Final Report Summary - SRNAGRAD (Robust developmental patterns generated by opposing gradients of mobile small RNAs.) Small RNAs are key regulators of gene expression, and work from the Timmermans’ lab recently showed that gradients formed by mobile small RNAs contribute significantly to developmental patterning. Organ polarity is an excellent model to study this role of small RNAs. Formation of stable, precisely defined boundaries between two distinct cell fates is a fundamental feature of plant and animal development. Such cell fate boundaries often serve as a source of positional information to coordinate further patterning and growth of the tissue or organ. For instance, the formation of the adaxial-abaxial boundary in developing leaves is critical for the differentiation of distinct cell types on the upper and lower sides of the leaf: adaxial palisade parenchyma cells are optimized for light capture whilst abaxial spongy parenchyma cells are optimized for gas exchange. In addition, formation of an adaxial-abaxial boundary is a prerequisite of the flattened outgrowth of the leaf blade. Failure to form or maintain a stable adaxial-abaxial boundary results in severe defects in plant development.One of the main determinants of leaf polarity is the abaxial cell fate promoting AUXIN RESPONSE FACTOR3 (ARF3) transcription factor. The expression of ARF3 is limited to the abaxial domain by the tasiR-ARFs, which belong to a plant specific class of small RNAs, the ta-siRNAs. tasiR-ARF biogenesis is limited to the two adaxial-most cell layers of leaves by the localized expression of its biogenesis factors, but cell-to-cell movement of this small RNA generates a gradient that dissipates abaxially across the leaf. Surprisingly, instead of creating a reverse gradient of ARF3 expression, this gradient generates an on-off transition of ARF3 target gene expression consistent with a possible dose-dependent read-out of the small RNA gradient. TasiR-ARFs are not the only mobile, gradient forming small RNAs with an important role in development. A miRNA, miR166 is produced from precursors expressed in the abaxial epidermal layer, but mature miR166 forms an adaxially dissipating gradient in leaves and restricts the expression of its targets to the adaxial side of the leaf. Several open questions concerning the formation and function of small RNA gradients remain: how are small RNA gradients formed? How do they create target gene expression patterns and developmental phenotypes? What is the importance of small RNA gradients? Our project addresses these important questions.To monitor how miRNAs gradients generate developmental patterns, we created transgenic lines expressing an artificial miRNA targeting the abaxial specific ARF3 transcription factor (miRARF). We expressed miRARF from different promoters to modulate the position and direction of the artificial small RNA gradient and investigate how these parameters affect the domain, level and boundary of target gene expression. To eliminate the endogenous regulation of ARF3 by tasiR-ARF, we introduced the miRARF transgenes into the rdr6 tasiR-ARF biogenesis mutant background. The ARF3 promoter is active throughout the developing leaf, such that in the rdr6 background, we are able to monitor the effect of miRARF on target gene expression in all cells of the leaf. We used the following promoters to drive miRARF expression: the adaxial epidermis-specific AS2 promoter, the abaxial-specific FIL promoter and the MIR166A promoter specific for the abaxial epidermal layer. As ARF3 depletion, misexpression and overexpression cause developmental phenotypes, we are also able to analyze the phenotypic effects of miRARF when expressed from the different promoters.The rdr6 mutant phenotype results from ARF3 misexpression and is characterized by formation of elongated, rectangular and downward curling leaves, whereas loss of ARF3/4 function results in formation of strongly upward curled leaves with ectopic outgrowths on the lower surface. I have found that lines expressing miRARF from the AS2, FIL and MIR166A promoters display a range of phenotypes from mildly adaxialized to nearly wild-type or weakly abaxialized. The latter are consistent with a moderate or incomplete suppression of the rdr6 phenotype. To directly visualize the effect of miRARF on target gene expression, I crossed the rdr6 miRARF lines into a transgenic line expressing a translational GUS fusion of the ARF3 gene expressed from its own promoter (pARF3:ARF3-GUS) also in the rdr6 background. Investigation of pARF3:ARF3-GUS expression patterns by GUS staining of selected lines has shown that miRARF expressed from the adaxial AS2 promoter silences GUS expression in the adaxial domain, but does not affect GUS activity in the abaxial domain. In contrast, miRARF expressed from the abaxially-specific FIL promoter efficiently silences ARF3-GUS expression in nearly the entire leaf, suggesting that miRARF moves from its site of biogenesis and spreads throughout the leaf. Interestingly, miRARF expressed from the MIR166A promoter in the abaxial epidermis creates an on-off target expression boundary positioned within the abaxial domain, 1-2 cell layers beyond the adaxial-abaxial boundary. These results suggest that gradients of mobile miRNAs are able to create on-off target gene expression boundaries at various positions in the leaf. Considering what we know about the strengths of these promoters, the results also suggest that the level of the small RNA affects the position of this expression boundary, consistent with a dose-dependent read-out of the small RNA gradient (Figure 1 a-c). I also performed in situ hybridization experiments to visualize the expression pattern of miRARF in the pMIR166A:miRARFrdr6 line, and found that it forms a decreasing concentration gradient from the abaxial epidermis through 3-4 cell layers, indicating that we were able to create an inverted small RNA gradient as compared to wild-type (Figure 1 d). Next I investigated the role of small RNA gradients in plant development by analysing the expression pattern of adaxial and abaxial reporter transgenes in wild-type, rdr6 and pMIR166A:miRARFrdr6 lines to see how the formation of proper small RNA gradients affects the sharpness and robustness of the adaxial-abaxial cell fate boundary. Interestingly, I have found that while GUS reporters for the precursors of the gradient forming small RNAs themselves (TAS3:GUS and pMIR166A:GUS) maintain a robust, wild-type expression pattern even in lines where the small RNA gradients are perturbed, the pAS2:GUS reporter shows a severe loss of robustness in its expression pattern, suggesting that properly formed small RNA gradients are essential for robust adaxial-abaxial polarity. To study how this loss of robustness on the cell fate level affects developmental robustness, I compared the variability of leaf parameters between wild-type, and lines with perturbed small RNA gradients. I found that the latter were significantly more variable, further supporting the idea that properly formed small RNA gradients contribute to developmental robustness. In the last year of the project, we compared the expression pattern of the pFIL:eGFP abaxial cell fate reporter between wild-type and pMIR166A:miRARFrdr6 to analyse the straightness and sharpness of the adaxial-abaxial boundary in lines with perturbed small RNA gradients. Intriguingly, we found that the boundary of pFIL:eGFP expression shows a significant loss of both sharpness and straightness in the pMIR166A:miRARFrdr6 line, indicating that properly formed small RNA gradients are essential to the formation of sharp and straight developmental boundaries. Next we asked whether this patterning property is unique to tasiR-ARFs/miRARFs, by analyzing target expression in lines in which position and orientation of the other gradient forming small RNA of the leaf, miR166, has been perturbed. In wild-type plants, the miR166 target PHABULOSA (PHB) transcription factor is transcribed throughout the leaf, but its expression is restricted to the adaxial side by miR166. To test the effect of perturbed miR166 gradients on adaxial-abaxial boundary formation, we expressed the YFP-fused, miR166 resistant point mutant form of PHB from its own promoter throughout the leaf, and expressed miR166* carrying the complementary mutation from the AS2 promoter to invert the gradient. I crossed this line to adaxial and abaxial cell fate reporters to monitor how inverting the miR166 gradient affects cell fate determination and adaxial-abaxial boundary formation. Similarly to what we have seen in the lines with the inverted miRARF gradient, expression pattern of the pAS2:GUS reporter becomes highly variable in the pPHB:PHB-YFP;pAS2:mi166* background, further supporting the notion that properly formed small RNA gradients provide robustness to developmental patterning.To test whether the ability to create on-off boundaries of target gene expression is inherent to gradients of mobile small RNAs, or it depends on the wiring of the downstream gene network, we wanted to test whether an artificial small RNA targeting GFP exhibits the same patterning behaviour as endogenous small RNA – target pairs. By performing in situ hybridization, I showed that similarly to miRARF, miRGFP forms a gradient by mobility throughout the leaf. When interrogating GFP target expression by confocal microscopy in collaboration with Dr. Damianos Skopelitis, intriguingly we found that irrespective of direction of miRGFP expression, it creates an on-off transition between GFP expressing and non-expressing cells (Figure 2a). Thus we concluded that gradients of mobile small RNAs possess an inherent ability to create on-off target expression boundaries. To further investigate how miRGFP patterns GFP expression, we quantified signal intensity of both miRGFP and GFP. We found that the gradient of miRGFP creates a surprisingly sharp boundary between GFP expressing and non-expressing cells (Figure 2b).Finally, we asked whether small RNAs regulate their targets in a dose-dependent manner. We created transgenic lines that express miRGFP in an estradiol-inducible manner from the adaxial epidermis-specific AS2 promoter. We found that as we increased miRGFP levels by longer induction times, the boundary of GFP expression moved increasingly towards the abaxial side of the leaf, suggesting that small RNA gradients regulate their targets in a dose-dependent manner. final1-figure1.jpg final1-figure2.jpg
