Polar auxin transport, mediated by two distinct plasma membrane-localized auxin influx and efflux carrier proteins/complexes, plays an important role in many plant growth and developmental processes including tropic responses to gravity and light, development of lateral roots and patterning in embryogenesis. We have previously shown that the Arabidopsis AGRAVITROPIC 1/PIN2 gene encodes an auxin efflux component regulating root gravitropism and basipetal auxin transport. However, the regulatory mechanism underlying the function of AGR1/PIN2 is largely unknown. Recently, protein phosphorylation and dephosphorylation mediated by protein kinases and phosphatases, respectively, have been implicated in regulating polar auxin transport and root gravitropism. Here, we examined the effects of chemical inhibitors of protein phosphatases on root gravitropism and basipetal auxin transport, as well as the expression pattern of AGR1/PIN2 gene and the localization of AGR1/PIN2 protein. We also examined the effects of inhibitors of vesicle trafficking and protein kinases. Our data suggest that protein phosphatases, sensitive to cantharidin and okadaic acid, are likely involved in regulating AGR1/PIN2-mediated root basipetal auxin transport and gravitropism, as well as auxin response in the root central elongation zone (CEZ). BFA-sensitive vesicle trafficking may be required for the cycling of AGR1/PIN2 between plasma membrane and the BFA compartment, but not for the AGR1/PIN2-mediated root basipetal auxin transport and auxin response in CEZ cells.
Molecular analyses of genes encoding the putative auxin influx and efflux carrier proteins, and immunolocalization studies of the encoded proteins support a model postulating that the direction of polar auxin transport is primarily mediated by auxin efflux carrier components that are asymmetrically localized in the plasma membrane (Gälweiler et al., 1998). Although, the auxin influx carrier protein AUX1 may also be involved (Swarup et al., 2001). The Arabidopsis AGRAVITROPIC 1 (AGR1, also named as EIR1/PIN2/WAV6; At5g57090; Chen et al., 1998; Luschnig et al., 1998; Müller et al., 1998; Utsuno et al., 1998), a component of the auxin efflux complex, was localized to the basal plasma membrane of epidermal, cortical and lateral cap cells in the root tip region, coinciding with the direction of root basipetal auxin transport (Boonsirichai et al., 2003; Müller et al., 1998).
Molecular mechanisms that underlie the asymmetric plasma membrane localization and activities of AGR1/PIN2 and other PIN proteins remain to be elucidated. An emerging model for polar auxin transport postulates that a brefeldin A (BFA)-sensitive vesicle trafficking pathway is involved in the regulation of PIN protein cycling between an unknown perinuclear compartment and the plasma membrane (Geldner et al., 2001, 2003, 2004; Steinmann et al., 1999; see Muday and Murphy, 2002 for review). Recently, it was shown that the cycling of PIN3 protein may be regulated by gravity (Friml et al., 2002a).
Using a tobacco cell culture system, Delbarre et al. (1998) measured the carrier-mediated auxin influx and efflux activities. Their data indicated that protein phosphorylation plays an essential role in maintaining auxin efflux, but not influx activities in tobacco cells. In separate studies, protein phosphatases have been identified as important regulators of polar auxin transport in Arabidopsis plants. The Arabidopsis RCN1 gene (root curling in the presence of NPA) encodes one of three A regulatory subunits of protein phosphatase 2A, a heterotrimeric serine/threonine protein phosphatase (DeLong et al., 2002; Luan, 2003). Loss-of-function mutations in the RCN1 gene result in plants defective in root curling, apical hook formation, and root and shoot elongation (Deruere et al., 1999; Garbers et al., 1996). Mutant plants also exhibit an elevated level of root basipetal auxin transport (Rashotte et al., 2001), and an impaired sensitivity to the polar auxin transport inhibitor 1-N-naphthylphthalamic acid (NPA) in auxin accumulation in hypocotyls (Garbers et al., 1996; Rashotte et al., 2001). On the other hand, loss-of-function mutations in PINOID (PID) gene, encoding a protein kinase, have inflorescence stems similar to that of the pin-formed (pin1) mutant (Benjamins et al., 2000; Christensen et al., 2000; Furutani et al., 2004). While pid mutants do not exhibit obvious defects in roots, over-expression of PID causes roots of the transgenic plants to be agravitropic (Benjamins et al., 2000; Christensen et al., 2000), and an apical-basal shift in the polarity of several PIN proteins (Friml et al., 2004). Taken together, these observations suggest that both protein phosphorylation and dephosphorylation play important roles in regulating polar auxin transport.
Previous genetic and physiological studies indicated that AGR1/PIN2 plays an important role in root gravitropism and basipetal auxin transport (Bell and Maher, 1990; Chen et al., 1998; Luschnig et al., 1998; Rashotte et al., 2001). To extend our previous analysis on AGR1/PIN2, we examined the effects of chemical inhibitors of protein kinases and phosphatases on AGR1/PIN2-mediated root gravitropism, basipetal auxin transport and auxin response in the root central elongation zone (CEZ; Mullen et al., 1998), using an auxin response reporter, DR5::β-glucuronidase (DR5::GUS; Ulmasov et al., 1997). Our studies indicate that protein phosphatases sensitive to cantharidin and okadaic acid are involved in regulating AGR1/PIN2-mediated root gravitropism, basipetal auxin transport, and auxin response in the CEZ. On the other hand, inhibitor studies indicated that BFA-sensitive vesicle trafficking may be required for the cycling of AGR1/PIN2 between the plasma membrane and perinuclear endosomal compartments, but not for the AGR1/PIN2-mediated root basipetal auxin transport and auxin response in CEZ cells.
AGR1/PIN2, a component of the auxin efflux carrier complex mediates root basipetal auxin transport and induces the expression of DR5 auxin response reporter in the root elongation zone
To further analyze AGR1/PIN2 function in root gravitropism and basipetal auxin transport (Chen et al., 1998; Luschnig et al., 1998; Müller et al., 1998; Rashotte et al., 2001; Utsuno et al., 1998), we tested root basipetal auxin transport using DR5 auxin response reporter consisting of multiple copies of an auxin responsive cis-element TGTCTC fused to Escherichia coliβ-glucuronidase gene (Ulmasov et al., 1997). Four to 5-day-old DR5::GUS Arabidopsis seedlings were transferred onto growth medium plates. Agar blocks of 1 mm diameter made of the same growth medium supplemented with 10 μm IAA or DMSO (0.1% v/v) solution were placed next to the root tips. After incubation for 1–1.5 h, DR5::GUS gene expression was assayed by histochemical staining. As shown in Figure 1(a), DR5::GUS was expressed in root tips of wild-type plants as previously reported (Sabatini et al., 1999). IAA application did not change the root-tip expression of the DR5::GUS gene (Figure 1b). However, it induced the DR5::GUS gene expression in CEZ cells approximately 400 μm away from the root tip. The gap in the GUS expression in the distal elongation zone (DEZ) was also GUS positive after overnight staining (data not shown).
The DR5::GUS gene expression in CEZ cells was specific to auxin treatments as it was not observed in wild-type plants treated with other signaling molecules including benzyl aminopurin (cytokinin), ACC (ethylene biosynthetic precursor), abscissic acid, GA3 (gibberillin) and brassinolide (brassinosteroid) (data not shown). Furthermore, the expression of the DR5::GUS gene in CEZ cells required a minimum of 10−7m IAA and incubation time of 60 min (data not shown).
As AGR1/PIN2 is expressed at high levels in the DEZ and to a lesser level in the CEZ, we tested whether AGR1/PIN2 plays a role in the auxin-induced DR5::GUS gene expression in CEZ cells, using the agr1-5 mutant (Chen et al., 1998). In agr1-5 mutant roots, DR5::GUS gene expression was stronger in the root tip and expanded into the lateral root-cap cells compared to the wild type (Figure 1a,c), consistent with a previous report that free auxin levels in the root tip were higher in the agr1 mutant than in the wild type (Ottenschläger et al., 2003). When root tips of agr1-5 DR5::GUS plants were treated with 10 μm IAA for 1–1.5 h, GUS staining in CEZ cells was negative or extremely weak (Figure 1d).
We also tested the effects of global auxin application on DR5::GUS gene expression in wild type and agr1-5 mutant plants. When wild type and agr1-5 seedlings were treated with an auxin solution for 1 h, the patterns of the DR5::GUS gene expression in CEZ cells were identical to that observed when seedlings were treated with auxin at root tips (Figure 1e–h; compare Figure 1b and f, d and h).
The reduced GUS activity in CEZ cells of the agr1-5 mutant was not due to a lack of cellular response to auxin in the mutant, because the expression of the DR5::GUS gene was readily induced in CEZ cells of the agr1-5 mutant when mutant seedlings were vacuum infiltrated with, or incubated in the auxin solution for an extended period of time (data not shown).
To evaluate whether the effect of auxin on DR5::GUS gene expression occurred at the level of gene expression, we prepared total RNA from the root elongation zone of 5-day-old light-grown seedlings treated with or without auxin. As shown in Figure 1(i,j), we found that GUS transcripts were absent in the root elongation zone of both wild type and agr1-5 mutant plants at T = 0 after the auxin treatment. At T = 60 min, GUS transcripts were significantly increased in the wild type (set as 100% relative level, normalized with the expression levels of the eIF4A gene). At this time point, GUS transcripts were 33 ± 13% lower in the agr1-5 mutant than in the wild type (Figure 1i,j; Student's t-test, P < 0.05). The reduction of GUS transcripts in the mutant roots remained similar over the 90-min period. At T = 120 min, GUS transcripts in the elongation zone of wild-type plants were increased by 86% from T = 0, whereas GUS transcripts in the agr1-5 mutant were only increased by 35% from T = 0 (Figure 1i,j).
The reduced GUS transcript level in the elongation zone upon auxin treatment may reflect reduced auxin level, auxin sensitivity, or a combination of both in the agr1 mutant. To differentiate these possibilities, we measured the rate of root basipetal auxin transport in wild type and agr1-5 mutant plants using a protocol modified from Rashotte et al. (2000). We found that basipetal auxin transport was 31 ± 10% lower in the agr1-5 mutant than in the wild type (Student's t-test, P < 0.05; n = 10; number of experiments = 3; Figure 1k), consistent with the report of Rashotte et al. (2001). Even though, this observation strongly supports the hypothesis that auxin levels in CEZ cells are reduced in the agr1 mutant. However, we cannot exclude the possibility that auxin sensitivity is also reduced in the mutant.
Cantharidin, okadaic acid and TIBA, but not BFA effectively blocked the auxin response reporter gene expression in CEZ cells
To examine cellular processes involved in the regulation of AGR1/PIN2-mediated root basipetal auxin transport, we tested the effects of various pharmacological compounds on DR5::GUS gene expression in CEZ cells upon auxin treatments. Four- to 5-day-old seedlings were transferred to the liquid growth medium supplemented with 0, 1, 5, 10, 25 and 50 μm of cantharidin, or 0, 0.03, 0.1 and 0.3 μm of okadaic acid, both of which are specific inhibitors of protein phosphatases 1 and 2A (Li et al., 1993; Luan, 2003). After incubation for 1 h, 10 μm of IAA was added to the medium. Seedlings were incubated for an additional 2 h before GUS activity was histochemically assayed. Pre-treatments with cantharidin (25 μm or higher) or okadaic acid (0.3 μm) effectively inhibited the auxin-induced DR5::GUS gene expression in CEZ cells (Figure 2d,e; Table 1). In contrast, cantharidin at 10 μm or lower and okadaic acid at 0.1 μm or lower did not completely interrupt the DR5::GUS expression in the root elongation zone (Table 1).
Table 1. Effects of a genetic mutation and various chemical inhibitors on root gravitropism, basipetal auxin transport and auxin-induced DR5::GUS gene expression in the central elongation zone (CEZ)
We also tested the effects of other chemicals known to affect root gravitropism on DR5::GUS gene expression in CEZ cells (Table 1). We found that a polar auxin transport inhibitor 2,3,5-triiodobenzoic acid (TIBA; 25 μm or higher) completely blocked the auxin-induced DR5::GUS expression in CEZ cells (Figure 2c). In contrast, pre-treatments with a protein kinase inhibitor staurosporin (10 μm; Figure 2h), or a vesicle trafficking inhibitor BFA (25 and 50 μm; Figure 2g; data not shown) did not significantly inhibit the auxin-induced DR5::GUS expression in CEZ cells (see Table 1). Surprisingly, NPA at 50 and 100 μm did not abolish DR5::GUS expression in CEZ cells (Figure 2f, Table 1). We obtained identical results in separate experiments using root-tip application of auxin (data not shown).
Complex regulation of root basipetal auxin transport by cantharidin and okadaic acid
To further confirm the effects of these chemicals on root basipetal auxin transport, we measured the movement of radio-labeled IAA in roots treated with or without these chemicals. As shown in Figure 3(a), when Arabidopsis seedlings were pre-treated with 5, 10 and 25 μm of cantharidin for 1.5 h, basipetal auxin transport rate was at 122 ± 14%, 106 ± 18% and 70 ± 5% of the control, respectively. Cantharidin at 5 μm slightly but significantly elevated the rate of basipetal auxin transport (Student's t-test, P < 0.05; n = 6–10; number of experiments = 3), whereas it did not significantly change the auxin transport at 10 μm. At 25 μm, however, cantharidin significantly reduced the rate of root basipetal auxin transport (Student's t-test, P < 10−7).
To test whether these dosage-specific effects were due to physiological effects rather than non-specific drug-related responses, we also tested basipetal auxin transport in seedlings pre-treated with various concentrations of okadaic acid, another inhibitor of protein phosphatases 1 and 2A. We found that seedlings pre-treated with 0.03, 0.1 and 0.3 μm of okadaic acid had auxin transport levels at 118 ± 11%, 117 ± 18% and 57 ± 9% of the control, respectively (Figure 3c). Similar to cantharidin, okadaic acid at 0.03 μm slightly, but significantly increased the rate of basipetal auxin transport in roots (Student's t-test; P < 0.05). On the other hand, okadaic acid at 0.3 μm significantly reduced the rate of auxin transport (Student's t-test; P < 10−5).
Previous reports showed that mutations in the regulatory A subunit (RCN1) of the heterotrimeric serine/threonine protein phosphatase 2A caused an elevated rate of root basipetal auxin transport and delayed the root gravitropic response when compared with wild-type plants (Rashotte et al., 2001). Furthermore, treatments of wild-type plants with cantharidin (10 μm) for 48 h elevated the rate of auxin transport and delayed the root gravitropic curvature response, phenocopying the rcn1 mutation (Rashotte et al., 2001). However, the negative effect of cantharidin at 25 μm on auxin transport was not reported previously. To test whether the effect of high doses of cantharidin is through RCN1 or other regulatory proteins such as other regulatory A subunit isoforms of PP2A (see Zhou et al., 2004), we examined the rate of basipetal auxin transport in the rcn1 mutant treated with various concentrations of cantharidin. As shown in Figure 3(b), the rate of auxin transport in rcn1 mutant plants treated with 0, 5, 10 and 25 μm cantharidin was 127 ± 18, 132 ± 16, 121 ± 11 and 80 ± 11% of the wild-type control (100%). In rcn1 roots, basipetal auxin transport was significantly elevated compared to the wild-type control (Figure 3a,b; Student's t-test; P < 0.005). In rcn1 roots treated with 5 and 10 μm cantharidin for 90 min, basipetal auxin transport was not significantly different from the DMSO control (Figure 3b; Student's t-test; P > 0.05 for both treatments). However, when the rcn1 mutant was treated with 25 μm cantharidin, basipetal auxin transport was significantly reduced compared to the DMSO control (Student's t-test, P = 4 × 10−6). Furthermore, treatments with 25 μm cantharidin slightly but significantly elevated root basipetal auxin transport in the rcn1 mutant compared to the wild type (Figure 3a,b; Student's t-test, P < 0.05). These data suggest that cantharidin primarily targets RCN1 regulatory activity at low concentrations, and affects both RCN1 and other targets at high concentrations.
It is known that TIBA and NPA are inhibitors of polar auxin transport. We tested the dosage effects of TIBA and NPA on basipetal auxin transport in Arabidopsis roots. We found that the rate of basipetal auxin transport in seedlings pretreated with 10, 25, and 50 μm of TIBA was 79 ± 17%, 61 ± 7% and 21 ± 4% of the control, respectively (Figure 3d), and 79 ± 16%, 61 ± 9% and 54 ± 17% of the control in seedlings treated with 10, 25 and 50 μm of NPA, respectively (Figure 3e). These observations are consistent with their proposed role as inhibitors of polar auxin transport.
As the vesicle trafficking inhibitor BFA did not inhibit DR5::GUS gene expression in CEZ cells upon auxin treatment (Figure 2g), we also tested the effect of BFA on basipetal auxin transport. As shown in Figure 3(f), the rate of root basipetal auxin transport in seedlings pre-treated with 25 and 50 μm of BFA was 120 ± 17 and 100 ± 13% of the control, respectively.
Both treatments with 25 μm cantharidin and the agr1-5 mutation reduced the rate of basipetal auxin transport by approximately 30%, and inhibited the DR5::GUS gene expression in CEZ cells upon auxin treatments. To test whether the cantharidin-sensitive phosphatases directly regulate AGR1/PIN2 protein, basipetal auxin transport was measured in the agr1-5 mutant treated with or without 25 μm cantharidin. If the inhibitory effect of 25 μm cantharidin on basipetal auxin transport and DR5::GUS gene expression results from the inhibition of AGR1/PIN2 activity, it is expected that basipetal auxin transport in agr1-5 mutant roots remains at a similar rate independent of the cantharidin treatment. On the contrary, if high doses of cantharidin affect auxin transport processes mediated by proteins other than AGR1/PIN2, basipetal auxin transport in agr1-5 mutant roots should be further reduced by cantharidin treatments. As shown in Figure 3(f), cantharidin treatments did not further reduce the rate of basipetal auxin transport in agr1-5 mutant roots, suggesting that cantharidin at high concentrations directly affects the activity of AGR1/PIN2 protein.
AGR1/PIN2 protein localization was not affected by cantharidin
Previously, we have shown that AGR1/PIN2 protein is predominantly localized at the basal end of root epidermal, cortical and lateral cap cells (Boonsirichai et al., 2003; Peer et al., 2004; also see Figure 4a,e), consistent with an earlier report (Müller et al., 1998). The basal localization of AGR1/PIN2 supports its role in root basipetal auxin transport. To dissect the inhibitory mechanism of cantharidin on auxin transport, we tested whether cantharidin affects AGR1/PIN2 protein localization. For these experiments, 3-day-old seedlings of wild-type plants (both Col-O and Ws ecotypes) were treated with various concentrations of cantharidin for 1.5 h. The localization of AGR1/PIN2 protein was then immunodetected with an affinity-purified anti-AGR1 antibody (Boonsirichai et al., 2003; also see Experimental procedures). As shown in Figure 4(c), treatments with cantharidin at concentrations of up to 50 μm did not change the basal location of AGR1 in roots. We also tested AGR1 localization in the rcn1 mutant. The pattern of AGR1 protein localization in the rcn1 mutant was similar to the wild type treated with or without cantharidin (Figure 4a,c,d). Taken together, these data suggest that altering protein phosphatase activity either by application of specific inhibitors or by mutating the A regulatory subunit of protein phosphastase 2A affects root basipetal auxin transport without affecting AGR1 protein localization.
As pre-treatments with BFA, a vesicle trafficking inhibitor, did not affect root basipetal auxin transport and DR5::GUS gene expression in CEZ cells, we tested the effect of BFA on AGR1 protein localization. We found that BFA at 25 μm caused the accumulation of AGR1 protein in perinuclear endosomal compartments (BFA bodies; Figure 4b,f), in agreement with our previous report (Boonsirichai et al., 2003). In seedlings treated with 50 μm BFA, a large amount of AGR1 protein was accumulated in the BFA bodies (Figure 4g). Despite changes in the AGR1 protein localization in root cells treated with BFA, a significant amount of AGR1 proteins was still detected in the plasma membrane. This might explain why BFA treatments did not affect basipetal auxin transport.
We further examined the effect of cantharidin on the promoter activity of AGR1/PIN2, using an AGR1 promoter GUS fusion reporter construct, as the AGR1::GUS reporter gene expression correlated well with the RNA in situ hybridization data previously reported (Chen et al., 1998). Five-day-old transgenic seedlings expressing the AGR1::GUS reporter gene were treated with DMSO solvent, cantharidin (25 μm), TIBA (50 μm), NPA (50 μm) or IAA (10 μm) for 90 min at room temperature. The GUS activity was histochemically stained. We found that pre-treatments with these chemicals did not change the pattern of AGR1 gene expression when compared with the control (Figure 4h–k; data not shown).
Cantharidin exhibits a dose-dependent inhibitory effect on root gravitropic response
Mutations in the AGR1/PIN2 gene render mutant plants defective in root basipetal auxin transport and root gravitropic response. To directly test the effect of cantharidin on root curvature response upon gravistimulation, we examined the root gravitropic response in wild-type seedlings treated with various concentrations of cantharidin. In the first experiment, seedlings were vertically grown under light on 0.5× MS growth medium, supplemented with 0, 10 and 25 μm cantharidin for 3 days. Plates were photographed and the orientation of root and hypocotyl tips was analyzed. We found that cantharidin at 10 and 25 μm caused both roots and hypocotyls to orient in directions significantly away from the vertical orientation (Figure 5a–c).
The inhibitory effect of cantharidin on the gravity response may be due to a total inhibition of root elongation. To clarify this, we measured root elongation at concentrations of cantharidin used for our gravitropic assays. Our data showed that cantharidin at concentrations of 5, 10 and 25 μm did not completely abolish root growth (Figure 5d). Therefore, the inhibitory effect of cantharidin on root gravitropism was not likely due to a lack of root growth. In addition, seedlings grown on the growth medium supplement with 1% sucrose and 25 μm cantharidin did not exhibit the swelling phenotype reported before (Baskin and Wilson, 1997). The reduced growth and agravitropic root phenotypes of the wild-type plants grown in the presence of 25 μm cantharidin are strikingly similar to the phenotypes of rcn1 a3 double mutants of two regulatory A subunit isoforms of PP2A (Zhou et al., 2004), suggesting that high doses of cantharidin may regulate protein phosphatase activities through multiple regulatory A subunit isoforms.
To quantify the effect of cantharidin on kinetics of root gravitropism of light-grown plants, 4- to 5-day-old vertically grown seedlings were transferred onto growth medium plates supplemented with 0, 1, 5, 10 and 50 μm of cantharidin. After incubation for 1–1.5 h, the plates were rotated by 90°. Root tip angles were then measured every 2 h for a period of 22 h. As shown in Figure 5(e), after 6 h of gravistimulation, the average root tip angle of seedlings treated with 0, 1, 5, 10 and 50 μm of cantharidin was 79 ± 11, 74 ± 11, 35 ± 11, 26 ± 9, and 10 ± 7 from the horizontal orientation, respectively. After 22 h of gravistimulation, the average root tip angle of seedlings treated with 0, 1, 5, 10 and 50 μm of cantharidin was 89 ± 6, 86 ± 3, 50 ± 11, 34 ± 14, 8 ± 5, respectively.
To elucidate the regulatory mechanisms of AGR1/PIN2, an auxin efflux regulator controlling root gravitropism and basipetal auxin transport, we explored the utility of an auxin response reporter system (DR5::GUS; Ulmasov et al., 1997), and a direct auxin transport assay (Rashotte et al., 2000; see Experimental procedures). Our analyses suggest that AGR1/PIN2 mediates the basipetal auxin transport in roots and is, at least partially, responsible for the induction of DR5::GUS reporter gene expression in CEZ cells. The auxin response in CEZ cells that we report here has been observed by other groups using the same auxin response reporter used in this study (e.g. Figure 1b in Hayashi et al., 2003; Figure 3 in Oono et al., 2003), as well as different auxin response reporters (e.g. Figure 7a in Fukaki et al., 2002; Figure 1 in Hayashi et al., 2003; Figure 1 in Oono et al., 2003). However, the cause of the auxin response has not yet been elucidated. Because TIBA, a polar auxin transport inhibitor and a genetic mutation in AGR1/PIN2 greatly reduced this auxin response, we hypothesize that a transient increase in the intracellular auxin concentration in CEZ cells due to AGR1/PIN2-mediated basipetal auxin transport is responsible for the DR5::GUS gene expression in these cells. This hypothesis is consistent with the pattern of the AGR1/PIN2 gene expression and the corresponding protein localization being primarily in DEZ cells, and to a lesser degree in CEZ cells (Figure 4a,i; Figure 5a–c in Chen et al., 1998).
Alternatively, AGR1/PIN2 may regulate the auxin sensitivity of CEZ cells. However, as direct auxin transport measurements indicate that basipetal auxin transport is significantly reduced in agr1-5 mutant roots (Figure 1k), it is likely that the DR5::GUS gene expression in the CEZ results from the AGR1/PIN2-mediated basipetal auxin transport. To investigate regulatory mechanisms that underlie AGR1/PIN2 biological functions, we tested the effects of several chemicals inhibitory of various cellular functions on auxin transport. We found that TIBA (25 μm or higher), a polar auxin transport inhibitor, cantharidin (25 μm or higher) and okadaic acid (0.3 μm), specific inhibitors of protein phosphatases 1 and 2A, significantly reduced the rate of basipetal auxin transport and the auxin-induced DR5::GUS gene expression in CEZ cells. It is noteworthy that the cantharidin concentrations required for inhibiting basipetal auxin transport and auxin response were higher than that required for inhibiting root gravitropism. The different dosage effects of protein phosphatase inhibitors are similar to that of polar auxin transport inhibitors, as the concentration of NPA or TIBA required for significant inhibition of root elongation and auxin transport is higher than that required for their inhibition of root gravitropism (Chen et al., 1998; this study). It is reasonable to speculate that even a slight perturbation in auxin transport would result in a visible agravitropic phenotype. On the other hand, only significant alterations in auxin transport could be detected in the auxin transport assays. Nonetheless, the inhibitory effect on basipetal auxin transport and the DR5::GUS expression in CEZ cells was dependent on a similar dose of cantharidin or TIBA, suggesting a close link between these two processes and implicating the involvement of cantharidin-sensitive protein phosphatases in the regulation of root basipetal auxin transport. Consistent with this hypothesis, a slight increase in auxin transport resulting from a long-term (2 day) treatment of wild-type plants with 10 μm cantharidin, or a mutation in RCN1 gene caused agravitropic defects (Rashotte et al., 2001). In this study using a slightly modified protocol, we found that a short-term (90 min) treatment with cantharidin (5 μm) or okadaic acid (0.03 μm) slightly elevated the rate of basipetal auxin transport (Figure 3a,c). Yet, cantharidin at 5 and 10 μm reduced the root gravitropic response (Figure 5b,d).
Reversible protein phosphorylation has been implicated in regulating auxin transport. Using a tobacco cell culture system, Delbarre et al. (1998) demonstrated that protein phosphorylation was essential to sustain the activity of the auxin efflux carrier, but not the influx carrier. Physiological studies of rcn1 mutants suggested that the reduced protein phosphatase 2A activity through mutations in RCN1 leads to defects in polar auxin transport, root curling and apical hook formation (Deruere et al., 1999; Rashotte et al., 2001). As discussed above, the long-term treatment of wild-type plants with 10 μm cantharidin mimics the phenotypes of rcn1 mutants, including an increased root basipetal auxin transport and altered root gravity response (Rashotte et al., 2001). Our data on the effect of cantharidin at low concentrations are in agreement with the report of Rashotte et al. (2001). However, the effect of cantharidin of higher concentrations on auxin transport has not been investigated previously. We report here that short-term treatments (60–90 min) with cantharidin (25 μm or higher) significantly reduced both root basipetal auxin transport and the auxin-induced DR5::GUS gene expression in CEZ cells, as well as the root gravitropic response. On the other hand, cantharidin at low concentrations slightly elevated basipetal auxin transport even after a short-term treatment (60–90 min). These data suggest that protein dephosphorylation may play two opposite roles in root basipetal auxin transport depending on the concentrations used. The serine/threonine protein phosphatase 2A specifically regulated by cantharidin (Li et al., 1993) consists of a catalytic subunit (C), and regulatory A and B subunits (Zhou et al., 2004). In the Arabidopsis genome, ORFs encoding five C, three A and 17 B subunits have been annotated, with the potential to form 255 PP2A isoforms (Zhou et al., 2004). Double mutant analyses among the three A subunit isoforms indicated that they perform partially overlapping biological functions (Zhou et al., 2004). It is likely that cantharidin at different concentrations regulates activities of different PP2A isoforms. This hypothesis is supported by the following observations. Low concentrations of cantharidin did not affect the rate of basipetal auxin transport in rcn1 mutant roots, suggesting that cantharidin at low concentrations directly targets the regulatory A subunit, RCN1 of PP2A, consistent with the previous report (Rashotte et al., 2001). In contrast, cantharidin at 25 μm significantly reduced auxin transport in the rcn1 mutant. However, the reduction in the rate of auxin transport by high doses of cantharidin is slightly but significantly lower in the rcn1 mutant than in the wild type, suggesting that cantharidin at high concentrations affects RCN1 and other targets.
The rate of basipetal auxin transport in agr1-5 mutant roots was reduced by approximately 30% when compared with the wild type. Treatments of agr1-5 roots with 25 μm cantharidin did not further reduce the rate of basipetal auxin transport, suggesting that AGR1/PIN2 is the direct target of the action of high concentrations of cantharidin. Future experiments aimed at identifying sites of phosphorylation involved in the post-translational modification of AGR1/PIN2 protein will provide clues to elucidate molecular mechanisms by which protein phosphatases regulate auxin transport and root gravitropism.
Surprisingly, the vesicle trafficking inhibitor BFA (up to 50 μm) did not significantly interrupt both basipetal auxin transport and the DR5::GUS gene expression in CEZ cells (Figures 2g and 3f), albeit treatments with BFA at 25 and 50 μm caused the accumulation of a significant amount of AGR1/PIN2 proteins in endosomal compartments (Figure 4b,f,g), and a mild defect in the gravitropic response (Table 1; Geldner et al., 2001; R. Chen, unpublished data). This suggests that the vesicle trafficking pathwayregulated by BFA plays a minor role in root basipetal auxin transport and the expression of the auxin response reporter gene in CEZ cells. However, BFA may affect the lateral auxin transport in the root cap, thereby inhibiting root gravitropism.
Interestingly, the phytotropin NPA (up to 50 μm) did not abolish the auxin-induced DR5::GUS gene expression in CEZ cells, despite the fact that it inhibited basipetal auxin transport by as much as 46% (Figure 3e). How NPA regulates auxin transport is still largely unclear. Previous studies suggested that the auxin efflux activity is mediated by a complex of two to three components, namely catalytic and NPA-binding regulatory subunits, and, possibly, a third labile component (see Lomax et al., 1995 for review). Recently, several NPA-binding proteins have been isolated and implicated in regulating polar auxin transport (Murphy et al., 2002; Noh et al., 2003). Protein–protein interaction studies using a yeast two-hybrid system suggested that the putative auxin efflux carrier PIN1 protein, a paralogue of AGR1/PIN2, does not directly interact with the NPA-binding protein, MDR1 (Blakeslee et al., 2004). Even more surprisingly, loss-of-function mutants of MDR1 and its close homologue AtPGP1 exhibited reduced auxin transport activity in stems, suggesting that proteins encoded by MDR1(AtPGP19) and AtPGP1 may be themselves auxin transporters (Noh et al., 2003).
Taken together, our data agree with previous observations and further suggest that NPA likely acts differently from TIBA, the latter of which may directly inhibit root basipetal auxin transport regulated by AGR1. Alternatively, it is possible that auxin distribution in roots is through more complex pathways than what has been hypothesized including basipetal, acropetal pathways (Lomax et al., 1995), as well as canalization and lateral transport (Blilou et al., 2005; Friml et al., 2003; Ottenschläger et al., 2003).
Flavonoids, plant secondary metabolites, have been implicated as endogenous inhibitors of polar auxin transport (Brown et al., 2001; Murphy et al., 2000). Consistent with their role in regulating auxin transport, the biosynthesis of flavonoids is highly dynamic and sensitive to environmental conditions. Flavonoids displace NPA binding in vitro (Jacobs and Rubery, 1988). Using the same DR5 auxin response construct, Buer and Muday (2004) reported that the basipetal auxin transport is elevated in roots of the flavonoid-deficient tt4(2YY6) mutant. However, the auxin-induced DR5::GUS gene expression in CEZ cells of wild-type plants was very weak compared to the pattern in tt4(2YY6) plants (G.K. Muday, Wake Forest University, NC, USA, personal communication). Flavonoids regulate steady-state transcript levels of PIN genes and the PIN protein localization (Peer et al., 2004). This regulation is further modulated by a variety of environmental cues and auxin (Peer et al., 2004). Investigation of the role of protein phosphorylation and dephosphorylation in auxin transport in flavonoid mutants promises to provide further insights into how auxin transport is regulated in planta.
Plant materials and growth conditions
Seeds of Arabidopsis thaliana (ecotypes Columbia, or Col-O, Wassilewskija, or Ws, and Estland) were obtained from the Arabidopsis Biological Resource Center (Ohio State University, Columbus). Seeds of rcn1 (Ws background) and DR5::GUS transgenic line (Col-O background) were provided by Alison DeLong (Brown University), and Jane Murfett and Thomas Guilfoyle (University of Missouri, Columbia), respectively. agr1-5 line (Estland ecotype) was as reported before (Chen et al., 1998). BC2 seeds of an agr1-5 [DR5::GUS] line were derived from a cross between agr1-5 (male) and DR5::GUS (female) plants and a second backcross of a resulting agr1-5 [DR5::GUS] F2 plant with the DR5::GUS plant. The agr1-5 [DR5::GUS] (BC2) and DR5::GUS lines were used in this study, except when specified otherwise. To construct the AGR1::GUS reporter, a 9897-bp genomic fragment containing the AGR1 coding region and its 5′ and 3′ flanking sequences were cloned into the EcoRI site of pBluescript® II KS- vector (Strategene, La Jolla, CA, USA). The resulting recombinant plasmid was used to amplify a 2132-bp 5′ flanking sequence of AGR1 in a PCR reaction with the following primers, 5′-ATTAACCCTCACTAAAG-3′ (T3 primer) and 5′-ACGCGTCGACGAGATATAGATGAG TATATGTGGTGTG-3′ (an introduced SalI sequence is underlined). The AGR1 5′flanking sequence was then cloned into the SalI site of pBI101.1 (Clontech, Palo Alto, CA, USA) and introduced into Agrobacterium tumefaciens GV3101. Transgenic plants were generated using a vacuum infiltration method. Seeds were surface sterilized by soaking in 20% bleach and 0.1% SDS solution for 12 min and subsequently rinsed with water three to four times. After incubation at 4°C for 1–2 days, seeds were germinated and grown vertically on petri dishes containing 0.5× Murashige minimal organic medium (Invitrogen, Grand Island, NY, USA; cat. no. 23118-060), supplemented with 1.5% sucrose (Sigma, St Louis, MO, USA) and 1.5% Phytagar (Invitrogen; cat. no. 10675-023), in climate-controlled growth rooms (24/20°C day/night temperature, 16/8 h photoperiod, 80 μE sec−1 m−2).
N-(1-naphthyl) phthalamic acid (NPA) was purchased from TCI America (Portlant, OR, USA; cat. no. N0067). 3–5(n)-3H-indoleacetic acid (3H-IAA; 25 Ci mmol−1) was from Amersham (Arlington Heights, IL, USA). 5-bromo-4-chloro-3-indolyl β-d-glucuronide cyclohexylammonium salt (X-Gluc CHA) was from Inalco SPA (Milan, Italy; code no. 1758-0600). Brefeldin A was from Molecular Probes (Eugene, OR, USA; cat. no. B7450) and Sigma (cat. no. B7651). All other chemicals were purchased from Sigma. Stock solutions were made in DMSO solvent and stored in −20°C. DMSO (0.1%) was used as mock control.
Gravitropism assay was as described previously (Chen et al., 1998). Briefly, 4-day-old vertically grown seedlings under light were transferred to new petri dishes containing the same growth medium supplemented with chemicals at concentrations indicated in the text. After incubating vertically for 1 h, seedlings were rotated by 90° and images of the displacement of root tips over a period of 22 h was captured with a Nikon Coolpix 4500 digital camera. The angle of curvature was measured from the digitized images using an NIH Image Analysis software.
GUS histochemical staining
GUS histochemical staining was as described previously (Boonsirichai et al., 2003). Briefly, 3- to 5-day-old light-grown seedlings were incubated in a GUS staining solution [100 mm Tris, pH7.5; 29 mg ml−1 NaCl; 1 mm K3Fe(CN)6; 20% MeOH] for 15 min to 2 h. The enzymatic reaction was stopped by transferring seedlings to a clearing solution (20% methanol and 1 N HCl) followed by incubation at 55°C for 30 min. Seedlings were then treated with a neutralization solution (7% NaOH, 60% EtOH) and incubated at room temperature overnight. Seedlings were subsequently treated with 30 and 10% EtOH for 30 min each and finally to 10% EtOH, 50% glycerol solution before microscopic observations.
Immunofluorescence protein localization
Immunolocalization of AGR1/PIN2 proteins was as described by Boonsirichai et al., (2003) with modifications. Briefly, 4-day-old seedlings were fixed in 4% (w/v) p-formaldehyde in PME buffer (50 mm PIPES-KOH, pH 6.9, 5 mm MgSO4, 10 mm EGTA), for 1 h at room temperature or overnight at 4°C. Samples were then incubated for 30 min in 0.5% (w/v) macerozyme R-10 and 0.1% (w/v) pectolyase Y-23 in PME for partial digestion of cell walls. Primary and secondary antibodies were diluted in PME containing 3% (w/v) BSA. Anti-AGR1 polyclonal antibodies were raised against a purified polypeptide containing amino acids 379–503 of the AGR1 protein, affinity-purified and used at 1:200 dilution. Fluorescein (FITC)-conjugated goat anti-rabbit IgG (Jackson ImmunoResearch, West Grove, PA, USA) was used at 1:300 dilution. After three 20-min washes in PME, samples were mounted in Vectashield (Vector Laboratories, Burlingame, CA, USA) and imaged with a MRC-1024 ES laser confocal scanning microscope (Bio-Rad, Hercules, CA, USA). Single optical sections were collected and Kalman averaged three times. Fluorescein was detected by exciting samples with the 488-nm laser line and measuring the emission at 522 nm. The captured images were pseudocolored green and processed/assembled using Adobe Photoshop 5.0 L.E. (Adobe Systems, Inc., Mountain View, CA, USA).
Semiquantitative RT-PCR assay for GUS expression
Seeds were germinated and grown vertically under light for 5 days on 0.5× MS medium, containing 1.5% phytagar and 1% sucrose. The seedlings were then transferred onto freshly prepared agar plates containing 10 μm IAA and incubated for 60, 90 and 120 min. For RNA extraction, root tissues were harvested from approximately 100 seedlings after removal of shoots, and root tips of approximately 0.3 mm in length. Total RNAs were isolated using the TRI reagent according to the manufacturer's instructions. The first-strand cDNAs were synthesized from 1 μg of total RNAs using Superscript II reverse-transcriptase (Invitrogen, Carlsbad, CA, USA) in a total reaction volume of 20 μl. One microliter of the cDNA mixture was used as the template in a 50 μl PCR containing gene-specific primers and Ex-Taq DNA polymerase (Takara, Madison, WI, USA).
To standardize the PCR amplification, all PCR mixtures included primers (5′-GCAAGAGAATCTTCTTAGGGGTATCTATGC-3′ and 5′-GGTGGGAGAAGCTGGAATATGTCATAG-3′) to amplify a 492bp-long eIF-4A (GeneBank ID: AC005287) as an internal control and primers (5′-ACGTCCTGTAGAAACCCCAA-3′ and 5′-CCCGCTTCGAAACCAATGCC-3′) for the amplification of GUS transcripts. A series of PCR reactions with a variable number of cycles were performed. Results from 20 cycles were reported. The amplified products were separated on 1% agarose gel, stained with ethidium bromide (10 μg ml−1), and captured as digital images under UV light. Fluorescence intensities of individual band were quantified from the digital images using LabWorks software. Transcript levels of the GUS gene were normalized against the eIF-4A internal control.
Basipetal auxin transport assay
Basipetal auxin transport assay was as described previously (Rashotte et al., 2000), with the following modifications. Four-day-old light-grown seedlings were transferred to growth medium plates supplemented with various concentrations of drugs and incubated for 60–90 min. Agar blocks of 1 mm diameter supplemented with 7.7 × 10−8m3H-IAA were placed immediately next to the root tips. After incubation for 1.5 h, agar blocks were removed. Root tips of approximately 0.3 mm in length were excised and discarded. Two consecutive 2 mm root segments were then excised from the remaining root tips and placed separately into two scintillation vials each containing 5 ml of scintillation fluid. The amount of radioactivity in the two root segments pooled from 6–10 seedlings was measured using a Beckman Coulter (Fullerton, CA, USA) LS6500 scintillation counter. The amount of the radioactivity was the average of three separate experiments ±standard deviation. Student's t-test (paired with two-tailed distribution) was used to evaluate statistical significance of the data.
We thank Tom Guifoyle and Jean Murfett (University of Missouri), Alison DeLong (Brown University), and the Arabidopsis Biological Resource Center (ABRC, Ohio State University), for providing DR5::GUS, rcn1, and wild-type seeds, respectively; Marilyn Roossinck, Alison DeLong, Hong-Liang Wang and two anonymous reviewers for insightful comments and suggestions on the manuscript. Financial support was from the Noble Foundation to R.C. and E.B.B., and the National Aeronautics and Space Administration to P.H.M. and E.B.B.