Pattern of auxin and cytokinin responses for shoot meristem induction results from the regulation of cytokinin biosynthesis by AUXIN RESPONSE FACTOR3

Abstract

De novo organ regeneration is an excellent biological system for the study of fundamental questions regarding stem cell initiation, cell fate determination, and hormone signaling. Despite the general belief that auxin and cytokinin responses interact to regulate de novo organ regeneration, the molecular mechanisms underlying such a cross talk are little understood. Here, we show that spatiotemporal biosynthesis and polar transport resulted in local auxin distribution in Arabidopsis (Arabidopsis thaliana), which in turn determined the cytokinin response during de novo shoot regeneration. Genetic and pharmacological interference of auxin distribution disrupted the cytokinin response and ATP/ADP ISOPENTENYLTRANSFERASE5 (AtIPT5) expression, affecting stem cell initiation and meristem formation. Transcriptomic data suggested that AUXIN RESPONSE FACTOR3 (ARF3) mediated the auxin response during de novo organ regeneration. Indeed, mutations in ARF3 caused ectopic cytokinin biosynthesis via the misexpression of AtIPT5, and this disrupted organ regeneration. We further showed that ARF3 directly bound to the promoter of AtIPT5 and negatively regulated AtIPT5 expression. The results from this study thus revealed an auxin-cytokinin cross talk mechanism involving distinct intermediate signaling components required for de novo stem cell initiation and shed new light on the mechanisms of organogenesis in planta.

Figure 1.

Regional establishment of auxin responses relative to WUS expression. A to C, DR5rev::GFP signals (green) in the edge region of the noninduced callus (SIM for 0 d; 94.3%; n = 122). D to F, Regional distribution of DR5rev::GFP signals in the calli grown on SIM for 2 d (83.3%; n = 108). G to J, Auxin distribution indicated by the DR5rev::GFP signals and WUS expression indicated by the pWUS::DsRed-N7 signal (magenta) in calli grown on SIM for 4 d (81.5%; n = 127). K to N, Traverse view of the callus grown on SIM for 4 d (80.9%; n = 110). O to R, DR5rev::GFP and WUS signals accumulate in the promeristem of calli grown on SIM for 6 d (80.0%; n = 100). S to U, Auxin distribution and WUS expression in the calli grown on NPA-containing SIM for 4 d (86.8%; n = 152). Bright-field (B, E, I, and T) and merged images of GFP (green) and DsRed (magenta) channels (J, N, and R) are shown. Chlorophyll autofluorescence is shown in blue (M, N, Q, and R). Bars = 100 µm.

Figure 2.

Regional establishment of cytokinin responses relative to WUS expression. A to C, TCS::GFP signals (green) in the edge of the noninduced callus (94.5%; n = 127). D to F, TCS::GFP signals beginning to distribute regionally in the calli grown on SIM for 2 d (85.9%; n = 128). G to I, Regional distribution of the TCS::GFP signals colocalizes with the pWUS::DsRed-N7 signal (magenta) in the calli grown on SIM for 4 d (89.2%; n = 222). J to M, Traverse views of the calli grown on SIM for 4 d. N to Q, TCS::GFP and WUS signals accumulated in the promeristem of calli grown on SIM for 6 d (80.0%; n = 110). R to T, Cytokinin response (TCS::GFP) and WUS (pWUS::DsRed-N7) expression in the callus grown on NPA-containing SIM for 4 d. U to W, Cytokinin responses (TCS::GFP) in the calli transformed with antisense PIN1 driven by an estrogen receptor-based transactivator, XVE (for fusion of the DNA-binding domain of the bacterial repressor LexA, the acidic trans-activating domain of VP16, and the regulatory region of the human estrogen receptor). The calli were incubated on estrogen-containing SIM for 4 d (82.8%; n = 116). A to I and N to W are longitudinal sections of calli. Bright-field (B, E, H, S, and V) and merged images of GFP (green) and DsRed (magenta) channels (I, M, and Q) are shown. Chlorophyll autofluorescence is shown in blue (L, M, P, and Q). Bars = 100 µm.

Figure 3.

Spatiotemporal expression of auxin biosynthetic genes during shoot induction. A, qRT-PCR analysis of YUC expression level in noninduced calli and in calli grown on SIM for 4 d. B, pYUC1::GUS signals in noninduced calli (91.7%; n = 121) and in calli grown on SIM for 4 d (84.0%; n = 156) or 6 d (81.5%; n = 162). C, pYUC4::GUS signals in the noninduced callus (85.9%; n = 198) and in the calli grown on SIM for 4 d (85.6%; n = 167) or 6 d (83.3%; n = 150). D to K, Localization of pYUC4::GFP (green) and pWUS::DsRed-N7 (magenta) signals in the noninduced callus (D; 90.2%; n = 122) and in calli grown on SIM for 4 d (E–G; 84.4%; n = 166) or 6 d (H–K; 80.3%; n = 152). D to K are longitudinal sections of calli. Bright-field (D and F) and merged images of GFP (green) and DsRed (magenta) channels (G and K) are shown. Chlorophyll autofluorescence is shown in blue (J and K). Bars = 1 mm (B and C) and 100 µm (D, G, and K).

Figure 4.

Expression of cytokinin biosynthetic AtIPT genes within the callus during shoot induction and following NPA treatment. A, Expression levels of AtIPTs in the noninduced calli and in calli grown on SIM for 4 d determined by qRT-PCR. B, pAtIPT5::GUS signals in the noninduced callus (85.9%; n = 199). C to E, Regional distribution of pAtIPT5::GUS signals in calli grown on SIM for 2 d (C; 90.5%; n = 199), 4 d (D; 89.5%; n = 190), or 6 d (E; 87.9%; n = 182). F to H, AtIPT5 expression patterns in the noninduced callus (F; 90.9%; n = 209) and in calli grown on SIM for 4 d (G; 89.5%; n = 190) or 6 d (H; 87.9%; n = 182). I to K, pAtIPT5::GUS signals distributed uniformly at the edge region of the calli grown on SIM for 0 d (I; 85.2%; n = 88), 4 d (J; 83.3%; n = 66), and 6 d (K; 84.9%; n = 63) when treated with NPA. F to K are longitudinal sections of calli. PM, Promeristem. Bars = 700 µm (B–E), 200 µm (F, G, I, and J), and 50 µm (H).

Figure 5.

Spatiotemporal expression of ARF3 and its mutational effects on shoot formation. A to E, In situ hybridization analyses showing the spatial expression of ARF3 in the noninduced callus (A; 87.5%; n = 112) or in calli grown on SIM for 2 d (B; 80.5%; n = 123), 4 d (C; 75.4%; n = 130), 6 d (D; 72.7%; n = 110), or 8 d (E; 73.2%; n = 123). F, Regenerated shoots from wild-type (WT) calli grown on SIM for 18 d (89.9%; n = 99). G, The arf3/ett2 mutant callus grown on SIM for 18 d showing no shoot regeneration. H, The arf3/ett2 mutant callus grown on SIM for 18 d with a few regenerated shoots (34.5%; n = 200). CL, Cauline leaf. Arrows indicate regenerated shoots, and arrowheads indicate ARF3 signals. A to E are longitudinal sections of calli. Bars = 100 µm (A–C), 50 µm (D and E), and 1 mm (F–H).

Figure 6.

AuxRE-dependent ectopic expression of AtIPT5 in arf3. A and B, pAtIPT5::GUS signals in the wild-type callus induced on SIM for 0 d (A; 89.5%; n = 190) or 4 d (B; 85.9%; n = 199). C and D, GUS signals in the calli of the arf3 mutant grown on SIM for 0 d (C; 82.7%; n = 168) or 4 d (D; 80.4%; n = 148). E, Schematic illustration of the AtIPT5 promoter. TGTCTC and TGTCNN on both the sense and antisense strands are indicated by blue and green bars, respectively; red plus signs denote point mutations (TGTC→TGGC). F and G, pAtIPT5m::GUS signals in the noninduced callus (F; 88.8%; n = 179) or in the calli grown on SIM for 4 d (G; 85.4%; n = 198). A to D, F, and G are longitudinal sections of calli. Bars = 80 µm.

Figure 7.

AtIPT expression is enhanced in the arf3 mutant. qRT-PCR analyses show that the expression of AtIPT3 (A), AtIPT5 (B), and AtIPT7 (C) was enhanced in the arf3 mutant. AtIPT5 was the most significantly enhanced under these conditions.

Figure 8.

ARF3 directly binds to the promoter of AtIPT5. A, Yeast one-hybrid analysis revealing the direct interaction between ARF3 and the AtIPT5 promoter. Yeast strains containing the PAtIPT5 promoter-AbAi or PAtIPT5m promoter-AbAi construct were grown on medium under selective (SD/−Leu; +100 ng mL⁻¹ AbA) or nonselective (SD/−Leu; −AbA) conditions. Full-length ARF3 cDNAs fused to pGADT7 AD are indicated on the plates, and the empty pGADT7 AD vector was used as a negative control. The p53-AbAi vector was used as a positive control in the kit (Clontech Laboratories). B, Interaction between ARF3 and the AtIPT5 promoter (red line) was determined by SPR analysis. The AtIPT5m promoter was introduced as a negative control (blue line). C, EMSA analysis showing the interaction between ARF3 and the AtIPT5 promoter. The retarded DNA-protein complex was competed using either wild-type (WT) probe or the mutated probes at a 5×, 25×, or 50× molar excess. D, The AuxREs TGTCTC and TGTCNN on both sense and antisense strands of the AtIPT5 promoter are indicated as blue and green bars, respectively. Red lines indicate fragments amplified in E. Fragment a (−356 to +39) includes the sequence used in the EMSA experiments (−155 to −130), and fragment b (−1,133 to −850) was used as a negative control. E, Enrichment of specific regions of the AtIPT5 promoter (fragments a and b) using anti-GUS, anti-GFP, and anti-MYC antibodies in pARF3::ARF3tasiR-GUS, pMP (ARF5)::MP-GFP, and 35S::6myc-ARF8 transgenic plants, respectively. Mouse IgG was used as a mock control. The fold enrichments of specific regions (fragments a and b) were detected by qRT-PCR analysis after normalization to the unrelated UBQ10 control sequence. Means were calculated from three biological replicates, and each biological sample was examined using three PCR technical replicates.