Can Ca²⁺ Fluxes to the Root Xylem Be Sustained by Ca²⁺-ATPases in Exodermal and Endodermal Plasma Membranes?

Abstract

The pathway of Ca²⁺ movement from the soil solution into the root stele has been a subject of controversy. If transport through the endodermis is assumed to be through the cytoplasm, the limiting factor is believed to be the active pumping of Ca²⁺ from the cytoplasm into the stele apoplast through the plasma membrane lying on the stele side of the Casparian band. By analogy, for similar transport through the exodermis, the limiting step would be the active pumping into the apoplast on the central cortical side of the layer. Such effluxes are mediated by Ca²⁺-ATPases. To assess whether or not known Ca²⁺ fluxes to the stele in onion (Allium cepa) roots could be supported by Ca²⁺-ATPases, the percentages of total membrane protein particles required to effect the transport were calculated using measured values of membrane surface areas, an animal literature value for Ca²⁺-ATPase V_max, plant literature values for Ca²⁺-ATPase K_m, and protein densities of relevant membranes. Effects of a putative symplastic movement of Ca²⁺ from the exo- or endodermis into the next cell layer, which would increase the surface areas available for pumping, were also considered. Depending on the assumptions applied, densities of Ca²⁺ pumps, calculated as a percentage of total membrane protein particles, varied tremendously between three and 1,600 for the endodermis, and between 0.94 and 1,900 for the exodermis. On the basis of the data, the possibility of Ca²⁺ transport through the cytoplasm and membranes of the exodermis and endodermis cannot be discounted. Thus, it is premature to assign an entirely apoplastic pathway for Ca²⁺ movement from the soil solution to the tracheary elements of the xylem. To verify any conclusion with certainty, more detailed data are required for the characteristics of exo- and endodermal Ca²⁺-ATPases.

The root endo- and exodermis are highly specialized layers that lie on the radial path of ion and water uptake in plant roots. Roots of virtually all vascular plants have an endodermis (Damus et al., 1997), and many angiosperm species also possess an exodermis (Perumalla et al., 1990; Peterson and Perumalla, 1990). When mature, these layers are characterized by having Casparian bands (deposits of lignin and suberin in the interstices of their anticlinal, primary walls) and often suberin lamellae (thin layers of suberin laid down on the inner faces of all the primary walls). In onion (Allium cepa), the exodermis is dimorphic, being composed of long and short cells (Kroemer, 1903). At maturity, all exodermal cells have Casparian bands; the long cells have suberin lamellae but the short cells tend to lack these structures, being functionally similar to the passage cells of the endodermis (Kamula et al., 1994). In the exodermis, suberin lamella formation severs the plasmodesmata, and long cells die. This differs from the situation in the endodermis where the plasmodesmata remain intact during and after suberin lamella deposition, and the cells remain alive (Ma and Peterson, 2000, 2001a).

Like the other essential ions absorbed by the root, Ca²⁺ moves radially through the root tissues (epidermis, exodermis [when present], central cortex, endodermis, pericycle, and perhaps through xylem parenchyma) before entering the lumena of the tracheary elements. Previous work has conclusively shown that Ca²⁺ passes through the mature, intact exodermis and endodermis of onion roots on its way to the root stele (Cholewa and Peterson, 2004). Does this ion move through the endodermis and exodermis apoplastically or symplastically (or both)?

The controversy concerning the pathway of Ca²⁺ through the root exodermis and endodermis springs from its normally low concentration in the cytoplasm (in the range of 10⁻⁷ m), from the question of the permeability of Casparian bands to ions, and from experimental results that favor one or the other pathway. (For a full discussion of these topics, see Cholewa, 2000; White, 2001; and Cholewa and Peterson, 2004.) Briefly, movement of Ca²⁺ through the cytoplasm of the exodermis and endodermis is deemed unlikely because of the low resting levels of the free ion compared to much greater levels in the apoplast (10⁻³ m to 10⁻⁴ m; for additional information, see Bjorkman and Cleland, 1991; Evans et al., 1991; and Sattelmacher, 2001). However, diverse tests have demonstrated the impermeability of Casparian bands to ions (de Rufz de Lavison, 1910; Baker, 1971; Robards and Robb, 1972; Nagahashi et al., 1974; Peterson, 1987), including Ca²⁺ (Kuhn et al., 2000; Cholewa and Peterson, 2004). Clarkson (1991) has proposed a model whereby Ca²⁺ moves through the endodermal cytoplasm: The ions enter through Ca²⁺ channels in the plasmalemma on the cortical side of the Casparian band and are pumped out by Ca²⁺-ATPases in the plasmalemma on the stele side of this structure. The same model may be applied to the mature exodermis. The postulate that ions cross membranes en route to the stele is supported by results of a previous study that showed that in onion, Ca²⁺ does not freely diffuse through the Casparian band of the exodermis and that inhibitors of Ca²⁺channels and Ca²⁺-ATPases reduce the delivery of the ion into the stele by 73% (Cholewa and Peterson, 2004). Clarkson's model, however, has been criticized by White (1998, 2001), who claims that unreasonably large quantities of Ca²⁺-ATPase (responsible for the limiting step) would need to be present in the endodermal plasmalemmas to explain modeled amounts of Ca²⁺ entering the root stele.

This contribution is a quantitative consideration of Ca²⁺ movement into the stele of the onion root, making all assumptions explicit. From the outset, it was assumed that all Ca²⁺ entering the root stele passed across the membranes and through the cytoplasms of the endo- and exodermal cells and that no Ca²⁺ moved through cells with suberin lamellae, the latter having been demonstrated by Moore et al. (2002) for Arabidopsis (Arabidopsis thaliana) endodermis. Two developmental stages of the root were considered. The younger was a zone where the endodermis had matured to the extent that it had Casparian bands; the exodermis was immature. The older was a region in which most of the endodermal cells had also developed suberin lamellae; the exodermis had matured (with Casparian bands in all cells and suberin lamellae in long cells). Measurements of net Ca²⁺ fluxes through both zones were available from a previous study (Cholewa and Peterson, 2004). Clarkson's model was extended to include symplastic movement through plasmodesmata from the layer in question into the innermost single layer of neighboring cells (the pericycle in the case of the endodermis, and the central cortex in the case of the exodermis). The question was asked, “What proportion of the plasmalemma proteins in the endodermis and exodermis (and adjacent layers when considered) would need to be Ca²⁺-ATPase to account for the known flux of Ca²⁺ into the onion root stele?”

RESULTS

Net Flux of Ca²⁺ into the Onion Root Stele

According to Cholewa and Peterson (2004), the amount of Ca²⁺ transported to the stele in the young root zone was 26.7 nmol (10 mm root length)⁻¹ 24 h⁻¹. In the mature zone, the amount had been reduced to 8.7 nmol (10 mm root length)⁻¹ 24 h⁻¹ (Table I).

Table I.

Data obtained from the literature

Parameter	Value	Reference
Flux of Ca2+ (24 h−1) into stele in a 10-mm length of root		Cholewa and Peterson (2004)
Young root zone	26.7 nmol
Old root zone	8.70 nmol
Area of tangential plasmalemmas in short cells of onion root exodermis in a 10-mm root length	2.05 mm2	Kamula et al. (1994)
Activity of Ca2+-ATPase	50 ions s−1	Garrahan and Rega (1990)
Protein packing density in maize (Zea mays; number of particles per μm2 membrane surface)		Robards et al. (1980)
Endodermal plasmalemma	2,985 μm−2
Central cortical cell plasmalemma	1,450 μm−2

Measurements of Relevant Membrane Surface Areas

Surface areas were calculated from measurements made of various cells in onion roots and assuming a 10 mm length of root (Table II). In the young zone, the surface area of the relevant endodermal plasma membrane (i.e. the membrane on the stele side of the Casparian band) was 11 mm², henceforth referred to as 100% of the original endodermal surface area. If it is assumed that Ca²⁺ moves through plasmodesmata into the pericycle, thus enlarging the symplastic compartment from which the ion can be pumped into the apoplast, the plasma membrane surface is increased to 33.9 mm² (i.e. to about 300% of the original endodermal surface area).

Table II.

Surface areas of selected membranes in a 10-cm length of onion root

Root Zone	Cell Layers Involved	Values from Measurements	Surface Areas of Relevant Membranes
			mm2
Young	Endodermis	Radial wall internal to Casparian band (root cross-section): 4.5 μm	11
		Tangential wall (root cross-section): 25 μm
		Length of cell: 155 μm
		Number of cells: 30
	Pericycle	Radial wall (root cross-section): 7 μm
		Tangential wall (root cross-section): 15 μm
		Length of cell: 155 μm
		Number of cells: 30
	Endodermis plus pericycle		34
Old	Endodermis		2.5
	Endodermis plus pericycle		26
	Exodermis short cells		2.05a
	Cortical parenchyma adjacent to exodermis	Diameter of cell: 29.7 μm
		Length of cell: 440 μm
		Number of cells: 90
	Exodermis short cells plus central cortical parenchyma layer		89

Values were obtained by measuring the walls adjacent to the desired membranes.

^aLiterature data. See Table I.

In the old root zone, structural changes had occurred in the endodermis. Its relevant membrane surface area was reduced compared to that in the young zone, owing to the development of suberin lamellae, which appear to be impermeable to Ca²⁺ (Moore et al., 2002). Because about 23 of the mean total of 30 endodermal cells developed suberin lamellae (M. Hayter and C. Peterson, unpublished data), the relevant membrane surface area through which Ca²⁺ could move became 2.5 mm² (23% of the original endodermal surface area; Table II). The effect of adding the membranes of the pericycle to those of the endodermis in the old region of the root was striking because the membrane area of the endodermis had been reduced by development of suberin lamellae in the majority of its cells. In this case, the membrane surface of the endodermis alone (2.5 mm²) was increased by more than 10-fold (to 26 mm²) as a result of addition of the pericycle.

In the case of the exodermis, suberin lamellae had developed in its long cells, leaving only the membranes of the short cells to consider. The relevant membrane surface area in this layer is 2.1 mm² (Table II), representing only 7% of the total exodermal surface area (assuming an exodermal diameter of 1 mm). The effect of including an additional cortical layer brought about a very striking increase in available membrane surface area because of the relatively small plasmalemma area of the exodermal short cells and the greater area of plasma membrane contributed by the large central cortical cells. The increase was from 2.1 mm² to 89 mm² (Table II).

Values for Ca²⁺-ATPase Activity and K_m

To the best of our knowledge, to date there is no data on the activity of purified Ca²⁺-ATPase from a plant source. A reasonable estimate of the pump's V_max of 50 ions pump⁻¹ s⁻¹ was obtained from animal literature (Table I). A range of K_m values of between 0.4 μm to 10 μm is known from plant literature (Table I).

Densities of Protein Packing in Plasma Membranes

From freeze fracture images, Robards et al. (1980) obtained data on protein particles in plasma membranes of endodermal and other root cortical cells. Using their values from cryoprotected membranes, the average for the P face is 1,450 particles μm⁻² for central cortical cells and 2,985 particles μm⁻² for endodermal cells (Table I).

Number of Ca²⁺-ATPase Molecules Required for Ca²⁺ Movement to the Stele

At a given protein density in the plasma membrane, the number of Ca²⁺-ATPase molecules required to sustain the known flux of Ca²⁺ into the onion root stele depends on the activity of the pumps and on the area of the membrane through which pumping occurs. The required numbers of Ca²⁺-ATPases were calculated for young and old zones of onion roots with various cell layers involved and with three levels of pump activity (i.e. for V_max, 1% and 33% V_max). As their activity is assumed to diminish, more pump molecules would be required to support Ca²⁺ efflux from the cytoplasm (Table III).

Table III.

Number of Ca²⁺-ATPase molecules required (per μm square membrane surface) to account for the observed rates of Ca²⁺ transport into the onion root stele

Root Zone	Cell Layer(s) Involved	Activity of Ca2+-ATPase Is Vmax	Activity of Ca2+-ATPase Ranges from 33% Vmax to 1% Vmax
Young	Endodermis	340	1,000–34,000
	Endodermis plus pericycle	110	330–1,100
Old	Endodermis	480	1,400–48,000
	Endodermis plus pericycle	46	140–4,600
	Exodermis	580	1,700–58,000
	Exodermis plus central cortical layer	14	420–1,400

Pump activity at V_max is 50 Ca²⁺ ions pump⁻¹ s⁻¹. Data also shown for 33% and 1% V_max, and for adding the membrane surface areas of cell layers adjacent to the endodermis and exodermis in young and old root zones.

Required Ca²⁺-ATPase as a Percent of Total Plasmalemma Protein Particles

Values from Tables I and III were used to calculate the percent of total membrane protein that would need to be Ca²⁺-ATPase to account for the known flux of Ca²⁺ into the onion root stele. It was assumed that both the plasma membranes of the endodermis and short cells of the exodermis have the numbers of proteins reported for the endodermis by Robards et al. (1980). It was further assumed that the cells of the cortex and pericycle had the numbers of proteins measured for cortical cells (Robards et al., 1980). Assuming a Ca²⁺-ATPase activity of V_max or 33% V_max, in no case did the required number of pumps exceed the total number of protein particles in the membranes of endodermis and exodermis alone (Table IV). Adding the plasmalemma surface area from one rank of the neighboring cells reduced the percentage of total proteins required to be Ca²⁺-ATPase by a factor of about 2 in the case of the endodermis in the young zone, about 5.5 in the old zone, and about 21 for the exodermis in the old zone (Table IV).

Table IV.

Ca²⁺-ATPases (percent of total membrane protein particles) required for Ca²⁺ movement into onion root stele

Root Zone	Cell Layer(s) Involved	Activity of Ca2+-ATPase Is Vmax	Activity of Ca2+-ATPase Ranges from 33% Vmax to 1% Vmax
Young	Endodermis	11	34–1,100
	Endodermis plus pericycle	5.7	17–570
Old	Endodermis	16	48–1,600
	Endodermis plus pericycle	2.9	8.8–290
	Exodermis	19	59–1,900
	Exodermis plus central cortical layer	0.94	2.9–94

Activity of Ca²⁺-ATPase pump at V_max is 50 Ca²⁺ ions pump⁻¹ s⁻¹. Effects of using 33% V_max and 1% V_max, and of adding the membrane surface areas of cell layers adjacent to the endodermis and exodermis are also considered.

If a value of 1% V_max is used in the calculation of these data, very different results are obtained. In all cases save one, the number of required particles exceeds the total protein particles expected to be in the membranes, and by a wide margin (Table IV).

DISCUSSION

To describe the pathway of Ca²⁺ movement from the soil solution to the stele correctly, it is necessary to know whether or not it is possible for this ion to pass through the cytoplasm of the mature endodermis and exodermis. The limiting step is considered to be the centripetal calcium efflux from the cytoplasm into the stele apoplast via the pump Ca²⁺-ATPase (see White, 2001). The feasibility of this step depends on the density of the pumps in the membrane, their activity, the total membrane area through which transport occurs (i.e. on the size and number of cells involved), and the quantity of Ca²⁺ moved. In onion, the amounts of Ca²⁺ moving into the stele in regions before and after maturation of the exodermis are 26.7 nmol 24 h⁻¹ and 8.7 nmol 24 h⁻¹, respectively (Cholewa and Peterson, 2004). Are there enough proteins in the inner tangential plasma membranes of the endodermis and exodermis to mediate the measured Ca²⁺ flux? The answer to this question depends entirely on the assumptions made regarding several parameters.

Activity and K_m of Ca²⁺-ATPase

Pump activity in situ is related to its V_max and K_m. To the best of our knowledge, the V_max for a plant-derived Ca²⁺-ATPase has not been measured for pumps purified from plasma membranes. Thus, the value of 50 Ca²⁺ ions (Ca²⁺-ATPase molecule)⁻¹ s⁻¹ reported for purified animal Ca²⁺-ATPase was used (Garrahan and Rega, 1990). V_max data can be calculated from plants, but only after making several assumptions. For example, if we assume a pump molecular mass = 120,000 D (see Evans and Williams [1998] for a range of values) and further assume that Ca²⁺-ATPase represents 0.1% by weight of the total membrane protein (Chen et al., 1993), then V_max values range between 3 and 57 (Table V).

Table V.

Estimated values for V_max Ca²⁺-ATPase activity in plasma membranes of plants

Plant Species	Organ	Vmax	Ca2+-ATPase Activity	References
		pmol Ca2+ mg protein−1 s−1	Ca2+ ions (pump molecule)−1 s−1
Maize	Leaf	475	57	Kasai and Muto (1990)
Bryonia dioica	Tendril	40	5	Liss and Weiler (1994)
Wheat	Root	25	3	Olbe and Sommarin (1991)

Primary data were derived from the literature. Assumptions: Ca²⁺-ATPase molecular mass = 120,000 D; Ca²⁺-ATPase constitutes 0.1% of the total membrane protein (Chen et al., 1993).

The rate at which Ca²⁺ is pumped from the cytoplasm into the apoplast of the stele also depends on the K_m of the Ca²⁺-ATPase and the concentration of free Ca²⁺ in the cytoplasm. According to Kiegle et al. (2000), this concentration is 0.1 μm in the endodermal cells of Arabidopsis. K_m values of Ca²⁺-ATPases from various plant sources range from 0.4 to 10 μm (Evans and Williams, 1998). Using this range and a free Ca²⁺ cytoplasmic concentration range of 1 to 2 μm, the extremes of Ca²⁺-ATPase activity in situ can be expected to vary from 1% to 33% V_max. If one uses a value of 33% V_max to calculate the proportion of protein particles that would need to be Ca²⁺-ATPase to sustain known Ca²⁺ fluxes, values of less than 100% are obtained (Table IV). In other words, the number of Ca²⁺-ATPase molecules does not exceed the total number or protein particles in the membrane. In this calculation, all values obtained were considerably higher than that of 0.1% cited by Chen et al. (1993) for carrot (Daucus carota) endoplasmic reticulum. However, the membranes of the endodermis and xylem parenchyma are known to be specialized with respect to their protein content, e.g. proton-ATPases (see Sattelmacher, 2001). The membranes of the short cells of a dimorphic exodermis have not yet been tested in this regard, but it is reasonable to assume that they are also specialized because the anatomy of the system dictates that all ions entering the root go through these cells (Ma and Peterson, 2000, 2001a, 2001b). To date, there are no quantitative measures of amounts of a specific protein in a plasma membrane from these cells. However, to take an extreme example from the animal literature, 70% of the protein in the skeletal muscle sarcoplasmic reticulum can be Ca²⁺-ATPases (Garrahan and Rega, 1990).

If one assumes the lower extreme of 1% V_max (cytoplasmic Ca²⁺ is 1 μm and the K_m of the Ca²⁺-ATPase is10 μm), it is clear that there would not be enough protein transporters (by a factor greater than 10) in the membranes of either the endo- or exodermis to support the previously measured Ca²⁺ flux. The dramatic discrepancies highlighted here by considering extreme cases illustrate the necessity of knowing, at sufficient accuracy, both kinetic parameters for plant Ca²⁺-ATPases and proportions of membrane protein that could be identified as Ca²⁺-pumps in cells that may be specialized in pumping this ion. There are animal cells that are highly specialized in Ca²⁺ pumping (see above). Garrahan and Rega (1990) report a broad range of K_m values, 0.039 to 4.5 μm, obtained from a variety of animal sources.

Membrane Areas Available for Ca²⁺ Efflux

If Ca²⁺ were to be transported through plasmodesmata, then the membrane surface area available for its passage back into the apoplast would be greatly increased. In this study, the effects of adding the membrane surfaces of one extra layer to the endodermis and exodermis were explored. Although it was assumed that these cells did not have the elevated protein particle densities characteristic of the endodermal cells, when their plasmalemma surface areas were added to those of the endodermis and exodermis, the proportion of total protein particles that would need to be Ca²⁺-ATPase was substantially reduced (Table IV). However, in cases where a pump activity of 1% V_max was assumed, the addition of another cell layer did not bring the required pump numbers below 100% of the total number of particles, except for one case that was still exceptionally high (Table IV).

If the symplastic movement were to involve more than one additional cell layer, the proportion of total membrane protein required to explain the flux of Ca²⁺ would become smaller. However, the idea of an increase of the area available for pumping implies that the movement of Ca²⁺ across plasmodesmata (either by diffusion or solvent drag) is not rate limiting. Furthermore, the problem with scenarios involving Ca²⁺ movement through plasmodesmata is that elevated levels of this ion are known to close these intercellular connections (Erwee and Goodwin, 1983; Holdaway-Clarke et al., 2000). The results of Moore et al. (2002), in which elevated levels of Ca²⁺ generated in endodermal cells lacking suberin lamellae were not propagated to endodermal cells with suberin lamellae, also argue against a facile symplastic transfer of Ca²⁺. If symplastic transport is to be invoked to explain the exit of calcium from the symplast to the apoplast, then the level of free Ca²⁺ in the cytoplasm and plasmodesmata must be kept low. This could be accomplished if a bound form of the ion could move through plasmodesmata.

If we assume that ions other than Ca²⁺ eventually transported in the xylem move radially into the root stele symplastically, then the plasma membranes of the endodermis-pericycle-xylem parenchyma complex must be responsible for transferring them into the apoplast. In onion, all these cells are known to be connected by plasmodesmata (Ma and Peterson, 2001b). It is proposed that a division of labor occurs within this complex so that the efflux of Ca²⁺ occurs mainly through the plasma membranes of endodermal cells, while the efflux of other ions that are easily mobile symplastically are transferred to the apoplast from cells (i.e. of the pericycle and xylem parenchyma) closer to the xylem conduits.

Comparison of Ca²⁺ Fluxes from the Endodermis of Onion and Arabidopsis

Kinetic data for Arabidopsis endodermal cells (from efflux of cytoplasmic Ca²⁺ following application of various stresses) does not agree well with that for onion. Using Figure 4(d) from Kiegle et al. (2000) illustrating the decline in Arabidopsis cytoplasmic Ca²⁺ following a cold shock, the average half-time of the decay is 4.8 s. It is assumed that all the Ca²⁺ leaving the cytoplasm was doing so through the inner tangential plasmalemma. Endodermal cell dimensions were taken from Figure 21 in Baum et al. (2002; mean tangential width = 21 μm; mean radial width = 13 μm). Assuming that the cell length is 10-fold greater than the radial width and that the cytoplasm occupies 5% of the cell volume, we arrive at the very low rate of 3.0 10⁻²³ mol Ca²⁺ (μm plasmalemma surface area)⁻² s⁻¹ compared to that of onion at 2.8 10⁻²⁰ mol Ca²⁺ (μm plasmalemma surface area)⁻² s⁻¹. In this calculation, two assumptions favoring the value obtained from Arabidopsis were made. First, the rate constant was calculated from a starting cytoplasmic free Ca²⁺ concentration of 0.74 μm, and second, it was assumed that all the Ca²⁺ leaving the cell did so through the inner tangential endodermal membrane. Thus, the actual rate of transport of Ca²⁺ through this membrane should be much lower than calculated. This difference between Arabidopsis and onion rates of Ca²⁺ transport may, in part, be explained by an inhibitory effect of the stress (cold) on Ca²⁺-ATPase activity.

Comparison of Ca²⁺ Fluxes through the Endodermis of Onion and Cereal Root

In earlier studies, White (1998, 2001) concluded that in cereals, the protein content of endodermal membranes is insufficient to account for Ca²⁺ efflux into the stele by Ca²⁺-ATPases. Was this conclusion reached because the flux of Ca²⁺ assumed to pass the endodermis of these roots was higher than that measured for onion? White used a Ca²⁺ flux value of 40 nmol h⁻¹ gfw⁻¹. To convert this weight-based value to the units of root length used in this study, a density of one was assumed, as well as a root diameter of 0.469 mm (taken from plate 84A in Esau, 1965). White's Ca²⁺ flux value becomes 0.71 nmol 10 mm⁻¹ 24 h⁻¹, which is considerably lower than the value for onion. However, to compare the fluxes in the two roots directly, they need to be expressed as a function of membrane surface area (since this parameter differs for roots of various endodermal diameters). The diameter of the endodermis in wheat (Triticum aestivum) is 0.17 mm (taken from plate 84A in Esau, 1965) so that the membrane surface in a 10 mm length of root would be 5.4 mm². The comparable surface area in an onion root is 11 mm² (Table II). According to this comparison, the flux of Ca²⁺ through the wheat root would be 0.13 nmol mm⁻² compared to 2.4 nmol mm⁻² in onion. Thus, the Ca²⁺ flux value used for White's calculations was not unusually large but was, in fact, substantially smaller than that used for onion in this study. The extent to which a maximum rate for cereal may have been underestimated due to suberin lamella development is unknown.

CONCLUSION

At present, it is not possible to completely solve the controversy of the pathway of Ca²⁺ movement to the xylem in terms of modeling its transport using kinetic data of Ca²⁺-ATPases of exo- and endodermal cells. Depending on the values assumed for various parameters, the results, nevertheless, indicate that such transport could possibly account for all the Ca²⁺ moving to the root stele. However, such a conclusion strongly rests on the grounds that there is sufficient protein present in the membrane. For a definitive answer, values need to be established for the following:

• (1) Activity of Ca²⁺-ATPase. This needs to be measured from purified protein preparations.

• (2) Proportion of total membrane protein that is Ca²⁺-ATPase. This really needs to be known for the plasma membranes of the endodermis and exodermis, since specialization of membrane function within the root is known to occur in the case of the endodermis and probably also occurs in the exodermis. Immuno methods at the ultrastructural level could provide some qualitative insight into this issue.

• (3) Mobility of Ca²⁺ through plasmodesmata. Although the available data indicate that free Ca²⁺ is immobile, work with a range of cytoplasmic concentrations may reveal a threshold level for the reaction of plasmodesmatal closure. One could then consider whether or not symplastic transport sufficient to explain the observed fluxes of Ca²⁺ is feasible.

In conclusion, when the feasibility of Ca²⁺ transport through the cytoplasm and membranes of the endodermal and exodermal cells was explored by kinetic analyses of Ca²⁺-ATPase activity, very diverse results were obtained. Because of the uncertainty created by this diversity and the results of a previous study (Cholewa and Peterson, 2004), which demonstrated that the Ca²⁺ loaded into the stele of this root was substantially reduced by known inhibitors of calcium channels and Ca²⁺-ATPases, it would be premature to conclude that a major apoplastic bypass is necessary to explain the delivery of Ca²⁺ to the root stele (White, 1998, 2001). If one accepts the available evidence that the Casparian bands of the endodermis and exodermis are virtually impermeable to ions, the young, adventitious onion root, with its endodermis and exodermis undisturbed by developing laterals, can provide a useful experimental system with which to measure the flux of Ca²⁺ and other ions through the cell-to-cell path.

MATERIALS AND METHODS

Proportions of membrane protein particles that would need to be Ca²⁺-ATPase to account for the transfer of Ca²⁺ to the stele of onion (Allium cepa) root, previously observed by Cholewa and Peterson (2004), were calculated after making measurements or assumptions regarding the parameters: area of membrane involved (measured), activity of Ca²⁺-ATPase at V_max (from animal literature) and K_m (from plant literature), and protein packing density in the plasmalemmas of the various cells involved (from plant literature). All values were normalized to a 10-mm length of onion root and a 24-h time period unless otherwise indicated. Calculated data were rounded to two significant figures.

Measurements of Relevant Membrane Surface Areas

Adventitious roots of onion (Allium cepa L. cv Ebeneezer) were obtained from bulbs sprouted in vermiculite. Growth conditions were as previously described (Cholewa and Peterson, 2004). Such roots were used to measure surface areas of plasma membranes through which Ca²⁺ may be moved from the cytoplasm to the apoplast by Ca²⁺-ATPases. These areas were assumed to equal those of the adjacent walls that were measured with the light microscope in freehand, cross sections of onion roots. The terminology traditionally applied to wall position (i.e. transverse and radial) was extended to describe the positions of the adjacent plasma membranes. Cell lengths were obtained from freehand longitudinal sections so that the plasmalemma covering the relevant parts of the transverse walls could be included. Two root zones termed “young” and “old” were considered. Previous research (Cholewa and Peterson, 2004) provided an anatomical characterization of these two zones, as well as calcium flux values from the ambient solution into the root stele for each zone. Areas of exodermal short-cell tangential plasmalemmas were obtained from the literature (Table I).

Young Root Zone

This region of the onion root was characterized by an endodermis with Casparian bands but no suberin lamellae, and an immature exodermis (i.e. with neither Casparian bands nor suberin lamellae). The first assumption was that all Ca²⁺ entering the stele moved from the cytoplasm of the endodermis into the apoplast through the plasma membranes on the stele side of the endodermis. The second assumption was that some Ca²⁺ moved through plasmodesmata that link the cytoplasm of the endodermis to that of the pericycle (Ma and Peterson, 2001a, 2001b) and then moved from the cytoplasm to the apoplast through the plasmalemmas on the stele side of the endodermis plus all the plasmalemmas of the pericycle cells.

Old Root Zone

In this zone, 23 of a total of 30 endodermal cells were assumed to have developed suberin lamellae; the exodermis had Casparian bands and the long cells also had suberin lamellae. Considering the endodermis alone, according to the first assumption, all the Ca²⁺ entering the stele moved from the cytoplasm of the endodermal passage cells (i.e. without suberin lamellae) to the apoplast through plasmalemmas on the stele side of the endodermis. Using the second assumption, symplastic movement (i.e. movement through plasmodesmata) of Ca²⁺ from the endodermis to the pericycle occurred followed by tangential symplastic movement within the pericycle. The membrane area of interest in this case included the plasmalemmas of the endodermal passage cells on the stele side and all the plasma membranes of the pericycle cells.

Assuming that all Ca²⁺ destined for the stele passed through the membranes of the short, exodermal cells, the relevant area with Ca²⁺-ATPase would be on the cortical side of the short cells. Using the second assumption, namely that Ca²⁺ was transferred symplastically to the first rank of central cortical cells and then tangentially symplastically throughout this layer, the surface area of interest was the sum of plasma membranes of the exodermal short cells plus those of the first rank of central cortical cells.

Calculation of Percent of the Total Protein Particles That Would Need to be Ca²⁺-ATPase to Account for the Observed Flux of Ca²⁺ into the Stele

The number of Ca²⁺-ATPase molecules required was calculated from the Ca²⁺ flux into the stele (i.e. the number of Ca²⁺ ions moved into the stele s⁻¹) and the activity (turnover) of each Ca²⁺-ATPase molecule (i.e. the number of Ca²⁺ ions moved across a plasmalemma by an individual Ca²⁺-ATPase molecule s⁻¹). The percent of the total membrane protein particles required to be Ca²⁺-ATPase to account for the observed Ca²⁺ flux was calculated from the above figure and the total number of protein particles in the membrane area through which the transport was assumed to occur.