The first step in root symbiosis is the molecular signal exchange between roots and nitrogen-fixing rhizobia or mycorrhizae. Legume-derived flavonoids induce the biosynthesis of Nod factors in rhizobia, and some symbiosis-related flavonoids are accumulated at the colonization site. This suggests that Nod factors promote flavonoid biosynthesis in a positive feedback loop (Liu and Murray, 2016; Panche et al., 2016; Sharma et al., 2020). These root-excreted flavonoids serve as chemo-attractants to facilitate the movement of rhizobia (e.g., Sinorhizobium meliloti) to root hairs (Hassan and Mathesius, 2012). Interestingly, some specific host flavonoids, such as luteolin and naringenin, were shown to induce Ca²⁺ transients in rhizobia, which subsequently activates bacterial Nod-related genes (i.e., nodA, nodB, and nodC) in Rhizobium leguminosarum cv. viciae (Moscatiello et al., 2010; Cui et al., 2019). These findings suggest Ca²⁺ signaling in bacterial symbionts plays a role during plant symbiotic microbe interaction. When studying arbuscular mycorrhizal fungi (AMF)-peanut symbiosis, it was noted that exogenous Ca²⁺ application improved the colonization of peanut roots by AMF and induced the expression of plant genes, including those genes regulating flavonoid biosynthesis (Cui et al., 2019). The results from these studies suggest Ca²⁺ signaling is important during the initial interactions between plants and symbionts.

Nod factor-induced Ca²⁺ spikes in root hairs is essential for plant root-nodule symbiosis. Earlier studies have revealed that S. fredii-derived NGR234 Nod factor was able to induce Ca²⁺ concentration increases within root hairs of nodulating legumes, such as Chamaecrista fasciculata, Acacia retinoides, Cytisus proliferus, Lupinus pilosus, and Medicago truncatula (Granqvist et al., 2015). However, Nod factor failed to trigger Ca²⁺ oscillations in the non-nodulating legume Cercis siliquastrum (Granqvist et al., 2015). This observation suggests that Ca²⁺ transients are common events in rhizobia-compatible plants. To further investigate Ca²⁺ transients in non-leguminous plants, Granqvist et al. (2015) observed that non-leguminous plants (e.g., Parasponia andersonii) could form a symbiotic relationship with rhizobia and exhibit Ca²⁺ spiking in response to NGR234 Nod factors. However, Trema tomentosa, a non-nodulating plant related to Parasponia, did not exhibit Nod factor-induced Ca²⁺ oscillations (Granqvist et al., 2015). These results suggest that Nod-factor-triggered Ca²⁺ oscillations are a common feature in response to symbiotic bacteria in nodulating species.

Nitrogen fixation in endosymbiotic plant-bacterial associations is limited to the Fabid clade (e.g., squash, bean/pea and rose families). The most well-studied bacterial associations are between legumes and Rhizobium. The association between a nitrogen-fixing filamentous bacteria (Frankia) and a diverse range of trees and woody shrubs is less well characterized. However, a recent study found that a novel symbiotic factor from Frankia CcI3 strain was resistant to chitinase treatment and had relatively low molecular weight (i.e., in the range 0.5–5 KDa) (Chabaud et al., 2016). The novel symbiotic factor could trigger Ca²⁺ spikes in root hairs and induce nodule inception (CgNIN) gene expression in the actinorhizal plant Casuarina glauca. This finding suggests that certain symbiotic responses, such as Ca²⁺ spiking, are conserved across plants that can form symbiotic associations in the Fabid clade (Chabaud et al., 2016).

Since nuclear Ca²⁺ oscillations are required for rhizobial and mycorrhizal symbioses, studies to better understand Ca²⁺ oscillations have mainly focused on ion channels and a pump located at the nuclear envelope (NE). The Ca²⁺ channels include CNGCs among which CNGC15s is crucial for the observed Ca²⁺ spiking triggered by rhizobia colonization in Medicago. The CNGCs are regulated by calmodulin (CaM) via its interaction with the CNGC isoleucine glutamine (IQ) motif. A recent study found that Ca²⁺-bound CaM2 regulates Ca²⁺ spiking by associating with the CNGC15s (e.g., CNGC15a, CNGC15b, and CNGC15c), which results in closed Ca²⁺ channels. The closing of the channel prevents it from releasing Ca²⁺ into the nucleoplasm, while a calcium pump (MCA8) drives calcium back to the nuclear envelope lumen; the opening and closing of the channels shape the nucleoplasmic calcium concentration (Cerro et al., 2022). A mutated CaM2, called CaM2^R91A, displayed increased binding affinity to CNGC15s. When CaM2^R91A was expressed in Medicago truncatula, the plants exhibited an increased Ca²⁺ oscillation frequency during early stage of colonization in both AM and rhizobia. Moreover, plants expressing CaM2^R91A showed enhanced Nod-factor-mediated induction of nodulation-related genes, such as NIN and NF YA1. Although the CaM2^R91A expressing plants were able to maintain enhanced bacterial symbiosis at later timepoints (14 and 28 dpi), they could not sustain AM intraradical hyphae and arbuscule formation in the roots (Cerro et al., 2022). Thus, the Ca²⁺-bound form of CaM2 plays an important role in modulating CNGC15 activity and the subsequent Ca²⁺ oscillations, but the downstream Ca²⁺-mediated signaling networks differ between AM and root nodule symbiosis ( Figure 1 ).

The nuclear pore complex (NPC) in Lotus japonicus is essential for Nod factor-induced nuclear Ca²⁺ oscillations (Kanamori et al., 2006). Nod factor triggered a weaker nuclear Ca²⁺ oscillation in mutant nup133 as compared to the wild-type control, and the nup133 mutants showed no mycorrhizal colonization and reduced nodulation by Rhizobium bacteria at permissive temperatures (Kanamori et al., 2006). In L. japonicus, the nucleoporin gene NUP85 was also required for Nod-factor-induced nuclear Ca²⁺ oscillation as well as bacterial nodulation and mycorrhizal colonization (Saito et al., 2007). The nuclear pore complex may mediate Nod factor-induced nuclear Ca²⁺ oscillations indirectly by modulating the transport of symbiosis-related mRNAs [such as Nod receptors and symbiosis-related Ca²⁺ channels (CASTOR, POLLUX/DMI1, CNGC15) or Ca²⁺pumps (MCA8)] from nucleus to ribosome for polypeptides or protein biosynthesis. Another possibility is that the NPCs regulate the localization of CASTOR, POLLUX/DMI1, CNGC15 to the nuclear membranes, although the biological mechanism involved deserves further study.

In addition to nuclear-localized Ca²⁺ channels and components of the nuclear pore complex (NUP85, NUP133), other cation channels are also required for nuclear Ca²⁺ oscillations. Medicago truncatula DMI1 and its two homologs CASTOR and POLLUX in L. japonicus were once thought to be potassium (K⁺) channels (Peiter et al., 2007; Charpentier et al., 2016) but more recent evidence shows that they were Ca²⁺ channels (Kim et al., 2019). Ca²⁺ binding to the CASTOR gating ring was required for root nodule symbiosis, and legumes carrying mutated CASTOR at either of two Ca²⁺ binding sites (D442A or E493Q) failed to form rhizobia-induced nodulation. This finding links defects in Ca²⁺ binding to Ca²⁺ channel regulation, which ultimately affects the legume-microbe symbiosis (Kim et al., 2019). However, this study was carried out in mammalian cells (HEK293), and to further characterize the function of DMI1 and clarify if it is a Ca²⁺ or K⁺ channel, future research should be performed in plant cells. Furthermore, Nod factor induced the association between C-terminal of DMI1 and N-terminal of CNGC15s. In addition, DMI1 associated with CNGC15s (CNGC15a, CNGC15b, CNGC15c) to form a complex protein in nuclear membranes which was required for the activation of nuclear Ca²⁺ spiking (Charpentier et al., 2016). The latest study further confirmed that the two cation channels, DMI1 and CNGC15, form a channel complex to regulate nuclear symbiotic Ca²⁺ oscillations and nodule development (Liu et al., 2022). Genetic testing showed that gain-of-function mutations in MtDMI1, DMI1 (S760N), displayed spontaneous nuclear Ca²⁺ spikes and constitutive activation of nodulation (Liu et al., 2022). The S760N mutation DMI1 caused nuclear Ca²⁺ oscillations in a CNGC15 dependent manner and spontaneous nodulation (Liu et al., 2022). These studies extend our understanding of activating cation channels complex to form nuclear Ca²⁺ oscillations during plant symbiotic microbe interaction.

Ca²⁺ pumps regulate Ca²⁺ changes in the nuclear region, like nuclear localized Ca²⁺ channels, during plant and symbiotic microbe interactions. Other than depending on the ion concentration or electrochemical gradient like Ca²⁺ channels, Ca²⁺ pumps consumed ATP to facilitate Ca²⁺ movement against it (Demidchik et al., 2018). A Ca²⁺ pump, MtMCA8, was involved in the formation of symbiosis-induced nuclear Ca²⁺ oscillation, and MCA8-silenced plants displayed decreased mycorrhizal colonization (Capoen et al., 2011). Unlike DMI1 being mainly distributed in the inner layer of nuclear membrane, MCA8 was equally localized at both inner and outer layers of the nuclear membrane and at the endoplasmic reticulum (ER). A hypothesis proposes that the inner-layer-localized MCA8 mediates the recapture of nuclear Ca²⁺ spikes, while the outer-layer- and the ER-localized MCA8 may reload the Ca²⁺ store at the ER or nuclear envelope from the cytoplasm (Capoen et al., 2011; Tian et al., 2020).

A key component of calcium regulation during plant symbiosis is the small guanosine triphosphatase (GTPase). The GTPase, belongs to the Rho/Rop family, directly regulates reactive oxygen species (ROS) production through activating the respiratory burst oxidase homolog B (RBOHB) (Wang et al., 2020). The ROS activates Ca²⁺ channels and triggers Ca²⁺ influxes, which subsequently activates plant immune responses (Wang et al., 2020), which suggests the GTPases play a role in plant defense (Rivero et al., 2019). Interestingly, the small GTPases [Rho-like GTPas (MtROPs)] and heterotrimeric G-proteins including Gα, Gβ, and Gγ subunits are also involved in root nodule symbiosis (Ke et al., 2012; Pandey, 2019; Bovin et al., 2021): the expressions of MtROP3, MtROP5 and MtROP6 were induced in rhizobia-infected roots (Liu et al., 2010); and genetic tests indicated that the Gα repressed nodule development, while the Gβ, Gγ and RGS promoted nodule development (Choudhury and Pandey, 2013). Further study revealed that ROP6 interacted with NFR5, but not with NFR1, to positively regulate infection thread development and nodulation formation in soybean (Ke et al., 2012). Another study revealed that ROP9 interacted with RACK1 and regulated root nodule development (Gao et al., 2021). Active NFR1 phosphorylated the regulator of G-protein signaling (RGS) proteins, which deactivated Gα, a negative regulator of root nodulation (Choudhury and Pandey, 2015). More studies are needed to reveal how small GTPases, together with Nod factor receptor complex, regulate symbiotic cytoplasmic and/or nuclear Ca²⁺ spiking.

In 1995, the Poovaiah laboratory cloned and characterized a novel protein kinase from lily, which turned out to be regulated by both Ca²⁺ and CaM. Hence, it was named Ca²⁺/CaM-dependent protein kinase (CCaMK; Patil et al., 1995). Unlike all the other Ca²⁺/CaM-dependent protein kinases (CaMKs) which were discovered in animal cells, the CCaMK reported from plants contained a C-terminal visinin-like domain, including three EF-hand motifs, which functioned as a Ca²⁺-sensitive molecular switch (Sathyanarayanan et al., 2001). The CCaMK involved in symbiosis is encoded by DMI3 in Medicago truncatula and LjCCaMK in L. japonicus (Patil et al., 1995). CCaMK is essential for root nodule formation and mycorrhizal associations. CCaMK has a serine/threonine kinase domain at the N-terminal and two Ca²⁺-mediated regulatory domains at the C-terminal. The two C-terminal domains include a visinin-like domain with three EF-hand motifs (i.e., identified as three Ca²⁺-binding domains) and one CaM-binding domain with autoinhibitory function (Sathyanarayanan et al., 2001; Yuan et al., 2017). Plants carrying the mutated CCaMK lacking the autoinhibitory domain exhibited spontaneous nodulation even without rhizobia infection (Gleason et al., 2006). This data indicates that the legume CCaMK is a master controller of nodulation and its autoinhibitory domain is important in regulating its activity. Further studies have shown that when basal levels of Ca²⁺ bind to CCaMK, the protein is kept in an inactive state. However, at elevated Ca²⁺ concentrations (e.g., Ca²⁺ spiking), Ca²⁺/CaM also binds to CCaMK, and the protein becomes activated (Miller et al., 2013). Thus, CCaMK is kept in an inactive state when there are no symbiotic microbes present. However, once symbiosis signals are perceived, Ca²⁺ spiking is induced, and the inactive CCaMK state is overridden by higher levels of Ca²⁺ and CaM binding (Miller et al., 2013).

Further studies in the Poovaiah laboratory and others revealed that a mutated CCaMK negatively affects root nodule symbiosis in Medicago truncatula (Sinharoy et al., 2009; Jauregui et al., 2017). Site-directed mutations in the CaM-binding domain of CCaMK altered its binding capacity to CaM, providing an effective approach to study how CaM regulates CCaMK during rhizobial symbiosis in Medicago truncatula. Mutating the tryptophan at position 342 to phenylalanine (W342F) increased the CaM-binding capability of the mutant, which underwent autophosphorylation and catalyzed substrate phosphorylation in the absence of Ca²⁺ and CaM. When the mutant W342F was expressed in ccamk-1 roots, the transgenic roots exhibited an altered nodulation phenotype. These results suggest that altering the CaM-binding domain of CCaMK could generate a constitutively activated kinase with a negative role in the physiological function of the CCaMK [(Jauregui et al., 2017) ( Figure 2 )].

The CCaMK phosphorylated symbiosis-related substrate has been identified, CYCLOPS in Lotus and interacting protein of DMI3 (IPD3) in Medicago truncatula (Lévy et al., 2004; Messinese et al., 2007; Yano et al., 2008). As with auto-active CCaMK, auto-active CYCLOPs causes spontaneous nodulation in the absence of rhizobia (Gleason et al., 2006; Hayashi et al., 2010; Miller et al., 2013; Singh et al., 2014). CCaMK phosphorylates CYCLOPS/IPD3 to form a complex that binds to promoter elements and induces the expression of symbiosis-involved genes (Yano et al., 2008; Singh et al., 2014). For example, in Lotus, CYCLOPS works with CCaMK and a DELLA transcription factor to regulate the expression of reduced (or required) arbuscular mycorrhiza1 (RAM1) (Gobbato et al., 2012; Pimprikar et al., 2016). RAM1 is a GRAS transcription factor which, when expressed, initiates the colonization of plant roots by arbuscular mycorrhiza (Gobbato et al., 2012). During Nod-factor signaling, CCaMK/IPD3 forms large complexes with two GRAS proteins, nodulation signaling pathway1 (NSP1) and NSP2, in addition to DELLA proteins. The DELLA proteins work as scaffolding to link the CCaMK-IPD3 complex with the NSP1-NSP2 complex, resulting in a complicated unit that regulates symbiotic signaling (Fonouni-Farde et al., 2016; Jin et al., 2016). This unit also activates the expression of two downstream transcription factors, NIN and ERN1 (ERF, required for nodulation 1) (Marsh et al., 2007). Furthermore, ERN1 and/or ERN2 regulate the expression of rhizobium-directed polar growth (RPG), cystathionine β-synthase like 1 (CBS1), nodule pectate lyase (NPL), and nuclear factor YA 1 (NF-YA1), while NIN regulates the expression of early nodulin 11 (ENOD11) and ENOD12 (Fonouni-Farde et al., 2016).

Another reported interactor of CCaMK is the Calf intestinal phosphatase 73 (CIP73) (Kang et al., 2011). CIP73 belongs to a large ubiquitin super family, and it contains a Scythe N ubiquitin-like domain. A report showed that CIP73 interacted with CCaMK in a Ca²⁺-independent manner (Kang et al., 2011). However, CIP73 is phosphorylated by CCaMK in a Ca²⁺/CaM-dependent manner (Kang et al., 2011). The cip73 silencing mutants displayed significantly reduced nodulation as compared to the wild-type control, indicating that it has a role in nodule formation (Kang et al., 2011). Notably, due to CIP73 having a scythe-N ubiquitin-like domain, it may be interesting to study whether the 26S proteasome mediates rhizobial/AM fungal infections. Known components of the Ca²⁺-mediated local symbiotic pathway in roots is described in Figures 1 and 2 .

To activate nodulation or arbuscule formation, the CaM-CCaMK-DELLA-CYCLOPS protein complex binds to the promoter of symbiosis-related genes (Pimprikar et al., 2016). The DELLAs are a key scaffold protein for symbiosis, but they are also critical transcription factors that regulate phytohormone signaling. Therefore, DELLAs may be the link between hormone signals and symbiosis (Liu et al., 2018). Studies about phytohormones involved in symbiosis focus on gibberellic acid (GA), auxin, cytokinin, and abscisic acid (ABA).

NFs-triggered the activation of nodulation requires an optimal level of GAs and exogenous high concentration (>0.01 μM) GA treatments inhibit AM and rhizobial symbioses (Ferguson et al., 2005; Ferguson et al., 2011). One hypothesis is that GA negatively regulates plant symbiosis through the disruption of the CCaMK-DELLA-CYCLOPS complex. This disruption occurs when GA-receptor GID1 (GA INSENSITIVE DWARF1) perceives GA, and then interacts with DELLA proteins (Nemoto et al., 2017). The GA-GID1-DELLA complex recruits a specific F-box protein that interacts with the SCF E3 ligase complex, resulting in the 26S proteasome-mediated ubiquitination and degradation of DELLA proteins. Recruiting E3 ligases to DELLAs (i.e., part of the CaM-CCaMK-DELLA-CYCLOPS complex) may lead to the degradation of the entire complex (Wang and Deng, 2011; Kudla et al., 2018).

Auxin has a positive role in nodulation (Suzaki et al., 2013; Breakspear et al., 2014; Bensmihen, 2015), and nodule numbers are regulated by shoot-to-root auxin transport (van Noorden et al., 2006). Auxin also seems to be positively involved in AM symbiosis (Hanlon and Coenen, 2011; Etemadi et al., 2014). However, a separate study revealed that indole-3-acetic acid (IAA), a class of auxin, promoted GA1 accumulation in pea (O’Neill and Ross, 2002); further studies are needed to extend our knowledge about GA and auxin crosstalk during plant symbiotic microbe interaction.

The role of Ca²⁺ in controlling cell division and growth is well recognized (Perris et al., 1968). It is becoming clear that there is also a linkage between cytokinin signaling and Ca²⁺ signaling. Cytokinin is an important hormone involved in symbiotic interactions between Rhizobium bacteria and leguminous plants. This interaction leads to the induction of the nitrogen-fixing nodule. It was proposed that cytokinin was the key differentiation signal for nodule organogenesis (Frugier et al., 2008). It was also proposed that cytokinin is involved in the regulation of NIN (Nodule Inception) expression to initiate nodule organogenesis and other transcriptional regulators through mechanisms operating both locally and systemically (Yano et al., 2008; Liu et al., 2019). Further study revealed that Ca²⁺ signaling involve cytokinin mediated nodule formation through regulating cytokinin biosynthesis (Reid et al., 2016; Reid et al., 2017). Nod factor induced cytokinin biosynthesis genes expression, including isopentenyl transferase 2 (LjIPT2) and lonely guy 4 (LjLog4), and CCaMK is required for this induction (Reid et al., 2017), although the underlying mechanism is still unclear.

ABA application can inhibit root nodulation, suggesting that ABA is a negative regulator of rhizobial symbiosis (Suzuki et al., 2004; Ding et al., 2008). Interestingly, arbuscule formation was compromised in an ABA biosynthesis-defective tomato mutant sitiens (Herrera-Medina et al., 2007) and further work in Medicago truncatula supported the idea that some components of ABA signaling were needed for AM symbiosis (Charpentier et al., 2014). Another study revealed that ABA contributed to root symbiosis in a dose-dependent manner: high concentrations of ABA repressed AM colonization, while low ABA (i.e., less than 200 μM) promoted AM development (Liu et al., 2018). ABA works in complex signaling pathways with other hormones, including GA. In fact, the interconnection between ABA and GA is illustrated by ABA negatively regulating GA biosynthesis-related gene expression and positively regulating GA catabolism (Nag et al., 2005; Martín-Rodríguez et al., 2016). Another study revealed that exogenous ABA application enhanced the stability of DELLA protein, even in the presence of GA (Achard et al., 2006). ABA maintains the stability and integrity of DELLAs and low doses of ABA may contribute to its positive impact on AM symbiosis (Bedini et al., 2018). High levels of ABA impair Ca²⁺ oscillations, which negatively affects symbiosis (Charpentier et al., 2014). Further studies could address whether SA and JA are also involved in root symbiosis through stabilizing the DELLA protein, although the underlying molecular mechanism remains unclear (Liu et al., 2018).

CCaMK also has a positive role in ABA-mediated responses (Ni et al., 2019; Chen et al., 2021). Work in rice showed that the type C protein phosphate (PP2C), also known as PP45, negatively affected CCaMK activity by dephosphorylating T263. However, ABA induced H₂O₂ accumulation suppressed the transcriptional expression of PP45 (Ni et al., 2019). Although this work was performed in rice, which does not form symbiotic relationships with rhizobia, it would be interesting to hypothesize that ABA is involved in the mediation of root symbiosis through CCaMK. Further studies are needed to better understand the interaction between CCaMK and ABA and their role in symbiosis.