TRANSCRIPTIONAL PROFILING OF DROSOPHILA MELANOGASTER TENDON CELLS AND MYOTUBES DURING MYOGENESIS by JULIANA MARIE VALERA B.S., Colorado State University, 2009 A thesis submitted to the Faculty of the Graduate School of the University of Colorado in partial fulfillment of the requirements for the degree of Master of Science Biology Program 2015
! ""! 2015 JULIANA MARIE VALERA ALL RIGHTS RESERVED
! """! This thesis for the Master of Science degree by Juliana Marie Valera has been approved for the Biology Program by Aaron N. Johnson, Chair Christopher Miller Christopher Phiel April 24, 2015
! "#! Valera, Juliana Marie (M.S., Biology) Transcriptional profiling of Drosophila melanogaster tendon cells and myotubes during myogenesis Thesis directed by Assistant Professor Aaron N. Johnson. ABSTRACT Myotube elongation is the process by which nascent myotubes extend and identify appropriate muscle attachment sites on tendon cells, forming strong myotendinous junctions. Proper assembly of muscle-tendon attachment is crucial for proper muscle function, and defects in the formation of stable attachments result in congenital myopathies. Although myotube elongation is an essential component of muscle morphogenesis and diversification, the molecular pathways that guide nascent myotubes toward tendon cells remain largely unknown. At the morphological level, tendon progenitor cell loss causes inappropriate myotube localization suggesting that tendon cells secrete essential myotube guidance cues. With the development of RNA-seq technology it is now possible to systematically screen an entire transcriptome for genes that are differentially expressed in two different cell types during different stages of development. This technology can be readily applied to muscle development in Drosophila. To identify guidance pathways, I purified populations of embryonic tendon cells and nascent myotubes by fluorescence activated cell sorting (FACS). I isolated celltype specific mRNAs from these cell populations and have deep sequenced and analyzed the RNA samples. Through the analysis of these samples, ligands that regulate the Fibroblast growth factor (FGF) and Sphingosine-1-phosphate (S1P) pathways have
! #! !! emerged as important players of muscle elongation. Further investigation of these ligands will provide unique insights into the molecular mechanisms that regulate myogenesis. The form and content of this abstract are approved. I recommend its publication. Approved: Aaron N. Johnson
! #"! DEDICATION I dedicate this work to my parents. Everything I am, you helped me to be. Thank you for your unconditional love, support, and encouragement.
! #""! ACKNOWLEDGMENTS My foremost gratitude belongs to my thesis advisor Dr. Aaron Johnson for his guidance and support throughout this challenging process. I am also grateful to Dr. Christopher Phiel and Dr. Christopher Miller for serving as members on my thesis committee. Their comments and questions were beneficial in my completion of this thesis. In addition I would like to thank the past and present lab members of the Johnson Lab for their assistance during this project and for their encouraging words and support. I am also very grateful to the core facilities at the CU Denver Anschutz Medical Campus, in particular the Flow Cytometry Facility, the Genomics and Microarray Core, and the Biostatistics and Bioinformatics Core for technical support. Finally, I would like to thank all of my friends for their understanding and encouragement in many moments of crisis.
! #"""! TABLE OF CONTENTS CHAPTER I. INTRODUCTION ...........................................................................................................1 Overview of muscle development.......1 Ver tebrate limb muscle morphogenesis...................2 Drosophila somatic muscle development................4 Drosophila somatic muscle elongation................6 Myotube guidance cues....7 II. AIMS OF THE THESIS...10 III. MATERIALS AND METHODS.........................................................................11 GAL4/UAS system.................11 Overnight embryo collection for fluorescence activated cell sorting....12 Cell isolation..13 Fluorescence activated cell sorting (FACS)......13 RNA extraction......14 RNA-sequencing Bioinformatics analysis of RNA-sequencing results.....15 Embryo dechorionation and fixation..... In vitro transcription reaction to generate riboprobe for in situ hybridization..17 RNA in situ hybridization..18 Fluorescent in situ hybridization IV. RESULTS........................................................................................24 FACS purification..24
! "$! RNA integrity.....25 RNA-seq uencing....26 Principal-component analysis....28 Validation of RNA-seq uencing data......30 Bioinformatics analysis of enriched transcripts.........32 Sphingosine-1-Phosphate (S1P) signaling.....36 Fibroblast growth factor (FGF) signaling......37 V. DISCUSSION.......................................................................................................41 Conclusions....41 VI. FUTURE DIRECTIONS.........................44 Further directions with RNA-seq uencing data......44 REFERENCES..................................................................................................................46
! $! !! LIST OF TABLES Table 1. Myotube RNA-seq uencing results of known myogenic transcripts..........31 2. Lipid metabolic process mRNA expression...36 3. FGF signaling pathway mRNA expression
! $" ! LIST OF FIGURES Figure 1. Vertebrate limb muscle morphogenesis...........................................................................3 2. Drosophila embryonic somatic musculature.......................................................4 3. Somatic muscle development..........................................................................................6 4. Somatic muscle elongation..................................................................................7 5. The UAS-GAL4 system.12 6. Robust, tissue-specific expression of eGFP for FACS......25 7. Comparison of quality of RNA from FACS sorted tendon cells and myotubes 8. PCA of RNA-sequencing data... 9. Somatic muscle RNA expression pattern..31 10. Functional annotation clustering of significantly enriched genes................34 11 Enriched GO biological process categories.....35 12. Biosynthesis and metabolism of Sphingosine-1-Phosphate (S1P).. 13. Pyramus (Pyr) and Thisbe (Ths) are expressed in tendon cells...39 14. Alignment of Drosophila CG42369 and hs FGF3...40 15. MHC expression in St 17 wildtype embryos...45
! $""! LIST OF ABBREVIATIONS 7AAD 7-Amino Actinomycin D ANOVA analysis of variance BDGP Berkeley Drosophila Genome Project BDSC Bloomington Drosophila Stock Center Btl Breathless CRISPR Clustered regularly interspaced short palindromic repeat crRNA CRISPR RNA DA Dorsal Acute DAVID Database for Annotation, Visualization and Integrated Discovery Dnt Doughnut on 2 DO Dorsal Oblique drl derailed DSB Double strand breaks DT Dorsal Transverse Duf Dumb-founded eGFP enhanced green fluorescent protein Egfr Extracellular growth factor receptor EOE End over end FACS Fluorescence activated cell sorting FC founder cell FCMs fusion-competent myoblasts FGFs Fibroblast growth factors
! $"""! FGFRs high affinity tyrosine kinase receptors FISH Florescent in situ hybridization FPKM fragments per kilobase of exon per million mapped reads GO Gene Ontology gRNAs Guide RNAs GSNAP Genomic Short-Read Nucleotide Alignment Program HR Homologous recombination Htl Heartless Ig Immunoglobulin domain ISH In situ hybridization Kon Kontiki LL Lateral Longitudinal LO Lateral Oblique 1sc lethal of scute LT Lateral Transverse MHC Myosin Heavy Chain MTJ Muscle-tendon junction NM Nemaline myopathy PCA Principal-component analysis PC1 Principal component 1 PC2 Principal component 2 PK Proteinase K pyr pyramus
! $"#! RIN RNA Integrity Number Robo Roundabout RNA-seq RNA Sequencing RT Room temperature SBM Segment Border Muscles Shh Sonic hedgehog Sns Sticks-and-stones S1P Sphingosine-1-phosphate Sr Stripe St Stage TALEN Transcription activator-like effector nuclease TCs Tendon cells ths thisbe TPCs Tendon precursor cells TPM2 tropmyosin2 TPM3 tropomyosin 3 twi Twist VA Ventral Acute VL Ventral Longitudinal VO Ventral Oblique VT Ventral Transverse Wun2 Wunen2
! %! CHAPT ER I INTRODUCTION Overview of muscle development Muscle is a contractile tissue that performs multiple functions from the facilitation of conscious and unconscious movements to the maintenance of posture in multicellular animals. The formation of muscular tissue during embryonic development depends on a complex series of events beginning with the specification of the mesoderm. Muscles distinct in type, size, shape, and attachment originate from the mesoderm and achieve their identities based on the regulation of signaling mechanisms, genetic networks, and molecular switches (Bate, 1993; Christ and Ordahl, 1995). Muscle develops in both vertebrates and invertebrates through the fusion of immature muscle cells called myoblasts. Myoblasts fuse with neighboring cells to form multinucleated myofibers. Each myofiber elongates, matures, and establishes an attachment to specific tendon cells. This process of elongation is an essential step of myogenesis and may underlie a number of muscle diseases. However, there are major knowledge gaps concerning the molecular regulators of this process (Schweitzer et al., 2010; Schejter and Baylies, 2010; Schnorrer and Dickson, 2004). Studies in vertebrate and invertebrate systems have provided insights into the understanding of the genetic, molecular, and cellular mechanisms that regulate myogenesis. An invertebrate model that has an enormous impact on the genetic analysis of myogenesis is the fruit fly, Drosophila melanogaster. Compared to vertebrate models, Drosophilas fast generation time and ex utero development are features that make it an ideal experimental organism. In addition, Drosophila is amendable to genetic
! &! manipulation and contains many similarities with vertebrates at both the morphological and molecular levels making it an indispensable model in the field of myogenesis. Orthologues of key genes in Drosophila also regulate vertebrate skeletal muscle development and the musculature of the Drosophila embryo body wall shares fundamental similarities in structure and function with vertebrate skeletal muscle (Taylor, 2006). The two are similar in that they are both striated, multinucleate, and under voluntary control. Vertebrate limb muscle morphogenesis In vertebrates, skeletal muscle originates from the lateral dermomyotome of somites, an epithelial sheet formed by the paraxial mesoderm (Christ and Ordahl, 1995; Chevallier et al., 1997). During vertebrate limb muscle morphogenesis, skeletal muscle progenitors migrate from the dermomyotome into the developing limb bud where they differentiate into myoblasts to form the dorsal and ventral muscle masses (Chevallier et al., 1997; Bismuth and Relaix, 2010) (Figure 1). As the limb develops distally, myoblasts undergo migration to form the distal musculature (Anderson et al., 2012; Hu et al., 2012). During myoblast migration, tendon cell progenitors stay in close proximity to myoblasts (Schweitzer et al., 2001). The tendon progenitors then segregate, locate their skeletal insertion sites, and mature into functional tendons (Kardon, 1998). Concurrently, mononucleate myoblasts fuse and elongate to form primary fibers which will identify and locate their muscle attachment sites and form strong myotendinous junctions (Wigmore and Evans, 2002; Schnorrer et al., 2004; Sieiro-Mosti et al., 2014). Many of the molecules that control vertebrate muscle precursor specification and progenitor cell migration from the dermomyotome to the limb have been identified
! '! (Anderson et al., 2012; Hu et al., 2012), but the molecular pathways responsible for guiding a nascent myotube toward tendon cells remain largely unknown. At the morphological level, tendon progenitor cell loss causes inappropriate myotube localization and tendon cell maturation requires close physical association with approaching myotubes (Kardon, 1998). Hence, myotube and tendon cell maturation require bidirectional signals between each cell type. Until the missing guidance cues that underlie the process of myotube elongation have been identified, we cannot fully understand the process of myogenesis. Figure 1. Vertebrate limb muscle morphogenesis. Diagram of myoblast migration and targeting to tendon cells during vertebrate limb muscle development in an avian embryo. Sonic hedgehog (Shh) signaling is required for muscle cell migration toward the distal limb bud (Hu et al., 2012).
! (! Drosophila somatic muscle development Insect somatic muscles are the muscles of the insect body wall that direct contraction of the exoskeleton. Each embryonic abdominal hemi-segment develops 30 distinct somatic muscles (Figure 2) consisting of 5 DO (dorsal oblique), 3 DA (dorsal acute), 1 LL (lateral longitudinal), 1 LO (lateral oblique), 4LT (lateral transverse), 4 VL (ventral longitudinal), 2 SBM (segment border muscles), 6 VO (ventral oblique), 1DT (dorsal transverse), 1VT (ventral transverse), and 3 VA (ventral acute) muscles (Hartenstein, 2006). B A F igure 2 Drosophila embryonic somatic musculature (A) Diagram of the 30 somatic muscles that develop in each segment of the Drosophila embryo. Somatic muscle is analogous to vertebrate skeletal muscle. (Reprinted from Ruiz Gomez et al., 1997). (B) Confocal micrograph of somatic muscles from three segments of a live Drosophila embryo. B
! )! Embryonic somatic muscle development in Drosophila requires a series of consecutive events (Figure 3). In the developing Drosophila embryo, expression of the transcription factor Twist (twi) results in the division of the mesoderm into two subdomains. The mesodermal subdomain that contains high levels of twi expression will give rise to the somatic mesoderm (Baylies and Bate, 1996). Further patterning of somatic mesodermal cells is accomplished by the expression of the proneural gene, lethal of scute (1sc) (Carmena et al., 1995). Notch-dependent lateral inhibition restricts the expression of 1sc to a single muscle progenitor cell within a precluster of cells that make up an equivalent group (Carmena and Baylies, 2006). This cell will undergo asymmetric division to form a founder cell (FC), which seeds the formation of specific muscles. The remaining cells within the cluster form fusion-competent myoblasts (FCMs) (Carmena et al., 1995). Each of the 30 FCs within a Drosophila hemi-segment will fuse with FCMs to form individual multinucleate myotubes (Bate, 1990). FCs express the Immunoglobulin domain (Ig) transmembrane protein Dumb-founded (Duf) while FCMs express the Ig transmembrane protein Sticks-and-stones (Sns) (Ruiz-Gomez et al., 2000; Bour et al., 2000). During myoblast fusion, the dissimilar transmembrane protein domains bind to one another, unifying the two cells. Tight association of cell lipid biolayers is achieved by increasing amounts of Myosin II, which generates the mechanical tension needed for membrane fusion of FCs to FCMs (Kim et al., 2015). The resulting myotube acquires a specific fate due to the differential expression of muscle identity genes, which are inherited by the founder cell of an individual muscle (Frasch, 1999; Baylies et al., 1998). For example, the muscle identity genes Kruppel (Kr) and slouch encode transcription
! *! factors whose expression determines the specification of the VA muscles 26 and 27 (Knirr et al., 1999; Ruiz-Gomez et al., 1997). Immediately after generation of the VA26 and VA27 founder cells, muscle VA26 will lose co-expression of Kr and slouch while muscle VA27s identity depends on maintaining co-expression of Kr and slouch until after fusion (Frasch, 1999). Drosophila somatic muscle elongation As myotubes mature, they undergo a process of myotube elongation in which nascent myotubes are guided toward attachment sites and establish firm connections with specialized epidermal cells called tendon cells (Schnorrer et al., 2004). Myotube elongation occurs concurrently with myotube formation (Figure 4). Founders fuse with neighboring FCMs as the nascent myotube extends its leading edges bi-directionally toward tendon precursor cells (TPCs). TPCs are defined autonomously in the ectoderm. The early growth response-like triple zinc finger transcription factor Stripe (Sr) is responsible for the differentiation of early TPCs into tendon cells (TCs) (Frommer et al., 1996). However, tendon cell development is biphasic and Sr alone is not sufficient for Fig ure 3 Somatic muscle development 30 somatic muscle myoblasts, or founder cells, are specified in e ach embryonic segment (3 founders shown). The founder cells fuse with neighboring cells to form nascent myotubes (second panel). The nascent myotubes then elongate toward tendon cells (third panel). After elongation, mature somatic muscles form intracellul ar attachments with tendon cells.
! +! tendon cell maturation. While tendon cells direct myotube migration, tendon-muscle association plays an equally important role in tendon cell differentiation (Becker et al., 1997). The establishment of an attachment between a myotube and the tendon precursor, results in the terminal differentiation of a tendon cell via Vein signaling (Yarnitzky et al., 1997). As a myotube is approaching a TPC, it releases the neuregulin-like ligand Vein. Vein accumulates at the muscle-tendon junction (MTJ) site and signals to TPCs through the extracellular growth factor receptor (Egfr). TPCs that receive the Vein signal express high levels of Sr, which are necessary for TC terminal differentiation (Yarnitzky et al., 1997). Myotube guidance cues In Drosophila, the only known molecular regulator of myotube elongation is Slit, a leucine-rich repeat protein that is expressed by segment border tendon cells and localizes to muscle attachment sites in the ectoderm (Rothberg et al., 1990; Kramer et al., 2001). Normally, ventral muscles attach to the epidermis and do not cross the ventral Figure 4. Somatic muscle elongation The molecular pathways that guide nascent myotubes toward tendon cells remain unknown.
! ,! midline. However, slit mutant embryos display ventral muscles that inappropriately overextend across the central nervous system, resulting in abnormal muscle patterning (Kidd et al., 1999). In addition, ventral muscles of slit mutant embryos frequently misattach to muscle attachment sites in the epidermis, dramatically altering muscle pattern (Kramer et al., 2001). These results argue that myogenesis is dependent on a muscles response to unknown guidance signals. Similarly, the expression of Slit in dorsal and ventral muscle masses of developing chick and mouse hind limbs demonstrate the mechanisms that regulate myotube elongation in both vertebrates and invertebrates are likely conserved (Vargesson et al., 2001). Several intracellular effectors of myotube elongation have been identified in muscle or tendon cells (Callahan et al., 1995; Schnorrer et al., 2007; Kramer et al., 2001). The Roundabout (Robo) protein is expressed in the ventral muscles and is the guidance receptor for Slit (Kramer et al., 2001). When Robo is expressed ectopically in LT muscles, these muscles are attracted toward segment border tendon cells (Kramer et al., 2001). The orphan receptors derailed (drl) and Doughnut on 2 (Dnt), members of the receptor tyrosine kinases (RTKs) family (Callahan, 1995), have also been shown to play a role in muscle guidance. Drl and Dnt coordinate to guide the LT muscles 21-23 toward correct attachment sites (Callahan, 1996; Lahaye et al., 2012). In another identified molecular system for muscle guidance, the transmembrane protein Kontiki (Kon) is expressed in a subset of myotubes and is required for recognition and attachment of the VL 1-4 muscles to their correct muscle attachment sites (Schnorrer et al., 2007). Although the ligand that binds to the Kon receptor is unknown, the glutamate receptor binding protein (Dgrip) has been shown to play a role in Kon signaling through its PDZ
! -! binding domain. Prevention of Kon-Dgrip interaction lead to defects in the VL1 muscles (Schnorrer et al., 2007). Directed cell migration and target recognition are critical for the development of the musculature. To date, the molecular mechanisms that control this process remain largely unknown. The identification of orphan receptors Kon, Drl, and Dnt provide insight into the mechanisms that underlie muscle development, however these studies have not identified the ligands that promote myotube elongation.!While Drosophila has had an enormous impact on the genetic analysis of myogenesis, the Drosophila embryo is a limiting factor in identifying myotube elongation pathways due to the difficulty of dissecting out tendon and muscle cells needed to perform downstream gene expression assays such as microarray analysis and RNA Sequencing (RNA-seq ). With the advancement of fluorescence activated cell sorting (FACS) technology, we can identify and isolate tendon or muscle cells in developing transgenic Drosophila embryos. Tendon cells labeled with the cell surface marker Stripe (Sr) and myotubes labeled with the cell surface marker RP298, express the enhanced green fluorescent protein (eGFP), enabling FACS sorting of cell-type specific markers via the GAL4 /UAS system. This alternative approach combined with RNA-seq, will provide an accurate snapshot of the mRNA transcriptome during muscle elongation, allowing for the identification of not only novel protein ligands, but perhaps unsuspected classes of ligands.
! %.! CHAPTER II AIMS OF THE THESIS The goal of this research project is to identify potential guidance pathways and intracellular effectors that regulate myotube elongation. I hypothesize that there are instructive interactions between tendon cells and muscle cells, in which a tendon cell releases signals that are received by the myoblast to facilitate myotube path finding and muscle-tendon attachment. To identify guidance molecules, I performed the following: Aim#1: Use fluorescence activated cell sorting (FACS) to profile the transcriptome in tendon cells and nascent myotubes during myotube elongation in developing Drosophila embryos. Aim#2: Identify candidate transcripts that show significant enrichment of tendon cells or myotubes. Aim#3: Validate candidate transcripts by in situ hybridization (ISH).
! %%! CHAPTER III MATERIALS AND METHODS GAL4/UAS system The UAS-GAL4 system drives efficient, tissue-specific gene expression in Drosophila. The GAL4 gene encodes the yeast transcription activator protein GAL4 and has the ability to activate transcription in flies from promoters that contain GAL4 binding sites (Fischer et al., 1988). Directed gene expression is accomplished by random insertion of the GAL4 gene into the Drosophila genome. A GAL4-dependent target gene is constructed by subcloning a sequence of interest behind an upstream activation sequence (UAS) that contains GAL4 binding sites. In the absence of GAL4, the target gene is silent. However when GAL4 is present, it binds to specific sites located on the UAS and activates gene transcription (Brand and Perrimon, 1993). Fly lines carrying a UAS.eGFP insertion have been shown to successfully isolate cell-type specific cells via FACS (Bryantsev and Cripps, 2012). To activate a target gene in a cell-specific pattern we obtained two distinct UAS.eGFP lines and tested their expression in tendon cells, using Sr.Gal4, and separately in migrating myotubes, using RP298.Gal4. Flies carrying the target (UAS-eGFP) were crossed to flies expressing Sr.GAL4 or RP298.Gal4 (Figure 5).
! %&! Overnight embryo collection for fluorescence activated cell sorting Myotube elongation occurs during embryonic stage (St) 12-13 (7-10 hour post fertilization at 25C). To ensure that embryos collected for FACS were at the same developmental stage, grape agar plates were collected every 2 hours for a total of 10 hours and aged for 9.5 hours at 25C. Embryos were independently collected from two genotypes: tendon cells expressing eGFP and myotubes expressing eGFP. Figure 5. The UAS-GAL4 system. The UAS-GAL4 method consists of a fly line that contains a GAL4 driver, and another fly line that contains a gene of interest downstream of a UAS binding site. When the two fly lines are crossed to one another, GAL4 protein binds to the UAS and activates transcription of the gene of interest.
! %'! Cell isolation Cells were isolated according to Bryantsev and Cripps, 2012. 1. Dechorinate embryos with 40% bleach for 2.5 minutes. Wash extensively with distilled water and transfer with PBST. 2. Use a low-magnification dissection microscope to locate and remove hatched larvae and old embryos. 3. Remove PBST and dounce homogenize ~1.0-1.5 x 103 embryos in 1ml of Insect Dissection Medium using 7 strokes. 4. Combine homogenates in a 15ml tube and spin at 500 x g for 5 minutes. 5. Resuspend in 1ml of Schneiders medium treated with 1mg/ml elastase. 6. Incubate end over end (EOE) at room temperature (RT) for 30 minutes. 7. Stop the enzymatic reaction with 4ml Schneiders medium containing 10% FBS, gravity-filter through a 70um cell strainer and centrifuge at 500 x g for 5 minutes. 8. Resuspend pellet in 1ml of Schneiders medium supplemented with 10% FBS, 10ug/ml 7AAD, 10ug/ml Hoechst, and 0.1% Pluronic F-68. 9. Incubate for 20 minutes at RT, keep on ice until sorting. Fluorescence activated cell sorting (FACS) 1. The cell suspension was sorted using a high-speed FACSAria cell sorter. Adjust machine settings to the single cell-sorting mode. 2. Replace the sheath fluid with 2L of Seecof saline containing: 6mM Na2HPO4 3.67mM KH2 PO4 106mM NaCl, 26.8 mM KCL, 6.4 mM MgCl2, 2.25mM CaCl2 pH 6.8, and additionally supplemented with 0.05% Pluronic F-68. 3. Gate for GFP-positive and Hoechst-positive events (Figure 6).
! %(! 4. Collect cells directly into a microcentrifuge tube containing 350ul of Seecof saline and vortex. RNA extraction RNA was isolated using an optimized protocol for RNeasy Mini kit (QIA-GEN). 1. Pellet cells at 1,000 x g for 3 minutes. Remove supernatant and add 700ul of Buffer RLT and vortex for 1 minute. 2. Transfer up to 700l of the sample, including any precipitate that may have formed, to an RNeasy spin column placed in a 2 ml collection tube. Close the lid gently, and centrifuge for 15 seconds at 10,000 x g. Discard the flow-through. 3. Add 700l Buffer RW1 to the RNeasy spin column. Close the lid gently, and centrifuge for 15 seconds at 10,000 x g to wash the spin column membrane. Discard the flow-through. 4. Add 500l Buffer RPE to the RNeasy spin column. Close the lid gently, and centrifuge for 15 seconds at 10,000 x g to wash the spin column membrane. Discard the flow-through. 5. Add 500l Buffer RPE to the RNeasy spin column. Close the lid gently, and centrifuge for 2 minutes at 10,000 x g to wash the spin column membrane. 6. Place the RNeasy spin column in a new 2ml collection tube, and discard the old collection tube with the flow-through. Close the lid gently, and centrifuge at full speed for 1 minute. 7. Place the RNeasy spin column in a new 1.5ml collection tube. Add 20l of RNase-free water directly to the spin column membrane. Close the lid gently, and incubate at RT for 1 minute.
! %)! 8. Centrifuge for 1 minute at 10,000 x g to elute the RNA. *Perform all steps of the procedure at RT. During the procedure, work quickly. RNA-sequencing RNA samples from tendon cells, non-sorted cells from tendon-GFP expressing lines, myotube cells, and non-sorted cells from myotube-GFP expressing lines were submitted to the University of Colorado Genomics and Microarray Core for library preparation. Strand-specific RNA libraries were prepared from samples using the Illumina TruSeq stranded-mRNA sample library kit. Libraries were sequenced on the Illumina HiSeq2000 platform using single-end 50bp chemistry. Bioinformatics analysis of RNA-sequencing results Bioinformatics analysis was performed by the University of Colorado Bioinformatics Core. Quality control of sequence data using a custom Python script removed low-quality bases. The remaining sequences were mapped to the Drosophila genome using Genomic Short-Read Nucleotide Alignment Program (GSNAP) (Wu and Nacu 2010). Cufflinks permitted the assembly of transcripts to the Drosophila genome (Trapnell et al., 2012). After strand alignment of sequences, the number of fragments per kilobase of exon per million mapped reads (FPKM) was determined. FPKMs from all six libraries were analyzed using princip al-component analysis (PCA) and the total number of times that each nucleotide was sequenced was normalized to the sum of numbers of all Drosophila transcripts in tendon cell and myotubes. Drosophila open reading frames (ORFs) that were transcribed differently in tendon cells than myotubes were identified based on their calculated fold changes in FPKMs and then by analysis of variance (ANOVA) (Baird et al., 2014).
! %*! mRNAs that showed increased enrichment in tendon cells and muscle cells compared to non-sorted cells were further analyzed by the Database for Annotation, Visualization and Integrated Discovery (DAVID) (Huang et al., 2009). DAVID was used to perform Gene Annotation Enrichment Analysis on upregulated transcripts. The analysis was used to identify Gene Ontology (GO) clusters that contain significantly enriched signaling component associated GO terms. Embryo dechorionation and fixation 1. A cage was set up containing about 100 adult virgins and 50 adult males. Cages were covered with a grape agar plate containing yeast paste. Cages were kept at 25C for optimal egg laying. After the adult flies have laid for 24 hours, the grape agar plate was changed. 2. Remove the embryos from the original plate by adding a small amount of water to the plate and by gently swirling a paintbrush over the surface of the agar to loosen the embryos. Pour the embryos into a cell strainer and rinse thoroughly to remove all residual yeast. 3. Dechorionation. Drosophila embryos are enclosed in a chorion. To dechorionate, place the cell strainer in a 40% solution of ultra bleach for 2.5 minutes. Swirl the basket several times to ensure all embryos are exposed to the bleach. Rinse the basket several times in cold water. Then rinse the basket with distilled water. Transfer the embryos to a 7ml scintillation vial with PBST. 4. Fix. Remove all PBST from the scintillation vial. Dilute 0.5ml of formaldehyde in 1.5ml of PEM and add fix to scintillation vial. Add 2ml heptane to the fix. Shake the vial by hand briefly. The embryos should now be at the interface of the
! %+! aqueous (fix) and non-aqueous (heptane) layers. If so, shake embryos at 250 RPM for 20 minutes. 5. Devitellination. Drosophila embryos are enclosed in a vitelline membrane that must be removed for all staining procedures. To devitellinate the embryos, remove the bottom (non-aqueous) layer from the scintillation vial. Use a 200ul pipet to remove all the liquid in the lower layer without removing the embryos. Add one volume of 2ml methanol and shake the embryos by hand for 2 minutes. Excessive shaking at this step will break the embryos. Stand the vial upright for one minute. Devitellinated embryos will settle to the bottom of the vial. Remove the embryos with a clean Pasteur pipet and transfer to a 1.5ml microfuge tube. Rinse the embryos twice with methanol. Store embryos in methanol at -20C. In vitro transcription reaction to generate riboprobe for in situ hybridization 1. Mix the following in a microfuge tube: 2l 10X labeling mix (Dig), 2l 10X transcription buffer, 1l RNase inhibitor, 1l polymerase 50U/l (specific to the promoter on 3 end of cDNA), 1g linear template DNA. Bring total reaction volume to 20l with DEPC sH2O. 2. Incubate reaction for 3 hours at 37C UNLESS the polymerase is Sp6, then incubate 30C. A. Add 1l Turbo DNase, incubate for 20min at 37C B. Precipitate the RNA with 2l 3M NaOAc (DEPC treated) and 50l 100% Ethanol. C. Resuspend the RNA in DEPC treated H2O and store at -80C. Probe concentration can be determined using a spectrophotometer.
! %,! 3. To optimize the probe concentration: A. Hybridizing 1ug of probe usually results in a clean in situ, but optimal probe concentration should be found empirically. B. Perform a series of in situs on wild type embryos using a range of probe concentrations to find the optimal probe concentration. Embryos that turn black immediately have too much probe. RNA in situ hybridization We first checked the Berkeley Drosophila Genome Project (BDGP) database for RNA expression patterns of our candidate transcripts. The BDGP has documented the expression pattern for thousands of genes identified in the Drosophila genome (Tomancak et al., 2007). If the gene expression of a candidate transcript was unreported, we performed RNA in situ hybridizations using the following protocol: 1. This protocol has been optimized for dig-labeled RNA probes. All rinses are 1ml unless otherwise noted. 2. Transfer 50l of embryos fixed in 4%formaldehyde for 20 min into RNase free siliconized microfuge tubes. Siliconized tubes ensure that approximately 90% of embryos are recovered at the end of the protocol. 3. Rinse the embryos several times with methanol then once with 1:1 methanol to xylenes and incubate 5 min EOE RT. Add 1ml xylenes and incubate 1-3 hours EOE RT. Xylenes clear the embryos and gives cleaner staining. Rinse 3X with methanol. 4. Hydrate the embryos. All solutions should be DEPC treated for the remainder of the hybridization. Add 0.5ml methanol and 0.5ml PBT and rotate EOE RT for
! %-! 5min. Remove methanol/PBT, add 1ml PBT and rotate EOE RT 5min. Remove PBT and briefly rinse with 1ml PBT. 5. Second fix. Add 875l PBT and 125l 16%formaldehyde to the embryos and incubate 5 min EOE at RT. Rinse three times with PBT. 6. Proteinase K (PK) treatment. PK degrades cell membranes allowing the riboprobe to readily diffuse into the cell. Riboprobes can still enter the cell without PK treatment, albeit less efficiently. The length of the PK treatment must be optimized as excess PK treatment will cause the embryos to clump. Incubating embryos in freshly diluted 20g/ml PK for 1 min EOE at RT works well. Following PK treatment immediately rinse embryos with four quick PBT washes. 7. Third fix. Add 875l PBT and 125l 16%formaldehyde to the embryos and incubate 30 min EOE at RT. Rinse three times with PBT. 8. Fourth fix. Add 875l PBT and 125l 16%formaldehyde to the embryos and incubate 20 min EOE at RT. Rinse three times with PBT. 9. Equilibrate embryos in prehyb. Incubate embryos in a 1:1 solution of prehyb to PBT for 5 min EOE RT then twice in prehyb for 5 min EOE at RT. 10. Prehybridization. Prewarm 1ml prehyb to 55C. Add the warmed prehyb to the embryos and incubate at 55 C in a stationary heat block for at least one hour. 11. Hybridization. To remove hairpins from the probe, boil the volume of probe to be hybridized then immediately cool the probe in an ice bath. Spin down the tube and return the probe to the ice bath. Remove as much of the prehyb as possible off of the embryos. Add probe then 100l prewarmed HYB. After adding the
! &.! HYB spin the embryos in a tube rack to mix the HYB. Be careful not to turn the tube over while mixing. Hybridize the probe for 20 hours at 55 C in water bath. 12. Probe rinse and rehydration. All washes at 55C. Add 1ml prewarmed prehyb and incubate 20 min. Wash embryos for 20min each in the following prehyb:PBT dilution: 4:1, 3:2, 2:3, 1:4. Then rinse twice with PBT EOE RT 10min each. 13. Block embryos in 5%NGS/PBT for 1-3 hours EOE RT. 14. Dilute !-Dig -AP (Roche 1093274) 1:2000 in 5%NGS/PBT. Preabsorbing the !Dig -AP does not improve the noise to signal ratio. Add 500l diluted antibody to embryos and incubate 2-3 hours RT EOE. Altering the !-Dig -AP concentration does not improve an excessive signal or a weak signal. If the signal is incorrect, then alter the probe concentration. 15. Rinse embryos 3 times PBT RT EOE 10min each. Rinse the embryos twice with Staining Buffer (this should be made the day of the reaction) for 10 min each. 16. Staining Solution. Make 1ml for each reaction. Add 8 l of NBT/BCIP stock solution for each ml of Staining Buffer to make the Staining Solution. Mix well and keep in the dark. 17. After embryos have been rinsed in the Staining Buffer, add 1ml Staining Solution. Transfer some embryos to dish and keep in the dark. Cover the tubes with foil and rotate EOE RT. Watch the color develop in the embryos in the dish using a dissecting scope. The reaction usually takes 10-60 min. Wait until a strong signal is achieved since the signal seems to weaken at higher magnifications. 18. Rinse embryos four times in PBT. Equilibrate in 70%glycerol at 4 C. Embryos are fully equilibrated after they sink to the bottom of the microcentrifuge tube.
! &%! Fluorescent in situ hybridization 1. Transfer 50l of embryos fixed in 4%formaldehyde for 20 min into RNase free siliconized microfuge tubes. 2. Rinse the embryos several times with methanol. Add 900ul of methanol and 100ul of 30% hydrogen peroxide and rotate EOE for 60 min. 3. Rinse the embryos several times with methanol then once with 1:1 methanol to xylenes and incubate 5 min EOE RT. Add 1ml xylenes and incubate 1-3 hours EOE RT. 4. Hydrate the embryos. All solutions should be DEPC treated for the remainder of the hybridization. Add 0.5ml methanol and 0.5ml PBT and rotate EOE RT for 5min. Remove methanol/PBT, add 1ml PBT and rotate EOE RT 5min. Remove PBT and briefly rinse with 1ml PBT. 5. Second fix. Add 875l PBT and 125l 16%formaldehyde to the embryos and incubate 5 min EOE at RT. Rinse three times with PBT. 6. Proteinase K (PK) treatment. Incubate embryos in freshly diluted 20g/ml PK for 1 min EOE at RT. Following PK treatment immediately rinse embryos with four quick PBT washes. 7. Third fix. Add 875l PBT and 125l 16%formaldehyde to the embryos and incubate 30 min EOE at RT. Rinse three times with PBT. 8. Fourth fix. Add 875l PBT and 125l 16%formaldehyde to the embryos and incubate 20 min EOE at RT. Rinse three times with PB T. 9. Equilibrate embryos in prehyb. Incubate embryos in a 1:1 solution of prehyb to PBT for 5 min EOE RT then twice in prehyb for 5 min EOE at RT.
! &&! 10. Prehybridization. Prewarm 1ml prehyb to 55C. Add the warmed prehyb to the embryos and incubate at 55C in a stationary heat block for at least one hour. 11. Hybridization. To remove hairpins from the probe, boil the volume of probe to be hybridized then immediately cool the probe in an ice bath. Spin down the tube and return the probe to the ice bath. Remove as much of the prehyb as possible off of the embryos. Add probe then 100l prewarmed HYB. After adding the HYB spin the embryos in a tube rack to mix the HYB. Be careful not to turn the tube over while mixing. Hybridize the probe for 20 hours at 55C in water bath. 12. Probe rinse and rehydration. All washes at 55C. Add 1ml prewarmed prehyb and incubate 20 min. Wash embryos for 20 min each in the following prehyb:PBT dilution: 4:1, 3:2, 2:3, 1:4. Then rinse twice with PBT EOE RT 10 min each. 13. Block embryos in 5%NGS/PBT for 1-3 hours EOE RT. 14. Dilute mouse !-Dig 1:200 and rabbit !-GFP 1:600 in 5%NGS/PBT. Preabsorb the antibody to improve the noise to signal ratio. Add 300l diluted antibody to embryos and incubate overnight at 4C EOE. 15. Rinse embryos four times with 1ml 5%NGS/PBT RT EOE 10 min each. 16. Dilute !mouse HRP 1:25 and !rabbit Alexa 555 1:50 in 5%NGS/PBT. Add 300l diluted antibody to embryos and incubate for 2 hrs RT EOE. 17. Rinse embryos three times in PBT EOE RT 10 min each. 18. TSA reaction. This should be done in the dark. After rinsing the secondary antibody off of the embryos, add 200ul TSA reaction buffer, 2ul diluted hydrogen peroxide solution, and 2ul of tyramide conjugated Alexa fluor. Rotate embryo s
! &'! EOE at RT for 5 minutes. If the staining reaction goes too long there will be high background staining in the embryos. Increase antibody concentrations if the TSA reaction does not produce a strong signal after 10 min. 19. After the TSA reaction has gone for the appropriate time, rinse the embryos three times with PBT, then once with PBT for 10 min EOE. The embryos can now be equilibrated in 70%Glyceroll/PBT. Store in the dark at 4 C.
! &(! CHAPTER IV RESULTS FACS purification To identify genes and other factors contributing to the process of myotube elongation, I used FACS to purify tendon cells and muscle cells from dissociated embryos at St 12-13 (8-10 hours post fertilization). St 12-13 is the point in development in which the myotubes of a Drosophila embryo are undergoing elongation. To accomplish purification of tendon cells and myotubes, we took advantage of a UASeGFP fly line. Embryos that expressed eGFP in a subset of cells under the control of a specific GAL4 promoter [tendon:GFP (Figure 6A) or myotube:GFP (Figure 6B)] were collected (see Methods and Materials for overnight embryo collection for fluorescence activated cell sorting), dounce homogenized, and dissociated into a cell suspension. Cells were stained with the viability marker Hoechst and the apoptotic marker 7-Amino Actinomycin D (7AAD). In combination, cells were defined as dead (Hoestch-negative, 7AAD-bright), apoptotic (Hoechst-dim, 7AAD-dim), and live (Hoechst-bright, 7AADnegative). Sr.GAL4-positive cells or Rp298-positive cells were isolated by flow cytometry based on GFP fluorescence. Cells that expressed a fluorescent intensity greater than or equal to 103 for both GFP fluorescence and Hoechst fluorescence represented sorted tendon cell samples or sorted myotube samples, while cells that expressed a fluorescent intensity less than 103 for GFP fluorescence and greater than or equal to 103 for Hoechst fluorescence represented non-sorted control cells (Figure 6 A-B). 2x105 to 3x105 cells were collected for each sample. Tendon cell and myotube purification by FACS was performed three times for each sample in order to allow triplicate RNA-seq
! &)! analysis. RNA integrity FACS has made it possible to isolate specific cell types of interest. Sorted cells are used as a source of mRNA for characterization of gene expression at the moment of RNA extraction. However, isolating high quality RNA from FACS sorted tendon cells and myotubes has presented difficulties. RNA is sensitive to degradation by ubiquitous RNases. As a result of these enzymes, shorter fragments of RNA occur in a sample and can compromise results of downstream applications that can result in misrepresentation of gene expression levels (Imbeaud et al., 2005). The challenge of isolating high quality RNA is further amplified due to hydrodynamic stress that is imposed on cells when they Fig ure 6 Robust, tissue-specific expression of eGFP for FACS. (A B) Confoca l micrographs of St13 embryos. A tendon cell driver (A; Sr.Gal4) and a somatic muscle driver (B; RP298.Gal4) were used to express eGFP. Representative FACS plots are shown for each genotype Each cell is represented by a dot and positioned on the x and y scales according to the fluorescent intensities detected for that cell.
! &*! pass the flow cytometer nozzle. Although existing RNA isolation protocols have yielded RNA that is adequate for microarray and RT-PCR, its use in transcriptional profiling by RNA-seq is limited due to the requirement of purified RNA with a RNA Integrity Number (RIN) greater than eight. By sorting directly into Seecof saline and extracting RNA immediately after the purification of tendon cells or myotubes, I was able to greatly improve the yield of high quality RNA as demonstrated by high 28S and 18S ribosomal RNA peaks (Figure 7 E-H). RNA-sequencing To characterize the transcriptome of tendon cells and myotubes during muscle elongation, cDNA libraries were constructed using the Illumina TruSeq stranded-mRNA sample preparation kit. A library kit was prepared for each sample (3 tendon, 3 nonsorted tendon control, 3 myotubes, 3 non-sorted myotube control). A total of 300 million single-end 50-bp reads were generated using the Illumina HiSeq2000 platform. The sequence reads were then mapped to the Drosophila reference genome using Genomic Short-Read Nucleotide Alignment Program (GSNAP). RNA-seq analysis compared the expression levels of genes in a test condition with those in a control condition. The following comparisons were made: tendon cells vs. embryo, muscle cells vs. embryo, and muscle cells vs. tendon cells. Differences in read counts of each comparison will represent either true, biological differences within the comparison, or experimental noise. To ensure accuracy of RNA-seq data, mRNA biological triplicates were obtained and submitted for each condition. Multiple readings for test and control groups allowed us to accurately measure differential gene expression and detect the amount of variability present in our measurements.
! &+! Figure 7. Comparison of quality of RNA from FACS sorted tendon cells and myotubes. (A -H) Electropherograms from Bioanalyzer showing viability and RIN numbers from two independent sorting events. Analyses shown compare traditional methods (A-D) of RNA extraction to the newer methods (E-H) described in this thesis. RIN values range from 10 (intact) to 1 (totally degraded).
! &,! Principal-component analysis To visualize sample to sample differences within a group and between conditions we used principal-component analysis (PCA). PCA was performed on fragments per kilobase of exon per million mapped reads (FPKM) based expression levels (Figure 8). In the case of tendon vs. embryo (Figure 8A), samples are distinguishable between tendon and non-sorted control cells, however samples within a group display minor variability. Although we expect to see as little variability as possible within a group, these samples were extracted from Drosophila in vivo. What we see is representative of true gene expression. Based on PCA, biological replicates within a group display similar gen e expression profiles and could be further analyzed.
! &-! !"#$%&'()'*+,'-.'/0,1 2&3 $&45"4# '6787) '/0 123!45"67"89:!7;<8;6=6>!%!/42%3!?=8959>=@!?9<8:=?!AB!>C="5!:95D=?>! #95"967=!96@!5=?E:>=@!"6!?=8959>";6!A=>F==6!7=::!>B8=?G!45"67"89:!7;<8;6=6>!&!/42&3!?=8959>=@!?9<8:=?!AB!>C=! 6=$>!:95D=?>!#95"967=!96@!5=?E:>=@!"6!?=8959>";6!;H!?9<8:=?!F">C"6!7=::?!>B8=?G!'
! '.! Validation of RNA-sequencing data Our FACS RNA-seq approach successfully identified known myogenic transcription factors, muscle sarcomeric proteins, and known myotube guidance molecules from the muscle vs embryo data set, validating RNA-seq data (Table 1). Amo ng the list of identified muscle transcription factors is twist (FC=7.07), which is involved in the determination of mesodermal fate (Baylies and Bate, 1996). Transcripts that encode various identity genes that are expressed in founder cells were also present in our RNA-seq data. Nautilus (FC=4.51), slouch (FC=4.30), spalt major (FC=3.16), apterous (FC=2.32), and pointed (FC=2.31) transcripts were significantly enriched. These identity genes work in a combinatorial manner to regulate target genes that control muscle morphology (Dohrmann et al., 1990). RNA-seq results identified the known sarcomeric proteins Troponin C at 73F (FC=6.67), Troponin C at 47D (FC=5.5), sarcomere length short (FC=5.33), Tropomyosin 2 (FC=5.20), Actin 57B (FC=5.07), Myosin light chain (FC=4.94), bent (FC=4.85), sallimus (FC=4.41), and actinin (FC=3.76). Sarcomeric protein localization is responsible for the formation of sarcomeres, the contractile units of a myofibril (Rui et al., 2010). In addition, the most recent identified orphan receptor Donut on two (FC=2.05) and its RTK family member, derailed (FC=1.64) were identified in the RNA-seq data. Both coordinate with one another to facilitate guidance of the LT muscles to their correct tendon attachment sites (Lahaye et al., 2012). The RNA expression of known muscle transcripts was checked on the BDGP website for the transcripts that were available. The BDGP has documented the expression pattern for thousands of genes identified in the Drosophila genome (Tomancak et al., 2007). BDGP in situ hybridization (ISH) images for available known
! '%! muscle transcripts showed RNA expression in the somatic muscle, further validating our RNA-seq data (Figure 9). These results show that the FACS cell isolation techniques and RNA extraction methods used to acquire total RNA from tendon or myotube cells, allowed for accurate representation of the transcriptome of tendon cells and myotubes by RNA-seq. (FC) Fold Change from Muscle vs Embryo ! "#$% &'() ! *+' &'() ! ,-. &'() +/0 &'() ! &120 ! &'() /* ! &'() ! &'() 3-'$ 4*+56)7 ! &'() &1 & &'() ! Figure 9. Somatic muscle RNA expression pattern. RNA in situs of (A) nautilus, (B) slouch, (C) Myocyte enhancer factor 2, (D) apterous, (E) pointed, (F) Myocardin-related transcription factor, (G ) Troponin C at 73 F, (H) sallimus, (I ) Glutamate receptor binding protein from the Berkeley Drosophila Genome Project /IJK43!@9>9A9?=G! Table 1. Myotube RNA seq uencing results of known myogenic transcripts.
! '&! Bioinformatics analysis of enriched transcripts The Database for Annotation, Visualization and Integrated Discovery (DAVID) was used to facilitate the functional annotation and analysis of enriched transcripts from the RNA-seq data. DAVID uses data-mining tools that combine biological annotation data with graphical displays to promote discovery through functional classification and biochemical pathway maps (Huang et al., 2009). Using DAVID, I examined transcripts that showed increased enrichment greater than or equal to 1.5 fold change in tendon cells or myotubes compared to non-sorted control cells. The DAVID database performed Gene Annotation Enrichment Analysis on upregulated transcripts from each data set (tendon cells vs. embryo, muscle cells vs. embryo, and muscle cells vs. tendon cells). This analysis uses controlled vocabulary from the Gene Ontology Consortium (GO) to graphically display the distribution of differentially expressed genes among functional categories (Dennis et al., 2003). The analysis identified GO clusters of genes that contain significantly enriched signaling component associated GO terms (Figure 10). Differentially regulated genes in tendon cells represented GO terms related to vesiclemediated transport, lipid metabolic process, and intracellular protein transport (Figure 10 A,C). I then determined the classification of GO terms into biological processes. DAVID identified a number of Biological Function GO categories that were significantly enriched in our upregulated gene lists (Figure 11). Enriched GO categories represent functions necessary for myogenesis. In tendon cells, Biological Process GO categories identified genes related to signal transduction, vesicle-mediated transport, and transmembrane transport (Figure 11 A,C). Significantly enriched Biological Process GO
! ''! categories for myotubes include signal transduction, cell adhesion, cell migration, and muscle attachment (Figure 11 B,D). Through visualization of functional annotation clusters and GO Biological Process categories, many of the genes upregulated in tendon cells or myotubes encode regulators of transcription and components of signaling pathways. From the bioinformatics analysis, I selected and conducted further analysis on candidate transcripts that had the best combination of upregulation and GO terms. These transcripts were enriched greater or equal to 2 fold change and contained GO terms that represented signaling processes. I focused on regulators of the Sphingosine-1-Phosphate (S1P) and Fibroblast growth factor (FGF) signaling pathways.
! '(! Fig ure 10. Functional annotation c lustering of significantly enriched genes. Transcripts that were enriched 1.5 fold change were subjected to Functional Annotation Clustering using DAVID. For comparisons of Tendon vs Muscle (C) and Muscle vs Tendon (D), tendon transcripts represent genes enriched 1.5 fold change and muscle transcripts represent genes 1.5 fold change. The Gene Ontology (GO) term for each cluster is shown.
! ')! ! !! Fig ure 11 Enric hed GO biological process categories Transcripts were classified into functional groups based on GO a ssignments as predicted for thei r involvement in biological process. Upregulated GO biologic process categories were selected from (A) Tendon vs Embryo, ( B) Muscle vs Embryo, (C) Tendon vs Muscle, and (D) Muscle vs Tendon lists. For comparisons of Tendon vs Muscle (C) and Muscle vs Tendon (D), tendon transcripts represent genes enriched 1.5 fold change and muscle transcripts represent genes 1.5 fold change. Categories have a p value < 0.05.
! '*! Sphingosine-1-Phosphate (S1P) signaling S1P is a sphingolipid mediator that is derived by the metabolism of the membrane phospholipid, sphingomyelin (Hannun et al., 2001) (Figure 12). S1P signaling has been shown to play a role in several developmental process es such as cardiogenesis (Kupperman et al., 2000), limb development (Chae et al., 2004) and cell migration (Spiegel and Milstien, 2003). Signaling is achieved by activation of G protein coupled S1P receptors (Alexander et al., 2011). To facilitate cell migration, S1P binds to the S1P2 and S1P3 G protein coupled receptors (Spiegel and Milstien, 2003). Interestingly, transcripts within the lipid metabolic process cluster identified S1P-pathway signaling genes Ceramidase (FC=4.91), Sphingomyelin synthase-related 1 (FC=1.91), and Sphingosine-1-phosphate lyase (FC=2.55) (Table 2). Wunen2 (Wun2) (FC=2.07), a lipid phosphate phosphatase (LPP) was also enriched in our screen. Lipid phosphate dephosphorylation of S1P by wun2, forms a gradient that provides directional cues to migrating germ cells in the soma. Germ cells follow the gradient toward lipid phosphates, migrating away from wun2 (Renault et al., Table 2. Lipid metabolic proces s mRNA expression (FC) Fold Change from Tendon vs Embryo
! '+! 2004). S1P signaling pathway transcripts would make good transcript candidates to pursue, based on their identification in muscle and chemotactic influences. Fibroblast growth factor (FGF) signaling Fibroblast growth factors (FGFs) are secreted proteins that signal by binding to high affinity tyrosine kinase receptors (FGFRs) on the surface of a cell (Klint et al., 1999). In vertebrates, the FGF pathway plays a role in a variety of developmental and cellular processes such as migration and differentiation (Itoh and Ornitz, 2004; Huang and Stern, 2005). In Drosophila, this pathway has been demonstrated to a play a role in the development of Drosophila musculature (Shisido et al., 1997; Stathopoulos et al., 2004). !"#$%&'9:)'';"-2<48=&2"2'746'>&87?-@"2>'-.'AB="4#-2"4&191 *=-2B=78&'CA9*D)'0!?7C=<9>"7!@"9D59C=!H;5<9>";6!;H!L%4!AB! ?8C"6D;C"6!>C=!E85=DE:9>=@!D=6=!:"?>!H5;C= !H;::;F"6D!>596?75"8>?!F=5=!?"D6"H"796>:B!=65"7C=@Q!!"#!$%& /R2S*G%-3P & '%(!)*+!$%&/ ',!$% 3!/R2S(G-%3P &96@! -(!*./!$0*.1& / -/! 3!/R2S&G,)3G!/R23! R;:@!2C96D=G
! ',! Drosophila contains the two FGF receptors (FGFR), Breathless (Btl) and Heartless (Htl). The receptor Htl is essential for the development of the somatic meso derm (Beiman et al., 1996). FGFR Htl receptor tyrosine kinase activation by thisbe (ths) and pyramus (pyr) leads to activation of the RAS-RAF-MAPK pathway, resulting in mesoderm migration (Stathopoulos et al., 2004). FGF signaling pathway genes, branchless (FC=3.29), breathless (FC=2.72), and heartless (FC=3.89) were within the cluster associated with the GO term signaling pathway (Table 3). Enrichment of transcripts in the FGF pathway led me to test the possibility of pyr and ths as possible ligands involved in the process of myotube guidance. Pyr and ths have been shown to coordinate with Htl to facilitate mesoderm cell dorsal migration and dorsal spreading (McMahon et al., 2010). Previous experiments have also shown RNA expression of ths in the muscle attachment sites of St 15 embryos (Stathopoulos et al., 2004). To visualize expression of ths and pyr RNA, I performed fluorescent in situ hybridizations (Figure 13). Strikingly, these results showed robust, specific expression in tendon cells during St 13-15 embryos. These results are the first to show co-localization of a tendon marker with ths and pyr. Table 3. FGF signaling pathway mRNA expression /R23!R;:@!2C96D=!H5;
! '-! Another transcript within our tendon list that showed increased enrichment was CG42369 (FC=126.87). A possible human orthologue of this gene is FGF3. An alignment of the Drosophila CG42369 to the human FGF3 domain resulted in 16.8% identical sequences and 32.0% similar sequences (Figure 14). A previous study performed an alignment of the highly conserved FGF3 domain sequences of ascidians and humans and reported 20.6% identical sequences and 38.9% similar sequences (Satou et al., 2002). While the percent identity may seem low, Ciona intestinalis FGF3 proteins share a rare conserved common motif that precedes the highly conserved FGF3 domain, suggesting Ci-FGF3 is an orthologue of the vertebrate FGF3 (Satou et al., 2002). Based on this data, FGF3 is a potential candidate of being another activator of the FGF pathway. Figure 13. Pyramus (Pyr) and Thisbe (Ths) are expressed in tendon cells. Sr.gal4>eGFP embryos co-labeled for Ths (A, C, E; green) and Pyr (B,D,F; green). Pyr and Ths colocalize to Stripe (Sr) (A-F, yellow, white arrowheads). ! ! ! ! ! ! ! ! !
! (.! ! !"#$%&'9E)',@"#4>&48'-.'!"#$#%&'()*+FE:GHI'746'=2'!F!G)'2:E?>9:!U!F9?! E?=@!>;!9:"D6!?=VE=67=?G!I:E=!A;$=@!5=D";6?!5=85=?=6>!9!595=!7;6?=5#=@! RKR'!<;>"H!96@!5=@!A;$=@!5=D";6?!5=85=?=6>!9!C"DC:B!7;6?=5#=@!RKR!@;<9"6G! K5==6!A;$=@!5=D";6?!5=85=?=6>!9!?"D69:!8=8>"@=G!I:97W!?C9@=@!9<"6;!97"@?! 95=!"@=6>"79:!96@!D5=B!?C9@=@!9<"6;!97"@?!95=!?"<":95G!
! (%! CHAPTER V DISCUSSION Conclusions Until recently DNA microarrays have been the technology of choice for whole genome analysis studies in muscle (Dobi et al., 2014) and tendon cells (Havis et al., 2014). Advances in next generation sequencing technologies have changed the way we can analyze a population of cells. RNA-seq is an efficient tool and has been shown to identify new transcripts and increase transcript sensitivity (Marguerat and Bahler, 2010). Using a combination of FACS, RNA-seq, and computational bioinformatics analysis, I was able to successfully purify tendon cells and myotubes from Drosophila embryos and perform transcriptional profiling on these cells. My RNA-seq analysis identified a number of known muscle transcripts, sarcomeric proteins, and molecular guidance molecules (Table 1), which validated our approach. Within the generated data, we also saw enrichment of transcripts belonging to the S1P and FGF signaling pathways. In recent years, the role of signal transduction pathways in regulating developmental processes has become apparent. Upregulated transcripts belonging to the S1P signaling pathway encode enzymes that generate the S1P ligand (Figure 12). S1P transcripts such as wun2, have known roles in chemotaxis. In the example of germ cell migration, after specification of primordial germ cells, germ cells migrate through and along somatic tissue to the gonad (Starz-Gaiano and Lehmann, 2001; Wylie, 1999). Proper migration to somatic gonadal precursors from the midgut relies on wun2 expression (Starz-Gaiano et al., 2001). wun2 expression in somatic cells hydrolyzes S1P which causes a gradient that facilitates germ cell migration away from
! (&! somatic cells that express wun2 and toward lipid phosphates (Renault et al., 2004). The repulsion mechanism in which wun2 facilitates proper germ cell migration could also underlie the mechanism of muscle elongation. A possibility is that wun2 in myotubes repulse the leading edge of a myotube towards tendons that are secreting lipid phosphates. It will be interesting to see if wun2 has a role in myotube guidance. FGF signaling contains tyrosine kinase receptors that are capable of autophosphorylation as well as phosphorylation of other substrates (Furdui et al., 2006). FGF signaling pathways have been shown to play roles in developmental processes such as tracheal branching through the FGFR Btl and its FGF ligand Bnl (Sutherland et al., 1996) and mesoderm migration through the FGFR Htl and its ligands Pyr and Ths (Stathopoulos et al., 2004). Pyr and Ths expression in tendon showed co-localization with the tendon marker Sr (Figure 13). To our knowledge, this is the first study to show this expression. These results may have important implications suggesting that FGF ligands are not only involved in the chemotaxis of the early mesoderm, but perhaps in myotube elongation as well. In vertebrates, FGF signaling is far more complex compared to FGF signaling in Drosophila, which contains only two receptors and three characterized ligands. The Htl ligands Pyr and Ths remained unknown until fairly recently (Stathopoulos et al., 2004) and still the possibility of undiscovered ligands for these receptors exist. Our RNA-seq data identified a novel transcript, CG42369 that showed high conservation with the hsFGF3 sequence (Figure 14). CG42369 was significantly enriched (FC=126.87) and could be a possible unknown ligand of the Htl receptor and a candidate regulator of myotube elongation.
! ('! Vertebrate muscle development shares fundamental similarities with Drosophila. Findings in one organism will advance our knowledge of muscle development in the other, giving insight into human conditions and diseases that are characterized by abnormal muscle development. This data set is the first to directly compare the transcriptome of Drosophila tendon cells and myotubes. The data produced from this project will be valuable in setting the foundation for further investigations of candidate transcripts that play a role in myotube guidance. The identification of these transcripts will give us a better understanding of the genes affected in patients with congenital myopathies and will provide information on the pathological processes underlying these diseases.
! ((! CHAPTER VI FUTURE DIRECTIONS Further directions with RNA-sequencing data The transcripts that show robust expression in tendon cells or muscle during St 12-13 will be subjected to mutational analysis. We will first test mutations that are currently available. Mutant lines will be obtained from the Bloomington Drosophila Stock Center (BDSC) or from the Drosophila community. To test if candidate transcripts function during muscle elongation, we will assay muscle morphology by Myosin Heavy Chain (MHC) expression in homozygous mutant embryos (Figure 15). Drosophila has a single MHC gene that encodes all sarcomeric heavy chain polypeptides (Bernstein et al., 1983). Homozygous mutants that demonstrate a tissue specific mechanism will display a phenotype in which muscles fail to elongate. We expect mutational analysis will validate our results and uncover novel mechanisms that regulate myotube elongation during muscle development. For mutations that are unavailable, we will use the clustered regularly interspaced short palindromic repeat (CRISPR/Cas9) system for targeted mutagenesis in Drosophila. CRISPR-mediated genome editing is induced by injecting plasmids encoding target specific guide RNAs (gRNAs) into Cas9 expressing embryos (Sebo et al., 2014). The specificity of this system is achieved by a CRISPR RNA (crRNA) that base pairs with a 20 nucleotide complimentary sequence within the DNA. Cas9 endonuclease is targeted to this sequence causing double strand breaks (DSB) in the DNA (Gasiunas et al., 2012).
! ()! The CRISPR/Cas9 system has emerged as an efficient alternative to traditional techniques such as gene targeting by homologous recombination (HR) and transcription activator-like effector nuclease (TALEN) which have low efficiency and are labor intensive (Rong and Golic, 2000; Liu et al., 2012). The CRISPR/Cas9 system is an attractive tool because the simplicity at which it can be reprogrammed to target different sites enables large-scale application of this technology and allows for the simultaneous generation of mutations in multiple genes (Basset and Liu, 2013). Figure 15. MHC expression in St 17 wildtype embryos. (A -B) By St17, somatic muscles express MHC.
! (*! REFERENCES Alexander, S.P., Mathie, A., Peters, J.A. (2011). Guide to receptors and channels. GRAC, 5th edition. Br. J. Pharmacol. 164, S1-S324. Anderson, C., Williams, V.C., Moyon, B., Daubas, P., Tajbakhsh, S., Buckingham, M.E., Shiroishi, T., Hughes, S.M., Borycki, A.G. (2012). Sonic hedgehog acts cellautonomously on muscle precursor cells to generate limb muscle diversity. Genes & Development. 26, 2103-2117. Baird, N.L., Bowlin, J.L, Cohrs, R.J., Gilden, D., Jones, K. (2014). Comparison of Varicella-Zoster virus RNA sequences in human Neurons and fibroblasts. J. Virol. 88, 10. Bassett, A.R. and Liu, J. (2013). CRISPR/CAS9 and Genome Editing in Drosophila. Journal of Genetics and Genomics. 41, 7-19. Bate, M. (1990). The embryonic development of larval muscles in Drosophila. Development. 110, 791-804. Bate, M. (1993). The mesoderm and its derivatives. The Development of Drosophila melanogaster. New York: Cold Spring Harbor Press. 1013-1090. Baylies, M.K. and Bate, M. (1996). twist: a myogenic switch in Drosophila. Science. 272, 1481-1484. Baylies, M.K., Bate, M., Ruiz Gomez, M. (1998). Myogenesis: A view from Drosophila. Cell. 93, 921-927. Becker, S., Pasca, G., Strumpf, D., Min, L., Volk, T. (1997). Recipricol signaling between Drosophila epidermal muscle attachment cells and their corresponding muscles. Development. 124, 2615-2622. Beiman, M., Shilo, B.Z., Volk, T. (1996). Heartless, a Drosophila FGF receptor homolog, is essential for cell migration and establishment of several mesodermal lineages. Genes and Dev. 10, 2993-3002. Bernstein, S., Mogami, K., Donady, J., Emerson, C. (1983). Drosophila myosin heavy chain is encoded by a single gene located in a chromosomal region of other muscle genes. Nature. 309, 393-397. Bismuth, K. and Relaix, F. (2010). Genetic regulation of skeletal muscle development. Experimental Cell Research. 316, 3081-3086.
! (+! Bour, B.A., Chakravarti, M., West, J.M., and Abmayr, S.M. (2000). Drosophila SNS, a member of the immunoglobulin superfamily that is essential for myoblast fusion. Genes Dev. 14, 1498-1511. Brand, A. and Perrimon, N. (1993). Targeted gene expression as a means of altering cell fate and generating dominant phenotypes. Development. 118, 401-415. Bryantsev, A.L. and Cripps, R.M. (2012). Purification of cardiac cells from Drosophila embryos. Methods. 56, 44-49. Callahan C.A., Bonkovsky, J.L., Scully, A.L., Thomas, J.B. (1996). derailed is required for muscle attachment site selection in Drosophila. Development. 122, 2761. Callahan, C.A., Muralidhar, M.G., Lundgren, S.E., Scully, A.L., Thomas, J.B. (1995). Control of neuronal pathway selection by a Drosophila receptor protein-tyrosine kinase family member. Nature. 376,171. Carmena, A., Bate, M., Jimenez, F. (1995). Lethal of scute, aproneural gene, participates in the specification of muscle progenitors during Drosophila embryogenesis. Genes Dev. 9, 2373-2383. Carmena, A. and Baylies, M. (2006). Development of the Larval Somatic Musculature. Muscle Development in Drosophila. 79-91. Chae, S.S., Paik, J.H., Allende, M.L., Proia, R.L., Hla, T. (2004). Regulation of limb development by sphingosine-1-phosphate receptor S1p1/EDG-1 occurs via the hypoxia/VEGF axis. Dev. Biol. 268, 441-447. Chevallier, A., Kieny, M., Mauger, A. (1997). Limb-somite relationship origin of the limb musculature. J. Embryol. Exp. Morphol. 41, 245-258. Christ, B. and Ordahl, C.P. (1995). Early stages of chick somite development. Anat. Embryol. 191(5), 381-396. Dennis, G.J., Sherman, B.T., Hosack, D.A., Yang, J., Gao, W., Lane, H.C., Lempicki, R.A. (2003). DAVID: Database for Annotation, Visualization, and Integrated Discovery. Genome Biology. 4:R60. Dobi, K.C., Halfon, M.S., Baylies, M.K. (2014). Whole genome analysis of muscle founder cells implicates the chromatin regulator Sin3A in muscle identity. Cell Rep. 8(3), 858-870. Dohrmann, C., Azpiazu, N., Frasch, M. (1990). A new Drosophila homeobox gene is expressed in mesodermal precursor cells of distinct muscles during embryogenesis. Genes Dev. 4, 2098-2111.
! (,! Fischer, J.A., Giniger, E., Maniatis, T., Ptashne, M. (1988). GAL4 activates transcription in Drosophila. Nature. 332, 853-865. Frasch, M. (1999). Controls in patterning and diversification of somatic muscles during Drosophila embryogenesis. Curr. Opin. Genet. Dev. 9, 522-529. Frommer, G., Vorbruggen, G., Pasca, G., Jaeckle, H., Volk, T. (1996). Epidermal egr-like zinc finger protein of Drosophila participates in myotube guidance. EMBO. 15, 16421649. Furdui, C.M., Lew, E.D., Schlessinger, J., Anderson, K.S. (2006). Autophosphorylation of FGFR1 kinase is mediated by a sequential and precisely ordered reaction. Mol Cell. 21, 711. Gasiunas, G., Barrangou, R., Horvath, P., Siksnys, V. (2012). Cas9-crRNA ribonucleoprotein complex mediates specific DNA cleavage for adaptive immunity in bacteria. Proc. Natl. Acad. Sci. 109, E2579-E2586. Hannun, Y.A., Luberto, C., Argraves, K.M. (2001). Enzymes of sphingolipid metabolism: from modular to integrative signaling. Biochemistry. 40, 4893-4903. Hartenstein, V. (2006). The Muscle Pattern of Drosophila. Muscle Development in Drosophila. 8-27. Havis, E., Bonnin, M., Olivia-Martines, I., Nazaret, N., Ruggiu, M., Weibel, J., Durand, C., Guerquin, M., Bonod-Bidaud, C., Ruggiero, F., Schweizter, R., Duprez, D. (2014). Transcriptomic analysis of mouse limb tendon cells during development. Development. 141, 3683-3696. Hu, J.K.H., McGlinn, E., Harfe, B.D., Kardon, G., Tabin, C. (2012). Autonomous and nonautonomous roles of hedgehog signaling in regulating limb muscle formation. Genes & Development. 26, 2088-2102. Huang, G.W., Sherman, B.T., Lempicki, R.A. (2009). Systematic and integrative analysis of large gene lists using DAVID Bioinformatics Resources. Nature Protoc. 4(1), 44-57. Huang, P. and Stern, M.J. (2005). FGF signaling in flies and worms: more and more relevant to vertebrate biology. Cytokine Growth Factor Rev. 16, 151. Imbeaud, S., Graudens, E., Boulanger, V., Barlet, X., Zaborski, P., Eveno, E., Mueller, O., Schroeder, A., Auffray, C. (2005). Towards standardization of RNA quality assessment using user-independent classifiers of microcapillary electrophoresis traces. Nucleic Acids Research. 5, 33-56. Itoh, N. and Ornitz, D.M. (2004). Evolution of the Fgf and Fgfr gene families. Trends Genet. 20, 563.
! (-! Kardon, G. (1998). Muscle and tendon morphogenesis in the avian hind limb. Development. 125, 4019-4032. Kidd, T., Bland, K.S., Goodman, C.S. (1999). Slit is the Midline Repellent for the Robo Receptor in Drosophila. Cell. 96, 785-794. Kim, J.H., Ren, Y., Ng, W.P., Li, S., Son, S., Kee, Y.S., Zhang, S., Zhang, G., Fletcher, D.A., Robinson, D.H., Chen, E.H. (2015). Mechanical tension drives cell membrane fusion. Developmental Cell. 32, 561-573. Klint, P. and Clasesson-Welsh, L. (1999). Signal transduction by fibroblast growth factor receptors. Front Biosci. 4D, 165-177. Knirr, S., Azpiazu, N., Frasch, M. (1999). The role of the NK-homeobox gene slouch (S59) in somatic muscle patterning. Development. 126(20), 4525-4535. Kramer, S.G., Kidd, T., Simpson, J.H., Goodman, C.S. (2001). Switching repulsion to attraction: changing responses to slit during transition in mesoderm migration. Science. 292 (5517), 737-740. Kupperman, E., An, S., Osborne, N., Waldron, S., Stainer, D.Y. (2000). A sphingosine-1phosphate receptor regulates cell migration during vertebrate heart development. Nature. 406, 192-195. Lahaye, L.L, Wouda, R.R., de Jong, A.W.M., Fradkin, L.G., Noordermeer, J.N. (2012). WNT5 Interacts with the Ryk Receptors Doughnut and Derailed to Mediate Muscle Attachment Site Selection in Drosophila melanogaster. PLo S One. 7(3), e32297. Liu, J., Li, C., Yu, Z., Huang, P., Wu, H., Wei, C., Zhu, N., Shen, Y., Chen, Y., Zhang, B., Deng, W., Jiao, R. (2012). Efficient and specific modifications of the Drosophila genome by means of an easy TALEN strategy. Journal of Genetics and Genomics. 39(5), 209-215. Marguerat, S. and Bahler, J. (2010). Rna-seq: From technology to biology. Cell Mol. Life Sci. 67, 569-579. McMahon, A., Reeves, G.T., Supatto, W., Stathopoulos, A. (2010). Mesoderm migration in Drosophila is a multi-ste p process requiring FGF signaling and integrin activity. Development. 137, 2167-2175. Renault, A.D., Sigal, Y.J., Morris, A.J., Lehmann, R. (2004). Soma-germ line competitition for lipid phosphate uptake regulates germ cell migration and survival. Science. 305(5692), 1963-1966.
! ).! Rong, Y.S. and Golic, K.G. (2000). Gene Targeting by Homologous Recombination in Drosophila. Science. 288, 2013-2018. Rothberg, J.M., Jacobs, J.R., Goodman, C.S. (1990). slit: an extracellular protein necessary for the development of midline glia and commissural axon pathways contains both EGF and LRR domains. Genes & Development. 4, 2169-2187. Rui, Y., Bai, J., Perrimon, N. ( 2010). Sarcomere formation occurs by the assembly of multiple latent protein complexes. PLoS Genet. 6(11): e1001208. Ruiz-Gomez, M., Coutts, N., Price, A., Taylor, M.V., Bate, M. (2000). Drosophila dumbfounded: a myoblast attractant essential for fusion. Cell. 102, 189-198. Ruiz-Gomez, M., Romani, S., Hartmann, C., Jackle, H., Bate, M. (1997). Specific muscle identities are regulated by Kruppel during Drosophila embryogenesis. Development. 124(17), 3407-3414. Satou, Y., Imai, K.S., Satoh, N. (2002). Fgf genes in the basal chordate Ciona intestinalis. Dev. Genes Evol. 212, 432-438. Schejter, E.D. and Baylies, M.K. (2010). Born to run: creating the muscle fiber. Curr. Opin. Cell Bio. 22(5), 566-574. Schnorrer, F. and Dickson, B.J. (2004). Muscle building: mechanisms of myotube guidance and attachment site selection. Developmental Cell. 7, 9-20. Schnorrer, F., Kalchhauser, I., Dickson, B.J. (2007). The transmembrane protein Kon-tiki couples to Dgrip to mediate myotube targeting in Drosophila. Developmental Cell. 12(5), 751-766. Schweitzer, R., Chyung, J.H., Murtaugh, L.C., Brent, V.R., Olson, E.N., Lassar, A., Tabin, C.J. (2001). Analysis of the tendon cell fate using Scleraxis, a specific marker for tendons and ligaments. Development. 128, 3855-3866. Schweitzer, R., Zelzer, E., Volk, T. (2010). Connecting muscles to tendons: tendons and musculoskeletal development in flies and vertebrates. Development. 137(17), 2807-2817. Sebo, Z., Lee, H.B., Peng, Y., Guo, Y. (2014). A simplified and efficient germlinespecific CRISPR/Cas9 system for Drosophila genomic engineering. Fly. 8;1, 52-57. Shishido, E., Ono, N., Kojima, T., Saigo, K. (1997). Requirements of DFR1/ Heartless, a mesoderm-specific Drosophila FGF-receptor, for the formation of heart, visceral and somatic muscles, and ensheathing of longitudinal axon tracts in CNS. Development. 124, 2119.
! )%! Sieiro-Mosti, D., De La Calle, M., Pele, M., Marcelle, C. (2014). A dynamic analysis of muscle fusion in the chick embryo. Development. 141, 3605-3611. Spiegel, S. and S. Milstien. (2003). Sphingosine-1-phosphate: an enigmatic signalling lipid. Nat. Rev. Mol. Cell Biol. 4, 397. Starz-Gaiano, M., Cho, N.K., Forbes, A., Lehmann, R. (2001). Spatially restricted activity of a Drosophila lipid phosphatase guides migrating germ cells. Development. 126, 983-991. Starz-Gaiano, M. and Lehmann, R. (2001). Moving towards the next generation. Mech. Dev. 105, 5-18. Stathopoulos, A., Tam, B., Ronshaugen, M., Frasch, M., Levine, M. (2004). pyramus and thisbe: FGF genes that pattern the mesoderm of Drosophila embryos. Genes Dev. 18, 687. Sutherland, D., Samakovlis, C., Krasnow, M.A. (1996). branchless encodes a Drosophila FGF homolog that controls tracheal cell migration and the pattern of branching. Cell. 87, 1091-1101. Taylor, M.V. (2006). Comparison of Muscle Development in Drosophila and Vertebrates. Muscle Development in Drosophila. 169-203. Tomancak, P., Berman, B.P., Beaton, A., Weiszmann, R., Kwan, E., Hartenstein, V., Celniker, S.E., Rubin, G.M. (2007). Global analysis of patterns of gene expression during Drosophila embryogenesis. Genome Biol. 8(7), R145. Trapnell, C., Roberts, A., Goff, L., Pertea, G., Kim, D., Kelley, D.R., Pimentel, H., Salzberg, S.L., Rinn, J.L, Pachter, L. (2012). Differential gene and transcript expression analysis of RNA-seq experiments with TopHat and Cufflinks. Nature Protocols. 7, 562578. Vargesson, N., Luria, V., Messina, I., Erskine, L., Laufer, E. (2001). Expression patterns of Slit and Robo family members during vertebrate limb development. Mechanisms of Development. 106, 175-180. Wigmore, P.M. and Evans, D.J. (2002). Molecular and cellular mechanisms involved in the generation of fiber diversity during myogenesis. Int Rev. Cytol. 216, 175-232. Wu, T.D. and Nacu, S. (2010). Fast and SNP-tolerant detection of complex variants and splicing in short reads. Bioinformatics. 26, 873. Wylie, C. (1999). Germ cells. Cell. 96, 165-174.
! )&! Yarnitzky, T., Min, L., Volk, T. (1997). The Drosophila neuregulin homolog Vein mediates inductive interactions between myotubes and their epidermal attachment cells. Genes Dev. 11, 2691-2700.