Distinct regulatory effects of the Na,K-ATPase gamma subunit

The two variants of the gamma subunit of the rat renal sodium pump, gamma(a) and gamma(b), have similar effects on the Na,K-ATPase. Both increase the affinity for ATP due to a shift in the enzyme's E(1) <--> E(2) conformational equilibrium toward E(1). In addition, both increase K(+) antagonism of cytoplasmic Na(+) activation. To gain insight into the structural basis for these distinct effects, extramembranous N-terminal and C-terminal mutants of gamma were expressed in rat alpha1-transfected HeLa cells. At the N terminus, the variant-distinct region was deleted (gammaNDelta7) or replaced by alanine residues (gammaN7A). At the C terminus, four (gamma(a)CDelta4) or ten (gamma(a)CDelta10) residues were deleted. None of these mutations abrogates the K(+)/Na(+) antagonism as evidenced in a similar increase in K'(Na) seen at high (100 mm) K(+) concentration. In contrast, the C-terminal as well as N-terminal deletions (gammaNDelta7, gamma(a)CDelta4, and gamma(a)CDelta10) abolished the decrease in K'(ATP) seen with wild-type gamma(a) or gamma(b). It is concluded that different regions of the gamma chain mediate the distinct functional effects of gamma, and the effects can be long-range. In the transmembrane region, the impact of G41R replacement was analyzed since this mutation is associated with autosomal dominant renal Mg(2+)-wasting in man (Meij, I. C., Koenderink, J. B., van Bokhoven, H., Assink, K. F. H., Groenestege, W. T., de Pont, J. J. H. H. M., Bindels, R. J. M., Monnens, L. A. H., Van den Heuvel, L. P. W. J., and Knoers, N. V. A. M. (2000) Nat. Genet. 26, 265-266). The results show that Gly-41 --> Arg prevents trafficking of gamma but not alphabeta pumps to the cell surface and abrogates functional effects of gamma on alphabeta pumps. These findings underscore a potentially important role of gamma in affecting solute transport, in this instance Mg(2+) reabsorption, consequent to its primary effect on the sodium pump.     

Expression and functional role of the gamma subunit of the Na, K-ATPase in mammalian cells

The functional role of the gamma subunit of the Na,K-ATPase was studied using rat gamma cDNA-transfected HEK-293 cells and an antiserum (gammaC33) specific for gamma. Although the sequence for gamma was verified and shown to be larger (7237 Da) than first reported, it still comprises a single initiator methionine despite the expression of a gammaC33-reactive doublet on immunoblots. Kinetic analysis of the enzyme of transfected compared with control cells and of gammaC33-treated kidney pumps shows that gamma regulates the apparent affinity for ATP. Thus, gamma-transfected cells have a decreased K'ATP as shown in measurements of (i) K'ATP of Na,K-ATPase activity and (ii) K+ inhibition of Na-ATPase at 1 microM ATP. Consistent with the behavior of gamma-transfected cells, gammaC33 pretreatment increases K'ATP of the kidney enzyme and K+ inhibition (1 microM ATP) of both kidney and gamma-transfected cells. These results are consistent with previous findings that an antiserum raised against the pig gamma subunit stabilizes the E2(K) form of the enzyme (Therien, A. G., Goldshleger, R., Karlish, S. J., and Blostein, R. (1997) J. Biol. Chem. 272, 32628-32634). Overall, our data demonstrate that gamma is a tissue (kidney)-specific regulator of the Na,K-ATPase that can increase the apparent affinity of the enzyme for ATP in a manner that is reversible by anti-gamma antiserum.     

Differential regulation of renal Na,K-ATPase by splice variants of the gamma subunit.

Sodium and potassium-exchanging adenosine triphosphatase (Na,K-ATPase) in the kidney is associated with the gamma subunit (gamma, FXYD2), a single-span membrane protein that modulates ATPase properties. Rat and human gamma occur in two splice variants, gamma(a) and gamma(b), with different N termini. Here we investigated their structural heterogeneity and functional effects on Na,K-ATPase properties. Both forms were post-translationally modified during in vitro translation with microsomes, indicating that there are four possible forms of gamma. Site-directed mutagenesis revealed Thr(2) and Ser(5) as potential sites for post-translational modification. Similar modification can occur in cells, with consequences for Na,K-ATPase properties. We showed previously that stable transfection of gamma(a) into NRK-52E cells resulted in reduction of apparent affinities for Na(+) and K(+). Individual clones differed in gamma post-translational modification, however, and the effect on Na(+) affinity was absent in clones with full modification. Here, transfection of gamma(b) also resulted in clones with or without post-translational modification. Both groups showed a reduction in Na(+) affinity, but modification was required for the effect on K(+) affinity. There were minor increases in ATP affinity. The physiological importance of the reduction in Na(+) affinity was shown by the slower growth of gamma(a), gamma(b), and gamma(b') transfectants in culture. The differential influence of the four structural variants of gamma on affinities of the Na,K-ATPase for Na(+) and K(+), together with our previous finding of different distributions of gamma(a) and gamma(b) along the rat nephron, suggests a highly specific mode of regulation of sodium pump properties in kidney.     

The gamma subunit modulates Na(+) and K(+) affinity of the renal Na,K-ATPase

The Na(+),K(+)-ATPase catalyzes the active transport of ions. It has two necessary subunits, alpha and beta, but in kidney it is also associated with a 7.4-kDa protein, the gamma subunit. Stable transfection was used to determine the effect of gamma on Na, K-ATPase properties. When isolated from either kidney or transfected cells, alphabetagamma had lower affinities for both Na(+) and K(+) than alphabeta. A post-translational modification of gamma selectively eliminated the effect on Na(+) affinity, suggesting three configurations (alphabeta, alphabetagamma, and alphabetagamma*) conferring different stable properties to Na, K-ATPase. In the nephron, segment-specific differences in Na(+) affinity have been reported that cannot be explained by the known alpha and beta subunit isoforms of Na,K-ATPase. Immunofluorescence was used to detect gamma in rat renal cortex. Cortical ascending limb and some cortical collecting tubules lacked gamma, correlating with higher Na(+) affinities in those segments reported in the literature. Selective expression in different segments of the nephron is consistent with a modulatory role for the gamma subunit in renal physiology.     

The fourth transmembrane segment of the Na,K-ATPase alpha subunit: a systematic mutagenesis study

The Na,K-ATPase is a major ion-motive ATPase of the P-type family responsible for many aspects of cellular homeostasis. To determine the structure of the pathway for cations across the transmembrane portion of the Na,K-ATPase, we mutated 24 residues of the fourth transmembrane segment into cysteine and studied their function and accessibility by exposure to the sulfhydryl reagent 2-aminoethyl-methanethiosulfonate. Accessibility was also examined after treatment with palytoxin, which transforms the Na,K-pump into a cation channel. Of the 24 tested cysteine mutants, seven had no or a much reduced transport function. In particular cysteine mutants of the highly conserved "PEG" motif had a strongly reduced activity. However, most of the non-functional mutants could still be transformed by palytoxin as well as all of the functional mutants. Accessibility, determined as a 2-aminoethyl-methanethiosulfonate-induced reduction of the transport activity or as inhibition of the membrane conductance after palytoxin treatment, was observed for the following positions: Phe(323), Ile(322), Gly(326), Ala(330), Pro(333), Glu(334), and Gly(335). In accordance with a structural model of the Na,K-ATPase obtained by homology modeling with the two published structures of sarcoplasmic and endoplasmic reticulum calcium ATPase (Protein Data Bank codes 1EUL and 1IWO), the results suggest the presence of a cation pathway along the side of the fourth transmembrane segment that faces the space between transmembrane segments 5 and 6. The phenylalanine residue in position 323 has a critical position at the outer mouth of the cation pathway. The residues thought to form the cation binding site II ((333)PEGL) are also part of the accessible wall of the cation pathway opened by palytoxin through the Na,K-pump.     

Functional role and immunocytochemical localization of the gamma a and gamma b forms of the Na,K-ATPase gamma subunit

The gamma subunit of the Na,K-ATPase is a member of the FXYD family of type 2 transmembrane proteins that probably function as regulators of ion transport. Rat gamma is present primarily in the kidney as two main splice variants, gamma(a) and gamma(b), which differ only at their extracellular N termini (TELSANH and MDRWYL, respectively; Kuster, B., Shainskaya, A., Pu, H. X., Goldshleger, R., Blostein, R., Mann, M., and Karlish, S. J. D. (2000) J. Biol. Chem. 275, 18441-18446). Expression in cultured cells indicates that both variants affect catalytic properties, without a detectable difference between gamma(a) and gamma(b). At least two singular effects are seen, irrespective of whether the variants are expressed in HeLa or rat alpha1-transfected HeLa cells, i.e. (i) an increase in apparent affinity for ATP, probably secondary to a left shift in E(1) <--> E(2) conformational equilibrium and (ii) an increase in K(+) antagonism of cytoplasmic Na(+) activation. Antibodies against the C terminus common to both variants (anti-gamma) abrogate the first effect but not the second. In contrast, gamma(a) and gamma(b) show differences in their localization along the kidney tubule. Using anti-gamma (C-terminal) and antibodies to the rat alpha subunit as well as antibodies to identify cell types, double immunofluorescence showed gamma in the basolateral membrane of several tubular segments. Highest expression is in the medullary portion of the thick ascending limb (TAL), which contains both gamma(a) and gamma(b). In fact, TAL is the only positive tubular segment in the medulla. In the cortex, most tubules express gamma but at lower levels. Antibodies specific for gamma(a) and gamma(b) showed differences in their cortical location; gamma(a) is specific for cells in the macula densa and principal cells of the cortical collecting duct but not cortical TAL. In contrast, gamma(b) but not gamma(a) is present in the cortical TAL only. Thus, the importance of gamma(a) and gamma(b) may be related to their partially overlapping but distinct expression patterns and tissue-specific functions of the pump that these serve.     

Cysteine-scanning mutagenesis study of the sixth transmembrane segment of the Na,K-ATPase alpha subunit

The accessibility of the residues of the sixth transmembrane segment (TM) of the Bufo marinus Na,K-ATPase alpha subunit was explored by cysteine scanning mutagenesis. Methanethiosulfonate reagents reached only the two most extracellular positions (T803, D804) in the native conformation of the Na,K-pump. Palytoxin induced a conductance in all mutants, including D811C, T814C and D815C which showed no active electrogenic transport. After palytoxin treatment, four additional positions (V805, L808, D811 and M816) became accessible to the sulfhydryl reagent. We conclude that one side of the sixth TM helix forms a wall of the palytoxin-induced channel pore and, probably, of the cation pathway from the extracellular side to one of their binding sites.     

Structural and functional features of the transmembrane domain of the Na,K-ATPase beta subunit revealed by tryptophan scanning

In oligomeric P2-ATPases such as Na,K- and H,K-ATPases, beta subunits play a fundamental role in the structural and functional maturation of the catalytic alpha subunit. In the present study we performed a tryptophan scanning analysis on the transmembrane alpha-helix of the Na,K-ATPase beta1 subunit to investigate its role in the stabilization of the alpha subunit, the endoplasmic reticulum exit of alpha-beta complexes, and the acquisition of functional properties of the Na,K-ATPase. Single or multiple tryptophan substitutions in the beta subunits transmembrane domain had no significant effect on the structural maturation of alpha subunits expressed in Xenopus oocytes nor on the level of expression of functional Na,K pumps at the cell surface. Furthermore, tryptophan substitutions in regions of the transmembrane alpha-helix containing two GXXXG transmembrane helix interaction motifs or a cysteine residue, which can be cross-linked to transmembrane helix M8 of the alpha subunit, had no effect on the apparent K(+) affinity of Na,K-ATPase. On the other hand, substitutions by tryptophan, serine, alanine, or cysteine, but not by phenylalanine of two highly conserved tyrosine residues, Tyr(40) and Tyr(44), on another face of the transmembrane helix, perturb the transport kinetics of Na,K pumps in an additive way. These results indicate that at least two faces of the beta subunits transmembrane helix contribute to inter- or intrasubunit interactions and that two tyrosine residues aligned in the beta subunits transmembrane alpha-helix are determinants of intrinsic transport characteristics of Na,K-ATPase.     

Structure-function relations of interactions between Na,K-ATPase,the gamma subunit,and corticosteroid hormone-induced factor

Corticosteroid hormone-induced factor (CHIF) and the gamma subunit of the Na,K-ATPase (gamma) are two members of the FXYD family whose function has been elucidated recently. CHIF and gamma interact with the Na+ pump and alter its kinetic properties, in different ways, which appear to serve their specific physiological roles. Although functional interactions with the Na,K-ATPase have been clearly demonstrated, it is not known which domains and which residues interact with the alpha and/or beta subunits and affect the pump kinetics. The current study provides the first systematic analysis of structure-function relations of CHIF and gamma. It is demonstrated that the stability of detergent-solubilized complexes of CHIF and gamma with alpha and/or beta subunits is determined by the trans-membrane segments, especially three residues that may be involved in hydrophobic interactions. The transmembrane segments also determine the opposite effects of CHIF and gamma on the Na+ affinity of the pump, but the amino acids involved in this functional effect are different from those responsible for stable interactions with alpha.     

Turnover of the catalytic subunit of Na,K-ATPase in HTC cells

A polyclonal antibody to the catalytic subunit of rat kidney Na,K-ATPase has been raised in rabbits and used to analyze the turnover of the subunit in the rat hepatoma cell line HTC. It had been shown previously (Baumann, H., and Doyle, D. (1978) J. Biol. Chem. 253, 4408-4418) that the membrane proteins of these cells displayed multicomponent turnover kinetics, the minority of the surface proteins turning over with a half-time of about 20 h and the remainder with a half-time of about 100 h. That the antibody precipitated both the alpha (catalytic) and beta (glycosylated) subunits of the Na,K-ATPase from Triton extracts of HTC cells could be demonstrated following metabolic labeling of the cells with either [3H]leucine or a mixture of [3H] mannose and [3H]fucose, but following labeling with [35S]methionine radioactivity was found only in the alpha subunit of the precipitates. Incorporation of [35S]methionine into the alpha subunit could be detected 2 min after addition of the isotope to the cell suspension. Then and at all times thereafter the label was recoverable only from the particulate fraction of a 150,000 X g 60-min centrifugation; no labeled alpha subunit was ever detected in the supernatant fraction. By quantitative densitometry of radioautographs of sodium dodecyl sulfate-polyacrylamide gels of labeled antibody precipitates, it could be shown in pulse-chase experiments that the specific activity of the alpha subunit remained unchanged for 3-4 h (transit time) after the pulse was initiated and that the activity subsequently decayed exponentially with a half-time of 18 h. In a population growing with a generation time (tG) of 33 h, this decay corresponds to a turnover rate constant of 0.49/tG. The catalytic subunit is among those membrane proteins with a rapid turnover rate.     

Residues within transmembrane domains 4 and 6 of the Na,K-ATPase alpha subunit are important for Na+ selectivity

The Na,K- and H,K-ATPases are plasma membrane enzymes responsible for the active exchange of extracellular K(+) for cytoplasmic Na(+) or H(+), respectively. At present, the structural determinants for the specific function of these ATPases remain poorly understood. To investigate the cation selectivity of these ATPases, we constructed a series of Na,K-ATPase mutants in which residues in the membrane spanning segments of the alpha subunit were changed to the corresponding residues common to gastric H,K-ATPases. Thus, mutants were created with substitutions in transmembrane domains TM1, TM4, TM5, TM6, TM7, and TM8 independently or together (designated TMAll). The function of each mutant was assessed after coexpression with the beta subunit in Sf-9 cells using baculoviruses. The enzymatic properties of TM1, TM7, and TM8 mutants were similar to the wild-type Na,K-ATPase, and while TM5 showed modest changes in apparent affinity for Na(+), TM4, TM6, and TMAll displayed an abnormal activity. This resulted in a Na(+)-independent hydrolysis of ATP, a 2-fold higher K(0.5) for Na(+) activation, and the ability to function at low pH. These results suggest a loss of discrimination for Na(+) over H(+) for the enzymes. In addition, TM4, TM6, and TMAll mutants exhibited a 1.5-fold lower affinity for K(+) and a 4-5-fold decreased sensitivity to vanadate. Altogether, these results provide evidence that residues in transmembrane domains 4 and 6 of the alpha subunit of the Na,K-ATPase play an important role in determining the specific cation selectivity of the enzyme and also its E1/E2 conformational equilibrium.     

Functional role of oxygen-containing residues in the fifth transmembrane segment of the Na,K-ATPase alpha subunit

The functional roles of Tyr771, Thr772, and Asn776 in the fifth transmembrane segment of the Na, K-ATPase alpha subunit were studied using site-directed mutagenesis, expression, and kinetics analysis. Nonconservative replacements Thr772Tyr and Asn776Ala led to reduced Na,K-ATPase turnover. Replacements at these positions (Asn776Ala, Thr772Leu, and Thr772Tyr) also led to high Na-ATPase activity (in the absence of K+). However, Thr772- and Asn776-substituted enzymes showed only small alterations in the apparent Na+ and K+ affinities (K1/2 for Na,K-ATPase activation). Thus, the high Na-ATPase activity does not appear related to cation-binding alterations. It is probably associated with conformational alterations which lead to an acceleration of enzyme dephosphorylation by Na+ acting at the extracellular space (Argüello et al. J. Biol. Chem. 271, 24610-24616, 1996). Nonconservative substitutions at position 771 (Tyr771Ala and Tyr771Ser) produced a significant decrease of enzyme turnover. Enzyme-Na+ interaction was greatly changed in these enzymes, while their activation by K+ did not appear affected. Although the Na+ K1/2 for Na,K-ATPase stimulation was unchanged (Tyr771Ala, Tyr771Ser), the activation by this cation showed no cooperativity (Tyr771Ala, nHill = 0.75; Tyr771Ser, nHill = 0.92; Control, nHill = 2.28). Substitution Tyr771Phe did not lead to a significant reduction in the cooperativity of the ATPase Na+ dependence (nHill = 1.91). All Tyr771-substituted enzymes showed low steady-state levels of phosphoenzyme during Na-activated phosphorylation by ATP. Phosphorylation levels were not increased by oligomycin, although the drug bound and inactivated Tyr771-substituted enzymes. No E1 left and right arrow E2 equilibrium alterations were detected using inhibition by vanadate as a probe. The data suggest that Tyr771 might play a central role in Na+ binding and occlusion without participating in K+-enzyme interactions.     

Structure of the 5th transmembrane segment of the Na,K-ATPase alpha subunit: a cysteine-scanning mutagenesis study

To study the structure of the pathway of cations across the Na, K-ATPase, we applied the substituted cysteine accessibility method to the putative 5th transmembrane segment of the alpha subunit of the Na,K-ATPase of the toad Bufo marinus. Only the most extracellular amino acid position (A(796)) was accessible from the extracellular side in the native Na,K-pump. After treatment with palytoxin, six other positions (Y(778), L(780), S(782), P(785), E(786) and L(791)), distributed along the whole length of the segment, became readily accessible to a small-size methanethiosulfonate compound (2-aminoethyl methanethiosulfonate). The accessible residues are not located on the same side of an alpha-helical model but the pattern of reactivity would rather suggest a beta-sheet structure for the inner half of the putative transmembrane segment. These results demonstrate the contribution of the 5th transmembrane segment to the palytoxin-induced channel and indicate which amino acid positions are exposed to the pore of this channel.     

Na,K-ATPase: radiation inactivation studies

Na,K-ATPase from duck salt gland and ox brain in the membrane-bound or solubilized form was studied by the radiation inactivation technique using ATP, CTP, GTP or p-NPP as substrates. The values of radiation inactivation size (RIS) were compared with the target size (TS) for the alpha-subunit of the enzyme obtained by an independent method as well as with analytical centrifugation data obtained for C12E8-solubilized enzyme. It was concluded that during ATP (CTP) hydrolysis the enzyme operates as an oligomeric structure; the complex formation requires the presence of K+ and adenosine triphosphate binding to the sites with a low affinity for the nucleotide. Specially designed experiments revealed that the degree of enzyme oligomerization increases with an increase in the microviscosity of the membrane lipid environment.     

Na/K-ATPase mimetic pNaKtide peptide inhibits the growth of human cancer cells

Cells contain a large pool of nonpumping Na/K-ATPase that participates in signal transduction. Here, we show that the expression of α1 Na/K-ATPase is significantly reduced in human prostate carcinoma as well as in several human cancer cell lines. This down-regulation impairs the ability of Na/K-ATPase to regulate Src-related signaling processes. A supplement of pNaKtide, a peptide derived from α1 Na/K-ATPase, reduces the activities of Src and Src effectors. Consequently, these treatments stimulate apoptosis and inhibit growth in cultures of human cancer cells. Moreover, administration of pNaKtide inhibits angiogenesis and growth of tumor xenograft. Thus, the new findings demonstrate the in vivo effectiveness of pNaKtide and suggest that the defect in Na/K-ATPase-mediated signal transduction may be targeted for developing new anticancer therapeutics.     

Expression of an ouabain-resistant Na,K-ATPase in CV-1 cells after transfection with a cDNA encoding the rat Na,K-ATPase alpha 1 subunit

We have used a gene transfer system to investigate the relationship between expression of the rat Na,K-ATPase alpha 1 subunit gene and ouabain-resistant Na,K-ATPase activity. A cDNA clone encoding the entire rat Na,K-ATPase alpha 1 subunit was inserted into the expression vector pSV2neo. This construct (pSV2 alpha 1) conferred resistance to 100 microM ouabain to ouabain-sensitive CV-1 cells. Hybridization analysis of transfected clones revealed the presence of both rat-specific and endogenous Na,K-ATPase alpha 1 subunit DNA and mRNA sequences. A single form of highly ouabain-sensitive 86Rb+ uptake was detected in CV-1 cells, whereas two distinct classes of ouabain-inhibitable uptake were observed in transfectants. One class exhibited the high ouabain sensitivity of the endogenous monkey Na,K-ATPase, while the second class showed the reduced ouabain sensitivity characteristic of the rodent renal Na,K-ATPase. Examination of the ouabain-sensitive, sodium-dependent ATPase activity of the transfectants also revealed a low affinity component of Na,K-ATPase activity characteristic of the rodent kidney enzyme. These results suggest that expression of the rat alpha 1 subunit gene is directly responsible for ouabain-resistant Na,K-ATPase activity in transfected CV-1 cells.     

A new variant of the gamma subunit of renal Na,K-ATPase. Identification by mass spectrometry, antibody binding, and expression in cultured cells

The gamma subunit is a specific regulator of Na,K-ATPase expressed mainly in kidney. On SDS-polyacryylamide gel electrophoresis, gamma runs as a doublet, but the origin and significance of the doublet is obscure. Mass spectrometry of the gamma chains of rat kidney Na, K-ATPase shows that gamma(a) (upper) has a mass of 7184.0 +/- 1 Da (carbamidomethyl cysteine), corresponding closely to that for the published sequence without the initiator methionine, while gamma(b) (lower) has a mass of 7337.9 +/- 1Da. Tryptic peptide mapping and sequencing by mass spectrometry reveals that the seven N-terminal residues of gamma(a), TELSANH, are replaced by Ac-MDRWYL in gamma(b), but otherwise the chains are identical. Antibodies raised against peptides TELSANHC and MDRWYLC recognize either gamma(a) or gamma(b) of the Na,K-ATPase, respectively. gamma(a) or gamma(b) cDNAs have been expressed in human embryonic kidney and HeLa cells. The major bands expressed correspond to gamma(a) or gamma(b) of renal Na, K-ATPase. Additional minor bands seen after transfection, namely gamma(a)' in human embryonic kidney and gamma(b)' in HeLa, are presumably cell-specific modifications. The present work clarifies earlier uncertainty regarding doublets seen in kidney and in transfected cells. In particular, the results show that renal Na, K-ATPase contains two variants of the gamma subunit with different sequences but otherwise are unmodified. We discuss the possible functional significance of the two variants.     

Modulation of cytokine production by human mononuclear cells following impairment of Na,K-ATPase activity

Cytokines, including TNFα and IL-1β, are central to the chronic inflammatory process and tissue damage that characterises diseases such as rheumatoid arthritis. The mechanisms responsible for long-term generation of these molecules are poorly understood. We have previously demonstrated impaired activity of Na,K-ATPase, a key enzyme regulating intracellular cation levels, on rheumatoid mononuclear cells. Mimicking this `defect' on normal mononuclear cells with ouabain has been shown to induce TNFα and, in particular, IL-1β production, whereas IL-6 synthesis was suppressed. A similar pattern of cytokine generation was noted when mononuclear cells were treated with the sodium ionophore, monensin. Induction of cytokine production was related to up-regulation of the appropriate mRNA, although enhanced secretion of processed IL-1β was also observed. The mechanism underlying these cellular responses appears to involve sodium/calcium exchange across the cell membrane. Impaired Na,K-ATPase activity might promote pro-inflammatory cytokine secretion in patients with rheumatoid arthritis.