Overexpression of a Rosa rugosa Thunb. NUDX gene enhances biosynthesis of scent volatiles in petunia

Plant materials

Rosa rugosa ‘Tanghong’, a representative Chinese traditional rose cultivar, was used as experimental material. The three-year-old clonal seedings planted in the field of Rose Resources of Yangzhou University (N32°23′27.64″, E119°25′10.23″) were selected as the flower source. According to previous division standard, the petals of flowers during five stages and different parts of full opening flowers were picked (Feng et al., 2014). For every sample, 3 biological replications were prepared with different ramets and 3–5 flowers of the same plant were selected for each replication.

RNA extraction and purification

Total RNA was isolated from different R. rugosa tissues according to the manufacturer’s instructions of MiniBEST Universal RNA Extraction Kit (TaKaRa, Japan). RNA samples were treated with DNase using DNaseI kit (TaKaRa, Japan) according to the manufacturer’s guidelines, and then quantified by a spectrophotometer (Eppendorf, Germany) at 230 nm, 260 nm and 280 nm.

Cloning of RrNUDX1 gene and sequence analysis

Full cDNA of RrNUDX1 gene was cloned by rapid amplification of cDNA end (RACE). The protocols of 3′ RACE and 5′ RACE were same as previously described in Feng et al. (2014) with the 3′ RACE nest primers (5′-AGCCAAACCATCGCAGTA-3′ for first round 5′-ATGGTTGGGGATGGTATG-3′for second round) and 5′ RACE gene-specific primer (5′-CTGCTGTGCCAATGCTGA-3′). The full-length cDNA of the RrNUDX1 gene was assembled and analyzed using DNAMAN 5.0 software.

Subcellular localization of RrNUDX1 gene

The cDNA was synthesized from 1 µg RNA using PrimeScript RT reagent Kit with gDNA Eraser (TaKaRa, Japan). Complete CDS of RrNUDX1 were isolated from the cDNA. CDS of RrNUDX1 and pBWA(V)HS-GFP vectors were digested by BsaI/Eco31I enzymes, and pBWA(V)HS-NUDX1-GFP fusion expression vector was constructed by using T₄DNA ligase connection, empty vector pBWA(V)HS-GFP was used as control. The two vectors were transferred into the competent cells of Agrobacterium tumefaciens LBA4404 by electroporation in MicroPulser electroporator (Bio-RAD, USA) at voltage of 2.4 KV/5 ms, and the positive bacteria were screened by Kanamycin and PCR. Then the screened positive bacteria were expanded in LB liquid medium to OD₆₀₀ = 0.6, and the lower epidermis of tobacco leaves was injected and infected with these bacteria fluid, the tobacco plants were cultured in low light. After 2 days, tobacco leaves were taken and the fluorescence of the transformed leaf cells was imaged using a confocal laser-scanning microscope (Olympus FV10 ASW).

Temporal and spatial expression analysis

The relative expression of RrNUDX1 were analyzed by real-time quantitative reverse transcription PCR (qRT-PCR). Rosa hybrid α- tubulin subunit actin gene (GenBank accession no. AF394915.1) was used as an internal reference gene (Actin-F: 5′-GCCACCATCAAGACCAAG-3′; Actin-R: 5′-ATCAATGCGGGAGAACAC-3′). Experiments of qPCR were performed with 12.5 µl SYBR^® Premix Ex Taq (2 ×) (TaKaRa, Japan), 1 µl Forward Primer (10 µM, 5′-GCGGTGGTAGTATGCCTGTT-3′), 1 µl 10 µM Reverse Primer (10 µM, 5′-TTCCTTCAGTTCCCTTGCTG-3′), 2 µl cDNA and 8.5 µl ddH₂O. PCR procedure is an initial incubation at 95 °C for 5 min, then followed by 40 cycles of 15 s at 95 °C, 30 s at 60 °C, and 30 s at 72 °C on a BIO-RAD CFX96 Real-Time System (Bio-Rad, USA). The expression level calculation according to the comparative threshold cycle (Ct) (2- Delta DeltaCt) method (Schmittgen & Livak, 2008) followed previous description (Feng et al., 2014).

Gas chromatography with mass spectrometry (GC-MS) analysis of headspace volatiles

1 g of fresh flower tissue and internal standard (3-nonanone, 0.8 µg µL⁻¹, Sigma, USA) were used for the headspace solide-phase microextraction for each sample. The extraction protocol and GC-MS protocol were the same as our previous method with minor parameter adjustments (Feng et al., 2010; Feng et al., 2014). The adjustments included mass spectra scanning at m/z 30–600 amu, more sensitive FFAP elastic quartz capillary vessel column (60 m ×0.32 mm I.D., 1.0µm film, Agilent Corporation, USA) and column temperature program (initial temperature at 50 °C for 1 min, and then increased at 5 °C/min to 120 °C, then increased at 8 °C/min to 200 °C, finally increased at 12 °C/min to 250 °C which was maintained for 7 min).

Qualitative and quantitative analysis of headspace volatiles

Significant data selection by the Xcalibur (a shareware of Thermo Electron Corporation, https://xcalibur.updatestar.com/) and quantitative analysis of the headspace compounds with internal standard 3-nonanone (0.8 µg µL⁻¹) were performed as lab self-built method (Feng et al., 2010).

Overexpression of RrNUDX1 gene in Petunia hybrida

The recombinant pCAMBIA1304-RrNUDX1 expression vector construction and transfection into A. tumefaciens EHA105 competent cell (TransGen, China) followed previous method (Sheng et al., 2018). Transformation to Petunia hybrida cv ‘Mitchell Diploid’ via Agrobacterium-mediated infection on leaf disc were refers to Guo et al. (2014). Petunia explant leaf discs (0.5 cm diameter) cut from 4-week-old seedlings of aseptic seed were pre-cultured on pre-cultured MS medium (30 g/L sucrose, 7 g/L Agar, 3.0 mg/L 6-BA and 0.2 mg/L IAA) at 28 °C for 2 d, then co-incubated with infection liquid where positive EHA105 (OD₆₀₀ 0.8–1.0) was resuspended to OD₆₀₀ = 0.3. After co-incubation for 5 min, the explants with no residual infection liquid were cultured on pre-cultured medium with additional 30 µMol/L acetosyringone in dark for 2 days. Explants were transferred to the screening medium (pre-cultured medium with suited selection pressure) to induce callus and the medium was renewed every 2 weeks to subculture resistant calluses. When the seedlings (2–3 cm adventitious buds) regenerated from the callus, the buds were cut off to subculture on the rooting 1/2 MS medium (30 g/L sucrose, 7 g/L agar, 0.1 mg/L NAA) with selection pressure. The hygromycin selection pressures were 7 mg/L and 6 mg/L at callus induction and buds rooting, respectively, and bacteriostat was 500 mg/L carbenicillin. The regeneration seedings growing up to five leaves with one sprout were transplanted into soil. The culture condition was 16 h light period, light intensity of 200 µmol m⁻²s⁻¹, temperature of 25 °C/23 °C and relative humidity of 70%. Callus GUS staining and PCR were used to select the genomic transgenic petunia. The overexpression level of RrNUDX1 in transgenic lines were detected by qRT-PCR. GC–MS was used to analyze the aroma composition and content of wild–type and transgenic petunia flowers in bloom.

Mathematics statistical methods

The average value of three replicates with standard error (SE) was used as data of each sample. In significance difference test between two sample, independent two-sample t-test was performed for two data groups. In significance difference test of pairwise comparison of more than three samples, one-way ANOVA (Analysis of Variance) and LSD (Least Significant Difference) was performed for multiple comparison. All the calculation was based on the SPSS 18.0 (IBM SPSS Modeler 18.0, https://www.ibm.com/support/pages/downloading-ibm-spss-modeler-180).

Isolation, sequence analysis and subcellular localization of RrNUDX1

We successfully isolated a NUDX1 gene (RrNUDX1, GenBank accession number KX096710.1) related to the monoterpenes biosynthesis of R. rugosa. The full cDNA of RrNUDX1 was 777 bp in length including a 453 bp coding sequence (CDS) which encodes 151 amino acids, the 5′ untranslated region (UTR)(68 bp) and the 3′ UTR (224 bp). RrNUDX1 had high (98%) sequence identity with homology gene of Rosa ×hybrida ‘Papa Meilland’ (RhNUDX1, JQ820249.1) with only 7 SNPs including 3 nonsynonymous SNPs in the CDS. RrNUDX1 protein had higher identity with homology protein of Rosa chinensis (AFW17224.1) (98.67%) than RhNUDX1 (M4I1C6.1) (98%) (Fig. S2). RrNUDX1-GFP fusion protein exhibited no signal in the nucleus of tobacco leaf cells, indicating that RrNUDX1 localized in the cytoplasm (Fig. 1).

RrNUDX1 expression and volatile monoterpenes accumulation in flower

Temporal and spatial expression of RrNUDX1 had a significant difference from budding to senescence in R. rugosa ‘Tanghong’ (Figs. 2A–2D). The expression level of RrNUDX1, which was low in the budding phase, rapidly increased from the early opening stage, reached the highest level at the half-opening stage, started to decrease from the blooming stage to 49.806% of that during the half-opening stage, and then markedly decreased during the senescence phase. The expression level of RrNUDX1 exhibited tissue specificity from the different organs of the flower, and the expression level reached the highest in the petals. However, the expression levels in the stamen and pistil were only 6.9% and 5.2% of that in the petals, respectively. Meanwhile, the expression levels in the anthocaulus, receptacle, and calyx were extremely low.

To verify the relations between the gene expression of RrNUDX1 and the accumulation of the main monoterpenoids in R. rugosa, we measured and determined the aroma component and content. Emphatical analysis was conducted on representative components, such as geraniol, citronellol, nerol (Figs. 2E–2F), and their acetate derivatives, including geraniol acetate, citronellyl acetate, and neryl acetate (Figs. 2G–2H). The contents of the main monoterpenes and their acetate derivatives initially increased and then decreased with the flower development. Geraniol, nerol, geranyl acetate, and neryl acetate reached the highest levels in the blooming stage, whereas citronellol and citronellyl acetate reached their highest levels during the half-opening stage. The total content of the six main aromatic components in the half-opening stage was 10.12 µg/g higher than that in the blooming stage. The six main aromatic components mainly originated from the petal and stamen. The amount of the three monoterpene alcohols in the petal was more than four (4.34) times that in the stamen, whereas the amount of the three acetate derivatives in the stamen was 192.21 µg/g, which was more than three times that in the petal. Several aroma components were detected in the stamen, whereas no monoterpene alcohols or acetate derivatives mentioned above were found in the anthocaulus, receptacle, and calyx.

Overexpression vector construction of RrNUDX1 gene and genetic transformation to petunia hybrida

The petiole callus transformed by 35S: RrNUDX1 vector were stained with varying degrees of blue by GUS staining (Fig. S3). RrNUDX1 was successfully integrated into the genome of five petunia lines by PCR detection of six randomly selected lines (Fig. S4) and expressed significantly (Fig. S5). The wild-type and transgenetic petunia plants significantly differed in phenotype under the same conditions. Compared with those of the wild type, the leaves of the petunia with RrNUDX1 gene transformed were significantly wider, with 10.83 mm and 8.19 mm larger width and length than the wild type; moreover, the transgenic plants grows faster compared with the wild-type plants (Fig. 3, Table 1).

GC–MS were used to detect the aroma components of the petunia flowers in bloom, and 10 major aroma components were selected for statistical analysis (the sum of the 10 aroma components accounted for 80.10% of the total aroma components in petunia) (Table 2). Results showed that the contents of all ester fragrance ingredients, except for benzyl benzoate, increased in the transgenic petunia plants. In particular, the increase in the content of methyl benzoate was the most significantly. In the transgenic Line1, the content of methyl benzoate was 63.04 µg/g, which was 1.69 times that in the wild-type plants. The alcohols were mainly benzyl alcohol and phenylethyl alcohol, and the contents of benzyl alcohol in transgenic Line1 and Line2 were 25.58% and 53.49% higher than those in the wild-type plants. In addition, the contents of the two main phenolic compounds also increased, and the contents of isoeugenol in the three transgenic Lines were 3.2, 4.3, and 4.9 times those in the wild-type plants.