The Arabidopsis ortholog of the DEAH-box ATPase Prp16 influences auxin-mediated development

Abstract

In animals and yeasts, the DEAH-box RNA-dependent ATPase Prp16 facilitates pre-mRNA splicing. However, in Chlamydomonas reinhardtii and Caenorhabditis elegans, Prp16 orthologs are not important for general pre-mRNA splicing, but are required for gene silencing and sex determination, respectively. The CLUMSY VEIN (CUV) gene, which encodes a unique Prp16 ortholog in Arabidopsis thaliana, influences auxin-mediated development. A loss-of-function cuv-1 mutation tells us that CUV does not facilitate splicing of pre-mRNA substrates indiscriminately, but differentially effects splicing and expression of genes. Here we show that CUV influences root-meristem maintenance and planar polarity of root-hair positioning, both of which are processes regulated by auxin. We propose that Arabidopsis PRP16/CUV differentially facilitates the expression of genes, including genes involved in auxin biosynthesis, transport, perception and signaling, and that in this way it influences auxin-mediated development.

In plants, organ and tissue development are directed by local auxin gradients associated with auxin maxima, which are defined by local auxin biosynthesis¹ and by polar auxin transport driven by the PIN family of auxin efflux proteins.^2,3 However, the mechanisms which regulate the expression of genes which drive auxin biosynthesis and polar auxin transport remain largely unknown.

As a model system to study auxin-mediated development, we have genetically analyzed vascular development in Arabidopsis.^4-6 clumsy vein-1 (cuv-1) was identified as an Arabidopsis mutant defective in leaf vascular development. cuv-1 is defective in auxin-mediated developmental processes including male-gametophyte transmission, apical-basal patterning in the embryo and gynoecium, stamen development, phyllotaxy, root-hair elongation and vascular development.⁶ CUV encodes an ortholog of the DEAH-box RNA-dependent ATPase pre-mRNA-processing factor 16 (Prp16), which is conserved in eukaryotic cells. Pre-mRNA splicing is catalyzed by the ribonucleoprotein complex spliceosome. Pre-mRNA splicing occurs through 2 sequential steps: cleavage of the 5′ splice site and exon ligation. In animals and yeasts, Prp16 mediates conformational change of the spliceosome, thereby facilitating the first-to-second step transition of splicing.^7,8 Prp16 also proofreads 5′ splice site cleavage during the first step of splicing.^9-12 A loss-of-function mutation in the Arabidopsis PRP16/CUV differentially affects splicing and expression of genes, including genes involved in auxin biosynthesis, distribution, perception and signaling.⁶ Here we present additional data showing that cuv-1 is defective in root-meristem maintenance and planar polarity of root-hair positioning, both of which are regulated by auxin.

Auxin and its directional cellular efflux facilitate root-meristem maintenance.^13,14 Roots of cuv-1 were shorter than those of wild type, implying that CUV functions in root development. Root hairs in cuv-1 emerged closer to the root tip than those in wild type (cf. Fig. 1B with A), indicating that the differentiation zone is closer to the root tip in cuv-1. The size of the root meristem in cuv-1 was smaller than that in wild type (Fig. 1C), suggesting that CUV is required for root-meristem maintenance.

Figure 1.

The cuv-1 mutation affects root-meristem activity and planar polarity of root-hair positioning. (A and B) Roots of 5-day-old wild-type (A) and cuv-1 (B) seedlings. (C) The root-meristem cell number of wild type and cuv-1. The root meristem size was measured by counting the number of cortex cells in a single file of cells extending from the quiescent center up to the first rapidly elongating cell.²⁸ Root-meristem size was significantly smaller in cuv-1 than in wild type. Asterisks indicate P values: ***, P < 0.001. Individual P values were as follows: 1.0 × 10⁻⁹ (wild type, n = 15; cuv-1, n = 20), 1.3 × 10⁻¹⁰ (wild type, n = 18; cuv-1, n = 25), 1.0×10⁻¹³ (wild type, n = 12; cuv-1, n = 23), 3.9 × 10⁻¹⁵ (wild type, n = 14; cuv-1, n = 24) for roots grown at 22˚C for 3, 5, 7 and 10 days, respectively. (D and E) Root-hair position in wild-type (D) and cuv-1 (E) epidermal cells. Arrowheads indicate apical and basal ends of cells. The apical and basal directions are oriented toward the aerial tissue and the root tip, respectively. Asterisks mark root-hair initiation sites. (F) Quantitative analysis of relative root-hair position in wild type and cuv-1. The number of cells (frequency) with relative hair positions in 20 classes from basal (0) to apical (1) ends was determined for 150 hair cells from 30 roots each of wild type and cuv-1. Significance (P value) was determined by a non-parametric, 2-sample Kolmogorov-Smirnov test (http://www.physics.csbsju.edu/stats/KS-test.n.plot_form.html).^29,30 A significant difference in frequency distributions was detected for wild type versus cuv-1 (^**P < 0.001).

Auxin promotes root-hair formation.^15,16 Either over-expression of PIN auxin-efflux proteins^17,18 or a loss-of-function mutation of AUX1 auxin-influx gene,¹⁹ which supposedly decrease auxin in cells, causes reduction of root-hair outgrowth. The Aux/IAA family contains genes which negatively regulate auxin signaling. Gain-of-function mutations of the Aux/IAA genes axr2-1/iaa7, slr-1/iaa14 and axr3-3/iaa17 result in very few root hairs: a low-auxin phenotype.^15,20-22 The cuv-1 suppressed the root-hairless phenotype in axr2-1, slr-1 and axr3-3, suggesting that CUV influences auxin-mediated root-hair outgrowth.⁶ Auxin influences not only root-hair outgrowth but also root-hair placement. In Arabidopsis, root hairs emerge close to the basal (root tip-oriented) ends of hair-forming epidermal cells (Fig. 1D).²³ While root-hair positions are shifted apically in low auxin-response axr2 mutant, exogenous application of auxin leads to a basal shift of root-hair position in wild type,²³ suggesting that auxin modulates the planar polarity of root-hair positioning. In cuv-1, root-hair position was shifted basally in hair-forming cells (compare Fig. 1E with D; Fig. 1F). Thus, the cuv-1 mutation confers a high auxin-response-like phenotype in root-hair outgrowth and positioning. Indeed, in cuv-1 roots, a local increase in auxin response was observed around the root differentiation zone.⁶ These findings suggest that in addition to root-hair outgrowth, CUV influences auxin-mediated planar polarity of root-hair positioning. Local auxin biosynthesis and redistribution are required for planar polarity establishment in roots.²⁴ Therefore, proper expression of genes involved in auxin biosynthesis and redistribution is important for auxin-mediated planar polarity. The additional data presented here further support the idea that CUV is required for auxin-mediated development.⁶

There is increasing evidence that different pre-mRNAs have different dependencies on conserved components of the pre-mRNA splicing machinery. Transcript-specific changes in splicing can occur when the activity of core spliceosomal components are modulated.^25,26 Therefore, tissue-specific or developmental-stage-specific modulation of splicing factors such as PRP16/CUV could be used as a means for tuning of gene expression. In fact, impairment of PRP16/CUV function differentially affects expression of genes involved in auxin biosynthesis, transport, perception and signaling, which we think results in concerted influences on auxin-mediated development.⁶ On the other hand, it has been suggested that the splicing machinery is limiting, implying competition among pre-mRNAs, suggesting that changes in the composition of the pre-mRNA pool influence splicing regulation.²⁷ Because splicing components have the capacity to exert transcript specificity, different pre-mRNAs most likely have different limiting splicing factors and competing pre-mRNAs. Thus, splicing efficiency of a pre-mRNA may be influenced by its own abundance and affinity for limiting splicing factor(s), those of competing pre-mRNAs, and the activity of splicing factors.²⁷ Pre-mRNAs that do not compete well for splicing would be more susceptible to changes in splicing factor activity and the relative amounts of competing pre-mRNAs. To further clarify the role of PRP16/CUV in auxin-mediated development, a comprehensive identification of genes affected by cuv mutations is required.

Disclosure of Potential Conflicts of Interest

No potential conflicts of interest were disclosed.

Acknowledgments

We thank Toshiharu Shikanai, Franck A. Ditengou, William Teale, and Elizabeth Nakajima for critically reading the article, and Kiyotaka Okada and members of the Shikanai laboratory for stimulating discussions.

Funding

This work was supported in part by a Grant-in-Aid for Scientific Research (24570047 to R.T.) from the Japan Society for the Promotion of Science.