INTRODUCTION

Top Abstract Introduction Methods Results Discussion References

EXPERIMENTAL EVIDENCE linking an endogenous digitalis-like Na⁺-K⁺-ATPase inhibitor to fluid and electrolyte homeostasis through regulation of renal sodium excretion and, in the case of dysregulated states, to the pathogenesis of a prevalent human disease, hypertension, stimulated efforts to isolate and characterize such a compound (4, 17). Until now, the only known specific regulators of mammalian Na⁺-K⁺-ATPase were plant-derived digitalis glycosides and bufodienolides found in amphibian species. Recently, a novel compound was isolated from bovine hypothalamus and structurally characterized as an isomer of the plant glycoside, ouabain (31). This hypothalamic inhibitory factor (HIF) had been shown to have biological and biochemical properties similar to but not identical to those of ouabain (21). Thus HIF is a potent inhibitor of the Na⁺ pump in renal tubular cells (7), has positive inotropic activity in cardiac myocytes (18), and has vasoconstrictive properties in isolated blood vessels (25), all consistent with the proposed role for regulation of body fluid volume status and cardiovascular physiology.

Important differences in biological activity from plant ouabain have also been shown. Thus there were differences in binding and dissociation kinetics in the renal cells (7); inotropic activity in cardiac myocytes occurred at concentrations approximately three orders of magnitude less than for ouabain, with evidence for less toxicity in the cardiac cells (18); and HIF produced potent biological activity in tissues containing isoforms of the Na⁺-K⁺-ATPase that are highly resistant to ouabain (21). Like ouabain, HIF is an -L-rhamnoside, but it differs in its steroid portion as a regio- or stereochemical isomer of the plant compound, and this structural distinction is presumed to account for the observed differences in biological properties (31).

Most recently, Zhao and co-workers (33) demonstrated that an Na⁺-K⁺-ATPase inhibitor purified from human plasma, concluded by Hamlyn and co-workers (19) to be indistinguishable from ouabain, is in fact structurally different from ouabain but identical to HIF, supporting the notion that the human sodium pump may be under specific physiological regulation by this mammalian analog of the digitalis glycosides (33).

Nothing is yet known about the biosynthetic pathway of HIF, and little is known about specific physiological stimuli that might lead to its production or cellular release. Crabos and co-workers (10) did, however, demonstrate that HIF was released by rat midbrain slices in vitro and that this release was inhibited by atrial natriuretic peptide. Some recent clinical studies have raised the possibility that the appearance of digitalis-like activity in human serum may be related to low oxygen tension. Thus increased levels of a digoxin-like immunoreactive factor and/or Na⁺-K⁺-ATPase inhibitory activity have been reported in the plasma of normal individuals at high altitude (12), in patients with chronic respiratory failure (15, 32), and in normal subjects undergoing voluntary hypoventilation (2). Furthermore, an altered response to O₂ administration in chronically hypoxic patients has been reported (11).

The aim of the current study was to investigate the effect of hypoxia on the release of endogenous Na⁺-K⁺-ATPase inhibitory activity from rat brain and adrenal tissues. We report here a remarkable increase in release from both tissues of an inhibitor that coelutes with purified bovine brain HIF in response to hypoxic challenge.

	METHODS

Top Abstract Introduction Methods Results Discussion References

Animals. Two groups of Wistar rats (250-350 g) were kept for 3 days in different experimental conditions. The first group was housed in a regular cage at room atmosphere; the second was kept in a special sealed cage under an artificial air, poor in oxygen (10%). The air composition in the latter was checked every day, and the range of oxygen levels was between 8.8 and 10.3% (CO₂ range, 0.09-0.19%; and N₂ range, 88.6-90.1%). Both groups of rats were maintained on the same diet (Rodent Laboratory Chow 5001; Purina Mills) and allowed free access to water.

Tissue incubations. Rats were anesthetized using pentobarbital sodium (6.5 mg/100 g). Blood samples were collected from each rat in tubes containing sodium heparin, and adrenals and brain were removed and washed twice in a chilled buffer solution [containing, in mM, 10 N-2-hydroxyethylpiperazine-N'-2-ethanesulfonic acid (HEPES), 114 NaCl, 5 KCl, 1.15 MgSO₄, 10 D-glucose, 25 NaHCO₃, 2.5 CaCl₂, and 1 NaH₂PO₄, pH 7.4].

Each brain was divided into two halves, the cortex was removed, and the midbrain was dissected free. Slices of midbrain tissue of each half (average weight 22.7 mg) were placed in a culture plate insert (Millicell CM, Millipore) and incubated (37°C) in the HEPES buffer medium (1,250 µl) for 3 h in two different incubators, one with a gas mixture of 5% CO₂-95% O₂ and the second with 5% CO₂-4% O₂-91% N₂ (Fig. 1). After 3 h of incubation, supernatants were collected, and fresh medium (1,180 µl) plus 70 µl of 3 M NaCl (final concentration 300 mM) were added. We performed the NaCl incubation to verify whether high NaCl concentration, a known stimulus for ouabain-like substance release in central nervous system following intrathecal administration in vivo (23, 24), would be an effective stimulus under our in vitro experimental conditions. Further release of HIF by NaCl stimulation could thus serve as a control for tissue viability subsequent to the initial low O₂ exposure. After 1 h of incubation in these conditions the supernatants were collected. All supernatants were acidified with 10 µl of 12 N acetic acid and heated to boiling to precipitate proteins. Samples were centrifuged at 16,000 g for 15 min, and the supernatants were collected and neutralized with 5 N NaOH to pH 7.4. The same procedure was followed for slices from adrenals. Average weight of adrenal tissue per Millicell CM plate was 5.37 mg. Peripheral blood was collected and centrifuged at 2,000 g for 15 min, and the plasma was separated, acidified, boiled, centrifuged, and neutralized to pH 7.4, as for the other samples.

View larger version (21K): [in this window] [in a new window]   Fig. 1.   Schematic representation of study protocol. From each rat, kept both in regular cages (21% O2, left) and in hypoxic sealed cages (10% O2, right), brain was removed, and half of each midbrain was subsequently incubated in vitro at either 95% or 4% O2. Supernatants were collected, and new medium rich in NaCl was added to study further release of Na+-K+-ATPase inhibitor. The same protocol was applied to adrenal tissue.

In all, there were 15 samples coming from each of 8 different conditions of midbrain incubations, and there were 3 samples from each of 8 adrenal incubations. The experimental conditions can be summarized as follows (see Fig. 1 for schematic): group 1, 15 supernatants from midbrain tissue of rats kept in regular cages with 21% O₂ and subsequently incubated in vitro with 95% O₂; group 2, 15 supernatants from midbrain tissue of rats kept in regular cages with 21% O₂ and incubated with 4% O₂; group 3, 15 supernatants from midbrain tissue of rats kept in special cages with 10% O₂ and incubated with 95% O₂; group 4, 15 supernatants from midbrain tissue of rats kept in special cages with 10% O₂ and incubated with 4% O₂; groups 5-8, 15 supernatants from midbrain tissue-treated as groups 1-4, respectively, but after the incubation with 300 mM NaCl; group 1A, 3 supernatants from adrenal tissue of rats kept in regular cages with 21% O₂ and incubated with 95% O₂; group 2A, 3 supernatants from adrenal tissue of rats kept in regular cages with 21% O₂ and incubated with 4% O₂; group 3A, 3 supernatants from adrenal tissue of rats kept in special cages with 10% O₂ and incubated with 95% O₂; group 4A, 3 supernatants from adrenal tissue of rats kept in special cages with 10% O₂ and incubated with 4% O₂; and groups 5A-8A, 3 supernatants from adrenal tissue corresponding to above conditions of groups 1A-4A, respectively, but following the subsequent incubation with 300 mM NaCl.

To purify and concentrate activity and determine retention time of inhibitory activity in comparison with HIF derived from bovine hypothalamus, supernatants from each tissue group and respective incubation condition were combined and chromatographed using CHP-20P resin in a reverse-phase technique. Plasma samples were chromatographed in the same manner.

CHP-20P chromatography. The method was a cornerstone in the complete purification of HIF from bovine hypothalamus as previously described (31). Briefly, 30 ml of CHP-20P resin (Mitsubishi) was activated by using 10 column volumes of high-performance liquid chromatography (HPLC)-grade methanol followed by 15 column volumes of water. The aqueous sample was loaded, and the column was eluted at a flow rate of 2 ml/min for 120 min, developed as a linear gradient of 0-100% HPLC grade methanol. Thirty 8-ml fractions were collected, and aliquots from each fraction were dried, reconstituted in buffer, and assayed for Na⁺-K⁺-ATPase inhibitory activity as ⁸⁶Rb⁺ uptake into human erythrocytes by a slight modification of the method previously described (8).

⁸⁶Rubidium uptake assay. Sodium pump activity was estimated as ouabain-sensitive ⁸⁶Rb⁺ uptake into human red blood cells from healthy donors. Briefly, erythrocytes were washed and suspended to 50% cells in a HEPES buffer, pH 7.4 (containing, in mM, 20 HEPES, 1 CaCl₂, 1 MgSO₄, 5 NaH₂PO₄, 138 NaCl, and 11 glucose). One-milliliter aliquots of each chromatographic fraction were dried and reconstituted with 50 µl of buffer and incubated at 37°C with 50 µl of erythrocyte suspension and 4 µCi/ml ⁸⁶RbCl for 90 min. The reaction was quenched with 750 µl of cold HEPES buffer (4°C). Unbound counts were separated from cells by spinning through a cushion of silicon/phthalate oil (1:1), and erythrocyte ⁸⁶Rb⁺ was determined by gamma emission.

	RESULTS

Top Abstract Introduction Methods Results Discussion References

Chromatographic concentration of Na⁺-K⁺-ATPase inhibitory activity. Figure 2 shows the chromatographic profiles for Na⁺ pump inhibitory activity from midbrain (Fig. 2A) and adrenal (Fig. 2B) supernatants under the various oxygenation conditions. Inhibitory activity under all circumstances chromatographed with a retention time from 76 to 96 min (10 collected fractions, flow rate 2 ml/min). This retention corresponded exactly with that found for bovine HIF (shaded regions in Fig. 2), suggesting but not proving that the inhibitor recovered from brain and adrenal supernatants is the same as the bovine hypothalamic inhibitor structurally characterized as a isomer of ouabain (31).

View larger version (30K): [in this window] [in a new window]   View larger version (37K): [in this window] [in a new window]   Fig. 2.   Sodium pump inhibition in fractions from CHP-20P chromatography under various in vivo and in vitro oxygen conditions. A: midbrain tissue slices (groups 1-4). B: adrenal tissue slices (groups 1A-4A). Shaded portion shows area of elution of bovine hypothalamic inhibitory factor (HIF) under identical chromatographic conditions. Note that the descriptions of groups 1-4 in A apply also to groups 1A-4A in B.

View larger version (18K): [in this window] [in a new window]   Fig. 3.   Na+-K+-ATPase inhibition caused by equal doses of purified tissue slice supernatants in the erythrocyte rubidium uptake (y-axis) and purified Na+-K+-ATPase inhibition (x-axis) assays. Inhibition is normalized for time of incubation and grams of tissue incubated. Linear regression analysis shows a highly significant correlation between values (r = 0.7, P < 0.0001).

Effects of oxygen tension and NaCl on tissular release of Na⁺-K⁺-ATPase inhibitory activity. There was an increase in inhibitory activity in supernatants from midbrain and adrenal tissue samples incubated in vitro under 4% O₂ in all examined conditions compared with paired tissues exposed to high O₂ tensions in vitro (Table 1, groups 1-4; groups 1A-4A). Subsequent exposure of both tissues to high NaCl concentration stimulated further release of inhibitor from midbrain (Table 1, groups 5-8) and adrenal slices (groups 5A-8A), and this was accentuated by in vitro hypoxia, except in midbrain tissues from animals exposed to chronic hypoxia in the in vivo phase of the experiment (group 7 vs. 8). Since equimolar addition of choline chloride did not produce any increase in endogenous inhibitor release (see METHODS; data not shown), the effect of NaCl seems to be due to a specific stimulus and not to change in osmotic strength of the bathing medium. This finding is consistent with results of intrathecal administration of NaCl in vivo (24).

	DISCUSSION

Top Abstract Introduction Methods Results Discussion References

Physiological studies of HIF, recently structurally characterized as an isomer of ouabain, have established this molecule as a candidate for an endogenous Na⁺-K⁺-ATPase regulator with a role in cardiovascular and renal physiology (7, 18, 25). The structural signature of HIF, which definitively distinguished it from ouabain, is the complete cancellation of circular dichroic spectral signals of acyl derivatives of the molecule, reflecting a unique structural arrangement in the steroid portion of the molecule (31).

Bovine hypothalamus has been a reliable source for the extraction of HIF, but which cells or tissues actually produce HIF is unknown. It was shown, however, that rat midbrain slices release HIF in vitro, and this release was modulated by atrial natriuretic peptide but not vasopressin (10). Other laboratories have suggested the adrenal as a site of synthesis of Na⁺-K⁺-ATPase inhibitory activity. In some instances this conclusion was based on indirect evidence using immuno-cross-reactivity to polyclonal anti-ouabain antibodies (19). The specificity of these findings has been questioned (13). We have not previously examined adrenal tissue for the presence of HIF.

Furthermore, nothing has been known about the stimuli for the release or synthesis of HIF from tissue except for the above-mentioned work of Crabos et. al. (10). Jandhyala and Ansari (24) and then Huang and co-workers (23) provided experiments showing that intrathecal administration of NaCl stimulated brain ouabain-like activity, and recently studies in humans provided evidence that hypoxia stimulates the appearance in blood of Na⁺-K⁺-ATPase inhibitory activity or immunoreactivity to anti-digoxin antibodies (2, 12, 15, 32). The purpose of the current work was to examine the effects of these two stimuli on the appearance of Na⁺-K⁺-ATPase inhibitory activity in midbrain and adrenal tissue by direct measurement of bioactivity on purified supernatants of the respective tissue slices incubated in vitro.

In vitro hypoxia proved to be a potent stimulus for the release from both tissues of an Na⁺-K⁺-ATPase inhibitor that coeluted exactly with bovine brain HIF on reverse-phase chromatography. The CHP-20P resin was chosen, since it had provided the crucial separation of HIF from more polar, nonspecific inhibitors of Na⁺-K⁺-ATPase (31). The amount of inhibitory activity released in the hypoxia experiments was inadequate to allow definitive identification of the inhibitor as HIF by the combination of liquid chromatography-mass spectroscopy and circular dichroic analysis of naphthoyl derivatives (31), but the identical chromatographic retention time and parallel behavior in both the rubidium uptake (sodium pump) and direct Na⁺-K⁺-ATPase inhibition assays are consistent with the inhibitor released from the tissues as being HIF. Using HPLC, Ferrandi and co-workers (14) also found coelution of rat midbrain, rat adrenal, and bovine hypothalamic Na⁺-K⁺-ATPase inhibitory activity purified by the method used to obtain HIF in our laboratory.

Two controls were used to address the possibility that low O₂ effected the release of the inhibitor by causing cell death. The first was the choice of O₂ concentration. The level of 4% O₂ represents a mild level of hypoxia, which was shown to not affect cell viability in adrenal tissue (29). The second was the subsequent addition of high NaCl concentration, a stimulus shown by other investigators to stimulate ouabain-like activity release in the central nervous system (23, 24). In our experiment, addition of 300 mM NaCl to cells already stimulated by hypoxia either in vivo or in vitro produced further release of the inhibitor, and this effect was not due to altered osmotic strength, since addition of equimolar choline chloride was without effect.

Other results also suggest that hypoxia represents a specific rather than toxic stimulus. Chronic in vivo hypoxia produced an increase in plasma levels of inhibitory activity, and in adrenal tissue in vivo hypoxia stimulated a large release of activity from cells never exposed to hypoxia in vitro (group 3A). The fact that baseline release of inhibitor from adrenal tissue never exposed to hypoxia (group 1A) was very modest suggests that the in vivo exposure to low O₂ tensions stimulated synthesis and storage of the inhibitor in this tissue, which was released by the subsequent in vitro hypoxia (group 4A).

As detailed in RESULTS, midbrain and adrenal tissues did not always behave in a parallel manner in responses to hypoxia. Both tissues produced relatively modest amounts of inhibitor under the conditions of atmospheric O₂ in vivo followed by high O₂ tensions in the in vitro phase of the experiments (groups 1 and 1A). In vivo hypoxia was associated with decreased release of inhibitor from midbrain on subsequent in vitro incubation at either high or low O₂ tensions, whereas adrenal tissue from rats exposed to chronic hypoxia in vivo responded with dramatic release of inhibitor under both low and high O₂ in vitro (groups 3A and 4A). NaCl was a stimulus to further promote inhibitor release in both tissues, except in adrenal tissue from animals living at atmospheric O₂ where subsequent hypoxia in vitro produced an apparently depleting release of inhibitor from responsive pools as reflected in relatively modest further response to NaCl (group 6A). In general, stimulated adrenal tissue released larger amounts of Na⁺-K⁺-ATPase inhibitor per gram than stimulated midbrain tissue. This may explain the lack of dose response seen with adrenal compared with midbrain extracts where the standard aliquots tested reached saturation of inhibition in the assay in the case of the adrenal activity.

Taken together, these differences between midbrain and adrenal tissue could be interpreted to mean that the adrenal might be the primary source of synthesis and release of HIF in response to relevant stimuli. In vivo hypoxia appears to deplete the midbrain of HIF, whereas it seems to stimulate synthesis and accumulation of the Na⁺-K⁺-ATPase inhibitor in the adrenal. Although this is entirely speculative at this point (see below), if local elaboration of an Na⁺ pump inhibitor were to represent a protective physiological response to tissue hypoxia, then a lower rate of synthesis and accumulation in brain tissue might contribute pathophysiologically to the vulnerability of cerebral tissue to hypoxic/ischemic damage. At any rate, this is the first direct demonstration that adrenal cells contain Na⁺-K⁺-ATPase inhibitory activity that cochromatographs with bovine brain HIF.

The physiological or pathophysiological role of hypoxia-induced Na⁺-K⁺-ATPase inhibitory activity remains largely to be explored. Several lines of work have suggested an association between hypoxia-induced changes in digitalis-like activity on the one hand and enhanced vasoreactivity on the other. Thus De Angelis et al. (12) reported a correlation between a digitalis-like substance, urinary Na⁺ excretion, and blood pressure in human subjects during prolonged exposure to hypoxia. Bender and co-workers (3) showed an increased total peripheral resistance in young men after 3 wk of hypoxia, and pressure response to catecholamine administration was enhanced in rats (1). Guazzi et al. (16) demonstrated an enhanced reactivity of the pulmonary circulatory system in hypertensive subjects in response to hypoxic stimulus. In pregnancy, high altitude is accompanied by enhanced vascular reactivity that was not due to decreased vasodilatory prostaglandin production (21). Our current findings that low O₂ tensions stimulate the tissular release of HIF or an HIF-like compound are consistent with the hypothesis linking hypoxia and vasoreactivity to a circulating Na⁺-K⁺-ATPase inhibitor (2, 12, 15, 32), since HIF has been previously shown to exert potent vasoconstrictive effects in vascular rings in vitro, even in species known to be highly resistant to ouabain (25).

At another level, it is interesting to speculate about hypoxic stimulus of an endogenous Na⁺-K⁺-ATPase inhibitor as part of a cellular protective response to O₂ deprivation. The underlying concept revolves around that fact in certain tissues a large proportion of cellular energy is used for ion pumping and that reductions in these fluxes could result in important energy savings. Brain tissue from certain turtle species is remarkably tolerant to hypoxia compared with mammalian brain. Suarez et al. (30) found glucose metabolic rates in turtle brain to be one-sixth that of rat brain, and Na⁺-K⁺-ATPase activity was found to be 2 to 2.5 times higher in rat brain compared with the turtle. Perez-Pinzon et al. (28) demonstrated downregulation of sodium channels in turtle brain in response to hypoxia. They hypothesized that the resulting decrease in Na⁺ flux could be associated with altered action potential threshold, electrical depression, and decreased ion pumping accounting for the reduced energy expenditure in response to anoxia characteristic of this species (28).

Decreased production of ATP is primordial to other mechanisms of cell injury during ischemia (6). Certain tissues particularly vulnerable to ischemic injury have high concentrations of Na⁺-K⁺ pumps that utilize a large fraction of cellular energy production. Thus, in brain, 40-50% of resting O₂ consumption and, in kidney tissue, >50% of resting O₂ consumption is utilized for active Na⁺ and K⁺ transport (9). A decrease in this Na⁺ pump activity in response to hypoxia could spare ATP for membrane stabilization and other cellular homeostatic processes. ATP sparing might result from multiple mechanisms, including the "channel arrest" discussed above. Local elaboration of an endogenous Na⁺-K⁺-ATPase inhibitor could likewise slow the pump, resulting in cellular ATP conservation. It is therefore of potential importance that hypoxia induced release of endogenous Na⁺-K⁺-ATPase inhibitory activity in brain and adrenal tissue, but our studies cannot bear directly on the question, because we did not measure ATP levels or ion fluxes in the two tissues. Consistent with the hypothesis, however, are the findings of Brezis et al. (5) that ouabain administration in the isolated perfused kidney is protective against hypoxic injury in the vulnerable, Na⁺-K⁺-ATPase-rich medullary thick ascending limb of Henle.

In summary, we have shown that hypoxia is a potent stimulus for the release of an endogenous inhibitor of Na⁺-K⁺-ATPase from both midbrain and adrenal tissues in vitro and that this release is in general further stimulated by high NaCl concentration. The substance cochromatographs with HIF and has Na⁺-K⁺-ATPase inhibitory activity in two bioassays parallel to HIF, a mammalian isomer of ouabain with demonstrated regulation of Na⁺ pump activity in cardiovascular and renal cells. Further studies will be necessary to explore the role of endogenous Na⁺-K⁺-ATPase inhibition in cellular mechanisms for adaptation to hypoxic stress.