Priming Treatments with Biostimulants to Cope the Short-Term Heat Stress Response: A Transcriptomic Profile Evaluation

Plant stress induced by high temperature is a problem in wide areas of different regions in the world. The trend of global warming is going to enhance the effects of heat stress on crops in many cultivation areas. Heat stress impairs the stability of cell membranes and many biological processes involving both primary and secondary metabolism. Biostimulants are innovative agronomical tools that can be used as a strategy to counteract the detrimental effect of abiotic stresses, including heat stress. In this work, two biostimulants based on Ascophyllum nodosum extracts (named Phylgreen) and based on animal L-α amino acids (named Delfan Plus) were applied as priming treatments to Arabidopsis thaliana plants subjected to heat stress exposure. Plants at the vegetative stage were treated with biostimulants 12 h before high temperature exposure, which consisted of maintaining the plants at 37 ± 1 °C for 4 h. Transcriptional profiles, physiological, and biochemical analyses were performed to understand the mode of action of the biostimulants in protecting the plants exposed to short-term heat stress. At a physiological level, chlorophyll, chlorophyll a fluorescence, phenolic index, total anthocyanins, reactive oxygen species (ROS) were measured, and significant variations were observed immediately after stress. Both biostimulants were able to reduce the oxidative damage in leaves and cell membrane. Transcriptomic data revealed that upregulated genes were 626 in Phylgreen and 365 in Delfan Plus, while downregulated genes were 295 in Phylgreen and 312 in Delfan Plus. Bioinformatic analysis showed that the biostimulants protected the plants from heat stress by activating specific heat shock proteins (HPS), antioxidant systems, and ROS scavengers. The results revealed that the biostimulants effectively induced the activation of heat stress-associated genes belonging to different transcription factors and HSP families. Among the heat shock proteins, the most important was the AtHSP17 family and in particular, those influenced by treatments were AtHPS17.4 and AtHPS17.6A, B, showing the most relevant changes.


Introduction
Heat stress is well known as one of the main abiotic stresses that affect the performance of cultivated crops, and its potential severity is going to be increased by global climate change [1]. Heat stress due to high temperature can negatively affect plant growth, development, and, more severely, the reproductive stages causing a decrease of crop yield [2]. The exposure of crops to heat stress is not always constant and its intensity generally was reported in plants subjected to drought, flooding, high light intensity, chilling, heat, salinity, air pollutants, and herbicides. Enhanced ROS generation has also been found in heat stressed plants with an increase of membrane lipid peroxidation [11]. H 2 O 2 is a stable ROS that has a pivotal role in response and defence in plants. The H 2 O 2 is the regulator of physiological and biochemical processes in plants under biotic and abiotic stresses [12]. At high concentrations, it can be toxic, while in low concentrations, it activates specific defence mechanisms in plants, allowing them to overcome stressful conditions [13].
Heat stress is a common abiotic stress during summer in Mediterranean areas and the severity of the stress is increased if associated with low water availability. Prediction of these events is not easy with the available forecast models. Thus, farmers need additional practical tools to alleviate crops from abiotic stresses [14][15][16]. To date, several biostimulant products are available on the market and they can be used as a good agronomic tool to prepare plants for incoming critical periods (priming) or to counteract the negative effects and help plants to recover after stress [15,17]. Priming is an agronomic strategy used in agriculture for inducing the seeds or plants to activate specific physiological or biochemical processes for improving germination or plant performance against specific events (i.e., germination for seeds, counteract abiotic stresses for plants, etc.). Biostimulants can act by increasing the capacity of the crops to face a stressful condition and avoid the negative effects on their performance. This priming effect is due to the activation of a complex mechanism of molecular interactions [18] that is triggered by the active components of the biostimulant [19] (Fleming et al., 2018). Previous studies have shown that Ascophyllum nodosum seaweed extracts can act as biostimulants maintaining crop productivity during and after exposure to abiotic stresses, by protecting the plant cells from the stress-induced oxidative damage [20]. However, it has been shown that, starting from the same raw material, the success of a seaweed-based biostimulant depends on the manufacturing procedures [21]. Biostimulants containing amino acids can also be used as a priming treatment for the activation of different physiological processes. The application of amino acid-based biostimulants (Delfan Plus) applied as a priming treatment was able to increase drought tolerance, mitigating stress, in the Arabidopsis thaliana model [19]. Understanding the mechanism of action of biostimulants and the plant response to their application in combination with abiotic stress conditions is crucial for the optimization and use of these novel agronomic tools [22].
The aim of the work was to evaluate the priming effect of the biostimulant products applied to Arabidopsis thaliana plants subjected to a short-term heat stress (37 • C). This was achieved by the identification of the transcriptional changes in stressed plants previously treated with biostimulants. The physiological, biochemical, and molecular approaches provided useful information for understanding the mode of action of the products when applied prior to the stress (priming/preventive effect). The hypothesis of this work was that biostimulants containing Ascophyllum nodosum extracts or containing amino acids could be applied as protective treatments against heat stress in plants. Biochemical and molecular analyses have been used to identify the activation of the specific clusters of genes that are responsible for the increase in plant tolerance and recovery in plant performance after stress exposure.

Transcriptional Changes
The RNA-seq analysis enabled identification of a total of 32,833 sequences corresponding to gene sequences in the Arabidopsis thaliana reference genome. Based on the count of each read, the expression of each gene was estimated as Log FC, compared to non-stressed control (CTR). The Venn diagram has been used for identifying the differentially expressed genes activated or repressed by the treatments. A total of 1756 DEGs were expressed under heat stress, while 1228 DEGs were found in response to Phylgreen application, and 844 Plants 2022, 11, 1130 4 of 20 in Delfan Plus application. Biostimulant treatments shared 1620 DEGs and only 174 were commonly expressed in the biostimulants and stress conditions (Table 1). The DEGs with a fold change (FC) > 2 and up-or downregulated genes are shown in the Venn diagram ( Figure 1A,B). Upregulated DEGs were 611 genes in CTRS, 626 genes in Phylgreen, and 365 in Delfan Plus ( Figure 1A), 14 were DEGs in all treatments compared to CTR. Downregulated DEGs were 229 in CTRS, 295 in Phylgreen, 312 in Delfan Plus, only one gene was common in all treatments ( Figure 1B). expressed under heat stress, while 1228 DEGs were found in response to Phylgreen application, and 844 in Delfan Plus application. Biostimulant treatments shared 1620 DEGs and only 174 were commonly expressed in the biostimulants and stress conditions ( Table 1).
The DEGs with a fold change (FC) > 2 and up-or downregulated genes are shown in the Venn diagram ( Figure 1A,B). Upregulated DEGs were 611 genes in CTRS, 626 genes in Phylgreen, and 365 in Delfan Plus ( Figure 1A), 14 were DEGs in all treatments compared to CTR. Downregulated DEGs were 229 in CTRS, 295 in Phylgreen, 312 in Delfan Plus, only one gene was common in all treatments ( Figure 1B).
The two biostimulants had different effects on transcription activation. Phylgreen and Delfan Plus showed a higher number of upregulated DEGs than downregulated. The total number of DEGs was higher in Phylgreen compared to the stressed control and Delfan Plus (Figure 2A, Supplementary Dataset S1).

Expression of Transcription Factors
Transcription factors are important regulators of gene expression, and their regulatory function is extremely important under stress conditions. Phylgreen and Delfan Plus

Expression of Transcription Factors
Transcription factors are important regulators of gene expression, and their regulatory function is extremely important under stress conditions. Phylgreen and Delfan Plus showed changes in the number of transcription factors that were found differentially expressed compared to the control. In plants treated with Delfan Plus the upregulated genes were 71, while 36 were downregulated ( Figure 2B). Only one gene, the at2g38250 that encodes for Homeodomain-like superfamily protein was in common among the treatments. Phylgreen and Delfan Plus shared 35 differentially expressed transcription factors. The transcription factors mostly affected by Phylgreen and Delfan Plus (Log FC > 5) belonged to the families MYB, bHLH, and DREB which are usually involved in abiotic stress plant responses. The AtbHLH and AtMYB1 genes regulate the activation of the shikimate pathway and were highly expressed in response to both biostimulants (Supplementary Dataset S2). The analysis of transcription activation between CTRS vs. Phylgreen showed that the ZINC FINGER PROTEIN 6 (AtZEP2, AtZFP6), LITTLE ZIPPER 2 (AtZPR2), zinc finger, AP2 domaincontaining transcription factor, and WRKY74 were upregulated in the Phylgreen compared to CTRS (Supplementary Dataset S1). The transcription activation in Delfan Plus compared to CTRS revealed that the transcription factors activated in plants under stress treated with Delfan Plus were AtWRKY72, WRKY31, ALCATRAZ (at5g67110), and CHP-rich zinc finger protein putative (Supplementary Data S1).

Heat Stress Response-Associated Genes
The analysis of genes specifically associated to heat stress response revealed a significant increment in the number of genes activated by all the treatments (stressed control and biostimulants) compared to untreated plants. In general, the biostimulants induced the overexpression of genes rather than the downregulation, as shown in Figure 2C.
The expression pattern observed in all treatments was similar and markedly different compared to the one of CTR. However, specific responses were induced by each of the treatments. All treatments induced or inhibited the expression (Log FC > 2 or < −2) of a few genes in a specific manner. Most of these genes encoded for heat shock proteins (HSP), which could be further used as marker for studying the specific mode of action of each biostimulant. In the stress-control (CTRS), the highest expressed genes were AtHSP17.6, HSP binding protein (at4g24190), and AtHSP90.7. The higher gene expression belonging to associated heat stress genes was found in the HPS17 families (AtHPS17.4, AtHPS17.6A, B), between Phylgreen and Delfan Plus treatments. Downregulated genes in the biostimulants treatments were AtBIP3 and heat shock-related protein (At2g03020).

Secondary Metabolism-Phenylpropanoid Pathway
The DEGs encoding for secondary metabolism were analyzed, and differences were observed in the phenylpropanoid pathway. Delfan Plus treatment based on RNAseq data after 16 h showed a downregulation of this biosynthetic route, while no difference was observed as a response to Phylgreen application ( Figure 2D). In CTRS, the AtACOS5 (Acyl-COA Synthetase 5) and 4-coumarate-CoA ligase genes were upregulated, while laccase/diphenol oxidase and transferase family protein were downregulated. In Phylgreen treated plants, AtCCR2, (cinnamoyl-CoA reductase), O-methyltransferase family 2 protein, and 4-coumarate-CoA ligase family protein were upregulated. No downregulated genes with FC < −2 were observed (Supplementary Dataset S2).
A mitochondrial AtHSP23-5 was upregulated in both biostimulants ranging from 3.4 to 4.3 FC. This gene could play an important role in heat stress tolerance.

DAVID Enrichment Analysis
The enrichment analysis was performed to identify classes of genes that were overrepresented in each treatment and an association with a specific biostimulant treatment could be highlighted. The results obtained from the enrichment analyses are shown as functional annotation clustering and gene function classification (Supplementary Dataset S3).
In CTRS, the enrichment analysis showed a fold enrichment > 1.0 15 and the top ten categories highlighted in CTRS treatments were SAMRT, INTERPRO, GOTERM, KEGG_PATHWAY, and UP_SEQ_FEATURE ( Table 2). The most significant terms enriched were the knottin, a scorpion toxin-like, palmitoyl protein thioesterase in the INTERPRO category. The thiolester and palmitoyl hydrolase activity were the most enriched terms in the GOTERM category. In the Phylgreen treatment the top ten categories were the same as CTRS except for the KEGG_PATHWAY and the most enriched terms were related to wounding response, following from AP2/ERF DNA binding protein. Delfan Plus treatment showed the GOTERM, UP_SEQ_FEATURE, and UP_Keywords as the most enriched categories, which include abiotic defence response and oxidoreductase or oxidation-reduction process.  The DAVID functional annotation clustering (FAC) considering an Enrichment Score > 2 identified one cluster for CTRS, 11 for Phylgreen (Supplementary Data S3) and two for Delfan Plus. The FAC analysis reported for the CTRS terms mainly belonged to fatty acid metabolism ( Table 3). The two highest enriched clusters identified in the Phylgreen treatment included transcriptional regulation terms and ethylene signalling pathways ( Table 3). The first class with an enrichment score of 5.5 included ROS, heat, and high light responsive genes. The second enrichment score of 4.5 included the WRKY transcription factor in the following terms DNA-binding region: WRKY, SM00774:WRKY, and IPR003657:DNAbinding WRKY. The Phylgreen showed the highest percentage of genes annotated in cluster 2 and with percentages ranging from 5 to 11.6% (Table 3). Table 3. DAVID functional analysis: functional annotation cluster (FAC) of differentially expressed genes in CTRS, Phylgreen, and Delfan Plus (Enrichment Score >2) recognized in DAVID database. Functional category, terms, p value, fold enrichment, and statistical significance (Bonferroni, Benjamini, and FDR). In the table, two annotation clusters (if available) with highest enrichment score for each treatment.

Histochemical Analysis
For all the samples, the presence of secondary metabolites of the phenylpropanoid pathway (mainly lignin) and ROS (H 2 O 2 ) by histochemical analysis on mature leaves, in non-stressed control (CTR), stressed control CTRS, and treated plants were analyzed. ROS were visualized by 3,3'-diaminobenzidine (DAB) staining. DAB is oxidized in the presence of peroxidase and hydrogen peroxide resulting in the deposition of a brown, alcohol-insoluble precipitate at the site of enzymatic activity (Figure 3). The CTR did not show any brown spots. The stressed control CTRS showed several brown spots along the leaf margins and spread onto the leaf blade, indicating the accumulation of ROS in different areas of the leaves. Phylgreen and Delfan Plus application strongly reduced the ROS accumulation and only a few small spots were visible, demonstrating the positive effects of these treatments against heat stress (Figure 4). In particular, leaves treated with Phylgreen (A) and Delfan Plus (B) only showed small patches of DAB signal along the leaf margins or on the leaf blade. It is interesting to note that on the leaf areas reached by the drops of the products Phylgreen and Delfan Plus used to treat the plants, a faint but clear DAB signal was present. Damaged cells induced by heat stress could be visualized using autofluorescence. The intensity of autofluorescence depends on the cell wall composition and construction. Changes in autofluorescence can be used only as a coarse indicator of cell wall products. High autofluorescence signals were found in CTRS, indicating the heat stress damage. In contrast, no damage was found in control plants, while almost no autofluorescence signals were found in two samples treated with Delfan Plus (B) and only a few spots were found in leaves treated with Phylgreen (A) (Figure 4).

Chlorophyll and Chlorophyll a Fluorescence
The effect of treatments was evaluated by the determination of non-destructive chlorophyll content and chlorophyll a fluorescence measurements. Immediately 4 h after stress (HAS), the highest chlorophyll content was found in Phylgreen treatment, 7.6 arbitrary units (a.u.), while CTR showed a value of 5.8 a.u. The CTRS showed intermediate values, while CTR and Delfan Plus showed similar chlorophyll content. Two days after stress (DAS), the plants did not show significant variations among treatments. After 6 days of recovery, an increase of chlorophyll was observed in the CTRS and treated plants ( Figure 6). Delfan Plus at the beginning of the experiments had lower chlorophyll but during recovery, it increased, reaching values similar to control and Phylgreen. The chlorophyll a fluorescence analysis, enables a non-destructive estimation of leaf functionality that can provide information on the health or stress conditions of plants. Statistical analysis showed that Fv/Fm, PI, and DIo/RC were not statistically different, while a difference was found in the time. The Fv/Fm ratio represents the maximum quantum efficiency of photosystem II and is an indicator of plant stress. The Fv/Fm ratio did not change among treatments, indicating that maximum quantum efficiency of PSII was not affected by the treatments ( Figure 6B). The performance index (PI) represents the measurement of the leaf functionality of the plants. Phylgreen after 4 h heat stress exposure showed a value of 2.4 a.u. similar to CTR, indicating that the stress did not affect leaf functionality ( Figure 6C). During the recovery period, the PI values ranged from 1.8 to 2.6 a.u. (Figure 6C). The dissipation energy per reaction centre (DIo/RC) indicates the energy lost as heat under stressful conditions. No differences among treatments were found ( Figure 6D).

Chlorophyll and Chlorophyll a Fluorescence
The effect of treatments was evaluated by the determination of non-destructive chlorophyll content and chlorophyll a fluorescence measurements. Immediately 4 h after stress (HAS), the highest chlorophyll content was found in Phylgreen treatment, 7.6 arbitrary units (a.u.), while CTR showed a value of 5.8 a.u. The CTRS showed intermediate values, while CTR and Delfan Plus showed similar chlorophyll content. Two days after stress (DAS), the plants did not show significant variations among treatments. After 6 days of recovery, an increase of chlorophyll was observed in the CTRS and treated plants ( Figure 6). Delfan Plus at the beginning of the experiments had lower chlorophyll but during recovery, it increased, reaching values similar to control and Phylgreen. The chlorophyll a fluorescence analysis, enables a non-destructive estimation of leaf functionality that can provide information on the health or stress conditions of plants. Statistical analysis showed that Fv/Fm, PI, and DIo/RC were not statistically different, while a difference was found in the time. The Fv/Fm ratio represents the maximum quantum efficiency of photosystem II and is an indicator of plant stress. The Fv/Fm ratio did not change among treatments, indicating that maximum quantum efficiency of PSII was not affected by the treatments ( Figure 6B). The performance index (PI) represents the measurement of the leaf functionality of the plants. Phylgreen after 4 h heat stress exposure showed a value of 2.4 a.u. similar to CTR, indicating that the stress did not affect leaf functionality ( Figure 6C). During the recovery period, the PI values ranged from 1.8 to 2.6 a.u. (Figure 6C). The dissipation energy per reaction centre (DIo/RC) indicates the energy lost as heat under stressful conditions. No differences among treatments were found ( Figure 6D).

Total Phenolics and Anthocyanins
Total phenols, expressed as the phenolic index, showed similar trends of anthocyanins for all the treatments. Data subjected to two-way ANOVA revealed that no significant differences were found. The phenolic index, after 2 days of recovery, ranged from 19.8 ABS 320nm /g FW to 17.7 ABS 320nm /g FW ( Figure 7A).
Statistical analysis of anthocyanins showed that there were no significant differences among treatments. Anthocyanin concentrations ranged from 21.5 to 24 mg/100 g ( Figure 7B). Statistical analysis was performed using two-way ANOVA and differences amongst means were determined using LSD test. Asterisks indicate significant differences, * for p < 0.05, ** for p < 0.01. In the A and D, significant differences for each timepoint have been highlighted using different letters.

Total Phenolics and Anthocyanins
Total phenols, expressed as the phenolic index, showed similar trends of anthocyanins for all the treatments. Data subjected to two-way ANOVA revealed that no significant differences were found. The phenolic index, after 2 days of recovery, ranged from 19.8 ABS320nm/g FW to 17.7 ABS320nm/g FW ( Figure 7A).
Statistical analysis of anthocyanins showed that there were no significant differences among treatments. Anthocyanin concentrations ranged from 21.5 to 24 mg/100 g ( Figure  7B). . Data are means with standard errors (n = 5). Statistical analysis was performed using two-way ANOVA and differences amongst means were determined using LSD test. Asterisks indicate significant differences, * for p < 0.05, ** for p < 0.01. In the A and D, significant differences for each timepoint have been highlighted using different letters.

Discussion
Biostimulants represent innovative tools to be used for protecting plants from abiotic stresses [15,22,23]. Several studies have shown that biostimulants can play a key role in helping crops against abiotic stresses, including heat stress [15]. A biostimulant containing a mixture of sugarcane molasses with yeast extract has been applied to tomatoes grown in high temperature (32 °C). The treated plants showed higher ascorbic acid in fruits and leaves. Positive effects were also found on yield [24]. The effect of treatments can be as-

Discussion
Biostimulants represent innovative tools to be used for protecting plants from abiotic stresses [15,22,23]. Several studies have shown that biostimulants can play a key role in helping crops against abiotic stresses, including heat stress [15]. A biostimulant containing a mixture of sugarcane molasses with yeast extract has been applied to tomatoes grown in high temperature (32 • C). The treated plants showed higher ascorbic acid in fruits and leaves. Positive effects were also found on yield [24]. The effect of treatments can be ascribed to the increase of the antioxidant compounds in plants that reduce the ROS accumulation. It is well known that ROS increase under heat stress and induce cell membrane damage with lipid peroxidation. Membrane destabilization, at leaf level, can reduce the electron transfer and the photosynthetic activity. At the physiological level, high temperatures can induce heat stress, while the severity of damage depends on crop sensitivity and exposure period. Heat stress is usually associated with a reduction in gas exchange and photosynthesis activity due to stomatal closure. The heat stress damage can affect the repair activity of photosystem II [25]. Chlorophyll a fluorescence measurement enables the non-destructive evaluation of plant health status and can provide useful information on the effect of treatments on heat stress tolerance induction. The Fv/Fm ratio and Fo have been successfully used as parameters for screening the tolerance of tropical fruit crops to heat stress [26]. Phylgreen treatment was the most active in increasing chlorophyll content under heat stress. This result can be explained by the effect of Ascophyllum nodosum extract that increases chlorophyll concentration and promotes growth [26]. After the recovery period, both biostimulants showed the same leaf functionality as non-stressed plants. Another mechanism used by plants to protect themselves from heat stress is the enhanced production of secondary metabolites and among them phenylpropanoids play a central role in plant response to heat stress [27][28][29]. Their antioxidant capacity is responsible for the neutralization of ROS generated by heat shock, thus preventing oxidative stress and cell damage [30]. The presence of secondary metabolites and ROS in controls and treated plants was also analyzed through histochemical analysis. Microscopic image analysis showed damage on the leaf blade in CTRS plants while Phylgreen and Delfan Plus treated plants avoided H 2 O 2 accumulation, demonstrating the positive effect of biostimulants to mitigate heat stress. The Phylgreen is an Ascophyllum nodosum extract and the antioxidant activities demonstrated in plants and in vitro can explain the reduced electrolyte leakage [31]. In CTR leaves, microscopic analysis failed to detect any signal of ROS and secondary metabolites. The values of phenolic post-stress index demonstrated that plants were able to cope with heat stress and also enhanced the leafy functionality as observed by chlorophyll a fluorescence.
Considering the analysis of cell dimension, the heat stress did not influence this parameter. In fact, the analysis of stomata dimensions showed no significant differences amongst the different samples and the slight differences observed in mesophyll cell dimension even if statistically significant, are more probably due to differences in dimension and developmental stage of the leaves analyzed than to the treatments. Considering all the histochemical analyses together, it could be observed that the Phylgreen and Delfan Plus treatments were shown to exert a positive effect on plant protection by reducing the consequences of thermal stress. The biostimulants induced a strong induction of differentially expressed transcription factors (TFs) compared to CTRS. Amongst the TFs associated with the heat responses, AtZat10 and AtZat12 were upregulated in both biostimulants with 2.4-2.7 and 2.7-3.0-fold change, respectively. The AtZat 10 is an HPS sensor of hydrogen peroxide and AtZat12 is a heat response transcription factor [32], which suggests that the biostimulant treatment induced the activation of the transcriptional machinery involved in stress protection. Both biostimulants also induced the AtHsfs, which is associated with the heat response regulation of plants. Since these TFs were not expressed in the CTRS, this could be the strategy of the plants to enhance the activation of specific genes in heat defence. The common expression of ATHsfs3 in both biostimulants has been found to induce thermotolerance in Arabidopsis [33]. AtHSF3 can be considered as the key regulator of the immediate stress-induced activation of heat stress gene transcription [34]. These high gene regulation inductions have also been highlighted by the DAVID results which revealed the categories affected by the different treatments. DAVID bioinformatics tools can provide information on over-representation of the GO category terms [35] and help in understanding the mode of action of the biostimulants. In fact, most of the GO terms evidenced by the DAVID analysis were those related to the transcriptional regulation, abiotic stress defence and secondary metabolism. At the transcriptional level, among the heat stress-associated genes, the AtHSP17.6A was upregulated in all treatments. This gene was the highest expressed in CTRS and literature confirmed that it rapidly increases after heat stress, while it is not detectable in vegetative tissues in the absence of stress [36]. The activation of AtHSP17.6A is the plant strategy to counteract heat stress. Biostimulants were able to strongly induce a higher number of HSPs and it could indicate the mode of action of these products to enhance abiotic or heat stress tolerance. Phylgreen and Delfan Plus showed similar results in the activation of the heat shock proteins (HSP) DEGs. The most expressed DEG was AtHSP17.4 in both biostimulants and this gene encodes for the small HPS (sHPS) protein localized in the cytosol [37]. These HPS are mainly expressed during development, embryo maturation, and germination under normal conditions. In heat stress conditions, the expression of AtHSP17.4 has been found in leaves and seems to have a protection function against the stress [36]. However, the expression of AtHSP17.4 is associated with the activation of ABI3 [38,39]. In the current data presented, the expression of AtABI3 was induced but at values from 0.5 to 0.8. Other highly expressed HPS genes were AtHSP17.6 CII, AtHSP17.8 CI, AtHSP18.2, and AtHSP70 [40]. All these genes have been associated with different abiotic stresses. The AtHSP17.6 has been induced by abscisic acid or osmotic stresses [37,41]. The expression of AtHSP17.8 is associated with drought and salt stress. Overexpression of AtHSP17.8 in Arabidopsis and lettuce increased tolerance against drought and high salinity stresses [42]. The expression of this gene seems to induce hypersensitivity to ABA with an improved crop sensitivity to water reduction.
In this work the AtHSP23-5 was upregulated in both biostimulants suggesting an important role in heat stress defence. This gene is located in the nuclear genome and studied from evolution in Brassicaceae [43], while a mitochondrial HPS22 has been found in Drosophila melanogaster under heat stress conditions [44].

RNA Isolation
RNA sampling was immediately performed at the end of heat stress treatment, after 4 h of heat stress and 16 h after biostimulant application. Total RNA was extracted from 100 mg of leaf powder using the Spectrum Plant Total RNA Kit (Sigma-Aldrich, Milan, Italy). RNA purity and integrity were assessed with an Agilent 2100 bioanalyzer-RNA 6000 Nano Chip (Agilent Technologies, Santa Clara, CA, USA) and quantified with a Nanodrop 8000 (Thermo Scientific, Waltham, MA, USA). RNA samples with A260/A280 ≥ 1.9 and RNA integrity number (RIN) ≥ 7 were used for the RNA sequencing and further library preparation (libraries corresponding to a biostimulant product and to stressed and unstressed controls).

RNA-Seq and Library Preparation
Illumina sequencing and de novo assembly were performed at BMR Genomics Labs (Padua, Italy). The de novo assembly was performed to verify if the biostimulants could induce the activation of new specific genes. Random primed cDNA libraries were prepared with a TruSeq RNA Sample Prep kit (Illumina, San Diego, CA, USA) and sequenced in two paired-end modes in one run on the Illumina HiSeq-2000 Platform, generating up to 60M reads of 100 bp per sample. Alignment of reads to the reference genome of Arabidopsis (Arabidopsis_thaliana TAIR10 v42) was performed using HISAT 2.1.0 tool [45,46]), while gene association was performed with feature Counts v 1.6.0 (DESeq2 package version: 1.34.0 DESeq2 package version: 1.34.0). Specific libraries were prepared for unstressed untreated control plants (CTR) and for stressed untreated control plants (CTRS) and for biostimulants (applied on stressed plants). RNAseq data have been stored and available in the NCBI database (Accession n. PRJNA777374). The validation of RNAseq data has been performed by qPCR gene expression analysis on the following up-and downregulated genes at2g29500, at5g12030, at5g12020, at1g09080, at3g46230, at1g72070, at5g59720, at5g37750, at5g37440, and at3g14200 (Supplementary Table S1) RPKM Log2 fold changes versus qPCR 2 −∆∆CT log2 fold changes were compared as reported by Li [47] and correlation coefficient was r = 0.58.

Bioinformatic Analysis
Gene annotation was performed using various platforms (including Blast2GO and Panther Tair tools). Genes differentially expressed were identified in each treatment and calculated as follows (Log2FC > 2, calculated as follow LOG ([A]; 2)-LOG ([B]; 2), A gene expression in the treatment vs. B gene expressed in the unstressed control) were selected and Venn diagrams were prepared to individuate common and unique responses in all the libraries. MapMan tool was used to display transcriptomic datasets onto diagrams of metabolic pathways and biological processes. This analysis allowed the identification of the most affected biochemical or signalling transduction pathways and the identification of the genes differentially expressed in the most interesting pathways for each treatment versus control unstressed/untreated. Enrichment of pathways, gene functions, and organelle associations based on Gene Ontology (GO) and other functional annotation data were identified using DAVID (http://david.abcc.ncifcrf.gov/, accessed 3 March 2021) [35]. The DAVID bioinformatics tool was also used to examine the biological significance of the transcriptome changes in the different treatments. Medium stringency was applied for the analyses. DAVID analysis identifies significantly enriched biological themes by examining for enrichment in over 40 different publicly available annotation categories (Supplementary  Table S2), analyzing up-and downregulated sets separately. Significance was determined using a modified Fisher's exact statistic (EASE score), and significantly enriched biological themes were identified as clusters of annotated terms and KEGG_PATHWAYs (https:// www.genome.jp/kegg/pathway.html, accessed on 20 January 2022). A cluster enrichment score of 1.3 for an annotation cluster is equivalent to non-log scale 0.05, and therefore scores of 1.3 or greater are considered enriched [35] Fold-enrichment scores were also used to indicate the magnitude of enrichment for individual terms and KEGG_PATHWAYs, and fold-enrichment scores greater than 2 are suggestive of an informative change [35,[46][47][48].

Non-Destructive Determination: Chlorophyll and Chlorophyll a Fluorescence
Non-destructive analysis was carried out in vivo on leaf tissue for the leaf functionality evaluation of Arabidopsis thaliana plants after treatments and stress applications (12 h of priming and 4 h after stress-HAS, 2 days after stress-DAS, and 6 days after stress). The chlorophyll content was colorimetrically measured in vivo using a non-destructive instrument (CL-01, Hansatech, UK). Chlorophyll a fluorescence was measured using a portable fluorometer (Handy PEA, Hansatech, Kings Lynn, UK). Leaves were dark-adapted using leaf clips; after 30 min, a rapid pulse of high-intensity light of 3000 µmol m −2 s −1 (600 W m −2 ) was applied to the leaf inducing fluorescence. Fluorescence parameters were calculated automatically by the used device, such as Fv/Fm, the variable fluorescence to maximum fluorescence. Starting from these parameters, OJIP test analyses [49] were performed to determine the following index: performance index (PI); dissipation of energy per cross-section (DIo/RC) and density of reaction centre at P stage (RC/CSm).

Phenolic Index and Anthocyanins Content
These parameters were determined two and six days after stress application. The phenolic index and the total anthocyanin content in leaf tissue were determined using spectrophotometric methods (Spectrophotometer, Thermo, Italy). Phenolic index was determined by a direct measurement of the leaf extract absorbance at 320 nm. About 20-30 mg of fresh leaf tissue (disk of 5 mm diameter) was weighed and 3 mL methanolic HCl (1%) were added. After overnight incubation, the supernatant was read at 320 nm. The values were expressed as mg/100 g fresh weight (FW). Anthocyanin concentrations were determined on samples of 20-30 mg of fresh leaf (disks of 5 mm diameter) extract using 3 mL of methanolic HCl (1%). Samples were incubated overnight at 4 • C in darkness. The concentration of cyanidin-3-glucoside equivalents was determined spectrophotometrically at 535 nm using an extinction coefficient (ε) of 29,600. Sampling was performed taking four biological replicates for each treatment.

Histochemical Analysis
Samples for the histochemical analysis were collected 4 h after heat stress. Arabidopsis leaves were detained in ethanol 100% and then observed using a Zeiss Axiophot D1 microscope, using ultraviolet epifluorescence (excitation filter 365 nm; dichroic mirror 395 nm, barrier filter 420 nm), equipped with an AxioCam MRc1 digital camera. Hydrogen peroxide (H 2 O 2 ) detection was performed as previously reported [50]. Arabidopsis leaves were submerged in 3,3 -Diaminobenzidine (DAB) solution (1 mg/mL DAB, pH 3.8) and left overnight in the dark. The leaves were then rinsed in water, transferred to tubes containing ethanol 96% that were left in boiling water until chlorophyll was completely removed. The leaves were observed using a Zeiss Axiophot D1 microscope, equipped with an AxioCam MRc1 digital camera [51]. Cell death was quantified as electrolyte leakage measurement, as reported by Roberts et al. [52]. Briefly, for each leaf analyzed, 4 leaf discs (diameter: 8 mm) were floated in water for 30 min, then transferred to tubes containing 4 mL distilled water. The conductivity of the solution was determined with an Orion Conductivity Meter (µSiemens/cm). For each histochemical analysis, three biological replicates were used obtaining similar results. To determine mesophyll and stomata cell size, mature leaves were collected and treated with a clearing solution (160 g chloral hydrate, 20 mL glycerol in 60 mL water). Cleared leaves were mounted on slides, and interference contrast images were taken using a Zeiss IMAGE R.D1 microscope equipped with an AxioCam MRc1 digital camera. For each treatment, 3 plants were analyzed, from each plant one leaf was cleared and at least 50 mesophyll/stomata cells were measured.

Statistical Analysis
Analytical data were subjected to two-way ANOVA performed by GraphPad Prism 6.0 (GraphPad Software, San Diego, CA, USA). Significant differences amongst means were determined using LSD multiple comparisons test. Specific details are also reported in figure legends.

Conclusions
In conclusion, results revealed that biostimulants effectively induced activation of heat stress-associated genes belonging to different transcription factors and HSP families. The specific cluster of genes confirmed the physiological and biochemical data observed, demonstrating that the biostimulants were able to reduce the heat damage in Arabidopsis plants by activating antioxidant systems and heat repair systems. In general, both treatments with Phylgreen and Delfan Plus treatments had a positive effect in counteracting heat stress and transcriptional and biochemical data observed in Arabidopsis thaliana can represent a starting point for further studies focusing on agricultural crops.
Supplementary Materials: The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/plants11091130/s1, Figure S1: Experimental plan; Figure S2: Category and terms reported in the DAVID analysis. Table S1: Set of primers used for the qPCR; Table S2