetylglucosaminyltransferase, a protein required for Nod factor assembly. Nod factors are molecules that determine host specificity and are indispensable for nodulation of compatible host legumes [23,37,101]. The nodC gene has been used to define the host range and symbiovars within the several rhizobial genera [3,9,37]. Nod genes are often considered not useful to classify species of rhizobia due to the possibility of natural horizontal gene transfer among bacterial strains in the soil. However, they prove useful for defining symbiovars. The nodC gene-based phylogeny of lupin-nodulating symbiotic rhizobia is presented in Figure 2. This analysis of the available nodC sequences from Lupinus spp. isolates allows for the identification of six different symbiovars: four of them include Bradyrhizobium isolates, one includes the Phyllobacterium and Microvirga strains, and one includes M. lupini. The majority of lupin-nodulating bacteria belong to the sv. genistearum, first described by Vinuesa et al. [73], which also includes some strains that nodulate other legume genera in the tribe Genisteae. The sv. genistearum is phylogenetically divergent from the rest of Bradyrhizobium symbiovars, which are integrated by Bradyrhizobium species that nodulate other legumes, namely sv. glycinearum [73], sv. retamae [102], sv. vignae [103], sv. sierranevadense [104], sv. centrosemae [105], sv. tropici [105], and sv. phaseolarum [105]. L. mariae-josephi-nodulating rhizobia, isolated in Spain, represent a new symbiotic lineage affiliated to sv. retamae, which also includes B. retamae [39]. Interestingly, all other rhizobial strains isolated from Lupinus spp. in the Iberian Peninsula belong to the genera B. japonicum and B. canariense, within sv. genistearum. Based on their symbiotic characteristics and nodC sequences, three strains isolated from L. micranthus nodules in Tunisia allowed for two new symbiovars to be defined: sv. lupini, which comprises two Bradyrhizobium isolates, and sv. mediterranense, which includes Microvirga and Phyllobacterium species [9]. Finally, Bradyrhizobium sp. LmicA16 and M. lupini Lut6 seem to represent to two new symbiovars, which appear distant from all other defined symbiovars. M. lupini is close to Rhizobium and Sinorhizobium (syn. Ensifer) strains, suggesting that its nodC was obtained from these rhizobial species through horizontal transfer. This is in agreement with a previous study in which a nodA gene sequence phylogenetic analysis placed M. lupini in the same clade with Rhizobium, Mesorhizobium, and Sinorhizobium species [84]. The nodC gene sequences of Ochrobactrum lupini and Phyllobacterium lupini are not available. However, their nodD gene sequences show a high similarity to the nodD gene of Rhizobium species [27,41], suggesting that their nodulation genes could have been acquired by horizontal gene transfer from this rhizobial genus.
When we compare the phylogenetic trees of the nodC gene ( Figure 2) and the 16S rRNA gene (Figure 1), we can appreciate some consistencies and inconsistencies. Most lupinnodulating Bradyrhizobium species group together within sv. genistearum, which suggests that they share a common evolutionary history or that a horizontal transfer of symbiotic genes between different Bradyrhizobium species associated with lupin and other Genisteae legumes has occurred. However, B. valentinum, as well as Bradyrhizobium sp. strains LmiH4, LmiT2, and LmicA16, which appear relatively close in the 16S rRNA tree, belong to three quite distant symbiovars according to their nodC sequences, all of them still close to bradyrhizobia that nodulate other legumes. Unlike the species in the genus Bradyrhizobium, non-rhizobial species isolated from lupin nodules appear to have obtained their symbiotic genes from other rhizobial genera via horizontal transfer [29]. Some Phyllobacterium and Microvirga strains share a common symbiovar in the proximity of Mesorhizobium species, suggesting that they acquired their symbiotic genes from rhizobia within the genus Mesorhizobium. These strains were isolated from L. micranthus in Tunisia and Morocco, so they share host plant and have close geographical origins. However, Microvirga lupini, isolated from L. texensis nodules in Texas, represents a completely different symbiovar. According to the nodC tree, it could have acquired this gene from phylogenetically close Rhizobium or Ensifer species.

nifH, a Gene Required for Effective Nitrogen Fixation
The nif genes, which codify for different nitrogenase enzyme complex subunits and several regulatory proteins, are required for nitrogen fixation [22]. They are present in rhizobia and in other bacteria capable of nitrogen fixation, including other symbiotic bacteria, such as Frankia; endophytes, such as Azospirillum; heterotrophic free-living bacteria, such as Azotobacter; and cyanobacteria, such as Anabaena, but they are also found in bacteria that are not able to fix nitrogen. The nifH symbiotic gene is one of the essential genes required to synthesize the enzyme nitrogenase, and it is considered as a highly conserved gene [106]. The nifH gene-based phylogenetic tree shown in Figure 3 includes the available nifH sequences of lupin-nodulating rhizobia and the nifH sequences with the highest similarity. Most strains associated with the genus Bradyrhizobium were isolated from lupin species growing in the Old World. American lupin microsymbionts isolated from L. albescens are somewhat distant from the Old World group. This is in agreement with a recent report in which a phylogenetic analysis of the nifD gene sequences of bradyrhizobia present in Brazilian soils was performed [107]. This analysis clustered the indigenous and the Old World-imported lupin-nodulating Bradyrhizobium strains in different clades. The fact that Old World genistoids are predominantly nodulated by Old World bradyrhizobia in American soils suggests that high symbiotic specificity is the reason why microsymbionts have been codispersed and adapted to the soils where their legume hosts have been introduced [107]. B. lupini appears closer to the American strains than to the Old World group, while B. valentinum is distant from the rest of lupin microsymbionts, and clusters together with B. retamae in concordance with the nodC gene phylogenetic analysis (Figure 2).
Regarding lupin-nodulating bacteria not belonging to the genus Bradyrhizobium, Microvirga tunisiensis groups together with M. lupini despite the fact that their respective nodC gene sequences seem to have different origins (Figure 2). They appear close to M. lotononidis and M. zambiensis, isolated from Listia angolensis nodules, a species within the only other genus besides Lupinus that forms lupinoid nodules. This suggests that Microvirga species share their own nifH gene. A phylogenetic analysis based on the nifD and nifH gene sequences of M. lupini, M. lotononidis, and M. zambiensis showed that these species are closely related to Rhizobium etli [84]. The nifH sequence of Ochrobactrum lupini isolated from L. honoratus nodules [27] shows high similarity to those of several Mesorhizobium species, which suggests that O. lupini obtained its symbiotic genes by horizontal transfer from Mesorhizobium.

Conclusions
Most Lupinus spp. microsymbionts isolated all over the world belong to the Bradyrhizobium lineage. While all lupin-nodulating rhizobia were initially referred to as Bradyrhizobium sp. (Lupinus) and sometimes as B. lupini, the advancement of molecular biology, bioinformatics tools, and gene sequence-based taxonomy has allowed for the identification of numerous species under the Bradyrhizobium sp. (Lupinus) umbrella, as shown in the phylogenetic analysis based on 16S rRNA gene sequences (Figure 1). This analysis groups lupin microsymbionts into five different groups. The largest one corresponds to the genus Bradyrhizobium, and comprises eight subgroups. Four additional clades include the fastgrowing bacteria Microvirga, Phyllobacterium, Ochrobactrum, and Devosia. Bacteria associated with the genera Agrobacterium, Rhizobium, and Neorhizobium have also been described as lupin microsymbionts according to a concatenated phylogenetic analysis which used the sequences of housekeeping genes. However, the 16S RNA gene sequences of these strains are not yet available.
It is a known fact that, in general, the phylogeny of symbiotic genes is not in congruence with that of core genes (16S rRNA and housekeeping genes) [21]. Symbiotic genes' phylogenies can be used to characterize isolates at the symbiovar level [37]. Divergences between phylogenies indicate that horizontal gene transfer of symbiotic genes might have occurred between species and/or genera. Full genome sequence analysis of symbiotic bacteria will likely increase our understanding of the legume-rhizobia interactions, and, most probably, will lead to a modification of the taxonomic criteria currently accepted for the definition of rhizobial species and genera. In fact, the taxonomy of bacteria that induce nodules in Lupinus spp. has been rearranged in recent years with the definition of new species and genera, the reclassification of some species in new genera, and the definition of new symbiovars. Moreover, a significant increase in the number of genera and species of lupin microsymbionts is expected to occur in the future, considering that most studies to date are mostly limited to Old World lupin species, while between 200 and 500 New World Lupinus species remain to be characterized.
The phylogenetic analysis of nodC sequences (Figure 2) suggests the existence of six different symbiovars: four of them include Bradyrhizobium isolates, one includes Phyllobacterium and Microvirga strains, and one includes Microvirga lupini. The majority of Lupinus spp. symbiotic bacteria in the genus Bradyrhizobium which are related to B. lupini and to the phylogenetically close B. cytisi/B. rifense lineages belong to the symbiovar genistearum. However, some lupin-nodulating Bradyrhizobium species appear to constitute new distant symbiotic lineages, namely sv. retamae and sv. lupini, and, possibly, a new symbiovar. Symbiovar mediterranense comprises Microvirga and Phyllobacterium strains, and M. lupini appears to represent another new symbiovar. Despite their taxonomic distance, it seems that the nodC symbiotic genes of all bradyrhizobia that nodulate lupin are more closely related to other Bradyrhizobium species than to other rhizobial genera. Contrary to lupinnodulating bradyrhizobia, it appears that other lupin microsymbiont species have acquired their nod genes from several rhizobial genera other than Bradyrhizobium. The nifH genebased phylogenetic tree (Figure 3) suggests that within the genus Bradyrhizobium, this gene seems to somehow be associated with the geographical origin of the strains. In the nifH tree, all Microvirga species cluster together, suggesting that, contrary to the foreign origin of their nodC gene, they share a common nifH gene which might not have been obtained from other rhizobia through horizontal transfer. The nifH gene of Ochrobactrum lupini is probably the result of a horizontal transfer event from Mesorhizobium species.
Several studies have reported the isolation of Paenibacillus, Micromonospora, Rahnella, Serratia, Raoultella, and Stenotrophomonas strains from lupin nodules; however, these bacterial isolates have not been confirmed to form effective nitrogen-fixing nodules. Most of these nodule endophytes have been reported to have PGPR traits. The horizontal gene transfer from symbiotic bacteria to endophytes represents a one-step evolution that might lead to the emergence of new symbiotic bacteria; therefore, it is ecologically and evolutionarily important [108]. Despite the physical proximity of endophytes and symbiotic bacteria in lupin nodules, it seems unlikely that they gain nodulating and nitrogen-fixing capacities from bradyrhizobia, as no horizontal transfer of symbiotic genes from Bradyrhizobium to other lupin-nodulating bacteria has been described to date. However, the transfer of nifH from soybean-nodulating B. japonicum to endophytic Bacillus has been reported [109], and still, the possibility exists that lupin nodule endophytes acquire symbiotic genes when sharing nodule occupancy with lupin microsymbionts other than Bradyrhizobium.
Overall, the data that are available so far and included in the present review suggest that horizontal transfer of specific symbiotic genes from different rhizobial genera represents a key mechanism, allowing lupin to form symbioses with selected bacteria initially considered as non-symbiotic or unable to nodulate lupin. These new lupin microsymbionts are probably better adapted to particular soil and climate conditions. Thus, the dissemination of indiscriminate symbiotic genes appears to be part of a bacterial cooperation strategy which benefits a biodiverse population rather than a particular species, and allows for the development of strain-specific legume-rhizobia symbioses in particular habitats, which would not be viable otherwise. A recent review [29] reported on the occurrence and importance of horizontal transfer of symbiotic genes within and between rhizobial genera, addressing the diversity of microsymbionts that are able to nodulate different legume genera. The authors did not report, however, on lupin and its nodulating bacteria. Lupin is a singular legume crop in many ways, and is, foremost, a promising yet underexploited plant protein source. It also presents great potential as a medicinal plant, owing to its seed particular protein, lipidic, and phenolic profiles. Its application in the remediation of poor and degraded soils and heavy metal-contaminated soils has also been proposed as very promising. Well-adapted microsymbionts have been shown to increase lupin tolerance to heavy metals, and certain inoculants have been reported to improve the nutritional and antioxidant characteristics of lupin seeds. In the present review, we summarized the current knowledge on the diversity of microsymbionts that nodulate Lupinus spp., and the important role that horizontal transfer seems to play. Our conclusions are in agreement with those of Andrews et al. [29]. Deeper insight into the nature of lupin microsymbionts, including comprehensive whole genome analyses, will undoubtedly provide new knowledge that might prove useful for the purpose of increasing lupin's stress tolerance, adaptability to marginal soils and climates, grain quality, and crop yields.