Up against the wall: arabinogalactan-protein dynamics at cell surfaces


There is still much to discover about plant cell walls and about how they are coordinately assembled, extended and moulded during organogenesis. Arabinogalactan proteins (AGPs) are proteoglycans that are abundant at cell surfaces and are widespread across plant taxa. They are implicated in several aspects of cell wall biology including the signalling of cell context and cell expansion, yet remain elusive in terms of their mechanistic involvement in cell processes (Gaspar et al., 2001; Showalter, 2001). Structural and immunochemical analyses reveal considerable heterogeneity to the arabinogalactan component that can account for over 90% of AGP mass (Showalter, 2001). Biochemical and bioinformatics approaches indicate tens of genes encoding AGP core proteins within a species (Gaspar et al., 2001). The collective of all AGPs within an organ are assumed, to some extent at least, to share a set of properties that relates to their function. A report in this issue of New Phytologist takes AGPs, as a collective, head on and presents an intriguing set of observations concerning their occurrence and dynamics at cell surfaces (Lamport et al., pp. 479–492).

What Lamport et al. have done is to quantify AGPs (as defined through the binding of the coloured synthetic phenyl glycoside known as β-glucosyl Yariv reagent (YR)), in cell surface compartments. YRs, by means of their capacity to specifically bind to AGPs, have been widely used to quantify AGPs, to locate them in plant materials and to explore AGP function by their capacity to aggregate AGPs when added to living systems (Nothnagel, 1997; Showalter, 2001). Using suspension-cultured cells adapted to grow in NaCl, Lamport et al. quantified YR-binding AGPs in plasma membrane, cell wall and growth medium fractions. A previous report had indicated lowered levels of plasma membrane AGPs in salt-adapted cells (Zhu et al., 1993) and this is confirmed in the current work that also goes on to demonstrate a dramatic (over five-fold) increase in AGPs in the growth medium of the salt-adapted cells relative to control medium (Lamport et al.). Classical AGPs, a major subset of all AGPs, are attached to plasma membranes by means of glycosyl-phosphatidyl inositol (GPI) anchors (Schultz et al., 2000), thus providing a potential mechanism for AGP release from plasma membranes into the cell wall by phospholipase action. The study has also indicated that large amounts of AGPs are released by sonic disruption of intact cells but not from the equivalent disruption of protoplasts (Lamport et al., 2006). This is interpreted as a release of AGPs from the periplasm (the region between the plasma membrane and the cell wall), distinct from cell wall itself. In short, Lamport et al. interpret their findings to indicate that AGPs move sequentially through the locations of the cell surface – plasma membrane, periplasm, cell wall and into the medium – and that the flux of AGPs through these compartments is dramatically increased in the salt-adapted cells. This is an interesting set of observations that raises important questions and fuels speculation upon what the properties and functions of cell surface AGPs may be.

‘… it is speculated that AGPs have the capacity to migrate through cell walls acting as plasticizers that increase wall extensibility.’

AGPs and the expanding cell

A range of studies involving YR activity, AGP occurrence in mutants and AGP core protein manipulation has implicated AGPs in cell expansion in both tip-growing and multicellular systems. An unanswered aspect of these studies is the mechanism of AGP action. Two possible functions are proposed in the Lamport et al. paper for AGPs released from the plasma membrane. Firstly, they are proposed to act as a perisplasmic buffer stabilizing the plasma membrane–cell wall interface. However, it seems likely that plasma-membrane-bound AGPs could fulfil such a function and that this relates to the plasmalemmal reticulum AGPs proposed to link cell walls to the cytoskeleton (Pickard & Fujiki, 2005). An occurrence of soluble periplasmic AGPs may relate more persuasively to the second proposed function. In this, it is speculated that AGPs have the capacity to migrate through cell walls, acting as plasticizers that increase wall extensibility. To account for the paradox that the smaller, slower-growing salt-adapted cells have the higher flux of AGPs across their surfaces, it is argued that other factors constrain cell expansion in these cells. The idea of a plasticizer, of course, implies an intimate interaction with cell wall polymer systems. Lamport et al. suggest that AGPs may act on the pectic network to increase cell wall porosity. It is of interest that Takeda & Fry (2004) have observed that AGPs have the capacity to increase the activity of xyloglucantransglycosylases and thus they could impact upon the assembly or properties of the cellulose–xyloglucan networks.

Evidence in support of the idea that a release of AGPs from plasma membranes is involved in cell expansion comes from studies on tip-growing systems, where AGPs can be specifically detected in the cell wall at the points of cell extension (Lee et al., 2005; Pereira et al., 2005). In the moss Physcomitrella patens, a specific tip-localised cleavage of AGPs from the plasma membrane resulting in their appearance in the cell wall (Lee et al., 2005 and as seen in Fig. 1) has been proposed to be a facet of apical cell extension. This tip-localised detection of AGPs is disrupted by YR that also blocks apical cell growth. It is also of interest that, in this moss system, high levels of AGPs accumulate in the growth medium (Lee et al., 2005).

Figure 1.

Micrograph showing indirect immunofluorescence labelling of an intact germinated Physcomitrella patens spore with a monoclonal antibody that binds to arabinogalactan proteins (AGPs). The LM6 arabinan epitope is specifically located in the cell wall at the tips of extending apical cells (arrows), indicating a tip-focused release of AGPs from the plasma membrane. For details, see Lee et al. (2005). Scale bar, 1 mm.

Questions of flux

A major unanswered question is how AGPs get from the membrane/periplasm to the growth medium, more specifically how they get through cell walls. Release from suspension cell cultures is not merely a by-product of cell expansion or proliferation, because Lamport et al. clearly indicate that AGP accumulation in growth medium can be uncoupled from both proliferation and cell enlargement in the smaller, slower-growing salt-adapted cells. Clearly, at the tip of pollen tubes/Physcomitrella apical cells, AGPs may be released through cell wall regions that have greatly elevated porosity. Indeed, as suggested above, if AGPs can effect a loosening of the cell wall polymer systems then they may effect their own movement through a cell wall. What controls the rate of AGP flux across the cell surface? Lamport et al. argue that the loss of AGPs into the medium from a more porous cell wall could feed back to increase the rate of release from the plasma membrane. Alternatively, it would seem possible that the control of AGP flux could lie directly with phospholipase action regulating the movement of AGPs into the cell wall.

Suspension-culture systems are clearly excellent for determining the dynamics of AGP release, but how closely does AGP release from suspension-cultured cells reflect AGP movement across cell surfaces in a meristem or an expanding organ? Suspension-cultured cells proliferate without coordinated division planes, whereas in a meristem cell proliferation and cell expansion are highly ordered. Are AGPs selectively lost from extending longitudinal cell walls in elongating systems equivalent to that proposed for the GPI-linked COBRA protein (Roudier et al., 2005)? Do AGPs accumulate in middle lamellae between such walls? These are interesting questions to which the observations of Lamport et al. should direct attention.

AGPs: modules, individuals and the collective

The hypothesis that AGPs act directly to modulate cell wall properties is a good one that should be readily testable. But can a common property of the collective of AGPs be reconciled with the diversity of individual AGPs in terms of their protein and glycan components? Much recent effort has been aimed at dissecting the function of individual AGP core proteins, and indeed subtle effects on diverse aspects of plant development have been identified following gene disruption (e.g. van Hengel & Roberts, 2003; Acosta-Garcia & Vielle-Calzada, 2004; Gaspar et al., 2004). Do the observed effects of specific AGP loss in diverse developmental processes all arise from a related impact on cell walls, albeit in a restricted cell type or cell wall region? Another aspect of AGP complexity is that AGP modules can occur in hybrid molecules with other protein domains (e.g. Johnson et al., 2003; Mashiguchi et al., 2004; Motose et al., 2004). Is there a particular property that an AGP module imparts to hybrid molecules? Does an AGP module confer on a protein, for example, residence properties at the plasma membrane or a capacity to move through a cell wall?

The mechanism of the impact of an individual AGP or AGPs collectively upon aspects of plant development remains elusive. The paper of Lamport et al. deals with (YR-binding) AGPs as a collective but describes a powerful system suitable for more detailed analyses and it raises a range of questions that will inform approaches to uncover the mechanism of AGP action. Can plasticizer activity be directly assayed in some way by the addition of AGPs to cell walls? Are all AGPs up-regulated equally in response to salt stress or is there up-regulation of a subset? The road ahead for AGP studies remains a challenge, but the insights that such studies could give into mechanisms of cell wall modulation and plant development remain great.