Crop Science Illumina
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Published online 1 July 2008
Published in Crop Sci 48:1470-1481 (2008)
© 2008 Crop Science Society of America
677 S. Segoe Rd., Madison, WI 53711 USA
This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Chapman, K. D.
Right arrow Articles by Kinney, A. J.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Chapman, K. D.
Right arrow Articles by Kinney, A. J.
Agricola
Right arrow Articles by Chapman, K. D.
Right arrow Articles by Kinney, A. J.
Related Collections
Right arrow Seed Quality
Right arrow Cotton
Right arrow Cell Biology & Molecular Genetics

Reduced Oil Accumulation in Cottonseeds Transformed with a Brassica Nonfunctional Allele of a Delta-12 Fatty Acid Desaturase (FAD2)

Kent D. Chapmana,*, Purnima B. Neogia, Kater D. Hakec, Agnes A. Stawskaa, Thomas R. Speedc, Matthew Q. Cottera, David C. Garrettb, Thomas Kerbyc, Charlene D. Richardsona, Brian G. Ayreb, Supriyo Ghoshd and Anthony J. Kinneye

a Center for Plant Lipid Research, Univ. of North Texas, Denton, TX
b Dep. of Biological Sciences, Univ. of North Texas, Denton, TX
c Delta and Pine Land Company, One Cotton Row, Scott, MS
d Bruker Optics, Inc., The Woodlands, TX
e Pioneer Crop Genetics, DuPont Experiment Station, Wilmington, DE

* Corresponding author (chapman{at}unt.edu).


   ABSTRACT
TOP
 NOTES
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
In an effort to better understand the mechanisms that regulate oil accumulation and packaging in seeds, transgenic cotton lines were generated using a Brassica napus nonfunctional delta-12 fatty acid desaturase (FAD2) gene under control of the phaseolin promoter. Seeds of numerous transgenic plant lines had reduced oil content compared with null-segregating siblings or nontransformed seeds. Seed oil content was quantified by 1H-NMR, and was reduced to 12% or less of seed weight in transgenics from 20% by weight in nontransformed controls. Light- and electron-microscopic analyses of severely lowered lines, showed a reduction in overall cotyledon thickness and a disruption in cellular and subcellular organization. Lipid bodies and protein bodies were fewer in transgenics, and their size and distribution in cells was different than that in nontransformed seeds. Coincident with reduced storage reserves, sucrose levels were elevated in transgenic seeds. The overall effect of oil suppression was to selectively reduce the size of the embryo, but the seed coat and fiber properties remained unaffected in seeds. In fact there was a significant increase in lint percentage in all oil-suppressed lines examined. Overall we propose that expression of the nonfunctional Bnfad2 allele in cottonseeds disrupts normal oil biosynthesis and provides for redirection of carbon reserves.

Abbreviations: ACP, acyl carrier protein • ER, endoplasmic reticulum • FAD, fatty acid desaturase • PCR, polymerase chain reaction • RT-PCR, reverse transcriptase–PCR • TAGs, triacylglycerols • TD-NMR, time-domain NMR


   INTRODUCTION
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
COTTON IS GROWN mostly for fiber production, but it is also the world's sixth largest source of vegetable oil. Refined cottonseed oil is composed of 26% palmitic (16:0), 2% stearic (18:0), 15% oleic (18:1), and 55% linoleic (18:2) acids (Jones and King, 1996). Fatty acids in plants are major structural components of biological membranes and storage oil, and the assembly of membrane and storage lipids utilizes the same cellular machinery. Intense efforts are underway by many laboratories to understand the mechanisms that regulate oil packaging in seeds and its relationship to membrane biogenesis, which involves three major biosynthetic events: (1) the synthesis of fatty acids in plastids, (2) the modification of fatty acids by enzymes located in the endoplasmic reticulum (ER), and (3) the packaging of these fatty acids into oil bodies (Fig. 1 ).


Figure 1
View larger version (36K):
[in this window]
[in a new window]

  		Figure 1. Higher plants synthesize fatty acids de novo in the stromal compartment of plastids. Acyl chains esterified to an acyl carrier protein (ACP) undergo chain elongation by the sequential addition of two carbon units, donated by malonyl ACP. Hydrolysis of the acyl-ACP thioester bond by acyl-ACP thioesterases (FATB and FATA) terminates acyl chain elongation. Fatty acids are exported from seed plastids to the endoplasmic reticulum (ER) for the synthesis of membrane glycerolipids or storage oils (TAGs). Fatty acid desaturases (e.g., FAD2), located in ER, modify the number of double bonds in the fatty acyl chains. In cottonseeds the major proportion of flux is toward the synthesis of TAG (with mostly 18:2 fatty acids) and packaging into oil bodies that are stored in the cytoplasm (diagram adapted from Somerville et al., 2000). Inset shows a portion of a cotton cotyledon cell visualized by conventional transmission electron microscopy, showing the typical structural organization of oil bodies (OB) and protein bodies (PB). Bar represents 1 micron.

 
An enzyme important for the synthesis of polyunsaturated fatty acids in higher plants is the delta-12 desaturase (FAD2) that inserts a double bond between carbons 12 and 13 of monosaturated oleic acid to produce linoleic acid (Liu et al., 2001). Endoplasmic-reticulum-membrane-bound fatty acid desaturases (FAD2 and FAD3) synthesize linoleic (18:2) and linolenic acids (18:3), both of which are common components of cellular membranes and commercial vegetable oils (Shanklin and Cahoon, 1998). There are multiple FAD2 genes in the cotton genome (Chapman and Pirtle, 2001). However, FAD2-1 is highly expressed in maturing embryos and is the main contributor of the polyunsaturated fatty acids in the seeds of cultivated cottons (Liu et al., 1999).

Genetic engineering strategies that target FAD2 provide opportunities for the dramatic alteration in seed polyunsaturated fatty acid composition, and several groups have reported the development of high-oleic transgenic cottonseed lines (Chapman et al., 2001; Liu et al., 2002; Sunilkumar et al., 2005). Our group suppressed endogenous activity of cottonseed FAD2 by expressing a nonfunctional fad2 allele from Brassica napus (Bnfad2; Chapman et al., 2001) (Fig. 2 ). As might be predicted, cottonseeds from these lines had elevated oleic acid (up to 45%) and reduced linoleic acid (down to 30%). Alternatively, RNA interference silencing of cottonseed FAD2 expression was accomplished by Liu et al. (2002), and cottonseed in fad2-suppressed lines had 80% or so oleic acid and concomitant reductions in palmitic and linoleic acid content. Efforts by Rathore and coworkers to elevate oleic acid content in cottonseeds by FAD2-antisense suppression (Sunilkumar et al., 2005) also resulted in high-oleic and low-linoleic phenotypes.


Figure 2
View larger version (40K):
[in this window]
[in a new window]

  		Figure 2. Panel A: Alignment of a portion of the amino acid sequence of the mutant (Broglie, 1999) and wild-type (AY577313) Brassica napus FAD2 protein products reveals two amino acid substitutions (red arrow) in one of the histidine boxes (blue line) that forms the catalytic site rendering the mutant protein catalytically inactive. Panel B: Identification of BnFAD2 transcripts by reverse transcriptase–polymerase chain reaction (RT-PCR) in mature seeds of transgenics. M, marker. Lanes 1–3: Coker, 36A null segregant, 11G null segregant (negative controls). Lanes 4–8: transgenic lines 36A-T2, 36A-T3, 36A-T4, 11G-5, 11E-8 showing amplification of a 528-bp fragment with BnFAD2-specific primers. Panel C: Amplification of actin transcripts (539 bp) by RT-PCR served as a control for all samples in panel B. Separate experiments where the avian reverse transcriptase was left out of the reaction showed no amplification of bands for either BnFAD2 or actin (not shown), confirming amplification of mRNA. Panel D: Tissue-specific expression studies (in 36A-T2 transgenic line) of BnFAD2 mutant allele by RT-PCR in young leaf (YL), mature leaf (ML), anther (An), pollen (Pol), seed (S) showed transcript only in 36A-T2 transgenic seed. Panel E: Amplification of stearoyl acetyl carrier protein desaturase (SAD1) transcripts (107 bp) by reverse transcriptase–polymerase chain reaction was used as a control for all the samples in panel D.

 
Subsequent characterization of the Bnfad2-containing lines (Chapman et al., 2001) revealed poor seed germination characteristics, a phenotype not reported with either RNA-mediated silencing approach (Liu et al., 2002; Sunilkumar et al., 2005), suggesting that expression of the nonfunctional BnFAD2 gene may be in part responsible for germination and seedling establishment defects. Here we show that expression of Bnfad2 in cottonseeds is associated with reduced embryo size, reduced seed oil and protein content, and increased sucrose content. However, these effects were embryo specific since seed coat weights and fiber properties were not adversely affected. Reduced oil content was associated with abnormal lipid body morphology and poor germination and seedling growth. The reduction in reserve content in embryos may afford a reallocation of resources in cottonseed since both fiber percentage and fiber content appeared to be elevated in these low-oil lines. Overall we propose that poor germination and seedling establishment in these Bnfad2 cotton lines is due to inefficient reserve accumulation during seed maturation and/or reduced capacity for reserve mobilization during germination and early seedling growth. These results suggest new strategies for increasing cellulose synthesis in this fiber crop.


   MATERIALS AND METHODS
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 REFERENCES
 
Plant Material
Seeds of primary transformed cotton plants (Chapman et al., 2001) were surface sterilized in 20% bleach, rinsed in sterile water, and germinated on moistened towels in covered containers at 30 to 35°C in the dark for 4 to 5 d. Seed germination was scored by radical emergence, and seedling growth was quantified by hypocotyl and radicle/root length. Seedlings were transplanted to soil in 4-inch pots for acclimation and growth in greenhouse conditions. At the two- to three-leaf stages, seedlings were genotyped by polymerase chain reaction (PCR) analysis (commercial service, Biodiagnostics, Inc., Madison, WI), transplanted into 12- or 20-inch pots, and supplemented with time-release fertilizer according to manufacturer's instructions (Osmocote Plus, Scotts International BV, Geldermalsen, The Netherlands). For initial trials with seven independent lines (T1 generation), plants were maintained in a glasshouse under evaporatively cooled conditions at 30 to 40°C during the day and 25 to 30°C at night. Five to 12 plants of each Bnfad2 line (and corresponding null controls) were placed randomly in the greenhouse to minimize local environmental effects. Seedlings for null segregants and controls were selected from a second planting that was delayed by 10 to 14 d to minimize growth differences attributed to seedling emergence. Plant height, flower position, and boll development were monitored and were similar after the seedling stage for nontransgenics and transgenics.

For follow-up yield comparisons of two lines, 36A and 11E (T3 generation), plants were maintained at 30°C in an air-conditioned glasshouse and watered for 30 min, twice daily. Daylight was supplemented and day length extended with twelve 400-W sodium vapor lamps. In all cases, controls included nontransgenic, null-segregating individuals (derived from the same line), a vector-control line (plants transformed with pBI121 (Chapman et al., 2001), and the wild-type background (Coker 312).

Polymerase Chain Reaction Analyses
The occurrence of the Bnfad2 sequence in cotton genomic DNA was assessed by PCR using conditions previously described (Chapman et al., 2001). Large numbers of samples were on occasion analyzed by a commercial service (Biodiagnostics, Inc.) using the same primers and identical conditions. For determination of Bnfad2 expression, total RNA was isolated and pooled from 10 ginned cottonseeds showing low oil content using the RNeasy Plant Mini Kit (QIAGEN, Valencia, CA), or from leaf, anther and pollen according to (Vicient and Delseny, 1999), with a clean-up step via the RNeasy Plant Mini Kit (QIAGEN), according to the manufacturer instructions. RNA was quantified and evaluated for purity by UV spectroscopy and agarose gel electrophoresis (Krieg, 1996). Bnfad2 transcripts were amplified by reverse transcriptase (RT)-PCR using the primers: BnFAD2 forward primer ATGCAAGTGTCTCCTCCCTCC and BnFAD2 reverse primer CGTTAACATCACGGTGCGTC. The two principal cottonseed oleosin genes, MATP6 and MATP7(Hughes et al., 1993), were amplified with the following primers: MATP6 forward primer AAGTCCGTGACCGCAAC and MATP6 reverse primer GCCCCACATACTCTGTC and MATP7 forward primer CGCTCTTACCGGACGTTTGAC and MATP7 reverse primer CCCCACCATATCTTGCATGCC. Controls without reverse transcriptase confirmed that only RNA was amplified in these reactions. Amplification of actin and SAD1 mRNAs were used as an endogenous control for expression analyses (Hoang and Chapman, 2002; Yang et al., 2005).

Microscopy
Seeds were imbibed in water for 1 h, after which the embryos were excised from their seed coats and the cotyledons diced into small pieces. Primary fixation was in 2% (v/v) glutaraldehyde in 100 mM potassium phosphate buffer (pH 6.8) with 5% dextrose for 2 h. Samples were rinsed in buffer and postfixed in 1% OsO4 at 4°C overnight. Samples were dehydrated through a graded ethanol series and flat embedded in Spurr's epoxy resin (Electron Microscopy Science, Hatfield, PA). For light microscopy, semithin (1 µm), cross sections of cotyledons were cut with an MT-6000 microtome (Boeckeler Instruments, Inc., Tucson, AZ) and placed on glass slides for staining in toluidine blue (0.5% w/v) followed by basic fuchsin (0.05% w/v). Images were captured with a Nikon Microphot-FX microscope (Nikon Instruments, Inc., Melville, NY). For transmission electron microscopy, ultrathin sections (60–90 nm) were cut using a diamond knife and mounted on 200-mesh copper grids. Sections were stained with saturated uranyl acetate and lead citrate (0.3%) and viewed at 80 kV on a JEOL JEM 100CXII transmission electron microscope (JEOL USA, Inc., Peabody, MA).

Confocal scanning fluorescence microscopy was used to visualize lipid bodies in cotyledon samples. Unfixed cotyledons were equilibrated in 80 mM PIPES pH 7 for 20 min before staining for 20 min in the neutral lipid-specific stain Nile red (0.16 µg/mL), prepared from an 8-mg/mL stock solution dissolved in DMSO. Samples were washed 3 x 5 min in 80 mM PIPES. All steps were performed in darkness. A final H2O wash was performed, and the sections were mounted in H2O under a cover slip and visualized with a Zeiss 200M optical microscope with CSU-10 Yokogawa confocal scanner from McBain Instruments and photographed with a Hamamatsu digital camera (Phoenix, AZ). Lipid bodies were viewed at 545/30-nm and 593/45-nm emission, taking advantage of the natural spectrum shift (Greenspan et al., 1985) with excitation at 488 nm and 568 nm. Autofluorescence was negligible in unfixed material under these conditions.

Nitrogen, Phosphorous, and Potassium Content
Seed nitrogen, phosphorous, and potassium were measured by a commercial service using the following accepted methods. Moisture content was estimated by AOAC method 4.1.06 from AOAC (1995). Total N was determined by AOAC method 4.2.08 from AOAC (1998). Total P and K were quantified by methods SW-6010B and SW-846, respectively, from USEPA (2007).

Oil and Protein Quantification
Seed oil and protein content were quantified in pooled seed samples or in single seeds by time-domain (TD)-pulsed 1H-NMR on a Bruker minispec mq20 (Bruker Optics, Billerica, MA). During TD-NMR experiments, the cottonseed samples were excited by radio-frequency pulses in resonance with the Larmor frequency of hydrogen (Callaghan, 1991). Hence the NMR data comprised only the signal from the hydrogen nuclei present in cottonseed molecules. The magnetization of the hydrogen nuclei present in water, oil, and protein molecules in the seeds show distinctly different relaxation patterns after the excitation pulse. This difference is utilized and the oil content in the seeds is noninvasively measured using a spin-echo pulse sequence (Todt et al., 2006). A pulse sequence extracting both the spin-lattice and spin-spin relaxation properties was used to analyze the protein content in cottonseeds using TD-NMR. The reference values of oil and protein content of different seed lines were obtained by conventional means (oil by gravimetric determination following total lipid extractions, and protein by Bradford assay following total protein extraction according to Ferguson et al. [1996]). A chemometric model was built through a partial least-square algorithm (Kramer, 1998) using the reference values and the TD-NMR signals. The protein content in the cottonseed samples was then measured noninvasively against this model.

Carbohydrate Analysis
Plants were genotyped at the seedling stage and were grown in 20-inch pots in an air-conditioned greenhouse. Ten ginned seeds from each of 12 Bnfad2 36A (T3) plants and 6 null-segregant plants were ground in liquid nitrogen and extracted at 80°C three times with 80% ethanol (5 mL, 3 mL, and 2 mL). Lactose (1 µmol g–1) was added at the time of extraction as a quantitative standard. Starch in the insoluble fractions was assayed with a commercially available kit (Megazyme, Bray, Ireland; amyloglucosidase/a-amylase method). For neutral sugars in the soluble fractions, the ethanol was removed under a stream of nitrogen, and the aqueous remainder extracted with chloroform (0.5 mL). The aqueous phase was brought to 5 mL H2O per gram starting material, and 500 µL was passed through a column consisting of (from top to bottom) AG50W cation exchange resin (H+ form; Bio-Rad, Hercules, CA), polyvinyl polypyrrolidone (Sigma, St. Louis, MO), and AG1 anion exchange resin (formate form; Bio-Rad; 250 µL, 100 µL and 250 µL, respectively), washed with 500 µL H2O, and the flow-through filtered through a 0.22-µm nylon HPLC filter (Costar, Cambridge, MA). Each sample was further diluted and sugars from roughly 20 µg starting material were resolved and quantified against standards by high-performance anion exchange chromatography with pulsed-amperometric detection using a CarboPac PA20 column at 40°C, 50 mM NaOH eluent, and quadruple waveform, as recommended by the instrument manufacturer (Dionex, Sunnyvale, CA). Glucose and galactose elute as a single peak under these conditions. Values were quantified relative to the lactose internal standard.


   RESULTS
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 REFERENCES
 
The Brassica napus nonfunctional fad2 (Bnfad2) allele contains two amino acid substitutions within one of the His boxes that comprise the catalytic site (Fig. 2), rendering the enzyme inactive, but not otherwise altering the open reading frame (Broglie et al., 1999). A binary vector, designated pZPHMCFd2 and harboring the Bnfad2 allele between the 5' and 3' flanking regions of the phaseolin gene (Chapman et al., 2001), was expressed in cottonseeds (Fig. 2). The expression of Bnfad2 resulted in a change in the percentages of oleic and linoleic acids in mature cottonseeds (Chapman et al., 2001), presumably through interference with endogenous cottonseed FAD2 activity during seed development and reserve accumulation. Unexpectedly, there was a reduction in seed oil content that was correlated with the severity of change in oleic acid content, such that seeds from transgenic lines with the highest oleic acid content had the lowest oil content (r2 = 0.91 for eight independent transgenic lines, Fig. 3 ). The oil content was reduced by 30 to 50% in the seeds of primary transformants (T1 seed) of seven independent lines.


Figure 3
View larger version (18K):
[in this window]
[in a new window]

  		Figure 3. Oil suppression in T1 seeds of Bnfad2 primary transformants. Eight independent transgenic lines of the modified oleic acid lines harboring the BnFAD2 construct (Chapman et al., 2001) revealed a reduction in seed oil content, and there was an inverse relationship between oil and oleic acid content. Seed oil content was quantified in 15-seed batches by 1H-NMR using refined cottonseed oil as a standard, and performed in triplicate on aliquots of seed previously measured for fatty acid composition. Numbers are averages from three samples for oil and a single representative fatty acid methyl ester sample, all from the same seed lots. The solid diagonal line is from a regression analysis of the values (r2 = 0.91), and the dashed lines represent the 95% confidence limit (Sigmaplot for Windows). Oil content was verified gravitimetrically from hexane extracts of pooled seed samples (eight seeds for each sample).

 
The RT-PCR analysis of RNA extracted from control (Coker 312, 36A null segregants, 11G null segregants) and several transgenic lines showed Bnfad2 expression only in 36A,11G, and 11E transgenic lines (Fig. 2). The presence of transcripts in transgenic line 36A T2, T3, and T4 generations (Fig. 2, panel B) confirmed that the Bnfad2 expression was heritable. The RT-PCR analysis of RNA extracted from different cotton tissues demonstrated the seed-specific expression of the Bnfad2 allele (Fig. 2, panel D).

The substantial reduction in oil content in numerous lines prompted a more detailed investigation of the physiological characteristics of the Bnfad2 transgenic seeds. Multiple individuals were germinated and grown from each of seven transgenic lines, along with corresponding null segregants (siblings lacking the Bnfad2 transgene), vector control (pBI-121) plants, and wild-type Coker 312 (transgenic background). Plants for these studies were typed at the seedling stage by PCR amplification of the Bnfad2 allele (Chapman et al., 2001), and due to seedling emergence differences between Bnfad2 transgenics and controls, germination of the null segregants and wild-type controls was delayed for 10 to 14 d relative to the Bnfad2 lines so that plant development was similar over the period of fruiting and seed development. (No growth differences were noted between BnFAD2 transgenics and nontransgenics after the seedling stage, about 28 d.)

The seeds of the transgenic lines (positive for the presence of the Bnfad2 allele) were significantly smaller (p < 0.001) than those of segregating controls (negative for the presence of the Bnfad2 allele) (Fig. 4 ). This difference in seed size was evident by visual inspection of fuzzy seed and by embryo morphology and was quantified as seed index from ginned, pooled seed samples from individual plants (Fig. 4). Bolls were harvested in two groups for each plant for comparisons: nodes 6, 7, and 8 as one group (representing early boll accumulation), and bolls from all other nodes (higher boll competition for assimilates) in a second group. Results were the same for both sets of comparisons (not shown), so there was no obvious positional or developmental influence for the Bnfad2 impact on seed size.


Figure 4
View larger version (54K):
[in this window]
[in a new window]

  		Figure 4. Upper panel: Bnfad2 transgenic lines (11E-8, 84a-8, 36A-13) showing differences in T2 seed and embryo size in comparison with nontransgenic (Coker 312) and null segregants (55B-7). Bottom panel: quantification of seed index (g/100 seed) for Bnfad2-expressing (polymerase chain reaction [PCR]-positive segregating siblings, red letters) and nonexpressing (PCR-negative segregating siblings, black letters) plants. Seeds were the selfed progeny derived from T1 individuals. For the data shown, seed cotton was pooled from bolls on nodes 6, 7, and 8 for each plant, and fiber removed by a table-top, 10-saw gin. Plants were genotyped by PCR at the seedling stage, and values are means and standard deviations of 5 to 12 individual plants from each line. For comparison, 10 plants each of the wild-type, Coker 312 background and vector control (pBI121) were grown under the same conditions. Seed index for each line was compared for statistical differences by a student's t-test. **p < 0.001 Bnfad2 lines vs. Coker 312.

 
The reduction in seed size was accompanied by the heritable reduction in seed oil content (Fig. 5 ). Average seed oil content for all transgenic lines was 10 to 12% by weight, compared with approximately 20% for controls (null-segregating siblings, vector controls, and Coker 312). As with seed size, this significant reduction in seed oil for all transgenic lines (p < 0.0001) was observed for bolls pooled from two different groups. Results in Figure 5 are from the group of bolls harvested from nodes 6, 7, and 8 that had developed earliest on the plants. On the other hand, the lint percentage (percentage of total seed cotton weight that was ginned fiber) of these seeds expressing the Bnfad2 allele was significantly higher than the corresponding null segregants or controls (Fig. 6 ; p < 0.004). To confirm that the changes in BnFAD2 seeds were confined to the embryo, samples of the ginned seeds for each transgenic line and the corresponding null siblings were separated into seed coats and embryos. Reductions in total seed mass (Fig. 4) were attributable to changes only in embryo mass (Fig. 7 ); the mass of the seed coats was not different between transgenics (+) and controls (–). Other seed constituents were impacted in transgenic cotton plants compared with nontransformed siblings. The average seed N, K, and P content trended lower in transgenic seeds (Table 1 ) on a seed weight basis (in fact, P was significantly lower at p < 0.01). Moreover, on a per fiber basis (extrapolated to per bale), transgenics contained significantly less seed nitrogen, K2O, and P2O5, suggesting that soil nutrient requirements would be substantially reduced for transgenic lines.


Figure 5
View larger version (43K):
[in this window]
[in a new window]

  		Figure 5. Comparison of average oil content (% by weight) for seeds harvested from transgenic Bnfad2-expressing (polymerase chain reaction [PCR]-positive segregating siblings, red letters) and nonexpressing (PCR-negative segregating siblings, black letters) plants. Seeds were the selfed progeny derived from T1 individuals. For the data shown, seed cotton was pooled from bolls on nodes 6, 7, and 8 for each plant, and fiber removed by a table-top, 10-saw gin. Seed oil was quantified by pulsed-field 1H-NMR on a Bruker minispec seed analyzer, using cottonseed oil for calibration. Plants were genotyped by PCR at the seedling stage, and values are means and standard deviations of 5 to 12 individual plants from each line (pooled samples of 15 seeds analyzed in duplicate for each plant). For comparison, 10 plants each of the wild-type, Coker 312 background and vector control (pBI121) were grown under the same conditions. Seed oil content for each line was compared for statistical differences by a student's t-test. **p < 0.00001 BnFAD2 lines vs. Coker 312.

 

Figure 6
View larger version (41K):
[in this window]
[in a new window]

  		Figure 6. Comparison of average lint percentage by weight (100 x grams lint/grams lint + grams seed) for bolls harvested from transgenic Bnfad2-expressing (polymerase chain reaction [PCR]-positive segregating siblings, red letters) and nonexpressing (PCR-negative segregating siblings, black letters) plants. Seeds were the selfed progeny derived from T1 individuals. For the data shown, seed cotton was pooled from bolls on nodes 6, 7, and 8 for each plant, and fiber removed by a table-top, 10-saw gin. Plants were genotyped by PCR at the seedling stage, and values are means and standard deviations of 5 to 12 individual plants from each line. For comparison, 10 plants each of the wild type, Coker 312 background, and vector control (pBI121) were grown under the same conditions. Lint percentage for each line was compared for statistical differences by a student's t test. **p < 0.004 BnFAD2 lines vs. Coker 312.

 

Figure 7
View larger version (82K):
[in this window]
[in a new window]

  		Figure 7. Comparison of total seed weight, embryo weight, and seed coat weight reveals that the reduction in seed size in Bnfad2 transgenic lines is attributed to altered embryo mass and not due to changes in seed coat weight. Data are averages from 15 seeds of each line, and the variability (standard deviation) was generally less than 10%. Samples were from plants used in Figures 4–6GoGo.

 

View this table:
[in this window]
[in a new window]

  		Table 1. Average seed index, seed oil, fiber, seed N, seed P, and seed K in lines expressing the Bnfad2 allele (pos, positive) versus those that were not (neg, negative). Seed N, P, and K were calculated on a seed weight basis and on a fiber basis (1 bale equals 480 lbs). Values for each line are the averages of T2 seed samples from 4 to 12 plants in each line. The p values were obtained by t test between the average values of the positive and negative lines.

 
Examination by bright-field microscopy of cross sections through cotyledons of mature cottonseeds (Fig. 8 ) showed that, compared with control seeds (Coker 312), those lines with severe reductions in seed oil content had thinner cotyledons with smaller or partially collapsed parenchyma cells, which is the principal location of seed oil (lipid) bodies. Ultrastuctural analyses of parenchymal cells by transmission electron microscopy confirmed the disruption in lipid body and protein body morphologies in the transgenic lines (Fig. 9A, 1–3 ). Visualization of seed lipid bodies by Nile red staining and confocal scanning fluorescence microscopy demonstrated the altered abundance, size, and subcellular distribution of lipid bodies in Bnfad2-expressing lines (Fig. 9B, 1 and 2). Oleosins have been suggested to stabilize oil bodies, especially during embryo desiccation (Murphy et al., 1989). Two seed maturation genes (MATP6 and MATP7) were analyzed by RT-PCR in transgenic and control lines (Fig. 10 ). MATP6 and MATP7 encode the major cottonseed oleosins, which, like in other oilseeds, comprise most of the protein in oil body half-unit membranes (Huang, 1992; Keddie et al., 1992). Transcript levels were generally similar for both oleosin genes in seeds of all lines (transgenic and nontransgenic), suggesting that the reduced oil content in Bnfad2 transgenics was not attributable to dramatically altered oleosin expression.


Figure 8
View larger version (167K):
[in this window]
[in a new window]

  		Figure 8. Comparison of cellular and subcellular features in cotyledons of Bnfad2 transgenic embryos to those of the nontransgenic Coker 312 background by bright-field microscopic analysis revealed changes in cotyledon thickness, cell organization/shape, and numbers of subcellular organelles in Bnfad2 lines. A. Coker-312 (oil, 20%; protein, 38%). B. 84a-8 (oil, 7%; protein, 10%). C. 11E-8 (oil, 4%; protein, 4%). All images at the same magnification; bar represents 50 µm.

 

Figure 9
View larger version (86K):
[in this window]
[in a new window]

  		Figure 9. Comparison of cellular and subcellular features in cotyledons of BnFAD2 transgenic embryos to those of the nontransgenic Coker 312 background. A. Conventional transmission electron microscopy of cotyledon cells of control (1) and transgenic (2, 3) lines (1, Coker 312; 2, 84a-8; 3, 11E-8). Bar represents 2 µm. B. Cells of trangenics had fewer lipid bodies and protein bodies. Confocal scanning fluorescence images of cotyledons of control (1, Coker 312) and transgenic (2, 84a-8) seeds stained with a neutral lipid-specific stain (Nile red) to reveal oil bodies. LB, lipid body; PB, protein body. Scale bar represents 10 µm.

 

Figure 10
View larger version (44K):
[in this window]
[in a new window]

  		Figure 10. Panels A and B: Reverse transcriptase–polymerase chain reaction (RT-PCR) analysis of major cotton oleosin (MATP6 and MATP7) genes showing transcript levels in control, nontransgenic (Coker 312, 36A null-segregant T2, 11G null-segregant T2) and transgenic lines (36AT2, 36AT3, 36AT4, 11GT2, 11ET2). Panel C: Amplification of actin transcripts (539 bp) by RT-PCR served as a control for all the samples. Levels of oleosin transcripts were generally similar between controls and transgenics. Here amplification was for 30 cycles (shown) but also was performed at 20 cycles and 35 cycles (not shown), and there was little difference in product amounts among the genotypes.

 
The reduced oil content and disrupted embryo morphology suggested that germination might be affected in the Bnfad2 lines. Indeed, when germination and seedling growth were scored for several transgenic and nontransgenic siblings of several Bnfad2 lines, there was a notable reduction in seed germination and seedling growth in transgenics (Table 2 ). Seed germination was reduced considerably in low-oil lines, and seedling growth was slowed as well. Although not shown here, once seedlings were transplanted and transferred to the glasshouse and true leaves were formed, there were no differences in growth between BnFAD2 transgenics and nontransgenic siblings by 21 to 28 d. Development of nodes, flowers, fruit, and plant height were essentially the same for transgenics and nontransgenics, consistent with the notion that the activity of the phaseolin promoter and the BnFAD2 allele was embryo specific.


View this table:
[in this window]
[in a new window]

  		Table 2. Summary of seed oil content, percentage germination, and 5-d-old seedling growth characteristics for selected transgenic (+) lines (T3) compared with controls (Coker 312 untransformed, pBI121 vector control, and null segregants [–]).

 
To confirm the heredity of seed characteristics and to obtain additional biochemical data for seeds, T3 plants of two representative lines (36A and 11E; 12 plants for each line) were compared with null-segregating siblings (six plants for each line) in a glasshouse trial. Single-seed oil measurements and PCR confirmed the identity of transgenic individuals and their null-segregating siblings. Bolls were harvested and combined for each plant, and average seed, fiber, oil, and protein yields were evaluated (Fig. 11 and Table 3 ). Total seed-cotton yield was similar between transgenics and nontransgenics. But like previous generations, there was an alteration in the oil, protein, and lint percentage, such that the Bnfad2-expressing lines had reduced oil and protein and increased lint, compared with null-segregating controls (Fig. 11, upper panel). On a per gram basis, the fiber content of transgenic seeds was higher than that in nontransgenics (Fig. 11, lower panel); for line 36A, this was a significant increase (p < 0.005). When overall yield parameters were summarized for these two lines, differences between transgenics and nontransgenic siblings derived from line 11E were significant for oil, protein, and seed size (p < 0.005, 0.05, and 0.05, respectively), but not for lint percentage (Table 3). The alteration in carbon allocation in line 36A showed significant changes in seed oil, protein, seed size, and overall fiber yield despite insignificant differences in total seed-cotton yield (Table 3). In fact, total seed cotton yield was not significantly different between 36A positive and 36A negative plants (p > 0.44); however, average fiber yield per plant was increased by 28% (p < 0.05). It is possible that a reduced demand for carbon in the embryo feeds back to ovule tissues, providing more soluble carbohydrate for cellulose synthesis in fiber cells. Indeed, measurements of soluble carbohydrate in Bnfad2 seeds showed a demonstrable increase in sucrose content compared with seeds from null-segregating siblings (Fig. 12 ). Sucrose, the principle form of carbon imported via the phloem (Tarczynski et al., 1992), was twice the level in Bnfad2 seeds than in null segregants, whereas levels of raffinose, a prominent trisaccharide thought to function as a storage reserve and compatible solute (Hendrix, 1990), was not altered. Of the minor sugars, galactinol, the galactosyl donor for the synthesis of raffinose and stachyose, was reduced in transgenic seeds, while trehalose glucose/galactose, fructose, and stachyose were unchanged. Starch is not a prominent storage reserve in mature cottonseed (Hendrix, 1990), and consistent with this, very little starch was present, and there was no significant difference in starch levels in seeds of Bnfad2 (1.27 ± 0.44 mg/g fresh wt.) and null segregants (1.54 ± 0.43 mg/g fresh wt.). Taken together, these results suggest that it is possible to influence sucrose partitioning in cottonseed tissues by manipulating embryo sink strength.


Figure 11
View larger version (21K):
[in this window]
[in a new window]

  		Figure 11. Upper panel: Comparison of average oil content, protein content, and lint for seeds harvested from Bnfad2 plants (pos, polymerase chain reaction [PCR]–positive segregating siblings) and control (neg, PCR-negative segregating siblings) plants. Seeds were the selfed progeny derived from T3 individuals (12 plants each for 36A and 11G positive; 6 plants each for the corresponding null segregants). Seed cotton was pooled from each plant, and fiber removed by a table-top, 10-saw gin. Seed oil and protein were quantified by pulsed-field 1H-NMR on a Bruker minispec seed analyzer. Seed protein was also verified by Bradford assay of total protein extracts (Ferguson et al., 1996). Values are averages and standard deviations. Lower panel: the amount of fiber per gram of seed was significantly higher (p < 0.003) in Bnfad2 plants than in corresponding controls.

 

View this table:
[in this window]
[in a new window]

  		Table 3. Summary yield characteristics for T3 plants derived from two independent transgenic lines, 36A and 11E, compared with null-segregating siblings. Averages are per plant ± standard deviation. n = 12 for transgenics (pos, positive) and 6 for nontransformed siblings (neg, negative).

 

Figure 12
View larger version (22K):
[in this window]
[in a new window]

  		Figure 12. Soluble carbohydrates were analyzed from pooled seed samples of T3 plants derived from line 36A (see Fig. 10 and Table 2). Plants (twelve 36A positive, six 36A negative) were polymerase chain reaction–typed at the seedling stage and were grown in 20-inch pots in an air-conditioned glasshouse. Total sugars were extracted from ginned seeds and identified and quantified by high-performance anion exchange chromatography with pulsed-amperometric detection according to a lactose standard. *p < 0.003; ** p < 0.00003.

 

   DISCUSSION
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
REFERENCES
 
Some indications of the mechanism(s) responsible for oil and protein suppression in Bnfad2 lines can be gleaned from comparing our results with those of others. The modification of cottonseed fatty acid composition has received considerable attention in recent years, and the generation of transgenic plants with a variety of oil compositions has been accomplished (Liu et al., 2008). In most cases, dramatic alterations in fatty acid composition (e.g., high oleic, low palmitic) have not impacted the physiology of the cottonseeds or seedlings in a negative manner, with the exception of high stearic acid levels (Liu et al.,2002). These observations are in line with other oilseed crops including sunflower, soybean, and canola (Cahoon et al., 2007). In fact, the modification of oleic acid content by two other biotechnology strategies (RNAi or antisense suppression of FAD2) in cotton have not shown any impact on seed reserve content or seed germination, even with oleic acid percentages up to 80% (Liu et al., 2008). Here oleic acid percentages were raised modestly to approximately 30%, far below those in RNAi lines (Liu et al., 2002). This argues that it is not the proportion of oleic acid in seeds, per se, that negatively impacts seed germination and seedling emergence in Bnfad2-expressing lines (Table 2). Instead, we propose that the reduction in seed reserve content (total oil and protein) is responsible for the poor seedling vigor, and that this is due to interference of the Bnfad2 allele with ER-mediated assembly of oil bodies during seed development.

Altered oil body morphology has been associated with poor seed viability and reduced seedling vigor, presumably due to the inefficient mobilization of oil reserves (Siloto et al., 2006). Although microscopic examination of the Bnfad2 cottonseeds revealed aberrant oil bodies (Fig. 9), the transcript levels for both of the oleosin genes (MATP6 and MATP7) were generally similar (Fig. 10). Overall numbers of oil bodies were substantially reduced in seeds expressing the Bnfad2 allele. Additionally, the size of oil bodies in Bnfad2 lines was quite variable, with some oil bodies nearly 8 µm in diameter in these cells (Fig. 9), similar to phenotypes in oleosin-suppressed lines (Siloto et al., 2006). It is likely that the disruption in normal oil body ontogeny and overall reduction in oil content in Bnfad2 lines were in part responsible for the poor seed germination and seedling vigor.

The subcellular abnormalities may result from a general overexpression of the nonfunctional fad2 allele, which disrupts the ER machinery involved in the biogenesis of oil bodies (and protein bodies). Indeed the Bnfad2 sequence showed the seed-specific expression (Fig. 2, panel C) and contains a C-terminal ER-retrieval motif (Table 4 ) identified by others (McCartney et al., 2004). Although not quantified, the endogenous cottonseed FAD2-1 gene was expressed in all low-oil lines examined (transcripts were detected by RT-PCR, not shown), indicating that the mechanism of oil suppression was not due to silencing of endogenous FAD2 expression but rather resulted from posttranscriptional mechanisms conferred by Bnfad2. Instead, a generalized disruption in oil body formation is more likely, which also might be expected to impact protein body biogenesis since these processes actively overlap in the ER of developing oilseed cotyledons (Murphy et al., 1989; Kinney et al., 2001). A somewhat related possibility is that a general unfolded protein response is induced in developing cottonseeds expressing the Bnfad2, and this interferes with ER-dependent formation of oil bodies and protein bodies, as has been reported for the floury2 mutants of maize (Boston et al., 1991). Still another explanation is that the Bnfad2 allele behaves as a dominant negative mutation, binding with endogenous cotton FAD2 (or other oil-modifying enzymes) and diminishing its catalytic efficiency. There is some circumstantial support for this scenario since expression of a Crepis palestina diverged FAD2 in Arabidopsis seeds resulted in elevated oleic acid content, and the coexpression of C. palestina FAD2 restored the normal levels of oleic acid (Singh et al., 2001). Future experiments with Bnfad2-specific antibodies, or epitope-tagged Bnfad2 transgenics will help address these possibilities.


View this table:
[in this window]
[in a new window]

  		Table 4. Predicted bipartite endoplasmic reticulum (ER) retrieval motif at C terminus for BnFAD2. Alignment of the C-terminal amino acid sequences of FAD2 from Brassica napus (Bnfad2), Gossypium hirsutum (GhFAD2-1, GhFAD2-2, GhFAD2-3), and Arabidospsis thaliana (AtFAD2) spanning the consensus ER retrieval motif (McCartney et al., 2004). Underlined sequences are the conserved aromatic amino acid–enriched residues in FAD2 that are necessary for ER retrieval. Italicized residues indicate identical or conserved amino acids in targeting sequence.

 
One intriguing aspect of the low-oil cotton lines was the alteration of carbon flux within the Bnfad2 cottonseeds. The reduction in percentage oil (and to some extent, protein) in the embryo was accompanied by an increase in percentage lint on the ovule surface (Figs. 5, 6, and 10; Table 3). This relationship was heritable and dependent on Bnfad2 expression. Furthermore, a substantial increase in sucrose content was measured in Bnfad2 seeds of T3 plants (Fig. 11), and in this line (36A) there was a significant 28% increase in total fiber (cellulose) yield per plant over the null segregants (Table 3). Taken together, these results suggest that the inefficient conversion of sucrose to oil in embryos results in greater availability of carbohydrate precursors for cellulose biosynthesis on the ovule surface. This represents an interesting interaction in carbon flux between filial and maternal tissues in seeds and might provide new strategies for increasing fiber yield in cotton plants, or other carbohydrate reserves in other oilseed crops, particularly if combined with a gene switch technology to facilitate normal seed germination and robust seedling vigor. Additional work will be required to understand the means of communication at the molecular level between embryo and seed coat, and these Bnfad2 lines will help to address these questions in the future.

One additional unanticipated impact associated with suppression of embryo oil and protein was the reduction in seed N, P, and K (Table 1). This may have significant implications in terms of fiber yield and soil fertility. The results indicate that substantially less soil nutrients would be required by Bnfad2-expressing transgenics to yield the same or more fiber. This could have a substantial positive effect for growers planting cotton in nutrient-poor soils or for those growers that wish to reduce fertilizer inputs in more nutrient rich soils. Additional trials with Bnfad2-expressing lines will be required to establish soil nutrient requirements for these plants, but alteration of sink strength has a clear impact on the accumulation of seed mineral reserves, suggesting that a targeted reduction in certain seed reserves may be an important strategy in some crops for the reduction of input costs.


   ACKNOWLEDGMENTS
 
This work was supported initially by a grant from the USDA-NRICGP (2001-01491) and, in part, by Delta and Pine Land Company.


   NOTES
TOP
 NOTES
ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
All rights reserved. No part of this periodical may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher. Permission for printing and for reprinting the material contained herein has been obtained by the publisher.

Received for publication November 13, 2007.


   REFERENCES
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 





This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Chapman, K. D.
Right arrow Articles by Kinney, A. J.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Chapman, K. D.
Right arrow Articles by Kinney, A. J.
Agricola
Right arrow Articles by Chapman, K. D.
Right arrow Articles by Kinney, A. J.
Related Collections
Right arrow Seed Quality
Right arrow Cotton
Right arrow Cell Biology & Molecular Genetics

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
The SCI Journals Agronomy Journal Vadose Zone Journal
Journal of Natural Resources
and Life Sciences Education		 Soil Science Society of America Journal
Journal of Plant Registrations Journal of
Environmental Quality		 The Plant Genome