Transcriptomic Regulation of Aquaporins During Seed Germination in the Marine Seagrass Cymodocea nodosa

TL;DR

This study explores how aquaporin genes are regulated during seed germination in the marine seagrass Cymodocea nodosa, revealing their role in water transport and germination progress.

Contribution

The study provides one of the first insights into aquaporin function during seed germination in marine seagrasses.

Findings

PIP, TIP, and SIP aquaporin transcript levels show differential expression during germination stages.

Expression changes are linked to activation responses related to germination progress.

C. nodosa is positioned as a model for studying aquaporin regulation by environmental factors.

Abstract

Seed germination is a key phase that transitions the seed from dormancy to active growth, where imbibition emerges as the initial event, followed by aquaporin-mediated regulation of cellular water that supports metabolic reactivation under favourable conditions. Aquaporins are small integral membrane proteins that facilitate the passive transport of water and small solutes across membranes and play key roles in plant development and physiology. In terrestrial plants, aquaporins are classified into five main types—PIPs, TIPs, NIPs, SIPs, and XIPs—with PIPs and TIPs being the most abundant and widely expressed. Whilst knowledge of seagrass aquaporins and their physiological roles remains limited, their functional involvement in seed germination is largely unknown. In this study of the marine seagrass Cymodocea nodosa, transcriptome assembly and analysis enabled the identification of aquaporin-encoding sequences, which were subsequently used for gene expression analysis during germination. Three well-defined seed stages of C. nodosa were analysed to assess expressions patterns of PIPs, TIPs, NIPs, and SIPs. Results showed that transcript levels of PIP, TIP, and SIP aquaporins exhibit differential expression patterns, and the changes are mainly associated with activation responses related to seed germination progress. Overall, this study provides one of the first foundations for investigating aquaporin function during seed germination in marine seagrasses, defining expression patterns across imbibition and early germination, and positioning C. nodosa as a valuable in vitro model for exploring the regulation of aquaporin activity by environmental and physiological factors.

Figures

Full text

1. Introduction

Seed development is a critical process involving coordinated physiological, biochemical, and molecular events through which a fertilized ovule develops into a mature, viable seed capable of germinating under favourable conditions. This process is tightly regulated by hormonal balance and signalling, particularly by abscisic acid (ABA) and gibberellins (GAs). During seed germination, water uptake, or imbibition, represents the first step of the process, as storage reserves are made available through hydrolytic enzyme activation [1,2]. The inward and outward movement of water within and across seed tissues can be modulated by the regulated activity of aquaporins, among other physical and physiological factors. In terrestrial plants, aquaporins (AQPs) play important roles in water transport and cell membrane permeability, processes essential for seed development and germination, as well as for transpiration and water movement in roots, stems, and leaves [3].

Aquaporins are a family of small (24–30 kDa) integral membrane proteins that facilitate the passive influx and efflux of water and small solutes across biological membranes in plants. A plethora of aquaporin isoforms have been identified and localised in different tissues, indicating their roles in water transport and in developmental and physiological processes. Plant aquaporins are classified into five major groups: plasma membrane intrinsic proteins (PIPs), tonoplast intrinsic proteins (TIPs), nodulin 26-like intrinsic proteins (NIPs), small basic intrinsic proteins (SIPs), and X-intrinsic proteins (XIPs) [4,5,6]. Moreover, aquaporins often display dual subcellular localisation and are grouped into several subfamilies, among which PIPs and TIPs are the most abundantly expressed [7].

Molecular mechanisms underlying aquaporin function in seagrasses remain largely unexplored. Available evidence, which is primarily derived from studies of vegetative tissues, suggests that aquaporins facilitate the transport of water and small solutes, a process essential for osmotic regulation and acclimation to marine conditions. Furthermore, genomic and expression studies have revealed multiple aquaporin subfamilies that play tissue- and condition-specific roles [8]. In this context, transcriptome annotation provides a comprehensive catalogue of expressed genes and their functions, serving as a foundation for transcriptomic approaches to identify regulatory pathways involved in dormancy release, water uptake, and metabolic activation during seed germination in terrestrial plants [9]. Such approaches are particularly valuable in marine angiosperms, whose genomes remain poorly characterised compared with those of terrestrial plants [10]. Transcriptomic datasets further allow for the exploration of the expression dynamics of the stress-related genes, aquaporins, and enzymes involved in reserve mobilisation. They also provide a conceptual framework—largely established in terrestrial systems—for interpreting molecular processes underlying seed rehydration, enzyme activation, and cellular expansion during germination [11,12]. Together, this knowledge serves as a foundation for extending molecular analyses to seed physiology in marine angiosperms, including analyses of gene regulatory networks, aquaporin function, hormonal control, and the timing of key cellular events underlying successful germination.

Cymodocea nodosa is a marine angiosperm that belongs to the monocot clade (order Alismatales). It produces dormant seeds which germinate and establish seedlings once dormancy is broken. Dormancy release is associated with water uptake and metabolic reactivation, leading to germination and seedling establishment [13,14]. Seed germination in C. nodosa progresses through three stages, namely, Stage 0, corresponding to the closed seed; Stage I, marked by rupture of the dorsal crest at the extreme, which represents the first sign of germination; and Stage II, characterised by the protrusion of the cotyledon and subsequently, the radicle, indicating completion of the process [13]. Germination is assumed to occur through gradual imbibition leading to the rehydration of the endosperm and the activation of metabolic processes. These events, in turn, promote vacuolisation, marking a transition from storage of reserve substances to mobilisation and cellular expansion. The developmental stages are regulated by intense metabolic activity, which provides energy for the mobilisation of stored reserves through enzymatic hydrolysis and produces sugars that sustain the early stages of germination [13].

Given that water uptake is a central driver of the metabolic and structural transitions that characterise seed germination, understanding the molecular regulation of water movement is essential to elucidate the mechanisms underlying dormancy release and early seed germination. The present work analyses, through droplet digital PCR, the gene expression of four groups of aquaporins annotated in the transcriptome of C. nodosa to shed light on molecular mechanisms regulating dormancy breakage and germination across the three defined stages of the seed.

2. Results and Discussion

2.1. Data Set of Transcriptomes

Three transcriptomes corresponding to Stage 0, Stage I, and Stage II were independently assembled and analysed, and a merged transcriptome was generated to make possible comprehensive comparative analyses across seed stages.

Sequencing yielded more than 4 Gbp (4,163,963,512 bp) of high-quality data, with an average read length of approximately 1800 bp and a mean read quality of Q41, corresponding to an expected error rate of one per 12,500 bases (Table 1). The metrics indicate excellent data quality suitable for reliable transcriptome assembly, isoform annotation, gene expression profiling, and functional analyses, including the identification of aquaporins [15,16,17,18,19].

The Full-Length Non-Concatemer (FLNC) reads, clustering, and isoform generation indicated optimal values for the analysis of the transcriptome of each stage and the merged transcriptome. The number of FLNC reads was consistent across stages, reflecting a relatively balanced contribution from all three stages in terms of both read counts and high-quality (HQ) isoform numbers, with no evident bias toward any stage (Table 2) [20,21,22,23]. The mean length of the FLNC reads was 1690 bp. The high number of HQ isoforms per sample (stage) indicates a highly comprehensive functional representation. In the merged transcriptome analysis, both HQ and low quality (LQ) isoforms accounted for a total of more than 238,000 isoforms. After removing redundancies, the final number of unique isoforms was 123,024 (Table 2).

BUSCO analysis yielded very good results for the transcriptome of the three stages and the merged transcriptome, showing very high percentages of complete and essential genes, minimal gene fragmentation, and relatively low proportions of single-copy or missing genes (Table 3). The high duplication rate indicates the presence of many isoforms and splicing variants, which is likely associated with increased alternative splicing or differential gene expression during the early germination stages [24,25,26,27].

2.2. Identification of Aquaporins Sequences

A total of 18 full length aquaporin-encoding genes were identified in the transcriptome of Cymodocea nodosa seeds from a total of 88 complete aquaporin transcripts (Figure 1). The number of aquaporins identified in C. nodosa is consistent with that in other studies, as it can range from 15–30, depending on tissue, physiological state, and assembling quality in monocotyledonous and other angiosperms [28]. Moreover, evidence show that, although angiosperm genomes contain dozens of aquaporin genes, only a small subset is significantly expressed in seeds during maturation and germination [11,29,30,31,32]. Thus, in terrestrial plants, the Oryza sativa genome encodes 33 aquaporins, of which 10 to 12 showed significant expression during germination [30], whereas 35 aquaporin have been identified in the genome of Arabidopsis thaliana (Table 4) [9,32]. Regarding aquaporin type, expression during seed germination in A. thaliana is predominantly associated with TIP isoforms, whereas PIP aquaporins generally show lower transcript abundance [32,33].

In the marine angiosperm seagrass Zostera marina, a genomic and transcriptomic study of the leaves, roots, and flowers identified 25 aquaporin genes in its genome. Through tissue-specific RNA-Seq, 24 of these 25 genes showed expression in the roots, vegetative tissue (leaves), and male and female flowers [8].

All full-length aquaporin-encoding genes identified in C. nodosa (18 sequences) retrieved from the UniProt database, of which 10 correspond to PIP, 4 to SIP, 3 to TIP, and 1 to NIP, together with 70 additional aquaporin-encoding genes from Alismatales, the same order to which C. nodosa belongs, were used for phylogenetic analysis to provide an appropriate phylogenetic framework for comparative analysis. The C. nodosa nucleotide sequences were translated into their corresponding amino acid sequences prior to analysis. A matrix of 86 aa sequences and 1161 positions was obtained after adding the three GPLs outgroup sequences and removing identical sequences using ALTER (Supplementary Tables S1 and S2). A minimum bootstrap support threshold of 70 was applied to assess the reliability and robustness of the maximum-likelihood analysis.

The maximum-likelihood analysis revealed relationships between C. nodosa aquaporins and aquaporin proteins from other Alismatales species. Cymodocea nodosa amino acid sequences clustered with predicted amino acid sequences corresponding to the NIP, PIP, TIP, and SIP aquaporin subfamilies retrieved from UniProt, with bootstrap support values ranging from 98 to 100 (Figure 1). The presence of these four subfamilies has been recognized in monocots and Brassicaceae [8]. As plant aquaporin subfamilies are typically represented by multiple gene copies, the sequence sets corresponding to PIPs, TIPs, SIPs, and NIPs identified in the Cymodocea nodosa transcriptome were considered to likely represent distinct aquaporin gene members. In contrast, aquaporin isoforms derived from alternative splicing or post-transcriptional variation were not addressed in this study, as subclade assignments may be influenced by inconsistent or provisional annotations (e.g., uncharacterised or putative aquaporins; Supplementary Table S1). Consistent with this approach, functional relationships among individual aquaporin isoforms remain poorly resolved and were therefore beyond the scope of the present work [35]. Overall, aquaporins expressed in C. nodosa seeds show a conserved subfamily composition and phylogenetic organisation typical of angiosperms.

2.3. Gene Expression of Aquaporins During Seed Germination

Seed germination in C. nodosa depends on water uptake to initiate metabolic processes, while dormancy release, when present, is required to permit germination [13,36]. Dormancy is considered, in functional terms, as a reversible state released prior to water uptake and metabolic activation. Water diffusion across cell membrane is facilitated by aquaporins, which are integral membrane proteins in charge of transporting water, as well as small and neutral solutes. A genome study in the eelgrass Zostera marina revealed 25 aquaporins, which were grouped in four subfamilies (NIP, TIP, SIP and PIP), distributed in roots, vegetative tissue, and flowers [8]. Beyond recognition of aquaporin types, it is a fact that nothing is known about the functional role of aquaporins during the development of marine seagrasses, despite its importance for seed survival. In line with this, antibody-based studies of aquaporins in Posidonia have primarily been linked to the regulation of water transport rather than to water-driven processes underlying seed germination [37,38]. Here, time-course analysis of C. nodosa seed germination reveals dynamic changes in the transcript abundance of PIP, TIP, and SIP aquaporins across germination stages, from the closed seed stage, through rupture of the dorsal crest as the first visible sign of germination, to cotyledon and subsequent radicle protrusion, representing the final stage of the germination sequence. In C. nodosa, one gene sequence of each aquaporin type, namely NIP, TIP (TIP2) and PIP (PIP4), is downregulated, while other gene sequences belonging to the PIP-, SIP- and TIP- aquaporin type are upregulated towards the end of the germination period (Stage II; Figure 2). The observation that aquaporin sequences exhibit significant differential expression patterns suggests that they may play roles in seed imbibition and in subsequent metabolic activation processes accompanying early seed germination in C. nodosa. This finding aligns with the known ability of aquaporins to transport a variety of solutes beyond water.

Changes in transcript abundance, specifically between two TIPs (upregulated TIP 1 and downregulated TIP2) and among the five PIP contigs (with only PIP4 being downregulated; Figure 2A,B) may be associated with physiological changes accompanying early seed germination. Beyond seed germination, other studies have reported differential PIP expression in angiosperm species, indicating subtle interactions between aquaporins and their water-channel activities [39]. Regarding SIPs, they are likely to fulfil comparable roles during seed germination, as their transcript levels increase by approximately twofold (Figure 2C). This is consistent with the pattern observed in plants, in which aquaporin transcript abundance rises toward the final stages of seed development, coinciding with increased metabolic activity [12]. In terrestrial plants, including Arabidopsis, SIPs help maintain water balance, whilst NIPs generally show low expression levels and are associated with floral development and nitrogen-fixing root nodules [7,40]. Furthermore, NIPs have been related to transient expression in specialised organs involved in the transport of small solutes [41]. In C. nodosa, the ca. 50% downregulation of NIP transcripts in Stage II seeds (10 transcript copies × μL^−1^) may represent a mechanism to limit inward water uptake once dormancy has been broken. These results may be explained by considering nitrogen acquisition in C. nodosa and other marine species. Unlike terrestrial plants, nitrogen acquisition in these species is mediated by associated microorganisms in the leaves rather than through root nodules [42]. Nonetheless, NIP expression in terrestrial plants can vary among tissues and development stages depending on their transport substrates, leading to their classification into three types (NIP-I, -II and -III) [43]. In C. nodosa, these potential NIP types could not be clearly distinguished in the phylogenetic analysis and thus appear as a single NIP sequence, possibly explaining their differential gene expression pattern. This observation is consistent with findings from the Zostera species, where no NIP sequences were identified [8].

Gene expression of aquaporins can be differentially regulated in a myriad of processes that accompany germination, such as fusion of protein storage vacuoles into large central lytic vacuoles, which contributes to increased turgor pressure and cell elongation [28,44]. Correlation between the expression of specific aquaporin types and the presence of distinct vacuole types has been reported in orthodox seeds such as Arabidopsis [45], rice [46], and broad beans [47], as well as in the recalcitrant seeds of the horse chestnut [48]. Water influx into the cell is primarily facilitated by PIPs as they act like water channels, while excess water is transported into the vacuole through tonoplast aquaporins (TIPs). TIPs not only regulate cell turgor pressure via controlled gating [7], but they are also thought to be involved in the maintenance of reactive species homeostasis during seed development, as degradation of polysaccharides, lipids and proteins is assumed to occur [49].

In C. nodosa, the presence of aquaporin-encoding transcripts was observed to be abundant throughout three stages of seed germination, with significantly higher levels during cotyledon emergence and elongation (Stage II) compared to those in seeds displaying an opened dorsal crest at the extreme end (Stage I) of C. nodosa. This increase ranges from 1.5 to 170-fold in transcript abundance in Stage II for different PIP contigs (Figure 2B). For the two TIP contigs, the transcript levels varied ca. 3-fold across the stages, showing an increase in TIP1 expression and a decrease in TIP2 as seeds germinate (Figure 2A). High transcript abundance of PIP and TIP aquaporins during the three seed germination stages of C. nodosa is consistent with their key roles in water balance reported for other plant physiological processes and also reflects the high evolutionary conservation of these aquaporin families in plants [50]. Taken together, these observations suggest that the expression of PIP and TIP aquaporins is coordinated and stage-dependent during seed germination in C. nodosa.

Additionally, it has been suggested that the process of seed germination involves changes in vacuolar remodelling. This includes the predominance of storage vacuoles in the early stages and the presence of lytic vacuoles during cotyledon emergence. These structural changes are accompanied by regulated water fluxes, suggesting that aquaporin function is modulated at multiple regulatory levels during germination. Plant growth regulators and post-translational modifications, including those influenced by ions such as Ca^2+^, can modulate aquaporin activity during germination [7]. Water uptake-related signalling is therefore regulated by both hormonal and physical cues, contributing to seed imbibition and dormancy release. At the transcriptional level, aquaporin expression has been reported to be regulated by abscisic acid (ABA), which inhibits amylase activity, and by gibberellins (GAs), which promote it [51]. Furthermore, aquaporin gene expression can be influenced by nutritional cues, such as nitrogen deficiency [52], and environmental factors, including light and temperature [53,54]. These observations indicate that seed germination is regulated by an interplay of physical, hormonal and metabolic cues, in addition to water transport.

Consistent with the aforementioned, metabolite analyses of C. nodosa seeds have revealed the presence of an inhibitor of brassinosteroid synthesis and pyrrolidines, the latter being degradation products of polyamines [55]. Polyamines are known to play roles during the growth and development of C. nodosa and have been described as an anti-senescence inducing factor [56]. Together, these findings point to the involvement of metabolic and hormone-related processes during dormancy release and early germination, acting alongside water uptake-associated cellular changes. Accordingly, although results suggest that different PIPs are involved in C. nodosa germination and that TIPs may regulate turgor pressure and degradation processes, it remains difficult to determine the precise time during germination when aquaporin functionality corresponds to vacuolization changes, embryo water uptake capacity, or hormonal control mechanisms. Aside from the temporal-expression pattern, it also can also be suggested that variations in aquaporin transcript abundance may also reflect spatial distribution differences, as previously reported among seeds, cotyledon and roots tissues [57].

In summary, the differential expression of aquaporins across germination stages indicates their involvement in water-related processes during dormancy release and early seed germination. Moreover, changes in the expression of PIP and TIP transcripts, coinciding with cotyledon emergence and radicle protrusion, suggest stage-dependent modulation of aquaporin expression and provide a molecular framework for interpreting processes related to water transport during germination in seagrass seeds. Further studies could explore aquaporin regulation in response to varying salinity levels, particularly considering that C. nodosa seed germination predominantly occurs in vitro at 15 psu [13].

3. Materials and Methods

3.1. Defined Stages of Cymodocea nodosa Seeds

Seeds of C. nodosa were collected on the south coast of Gran Canaria Island (Canary Islands, Spain). They were washed and cultivated in 15 psu seawater for 25–30 days, as previously described [13], to induce germination. Three different stages were identified according to the germination stage. The dormant stage (Stage 0, control) corresponds to an intact seed. Stage I refers seeds with an opening of dorsal crest at the extreme, which takes ca. 2 w from Stage 0, and finally, a cotyledon emerges and elongates in Stage II (Figure 3). Moreover, around day 25, the first roots are oppositely formed, and a complete seedling is developed by day 35, consisting of a bundle, three leaves and several seminal roots [13]. For each germination stage, ca. 15–20 seeds were withdrawn and treated as a pooled sample.

3.2. RNA Extraction and Reverse Transcription

Total RNA from pooled seeds of each stage (stages of seeds 0, I and II) was isolated using TRI-Reagent^®^ (Sigma, St. Louis, MO, USA), according to the manufacturer’s instructions. The isolated RNA samples were individually suspended in 20 μL of 1 M Tris-HCl (pH 8), 0.5 M EDTA, and treated with DNase (1 Umg-1, Promega, Madison, WI, USA) to destroy contaminating DNA. RNA was quantified using BioExperion with a minimum RIN quality of 7 or higher. Extracted RNA from each sample (~1 μg) was reverse transcribed in the presence of oligo (dT) and primers with randomly generated sequences from an iScript cDNA synthesis kit (Bio-Rad, Hercules, CA, USA). The reverse transcription procedure was carried out at 25 °C for 5 min, 42 °C for 30 min, and 85 °C for 5 min. The integrity of the cDNA was validated using a NanoDrop spectrophotometer (Thermo Fisher Scientific, Waltham, MA, USA). The products were kept at 4 °C until used for gene expression.

3.3. Construction of Seed cDNA Libraries

Three cDNA libraries (stages of seeds 0, I and II) were constructed and sequenced using SMRT Cell 1M Sequel II PacBio equipment. All libraries were sequenced independently to increase the deep sequencing and facilitate the search for aquaporin transcription differences among three seed stages.

The preparation of Iso-Seq transcriptomic libraries and sequencing reactions were carried out by the Genomics Section of the Central Support Service for Experimental Research (SCSIE) at the University of Valencia. PacBio includes an mRNA selection step based on the poly(A) tail, which enables the sequencing of high-quality full-length transcripts with complete poly(A) tails and facilitates the analysis of isoforms and alternative polyadenylation events.

To assess the completeness of the assembled transcriptome, a Benchmarking Universal Single-Copy Orthologs (BUSCO) [58] analysis was performed using the eukaryota_odb10 lineage dataset, which includes 255 conserved orthologous genes representative of eukaryotic species. BUSCO was run in transcriptome mode (euk_tran) to evaluate the presence of complete, fragmented, and missing genes within the assembled transcripts. This analysis allowed us to quantify the proportion of essential genes recovered, providing an objective measure of the completeness and quality of the transcriptome assembly. The results were reported as percentages of complete (single-copy and duplicated), fragmented, and missing BUSCOs, thus serving as a standard metric for transcriptome assembly evaluation. The Rapid Analysis of Transcriptome Data platform (TRAPID) webserver was used to assign annotations and GO terms to the predicted genes of the aquaporins, as well as to detect open reading frames (ORFs) and frameshift corrections at each transcript [59,60]. The TRAPID database, joined to PLAZA 4.5 plant database [59], also assigns functions based on sequence similarity.

3.4. Identification of Aquaporins Sequences

Full sequence transcripts, gene sequences with a start- and end-codon, identified as aquaporins, were selected from the transcriptome. Then, a maximum-likelihood (ML) [61] analysis was conducted to compare C. nodosa transcripts with those aquaporins identified in the UniProt database [62], utilizing a selection of completely identified aquaporins listed for the Alismatales order (Supplementary Table S1). Too short and too long sequences were eliminated to minimize artifacts as a long-branch attraction in the inferred phylogeny. Due to the close phylogenetic relationship between aquaporins and aquaglyceroporins (GLPs) [63,64,65], three GLPs were selected, one from Rhizophagus irregularis (I3W9F7) and two from* Botryotinia fuckeliana* (A0A384JPP3, A0A384JSZ0), and were used as an outgroup for the phylogenetic analysis of C. nodosa aquaporins.

An alignment dataset was constructed using MAFFT [66] from the amino acid sequences of the aquaporins protein-coding genes retrieved from UniProt DB [62], which is linked to the National Center for Biotechnology Information (NCBI, https://www.ncbi.nlm.nih.gov, accessed on 26 November 2025) and to the EMBL-EBI (European Molecular Biology Laboratory European Bioinformatics Institute, https://www.ebi.ac.uk/, accessed on 26 November 2025), and from the translated full sequence transcripts of aquaporins from C. nodosa transcriptome. Molecular databases at the UniProt were searched using “aquaporin” as a text query and filtered by “Taxonomy [OC]” (Alismatales [16360]). The ALTER webserver [67] was used to transform protein alignments in different format conversions and eliminate identical sequences (Supplementary Table S2).

The ML analysis was performed using the IQ-TREE webserver [68,69] with automatic detection of the substitution model [70], with Free Rate heterogeneity (+R) [71,72], Ultrafast bootstrap approximation [73] with 1000 bootstrap alignments, and 0.99 of the minimum correlation coefficients.

Once the C. nodosa sequences were clustered according to aquaporin type (i.e., PIP, SIP, TIP and NIP), alignments for each clade were performed through Sequencher v. 5.0.1 (Gene Codes Co.; Ann Arbor, MI, USA) to design pairs of specific primers with the corresponding nt sequences given in the transcriptome.

3.5. Gene Expression Through ddPCR

Gene expression of each aquaporin type (ie., PIP, SIP, TIP and NIP) was assayed through droplet digital PCR (ddPCR) for each of three germination stages for C. nodosa seeds. Specifically, gene expression was evaluated by designing five primer pairs for PIP (PIP1-PIP5), two for SIP (SIP1 and SIP2) and TIP (TIP1 and TIP 2), and one for NIP (Table 5). For quantification of each target transcript by ddPCR, QX200 ddPCR EvaGreen Supermix (Bio-Rad) was used, according to the manufacturer’s instructions. Briefly, for each sample, a PCR reaction mix (final volume, 20 μL) was prepared containing 1.5 μL of cDNA, 10 μL of QX200 ddPCR Eva Green Supermix, and 0.22 μL of each primer (10 μM), which was then loaded into a cartridge. Then, an oil droplet (70 μL) was loaded into each cartridge, and the cartridge was covered with a gasket. Each cartridge was individually introduced into the droplet generator, and finally, droplets of ~40 μL were transferred to the amplification plate. For each gene, three replicates were analysed for each seed stage. PCR amplification was performed with a C1000 Touch Thermal Cycler (Bio-Rad) using the following conditions: an initial step at 95 °C for 5 min; followed by 40 cycles of 95 °C for 30 s, an experimentally determined annealing temperature for each contig sequence for 1 min, and 72 °C for 45 s; a single step at 4 °C for 5 min; and a temperature ramping from 4 °C to 90 °C at a rate of 2 °C s^−1^ for 5 min. After amplification, each sample was quantified using QuantaSoft v1.7.4 software (Bio-Rad). Data from merged wells (corresponding to each group of replicates) were retrieved, and the concentration of each group is given as the average number of transcript copies per μL.

3.6. Data Analysis

Gene expression (transcript copies × μL^−1^) is expressed as mean ± standard deviation (SD; n = 3). Statistical comparisons of concentrations were performed using R software (https://www.r-project.org, accessed on 26 November 2025) [74]. Differences in aquaporin gene expression were assessed using one-way ANOVA, followed by post hoc Tukey’s HSD and Dunnett’s T3 tests, with significance set at p ≤ 0.05. Comparisons were conducted between seed Stages II and I, as well as between each seed stage and the control (Stage 0).