Assessing the Role of APNH, a Encoding for a Human Amiloride-Sensitive Na/H Antiporter, on the Interindividual Variation in Red Cell Na/Li Count ert ransport1

Christopher R.K. Dudley,2 Luis A. Giuffra, Anthony E.G. Rome, and Stephen T. Reeders

correction of SIC values for known covariotes. There C.R.K. Dudley, S.T. Reeders, Department of Internal was no significant association between mean SIC Medicine. Yale University School of Medicine, New- values and any of the three possible genotypes of Haven, CT the APNH locus either before or after regressing out L.A. Giuffra, ST. Reeders, Department of Human Ge- covariates (F= 0.64 and P> 0.52; F= 0.63 and P> netics, Yale University School of Medicine, New- 0.53, respectively). Linkage analysis similarly foiled Haven, CT to demonstrate a relationship between the squared A.E.G. Rome, Department of Nephrology, The Royal difference In SIC values and the identity by descent Hospital of St. Bartholomew, London, England status for APNH as well as other loci that map close to APNH (D1S57, RH, and ALPI). Taking these results (J. Am. Soc. Nephrol. 1991; 2:937-943) together, we conclude that mutations at the APNH locus are not responsible for the observed variation in SIC values. ABSTRACT Key Words: Blood pressure, intermediate phenotype, associ- The “reverse genetic” approach to essential hyper- ation study, sib pair method, linkage analysis tension Is complicated by the tact that blood pres- sure is a heterogeneous, quantitative, complex trait. One strategy is to use Intermediate phenotypes” that are not only associated with hypertension but S ince the original observation by Canessa et at. that also have a simple mode of inheritance, com- demonstrating an association between red cell patible with the action of a single gene. Red cell sodium-lIthium countertransport (SLC) and essential sodium-lithium countertransport (SIC) is one of the hypertension (1), a number of other groups have best characterized intermediate phenotypes for hy- reported similar findings (2.3). Genetic studies of pertension. The similarity in stoichiometry and kinet- countertransport levels have shown familial aggre- ics between SIC and Na/H exchange has led to gation and a higher concordance rate for monozygotic than for dizygotic twins (4.5). These findings suggest the proposal that the gene encoding the Na/H that SLC has a major heritable component and estI- antiporter (APNH) may be responsible for the individ- mates that up to 80 to 90% of the individual variance ual variance in SIC. We have tested this hypothesis in SLC is accounted for by inherited factors (5.6). by both an association study and Haseman and Complex segregation analysis has been performed to Elston’s sib pair method of linkage analysis, by using determine which of a variety of genetic models best a polymorphism at the APNH locus detected by de- explains the familial pattern of SLC values, Two such naturing gradient gel electrophoresis. Both analyti- studies provide evidence for the action of a single cal techniques were performed before and after major gene on a polygenic background (7,8). These results have important implications, not the least of which is that the same that determine SLC values may also be involved in the pathogenesis of ‘ReceIved December 18, 1990. Accepted June I, 1991. 2Correspondence to Df. C.RK. Dudley. Nuffteld Department of Clinical Medi- hypertension. The identification of such genes, cine, John Radcliffe Hospital, Headinglon, Oxford 0X3 9DU. England. therefore. may provide important insights into the l0466673/0204-0937$03.O0/0 mechanisms determining blood pressure variation Journal of the Ameflcan Society of Nephrology Copyright 0 1991 by the American Society of Nephroiogy and control.

Journal of the American Society of Nephromogy 937 APNH and Na/Li Countertransporf

The physIological role of the transport system that 7.4) at 4#{176}C.SLC activity was determined by measur- mediates In vitro erythrocyte SLC is not established. ing, in parallel and in duplicate, lithium efflux from The similarity In kinetics and stoichiometry between loaded cells into a sodium-contaIning medium (150 SLC and Na/H countertransport has led to the pro- mM NaC1, 10 mM glucose, 0.1 mM ouabain, and 10 posal that red cell SLC is carried out by the ubiqui- mM Tris-MOPS [pH 7.4J at 37#{176}C)and into a sodium- tously distributed Na/H antiporter (9,10). In sup- free medium (110 mM magnesium chloride, 10 mM port of this argument is the demonstration that both glucose, 0.1 mM ouabain, and 10 mM Tris-MOPS [pH SLC and Na/H exchange in sarcolemmal vesicles 7.4] at 37#{176}C)incubated at 37#{176}C.After 0, 20, 40 and from bovine artery are mediated by a single ethyli- 60 mm of incubation, duplicate samples were taken sopropyl amiloride-sensitive cation exchanger with from duplicate tubes, cooled quickly to 4#{176}C,and cen- affinity for sodium, lithium, and hydrogen ions (11). trifuged. Furthermore, Na/H antiport activity has been re- The lithium concentration in the supernatant was ported to be elevated in leukocytes and platelets from determined by atomic-absorption spectrophotome- patients with hypertension (12.13) as well as in ge- try. The hematocrit of each efflux medium was meas- netically hypertensive rats (14,15). ured at the end of the experiment by using a Coulter A flbroblast Na7H antiporter cDNA has been counter. The lithium efflux rates in the sodium-free cloned by complementation (16) and mapped by in and sodium-containing media were taken as the situ hybridization to lp (17). Linkage slopes of the linear regression lines calculated with studies have placed the genomic locus for this Na/ lithium efflux as a function of time. Correlation coef- H antiporter (designated APNH) close to the rhesus ficients greater than 0.98 were considered accepta- locus (RH) (18,19). It has been proposed that the ble. Net SLC activity was calculated by subtracting APNH gene product carries out SLC and is responsi- the lithium efflux rate in the sodium-free medium ble for the variation observed in SLC values (20,21). from that In the sodium-containing medium and was We have tested this hypothesis by association studies expressed as millimoles of lithium released from one and by the sib pair method of linkage analysis by liter of red cells per hour. The coefficient of variation using members from six large pedigrees in whom of repeated assays was 10%. SLC was measured. Because the power of the sib pair method is limited by a requirement for a highly in- APNH Polymorphism formative locus, we have also performed linkage analysis between SLC and other polymorphic loci A polymorphism in the APNH gene was detected by (namely D1S57, RH, and ALPL, the gene for alkaline denaturing gradient gel electrophoresis (DGGE). A ), which map close to APNH. 376-base-pair product of the polymerase chain re- action (PCR). corresponding to nucleotides -25 .e METHODS +351 of the coding sequence for APNH, (where + 1 is the first nucleotide of the translation initiation co- Subjects don) (16), was generated by using genomic DNA from As part of a broader study on the genetic determi- each individual as a template. This product was then nants of blood pressure, 190 individuals from six subjected to DOGE on a 35 to 65% gradient as de- large English kindreds were evaluated. Subjects were scribed previously (18). seen by a single investigator, weight was measured DGCIE is a powerful technique allowing two DNA in a standardized way, and blood was drawn for DNA fragments differing by as little as a single base sub- analysis and assay of SLC activity. Thirty-six sub- stitution to be physically separated from each other jects were receiving one or more antihypertensive (23). The method can be further enhanced by allow- agents at the time of the investigation. ing denatured PCR products to reanneal. creating heteroduplex DNA in regions where sequence poly- SIC Activity morphism exists as shown (Figure 1). PCR products SLC was measured as the rate of sodium-dependent from individuals homozygous for each allele were lithium efflux from lithium loaded red blood cells as subcloned into pBSII (Stratagene) and sequenced. Se- described by Canessa et at. (1) and as modified by quence comparison revealed that a single base-pair Smith et at. (22). In brief, washed red cells were substitution of cytosine for thymine at nucleotide 69 suspended at a hematocrit of approximately 20% In distinguishes two alleles, designated Cl and C2 (Fig- a loading solution consisting of 150 mM lithium chlo- ure 2). Although this substitution does not alter the ride, 10 mM glucose, and 10 mM Tris-morpholine- amino acid sequence, It is responsible for the poly- propanesulfonic acid (MOPS) (pH 7.4) and were In- morphism detected by DOGE. The cytosine at posi- cubated in a shaking water bath at 37#{176}Cfor 3 h. tIon 69 in the C2 allele confers greater stability on After loading, external lithium was removed by five the DNA duplex under denaturing conditions and washings with 75 mM magnesium chloride, 85 mM allows the DNA fragment to migrate further down sucrose, 10 mM glucose, and 10 mM Tris-MOPS (pH the denaturing gradient before “melting” occurs.

938 Volume 2 ‘ Number 4 #{149}1991 Dudley et al

Statistical Methods The possible genetic role of APNH in SLC variation was tested in two ways. Association Studies. Mean values for SLC were compared by analysis of variance (ANOVA) across the three genotypes (C1C1, C1C2, and C2C2) that 23 4 56 7 could be detected at the APNH locus by using the DGGE polymorphism described above. Linkage Analysis. The genotypes of members from all sibships together with the parental genotypes were identified for each locus examined. All possible sib pairs were formed, and the Identity by descent (i.b.d.) status at each locus (that is, the number of 4- identical alleles shared by each pair: 0, 1, or 2), was evaluated for each pair by inspection. In situations Heteroduplexes where either 0 or 1 allele were equally likely to be shared, the average number was scored I.e., 0.5. Homoduplexes Similarly, where 1 or 2 alleles i.b.d. were equally likely, 1.5 was scored. However, in cases where 0, 1, or 2 alleles are all equally likely, the use of 1 as an average value biases the data against linkage. Con- sequently, all cases where this occurred were ex- cluded from the data set. Linkage was tested both with a regression analysis as well as with the non- Figure 1. Alleles at the APNH locus detected by DGGE were parametric Spearman’s rank correlation coefficient named Cl and C2. Heat-denatured PCR products, allowed to reanneal before loading, create heteroduplexes in re- (24). In sibships with more than two individuals, all gions where sequence polymorphism exist(lanes 4 and 6). possible sib pairs were formed and treated as inde- Homoduplexes for Cl (lanes 1, 2, 5, and 7) and C2 (lane pendent observations in the analyses. This has been 3) are also shown. shown to have no effect on either the power or the significance of the tests for linkage (25). For each pair, the squared difference in SLC values was re- gressed on i.b.d. status at each locus (24). For a Cl Allele C2 Allele quantitative trait, linkage to a locus is suspected if the greater number of alleles shared Is accompanied G AT C G ATC by smaller differences between sib values for the trait under investigation (24). This relationship be- tween the proportion of alleles i.b.d. and the squared difference In SLC values for each sib pair is also presented by using Spearman’s rank correlation G coefficient. A significantly negative correlation im- plies the presence of linkage. G SLC values were also analyzed after regressing out concomitant variables (weight and number of anti- hypertensive medications) and after performing a square root transformation on the residual values to Figure 2. Sequence analysis of PCR products from individ- achieve normality. This is a very important step in uals homozygous for the Cl and C2 alleles, respectively. the analysis, because continuous variables can pres- reveal that the polymorphism detected by DGGE is due to ent marked skewness because of covariates or lack a cytosine-for-thymine substitution as indicated by the or- of normality. This skewness can mimic a major gene row. effect and artificially inflate the evidence for linkage. Therefore, results will be presented for both the raw data and for the corrected and normalized residual values. D1S57, RH, and ALPL Polymorphisms Rhesus phenotypes and D1557 and ALPL restric- RESULTS tion fragment length polymorphisms were typed in The general descriptive statistics for the individ- all individuals as described elsewhere (18). uals making up the data set are given in Table 1. The

Journal of the American Society of Nephrology 939 APNH and Na”/Li Countertransport

TABLE 1. General descriptive statistics of the population studied#{176}

Male Female Total

Age (yrs) 41.9 40.6 41.3 Na/LI countertransport 0.270 ± 0.094 0.254 ± 0.105 0.263 ± 0.099 (mmol/L/h) Weight (kg) 80.2 ± 12.3 67.9 ± 12.8 74.9 ± 13.9 Body mass Index (Kg/rn2) 26.1 ± 3.8 26.1 ± 4.4 26.1 ± 4.0

“Mean values ± SD are shown.

TABLE 2. Mean SIC values (±SD) compared by ANOVA across genotypes for APNH before and after correction of normalized residuals for covoriates

Antiporter genotype ANOVA CICI CIC2 C2C2

N 73 87 13 Mean Na/Ll countertransport 0.266 ± 0,1 0.255 ± 0.083 0.283 ± 0.088 F= 0.64

(mrnol/L/h) P = 0.52 N 66 74 11 Mean normalized residual Na/ 0.428 ± 0.104 0.412 ± 0.109 0.446 ± 0.102 F= 0.69

LI countertransport (rnrnol/L/ P = 0.50 h)

average value of SLC for the group as a whole is TABLE 3. Number of informative sib pairs and the similar to that reported In other populations (6.26). mean value of allele sharing for each locus In our data set, this value is strongly correlated with examined by using both uncorrected and weight (r = 0.286; N = 167; P = 0.0002) and is related normalized residual SIC values to the number of antihypertensive medications pre- With Uncorrected With Corrected scribed (F = 3.84; P = 0.0 106). There was no signif- SIC Values SIC Values icant association between mean SLC values and any Locus of the three possible genotypes of the APNH locus N mean i.b.d. N mean i.b.d. before correction for covariates was made (F = 0.64; P> 0.52) (Table 2). After regressing out the covariates APNH 40 0.8875 40 0.8875 and normalizing the residuals, the association be- D1S57 116 1.1336 97 1.1494 tween mean corrected SLC values and genotype (F = RH 81 1.0864 72 1.1041 0.63; P> 0.53) remained insignificant (Table 2). The ALPI 70 0.9857 65 0.9846 number of informative sib pairs used for each locus examined by using SLC values, both before correction and after normalizing the residuals, is shown in Table 3. together with the mean number of alleles DISCUSSION shared. Linkage analysis failed to demonstrate an effect of the I.b.d. status at each locus when re- We have tested the role of a Na7H’ antiporter, gressed on the squared difference in SLC values APNH, in SLC variation by two approaches: associa- (Table 4). After correcting for covariates and normal- tion studies and linkage analysis. The success of izing the residuals, the slopes of the regression lines association studies based on “silent” polymorphisms were still not significant (Table 5). Likewise, the use that do not change the codon sense, such as the of Spearman’s rank correlation coefficient failed to DGGE polymorphism described here, relies on the identify a significant relationship between the num- phenomenon of linkage disequilibrium. That is, the ber of alleles i.b.d. at each locus and the mean mutation which causes the altered phenotype (high squared difference in SLC values (both uncorrected or low SLC values) is presumed to lie so close to the and after regressing out covariates and normalizing mutation causing the DGGE polymorphism that they residual values) (Table 6). occur together on the saxne chromosome more often

940 Volume 2’ Number 4’ 1991 Dudley et al

TABLE 4. Number of alleles identical by descent and regression analysis on uncorrected SIC valuesa

No. of Alleles i.b.d. Regression Analysis Locus 0 0.5 1 1.5 2 T(9=0) Pr>ITI

APNH 7 13 7 8 5 0.0027 1.08 0.2861 0.93 0.49 0.64 0.89 1.42 D1S57 15 8 53 11 29 -0.0045 -1.30 0.1959 1.89 0.66 2.01 0.62 1.03 RH 12 11 28 11 19 -0.0015 -0.54 0.5874 1.64 0.60 1.34 0.90 1.08 ALPL 9 15 21 19 6 0.0037 1.07 0.2904 0.54 1.16 0.90 1.47 1.21

“For each locus, the number of sib pairs is shown above the mean squared difference In SLC x102.

TABLE 5. Number of alleles identical by descent and regression analysis on normalized residual SIC values0

No. of Alleles i.b.d. Regression Analysis Locus 0 0.5 1 1.5 2 T(it=0) Pr>ITI

APNH 7 13 7 8 5 0.0019 0.60 0.5489 1.05 0.46 1.87 0.98 1.06 D1S57 13 6 42 11 25 -0.0083 -1.93 0.0571 3.12 0.53 2.22 0.81 1.22 RH 10 11 23 10 18 -0.0007 -0.20 0.8404 1.95 0.63 2.13 0.83 1.55 ALPI 9 14 18 18 6 0.0070 1.19 0.2403 0.48 1.06 1.49 2.58 0.55

“For each locus, the number of sib pairs Is shown above the mean squared difference ri SLC x102.

TABLE 6. Spearman’s rank correlation coefficient for the relationship between the number of alleles i.b.d. at each locus and the mean squared difference in SLC values, both uncorrected and after regressing out covariates and normalizing residual values

Uncorrected SIC Values Corrected SIC Values Locus N r P>’r N r P>r

APNH 40 0.175 0.2796 40 0.126 0.4378 D1S57 116 -0.119 0.2030 97 -0.130 0.2025 RH 81 0.002 0.9809 72 -0.021 0.8554 ALPI 70 0.225 0.0602 65 0.138 0.2702

than would be predicted from the population fre- the two are not in linkage disequilibrium, no associ- quency of each alone. Ideally. such a study should ation will be detected. Include only unrelated individuals to avoid the con- In contrast, Haseman and Elston’s sib pair method founding effect of counting the same allele more than of linkage analysis does not rely on assumptions of once within families. Although this weakens the use- disequilibrium. When a sib pair shares two alleles fulness of an association study with related individ- i,b.d. at the polymorphic locus of APNH detected by uals. it still has a value as a preliminary analysis. DGGE, they are also likely to share large regions of Failure to detect an association In this type of study DNA on either side of the polymorphic site, including must be Interpreted with caution. The polymorphism the whole gene and regulatory region. Genetic linkage and the mutation causing the altered phenotype may with SLC can therefore be carried out. The failure to lie very close together (i.e., genetically linked), but, if demonstrate any Influence of APNH on SLC variation

Journal of the American Society of Nephrology 941 APNH and Na/Li Countertransport

when linkage analysis is performed confirms the REFERENCES findings of the association study. Together, these two 1. Canessa M, Adragna N, Solomon H, Connolly methods show no effect of APNH on SLC variation. T, Tosteson DC: Increased sodium-lithium coun- The led score method is a powerful technique for tertransport in red cells of patients with essen- linkage analysis because It uses all information in a tial hypertension. N Engi J Med 1980;302:772- pedigree. However, In this study, we have favored the 776. 2. Turner ST, Boerwinkle E, Johnson M, Richel- use of the sib pair method for linkage analysis. The son E, Sing CF: Sodium-lithium countertrans- Id score is a parametric statistic dependent on the port in ambulatory hypertensive and normoten- mode of Inheritance of the trait being studied. Al- sive patients. Hypertension 1 987;9:24-34. though an autosomal recessive model has been pro- 3. Wilhams RR, Hunt SC, Kuida H, Smith JB, Ash KO: Sodium-lithium countertransport in eryth- posed for SLC by two groups, this is based on a rocytes of hypertension prone families in Utah. number of assumptions (7,8). Furthermore, evidence Am J Epidemiol 1983;1 18:338-344. for such a major gene underlying SLC has not been 4. Kagamimori 5, Naruse Y, Takata M, Fujita T, demonstrated by a number of other investigators Watanabe M: Familial aggregation of red blood cell cation transport systems in Japanese fami- (6,27). The sib pair method of linkage analysis is lies. Am J Epidemiol 1985;122:386-390. independent of the mode of Inheritance of SLC val- 5. Lewitter FW, Canessa M: Red cell transport ues. Under these circumstances, therefore, It is a studies in adult twins. Am J Hum Genet better choIce for analysis. A limitation of the sib pair 1 985;36: 1 72S. method is the requirement for a highly polymorphic 6. Dadone MM, Hasstedt SJ, Hunt SC, Smith JB, Ash KO, Williams RR: Genetic analysis of so- locus in order to get maximum information on allele dium-lithium countertransport in 10 hyperten- sharing to avoid ambiguity. The DGGE polymorphism sIon prone kindreds. Am J Med Genet used here has a polymorphism information content 1984; 17:565-577. of 0.37. We have therefore also performed sib pair 7. Hasstedt SJ, Wu LL, Ash KO, Kuida H, Wil- liams RR: Hypertension and sodium-lithium linkage analysis between SLC and other highly pol- countertransport in Utah pedigrees: Evidence for ymorphic loci which map close to APNH, namely, major-locus inheritance. Am J Hum Genet D1S57, RH, and ALPL. The results with these loci 1 98;63: 14-22. also fail to show any significant effect on SLC. 8. Motuisky AG, Burke W, Billings PR, Ward RH: Although our results exclude APNH as the locus Hypertension and the genetics of red cell mem- brane abnormalities. Ciba Found Symp responsible for the variation in SLC values, they can 1987; 130: 150-166. not exclude it as the locus encoding a structural 9. Aronson PS: Red cell sodium-lithium counter protein performing erythrocyte. Na/LF’ exchange. transport and essential hypertension. N Engl J Furthermore, the difference in kinetics between the Med 1982;307:317. apical and basolateral Na7H antiporter in cultured 10. Funder J, Wieth JO, Jensen HE, Ibsen KY: The sodium/lithium exchange mechanism in essen- renal tubular cells (LLPCK 1) (28) suggests that either tial hypertension. Is it a sodium/proton exchan- posttranscriptional modifications of APNH occur or ger? In: Villarreal H, Sambhi MP, eds. Topics in other loci coding for physiologically distinct Na/H Pathophysiology of Hypertension. Boston: Mar- antiporters may exist. When identified, these loci will tinus Nijhoff: 1984:147-161. 11. Kahn AM, Allen JC, Cragoe EJ, Shelat H: So- also need to be tested as candidate genes for SLC. In dium-lithium exchange and sodium-proton ex- addition, the importance of the phosphoinosotide sig- change are mediated by the same transport sys- naling pathway in modifying Na/H antlporter activ- tem in sarcolemmal veslcles from bovine supe- ity is beginning to emerge (29). If erythrocyte Na7LI rior mesenterIc artery. Circ Res 1989;65:818- countertransport proves to be mediated by a Na/H 828. 12. Livine A, Balfe JW, Veitch R, Marquez-Julio A, antiporter, then genetic differences in the phosphol- Grinstein 5, Rothstein A: Increased platelet nositide regulating system may account for the ob- Na7H exchange rate in essential hypertension: served variation in SLC. Application of a novel test. Lancet 1987; 1:533- 536. 13. Ng LL, Dudley C, Bomford J, Hawley D: Leu- cocyte intracellular pH and Na/H antlport ac- tivity in human hypertension. J Hypertens ACKNOWLEDGMENTS 1 989;7:47 1-475. 14. Berk BC, Vallega G, Muslin AJ, Gordon HM, Canessa M, Alexander RW: Spontaneously hy- We thank the members of the six Oxfordshire families who partici- pertensive rat vascular smooth muscle cells in pated in this study. The expert technical assistance of Kay Yates culture exhibit increased growth and Na/H ex- and the secretarial skills of Sidney McGaughey are gratefully ac- change. J C1In Invest 1989;83:822-829. knowledged. We thank Professor J.G.G. Ledingham for his support 15. Ek TP, Deth RC: Elevated phosphollpase C and and encouragement. C.R.K.D. is supported by grant DK40703 from Na/H exchange activity in spontaneously hy- the NIH. L.A.G. is a fellow of the MacArthur Foundation Mental pertensive rats. Hypertension 1988; 12:331 - Health Research Network 1. S.T.R. is an assistant investigator with 332. the Howard Hughes Medical Institute. 16. Sardet C, Franchi A, Pouyssegur J: Molecular

942 Volume 2 ‘ Number 4 1991 Dudley et al

cloning, primary structure and expression of the single base changes In DNA: cleav- human growth factor-activatable Na/H anti- age and denaturing gradient gel electrophoresis. porter. Cell 1989;56:27l-280. In: Davies K, ed. Genome Analysis: A Practical 17. Mattel MG, Sardet C, Franchi A, Pousegur J: Approach. Washington, DC: IRL Press; The human amiloride-sensltive Na /H’ anti- 1988:95-140. porter: Localization to chromosome 1 by in situ 24. Haseman JK, Elston RC: The investigation of hybridization. Cytogenet Cell Genet l988;48:6- linkage between a quantitative trait and a marker locus. Behav Genet 1972;2:3-19. 18. Dudley CRK, Giuffra LA, Tlppett P, Kldd KK, 25. Amos CL, Elston RC, Wilson AF, Bailey-Wilson Reeders ST: The Na7H antiporter: A “melt” JE: A more powerful robust sib-pair test of link- polymorphism allows regional mapping to the age for quantitative traits. Gen Epidemiol short arm of chromosome 1. Hum Genet 1 989;6:435-449. 1 990;86:79-83. 26. Weder AB, Schork NJ: Mixture analysis of 19. Llfton RP, Sardet C, Pouyssegur J, Lalouel J- erythrocyte lithium-sodium countertransport M: Cloning of the human genomic amilorlde- and blood pressure. HypertensIon 1989; 13:145- sensitive Na/H antiporter gene, identification 150. of genetic polymorphisms, and localization on 27. Boerwinkle E, Turner ST. Weinshilboum R, the genetic map of chromosome lp. Genomics Johnson M, Richelson E, Sing CF: Analysis of 1990;7:131-l35. the distribution of erythrocyte sodium-lithium 20. Mahnensmith RL, Aronson PS: The plasma countertransport in a sample of representatives membrane sodium-hydrogen exchanger and its of the general population. Gen Epidemiol role in physiological and pathophysiological 1986;3:365-378. processes. Circ Res 1985;56:773-788. 28. Haggerty JG, Agarwal N, Reilly RF, Adelberg 21. Canessa M, Morgan K, Semplicini A: Genetic EA, Slayman CW: Pharmacologically different difference in lithium-sodium exchange and reg- Na/H antiporters on the apical and basolateral ulation of the sodium-hydrogen exchanger in surfaces of cultured porcine kidney cells (LLC- essential hypertension. J Cardiovasc Pharmacol PK1). Proc Nati Acad Sd USA 1988;85:6797- 1988; 1 2(suppl 3):S92-S98. 6801. 22. SmIth JB, 1rice AL, Williams RR, et at.: A 29. Owen NE, Knapik J, Strebel F, Tarpley WG, reproducible sodium-lithium countertransport Gorman RR: Regulation of Na4/H’ exchange in assay: The outcome of changing key laborato normal NIH-3T3 cells and in NIH-3T3 cells ex- parameters. Clin Chim Acta r982; 122:327-33 pressing the ras oncogene. Am J Physiol 23. Myers RM, Sheffield VC, Cox DR: Detection of 1 989;256:C756-C763.

Journal of the American Society of Nephrology 943