Introduction

Stomatal pores control the flow of gases and water vapor between plants and the environment. Each stomatal pore is encircled by two guard cells on aerial epidermal surfaces of nearly all land plants. The fossil record indicates that stomata are both ancient and are a significantly conserved feature of terrestrial plants (Beerling and Royer 2002). As plants transitioned from aquatic to terrestrial habitats, they encountered a range of environmental challenges, including desiccation (Waters 2003). In response, plants evolved a hydrophobic layer on the surface of the epidermis known as the cuticle to reduce water loss (Kong et al. 2020). Such cuticular surfaces limit carbon dioxide (CO 2 ) entry and water vapor loss, which led to the necessity for stomata in land plants (Riederer andMuller 2008, Clark et al. 2022).

Stomata respond quickly to changes in CO 2 concentrations in leaf intercellular air spaces. High CO 2 triggers stomatal closing, and low CO 2 induces stomatal opening (Heath 1948). The CO 2 concentration inside leaves (c i ) shifts rapidly on a daily basis, e.g. upon cloud cover or darkness when respiration and reduced photosynthesis quickly cause a doubling or more in c i (Hanstein et al. 2001, Roelfsema et al. 2002). Since the beginning of the Industrial Revolution, atmospheric CO 2 concentrations have shown a steep rise. Global atmospheric CO 2 has increased from 315 ppm to more than 420 ppm since 1958, according to recordings at the Mauna Loa Volcanic Observatory. The CO 2 concentration is predicted to be 800 ppm or higher by the end of this century. This increase in ambient CO 2 concentrations is expected to reduce stomatal pore apertures and stomatal density in many plant species (Medlyn et al. 2001, Ainsworth andRogers 2007). Stomata play physiologically important roles in various aspects of plant life, including the exchange of CO 2 and water vapor and facilitating transpiration, which can facilitate the uptake of water and nutrients from the soil (Hetherington and Woodward 2003). Furthermore, stomatal transpiration reduces leaf temperatures. Therefore, understanding stomatal responses to CO 2 concentrations has been one of the critical research areas concerning the impacts of frequently occurring weather extremes on plant growth and crop yield.

As autotrophic organisms, plants survive and grow due to photosynthesis. The biochemical pathways of photosynthesis to harness light energy for the conversion of CO 2 into sugars have been preserved through evolutionary transitions from cyanobacteria to chloroplasts in higher plants (Vermaas 2001). Yet photosynthesis has undergone significant evolutionary changes over an extended period of time, resulting in highly intricate and sophisticated mechanisms. Given that CO 2 is required in the process of carbon assimilation, the regulatory mechanisms that control stomata, i.e. the small openings on the surface of plants through which gases exchange between the atmosphere and the leaves, are inextricably linked to the process of photosynthesis.

In this review, we discuss the evolutionary and physiological aspects of stomatal movements in response to changes in CO 2 concentration and review the advances in research on stomatal CO 2 responses. Given the recent discovery of the stomatal primary CO 2 /HCO 3 -sensor in Arabidopsis, we describe the molecular mechanisms of how changes in CO 2 concentration can be perceived by stomatal guard cells and induce stomatal movements.

Evolutionary Aspects of Stomatal Function Stomata in non-vascular and vascular land plants

After the appearance of plants with stomata, plant lineages diverged into non-vascular plants such as bryophytes and vascular plants with roots and vascular tissue (Berry et al. 2010, Clark et al. 2022). As a non-vascular land plant, the moss (Physcomitrium patens) belongs to one of the most primordial diverging lineages that possess stomata (Caine et al. 2016). In P. patens, stomata are localized to the capsule on the sporophyte (Chater et al. 2016). It has been known that moss stomata play a critical role in the desiccation of the capsule, which leads to the dispersal of spores (Duckett et al. 2009, Chater et al. 2016). According to recent studies, the moss P. patens shares a common ancestral molecular machinery for stomatal development with Arabidopsis (Caine et al. 2016, Chater et al. 2016). Specifically, the presence of orthologs PpSMF1 and PpSCRM1 are necessary for moss stomata development, paralleling Arabidopsis regulatory processes (Pillitteri and Torii 2012, Chater et al. 2016, Simmons and Bergmann 2016). More recent studies have shown that moss stomata regulate CO 2 assimilation and water loss (Kubásek et al. 2021). Regarding the evolution of stomatal opening and closing, there are conflicting views on stomatal function. The phytohormone abscisic acid (ABA) induces stomatal closure in response to drought (Cutler et al. 2010). Brodribb and McAdam (2011) argued that bryophyte, lycophyte, and fern stomata lack ABA and elevated CO 2 responses. They proposed that stomatal responses in these groups occur solely via "hydropassive" mechanisms and that receptor/sensor-mediated "hydroactive" signaling mechanisms emerged only with gymnosperms and angiosperms (Brodribb andMcAdam 2011, McAdam andBrodribb 2012). Conversely, other studies suggest that early-developing lineages utilized additional hydroactive mechanisms for stomatal responsesopening and closing in response to ABA, CO 2 , and light, suggesting their presence in early land plants (Chater et al. 2011, Ruszala et al. 2011, Clark et al. 2022). Research on stomatal responses in intact fern plants has shown unequivocal CO 2 and ABA responses in some fern species depending on conditions, but a lack of or weak responses in other fern species (Horak et al. 2017). In seed plants, the presence of active regulators of stomatal responses has been substantiated by various findings (Tardieu and Davies 1993, Negi et al. 2008, Vahisalu et al. 2008).

History of atmospheric CO 2 concentration and the emergence of stomata

It has been suggested that a major reduction in atmospheric CO 2 concentration occurred 400 million years ago during the mid-Paleozoic era (Beerling 2005). A substantial reduction in CO 2 levels during the mid-Paleozoic was due to increased silicate weathering, with secondary contributions from enhanced sequestration of organic matter resulting from photosynthesis via the proliferation and dispersal of vascular terrestrial plants (Berner 1991, Berner 1992). The emergence of stomata ∼400 million years ago enabled plants to increase CO 2 diffusion linked to the assimilation of CO 2 necessary for photosynthesis, while simultaneously managing water loss through transpiration (Hetherington and Woodward 2003). This pivotal evolutionary advancement significantly aided the establishment of vascular plants in terrestrial ecosystems (Edwards et al. 1998, Duckett et al. 2009, Chater et al. 2016).

Evolution of CO 2 signaling components

Recent studies have identified MPK4/12 and HIGH LEAF TEMPERATURE 1 (HT1) as primary sensors in CO 2 signaling pathways, with CONVERGENCE OF BLUE LIGHT AND CO 2 1/2 (CBC1/2) acting downstream of HT1 to mediate stomatal responses to CO 2 (see Sections "Identification of key signaling protein kinases" and "HT1-MPK4/12 complex is the primary CO 2 /HCO 3 -sensor" for details; Takahashi et al. 2022). The MAPK gene family is conserved across all major plant lineages. However, MPK4 and MPK12 orthologs, which function in stomatal regulation, appear to have evolved more recently within angiosperms (González-Coronel et al. 2021, Ivsic et al. 2025). The Raf-like MAPKKK genes, HT1 and CBC1/2, have undergone distinct evolutionary processes. Phylogenetic analyses have shown that CBC1/2 are found in both gymnosperms and angiosperms, while CO 2 -sensing HT1 is found only in angiosperms (Ivsic et al. 2025). Given that non-angiosperms show a relatively weak (or absent) stomatal CO 2 response (Klein and Ramon 2019), a comparative phylogenetic analysis of the stomatal CO 2 /HCO 3 -sensing HT1-MPK12/4 module along with downstream signaling components across plant species would provide additional insights on CO 2 sensing and signaling mechanisms in non-angiosperm plants.

Stomatal CO 2 Signaling Mechanisms

Rapid and reversible stomatal movements in response to CO 2 Similar to the stomatal response to light and dark, the opening and closing of stomata in response to changes in CO 2 concentration are rapid and reversible (Fig. 1). In general, stomata open when the CO 2 concentration decreases and close in response Figure 1. Reversible stomatal response to changes in CO 2 concentration. Effect of ambient CO 2 concentration shifts on stomatal conductance of an Arabidopsis leaf was measured using a gas exchange analyzer (LI-6800, LI-COR, Nebraska, USA). The stomatal conductance was stabilized under 400 ppm CO 2 . The external CO 2 concentration was changed to 800 ppm and then returned to 400 ppm repeatedly as indicated above. Stomatal conductance data were collected every minute.

to CO 2 elevation. Although it is unlikely that atmospheric CO 2 concentrations drastically fluctuate throughout a single plant's life (Mott 1990), the CO 2 concentration in plant leaves can dynamically change due to variations in photosynthetic activity and respiration depending on light conditions (Mott 1990, Hanstein et al. 2001, Roelfsema et al. 2002). Stomatal guard cells can sense these changes in CO 2 concentration within the leaf. The rapid and reversible stomatal movements allow plants to control stomatal apertures in response to carbon demand for photosynthesis, which is advantageous in balancing the uptake of CO 2 for photosynthesis and water loss through the stomatal pores. The stomatal closure rate in Arabidopsis in response to high CO 2 concentrations is usually faster than the low CO 2 -induced stomatal opening rate (Fig. 1). Stomatal closure is achieved by rapid passive ion transport through ion channels, resulting in water efflux and a decrease in guard cell volume, whereas stomatal opening requires slower active proton pump-energized ion transport against concentration gradients into the cells and an increase in guard cell volume (Shimazaki et al. 2007). The signal transduction mechanisms underpinning stomatal opening and closing in response to CO 2 concentration must be rapid and reversible.

Identification of key signaling protein kinases

The CO 2 sensing and signal transduction mechanisms had long remained unclear. However, a high-throughput unbiased forward genetic screening system for mutants with impaired stomatal function was established in 2002 by measuring leaf surface temperature using infrared thermography (Merlot et al. 2002), and a pioneering study applied this methodology to stomatal CO 2 response research (Hashimoto et al. 2006). Hashimoto et al. (2006) used thermography to isolate recessive Arabidopsis mutants whose stomata did not open at low CO 2 concentrations and isolated a gene named HT1 (Hashimoto et al. 2006). The identification of the HT1 gene has greatly advanced plant CO 2 research, and the function of HT1 as a direct component of the primary stomatal CO 2 sensor was recently revealed, as discussed below.

The HT1 gene was identified as one of the protein kinase genes belonging to the subgroup C5 of the Arabidopsis Raflike gene family (MAPK Group 2002, Hashimoto et al. 2006). Strong recessive ht1 mutant alleles show a constitutive high CO 2 closed stomatal response and disrupt low CO 2 -induced stomatal opening (Hashimoto et al. 2006). Subsequently, genetic screening led to the identification of dominant HT1 point mutations (R102K and A109V) that caused the opposite phenotype to recessive ht1 mutant alleles (Hashimoto-Sugimoto et al. 2016, Hõrak et al. 2016). These dominant HT1 alleles caused a constitutive stomatal opening and disruption of high CO 2 -induced stomatal closing (Hashimoto-Sugimoto et al. 2016, Hõrak et al. 2016), in contrast to the constitutively closed stomata of the ht1 loss-of-function plants. The findings that the dominant point mutations in HT1 do not greatly affect HT1 kinase activity in vitro (Hashimoto-Sugimoto et al. 2016, Hõrak et al. 2016, Takahashi et al. 2022) suggest that proper regulation of HT1 kinase activity in response to CO 2 might be required for the stomatal CO 2 response.

Arabidopsis CBC1/2 kinases, which belong to the subgroup C7 of the Raf-like kinase family (MAPK Group 2002, Hiyama et al. 2017), have a similar CO 2 signaling phenotype to that of the HT1 protein kinase. CBC1 and CBC2 were identified by a phosphoproteomics approach. The cbc1 cbc2 double knockout mutant has constitutively closed stomata regardless of CO 2 concentration, similar to the ht1 loss-of-function mutants (Hiyama et al. 2017). CBC1 is phosphorylated in response to blue light (Hiyama et al. 2017). Although the physiological significance of the blue-light-induced phosphorylation of CBC1 remains unclear, CBC1 can function as a convergence point between stomatal blue light and CO 2 signaling pathways (Hiyama et al. 2017).

The mitogen-activated protein kinases, MPK4 and MPK12, belong to the subgroup B of the 20 MAP kinase genes in Arabidopsis (MAPK Group 2002), have a phenotypically opposite function to the HT1 kinase and CBC1/2 kinases. The mpk12-4 single gene loss-of-function mutant showed greater stomatal conductance and weak CO 2 sensitivity (Jakobson et al. 2016), and guard-cell-specific knockdown of the clade member, MPK4, in the mpk12-4 mutant caused a high stomatal conductance and largely disruption of CO 2 -induced stomatal closing (Tõldsepp et al. 2018). Interestingly, mpk12 was also found as a background mutation in some GABI-KAT lines. The phenotypic discrepancies among different calcium-sensing receptor (cas) alleles (Jakobson et al. 2016) and between different glutamate decarboxylase 2 (gad2) alleles (Piechatzek et al. 2024) in stomatal CO 2 response led to the identification of the lack of MPK12 as the genuine cause of their CO 2 -insensitive phenotypes. mpk12 loss-of-function mutants and natural variants showed reduced water use efficiency (WUE) (Des Marais et al. 2014, Jakobson et al. 2016), indicating the physiological importance of the stomatal CO 2 response in balancing water loss and carbon uptake. It was also reported that silencing the MPK4 ortholog in Nicotiana tabacum resulted in stomatal opening and CO 2 insensitivity (Marten et al. 2008). Interestingly, the stomata of mpk4/12 showed robust stomatal closure in response to ABA (Tõldsepp et al. 2018).

Second messengers

Possible second messengers and their functions in the stomatal CO 2 response have been discussed intensely in other papers (Assmann 1999, Vavasseur andRaghavendra 2005). We review this subject only briefly below:

Cytosolic calcium ion concentration. Calcium is a universal second messenger in signal transduction. Several studies showed that elevation in the cytosolic calcium concentration contributes to stomatal closure. In 1996, Webb et al. reported that intracellular Ca 2+ increased in response to high CO 2 concentrations in stomatal guard cells of Commelina communis using fura-2 fluorescence imaging (Webb et al. 1996). Calcium can mediate stomatal closure by regulating ion channels in guard cells (Schroeder and Hagiwara 1989). Intriguingly, modulation of calcium transients was observed during low and high CO 2 concentrations that lead to stomatal opening and closing in Arabidopsis guard cells (Young et al. 2006). It is still a mystery how calcium signaling can trigger a specific physiological response. One possible hypothesis is stimulus-specific regulation (priming) of Ca 2+sensitive proteins (Young et al. 2006, Dodd et al. 2010). Quintuple mutants in calcium-dependent protein kinases showed a slowed stomatal CO 2 response, whereas knockout of all five plasma membrane (PM)-targeted Ca 2+ -binding CBL proteins did not impair the stomatal CO 2 response (Schulze et al. 2021).

Cytosolic pH change. CO 2 lowers the pH in the solution when it dissolves. Since ion transporters in stomatal guard cells can be controlled by pH changes, cytosolic pH has been investigated as a possible second messenger in stomatal CO 2 responses. However, it remains unclear whether a decrease in pH mediates stomatal CO 2 signaling (Assmann 1999). The change in cytoplasmic pH caused by physiological concentrations of CO 2 seems to be almost negligible in guard cells (Brearley et al. 1997, Xue et al. 2011).

Malate. When the CO 2 concentration rises, the apoplastic carbon metabolite malate was reported to increase from about 1 mM up to about 3 mM in Vicia faba (Hedrich et al. 1994). Extracellular malate has been reported to activate the PM anion channels in guard cells, leading to anion efflux from guard cells and resulting in stomatal closure (Hedrich and Marten 1993). These findings led to a hypothesis that malate might be the CO 2 sensor that mediates stomatal movements in response to changes in CO 2 concentration (Hedrich and Marten 1993). However, other research reported that very high concentrations (40 mM or more) of apoplastic malate were needed for stomatal closure (Cousson 2000). Therefore, whether the increase in apoplastic malate explains stomatal CO 2 sensing has been unclear.

ABC transporters mediate export of molecules from the cytoplasm. In contrast, the PM-localized transporter ABCB14 was proposed to mediate influx of malate into guard cells (Lee et al. 2008). Loss-of-function mutants of the PM-localized transporter ABCB14 were reported to show a slightly lower stomatal conductance at a high CO 2 concentration than wild-type control plants (Lee et al. 2008). However, when isolated epidermis was exposed to high concentrations of CO 2 , CO 2 -induced stomatal closure similar to that of the wild type was observed. On the other hand, when the isolated epidermis was exposed to malate, the knockout plants closed stomata more tightly than the wild type. Malate produced in mesophyll cells may be involved in malate-dependent stomatal closure through the apoplastic pathway (Lee et al. 2008), but as discussed above, physiological concentrations of extracellular malate alone did have a consistent effect on CO 2 -regulated stomatal movements (Cousson 2000). Evidence from several laboratories suggests that crosstalk between guard cells and mesophyll cells is involved in CO 2 control of stomatal movements, with sugar supply to guard cells from the mesophyll playing an important role (Messinger et al. 2006, Flutsch et al. 2020a, 2020b).

CO 2 regulation of the PM potential in guard cells

Electrophysiological analysis has shown that the membrane potential of the guard cell PM becomes more positive (depolarizes) in response to high CO 2 (Brearley et al. 1997, Roelfsema et al. 2002). Consistent with these findings, electrophysiological analyses have shown that both S-type and R-type anion channels in guard cells are activated by CO 2 elevation (Raschke et al. 2003). Anion channel-mediated anion efflux causes depolarization of the guard cell PM.

It was suggested that the PM H + -ATPase is controlled by CO 2 (Edwards and Bowling 1985). Although this is likely, direct regulation of PM H + -ATPases during CO 2 signaling has not yet been reported, likely due to the difficulties in measuring this activity in the presence of other conductances and in biochemically investigating its CO 2 response in guard cells. Recent immunostaining has reported that the penultimate threonine residue of the PM proton pump of Arabidopsis thaliana can be dephosphorylated in response to the addition of NaHCO 3 (Ando et al. 2022). However, studies suggest that HT1 and CBC1/2 protein kinases do not function in phosphorylation of the penultimate threonine residue of the PM H + -ATPase (Hiyama et al. 2017, Ando et al. 2022). Further research is required to link early CO 2 signal transduction mechanisms to proton pump regulation.

What Is the Primary Stomatal CO 2 Sensor? CO 2 sensing in other organisms

Many organisms, including humans, sense CO 2 /bicarbonate for homeostasis and environmental responses. Soluble adenylyl cyclase (sAC) is an enzyme responsible for producing cyclic adenosine monophosphate (cAMP), one of the common second messengers of intracellular signal transduction across species. sAC was found to be activated by bicarbonate in mammals (Chen et al. 2000) and functions as a CO 2 /HCO 3 sensor (Rossetti et al. 2021). The crystal structure of the human sAC with bicarbonate ions was solved in 2014 (Kleinboelting et al. 2014). In insects such as mosquitoes and fruit flies, gustatory receptors sense CO 2 (Jones et al. 2007). However, no homologous proteins with similar functions have been found in higher plants, suggesting independent evolutions of CO 2 sensing components in plants and animals.

Rubisco

Most guard cells have chloroplasts with metabolic functions but likely with reduced photosynthetic activity compared to mesophyll cells (Reckmann et al. 1990, Tominaga et al. 2001, Lawson 2009). It has been demonstrated that photosynthesis in guard cells is required for maintaining stomatal turgor pressure rather than for CO 2 signaling (Azoulay-Shemer et al. 2015). A role for the weaker photosynthetic activity of guard cells in stomatal opening is supported by the finding that chlorophylldepleted guard cells can show collapsed closed stomatal apertures (Azoulay-Shemer et al. 2015). Rubisco is an enzyme that captures CO 2 as the first step of the Calvin cycle in C3 plants. Therefore, Rubisco could theoretically function as a guard cell CO 2 sensor candidate. Mutants in guard cell-expressed Rubisco small subunit genes did not show a clear impairment in high CO 2 -induced stomatal closing (Von Caemmerer et al. 2004, Hu et al. 2010). These findings suggest that the CO 2 sensor mechanism that controls stomatal closing is mediated by a different pathway.

Carbonic anhydrases and bicarbonate

Mechanisms that detect variations in CO 2 and bicarbonate concentrations regulate stomatal movements. Carbonic anhydrases (CAs) are enzymes that facilitate the reversible conversion of CO 2 and water into protons and bicarbonate ions. Double mutants in the beta-carbonic anhydrases βCA1 and βCA4 show a higher stomatal conductance and slowed CO 2 -induced stomatal closure in Arabidopsis (Hu et al. 2010). Notably, introducing a structurally unrelated mammalian CA into doublemutant plants by expression in guard cells effectively rescued the wild-type-like stomatal CO 2 response (Hu et al. 2010). This evidence suggests that the catalytic function of CA enzymes plays a critical role in "transponding" (Frommer 2010) the CO 2 signal within this pathway and implicates bicarbonate and/or protons as potential second messengers in the process of signal transduction. Hu et al. (2015) investigated the intracellular localization of beta-carbonic anhydrases βCA1 and βCA4 in guard cells, revealing that distinct cellular expression localizations can restore CO 2 -induced stomatal responses. Guard cell localization analyses and reaction-diffusion modeling led to the model that more than one cellular localization, including PM or cytoplasmic localization of CAs, enables CO 2 control of stomatal movements (Hu et al. 2015). A double mutant in maize beta-CA (βCA) showed impaired stomatal CO 2 responses (Kolbe et al. 2018). These maize beta-CAs are predicted to be located in the cytoplasm, consistent with this model. Moreover, mutation of the rice carbonic anhydrase βCA1 showed impairment in CO 2 -regulated stomatal movements consistent with these studies (Chen et al. 2017). Interestingly, different splice isoforms of rice βCA1 are targeted to different locations in guard cells that rescue the stomatal CO 2 response (Mao et al. 2025). Given that β-CAs use their enzymatic activity to accelerate the catalysis of CO 2 , this research suggested that the stomatal CO 2 sensors remained to be identified (Hu et al. 2010).

Direct modulation of SLAC1 anion channel by CO 2 /HCO 3 -

The SLAC1 (SLOW ANION CHANNEL-ASSOCIATED 1) gene encodes an S-type anion channel and triggers stomatal closure by releasing anions from guard cells (Negi et al. 2008, Vahisalu et al. 2008). S-type anion channel currents are enhanced by intracellular bicarbonate in wild-type Arabidopsis guard cells (Hu et al. 2010, Xue et al. 2011). The CO 2 /HCO 3 -signal is transduced to SLAC1 anion channel via protein kinases (Hashimoto et al. 2006, Xue et al. 2011, Hua et al. 2012, Hõrak et al. 2016). SLAC1 is activated by protein kinases (Geiger et al. 2009, Lee et al. 2009, Brandt et al. 2012). Electrophysiological reconstitution of the CO 2 response can be examined by measuring SLAC1-mediated anion channel currents in Xenopus laevis oocytes. For example, the co-expression of SLAC1 with either the OST1 (OPEN STOMATA 1) protein kinase or CPK6 (Ca 2+ -dependent protein kinase 6) or CPK23 in oocytes was found to activate SLAC1-mediated anion channel currents (Geiger et al. 2009, Lee et al. 2009). Reconstitution of ABA activation of SLAC1 was demonstrated by co-expression of ABA receptors, protein kinases, and PP2Cs (Brandt et al. 2012). The following reconstitution experiments indicated the potential for the existence of multiple CO 2 signaling pathways. Co-expression of CAs, PIP2;1 aquaporin, protein kinases, and SLAC1 showed that SLAC1 channels are enhanced at elevated CO 2 /HCO 3 -, and this was observed independent of which protein kinase (OST or CPK6 or CPK23) was co-expressed in oocytes. SLAC1 activity is further directly enhanced by ∼ =20% by elevation in the cytosolic bicarbonate concentration (Zhang et al. 2018). The R256 residue in SLAC1 is important for this direct modulation by CO 2 /HCO 3 - (Zhang et al. 2018). The structural resolution of the SLAC1 channel supports this finding, in which the R256 residue plays an important role in regulating the open probability of SLAC1 (Deng et al. 2021, Li et al. 2022). Moreover, the ten transmembrane-domain region of SLAC1 has been shown to function in CO 2 /HCO 3 sensing (Yamamoto et al. 2016). Gas exchange and patch clamp studies on transgenic plants with SLAC1-R256A demonstrated a partial impairment in CO 2 -regulated stomatal movement and HCO 3 --activated S-type anion channel currents in guard cell membranes, while low CO 2 elicited an intact stomatal opening response in SLAC1-R432A-expressing plants (Zhang et al. 2018). These observations suggested that the modulation of SLAC1 by CO 2 /HCO 3 -is modulatory and requires protein kinases that regulate SLAC1 and that other primary CO 2 sensors must exist that regulate protein kinases (Zhang et al. 2018).

HT1-MPK4

/12 Complex Is the Primary CO 2 /HCO 3 -Sensor CO 2 regulation of protein phosphorylation cascade The Arabidopsis HT1 and CBC1 protein kinases were identified as negative regulators in CO 2 -induced stomatal closure. The strong recessive ht1-2 allele shows no discernible low CO 2induced stomatal opening. Using recombinant proteins, Takahashi et al. found that the HT1 Raf-like kinase dramatically activates CBC1 protein kinase activity by phosphorylation at two amino acid residues inside or near the CBC1 kinase activation loop (Takahashi et al. 2022). The positive regulators MPK4/12 were found to inhibit HT1 activity (Hõrak et al. 2016, Jakobson et al. 2016, Takahashi et al. 2022). Addition of MPK4 or MPK12 revealed that adding NaHCO 3 as a source of CO 2 /bicarbonate reduced HT1-induced CBC1 kinase activation in vitro. When HT1 and MPK4/12 were mixed without the addition of CBC1, HT1 kinase activity decreased in the presence of NaHCO 3 , while another pair of protein kinases alone (HT1 and CBC1) did not respond to NaHCO 3 . The results above indicate that the HT1 and MPK4/12 combination can sense CO 2 /HCO 3 -(Takahashi et al. 2022).

CO 2 /HCO 3

-enhances HT1-MPK4/12 complex formation HT1 and MPK4/12 were reported to interact with each other directly (Jakobson et al. 2016, Tõldsepp et al. 2018). In vitro pull-down assays and quantitative blinded BiFC experiments in plant cells revealed that the MPK4/12-HT1 interaction depends on the presence of CO 2 /HCO 3 - (Takahashi et al. 2022). Several CO 2 -insensitive dominant mutations in the HT1 gene, including HT1-R102K (Hashimoto-Sugimoto et al. 2016) and HT1-A109V (Hõrak et al. 2016), impaired the interaction of these HT1 variants with MPK4 and MPK12, which can explain their dominant high CO 2 -insensitive phenotype. Moreover, the HT1-R102K and HT1-A109V isoforms disrupted the NaHCO 3dependent CBC1 inhibition in vitro (Takahashi et al. 2022), suggesting that CO 2 /HCO 3 --induced MPK4-HT1 complex formation is required for CO 2 -induced stomatal closure.

MAP kinase activity is not required for CO 2 sensing

Surprisingly, the kinase-inactive MPK12-K76R sustained the ability to mediate the CO 2 /HCO 3 --dependent CBC1 inhibition in vitro and rescued the CO 2 -less sensitive phenotype of the mpk12 knockout mutant (Takahashi et al. 2022). These data suggest that CO 2 sensing by HT1 and MPK4/12 does not seem to require the kinase activity of MPK4/12 (Takahashi et al. 2022). How the HT1 kinase is inhibited by MPK4/12 and elevated CO 2 /HCO 3 -remains unknown. The complex formation with MPK4/12 may induce HT1 structural change or block the interaction with CBC1 or ATP. This molecular sensor mechanism could be similar to the ABA receptor mechanism, in which ABA inhibits PP2C phosphatase activity through the ABA receptor Pyrabactin Resistance 1-Like or Reulatory Components of ABA Receptor (PYL/RCARs) that, upon binding ABA, directly interacts with the PP2C catalytic site (Cutler et al. 2010). The findings that the dominant HT1-R102K and HT1-A109V point mutations inhibit CO 2 /HCO 3 -dependent binding of these HT1 variants to MPK4/12 may indicate that these mutations could have a similar effect to the well-known ABA-insensitive PP2C mutations, abi1-1 and abi2-1, which disrupt ABA-dependent binding to PYL/RCARs, thereby impeding inhibition of their phosphatase activity. The proposed CO 2 sensing and signaling model overview is shown in Fig. 2.

CO 2 or HCO 3

-?

In the apoplastic space and cytoplasm, CO 2 can dissolve in water and exists as CO 2 and HCO 3 -forms depending on the prevailing pH. The conversion between CO 2 and HCO 3 -carbon species is rapidly catalyzed by βCAs. HCO 3 -can act as a signaling molecule in the stomatal CO 2 response (Xue et al. 2011). The inhibition of HT1 kinase activity by MPK4 and NaHCO 3 was observed in the reaction buffers adjusted to pH 7 or higher (Takahashi et al. 2022). Since HCO 3 -is the predominant form in such pH conditions, HT1 and MPK4/12 may perceive mainly HCO 3

-more so than CO 2 as a ligand. However, further research will be required to address this hypothesis.

Brachypodium BdMPK5 functions in CO 2 sensing

Grass stomata have unique structures consisting of two guard cells, flanked by subsidiary cells (Raissig et al. 2016). The molecular mechanisms mediating stomatal CO 2 signal transduction in grasses have remained largely unknown. A recent unbiased forward genetics infrared imaging CO 2 response screen revealed that the Brachypodium BdMPK5 gene, an orthologue of Arabidopsis AtMPK4, plays a central role in CO 2 -induced stomatal closure (Lopez et al. 2024). The BdMPK5 mutant identified by a genetic thermography screen causes a single amino acid substitution of D90N in the BdMPK5 protein. In vitro phosphorylation assays showed that the BdMPK5-D90N protein could not inhibit HT1 kinase activity, whereas the wild-type BdMPK5 protein could inhibit HT1 upon NaHCO 3 addition (Lopez et al. 2024). These results indicate that CO 2 sensing by MAP kinase and HT1 may be conserved in monocots. Stomata provide an avenue for CO 2 intake required for photosynthesis that largely occurs in leaf mesophyll cells. Stomatal conductance is positively correlated with CO 2 assimilation over wide ranges of environmental conditions in C3 and C4 plants (Wong et al. 1979). For example, at higher light intensities and warming temperatures, photosynthesis increases, as do stomatal apertures (Raschke 1970, Pankasem et al. 2024). Stomatal conductance and net CO 2 assimilation directly affect the intercellular CO 2 concentration (c i ). Direct measurements of leaf internal CO 2 concentration using microelectrodes demonstrated a wide c i range with a rapid decrease of c i to the range of <∼200 ppm in response to light exposure and increase in CO 2 assimilation. Furthermore, c i increased rapidly up to ∼three-fold within minutes after the onset of darkness due to a cessation of photosynthesis (Hanstein et al. 2001, Roelfsema et al. 2002). Such rapid and dynamic physiological ranges of leaf internal CO 2 concentrations result in feedback causing stomatal opening and closing responses to achieve the optimal balance of photosynthetic capacity and WUE (Ball et al. 1987).

Interactions between Stomatal CO 2 Sensing and Environmental

Light and CO 2 signaling in stomata

Red light and blue light induce stomatal opening. Red light is required at a much higher intensity than blue light to induce robust stomatal opening (Shimazaki et al. 2007). The blue light sensing and signaling mechanisms mediated by phototropins have been relatively well characterized and are reviewed elsewhere (Shimazaki et al. 2007). However, how the red light signal is perceived and transduced remains largely unknown and a matter of debate.

A body of evidence has shown that red light induces stomatal opening in Arabidopsis through its direct stimulation of leaf photosynthesis, primarily in mesophyll cells (Lawson et al. 2014). An increase in leaf photosynthesis under red light results in a reduction in c i and an increase in photoassimilates. A study showed that stomatal CO 2 signaling mechanisms play a pivotal role in mediating red light-induced stomatal opening. It was found that low CO 2 enhanced the stomatal blue light response (Assmann 1988). Hosotani et al. (2021) reported that a low CO 2 background partially bypasses the strong red light background required for stimulating a robust blue lightinduced stomatal opening. In addition, genetic studies performed by several independent groups showed that mutants lacking key CO 2 signaling players showed impaired red light responses (Matrosova et al. 2015, Hiyama et al. 2017, Taylor et al. 2024). Note, however, there is a parallel, molecularly less well understood red light signal transduction pathway in guard cells that mediates stomatal opening, independent of Ci, which is observed in mesophyll-free leaf epidermal strips of C. communis (Schwartz and Zeiger 1984). Moreover, experiments in Arabidopsis did not exclude an additional c i -independent component in red light-induced stomatal opening (Matrosova et al. 2015).

Several studies performed leaf gas exchange experiments where the leaf internal CO 2 concentration was controlled to a stable level and showed that stomata conductance increased in response to red light exposure (Messinger et al. 2006, Lawson et al. 2008). This suggests that an additional unknown red light sensing mechanism and signaling components can contribute to stomatal opening in response to red light and increased photosynthesis. Such low c i -independent mechanisms may, in part, explain the weak red light response in strong CO 2 signaling mutants (Matrosova et al. 2015, Taylor et al. 2024). Whether there is a biochemical interaction of c i -independent red light and CO 2 signaling mechanisms deserves thorough investigation in the future.

The leaves of cbc1/cbc2 double mutant plants lacking Raflike CBC1/CBC2 protein kinases, that function downstream of the HT1 Raf-like kinase, showed strongly impaired stomatal low CO 2 and red light-induced stomatal opening (Hiyama et al. 2017). HT1 functions as a subunit of the CO 2 /bicarbonate sensor that directly phosphorylates and activates CBC1 (Takahashi et al. 2022). HT1 can be activated in response to low CO 2 /HCO 3 -concentrations (Takahashi et al. 2022). Under red light, it is likely that the resulting low c i causes the level of HT1mediated CBC1 phosphorylation to be increased in guard cells of intact leaves. CBC1/CBC2 has been genetically linked to a light-induced deactivation of S-type anion channels in guard cells (Hiyama et al. 2017).

In addition, when whole leaves were exposed to red light, the phosphorylation level at various phosphosites of the PM H + -ATPases in stomatal guard cells increased, resulting in stomatal opening (Ando and Kinoshita 2018, Fuji et al. 2024, Hayashi et al. 2024). The penultimate threonine residue of H + -ATPases is rapidly dephosphorylated upon transition to darkness, elevated CO 2 or in the presence of NaHCO 3 (Ando et al. 2022). Taken together, this suggests that CO 2 and red light signal transduction mechanisms are likely to be integrated at the signaling events upstream of H + -ATPases, S-type anion channels, and other metabolic mechanisms, channels, and transporters in the guard cell signaling network, including the K + uptake and K + release channels in the PM of guard cells (Schroeder et al. 1987, Brearley et al. 1997).

Warming temperatures and stomatal CO 2 response

Increasing temperatures cause stomatal opening (Darwin 1898). However, the genetic and molecular mechanisms have been difficult to investigate, as increasing temperatures also increase the vapor pressure difference (VPD leaf ) between leaves and the atmosphere, which causes stomatal closing (Merilo et al. 2018). Experimental methods that hold VPD leaf stable have enabled the first insights into this pathway. Recently, stomatal CO 2 sensors and signaling mechanisms were found to mediate the stomatal response of intact leaves to warming temperatures. Stomatal opening mediated by a 10 • C temperature increase from 18 • C to 28 • C was severely impaired in mutants disrupting the functions of CO 2 sensors and early signaling components, including HT1, MPK12/MPK4, and CBC1/CBC2 (Pankasem et al. 2024). This temperature increase enhances CO 2 assimilation by rapid photochemical and biochemical processes, which in turn causes a reduction in the leaf's internal CO 2 concentration (Raschke 1970, Farquhar and Wong 1984, Pankasem et al. 2024). Indeed, warming-induced stomatal opening was abolished in mutant leaves of the CO 2 sensor component HT1. Combined genetic, physiological, and biochemical findings suggest that the warming-induced increase in leaf photosynthesis, followed by a reduction in c i , increases HT1 kinase activity (Takahashi et al. 2022, Pankasem et al. 2024) (Fig. 3).

Interestingly, stomatal conductance further increased when temperatures further increased to 34 • C, but CO 2 assimilation decreased (Urban et al. 2017, Diao et al. 2024, Pankasem et al. 2024). These findings show that additional photosynthesis-uncoupled mechanisms are triggered when plants are exposed to heat stress, most likely to facilitate evaporative cooling of leaves (Fig. 3). Several studies have recently proposed molecular mechanisms that could be linked to photosynthesis-uncoupled heat stress responses, including the stomatal blue light sensor phototropins (Kostaki et al. 2020). Furthermore, direct activation of PM H + -ATPases at 34 • C by high-temperature-responsive MAP4K, TARGET OF TEMPERATURE 3 was proposed (Xu et al. 2024). Another hypothesis proposes that heat causes rapid degradation of ABA in guard cells (Mills et al. 2024). The mechanisms mediating heat stress-induced stomatal opening deserve further investigation, given the increasing occurrence of heat waves in many regions.

CO 2 and ABA response

The interaction between ABA and CO 2 in stomatal closure has been debated for decades. Studies of the CO 2 sensor mutants ht1-2, ht1-A109V, and mpk12/mpk4GC demonstrated that ABA-induced early guard cell signal transduction occurs in first order independently of the early CO 2 signaling pathway (Hashimoto et al. 2006, Hõrak et al. 2016, Tõldsepp et al. 2018). This model is further supported by Boolean modeling (Gan et al. 2024) and physiological analyses showed that ABA induces stomatal closure at nominally zero CO 2 (Karanam et al. 2021). In contrast, alterations in stomatal CO 2 response kinetics observed in ABA signal transduction mutants point to a role for ABA in elevated CO 2 -induced stomatal closure (Xue et al. 2011, Chater et al. 2015, Hsu et al. 2018, Zhang et al. 2020, Seller and Schroeder 2023). A study suggested high CO 2 signal transduction occurs via the upregulation of ABA receptor transcripts and ABA signal transduction (Dittrich et al. 2019). However, high CO 2 -induced stomatal closing is completed within minutes, whereas transcript up-regulation and protein synthesis usually require more time. Stomatal CO 2 response phenotypes in the ABA receptor mutants reported by Dittrich et al. (2019) could not be confirmed (Zhang et al. 2020). In addition, elevated CO 2 was found not to activate ABA signal transduction based on the following findings: CO 2 elevation did not rapidly increase ABA concentrations in guard cells nor ABAinduced gene expression in guard cells nor activate OST/SnRK2s protein kinases in guard cells (Hsu et al. 2018, Zhang et al. 2020). Interestingly, though, plants maintain higher basal ABA levels and basal OST1/SnRK2s protein kinase activities in guard cells compared to other leaf cells under non-stressed conditions (Lahr and Raschke 1988, Hsu et al. 2018, Zhang et al. 2020). The current model suggests that basal ABA signaling amplifies and facilitates CO 2 -induced stomatal closure, with ABA and CO 2 signal transduction converging downstream of OST1/SnRK2s protein kinases in guard cells (Hsu et al. 2018, Zhang et al. 2020).

ABA induces massive chromatin remodeling in guard cells, and ABA-induced accessible regions in guard cells are substantially more abundant and significantly affect different accessible chromatin regions than those regulated by low or high CO 2 conditions, demonstrating that ABA and CO 2 cause highly distinct transcriptional programs in guard cells (Seller and Schroeder 2023). The above findings correlate with the role of ABA signaling in generating a 'memory' effect in guard cells during drought. ABA signaling is proposed to be required for building transcriptional memory in guard cells to prevent immediate reopening after re-watering plants and ensures a faster response to subsequent repeated dehydration stress (Virlouvet andFromm 2015, Seller andSchroeder 2023). Typically, plants cannot reopen their stomata until several days after re-watering has occurred. In contrast to drought, stomata can open and close rapidly and reversibly in response to fluctuations in intracellular CO 2 concentrations driven by the balance of photosynthetic activity and dark respiration throughout the day. Interestingly, several accessible chromatin regions upstream of some positive regulators of low CO 2induced stomatal opening are specifically repressed by ABA (Seller and Schroeder 2023), indicating a cross-talk mechanism between the ABA and CO 2 response at the level of chromatin remodeling, potentially enabling drought-induced stomatal closing to dominate over low CO 2 -induced stomatal re-opening (Seller and Schroeder 2023). These findings further indicate that basal ABA can modulate the stomatal CO 2 response by fine-tuning chromatin accessibility and the expression level of CO 2 signaling genes.

Stomatal development regulation by CO 2

Elevated CO 2 concentrations cause a reduction in stomatal density (Woodward 1987). Mutant leaves lacking the HIGH CARBON DIOXIDE encoding an enzyme that produces very long-chain fatty acids and wax in guard cells show an increased stomatal index under a high CO 2 concentration in contrast to the lower stomatal index of WT plants (Gray et al. 2000). Such an inverse phenotype was also found in mutant leaves of the negative regulators of stomatal development epf2 and erecta, disrupting genes encoding pro-peptide and a peptide ligand binding leucine-rich receptor protein kinase, respectively (Engineer et al. 2014). Interestingly, ca1/ca4 double mutant leaves lacking CA isoforms that show slowed stomatal conductance responses to CO 2 concentration changes (Hu et al. 2010, Kolbe et al. 2018) show increased stomatal density and index in Arabidopsis (Engineer et al. 2014).

Although ca1/ca4 showed a stomatal development phenotype (Hu et al. 2010, Engineer et al. 2014), other strong CO 2 signaling mutants, including ht1 mutants and mpk12/mpk4 mutants, do not show a clear difference in stomatal density (Hashimoto et al. 2006, Engineer et al. 2014, Hõrak et al. 2016, Tõldsepp et al. 2018). Therefore, the HT1-MPK12/MPK4 mediated CO 2 /bicarbonate sensing mechanism does not appear to function upstream of CO 2 -modulation of stomatal development. The physiological mechanisms by which CAs affect stomatal development require further analysis.

Dual-Role Kinase MPK4: Plant Immune Response and CO 2 Sensing MPK4 is a central MAP kinase in the pathogen response pathway

MPK4 plays important roles in many plant immune responses against pathogens. Arabidopsis mpk4 loss-of-function mutants show an extreme dwarf phenotype, salicylic acid accumulation, and enhanced immune-related gene expression, suggesting that MPK4 negatively regulates the immune response in Arabidopsis (Petersen et al. 2000). MPK4 is activated by a classical upstream MAP kinase cascade consisting of MAP3K (MEKK1) and MAP2K (MKK1/2) in immune signal transduction (Fig. 4) (Berriri et al. 2012). This suggests that MPK4 kinase activity is required for the plant pathogen response.

The extreme dwarf phenotype of MPK4 is not complemented by MPK12 expression under the control of the MPK4 promoter (Tõldsepp et al. 2018). MPK4 and MPK12 share 81% amino acid identity (BLASTP alignment), and both MPKs function in CO 2 sensing in Arabidopsis, but their immune response functions differ. The observation that recombinant MPK12 protein has a weaker kinase activity than MPK4 (Tõldsepp et al. 2018) is consistent with the inability of MPK12 to complement the mpk4 mutant's dwarf phenotype. To investigate how MPK4 and MPK12 proteins function differently may provide insight into how the immune response and the stomatal CO 2 response are specifically transduced via the same signaling component.

MPK4 homologs in P. patens

The moss P. patens has two homologs of Arabidopsis AtMPK4, named PpMPK4a and PpMPK4b, and it has been shown that PpMPK4a is required for immune responses (Bressendorff et al. 2016). This suggests that the involvement of MPK4 in plant immune responses seems to be an ancient function.

A previous study reported that P. patens closes stomata in response to high CO 2 concentrations (Chater et al. 2011). Furthermore, chloroplast orientation movement is regulated in response to CO 2 in the moss (Sugiyama and Terashima 2023). These findings suggest that P. patens has CO 2 sensing mechanism(s). However, whether moss stomata can respond to CO 2 is controversial (Brodribb and McAdam 2011). Another study from a different group reported that stomatal conductance does not respond to CO 2 concentration changes in several moss species (Kubásek et al. 2021). It is unknown whether the moss MPK4s function in CO 2 sensing. Investigating the function of PpMPK4a and PpMPK4b proteins in CO 2 sensing and the evolutionary history of HT1 may provide insight into the origin of the plant stomatal CO 2 response. However, recent phylogenetic analyses have shown that HT1 originated from angiosperms (Ivsic et al. 2025).

Gene duplication produces MPK4 family members with different roles

It has been suggested that the Arabidopsis MPK12 gene arose from an ancestral MPK4 gene in the Brassicaceae (Tõldsepp et al. 2018). The Arabidopsis genome includes MPK11, the closest homologous gene to MPK4 (89% amino acid identity in BLASTP alignment), in subgroup B of the MAP kinase gene family. MPK4 plays an important role in cell plate formation by interacting with microtubules during cell division, and MPK11 seems to play a partially redundant role in this process (Kosetsu et al. 2010). Surprisingly, MPK11 does not seem to function in the stomatal CO 2 response (Tõldsepp et al. 2018). P. patens MPK4b may have a more crucial or broader function than PpMPK4a since the authors could not isolate a knockout mutant in the moss PpMPK4b gene (Bressendorff et al. 2016). These imply that the MPK4 genes among plant species might have evolutionarily developed diversified functions by gene duplication.

Stomatal CO 2 sensing and plant immune response pathways

One of the open questions is why MPK4 plays important roles in two different environmental responses: immune response and stomatal CO 2 response. Stomata are important for the immune response since stomatal pores provide a pathway for pathogen entry into plant leaves. A previous study revealed that plants close stomata in response to pathogen elicitors, referred to as stomatal immunity (Melotto et al. 2006). On the other hand, pathogens secrete effectors that have been suggested to promote stomatal opening to facilitate invasion into the plant body. A recent study further showed that bacterial effectors also promote stomatal closure via ABA accumulation to induce an aqueous space in the apoplast (Roussin-Leveillee et al. 2022). Stomatal signaling may have evolved through interactions between plants and pathogens. It is unclear whether MPK4 is involved in the pathogen-induced responses in stomata. Recently, the cbc1 cbc2 mutant was reported to show resistance to pathogens (Goncalves Dias et al. 2024), which may be a direct pleiotropic effect of the reduced stomatal conductance of this double mutant that restricts the entry of the pathogens (Hiyama et al. 2017). More genetic and physiological evidence will be needed to understand the possible crosstalk of these two responses.

Future Research

The discovery of CO 2 /HCO 3 --dependent complex formation of HT1 and MPK4/12 has revealed the nature of plant stomatal CO 2 sensing. HT1-induced CBC1 activation and CO 2 -induced inhibition of HT1 kinase activity by binding MPK4/12 have provided insight into the primary CO 2 signaling core. These findings have also highlighted some questions to be addressed in the future. Solving the structure of the HT1-MPK4/12 protein complex will be necessary. The precise structure of the HT1-MPK4/12 complex would solve these important questions: (i) whether and, if so, how CO 2 and/or HCO 3 -bind to this protein complex? (ii) How does CO 2 /HCO 3 -enhance HT1-MPK4/12 complex formation? (iii) How do MPK4/12 and HT1 specifically bind to each other? And (iv) How does MPK4/12 inhibit HT1 kinase activity in the presence of CO 2 /HCO 3 -without MAP kinase activity (Takahashi et al. 2022)? Additionally, the direct downstream substrate(s) of CBC1/2 protein kinases and protein phosphatases that have been proposed to be required to dephosphorylate CBC1 and possibly HT1 (Takahashi et al. 2022) remain unknown.

Stomatal regulation of gas exchange is critical for plant WUE. Studies in crops have provided evidence that WUE can be improved at elevated CO 2 concentrations (Conley et al. 2001, Al-Salman et al. 2023, Sobuj et al. 2024, Radolinski et al. 2025). This effect is considered due to decreased stomatal conductance and transpirational water loss and potential increased carbon assimilation under high CO 2 concentration conditions. Note that WUE depends on conditions and plant species, given the impacts of heat stress and vapor pressure deficit on stomatal conductance and water use (Gray et al. 2016). Genetic evidence has documented that the stomatal CO 2 response is critical for improving instantaneous WUE (Hu et al. 2010, Des Marais et al. 2014, Xiao et al. 2024). The HT1 ortholog in rice likely functions in the stomatal CO 2 signaling pathway (Xiao et al. 2024). To investigate whether the HT1-MPK4/12 CO 2 /HCO 3 -sensor is conserved in other crops will be needed. The effect of high CO 2 concentrations on WUE could be affected by temperature and water stress (Gray et al. 2016, Hatfield and Dold 2019, Li et al. 2023, Pankasem et al. 2024, Sobuj et al. 2024). Further research on the molecular interactions between these signaling pathways would advance our knowledge.

Atmospheric CO 2 concentrations have fluctuated since stomata emerged 400 million years ago. It remains unclear when land plants acquired a stomatal movement CO 2 response. The evolutionary aspect of the HT1-MPK4/12 module and signaling factors such as CBC1 would be interesting subjects for additional analyses. Understanding how plants have responded to and acclimated to atmospheric CO 2 concentrations may provide a clue for future efficient plant breeding and engineering to improve crop yields and WUE of crops and trees in view of increasing CO 2 concentrations and weather extremes.