The whole agricultural sector is facing the forthcoming challenges to keep up productivity with a growing global population (Ray et al., 2013). Nowadays, high and continuous agricultural productivity is dependent on the use of chemical fertilizers and pesticides. Nevertheless, the excessive use of these compounds has adverse effects on human health and the environment (Carvalho, 2006). The recent progress in the use of either natural plant growth-promoting substances or microorganisms (also termed plant biostimulants) has allowed a reduction and optimized use of fertilizers. This combination of mineral nutrients and biostimulants allows a better nutrient use efficiency, a better crop tolerance against abiotic stresses, and indirectly, a better quality and an improved yield of the crops (Rouphael and Colla, 2020; du Jardin et al., 2020). Likewise, microorganisms/organisms or natural substances were also tested and used in agriculture as agents able to interfere by different means with the occurrence of plant diseases caused by pathogens. Natural substances encompass various types of biomolecules, which can be extracted from a vast number of plant species. Among these natural substances, plant elicitors are described as substances able to activate plant immune system, and further, to protect crop against various kind of pathogens and parasites both in conventional and organic agriculture (Wiesel et al., 2014; Jamiolkowska, 2020).

Plants have indeed developed an efficient immune system described in the zig-zag model from Jones and Dangl (2006), which can be triggered via (i) the perception of plant elicitors or pathogen/microbial/damage-associated molecular patterns (P/M/DAMPs) also known as pathogen-associated molecular pattern PAMP-triggered immunity (PTI) or (ii) specific pathogens’ effectors (effector-triggered immunity, ETI; Jones and Dangl, 2006). In field conditions, the mobilization of PTI by plant defence elicitors could lead to a pesticide reduction. In PTI, the recognition of plant elicitors by cell surface pattern-recognition receptors (PRRs) induces a series of early events, such as reactive oxygen species (ROS), nitric oxide production, and intracellular calcium influx. Then, intermediate events consist of activation of mitogen-activated protein kinases (MAPK) and phytohormone signaling (salicylic acid—SA, jasmonic acid—JA, and/or ethylene—ET; Bigeard et al., 2015). These signaling cascades trigger the induction of defense genes leading to the production of various defense-related compounds such as pathogensis-related (PR) proteins (Van Loon et al., 2006) or specialized antimicrobial compounds (Boller and Felix, 2009).

Many plant elicitors, also called PAMPs, which have been so far isolated and tested in both laboratory and field experiments, originated from microbes (MAMPs; e.g., flagellin; Zipfel et al., 2004), or plant themselves (DAMPs; oligogalacturonides; Hahn et al., 1981; Benedetti et al., 2015). A third category, called exogenous elicitors, which includes seaweed-based natural substances, was also reported to activate PTI. Interestingly, many macroalgae-based extracts were also reported to exhibit plant growth-promoting properties. Carrageenans are galactan-based polysaccharides commonly found in red macroalgae and were reported to exhibit plant-eliciting properties (Mercier et al., 2001). Laminarins, β-glucan-containing polysaccharides of brown macroalgae were also reported to exhibit plant-eliciting properties (Klarzynski et al., 2000; Aziz et al., 2003). Finally, ulvan polysaccharide, constitutive component of the cell walls of the green macroalgae Ulva genus was also reported to activate PTI (Cluzet et al., 2004; Jaulneau et al., 2010; Martin et al., 2020; Borba et al., 2021).

Ulva spp. belong to the class of Ulvophyceae, a group of green marine benthic algae, which dominates shallow marine environments and displays outstanding diversity regarding cytological and morphological characteristics (Wichard et al., 2015). Ulva spp. were shown to contain macro- and micronutrients, phytohormones, osmoprotectants, and other compounds with possible biological activities (Chbani et al., 2015; Shoubaky and Salem, 2016; Nabti et al., 2017). The presence of these compounds may accounts for well documented, Ulva extract-dependent, plant-growth promoting properties (Gireesh et al., 2011; Divya et al., 2015; Castellanos-Barriga et al., 2017; Paulert et al., 2021; Shefer et al., 2022). In addition to ulvans, Ulva spp. was also reported to contain fibrillar cellulose, mannan, or xylan polysaccharides in their cell walls (Domozych et al., 2012).

One class of proteins, the arabinogalactan proteins (AGPs), found in algal, moss, fern, and flowering plant cells walls and are strongly implicated in developmental processes (Lee et al., 2005; Nguema-Ona et al., 2012; Bartels and Classen, 2017; Renzaglia et al., 2017; Happ and Classen, 2019; Palacio-López et al., 2019) as well as in interaction with microorganisms (Nguema-Ona et al., 2013; Mareri et al., 2019). AGPs are proteoglycans consisting of two distinct moieties, the carbohydrate and the protein domain. The carbohydrate component typically accounts for 90%–98% of an AGP by weight and is rich in arabinose and galactose residues. The protein moiety, accounting for less than 10% of an AGP by weight is hydroxyproline-rich (Showalter, 2001; Seifert and Roberts, 2007). However, there is a wide range of variability in the structure and composition of both the carbohydrate and the polypetide parts. Based on the amino acid sequence and composition, AGPs were initially categorized into classical AGPs [consisting of a P/Hyp-rich domain heavily O-glycosylated, a hydrophobic C-terminal (C-ter) domain required for anchorage to the plasma membrane, and a signal peptide sequence] and non-classical AGPs (sometimes N-glycosylated and lacking the C-ter domain; Nguema-Ona et al., 2012). Non-classical AGPs also tend to be less heavily glycosylated (Showalter, 2001; Ma et al., 2018).

Although AGPs and AGP-like structures were reported to occur across the green and brown algae lineages, contrasting with the wealth of information available on land plant AGPs, much less is known about AGP occurrence, structure, and function in algae (Sørensen et al., 2011; Hervé et al., 2015; Palacio-López et al., 2019). Using immunocytochemistry and Yariv reagent, the presence of AGPs was described in several green microalgae of the freshwater-originated Charophyta division, specifically in Desmidiaceae, Coleochaetacea, Mesotaeniacea, Zygnemataceae, Chlorokybaceae, and Peniaceae families (Domozych et al., 2007, 2009; Eder et al., 2008; Sørensen et al., 2011; Palacio-López et al., 2019; Pfeifer et al., 2021). Furthermore, AGPs were detected as well in the Charale order, representing the multicellular algae with stem-like and leaf-like structures (Domozych et al., 2010). Within the Chlorophyta division, AGPs were also reported in Oedogoniaceae and Codiaceae families (Estevez et al., 2008, 2009; Fernández et al., 2010, 2015). Very recently, AGP-like glycoproteins were isolated for the first time from Ulva lactuca (Přerovská et al., 2021). In this study, Ulva AGP-like glycoproteins exhibited a contrasting reactivity with primary anti-AGP antibodies as well as with Yariv reagent when compared to AGP glycoproteins isolated from Solanum lycopersicum. While the amino acid analysis of the AGP-like glycoproteins purified by the β-d-glucosyl Yariv reagent showed a similarity between Ulva AGP-like glycoproteins and land plant AGPs, saccharide analysis revealed unique glycosylation of the Ulva lactuca AGP-like glycoproteins. Surprisingly, arabinose and galactose were not the most prevalent monosaccharides and the most outstanding was the presence of 3-O-methyl-hexose, which has never been described in the AGPs (Přerovská et al., 2021). Nevertheless, methylation of AGP glycans was previously reported (Bartels et al., 2017; Bartels and Classen, 2017; Happ and Classen, 2019; Temple et al., 2019; Pfeifer et al., 2020). Moreover, methylated glycoproteins and polysaccharides are widely distributed within algal cell walls (Ogawa et al., 2001; Bollig et al., 2007; Capek et al., 2008; Levy-Ontman et al., 2011; Staudacher, 2012; Mathieu-Rivet et al., 2014; Mócsai et al., 2019; Pfeifer and Classen 2020).

In the present study, an AGP-like enriched fraction from Ulva lactuca has been purified and chemically characterized. In order to find out if the Ulva AGP-like enriched fraction would exhibit plant eliciting properties, the fraction was tested for its ability to elicit the activation of PTI on oilseed rape (Brassica napus). Oilseed rape is widely grown in Europe, Canada, China, and Australia, and ranks second as oilseed production right after soybean (Neik et al., 2020; Raboanatahiry et al., 2021). Oilseed rape is confronted by a plethora of pathogenic agents, including Plasmodiophora brassicae, Leptosphaeria maculans, Sclerotinia sclerotiorumm, Hyaloperonospora parasitica, and others (Becker et al., 2017; Neik et al., 2017, 2020). Leptosphaeria maculans is a hemibiotrophic fungal pathogen causing blackleg disease, also called phoma stem canker (Lipková et al., 2021). The disease causes annually 10%–20% of yield losses (Van de Wouw and Howlett, 2020). The AGP-like enriched fraction was further tested for its ability to reduce the occurrence and the spread of blackleg disease on oilseed rape cotyledons. All along this work, the level of activation of PTI as well as the efficacy of the AGP-like enriched fraction was evaluated. In parallel, an ulvan was also isolated and its ability to both activate immune responses and to protect oilseed rape, was evaluated and compared to the AGP-like enriched fraction. Our results showed that the AGP-like enriched fraction was able to significantly activate PTI, and further, to protect the oilseed rape cotyledons from the occurrence and the spread of L. maculans. Interestingly, plants treated with the AGP-like enriched fraction showed a concentration-dependent reduction in the severity of L. maculans infection, while ulvan was effective only at the highest tested concentration. Likewise, the level of activation of PTI was more pronounced following the application of the AGP-like enriched fraction compared to the ulvan.

Seaweed extracts are already used in agriculture for their growth-promoting activity and ability to enhance plant stress tolerance for decades (Battacharyya et al., 2015). Although Ascophyllum nodosum products are the most commercially used (Sharma et al., 2014), Ulva sp. extracts have also high potential and have been the topic of various research papers (Dominguez and Loret, 2019). The most crucial challenge in seaweed biostimulant development is to choose the right extraction protocol, which will harvest all desired molecules with biostimulant activity. The protocol immensely affects the composition of final product and various protocols were established over the years. Although novel extraction technologies such as supercritical fluid extraction or microwave-assisted extraction are available, at the industrial level, the most common method is heating of algal biomass with potassium or sodium hydroxide solutions under pressure. Such harsh conditions can lead to the loss of some bioactive compounds plus result in uncontrolled fragmentation of polysaccharide chains, which consequently affect the biostimulant activity of the formulation (El Boukhari et al., 2020; Ali et al., 2021). Our results show, that even mild buffer extraction yields high amounts of proteins and saccharides with biostimulant activity within our extract. The eliciting activity of seaweed extracts is mainly attributed to the sulfated polysaccharides presented in their cell walls (Stadnik and Freitas, 2014). Nevertheless, algae contain a tremendous number of other biomolecules with potential biostimulant activity and AGPs belong to them. AGPs play a crucial role in higher plant defense responses and plant-microbe interactions (Nguema-Ona et al., 2013; Mareri et al., 2019), and their presence was confirmed also in brown and green algae (Estevez et al., 2009; Hervé et al., 2015; Ma et al., 2017; Přerovská et al., 2021). To test the ability of U. lactuca AGP-like glycoproteins to elicit defense responses and enhance stress resistance of higher plants, an AGP-like enriched sample containing approximately one-third of AGP-like glycoproteins was prepared by IEX chromatography. The ulvan extracted from U. lactuca was used for comparison. It activates plant immunity through the RBOH-dependent JA signaling pathway without inducing hypersensitive response (HR; Jaulneau et al., 2010; Freitas and Stadnik, 2015; Martin et al., 2020).

Up to now, seven studies focusing on the effects of different Ulva spp. extracts on plant infections with pathogens were published. Most of the studies used identical extraction protocol based on Cluzet et al. (2004), where dried Ulva material was autoclaved in distilled water, followed by ethanol precipitation. Some of the studies then call their product ulvan or some more generally as Ulva extract. In this study, ulvan was prepared according to Yaich et al. (2013) and based on excellent review of ulvan extractions done by Kidgell et al. (2019); our protocol should have high extraction yield, selectivity and low degradation. Only six studies also tried to analyze their products, mainly by means of spectrophotometric analyses, monosaccharide composition, and FT-IR analyses (Cluzet et al., 2004; Paulert et al., 2009, 2010; Jaulneau et al., 2011; Hernández-Herrera et al., 2014; Borba et al., 2019). Ulvan is mainly composed of rhamnose and glucuronic acid with the main repeating disaccharide unit (→4)-β-d-GlcAp-(1 → 4)-α-l-Rhap-(1→, in which glucuronic acid can be replaced to a certain extent by iduronic acid or xylose). Sulfation occurs mainly on C3 of the rhamnose and also C2 of the xylose or glucuronic acid (Lahaye and Robic, 2007; Kidgell et al., 2019). Although our results of ulvan analyses agrees with previously published data, outstanding is the presence of 3-O-methyl-hexose and unidentified saccharides, which possibly comes from contaminating glycoproteins. The most intriguing is the nature of negatively charged monosaccharide with retention time 36.47 min. Nevertheless, based on the knowledge of ulvan composition and the separation principle of the HPAEC/PAD technique, its long elution time indicates a strongly polar nature suggesting that we are dealing with a sulfated monosaccharide such as rhamnose-3-sulfate, which would correspond to its high content in ulvan (Templeton et al., 2012; Yaich et al., 2013). The composition of AGP-like enriched samples differs greatly from ulvan containing greater variety of monosaccharides and even more unidentified ones including rhamnose-3-sulfate. If true, the origin of this sulfated monosaccharide in the sample purified by ion-exchange chromatography, which should not contain ulvan, remains unknown. However, the correlation between sulfation and salt tolerance was previously proven (Aquino et al., 2011). Thus, the hypothetical presence of sulfated monosaccharide within AGP-like glycoproteins might be an adaptation to the marine environment. The other unidentified negatively charged monosaccharide with retention time 25.15 min is most probably 4-O-methyl-glucuronic acid, which would be in agreement with Pfeifer et al. (2020), who found out that Zostera marina AGPs contained high amounts of glucuronic acid and terminal 4-O-methyl-glucuronic acids, rare to land plant AGPs. We hypothesize, that the presence of unusual, modified monosaccharides might play an important role in eliciting plant resistance against the pathogen. The presence of galacturonic acid is interesting since it has rarely been described as part of AGP glycans (Tan et al., 2013). Though, high content of galacturonic acid and glucuronic acid was also identified in AGP-like glycoproteins of Micrasterias denticulata (Eder et al., 2008), which pointed out to the unique glycosylation of algal AGP-like glycoproteins. The high content of uronic acids was proposed as a specific adaptation to the marine environment, thanks to their calcium-binding capacity and the ability of calcium ions to protect plants from harmful effects of salt stress (Lahaye and Epstein, 1969, 1971; Cramer et al., 1985; Pfeifer et al., 2020). The ion-binding capacity of AGPs can be fine-tuned according to environmental factors (Lamport and Várnai, 2013; Pfeifer et al., 2020). Moreover, the essential role of pH-dependent periplasmic AGP–Ca²⁺ capacitor in signaling and normal plant development was reported (Lamport and Várnai, 2013; Lamport et al., 2014, 2018; Mizukami et al., 2016; Lopez-Hernandez et al., 2020). Importantly, big differences in composition can be found between the AGP-like enriched fraction prepared with the help of IEX chromatography and AGP-like glycoproteins obtained by Yariv precipitation from U. lactuca extract, whose composition was previously published (Přerovská et al., 2021). The discrepancies are caused most probably due to the distinct content of individual AGP-like glycoproteins, or the presence of contaminating compounds based on different purifications. For instance, the IEX purified product contained more proteins other than AGP-like glycoproteins, and possibly a large fraction of them is glycosylated making the data difficult to interpret. Moreover, Yariv precipitated AGP-like glycoproteins are highly enriched in AGP-like glycoprotein with molecular weight approximately 20 kDa compared to AGP-like enriched fraction prepared by IEX chromatography. Nevertheless, the composition of Yariv precipitated AGP-like glycoproteins still differed greatly from data known from land plant AGPs. Most importantly, the use of Yariv reagent is not suitable for large scale purification, due to the cost of Yariv reagent and mainly due to extremely low yields of the purification (Přerovská et al., 2021). Our results showed that pretreatment of B. napus plants with AGPs significantly reduced the development of L. maculans symptoms on cotyledons. To reveal the mechanism of fungus retardation induced by AGPs treatment, the direct antimicrobial activity tests of the compounds were performed in axenic cultures in vitro. Neither AGPs nor ulvan had any effect on the growth of L. maculans in this study. Unfortunately, no similar data are available for comparison. However, in the case of ulvan, various data can be found showing either no direct antifungal effect toward different pathogens (Freitas and Stadnik, 2012) or even enhancing the germination of conidia of C. lindemuthianum (Paulert et al., 2009). The latter corresponds to the enhanced growth of L. maculans, which can be explained in the same way. Simply, the polysaccharide can serve as a carbon source for the fungus. Nevertheless, since the compounds studied did not show any direct antifungal effect, but at the same time were able to reduce the severity of L. maculans infection at a certain concentration, it can be assumed that the protection is due to their elicitor activity. In general, elicitors trigger numerous signaling events that lead to the activation of the defense. Among the earliest is the ROS production of superoxide, hydroxyl radical and hydrogen peroxide. The latter plays a central role in biotic stress, including oxidative burst, cross-linking of cell wall proteins, callose deposition, signaling, defense gene expression, and hypersensitive response often manifested by systemic acquired resistance (Freitas and Stadnik, 2015; Waszczak et al., 2018). While AGPs caused a concentration-dependent production of H₂O₂, almost no H₂O₂ was produced after treatment with ulvan. Similar H₂O₂ accumulation was also found in B. napus cotyledons infiltrated with an oligosaccharide elicitor isolated from L. maculans mycelium (Kim et al., 2013).

These results are consistent with gene expression analysis, as ROS can potentiate the production of SA and SA-mediated signaling, leading to the expression of SA-responsive defense genes such as PR1. These findings agree with the proposed mode of action of A. nodosum extract (Stella Maris®; Cook et al., 2018). Moreover, the H₂O₂ produced could likely have direct antimicrobial activity, as the inhibitory effect of hydrogen peroxide on conidial germination and mycelial growth of L. maculans (Jindřichová et al., 2011) has been described previously. It is noteworthy that H₂O₂ production could be partly caused by phytotoxicity of AGPs, as necrosis formed after treatment with a high concentration (10 mg·ml⁻¹, data not shown). However, this fact is not a problem as even low concentrations lead to a significant reduction in the severity of infection without phytotoxic effects. Phytotoxicity phenomenon has already been described for other elicitors (Burketova et al., 2015; Trdá et al., 2019). Several articles describing the effect of ulvan treatment on H₂O₂ production showed a different response depending on the plants used and the priming of ROS production (Paulert et al., 2010; Abouraïcha et al., 2015; Freitas and Stadnik, 2015). Although our results seem to contradict the findings of Freitas and Stadnik (2015), where ulvan treatment resulted in higher increase of H₂O₂ production in A. thaliana compared to our findings, they used ulvan from Ulva fasciata and a different extraction methodology. Ulvan composition is highly dependent on the source species, ecophysiology, extraction, and processing procedure, which causes diverse bioactivity profiles (Kidgell et al., 2019). In order to determine the origin of the hydrogen peroxide produced, the expression of two NADPH oxidases, also called respiratory burst oxidase homologs (RBOHs), was analyzed. RBOH D and RBOH F were chosen, since they are known to be key players in stress responses in various plant pathosystems (Torres et al., 2002; Morales et al., 2016; Jasso-Robles et al., 2020). Both AGPs and ulvan caused a significant increase in the expression of RBOH F, but not RBOH D. Although these two enzymes cooperate during ROS generation, they are thought to play different roles in the regulation of hypersensitive response. While RBOH D is responsible for most of the ROS production during effector-triggered immunity, RBOH F is thought to control cell death (Torres et al., 2002). Moreover, a different expression pattern of these two NADPH oxidases has been demonstrated, with RBOH F being mainly expressed in leaves. Nevertheless, striking differences between their functions are evident in the literature depending on the pathosystem studied and even on the inoculation method, plant growth conditions or sampling time (Morales et al., 2016). Moreover, NADPH oxidases are not the only sources of hydrogen peroxide during defense responses. Polyamine oxidases and cell wall peroxidases also contribute (Kámán-Tóth et al., 2019; Jasso-Robles et al., 2020).

In addition to H₂O₂ accumulation, the treatment of plants with AGPs caused changes in defense genes transcription. The results suggest that increased resistance of B. napus against L. maculans elicited by AGPs is SA-dependent as indicated by elevated transcription of both SA-biosynthetic gene ICS1 and SA-responsive gene PR1. Since the expression of PAL did not differ from the control, it is probable that SA is synthesized exclusively via the pathway regulated by ICS1. In addition to the salicylic acid signaling pathway, the AGPs induced also expression of βCHI gene involved in JA/ET signaling. On the other hand, the activation of AOS transcription, the biosynthetic gene for jasmonic acid, by AGPs, was not observed. This resembles the signaling situation reported by Šašek et al. (2012b) within B. napus infection with L. maculans who showed, that the main signaling pathways involved in this pathosystem are SA and ET signaling and that the transcription of the related genes was significantly increased 7 days after pathogen recognition. This is further supported by the strong transcription of ACS2, the ethylene-biosynthetic gene, elicited by AGPs treatment. Our findings indicate elicitation of both SA-dependent and ET-dependent signaling pathways. Nevertheless, the ever-increasing discoveries of crosstalks in between the signaling pathways revealed the truly complex nature of plant responses (Bürger and Chory, 2019; Yang et al., 2019), even in the plant B. napus (Nováková et al., 2014). Besides our results correspond to the results of Cluzet et al. (2004), who described the increased expression of CHI, PR1, and PR10 genes after treatments with various Ulva extracts. The activation of PR1 transcription was also reported in B. napus cotyledons after treatment with oligosaccharide elicitor isolated from L. maculans mycelium (Kim et al., 2013) and protein elicitor isolated from L. maculans cultivation medium (Nováková et al., 2016). The high production of hydrogen peroxide together with a strong induction of SA and ET signaling pathways explains significant inhibitory effects of AGPs on the infection development.

Surprisingly, ulvan caused almost no significant changes in gene expression except for slightly increased levels of βCHI and VSP, consistent with its known mode of action (Jaulneau et al., 2010; Hernández-Herrera et al., 2016; Ramkissoon et al., 2017). Based on the results, the U. lactuca ulvan appeared to be efficient in the pathosystems studied at higher concentrations, which was further supported by an 80% decrease in L. maculans infection after treatment with a concentration of 10 mg·ml⁻¹ (Supplementary Figure 3). The variability in ulvan data may be due to the viscous nature of the concentrated samples, which causes uneven infiltration.

In addition to the major defense signaling pathways regulated by SA, JA, and ET, B. napus plants responded to elicitor treatment with a decrease in genes related to ABA. Both the NCED3 biosynthetic gene and the RD26 responsive gene were downregulated by both the AGPs and ulvan. This result is in accordance with previous findings of Jaulneau et al. (2010), who reported a transient decrease (2 days after treatment) in ABA-responsive genes in Medicago truncatula. On the other hand, Chen et al. (2013) found an increase in ABA in plants treated with a protein elicitor from oomycete Phytophthora boehmeriae, which lead to significant reduction of pathogen infection. These contrasting results indicate that the role of ABA in induced resistance by elicitors is not as straightforward as, e.g., the role of SA. The role of ABA in plant defense against pathogens is less defined and the data are less consistent compared with SA, JA, and ET signaling. Since ABA regulates stomata opening, it is suggested that ABA is an important phytohormone in protecting the host plant from pathogen penetration via the stomata. The possible positive role of ABA in the studied pathosystem B. napus—L. maculans was previously reported by Šašek et al. (2012b). Similar to ABA, several elicitors of different origins induced stomata closure and ROS production in guard cells (Allègre et al., 2009).

In conclusion, our study makes an important contribution to the understanding of the mechanisms behind the elicitor activity of U. lactuca extracts recently introduced in agriculture. In addition to the well-described polysaccharide ulvan, U. lactuca contains other compounds that elicit even stronger defenses against pathogens. We were able to prepare an AGP-like enriched fraction that efficiently induced resistance to the hemibiotrophic fungal pathogen L. maculans in cotyledons of B. napus. Examination of the signaling events revealed that the triggered defense mechanisms were regulated by H₂O₂, SA, and ET signaling. Proposed mechanisms of actions for both AGPs and ulvan are presented in the Figure 6. Since AGPs showed higher efficiency than ulvan, AGPs may have the potential to become a component of plant protection products in the future. Moreover, for their possible future application, our following research will be focused on increased penetration of AGPs to the plants, testing oligosaccharides produced from AGPs, and assessing their effect also on other pathosystems including monocot plants, which differ in their defense signaling.

The original contributions presented in the study are included in the article/Supplementary Material; further inquiries can be directed to the corresponding author.

TP contributed to data collection, designing research, and analyzing the data and writing the manuscript. BJ contributed to data collection, designing plant experiments, analyzing the data, and editing the manuscript. SH contributed to data collection (saccharide analysis). J-CY and VF contributed to designing research. LB contributed to designing research and editing the manuscript. PL contributed to data collection, designing research, and editing the manuscript. All authors contributed to the article and approved the submitted version.

The work was supported from European Regional Development Fund-Project “Centre for Experimental Plant Biology” (no. CZ.02.1.01/0.0/0.0/16_019/0000738) and by the grant of Specific University Research (A1_FPBT_2020_001).

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.