点击显示 收起
【摘要】 A previous study demonstrated that proximal tubule cells regulate HCO 3 - reabsorption by sensing acute changes in basolateral CO 2 concentration, suggesting that there is some sort of CO 2 sensor at or near the basolateral membrane (Zhou Y, Zhao J, Bouyer P, and Boron WF Proc Natl Acad Sci USA 102: 3875-3880, 2005). Here, we hypothesized that an early element in the CO 2 signal-transduction cascade might be either a receptor tyrosine kinase (RTK) or a receptor-associated (or soluble) tyrosine kinase (sTK). In our experiments, we found, first, that basolateral 17.5 µM genistein, a broad-spectrum tyrosine kinase inhibitor, virtually eliminates the CO 2 sensitivity of HCO 3 - absorption rate ( J ). Second, we found that neither basolateral 250 nM nor basolateral 2 µM PP2, a high-affinity inhibitor for the Src family that also inhibits the Bcr-Abl sTK as well as the Kit RTK, reduces the CO 2 -stimulated increase in J. Third, we found that either basolateral 35 nM PD168393, a high-affinity inhibitor of RTKs in the erbB (i.e., EGF receptor) family, or basolateral 10 nM BPIQ-I, which blocks erbB RTKs by competing with ATP, eliminates the CO 2 sensitivity. In conclusion, the transduction of the CO 2 signal requires activation of a tyrosine kinase, perhaps an erbB. The possibilities include the following: 1 ) a TK is simply permissive for the effect of CO 2 on J; 2 ) a CO 2 receptor activates an sTK, which would then raise J; 3 ) a CO 2 receptor transactivates an RTK; and 4 ) the CO 2 receptor could itself be an RTK.
【关键词】 kidney outofequilibrium solutions acidbase volume reabsorption signal transduction
RENAL TUBULE CELLS PLAY A central role in whole body acid-base balance by 1 ) secreting H + into the tubule lumen, thereby titrating NH 3 to NH 4 + and also creating titratable acid; and 2 ) moving an equivalent amount of "new HCO 3 - " into the blood, 70 mmol/day in humans, and thereby titrating the fixed acid produced by metabolism and generated by the gastrointestinal tract. At the same time, the tubules must also reabsorb nearly all of a much larger amount of HCO 3 - filtered in the glomeruli. The proximal tubule (PT) is the site of generation of 60% of the new HCO 3 - and the site of reabsorption of 80% of the filtered HCO 3 -, using the same transporters in both processes. The cell uses cytosolic carbonic anhydrase II ( 51, 52 ) to convert CO 2 + H 2 O to H + + HCO 3 -, and then extrudes the H + across the apical membrane ( 3, 10, 50 ) via Na/H exchangers ( 6, 7, 34 ) and H + pumps ( 22 ) and exports the HCO 3 - across the basolateral membrane, mainly via the electrogenic Na-HCO 3 cotransporter ( 11, 16, 47, 48 ). For the system to work properly, the PT cell must be able to respond to changes in whole body acid-base status by appropriately adjusting transporters activities.
To study how the PT senses changes in whole body acid-base status, our laboratory developed out-of-equilibrium (OOE) CO 2 /HCO 3 - solutions for altering, one at a time, basolateral [CO 2 ], [HCO 3 - ], or [H + ] (brackets denote concentration) ( 58, 59 ). We found that, at least in regard to acute acid-base disturbances, the PT responds not to changes in either basolateral (BL) pH (pH BL ) or intracellular pH (pH i ) but to changes in [CO 2 ] BL and [HCO 3 - ] BL ( 61 ). In the case of [CO 2 ] BL, increases cause HCO 3 - reabsorption to rise, whereas decreases have the opposite effect. Moreover, the data suggest that the PT cell senses [CO 2 ] BL directly, utilizing some sort of CO 2 sensor at or near the basolateral membrane. The key questions concern the nature of the CO 2 sensor and the mechanisms by which the cell transduces the CO 2 signal to an increase in HCO 3 - reabsorption.
The past two decades have seen major advances in understanding how organisms sense dissolved gases. For instance, nitric oxide binds to a heme moiety of soluble guanylyl cyclase ( 45 ), oxygen binds to a two-component receptor in bacteria, and ethylene binds to a receptor in Arabidopsis thaliana.
The bacterium Rhizobium meliloti senses oxygen using a two-component system, consisting of FixL and FixJ ( 20 ). FixL is a transmembrane protein with a COOH-terminal histidine kinase domain and a NH 2 -terminal heme-binding domain that blocks the histidine kinase when O 2 binds to the heme. When the [O 2 ] falls to microaerobic levels, the histidine kinase autophosphorylates a conserved His within its catalytic core, thereby activating FixJ, which, in turn, induces transcription of genes involved in nitrogen fixation ( 21, 45 ).
Ethylene acts like a hormone in plants, regulating such events as seed germination, fruit ripening, and leaf senescence ( 8, 18 ). In 1993 Chang et al. ( 13 ), working on the plant A. thaliana, found that mutations in the ETR1 gene block ethylene signaling. The deduced amino acid sequence of the COOH-terminal half of ETR1 is homologous to both components of the two-component systems. Later work showed that ethylene binds to an NH 2 -terminal hydrophobic domain of ETR1 ( 49 ) and that the binding involves copper as a cofactor ( 46 ). Downstream of ETR1, which is a histidine kinase, the ethylene signaling cascade involves the Raf-like kinase CTR1 ( 15, 29 ) and MAPK ( 37 ).
Because, like O 2 and ethylene, CO 2 is a small volatile molecule, we entertained the hypothesis that PT cells sense CO 2 using a comparable mechanism. Because mammalian cells do not have histidine kinases, we postulated that the sensor might be either a receptor tyrosine kinase (RTK) or a receptor-associated (i.e., soluble) tyrosine kinase (sTK). Furthermore, previous studies showed that both EGF and TGF-, both of which bind to receptors in the erbB family of RTKs, stimulate HCO 3 - and phosphate reabsorption, with a higher potency for TGF- on the HCO 3 - absorption rate ( J ) ( 41, 42 ). As a first step, in the present study we have examined the effect of tyrosine kinase inhibitors on the CO 2 -induced increase in J by the PT cell.
Our approach was to use OOE solutions to vary basolateral [CO 2 ] from 0 to 20% while keeping basolateral [HCO 3 - ] and pH fixed near their physiological values in isolated, perfused rabbit S2 PTs. We found that basolateral 17.5 µM genistein, a broad-spectrum tyrosine kinase inhibitor ( 1, 24 ), virtually eliminates the CO 2 sensitivity of J. Basolateral 4-amino-5-(4-chlorophenyl)-7-( t -butyl)pyrazolo[3,4- d ]pyrimidase (PP2), which inhibits the Src family ( 26 ) sTK, the Bcr-Abl sTK, as well as the Kit ( 53 ) RTK, is without effect at either 250 nM or 2 µM. On the other hand, basolateral 35 nM PD168393, which blocks RTKs in the erbB (i.e., EGF receptor) family by alkylating a cysteine residue in the ATP binding pocket ( 19 ), blocks the CO 2 -induced increase in J. Moreover, 10 nM basolateral BPIQ-I, which blocks erbB RTKs by competing with ATP ( 44 ), also eliminates CO 2 sensitivity. Thus the transduction of the CO 2 signal requires activation of a tyrosine kinase, possibly a member of the erbB family.
METHODS
The methods are similar to those in our previous studies ( 60, 61 ).
Biological Preparation
According to procedures approved by the Yale Animal Care and Use Committee, we perfused the PTs isolated from "pathogen-free" female rabbits (New Zealand White, Elite, Covance, Denver, PA) using methods similar to those described by Burg et al. ( 12 ) and later modified by Baum et al. ( 5 ) and by our laboratory ( 36, 59 - 61 ). To summarize, a rabbit weighing 1.4-2.0 kg was euthanized by a single overdose of 3 ml ( 20 mg) of intravenous pentobarbital sodium. An incision of the abdominal wall exposed the left kidney, which we rapidly removed and cut into coronal slices that we kept in cold (4°C) modified Hanks' solution ( solution 1 in Table 1 ). We hand-dissected a slice to obtain individual midcortical S2 segments of a PT. We cannulated the perfusion end of the tubule using concentric holding, perfusion, and exchange pipettes, and drew the collection end into a holding pipette. We randomized the orientation of the tubule between the two pipettes. However, the extreme proximal portion of the proximal straight tubule was always inside one pipette, the most distal portion of the proximal convoluted tubule was exposed to the bath solution between the two pipettes, and a more proximal portion of the proximal convoluted tubule was inside the second pipette. On the collection end, we used a calibrated collection pipette (volume 55 nl) to obtain samples of fluid. The mean length of perfused tubules in our J /fluid absorption rate ( J V ) experiments, as measured with an eyepiece micrometer, was 1.23 ± 0.02 mm ( n = 96 tubules). The mean luminal collection rate was 12.1 ± 0.2 nl/min ( n = 192 collection periods). We perfused the basolateral side of the tubule (i.e., "bath") at 7 ml/min with a solution at 37°C.
Table 1. Physiological solutions
Experimental Protocol and Solutions
Table 1 lists the compositions of the solutions, all of which were identical to those used in the aforementioned studies ( 60, 61 ).
We dissected PTs in Hanks' solution ( solution 1 ) at 4°C. The luminal perfusate always was solution 2, which contained dialyzed [ 3 H]methoxyinulin (MW 7,146, NET-086L, PerkinElmer Life Sciences, Boston, MA) as the volume marker. During a 20- to 30-min warm-up period, solution 3, which contained 2% albumin, flowed through the bath at 37°C. Following this warm-up period, we switched the bath to solution 4, 5, or 6, containing no drugs or DMSO, for the first of two collection periods. During the first 5-8 min of the first collection period, we discarded the first two collected samples before allowing the collected fluid to accumulate in the collection pipette. Subsequently, we began a series of three timed and calibrated collections, two samples for analysis of [ 3 H]methoxyinulin and one for analysis of total CO 2. We then switched to a second bath solution ( solution 4, 5, or 6 ) containing genistein (345834, Calbiochem, La Jolla, CA); PP2 (529573, Calbiochem); PD168393 (513033, Calbiochem); or BPIQ-I (203696, Calbiochem) plus 1:20,000 DMSO (D-5534, Sigma, St. Louis, MO). We then repeated the procedure outlined for the first collection periods. We generated OOE CO 2 /HCO3- solutions ( solution 5 and 6 ) by rapidly mixing streams of two dissimilar solutions (i.e., mixing solutions 5A and 5B to yield solution 5 and mixing solutions 6A and 6B to yield solution 6 ) ( 58, 61 ) and delivering the newly mixed solution to the tubule within 200 ms. All solutions had osmolalities of 300 ± 2 mosmol/kgH 2 O.
Measurement of J and J V
Our measurement of J (pmol·min -1 ·mm tubule length -1 ) and J V (nl·min -1 ·mm -1 ) was similar to that used by McKinney and Burg ( 32 ) and nearly identical to our previous approach (59-61). We determined total CO 2 in aliquots of the perfusate and collected fluid using a WPI "NanoFlo" device (World Precision Instruments, Sarasota, FL). The one difference from our previous study is that, because Diagnostic Kit 132-A (Sigma-Aldrich) for total CO 2 determinations was no longer commercially available, we generated our own reagents, a total of five components, according to the protocol described by Hall et al. ( 25 ) and Krömer et al. ( 30 ).
1 ) Made fresh on the day of analysis, this solution contained (in mM) 10 MgCl 2 6 H 2 O, 1 EDTA, 2 dithiothreitol, 100 Tris·HCl, 15 sodium azide (NaN 3 ), and 30 NaOH to adjust pH to 7.7. To eliminate any CO 2, we bubbled this solution with 100% O 2 for 30 min at 4°C. On the day of analysis, we added components 2-5 from concentrated stock solutions.
2 ) Phosphoenol pyruvic acid (151872, MP Biomedicals, Aurora, OH) was added to a final concentration of 2.2 mM from a 220 mM stock solution that also contained (in mM) 100 Tris·HCl and 15 NaN 3, pH 7.7 (stock stored at -20°C).
3 ) -NADH (10168, MP Biomedicals) was added to a final concentration of 0.32 mM from a 32 mM stock solution that also contained (in mM) 100 Tris·HCl, 15 NaN 3, pH 7.7 (powder stored at -20°C, and the stock solution made fresh).
4 ) Phosphoenol pyruvate carboxylase (153532, MP Biomedicals) was added to a final concentration of 0.275 U/ml from a 100 U/ml stock solution that also contained (in mM) 6 MgCl 2 6 H 2 O, 1 EDTA, 2 dithiothreitol, 40 Tris·HCl, and 15 NaN 3 as well as 10% glycerol, pH 7.7 (stock stored at -80°C).
5 ) Malate dehydrogenase (151581, MP Biomedicals) was added to a final concentration of 15.5 U/ml from a 10,000 U/ml stock solution in 70% (NH 4 ) 2 SO 4 (stock stored at 4°C).
We computed J V and J values from the equations described previously ( 59 ).
Data Analysis
The values that we report for J (or J V ) in the first collection period are unnormalized, mean values. The values that we report for J (or J V ) in the second collection period are normalized, mean values computed as described previously ( 60, 61 ). Briefly, in each experiment, we divided the J (or J V ) value obtained during the second collection period by the comparable value obtained during the first collection period; the result was a pair of second/first collection period ratios. We then multiplied 1 ) the second/first ratio for J (or J V ) in a particular experiment by 2 ) the unnormalized mean J (or J V ) value that we obtained during the first collection periods in a series of experiments following the identical protocol.
For comparisons of two means, two-tailed paired t -tests were performed using the Analysis Toolpack of Microsoft Excel. For comparisons of more than two means, one-way ANOVA and Dunnett's multiple comparison were performed using KaleidaGraph (Version 4, Synergy Software). Results are given as means ± SE, with the number of tubules ( n ) from which it was calculated.
RESULTS
Evaluation of DMSO
In this study, we added the inhibitors, predissolved in DMSO, to the bath solution at a final DMSO concentration of 1:20,000. Therefore, we first examined the effect of basolateral 1:20,000 DMSO on J and J V. Figure 1, A and B, summarizes experiments in which DMSO was present in the bath during both collection periods. During the first collection period, with equilibrated 5% CO 2 /22 mM HCO3- in the lumen ( solution 2 ) and in the bath ( solution 4 ), the J and J V values (grey bars) were similar to the historical averages ( 61 ). However, during the second collection period, when we switched the bath to an out-of-equilibrium solution ( solution 6 ) containing 20% CO 2, 22 mM HCO3-, and pH 7.40 (filled bars), the 20% CO 2 failed to increase J. Thus we suspected that an extended exposure to DMSO, even at a dilution of 1:20,000, reduces J during the second collection period.
Fig. 1. Evaluation of DMSO. A, C, and E : HCO3- reabsorption rate ( J ). B, D, and F : volume reabsorption rate ( J V ). Top and bottom : luminal solution was solution 2 throughout the experiment. During the first collection period ( left bars), the basolateral solution was always solution 4, with or without 1:20,000 DMSO, as indicated. During the second collection period ( right bars), the basolateral solutions always contained 1:20,000 DMSO added to either solution 4 (5% CO 2 ) or solution 6 (20% CO 2 ), as indicated. In the figure, the 2 bars represent the results of paired experiments; the same 11 tubules served as controls ( left bars) in C - F. Values are means ± SE, with nos. of tubules in parentheses. As indicated, the difference between left and right bars in C is statistically significant in a 2-tailed unpaired t -test. ** P < 0.01.
Figure 1, C and D, summarizes a study identical to that above, except that DMSO was present in the bath only during the second collection period. In these experiments, the switch from 5 to 20% CO 2 increased J ( P = 0.0032, 2-tailed t -test, unpaired) to its historical value, ( 61 ) which is 50% higher than the value in the first collection period.
As a final check, we examined the effect of adding DMSO during the second collection period when the bath contained 5% CO 2 throughout the experiment. Figure 1, E and F, shows that the addition of DMSO had no effect on either J or J V. Thus, in our remaining experiments, we added DMSO only in the second collection period, where its effects appear to be negligible.
Effects of Basolateral Genistein on Basolateral CO 2 Dependence of J and J V
To test the hypothesis that the PT's CO 2 -sensing mechanism may require a RTK or an sTK, we examined the effects of basolateral genistein ( 1, 24 ) on J and J V. We used OOE solutions to vary basolateral [CO 2 ] from 0 to 20%, while keeping [HCO3-] BL fixed at 22 mM and pH BL fixed at 7.40.
In the first group of experiments, we examined effect of basolateral 7 µM genistein with equilibrated 5% CO 2 /22 mM HCO3- in both the lumen ( solution 2 ) and the bath ( solution 4 ). During the first collection period, no drug was present ( in Fig. 2 ). During the second collection period, we added 7 µM genistein to solution 4 ( in Fig. 2 ). Although basolateral 7 µM genistein reduced the mean J by 25%, from 55 ± 3 to 41 ± 7 pmol·min -1 ·mm -1 ( n = 4 paired experiments), the difference was not statistically significant ( P = 0.07, 2-tailed t -test, paired). Then, we increased the concentration of basolateral genistein to 17.5 µM and repeated the above protocol in six paired experiments. Basolateral 17.5 µM genistein reduced the mean J by 45%, from 55 ± 3 to 30 ± 6 pmol·min -1 ·mm -1. In analyzing the effects of basolateral 17.5 µM genistein in basolateral 5% CO 2, we employed a one-way ANOVA for three groups: 1 ) control data ( in Fig. 2 A ), 2 ) 7 µM genistein ( in Fig. 2 A ), and 3 ) 17.5 µM genistein (filled pentagon in Fig. 2 A ). The overall P value was 0.003. Dunnett's multiple comparison shows that although the effect of basolateral 7 µM genistein on J was not statistically significant ( P = 0.48, n = 4), the effect of basolateral 17.5 µM genistein was significant ( P = 0.0015, n = 6).
Fig. 2. Effects of basolateral 7 or 17.5 µM genistein on the basolateral CO 2 concentration ([CO 2 ] BL ) dependence of J ( A ) and J V ( B ). B : J V. In both A and B, the luminal solution was solution 2 throughout the entire experiment. The basolateral solutions were solutions 5 (0% CO 2 ), 4 (5% CO 2 ), and 6 (20% CO 2 ) during the first collection periods and solutions 5, 4, and 6 plus 7 or 17.5 µM genistein during the second collection periods. Values are means ± SE, with nos. of tubules in parentheses. At 0 and 20% CO 2, the statistical comparisons between and filled pentagons at the same [CO 2 ] BL were made using a paired 2-tailed t -test (** P < 0.01). At 5% CO 2, the statistical comparison among,, and the filled pentagon at the same [CO 2 ] BL was made using a 1-way ANOVA for 3 groups; Dunnett's multiple comparison indicates that the only significant difference was between and filled pentagon.
Next, we extended our experiments with basolateral 17.5 µM genistein to [CO 2 ] BL levels of 0 and 20%. We employed the same protocol as for a [CO 2 ] BL of 5%; namely, in the first collection period 0% (or 20%) CO 2 was present without the drug, and in the second collection period 0% (or 20%) CO 2 was present with genistein. As shown in Fig. 2 A, at a [CO 2 ] BL of 0%, 17.5 µM genistein did not significantly affect J, which changed J from 28 ± 4 to 19 ± 2 pmol·min -1 ·mm -1 ( P = 0.072, 2-tailed t -test, paired). In contrast, at a [CO 2 ] BL of 20%, 17.5 µM genistein significantly reduced J from 72 ± 2 to 29 ± 2 pmol·min -1 ·mm -1 ( P = 0.000007, 2-tailed t -test, paired). An ANOVA shows that, in the presence of 17.5 µM genistein, the J data at [CO 2 ] BL values of 0, 5, and 20% (i.e., the 3 filled pentagons in Fig. 2 A ) are not statistically significant (overall P value was 0.13; Dunnett's multiple comparison shows P = 0.11 at [CO 2 ] BL levels of 0% and P = 0.96 at [CO 2 ] BL levels of 20%). Thus basolateral 17.5 µM genistein eliminates the stimulatory effect produced by basolateral CO 2 in the range 0-20%.
As far as J V is concerned, at a [CO 2 ] BL of 5%, the overall P value in a one-way ANOVA was 0.28, and Dunnett's multiple comparison indicates that, relative to the control condition, neither basolateral 7 µM genistein ( in Fig. 2 B, P = 0.66) nor basolateral 17.5 µM genistein (filled pentagon in Fig. 2 B, P = 0.45) had a significant effect on J V. In paired, two-tailed t -tests, the effects of basolateral 17.5 µM genistein on J V were not statistically significant at [CO 2 ] BL values of 0% ( P = 0.26) or 20% ( P = 0.80).
Effects of Basolateral PP2 on Basolateral CO 2 Dependence of J and J V
The genistein data suggest that the transduction of the basolateral CO 2 signal to an increase in J requires the activity of tyrosine kinase. Our next step was to examine the effect of PP2, a potent and relatively specific inhibitor of the Src family of sTKs ( 26, 53 ), on the basolateral CO 2 dependence of J ( Fig. 3 A ). The protocol for this and the remaining series of experiments was somewhat different from that in Fig. 2. As before, we perfused the lumen with solution 2 throughout the entire experiment. However, during the first collection period shown in Fig. 3, we always perfused the bath with solution 4 (equilibrated 5% CO 2 /22 mM HCO3-) without any inhibitor ( at [CO 2 ] BL = 5%). During the second collection period, we perfused the bath with solution 5 (i.e., 0% CO 2 ) ± inhibitor, 4 (i.e., 5% CO 2 ) ± inhibitor, or 6 (i.e., 20% CO 2 ) ± inhibitor. The control (i.e., drug-free) data at [CO 2 ] BL = 0% ( ) are from an earlier study ( 61 ), as are 13 of the control points at [CO 2 ] BL = 5% ( ) ( 61 ). The 5% control data are augmented by 59 points from the current study. Finally, seven of the control (i.e., drug-free) points at [CO 2 ] BL = 20% ( ) are from an earlier study ( 61 ), augmented by six DMSO points from Fig. 1 C in the present study. As described in METHODS, the values of J (or J V ) in the second collection period were normalized to the mean J (or J V ) value computed from 72 experiments during the first collection period.
Fig. 3. Effects of basolateral 250 or 2,000 nM 4-amino-5-(4-chlorophenyl)-7-( t -butyl)pyrazolo[3,4- d ]pyrimidase (PP2) on the [CO 2 ] BL dependence of J ( A ) and J V ( B ). In both A and B, the luminal solution was solution 2 throughout the entire experiment. The basolateral solutions were solution 4 (5% CO 2 ) during the first collection periods and solutions 5 (0% CO 2 ), 4 (5% CO 2 ), and 6 (20% CO 2 ), to which was sometimes added either 250 or 2,000 nM PP2, during the second collection periods. Values are means ± SE, with nos. of tubules in parentheses. At 0 and 5% CO 2, the statistical comparisons were made using a 1-way ANOVA for 3 groups (the 2 groups in this figure and the data for 35 nM PD168393 in Fig. 4 ). At 20% CO 2, the statistical comparisons were made using a 1-way ANOVA for 5 groups (the 3 groups in this figure and the data for 35 nM PD168393 and 10 nM BPIQ-I in Fig. 4 ). According to Dunnett's multiple comparison, none of the differences in this figure were statistically significant.
The J data obtained in the presence of basolateral 250 nM PP2, 50-fold greater than the published in vitro K i ( 26 ), at [CO 2 ] BL levels of 0, 5, or 20% at a fixed [HCO3-] BL of 22 mM and a fixed pH BL of 7.4, are summarized in Fig. 3A ( ). Because we used the same control data in our statistical analysis of the J (or J V ) data in Figs. 3 and 4, we applied a one-way ANOVA to analyze all of the data at 0% CO 2 in Figs. 3 and 4. Similarly, we applied a one-way ANOVA to analyze all of the data at 5% and again at 20% CO 2 in Figs. 3 and 4. The overall P value for three J groups was 0.87 for [CO 2 ] BL = 0% and was 0.018 for [CO 2 ] BL = 5%. For [CO 2 ] BL = 20%, the overall P value for five J groups was <0.0001. For Fig. 3 A, Dunnett's multiple comparison indicates that basolateral 250 nM PP2 had no significant effect on J at 0 ( P = 0.85), 5 ( P = 0.96), or 20% CO 2 ( P = 0.48). Even when we increased its concentration eightfold to 2 µM, PP2 did not have a significant effect ( P = 0.65) on J at a [CO 2 ] BL of 20% ( in Fig. 3 A ).
Fig. 4. Effects of basolateral 35 nM PD168393 or basolateral 10 nM BPIQ-I on [CO 2 ] BL dependence of J ( A ) and J V ( B ). The was moved to the right for legibility. In both A and B, the luminal solution was solution 2 throughout the entire experiment. The basolateral solutions were solution 4 (5% CO 2 ) during the first collection periods and solutions 5 (0% CO 2 ), 4 (5% CO 2 ), and 6 (20% CO 2 ), to which was sometimes added either 35 nM PD168393 or 10 nM BPIQ-I, during the second collection periods. Values are means ± SE, with nos. of tubules in parentheses. At 0 and 5% CO 2, the statistical comparisons were made using a 1-way ANOVA for 3 groups (the 2 groups in this figure and the data for 250 nM PP2 in Fig. 3 ). At 20% CO 2, the statistical comparisons were made using a 1-way ANOVA for 5 groups (the 3 groups in this figure and the data for 250 and 2,000 nM PP2 in Fig. 3 ). According to Dunnett's multiple comparison, the at 5% CO 2 is significantly different from the for the J data, and both the and filled diamond at 20% CO 2 are significant different from the for J data. * P < 0.05, ** P < 0.01.
We analyzed the J V data in the same way as the J data described above. For the ANOVA, the overall P value for three J V groups was 0.34 for [CO 2 ] BL = 0% and was 0.89 for [CO 2 ] BL = 5%. For [CO 2 ] BL = 20%, the overall P value for five J V groups was 0.78. Compared with the control situation with no added inhibitor ( in Fig. 3 B ), neither basolateral 250 nM PP2 ( in Fig. 3 B ) nor 2 µM PP2 ( in Fig. 3 B ) had a significant effect on J V at any value of [CO 2 ] BL.
Our data suggest that the J response of the rabbit S2 PT to basolateral CO 2 does not involve Src. PP2 is also inhibits the Bcr-Abl fusion protein ( 53 ), which forms as the Philadelphia chromosome creates a fusion protein of Bcr and Abl ( 23 ), which is an sTK, and the Kit family ( 53 ) of RTKs. Thus our data also make it unlikely that these kinases are involved in transducing the CO 2 signal.
Effects of Basolateral PD168393 and BPIQ-I on the Basolateral CO 2 dependence of J and J V
We next examined the effect of PD168393, which is a high-affinity inhibitor ( K i 0.7 nM) ( 19 ) of the erbB family of RTKs. The drug covalently reacts with a specific cysteine in the ATP binding pocket (see Table 2 ) and is thought to be rather specific. Our protocol was the same as in Fig. 3, and the control data in Fig. 4 A ( ) are the same as those presented in Fig. 3 A. The dependence of J on [CO 2 ] BL in the presence of basolateral 35 nM PD168393 is summarized in Fig. 4 A ( ). The statistical analysis of these J data was part of the same J ANOVA that we used to assess the PP2 data in Fig. 3 A. Compared with the control condition, PD168393 did not significantly affect J at [CO 2 ] BL of 0% ( P = 0.998), according to Dunnett's multiple comparison. However, at a [CO 2 ] BL of 5%, the drug significantly decreased J from 64 ± 4 to 39 ± 4 pmol·min -1 ·mm -1 ( P = 0.011). Moreover, at a [CO 2 ] BL of 20%, PD168393 significantly decreased J from 94 ± 2 to 23 ± 3 pmol·min -1 ·mm -1 ( P < 0.0001). Thus basolateral 35 nM PD168393 eliminates the PT's J response to basolateral CO 2 in the range 0-20%.
Table 2. Tyrosine kinases with a conserved cysteine residue in the ATP-binding pocket
To further examine the possibility that an erbB or a related tyrosine kinase might be involved in the CO 2 signal-transduction cascade, we examined the effect of a second high-affinity inhibitor of the erbB family, BPIQ-I, which acts by competing with ATP in the binding pocket ( K i 25 pM) ( 44 ). At the single [CO 2 ] BL of 20%, basolateral 10 nM BPIQ-I ( in Fig. 4 A ) significantly decreased J from 94 ± 2 to 36 ± 6 pmol·min -1 ·mm -1 ( P < 0.0001).
The statistical analysis of the J V data for PD168393 and BPIQ-I was also part of the same J ANOVA that we used to assess the PP2 data in Fig. 3 B. Compared with the control condition with no added inhibitors ( in Fig. 4 B ), neither PD168393 ( ) nor 10 nM BPIQ-I ( ) produced a significant effect on J V at any level of [CO 2 ] BL according to Dunnett's multiple comparison.
DISCUSSION
Use of Out-of-Equilibrium Solutions
Out-of-equilibrium solutions are a powerful tool for performing the experiments outlined in this paper. However, carbonic anhydrase (CA) activity at the extracellular surface of the basolateral membrane would tend to cause a local degradation of the OOE state. Blockade of this activity would independently reduce J and thus be impractical. On the other hand, as we raise [CO 2 ] from 0 to 5 to 20%, we clearly see major increases in J (e.g., see Fig. 3 ). Moreover, as shown previously, raising [HCO3-] BL from 0 to 44 mM causes major decreases in J. Thus any degradation of the OOE state by basolateral CAs must be minor. Nevertheless, to the extent that some degradation of the OOE state does occur, our data underestimate the true magnitude of the response of the tubule to isolated changes in [CO 2 ]. If a drug added to the PT inhibited all CAs, the result would be an enhancement of the OOE state but a decrease in J. We are unaware of any reports that the drugs used in the present study inhibit CAs.
Inhibitor Specificity
Our previous study ( 61 ) demonstrated that PT cells somehow sense acute decreases in [CO 2 ] BL and respond by lowering J and somehow sense acute increases in [CO 2 ] BL and respond by raising J. In addition, the PT cells change the reabsorption of other solutes ( J Other ) in a direction opposite to that of J, thereby minimizing changes in J V. A key question is how PT cells transduce the CO 2 signal. We hypothesized that an early element in the CO 2 signal-transduction cascade in PT cells is either a receptor tyrosine kinase or a receptor-associated (or soluble) tyrosine kinase.
Genistein is a broad-spectrum tyrosine-kinase inhibitor ( 24 ) that blocks both sTKs (e.g., Src family) and RTKs (e.g., erbB family) but exhibits no nonspecific inhibition with other kinases (PKA, PKC) at the concentration we used ( 24 ). As showed in Fig. 2 A, genistein eliminates the ability of the PT to respond to increases in [CO 2 ] BL. These data are consistent with the hypothesis that the CO 2 signal-transduction pathway in the PT cell requires either an sTK or an RTK.
PP2 was developed as a specific inhibitor of the Src family of sTKs ( 4 ), but it also inhibits Bcr-Abl ( 53 ), which is an sTK, as well as Kit, ( 53 ) which is an RTK. Using PTs or PT-like cell lines, others have found that 1-10 µM PP2 can significantly reduce a variety of physiological responses ( 2, 31, 43, 56 ). As shown in Fig. 3 A, even at a basolateral concentration that is 400-fold greater than its published K i for Src ( 26 ), PP2 did not reduce the CO 2 -stimulated increase in J.
PD168393 is a high-affinity, irreversible inhibitor of members of the erbB family of RTKs. It acts by alkylating a cysteine residue [i.e., Cys-773 in human EGF receptor (EGFR)] in the ATP binding pocket ( 19 ). Figure 4 A shows that, at a concentration 50-fold greater than its published K i, basolateral PD168393 totally eliminates the response to changes in [CO 2 ] BL. Although many authors regard PD168393 as a specific erbB inhibitor, it is in fact impossible to know how specific it is without assaying all tyrosine kinases. A search of human tyrosine kinases reveals a total of eight human tyrosine kinases that have a cysteine residue at a position comparable to Cys-773 in the ATP binding pocket of erbB1 ( Table 2 ). In principle, each of these tyrosine kinases, five RTKs and three sTKs, is a potential target of PD168393. Of these, erbB1 and erbB2 are known to be present in the PT ( 33, 35 ), and mRNA transcripts for erbB4 (as well as erbB3) have been reported for the whole kidney ( 38, 39 ). TEK or Tie2 ( 17 ), the angiopoietin receptor ( 55 ), is expressed almost exclusively in endothelial cells as well as certain cancer cells. EphB3 or Hek2 ( 9 ), a receptor for the ephrin-B family, plays a key role in neural development but is also expressed in the kidney. ITK plays an important role in T cell activation ( 28 ). BLK, which is primarily expressed in hematopoeitic cells ( 27 ), is a member of the Src family and thus should have been inhibited by PP2. Finally, JAK3 is expressed primarily in hematopoietic cells, where it interacts with cytokine receptors and plays a role in development and cell activation ( 54 ). Thus of the proteins listed in Table 2, erbB1, 2 and 4 as well as EphB3 are prime candidates as targets of PD168393 in our experiments.
BPIQ-I is also a high-affinity inhibitor of EGFR (i.e., erbB1) as well as erbB2 and erbB4. It acts by competing with ATP ( 44 ). At a concentration 50-fold greater than its published K i, basolateral BPIQ-I decreased the CO 2 -sensitive component of J by about two-thirds (i.e., decreased total J by 40%) at a [CO 2 ] BL of 20% (data not shown). At a concentration 400-fold greater than its published K i, BPIQ-I blocked virtually 100% of the CO 2 -sensitive component of J.
Our observation that both PD-168393 and BPIQ-I block the response to CO 2 increases the odds that, in our experiments, both drugs produce the observed effect by acting on a member of the erbB family.
Model
Previous work from our laboratory showed that simultaneously adding CO 2 and HCO3- to the basolateral, but not the luminal, side of the S3 segment of the rabbit PT triggers a fourfold increase in total H + extrusion ( 14 ). The results of our most recent study ( 61 ) indicate that this increase is due to basolateral CO 2. The data in the present study suggest that the tubule's response to altered [CO 2 ] BL requires one of a small group of tyrosine kinases (see Table 2 ). Figure 5 summarizes several potential mechanisms by which the PT cell might transduce the CO 2 signal. One possibility ( Fig. 5, 1 ) is that an RTK is simply permissive for the effect of CO 2 on J. Thus blocking the RTK would eliminate the response to CO 2 even though the RTK would not be downstream from the CO 2 receptor. Second ( Fig. 5, 2 ), a CO 2 receptor could activate an PP2-insensitive sTK, which would then raise J. Third ( Fig. 5, 3 ), a CO 2 receptor could transactivate an RTK ( 57 ), which would, in turn, raise J. In a variant of this third pathway, the CO 2 receptor could trigger a metalloproteinase to release an EGF-like proligand, which would, in turn, activate EGFR ( 40 ). In yet another variant of this third pathway, CO 2 could block the ability of a receptor protein tyrosine phosphatase to transinactivate an sTK or RTK. Finally ( Fig. 5, 4 ), the CO 2 receptor could itself be an RTK. Clearly, additional experiments would be required to establish the molecular identity of the tyrosine kinase involved in the response to CO 2.
Fig. 5. Model of how basolateral CO 2 might stimulate HCO3- reabsorption by the proximal tubule. On the basis of inhibitor studies, a receptor tyrosine kinase (RTK), an early element in a signal transduction cascade, could be 1 ) permissive, 2 ) activated by the CO 2 receptor, 3 ) transactivated by the CO 2 receptor, or 4 ) the CO 2 receptor itself. sTK, receptor-associated (or soluble) tyrosine kinase; NHE3, type 3 Na-H exchanger; NBCe1-A, Na-HCO3- cotransporter.
In the PT, the CO 2 sensor triggers not only a compensatory rise in J but also a decrease in the reabsorption of other solutes, thereby stabilizing J V ( 61 ). Thus we would predict that the PT cell can modulate acid-base transport without secondary effects on blood pressure. In Fig. 6 A ( connected by solid lines) are the control J data (i.e., obtained in the absence of any drugs) in the present study, replotted from Fig. 3 A. Also shown are the J Other values that we computed from the corresponding J and J V data ( connected by dashed lines).
Fig. 6. Effects of basolateral drugs on the calculated reabsorption of other solutes ( J Other ). A : drug-free control experiments. B : basolateral PP2 experiments. C : basolateral PD-168393 or BPIQ-I experiments. For each panel, the J data (points connected by solid lines) are taken from Figs. 3 or 4, whereas the J other values (points connected by dashed lines) were calculated from the corresponding J and J V values using the equation (2 x J + J other )/ J V = 300 mosmol/kgH 2 O.
In Fig. 6 B are the J data obtained in the presence of 250 nM PP2, replotted from Fig. 3 A ( connected by solid lines). Also shown are the J Other values that we computed from the corresponding J and J V data ( connected by dashed lines) and the data for 2,000 nM PP2 (adjacent ). For both control and PP2, the J Other values at 0% are significantly greater than at 5% CO 2, whereas those at 20% are not significantly different from those at 5%. Thus the trend is for J Other to fall as [CO 2 ] BL increases.
Finally, in Fig. 6 C are the J data obtained in the presence of 35 nM PD168393 (3 connected by solid lines) and 10 nM BPIQ-I (adjacent ), respectively, replotted from Fig. 4 A. Also shown are the corresponding J Other values ( connected by dashed lines and the adjacent ). For PD168393, the J Other value at 0% CO 2 is not significantly different from that at 5%, whereas the J Other value at 20% CO 2 is significantly different from that at 5%. For BPIQ-I, the J Other value at 20% CO 2 ( ) is similar to the PD168393 data at 0 and 5% CO 2. Thus with blockade of the CO 2 response, the trend is for J Other to be relatively stable as [CO 2 ] BL rises. In other words, the effective tyrosine kinase inhibitors not only blocked the J response to CO 2, they also tended to block the J Other response. These results are consistent with the hypothesis that the tyrosine kinase targeted by PD168393 and BPIQ-I are relatively early in the signal-transduction cascade, that is, before the bifurcation to stimulate J and inhibit J Other.
It is interesting to speculate that CO 2 sensors similar to that in the PT may be present in other cells that perform large amounts of acid-base transport, as in the choroid plexus, ciliary body, stomach, pancreatic ducts, intestines, and male genital tract. Thus understanding how the PT transduces the CO 2 signal may provide important clues about how such cells regulate acid-base transport.
GRANTS
This work was funded by National Institute of Diabetes and Digestive and Kidney Diseases Program Project Grant PO1-DK-17433. P. Bouyer was supported by a National Kidney Foundation fellowship.
ACKNOWLEDGMENTS
We thank Duncan Wong for computer support.
【参考文献】
Akiyama T, Ishida J, Nakagawa S, Ogawara H, Watanabe S, Itoh N, Shibuya M, and Fukami Y. Genistein, a specific inhibitor of tyrosine-specific protein kinases. J Biol Chem 262: 5592-5595, 1987.
Alexander LD, Alagarsamy S, and Douglas JG. Cyclic stretch-induced cPLA2 mediates ERK 1/2 signaling in rabbit proximal tubule cells. Kidney Int 65: 551-563, 2004.
Aronson PS. Mechanisms of active H + secretion in the proximal tubule. Am J Physiol Renal Fluid Electrolyte Physiol 245: F647-F659, 1983.
Bain J, McLauchlan H, Elliott M, and Cohen P. The specificities of protein kinase inhibitors: an update. Biochem J 371: 199-204, 2003.
Baum M, Quigley R, and Quan A. Effect of luminal angiotensin II on rabbit proximal convoluted tubule bicarbonate absorption. Am J Physiol Renal Physiol 273: F595-F600, 1997.
Bichara M, Paillard M, Leviel F, and Gardin J. Hydrogen transport in rabbit kidney proximal tubules-Na:H exchange. Am J Physiol Renal Fluid Electrolyte Physiol 238: F445-F451, 1980.
Bichara M, Paillard M, Leviel F, Pringent A, and Gardin JP. Na:H exchange and the primary H pump in the proximal tubule. Am J Physiol Renal Fluid Electrolyte Physiol 244: F165-F171, 1983.
Bleecker AB and Kende H. Ethylene: a gaseous signal molecule in plants. Annu Rev Cell Dev Biol 16: 1-18, 2000.
Bohme B, Holtrich U, Wolf G, Luzius H, Grzeschik KH, Strebhardt K, and Rubsamen-Waigmann H. PCR mediated detection of a new human receptor-tyrosine-kinase, HEK 2. Oncogene 8: 2857-2862, 1993.
Boron WF and Boulpaep EL. Intracellular pH regulation in the renal proximal tubule of the salamander. Na-H exchange. J Gen Physiol 81: 29-52, 1983.
Boron WF and Boulpaep EL. Intracellular pH regulation in the renal proximal tubule of the salamander: basolateral HCO3- transport. J Gen Physiol 81: 53-94, 1983.
Burg M, Grantham J, Abramow M, and Orloff J. Preparation and study of fragments of single rabbit nephrons. Am J Physiol 210: 1293-1298, 1966.
Chang C, Kwok SF, Bleecker AB, and Meyerowitz EM. Arabidopsis ethylene-response gene ETR1: similarity of product to two-component regulators. Science 262: 539-544, 1993.
Chen LK and Boron WF. Acid extrusion in S3 segment of rabbit proximal tubule. II. Effect of basolateral CO 2 /HCO3-. Am J Physiol Renal Fluid Electrolyte Physiol 268: F193-F203, 1995.
Clark KL, Larsen PB, Wang X, and Chang C. Association of the Arabidopsis CTR1 Raf-like kinase with the ETR1 and ERS ethylene receptors. Proc Natl Acad Sci USA 95: 5401-5406, 1998.
Cogan MG and Alpern RJ. Regulation of proximal bicarbonate reabsorption. Am J Physiol Renal Fluid Electrolyte Physiol 247: F387-F395, 1984.
Dumont DJ, Yamaguchi TP, Conlon RA, Rossant J, and Breitman ML. tek, a novel tyrosine kinase gene located on mouse chromosome 4, is expressed in endothelial cells and their presumptive precursors. Oncogene 7: 1471-1480, 1992.
Ecker JR. The ethylene signal transduction pathway in plants. Science 268: 667-675, 1995.
Fry DW, Bridges AJ, Denny WA, Doherty A, Greis KD, Hicks JL, Hook KE, Keller PR, Leopold WR, Loo JA, McNamara DJ, Nelson JM, Sherwood V, Smaill JB, Trumpp-Kallmeyer S, and Dobrusin EM. Specific, irreversible inactivation of the epidermal growth factor receptor and erbB2, by a new class of tyrosine kinase inhibitor. Proc Natl Acad Sci USA 95: 12022-12027, 1998.
Gilles-Gonzalez MA, Ditta GS, and Helinski DR. A hæmoprotein with kinase activity encoded by the oxygen sensor of Rhizobium meliloti. Nature 350: 170-172, 1991.
Gilles-Gonzalez MA and Gonzalez G. Heme-based sensors: defining characteristics, recent developments, and regulatory hypotheses. J Inorg Biochem 99: 1-22, 2005.
Gluck SL, Underhill DM, Iyori M, Holliday LS, Kostrominova TY, and Lee BS. Physiology and biochemistry of the kidney vacuolar H + -ATPase. Annu Rev Physiol 58: 427-445, 1996.
Goldman JM and Melo JV. Targeting the BCR-ABL tyrosine kinase in chronic myeloid leukemia. N Engl J Med 344: 1084-1086, 2001.
Groundwater PW, Solomons KR, Drewe JA, and Munawar MA. Protein tyrosine kinase inhibitors. Prog Med Chem 33: 233-329, 1996.
Hall NP, Cornelius MJ, and Keys AJ. The enzymatic determination of bicarbonate and CO 2 in reagents and buffer solutions. Anal Biochem 132: 152-157, 1983.
Hanke JH, Gardner JP, Dow RL, Changelian PS, Brissette WH, Weringer EJ, Pollok BA, and Connelly PA. Discovery of a novel, potent, and Src family-selective tyrosine kinase inhibitor. Study of Lck- and FynT-dependent T cell activation. J Biol Chem 271: 695-701, 1996.
Hsueh RC and Scheuermann RH. Tyrosine kinase activation in the decision between growth, differentiation, and death responses initiated from the B cell antigen receptor. Adv Immunol 75: 283-316, 2000.
Kanner SB and Perez-Villar JJ. Altering T-cell activation by targeting the multidomain tyrosine kinase Itk. Trends Immunol 24: 249-253, 2003.
Kieber JJ, Rothenberg M, Roman G, Feldmann KA, and Ecker JR. CTR1, a negative regulator of the ethylene response pathway in Arabidopsis, encodes a member of the raf family of protein kinases. Cell 72: 427-441, 1993.
Krömer S, Gardeström P, and Samuelsson G. Regulation of the supply of cytosolic oxaloacetate for mitochondrial metabolism via phosphoenolpyruvate carboxylase in barley leaf protoplasts. I. The effect of covalent modification on PEPC activity, pH response, and kinetic properties. Biochim Biophys Acta 1289: 343-350, 1996.
Liu J, Kesiry R, Periyasamy SM, Malhotra D, Xie Z, and Shapiro JI. Ouabain induces endocytosis of plasmalemmal Na/K-ATPase in LLC-PK 1 cells by a clathrin-dependent mechanism. Kidney Int 66: 227-241, 2004.
McKinney TD and Burg MB. Bicarbonate and fluid absorption by renal proximal straight tubules. Kidney Int 12: 1-8, 1977.
Mori S, Akiyama T, Yamada Y, Morishita Y, Sugawara I, Toyoshima K, and Yamamoto T. C-erbB-2 gene product, a membrane protein commonly expressed on human fetal epithelial cells. Lab Invest 61: 93-97, 1989.
Murer H, Hopfer U, and Kinne R. Sodium/proton antiport in brush-border-membrane vesicles isolated from rat small intestine and kidney. Biochem J 154: 597-604, 1976.
Nakanishi K, Sweeney W Jr, and Avner ED. Segment-specific c-ErbB2 expression in human autosomal recessive polycystic kidney disease. J Am Soc Nephrol 12: 379-384, 2001.
Nakhoul NL, Lopes AG, Chaillet JR, and Boron WF. Intracellular pH regulation in the S3 segment of the rabbit proximal tubule in HCO3--free solutions. J Gen Physiol 92: 369-393, 1988.
Ouaked F, Rozhon W, Lecourieux D, and Hirt H. A MAPK pathway mediates ethylene signaling in plants. EMBO J 22: 1282-1288, 2003.
Plowman GD, Culouscou JM, Whitney GS, Green JM, Carlton GW, Foy L, Neubauer MG, and Shoyab M. Ligand-specific activation of HER4/p180 erbB4, a fourth member of the epidermal growth factor receptor family. Proc Natl Acad Sci USA 90: 1746-1750, 1993.
Plowman GD, Whitney GS, Neubauer MG, Green JM, McDonald VL, Todaro GJ, and Shoyab M. Molecular cloning and expression of an additional epidermal growth factor receptor-related gene. Proc Natl Acad Sci USA 87: 4905-4909, 1990.
Prenzel N, Zwick E, Daub H, Leserer M, Abraham R, Wallasch C, and Ullrich A. EGF receptor transactivation by G-protein-coupled receptors requires metalloproteinase cleavage of proHB-EGF. Nature 402: 884-888, 1999.
Quigley R and Baum M. Effects of epidermal growth factor and transforming growth factor- on rabbit proximal tubule solute transport. Am J Physiol Renal Fluid Electrolyte Physiol 266: F459-F465, 1994.
Quigley R, Kennerly DA, Sheu JN, and Baum M. Stimulation of proximal convoluted tubule phosphate transport by epidermal growth factor: signal transduction. Am J Physiol Renal Fluid Electrolyte Physiol 269: F339-F344, 1995.
Ren L, Blanchette JB, White LR, Clark SA, Heffner DJ, Tibbles LA, and Muruve DA. Soluble fibronectin induces chemokine gene expression in renal tubular epithelial cells. Kidney Int 68: 2111-2120, 2005.
Rewcastle GW, Palmer BD, Thompson AM, Bridges AJ, Cody DR, Zhou H, Fry DW, McMichael A, and Denny WA. Tyrosine kinase inhibitors 9 synthesis and evaluation of fuse tricyclic quinazoline analogues as ATP site inhibitors of the tyrosine kinase activity of the epidermal growth factor receptor. J Med Chem 39: 918-928, 1996.
Rodgers KR. Heme-based sensors in biological systems. Curr Opin Chem Biol 3: 158-167, 1999.
Rodriguez FI, Esch JJ, Hall AE, Binder BM, Schaller GE, and Bleecker AB. A copper cofactor for the ethylene receptor ETR1 from Arabidopsis. Science 283: 996-998, 1999.
Romero MF, Fulton CM, and Boron WF. The SLC4 family of HCO3- transporters. Pflügers Arch 447: 495-509, 2004.
Romero MF, Hediger MA, Boulpaep EL, and Boron WF. Expression cloning and characterization of a renal electrogenic Na + /HCO3- cotransporter. Nature 387: 409-413, 1997.
Schaller GE and Bleecker AB. Ethylene-binding sites generated in yeast expressing the Arabidopsis ETR1 gene. Science 270: 1809-1811, 1995.
Schwartz GJ. Na + -dependent H + efflux from proximal tubule: evidence for reversible Na + -H + exchange. Am J Physiol Renal Fluid Electrolyte Physiol 241: F380-F385, 1981.
Schwartz GJ. Physiology and molecular biology of renal carbonic anhydrase. J Nephrol 15: S61-S74, 2002.
Sly WS and Hu PY. Human carbonic anhydrases and carbonic anhydrase deficiencies. Annu Rev Biochem 64: 375-401, 1995.
Tatton L, Morley GM, Chopra R, and Khwaja A. The Src-selective kinase inhibitor PP1 also inhibits Kit and Bcr-Abl tyrosine kinases. J Biol Chem 278: 4847-4853, 2003.
Yamaoka K, Saharinen P, Pesu M, Holt VE III, Silvennoinen O, and O'Shea JJ. The Janus kinases (Jaks). Genome Biol 5: 253, 2004.
Yancopoulos GD, Davis S, Gale NW, Rudge JS, Wiegand SJ, and Holash J. Vascular-specific growth factors and blood vessel formation. Nature 407: 242-248, 2000.
Yang T, Huang Y, Heasley LE, Berl T, Schnermann JB, and Briggs JP. MAPK mediation of hypertonicity-stimulated cyclooxygenase-2 expression in renal medullary collecting duct cells. J Biol Chem 275: 23281-23286, 2000.
Yarden Y and Sliwkowski MX. Untangling the ErbB signalling network. Nat Rev Mol Cell Biol 2: 127-137, 2001.
Zhao J, Hogan EM, Bevensee MO, and Boron WF. Out-of-equilibrium CO 2 /HCO3- solutions and their use in characterizing a new K/HCO 3 cotransporter. Nature 374: 636-639, 1995.
Zhao J, Zhou Y, and Boron WF. Effect of isolated removal of either basolateral HCO3- or basolateral CO 2 on HCO3- reabsorption by rabbit S2 proximal tubule. Am J Physiol Renal Physiol 285: F359-F369, 2003.
Zhou Y, Bouyer P, and Boron WF. Effects of angiotensin II on the CO 2 dependence of HCO3- reabsorption by the rabbit S2 renal proximal tubule. Am J Physiol Renal Physiol 290: F666-F673, 2006.
Zhou Y, Zhao J, Bouyer P, and Boron WF. Evidence from renal proximal tubules that HCO3- and solute reabsorption are acutely regulated not by pH but by basolateral HCO3- and CO 2. Proc Natl Acad Sci USA 102: 3875-3880, 2005.
作者单位：Department of Cellular and Molecular Physiology, Yale University School of Medicine, New Haven, Connecticut