5.2.1 Diffusion through membranes
Passive transport mediates a solute flux in the direction of a lower (electro)chemical potential. Diffusion is therefore by definition a passive process. Nevertheless, it is a crucially important phenomenon for all life. For example, diffusion of inorganic minerals in the soil solution is often a limiting factor for plant nutrition, whereas many smaller cells (e.g. bacteria) rely on diffusion for the cytoplasmic distribution of solutes. Substances can cross membranes by diffusion if they can dissolve in the oily interior of the membrane. Such lipophilic (or hydrophobic) substances include important compounds like O2, CO2, H2O2 and NH3. For example O2-CO2 gas exchange in lungs and in photosynthesising plant tissues operates by this process (Colour plate 5.2a). Another example is the plant hormone ethylene, which plays important role in stress response and in fruit ripening. These processes rely on the physical properties of lipid membranes and on the chemical and physical properties of the diffusing molecule. Such processes do not involve specific proteins (Colour plate 5.2a) and are therefore not submitted to any significant level of regulation.
An obvious disadvantage of relying on simple diffusion is the total lack of control which cells can exert on the flux of substances that move by this process. Organisms have therefore developed specific transport systems that allow diffusion across membranes to occur through dedicated proteins. Clearly, such systems can function only when the overall membrane permeability for the particular compound is relatively low, which is certainly the case for charged particles (ions) and also for many important substances such as sugars. Plant cells use facilitated diffusion through carrier-type transporters, through ion channels and through water channels. Ion and water channels will be dealt with in more detail below and the focus here is on some plant carrier systems that mediate facilitated diffusion.
Nitrogen is required by plants in vast quantities and many groups have studied the mechanism and identity of nitrogenous compounds that are taken up by plant roots. Most plants acquire N in either of two forms: as nitrate (NO3-) or ammonium (NH4+) ions, or a mixture of the two. The translocation of NH4 + is mediated by proteins of several transport families each containing multiple isoforms. Recently, one isoform of the tomato AMT (ammonium transporter) family (LeAMT1;1) was heterologously expressed in oocytes of Xenopus laevis (Ludewig et al., 2002), a convenient system to perform voltage clamp experiments (see Section 22.214.171.124). Inward current could be observed when micromolar amounts of NH4+ were added to the external medium showing (i) net positive charge was entering the oocyte interior and (ii) the transporter has a high affinity. Subsequently, the authors showed that the inward current was not sensitive to any other ion which might act as a coupled driving force. The conclusion, therefore, was that NH4+ is transported into the cell via a carrier-type uniport mechanism (Colour plate 5.2b), an excellent example of facilitated diffusion.
A second example of a putative carrier that mediates facilitated diffusion is HKT1. HKT1 is probably expressed in the plasma membrane and contains seven to nine transmembrane domains (TMD). In arabidopsis, there is only one HKT isoform, but in other species such as rice, the HKT family extends to seven or eight members. Initially, HKT1 was characterised as a K+:Na+ symporter (Rubio et al., 1995), but it has become increasingly evident that at least in arabidopsis, HKT1 only transports Na+ (Maser et al., 2002; Berthomieu et al., 2003). Thus, although there may be conditions where HKT1 can function as a K+:Na+ symport mechanism, it appears that in vivo, HKT1 binds a Na+ ion to both binding sites which makes it kinetically equivalent to a uniport mechanism that mediates facilitated Na+ diffusion. The physiological relevance of this system is still under debate. Antisense expression of this transporter in wheat led to a decrease in unidirectional Na+ influx (Laurie et al., 2003) providing direct evidence that it functions as a root Na+ uptake mechanism. In arabidopsis, there is the suggestion that AtHKT1 may function in Na+ translocation from root to shoot and vice versa (Maser et al., 2002; Berthomieu et al., 2003).
Ion channels (Colour plate 5.3) are integral membrane proteins that allow ions to pass when the protein is in the open state. Channels function as regulated pores and, simplistically put, all they do is be open (conduct) or closed. So, in contrast to many carriers and all pumps, ion channels are purely passive transporters. Opening and closing, or gating, of ion channels can depend on many factors but roughly speaking there are voltage-gated and ligand-gated channels and a third group is formed by mechanosensitive channels. Voltage-gated channels (Colour plate 5.3a-c) 'sense' the membrane potential and increase/decrease their open probability (Po) as a function of the membrane potential, whereas ligand-gated channels (Colour plate 5.3d) only open after binding an effector molecule, the ligand, which alters the protein conformation and thus leads to channel opening. Mechanosensitive channels play important roles in mammalian functions such as touch and hearing, whereas in plants they are envisaged to sense processes such as turgor-related changes in cell volume. In all ion channels, the pore domain constitutes the transmembrane aperture through which ions are conducted. The pore is usually water-filled, which makes it an attractive environment where ions can diffuse across the membrane. One region in the pore forms the selectivity filter (Colour plate 5.3a), an area that determines which ions are allowed through the pore.
There are three main characteristics that define an ion channel: channel conductance, channel selectivity and channel gating (Hille 2001). Conductance G is the reciprocal of resistance R defined in Ohm's law as:
with V the voltage expressed in volts (V) and I the current expressed in amperes (A). Conductance is expressed in siemens (S) and for ion channels is typically in the range of pS. The larger the conductance, the bigger the flux of ions that can pass through the channel. Often ion channel nomenclature refers to the most permeant ion, or the most permeant and physiologically relevant ion. Which ion permeates will depend on channel selectivity, which is typically defined by the selectivity filter, an area that provides a physical barrier and filters on the basis of ion size but also on the basis of electrostatic interactions between the permeating ion and charged residues of the channel protein. Thus, cation channels typically show negative residues lining the channel pore, whereas anion channels contain positive charges in this region.
In voltage-gated channels, opening of the pore strictly depends on changes in membrane potential Em. A decrease in membrane polarisation, depolarisation, stimulates opening of depolarisation-activated channels (Colour plate 5.3b), whereas increased membrane polarisation, hyperpolarisation, leads to opening of hyperpolarisation-activated channels. The link that connects changes in Em to channel gating is provided by a voltage sensor domain (Colour plate 5.3a), typically a transmembrane alpha helix with a large number of charged residues. Under the influence of changes in Em, movement of this domain translates into conformational changes in the channel protein leading to opening or closing of the pore. Although opening of voltage-gated channels requires a change in Em, there may be many factors other than membrane voltage that impact on the channel open probability, such as the presence of second messengers, the redox state or the phosphorylation state of the channel protein.
The ligand-gated ion channels constitute another main class of ion channel. In this type of transporter, gating occurs only after binding of specific compounds to the channel protein (Colour plate 5.3d). These compounds are called ligands, i.e. substances that bind to another compound to form a complex. The term agonist (literally meaning a 'competitor') is also frequently used since often different ligands can have similar actions and compete for the same binding site. In contrast, antagonists bind at the same site but have an opposite action to ligands. Ligand binding to the channel protein causes a conformational change that switches the channel from the closed to the open state or vice versa. Ligand binding is a minimum requirement for channel gating but, as for voltage-gated channels, overall gating properties of ligand-gated channels are often modulated by additional parameters which can include membrane polarisation (Hille, 2001). In the following section, some of the major classes of ion channel found in plants are briefly reviewed.
Potassium (K+) channels were the first class of ion channel described in plant cells and it was found that these consisted of two major categories, inward and outward rectifying (Maathuis et al., 1997; see also Colour plate 5.4a). The inward K+ channels are activated upon membrane hyperpolarisation and form a major pathway for K+ uptake, whereas the open probability of outward channels increases when the membrane depolarises, leading to loss of K+ from the cell. Examples of major inward and outward rectifying voltage dependent channels have been cloned (see Very and Sentenac, 2001 for a review) and their overall structure resembles that of mammalian Shaker-type channels with six TMD, a voltage sensing domain in the S4 region and a pore with selectivity filter in the S5-S6 part of the channel (Colour plate 5.4b). The selectivity filter contains the conserved GYGD motif, a sequence found in virtually all ion channels that are selective for K+, and the S4 region ensures that these K+ channels require a change in Em to transit from closed to open state. Functional channels are made up of four subunits either as homomers or as heteromers. Apart from Shaker-like channels, the major families of plant K+ channels also contain two or four TMD-structured channels and voltage-independent K+ channels.
Ca2+ channels have also been intensely studied in plants. In contrast to mammalian Ca2+ channels, plant Ca2+ channels are usually not very selective for Ca2+ and will also conduct many other cations such as K+ (White, 2000). Plant Ca2+ channels have mainly been characterised at the tonoplast where they are believed to play a role in Ca2+ signalling. These vacuolar Ca2+ channels consist mainly of ligand-gated channels that require IP3, NAADP or cADPR to be gated (White, 2000). However, one putative Ca2+ channel (TPC1) is voltage gated and has a 12 TMD structure (Peiter et al, 2005).
Non-selective ion channels allow passage of various ions and are often permeable to mono- and divalent ions. Plant non-selective cation channels have been characterised to some extent and generally found to be voltage independent (Demidchik et al.,
2002). Although some non-selective cation channels have been shown to be regulated by cyclic nucleotides (Leng et al., 2001; Maathuis and Sanders, 2001; Balague etal.,
2003), it is often not clear how their gating is controlled. These channels are believed to be important in signalling, in turgor regulation and in cation nutrition. They are of special relevance regarding plant stress since they may form a major conduit for the entry of toxic ions such as Na+ (Demidchik et al., 2002). Two gene families, cyclic nucleotide gated channels (CNGCs; Talke et al., 2003) and glutamate receptors (GLRs; Davenport, 2002), are believed to encode non-selective channels and are found in many plant species.
Only one family of anion channels has been described in plants (Hechenberger etal., 1996). Members of the ChLoride Channel (CLC) family show large homology to their mammalian counterparts that are mainly involved in cellular volume regulation. Plant CLCs contain 10-12 TMD and are strongly sensitive to Em with a Po vs. V relationship that is bell shaped. Little is known about the function of plant CLCs, but they are probably involved in early events during cell signalling which frequently involve a Cl- efflux, and possibly in nitrogen nutrition since some CLCs are also capable of transporting NO3- (Geelen et al., 2000). A further function may be in turgor regulation and turgor-driven movement when large amounts of both cations and anions are moved across tonoplast and plasmamembrane.
In all forms of life, water is the solvent for cellular solutes. In addition, water in plants is necessary to generate turgor and to provide a medium for mass flow of solutes (Chapters 9 and 10). Indeed, movement of water is intricately linked to that of ions and, as is the case for most ions, its transport is regulated and controlled. However, in contrast to ions, there are no known active transport mechanisms for water and thus water movement is always passive and directed towards a lower water potential. Although water permeability of phospholipid bilayers is substantial, it is now clear that biological membranes contain specific protein-based pathways for the movement of water. These water channels or aquaporins constitute a parallel and regulated pathway for water flux, both at the intracellular and the whole plant level (Chrispeels et al., 2001).
Structurally, aquaporins have a protein topology that is very similar to Shaker-type ion channels, with each subunit having six TMD. However, rather than one, aquaporin subunits have two pore-forming loops: one in the first half of the protein between TMD two and three and the other between TMD five and six (Colour plate 5.5). The second loop contains a cysteine residue where mercury can bind, leading to channel blockage. The aquaporin-specific signature motif NPA (asparagine, proline, alanine) is involved in the stabilisation of the pore-forming loops. Similar to Shaker ion channels, functional aquaporins consist of tetramers in which the eight pore loops combine to form transmembrane water-conducting paths. The water-conducting pore is around 0.30 nm wide at its narrowest point, a very close fit to the 0.28 nm size of a water molecule, and thus an efficient barrier to other substances. Apart from size exclusion, pore residues interact with the permeating water molecules mainly through H+ bonds. These structural properties ensure that aquaporins are highly selective for water molecules, which move through the protein in a single file, and that water channels are virtually impermeable to any charged species. The latter include protons, a remarkable property since protons can usually be transferred readily through water molecules. Despite this high level of substrate specificity, aquaporins can sustain very high transport rates of ~3 x 109 water molecules per second.
Although aquaporins more or less completely exclude charged species and are highly selective for water, several have been shown to be capable of transporting small non-charged solutes such as glycerol and urea. Sometimes a distinction is made between water-selective aquaporins and those that can conduct water and small solutes, the aquaglyceroporins. Comparative studies using an Escherichia coli isoform of each group showed that although both proteins were structurally and genetically highly similar, minor changes in the pore of the aquaglyceroporin result in a slightly wider selectivity filter and hence glycerol permeability (Wang et al., 2005).
In plants, aquaporins are predominantly expressed in vacuolar and plasma membranes and fall into four major classes, of which some are related to their membrane location. PIPs (plasma membrane intrinsic proteins) are expressed predominantly in the plasma membrane and are subdivided into the PIP1 and PIP2 subfamilies (Jo-hanson et al., 2001). In the tonoplast, TIPs (tonoplast intrinsic proteins) are present, with some isoforms specifically targeted to storage vacuoles, and others to lytic vacuoles. NIPs (NOD26-like intrinsic proteins) have largely unknown membrane locations apart fromNOD26 itself, which is expressed in root nodules. A fourth class comprises the SIPs (small basic intrinsic proteins), whose membrane location is also largely unknown (Johanson et al., 2001). These four classes tend to be members of big gene families with around 35 members in the model species arabidopsis. Such large gene families might suggest that functional redundancy is common for water channels but could also point to a need for isoform-specific expression patterns of aquaporins with finely tuned functional adaptations.
There is now good evidence that both transcriptional and post-transcriptional regulation of aquaporin activity takes place in plant membranes (Luu and Maurel, 2005). For example, root hydraulic conductivity can be seen to correspond closely to PIP1 mRNA levels in Lotus japonicus (Johanson et al., 2001) and the diurnal action of motor cells that move Samanea saman leaves closely mirrors levels in PIP2 (Johanson et al., 2001). In response to salt stress, root PIPs and TIPs are rapidly down-regulated but with a time difference between PIPs and TIPs (Maathuis et al., 2003). Post-translationally, both aquaporin glycosylation and phosphorylation appear to be major mechanisms for the regulation of hydraulic conductance. While glycosylation may be involved in the recruitment of protein to the relevant membrane (Vera-Estrella et al., 2004), phosphorylation directly enhances channel activity (Chrispeels etal., 2001; Johanson etal., 2001). Other factors that impact on aquaporin activity are H+ and Ca2+ with both these ions having a blocking effect on water transport through aquaporins.
Was this article helpful?