Gene networks and chromatin and transcriptional regulation of the phaseolin promoter in Arabidopsis.

The complete lack of seed storage protein expression in vegetative tissues and robust expression during embryogenesis makes seed development an ideal system to study tissue-specific expression of genes. The promoter for the Phaseolin (phas) gene, which encodes the major seed storage protein in bean (Phaseolus vulgaris), is activated in two sequential steps: Phaseolus vulgaris ABI3-like factor (Pv-ALF)-dependent potentiation and abscisic acid-mediated activation. In this study, a heterologous in vivo Pv-ALF/phas-GUS (for ß-glucuronidase) expression system in transgenic Arabidopsis thaliana leaves was used in conjunction with the powerful RNA-Seq approach to capture transcriptional landscapes of phas promoter expression. Remarkably, expression of over 1300 genes from 11 functional categories coincided with changes in the transcriptional status of the phas promoter. Gene network analysis of induced genes and artificial microRNA-mediated loss-of-function genetic assays identified transcriptional regulators RINGLET 2 (RLT2) and AINTEGUMENTA-LIKE 5 (AIL5) as being essential for phas transcription. Pv-ALF binding to the RLT2 and AIL5 promoter regions was confirmed by electrophoretic mobility shift assay. RLT2 and AIL5 knockdown lines displayed reduced expression of several endogenous seed genes, suggesting that these factors are involved in activation of endogenous Arabidopsis seed storage gene expression. Overall, the identification of these key factors involved in phas activation provides important insight into the two-step transcriptional regulation of seed-specific gene expression.

Plant SET domain-containing proteins: structure, function and regulation.

Modification of the histone proteins that form the core around which chromosomal DNA is looped has profound epigenetic effects on the accessibility of the associated DNA for transcription, replication and repair. The SET domain is now recognized as generally having methyltransferase activity targeted to specific lysine residues of histone H3 or H4. There is considerable sequence conservation within the SET domain and within its flanking regions. Previous reviews have shown that SET proteins from Arabidopsis and maize fall into five classes according to their sequence and domain architectures. These classes generally reflect specificity for a particular substrate. SET proteins from rice were found to fall into similar groupings, strengthening the merit of the approach taken. Two additional classes, VI and VII, were established that include proteins with truncated/interrupted SET domains. Diverse mechanisms are involved in shaping the function and regulation of SET proteins. These include protein-protein interactions through both intra- and inter-molecular associations that are important in plant developmental processes, such as flowering time control and embryogenesis. Alternative splicing that can result in the generation of two to several different transcript isoforms is now known to be widespread. An exciting and tantalizing question is whether, or how, this alternative splicing affects gene function. For example, it is conceivable that one isoform may debilitate methyltransferase function whereas the other may enhance it, providing an opportunity for differential regulation. The review concludes with the speculation that modulation of SET protein function is mediated by antisense or sense-antisense RNA.

MAK, a computational tool kit for automated MITE analysis.

Miniature inverted repeat transposable elements (MITEs) are ubiquitous and numerous in higher eukaryotic genomes. Analysis of MITE families is laborious and time consuming, especially when multiple MITE families are involved in the study. Based on the structural characteristics of MITEs and genetic principles for transposable elements (TEs), we have developed a computational tool kit named MITE analysis kit (MAK) to automate the processes (http://perl.idmb.tamu.edu/mak.htm). In addition to its ability to routinely retrieve family member sequences and to report the positions of these elements relative to the closest neighboring genes, MAK is a powerful tool for revealing anchor elements that link MITE families to known transposable element families. Implementation of the MAK is described, as are genetic principles and algorithms used in its derivation. Test runs of the programs for several MITE families yielded anchor sequences that retain TIRs and coding regions reminiscent of transposases. These anchor sequences are consistent with previously reported putative autonomous elements for these MITE families. Furthermore, analysis of two MITE families with no known links to any transposon family revealed two novel transposon families, namely Math and Kid, belonging to the IS5/Harbinger/PIF superfamily.

High rooting frequency and functional analysis of GUS and GFP expression in transgenic Medicago truncatula A17.

• An effective transformation method is described for Medicago truncatula A17, verifying its suitability as a model legume for functional genomics. • Media and culture methods are detailed that yielded an average frequency of 35% for recovery of transgenic shoots from cotyledonary node explants and 39% for root induction and regeneration of entire plants from 419 phosphinothricin-resistant shoots. • Fertile plants transgenic for both 35S-GFP and phas-GUS were obtained in five of eight independent experiments. The presence and stable inheritance of transgenes was confirmed by GFP or GUS expression and by genomic DNA blots. GFP expression driven by the normally constitutive CaMV 35S promoter diminished as the leaves matured. Although GUS was very strongly and uniformly expressed in seed cotyledons of most lines, one line exhibited an aberrant, patchy pattern. Additionally, weak GUS expression was evident in leaf veins from the normally stringently spatially regulated phas promoter. • Stably transformed, fertile, M. truncatula A17 plants were generated. The unconventional expression patterns for 35S-GFP and phas-GUS expression obtained in some transformants suggest the occurrence of novel epigenetic events.

PvALF and FUS3 activate expression from the phaseolin promoter by different mechanisms.

Transcription from the phaseolin (phas) promoter requires two major events: chromatin remodeling, mediated by PvALF, a B3 domain factor, and activation by an ABA-induced signal transduction cascade. Expression from phas is normally seed-specific, but high levels of expression in leaves can be obtained by ectopic expression of PvALF. Here, the system was used to compare the ability of PvALF and Arabidopsis FUS3, another B3 domain transcription factor that lacks the N-terminal activation and B1 domain present in PvALF, to activate phas expression in vegetative tissues. When compared to PvALF-mediated phas activation in the presence of ABA, a delay in phas activation was observed in the presence of both FUS3 and ABA in vegetative tissue. Significant differences in histone modifications at the phas promoter were mediated by FUS3 and PvALF, suggesting that they function through different epigenetic mechanisms. The relationship between PvALF and ABI5, a bZIP transcription factor, in mediating phas expression was also evaluated. Interestingly, over-expression of ABI5 rendered phas expression ABA-independent in the presence of PvALF. Changes in phas activity in different regions within seed embryos were demonstrated using abi5 mutants. Our results show that (1) redundant factors, such as PvALF and FUS3, employ different mechanisms to regulate their common target gene (phas); (2) ABI5, and possibly other redundant bZIP factors, act downstream of ABA in modulating phas expression in the presence of PvALF.

HideNseek, a post-genome approach to locate transgenes exemplified in Arabidopsis thaliana.

SUMMARY: Determination of transgene location is essential for investigating the effects of position on transgene expression levels and facilitates cloning of the resident gene affected by insertion. Currently used PCR-based approaches for determination of transgene location are relatively complicated and often fail when the transgene is duplicated, rearranged or fragmented. HideNseek is a new bioinformatics tool that allows computation of transgene locations, provided that a suitable genomic restriction enzyme digestion profile is available. Since the new approach is not based on the terminal sequences of the transgene insert, it is less sensitive to transgene duplication, rearrangement or fragmentation. HideNseek has been tested experimentally and by in silico simulation. The experimental example provided here shows that this simple approach is feasible, permitting rapid location of transgenes with little bench work. AVAILABILITY: available on request from the authors. SUPPLEMENTARY DATA: HideNseek input and output examples, experimental procedures and figures showing experimental results are provided as supplementary files: Supplementary material 1, 2, 3 and Supplementary figures (Figs 1 and 2), respectively. Supplementary data is available at Bioinformatics online.

Ordered histone modifications are associated with transcriptional poising and activation of the phaseolin promoter.

The phaseolin (phas) promoter drives copious production of transcripts encoding the protein phaseolin during seed embryogenesis but is silent in vegetative tissues, in which a nucleosome is positioned over its three-phased TATA boxes. Transition from the inactive state in transgenic Arabidopsis thaliana leaves was accomplished by ectopic expression of the transcription factor Phaseolus vulgaris ABI3-like factor (ALF) and application of abscisic acid (ABA). Placement of hemagglutinin-tagged ALF expression under the control of an estradiol-inducible promoter permitted chromatin immunoprecipitation analysis of chronological changes in histone modifications, notably increased acetylation of H3-K9 and H4-K12, as phas chromatin was remodeled (potentiated). A different array of changes, including acetylation of H3-K14 and methylation of H3-K4, was found to be associated with ABA-mediated activation. Thus, temporal separation of phas potentiation from activation revealed that histone H3 and H4 Lys residues are not globally hyperacetylated during phas expression. Whereas decreases in histone H3 and H4 levels were detected during ALF-mediated remodeling, slight increases occurred after ABA-mediated activation, suggesting the restoration of histone-phas interactions or the replacement of histones in the phas chromatin. The observed histone modifications provide insight into factors involved in the euchromatinization and activation of a plant gene and expand the evidence for histone code conservation among eukaryotes.

The dark side of green fluorescent protein.

Here, severe interference of chlorophyll with green fluorescent protein (GFP) fluorescence is described for medicago (Medicago truncatula), rice (Oryza sativa) and arabidopsis (Arabidopsis thaliana). This interference disrupts the proportional relationship between GFP content and fluorescence that is intrinsic to its use as a quantitative reporter. The involvement of chlorophyll in the loss of GFP fluorescence with leaf age was shown in vivo, by the removal of chlorophyll through etiolation or by ethanol extraction, and in vitro, by titration of a GFP solution with chlorophyll solutions of various concentrations. A substantial decrease in fluorescence in early development of medicago and rice leaves correlated with chlorophyll accumulation. In all three species tested, removal of chlorophyll yielded up to a 10-fold increase in fluorescence. Loss of GFP fluorescence in vitro was 4-fold greater for chlorophyll b than for chlorophyll a. Differences exist between plant species for the discrepancy between apparent GFP fluorescence and its actual level in green tissues. Substantial errors in estimating promoter activity from GFP fluorescence can occur if pigment interference is not considered.

A two-edged role for the transposable element Kiddo in the rice ubiquitin2 promoter.

Miniature inverted repeat transposable elements (MITEs) are thought to be a driving force for genome evolution. Although numerous MITEs are found associated with genes, little is known about their function in gene regulation. Whereas the rice ubiquitin2 (rubq2) promoter in rice (Oryza sativa) line IR24 contains two nested MITEs (Kiddo and MDM1), that in line T309 has lost Kiddo, providing an opportunity to understand the role of MITEs in promoter function. No difference in endogenous rubq2 transcript levels between T309 and IR24 was evident using RT-PCR. However, promoter analysis using both transient and stably transformed calli revealed that Kiddo contributed some 20% of the total expression. Bisulfite genomic sequencing of the rubq2 promoters revealed specific DNA methylation at both symmetric and asymmetric cytosine residues on the MITE sequences, possibly induced by low levels of homologous transcripts. When methylation of the MITEs was blocked by 5-azacytidine treatment, a threefold increase in the endogenous rubq2 transcript level was detected in IR24 compared with that in T309. Together with the observed MITE methylation pattern, the detection of low levels of transcripts, but not small RNAs, corresponding to Kiddo and MDM1 suggested that RNA-dependent DNA methylation is induced by MITE transcripts. We conclude that, although Kiddo enhances transcription from the rubq2 promoter, this effect is mitigated by sequence-specific epigenetic modification.

Assessment of penetrance and expressivity of RNAi-mediated silencing of the Arabidopsis phytoene desaturase gene.

RNA interference (RNAi) is of great value in plant functional genomics. However, the absence of RNAi phenotypes and the lack of uniform level of RNAi silencing has complicated gene identification. Here, the penetrance and expressivity of RNAi-mediated silencing of the phytoene desaturase (PDS) gene in Arabidopsis thaliana were examined quantitatively to provide a reference for the likely severity and distribution of silencing effects. Arabidopsis plants were transformed with an RNAi construct targeting PDS. Transgenic plants were examined for frequency of RNAi-mediated silencing and various silencing phenotypes. mRNA depletion level and RNAi expressivity were assayed by relative reverse transcription polymerase chain reaction (RT-PCR). High penetrance and variable expressivity of RNAi were demonstrated. An inverse correlation between PDS mRNA level and RNAi phenotype was seen. No direct relationship between copy number for the RNAi-generating transgene and phenotype was evident. Decreased RNAi penetrance in T2 plants was observed. It is suggested that variability in RNAi expressivity and postmeiotic decrease in RNAi penetrance constitute barriers for high throughput plant gene characterization.

Organization, not duplication, triggers silencing in a complex transgene locus in rice.

Despite the presence in nature of many functional gene families that contain several to many highly similar sequences, the presence of identical DNA sequence repeats is widely thought to predispose transgene inserts to homology dependent gene silencing (HDGS). The induction of transcriptional gene silencing (TGS) by RNAs homologous to promoter sequences has been reported recently in Arabidopsis and humans. However, mechanisms for TGS have not been studied in detail for rice, the most widely cultivated crop plant. Taking advantage of a well-characterized homozygous silenced transgenic rice line (siJKA), supertransformation was performed with a binary vector bearing mUbi1 and 35S promoter sequences identical to those in the resident transgenes. Analysis of the incoming and resident transgenes in the supertransformants revealed that the incoming mUbi1 transgene promoter was not silenced whereas the incoming 35S transgene promoter was silenced. That the resident silenced mUbi1-bar was not reactivated in these experiments as a result of passage through tissue culture and regeneration was established by the finding that regenerants from siJKA immature embryos were all silenced for mUbi1-bar. In a parallel experiment, when wild type rice calli were transformed with the same binary vector, neither of the incoming transgene promoters was silenced. Following 5-azacytidine (5-azaC) treatment of siJKA, aberrant RNA species corresponding to the 35S promoter, but not to the mUbi1 promoter, were detected. Nevertheless, no 21-25 nt RNAs corresponding to the 35S promoter sequence were detected. These results, together with detailed analyses of the progenies from the primary transformants and supertransformants, revealed that HDGS of the resident silenced locus was caused not by simple transgene duplication, but by aberrant transcripts derived from rearranged promoters present in siJKA. Practical consequences of this study include a justification for the use of multiple copies of a given promoter for transformation without inducing silencing, provided that their genomic integration does not result in aberrant transcription of the promoters.

MDM-1 and MDM-2: two mutator-derived MITE families in rice.

Numerous miniature inverted repeat transposable elements (MITEs) are present in the rice genome but their transposition mechanisms are unknown. In this report, we present evidence that two novel MITE families may have arisen from Mutator-related transposable elements and thus may use a transposition mechanism similar to that of Mutator elements. Two families of novel MITEs, namely, MDM-1 and MDM-2, were identified by searching for MITEs nested with Kiddo, a previously identified MITE family. MDM-1 and MDM-2 bear hallmarks of Mutator elements, such as long terminal inverted repeats (LTIRs), 9-bp target-site duplications (TSDs), and putative transposase binding sites. Strikingly, the MDM-1 family has a 9-bp terminus identical to that of a rice Mutator-like element ( MULE-9) and the MDM-2 family has an 8-bp terminus identical to that of the maize autonomous Mutator element MuDR. A putative transposase homologous to MURA protein is identified for the MDM-2 family. Thus, these two novel MITE families, with a total copy number of several hundred in rice, are designated Mutator-derived MITEs ( MDMs). Interestingly, sequence decay analysis of MDM families revealed a number of insertion site duplications (ISDs) in the alignment gaps, and widespread historical nesting events are proposed to account for the existence of these ISDs. In addition to its value for discovering new MITEs, the nesting analysis approach used in this study simultaneously identifies MITE insertion polymorphisms.

Plant chromatin: development and gene control.

It is increasingly clear that chromatin is not just a device for packing DNA within the nucleus but also a dynamic material that changes as cellular environments alter. The precise control of chromatin modification in response to developmental and environmental cues determines the correct spatial and temporal expression of genes. Here, we review exciting discoveries that reveal chromatin participation in many facets of plant development. These include: chromatin modification from embryonic and meristematic development to flowering and seed formation, the involvement of DNA methylation and chromatin in controlling invasive DNA and in maintenance of epigenetic states, and the function of chromatin modifying and remodeling complexes such as SWI/SNF and histone acetylases and deacetylases in gene control. Given the role chromatin structure plays in every facet of plant development, chromatin research will undoubtedly be integral in both basic and applied plant biology.

Module-specific regulation of the beta-phaseolin promoter during embryogenesis.

The phas promoter displays stringent spatial regulation, being very highly expressed during embryogenesis and completely silent during all phases of vegetative development in bean, Phaseolus vulgaris. This pattern is maintained in transgenic tobacco and, as shown here, Arabidopsis. Dimethyl sulphate in vivo footprinting analyses revealed that over 20 cis-elements within the proximal 295 bp of the phas promoter are protected by factor binding in seed tissues whereas none are bound in leaves. The hypothesis that this complex profile represents a summation of several module (cotyledon, hypocotyl, and radicle)-specific factor-DNA interactions has been explored by the incorporation of site-directed substitution mutations into 10 locations within the -295phas promoter. Only 2.6% of -295phas promoter activity remained after mutation of the G-box; the CCAAAT box, the E-box and the RY elements were also found to mediate high levels of expression in embryos. Whereas the CACA element has dual positive and negative regulatory roles, the vicilin box was identified as a strong negative regulatory element. The proximal (-70 to -64) RY motif was found to bestow expression in the hypocotyl while all the RY elements contribute to expression in cotyledons but not to vascular tissue expression during embryogenesis. RY elements at positions -277 to -271, -260 to -254, and -237 to -231 were found to orchestrate radicle-specific repression. The G-box appears to be the functional abscisic acid responsive element and the E-site may be a coupling element. The results substantiate the concept that autarkical cis-element functions generate modular patterning during embryogenesis. They also reflect the existence of both redundancy and hierarchy in cis-element interactions. Importantly, the virtually identical expression patterns observed for the two distantly related plants studied argue strongly for the generality of function for the observed factor-element interactions.

Sequence and spacing of TATA box elements are critical for accurate initiation from the beta-phaseolin promoter.

The beta-phaseolin (phas) gene, which encodes one of the major seed storage proteins of P. vulgaris, is tightly regulated at the transcription level resulting in strict tissue-specific and spatial expression during embryonic development. The phas proximal promoter contains a complex arrangement of core promoter elements including three TATA boxes as well as several putative initiator elements. To delineate the respective contributions of the core promoter elements to transcription initiation we have performed site-directed mutagenesis of the phas promoter. In vivo expression studies were performed on transgenic Arabidopsis harboring phas promoter mutants driving expression of the beta-glucuronidase (gus) reporter gene. Quantitative assessment of GUS activity in seeds bearing the promoter mutants indicated that both sequence and spacing of the TATA elements influenced the efficiency of transcription. Substitution, insertion or deletion mutations had no effect on histochemical staining patterns indicating that strict spacing requirements are not essential for correct spatial expression of phas during embryogenesis. Further evaluation of the phas promoter by in vitro transcription analysis revealed the presence of multiple TATA-dependent transcription initiation start sites. The distance between TATA elements and transcription start sites was maintained in insertion and deletion mutants through the creation of novel initiation sites, indicating that positioning of the TATA elements rather than DNA sequence was the primary determinant of start site location. We conclude that, while dispensable for proper spatial distribution, the complex architecture of the phas promoter is required to ensure high levels of accurate phas transcription initiation in the developing embryo.

S phase progression is required for transcriptional activation of the beta-phaseolin promoter.

Elucidating the mechanisms by which the transcription machinery accesses promoters in their chromatin environment is a fundamental aspect of understanding gene regulation. The phas promoter is normally constrained by a rotationally and translationally positioned nucleosome over its TATA region except during embryogenesis when it is potentiated by the presence of Phaseolus vulgaris ABI3-like factor (PvALF), a plant-specific transcription factor, and activated by an abscisic acid (ABA)-induced signal transduction cascade. Ectopic expression of PvALF and the supply of ABA in transgenic tobacco or Arabidopsis leaves can activate expression from phas. We confirmed by [3H]thymidine incorporation that active DNA replication occurred concomitant with the presence of PvALF and ABA. Arrest of DNA synthesis or S phase progression by infiltration of the leaves with replication inhibitors (hydroxyurea, roscovitine, mimosine) strongly inhibited transcriptional activation, especially the ABA-mediated activation step. Similarly, activation of endogenous Arabidopsis MAT and LEA genes in leaf tissue by the presence of ABA and ectopically expressed PvALF was inhibited by DNA replication arrest. No change in transcript levels on the arrest of replication was detected for abi1, abi2, and era1, negative regulators of the ABA signal transduction cascade or for cell cycle components ick1 and aip3. However, a reduction in transcript accumulation for the crucial ABA signaling effector, abi5, occurred upon DNA replication arrest (probably reflected in the decrease in MAT and LEA gene expression). Contrary to the conventional view that ABA inhibits DNA replication, our findings show that ABA acts in concert with S phase progression to activate gene expression.

Characterization of two rice DNA methyltransferase genes and RNAi-mediated reactivation of a silenced transgene in rice callus.

Two genomic clones ( OsMET1-1, AF 462029 and OsMET1-2, TPA BK001405), each encoding a cytosine-5 DNA methyltransferase (MTase), were isolated from rice ( Oryza sativa L.) BAC libraries. OsMET1-1 has an open reading frame of 4,566 nucleotides with 12 exons and 11 introns while OsMET1-2 has an open reading frame of 4,491 nucleotides with 11 exons and 10 introns. Although OsMET1-1 and OsMET1-2 have high sequence similarity overall, they share only 24% identity in exon 1, and intron 3 of OsMET1-1 is absent from OsMET1-2. As for other eukaryotic DNA MTases of the Dnmt1/MET l class, the derived amino acid sequences of OsMET1-1 and OsMET1-2 suggest that they are comprised of two-thirds regulatory domain and one-third catalytic domain. Most functional domains identified for other MTases were present in the rice MET1 sequences. Amino acid sequence comparison indicated high similarity (56-75% identity) of rice MET1 proteins to other plant MET1 sequences but limited similarity (approx. 24% identity) to animal Dnmt1 proteins. Genomic blot and database analysis indicated the presence of a single copy of OsMET1-1 (on chromosome 3) and single copy of OsMET1-2 (on chromosome 7). Ribonuclease protection assays revealed expression of both OsMET1-1 and OsMET1-2 in highly dividing cells, but the steady-state level of OsMET1-2 was 7- to 12-fold higher than that for OsMET1-1 in callus, root and inflorescence. The functional involvement of the rice DNA MTases in gene silencing was investigated using an RNAi strategy. Inverted repeat constructs of either the N- or C-terminal regions of OsMET1-1 were supertransformed into calli derived from a rice line bearing a silenced 35S-uidA-nos transgene. Restoration of uidA expression in the bombarded calli was consistent with the inactivation of maintenance methylation and with previous evidence for the involvement of methylation in silencing of this line.

Interaction of PvALF and VP1 B3 domains with the beta -phaseolin promoter.

The phas promoter is potently transcribed during embryogenesis but in vegetative tissues it is completely silenced by a rotationally positioned nucleosome. Ectopic expression in leaves of PvALF, a seed-specific transcription factor belonging to the plant-exclusive B3 domain-containing VP1/ABI3 family, leads to chromatin remodeling of the phas promoter, permitting transcriptional activation by the growth regulator abscisic acid (ABA). Specific interaction with RY elements present in 40-42 bp oligonucleotide probes has been shown in vitro for Arabidopsis ABI3 and the isolated B3 domain of maize VP1. Here, both in vivo and in vitro approaches were used to show physical interaction of the B3 domain of VP1 or PvALF to RY elements in the native phas promoter. In electrophoretic mobility shift assays, small changes in B3 domain concentration differentiated between RY element-specific and sequence non-specific DNA binding. Increased affinity of the PvALF B3 domain to RY elements was observed in the presence of histones and other basic proteins, possibly reflecting the ability of this B3 factor to interact with the phas promoter in its nucleosomal configuration.

The 5' UTR negatively regulates quantitative and spatial expression from the ABI3 promoter.

The involvement of transcription factors Arabidopsis abscisic acid-insensitive3 (ABI3), maize viviparous1 (VP1) and Phaseolus vulgaris ABI3-like factor (PvALF) in the spatial control of storage protein gene expression is well established. However, little insight exists as to how they are themselves regulated. To address this, a 5.15 kb ABI3 upstream sequence including a 4.6 kb full-length promoter and 519 bp of 5'-untranslated region (UTR) was used to drive either beta-glucuronidase (GUS) or green fluorescent protein (GFP) expression in Arabidopsis. Expression from the full-length (- 4630/ + 519ABI3 ) and various 5'-truncated promoters was detected during embryogenesis in all lines, except those transgenic for promoter elements shorter than 364 bp. Two upstream activating regions, -3600 to -2033 and -2033 to -882, enhanced GUS expression in seeds. The -882 to -364 region was sufficient to confer seed-specific expression of GUS when fused to a - 64/ + 6CaMV 35S minimal promoter. Expression from the ABI3 promoter constructs was seed-specific, except in the presence of exogenous abscisic acid (ABA) (>0.3 microM), when GUS expression was detected in seedling roots. Excision of a 405 bp region containing three upstream open reading frames (uORFs) from the 5'-UTR dramatically increased GUS expression and debilitated constraint of reporter expression in roots. Negative regulation of ABI3 expression by the 5'-UTR may involve a post-transcriptional mechanism analogous to that of tumor suppressor genes which also bear long, uORF-containing, 5'-UTRs, or through interactions with RNA-binding proteins.