doi:10.1152/ajplung.00128.2011 301:L702-L711, 2011. First published 5 August 2011;Am J Physiol Lung Cell Mol Physiol

Moya and Rodrigo IturriagaRodrigo Del Rio, Cristian Muñoz, Paulina Arias, Felipe A. Court, Esteban A.body is not prevented by antioxidant treatment

carotidenlargement and VEGF upregulation in the rat Chronic intermittent hypoxia-induced vascular

You might find this additional info useful...

46 articles, 21 of which can be accessed free at:This article cites http://ajplung.physiology.org/content/301/5/L702.full.html#ref-list-1

including high resolution figures, can be found at:Updated information and services http://ajplung.physiology.org/content/301/5/L702.full.html

at: can be foundAJP - Lung Cellular and Molecular Physiologyabout Additional material and information

http://www.the-aps.org/publications/ajplung

This infomation is current as of January 20, 2012. 

American Physiological Society. ISSN: 1040-0605, ESSN: 1522-1504. Visit our website at http://www.the-aps.org/.year (monthly) by the American Physiological Society, 9650 Rockville Pike, Bethesda MD 20814-3991. Copyright © 2011 by theintegrative aspects of normal and abnormal function of cells and components of the respiratory system. It is published 12 times a

publishes original research covering the broad scope of molecular, cellular, andAJP - Lung Cellular and Molecular Physiology

on January 20, 2012ajplung.physiology.org

Dow

nloaded from

Chronic intermittent hypoxia-induced vascular enlargement and VEGFupregulation in the rat carotid body is not prevented by antioxidant treatment

Rodrigo Del Rio,1 Cristian Muñoz,1 Paulina Arias,1 Felipe A. Court,2 Esteban A. Moya,1

and Rodrigo Iturriaga1

1Laboratorio de Neurobiología, Facultad de Ciencias Biológicas and 2Millennium Nucleus for Regenerative Biology,Facultad de Ciencias Biológicas, P. Universidad Católica de Chile, Santiago, Chile

Submitted 5 May 2011; accepted in final form 29 July 2011

Del Rio R, Muñoz C, Arias P, Court FA, Moya EA, Iturriaga R.Chronic intermittent hypoxia-induced vascular enlargement and VEGFupregulation in the rat carotid body is not prevented by antioxidanttreatment. Am J Physiol Lung Cell Mol Physiol 301: L702–L711, 2011. Firstpublished August 5, 2011; doi:10.1152/ajplung.00128.2011.—Chronic in-termittent hypoxia (CIH), a characteristic of sleep obstructive apnea,enhances carotid body (CB) chemosensory responses to hypoxia, butits consequences on CB vascular area and VEGF expression areunknown. Accordingly, we studied the effect of CIH on CB volume,glomus cell numbers, blood vessel diameter and number, and VEGFimmunoreactivity (VEGF-ir) in male Sprague-Dawley rats exposed to5% O2, 12 times/h for 8 h or sham condition for 21 days. We foundthat CIH did not modify the CB volume or the number of glomus cellsbut increased VEGF-ir and enlarged the vascular area by increasingthe size of the blood vessels, whereas the number of the vessels wasunchanged. Because oxidative stress plays an essential role in theCIH-induced carotid chemosensory potentiation, we tested whetherantioxidant treatment with ascorbic acid may impede the vascularenlargement and the VEGF upregulation. Ascorbic acid, which pre-vents the CB chemosensory potentiation, failed to impede the vascularenlargement and the increased VEGF-ir. Thus present results suggestthat the CB vascular enlargement induced by CIH is a direct effect ofintermittent hypoxia and not secondary to the oxidative stress. Ac-cordingly, the subsequent capillary changes may be secondary to themechanisms involved in the neural chemosensory plasticity inducedby intermittent hypoxia.

blood vessels; oxidative stress

THE OBSTRUCTIVE SLEEP APNEA (OSA) syndrome, a growingsleep-breathing disorder, is recognized as an independent riskfactor for systemic hypertension and other cardiovascular dis-eases (26, 41). Among disturbances produced by OSA, thechronic intermittent hypoxia (CIH) is considered the mainfactor for the development of hypertension (12, 41). Record-ings of carotid body (CB) chemosensory discharges in animalshave shown that exposure to intermittent hypoxia selectivelyincreases the normoxic carotid discharges and potentiates thechemosensory responses to acute hypoxia (7, 17, 29, 30, 36).Accordingly, it has been proposed that the enhanced carotidchemosensory responsiveness to hypoxia contributes to theCIH-induced hypertension (10, 17, 35, 40, 46).

Similarly to the CB chemosensory potentiation induced byCIH, it is well known that chronic sustained hypoxia (i.e., highaltitude) progressively increases carotid chemosensory dis-charges and responses to hypoxia (1, 8, 25). In addition,

long-term exposure to sustained hypoxia produces hypertrophyand/or hyperplasia of glomus cells and enlarges the CB sizeand the vascular area (6, 14, 20, 24, 32, 45). Indeed, a tenfoldincrease in the volume of the CB capillaries in hypoxic rats hasbeen reported (6, 20). In contrast, much less is known about themorphological consequences of CIH on the CB. The availableinformation indicates that exposure of adult rats for 10 days toCIH does not modify the number of glomus cells or the CB size(30), but other morphological changes have not been studied.

Because intermittent hypoxia-induced angiogenesis is a cru-cial compensatory mechanism for providing oxygen supply todifferent tissues during hypoxic conditions (18, 19), we hy-pothesized that long-term CIH may enlarge the capillary vas-cular network in the rat CB. A possible mediator of the effectof intermittent hypoxia on CB blood vessels is the vascularendothelial growth factor (VEGF). Upregulation of VEGF inthe CB has been associated to the increased blood vessel areain rats exposed to chronic sustained hypoxia (3, 5). VEGF alsopromotes the generation of new coronary collaterals, whichmay play a main role in protecting the myocardium of patientswith a long OSA history (22). However, it is not knownwhether CIH may modify the number and/or the size of theblood vessels or increase the VEGF expression within the CBof adult rats.

It has been proposed that increased levels of reactive oxygenspecies (ROS) play a critical role in the CIH-induced potenti-ation of the CB chemosensory responses to hypoxia (7, 17, 28,29). Indeed, Peng et al. (29) proposed that the superoxideradical contributes to the CIH-induced CB chemosensory po-tentiation. They found that pretreatment of rats with a super-oxide dismutase mimetic for 10 days before and during theexposure to CIH prevents the enhanced CB chemosensoryresponses to hypoxia. Recently, we tested the hypothesis thatoxidative stress contributes to the CB chemosensory potentia-tion in rats exposed to CIH for 21 days (7). We found that oralsupplementation of ascorbic acid effectively prevents the in-creased plasma lipid peroxidation and the 3-nitrotyrosine for-mation within the CB, as well as the potentiation of carotidchemosensory responses to hypoxia. These observations sup-port a main role for oxidative stress in the generation of CBchemosensory potentiation induced by intermittent hypoxia.

Cellular responses to intermittent hypoxia result in the ac-cumulation of the hypoxic-inducible transcription factor-�(HIF-1�) in the CB (31) and in O2-sensitive PC12 cells (47).Because VEGF is downstream of the HIF-1� signaling path-way (39), it is likely that increased HIF-1� levels in the CBmay contribute to the upregulation of VEGF in rats exposed toCIH. Alternatively, VEGF may be directly upregulated byoxidative stress intermediates such as superoxide radical and

Address for reprint requests and other correspondence: R. Iturriaga,Laboratorio de Neurobiología, Facultad de Ciencias Biológicas, P. Univer-sidad Católica de Chile, Casilla 114-D, Santiago, Chile (e-mail:riturriaga@bio.puc.cl).

Am J Physiol Lung Cell Mol Physiol 301: L702–L711, 2011.First published August 5, 2011; doi:10.1152/ajplung.00128.2011.

1040-0605/11 Copyright © 2011 the American Physiological Society http://www.ajplung.orgL702

on January 20, 2012ajplung.physiology.org

Dow

nloaded from

peroxynitrite (34). Thus it is plausible that the CIH-inducedoxidative stress may enhance the VEGF expression in the ratCB. For that reason, we hypothesized that treatment with anantioxidant may affect the expected overexpression of VEGFand the enlarged vascular area in the CB. Thus we tested theeffects of ascorbic acid on the vascular area enlargement andVEGF immunoreactivity (VEGF-ir) in the CB of rats exposedto CIH for 21 days. As a control of the known effects ofascorbic acid on the potentiated CB oxygen chemoreceptioninduced by CIH (7), we recorded carotid chemosensory re-sponses to several levels of PO2 in the same CBs employed toanalyze the morphological and VEGF-ir changes.

MATERIALS AND METHODS

Animals. Experiments were performed on 34 male Sprague-Dawleyrats, weighting initially �200 g. Rats were fed with standard rat chowdiet ad libitum and kept on a 12-h:12-h light/dark schedule (8:00am-8:00 pm). The experimental procedures to obtain CB tissuesamples and to record carotid sinus chemosensory discharges wereperformed under sodium pentobarbitone anesthesia (40 mg/kg ip),followed by additional doses when necessary to maintain a level ofsurgical anesthesia. The experimental protocol was approved by theBio Ethical Committee of the Facultad de Ciencias Biológicas of thePontificia Universidad Católica of Chile.

Experimental groups and chronic exposure to intermittent hypoxia.Rats were randomly assigned into four experimental groups. The firstgroup was exposed to CIH for 21 days. The second group wasexposed to CIH for 21 days and received ascorbic acid (1.25 g/l,CIH-AA) in the drinking tap water from the first day of CIH exposureas previously described (7). The third group was exposed to shamconditions (air to air cycles), and the fourth group was exposed tosham conditions and received ascorbic acid. The ascorbic acid solu-tion was freshly prepared everyday and preserved in dark containersto avoid oxidation. Fluid intake was not significantly different between allgroups (�10 ml/day per 100 g body wt). Rats were exposed to intermit-tent hypoxia as previously described (7, 17). Unrestrained, freely movingrats housed in individual chambers (12 cm � 35 cm, 2.2 l) were exposedto a CIH protocol consisting of hypoxic cycles of 5–6% inspired O2 for20 s, followed by room air for 280 s, applied 12 times/h, 8 h/dayduring 21 days or to sham conditions. The chambers had a rear N2

inlet and a front air extractor, which enables the recovery to normoxia.A computerized system controls the valve inlets and the alternatingcycles of the extractors. During hypoxia, the extractors stopped, andthe rear solenoid valves allow 100% N2 flows into the chambers. TheO2 level in the chambers was continuously monitored with an oxygenanalyzer (Ohmeda 5120). The CO2 in the chamber was maintainedlow by continuous air extraction. In the sham condition, the hypoxiccycles were replaced by flushing compressed air into the chambers.The room temperature was kept at 23–25°C, and the hypoxic andsham patterns were applied during the animal dark phase (8:00 am to4:00 pm).

In situ recording of carotid chemosensory discharges. The carotidchemosensory discharge was measured as previously described (7,17). Pentobarbitone-anesthetized rats were placed in supine positionand tracheotomized, and the rectal temperature was maintained at 38.0 �0.5°C with a regulated heating pad. One carotid sinus nerve wasdissected and placed on a pair of platinum electrodes and covered withwarm mineral oil. The neural signal was preamplified (Grass P511),filtered (10–500 Hz), and fed to an electronic spike-amplitude dis-criminator allowing the selection of action potentials of given ampli-tude above the noise. Selected action potentials were counted with afrequency meter to assess the CB chemosensory frequency of dis-charge (fx), expressed in Hz. This procedure was performed by anindependent unbiased observer that had no information regarding thenature of the treatment. During the CB recordings, the contralateral

carotid sinus nerve was cut to prevent vascular and ventilatory reflexresponses induced by the activation of the CB chemoreflexes. Thechemosensory discharge was measured at normoxia and during acutehypoxic stimuli maintained until the response was in semi-steady state(�20 s). Rats breathed spontaneously during the whole experiments.At the end of the experiments, rats were killed by an overdose ofsodium pentobarbitone (100 mg/kg ip).

Carotid body morphology. Pentobarbitone-anesthetized rats wereperfused intracardially with PBS (pH 7.4) for 10 min, followed bybuffered paraformaldehyde (PFA 4%, Sigma) for 10 min. The salineand fixative solutions were perfused at a constant pressure of 95.0Torr. The carotid artery bifurcations including the CBs were harvestedfrom the rats and postfixed by immersion in buffered-PFA 4% for 12h at 4°C. Carotid bifurcations were dehydrated in graded ethanolseries followed by xylol and then included in paraffin. The CBs wereserially sectioned at a thickness of 5 �m, mounted on silanized slides,deparaffinized in xylol, and hydrated through graded ethanol to water.To study the effect of CIH on the number of glomus cells, wemeasured the number of positive immunoreactive tyrosine hydroxy-lase (TH) cells, a recognized marker of chemoreceptor cells, in foursections per each CB. Tissue sections were exposed to UV light for 30min and treated with NaBH4 0.1% to quench the tissue autofluores-cence. The CB histological sections were incubated for 1 h in blockingserum solution (Vector Laboratory) followed by incubation with ananti-TH monoclonal antibody (1:150 in PBS containing 0.5% TritonX-100 and 1% BSA; cat. no. MAB318, Millipore) overnight at 4°C ina humidity chamber. After being washed with PBS, tissue sectionswere incubated for 1 h with Alexa-Fluor 488 rabbit antimouse IgG(1:200, Molecular Probes) in PBS with 1% BSA and 0.5% TritonX-100. Finally, sections were mounted in DAPI-containing media(Vectashield, Vector Laboratory) and visualized using a fluorescentmicroscope (Eclipse E, Nikon).

The Cavalieri’s unbiased volume estimator (13) was used to esti-mate the CB volume and the length. Briefly, an unbiased counting grid(3,000 �m2 area) was projected on CB histological sections of 5 �m,taken at regular intervals of 25 �m across the organ (Image J, NIH).To delimit the organ, the point counting was done in sections whereglomus cells were present. The morphometric measurements of thevascular area, number, and average diameter of blood vessels wereperformed on the largest CB sectional area.

Using immunohistochemical detection of �-smooth muscle actin(�-SMA), we identified the type of blood vessels (arterioles vs.capillaries) present in the largest CB sectional area. The CB sectionswere incubated in blocking serum solution (Vector Laboratory) fol-lowed by incubation with a mixture of an anti-�-SMA monoclonalantibody (1:100 in PBS containing 0.5% Triton X-100 and 1% BSA;cat. no. A5228, Sigma) and an anti-TH polyclonal antibody (1:150 inPBS containing 0.5% Triton X-100 and 1% BSA; cat. no. AB152,Millipore) overnight. After being washed with PBS, tissue sectionswere incubated for 1 h with a mixture of Alexa-Fluor 488 goat antimouse IgG (1:200, Molecular Probes) and Alexa-Fluor 546 goatanti-rabbit IgG (1:200, Molecular Probes). Finally, sections weremounted in DAPI-containing media (Vectashield, Vector Laboratory)and visualized using a fluorescent microscope (Eclipse E, Nikon).

In a separate series of experiments, the CBs from sham andCIH-exposed rats were dissected and processed for transmissionelectron microscopy to addresses whether CIH may change the dis-tance between the glomus cells and the blood vessels. For betterrecognition of the capillaries under electron microscopy, the CBs werefixed overnight by immersion in cold 2.5% glutaraldehyde at 4°C in10 mM phosphate buffer, pH 7.2. Fixing the CB by immersion allowsus to preserve red blood cells within the capillary lumen. The CBswere cut into three thick slices under a dissection microscope, andsamples were treated with 1% osmium tetroxide in cacodylate buffer(pH 7.2) for 30 min and then with 2% aqueous uranyl acetate,dehydrated in ethanol, and embedded in an Epon. Areas to beexamined by electron microscopy were selected, and ultrathin sections

L703INTERMITTENT HYPOXIA AND CAROTID BODY MORPHOLOGY

AJP-Lung Cell Mol Physiol • VOL 301 • NOVEMBER 2011 • www.ajplung.org

on January 20, 2012ajplung.physiology.org

Dow

nloaded from

were obtained with an ultramicrotome and placed on 300-mesh copperelectron microscopy grids. Finally, the sections were counterstainedwith uranyl acetate and lead citrate and examined using a transmissionelectron microscope (Phillips Tecnai 12 Bio Twin). The quantificationof the distance between the blood vessels and the chemoreceptor cellswas performed in two to three different sections per CB in 5–6-�mdiameter capillaries. Glomus cells were identified by the presence ofdense core vesicles using the ImageJ software (NIH).

Three-dimensional reconstruction of CB. In addition to volumeestimation, we performed a three-dimensional (3-D) reconstruction ofthe CB for visual confirmation of the CIH-induced changes on theorgan volume. Regions of interest, including blood vessels, connec-tive tissue, and parenchyma, were then interactively discriminated inthe digital images by tracing their profiles. Serial CB section arrange-ments were performed according to anteroposterior and dorsoventralaxes using the Adobe Photoshop CS3 software, and the tissue recon-struction was made using the Reconstruction 3D software. Finally, theimages were rendered using the 3D MAX image software.

Immunohistochemistry for VEGF. Anesthetized rats were perfusedintracardially with PBS at pH 7.4 for 10 min followed by buffered 4%PFA (Sigma). Carotid bifurcations containing the CB were dissectedand postfixed in the same fixative solution, dehydrated in ethanol,included in paraffin, cut in 5-�m sections, and mounted on silanizedslides. Deparaffinized samples were incubated with 0.3% H2O2 toinhibit endogenous peroxidase and then in blocking serum solution(Vector Laboratory). The slices were incubated with an anti-VEGFmonoclonal antibody (1:50 in PBS/BSA 1%; cat. no. sc-7269, SantaCruz Biotechnology) overnight at 4°C. Slices were incubated with auniversal biotinylated secondary antibody followed by a ready-to-usestabilized ABC reagent (Vectastain Elite ABC Kit, Vector Labora-tory), and revealed with 3, 3-diaminobenzidine tetrachloride (Sigma).Samples were counterstained with Harris Hematoxylin and mounted.Photomicrographs were taken at �100 with a CCD camera coupled toan Olympus CX 31 microscope (Olympus), digitized, and analyzedwith the ImageJ software (NIH). The VEGF-ir intensity, averagedfrom four fields for each sampled CB, was expressed as opticalintegrated intensity.

Statistical data analysis. Data were expressed as means � SE.Comparisons between two groups were performed with the Student’st-test, and differences between more groups were assessed with one-or two-way ANOVA tests, followed by Bonferroni post hoc compar-isons. The distributions of the blood vessel diameter from differentexperimental groups were statistically compared using the Kolmogorov-

Smirnov test. All analyses were done with the statistical significanceset at P � 0.05.

RESULTS

Effects of intermittent hypoxia on CB morphology. Figure 1illustrates the 3-D reconstructions of one CB from a ratexposed to CIH for 21 days and one CB from a sham rat. Ongross inspection, no differences in volume were found. Themorphometric analysis performed in seven CBs from seven CIH-treated rats and seven CBs from seven sham rats showed that bothaxial length and CB volume were not significantly different in theexperimental and sham control groups (Table 1). Figure 2 showsrepresentative CB histological sections taken from a CIH-treatedrat and a sham rat. The glomus cells of CBs from CIH rats showedno evidence of hyperplasia. The nuclei of glomus and sustentac-ular cells did not show metaphase spindles or other characteristicsof chromosome aggregation. The nuclei of the endothelial cellsalso showed no evidence of recent mitosis, as metaphase chro-mosome figures were absent (Fig. 2). The number of TH-positivecells, an established marker of glomus cells, was not modified bythe CIH exposure (see Table 1 and Fig. 3). The positive immu-noreactive TH cells in the CBs from CIH-treated rats werenormally distributed in clusters compared with sham rats, and noevidence of hyperplasia or hypertrophy was found in the CBsfrom the CIH-rats.

Intermittent hypoxia induced changes in CB vasculature.Figure 2 shows a representative example of the effects of CIHon the size of blood vessels in the CB. Because the measure-

Fig. 1. Three-dimensional reconstruction of 1 carotid body (CB) from a chronic intermittent hypoxia (CIH)-exposed rat and 1 CB from a sham rat. No changein the CB volume was observed in the CIH-exposed rats. Scale bars � 100 �m. Green bar, anteroposterior axis; blue bar, dorsoventral axis; red bar, lateralleft-right axis.

Table 1. Morphometric analysis of carotid bodies from CIHand sham rats

CB Length, �mCB Total Volume,

�m3 � 106Number of GlomusCells,/�m2 � 10�3

CIH 366.4 � 15.4* 20.0 � 2.4* 3.83 � 0.69*Sham 397.3 � 38.9* 16.6 � 2.3* 4.56 � 0.46*

Data are expressed as means � SE.. Sham are control rats. *P 0.05;chronic intermittent hypoxia (CIH) vs. sham. Student’s t-test. n � 7 carotidbodies (CBs) from 7 rats and in each group.

L704 INTERMITTENT HYPOXIA AND CAROTID BODY MORPHOLOGY

AJP-Lung Cell Mol Physiol • VOL 301 • NOVEMBER 2011 • www.ajplung.org

on January 20, 2012ajplung.physiology.org

Dow

nloaded from

ments of the vascular area were performed on the largest CBsectional area (i.e., the vascular center of the organ), most ofthe blood vessels included in the analysis were capillaries.Indeed, as is shown in Fig. 4, little positive staining for �-SMAis observed in the largest CB sectional area from a rat exposed

to CIH, indicating that most of the blood vessels were devoidof smooth muscle layers. The exposure to CIH resulted in asignificant increase of the CB vascular area expressed as apercentage of the sectional area (Fig. 5A) despite the fact thatthere were no significant differences between the sectional areastudied in both CIH-treated and sham rats. The mean CBsectional area studied was 48,300 � 6,350 and 43,050 � 3,350�m2, respectively, in the CIH and sham group (P 0.05;Student’s t-test; n � 7 CBs from 7 CIH and 7 sham rats). Theincreased CB vascular area found in CIH rats was due to asignificant increase in the mean area of the blood vessels (Fig.5B) but not in the number of vessels (Fig. 5C). Indeed, wefound that CIH modified the distribution of the blood vesseldiameter in the sectional area of the CB (Fig. 6, A and B). TheKolmogorov-Smirnov test showed a significant (P � 0.05)shift from small vessels (�10 �m diameter) toward largevessels (10 �m). Thus CIH increased the size of CB bloodvessels, rather than a generation of new blood vessels.

The quantification of the distance between the capillary vesselsand glomus cells under electron microscopy (see Fig. 7) showeda significant decrease in the CIH group compared with shamconditions (3.17 � 0.18 �m vs. 4.66 � 0.44 �m, in CIH andsham rats, respectively; n � 3 CB in each group, P � 0.01).Thus CIH did not modify the CB volume because the CIH-induced increased vascular area is accompanied by a decreasein the distance between the chemoreceptor cells and bloodvessels.

Increased VEGF-ir in the CB of rats exposed to intermittenthypoxia. Figure 8, A and B, shows CB histological sectionsstained with an anti-VEGF antibody and counterstained withhematoxylin to visualize nuclei morphology. CIH induced asignificant threefold increase in the integrated optical intensityof the VEGF immunoreactivity (P � 0.001, see Table 2). Thepositive VEGF-ir was mainly confined to clusters of cells,which have been defined as glomus cells in classical lightmicroscope studies. Indeed, most of the positive VEGF-irstaining was found in clusters of round to ovoid cells, withprominent nuclei (average diameter of �10 �m). It is worthnoting that not all the glomus cells expressed VEGF-ir,whereas a relatively small expression of VEGF was found invascular territories.

Fig. 2. Histological sections of 1 CB from a CIH-rat (A) and 1 from a sham rat(B). Sections were stained with Harris-Hematoxylin. Note that the glomus andendothelial cells did not show any evidence of hyperplasia or mitotic figures.Filled arrows, glomus cells nuclei; empty arrows, sustentacular cell nuclei;asterisks, blood vessels. Scale bar � 20 �m.

Fig. 3. Immunolocalization of tyrosine hy-droxylase (TH)-positive cells in 1 CB from aCIH rat and 1 CB from a sham rat. Noapparent differences in the number of TH-positive cells were observed. Green, TH-immunoreactive staining; blue, nuclei stain-ing with DAPI. Scale bar � 80 �m.

L705INTERMITTENT HYPOXIA AND CAROTID BODY MORPHOLOGY

AJP-Lung Cell Mol Physiol • VOL 301 • NOVEMBER 2011 • www.ajplung.org

on January 20, 2012ajplung.physiology.org

Dow

nloaded from

Effects of ascorbic acid on morphological and physiologicalchanges induced by CIH. Because oxidative stress plays anessential role in the CIH-induced potentiation of carotid che-mosensory responses to hypoxia, we tested whether ascorbicacid, which prevents the chemosensory potentiation induced byCIH (7), may reduce or impede the vascular enlargement andthe VEGF-ir upregulation. We found that ascorbic acid did notimpede the increment of the vascular area (Fig. 5, A–C), theshift toward large vessel diameter (Fig. 6C), and the increasedintegrated optical intensity of VEGF-ir induced by CIH (Fig. 8and Table 2). Nevertheless, ascorbic acid prevented the carotidchemosensory potentiation induced by CIH in the same CBsused to assess the effects of the antioxidant on the morphologyand VEGF-ir. It is worth noting that we did not find morpho-logical changes in two CBs from two different sham ratstreated with ascorbic acid. Indeed, the vascular area expressedas percentage of the sectional area in these CBs from sham ratstreated with the antioxidant was 13.1 � 1.8%, a value notsignificantly different compared with sham rats (P 0.05, Fig.5). In addition, we found that ascorbic acid did not modify themean vessel area (41.3 � 10.2 �m2 vs. 59.0 � 7.7 �m2 insham rats treated with and without ascorbic acid, P 0.05) andthe number of blood vessels (122.1 � 33.9 vs. 131.7 � 12.8 insham rats treated with and without ascorbic acid, P 0.05).

DISCUSSION

General. We studied the effects of intermittent hypoxia onthe CB size, number of glomus cells, number and distributionof blood vessel diameter, and VEGF-ir in rats exposed to CIHfor 21 days, which presented enhanced carotid chemosensorydischarges in normoxia and in response to acute hypoxia. Themain findings are the following: 1) CIH induced vascularenlargement in the adult rat CB but did not modify the volumeof the organ or the number of glomus cells; 2) the enlargementof the vascular area induced by CIH was due to the increasedsize of capillary blood vessels, rather than the formation of newvessels; 3) the vascular increased area in the CB from CIH ratswas associated to a decrease in the distance between thechemoreceptor cells and capillaries along with a marked in-crease in VEGF-ir; and 4) the treatment with ascorbic acid didnot impede the enlargement of the vascular area and theupregulation of VEGF-ir. Thus present results agree and ex-tend previous observations showing that exposure to CIH for10 days did not modify the CB volume or the number ofglomus cells in adult rats (29) and add new informationindicating that CIH promotes the upregulation of VEGF andthe enlargement of capillary blood vessels in the CB, which arenot blocked by antioxidant treatment with ascorbic acid.

Fig. 4. Immunohistochemical detection of�-smooth muscle actin (�-SMA) in 1 CBfrom a CIH-exposed rat. Note the presenceof little positive staining for �-SMA. Arrow,positive �-SMA blood vessel; asterisk, bloodvessels showing no staining for �-SMA.Scale bar � 60 �m.

L706 INTERMITTENT HYPOXIA AND CAROTID BODY MORPHOLOGY

AJP-Lung Cell Mol Physiol • VOL 301 • NOVEMBER 2011 • www.ajplung.org

on January 20, 2012ajplung.physiology.org

Dow

nloaded from

Morphological changes induced by intermittent hypoxia.The CB is the main arterial chemoreceptor that senses thearterial levels of PO2, playing a main role in cardiorespiratoryhomeostasis (15, 16). The CB morphology and function arelargely dependent on the inspired O2 level (Fig. 9). It is wellknown that long-term sustained hypoxia enlarges the CB,producing hyperplasia and/or hypertrophy of glomus cells (6,21, 24, 32). On the contrary, present results showed thatlong-term exposure to intermittent hypoxia did not modify thenumber of TH-positive cells or the volume of the CB of ratssubmitted to CIH for 21 days. Thus present results agree with

previous findings reported by Peng et al. (29), showing that thevolume and number of glomus cells in the adult rat CB wasunaltered by 10 days of CIH.

Present results showed that CIH did not modify the CBvolume but induced an increase in the vascular area. We foundthat CBs from rats exposed to CIH show a significant decreasein the distance between the chemoreceptor cells and bloodvessels (Fig. 7). This result partially explains the lack of effectof the vascular enlargement in the CB volume. The increasedvascular area induced by CIH did not result from the genera-tion of new blood vessels but from the enlargement of blood

Fig. 6. Distribution of blood vessel diameter in CIH-treated rats (A), sham rats(B), and CIH-treated rats with ascorbic acid (C). Intermittent hypoxia produceda significant shift on the distribution of blood vessel diameters toward largevessel (10 �m) diameters in the CBs from rats exposed to CIH. Note thatascorbic acid did not impede the CIH-induced shift in vessel diameters. Thedistribution of CB blood vessels from CIH rats was significantly differentcompared with the distribution of blood vessels of sham rats P � 0.05,Kolmogorov-Smirnov test; n � 7 CBs per group.

Fig. 5. Increased vascular area in CBs from CIH rats is likely due to bloodvessel enlargement. A: vascular area, expressed as a percentage of the CBsectional area. B: average area of blood vessel within the CB sectional area.C: number of blood vessels per field. **P � 0.01; *P � 0.05; n.sP 0.05;Bonferroni test after one-way ANOVA; n � 7 CBs for CIH and sham group,n � 5 in CIH-rats treated with ascorbic acid (CIH AA).

L707INTERMITTENT HYPOXIA AND CAROTID BODY MORPHOLOGY

AJP-Lung Cell Mol Physiol • VOL 301 • NOVEMBER 2011 • www.ajplung.org

on January 20, 2012ajplung.physiology.org

Dow

nloaded from

vessels. The twofold increase in the CB blood vessels areainduced by CIH found in this study was due to a significantshift in the distribution of blood vessels toward large diameters(10 �m) but not an increase in the number of blood vesselsin the CB from CIH rats (Figs. 5 and 6). Changes in thevascularization of an adult organ may result from two pro-cesses: angiogenesis defined as the formation of new vessels bysprouting from preexisting blood vessels, and arteriogenesis,which consisted of the growth of preexisting vessels (2, 38).Our results support the idea that CIH produced arteriogenesisrather than angiogenesis in the adult rat CB. The changes insize of blood vessels found here cannot be attributed to thefixative procedure of the CB tissue because sham and CIH ratswere perfused with the fixative solution at the same physio-logical pressure (�95.0 Torr). Thus the enlargement of thevasculature observed in the CIH group represents the effect ofintermittent hypoxia.

VEGF and intermittent hypoxia. To our knowledge, this isthe first study showing that adult rats exposed to intermittenthypoxia, which developed hypertension and enhanced CBchemosensory responses to hypoxia (7, 17), show a simulta-neous increase of the vascular area and the VEGF-ir in the CB.We found that the increased diameter of CB blood vessels wasassociated with an increase in VEGF expression, which is

considered to promote vascular remodeling in the CB from ratsexposed to chronic sustained hypoxia (9). Hypoxia is a well-known stimulus that upregulates the expression of VEGF andinduces posttranscriptional VEGF mRNA stability, playing acritical role in the formation of new blood vessels (11), and inthe remodeling of vascular beds under hypoxic conditions (23,44). Chen et al. (3, 5) found that sustained hypoxia producesVEGF overexpression in the rat CB, suggesting that VEGFplays a key role in the vascular effects induced by hypoxia inthe rat CB. In addition, Tipoe and Fung (42) reported thatthe adult rat CB expressed VEGF receptors in normoxic

Fig. 8. Representative photomicrographs showing positive immunoreactivityfor VEGF in CBs from a CIH rat (A), a sham rat (B), and a CIH rat treated withascorbic acid (C). Sections were counterstained with Harris-Hematoxylin.Scale bar � 20 �m.

Fig. 7. Exposure to CIH induces a decrease in the distance between glomuscells and capillaries. Representative images of a sham and a CIH-exposed ratCB, obtained by transmission electron microscopy. Asterisk represent glomuscells. Scale bar � 6 �m.

L708 INTERMITTENT HYPOXIA AND CAROTID BODY MORPHOLOGY

AJP-Lung Cell Mol Physiol • VOL 301 • NOVEMBER 2011 • www.ajplung.org

on January 20, 2012ajplung.physiology.org

Dow

nloaded from

conditions, whereas the exposure to sustained hypoxia in-duced the overexpression of VEGF type I and II receptors.Thus the increased levels of VEGF-ir in the CB of CIH ratsmay be involved in the growth of preexisting vessels. Most ofthe blood vessels analyzed in this study were capillaries be-cause we did not find positive staining for smooth muscles cellsin the central part of the CB (Fig. 4). Therefore, present resultssuggest that CIH enlarges the capillaries, but we cannot ex-clude the possibility that some arterioles and/or veins wereaffected by intermittent hypoxia. Because our measurementswere performed in the central part of the organ, a further studyusing three-dimensional histological reconstruction and spe-cific proliferation antibodies is required to solve this question.

CB potentiation and vascular changes induced by intermit-tent hypoxia: role of oxidative stress. The mechanisms under-lying the CB potentiation induced by CIH are not entirelyknown, but previous observations support the idea that CIHincreases the endothelinergic tone in the CB (37), which islikely secondary to an increased ROS level (28). We cannotpreclude that CIH may affect the release of other molecules(such as dopamine, histamine, ACh, and ATP), which havebeen proposed as transmitter or sensory modulators betweenthe glomus cells and the petrosal neurons (15, 16). ROS andreactive nitrogen species (RNS) have been proposed as keymediators of the cardiovascular alterations in animals exposedto CIH (4, 7, 28, 29). Studies performed in patients with OSAand CIH-exposed animals showed that the cyclic episodes ofhypoxia-reoxygenation produce systemic oxidative stress at-tributable to the accumulation of ROS and RNS, which arepotential sources of cellular damage. Peng et al. (29) providedevidence that the superoxide radical participates in the poten-tiation of the rat CB chemosensory responses to hypoxiainduced by CIH. Recently, we found that CIH increasedplasma lipid peroxidation, 3-nitrotyrosine formation, and iNOSexpression in the CB, enhanced CB chemosensory and venti-latory hypoxic responses, and produced hypertension (7). In-deed, ascorbic acid treatment prevents the increases in plasmalipid peroxidation and 3-nitrotyrosine formation within the CB,the enhanced CB chemosensory and ventilatory responses tohypoxia, as well as the hypertension. It is well known that otherantioxidants also prevent the CB chemosensory potentiationand the hypertension induced by CIH. Indeed, antioxidanttreatment with manganese (III) tetrakis(1-methyl-4-pyridyl)-porphyrin (MnTMPyP) in rats prevents the CB oxidativestress, the chemosensory potentiation, and the hypertension(28). Similarly, Tempol (43) impedes the elevation of arterialpressure in rats exposed to CIH. However, the effects of otherantioxidants on CIH-induced vascular enlargement in the adultrat CB have not been studied. Further studies are needed to

address the contribution of specific ROS on CIH-induced CBmorphological alterations.

Present results show that ascorbic acid did not impede theenlargement of vascular area and VEGF-ir in the adult rat CB,suggesting that these changes are separated from the functionalCB potentiation. A similar dissociation between functional andmorphological changes induced by sustained hypoxia wasreported by Pequignot et al. (33). They found that sustainedhypoxia induced the enlargement of the rat CB by increasingthe vascular area and by producing hypertrophy of glomic andinterstitial tissues. The dopamine and norepinephrine contentin the CB increased 40–50 times compared with the controls.Treatment with DL-propranolol abolished the vasodilatory ef-fect of hypoxia within the first week but did not prevent theother structural changes or the rise in catecholamine content.These results suggest that the vasodilatation induced by sus-tained hypoxia is controlled by -adrenergic receptors,whereas the structural and biochemical events occurring in theCB are controlled by different mechanisms. The fact thatascorbic acid did not prevent the increased expression ofVEGF-ir and the vascular enlargement suggests that intermit-tent hypoxia per se and not the increased oxidative stress in theCB produced by CIH was the primary stimulus for VEGFupregulation.

In summary, present results show that CIH increased thevascular area in the adult rat CB, mainly attributable to anenlargement of the capillary diameter associated with an in-creased VEGF-ir. Ascorbic acid treatment, which preventedthe CB chemosensory potentiation induced by CIH, failed toprevent vasculature enlargement and the increased VEGF-ir inthe CB of CIH-treated rats, suggesting that the increasedvascular area induced by CIH is a direct effect of intermittenthypoxia and not secondary to the oxidative stress. Thus ourresults suggest that capillary changes induced by intermittenthypoxia may be secondary effects from mechanisms involvedin CB chemosensory plasticity.

ACKNOWLEDGMENTS

We thank Dr. Julio Alcayaga for assistance in the construction and opera-tion of the hypoxic chambers and Ms. Mónica Pérez for excellent EMprocessing.

Table 2. Immunohistochemical VEGF-positive stainingquantification in CB

Sectional Area, �m2 Integrated Optical Density, a.u.

CIH 74,075 � 11,147 42.5 � 4.4Sham 66,483 � 4,596* 13.2 � 0.9†CIH AA 69,630 � 1,071* 40.3 � 2.1*

Data are expressed as means � SE. *P 0.05 compared with CIH group;†P � 0.001 compared with CIH group. Bonferroni test after one-wayANOVA; n � 5–6 CBs in each group. a.u., arbitrary units; CIH AA, CIHsupplemented with ascorbic acid.

Fig. 9. CB chemosensory responses to several levels of inspired PO2 in shamrats (dashed line and Œ), CIH-exposed rats (solid line and �), and rats exposedto CIH and treated with ascorbic acid (solid line and �) in 4 sham rats, 5CIH-treated rats, and 5 CIH-treated rats with ascorbic acid. fx, frequency ofcarotid chemosensory discharges, expressed in Hz. ***P � 0.001; *P � 0.05compared with sham rats. Bonferroni test after 2-way ANOVA.

L709INTERMITTENT HYPOXIA AND CAROTID BODY MORPHOLOGY

AJP-Lung Cell Mol Physiol • VOL 301 • NOVEMBER 2011 • www.ajplung.org

on January 20, 2012ajplung.physiology.org

Dow

nloaded from

GRANTS

This work was supported by grant 1100405 from the National Fund forScientific and Technological Development of Chile (FONDECYT) to R.Iturriaga and Millennium Nucleus no. P-07-011-F to F. Court. Rodrigo Del Riowas supported by a CONICYT AT-24091043 fellowship.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by theauthors.

REFERENCES

1. Bisgard GE. Carotid body mechanisms in acclimatization to hypoxia.Respir Physiol 121: 137–246, 2000.

2. Buschmann I, Schaper W. Arteriogenesis versus angiogenesis: twomechanisms of vessel growth. News Physiol Sci 14: 121–125, 1999.

3. Chen J, Dinger B, Jyung R, Stensaas L, Fidone S. Altered expressionof vascular endothelial growth factor and FLK-1 receptor in chroni-cally hypoxic rat carotid body. Adv Exp Med Biol 536: 583–591, 2003.

4. Chen L, Einbinder E, Zhang Q, Hasday J, Balke CW, Scharf SM.Oxidative stress and left ventricular function with chronic intermittenthypoxia in rats. Am J Respir Crit Care Med 172: 915–920, 2005.

5. Chen J, He L, Liu X, Dinger B, Stensaas L, Fidone S. Effect of theendothelin receptor antagonist bosentan on chronic hypoxia-induced mor-phological and physiological changes in rat carotid body. Am J PhysiolLung Cell Mol Physiol 292: L1257–L1262, 2007.

6. Clarke JA, Daly MB, Marshall JM, Ead HW, Hennessy EM. Quanti-tative studies of the vasculature of the carotid body in the chronicallyhypoxic rat. Braz J Med Biol Res 33: 331–340, 2000.

7. Del Rio R, Moya EA, Iturriaga R. Carotid body and cardiorespiratoryalterations in intermittent hypoxia: the oxidative link. Eur Respir J 36:143–150, 2010.

8. Di Giulio C, Huang WX, Mokashi A, Roy A, Cacchio M, Macr� MA,Lahiri S. Sustained hypoxia promotes hyperactive response of carotidbody in the cat. Respir Physiol Neurobiol 134: 69–74, 2003.

9. Dinger B, He L, Chen J, Stensaas LJ, Fidone SJ. Mechanisms ofmorphological and functional plasticity in the chronically hypoxic carotidbody. In: Oxygen Sensing, Responses and Adaptation to Hypoxia, editedby Lahiri S, Semenza GL, Prabhakar NR. New York: Dekker, 2003, p.439–465.

10. Feng J, Chen BY, Cui LY. Carotid body-mediated changes of sympa-thetic nerve and their relationships with hypertension. Chin Med J (Engl)121: 1732–1735, 2008.

11. Folkman J, Shing Y. Angiogenesis. J Biol Chem 267: 10931–10934,1992.

12. Gozal D, Kheirandish-Gozal L. Cardiovascular morbidity in obstructivesleep apnea, oxidative stress, inflammation, and much more. Am J RespirCrit Care Med 177: 369–375, 2008.

13. Gundersen HJG, Jensen EB. The efficiency of systematic sampling instereology and its prediction. J Microsc 147, 229–262, 1987.

14. Heath D, Jago R, Smith P. The vasculature of the carotid body. Cardio-vasc Res 17: 33–42, 1983.

15. Iturriaga R, Alcayaga J. Neurotransmission in the carotid body: trans-mitters and modulators between glomus cells and petrosal ganglion nerveterminals. Brain Res Rev 47: 46–53, 2004.

16. Iturriaga R, Varas R, Alcayaga J. Electrical and pharmacologicalproperties of petrosal ganglion neurons that innervate the carotid body.Respir Physiol Neurobiol 157: 130–139, 2007.

17. Iturriaga R, Moya EA, Del Rio R. Carotid body potentiation inducedby intermittent hypoxia: implications for cardiorespiratory changesinduced by sleep apnoea. Clin Exp Pharmacol Physiol 36: 1197–1204,2009.

18. Kalaria RN, Spoors L, Laude EA, Emery CJ, Thwaites-Bee D, FairlieJ, Oakley AE, Barer DH, Barer GR. Hypoxia of sleep apnoea: cardio-vascular and cerebral changes after intermittent hypoxia in rats. RespirPhysiol Neurobiol 140: 53–62, 2004.

19. Kanaan A, Farahani R, Douglas RM, LaManna JC, Haddad GG.Effect of chronic continuous or intermittent hypoxia and reoxygenation oncerebral capillary density and myelination. Am J Physiol Regul IntegrComp Physiol 290: R1105–R1114, 2006.

20. Laidler P, Kay JM. The effect of chronic hypoxia on the number andnuclear diameter of type I cells in the carotid bodies of rats. Am J Pathol79: 311–320, 1975.

21. Laidler P, Kay JM. A quantitative study of some ultrastructural featuresof the type I cells in the carotid bodies of rats living at a simulated altitudeof 4300 meters. J Neurocytol 7: 183–192, 1978.

22. Lavie L, Lavie P. Ischemic preconditioning as a possible explanation forthe age decline relative mortality in sleep apnea. Med Hypotheses 66:1069–1073, 2006.

23. Lee CG, Link H, Baluk P, Homer RJ, Chapoval S, Bhandari V, KangMJ, Cohn L, Kim YK, McDonald DM, Elias JA. Vascular endothelialgrowth factor (VEGF) induces remodeling and enhances TH2-mediatedsensitization and inflammation in the lung. Nat Med 10: 1095–10103,2004.

24. McGregor KH, Gil J, Lahiri S. A morphometric study of the carotidbody in chronically hypoxic rats. J Appl Physiol 57: 1430 –1438, 1984.

25. Nielsen AM, Bisgard GE, Vidruk EH. Carotid chemoreceptor activityduring acute and sustained hypoxia in goats. J Appl Physiol 65: 1796–1802, 1988.

26. Parati G, Lombardi C, Narkiewicz K. Sleep apnea: epidemiology,pathophysiology, and relation to cardiovascular risk. Am J Physiol RegulIntegr Comp Physiol 293: R1671–R1683, 2007.

27. Pawar A, Peng YJ, Jacono FJ, Prabhakar NR. Comparative analysis ofneonatal and adult rat carotid body responses to chronic intermittenthypoxia. J Appl Physiol 104: 1287–1294, 2008.

28. Pawar A, Nanduri J, Yuan G, Khan SA, Wang N, Kumar GK,Prabhakar NR. Reactive oxygen species-dependent endothelin signalingis required for augmented hypoxic sensory response of the neonatal carotidbody by intermittent hypoxia. Am J Physiol Regul Integr Comp Physiol296: R735–R742, 2009.

29. Peng YJ, Overholt JL, Kline D, Kumar GK, Prabhakar NR. Inductionof sensory long-term facilitation in the carotid body by intermittenthypoxia: implications for recurrent apneas. Proc Natl Acad Sci USA 100:10073–10078, 2003.

30. Peng YJ, Prabhakar NR. Effect of two paradigms of chronic intermittenthypoxia on carotid body sensory activity. J Appl Physiol 96: 1236–1242,2004.

31. Peng YJ, Yuan G, Ramakrishnan D, Sharma SD, Bosch-Marce M,Semenza GL, Pranhakar NR. Heterozygous HIF-1alpha deficiency im-pairs carotid body-mediated systemic responses and reactive oxygenspecies generation in mice exposed to intermittent hypoxia. J Physiol 577:705–716, 2006.

32. Pequignot JM, Hellström S, Johansson C. Intact and sympathectomizedcarotid bodies of long-term hypoxic rats: a morphometric ultrastructuralstudy. J Neurocytol 13: 481–493, 1984.

33. Pequignot JM, Hellström S, Johansson C. DL-propranolol inhibits thevascular changes in the rat carotid body Induced by long-term hypoxia.Virchows Arch 411: 331–336, 1987.

34. Platt DH, Bartoli M, El-Remessy AB, Al-Sharbrawey M, Lemtalsi T,Fulton D, Caldwell RB. Peroxynitrite increases VEGF expression invascular endothelial cells via STAT3. Free Radic Biol Med 39: 1353–1361, 2005.

35. Prabhakar NR, Peng YJ, Kumar GK, Pawar A. Altered carotid bodyfunction by intermittent hypoxia in neonates and adults: relevance torecurrent apneas. Respir Physiol Neurobiol 157: 148–153, 2007.

36. Rey S, Del Rio R, Alcayaga J, Iturriaga R. Chronic intermittent hypoxiaenhances cat chemosensory and ventilatory responses to hypoxia. JPhysiol 560: 577–586, 2004.

37. Rey S, Del Rio R, Iturriaga R. Contribution of endothelin-1 to theenhanced carotid body chemosensory responses induced by chronic inter-mittent hypoxia. Brain Res 1086: 152–159, 2006.

38. Schaper W. Colateral circulation: past and present. Basic Res Cardiol104: 5–21, 2009.

39. Semenza GL, Agani F, Iyer N, Kotch L, Laughner E, Leung S, Yu A.Regulation of cardiovascular development and physiology by hypoxia-inducible factor 1. Ann NY Acad Sci 874: 262–268, 1999.

40. Smith ML, Pacchia CF. Sleep apnoea and hypertension: role of chemore-flexes in humans. Exp Physiol 92: 45–50, 2007.

41. Somers VK, White DP, Amin R, Abraham WT, Costa F, Culebras A,Daniels S, Floras JS, Hunt CE, Olson LJ, Pickering TG, Russell R,Woo M, Young T. Sleep Apnea and cardiovascular disease. J Am CollCardiol 52: 686–717, 2008.

42. Tipoe Fung ML GL. Expression of HIF-1alpha, VEGF and VEGFreceptors in the carotid body of chronically hypoxic rat. Respir PhysiolNeurobiol 138: 143–154, 2003.

43. Troncoso Brindeiro CM, da Silva AQ, Allahdadi KJ, Youngblood V,Kanagy NL. Reactive oxygen species contribute to sleep apnea-induced

L710 INTERMITTENT HYPOXIA AND CAROTID BODY MORPHOLOGY

AJP-Lung Cell Mol Physiol • VOL 301 • NOVEMBER 2011 • www.ajplung.org

on January 20, 2012ajplung.physiology.org

Dow

nloaded from

hypertension in rats. Am J Physiol Heart Circ Physiol 293: H2971–H2976,2007.

44. Voelkel NF, Tuder RM. Hypoxia-induced pulmonary vascular remodel-ing: a model for what human disease? J Clin Invest 106: 733–738, 2000.

45. Wang ZY, Bisgard GE. Chronic hypoxia-induced morphological andneurochemical changes in the carotid body. Microsc Res Tech 59: 168–177, 2002.

46. Weiss JW, Liu MD, Huang J. Physiological basis for a causal relation-ship of obstructive sleep apnoea to hypertension. Exp Physiol 92: 21–26,2007.

47. Yuan G, Nanduri J, Khan S, Semenza GL, Prabhakar NR. Induc-tion of HIF-1alpha expression by intermittent hypoxia: involvement ofNADPH oxidase, Ca2� signaling, prolyl hydroxylases, and mTOR. JCell Physiol 217: 674 –685, 2008.

L711INTERMITTENT HYPOXIA AND CAROTID BODY MORPHOLOGY

AJP-Lung Cell Mol Physiol • VOL 301 • NOVEMBER 2011 • www.ajplung.org

on January 20, 2012ajplung.physiology.org

Dow

nloaded from