| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Department of Cell and Developmental Biology, University of Pennsylvania, Philadelphia, Pennsylvania 19104, USA
1Correspondence: B1 Anatomy-Chemistry Bldg, Department of Cell and Developmental Biology, University of Pennsylvania, Philadelphia, PA 19104-6058, USA. E-mail: armstroc{at}mail.med.upenn.edu
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONREFERENCES |
|---|
Key Words:
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONREFERENCES |
|---|
2
3)
, Porter recognized two important aspects of this
membrane system. The first is that the SR is simply a specialized
expression of the endoplasmic reticulum (ER) that pervades all cell
types, and the second is that the precise disposition of the SR in
relation to the myofibrils indicates its role in some aspect of the
control of contractility. Both of Porter’s insights are now fully
confirmed. It is known that the SR initially develops as ER containing
the usual garden variety generic ER proteins and that, as the muscle
cell differentiates, it becomes greatly enriched in SR-specific
proteins (4
, 5)
. Three proteins first purified from the SR
are the calcium ATPase or calcium pump protein responsible for pumping
calcium into the lumen of the SR during relaxation (6)
;
calsequestrin, the low-affinity calcium-binding protein that greatly
increases the SR lumen capacity for calcium (7)
; and the
ryanodine receptor (RyR), responsible for calcium release during muscle
activation (see ref 8
for a review). It was later found that all types
of cells contain cell-specific forms and/or analogues of these three
proteins that are responsible for handling calcium. In all cells, these
proteins tend to be grouped together. The SR is thus an extensive,
specialized domain of the muscle cell ER. Conversely, all cells contain
SR-like specialized domains but in much smaller amounts. Some nonmuscle
cells, such as the Purkinje cells of the cerebellum, actually have
extensive SR-like domains containing muscle-specific isoforms of RyRs
and calsequestrin (9
, 10)
.
As Porter, in his inimitable style, wrote: "the precise morphological
relation of the reticulum to the myofibrils... prompts the
suggestion that the system is functionally important in muscle
contraction." Understanding of the specific SR role in controlling
muscle contraction was shaped by critical publications in the exciting
period that immediately preceded and followed Porter’s descriptions of
the SR. The local stimulation experiments of A. F. Huxley,
performed in the late fifties and early sixties, indicated the presence
of a specific pathway allowing spread of the electrical event into the
fiber interior and thus provided a crucial link between surface
depolarization and the rapid activation of centrally located myofibrils
(see ref 11
for a review). The seminal 1959 paper by A. Weber gave the
first proof that calcium ions, now fully recognized as general
intracellular messengers, are responsible for the control of muscle
contraction (12)
. S. Ebashi and A. Weber influenced each
other across continents in constructing clear evidence that the
calcium-sequestering ability of the SR fully accounts for relaxation,
while W. Hasselbach precisely defined the SR calcium pumping action
(see ref 13
for a review).
Having closely followed L. D. Peachey in Porter’s laboratory, I
was fully aware of the search, initiated by A. F. Huxely’s
results, for a link between the muscle fiber plasmalemma and its
interior that would allow rapid spread of the electrical event into the
fiber interior (14)
. In their 1957 paper (3)
,
Porter and Palade had described a repeating structural unit, the triad,
which is located in precise relation to the bands of the myofibrils:
either opposite the Z line or opposite the A-I junction. The triad is
compose of 2 SR cisternae closely facing a central tubule. The location
of the triad (3)
and of its central tubule
(15
16
17)
coincide with the sites at which inward spread of
contractions resulting from local depolarization occurs
(11)
. Using painstaking serial sections,
Andersson-Cedergren (18)
demonstrated that the central
elements of the triad forms a continuous network across the muscle
fibers, hence the name transverse (T) tubules for these components of
the triad. So the eyes of electron microscopists were focused on the T
tubules, and I was primed for the discovery that greeted me toward the
end of my postdoctoral training period in Porter’s laboratory (1963).
I was looking at thin sections of skeletal muscle from guppies that had
been bred in Elisabeth Porter’s tanks and suddenly saw repeated
examples of wide-open mouths of T tubules at the fiber’s edge. It was
easy to get Porter’s attention: I walked into his laboratory at
10:00–11:00 PM, the usual end of his working day, when he could relax,
and showed him my negatives. Porter’s interest was immediately
aroused. My pictures were not to his high standards, so he took one of
my blocks, personally cut thin sections, and took a series of
micrographs, each a work of art (Fig. 1
). Porter’s micrographs constituted most of the illustrations in the
final publication showing that T tubules have the essential features
necessary for their role in spreading surface membrane depolarization
into the fiber interior (19, see also Fig. 1B
). Continuity
of the T tubule lumen with the extracellular spaces was also confirmed
in the same year by the penetration of extracellular space tracers
(20
, 21)
and a fluorescent dye (22)
.
Continuity of T tubules with the surface membrane is at the basis of
the large surface membrane capacitance of muscle fibers.
|
Junctional domains of the SR are associated either with the surface
membrane or with the T tubules (Figs. 1C, G
),
forming well-defined junctions with them called calcium release units
(see 23, 24 for reviews). During excitation-contraction (e-c) coupling,
exterior membranes are initially depolarized, and immediately after the
SR releases calcium into the myofibrillar spaces. The structural
question then becomes: what is the relation between SR and exterior
membranes at these specialized junctional sites that allows translation
of the electrical event into release of calcium from the SR during
muscle activation? Skipping ahead in time by ~25 years, the modern
structural question regarding this step in e-c coupling is: what is the
spatial relationship between proteins of the SR and of the exterior
membranes in the calcium release units, and what can be deduced from
this association? Four proteins of the junctional SR are well
identified: the ryanodine receptors (RyRs), or SR calcium release
channels (8)
; calsequestrin (7
, 25)
; triadin
(26)
; and junctin (27)
. RyRs are
homotetrameric calcium channels constituted of 4 identical subunits
with a large cytoplasmic domain and a combined mass of ~2,000 kDa,
which, when open, allow rapid escape of the SR calcium into the
myofibrillar spaces. The cytoplasmic domains of RyRs are visible in the
electron microscope as feet that connect the junctional SR surface to
exterior membranes (Fig. 1C, D
). RyRs are thus
strategically located for interacting with the surface membrane.
Calsequestrin is a luminal protein of the SR cisternae, located in
proximity of the junctional domains of the SR (Figs. 1B, C
). Its function is to increase the total capacity for
calcium of the SR lumen, while maintaining a relatively high free
calcium concentration, thus allowing a large gradient in ionic calcium
concentration between SR lumen and the myofibrils. Triadin is probably
the protein that links calsequestrin to the SR surface, keeping it in
proximity of the RyRs, and junctin may also have a similar role. Either
one of the latter 2 proteins, or others yet to be identified, is
responsible for keeping calsequestrin immobilized in close proximity to
the calcium release channels. One surface membrane protein is located
at the junctional domains of exterior membranes participating in
calcium release units (23)
: the L-type calcium channel,
also called dihydropyridine receptor (DHPR). DHPRs act as voltage
sensors as well as calcium channels, and their action is necessary for
initiating calcium release from the SR (28
, 29)
.
In skeletal muscle, DHPRs are grouped into tetrads, or groups of 4
components located at the corners of small squares (Figs. 1E, I
). DHPR tetrads form ordered arrays that
face ordered arrays of the SR feet in such a way that each of the 4
DHPRs composing the tetrad appears to be linked to a subunit of an
underlying foot (30
, 31)
. This direct structural relation
of the 2 major calcium release unit proteins helps support the so
called ‘mechanical’ hypothesis of e-c coupling, proposing that
voltage sensors in the T tubule membrane (later shown to be the DHPRs)
sense the depolarization and affect calcium release from the SR by a
direct molecular interaction (32)
. One of the puzzling
structural observations regarding the DHPR-RyR relationship is that
tetrads are associated with alternate feet (Fig. 1M
),
leaving a set of orphan feet (or RyRs) that are not directly linked to
DHPRs. Comparative work on a variety of muscles in vivo and
in vitro indicates that the alternate disposition is the
rule for skeletal muscle, and it does not depend on the presence of 2
types of RyR isoforms.
Recent exploration of the molecular and developmental basis of the
DHPR-RyR relationship is the work of two postdoctoral fellows in my
laboratory (Drs. Hiroaki Takekura and Feliciano Protasi) and involves
collaborations with Drs. P. D. Allen, K. G. Beam, B. E.
Flucher, and H. Takeshima. The first question that we considered is how
the alternate disposition of tetrads is produced during development of
the junctions. This question was explored in collaboration with Dr.
B. E. Flucher using cells of skeletal muscle origin in the BC3H1
cell line, which develop co-localized clusters of DHPRs, triadin and
RyRs at the cell periphery (33)
. Thin sections and
freeze-fracture of these cells show well differentiated calcium release
units containing extensive ordered arrays of feet and tetrads. The
tetrads have an alternate disposition. Many of the junctions have
tetrads with missing elements, and because we find these in early as
well as late days in culture, we assume that many represent junctions
in the process of formation. Interestingly, even when quite incomplete,
arrays of tetrads have the alternate disposition relative to arrays of
feet, indicating that this relationship is intrinsic to the
interactions between the 2 proteins.
The second question is whether DHPRs grouping into tetrads is unique to
skeletal muscle or is present in muscles that seem to have a different
basis for e-c coupling. Cardiac muscle contains isoforms of RyRs and
DHPRs different from those of skeletal muscle, and they differ
functionally in that permeation of calcium through the DHPRs seem to be
a requisite for e-c coupling (34
, 35)
. DHPRs and RyRs of
cardiac muscle are located at sites that appear co-localized at the
light microscope level (Fig. 1F
), just as in skeletal
muscle, and this juxtaposition is achieved early in the differentiation
of the e-c coupling apparatus (36
37
38)
. Freeze-fracture,
however, shows that the DHPRs disposition in cardiac muscle is distinct
from that in skeletal muscle (compare Figs. 1I, K
). The DHPRs are in proximity of the feet but do not seem
to be directly linked to them (Fig. 1N
) so that their
interaction may be indirect, via a transmitter (e.g., calcium).
The above information is directly relevant to understanding 2 mouse
models for the study of e-c coupling. In one model the main (
1)
subunit of the skeletal DHPRs is missing. The muscles do not contract
because of a lack of e-c coupling and develop poorly (hence the term
dysgenic). However, triads containing arrays of feet are formed,
indicating that the presence of RyRs is not needed for the formation of
SR-surface junctions. Clusters of DHPRs (detected by immunolabeling)
and DHPR tetrads (detected by freeze-fracture) are missing in dysgenic
muscle cells, but they are restored by transfection of cultured
dysgenic myotubes with cDNA for the DHPR, demonstrating that tetrads
are composed of DHPRs (31
, 39
, 40)
. The other model is a
targeted null mutation of the skeletal type RyR, which also results in
block of e-c coupling. Muscle fibers lacking RyR, unexpectedly, form
triads (41) (Fig. 1H
). The triads do not have feet, hence
the term dyspedic, but otherwise appear normal in thin sections. A cell
line (1B5) developed from a dyspedic mutation also lacks e-c coupling.
Using these cells, we have shown that despite the absence of feet,
dyspedic SR-surface junctions are formed and contain triadin and DHPRs
(42)
. However, in freeze-fracture the clusters of DHPRs
located at dyspedic junctions do not aggregate into tetrads (Fig. 1L
). Expression of cDNA coding for the skeletal muscle type
1 RyR in 1B5 cells results in the formation of peripherally located RyR
spots (Fig. 1P
) and in the clustering of DHPR in tetrads
(Fig. 1Q
). Hence DHPRs do not need the presence of RyRs to
cluster at SR-surface junctional sites, but they do need an interaction
with skeletal RyRs in order to form tetrads. This also indirectly
confirms that a link between RyR and DHPRs exists in skeletal muscle
fibers. Conversely, the lack of tetrads in cardiac muscle indicates
that either a RyR-DHPR link is not present or that the link is
different from that in skeletal muscle. Both the presence of the
RyR-DHPR link and its absence or difference have profound functional
implications for e-c coupling. K. R. Porter would have approved of
these developments in our understanding of the SR, because they are
based on precisely definable structure-function relationships.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONREFERENCES |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
