Genome-Wide Characterization and Expression Profiling of ABA Biosynthesis Genes in a Desert Moss Syntrichia caninervis

Syntrichia caninervis can survive under 80–90% protoplasmic water losses, and it is a model plant in desiccation tolerance research. A previous study has revealed that S. caninervis would accumulate ABA under dehydration stress, while the ABA biosynthesis genes in S. caninervis are still unknown. This study identified one ScABA1, two ScABA4s, five ScNCEDs, twenty-nine ScABA2s, one ScABA3, and four ScAAOs genes, indicating that the ABA biosynthesis genes were complete in S. caninervis. Gene location analysis showed that the ABA biosynthesis genes were evenly distributed in chromosomes but were not allocated to sex chromosomes. Collinear analysis revealed that ScABA1, ScNCED, and ScABA2 had homologous genes in Physcomitrella patens. RT-qPCR detection found that all of the ABA biosynthesis genes responded to abiotic stress; it further indicated that ABA plays an important role in S. caninervis. Moreover, the ABA biosynthesis genes in 19 representative plants were compared to study their phylogenetic and conserved motifs; the results suggested that the ABA biosynthesis genes were closely associated with plant taxa, but these genes had the same conserved domain in each plant. In contrast, there is a huge variation in the exon number between different plant taxa; it revealed that ABA biosynthesis gene structures are closely related to plant taxa. Above all, this study provides strong evidence demonstrating that ABA biosynthesis genes were conserved in the plant kingdom and deepens our understanding of the evolution of the phytohormone ABA.


Introduction
Abscisic acid (ABA) is a very important carotenoid-derived phytohormone and plays a significant role in plant growth and development under biotic or abiotic stress conditions [1]. As a sesquiterpene structure compound (C 15 ), ABA is synthesized directly in fungi [2,3], while in plants, it is usually produced through carotenoid cleavage (C 40 ), which is an indirect synthetic pathway ( Figure 1A) [4,5]. The first step in the ABA biosynthetic pathway is the conversion of zeaxanthin to all-trans-violaxanthin by zeaxanthin epoxidase (ZEP) [6]. ABA1 encodes ZEP in Arabidopsis thaliana [7]. All-trans-violaxanthin is then converted to 9 -cis-neoxanthin and 9 -cis-violaxanthin, which are, respectively, catalyzed by neoxanthin synthase (NSY) and an unknown isomerase [8]. ABA4 encodes NSY in A. thaliana [9]. The next step, 9 -cis-violaxanthin and 9 -cis-neoxanthin cleavage by a 9-cis-epoxycarotenoid dioxygenase (NCED) is generally considered the rate-limiting step in the ABA biosynthetic pathway [10]. The cleavage product xanthoxin is then converted to abscisic aldehyde by a short-chain alcohol dehydrogenase encoded by the ABA2 gene [11]. The final step, abscisic aldehyde oxidizes to ABA, is catalyzed by an abscisic aldehyde's oxidation (AAO) [12]. Arabidopsis ABA3 encodes molybdenum cofactor (MoCo), which is required for AAO The total number of ABA biosynthesis genes in all plants; the green word represents different plant taxa; the number of ABA biosynthesis genes was visualized using TBtools with a homogenization method of log2; the circle represents the specific number of ABA biosynthesis genes.

Identification and Characteristic of ABA Biosynthesis Genes in S. caninervis
Combining the results of BLASTP and HMMER, a total of 42 ABA biosynthesis genes were identified in the S. caninervis genome, including one ScABA1 gene, two ScABA4 genes, five ScNCED genes, twenty-nine ScABA2 genes, one ScABA3 gene and four ScAAO genes (Tables 1 and S1). ScABA1 protein sequence analysis showed that it encoded 681 amino acids (aa) with an average molecular weight (MW) at 73.9 kDa and isoelectric points (pI) at 8.15. Simultaneously, subcellular mapping prediction indicated that the ScABA1 gene was located in the chloroplast. The predicted protein lengths of ScABA4 genes (Sc_g04273/Sc_g01818) were 252 and 199 amino acids; ScABA4 proteins' molecular weights were 27.8 and 13.6 kDa, pI were 9.71 and 7.66, and they were located in chloroplasts and the endomembrane, respectively. The length of ScNCED proteins ranged from 540 to 615 amino acids, the predicted molecular weight ranged from 60.5 to 68.2 kDa, and pI ranged from 5.59 to 7.57, and these ScNCED genes were located in chloroplasts and cytoplasm. As S. caninervis had a number of ScABA2 genes, we chose the top five ScABA2 genes according to the E-values of the BLASTP screen for further comparison and analysis of their characteristics. The ScABA2 protein sequences ranged from 254 to 335 amino acids in length, their molecular weights ranged from 26.1 to 36.7 kDa, their pI ranged from 5.52 to 8.24, and most of them were located in the mitochondrion. ScABA3 protein sequence analysis showed that it was composed of 919 amino acids, its molecular weight was 101.8 kDa, pI was 7.1, and it was located in the extracellular space. Four ScAAO protein sequences analysis showed that their lengths ranged from 1273 to 1334 amino acids, molecular weights ranged from 138.5 to 144.9 kDa, pI ranged from 6.14 to 6.67, and most of them were located in the cytoplasm.
To comprehensively evaluate the characteristics of the ABA biosynthesis genes in S. caninervis, we also identified these genes in the model bryophyte P. patens genome. The result indicated that the ABA biosynthesis gene number in P. patens was the same as that in S. caninervis, except that P. patens had eight PpNCEDs and twenty PpABA2s (Tables 1 Expression of ABA biosynthesis genes is closely related to endogenous ABA levels, so they will directly determine ABA's action in different plant stress responses. Several studies have investigated the evolution of ABA biosynthesis-related genes across land plant lineages using bioinformatics, but very few are known outside of angiosperms, including species of gymnosperms, ferns, lycophytes, bryophytes, and chlorophytes; thus, this is an expanding field [15]. This research has revealed novel insights into the ABA biosynthesis pathway across different plant taxa, especially in bryophytes. Bryophytes are key organisms in plant evolution, and clarification of their ABA biosynthesis processes is important for understanding land plant evolutionary adaptation [16]. The moss Physcomitrella patens is the model plant in bryophyte research [17]; recent research on the P. patens ABA biosynthesis pathway has revealed it contain PpABA1, PpABA4, PpNCED, and PpABA3 genes and they play roles in plant stress response. For example, PpABA1 can enhance osmotic acclimation, and the transcript abundance of PpNCED is significantly increased under salt stress, while PpABA2 and PpAAO have not been reported [18][19][20]. In general, the data on bryophytes' ABA biosynthesis are currently scarce. One of the most striking aspects of ABA biosynthesis is the drastic increase in ABA levels during dehydration [21], and an ABA-induced increase in stress tolerance is also reported in other land plant lineages, including nonvascular bryophytes that diverged from vascular plants more than 480 million years ago. Desiccation tolerance (DT) moss is a type of bryophyte that can survive under 80-90% protoplasmic water loss [22]. ABA plays an important role in DT moss, as in angiosperms, ferns, lycophytes, and other desiccationtolerant plants [21]. Syntrichia caninervis Mitt. is a desert moss distributed in the arid land of Central Asia and North America [23,24]; it is one of the dominant species in the biological soil crusts of deserts, and it plays an important role in ecological restoration and biodiversity conservation in the desert ecosystem [25]. S. caninervis is the only DT moss for which its genome has been sequenced [26]; it propagates rapidly and can regenerate a large number of gametophytes within one month [27]. Moreover, S. caninervis contains a large number of resistance genes, such as ScELIPs, ScALDHs, and ScDREBs [28][29][30][31]. From the above discussion, it appears that S. caninervis is not only a model plant for the study of DT mechanism and excellent stress tolerance gene mining [32] but also a typical material to study the acquisition of the ABA biosynthesis genes.
As reported in the literature, S. caninervis is capable of ABA biosynthesis, and ABA concentrations increase under dehydration stress [32,33]. Nevertheless, until now, no data on S. caninervis ABA biosynthesis genes have been reported. In this study, we identified and analyzed one ScABA1 gene, two ScABA4 genes, five ScNCED genes, twenty-nine ScABA2 genes, one ScABA3 gene, and four ScAAO genes in the S. caninervis genome, and investigated their expression pattern under different abiotic stresses. In addition, we also identified and compared ABA biosynthesis genes in additional bryophytes, chlorophytes, and other plant taxa. The present study, therefore, focused on the ABA biosynthesis genes in early-diverging plant species, including bryophytes and chlorophytes, and recent progress in the model desiccation-tolerant moss S. caninervis and the model moss P. patens. With recent advances in our capacity to characterize gene information in different plant taxa, we are on the cusp of revealing the origins of these critical ABA biosynthesis genes and understanding how ABA may have shaped land plant adaptation.

Identification and Characteristic of ABA Biosynthesis Genes in S. caninervis
Combining the results of BLASTP and HMMER, a total of 42 ABA biosynthesis genes were identified in the S. caninervis genome, including one ScABA1 gene, two ScABA4 genes, five ScNCED genes, twenty-nine ScABA2 genes, one ScABA3 gene and four ScAAO genes (Tables 1 and S1). ScABA1 protein sequence analysis showed that it encoded 681 amino acids (aa) with an average molecular weight (MW) at 73.9 kDa and isoelectric points (pI) at 8.15. Simultaneously, subcellular mapping prediction indicated that the ScABA1 gene was located in the chloroplast. The predicted protein lengths of ScABA4 genes (Sc_g04273/Sc_g01818) were 252 and 199 amino acids; ScABA4 proteins' molecular weights were 27.8 and 13.6 kDa, pI were 9.71 and 7.66, and they were located in chloroplasts and the endomembrane, respectively. The length of ScNCED proteins ranged from 540 to 615 amino acids, the predicted molecular weight ranged from 60.5 to 68.2 kDa, and pI ranged from 5.59 to 7.57, and these ScNCED genes were located in chloroplasts and cytoplasm. As S. caninervis had a number of ScABA2 genes, we chose the top five ScABA2 genes according to the E-values of the BLASTP screen for further comparison and analysis of their characteristics. The ScABA2 protein sequences ranged from 254 to 335 amino acids in length, their molecular weights ranged from 26.1 to 36.7 kDa, their pI ranged from 5.52 to 8.24, and most of them were located in the mitochondrion. ScABA3 protein sequence analysis showed that it was composed of 919 amino acids, its molecular weight was 101.8 kDa, pI was 7.1, and it was located in the extracellular space. Four ScAAO protein sequences analysis showed that their lengths ranged from 1273 to 1334 amino acids, molecular weights ranged from 138.5 to 144.9 kDa, pI ranged from 6.14 to 6.67, and most of them were located in the cytoplasm. To comprehensively evaluate the characteristics of the ABA biosynthesis genes in S. caninervis, we also identified these genes in the model bryophyte P. patens genome. The result indicated that the ABA biosynthesis gene number in P. patens was the same as that in S. caninervis, except that P. patens had eight PpNCEDs and twenty PpABA2s (Tables 1 and S1). In total, the characteristics of ABA biosynthesis genes were similar in both bryophytes, which suggested that these genes were conserved in bryophyte species. The PpABA1 protein was 685 amino acids in length, with a predicted molecular mass of 74.4 kDa and pI of 7.85, and it was located in chloroplast too. The predicted protein lengths of PpABA4 genes (Pp3c12_21540/Pp3c3_37490) were 297 and 264 amino acids, PpABA4 proteins molecular weights were 33 and 29.2 kDa, pI were 10.03 and 9.69, and they were located in mitochondrion and chloroplast respectively. All PpNCED proteins ranged from 538 to 622 amino acids in length, molecular weights ranged from 61.2 to 69.9 kDa, pI ranged from 5.63 to 5.84, and they were mainly located in the cytoplasm. PpABA2 protein sequence analysis showed that they encoded 306-359 amino acids, molecular weight ranged from 31.7 to 39.3 kDa, putative pI ranged from 5.89 to 8.77, and they were dispersedly localized in the chloroplast, mitochondrion, and cytoplasm. PpABA3 protein sequence analysis showed that it encoded 940 amino acids, the molecular weight was 104.2 kDa, and the pI was 6.34. PpABA3 was located in chloroplast, which distinguished it from S. caninervis. PpAAO protein sequences analysis showed that they were 1283 to 1396 amino acids in length, predicted molecular mass ranged from 139.2 to 152.2 kDa, pI ranged from 5.86 to 6.46, and they were mainly located in the cytoplasm.

Gene Number of ABA Biosynthesis Genes in Different Plant Taxa
On the basis of the identification and characterization of the ABA biosynthesis genes in S. caninervis and P. patens, it can be concluded that the ABA biosynthesis genes existed in bryophytes. Nevertheless, the comprehensive analysis of ABA biosynthesis genes in chlorophyte species has not been reported. To better understand the evolution of the ABA biosynthesis genes, here, we summarized the number of ABA1, ABA4, NCED, ABA2, ABA3, and AAO genes among 19 representative species from chlorophytes to angiosperms ( Figure 1). As shown in Figure 1B, all of the ABA biosynthesis genes were present in different plant taxa, whether in lower or higher plants.
Specifically, ABA1 is a single gene in all plant species except Z. mays and P. trichocarpa ( Figure 1B), so the ABA1 gene quantity is very stable in different plant genomes. The result of the ABA4 gene number analysis indicated that the bryophyte was a transition taxon because all of the lower plant chlorophytes had two ABA4 genes, and most of the higher plant angiosperm had only one ABA4 gene; however, some bryophytes had one, and the other bryophytes had two ABA4 genes, such as P. patens, S. caninervis, and C. purpureus ( Figure 1B). According to the identification of the NCED genes in 19 plants, we found that all plant species had five NCED genes at least; angiosperms had the largest number of NCED genes, ranging from 5 to 26; chlorophytes had the lowest number of NCED, ranging from 5 to 7, and bryophyte plants were in the middle, ranging from 5 to 12 ( Figure 1B). The number of ABA2 genes suggested that bryophyte was in the middle ranging from 17 to 29, with the minimum in chlorophyte ranging from 4 to 11 and the maximum in angiosperms ranging from 12 to 56 ( Figure 1B). ABA2 belongs to the short-chain dehydrogenase/reductase (SDR) family; there are 56 ABA2/SDR genes in A. thaliana, but only one AtABA2 (AT1G52340) participates in ABA biosynthesis. Moreover, each plant had many ABA2 genes in our results, but we can not accurately confirm which ones are involved in ABA biosynthesis merely by using bioinformatic analysis. As ABA3 is a single gene, its number in 19 plants was similar to that of the ABA1 gene ( Figure 1B). Each chlorophyte had only one AAO, which differed from bryophytes (3-7 AAO genes) and angiosperms (2-6 AAO genes). All in all, the number of ABA biosynthesis genes was closely related to plant taxa.

Phylogenetic and Conserved Motif Analysis of ABA Biosynthesis Genes in Various Plants
Great progress regarding ABA biosynthesis genes has been made in angiosperms; however, little attention has been given to these genes in early-diverging plant species such as bryophytes and chlorophytes. To better understand the evolutionary relationships within all the ABA1, ABA4, NCED, ABA2, ABA3, and AAO genes in different plant taxa, phylogenetic and conserved motif analysis was conducted.
A maximum likelihood phylogenetic tree was constructed to explore the evolutionary relationship of 21 ABA1 genes from 19 species inferred from the amino acid sequences in IQTree v1.6.12 using the best-fit substitution model (WAG + I + G 4 ), which was automatically selected by the software according to the Bayesian information criterion scores and weights (BIC) with partitions. As shown in the phylogenetic tree (Figure 2A), the ABA1 protein sequences of S. caninervis, left bryophyte, lycophyte, and fern species were clustered together in one clade, which means that bryophyte species were closely related to each other. In addition, the ABA1 of three chlorophyte species and gymnosperm were clustered together in one clade, and angiosperm species were clustered together. The results of the conserved motif analysis showed that the ScABA1 protein was composed of the FAD_binding_3 and FHA domain, which was consistent in bryophytes even in different plant taxa ( Figure 2B). The A. trichopoda ABA1 protein is an exception because it did not contain the FHA domain. Moreover, most of the ABA1 proteins had two FAD_binding_3 domains, while a few of them had only one FAD_binding_3 domain, such as Z. mays, C. richardii, and S. fallax. Consequently, the phylogenetic and conserved domain results indicated that ABA1 is highly conserved in different plants. the FAD_binding_3 and FHA domain, which was consistent in bryophytes even in different plant taxa ( Figure 2B). The A. trichopoda ABA1 protein is an exception because it did not contain the FHA domain. Moreover, most of the ABA1 proteins had two FAD_bind-ing_3 domains, while a few of them had only one FAD_binding_3 domain, such as Z. mays, C. richardii, and S. fallax. Consequently, the phylogenetic and conserved domain results indicated that ABA1 is highly conserved in different plants. An ML phylogenetic tree of 28 ABA4 genes from 19 plant species was constructed in IQTree using the best-fit substitution model (JTT + F + I + G4), which was automatically selected by the software. As with ScABA1, ScABA4 was clustered together with other bryophyte, fern, and lycophyte species; among them, S. caninervis was closely associated with C. purpureus, B. argenteum, and P. patens. All three chlorophyte species, including six ABA4 proteins, formed one clade and had a close relationship. Moreover, the ABA4 genes of angiosperm species were clustered together and formed three clades. Meanwhile, a separate clade was formed for T. plicata (Thupl.29379874s0002), a gymnosperm, and another separate clade was formed for A. thaliana (AT1G67080) in the angiosperm ABA1 clade ( Figure 3A). In line with previous studies, ABA4 contains a DUF4281 domain, which is widely distributed across all plant taxa ( Figure 3B). At the same time, we found that there were shorter protein lengths in S. caninervis and S. moellendorffii. Above all, we demonstrated that ABA4 is highly conserved in the different plant taxa. An ML phylogenetic tree of 28 ABA4 genes from 19 plant species was constructed in IQTree using the best-fit substitution model (JTT + F + I + G 4 ), which was automatically selected by the software. As with ScABA1, ScABA4 was clustered together with other bryophyte, fern, and lycophyte species; among them, S. caninervis was closely associated with C. purpureus, B. argenteum, and P. patens. All three chlorophyte species, including six ABA4 proteins, formed one clade and had a close relationship. Moreover, the ABA4 genes of angiosperm species were clustered together and formed three clades. Meanwhile, a separate clade was formed for T. plicata (Thupl.29379874s0002), a gymnosperm, and another separate clade was formed for A. thaliana (AT1G67080) in the angiosperm ABA1 clade ( Figure 3A). In line with previous studies, ABA4 contains a DUF4281 domain, which is widely distributed across all plant taxa ( Figure 3B). At the same time, we found that there were shorter protein lengths in S. caninervis and S. moellendorffii. Above all, we demonstrated that ABA4 is highly conserved in the different plant taxa. In fact, we identified 206 NCED genes from 19 plant species, and the ML tree was constructed from the top five NCED genes of each species based on the E-value of BLASTP result in IQTree using the best-fit substitution model (LG + F + R9). We found that ScNCED genes in S. caninervis were always clustered together with C. purpureus; simultaneously, ScNCEDs had a close relationship with other bryophyte species. In addition, three chlorophyte species, including 15 NCED proteins, formed two clades, NCED proteins of fern, lycophyte, and gymnosperm were respectively clustered together, NCED proteins of angiosperm were scattered in the phylogenetic tree ( Figure S1A). The conserved motif analysis showed that all NCED proteins had one RPE65 domain, which was widely distributed across all plant taxa ( Figure S1B). In sum, the NCED gene is highly conserved, too.
ABA2 belongs to a multigene family, and we identified 471 ABA2 genes from 19 species. The ABA2 evolutionary tree was constructed from the top five ABA2 genes of each species as an NCED phylogenetic tree, and LG + R5 was the best-fit substitution model in IQTree for the construction of the ML phylogenetic tree of ABA2. The result showed that the classification of the phylogenetic tree was ambiguous because it did not cluster by plant taxa ( Figure S2A). Meanwhile, ScABA2 genes were always clustered together with different bryophyte species. Our conserved motif analysis showed that each ABA2 protein had an adh_short_C2, adh_short_C, and KR domain ( Figure S2B). In accordance with the phylogenetic and conserved domain results, ABA2 is very conserved in different plants.
An ML phylogenetic tree based on 20 ABA3 genes was constructed in IQTree using the best-fit substitution model (JTT + I + G4), which was automatically selected by the software. As shown in Figure 4A, ScABA3 was sister clade to C. purpureus, B. argenteum, P. patens and S. fallax and had a close relation. The ABA3 proteins of three chlorophytes, S. moellendorffii, and M. polymorpha, clustered together in one clade, and they were closely related to each other. The ABA3 of angiosperm species were clustered together; a separate In fact, we identified 206 NCED genes from 19 plant species, and the ML tree was constructed from the top five NCED genes of each species based on the E-value of BLASTP result in IQTree using the best-fit substitution model (LG + F + R 9 ). We found that ScNCED genes in S. caninervis were always clustered together with C. purpureus; simultaneously, ScNCEDs had a close relationship with other bryophyte species. In addition, three chlorophyte species, including 15 NCED proteins, formed two clades, NCED proteins of fern, lycophyte, and gymnosperm were respectively clustered together, NCED proteins of angiosperm were scattered in the phylogenetic tree ( Figure S1A). The conserved motif analysis showed that all NCED proteins had one RPE65 domain, which was widely distributed across all plant taxa ( Figure S1B). In sum, the NCED gene is highly conserved, too.
ABA2 belongs to a multigene family, and we identified 471 ABA2 genes from 19 species. The ABA2 evolutionary tree was constructed from the top five ABA2 genes of each species as an NCED phylogenetic tree, and LG + R 5 was the best-fit substitution model in IQTree for the construction of the ML phylogenetic tree of ABA2. The result showed that the classification of the phylogenetic tree was ambiguous because it did not cluster by plant taxa ( Figure S2A). Meanwhile, ScABA2 genes were always clustered together with different bryophyte species. Our conserved motif analysis showed that each ABA2 protein had an adh_short_C2, adh_short_C, and KR domain ( Figure S2B). In accordance with the phylogenetic and conserved domain results, ABA2 is very conserved in different plants.
An ML phylogenetic tree based on 20 ABA3 genes was constructed in IQTree using the best-fit substitution model (JTT + I + G 4 ), which was automatically selected by the software. As shown in Figure 4A, ScABA3 was sister clade to C. purpureus, B. argenteum, P. patens and S. fallax and had a close relation. The ABA3 proteins of three chlorophytes, S. moellendorffii, and M. polymorpha, clustered together in one clade, and they were closely related to each other. The ABA3 of angiosperm species were clustered together; a separate clade was formed for T. plicata (Thupl.29378725s0003), a gymnosperm, and another separate clade was formed for A. trichopoda (AmTr v1.0 scaffold00032), an angiosperm ( Figure 4A). The conserved motif analysis showed that ABA3 contained one MOSC, MOSC_N, Thioredoxin domain, and two Aminotran_5 domains in each plant taxon, except for B. argenteum, S. moellendorfii, and C. reinhardtii ( Figure 4B). Therefore, ABA3 is very conserved in the different plant taxa. clade was formed for T. plicata (Thupl.29378725s0003), a gymnosperm, and another separate clade was formed for A. trichopoda (AmTr v1.0 scaffold00032), an angiosperm ( Figure  4A). The conserved motif analysis showed that ABA3 contained one MOSC, MOSC_N, Thioredoxin domain, and two Aminotran_5 domains in each plant taxon, except for B. argenteum, S. moellendorfii, and C. reinhardtii ( Figure 4B). Therefore, ABA3 is very conserved in the different plant taxa. Finally, sixty-three AAO genes were identified for ML phylogenetic tree construction in IQTree using the best-fit substitution model (LG + F + R6). The result showed that the phylogenetic tree of AAO genes was different from other ABA biosynthesis genes because they did not cluster by plant taxa. However, we found that ScAAO proteins were always clustered together with C. purpureus AAO proteins, which were similar to other ABA biosynthesis genes ( Figure 5A). The conserved motif analysis suggested that each AAO protein had Fer2, Fer2_2, FAD_binding_5, Ald_Xan_dh_C, MocoBD1, MocoBD2, and CO_deh_flav_C domains ( Figure 5B). In consequence, the phylogenetic and conserved domain results confirmed that AAO is very conserved in the different plant taxa. Finally, sixty-three AAO genes were identified for ML phylogenetic tree construction in IQTree using the best-fit substitution model (LG + F + R 6 ). The result showed that the phylogenetic tree of AAO genes was different from other ABA biosynthesis genes because they did not cluster by plant taxa. However, we found that ScAAO proteins were always clustered together with C. purpureus AAO proteins, which were similar to other ABA biosynthesis genes ( Figure 5A). The conserved motif analysis suggested that each AAO protein had Fer2, Fer2_2, FAD_binding_5, Ald_Xan_dh_C, MocoBD1, MocoBD2, and CO_deh_flav_C domains ( Figure 5B). In consequence, the phylogenetic and conserved domain results confirmed that AAO is very conserved in the different plant taxa.

Structural Characterization of ABA Biosynthesis Genes in Various Plants
To discover the differences in ABA biosynthesis genes, we analyzed the structural characteristics of these genes among different plant taxa (Figures S3-S8) and compared the exon numbers in 19 plant species from different plant taxa ( Figure 6). Through analysis of the ABA1 gene structure, it was found that ScABA1 had two exons; unexpectedly, the ABA1 exon number in bryophytes had a great difference when compared with other plant taxa, as most of them only had one or two exons, such as S. caninervis, while M. polymorpha and S. fallax had 16 exons. In addition, gymnosperm, fern, lycophyte, and the majority of

Structural Characterization of ABA Biosynthesis Genes in Various Plants
To discover the differences in ABA biosynthesis genes, we analyzed the structural characteristics of these genes among different plant taxa (Figures S3-S8) and compared the exon numbers in 19 plant species from different plant taxa ( Figure 6). Through analysis of the ABA1 gene structure, it was found that ScABA1 had two exons; unexpectedly, the ABA1 exon number in bryophytes had a great difference when compared with other plant taxa, as most of them only had one or two exons, such as S. caninervis, while M. polymorpha and S. fallax had 16 exons. In addition, gymnosperm, fern, lycophyte, and the majority of angiosperm ABA1 genes had 16 exons. A. trichopoda ABA1 had 27 exons which was the most numerous, while P. patens and C. purpureus only had one exon, which were the least numerous. The ABA1 genes from three chlorophytes had 8-10 exons. Thus, the gene structure of ABA1 exhibited extreme variation (Figures 6 and S3). Through analysis of the ABA4 gene structure, we found that the majority of angiosperm species had six exons, bryophyte species had seven exons, and chlorophytes had 4-6 exons ( Figure S4). Overall, the variation in the ABA4 exon number was not obvious (Figure 6). NCED is a large gene family; we retained the top five NCED genes according to the E-value to further analyze the gene structure. Bryophyte NCED gene structures were different from other taxa because these genes differed greatly in their exon number. Some NCED genes had one or less than five exons; the others had 13-15 exons ( Figure S5). Taking S. caninervis, for example, the exon number of ScNCED genes was one (Sc_g07550), four (Sc_g09383), fourteen (Sc_g14527), fourteen (Sc_g11853), and thirteen (Sc_g07528), which shows a great variation. However, a similar gene structure was predicted between angiosperm species NCED genes composed by one exon, except that Zm00001d031086 had two exons in Z. mays; LOC_Os12g44310 had fourteen exons in O. sativa; Zosma01g05390 and Zosma01g01740 had two and fifteen exons in Z. marina; and AmTr_v1.0_scaffold00022 and AmTr_v1.0_scaffold00056 had fourteen and nine exons in A. trichopoda. At the same time, the NCED exon-intron structures of gymnosperm, fern, and lycophyte species were similar to those in angiosperm species. Distinctly, each chlorophyte NCED gene contained more than 10 exons ( Figure S5). Hence, NCED gene structures were varied in the different evolutionary branches ( Figure 6). Through analysis of the ABA2 gene structure, it was found that the ScABA2 exon number ranged from 1 to 6, and it was the same in each bryophyte. Most angiosperm, gymnosperm, fern, and lycophyte species contained two exons; a few of them differed, such as Zm00001d049277, which had three exons, while Potri.016G073800 had only one, Zosma03g33130 had eleven exons, and so on. The exon number of chlorophyte species ranged from 7 to 12 exons; consequently, chlorophytes had many more exons (Figures 6 and S6). In terms of the prediction result of the ABA3 gene structure, we found that the exon number of ABA3 genes was stabilized at 20-23 exons, with the exception of three chlorophyte species that contained 8-16 exons, which were obviously lower than in other plants (Figures 6 and S7). Based on the AAO gene structure, we revealed that the exon number showed a great difference in S. caninervis, ranging from 7 to 27, with a similar result in bryophytes. However, the majority of angiosperm, gymnosperm, fern, and lycophyte AAO genes had 10 exons. Moreover, the AAO genes of two chlorophyte species had thirty-two exons, which was the largest number, while Zm00001d019376 only had four exons, which was the lowest ( Figure S8). The exon number of AAOs had a great difference in each evolutionary branch, with huge variation (Figure 6). In conclusion, the results further indicated that the ABA biosynthesis gene structures were associated with plant taxa.
angiosperm ABA1 genes had 16 exons. A. trichopoda ABA1 had 27 exons which was the most numerous, while P. patens and C. purpureus only had one exon, which were the least numerous. The ABA1 genes from three chlorophytes had 8-10 exons. Thus, the gene structure of ABA1 exhibited extreme variation (Figures 6 and S3). Through analysis of the ABA4 gene structure, we found that the majority of angiosperm species had six exons, bryophyte species had seven exons, and chlorophytes had 4-6 exons ( Figure S4). Overall, the variation in the ABA4 exon number was not obvious (Figure 6). NCED is a large gene family; we retained the top five NCED genes according to the E-value to further analyze the gene structure. Bryophyte NCED gene structures were different from other taxa because these genes differed greatly in their exon number. Some NCED genes had one or less than five exons; the others had 13-15 exons ( Figure S5). Taking S. caninervis, for example, the exon number of ScNCED genes was one (Sc_g07550), four (Sc_g09383), fourteen (Sc_g14527), fourteen (Sc_g11853), and thirteen (Sc_g07528), which shows a great variation. However, a similar gene structure was predicted between angiosperm species NCED genes composed by one exon, except that Zm00001d031086 had two exons in Z. mays; LOC_Os12g44310 had fourteen exons in O. sativa; Zosma01g05390 and Zosma01g01740 had two and fifteen exons in Z. marina; and AmTr_v1.0_scaffold00022 and AmTr_v1.0_scaf-fold00056 had fourteen and nine exons in A. trichopoda. At the same time, the NCED exonintron structures of gymnosperm, fern, and lycophyte species were similar to those in angiosperm species. Distinctly, each chlorophyte NCED gene contained more than 10 exons ( Figure S5). Hence, NCED gene structures were varied in the different evolutionary branches ( Figure 6). Through analysis of the ABA2 gene structure, it was found that the ScABA2 exon number ranged from 1 to 6, and it was the same in each bryophyte. Most angiosperm, gymnosperm, fern, and lycophyte species contained two exons; a few of them differed, such as Zm00001d049277, which had three exons, while Potri.016G073800 had only one, Zosma03g33130 had eleven exons, and so on. The exon number of chlorophyte species ranged from 7 to 12 exons; consequently, chlorophytes had many more exons (Figures 6 and S6). In terms of the prediction result of the ABA3 gene structure, we found that the exon number of ABA3 genes was stabilized at 20-23 exons, with the exception of three chlorophyte species that contained 8-16 exons, which were obviously lower than in other plants (Figures 6 and S7). Based on the AAO gene structure, we revealed that the exon number showed a great difference in S. caninervis, ranging from 7 to 27, with a similar result in bryophytes. However, the majority of angiosperm, gymnosperm, fern, and lycophyte AAO genes had 10 exons. Moreover, the AAO genes of two chlorophyte species had thirty-two exons, which was the largest number, while Zm00001d019376 only had four exons, which was the lowest ( Figure S8). The exon number of AAOs had a great difference in each evolutionary branch, with huge variation (Figure 6). In conclusion, the results further indicated that the ABA biosynthesis gene structures were associated with plant taxa.

Chromosomal Locations and Collinearity Analysis of ABA Biosynthesis Genes in S. caninervis and P. patens
In order to understand the distribution traits of ABA biosynthesis genes, we chose S. caninervis [26] and another model moss, P. patens [34], as examples to evaluate the ABA biosynthesis gene locations in the chromosomes (Figure 7). In S. caninervis, a desiccationtolerant moss, ScABA1 and ScABA3 are each a single gene, both located on Chromosomes 10. In addition, ScABA4 genes are distributed on Chromosomes 10 and 12; ScNCED genes are distributed on Chromosomes 2, 5, 7, and 8; ScAAO genes are distributed on Chromosomes 1, 3, 6, and 11; and ScABA2 genes are distributed on every chromosome except for Chomosomes 8 and 13. In P. patens, a model moss plant, PpABA1 and PpABA3 are also each a single gene, distributed on Chromosomes 12 and 5, respectively; PpABA4 genes are distributed on Chromosomes 3 and 12; PpNCED genes are distributed on Chromosomes 6,12,16,18,21,22 and 25; and PpAAO genes are distributed on Chromosomes 2, 5, 19 and 20. The PpABA2 gene is a huge gene family, so it is distributed among half of all chromosomes.
Given that tandem duplicates play an important role in the creation of gene families, and NCED, ABA2, and AAO belong to a multigene family, we examined the tandem repeat clustering of ABA biosynthesis genes in two bryophyte genomes. Based on the results of ABA biosynthesis gene location, we found that two ScABA2 tandem repeat clusters (Sc_g14027/Sc_g14030, Sc_g03294/Sc_g03296) were located on Chromosome 3 and 11, one PpNCED tandem repeat cluster (Pp3c25_4816/Pp3c25_4810) and one PpABA2 tandem repeat cluster (Pp3c18_20230/Pp3c18_20320) located on Chromosomes 25 and 18 ( Figure 7). However, no tandem repeat cluster was found for AAO gene families; these genes were located on different chromosomes in both bryophytes. In conclusion, the ABA biosynthesis genes were evenly distributed and dispersed on the chromosomes of S. caninervis and P. patens. Furthermore, Chromosomes 27 and 13 are sex chromosomes in S. caninervis and P. patens, respectively, which similarly did not host the ABA biosynthesis genes. There was a high similarity in the distribution of ABA biosynthesis genes between S. caninervis and P. patens. Therefore, we further analyzed the collinearity of ABA biosynthesis genes in the genomes of the two bryophyte species.
Comparison of the ABA biosynthesis genes between S. caninervis and P. patens can provide insights into the evolutionary history of ABA (Figure 8). It was found that nine ABA biosynthesis genes were homologous between the two bryophytes. In S. caninervis, the ScABA1 gene can be found; as in P. patens, ABA1 as a single gene has a significantly collinear relationship among bryophyte species. Among six NCED genes in the S. caninervis genome, only one ScNCED (Sc_g14527) gene has homologous genes in P. patens. We found that seven ABA2 genes had homologous genes between S. caninervis and P. patens. This result suggested that ABA1, NCED, and ABA2 genes may be generated from the same ancestral gene and have the same gene function. However, the ABA4, ABA3, and AAO genes of the two bryophyte species were not homologous, and it was speculated that these genes are more likely to be an independent event in the evolution of bryophytes.

Expression Analysis of ABA Biosynthesis Genes in S. caninervis under Different Stresses
ABA plays a key role in many types of abiotic stress in plants. Based on the information analysis results, we detected the expression patterns of ABA biosynthesis genes in the resistance model moss S. caninervis under abiotic stress. It is very important to analyze the expression of ABA biosynthesis genes under dehydration, salt, cold, heat, and ABA stress ( Figure 9). Under dehydration stress, the expression of ScABA1 rapidly increased approximately 20-fold and was sustained, except for 6 h and 24 h; ScABA4 had a slight fluctuation within 2-fold, ScNCED was down-regulated, ScABA2 slightly increased between 2-and 3-fold, ScABA3 was significantly up-regulated at the early stage, and ScAAO rapidly increased 5-10-fold and then declined to low abundance at 24 h. Under salt stress, the ScABA1 transcript abundance gradually accumulated to 38-fold before 2 h and decreased to 21-31-fold; ScABA4 and ScABA2 had a slight alteration within 2-3-fold, ScNCED was barely changed, ScABA3 was expressed at the late stage at approximately 4-6-fold, ScAAO rapidly increased 9-fold in 1 h but subsequently declined. Under cold stress, ScABA1 maintained high expression from 14-to 62-fold and peaked at 12 h; ScABA4 sustained low expression and peaked at 24 h at approximately 5-fold; ScNCED was almost unchanged and showed negligible levels at 12 h; ScABA2 was up-regulated and peaked at 12 h; ScABA3 had a high expression at 1 h, 12 h, and 24 h, ScAAO was rapidly increased and sustained at 6-10-fold. Under heat stress, ScABA1 was rapidly and constantly upregulated from 16-to 30-fold; the transcript abundance of ScABA4, ScABA2, and ScABA3 increased but not significantly; ScNCED significantly decreased; ScAAO rapidly increased to 8-fold and then slightly declined to 5-6-fold. Under ABA stress, ScABA1 gradually increased to 8-fold at 12 h; the expression of ScABA4, ScABA3, and ScAAO was at negligible levels; ScNCED was up-regulated before 1 h but subsequently declined to low abundance, and ScABA2 had high expression at 24 h at approximately 6-fold.

Expression Analysis of ABA Biosynthesis Genes in S. caninervis under Different Stresses
ABA plays a key role in many types of abiotic stress in plants. Based on the information analysis results, we detected the expression patterns of ABA biosynthesis genes in the resistance model moss S. caninervis under abiotic stress. It is very important to analyze the expression of ABA biosynthesis genes under dehydration, salt, cold, heat, and ABA stress ( Figure 9). Under dehydration stress, the expression of ScABA1 rapidly increased approximately 20-fold and was sustained, except for 6 h and 24 h; ScABA4 had a slight fluctuation within 2-fold, ScNCED was down-regulated, ScABA2 slightly increased between 2-and 3-fold, ScABA3 was significantly up-regulated at the early stage, and ScAAO rapidly increased 5-10-fold and then declined to low abundance at 24 h. Under salt stress, the ScABA1 transcript abundance gradually accumulated to 38-fold before 2 h and decreased to 21-31-fold; ScABA4 and ScABA2 had a slight alteration within 2-3-fold, ScNCED was barely changed, ScABA3 was expressed at the late stage at approximately 4-6-fold, ScAAO rapidly increased 9-fold in 1 h but subsequently declined. Under cold stress, ScABA1 maintained high expression from 14-to 62-fold and peaked at 12 h; ScABA4 sustained low expression and peaked at 24 h at approximately 5-fold; ScNCED was almost unchanged and showed negligible levels at 12 h; ScABA2 was up-regulated and peaked at 12 h; ScABA3 had a high expression at 1 h, 12 h, and 24 h, ScAAO was rapidly increased and sustained at 6-10-fold. Under heat stress, ScABA1 was rapidly and constantly up-regulated from 16to 30-fold; the transcript abundance of ScABA4, ScABA2, and ScABA3 increased but not significantly; ScNCED significantly decreased; ScAAO rapidly increased to 8-fold and then slightly declined to 5-6-fold. Under ABA stress, ScABA1 gradually increased to 8-fold at 12 h; the expression of ScABA4, ScABA3, and ScAAO was at negligible levels; ScNCED was up-regulated before 1 h but subsequently declined to low abundance, and ScABA2 had high expression at 24 h at approximately 6-fold.

Discussion
With the continuous development of sequencing technology, the genomes of bryophytes and chlorophytes have been published rapidly [26,[34][35][36], and these lay a solid foundation for further analysis and research on the origin of ABA in early-diverging plant species and also provide sufficient information for the study of the ABA biosynthetic pathway. In this study, we analyzed ABA biosynthesis genes and, for these genes, identified their evolutionary relationships, conserved domain, gene structure, chromosome location, collinearity, and expression pattern in S. caninervis and 18 representative plants from different plant taxa for the first time.
Previous studies have shown that S. moellendorffii, P. patens, and C. reindardtii, the model plant of lycophytes, bryophytes, and chlorophytes, respectively, do not contain ABA2 and AAO genes, suggesting that ABA2 and AAO appear only in angiosperms; thus, they speculated upon the presence of an alternative enzymatic pathway to convert xanthoxin to ABA [19]. In addition, little attention has been paid to the ABA biosynthetic pathway in chlorophytes; only ABA1, ABA4, and ABA3 genes were previously reported

Discussion
With the continuous development of sequencing technology, the genomes of bryophytes and chlorophytes have been published rapidly [26,[34][35][36], and these lay a solid foundation for further analysis and research on the origin of ABA in early-diverging plant species and also provide sufficient information for the study of the ABA biosynthetic pathway. In this study, we analyzed ABA biosynthesis genes and, for these genes, identified their evolutionary relationships, conserved domain, gene structure, chromosome location, collinearity, and expression pattern in S. caninervis and 18 representative plants from different plant taxa for the first time.
Previous studies have shown that S. moellendorffii, P. patens, and C. reindardtii, the model plant of lycophytes, bryophytes, and chlorophytes, respectively, do not contain ABA2 and AAO genes, suggesting that ABA2 and AAO appear only in angiosperms; thus, they speculated upon the presence of an alternative enzymatic pathway to convert xanthoxin to ABA [19]. In addition, little attention has been paid to the ABA biosynthetic pathway in chlorophytes; only ABA1, ABA4, and ABA3 genes were previously reported [37]. On the whole, there is a poor understanding of the ABA biosynthetic pathway in early-divergence plants. Thanks to the release of more genomes, we first found that S. caninervis contained a complete set of ABA biosynthesis genes, and this was consistent in other bryophyte species ( Figure 1B). Unexpectedly, three chlorophyte species had all of the ABA biosynthesis genes too. Similar to past research, our results showed that the ABA1 and ABA3 genes are a single gene in bryophytes and chlorophytes, as in angiosperms [7,14]. Compared with higher plants, chlorophyte plants had many more ABA4 genes. On the contrary, the quantity of NCED, ABA2, and AAO genes was much lower in chlorophytes, which suggested that these genes perhaps increased during evolution ( Figure 1B). In total, both lower and higher plants had ABA biosynthesis genes, and the number of ABA biosynthesis genes had a close relationship with the plant taxa; simultaneously, the results further indicated that the ABA biosynthetic pathway was conserved in the plant kingdom.
Previous reports revealed that the ABA1, ABA4, and NCED genes were conserved in angiosperm species by domain analysis [37,38]. In S. caninervis, we found that the conserved domain of the ABA biosynthesis genes was the same as in the rest of bryophytes, chlorophytes, and other plant taxa; for example, ScABA4 proteins also had the DUF4281 domain (Figure 3), and ScNCED proteins also had the RPE65 domain ( Figure S1). Therefore, this result provided evidence that the domain of ABA biosynthesis genes was highly conserved in both lower and higher plants. On the basis of phylogenetic tree analysis, the single genes, such as ABA1, ABA4, and ABA3, were clearly clustered together by plant taxa (Figures 2-4), while NCED, ABA2, and AAO belonged to a multigene family and their classification in the phylogenetic tree was ambiguous (Figures 5, S1 and S2). Besides the AAO phylogenetic tree, ABA biosynthesis genes are closely associated with plant taxa.
In the highly conserved ABA biosynthesis genes, gene structure had unanticipatedly large changes, as indicated by the wide range of exon numbers between the plant taxa, with the exception of ABA4 (Figures 6 and S4). The exon number of ABA1 and ABA3 genes increased during plant evolution ( Figures S3 and S7); conversely, NCED, ABA2, and AAO had far fewer exons in higher plants ( Figures S5, S6 and S8). In addition, bryophytes proved to be a turning point in the variation in exon number; for example, the majority of angiosperm, gymnosperm, fern, and lycophyte NCED genes were composed of one exon, and chlorophyte species had more than 10 exons; while some bryophyte species had one exon, the rest of the bryophytes had 13-15 exons ( Figure S5). According to the ABA biosynthesis gene structure analysis, it can be concluded that bryophytes were the transitional form in plant evolution; in line with past research, bryophytes were the earliest diverging lineages of the extant land plants and the most original group of existing higher plants [39,40]. Consequently, the differences in ABA biosynthesis gene structures were affected by plant taxa.
On the basis of the analysis of the ABA biosynthesis genes distribution in S. caninervis and P. patens, these genes were evenly distributed and dispersed on the chromosomes, but they were not located on sex chromosomes in the two bryophytes (Figure 7). Through collinearity analysis, ABA1, NCED, and ABA2 genes showed homologous genes between the S. caninervis and P. patens genomes (Figure 8). This provides new evidence that the ABA biosynthetic pathway originated from the same ancestral gene and shared the same stress resistance function.
When organisms start to colonize terrestrial habitats, endogenous ABA is increased even under mild drought stress; then, desiccation-protecting mechanisms are stimulated [15,41]. In this study, an extremely desiccation-tolerant moss was chosen as a representative to examine whether the ABA biosynthesis genes of bryophytes respond to abiotic stress. In S. caninervis, all ABA biosynthesis genes changed their expression under five types of treatments to cope with abiotic stress (Figure 9). Moreover, the expression of ScABA1 concerning abiotic stress was more profound than that of other genes. However, it is worth noting that ABA synthesis genes had lower expression levels under 10 µM ABA than other treatments, except for the ScNCED and ScABA2 genes. Based on the transcriptional expression profiles of the ABA biosynthesis genes under different stresses, ABA could play a key regulatory role in abiotic stress in desiccation tolerance moss, as shown in a previous study [33].
In this article, we focused on the ABA biosynthesis genes in early-diverging plants, especially bryophytes. Although the number and structure of ABA biosynthesis genes were affected by plant taxa, these genes existed and were intact in both lower and higher plants.
Our results provided solid evidence that the ABA biosynthesis genes were conserved in the plant kingdom, which is critical in understanding the role of ABA in early-diverging plant species. On the other hand, ABA biosynthesis genes are involved in responding to all types of abiotic stress, which suggests that ABA plays an important role in responding to abiotic stress in bryophytes. While bryophytes such as S. caninervis gametophytes do not have a stoma, how does the ABA biosynthetic pathway participate in the stress resistance of bryophytes? The data are currently scarce, and more work is required to investigate ABA biosynthesis in bryophytes in the future.

Phylogenetic Analysis and Conserved Domain Prediction
For phylogenetic tree reconstruction, the protein sequences were aligned with the MUSCLE application in MEGA 7. Maximum likelihood (ML) analysis was constructed in IQTree v1.6.12 [59] using the best-fit substitution model automatically selected by the software according to the Bayesian information criterion scores and weights (BIC) with partitions [60], and an ultrafast bootstrap (UFB) with 1000 replicates. For conserved structural motifs, the protein sequences were identified using the PFAM website (http:// pfam.xfam.org/ (accessed on 22 January 2023)) with default settings. Next, the phylogenetic tree and conserved domain were trimmed and displayed using the EvolView website (https://www.evolgenius.info/evolview/#/ (accessed on 22 January 2023)).

Analysis of ABA Biosynthesis Gene Structure
To study the structural evolution of ABA biosynthesis genes, the regions of exons and introns were predicted in all ABA biosynthesis genes. For this comparative genomic analysis, the genomes and annotation files were retrieved from Phytozome. The ABA biosynthesis gene structures were shown through the Biosequence Structure Illustrator of TBtools v1.106 (https://github.com/CJ-Chen/TBtools/ (accessed on 22 January 2023)) [61].

Chromosome Location and Collinearity Analysis
The chromosome distribution of ABA biosynthesis genes was drawn with TBtools according to the location information from 19 plant genome databases [61]. MCScanx software was used to analyze the collinearity of ABA biosynthesis genes among P. patens and S. caninervis [26,34], and then the collinearity diagram was obtained through the Multiple Synteny Plot of TBtools.

Gene Expression Analysis of ABA Biosynthesis Genes under Abiotic Stress
Dry Syntrichia caninervis Mitt. gametophytes were collected from the Gurbantunggut Desert in Xinjiang, China (44 • 32 30 N, 88 • 6 42 E). Dry gametophytes were fully hydrated on filter paper saturated with sterile water in glass Petri dishes for 24 h at 25 • C, with light at a photosynthetic photon flux density (PPFD) of 100 µmol/m 2 /s, prior to the abiotic stress treatments, including dehydration, salt, cold, heat and ABA, and then the samples were collected at 0 h, 0.5 h, 1 h, 2 h, 6 h, 12 h, and 24 h in each treatment. For dehydration treatment, the fully hydrated gametophytes were slow dried in a closed desiccator with saturated sodium nitrite at room temperature (67% relative humidity, 25 • C). For salt and ABA treatment, the fully hydrated gametophytes were transferred to new Petri dishes with filter paper saturated with 8 mL 150 mM NaCl or 10 µM ABA solutions at room temperature (30% relative humidity, 25 • C). For cold and heat stress treatments, the fully hydrated gametophytes were placed in Petri plates on water-saturated filter paper and incubated at either 4 • C or 42 • C.
The gametophytes of each treatment at different timepoints were collected and frozen in liquid nitrogen for RNA extraction. Total RNA was isolated using the MiniBEST Plant RNA Extraction Kit (TaKara, Dalian, China) according to the manufacturer's protocol. The cDNA was synthesized from 1 µg of total RNA using the PrimeScript RT reagent kit (Takara, Dalian, China). Gene expression was performed in 96-well plates with the CFX96 Real-Time PCR Detection System (Bio-Rad, USA), with three technical replicates and two biological replicates [28]. α-TUB2 was used as an internal control [62], and the relative expression levels of ABA biosynthesis genes (ScABA1/Sc_g04625, ScABA4/Sc_g04273, ScNCED/Sc_g07550, ScABA2/Sc_g02085, ScABA3/Sc_g05025, ScAAO/Sc_g15857) were calculated by the 2 −∆∆CT method [63], and all primers are displayed in Table S2. Data are expressed as the mean ± SD. Error bars represent standard deviations.

Conclusions
In this study, we performed a comprehensive analysis of ABA biosynthesis genes, including identification, evolutionary relationships, gene and protein structures, and expression patterns. Our results demonstrated that S. caninervis had complete ABA biosynthesis genes, and we first found that chlorophytes had all of these genes. Although the ABA biosynthesis genes of chlorophytes were separated from those of other plant taxa in the phylogenetic tree, the conserved domain was the same in all plants. Furthermore, we found that the ABA biosynthesis gene structures showed huge variation between different evolutionary branches. Noticeably, ABA biosynthesis genes were not located on sex chromosomes and were evenly distributed and dispersed on the other chromosomes. Simultaneously, ABA1, NCED, and ABA2 genes had collinearity between the S. caninervis and P. patens genomes. In S. caninervis, all of the ABA biosynthesis genes responded to abiotic stress, and ABA also plays an important role in the stress response. In conclusion, this study not only identified the ABA biosynthesis genes in S. caninervis and other representative plants from different plant taxa but also provided sufficient evidence confirming that the ABA biosynthetic pathway was conserved in the evolutionary process.
Supplementary Materials: The following supporting information can be downloaded at: https://www.mdpi.com/article/10.3390/plants12051114/s1, Figure S1: Phylogenetic and deduced protein structure analysis of NCED; Figure S2: Phylogenetic and deduced protein structure analysis of ABA2; Figure S3: The intron/exon structure of ABA1 genes; Figure S4: The intron/exon structure of ABA4 genes; Figure S5: The intron/exon structure of NCED genes; Figure S6: The intron/exon structure of ABA2 genes; Figure S7: The intron/exon structure of ABA3 genes; Figure S8: The intron/exon structure of AAO genes. Table S1: The BLASTP screen result of ABA biosynthesis gene sequences; Table S2: Primer information for RT-qPCR of ABA biosynthesis genes.  Data Availability Statement: All of the accession numbers of ABA biosynthesis genes in this study can be found in Table S1.