CPR5-mediated nucleo-cytoplasmic localization of IAA12 and IAA19 controls lateral root development during abiotic stress

Significance

Plants have evolved numerous strategies to compromise a variety of harsh environments into the development for survival. Plasticity of lateral root development is crucial for plants to grow under abiotic stress conditions. Auxin is the major phytohormone controlling lateral root development. In this study, a novel mechanism is revealed underlying abiotic stress-mediated suppression of auxin signaling in lateral root development. Abiotic stresses induce nuclear localization of auxin signaling repressors by a suppression of their cytoplasmic export, thereby inhibiting auxin signaling and the development of lateral roots.

Abstract

Plasticity of the root system architecture (RSA) is essential in enabling plants to cope with various environmental stresses and is mainly controlled by the phytohormone auxin. Lateral root development is a major determinant of RSA. Abiotic stresses reduce auxin signaling output, inhibiting lateral root development; however, how abiotic stress translates into a lower auxin signaling output is not fully understood. Here, we show that the nucleo-cytoplasmic distribution of the negative regulators of auxin signaling AUXIN/INDOLE-3-ACETIC ACID INDUCIBLE 12 (AUX/IAA12 or IAA12) and IAA19 determines lateral root development under various abiotic stress conditions. The cytoplasmic localization of IAA12 and IAA19 in the root elongation zone enforces auxin signaling output, allowing lateral root development. Among components of the nuclear pore complex, we show that CONSTITUTIVE EXPRESSOR OF PATHOGENESIS-RELATED GENES 5 (CPR5) selectively mediates the cytoplasmic translocation of IAA12/19. Under abiotic stress conditions, CPR5 expression is strongly decreased, resulting in the accumulation of nucleus-localized IAA12/19 in the root elongation zone and the suppression of lateral root development, which is reiterated in the cpr5 mutant. This study reveals a regulatory mechanism for auxin signaling whereby the spatial distribution of AUX/IAA regulators is critical for lateral root development, especially in fluctuating environmental conditions.

Plants are subject to various environmental stresses that affect their physiology and development throughout their life cycle. To mitigate the consequences of these adverse growth conditions, plants have accordingly evolved protective mechanisms to withstand stress and ensure their normal growth and development (1–4). The plasticity of the root system architecture, which dictates the spatiotemporal properties of roots, is critical to support optimal growth and development, enabling plants to cope with environmental changes. Notably, the development of lateral roots is a key factor shaping the architecture of the root system, as it determines the overall root volume and surface area and directs the shape and direction of roots to optimize their foraging of crucial resources such as water and nutrients (3–6). Under stress conditions, plants generally adjust the development of lateral roots to avoid less favorable regions. For example, Arabidopsis thaliana roots can distinguish a wet patch of soil from a dry area and will develop lateral roots preferentially into areas with higher water contents (6–8). By contrast, plants will block lateral root formation in their aerial parts or when roots are not in contact with water (9, 10). Drought and osmotic stresses regulate stomatal movement, and salt stress disrupts ion homeostasis (11–16), which in turn is the outcome of differential gene expression in each stress conditions (17). Intriguingly, these stresses alter the development of lateral roots in a similar manner (2, 6, 7, 18–20). This led us to question if there is a common pathway operating under abiotic stresses to regulate lateral root development. Developmental plasticity of lateral roots is essential for plant survival and crop yield under abiotic stress conditions; importantly, this plasticity is mainly mediated by auxin (4–6, 21).

Indeed, auxin is a critical factor in lateral root formation, from priming to initiation, patterning, and emergence (5, 22–24). Decreasing auxin responses by inhibiting polar auxin transport block lateral root development, while exogenous application of auxin has the opposite effect, indicating that regulating the auxin response is both necessary and sufficient for the development of lateral roots (25). In addition, mutants in auxin signaling regulators show defects in lateral root development; the arf7/19 double mutant, which lacks AUXIN RESPONSE FACTOR7 (ARF7) and ARF19, does not develop lateral roots, whereas arf5 mutants and gain-of-function mutants of IAA12, IAA14, and IAA19 have severely compromised lateral root development (26–29). Lateral root priming, which occurs at the basal meristem and the elongation zone, refers to signaling events that involve the regulation of auxin responses or gene expression, leading to the specification of cells for pre-branch sites in lateral root formation (22, 23, 30, 31). After priming, several auxin signaling regulators (ARFs and IAAs) participate in the transcriptional control of lateral root formation-related genes (5, 28, 30, 32). Under stress conditions, plants reduce their lateral root development to adapt to their new environments (2, 19, 33). Although many studies have explored the mechanism of lateral root development, few have focused on the underlying mechanism that dictates how lateral root development is altered during abiotic stress.

Abiotic stress elevates the levels of reactive oxygen species (34, 35), which in turn repress the expression of the auxin transporter genes PIN-FORMED 1 (PIN1), PIN2, and PIN3 or the polarity of their encoded proteins (36–38). Furthermore, various stress conditions raise the transcript levels of the microRNAs miR393 and miRNA160, thereby attenuating the transcript levels of the auxin receptor genes TRANSPORT INHIBITOR RESPONSE 1 (TIR1), AUXIN SIGNALING F-BOX 2 (AFB2), and AFB3 (39, 40) and of ARF genes (40, 41), respectively. In water-deficit conditions, the SUMOylation of ARF7 also contributes to the inhibition of lateral root development by enhancing the activity of its associated repressor, IAA3 (8). Although these studies have demonstrated that abiotic stress inhibits auxin signaling during lateral root development at the transcriptional and posttranslational levels, a direct regulatory link between stress and regulators of auxin signaling remained to be firmly established. In this study, we show that the nucleo-cytoplasmic distribution of IAA12 and IAA19 is mediated by CONSTITUTIVE EXPRESSOR OF PATHOGENESIS-RELATED GENES 5 (CPR5) and is critical to modulate auxin signaling output and redirect spatiotemporal properties of lateral root development in response to stress.

Results

Abiotic Stresses Induce Nuclear Translocation of IAA12 and IAA19 at the Root Elongation Zone, Inhibiting Lateral Root Development.

To elucidate the spatiotemporal characteristics of root development under fluctuating environmental conditions, we observed lateral root development under various stresses; polyethylene glycol (PEG), mannitol, and NaCl are mainly used for drought stress, osmotic stress, and salt stress to seedlings, respectively (6, 42–46). We determined that the total number of lateral roots is much lower after treatment of Col-0 seedlings with PEG, mannitol, or NaCl compared to that in mock-treated seedlings (Fig. 1 A and B). Exposure to high concentrations of salt resulted in shorter primary roots, as did mannitol treatment to a lesser extent; however, PEG treatment did not show any significant change in elongation of the primary root. In addition, the density of lateral roots was substantially lowered after PEG, or mannitol treatments, but not after salt treatment. These observations indicated that various abiotic stresses can drastically affect root architecture. We then examined the output of auxin signaling, which is a major phytohormonal regulator of lateral root development (5, 22–24), under stress conditions. To this end, we characterized the fluorescence intensity of green fluorescent protein (GFP) when expressed from the auxin-responsive DR5v2:NLS-GFP transgene in the wild-type background. After treatment with PEG, mannitol, or NaCl, the relative GFP fluorescence in DR5v2:NLS-GFP transgenic lines was much weaker than in mock-treated seedlings, especially in the root elongation zone where lateral root development is initiated (30, 31) (Fig. 1C and SI Appendix, Fig. S1A).

Fig. 1.

Abiotic stresses inhibit lateral root development and induce the nuclear accumulation of IAA12 at the root elongation zone. (A) Inhibition of lateral root development of 10-d-old wild-type (Col-0) seedlings in various abiotic stress conditions. Col-0 seedlings were grown on half-strength MS medium containing 30% (w/v) PEG, 200 mM mannitol, or 100 mM NaCl for 3 d. (Scale bar, 1 cm.) (B) Lateral root number, primary root length, and lateral root density of 10-d-old Col-0 seedlings were grown under mock conditions or treated with PEG, mannitol, or NaCl (n = 15). Bar graphs show means ± SD with individual data points. Different lowercase letters indicate significant differences (P < 0.05, determined by one-way ANOVA with post hoc Tukey’s honest significant difference [HSD] test). (C and D) Abiotic stresses repress the auxin responses (C) and change the nucleo-cytoplasmic distribution of IAA12 (D) in the root elongation zone. Representative confocal images show the GFP fluorescence for DR5v2:NLS-GFP (C) and IAA12pro:gIAA12-GFP (D) transgenic lines in the roots of 8-d-old seedlings after mock, PEG, mannitol, or NaCl treatment. Roots were stained with propidium iodide (PI). Enlarged images show the fluorescence in the elongation zone of DR5v2:NLS-GFP 35S:H2A-mCherry (C) and IAA12pro:gIAA12-GFP 35S:H2A-mCherry (D) roots. H2A-mCherry was used as a nuclear marker. Red arrows indicate nuclear-localized IAA12-GFP. (Scale bars, 100 μm.) (E) Abiotic stresses induce the nuclear accumulation of IAA12 at the early stage of lateral root development. Confocal images represent GFP and mCherry fluorescence for IAA12pro:gIAA12-GFP 35S:H2A-mCherry transgenic seedlings at the lateral root initiation site after mock treatment and the indicated stress treatments. Red arrows indicate lateral root initiation cells. (Scale bar, 100 μm.) (F) Partitioning of IAA12 between nuclear and cytosolic fractions using the elongation zone of IAA12pro:gIAA12-GFP seedlings subjected to mock, PEG, mannitol, or NaCl treatment. T, 5% of total; C, cytoplasmic fraction; N, nuclear fraction. Protein abundance was determined using α-GFP antibody for IAA12. UGPase was used as a cytoplasmic marker and H3 as a nuclear marker.

Based on the above results, we hypothesized that the decrease in auxin signaling output might be directly related to the lateral root development impaired under abiotic stress. To explore this possibility, we examined how auxin signaling components regulating lateral root development respond to stress. Among ARFs and AUX/IAAs, ARF7, IAA12, and IAA19 play major roles in root and lateral root development; in addition, their encoding genes have a broad expression pattern in the root (26–28, 47, 48). We therefore generated transgenic lines harboring ARF7pro:gARF7-GFP, IAA12pro:gIAA12-GFP, IAA19pro:gIAA19-GFP, IAA12pro:gIAA12-Venus, IAA19pro:gIAA19-Venus, IAA12pro:gIAA12-HA, and IAA19pro:gIAA19-HA constructs, whereby each ARF or IAA genomic coding region is driven by their respective promoter and is expressed as an in-frame fusion with GFP. We detected ARF7-GFP fluorescence in the nucleus at the meristem region and observed the formation of condensates in the cytoplasm of cells at the root elongation zone, as previously reported (49). Notably, ARF7-GFP distribution did not change after PEG, mannitol, or NaCl treatment (SI Appendix, Fig. S1B). IAA12-GFP also localized to the nucleus in the meristem region but, unexpectedly, accumulated mainly in the cytoplasm in cells at the root elongation zone under normal growth conditions (Fig. 1D). Under normal conditions, IAA19-GFP also localized to the nucleus in the meristem region and to both the nucleus and the cytoplasm in cells at the root elongation zone (SI Appendix, Fig. S1C). Venus-tagged IAA12 and IAA19 exhibited the same subcellular localization pattern with GFP-tagged IAA12 and IAA19 (SI Appendix, Fig. S1F). However, treatment with PEG, mannitol, or NaCl resulted in the translocation of IAA12-GFP and IAA19-GFP to the nucleus at the root elongation zone (Fig. 1D and SI Appendix, Fig. S1C).

During the early stage of lateral root initiation, IAA12-GFP accumulated in the cytoplasm, but it was mainly localized to the nucleus after PEG, mannitol, or NaCl treatment (Fig. 1E). To better quantify the relative distribution of IAA12 and IAA19 between the nucleus and cytoplasm, we performed fractionation assays on the elongation zone of the transgenic lines that were mock-treated or exposed to each individual stress. Although IAA12-GFP, IAA19-GFP, IAA12-HA, and IAA19-HA were largely present in the cytoplasm under normal conditions, exposure to PEG, mannitol, or NaCl resulted in all proteins becoming localized mainly in the nucleus (Fig. 1F and SI Appendix, Fig. S1 D and E). We concluded from these observations that the abiotic stress-induced nuclear translocation of IAA12/19 suppresses the auxin responses required for lateral root development.

CPR5, A Component of the NPC, Selectively Induces the Translocation of AUX/IAA Proteins.

The nuclear pore complex (NPC) regulates the permeability of the nuclear pore and modulates auxin signaling (50–53), prompting us to examine whether the NPC is involved in the translocation of AUX/IAA proteins to control auxin responses. Accordingly, we tested five previously identified nucleoporin (NUP) genes that have mutant RNA sequence data or whose loss-of-function mutants were known to have altered auxin-responsive gene expression: MOS3 (MODIFIER OF SNC1 3), HOS1 (HIGH EXPRESSION OF OSMOTICALLY RESPONSIVE GENES 1), NUP75, ALADIN, and CPR5 (54–57). We cloned each NUP as an effector construct and co-transfected them individually into protoplasts along with either the auxin-responsive GH3pro:LUC reporter (with the GRETCHEN HAGEN 3 promoter driving the transcription of the firefly luciferase [LUC] reporter gene) or IAA12-GFP. Of the five NUPs tested, only CPR5 increased GH3 promoter activity, by about sevenfold over controls (Fig. 2A), and it also induced the translocation of IAA12-GFP, a primarily nucleus-localized protein, from the nucleus to the cytoplasm (Fig. 2B). By contrast, neither MOS3, HOS1, NUP75, nor ALADIN affected the auxin response or the subcellular localization of IAA12-GFP.

Fig. 2.

CPR5 mediates the cytoplasmic translocation of IAA12. (A) Only CPR5, among nucleoporins (NUPs) including MOS3, HOS1, NUP75, ALADIN, and CPR5, enhances auxin-responsive GH3pro:LUC reporter activity. Auxin-responsive GH3pro:LUC reporter activity was examined in protoplasts transfected with constructs expressing individual NUP genes, which have mutant RNA sequence data or alter the expression of auxin-responsive genes. Protein abundance was determined by immunoblot using α-HA antibody (Bottom). (B) Fluorescence microscopy analysis shows the enhanced cytoplasmic localization of IAA12-GFP in protoplasts co-transfected with CPR5, while other NUPs have no effect on subcellular localization of IAA12-GFP. NUPs were fused to mRFP at their N termini and to HA at their C termini. (Scale bar, 100 μm.) (C) CPR5 is the only inner ring NUP that specifically enhances auxin-responsive GH3pro:LUC reporter activity in protoplasts transfected with individual constructs encoding NUP35, NUP155, NUP205, CPR5, or CPR5-C (dominant negative form of CPR5). Protein abundance was determined by immunoblot using α-HA antibody (Bottom). (D) CPR5, but not other NUPs or CPR5-C, promotes the cytoplasmic translocation of IAA12-GFP. Fluorescence microscopy analysis was performed on protoplasts co-transfected with IAA12-GFP and mRFP-NUPs-HA. (Scale bar, 100 μm.) White-dashed lines in (B) and (D) show the enlarged images of a single cell showing IAA12-GFP. Bar graphs in (A) and (C) show means ± SD. Different lowercase letters indicate significant differences (P < 0.05, determined by one-way ANOVA with post hoc Tukey’s HSD test).

As CPR5 is an integral protein present in the inner ring of the NPC (57, 58), we further investigated the role of three additional NUP components of the inner ring, NUP35, NUP155, and NUP205 (57, 58), as well as CPR5-C, a dominant negative form of CPR5 (56). Co-transfecting constructs of each inner ring NUP and CPR5-C failed to alter auxin-responsive reporter activity or subcellular localization of IAA12-GFP (Fig. 2 C and D). We then tested whether CPR5-mediated translocation of AUX/IAA is specific for IAA12 by transfecting protoplasts with constructs of each of the 29 AUX/IAA genes alone or together with CPR5. We discovered that AUX/IAAs that belong to the same clade show a similar if not identical localization pattern (SI Appendix, Fig. S2A). For example, CPR5 induced the cytoplasmic translocation of IAA13 and IAA6, which are the most closely related to IAA12 and IAA19, respectively (SI Appendix, Fig. S2 B–D). In addition, IAA1, IAA2, IAA3, IAA4, IAA14, IAA16, IAA17, IAA18, IAA26, and IAA27 also translocated from the nucleus to the cytoplasm when CPR5 was co-transfected (SI Appendix, Fig. S2A). However, the other AUX/IAAs appeared to be insensitive to the presence of CPR5, as they accumulated either in the nucleus or in the cytoplasm regardless of CPR5 (SI Appendix, Fig. S2A). Interestingly, IAA5, which belongs to the clade closest to IAA6 and IAA19, did not respond to CPR5 in regard to its translocation (SI Appendix, Fig. S2E). These data suggest that CPR5, among multiple nuclear pore components, specifically controls the translocation of a subset of IAAs in auxin signaling.

CPR5 Activates Auxin Responses in Lateral Root Development via the Translocation of IAA12 and IAA19.

To confirm the role of CPR5-mediated translocation of IAA12 and IAA19 at the root elongation zone, we observed the fluorescence pattern derived from GFP expressed in the roots of transgenic lines harboring the IAA12pro:gIAA12-GFP and IAA19pro:gIAA19-GFP transgenes in the Col-0, cpr5, and CPR5-OE (35S:gCPR5-HA) backgrounds. In Col-0, IAA12-GFP and IAA19-GFP were nucleus-localized in the meristem region, but gradually accumulated in the cytoplasm closer to the elongation zone (Fig. 3A and SI Appendix, Fig. S3A). By contrast, in cpr5 roots, IAA12-GFP and IAA19-GFP were exclusively nucleus-localized, even in the elongation zone, whereas in CPR5-OE lines, we mainly observed IAA12-GFP and IAA19-GFP in the cytoplasm, from the meristem to the elongation zone (Fig. 3A and SI Appendix, Fig. S3A). We independently validated these results with nuclear and cytoplasmic fractionation assays using the elongation zone of the same lines used for imaging IAA12-GFP and IAA19-GFP in all three genotypes (Col-0, cpr5, and CPR5-OE lines) (Fig. 3B and SI Appendix, Fig. S3B). We observed that CPR5 is associated with the nucleus of CPR5-OE lines (Fig. 3B and SI Appendix, Fig. S3B). IAA12 and IAA19 mainly accumulated in the nucleus or cytoplasm in the cpr5 mutant and CPR5-OE lines, respectively (Fig. 3B and SI Appendix, Fig. S3B).

Fig. 3.

CPR5-mediated cytoplasmic localization of IAA12 at the root elongation zone alters the interaction between IAA12 and ARF7. (A) CPR5 affects the subcellular localization of IAA12. Representative confocal images show GFP fluorescence for IAA12pro:gIAA12-GFP in the roots of 8-d-old Col-0, cpr5, and CPR5-OE seedlings. Roots were stained with PI. Enlarged images show the fluorescence for the elongation zone of 8-d-old IAA12pro:gIAA12-GFP 35S:H2A-mCherry transgenic seedlings in the Col-0, cpr5, and CPR5-OE backgrounds. H2A-mCherry was used as a nuclear marker. Red arrows indicate nuclear-localized IAA12-GFP. (Scale bars, 100 μm.) (B) CPR5 enhances the cytoplasmic localization of IAA12. Total proteins from the elongation zone of IAA12pro:gIAA12-GFP transgenic lines in the Col-0, cpr5, and CPR5-OE backgrounds were separated into nuclear and cytoplasmic fractions. T, 5% of total; C, cytoplasmic fraction; N, nuclear fraction. Protein abundance was determined by immunoblot using α-HA and α-GFP antibodies for CPR5 and IAA12, respectively. UGPase was used as a cytoplasmic marker and H3 as a nuclear marker. (C) CPR5 reduces the interaction between ARF7 and IAA12. Co-immunoprecipitation using protoplasts co-transfected with mER7:GFP as internal control and the indicated effector constructs. ARF7 was pulled down with α-Flag antibody. Protein abundance was determined by immunoblot using α-Flag antibody for ARF7 and α-HA antibody for CPR5 and IAA12. (D) CPR5 interferes with IAA12-mediated suppression of ARF7-dependent transcription. Auxin-responsive GH3pro:LUC reporter activity was measured to examine the effect of CPR5 on IAA12 function. Protein abundance was determined by immunoblot using α-GFP antibody or α-HA antibody. (E) Cytoplasmic IAA12 does not properly repress ARF7-dependent transcription. Reporter assays using auxin-responsive GH3pro:LUC to examine the function of subcellular localization of IAA12 in ARF7-dependent transcription. (F) Cytoplasmic IAA12 is insensitive to auxin-mediated degradation. Immunoblot analysis of mRFP-IAA12-HA, mRFP-IAA12-NLS-HA, and mRFP-IAA12-NES-HA abundance in protoplasts transfected with the constructs after treatment with 100 μM CHX or together with 10 μM IAA for the indicated times (h). Protein abundance was determined by immunoblot using α-HA antibody. Coomassie Brilliant Blue R-250 staining of the RuBisCo Large subunit (rbcL) was used as a protein loading control. Bar graphs in (D) and (E) show means ± SD. Different lowercase letters indicate significant differences (P < 0.05, determined by two-way ANOVA with post hoc Tukey’s HSD test).

We then examined whether CPR5 directly modulates auxin signaling via translocation of IAA12. We co-transfected protoplasts with IAA12 and ARF7, encoding the interacting partner of IAA12, with or without CPR5. In the absence of CPR5, ARF7 was pulled down with IAA12 upon co-immunoprecipitation. Notably, co-transfection with CPR5 in protoplasts decreased the interaction between ARF7 and IAA12 (Fig. 3C). We also examined the effect of CPR5-mediated IAA12 translocation on auxin signaling by using the GH3pro:LUC reporter in protoplasts. ARF7 increased the transcriptional output of the GH3 promoter and further enhanced its activity when CPR5 was co-transfected (Fig. 3D). IAA12-GFP and IAA12-HA strongly decreased the activation of the reporter by ARF7; however, CPR5 partially compromised IAA12-mediated suppression of ARF7-dependent activation of the auxin-responsive GH3pro:LUC reporter. IAA19-GFP and IAA19-HA also exerted the same effect on the reporter activity as IAA12 proteins (Fig. 3D and SI Appendix, Fig. S3C). These data strongly suggest that the CPR5-induced exclusion of IAA12 from the nucleus activates auxin signaling. To test this idea, we generated constructs that encode IAA12 fused to a nuclear localization sequence (NLS [-PKKKRKVA-], IAA12-NLS) or a nuclear export sequence (NES [-LPPLERLTL-], IAA12-NES) (SI Appendix, Fig. S3D) and examined their effects on GH3 reporter activity. IAA12-NLS, like IAA12, strongly decreased the ARF7-mediated activation of the GH3 promoter; however, IAA12-NES only partially suppressed the ARF7-mediated activation of GH3pro:LUC, likely because some IAA12-NES remained in the nucleus (Fig. 3E and SI Appendix, Fig. S3D). To further understand how local IAA12 affects auxin signaling, we examined protein stability in protoplasts. The abundance of IAA12, IAA12-NLS, and IAA12-NES did not change in the presence of cycloheximide (CHX) (Fig. 3F). After auxin treatment, the abundance of IAA12 decreased gradually as much as those of IAA12-NLS, but not IAA12-NES (Fig. 3F). Compared to short-lived AUX/IAA, IAA17 (59–63), IAA12 and IAA19 proteins have much longer half-lives (SI Appendix, Fig. S3E). IAA12 and IAA19 also showed long protein half-lives in planta and were fully degraded after treatment of IAA for 2 h (SI Appendix, Fig. S3F). These data indicate that cytosolic IAA12 is insensitive to auxin and cannot suppress ARF7-mediated activation of auxin responses.

To further explore the function of CPR5 in auxin signaling, we introduced the auxin-responsive reporter construct DR5v2:NLS-GFP into the Col-0, cpr5, and CPR5-OE backgrounds. We detected abolished GFP fluorescence from the DR5v2:NLS-GFP reporter in the elongation zone of the cpr5 mutant (Fig. 4A and SI Appendix, Fig. S3G). In agreement with this observation, the relative fluorescence intensity of GFP was lower in the root of the cpr5 mutant than in that of Col-0 (Fig. 4B and SI Appendix, Fig. S3G). By contrast, in the CPR5-OE lines, the relative GFP intensity was much higher throughout the entire root and particularly in the elongation zone (Fig. 4 A and B and SI Appendix, Fig. S3G). The auxin signaling in the elongation zone and basal meristem region are essential for lateral root priming, which determines lateral root formation (22, 23).

Fig. 4.

CPR5 positively regulates auxin responses in lateral root development. (A) CPR5 enhances auxin responses in Arabidopsis. Representative confocal images show GFP fluorescence derived from the DR5v2:NLS-GFP transgene in the PI-stained roots of 8-d-old Col-0, cpr5, and CPR5-OE seedlings. (Scale bars, 100 μm.) (B) Relative GFP intensity of DR5v2:NLS-GFP 35S:H2A-mCherry in the roots of 8-d-old Col-0, cpr5, and CPR5-OE lines (n = 15). (C) CPR5-mediated cytoplasmic translocation of IAA12 and IAA19 promotes lateral root development. Representative images of lateral root development of 10-d-old Col-0, IAA12/19-NLS, cpr5, cpr5 IAA12/19-NLS, and CPR5-OE lines after mock or 20 nM IAA treatment. (Scale bar, 1 cm.) (D) Lateral root number, primary root length, and lateral root density of 10-d-old Col-0, IAA12/19-NLS, cpr5, cpr5 IAA12/19-NLS, and CPR5-OE seedlings after mock or 20 nM IAA treatment (n = 15). In (B) and (D), bar graphs show means ± SD with individual data points. Different lowercase letters indicate significant differences (P < 0.05, determined by one-way ANOVA with post hoc Tukey’s HSD test).

We next investigated the role of CPR5-mediated translocation of IAA12 and IAA19 in auxin signaling by characterizing the root phenotypes of Col-0, cpr5, and CPR5-OE lines in detail. IAA12/19 (IAA12pro:gIAA12-HA IAA19pro:gIAA19-HA) and IAA12/19-NES (IAA12pro:gIAA12-NES-HA IAA19pro:gIAA19-NES-HA) lines did not change the lateral root development compared to Col-0 (SI Appendix, Fig. S4 A and B). However, IAA12/19-NLS (IAA12pro:gIAA12-NLS-HA IAA19pro:gIAA19-NLS-HA) showed reduced lateral root number and primary root length compared to Col-0 (Fig. 4 C and D and SI Appendix, Fig. S4 A and B). The cpr5 mutant produced fewer lateral roots and had a shorter primary root compared to Col-0, as previously reported (64). cpr5 IAA12/19-NLS showed reduced lateral root number and primary root length compared to the cpr5 mutant (Fig. 4 C and D), whereas CPR5-OE lines showed the opposite phenotype, with more lateral roots and a longer primary root (Fig. 4 C and D). To our surprise, IAA12/19-NLS, cpr5, and cpr5 IAA12/19-NLS were even insensitive to exogenous auxin in regard to lateral root development, while CPR5-OE lines were sensitive to auxin but similar to Col-0 (Fig. 4 C and D). Therefore, the density of lateral roots was highly decreased in IAA12/19-NLS, cpr5, and cpr5 IAA12/19-NLS, but substantially increased in CPR5-OE lines. This suggests that CPR5-mediated translocation of IAA12 and IAA19 positively regulates auxin signaling and lateral root development. To further examine whether CPR5 is directly linked to canonical auxin signaling in lateral root development, we generated the arf7/19 cpr5 triple mutant and overexpressed CPR5 in the arf7/19 double mutant background (arf7/19 CPR5-OE). ARF7 and ARF19 are positive regulators of auxin signaling for lateral root development (26). Notably, similar to arf7/19, neither the arf7/19 cpr5 nor arf7/19 CPR5-OE lines failed to develop lateral roots (SI Appendix, Fig. S4 C and D), suggesting that CPR5-regulated lateral root development is dependent on the ARF7 and ARF19 pathway. Furthermore, compared to Col-0, IAA12/19-NLS and the cpr5 mutant were characterized by shorter hypocotyls when seedlings were grown at 29 °C and cpr5 IAA12/19-NLS showed even shorter hypocotyls than the cpr5 mutant, while CPR5-OE lines exhibited longer hypocotyl elongation under the same conditions (SI Appendix, Fig. S4 E and F). There were no significant changes in hypocotyl length in either the cpr5 mutant or CPR5-OE lines compared to Col-0 when grown at 22 °C (SI Appendix, Fig. S4 E and F). Hypocotyl elongation is induced by auxin-mediated cell expansion at high temperature (29 °C) compared to normal ambient temperature (22 °C) (65, 66). Together, these data strongly indicate that the CPR5-mediated cytoplasmic translocation of IAA12 and IAA19 increases the pool of free ARF7 in the nucleus, leading to direct activation of auxin signaling.

CPR5, Which Induces the Cytoplasmic Translocation of IAA12 and IAA19, Regulates Lateral Root Development in Response to Stress.

To investigate how abiotic stress induces the nuclear translocation of IAA12 and IAA19, leading to the inhibition of lateral root development, we examined the expression pattern of CPR5 under various stress conditions. Interestingly, CPR5 transcript levels were decreased following treatment with PEG, mannitol, or NaCl (Fig. 5A). CPR5 expression was previously reported to decrease during drought stress (67). We also determined the staining pattern from CPR5pro:GUS lines, carrying the β-glucuronidase (GUS) reporter gene driven by a 3-kb CPR5 promoter fragment, under control and stress conditions. We detected GUS staining at the root elongation zone (Fig. 5B), in cotyledons, and in hypocotyls (SI Appendix, Fig. S5 A and B). This GUS staining pattern drastically weakened upon treatment with PEG, mannitol, or NaCl. To further examine whether abiotic stress suppresses the activity of CPR5, we investigated the subcellular localization of IAA12-GFP in CPR5-overexpressing protoplasts and CPR5-OE transgenic lines after abiotic stress treatment. IAA12-GFP proteins were mainly localized in the cytoplasm in CPR5-overexpressing protoplasts and CPR5-OE lines under normal condition, but, interestingly, mostly localized in the nucleus after abiotic stress treatment (SI Appendix, Fig. S6 A and B). As the cpr5 mutant is known to accumulate high levels of biotic stress-related salicylic acid (SA) (64), we tested the effect of exogenous SA application on IAA12pro:gIAA12-GFP and IAA19pro:gIAA19-GFP lines; however, SA did not alter the nucleo-cytoplasmic localization of IAA12-GFP or IAA19-GFP at the root elongation zone (SI Appendix, Fig. S7 A and B). Together, these results suggest that abiotic stress-induced suppression of CPR5 expression and activity results in incremental localization of IAA12 and IAA19 to the nucleus.

Fig. 5.

Abiotic stress-mediated suppression of CPR5 expression is directly related to lateral root development. (A) The expression of CPR5 was strongly decreased upon abiotic stress exposure. Total RNA was isolated from 8-d-old Col-0 seedlings after mock, PEG, mannitol, or NaCl treatment for the indicated times (h) and transcript levels of CPR5 and UBQ1. UBQ1 was used as a control. (B) Representative images show CPR5pro:GUS expression patterns in the root after mock or the indicated stress treatments. (Scale bars, 100 μm.) (C) CPR5 controls lateral root development in response to stress. Representative images show lateral root development of 10-d-old Col-0, IAA12/19-NLS, cpr5, cpr5 IAA12/19-NLS, and CPR5-OE seedlings after mock or the indicated stress treatments. (Scale bars, 1 cm.) (D) Lateral root number, primary root length, and lateral root density of 10-d-old Col-0, IAA12/19-NLS, cpr5, cpr5 IAA12/19-NLS, and CPR5-OE seedlings after mock, PEG, mannitol, or NaCl treatment (n = 15). Bar graphs show means ± SD with individual data points. Different lowercase letters indicate significant differences (P < 0.05, determined by two-way ANOVA with post hoc Tukey’s HSD test). (E) Model depicting how CPR5-mediated nucleo-cytoplasmic localization of IAA12 and IAA19 regulates auxin responses and lateral root development during abiotic stress conditions. In normal conditions, CPR5 is expressed at the root elongation zone and promotes the cytoplasmic translocation of IAA12 and IAA19, which prevents the suppression of ARF-dependent transcription, resulting in high auxin responses and optimal lateral root development (Left). Under abiotic stress conditions, CPR5 expression is strongly decreased, leading to the accumulation of nucleus-localized IAA12 and IAA19 at the root elongation zone and, thus, the repression of ARF-dependent transcription, resulting in low auxin responses and reduced lateral root development under abiotic stress conditions (Right).

To confirm that CPR5-mediated cytoplasmic translocation of IAA12 and IAA19 proteins participates in plant responses to stress, we dissected lateral root development in Col-0, cpr5, and CPR5-OE lines after mock treatment or treatment with PEG, mannitol, or NaCl. The number of lateral roots, the length of primary roots, and the density of lateral roots were all decreased in IAA12/19-NLS, cpr5, and cpr5 IAA12/19-NLS lines, but increased in CPR5-OE lines, compared to those of Col-0 (Fig. 5 C and D). After exposure to stress, the number of lateral roots, the length of primary roots, and the density of lateral roots further decreased in IAA12/19-NLS, cpr5, and cpr5 IAA12/19-NLS lines, but reached numbers that were not significantly different from those of Col-0 exposed to stress (Fig. 5 C and D). All parameters remained higher in CPR5-OE lines compared to Col-0 under stress conditions, although they were lower than those of mock-treated CPR5-OE seedlings. Taken together, these results indicate that CPR5-mediated control of IAA12 and IAA19 subcellular localization is a major regulatory mechanism in auxin signaling to modulate lateral root development under abiotic stress conditions.

Discussion

Auxin regulates numerous aspects of plant growth and development through a simple signaling pathway (68–71); in particular, auxin functions as a final integrator of various signals in lateral root formation. Shortly, auxin induces the proteolysis of repressor AUX/IAA proteins, which alleviates their repression of ARFs and allows the transcriptional regulation of auxin signaling (68–70, 72). Apart from this canonical auxin signaling, as we describe here, there is also a CPR5-mediated regulatory mechanism of auxin signaling that determines the localization of IAA12 and IAA19 and subsequent release of ARFs to initiate lateral root development. We showed that CPR5 mediates the cytosolic translocation of IAA12 and IAA19, which in turn increases the pool of free ARFs in the nucleus, and maintains the output of auxin signaling under normal growth conditions at the root elongation zone, which determines lateral root founder cells and initiates lateral root primordia (5, 22).

Cytoplasmic IAA12 was highly stable even when auxin levels were high, which might be a prerequisite for a swift response to environmental changes in the nucleus irrespective of the canonical auxin signaling pathway. In fact, abiotic stress suppressed CPR5 transcription, resulting in the accumulation of IAA12 and IAA19 in the nucleus and the inhibition of auxin signaling in the context of lateral root development. Auxin signaling, especially the IAA14-ARF7/ARF19 and IAA12-ARF5 modules, is crucial for lateral root development (5, 28, 29, 73, 74). We determined that IAA12 also interacts with ARF7, as well as with ARF5, and could suppress ARF7-dependent transcriptional activation of auxin signaling, supporting the notion that various IAA-ARF modules, including the IAA12-ARF7 module, regulate lateral root development following various endogenous and environmental signals (5, 75–78).

The spatiotemporal control of phytohormone signaling regulators contributes to the regulation of diverse developmental processes in plants (49, 79–82). In Arabidopsis, more than 30 NUP proteins regulate the permeability of nuclear pores. Several NUP proteins have been reported to control auxin signaling by regulating efficient export of mRNA from the nucleus (51–53). For example, higher mRNA accumulation in the nucleus of mutants in NUP96, NUP160, and TRANSLOCATED PROMOTER REGION, which all suppress the effects of the auxin-resistant 1 mutant defective in auxin sensitivity, may contribute to the nuclear retention of transcripts encoding negative regulators of auxin signaling (51, 53). In addition, ARF7 and ARF19, two positive regulators of auxin signaling, form cytoplasmic condensates in the upper parts of roots and control auxin sensitivity (49). CPR5, a component of the NPC, might provide another level of regulation of auxin signaling by spatially segregating negative regulators AUX/IAAs as well as IAA12 and IAA19. The CPR5-mediated cytoplasm-localized AUX/IAAs, including IAA12 and IAA14, are mainly involved in lateral root formation and contain four conserved domains. Domain I is a co-repressor domain with an ethylene-responsive element binding-associated amphiphilic repression (EAR) motif that recruits chromatin remodeling components; domain II is an auxin-binding domain that induces the degradation of AUX/IAAs; and domains III and IV together harbor a PB1 (Phox and Bem1p) domain that binds to ARFs and AUX/IAAs (69, 72, 83). Notably, those IAAs that were constitutively localized to the cytosol and were CPR5-insensitive (IAA10, IAA11, IAA20, and IAA28–IAA34) carry various mutations in their NLS motifs (84) and do not contain entire conserved domains. For example, IAA20 and IAA29–IAA33 each lack either domain I or domain II (83). Endogenous NLS motifs within AUX/IAAs might be necessary for IMPORTIN-mediated cytoplasm-to-nucleus translocation. IAA8 and IAA9, which localized to the nucleus regardless of CPR5, have an abnormally long linker between domains I and II (83). However, it is also plausible that CPR5-mediated cytoplasmic AUX/IAAs are retained and function in the cytoplasm because of as-yet-unknown interacting factors that inhibit their nuclear transport. But, we used GFP-tagged [which has a very long protein half-life (85)] AUX/IAAs to show the changes in their subcellular localization patterns. Therefore, a more systematic study is needed to conclude CPR5-mediated changes in subcellular localization, if, any, of AUX/IAAs, tagged with Venus or HA, which have short protein half-life or short length. Nevertheless, our detailed set of experiments using Venus and HA tags demonstrate well that CPR5, among multiple nuclear pore components, controls the translocation of at least IAA12/19 proteins in auxin signaling.

CPR5 is a membrane-anchored protein in the inner ring part of the NPC and is surrounded by several other NUPs (51, 56, 86). We could not detect a direct interaction between CPR5 and IAA12, but CPR5 directly interacts with NUP155 and NUP93a (56), which might also associate with other NUPs. The expression patterns of NUP genes vary across different tissues (87–89), raising the possibility that different combinations of NPCs might determine the selectivity of translocated proteins in each tissue. CPR5 induces not only the translocation of AUX/IAAs but also that of the defense-related proteins NONEXPRESSER OF PR GENES 1 (NPR1), JASMONATE-ZIM-DOMAIN PROTEIN 1 (JAZ1), and ABA-INSENSITIVE 5 (56). As CPR5 is widely expressed in various tissues (67), CPR5-mediated translocation of target proteins could regulate a plethora of plant growth and development. The structural variation in AUX/IAAs, irrespective of their interacting partners and the expression patterns, is likely to be linked to their functionality in auxin signaling throughout the plant life cycle; however, the detailed regulatory mechanisms still remain to be elucidated.

Although the inhibition of auxin signaling is critical for preventing lateral root development under diverse stress conditions (4, 21), its underlying regulatory mechanism has not been fully elucidated. Here, we demonstrated that the active regulation of AUX/IAA localization by CPR5 might act as a common integrator of various abiotic stress signals to regulate the output of auxin signaling. That is, the lowered expression of CPR5 under abiotic stress allows IAA12 and IAA19 to be retained in the nucleus, thereby reducing the auxin response and suppressing the development of lateral roots (Fig. 5E). Our findings provide insight into auxin signaling regulatory mechanism under abiotic stress conditions, whereby the localization of IAA12 and IAA19 plays a critical role in lateral root development.

Materials and Methods

Plant Materials and Growth Conditions.

Arabidopsis thaliana Columbia-0 (Col-0) was used as the wild-type control in all experiments and genetic backgrounds to generate transgenic lines in this study. Plants were grown under long-day (22 °C, 16 h light/8 h dark) conditions. The cpr5 (SALK_021049) and arf7-1 arf19-1 (CS24625) mutants were provided by Arabidopsis Biological Research Center. All double and triple mutant plants were obtained through genetic crosses.

Plasmid Construction and Arabidopsis Transformation.

To construct plasmids for protoplast expression, coding sequences of CPR5 (At5g64930), CPR5-C (dominant negative form of CPR5) (56), MOS3 (At1g80680), HOS1 (At2g39810), NUP75 (At4g32910), ALADIN (At3g56900), NUP35 (At3g16310), NUP205 (At5g51200), ARF7 (At5g20730), and all IAA genes were PCR amplified without stop codon from Col-0 total cDNA. PCR amplicons were cloned into protoplast expression vectors (26) in frame with GFP and the hemagglutinin (HA), NLS-HA, NES-HA, Flag, or mRFP tags under control of the 35S C4PPDK promoter; all plasmids harbor an ampicillin resistance cassette. To generate the CPR5pro:GUS transgenic lines, a CPR5 promoter fragment (3,000 bp) including the 5′ untranslated region (5′ UTR) was PCR amplified from Col-0 genomic DNA and then cloned into pCAMBIA1303 upstream of the GUS reporter gene; resulting plasmid also carries a hygromycin resistance marker for plant selection. To generate the DR5v2:NLS-GFP transgenic lines, a fragment corresponding to DR5v2:NLS (90) was cloned into pCAMBIA1303 in frame with monomeric GFP (mGFP). To generate the IAA12pro:gIAA12-GFP, IAA19pro:gIAA19-GFP, and ARF7pro:gARF7-GFP transgenic lines, 3,000-bp promoter fragments including 5′ UTR and full-length genomic sequences for the respective genes were PCR amplified and cloned into pCAMBIA1303 in frame with mGFP. To generate CPR5-overexpressing (CPR5-OE) lines, full-length genomic sequence of CPR5 was PCR amplified without stop codon and cloned downstream of 35S C4PPDK promoter in frame with HA tag in the pPZP211, which has a kanamycin resistance marker for plant selection. To generate 35S:H2A-mCherry lines, full-length CDS sequence of HISTONE2A-6 (H2A, AT5G59870) was amplified by PCR without stop codon and cloned downstream of 35S C4PPCK promoter in frame with mCherry tag in the pCB302ES vector with a basta resistance marker for plant selection. All constructs were transformed into Col-0 via Agrobacterium tumefaciens-mediated transformation using floral dipping method (91). IAA12pro:gIAA12-GFP, IAA19pro:gIAA19-GFP, and DR5v2:NLS-GFP transgenes were introduced into the CPR5-OE lines and the cpr5 mutant by genetic crossing. Primers for cloning are listed in SI Appendix, Table S1.

An extended description of materials and methods used in this study is given in SI Appendix, SI Materials and Methods.

Supplementary Material

Appendix 01 (PDF)

Data, Materials, and Software Availability

All study data are included in the article and/or SI Appendix.