Dec 10, 1984 - to methacholine perfusion. We wished to determine whether the muscarinic receptor forms were static or subject to change. We found changes.
0022-3565/85/2323-0754$02.OO/O THE JOURNAL OF PHARMACOLOGY Copyright © 1985 by The American
AND
EXPERIMENTAL
Society
Vol. 232, No. 3 Printed in U.S.A.
THERAPEUTICS
for Pharmacology
and Experimental
Therapeutics
Cardiac Cholinergic Muscarinic Receptors: Affin ity Forms with Down-Regulation1 ROBERT
ROSKOSKI,
Departments Department
of Biochemistry (R.R., R.R.R., WE.) and Biometry of Chemistry (P.F.C.), North Texas State University,
Accepted
JR.,
for publication
RICK
December
A.
REINHARDT,
WOLF
ENSELEIT,
(W.D.J.), Denton,
Louisiana Texas
Changes
WILLIAM
D.
State UnWersity
JOHNSON Medical
Center,
in Multiple
and
PAUL
F. COOK2
New Orleans,
Louisiana
and
10,1984
ABSTRACT The isolated working rat heart exhibits dynamic changes in the cholinergic muscarinic receptor in response to perfusion with the acetylcholine congener methacholine. For example, perfusion with 4 zM methacholine for 2.5 hr mediates a statistically significant and reversible 1 0 to 1 5% decrease in receptor content in right atrium and left atrium and ventricle. The muscarinic receptor exhibits a single affinity state for antagonists but multiple affinity states for receptor agonists. When agonist 3H-antagonist competition experiments are performed, the concentration dependence of displacement by agonists is flattened and extends over more than 2 log units. From such curves, it is difficult to extract values for the relative proportions of the multiple receptors or their K0 values by inspection. We have developed a procedure
Although to control
the heart by the
exhibits
intrinsic
autonomic
rhythmicity,
nervous
system.
thetic innervation decreases heart rate whereas the sympathetic division mediates sponses. The postganglionic parasympathetic is acetylcholine
tissue
which
and the
muscarinic
interacts
myocardium
receptor
with
through
(Higgens
the the
it is subject
The and the
action
et at., 1973).
The
ofthe
July
3, 1984.
in part
by Grants
HL-24791
Recipient
of U.S.
Public
Health
Service
Research
Career
ABBREVIATIONS:
754
QNB, quinuclidinyl
benzilate;
1975; et at.,
Ward
and
Young,
1977),
as
well
as
rat
1979).
modified
(R. R.) and GM 31686 Development
GppNHp,
at.,
pacemaker
(P. F. C.). 2
et (Berrie
cholinergic
re-
characterization
AM 01155.
(Taylor
A variety of procedures have been used to calculate the KD and proportions of the multiple forms of receptor through the use of competition assays with radioactive and unlabeled agents (Hulme et aL, 1978). These procedures use
contractility opposite
and quantitation of this receptor was greatly facilitated by the development of [3H]QNB (a potent muscarinic receptor antagonist) as a radioligand by Yamamura and Snyder (1974). Hulme and co-workers (1978) reported that antagonists exhibit simple mass action binding to the rat brain muscarinic receptor. In contrast, the binding of agonists is more complex and is consistent with the existence oftwo or more affinity states (Birdsall et at., 1978). Similar results have been found with rabbit and porcine heart (Fields et at., 1978; Schimerlik and Searles, 1980), embryonic chick heart (Galper and Smith, 1980; Halvorsen and Nathanson, 1981; Galper et aL, 1982) and ileal smooth muscle Received for publication 1 This work was supported
formsof
heart
parasympa-
neurotransmitter
specialized
for plotting the competition curves so that the K1, values and proportion of multiple receptor forms can be estimated graphically. We determined these more accurately by computer using a nonlinear least-squares analysis. After perfusion of methacholine for 1 hr, there were increases in the KD of the low and high affinity the receptor in the right and left ama. After 2.5 hr of perfusion, the K1, of the high-affinity form increased further in the left atrium. In the right atrium, the two affinity states were converted into a single state of low affinity. Although there was a significant decrease in the amount of receptor in left ventricle, there were no changes in the K0 values or proportions of the two states. All changes in receptor reversed during an additional 2.5 hr of perfusion without methacholine.
Award
values
Eadie-Hofstee
analyses
(Molinoff
et
aL,
1981)
and
nonlinear least-squares curve fitting (Hancock et at., 1979). We have developed a general procedure for plotting the competition of one agent by another which requires the determination of bound and free radioligand concentration at each concentration of inhibitor and the KD of the radioligand in the absence of inhibitor (determined independently by a Scatchard analysis). Our procedure gives an intuitive grasp of the approximate KD values and proportions of the multiple states by inspection. We have also calculated the KD values and proportions of the multiple states using nonlinear least-squares fitting procedures. We reported previously that perfusion ofthe isolated working rat heart with 4 M methacholine for a 2.5-hr period brought about a 10 to 15% decrease in receptor in right and left atria and left ventricle, but not in right ventricle (Reinhardt and Roskoski, 1983). After an additional 2.5 hr in the absence of methacholine, the receptor content returned to control values. There was no detectable decrease in receptor content after 1 hr
5’-guanylylimidodiphosphate.
Cardiac
1985
of perfusion with methacholine. and proportions of the multiple in response
to methacholine
whether the muscarinic to change. We found proportions
in two
We
measured
affinity
perfusion.
receptor changes
heart
the
forms We wished
where
occurs
(right and left atria). We found no changes in KD values or in the proportions of the two states in right or left ventricle in response to methacholine perfusion even though down-regulation was observed in the latter chamber.
Materials Agonist Sprague
3H-antagonist Dawley
absence of 4 M
and Methods assays.
competition
rats
g) were
(200-250
perfused
Hearts in
the
from presence
male or
for the specified times as described previously (Reinhardt and Roskoski, 1983) using the procedure of Morgan et at. (1980). After perfusion, the hearts were dissected and stored frozen in liquid nitrogen. At the time of the assay, hearts were homogenized in Ringer’s solution (1:32 for atria and 1:16 for ventricle, w/v). Duplicate tubes containing a given concentration of [3H]QNB of about 25 pM and about 22 concentrations of unlabeled methacholine (from 1 M to 3 mM, final) in Ringer’s solution were prepared. Portions (25 d) of homogenate were added to initiate the reaction (1 ml final volume). After 30 mm at 37’C bound [3H]QNB was collected on G/FA glass fiber filters as noted previously (Reinhardt and Roskoski, 1983). Nonspecific binding was determined in the presence of 1 MM unlabeled QNB or atropine in duplicate tubes. [3H]QNB binding was also determined in the absence of added methacholine. Protein was determined by the procedure of Lowry et at. (1951) using bovine serum albumin as methacholine
standard. Theory. A number of methods for graphical analysis ofbinding data from competition assays are currently in the literature (Molinoff et at., 1981; Hancock et at., 1979). These analyses usually consist of plots based on the linear transform of the Michaelis-Menten equation developed by Eadie-Hofstee. Equation 1 describes the competitive binding of ligand I and ligand L with receptor
-
B
Kq(1
)
B[L] +
+
(1)
EL]
i)
+
or subject and their
down-regulation
a concave
K(K1 BK1+ [I]) B
=
B
-
[K(K1 I K1+ [J])]B
(K
+ q
Equation
3 can
be rearranged B.(i
In this case [U) is plotted
+
upward
curve will be obtained.
of [3H]QNB
=
(5) EL]
+
Extracting
estimates
of
B
methods.
considered four computational methods ofthe nonlinear model: 1) the grid-search or Taylor series method, 3) the gradient and 4) Marquardt’s method which is a
We
for estimating the parameters method, 2) the Gauss-Newton
or steepest-descent compromise
methods
method
between
Gauss-Newton
were extremely
sensitive
and
steepest
descent.
to the choice of initial
four
All
estimates
but
usually
resulted in final estimates which were in close agreement for test data. For convenience, the results reported here were calculated by the Gauss-Newton method using the BMDP program P3R (Dixon et at., 1981). The algorithm for obtaining estimates of parameters of the modeluses iterative techniques to minimize the residual sum of squares. Convergence was usually obtained in less than 15 iterations with initial estimates chosen from graphs of data plotted according to equation 4. A statistical decision regarding the number of receptor classes (i.e., i vs. i + 1) was aided by the use of partial F tests, inspection of plots of residuals of the fitted model and comparison of the asymptotic S.E.s of the estimated parameter to the estimates (i.e., the coefficient of variation of the estimates). Two parameters were estimated for each class of receptor in a given model. Let SS1 and SS1+1 denote the residual sums of squares for i and i + 1 classes, respectively, and r the degrees associated
with
SS1. Then
F_SiSS1)
gives a suitable
statistic
r
ssi
-
for testing
2
the hypothesis
that
the appropriate
number of classes is i rather than i + 1. This statistic is approximately distributed as F with 2 and r degrees of freedom. Student’s t test was used to determine the significance between means among
and analysis means (Zar,
of variance
was used
to determine
differences
The KD values were calculated apparent KD values by the procedure of Cheng and Prusoff correct for receptor sites occupied by [3H]QNB. 1984).
from
the
(1973)
to
Results FEll]
We examined (4)
is plotted on the ordinate and B.(K[I]/ on the absissa. B is obtained directly as the ordinate intercept with -1/K1 as the slope. Kq (66 pM) was obtained from more than 10 Seatchard analyses of the binding of the ligand (L) to receptor under our experimental conditions (present work and Reinhardt and Roskoski, 1983). In the case in which more than one class of receptor exists which has different affinities for a competing ligand (I), but the same affinity for ligand (L), the following equation obtains. + K/[L])
)
(3)
K11[L]
B,,.K[I] [L]
+
binding.
Statistical
as follows.
j#{231}\ B [L]J
\
(2)
Kq(1
and the apparent K1 which have known statistical properties requires an iterative fitting procedure as will be discussed in the following section. By visual inspection, the negativeof the reciprocal of the slope of each portion of the curve gives the apparent KD (as in a Scatchard analysis); the maximum ordinate intercept represents 100% of the multiple states and the extrapolated ordinate intercept of each affinity form yields an estimate of its proportion of the total binding sites. The total receptor (maximum binding) is not given by this competition procedure and can be determined independently by Scatehard analysis
where
B. and B represent the amount of L bound to receptor at any concentration below saturation and saturation, respectively, Kq is the dissociation constant for L from the R/L (receptor/ligand) complex and K1 is the dissociation constant of I from the Rh complex. If one rearranges the equation in Scatchard and Eadie-Hofstee form, respectively, equations 2 and 3 are obtained:
[L]
+
In equation 5, W and B” represent the amount ofL bound to receptor classes 1 and n, respectively (n represents the apparent number of classes of receptor), whereas K11 and K1,, represent the dissociation constant for R1/I and R5/I, respectively. All other terms have the same definition as above. When data are plotted in the form of equation 4,
of freedom
f_ [U
B5[L] +
B=
to determine
forms were static in the KD values
regions
receptor
755
Receptors
B’[L]
values
KD
of the
Muscarinic
regions 2.5 hr.
are required QNB from sion
the multiple
(fig.
to displace the
right
states
the
than
expected
that
subsaturating atrial
1). As reported
rat brain,
curve
such displacement tion or KD values rinic
affinity
in response to perfusion with 4 tM We found that higher concentrations
receptor
from
curves, of the for
after
by Hulme
from
the
multiple
control mass
affinity
If the
of [3H]
methacholine
et at. (1978)
a theoretical it is difficult
methacholine.
of methacholine
concentrations
receptor
obtained
of several rat heart methacholine for
hearts action
to estimate forms
data
perfu-
in the
of
is flatter
curve.
From
the proporof the
are
case
musca-
replotted
as
756
Roskoski
et al.
Vol. 232
accurate than that for the high-affinity state, which is the usual case for parameters obtained by graphical analysis of Scatchard-type plots. The KD values obtained after the procedure of Cheng and Prusoff (1973) were 2.9 ± 0.7 and 171 ± 17 tM,
a
a)
respectively.
After 2.5 hr of perfusion with 4 iM methacholine, higher concentrations of methacholine are required to inhibit [3H] QNB binding (fig. 1) from homogenates of right atria. When this is replotted, the results are shown in figure 2B. Computer fitting was best with a single affinity state with a KD of 166 ± 16 ,M (corrected KD = 120 ± 12 SM). There is a change from two states to a single state of intermediate affinity in the right atrium associated with receptor down-regulation. After an additional 2.5 hr of perfusion in the absence of methacholine, the amount of receptor returns to control levels and the two affinity states also return (table 1). These experiments indicate that the subtypes of the muscarinic receptor undergo reversible changes over relatively short time periods. We observed previously a decrease in receptor in left atrium
z a
io2
io-
[Methacholine], M Fig. 1. Competition of methacholine and [3H]QNB for right atrial receptor in heartS perfused with and without methacholine. Rat hearts were perfused 2.5 hr in the presence or absence of 4 M methacholine and the tissues were dissected and stored in liquid N2 as described previously (Reinhardt and Roskoski, 1 983). Right atrial homogenates were incubated with subsaturating concentrations (25 pM) of [3HJQNB and varying concentrations of methacholine as descilbed under “Materials and Methoris.” The points represent the experimental value and the curves were generated by computer fitting for 2 sites (0, control perfusion) and 1 site (#{149}, methacholine perfusion).
after
The
the
estimate
of the KD of the
low-affinity
state
was
of
perfusion
with
high-affinity
As in the
methacholine
(Reinhard
and
form
right
increased
from
4.4 to 17 M
and
the
low-
affinity form increased from 172 to 240 cM. The percentage of the high-affinity form, however, increases from 40 to 57%. Although comparable receptor decreases occurred in the right and left atrium, the response of the receptor affinity states differed. After an additional 2.5 hr of perfusion in the absence of methacholine, control state
1fM(47±2%);KD2=236±23M(53±2%).
visual
h
atrium, we find two affinity left atrium (fig. 3; table 1). We wanted to determine whether the two receptor subtypes were converted to a single affinity state after methacholine perfusion as we found in the right atrium. Instead, we found that the KD for
described in equation 4 (“Materials and Methods”), the results can be displayed in a more comprehensible fashion (fig. 2). The experimental data are plotted as open circles and the curved line gives the computer fit of the data assuming two affinity states. From the straight lines drawn by inspection, one can calculate the approximate KD values (from the negative reciprocal of the slopes) and the proportions (from the y-axis intercepts) thereby obtaining: K1,1 = 12 M (44%) and KD2 = 247 M (56%). By computer fitting using nonlinear least-squares analysis, the following uncorrected KD values were obtained: KD1=4±
2.5
Roskoski, 1983). states in control
the indicating
KD values that the
and proportions changes were
return reversible
to the (table
1).
more
5
4
z a z
a U
z a
z
0
a +
8
KQNB#{149} CMethacholineJ
I
[QNB] x 1O
(M2)
Fig. 2. MOdified Scatchard replot of the competition of methacholine and [3H]ONB (B) rat hearts. The data on the competition of methacholine and [3H]QNB 4. The points are the experimental values and the solid line is a computer fit lines were drawn by inspection. The negative reciprocal of the slope yields the !(. The ordinate intercept. used equation
Bobs#{149} KQNB#{149} [Methacholine]
/
[QNB]
x i#{252}15 (M2)
for the right atnal receptor in control (A) and methacholine from which figure 1 was derived were plOtted according to of the data from a 2-site (A) and 1-site (B) model. The broken percentages of each site are proportional to the length of the
Cardiac
1985
Muscarinic
757
Receptors
TABLE 1 Methacholine perfusion reversibly afters multiple affinfty states of the muscarinic receptor in selected heart regions The hearts were perfused in the absence of presence of 4 zM methacholine as described previously (Reinhardt and Roskoski. 1983). The K0 values and proportion the mulp$e
affinity
for methacholine
forms
Reon
were determined Perfuon
as decribed
Materials
under
and Methods.
The data represents
Total Receptor
Condeons
2.5 hr control 2.5 hr methacholine 2.5 hr methacholine +2.5 hr control 2.5hrcontrol
Leftatrium
Right ventricle
Right atrium Left atrium a C
252
2.5 hr methacholine 2.5 hr methacholine +2.5 hr control 2.5hrcontrol 2.5 hr methacholine 2.5 hr methacholine +2.5 hr control 2.5 hr control 2.5 hr methacholine 2.5 hr methacholine 2.5 hr control
Leftventricle
a
262 235
± ± ±
6#{176}
8.5
6.8 3.4
73.5±2.5 59 ± 2.5k 65.0 ± 8 3.5
84.5 ± 4 84.5 ± 4.5 258 ± 7
1 hr methacholine
279
±
pM
8
178 ± 15 (53 ± 2)
± 6.5b (1O0%)b ± 1.0(47 ± 3)
4.4±1.0(40±2)
± 7b ± 10
1 hr methacholine
Data from six hearts in each group (Reinhardt and ROSkOSki, different from control (P < .05). Sigr#{227}f’-’ntiy different from control and 2.5 hr of methacholine
120 3.3
285±7
±
of
determinations.
ofthree
isM
2.4 ± 0.9 (47 ± 2)
6
234 266
89.5
± S.E.M.
K,, (%)
md/mg
Right atrium
the mean
± ±
5.4±
174
1.0(30±2)
0.8 (24 0.9 (26 0.8 (28
8.8 8.4
± ±
0.8k (50 ± 2%) 09bc (40 ± 3%)
± 3)
202±15(70±2) 187 ± 15 (72
4.1 ± 1 .2 (28 ± 2) 5.4 ± 0.9 (32 ± 2) ± ± ±
14(53
201 ±15(60±2) 279 ± 17 (43 ± 3)#{176} 200 ± 14 (63 ± 2)
1 .6b (57 ± 3)b 0.9 (37 ± 2)
4.1 3.8 3.8
±
± 2) 195 ± 14 (68 ± 2)
± 3) ± 2) ± 2)
176
±
15 (76
±
3)
1 77 ± 14 (74 ± 2)
166
±
15 (72
±
2)
275 ± 20
273
±
(50 ± 2%) 16b (60 ± 3%)
1983). treatment (P