Evaluation of plant stress due to plasma-generated reactive oxygen and nitrogen species using electrolyte leakage

For the plant-stress measurement experiment, a previously developed humidified-air-plasma device ^15,24) is used and operated at fixed air $\left({F}_{{\rm{air}}}\right)$ and water $\left({F}_{{{\rm{H}}}_{{\rm{2}}}{\rm{O}}}\right)$ flow rates of 16 slm and 94 μl min⁻¹, respectively. As schematically shown in Fig. 1, the humidified-air-plasma device consists of a dielectric barrier discharge (DBD) plasma section operated at a 36–38 W discharge-coupling power and an air and water supply system. It should be mentioned that to scale the plasma-induced effects to remote targets, the plasma-generated RONS are transported via gas-flow convection, labeled as a plasma effluent gas (PEG) in Fig. 1. Further details of the humidified-air-plasma devices can be found in our previous works. ^15,24)

The densities of the gas phase RONS (O_3gas, NO_gas, NO_2gas, N₂O_5gas, N₂O_gas, H₂O_gas, H₂O_2gas, HNO_3gas) were measured ^15,24) using Fourier-transform infrared (FTIR) absorption spectroscopy with a 3.6 m optical-path gas cell; the typical RONS densities for the given conditions are shown in Table I along with the limits of detection. For the given conditions, the major RONS are O_3gas, H₂O_2gas, N₂O_5gas, HNO_3gas, and N₂O_gas. It should be noted that the RONS densities for the PEG described in Table I are measured by sampling the entire PEG at the exit of the humidified-air-plasma device, followed by mixing with ambient air, which dilutes the RONS densities by a factor of approximately ten 10 cm downstream from the target.

Seeds of Arabidopsis thaliana ecotype Columbia, were sown on hydrated soil, followed by vernalization in the dark at 4 °C for two to three days before germination, and grown at 23 °C under fluorescent light with a 10/14 h light/dark cycle in a growth chamber. After the cotyledons were fully opened (within 2 weeks from sowing), plants that were growing well were moved to individual experimental soil pods and then grown on for five to eight weeks $\left({T}_{{\rm{grown}}}\right).$ Three samples were exposed to the PEG to cause either plasma-generated RONS stress or were exposed to air flow as a control 10 cm downstream of the humidified-air-plasma jet's exit.

Three rosette leaves of each A. thaliana sample were cut within one hour of the PEG exposure (or air-flow exposure, in the case of control samples) and then immersed into 30 ml of 0.4 M sorbitol solution at a typical pH of 5.4 in polypropylene tubes for a specified sorbitol soaking time from 1 to 24 h. This caused exudate from the leaves, including electrolytes and nonelectrolytes. Large mature leaves at the scale of the overall rosette, i.e. outer leaves, and relatively small young leaves grown within a 2 cm diameter from the rosette's center, i.e. inner leaves, were sampled separately in the sorbitol solution. After the soaking period, the electrical conductivity of the leaves-soaked sorbitol solution $E{C}_{{\rm{sample}}}$ was measured to evaluate the electrolyte exudate that leaked from the leaves using a conductivity meter (ES-71, HORIBA) and a temperature-compensated electrode (9382-10D, HORIBA) with a cell constant of 1 cm⁻¹ for a measurement range from 1 μS cm⁻¹ to 100 mS cm⁻¹. The leaves-soaked sorbitol solutions were immersed in a boiling water bath for 25 min to extract the exudate from the leaves by heat treatment. The leaves-soaked sorbitol solutions were subsequently cooled at room temperature for the electrical conductivity measurement $\left(E{C}_{{\rm{total}}}\right)$ to evaluate the extracted electrolyte in the exudate. Typical values of $E{C}_{{\rm{total}}}$ for the inner and outer leaf samples were approximately 20 μS cm⁻¹ and 110–150 μS cm⁻¹, respectively. To compensate for the leaked electrolyte variation due to individual differences in the masses of the sampled leaves, the electrolyte leakage rate $\left(ELR\right)$ was calculated as follows:

$$\begin{eqnarray}&&ELR=100\,\displaystyle \frac{E{C}_{{\rm{sample}}}-E{C}_{{\rm{blank}}.{\rm{s}}}}{E{C}_{{\rm{total}}}-E{C}_{{\rm{blank}}.{\rm{t}}}}\,\left[{\rm{ \% }}\right],\end{eqnarray} \tag{ 1 }$$

where $E{C}_{{\rm{blank}}.{\rm{s}}}$ and $E{C}_{{\rm{blank}}.{\rm{t}}}$ are the measured electrical conductivities of the blank sorbitol solutions before and after the boiling-water immersion process, typically only 1 ∼ 3 μS cm⁻¹, respectively.

The non-linearity of the EC values relative to the electrolyte exudate concentration was tested by diluting the sorbitol solutions, and was mostly below 5%, including a measurement uncertainty of ± 1 μS cm⁻¹, thus the measured EC values and the ELR are considered to be linear to the concentration of the electrolyte exudate from the leaves, but it can be influenced by the ion composition in the exudate.

The ionic composition of the leaves-soaked sorbitol solution was analyzed by an ion chromatograph (DX-120, Dionex) with a 20 μl sample loop and a conductivity detector with a suppressor, provided by the Technical Division of the School of Engineering, Tohoku University. Three of the leaves-soaked sorbitol solutions from each A. thaliana sample were mixed for averaging purposes, followed by double dilution and PTFE-membrane filtering with a 0.22 μm pore size, prior to ion chromatography. Among a number of electrolytes, inorganic cations and anions were analyzed: Li⁺, NH₄ ⁺, Na⁺, K⁺, Ca²⁺, Mg²⁺, and F⁻, Cl⁻, SO₄ ²⁻, NO₃ ⁻, NO₂ ⁻, PO₄ ³⁻(Pi⁻), respectively. The ion concentrations in the leaves-soaked sorbitol solution for the inner and outer leaves, extracted by heat treatment, are shown in Fig. 2(a). The error bars are the sample's standard deviation (SD). The concentrations of NO₂ ⁻, F⁻, and Li⁺ remained below the detection limit of 0.05 mg l⁻¹ for the given samples. The measured ions contained in the heat-treated exudate typically account for 80% of the measured electrical conductivity, estimated using the limiting ion conductivity in water. ²⁹⁾ Therefore, there were undetected electrolytes in the leaves-soaked sorbitol solutions. It should be mentioned that while PO₄ ³⁻ was analyzed in the ion chromatograph, the conductivity estimation was made using H₂PO₄ ⁻, considering the sorbitol solution's pH of 5.4 and the pK_a values for the phosphoric acid of 2.14, 7.2, and 12.4.

Zoom In Zoom Out Reset image size

Three normalized ion concentrations are introduced to discuss the ion leakage due to plasma-generated RONS. The measured ion concentrations in molar units, converted from the measured values in mg l⁻¹, are totaled to obtain the total ion concentration. The extracted ion-composition rate is then calculated by normalizing the extracted ion concentrations by the total extracted-ion concentration [Fig. 2(b)]. In contrast, the leaked-ion concentrations are normalized with the total leaked-ion concentration to obtain the leaked-ion-composition rate. The leaked-ion concentrations are also normalized by the total extracted-ion concentration [normalized leaked ions (NLI)]. It should be noted that NLIs are different from the leaked-ion composition, dividing the individual leaked ions by the sum of the "leaked" ion concentrations (the total leaked-ion concentration).

A. thaliana samples with ${T}_{{\rm{grown}}}$ = 5 to 8 weeks are exposed to the plasma effluent gas (PEG) generated by the humidified-air-plasma device characterized in Table I. The PEG exposure time, ${t}_{e}$, is varied to modulate the RONS exposure stress, while the control samples are exposed to an air flow for 60 s. Right after the PEG exposure, no apparent change was observed, while within 24 h, some leaves appeared to be wilted, as shown in Figs. 3(a) and 3(b) for ${t}_{e}\geqslant 30{\rm{s}}.$ The leaves identified as wilting are indicated with red-flipped triangles in Fig. 3(b), which is a close-up view of Fig. 3(a). The outer leaves, defined as the mature rosette leaves, tend to show wilting damage symptoms. In contrast, the appearance change due to PEG-exposure stress was not clear for the inner leaves, i.e. the younger growing leaves within the dashed circle in Fig. 3(b).

Zoom In Zoom Out Reset image size

The RONS stress to the inner leaves, which do not show the symptoms of wilting damage at 24 h after the PEG exposure, could be significant, because the PEG flow, aligned to the rosette's center, has a characteristic diameter of approximately 20 mm even with a free jet, i.e. without the A. thaliana pod at 10 cm downstream. On the other hand, the RONS stress to the outer leaves may be similar to that of the inner leaves. With the A. thaliana pod in place, the PEG flow impinges on the soil's surface, and therefore the RONS densities over the soil's surface may be similar to those at the center of the PEG flow, while the upper side of the outer leaves might be exposed to lower RONS densities due to mixing with the ambient air. Therefore, the RONS stress to the inner leaves is expected to be similar or more severe than that to the outer leaves.

Figure 3(c) shows that the $ELR{\rm{s}}$ for the inner leaves from the ${T}_{{\rm{grown}}}$ = 7 weeks' samples are significantly higher for ${t}_{e}\geqslant 30$ s, after which the corresponding outer leaves are wilted. In addition, the $ELR$ tends to increase with ${t}_{e}$ for all ${T}_{{\rm{grown}}}$ conditions. Therefore, this significant $ELR$ for the inner leaves appears to coincide with the plasma-generated RONS stress.

Figure 4 shows the $ELR{\rm{s}}$ for both the inner and outer leaves from ${T}_{{\rm{grown}}}$ = 7-week samples for various sorbitol soaking times. The measured $ELR{\rm{s}}$ for both the inner and outer leaves are found to become significant relative to the control within one hour and then tend to increase with soaking time. This increasing trend of the $ELR{\rm{s}}$ with sorbitol-soaking time may result from the plant's response to the PEG exposure and/or the additional stress due to the sorbitol soaking. The $ELR$ for the outer leaves is also significant at one hour for ${t}_{e}$ = 30 s, whereas wilting is observed at 24 h. The lower $ELR$ level for the outer leaves may be due to the non-uniformity of the PEG flow and/or attributed to the growth stage of the leaves, as discussed in later sections. From Figs. 3 and 4, it can be concluded that the applied PEG exposure stresses can induce electrolyte leakage from A. thaliana and that the $ELR$ for the inner leaves might correlate well with the PEG exposure stress applied.

Zoom In Zoom Out Reset image size

It should be mentioned that, while the $ELR$ level appears not to be significantly dependent on ${T}_{{\rm{grown}}}$ in Fig. 3, wilting of the outer leaves was barely observed for ${T}_{{\rm{grown}}}$ = 8 weeks after 24 h. Also, the inner leaves, for which the expected RONS stress could be more severe than that of the outer leaves, were not wilted for ${t}_{e}=30\,{\rm{s}}$ within 24 h. Furthermore, the severe wilting is observed at the outer leaves, but the $ELR$ levels for the inner leaves are higher than those for the outer leaves. Therefore, the observed wilting-damage symptoms can hardly be explained by the $ELR$ level and do not completely reflect the RONS stress. An interpretation may be that a plant stress response to the plasma-generated RONS stress turns out to be visible as the wilting-damage symptom. This interpretation indicates that there are factors leading to the wilting other than the RONS stress; which would also not conflict with the observed $ELR$ levels in Fig. 4, which show a significant $ELR$ for the outer leaves at one hour, when clear wilting is not observed. Thus, the $ELR$ appears to have an advantage over the observation of wilting in evaluating RONS stress.

The ion compositions in the leaked exudate from the leaves are analyzed as described in Sect. 2.4. The normalized leaked ion (NLI) concentrations in Fig. 5(a) shows that the total leaked-ion concentrations (height of the stacked bar) appear to be closely related to the $ELR{\rm{s}}$ in Fig. 4. It should be mentioned that the detected leaked ions in Fig. 5(a) account for typically 50% to 80% of the measured electrical conductivity in Fig. 4, which indicates that the major ions in the leaked exudate are measured but that there are also undetected electrolytes.

Zoom In Zoom Out Reset image size

Figure 5(b) shows that the leaked-ion compositions corresponding to Fig. 5(a). The major leaked ions for all the PEG-exposed samples are K⁺, Na⁺, NO₃ ⁻, and SO₄ ²⁻. The sorbitol soaking time tends to increase the leaked-ion concentrations. Ca²⁺ and Pi⁻ then become detectable at 16 h, which may result from the plant's response to PEG stress and/or the additional sorbitol soaking stress, and the detection limit of the ion chromatograph.

Control samples for 1 and 3 h only show ion concentrations for NO₃ ⁻ and Na⁺ at around the detection limit of 0.05 mg l⁻¹. The leaked ions from the control samples at 16 h include Ca²⁺ and Mg²⁺ for both the inner and outer leaves and Pi⁻ and SO₄ ²⁻ for the outer leaves. On the other hand, K⁺, NH₄ ⁺, and Cl⁻ are below the detection limit for both controls and Pi⁻ and SO₄ ²⁻ are also below the detection limit for the control of the inner leaves. Therefore, with only the sorbitol-soaking stress, the leakage of K⁺, NH₄ ⁺, Cl⁻, Pi⁻, and SO₄ ²⁻ from the leaves makes a minor contribution. In addition, it appears that K⁺ is found to be the most dominant leaked ion for the PEG-exposed samples, which accounts for nearly 40% of the measured electrical conductivity.

The electrolyte leakage due to drought, high- and low-temperature stresses has been discussed with the membrane competence. ^{25–28,30–34)} For plasma-induced electrolyte leakage, the interpretation of the $ELR$ is not well known but the selective K⁺ leakage is similar to that observed for other stress sources. ³⁰⁾ To resolve the relation between the plasma-generated RONS stress and the leaked electrolyte, it is important to compare the RONS-induced leaked electrolyte composition with the leaked-ion composition caused by other types of stress or damage. Three outer leaves are gently compressed for a second to provide physical compression stress, followed by a sorbitol soaking similar to that used for other PEG-stress samples. The total ion concentration was 1.6 mg l⁻¹, similar to the typical value of 2 mg l⁻¹ obtained for the PEG-stress samples. The ion compositions extracted by heat treatment for the same physically stressed samples are also analyzed, as shown in Fig. 6. Figure 6 shows that the ions leaked due to physical stress contain all the ions in the extracted exudate at a detectable level and the leaked K⁺ composition is found to be below 30%, nearly half of that of the PEG samples in Fig. 5(b) and lower than the extracted K⁺ composition in Fig. 6. It should be mentioned that the K⁺ ion composition in Fig. 5(b) is nearly 80% and higher than the extracted-ion composition in Fig. 2, obtained from the same samples. Therefore, the electrolytes leaked due to the PEG exposure can hardly be explained by physical stress or damage alone.

Zoom In Zoom Out Reset image size

The $ELR$ has been called an "injury index" in former works that infer damage to membrane permeability from drought and temperature stresses, but it is known to be a separate index from cell viability. ^25–28) The given interpretation was that cells were injured but not killed, based on experiments on passive osmosis-based transport related to K⁺ and sugar. Recent reports on membrane transporters, including ion channels, indicate that a significant amount of K⁺ efflux can occur with ROS stresses such as those due to O₃, hydrogen peroxide (H₂O₂), and hydroxyl (OH) radicals. ^30–33) The PEG in Table I predominantly contains O_3gas and includes H₂O_2gas, and supplies precursors for HOONO_aq and HOONO_2aq, a potential donor of OH and HO₂ radicals. ^16,18) Therefore, the observed selective K⁺ leakage due to the PEG stress may also be interpreted as a result of the RONS stress response, and not only due to cell death. Because there are experimental reports of a plasma-generated OH-mediated Ca⁺ influx through the ion channel in animal cells right after plasma treatment, ^34,35) one hypothesis for the observed ion leakage due to plasma-generated RONS stress could be that the plasma-generated RONS, when transferred to the plant cell membrane, affect the ion transporters related to K⁺. This hypothesis does not conflict with the observation that the $ELR$ correlates with the plasma-generated RONS stress and miscorrelates with the subsequent wilting, which would have other additional factors.

The $ELR$ measurement is easy to perform and therefore it can be utilized for quantifying plasma-generated RONS stress, but some limitations have become clear in this paper. In Fig. 3, for the same ${t}_{e},$ the $ELR$ levels tend to decrease with ${T}_{{\rm{grown}}}.$ In Fig. 4, the $ELR$s for the outer leaves remain significantly lower than those of the inner leaves, which could also be due to a lower RONS stress, but the $ELR{\rm{s}}$ for the outer leaves do not exceed more than 50%, even at ${t}_{e}$ = 60 s. Furthermore, in Figs. 3 and 4, the $ELR$ levels appear to be different for exactly the same PEG-exposure conditions: ${T}_{{\rm{grown}}}$ = 7 weeks and ${t}_{e}$ = 30 s, though the difference of the $ELR{\rm{s}}$ is not significant (p > 0.05). This indicates that the absolute values of $ELR$ might not completely reflect the plasma-generated RONS stress alone, but also other factors at work in plants.

An explanation of the observed limitations can be made using the average volume of the vacuole, which may occupy the major part of the cell's volume in mature leaves. Supposing that the RONS stress mainly or primarily affects cell-membrane transport, as discussed for the ion-selective leakage, it can be speculated that the electrolytes in vacuoles bounded by tonoplasts might not leak out immediately, and therefore the $ELR{\rm{s}}$ become smaller due to a higher concentration of the total extracted electrolyte for the outer mature leaves. This proposed hypothesis can explain the $ELR$ difference due to ${T}_{{\rm{grown}}}$ and the inner and outer leaves in Figs. 3 and 4.

Other causes of the $ELR$ variation might be attributed to sample preparation. In Figs. 2 and 6(b), the extracted ion compositions are similar but not exact, e.g. NO₃ ⁻, which may be due to some minor differences in preparing the A. thaliana samples. Considering the ion selective-leakage mechanism, the ion composition of A. thaliana leaves before RONS stress may be an important factor. The variation might be also influenced by time-measurement delay, due to the sequential EC measurements.

Some of the above-mentioned limitations can be avoided by carefully preparing a considerable number of samples, choosing relevant leaves, and completing tests at once. The $ELR$ can then be utilized for quantifying the plasma-generated RONS stress and thereby could also be used to characterize the plasma-generated RONS supplied to plants.