Na,K-ATPase in the nuclear envelope regulates Na+: K+ gradients in hepatocyte nuclei

Evidence is emerging that the nuclear envelope itself is responsible for transport and signaling activities quite distinct from those associated with the nuclear pore. For example, the envelope has a Ca2+-signaling pathway that, among other things, regulates meiosis in oocytes. The nuclear envelope's outer membrane also contains K+ channels. Here we show that Na+/K+ gradients exist between the nuclear envelope lumen and both cytoplasm and nucleoplasm in hepatocyte nuclei. The gradients are formed by Na,K-ATPases in the envelope's inner membrane, oriented with the ATP hydrolysis site in the nucleoplasm. We further demonstrate nucleoplasm/cytoplasm Na+ and K+ gradients, of which only the Na+ gradient is dissipated directly by Na,K-ATPase inhibition with ouabain. Finally, our results demonstrate that nuclear pores are not freely permeable to sodium and potassium. Based on these results and numerous in vitro studies, nuclear monovalent cation transporters and channels are likely to play a role in modulation of chromatin structure and gene expression.

Lens epithelium and fiber Na,K-ATPases: distribution and localization by immunocytochemistry

PURPOSE

To use immunofluorescence and immunogold techniques to identify the catalytic subunits of the Na,K-ATPases of the lens and to determine their location in the cells of the epithelium and cortex of bovine and human lenses.

METHODS

Frozen sections of capsulated and decapsulated bovine and human lenses were prepared, blocked, and treated with affinity-purified polyclonal rabbit antibodies to the Na,K-ATPase catalytic subunit isoforms with subsequent treatment with fluorescein isothiocyanate-labeled goat anti-rabbit IgG and visualization of the fluorescence by light microscopy. An immunogold-labeled goat anti-rabbit IgG was used to detect, by electron microscopy, the binding of the same affinity-purified polyclonal antibodies to thin sections of decapsulated lenses that had been fixed and embedded in Lowicryl K4M. The results were confirmed by staining of western blot analysis of sodium dodecyl sulfate-polyacrylamide gel separations of enriched membrane preparations from bovine and human lenses.

RESULTS

The three common catalytic subunits of the Na,K-ATPases are present in the plasma membranes of lens epithelium, lens fibers, or both. The data indicate a polarized distribution of the alpha1 and alpha3 catalytic subunit isoforms in central epithelium. In the cortical fibers, the alpha2 isoform is present around the interdigitations. The alpha3 isoform is found in the interdigitation-free regions of human cortical fibers.

CONCLUSIONS

This unique distribution of Na,K-ATPases precludes the popular pump-leak model for lens monovalent cation homeostasis. The functional significance of the distribution of Na,K-ATPases in the lens epithelium and superficial fibers is currently under investigation.

Changes in Na,K-adenosine triphosphatase (ATPase) concentration and Na,K-ATPase-dependent adenosine triphosphate turnover in human erythrocytes in diabetes

The concentration of Na,K-adenosine triphosphatase (ATPase) and Na,K-ATPase-dependent adenosine triphosphate (ATP) turnover was measured in fasting blood samples of 20 subjects with insulin-dependent diabetes mellitus (IDDM), 22 subjects with non-insulin-dependent diabetes mellitus (NIDDM), and 20 nondiabetic subjects. [3H]ouabain binding was used to determine Na,K-ATPase concentration. There were 471 +/- 70 (mean +/- SD) ouabain binding sites per erythrocyte, normally distributed in the nondiabetic subjects. The number of ouabain sites per cell was lognormally distributed in the two populations of diabetic subjects. The mean of lognormal distributions of ouabain sites per cell was significantly lower in the IDDM group. The mean of the lognormal distribution for the NIDDM group was not significantly different from that of the nondiabetic subjects. Na,K-ATPase-dependent ATP turnover (molar activity) was 9,580 +/- 742 mol/mol minute (mean +/- SD) normally distributed in the nondiabetic population. A lognormal distribution was observed in the diabetic population. Means of the lognormal distributions were significantly different: 3.98 +/- 0.05 for the nondiabetic population and 3.13 +/- 0.48 for both diabetic populations. Changes in the concentration of Na,K-ATPase (ouabain sites per cell) and Na,K-ATPase-dependent ATP turnover did not correlate with hemoglobin A1C (HbA1C) or with blood glucose. This would suggest that elevated glucose concentrations do not directly cause decreased Na,K-ATPase function in the diabetic erythrocyte.

Nondestructive measurement of retinal glucose transport and consumption in vivo using NMR spectroscopy

The cellular events underlying various retinopathies are poorly understood but likely involve perturbation of retinal glucose metabolism. Current methods for assessing this metabolism are destructive, thus limiting longitudinal studies. We hypothesize that following an intravitreous injection, the clearance rate of a glucose analogue will be a nondestructive index of retinal glucose transport and metabolism in vivo. First, radiolabeled glucose analogues were injected into the vitreous. After 40 min, the dominant clearance path was posterior via the retina and was consistent with a facilitated transport mechanism. Next, either [6,6-2H2]glucose or 3-deoxy-3-fluoro-D-glucose was injected into the vitreous of rabbit eyes, and the clearance rate of each analogue was determined over 40 min using, respectively, 2H or 19F NMR. These rates were interpreted as a function of the retinal glucose transport and consumption. From the NMR data, the rate of retinal glucose consumption was approximately 16 times slower than the transport of glucose. These data demonstrate that NMR measurements of glucose analogue clearance rate from the vitreous can provide a nondestructive index of retinal glucose transport and consumption in vivo.

Na,K-ATPases of the lens epithelium and fiber cell: formation of catalytic cycle intermediates and Na+: K+ exchange

Previous studies suggest that an alpha 2-related isoform of the catalytic subunit is predominant in the lens fiber cells. The alpha 1 isoform is predominant in the lens epithelium (Garner, Horwitz and Enomoto, 1992). Data are presented to show that strophanthidin-sensitive K+ transport is sustained by both of these lens Na,K-ATPases. The K50 for strophanthidin inhibition of K+ transport was 1.4 +/- 0.5 x 10(-6) M for the alpha 1 isoform of the epithelium, 1.3 +/- 0.6 x 10(-7) M for the alpha 2-related isoform of the lens fiber cells. Na+ accumulation in cultured bovine lenses was strophanthidin sensitive. The K50 values for strophanthidin-sensitive Na+ accumulation were similar to those obtained for K+ transport. ARP turnover by the lens fiber cell Na,K-ATPase (1700 +/- 600 min-1) was lower than ATP turnover by the lens epithelium Na,K-ATPase (8000 +/0 1000 min-1). Both lens Na,K-ATPases form the (ouabain + Mg(2+) + phosphate)-dependent phosphoenzyme. Both lens Na,K-ATPases form the (ATP + Na(+) + Mg2+)-dependent phosphorylated intermediate. K+ does not effectively dephosphorylate the Na,K-ATPase of the lens fibers. K+ does cause dephosphorylation of the Na,K-ATPase of the lens epithelium. Interaction of the Na,K-ATPase with Mg2+ would appear to cause the monovalent cation insensitivity. The lens fiber cell Na,K-ATPase, like the lens epithelium Na,K-ATPase occludes two K+ ions. However, between the two major Na,K-ATPases of the lens, there would appear to be differences in the ATP dissolution of the K-occluded state.

Catalytic subunit isoforms of mammalian lens Na,K-ATPase

To identify the Na,K-ATPase isoforms present in the mammalian lens, seven antisera were prepared to selected peptide sequences of the catalytic (alpha) subunit. Three antisera were prepared to peptide sequences at the N-terminus of the three sequenced rat alpha isoforms. There is < 53% sequence homology among the isoforms in this region. Three antisera were prepared to peptide sequences at the ouabain binding site in the extracellular loop between membrane spanning sequences 1 and 2 of the sequenced rat alpha isoforms; sequence homology among the isoforms in this region is < 69%. An antiserum was also prepared to the carboxyl terminal region of the alpha 2 rat isoform. The sequenced isoforms (rat and human) in this region are > 94% homologous. The results from stains of Western blots of SDS-PAGE separations of lens membranes are presented. Alpha 1 is the predominant isoform of the epithelium. It is not found in cells of the central epithelium but is present in cells located more toward the equator. Alpha 3 is the catalytic subunit of the central 43% of the epithelium. The lens fiber cell membranes have a catalytic subunit that is related to the alpha 2 isoform. In the fiber cell a 98-100 kDa band stains with the antiserum to the alpha 2 N-terminus and the antiserum to the alpha 2 ouabain site. The antiserum to the alpha 2 C-terminus does not stain the 98-100 kDa band. (Preliminary reports of these results were presented at the 1992 and 1993 meetings of the Association for Research in Vision and Ophthalmology).

Cations, oxidants, light as causative agents in senile cataracts

Lens transparency is a function of regular cell shape, regular cell volume, minimal extracellular space, and minimal scatter elements. The cellular structure and molecular structure of the lens is reviewed. The importance of the cytoarchitecture especially the sutures, is discussed. The high cholesterol: phospholipid ratio of the lens fiber cell membranes is related to the functions of low permeability, low fluidity, and mechanical stability. Also reviewed are the contributions of the lens crystallins to lens clarity and to lens refractive index. The importance of intracellular and extracellular cation and water concentrations are reviewed. Finally the effects of systemic diseases, oxidation, and light on lens clarity are discussed relative to changes in lens fiber cell cation concentrations.

Na,K-ATPase and phospholipid degradation in bovine and human lenses

Na,K-ATPase, an enzyme intrinsic to the membrane of most cells, is inhibited in cataract. Na,K-ATPase, activity in clear non-cataractous lenses is found predominantly in the lens epithelium. The lens fiber cells would appear to be unique, among mammalian cells in that Na,K-ATPase activity is low if not absent. The study presented here indicates that Na,K-ATPase is present, often in high concentration, but progressively more functionally compromised as the fiber cells mature. The membrane lipid environment as causative agent in the loss of normal function of Na,K-ATPase, is considered in this study. The data indicate a correlation between increasing cholesterol/phospholipid ratio, increasing phospholipase A2 activity and decreasing Na,K-ATPase activity in normal clear lenses. The phospholipase A2 activity is higher in cortex and nucleus than in the epithelium of normal bovine and human lenses. The phospholipase A2 is Ca2+ dependent and may be membrane associated.

Na,K-ATPase of cultured bovine lens epithelial cells: H2O2 effects

Na,K-ATPase function was studied in cultured bovine lens epithelial cells under confluent and non-confluent conditions. The affinity of the Na,K-ATPase for the cardiac glycoside, ouabain, differs between the confluent and non-confluent cultures. The confluent cells have a higher affinity for ouabain than do the non-confluent cells. The ouabain affinity of the confluent cells is similar to that for the Na,K-ATPase isolated from the bovine axolemma and the bovine lens cortex. The ouabain affinity of the non-confluent cells is similar to that for the Na,K-ATPase of the renal medulla and bovine lens epithelium. Similar results are not found with confluent and non-confluent MDCK cells. H2O2 treatment of confluent and non-confluent lens epithelial cell cultures has differing effects on the Na,K-ATPase function. In the confluent cell preparations, H2O2 affects K(+)-dependent dephosphorylation of the intermediate phosphoenzyme. In the non-confluent preparations. H2O2 appears to inhibit K(+)-occlusion.

Na(+)-K(+)-ATPase and changes in ATP hydrolysis, monovalent cation affinity, and K+ occlusion in diabetic and galactosemic rats

This study showed that steady-state kinetics of ATP hydrolysis by Na(+)-K(+)-ATPase are altered in the BB Wistar diabetic rat and experimental galactosemia. Four days after onset, this change was not evident if NaCNBH3 was omitted during enzyme preparations (indicating reversibility). Ninety days after onset, NaCNBH3 reduction was not necessary to see the change in ATP hydrolysis kinetics (indicating nonreversibility). The change in steady-state ATP hydrolysis was similar to that reported earlier for Na(+)-K(+)-ATPase of the lens epithelium and kidney medulla of diabetic individuals and for two in vitro glycosylation models. Our study also showed that the affinities of Na(+)-K(+)-ATPase for K+ are altered, and Na(+)-K(+)-ATPase-dependent K+ occlusion is inhibited in diabetic and galactosemic animals. Because K+ occlusion is required for efficient K+ transport, this finding supports previous in vitro studies that indicated that glycosylation inhibits pump-dependent K+ transport. Furthermore, our study suggested an irreversible impairment of Na(+)-K(+)-ATPase function in the diabetic BB Wistar rat as early as 15 days after onset, even when blood glucose was maintained at 6.7 mM by daily insulin injection.

Nonenzymatic glycation of Na,K-ATPase. Effects on ATP hydrolysis and K+ occlusion

Glycation of the Na,K-ATPase in vitro (formation of Schiff base with glucose followed by reduction with NaCNBH3) shows the presence of three classes of reactive amino groups that differentially affect catalysis and cation binding. Reaction in the absence of ATP results in irreversible inhibition of enzyme activity with a t1/2 of 53 min. This is due to modification of one class of amino groups that affect the catalytic domain of the enzyme. In the presence of ATP, glycation first results in a shift in the steady state kinetics of ATP hydrolysis from substrate activation to Michaelis-Menten kinetics accompanied by an increase in the apparent affinity for K+ in the p-nitrophenylphosphatase reaction. This change in kinetic properties occurs with a t1/2 of 9 min and results in the complete loss of K+ occlusion. Incorporation of glucose is into the catalytic subunit, remote from the N-terminal end. Apparent total inhibition of K+ occlusion occurs with a stoichiometry 0.8 mol of glucose incorporated per mol of enzyme. Therefore, there is a rapidly reacting amino group that affects the cation binding domain of the Na,K-ATPase. More slowly, with a t1/2 of 9 h, the ATP hydrolysis kinetics change from Michaelis-Menten to substrate inhibition without recovery of K+ occlusion, showing that, in the E1 conformation, there is a third, slower reacting class of amino groups in the Na,K-ATPase that affects low affinity nucleotide interaction with the catalytic subunit.

Induction of de novo synthesis of crystalline lenses in aphakic rabbits

The mammalian lens, like other ectodermal tissues, can regenerate itself given the proper environment. Endocapsular phacoemulsification of adult rabbit lenses was performed. A lens capsular bag with posterior and anterior lens capsule relatively intact was left in the eye. Regrowth of material in the capsular bag was followed by slit lamp biomicroscopy and photography over a 12-month period. Histopathology of the new material showed regions of relatively normal epithelial cells and lens fibers as well as regions where growth was irregular. All major lens crystallin classes were present in the regenerated lens. Several specific crystallin subunits, known to arise by post-translational modification of primary gene products, were absent or present in abnormally low concentrations.

Stimulation of glucosylated lens epithelial Na,K-ATPase by an aldose reductase inhibitor

In diabetes, glucosylation of the Na,K-ATPase of the lens epithelium makes the pump inefficient. K+ transport and ATP hydrolysis (at near saturating ATP concentrations) are inhibited and the kinetics of ATP hydrolysis become substrate inhibition type. The AR inhibitor (AL1576, Alcon Laboratories) stimulates K+ transport and ATP hydrolysis by glucosylated bovine lens Na,K-ATPase. This inhibitor has a slight stimulatory effect upon the unmodified enzyme function also. The AR inhibitor is not able to prevent glucosylation of the pump in high-glucose-containing medium.

Direct stimulation of Na+-K+-ATPase and its glucosylated derivative by aldose reductase inhibitor

In the presence of 10(-8) M concentrations of the aldose reductase inhibitor AL 1576, there is a 20-30% increase in the rate of hydrolysis of near-saturating concentrations of ATP by bovine renal Na+-K+-ATPase. When bovine renal Na+-K+-ATPase is reacted with glucose 6-phosphate in the presence of 10(-8) M concentrations of AL 1576 or 10(-6) M concentrations of a second aldose reductase inhibitor, sorbinil, glucosylation occurs. Whereas sorbinil has no effect on ATP hydrolysis by the glucosylated Na+-K+-ATPase, 10(-8) M AL 1576 causes a shift in the kinetics of hydrolysis of ATP from substrate inhibition to normal substrate activation. The aldose reductase inhibitors interact with the enzyme at the low-affinity ATP-binding site.

ATP hydrolysis kinetics of Na,K-ATPase in cataract

The steady-state kinetics of hydrolysis of Mg2+ ATP by the epithelial Na,K-ATPase of individual human lenses were determined. Among the cataract lens population, four distinct kinetic types were observed: negative kinetic co-operativity. Michaelis-Menten kinetics, positive kinetic co-operativity, and substrate inhibition kinetics. Negative kinetic co-operativity and Michaelis-Menten kinetics were also observed in a group of presumably clear lenses from non-diabetic individuals ages 16-42 years. Substrate inhibition kinetics were found to be prevalent in individuals with mature onset diabetes. Substrate inhibition kinetics were also observed for Na,K-ATPase isolated from lenses which had been incubated in high glucose. It would appear that this modification leads to an inhibition of Na,K-ATPase-dependent K+ influx into these cultured lenses.

H2O2-modification of Na,K-ATPase. Alterations in external Na+ and K+ stimulation of K+ influx

Studies, at steady state, of the Na,K-ATPase dependent influx of K+ into bovine lenses in organ culture are used to characterize further the H2O2-modification of the Na+ pump. Control lenses display constants for interaction with external Na+ and K+ similar to those obtained for the erythrocyte. H2O2 treatment of the bovine lens leads to total loss of external Na+ stimulation and alteration of external K+ stimulation.

Glucose-6-phosphate modification of bovine renal Na,K-ATPase: a model for changes occurring in the human renal medulla in diabetes

The kinetics of hydrolysis of ATP were determined for the renal Na,K-ATPase, in the K+ conformation, modified with glucose-6-phosphate. There was a shift in the ATP hydrolysis kinetics from negative kinetic co-operativity for the control enzyme preparations to substrate inhibition kinetics for the modified enzyme preparations. The effect was reversible and stabilized after NaBH4 reduction. Approximately 4 moles of glucose-6-phosphate were incorporated per mole of Na,K-ATPase (based on MW of 150,000 daltons). Similar substrate inhibition kinetics were observed for the renal Na,K-ATPase isolated from several human subjects with mature onset diabetes.

Kinetic cooperativity change after H2O2 modification of (Na,K)-ATPase

The kinetics of hydrolysis of ATP and p-nitrophenylphosphate and the action of the allosteric effectors, Na+ and K+, upon the hydrolysis of these substrates were used to study the H2O2-modified, uncoupled (Na,K)-ATPase isolated from cultured bovine lenses ( Garner , W. H., Garner , M. H., and Spector , A. (1983) Proc. Natl. Acad. Sci. U.S.A. 80, 2044-2048). Pure bovine renal (Na,K)-ATPase was modified by H2O2 in 150 mM KCl and 20 mM MgCl2 to yield an enzyme with kinetic properties similar to the enzyme isolated from the H2O2-treated, cultured bovine lens. H2O2 modification changes the interaction of the ATP hydrolysis site from negative to positive kinetic cooperativity. H2O2 modification dramatically alters Na+ stimulation of ATP hydrolysis and Na+ inhibition of p-nitrophenylphosphate hydrolysis while having little effect upon K+ control of the hydrolysis of these two substrates.

H2O2-induced uncoupling of bovine lens Na+,K+-ATPase

A 1-hr exposure of bovine lenses in organ culture to H2O2 concentrations in the range found in the aqueous fluid of patients with cataracts inhibits 86Rb+ influx. At 1 mM H2O2, complete inhibition was observed and further investigated. Membrane permeability is slightly decreased. Although lactate concentrations increase 2-fold, lens ATP concentrations decrease approximately equal to 10%, suggesting that glycolysis may be stimulated but ATP production is not able to keep up with the demand for energy. Examination of epithelial cell Mg2+-stimulated Na+,K+-ATPase isolated from the cultured lenses indicates H2O2-induced modification. At 5 mM MgATP, ATP hydrolysis is accelerated 30%; at 3 mM MgATP, hydrolysis is normal; and at 0.75 mM MgATP, it is inhibited 75%. p-Nitrophenyl phosphate hydrolysis and eosin maleimide binding indicate that K+ control of the enzyme is modified. Thus, a very early effect of H2O2 upon the lens, well before the formation of opacity, appears to be the uncoupling of Na+ and K+ transport from ATP hydrolysis.

Biochemical evidence for membrane disintegration in human cataracts

Biochemical evidence is presented for the disintegration of the lens fiber plasma membrane in human cataracts. The intrinsic membrane proteins are found in both the water-soluble and water-insoluble nonmembrane fractions of the cataract lens but not in the normal tissue. Furthermore, in contrast to the normal lens, not all of the lipid found in the cataractous lens is isolated with the membrane fraction. In cataracts, both the membrane and membrane fragments are involved in covalent high molecular weight aggregates with an extrinsic membrane protein (43,000 daltons) and a cytoplasmic protein (gamma-crystallin).

Selective oxidation of cysteine and methionine in normal and senile cataractous lenses

The oxidation state of methionine and cysteine in normal and cataractous lenses is reported. In young lenses no oxidation was detected in any protein fraction examined. Only the intrinsic membrane fraction and membrane-related components showed evidence of oxidation in old (60-65 years of age) normal lenses. However, in a similar age group, with the development of cataract, progressive, dramatic changes were observed. With severe cataracts, 60% or more of the methionine in membrane-associated components was found in the methionine sulfoxide form, and methionine sulfone was observed in one case. Most of the cysteine was found oxidized to either the disulfide form or putative cysteic acid. Mixed disulfides with glutathione were observed. Oxidative changes in soluble components as illustrated by alpha-crystallin occurred more gradually. The data clearly support the viewpoint that extensive oxidation of lens proteins occurs with cataract and that it begins at the lens fiber membrane.

An extrinsic membrane polypeptide associated with high-molecular-weight protein aggregates in human cataract

A 43,000-dalton polypeptide has been isolated from the high-molecular-weight disulfide-rich fraction of the water-insoluble protein of human cataractous lenses. On the basis of immunochemical reactivity and fluorescent antibody binding, this polypeptide is localized in the membrane region of the lens cell. This observation suggests an interaction between the soluble lens proteins and membrane-associated polypeptides in the formation of large protein aggregates which may cause cataract.

Complete amino acid sequence of myoglobin from the pilot whale, Globicephala melaena

The complete amino acid sequence of the major component myoglobin from the pilot whale, Globicephala melaena, was determined by specific cleavage of the protein to obtain large peptides which are readily degraded by the automatic sequencer. The apomyoglobin was selectively cleaved at the two methionyl residues with cyanogen bromide and the acetimidated apomyoglobin was cleaved at the three arginyl residues by trypsin. From the sequence analysis of four of these peptides and the apoprotein, over 90% of the covalent structure of the protein was obtained. The remainder of the primary structure was determined by sequence analysis of three of the tryptic peptides isolated from the central cyanogen bromide fragment after modification of its single arginyl residue with 1,2-cyclohexanedione. This myoglobin differs from that of the Black Sea dolphin at four positions and from the myoglobin of the killer whale, Pacific common dolphin, and Atlantic bottlenosed dolphin at two positions. The above differences reflect the close taxonomic relationship of these five species of Cetacea. This sequence determination was aided by the use of a Texas Instruments 980A minicomputer system which performed peak integrations for all samples subjected to amino acid analysis.

Determination of the pK values for the alpha-amino groups of human hemoglobin

The rate of reaction between alpha-amino groups and cyanic acid was followed at 26 degrees and ionic strength 0.2 M as a function of pH of human hemoglobin Ao solutions to determine the pK and the pH-independent second order rate constant, kappa, for these groups in the alpha and beta chains. At a given point in time, the extent of the reaction was determined by employing the Beckmann Sequencer as a quantitative tool in which the yields of leucine and histidine in the second Edman degradation cycle were used to define the rates of reaction of the alpha and beta chains, respectively. From these results, the individual were evaluated (Garner, M.H., Garner, W.H., and Gurd, F. R.N. (1973) J. Biol. Chem. 248, 5451-5455). Values for pK for the alpha and beta chains were, respectively, 6.74 and 6.93 for cyanoferrihemoglobin, 6.95 and 7.05 for carboxyhemoglobin, and 7.79 and 6.84 for deoxyhemoglobin. Values for kappa, M- minus 1 S-minus 1, for the alpha and beta chains were, respectively, 12.5 and 17 for cyanoferrihemoglobin, 12 and 18 for carboxyhemoglobin, and 91 and 24 for deoxyhemoglobin. Limits of significance were estimated for both variables in each case. The pK results for valine 1alpha agree well with the value obtained by Hill and Davis (1967) J. Biol. Chem. 242, 2005-2012) for carboxyhemoglobin and with that of Kilmartin and Rossi-Bernardi ((1971) Biochem. J. 124, 31-45) for deoxyhemoglobin. Values obtained for sperm whale myoglobin were 7.77 for pK and 7.4 for kappa. The results are useful for the interpretation of the allosteric interactions of hemoglobin with hydrogen ions, with CO2, and with phosphate.