This Article Abstract Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Chi, Y. Articles by Bansal, V. K. Search for Related Content PubMed PubMed Citation Articles by Chi, Y. Articles by Bansal, V. K.

(Hypertension. 1996;27:456-464.)
© 1996 American Heart Association, Inc.

Articles

Correlations of Na⁺-Li⁺ Exchange Activity With Na⁺ and Li⁺ Binding and Phospholipid Composition in Erythrocyte Membranes of White Hypertensive and Normotensive Individuals

A Nuclear Magnetic Resonance Investigation

Yuling Chi; Duarte Mota de Freitas; Mary Sikora; Vinod K. Bansal

From the Department of Chemistry, Loyola University of Chicago (Y.C., D.M. de F.), and the Renal Disease and Hypertension Section, Loyola University Stritch School of Medicine, Maywood (M.S., V.K.B.), Ill.

	Abstract

Top Abstract Introduction Methods Results Discussion References

 
Abstract Enhanced Na⁺-Li⁺ exchange activity has been reported in red blood cells (RBCs) of white patients with essential hypertension compared with RBCs of normotensive individuals. To understand the factors responsible for this finding, we applied novel and conventional spectroscopic and kinetic methods to blood samples from 10 hypertensive and 10 normotensive individuals. We measured the kinetic parameters (V_std, V_max, and K_m) for RBC Na⁺-Li⁺ exchange by atomic absorption spectrophotometry and used ²³Na and ⁷Li nuclear magnetic resonance relaxation methods to measure Na⁺ and Li⁺ binding to RBC membranes as well as ³¹P nuclear magnetic resonance spectroscopy to measure membrane phospholipid compositions. We found significant differences between the two groups for the affinity of Na⁺ for the RBC membrane (0.202±0.054 mmol/L^-1 for hypertensive patients versus 0.296±0.071 mmol/L^-1 for normotensive subjects, P<.005). The kinetic parameters of RBC Na⁺-Li⁺ exchange (V_std, V_max, and K_m) were 0.32±0.09 and 0.66±0.17 mmol Li⁺/L cell·h and 160±62 mmol/L, respectively, for hypertensive patients versus 0.21±0.06 and 0.32±0.14 mmol Li⁺/L cell·h and 86±69 mmol/L for normotensive subjects (P<.05). The fractions of phosphatidylserine and phosphatidylethanolamine were 0.153±0.009 and 0.294±0.016 for hypertensive patients versus 0.138±0.013 and 0.325±0.018 for normotensive subjects (P<.05). The Na⁺ binding constants were negatively correlated with the K_m values for both the hypertensive (r=-.61, P=.01) and normotensive (r=-.43, P=.04) groups. Changes in lipid-protein interactions in the RBC membranes of hypertensive patients appear to be responsible for weaker Na⁺ binding to the membrane and for the faster rates of RBC Na⁺-Li⁺ exchange.

Key Words: hypertension, essential • membranes • erythrocytes • ions • nuclear magnetic resonance

	Introduction

Top Abstract Introduction Methods Results Discussion References

 
The transport pathways for Li⁺ ions across the human RBC membrane have been identified and characterized in detail; they include Na⁺-Li⁺ exchange or countertransport, Na⁺-Li⁺ cotransport, anion exchange, the Na⁺,K⁺-ATPase, and a residual Li⁺ pathway.¹ At physiologically relevant concentrations of Na⁺ and K⁺, Li⁺ entry into RBCs occurs via the anion exchange protein (when bicarbonate is present in the incubation medium), whereas Li⁺ efflux from RBCs is mediated primarily by Na⁺-Li⁺ exchange; the leak pathway contributes to both entry and extrusion of Li⁺.² Much attention has been devoted to the investigation of RBC Na⁺-Li⁺ exchange because abnormalities in the rates of Li⁺ transport mediated by this RBC membrane protein are associated with several diseases, such as essential hypertension,³ hyperlipidemia,⁴ and diabetes.⁴ Abnormal RBC Na⁺-Li⁺ exchange rates have also been reported for women taking oral contraceptives and in late pregnancy, as well as for bipolar patients treated with lithium carbonate.¹

Before 1990, the maximal rates of Na⁺-Li⁺ exchange in human RBCs were generally assayed by measurement of the rates of Li⁺ efflux from Li⁺-loaded RBCs into Na⁺-containing and Na⁺-free media; the maximal rates of RBC Na⁺-Li⁺ exchange were then calculated by subtraction of the measured rates of Li⁺ transport in the Na⁺-free medium from those measured in the Na⁺-containing medium.³ Mechanistic studies of RBC Na⁺-Li⁺ exchange have indicated that the RBC membrane transport protein is asymmetrical, with higher Na⁺ and Li⁺ affinities on the intracellular than extracellular side of the membrane.^{5 6} It was recently found that under standard transport assay conditions, Li⁺ is present at a saturating concentration at the internal binding site of the RBC membrane transport protein, but Na⁺ may not be saturating at the external site.^{4 7} The K_m values for extracellular Na⁺ are, at least for some individuals, of the same order of magnitude as the extracellular Na⁺ concentration (140 mmol/L) used in the standard transport assay, suggesting that the RBC Na⁺-Li⁺ exchange protein is far from saturated with Na⁺ on the extracellular side of the RBC membrane. Therefore, the rates under standard transport assay conditions may not be maximal rates of RBC Na⁺-Li⁺ exchange; variations in Na⁺ affinity (K_m) and V_max could change the observed rates. Only by varying the Na⁺ concentration in an isotonic suspension medium can one measure the true kinetic parameters of RBC Na⁺-Li⁺ exchange.^{4 7 8} In the present study, we attempted to identify the factors responsible for the elevated Na⁺-Li⁺ exchange rates in RBCs from hypertensive patients by using the new transport assay method developed by Rutherford and coworkers⁷ in conjunction with NMR methods we developed^{9 10 11 12} that probe the interactions of Li⁺ and Na⁺ ions with the phosphate head groups in the human RBC membrane.

Several investigators have measured RBC parameters from hypertensive and normotensive individuals by means of the methods that we also used in our study.^{3 7 13 14 15 16 17} However, this is the first report in which all available, modern spectroscopic and kinetic methods were used for the detailed investigation of the factors responsible for the fast rates of Na⁺-Li⁺ exchange in RBCs from white essential hypertensive patients. Moreover, some measurements, eg, ³¹P NMR measures of the phospholipid composition of RBC membranes from hypertensive patients, have never been reported previously, and the understanding of and experimental procedures for the measurement of the kinetic parameters of RBC Na⁺-Li⁺ exchange have changed drastically in the past 5 years.^{4 7 8 15} A reexamination of elevated RBC Na⁺-Li⁺ exchange in RBCs from some hypertensive patients is therefore warranted.

We investigated whether an abnormal phospholipid composition of the RBC membrane accounts for abnormal Li⁺ or Na⁺ interactions with the RBC membranes from hypertensive patients, resulting in elevated Na⁺-Li⁺ exchange rates. We hypothesized that alterations in the phospholipid compositions of the RBC membranes of hypertensive patients resulted in weaker Li⁺ or Na⁺ binding to the RBC membranes and in elevated Na⁺-Li⁺ exchange rates. Opposite trends should be observed for normotensive individuals.

	Methods

Top Abstract Introduction Methods Results Discussion References

 
Subjects
Whole blood samples from three male and seven female essential hypertensive patients (without hypothyroidism, hyperlipidemia, diabetes mellitus, obesity, pregnancy, or any other condition that could affect blood pressure)⁴ and from three male and seven female normotensive individuals were obtained through the Hypertension and Renal Section of the Loyola University Medical Center. The normotensive individuals were matched with the hypertensive patients according to sex and race. Because RBC Na⁺-Li⁺ exchange activity is elevated in white but not in black hypertensive patients,^{13 14} all individuals used in this study were white. The protocols for the experiments and procurement of human subjects and samples were approved by the Human Investigative Review Committee of the Loyola University Medical Center, Institutional Review Board Committee. The selection of essential hypertensive patients used in this study was based on medical histories. Some of the patients were receiving antihypertensive medication (verapamil, triamterene, atenolol, enalapril, clonidine, and/or hydrochlorothiazide) at the time of blood drawing. All normotensive individuals had diastolic pressures less than or equal to 80 mm Hg at the time of blood drawing. All hypertensive and normotensive individuals had no dietary restrictions and used salt freely; they were instructed to fast for a minimum of 12 hours before blood drawing. Triglyceride and cholesterol concentrations were measured for all samples at the clinical laboratory of the Loyola University Medical Center.

Materials
Choline chloride, glucose, sucrose, LiCl, NaCl, dimethyl sulfoxide, MeOH, chloroform, deuterated chloroform (CDCl₃), KCl, and deuterium oxide (99.8%) were supplied by Aldrich Chemical Co, Inc. HEPES, EDTA (99%), butylated hydroxytoluene, bovine serum albumin, ouabain, and the detergent octyl ß-D-glucopyranoside were purchased from Sigma Chemical Co. The protein assay dye reagent was obtained from Bio-Rad Laboratories.

Sample Preparation
Whole blood was collected by venipuncture into non–silicone-coated tubes with EDTA (K₃). RBCs were packed and concentrated by centrifugation at 3928g and washed at 4°C with choline washing solution (112 mmol/L choline chloride, 80 mmol/L sucrose, 10 mmol/L glucose, 10 mmol/L HEPES, pH 7.4) three times. Washed RBCs were added to lithium loading solution (150 mmol/L LiCl, 10 mmol/L glucose, 10 mmol/L HEPES, pH 7.4) at 0.10 hematocrit and incubated at 37°C for 3 hours. After incubation, extracellular Li⁺ was removed from the Li⁺-loaded RBCs by washing four times with choline washing solution. The final intracellular lithium concentration (9.09±1.07 mmol/L of cells) was determined by AA. The Li⁺-loaded RBCs were suspended at 0.10 hematocrit in six media containing 150, 100, 70, 40, 20, and 0 mmol/L NaCl that were made isotonic with choline chloride and also contained 10 mmol/L glucose, 0.1 mmol/L ouabain, and 10 mmol/L HEPES, pH 7.4. Aliquots were taken every 20 minutes from each of the Li-loaded RBC suspensions and centrifuged at 10 519g for 1.5 minutes at 4°C. The supernatants were collected and analyzed by AA.^{3 7}

Unsealed RBC membranes were prepared by lysing of the washed, packed RBCs with 5 mmol/L HEPES buffer, pH 8.0.¹⁸ The membranes were washed three more times with this same buffer and centrifuged at 51 948g at 4°C until they were pale white. Protein concentration was measured by the Bradford assay.^{19 20}

Phospholipids in the RBC membrane were extracted as follows^{12 21} : 17 mL methanol was added to 1 mL membrane and mixed for 10 minutes. Then 33 mL chloroform was added, and the sample was mixed for an additional 15 minutes. All extracting solvents contained 0.227 mmol/L butylated hydroxytoluene as an antioxidant. The resulting extract was filtered through a sintered glass funnel and washed with 50 mL of a chloroform/methanol mixture (2:1). The filtrate was washed at 0.2 times its total volume with 99 mmol/L KCl for removal of all nonlipid impurities. The bottom chloroform layer was collected and dried in a rotary evaporator at 30°C. The purified phospholipids were suspended in 3 mL of the solvent mixture solution (CDCl₃/MeOH/EDTA [0.2 mol/L], 125:8:3).

NMR, AA, and Hematocrit Measurements
²³Na and ⁷Li spin-lattice relaxation time (T₁) measurements were made at 79.4 and 116.5 MHz, respectively, on a Varian VXR-300 NMR spectrometer with 10-mm low- and high-frequency probes, respectively, at 37°C, with spinning. ³¹P NMR spectra were obtained at 121.4 MHz on the same instrument at 27°C. Analysis of phospholipids in RBC membranes was conducted by ³¹P NMR spectroscopy, as described.^{12 21} The phospholipid extract sample was placed in a 10-mm NMR tube and allowed to stand for a few minutes until the aqueous phase had separated. The spinning turbine was adjusted along the NMR tube so that only the chloroform phase was in the region of signal detection. Each ³¹P NMR resonance in the spectra of phospholipid extract samples was assigned by spiking the samples with pure phospholipids. The areas under the ³¹P NMR resonances in the spectra of phospholipid extracts of RBC membranes were integrated and normalized to 100% for the five major classes of phospholipids. The accuracy of the percentage composition measured by ³¹P NMR spectroscopy for PC, PS, PE, and sphingomyelin is within ±10% of the values displayed in Fig 4, whereas the accuracy of the PI values is within ±20%; the accuracy of the ³¹P NMR measurements is lower for the PI values because of their lower concentrations in RBC membranes and the associated problems of baseline noise.²² Total phospholipid amounts in the RBC membrane extracts were obtained by comparison of spectra of the extract samples alone with spectra of the extracts spiked with known amounts of pure PC.

View larger version (11K): [in this window] [in a new window]   Figure 4. Plots show phospholipid fractions for RBC membranes from hypertensive (circles) and normotensive (squares) individuals. Sph indicates sphingomyelin.

AA measurements of Li⁺ concentrations were conducted at 670.8 nm on a Perkin-Elmer 5000 spectrophotometer. Hematocrit measurements were obtained with a microcentrifuge (model MB IM116, IEC).

Calculations
The kinetic parameters of Na⁺-Li⁺ exchange, V_std, V_max, and K_m, were obtained from the following equations^{3 7} :

where Ht is hematocrit.

We used James-Noggle plots²³ to calculate the K_b values for Na⁺ and Li⁺ to the RBC membranes. We calculated these values by using the observed T₁ values and assuming a two-state (free and bound Na⁺ or Li⁺) model undergoing fast exchange^{9 10 11 12} :

where [M⁺] and [B] are the total concentrations of Na⁺ or Li⁺ and of membrane binding sites, respectively. The r² values for the curve fitting of the data used in the calculations of the K_b values and kinetic parameters reported in Figs 2 and 5 were greater than .90. Na⁺ in the exposed membrane suspension is not 100% visible by ²³Na NMR.²⁴ As for similar investigations,^{16 17} the K_b(Na) values that we report in this study are uncorrected for Na⁺ invisibility of the ²³Na⁺ NMR resonance; however, the visibility factor did not affect the conclusions drawn in this study.

View larger version (14K): [in this window] [in a new window]   Figure 2. Plots of Na+ (A) and Li+ (B) Kb values for RBC membranes from hypertensive (circles) and normotensive (squares) individuals.

View larger version (13K): [in this window] [in a new window]   Figure 5. Plots of Vstd (A), Vmax (B), and Km (C) values for RBCs from hypertensive (circles) and normotensive (squares) individuals.

Statistical Analyses
Each essential hypertensive patient was matched to a normotensive individual according to sex and race. The statistical significance of the variances between hypertensive and normotensive individuals for the kinetic parameters of Na⁺-Li⁺ exchange, Na⁺ and Li⁺ K_b values, and phospholipid composition of the RBC membrane were analyzed by one-way ANOVA with Tukey's conservative correction. We used a power analysis to determine the sample size needed for this investigation. We obtained correlation coefficients among these parameters by using a Spearman correlation; correlation coefficients greater than or equal to .40 were considered significant (P.05).

	Results

Top Abstract Introduction Methods Results Discussion References

 
Fig 1 shows the ²³Na and ⁷Li T₁ values observed on titration with Na⁺ (A) or Li⁺ (B) of unsealed membranes prepared from the RBCs of a hypertensive patient and a matched normotensive individual at the same membrane protein concentration. For the same Na⁺ concentrations, the ²³Na T₁ values were significantly higher (P<.05) for the unsealed membrane suspension from RBCs of a hypertensive patient than from those of the matched normotensive individual. For a given Li⁺ concentration, however, the ⁷Li T₁ values were not significantly different (P>.05) for the hypertensive patient and matched normotensive individual. Similar results were obtained for the rest of the matched pairs.

View larger version (11K): [in this window] [in a new window]   Figure 1. 23Na (A) and 7Li (B) T1 values for RBC membrane samples from a hypertensive patient (diamonds) and normotensive subject (squares), who were matched according to sex and race, in the presence of increasing Na+ or Li+ concentrations. Membrane protein concentration was 8.0±1.0 mg/mL for both samples.

We calculated Na⁺ and Li⁺ K_b values to RBC membranes by using James-Noggle plots.²³ One-way ANOVA with Tukey's conservative correction showed significant differences between hypertensive patients and normotensive subjects for the Na⁺ K_b values to the RBC membrane (0.202±0.054 versus 0.296±0.071 mmol/L^-1, n=10, P<.005) (Table and Fig 2). No significant difference was found between the two groups for the Li⁺ K_b values to the RBC membranes (Table and Fig 2).

View this table: [in this window] [in a new window]   Table 1. Na+-Li+ Exchange Parameters, Na+ and Li+ Binding Constants, Phospholipid Compositions, and Total Phospholipids in RBCs and Clinical Characteristics of Essential Hypertensive and Normotensive Individuals

Fig 3 shows a typical ³¹P NMR spectrum of a phospholipid extract from a human RBC membrane. The assignments of the phospholipid resonances are indicated in the spectrum and were obtained as described in "Methods." Two types of PE are present in human RBC membranes, and they can be discriminated by ³¹P NMR spectroscopy: regular PE has an alkyl ether on the glyceride backbone, whereas PE plasmalogen has an alkenyl ether. The areas under the two PE resonances were added to yield the PE composition. Under our current experimental conditions for phospholipid analysis, the percentage of chloroform in the solvent mixture was increased (see "Methods"), resulting in complete resolution of the PS resonance from the PE/sphingomyelin envelope of ³¹P NMR resonances. In contrast, in our previous ³¹P NMR study of phospholipid extracts of RBC membranes, the PS resonance was overlapping with the ³¹P NMR resonances of PE, PE plasmalogen, and sphingomyelin.¹² We therefore conclude that the resolution of ³¹P NMR spectra of phospholipid extracts from RBC membranes is improved significantly when the percentage of chloroform in the solvent mixture is increased.^{12 22}

View larger version (7K): [in this window] [in a new window]   Figure 3. 31P NMR spectrum of a phospholipid extract from human RBC membrane. Assignments of each resonance to the various phospholipids are shown. PEp indicates PE plasmalogen; Sph, sphingomyelin; and AAPC, alkylacyl PC. Positions of the resonances (chemical shifts) are reported relative to that of PC set at -0.84 ppm.22

By using ³¹P NMR spectroscopy (Table and Fig 4) and according to one-way ANOVA with Tukey's conservative correction, we found that the fraction of PS in RBC membranes was significantly higher for essential hypertensive than for normotensive individuals (0.153±0.009 versus 0.138±0.013, n=10, P<.05), whereas the fraction of PE was significantly lower (0.294±0.016 versus 0.325±0.018, n=10, P<.001). No statistically significant differences (P.05) were found between the hypertensive and normotensive control groups for the fractions of the other phospholipids. We also used the ³¹P NMR method to measure total phospholipid concentrations in the RBC membranes from hypertensive and normotensive individuals; total phospholipid concentrations did not differ significantly between the RBC membranes of the two groups (Table).

As shown in the Table and Fig 5, the RBC Na⁺-Li⁺ exchange rates measured by the standard method (V_std) and the kinetic parameters V_max and K_m, which we measured by varying the extracellular Na⁺ concentration, were significantly higher for essential hypertensive patients than for normotensive individuals (V_std, V_max, and K_m were 0.32±0.09 and 0.66±0.17 mmol Li⁺/L cell·h and 160±62 mmol/L for hypertensive patients versus 0.21±0.06 and 0.32±0.14 mmol Li⁺/L cell·h and 86±69 mmol/L for normotensive subjects, n=10, P<.05).

By using a power analysis, we found that a sample size of 10 for each group was sufficient to yield statistically significant differences with a power level of 0.90 and an value of 0.05 for all of the RBC parameters [V_std, V_max, K_m, K_b(Na), PS, and PE] that were also statistically significantly different by one-way ANOVA with Tukey's conservative correction. For a power level of 0.95, however, only the V_max, K_m, and PE values were sufficient to yield statistically significant differences between the two groups with an value of 0.05.

We calculated Spearman correlation coefficients for the RBC parameters within the hypertensive and normotensive groups. For the hypertensive group, we found that the V_max values were positively correlated with the V_std values (r=.49, P=.03), the K_b(Na) values were negatively correlated with the K_m values (r=-.61, P=.01), and the K_b(Na) values were positively correlated with the PE values (r=.52, P=.02). For the control group, we also found that the V_max values were positively correlated with the V_std values (r=.82, P=.01), the K_b(Na) values were negatively correlated with the K_m values (r=-.43, P=.04), and the K_b(Na) values were significantly positively correlated with the PE values (r=.42, P=.04). No other significant correlations were found among the remaining measured parameters within each group.

	Discussion

Top Abstract Introduction Methods Results Discussion References

 
Increased sodium-lithium countertransport activity was previously reported not only in hypertensive³ but also in hyperlipidemic^{15 25 26 27} patients. Abnormal membrane phospholipid composition in erythrocytes from hyperlipidemic patients has been investigated.^{28 29 30} To avoid the possible interference of hyperlipidemia in our investigation of essential hypertension, we selected hypertensive and normotensive individuals with normal triglyceride and cholesterol concentrations. The normal average values for cholesterol and triglyceride levels at the clinical laboratory of the Loyola University Medical Center were 5.43 mmol/L and less than 250 mg/dL, respectively. In the sample used in our study, three essential hypertensive patients and one normotensive individual had slightly elevated cholesterol levels; however, elevated triglyceride levels were found for only one normotensive individual. We attribute the slightly elevated cholesterol and triglyceride levels for some of our subjects to noncompliance with the instructions we gave for fasting at the time of blood drawing (see "Methods"). However, the triglyceride and cholesterol data listed in the Table indicate that triglyceride and cholesterol concentrations did not differ significantly between the two groups.

Na⁺ and Li⁺ ions are in the fast exchange domain on the ²³Na and ⁷Li NMR time scales^{9 10 11 12 16 17} ; the ²³Na or ⁷Li T₁ values (Fig 1) measured with RBC membrane samples of hypertensive and normotensive individuals therefore depict the weighted averages of free Na⁺ or Li⁺ in the suspension medium and of Na⁺ or Li⁺ bound to the RBC membranes. Free Na⁺ or Li⁺ exhibits relatively larger T₁ values than bound Na⁺ or Li⁺. The ²³Na or ⁷Li T₁ value increased in the presence of increasing concentrations (0 to 10 mmol/L) of Na⁺ or Li⁺ in membrane samples of both hypertensive and normotensive individuals because of increasing concentrations of free Na⁺ or Li⁺. These observations confirm that both ²³Na and ⁷Li T₁ values are sensitive to Na⁺ and Li⁺ binding to RBC membranes and that they can be used for the calculation of the Na⁺ and Li⁺ K_b values to RBC membranes from James-Noggle plots (see "Methods"). For Na⁺ concentrations lower than 10 mmol/L, the amount of Na⁺ bound to the same amount of RBC membrane was smaller for hypertensive than for normotensive individuals (Fig 1A); however, for Na⁺ concentrations greater than or equal to 10 mmol/L, the amount of Na⁺ bound to RBC membranes from both hypertensive and normotensive individuals appears to be the same because the observed ²³Na T₁ values are mostly determined by the fractions of free Na⁺. In contrast, for a given Li⁺ concentration, the amount of Li⁺ bound to the RBC membranes of hypertensive patients was not significantly different from that of normotensive individuals (Fig 1B). Our observed ²³Na T₁ data (Fig 1A) and the calculated K_b(Na) values (Fig 2 and Table) indicate that Na⁺ has a weaker affinity for RBC membranes of hypertensive than of normotensive individuals. However, our observed ⁷Li T₁ results (Fig 1B) and the calculated K_b(Li) values (Fig 2 and Table) indicate that Li⁺ has a similar affinity for RBC membranes of both hypertensive and normotensive individuals. Our calculated K_b(Na) values are in agreement with those previously calculated from ²³Na T₁ values measured in RBC membrane suspensions from hypertensive and normotensive individuals.^{16 17} Similarly, our calculated K_b(Li) values for RBC membranes from normotensive individuals are in agreement with those found in our previous investigation of lithium-treated bipolar patients and normal individuals.¹²

The head groups of the phospholipids PI and PS are negatively charged; these two anionic phospholipids reside primarily in the inner leaflet of the RBC membrane. Our previous ⁷Li NMR relaxation studies of Li⁺-loaded RBCs and of RBC components in the absence and presence of varying Mg²⁺ concentrations indicate that the primary binding sites for alkali metal ions in intact RBCs are present in the cytoplasmic side of the RBC membrane^{10 11} and that variations in Li⁺ or Na⁺ K_b values to RBC membranes may be associated with abnormal phospholipid composition.¹² Our ³¹P NMR data (Table and Fig 4) indicated that there were significant differences between the hypertensive and normotensive groups in the fractions of PS and PE but not in those of PC, sphingomyelin, and PI. The phospholipid percentage compositions of the human RBC membrane that we measured by ³¹P NMR spectroscopy are in general agreement with those measured by other researchers by two-dimensional thin-layer chromatography^{31 32} or high-performance liquid chromatography.³³ The slightly different percentage compositions measured for human RBC membranes by the different methods can be attributed to the fact that ³¹P NMR spectroscopy measures the total phosphate directly in each phospholipid head group; the visualization reagents used in thin-layer chromatography or the UV absorbance measurements in high-performance liquid chromatography result in the detection of only unsaturated fatty acids.^{34 35} These methodological differences presumably also account for the fact that in this study we found significant differences between the RBC membranes from the hypertensive and normotensive groups in the percentage compositions of PS and PE, but no significant differences were found previously with thin-layer chromatographic measurements.³¹ We also used ³¹P NMR spectroscopy to measure the total phospholipid concentrations in the RBC membranes from hypertensive and normotensive individuals and found that they were not significantly different (Table). However, the total phospholipid concentrations in the RBC membranes that we measured by ³¹P NMR spectroscopy were in good agreement with those previously reported.^{32 33} The statistically significant differences that we found in the PS and PE fractions are therefore not associated with variations in the total phospholipid concentrations of the RBC membranes from the two groups.

The AA-determined rates of RBC Na⁺-Li⁺ exchange measured under standard assay conditions (V_std) and in isotonic media containing varying Na⁺ concentrations (V_max) were found to be significantly higher for the hypertensive than normotensive individuals (Table and Fig 5); however, the exchange rates for women were not significantly lower (P>.09) than those for men in both the hypertensive and normotensive control groups. The sex variations and enhanced rates that we observed for the hypertensive patients are not in agreement with those previously reported, presumably because of the small sample size used in our study.^{3 7 36 37} Within the hypertensive group, one-way ANOVA did not show significant differences for the AA-determined V_std and V_max values of RBC Na⁺-Li⁺ exchange between the subgroups that were taking and not taking antihypertensive medication; these results are in agreement with those previously reported,^{37 38} which show that medications for blood pressure regulation do not affect the enhanced RBC Na⁺-Li⁺ exchange rates in hypertensive patients. However, a study³⁹ that appeared in print after we collected the blood samples for the present study showed that long-term antihypertensive therapy with angiotensin-converting enzyme inhibitors caused a significant decrease in RBC Na⁺-Li⁺ exchange rates. Three of the 10 patients used in our study were receiving angiotensin-converting enzyme inhibitors. However, we found no significant differences (P>.23) in the V_std and V_max values between these 3 and the remaining 7 patients; this difference between our study and that on the effect of angiotensin-converting enzyme inhibitor therapy on RBC Na⁺ transport³⁹ might be related to the smaller sample size we used. The AA-determined K_m values of Na⁺ to the extracellular side of RBC membranes were also significantly higher for hypertensive than for normotensive individuals (Table and Fig 5); the K_m values that we obtained for RBCs from both hypertensive and normotensive individuals are in general agreement with those previously measured in some^{15 40} but not all⁷ previous reports.

The K_b(Na) values that we calculated from the ²³Na T₁ values measured in Na⁺-containing suspensions of unsealed RBC membranes are measures of Na⁺ binding to the phosphate head groups of phospholipids and protein binding sites present in the internal and external leaflets of the RBC membrane. In contrast, the AA-determined K_m values that we obtained by varying the extracellular Na⁺ concentration in the transport assay are Na⁺ K_m values to the extracellular side of the membrane Na⁺-Li⁺ exchange protein. It is therefore not surprising that significant negative correlations existed between the K_b(Na) values and K_m values within the hypertensive and normotensive groups. A comparison of K_b(Na) and K_m requires taking the reciprocals of one of these sets of values; the values of K_b(Na) are larger than those of 1/K_m, presumably because of the additional number of membrane binding sites for Na⁺ that are probed in the ²³Na NMR relaxation measurements relative to the AA-determined transport measurements.

The authors of some recent studies^{40 41} have claimed that K_m values in the range of 160 mmol/L for RBC samples from hypertensive patients cannot be obtained accurately from measurements at an extracellular Na⁺ concentration of 150 mmol/L. This statement is only true provided that the estimates of K_m are derived from hyperbolic Michaelis-Menten plots (as opposed to K_m estimates originating from linear, double-reciprocal plots of 1/V versus 1/[Na⁺]). In our study, however, we did not simply conduct single measurements in suspensions containing 150 mmol/L Na⁺; we varied the extracellular Na⁺ concentration from 0 to 150 mmol/L, and we maintained the isotonicity of all suspension media at 0.15 mol/L with choline chloride (see "Methods"). The method recently described by Canessa and coworkers (Canessa et al⁴⁰ and Zerbini et al⁴¹ ) involves the use of hypertonic Na⁺ media (150 to 250 mmol/L). We do not favor that method because no data in these two studies supported the contention that the ionic permeability of RBC membranes (including that mediated by Na⁺-Li⁺ exchange) and viability of the cells were not affected by the harsh, hypertonic conditions.

Previous ³¹P and ²H NMR studies of phospholipid suspensions indicated that PS and PI interacted with Li⁺ and Na⁺ ions.^{42 43} We found significant differences between the two groups for the PS values by one-way ANOVA with Tukey's conservative correction (Table). Because of baseline noise and the small amounts of PI present in RBC membranes,^{12 22} it is possible to determine the PI content with an accuracy of only ±20%; the low accuracy of the PI content measured by the ³¹P NMR method indicates that differences in PI levels between the two groups, and any correlations between the PI content and kinetic transport parameters should be interpreted with caution. However, the PS content in RBC membranes is sufficiently high to be determined accurately by ³¹P NMR spectroscopy. We speculate that the higher PS percentage composition of RBC membranes from hypertensive patients relative to that from normotensive individuals might drive the positively charged extracellular Na⁺ ions to traverse the RBC membrane more quickly for essential hypertensive patients than for normotensive subjects. We previously reported that Li⁺-treated bipolar patients had lower Na⁺-Li⁺ exchange rates, which were negatively correlated with Li⁺ binding to RBC membranes, and that they had higher PS contents than did normal individuals.¹² The higher PS content in the inner leaflet of RBC membranes of Li⁺-treated bipolar patients is presumably responsible for a stronger interaction with intracellular Li⁺ and makes it more difficult for Li⁺ to be transported, resulting in lower rates of RBC Na⁺-Li⁺ exchange.¹²

In this initial, exploratory study, we did not recruit hypertensive or normotensive individuals who were on a specific diet. It is therefore possible that the phospholipid fractions as well as the triglyceride and cholesterol levels that we report for RBC membranes were influenced by the type of diet that the subjects were on at the time of blood drawing. It is important to note, however, that all of our subjects were instructed to have fasted for a minimum of 12 hours at the time of blood drawing; we estimate that only approximately 70% of the subjects complied with these instructions. We believe that this precaution in our experimental design should have minimized the contribution of individual diets on the lipid compositions of the RBC membranes that we measured.

In our study, there was a perfect match between the hypertensive and normotensive control groups in terms of race and sex. Although we attempted to match each essential hypertensive patient as closely as possible with a normotensive subject of similar age, this was not always achieved, as is apparent from the significantly higher average age of the hypertensive group relative to the average age of the normotensive group (Table). It is always difficult to obtain a perfectly age-matched sample of hypertensive and normotensive individuals because essential hypertension is known to be far more common in the elderly population.³⁶

We found significant differences between the hypertensive and normotensive groups in both systolic and diastolic pressure values (Table). None of our patients had isolated late-onset systolic hypertension. Every patient had diastolic hypertension, and each had been diagnosed to have essential hypertension for a minimum of 10 years. The statistically significant differences between the RBC parameters of the samples from hypertensive patients relative to those from normotensive individuals that we found in this study are therefore not related to late-onset systolic hypertension but rather to long-term diastolic hypertension.

We report here the results of our statistical analysis with a one-way ANOVA method with Tukey's conservative correction as well as the Spearman correlation coefficients. The ANOVA and Spearman statistical tests are appropriate for the small sample size used.⁴⁴ Despite the small sample size, we were able to obtain statistically significant differences for and significant correlations within several RBC parameters (V_max, K_m, and PE) by using one-way ANOVA with Tukey's conservative correction and Spearman correlation coefficients; the power analysis we conducted (see "Results") confirmed that our sample size was sufficient to yield significant differences at the 95% confidence level. However, future epidemiological studies with the methods described in this study should include a minimum sample size of 15 to 20 to confirm whether the significant differences in RBC parameters [V_std, K_b(Na), and PS] that we observed at a 90% confidence level also hold for a larger sample size. We did not attempt to use a larger sample size because our goal was to search for the RBC parameters that provide the most information on the origins of elevated Na⁺-Li⁺ exchange rates in RBCs from white essential hypertensive patients. The time-consuming processing of each sample prevented us from investigating a much larger sample size in this initial study. Our focus was not on investigating the mortality, epidemiology, or treatment efficacy for essential hypertensive patients; the data we accumulated therefore are not sufficient to allow us to draw conclusions about these aspects of essential hypertension in white patients. Nonetheless, the information we report here is important and significant because it identified more clearly than in previous studies the molecular abnormalities at the level of RBC membranes for white essential hypertensive patients who had elevated RBC Na⁺-Li⁺ exchange rates.

Alterations of the fatty acid composition of some phospholipids in essential hypertensive patients have been found to cause abnormal RBC Na⁺-Li⁺ exchange activity.^{1 28 29 30 45 46} More detailed investigations of the molecular species composition of phospholipids may therefore help us to understand the increased Na⁺-Li⁺ exchange activity in RBCs from essential hypertensive patients.

	Selected Abbreviations and Acronyms

 

AA = atomic absorption Kb = binding constant Km = dissociation constant NMR = nuclear magnetic resonance PC = phosphatidylcholine PE = phosphatidylethanolamine PI = phosphatidylinositol PS = phosphatidylserine RBC = red blood cell T1 = spin lattice relaxation time Vmax = maximal velocity Vstd = standard velocity

	Acknowledgments

 
This work was supported in part by US Public Health Service grant MH-45926 and by funds from Loyola University. We are grateful to Prof George Michel (Department of Psychology, DePaul University, Chicago, Ill) for advice with the statistical analysis of the data, and to Elisabeth Lanzl from The University of Chicago Medical Center for editing the manuscript.

	Footnotes

 
Reprint requests to Dr Duarte Mota de Freitas, Department of Chemistry, Loyola University of Chicago, 6525 North Sheridan Rd, Chicago, IL 60626. E-mail dfreita@orion.it.luc.edu.

Received October 9, 1995; first decision November 21, 1995; accepted November 30, 1995.

	References

Top Abstract Introduction Methods Results Discussion References

 
1. Duhm J. Pathways of lithium transport across the human erythrocyte membrane. In: Thelier M, Wissocq J-C, eds. Lithium Kinetics. Carnforth, UK: Marius Press; 1992:27-53.

2. Ehrlich BE, Diamond JM. Lithium fluxes in human erythrocytes. Am J Physiol. 1979;237:C102-C110. [Abstract/Free Full Text]

3. Canessa M, Adragna N, Solomon HS, Connolly TM, Tosteson DC. Increased sodium-lithium countertransport in red cells of patients with essential hypertension. N Engl J Med. 1980;302:772-776. [Abstract]

4. Rutherford PA, Thomas TH, Wilkinson R. Erythrocyte sodium-lithium countertransport: clinically useful, pathophysiologically instructive or just phenomenology? Clin Sci. 1992;82:341-352. [Medline] [Order article via Infotrieve]

5. Sarkadi B, Alifimoff JK, Gunn RB, Tosteson DC. Kinetics and stoichiometry of Na-dependent Li transport in human red blood cells. J Gen Physiol. 1978;72:249-265. [Abstract/Free Full Text]

6. Hannaert PA, Garay RP. A kinetic analysis of Na-Li countertransport in human red blood cells. J Gen Physiol. 1986;87:353-368. [Abstract/Free Full Text]

7. Rutherford PA, Thomas TH, Wilkinson R. Increased erythrocyte sodium-lithium countertransport activity in essential hypertension is due to an increased affinity for extracellular sodium. Clin Sci. 1990;79:365-369. [Medline] [Order article via Infotrieve]

8. Aronson JK. Methods for expressing the characteristics of transmembrane ion-transport systems. Clin Sci. 1990;78:247-254. [Medline] [Order article via Infotrieve]

9. Mota de Freitas D. Alkali metal NMR. Methods Enzymol. 1993;227:78-106. [Medline] [Order article via Infotrieve]

10. Rong Q, Espanol M, Mota de Freitas D, Geraldes CFGC. ⁷Li NMR relaxation study of Li⁺ binding in human erythrocytes. Biochemistry. 1993;32:13490-13498. [Medline] [Order article via Infotrieve]

11. Mota de Freitas D, Amari L, Srinivasan C, Rong Q, Ramasamy R, Abraha A, Geraldes CFGC, Boyd MK. Competition between Li⁺ and Mg²⁺ for the phosphate groups in the human erythrocyte membrane and ATP: an NMR and fluorescence study. Biochemistry. 1994;33:4101-4110. [Medline] [Order article via Infotrieve]

12. Mota de Freitas D, Abraha A, Rong Q, Silberberg J, Whang W, Borge GF, Elenz E. Relationship between lithium ion transport and phospholipid composition in erythrocytes from bipolar patients receiving lithium carbonate. Lithium. 1994;5:29-39.

13. Canessa M, Morgan K, Goldszer R, Moore TJ, Spalvins A. Kinetic abnormalities of the red blood cell sodium-proton exchange in hypertensive patients. Hypertension. 1991;17:340-348. [Abstract/Free Full Text]

14. Canessa M, Falkner B, Hulman S. Red blood cell sodium-proton exchange in hypertensive blacks with insulin-resistant glucose disposal. Hypertension. 1993;22:204-213. [Abstract/Free Full Text]

15. Rutherford PA, Thomas TH, Laker MF, Wilkinson R. Plasma lipids affect maximum velocity not sodium affinity of human sodium-lithium countertransport: distinction from essential hypertension. Eur J Clin Invest. 1992;22:719-724. [Medline] [Order article via Infotrieve]

16. Ong RL, Cheung HC. ²³Na-NMR studies of Na⁺ interaction with human red cell membranes from normotensives and hypertensives. Biophys Chem. 1986;23:237-244. [Medline] [Order article via Infotrieve]

17. Urry DW, Trapane TL, Andrews SK, Long MM, Overbeck HW, Oparil S. NMR observation of altered sodium interaction with human erythrocyte membranes of essential hypertensives. Biochem Biophys Res Commun. 1980;96:514-521. [Medline] [Order article via Infotrieve]

18. Steck TL, Kant JA. Preparation of impermeable ghosts and inside-out vesicles from human erythrocyte membranes. Methods Enzymol. 1974;31:172-180. [Medline] [Order article via Infotrieve]

19. Bollag DM, Edelstein SJ. Protein concentration determination. In: Bollag DM, Edelstein SJ, eds. Protein Methods. New York, NY: Wiley-Liss; 1991:50-55.

20. Fanger BO. Adaptation of the Bradford protein assay to membrane-bound proteins by solubilizing in glucopyranoside detergents. Anal Biochemistry. 1987;162:11-17. [Medline] [Order article via Infotrieve]

21. Meneses P, Glonek T. High resolution ³¹P NMR of extracted phospholipids. J Lipid Res. 1988;29:679-689. [Abstract]

22. Edzes HT, Teerlink T, Van der Knaap MS, Valk J. Analysis of phospholipids in brain tissue by ³¹P NMR at different compositions of the solvent system chloroform-methanol-water. Magn Reson Med. 1992;26:46-59. [Medline] [Order article via Infotrieve]

23. James TL, Noggle JH. ²³Na nuclear magnetic resonance relaxation studies of sodium ion interaction with soluble RNA. Proc Natl Acad Sci U S A. 1969;62:644-649. [Abstract/Free Full Text]

24. Nissen H, Jacobsen JP, Horder M. Assessment of the NMR visibility of intraerythrocyte sodium by flame photometric and ion-competitive studies. Magn Reson Med. 1989;10:388-398. [Medline] [Order article via Infotrieve]

25. Corrocher R, Steinmayr M, Ruzzenente O, Brugnara C, Bertinato L, Mazzi M, Furri C, Bonfanti F, De Sandre G. Elevation of red cell sodium-lithium countertransport in hyperlipidemias. Life Sci. 1985;36:649-655. [Medline] [Order article via Infotrieve]

26. Hunt SC, Williams RR, Smith JB, Ash KO. Associations of three erythrocyte cation transport systems with plasma lipids in Utah subjects. Hypertension. 1986;8:30-36. [Abstract/Free Full Text]

27. Carr SJ, Thomas TH, Laker MF, Wilkinson R. Elevated sodium-lithium countertransport: a familial marker of hyperlipidemia and hypertension? J Hypertens. 1990;8:139-146. [Medline] [Order article via Infotrieve]

28. Engelmann B, Streich S, Schonthier UM, Richter WO, Duhm J. Changes of membrane phospholipid composition of human erythrocytes in hyperlipidemias, I: increased phosphatidylcholine and reduced sphingomyelin in patients with elevated levels of triacylglycerol-rich lipoproteins. Biochim Biophys Acta. 1992;1165:32-37. [Medline] [Order article via Infotrieve]

29. Engelmann B, Schonthier UM, Richter WO, Duhm J. Changes of membrane phospholipid composition of human erythrocytes in hyperlipidemias, II: increases in distinct molecular species of phosphatidylethanolamine and phosphatidylcholine containing arachidonic acid. Biochim Biophys Acta. 1992;1165:38-44. [Medline] [Order article via Infotrieve]

30. Engelmann B, Duhm J, Schonthier UM, Streich S, Op den Kamp JAF, Roelofsen B. Molecular species of membrane phospholipids containing arachidonic acid and linoleic acid contribute to the interindividual variability of red blood cell Na⁺-Li⁺ countertransport: in vivo and in vitro evidence. J Membr Biol. 1993;133:99-106. [Medline] [Order article via Infotrieve]

31. Preiss R, Prumke HJ, Muller E, Schmeck G, Schmidt J, Banaschak H. Sodium flux and lipid spectrum in the erythrocyte membrane in essential hypertension. Int J Clin Pharmacol Ther Toxicol. 1982;20:105-112. [Medline] [Order article via Infotrieve]

32. Sengupta N, Datta SC, Sengupta D. Platelet and erythrocyte membrane lipid and phospholipid patterns in different types of mental patients. Biochem Med. 1981;25:267-275. [Medline] [Order article via Infotrieve]

33. Kurumi Y, Adachi Y, Itoh T, Kobayashi H, Nanno T, Yamamoto T. Novel high-performance liquid chromatography for determination of membrane phospholipid composition of rat hepatocytes. Gastroenterol Jpn. 1991;26:628-632. [Medline] [Order article via Infotrieve]

34. Pearce JM, Shifman MA, Pappas AA, Komoroski RA. Analysis of phospholipids in human amniotic fluid by ³¹P NMR. Magn Reson Med. 1991;21:107-116. [Medline] [Order article via Infotrieve]

35. Spillman T, Bretaudiere JP, Cotton DB, Lynn SC. Influence of phospholipid saturation on classical thin-layer chromatographic detection methods and its effect on amniotic fluid lecithin sphingomyelin ratio. Clin Chem. 1983;29:250-255. [Abstract/Free Full Text]

36. Laurenzi M, Trevisan M. Sodium-lithium countertransport and blood pressure: the Gubbio population study. Hypertension. 1989;13:408-415. [Abstract/Free Full Text]

37. Ramasamy R, Mota de Freitas D, Bansal VK, Dorus E, Labotka RJ. Nuclear magnetic resonance studies of lithium transport in erythrocyte suspensions of hypertensives. Clin Chim Acta. 1990;188:169-176. [Medline] [Order article via Infotrieve]

38. Canessa M, Adragna N, Bize I, Connolly T, Solomon H, Williams G, Slater E, Tosteson DC. Ouabain-insensitive cation transport in red cells of normotensive and hypertensive subjects. In: Zumkley H, Losse H, eds. Intracellular Electrolytes and Arterial Hypertension. New York, NY: Georg Thieme Verlag; 1980:239-249.

39. De la Sierra A, Insa R, Compte M, Martinez-Amenos A, Sierra C, Hernandez-Herrero G, Coca A. Effect of long-term antihypertensive therapy with angiotensin converting enzyme inhibitors on red cell sodium transport. Am J Hypertens. 1995;8:622-625. [Medline] [Order article via Infotrieve]

40. Canessa M, Zerbini G, Laffel LMB. Sodium activation kinetics of red blood cell Na⁺/Li⁺ countertransport in diabetes: methodology and controversy. J Am Soc Nephrol. 1992;3:S41-S49.

41. Zerbini G, Ceolotto G, Gaboury C, Mos L, Pessina AC, Canessa M, Semplicini A. Sodium-lithium countertransport has low affinity for sodium in hyperinsulinemic hypertensives. Hypertension. 1995;25:986-993. [Abstract/Free Full Text]

42. Merchant TE, Glonek T. ³¹P NMR of tissue phospholipids: competition for Mg²⁺, Ca²⁺, Na⁺ and K⁺ cations. Lipids. 1992;27:551-559. [Medline] [Order article via Infotrieve]

43. Roux M, Bloom M. Ca²⁺, Mg²⁺, Li⁺, Na⁺, and K⁺ distribution in the head group region of binary membranes of PC and PS as seen by deuterium NMR. Biochemistry. 1990;29:7077-7089. [Medline] [Order article via Infotrieve]

44. Siegel S. Nonparametric Statistics for the Behavioral Sciences. New York, NY: McGraw-Hill Publishing Co; 1956.

45. Engelmann B, Duhm J. Effect of cholesterol and dipalmitoyl phosphatidylcholine enrichment on the kinetics of Na⁺-Li⁺ exchange of human erythrocytes. J Membr Biol. 1991;122:231-238. [Medline] [Order article via Infotrieve]

46. Ollerenshaw JD, Heagerty AM, Bing RF, Swales JD. Abnormalities of erythrocyte membrane fatty acid composition in human essential hypertension. J Hum Hypertens. 1987;1:9-12.[Medline] [Order article via Infotrieve]

This article has been cited by other articles:

				C. G. Savopoulos, A. I. Hatzitolios, N. A. Katsiki, M. Baltatzi, M. Kosmidou, N. Raikos, D. P. Mikhailidis, A. G. Ziakas, G. Kaiafa, and N. Tsesmeli Sodium-Lithium Countertransport Activity in Healthy, Dyslipidemic, and Hypertensive Individuals Angiology, January 1, 2009; 59(6): 727 - 735. [Abstract] [PDF]

				S. N. Orlov, N. C. Adragna, V. A. Adarichev, and P. Hamet Genetic and biochemical determinants of abnormal monovalent ion transport in primary hypertension Am J Physiol Cell Physiol, March 1, 1999; 276(3): C511 - C536. [Abstract] [Full Text] [PDF]

				M. Pahor, M. B. Elam, R. J. Garrison, S. B. Kritchevsky, and W. B. Applegate Emerging Noninvasive Biochemical Measures to Predict Cardiovascular Risk Arch Intern Med, February 8, 1999; 159(3): 237 - 245. [Abstract] [Full Text] [PDF]

				J. Zicha, Z. Dobesova, and J. Kunes Plasma Triglycerides and Red Cell Ion Transport Alterations in Genetically Hypertensive Rats Hypertension, September 1, 1997; 30(3): 636 - 640. [Abstract] [Full Text]

This Article Abstract Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Chi, Y. Articles by Bansal, V. K. Search for Related Content PubMed PubMed Citation Articles by Chi, Y. Articles by Bansal, V. K.