Arabidopsis H+-PPase AVP1 Regulates Auxin-Mediated Organ Development

See allHide authors and affiliations

Science  07 Oct 2005:
Vol. 310, Issue 5745, pp. 121-125
DOI: 10.1126/science.1115711


The transport of auxin controls developmental events in plants. Here, we report that in addition to maintaining vacuolar pH, the H+-pyrophosphatase, AVP1, controls auxin transport and consequently auxin-dependent development. AVP1 overexpression results in increased cell division at the onset of organ formation, hyperplasia, and increased auxin transport. In contrast, avp1-1 null mutants have severely disrupted root and shoot development and reduced auxin transport. Changes in the expression of AVP1 affect the distribution and abundance of the P–adenosine triphosphatase and Pinformed 1 auxin efflux facilitator, two proteins implicated in auxin distribution. Thus, AVP1 facilitates the auxin fluxes that regulate organogenesis.

The phytohormone auxin [principally indole acetic acid (IAA)] plays a fundamental role in the formation of all plant organs, and gradients of auxin have been shown to be essential to polarity of development (1, 2). Because IAA is a weak acid (pKa 4.75, where Ka is the acid dissociation constant), a chemiosmotic model describes polar auxin uptake and efflux driven by a plasma membrane H+ gradient (3). In the acidic apoplast, an enrichment of the lipophilic protonated species of IAA facilitates its entry into the cell. The same gradient motivates efflux of anionic (non–lipid soluble) IAA retained in the neutral cytoplasm by means of polarly localized efflux complexes (4). Thus, the transporters responsible for setting cytoplasmic and apoplastic pH are likely to have key roles in driving this polar flux of auxin.

Plants have three distinct membrane H+-pumps capable of generating pH gradients (5). The P-type H+–adenosine triphosphatase (P-ATPase) is a single-subunit protein that energizes transport across the plasma membrane (PM) by extruding H+ from the cell (6). The vacuolar H+-ATPase (V-ATPase) complex, encoded by at least 26 genes, acidifies the vacuole and other intracellular trafficking compartments and is known to be required for embryonic development and cell expansion (7, 8). H+-pyrophosphatases (H+-PPases) are single-subunit proteins that also generate proton gradients in endomembrane compartments with the use of pyrophosphate (PPi) instead of ATP (9). The Arabidopsis genome contains one type I H+-PPase, AVP1 (10), and one type II, AVP2/AVPL1, which shares 35% amino acid identity with AVP1 (11). AVP1 has been traditionally viewed as a vacuolar H+-pump and has not previously been implicated in hormone transport or developmental regulation.

However, we observed that alterations in AVP1 expression produce plants with morphogenetic variations typical of hormonal defects. Thus, the AVP1 overexpressing plants (AVP1OX) AVP1-1 and AVP1-2 had more rosette leaves (3 and 16, respectively; table S1) and significantly greater leaf area (60 and 40%, respectively, P < 0.05) than wild-type plants (Fig. 1B and fig. S1) as a result of increased cell numbers (figs. S2 and S3). Furthermore, AVP1-1 and AVP1-2 exhibited enhanced root growth (Fig. 1C) and dry weight (2.6 and 9.4 heavier, P < 0.01; table S2) compared with wild-type plants. Thus, AVP1 appears to function in both shoot and root development.

Fig. 1.

Phenotypes associated with AVP1 gain- and loss-of-function mutants. (A) Seven-day-old avp1-1 plants grown on plates. Scale bar, 2.5 mm. (B) Rosettes from 45-day-old, soil-grown Col-0 (WT) and AVP1OX plants. (C) Root systems from 50-day-old, hydroponically grown Col-0 and AVP1OX plants. Scale bar in (B) and (C), 1 cm. (D) A 70-day-old avp1-1 plant grown on plates. (E) avp1-1 inflorescence. Scale bar in (D) and (E), 1 mm. (F) Wild-type flower. Scale bar, 0.5 mm. (G) Immunoblots of Col-0 and avp1-1 membrane proteins probed with antisera to AVP1 and V-ATPase. (H and I) Root apices from 7-day-old Col-0 (H) and avp1-1 (I). m, meristem; ci, columella initial; S1 to S3, stories of columella cells; tc, tip cells. (J) A 35-day-old avp1-1 root apex. (K) Root of AVP1-RNAi plant 7 days after induction. Scale bar for (H) to (K), 20 μm. (L and M) Six-day-old roots from Col-0 (L) and avp1-1 (M) grown on plates. (N) A 6-day-old Col-0 root grown with 25-μM NPA. Scale bar for (L) to (N), 200 μm. (O) avp1-1 mutant is complemented with pINDEX3-AVP1 construct after 12 days of induction. Plant at day 1 (inset). Scale bar, 0.5 cm.

To further assess the role of AVP1 in organ development, we analyzed avp1 loss-of-function mutants. avp1-1 is a recessive mutant with a transferred DNA insertion in the predicted fifth exon of the AVP1 gene. This lesion impairs full-length transcription (fig. S4). Consistent with avp1-1 being a knockout allele, the 81-kD AVP1 protein was not detected in Western blots of microsomal fractions isolated from avp1-1 homozygotes. (Fig. 1G). However, a faint band at ∼84-kD was observed in the avp1-1 microsomal fraction, consistent with detection of AVP2/AVPL1, which is ∼3 kD larger than AVP1 (11). Immunoblot analysis with an antibody raised against the V-ATPase B subunit detected the expected 60-kD band in both wild-type and avp1-1 plants (Fig. 1G).

Homozygous avp1-1 seedlings showed altered root and shoot development: Cotyledons were either normal (∼1%), cup-shaped (∼1%), or heart-shaped (∼98%) similar to cup-shaped cotyledon mutants (12) (Fig. 1A). Only 30% of the avp1-1 plants initiated flower development, and none developed fully (Fig. 1, D to F). Most formed small cuplike structures (Fig. 1E) that had no floral organs or occasionally contained a small pistil (<10%) (13).

Consistent with the increase in leaf size due to increased cell number seen in AVP1OX, rosette leaf size in avp1-1 was ∼20% that of the wild-type leaf size (Fig. 1D). However, the average size of avp1-1 mesophyll cells was unchanged from that of the wild-type cells (Col-0 = 782 ± 196 μm2, n = 476; avp1-1 = 741 ± 343 μm2, n = 251, P > 0.05, Student's t test, errors are SD), suggesting a reduction in cell number in this mutant. Vascular patterning and mesophyll cell organization along the leaf blade were also disordered, resulting in thicker leaves with uneven margins (Fig. 1D and fig. S5). No abnormalities in cell organization were visible in shoot apical meristems in avp1-1 during early development (fig. S6), but disrupted growth later became evident (Fig. 1, D and E).

The Arabidopsis root tip is a highly ordered structure with a well-defined root cap, three columella cells layers, and a layer of peripheral or tip cells covering the apical meristem (14) (Fig. 1H). However, the cells of the avp1-1 root apex were small, the columella was reduced to two layers, and the root cap was reduced and deformed (Fig. 1I), eventually collapsing (Fig. 1J), as reported in the auxin-related pin4 and tt4 mutants (15, 16). Disrupted cell elongation was also evident in the upper portions of avp1-1 roots (Fig. 1, L and M) which exhibited swelling similar to wild-type roots treated with the auxin-efflux inhibitor N-1-naphthylphthalamic acid (NPA) (Fig. 1N).

Comparison with AVP1 RNA interference (RNAi) lines and complementation confirmed that the avp1-1 phenotypes were due to a lesion in the AVP1 gene. Thus, AVP1 RNAi lines exhibited impaired shoot development and disordered root cell patterns similar to avp1-1 (Fig. 1K and fig. S7). Further, wild-type shoot and root phenotypes were restored in dexamethasone (DEX)–treated homozygous avp1-1 seedlings (Fig. 1O, inset) carrying a DEX-inducible AVP1 cassette (Fig. 1O and fig. S8). These observations strongly suggest that the organogenesis-related phenotypes seen in avp1-1 are attributable to the lesion in AVP1.

The phenotypes of both gain- and loss-of-function AVP1 mutants suggest that AVP1 functions in organ development. In situ hybridization revealed that AVP1 is expressed in all shoot meristems as well as the endodermal/pericycle ring of mature roots, developing leaves, sepals, petals, stamens, and carpels (Fig. 2, A to J). The reported expression patterns of the CAMATA and VOZ transcription factors (17) predicted to bind a 38–base pair cis-acting region of the AVP1 promoter are also consistent with these AVP1 expression patterns.

Fig. 2.

AVP1 expression patterns at the messenger RNA and protein levels. [(A), (C), (D), (F), (H) to (J)] AVP1 antisense probe. [(B), (E), (G)] AVP1 sense probe. Scale bar, 50 μm. (A) AVP1 is expressed in the parenchyma cells surrounding the vasculature in roots. (B) No signal was detected with AVP1 sense probe. (C) AVP1 is expressed in the shoot apical meristem and young leaf primordia at 13 days but restricted to the adaxial side of older leaves (D). (E) No signal was detected with AVP1 sense probe. (F) AVP1 is expressed in inflorescence meristems, floral primordia, and procambium cells (arrows). (G) No signal was detected with AVP1 sense probe. (H) Longitudinal section of a stage 6 flower. (I) AVP1 expression remains high in petals, anthers, and carpels in stage 10 flowers. (J) At stage 12, AVP1 is only detected in petals and carpel. SAM, shoot apical meristem; lp, leaf primordia; St 2, stage 2 floral meristem; St 4, stage 4 flower; IM, inflorescence meristem; Se, sepal; Pe, petal; St, stamen; Ca, carpel. (K to M) Confocal immunohistochemical images of 5-day-old seedlings. Scale bar, 50 μm. (K) AVP1 localization is punctuate in Col-0 root tip and at the plasma and vacuolar membranes in epidermal cells (inset). (L) AVP1 signal is increased in AVP1-2 root tip and epidermal cells (inset). (M) AVP1 signal is not observed in avp1-1 root tips.

To better understand how a H+-PPase could impact developmental programs, we analyzed the subcellular distribution of AVP1. In addition to vacuolar membrane (tonoplast) localization, immunolabeling revealed a punctuate AVP1 signal throughout wild-type and AVP1-2 young roots (Fig. 2, K and L). This signal was absent in avp1-1 (Fig. 2M) and strongly increased in cortical cells above the elongation zone in AVP1OX roots (Fig. 2, K and L, insets). Western blots of wild-type and AVP1OX root microsomal fractions separated by discontinuous sucrose density gradients detected AVP1 signal in tonoplast, endosome, and PM fractions (Fig. 3, A and B). The PM and the endosomal AVP1 signals in AVP1OX roots were 60 and 80% greater than Col-0, respectively (Fig. 3, A and B). Microsomal fractions from AVP1OX plants also exhibited increased PM P-ATPase protein abundance (∼70%) and activity (∼100%) (Fig. 3, C and D), whereas the decreased P-ATPase activity observed in avp1-1 seedlings (Fig. 3D) was consistent with a reduced P-ATPase signal detected in immunofluorescence imaging of avp1-1 root tips (Fig. 3F).

Fig. 3.

P-ATPase levels and activity in AVP1 gain- and loss-of-function mutant plants. (A) Microsomal fractions from Col-0 (WT) and AVP1-2 hydroponically grown roots were separated on discontinuous sucrose density gradient (22% tonoplast, 32% endosomes, and 38% plasmalemma) and immunoblots analyzed with polyclonal antisera to the PPi binding site of AVP1 and the P-ATPase and monoclonal antibodies to the V-ATPase B subunit (26). Relative densities of AVP1 (B) and P-ATPase (C) were quantified with Bio-Rad Quantity One software. Values are means ± SD of three independent experiments. (D) P-ATPase activity (micromol Pi per milligrams protein per minute) of microsomal fractions from Col-0, avp1-1, and AVP1-2. Data are means ± SD of three independent experiments. (E) Apoplastic pH was determined at the root elongation zone from Col-0, avp1-1, and AVP1-2 6-day-old seedlings with the use of Oregon Green. Values represent means ± SD of five measurements from ≥10 individual plants. (F) P-ATPase immunohistochemical localization at the plasma membrane in 5-day-old Col-0 and avp1-1 root tips. Scale bar, 50 μm.

The phenotypes of AVP1 gain- and loss-of-function mutants are consistent with alterations in auxin-mediated development but apparently do not reflect altered auxin synthesis. Total free auxin levels in AVP1OX and avp1-1 whole seedlings were comparable to those in the wild-type seedlings (13). However, the increased lateral root proliferation and rates of gravitropic bending observed in AVP1OX (fig. S10) are consistent with enhanced cellular auxin uptake and efflux (18). Consistent with this interpretation, free auxin levels in hypocotyls of AVP1OX were ∼160 ± 17% those of the wild type. Altered auxin uptake was also suggested when we observed that avp1-1 seedlings treated with either IAA or the synthetic auxin naphthalene acetic acid (NAA) formed calli rather than the hairy root systems usually seen in auxin-treated wild-type plants (Fig. 4A and fig. S11).

Fig. 4.

Auxin transport, PIN1 localization and trafficking model in AVP1 gain- and loss-of-function seedlings. (A) Shoot and root development of 53-day-old Col-0 (WT) and avp1-1 plants with 5 μM IAA. Scale bar, 1 cm. GUS staining in 6-day-old roots of Col-0 DR5::GUS (B) and avp1-1DR5::GUS (C). GUS staining in 6-day-old cotyledons of Col-0 DR5::GUS (D) and avp1-1DR5::GUS (E). Scale bar in (B) to (E), 50 μm. (F) Basipetal IAA transport from shoot apex to root-shoot junction and root tips of 5-day-old avp1-1 and AVP1-1 seedlings. Values are percentage of Col-0 control, and means ± SD of two assays of 10 seedlings each. PIN1 (G) localization in root tips of 5-day-old Col-0, avp1-1, and AVP1-1. Scale bar, 50 μm. AVP1 function regulates trafficking of H+-ATPase and PIN1. (H) Loss of AVP1 function results in decreased H+-ATPase and PIN1 abundance on the PM; overexpression results in increased abundance; H+-ATPase and PIN1 traffic in the same vesicles (23). Resultant changes in pH alter chemiosmotically driven auxin flux through the cells (arrows). In avp1-1, auxin levels do not fall below the threshold required for increased PIN1 redistribution (16).

To further investigate the effects of AVP1 activity on auxin distribution, we used the synthetic auxin-responsive DR5::GUS reporter (19). Wild-type seedlings transformed with DR5::GUS exhibited GUS signal in root apices, vascular tissue, and cotyledonary margins (Fig. 4, B and D) (2). In contrast, the DR5::GUS signal in avp1-1 roots was restricted to the meristem with no signal in the stele (Fig. 4C). Notably, a strong DR5::GUS signal was observed in the abnormally developed vascular systems in the avp1-1 fused cotyledons (Fig. 4E). These aberrant DR5::GUS patterns and altered vascular development observed in avp1-1 roots and cotyledons, respectively, further suggested to us that alterations in auxin transport and accumulation were occurring when AVP1 expression was altered (20).

To directly test this hypothesis, basipetal transport of IAA from the shoot tip to both the root-shoot junction and the root tip was assayed in AVP1 overexpression and loss-of-function mutants. Auxin transported to AVP1-1 root tips was 46% greater than in wild-type plants (P < 0.05) (Fig. 4F); transport to the root-shoot junction was approximately 20% less than in wild-type plants (P < 0.05), suggesting a higher rate of auxin transport in AVP1OX seedlings. Conversely, transport of auxin to the root tip and the root-shoot junction was significantly reduced in avp1-1 seedlings (27 and 50%, respectively, of transport in the wild-type seedlings; P < 0.05) (Fig. 4F). These data are consistent with the DR5::GUS patterns and suggest that auxin transport is responsive to AVP1 levels.

PM-localized H+-PPases are unlikely to function directly in apoplastic acidification (21). However, the altered P-ATPase abundance and activity seen with AVP1OX and avp1-1 plants could alter apoplastic pH and affect intracellular proton homeostasis and gradients and so alter the driving forces for auxin transport. Vacuolar, cytosolic, and apoplastic pH were therefore measured in root elongation zone cells of wild-type, AVP1OX, and avp1-1 seedlings to test whether alterations in AVP1 expression might directly or indirectly influence pH homeostasis. The apoplastic pH was significantly more acidic in AVP1OX (pH 5.1 ± 0.05) and more alkaline in avp1-1 (pH 5.8 ± 0.06) than in wild-type plants (pH 5.6 ± 0.05) (P < 0.05, Student's t test) (Fig. 3E). No difference in cytosolic pH was detected (fig. S9A). However, the vacuolar pH of the avp1-1 seedlings was more alkaline than that in wild-type seedlings (P < 0.05, Student's t test) (fig. S9B). The observed changes in P-ATPase activity and cell wall acidification suggest an explanation for at least some of the observed avp1-1 and AVP1OX phenotypes. Thus, altered apoplastic acidity in avp1-1 and AVP1OX would be expected to result in consequent changes in chemiosmotic motivation of polar auxin transport. Furthermore, because at normal apoplastic pH (5.5), only ∼16% of the extracellular IAA is protonated, even a small change in pH would be expected to have a pronounced effect on IAA uptake.

In addition to apoplastic acidification, other regulatory factors involved in auxin transport might be affected by altered AVP1 activity. Alterations in speciation and distribution of flavonoids are thought to regulate auxin transport (16). Altered flavonoid accumulations were observed in avp1-1 but appeared to be a response to altered auxin accumulations and PM P-ATPase activity (22) rather than a factor contributing to altered auxin transport (fig. S12).

The Pinformed 1 (PIN1) facilitator component of the auxin efflux carrier is also essential for polar auxin transport from the shoot tip to the root tip (4) and correlations between PIN-dependent auxin efflux, auxin gradients, and primordia development have been documented in organ formation (1). PIN1 has been shown to dynamically cycle together with P-ATPases between the PM and endomembrane compartments in a manner that is sensitive to auxin efflux inhibitors, the membrane trafficking inhibitor brefeldin A, and levels of auxin itself (4, 23, 24). PIN2 is also dynamically cycled, but by a different mechanism. To determine whether AVP1 activity influences subcellular localization of auxin efflux carriers, PIN1 and PIN2 were immunolocalized in root tips. PIN2 localization and abundance were similar to those in the wild-type plants (fig. S13). PIN1 distribution was also similar to that in the wild-type plants (23, 24) in avp1-1, but the signal intensity was decreased, suggesting decreased plasma membrane abundance. In AVP1OX, the PIN1 signal intensity increased, but was restricted to the apical portion of the vascular cylinder (Fig. 4G). PIN1 expression in avp1-1, as determined by quantitative real-time fluorescence polymerase chain reaction, was not different from expression in wild-type plants; however, it was decreased in AVP1OX (13). The AVP1OX results are consistent with auxin-responsive reduction of PIN1 expression in nonvascular tissues and increased PM stabilization of PIN1 and P-ATPases observed in root apical tissues (16, 24). Furthermore, the decreased PIN1 signal observed in avp1-1 cannot be attributed to reduced auxin transport to the root, as this is expected to result in an increased PIN1 signal (16). Instead, this result suggests that loss of AVP1 activity directly affects PIN1 trafficking or stability. A direct role in the increased PM abundance of PIN1 seen in AVP1OX vascular tissues also cannot be ruled out.

The data presented here indicate that although classically thought of as a tonoplast resident proton pump responsible for acidifying the vacuole, AVP1 also contributes to the regulation of apoplastic pH and to auxin transport, likely by mediating the trafficking of the PM P-ATPase and associated proteins, including PIN1 (Fig. 4H). Changes in intracellular auxin levels are also known to alter the expression of P-ATPase genes (25), establishing a feedback loop where AVP1 activity can regulate both targeting and level of the PM proton pump. Thus, in addition to its established role in the maintenance of vacuolar pH, our data reveal a previously unrecognized role for AVP1 in facilitating auxin transport and the regulation of auxin-related developmental processes.

Supporting Online Material

Material and Methods

Figs. S1 to S13

Tables S1 to S3


References and Notes

View Abstract

Navigate This Article