Photo-Protective Mechanisms and the Role of Poly (ADP-Ribose) Polymerase Activity in a Facultative CAM Plant Exposed to Long-Term Water Deprivation

The Crassulacean acid metabolism (CAM) pathway helps plants to alleviate the oxidative stress under drought, but the shift to CAM-idling may expose plants to the overproduction of reactive oxygen species causing cell damages. The facultative CAM species Portulacaria afra L., was subjected to long-term water deprivation to assess the photo-protective strategies and the poly (ADP-ribose) polymerase (PARP) activity during water stress and plant capability to recover from the stress. Measurements of titratable acidity, chlorophyll fluorescence emission, and antioxidant activity were performed during the stress and rewatering. Under water deprivation, plants shifted from C3 to CAM metabolism, reaching the CAM-idling status at the end of the stress period. The daily variation of the titratable acidity and PARP activity increased at the beginning of stress and declined with stress progression, reaching the lowest value at the end of stress treatment. H2O2 content, superoxide dismutase (SOD), peroxidase (POD), and catalase (CAT) activities increased with the severity of water stress. The photochemical processes remained high during the entire stress period indicating the presence of alternative sinks to CO2 fixation. The elevated activity of catalase under severe water stress suggests the occurrence of photorespiration in sustaining the photosynthetic electron transport under CAM-idling condition. The overall data indicate that scavenger enzymes, photorespiration and PARP activity modulation contribute to the strong resistance of P. afra to severe water stress, preserving the functioning of photosynthetic apparatus and ensuring plant recovery with rewatering.


Introduction
Crassulacean acid metabolism (CAM) represents a carbon fixation pathway typical of plants adapted to arid environments. In CAM plants the stomata are open during night and the CO 2 uptake occurs with organic acids synthesis (viz. malate). The CO 2 fixation is mediated by phosphoenolpyruvate carboxylase, PEPC (phase I). In the morning, during the early light period the simultaneous CO 2 fixation by PEPC and Rubisco occurs while stomata start to close (phase II). During the day, stomata are closed, and organic acids stored during night are decarboxylated. The released CO 2 is re-fixed by Rubisco and assimilated in C3 pathway (phase III). In the late light period, stomata reopen and

Leaf Water Status and Photosynthetic Pigment Content
The relative water content, RWC, decreased with the increase of the water stress ( Table 1). The lowest value (47%) (p < 0.01) was reached at 79 Days Without Water (DWW) and remained constant (47%) up to 94 DWW. With plant rewatering, RWC full recovered, returning to pre-stress values (89%). Photosynthetic pigment content (total chlorophyll and carotenoids), was not affected by water stress and remained almost constant during the whole stress period and plant rewatering, ranging from 30.1 ± 1.1 and 38.7 ± 2.0 µg cm −2 for total chlorophylls, and 7.5 ± 0.3 and 9.1 ± 0.4 µg cm −2 for total carotenoids. Table 1. Leaf relative water content (RWC) before (0 days) and during (60-94 days) water stress, and after the rewatering period of 14 day. Data are means (n = 10) ± SE.

Days Without Water (DWW)
Relative Water Content (%) Different letters indicate statistically significant differences (p ≤ 0.05) among days. Data were analyzed by one-way ANOVA repeated measurements followed by the Tukey's test.

Titratable Acidity Fluctuation
Before the water stress, plants showed a small fluctuation in titratable acidity from morning to night, suggesting that the C3 cycle was the main functioning photosynthetic pathway in P. afra and that limited CAM activity (likely CAM-cycling) occurred too ( Figure 1). Under stress, P. afra plants showed a large fluctuation of titratable acidity from morning to night (p < 0.01). The titratable acidity variation peaked (120 µeq NaOH g FW −1 ) at 15 DWW, as the consequence of the increased daytime organic acid degradation produced and accumulated during the night ( Figure 1). Our data suggest an active CAM pathway in this circumstance. With the progression of water stress, the fluctuation of titratable acidity decreased (p < 0.01) for the reduction of the nocturnal acid accumulation, reaching the lowest value (20-30 µeq NaOH g −1 FW) at 100 DWW. At this time P. afra plants seem to reach the CAM-idling status. With rewatering, the acid accumulation during the night significantly raised, determining a marked increase of daily variation of titratable acidity, which suggested that the CAM pathway was fully active.

Fluorescence Measurements
The quantum yield of photosystem II electron transport (Φ PSII ) ( Figure 2A) and the electron transport rate (ETR) ( Figure 2B) progressively decreased from 0 to 16 DWW (p < 0.01). With the progression of water stress, no further significant decline of Φ PSII and ETR was observed compared to the value reached at 16 DWW.
The photochemical quenching, qL, raised at 16 DWW (p < 0.01) and remained almost constant for the remaining DWW ( Figure 2C). The quantum yield of regulated energy dissipation (Φ NPQ ) increased few days after the imposition of water stress up to 60 DWW, then it decreased to pre-stress values ( Figure 2D). The quantum yield of non-regulated energy dissipation (Φ NO ) progressively raised with the severity of water stress, reaching the highest value (p < 0.01) at 80 and 100 DWW ( Figure 2E). On the contrary, the maximal PSII photochemical efficiency, F v /F m , gradually declined (p < 0.01) as water stress increased. The lowest value of F v /F m was observed at 100 DWW ( Figure 2F). Measurements carried out 14 days after plant rewatering showed the recovery of all photochemical parameters compared to Plants 2020, 9, 1192 4 of 14 pre-stress conditions (0 DWW) except for qL, which remained higher and Φ NO , which stayed lower compared to 0 DWW ( Figure 2).

H 2 O 2 Content, Non-Enzymatic and Enzymatic Antioxidants
The progressive water deprivation determined a significant increase (p < 0.01) of the leaf hydrogen peroxide (H 2 O 2 ) content ( Figure 3A). The maximum concentration was found at 56 DWW. At 100 DWW; no further increase of H 2 O 2 level was measured compared to 56 DWW.
The enzymatic ( Figure 3B-D) and non-enzymatic ( Figure 4) antioxidants exhibited a different trend. The activity of the antioxidant enzymes superoxide dismutase (SOD), peroxidase (POD), and catalase (CAT) significantly increased (p < 0.01) just after a few days from water stress imposition compared to pre-stress condition ( Figure 3B-D), showing a maximum (p < 0.01) of activity at 100 DWW. With the rewatering, the H 2 O 2 content decreased to pre-stress values, while the antioxidant enzyme activity remained higher and comparable to that measured at 100 DWW ( Figure 3).
On the contrary, the water-soluble ( Figure 4A) and lipid-soluble ( Figure 4B) antioxidant capacity decreased as water stress increased, showing a reduction of 42% (p < 0.01) at 100 DWW compared to 0 DWW. With rewatering, the water-soluble and lipid-soluble antioxidant capacity did not recover to pre-stress values but remained lower. In particular, the water-soluble antioxidant capacity did not change, while the lipid-soluble antioxidant capacity weakly increased (p < 0.01) compared to values measured at 100 DWW ( Figure 4).

PARP Activity
The PARP activity showed a similar trend compared to that observed for the daily fluctuation of titratable acidity ( Figure 5). More specifically, the PARP activity peaked at 15 DWW (p < 0.01) and progressively declined until 100 DWW, when plants were likely to be found in CAM-idling status. The lowest enzymatic activity was measured at 100 DWW. With rewatering, the PARP activity significantly increased (p < 0.001) compared to values measured at 0 and 100 DWW ( Figure 5).  Different letters indicate statistically significant differences in morning or evening acid content (p ≤ 0.05). Data were analyzed by one-way ANOVA repeated measurements followed by the Tukey's test. Data are means (n = 10) ± SE.

Fluorescence Measurements
The quantum yield of photosystem II electron transport (ΦPSII) (Figure 2A) and the electron transport rate (ETR) ( Figure 2B) progressively decreased from 0 to 16 DWW (p < 0.01). With the progression of water stress, no further significant decline of ΦPSII and ETR was observed compared to the value reached at 16 DWW.
The photochemical quenching, qL, raised at 16 DWW (p < 0.01) and remained almost constant for the remaining DWW ( Figure 2C). The quantum yield of regulated energy dissipation (ΦNPQ) Different letters indicate statistically significant differences in morning or evening acid content (p ≤ 0.05). Data were analyzed by one-way ANOVA repeated measurements followed by the Tukey's test. Data are means (n = 10) ± SE.

Correlations among Investigated Parameters
The correlation among the investigated parameters were reported in Table S1. Φ PSII showed a negative correlation with Φ NO (r = −0.968) and H 2 O 2 (r = −0.815) and a positive correlation with F v /F m (r = 0.974). Φ NO was negatively correlated to ETR (r = −0.968) and F v /F m (r = −0.994). SOD was positively correlated to POD (r = 0.980) and CAT (r = 0.987) and negatively correlated to water-soluble (r = −0.988) and lipo-soluble (r = −0.967) antioxidants. POD exhibited a positive correlation with CAT (r = 0.997) and a negative correlation with water-soluble (r = −0.983) and lipo-soluble (r = −0.970) antioxidants. CAT was negatively correlated to water-soluble (r = −0.975) and lipo-soluble (r = −0.968) antioxidants. Water-soluble and lipo-soluble antioxidants showed a positive correlation each other (r = 0.973).  The enzymatic ( Figure 3B-D) and non-enzymatic ( Figure 4) antioxidants exhibited a different trend. The activity of the antioxidant enzymes superoxide dismutase (SOD), peroxidase (POD), and catalase (CAT) significantly increased (p < 0.01) just after a few days from water stress imposition compared to pre-stress condition ( Figure 3B-D), showing a maximum (p < 0.01) of activity at 100 DWW. With the rewatering, the H2O2 content decreased to pre-stress values, while the antioxidant enzyme activity remained higher and comparable to that measured at 100 DWW ( Figure 3).
On the contrary, the water-soluble ( Figure 4A) and lipid-soluble ( Figure 4B) antioxidant capacity decreased as water stress increased, showing a reduction of 42% (p < 0.01) at 100 DWW compared to 0 DWW. With rewatering, the water-soluble and lipid-soluble antioxidant capacity did not recover to pre-stress values but remained lower. In particular, the water-soluble antioxidant capacity did not change, while the lipid-soluble antioxidant capacity weakly increased (p < 0.01) compared to values measured at 100 DWW ( Figure 4).

PARP Activity
The PARP activity showed a similar trend compared to that observed for the daily fluctuation of titratable acidity ( Figure 5). More specifically, the PARP activity peaked at 15 DWW (p < 0.01) and progressively declined until 100 DWW, when plants were likely to be found in CAM-idling status. The lowest enzymatic activity was measured at 100 DWW. With rewatering, the PARP activity significantly increased (p < 0.001) compared to values measured at 0 and 100 DWW ( Figure 5). The PARP activity showed a similar trend compared to that observed for the daily fluctuation of titratable acidity ( Figure 5). More specifically, the PARP activity peaked at 15 DWW (p < 0.01) and progressively declined until 100 DWW, when plants were likely to be found in CAM-idling status. The lowest enzymatic activity was measured at 100 DWW. With rewatering, the PARP activity significantly increased (p < 0.001) compared to values measured at 0 and 100 DWW ( Figure 5).

Discussion
The CO 2 concentration and the stomatal closure in phase III during the day result in the increase of water use efficiency in CAM plants but also in a potential oxidative stress, due to the sustained photosynthetic electron transport and the rise of internal O 2 concentration [21]. This condition is particularly dangerous under CAM-idling status, when gas exchanges are suppressed, the accumulation and decarboxylation of organic acids (i.e., malate/citrate) as well as the CO 2 fixation by Rubisco are reduced. In our experiment, the C3 pathway was the main functioning photosynthetic pathway in unstressed P. afra plants, while the reduced daily acid fluctuation is attributable to the CAM-cycling operating with absence of nocturnal CO 2 uptake [20].
When exposed to water stress, plants likely shifted to CAM right after water stress imposition and to CAM-idling at the end of the stress period (80-100 DWW). This behavior is suggested by the daily fluctuation in titratable acidity, which showed the lowest value (30 µeq NaOH g −1 FW) at 80-100 DWW, in concomitance with the lowest leaf RWC. Daily measurement of gas exchanges could have confirmed the evidence of both CAM and CAM-idling status in P. afra plants in response to prolonged water stress. However, even if we could not measure the gas exchanges, our hypothesis is supported by the recent work of Guralnick et al. [22], which demonstrated the development of CAM activity through the titratable acidity levels and diurnal acid fluctuations. The daily acid fluctuation measured in the present study is in agreement with values found by Guralnick and Ting [20] in P. afra plants exposed to prolonged water stress. These authors report values of daily fluctuation of titratable acidity of 40-50 µeq NaOH g −1 FW and no nocturnal gas exchange in stressed plants operating in CAM-idling mode. Based on this evidence, we hypothesized that P. afra plants at 100 DWW were in CAM-idling mode, while they moved to full CAM after rewatering under long photoperiod [20], on the basis of the greatest daily acid fluctuation measured after 14 days from water supply.
In our experimental conditions, endogenous H 2 O 2 content increases as water stress progresses, suggesting the occurrence of oxidative stress, even if the possibility that H 2 O 2 may act also as a signal involved in C3-CAM transition cannot be excluded [23]. In response to H 2 O 2 production, the activity of SOD, CAT, and POD scavenger enzymes promptly rose and remained high during the period of water stress and recovery. Our data agree with other studies that found an increase of antioxidant enzyme activity during the C3-CAM transition in response to mild water stress [7,24]. The unexpected result is the high level of CAT activity at 100 DWW when P. afra shifted likely to CAM-idling [24]. We suppose that the sustained CAT activity during the exacerbated water stress was effective in removing the H 2 O 2 also produced in the photorespiration [25]. Generally, catalase in peroxisomes neutralizes H 2 O 2 into O 2 and H 2 O [26], and its activity decreases when photorespiration is suppressed [27]. Thus, the elevated CAT activity in P. afra plants during the CAM-idling likely suggests the occurrence of photorespiration. It has been shown that some CAM species present a reduction in glycolate oxidase activity when they accumulate more malate, as a metabolic signal of photorespiration suppression [28,29]. As we observed a reduction in nocturnal organic acid (likely malic acid) accumulation with the progression of water stress, we supposed that photorespiration could be active during the CAM-idling status, contributing to keep oxidized the photosynthetic electron transport chain. The photorespiration could also provide CO 2 for the C3 metabolism during the daylight, as recently proposed by Guralnick and colleagues [22] for the C4 plant Portulaca grandiflora.
Converse to the enzymatic scavengers, the water-soluble and lipid-soluble antioxidant capacity decreased with the severity of stress. Generally, these antioxidants are increased in plants exposed to the water stress. Our data indicate that the non-enzymatic activities would not be responsible for controlling ROS when plants are in CAM and CAM-idling. The antioxidant system would have a protective function against ROS in plants which mainly utilize the C3 pathway. In fact, the antioxidant capacity remained low also after the rewatering, when plants were full CAM and maintained an elevated enzymatic antioxidant activity. Thus, we hypothesized that the enzymatic antioxidant activity would represent the main defense against ROS when plants shift from C3 to CAM-idling under severe water stress and to CAM after rewatering.
The high activity of scavenger enzymes, especially under strong water stress, suggests an intense activity of other photochemical processes than CO 2 fixation in sustaining the electron transport, as indicated by the high level of photochemical quenching (qL) [30]. The qL index started to increase soon after water stress imposition remaining elevated during the whole period of stress, despite Φ PSII and Φ NPQ reduction. With the progression of water stress, a gradual increase of quantum yield of non-regulated energy dissipation (Φ NO ) was also observed. This condition usually suggests the occurrence of higher photo-damage risks and is related to the rise of ROS production [31]. Based on our results, we assumed that the rise of Φ NO might represent a critical "regulatory" mechanism to guarantee the acclimation in P. afra under severe water stress, and more specifically under the CAM-idling status. On the other hand, the quantum yield of regulated energy dissipation (Φ NPQ ) seems to have a role in counteracting the water stress at the first time, as previously demonstrated by Cela et al. [10].
This "regulatory" mechanism might operate together with PARP activity and potentiate the plant defenses during the CAM-idling status when the risks of oxidative damages are elevated. The safety role of PARP enzymes in plants has been widely proved in several C3 species in response to different stress, such as high temperature, ionizing radiation, and drought [13,32,33]. However, it is the first time that the function of poly (ADP) ribosylation is considered in a CAM-inducible species, and in particular under CAM-CAM-idling transition. In response to prolonged water stress, the PARP activity in P. afra plants was downregulated and this reduction would allow leaves to become more tolerant to severe water stress. The amplified tolerance might occur by either keeping the cell energy at a high level through a reduction in NAD + breakdown and ATP overconsumption by PARP [13,14,34] or by ABA-dependent stress signaling [35]. The high NAD + levels as a consequence of the reduced PARP activity, would lead to the synthesis of the cyclic ADP-ribose (cADPR) responsible for the abscisic acid (ABA) production and hence ABA-responsive gene expression. Further research studies are needed to understand the primary or ancillary role of the PARP enzyme in the tolerance to abiotic stress in CAM plants as well as in the seasonal shift from C3 to CAM photosynthesis. Data reported in this study seem to indicate a close relationship between the photosynthetic metabolic pathway and PARP activity. The poly (ADP-ribose) polymerase activity was higher after rewatering compared unstressed plants. It is likely due to the fact that under full CAM the potential for the oxidative stress was greater than under C3 metabolism, as consequence of the elevated electron transport activity (ETR) and internal O 2 concentrations [2][3][4], which exposes plants to the oxidative stress. This would explain also the elevated activity of the enzymatic antioxidants under full CAM.

Plants Growth Condition and Experimental Design
Plants of Portulacaria afra (L.) Jacq. of three-years old were grown in a greenhouse under natural sunlight in pots of 5 L filled with a mixture of soil: peat (1:1 v:v) and irrigated to field water capacity. Ten plants were selected for water stress condition (water-stressed plants) and maintained without water up to 100 days (Days Without Water-DWW). Then, the stressed plants were watered for 14 days (RW-rewatering) to field water capacity to assess the plant capability to recover from stress. Trends of air temperature and relative humidity experienced by plants inside the greenhouse are shown in Figure 6. The physiological behavior of P. afra plants during stress and rewatering was monitored at 7, 16, 27, 34, 56, 70, 85, and 100 DWW and after 14 RW by the measuring of titratable acidity and chlorophyll fluorescence emission. In addition, in some critical moments of the water stress period (16,27,56, and 100 DWW) and during the recovery (14 RW), the plant capability to withstand the stress was evaluated following the enzymatic and non-enzymatic antioxidants and poly(ADPribose)polymerase activity (PARP).
All the biochemical and physiological analyses were conducted on mature current-year leaves collected at midday, when P. afra plants were likely in the phase III of CAM. Five independent leaves per plant were chosen for analyses and the results were averaged. Statistical analysis was carried out on the averages.

Leaf Water Status and Photosynthetic Pigment Content Measurement
The leaf water status was assessed measuring the relative water content RWC = (FW − DW)/(TW − DW), where FW and TW are the fresh and turgid weight, respectively. DW represents the dry weight after oven-drying the leaves at 80 °C to constant weight. TW was determined under a vaporsaturated atmosphere dipping the petiole in water and keeping leaves darkened at 4 °C for 24 h.
Photosynthetic pigments (total chlorophylls and carotenoids) were extracted from samples collected at midday, in ice-cold 100% acetone with a mortar and pestle and centrifuged at 5000 rpm for 5' (Labofuge GL, Heraeus Sepatech, Hanau, Germany). The absorbance of supernatants was quantified spectrophotometrically (UV-VIS Cary 100, Agilent Technologies, Santa Clara, CA, USA) at wavelengths of 470, 645, and 662 nm. The pigment content was calculated according to Lichtenthaler [36]. The physiological behavior of P. afra plants during stress and rewatering was monitored at 7, 16, 27, 34, 56, 70, 85, and 100 DWW and after 14 RW by the measuring of titratable acidity and chlorophyll fluorescence emission. In addition, in some critical moments of the water stress period (16,27,56, and 100 DWW) and during the recovery (14 RW), the plant capability to withstand the stress was evaluated following the enzymatic and non-enzymatic antioxidants and poly(ADP-ribose)polymerase activity (PARP).

Titratable Acidity Determination
All the biochemical and physiological analyses were conducted on mature current-year leaves collected at midday, when P. afra plants were likely in the phase III of CAM. Five independent leaves per plant were chosen for analyses and the results were averaged. Statistical analysis was carried out on the averages.

Leaf Water Status and Photosynthetic Pigment Content Measurement
The leaf water status was assessed measuring the relative water content RWC = (FW − DW)/ (TW − DW), where FW and TW are the fresh and turgid weight, respectively. DW represents the dry weight after oven-drying the leaves at 80 • C to constant weight. TW was determined under a vapor-saturated atmosphere dipping the petiole in water and keeping leaves darkened at 4 • C for 24 h.
Photosynthetic pigments (total chlorophylls and carotenoids) were extracted from samples collected at midday, in ice-cold 100% acetone with a mortar and pestle and centrifuged at 5000 rpm for 5' (Labofuge GL, Heraeus Sepatech, Hanau, Germany). The absorbance of supernatants was quantified spectrophotometrically (UV-VIS Cary 100, Agilent Technologies, Santa Clara, CA, USA) at wavelengths of 470, 645, and 662 nm. The pigment content was calculated according to Lichtenthaler [36].

Titratable Acidity Determination
For the determination of titratable acidity, eight leaves per plant were collected at the early morning and the evening. Leaves from the same plant were mixed together and grounded in de-ionized water with a mortar and pestle and titrated with 0.01 N NaOH to a pH 7.0 endpoint. Data were expressed as µeq NaOH g −1 FW.

Chlorophyll a Fluorescence Emission Measurements
Chlorophyll a fluorescence measurements were carried out at midday by a pulse amplitude modulated fluorometer (Junior-PAM, Walz, Germany). Rapid light response curves (RLC) were performed on attached leaves at different light intensities (125-1500 µmol m −2 s −1 ) to assess the efficiency of the photosynthetic apparatus in the utilization of the absorbed light in photochemistry. Leaves were darkened for 30' to determine the maximal PSII photochemical efficiency (F v /F m ). The basal fluorescence signal (F o ) was obtained by a weak beam of blue light at a frequency of 0.6 kHz, while the maximal fluorescence in the dark-adapted state (F m ) was measured by applying a 1-s light saturating pulse. During exposure to light, the steady-state fluorescence (F s ) was obtained by setting the measuring light to the frequency of 20 kHz; maximal fluorescence in the light-adapted state (F' m ) was measured by applying a 1-s light saturating pulse.
The partitioning of absorbed light energy was estimated according to the model proposed by Kramer et al. [37]. The allocation of photons absorbed by the PSII antennae to photosynthetic electron transport was estimated as Φ PSII = (F' m − F s )/F' m . The quantum yield of regulated energy dissipation and of non-regulated energy dissipation within PSII were calculated as Φ NPQ = 1 − Φ PSII − 1/ (NPQ + 1 + qL × (F m /F 0 − 1)) and Φ NO = 1/(NPQ + 1 + qL × (F m /F 0 − 1)), respectively. The coefficient of photochemical quenching, qL, was defined and estimated following Kramer et al. [37]. The electron transport rate (ETR) was calculated according to Krall and Edwards [38] as Φ PSII × PPFD × 0.5 × 0.84, considering the absorbed photon energy equally distributed between PSI and PSII and 0.84 the assumed light absorbance of the leaf.

Hydrogen Peroxide Determination
Among reactive oxygen species (ROS), the hydrogen peroxide (H 2 O 2 ) is a constitutive product within cells and the variation in its concentration may suggest the occurrence of possible stress conditions. The H 2 O 2 levels were measured following the colorimetric method described by Zhou et al. [39]. Briefly, leaves (0.3 g), frozen in liquid nitrogen, were grounded in a mortar with a pestle, in the presence of 3 mL of 5% TCA and 0.10 g activated charcoal. The homogenate was centrifuged at 10,000 rpm for 20 min at 4 • C, and the supernatant was added to 17 M ammonia solution and filtered. Then, a reaction mixture containing 1 mL of filtrate, 8 µg of catalase and 1 mL of colorimetric reagent was incubated for 10 min at 30 • C. The absorbance of the samples was determined spectrophotometrically at 505 nm. The colorimetric reagent contained 10 mg of 4-aminoantipyrine, 10 mg of phenol, 5 mg of peroxidase (150 U mg −1 ), dissolved in 50 mL of 100 mM acetic acid buffer (pH 5.6). The content of H 2 O 2 was calculated using a standard curve and expressed as µmol H 2 O 2 g −1 fresh weight (µmol H 2 O 2 g −1 FW).

Non-Enzymatic and Enzymatic Antioxidant Measurements
The water-soluble antioxidants were extracted in a mixture consisting of methanol, water, and formic acid (80:20:0.1) for 3 h in the dark. The supernatant (S1), obtained after centrifugation at 3500 rpm for 1 min, at 4 • C, which represents the hydrophilic extract, was stored at 4 • C. The pellet was subjected to two subsequent extractions. The hydrophilic extracts S1, S2, and S3, were combined and further centrifuged at 10,000 rpm for 5 min at 4 • C. The lipid-soluble antioxidants were extracted in acetone by performing the same experimental procedure above described. The soluble and fat-soluble extracts were directly used in the antioxidant assay. Free radical scavenging activity of whole examined samples was determined by 2,2'-azino-bis(3-ethylbenzthiazoline-6-sulphonic acid) (ABTS •+ ) radical cation decolorization assay as described by Re et al. [40] with minor modification. ABTS •+ cation radical was obtained by the reaction between 7 mM ABTS in water and 2.45 mM potassium persulfate (1:1). The reaction mixture was stored for 16 h in the dark at room temperature before use and utilized within two days. The ABTS •+ solution was diluted with methanol, to obtain an absorbance of 0.700 ± 0.050 at 734 nm. Then 5 µL of plant extract was added to 3.995 mL of diluted ABTS •+ solution, and the absorbance was measured at 734 nm. An appropriate solvent blank was run in each assay. All the measurements were carried out at least three times. Trolox solution (final concentration 0-15 µM) was used as a reference standard. The results were expressed as µmol Trolox g −1 fresh weight of tissue.

Poly (ADP-Ribose) Polymerase (PARP) Assay Activity
The ADP-ribosylating activity was determined as reported in Arena et al. [15]. In brief, the assay was carried out on isolated nuclei incubated in presence of 0.4 mM [ 32 P] NAD + (10,000 cpm/nmole) in a 500 mM Tris-HCl buffer, pH 8.0; 50 mM MgCl 2 and 10 mM DTT (reaction mixture). The reaction was stopped by adding ice-cold 20% trichloroacetic acid (w/v). Then, the mixture was filtered on Millipore filters (HAWPP0001, 0.45 µm) and subjected to various washes using 7% trichloroacetic acid. The radioactivity of insoluble acid material associated with the filter was measured in a liquid phase scintillator (Bechman LS 1701). The enzyme activity was expressed in enzymatic milliunit; 1 mU catalyzes the synthesis of 1 nmol ADP-ribose/min, at the optimal pH and temperature.

Statistical Analysis
Statistical analysis and graphical elaboration of the data were performed by of Sigma-Plot 12.0 software package (Jandel Scientific, San Rafael, CA, USA). Data were analyzed by one-way ANOVA repeated measurements followed by the Tukey's test. The results are reported as mean (n = 10) ± SE. Differences were considered statistically significant at p ≤ 0.05. Shapiro-Wilk and Kolmogorov-Smirnov tests were performed to check for normality. The correlations between selected parameters were investigated using Pearson's correlation test.

Conclusions
Our results showed that P. afra plants counteract the water stress by defense mechanisms operating both at the physiological and biochemical level. The scavenger enzymes and likely the photorespiration, dissipating the excess of absorbed light when the photosynthetic metabolism is down-regulated by severe water stress, preserve the photosynthetic apparatus integrity and guarantee a prompt recovery with the replacement of the optimal plant water status. With rewatering, the maximum photochemical efficiency is fully recovered, and H 2 O 2 and Φ NO are dropped to pre-stress values, confirming the relationship between hydrogen peroxide production and the non-regulated energy dissipation mechanisms. The regulation of PARP and antioxidant enzyme activity in synergy with the water-soluble and lipid-soluble antioxidant capacity and PSII photoprotective strategies, guarantees the tolerance of P. afra to severe and prolonged water stress. These strategies ensure a rapid physiological recovery with rewatering. Our results suggest that the PARP activity is modulated depending on the operating photosynthetic pathway.
Author Contributions: Conceptualization, L.V. and C.A.; investigation, L.V., E.V., G.C., A.D.M. and C.A.; writing-original draft preparation, L.V.; writing-review and editing, L.V. and C.A. All authors have read and agreed to the published version of the manuscript.
Funding: This research received no external funding.