Functional Analysis of the Rice Type-B Response Regulator RR22

Abstract

The phytohormone cytokinin plays a critical role in regulating growth and development throughout the life cycle of the plant. The primary transcriptional response to cytokinin is mediated by the action of the type-B response regulators (RRs), with much of our understanding for their functional roles being derived from studies in the dicot Arabidopsis. To examine the roles played by type-B RRs in a monocot, we employed gain-of-function and loss-of-function mutations to characterize RR22 function in rice. Ectopic overexpression of RR22 in rice results in an enhanced cytokinin response based on molecular and physiological assays. Phenotypes associated with enhanced activity of RR22 include effects on leaf and root growth, inflorescence architecture, and trichome formation. Analysis of four Tos17 insertion alleles of RR22 revealed effects on inflorescence architecture, trichomes, and development of the stigma brush involved in pollen capture. Both loss- and gain-of-function RR22 alleles affected the number of leaf silica-cell files, which provide mechanical stability and improve resistance to pathogens. Taken together, these results indicate that a delicate balance of cytokinin transcriptional activity is necessary for optimal growth and development in rice.

Keywords: cytokinin; grain yield; panicle; rice; root architecture; silica cell; trichome; type-B response regulator.

Figure 1

RR22 gain-of-function and loss-of-function lines. (A) Structural design of the RR22-OX cassette. A genomic version of RR22 was expressed from the Zea mays UBI1 promoter incorporating the first intron of the UBI1 gene. A sGFP fluorescent tag was encoded at the 3'-end of RR22. The intron/exon structure of the RR22 gene is shown below the cassette diagram. (B) Insertional positions for the four Tos17 alleles of rr22. (C) Relative protein expression levels of RR22-GFP in three independent RR22-OX lines as assessed by immunoblot using an anti-GFP antibody. A non-transgenic wild-type (WT) extract serves as a negative control. Ponceau S staining was used as a protein loading control. (D) Distribution of RR22-GFP fluorescent signal in root, dark-grown coleoptile, and leaf of RR22-OX L4. The WT root is a negative control. Scale bar = 100 μm. (E) Transcript levels of rr22 and ACT1 in the Tos17 rr22 mutant lines and corresponding WT siblings assessed by semiquantitative qPCR.

Figure 2

Cytokinin hypersensitive response of RR22-OX lines. (A) Expression of cytokinin-dependent genes based on qRT-PCR. Seven-day-old seedlings were treated with 1 μM BA or a NaOH vehicle control for 1 h, and gene expression examined in the shoots of the RR22-OX lines and their WT siblings (n = 3). Expression of the ACT1 gene was used for normalization. The fold change in gene expression is indicated in blue lettering for each BA-treated to untreated sample. (B,C) Root growth response to cytokinin. Seminal root length (B) and lateral root density (C) were determined for 7-day-old seedlings grown on Yoshida hydroponics medium supplemented with a vehicle control, 10, or 100 nM BA (n ≥ 8). For statistical analyses, ANOVA with a post hoc Holm multiple comparison calculation was performed for each line, involving RR22-OX and the WT sibling of the line. Different letters indicate significant differences at p < 0.05.

Figure 3

Shoot growth phenotypes of RR22-OX lines. (A) Shoots of RR22-OX lines and their WT siblings, taken at the initiation of flowering in the WT plants. Scale bar = 30 cm. (B) Plant height at flowering, number of tillers at flowering and at seed maturation. (C) Flag leaf length and width. (D) Epidermal long cell 1 (LC1) length and width of the flag leaves. Error bars show SE. The T-test was used for statistical comparison of each RR22-OX line to its WT sibling (^*p < 0.05; ^**p < 0.01; ^***p < 0.001; ns, not significant).

Figure 4

Leaf cell file phenotypes of RR22-OX, rr22, and rr21/22/23 mutant lines. (A) Representative images of the abaxial epidermis of the flag leaf. The region containing the silica/cork cell files is highlighted with a yellow box, and each silica/cork cell file indicated with a black arrow. Each stomata file is indicated with an asterisk (^*). (B) Quantification for the number of rows of adjacent silica/cork cell files. Impressions from five plants (two per plant) were examined and the number of rows of adjacent silica/cork cell files determined (n > 47). Error bars show SE. The T-test was used for statistical comparison of each mutant line to its WT sibling (^***p < 0.001).

Figure 5

Panicle phenotypes of RR22-OX lines. (A) Panicles of RR22-OX lines and their WT siblings. Scale bar = 5 cm. (B) Yield parameters of the RR22-OX lines, including panicle length, primary branch (PB) and secondary branch (SB) number, spikelets and grains per panicle, and percent seed set (n = 15). Error bars show SE. The T-test was used for statistical comparison of each RR22-OX line to its WT sibling (^***p < 0.001).

Figure 6

Panicle phenotypes of Tos17 rr22 lines. (A) Panicles of rr22-1, rr22-2, rr22-3, and rr22-4 lines and their WT siblings. Scale bar = 5 cm. (B) Yield parameters of the rr22 lines, including panicle length, PB and SB number, spikelets and grains per panicle, and percent seed set (n ≥ 6). The T-test was used for statistical comparison of each rr22 line to its WT sibling (^*p < 0.05; ^**p < 0.01; ^***p < 0.001; ns, not significant).

Figure 7

Trichome-related phenotypes of Tos17 rr22 lines. (A) Stigma brushes (scale bar = 0.5 mm) and trichomes on grain hulls (scale bar = 1 mm). (B) Stigma brush hair length and hull trichome length, with the five longest hairs or trichomes measured per stigma or hull, respectively (n = 50). (C) Expression analysis by qRT-PCR of the stigma-enriched genes GL3A and EXPA6 in carpels of rr22 mutants and their WT siblings (n = 4). Expression of the UBQ5 gene was used for normalization. Error bars are for SE. The T-test was used for statistical comparison of each rr22 line to its WT sibling (^*p < 0.05; ^**p < 0.01; ^***p < 0.001; ns, not significant).