Non-specific phospholipase C4 hydrolyzes phosphosphingolipids and phosphoglycerolipids and promotes rapeseed growth and yield

The NPC family in plants and NPC expression in rapeseed

To analyze the diversity and conservation of the NPC family in plants, we used Arabidopsis NPC sequences as queries to search the genomes of seven plant species encompassing bryophytes, lycophytes, monocots, and dicots. This analysis identified 43 NPC sequences, all of which contain a phosphatase domain (Pfam entry PF04185) with the four conserved motifs ENRSFDxxxG, TxPNR, DExxGxxDHV, and GxRVPxxxxxP. The NPC proteins from various plant species cluster into four clades, corresponding to NPC1, NPC2, NPC3–5, and NPC6 according to the nomenclature of the Arabidopsis NPC family (Figure 1A). Plant NPCs can be divided into two subtypes based on the presence of an N-terminal signal peptide (SP-NPC): NPC1, NPC2, and NPC6 contain an SP-NPC, whereas NPC3–NPC5 for the plant species analyzed lack the signal peptide but contain a conserved cysteine (Cys) in their C termini. The only exception is NPC5 from Arabidopsis, which lacks this Cys residue. We showed recently that this Cys in NPC4 was acylated and that this acylation was required for the membrane association of NPC4 and its hydrolysis of GIPC (Yang et al., 2021b). Bryophytes and lycophytes contain only SP-NPCs, whereas angiosperms (monocots and dicots) have the two NPCs, indicating that SP-NPCs are more ancient than the non-SP-NPCs.

The rapeseed genome contains 13 BnaNPC genes comprising two NPC1, three NPC2, two NPC3, two NPC4, and four NPC6 genes distributed on the A and C subgenomes (Figure 1A). Based on expression data from the Brassica napus multi-omics information resource (BnIR, http://yanglab.hzau.edu.cn/) (Liu et al., 2021a), BnaNPC1 and BnaNPC6 genes were highly expressed in various tissues, whereas BnaNPC2 genes were highly expressed in stems and developing seeds. BnaNPC3 genes showed high expression in roots, stems, flowers, silique walls, and early developing seeds, and BnaNPC4 genes were mainly expressed in roots and seeds at late stages of development under normal growth conditions (Figure S1A). We used reverse transcription quantitative polymerase chain reaction (RT-qPCR) to verify the bulk expression pattern of BnaNPC4 genes (Figure 1B), as BnaNPC4 genes from the A and C subgenomes cannot be distinguished by this technique due to high sequence similarity (Figure S2). To determine the expression of BnaNPCs under P_i deficiency, we analyzed the transcriptome data of leaves at the three-leaf and five-leaf stages from seedlings that had experienced 7 d of P_i deficiency (Chao et al., 2020), revealing that among of the NPC family, only BnaNPC4 genes display increased expression under P_i deficiency conditions (Figure S1B). RT-qPCR also verified that BnaNPC4 genes are highly induced in various tissues after 7 d of P_i deficiency (Figure 1C).

Genome editing and overexpression of NPC4 in rapeseed

To investigate the effect of BnaNPC4 on rapeseed growth in response to P_i deficiency, we used clustered regularly interspaced short palindromic repeat (CRISPR)/CRISPR-associated nuclease 9 (Cas9)-mediated genome editing to knock out BnaNPC4s in the A and C subgenomes. To this end, we designed a single guide RNA (sgRNA) targeting the second exon to edit both BnaC01.NPC4 and BnaA01.NPC4 because it is in a region identical in the two genes (Figure S2). We identified 20 plants that were positive for CRISPR/Cas9 editing based on detection of the T-DNA insertion carrying Cas9 in the T₀ transgenic generation. Sequencing of the target site confirmed the editing of these 12 plants. We obtained four homozygous mutant lines through self-pollination over several generations. Two of the resulting CRISPR/Cas9-edited lines, designated C8 and C10, each harbored a 1-bp insertion; line C9 had a 5-bp deletion, and line C11 carried a 2-bp deletion (Figure 1D, E). All mutations resulted in the introduction of frameshifts into the open reading frame and a premature termination of BnaNPC4 translation, suggesting that these mutants are loss-of-function alleles. We refer to these lines as knockout (KO) mutants hereafter.

In parallel, we overexpressed BnaC01.NPC4 by introducing the 35 S:BnaNPC4-GFP (an expression cassette consisting of BnaNPC4 cloned in-frame and upstream of the green fluorescent protein (GFP) sequence and driven by the cauliflower mosaic virus (35 S) promoter) construct into rapeseed using the complementary DNA (cDNA) of BnaC01.NPC4. We identified positive transgenic T₀ plants by PCR, yielding eight homozygous lines with a high level of BnaNPC4 transcript levels after self-pollination for several generations (Figure 1F). The relative BnaNPC4 transcript levels in overexpression (OE) lines OE3 and OE4 were approximately 1,200- and 400-fold higher under normal P conditions, respectively, than those of non-transgenic controls, with the other lines showing BnaNPC4 fold increases over controls ranging from several to dozens of times higher (Figure 1F). We also detected the BnaNPC4-OE-GFP protein in OE3 and OE4 lines by immunoblotting with an anti-GFP antibody (Figure 1G).

BnaNPC4 promotes growth and seed production in rapeseed

To determine the effect of BnaNPC4 on rapeseed growth, we characterized the phenotype of homozygous transgenic lines for BnaNPC4 (OE and KO) relative to that of non-transgenic control plants (referred to hereafter as wild type (WT)). The hypocotyl length of 7-d-old seedlings for BnaNPC4-OE lines was significantly longer than that of WT, whereas we observed the opposite for the KO mutants, C8 and C9 (Figure S3). In all genotypes, plant growth was slower in the absence of any supplied P_i than under P_i-sufficient conditions. However, compared to WT, the two BnaNPC4-OE lines displayed significant increases in growth, while that of the BnaNPC4-KO lines was impeded, under both conditions (Figure 2A–C). Under P_i-sufficient conditions, the average plant height, root length, and fresh weight increased by 8.1%, 11.2%, and 15.6% in BnaNPC4-OE lines, but decreased by 4.2%, 12.5%, and 10.8% in BnaNPC4-KO lines, compared to WT (Figure 2D–F). Under P_i-deficient conditions, the average plant height, root length, and fresh weight increased by 6.1%, 9.5%, and 22.6% in BnaNPC4-OE lines, but decreased by 8.8%, 10.8%, and 11.3% in BnaNPC4-KO lines, relative to WT (Figure 2D–F). These results indicate that BnaNPC4 promotes seedling growth under both P_i-sufficient and P_i-deficient conditions.

We grew WT, BnaNPC4-KO, and BnaNPC4-OE plants to maturity in soil with different levels of supplied P_i (50, 200, or 1,000 µmol/L) to monitor their agronomic traits. Under sufficient P_i supply (1,000 µmol/L), BnaNPC4-OE lines produced 10.4% more seeds than WT, whereas KO lines had a roughly equal number of seeds to WT. However, under limited P_i supply (50 and 200 µmol/L P_i), BnaNPC4-KO plants had a lower seed yield than WT. Indeed, when grown with 50 µmol/L or 200 µmol/L P_i, the seed yield of the KO lines was 15.1% and 23.1% lower on average, respectively, than that of WT, whereas the seed yield of BnaNPC4-OE lines was on average 25.3% and 17.7% higher, respectively, than that of WT (Figure 2G). BnaNPC4-OE plants were taller, with longer inflorescences, whereas KO plants were shorter and had shorter inflorescence than WT (Figure 2G). BnaNPC4-OE plants tended to have more effective branches (seed-bearing branches) and more siliques, whereas KO plants showed the opposite trend (Figure S4). Seed weight was similar among WT, KO, and OE plants (Figure S4). These results indicate that increased NPC4 expression promotes overall plant growth and seed production under P_i-sufficient and P_i-deficient conditions, and the effect of BnaNPC4-KO and BnaNPC4-OE on rapeseed growth and seed production is greater under P_i-deficient than P_i-sufficient conditions.

BnaNPC4 alters sphingolipid metabolism in rapeseed

To determine how BnaNPC4 affects plant growth, we analyzed the sphingolipid levels of BnaNPC4-KO and BnaNPC4-OE lines in rapeseed, as NPC4 in Arabidopsis hydrolyzes GIPC to produce hydroxyceramide (hCer) (Yang et al., 2021a, 2021b). Accordingly, we transferred 7-d-old seedlings that had germinated in distilled water to P_i-sufficient and P_i-deficient hydroponic conditions for 2 weeks (Figure 2A). Compared to those in WT roots, the GIPC levels in BnaNPC4-OE roots were 21.2% lower under sufficient P_i supply, but 36.3% lower in the P_i-deficient condition (Figure 3A, B). By contrast, BnaNPC4-KO roots accumulated 23.1% more GIPC under P_i sufficiency, and 54.2% more GIPC under P_i deficiency, than WT roots (Figure 3A, B). Conversely, the levels of the non-phosphosphingolipids hCer and GlcCer were higher in BnaNPC4-OE roots, but lower in BnaNPC4-KO roots, compared to WT roots (Figure 3C, D, E). In addition, we determined that the levels of multiple GIPC and hCer species were altered in BnaNPC4-OE and BnaNPC4-KO roots (Figure 3B, D; Table S1). The opposite patterns seen for GIPC and hCer levels in BnaNPC4-OE and BnaNPC4-KO lines suggest that BnaNPC4 hydrolyzes phosphosphingolipids in roots.

In leaves, BnaNPC4-OE plants displayed a 25% decrease in GIPC levels relative to WT (Figure 4A, B) and a 39.7% increase in hCer (Figure 4C, D). By contrast, BnaNPC4-KO mutant leaves had levels of GIPC and hCer similar to those of WT leaves (Figure 4A–D; Table S1). However, under P_i deficiency, BnaNPC4-OE and BnaNPC4-KO leaves showed an opposite change in GIPC levels, with total GIPC levels being 27.5% lower in the OE lines, but 17.9% higher in the KO lines, compared to WT leaves (Figure 4A, B). The hCer and GlcCer levels in leaves were comparable among the leaves of WT, BnaNPC4-OE, and BnaNPC4-KO plants (Figure 4C, D, E). The lack of effect on leaf sphingolipids under sufficient P_i in BnaNPC4-KO is consistent with the lower BnaNPC4 expression seen in leaves under sufficient P_i supply (Figures 1B, S1A). During P_i deficiency, NPC4 expression is induced in leaves, resulting in a greater effect in the NPC4-KO on leaf GIPC levels under P_i-deficient compared to P_i-sufficient conditions. These results support the notion that NPC4 hydrolyzes GIPC in leaves under P_i deficiency conditions.

BnaNPC4 changes phosphoglycerolipid composition

We also analyzed the effect of membrane glycerolipid composition in BnaNPC4-OE and BnaNPC4-KO plants. Under sufficient P_i supply, the roots of BnaNPC4-OE plants accumulated lower levels of PC (10.6%), PE (6.0%), and phosphatidylserine (PS) (15.5%) (Figure 5A–C), with most phosphoglycerolipid species showing significantly lower levels (Figure 5D–H; Table S1). Under P_i deficiency conditions, the roots of BnaNPC4-OE plants exhibited a further decrease in their phosphoglycerolipid levels and an increase in diacylglycerol (DAG) levels relative to WT roots. The levels of PC, PE, and PS in OE3 roots were 13.7%, 10.0%, and 30.8% lower than that of WT roots, respectively (Figure 5A–C), with most phosphoglycerolipid species displaying significantly lower abundances in roots (Figure 5D, E, H; Table S1). Conversely, the DAG level in BnaNPC4-OE roots was 27.7% higher than that of WT roots (Figure S5A). By contrast, the roots of BnaNPC4-KO mutant and WT plants had similar levels of PC, PE and PS under P_i-sufficient and P_i-deficient conditions (Figure 5A–C). For non-phosphoglycerolipids, compared to WT, the levels of total DAG and five DAG species were decreased in the roots of BnaNPC4-OE plants (Figures S5A, S6A; Table S1). BnaNPC4-OE roots displayed a 37.1% increase in monogalactosyldiacylglycerol (MGDG) levels (Figures S5B, S6B) and a 16.1% increase in DGDG levels (Figures S5C, S6C), with most molecular species showing significant increases in these BnaNPC4-OE lines (Figure S6; Table S1). Similarly, compared to WT, BnaNPC4-OE leaves had lower levels of PC (11.7%), PE (6.3%), and PS (19.0%) than WT leaves (Figure 5A–C), with most phosphoglycerolipid species displaying significant decreases (Figure 5F–H; Table S1). Under P_i deficiency conditions, BnaNPC4-OE leaves exhibited a further decrease in phosphoglycerolipid levels, with the abundances of PC, PE, and PS being 25.5%, 24.4%, and 8.2% lower, respectively, than those in WT leaves (Figure 5A–C), while most phosphoglycerolipid species displayed significant decreases (Figure 5F–H; Table S1).

By contrast, the leaves of BnaNPC4-KO mutant lines had levels of total PC, PE, and PS comparable to those of WT leaves under sufficient P_i supply (Figure 5A–C). The levels of total MGDG and DGDG were also similar between WT and BnaNPC4-KO leaves (Figure S5B, C), with the exception of DAG content, which was lower in BnaNPC4-KO leaves than in WT leaves (Figure S5A). As noted earlier, the lack of effect in the BnaNPC4-KO mutants is consistent with the lower BnaNPC4 expression detected in leaves under P_i-sufficient conditions (Figures 1B, S1A). By contrast, under P_i deficiency conditions, the leaves of the BnaNPC4-KO mutants accumulated more PC (6%), PE (5%), and PS (13.7%) than WT leaves (Figure 5A–C), with most phosphoglycerolipid species displaying significant increases (Figure 5F–H; Table S1). For non-phosphoglycerolipids, compared to WT, the levels of three DAG species were lower in the leaves of BnaNPC4-KO mutant plants (Figure S5A; Table S1), while total DGDG levels increased by 35.2% in the leaves (Figure S5C), with most non-phosphoglycerolipid species showing significant increases in the leaves of BnaNPC4-OE lines (Figure S6; Table S1). The greater effect under P_i-deficient conditions relative to P_i-sufficient conditions in the mutants support the idea that P_i-deficiency-induced BnaNPC4 contributes to glycerolipid hydrolysis. The opposite changes seen for phosphoglycerolipids in BnaNPC4-OE and BnaNPC4-KO mutant lines suggest that BnaNPC4 uses phosphoglycerolipids as substrate, thereby altering the overall phosphoglycerolipid composition, in rapeseed.

BnaNPC4 hydrolyzes both phosphosphingolipids and phosphoglycerolipids

To test whether BnaC01.NPC4 encodes an active enzyme hydrolyzing phosphosphingolipids and phosphoglycerolipids, we produced His-tagged BnaC01.NPC4 in Escherichia coli. The coding sequence of BnaC01.NPC4 is 1,611 bp, coding for 537 amino acids, and the target protein with His-taq (BnaC01.NPC4-6xHis) was 68 kDa (Figure 6A). We established that recombinant purified BnaC01.NPC4-6xHis hydrolyzes PC to produce DAG, whereas we detected no such activity when using a protein preparation obtained from E. coli cells transformed with the empty vector as control (Figure 6B). BnaNPC4 also hydrolyzed other phosphoglycerolipids tested, including PE, PG, PI, PS, and PA, to generate DAG (Figures 6C, S7).

When the phosphosphingolipid GIPC was used as substrate for recombinant BnaC01.NPC4-6xHis, we observed a decrease in the level of GIPC with a concomitant increase in the lipid product hCer (Figure 6D) and soluble head-group phosphate (Figure 6E). These results indicate that BnaNPC4 hydrolyzes both phosphoglycerolipids and phosphosphingolipids and can use various classes of phospholipids as substrates in vitro. This finding is consistent with our recent protein structure analysis of NPC4 revealing how the large exposed cleft of NPC4 enables the recognition of phospholipids with different molecular sizes (Fan et al., 2023).

BnaNPC4-KO and BnaNPC4-OE plants show opposite effects on the expression of genes involved in P_i responses and P_i content

During membrane lipid remodeling under P_i deficiency, the decrease in phospholipid contents is accompanied by an increase in non-phospholipids to maintain membrane integrity and function. DAG and hCer released by NPC4 can be converted to non-phospholipids (glycolipids and triacylglycerols (TAGs)) by non-phospholipid synthases, such as digalactosyldiacylglycerol synthase 2 (DGD2), monogalactosyldiacylglycerol synthase type C (MGDC), UDP-sulfoquinovose synthase 2 (SQD2), and diacylglycerol acyltransferase (TAG1). Similarly, the phosphorylated head groups produced by NPC can be dephosphorylated by phosphatases, such as PAP, phosphoethanolamine/phosphocholine phosphatase 1 (PEPC1), and phosphate starvation-induced gene 2 (PS2), to release P_i for cellular use (Angkawijaya and Nakamura, 2017). P_i is further redistributed for plant cellular and growth needs within by the plasma membrane phosphate transporters PHT1 and PHO1 and by intracellular transporters, such as PHO1, PHT2, PHT3, PHT4, and PHT5 (Abbas et al., 2018; Dissanayaka et al., 2021; Shi et al., 2021).

Thus, we monitored the transcript levels of selected genes involved in lipid metabolism and PSR. Compared to WT, the leaves of BnaNPC4-OE plants showed higher transcript levels of the non-phospholipid biosynthesis genes SQD2, DGD2, TAG1, and MGDC, the phosphatase genes GLYCEROPHOSPHODIESTER PHOSPHODIESTERASE 5 (GDPD5), MYO-INOSITOL POLYPHOSPHATE 5-PHOSPHATASE (IP5PII), PAP8, and PAP17, the P homeostasis regulator gene CATION EXCHANGER 3 (CAX3), and the P_i transporter genes PHO1;H3, PHT2;1, and PHT4;4 (Figure 7A). By contrast, the leaves of BnaNPC4-KO mutant plants had lower transcript levels of these genes (Figure 7A). The observed pattern in the expression levels of these genes in BnaNPC4-OE and BnaNPC4-KO lines suggests that BnaNPC4 promotes membrane phospholipid hydrolysis and increases the expression of genes involved in P_i metabolism, transport, and response.

To examine the effect of BnaNPC4 on P_i usage, we assayed the total P (TP) and free P_i content in the leaves of WT and BnaNPC4-OE and BnaNPC4-KO lines in rapeseed grown in soil under P_i-sufficient and P_i-deficient conditions. BnaNPC4-OE had higher levels of TP and free P_i under both P_i-sufficient and P_i-deficient conditions. By contrast, BnaNPC4-KO accumulated as much TP and free P_i as WT under sufficient P_i supply. However, under P_i-deficient conditions, BnaNPC4-KO plants showed lower TP and free P_i levels in their leaves compared to WT. In addition, the magnitude of the decrease in TP and free P_i increased with a decrease in P_i supply from 200 µmol/L to 50 µmol/L P_i (Figure 7B). This finding is consistent with the higher expression of phosphatase genes and P_i transporter genes in the BnaNPC4-OE lines (Figure 7A).

Plant material and growth conditions

The ‘Westar’ cultivar of rapeseed (Brassica napus) was used for genetic manipulations and physiological experiments. Hydroponic experiments were conducted using a modified plant culture Hoagland solution containing 5 mmol/L KNO₃, 1 mmol/L KH₂PO₄ (for P_i-sufficient conditions) or K₂SO₄ (for P_i-deficient conditions), 2 mmol/L MgSO₄, 5 mmol/L Ca(NO₃)₂, 46 μmol/L H₃BO₃, 0.32 μmol/L CuSO₄, 0.77 μmol/L ZnSO₄, 9.14 μmol/L MnCl₂, and 0.37 μmol/L Na₂MoO₄ with 50 μmol/L ethylenediaminetetraacetic acid (EDTA)-Fe (II), at pH 5.8. Seven-d-old seedlings germinated and grown in pure water were transferred to P_i-sufficient or P_i-deficient conditions for several weeks, at which point the photographs were taken. Rapeseed plants were grown in a growth room with a 16-h light/8-h dark photoperiod at 25°C/21°C (day/night), 50% humidity, and a light intensity of 200 µmol/m²/s.

Rapeseed plants were grown in an air-conditioned, climate-controlled (Argus Control Systems) greenhouse set at 20°C–21°C, 40%–60% humidity, and 14-h light/10-h dark photoperiod with supplemental light threshold of 566 µmol/m²/s¹ (with supplemental lights turning off when outside solar radiation was 566 µmol/m²/s¹ or above). The shade curtain was automatically pulled to 50% when sunlight was over 700 W/m² and was pulled down 100% when the sunlight was over 800 W/m². Seeds were sown in soil (vermiculite: perlite; 3:1, v/v). One seed per genotype was sown per pot (3.79 L), with 10 pots per genotype. Seven pots were placed on a shelf (52 × 118 cm; width × length). Therefore, each individual plant had a growth area of 876.6 cm². All pots were randomly positioned on the shelf. Plants were fertilized with Hoagland solutions supplemented with 1,000 µmol/L P, 200 µmol/L P, or 50 µmol/L P by dipping the pots in the solution for 5 min once a week (young seedlings), or every 5 d (adult plants). After fertilization, the pots were randomly placed back on the shelf. Such watering with fertilization continued until at least half of the seed pods turned brown and was stopped once all pods were brown. Leaves were harvested from 3-week-old seedlings for determination of total P and soluble inorganic P contents. The agricultural traits were measured with 8-week-old plants or mature plants.

Phylogenetic analysis

Amino acid sequences of NPCs using Arabidopsis proteins as the query were retrieved from UniProt (https://www.uniprot.org/). A multiple sequence alignment of NPCs was generated by ClustalW and the phylogenetic tree was reconstructed by the neighbor-joining method using MEGA 7.0. The phylogenetic tree was modified by EVOLVIEW (https://www.evolgenius.info/evolview/) (Zhang et al., 2012). The functional domains of NPCs were determined by European Bioinformatics Institute Search InterPro (http://www.ebi.ac.uk/interpro/search/sequence/).

RNA extraction and qRT-PCR

Total RNA was extracted from tissues, such as leaves, stems, roots, flowers, silique wall, and seeds, of rapeseed using Trizol reagent. The isolated RNA was used as template for first-strand cDNA synthesis through RT using an iScript kit (Bio-Rad). qPCR was performed with SYBR green fluorescent labeling of double-stranded DNA using a CFX Connect instrument (Bio-Rad). The expression level was normalized to that of ACTIN7 (BnaA06G0089200WE). PCR reactions were as follows: one cycle at 95°C for 1 min; 45 cycles at 95°C for 15 s, 60°C for 15 s, and 72°C for 20 s; and final extension of DNA at 72°C for 5 min. The qPCR primers and description of the genes are listed in Tables S2 and S3.

Protein production, purification, and NPC activity assay

The full-length coding sequence of NPC4 was amplified from first-strand cDNA prepared from total RNA of rapeseed using forward (5′-CGGAATTCATGGCACAGACGACTTTAGGCTCAG-3′, EcoRI) and reverse primers (5′-CCCTCGAGTCAATCATGACTAGCAAAGCAAGAGAAC-3′, XholI). After digestion with the appropriate restriction enzymes, the fragment was cloned into the vector pET28a(+). The resulting NPC4-6×His vector was transformed into the E. coli strain BL21 (DE3). The NPC4-6xHis fusion protein was produced and purified as described previously. Isopropyl β-D-1-thiogalactopyranoside (0.1 mmol/L final concentration) was added to the bacterial culture when it reached an optical density at 600 nm of 0.6 to induce NPC4 protein production for 16 h at 16°C (Yang et al., 2021a). The bacterial cells were harvested by centrifugation at 15,000 × g for 10 min at 4°C and resuspended in extraction buffer containing 50 mmol/L Tris-HCl, pH 7.3, 50 mmol/L NaCl, 5% (v/v) glycerol, 1 mmol/L dithiothreitol (DTT), and 0.5 mmol/L phenylmethylsuflonyl fluoride (PMSF). The bacterial cells were lysed by sonication and centrifuged at 15,000 × g for 30 min at 4°C to remove cell debris. For purification, Ni-NTA agarose (GE Healthcare) was used according to the manufacturer's instructions. Briefly, 0.5 mL of nickel-nitrilotriacetic acid (Ni-NTA) agarose was added to 3 mL bacterial supernatant and incubated for 1 h with gentle agitation at room temperature. Ni-NTA agarose beads were loaded onto a column and washed four times using wash buffer containing 50 mmol/L Tris-HCl, pH 7.3, 50 mmol/L NaCl, 5% (v/v) glycerol, 1 mmol/L DTT, 50 mmol/L imidazole and 0.5 mmol/L PMSF. The purified proteins were separated by 10% (w/v) sodium dodecyl sulfate – polyacrylamide gel electrophoresis, the resulting gels were stained with Coomassie Brilliant Blue, transferred to polyvinyl difluoride membranes, and then subjected to immunoblotting using His-tag antibodies (1:3,000 dilution; A-5588, Sigma).

NPC activity was assayed as previously described (Nakamura et al., 2005; Peters et al., 2010; Yang et al., 2021a). The reaction mixture contained 10 μg purified NPC4-6xHis and lipids that were emulsified in reaction buffer (25 mmol/L HEPES, pH 7.5, 10 mmol/L CaCl₂, and 10 mmol/L MgCl₂) by sonication on ice for 5 min in a final volume of 200 µL. The reaction was incubated at 30°C for 30 min and stopped by adding 200 µL of butanol followed by vigorously vortexing. The sample was centrifuged at 12,000 × g for 3 min at room temperature and 200 µL of the lower phase was used to measure the phosphate released from the head group, as determined by molybdenum blue staining. For thin-layer chromatography (TLC), the solvent from the organic, upper phase of the 200 µL butanol extract from the reaction was evaporated under a stream of N₂, and the residue was dissolved in 20 µL chloroform and separated on a TLC plate (Yang et al., 2021a). GIPC was chromatographed in a mixture of CHCl₃, CH₃OH, and 4 mol/L NH₄OH (9:7:2, v/v/v) with 0.2 mol/L ammonium acetate. DAG was chromatographed in a mixture of petroleum ether, diethyl ether, acetic acid (50:50:1, v/v/v) and visualized by copper sulfate staining as described (Kim et al., 2009).

Genomic editing and plant transformation of NPC4 in B. napus

Vectors for NPC4 OE and CRISPR/Cas9-mediated gene editing were constructed and introduced into the Westar cultivar using a method previously described (Dai et al., 2020). In brief, the full-length NPC4 coding sequence was cloned from first-strand cDNA prepared from total RNA of rapeseed leaves into the pMDC83 vector after digestion with SpeI and AscI. The primers for the amplification of NPC4 were as follows: forward primer (5′-GGACTAGTATGGCACAGACGACTTTAGGCTCAG-3′) and reverse primer (5′-TTGGCGCGCCTCAATCATGACTAGCAAAGCAAGAG-3′). The NPC4 knockout target site was designed to target the second exon of NPC4 by CRISPR-P2.0 (http://crispr.hzau.edu.cn/CRISPR2/) (Lei et al., 2014). Oligonucleotides for the target site annealed to form double-stranded DNA was inserted into pKSE401 (Zhang et al., 2020b). The vectors were transformed into the Agrobacterium tumefaciens strain GV3101. Transgenic plants were produced with the resulting plasmids via Agrobacterium-mediated transformation of rapeseed as previously described (Dai et al., 2020).

The homozygosity of the T-DNA insertion in NPC4-OE lines was verified by PCR-based genotyping using a T-DNA left border primer and gene-specific primers. NPC4 expression in WT and OE were analyzed by RT-PCR; NPC4 protein abundance was checked by immunoblotting with an anti-GFP antibody (1:3,000 dilution; sc-9996, Santa Cruz) conjugated with horseradish peroxidase (HRP). CRISPR/Cas9 knockout mutants of NPC4 were identified by Sanger sequencing of the target sites using upstream and downstream primers. Primers used for the PCR verification of NPC4-OE and mutant are listed in Table S2.

Lipids analysis

Lipids were extracted as previously described (Markham and Jaworski, 2007; Vu et al., 2014). Briefly, 10 to 30 mg of freeze-dried leaf and root samples were homogenized and resuspended in 3 mL extraction solvent H (isopropanol, heptane, water mixture, 55:20:25, v/v/v). The supernatant was collected and dried under a stream of nitrogen gas. Samples were heated in 1 mL of tetrahydrofuran (THF), methanol, water mixture (2:1:2, v/v/v) containing 0.1% (v/v) formic acid, dissolved by ultrasound sonication, and then centrifuged at 500 × g for 10 min to remove insoluble substances. Total lipids were analyzed using an Exion ultra-performance liquid chromatography system coupled with a triple quadrupole/ion trap mass spectrometer (6500 Plus QTRAP; SCIEX) according to a previous method with modifications (Vu et al., 2014; Yang et al., 2021a).

Measurement of P concentration

Total P and soluble P_i concentrations were measured based on a modified ammonium molybdate staining method (Feng et al., 2011). For total P measurement (Hazel et al., 2014), 10–20 mg freeze-dried samples were digested in 3 mL HNO₃ at 100°C for 1 h. After cooling for 10 min, 1 mL of HClO₄ was added, and the acid digests were heated to 120°C for 1 h until the tissues were completely broken down. After cooling, samples were diluted with 2 mL of double-distilled water. Aliquots (10–50 µL) of samples were diluted to 800 µL and then 150 μL of molybdate solution (17.55 g/L (NH₄)₆ Mo₇•4H₂O, 2 mol/L H₂SO₄) was added. After 10 min, 50 μL of green malachite solution (3.5 g/L polyvinylic alcohol, 0.35 g/L malachite green) was added and allowed to stand for 2 h. The absorbance of the reaction was measured at 650 nm and the total P concentrations were calculated from a standard curve constructed using standard solutions of KH₂PO₄. To determine soluble inorganic phosphorus concentrations (Staudinger et al., 2022), dried leaves were ground to powder and extracted with hot water (10% w/v) at 85°C for 40 min. The extracts were centrifuged at 5,000 × g for 2 min at room temperature and aliquots of the supernatant were diluted to 650 µL. The diluted sample was mixed with 50 μL 10% (w/v) ascorbic acid and 300 μL of a 0.44% (w/v) ammonium paramolybdate solution (in 1 NH₂SO₄). After the mixture was incubated at 42°C for 1 h, absorbance was determined at 820 nm from a standard curve with KH₂PO₄. Experiments included five biological repeats.

Statistical analysis

Values are means ±  SD. Different letters indicate differences at P < 0.05 among genotypes during different growth conditions based on a one-way analysis of variance (ANOVA). * is significant at P < 0.05; ** is significant at P < 0.01 compared to the control based on Student's t-test.