Arabidopsis thaliana ecotype Col-0 constitutively expressing intracellular Ca²⁺ indicator aequorin (pMAQ2) is a gift from M. Knight and the principles of how the active aequorin is formed can be found in Knight et al. (1991). Arabidopsis plants were grown in 150 mm × 15 mm round Petri dishes in half-strength Murashige and Skoog salts (MS; Gibco), supplemented with 1.5% (w/v) sucrose (Sigma), and 0.8% (w/v) agar (Becton Dickinson) adjusted to pH 6.0 with KOH in controlled an environmental room at 21 ± 2°C. The fluency rate of white light was ∼110 μmol m^-2 s^-1. The photoperiods were 16 h light/8 h dark cycles. Seeds were sterilized with 2.5% plant preservative mixture (Caisson Laboratories) and stratified at 4°C for 3 days in the dark, and then transferred to the growth room.

Arabidopsis thaliana plants expressing cytosolic apoaequorin were used for [Ca²⁺]_i measurements (Knight et al., 1991; Tang et al., 2007). Sixty-four seedlings were grown on half-strength MS medium for 8 days. Reconstitution of aequorin was performed in vivo by spraying seedlings with 3.3 mL of 10 μM coelenterazine (from Prolume) per Petri dish followed by incubation at 22°C in the dark for 8 h. Treatments and aequorin luminescence imaging were performed at room temperature using a ChemiPro HT system, which includes a cryogenically cooled and back-illuminated charge-coupled device (CCD) camera, liquid nitrogen autofiller, camera controller, and computer-equipped WinView/32 software (Roper Scientific) as described previously (Tang et al., 2007). The CCD camera has a 1300 × 1340 pixel resolution and is cooled to -120°C by the cryogenic cooler system prior to image recording. The recording was started 80 s prior treatments and luminescence images were taken every 20 s or continuous 7 min. The total remaining aequorin was estimated by treating plants with a discharging solution containing 0.9 M CaCl₂ in 10% (v/v) ethanol and recorded for 5 min until values were within 1% of the highest discharge value (Tang et al., 2007; Ranf et al., 2012; Yuan et al., 2014). The recorded luminescence images were analyzed using Meta Morph 7.7 and WinView/32. Here, the Ca²⁺ level depicted as L/L_max ratio correlates with the light emission from aequorin. To calculate the ratio, the actual aequorin luminescence, denoted as L, at any sampling point is normalized by the total remaining aequorin (Knight et al., 1996; Ranf et al., 2012). The experiments were carried out under room temperature between 22 and 24°C.

For stress treatments, Petri dishes were placed individually into the ChemiPro HT chamber and luminescence images were started 80 s prior the treatment and taken at 20 s intervals or 7 min continuously. The treatment solution (100 mL) at 1 μM concentrations of flg22 and Pep1 (Felix et al., 1999; Huffaker et al., 2006), which were synthesized by China Peptide¹ or 200 mM NaCl (Sigma) was added into Petri dish in the dark, and luminescence was recorded. For changes in bath solution, a four-channel peristaltic pump (Dynamax RP-1, Rainin) was used to perfuse Petri dish with water as indicated in the figures. Then, additional stress treatment was applied by adding 100 mL solution into Petri dish.

Changes in cytosolic Ca²⁺ concentration ([Ca²⁺]_i) can be monitored by the bioluminescent Ca²⁺-binding protein aequorin in vivo (Knight et al., 1991). Apoaequorin can be expressed in plants and spontaneously reconstitutes to functional holo-aequorin upon addition of the native luminophore coelenterazine (CTZ-n) or chemically modified derivatives, such as coelenterazine-h (CTZ-h), for enhanced sensitivity (Shimomura et al., 1993; Mithofer and Mazars, 2002).

To determine whether the mechanisms behind the increases of [Ca²⁺]_i induced by biotic and abiotic stresses are interrelated in Arabidopsis, we first attempted to identify the optimum concentrations of flg22 and Pep1 that ideally could be applied to generate about half of the maximum amplitude of [Ca²⁺]_i required for potential up- and down-regulation. Furthermore, we attempted to establish the kinetics of flg22- and Pep1-induced [Ca²⁺]_i increases so as to administer these stresses in different sequential combinations. To analyze flg22-induced increases in [Ca²⁺]_i, we treated Arabidopsis seedlings expressing aequorin with solutions containing 0 to 2 μM flg22. Aequorin bioluminescence images were recorded every 20 s for 600 s. The Ca²⁺ level correlates with the light emission from aequorin and is depicted as the ratio of L/L_max, where the actual aequorin luminescence (L) at any measurement point is normalized to the total remaining aequorin (Knight et al., 1996). Plants grown on the half-strength MS medium had an average basal [Ca²⁺]_i of 80 ± 21 nM (Dodd et al., 2010). As expected, the [Ca²⁺]_i increased in response to flg22 treatment (Figure 1A). The magnitudes of [Ca²⁺]_i increases were found to be dependent on the concentration of flg22, with a higher concentration of flg22 evoking a greater increase in [Ca²⁺]_i. The flg22 concentration required for a half-maximal response (i.e., the half of elicitor concentration which induced maximal [Ca²⁺]_i increase) was 1 μM, which was chosen as an optimum concentration for the subsequent analysis of interaction with Pep1- and NaCl-induced increases in [Ca²⁺]_i. Next, we determined the temporal dynamics of flg22-induced [Ca²⁺]_i increases under the imposed experimental conditions as a control for further comparison (Figure 1D). We found that [Ca²⁺]_i increased immediately after the application of 1 μM flg22, reached a peak at about 120 s, and then declined gradually (Figure 1D). Imaging aequorin bioluminescence for less than 20 s resulted in images with a low signal-noise ratio in our system. Thus, the temporal resolution was set at about 20 s, which was sufficient for the current study. At about 200 s, the [Ca²⁺]_i was reduced to a new resting level. Similarly, we analyzed the increases in [Ca²⁺]_i in response to Pep1. Seedlings were treated with different concentrations of Pep1 from 0 to 4 μM, and [Ca²⁺]_i was analyzed. As expected, Pep1 induced increases in [Ca²⁺]_i in a dose-dependent manner (Figure 1B). The [Ca²⁺]_i increases recorded after single treatments were consistent with the results described above (Figure 1C). The Pep1 concentration required to achieve a half-maximal response was around 1 μM, with the magnitude of [Ca²⁺]_i similar to that induced by 1 μM flg22. We then determined the temporal dynamics of the [Ca²⁺]_i increase induced by 1 μM Pep1. Following treatment with 1 μM Pep1, the [Ca²⁺]_i increased and reached a peak at 140 s (Figure 1E), then it took another 250 s for the [Ca²⁺]_i to reach a new basal level. In overall, it seems that the increases of [Ca²⁺]_i occur faster in response to flg22 than Pep1, but reset to a resting level 400 s after the treatment.

To further characterize the potential interaction between the different biotic stress stimuli-triggered [Ca²⁺]_i signaling, plants were treated either with the same or different stimulus. When the Arabidopsis seedlings were treated with 1 μM flg22, the level of [Ca²⁺]_i increased quickly to reach a peak, then decreased to the new resting level after 400 s (Figure 2A), as described in Figure 1A. A subtle increase in [Ca²⁺]_i could be detected in seedlings after they had been washed with deionized water at around 500 s (Figure 2A; green). Next, flg22 was added again, which resulted in a very minor increase in [Ca²⁺]_i. After 800 s, it decayed to a level similar to the previous resting level. Compared with the first flg22 treatment, which led to a large [Ca²⁺]_i increase, the second flg22 treatment resulted in a [Ca²⁺]_i increase that was only a fraction of the size of the first [Ca²⁺]_i increase. This observation suggests that the flg22-activated Ca²⁺ permeable channel may be desensitized or adapted by some unknown signaling elements upstream. To test whether the desensitization or adaptation occurs, we can (after waiting for 3 h) detect a normal [Ca²⁺]_i increase in response to flg22. This result suggests that desensitization of the channel is likely to happen, which agrees with the results reported by Heese’s group (Smith et al., 2014). Subsequently, we analyzed whether the MAMP-activated Ca²⁺ permeable channel was affected by the initial MAMP treatment. The second flg22 treatment was replaced by a treatment with 1 μM Pep1 at 600 s (Figure 2B). Interestingly, the peak of [Ca²⁺]_i induced by 1 μM Pep1 was clearly greater than that of 1 μM flg22 (P < 0.001). After 900 s, the [Ca²⁺]_i decreased to a new basal level (Figure 2B). The lower inhibition of the Pep1-induced [Ca²⁺]_i increase compared with the increase induced by the initial flg22 treatment suggests that the initial high level of [Ca²⁺]_i, which resulted from the flg22 treatment, inhibited flg22 to a greater extent than Pep1 (Figures 2A,B). By analogy, we used Pep1 as the first stimulus to treat the seedlings, and then analyzed the second treatment using flg22 or Pep1. When the second Pep1 was added to the Petri dish, following the first Pep1 treatment and water washing step at around 540 s the [Ca²⁺]_i level stabilized to a point similar to previous resting levels (Figures 2D,E). The column chart has also been used to clearly show the results, as in Figures 2C,F. However, when we used 1 μM flg22 to replace Pep1 at 600 s, the peak value was smaller but significantly higher than that induced by the second Pep1 treatment (Figures 2D,E). Similarly, our results suggested that the high [Ca²⁺]_i resulting from the initial Pep1 activation inhibited the MAMP-induced [Ca²⁺]_i to a greater degree than the PAMP-induced [Ca²⁺]_i (Figure 2).

Based on our study, we know that different biotic stresses may induce [Ca²⁺]_i increases via different channels. It is of great importance to further characterize the interaction between biotic and abiotic stress stimuli-triggered [Ca²⁺]_i signaling. For such a purpose, we treated the plants with both the biotic and abiotic stimulus. When the Arabidopsis seedlings were treated with 1 μM flg22, the level of [Ca²⁺]_i increased rapidly to reach a peak, and decreased to the new resting level after 400 s. A subtle increase in [Ca²⁺]_i could be detected in the seedlings after they were washed in deionized water at around 500 s (Figure 3A; green). Next, NaCl was added, which caused a sharp increase in [Ca²⁺]_i (Figure 3A). The [Ca²⁺]_i then decayed from 700 s to a level similar to previous resting level. Compared with the single NaCl treatment, which led to a large increase in [Ca²⁺]_i increase (Figure 3C), the NaCl treatment after the flg22 stimulus still resulted in an increase in [Ca²⁺]_i that was similar to a single NaCl-induced increase in [Ca²⁺]_i. This observation suggests that there may be no interaction between the NaCl-activated Ca²⁺ permeable channel (NaC) and flg22-activated Ca²⁺ permeable channel. To verify this hypothesis, we treated the seedlings with 200 mM NaCl. The [Ca²⁺]_i increased quickly to reach a peak and then decreased to the new resting level after 150 s (Figure 3C). Next, we added 1 μM of flg22, which caused an increase in [Ca²⁺]_i similar to the effect with a single flg22 treatment. Subsequently, we used a DAMP elicitor Pep1 instead of flg22 to determine whether a DAMP-induced [Ca²⁺]_i increase can affect an NaC. To clearly show the results, column chart were used in Figures 3B,D. As expected, it appeared that there was no interaction between the NaC and Pep1-activated Ca²⁺ permeable channel (Figures 4A–D). Based on this study, our results suggest that abiotic and biotic stress stimuli-activated Ca²⁺ permeable channels may be completely independent of each other.

To investigate thoroughly the relationship and interaction between [Ca²⁺]_i increases triggered by biotic and abiotic stresses, Arabidopsis seedlings were treated with 1 μM flg22, 1 μM Pep1, or 200 mM NaCl separately, or any two of these three elicitors. The [Ca²⁺]_i increases recorded after single treatments were consistent with the results described above. When plants were treated with 1 μM flg22 and 1 μM Pep1 together, the peaks of [Ca²⁺]_i were slightly larger than those induced by each individual stimulus (Figure 5A). However, when plants were treated with 200 mM NaCl together with 1 μM flg22 or 1 μM Pep1, the peaks of [Ca²⁺]_i were larger than those induced by each individual stimulus, showing an additive effect (Figures 5C,E). To further analyze the difference in [Ca²⁺]_i increases in response to both biotic and abiotic stresses, we calculate the [Ca²⁺]_i to clearly illustrate the data. As shown in Figure 5B, compared with individual treatment with Pep1, plants treated with flg22 and Pep1 together show only a slight increase in [Ca²⁺]_i. However, compared with individual treatment with NaCl, plants treated with flg22 and NaCl together, or Pep1 and NaCl together, show an increase in [Ca²⁺]_i that is almost equal to that shown with the combined treatment (Figures 5D,F). These results suggest that the flg22-induced and Pep1-induced [Ca²⁺]_i increases may operate through similar Ca²⁺ permeable channels, though biotic stress-induced [Ca²⁺]_i increases may occur as independent events. In other words, biotic and abiotic stresses may activate different Ca²⁺ permeable channels.

The co-treatment of biotic and abiotic stresses triggers an additive effect on the increase of [Ca²⁺]_i. Treated either by flg22 with NaCl or Pep1 with NaCl, the increase of [Ca²⁺]_i could indicate that biotic and abiotic stresses activate different Ca²⁺ permeable channels. While it is not clear whether the treatment of NaCl will affect the calcium signal process and peak value, which are caused by the flg22/Pep1 treatment. To answer this question, in our experiments, seedlings were treated either by 1 μM flg22 along with 200 mM NaCl or 1 μM Pep1 along with 200 mM NaCl. The results are shown in Figure 6. The shaded area in Figures 6A,D illustrates the difference between the co-treated plants and NaCl-only treated plants. The subtracted shaded area reveals that the dynamic process and peak value of [Ca²⁺]_i achieved by the treatment of using both flg22 and Pep1 are similar to those delivered by using the flg22 or Pep1 alone (Figures 6B,C,E,F). These results indicate that biotic and abiotic stresses induce the increase of [Ca²⁺]_i through different Ca²⁺ permeable channels. Such results are consistent with our prediction.