5
1. INTRODUCTION
1.1 Skeletal muscle fiber properties
In vertebrates three different muscular tissues can be distinguished: smooth, cardiac
and skeletal. Smooth muscle surrounds internal organs and major blood vessels, is
innervated by the autonomous nervous system and it is made of mono-nucleated cells
(called myocytes). The cells of cardiac muscle are mono-nucleated, and together with
skeletal muscles form the complex of striated muscles.
The functional units of skeletal muscles are the muscle fibers, long cylindrical
multinucleated, with about 100 nuclei cells, 10-50 µm of diameter, 1-10 mm of length in
rat muscles. The great majority of the cytoplasm of these cells is occupied by fascicles of
longitudinal filaments called myofibrils, subdivided in light and dark bands, called I band
and A band respectively. The basic structural and functional unit of skeletal muscle is
called sarcomer: this repeated element has a length of 2 µm. Adjacent sarcomeric units are
perfectly aligned, giving rise to the characteristic striated morphology that can be seen with
the microscope. Each sarcomer contains two kinds of filaments: thin filaments, made of
actin (that constitutes the light I band), thick filaments, made of myosin (dark A band).
Contraction speed is related to the ATPase activity of myosin, and to the enzymes of the
sarcoplasmic reticulum (SR) that sequester calcium (Ca
2+
). Myosins are composed of three
couple of subunits: two heavy chains (MyHC), two light chains (MLC) and two essential
or alkaline chains. The two C-terminal domains of the MyHC subunits are coiled to form
an α-helix tails, responsible for the association of many other MyHC dimers in a single
filament. ATPasic activity and actin binding site are localized within the globular region of
the protein (head).
1.1.1 Myofiber diversity
Muscle fibers vary considerably with respect to their morphological, biochemical
and physiological properties, enabling different muscles to fulfill a variety of functions,
from maintaining the body posture to performing a wide range of movements and motions.
Mammalian skeletal muscles are composed of two major fiber-types (I and II),
basically distinguished on the basis of the myosin heavy chain isoform (MyHC) that they
express, and differ in terms of size, metabolism, and contractile properties. MyHCs
6
determine the contractile properties of the fibers, while fiber metabolism determines their
fatiguability. Slow (type I) fibers are characterised by slow velocity of shortening and an
high content in mitochondria, exhibiting a predominantly oxidative metabolism and
resistance to fatigue. They are recruited for sustained, tonic contraction events. Fast (type
II) fibers fatigue and contract rapidly and thus have low resistance to fatigue. Type II fibers
are further grouped into three subtypes, IIA (or 2A), IIX/D (or 2X/D), IIB (or 2B) differing
with respect to their mitochondrial content. IIB fibers are the most glycolityc and most
fatigable ones while IIA, being rich in mitochodria, are more oxidative and relatively
slower. IIX/D fibers have intermediate properties between IIA and IIB fibers. Type II-fast
fibers are required for sudden rapid movements.
Although fibers are roughly divided into these general categories, a wide spectrum
of fiber-types exists, from the specialized extremes to the intermediate fiber-types,
expressing more than one MyHC isoform (e.g. I and IIA, or IIA and IIX, or IIX and IIB).
In addition to the variety of muscle fiber intrinsic composition, the pattern of expression of
the MyHC varies between species.
1.1.2 Myofiber plasticity
Myofiber composition is determined by developmental cues and neuronal activity.
During embryogenesis, primary myofibers can mature into slow or fast fiber independent
of innervation, with a mechanism not yet identified. When primary myofibers mature into
secondary myofibers, embryonic MyHC isoforms are progressively replaced by adult IIA,
IIX, and IIB isoforms (De Nardi et al., 1993). At this point innervation is required for
skeletal muscle growth and survival, and, in addition, controls fiber-type properties.
Skeletal muscle owns a remarkable capacity to remodel in response to specific
environmental and physiological stimuli. This powerful adaptability results from specific
patterns of nerve activity and the diversity of muscle fiber-types. Specific impulse patterns
delivered by motor neurons have been known for years to exert phenotypic changes on
muscle fibers which they innervate, thus remodelling myofibers and maintaining their size
and function. The determinant role of nerve activity in switching muscle fiber properties
has been demonstrated by a surgical switch of the innervation between the slow soleus and
the fast EDL muscles (cross-innervation). After 11-15 months soleus acquires most
properties of a fast muscle and EDL of a slow muscle (Barany and Close, 1971; Vrbova,
7
1963). These data indicate that, even if innervation has a great influence, probably there
are some intrinsic differences, related to different embryonic lineages (Pette and Staron,
1997). Subsequently, electrical stimulation experiments corroborated the cross-innervation
data, demonstrating that impulse patterns, applied directly on the nerve or on denervated
muscles, and mimicking the firing pattern of slow and fast motor neurons, could induce
changes in muscle fibe-type composition (Pette and Vrbova, 1992; Pette, 2001). A firing
pattern mimicking fast motor neuron activity can cause a slow-to-fast switch in the
direction I → IIA → IIX → IIB whereas an opposite switch, fast-to-slow, takes place using
a firing patter resembling that one of slow motor neurons and it follows the obligatory
sequence IIB → IIX → IIA → I. However, electrostimulation usually produces an
incomplete transformation of the muscle fiber-type because the range of adaptability is
limited by intrinsic differences between muscles, thus a fast muscle can adapt in the range
IIB ↔ IIX ↔ IIA and slow muscle in the range I ↔ IIA ↔ IIB (Ausoni et al., 1990). The
limit of these experimental approaches is that changes in protein (myofibrillar and SR
proteins as well as metabolic enzymes) expression in response to chronic electrical
stimulation require weeks or months or even years (Pette and Staron, 1997), while changes
at the transcript levels occur more rapidly (Huber and Pette, 1996). Also during
pathological conditions complete switches in muscle fiber-type appear, e.g. long term
spinal cord injury involves disappearance of type I fibers (Grimby et al., 1976). In fact
during denervation a “default” program takes place in both fast and slow muscles, giving
rise to a slow-to-fast switch (Butler-Browne et al., 1982), in addition to muscle atrophy, as
a consequence of the lack of electrical stimulation (Spector et al., 1985 a and b).
1.2 Ca
2+
-dependent events in skeletal muscle
1.2.1 Excitation-contraction coupling
Excitation-contraction coupling is the process whereby membrane depolarization
triggers force production (Fig. 1). Membrane depolarization, triggered by a neuronal
stimulus, activates Ca
v
1.1 in the plasma membrane of skeletal muscle, which are voltage-
gated L-type Ca
2+
channels, also known as dihydropyridine receptors (DHPRs), because
dihydropyridine is a blocker of this channel (Glossmann et al., 1985). Many L-type
channels are situated on t-tubules, deep invaginations of the plasma membrane (Fig. 1),
8
perpendicular to the long axis which allow depolarization to quickly reach the interior of
the cell (Curtis and Catterall, 1984). L-type channels are responsible for the excitation-
contraction coupling, due to the interaction of their α
1
subunit with the ryanodine receptor
RyR (in skeletal muscle the RyR1 isoform), thus connecting the tubular to the SR
membranes and forming the junctions between the t-tubule and SR (t-SR junctions). In
particular, every RyR1 is associated with a tetrade of four Ca
v
1.1 channel (Zalk et al.,
2007).
Fig. 1. Skeletal muscle excitation-contraction and excitation-transcription coupling. Activation of
muscle contraction results in depolarization of the plasma membrane and the transverse tubule (t-tubule)
system. The t-tubule system carries the signal to the interior of the fibre where the voltage-sensing
dihydropyridine receptor (DHPR) detects the change in membrane potential and transmits the signal to the
calcium-release channel, also known as the ryanodine receptor (RYR). Release of calcium from the internal
stores (sarcoplasmic reticulum; SR) into the myoplasm results in actin–myosin interaction, fibre shortening
and force production. Skeletal muscle calcium levels are also elevated by ligand-mediated activation of L-
type calcium channels by signalling molecules such as insulin-like growth factor-I, which result in influx of
calcium via the plasma membrane. Both sources of intracellular calcium are thought to play a role in
excitation–transcription coupling. [Ca
2+
]i, intracellular free calcium ion concentrations; CaM, calmodulin
(from Chin, 2004).
RyR1 is the major channel for Ca
2+
release from the SR, once membrane
depolarization reaches t-tubules. It is believed that an electromechanical coupling exists,
which converts an electrical signal into structural changes first in the Ca
v
1.1 channels.
These changes in turn induce a structural alteration in the RyR1s, which finally triggers the
opening of the RyR1s (Schneider and Chandler, 1987; Rios and Pizarro, 1991). Ca
2+
released from RyR1s binds to diverse targets, including troponin C, which triggers muscle
contraction through the actin-myosin system. Following contraction, Ca
2+
uptake from the
9
cytoplasm into the SR is carried out by the SR Ca
2+
-ATP-ase (SERCA) pump (MacLennan
et al., 1985).
1.2.2. Excitation-transcription coupling
The process whereby a given electrical impulse, causing plasma membrane
depolarization, leads to specific gene activation or inactivation programs is called
excitation-transcription coupling, in analogy with excitation-contraction coupling. The
molecular mechanisms responsible for coupling nerve activity to transcriptional changes
have not been fully defined. Motor neuron activity results in a marked elevation of
intracellular Ca
2+
and this increase seems to be the primary regulator of altered gene
expression in skeletal muscle (Chin et al., 2004). Type I fibers are stimulated by
frequencies of 10-30 Hz, whereas type II fibers by 80-150 Hz (Hennig and Lomo, 1985).
The frequency and the duration of the stimulus determine the amplitude and duration of the
Ca
2+
transients, which ultimately regulates the force produced by the muscle. It is thought
that the amplitude and duration of the Ca
2+
transient are decoded by muscle cells through
Ca
2+
-dependent molecular transducers (McCullagh et al., 2004; Serrano et al., 2001; see
paragraph 1.3.2), triggering different gene expression programs depending on the specific
Ca
2+
transient sensed by the fiber.
Ca
2+
acts as a second messenger in skeletal muscle and the downstream pathways,
which convert this signal into changes in protein expression, include the Ca
2+
-dependent
phosphatase calcineurin (Cn) (Chin et al., 1998; McCullagh et al., 2004; Serrano et al.,
2001), Ca
2+
/calmodulin-dependent kinases (CaMK) (Chin, 2004; Ojuka, 2003) and Ca
2+
-
dependent protein kinase C (PKC) (Rose et al., 2004).
Our attention has focused on Cn, which has emerged as a prominent pathway in the
excitation-transcription coupling process in the last few years (McCullagh et al., 2004;
Serrano et al., 2001; see paragraph 1.3.2), and on CaMKII pathways, another important
Ca
2+
-dependent pathway in skeletal muscle (Rose et al., 2006). We have also studied their
potential cooperation in the regulation of activity-dependent gene expression. Both
pathways are Ca
2+
- and calmodulin-dependent. Calmodulin (CaM) means CALcium
MODULated proteIN, in fact this protein senses Ca
2+
intracellular level, and once activated
by Ca
2+
, it activates in turn a number of target enzymes, including calcineurin and CaMK.
10
1.3 Calcineurin (Cn)
Cn is the only serine/threonine protein phosphatase that is Ca
2+
and CaM-dependent
but this property makes it one of the most common intracellular transducers of Ca
2+
signalling pathways. Its name derives from its ability to bind Ca
2+
and the predominant
localization in nervous tissue.
Cn is an heterodimer consisting of a 58-64 kDa catalytic subunit, called calcineurin
A (CnA; Fig. 2), and a 19 kDa regulatory subunit, named calcineurin B (CnB). Mammalian
CnA exists in three isoforms, CnA α, CnA β and CnA γ, while there are two isoforms of the
regulatory subunits, CnB1 and CnB2. CnA γ is expressed only in testis and brain, whereas
CnA α and CnA β are ubiquitous. CnB1 is expressed in the same tissues that express also
CnA α and CnA β, while CnB2 is expressed mainly in testis and also in brain (Kincaid,
1993).
Fig. 2. Functional domain organization of calcineurin A. A. Schematic representation of the three
mammalian isoforms of calcineurin A. The variable regions and 10-amino acid insert, resulting from
alternative splicing, in the α and β isoforms of mammalian calcineurin A are shown in black. B. Extended
representation of the regulatory domain; the amino acid sequences of the calcineurin B-binding helix, the
calmodulin-binding domain, and the auto inhibitory peptide are boxed. The numbering of the amino acids is
that of calcineurin A α. Residues critical for interaction with cyclophilin and FKBP are represented by white
on black letters (from Klee et al, 1998).
Cn was first purified in brain, being highly concentrated in neurons (Klee et al.,
1979), but it is also broadly distributed in other tissues. Regardless of its source, CnA is
always tightly bound to CnB (even in the presence of only nanomolar concentrations of
11
Ca
2+
; Klee et al., 1988). This two-subunit structure is unique among phosphatases and it is
conserved from yeast to man (Kincaid, 1993), being fundamental for the activity of Cn.
As mentioned before, Cn activity is controlled by Ca
2+
and CaM. CnB has a highly
conserved structure which is similar to that of CaM, with two lobes, each composed of two
adjacent Ca
2+
-binding loops connected by a flexible helix linker (Anglister et al., 1994).
With regard to sequence homology, CnB is similar to “EF-hand” Ca
2+
-binding protein.
CnB has its one high-affinity site occupied by Ca
2+
, even if Ca
2+
is present at nanomolar
concentrations (10
-7
M), but the enzyme is inactive. When Ca
2+
concentration rises
between 0.5 to 1 µM, the three low-affinity sites of CnB are also bound to Ca
2+
, and Cn is
weakly activated. This activity can be enhanced of about 20-folds by putting a equimolar
concentration of CaM (Klee et al., 1998); in fact, the higher the amount of CaM, the lower
the Ca
2+
concentration needed to fully activate Cn. Conversely the higher the Ca
2+
concentration, the lower the CaM necessary for the activation of Cn. In other words, the
same stimulus in terms of Ca
2+
concentration can produce different responses in
dependence of the CaM concentration.
Besides containing a CnB- and a CaM-binding site, CnA regulatory domain has an
autoinhibitory domain (AID; Fig. 2B), which inhibits the phosphatase activity (Hashimoto
et al., 1990) if Ca
2+
is not bound to CnB and CaM.
One of the hypothesis about the activation mechanism of Cn, suggests that the
binding between CaM and CnA causes conformational changes in CnA, so that a
displacement of the AID of CnA from the catalytic site is produced (Klee et al., 1988). The
mechanism through which CnB, binding to Ca
2+
, activates Cn is still not clear. It is well
established that CnB must be present to have the activation of the enzyme. Moreover, Ca
2+
must be bound to high-affinity sites of CnB, although the two subunits dissociate. Ca
2+
binding to low-affinity sites is apparently responsible not only for the CaM-independent
Cn activity (low), but it is also involved in the CaM-dependent Cn activity (high).
The activity of Cn is controlled not only by Ca
2+
and CaM but also by several Cn
inhibitors, which have been identified during the past few years. The function of Cn has
remained unclear until it was identified as the target of the immunosuppressants
cyclosporine A and FK506, inhibiting its catalytic activity (Kiani et al., 2000).
Physiological Cn inhibitors have been identified, which control the duration and intensity
of Cn activity. Among such endogenous inhibitory proteins are calcineurin-binding protein
12
1 (CABIN1, also known as CAIN), the A-kinase anchor protein AKAP79 (also known as
AKAP5) and members of the Down’s syndrome critical region (DSCR)/modulatory
calcineurin-interacting protein (MCIP) family of Cn inhibitors, which are also known as
calcipressins (Crabtree, 2001).
Cn is known to play an important role in the regulation of gene expression in
mammalian cells (Klee et al., 1998, Wang et al., 1999; Sussman et al., 1998; De La Pompa
et al., 1998; Graef et al., 2001). A well characterized example of Cn transcriptional control
is its role in the maintenance and induction of the slow fiber program in skeletal muscles
(Chin et al., 1998; Serrano et al., 2001; Meissner et al., 2006; Wu et al., 2000). McCullagh
et al. have shown that the latter effect of Cn is mediated by the nuclear factor of activated
T cells (NFAT; see paragraph 1.3.2), the main downstream substrate of Cn (see next
paragraph). Cn-NFAT pathway and the mechanism by which Cn is activated have been
extensively investigated in the immune system, in particular in lymphocytes. Indeed,
NFAT has been initially identified as an inducible nuclear factor that could bind to the
interleukin-2 promoter (IL-2) in activated T cells (Shaw et al., 1988). Briefly, upon binding
of T cell receptor (TCR) ligand, the intracellular concentration of Ca
2+
increases, thus
activating Cn, which in turn desphosphorylates NFAT and induces its translocation to the
nucleus where it activates the transcription of genes coding for cytokines, chemokines and
cell surface receptors that promote a productive immune response (Rao et al., 1997). The
ability of NFAT to interact with different transcriptional partners determines the activation
or deactivation of specific gene expression programs (Macian, 2005).
1.3.1 NFAT family
NFAT is a multigene family gene composed of five distinct elements, NFATc1,
NFATc2, NFATc3, NFATc4 and NFAT5. Even if the latter has extensive homology with
the other NFATs, it is not activated by Cn (Lopez-Rodriguez et al., 1999) and will not be
further considered in this thesis work. The length of all NFAT proteins is comprised
between 700 and about 1000 amino acids, with different splice variants. Each of the Cn-
activated NFAT proteins share a similar structure, roughly divided in three regions: a
moderately conserved N-terminal region (also known as NFAT homology region, or
NHR), which is the regulatory domain that controls NFAT cellular distribution and
Ricevi informazioni sui nostri servizi, sulle offerte e non perdere news e consigli su università e lavoro.
