Assay of adenylate cyclase in homogenates of control and Duchenne human skeletal muscle

Chnrco Chimicu Am, I I I (1981) 133- 146 0 Elsevier/North-Holland Biomedical Press

CCA

133

1674

Assay of adenylate cyclase in homogenates of control and Duchenne human skeletal muscle Depurtmenr

C. Cerri*, J.H. Willner**

and L.P. Rowland

und H. Il. Merritt

Reseurch Center for Muxulur

of Neurology

Clinicul

D.yssrrophy und

Reiured Doeuses. Columhru

Uninersiy

College of Physicruns and Surgeons, New Yorh, NY (U.S.A.)

(Received

August

5th, 1980)

Summary

The wide range of values reported for activity of adenylate cyclase (AC) in human skeletal muscle prompted re-evaluation of conditions used for homogenization and assay. Adenylate cyclase activity in the same normal muscle differed with different techniques of homogenization. In pH 7.5 isotonic Tris buffer, basal and catecholamine-activated activities declined rapidly in homogenates kept at 4°C. Loss of basal activity was prevented by addition of a chelator of divalent cations. Loss of response to isoproterenol was prevented by addition of guanylnucleotides. Enzyme activity was maximal at 37°C and pH 7.6. Enzyme activity was lower when theophylline was used to prevent degradation of labelled 3’,5’ cyclic adenosine monophosphate (cyclic AMP) than when unlabelled cyclic AMP was used to this purpose. Basal activity increased with increased MgCl, concentration up to 50 mmol/l, but isoproterenol-activated activity was maximal at 4 mmol/l MgCl,. AC was inhibited by exogenous adenosine, but addition of adenosine deaminase to the assay mixture did not increase AC activity. Based upon these observations, standardized procedures of homogenization and assay were devised and used to measure AC activity in muscles of boys with Duchenne muscular dystrophy: basal and isoproterenol-stimulated activities were abnormally low.

Introduction

The small amount of skeletal muscle obtained by biopsy for study of human diseases often makes it necessary to study enzymes without isolating them. For membrane-bound enzymes, the isolation of the membranes also risks loss of cytoplasmic or membrane-bound regulators of the enzymes.

* Recipient of a postdoctoral fellowship from the Muscular Dystrophy Association. ** Recipient of a Teacher-Investigator Award from NINCDS. Address correspondence to: C.G. Cerri M.D., Department of Neurology, College Surgeons, Columbia University, 630 W 168th New York, N.Y. 10032, U.S.A.

of Physicians

and

134

TABLE

I

ASSAY MIXTURE

CONDITIONS

EMPLOYED

FOR AC ASSAY IN HUMAN

SKELETAL

MUSCLE

First author

Mawatari

Susheela

Canal

Khokhlov

Takahashi

Tris (mmol/l) Mg (mmol/l) Inhibitor (mmol/l) ATP (mmol/l) Reg. syst.

40 3.3” 2A 1.0

ns. 3.3a 20T 2.0

80 3.0” IOT 1.0

10

5.0h 5T I.0

50 5.0b 10T 1.0

9v 7.4 10 2.9/16.1

Py

41 7.5 5 5.4/n.s.

PY 7.4 5-20 4.1/256’

4, 7.5 15 20/29

PH Time (mm) Activity (basal/catec)

KS.

n.s. n.s/ I .08

“MgSO.,. ‘MgClz. ’ 1.0 mmol/l EDTA, 10 mmol/l KC1 and I mmol/l d 1.5 mmoI/l EGTA added to the medium. Py, phosphoenol pyruvate/pyruvate kinase. A, CAMP. T, theophylline. n.s., not stated. Activity expressed as picomol non-collagen protein).

DTT added

d

to the medium.

of CAMP formed per min per mg protein

(Susheela

used total, all others

Although there have been several studies of AC activity in mammalian skeletal and heart muscle [l-2], there has been no comprehensive study of techniques for preparation of homogenates or of optimal assay conditions for human skeletal muscle. Rather, methods used for mammalian muscle were used directly for human muscle, with slight variation in different laboratories. Probably as a consequence of the diversity of techniques used (Tables I, II), values reported for AC of normal muscle varied widely [3- 111. Consequently, the reported abnormalities of AC in muscle of patients with Duchenne dystrophy [3-71 are problematic. We therefore reviewed common methods of muscle homogenization and evaluated conditions of assay of this enzyme in normal human and Duchenne skeletal muscle.

TABLE

II

HOMOGENIZATION

TECHNIQUES

First author

Mawatari

Buffer (mmol/l)

Gly-Gly

(2)

USED FOR HUMAN

SKELETAL

MUSCLE

Susheela

Canal

Khokhlov

Takahashi

Tris (80)

Tris (10)

Tris (IO)

none

Plus (mmol/l)

MgSQ., ( 1)

Sucrose ( IO)

EDTA (I)

Sucrose (25)

% Homogenate

5

n.s.

5

n.s.

16

PH

7.4

7.3

8.0

1.4

ns.

Homogenizer

glass/glass

n.s.

teflon/glassa

ns. b

glass/glass

aHomogenates filtered through gauze before assay. bHomogenates spun at 5000 rpm and pellet employed ns., not stated.

for assay.

135

Materials and methods S-Guanylylimido-diphosphate (GppNHp), adenosine triphosphate (ATP), creatine phosphate, creatine kinase (CK), pyruvate kinase, phosphoenol pyruvate, guanosine triphosphate (GTP), tris(hydroxymethyl)-aminomethane (Tris), sucrose, ethylene-diamine-tetraacetic acid (EDTA), adenosine, adenosine deaminase, dithiothreitol (DTT) and ethyleneglycol-bis(amino-ethyl ether)N’N’-tetraacetic acid (EGTA), were purchased from Sigma; [LY-~*P]ATPfrom International Chemicals and Nuclear Co. or New England Nuclear and cyclic [G-3H]AMP from New England Nuclear; all other reagents were purchased from Fisher and were of the highest purity available. Normal human pectoral muscle was excised during radical mastectomies and normal human quadriceps was obtained from amputations or from diagnostic biopsies of individuals who were ultimately deemed to be free of any disease on the basis of physical examination, serum enzymes, electromyography and muscle morphology. Muscles were obtained from six patients with Duchenne dystrophy as part of routine diagnostic biopsies; none of the muscles studied were used in a previous study from this laboratory [3]. Specimens were stored at - 50°C or in liquid nitrogen until used. Control and Duchenne muscles were stored for comparable periods of time. Activity in frozen muscle was stable for at least three months. Only quickly frozen muscle was studied: muscles left at room temperature for 30 min before homogenization lost up to 50% of AC activity. For assay, frozen muscle was transferred to the laboratory in liquid nitrogen, weighed without thawing, and dropped into 19 ~01s. of ice-cold homogenization buffer, in which it was allowed to thaw for 1 min before starting homogenization. Homogenizers and tubes used for homogenization were cooled continuously in ice-chilled water. Four different buffers were compared for effects on AC in homogenates: 10 mmol/l Tris-HCl, 0.15 mol/l KCl, pH 7.4 [lo], and 10 mmol/l Tris-HCl, 0.79 mol/l sucrose pH 7.4, with or without 0.4 mmol/l EGTA or 1 mmol/l EDTA [ 111. The homogenates were kept on ice until assays were performed at 10, 30, 60 and 90 min after the start of homogenization. Reactions were started by adding incubation mixture to homogenate. AC was assayed according to Salomon [12], with minor modifications. Unless otherwise stated, the incubation mixture contained 30 mmol/l Tris-HCl pH 7.5, 5 mmol/l MgCl,, 20 mmol/l creatine phosphate, 1 mmol/l cyclic AMP, 1 mmol/l ATP, [ar-32P]ATP (3-5 X IO6 dpm/tube), 100 U/ml creatine kinase and 50 ~1 of homogenate in a final volume of 100 ~1. Reactions were carried out at 37°C for 10 min and stopped by addition of 150 ~1 of 1N HClO,. Regardless of the method of homogenization, assays were linear with time for at least 15 min and linear with protein content between 0.15 and 0.40 mg of non-collagen protein. [G-3H]cAMP (3-5 X lo3 cpm/tube) was added to the assay mixture to monitor recovery, which varied from 60% to 80%, and [32P]cAMP formed during the reaction was isolated according to the method of White and Karr [ 131. After isolation, radioactive CAMP was counted in a Mark III Searle scintillation counter, with a dual-label program correcting for the cross-talk between tritium and phosphorus. To compare different techniques of homogenization, we used the buffer containing 10 mmol/l Tris-HCl, 0.79 mol/l sucrose, and 0.4 mmol/l EGTA, pH 7.4. We evaluated three methods of tissue dispersion: pulverization in a liquid N,-cooled

136

mortar, homogenization in either an all-glass or glass-Teflon Potter-Elvehjem tissue grinder; and a combination of ultrasonic and mechanical disruption, the Polytron PT 10 generator. With both glass and Teflon pestles, the clearance was the same (0.004 to 0.006 inch). Both pestles were driven by an electric motor (Krebs Elc and Mfg) at 5000 rpm. After 15-s periods of homogenization, the temperature of the buffer did not exceed 35°C; to allow the buffer to cool to 3.0” (the starting temperature), 15 s were allowed to elapse before a second 15-s stroke. In some experiments the homogenate obtained with the Potter-Elvehjem apparatus was further disrupted by sonication with a Sonic 300 dismembrator (Artec Co.). AC activity in these experiments was assayed exactly 10 min after homogenization. An aliquot was taken from each homogenate for determination of non-collagen protein [ 14,151. For statistical analysis of differences between homogenization techniques, experiments were planned to conform to a randomized block design. Each block consisted of at least 15 determinations, each performed in triplicate. Analysis of variance among determinations under different homogenization procedures was calculated according to the method of Ostle and Mensing [16]. K, and V were determined according to the method of Cleland [17]. Experiments were performed either at constant total Mg concentration (5.0; 2.0; 1.0 mmol/l), with variable ratios among ratio. Metal-ATP interacMg * ’ , MgATP, and ATP, or at constant MgATP/Mg*+ tions were calculated according to the method of Bartfai [18], and mixing instructions for each assay mixture were obtained with the help of a computer program, accounting for binding constants for each chemical present in solution. Osmolarity and ionic strength effects were examined with the use of mannitol and NaCl; neither of these chemicals changed AC activity as determined by correlation analysis. Statistical analyses, K, computations, and calculation of molarity of chemicals were performed in a Data General Nova 800 computer. Results

Homogenization Stability in homogenates In 10 mmol/l Tris, 150 mmol/l KC1 pH 7.4, or 10 mmol/l Tris, 0.79 mol/l sucrose pH 7.4, AC activity of homogenates declined rapidly at 4°C (Fig. 1). Addition of 4 mmol/l DTT had no effect on enzyme lability in either solution. When 1 mmol/l EDTA or 0.4 mmol/l EGTA was added, however, activity was constant for at least 1 h. If EDTA or EGTA were added only immediately before assay, they failed to restore activity lost during storage. Addition of Ca2+ to final concentrations of 10 to 100 pmol/l in excess of the calculated chelating capacity of EDTA or EGTA did not cause enzyme activity to decline progressively. However, EDTA or EGTA did not prevent another consequence of storage, progressive loss of AC responsiveness to stimulation by isoproterenol. The response to isoproterenol was reduced by 25% if homogenates were kept at 4°C for 15 min before assay. Addition of 10 mmol/l GppNHp, reliably, or 100 mmol/l GTP, more variably, protected the full catecholamine response (Fig.2). No differences were found if the nucleotides were added immediately before assay or directly in the homogenization buffer.

137

TRIS

0=

SUCrOSe

.=

EGTA -sucrose

A=

DTT-sucrose

TRIS TRIS

Minutes in ice before assay Fig. I. Basal AC activity after homogenate storage Polytmn at a setting of 7.5 for 10s. Assay conditions

at 40°C before assay. Muscle homogenized as stated in “Materials and methods” section.

with

Methods of tissue disruption Using buffers with EGTA, in which AC activity was stable for 1 h after homogenization, we compared procedures of homogenization. AC activity was not detected if frozen muscle was pulverized in a mortar cooled in liquid nitrogen and the powder re-suspended in buffer at 4°C with the help of a microspatula. With the Polytron, AC activity depended upon duration and intensity of homogenization (Table III). AC activity was highest after disruption for 5 or 10 s at maximal intensity. Reduced time or force increased variance among samples, and activity tended to be lower, probably due to incomplete homogenization. Prolonged homogenization resulted in loss of activity (Table III). With all-glass homogenizers, at least 15 s were needed to disperse the tissue adequately, and activity was less than the maximal activity obtained with the Polytron. More prolonged homogenization reduced activity. The Teflon pestle did not completely disrupt the muscle in less than 30 s. However, activity obtained after using a Teflon pestle for 30-90 s was as high as the maximal activity obtained with the Polytron (Table IV). 40-60% of the activity was lost if the homogenate, obtained with any of these methods, was sonicated for 15 s at a setting of 35; no further loss was found if the homogenate was sonicated another 15 s. Incubation conditions

m 2+ Ca2+, A TP, and regenerating system for A TP AC activity was not detected in the absence of Mg. With 0.5 mmol/l MgCl, and 1 mmol/l ATP, AC activity was present, but was not stimulated by isoproterenol. At 9

138

1

0= sucrose-TRIS

30

l = EGTA-sucrose-TRIS zt= EGTA- DTT-sucro,s-TRIS

20

.=

EGTA-sucross.GppNHp-TRIS

10 i

Minutes

in

ice

Fig. 2. Isoproterenol-activated as in Fig. I.

before assay

AC activity

after storage

of homogenates

at 4°C before

assay. Conditions

1 mmol/l MgCl,, basal activity increased and stimulation by isoproterenol became evident. At 4 mmol/l MgCl,, isoproterenol-stimulated activity was maximal; at higher concentrations of MgCl 2, isoproterenol-stimulated, activity declined progressively, although basal activity continued to increase at concentrations of MgCl, above 10 mmol/l (Fig..3). No differences were found between MgSO, and MgCl,. When either the MgATP/Mg2+ ratio or the total Mg concentration was constant, the concentration of Mg *+ had no effect on the K, for MgATP (Fig. 4). The K, in presence of GppNHp was identical to the K, in basal conditions (0.28 mmol/l MgATP) When Ca2’ was added to the assay mixture or to the homogenate in amounts exceeding the chelating capacity of EGTA (unbound Ca2+ concentration 10 or 100 pmol/l), AC activity was inhibited by 20% or 60%, respectively. The effect was the TABLE

III

ADENYLATE POLYTRON

CYCLASE

ACTIVITY

AFTER

HOMOGENIZATION

Time

Force

(s)

5.5

6.5

8.5

5 IO 20 30

n.d. 5.8 6.3 I+ 0.8 6.9 + I.7

n.d. 4.X I+ 2.5 4.4 k I.3 5.2 2 0.5

10.4 1.3 10.2 9.5

IN TSEG

BUFFER

IO k i -t +

2.6 4.9 0.7 2.3

12.2 II.1 9.6 10.8

n.d., not determined. Values as picomol of CAMP synthesized per min per mg of non-collagen protein. Mean? S.D. of at least three different determinations each performed in triplicate.

k 2 2 c

1.9 1.4 I.7 0.5

WITH

A

139

TABLE

IV

ADENYLATE CYCLASE THE POTTER-ELVEHJEM

ACTIVITY AFTER APPARATUS

Time

HOMOGENIZATION

IN TSEG

Glass

Teflon

8.7 i- 0.8 8.2 -c 2.3 9.222.1 8.6 i 4.4

n.d. 12.6 2 2.0 n.d. II.9 k 2.9

BUFFER

WITH

(s) 15 30 60 90

Values expressed as picomol of CAMP synthesized per min per mg of non-collagen Mean k S.D. of at least three determinations, each performed in triplicate. n.d., not determined.

protein.

same if Ca’+ was added to the assay mixture or to the homogenate. Phosphoenol pyruvate-pyruvate kinase as a regenerating system for ATP did not differ from creatine phosphate-CK. Temperature, ionic strength, osmolarity and pH Maximal basal and isoproterenol-stimulated activities were found at pH 7.8 (Fig. 5), respectively. AC activity was highest at 38°C. At 37°C it was more active than at 30°C and 7 times more active than at 40°C. Increasing the ionic strength in the assay, either with or without changing osmolarity,

.E

35

_

30

_

7.6 and 4 times 10 times did not

25

is z" g

20_

2 z 0

15 -

Z E 10 _

0 = basal O=

0

’

1

2

3

4

mmol/l

actiwty

isoproterenol

5

stimulated

6

activity

10

11

MgC12

Fig.3. AC activity at increasing concentrations of buffer with Polytron at setting IO for 5s. Reaction 20 mmol/l creatine phosphate, 100 p/ml CK. 30 mmol/l GppNH. Bars represent standard deviation

MgCI, in the assay. Muscle.homogenate in TSEG mixture contained I mmol/l ATP, 2 mmol/l CAMP. mmol/l Tris pH 7.5. with (0) or without (0) 0.1 of samples assayed six times.

140

l

I..

1

I

I.

2

3

4

5

6

mmdl

2.

7

8.

9

6

I

I

-

10 11 12

MgATP

Fig. 4. Lineweaver-Burke double reciprocal plot of AC activity r versus substrate. Activity assayed at 37°C. pH 7.6. Data are shown for fixed MgATP/Mg*+ ratio of 2.4 (A) and for constant MgCl, content (2 mmol/l) with (0) and without (0) 0.1 mmol/l GppNHp.

affect AC activity. However, values obtained with homogenization in sucrose buffers were 30 k 0.6% lower than values obtained with equi-osmolar KCl. Inhibition ofphosphodiesterase

1.0 mmol/l was the optimal concentration of unlabelled cyclic AMP which, added to the reaction mixture, inhibited degradation of 32P-cyclic AMP. Theophylline provided maximal apparent activity of AC at a concentration of 20 mmol/l, but AC activity in the same muscle was 28% lower than with 1.0 mmol/l cyclic AMP. With 30 mmol/l theophylline, AC activity was 46% of that obtained with 1.0 mmol/l CAMP.

24 __60

q=

basal

&

isoprotarenol

activity stimulated

activity

J 7.0

7.1

7.2

7.3

7.4

7.5

76

7.7

7.6

7.9

a0

6.1

6.2

6.3

6..

6.5

6.6

PH

Fig. 5. AC activity at various pH, assayed with 2 mmol/l MgCl,, ATP-regenerating system, 30 mmol/l Tris, with (A) or without mmol/l isoproterenol.

I mmol/l ATP. 7.5 mmol/l (0) 0.1 mmol/l GppNHp

CAMP, and 0.1

141

Guanyl nucleotides and isproterenol

Activation by GppNHp or GTP was maximal at concentrations of 0.1 mmol/l (Fig. 6). GppNHp at this concentration slightly increased basal activity (1.25 fold). IsoproterenolO.1 mmol/l and GppNHp 0.1 mmol/l together increased AC activity 3-4 fold at 1 mmol/l ATP and 2 mmol/l MgCl,. Adenosine

When exogenous adenosine was added to the homogenate, AC activity declined. 1.25 mmol/l adenosine reduced AC activity by 508, but increasing the adenosine concentration up to 2.5 mmol/l did not further reduce the activity. Inhibition by exogenous adenosine was reversed by adenosine deaminase, but use of 135 mU of adenosine deaminase in the assay mixture did not increase either basal or isoproterenol-stimulated activity in the absence of added adenosine. Assay results Normal muscle

To determine values for control human muscle under optimal conditions, we measured AC activity in 14 muscles, five obtained under general anesthesia and nine with local anesthesia. We homogenized muscles to concentrations of 3-5% (w/v) in Tris-KCl-EGTA buffer with a Potter-Elvehjem homogenizer with Teflon pestle. In a final volume of 100 ~1, AC was measured at 37°C pH 7.6, with 1.0 mmol/l ATP (30- 50 cpm/picomol), 2.0 mmol/l MgC12, 20 mmol/l creatine phosphate, 100

I

I

I

-7

-6

1

-4

Guanylnucleotidcs

Fig.6.

l

= guonoslntriphosphate guonosintrlphosphate

A=

guonormlmldodiphosphots

o=

guanorinimldodiphosphate

Stimulation

by

the ratio

plu~~isoproterenol

guanylnucleotides between

-3

molarity

0:

ordinate are

1

1

-5

stimulated

plus

isoprotcrenol

of basal activity

AC

activity.

and basal

Assay

activity.

conditions

as in

Fig. 1. Units in

142

TABLE

V

Control

Duchenne

Basal activity

Isoproterenol

Ratio

14

IO.88 r+ 3.61 (6.00 ~ 16.2)

41.6 r+ 23.3 (16.8 - 101)

4.0 k 1.9 ( I .5 - 7.3)

6

3.24 -t l.609 (1.69-6.51)

5.7 2 4.X” (1.7 - 14.4)

1.9 k 0.5h (1 - 2.7)

=p< 0.005. hp
protein.

Data are

U/ml CK, 35 mmol/l Tris, 1.25 U/ml adenosine deaminase and 1.O mmol/l cyclic AMP. Type of anesthesia had no significant effect on AC activity. Basal AC activity was 10.88 ? 3.61 (6.00- 16.2) isoproterenol-activated activity, assayed in the presence of lop4 mol/l GppNHp, was 41.6 -+ 23.3 (16.8101); the increase over basal was 4.0 ? 1.9 fold. Although the small number of control muscles studied precluded detailed analysis, there was no obvious effect of age of individual, between 4 and 62 years old, or sex, on basal or isoproterenol-stimulated AC activity. Reproducibility of this method was assessed by five assays on two different days of a piece of human vastus lateralis muscle (kept frozen in liquid nitrogen). Two pieces were homogenized using a setting of 10 for 5 s on the Polytron and three with the Teflon glass homogenizer for 30 s. Values of triplicate determinations of the same homogenate varied less than 5% of the mean. The standard deviation of values around a mean of 8.20 was 1.19. Duchenne muscle Basal and isproterenol-stimulated activities in Duchenne muscle homogenized in Tris-KCl-EGTA and kept in ice were stable for at least 45 min. The mean basal activity of AC in muscle of six patients with Duchenne dystrophy was significantly lower than in muscle of controls. Response to isoproterenol was also reduced in all Duchenne muscles (Table V) in presence of GppNHp. In three patients there was no stimulation at all. Discussion There have been detailed investigations of the stability and response to hormones and ions of AC in rabbit heart plasma membrane and in mammalian skeletal muscle membranes [ 1,2], but studies of human skeletal muscle homogenates have been limited. Matawari et al. [3] used an all-glass homogenizer, which gave lower AC values in our studies than either Teflon-glass homogenizers or the Polytron. They found that catecholamine stimulation of AC was Mg-dependent, that Ca2+ inhibited AC, and that the K, for ATP was 0.3 mmol/l, a value similar to the K, for MgATP which we calculated. In all other studies of human muscle homogenates [4-6,9,10], AC stability and conditions of assay were not evaluated, and theophylline was used in the assay system. Theophylline however, inhibits AC, as shown by Severson et al. [l], Shep-

143

pard [19], and this study. This effect probably is mediated by adenosine receptors

WI. Homogenization procedure Preservation of enzyme activity

We found that techniques of homogenization and time between assay and homogenization affected AC activity. The rapid loss of activity at 4°C was probably not due to degradation by proteolytic enzymes, because the temperature was too low and the time too short. Inclusion of a proteolytic inhibitor (peptostat) failed to prevent AC degradation in cardiac sarcolemmal membrane [l]. Oxidation of the enzyme or loss of free sulfhydryl groups was probably not the cause of enzyme lability, because DTT, which preserved AC activity in isolated liver cell membranes [21,22], offered no protection in muscle homogenates. EGTA, a relatively specific Ca2+ -chelator, prevented loss of activity in muscle homogenates kept at 4°C before assay. However, addition of exogenous calcium to the homogenates in excess of the chelating capacity of EGTA or EDTA did not remove this protective effect. How these chelators preserved AC activity is therefore unexplained. Calcium may have been removed irreversibly from critical sites in the membrane, a hypothesis that cannot be tested. Coupling of the adrenergic beta-receptor to AC depends on GTP [23]; and a subunit of AC has GTPase activity [24]. GTP may be lost from the AC complex either by diffusion into and dilution in the medium, or by GTPase activity. In either case, addition of either GTP or a GTPase-resistant analogue, GppNHp, would restore catecholamine activation. Muscle disruption methods

We found that an all-glass homogenizer yielded lower AC activity than Teflonglass or a Polytron, which gave comparable results. The lower activity of the all-glass apparatus was not due to heat, because the system was carefully chilled and homogenization was brief. The “smoothness” or “roughness” of the pestle surface in the Potter-Elvehjem or the rotation speed in the Polytron probably determined the size and extent of disruption of membrane particles. AC is a multicomponent enzyme that spans cell surface membranes; disruption of membranes may result in disruption of enzyme or loss of neighboring stabilizers or activators. The deleterious effect of sonication is consistent with this suggestion. Incubation conditions Magnesium effects

Because the affinity of Mg2+ for ATP increases with pH in the range of 7.0 to 8.6, an increase of pH within this range increases the concentration of MgATP. If the concentration of MgATP was kept constant by correcting for the effect of pH on the affinity constant, the basal activity of AC did not increase with increasing pH. Dependence of enzyme activity on pH was therefore partially due to variations in concentrations of MgATP, Mg2+ , and ATP. The influence of Mg on AC activity is bimodal: the substrate for AC is MgATP [25] rather than free ATP, but with constant concentration of MgATP, V increased

144

with increased [Mg2’]. If [MS*‘] is increased when [MgATP] is held constant, [ATP] decreases. If ATP inhibits AC, decreased [ATP] could result in apparent enzyme activation. Whether free ATP inhibits AC is controversial [26], and removal of ATP inhibition could not explain the different effects of Mg on basal and isoproterenol-stimulated activity. An allosteric regulatory site for Mg*+ would explain the increase in V when free [MS*‘] increased. The different effects of Mg*+ on basal and isoproterenolstimulated activity could be explained by localization of this allosteric site for Mg*+ in the catalytic subunit of AC. Localization of this Mg2+ site on the beta-receptor or the guanylnucleotide-linked coupling subunit would be inconsistent with the lack of effect on the Hill coefficient for Mg2+ by catecholamines or guanylnucleotides [27] and with the dependence of human granulocyte AC on intracellular Mg2+ concentration [27]. The effect of Mg2+ was the same on basal and guanylnucleotidestimulated activities, which also suggests that the Mg*+ binding site is located outside the guanylnucleotide binding subunit of AC. Narayanan et al. [28] suggested that the Mg-binding site participates in the coupling between regulatory and catalytic subunits. This facilitatory role of the Mg site could explain the different effects of Mg on basal and isoproterenol-stimulated activities. Adenosine effect Two adenosine receptors, one usually stimulating and the other always inhibiting AC activity, have been identified in mammalian tissues [29]. Our data suggest that human muscle contains an adenosine receptor that inhibits AC. The lack of effect of adenosine deaminase could be explained by either a low endogenous concentration of adenosine in our assay or a higher affinity of adenosine for the adenosine receptor than to adenosine deaminase. Adenosine deaminase would have been ineffective only if the Ki of the adenosine receptor was lo4 higher than the K, of adenosine deaminase for adenosine. AC activity in normul und Duchenne human muscle The values we determined for muscle AC activity, using a standard assay based upon these studies, are difficult to compare to previous reports, because activity is so dependent on techniques of homogenization and method of assay. In all previous studies of Duchenne muscle the total activity after addition of catecholamines was abnormally low, but basal activities were either low [4], normal [3,7], or not recorded [5], and values of normal activity differed (Table I). These discrepancies were probably due in part to the lability of AC and the variety of conditions used to measure it. Reduced basal activity of AC in Duchenne muscle could represent a “denominator problem”; because fat and connective tissue are increased in Duchenne muscle, non-collagen protein is usually used, but this assumes that there is no selective loss of contractile or cytoplasmic proteins (which probably does occur to a variable extent in different cases). For this reason, the loss of stimulation by isoproterenol may be more meaningful, especially because normal fat cell AC is sensitive to activation by beta-agonists. Our data confirm the original observation by Mawatari et al. [3] that the response of AC to catecholamines is abnormal in Duchenne muscle. We found that the change in AC activity was not due to increased lability of AC in muscle homogenates, a possibility left open in other reports.

145

The lack of responsiveness of AC to catecholamines is not likely to be the fundamental fault in Duchenne dystrophy. This abnormality, however, strengthens the belief that the genetic disorder affects the muscle surface membrane, providing an altered environment for this membrane-bound enzyme. Acknowledgement

This study was supported by Center Grants Association and NINCDS (No. NS 11766).

from the Muscular

Dystrophy

References 1 Severson,

D.I., Drummond, G.I. and Sulakhe, P.V. (1972) Adenylate cyclase in skeletal muscle. Kinetic properties and hormonal stimulation. J. Biol. Chem. 247, 2949-2958 2 Snyder, F.F. and Drummond, G.I. (1978) Activation and stabilization of cardiac adenylate cyclase by GTP analog and fluoride. Arch. Biochem. Biophys. 185, 1 IO- 125 3 Mawatari, S., Takagi, A. and Rowland, L.P. (I 974) Adenyl cyclase in normal and pathologic human muscle. Arch. Neurol. 30, 96- 102 4 Canal, N., Frattola, L. and Smime, S. (1975) The metabolism of 3’.5’ cyclic AMP in diseased muscle. J. Neurol. 208. 259-265 5 Susheela, A.K., Kaul, R.D., Sachdeva, K. and Singh, N. (1975) Adenyl cyclase activity in Duchenne dystrophic muscle. J. Neurol. Sci. 24, 361-363 6 Khokhlov, A.P. and Malakhovsky, V.K. (1978) Characterization of cyclic 3’-5’ AMP changes in patients with progressive muscular atrophy. Vopr. Med. Khim. 24, 754-758 7 Willner, J.H., Cerri, C.G., Somer, H. and Rowland, L.P. (1978) IVth International Congress on Neuromuscular Diseases, Montreal Abstract 478 8 Willner, J.H., Cerri, C.G. and Wood, D.S. (1979) Malignant hyperthermia: abnormal cyclic AMP metabolism in skeletal muscle. Neurology 29, 57 9 Takahashi, K., Takao, H. and Takai, T. (I 978) Adenylate cyclase in Duchenne and Fukuyama type muscular dystrophy. Kobe J. Med. Sci. 24, 193- 198 10 Reddy. N.B., Oliver, K.L., Festoff, B.W. and Engel, W.K. (1978) Adenylate system of human skeletal muscle. Biochim. Biophys. Acta 540, 371-388 11 Ivengar, R., Swartz, L.T. and Birnbaumer, L. (1979) Coupling of glucagon receptor to adenylate cyclase: requirement of a receptor-related guanylnucleotide binding site for coupling of receptor to the enzyme. J. Biol. Chem. 254, 1I19- 1123 12 Salomon, Y., Londos, C. and Rodbell, M. (1974) A highly sensitive adenylate cyclase assay. Anal. Biochem. 58, 54 I - 548 13 White, A.A. and Karr, D.B. (1977) Improved two step method for assay of adenylate and guanylate cyclase. Anal. B&hem. 85.451-460 14 Lilienthal, J.L., Zierlor, K.L. and Folk, B.P. (1950) A reference base and system for analysis of muscle constituents. J. Biol. Chem. 185. 501-508 N.J., Farr, A.L. and Randall, R.J. (1951) Protein measurement with the 15 Lowry, O.H., Rosebrough, folin-phenol reagent. J. Biol. Chem. 193, 265-275 Iowa University Press, Ames 16 Ostle, B. and Mensing, R.W. (1975) Statistic in Research, pp. 289-372, 17 Cleland, W.W. (1972) The statistical analysis of enzyme kinetic data. Adv. Enzymol. 7 1, I- 29 18 Bartfai, T. (1979) Preparation of metal-chelate complexes and the design of steady state kinetic experiments involving metal nucleotide complexes. In: Advances in Cyclic Nucleotide Research, Vol. IO (Brooker, G., Greengard, P. and Robison, G.A., eds.), pp. 129-242, Raven Press, New York stimulated adenyl cyclase by theophylline. Nature 19 Sheppard, H. (1970) Inhibition of norepinephrine (Lond.) 228, 5676-568 D.M.F., Schlegel, W. and Rodbell, M. (1978) Adenosine analogs inhibit 20 Londos, C., Cooper, adipocyte adenylate cyclase by a GTP-dependent process:basis for actions of adenosine and methylxanthines on cyclic AMP production and lipolysis. Proc. Natl. Acad. Sci. U.S.A. 75, 5362-5364 21 Neer, E.J. (1976) Interaction of soluble brain adenylate cyclase with manganese. J. Biol. Chem. 254, 2089-2096

146

22 Lin, M.C., Welton, A.F. and Berman, M.F. (1978) Essential role of GTP in the expression of adenylate cyclase activity after cholera toxin treatment. J. Cycl. Nucl. Res. 4, 159- 168 23 Nambi, P. and Drummond, G.I. (1979) Catecholamine and guanine nucleotide activation of skeletal muscle adenylate cyclase. Biochim. Biophys. Acta 583, 287-294 24 Levitzki, A. (1978) The role of coupling of adenylate cyclase to hormone receptors and its modulation. Biochem. Pharmacol. 27, 2083-2088 25 Garbers, D.L. and Johnson, R.A. (1975) Metal and metal-ATP interactions with brain and cardiac adenylate cyclase. J. Biol. Chem. 250, 8849-8856 26 Rodbell, M., Lin, M., Salomon, Y., Londos, C., Hartwood, J.P., Marton, B.R., Rendel, M. and Bermann, M. (1975) In: Advances in Cyclic Nucleotide Research, Vol. 5 (Drummond, G.I.. Greengard. P. and Robison, G.H., eds.), pp. 3-29, Raven Press, New York 27 Stoic, V. (I 979) Control of adenylate cyclase by divalent cations and agonists. Biochim. Biophys. Acta 569. 261- 276 28 Narayanan, N., Wei, J.W. and Sulakhe, P.V. (1979) Differences in the cation sensitivity of adenylate cyclase from heart and skeletal muscle: modification by guanyl nucleotides and isoproterenol. Arch. Biochem. Biophys. 197, 18-29 29 Premont, J., Guillon, G. and Bockaert, J. (1979) Specific Mg2+ and adenosine sites involved in a bireactant mechanism for adenylate cyclase inhibition and their probable localization on this enzyme catalytic component. Biochim. Biophys. Res. Commun. 90. 5 13- 5 19