Non-Coding RNAs in Legumes: Their Emerging Roles in Regulating Biotic/Abiotic Stress Responses and Plant Growth and Development

Noncoding RNAs, including microRNAs (miRNAs), small interference RNAs (siRNAs), circular RNA (circRNA), and long noncoding RNAs (lncRNAs), control gene expression at the transcription, post-transcription, and translation levels. Apart from protein-coding genes, accumulating evidence supports ncRNAs playing a critical role in shaping plant growth and development and biotic and abiotic stress responses in various species, including legume crops. Noncoding RNAs (ncRNAs) interact with DNA, RNA, and proteins, modulating their target genes. However, the regulatory mechanisms controlling these cellular processes are not well understood. Here, we discuss the features of various ncRNAs, including their emerging role in contributing to biotic/abiotic stress response and plant growth and development, in addition to the molecular mechanisms involved, focusing on legume crops. Unravelling the underlying molecular mechanisms and functional implications of ncRNAs will enhance our understanding of the coordinated regulation of plant defences against various biotic and abiotic stresses and for key growth and development processes to better design various legume crops for global food security.


Introduction
Legumes are the third largest family of flowering plants, and grain legumes are essential components of the human food diet, supplying 'plant-based dietary proteins' and essential micronutrients and vitamins [1][2][3]. Thus, legume crops serve as an essential component for sustaining global food security. Their ability to fix atmospheric nitrogen through symbiotically active bacteria in root nodules enriches soil nitrogen content and minimizes the use of chemical-based nitrogenous fertilizers, thus protecting the environment from pollution [1]. In the past, elucidating the function of protein-coding genes controlling biotic and abiotic stresses and developmental processes in plants has involved conventional breeding and biochemical and molecular approaches [4]. However, rapid progress in functional genomics, especially transcriptome sequencing by RNA-seq, has given us the opportunity to investigate RNAs that do not code proteins, known as ncRNAs, which control diverse biological functions in the plant kingdom [5]. These ncRNAs are classified as small ncRNAs, comprising miRNAs (21-24 nt long) [6], small interfering RNAs (siRNAs) [7], Piwi-interacting RNAs (piRNAs) (generally found in animals) [8] and lncR-NAs (>200 nt long) [9]. circRNA are another class of ncRNA generated from pre-mRNA splicing, featuring closed 3 and 5 ends covalently [10]. In addition to these ncRNAs, small nucleolar RNAs (snoRNAs), ribosomal RNAs (rRNAs), and transfer RNAs (tRNAs) known as housekeeping ncRNAs are also found in plant species [11]. The main classes of ncRNAs, illustrated in Figure 1, contribute to various plant development pathways and abiotic and

Deciphering the Molecular Mechanisms of ncRNAs Regulating the Response of Legumes to Water Stress
Drought stress is the most important abiotic stress globally, affecting all plant growth and developmental stages, and ultimately reducing crop yields [145]. Plants adapt to a water deficit environment by evoking various physiological, biochemical, metabolic, and molecular mechanisms [146]. Many QTL/genes contributing to drought tolerance have been investigated in various legumes [147]. Indeed, the participatory role of various regulatory ncRNAs and their corresponding target gene(s) controlling drought stress have been deciphered in various plant species, including legumes [55,95,113,125]. A plethora of novel drought-responsive miRNAs have been identified in legume crops-157 in cowpea [55], 143 and 128 in grass pea [121], and 284 in chickpea [92]-and 3457 high-confidence lncR-NAs have been identified in chickpea [66]. ncRNAs confer drought tolerance by regulating gene(s) encoding various regulatory TFs and osmoregulatory/osmoprotective compounds by activating hormone signalling and antioxidants that minimize oxidative stress/reactive oxygen species (ROS) activity in plants under water stress [55,113,121].
A study on the participatory role of conserved miRNAs-miR398a/b and miR408-in regulating water stress in pea revealed the downregulation of these miRNAs in root and shoot tissue under water deficit conditions [125]. However, the copper superoxide dismutase, CSD1 target gene of miR398a/b, was upregulated, suggesting an inverse relationship between the target gene and the involved miRNA controlling water stress in pea. Similarly, De la Rosa et al. (2019) [48] supported the upregulatory role of CSD1 and ADH1 mRNAs targeted by miR398 and miR2119 in common bean adapting to drought stress.
Considering the role of circRNAs attributing drought tolerance, Dasmandal et al. (2020) [148] uncovered numerous drought responsive differentially expressed circRNAs in chickpea and soybean. The authors also predicted three eTMs those acted as sponge for miR-NAs that target Glyma.18G065200.1 gene in soybean, and XM_004517122, XM_027336693 genes in chickpea. The functional role of these targeted genes was associated with hormone signalling and various transcription factors under drought stress [148]. Further mechanistic understanding of ncRNAs and the corresponding target gene(s) will enhance our understanding of ncRNAs regulating drought tolerance in legume crops.

Role of ncRNAs in Plant Adaptation to Salinity Stress
The rapid conversion of uncultivable land to cultivated land and the excessive use of irrigation water have increased salinity, which is a major challenge for crop growth, including legumes, and causes significant yield losses [149]. Plants orchestrate various biochemical and molecular mechanisms to survive the increasing salinity stress [149], including ncRNAs [15,82,83,102], which target genes related to photosynthesis, TFs regulating growth, genes related to salinity-responsive hormone signalling, genes that minimize the uptake of toxic ions, viz., Na + , and genes that limit ROS activity [15,83,95]. Paul et al. (2011) [114] investigated the role of miRNAs controlling salinity stress in cowpea and recovered 18 conserved miRNAs (e.g., miR160, miR156/157, miR159, miR169, miR172, miR408) from root tissue and identified 15 corresponding target gene(s) as TFs (e.g., ARF, SBP, AP2, TCP). Functional validation through quantitative real-time PCR (qRT-PCR) revealed the upregulation of seven miRNAs under salinity stress.
A genome-wide survey of lncRNA through transcriptome analysis in groundnut identified 1442 lncRNAs [102]; notably, TCONS_ 00292946 lncRNA was downregulated in roots within 12 h of salinity stress but upregulated at 24 h. Expression of TCONS_00176941 was upregulated within 12 h in roots and downregulated within 12 h of salinity stress in leaves, while TCONS_00011551 was upregulated under salinity stress [102].   [15] investigated the role of lncRNAs involved in regulating the salinity stress response and conferring tolerance by alleviating ROS-related stress in Medicago truncatula. The authors identified the functional role of various lncRNAs attributing to salinity tolerance, including TCONS_00116877, which induced the Medtr7g094600 gene encoding glutathione peroxidase to minimize ROS-derived stress in roots (see Table 2).
Alzahrani et al. (2019) [116] uncovered 1220 salt-responsive miRNAs by small RNA sequencing of two contrasting faba bean (Vicia faba) genotypes for salinity stress response (ILB4347 tolerant and Hassawi-3 sensitive). The Hassawi-3 genotype had 284 upregulated and 243 downregulated miRNAs, while ILB4347 had 298 upregulated and 395 downregulated miRNAs in the control and under salinity stress. The target gene(s) were predicted to encode TFs, laccases, superoxide dismutase, plantacyanin, and F-box proteins in addition to genes involved in hormone signal transduction, phosphatidylinositol signalling, and the MAPK signalling pathway [116].

Contribution of ncRNAs Attributing Plant Adaptation under Metal Toxicity Stress
Metal toxicity is an abiotic stress increasingly faced by plants due to rapid industrialization, excessive use of inorganic fertilizers, and overuse of irrigation water contaminated with heavy metals, especially cadmium and mercury [150]. Among the various complex molecular mechanisms, identifying the role of ncRNAs, including miRNAs and lncRNAs, is a potential approach for minimizing metal toxicity in plants [77,151].

Molecular Mechanisms of ncRNAs Regulating Nutrient Acquisition and Homeostasis in Legumes
Plants acquire essential nutrients by recruiting various physiological and molecular mechanisms via roots and soil for proper growth and development [153,154]. Of these mechanisms, the critical role of ncRNAs regulating the uptake of various macro-and micronutrients has been recognized [155,156].
Nitrogen (N)-serving as the source of various essential amino acids and acting as an important element for entire nitrogen metabolism-is a critical determinant for plant growth and development [157]. Emerging functional genomics approaches, viz., RNA-seq, can underpin the plethora of nitrate transporter QTLs, gene(s), and ncRNAs controlling N use efficiency (NUE) in plants [158]. However, the complete molecular mechanism of NUE/N homeostasis remains unclear in plants, including legumes.
Phosphorus (P) is the second most essential macronutrient required for basic biochemical and metabolic processes in plants, including legumes [161]. Plants usually uptake P in the form of inorganic phosphate (Pi). Thus, P deficiency limits overall plant growth and development. The involvement of several P-responsive ncRNAs has been elucidated in various plant species [14,17,160,[162][163][164]. Likewise, previously P-responsive miRNAs have been reported in common bean [60,144], white lupin [119], soybean [165], M. truncatula [132], alfalfa [166,167], and lupin (Lupinus albus) [119]. Several conserved regulatory miRNAs, such as miR399 [162,[168][169][170] and miR156, miR169, and miR2111 [160] regulating Pi homeostasis have been reported in Arabidopsis. Li et al. (2018) [13] confirmed the inductive role of miR399 (targeting phosphate transporter genes) and miR398 (targeting Copper chaperone for SOD) under low Pi stress in roots of Medicago sativa. However, the authors noted downregulation of miR156 (targeting SPL TF), miR159 (targeting MYB TF), miR160 (targeting auxin response factor TF), miR171 (targeting GRAS TF), and miR2643 (targeting MATE). The molecular mechanism involving IPS1 lncRNA serving as eTM for miRNA399 targeting PHO2 gene expression and controlling Pi homeostasis has been established in Arabidopsis [34]. Under low Pi conditions, upregulation of the PHR1 gene and miRNA399 inhibiting the PHO2 gene (encoding transcript causing Pi transporter degradation) enabled high Pi acquisition in Medicago sativa [17]. Downregulation of PDIL2 and PDIL3 lncRNAs enhanced transcript expression of Medtr1g074930, a Pi transporter gene, enabling high Pi uptake under low Pi conditions. However, PDIL1 lncRNA serves as a target mimicry for miR399, inhibiting the degradation of MtPHO2 transcripts that could downregulate the Pi transport gene and Pi uptake [17] (see Figure 2). To gain insight into the role of P-responsive circRNAs, Lv et al. (2020) [14] uncovered 120 differentially expressed cicr-RNAs by transcriptome sequencing of two contrasting P-responsive soybean genotypes at different P levels. Gene ontology (GO) enrichment analysis predicted that the putative role of the differentially expressed circRNAs is related to nucleoside binding, organic substance catabolic processes, and oxidoreductase activity [14]. Low P-responsive circRNAs could be targeted for improving phosphorus use efficiency in soybean. Thus, a complex network of ncRNAs and their corresponding target gene(s) play a central role in regulating Pi homeostasis in plants.

Regulatory Role of ncRNAs for Shaping Developmental Processes in Legume Species
Apart from various biotic and abiotic stresses, ncRNAs, including miRNAs (conserved and nonconserved) and lncRNAs, play a pivotal role in regulating plant growth and development and in various metabolic pathways, which has been investigated in various legume species [61,91,92,96,103,120,171,172]. Small, deep RNA sequencing analysis of seven chickpea tissues was used to investigate a comprehensive set of 440 known and conserved and 178 novel miRNAs targeting various TFs and gene(s) that control various developmental processes, including leaf, flower, pod, and root development and various metabolic processes in chickpea [92] (see Table 1). Subsequently, small RNA sequencing of chickpea leaves and flowers discovered 157 conserved and novel miRNAs that regulate various developmental processes and stress responses [12]. Of the identified miRNAs, miR156, miR159, miR160, miR162, miR164, miR172, miR408, and miR393 targeting SBP, MYB, ARF, DCL1, HD-zip, AP2, F-box protein, and plantacyanin encoding genes, respectively, contribute to various plant development processes [12] (see Table 2). The authors also disclosed the role of TAS3-derived tasiRNA targeting ARF2, ARF3, and ARF4 transcription factors controlling auxin response, and thus contributing to development pathways in chickpea. In this context, Jagadeeswaran et al. (2009) [51] identified and characterized Tas3-siRNAs from M.trucatula and also functionally validated three ARF genes targeted by these Tas3-siRNAs.
Considering ta-siRNA participating in regulating compound leaf and flower development in L. japonicus, Yan et al. (2010) [173] established the role of Reduced leaflet1 (REL1) and Reduced leaflet3 (REL3) genes encoding homologs of Arabidopsis (Arabidopsis thaliana) 'SUPPRESSOR OF GENE SILENCING3 and 'ARGONAUTE7/ZIPPY', respectively, key components required for ta-siRNA biogenesis. Positional cloning analysis of REL1 and REL3 genes revealed that the ta-siRNA pathway critically plays significant role in controlling compound leaf and flower development in L. japonicus [173]. Likewise, elucidating the role of trans-acting siRNA3 (TAS3) involved in leaf margin indentation and organ separation, Zhou et al. (2013) [174] examined that Mt-AGO7/LOBED LEAFLET1 is required for the biogenesis of ta-siRNA to suppress the expression of Auxin Response Factors. Evidence of lobed leaf margin and widely spaced lateral organ phenotype demonstrated in the ago7 mutant suggested that TAS3 plays a negative role in leaf margin and lateral organ development in M. truncatula [174].  [120] reported several miRNAs regulating floral development, viz., Ll-miR280, Ll-miR281, and Ll-miR285 (possibly targeting ARF6 and ARF8); Ll-miR445 and Ll-miR130 (targeting TCP4 and MYB33); and Ll-miR329/miR160-5p, Ll-miR332/miR160-5p, and Ll-miR333/miR160-5p miRNAs regulating flower abscission in yellow lupin (Lupinus luteus L.). Among the siRNAs identified from this study, Ll-siR173, Ll-siR4, and Ll-siR13 exhibited upregulation and downregulation of Ll-siR208, suggesting the active role of siRNA functioning in lupin pedicel [120]. Das et al. (2019) [57] explored a plethora of lncRNAs and target miRNAs forming an endogenous target mimicry leading to pod and seed development using transcriptome analysis of tissue collected during anthesis and pod development in pigeon pea. Functional validation revealed that sequestering Cc-miR160h by Cc_lncRNA-2830 enabled the transcription of XM_020377020 (encoding auxin response factor 18-like protein) during pod development at 10 and 20 days after anthesis (DAS). However, expression of Cc_lncRNA-2830 at 30 DAS decreased, which upregulated Cc-miR160h and degraded the XM_020377020 transcript [57] (see Figure 2).
As the entire underlying molecular mechanism for seed development, from embryogenesis and filling to maturation, remains elusive [98], several investigations have reported the involvement of various ncRNAs regulating seed development in grain legumes [92,93,98,102]. To investigate the contributory role of ncRNA involved in the seed development process, transcriptome sequencing of seed samples using an Illumina Genome Analyzer IIx uncovered 72 known and 39 new miRNAs involved in seed development, particularly embryogenesis, dormancy, and maturation, in common bean [98]. The notable miRNAs and the target genes involved in regulating seed development were MIR156 repressing SPL; MIR169 repressing NF-YA1 and NF-YA9; MIR399 inhibiting SUT1 related to sucrose transport; MIR399 inhibiting PHO2 contributing in phosphorus allocation; MIR160 repressing ARF10, ARF16, and ARF17; MIR167 inhibiting NCED1 associated with ABA synthesis; and MIR395 repressing SULTR2;1, APS contributing to sulphate assimilation and allocation during seed filling [98]. Likewise, genome-wide profiling of miRNAs using small RNA sequencing of seeds of two contrasting chickpea genotypes-Himchana1 (low seed weight) and JGK3 (high seed weight)-unfolded 113 known and 243 novel miRNAs controlling seed development in chickpea [93] (see Table 1). The target genes of identified miRNAs contributing to seed development were predicted to be SPL, GRF, MYB, ARF, HAIKU1, SHB1, KLUH/CYP78A5, and E2Fb. Low expression of Car-miR319 and Car-miR166 and upregulation of their corresponding target genes, bZIP and homeobox-REVOLUTA TFs, in JGK3 indicated their important role in seed size determination in chickpea [93]. The authors also located 19 miRNAs and 41 target genes in previously identified QTLs contributing to seed size.
The role of various conserved miRNAs, viz., miR167, miR390, miR164, miR399, miR156/157, miR1511, and mir319, and seven novel miRNAs, viz., NovmiR13, NovmiR12, and NovmiR04, regulating seed development in narrow-leafed lupin was confirmed in studies by DeBoer et al. (2019) [118]. Differential expression analysis revealed upregulation of Lan-miR-156a-2, Lan-miR-164-3, Lan-miR-167a/c, Lan-miR-319, Lan-miR-399b/c, NovmiR12, and Nov-miR13 in seeds, indicating their role in regulating seed development in lupin [118]. The role of miRNAs controlling genes related to sugar metabolism during seed development is worth mentioning [87,175]. In soybean, deep sequencing and degradome sequencing of developing soybean seed revealed several miRNAs targeting genes that contribute to seed development [87]. Among the identified miRNAs, functional validation of gma-miR1530 revealed its role in inhibiting the target transketolase gene that contributes to switching carbon assimilation to energy metabolism during seed development. Likewise, the pentatricopeptide repeat protein-encoding gene was targeted by Soy_3 and Soy_16, while Soy_25 (targeting Glyma05g33260 homolog of Arabidopsis "SUPPRESSOR OF GENE SILENCING 3") contributing to seed development was identified [87] (see Figure 2). A total of 484 miRNAs were recovered from small RNA sequencing of four contrasting soybean lines with high protein/high oil, high protein/low oil, high oil/low protein, and low protein/low oil [175]. Functional validation of selected miRNAs, including Glyma.13G035200 and Glyma.14G156400 (encoding alcohol dehydrogenase 1) targeted by Gma-miR2119, Glyma.04G178400 (encoding ADP-glucose pyrophosphorylase family protein) targeted by Gma-miR1521a, and Glyma.19G094000 (related to sugar synthesis and metabolism) targeted by miR156, using RT-qPCR indicated their significant role in controlling storage genes during seed development in soybean [175].
Computational analysis identified 347 candidate circRNAs in groundnut [110]; the differential expression of 29 circRNAs was upregulated in seeds collected from RIL 8107 at 35 days after flowering (DAF) and RIL 8106 at 35 DAF, confirming their contributory role in seed development [110]. Likewise, Ma et al. (2020) [111] detected 9388 known and 4037 novel lncRNAs in groundnut, of which 1437 lncRNAs were differentially expressed. Functional validation of selected lncRNAs confirmed their role in seed development. The participatory role of miR156, miR159, miR171, and miR14 (targeting genes related to cellular amino acid metabolism, fatty acid metabolism, and lipid metabolism) in groundnut is noteworthy [56].
To establish the role of the DCL2-dependent 22-nucleotide siRNA (derived from long inverted repeats) regulating chalcone synthase (CHS) genes attributing seed coat colour in soybean, a study conducted by Jia et al. (2020) [176] revealed that CRISPR/Cas9-driven loss-of-function mutants of DCL2 (GmDCL2a and GmDCL2b) caused changes in seed coat colour from yellow to brown in Gmdcl2a/2b mutants in soybean. Thus, this study confirmed that DCL2 controls soybean seed coat colour via generating siRNA from long inverted repeats [176].
Further identification of ncRNAs related to the development process, especially pod and seed development, and their precise function will provide better new avenues for improving pod and seed size and thus grain yield in legumes.

ncRNAs Orchestrating Nodulation, Symbiosis, and Root Development Processes
Legumes are unique due to their inherent ability of forming root nodules in association with active soil rhizobacteria that assist in fixing atmospheric nitrogen [1]. The underlying molecular mechanism and around 200 genes involved in fixing atmospheric nitrogen in soil through nodulation and symbiosis have been deciphered [177,178]. Likewise, evidence of small RNAs, including miRNAs involved in nodule development and root symbiosis, has been reported in various model legumes, viz., M. truncatula, L. japonicus, and soybean [49,51,76,133,[179][180][181][182][183]. The greater abundance of miR172 in root nodules than leaf tissue in Medicago truncatula [76], Lotus japonicus [138], common bean [60], and soybean [134] suggests its active role in nodulation. The role of MIR166 (targeting HD-ZIP III TF genes contributing to root nodule development) in Medicago truncatula was revealed by its overexpression, which downregulated HD-ZIP III, inhibiting symbiotic nodules and lateral root development [132]. Similarly, in soybean, miR166 and miR396 (targeting HD-ZIP III TF and cysteine protease gene, respectively) depicted downregulation during nodulation in soybean [49].
Considering the potential role of miRNAs involved in signalling pathways related to nodule infection and N 2 fixation, De Luis et al. (2012) [138] demonstrated that the induction of miR171c in root nodules targeting NSP2 TF is correlated with bacterial nodule infection. While the induction of miR397 is noted strictly in rhizobial bacteria-infected active N 2 fixing nodules, it participates in contributing to nitrogen fixation-related copper homeostasis and also targets the laccase copper protein family gene in Lotus japonicus [138]. Subsequently, the negative role of gma-miR171o and gma-miR171q miRNAs regulating soybean nodulation was functionally validated [184]. The authors demonstrated that the regulatory expression of two TF genes, GmSCL-6 and GmNSP2 (target genes of gma-miR171o and gma-miR171q miRNAs), plays an active role in the expression of NIN, ENOD40, and ERN genes involved in the nodulation process in soybean. Among the other miRNAs attributed to the nodulation process, the regulatory circuit of nodule development controlled by miRNA172-targeting AP2 and miRNA156-regulating miRNA172 expression in soybean has been investigated [49,134].
Various research groups [140,185,186] have suggested that the negative regulation of miR171h targeting MtNSP2 is needed for nodule formation and the mycorrhizal signalling pathway in Medicago truncatula. Overexpression of miR396b in roots of Medicago truncatula impaired root growth and diminished mycorrhizal colonization by targeting six growthregulating factor genes (MtGRF) and two bHLH79-like genes, indicating the significant role of miR396b in root growth and mycorrhizal colonization [139] (see Table 2). Further insights into the underlying complete molecular mechanism of miR172c controlling rhizobial infection and precise nodulation regulation were elucidated in soybean [135]. The authors postulated that the absence of rhizobia Nodule Number Control1 (NNC1) suppresses the transcription of ENOD40 genes in soybean. However, in the presence of rhizobia, nod factor receptors induced a signal cascade that evokes the upregulation of miR172c targeting the NNC1 gene. Thus, the inhibition of NNC1 allows transcription of ENOD40 genes leading to nodule organogenesis in soybean (Figure 3).  [81] supported that the inductive activity of miR172a in L. japonicus roots requires the presence of both active rhizobial bacteria and bacterial Nod factor signalling during the early stage of symbiotic infection. Possible targets of miR172a were predicted to be the RAP2-7-like1, AP2-like1, and AP2-like2 genes during bacterial symbiosis. Subsequently, Yan et al. (2015) [84] functionally demonstrated that the overexpression of miR393j-3p miRNA targeting a nodulin gene Early Nodulin 93 (ENOD93) significantly inhibited nodule formation in soybean. Turner et al. (2012) [85] monitored the high expression of Glyma10g10240 and Glyma17g05920 (targets of miRNA169), which encode HAP proteins that contribute to nodule development.
The role of miR169 in regulating nodule development (transition from meristematic to differentiated cells) in M. truncatula by targeting the MtHAP2-1 novel symbiosisspecific TF gene has been established [133] (Figure 3). Li et al. (2010) [129] supported the role of miR482, miR1512, and miR1515 with enhanced nodule numbers at the transgenic level, thus suggesting their role in nodule development in soybean. However,   [136] demonstrated that overexpression of miR156 in transgenic plants caused inhibited nodule development in Lotus japonicus. Similarly, in common bean, overexpression of miR319 the target TCP10 TF gene mRNA, which positively induces the action of the LOX2 gene involved in jasmonic acid synthesis [141], stimulated the nodule development but decreased the rhizobial infection process [141].
Furthermore, to gain deeper insight into the role of miRNAs regulating nodulation  [81] supported that the inductive activity of miR172a in L. japonicus roots requires the presence of both active rhizobial bacteria and bacterial Nod factor signalling during the early stage of symbiotic infection. Possible targets of miR172a were predicted to be the RAP2-7-like1, AP2-like1, and AP2-like2 genes during bacterial symbiosis. Subsequently, Yan et al. (2015) [84] functionally demonstrated that the overexpression of miR393j-3p miRNA targeting a nodulin gene Early Nodulin 93 (ENOD93) significantly inhibited nodule formation in soybean. Turner et al. (2012) [85] monitored the high expression of Glyma10g10240 and Glyma17g05920 (targets of miRNA169), which encode HAP proteins that contribute to nodule development.
The role of miR169 in regulating nodule development (transition from meristematic to differentiated cells) in M. truncatula by targeting the MtHAP2-1 novel symbiosis-specific TF gene has been established [133] (Figure 3). Li et al. (2010) [129] supported the role of miR482, miR1512, and miR1515 with enhanced nodule numbers at the transgenic level, thus suggesting their role in nodule development in soybean. However,   [136] demonstrated that overexpression of miR156 in transgenic plants caused inhibited nodule development in Lotus japonicus. Similarly, in common bean, overexpression of miR319 the target TCP10 TF gene mRNA, which positively induces the action of the LOX2 gene involved in jasmonic acid synthesis [141], stimulated the nodule development but decreased the rhizobial infection process [141].
Furthermore, to gain deeper insight into the role of miRNAs regulating nodulation and the symbiosis process, Sós-Hegedűs et al. (2020) [142] established and functionally validated the regulatory mechanism of the nodulation and symbiosis process through silencing of target NB-LRR genes by miR2118, miR2109, and miR1507 miRNAs in Medicago truncatula. During nodulation and symbiotic nitrogen fixation, the symbiotic bacteria upregulate miR2118, miR2109, and miR1507 miRNAs, at the cost of downregulating NB-LRR genes; consequently, the plant's innate immunity is compromised during symbiosis in nodules [142] (see Figure 2). Recently, Tsikou et al. (2018) [187] and Gautrat et al. (2020) [131] suggested that miR2111 targeting TOO MUCH LOVE (encoding F-box/kelch-repeat protein), a nodulation suppressor, could enhance nodulation. However, the prevalence of rhizobial inoculation/infection and nitrate treatment reduced the level of miR2111s in leaves and roots, depending on the shoot-acting hypernodulation and aberrant root 1 (HAR1) receptor. Moreover, describing the fine-tuning and autoregulation mechanism of nodulation, Gautrat et al. (2020) [131] postulated that the Clavata3/Embryo surrounding region 12 (CLE12) and the CLE13 signalling peptides synthesized in roots act through HAR1/super numeric nodule (SUNN) receptors to negatively regulate the action of miR2111 [130]. This miR2111 otherwise favours root symbiotic nodulation under nitrogen-starved conditions by C-terminally encoded peptide (CEP) produced in root and acts in shoot through the compact root architecture 2 (CRA2) receptor. Likewise, Okuma et al. (2020) [130] confirmed the regulatory role of HAR1-dependent miR2111s produced from the MIR2111-5 locus in shoots controlling root nodulation in Lotus japonicus using functional analysis. Apart from these model legumes, three A. hypogaea-specific miRNAs, ahy-miR3508 (targeting gene encoding pectinesterase), ahy-miR3509, and ahy-miR3516, were identified; however, it is not known whether they participate in the nodulation process [108]. In common bean, genome-wide transcriptome analysis using Genome Analyzer IIx and degradome analysis identified 185 mature miRNAs and 181 targets for these identified miRNAs [100]. Functional characterization of selected miRNAs, viz., miRNov153 targeting uridine kinase (Phvul.003 g180800), miR319 targeting TCP TF family member (Phvul.011 g156900), and miR-Nov494 targeting aldehyde dehydrogenase (Phvul.004G162200.1), were upregulated, but their corresponding target genes were downregulated, indicating their significant involvement in controlling nodule development in common bean [100].
Furthermore, these miRNAs, an abundance of 21-nucleotide phased siRNAs derived from PHAS loci corresponding to protein coding genes NB-LRRs, were noted in soybean nodule [90] and in common bean nodule [100]. Likewise, evidence of circRNAs involved in nodule development and rhizobial symbiosis has been reported in common bean [188]. The authors suggested their role of acting as eTM and regulating the transmembrane transport and positive regulation of kinase activity during nodule development and the nitrogen fixation process. Recently, Tiwari et al. (2021) [189] and Hoang et al. (2020) [190] comprehensively discussed the interplay of various miRNAs impacting hormone signalling and regulating various regulatory genes during rhizobial infection, nodule organogenesis, and nitrogen fixation. A thorough understanding of various gene networks and their interplay with regulatory ncRNAs and precise function in controlling nodulation and related processes during the symbiosis process will further illuminate our insights into legume symbiosis at the molecular level involving ncRNAs.

Conclusions and Future Perspectives
The discovery of ncRNAs and their functional annotation have received considerable interest for investigating the underlying molecular mechanisms controlling various biological phenomena in legumes and opened a new avenue for improving traits of interest. As ncRNAs are dynamic, they are rapidly being discovered and functionally characterized in various plant species, including legumes [19]. However, the complete characterization of discovered ncRNAs at the functional level and their target gene(s) is limited to model legumes, viz., M. truncatula, L. japonicus, and soybean. Other legumes also need attention for the investigation of novel ncRNAs and their functions. Emerging approaches including powerful deep transcriptome sequencing technologies and advances in computational biology will facilitate the discovery of more ncRNAs and annotate their function. Moreover, emerging approaches of genome editing technology, viz., CRISPR/Cas9, will enable the functional characterization of novel ncRNAs (through loss-of-function/gain-of-function analysis) or manipulation of miRNAs causing the reprogramming of gene expression that controlling various traits of breeding importance with high precision [21,130,191]. Thus, the artificial manipulation of ncRNAs controlling various breeding traits could help us develop designer crops for sustaining global food security under predicted climate change scenarios. Institutional Review Board Statement: Ethical review and approval were waived for this study due to it being an extensive literature search study.

Informed Consent Statement: Not applicable.
Data Availability Statement: Not applicable.