Introduction

Pollen grains present a highly specialized tissue of angiosperm plants. They can be regarded as two- or three-cellular organisms consisting of the vegetative cell which surrounds a generative cell or two sperm cells. Within the anther tissue these male gametophytes develop from a diploid pollen mother cell (microsporocyte) after meiotic divisions into haploid microspores that further divide by mitosis into the vegetative and the generative cell. In some species the generative cell further divides into two sperm cells in the pollen grain whereas in other species the generation of sperm cells occurs later. In a last step the anther tissue dehydrates and exposes the pollen grains to the environment where they wait for transport to the stigma (Bedinger 1992; Cresti et al. 1992; Tanaka 1993).

The primary task of the pollen grain is to deliver the sperm cells to the egg apparatus that in many cases is not on the same flower, even not on the same plant and sometimes many kilometres away. To reach the "females" the pollen grains are either carried by insects, birds or mammals or are wind-blown over large distances. During their transportation the pollen has to resist harsh environmental conditions and severe damage by their deliverers like mechanical stress, changes in temperatures and humidity. Therefore, they are covered by an additional cell wall, the exine consisting of sporopollenin which resists mechanical and chemical treatments. Furthermore, they are in a dehydrated, quiescent state with water contents sometimes less than 10 % of the fresh weight. This desiccation helps to maintain the pollen viability during transport.

To germinate and to grow a pollen tube as fast as possible after landing on the stigma, the pollen grains generate all mRNAs and proteins that they will need for a fast start during the last phase of pollen development (Mascarenhas 1990; Schrauwen et al. 1990). This was confirmed recently by an increasing number of transcriptomic (e.g., Becker et al. 2003; Bock et al. 2006; Wei et al. 2010) and proteomic studies on developing, mature and tube-growing pollen grains (e.g., Dai et al. 2006; Grobei et al. 2009; Pertl et al. 2009).

A large number of cellular processes are involved in polar growth of the pollen tube and have been studied intensively during the last decades. Initiation of pollen tube germination and the subsequent growth of pollen tubes can be viewed as a temporally and spatially fine-tuned network of interacting cellular processes and pathways like the cytoskeleton dynamics, vesicle trafficking and fusion, Ca2+, pH and phosphoinositide signalling, G-protein signalling and many more (Holdaway-Clarke and Hepler 2003; Cheung and Wu 2008; Cole and Fowler 2006; Michard et al. 2009; Steinhorst and Kudla 2013). It has been postulated that ion fluxes or transport across the plasma membrane may synchronise these individual processes leading to a concerted action for steady tube growth (Feijó et al. 1995, 2001). One of the best-characterized plasma membrane transport proteins in plants is the plasma membrane H+ ATPase (PM H+ ATPase, EC 3.6.3.6) which belongs to the family of P-type ATPases. This article reviews the current knowledge of the physiological role of the PM H+ ATPase in the pollen's life, emphasizes its central role, and discusses some open questions and knowledge gaps that need to be filled for a comprehensive understanding of the H+ pump's role during pollen germination and tube growth.

Activity and function of the plasma membrane H+ ATPase

The plasma membrane H+ ATPase consists of approximately 950 amino acids resulting in a molecular weight of ca. 105 kDa and a pI around 6.5. The protein contains ten transmembrane domains. Its N and C terminus face the cytoplasmic side and together with the small and the large loop more than 70 % of the protein resides in the cytosol and only a small part is exposed to the extracellular side (Fig. 1a). The major role of the PM H+ ATPase is the generation of an electrochemical gradient of H+ across the plasma membrane of plant cells by coupling the energy from ATP hydrolysis to the transport of H+ and thus energizing the plasma membrane for uptake of ions and nutrients. Furthermore, due to the export of H+ from the cytosol to the extracellular space, the PM H+ ATPase is involved in the regulation of cytoplasmic as well as cell wall pH. General and basic principles on the activity and function of the PM H+ ATPase can be found in other review articles in more detail (Serrano 1989; Briskin 1990; Palmgren 1998, 2001; Morsomme and Boutry 2000; Gaxiola et al. 2007). It has to be noted that two other H+ pumps, the V-type H+ ATPase and the H+ pyrophosphatase, which are mainly localized at the tonoplast, also contribute to the cytosolic pH homeostasis and thus may affect pollen tube growth, too.

Fig. 1
figure 1

Model of the lily pollen PM H+ ATPase LilHA1. a Membrane topology of LilHA1 (AY029190.2). Transmembrane domains were predicted by HMMTOP (http://www.sacs.ucsf.edu/cgi-bin/hmmtop.py) and the model was drawn with TOPO2 (http://www.sacs.ucsf.edu/TOPO2/). b Three-dimensional model of LiHA1 created by homology modelling (SWISS-Model, http://swissmodel.expasy.org/workspace) using the crystal structure of AHA2 (3b8cA; Pedersen et al. 2007). The complete CTAD of LilHA1 was modelled by I-TASSER (http://zhanglab.ccmb.med.umich.edu/I-TASSER). The last part of the CTAD starting as indicated by red bar in a is coloured bright red. Colours in b correspond with marked regions in a. A actuator domain, N nucleotide binding domain, P phosphorylation domain, R regulatory domain (=CTAD)

Full size image

The activity of the PM H+ ATPase was mainly characterized using plasma membrane vesicles that were isolated by sucrose density gradient centrifugation or aqueous two-phase partitioning (Larsson et al. 1987; Briskin and Hanson 1992; Palmgren 1998). These methods result in a stochastic mixture of inside (cytosolic side)-out and outside-out vesicles. Upon addition of ATP the H+ transport, changes in membrane potential (VM) and ATP hydrolysis can be monitored on inside-out vesicles to characterize the PM H+ ATPase activity (Briskin et al. 1987; Palmgren 1990; Briskin and Reynolds-Niesman 1991). Further purification of the pump protein needs careful solubilisation by detergents to preserve the structure and activity of the protein in reconstituted vesicles. It has to be noted that despite the large number of studies on the biochemical characterization of the PM H+ ATPase activity no specific inhibitor like, for instance ouabain for the Na+,K+ ATPase, has been found. Other H+ ATPases, such as the V-ATPases and the F-ATPases, can be specifically inhibited by bafilomycin A and oligomycin, respectively. Therefore, most studies use ortho-vanadate that blocks the formation of a phosphorylated intermediate in all P-type ATPases as well as in a number of phosphate-metabolizing enzymes, to inhibit the PM H+ ATPase activity. A less well-characterized component, erythrosine B, also inhibits the PM H+ ATPase activity.

Furthermore, the PM H+ ATPase can also be investigated in heterologous expression systems. The choice of the expression system is important because a total reduction of endogenous PM H+ ATPases may be lethal for the organism chosen to express this protein. The PM H+ ATPase is a so-called housekeeping enzyme and hence essential for the survival of the cells. Therefore, the endogenous yeast PMA1 gene was placed under control of a galactose-inducible promoter whereas in the presence of glucose the plant proton pump is expressed (Cid et al. 1987; Villalba et al. 1992; Palmgren and Christensen 1993; Regenberg et al. 1995). Alternatively, the expression of both yeast PM H+ ATPases, PMA1 and PMA2, was inactivated by disrupting their genes. Yeast cells could survive in galactose medium with a special plasmid in which PMA1 and URA3 were placed under the control of the GAL1 promoter whereas the introduced plant proton pump is controlled by the PMA1 promoter. Shifting from galactose to glucose medium induced an exclusive expression of the plant PM H+ ATPase while the PMA1-carrying plasmid was removed simultaneously by selection on URA3 suicidal medium (Morsomme and Boutry 2000; de Kerchove d'Exaerde et al. 1995). These strategies of recombinant expression of the plant-derived H+ ATPase also allow the addition of tags to the recombinant protein which simplifies the identification and the purification for further studies on reconstituted vesicle systems.

In general, a specific enzyme activity of the PM H+ ATPase of 1–2 μmol Pi min−1 mg−1 was measured in purified plasma membrane vesicles with a Km for MgATP ranging between 0.3 and 1.5 mM (Palmgren 1998; Morsomme and Boutry 2000). The pH optimum of the ATPase activity is slightly acidic ranging around pH 6.5–6.8 and the ATP hydrolysis can be stimulated by addition of potassium ions (10–50 mM; Buch-Pedersen et al. 2006). Usually, 1 H+ is transported per hydrolysed ATP but may vary due to specific regulations of the pump activity thus generating a pH gradient of 1.5 to 2 pH units and hyperpolarizing the plasma membrane even more negative than −200 mV (Serrano 1989; Morsomme et al. 1996; Palmgren and Harper 1999; Gaxiola et al. 2007).

Structure and regulation of the plasma membrane H+ ATPase

A C-terminally tagged PM H+ ATPase (AHA2) was expressed in yeast to allow its purification and crystallization. The crystal structure shows three large cytosolic domains N (nucleotide binding), P (phosphorylation domain) and A (actuator domain) in addition to a transmembrane domain with ten helices (Pedersen et al. 2007). Figure 1 shows a homology model of the lily pollen H+ ATPase isoform LilHA1 (access. no. AY029190.2) using the crystal structure of AHA2 as a template. To couple the transmembrane transport of H+ to ATP hydrolysis, the MgATP is bound to the N domain which then moves towards the P domain where a conserved asparagine residue (Asp 329 in AHA2) becomes phosphorylated. Dephosphorylation of the Asp residue involves a large conformation change of domain A simultaneously with conformational changes in the membrane region. The energy released from the hydrolysis of ATP is temporarily stored in a conformational change of the protein exhibiting its H+ binding side either to the cytosol (unphosphorylated) or to the extracellular space (phosphorylated) and thus transports H+ from one side to the other side of the membrane (Buch-Pedersen et al. 2009).

A moderate regulation of the overall activity of PM H+ ATPases in plant cells may occur at the transcriptional level in response to environmental changes (e.g., temperature, salts, sugar; see Lee et al. 2004; Camoni et al. 2006; Janicka-Russak and Klobus 2007). Alternatively, a regulated incorporation of exocytotic vesicles already containing the PM H+ ATPase protein increases the amount of ATPase in the plasma membrane and therefore increases H+ extrusion as has been shown after auxin addition (Hager et al. 1991). However, the most prominent short-term regulation in activity is at the post-translational level due to phosphorylation of the C-terminal autoinhibitory domain (CTAD) that results in the binding of 14-3-3 proteins. In this case the penultimate threonine (Thr) residue at the CTAD is phosphorylated which in turn enables the binding of a 14-3-3 dimer and thus, displacing the C terminus to allow a higher ATP turnover and H+ transport (Fuglsang et al. 1999; Svennelid et al. 1999; Camoni et al. 2000; Maudoux et al. 2000). A crystallographic model showed details of the interaction between the C terminus of the PM H+ ATPase and 14-3-3 proteins and explains how the fungal toxin fusicoccin irreversibly activates the ATPase activity by strengthening the interaction between 14-3-3 s and the PM H+ ATPase C terminus (Korthout and DeBoer 1994; Oecking et al. 1997; Baunsgaard et al. 1998; Fullone et al. 1998). A predicted structure of the CTAD which could not be crystallized, so far, was superimposed onto the LILHA1 model (Fig. 1b). This simple model implies that the C-terminal part (bright red) of the LilHA1-CTAD might reach into the gap between the A, N and P domain thus slowing down the conformational changes necessary for H+ transport. Note that this hypothesis is not confirmed yet by crystallization experiments, and other possible arrangements of the CTAD have been proposed (Axelsen et al. 1999; Palmgren 2001; Pedersen et al. 2007).

Using single particle analysis of electron microscopy images, a wheel-like structure formed by a H+ ATPase hexamer held together by six 14-3-3 proteins was identified that might be the functional unit of the PM H+ ATPase (Ottmann et al. 2007). It has to be noted that other regulatory mechanisms independent from the C terminus or phosphorylation or 14-3-3 protein binding have been reported but still need more investigations to show how these regulatory mechanisms are involved in the physiological function of the PM H+ ATPase (Borch et al. 2002; Morandini et al. 2002; Giacometti et al. 2004; Ekberg et al. 2010; Piette et al. 2011). In addition, phosphorylation of amino acid residues other than the penultimate Thr might inhibit the PM H+ ATPase activity (Lino et al. 1998; Fuglsang et al. 2007; Yang et al. 2010).

If the major regulation of the PM H+ ATPase activity occurs via phosphorylation the question raises which protein kinases and phosphatases are responsible for the phosphorylation and dephosphorylation, respectively? So far, this question is still under investigation and the outcome might result in a very variable picture: each stimulus may activate different signal transduction pathways that finally phosphorylate the PM H+ ATPase at different positions via diverse kinases resulting in a specific response to the respective signal. For instance, in Vicia faba guard cells the blue light activation of the PM H+ ATPase involves phosphorylation by a K252a-independent protein kinase and dephosphorylation by a Mg2+-dependent type 2C phosphatase both co-localizing with the PM H+ ATPase (Hayashi et al. 2010) whereas in Arabidopsis seedlings the phosphorylation by PKS5 inhibits the ATPase activity (Fuglsang et al. 2007). Even in the same organism (Nicotiana sp.), two different ATPase isoforms (PMA2 and PMA4) showed variability in their phosphorylation state (Bobik et al. 2010).

In addition to the described regulation of the PM H+ ATPase by phosphorylation and interaction with 14-3-3 s, a novel interacting protein, PPI1 (proton pump interactor 1) has been discovered (Morandini et al. 2002; Muniz Garcia et al. 2011). PPI1 also stimulates the ATPase activity by binding to the CTAD but to a different domain than 14-3-3 s (Viotti et al. 2005). So far, the role of PPIs in pollen was not investigated but might become important because high expression levels of the PPI1-mRNA (AT4G27500) can be observed when analyzing micro array data (www.genvestigator.com).

Activity, localization and regulation of the PM H+ ATPase in pollen

Pollen grains of many species can be easily cultured in synthetic culture media that contain sugars as an osmoticum, millimolar amounts of K+, Ca2+, boric acid and an acid pH. For instance, several germination media were reported for Lilium longiflorum, Nicotiana tabacum or Arabidopsis thaliana (Tupy et al. 1977; Dickinson 1978; Daher et al. 2009; Pertl-Obermeyer et al. 2013). During in vitro cultivation, an acidification of the germination medium could be observed due to H+ extrusion from the pollen grains (Southworth 1983; Rodriguez-Rosales et al. 1989; Tupy and Rihova 1984). Using in vivo (Rodriguez-Rosales et al. 1989; Certal et al. 2008; Pertl et al. 2010) and biochemical studies on pollen membrane fractions (Obermeyer et al. 1996; Pertl et al. 2001, 2005) it was demonstrated that the PM H+ ATPase is responsible for the medium acidification. Additionally, all antagonists of the PM H+ ATPase also inhibited pollen germination and tube growth, whereas stimulators of the PM H+ ATPase activity were also able to boost the germination frequency as well as the tube growth rates as summarized in Fig. 2 (see also Fricker et al. 1997; Certal et al. 2008). The inhibitors vanadate and erythrosine B can totally block lily pollen germination. A stimulation of germination by 20 % was achieved with fusicoccin, which promotes the binding of 14-3-3 proteins to the PM H+ ATPase (Pertl et al. 2001), whereas AICAR (5-aminoimidazole-4-carboxamide ribonucleoside monophosphate) which blocks the binding of 14-3-3 proteins to the PM H+ ATPase (Paul et al. 2005), prevents germination of pollen grains. Additionally, less well-known environmental parameters, such as the concentration of boric acid in the germination medium or even weak electrical AC fields, can increase the germination frequency of lily pollen and it was demonstrated that both parameters act on the PM H+ ATPase activity (Obermeyer et al. 1996; Plätzer et al. 1997). All studies showed a direct correlation between the PM H+ ATPase activity and the pollen germination as well as tube growth and thus demonstrate an active PM H+ ATPase as a prerequisite for successful pollen germination and tube growth.

Fig. 2
figure 2

Effect of inhibitors and promoters of the PM H+ ATPase activity on lily pollen germination. Pollen grains of Lilium longiflorum were cultivated in germination medium (10 % w/v sucrose, 1 mM KCl, 0.1 mM CaCl2, 1.6 mM H3BO3, pH 5.6) in the presence of the respective chemicals. For control experiments (0 μM), the appropriate solvent was added at the highest concentration (e.g., ethanol for fusicoccin). AICAR 5-aminoimidazole-4-carboxamide ribonucleoside monophosphate

Full size image

Furthermore, since the first studies on extracellular currents in lily pollen by Manfred Weisenseel and Lionel Jaffe it is well-known that pollen grains and especially tubes drive steady electric currents in the range of hundred picoamperes through themselves (Weisenseel et al. 1975). More recent investigations (reviewed by Feijó et al. 2001; Holdaway-Clarke and Hepler 2003; Michard et al. 2008) gave a detailed image of these currents: calcium ions enter the tip of a growing pollen tube generating a tip-localized gradient of the free cytosolic Ca2+ concentration ([Ca2+]cyt) that acts like a built-in navigation system (Obermeyer and Weisenseel 1991; Rathore et al. 1991; Pierson et al. 1994). Potassium ions (Messerli et al. 1999) and anions (Cl–) are flowing into the tube and in addition, chloride ions show a large efflux at the tube tip (Zonia et al. 2002). The currents carried by H+ can be distinguished as a large efflux at the pollen grain, small influxes along the tube and a large influx at the very tip followed by a remarkable efflux ca. 20 μm behind the tip (Feijó et al. 1999; Certal et al. 2008). This spatial difference in H+ fluxes along the tube's plasma membrane creates variations in the cytosolic pH. It is now generally accepted that growing pollen tubes show a slightly acidic pH in the tip cytosol and an alkaline band approximately at the beginning of the clear cap (Turian 1981; Feijó et al. 1999; Michard et al. 2008), although two studies were unable to monitor the pH gradients probably due to the high buffer capacity of the applied intracellular pH indicator (Parton et al. 1997; Fricker et al. 1997). Recently, it was shown that the H+ effluxes at the pollen grain as well as in the region of the alkaline band are mainly caused by a vanadate-inhibited PM H+ ATPase (Certal et al. 2008 and supplements therein) although the pH of the alkaline band is slightly different from the pH optimum of the PM H+ ATPase. To solve this mismatch in pH dependence, a detailed biochemical characterization of the transport properties of pollen-specific PM H+ ATPase isoforms is necessary.

Generally, the observed H+ efflux pattern correlates quite well with the localization of the PM H+ ATPase. Immunogold localization in lily pollen and localization of a NtAHA::GFP fusion protein in tobacco pollen showed an accumulation of the H+ ATPase in the plasma membrane of pollen grains where it is responsible for the H+ efflux (Obermeyer et al. 1992; Certal et al. 2008). No fluorescence-labelled PM H+ ATPases were detected in the PM of the tip region of the clear cap thus allowing a net influx of H+. Additionally, immunogold staining of the PM H+ ATPase was also observed in the ER and the Golgi apparatus thus demonstrating the ATPases' way to the plasma membrane through the secretory pathway. The distribution of the PM H+ ATPase in pollen ER and Golgi membranes was recently confirmed by a proteomic study used to identify membrane proteins from organelles of lily pollen (Pertl et al. 2009). The GFP::ATPase fusion protein could also be detected in the V-shaped cone of vesicles in the tip region of growing tubes (Certal et al. 2008) whereas no tip-localized vesicles were stained during immunogold labelling (Obermeyer et al. 1992). It still needs to be clarified where the PM H+ ATPase is incorporated into the plasma membrane. Is it delivered by the tip-targeted secretory vesicles to the tip zone or by another secretory pathway?

Regulation of pollen PM H+ ATPase is suggested to follow the same general scheme as in other tissue. Several isoforms of regulatory 14-3-3 proteins have been identified in lily pollen (Pertl et al. 2011) and a fusicoccin-induced accumulation of 14-3-3 s with the membrane fraction was observed (Pertl et al. 2001). Although the direct interaction between these two pollen proteins still needs to be investigated in detail, a regulatory role for 14-3-3 proteins in osmoregulation was shown in lily pollen (Pertl et al. 2010). A hyper-osmotic treatment causes a hyperpolarization of the plasma membrane and an increased acidification of the germination medium due to an increased activity of the PM H+ ATPase which, in turn, was caused by recruitment of more 14-3-3 proteins to the plasma membrane probably binding to the PM H+ ATPase. So far, all PM H+ ATPases known to be expressed in pollen contain all sequence motifs necessary for phosphorylation and subsequent binding of 14-3-3 proteins.

Expression of PM H+ ATPase in pollen

Because of its important role in energizing the plasma membrane, the H+ ATPase is probably expressed in every single cell of a plant with the exception of the generative cells of pollen grains (Gehwolf et al. 2002) where an active PM H+ ATPase would acidify the cytosol of the vegetative cell. The PM H+ ATPase belongs to the P3 sub-family of the P-type ATPase family (Pedersen et al. 2012) and consists of 12 isoforms in Arabidopsis thaliana, nine isoforms in Nicotiana plumbaginifolia, ten isoforms in rice, and ten isoforms in maize with still increasing information on isoforms as the number of sequenced organisms will increase in the future. A phylogenetic tree has been generated using the amino acid sequences of isoforms from Arabidopsis, Medicago, tobacco, tomato, potato, wheat, barley, rice, maize and lily (Fig. 3). It can be noticed that the PM H+ ATPase isoforms cluster in five subfamilies as it has been described earlier (Arango et al. 2003). However, a closer look reveals that isoforms mainly expressed in pollen can be found in subfamily IV. The PM H+ ATPase isoforms LilHA1 and LilHA2 have been identified from lily pollen (Gehwolf et al. 2002; Pertl et al. 2009), and NtHA1 was isolated from tobacco pollen (Certal et al. 2008). Furthermore, expression of isoform NpPMA5 (Nicotiana plumbaginifolia) was observed in pollen, too (Lefebvre et al. 2005). Analyzing the expression of the Arabidopsis thaliana isoforms AHA1 to AHA10 at the transcription level using micro array data stored at Genevestigator (www.genevestigator.com) also shows a fascinating picture on the tissue-specific expression of H+ ATPase isoforms (Fig. 4). In the majority of tissues AHA1 and AHA2 are the dominating isoforms, whereas in pollen their transcription level decreases dramatically. In pollen the isoform AHA8 shows by far the highest transcription level followed by AHA7, AHA6 and AHA9. This extreme difference between the transcription levels in pollen compared to sporophytic tissues might correspond to specific physiological functions of the PM H+ ATPase in pollen. This means that AHA8 is regulated by a special promoter tailored to the specific expression in pollen and that the AHA8 protein is specifically adapted to the life conditions of pollen grains and their special tasks. Especially the different hydration states of pollen grains may confront membrane proteins with a problem: it has been postulated that membrane lipids reorganize from the hexagonal (HII) phase to the lamellar (bilayer) phase (Simon 1974; Heslop-Harrison 1979) during the uptake of water, thus forcing membrane proteins to refold into the regenerating bilayer membrane again without losing their activity. One may therefore assume that specific isoforms of membrane proteins are expressed in pollen when compared with other water-containing tissues that face less dramatic changes in the hydration state. We favour this idea as all of the known pollen-specific PM H+ ATPase isoforms cluster in subfamily IV meaning that they have common features in their amino acid sequence that are lacking in other ATPase isoforms. Therefore, in addition to slightly variations in their protein structure, one might expect differences in the biochemical properties of the enzyme suiting its activity to the specific needs of pollen grains. These specific properties may help it to stay intact during the phase transitions of the lipid bilayer during de- and rehydration of the mature pollen grain or to allow a H+ export in the region of the alkaline band of the pollen tube that is actually out of the pH optima known from biochemically characterized isoforms like AHA2, NpPMA2 or NpPMA4. However, this hypothesis has to be proven by future experiments. Additionally, we hypothesize that all PM H+ ATPase isoforms from other organisms, which cluster in subfamily IV, are also mainly expressed in pollen.

Fig. 3
figure 3

Phylogenetic tree of PM H+ ATPase isoforms from several species. CLUSTAL X2.1 was used to align the PM H+ ATPase amino acid sequences and to construct the phylogenetic tree. The GeneBank accession numbers of the analyzed PM H+ ATPases are as follows: Arabidopsis thaliana: AHA1, NP_179486; AHA2,NP_194748; AHA3, NP_200545; AHA4, NP_190378; AHA5, NP_180028; AHA6, NP_178762; AHA7, NP_191592; AHA8, NP_189850; AHA9 NP_178181; AHA10, NP_173169; AHA11, NP_201073; Nicotiana tabacum: NtHA1, AY383599; Nicotiana plumbaginifolia: NpPMA1, Q08435; NpPMA2, Q42932; NpPMA3, Q08436; NpPMA4, Q03194; NpPMA5, AAV49160.1 (truncated); NpPMA6, AAD46186; NpPMA8, AAD46187; NpPMA9, AAD46188; Lilium longiflorum: LilHA1, AY029190.2; LilHA2, EF397610; Oryza sativa: OsHA1, D10207; OsHA2, D31843; OsHA3, AJ440001; OsHA4, AJ440002; OsHA5, AJ440216; OsHA6, AJ440217; OsHA7, AJ440218; OsHA8, AJ440219; OsHA9, AJ440220; OsHA10, AJ440221; Zea mays: ZmHA1, NP_001105360; ZmHA2, NP_001105470; Triticum aestivum: TaHA1, AAV71150; Solanum lycopersicum: SlHA1, NP_001234775; SlHA2, AAD55399; SlHA4, AAB17186; Solanum tuberosum: PHA1, CAA54046; PHA2, CAA54045; Hordeum vulgare: HvHA1, AY136627; Medicago truncatula: MtrHA1, CAB85494; MtrHA2, XM_003603484; MtrHA3, XM_003610033; MtrHA4, XM_003610032; MtrHA5, XM_003594906; MtrHA6, CAB85495. The annotation of the subfamilies follows Arango et al. (2003). Subfamily IV that was highlighted as pollen-specific expression is proposed for these isoforms

Full size image
Fig. 4
figure 4

Expression levels of Arabidopsis PM H+ ATPase isoforms. a Transcript levels of AHA1 to AHA10 of different Arabidopsis tissues were calculated from micro-array experiments using Genevestigator (www.genevestigator.com) and presented as log2 values. b Linear expression values of all 11 PM H+ ATPase isoforms in Arabidopsis pollen. Mean value ± SD

Full size image

Integration of signal transduction and metabolic pathways by the PM H+ ATPase

According to our current knowledge on biological systems, pollen germination and tube growth may be viewed as a complex but robust network in which cellular processes, signal transduction and metabolic pathways interact with each other thus forming a network that is well regulated in space and time. Whatever parameter is experimentally changed, either an entire cellular process (e.g., vesicle transport [Geitmann et al. 1996], actin polymerisation [Cai et al. 1997; see also Cheung and Wu 2008 for a recent review]) or a single enzyme activity (protein phosphatase 2A [Obermeyer et al. 1998] or Rho-GTPases [Kost 2008]) or even a small ion (tip-localized Ca2+ concentration; Malhó and Trewavas 1996), the system responds more or less in the same way: complete or temporary arrest of tube growth depending on the strength of the stimulus. A network may be seen as a large number of nodes linked by edges. A node with a large number of links to other nodes may become a central process in the network. As presented above the PM H+ ATPase may be such an important node. Modulation of the ATPase activity by independent and different parameters, leads immediately to a change in the germination frequency and in tube growth speed. By transition from a low activity state to a high activity state the pollen tube grows faster or slower if changing to a low activity state (summarized in Fig. 5). However, different independent signals and/or signal transduction pathways may have opposite effects on the PM H+ ATPase activity and thus the PM H+ ATPase has to be able to integrate signals from different sources. Priority signals might exist which can overrule other parameters and due to the length of the tube one might also assume that the PM H+ ATPase might be differentially regulated at different parts of the tube. For instance, a local increase in external pH (e.g., from pH 5.6 of the usual germination medium to pH 7.0) reduces the proton motif force at the plasma membrane locally and may result in a local activation of the H+ pump. In general, signal transduction pathways that modulate the phosphorylation state of proteins also change the PM H+ ATPase activity. So far, several phosphorylation sites have been identified in PM H+ ATPases (Rudashevskaya et al. 2012) and protein kinases and phosphatases that are involved in the phosphorylation of the H+ pump (Lino et al. 1998; Camoni et al. 1998; Fuglsang et al. 2006, 2007; Hayashi et al. 2010) were identified, but none could be found that specifically phosphorylates the penultimate threonine residue responsible for 14-3-3 binding. It is probable that metabolic pathways that affect the cytosolic ATP/ADP ratio (Obermeyer et al. 2013) may temporally lead to a change in the phosphorylation state of the PM H+ ATPase, and apart from that, low ATP concentrations will cause a decrease in H+ transport. Different external signals (hormones, light, salt stress) are well known to influence the H+ pump activity. Recently, changes in the water potential influenced the pollen PM H+ ATPase activity (Pertl et al. 2010) resulting in a 14-3-3-dependent (phosphorylation?) activation of the H+ pump and its contribution to osmosensing and osmoregulation. In this context, it has to be emphasized that regulation of the PM H+ ATPase may involve interaction with other plasma membrane proteins, e.g., receptor kinases, as has been shown for the brassinosteroid receptor BRI1 that interacted with AHA1 (Caesar et al. 2011). This research area on ATPase interacting proteins is just emerging, and no data have been obtained so far for pollen PM H+ ATPases and their regulation.

Fig. 5
figure 5

The PM H+ ATPase integrates several signal transduction and metabolic pathways in pollen germination and tube growth to initiate downstream responses (see text for details)

Full size image

However, it is important for the pollen grain and particularly, for the tube to respond to changes in environmental parameters as fast as possible to be able to maintain its optimal growth speed for reaching the ovules in time. Only pollen tubes which negotiate all obstacles on their way to the egg cells can contribute to fertilization and influence the next generation of plants. Based on the studies on the pollen PM H+ ATPase reviewed here, one may conclude that by modulating the ATPase activity, the pollen grain and tube is able to respond fast to small environmental perturbation on its way to the ovule. However, vast disturbances will definitely inhibit tube growth, but most small changes that may happen in the style tissue (e.g., pH differences, osmolarity, oxygen supply) may be compensated by speeding up or slowing down the tube growth speed via modulation of the PM H+ ATPase activity. Further research on the still unanswered questions on the pollen specificity of certain PM ATPase isoforms, their biochemical and structural properties, the trafficking of PM H+ ATPases to the pollen plasma membrane and putative regulatory interaction partners may reveal new features of this protein corresponding to the specific demands of a highly specialized cell, the pollen.