Abstract
The study reports on four different types of flag leaf rolling under soil drought in relation to the level of cell wall-bound phenolics. The flag leaf colonization by aphids, as a possible bioindicator of the accumulation of cell wall-bound phenolics, was also estimated.
The proteins of the photosynthetic apparatus that form its core and are crucial for maintaining its stability (D1/PsbA protein), limit destructive effects of light (PsbS, a protein binding carotenoids in the antennas) and participate in efficient electron transport between photosystems II (PSII) and PSI (Rieske iron–sulfur protein of the cytochrome b6f complex) were evaluated in two types of flag leaf rolling. Additionally, biochemical and physiological reactions to drought stress in rolling and non-rolling flag leaves were compared.
The study identified four types of genome-related types of flag leaf rolling. The biochemical basis for these differences was a different number of phenolic molecules incorporated into polycarbohydrate structures of the cell wall. In an extreme case of non-rolling dehydrated flag leaves, they were found to accumulate high amounts of cell wall-bound phenolics that limited cell water loss and protected the photosynthetic apparatus against excessive light. PSII was also additionally protected against excess light by the accumulation of photosynthetic apparatus proteins that ensured stable and efficient transport of excitation energy beyond PSII and its dissipation as far-red fluorescence and heat. Our analysis revealed a new type of flag leaf rolling brought about by an interaction between wheat and rye genomes, and resulting in biochemical specialization of flexible, rolling and rigid, non-rolling parts of the flag leaf. The study confirmed limited aphid colonization of the flag leaves with enhanced content of cell wall-bound phenolics.
Non-rolling leaves developed effective adaptation mechanisms to reduce both water loss and photoinhibitory damage to the photosynthetic apparatus under drought stress.
Introduction
Plants have developed several defence mechanisms against drought stress (Gadzinowska et al., 2019; Khan et al., 2019). One of them is inward leaf rolling especially visible in grasses and cereals (Kadioglu and Terzi, 2007). This mechanism increases drought resistance, e.g. in Ctenanthe setosa, a perennial herbaceous plant serving as a suitable model for studies on leaf rolling (Nar et al., 2009). Leaf rolling results from a variable degree of dehydration in different sites of a leaf cross-section (Baret et al., 2018), and is an effective mechanism limiting light capture, transpiration, leaf dehydration and aphid feeding damage (Kadioglu et al., 2012).
Drought stress makes the photosynthetic apparatus more sensitive to light (Valladares and Pearcy, 2002; Murata et al., 2007; Guidi et al., 2019). It was shown that the behaviour of the photosynthetic apparatus during leaf dehydration can be also interpreted based on the level of proteins [e.g. PsbS, a light-harvesting complex (LHC)-like photosystem II (PSII) protein that binds carotenoids in antennas; PsbA/D1, a protein of PSII; and PetC/Rieske iron–sulfur protein of the cytochtome (cyt) b6f complex] involved in photochemical reactions (Georgieva et al., 2009; Hura et al., 2018, 2019; Wu et al., 2020). The major component of PSII affected by water stress is D1 protein (Huang et al., 2019). Liu et al. (2006) demonstrated that the content of D1 protein in two wheat cultivars declined with increasing water stress.
Under water deficit, the photosynthetic apparatus absorbs more light quanta than can be converted into chemical energy (Takahashi and Murata, 2008). This results in damage to D1 (PsbA) protein in PSII, and such a photosystem is excluded from the photosynthetic electron transport chain until D1 is resynthesized (Gururani et al., 2015). Moreover, an important role for Rieske protein, a component of the cyt b6f complex, in electron transport between PSII and PSI under water stress was also indicated (Hura et al., 2018, 2019). However, most studies focus on the role of PsbS protein which, by binding carotenoids in antennas, takes part in non-photochemical quenching of the excess energy (Croce, 2015; Fan et al., 2015; Ruban, 2016; Daskalakis and Papadatos, 2017), or on the role of chlorophyll, whose changes in content limit the extent of the absorbed light (Verhoeven et al., 2001; Gururani et al., 2015; Townsend et al., 2018).
A negative effect of long-term photoinhibition is overproduction of reactive oxygen species (ROS) that oxidize lipids and photosynthetic pigments in chloroplast membranes (Hura et al., 2015, 2018, 2019). It has been shown that in sites of plastoquinone QA/QB activity, electron leak may occur, accompanied by a reduction of oxygen (O2) to superoxide radical, which is further dismutated to hydrogen peroxide (H2O2). Another possible site of ROS formation can be PSI, where overburdening of the electron transport chain causes a redirection of some electrons from ferredoxin to O2 and reduction of oxygen via the Mehler reaction to superoxide radical that is then spontaneously dismutated to H2O2 (Hura et al., 2015, 2018).
Considering the destructive effects of light on the photosynthetic apparatus under soil drought, a lack or only limited rolling of the flag leaf is a surprising and uncommon response to this stress factor. We noticed such responses of the flag leaves in some cultivars of winter triticale exposed to soil drought, and they were associated, for example, with the content of phenolics in the cell wall (Hura et al., 2012).
Incorporation of phenolic compounds into the cell wall structures requires the presence of peroxidases (Wakabayashi et al., 2012). The incorporated phenolics form esters and/or ether cross-bridges with carbohydrate components of the cell wall (Meyer et al., 2003). The phenolics most often incorporated into the cell wall of drought-exposed grasses and cereals are ferulic and p-coumaric acids (Meyer et al., 2003; Hura et al., 2016).
The phenolic bioactivity is associated with the presence of an aromatic six-membered benzene ring that can interact with both high-energy radiation, such as UV radiation and the short-wave range of visible radiation, and ROS or organic radicals. When the cell wall is saturated with phenolic compounds, it acts as a filter shielding the cell interior from high-energy radiation. Light causes photoinhibitory damage to the photosynthetic apparatus during drought (Nogués and Baker, 2000). Their ability to absorb UV radiation allows phenolic compounds to transform short-wave, high-energy radiation, which is destructive to cellular structures, into non-destructive blue fluorescence (Bilger et al., 2001). In this way, cell wall-bound phenolics can act as photoprotectors of the photosynthetic apparatus (Hura et al., 2018). Moreover, elevated levels of benzene rings in the cell wall strengthen its hydrophobic properties. This curbs water transport from the metabolically active symplast to the apoplast, and then its loss via the apoplast (Graça and Santos, 2007; Hura et al., 2016). Phenolics in the cell wall make it less elastic, tighter and stiffer (Hura et al., 2013), and this also makes the entire leaf stiffer. As a consequence, the leaves show no or only limited rolling during drought (Hura et al., 2012). However, it should be underlined that a lack of leaf rolling under drought may also possibly be due to leaf structural mutations, for example in the bulliform cells (Zhang et al., 2015). On the one hand, such a response can indicate sensitivity to the stress factor but, on the other hand, it may be due to additional and effective stress adaptation mechanisms that compensate the lack of leaf rolling. This raises questions concerning structural and functional adjustment of the photosynthetic apparatus (changes in pigment content, stabilization of protein/pigment–protein complexes or enhancement of non-photochemical quenching of excess energy) in the process of drought adaptation in non-rolling leaves.
Therefore, we assumed that some innate physiological, biochemical and molecular characteristics of the lines studied resulted in their different leaf rolling types. We also hypothesized that non-rolling leaves developed effective adaptation mechanisms to long-term drought, involving, for example, optimal use of light and dissipation of excess energy by the photosynthetic apparatus. To verify this hypothesis, we analysed the proteins of the photosynthetic apparatus that form its core and are crucial for maintaining its stability (D1 protein), limit destructive effects of light (PsbS, a protein binding carotenoids in the antennas) and participate in efficient electron transport between PSII and PSI (Rieske iron–sulfur protein of the cyt b6f complex) under drought stress (Hura et al., 2018, 2019). We also compared molecular and physiological responses to drought stress in rolling and non-rolling flag leaves. Water retention capacity was analysed for both types of leaf rolling under soil drought and compared with the content of phenolic compounds. The tendency of flag leaves to roll was assessed by determining the content of phenolic compounds in the cell wall of doubled haploid (DH) lines of winter triticale. Additionally, colonization of the flag leaves by aphids, as a possible bioindicator of phenolic accumulation in the cell wall structures, was estimated.
MATERIALS AND METHODS
Plant material
The study investigated 92 DH lines of triticale (× Triticosecale Wittmack) originated from the F1 hybrid ‘Hewo’ × ‘Magnat’, and two DH parental lines (Hura et al., 2017; Ostrowska et al., 2019): ‘Hewo’, the female parent (Plant Breeding Strzelce Ltd, Co., Poland) and ‘Magnat’, the pollen donor (DANKO Plant Breeders Ltd, Choryń, Poland).
Plant growth conditions
The seeds were placed into pots of 3.7 dm3 (nine plants per pot) filled with the same amount of soil and sand (1:3; v/v) mixture. The emerging seedlings were vernalized in cool chambers for 8 weeks at +4 °C, photosynthetic photon flux density (PPFD) 150 μmol m–2 s–1 and 10 h/14 h photoperiod. After that, the plants at the two-leaf stage were transferred into a greenhouse chamber. Air temperature in the chamber was 25–30/15–20 °C day/night, while relative air humidity oscillated around 30 %. The plants were also illuminated by high-pressure sodium lamps (400 W; Philips SON-T AGRO, Brussels, Belgium), and PPFD at the level of the flag leaves reached about 200–250 μmol m–2 s–1 (QSPAR Quantum Sensor, Hansatech Instruments Ltd, Kings Lynn, UK). The plants were supplied with full-strength Hoagland’s nutrient medium (Hoagland, 1948) once a week, and this schedule was kept until soil drought was started.
Conditions of soil drought and rehydration
Soil drought was applied individually to each of the investigated DH lines immediately after the flag leaves started to emerge. In this way drought stress was induced in very young and flexible flag leaves with cell walls built mainly of carbohydrates. Water content in the pots was gradually limited to about 30 % by withholding watering for 7 d. It was then retained at that level for the next 2 weeks. Water content in the control pots was maintained at about 75 %. During rehydration, after finishing the soil drought period, water content in the pots was immediately restored to 75 %. Soil moisture in the pots was controlled every day with a gravimetric method (plant weight in the pots was taken into account), between 08.00 h and 10.00 h, and soil water content in the pots was immediately restored to 30 % (drought) or to 75 % (control/rehydration).
Measurements
Analyses and measurements were made, among others, after 7, 14 and 21 d of limited watering and after 24 h of rehydration. Expansion of green aphids on the flag leaves of DH lines was evaluated on the fifth day of rehydration. The flag leaves collected for quantitative analysis were lyophilized (FreezeDry System/Freezone® 4.5, LABCONCO Kansas City, MO, USA).
Observation of flag leaf rolling
On the 21st day of soil drought, we examined rolling of the flag leaves in two parental lines and the remaining 92 DH lines. Various rolling patterns were photographed and described.
Pre-dawn leaf water potential (ΨW)
The leaf water potential was measured with a HR 33T psychrometer (WESCOR, Inc., Logan, UT, USA) coupled with C-52 leaf sample chambers (WESCOR, Inc.). The leaf discs (excised from a central zone of the leaf) with a diameter of 5 mm were left in C-52 chambers for 60 min. The measurements were performed in the dew point mode.
Leaf osmotic potential (ΨO)
The leaf osmotic potential was assessed HR 33T with a psychrometer (WESCOR, Inc.) equipped with C-52 leaf sample chambers (WESCOR, Inc.). The parameter was determined in the sap collected from the leaves with a syringe. Discs of filter paper soaked in the sap were left in the chambers for 35 min. The measurements were performed in the dew point mode.
Leaf water loss
Leaf water loss analyses were performed during the rehydration following 7, 14 and 21 d of drought. Leaf samples to analyse water loss rate were collected after 24 h of rehydration. The samples were weighed immediately (fresh weight of starting point for 0 h; A) (0 h), after 1, 2 and 3 h of incubation at 30 ° C (B) and following an oven incubation at 80 ° C for 48 h (dry weight; C). Water loss in the harvested leaves was determined using the following formula:
Soluble phenolics and cell wall-bound phenolics
About 5.0 mg of powdered samples were used to analyse both soluble and cell wall-bound phenolics. Soluble phenolics (SPhs) were extracted with 100 µL of 80 % (v/v) ethanol. Their total content was determined with the Folin–Ciocalteu assay as described before (Hura et al., 2017).
Cell wall-bound phenolics were extracted from insoluble material by basic hydrolysis in 1 n NaOH and kept for 24 h at 25 C (Hura et al., 2012, 2013). The residues were centrifuged and the resulting supernatant was retained. Then the supernatant (100 µL), H2O (900 µL), 25 % sodium carbonate solution (500 µL) and Folin–Ciocalteu reagent (125 µL) (twice diluted with water) were combined into a reaction mixture. Its absorbance was measured spectrophotometrically (Ultrospec 2100 Pro, Amersham Biosciences, Cambridge, UK) at 760 nm. The content of cell wall-bound phenolics was standardized against ferulic acid.
The content of cell wall-bound phenolics was converted into the number of molecules in 1 g of cell wall dry weight using the Avogadro number and the fact that 1 mol = 6.02 × 1023 phenolic molecules. For cell wall-bound phenolics, the assumed molar mass was 182.67 g mol–1. This is an average molar mass of four phenolic compounds most common in the cell walls of triticale, i.e. ferulic acid (194.18 g mol–1), p-coumaric acid (164.15 g mol–1), cinnamic acid (148.15 g mol–1) and sinapic acid (224.21 g mol–1) (Hura et al., 2016).
Activity of l-phenylalanine ammonia-lyase (PAL) and l-tyrosine ammonia-lyase (TAL)
PAL and TAL activity was measured with modified methods developed by Peltonen and Karjalainen (1995). The samples (about 100 mg f. wt) were homogenized in the extraction buffer (50 mm Tris–HCl pH 8.5, 14.4 mm 2-mercaptoethanol, 5 % w/v polyvinylpyrrolidone) at 4 °C. The reaction mixture included 2.5 mL of 0.2 % solution of l-phenylalanine (for PAL analysis) or l-tyrosine (for TAL analysis) in 50 mm Tris–HCl (pH 8.5) and 0.5 mL of the supernatant. Its incubation at 38 °C lasted for 24 h. Then, the absorbance at 290 nm (for PAL analysis) or 310 nm (for TAL analysis) was measured. The enzyme activity was indicated as nmol of cinnamic acid (for PAL analysis) or p-coumaric acid (for TAL analysis) synthesized for 1 h per 1 mg of protein [nmol h–1mg(prot.)–1]. Protein content was determined as described by Bradford (1976).
Soluble carbohydrate content
The colorimetric estimation of soluble carbohydrate (SC) accumulation was carried out using anthrone reagent (Ashwell, 1957) according to Gadzinowska et al. (2019).
Hydrogen peroxide content
Analysis of H2O2 content was based on a spectrofluorimetric assay with homovanillic acid, as previously described by Hura et al. (2015).
Leaf photosynthetic activity, carboxylation efficiency and water use efficiency
Photosynthetic rate (A), transpiration rate (E), stomatal conductance (gs) and intercellular concentration of CO2 (Ci) were assessed with an LCpro-SD infrared gas analyser (ADC BioScientific Ltd, UK). The measurement conditions in the chamber were as follows: 360 µmol CO2 mol–1 air, ambient air humidity (approx. 40 %), 28 °C and photosynthetically active radiation (PAR) intensity 600 µmol photons m–2 s–1. The measurements were taken between 10.00 h and 12.30 h on the central parts of the flag leaves.
We also determined the apparent carboxylation efficiency (A/Ci, i.e. the ratio of photosynthetic rate to intercellular concentration of CO2), instantaneous water use efficiency (WUE) as A/E (ratio of photosynthetic rate to transpiration intensity) and intrinsic WUE as A/gs (ratio of photosynthetic rate to stomatal conductance).
Chlorophyll fluorescence determination
Chlorophyll fluorescence transient (OJIP) was investigated with a Handy PEA fluorometer (Hansatech Ltd, Kings Lynn, UK). The quantum yield of electron transport from QA– to the end electron acceptors in PSI (φRo) was evaluated according to earlier reports (Tsimilli-Michael and Strasser, 2008; Strasser et al., 2010; Gadzinowska et al., 2019).
Modulated chlorophyll fluorescence was measured with an FMS 2 fluorometer (Hansatech Instruments). Appropriate parameters (Ф PSII, PSII quantum efficiency; ETR, the electron transport rate; and qNP, non-photochemical quenching) were calculated as previously described by van Kooten and Snel (1990) and Genty et al. (1989).
Emission spectra of blue, red and far-red fluorescence
Fluorescence emission spectra of blue (450 nm), red (690 nm) and far-red (740 nm) fluorescence were determined with a spectrofluorometer (Perkin-Elmer LS 50B, Norwalk, CT, USA). The intensity of red and far-red fluorescence was read between 650 and 800 nm after leaf excitation at 450 nm. The spectral slit width was set to 10 nm (excitation and emission).
Blue fluorescence intensity was read between 400 and 500 nm after leaf excitation at 337 nm. The excitation and emission monochromator slit width was adjusted to 10 nm.
Assimilation pigment content
Chlorophyll and carotenoid contents were analysed spectrophotometrically (Ultrospec II, Biochrom, Cambridge, UK) following extraction in 95 % ethanol. The concentrations of chlorophyll and carotenoids were then computed as per Lichtenthaler and Wellburn (1983).
Protein accumulation
The investigated proteins involved those of the photosynthetic apparatus (PsbS, Agrisera AS09 533; PsbA, Agrisera AS05 084; and PetC, Agrisera AS08 330), a protein responsible for photosynthetic fixation of CO2 (RbcL, Rubisco large subunit, form I and form II; Agrisera AS03 037), and a protein responsible for carbohydrate metabolism (SPS, sucrose phosphate synthase, global; Agrisera AS03 035A).
We then proceeded with protein extraction, electrophoresis, western blot and gel analysis following the procedures provided in details in our previous studies (Hura et al., 2018; Gadzinowska et al., 2019). Band density was analysed using Image Studio Litever 5.2 software (LI-COR) and normalized to the densest band that was treated as 100 %.
Colonization of flag leaves by aphids
Cereal aphids were transferred from the pots with DH lines not subjected to soil drought. They fed mainly on young parts of the donor plants (flag and sub-flag leaves) with high turgor pressure. The experiment was carried out in a greenhouse and involved DH lines previously exposed to soil drought (colonized plants). The observations were made on the fifth day of rehydration after soil drought, and after 4 d of contact between the donor and colonized plants. We only analysed colonization of the flag leaves as they showed the highest turgor pressure and the best performance following soil drought. The analysis involved four randomly selected pots with the lines representing four different types of flag leaf rolling (NRL, T1LR, T2LR and T3LR; see the Results) under soil drought. Therefore, each type of leaf rolling was represented by four pots harbouring six plants each, i.e. a total of 24 plants that were not colonized by aphids. The pots with donor plants also harboured six plants with leaves abundantly colonized by aphids. Every pot with the donor plants was surrounded on four sides with four pots with colonized plants, each showing a different type of leaf rolling during soil drought. The pots stood very close to each other ensuring a direct contact between the donor and colonized plants. There were four such groups made up of one pot with the donor plants and four pots with the colonized plants.
Statistical analysis
Statistical analysis was performed with Statistica v. 13.0 package (Statsoft Inc., Tulsa, OK, USA). Major effects of drought on physiological and biochemical parameters were analysed by analysis of variance (ANOVA). Prior to ANOVA, we checked the data for normality and homogeneity of variance. The Duncan’s multiple range test at a 0.05 probability level was chosen to determine the significance of differences among treatment means. Differences among two means were compared with Student’s t-test. The Pearson correlation coefficient between measured parameters was tested at a probability of P = 0.05.
RESULTS
Types of leaf rolling
On the 21st day of soil drought, we identified four different types of rolling in dehydrated flag leaves of triticale DH lines (Fig. 1). Seven DH lines and the parental line ‘Magnat’ showed no signs of the flag leaf rolling (NRL, non-rolling leaf). Possible consequences of this strategy may involve excessive exposure to light and rapid dehydration during soil drought. In 34 DH lines. the flag leaf was partially folded along the midrib (T1LR, type 1 of leaf rolling), which may result in limited light access and rapid dehydration of the flag leaves. Flag leaves of the other 15 DH lines were slightly folded along the midrib and their edges rolled inwards (T2LR, type 2 of leaf rolling). In this case, the consequences may involve limited access to light and possible slowing down of leaf dehydration. In the largest group of DH lines (n = 36) and the parent line ‘Hewo’, the flag leaves rolled so that their cross-section was nearly circular (T3LR, type 3 of leaf rolling). The consequences may involve significantly limited access of light and slowing down of leaf dehydration.
Types of flag leaf rolling identified in DH triticale lines on the 21st day of water stress.
It should be underlined that the flag leaves of DH lines varied in their biomass and water status within each type of leaf rolling. Also, non-rolled leaves of individual DH lines exhibited differences in these parameters. Moreover, the non-rolled leaves showed similar dry mass and/or similar water status to the rolled leaves. For further information on the leaf water potential (Ψ W) and flag leaf dry mass, see Supplementary data Tables S1 and S2, respectively.
Physiology of flag leaves in parental DH lines
The NRLs in the DH ‘Magnat’ line and the flag leaves in the DH ‘Hewo’ line showing type 3 of leaf rolling (T3LR) under soil drought also differed in their physiological activity (Fig. 2). When soil water content was optimal (C), the flag leaves of both parental lines were characterized by similar water potential (Ψ W), levels of H2O2 and activity of PSI and PSII determined by measuring the intensity of red (FI690) and far-red fluorescence (FI740). Soil drought induced a similar level of water stress (WS) in the flag leaves of both parental lines as evidenced by similar values of Ψ W. However, in NRL plants, we detected significantly lower levels of H2O2 and significantly lower intensity of red fluorescence emitted from PSII (the same as under optimal hydration of flag leaves) than in T3LR plants. A significant rise in the intensity of far-red fluorescence emitted from PSI occurred only in dehydrated NRL flag leaves.
Water potential (Ψ W), hydrogen peroxide level (H2O2) and intensity of red (FI690) and far-red (FI740) fluorescence emission on the 21st day of water stress in the flag leaves of the parental DH lines ‘Magnat’ and ‘Hewo’. Treatments: C, control; WS, water stress. Types of flag leaf rolling: NRL, non-rolling leaves; T3LR, type 3 of leaf rolling. Means ± s.e. (n = 7 for Ψ W, n = 10 for H2O2, FI690 and FI740). Means indicated with the same letters are not significantly different within measured parameters (Duncan’s multiple range test at P = 0.05).
At optimal soil water content (C), the flag leaves of both parental lines showed a similar intensity of photosynthesis (A) and a similar level of intercellular concentration of CO2 (Ci) (Table 1). However, under the same conditions (C), the ‘Hewo’ DH line exhibited a significantly higher transpiration intensity (E) accompanied by a significant increase in stomatal conductance (gs).
Changes in photosynthesis intensity (A), transpiration intensity (E), stomatal conductance (gs), intercellular concentration of CO2 (Ci), instantaneous water use efficiency (A/E), intrinsic water use efficiency (A/gs), and apparent carboxylation efficiency (A/Ci) on the 21st day of water stress in the flag leaves of ‘Magnat’ and ‘Hewo’ DH parental lines
| Parameters | NRL – ‘Magnat’ | T3LR – ‘Hewo’ | ||
|---|---|---|---|---|
| C | WS | C | WS | |
| A [µmol(CO2) m–2 s–1] | 32.2 ± 1.8a | 14.3 ± 2.06b | 35.4 ± 1.7a | 7.7 ± 0.4c |
| E [mmol(H2O) m– s–1] | 5.32 ± 0.23b | 2.43 ± 0.31c | 6.53 ± 0.20a | 1.32 ± 0.08d |
| gs [mol(H2O) m–2 s–1] | 0.47 ± 0.058b | 0.12 ± 0.023c | 0.72 ± 0.104a | 0.07 ± 0.008c |
| Ci [µmol(CO2) mol(air)-1] | 153.0 ± 9.7a | 107.3 ± 6.3b | 159.1 ± 16.4a | 92.4 ± 3.9b |
| A/E [µmol(CO2) mmol(H2O)-1] | 6.1 ± 0.3a | 5.8 ± 0.1a | 5.5 ± 0.4a | 5.9 ± 0.3a |
| A/gs [µmol(CO2) mol(H2O)-1] | 76.3 ± 9.1a | 124.6 ± 7.2b | 56.0 ± 7.6a | 119.5 ± 12.3b |
| A/Ci [mol(air) m–2 s–1] | 0.221 ± 0.023a | 0.133 ± 0.015b | 0.250 ± 0.035a | 0.084 ± 0.004c |
| Parameters | NRL – ‘Magnat’ | T3LR – ‘Hewo’ | ||
|---|---|---|---|---|
| C | WS | C | WS | |
| A [µmol(CO2) m–2 s–1] | 32.2 ± 1.8a | 14.3 ± 2.06b | 35.4 ± 1.7a | 7.7 ± 0.4c |
| E [mmol(H2O) m– s–1] | 5.32 ± 0.23b | 2.43 ± 0.31c | 6.53 ± 0.20a | 1.32 ± 0.08d |
| gs [mol(H2O) m–2 s–1] | 0.47 ± 0.058b | 0.12 ± 0.023c | 0.72 ± 0.104a | 0.07 ± 0.008c |
| Ci [µmol(CO2) mol(air)-1] | 153.0 ± 9.7a | 107.3 ± 6.3b | 159.1 ± 16.4a | 92.4 ± 3.9b |
| A/E [µmol(CO2) mmol(H2O)-1] | 6.1 ± 0.3a | 5.8 ± 0.1a | 5.5 ± 0.4a | 5.9 ± 0.3a |
| A/gs [µmol(CO2) mol(H2O)-1] | 76.3 ± 9.1a | 124.6 ± 7.2b | 56.0 ± 7.6a | 119.5 ± 12.3b |
| A/Ci [mol(air) m–2 s–1] | 0.221 ± 0.023a | 0.133 ± 0.015b | 0.250 ± 0.035a | 0.084 ± 0.004c |
Treatments: C, control; WS, water stress. Types of flag leaf rolling: NRL, non-rolling leaves; T3LR, type 3 of leaf rolling. Means ± s.e. (n = 9). Means marked with the same letters do not show significant differences within the measured parameters (Duncan’s multiple range test at P = 0.05)
Changes in photosynthesis intensity (A), transpiration intensity (E), stomatal conductance (gs), intercellular concentration of CO2 (Ci), instantaneous water use efficiency (A/E), intrinsic water use efficiency (A/gs), and apparent carboxylation efficiency (A/Ci) on the 21st day of water stress in the flag leaves of ‘Magnat’ and ‘Hewo’ DH parental lines
| Parameters | NRL – ‘Magnat’ | T3LR – ‘Hewo’ | ||
|---|---|---|---|---|
| C | WS | C | WS | |
| A [µmol(CO2) m–2 s–1] | 32.2 ± 1.8a | 14.3 ± 2.06b | 35.4 ± 1.7a | 7.7 ± 0.4c |
| E [mmol(H2O) m– s–1] | 5.32 ± 0.23b | 2.43 ± 0.31c | 6.53 ± 0.20a | 1.32 ± 0.08d |
| gs [mol(H2O) m–2 s–1] | 0.47 ± 0.058b | 0.12 ± 0.023c | 0.72 ± 0.104a | 0.07 ± 0.008c |
| Ci [µmol(CO2) mol(air)-1] | 153.0 ± 9.7a | 107.3 ± 6.3b | 159.1 ± 16.4a | 92.4 ± 3.9b |
| A/E [µmol(CO2) mmol(H2O)-1] | 6.1 ± 0.3a | 5.8 ± 0.1a | 5.5 ± 0.4a | 5.9 ± 0.3a |
| A/gs [µmol(CO2) mol(H2O)-1] | 76.3 ± 9.1a | 124.6 ± 7.2b | 56.0 ± 7.6a | 119.5 ± 12.3b |
| A/Ci [mol(air) m–2 s–1] | 0.221 ± 0.023a | 0.133 ± 0.015b | 0.250 ± 0.035a | 0.084 ± 0.004c |
| Parameters | NRL – ‘Magnat’ | T3LR – ‘Hewo’ | ||
|---|---|---|---|---|
| C | WS | C | WS | |
| A [µmol(CO2) m–2 s–1] | 32.2 ± 1.8a | 14.3 ± 2.06b | 35.4 ± 1.7a | 7.7 ± 0.4c |
| E [mmol(H2O) m– s–1] | 5.32 ± 0.23b | 2.43 ± 0.31c | 6.53 ± 0.20a | 1.32 ± 0.08d |
| gs [mol(H2O) m–2 s–1] | 0.47 ± 0.058b | 0.12 ± 0.023c | 0.72 ± 0.104a | 0.07 ± 0.008c |
| Ci [µmol(CO2) mol(air)-1] | 153.0 ± 9.7a | 107.3 ± 6.3b | 159.1 ± 16.4a | 92.4 ± 3.9b |
| A/E [µmol(CO2) mmol(H2O)-1] | 6.1 ± 0.3a | 5.8 ± 0.1a | 5.5 ± 0.4a | 5.9 ± 0.3a |
| A/gs [µmol(CO2) mol(H2O)-1] | 76.3 ± 9.1a | 124.6 ± 7.2b | 56.0 ± 7.6a | 119.5 ± 12.3b |
| A/Ci [mol(air) m–2 s–1] | 0.221 ± 0.023a | 0.133 ± 0.015b | 0.250 ± 0.035a | 0.084 ± 0.004c |
Treatments: C, control; WS, water stress. Types of flag leaf rolling: NRL, non-rolling leaves; T3LR, type 3 of leaf rolling. Means ± s.e. (n = 9). Means marked with the same letters do not show significant differences within the measured parameters (Duncan’s multiple range test at P = 0.05)
The WS induced by soil drought resulted in a significant drop in the investigated photosynthesis parameters in both DH parental lines. The intensity of photosynthesis (A) and transpiration (E) in T3LR flag leaves were significantly lower than in NRL leaves, while the values of gs and Ci did not differ significantly (Table 1).
Both DH lines exposed to water stress demonstrated a significant and similar increase in intrinsic WUE (A/gs). Such differences were not observed for instantaneous WUE (A/E) (Table 1). The values of photosynthesis intensity and intercellular concentration of CO2 were used to calculate apparent carboxylation efficiency (A/Ci). Under optimal soil water content, the flag leaves of both parental lines reached similar A/Ci ratio values. In dehydrated plants, the ratio decreased significantly, and the decrease was considerably greater in T3LR than in NRL plants (Table 1).
Cell wall-bound phenolics in parental lines
We analysed the content of cell wall-bound phenolics in the flag leaves of T3LR and NRL type and then used the Avogadro constant to convert it into the number of phenolic molecules in 1 g of cell wall dry weight (Fig. 3A). A significant increase in the number of phenolic molecules occurred only in the cell wall of NRL flag leaves of the ‘Magnat’ DH line. The difference between drought and control variants showed how many molecules of phenolic compounds were integrated into the cell wall framework of dehydrated leaves (0.75 × 1019 in ‘Hewo’; 17.04 × 1019 in ‘Magnat’). The calculation demonstrated that the cell wall structures of NRL flag leaves of the ‘Magnat’ DH line incorporated 16.29 × 1019 more molecules of phenolic compounds than those of T3LR leaves in the ‘Hewo’ DH line (Fig. 3B).
(A) The number of phenolic compounds incorporated into the cell wall structures on the 21st day of water stress for the flag leaves of parental DH lines ‘Magnat’ and ‘Hewo’. Types of flag leaf rolling: NRL, non-rolling leaves; T3LR, type 3 of leaf rolling. Means ± s.e. (n = 9). Means indicated with the same letters are not significantly different (Duncan’s multiple range test at P = 0.05). (B) Calculations of differences in the amount of incorporated phenolic molecules for control and stress treatments and for ‘Magnat’ and ‘Hewo’ DH lines subjected to water stress.
Water loss in parental lines
Figure 4A shows water loss in NRL flag leaves of the ‘Magnat’ line following 24 h rehydration after 7, 14 and 21 d of drought, respectively. In T3LR flag leaves of the ‘Hewo’ line, where the content of cell wall-bound phenolics remainded unaffected, water loss was analysed only following 24 h rehydration after 21 d of drought. After 7 d of soil drought, the average number of phenolic molecules in the NRL flag leaf cell wall reached 3.96 × 1019 (± 0.25 × 1019), after 14 d 11.64 × 1019 (± 0.93 × 1019) and after 21 d 20.56 × 1019 (± 1.40 × 1019). The flag leaves with the highest amount of incorporated phenolic molecules, i.e. those experiencing 21 and 14 d of drought, showed the slowest rate of water loss 24 h after their rehydration (Fig. 4A). Figure 4B indicates the correlation of water loss after 3 h of desiccation with the amount of phenolic compounds built into the cell wall structures of NRL flag leaves of the ‘Magnat’ line. The greater the amount of phenolics in the cell wall structures, the smaller the water loss was in 3TLR leaves.
(A) Water loss (%) in the flag leaves of the ‘Magnat’ DH line 24 h after rehydration following 7, 14 and 21 d of water stress. The water loss rate for the ‘Hewo’ DH line showing a low content of cell wall-bound phenolics was analysed only for the variant representing leaves 24 h after rehydration following 21 d of water stress. Treatments: C, control; WS, water stress. Mean values ± s.e. (n = 10). (B) Correlation between leaf water loss rate (% h–1) and the number of phenolic molecules incorporated into the cell wall structures of flag leaves of the ‘Magnat’ DH line 24 h after rehydration following 7, 14 and 21 d of water stress. The line denotes linear adjustment at a probability level of P = 0.05.
Photosynthetic apparatus activity in rolling (T3LR) and non-rolling (NRL) flag leaves
Likely consequences of non-rolling of the flag leaves in the ‘Magnat’ DH line in response to dehydration may involve not only a rapid water loss but also damage to the photosynthetic apparatus caused by excessive exposure to light (Fig. 5A). The analysis of key proteins of the photosynthetic apparatus performed on the 21st day of soil drought revealed a significant increase in the accumulation of PetC (Rieske protein) (Fig. 5B) and PsbA (D1) protein (Fig. 5D). The content of PsbS protein in dehydrated NRL flag leaves remained unaffected (Fig. 5C). These changes were accompanied by a significant decrease in both quantum efficiency of PSII (Ф PSII) and the electron transport rate (ETR), and an increase in non-photochemical quenching (qNP).
(A) Graphical representation of disadvantages of non-rolling leaves. Western blot analysis showing accumulation of Rieske iron–sulfur protein (PetC) of the cyt b6f complex (B), the carotenoid-binding protein in PSII (PsbS) (C) and D1 protein (PsbA) forming the core of PSII (D) on the 21st day of water stress in the flag leaves of the ‘Magnat’ parental line. The analysis was repeated three times (1–3). CBB, the share of equal amounts of total protein stained with Coomassie Brilliant Blue as a loading control in an SDS–PAGE stacking gel. The band density within each protein was normalized to the densest band that was assumed as 100 %. Asterisks represent differences significant at P = 0.05, Student’s t-test. Changes in the accumulation of individual proteins are presented together with changes in respective values of chlorophyll fluorescence parameters: ETR, electron transport rate; Ф PSII, PSII quantum efficiency; and qNP, non-photochemical quenching coefficient. Mean values ± s.e. (n = 10). Asterisks represent differences significant at P = 0.05, Student’s t-test.
The T3LR type of flag leaf rolling observed in the ‘Hewo’ line may limit both water transpiration and photoinhibitory damage to the photosynthetic apparatus (Fig. 6A). Similarly to the case in NRL leaves, dehydration of T3LR flag leaves was associated with a significant drop in the activity of the photosynthetic apparatus determined based on fluorescence parameters (ETR, Ф PSII and qNP). As regards proteins, dehydrated T3LR flag leaves showed no changes in PetC levels (Fig. 6A), while the accumulation of PsbS (Fig. 6C) and band I of PsbA (D1) dropped significantly (Fig. 6A), and the level of band II of this protein remained unaffected (Fig. 6A).
(A) Graphical representation of benefits of leaf rolling. Western blot analysis showing accumulation of Rieske iron–sulfur protein (PetC) of the cyt b6f complex (B), the carotenoid-binding protein in PSII (PsbS) (C) and D1 protein (PsbA) forming the core of PSII (D) on the 21st day of water stress in the flag leaves of the ‘Hewo’ parental line. The analysis was repeated three times (1–3). CBB, the share of equal amounts of total protein stained with Coomassie Brilliant Blue as a loading control in an SDS–PAGE stacking gel. The band density within each protein was normalized to the densest band that was assumed as 100 %. Asterisks represent differences significant at P = 0.05, Student’s t-test. Changes in the accumulation of individual proteins are presented together with changes in respective values of chlorophyll fluorescence parameters: ETR, electron transport rate; Ф PSII, PSII quantum efficiency; and qNP, non-photochemical quenching coefficient. Mean values ± s.e. (n = 10). Asterisks represent differences significant at P = 0.05, Student’s t-test.
Four types of flag leaf rolling in 92 lines of the DH population under soil drought. Correlation between the physiological activity and the number of phenolic molecules incorporated into the cell wall structures
Figure 7 shows the correlation between individual physiological parameters and the number of phenolic compounds incorporated into the cell wall structures for the four different types of flag leaf rolling during dehydration. The physiological parameters included water relations (water potential, Ψ W, and leaf water loss) and factors they depend on (osmotic potential, Ψ O, and gs), stress indicators (SPhs, blue fluorescence emission intensity FI440–450; and SCs, H2O2 level), and broadly understood plant productivity analysed based on chlorophyll fluorescence parameters (red fluorescence emission intensity; FI690, far-red fluorescence emission intensity, FI740; qNP, quantum yield of electron transport from QA– to the PSI end electron acceptors. φRo), biomass growth (flag leaf dry weight) or the levels of assimilation of pigments (sum of chlorophyll a and b, Chl a + b, and carotenoids).
Correlation between water potential (Ψ W), osmotic potential (Ψ O), stomatal conductance (gs), leaf water loss, soluble phenolic (SPh) content, soluble carbohydrate (SC) content, H2O2 content, emission intensity of blue (FI440–450)/red (FI690)/far-red (FI740) fluorescence, flag leaf dry weight, non-photochemical quenching coefficient (qNP), quantum yield of electron transport from QA– to the PSI end electron acceptors (φRo), chlorophyll (Chl a + b) content, carotenoid (Crt) content and the number of phenolic molecules incorporated in the cell wall structures of the flag leaves representing four different types of rolling (NRL, T1LR, T2LR and T3LR) on the 21st day of water stress in 92 DH lines of triticale. For all 92 DH lines, measurements for each parameter were performed in triplicate (92 × 3 = 276 measurement points). The lines denote linear adjustment at a probability level P = 0.05.
For the parameters of water relations, significant correlations were detected for water potential (Ψ W), osmotic potential (Ψ O) and leaf water loss. The results calculated for gs were not statistically significant, showed no correlation with the level of phenolic compounds in the cell wall and did not differentiate the types of flag leaf rolling in the investigated lines of DH populations. The highest number of phenolic molecules in NRL flag leaves correlated with the highest level of leaf hydration (Ψ W), the lowest levels of osmotic potential (Ψ O) and the slowest rate of water loss (Fig. 7).
For water stress indicators, statistically significant correlations were obtained for the level of SC, the content of H2O2 and the intensity of blue fluorescence emission (FI440–450), the main source of which is ferulic acid covalently bound to cell wall carbohydrates. In this group of parameters, the highest numbers of phenolic compounds in the cell wall of NRL and T1LR flag leaves correlated with the highest levels of SC, the lowest level of H2O2 and the highest intensity of blue fluorescence emission. The results on the content of SPhs were statistically insignificant and did not differentiate the types of flag leaf rolling (Fig. 7).
Also in Fig. 7, the correlations between the number of phenolic compounds in the cell wall structures of NRL, T1LR, T2LR and T3LR flag leaves, and chlorophyll fluorescence parameters, dry weight of flag leaves and the content of assimilation pigments are shown. The highest numbers of phenolic compounds in the cell wall of NRL and T1LR flag leaves correlated with the lowest red fluorescence (FI690) emission intensities, the highest far-red fluorescence (FI740) emission values, the lowest values of qNP, high values of the quantum yield of electron transport from QA– to the PSI end electron acceptors (φRo) and high levels of carotenoids. The results on flag leaf dry weight and chlorophyll content were insignificant and did not differentiate the type of rolling of dehydrated flag leaves (Fig. 7).
In T3LR flag leaves, a low content of phenolics in the cell wall significantly correlated with the lowest values of Ψ W, SCs, FI440–450, FI740, φRo and carotenoids, the highest values of Ψ O, qNP and FI690, the fastest rate of leaf water loss or high content of H2O2. The results calculated for T2LR leaves usually fell in between those for T3LR and T1LR/NRL leaves (Fig. 7).
T2LR type of leaf rolling during dehydration in 15 DH lines
The results presented in Fig. 8 represent the mean content of SCs and phenolic compounds in 15 DH lines with the T2LR type of flag leaf rolling on the 21st day of soil drought. Two samples were collected from all 15 flag leaves in 15 DH lines and they included the rolling and non-rolling part of the leaf. A repetition involved one sample taken from the rolling and non-rolling part of the same leaf from the same DH line, which is why the means in the graphs represent 15 repetitions.
Content of soluble carbohydrates (SC), soluble phenolics (SPh), number of phenolic compounds incorporated into the cell wall structures, content of RbcL (Rubisco large subunit) and SPS (sucrose phosphate synthase), and activity of PAL and TAL on the 21st day of water stress in the samples collected from non-rolling (1) and rolling (2) parts of flag leaves (the rolled parts were separated from the unrolled parts within a single leaf with a scalpel). CBB, the share of equal amounts of total protein stained with Coomassie Brilliant Blue as a loading control in an SDS–PAGE stacking gel. The band density within each protein was normalized to the densest band that was assumed as 100 %. Mean values ± s.e (n = 15 for SCs, SPhs, number of phenolic molecules in the cell wall; n = 3 for RbcL and SPS, n = 6 for PAL and TAL). Asterisks represent differences significant at P = 0.05, Student’s t-test.
The rolling parts of the flag leaves (edges) had a significantly higher level of SCs and SPhs than the non-rolling parts (Fig. 8). The analysis of cell wall-bound phenolics confirmed a considerably higher number of phenolic compounds bound to the cell wall carbohydrate components in the non-rolling parts of the leaves (Fig. 8).
The analysis of PAL and TAL activity demonstrated a significantly higher PAL activity in non-rolling parts of the leaves than in the rolling parts (Fig. 8). These measurements involved six samples constituting fragments of rolling and non-rolling parts of flag leaves randomly selected from DH lines showing the T2LR type of leaf rolling.
Analysis of the accumulation of RbcL and SPS indicated significantly higher levels of these proteins in the rolling parts of the flag leaves. The analyses were performed in three repetitions. One repetition consisted of samples taken from the rolling/non-rolling parts of five flag leaves from five different DH lines of the T2LR type of flag leaf rolling.
Flag leaf colonization by aphids in relation to content of cell wall-bound phenolics
Predominant species of the aphid population included Sitobion avenae Fabricius (about 90 %) and Rhopalosiphum padi Linnaeus (about 10 %). A significant increase in aphid colonization was observed in T3LR flag leaves that incorporated the lowest number of phenolic compounds during soil drought (Fig. 9). The aphids fed in the lowest abundance on NRL and T1LR flag leaves that incorporated the greatest number of phenolic molecules into their cell wall structures. The flag leaves of T2LR showed stronger aphid colonization than those of the NRL and T1LR type.
Colonization of the flag leaves by aphids on the fifth day of rehydration after soil drought that was also the fourth day of contact between the donor plants (DP) and DH lines that experienced drought and exhibited four types of flag leaf rolling: NRL, T1LR, T2LR and T3LR. Means ± s.e. (n = 24). Means indicated with the same letters show no significant differences within measured parameters (Duncan’s multiple range test at P = 0.05).
Discussion
Flag leaf rolling in cereals is one of visible symptoms of plant dehydration under soil drought (Hura et al., 2012; Cal et al., 2019). The most commonly observed type of rolling is T3LR, when the flag leaves roll into a tube with a circular cross-section (Fig. 1). Triticale, which we have been studying for >20 years, developed four different types of leaf rolling. This is particularly evident in large-scale soil drought experiments that include multiple genotypes or DH lines, such as a mapping population. These four types of flag leaf rolling in a single species may be surprising, especially in comparison with other cereals such as wheat, rye, barley or oats.
These four different types of flag leaf rolling may be explained by specific properties of triticale. It is an engineered species that did not evolve naturally but rather was bred as an intergeneric hybrid of wheat and rye (Kuleung et al., 2004; Ma and Gustafson, 2008). As such, it is likely to develop unique properties, e.g. controlling flexibility of flag leaves during dehydration. The biology of triticale has not been extensively studied, as it is a fodder crop with much smaller economic importance than wheat, rye or barley. Its responses to adverse environmental conditions, such as soil drought, are controlled not only by wheat and rye genomes but also by interactions of these genomes (Hura et al., 2017). For this reason, triticale may be a source not only of new and specific responses to soil drought but also of effective mechanisms increasing plant tolerance to soil water deficit. This is particularly important considering current climate changes, global warming and increasing frequency of extreme weather phenomena such as long-term periods without precipitation (Pokhrel et al., 2021). More and more hybrid cultivars of cereal plants are bred, e.g. oats with maize chromosomes, and their biology is intensively researched (Skrzypek et al., 2018; Juzoń et al., 2020). Therefore, studies focused on a classic hybrid triticale, known from the 1960s, may be a useful benchmark for investigating the molecular basis of interactions between different genomes or their parts in order to better understand the specific biology of the hybrids.
Our previous studies revealed different intensities of incorporating phenolic compounds into the cell wall structures in individual DH triticale lines exposed to soil drought (Hura et al., 2012, 2016). The phenolic compounds bind to the carbohydrate components of the cell wall by ester or ether bonds that create transverse bridges between polycarbohydrate chains (Fig. 10). This results in a loss of cell wall flexibility, as its structure becomes more dense, rigid and tight (Fry, 1982, 1983, 1986, 1987; Kamisaka et al., 1990; Wakabayashi et al., 1997). These changes also make the wall interior more hydrophobic. In this way, a phenolic-saturated cell wall is transformed into a chemical barrier that reduces water flow from the cell interior to the apoplast and the movement of water inside the apoplast towards its external environment (Hura et al., 2016) (Fig. 10). At the same time, the above-mentioned cell wall-bound phenolic-induced wall stiffness, and consequently leaf stiffness, affects leaf rolling under water stress conditions. The outcomes presented here indicate that an increased number of phenolic compound molecules in the cell wall (Fig. 3A) was accompanied by a slower rate of water loss in the flag leaves (Fig. 4A), and different types of flag leaf rolling (Fig. 1). This can be explained by the different content of phenolics accumulated in the cell wall. Triticale cell wall saturation with phenolic compounds is controlled mainly by the rye genome. Our previous study identified rye chromosomes 3R and 6R as those harbouring the loci associated with the process. A comparison of rye quantitative trait loci (QTLs) with the wheat genome revealed the presence of serine/threonine kinase and cytokinin oxidase/dehydrogenase 3 genes possibly associated with the integration of phenolics into the cell wall structures (Hura et al., 2017). Also under natural conditions of cereal growth, dehydrated flag leaves are stiff in rye and remain flexible in wheat. Lack of flag leaf rolling in triticale (NRL and T1LR type) exposed to soil drought correlates with the elevated content of phenolic molecules in the cell wall that prevent the leaf rolling. In the T3LR type, the leaves have low levels of cell wall-bound phenolics and thus maintain their flexibility. We therefore concluded that the NRL and T1LR type of flag leaf rolling are controlled mainly be the rye genome (RR), the T3RL type by the wheat genome (AABB) and the T2LR type by the interaction of wheat and rye genomes, i.e. the triticale genome (AABBRR) (Fig. 11).
Graphical representation of the function of phenolic compounds built into the cell wall structures in limiting water loss during soil drought.
Graphical representation of the role of the wheat (AABB) and/or triticale (RR) genome in controlling the rolling of dehydrated flag leaves in triticale.
Water deficit in leaf tissues boosts the emission of blue fluorescence (Hura et al., 2015). The source of this fluorescence are phenolic compounds (mainly ferulic acid) covalently bound (ester and/or ether bonds) to carbohydrate components of the cell wall (Lichtenthaler and Schweiger, 1998; Meyer et al., 2003). In our study, such a boost in blue fluorescence (FI410) emission intensity occurred in rigid NRL and T1LR leaves, i.e. those with the greatest number of phenolic compounds built into the cell wall during dehydration (Figs 3 and 7). Lack of the flag leaf rolling (NRL) is a specific example of plant response to soil drought. The photosynthetic apparatus in such leaves is exposed to photoinhibitory damage. Our experiment demonstrated a significant increase in Rieske and D1 proteins as well as a constant high level of PsbS protein in dehydrated flag leaves as compared with control (Fig. 5). T3LR plants with fully rolled dehydrated flag leaves showed a contrasting trend consisting of a drop in protein accumulation (Fig. 6). Increased accumulation of proteins in NRL plants is associated with a mechanism of adaptation of the photosynthetic apparatus to neutralizing excess excitation energy. A combination of these processes with an enhanced number of phenolic molecules incorporated into the cell wall compensates for the lack of flag leaf rolling during dehydration. Elevated levels of Rieske FeS protein increase electron transport rates and biomass yield (Simkin et al., 2017). At the same time, dehydration-induced accumulation of D1 indicates predominance of PSII repair processes related to stimulation of this protein translation over its degradation (Nishiyama et al., 2004; Murata et al., 2007). The increase in PetC and D1 accumulation in non-rolling dehydrated flag leaves served as a mechanism preventing PSII overloading with the excitation energy. It allowed for its efficient transfer to PSI and further dissipation as far-red fluorescence (FI740) (Baker, 1991; Huner et al., 1998; Agati et al., 2000; Hura et al., 2015). It is worth mentioning that maintaining high levels of PsbS protein in PSII of NRL flag leaves (Fig. 5) is also crucial to the regulation of the photosynthetic light harvesting (Brooks et al., 2014; Naranjo et al., 2016; Sacharz et al., 2017). This efficient adaptation mechanism curbed overproduction of ROS such as, for example, H2O2 (Figs 2 and 7) and stimulated photosynthetic activity of NRL flag leaves (Table 1).
Our study showed that the plants of 92 triticale DH lines, with four different types of leaf rolling (NRL, T1LR, T2LR and T3LR), clearly differ in their biochemical and physiological activity correlated with the number of phenolic molecules incorporated in the cell wall structures (Fig. 7). Similar differences in plant physiological activity were reported in other studies, but only for the degree of leaf rolling during drought stress (Terzi and Kadioglu, 2006; Sezgin et al., 2018; Cal et al., 2019). The research literature lacks reports on the physiology of flag leaves responding to soil drought with different types of rolling in the same species.
In 15 DH lines, we detected an unusual type of flag leaf rolling (T2LR) (Fig.1), which was probably due to an interaction of wheat and rye genomes during dehydration (Fig. 11). The rolled part of the flag leaves (edges) differed from the rigid, non-rolled part (central) in the accumulation of phenolic compounds and SCs, the activity of PAL and the levels of RbcL and SPS proteins (Fig. 8). In our opinion, the rolled part is controlled by the wheat genome, as indicated by the high content of carbohydrates and greater content of proteins responsible for CO2 fixation (RbcL) and carbohydrate metabolism (SPS) (Hura et al., 2018). The rigid, non-rolled part is controlled by the rye genome, as evidenced by higher activity of PAL indispensable for the formation of phenolic compounds (Sanchez-Rodriguez et al., 2011), and lower levels of soluble phenolic compounds due to their incorporation in the cell wall structures (greater number of molecules) (Fig. 8). in this way, the rolling part of the flag leaf exposed to soil drought becomes the source, and the rigid, non-rolling part serves as a sink intercepting carbohydrates to transform them into phenolic compounds (Arnold et al., 2004; Zhang et al., 2010). Kim et al. (2015) reported a similar type of a biochemical specialization concerning diterpenoid synthesis in different tissues of Stevia (Stevia rebaudiana) leaves.
Aphid colonization may serve as a biological indicator of the accumulation of cell wall-bound phenolics in the flag leaves of triticale (Fig. 9). As previously confirmed, increased content of phenolics boosts plant resistance to insect herbivores (Buanafina and Fescemyer, 2012; Czerniewicz et al., 2017; Selvaraj et al., 2020).
In summary, this is the first study reporting on four different types of flag leaf rolling in a single species during soil drought, which are genome related. The biochemical basis for these differences was a different number of phenolic molecules incorporated into polycarbohydrate structures of the cell wall. An especially interesting case were the flag leaves showing lack of rolling during dehydration. This was accompanied by considerable accumulation of cell wall-bound phenolics and involved a high risk of rapid water loss from cells and enhanced photoinhibitory damage to the photosynthetic apparatus. However, the cell wall-bound phenolics make the wall more rigid and dense, and reinforce its hydrophobic properties, thus limiting water loss from the non-rolling leaves. The phenolics saturating the cell wall in this type of flag leaves serve as photoprotectors limiting photoinhibitory damage to the photosynthetic apparatus. PSII was also protected against excess light in non-rolling flag leaves by the accumulation of photosynthetic apparatus proteins that ensured stable and efficient transport of excitation energy beyond PSII and its dissipation as far-red fluorescence and heat.
We demonstrated that a specific type of flag leaf rolling (rolled edges and flat central part) may result from interactions between wheat and rye genomes. This is also related to the biochemical specialization of the flexible, rolling and the rigid, non-rolling part of the flag leaf that differ in their production and utilization of SCs indispensable for the formation of phenolic compounds, including those integrated into the cell wall framework. Finally, the study confirmed limited aphid colonization of the flag leaves with enhanced content of cell wall-bound phenolics.
Further comprehensive investigations are required to pinpoint the regions (loci) of the triticale genome that are responsible for different types of flag leaf rolling under drought stress. It is also necessary to carry out a structural and functional characterization of those regions by identifying and sequencing genes within individual loci, and detecting transcripts and the corresponding proteins.
SUPPLEMENTARY DATA
Supplementary data are available online at https://academic.oup.com/aob and consist of the following. Table S1: the leaf water potential on the 21st day of water stress in the flag leaves of 92 DH lines. Table S2: the flag leaf dry weight on the 21st day of water stress for 92 DH lines.
Funding
This work was supported by a grant from The National Science Centre, Poland, grant no. 2018/31/B/NZ9/00298.
