Major mesodermal lineages

1
(No Transcript)
2
Major mesodermal lineages
3
Myoblasts arise from embryonic somites
from Gilbert, Developmental Biology
4
Condensation of mesenchymal cells into somites
from Gilbert, Developmental Biology
5
Myoblasts arise from embryonic somites
Somite formation
P
Dermomyotome and
myotome formation
A
(A-to-P direction)
from Rudnicki chapter in Rossant and Tam, Mouse
Development (2002)
6
Myoblasts arise from embryonic somites
Somites condensations of mesoderm cells in the
embryo, organized in pairs along A-P
(rostral-caudal) axis Some of these cells give
rise to muscle
chick embryo
from Bailey et al. (2001) Curr. Opin. Cell Biol.
7
(No Transcript)
8
Myoblasts arise from embryonic somites
MyoD, hi Myf5
Pax3,7
Newly formed somite receives signals from neural
tube (nt),notochord (ntc), surface ectoderm (se)
and lateral plate mesoderm (lpm).
Dermomyotome has differentiated from dorsal
region of somite. Dorsomedial lip (dml) of dm
forms next to neural tube. Ventral region
undergoes epithelial-to-mesenchymal transition to
form sclerotome (scl) which will later give rise
to vertebrae and ribs.
Lineage tracing experiments indicate that dorsal
aspect of somite is fated toform dermomyotome
(blue)
from Bailey et al. (2001) Curr. Opin. Cell Biol.
9
Myoblasts arise from the myotome of embryonic
somites
dm
(express Pax3)
mt
(express MyoD, Myf5)
Forelimb level somite. Myotome continues to form
subjacent to dm. Long-range muscle progenitor
cells (MPCs) delaminate from dm and invade limb
bud. Express Pax3 during migration, then
express MyoD and Myf5 as they differentiate
Cells from dermomyotome translocate through
medial and lateral lips (red) to form parts of
myotome (mt). Lateral lip of dm grows
ventro-laterally and will give rise to muscles of
ventral body wall (gray arrows).
from Bailey et al. (2001) Curr. Opin. Cell Biol.
10
(No Transcript)
11
(No Transcript)
12
Myogenic domains in the early embryo
(Ordahl and Williams, Bioessays 20357, 1998)
13
Different regions of the somites contribute
preferentially to different muscle groups of the
body
(medial portion of somite, gives rise to deep
muscles of back)
NOTE that these MPCs are migratory!
(lateral portion of somite, gives rise to body
wall, limbs, tongue)
from Rossant and Tam, Mouse Development (2002)
14
Skeletal Muscle Development
Steps in muscle differentiation
Subsets of myoblasts migrate to limbs and other
parts of developing animal to form "premuscle
masses" which eventually differentiate into
multinucleated (syncytial) mytotubes of skeletal
muscle
15
Myotube formation in skeletal muscle
As long as particular growth factors
(particularly FGFs) are present, myoblasts will
proliferate without differentiating.
When FGFs are depleted, myoblasts withdraw from
cell cycle, secrete fibronectin onto their ECM,
and bind through it via a specific integrin
(a5b1, their major FN receptor). The integrin-FN
interaction apparently induces a signal which is
critical for instructing the myoblasts to
differentiate further
Myoblasts align into chains via cadherins and
N-CAM (Ca dependent) and fuse (also Ca
dependent) to form multinucleated myotubes.
Fusion into multinucleated (syncytial) mytotubes
is mediated by metalloproteinases called
meltrins (relatives of fertilin, protein
implicated in sperm-egg fusion).
(Muscle fibers are formed from bundles of
myotubes.)
16
Myotube formation in cardiac muscle
Myoblasts do NOT fuse...
So then how do multinucleated myotubes form???
17
Myotube formation in cardiac muscle
Myoblasts do NOT fuse...
... instead, a string of mitoses within a single
myoblast results in a multinucleated myotube.
18
C3H 10T1/2 Cell System for Studying Muscle
Differentiation
myocyte
cell fusion
5-azacytidine
Azamyoblast
Myotube (muscle cell)
C3H 10T1/2 cell
Isolate DNA and transfect into untreated cells
5-AC incorporated into DNA in place of dC, cannot
be methylated
hypomethylated DNA may be more easily activated
5-AC treated 10T1/2 cells form myoblasts,
adipoblasts and chondroblasts
adapted from Lodish et al.
19
C3H 10T1/3 cells
5AC
myocytes (25-50)
adipocytes (7-28)
chondrocytes (1-7)
Classic work by Taylor and Jones Konieczny and
Emerson
20
High percentage of myocyte colonies implied that
undermethylation of one or a few loci suffices to
trigger determination to a particular lineage....
.... and this conclusion suggested a genetic
complementation approach to identifying myogenic
regulators
21
Clue to nature of activation by 5-AC
transfection of chicken cardiac a-actin gene into
(mouse) 10T1/2 cells vs aza-myoblasts
Transfected gene expressed in aza-myoblasts but
not 10T1/2 cells involvement of
regulatory locus and not only activation of
structural markers of myogenesis!
phenotypic (myogenic) conversion
22
MyoD (myogenic determination gene D) was the
first of the myogenic regulatory factors (MRFs)
to be cloned
Cloning scheme took advantage of the 10T1/2 cell
differentiation system
Lassar et al. (1986) derived stable myoblast cell
lines following treatment of 10T1/2 cells with
5-AC ("aza-myoblasts")
Then they transfected high molecular weight DNA
from the aza-myoblasts (plus a neoR gene) into
10T1/2 cells, selected for transformation (G418)
and scored different types of colonies
23
NOTE Aza-myoblast lines do not
differentiate unless maintained in
"differentiation medium" (very low serum
concentration, low growth factor content etc.)
(Idea that cell proliferation (growth) and
differentiation are antagonistic processes)
24
Criteria for scoring myogenic conversion
Concomitant activation of muscle specific markers
(myosin heavy and light chains, a subunit of
acetylcholine receptor, cardiac and skeletal
a-actins)
Fusion of myoblasts into myotubes
25
Lassar obtained myogenic conversion of about 1 in
15,000 transfected colonies (i.e. G418 resistant
colonies)
In contrast, frequency of spontaneous myogenic
conversion was found to be 2.5 x 10 /cell
generation (by plating cells at clonal density)
-8
26
Lassar et al. (Weintraub lab) made argument
against transfection of more than a single locus,
based on observed frequency of myogenic conversion

• Frequency of myogenic conversion was close to
  expected value for transfection of a single locus
  based on genetic complexity of mammalian DNA and
  the amount of transfected DNA taken up by each
  transformed cell
• 10 to 10
• (so 1/15,000 is in the ballpark)
• If two loci were required, one would expect
  frequency of 10 to 10

-4
-3
-8
-6
27
Not everyone accepted this argument and in a
sense was not entirely correct what if there
were two or more loci encoding proteins with
similar if not identical activities in this
assay? (Turned out to be the case.)
functional redundancy among family members
28
Nevertheless, the observed myogenic conversion,
along with very low probability of
transfecting two or more genes into the same
cell, suggested the authors were scoring for
activation (via 5-AC) of a regulator(s) of all
of these genes rather than for transfection of
the structural genes themselves
Statistical argument was important, because
mouse DNA was being transferred into a mouse
cell, so it was not possible to rule out
transfection of the structural genes themselves
29
Concept of the "master control gene"
from Slack and Tosh (2001) Curr. Opin. Genet. Dev.
30
Screen for aza-myoblast-specific genes
Expression Cloning Scheme Resulting in Discovery
of MyoD
Remove mRNAs
32
P-labeled cDNAS
aza-myoblast mRNAs (oligo dT)
Hybridize with excess of mRNAs from untreated C3H
10T1/2 cells
Subtractive Hybridization
(mRNAs common to both cell populations)
discard
(mRNAs specific
to azamyoblasts)
Clone enriched in aza-myoblast-specific cDNA
Screen myoblast cDNA library
use positives as probes
Assay for myogenic activity of MyoD cDNA
Cotransfect with plasmid carrying myoD cDNA and a
G418-resistance plasmid
Select on G418-containing medium
Stain with labeled anti-myosin antibody
C3H 10T1/2 cell
adapted from Lodish et al.
31
Aza-myoblast cDNA library
replica plate
screen Aza-myoblast library with
32
each subtracted P-labeled
cDNA probe
SUBTRACTED
PROBE
Aza-myoblast cDNA
muscle cDNA
untreated 10T1/2 cDNA
a
b
b
c
c
"a" is specific to myoblasts and is
down-regulated in terminally differentiated muscle
32
Aza-myoblast cDNA library
replica plate
SUBTRACTED
PROBE
Aza-myoblast cDNA
muscle cDNA
untreated 10T1/2 cDNA
a
a
b
b
c
c
"a" is activated in myoblasts and continues to be
expressed in differentiated muscle
33
muscle cDNA library
replica plate
SUBTRACTED
PROBE
Aza-myoblast cDNA
muscle cDNA
untreated 10T1/2 cDNA
d
a
a
b
b
c
c
"d" is activated during terminal differentiation
of muscle
34
muscle cDNA library
replica plate
SUBTRACTED
PROBE
Aza-myoblast cDNA
muscle cDNA
untreated 10T1/2 cDNA
d
a
a
b
b
c
c
e
e
e
"e" is not regulated during myogenic
differentiation
35
Colonies of myoD-transfected cells were
indistinguishable from C3H 10T1/2 cells treated
with 5-AC, and both types of cells exhibited
myotube-like properties.
MyoD cDNA could also convert a number of other,
non-muscle cell lines into muscle.
NOTE reminiscent of results of Blau's
heterokaryon studies (phenotypic conversion of a
variety of other cell types into muscle following
fusion with C2C12 cells)
36
MyoD was said to be a "master regulator" that
played a key role in muscle development by virtue
of its ability to regulate an entire program of
differentiation and to convert the phenotype of
many different non-muscle cell types to that of
muscle.
37
Similar approaches used to identify 3 other
myogenic regulatory factors (MRFs)
myogenin
Myf5
MRF4
(Also subtractive hybridization, hybridization
with degenerate probes, low-stringency
hybridization)
38
General structure of MyoD and the other MRFs
(Helix-loop-helix proteins)
Transactivation
DNA Binding
Dimerization
basic
HLH
NH2-
-COOH
(NLS 1)
(NLS 2)
Mef2c-interacting domain
NLS nuclear localization signal
39
Basic-HLH proteins are a large family of
transcription factors, with cell type-specific
members in different tissues (e.g.
erythroid, neural)
40
Not all bHLH proteins are cell-type specific a
subgroup of bHLH proteins (E-proteins) are widely
expressed
41
Cell type-specific basic-HLH proteins are
generally found as heterodimers with E-proteins
(E2A E12 or E47)
MyoD and E2A have similar but not identical
DNA-binding domains
42
DNA binding by helix-loop-helix proteins
basic
HLH
NH2-
-COOH
CANNTG
"E-box" DNA consensus sequence
43
DNA binding by helix-loop-helix proteins
CANNTG
Multiple E boxes are required for MyoD or other
MRFs to stimulate transcription
MyoD binds cooperatively to E-boxes in DNA
44
E-box sequence is not very specific present at
1/256 nucleotides in random sequence (frequent)
C A N N T G



1/4
1/4
1/4
1/4 1/256
What accounts for the specificity of MRF activity?
45
Myogenic specificity requires interaction of MRFs
such as MyoD with other transcription factors
combinatorial association with muscle
enhancer-binding factors (MEFs)
46
Structure of MEFs
NH2-
-COOH
(transactivation)
MADS
MEF
myogenin-interacting domain
MEFs are expressed in muscle and several other
tissues, bind a DNA sequence (MEF-box) DISTINCT
from that of the E-box
MEFs cannot induce myogenic conversion on their
own but enhance the ability of MRFs to do so
(physical interaction with MRF-E2A)
bHLH region of MRFs required for interaction with
MEFs but is buried in major groove of DNA --
perhaps functions to confer particular
conformation to other region(s) of MRF-E2A which
then interact with MEFs
47
If MEFs don't induce myogenesis on their own, how
does this transgene work?
lacZ
3X MEF sites
Expression detected in skeletal, cardiac, and
smooth muscle cells
48
MRF/E2A - MEF interactions
MEF dimer
E2A-MRF dimer
OR
E2A-MRF dimer
MEF dimer
OR
Lodish et al.
MRF-E2A-MEF complex can assemble in the
promoter/enhancer region of muscle-specific genes
containing an E-box, a MEF box, or both
AND
Promoters/enhancers of muscle-specific genes may
containing E-boxes, MEF boxes, or both
49
MyoD and related proteins can turn nearly any
cell into a muscle cell (in vitro), so synthesis
of MRFs must be inhibited at numerous steps to
allow other cell types to be produced!
Id proteins (Inhibitor of differentiation) lack
the basic region adjacent to the HLH domain that
is essential for specific DNA binding in MRFs and
other bHLH proteins. They prevent formation of
MyoD-E12/E47 heterodimers.
In proliferating myoblasts, the MyoD gene might
be activated (upregulated) by a MyoD homodimer or
by a heterodimer of MyoD with another protein
which is Id-insensitive.
50
A second type of myogenic inhibitor is expressed
in cells of the sclerotome
I-mf (Inhibitor of MyoD family) blocks the
nuclear localization signal of MyoD and thereby
sequesters it in the cytoplasm, where it is
unavailable to activate muscle-specific genes.
It also interferes with DNA binding by MyoD.
Chen et al. (1996)
51
I-mf inhibits DNA binding by MRFs
MRFs translated in vitro with E12 and incubated
together without I-mf or with increasing
concentrations of I-mf, then analyzed by EMSA
NLSMyf5 Myf5 fused with NLS from SV40 T-Ag
"nuclear rescue"
heterodimer complex



homodimer complex
Chen et al. (1996)
52
Neurogenesis requires regulatory proteins
analogous to bHLH myogenic proteins
Vertebrates neurogenin, NeuroD
Lodish et al.
53
Gene targeting approaches have been used to
determine when MRFs and MEFs function during
development in vivo
54
(No Transcript)
55
Null mutations in MyoD or myf5 result in delayed
development of specific subsets of skeletal
muscle, but muscles do form and animals are
viable
56
Myf5 and MyoD have overlapping functions
Functional redundancy results in more robust
developmental program and may provide more
flexibility in ability of cells in different
regions of the embryo to respond to extracellular
signals which regulate myogenesis
57
Model for genetic control of mammalian skeletal
muscle formation
Model based on KO expts in mice and
loss-of-function mutations in flies, consistent
with observation that myoblasts formed from
5AC-treated 10T1/2 cells express both MyoD and
Myf5 but that, as they begin to fuse into
myotubes, they express only myogenin
Somite
Myoblast determination
MRF4 is expressed later in development than MyoD,
Myf5 or myogenin and may play a role in
maintenance of differentiated state of muscle
cells
Myoblast
Differentiation
Myotube
Lodish et al.
58
Mice express multiple MEF genes
In Drosophila (one MEF gene) mutants carrying
null mutations in MEF, myoblasts form but do not
differentiate normally MEFS required for
differentiation but not determination of myoblasts
59
Phenotype
Genotype
myoD knockout
myf5 wild type
Normal
myf5 knockout
No myoblasts
coding sequence of Myf5 are replaced by those of
myogenin
Few myoblasts
myoD transcription-control region
myf5 transcription-control region
MyoD coding sequence
Myf5 coding sequence
Myogenin coding sequence
Lodish et al.
60
Myogenin and Myf5 are not biochemically
equivalent
Myf5 (and MyoD) both remodel chromatin more
effectively than does myogenin (chromatin
remodeling critical for normal cell
differentiation)
61
Results of knockout/knockin experiments
consistent with idea that different MRFs arose by
gene duplication and subsequent divergence via
mutation
from Rossant and Tam, Mouse Development (2002)
62
MRFs and MEFs associate with HATs and HDACs to
control activation and repression, respectively,
of the muscle differentiation program
Differentiation initiated
Model for roles of HATs and HDACs in control of
muscle differentiation
Undifferentiated myoblasts MRF (e.g. MyoD), MEF
activities repressed by various HDACs
During differentiation HDACs dissociate from
MRFs, MEFs -- coactivators with HAT activity
(e.g. 300/PCAF) associate with MRFs, MEFs,
resulting in activation of muscle program and
myotube formation
GRIP steroid receptor coactivator with
intrinsic HAT activity, also recruits p300 and
PCAF to target genes
from McKinsey et al. (2001) Curr. Opin. Genet.
Develop.
63
MyoD and HATs in muscle differentiation

In differentiating muscle, the HAT protein PCAF
but not p300 is critical for activation of MyoD
and muscle differentiation

p300 regulates MyoD by recruiting PCAF (serves as
molecular bridge)

MyoD is acetylated at three lysine residues by
PCAF (more efficient) or p300 in vitro
non-acetylatable mutants of MyoD have impaired
DNA binding, transcriptional and myogenic
activity.

Not yet clear relative contributions of MyoD
acetylation versus MyoD-mediated recruitment of
HATs to target chromatin (and subsequent
relaxation of chromatin)

Does GRIP associate with all MRFs and MEF2 family
members? Does stimulatory effect of GRIP require
its intrinsic HAT activity or is its ability to
recruit other coactivators more important? Not
yet known.
64
Regulation of HATs in muscle differentiation

Possible role for additional coactivators/represso
rs that regulate myogenesis by blocking
interactions between MRFs/MEFs and HATs/HDACs
(analogy with Cabin repressor in T cells which
blocks MEF2D activity by binding to its MADs box
and blocking interactions with p300)

Possible role for additional coactivators/represso
rs that regulate myogenesis by binding to and
repressing catalytic domains of p300 and PCAF
(e.g. the bHLH protein Twist, a myogenic
inhibitor)
65
Other possible mechanisms for regulating muscle
differentiation via chromatin structure

SWI/SNF complexes possess ATP-dependent
remodeling activity dominant negative SWI/SNF
enzymes block MyoD-dependent muscle
differentiation

Histone methylation controls gene silencing,
could provide mechanism for irreversible
repression of growth-associated genes in
terminally differentiated muscle cells
66
Complex signaling interactions regulate
myogenesis during development
Myotome induced by at least two distinct sets of
signals
Cells migrating from medial somite induced by
factors from NT (Wnt1, 3a Shh)
Cells migrating from lateral somite induced by
factors from epidermis (Wnts) and LPM (BMP4)
Wnt, BMP, ? other signals induce expression of
myogenic transcription factors
Other, inhibitory signals prevent a signal from
affecting an inappropriate group of cells
Shh activates MT and ST development, inhibits
BMP4 signal from LPM from extending medially and
ventrally -- prevents conversion of ST into muscle
Noggin produced by most medial portion of DM
prevents BMP4 (NT) from giving these cells
migratory characteristics of hypaxial
(ventro-lateral) muscle
from Rossant and Tam, Mouse Development (2002)
67
Myoblasts can and must respond to extracellular
signals that regulate proliferation (cell number)
and migration (location of muscle)
Differentiated muscle (myotube) cannot respond to
such signals
Cell cycle proteins (e.g. cyclins and Cdks)
influence transition from determined (myoblast)
to differentiated (myotube) state
Antagonism between positive and negative
regulators of G1 progression plays important role
in regulating myogenesis
68
(No Transcript)
69
MyoD is activated by different mechanisms in
different regions of the somite
Lateral (forms hypaxial muscles) extracellular
signals induce Pax3 (a homeodomain transcription
factor)
Pax3 activates MyoD and Myf5 (in absence of other
inhibitory transcription factors, e.g. those
expressed in sclerotome)
Medial (forms epaxial muscles) MyoD expressed
at low levels, then autoregulates to increase its
own expression
70
(No Transcript)
71
MPCs in adult skeletal tissue
modified from Rossant and Tam, Mouse Development
(2002)
72
Common regulators control formation of MPCs in
embryo and adult
muscle "side population" (FACS)
(Msx-1)
from Bailey et al. (2001) Curr. Opin. Cell Biol.
MPCs in lateral DM express Pax-3/-7 and
differentiate later than those in medial DM
extended proliferation required for formation of
sufficient number of MPCs to populate trunk and
limbs. Adult skeletal muscle is thought to
contain two distinct populations of MPCs,
satellite cells and MuSP ("side population")
cells.
73
The homeobox gene msx1 promotes the
de-differentiation of multinucleated myotubes
into mononucleated, proliferating cells that do
not express myogenic genes. Pools of these
mononuclear cells exhibit considerable plasticity
and can differentiate into cells that express
either myogenic, chondrogenic (cartilage),
osteogenic (bone) or adipogenic (fat) markers.
Clonal analyses will be required to determine
definitively whether or not all of these lineages
can arise from a single progenitor cell.
(see Odelberg et al., 2000 and Bailey et al.,
2001)
74
msx1 prevents differentiation of myoblasts to
myotubes (conditional, TetR regulated
Msx1-retrovirus vector introduced into C2C12
myoblasts)
msx1 reverses expression of muscle specific
proteins in mouse myotubes
msx1 induces disassembly of (cytokinesis in)
myotubes to mononucleated cells MyoD and cell
cycle inhibitor p21 are downregulated
75
msx1 reverses expression of muscle specific
proteins in mouse myotubes
Odelberg et al. 2000
76
Pools of mononucleated cells derived from
dedifferentiated mouse myotubes are multipotent
and can redifferentiate
collagen II
collagen X
alk. phos.
control
chondrogenic markers collagen II, X
osteogenic marker alkaline phosphatase
msx1
adipogenic markers Oil Red O, Nile Red stain
Oil Red O
Nile Red
myogenin
myogenic marker myogenin
control
msx1
Odelberg et al. 2000
77
msx1 induces transdetermination of C2C12 myoblasts
collagen II
collagen X
Alcian Blue
control
chondrogenic markers Alcian Blue stain, collagen
II, X
msx1
adipogenic marker Oil Red O
Oil Red O
alk. phos.
myogenin
osteogenic marker alkaline phosphatase
control
myogenic marker myogenin
msx1
Odelberg et al. 2000
78
Hypothetical scheme for control of
dedifferentiation of limb blastema during
regeneration in amphibians
Hughes 2001