GENETIC DISSECTION OF THE TRANSCRIPTIONAL HYPOXIA RESPONSE AND GENOMIC REGIONAL CAPTURE FOR MASSIVELY PARALLEL SEQUENCING by DOUGLAS WILLIAM TURNBULL A DISSERTATION Presented to the Department of Biology and the Graduate School ofthe University of Oregon in partial fulfillment of the requirements for the degree of Doctor of Philosophy September 2008 11 University of Oregon Graduate School Confirmation of Approval and Acceptance of Dissertation prepared by: Douglas Turnbull Title: "Genetic Dissection of the Transcriptional Hypoxia Response and Genomic Regional Capture for Massively Parallel Sequencing" This dissertation has been accepted and approved in partial fulfillment of the requirements for the Doctor of Philosophy degree in the Department of Biology by: Karen Guillemin, Chairperson, Biology Eric Johnson, Advisor, Biology Bruce Bowerman, Member, Biology Christopher Doe, Member, Biology Kenneth Prehoda, Outside Member, Chemistry and Richard Linton, Vice President for Research and Graduate Studies/Dean of the Graduate School for the University of Oregon. September 6, 2008 Original approval signatures are on file with the Graduate School and the University of Oregon Libraries. © 2008 Douglas William Turnbull iii iv An Abstract of the Dissertation of Douglas William Turnbull in the Department of Biology for the degree of to be taken Doctor of Philosophy September 2008 Title: GENETIC DISSECTION OF THE TRANSCRIPTIONAL HYPOXIA RESPONSE AND GENOMIC REGIONAL CAPTURE FOR MASSIVELY PARALLEL SEQUENCING Approved: _ Dr. Eric Johnson, Advisor When cells are faced with the stress of oxygen deprivation (hypoxia), they must alter their physiology in order to survive. One adaptation cells make during hypoxia entails the transcriptional activation of specific groups ofgenes as well as the concurrent repression ofother groups. This modulation is achieved through the actions of transcription factors, proteins that are directly involved in this transcriptional activation and repression. I studied the transcriptional response to hypoxia in the model organism Drosophila melanogaster utilizing DNA microarrays to examine the transcriptomes of five different mutant Drosophila strains deficient in the hypoxia-responsive transcription factors HIF-l, FOXO, NF1d3, p53, andMTF-l. By comparing hypoxia responsive gene expression in these mutants to that of wild type flies and subsequently identifYing binding sites for each transcription factor near putative target genes, I was able to identifY the v transcripts regulated by each transcription factor during hypoxia. I discovered thatFOXO plays an unexpectedly large role in hypoxic gene regulation, regulating a greater number of genes than any other transcription factor. I also identified multiple interesting targets of other transcription factors and uncovered a potential regulatory link: between HlF-l and FOXO. This study is the most in-depth examination of the transcriptional hypoxia response to date. I was also involved in additional research on transcriptional stress responses in Drosophila. Also included in this dissertation are two papers on which I was the second author. One paper identified a regulatory link: between the transcriptional responses to hypoxia and heat-shock. The other examined elevatedC~ stress (hypercapnia) in Drosophila, showing that this stress causes the down-regulation ofNFKB-dependent antimicrobial peptide gene expression. My studies of stress responses would not have been possible without well-described mutant fly strains. Another part of my dissertation research involved the creation of a method for characterizing new mutants for future studies. When researchers seek to identify the molecular nature of a mutation that causes an interesting phenotype, they must ultimately determine the specific responsible genomic sequence change. While classical genetic methods and other techniques can easily be used to roughly map the location ofa mutation in a genome, regions identified by these means are usually so large that sequencing them to precisely identify the polymorphism is laborious and slow. I have developed a technique that makes sequencing genomic regions ofthis size much easier. My technique involves capturing genomic regions by hybridization of fragmented genomic target DNA to biotinylated probes generated from fosmid DNA, which are subsequently vi immobilized and washed on streptavidin beads. Genomic DNA fragments are then eluted by denaturation and sequenced using the latest generation ofmassively parallel sequencing technology. I have demonstrated the effectiveness of this approach by sequencing a mutation-containing 336-kilobase genomic region from a Caenorhabditis elegans strain. My entire protocol can be completed in two days, is relatively inexpensive, and is broadly applicable to any situation in which one wants to sequence a specific genomic region using massively parallel sequencing. This dissertation includes both my previously published and my coauthored materials. CURRICULUM VITAE NAME OF AUTHOR: Douglas William Turnbull PLACE OF BIRTH: LaGrande, Oregon DATE OF BIRTH: March 18th, 1979 GRADUATE AND UNDERGRADUATE SCHOOLS ATTENDED: University of Oregon, Eugene, Oregon University ofPuget Sound, Tacoma, Washington DEGREES AWARDED: Doctor of Philosophy in Biology, 2008, University of Oregon Bachelor of Science in Biology, 2001, University of Puget Sound AREAS OF SPECIAL INTEREST: Gene Regulation Genomics PROFESSIONAL EXPERIENCE: Graduate Research Fellow, Department of Biology, University of Oregon, Eugene, Oregon, 2003-2007 Graduate Teaching Fellow, Department of Biology, University of Oregon, Eugene, Oregon, 2002-2003 vii GRANTS, AWARDS AND HONORS: Keck Foundation Training Grant, University of Oregon, 2007-2008 NIH Genetics Training Grant, University of Oregon,200S-2007 NIH Molecular Biology Training Grant, University of Oregon, 2003-2004 Murdock Summer Research Fellowship, University of Puget Sound, 2000 PUBLICATIONS: Baird, N.A., D.W. Turnbull, and E.A. Johnson. 2006. Induction of heat shock pathway during hypoxia requires regulation of Heat shock factor by Hypoxia- inducible factor-I. J Biol Chem 281: 38675-38681. viii IX ACKNOWLEDGMENTS I thank my advisor, Dr. Eric Johnson, for allowing me the freedom to design my own projects and experiments and for his seemingly endless supply of new scientific ideas. I also thank my family for fostering my curiosity and lifelong love of learning. Finally, I thank my wife Lisa for her love and support and for helping me to live each day to its fullest. xI dedicate this document to my father Valiant Richard Turnbull for the perfectionism and attention to detail I inherited from him. For this reason lowe my success in science and in life to him. Rest in peace Dad. xi TABLE OF CONTENTS Chapter I. INTRODUCTION TO HYPOXIC GENE REGULATION AND NEXT GENERATION SEQUENCING . II. GENETIC DISSECTION OF THE TRANSCRIPTIONAL HYPOXIA RESPONSE IN DROSOPHILA MELANOGASTER . Page 1 11 Materials and Methods . Results . 13 15 Discussion 23 III. HEAT SHOCK PATHWAY IS ACTIVATED BY THE DIRECT REGULATION 29 OF HEAT SHOCK FACTOR BY HIF-l DURING HYPOXIA .. Results 31 Discussion 41 Methods 44 IV. ELEVATED C02 SUPPRESSES SPECIFIC DROSOPHILA INNATE IMMUNE 49 RESPONSES DOWNSTREAM OF NF-kB PROTEOLYTIC ACTIVATION Results and Discussion 50 Concluding Remarks..... 63 Materials and Methods...... 63 V. GENOMIC REGIONAL CAPTURE FOR NEXT GENERATION SEQUENCING 69 Materials and Methods.......................................................................................... 71 Results 74 Discussion 81 VI. CONCLUSION........................................................................................................... 84 APPENDIX: ONLIJ'JE SUPPLEMENTAL MATERIAL FOR CHAPTER IV............. 87 xii Chapter Page REFERENCES.................................................................................................................. 90 --_ .._._-_._---------------- xlii LIST OF FIGURES R~ ~~ Chapter II 1. Example data plot from BAMarrayTM Bayseian ANOVA data analysis comparing the expression of 435 hypoxia-regulated transcripts in sima- and WT flies............................ 17 2. Hierarchical clustering of hypoxia modulated transcripts with significantly perturbed expression levels in at least one mutant strain 18 3. Examples of computationally identified conserved binding sites for putative hypoxia-responsive transcription factor targets 23 4. Schematic representation of a potential regulatory mechanism by which Sima likely contributes to the activation of dFOXO during hypoxia through transcriptional upregulation of Imp-L2 26 Chapter III 1. Hsftranscript levels are increased in a HIF-la-dependent manner.................. 33 2. Hsf gene region has conserved HREs and is a direct target of HIF-l............... 35 3. Hsp transcript levels are increased in an Hsf and HIF-la-dependent manner during hypoxia 37 4. Up-regulation of Hsps is Hsf dosage-dependent 38 5. Up-regulation of Hsps after reoxygenation is HIF-la-dependent and critical to survival 40 Chapter IV 1. Hypercapnia affects Drosophila development and physiology independently of neuronal CO2 sensing 52 2. Hypercapnia down-regulates specific antimicrobial peptides in Drosophila... 54 3. Hypercapnia increases mortality of bacterial infection in Drosophila 56 4. Hypercapnia suppresses antimicrobial peptide induction with rapid onset and recovery................................................................................ 59 5. Hypercapnia suppresses innate immune responses via a novel pathway......... 60 Chapter V 1. Schematic representation of regional capture protocol..................................... 76 2. Verification of captured library prior to NGS................................................... 79 3. Graphical representation of sequences from regionally captured samples aligned to a portion of the C. elegans genome............................................ 81 XIV Figure Page Appendix S1. Absolute levels of antimicrobial peptide RNAs in S2* cells, normalized to RP49............................................................................................................ 87 xv LIST OF TABLES Table Page Chapter II 1. Hypoxia modulated transcripts with significantly altered hypoxia expression levels in mutant strains , 19 Chapter V 1. Fosmid clones used for preparation of biotinylated probes 77 Appendix S1. Hypercapnia strongly down-regulates egg formation genes, but does not up-regulate genes induced by hypoxic, heat-shock, or oxidative stress responses , 88 S2. Hypercapnia does not affect cell viability........................................................ 89 CHAPTER I INTRonUCTION TO HYPOXIC GENE REGULATION AND NEXT GENERATION SEQUENCING Background of hypoxic gene regulation Molecular oxygen plays a multitude of essential roles in eukaryotic cellular physiology, perhaps most notably as the terminal electron acceptor in the electron transport chain of cellular respiration. Because of this critical role in cellular energy harvest, eukaryotic organisms have evolved several mechanisms to deal with situations where the supply of oxygen becomes scarce (hypoxia). One of the central hypoxic adaptations made by eukaryotic organisms .is the differential regulation of specific sets of transcripts. By varying the amounts of transcripts for various genes, cells are able to alter quantities of assorted proteins, and by doing so, change their physiology in ways that allow them to better cope with hypoxia. This hypoxia-induced transcriptional modulation accomplished through ihc action of hypoxia-responsi ve transcription factors, proteins that are directly involved in regulating the quantities of transcripts that are produced from specific sets of genes. ---------- 2 The transcription factor with the best-characterized role in hypoxic gene regulation is Hypoxia Inducible Factor-l (HIF-l) (1). HIF-l is a heterodimeric transcription factor composed of two subunits known as HIF-1 a and HIF-1~. HIF-1 ~ is a protein that is constitutively present in many cells, and dimerizes with several other transcription factors other than HIF-la in order to regulate transcription of genes involved processes as diverse as cellular responses to environmental toxins, to neural development (2). HIF-1 ~ is absolutely essential for the function of HIF-1 as a hypoxia- responsive transcription factor, but plays little role in the regulation of its activation. HIF-1 a, on the other hand, is a protein that is tightly regulated by cellular oxygen levels, and this regulation is the key step in the regulation of HIF-1 activity. When oxygen is present in a cell at normal levels (nonnoxia), HIF-la is hydroxylated at a proline residue by an enzyme known as HIF-l Proline Hydroxylase (HPH). This hydroxylation reaction is dependent on the presence of molecular oxygen, and serves as the direct link between oxygen abundance and HIF-Ia activity (3). When HIF-la is proline hydroxylated, it is recognized by Von Hippel Lindau tumor suppressor protein (VHL), which facilitates the ubiquitination and rapid proteasomal degradation of the HIF- lao This oxygen-dependent pathway prevents HIF-la protein from ever being present at normal oxygen concentrations long enough to dimerize with HIF-1 ~ and take part in transcriptional regulation. During hypoxia, however, HIF-1 a is not degraded, dimerizes with HIF-1~, and the dimeric HIF-1 transcription factor activates the expression of its target genes. 3 When it is acti ve in hypoxia, HIF-1 binds DNA sequence motifs with the core 5'- TRCGTG-3'. These elements are present, often in clusters, near genes that are regulated by HIF-1 under hypoxia. A number of genes have been identified as HIF-1 targets in mammalian and other model systems. HIF-1 is responsible for the transcriptional up- regulation of genes such as erythropoietin (EPO), and Vascular endothelial growth factor (VEGF) which increase oxygen delivery to hypoxic areas by stimulating red blood cell production, and the recruitment of blood vessels respectively in mammals(4). In the fruit fly Drosophila rnelanogaster, RIF-1 has been shown to regulate HPH, providing a feedback mechanism by which HIF-l activity is attenuated during hypoxia (5). Nathan Baird and J showed that in Drosophila, HIF-1 directly regulates the transcription of Heat shockfactor (HSF), a transcription factor that activates the transcription of heat shock protein genes (6). The full complement ofHIF-l regulated genes in Drosophila has not been exhaustively cataloged yet, however. Another transcription factor that has been shown to regulate gene expression during hypoxia is the Forkhead transcription factor FOXO. FOXO activity is inhibited by Insulin signaling, causing FOXO to activate gene expression during starvation conditions. Mechanistically, FOXO is phosphorylated by the kinase AKT in response to Insulin or other gro"'ih factors. This phosphor)' lation causes the export of FOXO from the nucleus, thereby inhibiting FOXO-mediated transcriptional activation (7). Interestingly, elevated FOXO activity is linked to increased longevity in Drosophila and Caenorhabditis elegans, suggesting that the genes regulated by this transcription factor have significant protective functions for cells (7). Indeed, several genes involved in 4 reactive oxygen species detoxification have been described as FOXO targets, in addition to a number of genes involved in protecting cell from other stresses (8). Recently, FOXO has been shown to activate gene expression in response to hypoxia in human cells (9). This study found that FOX03a, one of the four FOXO transcription factors in humans, is transcriptionally activated by HIF-I during hypoxia. This transcriptional up regulation contributes to the activation of FOXO-dependent gene expression during hypoxia. Interestingly, FOX03a was found to activate transcription of Cited2, gene that is known to down-regulate HIF-I activity, demonstrating that HIF-I and FOX03a regulate each other in a negative feedback loop. Nuclear factor kappa-B (NFKB) is another transcription factor that has been shown to activate gene expression in response to hypoxia. NFKB has been most thoroughly characterized as a central transcriptional regulator of the inflammatory response in mammals (10). Under normal conditions, NFKB is bound by the repressor protein IKB, which masks the nuclear localization sequence (NLS), prevent nuclear accumulation of the transcription factor. NFKB can be activated by a number of stimuli, including bacterial antigens, proinflammatory cytokines, oxidative stress, and ultraviolet radiation. Upon activation, IKB is phosphorylated, ubiquitinated, and degraded, exposing the NLS ofNFKB, which translocates in to the nucleus and activates transcription. Numerous NFKB target genes have been identified in multiple organisms, including those encoding inflanlmatory cytokines, chemokines, and cell adhesion molecules. In. Drosophila, there are three NFKB proteins, Dorsal (Dl), Relish (ReI), and Dorsal-related immunity factor (Dit) (11). Dl has been extensively studied for its role in development. 5 ReI and Dif play key roles in the innate immune system of the fly, where they transcribe genes encoding anti-microbial peptides in response to bacterial and fungual infection. Although numerous studies have shown that NFKB is activated by hypoxia, the mechanism by which this occurs is unclear (12) (13). There is evidence that hypoxia can induce the production of Nitric Oxide (NO) as well as other reactive oxygen species (ROS), and activation ofNFKB by ROS has been well documented (14) (15). These observations seem to support the hypothesis that NFKB activation by hypoxia takes place because of a mechanism involving some sort of hypoxia-related change in the redox state of the cellular environment. Although we are not sure of the mechanism responsible, our lab has observed multiple targets of ReI, as well as ReI itself being transcriptionally up regulated by hypoxia in Drosophila (16). Another stress-responsive transcription factor has been implicated in the hypoxia response is p53. This transcription factor is mainly known as a transcriptional activator of pro-apoptotic genes that is activated by DNA damage (17). A number of other stimuli can activate p53, including ultraviolet radiation, X-rays, low pH, and heat shock. In the absence of activating stimuli, p53 is ubiquitinated and targeted for degradation by the ubiquitin ligase MDM2. Hypoxia has been shown to activate p53 through direct binding and inhibition ofMDM2 by HIF-lu (18). Other researchers have reported that ROS play a role in the hypoxic activation of p53 (19). The role of p53 in Drosophila hypoxic gene regulation has not been previously examined. Because Drosophila possesses a p53 homologue, and ROS are known to be produced during hypoxia in the fly, it seems likely that p53 plays a role in hypoxic gene regulation in Drosophila. 6 Another hypoxia-responsive transcription factor that is likely to playa role in the hypoxia response of Drosophila is the Metal-responsive transcription factor-1 (MTF-1). MTF-1 is an evolutionarily-conserved transcription factor that possesses DNA binding zinc fingers which change conformation in response to fluctuations in cellular zinc concentration (20). MTF-1 activates the transcription of Metallothionein genes in response to increases in cytoplasmic metal ion levels. Metallothioneins are small, cysteine-rich proteins that bind zinc under physiological conditions. When other metal ions such as copper or cadmium enter the cytoplasm, Metallothioneins release their bound zinc and bind these more toxic ions with higher affinities. This mechanism allows MTF-1 to respond to a multitude of metals despite the fact that only zinc directly binds the transcription factor (21). Although the vast majority of work on MTF-1 has been directed toward understanding its role in cellular metal homeostasis, it has become clear that MTF-1 directs transcriptional changes in response to hypoxia as well. In human cells, the hypoxic transcriptional induction of Placenta growth factor (PIGF) was shown to be dependent on MTF-1 (22). Metallothionein genes have also been shown to be transcribed in response to hypoxia in an MTF-1-dependent manner (23). In Drosophila, our lab has observed transcriptional activation of Metallothionein genes during hypoxia, which is highly likely to require MTF-l (16). Drosophila has been well established as an extremely powerful model system for uncovering fundamental new facets of genetics and molecular biology. Numerous cellular pathways that have key roles in human disease are also present in Drosophila, and in-depth molecular biological characterization of these pathways is much easier and more rapid in the fly due to its short generation time and widely available genetic tools. 7 Each of the five hypoxia-responsive transcription factors that I described in the preceding pages is present in Drosophila. In this dissertation, I describe my work on understanding the transcriptional response to hypoxia using the Drosophila model system. I examined hypoxia-responsive gene expression in five different tly lines, deficient in HIF-l, FOXO, NFKB, p53, and MTF-l. By combining this analysis with bioinformatics, I was able to determine, to a large extent, the role of each transcription factor in the broader transcriptional hypoxia response. My work represents the first dissection of the transcriptional hypoxia response to this level of detail. This would not have been possible without use of the Drosophila model system. Background of genomic regional enrichment for sequencing It could be said that the blueprint for an organism lies in the sequence of its genome. Because the genomic sequence of any living thing contains so much useful information about the biology of that organism, a great deal of effort has been devoted to genome sequencing projects. The genome sequences produced by these projects make molecular biological studies enormously easier than they would be in the absence of reference genome sequences for the organisms being studied. The bulk of existing genomic sequence data has been produced using variations of the chain-terminator method ofDNA sequencing developed by Frederick Sanger and colleagues in the 1970's (24). In this method, DNA polymerase synthesizes a new strand of DNA from a specific primer that has been annealed to the DNA template that is to be sequenced. This synthesis takes place in the presence of a small concentration of dideoxynucleotides, which, when incorporated in to the newly synthesized strand, 8 prevent DNA polymerase from adding additional nucleotides and thus terminate the synthesis of the new chain ofnucleotides. This results in the synthesis of new DNA strands from the sequencing primers that are a multitude of lengths. In modem Sanger sequencing, the dideoxy terminator nucleotides also incorporate specific fluorophore moieties for each of the four dideoxynucleotides in the reaction. When the DNA synthesis is complete, the mixture of product strands is separated by size using electrophoresis to a resolution that allows the researcher to discern size differences of one nucleotide. By observing the size of each fragment and taking note of the fluorescent color ofthe chain-terminator nucleotide that was incorporated, the researcher can deduce the sequence of the newly synthesized DNA strand beginning at the 3' end of the sequencing primer. The reads produced by Sanger sequencing are typically less than one kilobase (kb) in length. This read length is 3,000,000 times smaller than the length of the human genome, for example, making the task of sequencing entire genomes using the Sanger approach laborious and costly. The desire to sequence new genomes more quickly as well as a need to determine sequence polymorphisms at the root of genetic diseases has led to the development of several types next generation DNA sequencing (NGS) technology. These NGS approaches all are based on the concept of sequencing massive numbers of random DNA fragments in paralleL Each sequencing read from a NGS run is much shorter than those produced in a Sanger sequencing run; the advantage that NGS provides lies in the number ofreads produced by these approaches. For example, the individual sequencing reads produced by Illurnina Genome Analyzer II NGS instrument 9 are 50 nucleotides or less in length, hbwever, this instrument produces roughly 96,000,000 of these reads which results in 4.8 gigabases ofraw sequence information. Due to sequencing errors and the necessity of significant overlap betweeQ. individual reads for assembly into a long continuous sequence, NGS instruments can still not sequence entire genomes of higher eukaryotes in single runs. In many cases, researchers are only interested in sequencing portions of a genome, such as when identifying disease related single nucleotide polymorphisms (SNPs) in clinical samples, or when attempting to identify the SNP responsible for an interesting phenotype in a mutant identified in a genetic screen. NGS instruments, however, do not sequence targeted genomic regions; random DNA fragments serve as the sequencing template resulting in reads that are scattered throughout the genome. Sanger sequencing allows researchers to select the region they would like to sequence based on the location of the sequencing primer, but sequencing large regions requires the design of numerous primers that each must be run in a separate sequencing reaction making the targeted sequencing oflarge n'gions with Sanger technology both slow and expensive. For these reasons, several groups are pur,;uing methods for selecting large genomic regions for sequencing byNGS. Two independent groups published methods for targeted sequencing of large genomic regions by NGS last year (25,26). Both ofthese papers describe essentially the same protocol in'Nhich custom DNA microarrays are designed and constructed to contain oligonucleotide sequences that are homologous to a large genomic region of interest. Genomic DNA is randomly sheared and hybridized to the custom microarray, and the sequences that are not homologous to those represented on the array are washed away. 10 The genomic fragments that hybridized to the array are then eluted by denaturation, amplified, and sequenced by NGS. This approach has proven to be an effective method of isolating and sequencing specific large genomic regions. I have developed an alternative method to the custom microarray based approach for capturing large genomic regions for sequencing by NGS. My method uses fosmid or bacterial artificial chromosome (BAC) DNA that is homologous to the genomic region of interest in place of the custom microarrays that are used in the previously published protocols. My method is significantly faster, and can be completed for a small fraction of the cost of microarray-based methods. Furthermore, no complete reference genome sequence for the target region is required by my technique. The low cost and simplicity of my technique will make current NGS technology an effective means of sequencing large genomic regions for more than only the most well funded research groups in the world. P.D. Etter contributed to the research described in Chapter II as a second author. N.A Baird was the first author of Chapter III. LT. Helenius and T. Krupinski were co- first authors and Y. Gruenbaum was the third author of Chapter IV. Bridge to Chapter II In the previous chapter I described what is known about the hypoxia-responsive transcriptional regulatory activities ofHIF-I, FOXO, NFKB, p53, and MTF-l. In Chapter II, I will report the identities of numerous genes that are regulated by each transcription factor during hypoxia in Drosophila melanogaster. 11 CHAPTER II GENETIC DISSECTION OF THE TRANSCRIPTIONAL HYPOXIA RESPONSE IN DROSOPHILA MELANOGASTER P.D. Etter contributed to this work by performing the genetic crosses and microarray experim~nts for two of the six genetic strains used. I conducted the remainder ofthe experiments as well as the data analysis and writing. Oxygen is required by all multicellular organisms for survival. When animals are deprived of oxygen, they undergo a complex series of behaviorial, developmental, physiological and molecular changes in order to survive the challenge. Hypoxia is also encountered by cells in tissues affected by heart attack, stroke, wounding, as well as within poorly vascularized tumors. Hypoxic tumors are often resistant to treatment, and indicate a poor prognosis for cancer patients (27). Due to the medical importance of hypoxia, a great deal of research has been devoted to understanding the mechanisms cells utilize to survive this stress. One mechanism widely utilized by organisms to cope with oxygen deprivation involves the modulation of gene expression in response to hypoxia. The transcription factor HIF-l has been well characterized as a regulator of hypoxia-responsive gene 12 expression (1). HIF-l is a heterodimeric transcription factor composed of an a and a ~ subunit. HIF-la is rapidly ubiquitinated and degraded in the presence of oxygen, but during periods of hypoxia the protein dimerizes with HIF-1~, which is constitutively expressed, to form the active HIFI transcription factor complex. HIF-l binds genomic regulatory elements known as hypoxia response elements (HREs), where it is transactivated and facilitates the transcriptional activation of genes that help cells survive hypoxia by either decreasing their demand for oxygen by shifting to glycolytic metabolism, increasing the supply of oxygen by recruiting blood vessels to the hypoxic area, or by stimulating the production ofred blood cells (1). In addition to HIF-l , other transcription factors are known to take part in regulating hypoxic gene expression. Nuclear factor kappa-B (NFKB) transcription factors playa large part in regulating gene expression in response to inflammatory stimuli and are also known to be activated by hypoxia (12) (13). Another transcription factor that has been shown to regulate gene expression in response to hypoxia is p53, which is a key regulator of genes that are responsible for the induction of apoptosis (28). The FOXO family of Forkhead transcription factors regulate gene expression in response to a multitude of stresses, and have recently been shown to be involved in regulating the transcriptional response to hypoxia in human cells(9). Although primarily characterized as a regulator of genes involved in the cellular response to heavy metal stress, the metal response element (MRE) binding transcription factor (MTF-l) has also been shown to respond to nitric oxide as well as hypoxia (29) (23). While NFKB, p53, FOXO, and 13 MTF-1 are all known to be hypoxia-responsive transcription factors, the comprehensive set of genes regulated by each during hypoxia have yet to be determined. Much of what is currently known in the field of genetics was gleaned from the study of the fruit fly Drosophila melanogaster. Drosophila has also proven to be an effective model system for the study of hypoxia. The Drosophila proteins Similar and Tango have been identified as functional homologues of mammalian RIF-1 a and RIF-1 ~ (30,31). Functional homologues of the hypoxia-responsive transcription factors NFKB, p53, FOXO, and MTF-1 are also present in Drosophila. Previously, our lab used microarrays to identify Drosophila genes that are transcribed in response to hypoxia (16). In the present study, we used available mutant Drosophila strains deficient in RIF-l to identify hypoxia-responsive transcripts that are regulated by each respective transcription factor. Materials and Methods Drosophila strains. In order to minimize differences in gene expression due to variations in the genetic backgrounds of the strains that were studied, both the wild type Oregon R control strain and each mutant were crossed to a strain bearing visible markers on each chromosome (yw;Sp/cCyO;dDp/TM3-Sb). The genotype of the wild type control strain wasyw;Sp/CyO;+. The genotypes of the mutant strains studied were as follows: yw;Sp/CyO;sima7607, yw;Sp/CyO,foxo21/foxo25, yw;Sp/CyO;p535A-1-4, yw;Sp/CyO;reze20, and yw;Sp/CyO/mtj-1140-IR. 14 Hypoxia treatment. Twenty male and female adult flies between the ages of 1 and 5 days post-ec1osion were collected from each genotype and placed in separate vials containing standard Drosophila medium for each hypoxia experiment. Six replicate populations were collected for each genotype. After a 24-hour recovery period following counting, three of the vials for the wild type strain and each mutant were placed in a sealed chamber at room temperature that was then flushed with a mixture of 0.5% O2 and 99.5% N2 (hypoxia). The three other vials for each respective genotype were placed in an identical chamber containing normal atmospheric oxygen concentrations (normoxia). The flies were incubated in the hypoxia and normoxia chambers for six hours, following which they were rapidly removed and frozen in liquid N2• RNA extraction and DNA microarray experiments. Following either hypoxia or normoxia treatment, frozen flies were homogenized in TRIzol reagent and RNA was extracted according to the manufacturer's protocol (Invitrogen). For each strain, 20 I-lg of RNA from each normoxia and hypoxia treatment were labeled for microarray hybridization using the SuperScript Direct cDNA Labeling System (Invitrogen), incorporating either Cyanine 3 (normoxia) or Cyanine 5 (hypoxia) conjugated dUTP (Perkin Elmer). In each strain, gene expression during hypoxia was assayed by combining labeled normoxia and hypoxia eDNA, and hybridizing the samples to DNA microarrays containing 16,416 oligonucleotides from the INDAC set representing the D. melanogaster transcriptorne (Illumina). Three microarrays were hybridized for each mutant, representing three independent biological repeats. Microarrays were scanned and analyzed using Gene Pix Pro 6.0 software (Molecular Devices). 15 Microarray Data Analysis. In order to identify the complete set of hypoxia- modulated transcripts, we identified transcripts in wild type flies that increased or decreased greater than 1.6 fold during hypoxia (average of 3 microarray data sets). We then examined the expression level of this set of transcripts in each mutant strain. In order to identitY genes that were differentially regulated during hypoxia in each mutant strain, we used BAMarrayTM 2.0 software, which utilizes Bayesian ANOVA analysis to identitY significant differences in transcript levels between multiple microarray data sets (32). Conserved binding sites for each transcription factor were identified using our own software (http://genomix.uoregon.edu/~eric-johnsonlcgi-binlGAS.pl). The motifs used for each transcription factor binding site search are as follows: Sima - TACGTG; dFOXO - RTAAACAA; Rel- GGGAHNYMY; p53 - RRRCWWGYYY; MTF-1 - TGCRCNC. Groups of2 or more binding sites within 2 kb of an open reading frame that are conserved in D. melanogaster as well as at least three other Drosophilid genomes were considered likely to be functional binding sites for the transcription factor being queried. Results Hypoxia-induced changes in gene expression-DNA microarrays were used to compare transcriptional profiles of adult Drosophila melanogaster flies that had been exposed to hypoxia to those of flies that were kept at atmospheric oxygen concentrations. Upon treatment with hypoxia, 435 transcripts were either up- or down-regulated greater than 1.6 fold (3.05% of the transcripts represented on the array). These results are in agreement with the set of transcripts that we previously identified as being induced by 16 hypoxia (16). In order to identify transcription factors involved in the regulation of the identified hypoxia modulated transcripts, we examined hypoxia-responsive gene expression in mutant strains deficient in Sima, dFOXO, NFKB, p53, and MTF-l. Hypoxia-responsive gene regulation in mutanr strains-- In order to minimize differences among the mutants due to variation in their genetic backgrounds, we performed a series of crosses to make each mutant genetically identical to the wild-type strain, with the exception of the chromosome containing each respective mutation. We then exposed each mutant strain to the same hypoxia treatment as the wild type flies and identified transcripts that were either induced or repressed by hypoxia using microarrays. Baesian ANOVA analysis was then used to identify transcripts that were differentially regulated under hypoxia between wild type flies, and each mutant strain (Figure 1) (32). Of the 435 transcripts that were induced or repressed 1.6 fold or greater during hypoxia in wild type flies, 190 were identified as significantly different in anyone of the mutant strains by Bayesian ANOVA analysis. We analyzed the expression or each of these 190 genes by hierarchical clustering (Figure 2). Interestingly, dFOXO mutants exhibit the greatest difference in hypoxia-induced gene regulation when compared to wild type flies by hierarchical clustering. Sima mutants exhibit the second largest difference in hypoxic gene regulation from wild type flies, while p53, rei, and MTF-l exhibit patterns in gene regulation that are more similar to wild type. Figure J. Example data plot from BAMarrayTM Bayseian ANOYA data analysis comparing the expression of 435 hypoxia-regulated transcripts in simo- and WT fl ies. Genes are class ified as significantly di fferentially expressed between two data sets when their Zcut values are large, and their posterior variances are close to I. BAMarray™ automates the selection of such genes using a rule based on the specific data set. For details of BAMarra/M statistics, please see {Ishwaran et aI., 2006, BMC Bioillfonnatics, 7, 59}. 18 I~. 000 0000000 .. '0000~N.---i .... ! I IO...--if'JM ;J~~~::f;;'j::::::. ...''';.·.1·... ·111· .1.'.' '.', _. ~~u... - ",'- 'I'· , ......I' ,. ,. ",. II"",,·, '''.1',' ... ,q ....... , ,.,.. .. ' "",., , . . ~,.. .'.....,," ~; ". ,!~:~~*~~;~;,:,.' Figure 2. Hierarch ical clustering of hypox ia modulated transcripts with significantly perturbed expression levels in at least one mutant strain. Rows represent the average expression levels of transcripts from 3 biological replicates for the strain indicated in each column. During hypoxia, dFOXO mutants exhibit the greatest difference from wild type flies, followed by sima mutants. 19 In order to refine the lists of target genes to include only direct transcriptional targets of each transcription factor under hypoxia, and to eliminate genes indirectly affected by the loss of that factor, we used a bioinformatic approach (Figure 3). Genes were considered strong candidate targets of a given transcription factor under hypoxic conditions if they showed significantly altered expression levels in that mutant under hypoxia and also contained two or more conserved binding sites for the affected transcription factor within two kilobases of the open reading frame. Based on these criteria, we identitied 47 hypoxia modulated target genes for dFOXO, 24 targets of Sima, 16 targets of Relish, 14 targets ofp53, and 14 MTF-I targets (Table 1). These numbers are in accordance with the differences in hypoxic transcriptional modulation between each mutant strain and wild type based on hierarchical clustering, with FOXO being required for the hypoxic modulation of the greatest number of transcripts, followed by Sima, ReI, p53, and MTF-1 (Figure 2). Table 1. Hypoxia modulated transcripts with significantly altered hypoxia expression levels in mutant strains Transcript dFOXO-dependent transcripts CG18466 NAD-dependent methylenetetrahydrofolate dehydrogenase CGl0383 CG7219 Proteinase inhibitor 14 CG5966 lipid metabolism CG8709 lipid metabolism CG7850 puckered CG5953 Fold hypoxic induction inWT 3.94 3.66 336 3.01 2.93 2.85 2.77 Fold hypoxic induction in mutant 1.06 1.25 0.00 1.00 1.42 1.93 0.00 Conserved binding sites 4 5 2 2 6 8 8 20 CG3090 Sox box protein 14 2.66 0.00 2 CG13868 2.39 0.79 7 CG3705 astray 2.39 1.05 2 CG14801 nucleic acid metabolism 2.22 0.00 8 CG32369 carbohydrate metabolism 2.22 0.00 6 CG15423 2.22 1.24 2 CG11796 amino acid catabolism 2.20 0.69 3 CG7554 comm2 2.17 0.00 2 CG5295 2.13 1.41 3 CG 12489 Defense repressor 1 2.04 0.00 10 CG 1600 Zinc-containing alcohol dehydrogenase superfamily 2.01 1.15 6 CG31811 centaurin gamma 1A 1.96 1.29 5 CG5467 1.93 0.00 3 CG5246 proteolysis Peptidase S1 1.93 1.13 2 CG16926 1.91 0.91 2 CG12358 polyA-binding protein interacting protein 2 1.89 1.03 3 CG32972 cell adhesion 1.84 0.00 4 CGl0580 fringe 1.82 0.00 6 CG7525 Tie-like receptor 1.78 0.00 2 CG15162 Misexpression suppressor ofras 3 1.73 0.96 3 CG12845 Tetraspanin 42Ef 1.73 1.07 2 CG8222 PDGF- and VEGF-receptor related Pvr 1.72 1.03 2 CG31319 RhoGAP88C 1.71 0.00 3 CG3234 timeless 1.71 0.00 2 CG5248 locomotion defects 1.69 0.00 3 CG13388 A kinase anchor protein 200 1.67 1.13 2 CG1921 sprouty 1.66 1.04 3 CG5461 bunched 1.64 1.00 19 CG4993 PRL-l 1.59 1.03 5 CG8095 scab 1.59 1.06 2 CG8127 Ecdysone-induced protein 75B 1.57 1.05 24 CG14401 1.57 0.83 3 CG10637 Numb-associated kinase 1.53 0.00 3 CG7210 kelch 1.53 0.92 2 CG17278 Proteinase inhibitor] I 1.52 0.00 2 CG8098 Picot 0.63 0.93 4 CG13607 0.62 1.06 2 CG7720 cation transporl 0.54 1.01 7 CG12602 cation transport 0.42 0.74 2 Sima-depenndent transcripts CG31449 Heat-shock-protem-70Ba 4.91 3.89 2 CGI0160 Ecdysone-inducible gene L3 3.06 1.09 2 CG15009 Ecdysone-inducible gene L2 2.13 0.92 6 CG7224 1.60 0.30 3 CGl1652 DPH2 protein 68D2 1.56 0.15 2 CG31769 1.29 -0.07 2 21 CG31543 HIF prolyl hydroxylase 1.16 -0.02 9 CG1333 ErolL 1.05 -0.04 2 CG5467 0.95 0.10 4 CG6246 nubbin 0.95 0.41 4 CG 12358 polyA-binding protein interacting protein 2 0.92 0.27 5 CG32972 Beta-Ig-H3/fasciclin Protein kinase-like 35Al 0.88 0.00 5 CG5748 Heat shock factor 0.86 0.08 3 CG30069 cell cycle 0.85 -0.05 3 CG 10746 fledgling ofKlp38B 0.84 -0.22 4 CG8913 oxygen and reactive oxygen species metabolism 0.83 0.22 3 CG3234 timeless 0.77 -1.20 2 CG9078 infertile crescent 0.66 0.11 2 CG 10827 Alkaline phosphatase -0.66 0.00 2 CG12385 Peptidase SI -0.70 -0.17 2 CG13997 -1.21 -0.06 2 CG12602 cation transport -1.24 -0.63 3 CGI0129 nudel -1.97 0.00 3 Relish-dependent transcripts CG32130 apoptosis proteolysis Peptidase SI 4.72 7.00 5 CG 18466 NAD-dependent methylenetetrahydrufolate dehydrogenase 3.94 1.75 4 CG3896 defense response 2.11 1.41 2 CG 11992 Relish 2.03 1.00 4 CG5467 1.93 1.08 3 CG32972 cell adhesion 1.84 1.00 3 CGI7631 1.64 1.06 2 CG 10248 Cytochrome P450-6a8 1.53 2.00 6 CG170i 0.64 0.94 3 CG8098 Picot 0.63 0.98 2 CGI0827 0.63 1.00 2 CG 12387 Trypsin 0.58 0.88 2 CG 17876 Amylase distal 0.55 1.00 4 CG8083 cation transport 0.48 0.70 2 CG 12602 cation transport 0.42 0.67 2 MTF-l-dependent tra!1scripts CG4463 Heat shock protein 23 5.13 10.80 3 CG5550 defense response 5.10 8.30 2 CG14847 1.84 0.95 3 CG4919 Glutamate-cysteine ligase modifier 1.82 1.00 2 CG 1399 Leucine-rich repeat 1.77 1.88 8 CG3234 timeless 1.71 1.13 5 CG32041 Heat shock gene 67Bb 1.69 4.63 5 CG9568 1.67 2.63 2 CG4312 Metallothionein B 1.59 0.00 5 -------~--~ ~-- 22 CG5059 L56 2.35 4 CG10725 chitin metabolism 1.53 0.00 2 CG 10827 Alkaline phosphatase 0.63 0.94 4 CGl941 0.60 0.99 2 CG17876 Amylase distal Amy-d 0.55 0.00 3 p53-dependent transcripts CG8127 Ecdysone-induced protein 75B 1.53 2.27 12 CG30069 cell proliferation 1.80 0.94 8 CG7720 cation transport 0.54 0.91 5 CG 10827 mesoderm development 0.63 1.03 4 CG32972 cell adhesion 1.84 0.00 3 CG 10580 fringe 1.82 0.00 3 CG13503 actin filament organization 1.73 0.00 3 CG31319 RhoGAP88C 1.71 0.00 3 CG 12602 cation transport 0.42 0.63 3 CG5246 proteolysis Peptidase 81 1.93 0.00 2 CG18402 Insulin-like receptor InR 1.91 1.30 2 CG6437 GlcT-l 1.68 1.11 2 CG1941 0.60 0.00 2 CG 10129 nudel Toll signaling pathway 0.26 0.00 2 23 A .ll:l)h ..<. ;;Lw r~<. Lo",~\ ~tr(' -;-:l""'~ ;rt,o, .:":':.' ~ ;,. . :-":l.' ~:.:o:... :-: ..•.t;. ,'-XI, .•;"\7,t n.~. ,";1:.' {N,... I"XI. :.J.; ""r.. : ~. to", ." . ..~I~··'n...t'I.P::··': .:':·':""J,·IJJ>J,J.l;/.~ r~:·"""\I:';";I.·.~·".:~~·.J;J...·~'.·.:·".'. - ·:';:"~v..\;;·tC' ·.t""7ia,lMI·"_'J;"~M ;(;I·.M.\;t::' ·,~,,:-~ /..·.':·J..O.....'., :t\·)J\.\::'·... :·.v.I·f.~ '.A~U.I..~.~· .~~·A"'·.: ll.", ':. :U'~'lA: tt I ..•••.!:;.'.\ :-:.1..J.....-;- rt " r~·J<.·., ~;"l ..Jl,,'r;:; h '" '" - ("'i':!:~... :~ ·1"l:'~.U•• :l.1.';\:-.•~ ... ::.l.:"'....o;-:; :·~~: , .(}.M·~':":.A .(',":-"-":':\': : V p .....·l"cn·... :::.::1.::: ..•:I.'.~,·_,:::":"~... ::Io~-: r;,." =re.7:' ... ':..' .:,,~ ~:: ..~ __ l";-to:n:_'~.'.rl.·.· .. ~ "'c :::t~,"':';:::I.J.r:.·x.'V.·.·.;·.·.:..·.: ;;'.'·'.7...: O'-:.·.·.{·.·N,·r."..;; :::r.·.:~·::(,· r.ru:J.-;I.:::.: ~-:;;,••\7'l;';.,\.:.J.r.\V; 1::1;:;. r ". ;':.1. \,1, \1.·.r:_~ II ;,. ~:t 1:\.... :U•••t,'1.')''''~ 1, ':tA ::. ... fI.1ti.l>." ., \. :~~. ~'X'" (':.I.G.'~ ::, ,-J, ... .~~ ~ \..... : ;-. t ( : ~:. ~ ::- r ~. r.·. ::;l.' l. :.'\ L·~• ....:;. -- _•. B .... -. Hsf-RA lill:r. "It!. ~-;:; :t~,~••J.":'.. :-:-:;1.'."1.'".,1.'.;-'.,, ..~:f ~-:-;t: ......" -:':":\....'7-:.'I,••~;~., .. tH~ '.7:'\:",' ,' '7'\.\ .. ::. '''',1.1..' '. ~ ,.1 -;::1,,: Il}.·.'~. ~.·.... -;-::..u~IJo7 : :- I c Hypoxill ... HIF-lII.RNAi HypoxiaNonnoxia B Figure 2. The I-Isfgene region has conserved I-IREs and is a direct target of I-lIF-I. A) An alignment of the nucleotide sequence ofthe second intron the H.~lgene in eight Drosophila species shows two HREs are fully conserved. B) The genomic sequence containing the two HREs in the H4intron were cloned up- stream of the minimal promoter of the Green ]-I Pel ican reporter vector. This construct was transfected into [(c I C>7 cells split into three conditions: normoxia, hypoxia or hypoxia with H1F-1 a RNAi. The reporter was not activated by normox ia but hypox ia induced expression of the GFP reporter. The hypoxia activation of the reporter was eliminated by the addition of 1-1 IF- Ia RNAi. C) Fluorescence was measured using an ISS PC I spectrofluorometer and normalized by cell number. Quantification confirmed a significant increase in fluorescence during hypoxia and a significant decrease from the hypoxic induction when HIF-l a RNAi was added to hypoxic cells. D) Chromatin immunoprecipitation and PCR showing enrichment of the genomic region contain ing the two f-l REs within the Hsfgene in epitope-tagged HIF-I a transfected versus untransfected KC167 cells. The Hph gene and AClin5c genes were used as positive and negative controls respectively. 36 were not completely eliminated in hypoxic cells treated with HIF-1u dsRNA, presumably because the hypoxic stress activated the basal (normoxic) levels of Hsf protein already present in the cells. No HREs were found near any of the Hsp genes, therefore it is unlikely that HIF-1 was directly up-regulating these genes during hypoxia. We tested if the up-regulation of Hsps during hypoxia was dependent on Hsf. Cells were treated with control or HsfRNAi and placed in normoxic and hypoxic conditions. When Hsf was removed through RNAi, Hsp transcripts were eliminated completely, compared to the strong induction seen in cells treated with control dsRNA (Fig.3B). Real-time peR was used to more accurately quantify the results from both of the RNAi experiments. HIF-lu RNAi reduced the up-regulation ofHsps during hypoxia, yet Hsf RNAi completely removed Hsp transcripts (Fig. 3C). From these results, we can discern that Hsfregulates Hsps, while HIF-l regulates Hsf The lack of strong Hsp up-regulation in hypoxic HIF-1 knockdown cells suggests that the HIF-l-mediated increase in Hsftranscript levels is an important step in regulating the sensitivity and activity of the heat shock response pathway. The functional impact of an increase in Hsf transcript levels in hypoxia was tested by assaying the response to hypoxia of a fly heterozygous for the null Hsj mutation (59), and therefore containing only a single wild-type copy of Hsf After exposure to hypoxia, these flies had reduced levels of Hsftranscripts compared to wild-type Oregon R flies as measured by real-time PCR (Fig. 4). The heterozygous flies with a reduction in Hsftranscripts also showed a strong reduction in Hsp26, Hsp27 and Hsp68 transcript levels compared to the control flies, although two Hsp 70 genes had normal levels of induction. 37 A Hypoxia + ... HIF· Ia. RNAi + HIF-1. -:9 .~ 20% '0 ~0 0% Hsf Hsp26 Hsp27 Hsp68 Hsp70Ab Hsp70Ba Figure 4. L'p-rcguiation ofJr'Ps is Hsf dosage..dependent. Re,1!-time peR of flies heterozygous for a null Hs/mutation shov, a sigwfk,mt cC'du.;tion ill }h[ Hsp26, lIsp2} and Hsp68 transcript abundance compared to wild-type flies ;tIter hypoxia \1' < 0.05). Tris demonstrates that l-hftranscript abundance is critical to the magnituck of Hsr 1" o\luct'on. Standard ern,!, of the mean is shown. These findings suggest that Hsf abundance impacts the up-regulation of some Hsps in a dose-dependent manner during hypoxia. Lower Hsf transcript abundance than the levels normally achieved during hypoxia are insufficient for the full up-regulation of Hsps. Full inducTion of/i<;ps und viability during reoxygenatioll is dependent on increased Hsf levels During the return. to normal oxygen conditions., Hsp levels remain high and are critical to tissue survival during this reoxygt~natioll (60) (61). The effect of the HIF-I- dependent increase in Hsf level on I-lsI' expression persists during reoxygenation. KC167 39 tissue culture cells \vith HIF- i a knocked down by RNAi had little increase 'in Hsp expression after hypoxia treatment a..1'J.d a reoxygenation period (Fig. 5A). Thus, the up- regulation of lisjduring hypoxia is critical to the high levels of Hsp transcripts during reoxygenation, as well as hypoxia. Furthermore, we examined the functional importance in vivo of increased Hsf transcript abundance by assaying larval survival under hypoxia and reoxygenation stress. First instal' larvae were reared in a regime of alternating hypoxia and reoxygenation. The lis! heterozygotes had greatly reduced survival compared to larvae reared in normoxia (Fig. 5B). Control wild-type larvae showed no significant difference in survival between normoxia and the hypoxia and reoxygenation environments. These findings demonstrate the dosage importance of HsE transcript levels for coping with hypoxia and reoxygenation at the organismalleveL 40 Hsp27 ....•............•. Actin5c ........ wt Hsf +/- 0% 100% 80% co > -;:; 60% ... ::J (j) co > ... ~ 40% ;:R0 20% B + + + .... Hsp68 Reoxygenation HIF-1a RNAi r----------.., Hsf Hsp26 A Figure 5. Up-regulation of Hsps atter reoxygenation is HIF-la-dependent and critical to survival. A) RT- peR of transcripts involved in the heat shock pathway are up-regulated after hypoxia. HIF-la RNAi reduces the increase in Hsp transcripts. Actin5c used as a control. B) Larvae reared in either normoxia Or hypoxia with a reoxygenation period each day were allowed to develop into pupae. Development of wild- type larvae was minimally affected by the hypoxic and reoxygenation stress. However, half as many Hsf- larvae reached the pupal stage when faced with repetitive hypoxia and reoxygenation compared to normoxic larvae_ The reductiOn in survival was significant (p < 0.05). Standard error of the mean is shown. 41 Taken together these experiments show the sequential order and importance of the hypoxia response. During hypoxia, HIF-I directly up-regulates Hsj, which in tum up- regulates the whole family of Hsps. Without the HIF-I regulated increase in Hsf, Hsps transcript levels never reach full induction during hypoxia or reoxygenation, and organismal viability is reduced. Discussion Up-regulation of Hsps during hypoxia is part of the canonical low-oxygen stress response seen in Drosophila (16), C. elegans (47), and mammalian tissues (62). This study provides evidence that the up-regulation of Hsfduring hypoxia surprisingly requires the activity of HIF-1, the effector of the low oxygen response. The transcriptional control of Hsfby HIF-1 has a functional impact on the activity of the heat shock response during hypoxia and the return to normal oxygen levels. Cells lacking HIF-lor with reduced dosage of Hsf only increase Hsp transcript production slightly during low oxygen exposure and reoxygenation. The decreased production of Hsps reduces viability in flies experiencing hypoxia and reoxygenation, demonstrating that the full induction of the heat shock response is essential to counter the diverse physiological stresses associated with low oxygen. Thus, we propose a model where HIF-l directly up-regulates Hsfduring hypoxia, and the increased Hsf abundance in tum allows Hsfto further up-regulate Hsps during low oxygen exposure and also after the return to normal oxygen levels. The regulation of Hsfby HIF-l provides a clear example of how cross-regulation between physiological 42 stress response pathways can allow one pathway to sensitize the second and elicit a response under conditions where normally it would not be activated. Complex regulation o/physiological re.sponse pathways Cross-regulation bct\veen physiological pathways appears to be a feature of the low oxygen response. It has been shown that the insulin pathway can dramatically affect the HIF-l pathway (63). Through the actions of the phosphatidylinositol 3-kinase/Akt pathway, HIF-Io. translation is increased in a manner that outpaces the naturally normoxic degradation ofHIF-la (64). This leads to HIF-l activation even when oxygen is present and up-regulating its downstream targets. Recently, it has been shown that transforming gro~ih factor-(31 activates the HTF-l pathway by reducing the levels of prolyl hydroxylases that tag HIF-IO'. fi)r degradation. Interestingly, it is also known that Hsp90 plays are role in stabilizing H1.F-la (65) (66). This mechanism is independent of the canonical oxygen-dependent regulation of HIF-1 a and was the first evidence of any link between the heat shock and hypoxia stress pathways. The cross-regulation between HIF-1 and Hsf found here is a new type of control, where the transcriptional eilector of the low oxygen response directly regulates the transcript level of the effector of the heat shock response in order to sensitize the pathway. Interestingly, it has been already shown that HIF-l and Hsfpathways have regulatory interactions, hut in response to heat. Studies using C. elegans and rats showed that HLF··J aCiivity was essential for heat acclimation (67) (68). Our findings may explain the mechanism behind this phenomenon, in that the increase in metabolic activity 43 during high temperature may cause oxygen scarcity, thus stabilizing HIF-I and increasing Hsf transcript levels. Transcriptional control o.lthe heat shock response The activity of the heat shock pathway has been shown to be controlled by the trimerization and post-translationalmoditication of Hsfprotein subunits (51). Our results indicate that transcriptional control of Hslis a means of further regulation of heat shock pathway activity. This transcriptional regulatory step is controlled by HIF-1, supporting a model in which the HIF-l pathway causes increased H.sftranscription during hypoxia as a means to increase the cellular abundance of Hsf and increase the sensitivity of the heat shock pathvvay. In addition. the control of heat shock response sensitivity by HIF-I, the regulator of the low oxygen response. suggests that strcss response pathways can assimilate complex new fWlctions by regulating the transcriptional activators of other stress pathways, Disease implications It has been shmvn that the increase in Hsp levels are critical for cell survival during hypoxia and the subsequent reoxygenation (61) (69). Our results indicate that it is through the fHF-l padmay that tht~ cell achieves this IIsp increase and is a means to protect against tnr stress of hypoxia. HIF-I ac(;umulation and activity have been linked to tumor progression and various Hsps have also been shown to be crucial to cancer survival (70), thus, the hypoxic and beat shock response pathways play important roles in the pathophysIOlogy of cancer. Our finding that the activity ofHIF-I controls the output 44 of the heat shock pathway offers possible therapeutic approaches for mitigating hypoxic tissue damage and tumor growth by targeting this novel regulatory link. Methods Cell culture and hypoxia treatments Drosophila melanogaster KC167 tissue culture cells were obtained from the Drosophila Genomics Resource Center. Cells were maintained in Schneider's Drosophila Medium (Gibco), supplemented with 5% heat-inactivated Fetal Bovine Serum (Gibco). For hypoxia experiments, cells were incubated for 6 hours in chambers flushed with 0.5% 02 gas. The reoxygenation step consisted of a 15 minute return to normal oxygen levels. RNA interference RNAi was performed as previously reported (71). The following primer pairs were used to generate template DNA: control Green Fluorescent Protein (GFP) (5'- GCCACAAGTTCAGCGTGTCC and 5'-GCTTCTCGTTGGGGTCTTTC), HIF-la (sima) (5'-CTGCGGGACTATCATAACAACC and 5'- AGGCTCAAAATCAATCTTTTGG), alternate HIF-la (5'- GCATCACATCAAAGAGTCCCGAG and 5'-TCCGCAACCGTAACACCACTAC) and Hsf (5'-TGCCAAACAGTCCGCCTTATTAC and 5'-TGCTTTCCAAGTGCCCGTG). The T7 promoter sequence (5'-TAATACGACTCACTATAGGGAGA) was added 5' to all above primers when ordered (IDT). 45 Reverse Transcription-PCR Total RNA was isolated using standard Trizol protocols. Instructions from Superscript III One-Step RT-PCR System with Platinum Taq (Invitrogen) were followed using 1 !lg of total RNA and 21 cycles of amplification were used for each test. The following primer pairs were used: HIF-1u (5'-CGAACTCGGTACTAAAGAACCTGC and 5'-GGGTCCTACTTTCACGCAAGG), Hsf (5'-ATCTGCTGCGTGGCGATG and 5'-CGTCCGTGTCCAAAATGTCG), Hsp26 (5'-ATGGCGTGCTCACCGTCAGTATTC and 5'-GGATGATGTTGGATGATGATGGCTC), Hsp27 (5'- AGGAGGAAGAAGACGACGAGATTCG and 5'- CATTGGGTGTGTTGTGGTGTGTCC), Hsp68 (5'- TTCACCACCTATGCCGACAACCAG and 5'- TCACATTCAGGATACCGTTTGCGTC), Hsp70Ab (5'- TCCATTCAGGTGTATGAGGGCG and 5'-CGTTCAGGATTCCATTGGCGTC), Hsp70Ba (5'-ACGATGCCAAGATGGACAAGGG and 5'- CGTCTGGGTTGATGGATAGGTTGAG) and Actin5c (5'- GGATGGTCTTGATTCTGCTGG and 5'-AGGTGGTTCCGCTCTTTTC). Real-time PCR Total RNA was isolated using standard Trizol protocols. cDNA was synthesized following the SuperScript III Reverse Transcriptase protocol (Invitrogen). Real-time PCR was performed using the Sybr Green PCR Master Mix (Applied Biosystems) and an 46 ABI PRISM 7900HT detection system (Applied Biosystems). The supplied analysis software was used for data interpretation. The following primer pairs were used: Hsf (5'- ACACCGCAGCCTCACATTATGACC and 5'... ATTTCCCTGGAGCAGCAAGTCCTC), Hsp27 (5'- AGGAGGAAGAAGACGACGAGATTCG and 5'- CATTGGGTGTGTTGTGGTGTGTCC), Hsp68 (5'- TTCACCACCTATGCCGACAACCAG and 5'- TCACATTCAGGATACCGTTTGCGTC), Hsp70Ab (5'- TCCATTCAGGTGTATGAGGGCG and 5'-CGTTCAGGATTCCATTGGCGTC), Hsp70Ba (5'-ACGATGCCAAGATGGACAAGGG and 5'- CGTCTGGGTTGATGGATAGGTTGAG) and Actin5c (5'~ TGCTGGAGGAGGAGGAGGAGAAGTC and 5'- GCAGGTGGTTCGCTCTTTTCATe). Hypoxia reporter construction A small region of the Hsfintron containing the possible hypoxia regulatory elements was cloned into the Green H Pelican reporter vector. KC167 cells were transfected using the Effectene kit (Qiagen) with this reporter and put in normoxia, hypoxia or hypoxia with HIF-la R.~Ai. Images were taken using a Nikon Eclipse TE2000-U microscope and MetaVue image capture software. 47 Chromatin immunoprecipitation KCl67 cells were transfected using the Effectene kit (Qiagen) with a pAc5.1/Sima plasmid, which contains the full-length HIF-Iu (sima) cDNA sequence with a c-terminal V5 epitope tag under the control of the Actin5c promoter. An equal quantity of mock transfected cells was used as a control and all purification steps were carried out in parallel with the control and experimental cells. 24 hours after transfection, control and experimental cells were incubated in hypoxia at room temperature for 24 hours. DNA isolation and purification procedures followed standard V5 protocol. PCR was used to detect Hsj, Hph, and Actin5c genomic regions in each of the samples of isolated DNA. 35 cycles were used to amplify 2 fll of template from each sample using the following primers: Hsf(5/-CTCCCACCACATACCGCTAATC and 5/- AAAAGCCAACTGAATGACCAAGG), Hph (5/-CCTTCTCACACTCCCTTCGCTG and 5/-CACTCTCTGCCAAGCCAAACC), Actin5c (5'- TGTGTGTGAGAGAGCGAAAGCC and 5'-CTGGAATAAACCGACTGAAAGTGG). Larval survival First instar larvae of wild-type or flies with only one copy of the Hsfgene were counted and placed 10 per vial of food. These vials were split and placed into groups for normoxic or hypoxia and reoxygenation stresses. This experimental group was maintained at 0.5% oxygen for 23 hours, then placed in ambient oxygen for one hour, before returning to hypoxia. This cycle was repeated daily and after two weeks the vials were scored for survival. 48 Bridge to Chapter IV By identifying HSF as a target of HIF-1, we described a mechanism by which hypoxia activates a transcriptional response more commonly associated with heat stress. In Chapter IV, more cross talk between stress pathways is reported. Here, we showed that hypercapnia causes the repression ofNF-KB-dependent antimicrobial gene expression. -~~----------------- 49 CHAPTER IV ELEVATED C02 SUPPRESSES SPECIFIC DROSOPHILA INNATE IMMUNE RESPONSES DOWNSTREAM OF NF-kB PROTEOLYTIC ACTIVATION Reproduced from LT. Helenius·, T. Krupinski·, D.W. Turnbull, Y. Gruenbaum, N. Silverman, E.A. Johnson, P.H.S. Sporn, 1.1. Sznajder, G.J. Beitel. Submitted to Journal of Experimental.Mediczne, 2008 *Authors contributed equally to this work I performed the experiments and analyzed the data for the gene expression microarray results described in this paper. I also helped with interpretation of the data and the design of subsequent experiments based on it. Although an average human produces 450 liters of CO2 per day (72) and elevated levels of CO2 (hypercapnia) in the pulmonary and/or circulatory system are associated with worse outcomes in chronic obstructive pulmonary disease (COPD) (73) (74), the cellular pathways that sense and respond to C02 are poorly understood (reviewed in (75) (76») and a comprehensive description of a CO2 response pathway is lacking in any system. We sought to develop a genetically and molecularly tractable system for defining non-neuronal CO2 response pathways and chose Drosophila because of extensive conservation of the hypoxia (31), nitric oxide (NO) (77) and nearly all other major signal 50 transduction pathways between Drosophila and mammals. Further, Drosophila possess a well-characterized, multi-component innate immune system controlled by conserved signaling pathways that include NF-kB-family transcription factors (78). Genetic and high throughput RNAi studies using whole Hies and S2 cell culture have contributed to understanding the human innate immune system (reviewed in (79)), including identification of Toll, the founding member of the Toll-like-receptor (TLR) family, and most recently of the Akirin proteins that are required for NF-kB-dependent gene expression (80). In this report we show that Drosophila have specific physiological responses to hypercapnia which include a suppression of particular NF-kB-regulated antimicrobial peptides that is remarkably parallel to the suppression of specific NF-kB-regulated innate immune/inflammatory cytokines in human macrophages (see accompanying report by Wang et al. [refj). These results establish Drosophila as a general model for defining non-neuronal C02 signaling pathways and as a specific model for investigating suppression ofNF-kB effectors by hypercapnia. Results and Discussion Hypercapnia causes specific effects on Drosophila physiology independent of known neuronal CO2 sensing pathways. To determine whether hypercapnia affects Drosophila physiology, we exposed Hies to 13% and 19.5% CO2 (~pC02 94 and 140 mmHg), while maintaining 02 at 21% (see Materials and methods). These CO2 levels are well below the ~30% concentration at which CO2 becomes anesthetic. Hypercapnia causes dose-dependent defects in 51 embryonic development. 13% C02 results in 20% of embryos having moderate defects such as malformations of the airway system (compare Fig. IA and B) and also significantly slows hatching of eggs laid by normocapnic mothers (Fig. ID,E). 19.5% CO2 severely disrupts embryonic development with ~30% of exposed embryos showing large-scale patterning and morphogenesis defects (Fig. IC), and more than 70% of eggs failing to hatch (Fig. IE). In contrast, adult Drosophila survive exposure to 13% and 19.5% CO2 for many weeks (data not shown), consistent with the essentially normal postnatal development of mice exposed to 12% CO2 (81). The effects of hypercapnia on vertebrate embryonic development have not been reported, but these results suggest that embryonic and adult metazoans may have different sensitivities to hypercapnia. However, exposure of adult flies to hypercapnia is not without adverse consequences, as hypercapnia causes a dose-dependent reduction in the number of eggs females lay (Fig. 1F). Importantly, flies homozygous for a null mutation in the neuronal CO2 receptor Gr63a (82) (83) are as sensitive to hypercapnia as wild-type flies in the egg hatching and laying assays (Fig. ID-F). Thus, many physiological effects of hypercapnia are mediated by an as yet uncharacterized CO2 response pathway(s). 52 I:IWT _Gr63. Egg laying Blot anti.AT!'" air 13% 19.5% lC02l ATP" endocytosis in S2 cellsGElI9 hatching (48hj • I::::IWT -.sr63a Egg hatcbing 124nj F air 13% 195% [C~l D 1 E Figure 1. Hypercapnia affects Drosophila development and physiology independently of neuronal CO2 sensing. (A-C) Drosophila embryonic development is disrupted by hypercapnia. The embryonic tracheal (airway) system, visualized by a lumenal stain, is an excellent readout of general morphogenesis because tracheal branch guidance requires interaction with multiple tissues. Culturing wild-type embryos in 13% CO2 causes moderate defects in 20% of embryos (B, n=111), while 19.5% CO2 causes severe defects in 28% of embryos (C, n=71) and increases the prevalence of moderate defects to 41%. Moderate defects (B) include breaks in the main airways (arrowheads) and missing or ectopic interconnections of dorsal branches (arrows), while severely abnormal tracheal development reveals gross embryonic patterning and/or morphogenesis defects (C). (D,E) Hypercapnia (24h, D) decreases hatching of eggs laid in normocapnia by wild-type (WT) or mutant flies homozygous for a null allele in the neuronal CO2 receptor, Gr63a. In 13% CO2 over 90% of WT embryos hatch after 48h (E), but in 19.5% CO2 only ~30% hatch. (F) Hypercapnia reduces the number of eggs laid in 48h by WT and Gr63a females mated in normocapnia. (G) As in mammalian cells hypercapnia (lh) causes endocytosis of the Na,K-ATPase in 82 cells. * = p < 0.05, ** = P < 0.005 between CO2 condition and air; error bars are standard deviation. We next asked whether Drosophila share with mammals any specific response to hypercapnia. We previously reported that in human alveolar epithelial cells, hypercapnia causes endocytosis of the Na,K-ATPase (84). Analysis of surface abundance of Na,K- ATPase in Drosophila 82 cells reveals that hypercapnia also causes dose-dependent endocytosis of the sodium pump in 82 cells (Fig. 1G). As Gr63a is expressed only in specific CO2-sensing olfactory neurons (82), the responsiveness of the hematopoietic 82 cell line to hypercapnia further supports the existence of cell autonomous C02 responses that do not depend on the neuronal Gr63a sensing pathway. That both Drosophila and 53 human cells endocytose their Na,K-ATPase in response to hypercapnia suggests that some C02 responses are conserved between mammals and flies. Hypercapnia causes specific effects on gene expression To investigate the molecular basis of the physiological effects of hypercapnia and to identify CO2-responsive promoters to use as markers for dissecting C02 signaling pathways, we next performed microarray analysis on adult flies. Similar to the limited changes in gene expression seen in neonatal mice raised in CO2 (81), exposure of adult flies to 24h of 13% CO2 causes fewer than 500 genes to be up- or down-regulated more than 1.5-fold, and fewer than 10 by more than 10-fold (data not shown). Therefore, in Drosophila CO2 causes specific rather than global changes in gene expression. Consistent with the dramatic decrease in fecundity of female flies during hypercapnia, the chorion and vitelline membrane genes required for egg production are among the genes most highly down-regulated by elevated C02 (Table Sl). Notably, C02 does not induce genes characteristic of hypoxic, heat shock or oxidative stress responses (Table Sl), indicating that the responses to elevated C02 are mediated by distinct pathways. Most intriguingly, a subset of the antimicrobial peptide (AMP) genes that are effectors of the Drosophila innate immune system are down-regulated by elevated C02 levels (Fig. 2A). Given the prevaLence of pulmonary infections in hypercapnic COPD patients (85), we focused on defining the effects of hypercapnia on the highly conserved NF-kB-regulated innate immune system of Drosophila and determining whether hypercapnia could render a host more susceptible to infection by a pathogen. 54 E 8' 2. Mult fly q?CR ;! ~;- 1. " m - Qi i 1. ~~ .~ ,., 0.• n. ~ 0 0.11 i eM o StaphyIOCf)CCUS 8ureus 19r11Wtll1n l.!IJ GEnterococcus faecafis (grOWt!lIll1.6) ••" )V rr'Il Q I"SOO, //)7.... .>rt .;/;" ::: •.A.C'-f:::: Y llI,t& ..ner VlOCJ\a~OO StaphylocCK;CIIS aUfQUS (P"U.00811 E Figure 3. Hypercapnia increases mortality of bacterial infection in Drosophila. (A-D) Hypercapnia does not increase mortality of flies inoculated with PBS (A) or with E. coli (B), which is not a Drosophila pathogen, but does significantly increase mortality after inoculation with the Drosophila natural pathogen E. faecalis (C), or the human opportunistic pathogen S. aureus (D) at CO2 levels as low as 7%. Unless otherwise noted, flies were exposed to indicated CO2 level for 24h before inoculation and returned to hypercapnia until end of assay. We show representative results for the lowest CO2 levels at which significant effects on mortality were consistently observed. All experiments were done in triplicate and the trial with the middle p-value is shown. (E) Pre-treatment of flies with 9% CO2 prior to S. aureus infection in air is sufficient to increase mortality even when flies are cultured in air during and after inoculation. (F,G) Effects of hypercapnia on bacterial growth. Note that S. aureus growth is dramatically reduced in 7% CO2 even though hypercapnia increases mortality of flies infected with S. aureus. Error bars smaller than the data-point symbols are not shown. (A-E) Kaplan-Meier survival curves, p-value determined by log-rank analysis. ---------------- 57 Although the above results demonstrate that hypercapnia can decrease survival during infection, they do not distinguish whether decreased survival results from suppression of host defense, from enhanced bacterial pathogenicity, or both. We definitively demonstrate that increased mortality results at least in part from direct effects of CO2 on the fly immune system by exposing flies to 9% CO2 for 24h and returning them to normocapnia after inoculation with S. aureus. This hypercapnic pre-treatment of the host significantly increases mortality after pathogen infection although the pathogen was never exposed to hypercapnia (Fig. 3E). We also investigated whether hypercapnia alters bacterial growth rates. Growth of E. faecalis in media is not significantly affected by elevated CO2 (Fig. 3F). However, despite the markedly increased mortality of S. aureus-infected flies, growth of S. aureus is actually reduced several-fold by even 7% CO2 (Fig. 3G). As decreased bacterial growth would have been expected to improve fly survival, the observed increase in mortality further supports the conclusion that hypercapnia suppresses Drosophila innate immune responses. Importantly, the decreased growth rates of some bacteria in elevated CO2 may provide an explanation for the existence of hypercapnic immune suppression. Because hyperactivation of immune responses in Drosophila or humans can cause pathology beyond that of an infection alone, immune responses may be deliberately attenuated during hypercapnia to prevent overreaction when pathogen growth is reduced. The observed increase in mortality during hypercapnia could result from an imperfect match between the degree of immune suppression and changes in bacterial growth. Another, and not exclusive, possibility is that CO2 serves as a readout of metabolic activity with 58 elevated C02 indicating excessive metabolic load. Thus, hypercapnia may suppress select metabolically demanding functions including immune responses (89) and egg-laying to allocate energy to more immediate needs such as powering the flight muscles, which are among the most metabolically active tissues known. Hypercapnia suppresses, not delays, innate immune responses To further investigate the nature of hypercapnic immune suppression, we determined how hypercapnia affects the kinetics of innate immune responses. Hypercapnia does not alter the timing of AMP induction in PGN-challenged S2* cells, but the magnitude of the responses at any time is reduced (Fig. 4A). Thus, hypercapnia suppresses rather than delays innate immune responses. In S2* cells, this suppression exhibits rapid onset and recovery, as shifting cells from air to 13% C02 simultaneous with PGN challenge causes a two-fold reduction in Dpt induction, and increasing the pretreatment time beyond 5h does not further increase the suppression (Fig. 4B). Conversely, CO2-treated cells challenged upon shift to normocapnia have normal induction of Dpt (Fig. 4C). The effects of hypercapnia on AMP induction in S2* cells closely parallel the effects on cytokines in human macrophages as detailed in the accompanying report by Wang et al [ref], supporting the existence of a conserved molecular mechanism of hypercapnic . .Immune suppreSSIOn. A S2"ceUs (PGN·cbattenged) -.-air -Ir13% C02 59 1 3 5 Hours of PGN challenge 1 B 52" celts {PGW-cball@nged) iij U) sa III 'i en .\It' ~ _ \)I\!'ill~ ~~ "'co. ~ c i .si 1.004..... ~..... 52" cells fPGN-cbalie ng@d) Figure 4. Hypercapnia suppresses antimicrobial peptide induction with rapid onset and recovery. (A) Hypercapnia suppresses, not delays, antimicrobial peptide (AMP) induction in vitro as assayed by qPCR of Diptericin (Dpt) in PGN-challenged S2* cells. (B) qPCR of Dpt in PGN-challenged S2* cells shows that the onset of immune suppressive effects of hypercapnia is rapid as shifting cells to hypercapnia at the time of PGN challenge results in ~50% suppression of Dpt with maximal suppression reached by 5h. (C, Left) Recovery from immune suppression is also rapid as challenging S2* cells with PGN in air immediately after removal from CO2 results in full induction of Dpt. (C, right) As in unchallenged flies and S2* cells, Drosomycin is not strongly regulated by hypercapnia. * = p < 0.05, ** = P < 0.005; error bars are standard deviation. 60 PGN 110 O,~ll 1Ii no 0511 111 , "," r...;,;,,:,.-.:.:..:...;--'-"-----------.k-I50kD tjrp-R~. I I (~"oll'","' i S:ICO: :> ... ... ... ... ( ) ...... ( ) ... 336 kb Figure 3. Graphical representation of sequences from regionally captured samples aligned to a portion of the C. e/egans genome. The area of chromosome V that was targeted for capture is indicated in red bordered by blue Iines. Black bars beneath the zoomed in area of chromosome V indicate sequence reads. 31% of the total sequence reads that aligned uniquely to the genome were within the target region. Other reads were scattered randomly throughout the genome. Discussion One major drawback of NOS approaches is that they do not provide an easy means of sequencing specific genomic areas since the sequencing reaction does not proceed from a fixed gene-specific primer as it does in Sanger sequencing. Because of this limitation, a method for capturing specific genomic regions for NOS by DNA microarray hybridization was previously developed. Here, I have described an alternative to the array-based capture method, which is cheaper, faster, and easier to perform. 82 I have demonstrated the effectiveness of this protocol by capturing and sequencing a 336 kb region of the C. elegans genome from a mutant strain isolated in a genetic screen. Despite the fact that a large number of sequences obtained were from outside the region of interest, the enrichment was sufficient to sequence the mutation- containing region with an average 24x coverage. The data from the sequencing run was obtained quite recently, so we have only begun to analyze the data. In fairly short order, we will be able to visualize SNPs between the obtained aligned sequences and the C. elegans reference genome, which should make identification ofthe SNP responsible for the mutant phenotype quite easy. This regional capture technique has applications beyond the characterization of mutants from genetic screens in model organisms. Using a fraction of a lane in a GA-II NGS instrument, my protocol was able to provide 24x coverage of the targeted region. This depth of coverage would have increased substantially if the entire lane had been dedicated to sequencing the captured region. With increased depth of coverage, the captured region could be assembled without the aid of a reference genome, and this protocol would be a viable method of capturing and sequencing genomic regions of non- model organisms with incompletely sequenced genomes. A variation on this protocol is currently being evaluated as a means of mapping transposon insertion sites in maize. Indeed, researchers should be able to make use of this protocol, or a modification of it, for any situation in which they would like to enrich for a specific genomic location for NGS. 83 Bridge to Chapter VI In Chapter V I described a technique I have developed for sequencing specific DNA regions using NGS. This protocol is considerably faster and less expensive than existing techniques. In Chapter VI I will summarize the findings from Chapters II, III, IV, and V, and highlight the broader significance of these results 84 CHAPTER VI CONCLUSION Stress-responsive gene regulation A key adaptation organisms make in order to survive physiological stresses is the differential regulation of gene expression. This regulation is directly carried out by stress-responsive transcription factors. One physiological stress that all eukaryotic organisms must cope with is hypoxia. In Chapter II I have identified genes that are regulated by hypoxia, and have determined the roles of five different hypoxia-responsive transcription factors in the regulation of this set of hypoxia-modulated transcripts. This was achieved by examining hypoxic gene expression in D. meianogaster mutants lacking the hypoxia-responsive transcription factors HIF-l, FOXO, Rei, p53, and MTF-l , and by refining the lists of affected transcripts using bioinformatic analysis. I have identified several previously characterized target genes of the five hypoxia-responsive transcription factors, as well as genes that have never been previously described as targets. I have also identified a mechanism by which HIF-l can contribute to the activation of FOXO during hypoxia through the HIF-l-dependent transcriptional up regulation of Imp-L2. Finally, I 85 have determined that FOXO plays a much larger role in hypoxic gene regulation than has previously been thought. One set of genes that are highly transcriptionally up regulated in response to hypoxia are those encoding heat shock proteins. In Chapter III, we showed that HIF-l plays a significant role in the transcriptional up regulation of Heat shock protein (HSP) genes during hypoxia. This occurs because HIF-l directly up regulates the transcription of the Heat shock factor (HSF) gene during hypoxia, and this up regulation is necessary for the full transcriptional induction of HSP's. We showed that this HIF-I-dependent up- regulation of HSF during hypoxia is functionally significant because flies with reduced HSF levels during hypoxia are less viable than wild-type flies. Exposure to elevated CO2 levels (hypercapnia) is another stress that eukaryotes adapt to by modulating gene expression. In Chapter IV, we identified genes that are differentially regulated during hypercapnia in Drosophila. We found that the transcription of immune response genes that are known targets of the transcription factor Relish (ReI) is significantly repressed by hypercapnia. Interestingly, we showed that this transcriptional repression occurs by a pathway that is either downstream or in parallel to the canonical ReI activation pathway. Our results have medical relevance because hypercapnia is frequently experienced by patients suffering from chronic pulmonary obstructive disease and the suppression of immune response genes may promote bacterial infections in patients experiencing this physiological stress. This study also establishes Drosophila as a powerful model system for the study of hypercapnia. 86 Regional capture for next generation sequencing The Sanger method of DNA sequencing is a well-established technology that is widely available to researchers at numerous institutions across the world. However, the amount of sequence data produced per Sanger sequencing run is limited to less than 1000 bases, making sequencing large DNA fragments with this technology costly and slow. To address this limitation, next generation sequencing (NGS) techniques have been developed in recent years that produce much more sequence data per run. The downside of these new approaches is that the sequence reads produced by NGS instruments are not targeted to a specific region of the template DNA making the sequencing of specific regions of complex genomic DNA samples impossible with these techniques. Custom DNA microarray-based regional capture schemes have been developed to facilitate the enrichment of specific DNA regions for NGS, but these approaches are slow and costly. In Chapter V I have described a regional capture protocol that provides enrichment of target DNA sequences that is comparable to the array-based schemes. My approach is considerably faster and less expensive, however because biotinylated fosmid DNA is used in place of custom DNA microarrays as the probe for the hybridization-based capture. I have demonstrated the effectiveness of my protocol by capturing a genomic DNA region from a C. elegans strain that contains a mutation causing SNP. Sequencing this region by NGS provided an average of 24x coverage for every base in the targeted region, which was more than sufficient for identifying the SNP. My regional capture protocol is broadly applicable to any situation in which a researcher desires to enrich for a specific region of a complex DNA sample for sequencing by NGS. 87 APPENDIX ONLINE SUPPLEMENTAL MATERIAL FOR CHAPTER IV ..Air; uninduced IIIlIAir; PGN fZ?JC02; uninduced mC02 ;PGN 30 26 22 18 14 10 < 6z 0::: 0> 2~ 1.25a.. 0::: -a.. ~ 1.00 0.75 0.50 0.25 0.00 Drs Att Opt Cee Def Orm Mtk Figure 81. Absolute levels of antimicrobial peptide RNAs in 82* cells, normalized to RP49. CO2 = 13%, 1Dh duration before RNA extraction; PON = 5h before RNA extraction. This is the same experiment as in Figure 2C and D. AMP abbreviations are as in Figure 2. 88 Table 81. Hypercapnia strongly down-regulates egg fonnation genes, but does not up-regulate genes induced by hypoxic, heat shock, or oxidative stress responses. CG6517 Chorion protein 18 CG6524 Chorion protein 19 CG6533 Chorion protein 16 CG 11213 Chorion protein 38 CG 1778 Chorion protein 36 CG6519 Chorion protein 15 CG2175 Defective chorion 1 CG9271 Vitelline membrane 34Ca CG9048 Vitelline membrane 26Aa CG 15349 Chorion protein a CG15351 Chorion protein c CG9046 Vitelline membrane 26Ab B. H oxia-induced enes§ CG 11765 Peroxiredoxin CG 11949 Coracle CGl2896 CG 10160 ImpL3 CG6494 Hairy h CGl3499 CG32130 CG 10078 Phosphoribosylamidotransferase 2 CGl600 CG8846 Thor Fold induction 13% CO2 C. Heat shock-induced enes~ -42.3 CG16749 -38.5 CG18493 -28.1 CG9470 Metallothionein A -26.5 CG3106 -20 CG6730 Cyp4d21 -lOA CG1946 -9.8 CG9080 -7.2 CG6602 -5.9 CG3360 Cyp313al CG13833 (up 2.9-fold in heat -5.8 shock) -4.8 -4 D. Oxidative stress-induced enest -1.2 CG4533 1(2)efl -1. 3 CG5164 GstEl 1.1 CG 10794 Diptericin B 1 CG31628 ade3 -1.3 CG18466 Nmdmc -1.1 CG7632 -1.2 CGI0146 Attacin A -1.6 CG8772 nemy 1.1 CG 14121 1.2 CG1383 Defensin Fold induction 1 1.1 1 -1 -1.2 1 -1.1 -1.1 -1.1 1.1 -1.2 1 -104 -1 1.1 -1.1 -2.3 1 -1 -2.1 Average fold induction in 13% CO2 (24h, CO2 versus air) of genes specific for egg fonnation (A) and selected genes reported to be highly induced by previously characterized stresses (B-D). Fold induction is the average of 3 separate microarray experiments of 5-day old flies. Stress genes are shown in decreasing order of induction magnitude for particular stresses with maximal and minimal values shown in parentheses. See references below for fold up-regulation by hypoxia§ (range: CG8846 Thor 4.7-fold to CG 11765 Peroxiredoxin 8A-fold), heat shock~ (range: CG13833 2.9-fold to CG 16749 41-fold) and oxidative stresst (range: CG 1383 Defensin 1.6-fold to CG4533 1(2)efl 7-fold). §Hypoxia-induced genes reported by G. Uu et al., Physiol. Genomics 25: 131-141,2006 ~Heat shock-induced genes reported by lG. Sorensen et al., Cell Stress & Chaperones, (2005) 10 (4) 312- 328 tOxidative stress-induced genes reported by G.N. Landis et al., PNAS 2004;101(20):7663-8 Table S2. Hypercapnia does not affect cell viability. Trypan blue dye exclusion assay performed with S2* cells shows that hypercapnia and the other experimental conditions used do not significantly alter cell viability. If indicated, cells were primed for immune response with 20-hydroxyecdysone (ecd) at Oh, placed in 13% CO2 at 15h, and/or stimulated with E. coli PON at 20h. Rapid cell death by strong oxidative stress shown for comparison. t standard deviation of triplicate experiments in parentheses. Time at Experimental condition measurement (h) % cells containing dye t Air 0 6.9 (0.13) Air 23 5.6 (0.74) Air + ecd 15 5.6 (0.82) Air + ecd 23 4.4 (1.07) 13% C02 + ecd 23 5.7 (1.30) air + ecd + PGN 23 5.2 (1.54) 13% C02 + ecd + PGN 23 6.0 (0.98) Air + 100mM t-butyl hydroperoxide 1.5 31 Air + 1M t-but I h dro eroxide 1.5 100 89 90 REFERENCES 1. Semenza, G.L. 2007. Hypoxia-inducible factor 1 (HIF-l) pathway. Sci STKE. 2007:cm8. 2. Kewley, R.J., M.L. Whitelaw, and A. Chapman-Smith. 2004. The mammalian basic helix-loop-helixlPAS family of transcriptional regulators. Int J Biochem Cell Bioi. 36:189-204. 3. Wang, G.L., B.H. Jiang, E.A. Rue, and G.L. Semenza. 1995. Hypoxia-inducible factor 1 is a basic-helix-loop-helix-PAS heterodimer regulated by cellular 02 tension. Proc Natl Acad Sci USA. 92:5510-5514. 4. Stockmann, C., and J. Fandrey. 2006. Hypoxia-induced erythropoietin production: a paradigm for oxygen-regulated gene expression. Clin Exp Pharmacol Physiol. 33:968-979. 5. Centanin, L., P.l. Ratcliffe, and P. Wappner. 2005. Reversion oflethality and growth defects in Fatiga oxygen-sensor mutant flies by loss of hypoxia-inducible factor-alpha/Sima. EMBO Rep. 6:1070-1075. 6. Baird, N.A., D.W. Turnbull, and E.A. Johnson. 2006. Induction of the heat shock pathway during hypoxia requires regulation of heat shock factor by hypoxia- inducible factor-I. J Bioi Chern. 281 :38675-38681. 7. Accili, D., and K.C. Arden. 2004. FoxOs at the crossroads of cellular metabolism, differentiation, and transformation. Cell. 117:421-426. 8. Brunet, A., L.B. Sweeney, l.F. Sturgill, K.F. Chua, P.L. Greer, Y. Lin, H. Tran, S.E. Ross, R. Mostoslavsky, H.Y. Cohen, L.S. Hu, H.L. Cheng, M.P. Jedrychowski, S.P. Gygi, D.A. Sinclair, F.W. Alt, and M.E. Greenberg. 2004. Stress-dependent regulation ofFOXO transcription factors by the SIRTI deacetylase. Science. 303:2011-2015. 9. Bakker, W.J., 1.S. Harris, and T.W. Mak. 2007. FOX03a is activated in response to hypoxic stress and inhibits HIFl-induced apoptosis via regulation ofCITED2. Mol Cell. 28:941-953. 91 10. Chen, F., V. Castranova, X. Shi, and L.M. Demers. 1999. New insights into the role of nuclear factor-kappaB, a ubiquitous transcription factor in the initiation of diseases. Clin Chem. 45:7-17. 11. Minakhina, S., and R. Steward. 2006. Nuclear factor-kappa B pathways in Drosophila. Oncogene. 25:6749-6757. 12. Chandel, N.S., W.C. Trzyna, D.S. McClintock, and P.T. Schumacker. 2000. Role of oxidants in NF-kappa B activation and TNF-alpha gene transcription induced by hypoxia and endotoxin. J Immunol. 165:1013-1021. 13. Koong, A.C., E.Y Chen, and A.J. Giaccia. 1994. Hypoxia causes the activation of nuclear factor kappa B through the phosphorylation of I kappa B alpha on tyrosine residues. Cancer Res. 54:1425-1430. 14. Wingrove, J.A., and P.B. O'Farrell. 1999. Nitric oxide contributes to behavioral, cellular, and developmental responses to low oxygen in Drosophila. Cell. 98:105- 114. 15. Galkin, A., A. Higgs, and S. Moncada. 2007. Nitric oxide and hypoxia. Essays Biochem. 43 :29-42. 16. Liu, G., J. Roy, and E.A. Johnson. 2006. Identification and function of hypoxia- response genes in Drosophila melanogaster. Physiol Genomics. 25: 134-141. 17. Slee, E.A., D.J. O'Connor, and X. Lu. 2004. To die or not to die: how does p53 decide? Oncogene. 23:2809-2818. 18. Chen, D., M. Li, J. Luo, and W. Gu. 2003. Direct interactions between HIF-l alpha and Mdm2 modulate p53 function. J Bioi Chem. 278:13595-13598. 19. Chandel, N.S., M.G. Vander Heiden, C.B. Thompson, and P.T. Schumacker. 2000. Redox regulation ofp53 during hypoxia. Oncogene. 19:3840-3848. 20. Li, Y, T. Kimura, J.B. Laity, and G.K. Andrews. 2006. The zinc-sensing mechanism of mouse MTF-l involves linker peptides between the zinc fingers. Mol Cell Bioi. 26:5580-5587. 21. Zhang, B., O. Georgiev, M. Hagmann, C. Gunes, M. Cramer, P. Faller, M. Vasak, and W. Schaffner. 2003. Activity of metal-responsive transcription factor 1 by toxic heavy metals and H202 in vitro is modulated by metallothionein. Mol Cell Bioi. 23:8471-8485. 22. Green, C.J., P. Lichtlen, N.T. Huynh, M. Yanovsky, K.R. Laderoute, W. Schaffner, and B.J. Murphy. 2001. Placenta growth factor gene expression is induced by 92 hypoxia in fibroblasts: a central role for metal transcription factor-I. Cancer Res. 61 :2696-2703. 23. Murphy, RJ., G.K. Andrews, D. Bittel, D.J. Discher, J. McCue, C.J. Green, M. Yanovsky, A. Giaccia, R.M. Sutherland, K.R. Laderoute, and K.A. Webster. 1999. Activation of metallothionein gene expression by hypoxia involves metal response elements and metal transcription factor-I. Cancer Res. 59:1315-1322. 24. Sanger, F., S. Nicklen, and A.R. Coulson. 1977. DNA sequencing with chain- terminating inhibitors. Proc Nat! Acad Sci USA. 74:5463-5467. 25. Okou, D.T., K.M. Steinberg, C. Middle, D.J. Cutler, T.J. Albert, and M.E. Zwick. 2007. Microarray-based genomic selection for high-throughput resequencing. Nat Methods. 4:907-909. 26. Albert, T.J., M.N. Molla, D.M. Muzny, L. Nazareth, D. Wheeler, X. Song, T.A. Richmond, C.M. Middle, M.J. Rodesch, C.J. Packard, G.M. Weinstock, and R.A. Gibbs. 2007. Direct selection ofhuman genomic loci by microarray hybridization. Nat Methods. 4:903-905. 27. Brown, J.M., and A.J. Giaccia. 1998. The unique physiology of solid tumors: opportunities (and problems) for cancer therapy. Cancer Res. 58: 1408-1416. 28. Hammond, E.M., and A.J. Giaccia. 2005. The role ofp53 in hypoxia-induced apoptosis. Biochem Biophys Res Commun. 331 :718-725. 29. Chung, M.J., C. Hogstrand, and S.J. Lee. 2006. Cytotoxicity of nitric oxide is alleviated by zinc-mediated expression of antioxidant genes. Exp Bio! Med (Maywood). 231:1555-1563. 30. Bacon, N.C., P. Wappner, J.F. O'Rourke, S.M. Bartlett, R Shilo, C.W. Pugh, and P.J. Ratcliffe. 1998. Regulation of the Drosophila bHLH-PAS protein Sima by hypoxia: functional evidence for homology with mammalian HIF-1 alpha. Biochem Biophys Res Commun. 249:811-816. 31. Lavista-Llanos, S., L. Centanin, M. Irisarri, D.M. Russo, J.M. Gleadle, S.N. Bocca, M. Muzzopappa, P.J. Ratcliffe, and P. Wappner. 2002. Control of the hypoxic response in Drosophila melanogaster by the basic helix-loop-helix PAS protein similar. Mo! Cell Bio!. 22:6842-6853. 32. Ishwaran, H., J.S. Rao, and D.R Kogalur. 2006. BAMarraytrade mark: Java software for Bayesian analysis ofvariance for microarray data. BMC Bioinformatics.7:59. 93 33. Honegger, B., M. Galic, K. Kohler, F. Wittwer, W. Brogiolo, E. Hafen, and H. Stocker. 2008. Imp-L2, a putative homolog of vertebrate IGF-binding protein 7, counteracts insulin signaling in Drosophila and is essential for starvation resistance. JBiol.7:10. 34. Murphy, B.J., T. Kimura, B.G. Sato, Y. Shi, and G.K. Andrews. 2008. Metallothionein induction by hypoxia involves cooperative interactions between metal-responsive transcription factor-l and hypoxia-inducible transcription factor- 1alpha. Mol Cancer Res. 6:483-490. 35. Krieg, A.J., E.M. Hammond, and A.J. Giaccia. 2006. Functional analysis ofp53 binding under differential stresses. Mol Cell Bioi. 26:7030-7045. 36. Pear, W.S., and M.C. Simon. 2005. Lasting longer without oxygen: The influence of hypoxia on Notch signaling. Cancer Cell. 8:435-437. 37. Semenza, G.L. 1999. Regulation of mammalian 02 homeostasis by hypoxia- inducible factor 1. Annu Rev Cell Dev Bioi. 15:551-578. 38. Hon, W.C., M.l. Wilson, K. Harlos, T.D. Claridge, C.J. Schofield, C.W. Pugh, P.H. Maxwell, P.J. Ratcliffe, D.l. Stuart, and E.Y. Jones. 2002. Structural basis for the recognition of hydroxyproline in HIF-l alpha by pVHL. Nature. 417:975-978. 39. Huang, L.E., J. Gu, M. Schau, and H.F. Bunn. 1998. Regulation of hypoxia- inducible factor lalpha is mediated by an 02-dependent degradation domain via the ubiquitin-proteasome pathway. Proc Natl Acad Sci USA. 95:7987-7992. 40. Ivan, M., K. Kondo, H. Yang, W. Kim, J. Valiando, M. Ohh, A. Salic, J.M. Asara, W.S. Lane, and W.G.J. Kaelin. 2001. HIFalpha targeted for VHL-mediated destruction by proline hydroxylation: implications for 02 sensing. Science. 292:464-468. 41. Jaakkola, P., D.R. Mole, Y.M. Tian, M.l. Wilson, J. Gielbert, SJ. Gaskell, A. Kriegsheim, H.F. Hebestreit, M. Mukherji, C.J. Schofield, P.H. Maxwell, C.W. Pugh, and P.J. Ratcliffe. 2001. Targeting ofHIF-alpha to the von Hippel-Lindau ubiquitylation complex by 02-regulated prolyl hydroxylation. Science. 292:468- 472. 42. Huang, L.B., and H.F. Bunn. 2003. Hypoxia-inducible factor and its biomedical relevance. J Bioi Chern. 278:19575-19578. 43. Pugh, C.W., and P.J. Ratcliffe. 2003. Regulation of angiogenesis by hypoxia: role of the HIF system. Nat Med. 9:677-684. 94 44. Vogelstein, B., and K.W. Kinzler. 2004. Cancer genes and the pathways they control. Nat Med. 10:789-799. 45. Semenza, G.L. 2002. HIF-l and tumor progression: pathophysiology and therapeutics. Trends Mol Med. 8:S62-7. 46. Pouyssegur, 1., F. Dayan, and N.M. Mazure. 2006. Hypoxia signalling in cancer and approaches to enforce tumour regression. Nature. 441:437-443. 47. Shen, C., D. Nettleton, M. Jiang, S.K. Kim, and J.A. Powell-Coffman. 2005. Roles of the HIF-l hypoxia-inducible factor during hypoxia response in Caenorhabditis elegans. J Biol Chem. 280:20580-20588. 48. Welch, W.J. 1993. Heat shock proteins functioning as molecular chaperones: their roles in normal and stressed cells. Philos Trans R Soc Lond B Biol Sci. 339:327- 333. 49. Westwood, J.T., J. Clos, and C. Wu. 1991. Stress-induced oligomerization and chromosomal relocalization of heat-shock factor. Nature. 353:822-827. 50. Wu, C. 1995. Heat shock transcription factors: structure and regulation. Annu Rev Cell Dev Biol. 11 :441-469. 51. Orosz, A., J. Wisniewski, and C. Wu. 1996. Regulation of Drosophila heat shock factor trimerization: global sequence requirements and independence of nuclear localization. Mol Cell Biol. 16:7018-7030. 52. Pelham, H.R. 1982. A regulatory upstream promoter element in the Drosophila hsp 70 heat-shock gene. Cell. 30:517-528. 53. Nambu, J.R., W. Chen, S. Hu, and S.T. Crews. 1996. The Drosophila melanogaster similar bHLH-PAS gene encodes a protein related to human hypoxia-inducible factor 1 alpha and Drosophila single-minded. Gene. 172:249-254. 54. Semenza, G.L., B.B. Jiang, S.W. Leung, R. Passantino, J.P. Concordet, P. Maire, and A. Giallongo. 1996. Hypoxia response elements in the aldolase A, enolase 1, and lactate dehydrogenase A gene promoters contain essential binding sites for hypoxia-inducible factor 1. J Biol Chem. 271:32529-32537. 55. Aprelikova, 0., G.V. Chandramouli, M. Wood, J.R. Vasselli, 1. Riss, J.K. Maranchie, W.M. Linehan, and 1.C. Barrett. 2004. Regulation ofHIF prolyl hydroxylases by hypoxia-inducible factors. J Cell Biochem. 92:491-501. 56. Bray, N., and L. Pachter. 2004. MAVID: constrained ancestral alignment of multiple sequences. Genome Res. 14:693-699. 95 57. Barolo, S., L.A. Carver, and J.W. Posakony. 2000. GFP and beta-galactosidase transformation vectors for promoter/enhancer analysis in Drosophila. Biotechniques. 29:726, 728, 730, 732. 58. GOIT, T.A., T. Tomita, P. Wappner, and H.F. Bunn. 2004. Regulation ofDrosophila hypoxia-inducible factor (HIF) activity in SL2 cells: identification of a hypoxia- induced variant isoform of the HIFalpha homolog gene similar. J Bioi Chern. 279:36048-36058. 59. Jedlicka, P., M.A. Mortin, and C. Wu. 1997. Multiple functions of Drosophila heat shock transcription factor in vivo. EMBO J. 16:2452-2462. 60. Donnelly, TJ., R.E. Sievers, F.L. Vissem, W.J. Welch, and C.L. Wolfe. 1992. Heat shock protein induction in rat hearts. A role for improved myocardial salvage after ischemia and reperfusion? Circulation. 85:769-778. 61. Kabakov, A.E., K.R. Budagova, A.L. Bryantsev, and D.S. Latchman. 2003. Heat shock protein 70 or heat shock protein 27 overexpressed in human endothelial cells during posthypoxic reoxygenation can protect from delayed apoptosis. Cell Stress Chaperones. 8:335-347. 62. Morimoto, R.1. 1993. Cells in stress: transcriptional activation of heat shock genes. Science. 259:1409-1410. 63. Zelzer, E., Y. Levy, C. Kahana, B.Z. Shilo, M. Rubinstein, and B. Cohen. 1998. Insulin induces transcription of target genes through the hypoxia-inducible factor HIF-1alpha/ARNT. EMBO J. 17:5085-5094. 64. Mottet, D., V. Dumont, Y. Deccache, C. Demazy, N. Ninane, M. Raes, and C. Michiels. 2003. Regulation of hypoxia-inducible factor-1alpha protein level during hypoxic conditions by the phosphatidylinosito13-kinase/Akt/glycogen synthase kinase 3beta pathway in HepG2 cells. J Bioi Chern. 278:31277-31285. 65. Isaacs, J.S., Y.J. .lung, E.G. Mimnaugh, A. Martinez, F. Cuttitta, and L.M. Neckers. 2002. Hsp90 regulates a von Hippel Lindau-independent hypoxia-inducible factor-l alpha-degradative pathway. J Bioi Chern. 277:29936-29944. 66. Minet, E., D. Mottet, G. Michel, 1. Roland, M. Raes, J. Remacle, and C. Michiels. 1999. Hypoxia-induced activation ofHIF-l: role ofHIF-lalpha-Hsp90 interaction. FEBS Lett. 460:251-256. 67. Maloyan, A., L. Eli-Berchoer, G.L. Semenza, G. Gerstenblith, M.D. Stem, and M. Horowitz. 2005. HIF-lalpha-targeted pathways are activated by heat acclimation 96 and contribute to acclimation-ischemic cross-tolerance in the heart. Physiol Genomics.23:79-88. 68. Treinin, M., J. Shliar, H. Jiang, lA. Powell-Coffman, Z. Bromberg, and M. Horowitz. 2003. HIF-1 is required for heat acclimation in the nematode Caenorhabditis elegans. Physiol Genomics. 14:17-24. 69. Nakano, M., D.L. Mann, and A.A. Knowlton. 1997. Blocking the endogenous increase in HSP 72 increases susceptibility to hypoxia and reoxygenation in isolated adult feline cardiocytes. Circulation. 95:1523-1531. 70. Ciocca, D.R., and S.K. Calderwood. 2005. Heat shock proteins in cancer: diagnostic, prognostic, predictive, and treatment implications. Cell Stress Chaperones. 10:86-103. 71. Clemens, lC., C.A. Worby, N. Simonson-Leff, M. Muda, T. Maehama, B.A. Hemmings, and J.E. Dixon. 2000. Use of double-stranded RNA interference in Drosophila cell lines to dissect signal transduction pathways. Proc Natl Acad Sci U SA. 97:6499-6503. 72. Cherniack, N.S., and G.S. Longobardo. 1970. Oxygen and carbon dioxide gas stores of the body. Physiol Rev. 50:196-243. 73. Groenewegen, K.H., A.M. Schols, and E.F. Wouters. 2003. Mortality and mortality- related factors after hospitalization for acute exacerbation of COPD. Chest. 124:459-467. 74. Connors, A.F.J., N.V. Dawson, C. Thomas, F.E.J. Harrell, N. Desbiens, W.J. Fulkerson, P. Kussin, P. Bellamy, L. Goldman, and W.A. Knaus. 1996. Outcomes following acute exacerbation of severe chronic obstructive lung disease. The SUPPORT investigators (Study to Understand Prognoses and Preferences for Outcomes and Risks of Treatments). Am J Respir Crit Care Med. 154:959-967. 75. Hetherington, A.M., and J.A. Raven. 2005. The biology of carbon dioxide. Curr Bioi. 15:R406-10. 76. Bahn, y'S., and F.A. Muhlschlegel. 2006. C02 sensing in fungi and beyond. Curr Opin Microbiol. 9:572-578. 77. Foley, E., and P.H. O'Farrell. 2003. Nitric oxide contributes to induction of innate immune responses to gram-negative bacteria in Drosophila. Genes Dev. 17: 115- 125. 78. Lemaitre, B., and J. Hoffmann. 2007. The host defense of Drosophila melanogaster. Annu Rev Immunol. 25:697-743. 97 79. Hoffmann, J.A., and 1M. Reichhart. 2002. Drosophila innate immunity: an evolutionary perspective. Nat Immunol. 3:121-126. 80. Goto, A., K. Matsushita, V. Gesellchen, L. El Chamy, D. Kuttenkeuler, O. Takeuchi, J.A. Hoffmann, S. Akira, M. Boutros, and J.M. Reichhart. 2008. Akirins are highly conserved nuclear proteins required for NF-kappaB-dependent gene expression in drosophila and mice. Nat Immunol. 9:97-104. 81. Li, G., D. Zhou, A.G. Vicencio, J. Ryu, J. Xue, A. Kanaan, O. Gavrialov, and G.G. Haddad. 2006. Effect of carbon dioxide on neonatal mouse lung: a genomic approach. J Appl Physiol. 101:1556-1564. 82. Jones, W.D., P. Cayirliog1u, LG. Kadow, and L.B. Vosshall. 2007. Two chemosensory receptors together mediate carbon dioxide detection in Drosophila. Nature. 445:86-90. 83. Kwon, J.Y., A. Dahanukar, L.A. Weiss, and 1R. Carlson. 2007. The molecular basis of C02 reception in Drosophila. Proc Natl Acad Sci USA. 104:3574-3578. 84. Vadasz, 1., L.A. Dada, A. Briva, H.E. Trejo, L.C. Welch, J. Chen, P.T. Toth, E. Lecuona, L.A. Witters, P.T. Schumacker, N.S. Chandel, W. Seeger, and J.L Sznajder. 2008. AMP-activated protein kinase regulates C02-induced alveolar epithelial dysfunction in rats and human cells by promoting Na,K-ATPase endocytosis. J Clin Invest. 118:752-762. 85. Sethi, S., and T.F. Murphy. 2001. Bacterial infection in chronic obstructive pulmonary disease in 2000: a state-of-the-art review. Clin Microbiol Rev. 14:336- 363. 86. Shirasu-Hiza, M.M., and D.S. Schneider. 2007. Confronting physiology: how do infected flies die? Cell Microbiol. 9:2775-2783. 87. Cox, C.R., and M.S. Gilmore. 2007. Native microbial colonization of Drosophila me1anogaster and its use as a model of Enterococcus faecalis pathogenesis. Infect Immun.75:1565-1576. 88. Sethi, S. 2004. Bacteria in exacerbations of chronic obstructive pulmonary disease: phenomenon or epiphenomenon? Proc Am Thorac Soc. 1: 109-114. 89. Schneider, D.S. 2007. How and why does a fly turn its immune system off? PLoS Biol.5:e247. 90. Bogdan, C. 2001. Nitric oxide and the immune response. Nat Immunol. 2:907-916. 98 91. Lang, J.D.J., P. Chumley, J.P. Eiserich, A. Estevez, T. Bamberg, A. Adhami, J. Crow, and B.A. Freeman. 2000. Hypercapnia induces injury to alveolar epithelial cells via a nitric oxide-dependent pathway. Am J Physiol Lung Cell Mol Physiol. 279:L994-1002. 92. Lang, J.D., M. Figueroa, K.D. Sanders, M. AsIan, Y. Liu, P. Chumley, and B.A. Freeman. 2005. Hypercapnia via reduced rate and tidal volume contributes to lipopolysaccharide-induced lung injury. Am J Respir Crit Care Med. 171:147-157. 93. Briva, A., r. Vadasz, E. Lecuona, L.C. Welch, J. Chen, L.A. Dada, H.E. Trejo, V. Dumasius, Z.S. Azzam, P.M. Myrianthefs, D. Batlle, Y. Gruenbaum, and J.r. Sznajder. 2007. High C02 levels impair alveolar epithelial function independently ofpH. PLoS ONE. 2:eI238. 94. Choe, K.M., H. Lee, and K.V. Anderson. 2005. Drosophila peptidoglycan recognition protein LC (PGRP-LC) acts as a signal-transducing innate immune receptor. Proc Natl Acad Sci USA. 102:1122-1126. 95. Ramet, M., P. Manfruelli, A. Pearson, B. Mathey-Prevot, and R.A. Ezekowitz. 2002. Functional genomic analysis of phagocytosis and identification of a Drosophila receptor for E. coli. Nature. 416:644-648. 96. Stoven, S., r. Ando, L. Kadalayil, Y. Engstrom, and D. Hultmark. 2000. Activation of the Drosophila NF-kappaB factor Relish by rapid endoproteolytic cleavage. EMBO Rep. 1:347-352. 97. Ni Chonghaile, M., B. Higgins, and J.G. Laffey. 2005. Permissive hypercapnia: role in protective lung ventilatory strategies. Curr Opin Crit Care. 11 :56-62. 98. Vittimberga, F.J.J., D.P. Foley, W.C. Meyers, and M.P. Callery. 1998. Laparoscopic surgery and the systemic immune response. Ann Surg. 227:326-334. 99. O'Croinin, D.F., A.D. Nichol, N. Hopkins, J. Boylan, S. O'Brien, C. O'Connor, J.G. Laffey, and P. McLoughlin. 2008. Sustained hypercapnic acidosis during pulmonary infection increases bacterial load and worsens lung injury. Crit Care Med.36:2128-2135. 100. Margulies, M., M. Egholm, W.E. Altman, S. Attiya, J.S. Bader, L.A. Bemben, J. Berka, M.S. Braverman, Y.J. Chen, Z. Chen, S.B. Dewell, L. Du, J.M. Fierro, X.V. Gomes, B.C. Godwin, W. He, S. Helgesen, C.H. Ho, G.P. Irzyk, S.C. Jando, M.L. Alenquer, T.P. Jarvie, K.B. Jirage, J.B. Kim, J.R. Knight, J.R. Lanza, J.H. Leamon, S.M. Lefkowitz, M. Lei, J. Li, K.L. Lohman, H. Lu, V.B. Makhijani, K.E. McDade, M.P. McKenna, E.W. Myers, E. Nickerson, J.R. Nobile, R. Plant, B.P. Puc, M.T. Ronan, G.T. Roth, G.J. Sarkis, J.F. Simons, J.W. Simpson, M. Srinivasan, K.R. Tartaro, A. Tomasz, K.A. Vogt, G.A. Volkmer, S.H. Wang, Y. Wang, M.P. Weiner, 99 P. Yu, R.F. Begley, and J.M. Rothberg. 2005. Genome sequencing in microfabricated high-density picolitre reactors. Nature. 437:376-380. 101. Bennett, S. 2004. Solexa Ltd. Pharmacogenomics. 5:433-438. 102. Mangiapan, G., M. Vokurka, L. Schouls, J. Cadranel, D. Lecossier, J. van Embden, and AJ. Hance. 1996. Sequence capture-PCR improves detection of mycobacterial DNA in clinical specimens. J Clin Microbiol. 34:1209-1215. 103. Meyer, Q.C., S.G. Burton, and D.A Cowan. 2007. Subtractive hybridization magnetic bead capture: a new technique for the recovery of full-length ORFs from the metagenome. Biotechnol J. 2:36-40. 104. Bashiardes, S., R. Veile, C. Helms, E.R. Mardis, AM. Bowcock, and M. Lovett. 2005. Direct genomic selection. Nat Methods. 2:63-69. 105. St John, J., and T.W. Quinn. 2008. Rapid capture of DNA targets. Biotechniques. 44:259-264.