The cv. Nipponbare was used to determine the expression patterns of OsBADH2. Total RNAs were extracted from roots, stems, leaves and flowers of mature plant, as well as embryos, seedling roots, seedling stems, seedling leaves. Messenger RNA levels were analyzed by semi-quantitative RT-PCR. The results show OsBADH2 is expressed constitutively, with less expression abundance in mature roots (Figure 1).

We examined genetic variation in a 463 bp portion of OsBADH2 gene in 13 fragrant and six non-fragrant rice accessions that are representatives of cultivated varieties across China (Table 1). Multiple mutations identical to fgr allele identified by Bradbury et al. 22 were also observed in all the 13 fragrant accessions (data not shown). These mutations contain a total of six SNPs and eight deletions within a 25 bp region of the exon 7 as compared to the wild type allele in the tested non-fragrant varieties.

To dissect the physiological role of OsBADH2, we generated a large number of transgenic rice plants expressing OsBADH2-RNAi (Figure 2A). Construct was introduced into non-fragrant rice cv. Nipponbare by Agrobacterium tumefaciens-mediated T-DNA transfer. The majority of OsBADH2-RNAi primary transgenics (T0) resulted in PCR amplification with primers designed to the hpt (hygromycin phosphotransferase gene) marker. Preliminary aroma evaluation revealed that 21 out of 97 hpt-positive T0 plants showed abundant fragrance emission.

Ten T1 segregating populations derived from independent T0 plants with obvious aroma production were selected and cultivated in the university's farm. Leaves from the first and second crops and grains from the first crop of individual T1 plants were used for sensory evaluation of aroma production. As a result, the presence and absence of fragrance is unambiguously segregated within T1 populations and well consistently expressed in leaves and grains from the same plants. In all the 10 T1-segregating populations, we observed a correlation of OsBADH2-RNAi transgene integration with apparently sensory aroma production.

To investigate property of the emitted aromatic compound from the transgenic lines, a gas chromatography-mass spectrometry (GC/MS) analysis was conducted for Thai Hom Mali 105 (aromatic rice from Tailand), transgenic line (OsB2-Rc) and the wild type control. As shown in Figure 3, a considerable aroma production of 2-acetyl-1-pyrroline (2AP) was detected in both Thai Hom Mali 105 and transgenic line (OsB2-Rc), by contrast to the undetectable wild type control.

In order to test if the aroma production is co-segregated with the presence of T-DNA insertion in the transgenic segregating progeny, T2 mature leaves from five OsBADH2-RNAi repression plants and four non-fragrant wild-type plants segregated out from the same transgenic line (OsB2-Rc) were collected and subjected to the GC-MS analysis. As a result, 2AP accumulation was apparently enhanced in all the downregulation plants, by contrast to the undetectable wild-type plants without fragrance (Figure 3D–F).

To ensure that the observed phenotypes are correlated with reduced endogenous transcript, total RNA was extracted from leaves of wild type plants and four independent OsBADH2-RNAi repression lines. Analysis of semi-quantitative RT-PCR and real-time quantitative RT-PCR revealed a considerable reduction in endogenous OsBADH2 transcript levels in OsBADH2-RNAi repression lines compared to that of wild-type plants (Figure 2B).

To determine if the expression of OsBADH1 gene is affected in the down-regulated OsBADH2 transgenic lines, we conducted real-time PCR analysis to detect the OsBADH1 cDNA levels in the OsBADH2-RNAi repression lines. The OsBADH1 gene is the closest available homolog of OsBADH2 and is almost unaffected in the down-regulated OsBADH2 lines (Figure 2B).

To investigate the effects of down-regulated OsBADH2 on crop productivity, the plant height and 1000-grain weight of OsBADH2-deficient and wild type plants derived from four independent T2 transgenic lines were measured. The results showed the reduction in plant height as well as 1000-grain weight in the OsBADH2-deficient plants is differentially detectable when compared to the wild-type plants segregated out from the transgenic progenies (Table 2).

To further address the physiological functions of the OsBADH2 gene, wild type seeds and T2 seeds derived from four independent transgenic lines were germinated in various salt-stressed conditions. No significant difference in sprouting index and germination rates between wild type and OsBADH2 repression lines was observed (Figure 4A, B). Nevertheless, the seedling growth rates (scored by shoot length and weight, root length and weight and root number) in different salt concentrations were more or less inhibited in OsBADH2-deficient lines as compared to wild type (Figure 4C–G).

Thirteen fragrant and six non-fragrant cultivars from across China were used to determine the allelic variation of OsBADH2 locus. The plants were grown in farm's field at the Sichuan University. Primary transformants (T0) were first planted in the artificial climate incubators (BINDER, Tuttlington, Germany) under standard conditions (28°C day, 20°C night; 12 h light, 12 h dark), and transplanted into the field 5 weeks later. Wild type and the transgenic progeny plants were grown side by side in farm's field at the Sichuan University. After harvesting, the T1 stubs were transferred into pots and grown in a growth chamber (25°C, 12/12 h photoperiod at 200 μmol photons m^-2 s^-1). Tillers were regenerated from the stubs. To investigate the germination and seedling phenotypes, wild type and T2 seeds derived from the T1 transgenic plants were germinated under various concentrations of NaCl in the growth chamber.

DNA manipulations were carried out by using standard procedures (Molecular Cloning). Sequences from OsBADH2 cDNA (GenBank accession no. AK071221) were amplified by RT-PCR for construction of CaMV35S-OsBADH2-RNAi vector. An inverted-repeat fragment was constructed into vector pSK int 33 and transferred into pHB (driven by 2 × CaMV35S promoter 34) at the BamH I and Xba I restriction sites by using PCR with primers (B2IF1: 5' GCTCTCGAGTCTAGACCAATGGCCAGATTTGCAGT 3'; B2IR1: 5' GCACAAGCTTTGCGAGCAGTTCACCCAGAT 3'; B2IF2: 5' GGTGGATCCCCAA TGGCCAGATTTGCAGT 3'; B2IR2: 5' CACGAATTCTGCGAGCAGTTCACCCAGAT 3') introducing unique restriction sites at the product ends.

Transgenic plants were generated by Agrobacterium tumefaciens-mediated transformation according to the described procedure 35, and transformed lines were first selected for hygromycin (50 mg L^-1) resistance and then analyzed by PCR to determine the presence of T-DNA integration. Primers designed to the hygromycin phosphotransferase gene (hpt) of pHB for confirmation of integration are 5' TAGGAGGGCGTGGATATGTC 3' and 5' TACACAGCCATCGGTCCAGA 3' (GenBank accession No.E00777). PCR was performed by using Taq DNA Polymerase (Takara, Dalian, China) in MJ Mini™PCR (BIO-RAD, Hercules, California, USA), following the instruction given by the manufacturer.

Genomic DNA was extracted from mature rice leaves to determine the allelic variation of OsBADH2 locus in different varieties (Table 1). A 463 bp portion of OsBADH2 gene was amplified by PCR with primers designed for the OsBADH2 (5' ACATAGTGACTGGATTAGGTTCTG 3' and 5' CATCAACATCATCAAACACCACT 3'). This region of the gene includes the previously identified mutations putatively responsible for the aroma production 22. PCR products were cloned into the pMD18-T Vector (Takara, Dalian, China). Cycle sequencing was performed with the ABI Prism BigDye Terminators v2.0 cycle sequencing reaction kit (Applied Biosystems, Foster City, California, USA). Sequences were determined with an ABI Prism 377 genetic analyzer (Applied Biosystems). The sequence analysis was performed using the software DNAMEN v5.0 (Lynnon Biosoft Inc., Vandreuil, Quebec, Canada).

For semi-quantitative RT-PCR, total RNAs were extracted using Trizol reagent following the protocol provided by the manufacturer (Invitrogen, Carlsbad, CA) and treated with DNase (TaKaRa, Dalian, China). About 1 μg of total RNA from each sample was used for first-strand cDNA synthesis (Toyobo, Osaka, Japan). ACTIN gene was employed as positive internal control 36. Primers used for the RT-PCR analysis were designed for the ACTIN (5' AAGATCCTGACGGAGCGTGGTTAC 3' and 5' CTTCCTAATATCCACGTCGCACTTC 3'), and OsBADH2 (5' CCAATGGCCAGATTTGCAGT 3' and 5' TGCGAGCAGTTCACCCAGAT 3'), OsBADH1 (5' TGCGAACGCTGGTCAAGTCT 3'; 5' ATCACAGCGCCAGCTAGACC 3'). For real-time quantitative RT-PCR, total RNAs were treated with DNase (TaKaRa, Dalian, China) and about 1 μg of total RNA from each sample was used for first-strand cDNA synthesis (Toyobo, Osaka, Japan). Primers for real-time quantitative RT-PCR were designed for OsBADH2 (5' TGAAGCCGGTGCTCCTTTGT 3'; 5' CACTATAGGACTTTTTCCACCAAG 3'), OsBADH1 (5' GGTCAAGCCTGTTTCGTTAGAG 3'; 5' CAACCAACCTATCCAAGAATCGCT 3'), and the control ACTIN (5' ACCTTCAACACCCCTGCTAT 3'; 5' CACCATCACCAGAGTCCAAC 3'). The real time quantitative PCR was carried out in a total volume of 25 μL containing 12.5μL of SYBR^® Premix Ex Taq TM (TaKaRa, Dalian, China), 0.2 μM of each primer, and 7.5 μL of 1:65 diluted cDNA. Thermal cycling consisted of a hold at 95°C for 40 seconds followed by 40 cycles of 95°C for 5 seconds, 57°C for 10 seconds and 72°C for 15 seconds. After amplification, samples were kept at 95°C for 1 min and 55°C for 1 min. Then the temperature was raised gradually by 0.5°C every 10 seconds to perform the melt-curve analysis. Each sample was amplified in triplicate and all PCR reactions were performed on the iCycler^®PCR system (BIO-RAD, Hercules, California, USA). REST software 37 was used to quantify the OsBADH2 and OsBADH1 mRNA levels with ACTIN normalization by the 2-Ct method. To confirm the specificity of the PCR reaction, PCR products were electrophoresed on 1% agarose gel to verify accurate amplification product size.

Determination for the presence or absence of aroma was made according to previously described methods 38. Basmati370 was used as a control for the sensory aroma evaluation. At tillering stage of transgenic plants, 2 g fresh leaves were excised from individual plants, cut into 5 mm long pieces and kept in petri dishes mixed with 10 ml of 1.7% potassium hydroxide (KOH) solution. The petri dishes were kept under room temperature for about 10 minutes. They were then opened one by one, and the samples were sniffed and scored for presence or absence of aroma emission. To confirm the presence or absence of aroma, leaf tissue from tillers regenerated from stubs of T1 plants were soaked in KOH solution and evaluated for the aroma emission. In addition, about 20 grains harvested from T1 plants were placed into 5-ml screw-cap tube containing 1 ml of fresh water and were then incubated at 65°C for 2 hours. The aroma was evaluated after storage at 4°C for 20 min 21. All the samples were sniffed by three well-trained panelists. The samples were classified into two categories in presence or absence of aroma.

The hulled rice grains and/or mature leaves from the OsBADH2-RNAi transgenic and wild-type japonica cv. Nipponbare were collected, and Thai Hom Mali 105, a commercial aromatic variety from Thailand purchased from a retail store, was used as control 39. The 2, 4, 6-trimethylpyridine (TMP, Sigma Aldrich Chemical Co., Germany) was used as an internal standard 3940. It was dissolved in a precisely measured volume of 0.1 M HCl to give an internal standard solution with 2.00 ppm concentration of TMP. The milled rice grains/leaves (30 g) were added to a 250 mL flask containing 120 mL of internal standard solution. The mixture was stirred for 2 h before filtration. 9 mL of 1.0 M NaOH was added to the filtrate to make the solution slightly basic, then transferred to two 50 mL centrifuge tubes and centrifugated at 8000 rpm for 10 min. About 80 mL of the supernatant liquor was transferred to a 250 mL pear-shaped separatory funnel. Then 120 mL of dichloromethane was immediately added as an organic solvent. The extraction was conducted twice, resulting in 240 mL of dichloromethane solution. After drying with anhydrous sodium sulfate, the extract was concentrated to 1 mL using a rotary evaporator under reduced pressure and a temperature of 26°C. The concentrated extract was transferred to a tube and left open to the air at room temperature until its volume decreased to 0.1 mL, from which 1 μL was taken for qualitative analysis.

Samples were determined on a GC/MS system (GC-MSQP2010W, Shimadzu, Japan). Helium gas (purity 99.999%) at a pressure of 80 Kpa was used as the GC carrier gas. The injector and the GC/MS interface temperatures were set at 170°C and 250°C, respectively. The temperature of Rtx-wax capillary colum (30 m × 0.25 mm id, film thickness 0.25 μm, Restek, Bellefonte, PA) was programmed starting at 40°C after injection of samples. With the initial temperature of 40°C held for 2 min, it was ramped to 80°C at 6°C/min. After hold for 1 min in 80°C, it increased to 120°C at a rate of 4°C/min, then mounted up to 200°C at 8°C/min and held there for 20 min. The effluent from the capillary column went directly into the masspectrometer, operated in the electron impact (EI) mode with an ionization voltage of 70 eV, and the ion source temperature was 200°C 3940.

To investigate the effects of the disrupted OsBADH2 on crop performance, agronomical traits including plant height and 1000-grain weight were measured by using T2 fragrant/non-fragrant segregated progeny plants derived from four independent primary transformants.

To compare grain sprouting rates between wide type and OsBADH2-RNAi repression lines, 50 grains from each of three wild type controls (WTa, b and c) and four OsBADH2-deficient lines (OsB2-Ra, b, c and d) were placed and incubated in Petri dishes containing two pieces of filter paper moistened with 0, 50, 100 mM NaCl for 8 days. Sprouting index (∑R_GT/t, R_GT, the number of spouted grains in the t day) and germination rates (G/N × 100%, G, the number of spouted grains; N, the total number of grains) were measured.

For seedling growth rate experiments, 30 grains from each of three wild type controls (WTa, b and c) and four OsBADH2-deficient lines (OsB2-Ra, b, c and d) were incubated in distilled water for 7 days, and subsequently subjected to different salt concentrations (0, 50, 100 mM NaCl) for another 7 days. The length of shoots and roots, the number of roots and the fresh weight of shoots and roots were examined, respectively.