Abstract

   Hemp possesses significant healthcare value due to its rich composition
   of unsaturated fatty acids and a distinctive golden ratio of linoleic
   acid to α-linolenic acid. As a promising special-oil crop, it holds
   substantial potential for development and utilization. However, the
   regulatory mechanisms underlying its lipid metabolic pathways remain
   poorly understood. In this study, the independently bred hemp seed
   variety Longdama No. 9 was used to construct a regulatory network of
   the fatty-acid and lipid metabolisms through integrative transcriptomic
   and lipidomic analysis. Transcriptomic profiling revealed
   differentially expressed genes (DEGs) involved in lipid biosynthesis
   across various tissues. In leaves, DEGs associated with glycerolipid
   synthesis were generally upregulated compared to in roots and seeds. In
   seeds, DEGs involved in fatty-acid synthesis and triacylglycerol (TAG)
   assembly were predominantly upregulated. Meanwhile, root tissues showed
   a higher abundance of upregulated DEGs related to phospholipid
   biosynthesis. Lipidomic analysis further highlighted tissue-specific
   lipid distributions. The galactolipid monogalactosyldiacylglycerol
   (MGDG) was most abundant in the leaves. While phosphatidylglycerol (PG)
   had the highest molar percentage in the seeds, most other major
   phospholipids were predominantly found in the roots. The prevalence of
   the C36:6 molecular species in the MGDG and digalactosyldiacylglycerol
   (DGDG) indicates that hemp is a typical 18:3 plant. The combined
   transcriptomic and lipidomic analysis revealed that tissue-specific
   transcriptional regulation contributes to the unique lipid profile of
   hemp. These findings provide valuable insights into the regulation of
   lipid metabolism in hemp and identify key genes involved in oil
   biosynthesis, which can lay a theoretical foundation for the
   development and utilization of hemp as a special-oil crop.

   Keywords: Cannabis sativa L., lipid metabolic, lipidomic,
   transcriptomic, regulatory network

1. Introduction

   Hemp (Cannabis sativa L.) refers to cannabis cultivars in which the
   tetrahydrocannabinol (THC) content in the dry matter of the upper
   leaves and inflorescences during flowering remains below 0.3%, thereby
   eliminating its suitability for direct drug use [[40]1] Currently, it
   is currently cultivated on a commercial scale across Canada, the United
   States, and China. As a high-yield, low-input, and environmentally
   benign crop, it serves multiple supply chains, including food,
   pharmaceuticals, construction, and specialty chemicals [[41]2] Among
   the oldest domesticated plant species, it integrates substantial
   economic returns with pronounced pharmacological potential. The
   therapeutic relevance of hemp is primarily attributed to cannabinoids,
   which are biosynthesized and sequestered within the glandular trichomes
   of female inflorescences. These phytocannabinoids act as exogenous
   ligands of the endocannabinoid system, thereby modulating metabolic,
   immune, and neurophysiological homeostasis [[42]3,[43]4].
   Concomitantly, they exhibit well-documented antibacterial,
   anti-inflammatory, and anxiolytic properties [[44]5]. Collectively,
   industrial hemp represents a paradigmatic, sustainable cash crop whose
   versatile applications in agriculture and manufacturing are
   increasingly recognized as a catalyst for resilient economic growth on
   a global scale.

   In recent years, the escalating scientific and consumer recognition of
   the superior nutritional profile and associated health benefits
   inherent to hemp seeds has positioned them as a high-value commodity
   with significant agronomic and commercial development potential
   [[45]6]. Among all hemp tissues, only the seeds are currently permitted
   for food use, as they are intrinsically devoid of cannabinoids; any
   trace cannabinoid residues present result exclusively from
   adventitious, mechanically induced contamination during harvesting and
   downstream processing. Hemp seeds contain 25–35% lipids, with a
   distinctive fatty-acid profile comprising 70–90% polyunsaturated fatty
   acids (PUFAs) [[46]7]. Among these, linoleic acid (18:2, n-6; LA) is
   the most abundant, representing more than half of the total fatty-acid
   content, followed by α-linolenic acid (18:3, n-3; ALA) [[47]8]. As a
   rich dietary source of these two essential fatty acids (EFAs), which
   cannot be synthesized by the human body and must be obtained through
   the diet, hemp seed oil offers considerable nutritional benefits. This
   makes hemp a promising candidate for high-quality edible oil
   production, with significant potential for development and utilization
   [[48]2,[49]9].

   Triacylglycerol (TAG) is the primary constituent of vegetable oils and
   serves as the main storage form of carbon and energy in plants. It is
   predominantly accumulated in specific tissues and plays a vital role in
   seedling development, lipid homeostasis, and the reproductive processes
   [[50]10]. Fatty acids form the fundamental building blocks for TAG
   synthesis. In plants, TAG biosynthesis involves two major stages:
   fatty-acid synthesis within plastids and subsequent TAG assembly in the
   endoplasmic reticulum (ER) [[51]11]. While the specific TAG
   biosynthetic pathways may vary across species, it is estimated that
   over 90% of seed fatty acids are funneled into TAG via PC intermediates
   [[52]12]. Otherwise, the cell membrane is primarily composed of three
   major classes of lipids: glycerolipids, sphingolipids, and sterols
   [[53]13]. Among these, glycerolipids serve as the principal structural
   components, with variations in the fatty acyl chain composition and
   ratios across different membrane systems. Phospholipid biosynthesis
   mainly occurs in the endoplasmic reticulum (ER), and common plant
   phospholipids include phosphatidic acid (PA), PC,
   phosphatidylethanolamine (PE), phosphatidylserine (PS),
   phosphatidylglycerol (PG), and phosphatidylinositol (PI) [[54]14].
   Glycolipids in plants exhibit structures similar to phospholipids and
   are categorized, based on their saccharide head groups, into
   monogalactosyldiacylglycerol (MGDG), digalactosyldiacylglycerol (DGDG),
   and sulfoquinovosyldiacylglycerol (SQDG) [[55]15,[56]16]. MGDG and DGDG
   are predominant in photosynthetic tissues and are unique to
   photosynthetic membranes. Alongside PG, they are essential components
   of the plasma membrane and play a critical role in plant photosynthesis
   [[57]17]. Phospholipids are transported from the ER to chloroplasts,
   where they provide diacylglycerol (DAG) precursors for glycolipid
   biosynthesis. The composition of membrane lipids is intricately linked
   to the lipid synthesis, transport, turnover, remodeling, and
   degradation processes [[58]18].

   While the biosynthetic pathways of plant oils and associated gene
   functions have been extensively characterized in many higher plants,
   particularly in conventional oil crops [[59]19], genetic resources and
   functional studies in hemp remain limited. Gene-editing strategies
   targeting fatty-acid composition and oil content have become central to
   breeding elite oilseed germplasms. In recent years, emerging
   special-oil crops such as hemp have gained attention as valuable
   biological resources for identifying key genes involved in oil
   biosynthesis and metabolic improvement [[60]20,[61]21]. Meanwhile,
   although interest in cannabis-derived foods intended for human
   consumption is intensifying, specific regulation concerning the
   analytical parameters for the quality assessment of hemp seed oils is
   still lacking [[62]22]. However, due to strict regulatory oversight and
   its relatively recent industrialization, hemp still lacks comprehensive
   genetic information. To address this gap, the present study employed
   transcriptomic and lipidomic analysis to investigate the
   lipid-metabolism pathways of hemp and identify critical genes involved
   in oil biosynthesis. These findings aim to establish a foundation for
   deeper insights into the regulatory mechanisms and functional genomics
   of lipid metabolism in hemp. These findings will facilitate the
   targeted enhancement of lipid profiles in hemp, substantially elevate
   seed oil content, and furnish valuable genetic resources for the
   development of specialty oil crops. Furthermore, these studies could
   contribute to standardizing hemp seed oil quality.

2. Materials and Methods

2.1. Plant Materials and Growth Conditions

   The experimental material used in this study was Longdama No. 9, a
   seed-type hemp variety independently developed by the Institute of
   Industrial Crops, Heilongjiang Academy of Agricultural Sciences
   (Harbin). Plants were cultivated in a greenhouse under controlled
   conditions: a 14 h light/10 h dark photoperiod, day/night temperatures
   of 22 °C/18 °C, light intensity of 20,000 lux, and relative humidity
   maintained at 70 ± 1% [[63]23]. A total of 120 pots of plants under the
   same culture conditions were cultivated. Tissue samples were collected
   from leaves and roots six weeks after emergence and from seeds 14 days
   post-flowering during the developmental stage. Each sample set had 3
   biological replicates; each replicate was made of three mixed samples.
   The collected samples were immediately sealed in cryogenic tubes,
   flash-frozen in liquid nitrogen, and stored at −80 °C until further
   analysis.

2.2. RNA Sequencing and Quantitative Real-Time PCR (qRT-PCR)

   Total RNA was extracted from hemp tissues using the TRIzol reagent. RNA
   quality was assessed via agarose gel electrophoresis and an
   ultra-micro-UV-visible spectrophotometer (Thermo Fisher Scientific,
   Waltham, MA, USA), ensuring OD260/280 values between 1.8 and 2.0.
   Qualified RNA samples were used for cDNA library construction and
   subsequently sequenced on the Illumina HiSeq 4000 high-throughput
   sequencing platform. DEGs were annotated by aligning them against
   public databases, including Swiss-Prot, KEGG, KOG, and Pfam, to
   identify homologous protein sequences and assign functional
   annotations. Gene expression levels were quantified using the
   featureCounts tool (version 2.0.3), and the FPKM (fragments per
   kilobase of transcript per million mapped reads) method was employed to
   estimate transcript abundance. Differential expression analysis between
   various hemp tissues was conducted using the DESeq2 package, with
   p-value adjustment via the Benjamini–Hochberg method. DEGs were defined
   based on these criteria: Padj < 0.05 and |log2 (Fold Change) | ≥ 1.
   Enrichment analyses for Gene Ontology (GO) terms and KEGG pathways were
   then performed on the identified DEGs. To validate the RNA-seq data,
   twelve DEGs were selected for quantitative real-time PCR (qRT-PCR)
   analysis. Gene-specific primers were designed using Primer6 software
   ([64]Table S1). Reverse transcription was carried out using the
   SuperScript IV First-Strand Synthesis System (Thermo Fisher Scientific,
   Cat. No. 18090050) according to the manufacturer’s instructions.
   qRT-PCR was performed using TOYOBO’s real-time quantitative detection
   kit (Cat. No. QKD-201), with Actin used as the internal reference gene.
   Relative expression levels were calculated using the 2^−ΔΔCt method.

2.3. Lipid Extraction and Identification

   The hemp lipid extraction was slightly modified based on previous
   studies [[65]24]. Briefly, tissue samples from different parts of the
   hemp plant were immersed in preheated isopropanol containing 0.01%
   butylated hydroxytoluene (BHT) at 75 °C for 15 min. Following this, 1.5
   mL of chloroform and 0.6 mL of distilled water were added to each
   sample. The mixture was thoroughly vortexed and subjected to shaking
   extraction at 22 °C for 1 h at 180 rpm. After centrifugation, the
   supernatant was transferred to a clean tube, and 4 mL of a
   chloroform:methanol mixture (2:1, v/v) containing 0.01% BHT was added.
   The mixture was then vortexed and extracted at 22 °C for 30 min at 180
   rpm. This extraction process was repeated three to four times. The
   combined extracts were washed with 1 mL of 1 M KCl, and the supernatant
   was discarded following centrifugation. Subsequently, 2 mL of distilled
   water was added and vortexed thoroughly, and the supernatant was again
   removed. The final lipid extract was dried under a stream of nitrogen
   gas and stored at –20 °C for later use.

   The Kansas Lipidomic Research Center (KLRC, Manhattan, KS, USA) was
   commissioned to perform a comprehensive lipidomic analysis by
   electrospray ionization tandem mass spectrometry (ESI-MS/MS). This
   approach precisely quantified fluctuations in lipid metabolite levels
   while elucidating fatty-acid composition and structure. The profiles of
   the lipids were systematically drawn using precursor (Prec) and neutral
   loss (NL) strategies. The sample was introduced into the electrospray
   ionization source. A series of peaks in the lipid content were detected
   by electrospray ionization. The peaks on the spectrum were
   quantitatively compared with a set of internal standards. Data for each
   lipid molecular species were normalized and displayed as the mole
   percent of the total lipid analyzed.

2.4. Statistical Analysis and Data Visualization

   Statistical analyses and graph generation were conducted using GraphPad
   Prism 9.0 software. Results are presented as means ± standard deviation
   (SD). Construction of the regulatory network diagrams was carried out
   using a combination of TBtools-II (v 2.315) and Microsoft Office
   PowerPoint 2016.

3. Results

3.1. Transcriptomic Analysis of Hemp

   To elucidate the transcriptional regulatory mechanisms of the genes
   involved in lipid metabolism in hemp, RNA-seq analysis was conducted on
   nine samples derived from developing seeds, leaves, and root tissues.
   The raw sequencing data have been deposited in the NCBI database in
   FASTQ format under BioProject accession number PRJNA899681. After
   removing redundant sequences, a total of 60.7 Gb of clean data was
   obtained. The clean data per sample ranged from 6.59 to 7.37 Gb, with
   an average sequencing error rate of approximately 0.1% and alignment
   rates exceeding 85.78%. The proportion of uniquely mapped reads ranged
   from 81.58% to 86.61%, with Q30 base percentages exceeding 94.41% and
   GC content ranging between 43.03% and 44.43% ([66]Table S2). These
   metrics indicate high-quality sequencing data suitable for downstream
   analysis. The reliability of the RNA-seq results was further validated
   through qRT-PCR analysis ([67]Figure S1).

3.2. The GO and KEGG Enrichment Analysis

   DEGs in three tissues, leaves (L), roots (R), and developing seeds (S)
   were screened and compared ([68]Figure 1A). The comparison between
   seeds and roots (S vs. R) revealed the highest number of DEGs, totaling
   11,071, comprising 5151 upregulated and 5920 downregulated genes, with
   downregulated genes being more predominant. In contrast, the leaf vs.
   root (L vs. R) comparison showed the fewest DEGs, with 7922 genes
   identified: 4047 upregulated and 3875 downregulated. In the seed vs.
   leaf (S vs. L) group, a total of 9843 DEGs were detected, including
   4408 upregulated and 5435 downregulated genes, with the downregulated
   genes outnumbering the upregulated ones.

Figure 1.

   [69]Figure 1
   [70]Open in a new tab

   Quantitative statistics of differential genes and analysis of
   transcription data annotation results. (A) Venn diagrams illustrating
   the distribution of DEGs across three tissue types: leaf (L), root (R),
   and developing seed (S). The green diagram represents the total DEGs,
   the orange diagram highlights upregulated DEGs, and the blue diagram
   shows downregulated DEGs. (B) Functional classification of DEGs based
   on GO and KEGG pathway annotations. The bar charts display the top five
   terms in each GO category (biological process, cellular component,
   molecular function) and KEGG pathways. (C) DEGs involved in metabolic
   pathways, with different colors representing different tissue
   comparison groups.

   Transcriptome data were aligned to the KEGG database, resulting in the
   annotation of 3130 DEGs across 125 metabolic pathways. It is noteworthy
   that some DEGs were involved in multiple pathways. Among these, the
   “Metabolism” category accounted for the majority, with 2252 DEGs (72%),
   followed by “Genetic Information Processing,” comprising 28%.
   Additional pathway categories included “Environmental Information
   Processing” (10%, 305 genes), “Cellular Processes” (6%, 196 genes), and
   “Organismal Systems” (5%, 166 genes) ([71]Figure 1B). Among the
   metabolism pathways, the top five pathways with the most annotated
   genes were carbohydrate metabolism, amino acid metabolism, lipid
   metabolism, biosynthesis of other secondary metabolites, and energy
   metabolism pathways, accounting for 28% (508), 17% (305), 16% (292),
   15% (271), and 12% (219), respectively ([72]Table S3). A similar
   distribution of metabolic pathway involvement was observed across the
   different comparison groups, with carbohydrate metabolism consistently
   representing the largest proportion. For DEGs associated with lipid
   metabolism, the highest number was identified in the “L vs. R”
   comparison group, totaling 192 genes ([73]Figure 1C). GO annotation of
   the hemp transcriptome data classified the DEGs into three main
   categories: cellular components (CCs), molecular functions (MFs), and
   biological processes (BPs) ([74]Figure 1B). Within the cellular
   component category, 8987 DEGs were annotated, with the top five
   subcategories being integral components of membranes (2126 genes, 24%),
   plasma membranes (1383 genes, 15%), chloroplasts (1208 genes, 13%),
   membranes (531 genes, 6%), and chloroplast stromata (405 genes, 5%). In
   the molecular function category, 8250 DEGs were identified. The most
   represented subcategories included DNA-binding transcription factor
   activity (790 genes, 10%), sequence-specific DNA binding (428 genes,
   5%), heme binding (235 genes, 3%), structural constituents of ribosomes
   (225 genes, 3%), and oxidoreductase activity (189 genes, 2%). As for
   biological processes, 8030 DEGs were annotated. The major subcategories
   were the defense response (363 genes, 5%), the response to salt stress
   (299 genes, 4%), and translation (191 genes, 2%).

3.3. Lipid Metabolism-Related Genes

   Based on the annotation results of the hemp transcriptome data, further
   analysis of the gene annotation associated with lipid metabolism found
   that DEGs related to lipid metabolism were mapped to fourteen
   lipid-metabolism pathways (full list and detailed information are
   listed in [75]Supplementary Table S4). Among these, the most enriched
   pathways were glycerophospholipid metabolism, glycerolipid metabolism,
   and fatty-acid biosynthesis, with 71, 57, and 43 DEGs, respectively,
   representing 24%, 20%, and 15% of the lipid metabolism-related DEGs
   ([76]Figure 2A). Tissue-specific pairwise comparisons revealed that the
   “S vs. R” group exhibited the highest number of lipid
   metabolism-related DEGs, totaling 192, of which 41 were involved in
   glycerolipid metabolism (21%). In the “S vs. L” group, 169 DEGs were
   annotated, including 34 related to glycerophospholipid metabolism
   (20%). The “L vs. R” comparison revealed 121 DEGs, with 29 associated
   with glycerolipid metabolism (23%) ([77]Figure 2B). We found that
   across all tissue comparisons, glycerophospholipid metabolism,
   glycerolipid metabolism, and fatty-acid biosynthesis consistently
   showed high numbers and proportions of annotated DEGs ([78]Figure 2B).
   Notably, in comparisons involving developing seeds and root tissues,
   glycerophospholipid- and glycerolipid-metabolism pathways exhibited the
   highest DEG representation.

Figure 2.

   [79]Figure 2
   [80]Open in a new tab

   KEGG function classification and lipid-metabolism pathway enrichment
   analysis. (A) Distribution of metabolism-related DEGs annotated in the
   hemp transcriptome, with each color representing a different metabolic
   pathway. (B) DEGs associated with lipid-metabolism pathways across
   different tissue comparisons. Distinct colors indicate different
   comparison groups.

   To further investigate the regulatory patterns of lipid metabolism in
   hemp, genome-wide screening of lipid-related genes was conducted based
   on previously published studies [[81]23,[82]25]. Differentially
   expressed genes (DEGs) associated with lipid metabolism were identified
   across various hemp tissues using a threshold of |Log2FC| ≥ 1.
   Transcriptome analysis revealed a total of 696 genes involved in lipid
   metabolic pathways ([83]Figure S2 and Table S5). Among the tissue
   comparisons, the “S vs. R” group exhibited the highest number of
   lipid-related DEGs, with 489 genes identified, 302 of which were
   upregulated, significantly outnumbering the 187 downregulated genes. In
   the “L vs. R” comparison, 341 lipid-related DEGs were detected,
   including 185 upregulated and 156 downregulated genes. The “L vs. S”
   group showed the lowest number of DEGs, with 304 genes, nearly equally
   divided between upregulated (153) and downregulated (151) genes. Most
   genes associated with fatty acids and photosynthetic-membrane lipid
   synthesis are highly expressed in leaf tissues, and genes associated
   with TAGs are highly expressed in developing seeds.

3.3.1. Genes Associated with Triglyceride Biosynthesis

   Triacylglycerol (TAG) serves as the primary storage form of oil in
   hemp, making the TAG biosynthetic pathway critical for oil
   accumulation. To elucidate the regulatory mechanisms of TAG
   biosynthesis across different tissues, differentially expressed genes
   (DEGs) involved in the TAG metabolic pathway were identified, including
   GPDH (glycerol-3-phosphate dehydrogenase), GPAT (glycerol-3-phosphate
   acyltransferase), LPAT (lysophosphatidyl acyltransferase), PAP
   (phosphatidic acid phosphatase), DGAT, and PDAT. Three GPDH genes
   (LOC115710084, LOC115710085, LOC115698811) exhibited consistent
   upregulation in both the “S vs. L” and “S vs. R” comparisons, with
   LOC115698811 showing Log2FC values of 5.12 and 5.55, respectively. One
   GPAT gene (LOC115697313) was also significantly upregulated in seeds,
   with Log2FC values of 2.69 (“S vs. L”) and 3.23 (“S vs. R”). Strong
   induction of DGAT genes was observed in both comparisons, with WSD1
   showing the highest expression levels and Log2FC values, of 11.53 and
   11.12, respectively ([84]Table S6). Additionally, an upregulated PDAT
   gene showed expression increases of 5.89 (“S vs. L”) and 2.38 (“S vs.
   R”). Furthermore, two FAD2 genes (LOC115718446, LOC115719000) and two
   FAD3 genes (LOC115720710, LOC115697681) were identified as upregulated
   in the seed tissue. Notably, the FAD2 genes and FAD3 (LOC115720710)
   displayed high expression levels, with Log2FC values exceeding 5
   ([85]Figure 3 and [86]Table S6). In addition, five SAD genes were
   highly expressed in developing seeds (LOC115705664, LOC115705684,
   LOC115707488, LOC115698723, LOC115708035). Among these, LOC115705664
   exhibited the most pronounced upregulation, with Log2FC values of 9.38
   (“S vs. L”) and 7.49 (“S vs. R”). These expression patterns may
   contribute to the elevated polyunsaturated fatty-acid content observed
   in industrial hemp.

Figure 3.

   [87]Figure 3
   [88]Open in a new tab

   Schematic representation of lipid metabolism in hemp. (A) Visualization
   of the TAG and photosynthetic-membrane lipid metabolic regulatory
   network. Gene expression data were integrated into the corresponding
   metabolic pathways, with differential expression levels of key
   lipid-related genes illustrated as heat maps. The colors indicate FPKM
   values across different tissues, representing variations in gene
   expression and corresponding lipid metabolite types. (B) Heat map
   showing the expression profiles of genes involved in lipid biosynthesis
   pathways in hemp. The color gradient reflects expression intensity,
   with red indicating high expression and blue indicating low expression.

3.3.2. Analysis of Genes Involved in Photosynthetic-Membrane Lipid Synthesis

   Further screening identified DEGs involved in the biosynthesis of
   photosynthetic-membrane lipids, including MGD
   (monogalactosyldiacylglycerol synthase), DGD
   (digalactosyldiacylglycerol synthase), and SQD
   (sulfoquinovosyldiacylglycerol synthase), as well as FAD4, FAD6, and
   FAD7/8. Genes associated with galactolipid metabolism exhibited
   upregulated expression in both the “L vs. R” and “L vs. S” comparison
   groups. Specifically, MGD (LOC115707728) showed Log2FC values of 2.29
   (“L vs. R”) and 1.29 (“L vs. S”); DGD (LOC115698658) had Log2FC values
   of 1.18 and 1.05, respectively; and SQD (LOC115697608) exhibited Log2FC
   values of 1.57 (“L vs. R”) and 1.39 (“L vs. S”). Additionally, FAD4
   (LOC115698783) was markedly upregulated in both comparisons, with
   Log2FC values of 7.64 and 3.41, respectively. FAD6 (LOC115725117) also
   showed significant upregulation, with Log2FC values of 2.67 (“L vs. R”)
   and 1.78 (“L vs. S”). Similarly, FAD7 was significantly upregulated in
   leaf tissues, with Log2FC values of 4.53 and 2.30 in the “L vs. R” and
   “L vs. S” comparisons, respectively ([89]Figure 3 and [90]Table S6).
   FAD6 (LOC115725117) exhibited significant upregulation in both the “L
   vs. R” and “L vs. S” comparison groups, with Log2FC values of 2.67 and
   1.78, respectively. Similarly, FAD7 showed marked upregulation in the
   same comparisons, with Log2FC values of 4.53 and 2.30, respectively.

3.3.3. Analysis of DEGs in Fatty-Acid Metabolic Pathways

   Numerous DEGs related to fatty-acid synthase were identified in the
   hemp transcriptome. ACCase, a key rate-limiting enzyme in fatty-acid
   metabolism, was represented by one upregulated gene (LOC115712810) in
   both the “S vs. L” and “S vs. R” comparison groups, with Log2FC values
   exceeding 2 ([91]Figure 4, [92]Table S6). Genes encoding β-ketoacyl-ACP
   synthases (KASI, KASII, and KASIII) were expressed at notably higher
   levels in developing seeds compared to other tissues. In particular,
   two KASII genes (LOC115707135 and LOC115707136) showed strong
   upregulation (Log2FC > 6). KASIII (LOC115717112), which catalyzes the
   initial step in fatty-acid elongation, also exhibited upregulated
   expression, with Log2FC values of 1.42 (“S vs. L”) and 1.96 (“S vs.
   R”). Five SAD genes were significantly upregulated in seed tissues
   (LOC115705664, LOC115705684, LOC115707488, LOC115698723, LOC115708035).
   Among these, LOC115705664 showed the highest expression levels, with
   Log2FC values of 9.38 and 7.49 in the “S vs. L” and “S vs. R” groups,
   respectively, potentially contributing to the high content of
   unsaturated fatty acids in hemp. Additionally, FATA and FATB
   thioesterase genes were upregulated in both tissue comparisons, with
   FATA (LOC115714069) showing particularly elevated expression (Log2FC =
   2.17 in “S vs. L” and 1.97 in “S vs. R”), suggesting a role in
   promoting the accumulation of unsaturated fatty acids ([93]Figure 4,
   [94]Table S6).

Figure 4.

   [95]Figure 4
   [96]Open in a new tab

   Transcriptomic profiling of lipid metabolism-related genes in the
   fatty-acid biosynthesis pathways of hemp. (A) Schematic representation
   of the fatty-acid biosynthesis pathway in hemp, with gene expression
   data overlaid onto the corresponding metabolic pathway diagram.
   Differential expression levels of key lipid-related genes are
   visualized as heat maps, indicating tissue-specific variations in gene
   activity and associated lipid metabolite types. The different colors
   were used to indicate the FPKM values of lipid metabolism genes in
   different tissues. (B) Heat map showing the expression patterns of
   genes involved in fatty-acid biosynthesis. The color gradient indicates
   expression levels, where red represents high expression and blue
   represents low expression.

3.4. Lipidome of Hemp

   Lipidomic analysis was conducted to profile the major lipid classes and
   fatty-acid composition in hemp leaves, developing seeds, and root
   tissues. A total of 11 lipid components were identified, including two
   galactolipids (MGDG, DGDG); six phospholipids, including PC, PI, PE,
   PS, PA, and PG; and 3 lysophospholipids, including LPA, LPC, and LPG
   ([97]Figure 5). In developing seed tissues, PC was the most abundant
   phospholipid, comprising 30.6% of the total lipid content.
   Galactolipids collectively accounted for 44.6%, with MGDG representing
   68% of this category, making it the dominant galactolipid.
   Lysophospholipids were present at 1.3%, a higher proportion than in
   leaf or root tissues. In leaf tissues, galactolipids (MGDG and DGDG)
   were predominant, accounting for 65.4% of the total lipids, with MGDG
   alone contributing nearly 50%. Phospholipids represented 34.2% of the
   total lipids, with PC being the most abundant at 16.6%. Other lipid
   components were detected at lower levels. The elevated levels of
   galactolipids and PG, key constituents of the photosynthetic membrane,
   suggest their important role in photosynthesis. In contrast, root
   tissues showed the highest phospholipid content, comprising 84.9% of
   the total lipids. PC was the most abundant at 40.4%, followed by PA at
   21.9%. In comparison, galactolipids and PG were present at much lower
   levels, at 14.1% and 3.3%, respectively, indicating a reduced role of
   photosynthetic-membrane lipids in roots.

Figure 5.

   [98]Figure 5
   [99]Open in a new tab

   The concentrations of each lipid class of hemp in different tissues.
   Pie charts illustrating the proportional distributions of major lipid
   classes in developing seeds (A), leaves (B), and roots (C) of hemp.
   (D–F) Pie charts showing the distribution of lysophospholipid
   subclasses in seeds (D), leaves (E), and roots (F). Distinct colors
   represent different lipid classes. (G) Heat map displaying the relative
   concentration of each lipid class across hemp tissues. The color
   gradient indicates the concentration levels, with red denoting high
   abundance and blue indicating low abundance.

3.5. Analysis of Molecular Species Types of Lipids in Different Tissue Sites

   In addition to profiling overall lipid classes, the molecular species
   of the major lipid-associated fatty-acid side chains in hemp were
   analyzed. The predominant lipid molecular species in hemp were found to
   be C34 and C36, with galactolipids mainly consisting of C36
   species—particularly C36:6, which emerged as the most abundant subtype
   ([100]Figure 6). Within MGDG, the molar percentage of C36:6 in leaf
   tissue reached as high as 88%, significantly higher than in seeds or
   roots. For DGDG, the dominant molecular species in seeds and leaves
   were C34:3 and C36:6, while in the roots, C36:5, C36:4, C34:2, and
   C34:1 were more prevalent. Overall, the predominance of C36:6 in both
   MGDG and DGDG across tissues suggests that these galactolipids are
   primarily synthesized via the eukaryotic (endoplasmic
   reticulum-derived) pathway, indicating that hemp is a typical 18:3
   plant. The phospholipid species in hemp were also dominated by C34 and
   C36 molecules, with C34:2 and C36:4 being the most abundant. In PG, C34
   was the primary species, with C34:4, C34:3, and C34:2 dominating across
   tissues. In PC, C36:4 and C34:2 were the major species in developing
   seeds and roots, while C34:2 was predominant in leaves. The molecular
   profile of PE closely resembled that of PC, with C36:4 and C34:2 being
   most abundant in developing seeds. For PI, the composition was similar
   to that of PG, with C34 species predominating, especially C34:2 and
   C34:3. C34:2 was particularly dominant in seeds and root tissues. PS,
   though present in lower amounts, was enriched in the C36:5 species
   across tissues compared to other lipid classes. In PA, C34:2 and C36:4
   were predominant in seeds and roots, while C34:2, C34:3, C36:4, and
   C36:5 were the major molecular species in leaves.

Figure 6.

   [101]Figure 6
   [102]Open in a new tab

   The distributions of glycerolipid and phospholipid lipid molecular
   species in the different tissues of hemp. This chart illustrates the
   compositions of various lipid molecular species in hemp leaves, seeds,
   and roots. Different colors represent distinct lipid molecular species.

4. Discussion

   Currently, in comparison to major staple crops, hemp has relatively
   limited genetic resources and molecular biology research. In this
   study, we generated transcriptomic data from various hemp tissues using
   RNA-seq, yielding a total of 60.7 GB of high-quality transcriptome
   data. This dataset was used to explore the regulatory mechanisms
   governing the genes involved in lipid-metabolism pathways. The
   identified DEGs in hemp span a broad range of biological processes and
   metabolic pathways and include many with known functions related to
   lipid metabolism. The tissue-specific distribution of lipid classes
   follows a conservative pattern across diverse plant species, including
   rice, tomato, maize, and sorghum; glycolipids are markedly enriched in
   photosynthetic leaf tissues, whereas phospholipids predominate in root
   systems [[103]26,[104]27,[105]28,[106]29]. By integrating
   transcriptomic and lipid metabolomic analysis across different hemp
   tissues, we constructed a proposed molecular model of lipid metabolism
   ([107]Figure 3 and [108]Figure 4).

   In developing seeds of hemp, genes associated with TAG were found to be
   positively expressed ([109]Figure 7). GPAT, a central enzyme linking
   fatty-acid and glycerolipid biosynthesis, catalyzes the key
   rate-limiting initial step of glycerolipid biosynthesis and has been
   shown to play a critical role in oil biosynthesis in various plant
   species [[110]30,[111]31,[112]32]). In this study, the gene encoding
   GPAT was markedly upregulated in developing seeds, suggesting that GPAT
   is crucial for oil accumulation during hemp seed development. The LPAT
   catalyzes the second acylation step of the Kennedy pathway, converting
   LPA to PA. PA is subsequently converted to DAG and finally to TAG under
   the catalysis of DGAT [[113]33]. In general, DGAT1 is broadly involved
   in TAG biosynthesis in most plant species, while DGAT2 is more
   specialized in accumulating unusual fatty acids, such as DHA and EPA
   [[114]34,[115]35], and DGAT expression affects seed development and oil
   content [[116]34,[117]36]. In the present study, one LPAT gene and five
   DGAT genes were identified as significantly upregulated in developing
   hemp seeds ([118]Table S6), further supporting their functional
   relevance in hemp oil biosynthesis and accumulation. PC plays a pivotal
   role in lipid metabolism in plant cells. It not only serves as a major
   structural component of cellular membranes but also acts as a crucial
   acyl donor for lipid biosynthesis and an indispensable substrate for
   fatty-acid modifications, including desaturation and hydroxylation
   [[119]37]. The “Lands cycle”, a key pathway for rapid PC turnover and
   acyl remodeling, is mainly mediated by enzymes such as FAD2, PLA, and
   LPCAT [[120]38]. In the hemp transcriptome, simultaneous upregulation
   of FAD2, PDAT, and LPA contributes to the accumulation of 36:4 PC and
   18: 2 TAGs, which may be an important reason for the high unsaturated
   fatty-acid content in hemp seeds.

Figure 7.

   [121]Figure 7
   [122]Open in a new tab

   Proposed model for hemp lipid metabolism pattern. (A) Schematic
   representations of hemp plant structure and sampled tissues: (a) whole
   hemp plant, (b) leaf tissue, (c) developing seed tissue, and (d) root
   tissue. (B) Transcriptomic analysis reveals distinct expression
   patterns of lipid-related genes across different tissues. A heatmap
   displays the expression levels of genes involved in key lipid
   biosynthesis pathways, with varying colors indicating differential
   expression. This provides a clear visual comparison of tissue-specific
   gene activity related to lipid metabolism.

   Leaves, serving as the primary site for fatty-acid biosynthesis,
   exhibit an intimate association between their lipid metabolism and
   photosynthetic processes. In plant chloroplasts, the gene encoding
   ACCase was found to be significantly upregulated in developing seeds
   compared to other tissues, indicating its central role in supplying
   fatty-acid precursors for oil production. This observation is
   consistent with previous studies demonstrating that overexpression of
   ACCase significantly increases oil content in Chlamydomonas reinhardtii
   [[123]39]. KAS, specifically KASI, KASII, and KASIII, is critical for
   catalyzing condensation reactions during fatty-acid chain elongation
   [[124]40]. In the current study, genes encoding KAS were generally
   upregulated in both leaves and developing seeds of hemp, suggesting a
   prominent role of KAS in driving fatty-acid biosynthesis in these
   tissues. Hemp seeds are particularly rich in unsaturated fatty acids,
   which can comprise over 90% of the total seed oil content and which are
   recognized for their significant nutritional and health benefits
   [[125]41]. Fatty-acid desaturation predominantly takes place in the
   chloroplasts, where the rate-limiting enzyme SAD catalyzes the
   conversion of stearoyl-ACP to oleoyl-ACP. The critical function of SAD
   in unsaturated fatty-acid biosynthesis has been well-documented in
   several oil-rich crops [[126]42,[127]43,[128]44]. In this study,
   transcriptome analysis revealed that the SAD gene was significantly
   upregulated in developing hemp seeds, which could correspond to the
   accumulation of unsaturated fatty acids in hemp. MGDG and DGDG, primary
   constituents of photosynthetic-membrane lipids, are essential for
   chloroplast membrane integrity and photosynthetic efficiency [[129]45].
   Galactolipid biosynthesis occurs via two distinct pathways: the
   prokaryotic pathway and the eukaryotic pathway [[130]46]. For instance,
   in species like Arabidopsis thaliana and spinach, both pathways
   contribute to the synthesis of MGDG and DGDG, resulting in a prominent
   presence of C16 fatty acids, which are known as”16:3 plants”. In
   contrast, plants such as maize and wheat predominantly utilize the
   eukaryotic pathway, which leads to C36:6 becoming the dominant
   molecular species in their galactolipids, categorizing them as “18:3
   plants” [[131]47,[132]48]. In the present study, C36:6 was identified
   as the major galactolipid species in hemp leaves, indicating that hemp
   is a typical 18:3 plant. Supporting this, transcriptome analysis
   revealed that several genes encoding MGD and DGD were significantly
   upregulated in leaf tissues compared to other plant parts ([133]Figure
   3 and [134]Table S6), further highlighting the active galactolipid
   biosynthesis in photosynthetically active tissues of hemp.

   Lipid metabolism in root tissue focuses on phospholipid accumulation
   and adverse responses. Genes closely related to phospholipid
   biosynthetic pathways, such as PSD, PECT, and AAPT, showed increased
   expression in the root tissues. Consistently, a greater accumulation of
   phospholipids was detected in roots relative to other tissues. Previous
   studies have shown that phospholipid metabolism-related genes play
   essential roles in plant responses to abiotic stresses
   [[135]38,[136]49]. Thus, given that the roots are the primary organ in
   direct contact with the soil environment, the enrichment of
   phospholipids and the root-specific expression of these metabolic genes
   may contribute to the resilience of hemp to various environmental
   stressors, such as salinity, heavy metals, and alkaline conditions.

   In this study, an integrated transcriptomic and lipidomic analysis was
   conducted to investigate the regulatory patterns of lipid metabolism in
   different tissues of hemp. This study revealed that the composition and
   relative abundance of lipid species in hemp exhibit strong tissue
   specificity. Importantly, the expression patterns of genes involved in
   the biosynthesis and regulation of these lipid components also
   displayed a corresponding tissue-specific trend. Genes associated with
   the metabolism of lipid classes, such as glycerolipids, phospholipids,
   and galactolipids, were differentially expressed in tissues where those
   lipid types were enriched. These findings strongly suggest that the
   observed tissue-specific lipid distribution in hemp is driven by
   differential gene expression and the activity of key enzymes within
   lipid metabolic pathways. Elucidating the precise identity of these
   coding genes and their cognate metabolic pathways will furnish
   essential genetic resources and mechanistic insights for the targeted,
   genome-edited refinement of crop lipid profiles. Furthermore, this
   study integrates multi-tissue transcriptomic and lipidomic data to
   systematically characterize the tissue-specific regulatory network
   governing lipid synthesis in hemp, providing a verifiable theoretical
   framework for the “gene-metabolite-phenotype” association. Building on
   these findings, future work can further explore latent regulatory
   factors and their functional variants by combining genome-wide
   association studies with multi-omics modeling. Additionally, synthetic
   biology approaches can be employed to construct high-yield,
   high-specificity lipid cell factories, facilitating the paradigm shift
   from empirical breeding to design-based breeding in hemp. These
   strategies not only lay the theoretical and technical groundwork for
   the high-value utilization of this species in the food, pharmaceutical,
   and bioenergy sectors but also offer a transferable methodological
   pathway for lipid engineering in other plants.

5. Conclusions

   In conclusion, this study conducted transcriptomic and lipidomic
   integration analysis on developed hemp seeds, leaves, and root tissues.
   The results revealed that the DEGs associated with lipid metabolism
   were primarily enriched in glycerophospholipid metabolism, glycerolipid
   metabolism, and fatty-acid biosynthesis. In parallel, lipidomic
   profiling identified 11 distinct lipid classes across the tissues.
   Leaves were dominated by galactolipids, accounting for 65.4% of the
   total lipids, with C36:6 as the predominant species—consistent with the
   characteristic “18:3 plant” lipid profile. The roots showed a high
   abundance of phospholipids, representing 84.9% of their lipid
   composition. The seeds, meanwhile, were rich in PC and galactolipids.
   Transcriptomic analysis revealed that the lipid composition of hemp
   seeds is closely linked to the expression of genes involved in lipid
   metabolism, exhibiting a clear tissue-specific expression pattern.
   Genes associated with TAG biosynthesis were predominantly and highly
   expressed in developing seeds, whereas genes involved in galactolipid
   synthesis were most abundant in leaf tissues and those related to
   phospholipid biosynthesis showed the highest expression in root
   tissues. SAD, FAD2, and FAD3 show high expression levels in developing
   seeds. This likely contributes to the cumulation of unsaturated fatty
   acids, particularly PUFAs, in hemp. The findings of this study offer
   valuable insights into the molecular regulation of lipid metabolism in
   C. sativa and establish a theoretical foundation for future research.
   Furthermore, these results provide important genetic resources for the
   improvement of oil yield and quality in hemp and other oilseed crops
   through targeted breeding or biotechnological approaches.

Acknowledgments