Avian Muscle Growth and Development Thesis Presented In

Avian Muscle Growth and Development

Thesis

Presented in Partial Fulfillment of the Requirements for the Degree Master of Science in the Graduate School of The Ohio State University

Jacqueline R. Griffin, B.S.

Graduate Program in Animal Sciences

The Ohio State University

2014

Thesis Committee:

Michael S. Lilburn, Advisor

Macdonald Wick

Michael E. Davis

Copyrighted by

Jacqueline R Griffin

2014

Abstract

Numerous factors can influence avian embryonic development resulting in a relatively low correlation between chronological age (incubation days) and physiological stage of development. The prevailing staging system is based on visual morphological embryonic characteristics to establish developmental stages that are independent of chronological age (incubation days) and embryo size (Hamburger and Hamilton, 1951).

In addition to morphological staging, it is fundamentally important to define a staging system for temporal transcriptional events so as to better understand the fundamental molecular biological mechanisms that are responsible for embryonic skeletal myogenesis.

The developmental fast skeletal myosin isoforms (MyHC), the predominant proteins in the Pectoralis major (PM), are expressed as a cadre of highly specific temporal and spatial developmental isoforms. Our hypothesis is that the temporal transcription of

MyHC isoforms is correlated with the transcription of muscle-specific genes that are critical to PM muscle growth and development and can be used as molecular markers during muscle development. To test this hypothesis, our primary goal was to use a novel molecular method, based on a quantitative analysis of the transcriptome, to characterize the developmental stages in embryonic PM in the SCWL and compare the data with what has been reported in the literature (Tidyman et al., 1997). Results confirmed that the pattern of temporal transcription of the MyHC isoforms obtained in this study was consistent with Tidyman et al. (1997), thus giving us the tool we need to explore our

hypothesis that the temporal transcription of the MyHC isoforms can be used as molecular markers for muscle development. To test the hypothesis, we used embryos from a control Single Comb White Leghorn (SCWL) line and an intermediate dystrophic

Low Score Normal (LSN) line that exhibits breast muscle anomalies post-hatch. Tissue samples from the embryonic PM were collected at embryonic days (ED) 5 through 19 from the 2 unique poultry lines. Total RNA was isolated and gene transcription was quantified using NanoString Technologies®, which digitally detects and quantifies selected target mRNA. Data were analyzed using the LOESS smoothing function at a

95% confidence level. Results from this study revealed differences in the temporal transcription of the 3 embryonic MyHC isoforms, Cemb1, 2 and 3, and the neonatal

MyHC isoform, Cneo in both SCWL and LSN embryos. Line differences were primarily due to shifts in peak transcription, with significant changes occurring between the SCWL and LSN at ED14-15, consistent among all MyHC isoforms. In addition, there were genetic line differences in the quantitative temporal transcription of 4 known muscle- specific genes, decorin, MTR2, Myf6 (MRF4) and Six4. Furthermore, there was a positive correlation between Cemb1 and Myf6 (MRF4) in both the SCWL and LSN from

ED5 through ED19. These data allowed us to describe the temporal transcription of regulatory factors at identical cellular development stages in the 2 unique genetic lines.

If this approach is applied to broiler genotypes, it could lead to new, enhanced selection tools or aid in identifying the developmental points at which factors such as incubation environment influence the earliest stages of myogenesis.

iii

Vita

2012………………………………...... ……B.S. Animal Sciences, The Ohio State

University

2012 to present………………………..…….Graduate Teaching Associate, Department of

Animal Sciences, The Ohio State University

Fields of Study

Major Field: Animal Science

Table of Contents

Abstract……………………………………………………………………………...... ii

Vita………………………………………………………………………………….….....iv

List of Tables……………………………………………………………………………..vi

List of Figures……………………………………………………………………………vii

Chapter 1: Review of the Literature...……………………………………………………..1

Chapter 2: Molecular Staging based on the Temporal Myosin Heavy Chain

Transcription.…………………………………………………………………………….19

Chapter 3: Transcriptional Events Underlying Avian Myogenesis……………………...31

References……………………………………………………………………………..…42

List of Tables

Table 1.1. Specificity of monoclonal antibodies specific to four MyHC isoforms…….....6

Table 1.2. Outline of all 33 target genes used in this experiment……….……………….17

Table 3.1. Outline of reported genes, their corresponding NCBI reference, and target

sequence…………….………………………………………………………..34

List of Figures

Figure 1.1. Temporal transcription of myosin as five developmental isoforms expressed

during embryonic and post-hatch growth in the Single Comb White Leghorn

chicken………………………………………………………………………...1

Figure 1.2. Distribution of antibody epitopes on a single myosin molecule…….……...... 5

Figure 1.3. Schematic of the temporal differences in the expression of the myosin heavy

chain isoforms in slow growing egg layer and fast growing broilers selected

for increased breast yield…………………………………………….………..6

Figure 1.4. Delayed expression of the adult MyHC isoform expression in restricted fed

poults……………………………………………………………………..…..7

Figure 2.1. Diagram of NanoString® probe architecture used for nCounter® gene

transcription analysis system…………………………………….…….……25

Figure 2.2. Relative transcription of the developmental myosin isoforms in the SCWL

previously published by Tidyman et al. (1997) (Plot A) and in the SCWL

(Plot B) taken from preliminary data…..…………………………………..28

Figure 2.3. Relative transcription of the developmental myosin isoforms in the SCWL

at embryonic days 5 through 19……………………………………………29

Figure 3.1. Quantitative transcription of Cemb1 in the SCWL and LSN at embryonic

days 6 through 19……………….……………………………….…………35

vii

Figure 3.2. Quantitative transcription of Cemb2 in the SCWL and LSN at embryonic

days 7 through 19……………….…..……………………………………...35

Figure 3.3. Quantitative transcription of Cemb3 in the SCWL and LSN at embryonic

days 6 through 19………………………………….………………...……36

Figure 3.4. Quantitative transcription of Cneo in the SCWL and LSN at embryonic days 7

through 19……………………………………………..……….………….36

Figure 3.5. Quantitative transcription of Cemb1 and MRF4 in the SCWL and LSN at

embryonic days 5 through 19…………………………………………..…36

Figure 3.6. Quantitative transcription of Six4 in the SCWL and LSN at embryonic days 5

through 18…………………………………………….……………………..39

Figure 3.7. Quantitative transcription of MTR2 in the SCWL and LSN at embryonic days

5 through 18…………………...………………….…………………………39

Figure 3.8. Quantitative transcription of decorin in the SCWL and LSN at embryonic

days 5 through 18..…………………………………...……………………..40

viii

Chapter 1: Review of the Literature

The research described herein is of great practical importance to the poultry industry because an estimated 9 billion broiler chickens are raised annually with consumer expenditures for chicken at approximately $70 billion annually (National

Chicken Council, 2012). Within the poultry industry, broiler chickens have been continually selected for increased body weight and proportionately larger breast muscles.

The economic value and sustainability of the poultry industry is based primarily on the growth of a single muscle, the Pectoralis major (PM) breast muscle. The PM is the predominant breast muscle in poultry and the continued selection for breast muscle size has increased the potential for correlated negative effects on cooked muscle products.

There is a driving force within the broiler industry to increase breast yield while, at the same time, decrease processing costs and potential meat quality problems, thus emphasizing the importance of understanding those physiological factors underlying its development.

There have been a number of papers that have Figure 1.1. Temporal addressed the development of the PM post-hatch; however, transcription of myosin as five developmental isoforms the embryonic transcriptional events that control expressed during embryonic and post-hatch growth in the Single Comb White Leghorn myogenesis and ultimately the size and functionality of the chicken (Tidyman et al., 1997). PM remain largely unresolved. The PM in poultry is made

up exclusively of fast-twitch white fibers. The growth and development of the PM is associated with the temporal expression of fast skeletal muscle specific proteins, predominantly myosin. Fast skeletal muscle myosin is a hexameric protein composed of 2 identical heavy chains (MyHC) and 4 light chains. In chickens the myosin heavy chains are temporally expressed as a series of 6 developmental isoforms.

In Figure 1.1. the temporal transcription of the fast skeletal muscle MyHC in the PM are shown in sequence: 1) embryonic I (Cemb I), 2) embryonic II (Cemb II), 3) embryonic

III (Cemb III), 4) neonatal (Cneo) and 5) adult (Cadult; Tidyman et al., 1997).

While many embryonic research studies involve simply setting eggs in the incubator at the same time followed by sampling at chronological days thereafter, there are several shortcomings associated with simply comparing embryos at similar days of incubation. There is inherent variability in embryonic development due to a number of factors including egg size, normal variability in temperature within the incubator and age of the hen. These and other factors thus contribute to the relatively low correlation between chronological age (incubation days) and physiological stage of development

(Hamburger and Hamilton, 1951). This also does not include any potential genetic differences in the rate of development between different strains that have been used to study avian embryonic development.

Currently, there are no known biomarkers for the transcriptional events that are related to myogenesis and meat quality that could be used for improved genetic selection tools that could be critically important to the economic viability of poultry producers.

Precise knowledge of these transcriptional events could lead to the identification of biomarkers for the selection of breed stock that give rise to offspring with desired 2

quantitative traits such as increased breast yield without the tradeoff of processed product quality issues. Thus, in order to better understand the molecular and cellular events that drive skeletal muscle development, this study used a comparative development approach in transcriptomic analysis of the embryonic PM from 2 unique poultry lines: 1) the normal Single Comb White Leghorn (SCWL) and an intermediate dystrophic line, Low

Score Normal (LSN) that exhibits muscle aberrations post-hatch.

1.1. Avian Models of Muscle Growth

The chick embryo has been used for growth and development studies since

Aristotle first discovered the chick embryo as the ideal object for embryologic studies

(Hamburger and Hamilton, 1951). Hamburger and Hamilton (1951) used SCWL embryos to develop a staging system for chick embryonic development based on external embryo characteristics at similar developmental stages. This system is independent of chronological age (incubation days) and size of the embryo. Due to the overwhelming amount of traditional literature using the SCWL for studies on embryogenesis and myogenesis, it was decided that the SCWL would be the most appropriate control line for use in our studies.

Inherited muscular dystrophy in chickens was first observed in a commercial flock of New Hampshire chickens and birds exhibiting this anomaly were sent to the

University of California at Davis in 1954 (Julian, 1973; Wilson et al., 1979). The first report on this condition was in 1956 after it was determined to be inherited as an autosomal recessive trait (Julian, 1973).

Two lines were developed at the University of California at Davis from the original New Hampshire dystrophic chickens, lines 304 and 307. Line 304 exhibited 3

early pectoral muscle hypertrophy as a result of selection for early onset of the anomaly, whereas, in Line 307, early pectoral muscle atrophy was observed as a result of selection for high lipid content. Line 304 was crossed with an inbred normal SCWL line (03) and their progeny were repeatedly backcrossed with Line 03, resulting in Line 433 that exhibits early muscle fiber atrophy (Wilson et al., 1979).

The LSN genetic line, used in the experiments reported herein, was established by outcrossing the original New Hampshire dystrophic chickens (Line 301) with the SCWL and this line was sent to the University of Connecticut (Wilson et al., 1979). In 1979, this line was given the name Low Score Normal (LSN) to distinguish these chicks from normal chicks based on their inability to right themselves after being repeatedly placed on their backs (exhaustion score). At 2 to 3 months of age, normal chicks have an exhaustion score of 15 to 20 whereas LSN chicks had exhaustion scores of 6 or fewer

(Velleman et al., 2001). In addition to having a weakened PM, the LSN phenotype has decreased PM muscle mass when compared to the SCWL line (Velleman et al., 1993;

1996). At 1 week post-hatch, the weight of the PM, expressed as a percentage of body weight, was 1.9% in LSN chicks versus 3.2% in normal birds (Velleman et al., 2001).

In addition to phenotypic differences post-hatch, the LSN exhibits transcriptional and morphological differences, in vitro. A study comparing transcriptional events in fertilized LSN and SCWL eggs revealed an altered expression of myogenic events in the

LSN with an increased expression of TGF-β1 and decreased expression of MyoD at

17ED when compared to the SCWL (Li et al., 2009). LSN satellite cells exhibit an impaired ability to fuse and decreased proliferation and differentiation compared to normal satellite cells (Li et al., 1997). In addition, cultured satellite cells from the LSN 4

exhibit a decrease in myotube length and number of nuclei per myotube (Velleman et al.,

1999). In vitro studies show delayed myogenesis in the LSN when compared to the

SCWL, characterized by delayed expression of the ventricular MyHC isoform in satellite cells (Wick et al., 2003).

These observations make the LSN a useful model for studying muscle development due to the known phenotypic differences post-hatch (i.e., exhaustion score; significant reduction in muscle mass). The LSN also serves as a useful model for more in depth studies on the transcriptional events underlying embryonic myogenesis due to the delayed expression of myogenic events and vMyHC in conjunction with altered myotube morphology in LSN satellite cells when compared to normal SCWL satellite cells.

1.2. Myosin

Myosin is the major structural protein of the contractile apparatus found in the PM muscle, the predominant breast muscle in poultry. Myosin diversity is primarily the result of the differential expression of multiple Figure 1.2. Distribution of antibody isoforms of the myosin heavy chain (MyHC) epitopes on a single myosin molecule (Moore et al., 1992). subunit (Moore et al., 1992; Tidyman et al.,

1997). The temporal and tissue specific expression of the MyHC isoforms during PM muscle development in poultry has been well documented at the protein level using immunochemical analyses (Bandman, 1985; Cerny and Bandman, 1987; Bandman and

Bennett, 1988; Maruyama et al. 1993) and quantitative northern blot analyses (Tidyman

et al., 1997). The temporally expressed MyHC isoforms are shown in Figure 1.1. and their approximate epitope locations in Figure 1.2.

Detection of MyHC isoforms during embryogenesis and post-hatch has primarily been done using 5 fast MyHC isoform specific monoclonal antibodies. An outline of the specificity of the monoclonal antibodies for the myosin isoforms used in our study is shown below in Table 1.1.

EB 165 2E9 AB8 NA4 Cemb1 + - - + Cemb2 + - - + Cemb3 + - - + Neonatal - + - + Adult + - + + Table 1.1. Specificity of monoclonal antibodies specific to four MyHC isoforms (Moore et al., 1992).

The monoclonal antibody epitopes are distributed throughout the rod portion of the myosin molecule (Figure 1.2.) and none bind to identical locations, thus reinforcing the uniqueness of each MyHC isoform with respect to the amino acid sequence of the myosin rod (Moore et al., Figure 1.3. Schematic of the temporal differences in the expression of the myosin heavy chain isoforms in slow 1992). To better understand the biology underlying growing egg layer (solid line) and fast growing broilers selected for increased the temporal transcription of the MyHC isoforms, breast yield (dotted line) (Wick et al, 2003; Reddish et al, 2005; Lee et al., we must consider their transcription in relation to 2012). overall muscle growth and development. To date, there have been reports on the relationship between MyHC isoform appearance and myogenic development of the PM

in both chickens and turkeys (Bandman and Bennett, 1988; Cerny and Bandman, 1987;

Huffman et al., 2012). The post–hatch expression and transition of myosin from the neonatal to adult isoforms can be influenced by genetic selection (Figure 1.3.). For example, there is a shorter window or compressed expression pattern in fast growing broilers that have been selected for increased breast muscle yield when compared to slow growing SCWL chicks (Wick et al, 2003; Reddish et al, 2005; Lee et al., 2012).

The expression of the MyHC isoforms can also be influenced by environmental factors (e.g., plane of nutrition). Turkeys have the same temporal isoform expression sequences as chickens and their isoforms are immunologically similar (Maruyama et al., Figure 1.4. Delayed expression of the adult 1993). In turkey lines selected for differential MyHC isoform expression in restricted fed poults (Huffman et al., 2012). growth, restricted feeding of poults from the growth selected line delays the expression of the adult MyHC isoform (Figure 1.4.) when compared with ad libitum fed poults (Huffman et al., 2012).

The diverse transcription pattern of the MyHC isoforms is of interest due to the lack of understanding of their relationship with overall muscle growth and development.

Although the functional diversity of the MyHC isoforms remains unresolved, identification of the individual developmental MyHC isoform genes on a transcriptional level has been done through the isolation of cDNA clones.

The ventricular MyHC isoform (Cvent) is the first to be expressed in the embryo and repairing PM muscle (Bourke et al., 1991). Following expression of Cvent, the first 7

embryonic MyHC isoform expressed is Cemb1. Cemb1 was verified using mRNA of 14- day old embryonic skeletal muscle (Umeda et al., 1983). The resulting 2 clones from the embryonic MyHC contained mRNA sequences for 2 different, but homologous, fast skeletal MyHCs. This suggested that there was diversity within the MyHC at the protein level and is related to differences at the mRNA level (Moore et al., 1992). Umeda et al.

(1983) reported that the initial embryonic MyHC isoform, Cemb1, was clone p251 and it was predominantly expressed in embryonic tissue and not present in adult leg or breast muscle. Cemb1 was later identified using transcript specific probes as the corresponding genomic clone N118 (Robbins et al., 1986).

The second embryonic MyHC isoform, Cemb2, was later identified by Umeda et al. (1983) as the second of the 2 mRNA sequences expressed in 14-day old embryonic fast skeletal muscle and corresponded to the sequences of clone 110. The 110 clone had reduced expression when compared with the 251 clone during embryonic development and also differed from the 251 clone in that it was expressed post-hatch in leg muscle but was only present in breast muscle during embryonic development (Umeda et al., 1983).

The final embryonic MyHC isoform, Cemb3, was characterized using transcriptional analysis and was shown to be the predominant MyHC isoform expressed during late embryogenesis. It was homologous to the genomic clone pCM4.1 (Lagrutta et al., 1989).

All of the clones shared regions of structural gene homology, which suggests that these conserved sequences are a result of structural portions of the MyHC gene that participate in the ATPase and actin-binding activities of the molecule (Molina et al.,

1987). The high degree of homology between the genes in the MyHC family emphasizes 8

the difficulty associated with resolving the differential patterns of expression and subsequently the tissue-specific mechanisms controlling the temporal expression of these developmental skeletal fast muscle MyHC isoforms. Along with the temporal transition of the MyHC isoforms, defining subtle changes in myogenic transcriptional genes or the simultaneous transcription of known myogenic transcriptional genes with the MyHC isoforms could be used to identify crucial transcriptional events that are necessary for the successful growth and development of the PM breast muscle in poultry.

1.3. Myogenesis

Myogenesis is a tightly controlled process that is governed by multiple transcription factors and is associated with the temporal transcription of the MyHC isoforms (Maruyama et al., 1993). The integration of positive and negative signals at the gene level regulates the crucial balance between cell proliferation and differentiation necessary for the successful growth and development of skeletal muscle. In the mouse, the onset of muscle formation corresponds to the primary stage of myogenesis, followed by the formation of secondary myofibers beginning at approximately 14 days (Miller,

1992). Secondary muscle fiber formation has been shown to correlate with the onset of innervation, beginning at embryonic days 14-15, playing a role in the subsequent maturation of muscle fibers and fiber type identity (Hughes et al., 1993; Zhang et al.,

1998).

Skeletal muscle growth and development relies on the successful transition and maintenance of each phase of myogenesis. These processes rely on the precise transcription of a complex network of genes to ensure the accurate formation and maturation of muscle fiber and fiber type identity. Several key muscle-specific genes that 9

facilitate this process include myogenic regulatory factors (MRFs), growth factors, E proteins, proteoglycans, and various enzymes.

1.3.1. MRFs

Gene expression in skeletal muscle is controlled by a family of basic helix-loop- helix (bHLH) transcription factors known as myogenic regulatory factors (MRFs).

Included in these MRFs are Myf5, Myf6 (MRF4), MyoD and myogenin (Rudnicki et al.,

1993). During avian and mammalian embryonic development, muscle formation starts with a population of skeletal myogenic progenitor cells formed from somites that differentiate and give rise to the dermomyotome (Hamburger and Hamilton, 1951; Parker et al., 2003). The migration and proliferation of myogenic progenitor cells are critical in maintaining a large population of cells that will eventually differentiate into muscle cells and subsequently muscle fibers. Aiding in this process are two genes belonging to the paired box (PAX) family of transcription factors, Pax3 and Pax7 expressed in this dermomyotome. Pax3 induces the expression of the homeobox protein Six1 that works with Six4 in the delineation and migration of the myogenic progenitor cell population, thus maintaining a proliferative undifferentiated population (Buckingham, 2001). In the absence of functional Pax3, skeletal myogenesis is abolished. Pax7 is expressed by activated myogenic satellite cells during muscle growth and by quiescent satellite cells during muscle regeneration.

Once the population of myogenic progenitor cells is established, muscle development continues with determination, the process of irreversible specification, whereby a cell is committed to differentiate autonomously, from mesodermal progenitor cells to myoblasts (Buckingham, 2001). These signaling pathways subsequently 10

influence the onset of myogenesis through the activation of the myogenic determination genes Myf5 and MyoD, which direct cells to the skeletal myocyte lineage through determining the fate of a myogenic progenitor cell as a skeletal myoblast (Buckingham,

2001). Embryonic studies in mice show that in the absence of both Myf5 and MyoD, there is an absence of markers for skeletal muscle including MyHC, embryonic MyHC, and fetal MyHC and thus no formation of skeletal muscle (Rudnicki et al., 1993), thus emphasizing their importance in muscle specification (Parker et al., 2003). After determination, myoblasts divide and will eventually differentiate into muscle cells. Myf5 and MyoD are expressed in dividing myoblasts and are required for commitment of the proliferation of somatic cells to the myogenic linage, and without their expression myoblast cell growth is compromised (Buckingham, 2001).

Once myogenic cells have been specified, skeletal muscle development requires the differentiation and fusion of myoblasts, to form multinucleated myotubes and myofibers (Parker et al., 2003). Myogenin and Myf6 (MRF4) are expressed later in myogenesis during the activation of muscle specific genes and are required for the committed cells to differentiate further into myotubes and to mature myofibers (Rudnicki et al., 1992). In the absence of Myogenin or Myf6 (MRF4), muscle differentiation does not occur (Buckingham, 2001). More specifically, the absence of myogenin results in a normal number of undifferentiated myoblasts but a loss of differentiated muscle fibers

(Parker et al., 2003). Myf6 (MRF4) is important for terminal differentiation; the fusion of myoblasts to multinucleated myotubes, as well as myofiber maintenance (Parker et al.,

2003).

SMYD1 is a histone methyltransferase essential for embryogenesis through the regulation of myogenesis. SMYD1 has been shown to be a myogenic activator as it is recognized by the early myogenic regulator protein myogenin, thus promoting myogenic differentiation during skeletal myogenesis (Li et al., 2009).

The entire process of skeletal muscle differentiation is controlled by 4 highly related MRF proteins, which have distinct overlaps in expression during muscle development. MRFs work in conjunction with E proteins to regulate gene expression during myogenesis via a shared homologous bHLH domain required for DNA binding and dimerization (Londhe et al., 2011). Each muscle specific gene has a unique temporal association of MRFs and E proteins (Londhe et al., 2011). The E proteins involved in skeletal myogenesis include E12, E47 and HEB. These E proteins exist as a subgroup of the bHLH family of transcription factors that are expressed at all times during development and are all expressed by proliferating cells. More specifically, E12 and E47 are associated with myogenin in undifferentiated cells. As differentiation begins, HEB is associated with myogenin, MyoD and Myf5. This makes HEB a critical protein relative to inducing differentiation (Londhe et al., 2011).

1.3.2. Growth Factors

In addition to MRFs, growth factors including fibroblast growth factor (FGF) and insulin-like growth factor (IGF) can increase the expression of muscle specific genes, thus promoting skeletal muscle myogenesis. FGF expression stimulates progenitor cell proliferation and inhibits differentiation, whereas IGF is produced by the neural tube and somites and promotes differentiation (Buckingham, 1994).

1.3.3. Extracellular Matrix

The extracellular matrix (ECM) is a structural scaffold within tissues in which various cells are embedded. The ECM contains both collagen and proteoglycan components, which sanctions its ability to influence tissue structure, function, development, or gene expression, through the regulation of cell behavior via the interaction with each other, other extracellular matrix molecules (proteoglycans), and interaction with growth factors (Velleman, 1999). Different tissues have unique extracellular matrices that change as an animal ages.

Development and growth of skeletal muscle involves the precise regulation of cell adhesion and cell-cell recognition (Miller, 1992; Buckingham, 1994) through the interaction of muscle cells with the connective tissue ECM proteoglycans (Velleman,

1999), which are involved in tissue differentiation and collagen fibrillogenesis

(Velleman, 1999). A collagen molecule consists of a triple helical arrangement containing 3 collagen alpha chains. Collagen molecules aggregate in parallel fashion to form a fibril, where the individual collagen molecules overlap each other by three quarters of their length, leaving a collagen overlap zone and a gap zone within the fibril molecule. Decorin is a proteoglycan involved in the maturation of collagen fibrils and the formation of collagen crosslinks through binding of the decorin core protein to collagen in the gap zone between adjacent collagen molecules (Velleman, 1999), thus making it a regulator of cell attachment and migration to these gap regions (Velleman,

1999; Li et al., 2008). Decorin expression is of interest because of known expression differences in control SCWL and LSN embryos during late PM embryogenesis

(Velleman, 1996). Muscle development in the LSN PM is characterized by an increase in 13

decorin concentration within ECM expression relative to normal SCWL PM at embryonic day 20 (Velleman et al., 1996).

With regard to early embryogenesis, in vitro studies have been done using chicken myogenic satellite cells isolated from the PM muscle of normal control chickens.

To determine the effect of decorin on proliferation, Li et al. (2008) transiently transfected a chicken full-length decorin cDNA into normal, control PM satellite cells. This increased decorin expression, resulting in a subsequent increase in muscle satellite cell proliferation. Decorin over-expression resulted in a 100-fold increase in mRNA expression and a 50% increase in protein synthesis compared to control cells. In addition, the decorin transfected satellite cells aligned with each other, fused, and formed multinucleated myotubes, while control satellite cells were still proliferating as single cells. During differentiation, both decorin over-expression and control satellite cells fused to form multinucleated myotubes at the same rate.

During satellite cell proliferation and differentiation, cells secrete the growth factor transforming growth factor beta1 (TGF-β1) and decorin into the ECM. Decorin can interact with TGF-β1 to regulate the effect of TGF-β1 on muscle satellite cell proliferation. However, decorin can also function in a growth-factor independent manner mediating cell growth and differentiation in a different manner. Li et al. (2008) showed that over-expression of decorin reduced responsiveness to TGF-β1 during proliferation, suggesting that decorin induces satellite cell proliferation by inhibiting cell responsiveness to TGF-β1.

In addition to decorin, syndecan and glypican are 2 other proteoglycans that are expressed within the ECM during myogenesis. Both can serve as low affinity receptors 14

for FGF, a potent stimulator of muscle cell proliferation and strong inhibitor of differentiation. Increased syndecan expression results in enhanced FGF signaling, thus increasing the proliferation of myoblasts available to form muscle fibers (Velleman,

2004); however, these cells were unable to fuse during differentiation, making it an inhibitor of differentiation. In contrast, increased expression of glypican decreases FGF signaling, thus promoting differentiation and increased myotube formation (Velleman,

2004). All of these processes are critical to the formation and stabilization of skeletal muscle, as they facilitate accurate myoblast alignment and fusion into multinucleated myotubes.

1.3.4. Hyperplasia/Hypertrophy

Gene expression requires an adequate supply of methyl groups for methylation of

DNA. There are 3 key enzymes that are involved in DNA methylation include methylenetetrahydrofolate reductase (MTHFR), methionine synthase (MTR) and methionine synthase reductase (MTRR; Terruzzi et al., 2011). DNA hypomethylation causes cell enlargement and an increase in the number of cells, cell hypertrophy and hyperplasia respectively, leading to a higher rate of DNA synthesis. Myogenic proteins such as Myf5, MyoD, Myf6 (MRF4) and MyHC are largely induced in hypomethylated cells, cells that have a reduced expression of these key enzymes involved in DNA methylation (Terruzzi et al., 2011). The diameter and length of hypomethylated myotubes are greater when compared to non-hypomethylated control myotubes.

Therefore, decreased expression of these enzymes (hypomethylation) leads to an increase in muscle cell size through the increased activation of the myogenic proteins that promote

proliferation and differentiation of myoblasts, thus promoting overall muscle growth and increase of muscle mass (Terruzzi et al., 2011).

In addition to discussing the positive regulators of muscle development, it is also important to consider the expression of negative regulatory factors. Transforming growth factor beta1 (TGF-β1) is an inhibitor of skeletal muscle myoblast proliferation and differentiation via its effect on cell adhesion and the expression of myogenin. TGF-β1 binds to the core protein of decorin. When TGF-β1 is bound to the decorin core protein, its activity is suppressed, making decorin a negative regulator of TGF-β1 (Velleman,

1999). During LSN PM development, TGF-β1 is up regulated from embryonic day 20 through 1 week post-hatch and this increase in TGF-β1 expression corresponds to increased decorin synthesis (Velleman, 1999).

Myostatin is a member of the TGF-β1 family that prevents excessive muscle growth by limiting the proliferation of PAX-positive progenitor cells during development

(Buckingham, 2006). Short-term deletion of myostatin enhances muscle regeneration and promotes satellite cell self-renewal (Punch et al., 2009).

Table 1.2. outlines the regulatory genes selected for our studies and their corresponding references and known associations with myogenesis.

Expression during skeletal muscle Gene Name Type differentiation (D) or Source proliferation (P), if known. Cvent MyHC isoform Tidyman 1997 Cemb 1 MyHC isoform Tidyman 1997 Cemb 2 MyHC isoform Tidyman 1997 Cemb 3 MyHC isoform Tidyman 1997 Cadult MyHC isoform Tidyman 1997 Cneo MyHC isoform Tidyman 1997 (Myf6) MRF4 Transcription factor + (D) Rudnicki 1993 Myf5 Transcription factor + (D) Rudnicki 1993 MyoD Transcription factor + (D) Rudnicki 1993 Myogenin Transcription factor + (D) Rudnicki 1993 Six1 Transcription factor Indirectly + (D) Buckingham 2001 Six4 Transcription factor Indirectly + (D) Buckingham 2001 Pax3 Transcription factor + (D) Buckingham 2001 Pax7 Transcription factor Adult regeneration Buckingham 2001 TGF-β1 Growth factor - (P & D) Buckingham 2001 IGF Growth factor + (P & D) Buckingham 1994 FGF Growth factor + (P) & - (D) Buckingham 2001 Myostatin Growth factor - (P & D) Buckingham 2006 Smyd1 -P & +D Li 2009 Decorin Proteoglycan + (P & D) Velleman 1996 MTHFR1 DNA methylation Terruzi 2011 MTR2 DNA methylation Terruzi 2011 MTRR3 DNA methylation Terruzi 2011 E47 bHLH protein4 + (D) Londhe 2011 E12 bHLH protein + (D) Londhe 2011 HEB bHLH protein + (D) Londhe 2011 Β-actin Housekeeping + (D & P) NanoString® GAPDH Housekeeping + (D & P) NanoString® HPRT15 Housekeeping + (D & P) NanoString® HM BS6 Housekeeping + (D & P) NanoString® RPL47 Housekeeping + (D & P) NanoString® Syndecan-4 Proteoglycan + (P) Velleman 2011 Glypican Proteoglycan +(D) Velleman 2011 Table 1.2. Outline of all 33 target genes used in this experiment. 1 methylenetetrahydrofolate reductase (MTHFR) 2 methionine synthase (MTR) 3 methionine synthase reductase (MTRR) 4 Basic helix-loop helix (bHLH) 5 Hypoxanthine phosphoribosyltranserase 1 (HPRT1) 6 Hydroxymethlbilane synthase (HMBS) 7 Ribosomal protein L4 (RPL4) 17

1.4. Discussion

The objective of the study described herein was to define the molecular transcription patterns in 2 unique genetic lines during embryonic PM muscle development. Embryonic muscle development in these lines has not been previously characterized. The goal was to target specific stages of embryonic development via the temporal transcription of the developmental MyHC embryonic isoforms and use this temporal pattern as a developmental clock to further define transcription factors underlying myogenesis. To accomplish this, embryonic gene transcription data from the genes in Table 1.2. were obtained from 2 experimental lines: normal SCWL and the LSN line. The goal is that the system we are proposing can subsequently be used for future molecular analyses of embryonic muscle development in modern broiler genotypes that have been selected for extremes in breast muscle development.

Chapter 2: Molecular Staging based on Temporal Myosin Heavy Chain Transcription

Abstract

PM, are expressed as a cadre of highly specific temporal and spatial developmental isoforms. The hypothesis is that the temporal transcription and translation of MyHC isoforms is correlated with the transcription of muscle-specific genes that are critical to

PM muscle growth and development and can be used as molecular markers during muscle development. The goal of this study was to use a novel molecular method, based on a quantitative analysis of the transcriptome, to characterize the developmental stages in embryonic PM in the SCWL and compare the data with what has been reported in the literature (Tidyman et al., 1997). Similar transcription patterns would thus validate the use of the novel technology for characterizing embryonic PM muscle development in 19

poultry. Tissue samples from the embryonic PM were collected daily from days 5 through 19. Total RNA was isolated and gene transcription quantified using

NanoString®, which digitally detects and quantifies selected target mRNA. Data were analyzed using the LOESS smoothing function at a 95% confidence level. Results confirmed that the pattern of temporal transcription of the MyHC isoforms obtained in this study was consistent with Tidyman et al. (1997) but with much greater resolution.

These data confirm the fact that the NanoString Technology® results in a similar transcription pattern as reported by Tidyman et al. (1997) and thus gives us the tool we need to explore our hypothesis that the transcription patterns of the Cemb isoforms can be used as molecular clocks for embryonic fast skeletal muscle growth and development.

2.1. Introduction

The primary goal of this experiment was to validate the use of the novel

NanoString® molecular technology, which to date, has not been used to characterize transcriptional events in commercial avian species. To accomplish this, we compared the temporal transcription of the MyHC isoforms during myogenesis in embryos from a control SCWL genetic line with the data from Tidyman et al. (1997). These authors defined the temporal transcription of the MyHC isoforms using RNA from the PM of

SCWL starting at embryonic day 10 through 240 days post-hatch. The relative percentages of MyHC isoform transcripts in the PM in that study were derived from RNA dot blot hybridizations using transcript specific oligonucleotide probes. The experiment described herein was designed to emulate the study by Tidyman et al. (1997) as much as possible in order to make accurate comparisons with the NanoString® molecular technology for gene transcription analysis. 20

2.2. Objective

The objective of this study was to use a novel molecular method, based on a quantitative analysis of the transcriptome, to characterize the temporal transcription of

MyHC isoforms during embryogenesis in embryonic PM in the control SCWL line.

2.3. Hypothesis

The hypothesis is that the temporal transcription of MyHC isoforms will correspond with what has been reported in the literature (Tidyman et al., 1997), thus validating the use of the NanoString® technology.

2.4. Methods

Along with trying to emulate the sampling methodology (embryonic age) and genotype used by Tidyman et al. (1997), NanoString® technology has been reported to provide precise, accurate and reproducible gene transcription analysis (Reis et al., 2011).

This is primarily due to its technical ability to digitally quantify each target gene, including 5 housekeeping genes, using unique gene specific probes. Among gene specific probes, every code set contains probes designed against 14 External RNA

Controls Consortium (ERCC), a common set of external RNA controls developed to account for normal variability in RNA transcription, including the quality of starting material, level of cellular and RNA yield, platform employed, and human error. Among these 14 sequences, 6 are used as positive hybridization controls and 8 are designed as negative controls (NanoString Technologies®). Quantitative data are collected from these control probes and used to normalize the raw quantitative data collected from each target gene to account for variation.

2.4.1. Pectoralis major (PM) Tissue Extraction

SCWL embryos were incubated at the industry standard of 37.50C. One egg was removed from the incubator on chronological days 5 through 19. The shell was cracked at the air cell to expose the shell membrane. Embryos were removed, detached from the yolk sac, immediately decapitated and placed on a sterile petri dish. The PM breast muscle was exposed and removed from each side of the sternum using a sterile scalpel blade. The PM tissue was flash frozen in liquid nitrogen and directly placed in a -800C freezer to prevent RNA degradation prior to RNA extraction.

2.4.2. RNA Isolation

RNA isolation was performed using a Norgen® Animal Tissue Isolation Kit, according to manufacturer’s recommendation (Norgen® Biotek Corp, Thorold Ontario,

Canada). Briefly, samples were taken from -80oC storage, immediately submerged in liquid N and pulverized in a frozen mortar with a frozen pestle, all the while staying submerged in liquid N.

• Cell Lysate Preparation

The PM muscle sample was lysed to easily separate unwanted cellular debris, including proteins and organic material, from RNA. To accomplish this, 10 mg of PM tissue was placed in a sterile cryovial tube with 300ul of the Norgen® Lysis Buffer.

Homogenization was repeated to ensure maximum tissue surface area exposure to the lysis buffer. Following homogenization, 600ul of RNase Free Water and 20ul of reconstituted Proteinase K was added, followed by water bath incubation at 55oC for 15 minutes, vortexing every 5 minutes. This was to ensure further protein catabolism. After incubation, the sample was spun in a microcentrifuge for 1 minute at 14,000 x g to 22

separate the unwanted cellar debris (pellet) from the RNA. The supernatant, containing the RNA was removed and transferred to a new microcentrifuge tube, at which point

450ul of 95% ethanol was added to precipitate the nucleic acids for column binding.

• Binding RNA to Column

The Norgen® column was assembled, at which point 650ul of the ethanol- supernatant mixture (containing the RNA) was added to the top of the column and spun for 1 minute at 14,000 x g. The flow-through (ethanol and unwanted cellular debris) was discarded, leaving the purified RNA bound to the column. This step was repeated up to 3 times until all of the ethanol-supernatant mixture had passed through the column.

• On-Column DNAse Treatment

Possible DNA contamination can occur during nucleic acid precipitation.

Therefore, DNAse treatment ensures zero DNA contamination. To accomplish this,

100ul of Norgen® Enzyme Incubation Buffer and 15ul of DNAse1 was added to the column and spun for 1 minute at 14,000 x g. The entire flow-through of DNAse and buffer (115ul) was pipetted back onto the column and incubated at room temperature for

15 minutes to ensure all DNA was eluted in the flow-through.

• Column Wash

Four hundred microliters of the Wash Solution was subsequently added to the column and the column was spun down at 14,000 x g for 1 minute. The flow-through was discarded, an additional 400ul of Norgen® Wash Solution was added to the column and spun down again at 14,000 x g for 1 minute. The flow-through was discarded and the column was spun for 2 minutes at 14,000 x g to thoroughly dry the column.

• RNA Elution 23

The column was placed into a clean elution tube (provided with the kit), 50ul of

Norgen® Elution Buffer was added to the column and spun down at 200 x g for 2 minutes followed by 1 minute at 14,000 x g. This process was repeated again for a second elution in a separate tube.

• RNA Evaluation

Every RNA sample was evaluated for quality and concentration using gel electrophoresis and nanoDrop® respectively in preparation for downstream analysis using an nCounter® gene expression CodeSet design by NanoString® Technologies.

2.4.3. NanoString Technologies

Gene transcription analysis was quantified using an nCounter® gene expression

CodeSet designed by NanoString® Technologies. This allows for direct digital quantification of each target gene, resulting in a direct comparison of our 2 unique genetic lines. nCounter® probes are designed to provide a single-tube, ultra sensitive, reproducible, and highly multiplexed method for detecting nucleic-acid targets across all levels of biological expression. This technology does not require the conversion of mRNA to cDNA by reverse transcription or the amplification of the resulting cDNA by

Polymerase Chain Reaction (PCR) and has been shown to achieve superior gene expression quantification results when compared to RT-PCR (Reis et al., 2011).

2.4.3.1. Probe Architecture

The nCounter® analysis system uses unique color-coded molecular barcodes

(Figure 2.1.) that hybridize directly to nucleic acids through the use of gene specific color-coded probe pairs. In solution, the probes capture and count specific nucleic acid molecules in a complex mixture, through hybridizing directly to the target mRNA for 24

digital detection. The digital color-coded barcodes consist of unique combinations of 4 spectrally non-overlapping dyes arranged at 7 contiguous regions, making it possible to generate hundreds of unique transcripts, corresponding to 1 target gene, in a single reaction. Each target gene of interest is detected using a pair of reporter and capture probes carrying 35- to 50-base target-specific sequences. Each reporter probe is the

“barcode”, carrying a unique color code assigned to each target mRNA sequence at the 5’ end that enables the molecular barcoding for detection of the target mRNA in solution.

The capture probe carries a biotin at the 3’ end providing a molecular handle for attachment of target mRNA for immobilization to facilitate downstream digital detection.

Figure 2.1. Probe architecture showing capture probe, reporter probe and target mRNA complex.

2.4.3.2. Hybridization

While in the solution-phase, the capture probe and reporter probe hybridize to a complementary target mRNA in solution via the gene-specific sequences providing a digital count of target mRNA molecules. After the solution-phase, the tripartite molecule is affinity-purified first by the 3’-repeat sequence and then by the 5’-repeat sequence to remove excess reporter and capture probes, respectively.

2.4.3.3. nCounter Assay

Attachment occurs when the remaining hybridized probe/mRNA structures bind to a streptavidin-coated nCounter® cartridge via the biotinylated 3’ end of the capture probes. Following attachment, voltage is applied to elongate and align the reporter complexes. Biotinylated anti-5’ oligonucleotides that hybridize to the 5’-repeat sequences are added. The reporters are then immobilized by the binding of the anti-

5’oligonucleotides to the slide surface via the biotin. Finally, the immobilized reporters are prepared for imaging and counting using a proprietary reagent and the sample cartridge is placed into the digital analyzer for data collection.

2.4.3.4. nSolver Data Analyses

Using nSolver software provided by NanoString®, housekeeping gene normalization was performed in order to adjust the digital counts of all probes that were not expected to vary between samples. A total of 5 housekeeping genes were selected based on NanoString® recommendations. A normalization factor was calculated based on the average of the geometric mean of all 5 housekeeping genes for each data point.

The average of the geometric means across all data points was used as a reference against which each lane was normalized. A normalization factor was then calculated for each of the lanes based on the geometric mean of counts for the housekeeping genes in each lane relative to the average geometric mean of counts for the housekeeping genes.

2.5. Statistical Analyses

Gene transcription measurements were regressed on days of incubation using the

LOESS (i.e., Local regression) procedure of SAS V9.2, a non-parametric regression method that combines multiple regression models in a k-nearest neighborhood model

(Cleveland and Devlin, 1988). LOESS regressions allow great flexibility in estimating response curves because they make no assumptions about the parametric form of the regression. The optimal smoothing parameter was determined using cross-validation.

Curves were fitted for each line of birds. As with conventional regression methods (e.g., least-squares), LOESS produces prediction estimates as well as standard errors of these estimates. Comparisons of genetic lines on given days were done using t-tests based on the predictions and standard errors of each line. Rates of change in gene transcription were calculated using finite forward and backward differences (half-days). Significance was declared at P < 0.05.

2.6. Results

The relative percentage of total MyHC isoform transcription was calculated and graphed (Figure 2.2.) with 5% error bars in red, for accurate graphical comparison to

Tidyman et al. (1997). Given that we utilized a new technology, transcriptional gene analysis was conducted to confirm the temporal transcription of MyHC isoforms during embryogenesis, using comparisons with what is reported in the literature. The temporal transcription of the MyHC isoform data published by Tidyman et al. (1997) is shown in

Figure 2.2. Plot A. This report has served as the positive control for our preliminary research as this study was done using RNA from the PM of SCWL chicks (Tidyman et al., 1997), similar to the control experimental line used in the current experiment. It is 27

important to emphasize the use of the SCWL as the control genetic line because a critical component of the work was to confirm, using the NanoString® technology, what has already been reported. In Figure 2.2., Plot B was generated from our preliminary data of the MyHC relative transcription at embryonic days (ED) 10, 12, 14, 16, 18 and 19. The full analysis started at ED5 and included the intermittent days beyond what is shown in

Figure 2.2. For comparison purposes, days included are those published by Tidyman et al. (1997).

In addition to the data used to compare with Tidyman et al. (1997), full temporal

MyHC isoform transcription data included ED5 through 19 (Figure 2.3.).

Although all MyHC isoforms were detected to some degree during embryogenesis, their temporal transcription reveals key transition Figure 2.2. Relative transcription of the points that may prove useful in developmental MyHC isoforms in the SCWL previously published by Tidyman et al. (1997) (Plot A) and in the SCWL (Plot B) taken from characterizing developmental stages in preliminary data. Plot B significance determined at P < 0.05 using LOESS Smoothing function in SAS. the embryonic PM in the SCWL, which is the basis of the hypothesis we propose to test.

The results from the relative MyHC isoform transcription data during embryogenesis show Cemb1 was present in the greatest relative amount at 7ED with 74% of total MyHC mRNA, Cemb2 was greatest at 13ED with 31% of total MyHC mRNA,

Cemb3 was greatest at 17ED with 54% total MyHC mRNA, and Cneo was greatest at

19ED with 14% total MyHC mRNA (Figure 2.3.). 28

Figure 2.3. Relative transcription of the developmental MyHC isoforms in the SCWL at ED6 through 19. Significance determined at P < 0.05 using LOESS Smoothing function in SAS.

2.7. Discussion

The results from this study confirm that the temporal transitions of MyHC isoform transcription in the control SCWL line are consistent with Tidyman et al. (1997), therefore validating the use of the Nanostring Technology®. We hypothesize that the temporal transcription of developmental MyHC isoforms in the SCWL may be used to target specific stages of embryonic development and subsequently be used as developmental clocks to further define transcription factors underlying myogenesis.

With the confidence gained from the comparison data, we further characterized the MyHC isoform transcription in the LSN line, to test the hypothesis that differences existed in the transcription patterns between the 2 unique genetic lines. In addition, gene transcription analysis of muscle-specific genes known to be critical to the growth and development of skeletal muscle were quantified to reveal any transcriptional correlations with embryonic MyHC isoform transcription and/or differences in gene transcription between the SCWL and LSN.

Chapter 3: Transcriptional Events Underlying Avian Myogenesis

Abstract

In Chapter 2 we reported on a novel quantitative transcriptomic method to establish the temporal transcription of the developmental MyHC isoforms. We hypothesized that this method can be used to develop molecular stages for use in correlating the temporal transcription of MyHC isoforms with the transcription of regulatory genes critical to PM muscle growth and development. To test this hypothesis, we used embryos from a control SCWL line and an intermediate dystrophic LSN line that has been reported to exhibit breast muscle anomalies post-hatch. The previously established MyHC isoform transcription patterns were used to investigate the relationship between temporal MyHC isoform transcription and the correlated transcription of known myogenic transcription and regulatory factors in the embryonic PM in the 2 unique poultry genotypes. Tissue samples from the embryonic PM were collected at embryonic days 5 through 19 from SCWL and LSN embryos. RNA was isolated and transcription was quantified using NanoString Technologies®, which digitally quantifies selected target genes. The data were analyzed using the LOESS smoothing function at a 95% confidence level. There were differences in the quantitative temporal transcription of the

3 embryonic MyHC isoforms, Cemb1, 2 and 3, and the neonatal MyHC isoform, Cneo in both SCWL and LSN embryos. Line differences were primarily due to shifts in peak transcription, with significant changes occurring between the SCWL and LSN at

embryonic days (ED) 14-15 and this was consistent among all MyHC isoforms. In addition, there were genetic line differences in the quantitative temporal transcription of 4 known muscle-specific genes, decorin, MTR2, Myf6 (MRF4) and Six4. Furthermore, there was a positive correlation between Cemb1, the first embryonic MyHC isoform, and

Myf6 (MRF4) in both the SCWL and LSN from ED5 through ED19. The transcription patterns of both Myf6 (MRF4) and Cemb1 were significantly different in the SCWL and

LSN embryos on ED11 through ED18. These data allowed us to describe the temporal transcription of regulatory factors at identical cellular development stages in the 2 unique genetic lines. If this approach is applied to broiler genotypes, it could lead to new, enhanced selection tools or aid in identifying the developmental points at which factors such as incubation environment influence the earliest stages of myogenesis.

3.1. Introduction

The quantitative transcriptomic results from the control SCWL line (Chapter 1) provided critical preliminary data relative to confirming the known temporal transcription of MyHC isoforms reported in the literature. With the confidence gained from the control SCWL MyHC isoform transcription data and the technology used to obtain these results, we further characterized the temporal transcription of the MyHC isoforms in the

LSN line to test the hypothesis there are genotype differences in the temporal transcription of the MyHC isoforms. The gene transcription of several other known muscle-specific regulatory genes was also determined to study the potential correlation between the MyHC isoform transcription patterns and key regulatory genes as well as any genotype differences that might exist.

3.2. Objective

The objective of this study was to use a novel molecular method, based on a quantitative analysis of the transcriptome, to characterize the developmental stages in embryonic PM in the SCWL and the LSN line during myogenesis.

3.3. Hypotheses

Hypothesis 1. Genotype differences exist in the temporal transcription of the embryonic

MyHC isoforms in SCWL and LSN during embryonic myogenesis.

Hypothesis 2. The temporal transcription of MyHC isoforms is correlated with the transcription of muscle-specific genes unique to each genetic line.

3.4. Methods

NanoString Technologies® was used for gene transcription analysis (refer to

Chapter 2 methods; section 2.4. through 2.4.3.4). The methodology used for selecting genes for transcriptional analysis was based on their relationship and influence on the process of skeletal muscle growth and development during embryogenesis. An outline of the target sequences used for quantitative gene transcription analysis is shown in Table

3.1. The sequences included are from those genes for which significant differences between the SCWL and LSN were observed. Target gene sequences were selected from the complete gene sequence provided in the National Center for Biotechnology

Information (NCBI) database. Each target gene sequence was cross-hybridized against each other to ensure zero cross reactivity of probes during the solution phase of the nCounter® analysis system.

Gene NCBI Accession # Target Sequence Cemb1 MYH1emb.1 TTCCCTGGAAGATCAACTCAGTGAAATTAAGACAAAGGAAGAGGAGCAACAGC GCACCATTAATGACATCAGTGCACAGAAAGCTCGTCTACAAACAGAG Cemb2 NM_204228.1 GCTGAAGGCAGAAAGAAGCTACCACATATTTTATCAAATTATGTCCAACAAGA AACCAGAGCTAATTGGCATGCTTCTAATTACCACCAATCCATACGAC Cemb3 NM_001113709.1 CAGAGGCTGGTGGTGGAGGCAAAAAGGGTGGCAAGAAGAAGGGTTCTTCTTTC CAAACAGTTTCTGCTCTTTTCCGGGAGAACTTAAACAAGCTGATGAC Cneo M74087.1 AGTGTGACCTGAGGCATGCATAAAATGTGAACTCTGTGTTGCTTTTTTATGTC ATTGTCATCTATGTCTAGGTAATAAAGAGAGTAGAGACCTTTGCATA Myf6 NM_001030746.1 TGTACCGGCGGACGGTCGCCCTGCGCCCAGCCCGAGCAGTCCATTCCTGTCAG GGTGAAGCCGTTCGGGCTGCGTTTCGCGTCGTCGTGCTGAGACGTCC (MRF4) decorin NM_001030747.1 CTCAGCTTTGAGAGCTCCTGTTGCAAATCCCTGGATTAAAAGGTTCTGCCTGG AGTTGATGAACTGAGCCATGAGGCTAGTTCTCCTCTTCGTCCTACTG MTR2 NM_001031104.1 CAGTTCTGCGGGAGCGCATAATGATTTTGGATGGAGGCATGGGTACCATGATC CAGCAGCACGCTCTGTCAGAAGAAGATTTCCGAGGGCATGAATTTAA Six4 AJ133778.1 GCGTGGGGGCTTCGCAGCCGGTAACTTTAAATTCACCCAAAACCACTTCAAGT GCTGTGAGCAACGGGGTGTCCATCACTGACATCATGTCGTCTTCTTC Table 3.1. Outline of reported genes, their corresponding NCBI reference and target mRNA sequence.

3.5. Statistical Analyses

Gene transcription measurements were regressed on days of incubation using the

LOESS (i.e., Local regression) procedure of SAS V9.2, a non-parametric regression method that combines multiple regression models in a k-nearest neighborhood model

Curves were fitted for each line of birds. As with conventional regression methods (e.g., least-squares), LOESS produces prediction estimates as well as standard errors of these estimates. Comparisons of genetic lines on given days were done using t-tests based on the predictions and standard errors of each gene/time/line. The rates of change in gene transcription were calculated using finite forward and backward differences (half-days).

Significance was declared at P < 0.05.

3.6. MyHC Isoform Results

Significant differences existed between the SCWL and LSN embryos in the

MyHC isoform transcription patterns of the 3 embryonic MyHC isoforms Cemb1, 2 and

3 and the neonatal MyHC isoform, Cneo. Significant genotype differences were found for the temporal transcription of Cemb1 on ED9 through ED19 (Figure 3.1.), Cemb2 on

ED14 through ED19 (Figure 3.2.), Cemb3 on ED6 through ED17 (Figure 3.3.), and

Cneo on ED14 through ED19 (Figure 3.4.). In addition, the data suggested similar transcription patterns between the first embryonic MyHC isoform, Cemb1, (Plot A) and the known myogenic transcriptional gene Myf6 (MRF4) (Plot B), as shown in Figure

3.5.

Figure 3.1. Quantitative transcription of Figure 3.2. Quantitative transcription of Cemb2 Cemb1 in the SCWL and LSN. Plot generated in the SCWL and LSN. Plot generated from from predicted transcription values determined predicted transcription values determined by by LOESS smoothing function in SAS. An LOESS smoothing function in SAS. An asterisk indicates a difference (P < 0.05) asterisk indicates a difference (P < 0.05) between the SCWL and LSN. between the SCWL and LSN.

Figure 3.3. Quantitative transcription of Figure 3.4. Quantitative transcription of Cneo Cemb3 in the SCWL and LSN. Plot generated in the SCWL and LSN. Plot generated from from predicted transcription values determined predicted transcription values determined by by LOESS smoothing function in SAS. An LOESS smoothing function in SAS. An asterisk indicates a difference (P < 0.05) asterisk indicates a difference (P < 0.05) between the SCWL and LSN. between the SCWL and LSN.

A B

Figure 3.5. Plot A shows the quantitative transcription of Cemb1 in the SCWL and LSN. Plot B shows quantitative transcription of MRF4 in the SCWL and LSN. Plots generated from predicted transcription values determined by LOESS smoothing function in SAS. An asterisk indicates a difference (P < 0.05) between the SCWL and LSN.

3.7. MyHC Discussion

All of the MyHC isoforms exhibited different transcription patterns in the SCWL and LSN embryos starting at ED14-15. This suggests that this is a critical time point during myogenesis, when differential transcription between 2 unique genetic lines is initiated. Quantitative transcription patterns of Cemb1 and Cemb2 were similar in the

SCWL and LSN during the initial stages of myogenesis (ED5 through ED13). Starting at

ED15, Cemb1 transcription in the SCWL plateaued, whereas transcription in the LSN continued to increase (Figure 3.1.). At ED15, however, Cemb2 transcription in the

SCWL plateaued, whereas transcription in the LSN decreased (Figure 3.2.). Quantitative transcription patterns of Cemb3 and Cneo were similar in the SCWL and LSN during the initial stages of myogenesis (ED5 through ED13), but starting at ED14, Cemb3 transcription was delayed in the SCWL when compared to the LSN (Figure 3.3.).

Alternatively, at ED14, Cneo transcription was delayed in the LSN when compared to the

SWCL (Figure 3.4.). As previously mentioned, our data revealed significant differences between the SCWL and LSN in the quantitative transcription of the known MRF, Myf6

(MRF4), (Figure 3.5. Plot B) from ED11 through ED18. Myf6 (MRF4) transcription patterns are similar to those of Cemb1 in both lines (Plot A) suggesting a correlation exists between these 2 genes during the second phase of embryonic myogenesis.

These data support our hypothesis that there are differences in the temporal transcription of MyHC isoforms between the SCWL and LSN during embryonic myogenesis. The temporal transcription of MyHC isoforms is correlated with the transcription pattern of muscle-specific genes within each genetic line, although the correlated transcription patterns are unique within each genotype. Further studies are 37

needed to determine the significance of these correlations as it pertains to phenotypic characteristics of the PM post-hatch. To better understand the biology underlying the temporal transcription of the MyHC isoforms, we must consider this transcription in relation to overall muscle growth and development. This would be particularly important relative to the known MyHC isoform shifts at ED14-15, and might give some insight into what else is occurring during this critical time point when differential transcription between the SCWL and LSN is initiated.

As stated in Chapter 1, section 1.3, primary muscle fibers form during the first stage of skeletal myogenesis at approximately ED13-14, followed by the formation of secondary myofibers that form parallel to the primary fibers starting at approximately

ED14 (Miller, 1992). These data from the current study suggest that the significant shifts in MyHC isoform transcription between the SCWL and LSN initially occur starting at

ED14-15, during the second stage of myogenesis, when secondary myotubes are expected to form from primary myotubes (Miller, 1992). Differential transcription of the

MyHC isoforms between the 2 unique lines continues throughout the remainder of embryogenesis.

To better understand the genetic shifts in transcription patterns, phenotypic differences post-hatch are equally important to consider. As stated in Chapter 1, section

1.1., the LSN is a partially dystrophic line that exhibits muscle abnormalities post-hatch.

These differences include a weakened PM muscle (Velleman et al., 2001) and a significant reduction in PM muscle mass relative to the SCWL control line (Velleman et al., 1993; 1996). In vitro, the LSN exhibits morphological differences with a decrease in myotube length and number of nuclei per myotube (Velleman et al., 1999). These 38

phenotypic differences may be influenced by the differential transcriptional events taking place in the SCWL and LSN embryos. The differential MyHC isoform transcription initiated during the transition from the first to second phase of myogenesis suggests that this may be a critical developmental stage.

3.8. Transcriptional Analysis Results

Data from the full gene transcriptional analysis described in Table 1.2. revealed 3 genes that had significantly different embryonic transcription patterns in the SCWL and

LSN. These genes included Six4, decorin, and MTR2. Significant differences were observed for Six4 on ED5 through ED18 (Figure 3.6.), MTR2 on ED5 through ED18

(Figure 3.7.), and decorin on ED17 and ED18 (Figure 3.8.).

Figure 3.6. Quantitative transcription of Six4 Figure 3.7. Quantitative transcription of in the SCWL and LSN. Plot generated from MTR2 in the SCWL and LSN. Plot predicted transcription values determined by generated from predicted transcription values LOESS smoothing function in SAS. An determined by LOESS smoothing function in asterisk indicates a difference (P < 0.05) SAS. An asterisk indicates a difference (P < between the SCWL and LSN. 0.05) between the SCWL and LSN.

Figure 3.8. Quantitative transcription of decorin in the SCWL and LSN. Plot generated from predicted transcription values determined by LOESS smoothing function in SAS. An asterisk indicates a difference (P < 0.05) between the SCWL and LSN.

3.9. Transcriptional Analyses Discussion

The quantitative transcription of Six4 in the SCWL exhibited shifts in peak transcription compared to the LSN (Figure 3.6.). The quantitative transcription of MTR2 in the SCWL was uniform throughout embryogenesis, whereas transcription in the LSN was not consistent and was significantly decreased from ED5 through ED8 when compared to the SCWL (Figure 3.7.). The quantitative transcription of decorin in the

SCWL progressively increased throughout embryogenesis. Transcription in the LSN was similar until ED16 when it subsequently declined (Figure 3.8.), another example of differential gene transcription during the course of embryonic myogenesis in the SCWL and LSN embryos. This supports our hypothesis that the transcription of muscle-specific genes is unique to each line. Future studies are needed to determine the significance of the differential gene transcription in Six4, MTR2 and decorin, between the SCWL and

LSN lines.

Conclusion

The goal of this study was to target specific stages of embryonic development via temporal transcription of the developmental MyHC isoforms and to use this temporal pattern as a developmental clock to further define transcription factors underlying myogenesis. Results from the temporal transcription of the developmental fast skeletal

MyHC isoforms in the SCWL and LSN emphasize the ability and potential to serve as a molecular staging mechanism for skeletal myogenesis. Defining key MyHC isoform transitions and the potential transcriptional events driving their transcription will aid in further understanding the physiological factors underlying muscle growth and development.

This study suggests that a critical time point during embryonic myogenesis is approximately ED14-15, at which point genotype differences were observed between the

SCWL and LSN genetic lines in the temporal transcription of the MyHC isoforms. The differential transcription of select muscle specific genes indicates there are genetic differences on the transcriptional level between the 2 unique lines that could potentially be driving the temporal transcription of MyHC isoforms. Further transcriptional and morphological studies during this critical time frame and early post-hatch are of interest to further define transcriptional events underlying embryonic myogenesis in the PM that give rise to phenotypic differences post-hatch in lines selected for differential growth in breast muscle.

References

Bandman, E. 1985. Myosin isoenzyme transitions in muscle development, maturation, and disease. International Review of Cytology 97:97-131.

Bandman, E., and T. Bennett. 1988. Diversity of fast myosin heavy chain expression during development of gastrocnemius, bicep brachii, and posterior latissimus dorsi muscles in normal and dystrophic chickens. Developmental Biology 130:220-31.

Buckingham, M. 1994. Muscle: the regulation of myogenesis. Current Opinion in Genetics and Development 4:745-751.

Buckingham, M. 2001. Skeletal muscle formation in vertebrates. Current Opinion in Genetics and Development 11:440-448.

Cerny, L., and E. Bandman. 1987. Expression of myosin heavy chain isoforms in regenerating myotubes of innervated and denervated chicken PM muscle. Developmental Biology 119:350-62.

Cleveland, W. S., and S. J. Devlin. 1988. Locally-weighed regression: an approach to regression analysis by local fitting. Journal of the American Statistician Association 83:596-610.

Hamburger, V., and H. L. Hamilton. 1951. A series of normal stages in the development of the chick embryo. Journal of Morphology 88:49–92.

Huffman, K., I. Zapata, J. M. Reddish, M. S. Lilburn, and M. Wick. 2012. Feed restriction delays developmental fast skeletal muscle myosin heavy chain isoforms in turkey poults selected for differential growth. Poultry Science 91:3178-3183.

Hughes, S. M., J. M. Taylor, S. J. Tapscott, C. M. Gurley, W. J. Carter, and C.A. Peterson. 1993. Selective accumulation of MyoD and Myogenin mRNAs in fast and slow adult skeletal muscle is controlled by innervation and hormones. Development 118:1137- 1147.

Julian, L. M. 1973. Animal model: Hereditary muscular dystrophy of chickens. American Journal of Pathology 70:273-276.

Lagrutta, A. A., J. G. McCarthy, C. A. Scherczinger, and S. M. Heywood. 1989. 42

Identification and developmental expression of a novel embryonic myosin heavy-chain gene in chicken. DNA 8:39-50.

Lee, A., Y. Suh, M. P. Wick, and K. Lee. 2012. Temporal myosin heavy chain isoform expression transitions in broiler chickens compared with single comb white leghorns. Poultry Science 91:2872-2876.

Li, Z., S. G. Velleman, D. C. McFarland, J. E. Pesall, K. K. Gilkerson, N. H. Ferrin, and Y. Yun. 1997. Characterization of satellite cells derived from chickens with the Low Score Normal (LSN) muscle weakness. Cytobios. 91:75–85.

Li, X., D. C. McFarland, and S. G. Velleman. 2008. Extracellular matrix proteoglycan decorin-mediated myogenic satellite cell responsiveness to transforming growth factor-β1 during cell proliferation and differentiation. Domestic Animal Endocrinology 35:263- 273.

Li, Xi., and S. G. Velleman. 2009. Effect of transforming growth factor-β1 on embryonic and posthatch muscle growth and development in normal and low score normal chicken. Poultry Science 88:265-275.

Londhe, P., and J. K. Davie. 2011. Sequential association of myogenic regulatory factors and E proteins at muscle-specific genes. Skeletal Muscle Journal 1:1-18.

Maruyama, K., N. Kanemaki, W. Potts, and J. D. May. 1993. Body and muscle growth of domestic turkeys (meleagris gallopavo) and expression of myosin heavy chain isoforms in breast muscle. Growth Development and Aging 57:31-43.

Miller, J. B. 1992. Myoblast diversity in skeletal myogenesis: how much and to what end? Cell 69:l-3.

Molina, M. I., K. E. Kropp, J. Gulick, and J. Robbins. 1987. The sequences of an embryonic myosin heavy chain isolation of its corresponding cDNA. The Journal of Biological Chemistry 14:6478-6488.

Moore, L. A., M. J. Arrizubieta, W. E. Tidyman, L. A. Herman, and E. Bandman. 1992. Analysis of the chicken fast myosin heavy chain family. Journal of Molecular Biology 225:1143-1151.

Parker, M. H., P. Seale, and M. A. Rudnicki. 2003. Looking back to the embryo: defining transcriptional networks in adult myogenesis. Nature Reviews 4:495-505.

Punch, V. G., A. E. Jones, and M. A. Rudnicki. 2009. Transcriptional networks that regulate muscle stem cell function. WIREs Syst Biol Med. 1:128-140.

Reddish, J. M., M. Wick, N. R. St-Pierre, and M. S. Lilburn. 2005. Analysis of myosin 43

isoform transitions during growth and development in diverse chicken genotypes. Poultry Science 84:1729-1734.

Reis, P. P., L. Waldron, R. S. Goswami, W. Xu, Y. Xuan, B. Perez-Ordonez, P. Gullane, J. Irish, I. Jurisica, and S. Kamel-Reid. 2011. mRNA transcript quantification in archival samples using multiplexed, color-coded probes. Biotechnology 108:7956-7961.

Robbins, J., G. A. Freyer, D. Chisholm, and T. C. Gilliam. 1982. Isolation of multiple genomic sequences coding for chicken myosin heavy chain protein. The Journal of Biological Chemistry 257:549-556.

Rudnicki, M. A., P. N. J. Schneglesber, R. H. Stead, T. Braun, H-H. Arnold, and R. Jaenisch. 1993. MyoD or myf-5 is required for the formation of skeletal muscle. Cell 75:1351-1359.

Terruzzi, I., P. Senesi, A. Montesano, A. L. Torre, G. Alberti, S. Benedini, A. Caumo, I. Fermo, and L. Luzi. 2011. Genetic polymorphisms of the enzymes involved in DNA methylation and synthesis in elite athletes. Physiological Genomics 43:965-973.

Tidyman, W., L. Moore, and E. Bandman. 1997. Expression of fast myosin heavy chain transcripts in developing and dystrophic chicken skeletal muscle. Developmental Dynamics 208:491–504.

Umeda, P. K., C. J. Kavinsky, A. M. Sinha, H. J. Hsu, S. Jakovcic, and M. Rabinowitz. 1983. Cloned mRNA sequences for two types of embryonic myosin heavy chains from chick skeletal muscle. The Journal of Biological Chemistry 258:5206-5214.

Velleman, S. G., S. M. Brown, S. K. Gustafson, L. C. Faustman, P. A. Beaurang, F. Craft, and R. E. Hausman. 1993. Partial characterization of a novel avian defect affecting adult muscle function. Muscle and Nerve 16:881.

Velleman, S. G., J. D. Yeager, H. Krider, D. A. Carrin, S. D. Zimmerman, and R. J. McCormick. 1996. The avian low score normal muscle weakness alters decorin expression and collagen crosslinking. Connective Tissue Research 34:33-39.

Velleman, S. G. 1999. The role of the extracellular matrix in skeletal muscle development. Poultry Science 78:778-784.

Velleman, S. G., and D. C McFarland. 1999. Myotube morphology, expression and distribution of collagen type I during normal and low score normal avian satellite cell myogenesis. Dev. Growth Differ. 41:153–161.

Velleman, S. G., and K. E. Nestor. 2001. Mode of inheritance of the low score normal condition in chickens. Poultry Science 80:1273-1277.

Velleman, S. G., X. Liu, C. S. Coy, and D. C. McFarland. 2004. Effects of syndecan-1 and glypican on muscle cell proliferation and differentiation: implications for possible functions during myogenesis. Poultry Science 83:1020-1027.

Wick, M., J. M. Reddish, N. R. St. Pierre, and M. S. Lilburn. 2003. Biochemical analyses of muscles from poultry bred for rapid growth. Poultry Science 82:1980-1984.

Wick, M., S. G. Velleman, C. S. Coy, D. C. McFarland, C. I. Pretzman, N. R. St-Pierre, and J. C. Sawdy. 2003. Ventricular myosin heavy chain isoform expression is altered in vitro in low score normal chickens. Comparative Biochemistry and Physiology Part A 136:401-408.

Wilson, B. W., W. R. Randall, G. T. Patterson, and R. K. Entrikin. 1979. Major physiologic and histochemical characteristics of inherited dystrophy of the chicken. Annals New York Academy of Sciences 317:224-246.

Zhang, M., and I. S. McLennan. 1998. Primary myotubes preferentially mature into either the fastest or slowest muscle fibers. Developmental Dynamics 213:147-157.