Cytosolic Concentrations and Transmembrane Fluxes of NH4 +/NH3. An Evaluation of Recent Proposals

Ammonium (NH4 +) is an important, perhaps underestimated, nitrogen source for plant growth (Bloom, 1988; Glass, 1988; Kronzucker et al., 1997). In many agricultural soils it is present in millimolar concentrations in soil solution (Wolt, 1994), whereas in mature forest soils NH4 +:NO3 −ratios as high as 56:1 have been reported (Stark and Hart, 1997). In many such soils, NH4 + may constitute the only inorganic nitrogen source available to plants (Van Cleve et al., 1993). Physiological studies have demonstrated that NH4 + fluxes commonly exceed those of NO3 − from equimolar solutions, and when both ions are present, NO3 − influx is reduced within minutes of exposure to NH4 + (Lee and Drew, 1989; Gazzarini et al., 1999; Kronzucker et al., 1999). As a byproduct of photorespiration and the shikimic acid pathway, as well as through the turnover of various nitrogen pools, large quantities of NH4 + are also generated internally (Joy, 1988; Feng et al., 1998). Although much of this NH4 + is re-assimilated through the activity of Gln synthetase (GS), many plants show high values of NH4 + efflux, and some develop symptoms of toxicity when exposed to millimolar concentrations of NH4 + in nutrient media, particularly when NH4 + is the sole source of nitrogen. The actual mechanism of NH4 + toxicity is still unclear, despite extensive studies (Gerendas et al., 1997). For these reasons it is important to characterize subcellular compartmentation and fluxes of NH4 +. Because of the potential toxicity of NH4 +/NH3, it is often assumed that plant cytosolic NH4 + concentrations ([NH4 +]c) are maintained at very low (sub-millimolar) levels, through the high activity and high affinity of GS for NH4 + (Pearson and Stewart, 1993;Gerendas et al., 1997). This is despite the fact that millimolar values for [NH4 +]chave been measured using NMR, efflux analysis, ion-specific microelectrodes, and tissue fractionation (Fentem et al., 1983; Lee and Ratcliffe, 1991; Wang et al., 1993a; Wells and Miller, 2000), and that large plasma membrane fluxes of NH4 + have been reported (Lee and Ayling, 1993; Wang et al., 1993b; Min et al., 1999). Nevertheless, in a recent review of NH4 + transport,Howitt and Udvardi (2000) conclude that some estimates of [NH4 +]c are probably orders of magnitude too high. First, they argue that “it is difficult to reconcile the estimates of cytosolic NH4 + concentrations made by this group (referring to the data of Wang et al., 1993a) with the high affinity of cytosolic GS for NH4 +(K m = 10–20 μm).” Second, they cite the findings of a single NMR study by Roberts and Pang (1992) who estimated [NH4 +]cto be 3 to 10 μm for external [NH4 +] ([NH4 +]o) up to 1 mm. Howitt and Udvardi (2000) go on to suggest that both influx and efflux of NH3 are mediated by passive diffusion through a reversible low-affinity transport system (LATS), previously considered to be an NH4 + transporter. Finally, they suggest that NH4 + entry to the vacuole is via passive permeation of NH3 and acid trapping of NH4 +. These speculations are at variance with a large body of evidence in the literature and therefore need to be critically examined. Howitt and Udvardi (2000) suggest that [NH4 +]cmust be considerably lower than the 3.6 mm estimated by Wang et al. (1993a) for plants grown at low external [NH4 +] because of the reported low K m values for NH4 + (10–20 μm) of GS. However, it must first be emphasized that K m values of enzymes determined in vitro might be very different from those operating in vivo because of regulation by allosteric effectors. Furthermore, the extensively studied glycolytic enzymes haveK m/[substrate]c ratios that vary from 0.02 to 333 (Fersht, 1985), and similar deviations ofK m/[substrate]c from unity can be found for the enzymes nitrate reductase and pyruvate kinase. In the first case,K m/[NO3 -]ccan be as low as 0.01 (Lee and Clarkson, 1986; Siddiqi et al., 1991;Miller and Smith, 1992; Kanayama et al., 1999), whereas in the second, the ratio can be as low as 0.02 (Memon et al., 1985a, 1985b; Walker et al., 1996). Furthermore, catalytic rates of enzymes that are subject to allosteric regulation are typically determined by the concentration of regulatory molecules rather than by substrate concentration. For example, the hexokinase of erythrocytes has a K m value of 0.1 mm for Glc, despite a cellular Glc concentration around 5 mm. Hexokinase activity is not regulated by the available Glc but as a result of allosteric inhibition from its product Glc-6-phosphate, with the result that large changes of Glc concentration hardly alter the rate of glycolysis (Fersht, 1985). GS is also known to be a highly regulated enzyme, located at a pivotal biochemical position to regulate plant nitrogen assimilation (Eisenberg et al., 2000), and available data indicate that [NH4 +]ccan vary considerably even in the same system under different conditions of [NH4 +]osupply (Lee and Ratcliffe, 1991; Roberts and Pang, 1992; Wang et al., 1993a; Kronzucker et al., 1995). The rationale that an enzyme'sK m should necessarily track ambient substrate concentration to optimize reaction rate, therefore, fails to apply to such enzymes. Roberts and Pang (1992), cited by Howitt and Udvardi (2000) in support of micromolar values of [NH4 +]c, incubated excised maize root tips in solutions containing from 0 to 10 mm NH4 +, and used NMR signals from 31P and13C to estimate cytosolic and vacuolar pH. [NH4 +]cwas then calculated from total root NH4 + to be from 2 to 438 μm (the upper limit being already ≥20 times higher than the reported GS K m for NH4 +), on the presumption that NH3 rapidly equilibrated across the tonoplast and was trapped as NH4 + according to cytosolic and vacuolar pH values. In contrast, Lee and Ratcliffe (1991)used a more direct 14N-NMR method to obtain [NH4 +]cof 3 to 8 mm, almost 3 orders of magnitude higher than the Roberts and Pang results, when comparing values obtained at the same [NH4 +]o. All other methods (tissue fractionation, efflux analysis, and NH4 +-specific microelectrodes) have yielded [NH4 +]cvalues in the millimolar range (Kronzucker et al., 1995, and references therein; Wells and Miller, 2000). Transport of NH4 + across the plasma membrane is biphasic (Ullrich et al., 1984; Wang et al., 1993b), corresponding to an active, saturable, high-affinity transport system and a passive, non-saturable LATS. Howitt and Udvardi (2000), however, propose that LATS transport occurs via a reversible NH3 transporter. In support of this proposal, the authors cite a study of 14C-methylamine (CH3NH2/CH3NH3 +) uptake by Phaseolus vulgaris leaves (Raven and Farquhar, 1981), but in fact this study showed that the relationship of influx to external pH and the equilibrium concentration of methylamine were “far higher than could be explained by the transport of CH3NH2 alone.” Raven and Farquhar concluded that methylamine uptake, at least at pH values below 7, was predominantly as CH3NH3 +, and driven by the membrane electrical potential difference (Δψ). A, Directions of passive NH4 + (dotted arrows) and NH3 (solid arrows) fluxes as functions of external [NH4 +], (all concentrations are in μmol L−1), calculated from the Nernst equation, with cytosolic [NH4 +] set at 10 μm and vacuolar [NH4 +] at 20,000 μm. External [NH4 +] was set at values from 10 to 10,000 μm. The pKa for dissociation of NH4 + was set at 9.32; external, cytosolic, and vacuolar pH values were set at 5, 7, and 5, respectively; plasma membrane and tonoplast ΔΨ were set at −100 and +10 mV, respectively. B, As in Figure 1A, except that cytosolic [NH4 +] is set at 10,000 μm. A further prediction arising from the suggestion that NH3 efflux occurs passively via LATS is that NH3 efflux should diminish as [NH4 +]o(and thus [NH3]o) increases. However, measurements of 13N tracer efflux have demonstrated that this flux increased from approximately 10% of influx to as high as 86% of influx as [NH4 +]ois increased from 10 or 100 μm to 1 or 1.5 mm(Wang et al., 1993a; Kronzucker et al., 1995; Min et al., 1999). Such findings are inconsistent with passive efflux of NH3. To evaluate the hypotheses of Howitt and Udvardi, we have constructed a model for the directions of passive fluxes of NH3 and NH4 + (Fig. 1, A and B), assuming [NH4 +]cof 10 μm (Howitt and Udvardi, 2000) or 10,000 μm (Wang et al., 1993a). In this model, [NH4 +]owas varied from 10 to 10,000 μm, pH values were set at 5, 7, and 5, respectively, for external solution, cytosol, and vacuole, Δψ values of −100 and +10 mV were selected for plasma membrane and tonoplast, respectively, (Wang et al., 1994), and [NH4 +]vwas set at 20,000 μm. This value was selected to represent a midpoint of a quite wide range of literature values (Lee and Ratcliffe, 1991; Roberts and Pang, 1992; Wang et al., 1993a; Wells and Miller, 2000). The quantitative analyses of the distribution of NH4 +/NH3, and directions of passive fluxes of these molecules that follow, are therefore based upon realistic values of pH, [NH4 +]o, ΔΨ, and [NH4 +]v. These values may vary somewhat under natural conditions but differences corresponding with orders of magnitude are unlikely. The model predicts that when [NH4 +]cis 10 μm (Fig. 1A), influx of NH4 + is passive until [NH4 +]ofalls below approximately 100 nm, a rare situation in most soils. Active influx of NH4 +would therefore be virtually unnecessary. This is a surprising conclusion given the abundant physiological and molecular evidence suggesting that the high-affinity transport system actively transports NH4 + at [NH4 +]oof up to 655 μm (Ullrich et al., 1984; Wang et al., 1994;Gazzarini et al., 1999). In contrast, when [NH4 +]cis 10,000 μm (Fig. 1B), active influx of NH4 + is required in the range of [NH4 +]o≤ 500 μm. This corresponds with the [NH4 +]oat which a break between high- and low-affinity influx systems (Ullrich et al., 1984; Wang et al., 1993b; Kronzucker et al., 1996) and changes in membrane depolarization are observed (Wang et al., 1994). Figure1A also indicates that if [NH4 +]cis 10 μm, passive NH3influx is not energetically feasible unless [NH4 +]ois ≥1 mm. It is entirely impossible if [NH4 +]cis 10 mm (Fig. 1B), because the gradient for NH3 is always from cytosol to cell wall. In this regard, it is instructive to examine the effects of pH on13NH3/13NH4 +influx. Published data reveal that low-affinity putative13NH4 +fluxes declined from 18.63 to 11.44 μmol g−1h−1 as external pH increased from 4.5 to 7.5 in rice roots (Wang et al., 1993b). Thus, despite a 1,000-fold increase of [13NH3]o, tracer influx actually declined by 40%; a similar decline of13NH4 +fluxes between pH 5 and 7, was observed by Kosegarten et al. (1997)using rice roots. The results argue against NH3influx at typical values of soil pH. This is not to suggest that NH3 influx may not occur as the pKa for NH3/NH4 +(9.32) is approached or exceeded. However, this situation is rare in most soils. With respect to vacuolar accumulation of NH4 +, Howitt and Udvardi (2000)propose that the flux of NH4 +from cytosol to vacuole must be active. This is confirmed in our model for conditions where [NH4 +]cis 10 μm, or when [NH4 +]cis 10 mm and vacuolar [NH4 +] ([NH4 +]v) exceeds 5 mm. However, the authors consider it more likely that passive transport of NH3 across the tonoplast, and acid trapping in the vacuole, provide the mechanisms for vacuolar NH4 + accumulation. This situation might apply when [NH4 +]vis ≤1 mm, corresponding to an [NH3]v ≤ 0.048 μm. Above this value, however, as shown in Figure 1A, the gradient for NH3 permeation across the tonoplast is in the opposite direction (from vacuole to cytosol), requiring active transport of NH3 to the vacuole. The data reported by Lee and Ratcliffe (1991), Roberts and Pang (1992), and Wang et al. (1993a) provide [NH4 +]vvalues that are >1 mm. This leads to the unlikely scenario whereby transport of NH3 to the vacuole requires an active NH3 flux, not a passive flux as they propose. Together with a passive leak of NH3 in the opposite direction, this scenario would result in a futile cycling of NH3 between cytosol and vacuole. In contrast, when [NH4 +]c= 10,000 μm (Fig. 1B), the flux of NH3 to the vacuole will always be passive, and remobilization of NH4 + to the cytosol is also passive. [NH4 +]cvalues in the millimolar range are confirmed by four different methodologies. The single literature value for [NH4 +]cthat conforms to a micromolar value (Roberts and Pang, 1992) is predicted from measurements of cellular pH and total-tissue [NH4 +] in roots, rather than being based upon direct measurements of [NH4 +]c. Conjectures regarding cytosolic [ion] based upon in vitro measurements of enzyme K m values are untenable. If [NH4 +]cwere 10 μm, then (a) Active transport of NH4 + would be unnecessary, unless [NH4 +]ofalls below 100 nm; (b) Passive influx of NH3 is feasible at [NH4 +]o≥ 1000 μm, but the pH profile for13N tracer influx appears to contradict this possibility; (c) Passive efflux of NH3 is thermodynamically feasible, but appears to be contradicted by the observation that 13N efflux increases as [NH4 +]o(and hence [NH3]o) is increased; and (d) If [NH4 +]c is 10 μm, NH3 can only traverse the tonoplast passively as long as [NH4 +]v≤ 1 mm, an unlikely situation when NH4 + is available in external media.

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.