|UFDC Home||myUFDC Home | Help|
This item has the following downloads:
INVOLVEMENT OF BMP6 AND E2F5 INT SKELETAL MYOGENESIS
JENNELLE ROBINT MCQUOWN
A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
@ 2007 Jennelle Robin McQuown
To Bill and my family for all of their love and support. I would be lost without you guys! All of
my love, J.
I would like to thank my maj or professor, Dr. Sally Johnson, for serving as my supervisory
committee chair. I would also like to thank my committee members (Dr S. Paul Oh, Dr. Lori
Warren, and Dr. Lokenga Badinga) for their support and advice. I could not have asked for a
more supportive and encouraging group. I would also like to thank the members of Dr. Sally
Johnson' s lab for their assistance with my dissertation proj ect. I also count myself lucky to have
a great group of friends and family who served as my sounding board, my cheering squad, and
my source of inspiration.
The completion of my Ph.D. has definitely been a family goal. Without the love and
support of my family, especially my mom and dad, Linda and Dan McQuown, I might have not
achieved this accomplishment. Another huge source of encouragement and support was William
Pittsley. I feel that I share this degree with him since he has gone through every triumph, every
failure, and every fear and hope along with me, and I am eternally grateful. I also send prayers
up to my Grandpa Cossentine and Grandpa McQuown whom weren't able to see me finish this
degree but I know that they're both smiling down and sending me their love. For my Grandma
McQuown and Grandma Cossentine, I am so thankful that they are here to encourage me, send
me their love, and call me the first Dr. McQuown of the family!i Finally, I would like to thank all
of my wonderful friends. I feel lucky to have such a great cheering squad!
TABLE OF CONTENTS
ACKNOWLEDGMENTS .............. ...............4.....
LIST OF TABLES ................ ...............8............ ....
LIST OF FIGURES .............. ...............9.....
AB S TRAC T ............._. .......... ..............._ 1 1..
1 LITERATURE REVIEW ................. ...............13...............
Introducti on ................. ...............13......__ ......
Skeletal Muscle Biology .................. .... ....._._ ..... .............. ...... .........1
The SR and Transverse Tubule Systems within Skeletal Muscle ..........._...._ ................14
Maj or Fiber Type Classification ................. ...............15........... ...
Muscle Architecture .............. ...............16....
Skeletal Muscle Function .............. ...............19....
Skeletal Muscle Development ................... .......... ...............22......
Myogenic Regulatory Factors (MRF s) ................. ...............25........... ...
Satellite Cells............... ....... ............2
Satellite Cell Marker Proteins .............. ...............28....
c-M et .................. ........... ...............29.......
Syndecans 3 and 4 .............. ...............30....
M -cadherin .............. ...............30....
CD3 4............... ...............3 0.
Pax7 ................. .... ..... .... .......... ............3
Myocyte nuclear factor (MNF) .............. ...............32....
N otch .............. .. ... ............. .... .........3
Growth Factor Effects on Satellite Cells .............. ...............34....
Insulin-like Growth Factor I (IGF-I) .............. ...............35....
Fibroblast Growth Factors (FGFs) .............. ...............36....
Hepatocyte Growth Factor (HGF) .................. ........... .... ...............37....
Transforming Growth Factor Beta (TGFP) Superfamily .............. .....................3
M yostatin ................... ... ... ....... ............ .............3
Bone Morphogenic Proteins (BMPs) .............. ... .... ... .............4
BMP Function in vivo: Lessons Learned from Knockouts .............. .....................4
Current BMP6 Studies............... ...............45
Ras/Raf in Skeletal Muscle............... ...............46.
The Raf Family ............. ..... __ ...............47..
E2F Fam ily .................. ... ............ ... ....................4
E2F Signaling Mechanism and Involvement of Pocket Proteins .............. ........_......49
Cell Cycle Repression ...................... ..... .............5
E2Fs and Pocket Proteins in Skeletal Muscle .............. ...............51....
2 MATERIALS AND METHODS .............. ...............56....
M yoblast Cell Culture................ ...... ..... ................5
Growth Factor Treatment, and BrdU Pulsing and Fixation............... ...............57
Plasmids and Transfections .............. ...............57....
BrdU Staining and BrdU Incorporation............... .............5
Immunofluorescent E2F5 Staining ................. ...............58._._. ......
RNA Isolation and Nylon Arrays .............. ...............59....
Western Blots................ .. ...............59
Alkaline Phosphatase Staining .............. ...............60....
p38 Inhibition Assays .............. ...............61....
Apoptosi s Analysi s ................ ...............61................
Stati sti cs ................. ...............61........... ....
3 DIFFERENTIAL EXPRESSION OF TGFP SUPERFAMILY MEMBERS DURING
SKELETAL MYOGENESIS .............. ...............63....
O bj ective ............... .. ........ .... .... ...._........ ... ..........6
Differential Transcriptional Activity in Myoblasts versus Myofibers ................. ...............63
Differentiation TGFP Gene Expression in Myoblasts and Myofibers ................. ...............64
Discussion ................. ...............65.................
4 IMPACT OF BMP6 ON SKELETAL MYOGENESIS ................ ............................75
Obj ective ................... ............ ........ ........ .. .................7
Inhibition of Skeletal Myogenic Differentiation by BMP6 ....._____ .........__ ..............75
Dose-Dependent Effects of Recombinant BMP6 on Myoblasts ................. .....................76
Induction of ALP Activity in Response to BMP6 ....___......_____ ..... ....___........7
BMP6 Induces Rapid Transdifferentiation in Myoblasts ................ ......... ................77
BMP6 Does Not Alter Proliferation Rates of Myoblasts ................. ......... ................77
BMP6 is not Anti-Apoptotic............... .............7
Discussion ................. ...............78.................
5 BMP6 SIGNALING DURING SKELETAL MYOGENESIS............... ...............9
Obj ective .................. .. ........... ......... .... .... ..........9
Analysis of BMP Signaling Systems in Myoblasts ........... ................... ......... .......... .....9
Impact of Notch Inhibitor on BMP6-Mediated Inhibition of Differentiation .........._...........98
Discussion ........._..... ...._... ...............99.....
6 IMPACT OF E2F5 ON SKELETAL MYOGENESIS ......____ ........_ ...............113
O bj ective ............... ...... .........__ ...... ... ... .......... ......... 1
BMP6 Treatment does not cause E2F5 Nuclear Accumulation in Myoblasts ..................113
Presence of E2F5 in Satellite Cell Position ....._ .....___ ........__ ...........1
E2F5 Does Not Inhibit Myofiber Differentiation.............___ ...... ... .. .. .........__ ......11
pRb does not interact with E2F5 to Exert Inhibitory Effects on Muscle Specific Activity .115
E2F5 is Transcriptionally Active ................. ...............115...............
Discussion ................. ...............116................
7 SUMMARY AND CONCLUSIONS ................ ...............126...............
A GENE ARRAY LAYOUT AND TABLE ................. ...............130..............
B SUMMARY OF ABBREVIATIONS .............. ...............137....
LITERATURE CITED ................. ...............141.............
BIOGRAPHICAL SKETCH ....___ ................ ......._. ..........18
LIST OF TABLES
1-1. Receptor and R-Smad specificity for TGFP superfamily members ................ ................. 53
1-2. TGFP superfamily transgenic knockouts .............. ...............54....
LIST OF FIGURES
3-1. Differential transcriptional activity in myoblasts versus myofibers ............... ...............69
3-2. Intact RNA and cDNA probe synthesis............... ...............7
3-3. Gene expression profies in embryonic and satellite cell myofibers ...........__.................71
3-4. Differential gene expression in myoblasts and myofibers............... ...............7
3-5. Relative BMP gene expression in myoblasts and myofibers .............. ....................7
3-6. BMP6 does not undergo autocrine gene regulation ........._._.._......_.. .................7
4-1. Biochemical inhibition of skeletal myogenesis by BMP6............... ...............88..
4-2. Inhibition of skeletal myogenesis by BMP6............... ...............89..
4-3. BMP6 dose response curve............... ...............90.
4-4. BMP6 induction of alkaline phosphatase activity .............. ...............91....
4-5. BMP6 does not induce alkaline phosphatase (ALP) activity in fibroblasts. .......................92
4-6. BMP6 induces rapid transdifferentiation in myoblasts .............. ...............93....
4-7. BMP6 treatment does not alter my oblast proliferation .............. ...............94....
4-8. BMP6 is not anti-apoptotic............... .............9
5-1. Potential BMP6 signaling pathways affecting skeletal myogenesis .............. ..................106
5-2. Verifieation of BMP and TGFP signaling axis. ............. ...............107....
5-3. BMP6 treatment increases p38 phosphorylation ......___ .... ... ..__ .. ....__ .......10
5-4. p38 inhibition and BMP6 treatment result in additive inhibition of muscle specific
reporter activity. ........... ..... .._ ...............109...
5-5. BMP6 and p38 signal through independent pathways to influence myogenic
different ati on ................. ...............110._._._ .....
5-6. p38 signaling does not play a significant role in transdifferentiation. ........._.._... ...............111
5-7. BMP6 inhibition of differentiation is mediated in part by Notch. .............. .............. .1 12
6-1. BMP6 treatment does not cause E2F5 nuclear accumulation in myoblasts ......................121
6-2. Presence of E2F5 in satellite cell position............... ...............12
6-3. E2F5 does not inhibit myogenic differentiation ........................... ........._. ......12
6-4. pRb does not interact with E2F5 to exert inhibitory effects on muscle specific activity ....124
6-5. E2F 5 i s transcriptionally active ................ ...............125......_.__..
7-1. Proposed model of BMP6 on myogenic differentiation and interaction with Notch..........129
Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy
INVOLVEMENT OF BMP6 AND E2F5 INT SKELETAL MYOGENESIS
Jennelle Robin McQuown
Chair: Sally E. Johnson
Major: Animal Sciences
An amazing dynamic exists within skeletal muscle that is required for the development and
maintenance of the musculature system in response to stimuli. A differential BMP- and TGFP-
responsiveness and gene expression profile in myoblasts versus myofibers, suggests that TGFP
superfamily members play distinct regulatory roles at specific stages of myogenesis. Treatment
of myoblasts with BMP6 results in an increase in alkaline phosphatase activity in a dose-
dependent and time-dependent manner. Exogenous BMP6 treatment results in inhibition of
myoblast differentiation as observed by significant inhibition of muscle reporter activity, muscle
specific protein synthesis, and myoblast fusion. BMP6 treatment does not alter proliferation
rates or play an anti-apoptotic role in myoblasts. Inhibition of p3 8 activity and BMP6 treatment
caused significant inhibition of TnI-Luc activity and muscle specific protein expression markers
suggesting an additive effect through independent signaling cascades. Interestingly, the
combination of BMP6 and Notch inhibition partially restores MyHC expression in fibers. This
suggests that the myogenic inhibitory effect observed in the presence of BMP6 is not a cell cycle
effect and not a direct target of the RaflMEK/ERK signaling axis, but is partially mediated by
functional Notch signaling. Therefore, BMP6-mediated inhibition of differentiation is regulated
or partially controlled by the Notch signaling pathway. Ectopic BMP6 treatment did not induce
E2F5 nuclear accumulation in myoblasts further suggesting BMP6 is not causing a cell cycle
effect. E2F5 is present in non-dividing cells in vivo outside the dystrophin border in the putative
satellite cell position, although E2F5 did not impact myogenic differentiation. Additionally, the
inability of E2F5 to inhibit differentiation is not due to insufficient pocket protein function.
Multiple signaling axes are key factors during Raf-imposed block to myogenic differentiation.
Further understanding of the regenerative ability in skeletal muscle in response to stimuli would
be useful for patients or animals that experience severe muscle trauma, and for individuals with
skeletal muscle disorders, and would provide additional information for human therapeutic and
agricultural applications that would benefit from enhanced muscle mass.
Within an organism, skeletal muscle functions in locomotor activity, postural behavior,
and breathing. Skeletal muscle will undergo injury in response to direct trauma or as an indirect
result of neurological dysfunction or genetic defects, and if not repaired, muscle mass and
locomotion abilities may be lost, and death can possibly occur (Charge and Rudnicki, 2004). An
amazing dynamic exists within skeletal muscle that is required for the maintenance of the
musculature in regards to the states of atrophy, injury, and subsequent repair. Understanding the
regenerative ability of skeletal muscle in response to these stimuli are important areas of
research. For instance, patients who experience severe muscle trauma, individuals with genetic
skeletal muscle disorders like muscular dystrophy, and improving the recovery of astronauts
experiencing a weightless environment would benefit from such studies. Many changes to
muscle physiology occur in an adaptive response to decreased or increased usage of skeletal
muscle (Stein and Wade, 2005). The adaptive responses of decreased usage includes a shift in
myosin isoforms from slow to fast fiber type, replacement of protein with fat within the muscle,
shifts in energy source for metabolism away from lipids to glucose, a loss of bone mineral
density, and an increase in bone resorption (Shackelford et al., 2004; Stein and Wade, 2005).
Skeletal Muscle Biology
The mechanics of skeletal muscle contraction begin with a neural stimulus for contraction.
This results in a generation of an action potential in a motor neuron causing the release of
acetylcholine into the synaptic cleft of the neuromuscular junction (NMJ). This released
acetylcholine binds with the receptors on the motor end-plate, producing an end plate potential,
which leads to the depolarization that is conducted down the transverse tubules deep into the
muscle fiber. Depolarization results in calcium being released from the sarcoplasmic reticulum
(SR) (Crouch, 1985).
In the resting state, myosin cross-bridges remain connected to actin in a weak binding
state. When depolarization signal reaches the SR, calcium is released into the sarcoplasm, which
binds to troponin, causing a shift in troponin position to uncover the "active" sites on actin. The
"energized" or "cocked" myosin cross-bridge then forms a strong bond at the active sites on
actin. Inorganic phosphate released from the myosin cross-bridge energizes the cross-bridge to
allow it the ability to pull the actin molecules. Cross-bridge movement is completed by the
release of ADP from the myosin cross bridge. At this point, the myosin cross-bridges are in
what is considered the "strong binding state" with actin. Attachment of ATP to myosin allows
the cross-bridge to break the strong binding state and form a "weak binding state". ATP is then
broken down to ADP + Pi + energy. The energy is released and used to "energize" the myosin
cross-bridges. This cycle can repeat as long as calcium and ATP are present. The cycle is
stopped when the SR actively removes calcium from the sarcoplasm (Lieber, 2002).
The SR and Transverse Tubule Systems within Skeletal Muscle
Within the muscle, two membrane systems are present to activate the filaments. These two
systems are called the transverse tubule (T) system and the SR system. The T-system consists of
invaginations in the surface membrane physically contiguous with the sarcolemma, or holes that
extend into the fiber and transversely crosses the long axis of the fiber. The T-system conveys
activation signals received from the motor neuron into the myofibrils. Since motor neurons are
not in direct contact, the T-system helps increase activation time versus a simple diffusion
system. The release of calcium switches on muscle activation and the removal of calcium from
the myofilament cause muscle relaxation. Embedded in the SR membrane are specific calcium
channels and pumps that control this calcium release and uptake. The SR envelops each
myofibril to provide contact between the activation and force generation systems. The SR also
contacts the T-system, thus acting as a middle man in skeletal muscle activation and relaxation.
The T and SR systems are arranged in a structure called a "triad", which is a single T tubule
surrounded by two SR tubules located at each Z disk and the triad functions as the interface
between the extracellular and intracellular surfaces of the fiber (Lieber, 2002).
Additionally, the SR and T systems are linked by junctional feet structures. Junctional feet
are composed primarily of two components: ryanodine receptors (RyR) and dihydropyridine
receptors (DHP). RyRs are voltage sensing proteins embedded in the SR that sense action
potential traveling across the T system. DHPs are calcium release channels through which the
myoHilament receives activating calcium (Lieber, 2002).
Major Fiber Type Classification
Skeletal muscles can be classified as either red or white based on the maj or proportion of
red or white fibers they contain. Few muscles are composed solely of one fiber type; most are of
a heterogeneous composition. The red coloration of red fibers is due to higher myoglobin
content as compared to white fibers and the myoglobin allows for oxygen storage required for
oxidative metabolism. Conversely, white fibers have a high content of glycolytic enzymes and
low levels of oxidative enzyme activity (Smith, 1972).
Another method of skeletal muscle fiber designation is based upon contraction rates: slow-
twitch vs. fast-twitch fibers or type I vs. type II. There are four different adult myosin isoforms
in skeletal muscle: type I, IIA, IIX(D), and IIB with type I and IIA designated as red muscle
fibers and type IIX(D) and IIB designated as white muscle fibers (Barany, 1967; Roy et al.,
1984). Slow-twitch fibers, which are also called Type I fibers, are characterized by higher
number of mitochondria, larger mitochondria, and the mitochondria are located in two general
areas: the subsarcolemma, and the intermyofibrillar. The mitochondria within Type I fibers are
more efficient than in Type II fibers. For example, in the subsarcolemma, the mitochondria are
nine times more efficient in Type I and in the myofibrillar, they are five times more efficient in I.
These fibers are surrounded by numerous capillaries and have a higher concentration of
myoglobin, which gives the red muscle fiber types their visual color. Type I fibers have a large
capacity for aerobic metabolism and can use ATP more efficiently than Type II without lactic
acid buildup. The lipid content is greater in red muscle fibers, which serves as a metabolic fuel
source and while these fibers contract slower, they can contract for a longer time. These fibers
play a role in posture because they are less easily fatigued, as long as oxygen is available and due
to their higher resistance to fatigue and allowance for a higher maximal oxygen uptake (VO2
max) they are selected for in endurance type training (Evans et al., 1994; Van Swearingen and
Lance-Jones, 1995; Lieber, 2002).
Fast twitch or Type II fibers are glycolytic and can produce force at a higher rate. This is
called a phasic mode of action since contraction occurs in short bursts and these fibers are more
easily fatigued. White fibers have a more extensively developed SR and T-tubule system along
with more narrow Z disks versus white fibers. They also have fewer and smaller mitochondria
than Type I and the glycolytic metabolism that predominates in these fibers can occur both
aerobically or anaerobically. White fibers have a lower capillary density than red fibers.
Functional implications demonstrate that Type II fibers are increased during strength training due
to their ability to produce force at a higher rate (Zhang and McLennan, 1998; Lieber, 2002).
The anatomical and biomechanical properties of skeletal muscle determine the efficiency
and overall ability of the muscle to generate force and movement. The largest functional unit of
contractile filaments is the myofibril, which is a string of sarcomeres. The muscle fiber is
composed of myofibrils arranged side-by-side and the groups of muscle fibers are sheathed in
connective tissue termed the perimysium (Lieber, 2002).
The number of sarcomeres within a fiber depends on the muscle fiber length and diameter
of the fiber. They are the most important determinant of muscle fiber function. The total
distance of myofibrillar shortening is equal to the sum of the shortening distance of each
sarcomere. This allows for whole muscles to be able to shorten up to several centimeters even
though each sarcomere can only shorten ~1 Clm giving the muscle a tremendous ability to adapt.
Within the sarcomere are two maj or types of contractile filaments, the thick filaments and the
thin filaments. These filaments are large polymers of myosin and actin proteins. In thick
filaments, which are the myosin-containing filament, a "feathered" appearance is observed with
proj sections coming out at either end of the filament because one molecule rotates ~600 relative to
molecules on either side. These filaments generate the tension during muscle contraction and the
thin filaments or actin-containing filament regulates the tension generated (Gordon et al., 1966a;
Gordon et al., 1996b; Lieber, 2002). Tension generation is also a function of the magnitude of
the overlap between the actin and myosin filaments. Passive tension plays a role in providing
resistance in the absence of muscle activation. The source of passive tension is due to the protein
titin, which connects the thick myosin filaments end to end (Magid and Reedy, 1980; Labeit and
Striated skeletal muscle is composed of bundles of enormous, multi-nucleated cells and the
striations arise from a repeating pattern in the myofibrils. Active interdigitations of these
filaments produce muscle shortening and give muscle its striated appearance. Some
characteristics of striated muscle are its ability to contract and generate tension when activated
and its ability to return to its original length and form after contraction or stretching ceases.
These properties are primarily the result of separate sets of filament systems: contractile actin
and myosin filaments and viscoelastic titin and intermediate filaments (Vigoreaux, 1994; Knupp
et al., 2002).
Within the thin filaments, approximately 5% of the myofibrillar protein is composed of
troponin molecules, which are located approximately every seven or eight actin molecules along
the actin filament. Troponins are the regulatory proteins responsible for turning on contraction.
Troponin is composed of three subunits: Tn-I, Tn-C, and Tn-T. Tn-I exerts an inhibitory
influence on tropomyosin when calcium is not present. Tn-C: binds calcium during contraction.
Tn-T: binds troponin to tropomyosin. Tropomyosin is a long, rigid, and insoluble rod-shaped
molecule that stretches along in close contact with each strand of the thin filament. Within each
groove of the actin super-helix lays a strand of tropomyosin and a single molecule extends the
length of seven G-actin molecules (Lieber, 2002).
The dark striations observed in skeletal muscle are a lattice of thick filaments termed the A
band because it appears dark under a microscope. The light striations are termed I bands, which
are regions of the myofibril containing only thin filaments and Z disks. The Z disks are the
structure that attach to the thin filaments to act as actin-anchoring structures and function as a
common link to mechanically integrate contractile and elastic elements (Schroeter et al., 1996).
They are also involved in transmission of active and passive forces and differences in Z band
structure have been described for distinct classes of muscle and fiber types (Vigoreaux, 1994).
Thin filaments extend from two adj acent Z lines to interdigitate with the thick filaments. Thick
and thin filament arrays form the contractile system and are the repeating unit in each myofibril
(Bloom and Fawcett, 1968).
Intermediate Filaments (IFs) are one of the three maj or classes of cytoskeletal proteins
along with microtubules and microfilaments. The IF proteins and IF-associated proteins such as
desmin, vimentin, and nestin, are involved in maintaining the structural and functional integrity
of all muscle types and align themselves in a head-to-tail dimer along the central rod domain.
For example, in the skeletal muscle injury model, nestin and vimentin IF proteins begin to be
expressed 6 hours post-injury in myoblasts and their maximal expression was observed around 3-
5 days post injury. Thereafter, vimentin expression ceases completely, whereas nestin is found
to remain only in the sarcoplasm next to neuromuscular and myotendinous junctions. At 6-12
hours post-injury, desmin expression is upregulated and becomes the predominant IF protein in
myofibers, coinciding with the appearance of cross-striations. Nestin and vimentin are essential
during the early phases of myofiber regeneration and desmin is responsible for maintenance of
myofibrils in mature myofibers (Vaittinen et al., 2001).
A protofibril is a complex of eight IF monomers and the network of IFs form a lattice of
connections that link the different parts of the muscle fiber. These networks extend from the Z-
disk to adjacent myofibrils and from the sarcolemma to the Z-disc. Additionally, IFs are
associated with the neuromuscular junction, nuclear membrane, and mitochondria located
between adj acent myofibrils (Agbulut et al., 2001; Reipert et al., 1999).
Skeletal Muscle Function
Contraction velocity is based on the efficiency of the energy metabolism. Velocity is
fastest when ATP is generated from an oxidative rather than a glycolytic pathway. In addition,
contraction velocity is also dependent on the type of myosin present in the fiber. The four
Myosin Heavy Chain (MyHC) isoforms are the products of a multigene family whose expression
during development and in the adult are regulated by neuronal, hormonal, and mechanical
factors. Most single fibers (~86%) express only one MyHC isoform but there are examples of
co-expression of MyHC isoforms and the MyHC IIA and IIX/D isoforms are predominately
expressed among fibers (Rivero et al., 1999). In terms of shortening velocity, the gradient
observed for MyHC isoforms is I < IIA < IIX/D < IIB (Bottinelli et al., 1991; Galler et al., 1994).
There are also phenotypic differences associated with MHC isoform composition in terms
of metabolic and size properties of the muscle fiber types (Bar and Pette, 1988; Schiaffino et al.,
1989). For example, motor units containing more MyHC I or IIA are more resistant to fatigue
than motor units containing MyHC IIX/D or IIB. Additionally, MHC IIB fibers are generally
larger than MyHC I or IIA (Rivero et al., 1999).
The globular head on the myosin has ATPase activity. This activity determines how fast
the muscles contract, and is determined by how quickly ATP is hydrolyzed. In terms of the
force-velocity relationship, this illustrates that the amount of force generated by a muscle is
highly proportional to its velocity or velocity is dependent on how much force is resisting the
muscle. Maximum contraction velocity is termed Vmax, which is one of the most commonly used
parameters to characterize muscle, and is relative to both fiber type distribution and architecture
Function is largely determined by a muscle's architecture. While fiber size between
muscles does not vary greatly, there are maj or architectural differences between different muscle
groups and these are the best predictors of force generation (Burkholder et al., 1994). Skeletal
muscle architecture is defined as the arrangement of muscle fibers within the muscle relative to
the axis of force generation. There are three classifications of muscle arrangements termed
parallel, uni-pennate, and multi-pennate. Parallel muscles are composed of fibers that extend
parallel to the muscle's force-generating axis. Uni-pennate muscles have fibers that are
orientated at a single angle relative to the force-generating axis typically in an angle of 0 to
30 degrees. Most muscles are multi-pennate, where the fibers are orientated at several angles
relative to the axis of force generation (Lieber, 2002). Other parameters that help characterize
the properties of various muscles or muscle groups are the sarcomere length, which also
translates into fiber length, range of movement (ROM), and the physiological cross-sectional
area (PCSA). The PCSA of a muscle represents the sums of the cross-sectional areas of all of
the muscle fibers within the muscle and is directly proportional to the maximum titanic tension
generated by the muscle (Powell et al., 1984).
Tying together some of these parameters give a better indication of functional properties of
specific muscle groups. For example, the quadriceps muscles have relatively high pennation
angles, large PCSAs, and short fibers. Therefore, this makes them suited for the generation of
large forces. Hamstring muscles like the sartorious, semitendinosus, and the gracilis muscles
have low PCSAs and extremely long fibers, which allow for large excursions at low forces.
Conversely, the soleus muscles have high PCSAs and short fiber lengths so that these muscles
can generate high forces with small excursions making these suitable muscle groups for a
postural stabilization role (Edman et al., 1985; Woittiez et al., 1985).
There is also different terminology related to different types of contractions. For example,
isometric or static contractions are when the muscle tension increases but the body part does not
move. Concentric contractions are when the action of the muscle results in muscular shortening
as the body part moves and the force is applied to the muscle as it contracts. Eccentric
contractions occur when the muscle is activated and force is produced as the muscle lengthens.
Therefore, these types of contractions are more likely to result in muscle injury or soreness
versus concentric contractions. Additionally, muscle strengthening regimens are greatest when
exercises using eccentric contractions are utilized (Lieber and Friden, 2002). These factors
demonstrate why muscle architecture is also a consideration in training regimes for human and
animal athletes. As more information is generated on the basic understanding of muscle form
and function, there can be further benefits for athletic training and performance.
Skeletal Muscle Development
Skeletal muscle development is a highly complex and orchestrated process, which is still
not completely understood, both at the macroscopic and molecular levels. Skeletal muscle cells
are derived from the somites, except for the craniofacial muscles in mammals and birds (Armand
et al., 1983). Somites are balls of epithelial cells formed from the paraxial mesoderm in pairs on
either side of the neural tube, and differentiate into two regions the dermomyotome and
sclerotome. Due to their interactions with surrounding tissues, the ventromedial portion of the
somite gives rise to the sclerotome, which is the precursor to the ribs and axial skeleton, and the
dorso-lateral part forms the dermomyotome, which is where the myogenic precursors and dermis
originate in embryogenesis (Braun et al., 1992; Rudnicki et al., 1993; Christ and Ordahl, 1995).
Within the dermomyotome, the epaxial muscles and subsequent body wall muscles are formed
on the dorso-medial lip. The hypaxial muscles, which consist of the limbs, tongue, diaphragm
and ventral wall musculature, are formed on the dorso-lateral lip. Endothelial precursors are also
derived from the somites (Chevallier et al., 1977; Beddington and Martin, 1989; Ordahl and
LeDouarin, 1992; Wilting et al., 1995; Kardon et al., 2002).
Multiple pathways mediate the embryonic processes throughout skeletal muscle formation.
In terms of limb bud development, the progenitors of this process evolve by delamination and
migration distally into the developing limb bud in response to molecular signals from the
adjacent lateral plate mesoderm (Chevallier et al., 1977; Christ et al., 1977; Solursh et al., 1987;
Hayashi and Ozawa, 1995). Induction of premyogenic cells in the limb to delaminate is
mediated by hepatocyte growth factor (HGF) or scatter factor and by fibroblast growth factor
(FGF). In experiments where both FGF2 and HGF were applied ectopically in the interlimb
flank mesenchyme of chicks, delamination of the lateral dermomyotome was observed (Brand-
Saberi et la., 1996; Heymann et al., 1996). Additionally, when c-met, the tyrosine kinase
receptor for HGF, is inactivated, delamination in mice is prevented. Both HGF and FGF2 and
FGF4 are also found to promote migration by acting as a chemotactic source resulting in cell
migration towards the distal tip of the limb bud (Itoh et al., 1996; Takayama et al., 1996; Webb
et al., 1997; Scaal et al., 1999).
Transcription factors are involved in delamination and migration, with Pax3 and Lbx1
being initially expressed in the lateral dermomyotome. Pax3 is necessary for the epithelial-
mesenchymal transformation of the lateral dermomyotome (Daston et al., 1996). In Pa-x3 null
mice, which are also called Splotch mice, c-met expression is either significantly reduced or
absent in the lateral dermomyotome. This suggests that Pax3 is upstream of c-met, since the
dermomyotome is disorganized and the limb myogenic cells do not migrate (Epstein et al., 1996;
Yang et al., 1996; Mennerich et al., 1998; Tremblay et al., 1998). Inactivation of Lbx1 causes
premoygenic cells to delaminate properly but have dysfunctional migration. The resulting
phenotype exhibits minimal hindlimb musculature and the extensors (or dorsal muscles) are
missing in the forelimb. These observations are similar to experiments where Gab1 is
inactivated, which is also involved in c-met signaling and is also believed to be a result of cells
being unable to respond to limb migratory cues (Schafer and Braun, 1999; Brohmann et al.,
2000). Therefore, as these premyogenic cells migrate distally towards the limb bud, they begin
to switch on the myogenic determination bHLH transcription factors, M~yoD and M~yf5, and form
the dorsal and ventral muscle masses. The myogenic cells then terminally differentiate into slow
or fast fibers, which express the relevant form of slow or fast MyHC. The fast fibers are needed
for movement with the slow fibers helping maintain posture (Francis-West et al., 2003).
Another maj or player in skeletal muscle formation is Sonic hedgehog (Shh), which is
expressed in the notochord and floor plate of the neural tube and induces the formation of the
sclerotome. Along with the Wnt proteins (1, 3a, 7a, 11), Shh also promotes myogenesis by
activating expression of Myf5 in the epaxial myotome through its interaction with Wntl. Wnt1 is
expressed in the dorsal neural tube and the dorsal ectoderm. Wnt7a is expressed in the dorsal
ectoderm and found to activate the expression ofhyoD in hypaxial myotome (Schmidt et al.,
2000; Petropoulos and Skerjanc, 2002).
The bone morphogenic proteins (BMPs) are expressed in the lateral plate mesoderm and
limb bud ectoderm. Noggin, an antagonist for BMP, is expressed in the dorsomedial lip of
dermomyotome. BMPs and Noggin regulate Pa-x3 in proliferating muscle precursor cells
(MPCs), and Pax3 is critical for the downstream activation for Myogenic Regulatory Factors
(MRFs) (Amthor et al., 1998). The levels of BMPs within the somite, in the presence or absence
of its antagonists, controls the expression of Pax3 and can either promote embryonic muscle
growth by expanding Pax3-expressing muscle precursor cells or restricting development by
inducing apoptosis (Amthor et al., 2002). In the limb, BM~P2 and BM~P7 expression is increased
by Shh. This causes a stimulation of muscle growth and a delay in muscle differentiation by
allowing these BMPs to inhibit activation of MyoD and Myf5 by Pax3 (Tajbakhsh et al., 1997).
The BMP signals originating from the lateral plate delay the activation of myogenic bHLH gene
expression (Pourquie et al., 1996). In the presence of Noggin, BMPs are sequestered, Pax3
expression is repressed, and MyoD expression is dramatically expanded, inducing formation of a
lateral myotome (Duprez et al., 1998; Reshef et al., 1998; Kriiger et al., 2001). Conversely,
another BMP antagonist, follistatin, promotes Pax3 expression, which transiently delays muscle
differentiation and exerts a proliferative signal during muscle development (Amthor et al., 2002).
Myogenic Regulatory Factors (MRFs)
At the molecular level, MRFs mediate the process of myogenic determination and muscle-
specific gene expression enabling multipotent, mesodermal cells to give rise to mononucleated
myoblasts, withdraw from the cell cycle and differentiate into multinucleated muscle fibers,
which are the framework of whole muscle (Cooper and Konigsberg, 1961; Stockdale and Holtzer
1961; Stockdale et al., 1999). MRFs are a family of skeletal muscle specific bHLH transcription
factors and contain the genes, Myf5, M~yoD, M~RF4, and myogenin. When these factors are
overexpressed in non-myogenic cell lines, they cause them to differentiate into myogenic cells
(Davis et al., 1987; Wright et al., 1989; Braun et al., 1989). M~yoD and M~yf5 are expressed in
proliferating, undifferentiated cells and when cells begin to terminally differentiate, myogenin
expression is induced (Smith et al., 1993, 1994; Andres and Walsh, 1996). M~RF4 is expressed in
both early stages of myogenesis, during muscle development, as well as in adult muscle tissue
(Hinterberger et al., 1991).
Another family of genes, the MEF2 family augments the myogenic activities of MRFs.
(Shen et al., 2003). MyoD and MEF2 interact to control myoblast specification, differentiation,
and proliferation and MRF4 acts in embryonic cells to control myogenic determination (Kassar-
Duchossoy et al., 2004). However, myogenin expression is required for terminal differentiation
before birth. In addition, the expression of skeletal muscle a-actin is a marker of late
differentiation following myogenin expression (Knapp et al., 2006).
Further mediation of MRF activities is achieved through their heterodimer formation with
another family of bHLH proteins, the E proteins (Lassar et al., 1991; Shirakata et al., 1993). The
E proteins are class I, bHLH proteins. These class I proteins form heterodimers with the class II
bHLH proteins such as MyoD. This heterodimer binds to specific DNA regions (E boxes) that
contain the sequence CANNTG and activate transcription of muscle target genes (Yutzey and
Konieczny, 1992). E47 is an alternate splice product of the E2A gene expressed early in
myogenic differentiation (Sun and Baltimore, 1991; Quong et al., 1993).
Emerging data demonstrates that the intracellular signaling pathway, p38 mitogen-
activated protein kinase (MAPK), participates in several stages of the myogenesis. The most
documented role of p3 8 is its cooperation with transcription factors from the MyoD and MEF2
families in the activation of muscle specific genes. This interaction contributes to the temporal
gene expression during differentiation (Cuenda and Cohen, 1999; Zetser et al., 1999).
Withdrawal of myoblasts at the G1 stage is necessary for differentiation to occur. This is
partially mediated by p38 kinase and c-Jun N-terminal protein kinases (JNKs) signaling through
its induction of p21 expression in myoblasts (Puri et al., 2000; Mauro et al., 2002). The
MyoD/E47 heterodimer is regulated by p38 MAPK and MRF4 is phosphorylated by p38 MAPK
on Ser-3 1 and Ser-42 resulting in a reduced transcriptional activity. The modulation of MRF4
activity also results in selective silencing of muscle-specific genes in terminal differentiation
(Suelves et al., 2004). During somite development in mice, p38 MAPK plays a crucial role in
activating MEF2 transcription factors. In Xenopus laevis, p38 signaling is required for early
expression of2~yf5 and for the expression of several muscle structural genes (Keren et al., 2005).
Negative myogenic regulators exist such as, Id1 and Twist. Id1 is a dominant negative
HLH protein that prevents the interaction of the E proteins with MyoD or Myf5 by preferentially
binding to the E proteins (Norton et al., 1998). Twist complexes with E proteins to repress
myogenic differentiation. Twist also can directly block transcriptional activity of MyoD and
MEF2 by binding to E-box sequences (Hebrok et al., 1997; Puri and Sartorelli, 2000).
In order for muscle differentiation to occur, myoblast cell cycle arrest is critical. MyoD
promotes cell cycle arrest by inducing p21, which is a cyclin-dependent kinase (cdk) inhibitor
(Guo et al., 1995; Halevy et al., 1995). Induction of p21 is subsequent to myogenin expression
and it is believed that high levels of p21 are required for cells to remain in a post-mitotic state.
MyoD also induces retinoblastoma (Rb), which is a negative regulator of G1 progression
(Martelli et al., 1994). There is an intricately regulated pattern of cell cycle proteins that are
temporally expressed based on a variety of molecular interactions and further discussion of these
events is found in a later section. Briefly, G1 progression is controlled by Cdk-mediated
phosphorylation of the Rb protein, which allows for the release and activation of E2F 1
(Lundberg and Weinberg, 1998). E2F 1 mediates S-phase progression by allowing for expression
of genes required for DNA replication and mitosis (Ishida et al., 2001). Withdrawal from the
cell cycle is mediated by p21, which directly inhibits Cdk complexes and interferes with S-phase
progression by binding to and inhibiting the activity of proliferating cell nuclear antigen
(PCNA), a subunit of the DNA polymerase (Dotto, 2000).
More than 40 years ago, Alexander Mauro first identified muscle progenitor cells in frog
skeletal muscle, which were identified as satellite cells based on their observed location adj acent
to mature muscle fibers as seen by electron microscopy (Katz, 1961; Mauro, 1961). These
undifferentiated cells have a degree of plasticity, demonstrating properties of stem cells such as,
yielding all of the specialized cell types from which they originate and having the ability to self-
renew. In mature muscle, in response to injury, hypertrophy, routine maintenance or disease,
satellite cells serve as a reserve of muscle stem cells that are activated from their quiescent or Go
state to re-enter the cell cycle and provide myonuclei for growth or repair (Moss and Leblond,
1971). Self-renewal ability in satellite cells may occur by either a stochastic event or through
asymmetric cell division, and this self-renewal is required for maintenance of their own
population pool (Collins et al., 2005). Satellite cells are derived from a Pax3/Pax7 population of
progenitor cells located in fetal muscle (Relaix et al., 2006).
In a quiescent state, satellite cells are characterized by a low nucleus/cytoplasm ratio, few
organelles, and high amount of heterochromatin or condensed chromosomes, resulting in
minimal metabolic activity or transcription (Schultz, 1976). Quiescent satellite cells can also be
distinguished from activated cells because there are numerous morphological changes that occur
upon activation. During activation, cytoplasmic extensions become apparent, along with an
increase in cytoplasmic volume of the activated cell, the amount of heterochromatin decreases
and organelles such as the Golgi, endoplasmic reticulum, ribosomes, and mitochondria start to
appear (Schultz et al., 1978). Additionally, at the molecular level, MyoD expression is turned on,
and a CD34 isoform switch is observed, along with co-expression of Pax7, M-cadherin, and
Myf5. Cellular proliferation and division is indicated by expression of proliferation markers
such as PCNA, and induction of myogenin expression designates cells undergoing myogenic
differentiation (Fuchtbauer and Westphal, 1992; Grounds et al., 1992; Yablonka-Reuveni et al.,
1994; and Zammit et al., 2004).
Satellite Cell Marker Proteins
In order to distinguish satellite cells from other mononucleated cells within skeletal
muscle, protein markers have been identified which are specific to satellite cells alone or in
combination with other markers that allow for the specific identification of a pure population of
satellite cells. Some of the most studied are CD34, M-cadherin, Pax7, syndecan-3/4, VCAM1,
and c-met (Rosen et al., 1992; Beauchamp et al., 2000; Seale et al., 2000; Cornelison et al.,
2001). Pax7 is currently the best marker for identifying quiescent satellite cells (Seale et al.,
2000). Markers like CD34 are useful in identifying satellite cells on isolated myofibers, but are
not specific or unique to satellite cells and a co-staining or a combination of markers is needed to
definitively identify these cells (Beauchamp et al., 2000). Although, the truncated form of CD34
(Beauchamp et al., 2000) and the P isoform of myocyte nuclear factor (MNF) are specific to
quiescence (Garry et al., 1997; Yang et al., 1997). There is also a level of heterogeneity that
exists within the satellite cell population in terms of immunohi stochemical (IHC) cell staining
and in vitro clonal analyses (Schultz and Lipton, 1982; Baroffio et al., 1996; Molnar et al., 1996).
For example, most satellite cells are positive for both CD34 and M-cadherin and most of these
cells are also positive for Myf5 (Beauchamp et al., 2000), a subpopulation of ~20% are not
positive for these markers. Interestingly, functional studies suggest that this subpopulation of
cells serves as a reserve for satellite cell replenishment (Rantanen et al., 1995; Schultz, 1996).
c-Met is a transmembrane tyrosine kinase receptor, and represents the activated HGF
receptor, which is found in all quiescent satellite cells but not expressed in myofibers (Allen et
al., 1995; Cornelison and Wold, 1997; Tatsumi et al., 1998). However, it is not restricted to
satellite cells, and is detected in both resting and regenerating muscle. The c-met knockout
mouse demonstrates that c-met is necessary for proper limb, diaphragm, and skeletal muscle
formation. Following a crush injury, the mononucleated cells surrounding the necrotic fiber
express the c-met marker. Additionally, cells negative for another stem cell marker, CD34, but
positive for c-met (CD34-/c-met+) are still capable of giving rise to myotubes in culture
(Beauchamp et al., 2000).
Syndecans 3 and 4
Syndecans 3 and 4 have overlapping expression patterns in skeletal muscle with c-met and
are cell surface trans-membrane heparin sulfate proteoglycans. They function in FGF signaling
and the expression of Syndecan 3 and 4 is consistent with that of satellite cell position (between
basal lamina and sarcolemma) and proliferating MPCs (Cornelison et al., 2001). Syndecan 3 and
4 knockouts have severe defects in satellite cell activation and muscle regeneration (Comelison
et al., 2004). Syndecans play a role in signaling through the ERKl/2 MAP kinase pathway.
Primary mouse satellite cells require heparin sulfate for normal proliferation (Comelison et al.,
2001). Activation and initiation of myogenesis is delayed in vitro when syndecan signaling
events are blocked (Comelison et al., 2004).
M-cadherin is a calcium-dependent adhesion molecule, which is used as a marker of
quiescent satellite cells and activated myogenic precursors, but is not expressed in differentiated
myotubes (Irintchey et al., 1994; Moore and Walsh, 1993; Beauchamp et al., 2000; LaFramboise
et al., 2003). It is expressed in some, but not all quiescent satellite cells, demonstrating
heterogeneity within satellite cell compartments (Cornelison and Wold, 1997). Those that do
express M-cadherin also express CD34 (hematopoietic stem cell marker) (Beauchamp et al.,
2000). Studies suggest that M-cad+/CD34+ satellite cells are also Myf5-positive (Beauchamp et
al., 2000). Another observation is that the number of M-cadherin positive cells increases upon
satellite cell activation (Comelison and Wold, 1997).
CD34 is expressed in cells that express no cardiac, hematopoetic, or skeletal muscle
mRNA transcripts, indicating a lack of lineage. This follows with the idea that only a sub-
population of satellite cells expresses the CD34 marker and designates a sub-population of
satellite cells that are less committed to the myogenic lineage (Beauchamp et al., 2000). Cells
that express CD34, don't express Pax7 and M~-cadherin, and both CD34 positive and negative
progenitor cells can give rise to myotubes in vitro. One study examined CD34-positive versus
CD34-negative primary myoblasts and found that CD34-positive primary myoblasts were more
efficient in participating in regeneration but CD34-negative cells had a higher fusion index in
vitro (Beauchamp et al., 2000).
Pax7 is a member of the paired box family of transcription factors, and is localized to
nuclei situated in discreet peripheral locations within resting adult skeletal muscle (Seale et al.,
2000; Relaix et al., 2005). The number of cells positive for Pax7 is believed to correlate well
with the expected number of satellite cells. This is because Pax7 expression co-localizes with
myostatin, c-met, and m-cadherin in satellite cells resting between basal lamina (Seale et al.,
2000; LaFramboise et al., 2003; McCroskery et al., 2003; Halevy et al., 2004). In addition,
myogenic cells lines, which model quiescent, undifferentiated myoblasts, appear to be uniquely
marked with high levels of Pax7. Furthermore, the expression seen in proliferating primary
myoblasts is down-regulated upon myoblast differentiation. Although loss of Pax7 does not
induce differentiation, indicating that other factors must be present or absent for myoblasts to
commence terminal differentiation (Seale et al., 2000; Olguin and Olwin, 2004).
Pax7 inhibits myogenic conversion induced by MyoD. It is believed that Pax7 is indirectly
interfering with its function or competes for proteins necessary for MyoD-dependent
transcription since the presence of Pax7 cannot inhibit the effects of MyoD-E47 heterodimers
(Olguin and Olwin, 2004). Yet, overexpression of Pax7 in satellite cells does induce cell cycle
exit, prevention of BrdU incorporation, a decrease in M~yoD expression, and prevention of
myogenin induction (Olguin and Olwin, 2004). In addition to its impact on MyoD, it is believed
that myoblasts expressing the Pax7+/MyoD- phenotype may return to quiescence to replenish
satellite cell pools while myoblasts that acquire MyoD proliferate and fuse to form myotubes
(Zammitt et al., 2004; Relaix et al., 2005). Alternatively, one study showed that Pax7 may have
anti-apoptotic functions. Myoblasts transduced with dominant negative Pax7 lead to cell death
(Relaix et al., 2006).
Myocyte nuclear factor (MNF)
MNF is a winged helix transcription factor, also identified as Foxkl, expressed in both
cardiac and skeletal muscle, and in quiescent satellite cells. The number of MNF positive cells
increases when muscle is induced to regenerate, and while detected in regenerating myotubes,
M~NF is down-regulated during late stages of regeneration (Garry et al., 1997). There are two
isoforms of MNF, MNF-oc and MNF-P, and both are detectable in skeletal muscle. Alpha is
expressed in proliferating MPCs, while beta is limited to quiescent satellite cells and down-
regulated upon activation (Yang et al., 2000). In quiescent satellite cells, high MNF-P
expression allows for the formation of repressive complexes with mSin3 family members to
repress targeted gene transcription (Garry et al., 2000; Yang et al., 2000). Upon satellite cells
activation, isoform switching disrupts the repression by mSin3, allowing for the targeted genes to
become active (Yang et al., 1997).
Notch signaling plays a role in cellular homoestasis and cell fate determination (Hing et
al., 1994). In mammals, there are four genes, Notchl, Notch2, Notch3, and Notch4 and five
ligands, Dll-, Dll-3, Dll-4, Jagged-1, and Jagged-2 (Chitnis et al., 1995; Lindsell et al., 1995; Li
et al., 1998). The Notch receptors are transmembrane receptors and when bound by the extra-
cellular ligands, Delta, Serrate, or Lag-2, this results in a cleavage of the intracellular domain of
the receptors in the cytoplasm. This cleavage product is the active form of Notch, NICD (Notch
Intracellular Domain), which translocates to the nucleus where it binds to the family of
transcriptional repressors, CSL (also known as RBP-JK, CBF-1, Suppressor of Hairless, and Lag-
1). This interaction converts them into transcriptional activators (Nye et al., 1994; Ahmad et al.,
The molecule Numb is an inhibitor of Notch but little is known about the regulation of
Numb at the transcriptional level. Numb localizes to one pole during cell division in crescent-
shaped patterns so that only one daughter cell receives Numb. This asymmetric expression is
believed to play a role in determining which cell goes on to differentiate into a myofiber and
which cell is going to replenish the satellite cell pool. Numb and Pax3 protein expression are
mutually exclusive in satellite cells and a Pax3+/Numb- cell indicates a satellite cell that is less
committed to a specific phenotype. The daughter cell retaining Numb expression is more
committed to progressing along the myogenic lineage (Conboy and Rando, 2002).
Constitutively active Notchl signaling causes up-regulation of Pax3 and down-regulation
of Myf5, M~yoD, and desmin and a reduction in Pax7 expression. In response to injury, activated
Notch expression exhibits a highly localized expression at the site of injury, but not at more
distal regions along the fiber. This suggests that it might play a role in targeting or recruiting
specific factors to the injury site to promote regeneration (Conboy and Rando, 2002). The
efficiency of tissue regeneration decreases in response to aging and it was proposed that this is
due to age-related changes in satellite cell activity. Further analyses demonstrated insufficient
up-regulation of Delta, resulting in diminished Notch activation in aged, regenerating muscle.
When Notch was inhibited in young muscle, impaired regeneration was observed. Conversely,
when Notch was introduced into old muscle the regenerative potential was restored (Conboy et
To determine if systemic factors play a role in aged progenitor cells from specific tissues,
parabiotic pairings between young and old mice were studied (Conboy et al., 2005).
Interestingly, the exposure of satellite cells in old mice to young mouse serum resulted in
enhanced expression of Delta, increased Notch activation, and enhanced proliferation in vitro,
supporting the notion that there are systemic factors that change with age that impact progenitor
cell activity (Conboy et al., 2005).
Growth Factor Effects on Satellite Cells
There are multiple growth factors that regulate satellite cell activation and growth.
When myofibers are injured, the release of cytokines is stimulated, and growth factors, such as
HGF (Allen et al., 1995; Tatsumi et al., 1998) and FGFs, cause an activation and expansion of
satellite cells so that they will re-enter the cell cycle and rapidly proliferate, although activation
of satellite cells has been found to be delayed over time in response to aging (Sheehan and Allen,
1999; Clarke et al., 1993; Johnson and Allen, 1995; Cornelison et al., 2001, 2004).
Transforming growth factor beta (TGFP) inhibits cell proliferation to some extent, although its
more significant effect is in its inhibition of differentiation and fusion (Florini et al., 1986;
Greene and Allen, 1991; Rao and Kohtz, 1995; Stewart et al., 2003; Allegra et al., 2004).
MyoD, Myf5, and Pax7 are markers of the myoblasts at this stage (Davis et al., 1987; Wright et
al., 1989; Braun et al., 1989; Rhodes and Konieczny, 1989; Seale et al., 2004). Factors such as
TGFP, HGF, and FGF inhibit myoblasts from undergoing differentiation and formation of
myofibers (Gospodarowicz et la., 1976; Florini et al., 1986; Olsen et al., 1986; Massague et al.,
1986; Miller et al., 2000).
Skeletal myoblasts serve as a good model to study intracellular signaling cascades as they
undergo morphological changes characteristic of various cellular processes. Previous reports
demonstrated the significance of the Raf kinase signaling axis (Bennett and Tonks, 1997;
Coolican et al., 1997; Dorman and Johnson, 1999; Samuel et al., 1999; Dorman and Johnson,
2000; Winter and Arnold, 2000; DeChant et al., 2002). Elevated Raf levels are implicated in
causing repression of myoblast differentiation and low-levels enhancing differentiation. These
findings highlight the importance of time and duration of Raf signal transmission (DeChant et
al., 2002). The molecular basis for inhibition of differentiation may involve direct modification
of E-proteins and/or induction of TGFP-like gene expression. Additionally, inhibition of
myogenic differentiation has been demonstrated when cells were treated with various growth
factors such as FGFs (1, 2, 4, 6, and 9), TGFP, and high concentrations of serum (Allen and
Boxhorn, 1989; Sheehan and Allen, 1999).
Insulin-like Growth Factor I (IGF-I)
IGF-I promotes cell recruitment to injured muscle by coordinating a regenerative response
to muscle injury. Treatment of isolated satellite cells with IGF-I increases their proliferation
rates in response to the JAK/STAT pathway. This growth factor induces both proliferation and
subsequent differentiation of satellite cells via the type I receptor (Allen and Boxhorn, 1989;
Adams and McCue, 1998; Kamanga-Sollo et al., 2004). Activation of myoblast proliferation is
mediated through the MAPK pathway and induction of differentiation signals through the
phosphatidylinositol-3 -kinase (PI3K) pathway. In overloaded skeletal muscles, IGF-I peptide
levels increase (Adams and Haddad, 1996). In vivo evidence demonstrates that when the tibialis
anterior (TA) muscle of a rat is infused with IGF-I, there are measurable increases in muscle
protein, muscle DNA content, and absolute weight of the treated muscle (Adams and McCue,
1998). Interestingly, IGF-l is up-regulated in regenerating muscles and in aged rats and
application of IGF-I rescued approximately 46% of lost muscle mass and increased the
proliferation potential of satellite cells from atrophied gastrocnemius muscle (Chakravarthy et
al., 2000). In mdx~ mice (dystrophin null), high levels of muscle-specific IGF expression resulted
in an increase of approximately 40% in muscle mass. A subsequent increase in force generation
also was observed along with an elevation of signaling pathways associated with muscle
regeneration and protection against apoptosis (Barton et al., 2002). IGF-I serves as a strong
mitogen of satellite cells even in the presence of a strong growth inhibitor, TGFP, yet the
presence of TGFP is able to inhibit IGF-I-mediated satellite cell differentiation (Allen and
Boxhorn, 1989). Additionally, IGF-I increase the magnitude of the proliferative response
elicited by FGF2 and stimulate differentiation when treated alone (Greene and Allen, 1991).
Fibroblast Growth Factors (FGFs)
In muscle tissue, there are several FGFs expressed and released in response to injury
(Anderson et al., 1995). FGF2 and FGF6 are potent enhancers of muscle precursor cell (MPC)
expansion and satellite cells by increasing proliferation and promoting muscle regeneration
(Johnson and Allen, 1993; Lefaucheur and Sebille, 1995; Floss et al., 1997; Sheehan and Allen,
1999; Yablonka-Reuveni et al., 1999). FGF receptors (FGFR) 1, 2, 3, and 4 are expressed in
proliferating rat satellite cells with FGFR1 and 4 being the most prominent (Sheehan and Allen,
1999). FGF2 is a heparin-binding growth factor that increases satellite cell proliferation and
PCNA expression (Johnson and Allen, 1993). Treatment with FGF2 elicits a greater mitogenic
response than IGF-I or TGFPl and is as potent as HGF in stimulating satellite cell proliferation
(Sheehan and Allen, 1999). Combinations of FGF2 and HGF are additive with regard to
proliferation, and after injury, disrupted myofibers express FGF2, particularly in regions of
hyper-contraction (Anderson et al., 1995).
FGF2 is present in newly formed myotubes and inj section of FGF2 into the TA muscle of
male mdve mice during the first round of spontaneous necrosis results in enhanced satellite cell
proliferation. This is due to increasing the number of satellite cells that enter the cell cycle but
inhibits the differentiation of satellite cells (Lefaucheur and Sebille, 1995). The FGF6 knockout
in certain genetic backgrounds exhibits a reduced regenerative capacity after crush injury and
when they are interbred with mdx~ mice, a severe dystrophic phenotype is observed. FGF6 is up-
regulated in response to skeletal muscle injuries and helps completely restore experimentally
damaged skeletal muscle (Floss et al., 1997). Higher levels of FGF2 and FGFR1 are expressed
in the pectoralis major of turkey. Faster proliferating satellite cells from pectoralis major of
turkey express higher levels of FGF2 and FGFR1 by comparison to slower proliferating cells.
They also show a greater mitogenic response to FGF2, suggesting that FGFs may play a role in
proliferative rates of muscle cells (McFarland et al., 2003).
Hepatocyte Growth Factor (HGF)
HGF, also called Scatter Factor (SF), is a critical activator of satellite cells found initially
in crushed muscle extract (CME) (Tatsumi and Allen, 2004; Tatsumi et al., 2002). Direct
inj section of HGF into muscle results in activation of quiescent satellite cells, even in the absence
of trauma, and when an anti-HGF antibody is incubated with CME, the activation capacity of the
CME is abolished (Tatsumi et al., 1998; Tatsumi et al., 2002). This growth factor binds to the c-
Met receptor and signals through the PI3K pathway to promote cell survival and MAPK
pathways to stimulate the mitogenic effect (Allen et al., 1995). Placenta, liver, and muscle all
express HGF (Brand-Saberi et al., 1996). In addition, HGF is present in basal lamina of skeletal
muscle fibers, which provides a reservoir of HGF within skeletal muscle (Tatsumi and Allen,
2004). The HGF (-/-) mouse is embryonic lethal, and without HGF signaling, skeletal muscle
cells cannot migrate from the somite during embryogenesis (Schmidt et al., 1995; Bladt et al.
Transforming Growth Factor Beta (TGFP) Superfamily
The TGFP superfamily is involved in cellular proliferation, differentiation, migration, and
apoptosis. TGFP is one of the most potent negative regulators of proliferation and differentiation
of satellite cells (Florini et al., 1986; Greene and Allen, 1991; Rao and Koht, 1995; Stewart et al.,
2003; Allegra et al., 2004). Inhibition of myoblast fusion is dose-dependent and reversible
(Florini et al., 1986; Stewart et al., 2003).
The TGFP superfamily can be divided into three groups: the TGFPs, the activins/inhibins,
and the bone morphogenic proteins (BMPs). All three groups of growth factors signal through
serine/threonine kinase receptors (Massague et al., 1994). In the presence of growth factors,
ligands bind to a Type II receptor dimer located on the plasma membrane, which causes auto-
phosphorylation of the Type II dimer, recruitment of a Type I receptor dimer, and subsequent
phosphorylation of this dimer (Wrana et al., 1992; Attisano et al., 1993; Ebner et al., 1993;
Wieser et al., 1993; Wrana et al., 1994). This phosphorylation event recruits the receptor-
regulated Smads or R-Smads, which then undergo phosphorylation (Aoki et al., 2001). The R-
smads form a complex with the common-partner Smads (Co-Smads), or Smad4. The R-
Smad/Co-Smad complex translocates to the nucleus and binds to the DNA altering transcription
of target genes (Derynck et al., 1996; Liu et al., 1996; Meersseman et al. 1997; Nakao et al.,
1997). Additional co-activators and co-repressors lend regulation to the system (Wotton et al.,
Specificity of TGFP signaling is mediated by the different types of Type II and Type I
receptors present in the target cell and ligand affinities for these receptors (Wrana et al., 1992;
Attisano et al., 1993; Ebner et al., 1993; Wieser et al., 1993; Wrana et al., 1994). Additionally,
R-Smads (Smads 1, 2, 3, 5, and 8) are ligand-specific. Smads are cytoplasmic when inactive and
transfer to the nucleus upon phosphorylation (Derynck et al., 1996; Liu et al., 1996; Meersseman
et al. 1997; Nakao et al., 1997). I-Smads or inhibitory Smads (Smad6 and Smad7) bind to the
receptor and prevent phosphorylation and signaling activities (Imamura et al., 1997; Nakao et al.,
1997; Horiki et al., 2004). Negative regulation of TGFP signaling is achieved by follistatin,
which binds activin and prevents receptor docking (Nakamura et al., 1990). BMPs (2, 4, and 7)
form a trimeric complex between ligand, receptor, and follistatin to inhibit BMP actions (lemura
et al., 1998). Table 1-1 summarizes the receptor and R-Smad specificity for TGFP superfamily
members (ten Dijke et al., 1994a; ten Dijke et al., 1994b; Koening et al., 1994; Macias-Silva et
al., 1998; Yamashita et al., 1995; Rosenzweig et al., 1995; Liu et al., 1995; Nohno et al, 1995).
Myostatin (MSTN), also called growth and differentiation factor 8 (GDF8) is a member of
the TGFP superfamily and is responsible for maintaining muscle size. Mouse devoid of MSTN
possesses a larger muscle mass characterized by hypertrophic fibers. M~STNis expressed in
multiple tissues with the greatest mRNA levels found in muscle (Grobet al., 1997; Kambadur et
al., 1997; McPherron and Lee, 1997). MSTN circulates in the blood in an inactive form until the
pro-domain is cleaved away by a furin enzyme (McPherron et al., 1997; Lee and McPherron,
2001; Zimmers et al., 2002). Treatment with MSTN inhibits proliferation of muscle precursor
cells. Fluorescence Activated Cell Sorting (FACS) analysis determined that MSTN prevents
progression of myoblasts from G1 to S-phase transition of the cell cycle. Sub sequently,
upregulation of p21 and a decrease of Cdk2 protein and activity is observed, which results in an
accumulation of hypophorylated Rb and arrest in G1 (Thomas et al., 2000). This same cell cycle
arrest is observed in MSTN treated satellite cells. MSTN is thought to be required for
maintaining satellite cells in their quiescent state, until inhibited by a stimulus such as injury
(McCroskery et al., 2003). In MSTN-deficient satellite cells, a higher number of satellite cells
are activated compared to wildtype counterparts and addition of MSTN to myofiber explant
cultures inhibits satellite cell activation (McCroskery et al. 2003).
In theM2ST1V knockouts, hypertrophy and hyperplasia are observed (McPherron et al.,
1997). These animals exhibit an increased number of satellite cells versus wild type
counterparts, and the satellite cells present in the knockout animals have an increased
proliferation rate (McCroskery et al., 2003). Gene ablation also results in enlarged hearts of the
mutant animals. Myostatin signals through the Activin Receptor II and transgenic animals
carrying a dominant negative receptor have a three-fold increase in muscle mass compared with
wild type animals (Lee and McPherron, 2001).
Bone Morphogenic Proteins (BMPs)
BMPs are the largest group of family members in the TGFP superfamily (Reddi and
Huggins, 1972; Wozney et al., 1988). As the name implies, BMPs induce bone or cartilage
formation ectopically and initiate osteoblast differentiation (Urist, 1965; Gitelman et al., 1995).
BMPs are 30-3 8kDa homodimers that are synthesized as prepropeptides of approximately
400-525 amino acids. BMPs inhibit the myogenic differentiation of C2C12 cells, and convert
their differentiation pathway into that of osteoblast lineage cells (Katagiri e al., 1994).
Additionally, BMPs play a role in somite development by abrogating premature initiation of
myogenesis in the presomitic mesoderm (Pourquie et al., 1996). A BMP inhibitory signal is
believed to prevent the premature expression ofhyoD before somites are formed (Linker et al.,
2003). In pre-myogenic cells, BMP2, 4, and 7 have dose-dependent effects with low
concentrations maintaining a Pax3-expressing proliferative population and delaying
differentiation. Conversely, high concentrations of these BMPs prevent muscle development
(Amthor et al., 1998). In vivo, BMP2 expression in skeletal-derived cells prevents myogenic
differentiation and promotes osteogenic differentiation (Musgrave et al. 2001).
BMPs mediate non-osteogenic processes. During the developmental and differentiation
processes of the embryo, BMPs regulate epithelial-mesenchymal interactions, cell fate
specification, dorsoventral patterning, apoptosis, and the secretion of extracellular matrix
components (Vainio et al., 1993; Amthor et al., 1998; Weaver et al., 1999; Angerer et al., 2000;
Higuchi et al., 2002; Tiso et al., 2002). More specifically, in vertebrates, BMP2, 4, and 7 are
found to direct the development of neural crest cells into their ultimate phenotypes (Wilson and
Hemmati-Brivanlou, 1995; Miya et al., 1997).
In terms of signaling, there are Type I and Type II receptors specific for BMPs and
Smads1, 5, and 8 are the downstream molecules phosphorylated by the ligand-receptor
complexes (see Table 1-1). At the transcriptional level, BMPs regulate target genes such as,
Runx2, osteopontin, osteonectin, and osteocalcin specific for osteogenesis (Ahrens et al., 1993).
Smad6 is an I-Smad. It functions by binding to type I BMP receptors thus preventing the
activation of BMP Smads 1/5/8 (Imamura et al., 1997). Smad6 overexpression in chondrocytes
results in delayed differentiation and maturation (Horiki et al., 2004).
Noggin and other cystine knot-containing BMP antagonists associate with BMPs to block
their signaling. Noggin overexpression in osteoblasts, results in osteoporosis in mice (Devlin et
al., 2003; Wu et al., 2003). Noggin is a secreted peptide, which is expressed in condensing
cartilage and immature chondrocytes. Ablation of Noggin is embryonic lethal at 18.5 days post
coitum (dpc) and results in severe hyperplasia of the cartilage with multiple j oint fusions, severe
defects in somitogenesis, and multiple skeletal defects (Brunet et al., 1998; McMahon et al.,
1998). Noggin is found to have different affinities for different BMP family members making its
regulation of BMP-mediated processes more complex. For example, in the Noggin null
transgenic, different bones have varying responses to the de-repression of BMP signaling
(Zimmerman et al., 1996; Chang and Hemmati-Brivanlou, 1999). Depending on location and/or
embryonic origin of the bones, one observes inhibition, delay or acceleration of ossification
processes (Tylzanowski et al., 2006).
Another BMP negative regulator is Tob, which suppresses the activity of receptor-
regulated Smads (1/5/8) (Bradbury et al., 1991; Fletcher et al. 1991; Yoshida et al., 2000). Tob
null mice exhibit enhanced BMP signaling and increased bone formation (Yoshida et al., 2000).
Negative BMP regulation also is achieved by mechanisms that target elements of BMP signaling
for degradation. The Hect domain E3 ubiquitin ligase, Smurfl, targets Smad1 and 5 for
degradation, in addition to interacting with and mediating the degradation ofbone-specific
transcription factors such as Runx2 (Zhao et al., 2003). Smurfl can also target type I BMP
receptors for degradation by interacting with Smad6. They form a complex that can be exported
from the nucleus where it interacts with the BMP receptors to promote degradation (Murakami et
al., 2003). Furthermore, when Smurfl is overexpressed in osteoblasts, postnatal bone formation
is inhibited (Zhao et al., 2004).
While BMPs are capable of redirecting muscle mesenchyme cells to differentiate into bone
tissue and stimulate bone formation in vivo (Urist, 1965), previous reports have demonstrated
that mutations of BMP family members results in disruption of skeletal development (Kingsley
et al., 1992; Storm and Kingsley, 1999). Moreover, BMPs have opposing activities that are
concentration-dependent. In the presence of low BMP levels, myogenic precursor cells are
maintained in a proliferative state in developing limb bud, while high BMP levels induce cell
death. Thus, BMPs can both stimulate and restrict muscle growth suggesting that a
concentration gradient of BMPs is needed for the correct determination and maintenance of the
myogenic program (Amthor et al., 1998; Amthor et al., 2002). While BMPs signal through the
Smads, a multitude of different transcription factors are recruited, accounting for the diversity of
functional BMP responses. One set of genes involved in skeletal development are the Hox
family of transcription factors. In analyses of gain- or loss-of-function and naturally occurring
mutations of Hox genes, BMPs play a central role in embryonic skeletal patterning (Manley and
Capecchi, 1997; Yueh et al., 1998). Furthermore, several of the Hox genes interact with Smads
and there is evidence that Hox gene expression may be regulated by BMPs further suggesting
that BMPs are a critical mediator of skeletal myogenesis (Ladher et al., 1996; Tang et al. 1998;
Liu et al., 2004).
BM~P6, a member of the BMP subfamily, was originally isolated from phage plaques of a
hgl0-based cDNA library derived from 8.5-dpc murine embryos. The murine library was
hybridized with a 32P-labeled partial Xenopus laevis Vg-1 cDNA under low-stringency
conditions (Derynck et al., 1988). Northern blot analyses of murine tissues illustrated that BM~P6
was present in muscle (Lyons et al., 1989). The human and bovine homologs were isolated from
bone and designated BM~P6 (Celeste et al., 1990). Expression predominates in mature
chondrocytes during endochondral ossification and BMP6 treatment stimulates chondrogenic
and osteogenic phenotypes in vitro and induction of cartilage and bone formation in vivo
(Gitelman et al., 1994; Gitelman et al., 1995; Yamaguchi et al., 1996). Adult muscles express
BMP6 (Lyons et al., 1989) and BMP6 expression increases in Raf-arrested myoblasts (Wang et
al., 2004). In keratinocytes, treatment with BMP6 significantly decreases DNA synthesis
(D'Souza et al., 2001) and triggers differentiation programs (Tennenbaum et al., 1996). In these
differentiated keratinocyte cultures, E2F5 protein levels are significantly increased (D' Souza et
al., 2001). Treatment of keratinocytes with TGFPl causes reversible cell cycle arrest without
activating the differentiation program (Pierce et al., 1998; Dicker et al., 2000).
BMP Function in vivo: Lessons Learned from Knockouts
A compilation of the various BMPs and receptors targeted knockouts and their resulting
phenotypes are presented in Table 1-2. BM~P5 null mice are called short ear mice, because these
animals have reduced ear size in comparison to wild type animals, in addition to exhibiting
reduced vertebral processes, and a reduced number of ribs and sesamoid bones (Green, 1968;
Kingsley et al., 1992; King et al., 1994). Mice with a homozygous BM~P7 deletion are found to
die at birth due to renal failure because of hypoplastic/dysplastic kidneys (Dudley et al., 1995;
Luo et al., 1995; Wawersik et al., 1999). BM~P6 null mice are viable and fertile and exhibit no
maj or defects, except for a delay in ossification of the developing sternum (Solloway et al.,
1998). It is believed that BM~P2 may be functionally compensating for BM~P6 ablation since
BM~P2 and BM~P6 are required for some overlapping or redundant functions (Solloway et al.,
Utilization of BMPs in a clinical setting was demonstrated in various therapeutic
interventions such as, bone defects, non-union fractures, spinal fusion, osteoporosis, and root
canal surgery. Multiple studies utilizing BMP2 demonstrate its ability to promote healing of
severe long bone defects in rats, rabbits, dogs, sheep and non-human primates (Murakami et al.,
2002). Adenoviral administration of BMP2 mixed with a bioresorbable polymer and
mesenchymal stems cells was able to repair bone defects (Chang et al., 2003). RhBMP2
administered systemically increased mesenchymal stem cell activity and reversed age-related
bone loss due to ovariectomies in different mouse models (Turgeman et al., 2002) demonstrating
a potential application of BMPs in osteoporosis treatment. Interestingly, rhBMP2 has been used
as a complete bone graft substitute in spinal fusion surgery and in some cases BMP2 has been
more efficacious in promoting successful bone fusion as compared to autogenous bone grafts.
BMP2 was used in other fusion applications such as intervertebral and lumbar posterolaterial
fusions (Sandhu, 2003). BMPs can be used in dental applications since BMP2 induces bone
formation around dental implants used in periodontal reconstruction. It has been suggested that
BMPs could serve as an alternative to root canal surgeries (Schwartz et al., 1998; Cochran et al.,
Current BMP6 Studies
Recently, BMP6 was able to induce matrix synthesis and induce differentiation of bovine
ligaments fibroblasts providing another potential source of chondrocytes for tissue repair
(Bobacz et al., 2006). Mesenchymal stem cells (MSCs) treated with BMP6 in combination with
parathyroid hormone (PTH) and vitamin D(3) increased osteocalcin production. Osteocalcin is
used as a marker for bone formation and in these MSCs, BMP6 enhanced calcium formation
(Sammons et al., 2004). TGFP and BMPs may act in an antagonistic manner towards each other.
TGFP inhibits chondrocyte maturation relatively early in differentiation by down-regulating
bmp6, ihh (Indian hedgehog), and colX(collagen type X), which are genes specifically expressed
by hypertrophic chondrocytes (Vortkamp et al., 1996; Ferguson et al., 2004).
BMPs are implicated in the development of cancers of the GI tract, breast, and prostate
(Howe et al., 2001; Pouliot and Labrie, 2002; Brubaker et al., 2004). Currently, much focus is
being placed on TGFP superfamily members as targets for cancer therapy due to their ability to
suppress tumor progression by inhibiting growth of neoplastic tissues. BMP6 in the early stages
of carcinogenesis inhibits benign and malignant skin tumor formation (Wach et al., 2001).
BMP6 exerts an anti-proliferative and pro-apoptotic effects in multiple myeloma (Kawamura et
al., 2000; Hj ertner et al., 2001; Kersten et al., 2005).
Ras/Raf in Skeletal Muscle
A key regulator of signal transduction pathways controlling cell proliferation,
differentiation, and oncogenesis is Ras, a monomeric guanine nucleotide-binding protein with
intrinsic GTPase activity (Marshall, 1996). Ras is a molecular switch that cycle between a GTP-
bound active and a GDP-bound inactive state (McCormick, 1993). Oncogenic H-Ras inhibits the
differentiation of muscle cells independent of their continued proliferation (Olson et al., 1987).
Ras inhibits myogenesis by disrupting MRF function, resulting in the inhibition of differentiation
of muscle cells independent of proliferation (Ramocki et al., 1998). Ras proteins are localized at
the cytoplasmic face of the plasma membrane and activated by a large number of extracellular
stimuli, such as growth factors and hormones (McCormick, 1993; Bar-Sagi, 2001). A family of
Ras effector molecules specifically binds to and is activated by Ras-GTP, which then carries out
the downstream functions of activated Ras (Marshall, 1996). Some effectors of Ras are the Ral
GTPase signaling pathway involved in the G1 to S progression (Ramocki et al., 1998) and the
PI3K/AKT pathway involved in cell survival (Murphy et al., 2002). All of the major Ras
effectors have been tested for a role in the Ras-mediated inhibition of skeletal myogenesis but
none of them are able to duplicate the effects of oncogenic Ras (Ramocki et al., 1997; Weyman
et al., 1997).
The Raf Family
The Raf family consists of three serine/threonine specific kinases and are the best
characterized effectors of Ras. Members include A-Raf, B-Raf, and c-Raf, also called Raf-1
(Marshall, 1996). There are three conserved regions (CR1-3) within Rafs consisting of two N-
terminal regulatory domains (CR1 and CR2) and a C-terminal catalytic kinase domain (CR3).
Rafs lead to the activation of the extracellular signal-related kinase (ERK) pathway, which
mediates cellular proliferation and differentiation (Marshall, 1996).
Inactive Raf is cytosolic and translocates to the plasma membrane following activation by
Ras (Chong and Guan, 2003). The activation of Raf is important in mediating growth factor
gene expression and a maj or function of Raf protein kinase is to phosphorylate MEK 1 and
MEK2. Subsequently, ERK1 and ERK2 are phosphorylated on tyrosine and threonine residues.
These activated ERKs then either phosphorylate numerous cytoplasmic targets or migrate to the
nucleus to activate transcription factors such as c-fos and Elk1 (Huang et al, 1993). The
Raf/MEK/ERK signaling cascade is required for cell cycle progression and overexpression
causes cell transformation (Kolch et al., 1991; Cowley et al., 1994; Mansour et al., 1994). The
sustained activation of this pathway is implicated in differentiation with prolonged ERK
activated leading to differentiation of PC12 cells (Qui and Green, 1992). Raf-activated MEK-
ERK cascades are likely participants in apoptosis because potent ERK activation can protect
cells from apoptosis (Le Gall et al., 2000).
Signaling through the Raf/MEK/ERK controls several aspects of myogenesis.
Overexpression of a constitutively active Raf and MEK results in reduced muscle gene
transcription and these myoblasts are unable to form myocytes (Dorman and Johnson, 1999;
Dorman and Johnson, 2000; Samuel et al., 1999; Winter and Arnold, 2000). Down-regulation of
ERK activity by overexpression of MAPK phosphatase I (MKP-I), or through the use of
synthetic inhibitors, upregulates muscle specific gene transcription (Bennett and Tonks, 1997;
Dorman and Johnson, 1999). By contrast, other reports have found no effect of ERK activity on
muscle gene expression, myoblast fusion, and myoblast differentiation (Weyman et al., 1997;
Jones et al., 2001). Myoblasts expressing low levels of Raf differentiate more efficiently than
control cells again suggesting that ERKl/2 promotes myogenesis (Wang et al., 2004). In
differentiating C2 myoblasts, ERKl/2 activity increases MyoD expression and transcriptional
activity (Gredinger et al., 1998). Activated ERK1 is associated with myoblast proliferation and
activated ERK2 is associated with myoblast differentiation (Sarbassov et al., 1997). ERK1 and
ERK2 are both ubiquitously expressed in most tissues and similar in structure. ERK1 null mice
are viable, fertile, and of normal size (Nekrasora et al., 2005), whereas, ERK2 null mice are
embryonic lethal (Yao et al., 2003).
The maj or function of the E2F factors is cell cycle progression, in addition to playing roles
in metabolic activities such as proliferation, differentiation, and apoptosis (Fujita et al., 2002).
Genes expressed at the G1/S transition contain E2F-binding sites within their promoters
(DeGregori et al., 1995). E2F members are conserved throughout evolution from invertebrates
to mammals and there are currently nine family members, which demonstrate tissue-specific
activities (Dynlacht et al., 1994; D' Souza et al., 2001). Two major groups, activators and
repressors, comprise the E2F family. E2F l, E2F2, and E2F3a are involved in positive cell cycle
control and S-phase entry of quiescent cells. These members preferentially bind pRb with
expression peaking in late G1 and association with E2F-regulated promoters during the G1-S
transition. Ectopic expression of these members results in S-phase induction in serum-starved
cells (DeGregori et al., 1997; Lukas et al., 1996; Leone et al., 1998; Humbert et al., 2000).
E2F3b, E2F4 and E2F5 are classified as repressor proteins, due to their negative control of cell
cycle progression. E2F4 and E2F5 predominately bind pl07 and pl30 pocket proteins
(DeGregori et al., 1997; Leone et al., 2000). E2F6, E2F7, and E2F8 also are classified as
repressors but through a mechanism independent of pocket protein interaction because they lack
the Rb-binding sequence at the C-terminus (Campanero et al., 2000). Members of the E2F
family demonstrate differential mRNA expression throughout the cell cycle. For example, E2F 1
and E2F2 mRNA increase in late G1, and peak at the G1/S transition, although E2F2 mRNA
expresses at a lower level than E2Fl1. During quiescence, E2F3 and E2F5 mRNA are expressed,
with E2F 1 barely present. In early-to-mid G1, E2F3 and E2F5 mRNA levels rise (Sardet et al.,
1995; Pierce et al., 1998).
Members of the E2F family are grouped based on their homologous DNA binding domain
(DBD). All of the E2Fs, except E2F7 and 8, and both DP proteins have a conserved DBD and a
dimerization domain. E2F4 and E2F5 demonstrate a 72% amino acid identity to each other and
a 35% amino acid identity to E2Fl1-3 (Vaishnav et al., 1998). E2F4 and E2F5 make up a
subclass of the E2F family since their N-terminus lacks the cyclin A-binding domain found in
the other members. The repressive E2Fs contain the Rb-binding sequence and have nuclear
export sequences instead of nuclear localization sequences in the N-terminus (Helin et al., 1993;
Sardet et al., 1995).
E2F Signaling Mechanism and Involvement of Pocket Proteins
The E2F proteins signal via the formation of active heterodimer complexes with DP
proteins (Helin et al., 1993). E2Fs can dimerize with either DP1 or DP2, except for E2F7 which
binds DNA in a DP-independent manner (Ormondroyd et al., 1995). These E2F complexes
promote gene expression required for G1 progression and DNA replication (Chen et al., 2004).
The mammalian Rb family of proteins (pRb, pl07, and pl30) is also referred to as the
pocket proteins. The pocket protein motif allows for interactions with cellular proteins that
possess a LXCXE peptide motif (Lee et al., 1998). The LXCXE motif is present in E2Fs, the D-
type cyclins (Gill et al., 1998), and HDACs. The pocket domain is separated into two
functionally conserved regions, the A and B pockets. The spacer region is specific to each
pocket protein and contains binding sites for cyclin/CDK complexes in pl07 and pl30 but not in
pRb. pl30 and pl07 contain approximately 50% amino-acid identity and 30-35% identity to
pRb (Ewen et al., 1991; Hannon et al., 1993). Overexpression of cDNAs coding for pocket
proteins induces growth arrest at G1 (Classon et al., 2000).
Rb was the first tumor suppressor gene cloned and loss of Rb function serves as a hallmark
of oncogenic progression. pRb is a major G1 checkpoint and inhibits S-phase entry. pRb
promotes terminal differentiation, cell cycle exit and tissue specific gene expression (Dunaief et
al., 1994; DeCaprio et al., 1989; Dick et al., 2000).
Cell Cycle Repression
Pocket protein dephosphorylation occurs from anaphase to G1 or in response to growth
inhibitory signals (Ludlow et al., 1993). Phosphorylation events can also lead to permanent
inactivation of the pocket protein and possibly target it for degradation (Ma et al., 2003). The
type of functional effect that pocket proteins have on a cell is dependent on what type of co-
accessory proteins interact with the target gene (Stevaux et al., 2005). To repress gene
transcription needed for the G1 to S-phase transition, the pRb binds directly to the transactivation
domain of E2F. pRb also recruits chromatin remodeling factors such as histone deacetylase I
(HDAC1), SUV39H1, hBRM, and BRGl. These factors act on the surrounding nucleosome
structure to remodel it, and promote histone acetylation/deacetylation and methylation events
(Shao and Robbins, 1995). HDAC1 is recruited to E2F complexes by pRb to function in
repressing cyclin E gene expression (Magnaghi-Jaulin et al., 1998). SUV38H1 is a
methyltransferase that methylates K9 of histone H3 and cooperates with pRb in the repression of
E2F-responding promoters (Rea et al., 2000). hBRM and BRG1 are the mammalian homologs
of SNF2/SWI2, which are yeast chromatin remodeling complexes and they associate with pRb
(Strober et al., 1996; Kang et al., 2004).
E2Fs and Pocket Proteins in Skeletal Muscle
pRb is required for muscle differentiation and for transcription of myogenic bHLH factors.
Functional Rb is required for the activity of MyoD-mediated transcriptional activation of
myogenic genes (Gu et al., 1993). In models of inactive Rb, myoblast differentiation was
inhibited in vitro and terminally differentiated myotube nuclei were able to reenter the cell cycle.
In the absence of pRb, skeletal muscle cells exhibited ectopic DNA synthesis and/or apoptosis,
and pRb null mice exhibit severe defects in skeletal muscle (Schneider et al., 1994; Novitch et al,
1996; Zacksenhaus et al., 1996; Novitch et al., 1999; de Bruin et al., 2003; Wu et al., 2003).
Interestingly, E2Fs in complex with pl30 accumulate in cells such as, myoblasts, and
melanocytes undergoing terminal differentiation (Shin et al., 1995). In rat L6 myoblasts, pl07 is
normally involved in regulation of E2F proteins during cell cycle progression. Exponentially
growing L6 myoblasts demonstrate complexes of E2F and bound pl07 throughout the cell cycle.
During the differentiation of L6 cells, pl07 levels are reduced and pl30 levels are greatly
increased, suggesting that pl30 is a differentiation-specific regulator of E2F activity (Kiess et al.,
In summary, skeletal myogenesis is regulated by a multitude of growth factors and
signaling events. Since most studies have focused on the role of BMPs in embryonic
development, little is known about its role in postnatal muscle. E2Fs are critical mediators of
cell cycle progression. Satellite cells are requisite for growth and repair of skeletal muscle. It is
not well understood how satellite cells are activated to enter the cell cycle nor how they exit the
cell cycle to return to quiescence. E2F5 appears to have some role in Raf-arrested myoblasts and
BMP6 expression is high in these Raf-arrested cells. Therefore, these data will further assess the
role of BMP6 and E2F5 in regulation of skeletal myogenesis and/or satellite cell biology.
Table 1-1. Receptor and R-Smad specificity for TGFP superfamily members
Ligand Type II Receptor Type I Receptor R-Smad
Activin Act-R-II/IIB ALK4 Smad2,3
Table 1-2. TGFS
Bmp 11 (Gdfl1)
superfamily transgenic knockouts*
Delayed primitive streak, small allantois,
lack of amnion, heart defects, decreased
number of PGCs.
Increase bone density
Lack of allantois and PGCs, posterior
truncation, heart defects, and lack of
optic vesicle; heterozygotes-cystic
kidney, craniofacial malformations,
Short ear phenotype including defects in
skeleton, lung, and kidney
Delayed sternum ossification
Skeletal defects, kidney agenesis, eye
Defects in spermatogenesis and
Defects in PGC formation, testis cord
formation, and spermatogenesis
Defects in A-P patterning of axial
Hydrocephalic abnormalities growth
defects in seminal vesicle
Subfertile due to defects in oogenesis
Defects in left/right asymmetry
Brachypodism shortenedd skeleton in
limbs and reduced number of digit
Skeletal muscle hypertrophy
Sterile due to defects in oogenesis
No gastrulation (lack of primitive
streak): anterior neural pattern defects:
placenta defects (increased number of
trophoblast giant cells)
Abnormal left-right axis (left isomerism)
Extended streak, excessive mesoderm,
Lack whiskers and low incisors and cleft
Defects in eyelid development and
reproductive tract present in males)
Hyperactive immunity and defects in
Zhang and Bradley, 1996: Ying and
Bahamonde and Lyons, 2001; Daluiski
et al., 2001
Winnier et al., 1995: Dunn et al., 1997:
Funrta and Hogan, 1998: Lawson et al.,
1999; Ying et al., 2000
Green, 1968: Kingsley et al., 1992:
King et al., 1994
Solloway et al., 1998
Dudley et al., 1995: Luo et al., 1995:
Wawersik et al., 1999
Zhao et al., 1998
Zhao et al., 1996: Ying et al., 2000;
Yao et al., 2002
McPherron et al., 1999
Lee et al., 1998: Settle et al., 2001
Yan et al., 2001
Rankin et al., 2000
Storm et al., 1994: Storm and Kingsley,
McPherron et al., 1997
Dong et al., 1996
Conlon et al., 1994: Varlet et al., 1997:
Lowe et al., 2001; Ma et al., 2001;
Zhou et al., 1993
Meno et al., 1998
Meno et al., 1999
Matzuk et al., 1995b
Vassalli et al., 1994
Matzuk et al., 1992
Behringer et al., 1994
Shull et al., 1992: Dickson et al., 1995
Table 1-2 continued.
Perinatal lethality due to multiple defects
in heart, lung, limb, spinal column, eye,
inner ear, and urogenital system
Defects in embryonic angiogenesis
Defects in mesoderm formation as a
result of defective visceral endoderm
Defects in epiblast proliferation and no
mesoderm formation in null mutants,
impaired cardiac and limb development
in conditional mutants
Defects in epiblast differentiation and
lack ofmesoderm formation
Defects in seminal vesicle development,
female reproduction, and limb skeletal
Defects in gastrulation/lack of mesoderm
Defects in vasculogenesis and
Deficiency in reproduction due to
suppressed FSH and mild defects in
Defects in axial patterning and left-right
asymmetry (45% right isomerism)
Sanford et al., 1997
Proetzel et al., 1995
Oh et al., 2000
Gu et al., 1999; Mishina et al., 1999
Mishina et al., 1995: Ahn et al., 2001;
Gaussin et al., 2002
Gu et al., 1998
Baur et al., 2000; Yi et al., 2000; Yi et
Beppu et al., 2000
Oshima et al., 1996
Matzuk et al., 1995a
Oh and Li, 1997
*Adapted from Zhao, G.H. 2003. Consequences of Knocking out BMP signaling in the mouse.
Genesis 35:43-56. Table 1, pgs.45-46.
MATERIALS AND METHODS
Myoblast Cell Culture
C2C12 skeletal muscle satellite cells (Blau et al., 1985) were grown in Dulbecco's
modified Eagle's medium (DMEM) containing 10% fetal bovine serum (FBS), 1% v/v
penicillin/streptomycin, and 0.1% v/v gentamycin. 23A2 skeletal muscle cells (Konieczny and
Emerson, 1984) were maintained in Basal Medium Eagle (BME) supplemented with 15% FBS,
1% v/v penicillin/streptomycin, 0.1% v/v gentamycin reagent solution, 1% v/v L-glutamine.
23A2RafERDD myoblasts were derived from the parental 23A2 myoblasts and stably express a
tamoxifen-inducible chimeric Raf protein. The estrogen receptor fused to the Raf kinase domain
is unstable in the absence of the estrogen analog 4-hydroxytamoxifen (4HT). Addition of 4HT
binds to the estrogen receptor and allows for a dose-dependent increase in Raf protein expression
and kinase activity that initiates downstream ERKl/2 activation (Wang et al., 2004).
23A2RafERDD myoblasts were maintained in the same media as 23A2 cells with the addition of
10mM puromycin (Wang et al., 2004). All culture media, supplements, and sera were purchased
from Invitrogen, Carlsbad, CA. For induction of stable Raf expression, cells were washed with
phosphate buffered saline (PBS), treated with 10 Clg/ml protamine sulfate (CalBioChem, San
Diego, CA) in serum-free BME for 10 minutes. Cells were starved in serum free BME for one
hour and treated with 1CIM 4-hydroxytamoxifen (4HT; Sigma, St. Louis, MO) in 2% FBS BME.
Cells for immunofluorescence were cultured on 35 mm glass-bottomed plates (World Precision
Institute, Sarasota, FL) coated with 10% v/v BD Matrigel Matrix HC (BD Biosciences, San Jose,
Growth Factor Treatment, and BrdU Pulsing and Fixation
C2C12 cells were seeded on gelatin-coated tissue culture plates at a concentration of 1x105
cells/well in a 6-well cluster. Myoblasts were treated in the absence and presence of human
recombinant BMP6 (R&D Systems, Minneapolis, MN) at final concentration of 25 ng/ml for 48
hours in 2% FBS DMEM. When necessary, cells were pulsed with 10 C1M bromodeoxyuridine
(BrdU) during the last thirty minutes of treatment and then fixed in 70% Ethanol (EtOH) at 40C
for 30 minutes.
Plasmids and Transfections
Semi-confluent myoblasts were transfected by calcium phosphate precipitation formation.
In brief, each well received DNA precipitate containing 1 Clg of luciferase reporter, 0.5 Clg of
activator plasmid, 0.5 Clg of kinase, and 50 ng of pRL-tk, a Renilla luciferase plasmid used as a
monitor of transfection efficiency. The luciferase reporter plasmids used in the study were a
multimerized BMP response element (BRE-luc) (Korchynskyi and ten Dijke, 2002), the internal
response element of the quail troponin I gene (Tnl-luc) (Johnson et al., 1996), a multimerized
TGFP responsive element (3TP-luc) (Wrana et al., 1992), an Immunoglobulin E box reporter
plasmid (C1E5-Luc) (Johnson et al, 1996), and E2F-TA-luc (Clontech, Mountain View, CA).
Activator plasmids used were CMV-E47 (Page et al., 2004), CMV-E2F5 (OriGene Technologies
Inc., Rockville, MD), and pEM-MyoD (Page et al., 2004). Cells were maintained in growth
media or differentiation-permissive medium supplemented with human recombinant BMP6, and
TGFP (R&D Systems, Minneapolis, MN) for 48 hours prior to lysis and measurement of
luciferase and P-galactosidase or Renilla luciferase activities (Promega, Madison, WI). Final
concentrations of BMP6 were 1, 10, 25, or 100 ng/ml. TGFPl was supplemented at 10 ng/ml.
BrdU Staining and BrdU Incorporation
Cells were fixed in 70% ethanol for 10 minutes at room-temperature, followed by
incubation in 2N Hydrochloric acid (HC1) for 1 hour at 370C. Cells were washed well with
Phosphate Buffered Saline (PBS) and then incubated in blocking buffer (PBS containing 5%
horse serum (HS), and 0.1% Triton X-100) for 1 hour. Antigen was detected by incubation with
Biotynylated Anti-BrdU IgG (H+L) (BA-2000) at 1:100 (Vector Laboratories, Burlingame, CA)
for 1 hour, followed by incubation with HRP-Streptavidin at 1:100 (Vector Laboratories,
Burlingame, CA) for 1 hour. Antibodies were diluted in blocking buffer (PBS containing 2%
Horse Serum, and 0. 1% Triton X-100). Visualization of BrdU staining was accomplished by
addition of one part 3,3' -Diaminobenzidine tetrahydrochloride (DAB) and five parts Nickel
Chloride (NiC1) in the presence of H202. Percentages of BrdU incorporation were calculated by
dividing the number of BrdU positive nuclei by the total number of nuclei. The averages of a
minimum of six microscopic fields at 200X per treatment are shown.
Immunofluorescent E2F5 Staining
23A2RafERDD CellS were permeabilized for twenty minutes (lX PB S, 0.01% Triton X-
100), incubated in blocking buffer for one hour, and then incubated overnight at 40C with E2F5
antibody (sc-999) (Santa Cruz Biotechnology, Santa Cruz, CA) diluted 1:50 in blocking buffer
(lX PBS, and 2% Horse Serum). Next, cells were incubated with Alexafluor 488 conjugated
anti-rabbit diluted 1:100 in blocking buffer (Molecular Probes, Carlsbad, CA) for 45 minutes at
room temperature. Secondary antibody controls also were included to demonstrate the
specificity of an antibody's expression patterns. Hoechst dye (1 Clg/ml final concentration) was
included as a nuclear stain. Representative photomicrographs at 630X under oil immersion were
captured with a Nikon 1200DMX digital camera.
RNA Isolation and Nylon Arrays
Total RNA was isolated from C2C12 myoblasts and myofibers and 23A2 myofibers with
the use of a Stratagene RNA Easy Kit. Equal amounts (3 ug) of pooled RNA were reverse
transcribed with Superscript II (BD BioScience, San Jose, CA) in the presence of biotin-UTP.
Biotinylated cDNAs were used to probe a RT2 Gene Expression Assay, a nylon mini-arrays
containing 96 genes coding for various members of the TGFP superfamily, their receptors,
signaling intermediates, and transcriptional regulators (SuperArray Bioscience Corporation,
Frederick, MD). Array membranes were pre-hybridized with 100 Cpg/ml heat-denatured salmon
sperm DNA in GEAhyb Hybridization Solution (SuperArray Bioscience Corporation, Frederick,
MD) for 2 hours at 600C. Biotinylated probes were added to arrays in GEAhyb solution to
hybridize overnight (O/N) at 600C. Following hybridization, the membranes were washed twice
in 2X SSC (saline sodium citrate), 1% SDS (sodium dodecyl sulfate) and twice in 0.1X SSC,
0.5% SDS for 15 minutes each at 600C. After washing, membranes were blocked with
GEAblocking solution Q for 40 minutes at room temperature. Alkaline phosphatase-conj ugated
streptavidin (AP) was diluted 1:8,000 with Buffer F and incubated with the membrane for 10
minutes at room temperature. Membranes were washed four times with lX Buffer F for 5
minutes each and then rinsed twice with Buffer G and visualized with CDP-Star
chemiluminescent substrate for 2-5 minutes at room temperature (SuperArray Bioscience
Corporation, Frederick, MD) and exposure to X-ray film (XAR-5, Kodak).
C2C12 myoblasts were differentiated in the presence or absence of 100 ng/ml BMP6.
After 48 hours, the cells were lysed in 4X SDS-PAGE sample buffer (250 nM Tris pH 6.8, 8%
SDS, 40% glycerol, 0.4% P-mercaptoethanol) and protein concentrations were measured (Bio-
Rad; Bradford, 1976). Equal amounts of protein were electrophorectically separated through
denaturing gels and transferred to nitrocellulose membrane. The blots were incubated with 5%
nonfat dry milk or 5% bovine serum albumin (BSA) in TBST (10mM Tris, pH 8.0, 150nM NaC1,
0.1% Tween 20) to block nonspecific binding sites. Primary antibodies were diluted in blocking
buffer, and the blots were incubated overnight at 40C with shaking. Antibodies and dilutions
included the following: anti-myogenin (F5D ascites, Developmental Hybridoma Bank,
University of Iowa, 1:5,000), anti-myosin heavy chain (MF20 hybridoma supernatant,
Developmental Hybridoma Bank, University of Iowa, IA, 1:5), anti-troponin T (R & D Systems,
Minneapolis, MN, 1:2,500), anti-desmin (D3, hybridoma supernatant, Developmental
Hybridoma Bank, University of Iowa, IA, 1:10), anti-phospho-Smadl1/Smad5/Smad8 (Cell
Signaling Technology, Beverly, MA, 1:1,000), anti-p38 (1:5,000), and anti-phospho-p38
(1:5,000) (Cell Signaling Technology, Beverly, MA, 1:2,000). After three times of washing with
TBST for 15 minutes each, the blots were reacted with the appropriate peroxidase conjugated
secondary antibody diluted 1:5,000 in blocking buffer for 60 minutes at room temperature.
Visualization of protein bands was accomplished by chemiluminescence (ECL, Amersham
Biosciences, Piscataway, NJ) and exposure to X-ray film (XAR-5, Kodak).
Alkaline Phosphatase Staining
C2C12 myoblasts were treated for 0, 1, 24 or 48 hours with 100 ng/ml BMP6 or vehicle
alone (control). Cultures were fixed with 4% paraformaldehyde and alkaline phosphatase (ALP)
activity measured by reaction with 165 Cpg/ml 5-bromo-4-chloro-3-indolyl-phosphate (BCIP) and
330 Cpg/ml Nitro Blue Tetrazolium (NBT) color development substrates (Promega, Madison, WI)
for 18 hours at 37 C. Representative photomicrographs at 100X were captured under bright field
conditions with a 1200DMX digital camera (Nikon).
p38 Inhibition Assays
Myoblasts were treated with 10 CIM SB202190 (Upstate Biotechnology, Charlottesville,
VA), an inhibitor of p3 8 for 1, 24, and 48 hours. SB202190 is a cell permeable pyridinyl
imidazole that potently inhibits p3 80/SAPK2a and p3 8P/SAPK2b. This specific p3 8 inhibitor
does not effect ERK activity or SAPK/JNK MAP kinases. After treatment, cells were Eixed with
4% paraformaldehyde and assessed for ALP activity measured by reaction with NBT + BCIP or
lysed and assayed for Tnl-Luc activity.
Myoblasts were treated with 25 CIM staurosporine (Sigma, St. Louis, MO) for 3 hours.
Cells were lysed in 4X SDS-PAGE sample buffer and equal amounts of protein were
electrophorectically separated through denaturing gels and transferred to nitrocellulose
membrane. The blots were incubated with 5% nonfat dry milk in TBST and primary antibodies
were diluted in blocking buffer. The blots were incubated overnight at 40C with shaking.
Antibodies and dilutions included the following: anti-Poly (ADP-ribose) polymerase (PARP)
(1:1000, Cell Signaling Technology, Inc., Danvers, MA) and anti-Bcl-2 (1:500, BD Biosciences,
San Jose, CA). After three times washing for 15 minutes each with TBST, the blots were reacted
with the appropriate peroxidase conjugated secondary antibody diluted 1:5,000 in blocking
buffer for 60 minutes at room temperature. Visualization of protein bands was accomplished by
chemiluminescence and exposure to X-ray film.
All data presented represents at least three independent experiments with a minimum of
two to three replicates per treatment group. All numerical data were compared to appropriate
controls and each other as indicated for each experiment and analyzed following the General
Linear Models (GLM) Procedures of the Statistical Analysis System (SAS) (SAS, 1988).
Differences between treatments were calculated using predicted differences between the least-
squares means of treatment divided by standard deviation. Classes were designated as cells,
treatment (trt), replicates (rep), and relative luciferase units (rlu). The statistical model included
rlu equals treatment and replicates (rlu = trt rep) as main effects. Data was presented as Means +
SEM. Treatments were considered significantly different when P < 0.05.
DIFFERENTIAL EXPRESSION OF TGFP SUPERFAMILY MEMBERS DURING
Our previous work demonstrated that activated Raf, a key regulator of the MEK/ERK
pathway inhibits myogenesis based on its overall signaling intensity (Ramocki et al., 1997;
Dorman and Johnson, 1999; Page et al., 2004; Wang et al., 2004). In myoblasts exhibiting
strong Raf/Extracellular-Signal Regulated Kinase (ERK) signaling, TGF/73, GDF8, and BM~P6
are up-regulated, suggesting that these TGFP family proteins may serve as autocrine inhibitors of
differentiation (Wang et al., 2004). These proteins are members of the TGFP superfamily, which
consist of the BMP, activin, and TGFP subfamilies. Skeletal myogenesis is intricately regulated
by differential gene expression of these ligands. TGFP1 and GDF8 are inhibitors of proliferation
and differentiation (Florini et al., 1986; Greene and Allen, 1991; Rao and Kohtz, 1995; Stewart
et al., 2003; Allegra et al., 2004). Ablation ofM~STN in mouse and the naturally occurring
mutations found in some Belgium Blue cattle result in greater muscle mass (Grobet al., 1997;
Kambadur et al., 1997; McPherron and Lee, 1997). Much less is known about BMP effects on
myogenesis. The objective of these experiments was to examine TGFP signaling components
during myogenesis in embryonic myoblasts and a postnatal satellite cell derivative.
Differential Transcriptional Activity in Myoblasts versus Myofibers
Autocrine activity during the transition from myoblasts to myofibers was examined in
C2C 12 cells transfected with a TGFP response element promoter reporter (3 TP-Lux) (Wrana et
al., 1992), a BMP responsive reporter gene (BRE-Luc) (Korchynskyi and ten Dijke, 2002) or a
muscle reporter gene (Tnl-Luc). Myoblasts were maintained in differentiation-permissive
medium for 48 hours prior to lysis, and luciferase activity measured. As expected, Tnl-Luc
levels were barely detectable in myoblasts and increased approximately 12-fold during the
conversion to myofibers (Figure 3-1). Evaluation of BMP and TGFP-driven transcription varied
in myogenic cells. In myoblasts, 3TP-Lux relative activity is decreased during differentiation.
Conversely, BMP-directed transcriptional activity doubled in myofibers as compared to
myoblasts (Figure 3-1). The different reporter levels observed in myoblasts and myofibers
suggests that autocrine loops are present that may contribute to distinct stages of myogenesis.
Differentiation TGFP Gene Expression in Myoblasts and Myofibers
Numerical changes in reporter activity suggested differential expression of TGFP proteins
during myogenesis. RT2 Gene Expression Assays were used to quantify changes in relative
TGFP superfamily member gene expression profies in myoblasts and myofibers. Total RNA
was isolated from 23A2 (embryonic myoblasts) and C2C12 (adult satellite cells) myofibers and
evaluated by ethidium bromide impregnated formaldehyde agarose gels. Distinct 18S and 28S
bands demonstrate intact RNA (Figure 3-2A). Equal amounts of RNA were reverse transcribed
in the presence of biotin-UTP (Figure 3-2B). The TGFP miniarray was hybridized with
biotynlated cDNA according to manufacturer' s recommendations (SuperArray). Results
demonstrate Statl, Noggin, and Runx1 transcripts are more abundant in 23A2 myofibers than
C2C12 myofibers. By contrast, Bnp3, Bnprla, Itgb5, and Igfbp3 expression is greater in
C2C12 myofibers than 23A2 myofibers (Figure 3-3). Appendix A provides a complete list of
genes and spotting locations.
Total RNA was isolated from proliferating C2C12 myoblasts or differentiated myofibers
and relative gene expression was determined by nylon array. Results demonstrate that BMP
ligands and receptors, TGFP ligands, and Runxs are expressed differentially during the transition
from myoblasts to myofibers. Runx1 mRNA levels are low in mononucleated myoblasts and
increase in multinucleated myofibers; Runx2 mRNA levels are high in myoblasts and low in
myofibers and TGFP3 is low in myoblasts and increases in myofibers (Figure 3-4).
The BMP ligands are inductive and inhibitory in myogenesis (Tzahor et al., 2003).
Expression ofBM~P1, 2, 3, 4, and 6 were detectable in both myoblasts and myofibers. Closer
examination of these BMPs shows that BMP3, BMP4 and BMP6 mRNA levels are greater in
mononucleated myoblasts than multinucleated myofibers. BM~P2 expression does not appear to
differ between myoblasts and myofibers. BMP1 mRNA slightly increases from myoblast to
myofibers (Figure 3-5). Relative amounts ofBM~P1, 3, 4, and 6 messages differ in myoblasts
versus myofibers suggesting these genes may be involved in distinct stages of muscle formation.
BM~P6 expression is greater in myoblasts than myofibers. Autocrine regulation of BMP6
was examined in C2C12 myofibers. Results demonstrated that BM~P6 levels, and other BMP
ligands, are not affected by ectopic BMP6 treatment (Figure 3-6). In summary, differential gene
expression of TGFP superfamily members is detected in different myoblast cell lines and during
various stages of myogenesis. BMP6 mRNA is abundant in myoblasts but does not up-regulate
it own expression.
Assessment of autocrine activity during the transition from myoblast to myofiber by
evaluation of BMP-responsive, TGFP-responsive or muscle specific reporter elements
demonstrated the presence of autocrine loops that may contribute to distinct stages of
myogenesis. Tnl-Luc activity drastically increased during the conversion of myoblasts to
myofibers, which confirmed successful differentiation. TGFP-sensitivity was high in myoblasts
and low in myofibers, supporting the idea that a TGFP-like protein is inhibiting differentiation.
Conversely, BMP-sensitivity was low in myoblasts and high in myofibers. The different reporter
levels observed in myoblasts and myofibers further suggests that there are differential responses
for TGFP and BMPs at different stages of skeletal myogenesis or myoblasts versus myofibers.
23A2 cells are embryonic myoblasts that were derived from C3H10T1/2 embryonic
fibroblasts treated with 5-Azacytidine (Konieczny and Emerson, 1984). C2C12 are adult
satellite cells derived from murine limb muscle (Blau et al., 1985). While both of these
immortalized cell lines exhibit myogenic properties, there are also marked differences between
23A2 and C2C12 cells. Comparison of embryonic myoblasts and adult satellite cells
demonstrated detectable gene expression differences. Transcripts for Statl, Noggin, and Runx1
are greater in embryonic myoblasts than adult satellite cells. Stat1 is involved in modulating
anti-proliferative and growth arrest signals by inducing expression of cell cycle inhibitors,
p21wAFl/CIP and pro-apoptotic signals (Durbin et al., 1996; Meraz et al., 1996). The Stat1
knockout mouse demonstrates no overt developmental abnormalities although the mice have a
significant increase in bone mineral density and bone mineral content (Xiao et al., 2004).
Greater levels of Stat1 in 23A2 myofibers may explain why ALP induction is greater in C2C12
than 23A2 myoblasts since it may act as a negative regulator of osteogenic activity.
Conversely, Bnp3, Bnprla, Itgb5, and Igfbp3 transcripts are greater in adult satellite cells
than embryonic myoblasts. Bnp3 is the most abundantly expressed member of the BMP
subfamily in both mononucleated myoblasts and multinucleated myofibers assessed by nylon
array. BMP3 is also the most abundant BMP in adult bone and a maj or component of
osteogenin, which does have osteogenic activity (Wozney and Rosen, 1993; Luyten et al., 1989).
However, recombinant BMP3 is unable to induce an osteogenic response in multiple cell lines
(Bahamonde and Lyons, 2001; Daluiski et al., 2001). Knockout models demonstrate that Bnp3
is a negative determinant of bone density with null mice exhibiting twice as much trabecular
bone as wildtype counterparts. Additionally, BMP3 inhibits BMP2-responsiveness in
osteoprogenitor cells and acts as an antagonist of osteogenic BMPs by activating the
TGFP/activin pathway, which would antagonize BMP signaling (Bahamonde and Lyons, 2001;
Daluiski et al., 2001). Therefore, the abundant gene expression of BMP3 would be important for
skeletal muscle formation because high levels ofBmp3 play an essential role in modulating
osteogenic BMPs. Specifically, Bmp3 would help inhibit osteogenic conversion of myoblasts
through competition for signaling components similar to both the TGFP/activin and BMP
pathways such as Smad4 (Heldin et al., 1997).
Gene expression profies for myoblasts and myofibers demonstrated differential gene
expression such as Runx1 and TGF/73 are greater in myofibers than myoblasts and Runx2 is
greater in myoblasts than myofibers. Results suggest that these genes play a role in cell origin
and regulate distinct stages of skeletal myogenesis. Runx2 (or Cbfal) is the master regulator for
bone development and the Runx2 knockout mouse demonstrates a complete lack of bone
formation and chondrocyte hypertrophy in most of the skeleton (Shum and Nuckolls, 2002).
Based on nylon array results, Runx2 is greater in myoblasts than myofibers and this may be a
function of the ability of myoblasts to undergo transdifferentiation in response to BMP6.
Alternatively, myofibers and fibroblasts do not undergo transdifferentiation, which may be due
to lower levels of Runx2. Nylon arrays demonstrated that there are differential gene expression
profiles between 23A2 and C2C12 myofibers. Additionally, both lines were found to undergo
transdifferentiation in response to ectopic BMP6 treatment, but to different magnitudes which
also demonstrates differences in the molecular signaling of these immortalized lines.
In conclusion, these nylon arrays allow for a glimpse of differential gene expression of
embryonic myoblasts and postnatal satellite cells, and different stages of skeletal myogenesis.
Verification of these results by Real-Time PCR analyses are needed with specific primers for
each uniquely expressed gene. While previous literature demonstrates a role for these BMPs in
embryonic development and somite patterning (Winnier et al., 1995; Zhang and Bradley, 1996;
Dunn et al., 1997; Solloway et al., 1998; Daluiski et al., 2001; Ying and Zhao, 2001), the role of
BMPs in adult myogenic cells has not been highly researched and this would determine which
TGFP superfamily members, specifically BMP genes, are important for molecular regulation of
myogenic differentiation and fusion.
Figure 3-1. Differential transcriptional activity in myoblasts versus myofibers. C2C12
myoblasts (1 x 105) were transfected with 1 Cpg of 3TP-luc (TGFP-specific), BRE-luc
(BMP-responsive), or Tnl-luc (Muscle-specific). Cells were maintained in growth
medium (myoblasts) or differentiation-permissive medium (myofibers) for 48 hours.
Cells were lysed, and luciferase and Renilla luciferase activities were measured.
Means and SEM are from three independent experiments. Different letters indicates a
significant difference, P<0.05.
23A2 biotin-cD>NA IJ g a
C2C12 biotin-cDNA .g g. 1$
1:20 1:80 1:320 1:1280 1:5120 dilution
Figure 3-2. Intact RNA and cDNA probe synthesis. Total RNA was isolated from 23A2 and
C2C12 myofibers. 15 Clg was electrophoretically separated through formaldehyde
agarose gels. RNA integrity was visualized with ethidium bromide, 18S and 28S
ribosomal RNA are noted (A). One Clg of RNA was reverse transcribed in the
presence of biotin-UTP. Serial dilutions of biotin cDNA were spotted to nylon and
incubated with streptavidin HRP. Chemiluminescence demonstrates biotin-UTP
e as ~Itgb
Figre -3 Gene exrsso prfie in emroniad a teltecl myofes ToalN a
Figu Sgnfian g-. ene exprsiproiession dfemrenic arindicated in thel yolumnbetwee TtheRN two
Msolae rpogenetic Poeibyni ReeporieA la, Collal Prclagen, tye I, alph 1,fr Itb5,
Inernd betual 5, 1gfp(3, Insli n-like gevrowt fatosrbindin prtin 3,e pRunx1, Rubitnt
re .Bitnlated trncrpi n s fact re 1,a d Statl, ea GP/M Signal tasdcr n aci at r of
tSprrasriptiFolon in 1.rdztotebosweeicbtdwt vd
rt a :
rt "It2 :
-- Activin Receptors
- BMP Ilganids
- BMP liganids and receptors
- BMP antagonists & collagens
- ids and SMAD target genes
ECMs and Development
-T~GFB ligands and receptors
Symbol Le gend
S= Runx2 (Cbfal)
Figure 3-4. Differential gene expression in myoblasts and myofibers. Total RNA was isolated
from C2C12 myoblasts (A) or myofibers (B) and equal amounts (3 Cpg) of RNA were
reverse transcribed in the presence of biotin-UTP. Biotinylated cDNAs were used to
probe a TGFP/BMP Signaling Array (SuperArray). Following hybridization, the
blots were incubated with avidin-peroxidase and visualized by chemiluminescence.
Representative blots are shown. Gene Classifieation Groups are indicated on the
right of the diagram. Abbreviations: BMP, Bone Morphogenic Protein, TGFP,
Transforming Growth Factor beta, GDFs, Growth and Differentiation Factors, Ids,
Inhibitor of DNA binding, SMAD, MAD homolog, ECMs, Extracellular Matrix
Figure 3-5. Relative BMP gene expression in myoblasts and myofibers. Total RNA was
isolated from C2C12 myoblasts and myofibers and equal amounts (3 Gig) of RNA
were reverse transcribed in the presence of biotin-UTP. Biotynlated cDNAs were
used to probe a TGFP/BMP signaling array (SuperArray). Following hybridization,
blots were incubated with avidin-peroxidase and visualized by chemiluminescence.
Representative subsections of blots for BMP 1, 2, 3, 4, 5, 6, 10, and 15 are shown.
(A) Myoblasts; (B) Myofibers; (C) Schematic location of genes.
G = GAPDH
Figure 3-6. BMP6 does not undergo autocrine gene regulation. C2C 12 myoblasts were
differentiated for 48 hours in the absence (A) or presence (B) of 100 ng/ml BMP6.
Total RNA was isolated and equal amounts (3 Gg) were reverse transcribed in the
presence of biotin-UTP. Biotynlated cDNAs were used to probe a TGFP/BMP
signaling array (SuperArray). Following hybridization, blots were incubated with
avidin-peroxidase and visualized by chemiluminescence. Schematic location of BMP
IMPACT OF BMP6 ON SKELETAL MYOGENESIS
In Raf-arrested myoblasts, removal of TGFP1 biological activity does not restore the
myogenic program suggesting that another factor is mediating this inhibition (Wang et al., 2004).
Ectopic GDF8 treatment does not inhibit 23A2 myoblast differentiation. BMP6 mRNA is
present in myoblasts and absent in myofibers suggesting an inhibitory action during myogenesis
(Derynck, 1989). The balance of proliferation, cell differentiation, and apoptosis mediates the
pool of myoblasts available for skeletal myogenic maintenance. The objective of these
experiments was to measure BMP6 effects on myoblast proliferation, differentiation, and
Inhibition of Skeletal Myogenic Differentiation by BMP6
To further examine the role of BMP6 in embryonic myoblasts and satellite cells, 23A2
myoblasts and C2C12 satellite cells were transiently transfected with Tnl-Luc and treated with 0
or 100 ng/ml BMP6 for 48 hours in differentiation media prior to lysis and measurement of
luciferase activity. Results show that treatment with 100 ng/ml BMP6 resulted in more than an
80-fold inhibition of muscle specific reporter (Tnl-Luc) activity in both myogenic cell types
(Figure 4-1, P<0.05). Due to the significant biochemical inhibition of differentiation by BMP6,
assessment of morphological changes was performed. C2C 12 myoblasts were differentiated in
the presence of 100 ng/ml BMP6 for 48 hours. Subsequently, cells were Eixed and
immunostained for MyHC. Results show very few multinucleated fibers in response to BMP6
treatment (Figure 4-2A). Total cell lysates were isolated from a second set of C2C12 myofibers
treated in an analogous manner. Equal amounts of protein (10Clg) were analyzed by Western
blot for contractile and regulatory protein expression. Inhibition of the muscle contractile protein
MyHC, in addition to, myogenin, and troponin T protein markers was observed in response to
BMP6 treatment (Figure 4-2B). Thus, BMP6 inhibits the complete differentiation program.
Dose-Dependent Effects of Recombinant BMP6 on Myoblasts
BM~P 2, 4, and 7 have opposing activities that are concentration-dependent during
embryonic muscle growth (Amthor et al., 1998; Amthor et al., 2002). Due to the high
concentration of BMP6, a retrospective dose-response experiment was performed (1 ng/ml to
50 ng/ml BMP6). Biochemical differentiation was measured following transient transfection of
Tnl-Luc into C2C12 myoblasts. Cells were treated for 48 hours in differentiation media with
increasing concentrations of BMP6. Analysis of luciferase reporter activity shows an
approximate 20% decrease of Tnl-Luc activity at 25 ng/ml (P<0.05) and an approximate 50%
decrease at 50 ng/ml (P<0.001) (Figure 4-3A). Parallel plates were lysed for protein analyses.
Equal amounts of total cell proteins were separated by SDS-PAGE and transferred to
nitrocellulose. Blots were probed with anti-MyHC and anti-myogenin. Results demonstrate that
MyHC protein expression begins to decrease at 50 ng/ml BMP6 (Figure 4-3B). Interestingly,
myogenin protein expression begins to decrease at 10 ng/ml BMP6, further at 25 ng/ml and is
completely absent in response to 50 ng/ml BMP6 (Figure 4-3B).
Induction of ALP Activity in Response to BMP6
BMPs induce bone or cartilage formation ectopically (Urist, 1965; Gitelman et al., 1995).
BMP2 inhibits the myogenic differentiation of C2C12 cells, by converting their differentiation
pathway into that of osteoblast lineage cells (Katagiri et al., 1994). To determine if the block to
differentiation is associated with transdifferentiation, assessment of alkaline phosphatase (ALP)
activity of myofibers treated with BMP6 was measured. C2C12 myoblasts were treated with
vehicle only or increasing amounts of BMP6 in differentiation media for 48 hours. Cells were
Eixed and histologically stained for ALP activity, a marker enzyme of osteogenic cells. As
shown in Figure 4-4, as little as 5 ng/ml BMP6 was sufficient to induce ALP activity.
To determine if the effect of BMP6 is unique to the committed myoblast, C3H10T1/2
fibroblasts were treated with 0, 1, 10, and 100 ng/ml BMP6 in differentiation media for 48 hours.
Cells were Eixed and histologically stained for ALP activity. Results indicated that BMP6 does
not cause ALP induction in C3H10T1/2 fibroblasts (Figure 4-5).
BMP6 Induces Rapid Transdifferentiation in Myoblasts
C2C 12 myoblasts were treated with 100 ng/ml BMP6 for 24 and 48 hours. Sub sequently,
cells were Eixed, and ALP activity was measured. Results show that ALP activity is observed as
early as 24 hours and further increases at 48 hours (Figure 4-6). C2C 12 cells treated with vehicle
alone also demonstrated a slight induction of ALP at 24 and 48 hours indicating endogenous
BMP6 Does Not Alter Proliferation Rates of Myoblasts
Cell cycle withdrawal or inhibition of proliferation is required for differentiation and
fusion of myofibers. To determine if the block to myogenic differentiation by BMP6 is due to
altered cell proliferation, C2C12 myoblasts were treated with 25 ng/ml BMP6 for 48 hours.
Cells were pulsed with BrdU during the final thirty minutes of the treatment interval, fixed with
methanol, and immunostained for BrdU (Figure 4-7A). The numbers of nuclei were not different
between control and BMP6 treated cells (Figure 4-7B, P<0.05). Therefore, exposure of
myoblasts to BMP6 does not alter cellular proliferation.
BMP6 is not Anti-Apoptotic
Previous reports indicate that BMP2 and 4 promote cell survival in pluripotent
mesenchymal cells by inhibiting TNF-mediated apoptosis (Chen et al., 2001). BMP6 was
demonstrated to partially restore survivability in human mesenchymal stem cells (hMSCs)
induced by the BMP antagonist sclerostin (Sutherland et al., 2004). Bcl-2 is an anti-apoptotic
protein expressed abundantly in satellite cells (Krajnak et al, 2006). The protective actions of
BMP6 were examined in C2C12 myoblasts treated with staurosporine. C2C12 satellite cells
were treated with increasing concentrations of BMP6 for 48 hours and Bcl-2 protein content
measured by Western blot. Results demonstrate no changes in Bcl-2 protein concentration
following BMP6 treatment (Figure 4-8A). Subsequently, myoblasts were treated for 3 hours
with BMP6, 25 C1M staurosporine or BMP6 and staurosporine. Cells were lysed and assayed for
Poly (ADP-ribose) polymerase (PARP) protein expression by Western blot analysis. Results
demonstrate that staurosporine stimulates PARP cleavage, a hallmark of apoptosis (Figure 4-8B).
BMP6 does not prevent PARP cleavage in C2C12 myoblasts treated with staurosporine, thus
precluding an anti-apoptotic function (Figure 4-8B).
In summary, BMP6 significantly inhibits the complete differentiation program as observed
by biochemical suppression of muscle specific reporter activity, and morphological disruption of
myofiber formation and muscle-specific protein expression. This inhibitory effect is dose
dependent and results in rapid transdifferentiation of committed mesodermally-derived
myoblasts to an osteogenic lineage. BMP6 does not inhibit differentiation by promotion of a
proliferative state. Nor does BMP6 serve as an anti-apoptotic factor in myoblasts.
BM~P6 is greater in myoblasts than myofibers, and up-regulated in Raf-arrested myoblasts.
Therefore, the function of BMP6 on proliferation, differentiation, and apoptosis in skeletal
myogenesis was evaluated. Treatment of 23A2 and C2C12 myoblasts with 100 ng/ml BMP6
significantly inhibited muscle specific activity (Tnl-luc). Muscle fiber fusion and muscle
specific protein expression of myosin heavy chain (MyHC), myogenin and Troponin T are also
decreased in response to BMP6 treatment. Results suggest that BMP6 inhibits the complete
Other members of the TGFP superfamily can also inhibit differentiation. One of the most
notable examples is GDF8 or myostatin. M~yostatin is mutated in double muscled cattle breeds,
Belgium Blue and Piedmontese (Grobet et al., 1997; Kambadur et al., 1997; McPherron and Lee,
1997). It is predominately expressed in the muscle and negatively regulates myogenic
proliferation. The myostatin knockout mouse also demonstrates two times larger muscle mass as
compared to wild type counterparts (Thomas et al., 2000). Therefore, TGFP superfamily
members are critical mediators of skeletal myogenesis.
BMP6 is a morphogen, and morphogens are characterized by exhibiting different effects at
different levels or concentrations. In pre-myogenic cells, BMP2, 4, and 7 have dose dependent
effects with low concentrations maintaining a Pax3-expressing proliferative population and
delaying differentiation. Conversely, high concentrations of these BMPs prevent muscle
development (Amthor et al., 1998). In the presence of low BMP levels, myogenic precursor
cells are maintained in a proliferative state in developing limb bud, while high BMP levels
induce cell death. Thus, BMPs can both stimulate and restrict muscle growth (Amthor et al.,
1998; Amthor et al., 2002). This suggests that a concentration gradient of BMPs is needed for
the correct determination and maintenance of the myogenic program (Centrella et al., 1994;
Alliston et al., 2001; Reddi, 1994). 23A2 myoblasts treated with increasing concentrations of
BMP6 were measured for muscle specific reporter activity and MyHC and myogenin protein
expression. Results demonstrated that muscle reporter activity is significantly decreased at both
25 ng/ml and 50 ng/ml. MyHC protein expression began to decrease at 50 ng/ml BMP6 and
myogenin decreased at 10 ng/ml BMP6. This suggests that BMP6 exhibits dose dependent
effects on contractile and regulatory muscle proteins. Smad1 and Smad5 binding sites are found
within the promoter of myogenin, which allow BMP6 to induce downstream Smad signaling.
C2C12 cells transiently transfected with Smad1 and Smad5 were able to induced ALP activity
and decrease myogenin/chloramphenical acetyltransferase (myogenin-CAT) activity. Although
in NIH3T3 fibroblasts, Smad1 and Smad5 decreased myogenin-CAT but did not induced ALP
activity, which demonstrates that Smad1 and Smad5 are involved in the BMP signaling that
inhibits myogenic differentiation and induces transdifferentiation. Furthermore, the conversion
of these two differentiation pathways is regulated independently at the transcriptional level
(Yamamato et al., 1997). Additionally, high levels of BMP6 impact or repress stages of skeletal
myogenesis prior to terminal differentiation.
Promotion of osteogenic differentiation by BMP2 expression in skeletal muscle-derived
C2C12 cells (Musgrave et al. 2001) demonstrated that myoblasts can undergo
transdifferentiation from a myogenic to osteogenic cellular lineage in response to BMP. ALP
staining measures osteogenic activity and ectopic BMP6 treatment of C2C12 myoblasts
demonstrated a dose dependent induction of ALP activity. Additionally, when C2C12 myoblasts
were treated for 24 and 48 hours, induction of ALP activity in response to BMP6 treatment
demonstrated a time dependent transdifferentiation of myofibers into an osteogenic lineage.
The substantial induction in the numbers of ALP positive cells in C2C12 cultures maintained in
low serum at 24 and 48 hours also suggests a basal level of osteogenic activity in C2C12
myofibers. These cells are mononucleates and not myofibers. Comparison of gene transcripts
between myoblasts and myofibers by nylon array suggests that greater transcripts ofRunx1 could
explain why these cells demonstrate a basal level of osteogenic activity. Furthermore, BM~P1, 3,
and 4 transcripts are expressed in both untreated and BMP6 treated myoblasts, but only BM~P4
displays osteogenic activity, which could contribute to endogenous ALP activity of C2C12
myoblasts. BMPs 2, 4, 6, 7, and 9 are characterized as osteogenic BMPs, with BMPs 2, 7, 6, and
9 displaying the highest osteogenic activity when applied ectopically in vitro and in vivo (Luu et
BMP 1 is a metalloprotease that regulates deposition of fibrous extracellular matrix (ECM)
in vertebrates and does not display osteogenic activity (Bond and Beynon, 1995). It provides
procollagen C proteinase (PCP) activity to cleave the C propeptides of procollagens I-III to yield
the major fibrous components ofECM (Kessler et al., 1996; Li et al., 1996; Suzuki et al., 1996;
Scott et al., 1999). In bone, BMP1 co-purifies with TGFP-like BMPs from osteogenic extracts
of bone and is believed to coordinate the deposition of ECM with the activation of certain BMPs
in early development and later in the development of bone and other tissues (Wozney et al.,
1988; Scott et al., 2000). Furthermore, BMP3 would not contribute to endogenous ALP activity
because as mentioned previously, recombinant BMP3 does not display osteogenic activity and
acts as an antagonist of osteogenic BMPs (Bahamonde and Lyons, 2001; Daluiski et al., 2001).
Conversely, exposure of C3H10T1/2 fibroblasts to a BMP6 dose-response curve did not
induce ALP activity at any dosage for 1, 10, and 100 ng/ml BMP6. The differential responses
observed in C2C12 and C3H10T1/2 cells suggest that BMP6-mediated transdifferentiation is
specific to lineage restricted cells. Similar results of a significant ALP induction in C2C 12 cells
by BMPs (2, 4, 6, 7, and 9) was reported by other groups. C2C12 cells demonstrated the most
potent ALP induction (Ebisawa et al., 1999). Therefore, the origin of the cell line may determine
if the cells will respond to BMPs and/or the overall magnitude of the ALP response. Yang et al.,
(2003) did demonstrate induction of ALP activity and stimulation of osteoblast marker genes by
300ng/ml of recombinant BMP6, whereas, our experiments never used a dosage higher than
100ng/ml BMP6. Another group demonstrated an induction of ALP activity in C3H10T1/2
fibroblasts in response to 100 ng/ml BMP2, suggesting that other BMP family members or
higher dosages of BMP6 may have different impacts on fibroblasts. Additionally, BMP2-
mediated ALP induction in C3H10T1/2 fibroblasts was increased in the presence of HGF (Imai
et al., 2005). Therefore, the addition of HGF with BMP6 might result in a detectable ALP
induction in C3H10T1/2 fibroblasts.
Comparison of BMP2 and BMP6 finds that both are expressed in skeletal muscle cells but
are in different classes of the BMP subfamily (Lyon et al, 1989). BMP2 is a member of the DPP
class along with BMP4, whereas, BMP6 is a member of the 60A class along with BMP5, 7, 8A,
and 8B (Gitelman et al., 1997). Structurally, BMP2 contains 3 exons, two of which encode the
precursor protein (Feng et al., 1994). The non-coding exon refers to an exon located on the 5'
flanking region of the DNA that has been shown to serve as an alternate promoter, which
suggests that BMP2 is regulated in both a developmental and tissue-specific manner (Gitelman et
al., 1994). BMP6 is composed of 7 coding exons with the mature protein encoded by 3 full
exons and a portion of the fourth exon and is also developmentally regulated. While BMP2 and
BMP6 are members of the same subfamily, there is little similarity in the localization of the
intron-exon structures further demonstrating differences at the structural level (Gitelman et al.,
In vivo analyses demonstrates that BM~P6 null mice are viable and fertile and exhibit no
maj or defects in known BMP6-expressing tissues, except for a delay in ossification restricted to
the developing sternum (Solloway et al., 1998). It is believed that BM~P2 may be functionally
compensating for BM~P6 ablation since BM~P2 and BM~P6 are required for some overlapping or
redundant functions (Solloway et al., 1998). Based on nylon arrays, BM~P6 is greater in
myoblasts than myofibers, yet BM~P2 and BM~P4 transcripts don't appear to differ in myoblasts
and myofibers. Conversely, when the muscles of athymic nude rats were inj ected with
adenoviral vectors for BMP6 (AdBMP6), a rapid tissue calcifieation was observed. The
induction of bone was produced through mechanisms similar to both intramembranous and
endochondral ossifieation pathways and AdBMP6 was even more potent than the prototypical
adenoviral vector AdBMP2 (Jane et al., 2002). These studies utilized titers of BMP6 that would
not mimic physiological conditions but do demonstrate the powerful osteogenic activity of
BMP6. BM~P2 null mice exhibit multiple developmental defects including a delayed primitive
streak, small allantois, lack of amnion, heart defects and a decreased number of primordial germ
cells (Zhang and Bradley, 1996; Yin and Zhao, 2001). Therefore, BMP2 and BMP6 appear to
have different biological functions.
It has also been suggested that receptor oligomerization determines BMP2 signaling
pathways. Nohe et al., (2002) demonstrated that binding of BMP2 to preformed receptor
complexes activates the Smad pathway. Conversely, BMP2-induced recruitment of receptors
activates a Smad-independent pathway, which results in the induction of ALP activity via p38
MAPK (Nohe et al., 2002). These different receptor complexes may also recruit different
adaptor proteins such as, XIAP (Yamaguchi et al., 1999), BRAM-1 (Kurozumi et al., 1998), and
FKBPl2 (Wang et al., 1996). BMP2 treatment of C3H10T1/2 cells stimulates ERK1 and ERK2
during osteoblastic differentiation (Lou e al., 2000), and ERK activation can inhibit nuclear
translocation of Smad1, which would block the Smad pathway (Kretzschmar et al., 1997). Our
experiments demonstrate a strong Smadl/5/8 activation by BMP6 and only a slight p38
activation suggesting that BMP6 induces ALP activity through a different signaling mechanism
Characterization of BMP6 and BMP2 in C2C12 cells demonstrates that BMP6 strongly
binds to activin receptor-like kinase (ALK)-2 or ActR-I (Ebisawa et al., 1999). ALK2 forms
complexes with receptors like, BMPR-II or ActR-II. BMP6 can also weakly bind to ALK3,
which also can bind BMP2 (Ebisawa et al., 1999) but BMP2 preferentially binds to BMPRIA
and IB. C2C12 cells only express mRNA for BMPRIA (Akiyama et al., 1997). C3H10T1/2
fibroblasts express both BMPRIA and BMPRIB but BMPRIA expression levels are
endogenously higher. Transfection experiments with BMPRIA and BMPRIB in C3H10T1/2
cells also demonstrate that the dominant role in BMP2 mediated osteogenic development was
mediated by BMPRIA, with BMPRIB only partially influencing osteogenic development (Kaps
et al., 2004). Additionally, type IB and IA BMP receptors appear to transmit different signals
during the specification and differentiation of mesenchymal lineages (Kaps et al., 2004).
Truncation and overexpression of BMP receptors, BMPRIA and BMPRIB, have demonstrated
that overall receptor levels expressed in cells play a critical role in specification and
differentiation of osteoblasts by BMP2 (Chen et al., 1998). Therefore, while BMP2 and BMP6
can interact with the same BMP receptor, they demonstrate the strongest affinity for different
type I receptors. Therefore, differences in signaling could be based on like receptor
oligomerization, cross-talk with other signaling pathways, competition for signaling components,
interaction with additional proteins, and activation of different downstream transcriptional
targets, which would result in different biological responses of BMP2 and BMP6.
Transdifferentiation of myogenic cells to an osteogenic linage results in induction of
specific markers at the transcriptional level. Therefore, measurement of genes such as, Runx1,
Runx2, osteopontin, osteonectin, and osteocalcin (Ahrens et al., 1993) would determine if the
cells were exhibiting more "bone-like" characteristics and what genes are up-regulated in BMP6-
treated myoblasts. Interestingly, based on the nylon array assessment of myoblasts versus fibers,
Runx1 appeared to be greater in myofibers, yet Runx2 appeared to be greater in myoblasts.
Runx2 is required for later stages of chondrocyte and osteoblast differentiation, while Runx1
mediates early events of endochondral and intramembranous bone formation (Smith et al., 2004).
These results would need to be confirmed by Real-Time PCR and may have implications in
transcriptional regulation of the myoblast to myofiber transition and/or transdifferentiation of
myoblasts to osteoblasts. Since myogenic precursor cells (MPCs) also have the ability to form
skeletal muscle, bone, or cartilage, the differential expression of these transcription factors may
demonstrate how these precursor cells determine their ultimate cell fate.
Inhibition of embryonic skeletal muscle differentiation through promotion of a
proliferative state was monitored. Others have suggested that Raf inhibits skeletal myogenesis
by keeping myoblasts in a proliferative state (Samuel et al., 1999). Since BMP6 expression is
high in Raf-arrested myoblasts, it was proposed that BMP6 may also be promoting a
proliferative state of cells. When myoblasts were treated with BMP6, the number of BrdU
positive nuclei or cells in S-phase was not significantly different from untreated myoblasts. The
percent BrdU incorporation was calculated by dividing the number of BrdU positive nuclei by
the total number of cells and taking the average of six Hields. There also was no significant
difference in BMP6 treated percent BrdU incorporation versus control myoblasts. Therefore,
exposure of myoblasts to BMP6 does not inhibit proliferation suggesting that BMP6-mediated
inhibition of differentiation is not mediated through modulation of proliferative events.
Conversely, another group did demonstrate an inhibition of proliferation by [3H] thymidine
incorporation assays in a dose-dependent manner (Ebisawa et al., 1999). Comparatively, this
study did a dose response up to 1000 ng/ml BMP6 and only observed approximately a 20%
decrease at 300ng/ml, whereas our analyses assessed BrdU incorporation at 25 ng/ml, which is
considered a more physiological dose.
Apoptosis is a mechanism of programmed cell death that promotes tissue turnover,
embryonic development, and immunological defense mechanisms (Kerr et al., 1972; Adams,
2000; Siu et al., 2005). The intrinsic and extrinsic signaling pathways are the two principal
pathways involved in apoptosis. The intrinsic route uses cell signaling pathways to alter
mitochondrial function. Permeabilization of the outer mitochondrial membrane results in release
of cytochrome c, thereby forming an apoptosome. This macromolecular complex activates
caspase-9, mediated by apoptotic protease activign factor (APAF)-1 (Roy and Nicholson, 2000).
Conversely, the extrinsic pathway sends 'death ligand' signals of the the Tumor Necrosis Factor
(TNF) or Fas families through appropriate receptors that activate caspase-8 in conjunction with
the adaptor molecule, Fas-associated death domain (FADD). At this point, intrinsic and extrinsic
cascades converge to activate effector caspases (caspase-3 and -7), which cause the proteolytic
degradation of cellular material (Roy and Nicholson, 2000).
BMP2, BMP4, and BMP6 promote cell survival in mesenchymal cells (Chen et al., 2001;
Sutherland et al., 2004). Yet when myoblasts were treated with ectopic BMP6, Bcl-2 protein
expression, an anti-apoptotic marker, did not change in response to increasing amounts of BMP6.
Additionally, BMP6 did not prevent PARP cleavage in response to staurosporine, a common
inducer of apoptosis. This suggests that BMP6 does not have an anti-apoptotic function in
skeletal myoblasts. BMP6 was able to partially protect hMSC cells from sclerostin-induced
apoptosis by decreasing caspase activity (Sutherland et al., 2004). Sclerostin is a BMP
antagonist that binds to BMPs and blocks downstream signaling pathways. Furthermore, the
protective effect of BMP6 observed was only a partial block of sclerostin-mediated apoptosis
suggesting that additional factors are involved (Winkler et al., 2003). Therefore, the protective
effect of BMP6 observed in this scenario may be a result of additional signaling components not
present in myoblasts or is due to an alternative apoptotic signaling cascade specific to Sclerostin
or different from staurosporine-induced apoptosis.
In conclusion, BMP6 inhibits the complete differentiation program in myoblasts. BMP6
treatment results in a rapid transdifferentiation of myoblasts that is specific to a committed
mesodermal derived myogenic cell. BMP6 does not appear to involve modulation of
proliferation rates of myoblasts. Nor does BMP6 appear to demonstrate a survival role in
:E~ ~ ~ I iC2C12
Figure 4-1. Biochemical inhibition of skeletal myogenesis by BMP6. Myoblasts were
transiently transfected with 2 Clg of Tnl-Luc reporter construct and 50 ng pRL-tk
(Renilla) and treated 0 or 100ng/ml BMP6 in differentiation media for 48 hours.
Cells were lysed and assessed for Tnl-Luc activity in both 23A2 and C2C12
myofibers. Reporter luciferase activity was normalized to the amount ofRenilla
luciferase activity and vehicle only was set to 100%. Data represents the mean and
standard error of the mean (SEM) of three independent experiments. Different letters
indicates a significant difference, P<0.05.
ce-myo ge nm
| a-Troponin T
- -- ~ cl-desmin
Figure 4-2. Inhibition of skeletal myogenesis by BMP6. C2C12 myoblasts were treated with
100 ng/ml BMP6 for 48 hours. Myoblasts were fixed and immunostained for MyHC
expression. Representative microscopic images (200X) are shown (A). Parallel
plates were lysed and analyzed for muscle specific proteins by Western blot (B).
Tubulin expression was used as a loading control.
O 1 5 10 25 50
O 1 5 10 25 50 BMP6 (ng/ml)
,~~9 ggp e ii-MyHC
e. .. a-myogemni
Figure 4-3. BMP6 dose response curve. C2C12 myoblasts (1x105) were transiently transfected
with 2 Clg of Tnl-Luc reporter construct and 50 ng pRL-tk (Renilla). Cells were
maintained in differentiation media supplemented with BMP6 for 48 hours. Cells
were lysed and luciferase activities were measured (A). Reporter luciferase activity
was normalized to the amount of Renilla luciferase activity and the control (vehicle
only) was set to 100%. Data represents the mean and standard error of the mean
(SEM) of three independent experiments. Different letters indicates significance at
p<0.05. Total cell lysates were analyzed by Western blot for MyHC and myogenin
protein expression (B). Tubulin expression was used as a loading control.
Figure 4-4. BMP6 induction of alkaline phosphatase activity. C2C12 myoblasts were treated
with 0, 5, 10, 25, 50, and 100 ng/ml BMP6 for 48 hours. Cells were fixed and ALP
activity detected colorimetrically. Representative photomicrographs at 200X are
Figure 4-5. BMP6 does not induce alkaline phosphatase (ALP) activity in fibroblasts.
C3H10T1/2 fibroblasts were treated with 0, 1, 10, and 100 ng/ml BMP6 for 48 hours.
Cells were fixed, and stained for ALP activity. Representative photomicrographs at
200X are shown.
In ml B1IM 6
10n ml hlVII%
;:~vi' i ...C
Figure 4-6. BMP6 induces rapid transdifferentiation in myoblasts. C2C12 myoblasts were
treated for 1, 24 or 48 hours with vehicle alone (control) or 100 ng/ml BMP-6.
Cultures were fixed with 4% paraformaldehyde and ALP activity measured
colorimetrically. Representative photo-micrographs at 200X demonstrate intense ALP
staining as early as 24 hours of treatment.
Vphiplp Only Ce *i
~ 25 -
QI ao -
Figure 4-7. BMP6 treatment does not alter myoblast proliferation. C2C12 myoblasts were
treated with 25 ng/ml BMP6 in differentiation medium for 48 hours, pulsed with
10 CtM BrdU for 30 minutes, fixed with 70% ethanol at 40C for 30 minutes and
immunostained for BrdU expression (A). Representative photomicrographs at 100X
shown. % BrdU incorporation was calculated by dividing the number of BrdU
positive nuclei by the total number of nuclei (B). The averages of a minimum of six
fields per treatment are shown.
0 1250 100 BMP6 (ng/ml)
.. -- a r-Bcl-2
+ -+ BMP6
Figure 4-8. BMP6 is not anti-apoptotic. C2C12 myoblasts were placed in differentiation
permissive media with increasing concentrations of BMP6 for 48 hours. Whole cell
lysates were prepared and analyzed for expression of Bcl-2 by Western blot (A).
C2C12 myoblasts were treated with 100ng/ml BMP6 in the absence or presence of
25 C1M staurosporine (STS). Whole cell lysates were assessed for PARP cleavage
and tubulin protein expression (B).
BMP6 SIGNALING DURING SKELETAL MYOGENESIS
During skeletal myogenesis, BMP6 actions can be mediated by multiple signaling
mechanisms (Figure 5-1). The specifieity of the BMP6 and TGFP1 signaling responses are
achieved by different types of Type I and Type II receptors and R-Smads (Wrana et al., 1992;
Attisano et al., 1993; Ebner et al., 1993; Wieser et al., 1993; Wrana et al., 1994). Traditionally,
BMPs signal through Smads 1/5/8, while TGFP1 signals through Smads 2/3. BMPs also initiate
p38 signaling and within muscle, this regulation is specific to the p3800 isoform. A third, more
novel regulation, is through cross-talk with Notch signaling. The objective of these experiments
was to validate the presence and activation of three putative intracellular signaling cascades
induced by BMP6.
Analysis of BMP Signaling Systems in Myoblasts
The ability of BMP6 and TGFP1 to induce Smad phosphorylation and nuclear translocation
was evaluated. In brief, C2C12 myoblasts were treated with 10 ng/ml TGFP1 and 25 ng/ml
BMP6 for 48 hours. Total cell lysates were analyzed by Western blot for total and phospho-
Smadl/5/8 expression. Myoblasts treated with vehicle alone demonstrated lows levels of
phospho-Smadl/5/8, which were increased in response to both 10 ng/ml TGFP1 and BMP6
(Figure 5-2A). No differences in total Smadl/5/8 or tubulin were evident. Parallel plates of
C2C12 myoblasts treated as described were Eixed and immunostained for phospho-Smadl/5/8.
In both instances, the signaling molecules are located in the nucleus (Figure 5-2B). Therefore,
the archeotypical Smad signaling system is intact and functional in myoblasts. These results also
support previous observations that demonstrated expression of Smadl/5/8 in C2C12 cells.
Subsequent phosphorylation of Smad5 and weak phosphorylation of Smad1 was also observed
by BMP6. Smad8 was found to be constitutively phosphorylated in C2C12 cells (Ebisawa et al.,
In addition to signaling through the Smad proteins, BMPs also may utilize components of
the MAPK and Janus kinase (JAK)/signal transducers and activators of transcription (STAT)
family of signaling proteins, culminating in activation of JNK or p3 8 (von Bubnoff and Cho,
2001). p38 signaling is a requirement for muscle formation (Zetser et al., 1999; Lee et al.,2002).
The interplay of BMP6 signaling and p38 kinase activity during myogenesis was examined.
C2C12 myoblasts were treated with increasing amounts of BMP6 and assessed for total and
phospho-p38 protein expression by Western blot. Results demonstrate that while levels of total
p3 8 don't change in response to BMP6, phospho-p3 8 is slightly induced in myoblasts treated
with 25 and 50 ng/ml BMP6 (Figure 5-3).
C2C 12 myoblasts were transiently transfected with Tnl-Luc activity and treated with
100 ng/ml BMP6, 10 CIM SB202190 or a combination of both for 48 hours. Results show
100 ng/ml BMP6 significantly inhibited Tnl-Luc activity, as observed previously (Figure 5-4A,
P<0.05). Treatment with 10 C1M SB202190 inhibited Tnl-Luc activity (Figure 5-4A, P<0.05).
Importantly, the combination of BMP6 and SB202190 demonstrated an additive effect indicating
independent pathways (Figure 5-4A, P<0.0001). Morphological observations showed myofibers
in control cells, few myofibers in SB202190 treated, and no myofibers in either BMP6 treated or
the combination of BMP6 and SB202190 (Figure 5-4B).
C2C12 myoblasts were treated with BMP6, SB202910, and BMP6 with SB202910 for 48
hours in differentiation media. Cells were lysed and assayed for MyHC, myogenin, Troponin T,
and desmin protein expression by Western blot. Results demonstrate that muscle specific protein
markers (MyHC, myogenin, and Troponin T) are reduced in the presence of BMP6, p38 inhibitor
(SB202910), and the combination (Figure 5-4). C2C12 myoblasts treated with both BMP6 and
SB202910 demonstrated a more severe inhibition of myogenesis.
The importance of p3 8 signaling during BMP6-mediated transdifferentiation was
examined. Parallel plates of C2C12 myoblasts treated with BMP6, SB202190, and the
combination of BMP6 and SB202190. ALP activity was detected colorimetrically. Results
show that inhibition of p3 8 function does not alter ALP induction in response to BMP6 treatment
(Figure 5-6). This further demonstrates that BMP6 is signaling through an independent pathway
to cause inhibition of differentiation and transdifferentiation.
Impact of Notch Inhibitor on BMP6-Mediated Inhibition of Differentiation
A third signaling mechanism used by BMP ligands involves Notch. Notch is a
transmembrane receptor that is cleaved on the intracellular surface to release a proteolytic
fragment that translocates to the nucleus and affects gene transcription (Nye et al., 1994; Ahmad
et al., 1995). Notch inhibits myogenic differentiation similar to BMP6 (Kopan et al., 1994;
Takahashi et al., 1994). BMP4 repression of muscle gene expression involves Notch activation
(Dahlqvist et al., 2003). The involvement of BMP6 in Notch controlled signaling was examined
in C2C12 myoblasts. In brief, myoblasts were induced to differentiate in the presence or absence
of 25 ng/ml BMP6 and 10 CIM L685,458, a Notch inhibitor, for 72 hours. Cells were fixed and
immunostained for MyHC. Results show that MyHC was expressed in control cells and
significantly inhibited in BMP6 treated (Figure 5-7A and 5-7B, P<0.05). Treatment of C2C12
myoblasts with L685,458 alone does not alter MyHC expression and interestingly, the
combination of BMP6 and L685,458 partially restores MyHC expression (Figure 5-7A and 5-7B,
P<0.05). This suggests that the BMP6-mediated inhibition of differentiation is partially
controlled by the Notch signaling pathway.
In summary, differential BMP- and TGFP-responsiveness and gene expression exists in
myoblasts versus myofibers. This suggests that TGFP superfamily members play different
regulatory roles at various stages of myogenesis as myoblasts undergo terminal differentiation.
Treatment of myoblasts with BMP6 results in a dramatic increase of ALP activity in a dose-
dependent and time-dependent manner. Exogenous BMP6 treatment of skeletal myoblasts
results in inhibition of myoblast differentiation as observed by significant inhibition of muscle
reporter activity, muscle specific protein synthesis, and myoblast fusion. BMP6 treatment does
not alter proliferation rates of myoblasts or promote a proliferative state to inhibit differentiation.
Inhibition of p3 8 activity combined with BMP6 treatment caused inhibition of TnI-Luc activity
that was greater than either treatment alone suggesting an additive effect between BMP6 and
inhibition of p3 8 activity. Interestingly, the combination of BMP6 and Notch inhibition by
L685,458 partially restores MyHC expression in fibers. This suggests that the myogenic
inhibitory effect observed in the presence of BMP6 is partially mediated by functional Notch
BMPs can be mediated by multiple signaling pathways and these experiments validated the
presence and activation of these cascades in skeletal myoblasts. BMPs signal through
serine/threonine kinase receptors (Massague et al., 1994). In the presence of growth factors,
ligands bind to a Type II receptor dimer located on the plasma membrane, which causes auto-
phosphorylation of the Type II dimer and recruitment and transphosphorylation of a Type I
receptor dimer (Wrana et al., 1992; Attisano et al., 1993; Ebner et al., 1993; Wieser et al., 1993;
Wrana et al., 1994). This phosphorylation event recruits the receptor-regulated Smads (R-
Smads), which then undergo phosphorylation (Aoki et al., 2001). The R-Smads form a complex
with the common-partner Smad (Co-Smad), Smad4. The R-Smad/Smad4 complex translocates
into the nucleus and binds to DNA causing activation of target genes (Derynck et al., 1996; Liu
et al., 1996; Meersseman et al. 1997; Nakao et al., 1997). Co-activators and co-repressors lend
additional regulation to the system (Wotton et al., 1999). BMPs signal through Smadsl/5/8 and
the TGF~s typically signal through Smads2/3. BMP-activated Smads, most importantly Smad5,
are necessary for inhibition of myogenic differentiation and osteoblastic induction in C2C12
cells (Lee et al., 2000). Assessment of functional Smad 1/5/8 activation was observed in the
absence and presence of TGFP1 or BMP6 treatment by immunostaining for phospho-Smadl/5/8
expression. Untreated C2C12 myofibers demonstrated low levels of phospho-Smadl/5/8 protein
expression, which were increased in response to both 10 ng/ml TGFP1 and 25 ng/ml BMP6
observed by Western analyses. Additionally, nuclear phospho-Smadl/5/8 protein expression
was induced in response to 100 ng/ml BMP6 and 25 ng/ml TGF~lby immunoflourescence.
Therefore, induction of phospho-Smadl/5/8 protein expression in response to BMP6 or TGFPI
treatment demonstrates functional Smadl/5/8 activation in C2C12 myofibers and this induction
of phospho-Smadl/5/8 is not due to an increase in total Smadl/5/8 protein expression.
Since both BMP6 and TGFP1 induce functional Smadl/5/8 activation in myofibers, it may
be argued that similar downstream effects could result from either ligand treatment. While
BMP6 induces dramatic transdifferentiation of myoblasts, this same effect is not observed in
response to TGP1 treatment. Previous reports have demonstrated that TGFP1 and BMP6 ligand
bind to different Type II receptors and induce different target genes. For example, transfection
of Smad1 and Smad5 into C2C12 myoblasts and NIH3T3 fibroblasts resulted in decreased
myogenin promoter activity in both lines but ALP activity induction was only observed in
C2C12 myoblasts and not in NIH3T3 fibroblasts. These results demonstrated that these