SUMMARY
WHIRLY1 is a chloroplast-nucleus located DNA/RNA-binding protein with functions in development and stress tolerance. By overexpression of HvWHIRLY1 in barley, lines with a 10-and two lines with a 50-fold accumulation of the protein were obtained. In these lines, the relative abundance of the nuclear form exceeded that of the chloroplast form indicating that over-accumulating WHIRLY1 exceeded the amount that chloroplasts can sequester. Growth of the plants was shown to be compromised in a WHIRLY1 abundance-dependent manner. Over-accumulation of WHIRLY1 in chloroplasts had neither an evident impact on nucleoid morphology nor on the composition of the photosynthetic apparatus. Nevertheless, oeW1 plants were found to be compromised in the efficiency of photosynthesis. The reduction in growth and photosynthesis was shown to be accompanied by a decrease in the levels of cytokinins and an increase in the level of jasmonic acid. Gene expression analyses revealed that already in non-stress conditions the oeW1 plants had enhanced levels of pathogen response (PR) gene expression indicating activation of constitutive defense. During growth in continuous light of high irradiance, PR1 expression further increased in addition to an increase in the expression of PR10 and of the gene encoding phenylalanine lyase (PAL), the key enzyme of salicylic acid biosynthesis in barley. The activation of defense gene expression in oeW1 plants coincided with an enhanced resistance towards powdery mildew, which in barley is independent of salicylic acid. Taken together, the results show that over-accumulation of WHIRLY1 in barley to levels of 10 or more, amplified the tradeoff between growth and stress resistance.
INTRODUCTION
WHIRLY proteins are multifunctional DNA/RNA binding proteins localized to the DNA-containing organelles and the nucleus of higher plants (Krupinska et al., 2022). Investigations with mutants and knockdown plants have shown that WHIRLIES affect developmental processes and stress tolerance (Krupinska et al., 2022).
Initially, the WHIRLY1 protein has been identified as a transcriptional activator of the pathogen response gene PR10a in potato (Desveaux et al., 2000). Its binding to the promoters of PR genes that are enriched in elicitor response elements (ERE) was shown to depend on salicylic acid (SA) (Desveaux et al., 2004). In recent years it has been shown that the role of SA is not limited to pathogen defense but that SA has an essential role in the regulation of redox homeostasis and thereby affects plants’ responses towards abiotic and biotic stress (Mateo et al., 2006, Karpinski et al., 2013). Accordingly, the abundance of WHIRLY1 as a critical protein in SA signaling was shown to impact the plants’ tolerance to diverse abiotic stress situations as well as pathogen defense. In whirly1 (why1) Tilling mutants of Arabidopsis, in which the binding of WHIRLY1 to the promoter of PR1 was reduced, resistance to Peronospora parasitica was relieved, too (Desveaux et al., 2004). Very recently, it has been reported that overexpression of WHIRLY1 from Vitis vinifera under the control of a strong pathogen response promoter enhances resistance towards Phytophthora capsica (Lai et al., 2022).
Besides its positive effect on defense, WHIRLY proteins were also found to promote tolerance towards abiotic stress. In tomato plants, overexpression of WHIRLY1 was shown to enhance thermotolerance by upregulating the expression of the HSP21.5A gene which has an ERE in its promoter and encodes an endoplasmic reticulum-localized heat shock protein (Zhuang et al., 2020a). Another study by the same research group showed that the plants overexpressing SlWHIRLY1 had an enhanced chilling tolerance (Zhuang et al., 2019). Vice versa, tomato plants with an RNAi-mediated knockdown of SlWHIRLY1 showed reduced resistance to chilling (Zhuang et al., 2019) and heat (Zhuang et al., 2020a). In maize and barley, it has been demonstrated that a reduction of WHIRLY1 negatively affects chloroplast development (Prikryl et al., 2008, Krupinska et al., 2019). Furthermore, barley plants deficient in WHIRLY1 were shown to be compromised in light acclimation (Saeid Nia et al., 2022).
Intriguingly, WHIRLY1 in barley was shown to locate in both, chloroplasts and nucleus of the same cell (Grabowski et al., 2008). In transplastomic tobacco plants synthesizing a tagged WHIRLY1 protein, this tagged protein was found in the nucleus indicating a translocation of WHIRLY1 from chloroplasts to the nucleus. In these plants, the expression of PR genes was enhanced (Isemer et al., 2012b). It has been hypothesized that storage of a transferable resistance protein such as WHIRLY1 in plastids might allow the plants to react immediately to pathogen attack avoiding the time and costs of gene expression (Krause and Krupinska, 2009). The translocation was suggested to occur in response to stress-associated redox changes in the photosynthetic apparatus (Foyer et al., 2014). WHIRLY1 is a positive regulator of both plant development and stress tolerance. Hence, WHIRLY1 promotes two traits that usually are inversely correlated. Indeed, enhanced stress tolerance coincides with lower growth and productivity (Herms and Mattson, 1992). This tradeoff is thought to be caused by resource restrictions demanding a prioritization of either growth or defense in response to environmental factors (Huot et al., 2014). The tradeoff is seemingly inevitable because the energy required for resistance is no longer available for biomass accumulation and production of seeds (Karasov et al., 2017). Thousands of genes are typically activated to fight a pathogen or cope with another stressful situation. Among others, the tradeoff between growth and defense is regulated by crosstalk between defense and growth hormones (Huot et al., 2014). Regulation of the level of free auxin is a significant determinant of adaptive growth in response to biotic and abiotic stress (Park et al., 2007). Recently, it has been demonstrated that MAP kinases activated during the immune response are involved in the downregulation of the expression of photosynthesis-associated genes, thereby exerting a negative impact on growth (Su et al., 2018).
Several studies with different dicot species have clearly shown a positive impact of over-accumulating WHIRLY1 on stress tolerance, however, without reporting effects on development and growth in these plants. Regarding the multifunctionality of WHIRLIES (Krupinska et al., 2022a) it is expected that other physiological parameters are altered besides stress tolerance. This study aimed to investigate the impact of a much higher abundance of multifunctional WHIRLY1 on plant growth and stress tolerance.
RESULTS
Overexpression of HvWHIRLY1 altered the abundance of HvWHIRLY1 in chloroplasts and the nucleus
By transforming barley with HvWHIRLY1 under the control of the constitutive UBIQUITIN 1 promoter of maize (Figure 1a), three homozygous lines were selected and used for characterization. Immunoblot analysis with the HvWHIRLY1 specific antibody (Grabowski et al., 2008) revealed that in primary foliage leaves of line oeW1-14, the level of HvWHIRLY1 was enhanced by a factor of about 50 (Figure 1b) as it was also in line oeW1-2 (Figure S1a). For comparison, in line oeW1-15 the level of WHIRLY1 was enhanced by a factor of 10 (Figure 1b).
WHIRLY1 is dually located in chloroplasts and nucleus. To investigate whether the over-accumulation of the protein occurred in both compartments and whether the relative distribution between chloroplasts and nucleus is altered by the overexpression of WHIRLY1 leaves (Fig. 2a) WHIRLY1 abundance was immunologically investigated in chloroplast and nuclei fractions prepared from primary foliage leaves of the oeW1-15 and the oeW1-14 lines. Theoretically, excess WHIRLY1 could accumulate only in chloroplasts or could also accumulate in nuclei, whereby the ratio between chloroplast and nuclear WHIRLY1 could be similar to in WT plants or could be shifted towards the nuclear form (Figure 2a). Immunoblots with the specific antibody for HvWHIRLY1 showed that the abundance of HvWHIRLY1 was enhanced in both chloroplasts and nuclei (Figure 2b, Figure S1b). Thereby the relative increase in quantity in both lines was higher in nuclei than in chloroplasts (Fig. 2c). Considering that the proteins in chloroplasts and nucleus have the same molecular weight, the higher abundance in the nucleus likely results from an enhanced flux of protein from chloroplasts to the nucleus. This result suggests that the capacity to sequester HvWHIRLY1 in chloroplasts is saturated, and relatively more WHIRLY1 is transferred to the nucleus (Figure 2).
Growth of barley oeW1 plants
To investigate whether WHIRLY1 overaccumulation has consequences for growth, the lengths of primary foliage leaves were measured every day until they were fully expanded (Figure 3). An apparent growth reduction correlated with an abundance of WHIRLY1. Growth curves showed that the seedlings of oeW1-10 were longer than those of the oeW1-50 line (Figure 3). The growth kinetics did not differ between WT and oeW1 seedlings. The maximal expansion of the primary foliage leaves was terminated at the same time after sowing, i.e. at 9 days (Fig. 2a).
Characterization of the photosynthetic apparatus
To investigate whether changes in the photosynthetic apparatus are responsible for the reduced growth of the oeW1 seedlings, the functionalities of the two photosystems were examined during the development of barley seedlings in a daily light/dark regime. Chlorophyll fluorescence measurements revealed that the maximal quantum yield of photosystem II, FV/FM, which is a measure of photosystem II efficiency, in WT seedlings was already relatively high at 7 das (0.7) and increased to nearly 0.8 at 10 das. In comparison, FV/FM of the oeW1-50 leaves stayed rather low, reaching a maximal value of about 0.5 at 10 das (Figure 4a).
For comparison, in oeW1-10 seedlings, FV/FM had a value of 0.6 at 7 das and a value of 0.7 at 9 and 10 das (Fig. 4a). In contrast to FV/FM, which barely changed in wild-type leaves during the growth period investigated, the capacity of photosystem I measured by the maximal absorbance change of P700 (PM) increased from 0.3 at 7 das until 0.7 at 10 das (Figure 4a). The values stayed significantly lower in both oeW1 lines, whereby oeW1-14 seedlings had lower values than oeW1-15 seedlings (Fig. 4a).
In addition to the efficiency of photosystem II, the maximum electron transport rate of photosystem II (ETRmax) was reduced in the primary foliage leaves of oeW1 seedlings measured on different days after sowing (7-10 das). The results showed that oeW1-14 leaves had only about 50% of the electron transport capacity of WT leaves, while oeW1-15 leaves had about 80% of the WT level (Figure 4a). Whereas the transport rate in WT and oeW1-15 leaves was maximal already at 9 das, it still increased in oeW1-14 leaves from 9 das until 10 das. An analysis of the partitioning of absorbed energy in photosystem II revealed that the quantum yield of photosystem II (Ф(II)) was lower in oeW1 plants at all stages of development compared to WT plants. The remaining fraction was dissipated mainly as heat or fluorescence (Ф(NO)). Only a small fraction was used for non radiative dissipation in the oeW1 leaves (Figure 4b).
To investigate putative differences in the composition of the photosynthetic apparatus, the concentrations of pigments and the relative abundances of photosynthesis-associated proteins were analyzed. Pigment analyses by HPLC showed that the chlorophyll content per leaf area of primary foliage leaves from seedlings of the oeW1-50 leaves was lower than the chlorophyll contents of line oeW1-15 and the wild-type, which were similar (Figure 5). The ratio of chlorophyll a/b was reduced in both oeW1 lines (Figure 5) and was independent of the leaves developmental stage.
The content of xanthophyll cycle pigments was similar in all genotypes. However, the xanthophyll pool/chlorophyll ratio was higher in the leaves of the oeW1 lines, in particular in the oeW1-50 leaves.
Protein extracts from primary foliage leaves of WT and oeW1 seedlings were immunologically analyzed for the levels of central photosystem I (PsaA), photosystem II (D1/PsbA) proteins, and two light-harvesting proteins, i.e. LHCA1 and LHCB1, respectively. As already reported, the abundance of WHIRLY1 in the WT declined with the increasing age of the leaves (Kucharewicz et al., 2017, Krupinska et al., 2019). The levels of all tested proteins were similar in the WT and the oeW1 lines. (Figure 6). The staining also indicates that the two subunits of RubisCO had the same abundance as in the WT and that the levels are stable during development from 7 das until 12 das.
In order to investigate gene expression in chloroplasts from oeW1 leaves in comparison to WT leaves, mRNA levels of genes encoding central components of the photosynthetic apparatus were analyzed by RT-PCR (Figure S2). While mRNA levels of all genes declined during the development of WT leaves, the mRNAs stayed at relatively high levels during the development of the oeW1 leaves. While in RNAi-W1 plants, plastid gene expression was mainly due to the activity of the nuclear-encoded RNA polymerase (NEP) (Krupinska et al., 2019), in the oeW1 lines transcripts of both NEP (rpoB, clpP) and PEP (psbE) were present at higher levels than in WT plants. This result indicates that overexpression of WHIRLY1 did not hamper transcription in chloroplasts.
Chloroplast ultrastructure and nucleoid morphology
When primary foliage leaves of barley were fully expanded (10 das), ultrathin sections from WT and oeW1-14 seedlings grown in a daily light/dark cycle were fixed for ultrastructural analyses by transmission electron microscopy. While mitochondria and peroxisomes looked rather similar in WT and oeW1 samples (Figure 7a), chloroplasts showed noticeable morphological differences (Figure 7b). Chloroplasts of oeW1 plants apparently contained more plastoglobules (Figure 7). Plastoglobules are lipoprotein particles surrounded by a lipid monolayer, which is contiguous with the outer leaflet of thylakoid membranes. They contain mainly isoprenoid-derived lipophilic compounds and function in remodeling the lipid phase of thylakoids (van Wijk and Kessler, 2017). An increase in the number and/or size of plastoglobules was reported to indicate excess light in the photosynthetic apparatus (Brehelin et al., 2007, Rottet et al., 2016) and was observed during various stressful situations (Lichtenthaler, 2013, van Wijk and Kessler, 2017). Thylakoids in the chloroplasts of oeW1 leaves showed a tendency to swell. Following the lower photosynthetic activity of oeW1 leaves (Figure 3), chloroplasts of the oeW1 plants did not contain starch grains which were frequently observed in the wild-type chloroplasts.
Considering that WHIRLY1 is a major nucleoid-associated protein (Pfalz et al., 2006, Krupinska et al., 2022b), special attention was committed to the structure of nucleoids in mature chloroplasts of the primary foliage leaves of WT and oeW1 seedlings. Ultrastructural analyses did, however, not reveal apparent differences between nucleoids in WT and oeW1 sections (Fig. 6c-d). In addition, nucleoids of mature chloroplasts were also visualized by light microscopy after staining sections with SYBR Green (Figure 7e). By this procedure, neither differences in size nor the distribution of nucleoids were observed between the two genotypes.
Previous studies revealed a profound impact of WHIRLY1 on the packaging of plastid DNA (Krupinska et al., 2014) and bacterial nucleoids (Oetke et al., 2022). Considering that WHIRLY1 also plays a significant role in chloroplast development (Prikryl et al., 2008, Krupinska et al., 2019, Krupinska et al., 2022), it might be possible that putative differences in nucleoid morphology depend on the developmental stage of plastids. Using the developmental gradient of the leaves of small-grained cereals (Boffey et al., 1980), putative development-related changes in nucleoid morphology were investigated by staining sections prepared from the base and the middle part of the leaves of WT, oeW1-2 having a 50-fold over-accumulation as in oeW1-14 plants, and oeW1-15 seedlings with a 10-fold over-accumulation of WHIRLY1, respectively. Although WHIRLY1 abundances in chloroplasts are dramatically different between WT and an oeW1-50 line, no apparent differences in nucleoid morphology were observed. In all genotypes, nucleoids in undifferentiated plastids at the base are arranged like pearls on a string. In contrast, nucleoids in mature chloroplasts are dispersed inside the chloroplasts (Figure S3) due to their attachment to thylakoid membranes (Powikrowska et al., 2014).
Hormone levels and defense-related gene expression
To elucidate whether the reduced growth of the oeW1 plants was related to changes in the levels of growth hormones, cytokinins and auxins were determined at 7 and 10 das in primary foliage leaves of plants grown at a light intensity of 100 μmol m-2 s-1. These measurements revealed that independent of the developmental stage, the levels of the cytokinin N6- isopentenyl adenosine (iPR) were reduced by about 30% or 60% in the primary foliage leaves of oeW1-15 or oeW1-14 plants, respectively (Figure 8). For comparison, indole-3-acetic acid (IAA) levels were similar among the lines. At 10 das a reduction in the level of IAA by about 20% was measured in the leaves of the oeW1-14 line (Figure 8).
For comparison, the levels of hormones involved in defense were determined. While free salicylic acid was too low to be determined, its catabolite and storage compound were detectable. It has been shown that in barley during pathogen defense, SA is not produced via the isochorismate (ICS) pathway (Vlot et al., 2009, Rekhter et al., 2019), but rather by the phenylpropanoid pathway controlled by phenylalanine lyase (PAL) (Qin et al., 2019). The levels of 3,4-dihydroxy benzoic acid were about twofold higher in young leaves of the oeW1-14 leaves compared to the other lines, but this difference disappeared when leaves were collected at 10 das (Figure S6). Levels of SA glucosides which is a storage form of SA (Vlot et al., 2009), showed a tendency to be higher in the leaves from oeW1-14 plants, in particular in young leaves (7 das) (Figure S6). The determination of jasmonic acid (JA) revealed that at 8 das overexpression of HvWHIRLY1 significantly increased its level by a factor of 3 or 5 in oeW1-15 or oeW1-14 lines, respectively. Thereby the basis level measured in the WT leaves was slightly enhanced at 10 das, compared to 8 das (Figure 8). Taken together, the results revealed that over-accumulation of WHIRLY1 induced reciprocal changes in the levels of iPR and JA which might cause a shift from growth to defense.
To investigate whether according changes in gene expression accompanied the transition from development to defense, mRNA levels of key enzymes in the biosynthesis of defense hormones were determined besides the levels of WHIRLY1 mRNA and PR1 mRNA by quantitative real-time PCR. The result showed that HvWHIRLY1 had an up to 20-fold higher mRNA level in primary foliage leaves of oeW1-14 seedlings compared to the WT. PR1 is known as a marker of SAR (Linthorst, 1991). While in Arabidopsis and other dicots, it was reported to be a target gene of salicylic acid (Van Loon and Van Strien, 1999, Golshani et al., 2015), PR1 in rice was shown to accumulate in response to JA (Rakwal and Komatsu, 2000, Jwa et al., 2006). In this study, barley PR1 expression was upregulated in WT and oeW1 seedlings when the leaves became fully expanded. While in the WT, PR1 expression was upregulated by a factor of 6 at 10 das and by a factor of 11 at 12 das, in oeW1 seedlings, expression of the gene was highly upregulated by a factor of 120 at 10 das (Figure 9). Upregulation of PR1 in the oeW1 seedlings was neither accompanied by upregulation of the gene encoding PAL nor ICS, which are the key enzymes of the two pathways of salicylic acid biosynthesis (Vlot et al., 2009) (Figure 9).
Expression of the general stress associated HPL gene encoding hydroperoxide lyase, a chloroplast protein of the oxylipin pathway shown to protect against photoinhibition (Savchenko et al., 2017), was only upregulated by about 50% in fully expanded primary foliage leaves of the oeW1 seedlings. This gene was chosen because it is known to be a stress indicator gene regulated by retrograde signaling during stress in Arabidopsis (Xiao et al., 2012, Xiao et al., 2013). Its expression barely changed in WT seedlings during normal growth (Figure 9). In contrast to PR1, the expression of THIO1, another barley defense gene (Leybourne et al., 2022), was downregulated during growth in both WT and oeW1 seedlings (Figure S4).
Response of oeW1 plants to high light
In Arabidopsis, defense signaling is also activated in response to high light (Mateo et al., 2006, Karpinski et al., 2013). To induce high light stress, oeW1-14 and WT seedlings were grown in continuous light of 350 μmol m-2 s-1 (HL) and were compared to seedlings grown at only 100 μmol m-2 s-1 (LL) as described previously (Swida-Barteczka et al., 2018). Growth at high light leads to a decrease in the chlorophyll content of both WT and oeW1 plants. The reduction in chlorophyll content of oeW1 plants was significantly more pronounced than the WT (Figure 10a) but was not as prominent as in the case of the WHIRLY1 knockdown plants prepared by RNAi (Swida-Barteczka et al., 2018). In WT seedlings, FV/FM was not affected by higher irradiance during growth, while it even slightly increased in the case of oeW1-14 seedlings (Figure 10a).
When WT seedlings were grown at HL, PR1, and PR10 expression levels were elevated compared to the levels determined in LL-grown plants. This result follows the idea that SA is involved in response to HL. Overexpression of WHIRLY1 led to a dramatic increase in the expression of both PR genes (Figure 10b). Moreover, over-accumulation of HvWHIRLY1 led to enhanced expression of PAL, which was more pronounced at HL than at LL (4-fold in comparison to LL) (Figure 10b). Expression of PAL but not of ICS was also slightly enhanced in the WT at HL. In comparison, ICS expression was enhanced in oeW1 plants only at LL, but not at HL. The high expression of ICS at LL could be related to an increased demand for phylloquinone (Qin et al., 2019).
In addition, the expression of genes encoding two key enzymes of the two branches of the oxylipin biosynthesis in chloroplasts (Savchenko et al., 2017) was determined, i.e. HPL leading to the biosynthesis of aldehydes and allene oxide synthase (AOS), a key enzyme of JA biosynthesis (Delker et al., 2006). In the WT, HPL was upregulated by 4-fold, while AOS was upregulated by a factor of 5.5 (Figure 10b). In oeW1 seedlings, expression of HPL was already enhanced at LL and was only upregulated by 40% in HL compared to LL. Indeed, the expression levels at HL were identical between WT and oeW1 plants. The expression of AOS is upregulated likewise in HL in both the WT and the oeW1 plants.
Response of oeW1 plants to powdery mildew
To investigate the impact of WHIRLY1 accumulation on pathogen resistance, leaves were inoculated with spores of the powdery mildew fungus Blumeria graminis, an important barley pathogen. The susceptibility to powdery mildew was compared among WT, oeW1-14, oeW1-2 (two lines over-accumulating WHIRLY1 by a factor of 50), and two barley plants with an RNAi-mediated knockdown of HvWHIRLY1, W1-1 and W1-7 (with 10% and 1% of the protein in WT, respectively), which had been used in several investigations before (Krupinska et al., 2014b, Krupinska et al., 2019). Both oeW1-14, oeW1-2 were less susceptible to powdery mildew than the WT, as determined by estimating the percentage of the leaf surface infected by the fungus (Figure 11). Inversely, the leaves of the WHIRLY knockdown plants (W1-1 and W1-7) were more susceptible to inoculation with powdery mildew spores. The results show that a high abundance of WHIRLY positively affects the resistance towards powdery mildew.
DISCUSSION
Overexpression of WHIRLY1 in barley resulted in an up to 50-fold higher abundance of the WHIRLY1 protein, an improved tolerance towards powdery mildew, and diminished growth, indicating a typical tradeoff between growth and defense (Herms and Mattson, 1992, Huot et al., 2014). Although the tradeoff has often been explained by the competition of energy requirements of defense responses in relation to those for growth and reproduction, this apparently plausible explanation has also been questioned. Instead, the dilemma between development and defense was shown to stem from antagonistic crosstalks between growth and defense-related hormones (Karasov et al., 2017), which can be uncoupled in mutants (Campos et al., 2016).
Impact of WHIRLY1 overexpression on growth and photosynthesis
For oeAtWHIRLY1 plants, no obvious phenotype has been reported (Isemer et al., 2012a). A more detailed characterization has been performed with tomato lines overexpressing SlWHIRLY1 (Zhuang et al., 2019). In these plants, the mRNA level increased dramatically by factors of about a thousand. In contrast, the protein level was only enhanced by an estimated factor of approximately five (estimated from Figure 2 in Zhuang et al. 2019). No significant difference was observed in the phenotypes between tomato oeSlWHIRLY1 lines and the wild type at standard growth conditions. However, under chilling conditions, the oeSlWHIRLY1 lines grew better than the wild-type (WT) coinciding with a reduced level of ROS, as shown by fluorescence after staining with H2DCFDA (Zhuang et al., 2019). At the ultrastructural level, the oeSlWHIRLY1 plants were shown to retain intact grana thylakoids and to accumulate less starch in chilling conditions. However, in contrast to the barley lines, overexpressing HvWHIRLY1 (oeW1), under control conditons the abundance of starch grains did apparently not differ between WT and oeSlWHIRLY1 plants (Zhuang et al., 2019). Also in contrast to the barley oeW1 lines, the oeSlWHIRLY1 plants showed no difference in FV/FM at 25°C and even higher FV/FM values under chilling conditions (Zhuang et al., 2020b). Also, in contrast to the barley oeW1 plants, RubisCO content was higher in the oeSlWHIRLY1 plants than in the WT, both at 25°C and 4°C (Zhuang et al., 2020b). Under heat stress, the oeSlWHIRLY1 plants showed less wilting than WT tomato plants coinciding with increased sugar content and a reduced level of ROS (Zhuang et al., 2020a).
In contrast to the barley oeW1 plants, the two WHIRLY1 overexpressing dicot species investigated didn’t show a pronounced decrease in growth under standard conditions. Compared to the barley lines used in this study, over-accumulation of the protein in tomato is relatively low and could be a reason for the discrepancies between barley and tomato. Alternatively, the growth-related difference between WHIRLY1 over-accumulation in barley on the one hand, and tomato or Arabidopsis, on the other hand, could be due to differences in the impact of WHIRLY1 proteins on chloroplast nucleoid architecture. Only in monocots WHIRLY1 proteins were shown to have a specific PRAPP motif required for the compaction of nucleoids (Oetke et al., 2022). However, despite the over-accumulation of WHIRLY1, nucleoids did not show differences in their compactness and organization between WT and oeW1 plants as investigated by DNA staining (Figure 7, S3). This result is in line with the almost normal levels of plastid-encoded mRNAs (Figure S2) and the unvaried protein composition of the photosynthetic apparatus (Figure 6). Regarding these results, it is rather unlikely that alterations in the nucleoid compactness and the composition of the photosynthetic apparatus are responsible for the reduced growth of barley plants over-accumulating WHIRLY1.
On the other hand, the efficiencies of both photosystems, the maximal electron transport rate (ETRMAX), and the quantum yield of photosystem II (Φ(II)) were reduced in plants over-accumulating WHIRLY1. Inversely, the loss of absorbed energy by heat and fluorescence was enhanced in oeW1 plants indicating a malfunctioning of the photosynthetic apparatus. Potentially, changes in the hormone equilibrium could underlie the lower functionality of the photosynthetic apparatus (Muller and Munne-Bosch, 2021, Cackett et al., 2022). In this regard, the barley oeW1 plants might be comparable with mutants showing constitutive defense signaling. Arabidopsis mutants with constitutive expression of pathogenesis-related proteins (cpr) showed a dwarf phenotype (Zhang et al., 2003, Heidel et al., 2004). To investigate whether the impaired growth is a consequence of deteriorated photosynthesis or energy-consuming defense mechanisms, Mateo et al., (2006) investigated the photosynthetic properties of cpr mutants in comparison to the WT. Similar to the Arabidopsis cpr mutants, barley seedlings overexpressing WHIRLY1 have a reduced FV/FM, a higher ratio of VAZ pool pigments to chlorophylls, and reduced starch content as a consequence of the reduced capacity of the photosynthetic apparatus (Figure 5, 7).
Overexpression of WHIRLY1 caused changes in the equilibrium of hormones
Over-accumulation of WHIRLY1 indeed caused a shift in the hormone equilibrium. While the level of the cytokinin isopentenyl riboside (iPR) was reduced, the level of jasmonic acid (JA) is enhanced in oeW1 plants. Cytokinins are well-known for their positive impact on cell division and expansion during leaf development and growth (Brzobohaty et al., 1994, Wu et al., 2021). Moreover, cytokinins were shown to promote chlorophyll biosynthesis, assembly, and functioning of the photosynthetic complexes (Yaronskaya et al., 2006) and to play a role in responses to stress (Albrecht and Argueso, 2017, Cortleven et al., 2019). Cytokinins were found to regulate more than 100 genes involved in photosynthesis, including the genes of RubisCO and LHCs (Brenner and Schmulling, 2012) and those encoding sigma factors required for plastid gene transcription by PEP (Danilova et al., 2017). Applying cytokinins to wheat leaves increased endogenous cytokinin content and photosynthesis parameters Φ (PSII), FV/FM, and ETR, whereas inhibition of cytokinin biosynthesis had opposite effects (Yang et al., 2018). The published data suggest that a decreased cytokinin level led to an inactivation of photosystem II reaction centers (Muller and Munne-Bosch, 2021). Hence, the reduced efficiencies of the photosystems, together with the decreased ETR and quantum yield of photosystem II in the oeW1 leaves, are potentially caused by the decrease in the level of iPR. Under HL, cytokinins were reported to promote D1 repair (Cortleven et al., 2019). Whereas in the barley plants grown at continuous light of low irradiance, cytokinin levels were low in all genotypes, at high irradiance, the level increased in the WT but not in the oeW1 plants (Figure 8).
Nevertheless, FV/FM was higher in the oeW1-14 plants in HL compared to LL. This might indicate that oeW1 plants, compared to WT plants, have a better capacity to respond to HL. A similar finding has been reported for the response of tomato plants overexpressing WHIRLY1 towards chilling (Zhuang et al., 2020b).
In comparison to the cytokinin level, the level of the major auxin IAA was less affected in the barley oeW1 plants. This coincided with similar expression levels of selected genes responding to auxin, i.e. PIN1 and TIR1 (Figure S 4). Expression of genes related to auxine biosynthesis such as the YUCCA genes was neither detectable in the WT nor in the oeW1 plants. The level of JA was enhanced in oeW1 plants under standard growth conditions (Figure 8) and during growth in continuous light of low irradiance (Figure S6). It is known that a rise in JA has a negative impact on photosynthesis (Attaran et al., 2014, Muller and Munne-Bosch, 2021) and growth (Staswick et al., 1992). Recently it has been shown that the treatment of barley leaves with JA affects photosynthesis at the level of the oxygen-evolving complex (Kurowska et al., 2020). During growth in continuous light of high irradiance, the level of JA increased in the wild type. As a consequence, the differences among the genotypes measured at low light irradiance disappeared (Figure S6).
In leaves collected under standard growth conditions, the higher expression of defense-related genes such as PR1 and HPL, the latter of which has been proposed as general stress indicator genes (Savchenko et al., 2017), indicates that over-accumulation of WHIRLY1 activates defense signaling. Unexpectedly, the level of salicylic acids (SA) stayed below the method’s lowest quantification limit, i.e. 0.1 μM. If the over-accumulation of WHIRLY1 would have induced its synthesis, SA would have increased above a level of 1 μM. The SA-related compounds also did not show changes associated with WHIRLY1 quantities. It may be supposed that SA signaling is not affected by the overexpression of WHIRLY1 in barley. While in barley, only a limited number of pathogens induced an increase in the level of SA, all tested pathogens induced the expression of PR genes, including PR1 (Vallelian-Bindschedler et al., 1998). Obviously, SA in barley is not always required for defense-related gene expression. In the barley plants grown at continuous high light, also PR10 expression was enhanced. This gene might be expressed in response to the simultaneous presence of JA and light as reported for rice (Rakwal et al., 2001, Zheng et al., 2021). A minor contribution of SA to the defense response cannot be excluded considering that expression of PAL encoding the key enzyme of salicylic acid biosynthesis is activated in HL both in the WT and much more in the oeW1-50 plants (Figure 10b).
In Arabidopsis, WHIRLY1 was shown to be involved in salicylic acid (SA) signaling independent of NPR1 in the cytosol (Desveaux et al., 2004, Vlot et al., 2009, An and Mou, 2011, Carella et al., 2015). NPR1 is known to translocate from the cytosol to the nucleus upon binding of salicylic acid and thioredoxin-mediated reduction (Mou et al., 2003). It has been proposed that WHIRLY1 is translocated from chloroplasts to the nucleus in a similar fashion upon stress-associated redox changes in the photosynthetic apparatus (Foyer et al., 2014), whereby the mechanism of translocation remains unknown (Krupinska et al., 2022). In the oeW1 plants described in this study, the level of nucleus-located WHIRLY1 is highly upregulated even in the absence of stress. Consequently, in the barley oeW1 plants, defense signaling is constitutively activated, as evident by the expression of PR1 in fully expanded leaves of seedlings grown under standard growth conditions (Figure 9) and during continuous illumination of low irradiance (Figure 10b). It is obvious that the WHIRLY1-activated defense signaling is mediated by JA rather than by SA. This result is in accordance with reports on JA-dependent defense activation involving PR1 in rice (Yang et al., 2013).
By the growth of the oeW1 plants at high irradiance, a dramatic increase in expression of PR1 (450-fold instead of 70-fold in the wild type) and PAL (6-fold instead of only 50% in the WT) was observed (Figure 10b). This indicates that the oeW1 plants are capable of further enhancing defense responses. Considering that the abundance of WHIRLY1 is already high in non-stress conditions, it is unlikely that the higher expression of defense genes is caused by a further increase in WHIRLY1-dependent transcription of these genes. Rather WHIRLY1 abundance may intensify the binding of activating factors to the promoter of PR1 under certain conditions. Recently, it has been demonstrated that NPR1-mediated PR1 gene expression requires the formation of an activating complex consisting of histone acetyltransferase (HAC), NPR1, and a TGA transcription factor (Jin et al., 2018). Potentially, WHIRLY1 might regulate the accessibility of promoters for defense-associated transcription factors (Krupinska et al., 2014a, Krupinska et al., 2022a).
Surprisingly, in barley plants overexpressing WHIRLY1, the gene encoding isochorismate synthase (ICS) is activated at control conditions. This high expression could be related to a demand for phylloquinone which is essential for electron transfer in photosystem I. A barley ics mutant was reported to be deficient in phylloquinone, whereas it was not altered in the basal level of salicylic acid (Qin et al., 2019). Salicylic acid biosynthesis may proceed by two possible pathways, the ICS and PAL pathways, which both start from chorismate in chloroplasts (Lefevere et al., 2020). In Arabidopsis, only 10% of SA is produced by the PAL pathway, while 90% is produced by the ICS pathway (Garcion et al., 2008). By contrast in barley, ICS expression during HL exposure is lower than at LL (Figure 10b), while PAL expression is enhanced by a factor of 6. This is in accordance with the idea that in barley during stress the PAL pathway of SA biosynthesis is more critical than the ICS pathway. Since, in contrast to PR1, the expression levels of PAL (Figure 10) and of the defense gene THIO1 (Figure S4) were not elevated by overexpression of WHIRLY1 under normal growth conditions, it is unlikely that these genes are directly regulated by WHIRLY1. In contrast, PR1 and HPL were activated in the oeW1 plants both under normal growth conditions and at HL and, therefore might be directly activated by WHIRLY1.
The role of chloroplast-nucleus located WHIRLY1 in the growth-defense tradeoff
The reduced growth of oeW1 plants and the enhanced resistance towards powdery mildew indicate that overexpression of WHIRLY1 shifts the balance between growth and resistance to the latter. In recent years hormone crosstalk has emerged as a major player in regulating the growth-defense tradeoff (Huot et al., 2014). Although the antagonistic crosstalk between SA and the growth hormone auxin mostly has been reported to determine the tradeoff between growth and defense (Huot et al., 2014), overexpression of WHIRLY1 in barley had more impact on the levels of cytokinins and JA than on those of auxin and SA, suggesting that in this species the tradeoff is regulated by cytokinin and JA However, most studies on hormonal interactions during growth and defense have been performed with Arabidopsis. It is likely, that hormonal interactions in monocot plants are different, as has been reported for rice (De Vleesschauwer et al., 2013). It has been postulated that during immune responses in rice, NPR1-dependent SA-signaling is activated by JA binding to the COI1 receptor without a change in the level of SA (Yang et al., 2013). This model is in accordance with earlier reports on barley infection by powdery mildew, in which sensitivity to powdery mildew was found to be not accompanied by accumulation of SA (Vallelian-Bindschedler et al., 1998, Hückelhoven et al., 1999).
The high accumulation of WHIRLY1 in chloroplasts of the barley oeW1 plants had neither consequences for nucleoid organization nor plastid gene expression. Hence the reduced growth was likely not caused by changes in the plastid gene expression machinery but rather by the enhanced level of nucleus-located WHIRLY1, inducing changes in gene expression that eventually lead to a rewiring of hormonal homeostasis. The identical molecular weights of chloroplast-located WHIRLY1 and nucleus-located WHIRLY1 clearly indicate that both pools of WHIRLY1 had been processed to the mature form inside chloroplasts. Hence WHIRLY1 over-accumulating in the nucleus was transferred from chloroplasts to the nucleus as demonstrated before with transplastomic tobacco plants synthesizing WHIRLY1 inside chloroplasts (Isemer et al., 2012b). In another previous study, it had been shown that Arabidopsis plants accumulating WHIRLY1 outside the chloroplasts behave like a WHIRLY1-deficient mutant (Isemer et al., 2012a). Hence, the nuclear activities of WHIRLY1 require its preceding presence in chloroplasts. Whether WHIRLY1 undergoes a modification inside chloroplasts and how its transfer to the nucleus is mediated remains to be determined. Taken together, the findings of this study suggest that the WHIRLY1-mediated adjustment of hormonal homeostasis is controlled by chloroplasts which are crucial sensors of environmental information (Pfalz et al., 2012, Zhang et al., 2020).
According to the elevated PR1 expression in the absence of stress, barley oeW1 plants showed a constitutive defense response. To avoid a negative impact on growth, the expression of resistance genes might be restricted to the time of stress perception and the subsequent defense response (Karasov et al., 2017). Sequestering of WHIRLY1 in chloroplasts is a means to avoid its nuclear activity under non-stress conditions and to allow a fast response to stress only under conditions that induce the transfer of WHIRLY1 from chloroplasts to the nucleus (Krause and Krupinska, 2009). However, the 10 to 50 times higher level of WHIRLY1 in the oeW1 plants obviously exceeded the capacity for WHIRLY1 sequestration by chloroplasts. It remains to be tested whether a moderate increase in WHIRLY1 accumulation in the chloroplast is possible without transfer to the nucleus in non-stress conditions, thereby avoiding the constitutive expression of PR1 in the nucleus.
EXPERIMENTAL PROCEDURES
Plant material and growth conditions
Transgenic barley plants overexpressing HvWHIRLY under the control of the maize UBIQUITIN 1 promoter were generated by the transformation of barley immature embryos by Agrobacterium tumefaciens as described (Hensel et al., 2008). The pENTR/TOPO Gateway vector (Invitrogen, Karlsruhe, Germany) was used for the transfer to the pIPKb007 binary vector using Gateway™ LR as described (Himmelbach et al., 2007). Plantlets resistant to hygromycin were transferred into soil and cultivated in a greenhouse. Additionally, PCR analyses with primers (Krupinska et al. 2014, Supplementary Table 1) for the hygromycin resistance cassette were performed to verify the transgene integration. As control plants, the barley cultivar “Golden “Promise’ and for powdery mildew assays, the HvWHIRLY1 knockdown plants (RNAiW1-7) (Krupinska et al., 2014b) were used.
Barley grains were sown on soil (Einheitserde ED73, Tantau, Ütersen, Germany) and transferred for three days in a dark and cold chamber (6°C) to synchronize germination. Thereafter, the grains were transferred to a climate chamber where the seedlings were grown either in a standard daily light/dark cycle (16:8) as described (Krupinska et al., 2019) or in continuous light of different irradiances (100 or 350 μmol photons m-2 s-1) as also described previously (Swida-Barteczka et al., 2018).
Quantum yields of the photosystems and electron transport rate
The maximum quantum yield of photosystem II, FV/FM, and the maximum P700 (PM) signal were measured in parallel by Dual-PAM-100 (Walz GmbH, Effeltrich, Germany). The leaves were kept for about 10-15 minutes under low light (20-40 μmol m-2 s-1) before starting the measurement. The measurement was done at 13 different light levels, starting from zero and gradually increasing during six minutes to 1600 μmol m-2 s-1. In between of these light levels, there was a step with 60 μmol m-2 s-1 which is similar to the growth light in the climate chamber. The quantum yields of photosystem II as well as of non-radiative and radiative dissipation were calculated as follows (Klughammer and Schreiber, 2008): Ф(II) = (FM′−F)/FM′, Ф(NPQ) = F/FM′- F/FM, Ф(NO) = F/FM.
Determination of pigments by high-performance liquid chromatography
For the analysis of pigments, one cm long leaf segments excised from the area between 1.5 and 3 cm below the leaf tip were immediately frozen in liquid nitrogen and kept at −80°C. Pigments were extracted and HPLC analysis was performed as described (Saeid-Nia et al., 2022). To calibrate the detector (Nichelmann et al., 2016), pure carotenoid extracts (except antheraxanthin) were prepared through thin-layer chromatography (modified after Lichtenthaler and Pfister 1978). Afterwards, the concentrations of the pure pigment solutions were determined by spectrophotometry using the extinction coefficients provided by Davies (1976).
Immunoblot analyses
Total proteins were extracted from ground leaf material and subjected to SDS-PAGE, as reported (Krupinska et al., 2014, 2019). Proteins were transferred onto the nitrocellulose membrane by semi-dry electroblotting. Antibodies against PsaA (AS06172), PsbA/D1 (AS01016), LHCA1 (AS01005), and LHCB1 (AS01004) were purchased from Agrisera. The antibody directed towards HvWHIRLY1 was prepared against a synthetic peptide and can be purchased from Agrisera (AS163953). Immunoreactive complexes were visualized using a peroxidase-couples secondary antiserum with chemiluminescence detection kits (ECL Select, Amersham, USA; Lumigen, Southfield, MI, USA). The ChemiDoc MP Imaging Systems and the Image Lab 6.1 software (Bio-Rad Laboratories, Munich, Germany) were used for the quantification of signal intensities.
Determination of hormones
Leaf samples of ca. 30 mg (fresh weight) were weighed into 2 ml safe lock tubes (Eppendorf AG, Germany) and kept at -80°C until analysis. Empty tubes were used as blanks. Before extraction, two 3 mm ceria-stabilized zirconium oxide beads were placed into each tube. The samples were extracted and purified as described by Šimura et al. (2018) with minor modifications (Simura et al., 2018). The absolute quantification of all targeted phytohormones, excluding salicylates, was performed as described (Eggert and von Wiren, 2017).
The analysis of salicylates was performed using UHPLC-HESI-HRMS (Vanquish UPLC) coupled to QExactive Plus Mass Spectrometer (San Jose, CA, USA). The MS was equipped with a HESI source operating in negative ion mode. Salicylates baseline separation was achieved on a reversed-phase Acquity UPLC® HSS T3 column (10 Å, 2.1 × 100mm, 1.8μm, Waters) using a gradient elution of A (Water, 0.1% FA) and B (ACN, 0.1% FA) as follows: 0– 5min, 5% B; 5–10min, 5% to 80% B. Additional five minutes were added for column washing and equilibration (total run time, 15min). The column temperature was set at 45°C and the flow rate at 0.5 ml·min-1. The injection volume was 5μl. Source values were set as follows: Spray voltage 2.5kV; capillary temperature 255°C; S-lens RF level 40; Aux gas heater temp 320°C; Sheath gas flow rate 47; Aux gas flow rate 11. For spectra acquisition, a Full MS/dd-MS2 experiment was performed. Resolution in Full Scan was set as 70000. For MS/MS experiments, resolution 17,500 and NCE 40V were used. The identification of compounds found in extracts was based on a comparison of their retention times, MS2 spectrum and exact mass with standards.
RNA isolation and real-time PCR analysis
Total RNA was isolated from primary foliage leaves of seedlings using the peqGOLD-TriFast reagent (Peqlab Biotechnology, Erlangen, Germany) according to the manufacturer’s protocol. cDNA biosynthesis and real-time PCR were performed as described previously (Krupinska et al., 2019). Data were normalized to the level of the ADP-ribosylation factor 1 mRNA (Rapacz et al., 2012), to cytosolic GAPDH or to mRNA of the barley histone acetyltransferase (HORVU.MOREX.r2.1HG0027750), which has the alternative name GENERAL CONTROL NONREPRESSIBLE 5 (GCN5).
Transmission electron microscopy
Leaf segments from primary foliage leaves (2×2mm) at a position of 2 cm below the leaf tip were fixed and processed as described (Krupinska et al., 2014b).
Staining and localization of nucleoids
Leaf cross-sections were produced by hand or by a hand microtome from the primary foliage leaf of plants grown for 7 days. Sections were fixed by 4% (w/v) paraformaldehyde in phosphate-buffered saline (PBS) overnight at 4°C. After washing with PBS containing 0.12 %(w/v) Glycin, the sections were stained with SYBR®Green (1:5000, S7563 InvitrogenTM) for 45 min in darkness at room temperature. After washing with 1x PBS for 15 min, the sections were transferred onto a slide, capped with PBS/glycerol (v/v: 1:1), and a coverslip. Imaging was done at Leica SP5 confocal microscope system with an HCX PL APO CS 63.0 × 1.2 W objective. Excitation was done by an argon laser line 488 (5% power). Emission was detected between 510-570 nm (HV750) and 690-760 nm (HV480). A minimum of five images out of different regions of the specimen were taken from each sample. Image analysis, coloring, and composition were done by ImageJ 1.53q.
Infection with powdery mildew
Five plants were grown in 12 cm pots in compost soil. In an inoculation device, transgenic lines with two pots each were arranged, with three pots containing wild type. While rotating in the inoculation tower, the fourteen-day-old seedlings were inoculated with Blumeria graminis spores (isolate CH4.8) until a spore density of approx. 10 spores per mm2 have been reached. The disease scored 7 d after inoculation, as described (Schweizer et al., 1995).
SUPPORTING INFORMATION
Additional Supporting Information may be found in the online version of this article:
Figure S1. (a) The abundance of WHIRLY1 in primary leaves of the lines oeW1-2 and oeW1-14. (b) Immunoblots showing the distribution of WHIRLY1 between chloroplasts and nucleus.
Figure S2. Relative levels of mRNAs of HvWHIRLY1 and selected plastid genes (rpoB, clpP, psaA, psbA, psbE) were determined by RT-PCR.
Figure S3. Nucleoid morphology in different parts of primary foliage leaves from wild type, oeW1-50 and oeW1-10 plants.
Figure S4. Expression of PIN1, TIR1, THIO1and GR1 in primary foliage leaves from oeW1-2 and oeW1-14 grown in a daily light/dark cycle.
Figure S5. Expression of defense genes in primary foliage leaves from oeW1-2 and oeW1-14 grown in continuous light of low (LL) or high irradiance (HL).
Figure S6. Hormone levels supplementing Figure 8. (a) Levels of SAG and DHBA in primary foliage leaves from oeW1-2 and oeW1-14 grown in a daily light/dark cycle, (b) levels of iRP, IAA and JA in primary foliage leaves from oeW1-2 and oeW1-14 grown in continuous light of low (LL) or high irradiance (HL).
ACKNOWLEDGEMENTS
We thank Sabine Sommerfeld (IPK Gatersleben) and Susanne Braun (Institute of Botany, CAU, Kiel) for their excellent technical assistance. We are grateful to the German Research Foundation for financial support (KR1350-19-1).