The NADPH oxidase RBOHD is a Ca2+ channel
Hydrogen peroxide (H2O2)-induced the significant elevation of cytosolic free Ca2+ concentration ([Ca2+]cyt) in plants carrying calcium indicator aequorin (Col-0AEQ) was completely blocked by the calcium channel inhibitor LaCl3, suggesting that calcium channels are essential for Ca2+ influx across the plasma membrane during ROS signaling (Fig 1a, Extended Data Fig. 1). RBOHD in Arabidopsis thaliana (RBOHD, or AtRBOHD) and other plants localized on the plasma membrane has been extensively studied and is predicted to couple with an unidentified calcium channel for encoding Ca2+-ROS waves in numerous proposed models 2,10-12. We used rbohD mutant carrying a transposon insertion that results in no detectable full-length transcripts and lacks the specific ROS spike induced by pathogen-associated molecular pattern (PAMP) factor flg22 (Extended Data Fig. 2 and 3). Interestingly, the ROS-induced spike of [Ca2+]cyt dramatically decreased in rbohDAEQ in dependence of extra high Ca2+ (Extended Data Fig. 4), further supporting that RBOHD potentially integrates Ca2+-ROS signals.
To explore the physiological role of RBOHD in relation to calcium, we first examined the phenotypic responses of WT, rbohD and cngc2 plants under varying concentrations of additional calcium in the growth hydroponic media. Compared to WT, rbohD mutants exhibited significantly reduced plant size under low supplemental calcium conditions, whereas these phenotypes recovered under high calcium supplementation (Fig. 1b, c), contrasting with the phenotype observed in cngc2 controls 15. This highlights the critical role of RBOHD in sustaining plant growth in a Ca2+ concentration-dependent manner, thereby establishing a direct link between plant RBOHD function and calcium regulation.
To determine whether RBOHD exhibits calcium channel activity, we performed two-electrode voltage clamp (TEVC) recording assay in Xenopus oocytes, a widely adopted method for studying ion channels 15,16. RBOHD-GFP fusion protein localizes at the plasma membrane of the oocytes (Extended Data Fig. 5). Surprisingly, upon injection of RBOHD capped RNA (cRNA), oocytes displayed significantly large currents at -140 mV in the presence of 30 mM CaCl2 (Fig. 1d, e), which was significantly higher than that observed in the positive control CNGC14 17,18, whereas the negative control showed almost no detected current. Thus, we report for the first time that Arabidopsis NADPH oxidase RBOHD functions as a calcium channel.
To further validate the calcium channel activity of RBOHD, we conducted additional experiments in human embryonic kidney 293T (HEK293T) cells and the Ca2+-uptake deficient yeast mutant cch1 mid1 respectively. Stimulation of HEK293T cells expressing the calcium indicator GCaMP6s with exogenous 400 μM CaCl2 elicited distinct calcium-induced specific spikes in both RBOHD- and CNGC14-expressing cells (Fig. 1f). Furthermore, we observed yeast mutant cch1 mid1 was rescued by RBOHD similar to the positive control CNGC14 and wild-type BY4741 (Fig. 1g). In summary, integrating data from TEVC recording, HEK293T and yeast systems reveal a novel function of Arabidopsis NADPH oxidase RBOHD as a calcium channel.
RBOHD behaves as a non-selective cation channel
Calcium channels, such as transient receptor potential (TRP) channels, are known for transporting various ions 19. To investigate whether RBOHD exhibits similar capabilities, we assessed the effects of multiple cations in Xenopus oocytes using the TEVC recording assay. The results demonstrated that RBOHD can transport not only monovalent ions like K+ and Na+, but also bivalent ions like Ca2+ and Mg2+. However, it exhibitedvery weak transport activity of Ba2+ (Fig. 2a, b). This indicates that RBOHD possesses characteristics of transporting multiple cationic substrates, similar to ion channels such as human TRPs and Arabidopsis CSCs/OSCAs 19-21.
To further characterize the fundamental features of the calcium channel RBOHD, we analyzed its dependence on Ca2+ concentration and its pharmacological properties using the TEVC system. Increasing the CaCl2 concentration from 5 mM to 30 mM resulted in a corresponding increase in current signals from 1.5 μA to 5 μA, demonstrating that RBOHD channel activity is Ca2+ concentration-dependent (Fig. 2c, d). However, when the calcium channel inhibitor LaCl3 was added to the TEVC assay, the calcium channel activity of RBOHD was significantly reduced to levels comparable to negative controls (Fig. 2e, f). Furthermore, treatment with DIDS (4,4'-Diisothiocyanatostilbene-2,2'-disulfonate), which blocks Cl- channels known to amplify Ca2+ signals in TEVC 16, resulted in a significantly stronger primary Ca2+ signal compared to negative controls (Fig. 2e, f). In conclusion, RBOHD functions as a typical non-selective cation channel that is permeable to Ca2+ and other cations.
Human and yeast NOXs are universal Ca2+ and cation channels
NOXs are predicted to have emerged in the early stages of eukaryotic evolution, with mammalian and fungal NOXs sharing conserved protein domains similar to those found in plant RBOHs (7 and Fig. 3a, b). We aimed to determine whether NOXs from human (Homo sapiens), yeast (Saccharomyces cerevisiae) and other species, exhibit general channel activity similar to RBOHD across eukaryotes. We selected Arabidopsis and rice RBOHs, Human NOX2/NOX5/DUOX1 and yeast YNO1 which is essential for responding to stresses like H2O2, apoptosis and actin function 7,22. We also included SpNOX from the bacterium Streptococcus pneumoniae, given its potential role in ROS-related functions and its status as an exceptionally rare NOX homolog in prokaryotes 23.
We evaluated the calcium channel activity of various NOXs in Xenopus oocytes by TEVC recording assay. Among the tested NOXs, AtRBOHC/F, closely related to AtRBOHD, exhibited calcium channel activity, whereas AtRBOHE/H, OsRBOHA/B, and HsNOX5/DUOX1 did not (Fig 3c, d, Extended Data Fig. 6). Using CNGC14 as a positive control, we observed significant currents at -140 mV for HsNOX2, ScYNO1 and AtRBOHC in the presence of 30 mM CaCl2 (Fig. 3c, d). These results were further validated through independent yeast system to access calcium channel activity. HsNOX2, ScYNO1 and AtRBOHC were able to complement the growth phenotype of yeast mutant cch1 mid1, like CNGC14 and controls (Fig. 3e). However, SpNOX showed no calcium channel activity in either the TEVC or yeast assays (Fig. 3c, d, e). In summary, Arabidopsis, human and yeast NADPH oxidases universally exhibit calcium channel activity in eukaryotes.
To further investigate ion selectivity, we conducted assays similar to those used for RBOHD (Fig. 2a, b). The results showed that ScYNO1 and HsNOX2 can also transport the monovalent cation like Na+ and K+, as well as bivalent cations like Ca2+ and Mg2+, but exhibited very low activity for Ba2+ (Fig. 3f, g and Extended Data Fig. 7), consistent with the findings for RBOHD (Fig. 2a, b). Moreover, the signals observed with CaCl2 and Ca-Glu in the HsNOX2 TEVC assays were similar (Fig. 3g), suggesting that the cation Ca2+, rather than anions, plays a critical role in NOX channel activity. Collectively, these findings demonstrate that eukaryotic NADPH oxidases function as Ca2+/cation channels.
The function of RBOHD is regulated by protons and depends on external Ca2+ concentrations
To assess the function of calcium channel NOXs in response to environmental stimuli and stress signals, we first tested the calcium permeability of RBOHD using the TEVC assay under different pH values. The results showed that when pH was reduced from 7.4 to 5.6, the currents of RBOHD increased from an undetectable background level like negative controls to a high level of 1.5 μA, whereas there was no change was observed in the control CNGC14 (Fig. 4a, b). This suggests that lower pH, or more protons H+, is required for switching on the calcium channel activity of RBOHD. When the pH value ranged from 10 to 4.5, wild-type roots became longer with the increase of H+ concentration and showed a gradient change. However, rbohD roots were relatively insensitive to pH changes (Fig. 4c, d), indicating that RBOHD function is dependent on ambient pH or proton concentration. Furthermore, to assess the calcium-dependent physiological function of RBOHD, we observed the increasement of additional Ca2+ enhanced a significant difference in root elongation between WT and rbohD, especially at low pH 5.8 versus high pH 8 (Fig. 4e, f). This further supports that the function of the RBOHD channel on root elongation is not only dependent on pH, but also requires sufficient exogenous Ca2+.
To further test the function of RBOHD on stress responses, we found that rbohDAEQ seedlings treated with flg22 resulted in significantly lower [Ca2+]cyt spikes compared to controls especially under low extra Ca2+ conditions (Extended Data Fig. 8). Compared to controls, rbohD plants were more susceptible to bacteria Pst DC3000, and also showed inhibited stomata closure in response to flg22, particularly under insufficient extra Ca2+ supply (Extended Data Fig. 8). When adding SA, rbohD roots were extremely insensitive to the changing low or high extra calcium (Extended Data Fig. 8). Collectively, it demonstrates that the calcium-dependent function of RBOHD integrates Ca2+-ROS signals in stress responses.
Taken together, NADPH oxidases mediated Ca2+ influx are widely important for growth and defense in eukaryotic organisms. The function of NOX channel activity is activated by protons and depends on calcium concentrations. Therefore, we speculate that Ca2+-ROS coupling signaling stimulated by multiple internal and external cues could require NOX as a key hub to recognize, integrate and transduce signals in cells(Fig. 4g).