Transgenic Tobacco Overexpressing Brassica juncea HMG-CoA Synthase 1 Shows Increased Plant Growth, Pod Size and Seed Yield

Seeds are very important not only in the life cycle of the plant but they represent food sources for man and animals. We report herein a mutant of 3-hydroxy-3-methylglutaryl-coenzyme A synthase (HMGS), the second enzyme in the mevalonate (MVA) pathway that can improve seed yield when overexpressed in a phylogenetically distant species. In Brassica juncea, the characterisation of four isogenes encoding HMGS has been previously reported. Enzyme kinetics on recombinant wild-type (wt) and mutant BjHMGS1 had revealed that S359A displayed a 10-fold higher enzyme activity. The overexpression of wt and mutant (S359A) BjHMGS1 in Arabidopsis had up-regulated several genes in sterol biosynthesis, increasing sterol content. To quickly assess the effects of BjHMGS1 overexpression in a phylogenetically more distant species beyond the Brassicaceae, wt and mutant (S359A) BjHMGS1 were expressed in tobacco (Nicotiana tabacum L. cv. Xanthi) of the family Solanaceae. New observations on tobacco OEs not previously reported for Arabidopsis OEs included: (i) phenotypic changes in enhanced plant growth, pod size and seed yield (more significant in OE-S359A than OE-wtBjHMGS1) in comparison to vector-transformed tobacco, (ii) higher NtSQS expression and sterol content in OE-S359A than OE-wtBjHMGS1 corresponding to greater increase in growth and seed yield, and (iii) induction of NtIPPI2 and NtGGPPS2 and downregulation of NtIPPI1, NtGGPPS1, NtGGPPS3 and NtGGPPS4. Resembling Arabidopsis HMGS-OEs, tobacco HMGS-OEs displayed an enhanced expression of NtHMGR1, NtSMT1-2, NtSMT2-1, NtSMT2-2 and NtCYP85A1. Overall, increased growth, pod size and seed yield in tobacco HMGS-OEs were attributed to the up-regulation of native NtHMGR1, NtIPPI2, NtSQS, NtSMT1-2, NtSMT2-1, NtSMT2-2 and NtCYP85A1. Hence, S359A has potential in agriculture not only in improving phytosterol content but also seed yield, which may be desirable in food crops. This work further demonstrates HMGS function in plant reproduction that is reminiscent to reduced fertility of hmgs RNAi lines in let-7 mutants of Caenorhabditis elegans.

In agriculture, it is desirable to increase seed yield because grains represent significant sources of food, and the relevant key genes must be identified. Plant isoprenoids including sterols and BRs are essential in plant growth and reproduction [6,[22][23][24] and genes from the BR-specific biosynthetic pathway, including DWF4 and DWF5, affect seed production [22][23][24]. Transgenic Arabidopsis overexpressing DWF4 showed better vegetative growth and seed yield [23], while the Arabidopsis dwf5 mutant demonstrated a dwarf phenotype accompanied by abnormal seeds [22]. The genes in the first and third steps of the MVA pathway also affect plant growth and development. RNAi lines of Arabidopsis downregulated for cytoplasmic ACETOACETYL-COA THIOLASE2 (AACT2) displayed reduction in apical dominance, seed yield and root length, accompanied by sterility and dwarfing [25]. Also, the Arabidopsis hmgr1 mutant is dwarf-like and male sterile, and has a lower sterol content [26].

Plant materials and growth conditions
Wt tobacco (N. tabacum L. cv. Xanthi) obtained from the Institute of Molecular and Cell Biology (Singapore) was used in this study. Tobacco plants were grown at 25uC (16 h light)/22uC (8 h dark). Tobacco seedlings were cultured in Murashige and Skoog (MS) medium [40].

Generation of transgenic plants overexpressing HMGS
Plasmids pBj134 (wtBjHMGS1) and pBj136 (S359A) were used in Agrobacterium-mediated leaf disc transformation of N. tabacum [4,41]. The binary vector pSa13 [42] was used as vector control in transformation. T 1 transgenic tobacco seeds were selected on MS containing kanamycin (50 mg ml 21 ) and verified using PCR and DNA sequence [4]. T 2 homozygous plants with a single-copy transgene were compared in mRNA expression, metabolite composition, plant growth and seed yield.

Western blot analysis
Total protein was extracted [43] from 21-d-old tobacco leaves. Protein concentration was determined using the Bio-Rad Protein Assay Kit I (Bio-Rad). Protein (20 mg per well) separated on 12% SDS-PAGE was transferred onto Hybond-ECL membrane (Amersham) using a Trans-Blot H cell (Bio-Rad). Antibodies raised against the synthetic peptide (DESYQSRDLEKVSQQ) corresponding to BjHMGS1 amino acids 290 to 304 were used in western blot analyses [4,44]. Cross-reacting bands were detected using the ECL TM Western Blotting Detection Kit (Amersham).

Northern blot analysis
Tobacco total RNA was extracted from 21-d-old tobacco leaves using TRIzol reagent (Invitrogen). RNA (20 mg per well), separated on 1.3% agarose gels containing 6% formaldehyde, was transferred to Hybond-N membrane (Amersham) for northern blot analysis [45]. Digoxigenin-labelled probes were synthesized using the PCR Digoxigenin Probe Synthesis (Roche) with primer pairs ML276 and ML860 for BjHMGS1. Primers are listed in Table S1.

Southern blot analysis
Genomic DNA (40 mg) from 4-week-old tobacco leaves prepared by the CTAB method [46] was digested by EcoRI and separated on 0.7% agarose gel by electrophoresis, together with a 1-kb plus DNA standard ladder (Invitrogen). DNA was transferred from the agarose gel onto Hybond-N membrane (Amersham) by capillary transfer [47]. Southern blot analysis of tobacco using a 32 P-labelled full-length of BjHMGS1 cDNA probe with primer pair ML264 and ML860 was performed [4]. Primers are listed in Table  S1.

Extraction and quantitative analysis of sterols
For sterol profiling, freeze-dried materials from 20 mg of 60-dold soil-grown tobacco leaves and 10 mg of 20-d-old MS platecultured tobacco seedlings were used. Extraction and quantitative analysis of sterols were carried out as described [4,48]. GC-MS analysis (GC: Hewlett Packard 6890 with an HP-5MS capillary column: 30 m long, 0.25 mm i.d., film thickness 0.25 mm; MS: Hewlett Packard 5973 mass selective detector, 70 eV) was used to determine sterol content, with He as the carrier gas (1 ml/min). The column temperature program used included a fast rise from 60uC to 220uC (30uC/min) and a slow rise from 220uC to 300uC (5uC/min), then kept at 300uC for 10 min. The inlet temperature was 280uC. Compounds were identified using the National Institute of Standards and Technology (NIST) libraries of peptide tandem mass spectra (Agilent, USA). The sterol masses were determined by comparison of the peak area of each compound with that of the internal standard (lupenyl-3,28-diacetate). Two independent lines for each OE genotype were analysed. Five independent repeats (samples) for each independent line were used for sterol extraction. Each sample was injected twice in GC-MS analyses and an average of the sterol mass was taken. Sitosterol, campesterol and stigmasterol contents in transgenic tobacco HMGS-OEs were compared to those in vector (pSa13)-transformed plants following previous reports [4,17].

Seed germination assay
Tobacco seeds collected simultaneously from vector (pSa13)transformed control and HMGS-OE lines were sterilized in 20% bleach, 70% ethanol and then spread on MS medium agar plates supplemented with kanamycin (50 mg/l). About 30 tobacco seeds were sown on one plate. Five duplicate plates were used for each independent line [4]. All the plates were incubated at 4uC for 4 days and transferred to a culture room for 2 days under a photoperiod of 22uC 8-h dark and 23uC 16-h light. Subsequently, the number of germinated seeds was counted every 12 h for 60 h using a dissecting microscope. The emergence of the radicle was defined as germination [4]. The germination rates were calculated and compared using the Student's t-test. Two independent lines of OE-wtBjHMGS1 (''401'' and ''402'') and two independent lines of OE-S359A (''603'' and ''606'') were tested in seed germination assays. The experiment to measure seed germination was repeated twice.

Growth rate measurements
Growth rate was measured according to previous reports [49][50][51][52][53]. Four-d-old seedlings were transferred onto fresh MS plates placed vertically for a further 10-d growth. The dry weight of 14d-old seedlings was then measured. Five seedlings were grouped for weight measurements and a total of 30 groups were analysed per individual line.
For greenhouse plants, 7-d-old tobacco seedlings of similar size were transferred from MS medium to soil for further growth rate measurements. The height of 80-, 98-and 210-d-old tobacco were measured. As 80-d-old plants did not have flowers, the height measurement did not include the inflorescence. However, 98-and 210-d-old plants were flowering and the height measurement included the inflorescence. For 98-d-old tobacco, measurements of leaf fresh weight, length and width of the four bottom-most leaves were also analysed for the vector-transformed control, OE-wtBjHMGS1 and OE-S359A. Two independent lines from each OE construct were analysed for 80-d-old tobacco plants and three independent lines from each OE construct were analysed for 98and 210-d-old tobacco plants. For each line, six plants were used.

Comparison in tobacco seed yield
Seed yield was measured [49,51,52,54] to test the differences between HMGS-OEs (OE-wtBjHMGS1 and OE-S359A) and the vector-transformed control. Ten plants each from two independent lines from each OE construct were examined and T 2 homozygous seeds of each line were germinated on MS. Fourteend-old seedlings were transferred to soil in a greenhouse. Pods (30 per group) were harvested at maturity from each of 10 plants per line to determine total dry pod weight, average dry pod weight, total dry seed weight and total seed number. The experiment to measure seed yield was repeated twice (2-3 groups were analysed for each repeat).
To further determine if increase in seed size occurred, the dry weight of 100 seeds from each line was measured and 29 repeats were carried out per line. The average dry weight was calculated from 30 measurements of 100 seeds per line.

RNA analysis
Total RNA (5 mg) of 20-d-old tobacco seedlings and 14-d-old Arabidopsis were extracted using RNeasy Plant Mini Kit (Qiagen) and were reverse-transcribed into first-strand cDNA using the SuperScript First-Strand Synthesis System (Invitrogen). Quantitative Reverse Transcription-PCR (qRT-PCR) was carried out with a StepOne Plus Real-time PCR System (Applied Biosystems) and FastStart Universal SYBR Green Mater (Roche). The conditions for qRT-PCR were as follows: denaturation at 95uC for 10 min, followed by 40 cycles of 95uC for 15 s and 60uC for 1 min. Three experimental replicates for each reaction were carried out using gene-specific primers and tobacco ACTIN and Arabidopsis ACTIN2 were used as internal controls. The relative changes in expression from three independent experiments were analysed [55]. Primers for qRT-PCR are listed in Table S1.

Statistical analysis
Analyses of data in this work was carried out using the Student's t-test to determine any significant differences between means.

Molecular analyses of transgenic tobacco HMGS-OEs
The presence of wt and mutant BjHMGS1 in transgenic tobacco was verified by PCR ( Figure S1A Figure 2A). As the peptide used to generate anti-BjHMGS1 antibodies shows 100% homology to tobacco HMGS (GenBank accession number EF636813), a faint band was detected in the vector (pSa13)-transformed control ( Figure 2A). Northern blot analyses revealed that transgenic lines verified by western blot analysis expressed BjHMGS1 mRNA ( Figure 2B). Single-insertional lines identified by Southern blot analyses ( Figure S2) were selected for further experiments.
The % increase of sterols between transgenic tobacco (observed herein) and transgenic Arabidopsis (OE-wtBjHMGS1 and OE-S359A) [4] were also compared ( Table 2 and S2). A similar trend was observed in transgenic Arabidopsis and tobacco seedlings; OE-S359A transformants displayed higher increase than the OE-wtBjHMGS1 not only in each sterol (campesterol, stigmasterol and sitosterol) but also in total sterol (Table S2). OE-S359A transformants also showed similar increase over the OE-wtBjHMGS1 in both Arabidopsis and tobacco leaves for stigmasterol and total sterol ( Table 2 and S2).

Tobacco HMGS-OE seeds germinated earlier
As seeds from Arabidopsis HMGS-OEs were observed to germinate earlier than the vector (pSa13)-transformed control [4], the germination of tobacco HMGS-OE seeds was investigated. Tobacco seeds of OE-wtBjHMGS1 and OE-S359A not only germinated earlier but also displayed significantly higher germination rates than the control at 60 to 120 h post-germination ( Figure S3). Also, OE-S359A germinated faster than OE-wtBjHMGS1 ( Figure S3).
Growth differences in height ( Figure 4A-B) and leaf size ( Figure 4C-D) between 98-d-old HMGS-OEs (OE-wtBjHMGS1 and OE-S359A) and vector (pSa13)-transformed plants were also Table 1. Sterol profiles of tobacco HMGS-OE seedlings and leaves ( mg/mg dry weight).  (Figure 4). Both OE-wtBjHMGS1 and OE-S359A had a significant increase (91% and 97%, respectively) in height over the vector-transformed control ( Figure 4B). Leaf fresh weight and size (length and width) ( Figure 4C-D) in some of the OE-wtBjHMGS1 lines and all three OE-S359A lines were significantly heavier and bigger, respectively, than the control at similar age ( Figure 4D). Furthermore, growth differences in height between 210-d-old HMGS-OEs (OE-wtBjHMGS1 and OE-S359A) and vectortransformed plants were also observed ( Figure 5). OE-wtBjHMGS1 showed a significant increase (21%) in height over the control, while OE-S359A displayed an even higher increase (45%) ( Figure 5B).

Change in expression of isoprenoid biosynthesis genes in tobacco HMGS-OEs
qRT-PCR was performed to check the effect of BjHMGS1 overexpression on the expression of genes downstream of HMGS in tobacco HMGS-OE seedlings and to explore possible molecular mechanism of HMGS function in plant growth and seed production. The results from qRT-PCR revealed that the expression of NtHMGR1, NtIPPI2, NtSQS, NtSMT1-2, NtSMT2-1, NtSMT2-2 and NtCYP85A1 was significantly higher than in the vector (pSa13)-transformed control for both OE-wtBjHMGS1 and OE-S359A tobacco seedlings with the exception of NtSQS, NtSMT1-2, NtSMT2-2 and NtCYP85A1 in one OE-wtBjHMGS1 line (401) (P,0.01) (Figure 7). However, there was no difference in the expression of NtHMGR2 between all the HMGS-OE lines and the vector-transformed control (Figure 7). For the expression of NtFPPS, there was no disparity amongst the two lines of OE-wtBjHMGS1 (401 and 402) and the vector-transformed control, while the expression of NtFPPS in another OE-wtBjHMGS1 line (404) and in two OE-S359A lines (602 and 606) was slightly higher than the control (P,0.05) (Figure 7). Conversely, the expression of NtIPPI1, NtGGPPS1, NtGGPPS3 and NtGGPPS4 were downregulated in tobacco HMGS-OE seedlings (P,0.01) (Figures 7-8) while the expression of NtGGPPS2 was higher than the control (P,0.05) in two OE-wtBjHMGS1 lines (402 and 404) and two OE-S359A lines (602 and 606) (Figure 8). Observations that (i) NtSQS expression in all three OE-S359A lines was higher than all three OE-wtBjHMGS1 lines, (ii) NtHMGR1 and NtCYP85A1 expression in all three OE-S359A lines were higher than two (''401'' and ''402'') of three OE-wtBjHMGS1 lines, and (iii) NtSMT2-1 expression in two (''602'' and ''603'') of three OE-S359A lines was higher than two (''401'' and ''402'') of three OE-wtBjHMGS1 lines suggest that the differences in expression levels of NtSQS, NtHMGR1, NtSMT2-1 and NtCYP85A1 in OE-wtBjHMGS1 and OE-S359A do correspond to the expected Table 2. Increase (%) of sterol composition in tobacco HMGS-OE seedlings and leaves in comparison to vector (pSa13)-transformed control and elevation of OE-S359A over OE-wtBjHMGS1. differences in enzyme activities between recombinant wtBjHMGS1 and S359A [37].

New observations from tobacco HMGS-OEs
Our investigations on the overexpression of HMGS in transgenic tobacco revealed new observations not previously evident in Arabidopsis HMGS-OEs including the upregulation of NtIPPI2, NtSQS and NtGGPPS2 and downregulation of NtIPPI1, NtGGPPS1, NtGGPPS3 and NtGGPPS4 (Figures 7-8). However, similar to findings from Arabidopsis HMGS-OEs, enhanced NtHMGR1, NtSMT1-2, NtSMT2-1, NtSMT2-2 and NtCYP85A1 expression in tobacco HMGS-OEs was seen (Figure 7). Other new findings from tobacco HMGS-OEs included growth stimulation in the tobacco HMGS-OE lines, confirming the positive role of HMGS overexpression in plant growth. Furthermore, tobacco HMGS-OEs show increased pod size and seed yield ( Figure 6), indicative of a specific HMGS function in seed production. Improved growth, pod size and seed yield of OE-S359A in comparison to OE-wtBjHNMGS1 may be attributed to the higher NtSQS expression (Figure 7) and sterol content in OE-S359A transformants ( Table 1).

Function of HMGS in reproduction and development
In plants, the floral organs are involved in reproduction. HMGS has been shown to play a crucial role in floral development [4,34,35,37]. In Arabidopsis, higher AtHMGS expression had been observed in flowers than seedlings or leaves from RT-PCR analysis [4]. Using mutants in HMGS, AtHMGS was demonstrated essential for pollen fertility and proper development of tapetum-specific organelles in Arabidopsis [35]. In B. juncea, northern blot analysis had previously revealed that BjHMGS1 mRNA was highly expressed in flowers and seedling hypocotyls [34] and in situ hybridization analysis had shown that HMGS mRNA was predominantly localized in the stigmata and ovules of flower buds and in the piths of seedling hypocotyls [37]. BjHMGS1 and BjHMGS2, but not BjHMGS3 and BjHMGS4 expression was  detected in the floral buds as examined by RT-PCR analysis [37]. The effect on the overexpression of BjHMGS1 in transgenic tobacco observed herein further extends the significance of HMGS in reproduction related to seed production as well as to whole plant development (Figures 3-6). More interestingly, OE-S359A lines were found to display greater effect in growth, pod size and seed yield than OE-wtBjHNMGS1 (Figures 3-6). OE-S359A, which was expected to possess higher HMGS activity than OE-wtBjHMGS1, caused higher expression of tobacco native genes downstream of HMGS such as NtSQS, NtHMGR1, NtSMT2-1 and NtCYP85A1 (Figure 7), and increased sterol levels, which more effectively enhanced seed production in comparison to OE-wtBjHMGS1.
Besides HMGS, other enzymes in the early steps of the MVA pathway are important in these development processes. It has been observed that both hmg1/hmg1 and HMG1/hmg1 hmg2/hmg2 Arabidopsis mutants deficient in HMGR activity are male sterile [26,57]. The hmg1hmg2 male gametophytes in the HMG1/hmg1 hmg2/hmg2 mutant were lethal [57]. Furthermore, the characterization of Arabidopsis AACT1 and AACT2 led to suggest a specific role of AACT2 in catalyzing the first step of the MVA pathway [58], while AACT1 is rather involved in the peroxisomal fatty acid degradation process, like in tobacco seedlings [59]. Arabidopsis AACT2 RNAi lines further showed reduction in apical dominance, seed yield and root length accompanied by sterility and dwarfing [25]. These studies using the AACT RNAi lines, and mutants in HMGS and HMGR together with observations herein confirm the significance of the MVA pathway in plant reproduction and development.
Recently, two genes were cloned and characterized from two miRNA-action deficient (MAD) mutants; MAD3 encodes the MVA pathway enzyme HMGR1, while MAD4 encodes sterol C-8 isomerase in dedicated sterol biosynthesis [60]. Their results showed that the lack in HMGR1 catalytic activity is sufficient to inhibit miRNA activity and that sterol is essential for the normal activity of plant miRNAs [60]. Furthermore, their results implied that besides sterols, other isoprenoids may also affect the normal function of miRNA [60]. It has been reported that Caenorhabditis elegans HMGS1 (CeHMGS1) plays an important role in the miRNA pathway; CeHMGS1 regulates the function of many, if not all, miRNAs at multiple tissues and stages during C. elegans development [61]. Furthermore, CeHMGS1 affects the fertility of C. elegans in the miRNA defective let-7 worms [61]. This effect on fertility is reminiscent of our observations on tobacco HMGS-OEs herein on seed production which represents fertility in plants.

Effects of HMGS in regulating isoprenoid biosynthesis genes in tobacco HMGS-OEs
In transgenic Arabidopsis, the overexpression of wt and mutant (H188N, S359A and H188N/S359A) BjHMGS1 caused a feedforward effect in the upregulation of several genes in sterol biosynthesis including HMGR, SMT2, DWF1, CYP710A1 and BR6OX2 [4]. This study using tobacco HMGS-OEs demonstrated that some differences exist between tobacco and Arabidopsis HMGS-OEs in the expression of genes encoding HMGR and SMT (cf. Figure 1). Although HMGR is considered to be the ratelimiting enzyme in the MVA pathway in plants [62], only NtHMGR1 but not NtHMGR2 was upregulated in tobacco HMGS-OEs (Figure 7). This can perhaps be attributed to some differences in the localization and function of NtHMGR1 and NtHMGR2 [63,64]. NtHMGR1 is a house-keeping gene that likely participates in sterol biosynthesis, plant growth and development, while NtHMGR2 is stress-inducible [63,64]. Also elicitor-inducible HMGR activity is known to be associated with defence-related sesquiterpenoid accumulation in tobacco cell suspension cultures [65]. Thus it was not surprising that rather than NtHMGR2, NtHMGR1 was upregulated in seedlings undergoing rapid growth and development.
Isopentenyl diphosphate isomerase (IPPI) catalyses the interconversion of IPP and its allyl isomer dimethylallyl diphosphate (DMAPP) and provides the first key intermediate for the biosynthesis of all kinds of isoprenoids including sterols in the MVA pathway and carotenoids in the MEP pathway [1,3,12,66] (and references cited therein) (cf. Figure 1). IPP is most likely involved in cross-talk between the cytosolic MVA pathway and the plastidial MEP pathway [13,14]. AtIPPI1 and AtIPPI2 have been reported to be critical to sterol biosynthesis in the MVA pathway and Arabidopsis development [67]. Analysis of the expression of the two NtIPPI genes in tobacco HMGS-OE seedlings revealed that NtIPPI1 was downregulated, while NtIPPI2 was upregulated (Figure 7). Their corresponding proteins are apparently differentially localized in tobacco [68]. NtIPPI1 is targeted to the chloroplast, while NtIPPI2 is cytosolic, similar to BjHMGS1 [37,68]. Possibly, upregulation of BjHMGS1 and NtIPPI2 in the cytosol of tobacco HMGS-OE seedlings promoted cross-talk between the MVA and MEP pathways. The MEP pathway  produces simultaneously IPP and DMAPP, and plastidial NtIPPI1 is possibly needed to adjust the ratio of starter DMAPP to elongation units IPP for longer prenyl chains. If IPP is imported from the cytosol because of ''overproduction'', then plastidial NtIPPI1 would be downregulated.
FPPS catalyses the condensation of two molecules of IPP with DMAPP to form farnesyl diphosphate (FPP) (C 15 ) (cf. Figure 1), which provides the key precursor for the biosynthesis of essential isoprenoids such as sesquiterpenes, ubiquinones, polyterpenes, dolichols and sterols [69,70]. In plants, FPPS isozymes that are encoded by a small gene family, exert differential roles, based on their subcellular localisation [69,71]. NtFPPS expression was slightly elevated in seedlings of only one OE-wtBjHMGS1 line (Figure 7). Given that NtFPPS functions as the key provider of the universal product FPP in the biosynthesis of many C-15 related products, a moderate change in NtFPPS mRNA in the HMGS-OE lines may not be significant enough to affect sterol accumulation. Also, other NtFPPS isogenes or post-translational regulation may be involved [72][73][74].
SQS catalyses the biosynthesis of squalene by the reductive dimerization of two FPP molecules (cf. Figure 1), and represents the first committed step in the biosynthesis of sterols, BRs and triterpenes [75][76][77][78][79]. The change in NtSQS expression in seedlings was the most dramatic, with a 2.1-fold increase in two lines of OE-wtBjHMGS1 and 36.5-fold in OE-S359A, in comparison to the vector-transformed control (Figure 7). The increase of NtSQS mRNA in OE-S359A seedlings was also much higher (11.1-fold) than OE-wtBjHMGS1 ( Figure 7). Interestingly, NtSQS expression and NtSQS activity have been detected predominantly at the shoot apical meristem (SAM) rather than leaves or roots, implying that sterol biosynthesis occurs especially in the SAM [77]. Furthermore, the SAM is critical in plant growth and development, and stem cells from the SAM continuously generate all the aerial organs and tissues of a plant [80]. Results from qRT-PCR (Figure 7) herein support a role for NtSQS in HMGS-associated sterol accumulation related to growth and seed yield. Also, enhanced sterol accumulation, growth and seed yield in OE-S359A, over OE-wtBjHMGS1 (Figure 7), corresponded to higher NtSQS expression (Figure 7). Consistently, Arabidopsis SQS (AtSQS) displayed higher expression in HMGS-OEs than the vectortransformed control; and AtSQS expression in OE-S359A was higher than OE-wtBjHMGS1 ( Figure S4). However the elevation of NtSQS in tobacco OE-S359A over OE-wtBjHMGS1 (Figure 7)  was greater in comparison to AtSQS in Arabidopsis OE-S359A ( Figure S4). Furthermore, our results correspond well to a recent study on the overexpression of Glycine max SQS1 (GmSQS1) in Arabidopsis that yielded a 50% increase of seed sterol content [81]. An enhanced flux of MVA to FPP might present some risk as phosphatases always being present might liberate farnesol, which can be quite toxic to cells [82]. Thus SQS could remove a potentially dangerous intermediate and get it channelled into the synthesis and accumulation of chemically inert sterols and their derivatives.
In the MEP pathway, GGPPS catalyses the consecutive condensation of three molecules of IPP and one DMAPP to generate the 20-carbon geranylgeranyl diphosphate (GGPP) (cf. Figure 1), which is the universal key intermediate for the biosynthesis of carotenoids and of abscisic acid as derivative, of gibberellins, chlorophylls, tocopherols, phylloquinone, plastoquinone, dolichols, polyprenols and oligoprenols [12,83]. Although four GGPPS-like cDNAs have been reported from tobacco [84], only NtGGPPS2 was upregulated in two lines of OE-wtBjHMGS1 and all three lines of OE-S359A, while NtGGPPS1, NtGGPPS3 and NtGGPPS4 were observed to be downregulated in all the HMGS-OE seedlings (Figure 8), implying that HMGS overexpression had a positive effect on NtGGPPS2 expression and a negative role on NtGGPPS1, NtGGPPS3 and NtGGPPS4 expression. However, it cannot be discounted that NtGGPPS1, NtGGPPS3 and NtGGPPS4 may be subject to other modes of regulation such as post-translational modification that has been reported for AtGGPPS3, AtGGPPS7, AtGGPPS9 and AtGGPPS10 [85,86]. Most recently, a new relationship between the MVA pathway and the MEP pathway has been proposed in which the monoterpene S-carvone inhibited the production of MVA-derived capsidiol, a cellulose-induced sesquiterpenoid phytoalexin in tobacco by down-regulation of MEP-pathway dependent protein isoprenylation [87].