Ubiquitin-conjugating enzymes (E2) are known to show both spatial and temporal regulated expression, because their role in ubiquitination process, which can be associated to different molecular, cellular and physiological processes. With the goal of identifying a promoter and/or a regulatory region in cotton with new properties, we first analyzed Gossypium
spp E2 family in GenBank. Among the seven members found, one of them (GhGDRP85) seems to be a truncated version of the E2 homologues due to a premature stop codon. The stop codon is present in an uncommon region that seems to be an unspliced intron. Indeed, the GhGDRP85
mRNA possesses an internal 274-bp sequence that is spliced as an intron in an alternative Gossypium E2
transcript. This data was confirmed by qPCR using a reverse primer annealing in this region. The premature termination of GhGDRP85 protein prevents the conserved cysteine residue (Cys85) of the active site responsible for ubiquitin transfer to be encoded [33
]. The truncated E2 protein may act as a competitive inhibitor of the functional E2, and thereby down-regulate the ubiquitination pathway. Therefore, it seems that GhGDRP85 protein has evolved from E2 family, but as a result of alternative splicing that ultimately encodes a truncated E2-related protein. We observed intron junctions that could alternatively spliced over this 274-bp intron from the GhGDRP85
primary transcript under specific conditions, developmental stages and/or in certain tissue types. The alternative mRNA encodes the Cys85 along an E2 protein that shares very high sequence identity (from 2E-98 to 3E-109, 97-100%) with E2 family members, including UBC2, 8, 9, 10, 11, 28 and 30.
Thus, the GhGDRP85 transcript abundance in cotton tissues was evaluated to verify its regulatory sequence activity and its biotechnology usefulness. The GhGDRP85 transcript abundance was approximately 10-fold greater than the reference genes GhPP2A1 and GhUBQ14 in four analyzed tissues (leaf, stem, branch and flower). These results showed that the GhGDRP85 transcript is highly abundant in most cotton tissues, especially in roots, which the mRNA levels are 200-fold higher than the reference genes (Figure ).
To obtain further detail about the expression activity of the GhGDRP85 regulatory region, we isolated the corresponding upstream promoter (uceApro2) within the GhGDRP85 regulatory region (which included uceApro2 and a 5'UTR, named uceA1.7). The activity and the spatial gene expression driven by the constructs uceA1.7, uceApro2 and 35SdAMV nucleotide sequences fused to the GUS reporter were evaluated in the respective transgenic Arabidopsis plants.
Transgenic plants harboring uceApro2::GUS present basal levels of GUS activity in leaf, stem, flower bud and root, although GUS activity in root was approximately 4-fold higher when compared to other tissues (Figure ). The PlantCARE software did not identify any cis-acting element that would explain such high levels of expression in roots (Table ). Therefore, further analyses are still needed to locate root-specific cis-acting elements.
The remarkable difference in expression levels (Figures and ) driven by the uceA1.7 and uceApro2 constructs can be directly associated to the presence of the GhGDRP85 5'UTR. The presence of this region triggers an important quantitative effect, increasing from 11 to 15-fold expression in leaf, stem and flower bud tissues. In root the presence the 5'UTR result in an increase of only 7-folds, however it is also possible that the detection sensitivity of the GUS assay has reached its saturation point in the root tissues. On the other hand, some putative light-responsive cis-acting elements observed in 5'UTR could be responsible for the higher GUS expression in light-exposed tissues (motifs (p) and (r) in Figure and Table ).
In addition to the numerous regulatory elements present along the 5'UTR to regulate the transcription process, this sequence may also possess other elements that increase favorable effects on mRNA stability, processing, nucleus-cytoplasm translocation and translational apparatus assembly [13
], such as an intron just before the first exon that could characterize intron-mediated enhancement (IME) [30
]. Taken together, these findings suggest that the presence of the 5'UTR downstream of the uceApro2 promoter likely accounts for the observed increase in expression levels driven by uceA1.7. This phenomenon has already been described in previous reports of gene expression analyses in pea and conifers [28
The uceA1.7 sequence drives GUS expression to leaves and stems in comparable levels to the strong constitutive enhanced plant promoter CaMV35S. However, uceA1.7 drives higher GUS expression in roots (2-fold) and flowers (7-fold). 35SdAMV:GUS plants present a 10-fold decrease of GUS expression in flower bud, relative to leaf. It is noteworthy that GUS staining in leaf (Figure ) driven by uceA1.7 seems to be higher than that driven by CaMV35S. However this observation does not match to the quantitative fluorometric data presented on Figure . This kind of variance is commonly observed with GUS staining assays. Furthermore, the CaMV35S promoter, which has been widely used for plant expression systems, induces low and variable expression in floral organs [9
], rendering the production of plants expressing foreign genes on floral organs unpredictable. Due to the reported reduced gene expression driven by CaMV35S promoter in different plants upon nematode infection [36
], we intend to perform the analyses of uceA1.7 driven expression at nematode feeding sites.
The GhGDRP85 expression patterns in cotton plants are comparable to the GUS expression assays in transgenic A. thaliana plants. Despite of the similarity of expression levels measured in leaf, stem and flower tissues, the levels of GhGDRP85 transcript abundance in roots were much higher than the levels of GUS measured in the fluorometric assays. If there is no GUS detection limit threshold, we suggest that this difference might be due to the absence of cotton-specific trans-acting factors in Arabidopsis. Alternatively, the cloned uceA1.7 sequence might not contain cis-acting enhancer elements that are located far up- or downstream and present in the native GhGDRP85 gene regulatory region.