The Analysis of Pinus pinaster SnRKs Reveals Clues of the Evolution of This Family and a New Set of Abiotic Stress Resistance Biomarkers

Climate change is increasing the intensity and incidence of environmental stressors, reducing the biomass yields of forestry species as Pinus pinaster. Selection of new stress-tolerant varieties is thus required. Many genes related to plant stress signaling pathways have proven useful for this purpose with sucrose non-fermenting related kinases (SnRK), conserved across plant evolution and connected to different phosphorylation cascades within ABA- and Ca2+-mediated signaling pathways, as a good example. The modulation of SnRKs and/or the selection of specific SnRK alleles have proven successful strategies to increase plant stress resistance. Despite this, SnRKs have been barely studied in gymnosperms. In this work P. pinaster SnRK sequences (PpiSnRK) were identified through a homology- and domain-based sequence analysis using Arabidopsis SnRK sequences as query. Moreover, PpiSnRKs links to the gymnosperm stress response were modeled out of the known interactions of PpiSnRKs orthologs from other species with different signaling complexity. This approach successfully identified the pine SnRK family and predicted their central role into the gymnosperm stress response, linking them to ABA, Ca2+, sugar/energy and possibly ethylene signaling. These links made the gymnosperm kinases promising candidates into the search for new stress resistance-related biomarkers, which would be useful into future breeding strategies.


Introduction
The ever-increasing intensity of environmental stressors and the changes into their seasonal incidence, both associated to climate change, have a negative and global impact in plant biomass yields [1][2][3][4]. Pinus pinaster is a gymnosperm species with a key environmental and forestry relevance into the northwestern Mediterranean area, where it is used for afforestation due to its fast growth, high phenotypic plasticity and stress tolerance, soil stabilization capacity and the quality of its timber and oleoresins [5,6]. Moreover, this species is also a good example of the climate-associated reductions in plant biomass yield related to regional climatic changes, increasing abiotic-heat and drought-and biotic stress pressure, which are also a main constraint for other Pinus species [7][8][9][10].
Multiple efforts have been started globally to avoid climate change effects over plant production directed to the use of better management strategies and the selection of more stress tolerant and productive plant varieties [3]. Despite new stress tolerant varieties are required to cope with the climate change effects and/or increase productivity, their generation faces the complexity of plant traits as stress resistance, growth speed and wood quality. Compared to crops, in tree species and particularly in gymnosperms as P. pinaster the complexity problem add up to the large and poorly characterized species, promising species for this purpose. The modulation of the PpiSnRK orthologs of these species and/or their substitution for their P. pinaster counterparts would allow the characterization of the P. pinaster PpiSnRKs, reducing the required time-consuming transformation and/or selection strategies in the tree species. Moreover, the identification of the orthology links with these plant and microalgae species, with different positions in relation to the last microalgae-plant common ancestor, would trace down the evolution of the complex plant SnRK families and evidence many of the specific features of the gymnosperms SnRK families.
The known relevance of SnRKs for the enhancement of stress resistance and biomass production in plants, and the lack of knowledge of this kinase family in gymnosperms, has motivated the description of the SnRK family of the gymnosperm species P. pinaster. PpiSnRKs were described, and their interactions modeled through an approach based on sequence analysis and protein interaction prediction, comparing the pine SnRKs with their orthologs from different plant species. These approaches would give hints about the evolution of SnRK family in gymnosperms and predict the specific role of the identified PpiSnRKs, which would ultimately contribute to the selection of specific ones related to some traits desired in forestry. These selected PpiSnRKs and or their related proteins could be characterized in simpler plant systems and then possibly used as biomarkers to drive future P. pinaster breeding programs.

Description of the P. Pinaster PpiSnRK Family
P. pinaster transcriptome and proteome [25] available at Plaza [26] were used for the identification of the gymnosperm SnRK orthologs to the Arabidopsis SnRK family [14]. All of the Arabidopsis sequences of the family were used as query into BLAST-and BLASTP-based searches against P. pinaster transcriptome and proteome using the default NCBI BLAST and BLASTP parameters. Only hits with e-values lower than e -25 were considered for further analyses. Non-SnRK sequences were filtered out from homology results through the analysis of their domain composition using the version 75.0 of the Inter Pro Scan database [27]. Only P. pinaster sequences with the Arabidopsis SnRKs canonical domain composition-including Ser/Thr kinase domains (PTHR24343, PTHR43895), UBA (IPR015940), KA1/ αCTD (IPR001772), Immunoglobulin E-set (IPR014756), AMPK glycogen binding subunit (IPR032640), ASC (IPR006828), Immunoglobulin E-set (IPR014756), CBS (IPR000644) and NAF/FISL (IPR018451) domains (Table S1)-were maintained for further analyses as previously specified [28]. Pinus pinaster PpiSNRK transcript sequences were used for the identification of possible PpiSNRK isoforms. PpiSNRK transcripts were aligned with MUSCLE alignment algorithm using default parameters, considering a group of aligned transcripts isoforms when sharing 3 and 5 UTR regions. Pinus pinaster PpiSNRK transcripts [25] and P. taeda [26] transcript orthologs to the Arabidopsis SnRK sequences were used for the curation of the incomplete and fragmented P. pinaster PpiSnRK protein models. The P. pinaster and P. taeda transcript sequences were translated, and the resulting protein sequences aligned using the M-Coffee consensus alignment method with default parameters to their respective PpiSNRK protein models for their curation.

Alignment of SnRK Sequences
PpiSnRK sequences were aligned with other SnRK sequences belonging to Arabidopsis and different microalgae species [28] (Table S2) using the M-Coffee consensus alignment method with default parameters [29]. Alignment was filter-curated through the Transitive Consistency Score algorithm (TCS) using default parameters [30]. Curated alignment distances were used in the generation of a maximum likelihood (ML) tree whose consistency was evaluated through the Transfer Bootstrap Expectation score (TBE, 500 replicates) included in the FastTree workflow within the Booster platform [31] due to the high number of involved sequences. SnRK2 sequences from microalgae and different plant species including P. pinaster PpiSnRK2, Arabidopsis, Amborella and Selaginella (Table S2) were aligned using M-Coffee multiple alignment Agronomy 2020, 10, 295 4 of 13 method with default parameters [29]. Different alignments were made by changing the included plant species set to test the resulting trees topology consistency. TCS weight-curated alignment distances were used to build rooted ML trees whose consistency was evaluated through Felsestein's bootstrap (FBP, 500 replicates) into the PhyML-SMS workflow within the booster platform [31].

Bioinformatic and Statistical Analyses
All the procedures for the identification and classification of P. pinaster PpiSnRK were performed locally employing the bioinformatics suite Geneious v7 (Biomatters Inc.), with the exception of Inter Pro Scan [27] searches, M-Coffee alignments and TCS-based alignment curation, and ML trees construction that were respectively performed at the European Bioinformatics Institute (ebi.ac.uk), T-Coffee (tcoffee.crg.cat) and booster (booster.pasteur.fr) websites.
Protein-protein functional interactions between the pine PpiSnRK2 sequences were inferred using the STRING v11 [32] database. As P. pinaster or other gymnosperms were not available at the interaction database, PpiSnRK2 orthologs from related STRING-available species as Amborella, Selaginella and Arabidopsis were separately used as query into the STRING website (string-db.org). Three networks were made (one for each species) showing high confidence (over 0.7 STRING interaction score), known and predicted associations between proteins.

Identification of the P. pinaster SnRK Family Members
The search of the Arabidopsis SnRKs [14] in P. pinaster transcriptome and proteome by using BLAST and BLASTP revealed a large protein set including SnRK homologs and non-SnRK sequences sharing the conserved Ser/Thr kinase domain. The analysis of the domain composition of these proteins allowed to unequivocally distinguish the PpiSnRK sequences attending to the exclusive SnRK domains surrounding the conserved kinase domain and/or the specific features of their kinase domain. The combination of BLAST-based search and domain analysis over P. pinaster available sequence data resulted in the confident identification of 7 PpiSnRK1-associated regulatory subunits and 45 PpiSnRK kinases belonging to the subfamilies 1, 2 and 3, including multiple isoforms (Table S1).
Agronomy 2020, 10, x FOR PEER REVIEW 5 of 13 The third cluster, PpiSnRK2 sequences (Ser/Thr kinase (PTHR24343) domain), was divided into SnRK2(A) and (B) clusters as previously described [28] with all the P. pinaster PpiSnRK2 and Arabidopsis SnRK2 sequences falling into the SnRK2(A) cluster ( Figure 1, Table S1). SnRK2(A) also included a small group of plant-like microalgae SnRK2s conforming the Group 2 of SnRK2(A) ( Figure  1). P. pinaster sequences as PpiSnRK2.5, and PpiSnRK2.6 isoforms were confidently clustered along Arabidopsis SnRK2.1, 2.4, 2.5, 2.9 and 2.10 sequences, but the tree failed to cluster remaining ones with any of the Arabidopsis SnRK2 sequences ( Figure 1).  (1). Tree divided the sequences into the tree SnRK families (SnRK1, 2, 3). SnRK1 subfamily was divided into SnRK1 like and SnRK1 sequences, SnRK3 sequences was divided into five different subgroups and SnRK2 sequences were divided between SnRK2A and SnRK2B. P. pinaster PpiSnRK2 sequences were all included into the SnRK2A group along Arabidopsis SnRK2s.   (1). Tree divided the sequences into the tree SnRK families (SnRK1, 2, 3). SnRK1 subfamily was divided into SnRK1 like and SnRK1 sequences, SnRK3 sequences was divided into five different subgroups and SnRK2 sequences were divided between SnRK2A and SnRK2B. P. pinaster PpiSnRK2 sequences were all included into the SnRK2A group along Arabidopsis SnRK2s.

Description of the P. pinaster PpiKIN2 Subfamily Structure and Sequence Features
M-Coffee-based trees of P. pinaster PpiSnRK2, and other plant and microalgae SnRK2 orthologs confidently separated plant SnRK2 and microalgae CKIN2/SnRK2 (Figure 2) as the previous analysis ( Figure 1). Arabidopsis SnRK2 sequences were confidently grouped into the three previously described groups, namely I (ABA insensitive), II (Low ABA sensitivity) and III (high ABA sensitivity) [34] into all SnRK2 trees ( Figure 2). Pinus pinaster PpiSnRK2.5, 2.6.1 and 2.6.2, and PpiSnRK2.1, 2.2 and 2.3 fell within Groups I and III respectively into all tested tree topologies. Conversely, PpiSnRK2.4 isoforms did not cluster consistently with any of the previous groups ( Figure 2). Most P. pinaster and Amborella (phylogenetically closer to pine than Arabidopsis) SnRK2 were consistently clustered within I and III Groups along the Arabidopsis sequences. On the other hand, the consistency of Group II was lower across trees and Arabidopsis SnRK2(II) sequences-SnRK2.7 and SnRK2.8-were the only constantly present in this branch ( Figure 2). Amborella SnRK2(II) sequence only clustered with Arabidopsis SnRK2(II) sequences into one of the trees (Figure 2b), and P. pinaster PpiSnRK2.4 isoforms were either equidistant to all groups (Figure 2a,c) or to II and III groups (Figure 2b). Selaginella sequences had a basal position in the plant SnRK2 branch of the SnRK2/CKIN2 tree and none of its sequences were clustered within the previously defined groups (Figure 2c). Furthermore, built trees were unable to converge into a basal SnRK2 group (Figures 1 and 2).
The ABA-sensitive cluster III of SnRK2 was highly similar between Arabidopsis and P. pinaster. PpiSnRK2.1, 2.2 and 2.3 had highly conserved regulatory domains I and II (ABA box) after the characteristic SnRK Ser/Thr kinase domain, supporting their involvement in the gymnosperm ABA signaling (Figure 3c). P. pinaster PpiSnRK2.5, 2.6.1 and 2.6.2, within the Group I SnRK2, also conserved a regulatory domain I after the Ser/Thr kinase domain (Figure 3a). PpiSnRK2.4 isoforms conserved the regulatory domain I, which was followed by a C-terminal sequence with features observed in the regulatory domain II of Group II and III SnRK2 sequences (Figure 3b). Selaginella SnRK2 homolog sequences had a basal position in the SnRK2 tree and did not cluster with any of the defined SnRK2 groups (Figure 2c). In spite of this, these sequences showed analogy to Arabidopsis ABA sensitive II and III SnRK2. Selaginella SnRK2 had a conserved Domain I followed by an aspartate-rich domain II-like. These acidic aspartate residues are characteristic of the SnRK2s terminal domains of Groups II and III (Figures 2 and 3, Figure S1).

Arabidopsis, Amborella and Selaginella Orthologs to P. Pinaster PpiSnRK2 Are Connected to Stress Signaling and Metabolism Modulation
The Arabidopsis, Amborella and Selaginella orthologs to P. pinaster PpiSnRK2 were used to build three different STRING-based protein interaction networks from which a network outline was created pointing to the potential function of the pine PpiSnRK2 sequences (Figure 4, Figure S1). All the PpiSnRK2 orthologs across the different included species shared a connection to the SnRK1-related regulatory PpiSnRKβγ subunit and different starch modulatory enzymes, and to many PP2C and bZIP elements (Figure 4, Figure S1). In Amborella and Arabidopsis PP2Cs were functionally connected to the ABA receptors PYL/PYR/RCAR, which were also directly associated to these species group III SnRK2 kinases (Figure 4, Figure S1). The ABA receptor Mg chelatase (GUN5 or CHLH) and different ubiquitination-related elements were other conserved associations across the different species SnRK2 sequences. Despite the multiple connections conserved between the different species SnRK orthologs Selaginella and Amborella showed exclusive interactions. A protein phosphatase 2C and cyclic nucleotide-binding/kinase domain-containing protein (PP2C-L) was associated in the Selaginella and Amborella STRING networks (Figure 4, Figure S2b,c) with the conserved cluster of SnRK1, its associated regulatory βγ subunits and starch regulatory enzymes, while Amborella SnRK2(III) and Selaginella OST2.1 were connected to ethylene-and protein glycosylation-related elements respectively (Figure 4, Figure S1). were the only constantly present in this branch ( Figure 2). Amborella SnRK2(II) sequence only clustered with Arabidopsis SnRK2(II) sequences into one of the trees (Figure 2b), and P. pinaster PpiSnRK2.4 isoforms were either equidistant to all groups (Figure 2a,c) or to II and III groups ( Figure  2b). Selaginella sequences had a basal position in the plant SnRK2 branch of the SnRK2/CKIN2 tree and none of its sequences were clustered within the previously defined groups (Figure 2c). Furthermore, built trees were unable to converge into a basal SnRK2 group (Figures 1 and 2). The ABA-sensitive cluster III of SnRK2 was highly similar between Arabidopsis and P. pinaster. PpiSnRK2.1, 2.2 and 2.3 had highly conserved regulatory domains I and II (ABA box) after the characteristic SnRK Ser/Thr kinase domain, supporting their involvement in the gymnosperm ABA signaling (Figure 3c). P. pinaster PpiSnRK2.5, 2.6.1 and 2.6.2, within the Group I SnRK2, also  by an aspartate-rich domain II-like. These acidic aspartate residues are characteristic of the SnRK2s terminal domains of Groups II and III (Figures 2 and 3, Figure S1).

Arabidopsis, Amborella and Selaginella Orthologs to P. Pinaster PpiSnRK2 Are Connected to Stress Signaling and Metabolism Modulation
The Arabidopsis, Amborella and Selaginella orthologs to P. pinaster PpiSnRK2 were used to build three different STRING-based protein interaction networks from which a network outline was created pointing to the potential function of the pine PpiSnRK2 sequences (Figure 4, Figure S1). All the PpiSnRK2 orthologs across the different included species shared a connection to the SnRK1related regulatory PpiSnRKβγ subunit and different starch modulatory enzymes, and to many PP2C and bZIP elements (Figure 4, Figure S1). In Amborella and Arabidopsis PP2Cs were functionally connected to the ABA receptors PYL/PYR/RCAR, which were also directly associated to these species group III SnRK2 kinases (Figure 4, Figure S1). The ABA receptor Mg chelatase (GUN5 or CHLH) and different ubiquitination-related elements were other conserved associations across the different species SnRK2 sequences. Despite the multiple connections conserved between the different species SnRK orthologs Selaginella and Amborella showed exclusive interactions. A protein phosphatase 2C and cyclic nucleotide-binding/kinase domain-containing protein (PP2C-L) was associated in the Selaginella and Amborella STRING networks (Figure 4, Figure S2b,c) with the conserved cluster of SnRK1, its associated regulatory βγ subunits and starch regulatory enzymes, while Amborella SnRK2(III) and Selaginella OST2.1 were connected to ethylene-and protein glycosylation-related elements respectively (Figure 4, Figure S1).

Discussion
Despite the great divergence between P. pinaster and Arabidopsis, the identified pine PpiSnRK and Arabidopsis SnRK families had similar size, structure, and sequence features, and potentially also similar interactions and roles. These parallelisms could be extended to the SnRK2 subfamilies of Amborella-a sequenced embryophyte close to the last gymnosperm-angiosperm common ancestor-and Selaginella, an also sequenced protovascular plant more ancient than gymnosperms. The observed similarities between these distant species point to the early origin of the SnRK family configuration and function known in Arabidopsis and other land plants [14,28]. Furthermore, the conservation evidences and the simpler SnRK families of these ancient model species motivated their election for the characterization of the pine PpiSnRK elements.
Pinus pinaster and Arabidopsis share large and similar SnRK3 subfamilies. A reduced group of these SnRK3 sequences could be considered founder as were close in the sequence tree to Agronomy 2020, 10, 295 9 of 13 microalgae SnRK3/CKIN3 (Figure 1). Interestingly, between the plant sequences included in this group are Arabidopsis SnRK3.11 or SOS2, a salt stress responsive kinase within the SOS pathway [33], osmostress-sensitive SnRK3.13 [35] and P. pinaster PpiSnRK3.3 and PpiSnRK3.4. This closeness points at once to the ancient (microalgae) origin of this pathway and by analogy to the possible involvement of both microalgae and P. pinaster PpiSnRK3.3 and PpiSnRK3.4 into salt and/or osmotic stress responses.
Many of the remaining Arabidopsis SnRK3 sequences, more distant from the microalgae ones, and clustering into the 3-5 SnRK3 clusters within the sequence tree (Figure 1) are involved in ABA stress signaling and expressed under certain developmental times and in specific tissues [36]. Thus, the diversification of the plant SnRK3 subfamily from the founder elements might have been related to the increase in structure and life cycle complexity of land plants and/or to their adaption to the more stressing land environment. Same explanation and function thus could be applied to the P. pinaster PpiSnRK3 sequences clustering along with the Arabidopsis ones in the 3-5 groups ( Figure 1).
As SnRK3 and PpiSnRK3 subfamilies, PpiSnRK2 subfamily had a group division highly similar to Arabidopsis SnRK2. The SnRK2 subfamily sequence and structure similarity could also be expanded to Amborella, at the base of angiosperm group. Interestingly, although Amborella has a small SnRK2 family (three members), its members represent the three Arabidopsis SnRK2 groups. Again, the shared SnRK complexity and features between these distantly related species points to ancient duplication/divergence events-before angiosperm and gymnosperm division-but also to the conservation of these elements giving origin to the observed family similarity and pointing to their shared relevance and function. Besides this, the small size of Amborella SnRK2 family also points to gymnosperm and angiosperm independent SnRK2 duplication events. These independent duplication events could explain the observed grouping issues between elements sharing group II SnRK2 features as Arabidopsis SnRK2.7, SnRK2.8, Amborella AmSnRK2(II) and P. pinaster PpiSnRK2.4 isoforms.
On the other hand, Selaginella, a non-seed plant, has a simpler family structure exclusively composed by elements close to Arabidopsis SnRK2 II and III groups. The absence of ABA-insensitive group I SnRK2 sequences in Selaginella and their presence in Amborella, Arabidopsis and P. pinaster makes them the most recent SnRK2s. Group I SnRK2 would have originated at some point between vascular plant and seed plant origin. Moreover, the Selaginella SnRK2 subfamily points to the earlier origin of the ABA-sensitive SnRK2s and thus to the early involvement of ABA in SnRK2-based signaling pathways. This SnRK2(III) exclusive subfamilies have also been found in the moss Physcomitrella patens and charophyte algae supporting also the earlier origin of the ABA-sensitive elements [37]. Strikingly, C. reinhardtii CKIN2.8, a plant-like SnRK2, expression is not sensitive to exogenous ABA treatments [28].
Approaches including sequence and domain/motif analysis have been widely used along tree-based sequence clustering for the identification and description of many plant gene families, including the SnRK family. These strategies combined sequence-based methods with the evaluation of the identified sequences expression patterns in order to link them to particular stresses [28,38,39]. Other works [28,40] complemented this approach with the prediction of the proteins associated to the newly identified ones through databases such as STRING containing known relations between the orthologs to these elements. The involvement of apple bHLH transcription factors and C. reinhardtii CKIN/SnRK into their stress response system was predicted using this database. STRING predicted the link with ABA signaling of some of the identified apple bHLH genes and the link of C. reinhardtii CKIN/SnRKs to Ca 2+ and PP2C signaling, and to the modulation of carbon and nitrogen metabolism [28,40].
STRING database was also chosen to predict the PpiSnRK2 sequences function and associations with the gymnosperm stress response system and metabolism using data from related plant and microalgae species with different positions within the Plantae species tree and different stress response system complexity. The overlap of the known and predicted STRING-based protein-protein interaction networks draw from the P. pinaster PpiSnRK2 orthologs in Arabidopsis, Amborella and Selaginella ( Figure S2) outlined a network model of the possible associations of the P. pinaster PpiSnRKs with other stress related elements (Figure 4). These networks showed the conserved SnRK2 connections with PP2C, bZIP transcription factors, ABA sensing/signaling and carbon metabolism modulation in Arabidopsis, Amborella and Selaginella ( Figure S2, Figure 4) and microalgae [28]. This conservation makes highly probable the involvement of the P. pinaster kinases with these signaling elements. SnRK1-2 related bZIP transcription factors are involved in plant abiotic and energetic stress responses in different plant species [41,42]. The overexpression of the Arabidopsis bZIP ABF2 confers tolerance to multiple abiotic stresses [43]. Interestingly, the pine PpiSnRK2 could have conserved connections to ethylene signaling and protein glycosylation modulation not observed into the Arabidopsis family.
The prediction of the PpiSnRK interactions using species more ancient than P. pinaster and Arabidopsis (Amborella, Selaginella and microalgae) have also allowed the identification of possible SnRK2 interactions to ethylene signaling, protein glycosylation and alternative central regulators as a PP2C-like protein. Despite these connections were not conserved in the Arabidopsis network, they would have been inherited by the P. pinaster sequences. To overcome pine experimental limitations, model species as Arabidopsis, but specially Selaginella and Amborella-with simpler but close SnRK families-would contribute to the characterization of PpiSnRKs. The knowledge of the specific role of these kinases within the gymnosperm stress response and the identification of their related elements would provide promising targets to drive new selection and breeding strategies. Between these targets would be the PpiSnRKs themselves, orthologs to the conserved bZIP transcription factors and their modulated genes but also novel elements, which are absent from the characterized angiosperms SnRK families. These elements would be interesting candidates for screening programs aimed at the identification of P. pinaster genotypes carrying specific sequence changes linked to higher endurance to environmental stress. Previous works based on the characterization of the genotype and secondary metabolism profile among Mediterranean P. pinaster populations adapted to different environmental conditions found differences related to the individuals tolerance to stress [5,6]. The candidates found in this work are also an interesting base to identify novel stress associated variations, which could be used as biomarkers of better stress endurance and contribute to breeding/selection programs for the generation of more stress tolerant varieties both in P. pinaster and other close gymnosperm species.

Conclusions
This work successfully identified and described the SnRK family members of P. pinaster, a non-sequenced tree species with a large genome. The success of this approach was based on the search of the orthologs to the well described and characterized Arabidopsis SnRKs into the available transcriptome/proteome of P. pinaster and other close Pinus species. The identification of the gymnosperm SnRKs (PpiSnRK) search was complemented by the prediction of their associations with other stress response-related elements. This prediction required the comparison of the PpiSnRK elements with other known plant SnRK families including Arabidopsis and other plant species such as Amborella and Selaginella with simpler SnRK families. The pine PpiSnRK family showed a high resemblance to those of Arabidopsis but also to Amborella and Selaginella, giving hints about the pine family evolution, structure and function. SnRK2 subfamilies of Arabidopsis, Amborella and P. pinaster shared the same structure divided into ABA sensitive (Groups II, III PpiSnRK2.4 and PpiSnRK2.1, 2.2, 2.3 respectively) and insensitive (group I PpiSnRK2.5, 2.6) elements. This similarity was extended to the SnRK3 subfamily, PpiSnRK3.3 and 3.4 closeness to salt-sensitive Arabidopsis SOS2 pointed to the involvement of these PpiSnRK3 into the P. pinaster osmotic and/or salt stress response. Moreover, STRING protein interaction networks over related species suggested the conservation of the gymnosperm PpiSnRK2 interaction with known SnRK2 interactors as PP2C, bZIP and SnRK1 while highlighting possible novel interactions. These results supported the use of these model species especially the ones with simpler but close SnRK systems as Amborella for the characterization of the gymnosperm SnRKs, overcoming the species experimental difficulties. The knowledge of the specific role of these kinases within the gymnosperm stress response and the identification of their related elements would provide promising targets to drive new selection and breeding strategies. Between these targets would be the PpiSnRKs themselves, orthologs to the conserved bZIP transcription factors and their modulated genes but also novel elements, which are absent from the characterized angiosperm SnRK families. These elements would be interesting candidates for screening programs aimed at the identification of P. pinaster genotypes carrying specific sequence changes linked to higher endurance to environmental stress. These candidates would be used as biomarkers of better stress endurance and contribute to breeding/selection programs for the generation of more stress tolerant varieties both in P. pinaster and other close gymnosperm species.

Conflicts of Interest:
The authors declare no conflict of interest.