Root Growth Adaptation is Mediated by PYLs ABA Receptor‐PP2A Protein Phosphatase Complex

Abstract Plant root architecture dynamically adapts to various environmental conditions, such as salt‐containing soil. The phytohormone abscisic acid (ABA) is involved among others also in these developmental adaptations, but the underlying molecular mechanism remains elusive. Here, a novel branch of the ABA signaling pathway in Arabidopsis involving PYR/PYL/RCAR (abbreviated as PYLs) receptor‐protein phosphatase 2A (PP2A) complex that acts in parallel to the canonical PYLs‐protein phosphatase 2C (PP2C) mechanism is identified. The PYLs‐PP2A signaling modulates root gravitropism and lateral root formation through regulating phytohormone auxin transport. In optimal conditions, PYLs ABA receptor interacts with the catalytic subunits of PP2A, increasing their phosphatase activity and thus counteracting PINOID (PID) kinase‐mediated phosphorylation of PIN‐FORMED (PIN) auxin transporters. By contrast, in salt and osmotic stress conditions, ABA binds to PYLs, inhibiting the PP2A activity, which leads to increased PIN phosphorylation and consequently modulated directional auxin transport leading to adapted root architecture. This work reveals an adaptive mechanism that may flexibly adjust plant root growth to withstand saline and osmotic stresses. It occurs via the cross‐talk between the stress hormone ABA and the versatile developmental regulator auxin.

. ABA and stresses affect root architecture. a) The effect of ABA on lateral root organogenesis. Density of lateral root primordia and emerged lateral root was quantified in 7-  These phosphopeptides all had Mascot score higher than 30 (P <0.05). b) The statistical analysis of the abundance changes of the phosphopeptide 257-273. The fold change values were averages of three biological replicates. Error bars represent ± SE. c) An in vivo phosphorylation profile after Phos-tag mobility shift assay. Protein extracts from 7-d-old seedlings treated with 30 × 10 -6 M ABA or ethanol (solvent control) for 2 h were separated in a Phos-tag gel. Strongly shifted upper bands should be phosphorylated PIN1, and the weak lower bands should be non-phosphorylated PIN1. Both phosphorylated and unphosphorylated PIN1 proteins (arrowheads indicated) were detected with anti-PIN1 antibody. pin1 mutant were used as a negative control. The strong lower nonspecific bands are indicated by asterisks. Considering that all bands shifted abnormally for phos-tag gels, the protein markers do not tell the real mobility for proteins. + and -indicate incubated with or without treatment. d-g) Quantification of root gravitropic bending in phospho-deficient and phospho-mimic lines under salt and mannitol treatments. The root growth angle was measured in 5-d-old seedlings gravistimulated with or without 50 × 10 -3 M NaCl or 250 × 10 -3 M mannitol for 24 h (n ≥ 11 roots). Two independent transgenic lines PIN2-Dendra S1,2,3D-1 and PIN2-Dendra S1,2,3D-2 were used (f,g). h,i) Quantification of lateral root density in phospho-deficient and phosphomimic lines under salt and mannitol treatments. Seven-day-old transgenic lines germinated on MS medium supplemented with or without 20 × 10 -3 M NaCl or 25 × 10 -3 M mannitol were used for lateral root density quantification (n ≥ 11 roots). Error bars represent ± SE. Means with different letters are significantly different at P < 0.05 (Fisher LSD test) (d-i). Three independent experiments were performed with similar results. showing that PID interacts with PP2AC3/C4. As a negative control, PYL1-Flag did not coimmunoprecipitate with RCN1-GFP. The PP2AC3/C4-Flag and PID-GFP plasmids were co-transformed into Arabidopsis protoplasts. Total isolated proteins were immunoprecipitated with anti-Flag agarose and detected by immunoblotting with anti-GFP antibody. b,c) A LCI assay showing that PID interacts with PP2AC4 in plant cells. Relative fluorescence was quantified by ImageJ (c). Scale bar, 1 cm. Error bars represent ± SE. ( * * ) P < 0.01 (Student's t-test). d) Strongly enhanced phenotypes of pp2ac3 pp2ac4 35S::PID seedlings as compared to either parental line of the same age. Scale bar, 0.5 cm. e,f) Co-IP assays showing that PIN2HL interacts with PP2AC3/C4 (e) and also with PP2AC1/C2/C5 (f). As a negative control, PYL1-Flag did not coimmunoprecipitate with RCN1-GFP. The PP2ACs-GFP and PIN2HL-Flag plasmids were co-transformed into Arabidopsis protoplasts.
Coimmunoprecipitation and immunoblotting analysis were performed as in (a). + andindicate incubated with or without extracts, respectively (a,e,f). g) A yeast two-hybrid assay showing that PIN1HL interacts with PP2AC4. AD: B42 activation domain; BD: LexA DNAbinding domain. Three independent experiments were performed with similar results.
Representative images are shown. Figure S5. PYLs regulate PP2A activity and thus PID-mediated PIN phosphorylation. a) An in vivo PP2A activity assay. The total protein used for the assay was extracted from 10-d-old seedlings. The phosphatase activity of wild type was set to 100%. b) An in vitro enzyme activity assay of PP2AC4. The concentrations of PYLs and ABA were 8 × 10 -6 and 5 × 10 -6 M, respectively. The phosphatase activity of PP2AC4 alone was set to 100%. PP2A activity was always measured after addition of 1 × 10 -3 M EDTA to inhibit the activities of PP2C (a,b).
OA was used as a phosphatase inhibitor (a,b). c) Protein purity of PYLs isolated from E.coli and PP2AC3 immunoprecipitated from Arabidopsis protoplasts. These purified proteins were used in different assays. d) Kinetic-dependent curve of PP2AC3 phosphatase activity with or without PYLs and ABA in vitro. PP2AC3 phosphatase activity was measured using different concentrations of substrate phosphopeptide (6.25 × 10 -6 , 12.5 × 10 -6 , 25 × 10 -6 , 50 × 10 -6 , 100 × 10 -6 and 200 × 10 -6 M). The concentration of PYLs was 4 × 10 -6 M. The calculated Km and Vmax of PP2AC3 alone, PP2AC3 and PYLs without or with ABA treatment was 19.9 × 10 -6 , 14.1 × 10 -6 or 21.0 × 10 -6 M, and 26.9, 31.5 or 27.1 nmol/min, respectively. The data were curve-fitted using the Origin2018 software. e) ABA-bound PYR1 inhibiting the phosphatase activity of ABI1 in vitro. The different molecular ratios between recombinant ABI1 and PYR1 proteins used for the phosphatase activity assay were 1:0, 1:1, 1:2, and 1:4, respectively. The concentration of applied ABA was 5 × 10 -6 M. 5 × 10 -6 M OA was applied to the reaction to inhibit the activity of PPP family Ser/Thr-specific phosphoprotein phosphatases, such as PP1 and PP2A. The phosphatase activity of untreated ABI1 was set to His-tagged PYL1 protein pulled down with GST, GST-PP2AC3, or GST-PP2AC4 was detected by anti-His antibody. f) An in vivo PP2A activity assay in the pp2ac3 pp2ac4 double mutant. g) An in vitro phosphorylation assay of PIN2HL in the pp2ac3 pp2ac4 double mutant. Equal amounts of total protein extracts from 5-d-old seedlings treated or not with 30 × 10 -6 M ABA for 4 h were co-incubated with heterologously expressed GST-PIN2HL, and then used for an in vitro phosphorylation assay. + and -indicate incubated with or without substrate, extracts, or ABA treatment, respectively (a,b,e,g). h) mRNA detection of PP2AC3 and PP2AC4 in wild-type roots by reverse transcription polymerase chain reaction (RT-PCR).
ACTIN2 was used as a reference control gene. RNA was isolated from roots of 5-d-old wildtype seedlings treated or not with 30 × 10 -6 M ABA for 4 h. i,j) Increased tolerance of pp2ac3 and pp2ac4 mutants to the ABA-mediated inhibition in root gravitropic bending (i) and lateral root primordia development (j) (n ≥ 10 roots). k) NaCl-insensitive phenotypes of pp2ac3 and pp2ac4 mutants in terms of root gravitropic bending (n = 13 roots). l) Tolerance of pp2ac3 and pp2ac4 mutants under salt stress. Five-day-old seedlings grown on MS medium were transferred onto MS medium supplemented with 175 × 10 -3 M NaCl for 4 d. Survival rates of pp2ac3 and pp2ac4 were evaluated after salt treatment (n = 180 seedlings). m) PP2AC2 transcription in wild type and pp2ac2 mutant detected by quantitative real-time PCR (qRT-PCR). The expression of PP2AC2 in the wild type was set to 1. ACTIN2 is the internal control. RNA was isolated from 5-d-old seedlings. n-p) ABA response in pp2ac1, pp2ac2, and pp2ac5 mutants. Sensitivity of pp2ac1, pp2ac2, and pp2ac5 to ABA is almost similar to that of wild type in root gravitropism (n) and lateral root organogenesis (o, p) (n ≥ 10). Error bars (d,f,i-p) represent ± SE. ( * ) P < 0.05, ( * * ) P < 0.01 (Student's t-test). Three independent experiments were performed with similar results. Representative images are shown.

Root gravitropism evaluations
Five-day-old seedlings grown vertically on MS medium were transferred to solid medium supplemented with 30 × 10 -6 M ABA or the equivalent amount of ethanol as control. This concentration is commonly used for root-bending assay. [79][80][81] For root gravitropic index assays, the started positions of the root tips were marked and the plates were photographed after 5 d of vertical growth on ABA or control medium. Downward growth was recorded from the marked point. The vertical length and root length were measured with ImageJ (https://imagej.nih.gov/ij/download.html), and a root vertical growth index (VGI = vertical length / root length) was calculated as described previously. [82] For root gravitropic bending assays, the plates were immediately turned 90° compared with the original vertical position.
After 24 h of the gravistimulation in darkness, the root growth angle was measured using ImageJ. [34] For the DR5 translocation assays, seedlings were gravistimulated at 90° for 4 h in darkness after 16 h of vertical growth on ABA or control medium. For PIN2 translocation assays, seedlings were gravistimulated at 90° for 4 h in darkness immediately after transferring to ABA or control medium. Confocal imaging was performed immediately and ImageJ was used to measure fluorescence intensity.

Lateral root density evaluations
7-day-old seedlings germinated on the MS medium supplemented or not with 0.3 × 10 -6 M ABA. Lateral root initiation and developmental progression analyses were performed as described previously. [49,83] Auxin transport assays Basipetal auxin transport was measured in 6-d-old plants as described. [84]

In situ expression and localization analyses
Whole-mount immunolocalization in Arabidopsis roots was performed as described. [85] The

Y2H assays
The full-length coding sequences of PYL1 and PP2ACs were fused into pGBKT7 (binding domain, BD) and pGADT7 (activation domain, AD) vectors, respectively. These plasmids were co-transformed into yeast strain AH109 according to the standard yeast PEG transformation method. Transformed yeast cells were separately sprayed onto 2D synthetic dropout medium lacking Trp/Leu and 4D selective medium lacking Trp/Leu/His/Ade, and incubated for 4-5 d at 30°C. For ABA treatment, 4D selective medium was supplemented with 50 × 10 -6 M ABA. LexA-based Y2H assays were performed as described previously. [86] Co-IP assays Full-length PYR1, PYL1, and PYL2 were fused into the 35S::Flag vector with an N-terminal Flag tag, and PYL4, PP2ACs, PID, and PIN2HL were fused into the pCAMBIA1300 vector with a C-terminal GFP or Flag tag. The primers used to clone the constructs are listed in Table S1. Arabidopsis wild-type protoplasts transformed with an equivalent amount of the relevant constructs were incubated in 1 mL of W5 buffer (154 × 10

In vitro pull-down assays
The pull-down assay was performed as previously described with some modifications. [31] Briefly, the full-length coding sequence of PP2AC3, PP2AC4 or PYL1 was cloned into the pGEX-4T-1 vector or the pET28a vector. GST, GST-PP2AC3, GST-PP2AC4 proteins were purified on Glutathione Sepharose 4B beads (GE) and PYL1-His protein was purified on Ni Sepharose (GE). 10 μg of GST or GST-PP2AC3, GST-PP2AC4 proteins was incubated with 10 μL Glutathione Sepharose 4B beads mixed with 1×PBS buffer containing 0.1% Triton X-100 at 4°C for 1 h. The buffer was then removed after centrifugation. 1 μg of PYL1-His protein mixed with the fresh binding buffer was incubated at 4°C and rotated for 2 h. The beads were subsequently washed four times with 1×PBS buffer containing 0.1% Triton X-100. The pulled-down proteins were eluted in 5 × SDS loading buffer at 100°C for 5 min, and then separated on 10% SDS-PAGE gels, eventually detected by immunoblotting.

LCI assays
The full-length coding sequence of PP2AC4 was amplified and fused to the C-terminal of the pCAMBIA-cLUC vector, and the full-length coding sequence of PYL1 and PID were amplified and fused to the N-terminal of the pCAMBIA-nLUC vector, respectively. The LCI assay was performed as previously described. [87] All fusions were transformed into

In vitro phosphorylation assays and mass spectrometry analyses
The His-PIN1HL and GST-PIN2HL C-terminal fusions were constructed by fusing the hydrophilic loops of PIN1 and PIN2 into the pET30a and pGEX-4T-1 vectors, respectively.

In vivo and in vitro PP2A activity measurements
The Serine/Threonine Phosphatase Assay System (Promega) was used to determine the PP2A activity according to the manufacturer's instructions. For the in vivo assays, PP2A activity was measured in total protein extracts from 10-d-old seedlings. 1 × 10 -3 M EDTA was always added to the reaction to inhibit the activity of PP2C. For in vitro assays, PYLs were cloned into the pET28a or pGEX-4T-1 vector. Recombinant proteins of PYLs-His or GST-PYLs were expressed in E. coli BL21 strain and extracted as described for in vitro phosphorylation assays. Arabidopsis wild-type protoplasts were transiently transformed with 35S::PP2AC3-Flag and 35S::PP2AC4-Flag, and proteins were extracted as described for co-IP assays. After washing anti-Flag beads for 5-6 times, elution buffer (containing 100-500 μg mL -1 Flag peptide in TBS buffer) was mixed with the anti-Flag beads, which was incubated at 4°C and In vitro PP2C activity measurements PP2C activity was measured using a Serine/Threonine Phosphatase Assay System (Promega) as described previously. [36] The buffer used for PP2C activity measurement contained 250 ×