The apoplastic pH is a key determinant in the hypocotyl growth response to auxin dosage and light

Plant elongation growth is strongly influenced by light through interaction with phytohormones. Higher light intensity usually results in a shorter plant stature, whereas darkness or shade promotes a long and slender plant stature¹. At the cellular level, elongation growth is driven by cell expansion, a process predominantly controlled by the core phytohormone auxin in higher plants². While auxin is necessary for cell expansion, it can also inhibit the process when supplied in excessive amounts. As early as 1933, an optimal concentration of auxin that promotes the elongation of excised sections of the Avena coleoptile was reported, above which auxin inhibits growth³. Using pea epicotyl or Arabidopsis hypocotyl, studies found that when endogenous auxin was removed by decapitation, low concentrations of the auxin analogues 1-naphthaleneacetic acid (NAA) and indole-3-acetic acid (IAA) induce elongation, but high concentrations inhibit elongation^4,5,6. Today, the nature of the dosage-dependent reversal of auxin responses in hypocotyl elongation remains unclear. Although the biphasic effect or dual effect of auxin most often refers to the dosage-dependent reversal, light conditions can also reverse the hypocotyl response to auxin. Arabidopsis mutants with higher auxin levels exhibit longer hypocotyls in the light but shorter hypocotyls in the dark^7,8,9. As early as 1949, light has been observed to decrease auxin-induced growth in pea epicotyl⁶. Today, studies show that light modulates auxin homeostasis and antagonizes auxin signalling at multiple levels¹⁰. However, why auxin tends to promote hypocotyl elongation in the light but represses it in the dark has yet to be explained.

The ‘acid growth theory’, proposed over 50 years ago, has gained increasing support as a key mechanism behind auxin-mediated cell expansion^{5,11,12,13,14,15}. In brief, cell expansion requires an increased extensibility of the cell wall and cell turgor pressure. Auxin activates plasma membrane proton pump H⁺-ATPases (AHAs in Arabidopsis) by phosphorylating the C-terminal penultimate threonine residue (pT-AHA), for example, pThr947 of AHA2 (ref. ⁴), leading to proton extrusion into the extracellular space. The subsequent apoplastic acidification activates cell wall modification proteins and results in cell wall loosening and remodelling^14,16. Meanwhile, H⁺-ATPase activation is accompanied by plasma membrane hyperpolarization, which energizes the inward rectifying K⁺ channel to allow K⁺ uptake and water influx, thereby increasing turgor pressure to drive cell expansion^2,17. The ABP1-TMK system has been shown to phosphorylate pT-AHA^15,18,19, while the dephosphorylation of pT-AHA is controlled by D-clade type 2C protein phosphatases (PP2C-Ds), which inactivate the H⁺-ATPases¹³. Through the TRANSPORT INHIBITOR RESPONSE1/AUXIN SIGNALING F-BOX PROTEIN (TIR1/AFB)-mediated transcriptional pathway, auxin rapidly and transiently induces three large families of transcripts: SMALL AUXIN-UP RNAs (SAURs), GH3-related and AUXIN/INDOLE-3-ACETIC ACID (Aux/IAA)²⁰. Among them, the SAUR genes, which encode small and short-lived proteins, serve as output effectors to mediate auxin responses at the cellular level, such as cell expansion^21,22.

SAUR19 and several members of the SAUR family function by interacting with and blocking PP2C-D phosphatase activity, thereby indirectly inducing the activity of H⁺-ATPase and causing apoplastic acidification^13,23,24. Through this mechanism, SAUR19, SAUR61 and SAUR50 subfamilies have been shown to induce cell expansion associated with hypocotyl elongation^25,26,27,28. Further studies show that auxin-induced hypocotyl elongation requires TIR1/AFB-mediated nuclear pathway and new protein synthesis⁵. In particular, SAUR proteins are crucial downstream mediators of auxin-induced growth, and constitutive expression of SAUR19 confers constitutive auxin-independent growth^5,29. Point mutations in PP2C-Ds that specifically evade binding and inhibition by SAURs result in short hypocotyls, dwarfism and lack of auxin response²⁴, underscoring the essential role of the SAUR-PP2C-D module in auxin-stimulated growth. However, the repression of hypocotyl elongation elicited by a high dosage of auxin is poorly understood, as is the mechanism for the dramatic reversal of the hypocotyl response to auxin by light.

Auxin’s effects on hypocotyl are dosage- and light-dependent

The light- and dosage-dependent biphasic effect of auxin was investigated using Arabidopsis hypocotyls. When grown in media containing increasing concentrations of NAA, dark-grown seedlings (Col-0) displayed shortening hypocotyls, whereas light-grown seedlings displayed longer hypocotyls, with the longest occurring at 10 μM NAA (Fig. 1a–d). This result recapitulated the phenomenon that auxin typically inhibits hypocotyl elongation in the dark but promotes it in the light. Similarly, IAA inhibits hypocotyl elongation in dark-grown seedlings but has no effect in the light, probably due to its instability under light conditions (Supplementary Fig. 1a–d).

a–d, Wild-type Col-0 seedlings were grown on MS plates containing the indicated concentrations of NAA in the dark for 4 days or in the light for 6 days. NAA inhibited elongation of dark-grown hypocotyl but promoted elongation of light-grown hypocotyl. Scale bar, 1 mm (a,c). Hypocotyl lengths were measured (b,d) and statistical significance was calculated using one-way analysis of variance (ANOVA) with Tukey’s test for multiple comparisons (P < 0.05). For boxplots here and in the rest of the figures, the central line represents the median, the bottom and top edges of the box correspond to the 25th and 75th percentiles, respectively, and the whiskers extend to the highest and lowest data points within the group. Different lowercase letters above the bars indicate statistically significant differences between groups. e,f, The response of hypocotyls to auxin dosage depends on both auxin concentration and treatment duration. e, Col-0 seedlings grown in the dark for 2 days were treated with the auxin analogue picloram (Pic) or DMSO (Mock) in darkness for the indicated periods. f, Seedlings grown in the light (100 μmol m⁻² s⁻¹) for 3.5 days were treated with Pic or DMSO under dim light (3.2 μmol m⁻² s⁻¹) for the indicated periods. The hypocotyl lengths were measured and statistical significance was assessed using two-way ANOVA and Tukey’s multiple comparisons test (P < 0.05). g–j, Time-lapse imaging and quantitative measurement of hypocotyl growth rate in response to auxin. g,h, Two-day-old dark-grown seedlings (Col-0) were transferred to MS medium plates supplemented with the indicated Pic or DMSO. Data are shown as means ± s.e.m. (n = 13, 13, 13, 9, 14 and 16 for each treatment from Mock to 50 μM Pic). i,j, Wild-type (Col-0) seedlings were grown on MS plates supplemented with 50 μM l-kynurenine (l-kyn) and 50 μM yucasin for 2 days to inhibit endogenous auxin synthesis, then transferred to MS plates containing NAA or equimolar amounts of ethanol (Mock). Data are shown as means ± s.e.m. (n = 18, 14, 19, 20, 10, 11 and 15 for each treatment from Mock to 50 μM NAA).

Source data

Full size image

To examine the effect of treatment duration, we performed a series of experiments varying auxin concentration and treatment timepoint. In the dark, treatment with a low concentration of the auxin analogue picloram (0.1 μM Pic) for 3 h slightly enhanced hypocotyl elongation, but prolonging the treatment to 36 h inhibited hypocotyl elongation (Fig. 1e). Similarly, in the light, treatment with 50 μM Pic enhanced hypocotyl elongation in the first 36 h but became inhibitory after 60 h (Fig. 1f). Time-lapse imaging of etiolated seedlings revealed that inhibition of hypocotyl growth rate by high concentrations of auxin, such as 5 and 50 μM Pic or IAA, occurred within 30 min of treatment (Fig. 1g,h and Supplementary Fig. 1e,f). It was difficult to detect the promotion effect of auxin in these experiments, probably due to high endogenous levels of auxin in etiolated seedlings²⁵. In early studies decades ago, coleoptiles or epicotyls were first decapitated to eliminate endogenous supply of auxin from the shoot apex, allowing for easier observation of growth promotion by applied auxin^3,6. Instead of decapitation and physically damaging the hypocotyl, we pretreated dark-grown seedlings with the auxin biosynthetic inhibitors l-kynurenine and yacasin before conducting time-lapse hypocotyl imaging (Fig. 1i,j). The data showed that high concentrations of auxin (5 μM and 50 μM NAA) initially promoted dark-grown hypocotyl elongation but became inhibitory after a few hours of the treatment (Fig. 1i,j). Together, these results indicate that the response of hypocotyls to auxin depends not only on auxin concentration but also on the cumulative duration of the exposure to auxin. Here we use the term ‘dosage’ to refer to the combined effect of auxin concentration and the duration of treatment.

Light- and dark-hypocotyl transcriptomes respond to auxin dosage

Auxin-induced hypocotyl elongation requires TIR1/AFB-mediated transcription and new protein synthesis⁵. However, how high dosages of auxin cause growth inhibition of the hypocotyl is unclear. We considered two possible scenarios that may result in the reversal of auxin’s effect from promotion to inhibition. One possibility involves dosage- or light-dependent transcriptional reprogramming, in which auxin-induced growth-stimulating or inhibitory genes may reverse their expression trajectories between low and high dosages of auxin, or between light and dark conditions, to account for their opposite responses. The second possibility is that auxin-responsive transcriptomes do not show a reversal but instead, the downstream cell expansion activities level off and reverse during the persistent stimulation of auxin. To distinguish these possibilities, we conducted two sets of transcriptome analyses: (1) dissection of hypocotyls of etiolated seedlings treated with increasing concentrations of Pic or IAA and increasing treatment durations for auxin dosage responses (Supplementary Table 1 and Supplementary Fig. 2a,b); and (2) dissection of hypocotyls from etiolated or light-grown seedlings treated with 5 μM or 50 μM Pic to test the effect of light on auxin responses (Supplementary Table 2 and Extended Data Fig. 1a).

Comparing the auxin-responsive transcriptomes between dark- and light-grown hypocotyls, we found that 5 μM Pic induced a weaker response in light-grown hypocotyls than in the corresponding dark-grown hypocotyls (Extended Data Fig. 1a). Other than this, the general auxin-induced transcriptomic profiles looked similar between dark- and light-grown hypocotyls despite their opposite responses in hypocotyl elongation (Extended Data Fig. 1a). Only 4 differentially expressed genes (DEGs) showed opposite regulation by Pic in dark vs light hypocotyls (Extended Data Fig. 1b,c). We paid particular attention to the SAUR gene family, of which several members have been shown to induce cell expansion and hypocotyl elongation^13,25,30, including SAUR19 (SAUR15,19–29), SAUR61 (SAUR61–68) and SAUR9,10,14,50. Remarkably, these SAUR subfamilies were all induced in a concentration-dependent manner by picloram in both dark- and light-grown hypocotyls (Fig. 2a and Supplementary Table 3), even though dark-grown hypocotyls were inhibited by the treatment. In addition, Aux/IAA, PP2C-D and ACS genes displayed similar auxin-responsive expression patterns between dark and light-grown hypocotyls, with only quantitative differences, such as a weaker response to 5 μM Pic in the light compared with the dark (Extended Data Fig. 1d–f and Supplementary Table 3).

a, Heat map of Pic-responsive SAUR expression in dark- and light-grown hypocotyls. Wild-type seedlings grown in dark or light conditions were treated with Pic or DMSO for 45 min. Hypocotyls were then dissected for RNA sequencing. b,c, High-order saur mutant exhibited weakened auxin-elicited repression of hypocotyls in the dark. Col-0, saur19–24,26,27 (saur19^Octuple), saur7,13,15,19,20,22,24–29,73 (saur19^Tredecuple), saur6,12,14,16,50 and saur61–68,75 (saur61^Nonuple) were grown on solid MS media containing 5 μM Pic or DMSO in the dark for 3 days. The growth inhibition rate was calculated as (Mock − Pic)/Mock × 100%. d,e, High-order saur mutants exhibited weakened auxin-induced hypocotyl elongation in the light. Seedlings were grown on solid MS medium containing 5 μM Pic or DMSO in the light for 6 days. The growth rate was calculated as (Pic − Mock)/Mock × 100%. b,d, Significant differences were assessed using two-way ANOVA followed by Tukey’s post test (P < 0.01). c,e, Data from 3 biological replicates were analysed (means ± s.e.m.). For each replicate, n > 20. Statistical significance was assessed using one-way ANOVA with Tukey’s multiple comparisons test (P < 0.05). f, Relative expression levels of SAUR15 in hypocotyls of multiple transgenic 35S:SAUR15-GFP lines compared to the wild type. The seedlings were grown in the dark for 3 days (Dark) or under 0.8 μmol m⁻² s⁻¹ white light for 4 days and then transferred to 80 μmol m⁻² s⁻¹ white light for another 3 days (Light). Hypocotyls were dissected for RNA analysis. Data shown are means ± s.d. of 3 biological replicates. PP2A was used as an internal control. g–j, Hypocotyl phenotypes of SAUR15 overexpression lines with increasing transgene expression levels. Scale bar, 1 mm. Significant differences were assessed using one-way ANOVA with Tukey’s post test. The seedlings were grown in the dark for 3 days (g,h) or in the light for 6 days (i,j).

Full size image

The dosage-dependent transcriptomes were determined using hypocotyls of etiolated seedlings treated with 0.5 μM, 5 μM or 50 μM Pic (or IAA) for either 3 h or 12 h (Supplementary Table 1 and Supplementary Fig. 2a,b). In general, genes induced or repressed by 0.5 μM or 5 μM tended to be continuously and further induced or repressed by 5 or 50 μM, respectively. A progressive regulation (mostly upregulation) by increasing concentrations of Pic or IAA was consistently observed for Aux/IAA, PP2C-D, ACS and majority of the SAURs genes (Extended Data Fig. 2 and Supplementary Table 4). Between the 3- and 12-h treatment timepoints, the vast majority of genes showed similar response patterns (Supplementary Fig. 2h,i and Supplementary Table 6), although several genes changed the direction of their auxin response. These included genes related to terpenoid metabolism, detoxification and secondary metabolite process (Supplementary Fig. 2c–g and Table 5). On the other hand, the Aux/IAA, PP2C-D and ACS family genes showed similar pattern of regulation by auxin after prolonged exposure for 12 h (Extended Data Fig. 2d–i). In addition, the SAUR19 and SAUR61 subfamilies remained responsive to auxin, although the amplitude of transcript induction moderately subsided at the 12-h mark (Extended Data Fig. 2a,b). In SAUR15 Pro:SAUR15-GFP transgenic plants, the SAUR15-GFP protein persistently accumulated following picloram and IAA treatments at 4 and 16 h (Extended Data Fig. 2c).

In summary, the hypocotyl transcriptional programme was progressively activated in response to increasing auxin concentrations and displayed a similar pattern between dark- and light-grown hypocotyls. We did not find large-scale transcriptional reprogramming corresponding to the dramatic reversal of the hypocotyl response due to changes in auxin dosage or between dark and light conditions. However, it is possible that certain specific transcriptional changes may be overlooked in this analysis, which may have contributed to the reversed hypocotyl growth phenotype.

SAURs are key mediators of the biphasic hypocotyl response

To address the role of SAURs in auxin’s dual effect on hypocotyl elongation, we examined transgenic plants constitutively expressing SAUR50 or SAUR66-GFP. Similar to the reported SAUR19 constitutive expression plants^5,29, light-grown SAUR50 or SAUR66-GFP seedlings exhibited a constitutive auxin phenotype characterized by elongated hypocotyl and insensitivity to auxin (Extended Data Fig. 3). To create SAUR loss-of-function mutants, multiple related SAUR genes were simultaneously inactivated in an attempt to overcome genetic redundancy among SAURs. Using a multigene-editing CRISPR/Cas9 system³¹ (Supplementary Fig. 3), we generated high-order mutants of the SAUR19 and SAUR63 subfamilies: saur19^Octuple (saur19–24,26,27), saur19^Tredecuple (saur7,13,15,19,20,22,24–29,73) and saur61^Nonuple (saur61–68,75), in addition to previously generated mutant saur6,12,14,16,50 (ref. ³²). It should be mentioned that these targeted genes represent less than half the number of SAURs induced by 5 μM Pic, as it induces 27 SAUR genes at 45 min, 37 at 3 h, and 22 at 12 h (Supplementary Tables 3 and 4). Nevertheless, we observed weak but consistent auxin-mediated hypocotyl phenotypes in these mutants. Except for saur19^Octuple, the saur mutants exhibited shorter hypocotyls when grown in the dark (Fig. 2b and Supplementary Fig. 4a), indicating their importance in the elongation of etiolated hypocotyls. While picloram inhibited hypocotyl growth in all genotypes in the dark, the inhibition rates of saur6,12,14,16,50 and saur61^Nonuple appeared significantly compromised compared with those of the wild type, although the saur mutants had shorter hypocotyls without exogenous auxin treatment (Fig. 2c). Under light conditions, all of the mutants had normal hypocotyl length. However, while Pic (5 μM) promoted hypocotyl elongation in wild-type seedlings, all saur mutants showed significantly reduced stimulation (Fig. 2d,e and Supplementary Fig. 4b). In particular, saur19^Tredecuple showed strong insensitivity to auxin (Fig. 2e). These results show that SAURs are critical mediators in auxin’s biphasic control of hypocotyl elongation.

Increasing SAUR levels cause biphasic hypocotyl growth

Because auxin progressively induces SAURs, and SAUR proteins act at the cellular level to induce cell expansion and hypocotyl elongation, we wondered whether simply manipulating the levels of SAUR expression could phenocopy the dual effects of auxin on hypocotyl elongation. We focused on three SAUR subfamilies known to function in cell expansion and obtained multiple transgenic lines expressing a representative SAUR gene at drastically different levels. Ten independent 35S:SAUR15-GFP lines that overexpressed the transgene from approximately 10-fold (line #65) to over 2,500-fold (line #50) of the endogenous SAUR15 level were examined (Fig. 2f). In the dark, seedlings with moderate overexpression displayed longer hypocotyls than wild-type seedlings until SAUR15 was overexpressed 130-fold (line #54), and thereafter, greater SAUR15 overexpression corresponded with progressively shorter hypocotyls (Fig. 2g,h). The extreme transgenic line #50, which showed overexpression of more than 2,500-fold compared with the endogenous level, exhibited a stunted hypocotyl resembling that observed after growth under a high dosage of auxin (Figs. 2g,h and 1a,b, and Supplementary Fig. 1a). In the light, except for line #65, all of the other SAUR15-overexpressing lines displayed longer hypocotyls than the wild type (Fig. 2i,j). The growth-promoting effect of SAUR15 levelled off at just over 900-fold overexpression (line #41), and any additional SAUR15 production, such as in lines #145 and #50, was no longer more effective in promoting elongation (Fig. 2i,j).

A similar pattern of correlation was also observed between SAUR expression levels and hypocotyl phenotypes in transgenic lines overexpressing SAUR19, SAUR16, SAUR50, SAUR63 and SAUR66 (Supplementary Fig. 5). In all cases, hypocotyl elongation was promoted by moderate overexpression of SAURs and inhibited by excessive overexpression of SAURs. Similar to auxin, these SAURs promoted hypocotyl elongation until the level reached a dosage threshold. Thereafter, higher SAUR levels gradually weakened the growth-promoting effect, eventually leading to inhibition.

Plasma membrane (PM) H⁺-ATPase activation induces biphasic hypocotyl elongation

SAURs promote cell expansion via the SAUR-PP2C.D-AHA pathway, which results in increased phosphorylation of PM H⁺-ATPases C-terminal penultimate threonine (pT-AHA)^4,13. Using an antibody specific to phosphorylated pT-AHA³³, we observed that pT-AHA phosphorylation was induced by Pic treatment in etiolated hypocotyls and that the phosphorylation level increased from 5 μM to 50 μM Pic (Fig. 3a). Further, we utilized a non-invasive micro-test (NMT) system to monitor H⁺ flux from the epidermal cells of the upper hypocotyl, where active elongation takes place (Fig. 3b). The data showed that treatment with 0.5 μM and 5 μM IAA steadily increased H⁺ fluxes at 3 h and continued to induce H⁺ extrusion after 12 h of treatment (Fig. 3c). Remarkably, SAUR15 overexpression caused a progressive increase in pT-AHA phosphorylation (Fig. 3d) and H⁺ flux rates (Fig. 3e) as SAUR15 expression levels rose from line #54 (~130-fold overexpression) to line #102 (~500-fold overexpression) and line #145 (>1,600-fold overexpression). Noticeably, the highest H⁺ fluxes, as observed after 12-h treatment with 5 μM IAA in the dark or SAUR15 overexpression line #145, corresponded to conditions of hypocotyl inhibition (Supplementary Fig. 1e,f and Fig. 2g,h).

a, Picloram activates PM H⁺-ATPases in hypocotyls. Col-0 seedlings grown in the dark for 3 days were treated with the indicated concentration of picloram for 3 h. The hypocotyls were dissected for protein analysis. Antibodies against phosphorylated penultimate Thr947 of AHA2 (pThr947) or total AHA2 were used for immunoblotting. RPN6 was used as a control. b, The testing zone of H⁺ flux was located below the apical hook, spanning ~200–300 μm or 4–5 layers of hypocotyl epidermal cells (dashed box). c, Two-day-old etiolated Col-0 seedlings were treated with IAA or Mock for 3 or 12 h and subjected to H⁺ flux recording over a period of 320 s. Data are presented as means ± s.e.m. (for 3 h: Mock n = 6, 0.5 μM IAA n = 6, 5 μM IAA n = 8; for 12 h: Mock n = 7, 0.5 μM IAA n = 8, 5 μM IAA n = 6). d, Col-0 and three 35S:SAUR15-GFP transgenic lines (#54, #102 and #145) were grown in the dark for 2.5 days. Hypocotyls from these seedlings were collected for immunoblotting. e, SAUR15 overexpression promoted H⁺ efflux in hypocotyl epidermal cells. Two-day-old dark-grown Col-0 seedlings and 35S:SAUR15-GFP lines (#54 and #145) were used to record H⁺ flux. Data are presented as means ± s.e.m. (n = 5). f,g, Effects of fusicoccin on hypocotyl growth of etiolated seedlings. f, Col-0 seedlings were grown on MS plates containing the indicated concentrations of fusicoccin (FC) for 3 days in darkness. Scale bar, 1 mm. g, Hypocotyl lengths. Significant differences were assessed using one-way ANOVA with Tukey’s multiple comparisons test (P < 0.0001). h,i, Time-lapse imaging and hypocotyl growth measurements in response to FC. Two-day-old dark-grown Col-0 seedlings were transferred to MS medium plates supplemented with the indicated concentrations of FC or DMSO. Data are shown as means ± s.e.m. (n = 11, 11, 12 and 14 for each treatment from Mock to 20 μM FC).

Full size image

To directly investigate how activation of PM H⁺-ATPase affects hypocotyl elongation, we utilized the fungal toxin fusicoccin (FC), a chemical substance that keeps PM H⁺-ATPase in an active state³⁴. Increasing concentrations of FC elicited increased pT-AHA phosphorylation, H⁺ efflux in hypocotyl epidermal cells and a reduction in apoplastic pH (Extended Data Fig. 4a–d). Notably, hypocotyl growth exhibited a biphasic pattern relative to increases in the FC doses, as 0.25 μM FC promoted hypocotyl elongation, whereas high concentrations (1 μM and 5 μM) inhibited it (Fig. 3f,g). Time-lapse imaging of hypocotyl growth showed that high concentrations of FC strongly promoted hypocotyl elongation initially, then after a couple of hours, caused a reduction in growth rate compared with mock or low-concentration FC (0.25 μM) treatments (Fig. 3h,i and Extended Data Fig. 4e). Thus, increasing proton pump activity can also result in a biphasic response in hypocotyl elongation: moderate PM H⁺-ATPase activity promotes hypocotyl elongation, while uncontrolled stimulation hinders it.

Rising auxin dosage gradually acidifies hypocotyl apoplast

Studies have shown that auxin can induce apoplastic acidification in hypocotyls/coleoptiles^5,35, which triggers the acid growth mechanism. We wondered whether apoplast pH would continue to decrease in the case of auxin overdose. Utilizing the pH-sensitive dye 8-hydroxypyrene-1,3,6-trisulfonic acid trisodium salt (HPTS) staining assay³⁶, we found that 0.5 μM and 5 μM IAA steadily reduced the apoplastic pH of hypocotyl epidermis (Fig. 4a,b), correlating with the H⁺ flux rates (Fig. 3c). In addition, an apo-pHusion marker line was used, which assesses apoplastic pH on the basis of the ratio of EGFP intensity (quenched by low pH) to mRFP1 intensity (insensitive to pH). At concentrations that decreased hypocotyl growth of etiolated seedlings (Fig.1e,g,h), increasing Pic treatment from 0.5 μM to 5 μM caused a progressive reduction of the apoplastic pH (Fig. 4c,d). Similar results were obtained with 1 μM and 10 μM of NAA treatments (Supplementary Fig. 6). Similar to auxin, plants with higher SAUR expression levels showed lower apoplastic pH despite shorter hypocotyls (Extended Data Fig. 5). Consistent with this pattern, 10 μM FC caused a reduction in hypocotyl apoplastic pH (Extended Data Fig. 4c,d), while hypocotyl growth was inhibited (Fig. 3h,i). These results clearly show that higher concentrations of auxin cause the apoplast to be more acidic, even though the hypocotyls were strongly inhibited by auxin.

a–d, Higher auxin dosage induces greater apoplastic acidification in the hypocotyl epidermis. a,b, HPTS staining images (a) and quantification of relative apoplastic pH based on ratiometric values (458 nm/405 nm) of HPTS fluorescence (b). Col-0 seedlings grown in the dark for 2 days were treated with IAA or Mock for 3 or 12 h. Statistical significance was analysed using two-way ANOVA with Tukey’s multiple comparison test (P < 0.05, n ≥ 24). c,d, Assessment of apoplastic pH changes using the apo-pHusion marker line. apo-pHusion seedlings grown in darkness for 2.5 days were treated with Mock or Pic for 3 or 12 h. The relative apoplastic pH was estimated on the basis of the ratio of EGFP intensity (quenched by low pH) to mRFP1 intensity (insensitive to pH). Significant differences were assessed using two-way ANOVA with Tukey’s post test (P < 0.05, n ≥ 21). e, yuc1D showed higher PM H⁺-ATPase activities specifically in hypocotyls. Col-0 and yuc1D seedlings were grown in the dark for 3 days. Whole seedlings (W) or dissected hypocotyls (H) were used for protein extraction and immunoblotting with pThr947 and AHA2 antibodies. RPN6 was used as a loading control. f, H⁺ flux rates in the hypocotyl epidermal cells of Col-0 and yuc1D seedlings grown in the dark for 2 days. Data are shown as means ± s.e.m. (Col-0 n = 8, yuc1D n = 7). g,h, yuc1D displayed lower apoplastic pH in hypocotyl epidermal cells. Two-day-old dark-grown Col-0 and yuc1D were used for HPTS staining (g). The ratiometric quantification of HPTS staining (458 nm/405 nm) indicating the relative apoplastic pH is shown in h. Statistical significance was analysed using an unpaired two-tailed t-test (*P = 0.0150, n = 21). a,c,g, Scale bar, 500 μm. i, Schematic summary showing that steps of the auxin-induced nuclear pathway, including transcriptional responses and induction of SAURs, PM H⁺-ATPase activation, H⁺ efflux and apoplastic acidification in the hypocotyl, showed a progressive monophasic response to upstream signals (upward green arrows), while the subsequent hypocotyl elongation response was biphasic (up-and-down red arrows).

Full size image

To test whether endogenous auxin works in the same way as applied auxin, we examined yuc1D, a mutant characterized by high endogenous auxin level⁷. The yuc1D seedlings displayed shorter hypocotyls in the dark compared with wild-type seedlings, and applying exogenous auxin further impeded hypocotyl growth in the dark (Supplementary Fig. 7). Compared with the wild type, the etiolated yuc1D had an elevated level of pT-AHA phosphorylation specifically in the hypocotyl, but not in the whole seedling (Fig. 4e). Correspondingly, the yuc1D hypocotyl exhibited a higher H⁺ flux rate (Fig. 4f) and a lower apoplastic pH compared with that of dark-grown wild type (Fig. 4g,h).

We used loss-of-function saur mutants to evaluate the role of SAURs in auxin-induced apoplastic acidification. The saur high-order mutants, to various degrees, showed compromised NAA-induced pT-AHA phosphorylation, H⁺ extrusion rates and apoplastic acidification (Extended Data Fig. 6). Taken together, these results demonstrate that increasing auxin dosages progressively induce SAUR expression, which subsequently causes steady activation of PM H⁺-ATPases and reduction of apoplastic pH, all in a monophasic manner, but the response of hypocotyl exhibits a biphasic pattern (Fig. 4i). This hints at the possibility that the biphasic hypocotyl response to auxin happens as a result of cellular response to the increasingly acidified apoplast.

A low-pH threshold divides the biphasic growth responses

We next investigated how apoplastic pH affects hypocotyl elongation and attempted to influence extracellular pH by submerging 2-day-old etiolated seedlings (actively undergoing hypocotyl elongation driven by endogenous auxin) in liquid MS media with different pH values. Results from time-lapse imaging measurements and conventional methods showed that lowering the medium pH from 9.8 to 5.3 enhanced hypocotyl growth (Fig. 5a,b and Supplementary Fig. 8a–d), consistent with the acid growth theory established by early studies using abraded sections of maize or oat (Avena sativa L.) coleoptiles or pea (Pisum sativum L.) epicotyls^11,37. In our assay, the optimal pH range supporting endogenous auxin-mediated hypocotyl elongation was around pH 5.3, whereas further acidification below pH 4.4 and down to pH 3.0 gradually impeded hypocotyl elongation (Fig. 5a,b and Supplementary Fig. 8e). Similar results were obtained in repeats of this assay using different pH-adjusting buffers (Supplementary Fig. 8a–d). The SAUR15 overexpression line #145, presumably with a lower in vivo apoplastic pH than the wild type, was more sensitive to acid treatment and showed growth inhibition in media with pH of 4.8 and below (Supplementary Fig. 8e). Although this assay could not pinpoint the actual apoplast pH of the hypocotyl, the data support the hypothesis that the acid growth mechanism has a low-pH limit, above which apoplastic acidification stimulates growth, but below which further acidification inhibits cell elongation (Fig. 5c). We propose that a fast-growing etiolated hypocotyl may have an apoplastic pH close to the low pH threshold and is therefore sensitive to applied auxin by falling below the threshold to the auxin-inhibiting phase (Fig. 5c).

a,b, Hypocotyl growth under varying pH treatments. Two-day-old dark-grown wild-type (Col-0) seedlings were transferred to MS medium plates adjusted to the indicated pH levels. Data are shown as means ± s.e.m. (n = 4, 9, 13, 13 and 16 for treatments ranging from pH 3.0 to pH 9.8). c, A model illustrating the biphasic relationship between hypocotyl elongation, auxin levels (green line and shading) and apoplastic pH in dark-grown hypocotyls. Higher auxin levels correlate with decreasing apoplastic pH. Young etiolated hypocotyl has an apoplastic pH near the low pH threshold (dashed line) optimal for elongation. Factors promoting PM H⁺-ATPase activity, such as auxin treatment, cause the apoplastic pH to drop below the threshold, entering the auxin-inhibiting phase. Conversely, factors inhibiting PM H⁺-ATPase activity raise the apoplastic pH, moving towards the auxin-stimulating phase. d,e, The PM H⁺-ATPase inhibitor protonstatin-1 (PS-1) partially rescued auxin-elicited repression of hypocotyls. Hypocotyl phenotypes (d) and measurements of Col-0 seedlings grown in darkness for 3 days on MS medium containing Pic and PS-1 (e). Significant differences were assessed using one-way ANOVA with Tukey’s correction (P < 0.001). Scale bar, 1 mm. f,g, Hypocotyl phenotypes (f) and measurements of Col-0 and 35S:SAUR15-GFP seedlings grown in darkness for 3 days on MS medium containing PS-1 or DMSO (g). Scale bar, 1 mm. Significant differences were assessed using two-way ANOVA with Tukey’s post test (P < 0.05). h,i, The overexpression of PP2C-D1 rescued auxin-mediated inhibition of hypocotyls. Hypocotyl phenotypes (h) and measurements of Col-0 and 35S:PP2C-D1-GFP seedlings grown in the dark for 3 days on MS medium containing 1 μM Pic or Mock (i). Scale bar, 1 mm. Significant differences were assessed using two-way ANOVA with Tukey’s post test (P < 0.0001). j,k, Time-lapse imaging and hypocotyl growth measurements under NAA treatment at pH 5.8 (j) or 8.8 (k). Etiolated Col-0 seedlings were grown on MS medium at pH 5.8 for 2 days, then transferred to MS medium plates at either pH 5.8 or pH 8.8 containing the indicated concentrations of NAA or ethanol (Mock). Data are presented as means ± s.e.m. (for pH 5.8, n = 18, 19, 21, 16, 19, 17 and 19 for treatments ranging from Mock to 50 μM NAA; for pH 8.8, n = 22, 22, 21, 33, 18, 22 and 13 for treatments ranging from Mock to 50 μM NAA).

Full size image

According to this hypothesis, high-dosage auxin or SAUR overexpression inhibits hypocotyl elongation due to over-acidification of the apoplast, which results from overactivation of the proton pump. We performed the following experiments to test whether suppressing PM H⁺-ATPase activity may alleviate auxin-mediated inhibition of the hypocotyl. Etiolated seedlings were grown on medium containing 1 μM and 5 μM Pic with or without protostatin-1 (PS-1), an inhibitor of PM H⁺-ATPases³⁸. PS-1 by itself caused shorter hypocotyls, as normal hypocotyl elongation requires PM H⁺-ATPases activity. Pic treatments inhibited hypocotyl elongation, but PS-1 (0.5 μM) completely negated the growth inhibition elicited by 1 μM Pic and significantly alleviated the inhibition induced by 5 μM Pic (Fig. 5d,e). Similarly, PS-1 negated the excessive SAUR15-induced inhibition of hypocotyls in line #102 (Fig. 5f,g), as predicted by the model illustrated in Fig. 5c.

PP2C-D1 is an endogenous natural inhibitor of PM H⁺-ATPases and its overexpression causes short hypocotyls and dwarfism due to deficient cell expansion¹³. The result showed that PP2C-D1 completely suppressed the auxin-induced inhibition of hypocotyls. While 1 μM Pic inhibited wild-type hypocotyls, it did not inhibit those of PP2C-D1 overexpressing lines but instead promoted elongation in line #6 of 35S:PP2C-D1-GFP (Fig. 5h,i). This is presumably because PP2C-D1 could counteract the effect of Pic, thereby increasing the apoplastic pH from the auxin-inhibiting phase into the auxin-promoting phase (Fig. 5c). Taken together, these results unequivocally support the hypothesis that while PM H⁺-ATPases mediate auxin-induced hypocotyl elongation, their overactivation underlies the inhibition of hypocotyl elongation by high-dosage auxin.

We tested whether a high-pH buffer may neutralize the apoplast over-acidification caused by high-dose auxin and thereby alleviate the growth inhibition, by conducting a buffer interference experiment similar to that shown in Fig. 5a,b. In normal medium of pH 5.8, 0.1 μM NAA promoted hypocotyl elongation, while all higher concentrations inhibited it (Fig. 5j and Supplementary Fig. 9a). Remarkably, in the pH 8.8 medium, the inhibition caused by 0.5 μM NAA was almost completely negated and 0.25 μM NAA became a growth-promoting concentration, similar to 0.1 μM NAA (Fig. 5k and Supplementary Fig. 9b). In fact, a sequential increase in medium pH could gradually compromise NAA-induced inhibition of etiolated hypocotyls (Extended Data Fig. 7). These data provide strong evidence supporting the hypothesis that hypocotyl inhibition by auxin overdose is due to over-acidification of the apoplast (Fig. 5c).

Light increases hypocotyl apoplastic pH, opposing auxin

Light irradiation changes the hypocotyl response to auxin (Fig. 1a–f). To carefully investigate the light effect, we grew Arabidopsis seedlings under several dim light conditions in media containing different concentrations of Pic and found that increasing the quantity of light illumination gradually turned auxin from being an inhibitor to a stimulator of hypocotyl elongation (Fig. 6a). In this experiment, the optimal concentration of Pic to enhance hypocotyl elongation shifted from 0 in the dark to 0.01 μM Pic under dim light-0.5 (0.5 μmol m⁻² s⁻¹), to 0.1 μM Pic under dim light-1 (1 μmol m⁻² s⁻¹), and greater than 10 μM Pic under moderate light (5 μmol m⁻² s⁻¹) (Fig. 6a). This result indicates that light quantity incrementally changes the response profile of hypocotyls to auxin dosage towards greater tolerance against auxin-mediated inhibition of hypocotyls (Fig. 6b).

a, Increasing light irradiation gradually elevated the maximum Pic concentrations that promote hypocotyl elongation. Seedlings were grown on MS medium containing Pic or Mock in darkness for 3 days or under the indicated light intensities for 6 days. Hypocotyls were measured and significant differences assessed using two-way ANOVA with Tukey’s multiple comparisons test. The optimal concentration of Pic for each light condition is marked with an asterisk. b, A schematic diagram illustrating how light changes the auxin-dosage response in hypocotyls. The dashed lines indicate the highest promoting dosage points under different light conditions. c,d, Light-grown hypocotyls have a higher apoplastic pH than dark-grown hypocotyls, and auxin induced apoplastic acidification in both conditions. Dark samples were grown in the dark for 3 days. Light samples were grown under 1 μmol m⁻² s⁻¹ light for 4 days, then transferred to 80 μmol m⁻² s⁻¹ light for another 4 days. Seedlings were treated with Mock or indicated concentrations of Pic for 2 h before HPTS staining (c). Scale bar, 500 μm. The ratiometric quantification (458 nm/405 nm) of HPTS staining is shown in d. Significant differences were assessed using two-way ANOVA with Tukey’s multiple comparisons test (P < 0.05, n ≥ 14). e, Light-grown hypocotyls exhibited lower phosphorylation levels of PM H⁺-ATPases at Thr947 than dark-grown hypocotyls. Hypocotyls were dissected from Col-0 treated with Mock or 50 μM Pic for 1 h for immunoblotting. The ratio of pThr947 AHA2 to AHA2 protein levels is marked below each lane. f,g, Hypocotyls of dark-grown cop1-6 and pifq display higher apoplastic pH than those of wild-type seedlings. Seedlings were grown in the dark for 2 days and used for HPTS staining (f). Scale bar, 500 μm. Statistical significance (g) was analysed using one-way ANOVA with Tukey’s correction (P < 0.05, n ≥ 32). h, A model illustrating how light and auxin antagonistically regulate extracellular pH and hypocotyl elongation. Light irradiation increases the apoplastic pH of hypocotyls, while auxin decreases it. Acidification of the apoplast below the pH threshold (dashed line) inhibits cell expansion.

Full size image

We monitored the relative apoplast pH of dark- and light-grown hypocotyls treated with 5 μM or 50 μM Pic using HPTS staining. The results showed that light-grown hypocotyls exhibited markedly higher apoplastic pH values than dark-grown hypocotyls (Fig. 6c,d). The same result was obtained using the Apo-pHusion line treated with NAA (Supplementary Fig. 10a,b). The 2-h treatments with Pic or 10 μM NAA lowered the apoplast pH of both light- and dark-grown hypocotyls, although the apoplastic pH of auxin-treated light-grown hypocotyls were not lower than those of dark-grown mock-treated hypocotyls (Fig. 6c,d). This implies that their apoplastic pH did not drop below the low-pH threshold and remained within the auxin-stimulating range (Fig. 6h). In contrast, dark-grown hypocotyls started with a lower initial apoplastic pH (Fig. 6c,d), and Pic treatment further lowered the apoplastic pH below the threshold into the auxin-inhibitory range (Fig. 5c and 6h). This may explain why the same dose of auxin tends to promote elongation in light-grown hypocotyl but inhibits it in etiolated hypocotyl (Fig. 6a,h).

Consistent with the higher apoplastic pH, light-grown hypocotyls had markedly lower pT-AHA phosphorylation and, therefore, a reduced PM H⁺-ATPase activation level compared with dark-grown hypocotyls, and Pic treatment induced pT-AHA phosphorylation in both light and dark samples (Fig. 6e). Interestingly, light has been shown to activate PM H⁺-ATPase in the stomatal guard cells of leaves^39,40, which is opposite to its effect on hypocotyls (Fig. 6e), indicating that the negative influence of light on AHA activity and the apoplastic pH is probably hypocotyl specific. This result agrees with the observation that light-grown hypocotyls intrinsically have a higher apoplastic pH than dark-grown hypocotyls (Fig. 6c,d). The higher initial apoplastic pH of light-grown hypocotyls would provide a cushion to offset auxin-induced acidification, allowing the apoplast to remain in the growth-stimulating pH range (Fig. 6h). These results also reveal that a consequence of light irradiation is the alkalization of the apoplastic pH of hypocotyls.

Light and auxin differentially regulate SAURs to impact apoplastic pH

The light- and auxin-regulated transcriptomes share extensive overlapping target genes, among which are SAUR genes^10,25,41. We reanalysed SAUR gene expression in hypocotyls under continuous dark vs light conditions and during the dark-to-light transition²⁵. Within 6 h of dark-to-light transition, the cumulative expression level of all SAUR genes in the hypocotyl decreased by more than 3-fold (Extended Data Fig. 8a). The SAUR19 and SAUR61 subfamilies, which are associated with cell expansion, collectively decreased by nearly 7- and 11-fold, respectively (Extended Data Fig. 8b). These SAUR genes are stimulated by auxin (Fig. 2a and Extended Data Fig. 2a,b) and repressed by light in the hypocotyl. Upon the dark to light transition, turning off these SAURs could presumably serve as a brake on PM H⁺-ATPase-mediated proton extrusion and stop further acidification of the apoplast, which would help raise the apoplastic pH and decelerate hypocotyl elongation during de-etiolation.

Phytochrome interacting factors (PIFs) are key transcription factors directly activating SAUR expression in darkness, and the pifq (pif1 pif3 pif4 pif5) mutant displays a massive decline in overall SAUR transcript levels in the hypocotyl²⁵. We determined the apoplastic pH in the hypocotyls of dark-grown pifq as well as cop1-6, in which PIFs are unstable. The results showed that dark-grown pifq and cop1-6 had higher hypocotyl apoplastic pH values than dark-grown wild type (Fig. 6f,g). We suggest that the elevated apoplastic pH of pifq and possibly cop1-6 may contribute to their short hypocotyl phenotype in the dark.

The hypocotyl transcriptomes of seedlings grown under continuous dark and light conditions revealed that cumulative SAUR expression levels in hypocotyls were generally lower in the light than in the dark, particularly SAUR9–12 and most members of the SAUR19 subfamily (Extended Data Fig. 8c–f). When SAUR50 or SAUR66 was overexpressed, compensating for the reduction in SAUR levels in the light, longer hypocotyls with a significant attenuation in auxin-induced growth were observed (Extended Data Fig. 3). The SAUR61 subfamily, along with SAUR15 and SAUR29, showed higher transcript levels in light-grown hypocotyls (Extended Data Fig. 8e), indicating that these genes may have important functions in light-grown plants. Apart from the SAURs, other genes were antagonistically regulated by light and auxin in the hypocotyls of continuously light-grown plants (Supplementary Table 7), which may also contribute to the light–auxin antagonism in the control of hypocotyl growth.

Auxin promotes or inhibits hypocotyl/coleoptile elongation depending on its dosage and light conditions. This biphasic effect (or dual effect) of auxin has been known for nearly a century but has not been clearly explained. In this study, we provide evidence showing that how hypocotyls respond to auxin primarily depends on the apoplastic pH at the cell wall, although other factors may also be involved. We demonstrate that, regardless of the concentrations or light conditions, auxin always induces acidification of the apoplast in the hypocotyl, but the cellular response to increasing acidity is biphasic. Hypocotyl cells normally elongate in response to acidification through the acid growth mechanism, except when the apoplastic pH falls below a critical low-pH threshold, under which further auxin-induced acidification impedes hypocotyl elongation. We propose that the biphasic response to auxin is essentially a result of the cellular response to decreasing apoplastic pH driven by auxin, predominantly through the SAUR-PP2C.D-AHA pathway, and that the turning point between growth-promoting and growth-inhibiting phases corresponds to the low-pH threshold of the acid growth mechanism (Fig. 5c).

The data showed that the hypocotyl reacts to the cumulative total of auxin dosage, which takes into account both auxin concentration and the duration of the exposure (Fig. 1e,f,i,j). This action profile seems to match the way apoplastic pH changes, as it reflects the collective result of the rate of H⁺ efflux and the duration of H⁺ extrusion. Nonetheless, we noticed a subtle difference in the hypocotyl response to auxin and to FC. High concentrations of FC induced a more pronounced initial growth than did high concentration of auxin before inhibiting growth (Figs. 1g–j and 3h,i). This could indicate that auxin is more effective at triggering the inhibitory response of hypocotyls than the simple activation of AHAs by FC. It is possible that auxin may affect other process(es) in addition to AHA activation, which facilitate hypocotyl growth inhibition. For example, high concentrations of auxin have been shown to trigger the cleavage of the TMK1 C-terminal fragment, which translocates into the nucleus to interact with and phosphorylate IAA32/34 (ref. ⁴²). The model proposed here does not exclude the involvement of TMK cleavage and the subsequent events it triggers. It is also possible that auxin-induced SAURs may have other functions apart from regulating AHA activities, or that auxin may utilize different output effectors to control cell elongation. Nevertheless, we suggest that apoplastic over-acidification is a primary mechanism of auxin-induced hypocotyl inhibition on the basis of the following lines of evidence. First, high auxin dosage-induced inhibition of hypocotyl elongation can be significantly rescued by means that prevent the acidification process, such as AHA inhibitor PS-1 or overexpression of the endogenous AHA inhibitor PP2C-D1 (Fig. 5d,e,h,i). Second, increasing ambient pH of the apoplast significantly suppressed the ability of high auxin concentrations to inhibit hypocotyl elongation (Fig. 5j,k, Extended Data Fig. 7 and Supplementary Fig. 9). Third, the specific concentration of auxin that triggers the inhibition response is undefined, but rather depends on light conditions of hypocotyl growth and possibly other factors. For example, 1 μM IAA/NAA/Pic is considered a ‘high concentration’ for etiolated hypocotyls, because it triggers growth inhibition; but the same 1 μM is not considered a high concentration for light-grown hypocotyls, as it promotes elongation. This variability would argue against auxin concentration itself being the switch for the biphasic change of hypocotyl growth. On the other hand, our model on apoplastic pH as a key switch of biphasic responses, influenced by auxin and light in opposite directions, can well explain this phenomenon (Fig. 6h).

The revelation of the detrimental consequence of persistent uncontrolled activation of auxin sheds light on why the auxin nuclear pathway has built-in layers of negative feedback control mechanisms to prevent overstimulation of the system. Along with SAUR, auxin rapidly induces AUX/IAA and GH3 genes²⁰. AUX/IAA proteins act as repressors to block auxin-mediated transcription once auxin subsides; and GH3s mediate the removal of free auxin by conjugating excess auxins to amino acids²⁰. In addition, PP2C-D1, D2 and D7, which inhibit PM H⁺-ATPases, are transcriptionally co-induced with SAURs by auxin (Extended Data Figs. 1e and 2f). We speculate that they might serve to immediately switch off the proton pumps and terminate the acidification process as soon as SAURs decline. Pertinent to this, SAURs are innately short lived at both messenger RNA and protein levels²¹, which would help to maintain sensitivity to auxin and avoid overdosage. A case demonstrating the devastating consequence of auxin overdosage may be manifested by synthetic auxinic herbicides such as 2,4-D, which plants cannot metabolize unlike IAA⁴³. It is plausible that the herbicidal activity of 2,4-D is caused by uncontrolled, persistent stimulation of auxin signalling, leading to over-acidification of the extracellular matrix in the hypocotyl, among other issues.

At present, how overly acidic extracellular conditions cause growth inhibition remains unsolved. We found that when the medium pH drops below 4.4, hypocotyl elongation is inhibited within 30 min of treating etiolated Arabidopsis seedlings (Fig. 5a,b and Supplementary Fig. 8). Acid-induced growth involves the activation of expansins and other enzymes necessary for loosening and remodelling the cell wall^14,16,44. In addition, co-activation of inward rectifying K⁺ channels is necessary for the uptake of water to increase the turgor pressure¹⁷. Low pH limits in acid-induced growth had in fact been reported in early studies. Using Avena coleoptile segments, it was observed that cell growth reached its optima at pH 3.0, then dropped below pH 2.6 (ref. ¹¹). Similar findings were reported in experiments using sections of auxin-starved hypocotyls of Helianthus, where pH 3.5 was found to be less effective than pH 4.0 for cell growth¹². Evidence suggests that isolated hypocotyl cell wall materials from various plants show optimal extension at or below pH 4.0 and seem to have a high tolerance for acidic conditions^11,44. Acidification may also trigger pH-dependent changes in the biochemical or physical-chemical properties of pectin and other components of the cell wall^14,44. Obviously, a comprehensive understanding of these issues requires further experimentation.

A hallmark of light activity is the inhibition of hypocotyl and stem elongation. The short hypocotyl phenotype caused by light irradiation is a natural process, in contrast to the short hypocotyl due to auxin overdosage. Our data support the notion that light and auxin antagonistically regulate the apoplastic pH of hypocotyl cells, with auxin acting to decrease and light acting to increase the apoplastic pH (Fig. 6h). By raising the apoplastic pH of the hypocotyl, light would inhibit its elongation according to the acid growth theory¹⁴. Supporting this idea, compared with etiolated hypocotyls, light-grown seedlings or dark-grown pifq and cop1-6 mutants, which exhibit shorter hypocotyls, have relatively higher apoplastic pH values in their hypocotyl epidermis compared with dark-grown wild-type seedlings (Fig. 6c–g). Certainly, cell wall composition and structure change as plants de-etiolate and would also affect elongation. It should be emphasized that this model is based on data for hypocotyl elongation but is unfit for roots and leaves. In roots, auxin induces proton influx and apoplastic alkalization¹⁸, opposite to its action in hypocotyls. Although light irradiation inhibits the elongation of hypocotyls and stems, light is not associated with the inhibition of leaf expansion or root growth, but quite the opposite. Also, the asymmetric cell growth of the apical hook is associated with asymmetric inactivation of PM H⁺-ATPase at the concave side by PP2C-D1 (ref. ⁴⁵), and simultaneous activation at the convex side⁴⁵ probably through SAUR proteins such as SAUR57 (ref. ³²), which does not involve auxin overactivation. Therefore, there is a clear organ specificity in the proposed model of how light and auxin regulate cell expansion.

Accumulating studies have shown that light antagonizes auxin signalling at multiple levels, ranging from modulating auxin homeostasis to photoreceptor interference of AUX/IAA protein degradation, and the functional interaction of PIF and ARF transcription factors^10,41,46. In fact, many SAUR genes are direct transcription targets of PIFs and ARF6 in controlling hypocotyl elongation^25,41. Through these interplays, we suggest that light downregulates the same SAUR-PP2C.D-AHA pathway, which contributes to the increase in the apoplastic pH of hypocotyls. Our study highlights the cell wall as a critical location for converging light and auxin signalling activities in regulating plant elongation growth.

Plant materials and growth conditions

The yuc1D and saur6 saur12 saur14 saur16 saur50 (saur6,12,14,16,50) mutants were described previously^7,32. All plants used in this study were of the Columbia-0 (Col-0) ecotype. The seeds were surface sterilized using 15% NaClO for 5–10 min and washed 3–5 times with sterile distilled water before being sown on Murashige and Skoog basal salt medium (4.4 g l⁻¹, pH 5.8) supplemented with 1% (w/v) sucrose and 0.6% (w/v) agar. The seeds were kept at 4 °C for 4–5 days of stratification and then exposed to 80 μmol m⁻² s⁻¹ white light for 6–12 h to promote germination before either being transferred to darkness or growing under light for phenotype analysis. To dissect light-grown hypocotyls, the seeds after stratification were grown under 0.5 μmol m⁻² s⁻¹ white light for 4 days and then transferred to 80 μmol m⁻² s⁻¹ white light for another 4 days.

Plasmid construction and generation of transgenic lines

To generate transgenic plants, 35S:SAUR15-GFP, 35S:SAUR63-GFP, 35S:SAUR66-GFP, SAUR15, SAUR63 and SAUR66 were amplified, cloned and inserted into the Xba I and Xho I restriction sites of the pJim19-GFP (Barsta) vector²⁵. To construct 35S:SAUR16 and 35S:SAUR50, SAUR16 and SAUR50 were amplified from Arabidopsis genomic DNA, cloned and inserted into the pJim19 (Basta) vector⁴⁷ using Xba I and Sac I. To generate 35S:SAUR19-GFP, the construct from a previous study²⁸ was used. To generate SAUR15 Pro:SAUR15-GFP, the SAUR15 3’ UTR was cloned from genomic DNA using primers SAUR15 3’ UTR F_SpeI and SAUR15 3’ UTR R_SacI. It was then inserted into pJim19-GFP (Basta), which had been digested with Spe I and Sac I, resulting in the generation of pJim19-GFP-SAUR15 3’ UTR. Subsequently, the genomic fragment of SAUR15 without the stop codon was amplified using primers SAUR15 Promoter_ SbfI F and SAUR15 CDS without SC_XbaI R. This fragment was cloned into the Sbf I and Xba I restriction sites of pJim19-GFP-SAUR15 3’ UTR (Basta), yielding pJim19-SAUR15 Pro:SAUR15-GFP-SAUR15 3’ UTR (Basta). The primers used to clone SAURs are listed in Supplementary Table 8. To generate saur19–24 saur26 saur27 (saur19^Octuple), saur7 saur13 saur15 saur19 saur20 saur22 saur24–29 saur73 (saur19^Tredecuple) and saur61–68 saur75 (saur61^Nonuple) mutants, the novel CRISPR/Cas9 gene editing system for efficiently generating multiple mutants in Arabidopsis³¹ was used. The construction of these CRISPR/Cas9 vectors was performed as described previously. Briefly, small guide RNA (sgRNA) was inserted into the Bbs I restriction site of pAtU6-M and multiple sgRNA cassettes were then digested and ligated using Spe I/Sal I and Nhe I/Sal I. The tandem sgRNA cassette was subsequently digested, cloned and inserted into pUBQ10:Cas9-P2A-GFP (Gent) or pUBQ10:Cas9-P2A-GFP (Hyg) using Kpn I and Sal I. All the constructs were transformed into Arabidopsis by the floral dip method using Agrobacterium GV3101.

RNA-seq experiments and analysis

For the dark vs light auxin response set, light-grown seedlings were initially grown under dim light (1 μmol m⁻² s⁻¹) for 4 days, then transferred to 80 μmol m⁻² s⁻¹ white light for another 4 days. For dark-grown samples, cold-stratified seeds were first exposed to 80 μmol m⁻² s⁻¹ light for 12 h and then transferred to darkness for 3.5 days. These seedlings were collected and submerged in MS liquid medium supplied with DMSO (Mock) or the indicated concentrations of picloram. At the beginning of the treatment, the seedlings were vacuumed for 10 min and then returned to normal air pressure for another 35 min. Subsequently, hypocotyls of these treated seedlings were dissected and collected.

For RNA sequencing of 3 and 12 h picloram or IAA treatments, after 12 h of 80 μmol m⁻² s⁻¹ light exposure, wild-type (Col-0) seeds were transferred to darkness and grown for 3 days. The seedlings were then submerged in liquid MS medium containing specified concentrations of picloram, equimolar amounts of DMSO (Mock) or IAA, as well as equimolar amounts of ethanol (Mock), for 3 and 12 h, respectively. Hypocotyls were dissected and collected using microsurgical scissors and tweezers. Total RNA was extracted from the hypocotyls using a TaKaRa MiniBEST Plant RNA Extraction kit (Takara, 9769). Three or four biological replicates of each treatment group were prepared.

RNA-seq was performed on the Illumina HiSeq 4000 platform. A total of 125-bp paired-end sequencing reads were mapped to the Arabidopsis reference genome (TAIR10) using HISAT2, and reads mapped within exons were counted using HTSeq. DEGs were calculated using the DESeq2 package in R. A 1.5-fold change [log₂(fold change) = 0.585] (P_adj < 0.1) was set as the threshold for the identification of DEGs. Venn diagrams were generated online using Venny 2.1.0 (https://bioinfogp.cnb.csic.es/tools/venny/index.html). Gene Ontology (GO) annotation was performed using the R software package. The transcriptomic data have been deposited to the NCBI database under the dataset identifier PRJNA1054223.

RT–qPCR

Total RNA was extracted using a TaKaRa MiniBEST Plant RNA Extraction kit (Takara, 9769). The reverse transcription assay (YEASEN, 11141ES60) was conducted using 2 µg of total RNA. RT–qPCR was performed on a 7500 Fast Real-Time PCR System (Applied Biosystems) using Hieff qPCR SYBR Green Master Mix (Low Rox Plus) (YEASEN, 11202ES08). The gene-specific RT‒qPCR primers are listed in Supplementary Table 8. The relative gene expression values shown are the averages of at least three biological replicates, and PP2A was used as a loading control.

H⁺ flux measurement

The net H⁺ flux was measured using an NMT system (NMT100-SIM-XY; YoungerUSA, Xuyue Sci & Tech). As illustrated in Fig. 3b, the testing area comprises ~4–5 layers of epidermal cells (spanning ~200–300 μm) within one-eighth of the length of the hypocotyl below the apical hook. Two-day-old dark-grown seedlings were subjected to treatment with specific chemicals, or etiolated seedlings of the indicated genotype were incubated in the test solution (0.1 mM CaCl₂, 0.1 mM KCl, 0.3 mM HEPES pH 7.5) for 30 min. The microsensor was calibrated using calibration solutions of pH 7.0 and 8.0 in a liquid containing 0.1 mM CaCl₂, 0.1 mM KCl and 0.3 mM HEPES. Microsensor assembly: Initially, ~1 cm of H⁺ filling solution (15 mM NaCl, 40 mM KH₂PO4 pH 7.0) was introduced into the capillary-like hollow sensor (XY-CGQ-02, Xuyue Sci & Tech). Subsequently, a 50 μm long H⁺ LIX Holder (XY-SJ-H-10, Xuyue Sci & Tech) was drawn in and an AgCl/Ag electrode was then inserted. The net H⁺ flux was determined by quantifying the disparity in H⁺ concentration between two positions: one adjacent to the hypocotyl epidermal cells and the other situated 30 μm away from the epidermal cells.

Time‐lapse imaging and quantitative measurement of hypocotyl growth

The kinetic analysis of hypocotyl growth was conducted using a commercial seedling phenotyping platform, DynaPlant Seedlings (Microlens Technologies). Two-day-old dark-grown wild-type (Col-0) seedlings were transferred to fresh MS medium containing 1.2% agar and supplemented with the specified chemicals and/or adjusted to various pH levels. Subsequently, the seedlings were placed on the surface of the MS medium and the culture plates were positioned vertically for cultivation in darkness. Photographs were taken for each seedling at intervals of 6 min for FC treatment, 9 min for pic and IAA treatments, and 15 min for NAA and different pH treatments. Data analysis was performed using DynaPlant Analysis software (v.2.2.1), as provided by the manufacturer. The new growth length (L_new) was calculated as the difference between the length at a specific moment (L_t) and the initial length (L₀). In addition, the growth rate (R) was determined using the formula R = (L_t2−L_t1)/(t₂−t₁), where L_t represents the length at time t. The curve smoothing option was set to 3, indicating a moving average of the growth rate within a total of 3 measurement cycles before and after the time specified by the data points. ‘t₀’ represents the time of seedling transfer to the new plate and the start of the imaging programme. Except for Fig. 5a and Supplementary Fig. 8a,c, where t₀ = 1 h, all other t₀ values are 0.5 h.

Microscopic observation and analyses

For confocal microscopy analysis, dark-grown 35S:SAUR15-GFP or apo-pHusion⁴⁸ seedlings treated with the indicated concentrations of picloram or NAA were used to acquire fluorescence images using a confocal laser scanning microscope (Zeiss LSM710). The fluorescence of SAUR15-GFP was excited with a 488 nm laser and detected within the range of 493–598 nm using ZEN 2012 software. For apo-pHusion, EGFP fluorescence was excited at 488 nm and detected between 493 nm and 570 nm, while mRFP1 fluorescence was excited at 543 nm and detected between 582 nm and 654 nm. Additional details on confocal images can be found in the HPTS staining section. Signal intensity was quantified using ImageJ (https://imagej.net/software/fiji/).

Protein extraction and immunoblotting

To analyse the phosphorylation levels of H⁺-ATPase, hypocotyls were dissected from whole seedlings for protein extraction. Hypocotyl samples were ground in liquid nitrogen and the powder was suspended using protein extraction buffer (25 mM Tris-HCl pH 7.5, 150 mM NaCl, 1 mM EDTA, 1 mM dithiothreitol, 1 mM phenylmethylsulfonyl fluoride, 1× Roche PhosSTOP cocktail, 1× Roche protease inhibitor cocktail, 1% Triton X-100). The extracts were incubated on ice for 30 min and then centrifuged at 10,000 × g for 10 min at 4 °C to discard the precipitate. The supernatant was transferred into a new centrifugal tube containing SDS loading buffer at room temperature and used for SDS‒PAGE analysis. The prepared protein samples were not boiled before being loaded onto 8% (w/v) SDS‒PAGE gels to separate and evaluate the phosphorylation levels of penultimate threonine H⁺-ATPase. The anti-AHA2 (1:5,000 dilution) and anti-pThr947 AHA2 antibodies (1:3,000 dilution) were shared by T.K. and were used as described previously³³. Anti-RPN6 antibody (1:5,000 dilution, lab stock) and anti-HSP90 antibody (1:4,000 dilution, AS11 1629, Agrisera) were used as a loading control.

HPTS staining and imaging

HPTS staining and confocal imaging were performed according to ref. ³⁶ with some modifications. HPTS was dissolved in double-distilled water to obtain 100 mM stock solution and frozen at −20 °C. Three-day-old dark-grown seedlings were submerged in liquid MS medium containing 1 mM HPTS with 0.01% Triton X-100 under 0.02 MPa vacuum for 5 min¹⁵ and then incubated for 30 min under normal atmospheric conditions. Seedling imaging was performed using a Zeiss 710 confocal microscope. Cell wall fluorescence signals for the protonated HPTS form (excitation 405 nm, emission peak 514 nm) and the deprotonated HPTS form (excitation 458 nm, emission peak 514 nm) were obtained. The fluorescence images from multiple actively elongating hypocotyl cell populations were recorded for statistical analysis.

Hypocotyl length measurements

Except for the time-lapse imaging and quantitative measurement assay of hypocotyl growth, where hypocotyl lengths were analysed using DynaPlant Analysis software, all other measurements of hypocotyl lengths were conducted as follows: seedlings were treated with the indicated chemicals or pH and then placed onto 1% (w/v) agar plates for imaging using an Epson perfection V850 Pro scanner. The hypocotyl length was measured using ImageJ software.

Chemicals and treatments

Picloram (Sigma-Aldrich, P5575), fusicoccin (ChemCruz, 200754), PS-1 (provided by Dr Xiaoguang Lei, PKU), Yucasin⁴⁹ (provided by Dr Tomokazu Koshiba, TMU) and l–kynurenine⁵⁰ (Sigma-Aldrich, K8625) were dissolved in DMSO to create stock solutions and stored at −20 °C. IAA (Sigma-Aldrich, I3750) was dissolved in anhydrous ethanol to prepare stock solutions, which were also stored at −20 °C. The stock solutions were prepared with the following concentrations: picloram (100 mM), fusicoccin (5 mM), PS-1 (100 mM), yucasin (100 mM), l–kynurenine (100 mM) and IAA (100 mM). The working solution was diluted from the stock solution. For hypocotyl length analysis, except for Fig. 1e,f, Extended Data Figs. 4e and 7 and Supplementary Fig. 7d,e where auxin or fusicoccin were added to liquid MS and seedlings were immersed in the liquid MS for treatment, all other hypocotyl length analyses involved adding specified chemicals to solid MS medium, and seedlings were grown on media containing the specified chemicals. For RNA-seq, H⁺ flux measurement and HPTS staining, specific chemicals (picloram, fusicoccin, IAA, NAA) were added to liquid MS medium (pH 5.8) and the seedlings were immersed in this liquid MS medium for the designated treatment.

Different pH media and treatments

The liquid MS medium was prepared by dissolving 4.4 g l⁻¹ of Murashige and Skoog Basal Medium powder (Sigma-Aldrich, M5519) and adding 1% (w/v) sucrose. The pH meter (Mettler Toledo, EL20) was calibrated using standard buffer solutions of pH 4.01 (Mettler Toledo, 51350004), pH 7.00 (Mettler Toledo, 51350006) and pH 9.21 (Mettler Toledo, 51350008), and then used to monitor the pH of the liquid MS medium. The initial pH of the liquid MS medium was ~3.95. Unless otherwise specified, the liquid MS and MS medium plates used in this study were adjusted to pH 5.8 using potassium hydroxide. For Fig. 5a,b and Supplementary Fig. 8e, the pH of the MS medium was adjusted to the desired value using potassium hydroxide or hydrochloric acid. For Supplementary Fig. 8e, seedlings were submerged in the specified pH of liquid MS medium and incubated for 24 h in darkness. Afterwards, the seedlings were removed from the medium and subjected to further analysis. For Supplementary Fig. 8a–d, 1 mM citrate and 2 mM Na₂HPO₄ were added to the MS medium, or 2 mM glycine was added. The pH was adjusted using either sodium hydroxide or hydrochloric acid to the desired value.

Statistics and reproducibility

For each boxplot, the central line represents the median, the bottom and top edges of the box correspond to the 25th and 75th percentiles, respectively, and the whiskers extend to the highest and lowest data points within the group. Different lowercase letters above the bars indicate statistically significant differences between groups. Statistical data are provided in Supplementary Table 9. The statistics for hypocotyl length and fluorescence ratio both denote n as the number of seedlings used in the analysis. All western blot experiments were independently repeated at least twice, yielding consistent results.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.