Stimulation of A_2a adenosine receptor abolishes the inhibitory effect of arachidonic acid on the basolateral 50-pS K channel in the thick ascending limb

Abstract

The basolateral 50-pS K channels are stimulated by a cAMP-dependent pathway and inhibited by cytochrome P-450-omega-hydroxylase-dependent metabolism of arachidonic acid (AA) in the rat thick ascending limb (TAL). We now used the patch-clamp technique to examine whether stimulation of adenosine A_2a receptor modulates the inhibitory effect of AA on the basolateral 50-pS K channels in the medullary TAL. Stimulation of adenosine A_2a receptor with CGS-21680 or inhibition of phospholipase A₂ (PLA₂) with AACOCF3 increased the 50-pS K channel activity in the TAL. Western blot demonstrated that application of CGS-21680 decreased the phosphorylation of PLA₂ at serine residue 505, an indication of inhibiting PLA₂ activity. In the presence of CGS-21680, inhibition of PLA₂ had no further effect on the basolateral 50-pS K channels. The possibility that CGS-21680-induced stimulation of the basolateral 50-pS K channels was partially achieved by inhibition of PLA₂ in the TAL was also supported by the observation that CGS-21680 had no additional effect in the presence of AACOCF3. Moreover, stimulation of adenosine A_2a receptor with CGS-21680 also abolished the inhibitory effect of AA and 20-hydroxyeicosatetraenoic acid (20-HETE) on the 50-pS K channels. The effect of CGS-21680 on AA and 20-HETE-mediated inhibition of the 50-pS K channels was mediated by cAMP because application of membrane-permeable cAMP analog, dibutyryl-cAMP, not only increased the 50-pS K channel activity but also abolished the inhibitory effect of AA and 20-HETE. We conclude that stimulation of adenosine A_2a receptor increased the 50-pS K channel activity in the TAL, an effect that is achieved by suppression of PLA₂ activity and 20-HETE-induced inhibition.

basolateral k channels in the thick ascending limb (TAL) play an important role in the regulation of transepithelial transport (9, 17). They are responsible for generating basolateral membrane potential, which is essential for Cl diffusion across the basolateral membrane in the TAL. The physiological importance of the basolateral K channels in maintaining transepithelial membrane transport in the TAL and distal nephron is best demonstrated in SeSAME disease (Seizures, Sensorineural deafness, Ataxia, Mental retardation, and Electrolyte imbalance). This disease is the result of defective gene product encoding KCNJ10, an inwardly rectifying K channel (28), which is also expressed in the basolateral membrane of TAL and distal nephron (19). The renal phenotypes of SeSAME disease are hypokalemia, metabolic alkalosis, and hypomagnesemia. Presumably, defective basolateral K channels cause membrane potential depolarization thereby decreasing Cl diffusion across the basolateral membrane. Consequently, an increase in intracellular Cl concentrations results in the inhibition of the apical Na entry through Na-K-Cl cotransporter thereby suppressing Na absorption and diminishing the lumen-positive potential in the TAL. A decrease in the lumen-positive potential results in inhibition of magnesium absorption through the paracellular pathway in the TAL (16). Also, the inhibition of the Na absorption in the TAL increases Na delivery to the distal nephron, accordingly, stimulating Na absorption in the expanse of K in the connecting tubule and causing K wasting. Therefore, an alteration in the basolateral K channel activity has a significant effect on transepithelial transport not only in the TAL but also in the distal nephron segment. We and others demonstrated that a 50-pS K channel was highly expressed in the basolateral membrane of the TAL and most likely played an important role in generating the basolateral membrane potential (14, 25). Previous studies demonstrated that cAMP increased, whereas arachidonic acid (AA) inhibited the 50-pS K channels by cytochrome P-450 (CYP)-omega-hydroxylase-dependent metabolism (11, 14). Furthermore, stimulation of adenosine A_2a receptor activated the basolateral 50-pS K channels by a cAMP-dependent protein kinase A pathway. However, it has not been explored whether the stimulation of adenosine A_2a receptor modulates the AA-induced inhibition of the 50-pS K channels. Therefore, the aim of the present study is to examine whether stimulation of adenosine A_2a receptor modulates the inhibitory effect of AA-dependent pathway on the 50-pS K channels.

METHODS

Preparation of the TAL.

Sprague-Dawley rats of either sex (5–6 wk) were purchased from the animal facility in the second affiliated hospital of Harbin Medical University. The animals were kept on a normal rat chow and were free to access water. Rats were killed by cervical dislocation (body wt <90 g) and the kidneys were removed for dissecting the medullary TAL. The isolated medullary TAL tubules were incubated in a buffer solution containing type 1A collagenase (1 mg/ml; Sigma, St. Louis, MO) at 37°C for 45 min. The composition of the buffer solution was (in mM) 140 NaCl, 5 KCl, 1.5 MgCl₂, 1.8 CaCl₂, and 10 HEPES (pH 7.4). The collegenase-treated TAL was transferred onto a 5×5-mm cover glass coated with polylysine (Sigma) to immobilize the tubule. The cover glass was placed in a chamber mounted on an inverted microscope (Nikon) and the tubules were superfused with HEPES-buffered NaCl solution described above. The animal-using protocol has been approved by an independent animal-using committee in Harbin Medical University.

Patch-clamp technique.

Patch-clamp electrodes were made using a Narishige (P-81) puller with thick-wall glass capillaries (Degan, Minneapolis, MN) and had resistance of 4–6 MΩ when filled with 140 mM KCl. An Axon 200A patch-clamp amplifier was used for the patch-clamp experiments. The current was low-pass filtered at 0.5 kHz and digitized by an Axon interface (Digidata 1200). Data were sampled by an IBM-compatible Pentium computer at a rate of 4 kHz and analyzed by using a pClamp software system 9 (Axon Instruments, Burlingame, CA). The channel activity was defined as NP_o, a product of channel open probability (P_o) and channel number (N). The NP_o was calculated from data samples of 90-s duration in the steady state as follows: NP_o = ∑ (1t₁ + 2t₂ +…it_i), where t_i is the fractional open time spent at each of the observed current levels.

The pipette solution contained (in mM) 140 KCl, 1.8 mM MgCl₂, and 10 HEPES (pH 7.4).

Western blot.

Protein samples (50 μg) obtained from medullary TAL were separated by electrophoresis at 10% SDS-polyacrylamide gels and transferred to nitrocellulose membranes. The membranes were blocked with 5% nonfat dry milk in Tween-Tris-buffered saline (TBS-T), rinsed, and washed with 0.1% Tween-TBS. The membranes were incubated with the primary antibody at 4°C for 12 h. After being washed four times (10 min for each wash) with TBS-T, the membrane was incubated with the secondary antibody for an additional hour. ECL plus (Amersham Pharmacia Biotech) was used to detect the protein bands.

Chemicals and statistics.

CGS-21680, AACOCF3, AA, 20-hydroxyeicosatetraenoic acid (20-HETE), and dibutyryl-cAMP were purchased from Sigma. Antibody reacting with serine residue 505 of phospholipase A₂ (PLA₂) was purchased from Santa Cruz Biotechnology. Data are shown as means ± SE. We used t-tests to determine the significance of the difference between the control and experimental periods. If the P value was <0.05, the difference was considered to be significant.

RESULTS

We confirmed the previous finding that stimulation of adenosine A_2a receptor increased the activity of the basolateral 50-pS K channels (14). Addition of 5 μM CGS-21680 increased the basolateral K channel activity (NP_o) from 0.20 ± 0.03 to 0.53 ± 0.06 (Fig. 1), an effect observed within 5–10 min after adding CGS-21680 to the bath. Our previous study demonstrated the 50-pS K channel activity was constitutively inhibited by CYP-omega-hydroxylase-dependent metabolites of AA because inhibition of CYP-omega-hydroxylase increased the 50-pS K channels in the TAL (11). The notion that AA inhibits the 50-pS K channels was also supported by the experiments in which inhibiting PLA₂ with AACOCF3 mimicked the effect of blocking CYP-omega-hydroxylase and increased the basolateral 50-pS K channel activity. Figure 1 is a channel recording showing that application of 5 μM AACOCF3 stimulated the 50-pS K channels and increased NP_o from 0.18 ± 0.05 to 0.48 ± 0.1 (n = 4, P < 0.01; Fig. 1B). Next, we examined whether stimulation of adenosine A_2a receptor and inhibition of PLA₂ could have additive effect on the 50-pS channels. We treated the TAL with 5 μM CGS-21680 for 15 min followed by examining the effect of AACOCF3 on the 50-pS channels in cell-attached patches (Fig. 2). From inspection of Fig. 2, it is apparent that the incubation of TAL with CGS-21680 increased basal level of the 50-pS K channel activity (NP_o, 0.45 ± 0.1). However, inhibition of PLA₂ with 5 μM AACOCF3 failed to further stimulate the 50-pS K channel activity (NP_o, 0.5 ± 0.1, n = 5), suggesting that the effect of stimulating adenosine A_2a receptor and the effect of inhibiting PLA₂ were not additive. This notion was also supported by experiments in which we examined the effect of CGS-21680 on the 50-pS K channels in the TAL treated with AACOCF3 (Fig. 3). Application of 5 μM CGS-21680 did not significantly increase NP_o in the presence of AACOCF3 (0.30 ± 0.08 before CGS-21680 and 0.32 ± 0.09 after CGS-21680, n = 6). This strongly suggests the possibility that the stimulation of adenosine A_2a receptor-induced increase in the 50-pS K channel activity might partially be resulted from inhibiting PLA₂ activity thereby diminishing the AA-mediated inhibition of the K channels. It has been shown that phosphorylation of PLA₂ at serine residue 505 increases its catalytic activity (21, 22). Therefore, we employed the specific antibody that reacts with phosphorylated PLA₂ at serine residue 505 (phosph-PLA₂) to examine whether application of CGS-21680 affected the phosphorylation level of PLA₂ at serine residue 505. We incubated the isolated medullary TALs with CGS-21680 at 1 to 20 μM for 5 min and the cell lysates obtained from CGS-21680-treated TALs were subjected to electrophoresis followed by adding the phosph-PLA₂ antibody. Figure 4A is a typical Western blot from three such experiments showing that CGS-21680 treatment significantly decreased the phosphorylation level of PLA₂ at serine residue 505 by 40 ± 10% (5 μM), 50 ± 10% (10 μM), and 60 ± 15% (20 μM), respectively. Also, the 5-min incubation was sufficient to have a maximal effect on the PLA₂ phosphorylation (Fig. 4B, left) since prolonged incubation (15, 30 min) of the TAL with 5 μM. CGS-21680 did not further decrease the phosphorylation of PLA₂ (Fig. 4B, right). The results strongly suggest that stimulation of adenosine A_2a receptor inhibits the PLA₂ activity thereby minimizing AA-induced inhibition of the 50-pS K channels.

Fig. 1.

A: effect of AACOCF3 (5 μM) on the 50-pS K channel activity in the basolateral membrane of the thick ascending limb (TAL). The experiments were performed in a cell-attached patch. Top trace shows the experimental time course. Two parts of the trace, indicated by numbers, are extended to show the fast time resolution. The holding potential was 0 mV, and the channel closed current is indicated by “C.” B: bar graph summarizes the experiments in which the effect of CGS-21680 and AACOCF3 (5 μM) on the basolateral 50-pS K channels was examined in cell-attached patches.

Fig. 2.

Channel recording showing the effect of 5 μM AACOCF3 on the 50-pS K channels in the TAL treated with CGS-21680 (5 μM). The experiments were performed in a cell-attached patch. The holding potential was 0 mV, and the channel closed current is indicated by C. Two parts of the trace, indicated by numbers, are extended to show the fast time resolution.

Fig. 3.

Effect of CGS-21680 (5 μM) on the basolateral K channel activity in the presence of AACOCF3 (5 μM). Top trace shows the experimental time course. Two parts of the trace, indicated by numbers, are extended to show the fast time resolution. The holding potential was 0 mV, and the channel closed current is indicated by C. The experiments were performed in a cell-attached patch and the filter rate was 200 Hz.

Fig. 4.

A: Western blot showing that CGS-21680 treatment (1–20 μM) decreased phosphorylation of phospholipase A₂ (PLA₂) at serine residue 505 in the isolated medullary TAL. The tubules were treated with CGS-21680 at different concentrations for 15 min. B: effect of CGS-21680 on phosphorylated PLA₂ (phospho-PLA₂) on serine residue 505 at 5-min (left) and 15- and 30-min incubation (right).

To further determine whether stimulation of adenosine A_2a receptor also modulated the inhibitory effect of AA and 20-HETE, we examined the effect of AA on the 50-pS K channels in the TAL treated with 5 μM CGS-21680 for 15 min. Figure 5A is a channel recording demonstrating that stimulation of adenosine A_2a receptor with 5 μM CGS-21680 abolished the inhibitory effect of 5 μM AA, which inhibited the 50-pS K channels in the absence of CGS-21680 (11). In the presence of CGS-21680, AA failed to inhibit the 50-pS K channels and the mean NP_o (0.44 ± 0.06) was not significantly different from the control (CGS-21680, NP_o, 0.52 ± 0.1, n = 5; Fig. 5B). We also examined the effect of 20-HETE on the 50-pS K channels in the TAL treated with CGS-21680 (Fig. 5). From inspection of Fig. 5, it is apparent that 20-HETE also failed to inhibit the 50-pS K channels in the presence of CGS-21680 (NP_o, 0.50 ± 0.06, n = 5; Fig. 5B), suggesting that stimulation of adenosine A_2a receptor blocked the effect of 20-HETE on the 50-pS K channels. Thus, results indicate that the stimulatory effect of CGS-21680 was not completely the result of suppressing PLA₂ activity and that stimulation of adenosine A_2a receptor also affected the inhibitory effect of AA/20-HETE on the 50-pS K channels. Because 1 μM CGS-21680 only modestly reduced the PLA₂ phosphorylation (Fig. 4A), we suspect that AA should still inhibit the 50-pS K channels in the TAL treated with 1 μM CGS-21680 (Fig. 6). This notion was supported by the patch-clamp experiments showing that AA inhibited the 50-pS K channels and reduced NP_o from 0.42 ± 0.05 (1 μM CGS-21680) to 0.26 ± 0.04 (5 μM AA+ 1 μM CGS-21680; n = 6).

Fig. 5.

A: channel recording showing the effect of 5 μM arachidonic acid (AA; top) or 200 nM 20-hydroxyeicosatetraenoic acid (20-HETE; bottom) on the 50-pS K channels in the TAL treated with 5 μM CGS-21680. The holding potential was 0 mV, and the channel closed current is indicated by C. The experiments were performed in a cell-attached patch. B: bar graph summarizes the experiments in which the effect of AA (5 μM) and 20-HETE (200 nM) on the 50-pS K channels was examined in cell-attached patches of CGS-21680-treated TAL.

Fig. 6.

Channel recording showing the effect of 5 μM AA on the 50-pS K channels in the TAL treated with 1 μM CGS-21680. The holding potential was 0 mV, and the channel closed current is indicated by C. The experiments were performed in a cell-attached patch. Two parts of the trace are extended to demonstrate the channel activity at a fast time resolution.

Stimulation of adenosine A_2a receptor is expected to increase the intracellular cAMP level and activate protein kinase A (1). We next examined whether application of cAMP mimicked the effect of stimulating adenosine A_2a receptor and abolished the inhibitory effect of AA and 20-HETE on the 50-pS K channels. Figure 7 is a channel recording demonstrating that dibutyryl-cAMP (100 μM) stimulated the 50-pS K channels and increased channel activity from 0.22 ± 0.06 to 0.44 ± 0.07 (n = 5). Moreover, application of 5 μM AA failed to inhibit the K channels and the channel activity was 0.42 ± 0.13 in the presence of dibutyryl-cAMP (n = 6; Fig. 8B). Also, the treatment of the TAL with dibutyryl-cAMP abolished the effect of 20-HETE on the basolateral 50-pS K channels. Figure 8A is a channel recording demonstrating that 20-HETE failed to inhibit the 50-pS K channels in the medullary TAL treated with dibutyryl-cAMP. Data summarized in Fig. 8B demonstrate that channel activity (NP_o) in the presence of 200 nM 20-HETE was 0.34 ± 0.1, a value not significantly different from the control (dibutyryl-cAMP) NP_o (0.44 ± 0.07). Therefore, cAMP abolished both AA- and 20-HETE-induced inhibition of the 50-pS K channels.

Fig. 7.

Effect of AA (5 μM) on the basolateral K channel activity in the presence of dibutyryl-cAMP (DB-cAMP; 100 μM). Top trace shows the experimental time course. Three parts of the trace, indicated by numbers, are extended to show the fast time resolution. The holding potential was 0 mV, and the channel closed current is indicated by C. The experiments were performed in a cell-attached patch.

Fig. 8.

A: channel recording showing the effect of 20-HETE (200 nM) on the basolateral K channel activity in the presence of DB-cAMP (100 μM). The holding potential was 0 mV, and the channel closed current is indicated by C. The experiments were performed in a cell-attached patch. B: bar graph summarizes the experiments in which the effect of AA (5 μM) and 20-HETE (200 nM) on the 50-pS K channels was examined in cell-attached patches of DB-cAMP-treated TAL. NS, not significant.

DISCUSSION

Adenosine plays an important role in regulating renal hemodynamics, tubuloglomerular feedback, and renin release and is also involved in regulating renal epithelial transport in several nephron segments including the proximal tubule and TAL (15, 18, 24, 27, 30). Molecular biology, protein chemistry, and functional studies demonstrated that both adenosine A₁ or A_2a receptors are present in the TAL (4, 34). The physiological significance of adenosine A₁ receptor in modulating the membrane transport in the TAL has been suggested by experiments in which stimulation of adenosine A₁ receptor inhibited Cl absorption by 50% in the isolated, perfused medullary TAL (2). Furthermore, micropucture experiments performed in the adenosine A₁ receptor knockout mice indicated the role of adenosine A₁ receptor in inhibiting Na absorption in the water-impermeable loop of Henle's (32). Increased Na intake has been shown to stimulate adenosine formation in the interstitial fluid of renal outer medulla (29) thereby inhibiting Na transport in the TAL by activating adenosine A₁ receptor. Therefore, adenosine-induced inhibition of the Na absorption may play a role in preventing Na retention during increasing Na intake. Also, renal hypoxia has been shown to stimulate adenosine release in the medullary TAL (3), an effect that was blocked by inhibiting Na transport. It has been speculated that an increase in adenosine release should inhibit Na transport in the TAL through activating adenosine A₁ receptor thereby minimizing hypoxia-induced renal damage.

In contrast to adenosine A₁ receptor, the physiological importance of adenosine A_2a receptor in regulating renal epithelial transport is less explored. Our previous study demonstrated that stimulation of adenosine A_2a receptor enhanced the apical and basolateral K channel activity in the TAL (14, 20). We speculate that stimulation of adenosine A_2a receptor may serve as a local regulatory factor that links the activity of apical Na-K-Cl cotransporter to the basolateral Cl transport. Stimulation of the apical Na-Cl-K cotransporter increases the Na-K-ATPase activity and enhances ATP hydrolysis thereby increasing AMP generation and adenosine formation through cytosolic 5′-nucleotidase or membrane-associated endo-5′nucleotidase-dependent metabolism (33). Indeed, it has been reported that stimulation of Na transport enhances adenosine release in the macula densa (31). Since epithelial cells in the TAL have similar transport properties as those in the macula densa, it is conceivable that stimulation of Na transport could also increase intra- or extracellular generation of adenosine, which in turn activates adenosine A_2a receptor in the TAL. Consequently, stimulation of adenosine A_2a receptor activates the basolateral 50-pS K channels and increases the driving force for Cl exit. Thus, activation of adenosine A_2a receptor should have a positive feedback mechanism linking the apical NaCl transport to the basolateral Cl transport. This mechanism would prevent an increase in intracellular Cl concentrations, which have been shown to inhibit the activity of the Na-K-2Cl cotransporters (8).

In addition, stimulation of adenosine A_2a receptor may play a role in regulating cell volume and preventing cell swelling in the medullary TAL, which is bathed in the extracellular fluid having a high-NaCl concentration. A high extracellular NaCl concentration has been shown to stimulate adenosine release and to increase the intracellular cAMP formation presumably through activation of adenosine A_2a receptor in the medullary TAL (1). Consequently, an increase in cAMP should stimulate the basolateral K and Cl channel activity. Therefore, stimulation of adenosine A_2a receptor should favor Cl exit thereby maintaining a normal intracellular osmolarity. We speculate that both adenosine A₁ and A_2a receptors are involved in regulating the membrane transport in the TAL. Relevant to our speculation is the report that adenosine regulates vascular function of the glomerular afferent arteriol through the coordination between adenosine A₁ and A_2a receptors (5).

The CYP-omega-hydroxylase-dependent AA metabolism also plays an important role in regulating Na and HCO₃⁻ transport in the TAL (10, 23, 26). Application of AA inhibited both apical Na-K-Cl cotransporter, 70-pS K channels, and basolateral K and Cl channels in the TAL (7, 11, 13, 35). The inhibitory effects of AA on the cotransporters and ion channels were mediated by CYP-omega-hydrolase-dependent metabolites of AA such as 20-HETE because inhibition of CYP-omega-hydroxylase abolished the effect of AA. The CYP-omega-hydroxylase-dependent AA metabolism mediates the effect of Ca²⁺-sensing receptor on the apical K channels in the TAL. A previous study showed that stimulation of Ca²⁺-sensing receptor activated PLA₂ and inhibited the apical 70-pS K channels in the TAL (36). The finding that inhibition of PLA₂ increased the basolateral 50-pS K channel activity suggests that PLA₂ activity was involved in regulating the basal activity of basolateral 50-pS K channels in the TAL. Two lines of evidence suggest that the stimulation of adenosine A_2a receptor-induced increase in the basolateral 50-pS K channels was at least partially through inhibiting PLA₂ activity: 1) the effect of CGS-21680 and AACOCF3 was not additive and 2) incubation of isolated medullary TAL with CGS-21680 decreased the phosphorylation of PLA₂ at serine residue 505. Moreover, the finding that application of AA or 20-HETE failed to block the basolateral K channels in the TAL treated with CGS-21680 or dibutyryl-cAMP suggests a cross-talk between cAMP-dependent stimulatory pathway and 20-HETE-dependent inhibitory pathway in regulating the basolateral 50-pS K channels in the TAL.

Such a cross-talk has also been observed in regulating the apical K channels in the TAL, which was inhibited by 20-HETE and stimulated by a cAMP-dependent pathway (12, 20). However, 20-HETE plays a predominant role in regulating apical K channels since stimulation of adenosine A_2a receptor failed to activate the apical K channels when CYP-omega-hydroxylase-dependent AA metabolism was active such as during K depletion (20). Therefore, it is possible that some global physiological factors other than cAMP-dependent or 20-HETE-dependent pathway play a role in regulating the cross-talk between activation of adenosine A_2a receptor and CYP-omega-hydroxylase-dependent pathway. Since K depletion is known to suppress the Na transport in the TAL (6), it is physiologically important that the inhibitory effect of 20-HETE plays a dominant role under K-depleted conditions. On the other hand, to effectively link the apical and basolateral Cl transport requires that cAMP-dependent signaling plays a dominant role in regulating the basolateral K channels to synchronize the membrane transport.

Figure 9 is a cell model illustrating the mechanism by which adenosine stimulates basolateral K channels in the TAL. An increase in Na delivery to the TAL is expected to stimulate Na transport and to increase ATP consumption, which is expected to enhance adenosine formation. Stimulation of adenosine A_2a receptor activates adenylate cyclase and cAMP generation, which inhibits PLA₂ activity thereby decreasing the AA release. Moreover, stimulation of cAMP-dependent pathway suppresses the 20-HETE-mediated inhibition of the 50-pS K channels. However, the mechanism by which cAMP suppresses the effect of 20-HETE is not understood. It is possible that cAMP-PKA-induced stimulation of the 50-pS K channel may cause the conformation changes of the 50-pS K channel thereby indirectly preventing the effect of 20-HETE on the basolateral K channels. As consequence of activation of basolateral K channels, adenosine hyperpolarizes the basolateral membrane potential and increases the driving force for Cl exit. Therefore, stimulation of adenosine A_2a receptor synchronizes the apical Cl entry to the basolateral Cl exit. We conclude that stimulation of adenosine A_2a receptor activates the basolateral 50-pS K channels by increasing cAMP generation, which suppresses PLA₂ activity and the inhibitory effect of 20-HETE.

Fig. 9.

Cell scheme illustrating the mechanism by which stimulation of adenosine A_2a receptor suppresses PLA₂ activity and 20-HETE effect thereby stimulating the 50-pS K channels. AC, adenylate cyclase.

GRANTS

This work was supported by the Chinese National Nature Science Foundation no. 31071017 and no. 30770800 (to R. M. Gu), the Research Fund of High Education for Doctoral Program no. 20040226007 (to R. M. Gu), and National Institutes of Health Grant HL34300 (W. H. Wang).

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the author(s).