Ubiquitous miR159 repression of MYB33/65 in Arabidopsis rosettes is robust and is not perturbed by a wide range of stresses

Abstract

Background: The microR159 (miR159) - GAMYB pathway is conserved in higher plants, where GAMYB, expression promotes programmed cell death in seeds (aleurone) and anthers (tapetum). In cereals, restriction of GAMYB expression to seeds and anthers is mainly achieved transcriptionally, whereas in Arabidopsis this is achieved post-transcriptionally, as miR159 silences GAMYB (MYB33 and MYB65) in vegetative tissues, but not in seeds and anthers. However, we cannot rule out a role for miR159-MYB33/65 pathway in Arabidopsis vegetative tissues; a loss-of-function mir159 Arabidopsis mutant displays strong pleiotropic defects and numerous reports have documented changes in miR159 abundance during stress and hormone treatments. Hence, we have investigated the functional role of this pathway in vegetative tissues.

Results: It was found that the miR159-MYB33/65 pathway was ubiquitously present throughout rosette development. However, miR159 appears to continuously repress MYB33/MYB65 expression to levels that have no major impact on rosette development. Inducible inhibition of miR159 resulted in MYB33/65 de-repression and associated phenotypic defects, indicating that a potential role in vegetative development is only possible through MYB33 and MYB65 if miR159 levels decrease. However, miR159 silencing of MYB33/65 appeared extremely robust; no tested abiotic stress resulted in strong miR159 repression. Consistent with this, the stress responses of an Arabidopsis mutant lacking the miR159-MYB33/65 pathway were indistinguishable from wild-type. Moreover, expression of viral silencing suppressors, either via transgenesis or viral infection, was unable to prevent miR159 repression of MYB33/65, highlighting the robustness of miR159-mediated silencing.

Conclusions: Despite being ubiquitously present, molecular, genetic and physiological analysis failed to find a major functional role for the miR159-MYB33/65 pathway in Arabidopsis rosette development or stress response. Although it is likely that this pathway is important for a stress not tested here or in different plant species, our findings argue against the miR159-MYB33/65 pathway playing a major conserved role in general stress response. Finally, in light of the robustness of miR159-mediated repression of MYB33/65, it appears unlikely that low fold-level changes of miR159 abundance in response to stress would have any major physiological impact in Arabidopsis.

Keywords: Arabidopsis; GAMYB-like; Stress; Viral silencing suppressors; miR159.

Fig. 1

MiR159 constitutively silences MYB33/65 throughout rosette development. a–c Time-course transcript profiling of the miR159-MYB pathway in rosettes. The relative miRNA and mRNA levels were measured in rosettes approximately every 10 days throughout its development. The miR159 levels were normalized to sno101, the mRNA levels were normalized to CYCLOPHILIN. Values are the mean of three technical replicates with error bars representing the Standard Deviation (SD). Significant differences in values from the previous measurement are indicated with an _*, as determined by the Students T-test. d Time-course GUS-staining assay for rosettes of MIR159b:GUS and mMYB33:GUS transgenic lines. Staining was carried out on ten individual rosettes per time point, at ten-day interval during plant growth; only the first and last staining results are shown

Fig. 2

Morphological and molecular phenotypes via induced inhibition of miR159. a Application of 17-β-estradiol to 21-day-old XVE-MIM159 transformants induced leaf-curling (red circled). The representative picture was taken when plants were 35-day-old short-day grown plants. b qRT-PCR of MYB33, MYB65 and CP1 mRNA levels in 35-day-old XVE-MIM159 rosettes with either mock (M) or inducer (I) treatments. mRNA levels were normalized to CYCLOPHILIN. Values are the mean of three technical replicates with error bars representing the SD. Significant differences in values from the untreated are indicated with an _*

Fig. 3

Morphological and molecular analysis of stressed Arabidopsis miR159-MYB pathway mutants. a Phenotypic comparison of rosettes of Col, mir159ab.myb33.myb65 and mir159a plants treated with ABA, high temperature, high light, drought and cold. Plants were grown for two weeks at 21 °C and then subjected to three weeks of stress treatment (b) Taqman microRNA assays measuring miR159a levels in wild-type plants subjected to the above treatments. Levels were normalized to sno101. Values are the mean of three technical replicates with error bars representing the SD. Significant differences in values from the control are indicated with an _*

Fig. 4

Morphological and molecular analysis of low-temperature effect on the Arabidopsis miR159-MYB pathway mutants. a Phenotypic comparison of rosettes of Col, mir159a, mir159ab and mir159ab.myb33.myb65 plants stressed with low-temperature. Plants were grown for three weeks at 21 °C and then grown for eight weeks at 4 °C. b qRT-PCR analysis of MYB33, MYB65 and CP1 mRNA levels in the above rosettes. The mRNA levels were normalized to CYCLOPHILIN. Values are the mean of three technical replicates with error bars representing the SD. Significant differences in values from wild-type is indicated with an _*

Fig. 5

Constitutive expression of VSSs does not strongly perturb the miR159 silencing of MYB33/65. a Different phenotypes developed in 28-day-old 35S-P19 and 35S-P0 primary transformants with wild-type (Col) grown alongside as a control. b The representative classification of developmental defects among 35S-P0 primary transformants. Class I: wild-type-looking; Class II: mild reduction in rosette size and partially curled leaves; Class III: all leaves curled group; Class IV: severely stunted and all leaves curled. c qRT-PCR analysis of relative mRNA levels in the different classes. Significant difference in values from the control is indicated with an _*. d Comparison of P0 mRNA levels between 35S-P0(Col) and 35S-P0(myb33.myb65) with the same classified phenotypes. The RNA samples were extracted from 26-day-old plants. Col and myb33.myb65 were used as controls. P0 mRNA levels were normalized to UBIQUITIN (At4g05320), while MYB33/65 and CP1 were normalized to that of CYCLOPHILIN. Values are the mean of three technical replicates with error bars representing the SD. Significant differences between 35S-P0(Col) and 35S-P0(myb33,65) values are indicated with an _*

Fig. 6

TuMV infection does not appear to strongly perturb miR159 silencing of MYB33/65. a Morphological comparison between TuMV-infected Col and myb33.myb65 rosettes (21-day-post infection). Plants were inoculated with either Na₂PO₄ (mock) or TuMV. b qRT-PCR analysis of relative mRNA accumulations in rosettes with TuMV-symptoms being classified as either mild (M) or severe (S). All mRNA levels were normalized to CYCLOPHILIN. Error bars represent the SD of three technical replicates. c Analysis of mature miR159 levels in three TuMV-infected rosettes, T1-T3. The miR159 levels were normalized to sno101. Values are the mean of three technical replicates with error bars representing the SD