Am J Physiol Cell Physiol Fuel your research with LabChart
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Am J Physiol Cell Physiol 291: C282-C289, 2006. First published March 29, 2006; doi:10.1152/ajpcell.00564.2005
0363-6143/06 $8.00
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
291/2/C282    most recent
00564.2005v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (3)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Ibla, J. C.
Right arrow Articles by Colgan, S. P.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Ibla, J. C.
Right arrow Articles by Colgan, S. P.

MEMBRANE TRANSPORTERS, ION CHANNELS, AND PUMPS

Transcriptional repression of Na-K-2Cl cotransporter NKCC1 by hypoxia-inducible factor-1

Juan C. Ibla,1,2,* Joseph Khoury,1,2,* Tianqing Kong,2 Andreas Robinson,2 and Sean P. Colgan2

1Department of Anesthesia Perioperative and Pain Medicine, Children’s Hospital, and 2Center for Experimental Therapeutics and Reperfusion Injury, Department of Anesthesia Perioperative and Pain Medicine, Brigham and Women’s Hospital, Harvard Medical School, Boston, Massachusetts

Submitted 4 November 2005 ; accepted in final form 19 March 2006


   ABSTRACT
TOP
 ABSTRACT
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 GRANTS
 REFERENCES
 
Tissue edema is commonly associated with hypoxia. Generally, such episodes of fluid accumulation are self-limiting. At present, little is known about mechanisms to compensate excessive fluid transport. Here we describe an adaptive mechanism to dampen fluid loss during hypoxia. Initial studies confirmed previous observations of attenuated electrogenic Cl– secretion after epithelial hypoxia. A screen of known ion transporters in Cl–-secreting epithelia revealed selective downregulation of Na-K-2Cl cotransporter NKCC1 mRNA, protein, and function. Subsequent studies identified transcriptional repression of NKCC1 mediated by hypoxia-inducible factor (HIF). Chromatin immunoprecipitation analysis identified a functional HIF binding site oriented on the antisense strand of genomic DNA downstream of the transcription start site corresponding to the NKCC1 5'-untranslated region. Additional in vivo studies using conditional Hif1a-null mice revealed that the loss of HIF-1{alpha} in Cl–-secreting epithelia results in a loss of NKCC1 repression. These studies describe a novel regulatory pathway for NKCC1 transcriptional repression by hypoxia. These results suggest that HIF-dependent repression of epithelial NKCC1 may provide a compensatory mechanism to prevent excessive fluid loss during hypoxia.

transcription; ion transport; chloride secretion; epithelia

MUCOSAL SURFACES, such as the lung and colon, are lined by a monolayer of epithelia that provides tissue barrier and vectorial ion transport function. Recent reports indicate that mucosal epithelial cells are end point targets for a number of diseases, including ischemia-reperfusion injury, mucosal inflammation, and sepsis (14). It is therefore important to understand mechanisms of epithelial function and adaptation to such adverse conditions. From this standpoint, it is now appreciated that a common feature in a variety of disease processes is diminished oxygen delivery and/or oxygen availability (hypoxia) (31). For example, it was recently shown that epithelial cells lining the colon lack normal oxygenation during modeled colitis and that transcriptional pathways mediated by hypoxia-inducible factor (HIF)-1 appear to protect the epithelium during ongoing inflammatory processes (16). Such observations place the epithelium in a central locale to coordinate adaptive responses to hypoxia (15).

As stated above, a primary function of the epithelium is vectorial ion transport, the transport event responsible for mucosal hydration (1, 13). This process occurs through a coordinated series of membrane transporters. In Cl–-secreting epithelium, the rate-limiting (1) step is entry of Cl– via the Na-K-2Cl cotransporters, a family of proteins that mediate the electroneutral transport of Na+, K+ and Cl– across cellular membranes (13). These cotransporters are distributed over a wide variety of tissues, and, with the exception of renal epithelia, are localized along the basolateral membrane (13). Because of their particular ability to regulate both anionic and cationic fluxes, the Na-K-2Cl cotransporters serve a number of vital physiological functions. In the majority of cell types Na-K-2Cl cotransporters are involved in the regulation of cell volume and are functionally activated by cell shrinkage (13). In coordination with other solute transport pathways, Na-K-2Cl cotransporters play a key role in salt transport by secretory epithelia. Specifically, the apical/basolateral distribution determines whether Na-K-2Cl cotransporters participate in active fluid absorption or secretion depending on the particular organ and their localization (13).

We (37) and others (10, 22, 26) have previously studied the influence of hypoxia on epithelial electrogenic Cl– secretion and water transport. These studies revealed that hypoxia elicits a direct, noncytotoxic attenuation of stimulated Cl– transport. Here we identify molecular mechanisms of attenuated Cl– secretion by hypoxia. Specifically, these studies reveal the transcriptional repression of the major Na-K-2Cl cotransporter (NKCC1). Extensions of these studies identified a functional binding site for HIF-1{alpha} in the genomic fraction corresponding to the 5'-untranslated region (5'-UTR) of NKCC1. Such studies provide new insight into adaptive responses of epithelial cells and identify target molecules for their potential therapeutic use in mucosal disease.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 GRANTS
 REFERENCES
 
Cell culture. T84 intestinal epithelial cells were grown and maintained as confluent monolayers in 1:1 Dulbecco’s modified Eagle’s medium (DMEM)-Ham’s F-12 medium with 10% FBS at 37°C in 5% CO2 in room air. For electrophysiological measurements described below, T84 cells were plated on collagen-coated permeable supports as previously described in detail (11). Monolayers were grown on 0.33-cm2 or 5.0-cm2 ring-supported polycarbonate filters (Costar, Cambridge MA) unless otherwise noted, and they were used 6–12 days after plating as described previously (21). Epithelial cultures were exposed to hypoxia as described previously (5). Growth media were replaced with fresh media equilibrated with hypoxic gas mixture, and cells were placed in a hypoxic chamber (Coy Laboratory Products, Ann Arbor, MI). Oxygen concentrations were as indicated (normoxia equal to PO2 147 Torr and hypoxia equal to PO2 20 Torr), with the balance made up of N2, CO2 (constant PCO2 35 Torr), and water vapor from the humidified chamber.

Electrophysiological measurements. To measure agonist-stimulated short-circuit current (Isc), transepithelial potential, and resistance, we use a commercially available voltage clamp (Iowa Dual Voltage Clamps, Bioengineering, University of Iowa) interfaced with an equilibrated pair of calomel electrodes and a pair of Ag-AgCl electrodes, as described in detail elsewhere (11). Cl– secretory responses are expressed as a change in Isc [peak Isc minus baseline Isc; designated (Isc)] necessary to maintain zero potential difference across the monolayer.

Transcriptional analyses. The transcriptional profile of epithelial cells subjected to hypoxia was assessed in RNA derived from control epithelia (T84 cells) or epithelia exposed to hypoxia (PO2 20 Torr and PCO2 35 Torr for indicated periods of time) with quantitative genechip expression arrays (Affymetrix, Santa Clara, CA) (20). mRNA expression was quantified by real-time PCR (iCycler; Bio-Rad Laboratories, Hercules, CA), as described previously (27). The primer set containing SYBR green spanning exons 4 and 5 of the human NKCC1 was sense 5'-TTTGTAAGAGGAGGAGGAGCA-3' and antisense 5'-TTCTGCAAATCCAACCACAT-3' and primers for human beta-actin sense 5'-TGACCCAGATCATGTTTGAGA-3' and antisense 5'-AGTCCATCACGATGCCAGT-3'.

Protein fractionation, Western blotting, and cell surface biotinylation. Cells were pelleted in PBS, and the pellet was used for protein extraction with a cell lysis buffer (Cell Signaling Technologies, Beverly, MA) containing (mM) 20 Tris·HCl (pH 7.5), 150 NaCl, 1 Na2EDTA, 1 EGTA, 2.5 sodium pyrophosphate, 1 beta-glycerophosphate, 1 Na3VO4, and 1 PMSF, with 1% (vol/vol) Triton X-100 and 1 µg/ml leupeptin. For Western blot analyses, samples were subjected to Tricine-SDS gel electrophoresis (7.5% or 10%) and transferred to nitrocellulose membranes (Bio-Rad Laboratories). Blots were sequentially probed with anti-human NKCC1 antibody (Santa Cruz Biotechnology, Santa Cruz, CA) and horseradish peroxidase-conjugated secondary antibodies. Similarly, blots were probed with anti-human beta-actin antibody (Santa Cruz Biotechnology) as a control for loading. Chemiluminescent substrate was detected on X-OMAT film (Kodak). T84 monolayers were exposed to normoxia or 24–72 h of hypoxia, lysed in low-stringency lysis buffer, and immunoprecipitated overnight with NKCC1 antibody. Beads were washed three times in low-stringency lysis buffer, biotinylated (1 mM biotin, 25 min, 4°C), washed again, and resuspended in 150 mM NH4Cl (in lysis buffer) for 30 min at 4°C. Washed beads were resuspended in sample buffer, boiled (5 min), and separated on 10% SDS-PAGE. Proteins were transferred to nitrocellulose membrane and probed for biotinylated proteins with avidin-peroxidase.

5'-Rapid amplification of cDNA ends, cloning of 5'-UTR, mutation of HIF-1{alpha} transcription binding site, and mutation of transcription start site. RNA was isolated from the T84 human epithelial cell line with TRIzol (Invitrogen, Carlsbad, CA) for RNA ligase-mediated rapid amplification of cDNA ends (RLM-RACE). 5'-RLM-RACE cDNA was prepared with a FirstChoice kit (Ambion, Austin, TX) per the manufacturer’s protocol. 5'-RACE experiments were carried out with FirstChoice 5' forward primers and NKCC1-specific reverse primers (outer: 5'-CGTTCTCGGAAACCAGGTCCA-3'; inner: 5'-ACAGCCGCAAGACCAGGTGAGAGAGTCTT-3') with SuperTaq recombinant DNA polymerase (Ambion). The resulting band was cloned into a TA cloning vector (pCR 2.1, Invitrogen) and sequenced with T7 primer (Biopolymers Facility, Harvard Medical School). Cloning of the 5'-flanking region and adjacent 5'-UTR was performed with genomic DNA extracted from HeLa cells by a commercially available method (Qiagen, Valencia, CA). A PCR fragment was generated extending 510 bp from position –324 upstream of the transcription start site (TSS) (19) to +186 with primers: sense 5'-GATCCTCCCGGCGACATCACTG-3' and antisense 5'-TGCCCGACCCCCGCCGGAACGT-3'. The gel-purified band was then cloned into a pGL3 luciferase reporter vector (Promega, Madison, WI). Transcription factor analysis of this sequence was performed with online software (MatInspector, Genomatix). The identified HIF-1{alpha} transcription factor binding site was mutated with a commercially available kit (QuikChange site-directed mutagenesis kit, Stratagene). The TSS of NKCC1 was identified as described by the 5'-RACE and mutated similar to the HIF-1{alpha} transcription binding site mutation.

Plasmid transfection and luciferase reporter assays. HeLa cells were passaged into 24-well plates and allowed to attach for 20 h. Cells were then subjected to hypoxia (PO2, 20 Torr), and transient transfection of HeLa cells was carried out at 60% confluence in DMEM (Invitrogen) at a concentration of 0.5 µg of plasmid per well of a 24-well plate with preequilibrated hypoxic medium. Plasmids were transfected with Polyfect (Qiagen) at a concentration of 4 µl of reagent per 1 µg of DNA. Six hours after transfection, plates were either returned to normoxia conditions or maintained in hypoxia for an additional 18–48 h. At the desired time point after hypoxia, transfected cells were lysed in 100 µl of passive lysis buffer (Promega), and 20 µl was used in a luciferase assay with Luciferase Assay Reagent II (Promega). Luminescence readings were obtained with a luminometer (Turner Designs TD-20/20, Sunnyvale, CA). Relative luciferase units were recorded, and values were corrected to the total protein concentration (Lowry method) and verified by cotransfection with Renilla for each sample.

[3H]bumetanide binding assay. [3H]bumetanide binding to basolateral cotransporter sites was performed as described elsewhere (7) with T84 cells grown on 5-cm2 permeable supports subjected to normoxia or hypoxia as described above. In all experiments, paired monolayers exposed to excess unlabeled bumetanide were assayed to determine specific [3H]bumetanide binding.

Chromatin immunoprecipitation assay. Chromatin immunoprecipitation (ChIP) was performed as described previously (17) with minor modifications. Briefly, HeLa cells were grown to confluence and fixed with 1% formalin, and chromatin was isolated and sheared. Sheared chromatin was incubated with an anti-HIF-1 antibody (Transduction Laboratories, Lexington, KY) and immune complexes were precipitated with protein G Sepharose beads. Cross-linking was then reversed, and DNA was purified with a PCR purification kit (Qiagen). HIF binding to the NKCC1 5'-UTR DNA was quantified by standard PCR using primers (sense 5'-ACACTCGCGCGCTCGCTC-3', antisense 5'-GGCCGCGGCTCCATAGCTGC-3') designed to amplify a 217-bp region spanning the HIF-1 binding site located from position 1 bp to +202 bp with respect to the TSS. Chromatin incubated with a nonspecific antibody, beads without antibody, or water alone was used to control for nonspecific binding of DNA.

In vivo assessment of NKCC1. Colonic mucosal scrapings (enriched in epithelial cells) were obtained from 6- to 8-wk-old conditional Hif1{alpha}-mutant mice or littermate controls, as described previously (16). Scrapings were homogenized in RNAlater (Qiagen) with a 22-gauge syringe (Becton Dickinson) and a Qiashredder column (Qiagen). RNA extraction including DNAse digestion was performed with an RNeasy kit (Qiagen). Reverse transcription was done with an iScript cDNA synthesis kit (Bio-Rad Laboratories). Amplification was performed on an i-Cycler IQ real-time PCR detection system (Bio-Rad Laboratories) with the gene-specific primers for NKCC1 described above and mouse beta-actin: sense 5'-CTAGGCACCAGGGTGTGAT-3' and antisense 5'-TGCCAGATCTTCTCCATGTC-3'. Cycle parameters were 3 min at 95°C and then 40 cycles with 45 s at 94°C, 45 s at 57°C, 45 s at 72°C, and 5 s at 81.5°C (fluorescence data collection point), followed by repetitive melting cycles to establish product specificity. Comparison of gene expression in a semiquantitative manner was performed based on the mathematical model of Pfaffl (27). All procedures involving animals were performed according to National Institute of Health guidelines for use of live animals and were approved by the Institutional Animal Care and Use Committee at Brigham and Women’s Hospital.

Data analysis. Cl– secretion and luciferase reporter data were compared by two-factor ANOVA or Student’s t-test, where appropriate. Values are expressed as means and SD from at least three separate experiments.


   RESULTS
TOP
 ABSTRACT
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 GRANTS
 REFERENCES
 
Hypoxia attenuates electrogenic Cl– secretion. Initially, we utilized a cell culture model using Cl– secretory epithelia (T84) grown as monolayers on permeable supports to examine influences of hypoxia on electrogenic Cl– secretion. As shown in Fig. 1, and as we demonstrated previously (37), subjection of epithelia to normobaric hypoxia (PO2 20 Torr) resulted in attenuated electrogenic Cl– secretion. Indeed, both receptor-mediated (adenosine, 30 µM) and non-receptor-mediated (forskolin, 3 µM) stimulation of Cl– secretion, measured as a function of short-circuit current (Isc) by a voltage clamp, were attenuated in a time-dependent manner (P < 0.025 by ANOVA for each), with >80% decrease in Isc observed at 36–48 h of hypoxia. Such attenuated decreases in Cl– secretion are also reflected as a decrease in transepithelial water transport (37). Of note, and as previously shown, such decreases in electrogenic Cl– secretion occur in the absence of a significant fall in transepithelial resistance during normoxic conditions (P < 0.025), a sensitive measure of epithelial toxicity (5). Such data confirm previous work addressing epithelial ion transport properties and provide the basis to define these principles at the molecular level.


Figure 1
View larger version (8K):
[in this window]
[in a new window]
  		Fig. 1. Influence of hypoxia on receptor- and nonreceptor-stimulated Cl– secretion. T84 intestinal epithelia were subjected to normoxia (PO2 147 Torr) or hypoxia (PO2 20 Torr) for indicated periods of time and examined for forskolin (A, 3 µM)- or adenosine (B, 30 µM)-stimulated Cl– secretion (measured as a short circuit current, Isc). Data are pooled from 8–12 individual monolayers from 3 experiments in each condition, and results are expressed as mean ± SE Isc (*P < 0.01 compared with normoxia). Time points in normoxia did not change significantly (P = 0.025).

 
Influence of hypoxia on epithelial membrane ion channels. As an initial attempt to gain insight into mechanisms of attenuated Cl– secretion, we transcriptionally profiled epithelial RNA derived from normoxic and hypoxic (6 and 18 h) epithelia. These studies were insightful insomuch as they identified a selective loss of NKCC1 by hypoxia. Indeed, in Cl–-secreting epithelium, well-established membrane pathways exist for Cl– uptake through NKCC1 and Cl– efflux through Cl– channels, including the cystic fibrosis transmembrane regulator.

As shown in Fig. 2A, microarray analysis revealed that hypoxia repressed NKCC1 mRNA in a time-dependent fashion (P < 0.025 by ANOVA), RNA transcript downregulation at 6 h (2.1 ± 0.5 fold) and 18 h (8.7 ± 1.2 fold) of hypoxia compared with normoxic controls. This pattern of expression was confirmed with real-time PCR (Fig. 2B), showing a 2.1 ± 0.1- and 3.2 ± 0.2-fold repression at 6 and 18 h of hypoxia, respectively. Notably, mRNA transcript levels of other membrane channels important in coordinated Cl– secretion were not significantly changed by hypoxia (data not shown). These results implicate specific downregulation of NKCC1 by hypoxia.


Figure 2
View larger version (10K):
[in this window]
[in a new window]
  		Fig. 2. Repression of NKCC1 mRNA by hypoxia. A: confluent T84 cells were exposed to normoxia (Nx) or hypoxia (6- or 18-h exposure to PO2 20 Torr), and the relative expression of NKCC1 was quantified from total RNA by microarray analysis. Data are expressed as mean ± SD relative fluorescent units. B: real-time PCR analysis was used to confirm microarray data indicating hypoxia-mediated repression of NKCC1. Data were calculated relative to internal control (beta-actin) and are expressed as mean ± SD fold change. Results are derived from 3 experiments in each condition (*P < 0.01 compared with normoxia).

 
NKCC1 protein expression during hypoxia. We verified the above findings of hypoxia-mediated NKCC1 repression at the protein level. As shown in Fig. 3A, we examined surface NKCC1 protein with [3H]bumetanide, a method used previously to determine specific cell surface NKCC1 (7). These studies revealed a hypoxia time-dependent loss of surface NKCC1 (as reflected by [3H]bumetanide binding) through 48 h of hypoxia (P < 0.01 by ANOVA). Decreased surface NKCC1 was evident as early as 24 h of hypoxia (53 ± 6.2% decrease; P < 0.01) and was maximal at 48 h (e.g., no further decrease at 72 h; data not shown). In addition, we examined the relative contribution of the bumetanide-sensitive component of Cl– secretion after forskolin (100 nM) stimulation. Our analysis indicated that the bumetanide-sensitive component, measured as an Isc, in T84 cells used for these experiments corresponds to ~84.2 ± 4.2% of total current generated by agonist stimulation. Importantly, bumetanide was less effective in cells subjected to hypoxia (51.0 ± 4.1% inhibition; P < 0.025) than those subjected to normoxia, thereby supporting our hypothesis that hypoxia results in a functional repression of NKCC1.


Figure 3
View larger version (38K):
[in this window]
[in a new window]
  		Fig. 3. Influence of hypoxia on NKCC1 protein. A: confluent T84 cells were exposed to normoxia or hypoxia (48-h exposure to po2 20 Torr). Specific [3H]bumetanide binding was used to examine surface expression of NKCC1. Data are expressed as mean ± SE bound [3H]bumetanide from 3 experiments in each condition (*P < 0.01 compared with normoxia). B: confluent T84 monolayers were subjected to indicated periods of hypoxia in the presence or absence of the transcription inhibitor 5,6-dichlorobenzimidazole riboside (DRB). Cells were cooled to 4°C, scraped, and solubilized as described in MATERIALS AND METHODS. Lysates were separated by SDS-PAGE under nonreducing conditions, and blots were probed with polyclonal rabbit anti-NKCC1, labeled with peroxidase-conjugated secondary Ab, and developed by enhanced chemiluminescence (ECL). To control for loading, blots were stripped and reprobed with anti-beta-actin. One of three representative experiments is shown. C: confluent T84 monolayers were exposed to indicated times of normoxia and hypoxia, surface biotinylated, and immunoprecipitated with anti-NKCC1 antibody. Blots were probed with avidin-peroxidase and developed by ECL. Representative experiment of 3 is shown.

 
Further examination of total NKCC1 by Western blot (representative blot shown in Fig. 3B) revealed a loss of NKCC1 with increasing time of hypoxia. By Western blot, hypoxia-mediated repression of NKCC1 protein was evident at time points beyond 24 h, with maximal decreases at 48 and 72 h of hypoxia exposure compared with normoxia. Additional studies were performed to determine whether NKCC1 repression by hypoxia requires transcriptional activity. To do this, T84 cells were subjected to normoxia or hypoxia (48 h) in the presence or absence of the transcriptional inhibitor 5,6-dichlorobenzimidazole riboside (DRB; 3 µM final concentration) and assessed for changes in NKCC1 expression by Western blot. As shown in Fig. 3B, preincubation of T84 cells with DRB inhibited the hypoxia-dependent loss of NKCC1. Consequently, hypoxia-mediated repression of NKCC1 requires new mRNA synthesis.

Additionally, basolateral biotinylation and immunoprecipitation of NKCC1 during normoxia and hypoxia was performed in confluent, high-resistance (>1,000 ohm·cm2) T84 cells. As shown in Fig. 3C, it is readily evident that surface protein levels decrease with increasing time in hypoxia. Comparison with normoxic controls at the same time points provides strong evidence for NKCC1 dependence on hypoxia, consistent with our results obtained for mRNA and total protein. Although the surface expression of NKCC1 decreases between 12 and 24 h, the total protein expression decreases only after 24 h. In subsets of experiments, we found that with fractionation of membrane, cytoplasm, and nuclear lysates, no differences exist in the percentage of total retrieval of NKCC1 between normoxia and hypoxia (data not shown). From these findings, we concluded that the turnover of surface protein is likely more rapid than intracellular NKCC1.

Transcriptional regulation of NKCC1. In view of the likelihood of a transcription-mediated repression of NKCC1 during hypoxia, attention was directed at the 5'-UTR for potential hypoxia-regulated transcription factor sequences. Our 5'-RACE results confirmed the findings of Liedtke et al. (19) identifying the TSS 190 bp upstream of the start codon. We then cloned a genomic fraction extending from –324 bp to +186 bp (sequence shown in Fig. 4A) into pGL3 luciferase reporter vector and examined transcriptional activity in hypoxia after transient transfection in HeLa cells. As shown in Fig. 4B, analysis of luciferase reporter activity revealed that hypoxia represses NKCC1 activity of the genomic DNA downstream of the TSS corresponding to 5'-UTR by 71 ± 12% (P < 0.025) relative to cells incubated in normoxia, suggesting the possibility that hypoxia acts directly on this genomic fraction of NKCC1. To further characterize the role of the cloned 5'-flanking region and adjacent 5'-UTR for NKCC1 transcriptional control, we created a 5-bp mutation of the TSS (mutation from ACACT to TGAGA, construct termed NKCC1{Delta}TSS), using the wild-type NKCC1 construct. Adjusted relative luciferase units relative to the expression of the NKCC1{Delta}TSS during normoxia showed a significant loss of activity (89 ± 3.7%; P < 0.01) compared with wild-type NKCC1 during normoxia (Fig. 4C). These results corroborate the importance of this region of NKCC1 and the neighboring HIF-1{alpha} binding site for the transcriptional control of NKCC1 during hypoxia.


Figure 4
View larger version (31K):
[in this window]
[in a new window]
  		Fig. 4. Influence of hypoxia on activity of the NKCC1 5' genomic fraction. A: partial sequence and orientation of the hypoxia response element within the NKCC1 5' genomic fraction corresponding to the 5'-untranslated region (5'-UTR), relative to the transcription start site (TSS, italic text) and start codon (bold ATG). HIF, hypoxia-inducible factor (underlined text). B: confluent HeLa cell monolayers were transiently transfected with luciferase plasmids expressing sequence corresponding to NKCC1 5' genomic fraction (bp –324 to +186 relative to the transcription initiation start site). Cells were then exposed to hypoxia or normoxia for 24 h and assessed for luciferase activity. All transfections were normalized to protein levels and confirmed by cotransfection with Renilla plasmid. Data are means ± SE from 3 separate experiments. C: the TSS located at position 190 bp upstream of the start codon (ATG) was eliminated in the wild-type (WT) luciferase construct by site-directed mutagenesis. The resulting construct (NKCC1{Delta}TSS) was cotransfected along with Renilla luciferase plasmid into HeLa cells. Data shown are representative experiments from n = 3 and are normalized for background vector (empty PGL3), total protein and verified by cotransfection with Renilla. Data are means ± SD (*P < 0.025 compared with wild-type NKCC1 construct).

 
Analysis of our cloned region of NKCC1 revealed the existence of a potential binding site for HIF-1{alpha} in reverse orientation (binding site core sequence 5'-GACGT-3') at position –38 relative to TSS (Fig. 4A). Others have reported that the binding site for HIF can exist within the 5'-UTR (4). On the basis of these findings, and known epithelial expression of HIF-1{alpha} (14), we determined whether the genomic DNA downstream of the TSS corresponding to the 5'-UTR of NKCC1 would bind HIF-1{alpha}. For these purposes, we used ChIP analysis to examine binding of HIF-1{alpha} to the NKCC1 5' genomic fraction spanning the putative HIF-1{alpha} binding site in intact cells. As shown in Fig. 5A, this analysis revealed a prominent band of 217 bp in nuclei derived from hypoxic, but not normoxic, cells. No bands were evident in the beads-only control, and preimmunoprecipitation samples revealed equivalent DNA input. Such results indicate that hypoxia induces HIF-1{alpha} binding to the proximal 217 bp relative to the ATG of NKCC1.


Figure 5
View larger version (19K):
[in this window]
[in a new window]
  		Fig. 5. Binding and functional role of HIF-1 on NKCC1 expression. A: chromatin immunoprecipitation was used to examine HIF-1{alpha} binding to a genomic fraction corresponding to the NKCC1 5'-UTR in normoxic and hypoxic HeLa cells. Reaction controls included PCR performed with whole HeLa cell genomic DNA (input), antibody controls, and samples precipitated by protein G Sepharose beads alone (B) or beads wash (W). B: the HIF-1 binding site located at position +147 to +151 bp with respect to the TSS was eliminated in the wild-type luciferase construct by site-directed mutagenesis. The resulting construct (NKCC1{Delta}HIF) was cotransfected along with Renilla luciferase vector into HeLa cells, followed by 24 h of hypoxia. Data shown are a representative experiment from n = 3 and are normalized for background vector (empty PGL3, not shown), total protein and verified by cotransfection with Renilla. Data are means ± SD. *P < 0.025 C: real-time PCR analysis of murine epithelial NKCC1 in conditional Hif1a-mutant (–/–) and littermate control (+/+) animals. Data were calculated relative to internal control (beta-actin) and are expressed as mean ± SD relative NKCC1 mRNA (*P < 0.025), where transcript levels in control animals were normalized to 1. Results are derived from 3 animals in each genotype.

 
To further define the role of HIF-1{alpha} in NKCC1 repression during by hypoxia, a 5-bp mutation of the central HIF-1{alpha} binding site (mutation from GACGT to TGAAG, construct termed NKCC1{Delta}HIF) was created on the wild-type sequence and confirmed by sequencing. HeLa cells were then transfected with wild-type and NKCC1{Delta}HIF and exposed to hypoxia. As can be seen in Fig. 5B, although hypoxia repressed activity of the wild-type construct by 76 ± 14% (P < 0.01), mutation of the central HIF binding site resulted in a complete loss of hypoxia-mediated repression (7 ± 3% loss of luciferase activity; P = not significant). No consistent differences between the level of expression of wild-type NKCC1 and {Delta}HIF constructs were observed at normoxic conditions (P = 0.11). These results demonstrate that disruption of the HIF-1{alpha} binding site obliterates the NKCC1 repressive response to hypoxia. Together, these findings identify HIF-mediated repression of NKCC1 in hypoxia.

Role of HIF in NKCC1 expression in vivo. As proof of principle for these findings of HIF-1-mediated repression of NKCC1 expression, we examined NKCC1 mRNA levels in intestinal epithelia derived from conditional Hif1a-knockout animals (16). Intestinal epithelia from these mice lack detectable Hif1a expression in >70% of cells (16). Consistent with our hypothesis that HIF-1 transcriptionally represses NKCC1, real-time PCR analysis revealed a 9.7 ± 1.3-fold increase in NKCC1 expression in intestinal epithelia derived from animals conditionally lacking Hif1a (Fig. 5C) relative to their littermate controls (P < 0.01). Such findings support our in vitro findings and indicate the likelihood that HIF-1 also regulates murine NKCC1.


   DISCUSSION
TOP
 ABSTRACT
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
GRANTS
 REFERENCES
 
Epithelial cells of various mucosal organs are targets for hypoxia-mediated damage in a number of diseases (15). In particular, epithelial cells represent a functional mucosal surface with high turnover and high metabolic activity overlying a rich vascular bed and a large number of immune-related cell types. At these anatomic sites, even small perturbations of blood flow can result in rapid metabolic changes characteristic of ischemia and resultant hypoxia. Here we address molecular aspects of epithelial adaptation to hypoxia, with particular emphasis on hypoxic regulation of the ion transporter NKCC1.

Initial studies identified functional decreases in agonist-stimulated Cl– secretion in association with selective repression of NKCC1 by transcriptional profiling. In Cl–-secreting epithelia, such as those used here, expression levels of NKCC family members are rate limiting for Cl– uptake and therefore serve as a rate-limiting step for Cl– transport through various membrane Cl– channels (13). These findings of diminished ion transport in hypoxia confirm previous work by us (37) and others (10, 22, 26) in various epithelial cell types. Although our present results clearly demonstrate transcription-dependent downregulation of NKCC1, it is also likely that nontranscriptional mechanisms could also contribute to functional regulation of NKCC1 in Cl–-secreting epithelia. For example, activation of protein kinase C (PKC)-{epsilon} results in rapid internalization and degradation of surface NKCC1 (3, 12). Because hypoxia/ischemia has been shown to activate PKC-{epsilon} both in vitro and in vivo (9, 23, 33), it is possible that this mechanism contributes to the initial phase of attenuated Cl– secretion. Moreover, NKCC1 function can be directly regulated by threonine phosphorylation at the amino terminus (13). We do not presently know whether NKCC1 phosphorylation patterns differ in response to hypoxia.

Presently, little is known about the transcriptional regulation of NKCC1 family members. A number of studies have addressed the influence of inflammatory and signaling mediators on ion transporters. For example, inflammatory cytokines such as interferon-{gamma} and interleukin-4 have been shown to attenuate Cl– secretion, and such changes are associated with decreased expression of NKCC1 (2, 7, 8, 28, 35). These findings confirm work in the past indicating that select cytokines elicit an epithelial "phenotypic switch" whereby cytokines initiate a loss of epithelial characteristics and a gain of immune accessory function (6). These studies, however, have provided very little insight into mechanisms of NKCC1 regulation. From this perspective, we pursued our microarray findings to address transcriptional repression of NKCC1 by hypoxia. Important initial insight was gained by using the transcription inhibitor DRB, and on the basis of these findings we cloned and functionally characterized a genomic fraction corresponding to the NKCC1 5'-UTR. Studies using these luciferase constructs confirmed initial findings of NKCC1 repression by hypoxia and implicated transcription factor control of NKCC1 repression by hypoxia.

The use of the transcriptional inhibitor DRB provided two lines of evidence. First, the use of DRB demonstrated that the repression of NKCC1 by hypoxia is transcriptionally dependent and provided strong evidence to search for transcription factor that might regulate this function. Second, it revealed that such transcriptional dependence manifested at the protein level was not likely a result of mRNA stabilization. Given the temporal and robust hypoxia response observed in the repression of NKCC1, a candidate regulator was HIF-1, a heterodimeric transcription factor whose activation is dependent on stabilization of an O2-dependent degradation domain (29). It is recently appreciated that HIF can function as both a transcriptional activator and a transcriptional repressor (24). Therefore, a search of the cloned NKCC1 gene 5'-UTR identified one potential HIF-1 binding DNA consensus motif located in the NKCC1 5'-UTR. However, the existence of a HIF-1{alpha} binding consensus is not evidence for HIF-1{alpha}-mediated response; instead, the hypoxia response element (HRE) is defined as a cis-acting transcriptional regulatory sequence located within 5'-flanking, 3'-flanking, or intervening sequences of target genes (30). Therefore, two strategies were used to define the role of HIF-1 in repression of NKCC1. First, mutational analysis of NKCC1 5' genomic fraction constructs defined a functional HRE located in the distal 5' region of the NKCC1 promoter (from +147 bp to +151 bp relative to the major TSS). Second, ChIP analysis confirmed HIF-1{alpha} binding to the NKCC1 5' genomic fraction. At present, we do not know whether this binding is direct or indirect, and we do not know the nature of HIF-mediated transcription. Although previous work with peroxisome proliferator-activated receptor-{alpha} gene implicated potential repressor activity with HIF-1{alpha} binding sites oriented on the antisense strand (25), implying some degree of transcriptional directionality, there is not direct evidence for such a mechanism. Koshiji et al. (18) recently reported that Sp1 functions as a molecular switch to recruit HIF-1{alpha} and repress expression of mutS homolog (MSH-2–6). Our analysis of the NKCC1 5'-UTR and 5'-flank sequence have revealed at least one Sp1 binding site (positions –87 to –78; not verified experimentally) in the general vicinity of the HRE, and therefore it is possible that Sp1 could contribute to the transcriptional repression of NKCC1 within this region. Moreover, although a recent study comparing transcriptional responses between hypoxia and constitutively active HIF-1{alpha} identified a large cohort of transcriptionally repressed genes (24), no unique patterns of HRE expression were noted. Thus more work will be necessary to define the precise nature of HIF-mediated transcriptional repression.

A number of recent studies have implicated epithelial cells in adaptive responses to hypoxia in vitro and in vivo. As an example, we (15) and others have described an adaptive hypoxia response involving a number of barrier-related genes specifically regulated by HIF. This hypothesis was tested in a conditional HIF1a-mutant mouse line and revealed a protective role for murine HIF-1a in a mucosal inflammation model (16). Using this same mouse line, we confirmed our hypothesis that HIF transcriptionally represses NKCC1. Indeed, our findings revealed that colonic epithelial cells lacking Hif1a express increased NKCC1 transcript. Of note, these were unmanipulated mice and these studies reveal, as previously suggested (16, 36), that HIF may regulate homeostatic gene expression (i.e., "physiological hypoxia") in tissues with normally low PO2 values (e.g., colon) (32). Sequence analysis of NKCC1 in other species revealed similar HIF-1{alpha} binding sites consistent with HREs in monkey, rat, and mouse, suggesting that this sequence is evolutionarily conserved. More primitive species (Caenorhabditis elegans, Drosophila) were less convincing, but so too is the expression of NKCC1 in these species. In diseases of the intestine involving a degree of ischemia and/or hypoxia (e.g., ischemic colitis), the most common early and persistent symptom is diarrhea (34). Because electrogenic Cl– secretion, in excess, can result in secretory diarrhea, it is possible that the HIF-dependent repression of NKCC1 represents an adaptive response in the intestine to diminish fluid loss resulting from overstimulation of water transport. Together, the present findings represent a novel, HIF-dependent adaptive response likely to be important in hypoxia compensation of mucosal tissues. Extensions of these studies may provide therapeutic potential to prevent excessive fluid loss during hypoxia.


   GRANTS
TOP
 ABSTRACT
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 GRANTS
REFERENCES
 
This work was supported by a grant from the Foundation for Anesthesia Education and Research to J. C. Ibla, National Institutes of Health (NIH) Grants HL-60569 and DK-50189 (to S. P. Colgan), and Project 2 of NIH P50-DE-01691.


   FOOTNOTES
 

Address for reprint requests and other correspondence: S. P. Colgan, Brigham and Women’s Hospital and Harvard Medical School, Thorn Bldg. 704, 20 Shattuck St., Boston, MA 02115 (e-mail: colgan{at}zeus.bwh.harvard.edu)

The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

* J. C. Ibla and J. Khoury contributed equally to this work. Back


   REFERENCES
TOP
 ABSTRACT
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 GRANTS
 REFERENCES
 
1. Barrett KE and Keely SJ. Chloride secretion by the intestinal epithelium: molecular basis and regulatory aspects. Annu Rev Physiol 62: 535–572, 2000.[CrossRef][ISI][Medline]

2. Bertelsen LS, Eckmann L, and Barrett KE. Prolonged interferon-{gamma} exposure decreases ion transport, NKCC1, and Na+-K+-ATPase expression in human intestinal xenografts in vivo. Am J Physiol Gastrointest Liver Physiol 286: G157–G165, 2004.[Abstract/Free Full Text]

3. Calvo Del Castillo I, Fedor-Chaiken M, Song JC, Starlinger V, Yoo J, Matlin KS, and Matthews JB. Dynamic regulation of NKCC1 surface expression by protein kinase C-{epsilon} in chloride secretory epithelia. Am J Physiol Cell Physiol 289: C1332–C1343, 2005.[Abstract/Free Full Text]

4. Camenisch G, Stroka DM, Gassmann M, and Wenger RH. Attenuation of HIF-1 DNA-binding activity limits hypoxia-inducible endothelin-1 expression. Pflügers Arch 443: 240–249, 2001.[CrossRef][ISI][Medline]

5. Colgan SP, Dzus AL, and Parkos CA. Epithelial exposure to hypoxia modulates neutrophil transepithelial migration. J Exp Med 184: 1003–1015, 1996.[Abstract/Free Full Text]

6. Colgan SP, Furuta GT, and Taylor CT. Cytokines and epithelial function. In: Microbial Pathogenesis and the Intestinal Epithelial Cell, edited by Hecht GA. Washington, DC: American Society for Microbiology, 2003, p. 61–78.

7. Colgan SP, Parkos CA, Matthews JB, D’Andrea L, Awtrey CS, Lichtman A, Delp C, and Madara JL. Interferon-{gamma} induces a surface phenotype switch in intestinal epithelia: downregulation of ion transport and upregulation of immune accessory ligands. Am J Physiol Cell Physiol 267: C402–C410, 1994.[Abstract/Free Full Text]

8. Colgan SP, Resnick MB, Parkos CA, Delp-Archer C, Bacarra AE, Weller PF, and Madara JL. Interleukin-4 directly modulates function of a model human intestinal epithelium. J Immunol 153: 2122–2129, 1994.[Abstract]

9. Conte M, Soper B, Chang Q, and Tepperman B. The role of PKC{delta} and PKC{epsilon} in the neonatal rat colon in response to hypoxia challenge. Pediatr Res 55: 27–33, 2004.[CrossRef][ISI][Medline]

10. Dada LA, Chandel NS, Ridge KM, Pedemonte C, Bertorello AM, and Sznajder JI. Hypoxia-induced endocytosis of Na,K-ATPase in alveolar epithelial cells is mediated by mitochondrial reactive oxygen species and PKC-{zeta}. J Clin Invest 111: 1057–1064, 2003.[CrossRef][ISI][Medline]

11. Dharmsathaphorn K and Madara JL. Established intestinal cell lines as model systems for electrolyte transport studies. Methods Enzymol 192: 354–389, 1990.[Medline]

12. Farokhzad OC, Sagar GD, Mun EC, Sicklick JK, Lotz M, Smith JA, Song JC, O’Brien TC, Sharma CP, Kinane TB, Hodin RA, and Matthews JB. Protein kinase C activation downregulates the expression and function of the basolateral Na+/K+/2Cl– cotransporter. J Cell Physiol 181: 489–498, 1999.[CrossRef][ISI][Medline]

13. Gamba G. Molecular physiology and pathophysiology of electroneutral cation-chloride cotransporters. Physiol Rev 85: 423–493, 2005.[Abstract/Free Full Text]

14. Karhausen J, Haase VH, and Colgan SP. Inflammatory hypoxia: role of hypoxia-inducible factor. Cell Cycle 4: 256–258, 2005.[ISI][Medline]

15. Karhausen J, Ibla JC, and Colgan SP. Implications of hypoxia on mucosal barrier function. Cell Mol Biol (Noisy-le-grand) 49: 77–87, 2003.

16. Karhausen JO, Furuta GT, Tomaszewski JE, Johnson RS, Colgan SP, and Haase VH. Epithelial hypoxia-inducible factor-1 is protective in murine experimental colitis. J Clin Invest 114: 1098–1106, 2004.[CrossRef][ISI][Medline]

17. Kong T, Eltzschig HK, Karhausen J, Colgan SP, and Shelley CS. Leukocyte adhesion during hypoxia is mediated by HIF-1-dependent induction of beta2 integrin gene expression. Proc Natl Acad Sci USA 101: 10440–10445, 2004.[Abstract/Free Full Text]

18. Koshiji M, To KK, Hammer S, Kumamoto K, Harris AL, Modrich P, and Huang LE. HIF-1{alpha} induces genetic instability by transcriptionally downregulating MutS{alpha} expression. Mol Cell 17: 793–803, 2005.[CrossRef][ISI][Medline]

19. Liedtke CM, Wang X, and Smallwood ND. Role for protein phosphatase 2A in the regulation of Calu-3 epithelial Na+-K+-2Cl–, type 1 co-transport function. J Biol Chem 280: 25491–25498, 2005.[Abstract/Free Full Text]

20. Lockhart DJ, Dong H, Byrne MC, Follettie MT, Gallo MV, Chee MS, Mittmann M, Wang C, Kobayashi M, Horton H, and Brown EL. Expression monitoring by hybridization to high-density oligonucleotide arrays. Nat Biotechnol 14: 1675–1680, 1996.[CrossRef][ISI][Medline]

21. Madara JL, Parkos CA, Colgan SP, MacLeod RJ, Nash S, Matthews J, Delp C, and Lencer WS. Cl– secretion in a model intestinal epithelium induced by a neutrophil-derived secretagogue. J Clin Invest 89: 1938–1944, 1992.[ISI][Medline]

22. Mairbaurl H, Wodopia R, Eckes S, Schulz S, and Bartsch P. Impairment of cation transport in A549 cells and rat alveolar epithelial cells by hypoxia. Am J Physiol Lung Cell Mol Physiol 273: L797–L806, 1997.[Abstract/Free Full Text]

23. Mammen JM, Song JC, Yoo J, Kim PS, Davis HW, Calvo MI, Worrell RT, Matlin KS, and Matthews JB. Differential subcellular targeting of PKC-{epsilon} in response to pharmacological or ischemic stimuli in intestinal epithelia. Am J Physiol Gastrointest Liver Physiol 288: G135–G142, 2005.[Abstract/Free Full Text]

24. Manalo DJ, Rowan A, Lavoie T, Natarajan L, Kelly BD, Ye SQ, Garcia JG, and Semenza GL. Transcriptional regulation of vascular endothelial cell responses to hypoxia by HIF-1. Blood 105: 659–669, 2005.[Abstract/Free Full Text]

25. Narravula S and Colgan SP. Hypoxia-inducible factor 1-mediated inhibition of peroxisome proliferator-activated receptor {alpha} expression during hypoxia. J Immunol 166: 7543–7548, 2001.[Abstract/Free Full Text]

26. Papen M, Wodopia R, Bartsch P, and Mairbaurl H. Hypoxia-effects on Cai-signaling and ion transport activity of lung alveolar epithelial cells. Cell Physiol Biochem 11: 187–196, 2001.[CrossRef][ISI][Medline]

27. Pfaffl MW. A new mathematical model for relative quantification in real-time RT-PCR. Nucleic Acids Res 29: e45, 2001.

28. Rocha F, Musch MW, Lishanskiy L, Bookstein C, Sugi K, Xie Y, and Chang EB. IFN-{gamma} downregulates expression of Na+/H+ exchangers NHE2 and NHE3 in rat intestine and human Caco-2/bbe cells. Am J Physiol Cell Physiol 280: C1224–C1232, 2001.[Abstract/Free Full Text]

29. Schofield CJ and Ratcliffe PJ. Oxygen sensing by HIF hydroxylases. Nat Rev Mol Cell Biol 5: 343–354, 2004.[CrossRef][ISI][Medline]

30. Semenza GL. Hypoxia-inducible factor 1: master regulator of O2 homeostasis. Curr Opin Genet Dev 8: 588–594, 1998.[CrossRef][ISI][Medline]

31. Semenza GL, Agani F, Feldser D, Iyer N, Kotch L, Laughner E, and Yu A. Hypoxia, HIF-1, and the pathophysiology of common human diseases. Adv Exp Med Biol 475: 123–130, 2000.[ISI][Medline]

32. Sheridan WG, Lowndes RH, and Young HL. Intraoperative tissue oximetry in the human gastrointestinal tract. Am J Surg 159: 314–319, 1990.[CrossRef][ISI][Medline]

33. Song MS, Park YK, Lee JH, and Park K. Induction of glucose-regulated protein 78 by chronic hypoxia in human gastric tumor cells through a protein kinase C-{epsilon}/ERK/AP-1 signaling cascade. Cancer Res 61: 8322–8330, 2001.[Abstract/Free Full Text]

34. Sreenarasimhaiah J. Diagnosis and management of ischemic colitis. Curr Gastroenterol Rep 7: 421–426, 2005.[Medline]

35. Sugi K, Musch MW, Field M, and Chang EB. Inhibition of Na+,K+-ATPase by interferon {gamma} down-regulates intestinal epithelial transport and barrier function. Gastroenterology 120: 1393–1403, 2001.[CrossRef][ISI][Medline]

36. Synnestvedt K, Furuta GT, Comerford KM, Louis N, Karhausen J, Eltzschig HK, Hansen KR, Thompson LF, and Colgan SP. Ecto-5'-nucleotidase (CD73) regulation by hypoxia-inducible factor-1 (HIF-1) mediates permeability changes in intestinal epithelia. J Clin Invest 110: 993–1002, 2002.[CrossRef][ISI][Medline]

37. Taylor CT, Lisco SJ, Awtrey CS, and Colgan SP. Hypoxia inhibits cyclic nucleotide-stimulated epithelial ion transport: role for nucleotide cyclases as oxygen sensors. J Pharmacol Exp Ther 284: 568–575, 1998.[Abstract/Free Full Text]




This article has been cited by other articles:


Home page
 Am. J. Physiol. Cell Physiol.Home page
F. O'Mahony, F. Toumi, M. S. Mroz, G. Ferguson, and S. J. Keely
Induction of Na+/K+/2Cl- cotransporter expression mediates chronic potentiation of intestinal epithelial Cl- secretion by EGF
Am J Physiol Cell Physiol, June 1, 2008; 294(6): C1362 - C1370.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
291/2/C282    most recent
00564.2005v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (3)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Ibla, J. C.
Right arrow Articles by Colgan, S. P.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Ibla, J. C.
Right arrow Articles by Colgan, S. P.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Visit Other APS Journals Online
Copyright © 2006 by the American Physiological Society.