Network analysis of temporal effects of intermittent and ...€¦ · Network analysis of temporal...

Network analysis of temporal effects of intermittent and sustained hypoxiaon rat lungs

Wei Wu,1,2* Nilesh B. Dave,2* Guoying Yu,1,2 Patrick J. Strollo,2 Elizabeta Kovkarova-Naumovski,1,2

Stefan W. Ryter,2 Stephen R. Reeves,3 Ehab Dayyat,3 Yang Wang,3 Augustine M. K. Choi,2 David Gozal,3

and Naftali Kaminski1,2

1Dorothy P. and Richard P. Simmons Center for Interstitial Lung Disease, 2Division of Pulmonary, Allergy, and CriticalCare Medicine, School of Medicine, University of Pittsburgh, Pittsburgh, Pennsylvania; and 3Kosair Children’s HospitalResearch Institute and Division of Pediatric Sleep Medicine, Department of Pediatrics, University of Louisville,Louisville, Kentucky

Submitted 2 November 2007; accepted in final form 23 September 2008

Wu W, Dave NB, Yu G, Strollo PJ, Kovkarova-Naumovski E, RyterSW, Reeves SR, Dayyat E, Wang Y, Choi AMK, Gozal D, Kaminski N.Network analysis of temporal effects of intermittent and sustained hypoxia onrat lungs. Physiol Genomics 36: 24–34, 2008. First published September 30,2008; doi:10.1152/physiolgenomics.00258.2007.—The molecular networksunderlying the lung response to hypoxia are not fully understood. Weemployed systems biology approaches to study temporal effects ofintermittent or sustained hypoxia on gene expression in rat lungs. Weobtained gene expression profiles from rats exposed to intermittent orsustained hypoxia lasting 0–30 days and identified differentiallyexpressed genes, their patterns, biological processes, and regulatorynetworks critical for lung response to intermittent or sustained hyp-oxia. We validated selected genes with quantitative real-time PCR.Intermittent and sustained hypoxia induced two distinct sets of genesin rat lungs that displayed different temporal expression patterns.Intermittent hypoxia induced genes mostly involved in ion transportand homeostasis, neurological processes, and steroid hormone recep-tor activity, while sustained hypoxia induced genes principally par-ticipating in immune responses. The intermittent hypoxia-activatednetwork suggested a role for cross talk between estrogen receptor 1(ESR1) and other key proteins in hypoxic responses. The sustainedhypoxia-activated network was indicative of vascular remodeling andpulmonary hypertension. We confirmed the temporal expressionchanges of 12 genes (including the Esr1 gene and 4 ESR1 targetgenes) in intermittent hypoxia and 8 genes in sustained hypoxia withquantitative real-time PCR. Conclusions: intermittent and sustainedhypoxia induced distinct gene expression patterns in rat lungs. Thefunctional characteristics of genes activated by these two distinctperturbations suggest their roles in the downstream physiologicaleffects of intermittent and sustained hypoxia. Our results demonstratethe discovery potential of applying systems biology approaches to theunderstanding of mechanisms underlying hypoxic lung response.

temporal gene expression; microarray analysis; hypoxic response;systems biology

HYPOXIA HAS PROFOUND physiological effects. Traditionally,studies examining pathophysiological effects of hypoxia havemainly focused on sustained (or continuous) hypoxia (SH),particularly in the context of high-altitude physiology. More

recently, with the increased interest in sleep-disordered breath-ing, the physiological effects of intermittent hypoxia (IH) havealso been studied (27), and response differences between thesetwo hypoxic paradigms have been outlined.

In the clinical setting, IH and SH are found in distinctlydifferent conditions. IH is primarily associated with sleep-disordered breathing, which has been increasingly recognizedas a highly prevalent and clinically important set of disorders(25). In contrast, SH is associated with chronic obstructivepulmonary disease, pulmonary fibrosis, and other parenchymallung disorders that adversely affect alveolar gas exchange (16).The morbid consequences of these two conditions also differ:while IH is mainly associated with systemic hypertension,cardiovascular morbidity and mortality, and neuronal and hu-moral disorders, long-term SH most commonly leads to pul-monary hypertension and right ventricular (RV) failure (10, 26,28, 29, 32, 33).

In physiological terms, IH and SH exert somewhat differenteffects on the vascular, cardiac, and neurological systems. IHinduces systemic arterial blood pressure (BP), increased sym-pathetic nervous activity, and respiratory long-term facilitation(15, 20), while SH may induce pulmonary vascular remodeling(e.g., vascular wall thickening) and elevated pulmonary arterypressures, without evidence for altered systemic BP (13, 32).Both IH and SH can enhance the hypoxic ventilatory response;however, SH appears to induce higher early ventilatory in-crease and more rapid ventilatory decline than IH (27).

At the molecular level, IH and SH can both induce a varietyof responses in animals, which is believed to be modulatedmainly by hypoxia-inducible factor (HIF)-1. HIF-1 can inducehypoxic responses to increase O2 delivery and energy produc-tion, which is critical for animals to survive in hypoxia (1, 31,40). Similar responses have also been implicated in pathophys-iological mechanisms of vascular remodeling and pulmonaryhypertension observed in patients suffering from SH (35).Despite the important role of HIF-1 in hypoxia, the hypoxiceffects elicited by HIF-1 alone cannot fully explain physiolog-ical abnormalities occurring in patients suffering from IH,suggesting that other regulators may play key roles in IH. Amajority of studies on hypoxic effects in animals have exam-ined the response of the brain, carotid body, and cardiovascularsystem to IH or SH.

* W. Wu and N. B. Dave contributed equally to this manuscript.Address for reprint requests and other correspondence: W. Wu, Dorothy P.

and Richard P. Simmons Center for Interstitial Lung Disease, Div. of Pulmo-nary, Allergy, and Critical Care Medicine, School of Medicine, Univ. ofPittsburgh, Pittsburgh, PA 15213 (e-mail: [email protected]); N. Kaminski,Dorothy P. and Richard P. Simmons Center for Interstitial Lung Disease, Div.of Pulmonary, Allergy, and Critical Care Medicine, School of Medicine, Univ.of Pittsburgh, Pittsburgh, PA 15213 (e-mail: [email protected]).

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

Physiol Genomics 36: 24–34, 2008.First published September 30, 2008; doi:10.1152/physiolgenomics.00258.2007.

1094-8341/08 $8.00 Copyright © 2008 the American Physiological Society24

by 10.220.33.1 on August 10, 2017

http://physiolgenomics.physiology.org/

Dow

nloaded from


In this study, we employed gene expression profiling toidentify temporal expression patterns of genes in rat lungs inresponse to either IH or SH. Systems biology approaches wereapplied to globally characterize distinct molecular programsand potential regulatory networks that play important roles inIH and SH.

MATERIALS AND METHODS

Experimental Procedures

Adult inbred Sprague-Dawley male rats (weights 250–275 g) wereplaced in oxygen- and carbon dioxide-controlled chambers. Rats weresubjected to either IH or SH exposure as previously described for1–30 days of duration (11, 27). Briefly, rats subjected to the IHtreatment were exposed to 90-s cycles of 10% inspired oxygenfollowed by 90-s of 21% inspired oxygen for 12 h of simulateddaylight, when rats sleep; for the remaining 12 h during the night, therats were exposed to 21% inspired oxygen (11). Rats subjected to theSH treatment were exposed to 10% inspired oxygen at all times.Carbon dioxide concentrations were monitored and maintained at�0.01% by altering chamber ventilation (27). All experimental pro-tocols were approved by the Institutional Animal Use and CareCommittee at the University of Louisville and were in accordancewith National Institutes of Health requirements for the care and use oflaboratory animals.

Measurement of Systemic Blood Pressure

Animals (n � 6) were placed into a Plexiglas cylindrical restrainerwith a size-appropriate cuff around the tail. Automated occlusionplethysmography (tail cuff method) was used, which involved ananalog-to-digital signal converter and a digital acquisition and anal-ysis system (Kent Scientific, Torrington, CT). During a 10-minaccommodation period, body temperature was raised to 38–39°C witha heating pad, so as to increase blood flow in the tail. Heating wasdiscontinued once the BP signal was detectable. At least 8–10 BPreadings were obtained from each rat at each time point, and allmeasurements were conducted at the same time of day, namely, 10:00AM; these values were then averaged to yield the mean BP for thattime point for each rat. Systolic and diastolic pressure were measuredfrom six rats before initiation of hypoxic exposures, and then at thedesignated time points for up to 30 days. The time course for BPchanges throughout the experimental period was plotted for eachexperimental group.

Telemetric Measurement of Right Ventricular Pressure

RV blood pressure (RVP) in rats (n � 6) was measured with animplanted radiotelemetry system (Dataquest A.R.T. 2.1; Data Sci-ences). The transmitter (model TA11PA) was connected to a fluid-filled sensing catheter and transferred the signals to a remote receiver(model RPC-1) and a data-exchange matrix connected to a computer.Under intraperitoneal ketamine and xylazine anesthesia, the sensingcatheter was inserted into the jugular vein and forwarded to the RV.The waveform was displayed on the computer and used to ensurecorrect positioning of the catheter. Animals were allowed to recoverand were housed individually in standard rat cages. RVPs wererecorded at 1-h intervals from the time of implantation and arepresented as the mean RVPs averaged over the 24-h period.

Tissue Acquisition and RNA Extraction

Rats were exposed to either IH or SH and were killed after 1, 3, 7,14, or 30 days. Rats exposed to room air (RA) served as controls. Ratswere anesthetized with a pentobarbital overdose (100 mg/kg ip) andeuthanized by decapitation. Whole lungs were then rapidly removed,snap frozen over liquid nitrogen, and stored at �70°C until needed forfurther use. Total RNA was extracted as previously described (24) and

used for microarray experiments and quantitative real-time polymer-ase chain reaction (qRT-PCR).

Microarray Protocol

Microarray experiments were performed with a protocol recom-mended by the manufacturer of the arrays and as previously describedby us (43). Briefly, extracted RNA from each rat was used as atemplate to generate cRNA, which was then labeled with a fluorescentdye and hybridized to a CodeLink UniSet Rat I Bioarray presynthe-sized with oligonucleotide probes (Applied Microarrays). Microarrayswere washed and scanned as recommended.

Quantitative Real-Time PCR

The detailed protocol for qRT-PCR can be found in Ref. 24, withmodification. Briefly, cDNA was synthesized from 2 �g of total RNA.This was performed with SuperScript First-Strand Synthesis System(Invitrogen, 11904-018). The qRT-PCR was performed in a 384-wellformat, under standard cycling conditions, with reagents, primers, andprobes for TaqMan Gene Expression Assays (Applied Biosystems).Finally, the ABI Prism 7900HT instrument and Sequence DetectionSoftware v2.2 (SDS 2.2) (Applied Biosystems) were used for PCRquantification and measurement of the PCR amplification product.Gene-specific primers and probes used were as follows. Estrogenreceptor 1 (Esr1): Rn00562166_m1; Potassium inwardly rectifyingchannel, subfamily J, member 6 (Kcnj6): Rn00755103_m1; Puri-nergic receptor P2X, ligand-gated ion channel, 2 (P2rx2):Rn00586491_m1; Mitogen-activated protein kinase 1 (Mapk1):Rn00671828_m1; Amphiregulin (Areg): Rn00567471_m1; L-typevoltage-dependent calcium channel, �1C subunit (Cacna1c):Rn00709288_m1; Nuclear receptor subfamily 0, group B, member1 (Nr0b1): Rn00584062_m1; Cytochrome P-450, family 26, sub-family b, polypeptide 1 (Cyp26b1): Rn00710377_m1; CytochromeP-450, family 2, subfamily b, polypeptide 15 (Cyp2b15):Rn00755182_g1; ATPase, Na�/K� transporting, �2 polypep-tide (Atp1a2): Rn00688124_gH; Bone morphogenic protein 2(Bmp2): Rn00567818_m1; MAD homolog 1 (Drosophila) (Smad1):Rn00565555_m1; Endothelin 1 (Edn1): Rn00561129_m1; Matrixmetallopeptidase 9 (Mmp9): Rn01423075_g1; CD8b molecule(Cd8b): Rn580581_m1; Bone morphogenic protein receptor, type II(serine/threonine kinase) (Bmpr2): Rn01437210_m1; Glucuronidase,�: Rn00566655_m1.

Microarray Data Preprocessing

In our initial experimental design, there were three biologicalreplicates for each time point (i.e., microarray data were obtainedfrom RNAs extracted from three different rats). However, because ofquality issues, some microarrays were eliminated before data prepro-cessing, and therefore there are fewer than three biological replicatesfor some time points (see Supplemental Table E1 for details).1 Rawmicroarray data were preprocessed with a protocol previously de-scribed by us (43) and also available in the Supplemental Data. Theprocessed microarray data was log2-transformed and normalized witha statistical method, CyclicLoess, to minimize unwanted noise in thedata (43). We previously showed (43) that CyclicLoess is one of themost effective normalization methods for reducing intensity-depen-dent dye effects in CodeLink Bioarray data. Finally, the log2-trans-formed, normalized data were used as input to the downstreamcomputational analyses. The complete microarray data set is availableat the Gene Expression Omnibus (GEO) database (http://www.ncbi.nlm.nih.gov/geo; GSE8705).

1 The online version of this article contains supplemental material.

25GENE EXPRESSION ANALYSIS IN HYPOXIA

Physiol Genomics • VOL 36 • www.physiolgenomics.org

by 10.220.33.1 on August 10, 2017


Dow

nloaded from


Computational Analyses

Local regression and spectral analysis. We have developed aflexible yet rigorous nonparametric statistical procedure, local regres-sion and spectral analysis (LRSA), which estimates genes differen-tially expressed at different time points with respect to the control(RA, day 0) with a nonparametric local regression smoothing modeland identifies temporal expression patterns of these genes with aspectral clustering algorithm. The LRSA procedure is described indetail in Ref. 44. Briefly, LRSA is a two-step procedure. In the firststep, temporal expression data for each transcript in each microarraydata set are first fitted with a local polynomial quadratic (degree � 2)model within a smoothing parameter or bandwidth hi, where hi isselected optimally with the generalized cross-validation (GCV) score(36). We chose to use the degree of 2, instead of 1, for the localpolynomial model since our results showed that the former choice ledto more powerful results and yielded lower false discovery rates(FDRs) than the latter choice (data not shown). To determine whethera gene is differentially expressed over time with respect to the controltime point, we calculated the simultaneous 95% confidence intervalsfor the fitted (or expected) intensity values at time point k with amethod due to Faraway and Sun (7); this method guarantees that theseconfidence intervals are simultaneously valid for all the time points inthe data, and it also allows the estimate of heteroscedasticity in thedata (7, 36, 41). Since there was only one biological replicate at sometime points in both IH and SH data, we used a modified proceduredescribed in detail in the Supplemental Data to estimate the simulta-neous 95% confidence intervals at these time points. In addition, weused external control probes on the CodeLink arrays as a surrogate toestimate the FDR associated with LRSA, and thus to guide thedetection of differentially expressed genes; we call the surrogatemetric “FDR for external control probes” (FDREC). In the Supple-mental Data, we compare P values obtained at some test time point Krelative to the control time point for both external control probes andregular transcripts on the arrays; it can be seen that P values for thesedifferent types of probes follow similar distributions. In the secondstep, LRSA clusters the differentially expressed genes resulting fromthe first step with a spectral clustering algorithm. A web server foridentifying temporal differentially expressed genes with LRSA canbe accessed at http://www.pitt.edu/�wew16/HypoxiaNetwork/index.html.In addition, our program written in R and Matlab for the LRSAprocedure as well as the results yielded by computational analyses inthis work are also available from the same web site.

To identify differentially expressed genes during either IH or SH,we applied LRSA without multiple testing control to the two hypoxiadata sets, respectively. We have shown that when applied to the IHdata set LRSA without multiple testing control is much more powerfuland yet yields the same FDREC as LRSA with multiple testing control(see our unpublished work in Ref. 44 and additional details in theSupplemental Data). We determined a gene as differentially expressedin the IH data set only if its expression at time point k relative to thecontrol time point (day 0) satisfied: 1) P � 0.05 and 2) fold change �2 (on the original data scale). We showed in Ref. 44 that when thesecriteria were used we detected differentially expressed genes from theIH data set with high power and FDREC � 0. To identify differentiallyexpressed genes from the SH data set, we employed the same criteriaas we did for the IH data set on array data obtained from rats exposedto SH for 1, 7, and 30 days; we used a slightly higher fold change, foldchange � 2.5, to determine differentially expressed genes from thedata obtained from rats with 3-day SH exposure since the data have ahigher variability. Because the quality of the arrays obtained from ratsexposed to SH for 14 days was not reliable, we did not include themin our analysis. Differentially expressed genes identified from the SHdata set also satisfied FDREC � 0.

Gene Ontology statistical analysis. To identify functional groups ofgenes enriched among differentially expressed genes in rat lungs inresponse to either IH or SH, we performed Gene Ontology (GO)

analysis with the GOstat program (3). The GOstat program finds theenriched functional groups using Fisher’s exact tests with the aid ofthe GO annotation (3). The list of the genes on the CodeLink UniSetRat I Bioarrays was used as the reference gene list for GOstat. Todetermine whether a functional group is significantly enriched, weemployed the following criteria: 1) the unadjusted P value � 0.005,which amounted to P � 0.1 with the FDR-controlling procedure ofBenjamini and Hochberg, and 2) the number (ni) of the hypoxia-induced genes in GO group i satisfies ni 10, or the percentage(pi%) of the hypoxia-induced genes assigned to GO group i (i.e.,pi% � ni/Ni, where Ni is the total number of the genes in GO groupi) satisfies pi 20.

Identifying potential ESR1 target genes in rats. To determinepotential ESR1 target genes among differentially expressed genes inrat lungs during either IH or SH, we identified all such genes in the ratby using the lists of the potential human and mouse target genes ofESR1 published by Bourdeau et al. (4) and also available at http://www.mapageweb.umontreal.ca/maders/eredatabase. Specifically, wefirst mapped all human and mouse ESR1 target genes to their rathomolog genes; then we identified a gene as a potential ESR1 targetgene in the rat only if it appears on both lists of the rat homologs ofhuman and mouse ESR1 target genes.

Network analysis. To identify regulatory networks activated byeither IH or SH, genes differentially expressed during each type ofhypoxia were analyzed with Ingenuity Pathways Analysis (IPA)software (Ingenuity Systems, http://www.ingenuity.com). We havepreviously employed the IPA network analysis to detect informativeregulatory networks in asthma (21). The final IH- or SH-activatednetwork was formed by merging 14 subnetworks identified by IPAthat were assigned the highest scores; the network consisted of �500genes, the biggest network allowed to be identified by IPA. Additionalinformation on the procedure of the network analysis can be found inthe Supplemental Data.

Disease relevance analysis of a network. This analysis identifiedthe physiological malfunctions significantly associated with the de-tected network and was performed with IPA. Fisher’s exact test wasused to calculate the P value determining the probability that eachmalfunction associated with the network is due to chance alone. Wedetermine an association as significant if 1) P � 0.05 and 2) there aremore than a cutoff number (20 or 30 was used in this work) of genesin the network involved in the association.

RESULTS

Physiological Responses of Rats to IH or SH

IH and SH elicited different physiological responses as far assystemic BP and RVP. Systemic BP was increased slightly inIH after 30 days of exposure (P � 0.05, relative to RA controlrats) but remained unaltered in SH (Fig. 1A). Mean RVPs wereincreased in rats exposed to either IH or SH (P � 0.05 relativeto RA control rats), but the increase was significantly greater inSH than in IH (Fig. 1B).

Comparison of Expression Patterns of DifferentiallyExpressed Genes in IH and SH

Using the nonparametric LRSA procedure developed by usto analyze time course expression data, we detected two sets ofgenes differentially expressed over time in rat lungs induced byIH and SH, respectively (Supplemental Tables E2 and E5;the gene lists can also be found at http://www.pitt.edu/�wew16/HypoxiaNetwork). These two sets of genes exhib-ited different expression patterns (Fig. 2, A and C) andparticipated in different functions as well as forming differ-ent networks as suggested by multiple computational anal-

26 GENE EXPRESSION ANALYSIS IN HYPOXIA


by 10.220.33.1 on August 10, 2017


Dow

nloaded from


yses. In the following, we present gene expression patternsand functional and network analyses of the differentiallyexpressed genes activated during IH and SH in detail.

Temporal Gene Expression Patterns in IH

We identified 1,525 genes differentially expressed in ratlungs during IH with LRSA (Supplemental Table E2). These

genes exhibited seven temporal expression patterns (Fig.2A), which appeared to be homeostatic: expression of genespeaked at earlier time points and converged to the controllevel at later time points. An analysis of the numbers of thegenes induced over time during IH revealed a bell-shapeddistribution (Fig. 2B), which indicated that a vast majorityof the genes were induced on day 7, while only a few geneswere induced at earlier (days 1 and 3) or later (days 14 and30) time points.

The upregulated genes in response to IH reside in clusters1–4 (Fig. 2A and Supplemental Fig. E1A) and appeared to beinduced in a time-dependent manner: genes in clusters 1 and 2exhibited early-to-middle patterns, cluster 3 an early acutepattern, and cluster 4 a middle-to-late pattern. The downregu-lated genes in IH reside in clusters 5–7 (Fig. 2A and Supple-mental Fig. E1A); unlike the upregulated genes, the downregu-lated genes appeared to have more uniform patterns.

Functional Theme Analysis of Genes Enriched in IH

GO enrichment analysis revealed a large number of func-tional groups of genes significantly enriched in the upregulatedclusters (P � 0.1 with FDR, Fig. 3A and Supplemental Fig.E2), which contained genes involved in system development,neurological process, ion transport and homeostasis, G protein-coupled receptor signaling, behavior, oxidoreductase, and ste-roid hormone receptor activity; in contrast, there were fewfunctional groups enriched in the downregulated clusters,which included genes participating in nucleic acid binding andribonucleoprotein complex (Fig. 3B).

The majority of the enriched upregulated genes resided inthe two clusters—clusters 1 and 2 (Fig. 3A). Specifically, genesin cluster 1 displayed a cyclically regulated pattern that peakedon days 1 (weakly) and 7 (strongly) (Fig. 2A and SupplementalFig. E1A). Of all the clusters induced by IH, this clustercontained the largest number of significantly enriched func-tional groups of genes, which overlapped mostly with thoseenriched among all upregulated genes induced by IH. Interest-ingly, six of the eight genes encoding the steroid hormonereceptors (including ESR1 and ESR2) induced by IH werecoexpressed in this cluster and were significantly overrepre-sented (P � 0.02 with FDR).

Genes in cluster 2 exhibited a typical bell-shaped expressionpattern with a single peak on day 7 (Fig. 2A and SupplementalFig. E1A); this cluster was significantly enriched with thoseparticipating in K� transport and neuronal activities. Since ionchannels (especially K� channels) are important for neuronal

Fig. 1. Systemic diastolic (DBP) and systolic (SBP) bloodpressures and mean right ventricular blood pressure (RVP) inrats exposed to intermittent (IH) and sustained (SH) hypoxia.SBP and DBP (A) and RVP (B) were all measured in ratsexposed to IH or SH for 0 [room air (RA)], 1, 3, 7, 14, and 30days. Data are means SE (n � 6). Significant differences:*P � 0.05 compared between rats exposed to either IH or SHand RA control rats; �P � 0.05 compared between ratsexposed to IH and SH.

Fig. 2. Temporal expression profiles of genes differentially expressed in ratlungs in response to either IH or SH. A: heat map of the expression profile ofdifferentially expressed genes in IH. d, Day; cl, cluster. B: histogram of genesdifferentially expressed during IH on each measured day (D). C: heat map ofthe expression profile of differentially expressed genes in SH. D: histogram ofgenes differentially expressed during SH on each measured day.



by 10.220.33.1 on August 10, 2017


Dow

nloaded from


signaling, genes in this cluster were likely to mediate neuronalresponses to IH in rat lungs.

ESR1-Regulated Genes Are Enriched Among DifferentiallyExpressed Genes in IH

Since genes encoding steroid hormone receptors were over-represented among the IH-induced genes, we postulated thatESR1 plays regulatory roles in hypoxic responses to IH. To testthis possibility, we examined the differentially expressed genesinduced by IH. To our surprise, we found 248 potential ESR1target genes (Supplemental Table E3) (4); most of these re-sided in clusters 1, 2, and 6, which, interestingly, bore similarexpression patterns (i.e., clusters 2 and 6 displayed almost themirror image of one another). GO analysis revealed that theESR1 target genes in the upregulated clusters primarily partic-ipated in the following activities: response to stress, ion trans-port, neurological system process, G protein-coupled receptorsignaling, oxidoreductase activity, and steroid hormone recep-tor activity (Fig. 4A and Supplemental Fig. E3). Significantly,most of these functional groups overlapped with those foundamong the upregulated genes in IH, suggesting that ESR1plays key regulatory roles during IH. The ESR1 target genes inthe downregulated clusters were mainly involved in ubiquitin-dependent protein catabolic process, cytokine activity, andribonucleoprotein complex (Fig. 4B).

Verification of Genes Differentially Expressed Over Timein IH

To verify our findings from microarray analysis, we per-formed both literature survey and qRT-PCR. Many of the

differentially expressed genes we identified in the presentexperiments have been reported previously to be inducible byhypoxia (9, 30, 42), which included the upregulated Nfkb1,Ho-2, and Htr5a and the downregulated Kcnma1. These geneswere therefore not verified experimentally.

We validated differential expression of Esr1 and 11 othergenes including 4 known ESR1 target genes: P2rx2, Areg,Bmp2, and Smad1 (4, 5, 23) as well as Kcnj6, Mapk1, Cacna1c,Nr0b1, Cyp26b1, Cyp2b15, and Atp1a2 with qRT-PCR. Con-sistent with our microarray results (Fig. 5A, left), we observeda cyclic pattern of Esr1 in rat lungs in response to IH, but notto SH, with qRT-PCR (Fig. 5A, right). The other 11 selectedgenes were also confirmed to be significantly differentiallyexpressed in IH, and their expression patterns measured byqRT-PCR agreed well with those revealed by LRSA frommicroarray data (Fig. 5, B–L).

Regulatory Networks in IH

Using network analysis, we detected a regulatory networkamong differentially expressed genes in IH (Fig. 6A andSupplemental Fig. E4). The network contained �500 genes—the biggest network allowed to be identified by the IPAsoftware. This network contained a few hub nodes, each ofwhich formed direct connections with a large number of othernodes in the network (Fig. 6A and Supplemental Fig. E4A). Itcan be seen that ESR1 was one of the hub nodes in the network(shown in detail in Fig. 6B), which connected closely withother hub nodes including ARNT (also known as HIF-1�),VEGF (encoding vascular endothelial growth factor), NF-�B,JUN, MAPK1, as well as stress proteins including HSP70; it

Fig. 3. Functional groups of genes significantly enriched among differentially expressed genes in rat lungs during either IH or SH. Functional groups of genes enrichedamong the clusters upregulated in IH (A), downregulated in IH (B), upregulated in SH (C), and downregulated in SH (D) are shown. Within the parentheses beloweach Gene Ontology (GO) functional group name, ni, pi%, cl. c represent the number (ni) of the hypoxia-induced genes in GO group i, the percentage (pi%) ofthe hypoxia-induced genes assigned to the GO group i (i.e., pi% � ni/Ni, where Ni is the total number of the genes in GO group i), and the cluster(s) (cl. c)significantly enriched with the genes in the same GO group i. GO groups in this figure satisfied 1) P value � 0.1 with the false discovery rate (FDR) controland 2) ni 10 or pi 20. , Functional groups that satisfied P � 0.01 without the FDR correction but did not satisfy P � 0.1 with the FDR correction.



by 10.220.33.1 on August 10, 2017


Dow

nloaded from


also downregulated the genes involved in transcriptional andsplicing machinery such as Med20 and SF3A1. In addition,ESR1 interacted closely with the three other steroid hormonereceptors that were also differentially expressed in IH [ESR2,androgen receptor (AR), and nuclear receptor NR0B1], sug-gesting that there was cross talk between these proteins in thenetwork.

Interestingly, genes in this network have been previouslyassociated significantly with cancer (P � 1E-19), hematologicdisorder (P � 6E-18), cardiovascular disorder (P � 3E-12),gastrointestinal disease (P � 4E-10), as well as inflammatory,skeletal and muscular, connective tissue, genetic, neurological,and behavior disorders (Supplemental Fig. E5A and Supple-mental Table E4). Furthermore, our analysis showed that ESR1has been directly associated with most of these disorders(Supplemental Fig. E5A and Supplemental Table E4).

Temporal Gene Expression Patterns in SH

We also analyzed temporal expression data obtained fromrats exposed to SH and detected 936 differentially expressedgenes induced by SH (Supplemental Table E5). These genesdisplayed six temporal expression patterns (Fig. 2C), and many

of them were differentially expressed in an acute manner (Fig.2C and Supplemental Fig. E1B). Examination of the genesinduced over time in SH showed that a majority of the geneswere induced on day 1, while only a few genes were inducedat later time points (days 14 and 30) (Fig. 2D).

GO analysis revealed that the upregulated clusters (clusters1–3, see Supplemental Fig. E1 for details) in SH were signif-icantly enriched with the functional group of genes involved inimmune response; the same group of genes was also coex-pressed and enriched in cluster 2 (Fig. 3C), while the down-regulated clusters (clusters 5 and 6, Supplemental Fig. E1)were overrepresented by genes involved in RNA binding andprocessing (Fig. 3D). Our results also showed that cluster 1was significantly overrepresented by genes involved in organ(particularly blood vessel) morphogenesis; since pulmonaryvascular remodeling is a known effect of SH in rats, genes inthis cluster might play roles in this process.

Despite the finding that gene expression patterns in re-sponse to IH and SH were distinctly different in rat lungs,we found that a considerable number of the genes weredifferentially expressed in both types of hypoxia (Supple-mental Fig. E7A). It appeared that the genes upregulated inboth IH and SH were significantly overrepresented by thoseparticipating in system development, anatomic structuremorphogenesis, and cell cycle (Supplemental Fig. E7B),while the downregulated genes were overrepresented bythose involved in pyrophosphatase and ATPase activities(Supplemental Fig. E7C).

Verification of Genes Differentially Expressed in Rat LungsDuring SH

Many differentially expressed genes we identified in SHhave been previously reported to be inducible in hypoxia,which included Edn1 and Bmpr2, whose differential expres-sion has been associated with pulmonary hypertension devel-oped in animals exposed to SH (35, 37). It is unknown,however, how these genes were differentially expressed overtime in response to hypoxia exposure.

We validated the temporal expression patterns of the eightgenes with qRT-PCR. It can be seen that Edn1 exhibited acyclic temporal pattern that peaked on days 1 (strongly) and7 (weakly) in microarray data (Fig. 7A, left). We confirmedthis expression pattern of Edn1 in rats exposed to SH withqRT-PCR (Fig. 7A, right). The significant downregulationof Bmpr2 over time was also confirmed in rats exposed toSH with qRT-PCR (Fig. 7G). However, unexpectedly, thedownregulation of Bmpr2 was also observed in rats exposedto IH with qRT-PCR (Fig. 7G, right), despite the fact thatwe did not detect this in the microarray data. The expressionpatterns of the six other genes, Bmp2, Mmp9, Cd8b,Cyp26b1, Cacna1c, and Smad1, were also confirmed byqRT-PCR (Fig. 7, B–H).

Regulatory Networks in SH

Our network analysis revealed a regulatory network acti-vated by SH (Fig. 8 and Supplemental Fig. E8). This networkcontained the hub nodes including VEGF, NF-�B, and TGF-�,all of which have been implicated in vascular remodeling andpulmonary hypertension during hypoxia (39), as well as EDN1,

Fig. 4. Functional groups of genes significantly enriched among potentialestrogen receptor 1 (ESR1) target genes in rat lungs during IH. The functionalgroups of genes (shown in black) enriched among potential ESR1 target genesin the upregulated clusters (A) and the downregulated clusters (B) (seeSupplemental Fig. E1 for details) are shown. The functional groups overrep-resented among all the upregulated genes (shown in Fig. 3A) but not amongESR1 target genes are shown in gray. , Functional groups that satisfied P �0.01 without the FDR correction but did not satisfy P � 0.1 with the FDRcorrection.



by 10.220.33.1 on August 10, 2017


Dow

nloaded from


IL1B, MMP9, and PI3K. Unlike the IH-induced network,ESR1 was not involved in this network.

Finally, disease relevance analysis revealed that this net-work is significantly associated with physiological malfunc-tions including cancer (P � 2E-18), hematologic disease(P � 3E-16), cardiovascular disorder (P � 2E-8), colorectalcancer (P � 2E-12), and skeletal and muscular disorder(P � 2E-7) (Supplemental Fig. E5B, Supplemental Table

E6). These abnormalities were also associated with theIH-activated network; however, our analysis indicated that amajority (more than two-thirds) of the genes associated witheach malfunction were unique to either IH or SH, suggestingthat different molecular mechanisms were underlying thesame diseases associated with these two different types ofhypoxia (Supplemental Table E6). In addition, the SH-activated network is also significantly enriched with genes

Fig. 5. Verification of genes differentially ex-pressed during IH. A, left: temporal expressionpattern of the Esr1 gene (residing in cluster 1)fitted by local regression and spectral analysis(LRSA). Each black dot in the plot representsthe expression level of the gene (shown ony-axis, relative to expression of Esr1 in RAcontrol rats) in each individual rat exposed toIH for the designated days (shown on x-axis)in the normalized microarray data; solid curveshows the smoothing curve fitted to the mi-croarray data by LRSA and dotted curvesshow the simultaneous 95% confidence inter-vals estimated for the fitted data. Right: ex-pression level of Esr1 (relative to expressionof Esr1 in RA control rats) measured byquantitative real-time PCR (qRT-PCR). Datafor rats in response to RA or to IH or SH for 1,3, 7, 14, or 30 days are shown. Values aremeans SE. Significant differences: *P � 0.05compared with RA control rats. B–L: other ver-ified IH-induced genes are Kcnj6, P2rx2,Mapk1, Areg, Cacna1c, Nr0b1, Cyp26b1,Cyp2b15, Atp1a2, Bmp2, and Smad1.



by 10.220.33.1 on August 10, 2017


Dow

nloaded from


associated with rheumatic disease (particularly arthritis),autoimmune disease, and endocrine system disorders (par-ticularly diabetes) (Supplemental Fig. E5B and Supplemen-tal Table E6), which were specific to SH, but not to IH.

DISCUSSION

We employed a systems biology approach to study temporaleffects of IH and SH by examining time-dependent geneexpression profiles from whole lung tissues of rats exposed toeither IH or SH. Our results demonstrate that IH and SHinduced distinct sets of genes with different temporal expres-sion patterns. IH induced genes mostly involved in ion trans-port and homeostasis, neurological processes, system develop-ment, and steroid hormone receptor activity, while SH inducedgenes principally participating in immune responses. The IH-activated network suggested a role for cross talk between ESR1and other key proteins in hypoxic responses (e.g., ARNT and

MAPK1) during IH, whereas the SH-activated network wasindicative of vascular remodeling and pulmonary hypertension.We confirmed the temporal changes in the expression of 12genes (including Esr1 and 4 other ESR1 target genes) in IH and8 genes in SH with qRT-PCR. Our data suggest that ESR1 maybe a regulator of lung hypoxic response to IH.

Several lines of evidence have emerged from our analysessuggesting that ESR1 is a potential key regulator that modu-lates the expression of the genes responsible for major hypoxicresponses to IH. First, we verified the differential expression ofthe four ESR1 target genes during IH, P2rx2, Bmp2, Areg, andSmad1, with qRT-PCR. These genes participate in diverseactivities in various tissues where ESR1 plays important roles:P2X2 is involved in neuronal responses to hypoxia in carotidbody; BMP2 is implicated in bone development (6); amphi-regulin is a key mediator of ESR1 function in mammary glanddevelopment (5); and the direct contact of Smad1 with ESR1 isimplicated in the tumorigenesis of pituitary prolactinoma (23).Additionally, we identified 248 ESR1 target genes amonggenes induced by IH, and they were enriched with the samefunctional groups enriched in all upregulated genes during IH(Fig. 4A and Supplemental Fig. E3). Significantly, most ofthese activities have been well established to mediate importantdownstream effects of ESR1 signaling and action (12, 14, 17,38). Furthermore, our network analysis illustrated that ESR1interacts closely with numerous genes in the IH-activatednetwork (Fig. 6B). Finally, ESR1 interacts directly with AKT,MAPK1, and PI3K in the IH-activated network; these proteinkinases are known to modulate the activity of ESR1, and thusallow synergistic regulation of gene expression by these sig-naling pathways (18). All of the above-described evidencesuggests that ESR1 plays major regulatory roles in rat lung inresponse to IH. This finding, however, is both interesting andunexpected, since all the rats used in this work were male.Indeed, there is a recent awareness that ESRs play importantroles in biological processes in male animals, including vas-cular homeostasis and protection (17) and bone metabolism(34); there has been no report that they play key regulatoryroles in the context of diverse hypoxic adaptive or injury-relayed processes in male animals.

Our analyses also suggest several additional important find-ings. First, in the detected network induced by IH, ESR1interacted closely with other hub nodes previously known toplay key roles in hypoxia and/or ESR signaling, such asARNT, VEGF, MAPK1, and EDN1, suggesting that there isinterplay between ESR1 and these important proteins duringlung hypoxic responses to IH. Additionally, our disease rele-vance analysis showed that ESR1 is directly involved in manydisorders including cancer, hematologic, cardiovascular, andinflammatory disorders, and neurological, behavior, and met-abolic disorders (Supplemental Fig. E5A). Most notably, theseare also the disorders associated with patients with sleep-disordered breathing (33). These results are both interestingand potentially important—they suggest that ESR1 may notonly mediate crucial hypoxic responses to IH in the animal butalso be directly involved in pathophysiological mechanismsunderlying abnormalities associated with patients with sleep-disordered breathing.

In this context it is worth noting the potential effect of ESR1 onregulating neuronal responses in rat lungs during IH. Our datashowed that genes participating in neuronal processes were over-

Fig. 6. Gene regulatory network detected in rat lungs during IH. A: graphicalrepresentation of the molecular relationships between nodes (i.e., genes and/orgene products) in the network identified in rat lungs exposed to IH for 7 days.An edge (line) represents the biological relationship between 2 nodes. Theintensity of the node color indicates the magnitude of the upregulation (red) ordownregulation (green) of the gene at a specific time point relative to itsexpression at the control time point (day 0). The direct connections of ESR1with other nodes are shown by blue lines. Additional views of this network canbe seen in Supplemental Fig. E4. B: interaction of ESR1 with other nodes inthe identified network shown in A.



by 10.220.33.1 on August 10, 2017


Dow

nloaded from


represented among the potential target genes of ESR1 in hypoxia.Some of these genes are known to mediate crucial hypoxicresponses in the brain and the carotid body. For example, increas-ing evidence has shown that the purinergic receptor P2X2 is oneof the earliest proteins responding to hypoxia in mammals formediating adaptive changes in breathing in the carotid body. Wehave identified and confirmed upregulation of the P2rx2 gene(encoding P2X2) in rats exposed to IH in microarray data andwith qRT-PCR and thus provide initial evidence that the expres-sion of P2rx2 is induced in the rat lung during IH. Besides, wehave identified upregulation of the genes encoding the glutamatereceptor (Grik5) and the dopamine receptor D4 (Drd4)—both ofwhich are receptors for excitatory neurotransmitters and critical

proteins involved also in adaptive changes in breathing in thecentral nervous system (CNS) (8, 22). In addition, genes encodingreceptors for important neurotransmitters known to be active inthe brain have also been identified as upregulated in IH andpotentially regulated by ESR1, which included Gabrr2 (encodingan inhibitory GABA receptor A), Chrnb2 (encoding a nicotiniccholinergic receptor), and Adra1d (encoding an adrenergic �-re-ceptor). In particular, upregulation of the nicotinic cholinergic andadrenergic receptors has been associated with increased heartbeatand systemic hypertension: both symptoms have been reported inpatients with sleep-disordered breathing owing to IH (33). Theseresults are also consistent with previous studies indicating thatESR1 mediates neuroprotection by modulating expression of

Fig. 7. Verification of genes differentially expressed during SH. A, left: temporal expression pattern of the Edn1 gene (residing in cluster 1) fitted by LRSA.Details of the plot can be found in Fig. 4 legend. Right: expression level of Edn1 (relative to its expression in RA control rats) measured by qRT-PCR. Datafor rats in response to RA or to IH or SH for 1, 3, 7, 14, or 30 days are shown. B–H: verification of other genes differentially expressed in SH: Bmp2, Mmp9,Cd8b, Cyp26b1, Cacna1c, Bmpr2, and Smad1. Data are means SE. Significant differences: *P � 0.05 compared with RA control rats.



by 10.220.33.1 on August 10, 2017


Dow

nloaded from


neuronal genes (18) but so far have not been demonstrated in thelung.

An interesting feature revealed by our GO analysis of IH geneexpression patterns is that while a large number of functionalgroups were significantly overrepresented in upregulated clusters,only a few were overrepresented in downregulated clusters. Thisis consistent with previous observations suggesting that whenanimals are exposed to hypoxia specific responses essential tosurvival are elicited, while nonessential biological processes areshut off (1). Our data suggest that this is indeed the case. First, GOanalysis showed that the overrepresented groups in the upregu-lated genes are underrepresented in the downregulated genes,suggesting that essential functional groups are indeed mostlyupregulated during IH. Second, there were significantly fewer(and no additional) functional groups enriched among all differ-entially expressed genes during IH compared with those enrichedin the upregulated clusters, which demonstrates again the dilutingeffect of the downregulated clusters (i.e., the absence of theessential functional groups in these clusters) and also indicatesthat the up- and downregulated clusters contain genes involved indifferent biological processes.

The comparison of SH to IH yielded some meaningful differ-ences. Compared with IH, our results showed that 1) SH induceda different set of genes in rat lungs, most of which displayed acutepatterns that peaked on day 1—this may just reflect the dose effectof hypoxia during IH and SH, that is, the more severe exposure ofhypoxia during SH may lead to more acute responses in animals;2) the SH-induced genes were overrepresented by those involvedin immune response; 3) there was no obvious involvement ofESR1 in the SH-activated network; and 4) disease relevanceanalysis revealed that the SH-activated network is specifically,significantly associated with rheumatic disease and endocrinesystem disorders (particularly diabetes). Despite the finding thatIH and SH elicited distinct hypoxic responses in rat lungs, therewere a considerable number of common genes induced by bothtypes of hypoxia. Most of the common genes were downregulatedin IH and SH and were enriched with the functional groupsinvolved in pyrophosphatase and ATPase activities, suggestingthat reduction of energy consumption is a common hypoxicresponse in both types of hypoxia.

An interesting attribute of the gene expression patterns ob-served in our study is that many genes induced in IH or SH seemto display homeostatic patterns, that is, the expression of the genes

mostly returned to their baseline values after induction at the earlytime points. These results are in contrast to our physiological data,which showed that systemic BPs were increased continuously inrats exposed to IH and that RVPs were also elevated in rats duringeither IH or SH, albeit at different levels (Fig. 1). One possibleexplanation for this discrepancy is that the genes required forinducing the increase in systemic BP and RVP in rat lungs aredifferent from those required for maintaining them. Indeed, wecan see that in the IH- or SH-induced networks (SupplementalFigs. E4A and E8A) genes differentially expressed during hypoxia(i.e., nodes shown in red or green) can not only affect expressionof each other but also modulate expression of proteins whosemRNAs were not differentially expressed (i.e., nodes shown inwhite) during early hypoxic response. These proteins may in turnaffect the physiology of the animals long after the expression ofthe majority of genes goes back to the original level. A similarphenomenon has been observed in other physiological responsesduring hypoxia. For example, the activation of a serotonin recep-tor (5-HT2A) is required for the initiation but not the maintenanceof long-term facilitation elicited during IH, yet other proteins arerequired for maintaining long-term facilitation in the same ani-mals (2, 8, 19). While our study is unique in that we studiedtemporal gene expression patterns in response to IH or SH up to30 days, our results also suggest that to fully understand thelong-term complex regulation of the response to IH or SH,temporal gene expression patterns beyond 30 days should bestudied.

In this study we aimed to provide a systems-level view of thelung response to IH and SH. We identified global expressionpatterns that were distinct to these types of stimuli and validatedsome of the genes that represented these patterns and our keybiological insights. It is important to note that the lung is acomplex and highly dynamic organ, and thus it would be impor-tant to determine the cellular populations that express thesechanges in gene expression. While such an extensive localizationeffort was beyond the scope of this work, we believe that the wideavailability of our results will encourage scientists and investiga-tors to identify and explore these cellular response niches.

We have attempted here to identify molecular mechanisms under-lying hypoxic responses in the rat lung during IH or SH with asystems biology approach. While previous studies on hypoxia havemainly focused on the hypoxic effects on the CNS and cardiovascularsystem, our work provides the first comprehensive genomewideinvestigation of the hypoxic effects of IH and SH on the rat lung; itdemonstrates the wealth of the global transcriptional response of thelung to IH and SH and suggests that the lung may participate in thedownstream physiological effects of IH and SH. Our findings alsosuggest that ESR1 is a regulator of hypoxic responses in IH but notin SH and that genes critical for hypoxic responses in the brain andcardiovascular system may also play important and active roles in thelung. Our results provide clues and insights into novel molecularmechanisms underlying IH and SH responses, thereby warrantingfuture exploration of such novel pathways.

ACKNOWLEDGMENTS

We thank Dr. Thomas Richards and Lara Chensny from the Dorothy P. andRichard P. Simmons Center for Interstitial Lung Disease, Division of Pulmo-nary, Allergy and Critical Care Medicine, School of Medicine, University ofPittsburgh, and Emeka Ifedigboe from the Division of Pulmonary, Allergy andCritical Care Medicine, School of Medicine, University of Pittsburgh, for theirhelp in microarray experiments and in generating microarray data.

Fig. 8. Gene regulatory network detected in rat lungs during SH. Plot illus-trates the molecular relationships between genes and/or gene products in thenetwork identified in rat lungs in response to SH for 1 day. The details on thegraphical representation of the network can be found in Fig. 6 legend.Additional views of this network can be seen in Supplemental Fig. E8.



by 10.220.33.1 on August 10, 2017


Dow

nloaded from


Present addresses: N. B. Dave, Sleep and Breathing Disorders Ctr., Div. ofPulmonary and Critical Care Medicine, University of Texas SouthwesternMedical Center, Dallas, TX 75390; S. W. Ryter and A. M. K. Choi, Div. ofPulmonary, Allergy and Critical Care Medicine, Brigham and Women’sHospital, Harvard Medical School, Boston, MA 02115.

GRANTS

W. Wu’s work was funded by National Institutes of Health (NIH) Grant P50-HL-084932 and National Science Foundation Grant CCF-0523757. N. Kaminski’s workwas funded by NIH Grants HL-073745, HL-0793941, and HL-0894932. N. Kaminskiwas also supported by a generous donation from the Simmons Family. N. B. Dave’swork was funded by NIH Grant 1-F32-HL-78164-2. D. Gozal is supported by NIHGrants HL-65270, HL-69932, and SCOR 2P50-HL-60296-06, The Children’s Foun-dation Endowment for Sleep Research, and the Commonwealth of Kentucky Chal-lenge for Excellence Trust Fund.

DISCLOSURES

D. Gozal serves on the national speaker bureau for Merck Company and hasreceived honoraria for lectures in 2006 and 2007.

REFERENCES

1. Acker T, Acker H. Cellular oxygen sensing need in CNS function: physio-logical and pathological implications. J Exp Biol 207: 3171–3188, 2004.

2. Baker-Herman TL, Mitchell GS. Phrenic long-term facilitation requiresspinal serotonin receptor activation and protein synthesis. J Neurosci 22:6239–6246, 2002.

3. Beissbarth T, Speed TP. GOstat: find statistically overrepresented GeneOntologies within a group of genes. Bioinformatics 20: 1464–1465, 2004.

4. Bourdeau V, Deschenes J, Metivier R, Nagai Y, Nguyen D,Bretschneider N, Gannon F, White JH, Mader S. Genome-wide iden-tification of high-affinity estrogen response elements in human and mouse.Mol Endocrinol 18: 1411–1427, 2004.

5. Ciarloni L, Mallepell S, Brisken C. Amphiregulin is an essential medi-ator of estrogen receptor alpha function in mammary gland development.Proc Natl Acad Sci USA 104: 5455–5460, 2007.

6. Deroo BJ, Korach KS. Estrogen receptors and human disease. J ClinInvest 116: 561–570, 2006.

7. Faraway JJ, Sun J. Simultaneous confidence bands for linear regressionwith heteroscedastic errors. J Am Stat Assoc 90: 1094–1098, 1995.

8. Feldman JL, Mitchell GS, Nattie EE. Breathing: rhythmicity, plasticity,chemosensitivity. Annu Rev Neurosci 26: 239–266, 2003.

9. Giordano FJ. Oxygen, oxidative stress, hypoxia, and heart failure. J ClinInvest 115: 500–508, 2005.

10. Gozal D. Sleep-disordered breathing and school performance in children.Pediatrics 102: 616–620, 1998.

11. Gozal D, Daniel JM, Dohanich GP. Behavioral and anatomical corre-lates of chronic episodic hypoxia during sleep in the rat. J Neurosci 21:2442–2450, 2001.

12. Kelly MJ, Wagner EJ. Estrogen modulation of G-protein-coupled recep-tors. Trends Endocrinol Metab 10: 369–374, 1999.

13. Kourembanas S, Morita T, Liu Y, Christou H. Mechanisms by whichoxygen regulates gene expression and cell-cell interaction in the vascula-ture. Kidney Int 51: 438–443, 1997.

14. Le Mellay V, Grosse B, Lieberherr M. Phospholipase C beta andmembrane action of calcitriol and estradiol. J Biol Chem 272: 11902–11907, 1997.

15. Manukhina EB, Downey HF, Mallet RT. Role of nitric oxide incardiovascular adaptation to intermittent hypoxia. Exp Biol Med (May-wood) 231: 343–365, 2006.

16. McKay LC, Janczewski WA, Feldman JL. Sleep-disordered breathingafter targeted ablation of preBotzinger complex neurons. Nat Neurosci 8:1142–1144, 2005.

17. Meyer MR, Haas E, Barton M. Gender differences of cardiovasculardisease: new perspectives for estrogen receptor signaling. Hypertension47: 1019–1026, 2006.

18. Mhyre AJ, Dorsa DM. Estrogen activates rapid signaling in the brain:role of estrogen receptor alpha and estrogen receptor beta in neurons andglia. Neuroscience 138: 851–858, 2006.

19. Mitchell GS, Baker TL, Nanda SA, Fuller DD, Zabka AG, HodgemanBA, Bavis RW, Mack KJ, Olson EB Jr. Intermittent hypoxia andrespiratory plasticity. J Appl Physiol 90: 2466–2475, 2001.

20. Neubauer JA, Sunderram J. Oxygen-sensing neurons in the centralnervous system. J Appl Physiol 96: 367–374, 2004.

21. Novershtern N, Itzhaki Z, Manor O, Friedman N, Kaminski N. Afunctional and regulatory map of asthma. Am J Respir Cell Mol Biol 38:324–336, 2008.

22. Nurse CA. Neurotransmission and neuromodulation in the chemosensorycarotid body. Auton Neurosci 120: 1–9, 2005.

23. Paez-Pereda M, Giacomini D, Refojo D, Nagashima AC, Hopfner U,Grubler Y, Chervin A, Goldberg V, Goya R, Hentges ST, Low MJ,Holsboer F, Stalla GK, Arzt E. Involvement of bone morphogeneticprotein 4 (BMP-4) in pituitary prolactinoma pathogenesis through aSmad/estrogen receptor crosstalk. Proc Natl Acad Sci USA 100: 1034–1039, 2003.

24. Pardo A, Gibson K, Cisneros J, Richards TJ, Yang Y, Becerril C,Yousem S, Herrera I, Ruiz V, Selman M, Kaminski N. Up-regulationand profibrotic role of osteopontin in human idiopathic pulmonary fibrosis.PLoS Med 2: e251, 2005.

25. Polotsky VY, O’Donnell CP. Genomics of sleep-disordered breathing. ProcAm Thorac Soc 4: 121–126, 2007.

26. Prabhakar NR, Fields RD, Baker T, Fletcher EC. Intermittent hypoxia:cell to system. Am J Physiol Lung Cell Mol Physiol 281: L524–L528, 2001.

27. Reeves SR, Gozal E, Guo SZ, Sachleben LR Jr, Brittian KR, LiptonAJ, Gozal D. Effect of long-term intermittent and sustained hypoxia onhypoxic ventilatory and metabolic responses in the adult rat. J ApplPhysiol 95: 1767–1774, 2003.

28. Row BW, Kheirandish L, Cheng Y, Rowell PP, Gozal D. Impairedspatial working memory and altered choline acetyltransferase (CHAT)immunoreactivity and nicotinic receptor binding in rats exposed to inter-mittent hypoxia during sleep. Behav Brain Res 177: 308–314, 2007.

29. Row BW, Kheirandish L, Neville JJ, Gozal D. Impaired spatial learningand hyperactivity in developing rats exposed to intermittent hypoxia.Pediatr Res 52: 449–453, 2002.

30. Ryter SW, Alam J, Choi AM. Heme oxygenase-1/carbon monoxide: frombasic science to therapeutic applications. Physiol Rev 86: 583–650, 2006.

31. Semenza GL. Regulation of mammalian O2 homeostasis by hypoxia-inducible factor 1. Annu Rev Cell Dev Biol 15: 551–578, 1999.

32. Semenza GL. Regulation of physiological responses to continuous andintermittent hypoxia by hypoxia-inducible factor 1. Exp Physiol 91:803–806, 2006.

33. Shamsuzzaman AS, Gersh BJ, Somers VK. Obstructive sleep apnea:implications for cardiac and vascular disease. JAMA 290: 1906–1914, 2003.

34. Smith EP, Boyd J, Frank GR, Takahashi H, Cohen RM, Specker B,Williams TC, Lubahn DB, Korach KS. Estrogen resistance caused by amutation in the estrogen-receptor gene in a man. N Engl J Med 331:1056–1061, 1994.

35. Stenmark KR, Fagan KA, Frid MG. Hypoxia-induced pulmonary vas-cular remodeling: cellular and molecular mechanisms. Circ Res 99:675–691, 2006.

36. Sun J, Loader CR. Simultaneous confidence bands for linear regressionand smoothing. Ann Stat 22: 1328–1345, 1994.

37. Takahashi H, Goto N, Kojima Y, Tsuda Y, Morio Y, Muramatsu M,Fukuchi Y. Downregulation of type II bone morphogenetic proteinreceptor in hypoxic pulmonary hypertension. Am J Physiol Lung Cell MolPhysiol 290: L450–L458, 2006.

38. Valverde MA, Rojas P, Amigo J, Cosmelli D, Orio P, Bahamonde MI, MannGE, Vergara C, Latorre R. Acute activation of Maxi-K channels (hSlo) byestradiol binding to the beta subunit. Science 285: 1929–1931, 1999.

39. Voelkel NF, Tuder RM. Cellular and molecular mechanisms in the pathogenesisof severe pulmonary hypertension. Eur Respir J 8: 2129–2138, 1995.

40. Wang GL, Semenza GL. Purification and characterization of hypoxia-inducible factor 1. J Biol Chem 270: 1230–1237, 1995.

41. Wasserman L. All of Nonparametric Statistics. New York: Springer,2006, p. 268.

42. Weir EK, Olschewski A. Role of ion channels in acute and chronicresponses of the pulmonary vasculature to hypoxia. Cardiovasc Res 71:630–641, 2006.

43. Wu W, Dave N, Tseng G, Richards T, Xing EP, Kaminski N. Com-parison of normalization methods for CodeLink Bioarray data. BMCBioinformatics 6: 309, 2005.

44. Wu W, Dave NB, Kaminski N. LRSA: a new computational method foranalyzing time course microarray data. arXiv:08083399v1 [statAP], 2008[arXiv.org].



by 10.220.33.1 on August 10, 2017


Dow

nloaded from


Network analysis of temporal effects of intermittent and ...€¦ · Network analysis of temporal...

Documents

Transcript of Network analysis of temporal effects of intermittent and ...€¦ · Network analysis of temporal...