A New Opaque Variant of Maize by a Single Dominant RNA-Interference-Inducing Transgene

Corresponding author: Joachim Messing, Rutgers, The State University of New Jersey, 190 Frelinghuysen Rd., Piscataway, NJ 08854-8020., messing{at}mbcl.rutgers.edu (E-mail)

Communicating editor: J. A. BIRCHLER

	ABSTRACT

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

In maize, -zeins, the main protein components of seed stores, are major determinants of nutritional imbalance when maize is used as the sole food source. Mutations like opaque-2 (o2) are used in breeding varieties with improved nutritional quality. However, o2 works in a recessive fashion by affecting the expression of a subset of 22-kD -zeins, as well as additional endosperm gene functions. Thus, we sought a dominant mutation that could suppress the storage protein genes without interrupting O2 synthesis. We found that maize transformed with RNA interference (RNAi) constructs derived from a 22-kD zein gene could produce a dominant opaque phenotype. This phenotype segregates in a normal Mendelian fashion and eliminates 22-kD zeins without affecting the accumulation of other zein proteins. A system for regulated transgene expression generating antisense RNA also reduced the expression of 22-kD zein genes, but failed to give an opaque phenotype. Therefore, it appears that small interfering RNAs not only may play an important regulatory role during plant development, but also are effective genetic tools for dissecting the function of gene families. Since the dominant phenotype is also correlated with increased lysine content, the new mutant illustrates an approach for creating more nutritious crop plants.

THE nutritional value of cereal grain protein is a critical constraint in its use for animal and human consumption. Protein deposits in the specialized tissue of cereal endosperm are mostly deficient in essential amino acids, particularly lysine, tryptophan, and methionine. Because protein supplementation to correct such deficiencies is costly and wasteful of energy in animal nutrition, it is not feasible in vast areas of developing countries that rely on cereals as a sole food resource. Therefore, the improvement of cereal storage proteins for increased amounts of protein-bound essential amino acids has been recognized as a primary objective in breeding.

In maize, several approaches have been undertaken to correct this problem (MAZUR et al. 1999 ). A major breakthrough has been the isolation of the opaque-2 mutant (o2), with superior nutritional properties, albeit undesirable agronomic traits (MERTZ et al. 1964 ). More recently, the generation of quality-protein maize incorporated the enriched o2 protein quality into breeding materials devoid of the endosperm soft texture and pest susceptibility that characterize o2 (CORDOVA 2000 ). The molecular identification of the underlying defect in o2 variants provided the first example of a transcriptional activator gene that directly controls the transcription of maize storage protein or zein genes (SCHMIDT et al. 1987 ; LOHMER et al. 1991 ). It is now well established that O2 plays an important role in multiple endosperm-specific metabolic pathways (DAMERVAL and LE GUILLOUX 1998 ; HUNTER et al. 2002 ).

In maize, zeins, the core of protein endosperm reserves, have been the subject of intense studies due to their abundance, complexity, and impact in the overall nutritional value of maize seed. They have been classified into four subfamilies of -, ß-, -, and -zeins on the basis of their primary structure and differential solubility (UEDA and MESSING 1993 ). Zeins are encoded by single- or low-copy gene loci, with the exception of -zeins. The large -zein component, accounting for >70% of all zein protein, is composed of multiple active genes clustered in several chromosomal locations. Genomics approaches are best suited to identify and characterize the entire -zein gene complement (SONG et al. 2001 , SONG et al. 2002 ; WOO et al. 2001 ; HUNTER et al. 2002 ; SONG and MESSING 2002 ), which has been subdivided into 19- and 22-kD zein subfamilies. From a nutritional point of view, the exceedingly large proportion of codons for hydrophobic amino acids in -zeins is mostly responsible for the imbalance of maize protein reserves. Therefore, the reduction in -zein protein accumulation with biased amino acid content could provide a correction to this imbalance. The o2 mutation differentially inhibits the transcription of the 22-kD -zein component, which is thus greatly reduced in the endosperm (KODRZYCKI et al. 1989 ). Therefore, o2 mutants could be used to achieve such regulation. Unfortunately, several disadvantages have become apparent with the characterization of o2. First, O2 regulates genes other than 22-kD zeins. As mentioned above, the alteration of one or several of these transcripts generates deleterious phenotypes in the plant and grain product. Second, it has been reported that several 22-kD zeins are independent of o2 (CICERI et al. 2000 ; SONG et al. 2001 ), preventing an efficient repression of all 22-kD zein genes. This prompted the question of whether this effect can be reproduced independently of the o2 mutation.

Antisense and RNA interference (RNAi) technologies provide the means to target individual plant components for depletion (WATERHOUSE et al. 1998 ; CHUANG and MEYEROWITZ 2000 ; SCHWEIZER et al. 2000 ). However, in many cases reduction of gene expression is not complete and it is uncertain whether such an approach would be sufficient to achieve a visible phenotype. Since this has not been tested yet in maize seeds, we adopted a system for gene expression restricted to hybrid transgenic plants, first established in Arabidopsis (MOORE et al. 1998 ). This system is composed of the introgression of sequences encoding a modified prokaryotic transcriptional activator from a first group of transgenic plants into a second set with a background of a reporter gene of interest downstream of a plant promoter restricted to a TATA box and modified cis-acting prokaryotic elements, which are recognized and bound by the heterologous transcriptional activator with high affinity. Only the cross between both kinds of plants activates the expression of the reporter transgene (Fig 1A). We report here the successful evaluation of these approaches by obtaining a dominant opaque phenotype through RNAi but not antisense constructs, the stable and heritable expression of recombinant sequences in maize that produce a specific reduction in 22-kD -zein proteins. Moreover, the amino acid composition in transgenic opaque seeds had higher levels of lysine and reduced levels of leucine, alanine, and glutamine, confirming the shift in amino acid balance.

View larger version (17K): In this window In a new window Download PPT slide   Figure 1. Maize transactivation system. (A) Schematic representation of the transactivation system. Two sets of maize lines were produced: an activator line for constitutive expression of a chimeric transcriptional activator (KI), designed for high-affinity recognition and stimulation of a synthetic promoter (TATA box and 2x op, duplicated modified lacI operator binding sites; T, polyadenylation signal sequence), and a target line with no detectable transgene expression at first, since it is under the control of the synthetic promoter not recognized by any maize transcriptional activator (MOORE et al. 1998 ). In addition, both kinds of sequences include the elements for stable expression in maize. In the progeny of a cross between these lines where the active and passive sequences coexist, transactivation will take place with accumulation of the transcriptional activator and induction of the target gene. Regulation of the latter will depend on the characteristics of the promoter (P) used in the activator line. (B) Transactivation system antisense construct (A3). op, modified operator sequences of the E. coli lac operon (MOORE et al. 1998 ); (-50)-35S/TATA, (-50) promoter and TATA box of 35S CaMV; -zein, 5' UTR and coding region of the maize azs22.8 -zein cDNA, inserted in the antisense orientation in the SmaI site of the pOP polylinker; T-35S, 3' UTR of 35S CaMV. The right half of the scheme shows the pUbi-bar plasmid (CHRISTENSEN and QUAIL 1996 ) used in all the experiments either as part of an entire construct (transactivation system) or in cotransformation experiments (RNAi experiments). P-Ubi-in, promoter and first intron of the maize ubiquitin gene; Bar, coding region of the phosphinothricin acetyltransferase gene; T-nos, 3' UTR of the nopaline synthase gene. (C) Construct for constitutive expression of the hybrid modified E. coli/S. cerevisiae KI-His (MOORE et al. 1998 ). P-35S, promoter of 35S CaMV; KI, coding region of pKI-His; T-35S, 3' UTR of 35S CaMV. ORFs are shaded or solid boxes; regulatory sequences are represented by open boxes. Arrows indicate the direction of transcription.

	MATERIALS AND METHODS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Plant material:
High type II callus maize (Zea mays) seed lines A and B were kindly provided by Kan Wang, Plant Transformation Facility, Iowa State University. W64A and W64Ao2 lines were from our own collection. All plants were grown under greenhouse conditions.

Transformation constructs:
For the transactivation system, constructs were derived from those previously generated for Arabidopsis thaliana (Fig 1A; MOORE et al. 1998 ). For the reporter lines of the 22-kD -zein, the entire azs22.8 coding region (SONG et al. 2001 ) was purified as an EcoRI-NdeI fragment, filled in with Klenow enzyme, and cloned in the antisense orientation into the SmaI polylinker site of pOp (MOORE et al. 1998 ), downstream of duplicated modified operator sequences of the lac operon followed by a minimal (-50) 35S cauliflower mosaic virus (CaMV) promoter and upstream of a 35S CaMV terminator. The entire insert was purified, subcloned into the SmaI site of pBluescript SKII, and reexcised with HindIII for ligation in the HindIII site of pUbi-bar with the transcription units in opposite orientation (Fig 1B, construct A3). The entire insert of pKI-His, including the 35S CaMV promoter, a coding region composed of the engineered version of a translational fusion between the Escherichia coli lacI and the Saccharomyces cerevisiae Gal4 genes and the 35S CaMV terminator (MOORE et al. 1998 ), was excised by PstI digestion, end filled, and ligated to pUbi-bar (CHRISTENSEN and QUAIL 1996 ), a maize expression cassette with the Bar gene as a selectable marker, driven by the maize ubiquitin1 promoter and first intron and 3' end regulatory sequences of the nopaline synthase gene (Fig 1Cconstruct KI).

Two constructs were designed for RNAi experiments. An inverted repeat (IR; Fig 3A) construct was generated following the guidelines of TAVERNARAKIS et al. 2000 . A 483-bp XbaI-NcoI including 5' untranslated region (UTR) sequences and the amino-terminal half of azs22;8 -zein cDNA was ligated to XbaI-digested, dephosphatased pJM2170 (LAI and MESSING 2002 ) between 1.1 kb of the 27-kD zein promoter sequence and the 35S CaMV terminator, so that the coding sequences established an inverted repeat without intervening sequences. E. coli Sure strain (Stratagene, La Jolla, CA) was transformed with this construct to reduce the frequency of rearrangements, and supercoiled plasmid DNA was purified and directly used for cotransformation experiments together with pUbi-bar. To increase the stability and ease of the detection of the transgene and to increase the efficiency of silencing (LEVIN et al. 2000 ), an additional construct with the green fluorescent protein (GFP) gene-coding region as intervening sequence was made (Fig 4A, LIN). A 313-bp fragment of the amino-terminal half of the -zein cDNA was amplified by using primers R22-1 (5'-GGCGAGCGTCTACAACAACC) and R22-2 (5'-CACAACGAGAGGGCTAGATGAAAG) and directly cloned into pGEM-T Easy (Promega, Madison, WI). The insert was excised with SacII and SpeI and subcloned into the corresponding sites of a pBluescript SKII-derived plasmid carrying the GFP-coding region (CLONTECH, Palo Alto, CA), cloned between the BamHI and EcoRI sites. The same -zein PCR fragment was excised with NotI and cloned into the NotI site immediately after the GFP stop codon and upstream of the terminator, recombinant colonies with the right orientation were selected, and the plasmid DNA was digested with SacII to isolate the entire insert. This was gel purified, filled in with T4 DNA polymerase, and ligated to phosphorylated XbaI linkers (no. 1081, New England Biolabs, Beverly, MA). The excess linkers were eliminated by agarose gel electrophoresis followed by spin-columm agarose removal. The purified insert was digested with XbaI and then ligated to an XbaI-linearized, dephosphatased, pBluescript SKII-based vector between a 1.3-kb promoter fragment of the zp22/6 zein gene (SONG et al. 2001 ) and the 35S CaMV terminator. This promoter sequence has been determined to strongly activate the expression of transgenic reporter genes in maize endosperm (G. SEGAL and J. MESSING, unpublished results). This construct was used as supercoiled DNA in cotransformation, together with pUbi-bar.

View larger version (96K): In this window In a new window Download PPT slide   Figure 2. Transactivation system in maize. (A) Southern blot analysis of A3 lines. Leaf DNA from putative transgenic plants was digested with EcoRI, blotted onto a nylon membrane, and hybridized with a 444-bp PCR fragment of the Bar gene. A 1.5-kb fragment is recognized by this probe if the plant is transgenic. Selected A3-B line is designed with B at the top. (B) Southern blot analysis of KI lines leaf DNA digested with EcoRI and hybridized with the Bar gene probe. A 1.5-kb fragment is recognized by this probe if the plant is transgenic. KI-GR-11 line characterized in C is identified by an asterisk at the top. Similar results were obtained with a 1.6-kb XbaI fragment of the coding region of KI-His (data not shown). (C) PCR and SDS-PAGE protein analysis of seed derived from lines KI-GR-11, expressing KI-His and A3-B, a target line with an antisense 22-kD zein sequence, and progeny from their cross. DNA and total zeins were extracted from individual seeds. PCR was conducted with specific primers for Bar (top) and KI (middle), and proteins were fractionated and stained. M, 1-kb ladder DNA molecular weight marker (Invitrogen, Carlsbad, CA). The migration of the 22-kD zein is indicated by an arrow. BA, untransformed parental line.

View larger version (74K): In this window In a new window Download PPT slide   Figure 3. Analysis of transgenic plants derived from a simple IR construct. (A) Construct with simple IRs. P-27kD, promoter of the maize 27-kD -zein gene (LAI and MESSING 2002 ). -zein, XbaI-NcoI fragment of the 5' UTR and coding region of the maize 22-kD -zein cDNA, ligated to an XbaI site 3' of P-27kD. Maize selection cassette is as in Fig 2. (B) Southern blot hybridization of DNA extracted from putative transgenic plants with the Bar probe. (C) The same blot as in A was stripped of signal and hybridized with a 203-bp XbaI-EcoRI probe of the 3' UTR of the 35S CaMV gene. This probe is unique to the nonselectable plasmid of this experiment, indicating cotransformation. (D) SDS-PAGE analysis of total zeins extracted from T0 seeds of putative transgenics. BA, untransformed parental line. The migration of the 22-kD zein is indicated by an arrow. Two seeds from lines IR-C and IR-D with suppressed 22-kD zeins are identified. (E) PCR and SDS-PAGE analysis of DNA and total zeins extracted from T1 seeds derived from line IR-D in C. Three IR-D plants were either selfed or outcrossed to BA, and individual seeds from these progeny were analyzed. PCR was performed with primers specific to the Bar transgene. BA, untransformed parental line. o2, W64Ao2 seed. M, 1-kb ladder DNA molecular weight marker (Invitrogen). The migration of the 22-kD zein is indicated by an arrow. (F) Photograph of progeny seeds from IR-D, plant D3 in E, showing opaque phenotype. o2, W64Ao2. BA, untransformed parental seed.

View larger version (44K): In this window In a new window Download PPT slide   Figure 4. Analysis of transgenic plants derived from a construct with inverted repeats separated by a linker (LIN). (A) RNAi construct with linker (LIN). P-zp22/6, 1.3-kb promoter fragment of the maize zp22/6 gene (SONG et al. 2001 ; G. SEGAL and J. MESSING, unpublished results). -zein, PCR fragment of the coding region (see MATERIALS AND METHODS), cloned in opposing fashion flanking the nonactive pGFP-coding region (CLONTECH). (B) Southern blot hybridization of DNA extracted from putative LIN transgenics, digested with BamHI, and hybridized with a 763-bp XbaI fragment spanning the entire GFP coding region. Since the construct has a unique BamHI site, this probe uncovers a different restriction fragment length polymorphism pattern for independent transgenic events. Migration of a MW marker (1-kb DNA ladder, Invitrogen) is indicated in kilobases on the right. (C) PCR and SDS-PAGE analysis of DNA and total zeins extracted from T0 transgenic seed. PCR was performed with primers specific to the nonactive GFP sequence in the transgene. BA, untransformed parental line. M, 1-kb ladder DNA molecular weight marker (Invitrogen). The migration of the 22-kD zein is indicated by an arrow. (D) Photograph of seeds from line LIN-O in C showing an opaque phenotype. o2, W64Ao2. BA, untransformed parental line.

Maize transformation:
Transgenic plants were generated by particle bombardment of highly embryogenic type II callus according to FRAME et al. 2000 . Immature 1.0- to 2.0-mm embryos were aseptically dissected, incubated in callus-inducing medium at 25° in the dark, and transferred every week. Once a friable, highly proliferating type II callus had developed, it was arranged in the center of osmotic-medium plates and shot with DNA-coated gold particles (FRAME et al. 2000 ). We used single plasmids containing both the selection elements and the gene of interest (transactivator system) or cotransformation (RNAi experiments). After 2–4 months of bialaphos selection, rapidly growing callus lines were screened by the chlorophenol red assay (KRAMER et al. 1993 ) and plants were regenerated. Transgenesis was confirmed by Southern analysis. T₀ plants were either selfed or outcrossed to untransformed B x A hybrids.

Molecular procedures:
Southern analysis was performed as previously described (DAS et al. 1990 ). As a preliminary characterization, all putative transgenic plant DNA was hybridized to a 444-bp PCR probe of the Bar gene-coding region. To further analyze the KI transactivation plants, a 1.6-kb XbaI fragment spanning almost the entire transactivator open reading frame (ORF) was used. LIN line DNA was analyzed by labeling a 763-bp XbaI sequence of the entire pGFP (CLONTECH) insert. Total zeins were extracted from mature kernels with 70% ethanol and 2% ß-mercaptoethanol and fractionated by SDS-PAGE (DAS et al. 1990 ). PCR reactions were carried out in 50 µl and 0.5 mM of each of the following primers: for KI, KI-1 (5'-CAGGGCCAGGCGGTGAAG) and KI-2 (5'-TTTTTGGGTTTGGTGGGGTATCT); for bar, Bar-for (5'-GACATGCCGGCGGTCTGCACCATCG) and Bar-rev (5'-GCCAGAAACCCACGTCATGCCAG); for GFP, GFP-1 (5'-CGTCCAGGAGCGCACCATCTTCTT) and GFP-2 (5'-ATCCGCGTTCTCGTTGGGGTCTTT); and for -zein, R22-1 and R22-2 as described above. PCR conditions were 94° for 3 min, followed by 30 cycles of 30 sec at 94°, 30 sec at 57°, and 1 min at 72°, with a final extension for 10 min at 72°. A total of 10% DMSO was added in the PCR reactions to amplify Bar.

Amino acid composition analysis:
Protein-bound amino acid seed composition was determined in samples submitted to the New Jersey Feed Laboratory, Trenton, New Jersey.

	RESULTS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

A transactivation system reduces 22-kD zein accumulation in hybrid transgenic plants:
An outline of the components for a maize transgenic system regulated by transactivation is presented in Fig 1A. Two kinds of transgenic plants were required, one expressing a transcriptional activator that recognizes a unique synthetic promoter and the other with a chimeric gene consisting of the synthetic promoter controlling the gene of interest. When the transgenic lines are crossed, the synthetic promoter is recognized and the target transgene is transcribed. For the purpose of 22-kD antisense inhibition, we generated two sets of transgenic maize plants. One, termed A3, provided the target moiety of the system. Under an enhancerless 35S CaMV promoter, consisting mainly of the TATA box with two altered lac gene operator sequences for increased affinity, the entire azs8 22-kD zein cDNA (SONG et al. 2001 ) in the antisense orientation was cloned (Fig 1B). The other set of lines involved the stable integration of sequences encoding a modified transcriptional factor (MOORE et al. 1998 ) that is under the control of the constitutive 35S CaMV promoter and that can specifically bind to the altered lac operator sequences (Fig 1C, KI). However, since the lac operator is absent in the maize genome, expression of the transcriptional activator cannot result in transcription of any endogenous gene. This transgenic line should be useful for any future transactivation experiments in maize, which would require only the transformation of one line with a chimeric gene containing the synthetic promoter and a gene of interest. We did not expect to detect any antisense gene activity in these separated sets of transgenic plants. On the other hand, in a cross between both classes of plants, the transcriptional activator derived from KI should recognize and bind the upstream cis-acting elements from A3 and activate the transcription of the antisense -zein, which would affect the accumulation of 22-kD protein in the hybrid transgenic endosperm. Several lines were established for each kind of construct, as shown by Southern blot analysis (Fig 2A and Fig B), with copy number ranging from 1 to >10. For transactivation experiments, 21 hybrid transgenic lines combining diverse transactivator and -zein antisense events were screened by SDS-PAGE. Most seeds analyzed uncovered little or no reduction in the 22-kD zein level (data not shown). Line KI-GR-11 (designated by an asterisk in Fig 2B), which appears to have one entire copy of the transgene, crossed with A3-B, the line containing the antisense gene, revealed a significant reduction and was further analyzed. Individual T₁ seeds, originated from the cross of two selfed T₀ plants of these lines, were analyzed side by side from the parental lines by PCR and SDS-PAGE (Fig 2C). It is apparent that visible reductions in the 22-kD zein have occurred only in the seed resulting from the combination of KI with A3 in the same plant, while no change can be detected in the parental lines. No additional phenotype was detected in this progeny, which germinated and developed normally in every other respect. Inhibition of the 22-kD zein was not enough in this case to detect an opaque phenotype. These results show that, while the transactivation system works well in maize, the substantial residual level of 22-kD zein protein in hybrid transgenic plants proves that antisense constructs were unable to suppress 22-kD zein synthesis sufficiently to induce an opaque phenotype.

A simple inverted repeat RNAi construct triggers complete disappearance of the 22-kD zein:
Since antisense RNA expression did not achieve the levels of 22-kD zein reduction equivalent to the knockout of O2, an alternative approach was undertaken. More recently, it has been shown that RNAi technology is an additional penetrant and an efficient way of depleting individual mRNAs in eukaryotes (TAVERNARAKIS et al. 2000 ; MATZKE et al. 2001 ). It is believed that in contrast to antisense RNA, RNAi triggers the post-transcriptional gene-silencing pathway, which seems to be conserved in plants. Therefore, we made a simple construct (see MATERIALS AND METHODS), composed of an inverted repeat of a 22-kD zein cDNA fragment flanked by the endosperm-specific 27-kD -zein promoter and the 35S CaMV polyadenylation signal sequences (Fig 3A). Because of the concern that an RNAi construct with inverted repeats would be difficult to clone into a large plasmid, the silencing plasmid was used in cotransformation experiments with a second plasmid for expression of a maize selectable marker. The well-established pUbi-bar maize cassette of the 5' maize ubiquitin gene regulatory sequences and the NOS 3' UTR driving the expression of the phosphinothricin acetyltransferase (Bar) gene, for selection of transgenic callus with the herbicide bialaphos (CHRISTENSEN and QUAIL 1996 ), was mixed with the RNAi-inducing plasmid for gold particle coating and bombardment of the embryogenic callus. Southern blot analysis of the regenerated plants with either the bar gene or the 35S CaMV 3' UTR as a probe (Fig 3B and Fig C) revealed that a majority of the transgenic plants obtained had insertions of both the selectable and the cotransformed, nonselectable plasmids in 1–10 copies. Total zein extracts from T₀ mature seed products of self-fertilization or outcrosses to untransformed parental lines were analyzed by SDS-PAGE (Fig 3D). We could identify lines (IR-C and IR-D) with almost complete suppression of the 22-kD zein without compensatory changes in the other zein components. Differential levels of the larger -zein are due to segregation of the B and A lines that were used for transformation. The reduction of the 22-kD zein in these transgenic plants is much stronger than in those with antisense constructs and is at least as pronounced as the one observed with mutant o2 seeds (Fig 3E). Residual protein detected at the position of the depleted 22-kD zeins is probably a divergent 19-kD -zein, such as z448F14 (SONG and MESSING 2002 ), that comigrates with the targeted proteins. From a total of 13 lines examined, 5 were found to segregate seeds with normal and reduced 22-kD zein phenotypes. We observed that in the case of progeny derived from selfed plants of the transgenic line IR-D, the affected seeds appeared at the ratio of 3:1 relative to normal seeds, the expected frequency for a single dominant trait (data not shown). We next examined whether the transgene is stably transmitted. Seeds from T₀ IR-D plants were planted and the respective plants were either selfed or outcrossed. DNA and zein analysis from the resulting T₁ seeds (Fig 3E) provided evidence for 22-kD zein depletion in transgenic seeds from all these plants, indicating that transgene suppression is stable for at least two generations. More importantly, the quantitative changes in relative zein abundance were concomitant with the production of an opaque phenotype in the transgenic seed (Fig 3F). The appearance of a characteristic chalky texture across the entire endosperm cosegregated with protein alteration. Therefore, RNAi appears to be a useful genetic tool for studying the regulation of gene families in maize.

RNAi produced by 22-kD zein repeats separated by a linker sequence:
Since RNAi appeared to simultaneously and specifically silence all members of a multigene family, two additional parameters of this approach were investigated. RNAi can be induced by either a direct inverted repeat or inverted repeats separated by a neutral sequence. In fact, repeats separated by either a silent coding region or an intron have consistently yielded deeper and more stable gene silencing (CHUANG and MEYEROWITZ 2000 ; WESLEY et al. 2001 ). If this is the case, then one might not need a strong promoter. Therefore, a different endosperm-specific promoter from the zp22/6 zein gene (SONG et al. 2001 ) was selected for the synthesis of the RNAi transcripts. A 1.3-kb sequence comprising this promoter was fused to a RNAi sequence of inverted 22-kD coding sequences interrupted by the entire GFP-coding region as shown in Fig 4A. This RNAi construct was introduced into maize by the cotransformation method with a pUbi-bar plasmid as a selectable marker as described above. Southern analysis of DNA extracted from transgenic plant leaves (Fig 4B) corroborated the efficacy of cotransformation, since most of the plants selected presented one to multiple copies of the cotransformed, nonselectable transgene. There was no evidence of GFP expression in transgenic tissues, which is readily detectable when this reporter gene is placed under the control of the zp22/6 promoter (G. SEGAL and J. MESSING, unpublished results). The presence of GFP sequences in transgenic seed was established by PCR. SDS-PAGE of zeins extracted from mature seeds (Fig 4C) revealed an almost complete absence of 22-kD zeins in seeds from some lines, segregating with seeds with a normal phenotype in an apparent 3:1 ratio when originated from selfed plants, without appreciable change in the remaining zein components. The zein depletion pattern observed was indistinguishable from that of the IR lines with simple inverted repeats. At least three lines out of eight examined presented evidence of transgene-induced zein silencing. Again, the protein changes coincided with the appearance of an opaque endosperm (Fig 4D), segregating within families between normal and 22-kD zein-depleted seeds. These results show that gene silencing does not require high levels of RNAi transcripts and can be achieved with inverted repeats separated by other sequences, which are much easier to construct in E. coli. Again, the opaque phenotype can be achieved by a single, dominant Mendelian trait.

Amino acid composition of transgenic seeds:
One of the hallmarks of the opaque seed phenotype is an increase in the amino acid lysine. Therefore, transgenic seeds with an opaque phenotype were chosen for amino acid analysis. T₀ seeds from two transgenic events, IR-D and LIN-G, exemplifying the two types of RNAi construct with strong opaque phenotype, were selected and subjected to a preliminary amino acid composition analysis. It is well established that seed chemical composition is very variable and dependent on the genetic background. Since transformation has been carried out with hybrids of A188 and B73, future analysis will depend on backcrossing the transgene with the RNAi construct to different genetic backgrounds. Still, a preliminary amino acid analysis provides the first indication of the total amino acid changes due to the reduction of the 22-kD zeins. To illustrate the impact of the two types of RNAi constructs, the deviation of the concentration of each amino acid relative to the untransformed line was determined. Prominent deviations are summarized in Table 1. We could detect specific changes characteristic of -zein-deficient phenotypes. Indeed, the concentration of lysine was markedly higher in both transgenic lines. In general, the increase in basic and acidic amino acids and in glycine, and the relative reduction in glutamine, leucine, alanine, and serine, are expected if 22-kD zeins are reduced. However, some amino acid changes varied greatly between the two transgenic events (i.e., methionine). Changes could have been compounded by the reduction of 22-kD zein proteins and the association of the transgenic event to two different segregating genetic backgrounds and/or specific regulatory mechanisms interacting with them (e.g., CHIBA et al. 1999 ). Taken together, our data provide evidence that a new dominant opaque phenotype has been created in maize that, unlike the recessive mutation that prevents the transcriptional activation of a subclass of -zein storage protein promoters, uses a post-transcriptional gene-silencing approach.

	DISCUSSION

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Opaque-2 maize has attracted considerable attention due to its outstanding nutritional quality and the characteristic soft texture it confers to the endosperm. Once the molecular identity and the role of O2 as a transcriptional activator in 22-kD zein transcription were documented (SCHMIDT et al. 1992 ; UEDA et al. 1992 ), it was postulated that the alteration in the relative proportion of zeins was the cause of seed opacity. The changes in protein body structure detected in o2 relative to normal maize and the effect of o2 modifier genes, which restore the vitreous endosperm (GEETHA et al. 1991 ), are in agreement with this interpretation. On the other hand, O2 is responsible for pleiotropic effects in a variety of endosperm metabolic pathways (LOHMER et al. 1991 ; HABBEN et al. 1995 ; MADDALONI et al. 1996 ; KEMPER et al. 1999 ; HUNTER et al. 2002 ). Whether one or more of these alterations indirectly cause this phenotype cannot be resolved from mutations affecting the expression of the O2 locus. To separate effects on the transcription of storage protein genes from other genes by O2, a dominant mutation that would result in the selective repression of storage protein synthesis in the presence of normal O2 expression was needed.

We first used the expression of antisense sequences of 22-kD zein genes to block the expression of these storage proteins. Since we were concerned that constitutive expression of 22-kD antisense constructs might affect the regeneration and/or fertility of the resulting transgenic plants, we applied to maize for the first time a transgene expression concept that depends on crossing two different maize lines. Such a concept has previously been tested in Arabidopsis (MOORE et al. 1998 ). Two transgenic components are stably integrated in the plant genome, one encoding a potent transcriptional activator, KI-His, and the other, pOp, with a passive reporter or gene of interest, under the control of an enhancer-less promoter and with duplicated modified cis-acting operator sequences from the E. coli lac operon. Either after double transformation or as a result of the cross between both kinds of plants, the transcription factor binds the operator sequences with high affinity and activates transcription of the target gene. Thus, the process of transformation and generation of transgenic seeds becomes independent of the nature of the gene introduced. We demonstrated the feasibility of this approach by generating sets of maize transgenics, one ubiquitously expressing KI-His and the other with an antisense version of a 22-kD zein cDNA downstream of the operator with an appropriate distance to a TATA box sequence. Indeed, we could show that the 22-kD protein accumulation is reduced (Fig 2C). Moreover, this change is restricted to this zein component and the progeny of the seeds produced from the cross between the two kinds of plants in the system, whereas the parental transgenic lines remained unaltered compared to the untransformed parental line. The nature of these changes, although limited since no other phenotypic alteration is apparent, make us believe that this system will be a useful complement to other inducible expression systems in maize (MOORE et al. 1998 ; OUWERKERK et al. 2001 ; SHIMIZU-SATO et al. 2002 ). Furthermore, it would be possible to generate a variety of transactivator lines, driven by well-characterized promoters, for functional characterization of genes throughout development, as in the study of defective kernel mutant transgenes. Likewise, an allelic series of a target gene could be produced by crossing multiple target gene events with a particular transactivator line. Thus, variable levels of expression could be correlated with phenotypic variation. Since the elements included in the construction of pKI are generally recognized in transgenic monocots, it is likely that this system could be directly introduced in other cereals as well, together with suitable reporter lines.

The efficacy of antisense RNA for gene suppression was already known to be variable (WATERHOUSE et al. 1998 ; LEVIN et al. 2000 ). In our study, it does not constitute a genetic tool since transgenic seeds do not show a visible phenotype. In contrast, expression of DNA sequences, which generate RNA hairpin structures and double-stranded RNA, proved to be a consistent method to considerably reduce a gene product. Analysis of transgenic seeds revealed a higher frequency of events with strong target inhibition in the RNAi constructs compared to the regulated antisense experiment. A minimal requirement of a 26-nucleotide pairing for homology-based depletion (PARRISH et al. 2000 ) makes it possible to control entire multigene families sharing at least short regions of almost identical sequence. With 90–100% of nucleotide sequence identity over the entire open reading frame (Table 2), 22-kD zein genes are a priori suitable targets for these manipulations. RNAi was very effective not only in reducing 22-kD zein expression, but also in giving rise to an opaque phenotype of transgenic seeds. Moreover, the opaque phenotype was correlated with the segregation of a single dominant Mendelian factor and the reduction of zein synthesis. So far, opaque phenotypes have been either recessive or semidominant. While o2 requires a homozygous state of the allele, recreation of the floury-2 phenotype by expression of a defective -zein protein illustrates the potential of a unique transgenic change promoting deep endosperm modification in a semidominant fashion (COLEMAN et al. 1997 ). Here, we present the first maize dominant opaque phenotype generated by transgenic means. Together with Mc and DeB30, they are the only dominant opaque variants described thus far. The difference between 22-kD zein protein levels in mature seed of antisense and RNAi constructs is quite striking and at least partially based on the effectiveness of double-stranded RNA structures in cis vs. in trans to trigger post-transcriptional gene silencing (reviewed by BOSHER and LABOUESSE 2000 ; HSIEH and FIRE 2000 ; HUTVAGNER and ZAMORE 2002 ), which appears to be well conserved in evolution (reviewed by MATZKE et al. 2001 ). Furthermore, we recently could show that plants produce specific double-stranded RNAs as well as a Dicer homolog as part of their normal repertoire of factors controlling gene expression (PARK et al. 2002 ). These small transcripts are also referred to as microRNAs (miRNAs). The preservation of the degradation machinery for double-stranded RNA gives rise to the possibility that aspects of normal endosperm development depend on regulatory small interfering RNAs (HAMILTON et al. 2002 ; LLAVE et al. 2002 ; ZAMORE 2002 ).

View this table: In this window In a new window   Table 2. Nucleotide sequence homology of -zein genes

In contrast to recessive o2 variants, our transgenic constructs triggered the disappearance of the entire 22-kD -zein subfamily. In the maize inbred BSSS53, this subfamily consists of 22 + 1 clustered genes, 7 of them appearing to be active (SONG et al. 2001 ), with >90% nucleotide sequence identity among their translated regions (Table 2). This level of homology makes possible the specific abrogation of the targeted proteins as if they were the product of one gene. Their closest relatives, the 19-kD zeins, with <50% identity (Table 2) and short regions of homology do not favor the accumulation of heterologous double-stranded RNA and are thus unaltered. Moreover, some of the 22-kD zeins are still synthesized in homozygous BSSS53(o2), indicating the ineffectiveness of o2 variants in achieving a complete repression of all 22-kD zein genes (SONG et al. 2001 ). Therefore, the simulation of miRNAs promises to be a particularly useful genetic tool in cases where redundancy of regulatory factors prevents typical genetic strategies, such as in many polyploid crop plants that differ from a typical diploid plant such as Arabidopsis.

The specificity of this approach is further illustrated by the synthesis and accumulation of protein from the other zein subfamilies. Surprisingly, there have been fewer compensatory effects from the lack of 22-kD zein proteins on the other storage proteins. Such changes were detected in maize opaque mutants (HUNTER et al. 2002 ) as well as in rice seeds engineered for low glutelin accumulation (MARUTA et al. 2001 ) and deemed important due to the strong interactions between zein proteins of different subfamilies during protein body formation (BAGGA et al. 1997 ; KIM et al. 2002 ).

Frequently, the introduction of an opaque mutation improves dramatically the nutritional value of maize seed protein, but ultimately deleterious effects hinder either the agronomical or the industrial quality of the final product. Another appealing possibility for enhancing the nutritional value of cereals is to quantitatively modify individual seed components by genetic engineering (recently reviewed by GALILI and HOFGEN 2002 ). This includes the modification or replacement of metabolic enzymes to change their activity or regulatory properties (KARCHI et al. 1994 ; SHAVER et al. 1996 ; CHIBA et al. 1999 ; ZHU et al. 2001 ) or the direct alteration of the amount or composition of structural proteins (BLECHL and ANDERSON 1996 ; MARUTA et al. 2001 ; LAI and MESSING 2002 ). Potentially, other proteins, which appear correlated with enhanced amino acid seed composition, could become attractive targets for molecular intervention (HABBEN et al. 1995 ). Here, we show that it is possible to produce intended changes in the chemical composition of the maize endosperm by stable disruption of one component. The amino acid changes in the lines with most intense silencing of the 22-kD zein (Table 1), although modest, are directly related to the relative abundance of the protein in untransformed maize and to Mendelian seed segregation of the transgenes analyzed. To our knowledge, this is one of the first reports of stable alteration of a cereal seed component by RNAi. In addition, we were successful in eliminating the abundant gene product derived from a large multigene family with at least two distinct chromosomal locations. It is unlikely that such a level of modification could be achieved in the present paradigm by regulation at the transcriptional level. In conclusion, considering the degree of gene silencing achieved, it would be advantageous to combine the transactivator hybrid system with constructs designed to generate transcript depletion by RNAi for functional dissection of seed metabolic pathways and biotechnological applications in maize.

	ACKNOWLEDGMENTS

We thank Barbara Miesak More for critical reading of the manuscript, Kathy Ward for technical assistance, Mike Peterczak for greenhouse and field management, Ian Moore for providing pOp and pKI-His, and Kan Wang and staff of the Plant Transformation Facility (Department of Agronomy, Iowa State University, Ames, Iowa) for providing high type II seeds and technical advice on maize transformation. This work has been supported by Department of Energy grant no. DE-FG05-95ER20194 to J.M.

Manuscript received March 10, 2003; Accepted for publication May 7, 2003.

	LITERATURE CITED

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

BAGGA, S., H. P. ADAMS, F. D. RODRIGUEZ, J. D. KEMP, and C. SENGUPTA-GOPALAN, 1997  Coexpression of the maize delta-zein and beta-zein genes results in stable accumulation of delta-zein in endoplasmic reticulum-derived protein bodies formed by beta-zein. Plant Cell 9:1683-1696.[Abstract]

BLECHL, A. E. and O. D. ANDERSON, 1996  Expression of a novel high-molecular-weight glutenin subunit gene in transgenic wheat. Nat. Biotechnol. 14:875-879.[Medline]

BOSHER, J. M. and M. LABOUESSE, 2000  RNA interference: genetic wand and genetic watchdog. Nat. Cell Biol. 2:E31-E36.[Medline]

CHIBA, Y., M. ISHIKAWA, F. KIJIMA, R. H. TYSON, and J. KIM et al., 1999  Evidence for autoregulation of cystathionine gamma-synthase mRNA stability in Arabidopsis. Science 286:1371-1374.[Abstract/Free Full Text]

CHRISTENSEN, A. H. and P. H. QUAIL, 1996  Ubiquitin promoter-based vectors for high-level expression of selectable and/or screenable marker genes in monocotyledonous plants. Transgenic Res. 5:213-218.[Medline]

CHUANG, C. F. and E. M. MEYEROWITZ, 2000  Specific and heritable genetic interference by double-stranded RNA in Arabidopsis thaliana.. Proc. Natl. Acad. Sci. USA 97:4985-4990.[Abstract/Free Full Text]

CICERI, P., S. CASTELLI, M. LAURIA, B. LAZZARI, and A. GENGA et al., 2000  Specific combinations of zein genes and genetic backgrounds influence the transcription of the heavy-chain zein genes in maize opaque-2 endosperms. Plant Physiol. 124:451-460.[Abstract/Free Full Text]

COLEMAN, C. E., A. M. CLONE, R. HIGGINS, M. A. LOPES, and B. A. LARKINS, 1997  Expression of a mutant alpha-zein creates the floury2 phenotype in transgenic maize. Proc. Natl. Acad. Sci. USA 94:7094-7097.[Abstract/Free Full Text]

CORDOVA, H., 2000 Quality protein maize: improve nutrition and livelihoods for the poor. CIMMYT Maize Res. Highl. 1999–2000: 27–31.

DAMERVAL, C. and M. LE GUILLOUX, 1998  Characterization of novel proteins affected by the o2 mutation and expressed during maize endosperm development. Mol. Gen. Genet. 257:354-361.[Medline]

DAS, O. P., M. CRUZ-ALVAREZ, S. CHAUDHURI, and J. MESSING, 1990  Molecular methods for genetic analysis of maize. Methods Mol. Cell. Biol. 1:213-222.

FRAME, B., H. ZHANG, S. COCCIOLONE, L. SIDORENKO, and C. DIETRICH et al., 2000  Production of transgenic maize from bombarded type II callus: effect of gold particle size and callus morphology on transformation efficiency. In Vitro Cell Dev. Biol. Plant 36:21-29.

GALILI, G. and R. HÖFGEN, 2002  Metabolic engineering of amino acids and storage proteins in plants. Metab. Eng. 4:3-11.[Medline]

GEETHA, K. B., C. R. LENDING, M. A. LOPES, J. C. WALLACE, and B. A. LARKINS, 1991  opaque-2 modifiers increase gamma-zein synthesis and alter its spatial distribution in maize endosperm. Plant Cell 3:1207-1219.[Abstract/Free Full Text]

HABBEN, J. E., G. L. MORO, B. G. HUNTER, B. R. HAMAKER, and B. A. LARKINS, 1995  Elongation factor 1 alpha concentration is highly correlated with the lysine content of maize endosperm. Proc. Natl. Acad. Sci. USA 92:8640-8644.[Abstract/Free Full Text]

HAMILTON, A., O. VOINNET, L. CHAPPELL, and D. BAULCOMBE, 2002  Two classes of short interfering RNA in RNA silencing. EMBO J. 21:4671-4679.[Medline]

HSIEH, J. and A. FIRE, 2000  Recognition and silencing of repeated DNA. Annu. Rev. Genet. 34:187-204.[Medline]

HUNTER, B. G., M. K. BEATTY, G. W. SINGLETARY, B. R. HAMAKER, and B. P. DILKES et al., 2002  Maize opaque endosperm mutations create extensive changes in patterns of gene expression. Plant Cell 14:2591-2612.[Abstract/Free Full Text]

HUTVÁGNER, G. and P. D. ZAMORE, 2002  RNAi: nature abhors a double-strand. Curr. Opin. Genet. Dev. 12:225-232.[Medline]

KARCHI, H., O. SHAUL, and G. GALILI, 1994  Lysine synthesis and catabolism are coordinately regulated during tobacco seed development. Proc. Natl. Acad. Sci. USA 91:2577-2581.[Abstract/Free Full Text]

KEMPER, E. L., G. C. NETO, F. PAPES, K. C. MORAES, and A. LEITE et al., 1999  The role of Opaque2 in the control of lysine-degrading activities in developing maize endosperm. Plant Cell 11:1981-1994.[Abstract/Free Full Text]

KIM, C. S., Y.-M. WOO, A. M. CLORE, R. J. BURNETT, and N. P. CARNEIRO et al., 2002  Zein protein interactions, rather than the asymmetric distribution of zein mRNAs on endoplasmic reticulum membranes, influence protein body formation in maize endosperm. Plant Cell 14:655-672.[Abstract/Free Full Text]

KODRZYCKI, R., R. S. BOSTON, and B. A. LARKINS, 1989  The opaque-2 mutation of maize differentially reduces zein gene transcription. Plant Cell 1:105-114.[Abstract/Free Full Text]

KRAMER, C., J. DIMAIO, G. K. CARSWELL, and R. D. SHILLITO, 1993  Selection of transformed protoplast-derived Zea mays colonies with phosphinothricin and a novel assay using the pH indicator chlorophenol red. Planta 190:454-458.

LAI, J. and J. MESSING, 2002  Increasing maize seed methionine by mRNA stability. Plant J. 30:395-402.[Medline]

LEVIN, J. Z., A. J. DE FRAMOND, A. TUTTLE, M. W. BAUER, and P. B. HEIFETZ, 2000  Methods of double-stranded RNA-mediated gene inactivation in Arabidopsis and their use to define an essential gene in methionine biosynthesis. Plant Mol. Biol. 44:759-775.[Medline]

LLAVE, C., Z. XIE, K. D. KASSCHAU, and J. C. CARRINGTON, 2002  Cleavage of Scarecrow-like mRNA targets directed by a class of Arabidopsis miRNA. Science 297:2053-2056.[Abstract/Free Full Text]

LOHMER, S., M. MADDALONI, M. MOTTO, N. DI FONZO, and H. HARTINGS et al., 1991  The maize regulatory locus Opaque-2 encodes a DNA-binding protein which activates the transcription of the b-32 gene. EMBO J. 3:617-624.

MADDALONI, M., G. DONINI, C. BALCONI, E. RIZZI, and P. GALLUSCI et al., 1996  The transcriptional activator Opaque-2 controls the expression of a cytosolic form of pyruvate orthophosphate dikinase-1 in maize endosperms. Mol. Gen. Genet. 250:647-654.[Medline]

MARUTA, Y., J. UEKI, H. SAITO, N. NITTA, and H. IMASEKI, 2001  Transgenic rice with reduced glutelin content by transformation with glutelin A antisense gene. Mol. Breed. 8:273-284.

MATZKE, M. A., A. J. MATZKE, G. J. PRUSS, and V. B. VANCE, 2001  RNA-based silencing strategies in plants. Curr. Opin. Genet. Dev. 11:221-227.[Medline]

MAZUR, B., E. KREBBERS, and S. TINGEY, 1999  Gene discovery and product development for grain quality traits. Science 285:372-375.[Abstract/Free Full Text]

MERTZ, E. T., L. S. BATES, and O. E. NELSON, 1964  Mutant gene that changes protein composition and increases lysine content of maize endosperm. Science 145:279-280.[Abstract/Free Full Text]

MOORE, I., L. GÄLWEILER, D. GROSSKOPF, J. SCHELL, and K. PALME, 1998  A transcription activation system for regulated gene expression in transgenic plants. Proc. Natl. Acad. Sci. USA 95:376-381.[Abstract/Free Full Text]

OUWERKERK, P. B. F., R. J. DE KAM, J. H. C. HOGE, and A. H. MEIJER, 2001  Glucocorticoid-inducible gene expression in rice. Planta 213:370-378.[Medline]

PARK, W., J. LI, R. SONG, J. MESSING, and X. CHEN, 2002  CARPEL FACTORY, a Dicer homolog, and HEN1, a novel protein, act in microRNA metabolism in Arabidopsis thaliana.. Curr. Biol. 12:1484-1495.[Medline]

PARRISH, S., J. FLEENOR, S. XU, C. MELLO, and A. FIRE, 2000  Functional anatomy of a dsRNA trigger: differential requirement for the two trigger strands in RNA interference. Mol. Cell 6:1077-1087.[Medline]

SCHMIDT, R. J., F. A. BURR, and B. BURR, 1987  Transposon tagging and molecular analysis of the maize regulatory locus opaque-2.. Science 238:960-963.[Abstract/Free Full Text]

SCHMIDT, R. J., M. KETUDAT, M. J. AUKERMAN, and G. HOSCHEK, 1992  Opaque-2 is a transcriptional activator that recognizes a specific target site in 22-kD zein genes. Plant Cell 4:689-700.[Abstract/Free Full Text]

SCHWEIZER, P., J. POKORNY, P. SCHULZE-LEFERT, and R. DUDLER, 2000  Double-stranded RNA interferes with gene function at the single-cell level in cereals. Plant J. 24:895-903.[Medline]

SHAVER, J. M., D. C. BITTEL, J. M. SELLNER, D. A. FRISCH, and D. A. SOMERS et al., 1996  Single-amino acid substitutions eliminate lysine inhibition of maize dihydrodipicolinate synthase. Proc. Natl. Acad. Sci. USA 93:1962-1966.[Abstract/Free Full Text]

SHIMIZU-SATO, S., E. HUQ, J. M. TEPPERMAN, and P. H. QUAIL, 2002  A light-switchable gene promoter system. Nat. Biotechnol. 20:1041-1044.[Medline]

SONG, R. and J. MESSING, 2002  Contiguous genomic DNA sequence comprising the 19-kD-zein gene family from Zea mays.. Plant Physiol. 130:1626-1635.[Abstract/Free Full Text]

SONG, R., V. LLACA, E. LINTON, and J. MESSING, 2001  Sequence, regulation, and evolution of the maize 22-kD zein gene family. Genome Res. 11:1817-1825.[Abstract/Free Full Text]

SONG, R., V. LLACA, and J. MESSING, 2002  Mosaic organization of orthologous sequences in grass genomes. Genome Res. 12:1549-1555.[Abstract/Free Full Text]

TAVERNARAKIS, N., S. H. WANG, M. DOROVKOV, A. RYAZANOV, and M. DRISCOLL, 2000  Heritable and inducible genetic interference by double-stranded RNA encoded by transgenes. Nat. Genet. 24:180-183.[Medline]

UEDA, T. and J. MESSING, 1993  Manipulation of amino acid balance in maize seeds. Genet. Eng. 15:109-130.

UEDA, T., W. WAVERCZAK, K. WARD, N. SHER, and M. KETUDAT et al., 1992  Mutations of the 22- and 27-kD zein promoters affect transactivation by the Opaque-2 protein. Plant Cell 4:701-709.[Abstract/Free Full Text]

WATERHOUSE, P. M., M. W. GRAHAM, and M. B. WANG, 1998  Virus resistance and gene silencing in plants can be induced by simultaneous expression of sense and antisense RNA. Proc. Natl. Acad. Sci. USA 95:13959-13964.[Abstract/Free Full Text]

WESLEY, S. V., C. A. HELLIWELL, N. A. SMITH, M. B. WANG, and D. T. ROUSE et al., 2001  Construct design for efficient, effective and high-throughput gene silencing in plants. Plant J. 27:581-590.[Medline]

WOO, Y.-M., D. W.-N. HU, B. A. LARKINS, and R. JUNG, 2001  Genomics analysis of genes expressed in maize endosperm identifies novel seed proteins and clarifies patterns of zein gene expression. Plant Cell 13:2297-2317.[Abstract/Free Full Text]

ZAMORE, P. D., 2002  Ancient pathways programmed by small RNAs. Science 296:1265-1269.[Abstract/Free Full Text]

ZHU, X., G. TANG, F. GRANIER, D. BOUCHEZ, and G. GALILI, 2001  A T-DNA insertion knockout of the bifunctional lysine-ketoglutarate reductase/saccharopine dehydrogenase gene elevates lysine levels in Arabidopsis seeds. Plant Physiol. 126:1539-1545.[Abstract/Free Full Text]

This article has been cited by other articles:

				S. Ufaz and G. Galili Improving the Content of Essential Amino Acids in Crop Plants: Goals and Opportunities Plant Physiology, July 1, 2008; 147(3): 954 - 961. [Full Text] [PDF]

				S. P. Moose and R. H. Mumm Molecular Plant Breeding as the Foundation for 21st Century Crop Improvement Plant Physiology, July 1, 2008; 147(3): 969 - 977. [Full Text] [PDF]

				M. Dafny-Yelin, S.-M. Chung, E. L. Frankman, and T. Tzfira pSAT RNA Interference Vectors: A Modular Series for Multiple Gene Down-Regulation in Plants Plant Physiology, December 1, 2007; 145(4): 1272 - 1281. [Abstract] [Full Text] [PDF]

				M. Hansen, M. Lange, C. Friis, G. Dionisio, P. B. Holm, and E. Vincze Antisense-mediated suppression of C-hordein biosynthesis in the barley grain results in correlated changes in the transcriptome, protein profile, and amino acid composition J. Exp. Bot., November 1, 2007; 58(14): 3987 - 3995. [Abstract] [Full Text] [PDF]

				T. Chung, C. S. Kim, H. N. Nguyen, R. B. Meeley, and B. A. Larkins The Maize Zmsmu2 Gene Encodes a Putative RNA-Splicing Factor That Affects Protein Synthesis and RNA Processing during Endosperm Development Plant Physiology, June 1, 2007; 144(2): 821 - 835. [Abstract] [Full Text] [PDF]

				K. McGinnis, N. Murphy, A. R. Carlson, A. Akula, C. Akula, H. Basinger, M. Carlson, P. Hermanson, N. Kovacevic, M. A. McGill, et al. Assessing the Efficiency of RNA Interference for Maize Functional Genomics Plant Physiology, April 1, 2007; 143(4): 1441 - 1451. [Abstract] [Full Text] [PDF]

				R. Song, G. Segal, and J. Messing Expression of the sorghum 10-member kafirin gene cluster in maize endosperm Nucleic Acids Res., December 29, 2004; 32(22): e189 - e189. [Abstract] [Full Text] [PDF]

				J. Lai, N. Dey, C.-S. Kim, A. K. Bharti, S. Rudd, K. F.X. Mayer, B. A. Larkins, P. Becraft, and J. Messing Characterization of the Maize Endosperm Transcriptome and Its Comparison to the Rice Genome Genome Res., October 1, 2004; 14(10a): 1932 - 1937. [Abstract] [Full Text] [PDF]