Phenotypic plasticity of adventitious rooting in Arabidopsis is controlled by complex regulation of AUXIN RESPONSE FACTOR transcripts and microRNA abundance

Abstract

The development of shoot-borne roots, or adventitious roots, is indispensable for mass propagation of elite genotypes. It is a complex genetic trait with a high phenotypic plasticity due to multiple endogenous and environmental regulatory factors. We demonstrate here that a subtle balance of activator and repressor AUXIN RESPONSE FACTOR (ARF) transcripts controls adventitious root initiation. Moreover, microRNA activity appears to be required for fine-tuning of this process. Thus, ARF17, a target of miR160, is a negative regulator, and ARF6 and ARF8, targets of miR167, are positive regulators of adventitious rooting. The three ARFs display overlapping expression domains, interact genetically, and regulate each other's expression at both transcriptional and posttranscriptional levels by modulating miR160 and miR167 availability. This complex regulatory network includes an unexpected feedback regulation of microRNA homeostasis by direct and nondirect target transcription factors. These results provide evidence of microRNA control of phenotypic variability and are a significant step forward in understanding the molecular mechanisms regulating adventitious rooting.

Figure 1.

ARF6, ARF8, and ARF17 Regulate Adventitious Root Development. (A) Dark grown wild-type seedlings. Hypocotyls have reached 6 mm. Bar = 5 mm. (B) Wild-type seedlings 7 d after transfer to the light. Arrows indicate adventitious roots on the hypocotyls. Bar = 5 mm. (C) Adventitious roots were counted in seedlings that were first etiolated in the dark, until their hypocotyls were 6 mm long, and then transferred to the light for 7 d. Data from three independent biological replicates, each of at least 30 seedlings, were pooled and averaged. Error bars indicate ± se. A one-way analysis of variance (ANOVA) combined with the Tukey's multiple comparison test indicated that the values marked with one asterisk were significantly different from wild-type values and those marked with two asterisks were significantly different from values obtained from single mutants or ARF17-OX lines (P < 0.01; n > 90). (D) The lateral roots of the same seedlings were counted, and the lengths of their main roots were measured. No significant differences in either the length or number of lateral roots were observed between any of the lines (see Supplemental Figures 1A and 1B online). For simplicity, we show the mean lateral root density, expressed as the number of lateral roots divided by the length of the main root, of at least 30 plants of each line described in Methods. The experiments were repeated three times. Error bars indicate ± se. A one-way ANOVA combined with the Tukey's multiple comparison test indicated that the values marked with one asterisk were not significantly different from wild-type (Columbia-0 [Col-0]) values (P = 0.99; n > 45) and those marked with two asterisks were not significantly different from wild-type (Wassilewskija [Ws]) values (P = 0.88; n > 45).

Figure 2.

Expression Patterns of MIR160c, MIR167a, MIR167b, MIR167c, MIR167d, ARF6, ARF8, and ARF17 during the Early Stages of Adventitious Root Formation. (A) to (E) GUS staining of promMIR160c:GUS, promMIR167a:GUS, promMIR167b:GUS, promMIR167c:GUS, and promMIR167d:GUS (arranged from left to right in each panel) in seedlings grown in the dark until their hypocotyls were 6 mm long (A), after an additional 48 h (B) and 72 h (C) in the dark or after transfer to the light for 48 h (D) and 72 h (E). (F) to (J) GUS staining of promARF6:GUS, promARF8:GUS, and promARF17:GUS (arranged from left to right in each panel) in seedlings grown in the dark until their hypocotyls were 6 mm long (F), after an additional 48 h (G) and 72 h (H) in the dark, or 48 h (I) and 72 h (J) after their transfer to the light. Bars = 5 mm in (A) to (J). (K) Close-up image of promARF6:GUS hypocotyl from seedling shown in (J). (L) to (N) Close-up up images from the same seedlings as in (J); young adventitious root primordia of promARF6:GUS (L), promARF8:GUS (M), and promARF17:GUS (N) plants after 72 h in the light. Bars = 0.5 mm in (K) and 50 μm in (L) to (N). (O) Quantification by real-time RT-PCR of the steady state level of miRNA species miR160abc, miR167ab, miR167c, and miR167d in the different organs (cotyledons, apical meristem, hypocotyls, and root) of wild-type seedlings etiolated and then transferred to the light for 72 h as in (E) and (J). (P) to (R) Quantification by real-time RT-PCR of ARF6, ARF8, and ARF17 transcripts in roots (P), hypocotyls (Q), and cotyledons (R) of wild-type seedlings etiolated as in (F) (black bars) and after transfer to the light for 72 h as in (J) (white bars). T, total transcript; UC, uncleaved transcript. (S) Confirmation by real-time RT-PCR of ARF6, ARF8, and ARF17 transcripts in the apical meristem region from seedlings transferred to the light for 72 h as in (J). (O) to (S) Expression values are expressed relative to the expression level of APT1 used as a reference gene as described in Methods. Error bars indicate ± se obtained from three independent RT-PCR experiments.

Figure 3.

Quantification by Real-Time RT-PCR of ARF6, ARF8, and ARF17 Transcripts in Hypocotyls of ARF Mutant Lines Reveals Regulatory Loops. (A) and (B) Comparison of ARF6, ARF8, and ARF17 transcript levels found in the arf10-3 mutant, arf16-3 mutant, arf10-3 arf16-3 double KO, and MIR160c-OX line showing the importance of the balance between these transcript levels for control of adventitious rooting. (C) Steady state levels of both the total (T) and uncleaved (UC) ARF transcripts were quantified in the hypocotyls of representative ARF6-OX, ARF8-OX, and ARF17-OX lines and arf6-3 and arf8-7 KO mutants. Transcript abundance was quantified in the hypocotyls of representative ARF mutant or overexpressing lines etiolated and transferred to the light for 72 h. Gene expression values shown are relative to the expression in the wild type, for which the value is set to 1. Error bars indicate ± se obtained from three independent RT-PCR experiments. A one-way ANOVA combined with the Dunnett's comparison test indicated that the values marked with an asterisk were significantly different from wild-type value (P < 0.01; n = 3). All quantifications were repeated using two additional independent biological replicates and gave similar results.

Figure 4.

Schematic of the miRNA Maturation Process That Leads to the Cleavage of Target mRNAs. pri-miRNAs are mostly transcribed by RNA polymerase II from miRNA encoding genes. The pri-miRNAs are processed into mature miRNAs through a reaction driven by the action of the C2H2-zinc finger protein SE, the double-stranded RNA binding protein HYL1, DCL1, and nuclear cap binding complex. Mature RNA duplexes excised from pre-miRNAs are methylated by HEN1 and exported to the cytoplasm possibly through the action of the HST1 protein. The guide miRNA strand is then incorporated into AGO1 protein to carry out the cleavage of target mRNAs.

Figure 5.

Real-Time RT-PCR Assessment of Posttranscriptional Regulation of miR160 and miR167 by ARF6, ARF8, and ARF17. Steady state levels of pri-miRNAs (pri-miR) ([A] and [D]) and mature miRNAs ([B] and [D]) were quantified in the hypocotyls of representative ARF mutant or overexpressing lines etiolated and transferred to the light for 72 h, as were DCL1 and AGO1 transcripts (C). Steady state levels of HYL1, HEN1, SE, and HST1 were quantified in the same conditions (E). Gene expression values shown are relative to the expression in the wild type, for which the value is set to 1. Error bars indicate ± se obtained from three independent RT-PCR experiments. A one-way ANOVA combined with the Dunnett's comparison test indicated that the values marked with an asterisk were significantly different from the wild-type value (P < 0.01; n = 3). All quantifications were repeated using two additional independent biological replicates and gave similar results.

Figure 6.

A Model Integrating the Regulatory Loops between ARF and miRNA Genes in the Control of Adventitious Rooting Based on Results Obtained in This Study. Adventitious root initiation is controlled by a subtle balance of activator and repressor ARF transcripts, which is maintained by a complex regulatory network. ARF6 has both a positive and a negative effect on ARF8 and ARF17 transcript levels. It regulates positively ARF8 and ARF17 total transcript levels, whereas it has a negative effect on their uncleaved transcript amount by modulating positively miR160 and miR167s abundance, which drives degradation of ARF17 and ARF8 transcripts, respectively. By regulating miR167s, it also regulates its own uncleaved transcript level. Moreover, ARF8 regulates negatively both ARF17 total transcript amount and miR167s abundance and by consequence ARF6 and its own uncleaved transcript level. In turn, ARF17 represses ARF6 total transcript abundance. In addition, ARF17 regulates positively the pool of miR167s and thereby has a negative effect on ARF6 and ARF8 uncleaved transcript abundance. ARF17 regulates its own uncleaved transcript abundance by feedback regulation of miR160 level. ARF6 and ARF8 are positively regulated by light. Nevertheless, since ARF6 regulates ARF8 transcript abundance, we suggest that light induction of ARF8 may be driven by ARF6. ARF17 is repressed by light.