• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of pnasPNASInfo for AuthorsSubscriptionsAboutThis Article
Proc Natl Acad Sci U S A. Feb 15, 2005; 102(7): 2649–2654.
Published online Feb 4, 2005. doi:  10.1073/pnas.0406377102
PMCID: PMC548969
Plant Biology

A plasma membrane H+-ATPase is required for the formation of proanthocyanidins in the seed coat endothelium of Arabidopsis thaliana


The plasma membrane in plant cells is energized with an electrical potential and proton gradient generated through the action of H+ pumps belonging to the P-type ATPase superfamily. The Arabidopsis genome encodes 11 plasma membrane H+ pumps. Auto-inhibited H+-ATPase isoform 10 (AHA10) is expressed primarily in developing seeds. Here we show that four independent gene disruptions of AHA10 result in seed coats with a transparent testa (tt) phenotype (light-colored seeds). A quantitative analysis of extractable flavonoids in aha10 seeds revealed an ≈100-fold reduction of proanthocyanidin (PA), one of the two major end-product pigments in the flavonoid biosynthetic pathway. In wild-type seed coat endothelial cells, PA accumulates in a large central vacuole. In aha10 mutants, the formation of this vacuole is impaired, as indicated by the predominance of multiple small vacuoles observed by fluorescence microscopy using a vacuole-specific dye, 5-(and -6)-carboxy 2′,7′-dichlorofluorescein diacetate. A similar vacuolar defect was also observed for another tt mutant, tt12, a proton-coupled multidrug and toxic compound extrusion transporter potentially involved in loading provacuoles with a flavonoid intermediate required for PA production. The endothelial cells in aha10 mutants are otherwise healthy, as indicated by the lack of a significant decrease in (i) the accumulation of other flavonoid pathway end products, such as anthocyanins, and (ii) mRNA levels for two endothelium-specific transcripts (TT12 and BAN). Thus, the specific effect of aha10 on vacuolar and PA biogenesis provides genetic evidence to support an unexpected endomembrane function for a member of the plasma membrane H+-ATPase family.

Keywords: flavonoid, P-type ATPase, proton pump

Proton pumps are used in cells to energize specific membranes with an electrochemical potential and acidify compartments or an extracellular space. In plant cells, three types of proton pumps [P-type, V-type, and pyrophosphatase (PPase)] are all translated on rough endoplasmic reticulum and move through trafficking pathways to provide proton-pumping activity at diverse locations, including the plasma membrane, vacuoles, Golgi, and endosomes. The P-type H+-ATPases are thought to function exclusively at the plasma membrane (1). The V-type and PPase pumps are most abundant at endomembrane locations and are thought to function primarily in the acidification of vacuoles. Here, we provide genetic evidence for an unexpected function of a plasma membrane P-type H+-ATPase on an endomembrane system.

The P-type ATPase superfamily has five distinct branches (Fig. 1a) and includes pumps that translocate Na+/K+, H+, Ca2+, and heavy metals. They all share a common enzymatic mechanism involving a phospho–aspartate intermediate. The H+-ATPase branch (P3a) is found in plants and fungi but not in animals. Plants and fungi use H+-ATPases to energize their plasma membranes with an electrochemical gradient, but animals use a Na+/K+-ATPase. In Arabidopsis and rice, there are 11 and 10 proton-pump isoforms, respectively, which can be divided into five clusters (2, 3). The functional differences between these clusters have not been determined but may delineate biochemical functions or tissue specificity. The only representative of cluster 3 in Arabidopsis is autoinhibited H+-ATPase isoform 10 (AHA10). This isoform was previously shown to be expressed primarily in the integument layers of developing seeds (4), suggesting a unique seed-specific function.

Fig. 1.
The Arabidopsis P-type ATPase superfamily and map of AHA10. (a) Relationship tree showing the autoinhibited H+ P-type ATPase (AHA) family in Arabidopsis. The numbers represent the subfamilies of the plant AHAs identified by Arango et al. (2) and Baxter ...

In the development of the Arabidopsis seed, the embryo is nourished and protected by the maternally derived integument layers (5). In early stages of development, transfer cells from the integument are thought to be essential in the source-sink translocation of nutrients. In later stages, integument layers differentiate into a seed coat endothelium that protects the dormant embryo from extreme environmental stresses (e.g., cold, desiccation, and intense sunlight) until conditions are appropriate for germination. One important aspect of this differentiation is the synthesis of flavonoids, which are critical to UV protection, seed coat hardening, and desiccation tolerance (6, 7). In seeds, the two most abundant flavonoid end products are proanthocyanidins (PAs) and anthocyanins, whose biosyntheses diverge after the production of cyanidin (8). Both products accumulate in a large central vacuole within seed coat endothelial cells (9).

The synthesis and function of seed coat flavonoids has been actively investigated in Arabidopsis by using forward genetics. Mutations in this pathway have been identified by screening for mutants producing pale-colored seeds [transparent testa (tt) phenotype] and/or lacking condensed tannins [tannin deficient seed (tds) phenotype] (1012). The 15 mapped tt/tds mutants include six regulatory transcription factors, seven enzymes in the flavonoid biosynthesis pathways, and one GST. Of these 15 mutants, only one (tt12) disrupts a membrane transporter. The TT12 transporter is a member of the multidrug and toxic compound extrusion (MATE) family of transporters (13) and is proposed to function as a H+-coupled antiporter to load flavonoids (probably an intermediate specific to the PA pathway) into vacuoles within the developing seed coat endothelial cells. Evidence from maize mutants indicates that other flavonoids, such as anthocyanin, are transported into vacuoles or provacuolar compartments through an ATP-binding cassette (ABC)-type transporter (MRP3) (14), which is directly driven by ATP hydrolysis and does not depend on a H+ gradient across the membrane. At present, the MATE (TT12) and ABC-type transporter (MRP3) are the only two flavonoid transporters characterized at a genetic level (15).

Here, we report a tt phenotype associated with four independent T-DNA (transfer DNA) gene disruptions of AHA10, a member of the plasma membrane P-type proton-pump gene family. This tt phenotype results from a dramatically reduced accumulation of PA. At present, the only other loss-of-function knockout reported for a plant P-type H+-ATPase results in pollen lethality (39), supporting the expectation that plasma membrane proton pumps are essential for plant cells (16). In contrast, the aha10 mutation does not appear to affect the viability of seed coat endothelial cells, suggesting that these cells express other isoforms that provide a genetic redundancy for essential plasma membrane proton-pump functions. Instead, the phenotype of the aha10 seed coat endothelial cells reveals a specific disruption in (i) the maturation of a flavonoid storage vacuole and (ii) the accumulation of PA, one of the two most abundant end-product flavonoid pigments. These observations suggest that at least one member of the plasma membrane proton-pump family (i.e., AHA10) has an alternative function in an endomembrane location.

Materials and Methods

Plant Lines and Growth Conditions. Plant lines containing T-DNA insertions in AtAHA10 (At1g17260) in the Wassilewskija (Ws) background were identified from the Wisconsin collection by PCR screening (aha10-1 and aha10-2) and were confirmed by sequencing (17). Two additional lines in the Columbia (Col) background (aha10-4, SALK_019569 and aha10-5, SALK_048598.53.35) were obtained from the SIGnAL collection (18) and verified by PCR analysis with the left border primer. Seed for a tt12 mutant (Ws ecotype) was obtained from the Arabidopsis Biological Resource Center (CS5740). Plants were grown in soil under 24-h light conditions in a growth room at 20°C. All experiments were conducted on wild-type and mutant siliques or on seeds from plants that were grown side-by-side in the same soil container.

Complementation. The AHA10 gene was cloned as an XhoI/SpeI fragment from a genomic cosmid clone pcAXA1 (4) into a pBin vector (PS249) encoding a glufosinate-ammonium pestanol (BASTA)-resistance gene for plant expression (19). Mutant aha10-1 plants were transformed by using an Agrobacterium-mediated floral dip infiltration method (20). The T1 transgenic plants were selected on 0.5× Murashige and Skoog medium containing 10 μg/ml BASTA.

p-Dimethylaminocinnamaldehyde (DMACA) Assay. Seeds were incubated in 2% (wt/vol) DMACA in 3 M HCl/50% (vol/vol) methanol for 4 days (10). The seeds were then washed four times in 70% ethanol and visualized with a light microscope.

Flavonoid Quantitation. Nonanthocyanin flavonoids were extracted from seeds and analyzed as described in refs. 9 and 10, except that HPLC fractions were separated and analyzed by using a Supelco C18 column on a Waters Alliance 2900S HPLC with diode array monitoring of 200–600 nm. Briefly, the seeds were pulverized and extracted with acetone, Folch-partitioned, and extracted with hexane. For the quantitation of PA, extracts were further subjected to acid hydrolysis in the presence of phloroglucinol (J. T. Baker). Epicatechin analysis was conducted before phloroglucinol hydrolysis, because this hydrolysis releases the terminal epicatechin of the PA polymer. Authentic quercetin, kaempferol, and catechin standards were obtained from Sigma. Cyanidin and pelargonidin were a gift from Dr. Ralph Nicholson (Purdue University), and epicatechin was extracted from black tea (21). Cyanidin-based standards procyanidin B2 and B3 and pelargonidin-based standard pelargonidin chloride for phloroglucinol analysis were obtained from standard collections of Dr. Ralph Nicholson or from ChromaDex (Santa Ana, CA). We did not observe any peaks from Arabidopsis PA corresponding to catechin- or pelargonidin-derived adducts. Although the extraction of flavonoids from mature seeds was not complete, as indicated by residual pigment that was recalcitrant to solubilization, the ratios of flavonoids extracted in mature and developing seeds were the same (data not shown). The residual pigment was not extractable with isopropanol, chloroform/methanol/water 1:1:1, or ethyl acetate.

The anthocyanins were extracted from seeds by grinding in methanoic acid and were allowed to sit overnight at 4°C in the dark, Folch-partitioned (22), and analyzed by HPLC as described above.

Visualizing Vacuoles. Vacuoles were observed by using the fluorescent marker 5-(and -6)-carboxy 2′,7′-dichlorofluorescein diacetate (Molecular Probes). Seeds were collected from mature green siliques in which embryos had developed past the torpedo stage (i.e., early mature stage). Seeds were dissected and prepared as described in ref. 9, and images were observed on a Type DM/RB microscope (Leica, Deerfield, IL). A Nikon E5000 digital camera was used to capture images.

Quantitative RT-PCR. Total RNA was extracted from the first 12 green siliques closest to the apical meristem by using RNAEasy kits (Qiagen, Valencia, CA). First-strand cDNA was synthesized by using a poly(dT) primer and Moloney murine leukemia virus reverse transcriptase (Promega) according to the manufacturer's protocol. Taqman assays (Applied Biosystems) were performed on a Bio-Rad MyIQ Single Color Real Time PCR Detection system (Bio-Rad). Amplification reactions for transcripts encoded by IPP2 (At3g02780) and APX3 (At4g35000) were used as normalization controls. The sequences of the primers and probes used are as follows. For quantifying the BANYULS (BAN) transcript: primers, F1 5′-AGCTTATCTCTCTCGATGTCTTTCATC-3′ and R1 5′-CAGATAGCTTCTGCATTTCCTTGA-3′; probe, 5′-ACCGGTCACATGCATTTCTTTCCCG-3′. For quantifying the TT12 transcript: primers, F1 5′-TTGGGATATGCAGTTCATGCTT-3′ and R1 CCCGCTCCTAGCTCATTGC-3′; probe, 5′-TCTAAGTGCAGCAATCAGTGTGCGAGTG-3′. For quantifying the APX3 transcript: primers, F1 5′-GCCGTGAGCTCCGTTCTCT-3′ and R1 5′-TCGTGCCATGCCAATCG; probe, 5′-CGCGAACAAGAACTGTGCTCCTATCATG-3′. For quantifying the IPP2 transcript: primers, F1 5′-GTATGAGTTGCTTCTCCAGCAAAG-3′ and R1 5′-GAGGATGGCTGCAACAAGTGT-3′; probe, 5′-AAACACAAAGGTTACGTTCCCTCTAGTGTGGACT-3′.


AHA10 Gene Disruptions. As part of our effort to determine the in planta functions of all 11 Arabidopsis AHAs, we identified four independent T-DNA insertion plant lines harboring gene disruptions of AHA10. Two lines were obtained in the Ws ecotype from the Wisconsin collection (aha10-1 and aha10-2) (17) and two in the Col ecotype from the Salk collection (aha10-4 and aha10-5) (18) (Fig. 1b). Homozygous mutant plants were identified for all four mutations, indicating that the gene disruptions were not lethal. The chromosomal location of this aha10/tt mutation did not correspond to any well characterized tt mutant (23). PCR genotyping of segregating aha10-2 populations revealed normal Mendelian segregation of the aha10 mutation (χ2 = 0.644, P = 0.7247, n = 118). Extensive phenotypic analysis was restricted to the aha10-1 and aha10-2 alleles in the Ws background, although both Colbackground mutants showed a similar tt phenotype, as described below. The T-DNA insertion site in aha10-2 was sequenced and shown to be located in a coding sequence exon. The other three aha10 insertions were located in introns, as determined by border sequences.

Seeds from aha10 Plants Have a tt Phenotype. To assay for phenotypes in aha10 mutant lines, we conducted a variety of assays for plant growth and development under different growth conditions (see the supporting information, which is published on the PNAS web site). The only consistent difference observed was in seed coat color. Seeds from all four homozygous T-DNA insertion lines had pale seed coats (tt phenotype) under all growth conditions tested (Fig. 2 a and b). This phenotype was not observed in heterozygotes, indicating that all four gene disruptions represent recessive mutations.

Fig. 2.
AHA10 loss of function results in a tt phenotype with reduced PA and altered vacuoles. (a and b) Light microscope images of wild-type (a) and aha10 (b) seeds. (c and d) DMACA staining of Ws and aha10-1 seeds. DMACA changes color when it binds to 3′,4′ ...

The AHA10 Gene Complements. To confirm that the tt phenotype resulted from a loss of AHA10 function, we transformed aha10-1 plants with a genomic clone containing the AHA10 gene with 1.45 kb of upstream sequence and 1 kb of downstream sequence. Seeds from all 11 independent transgenic plant lines had a restored wild-type phenotype, as seen by a visual inspection of seed color as well as by staining for PA using DMACA (data not shown). Because our genomic clone did not contain any other predicted genes, complementation provides strong corroboration that AHA10 itself is required for normal pigmentation of Arabidopsis seed coats.

aha10 Mutants Have Normal Anthocyanin Levels but Reduced PA. The first tt mutants in Arabidopsis defined many of the key enzymes and proteins in flavonoid biosynthesis (11, 12, 24) (Fig. 3). Although the flavonoid pathway produces a wide variety of compounds, colored pigments in the Arabidopsis seed coat result primarily from the accumulation of anthocyanins and PA (8). The biosynthesis of both of these end-product pigments includes many shared upstream intermediates but diverges into two specific pathways after the production of cyanidin (Fig. 3).

Fig. 3.
Flavonoid pathway in Arabidopsis. Pathway diagram is derived from ref. 9. The known elements of the flavonoid pathway are indicated. The protein names are capitalized; the corresponding mutant line in Arabidopsis is in italics. The different chemical ...

To determine whether the PA and anthocyanin branches of the pathway were both affected in aha10 mutants, we quantified extractable levels of anthocyanin and PA in mature seeds. Although wild type and mutants showed comparable levels of anthocyanin, the mutants showed a >100-fold decrease in PA (Table 1). This significant decrease in PA (P < 0.05) was confirmed by staining seeds with DMACA (Fig. 2 c and d), which stains PA much more strongly than it stains other flavonoids. As expected, wild-type seeds stained black, indicating the presence of PA, whereas the aha10 seeds showed no detectable staining. Thus, aha10 seeds have a defect in flavonoid production that affects the PA pathway but not the anthocyanin pathway.

Table 1.
Quantitation of flavonoids in mature seeds shows mutants with decreased PA

A Disruption in PA Production Occurs Downstream of Epicatechin. To delineate the point at which PA production is disrupted in the aha10 mutants, we analyzed the flavonoid profiles of mature seed extracts by using HPLC. Of the seven extractable flavonoids that were quantified (Table 1), the end-product PA was the only flavonoid to consistently show a dramatic reduction (P < 0.05). In contrast, no significant changes were observed in anthocyanin (discussed above), kaempferol, or pelargonidin (P > 0.05). The colorless flavonol quercetin increased slightly (P < 0.05). These results provide strong evidence that the primary defect in aha10 mutants is specific to the PA biosynthetic branch.

In agreement with Abrahams et al. (9), we detected only epicatechin from our PA hydrolysis, indicating that Arabidopsis PA is composed exclusively of epicatechin monomers. This composition makes cyanidin the intermediate at the branch point that feeds into both the anthocyanin and PA pathways. Because anthocyanin was found at wild-type levels in the mutant, we hypothesized that the levels of the precursor cyanidin would not be reduced in the aha10 mutant. This expectation was verified, with cyanidin levels actually showing a slight increase (2.7-fold, P < 0.05).

In wild type, cyanidin is directly converted to epicatechin (the monomer unit of PA) in the first committed step of the PA pathway (25). This step is catalyzed by the enzyme BAN. To evaluate whether this PA-specific step was normal, we quantified soluble epicatechin directly from seed extracts. No free epicatechin was detectable in wild type, consistent with this intermediate occurring at very low levels. In contrast, significant levels of free epicatechin were detected in the mutants. Thus, the aha10 defect appears to cause a blockage resulting in the accumulation of epicatechin.

Vacuolar Biogenesis Is Altered in aha10 and tt12 Seed Coat Endothelial Cells. PA accumulates in the central vacuole of seed coat endothelial cells (5). To determine whether the loss of AHA10 affects the formation of these vacuoles, we performed microscopy with 5-(and -6)-carboxy 2′,7′-dichlorofluorescein diacetate (Fig. 2 e and f), a dye previously used to visualize plant vacuoles (9). When developing seeds (in which embryos are in the early mature stage) were examined, a large central vacuole was observed in the wild-type seed coat endothelial cells. In contrast, the equivalent staged cells in aha10 seeds showed multiple smaller vacuoles. The endothelial cells in Fig. 2 are representative of images most commonly observed for both wild type and mutant. Although small vacuoles were observed in 96% of the mutant endothelial cells (n = 77), only 11% of cells from wild-type controls showed comparable morphologies (n = 64). As the seeds matured, the differences between wild type and mutant diminished, indicating that the mutation resulted in a developmental delay, as opposed to a complete block in vacuolar biogenesis. We also observed this phenotype in the developing seeds of the tt12 mutant, where 94% of the cells observed had small vacuoles (n = 383).

aha10 Does Not Reduce the Accumulation of BAN and TT12 mRNA. To test for an aha10-related deficiency in general functions of seed coat endothelial cells, we conducted quantitative RT-PCR for two mRNAs (BAN and TT12). Both BAN and TT12 are specifically expressed in seed coat endothelial cells and function in the PA-specific branch of flavonoid biogenesis. BAN catalyzes the first committed step in PA synthesis (Fig. 3) (25), whereas TT12 encodes a MATE transporter that potentially functions to transport epicatechin into vacuoles (13). To evaluate the relative expression levels of these two endothelial cell-specific transcripts in mutants and wild type, we compared their expression values with two mRNA transcripts that are expressed in many cell types of the plant (IPP2 and APX3).

The RNA template used for RT-PCR was extracted from the uppermost 12 green siliques located closest to the apical meristem (Fig. 4). In each of three independent experiments, the relative expression levels of BAN and TT12 were slightly higher in the mutant than in the wild type. This finding shows that the expression of these two PA pathway genes is not the rate-limiting step in PA biosynthesis in an aha10 mutant. More generally, it provides further evidence that the aha10 disruption does not compromise essential cellular functions (e.g., the plasma membrane electrical potential and proton motive force) that would result in a global decrease in gene expression.

Fig. 4.
Expression levels are not decreased for two PA pathway-specific genes. Quantitative RT-PCR of BAN and TT12 transcripts in green siliques. The levels of BAN, TT12, and a control gene (IPP2) were compared with the APX3 gene for each RNA sample. In controls, ...


aha10 Disrupts the Biosynthesis of PA. Evidence presented here shows that four independent homozygous gene disruptions of a plasma membrane proton pump AHA10 (Fig. 1) all result in a tt phenotype (Fig. 2). These mutants appear to result from a loss of AHA10 function, as indicated by the ability to complement aha10-1 by using an AHA10 genomic clone. The underlying cause of this tt phenotype appears to be related to a 100-fold reduction in PA (Table 1). PA is produced in seed coat endothelial cells and accumulates in large central vacuoles. Because AHA10 is classified as a plasma membrane proton pump, the aha10/tt phenotype provides an unexpected connection between the function of a putative plasma membrane proton pump and the accumulation of a specific flavonoid in the central vacuole.

Although the exact composition of the colored pigments in the Arabidopsis seed coat is unknown, it has been attributed to anthocyanin and PA (8, 13). A complete loss of anthocyanidins and anthocyanins results in a yellow seed, as observed in a chalcone synthase-deficient mutant that produces no flavonoids (24) and the dihydroflavonol 4-reductase mutants that only produce flavonols (26). In contrast, the aha10 mutants show only a slight reduction in seed color. PA is the only major end-product flavonoid that is missing in aha10 mutants, suggesting that it is not PA, but anthocyanins or an unextracted pigment, that accounts for the major contribution to seed coat color in Arabidopsis.

The aha10 mutations do not completely eliminate any flavonoid biosynthetic branch in seeds, as evidenced by the detection of all three primary end products of the flavonoid biosynthetic pathway (PA, anthocyanin, and flavonols). Nevertheless, aha10 dramatically and specifically affects the PA-biosynthetic branch. This phenotype was revealed by the quantitation of seven different extractable seed flavonoids (Table 1). In this analysis, aha10 mutants showed either no significant change or increased levels of the end products anthocyanin and the flavonols kaempferol and quercetin, in contrast to a 100-fold reduction in the end product PA. Elevated concentrations were also observed for cyanidin, an intermediate that lies at the fork in the branched pathway leading to either PA or anthocyanin. This finding indicates that the defect in aha10 occurs at a step downstream of cyanidin biosynthesis.

In the downstream conversion of cyanidin to PA, epicatechin is the only known chemical intermediate. The biosynthetic steps leading to epicatechin are posited to occur in the cytosol, although the enzymatic machinery is membrane-attached (2729). Epicatechin is thought to be the intermediate that is transported into vacuoles (5). Once transported into a vacuole, condensation of epicatechin into PA may occur nonenzymatically in response to an acidic environment (30). In wild type, we did not observe any unpolymerized epicatechin, consistent with it being a low-abundance and transient intermediate. In contrast, we detected significant levels of unpolymerized epicatechin in mutants. This increase in epicatechin is consistent with a pathway “backup” resulting from a block in epicatechin transport into vacuoles or its conversion into PA.

aha10 and tt12 Disrupt Vacuolar Biogenesis. In different plant cells, there can be multiple vacuoles with distinct functions (31, 32). Some plant vacuoles may originate directly from the endoplasmic reticulum; others may require contributions from the Golgi or plasma membrane. Very little is known about the biogenesis of the specialized vacuoles in seed coat endothelial cells (15).

In aha10 mutants, we observed a predominance of small vacuoles, as opposed to one large central vacuole, in young developing seed coat endothelial cells (Fig. 2). This phenotype appears similar to those observed for gene disruptions of LDOX (leucoanthocyanidin dioxygenase) (9) and TT12-MATE (shown in this study). LDOX is an oxygenase that converts leucocyanidin to cyanidin. In the ldox mutant, the production of downstream flavonoids, such as epicatechin and PA, are blocked, and the endothelial cells develop with an abundance of small vacuoles. Although it is not clear how the ldox mutation prevents normal vacuolar biogenesis, two general mechanisms have been proposed. First, the ldox mutation may result in the buildup of a toxic upstream intermediate that somehow disrupts vacuolar biogenesis. Alternatively, it may block the production of a downstream intermediate that is required as a positive metabolic trigger for provacuolar maturation. Although the mechanism by which aha10 and tt12 disrupt vacuolar biogenesis is also unclear, all three mutations could cause a buildup in the same toxic intermediate or block the synthesis or accumulation of a positive metabolic trigger.

A Model for AHA10's Function in Flavonoid and Vacuolar Biogenesis. Although the expected function of a plasma membrane proton pump is to energize the plasma membrane and thereby enable signaling and solute transport at the cell surface, these established functions do not readily explain the specific vacuolar and PA biosynthesis defects observed for aha10 mutants. According to the current paradigm, a significant reduction in proton-pump activity at the plasma membrane would result either in cell death or a general down-regulation of all cellular functions. However, three lines of evidence indicate that these established functions were not significantly impacted by the loss of AHA10. First, the aha10 mutation was not lethal to seed coat endothelial cells, as evidenced by the production of brown-colored seeds. A lethal effect on this cell layer would be expected to produce a blond-colored seed (5). Second, mutant cells still produced normal or increased levels of non-PA pathway flavonoids, such as anthocyanins and flavonols (Table 1). This observation is most easily explained as a specific block in the PA biosynthesis pathway, rather than a general starvation for initial precursors, which would be expected from a general reduction in nutrient and solute uptake across the plasma membrane. Third, the levels of two seed coat endothelium-specific mRNAs (BAN and TT12) are expressed at normal to increased levels in the mutant (Fig. 4). Thus, the seed coat endothelial cells in aha10 mutants appear to have normal plasma membrane functions that supply the cells with adequate nutrients and enable the signaling systems required for their differentiation as pigment-producing cells.

To explain the aha10 phenotype, we offer a model that posits a unique and direct effect of AHA10 on an endomembrane system. In this model, AHA10 still functions as a proton pump but is recruited from the endoplasmic reticulum or plasma membrane to acidify endosomes or a specific provacuolar compartment. The AHA10-driven acidification may have three consequences. First, acidification may be required to establish a pH gradient and drive uptake (“cargo-loading”) of a flavonoid intermediate through a proton-coupled transporter (e.g., possibly epicatechin through the MATE-TT12 transporter). The uptake of PA precursors into vesicles is not only important for compartmentalizing of PA biosynthesis, but also may function to trigger vacuolar maturation. In membrane-trafficking studies, there are many examples in which the loaded “cargo” triggers vesicle maturation or fusion (33, 34). Second, acidification may be required to promote a nonenzymatic conversion of epicatechin to PA. Without this conversion, epicatechin concentrations may reach toxic levels that inhibit vesicle functions, including provacuolar maturation. Third, acidification by itself may be required for proper membrane trafficking, as seen in many endomembrane-trafficking pathways characterized in animal cells (35, 36). Thus, the proposed AHA10-driven acidification of vesicles provides a reasonable hypothesis to explain the aha10 mutant defects in vacuolar biogenesis and PA biosynthesis.

In mammalian systems, V-type ATPases (proton pumps) drive the acidification of endoplasmic reticulum-derived endomembrane compartments and endosomes originating from the plasma membrane (3537). For this reason, models for acidification of analogous endomembrane compartments in plants have assumed a similar role for plant V-type ATPase endosomes (38). However, unlike animals, plants also express two additional types of proton pumps, P-type H+-ATPases and PPases. Thus, it is possible that some combination of these multiple pumps are used throughout the various endomembrane systems in plant cells. Because P-type H+-ATPases represent the major source of proton pumping at the plant plasma membrane, it is reasonable to hypothesize that such pumps would provide a significant contribution to the proton-pumping capacity of PM-derived endosomal vesicles, some of which we speculate may participate in PA biosynthesis.

Our results do not rule out many alternative models, especially those in which AHA10 has an indirect or unique function not yet assigned to a known proton pump. The working model proposed above provides a logical explanation based on known activities of a proton pump. However, regardless of the mechanism, the aha10/tt phenotype is significant because it provides genetic evidence for an unexpected endomembrane function for a member of the plasma membrane proton-pump family. Although it is not clear whether AHA10 is the only member of this family to have an endomembrane function, our results support the hypothesis that some plant cells may use a P-type proton pump to help acidify endosomal or vacuolar compartments, in contrast to animal cells, which rely exclusively on V-type pumps.

Supplementary Material

Supporting Information:


We give special thanks to Kathy Truong, Greta Granstedt, and Steve McDowell for technical assistance. This work was supported by grants from the Human Frontier Science Program Organization (RG00268/2000-M to J.F.H.), Department of Energy (DE-FG03-94ER20152 to J.F.H.), National Science Foundation (0077378-DBI to J.F.H., 0132803 to A.S.M., and 0116266 to J.C.Y.), National Institutes of Health (1R01 GM070813-01 to J.F.H.), and Department of Agriculture National Research Initiative (02-35304-12290 to A.S.M.).


Author contributions: J.C.Y., W.A.P., A.S.M., and J.F.H. designed research; J.C.Y., G.A., N.F., N.B., T.C., W.A.P., S.P.H., and A.S.M. performed research; J.C.Y., W.A.P., S.P.H., A.S.M., and J.F.H. analyzed data; and J.F.H. wrote the paper.

This paper was submitted directly (Track II) to the PNAS office.

Abbreviations: AHA10, autoinhibited H+-ATPase isoform 10; PA, proanthocyanidin; MATE, multidrug and toxic compound extrusion; Ws, Wassilewskija; DMACA, dimethylaminocinnamaldehyde; tt, transparent testa; T-DNA, transfer DNA; LDOX, leucoanthocyanidin dioxygenase; BAN, banyuls.


1. Sze, H., Li, X. & Palmgren, M. G. (1999) Plant Cell 11, 677-690. [PMC free article] [PubMed]
2. Arango, M., Gévaudant, F., Oufattole, M. & Boutry, M. (2003) Planta 216, 355-365. [PubMed]
3. Baxter, I., Tchieu, J., Sussman, M. R., Boutry, M., Palmgren, M. G., Gribskov, M., Harper, J. F. & Axelsen, K. B. (2003) Plant Physiol. 132, 618-628. [PMC free article] [PubMed]
4. Harper, J. F., Manney, L. & Sussman, M. R. (1994) Mol. Gen. Genet. 244, 572-587. [PubMed]
5. Debeaujon, I., Nesi, N., Perez, P., Devic, M., Grandjean, O., Caboche, M. & Lepiniec, L. (2003) Plant Cell 15, 2514-2531. [PMC free article] [PubMed]
6. Rao, M. V., Paliyath, G. & Ormrod, D. P. (1996) Plant Physiol. 110, 125-136. [PMC free article] [PubMed]
7. Li, J., Ou-Lee, T. M., Raba, R., Amundson, R. G. & Last, R. L. (1993) Plant Cell 5, 171-179. [PMC free article] [PubMed]
8. Chapple, C. C., Shirley, B. W., Zook, M., Hammerschmidt, R. & Somerville, C. (1994) in Arabidopsis, eds. Meyerowitz, E. M. & Somerville, C. (Cold Spring Harbor Lab. Press, Plainview, NY), pp. 989-1030.
9. Abrahams, S., Lee, E., Walker, A. R., Tanner, G. J., Larkin, P. J. & Ashton, A. R. (2003) Plant J. 35, 624-636. [PubMed]
10. Abrahams, S., Tanner, G. J., Larkin, P. J. & Ashton, A. R. (2002) Plant Physiol. 130, 561-576. [PMC free article] [PubMed]
11. Debeaujon, I., Leon-Kloosterziel, K. M. & Koornneef, M. (2000) Plant Physiol. 122, 403-414. [PMC free article] [PubMed]
12. Koornneef, M., Luiten, W., de Vlaming, P. & Schram, A. (1982) Arabidopsis Inf. Serv. 19, 113-115.
13. Debeaujon, I., Peeters, A. J., Leon-Kloosterziel, K. M. & Koornneef, M. (2001) Plant Cell 13, 853-871. [PMC free article] [PubMed]
14. Goodman, C. D., Casati, P. & Walbot, V. (2004) Plant Cell 16, 1812-1826. [PMC free article] [PubMed]
15. Grotewold, E. (2004) Planta 219, 906-909. [PubMed]
16. Morsomme, P., Slayman, C. W. & Goffeau, A. (2000) Biochim. Biophys. Acta 1469, 133-157. [PubMed]
17. Young, J. C., Krysan, P. J. & Sussman, M. R. (2001) Plant Physiol. 125, 513-518. [PMC free article] [PubMed]
18. Alonso, J. M., Stepanova, A. N., Leisse, T. J., Kim, C. J., Chen, H., Shinn, P., Stevenson, D. K., Zimmerman, J., Barajas, P., Cheuk, R., et al. (2003) Science 301, 653-657. [PubMed]
19. Becker, D., Kemper, E., Schell, J. & Masterson, R. (1992) Plant Mol. Biol. 20, 1195-1197. [PubMed]
20. Clough, S. J. & Bent, A. F. (1998) Plant J. 16, 735-743. [PubMed]
21. Bailey, R., Nursten, H. & McDowell, I. (1993) J. Sci. Food Agric. 63, 455-464.
22. Adamse, P., Peters, J., Jaspers, P., van Tuinen, A., Koornneef, M. & Kendrick, R. (1989) Photochem. Photobiol. 50, 107-111.
23. Sagasser, M., Lu, G. H., Hahlbrock, K. & Weisshaar, B. (2002) Genes Dev. 16, 138-149. [PMC free article] [PubMed]
24. Shirley, B. W., Kubasek, W. L., Storz, G., Bruggemann, E., Koornneef, M., Ausubel, F. M. & Goodman, H. M. (1995) Plant J. 8, 659-671. [PubMed]
25. Xie, D. Y., Sharma, S. B., Paiva, N. L., Ferreira, D. & Dixon, R. A. (2003) Science 299, 396-399. [PubMed]
26. Shirley, B. W., Hanley, S. & Goodman, H. M. (1992) Plant Cell 4, 333-347. [PMC free article] [PubMed]
27. Xie, D. Y., Sharma, S. B. & Dixon, R. A. (2004) Arch. Biochem. Biophys. 422, 91-102. [PubMed]
28. Saslowsky, D. & Winkel-Shirley, B. (2001) Plant J. 27, 37-48. [PubMed]
29. Burbulis, I. E. & Winkel-Shirley, B. (1999) Proc. Natl. Acad. Sci. USA 96, 12929-12934. [PMC free article] [PubMed]
30. Marles, M. A., Ray, H. & Gruber, M. Y. (2003) Phytochemistry 64, 367-383. [PubMed]
31. Paris, N., Stanley, C. M., Jones, R. L. & Rogers, J. C. (1996) Cell 85, 563-572. [PubMed]
32. Surpin, M. & Raikhel, N. (2004) Nat. Rev. Mol. Cell. Biol. 5, 100-109. [PubMed]
33. Aridor, M. & Traub, L. M. (2002) Traffic 3, 537-546. [PubMed]
34. Petäjä-Repo, U., Hogue, M., Bhalla, S., Laperrière, A., Morello, J. P. & Bouvier, M. (2002) EMBO J. 21, 1628-1637. [PMC free article] [PubMed]
35. Gu, F. & Gruenberg, J. (2000) J. Biol. Chem. 275, 8154-8160. [PubMed]
36. Zeuzem, S., Zimmermann, P. & Schulz, I. (1992) J. Membr. Biol. 125, 231-241. [PubMed]
37. Nishi, T. & Forgac, M. (2002) Nat. Rev. Mol. Cell. Biol. 3, 94-103. [PubMed]
38. Nebenfuhr, A., Ritzenthaler, C. & Robinson, D. G. (2002) Plant Physiol. 130, 1102-1108. [PMC free article] [PubMed]
39. Robertson, W. R., Clark, K., Young, J. C. & Sussman, M. R. (2004) Genetics 168, 1677-1687. [PMC free article] [PubMed]

Articles from Proceedings of the National Academy of Sciences of the United States of America are provided here courtesy of National Academy of Sciences
PubReader format: click here to try


Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...


  • Compound
    PubChem Compound links
  • Gene
    Gene links
  • GEO Profiles
    GEO Profiles
    Related GEO records
  • MedGen
    Related information in MedGen
  • Pathways + GO
    Pathways + GO
    Pathways, annotations and biological systems (BioSystems) that cite the current article.
  • PubMed
    PubMed citations for these articles
  • Substance
    PubChem Substance links
  • Taxonomy
    Related taxonomy entry
  • Taxonomy Tree
    Taxonomy Tree

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...