KAI2 Can Do: Karrikin Receptor Function in Plant Development and Response to Abiotic and Biotic Factors

Abstract

The α/β hydrolase KARRIKIN INSENSITIVE 2 (KAI2) functions as a receptor for a yet undiscovered phytohormone, provisionally termed KAI2 ligand (KL). In addition, it perceives karrikin, a butenolide compound found in the smoke of burnt plant material. KAI2-mediated signaling is involved in regulating seed germination and in shaping seedling and adult plant morphology, both above and below ground. It also governs responses to various abiotic stimuli and stresses and shapes biotic interactions. KAI2-mediated signaling is being linked to an elaborate cross-talk with other phytohormone pathways such as auxin, gibberellin, abscisic acid, ethylene and salicylic acid signaling, in addition to light and nutrient starvation signaling. Further connections will likely be revealed in the future. This article summarizes recent advances in unraveling the function of KAI2-mediated signaling and its interaction with other signaling pathways.

Introduction

Karrikins (KARs) are butenolide compounds that were first identified as active components of smoke from burnt plant material that can stimulate the germination of dormant seeds from fire-following plants (Flematti et al. 2004). Critical for understanding KAR perception was the observation that KARs can trigger seed germination and seedling growth responses also in model plants such as Arabidopsis thaliana, which do usually not experience fire (Nelson et al. 2009). Reverse genetics in Arabidopsis identified KARRIKIN INSENSITIVE 2 (KAI2), encoding an α/β hydrolase protein, as being required for KAR-mediated seed germination (Waters et al. 2012). Subsequently, evidence from in vitro binding assays and co-crystallization of a protein-ligand complex suggested that KAI2 may have the ability to perceive KARs (Guo et al. 2013, Kagiyama et al. 2013). However, circumstantial evidence suggests that KAR may have to be metabolized before it can be perceived (Waters et al. 2015a, Sepulveda et al. 2022). Importantly, developmental phenotypes in the absence of fire and marker gene responses to Arabidopsis extracts suggest that KAI2 perceives an endogenous molecule provisionally termed KAI2 ligand (KL) (Conn and Nelson 2015, Sun et al. 2016) which likely represents a phytohormone. Fourteen years since the first description of smoke-derived KARs and 11 years after the identification of the receptor KAI2, the identity of KL remains elusive.

Signal transduction via KAI2 requires an F-box protein from the Skp, Cullin, F-box (SCF)-type E3 ubiquitin ligase complex called MORE AXILLIARY GROWTH 2 (MAX2) (Waters et al. 2012). Activation of KAI2 in the presence of SCF^MAX2 leads to ubiquitination and subsequent degradation of KAR response–specific members of the SUPPRESSOR OF MAX2-LIKE (SMXL) protein family (SMAX1 and SMXL2) (Carbonnel et al. 2020a, Khosla et al. 2020, Wang et al. 2020, Zheng et al. 2020). SMAX1 was discovered in a suppressor screen for KAR signaling–related max2 phenotypes revealing its role as a negative regulator of KAR signaling (Stanga et al. 2013).

MAX2 is also involved in signal transduction upon perception of carotenoid-derived plant hormones called strigolactones (SLs) (Nelson et al. 2011, Wang et al. 2022), and max2 mutants show both KAR- and SL-related phenotypes. SLs are perceived by the KAI2-related receptor DWARF14 (D14) which together with SCF^MAX2 leads to the degradation of SL response–specific members of the SMXL protein family (SMXL6,7,8 in Arabidopsis and D53 in rice) (reviewed in Kyozuka et al. 2022). The use of max2 to study SL responses has led to the mischaracterization of some of its KAR outputs as SL effects, highlighting the need to use specific kai2 and d14 mutants (Scaffidi et al. 2014, Villaécija-Aguilar et al. 2019). Furthermore, the synthetic SL analog rac-GR24 routinely used to induce SL-dependent responses contains two stereoisomers, of which GR24^5DS specifically activates D14, while the other GR24^ent-5DS activates KAI2 (Scaffidi et al. 2014).

Here, we focus on KAR/KL signaling and review the current knowledge on the biological roles of KAI2 in plant development, stress tolerance and interaction with microorganisms. Additionally, we highlight that KAI2 is an indispensable player in the intricate interlacing of phytohormone signaling pathways.

KAI2 Regulates Multiple Aspects of Plant Growth and Development

Germination

Consistent with the role of KARs in inducing seed germination, mutations in KAI2 result in enhanced primary dormancy of Arabidopsis seeds (Waters et al. 2012). Gibberellins (GAs) and abscisic acid (ABA) are major and antagonistic regulators promoting seed germination and dormancy, respectively (Finch-Savage and Leubner-Metzger 2006). The htl-3 mutant (mutated in KAI2) is impaired in rac-GR24-mediated induction of a key ABA catabolism gene CYP707A and the gene encoding the transcription factor WRKY33, which regulates CYP707A expression, possibly resulting in ABA accumulation and promotion of dormancy (Fig. 1A). However, wrky33 and cyp707a mutants partially retain germination responses to rac-GR24, indicating that these genes are not the sole target of KAI2 signaling (Brun et al. 2019).

The promotion of seed germination upon KAR treatment requires GA biosynthesis. But KAR₁ partially induces germination in the GA signaling mutant sleepy (sly) suggesting only partial dependence on canonical GA signaling (Nelson et al. 2009) (Fig. 1A). Interestingly, Arabidopsis smax1 seeds germinate even in the presence of GA biosynthesis inhibitors, suggesting that SMAX1 may be required for suppression of germination by DELLA proteins, the proteolytic targets (and repressors) of GA signaling (Bunsick et al. 2020).

Hypocotyl growth

KAI2 signaling regulates the inhibition of hypocotyl growth, e.g. during photomorphogenesis. Arabidopsis kai2 seedlings show an elongated hypocotyl compared to the wild type in dark-to-light transitions but not in continuous darkness, as well as a reduced cotyledon size (Sun and Ni 2011, Waters et al. 2012). In turn, smax1 smxl2 double mutants show shorter hypocotyls and larger cotyledons than wild-type plants, similar to the wild type after KAR treatment (Stanga et al. 2016). How KAI2 signaling suppresses hypocotyl elongation exactly, is not yet known. There are first indications that the interaction with other phytohormone signaling pathways plays a role in this process.

KAI2 signaling regulating hypocotyl elongation appears to be partially controlled by light. Seedlings grown in the dark after germination show an increase in KAI2 expression when transferred to red light. This short-term increase depends on HY5, encoding a transcription factor that positively regulates responses to light (Sun and Ni 2011, Waters and Smith 2013). However, wild-type and hy5 seedlings grown in continuous darkness or continuous light display the same KAI2 transcript accumulation, suggesting that HY5 is only required for KAI2 induction during the dark-to-light transition (Waters and Smith 2013). The hypocotyl response to KAR treatment is partially impaired in hy5 mutants (Nelson et al. 2010, Waters and Smith 2013), while kai2 and max2 mutations have additive effects to hy5 (Waters and Smith 2013). This suggests that HY5 regulates pathways that suppress hypocotyl elongation in parallel with KAI2 signaling. Additionally, members of the BBX zinc finger family in Arabidopsis, which interact with HY5 and positively regulate photomorphogenic responses, are partially required for KAI2-mediated hypocotyl reduction (Bursch et al. 2021) (Fig. 1B).

Kim et al. (2022) demonstrated that the dark-to-light transition is accompanied by accumulation of SMAX1 in Arabidopsis seedlings, while in the dark, SMAX1 accumulation is low and decreases with time. This suggests that KAI2 signaling plays a more important role in hypocotyl growth regulation in the light than in the dark. In the light, SMAX1 supports hypocotyl elongation by preventing accumulation of DELLA proteins in hypocotyl cell nuclei (Kim et al. 2022). Light-mediated inhibition of hypocotyl growth is hampered in DELLA-deficient mutants, while light exposure leads to a decrease in endogenous GA levels and subsequent DELLA accumulation (Achard et al. 2007). KAR treatment results in light and KAI2 signaling–dependent accumulation of DELLA proteins in hypocotyl cells resulting in cessation of growth (Kim et al. 2022) (Fig. 1B). Thus, additional environmental stimuli or the composition of the light spectrum may regulate the production of KL and thereby the level of KAI2 activity, SMAX1 degradation, DELLA accumulation and hypocotyl growth. In addition, independent pathways may influence SMAX1 stability and thereby hypocotyl elongation (Khosla et al. 2020).

During the dark-to-light transition, KAI2 also modulates auxin signaling. It likely facilitates the rapid movement of auxin produced at the shoot apex toward the root meristem by fine-tuning the abundance of auxin transport proteins of the PIN family (Hamon-Josse et al. 2022) (Fig. 1B). Reduction in PIN3, PIN4 and PIN7 abundance in the hypocotyl and an increase in PIN1, PIN3 and PIN7 abundance at the root meristem are thought to stall hypocotyl elongation while promoting root growth. This remodeling of PIN abundance is absent from kai2 mutants resulting in auxin build-up in the hypocotyl causing an increase in hypocotyl length.

Interestingly, for kai2a kai2b double mutants of the model legume Lotus japonicus no hypocotyl phenotype has been observed, indicating differences in the role of KAI2 in this process among species (Carbonnel et al. 2020b) In rice, mesocotyl elongation in the dark is controlled by D14L (encoded by the rice ortholog of AtKAI2) and D14 in an additive manner (Gutjahr et al. 2015, Kameoka and Kyozuka 2015, Zheng et al. 2020). Interestingly, mutation of smax1 in rice suppresses the long mesocotyl phenotype of d14 and d14l alike (Choi et al. 2020, Zheng et al. 2020), suggesting that both D14 and D14L target SMAX1 in the context of hypocotyl elongation. Indeed, also in Arabidopsis, D14 can target SMAX1 upon rac-GR24 treatment (Li et al. 2022a). However, not only SMAX1 but also D53 seems capable of promoting rice mesocotyl elongation as the rice d53 mutant, which expresses a degradation-resistant D53 has similarly long mesocotyls as d14 (Zheng et al. 2020). A d53 knockout mutant was to our knowledge not tested. Together, this suggests that SMAX1 can be targeted by both D14L and D14. Furthermore, SMAX1 and D53 may be able to interact with the same target proteins in the context of mesocotyl growth, but it is possible that wild-type D53 does not normally accumulate in the tissues relevant for mesocotyl elongation (d53 knockout mutant analysis is necessary to investigate that).

Leaf and shoot development

Above-ground morphological phenotypes of Arabidopsis kai2 mutants at a later stage of development include elongated petioles and increased blade length and width contrary to the effects of d14 (Sun and Ni 2011, Waters et al. 2012, Soundappan et al. 2015, Bennett et al. 2016). Thus, the ratio of KAI2 and SL signaling potentially regulates mature leaf morphology in Arabidopsis.

Brachypodium distachyon kai2 mutants exhibit longer internodes resulting in an elongated appearance (Meng et al. 2022). Green revolution involved the integration of dwarfing traits in crops which facilitated bearing heavy grain without experiencing yield loss due to lodging. The genes for these traits regulate GA production or sensitivity (Hedden 2003). Given that there seems to be an interaction between KAI2-mediated signaling and GA signaling, it is possible that GA plays a role in promoting internode growth in Brachypodium kai2. Currently, there is no report of this phenotype from other plant species, and no molecular data explaining the phenotype are available.

Root and root hair development

Roots and root hairs help in anchorage and acquisition of water and mineral nutrients from the soil. Negative gravitropism ensures that roots grow straight down but additional directional regulation is necessary to maximize nutrient retrieval while avoiding potential stressors such as obstacles, excess salt or dry soil patches (Su et al. 2017). KAI2 regulates multiple aspects of root system architecture including primary root length, root skewing and waving on agar surfaces, emergence of adventitious and lateral roots, lateral root number and root hair length and density with some variations among species (Swarbreck et al. 2019, 2020, Villaécija-Aguilar et al. 2019, 2022, Carbonnel et al. 2020a, b, Hamon-Josse et al. 2022, Meng et al. 2022).

Arabidopsis kai2 mutants show a decreased root hair length and density (Villaécija-Aguilar et al. 2019). This phenotype is conserved in B. distachyon kai2 and BdKAI2 can restore root hair growth in Arabidopsis kai2 (Meng et al. 2022). Interestingly, root hairs of Lotus japonicus and pea kai2a kai2b double mutants (in genomes of these legumes KAI2 is duplicated) have the same length as wild-type root hairs (Carbonnel et al. 2020a, Guercio et al. 2022). However, smax1 mutants of L. japonicus show an increase in root hair length like the smax1 smxl2 double mutant of Arabidopsis (Villaécija-Aguilar et al. 2019, Carbonnel et al. 2020a). Similarly, Brachypodium kai2 shows an increase in the distance of the emergence of first root hairs from the quiescent center, but this is not observed for the Lotus mutants. However, Lotus smax1 shows a decrease in this distance (Carbonnel et al. 2020a, Meng et al. 2022). The conservation of smax1 root hair phenotypes but the absence of kai2 root hair phenotypes in L. japonicus and pea suggests that in these legumes redundant factors may influence the stability or activity of SMAX1.

Carbonnel et al. (2020a) and Villaécija-Aguilar et al. (2022) described the molecular mechanism of KAI2-mediated-signaling in root system and root hair development. Carbonnel et al. (2020a) demonstrated that in L. japonicus, SMAX1 negatively regulates the expression of 1-AMINOCYCLOPROPANE CARBOXYLIC ACID (ACC) SYNTHASE 7 (ACS7), involved in ethylene biosynthesis, to regulate root system architecture. smax1 mutants produce higher amounts of ethylene than the wild type and exhibit increased expression of ACS7 causing longer root hairs and shorter primary roots. The phenotypes could be fully rescued by treating with the ethylene biosynthesis inhibitor 2-aminoethoxyvinyl glycine and the ethylene signaling inhibitor, AgNO₃ (Schaller and Binder 2017, Carbonnel et al. 2020a). Congruously, the short root hair phenotype of Arabidopsis kai2 and acs7 mutants was rescued by treatment with the ethylene precursor ACC (Villaécija-Aguilar et al. 2022). KAI2 is also required for accumulation of the auxin importer AUXIN TRANSPORTER PROTEIN 1 (AUX1) in the epidermis above the lateral root cap (LRC), the LRC, the root tip and the stele; and of the auxin exporter PIN-FORMED2 (PIN2) in the meristematic zone of the root tip. This seems to optimize the auxin level in epidermal cells above the LRC, promoting root hair elongation. Furthermore, AUX1 and PIN2 are required for KAR- and ACC-mediated root hair elongation, placing them downstream of KAR as well as ethylene signaling. Thus, a KAI2-ethylene-auxin-signaling cascade commands root hair growth in Arabidopsis (Villaécija-Aguilar et al. 2022) (Fig. 1C).

Lateral root formation differs from other root traits in Arabidopsis, as it is regulated by both KAI2 and D14 in an additive manner with both mutants showing increased lateral root density (Villaécija-Aguilar et al. 2019). Formation of junction roots, a type of adventitious roots in Arabidopsis, is also under the control of both KAI2 and D14 signaling but with a stronger impact of KAI2 (Swarbreck et al. 2020). In the grass B. distachyon, mutation of KAI2 alone leads to a dramatic increase in lateral roots (Meng et al. 2022), but the role of D14 has not been investigated. Rice d14l seems to have increased large lateral root density, but the significance of this difference to the wild type was not determined (Chiu et al. 2018). Interestingly, L. japonicus kai2a kai2b mutants do not show an increase in lateral roots (a difference to the wild type only becomes apparent upon karrikin treatment, Carbonnel et al. 2020b) but the smax1 mutant does (Carbonnel et al. 2020a). This, along with differences in root hair phenotypes compared to other plants, suggests diversification of KAI2 function in L. japonicus in defining root system architecture or variation in nutritional optima among species, which would shift the role of KAI2 to either positive or negative for lateral root development.

Role of KAI2 in bryophyte development

KAI2 orthologs can be traced back to charophyte algae (Delaux et al. 2012) and possibly arose via horizontal gene transfer of bacterial RsbQ (Wang et al. 2022). It is not yet clear which function KAI2 performs in algae. By contrast, the role of KAI2 in bryophytes is beginning to be revealed. During the early evolution of land plants, the KAI2 lineage split, giving rise to two superclades, the eu-KAI2 clade containing the characterized KAI2 sequences from angiosperms and homologs from other land plants and the DDK (D14/DLK2/KAI2) clade containing SL receptor D14 homologs, divergent bryophyte KAI2s and D14-LIKE 2 (DLK2) homologs (Bythell-Douglas et al. 2017). In the moss Physcomitrium patens, eu-KAI2 mutants are smaller than the wild type but have bigger gametophores. Additionally, they have longer gametophores when grown under continuous red light pointing to a role of KAI2 in photomorphogenesis (Lopez-Obando et al. 2021). These phenotypes are similar to that of Ppmax2 suggesting MAX2-dependent signal transduction via PpKAI2L (eu-KAI2) (Lopez-Obando et al. 2018). Interestingly, other homologs of PpKAI2L function in a MAX2-independent but PpCCD8-dependent manner, thereby regulating SL-specific responses (Lopez-Obando et al. 2021). This suggests that neo-functionalization of KAI2 to perceive SLs occured independently in mosses (Lopez-Obando et al. 2016).

The liverwort Marchantia polymorpha has two eu-KAI2 orthologs, MpKAI2a and MpKAI2b, and one each of MpMAX2 and MpSMXL (Waters et al. 2012, Mizuno et al. 2021). Phenotypic analysis of Mpkai2 mutants revealed morphological defects including reduced thallus area, increased angle of thallus curving, reduced gemma area and gemma formation and reduced number of cells in gemma. Mutants were also impaired in photomorphogenic responses and showed uninhibited growth in the dark. MpKAI2a and MpKAI2b show possible physical interactions with only GR24^ent-5DS but they do not respond to KARs (with only MpKAI2A seeming to regulate downstream responses) (Mizuno et al. 2021, Komatsu et al. 2023), indicating that KL is different from KARs found in smoke, at least in Marchantia.

Mpsmxl mutation can fully suppress the thallus angle, the gemma cell number and partially the thallus area phenotype. However, it shows the same reduction in the gemma area as Mpkai2a (Mizuno et al. 2021), suggesting that this trait is regulated by KAI2 in an SMXL-independent fashion. In fact, SMXL genes first appeared in mosses and liverworts and later duplicated and diversified in synchrony with the diversification of the receptor proteins (Moturu et al. 2018, Walker et al. 2019). Thus, in charophytes, KAI2 must have functioned independently of SMXLs and it seems that remnant SMXL-independent functions still exist in land plants.

Roles of KAI2 in Abiotic Stress Responses

In recent years, evidence is accumulating that KAI2 signaling not only regulates plant development per se but especially as a reaction to environmental conditions.

Heat tolerance

Global surface temperatures tend to fluctuate due to various factors but there has been an unusual trend toward warming in the last 20 years (Neukom et al. 2019). Plants deploy various mechanisms to evade serious consequences from an increase in ambient temperatures beyond the optimum (Wang et al. 2017, Wen et al. 2019) and display developmental responses to elevated temperatures, a phenomenon called thermomorphogenesis (Delker et al. 2022).

When Arabidopsis seeds are exposed to high ambient temperatures (30°C) for 4 d and transferred back to optimal conditions (20°C), kai2 seeds experience a reduced germination success compared to wild-type seeds and compared to kai2 seeds continuously incubated at 20°C (Wang et al. 2018). Wild-type seeds recover very well after high-temperature stress suggesting a positive role of KAI2 in heat tolerance. Interestingly, treatment with KAR₂ strongly inhibits seed germination in the wild type at higher temperatures. With the role of KAI2 in regulating seed dormancy already established, Wang et al. (2018) argued that in wild-type seeds, KAI2 could promote seed germination in favorable conditions but prevent seed germination in unfavorable conditions. This is conceivable under the assumption that SMAX1-interacting proteins change depending on the environmental condition (e.g. by their availability), and SMAX1 may block promoters of germination under favorable conditions and inhibitors of germination under unfavorable conditions.

KAI2 is also required for high-temperature tolerance post-germination. Ten-day-old Arabidopsis kai2 plants when exposed to 40°C for 5 d show higher mortality rates than wild type. This is accompanied by compromised cell membrane integrity and inability to regulate leaf surface temperature in kai2 (Abdelrahman et al. 2023). Plants tend to close stomata in events of high-temperature stress to not excessively lose water (Marchin et al. 2022). It has previously been shown in Arabidopsis kai2 that stomatal apertures are larger which can possibly exacerbate the effects of heat (Li et al. 2017, Marchin et al. 2022). Transcriptional data analysis revealed lower expression of a battery of genes in kai2 which are involved in plant thermo-tolerance. This included heat-shock protein-related genes like HSP70, HSP90 and HSP101, membrane fluidity thermosensors of CYCLIC NUCLEOTIDE GATED CHANEL gene family and WRKY transcription factors among others (Abdelrahman et al. 2023).

Arabidopsis KAI2-signaling mutants, when exposed to high-temperature stress of 28°C, have impaired thermomorphogenic hypocotyl growth, with kai2 having a longer and smax1 having a shorter hypocotyl compared to the wild type under stress (Park et al. 2022). PHYTOCHROME B (PHYB) has been established as a thermosensor that negatively regulates PHYTOCHROME-INTERACTING FACTOR 4 (PIF4) activity in promoting hypocotyl growth in response to high-temperature stress (Jung et al. 2016). Interestingly, the expression of a constitutively active phyB in kai2 or max2 background reduces thermomorphogenic hypocotyl growth, while inactive phyB is epistatic to smax1 (Park et al. 2022). Furthermore, Arabidopsis pif4 can suppress hypocotyl growth in kai2, whereas supplementary expression of the wild type version can rescue the attenuated response of smax1. This suggests that the PHYB-PIF4 module acts downstream to KAI2 signaling (Fig. 1D).

Protein–protein interaction assays in yeast and in planta show physical interactions between SMAX1 and PHYB, but the exact mechanism of how SMAX1 could regulate PHYB action is not yet clear (Park et al. 2022). The stability of both PHYB and PIF4 is not altered in smax1; however, the expression of PIF4-regulated YUCCA8 (YUC8, an auxin biosynthetic gene induced by high-temperature stress; Sun et al. 2012) is significantly reduced. Together with enhanced YUC8 induction in kai2 and reduction in smax1, this points to SMAX1 somehow counteracting PHYB inhibition of PIF4, probably via another protein (Park et al. 2022). Unraveling the identity of this protein could be interesting for future research.

Shade avoidance

Dense canopies raise the need for plants growing underneath to exercise adaptive responses to maximize light capture. The filtering of light by canopy leaves results in a decrease in the ratio of red to far-red wavelengths in the light reaching lower leaves and plants in the understorey. This triggers morphological adaptations cumulatively called shade avoidance responses (Tao et al. 2008, Hersch et al. 2014). Shade-intolerant plants exhibit elongation of hypocotyls, petioles and stems and an impediment in leaf development (Franklin 2008, Ciolfi et al. 2013). These morphological changes can come at a cost of yield which makes shade avoidance undesirable for densely sown crops.

KAI2 signaling negatively regulates the shade avoidance response in Arabidopsis. In fact, kai2, max2 and lines overexpressing SMAX1 show significantly increased hypocotyl growth, leaf area reduction and petiole length under shading, whereas lines overexpressing KAI2 and MAX2, KAR2-treated Col-0 plants and smax1 mutants are inhibited in responding to simulated shade. d14, max3 and max4 mutants show no change compared to the wild type, suggesting that KAI2 signaling but not D14 signaling contributes to regulating shade avoidance response (Xu et al. 2022).

In agreement with KAI2 regulating auxin abundance in hypocotyls during the dark-to-light transition (Hamon-Josse et al. 2022), hypocotyl growth in response to shade has been linked to over-accumulation of auxin as a result of increased abundance of auxin transport proteins PIN3 and PIN7 also in shade avoidance (Xu et al. 2022) (Fig. 1E). This suggests that KAI2 signaling may generally regulate auxin distribution and signaling in several responses to environmental conditions.

Response to drought

Climate change and shortage of fresh water to irrigate farms often result in farms plagued by drought. Thus, unraveling the molecular mechanisms underlying tolerance has become very important to enable breeding drought-resistant crop varieties (Hu and Xiong 2014).

KAI2-mediated signaling plays a role in drought stress resistance (Fig. 1F). Initially, this has been shown for max2 mutants (Bu et al. 2013, Ha et al. 2014), but the link with either KL or SL signaling was not clear. Li et al. (2017) showed that the Arabidopsis kai2 mutant is hypersensitive to drought. It displayed increased stomatal opening and cell membrane damage resulting in increased water loss upon drought. In addition, kai2 was found to be less sensitive to ABA, an important hormone regulating drought tolerance (Osakabe et al. 2014, Li et al. 2017). Additionally, kai2 showed impaired accumulation of anthocyanins and a reduced leaf cuticle which contributes to drought hypersensitivity. Some of these phenotypes were co-regulated by D14 albeit with varying levels of D14 contribution (Li et al. 2017, 2020). Consistent with a major contribution of KAI2, Arabidopsis smax1smxl2 mutants show increased resistance to drought (Feng et al. 2023). The double mutant exhibits slower leaf water loss, most likely a result of decreased leaf cuticular permeability and stomatal aperture and increased ABA responsiveness. Additionally, smax1smxl2 accumulates less reactive oxygen species (ROS), possibly from increased GLUTATHIONE PEROXIDASE3 (GPX3) and GPX7 expression. Furthermore, increased root hair length and root-to-shoot ratio in smax1smxl2 compared to the wild type demonstrate a stronger drought tolerance response (Carbonnel et al. 2020a, Feng et al. 2023). In B. distachyon however, Bdkai2 experienced only a minor increase in water loss compared to the wild type, but the wild type BdKAI2 could complement the Atkai2 water loss phenotype depending on the expression levels of the transgene (Meng et al. 2022). Thus, the need for KAI2 in drought stress response may differ among species.

Osmotic stress tolerance

Similar to high-temperature stress, Arabidopsis kai2 seeds germinate less under osmotic stress compared to the wild type, whereas germination of wild-type seeds subjected to osmotic stress is inhibited by KAR₂ treatment (Wang et al. 2018). This suggests that KAI2 also functions in osmotic stress tolerance confirming its role in mediating germination only under favorable conditions (Fig. 1G). GA treatment can rescue seed germination in kai2 under osmotic stress (Wang et al. 2018) but whether it can overpower the inhibitory effects of KAR₂ has not yet been assessed.

KAI2 continues to be important also post-germination for tolerance to osmotic stress. kai2 plants exhibit increased growth impairment, decreased survival rates and reduced biomass compared to wild-type Arabidopsis under salt stress (Mostofa et al. 2023). How KAI2 enhances the chances of survival under these conditions has been investigated by Mostofa et al. (2023). They demonstrate that KAI2 improves the ability of the plant to regulate shoot Na⁺ homeostasis, possibly by positive regulation of ion transporter genes such as SALT OVERLY SENSITIVE and HIGH-AFFINITY POTASSIUM TRANSPORTER 1;1. Furthermore, wild-type plants exhibit less oxidative stress–induced damage compared to kai2. This can result from higher expression and stronger activity of antioxidant enzymes like superoxide dismutase, catalase and ascorbate peroxidase among others in the wild type compared to kai2. In addition, genes involved in SL, ABA, jasmonic acid and salicylic acid biosynthesis and signaling are induced under salt stress in the wild type, whereas the induction is hampered in kai2 (Mostofa et al. 2023). This KAI2-dependent activation of other phytohormone pathways enhances the arsenal available to survive stressful conditions.

In the future, it will be interesting to decipher how the perception of abiotic stress activates KAI2 signaling. Since KAR treatment can enhance tolerance to various stresses (Wang et al. 2018, Shah et al. 2020, 2021), it is possible that stress signaling regulates KL biosynthesis.

Phosphate starvation responses

Phosphorus is a vital macronutrient for plant growth and development. The slow release of inorganic phosphate (P_i) from its soil reservoirs makes it poorly available to plants (Shen et al. 2011). One of the ways plants respond to phosphate shortage is by increasing root hair length allowing more efficient foraging of the accessible soil volume (Lynch 2011). KAI2 contributes to the root hair elongation response of Arabidopsis to low P_i (Villaécija-Aguilar et al. 2022). Mutation of kai2 and max2 results in a dampened root hair elongation response to reduced P_i concentration in growth media, while smax1 smxl2 mutants show increased root hair length even at high phosphate. kai2, max2 and smax1 smxl2 roots can still sense phosphate starvation, evident from induced expression of P_i starvation marker genes PHOSPHATE TRANSPORTER 1;4 and INDUCED BY PHOSPHATE STARVATION 1, and a slight root hair growth response to variations in P_i concentrations, which is though much weaker than in the wild type. Additionally, P_i starvation induced the expression of KAI2, MAX2 and DLK2 in a PHOSPHATE STARVATION RESPONSE 1 (PHR1)- and PHR1-LIKE (PHL1)-dependent manner (both being central transcription factors in phosphate starvation signaling; Paries and Gutjahr 2023) (Fig. 1C), suggesting that phosphate signaling regulates the abundance of KL receptor. Thus, KAI2 is essential for root hair growth responses to phosphate starvation stress. It will be exciting to learn in the future, whether KAI2 signaling also participates in regulating deficiency responses to other important nutrients.

Regulation of Biotic Interactions

Letting friends in

A majority of land plants tackle nutrient deficiency by forming mutualistic relationships with Glomeromycotina fungi, termed arbuscular mycorrhiza (AM). The fungi colonize the root cortical cells forming highly branched hyphal structures, the arbuscules, which are sites for the transfer of nutrients to the plant such as phosphorus and nitrogen. In return, the fungus receives carbohydrates and lipids (Keymer and Gutjahr 2018, Wipf et al. 2019). Each step in root colonization is precisely regulated at the cellular and molecular level (Gutjahr and Parniske 2013).

KAI2 plays an important role for colonization of rice roots. Deletion or mutation of D14L (rice ortholog of Arabidopsis KAI2), results in inability of AM fungi to penetrate the root epidermis (Gutjahr et al. 2015; Choi et al. 2020) (Fig. 1I). Osd14l is impaired in gene expression responses to germinating spore exudates, suggesting that OsD14L functions upstream of or in perception of fungal signaling molecules, or in transcriptional regulation of the response to these. A similar phenotype has been reported for Petunia kai2a, Brachypodium kai2, Barley d14l and Medicago kai2a kai2b (Liu et al. 2019, Meng et al. 2022, Li et al. 2022b). Consistently, rice and Medicago smax1 exhibit a significant increase in colonization levels (Choi et al. 2020, Li et al. 2022b). However, KAI2 is not required for AM development in Marchantia paleacea (Kodama et al. 2022), suggesting that KAR signaling may have been wired to AM signaling later in evolution in the vascular plant lineage.

Although no direct targets of SMAX1 acting in AM development have yet been determined, Choi et al. (2020) and Li et al. (2022b) place KAI2 signaling upstream of the regulatory network promoting AM development. Choi et al. (2020) reported transcriptional upregulation of SL biosynthesis pathway genes in non-inoculated smax1 roots. This included GRAS transcription factor NODULATION SIGNALING PATHWAY 2 (NSP2), known to be required for the expression of SL biosynthesis genes (Liu et al. 2011), beta carotene isomerase DWARF 27 (D27), CAROTENOID CLEAVAGE DIOXYGENASE 7 and 8 (CCD7 and CCD8), which are required for the production of SL precursors, and CYP711A2, which is involved in the production of orobanchol (Zhang et al. 2014). An increase in the production of a rice canonical SL 4-deoxyorobanchol could also be observed in the roots. If suggestions of a possible negative impact of SL sensing on AM colonization are true (Yoshida et al. 2012, Gutjahr et al. 2015, Li et al. 2022b), then a negative feedback loop could be assumed, which could be addressed by observing colonization levels in smax1 d53 double mutants. Nonetheless, smax1 and d53 phenotypes are largely mutually exclusive (Soundappan et al. 2015).

The list of genes upregulated in the smax1 mutant also contains regulators of root colonization which are part of the so-called common symbiosis signaling network, which also operates in root nodule symbioses (RNSs), such as SYMBIOSIS RECEPTOR-LIKE KINASE (SYMRK), Ca²⁺/CALMODULIN-DEPENDENT PROTEIN KINASE (CCaMK) and CYCLOPS (Oldroyd 2013), and receptor-like kinases, such as LysM-RECEPTOR-LIKE KINASE 2 (LysM-RLK2) [NOD FACTOR RECEPTOR 5 (NFR5)], which may be involved in the perception of fungal lipochitooligosaccharide (LCO) signals (He et al. 2019) (Fig. 1I). The perception of fungal LCOs by plants triggers a characteristic calcium spiking signal. This signal initiates a cellular development program in root cells to accommodate AM fungi (Gutjahr and Parniske 2013, Feng et al. 2019). Li et al. (2022b) observed that this spiking was abolished in the rice d14l mutant. Furthermore, barley d14l was impaired in the induction of RECEPTOR-LIKE KINASE 10 (RLK10) (barley homolog of NFR5) and KAR treatment could induce RLK10 in a D14L-dependent manner (Li et al. 2022b). Together with the reduced SL exudation, which is required in normal amounts to activate the fungus, this may explain the impaired root colonization of kai2 mutants.

Consistent with the role of common symbiosis signaling in RNS and the role of SMAX1 in suppressing the expression of common symbiosis genes, the symbiosis between nitrogen-fixing rhizobia and legumes may also be affected by KAI2 signaling. RNAi of KAI2 and MAX2 in soybean led to a slight reduction in nodule formation; however, this was also the case for D14 knockdown, although to a lesser extent (Ahmad et al. 2020). Knockout mutants will be required to clearly define the role of KAI2- and D14-mediated signaling in RNS development.

Keeping foes out

Phytohormone signaling and cross-talk thereof play essential roles in mediating plant responses to pathogen attacks and KAI2-mediated signaling is no exception. Arabidopsis kai2 and max2 mutants are more susceptible to the hemibiotrophic bacterium Pseudomonas syringae pv. tomato (Pst), whereas surprisingly smax1 as well as smxl6/7/8 can suppress the susceptibility phenotype of max2. However, the resistance phenotype is specific to KAI2 signaling, and the authors suggest that SMXL6/7/8 are somehow required for SMAX1 function (Zheng et al. 2023). The susceptibility phenotype of kai2 and max2 was attributed to larger stomatal openings and an impaired ability to close stomata in response to pathogen attack leading to unconstrained bacterial entry into the apoplast (Piisilä et al. 2015, Zheng et al. 2023) (Fig. 1H).

Beyond the physical barrier, max2 plants are compromised also at a molecular level. Transcriptional analysis of the mutant showed reduced expression of genes involved in salicylic acid–dependent resistance such as ISOCHORISMATE SYNTHASE 1 (ICS1), SYSTEMIC ACQUIRED RESISTANCE DEFICIENT 1 (SARD1), NON-RACE-SPECIFIC DISEASE RESISTANCE 1 (NDR1), PHYTOALEXIN DEFICIENT 4 (PAD4), TGA1A‐RELATED GENE 3 (TGA3) and NAC TRANSCRIPTION FACTOR-LIKE 9 (NTL9) to list a few in the initial stages of the infection, and this was rescued by smax1 (Zheng et al. 2023) (Fig. 1H). SA-dependent defense response is essential for the establishment of local and systemic resistance primarily against biotrophic and hemibiotrophic pathogens (Spoel et al. 2007). SA treatment can enhance disease resistance in max2 but not to the level of the wild type, suggesting that SA acts downstream to KAI2 signaling (Zheng et al. 2023). Still, no direct link between the two pathways is yet known and will be interesting to pinpoint in future research.

While KAI2 signaling acts against P. syringae in Arabidopsis, it does not appear to contribute to resistance against fungal pathogens Magnaporthe oryzae (in rice roots) and Botrytis cinerea (in Arabidopsis leaves) (Gutjahr et al. 2015, Piisilä et al. 2015). Perhaps its role is limited to defense against pathogens that enter the plant through stomata. Further research with a range of pathosystems will reveal the breadth of KAI2 functions in pathogen defense.

Conclusion and Outlook

Since the discovery of the KAR receptor KAI2 (Waters et al. 2012), our knowledge of KAI2-mediated signaling has rapidly expanded. It has been firmly established that KAI2 signaling is required for normal plant growth and development and plays a central role in responses to abiotic conditions and interactions with microbes, especially with AM fungi. Moreover, first evidence for KAI2 signaling and cross-talk with other phytohormone signaling pathways is emerging. However, equally large gaps in our knowledge remain yet to be filled.

We still lack knowledge on the identity of KL and enzymes involved in its biosynthesis. Targeted in silico modeling led to a sesquiterpene lactone (Rahimi and Bouwmeester 2021), while experimental evidence suggests that the structure of KL might contain a hydrolyzable desmethyl-butenolide ring, without a 4ʹmethyl group (Yao et al. 2021). Desmethyl-GR24^ent-5DS (dGR24^ent-5DS) activates KAI2 signaling more strongly in Arabidopsis, B. distachyon and M. polymorpha as compared to GR24^ent-5DS and does not induce signaling via D14 (Yao et al. 2021). Additionally, Selaginella moellendorffii KAI2 binds dGR24^ent-5DS in vitro and shows activation in a heterologous system, which otherwise does not occur upon either KAR or GR24^ent-5DS treatment (Waters et al. 2015b, Yao et al. 2021, Meng et al. 2022).

It has been speculated that NSP transcription factors may regulate the KL biosynthesis pathway (Li et al. 2022b). They are required for the expression of genes involved in apocarotenoid biosynthesis. Additionally, the promotion of AM colonization of M. truncatula by ectopic expression of NSP2 requires KAI2, and inhibition of colonization in nsp2 nsp2l double mutants is suppressed by smax1 mutation (Li et al. 2022b). It will be interesting to learn in the future if KL is really an apocarotenoid or derived from another biosynthetic pathway.

Despite some indications of KAI2-signaling interactions with other phytohormone pathways, we certainly do not know the whole range of interactions and in most cases, direct molecular connections to these other pathways remain to be uncovered. Particularly, we are missing knowledge about SMAX1 targets. These may differ depending on the scenario regulated by SMAX1. SMAX1 interactomes in different tissues and under different conditions may lead us to these targets. Major exciting questions remain to be answered.

Data Availability

No new datasets were generated or analyzed in this study.

Funding

Transregio Collaborative Research Center 356 ‘Genetic diversity shaping biotic interactions of plants’ (491090170) of the Deutsche Forschungsgemeinschaft (DFG); Max Planck Society.

Author Contributions

K.V. prepared the figure with input from C.G. K.V. and C.G. wrote the review text.

Disclosures

The authors have no conflicts of interest to declare.

References