Respiratory Physiology & Neurobiology 137 (2003) 125 /140

www.elsevier.com/locate/resphysiol

The biophysics of asthmatic airway smooth muscle

Newman L. Stephens *, Weilong Li, He Jiang, H. Unruh, Xuefei Ma
Department of Physiology, Faculty of Medicine, University of Manitoba, 730 William Ave, Winnipeg, MB, Canada R3E 3J7

Accepted 2 April 2003

Abstract

It is clear that significant advances have been made in the understanding of the physiology, biochemistry and
molecular biology of airway smooth muscle (ASM) contraction and how the knowledge obtained from these
approaches may be used to elucidate the pathogenesis of asthma. Not to belittle other theories of smooth muscle
contraction extant in the field, perhaps the most outstanding development has been the formulation of plasticity theory.
This may radically alter our understanding of smooth muscle contraction. Its message is that while shortening velocity
and capacity are linear functions of length, active force is length independent. These changes are explained by the ability
of thick filament protein to depolymerize at short lengths and to increase numbers of contractile units in series at
lengths greater than optimal length or Lref. Other advances are represented by the report that the major part of ASM
shortening is complete within the initial first 20% of contraction time, that the nature and history of loading determine
the extent of shortening and that these findings can be explained by the finding that the crossbridges are cycling four
times faster than in the remaining time. Another unexpected finding is that late in the course of isotonic relaxation the
muscle undergoes spontaneous activation which delays relaxation and smoothes it out; speculatively this could
minimize turbulence of airflow. On the applied front evidence now shows the shortening ability of bronchial smooth
muscle of human subjects of asthma is significantly increased. Measurements also indicate that increased smooth
muscle myosin light chain kinase content, via increased actomyosin ATPase activity could be responsible for the
changes in contractility.
# 2003 Elsevier B.V. All rights reserved.

Keywords: Airway, smooth muscle; Disease, asthma; Mammals, dog, humans; Muscle, smooth, airway, plasticity

1. Introduction has been disputed (Otis, 1983). However, there has

For a considerable period of time, the impor-
tance of airway smooth muscle (ASM) in the
regulation and regional distribution of ventilation
* Corresponding author. Tel.:
/1-204-789-3526; fax:
/1-
204-789-3941.
E-mail address: nstephe@ms.umanitoba.ca (N.L. Stephens).
1569-9048/03/$ - see front matter # 2003 Elsevier B.V. All rights reserved.
doi:10.1016/S1569-9048(03)00142-3
emerged a consensus that changes of smooth
muscle contractility, as described below, are now
thought to contribute importantly to excessive
airway narrowing in asthma. In this context it
was initially though that the increased narrowing
of the airways was not due to altered contractility
of the AMS but to release of agonists such as
histamine, leukotrienes and acetylcholine in in-
creased amounts from the sensitized mast cell
126 N.L. Stephens et al. / Respiratory Physiology & Neurobiology 137 (2003) 125 /140
during the course of the antigen /antibody reac-
tion. In a homely analogy the horse (ASM) itself
was not altered; it was just being whipped (acet-
ylcholine, histamine) harder. This is an important
point because at the time it was made the issue of
primary alteration in contractility had been dis-
missed.
However, it has been reported that the contrac-
tility of ovalbumin-sensitized, isolated canine tra-
cheal smooth muscle (STSM) was increased; both
maximum capacity (DLmax) and velocity (Vo) of
shortening were increased to a level that could
account for the bronchospasm of asthma (Anto-
nissen et al., 1979; Jiang and Stephens, 1992).
Maximum force development (Po) was however
unchanged. Others have reported similar findings
(Mitchell et al., 1994).
Maximum isometric force (Po) is the parameter
most often studied in animal models of asthma.
Under these conditions what is being measured is
closely correlated to changes of muscle stiffness. It
cannot provide any information regarding ASM
shortening which is the relevant parameter for
bronchospasm.
As a background for the mechanisms underlying
these changes certain relevant aspects of the basic
physiology and biochemistry of ASM smooth
muscle, that have developed in recent times, will
be reviewed.
2. Smooth muscle physiology
2.1. Latch bridges
Twenty years ago a major advance was made by
Murphy’s (Dillon et al., 1981)group which re-
ported that smooth muscle cycling cross bridges
were not kinetically homogeneous during contrac-
tion. In early contraction they were referred to as
normally cycling cross-bridges (NBR); they cycled
at a relatively faster rate (though 30 /50-fold
slower than in skeletal muscle) than bridges cycling
later in the contraction. The latter were called
latch bridges (LBR). On the basis of bioenergetic
studies (Siegman et al., 1997) they were re-named
slowly cycling cross bridges. LBRs cycled at one
quarter the velocity of NBR. They developed as
the result of partial dephosphorylation of the 20
kDa myosin light chain (MLC20), and were
responsible for most of the force development of
the muscle. Their economy of energy utilization
was also very high.
Thus the detection of time-dependent functional
heterogeneity in cross-bridges in smooth muscle
contraction was recognized as a unique feature.
The most recent advance made in the field of
smooth muscle contractility is that made by
Pratusevich et al. (1995). This group has developed
a radically different model of smooth muscle
contraction in which the most important feature
is that active isometric force is independent of
muscle length over a very wide range of lengths
extending from 0.5 to 1.5 Lo where Lo is the
muscle’s optimal length. From this one must
conclude that the Frank/Starling hypothesis
(Frank, 1904) does not apply to smooth muscle.
The implications of this for airway caliber regula-
tion could be quite considerable. Plasticity theory
will be dealt with below.
Because plasticity theory is, at this time just
that, and awaits universal acceptance, much of our
treatment of smooth muscle mechanics incorpo-
rates the conventional view (Dillon et al., 1981)
and the modern theories of smooth muscle con-
traction extant in the field (Chan et al., 2000;
Fredberg et al., 1999; Gunst et al., 1995; Morano
et al., 2000) the most important of which, in our
view is plasticity theory.
2.2. Importance of early phase of shortening
We have confirmed and extended the findings
reported in the preceding paragraph. To bring
home the importance of what happens in the early
phase of shortening, Fig. 1 shows a record of
length change versus time in a lightly loaded
muscle contracting isotonically. Clearly 75% of
the total shortening is complete within 2 sec; the
muscle’s contraction time is 10 sec (Stephens and
Jiang, 1995; Stephens et al., 1999). We feel this is
an important finding as it affords new insight into
the shortening process in smooth muscle. The
preponderant early shortening is due to the
operation of normally cycling crossbridges in this
early phase. These are almost four times as rapid
 N.L. Stephens et al. / Respiratory Physiology & Neurobiology 137 (2003) 125 /140
Fig. 1. Isotonic shortening */with load/0.1 Po. Record of
shortening versus time elicited from a muscle shortening
isotonically with a light preload. It is clear that more than
75% of total shortening is complete within the first 2 sec of
stimulation. DLmax /maximal shortening.
as bridges activated later (the so-called latch or
slowly cycling crossbridges).
Fig. 2 shows a plot of the difference in short-
ening between control (CTSM) and sensitized
(STSM) canine tracheal smooth muscle. It is
apparent that 90% of the increase in shortening
is complete within 1.5 s. Therefore bronchospasm,
Fig. 2. Plot of difference in isotonic shortening between control
(con) and sensitized (sen) strips of ASM. CASM, control of
airway smooth muscle; SASM, sensitized airway smooth
muscle.
127
as judged from these data, is fully developed at this
time. It follows that all studies relating to asth-
matic-bronchospasm should focus on the first 2
sec. It must be acknowledged that this depends on
what is the load on the muscle. At loads greater
than 0.4 Po shortening would require more than 2
sec for completion and the importance of early
shortening would be less.
Studies of Po at steady state, i.e. at 10 s, provide
no information about ASM narrowing, only about
stiffening of the wall. Yet the majority of studies
carried out on isolated ASM have been conducted
on isometric preparations (deJongste et al., 1987;
Mapp et al., 1989; Shioya et al., 1987). With
respect to shortening it is certainly true that
measurements made at 10 sec in STSM and
CTSM will show the expected differences but will
not detect the fact that this was completed in 2 sec.
At 2 sec it is important to note that the major
regulatory enzyme is myosin light chain kinase
(smMLCK). At 10 sec the key enzyme is myosin
light chain phosphatase which by partially (by
75%) dephosphorylating the MLC20, develops
LBRs. Thus measurements made at 10 sec would
have completely missed the bus. We ourselves have
found (Jiang et al., 1992) that at 10 sec in STSM
neither Po or Vo are altered.
2.3. Force /velocity relationships
These have been reported for a variety of
smooth muscles (Jiang and Stephens, 1990; Mitch-
ell and Stephens, 1983; Seow et al., 2000; Stephens,
1985; Stephens et al., 1988) and show similar
qualitative characteristics. Jiang and Stephens
(1990, 1992) and Mitchell et al. (1994) have
reported curves for canine TSM and bronchial
smooth muscle (BSM). Smooth muscles show one-
thirtieth to one-fiftieth the velocity of fast skeletal
muscle. However, their shortening capacity is
much greater. Surprisingly their maximum force
is the same; surprising because smooth muscle has
only one-quarter the myosin content of skeletal.
Part of the explanation may be that since the
former cycles very slowly, contact time in a cycle
between actin and myosin is much longer thus
allowing for greater force development
128 N.L. Stephens et al. / Respiratory Physiology & Neurobiology 137 (2003) 125 /140

Fig. 4. Force /velocity curve incorporating loads greater than

Po; these forcibly stretch the muscle. The stretching velocities
are denoted by negative values below the abscissa. Shortening
velocities are shown as positive values.

Fig. 3. Force /velocity and power curve shown for canine

trachealis. The right hand ordinate (product of load and
velocity) refers to the power curve.

Fig. 3 shows force /velocity and power curves

for canine TSM. Maximum isometric force (Po) is
about 3 kg/m2 which is the same as that for
skeletal and cardiac muscle. Maximum velocity of
shortening (Vo) is 0.15 cm/s. When this is normal-
ized with respect to optimal length (Lo), Vo is
about one-fiftieth of that for skeletal muscle. The
power curve shows the same qualitative character-
istics as skeletal muscle. Fords’ group has shown
that an ideal index of contractility is power (Seow
et al., 2000).
Force /velocity relationships for P
/Po. This
represents forcible stretching of the muscle. Hanks
and Stephens (1981) reported data for such Fig. 5. Comparison of Hill’s equation and modified equation to
experiments in TSM; they show behaviour similar experimental data points. A better fit is achieved by using
modified force /velocity equation, especially at high load end of
to that of skeletal muscle (see Fig. 4).
curve; a , b , a , b and g are constants.
Force /velocity at P approaching Po. Stephens
et al. (1997) have reported that at heavy loads the external recoil force on the contractile elements
force /velocity deviated downward from the ex- (CE) that has the same direction of shortening as
pected rectangular hyperbola. Fig. 5 shows such that of the CE with the same magnitude of resting
force /velocity curves. Deviation of velocities from tension. So the conventional Vo is the maximum
the hyperbola is seen at high loads (100 /130 mN/ velocity of shortening of the muscle rather than of
mm2). As the TSM’s parallel elastic component its CE. To eliminate the effect of this PEC, load
(PEC) is stretched slightly at Lo, there is a small was reduced to 0.80 Lo such that resting tension
 N.L. Stephens et al. / Respiratory Physiology & Neurobiology 137 (2003) 125 /140 129

was zero. A non-hyperbolic relationship was still

obtained. In place of the convention hyperbolic
equation developed by Hill (1938) (P
/a )(v
/b )/
(Po
/a )b where P /force, v/velocity and a , b
and Po are constants, a new equation was devel-
oped:
V b(Po P)=[(a
P)(gP)];
in which g is a new constant, a and b/g are
approximation of Hill’s a and b, and g
/Po. As
Po approaches g , V , calculated from the modified
equation, approaches infinity, which suggests that
g can be interpreted as the maximum force a cross
bridge can bear.

2.4. Models of smooth muscle contraction

Over the last 10 years several models of smooth

muscle contraction have been developed (Chan et
al., 2000; Fredberg et al., 1999; Gunst et al., 1995;
Morano et al., 2000; Stephens et al., 1986). The
best of these is the Plasticity Theory model
formulated by Pratusevich et al. (1995) since it
accounts for the totality of results of biochemical
studies.

2.4.1. Plasticity theory

This theory which is substantiated by functional
and structural studies bids fair to radically alter
our notions of how smooth muscle contracts. It is
iconoclastic in showing that in smooth muscle the
active length /force curve does not conform to the
pattern of the Frank /Starling relationship. Over a
fairly wide range of length*/from 0.5 to 1.5 Lo */
active force is independent of length. Over this
same range both velocity and compliance show a
linear relationship to muscle length. This beha-
viour can only be accounted for by the theory of
smooth muscle plasticity which states that during
the process of adaptation induced by multiple
stimulations at any given length greater than Lo,
more contractile elements, are added in series to
the muscle cell. At lengths less than Lo, elements
are removed.
Fig. 6 shows the behaviour of P , Vo, and
compliance as a function of muscle length, as
published by Pratusevich et al. (1995).
Fig. 6. Isometric force (a), velocity scale factor (b) and
compliance (c) plotted as a function of central segment length.
The lower set of points in (a) represents rest force. Experiments
were conducted on tracheal smooth muscle.
At short muscle length */0.5 Lo */where nor-
mally overlap of the actin filaments occurs there is
a reduction in force development. However,
Fords’ recent work shows that reduction in active
force at this reduced length does not occur; this is
explained by the operation of plasticity. According
to this, plastic remodeling of the contractile
machinery occurs. Its magnitude is a function of
successive stimulations of the muscle. This results
in a progressive increase in active force, till within
5 or 6 stimulations, the force developed at 0.5 Lo
130 N.L. Stephens et al. / Respiratory Physiology & Neurobiology 137 (2003) 125 /140
becomes equal to that at 1.0 Lo. It is speculated
that in the course of remodeling at this length
some contractile units disappear. The remainder
then elongate and so eliminate thin filament
double overlap thus providing optimal positioning
for attachment of a maximum number of cross-
bridges. The reverse occurs at 1.5 Lo where the
actins are pulled far apart. During adaptation new
contractile units in series are synthesized which
restores units in the contractile machinery to
optimal overlap. Thus over a length range from
0.5 to 1.5 Lo the classical ascending and descend-
ing limbs of the length/force curve are replaced by
a horizontal line which demonstrates the length-
independence of active force.
Plasticity theory provides an interesting expla-
nation for the slow cycling, force-producing
LBR’s. At 0.5 Lo due to reduced numbers of
contractile units in series, longer thick filaments
likely develop. Since they bear more cross bridges
in parallel, active force increases. Thus plasticity
theory could account for the maintained max-
imum active force, and reduced Vo and DLmax at
0.5 Lo. The theory also indicates that as length
increases from 0.5 to 1.5 Lo, Po remains constant
while Vo and DLmax show a linear increase (see
Fig. 6).
2.4.2. Other models
There are other models that have been put
forward to account for smooth muscle contrac-
tion. Gunst for example has developed a plasticity
theory based on behaviour of the contractile
apparatus that is regulated by re-arrangement of
the components of the cytoskeleton, chiefly a-
smooth muscle actin, b-actinin, paxillin and talin
(Gunst et al., 1995). Hai has put forward another
model based on memory properties of the con-
tractile machinery in which the contractile re-
sponse depends not on the final length of the
muscle but on its initial length (Chan et al., 2000).
Morano et al. (2000) have adduced yet another
model in which smooth muscle and non-muscle
myosin heavy chains co-exist in the same smooth
muscle cell. While the former is responsible for the
initial rapid, phasic component of the contractile
response, the latter regulates the later tonic
component. Their conclusions stemmed from
knock-out experiments. Finally there is a model
developed by Fredberg et al. (1999) in which
perturbations in myosin light chain structure
regulated contraction. It is crucial to point out
that while all will account for the length-indepen-
dence of isometric force development only Ford’s
model accounts for the findings that while force is
length-independent, shortening capacity and velo-
city are linearly related to length. The rather
staggering conclusion from all of this is that in
the absence of consideration of plasticity, smooth
muscle contraction data are flawed. This leads to
the conclusion that all smooth muscle contraction
data extant need to be re-evaluated. In only those
instances where studies were conducted isometri-
cally at optimal length (Lo) validity still exists.
Plasticity theory also showed that as muscle length
is reduced and held at the shortened length for
about 18/24 h, active and resting curves are
exactly the same as those at Lo, except they have
shifted to the left. This has implications for
asthma, because it suggests that at reduced length
active force would not diminish and so bronchial
narrowing would not be attenuated. Similar me-
chanical behaviour is seen in skeletal muscle, for
example, the diaphragm in emphysema.
2.5. Cross-bridges in ASM: time-dependent
behaviour
The time-dependent behaviour of cross bridges
has been reported by us (Stephens and Jiang, 1995)
(see Fig. 7). These show a record of lightly pre-
loaded isotonic shortening in a strip of canine
tracheal smooth muscle stimulated electrically.
During the course of successive contractions zero
load clamps were applied at different times. Both
transient (stemming from elastic elements) and
steady state responses are seen. From the latter
maximum velocity was derived. Compartmental
analysis of a semi-logarithmic plot showed that the
velocities during the first 3 /4 sec were approxi-
mately three times faster than those developing
between 4 and 10 s. Thus it seems that two
populations of cross bridges are operative. We
have stated above that the faster cross bridges are
responsible for 75% of total shortening. However,
these data are derived from records obtained
 N.L. Stephens et al. / Respiratory Physiology & Neurobiology 137 (2003) 125 /140
during the course of isotonic shortening and some
of the reduction in Vo could be due to reduced
activation at short lengths (Stephens et al., 1986).
Where zero load clamp studies were carried out
during the course of an isometric contraction, two
types of cross-bridges were again seen, but the
change in velocity was not 300% but only 30%. In
ASM therefore the slow bridges are only slightly
slower than normally cycling cross-bridges. This is
unlike vascular smooth muscle where Murphy’s
group (Dillon et al., 1981) has reported that the
difference is 400% during an isometric contraction.
We have no explanation for the discrepancy other
than the difference in tissues.
A very recent and thought-provoking explana-
tion for the behaviour of slowly cycling cross-
bridges has been put forward by Seow et al. (2000);
it stems directly from their plasticity theory. They
suggest that at lengths below optimal, their Lref,
contractile units are removed by virtue of thick
filament depolymerization. The actin double over-
lap at short lengths is thus eliminated and maximal
force development becomes equal to the Po of Lref.
The remaining thick filaments necessarily lengthen
resulting in more cross-bridges in parallel per half
sarcomere. This would result in increased force
development and account for all the behaviour
imputed to latch bridges.
2.6. Relaxation in ASM: development of a
relaxation index
Studies of both vascular and ASM from models
of spontaneous hypertension and allergic bronch-
ospasm have shown that relaxation is considerably
diminished and suggested this contributes consid-
erably to luminal narrowing of the vessels and
production of the disease.
Relaxation in ASM has been little studied. Such
studies as exist describe isometric relaxation only
(Jewell and Wilkie, 1960) and this of course does
not directly relate to ASM dilatation but to
changes in wall compliance. For this reason we
conducted studies on isotonic force /velocity
curves in canine tracheal smooth muscle. Fig. 8
shows force (panel 8A) and displacement (panel
8B) data from after-loaded isotonic, force /velo-
city curves in CTSM. In each of the curves in the
131
Fig. 7. Isotonic shortening with load/0.1 Po. Record of
shortening versus time elicited from a muscle shortening
isotonically with a light preload. It is clear that better than
75% of total shortening is complete within 2 sec of stimulation.
upper panel, the pathway of isotonic relaxation at
different loads can be seen; it follows cessation of
the electrical stimulus. We have used the half time
for relaxation (t1/2r ) as an index of relaxation. This
index shows relaxation is load-dependent. Just as
is done for assessing velocity, relaxation must be
measured at zero load. Fig. 9 shows a plot of the
various t1/2r ’s versus load. A single exponential fits
the data. From the deduced equation zero load
relaxation can be obtained (Jiang and Stephens,
1992; Stephens et al., 1988).
An outstanding problem with this approach is
that the length of the muscle’s contractile element
at the onset of relaxation is also load dependent.
This would affect the magnitude of the relaxation
index and render it inaccurate. To correct for this,
relaxation at different loads was measured at the
peak of lightly loaded-isotonic shortening. Each
relaxation would therefore start at the same
contractile element length (see Fig. 10). The
t1/2r ’s for the different loads were computed and
plotted versus load. From this plot, zero load
relaxation was computed. It was essentially the
same as for the data shown in Fig. 9 indicating
that over the range of loads applied the contractile
elements did not change enough to affect relaxa-
tion analysis of the various phases of relaxation.
2.6.1. Analysis of the various phases of relaxation
Inspection of the isotonic relaxation under the
lightest loads in Figs. 8 and 10, shows that the
132 N.L. Stephens et al. / Respiratory Physiology & Neurobiology 137 (2003) 125 /140

Fig. 8. Isotonic relaxation of bronchial smooth muscle after 10 sec of electrical stimulation; force (Panel A) and displacement (Panel B)
are shown as functions of time. The muscle strip was allowed to shorten and relax under different loads. The time needed for muscle to
re-elongate to one-half of its shortening from Lo was designated as the half-relaxation time (T1/2) for that load. Panel C: Length in mm
(shortening increasing upwards) versus time. Line a represents maximum slope (at 200 ms after completion of fast transient) of the slow
transient of the response to zero-load clamp. Stimulus: EFS, 18 V, 60 cycle AC, 10 sec duration. F Clamp: Zero load clamps. Lower
panel: Velocity versus time derived from upper panel, b represents maximum velocity. Time is in seconds.

curves seem to consist of three components. In
Fig. 8 in which relaxation runs from the end of
stimulation for a succeeding 10 sec (from 12 to 22
sec), the curve is convex upwards. This is likely due
to the tailing-off of ATP hydrolysis resulting from
waning actin-activated myosin Ca2
/Mg2
-AT-
Pase activity. This has been monitored by follow-
ing NADH oxidase activity fluorometrically. The
zero load clamps shown in Fig. 8C indicate a
progressive diminution of velocity of shortening of
the contractile elements (the slow transient). Fol-
lowing phase i, there is a linear component (phase
ii) extending from 22 to 25 sec after cessation of
stimulation is seen. We speculated that this stems
from recoil of the muscle’s internal resistor. The
compliance curves for this internal resistance to
shortening have been published by Stephens et al.
(1988). The steady state transients of the zero load
clamps show that the velocity for the contractile
element is zero and therefore the response to the
clamp is entirely elastic.
From 22 sec to the end of relaxation which
occurs at 40 s, the curve shows a convexity which
is downward. This is phase iii. This could be due to
 N.L. Stephens et al. / Respiratory Physiology & Neurobiology 137 (2003) 125 /140 133

Fig. 9. T1/2 values for different loads fitted to exponential function. Zero-load half time was obtained by mathematical extrapolation.
Coefficient of determination, R2, of 0.9896 indicate good fit of curve. t , half time; a , zero-load half time; k , constant; P , tension.

viscosity of the internal resistor or to a sponta-
neous length-dependent re-activation of the
smooth muscle contractile apparatus. The zero
load clamps show that the velocity of the contrac-
tile element increases and is maintained for
another 15 s. This could be once again due to
either physical factors or to reactivation. Fig. 11
shows records of intracellular [Ca2
]i versus time.
At the time of the convex downward phase iii, an
increase in Ca2
concentration is seen which could
be responsible for re-activation of the smooth
muscle by phosphorylation of the smooth muscle’s
regulatory light chain (MLC20). Fig. 12 shows the
time course of MLC20 phosphorylation. An
increase in this is seen at the same time when
Ca2
levels are increasing and the zero load
clamps show muscle reactivation. These reactiva-
tion mechanisms account for the slowing of the
relaxation rate and the downward convexity of the
relaxation curve in phase iii. At a highly spec-
ulative level, this slowing in the relaxation could
forestall the development of turbulence in the
flowing blood or air. It could also have important
implications for failure of in vivo restoration of
length and of reversal of plasticity. We have
obtained some preliminary data (unpublished
observations) to indicate that the prolongation of
relaxation in sensitized ASM is predominantly due
to prolongation of phase iii.
3. Smooth muscle pathophysiology: biophysics and
biochemistry of asthmatic ASM
This is an important area of current research,
firstly, because the prevalence of asthma is high
134 N.L. Stephens et al. / Respiratory Physiology & Neurobiology 137 (2003) 125 /140

Fig. 10. Isotonic load clamp */signal versus time. Force and displacement versus time records of a muscle strip allowed to shorten
under a light preload and at a specific point in time quickly clamped to different heavier loads. Thereafter, muscle strips relax faster.
Values of T1/2CE for these relaxation curves were obtained and plotted against their respective loads from a single experiment.

(10% of the North American population is asth-
matic) in spite of earlier diagnosis and better
treatment and, secondly, because its mortality is
also increasing.
Major emphasis has been rightly placed on
research on human tissues and cells. Our own
experience has shown that the changes in mechan-
ical properties of asthmatic ASM in humans are
almost exactly the same as those of sensitized
canine ASM. We feel, therefore, that the canine
model of allergic bronchospasm is perfectly ade-
quate.
As highly desirable as it is to work on human
tissues there are practical limitations. It is extre-
mely difficult to obtain adequate numbers of
samples to arrive at statistically significant con-
clusions.
3.1. Results from animal models
3.1.1. Length /force studies
We have carried out research on ASM (both
tracheal and bronchial) obtained from ragweed
pollen-sensitized dogs. These are highly and re-
producibly sensitized and possess hyperreactive
ASM (Antonissen et al., 1979; Jiang et al., 1990;
Stephens et al., 1988, 1999).
Studies of Shortening of Tracheal (TSM) and
Bronchial (BSM) Smooth muscle Fig. 13 shows
isometric length /force curves obtained from TSM
from sensitized and litter-mate control dogs. At
the lowest load which corresponds to the resting
load needed to stretch the muscle to its optimal
length (Lo) the deduced mean shortening (deduced
because these were isometric studies) of the
sensitized TSM is about 15% greater than that of
the control (P B/0.05). The degree of shortening
equals about a 80% reduction in luminal area. This
could easily account for severe bronchospasm.
These differences were statistically significant
(P B/0.05) and prove that the mechanical proper-
ties of the STSM have indeed changed. In the early
days, we used TSM as a model for BSM and
concluded that central airway in asthma possessed
increased shortening ability. Subsequently we
conducted experiments on isolated bronchial
smooth muscle and confirmed the TSM findings.
 N.L. Stephens et al. / Respiratory Physiology & Neurobiology 137 (2003) 125 /140 135

Fig. 11. Upper trace is a record of isotonic shortening in a lightly loaded muscle. Shortening is depicted downwards. The stimulus was
turned off at peak shortening. The lower trace represents the corresponding Fura-2 (Ca2
) concentration late in relaxation.

3.2. Force /velocity studies difference between the muscles for maximum
isometric force (Po), maximum capacity (DLmax)
Fig. 14 shows force /velocity curves obtained and velocity (Vo) of shortening. At 2 sec, both
from sensitized and control TSM. The muscles DLmax and Vo were significantly increased,
were after-loaded and shortened isotonically . Ana- whereas Po was unchanged.
lysis of the curves revealed that the sensitized TSM
showed a greater maximum shortening velocity
(Vo) than the control. These curves are subject to
criticism since as the applied load for the different
contractions was increased, the accompanying
shortening necessarily occurred at later times.
Whereas at low force /load, shortening occurred
within 2 sec, at high loads, shortening commenced
at about 6 s. Thus, velocities measured at low
loads and early in the shortening were subserved
by normally cycling cross-bridges, those at 6 sec
were subserved by latch-bridges. Hence, the force /
velocity curve is a composite. However, curves
were also obtained by quick-release methods at 2
sec an 10 s. At 10 s, there was no significant Fig. 12. Time course (0 /60 s) of MLC 20 phosphorylation.
136 N.L. Stephens et al. / Respiratory Physiology & Neurobiology 137 (2003) 125 /140
Fig. 13. Length /tension curves for control and sensitized (test) tracheal smooth muscles, expressed in percentized units.
3.3. The importance of Vo in study of asthmatic
bronchospasm
It may be questioned whether Vo is a relevant
parameter to study since shortening would be
basically only time-dependent. Since we showed
that 90% of the increase shortening of the sensi-
tized TSM is complete within 1.5 s, clearly Vo
becomes a limiting factor, much as it is in the
heart.
These findings obtained on sensitized TSM were
also seen in sensitized BSM (Jiang et al., 1990;
Jiang and Stephens, 1990).
3.4. Molecular mechanisms accounting for increase
in Vo
3.4.1. Actin-activated myosin light chain Mg2
-
ATPase activity
We conducted measurements of enzyme activity
in sensitized and control (TSM) and found that it
was increased significantly (P B/0.05) (see Table
1). This could account for the increased velocity of
shortening (Kong et al., 1986). These studies were
conducted in homogenates containing optimal
levels of MLCK and so the results obtained
expressed optimally stimulated acto-myosin light
chain ATPase activity.
3.4.2. Myosin light chain (20 kDa)
phosphorylation
Since this is the major regulator of smooth
muscle contraction, we measured its time course.
Fig. 12 shows a time plot of MLC20 phosphoryla-
tion. Phosphorylation of the sensitized TSM was
significantly increased (Stephens and Jiang, 1995).
This could account for the increased actin-acti-
vated Mg2
-ATPase activity.
3.4.3. Content of smMLCK
To account for increased phosphorylation of
MLC20, we measured the total and specific
 N.L. Stephens et al. / Respiratory Physiology & Neurobiology 137 (2003) 125 /140 137

Fig. 15. Total specific MLCK activity in sensitized and control

Fig. 14. Force /velocity curves for control and sensitized (test)
canine tracheal smooth muscle. The linearized transforms (Po/
P )/V versus load are also plotted. They confirm the hyperbolic
nature of the parent curves. From the slopes and intercepts of
these lines the a and b constants for the force /velocity
hyperbola were obtained.
Table 1
ATPase activity of canine tracheal smooth muscle myofibrils
ATPase activity of myofibrils (mMPi mg per
mosin min 1
TSM and BSM. Left panel: the incorporation rate of 32P into
MLC (MLCK activity) in TSM and BSM was significantly
higher than in control. Right panel: When expressed as the
incorporation rate of 32P into MLC/:gMLCK/min, i.e. specific
activity of MLCK the difference between sensitized and control
muscles disappeared. FWT, fresh weight of tissue.
(Liu and Stephens, 1995) in either contraction
mode between the control and sensitized TSM.
From these data, we concluded that sensitized
ASM showed increased shortening ability induced
by increased phosphorylation of the 20 kDa
myosin light chain and increased activation of
the actin-activated Ca2
/Mg2
myosin ATPase
activity.

STST, n/7 499.299/11.49

CTSM, n/ 345.009/6.81
7

activity of smMLCK. Fig. 15 shows that while

specific activity was unchanged, total activity was
increased and it is this that is responsible for
increase phosphorylation. Since smMLCK is regu-
lated by the 4Ca2
/calmodulin complex, we
measured the content of the latter and its activity
by its ability to stimulate PDE activity. No
significant difference was found (Liu and Ste-
phens, 1995).

3.4.4. Cellular Ca2
concentration
Fig. 16 shows plots of intracellular Ca2
con-
centration (ionic activity) versus time in control
and sensitized TSM contracting isotonically and
isometrically. No significant difference was found
Fig. 16. Typical Ca2
transients during EFS of sensitized and
control TSM. No difference was found between sensitized and
control. Isotonic shortening records in both sensitized and
control groups showed lower Ca2
transients than those of
isometrics. Ca2
concentration was normalized using standard
Ca2
solution from Molecular Probes.
138 N.L. Stephens et al. / Respiratory Physiology & Neurobiology 137 (2003) 125 /140

3.5. Motility assay studies

To determine the Ca2
/Mg2
ATPase activity

of the smooth muscle myosin heavy chain
(smMHC), a motility assay was carried out. The
methodology employed was developed by Ume-
moto and Sellers (1990). Basically it measured the
ability of isolated myosin subfragment I obtained
from CTSM and STSM, to translate actin fila-
ments on a glass slide under an inverted micro-
scope; images were captured by coupling to a
video camera. From this, the velocity of actin
translation was derived. Since the system was
unloaded, zero load velocity (Vo) was obtained.
Zero-load velocity is conventionally used as an
index of acto-myosin Ca2
/Mg2
ATPase activ- Fig. 17. Actin filaments moving. Frames 1, 2 and 3 show
ity. All other variables were held constant. In this filaments in motion. The horizontal twisted ‘rope-of-pearls’
context, the extent of phosphorylation of the structures seen towards the top of the picture are the
MLC200 is held constant by introducing equiva- tetramethyl-rhodamine stained actin filaments which are trans-
lated by subjacent myosin heads. Under fluorescence micro-
lent amounts of smMLCK. The levels of calcium scopy these produce a strong signal. Attached to these filaments
were also the same. Fig. 17 shows successive video- and extending vertically downwards are the myosin heads (two
recorder frames. The arrows indicate movement of heads for molecule) with a small segment of the rod-like tail.
an actin filament. These tails are attached to Y-shaped structures, the antibodies
which were raised against a sequence in the tails. The lowest
Table 2 shows mechanical data from isolated
part of the antibody (the handle of the Y) adheres to the
single smooth muscle myocytes obtained from nitrocellulose coating the slide. The surface of the slide not
control and ragweed-pollen sensitized canine tra- involved in binding down the antibody is coated with bovine
cheal smooth muscle. No significant difference in serum albumen.
Vo was detected. The level of phosphorylation of
the 20 kDa myosin light chain was the same in the
two groups. We concluded that the myosin heavy
Table 2
chain’s actin activated myosin Ca2
/Mg2
AT-
Results of motility assay
Pase activity was unaltered. We also concluded
that the extent of activation of the myosin Vo (mm/sec)9/SD
ATPase, as evidenced by extent of the MLC20 Sensitized canine ASM myosin 08569/0.173 (n/58)
phosphorylation, was responsible for the increased Control canine ASM myosin 0.7549/0.169 (n/60)
Vo seen in intact sensitized ASM. Fig. 18 shows a
dose /response curve for control ASM. The dose is
that of smMLCK concentration and the response degree of activation of the enzyme that accounts
that of Vo obtained from the motility assay. The for increased Vo.
curve is sigmoidal and supports the role of extent
of MLC20 phosphorylation in determining the
increased Vo of sensitized ASM. Our previous 3.5.1. Results from studies on human asthma
work on sensitized ASM strips had shown that Vo, The major question in ASM research has been
DLMAX, smMLCK content, and extent of MLC20 (a host of results from animal models notwith-
phosphorylation are increased. The motility assay standing): is the contractility of bronchial smooth
results demonstrate that the ATPase properties of muscle muscles from human subjects of asthma
smMHC itself are unaltered in asthma. It is the increased. The research cannot be conducted in
 N.L. Stephens et al. / Respiratory Physiology & Neurobiology 137 (2003) 125 /140
Fig. 18. Dose /response curves of myosin light chain kinase
(MLCK) content versus maximum velocity of actin filament
translation.
smooth muscle because of the great difficulties in
obtaining specimens in adequate numbers.
We have been successful in obtaining endobron-
chial biopsy material from asthmatics and
matched controls, in sufficient numbers to be
able to test for significant differences. The study
was approved by the Laval Hospital Ethics
Committee (Laval Hospital, Ste. Foy, Quebec,
Canada). Bronchial smooth muscle cells (BSMC)
were isolated from the samples by enzymatic
methods. For mechanical studies, BSMC were
obtained from five asthmatic and five normal
Fig. 19. Measurement of mRNA for smMLCK-108. The upper
panel shows smMLCK mRNA bands obtained by 5% poly-
acrylamide gel electrophoresis. Lane 1 presents molecular
weight marker bands. Lanes 2 /7 represent asthmatics, while
8 /10 represent non-asthmatic.
139
subjects. Their electrically-elicited shortening ca-
pacities were measured under an inverted micro-
scope coupled to a video-recorder and a data
analysis computer system. Statistically significant
increases in DLmax (39.05% of rest length9/1.99 SE
in asthmatic cells and 28.6%9/1.15 SE, in controls)
and Vo (7.2% of rest length per second9/0.8% SE
in asthmatic cells and 4.759/0.76% SE in controls)
were seen. The change in cross-sectional area of
the bronchial wall muscle and reduction in lumen
size resulting from this shortening change was
sufficiently large to account for the bronchospasm
of asthma. To elucidate the mechanism, we
measured mRNA abundance of smooth muscle
types of myosin heavy chain (smMHC), and
smooth muscle myosin light chain kinase
(smMLCK). RT-PCR using appropriate primers,
was employed to amplify these messages. Fig. 19
shows smMLCK mRNA bands in agarose gel.
Bands 1 /7 from asthmatic cells are denser and
larger than those of the control. Analysis revealed
that the content of message for smMLCK in
asthmatics (0.406 arbitrary densitometric units9/
0.021 SE, n /7) was higher (P B/0.05) than in
controls (0.04 arbitrary units9/0.008 SE, n/11).
These results show that the increased bronchos-
pasm seen in asthmatic humans could be due to
increased shortening ability of the ASM cell.
4. Conclusion
It is evident that considerable advances have
been made in the understanding of the physiology
and biochemistry of ASM. This is true at both
basic and applied levels. Perhaps the most out-
standing development in the field has been the
formulation of the plasticity theory. The demon-
stration that contractility of ASM is significantly
increased in human subjects of asthma was long
overdue.
Acknowledgements
The research carried out was made possible by
operating grants from the Canadian Institutes of
Health Research and the Manitoba Institute of
140 N.L. Stephens et al. / Respiratory Physiology & Neurobiology 137 (2003) 125 /140
Child Health. Thanks are due to Judy Olfert for
expert word processing.
References
Antonissen, L.A., Mitchell, R.W., Kroeger, E.A., Kepron, W.,
Tse, K., Stephens, N.L., 1979. Mechanical alterations of
airway smooth muscle in a canine asthmatic model. J. Appl.
Physiol. 46, 681 /687.
Chan, W.L., Silberstein, J.S., Hai, C.M., 2000. Mechanical
strain memory in airway smooth muscle. Am. J. Physiol.
278, C895 /C904.
deJongste, d.C., Mons, H., Bonta, H., Kerrbijn, K.F., 1987. In
vitro responses of airways from an asthmatic patient. Eur. J.
Respir. Dis. 1, 23.
Dillon, P.F., Aksoy, M.O., Driska, S.P., Murphy, R.A., 1981.
Myosin phosphorylation and the crossbridge cycle in
arterial smooth muscle. Science 211, 495 /497.
Frank, O., 1904. Thermodynamik d Muskels. Ergebn d Physiol.
3, 2.
Fredberg, J.J., Inouye, D.S., Mijailooich, S.M., Butler, J.P.,
1999. Perturbed equilibrium of myosin binding in airway
smooth muscle and its implications in bronchospasm. Am.
J. Resp. Crit. Care Med. 159, 1 /9.
Gunst, S.J., Meiss, R.A., Wu, M.-F., Rowe, M., 1995.
Mechanisms for the mechanical plasticity of tracheal
smooth muscle. Am. J. Physiol. 268, C1267 /C1276.
Hanks, B.S.R., Stephens, N.L., 1981. Mechanics and energetics
of lengthening of active airway smooth muscle. Am. J.
Physiol. 241, C42 /C47.
Hill, A.V., 1938. Heat of shortening and dynamic constants of
muscle. Proc. R. Soc. Lond. Biol. Sci. 126, 136 /195.
Jewell, B.R., Wilkie, D.R., 1960. The mechanical properties of
relaxing muscle. J. Physiol., Lond 152, 30 /47.
Jiang, H., Rao, K., Halayko, A.J., Kepron, W., Stephens, N.L.,
1990. Mechanical alterations in sensitized canine bronchial
smooth muscle. FASEB J. 4, A444.
Jiang, H., Stephens, N.L., 1990. Contractile properties of
bronchial smooth muscle with and without cartilage. J.
Appl. Physiol. 69, 120 /126.
Jiang, H., Rao, K., Halayko, A.J., Kepron, W., Stephens, N.S.,
1992. Bronchial smooth muscle mechanics of a canine
model of allergic airway hyperresponsiveness. J. Appl.
Physiol. 72, 39 /45.
Jiang, H., Stephens, N.L., 1992. Isotonic relaxation of sensi-
tized bronchial smooth muscle. Am. J. Physiol. 262, L344 /
L350.
Kong, S.K., Shiu, R.P.C., Stephens, N.L., 1986. Studies of
myofibrillar ATPase in pulmonary smooth muscle. J. Appl.
Physiol. 60, 92 /96.
Liu, G., Stephens, N.L., 1995. Increased Ca2
/ /Calmodulin
complex in ragweed pollen-sensitized canine tracheal
smooth muscle. Am. J. Respir. Crit. Care Med. 151, A288.
Mapp, C.E., Chitano, P., DeMarzo, P., 1989. Responses to
acetyl-choline and myosin content of isolated canine air-
ways. J. Appl. Physiol. 67, 1331 /1335.
Mitchell, R., Stephens, N.L., 1983. Maximum shortening
velocity of smooth muscle zero load-clamp vs afterloading.
J. Appl. Physiol. 55, 1630 /1633.
Mitchell, R.W., Ruhlmann, E., Magnussen, H., Rabe,
A.R.L.F., 1994. Passive sensitization of human bronchi
augments smooth muscle shortening velocity and capacity.
Am. J. Physiol. 267, L218 /222.
Morano, I., Chai, G.-X., Baltas, L.G., Valeria, L.Z., Lutsch,
G., Koff, M., Haase, H., Border, M., 2000. Smooth muscle
contraction without smooth muscle myosin. Nature Cell.
Biol. 2, 371 /375.
Otis, A.B., 1983. A perspective of respiratory mechanics. J.
Appl. Physiol. 54, 1183 /1187.
Pratusevich, V.R., Seow, C.Y., Ford, L.E., 1995. Plasticity in
canine airway smooth muscle. J. Gen. Physiol. 339, 1493 /
1497.
Seow, C.Y., Pratusevich, V.R., Ford, L.E., 2000. Series-to-
parallel transition in the filament lattice of airway smooth
muscle. J. Appl. Physiol. 89, 869 /876.
Shioya, T., Pollack, B., Munoz, N., Leff, A., 1987. Distribution
of airway contractile responses in major resistance airways
of the dig. Am. J. Pathol. 129, 102 /108.
Siegman, M.J., Butler, T.M., Mooers, S.U., Trinkle-Mulcahy,
L., Narayan, S., Adam, L., Chacko, L.S., Haase, H.,
Morano, I., 1997. Hypertrophy of colonic smooth muscle:
contractile proteins, shortening velocity and regulation. Am.
J. Physiol. 272, G1571.
Stephens, N., Wang, J., Halayko, A.J., 1997. Airway smooth
muscle contraction. In: Barnes, P.J., Grunstein, M.M., Leff,
A.R., Woolcock, A.J. (Eds.), Asthma. Lippincoot-Raven,
Phila, pp. 759 /800.
Stephens, N.L., 1985. Airway smooth muscle physiology,
bronchomotor tone, pharmacology and relation to asthma.
In: Weiss, E.B., Segal, M.S., Stein, M. (Eds.), Bronchial
Asthma. Little Brown, Boston, MA, pp. 96 /110.
Stephens, N.L., Kagan, M.I., Packer, C.S., 1986. Time depen-
dence of shortening velocity in tracheal smooth muscle. Am.
J. Physiol. 251, C435 /C442.
Stephens, N.L., Kong, S.K., Seow, C.Y., 1988. Mechanisms of
increased shortening of sensitized airway smooth muscle. In:
Armour, C.L., Black, J. (Eds.), Mechanisms in Asthma.
Alan R Liss, New York, pp. 231 /254.
Stephens, N.L., Jiang, H., 1995. Basic physiology of airway
smooth muscle. In: Busse, W., Holgate, S. (Eds.), Asthma
and Rhinitis. Blackwell Scientific, pp. 1087 /1115.
Stephens, N.L., Wang, Y., Li, W., Ma, X., 1999. Airway
hyperreactivity: direct smooth muscle approach. Pulm.
Pharm. Therap. 12, 97 /101.
Umemoto, S., Sellers, J.R., 1990. Characterization of in vitro
motility assays using smooth muscle and cytoplasmic
myosins. J. Biol. Chem. 265, 14 864 /14 869.