Annual Reviews
www.annualreviews.org/aronline
Annu. Rev. Biochem. 1980.49:533-564. Downloaded from arjournals.annualreviews.org
by LIBRARY CONTINUATIONS on 01/30/06. For personal use only.
Ann.Rev.Biochcm.
1980.49;533-64
Copyright
©1980by Annual
Reviews
Inc. All rightsreserved
BIOCHEMICAL PROPERTIES
OF HORMONE-SENSITIVE
1ADENYLATE CYCLASE
Elliott
012051
M. Ross
Departments of Pharmacologyand Biochemistry, University of Virginia School
of Medicine, Charlottesville, Virginia 22908
Alfred
G. Gilman
Department of Pharmacology, University of Virginia School of Medicine,
Charlottesville, Virginia 22908
CONTENTS
PERSPECTIVES
AND
SUMMARY
..........................................................................
PROTEINCOMPONENTS
OF HORMONE-SENSITIVE
ADENYLATE
CYCLASE
............................................................................................
Hormone
Receptors
and.4denylate
Cy¢lase
............................................................
Hormone
Receptors
Linkedto Adenylate
Cyclase
..................................................
SizeandShape
of Adenylate
Cy¢lase
......................................................................
Resolution
of Catalytic
andRegulatory
Proteins
....................................................
Properties
oftheCatalytic
Protein
(C)....................................................................
TheGuanine
Nucleotide-Binding
Regulatory
Protein(G/F)..................................
UNC
Lesion:Coupling
Factoror Covalent
Modification
........................................
Calcium-Dependent
Regulatory
Protein
..................................................................
Other
Protein
Factors
.................................................................... ’ ............................
Reconstitution
of Hormone-Sensitive
Activity..........................................................
534
535
535
537
538
540
542
543
545
546
547
548
1Abbreviationsused: Gpp(NH)p,
guanyl-5’-yl-imidodiphosphate;GTP-~-S,guanosine-5’O-(3-thiotriphosphate);CDR,calcium-dependent
regulatory protein (calmodulin);C, catalytic protein of adenylate cyelase; G/F, guaninenucleotide-bindingregulatory protein of
adenylatecyclase.
Weuse the word"hormone"to represent any hormone,autaeoid, neurotransmitter,or drug
that can activate adenylatecyclase in a receptor-mediatedfashion.
533
006-4154/80/0701-0533501.00
Annual Reviews
www.annualreviews.org/aronline
534
ROSS & GILMAN
EFFECTS
OFLIPIDS
AND
MEMBRANE
STRUCTURE
....................................
REGULATION
OFADENYLATE
CYCLASE
ACTIVITY
..................................
Regulation
byGuanine
Nucleotides
..........................................................................
Catecholaraine-Stiraulated
GTPase
..........................................................................
ARegulatory
GTPase
Cycle
......................................................................................
CONCLUSION
AND
QUESTIONS
............................................................................
Annu. Rev. Biochem. 1980.49:533-564. Downloaded from arjournals.annualreviews.org
by LIBRARY CONTINUATIONS on 01/30/06. For personal use only.
PERSPECTIVES
549
553
554
555
557
559
AND SUMMARY
Adenosine-3’:5’-monophosphate (cyclic AMP)is now recognized as
ubiquitous regulatory molecule, controlling diverse metabolic processes in
both prokaryotic and eukaryotic organisms. In animals, its principal role
is as an intracellular "second messenger"in the transduction of information
carried by numeroushormones,and its synthesis is catalyzed almost exclusively by the hormone-sensitive adenylate cyelase system [ATPpyrophosphate-lyase (cyclizing), E.C.4.6.1.1]. Hormone-sensitiveadenylate cyclase
activity is found in almost all animal cells (some erythrocytes and cultured
cells are exceptions) and, dependingupon the cell, can be stimulated by one
or more of a large number of hormones. These include various biogenic
amines, proteins, polypeptides, and someprostaglandins. Recently, interest
has also focused on negative hormonal control of adenylate cyclase by
opiates, c~-adrenergic amines, adenosine, and acetylcholine.
Because of the importance of cyclic AMPas a second messenger, interest
in adenylate cyclase has centered on the regulation of its activity by hormones and other ligands. As assayed in plasma membranepreparations,
adenylate cyclase displays a "basal" activity which varies enormouslyaccording to tissue and assay procedure. It is unclear whether this represents
the true activity of the unperturbed enzymeor slight stimulation of an
initially inactive enzymeby regulatory ligands (provided as impurities in the
membranepreparation or in the ATPused as substrate). This basal activity
can be elevated by the addition of appropriate hormones or analogues
thereof to the assay mixture, but the extent of hormonalactivation assayed
in vitro is generally less than that observed whenthe synthesis of cyclic
AMPis studied in intact cells or tissues. The concentration of hormones
required to stimulate the enzymeis also frequently increased after homogenization of tissues.
Hormonalstimulation of adenylate cyclase also requires the presence of
a guanine (or related purine) nucleotide in addition to substrate. This is
general requirementfor all cells studied in detail and has led to the finding
that various analogues of GTP, such as Gpp(NH)por GTP-T-S,can stimulate adenylate cyclase activity in the absence of hormones,as can GTPitself
under someconditions. Fluoride is another ubiquitously stimulatory ligand
of eukaryotic adenylat¢ cyclase. Activation usually requires greater than
millimolar concentrations of fluoride and is irreversible or only slowly
Annual Reviews
www.annualreviews.org/aronline
Annu. Rev. Biochem. 1980.49:533-564. Downloaded from arjournals.annualreviews.org
by LIBRARY CONTINUATIONS on 01/30/06. For personal use only.
HORMONE-SENSITIVE ADENYLATECYCLASE
535
reversible. Cholera toxin and several other bacterial toxins also stimulate
adenylatc cyclas¢, apparently by catalyzing the covalent modification of one
of the components of the enzyme.
This review discusses primarily the biochemical basis of these regulatory
phenomenaand howthe composition and structure of the system are mirrored in its regulation. The progress of research in this area reflects three
challenging properties of the enzyme.First, adenylate cyclase appears to be
composed exclusively of intrinsic membraneproteins and depends upon
their proper integration in a membranefor hormonal regulation. Thus,
while the catalytic, guanine nucleotide-binding, and hormone-bindingproteins may be solubilized with detergents and assayed according to their
individual activities, hormonalstimulation of adenylate cyclase is observed
onlyin intactmembranes.
Second,
theproteins
of adcnylatc
cyclasc
arc
scarce,and probably
noneexistsin a concentration
greaterthanI0
pmol/mg
membrane
protein;
somearclO0-fold
lessabundant.
Third,severaloftheproteins
arcquite
labile
indetergent
solution,
andfurther
apparentlability
isprobably
caused
by theirdissociation
during
manipulation.
Nevertheless,
thepastseveral
yearshavesccna majorincrease
in our
understanding
of theindividual
components
of thesystemandthemechanismoftheirinteraction.
Itisnowclearthathormone-sensitive
adcnylatc
cyclasc
is composed
ofat least
threeproteins:
a catalytic
protein
thatis
relatively
inactive
anddisplays
noneoftheregulatory
properties
described
above,
a guanine
nuclcotidc-binding
protein
thatmediates
theaction
of the
various regulatory ligands, and one or more hormonereceptors. Our understanding of hormonal stimulation of adcnylatc cyclasc suggests that the
regulatory protein is the proximal stimulator of the catalyst and that the
receptor-hormone complex acts by mediating the binding of pudne nucleotides to a regulatory site.
PROTEIN COMPONENTS OF HORMONE-SENSITIVE
ADENYLATE CYCLASE
HormoneReceptors and Adenylate Cyclase
It is nowclear that receptors for hormonesare indeed individual proteins,
distinct from adenylate cyclase. This idea derived first from kinetic studies
of the activation of adenylate eyclase, particularly in membranesof adipocytes. In these cells multiple hormones,which bind to different receptor
sites, seem to competefor a fixed numberof molecules of adenylate cyclase
(1, 2). Other arguments for the nonidentity of hormone receptors and
enzymehave stemmedfrom observations of their independent ontogenetic
regulation (3-6). More direct experimental approaches used chemical
genetic manipulation to resolve receptor and enzyme. Schrammshowed
Annual Reviews
www.annualreviews.org/aronline
Annu. Rev. Biochem. 1980.49:533-564. Downloaded from arjournals.annualreviews.org
by LIBRARY CONTINUATIONS on 01/30/06. For personal use only.
536
ROSS & GILMAN
that N-ethylmaleimide, which was knownto inactivate adenylate cyclase (7,
8), did not destroy the ligand-binding activity of/~-adrenergic receptors of
avian erythrocytes (8). Conversely, N,N-dicyclohexylearbodiimide inactivated binding sites for the fl-adrenergic ligand iodohydroxybenzylpindolol
at concentrations that did not inhibit the enzyme(9). Using a genetic
approach, Insel et al (10) demonstratedthat two clones of cultured cells that
are phenotypically deficient in adenylate cyclase (HTCrat hepatomaand
an $49 mouselymphomavariant) both retain/3-adrenergie receptors (10).
While each does in fact retain one of the two proteins necessary for adenylate cyclase activity, the data clearly showedthat B-adrenergicreceptors are
distinct from assayable adenylate cyclase.
A more direct demonstration that adenylate cyclase and hormonereceptors are distinct proteins camefrom the cell fusion experiments of Orly,
Sehramm,and co-workers (11-14). Sendai virus was used to fuse Friend
erythroleukemia cells to turkey erythrocytes that had been treated with
N-ethylmaleimide to inactivate adenylate cyelase. The Friend cells have
adenylate cyclase but lack B-adrenergic receptors. Plasma membranesfrom
the resultant erythrocyte-Friend cell heterokaryons displayed catecholamine-stimulatable adenylate cyclase activity. Cell fusion and membrane
preparation were performed in the presence of cycloheximide to prevent de
novo protein synthesis, which indicates that the source of stimulation was
the interaction of Friend cell enzymewith erythrocyte ~-adrenergic receptor. This receptor-enzyme interaction takes place rapidly (~ 5 min) in the
heterokaryon membrane(13). Cell-to-membranefusion rather than cell-tocell fusion can yield similar reconstitution of hormonalstimulation (14, 15),
and membranesfrom various cell types containing a variety of receptors
have nowbeen used successfully in this procedure. Thus the central postulate of the "floating receptor" modelof the regulation of adenylate cyclase
appears to be essentially accurate: hormonereceptors and adenylate cyclase
moleculesare discrete proteins, relatively free to diffuse and interact in the
plane of the bilayer.
Clear physical separation and molecular characterization of adenylate
cyclase and its related hormonereceptors were first achieved by Limbird
& Lefkowitz (16) and by Haga et al (17). These groups solubilized
adenylate cyelase activity and ~-adrenergic receptor sites with reasonably
good recoveries, and separated them both by gel exclusion chromatography
and sucrose density gradient eentrifugation in detergent solution. A similar
strategy was used by Vauquelin et al (18), whoused a $-adrenergic a~nity
matrix to separate receptor from enzyme. Separation of other receptors
from adenylate cyclase has been less clear cut, presumably for technical
reasons. Welton et al (19) showedthat the hepatic glucagon receptor did
not exactly cofractionate with adenylate cyclase during gel exclusion
chromatography. WhenDufau et al (20) chromatographed detergent ex-
Annual Reviews
www.annualreviews.org/aronline
HORMONE-SENSITIVE ADENYLATECYCLASE
537
tracts of testis and ovary on agarose gel, they foundseveral peaksof receptor
binding activity for LH/hCG,one of which coincided with adenylate cyclase. Whetherthis was fortuitous or representative of real association of
proteins is unclear, as is the significance of the multiple peaks.
Annu. Rev. Biochem. 1980.49:533-564. Downloaded from arjournals.annualreviews.org
by LIBRARY CONTINUATIONS on 01/30/06. For personal use only.
Hormone Receptors
Linked
to Adenylate
Cyclase
The ligand-binding properties of a wide variety of hormonereceptors that
act via adenylate cyclase have been studied in the membranes
of target cells
by the use of appropriate radioactive ligands; a rather large numberof
receptors have also been characterized after detergent solubilization. The
/~-adrenergic receptor has been partially purified (18, 21), and there is one
brief report on the possible purification of a receptor for LH(22). In several
cases [ADH(23), LH(24), FSH(25), PTH(26), and B-adrenergic
27, 28)] ligand binding characteristics of the soluble receptor are essentially
unaltered from those of the membrane-boundprotein. In the case of the
glucagon receptor (19), it has not been possible to bind ligands to the
solubilized protein, although a receptoroligand complexhas been solubilized
from hepatic membranesthat were first incubated with lzSI-glueagon. This
mayreflect a stabilizing effect of ligand uponthe receptor, and the relative
instability of the unliganded, detergent-solubilized protein, as was noted to
a lesser extent with the receptor for LH(24) and ADH
(23). It is interesting
that ~-adrenergic receptors solubilized with digitonin (a saponin) can bind
agonist or antagonist ligands in solution (16, 27, 28), whereaslinear alkylpolyethyleneoxidedetergents (Lubrol, Brij) have permitted only the solubilization of a receptor-ligand complex(17).
Detergent-solubilized preparations of LHand B-adrenergic receptors
have been characterized with respect to their hydrodynamicproperties, and
both appear to be rather asymmetricproteins that bind fairly large amounts
of detergent in solution (Table 1). [The data of Abou-Issa &Reichert (25)
on receptors for FSHfrom calf testis, in which no correction for detergent
binding was made,are not easily interpretable but are not grossly discrepant
from those shown.] Detergent binding to a protein can be taken to indicate
the presence of significant hydrophobic surface area, which suggests that
these molecules mayinteract with or span the hydrocarbon portion of the
plasma membranebilayer (30). However, the amount of detergent bound
per receptor molecule is in each case about that found in one micelle [110
molecules of Lubrol PX, 140 molecules of Triton X-100 (30, 31)]. This
also consistent with the interaction of a small region of the protein with a
single micelle. It is thus possible that these receptors are localized on the
outer face of the plasma membrane
and interact with the bilayer via a small
hydrophobic region, as is the case for cytochrome b5 and cytochrome b5
reductase (32).
Annual Reviews
www.annualreviews.org/aronline
538
ROSS & GILMAN
Table 1 Hydrodynamic properties
of hormone receptors
Receptor
afl-adrenergic
Annu. Rev. Biochem. 1980.49:533-564. Downloaded from arjournals.annualreviews.org
by LIBRARY CONTINUATIONS on 01/30/06. For personal use only.
Stokes radius (~)
S2o,w (S)
f/fo
M
r
Detergent bound
(g/g protein)
bGonadotropin
64
3.1
1.8
7.5 x 104
64
6.5
1.6
1.6 × 105
0.7 (Lubrol PX)
0.22 c (Triton X-100)
aFrom (17). Source: $49 lymphoma cell.
bFrom(24, 29). Sources: rat testis or ovary.
CCalculated from (29). Partial specific volumes determined by CsCi gradient centrifugation in presence of Triton X-100.
Size and Shape of Adenylate Cyclase
While adenylate eyelase was solubilized by nonionic detergents as early as
1962 (33), it has been only recently that careful hydrodynamicstudies
the size and shape of the solubilized enzymehave been undertaken. In 1974
Neer (34) used relatively standard methods to determine the sedimentation
coefficient and Stokes radius of rat renal adenylate eyelase (Table 2) and
calcdated a molecular weight of 159,000 for that enzyme. These studies
were performed in the presence of the nonionic detergent Triton X-100
Osoocytylphenoxypolyethyleneoxide), which was used to solubilize the enzyme. Adenylate cyclase displayed an apparent partial specific volumeof
0.74 in these experiments, suggesting that little detergent was boundto it.
(This is a typical partial specific volumefor a protein; that of Triton X-100
is 0.94.) Neer argued, therefore, that the renal enzyme has a minimal
hydrophobie surface area and probably does not penetrate the lipid bilayer
significantly.Bysimilarexperimental
approaches,
Neer(3 5), Hagaet al (17),
and Stengel &Hanoune(36) have shownthat adenylate eyelase from brain,
$49 lymphomacells, and liver are larger proteins (Mr "~ 2-3 X 10~). These
enzymesalso have larger apparent partial specific volumes, which suggests
¯ significant detergent binding and, hence, a greater relative hydrophobic
surface area. In an alternative approach, Schlegd et al (37, 38) used target
size theory to estimate the size of rat hepatic adenylate cyclase in intact
membranes. Whenthe enzyme was assayed with MgATPas substrate, a
target size of 2.3 - 2.4 X 10~ was obtained, in surprisingly goodagreement
with hydrodynamicmeasurements of solubilized enzyme. However, nonlinear decay curves suggested the presence of large aggregates within the
membranewith sizes similar to those determined by Houslay et al (39).
These large targets were hypothesized to represent catalytic protein in
association with multimedcregulatory protein. Detailed studies of the size
and shape of adenylate cyelase have not been undertaken with enzymefrom
manydifferent sources (el Table 2), so that the generality of these measure-
Annu. Rev. Biochem. 1980.49:533-564. Downloaded from arjournals.annualreviews.org
by LIBRARY CONTINUATIONS on 01/30/06. For personal use only.
Annual Reviews
www.annualreviews.org/aronline
HORMONE-SENSITIVEADENYLATECYCLASE
x
8
~ ~~
z
~=~
539
Annual Reviews
www.annualreviews.org/aronline
540
ROSS & GILMAN
Annu. Rev. Biochem. 1980.49:533-564. Downloaded from arjournals.annualreviews.org
by LIBRARY CONTINUATIONS on 01/30/06. For personal use only.
ments is unknown.However,the interaction of the regulatory and catalytic
proteins of the enzymederived from diverse tissues and animals (see below)
suggests that most tissues probably have structurally homologousenzymes.
Resolution of Catalytic
and Regulatory Proteins
It had been hypothesized for several years, without muchexperimental
support, that hormone-sensitive adenylate cyclase is composedof a regulatory protein in addition to a catalytic protein and hormonereceptors. When
the requirement for a guanine nucleotide for hormonal activation was
noted, it was natural that the nucleotide binding site was assumedto be on
this regulatory protein. The existence of these two distinct proteins was
recently demonstrated by two groups. Pfeuffer made use of the atfinity of
the regulatory protein for GTPto resolve it from the catalyst (42, 43) while
Ross & Gilman took advantage of somatic cell variants that were genetically deficient in one or the other protein (41, 44). Properties. of these
proteins are summarizedin Table 3.
Pfeuffer (43) found that whena Lubrol PXextract of pigeon erythrocyte
membraneswas passed over a columnof .GTP-substituted agarose, stimulation of adenylate cyclase activity in the extract was decreased whenmeasured in the presence of Mg2+ and either NaFor Gpp(NH)p.Elution of the
column with either GTPor Gpp(NH)pyielded a fraction that would combine with the unadsorbedfraction to restore activity, but that itself had
virtually no adenylate cyclase activity. If elution was performed with GTP,
stimulation of the recombined fractions by NaF could be demonstrated,
while elution with Gpp(NH)pyielded a reconstituted mixture that was
typically activated by that nucleotide. While this separation was not complete (i.e. somebasal and stimulatable activity remainedin the unadsorbed
fraction), this was the first concrete argumentfor the involvementof two
separate proteins in fluoride- or guanine nucleotide-stimulatable adenylate
cyclase activity. Since the affinity chromatographywas based on binding of
a regulatory ligand (GTP), and because the residual activity in the unbound
fraction had a differentially depleted response to fluoride and Gpp(NH)p,
Pfeuffer assumedthat it was the regulatory subunit of the enzymethat was
boundto the agarose and that the catalytic protein remained in the unbound
fraction. In this same report, Pfeuffer also showedthat solubilization of
membranesunder conditions that yielded the resolution described above
wouldalso solubilize a 42,000-dalton protein that had been labeled with the
photoattinity ligand GTP-~-azidoanilide. Circumstantial data argued
strongly that this protein was responsible for the reconstitution of activity
in the unboundfraction from the atfinity column(43, 45).
Ross & Gilmanresolved the two proteins of adenylate eyclase by virtue
of their differing thermal stabilities and their presenceor absencein different
clonal cell lines. They had previously shownthat a detergent extract of
Annual Reviews
www.annualreviews.org/aronline
HORMONE-SENSITIVE
ADENYLATE CYCLASE
541
Table 3 Protein componentsof hormone-sensitiveadenylate cyclase
Annu. Rev. Biochem. 1980.49:533-564. Downloaded from arjournals.annualreviews.org
by LIBRARY CONTINUATIONS on 01/30/06. For personal use only.
Hormonereceptor
Containshormonebinding site on extracellular face
Oneor moredifferent receptors per target cell
Catalytic Protein (C)
2+ " ATPless than 10%of activity with Mn
2+ ¯ ATP
Activity with Mg
Not stimulated by hormones,fluoride, or guanine nucleotides
Mr = 190,000
Sensitive to mild heating, low concentrationsof sulfhydryl reagents
Guaninenucleotide-binding regulatory protein (G/F)
Confers uponC ability to use MgATP
as substrate
Mediatesregulation of C’s activity by fluoride and guaninenucleotides
Binds guaninenucleotides and fluoride
Probable GTPase
Contains 41,000-45,000mol wt cholera toxin substrate
Absent in cyc- $49 lymphoma
cells
Morestable to heat or sulfhydryl reagentsthan is C
plasma membranes that contained adenylate cyclase could, under appropriate conditions, recombine with membranes of a phenotypically adenylate
cyclase-deficient
$49 lymphomacell (denoted cyc-) to yield hormone-sensitive activity (46). 2 The eye- variant ceils retain fl-adrenergic.receptors (10)
but lack adenylate cyclase activity assayable in the presence of MgATP,
which suggests that the mechanismof the reconstitution may be the interaction of solubilized enzyme with B-adrenergic receptors in or on the eyemembranes. However, thermal denaturation of the soli~ble enzymatic activity at 30° led to only slightly decreased levels of activity in the reconstituted
mixture. Thus a heat-inactivated
detergent extract of plasma membranes
could combine with the inactive eye- $49 membranesto yield relatively high
levels of adenylate cyclase activity that could be stimulated by fluoride,
Gpp(NH)p, or hormone. Similarly, the heated extract could reconstitute
soluble fluoride- or Gpp(NH)p-stimulatable activity upon combination with
a detergent extract of eye- membranes. These authors argued that the eyemembranes (or extracts therefrom) were supplying a heat-labile factor intrinsic to adenylate cyclase which was destroyed by the heating of the
complementary extract; and that the heated extract was hypothesized to
provide a second, more stable component that the eye- cells lacked. Sen2Clonesof eye- variant S49cells wereselectedfromwild-typecells byBourneet al (47). $49
cells are killed by elevatedintracellular concentrationsof cyclic AMP
(48). Thereforeagents
that activate adenylatecyelasein vivo, such as cholera toxin or fl-adrenergic agonists, may
be used for suchselection.
Annual Reviews
www.annualreviews.org/aronline
Annu. Rev. Biochem. 1980.49:533-564. Downloaded from arjournals.annualreviews.org
by LIBRARY CONTINUATIONS on 01/30/06. For personal use only.
542
ROSS & GILMAN
sitivity to proteases, sulfhydryl reagents, and temperature suggested that
both factors were proteins.
Ross et al (41) were led to propose that the more heat-labile protein,
which is retained in cyc- $49 cells, is the catalytic protein of adenylate
cyclase. This suggestion derived from the finding that cyc- membranes(or
extracts thereof) contain a Mn2+-dependentadenylate cyclase activity. The
protein that displays this Mn2+-dependentactivity has hydrodynamicproperties identical to those of the protein that is active in the reconstitution.
The two activities are also similarly thermolabile and are similarly stabilized
and labilized by a variety of mixtures of nuclcotides and divalent cations,
which suggests that the rcconstitution factor is a Mn2+-dcpcndcnt
adcnylate
cyclase. It is therefore likely that the Mn2+-dependentactivity is a nonphysiological manifestation of thc catalytic protein of adcnylatc cyclasc.
This protein is referred to as C.
Fromthese and following studies it is apparent that a minimumof two
proteins arc necessary for the expression of physiological adcnylatc cyclasc
activity and that a third protein, the receptor, is the site of hormonebinding.
If these proteins are not tightly bound to each other, the prospects of
purifying "holoadcnylatc cyclasc" maybc bleak. Homeyct al (49, 50) have
reported about 5000-fold purification of the cardiac enzymeafter stabilization of the complexwith sodiumfluoride, but this success is nearly unique
and has not yet bccn exploited further. Various claims of the purification
of adenylatc cyclasc to homogeneity(51, 52) arc not convincing, in part
because of the very low specific-activities of the products.
Properties of the Catalytic Protein
¯ Little is currently knownabout the physical properties of the catalytic
protein (C) of adenylatc cyclase. This paucity of information is duc to the
lability and difficulty of preparation of resolved C. The best characterized
preparation of C is in a crude Lubrol 12A9extract of the cyc- $49 cell
plasma membrane(41, 44). More promising preparations may result from
efforts by Pfeuffer to purify C from the unboundfraction in his affinity
chromatographic procedure (53) and from Ross’s separation of C from G/F
by gel exclusion chromatographyof a somewhatstabilized cholatc extract
of hepatic plasma membranes(E. M. Ross, in preparation). Londos ct
(54) have also reported the solubilization with deoxycholate of 2+dependentadcnylatc cyclase activity from liver, but the ability to reconstitute activity in the presence of MgATPupon addition of G/F was not
explored. Ross ct al (41) reported a molecular weight of C from cyc- $49
cells of 1.9 X 105 (see Table2), and Schlegcl et ai (38) found, by target
analysis, that Mn2+-dependent
activity from liver has a volumecorresponding to a mass of 1.5 X 105 daltons. Nothing is knownof a possible subunit
composition of C. Determination of partial specific volumemakes it seem
Annual Reviews
www.annualreviews.org/aronline
Annu. Rev. Biochem. 1980.49:533-564. Downloaded from arjournals.annualreviews.org
by LIBRARY CONTINUATIONS on 01/30/06. For personal use only.
HORMONE-SENSITIVE ADENYLATECYCLASE
543
likely that C has a relatively large hydrophobicsurface area, as is consistent
with the ability of cholate-solubilized C to be reincorporated into monolamellar phospholipid vesicles upon the removal of detergent (E. M. Ross, in
preparation).
The MnATP-dependent
adenylate cyclase activit~ of C is sensitive to
several proteases and sulfhydryl reagents (44). Evidently a second, more
reactive, cysteine residue is necessary for the interaction of C with the
regulatory protein, since N-ethylmaleimidedestroys the reconstitutive activity of C at a concentration more than tenfold below that which inhibits
catalysis (41).
C must have a substrate site for ATP,presumably as a divalent cationATPcomplex,but it is not knownif this is the samesite at whichthe effects
of nucleotides and metals on stability are exerted. There is at least one
regulatory divalent cation-binding site associated with adenylate cyclase
(55), but it need not be located on C. There is someevidence for a divalent
cation site on the regulatory protein (56).
It should be noted that there exist two other Mn2+-dependentadenylate
cyclase activities in mammalian
tissues, but neither appears to be related to
C. Mittal & Murad(57) showedthat guanylate cyclase, upon activation
free radicals, can utilize ATPas an alternative substrate and catalyze the
formation of cyclic AMP.However, guanylate cyelase from $49 cells is
more thermostable than is C (41), and Mn2+-dependentadenylate cyclase
activity attributable to C is muchhigher than is guanylate cyclase in the
same preparation of eye- membranes.A crude preparation of the guaninenucleotide-binding regulatory protein also failed to confer upon hepatic
guanylate cyclase the ability to utilize MgATP
as substrate. A soluble
Mn2+-dependent, hormone-insensitive adenylate cyclase has also been
found in testis (40, 58). However,it does not interact productively with the
regulatory protein (41) and is muchsmaller than C (40, 41).
Guanine Nucleotide-Binding
Regulatory Protein (G/F)
Muchmore is knownabout the molecular characteristics of G/F than is
knownabout C because of the former’s greater stability and ease of preparation and the ease with which it (or, at least, one of its subunits) can
radioactively labeled. Several different preparations of G/F, resolved from
C, are nowavailable in different states of purity. The G/F protein of pigeon
ery~hrocytes was initially separated from C by affinity chromatographyon
GTP-substituted agarose by Pfeuffer (43, 45) and subsequently by Spiegel
et al (59). Purification by this procedure (about 60-fold) can be improved
upon by subsequent sucrose density gradient centrifugation (43). Mammalian G/F was resolved from plasma membranesof various tissues and
cultured cells by the thermal or chemicalinactivation of C and was initially
characterized in such crude preparations (41, 44). t3/F from rabbit liver
Annual Reviews
www.annualreviews.org/aronline
Annu. Rev. Biochem. 1980.49:533-564. Downloaded from arjournals.annualreviews.org
by LIBRARY CONTINUATIONS on 01/30/06. For personal use only.
544
ROSS & GILMAN
plasma membraneshas now been partially purified (perhaps 1000-3000fold over the activity in crude extracts) following solubilization by cholate
(J. K. Northup, P. C. Sternweis, and A. G. Gilman, unpublished data).
Whilethese techniqu, es denature the activity of the catalytic protein, it is
likely that C is physically removedduring the early steps in the preparation.
G/F is also found free of C in the plasma membranesof certain clones of
HTChepatomacells and in detergent extracts therefrom (41). These cells
are phenotypically very similar to the cyc- $49 cell variants, and the existence of these two complementaryclones helps to substantiate the in vivo
requirement for both the G/F and C proteins for adenylate cyclase activity.
Initial data on the composition of G/F came from the work of Pfeuffer
(43), who used [32p]GTP-~-azidoanilide to label pigeon erythrocyte membranes. Of four specifically labeled proteins, the fractionation of a 42,000dalton band on dodecyl sulfate-polyacrylamide gels was consistent with its
involvement with adenylate cyclase. Further support for the involvement
of a 42,000-dalton protein with G/F function derives from studies using
cholera toxin. Strong evidence, reviewed in the previous volume(60), suggests that the crucial step in the stimulation of adenylate cyclase activity
by toxin is the ADP-ribosylationof the protein involved with regulation of
activity by guanine nucleotides. Using [32p]NADas the substrate for the
toxin, both Gill & Meren(61) and Cassel & Pfeuffer (45) were able to label
primarily a 42,000-dalton protein in pigeon erythrocyte membranesunder
conditions where stimulation of adenylate cyclase is optimal. ADP-ribosylation of this protein has both time and temperature dependence and requirements for GTPand a cytosolic protein that are strikingly similar to
those observed for activation of adenylate cyclase (62, 63). The 42,000dalton protein also binds to GTP-agaroseand is eluted in parallel with
reconstitutive G/F activity (45). Johnsonet al (64) have similarly labeled
wild-type $49 cell plasma membraneswith cholera toxin and [32p]NAD.
Johnson et al (65) and Howlett et al (67) demonstrated that G/F, rather
than C, was the probable site of action of the toxin. It was therefore of
interest that Kaslowet al (64, 66) could label a 45,000-dalton protein
membranesof wild-type $49 cells (phenotype +, G/F+), o r HTC cells a nd
humanerythrocytes (C-, G/F+) +,
but not in the cyc- $49 cell variant (C
G/F-), which supports the association of this protein with G/F function.
A 45,000-dalton protein in hepatic plasma membranes is also ADPribosylated by cholera toxin, and coelectrophoreses with a major Coomasie
blue-stained band in untreated, partially purified preparations of G/F. This
protein is also labeled whencholera toxin is used to label eye- membranes
that have been previously reconstituted with hepatic G/F (67a).
Initial hydrodynamicstudies of native solubilized G/F are not consistent
with the 4.2 X 104 - 4.5 X 104 molecular weight obtained in dodecyl sulfate.
Howlett et al (56) find varying molecular sizes for G/F from $49 lymphoma
Annual Reviews
www.annualreviews.org/aronline
Annu. Rev. Biochem. 1980.49:533-564. Downloaded from arjournals.annualreviews.org
by LIBRARY CONTINUATIONS on 01/30/06. For personal use only.
HORMONE-SENSITIVE ADENYLATECYCLASE
545
cells dependinguponthe presence or absence of regulatory ligands. Calculations of Mrrange from 9 X 104 - 1.3 X l05. Pfeuffer (53) has also determined
that the S2o, w of G/F varies with ligand present (GTP-’z-Sversus GDP)
a range consistent with a monomer-dimerinterconversion. Whether these
discrepancies truly represent the formation of a dimer or trimer or the
association of dissimilar proteins is uncertain. An intriguing and perhaps
related finding by Johnson et al (64) is that cholera toxin catalyzes the
ADP-ribosylation of two other proteins in wild-type $49 cell plasma membranes (Mr ~" 5.2 - 5.5 X 104) which are also absent in the cyc- (i.e’.
G/F-deficient) variant. Whether these proteins are related to G/F or not
is unclear, since humanerythrocytes display G/F activity but lack the
larger proteins as detected by cholera toxin-catalyzed ADP-ribosylation
(66). The kinetics of the thermal denaturation of G/F activity also led
speculation as to its possible subunit structure. It was found that, upon
heating at 50° C, the ability of crude G/F to reconstitute Gpp(NH)pstimulated activity decayedabout twice as fast as its ability to reconstitute
fluoride-stimulated activity. It was hypothesized that one protein, "F",
might restore Mg2+-dependent,
fluoride-stimulated activity, and that a second protein, "G", might mediate the guanine nucleotide responses of an
"F.C" complex (41, 44). The copurification of these two activities now
makesthis idea less appealing, as does the finding that GTPor a mixture
of fluoride plus ATPstabilizes both activities.
It is likely that G/F represents the binding site for ligands that activate
adenylate cyclase in the presence of MgATP.
The affinity chromatographic
procedure of Pfeuffer and his affinity labelling of a 42,000-dalton protein
both imply that G/F contains a binding site for GTP(43, 59). This is also
supported by the findings that G/F activity is stabilized by GTPor
Gpp(NH)p(41), and that its sedimentation coefficient is decreased in
presence of GTP-’/-S or Gpp(NH)p(53, 56). Similarly, fluoride plus
2+) decreases the sedimentation coefficient of
divalent cation (Mn2+ or Mg
G/F (56), whichsuggests that fluoride also binds to this protein. Effects
guanine nucleotides and fluoride upon resolved G/F are generally reversible, in contrast to their irreversible or poorly reversible effects uponintact
adenylate cyclase (67).
UNC Lesion:
Coupling
Factor
or Covalent
Modification
Selection of $49 lymphomacells in mediumcontaining/~-adrenergic agonists yields clones of a secondresistant phenotypein addition to cyc-. These
cells retain plasma membraneadenylate cyclase activity, which is stimu2+, and the intact cells
lated by fluoride or GppfNH)p
in the presence of Mg
respond to cholera toxin with increased production of cyclic AMP.However, these cells have lost responsivenessboth to/~-adrenergie agonists, the
Annual Reviews
www.annualreviews.org/aronline
Annu. Rev. Biochem. 1980.49:533-564. Downloaded from arjournals.annualreviews.org
by LIBRARY CONTINUATIONS on 01/30/06. For personal use only.
546
ROSS & GILMAN
selecting agent, and also to prostaglandins E1 and F-,2. Since they retain
fl-adrencrgic receptors, as assayed by the binding of [125I]iodohydroxybcnzylpindolol, the lesion appears as an uncoupling of enzymeand receptor-hence the name UNC,for uncoupled phenotype (68). The UNCadenylate
cyclase system is thus similar to the adenylate cyclase systems of temporarily refractory astrocytomacells (69, 70), adipocytes of hypothyroidrats
(71), and those systems artificially uncoupled by treatment with phospholipases (72), the polyene antibiotic filipin (73-76), or certain other amphiphilic compounds(73). The lesion is not total, since a small response
prostaglandins (10% as opposed to tenfold in wild type) is noted, and
isoproterenol slightly stimulates the rate of activation of the enzymeby
Gpp(NH)p(68). The cyc- lesion (loss of the (3/F protein) and
lesion are not complementarywith regard to hormonal stimulation of the
enzymeas assayed by reconstitution protocols (15, 46, 77) or in somatic cell
hybrids (78). It can be inferred, therefore, that either cyc- is deficient both
in G/F and a putative "UNCfactor" or that (3/F in UNCcells is somehow
defective. Sternweis & Gilman demonstrated that a crude preparation of
(3/F from wild-type $49 cells or rabbit liver can restore responsiveness to
hormone to UNCplasma membranes(77), and the ability to reconstitute
hormone responses in UNCmembranes cofractionates several thousand
fold with (3/F (P. C. Sternweis and A. (3. (3ilman et al, unpublished).
UNCplasma membranesare labelled with [32p]NADand cholera toxin, it
is also observed that the 45,000-dalton protein characteristic of G/F is
shifted to a moreacidic isoelectric point (67a). Theseresults, taken together,
provide a strong argument that the UNClesion represents a modification
(or lack of required modification) of the G/F protein such that it can
longer fulfill its role as a coupling factor betweenreceptor and C. The UNC
lesion also abolishes control by guanine nucieotides of the affinity of hormone-bindingto receptor (68). This loss is also restored by rcconstitution
with crude G/F (77). It is tempting to speculate that the molecular defect
found in the UNCvariant is the site of physiological regulation of G/F
function in someof those physiologically uncoupled states mentionedabove
(69-71).
Calcium-Dependent
Regulatory
Protein
The calcium-dependent regulatory protein (CDR,calmodulin) is an acidic,
low-molecular-weight, ubiquitous, Ca2÷-binding protein of mammalian
cells. It is recognized as the mediator of Ca2+-dependentcontrol of an
increasingly large numberof enzymes(79, 80). In 1975, Brostromet al (81)
and Cheunget al (82) showedthat adenylate cyclase in Lubrol extracts
cerebral cortex particulate fractions displayed a requirement for CDRfor
the stimulatory effects of Ca2+. The requirement in the extracts was demon-
Annual Reviews
www.annualreviews.org/aronline
Annu. Rev. Biochem. 1980.49:533-564. Downloaded from arjournals.annualreviews.org
by LIBRARY CONTINUATIONS on 01/30/06. For personal use only.
HORMONE-SENSITIVE ADENYLATECYCLASE
547
strated after the chromatographic removal of endogenousCDR.Elution of
endogenous CDRby EGTAcan also allow the demonstration of regulation
of membrane-boundenzymeby CDR(83-85). There is a wealth of kinetic
evidence that CDR.Ca2+ is the active species and that free CDRhas little
effect. Brostrom ct al (86) have also demonstrated an enhancement
Ca2+ of the hormone-stimulated accumulation of cyclic AMPin C6 glioma
cells, which supports the physiological relevance of the action of CDR.
CDRhas not, however, been shownto have any effect in preparations from
a number of cells whose adenylate cyclase does not normally respond to
Ca2+ (86, 87).
2+ is just beginning to be studied
The mechanismof the effect of CDR-Ca
in detail. Storm and co-workers have recently separated two fractions of
brain adenylate cyclase by virtue of the affinity of one fraction for CDRsubstituted agarose (88). They find that the species that does not bind
2+ and is also unresponsive to Gpp(NH)pand
unresponsive to CDR-Ca
fluoride. Someresponse to these ligands can be restored to this fraction by
the addition of a diluted crude membrane
extract, and the authors speculate
that t3/F is the protein that restores stimulation by t3pp(NH)p,fluoride,
2+ (89). The interaction of CDR-Ca
2+ with G/F is also consisand CDR-Ca
2+ is required
tent with the finding by Moss & Vaughan(90) that CDR-Ca
for the activation of detergent-solubilized rat brain adenylate cyclase by
2+ activates some adenylate cyclases and not
cholera toxin. WhyCDR-Ca
others is of great interest and is presumablysubject to analysis by reconstitution protocols. It is unknownwhether unresponsive systems lack the
2+ or whether CDRis so tightly
mechanism for interaction with CDR-Ca
boundto "unresponsive" cyclase as to be unremovable,as is the case with
the phosphorylase kinase-CDR complex (91). The lack of sensitivity
unresponsive systems to EGTAor to a wide range of Ca2+ concentrations
argues for the former explanation (87).
Other Protein Factors
During preparation of mammalian plasma membranes, a variable and
sometimessubstantial amountof adenylate cyclase activity or responsiveness to activators is lost. Manyinvestigators have noticed amelioration of
this loss by resuspension of the membranesin the cytosolic fraction (e.g.
92-95). Mostof these findings probably relate to loss of nucleotides or metal
ions. Pecker & Hanoune(92, 93) reported that the stimulatory activity
rat liver cytosol was at least 80%sensitive to treatment with phosphatase
and was less active than GTP. Katz et al (94) have described a cytosolic
fraction of rat liver with similar properties, but have claimed that it was
sensitive to proteases and not dialyzable. The absence of chromatographic
characterization and the lack of a linear assay makethe data difficult to
Annual Reviews
www.annualreviews.org/aronline
Annu. Rev. Biochem. 1980.49:533-564. Downloaded from arjournals.annualreviews.org
by LIBRARY CONTINUATIONS on 01/30/06. For personal use only.
548
ROSS & GILMAN
evaluate. A numberof investigators have noted alterations in the activity
of adenylate cyclase upon extraction or treatment with detergents and have
interpreted them in terms of the removal of protein factors. Bradhamfound
that extraction of rat brain membranes with Lubrol PX decreased the
responsiveness to fluoride ion of activity that remainedin the pellet, and
return of the detergent extract yielded a responsive preparation (96). However, the extract was itself stimulatory after treatment with trypsin or heat
and data on .the recovery of total activity are not available. Similar experiments by Cuatrecasas’s group, who used multiply extracted brain membranes and solubilized liver adenylate cyclase, are also unclear (97, 98).
Both groups may have removed some G/F, some CDR, or both. It cannot
be stressed too often, however,that adenylate cyclase activity and its responsiveness to stimulators are exquisitely sensitive both to detergent concentration and to the ratio of detergent to lipid and protein. This sensitivity
can be biphasic (stimulatory and inhibitory) and can be reflected differentially when different activators [e.g. fluoride, Gpp(NH)p]are used.
almost arbitrary ratio of activities assayed in the presence of these two
ligands can be achieved by manipulation of the concentration of detergents
and salts. Hence,it is absolutely essential to monitor total activity and to
use appropriate detergent and detergent plus protein controls.
Reconstitution
of Hormone-Sensitive
Activity
Reconstitution of hormone-sensitive adenylate cyclase from purified components is an obvious prerequisite to detailed studies of the mechanismof
their interaction. Progress toward this goal has so far been limited.
Sehramm’s group and others have utilized membranefusion to allow the
interaction of proteins present in the membranes
of different cells (11-15,
99) (see above). After fusion of complementarycells or of membranesfrom
different cells, the relevant proteins apparently diffuse laterally throughout
the hybrid membranesuch that they can interact productively. While protocols of this sort have potential as a valuable assay for the presence of
receptors, G/F and C, they are limited in that intact membranesmust be
used. Anattractive developmentwouldof course be the insertion of solubilized, purified proteins into monolamellarliposomes prior to reconstitution
by fusion of different liposomal preparations or of liposomes with membranes.
Ross & Gilman (41, 46) have demonstrated that membranesof cyc- $49
lymphomacells, which are deficient in G/F protein, can be reconstituted
by the addition of detergent-solubilized G/F to yield hormone-stimulatable
activity. The physical mechanismby which the G/F reassociates with the
cyc- membraneis obscure and is possibly dependent on the detergent used
to solubilize G/F. If G/F solubilized by Lubrol 12A9is mixed with cyc-
Annual Reviews
www.annualreviews.org/aronline
Annu. Rev. Biochem. 1980.49:533-564. Downloaded from arjournals.annualreviews.org
by LIBRARY CONTINUATIONS on 01/30/06. For personal use only.
HORMONE-SENSITIVE ADENYLATECYCLASE
549
membranes, hormone-stimulated activity can be assayed in the mixture.
However,if the mixture is centrifuged, G/F activity remains soluble and
can be separated from the apparently unaltered cyc- membranes(46). Stable
attachment of Lubrol-solubilized G/F to the cyc- membranesis only observed if the mixture is first incubated with an irreversible activator of the
enzyme [Gpp(NH)por a mixture of fluoride plus ATPor ADP]and yields
an irreversibly activated enzymethat cannot be stimulated further by hormones(67). If G/F is solubilized with cholate, however,mixing of G/F and
cyc- membranesin the presence of MgATP
(followed by dilution of cholate
and incubation above 4° C) yields stable binding of G/F to the membrane
without permanent activation of the enzyme. Sternweis & Gilman(77) have
used this protocol to reconstitute what is in most functional respects a
wild-type membrane, using either eye- or UNCmembranesas a starting
point.
The reconstitution of solubilized catalytic protein or of hormonereceptors into either depleted membranesor phospholipid vesicles has proven
more ditficult, and there is only one plausible report of the reconstitution
of hormone-stimulated activity using totally soluble proteins. Hoffmann
(100, 101) has described the reconstitution of dopamine-stimulated adenylate cyclase activity by a cholate-dilution protocol. The starting material is
a cholate extract of membranesfrom bovine caudate nucleus mixed with
asolectin. Other data suggest that the componentsof the system can also
be partially resolved before reconstitution and that substitution of different
detergents is feasible (101). This is quite a promising technical advance,
since, if it is generallyapplicable, it will allowthe use of increasinglypurified
proteins and lipids in reconstituted systems.
EFFECTS
OF
LIPIDS
AND MEMBRANE
STRUCTURE
It is clear that at least someof the protein componentsof the adenylate
cyclase system must be bound to an appropriate membranestructure if the
enzymeis to respond to hormones. However,the role of the structure and
composition of the plasma membranein the regulation of adenylate cyclase
activity is one of the most confusing and least carefully documentedareas
in the study of this enzyme.The basic observation is that solubilization of
the plasma membraneor the addition of any numberof agents that disrupt
membranestructure causes a loss of responsiveness to hormone. While
solubilizations of active Mg2+-dependentadenylate cyclase, C, G/F, and
receptors have been documented, there is to our knowledge no report of
soluble, hormone-stimulatableactivity in which both the specificity of the
effect of hormoneand the true solubility of the preparation have been
demonstrated. The latter criterion has been conspicuously absent from the
Annual Reviews
www.annualreviews.org/aronline
Annu. Rev. Biochem. 1980.49:533-564. Downloaded from arjournals.annualreviews.org
by LIBRARY CONTINUATIONS on 01/30/06. For personal use only.
550
ROSS & GILMAN
literature until recently; merefailure to pellet activity at 100,000 X g is
inadequate.
The role of the membraneprobably can be best considered in terms of
what a membraneprovides: a permeability barrier and a topologically
closed asymmetric surface, a mechanismfor the variable and regulatable
association and segregation of molecules within a structured millieu, a
hydrophobic environment and a stable hydrophobic-hydrophilic interface,
and the opportunity to interact with a variety of specific lipid molecules.
Each of these properties has been shownto be important for some enzyme
or transport system.
There is little evidence to suggest that coupling betweenhormonereceptors and adenylate cyclase is obligately involved with plasma membrane
transport or with the existence of a membranepotential, 3 although cyclic
AMPitself is involved in the regulation of numeroustransport processes
(105). While Moore & Wolff (73) have demonstrated that various ionophores uncouple the TSH-stimulated adenylate cyclase system at lower
concentrations than those needed to inhibit enzymatic activity, these concentrations are higher than those needed for ionophoric activity in other
membranesystems and the effect possibly reflects disruption of bilayer
structure. Wehave also found that a number of proton and cation ionophotos do not prevent stimulation of adcnylatc cyclasc by catecholamines
in membranes
of wild-type $49 cells at concentrations that arc sufficient to
uncouple oxidative phosphorylation (El M. Ross, unpublished). Since the
ionophores and such unrelated compoundsas filipin [a polyene antibiotic
that can grossly disrupt membranesby interaction with cholesterol (106)],
nonionic detergents, cholatc, and phenothiazines all cause qualitatively
similar multiphasic effects on adenylate cyclase as a function of increasing
concentration (stimulation, uncoupling of hormonalactivation, and inhibition), a primary action on the structure of the bilaycr seemsmorelikely than
does an increase in membranepermeability. Arguing for a direct relationship between ion gradients and adcnylatc cyclase, Grollman ct al (107)
showedthat thyrotropin causes hyperpolarization of cultured thyroid cells
and, in the presence of 50 mMchloride ion, of plasma membranevesicles
therefrom. This effect preceded activation of adenylate cyclase by 4-5 min.
Although the proton ionophore CCCP(2.5 ktM) abolished the membrane
potential, its effect on TSH-stimulatedadenylate cyclase was not reported;
any causal relationship between effects of TSHon membranepotential and
on adenylate cyclase is thus uncertain.
A number of membrane-bound enzymes have been shown to require a
specific lipid for activity or to be specifically influenced by one or more
3Thesituation is probablydifferentamong
fungiandbacteria(102-104).
Annual Reviews
www.annualreviews.org/aronline
Annu. Rev. Biochem. 1980.49:533-564. Downloaded from arjournals.annualreviews.org
by LIBRARY CONTINUATIONS on 01/30/06. For personal use only.
HORMONE-SENSITIVE ADENYLATECYCLASE
551
lipids, and such suggestions have been advanced for adenylate cyclase.
There exist numerousreports on the stimulatory effects of phospholipids,
particularly acidic phospholipids, on preparations of adenylate cyclase that
were previously perturbed by detergents, organic solvents, or phospholipases (108-111). Mostof these early findings, including Levey’sprovocative
reports on hormone-specific requirements for phospholipids (112, 113),
have been neither reproducednor exploited further. In general, no particular effort was madeto promote the reassociation of the lipid (frequently
added in bulk form) with the membraneor with soluble adenylate cyclase,
nor are there data to indicate whether added lipid interacted with the
membraneor merely bound detergent (including endogenous fatty acids,
lysophosphatides, etc). The best-documentedstudy of this sort is probably
that of Rubalcava &Rodbell (72), whoshowedthat treatment of rat liver
plasma membraneswith a phospholipase C with somespecificity for acidic
phospholipids abolished stimulation of adenylate cyclase by glucagon without inhibiting fluoride-stimulated activity. Concurrently, the Ka for binding
of glucagon was increased and the effect of GTPon glucagon binding was
lost (72, 76). Similar results were obtained with a phospholiPase A2.
neither ease was there a report of restoration of activity after treatment.
An alternative approach to the question of requirements for lipids has
been to alter metabolically the phospholipid composition of intact plasma
membranes.Engelhard et al (114, 115) grew LMfibroblasts in media supplemented with different ethanolamine/choline analogues. They found a
correlation betweenthe degree of substitution of the ethanolaminenitrogen
in membranephospholipids and the prostaglandin El-stimulated adenylate
cyelase activity. However,when GTPwas added to the assay the effect was
muchless striking, which suggests that variable contamination of membranes
with endogenous GTPmay have occurred. A variation of the K
m
for ATPas a function of phospholipid head group was also noted. The data
of Hirata et al (116) on effects of the methylation of phosphatidylethanolamine to phosphatidylcholine on hormonal stimulation of the enzyme are
intriguing but preliminary. A specific role of cholesterol is even less clear
than is that of phospholipids. Klein et al (117) noted a monotonicdecrease
in adenylate cyclase activity with increasing amountsof cholesterol in some
(but not other) clones of cultured kidney cells, but Hanski& Levitzki (118)
noted the opposite effect in turkey erythroeytes.
The lateral distribution and association of the protein components of
adenylate cyclase maybe crucial to hormonal regulation of adenylate cyclase. The notion of an enzyme(or perhaps free G/F, free C, and G/F plus
C) and a receptor "floating" in a fluid mosaicbilayer is probably adequate,
but their absolute freedomof motion is unclear. Physically, this motion has
not been defined at all. The apparent ability of multiple receptors to corn-
Annual Reviews
www.annualreviews.org/aronline
Annu. Rev. Biochem. 1980.49:533-564. Downloaded from arjournals.annualreviews.org
by LIBRARY CONTINUATIONS on 01/30/06. For personal use only.
552
ROSS & GILMAN
pete for a pool of adenylate cyclase molecules, as occurs in the multireceptor
adipocyte~ system(2), and the ability of receptor from one cell to activate
enzyme from another in the membraneof a heterokaryon within minutes
of fusion (13) argue for at least relative freedom. On the other hand,
Sahyounet al (119) have found that pretreatment of frog erythrocytes with
isoproterenol alters the distribution of fl-adrenergic receptors amongsubsequently prepared membranefractions, which suggests some restriction of
their lateral distribution. Vargaet al (120) were able to demonstrateclustering of MSHreceptors by labelling
Cloudman melanoma cells with
12~I-MSH,either at 0°C or after fixation with paraformaldehyde. Maguire
et al (121) have also argued against totally randomdistribution of fl-adrenergic receptors and enzymebased on their apparent stoichiometric relationship in small membranevesicles prepared from $49 cells.
The effects of the physical state of the plasma membranebilayer on the
adenylate cyclase system are being studied in increasing detail. Several
groups have tried to correlate adenylate cyclase activity and its regulation
with the physical state of endogenous membranelipids by studying the
temperature dependenceof catalysis in the presence of various activators
(113, 122-i25). Numerouslinear, upward and downwardcurved, and multiphasic Arrhenius plots have been produced, but few general conclusions
can be drawnfrom these studies. Moredirect attempts to relate the fluidity
of the bilayer to regulation of adenylate cyclase have involved enrichment
of isolated membraneswith different exogenousphospholipids. Houslay ¢t
al (126, 127) enriched hepatic plasma membraneswith synthetic phosphatidylcholines and were able to relate, qualitatively, changes in the temperatures of inflections of Arrhenius plots of glucagon-stimulated activity
with the thermotropic properties of the lipids used. No inflection was observed when fluoride-stimulated activity was measured. This study was the
first to suggest strongly that the state of the bilayer might mediate a ratelimiting process in hormonalactivation. Recently, Bakardjieva et al (128)
enriched membranesof Chang liver cells with different phospholipids and
comparedadenylate cyclase activity of membraneshigh in dimyristoylphosphatidylcholine
(DMPC) with membranes high in dipalmitoylphosphatidylcholine (DPPC)or dioleoylphosphatidylcholine (DOPC).At 37°C,
above the gel-liquid crystal transition temperature (T,~) of DMPC
and
DOPC,the membranesenriched in these lipids had lower adenylate cyclase
activity than did control preparations or DPPC-enrichedmembranes. At
17°C, below the Tm of DMPCbut above that of DOPC,the membranes
enriched in DMPC
or DPPChad roughly equivalent activities that were
higher than those of the DOPC-endchedmembranes. It thus appeared as
though a relative decrease of activity occurred above Tm. Since the differ-
Annual Reviews
www.annualreviews.org/aronline
Annu. Rev. Biochem. 1980.49:533-564. Downloaded from arjournals.annualreviews.org
by LIBRARY CONTINUATIONS on 01/30/06. For personal use only.
HORMONE-SENSITIVE ADENYLATECYCLASE
553
ence appeared to be most marked when isoproterenol-stimulated activity
was measured, binding to B-adrenergie receptors was measured and found
to selectively de~rease at temperatures abovethe T,n of the added lipid. It
is not knownwhether the changes in activity and in the numberof assayable
fl-adrenergic receptors occur by the same or different mechanisms.
A third finding with relevance to the effects of membranefluidity upon
adenylate cyclase is that of Levitzki and co-workers, whodemonstratedthat
cis-vaccenic acid causes an increase in the rate of activation of turkey
erythrocyte adenylate cyclase by Gpp(NH)pplus epinephrine. These experiments were based on the finding of Orly & Schramm(129) that A9-10
Al1-12 cis-monounsaturated fatty acids stimulate adenylate cyclase in
these membranes. Levitzki and co-workers (130, 131) found that the increase in the rate constant for activation by Gpp(NH)pplus epinephrine
was inversely related to the "microviscosity" of the bilayer, as assayed by
the fluorescence anisotropy of 1,6-diphenyl-l,3,5-hexatriene (132, 133).
Since Gpp(NH)pdoes not markedly stimulate activity by itself in these
membranes,the authors interpreted their results as indicating that collision
of receptor and enzyme, controlled by their rates of lateral diffusion and
hence by the viscosity of the membrane,was the rate-limiting factor in
activation by Gpp(NH)pplus hormone. These experiments do not of course
indicate whichmotions are limited, which proteins are involved, or even if
the apparent effect of microviscosity is on the motion or conformation of
a protein. However,taken together with the studies of Bakardjieva et al
(128) and Houslayet al 026, 127), they begin to suggest the major role that
the structure of the bilayer plays in the regulation of the system.
REGULATION
ACTIVITY
OF
ADENYLATE
CYCLASE
Most studies of the regulation of adenylate cyclase activity by hormones,
nucleotides, and other ligands have been undertaken with a view toward
understanding the physiological regulation of cyclic AMPmetabolism in a
particular tissue. Consequently, there exists a great mass of data derived
from manydifferent cells, frequently obtained using crude homogenates,
and only rarely directed toward understanding the primary biochemical
mechanismof that regulation. In this section we draw upon only a small
numberof studies in an attempt to organize what mechanistic information
is available and to correlate it with knowledgeof the individual proteins.
Regulation by divalent cations has been discussed elsewhere (55), and
still knowfrustratingly little about the molecularactions of fluoride. Hence,
Annual Reviews
www.annualreviews.org/aronline
554
ROSS & GILMAN
Annu. Rev. Biochem. 1980.49:533-564. Downloaded from arjournals.annualreviews.org
by LIBRARY CONTINUATIONS on 01/30/06. For personal use only.
we will stress the interacting effects of hormonesand guanine nucleotides
on the activity and association of G/F and C.
Regulation
by Guanine Nucleotides
Muchof our information on the regulation of hormone-stimulatable adenylate cyclase stems from observations on the effects of guanine nucleotides,
first noted by Cryer et al (134) and studied moreextensively by Rodbell and
co-workers (135-138). GTP, GDP, or ITP by themselves display diverse
inhibitory and sfimulatory effects on adenylate cyclase activity, and these
vary with cell type, pH, divalent cation concentration, ionic strength, presence of detergent, and probably a dozen other variables. The key observation, however, is that the presence of some guanine (or related purine)
nucleotide is an absolute requirement for the stimulation of adenylate cyclase activity by hormones(136). This requirement has been observed
plasma membranesof a number of different cell types (121) but is often
difficult to demonstrate due to the presence of contaminating guanine nucleotides in the ATPused as substrate (139) or in the membranepreparation
(140), and to the possible ability of ATPitself to serve this function at high
concentrations. This effect of guanine nucleotides is also suggested to be
significant in vivo by the finding that depletion of intracellular GTPby
growth of cells in the presence of mycophcnolicacid decreases the intracellular accumulation of cyclic AMPin response to hormones (141-143).
addition to the requirement of guanine nucleotides for hormonalactivation
of adenylate cyclase, the concentration of hormonenecessary to stimulate
activity is altered by the identity of the nucleotide used in the assay. Thus,
isoprotcrenol causes half-maximal stimulation of adenylate cyclase in $49
cell membranesat 100 nMin the presence of GTP, 50 nMin the presence
of ITP, 20 nMin the presence of XTF, and 15 nM in the presence of
Gpp(NH)p(at t=0; see below) (140).
Rodbell and co-workers, using [12~I]iodoglucagon, were also first to observe that those purine nucleotides that permit hormonal activation of
adenylate cyclase also frequently decrease the affinity of the receptor for
hormone (144). This effect has now been described in some detail for
receptors for glucagon (19, 76, 145), prostaglandin l ( 146, 1 47), a
B-adrenergic agonists (140, 148-150), and has also been observed with
receptors for TSH(73) and FSH(151). The effect is specific for agonist
antagonist) ligands of the receptor (148, 149). It should be noted that this
interaction of regulatory nucleotides and hormonesis not competitive but
rather is a negative heterotropic binding interaction. Thus GTP,ITP, and
Gpp(NH)p
all decrease the affinity of the receptor for hormonesto an equal
maximalextent, although with varying potency (140, 144, 147, 149), and
act by increasing the dissociation rate constant for hormone(19, 144, 150).
Annual Reviews
www.annualreviews.org/aronline
Annu. Rev. Biochem. 1980.49:533-564. Downloaded from arjournals.annualreviews.org
by LIBRARY CONTINUATIONS on 01/30/06. For personal use only.
HORMONE-SENSITIVE ADENYLATECYCLASE
555
It has beeninferred by us and others that the ability of nucleotides to alter
binding al~nity for hormonereflects association of receptor with adenylate
cyclase [most probably with G/F (41)]. This inference is based on several
observations: (a) the effect is lost whenthe enzymeis inactivated with
sulfhydryl reagents (O/F must bc inactivated--inactivation of C is insuflicient) (75, 150); (b) it is absent in UNCand cyc- variant S49 cells in
G/F is either absent or altered (68, 140); (c) it is restored by reconstitution
of UNCor eye- membraneswith crude preparations of G/F (77); (d) it
lost whenthe fl-adrenergic receptor and adenylate cyclase are uncoupled
by the polyeneantibiotic filipin (74, 75); (e) it is lost uponsolubilization
(and hence, uncoupling) of/~-adrenergic receptors (150); and (f) it is
only in those fl-adrenergic systems that display highly ctficient coupling
between hormonebinding and activation of adenylate cyclase (121).
More information on the mechanismof hormonal control came from the
use of poorly hydrolyzable analogues of GTP[Gpp(NH)p, Gpp(CH2)p,
t3TP-T-S]. Although Gpp(NH)pwas introduced as a substitute for GTP
in assays where hydrolysis of nucleotide was likely, it quickly became
apparent that it was unique in that Gpp(NH)palmost invariably activated
adcnylatc cyclasc in the absence of hormone,the activation was frequently
greater than that caused by fluoride or by GTPplus hormone, activation
was either irreversible or poorly reversible, and adenylate cyclase was much
more stable after activation (42, 138, 152-154). While hormones and
Gpp(NH)pmay appear to act synergistically,
it was shown by Bennett,
Jacobs & Cuatrecasas (155, 156) and others (140, 153) that the effect
hormoneis only to increase the rate of activation by Gpp(NH)p
rather than
to increase the final activity attainable. The intensively studied turkey erythrocyte enzyme, where activation by Gpp(NH)pappears to be totally
dependent on hormone, is probably only an extreme case (129, 153).
have already stressed that the irreversibility of activation by Gpp(NH)p
and
the effect of hormonesupon the rate of activation makethe quantitative
study of the kinetics of activation by this nucleotide somewhattreacherous
(121). There is, in essence, no defined Kact for hormoneunder these conditions unless extrapolation to zero time is made(140), and few have tried
quantitate accurately the effects of hormoneon activation kinetics (131,
140, 157-159). Furthermore, any calculation of an incremental activity due
to hormoneis a function of time. Manymisleading conclusions have been
generated by the application of equilibrium assumptions to the analysis of
data gathered using Gpp(NH)p.
Catecholamine-Stirnulated
GTPase
In spite of experimental pitfalls,
the data obtained with Gpp(NH)p,
Gpp(CH2)p,and ITP led a numberof investigators to speculate on the role
Annual Reviews
www.annualreviews.org/aronline
Annu. Rev. Biochem. 1980.49:533-564. Downloaded from arjournals.annualreviews.org
by LIBRARY CONTINUATIONS on 01/30/06. For personal use only.
556
ROSS & GILMAN
of either GTPhydrolysis (140, 160, 166) and/or phosphotransferase reactions (161) as a mechanismof regulation of adenylate cyclase. Moreimportantly, it led Cassel & Selinger (162) to discover a catecholamine-stimulated
GTPaseactivity in turkey erythrocyte membranesand to relate this activity
to the regulation of adenylate cyclase. The specific activity of the catecholamine-stimulated GTPase is low (less than 10 pmol min-1 mg-l) and is
assayed above a somewhat larger basal activity. App(NH)pmust be included in the assay to inhibit nonspecific nucleoside triphosphatases, and
even this strategy is insufficient to assay the enzymein hepatic plasma
membranes,where background activity is 50-fold greater than that in turkey erythrocyte membranes. Nevertheless, these authors showed that the
pharmacological specificity, Kmfor GTP, and inhibition by either phospholipases, detergent, or GTP-T-Swere all consistent with the involvement
of this activity with the hormonalcontrol of adenylate cyclase (162, 163).
Perhaps their most provocative finding, however, was that treatment of
plasma membraneswith cholera toxin markedly inhibited the catecholamine-stimulated GTPaseactivity without decreasing the "basal" or background level (164). The dependenceof this inhibition on the concentrations
of toxin and NAD
is similar to that observed for the activation of adenylate
cyclase. The decrease in GTPaseactivity is also consistent with the observations of others that cholera toxin alters the regulation of adenylate cyclase
by GTPso as to make GTPappear similar to the poorly hydrolyzablc
Gpp(NH)p;i.e. treatment with toxin causes activation by GTPto be greater
and longer lasting, increases the potency of hormonein the presence of
GTP, and decreases the incremental activity caused by hormones (140,
165-168).
These data led Cassel &Selinger to propose an explicit model for the
regulation of adenylate cyclase by hormones
andguaninenucleotidesbased
upon the binding and hydrolysis of GTP(164, 165); the model was quite
similar to the proposal of Levinson &Blume (166). Drawingheavily upon
knowledgeof the interaction of the prokaryotic translational elongation
factors Tu and Ts (169), Cassel & Selinger proposed a regulatory GTPase
cycle in which GTP-ligandedadenylate cyclas¢ was the active species and
hydrolysis of GTPto GDPwas the primary mechanism of inactivation.
Fractional activation of adenylate cyclase wouldthus be proportional to the
steady-state concentration of the GTP-liganded state. According to the
model, inhibition of GTPhydrolysis by cholera toxin would sustain activation, as would the binding of a nonhydrolyzable GTPanalogue such as
Gpp(NH)p. Conversely, hormone was proposed to promote binding
GTP, thereby stimulating both GTPaseactivity and adenylate cyclase activity. This facilitation of GTPbinding was suggested to represent the
"opening" of the binding site such that GDPcould dissociate and GTP
could bind. Using this model, Cassel & Selinger have been able to relate
Annual Reviews
www.annualreviews.org/aronline
Annu. Rev. Biochem. 1980.49:533-564. Downloaded from arjournals.annualreviews.org
by LIBRARY CONTINUATIONS on 01/30/06. For personal use only.
HORMONE-SENSITIVE ADENYLATECYCLASE
557
activation and inactivation rate constants for adenylate cyclase to steadystate activities of the enzyme(165, 170). As expected from their model,
cholera toxin decreases the inactivation rate constant without altering the
rate of activation by GTPplus hormone(165), and membraneslabeled with
[3H]GTPcould be shown to release [3H]t3DP upon the addition of hormone (171). Simultaneous addition of Gpp(NH)pplus hormone to
membranesled to the concomitant release of [3H]GDPand reactivation of
enzyme (172).
At this point, it should be mentioned that hormone-stimulated GTPase
activity still has been assayed only in turkey erythrocytes and only with
/~-adrenergic agonists. No rigorous proof of the coupling of GTPhydrolysis
to the inactivation of adenylate cyclase exists. Such proof will probably be
available only whenthe activities of purified G/F, C, and receptors can be
studied in properly reconstituted systems. The supporting evidence is striking, however, and the concept of a regulatory GTPasecycle has provided
a motivating and organizing hypothesis for the analysis of the regulation of
adenylate cyclase; such a paradigmhas been absent for twenty years.
A Regulatory
GTPase Cycle
The elegant studies by Cassel & Selinger and data from other laboratories
can nowform the basis of a muchmore detailed description of the regulatory GTPasecycle. In this section we will try to relate the steady-state’ and
kinetic data on the binding of regulatory ligands to their possible effects on
the activities and interactions of the proteins discussed in earlier sections.
Wewill stress, as a unifying concept, the positive and negative effects of the
binding of one ligand to a componentof the adenylate cyclase system upon
the binding of a second ligand or protein and the effects of such binding
upon enzymatic activity.
A schematic diagram of the proposed cycle is shown in Figure 1. The
central features are those suggested by Cassel &S elinger (165) and Levinson
& Blume(166). Details are as follows:
1. It is assumedthat G/F is the only site of binding of GTPand is the
site of GTPhydrolysis. Rodbell and co-workers (173) have proposed multiple binding sites for GTP,but their argumentsdo not seem to us compelling
whenit is realized that free G/F and G/F bound to receptors (R) or to
can have strikingly different properties. It is assumed G/F-GTPis the
active species and is the proximal and sole stimulator of the adenylate
cyclase activity of C. (G/F-GTP-R-His presumably too short-lived to be
significant.) At present, there are few data to suggest whether GTPpromotes the association of G/F and C (i.e. that the binding of C and guanine
nucleotides to G/F is positively cooperative) or whether the nucleotide
merely activates a preexistent G/F-Ccomplex. Pfeuffer’s demonstration of
an increased sedimentation coefficient of G/F in the presence of both active
Annual Reviews
www.annualreviews.org/aronline
558
ROSS & GILMAN
Annu. Rev. Biochem. 1980.49:533-564. Downloaded from arjournals.annualreviews.org
by LIBRARY CONTINUATIONS on 01/30/06. For personal use only.
_
~/ ’,~
t; ~"--#GIF’GTP’-
kI
G/F-GTP.R’H
"- GIF’GDP
G/F’GDP-R.H
G/F.N.N/k5
Figure
] A hypothetical
regulatory
GTPase cycle
in which hormone-receptor
complex
(H.R)acts bycatalizingthe displacement
of GDP
fromG/Fby means
of negativeheterotropic
binding.
C and GTP-T-$suggests that GTP-~-Spromotes stable binding of 12 to
O/F (53). Schlegel et al (38) found increases in the apparent size of adenylate cyelase upon treatment with Gpp(NH)p,which also argues for a positive interaction of the binding of (2 and GTPto ~/F. The data of Limbird
et al (1"/4) suggest, however, that the G/F-C complex may be relatively
stable in the unliganded state.
2. Hydrolysis of GTPby G/F (k 1) causes the inactivation of adenylate
cyclase, again as proposed by Cassel & Selinger. C maydissociate at this
point. NormallykI is not rate-limiting, but the use of a nonhydrolyzable
GTPanalogue or treatment of G/F with cholera toxin decreases kl (165),
which increases the concentration of G/F-GTP.It is unknownif free G/F
has GTPaseactivity or whether the activity is expressed only when G/F
is associated with another protein (e.g. C). In an analogous system, free
elongation factor Tu is nearly inactive but becomesan active GTPasewhen
bound to the ribosome (169). However,Cassel & Selinger (162) were
to inactivate adenylate cyclase with N-ethylmaleimidewithout inhibiting
GTPaseactivity.
3, In the absence of hormone,k_2 is the rate-limiting step in the regeneration of G/F-GTP,and k2~ k_2. In turkey erythroeytes, k_2 "~ 0, so that
after one cycle of GTPhydrolysis all G/F is in the G/F-GDPform and
adenylate cyclase is inactive. Gpp(NH)palone cannot activate adenylate
cyclase in turkey erythrocytes because there is no free G/F. In mammalian
calls, where Gpp(NH)palone does activate the enzyme, k_2 is assumed
be finite but low. In $49 lymphomacell membranes, k_z can be assumed
to be at least 0.06 rain-I, the rate constant for activation of adenylateeyclase
by Gpp(NH)p (41).
Annual Reviews
www.annualreviews.org/aronline
Annu. Rev. Biochem. 1980.49:533-564. Downloaded from arjournals.annualreviews.org
by LIBRARY CONTINUATIONS on 01/30/06. For personal use only.
HORMONE-SENSITIVE ADENYLATECYCLASE
559
4. This model differs substantially from that of Cassel &Selinger only
in the mechanismby which hormoneis proposed to facilitate regeneration
of G/F-GTP, and reflects the suggestion of Blume and co-workers (160,
166) that dissociation of GDPrather than binding of GTPis the regulated
step. It is generally compatible with the concept of "collision coupling"
suggested by Levitzki and co-workers (158). Wesuggest that the only
relevant action of hormoneis the negative heterotropic nature of its binding
to G/F-R with respect to guanine nucleotides. By this we mean that creation of a G/F-R-Hcomplexdecreases the affinity of G/F-Rfor nucleotide.
Webase this suggestion on the negative heterotropic effect of guanine
nucleotides upon hormonebinding, where finite increases in both the Kd
(144-150) and the dissociative rate constant (144, 147, 150) for hormone
have been documented. From thermodynamic considerations,
such an
effect must be reciprocal; i.e. if nucleotide increases the Kd for hormone,
then hormonemust increase the Kd for nucleotide. This concept is consistent with the demonstration by Cassd & Selinger of the catecholaminestimulated dissociation
of [3H]GDP or [3H]Gpp(NH)p from turkey
erythrocyte membranes(171, 172); also consistent is the requirement for
hormonenoted by Pfeuffer (42, 43) and Spiegel et al (59) to allow
exchangeofGDP(a tight ligand) with GMP
(a loose ligand) prior to affinity
chromatographyof G/F. It should be noted that the reaction path k4" k_5
¯ k6’k_7 is thermodynamicallyequivalent to the path k_2¯ k3. The receptorcatalyzed path is kinetically muchfaster. Formation of the unstable G/F-RH-GDPcomplex promotes the rapid dissociation (k_5) of what would
otherwise be a slowly dissociating (k_2) nueleotide ligand. [See (175) for
detailed discussion of the general kinetic implications of negative heterotropic binding.]
It can be questioned whysuch diverse effects of purine nueleotides are
observed if only one binding protein is involved (173). It is dear that G/F
behaves differently when free or whencomplexedwith C. Thus activation
of adenylate cyelase is generally irreversible when intact membranesor
unfraetionated extracts are treated with Gpp(NH)por fluoride but the
effects of these ligands uponfree G/F are reversible (56, 67). Similarly, the
reversibility of the effect of Gpp(NI-I)p upon hormonebinding (140),
opposedto its irreversible activation of enzyme,can be assumedto reflect
action of that nueleotide on C~/F-R as opposed to G/F-C. A number of
other discrepancies (173) are similarly explicable. Onemust alwaysconsider
whether free G/F, G/F-R, or G/F-Cis the species under study in a particular experiment.
CONCLUSIONS
AND QUESTIONS
As we can describe it at this time, hormone-sensitive adenylate eyelase
appears to be composedof at least three interacting proteins. The catalyst
Annual Reviews
www.annualreviews.org/aronline
Annu. Rev. Biochem. 1980.49:533-564. Downloaded from arjournals.annualreviews.org
by LIBRARY CONTINUATIONS on 01/30/06. For personal use only.
560
ROSS & GILMAN
by itself is relatively inactive, but can be stimulated by a guanine nucleotidebinding protein,
G/F, when G/F binds GTP. Upon hydrolysis
of GTP,
activation is terminated until G/F-GTP is regenerated. This regeneration
is catalyzed by the receptor-hormone complex, in that the rate of dissociation of GDPis stimulated by the binding of the hormone-receptor complex
to G/F. There is as yet no evidence for the direct interaction of receptor
with the catalyst. Evidently, the interaction of G/F with receptor can occur
only in a relatively
unperturbed membrane, since only then can hormonal
stimulation of C be observed.
The study of the biochemistry of adenylate cyclase is just beginning.
Details are missing and questions proliferate.
The kinetics and thermodynamics of the multiple interactions
of ligands and proteins are yet to be
quantitated. The variability
of hormonal responses among different tissues
is yet to be related to these basic parameters in any meaningful way. The
permissive role of an intact plasma membrane for the interaction
of G/F
and R has barry been approached. What is the mechanism of action of
fluoride on G/F?. What is the nature and significance of the hypothetically
modified G/F in the UNCvariant? Beyond this level, biochemical studies
of long-term regulation (refractoriness)
of adenylate cyelase by hormones
have been initiated in several laboratories. Howare the synthesis and stoichiometry of G/F, C, and receptors coordinated? We are just beginning to
learn to phrase these questions in a meaningful way.
ACKNOWLED(3MENTS
We would like to thank many of our colleagues
publications.
We thank Mrs. Wendy Deaner for
the manuscript.
Our own studies have been
GM26445, AM22125, and NS 10193, and by
Grant BC240.
for sending us preprints of
help in the preparation of
supported by USPHSgrants
Amerlean Cancer Society
Literature Cited
I. Perkins, J. P. 1973.Adv.CyclicNucleotide Re~3:1--64
2. Birnbaumer,L., Pohl, S. L., Krans,H.
M.J., Rodbell,M. 1970.Adv. Biocher~
PsychopharmacoL
3:185-208
3. Harden,T. K., Wolfe,B. B., Sporn,J.
R., Perkins,J. P., Molinoff,P. B. 1977.
Brain Re&125:99-108
4. Charness, M.E., Bylund,D. B., Beckman,B. S., Hollenberg,M. D., Snydcr,
S. H. 1976. Life ScL 19:243-50
5. Bilezekian, J. P.., Spiegel, A. M.,
Brown,E. M., Aurbach, G. D. 1977.
Mol. Pharmacol.13:775-85
6. Harden,T. K., Foster,S. J., Perkins,J.
P. 1979. d. Biol. Chem.254:4416-22
7. Oye, I., Sutherland, E. W.1966. Biochim. Biophys. Acta 127:347-54
8. Schramm,M., Naim, E. 1970. J. BioL
Chem.245:3225-31
9. Schramm,
M. 1976. J. Cyclic Nucleotide
Re~ 2:347-58
10. Insel, P. A., Maguire,M. E., Gilman,
A. G., Bourne, H, R., Coffino, P.,
Melmon,K. L. 1976. MoLPharmacoL
12:1062-69
11. Orly, J., Schramm,
M.1976. Proc. NatL
Acad. Sci. USA73:4410-14
Annual Reviews
www.annualreviews.org/aronline
Annu. Rev. Biochem. 1980.49:533-564. Downloaded from arjournals.annualreviews.org
by LIBRARY CONTINUATIONS on 01/30/06. For personal use only.
HORMONE-SENSITIVE
12. Schramm, M., Orly, J., Eimed, S.,
Korner, M. 1977. Nature 268:310-13
13. Schulster, D., Orly, J., Seid¢l, G.,
Schramm, M. 1978. J. Biol. Chert
253:1201-6
14. Schramm, M. 1979. Proc. Natl. Acad.
Sci. USA 76:1174-78
15. Schwarzmeier, J. D., Gilman, A. G.
1977. J.
Cyclic Nucleotide Res.
3:227-38
16. Limbird, L. E., Lefkowitz, R. J. 1977.
Z Biol. Chem. 252:799-801
17. Haga, T., Haga, K., Gilman, A. G.
1977. J. Biol. Chem. 252:5776-82
18. Vauquelin, G., Geynet, P., Hanouae,J.,
Strosberg, A. D. 1977. Proc. Natl..4cad.
Sci. USA 74:3710-14
19. Welton, A. F., Lad, P. M., Newby,A.
C., Yamamura, H., Nicosia, S., Rodbell,
M. 1977. J. Biol. Chem.
252:5947-50
20. Dufau, M. L., Baukal, A. J., Ryan, D.,
Catt, K. J. 1977. Mol. Cell. Endocrinol.
6:253-69
21. Caron, M.(~., Srinivasan, Y., Pitha, J.,
Kociolek, K., Lefkowitz, R. J. 1979. J.
Biol. Chem. 254:2923-27
22. Dufau, M. L., Ryan, D. W., Bankal, A.
J., Catt, K. J. 1975. J. Biol. Chem.
250:4822-24
23. Roy, C., Rajedson, R., Bockaert, J.,
Jard, S. 1975. J. Biol. Chem.
250:7885-93
24. Dufau, M. L., Charreau, E. H., Catt, K.
J. 1973. J. Biol. Chem. 248:6973-82
25. Abou-Issa,
H., Reichert,
L. E. Jr. 1977.
J. Biol. Chem.
252:4166-74
26. Malbon,C. C., Zull, J. E. 1977. J. Biol.
Chem. 252:1079-83
27. Caron, M. G., Lefkowitz, R. J. 1976. J.
Biol. Chem. 251:2374-84
28. Limbird, L. E., Lefkowitz, R. J. 1978.
Proc. Natl. dcad. Sci. USA 75:228-32
29. Dufau, M. L., Charreau, E. H., Ryan,
D., Catt, K. J. 1974. FEBS Lett.
39:149-53
30. Tanford, C., Reynolds, J. A. 1976. Biochim Biophy~ Acta 457:133-70
31. Helenius, A., Simons, K. 1975. Biochim. Biophys. Acta 415:29-79
32. Strittmatter, P., Rogers,M. J., Spatz, L.
1973. J. Biol. Chem. 247:7188-94
33. Sutherland, E. W., Rail, T. W., Menon,
T. 1962. J. Biol. Chert 237:1220-27
34. Neer, E. J. 1974. J. Biol. Chem.
249:6527-31
35. Neer, E. J. 1978. J. Biol. Chem.
253:1498-1502
36. Stengel, D., Hanoune,J. 1980. Eur. J.
Biochem. 102:21-34
37. Kempner, E. S., Schlegd, W. 1979.
Anal. Biochem. 92:2-10
ADENYLATE
CYCLASE
561
38. Schlegel, W., Kempner,E. S., Rodbell,
M. 1979. Z Biol. Chert 254:5168-76
39. Houslay, M. D., Ellory, J. C., Smith, G.
A., Hesketh, T. R., Stein, J. M., Warren, G. B., Metealfe, J. C. 1977. Biochira. Biophy~ Acta 467:208-19
40. Neer, E. J. 1978. J. Biol. Chert
253:5808-12
41. Ross, E. M., Howlett, A. C., Ferguson,
K. M., Gilman, A. G. 1978. J. Biol.
Chera. 253:6401-12
42. Pfeuffer, T., Helmreich, E. J. M. 1975.
Z Biol. Chem. 250:867-76
43. Pfeuffer, T. 1977. J. Biol. Chem.
252:7224-34
44. Ross, E. M., Gilman, A. G. 1977. Z
Biol. Chera. 252:6966-6970
45. Cassd, D., Pfeuffer, T. 1978. Proc. Natl.
Acad. ScL USA 75:2669-73
46. Ross, E. M., Gilman, A. G. 1977. Proc.
NatL dcad. Sci. USA 74:3715-19
47. Bourne, H. R., Coffmo, P., Tomkins, G.
M. 1975. Science 187:750-52
48. Daniel, V., Litwack, G., Tomkins, G.
M. 1973. Proc. NatL Acad. ScL USA
70:76-79
49. Homey,C. J., Wrenn, S. M., Haber, E.
1977. J. Biol. Chert 252:8957--64
50. Homey,C. J., Wrenn, S. M., Haber, E.
1978. Proc. Natl. dcad. Sc£ USA
75:59-63
51. Stellwagen, E., Baker, B. 1976. Nature
261:719-20
52. Swisloeki, N. I., Magnuson,T., Tierney, J. 1977. Arch. Biochem. Biophyz
179:157-65
53. Pfeuffer,
T. 1979. FEBS Lett.
101:85-89
54. Londos, C., Lad, P. M., Nielsen, T. B.,
Rodbell, M. 1979. J. Supramol. Struct.
10:31-37
55. Garbers, D. L., Johnson, R. A. 1975. J.
Biol. Chem. 250:8449-56
56. Howlett, A. C., Gilman, A. G. 1980.
J. Biol. Chem. In press
57. Mittal, C. K., Murad,F. 1977. J. Biol.
Chert 252:3136-40
58. Braun,
R. F. 1975. Proc. Natl.
Acad. T.,
Sc£ Dods,
USA 72:1097-1
59. Spiegel, A. M., Downs,R. W. Jr., Aurbach, G. D. 1979. Z Cyclic Nucleotide
ges, 5:3-17
60. Moss, J., Vaughan, M. 1979. Ann. Bey.
Biochert 48:581-600
61. Acad.
Gill, D.Sc£
M.,USA
Meren,
R. 1978. Proc. Natl.
75:3050-54
62. Enomoto, K., Gill, D. M. 1979. J. Supramol. StrucL 10:51-60
63. Enomoto,K., Gill, D. M. 1980. J. Biol.
Chem. 255:1252-58
64. Johnson, G. L., Kaslow, H. R., Bourne,
H. R. 1978. J. Biol. Chert 253:7120-23
Annual Reviews
www.annualreviews.org/aronline
Annu. Rev. Biochem. 1980.49:533-564. Downloaded from arjournals.annualreviews.org
by LIBRARY CONTINUATIONS on 01/30/06. For personal use only.
562
ROSS
& GILMAN
65. Johnson, G. L., Kaslow, H. R., Bourne,
H. R. 1978. Proc. NatL Acad. ScL USA
75:3113-17
66. Kaslow, H. R., Farfel, Z., Johnson, G.
L., Bourne, H. R. 1979. Mol. PharmacoL 15:472-83
67. Howlett, A. C., Sternweis, P. C., Macik,
B. A., VanArsdalc, P. M., Gilman, A.
G. 1979. d. Biol. Chem. 254:2287-95
67a. Schliefer, L. S., Garrison, J. C., Sternweis, P. C., Northrup, J. 1980. d. Biol.
Chem. In press
68. Haga, T., Ross, E. M., Anderson,H. J.,
Gilman, A. G. 1977. Prec. NatL Acad.
Sc~ USA 74:2016-20
69. Harden, T. K., Su, Y.-F., Perkins, J. P.
1979. d. Cyclic Nucleotide Re& 5:99106.
70. Su, Y.-F., Harden, T. K., Perkins, J. P.
1979. J. Biol. Chem. 254:38-41
71. Malbon, C. C., Moreno, F. J., Cabelli,
R. J., Fain, J. N. 1978. d. Biol. Chem.
253:671-78
72. Rubalcava, B., RodbeH, M. 1973. d.
Biol. Chem. 248:3831-37
73. Moore, W. V., Wolff, J. 1974. J. Biol.
Chem. 249:6255-63
74. Limbird, L. E., Lefkowitz, R. J. 1976.
Mol. Pharmacol. 12:559-67
75. Howlett, A. C., VanArsdale,P. M., Gilman, A. G. 1978. Mol. Pharmacol.
14:531-39
76. Lad, P. M., Preston, M. S., Welton, A.
F., Nielsen, T. B., Rodbell, M. 1979.
Biochim. Biophy~ Acta 551:368-81
77. J.
Sternweis,
P. C.,
Gilman, A. G. 1979.
Biol. Chem.
254:3333-40
78. Naya-Vigne, J., Johnson, G. L.,
Bourne, H. R., Coffmo,P. 1978. Nature
272:720-22
79. Wolff, D. J., Brostrom, C. O. 1979. Adv.
Cyclic Nucleotide Rez 11:27-88
80. Cheung, W. Y., Lynch, T. J., Wallace,
R. W. 1978. Adv. Cyclic Nucleotide Rez
9:233-51
81. Brostrom,C. O., Huang, Y.-C., Breekenridge, B. M., Wolff, D. J. 1975. Proc.
Natl. Acad. ScL USA 72:64-68
82. Cheung, W. Y., Bradham, L. S., Lynch,
T. J., Lin, Y. M., Tallant, E. A. 1975.
Biochem.
Biophyz Rez Commun.
66:1055-62
83. Brostrom, C. O., Brostrom, M. A.,
Wolff, D. J. 1977. J. Biol. Chera.
252:5677-85
84. Lynch, T. J., Tallant, E. A., Cheung,W.
Y. 1977. Arcl~ Biochem. Biophy~
182:124-33
85. Brostrom, M. A., Brostrom, C. O.,
Breckenddge, B. M., Wolff, D. J. 1976.
251:4744-50
86. Brostrom, M. A., Brostrom, C. O.,
Wolff, P. J. 1979. Z Biol. Chem~
254:7548-57
87. Burgess, W. H., Howlett, A. C., Kretsinger, R. H., Gilman, A. G. 197g. J.
Cyclic Nucleotide Re~ 4:175-81
88. Westcott, K. R., LaPorte, D. C., Storm,
D. R. 1979. Proc. Natl. Acad. ScL USA
76:204-8
89. Toscano, W. A. Jr., Westcott, K. R.,
LaPorte,
C., Storm,
D.76:5582-86
R. 1979.
Proc.
Natl.D. Acad.
Sc~ USA
90. Moss, J., Vaughan,M. 1977. Proc. Natl.
Acad. Scl.USA 74:4396-4400
9 I. Cohen,
P.,Burchell,
A.,Foulke~,
J. G.,
Cohen, P. T. W,, Vanaman, T. C.,
Nairn, A. C. 1978. FEBS Lett.
92:287-93
92. Pecker, F., Hanoune,J. 1977. J. Biol.
Chem. 252:2784-86
93. Pecker, F., Hanoune, J. 1977, FEBS
Let~ 83:93-98
94. Katz, M. S., Kelly, T. M., Pifieyro, M.
A., Gregerman, R. I. 1978. J. Cyclic
Nucleotide Rez 4:389-407
95. Sanders, R. B., Thompson,W. J., Robison, G. A. 1977. Biochim. Biophy&Acta
498:10-20
96. Bradham,L. S. 1977. J. Cyclic Nucleotide Rez 3:119-28
97. Sahyoun, N., Schmitges, C. J., LeVine,
H. III. Cuatrecasas, P. 1977. Life
2h1857-63
98. Hebdon, M., LeVine, H. III, Sahyoun,
N., Schmitges, C. J., Cuatrecasas, P.
1978. Proc. Natl. Acad. ScL USA
75:3693-97
99. Pike, L. J., Limbird, L. E., Lcfkowitz,
R. J. 1979. Nature 280:502-4
100. Hoffmann, F. M, 1979. Z Biol. Chem.
254:255-58
101. Hoffmann, F. M. 1979. Biocher,~ Biophyz Res, Commun. 86:988-94
102. Gonzalez, J. E., Petcrkofsky, A. 1977.
Z Supmmol. Struck 6:495-502
103. Peterkofsky, A., Gazdan, C. 1979. Proc.
Natl. .4cad. ScL US.4 76:1099-1103
104. Pall, M. L. 1977. £ BIOL Chem.
252:7146-50
105. Strewlcr, G. J., Orloff, J. 1977. Adv. Cyclic Nucleotide Rez 8:311-61
106. De Kruyjff, B., Gerritsen, W. J., Oerle~
roans, A., Demel, R. A., van Deenen, L.
L. M. 1974. Blochira. Biophys, Acta
339:30-43
107. Grollman, E. F., Lee, G., Ambesi-Impiombato,F. S., Meldolesi,M. F., Aloj, S.
M., Coon, H. G., Kaback, H. R., Kohn,
L. D. 1977. Proc. NatL .4cad. Sc£ USA
74:2352-56
108. SuHmovici, S.,Bartoov, B.,Lunenfeld, B.
1975. Biochim. Biophyz Acta 377:454-62
Annual Reviews
www.annualreviews.org/aronline
Annu. Rev. Biochem. 1980.49:533-564. Downloaded from arjournals.annualreviews.org
by LIBRARY CONTINUATIONS on 01/30/06. For personal use only.
HORMONE-SENSITIVE
109. Kempcn,H. J. M., DePont, J. J. H. H.
M., Bonting, S. L. 1974. Biochim. Biophys, Acta 370:573-84
1,10. Rethy, A., Tomasi, V., Trevisani, A.
1971. Arch. Biochem. Biophyz
147:36-40
111. Pohl, S. L., Krans, H. M. J., Kozyreff,
V.,
Birnbaumer,
L., Rodb¢ll, M. 1971.
J. Biol.
Chert~ 246:4447-54
112. Levey, G. S. 1971. J. Biol. Chem.
246:7405--10
1!3. Levey, G. S. 1971. Biochem. Biophy~
Rex Commu~ 43:108-13
114. Engelhard, V. H., Esko, J. D., Storm,
D.
R., Sc£
Glaser,
M. 1976. Proc. Natl.
Acad.
USA 73:4482-86
1!5. Engelhard, V. H., Glaser, M., Storm,
D. R. 1978. Biochemistry 17:3191-3200
1 i6. Hiram,F., Strittmatter, W. J., Axelrod,
J. 1979. Proc. Natl. Acad. ScL USA
76:368-72
117. Klein, I., Moore, L., Pastan, I. 1978.
Biochira. Biophy~ Acta 506:42-53
118. Hanski, E., Levitzki, A. 1978. LifeSc£
22:53-60
119. Sahyoun, N., Hollenberg, M. D., Bennett, V., Cuatreeasas, P. 1977. Proc.
Natl. Acad. Sc£ USA 74:2860-64
120. Varga, J. M., Saper, M. A., Lerner, A.
B., Fritsch, P. 1976. ,L. Supramo£
Struct. 4:45-49
121. Maguire, M. E., Ross, E. M., Gilraan,
A. G. 1977. Adv. Cyclic Nucleotide Re&
8:1-83
122. Kreiner, P. W., Keirns, J. J., Bitensky,
M. W. 1973. Proc. Natl. Acad. Sc£ USA
70:1785-1789
123. B~ir, H.-P. 1974. Mol. Pharmacol.
20:597-604
124. Rene, E., Peeker, F., Stengel, D., Hanoune, J.
1978. d. Biol. Chem.
253:838-4l
125. Salmon, C., Dderue-LaBelle, N., Fontaine,
Y.-A. 1974. Biochimie
56:1229-38
126. Houslay, M. D., Metealfe, J. C., Warren, G. B., Hesketh, T. R., Smith, G. A.
1976. Biochim.
Biophys,
Acta
436:489-94
127. Houslay, M. D., Hesketh, T. R., Smith,
G. A., Warren, G. B., Metealfe, J. C.
1976. Biochim. Biophy~ Acta 436:495504
128. Bakardjieva, A., Galla, M. J., Helmreich, E. J. M. 1979. Biochemistry
18:3016-23
129. Orly, J., Schramm,M. 1975. Proc. Natl.
,4cad. Sc£ USA 72:3433-37
130. Rimon,G., Hanski, E., Braun, S., Levitzld, A. 1978. Nature 276:394-96
131. Hanski, E., Rimon, G., Levitzld, A.
1979. Biochemistry 18:846-53
ADENYLATE
CYCLASE
563
132. Shinitzky, M., Barenholz, Y. 1974. d.
Biol. Chem. 249:2652-57
133. Chen, L. A., Dale, R. E., Roth, S.,
Brand, L. 1977. .L. Biol. Chem.
252:2163-69
134. Cryer, P. E., Jarett, L., Kipnis, D. M.
1969. Biochim.
Biophys.
Acta
177:586-89
135. Rodbell, M., Lin, M. C., Salomon, Y.,
Londos, C., Harwood,J. P., Martin, B.
R., Rendell, M., Berman, M. 1975. Adv.
Cyclic Nucleotide Re~ 5:3-29
136. Rodbdl, M., Birnbanmer, L., Pohl, S.
L., Krans, H. M. J. 1971. J. Biol. Chem.
246:1877-82
137. Rodbell, M. 1975. J. Biol. Chem.
250:5826-34
138. Londos, C., Salomon, Y., Lin, M. C.,
Harwood, J. P., Schramm, M., Wolff,
J., Rodbell, M. 1974. Proc. Natl./lcad.
ScL USA 71:3087-90
139. Kimura, N., Nagata, N. 1977. J, Biol.
Chert~ 252:3829-35
140. Ross, E. M., Maguire,M. E., Sturgill,
T. W., Biltonen, R. L., Gilman, A. G.
1977. J. Biol. Chem. 252:5761-75
141. J..
Franklin,
T. J.,77:113-17
Twose,P. A. 1977. F~ur.
Biochem.
142. Smith, C. M., Henderson, J. F., Baer,
H. P. 1977. J. Cyclic Nucleotide Res.
3:347-54
143. Johnson, G. S., Mukku,V. K. 1979. J.
Biol. Chem. 254:95-100
144. Rodbell, M., Krans, H. M. J., Pohl, S.
L., Birnbaumer,L. 1971. J. Biol. Chem.
246:1872-76
145. Lin, M. C., Nicosia, S., Lad, P. M.,
Rodbell, M. 1977. Z Biol. Chere~
252:2790-91
146. Moore, W. V., Wolff, J. 1973. Z Biol.
Chem. 248:5705-11
147. Lefkowitz, R. J., Mullikln, D., Wood,
C. L.,
Gore,
T. B.,
Mukherjee, C. 1977.
J..
Biol.
Chert
252:5295-5303
148. Maguire, M. E., VanArsdale, P. M.,
Gilman, A. G. 1976. Mol. Pharmacol.
12:335-39
149. Lefkowitz, R. J., Mullikin, D., Caron,
M. G. 1976. ,L. Biol.
Chem.
251:4686-92
150. d.
Williams,
L. T.,252:7207-13
Lefkowitz, R. J. 1977.
BioL Chem.
151. Abou-Issa,
H., Reiehert,
L. E. Jr. 1976.
J. BioL Chem.
251:3326-37
152. Schramm, M., Rodbell, M. 1975. J.
BioL Chem. 250:2232-37
153. Spiegel, A. M., Brown,E. M., Fedak, S.
A., Woodard, C. J., Aurbach, G. D.
1976. J. CyclicNucleotideRes. 2:47-56
Spiegel,
M.,249:7630-36
Aurbach, G. D. 1974.
154. J.
Biol. A.
Chem.
Annual Reviews
www.annualreviews.org/aronline
Annu. Rev. Biochem. 1980.49:533-564. Downloaded from arjournals.annualreviews.org
by LIBRARY CONTINUATIONS on 01/30/06. For personal use only.
564
ROSS & GILMAN
155. Jaeobs, S., Bennett, V., Cuatreeasas,
1976. J. Cyclic Nucleotide Rex
2:205-23
156. Bennett, V., Cuatreeasas, P. 1976. J.
Membr.Bio£ 27:207-32
157. Sevilla, N., Steer, M.L., Levitzki, A,
1976. Biochemistry15:3493-99
158. Tolkovsky,A. M., Levitzki, A. 1978.
Biochemistry 17:3795-3810
159. Tolkovsky,A. M., Levitzki, A. 1978.
Biochemistry 17:3811-17
160. Blume,A.J., Foster,C. J. 1976.J. Biol.
Cher~251:3399-3404
161. Cuatrecasas,P., Jaeobs,S., Bennett,V.
1975. Proc. Nat£ Acad. Sc~ USA
72:1739-43
162.Cassel, D., S¢linger, Z. 1976.Biochin~
Biophys,/lcta 452:538-51
163.Cassel,D., Selinger,Z. 1977.Biochera.
Biophys, Rex Commun.77:868-73
164. Cassel,
D., Selinger,
Z. 1977.Proc.NatL
Acad~Sc1
USA74:3307-11
165. Cassel,D., Levkovitz,H., Selinger, Z.
1977. J. Cyclic NucleotideRex 3:393406
166. Levinson,S. L., Blume,A. J. 1977. J.
Biol. Chem.252:3766--74
167. Flores, J., Sharp, G. W.G. 1975. J.
Clit~ lnvest. 56:1345-49
168. Johnson, G. L., Harden, T. K., Perkins, J. P. 1978. J. Biol. Chem.253:
1465-71
169. Lueas-Lenard,J., Beres, L. 1974. The
Enzymes10:53-83
170. Cassel, D., Eckstein, F., Lowe,M., Selinger, Z. 1979. d. Biol. Chem.254:
9835-38
171. Cassel,D., Selinger,Z. 1978.Proc.Natl.
Acad. Sc£ USA75:4155-59
172. Cassel, D., Selinger, Z. 1977.J. Cyclic
Nucleotide Reg 3:11-22
173. Lad, P. M., Welton,A. F., Rodbell,M.
1977. J. Biol. Chem.252:5942-46
174. Limbird,L. E., Hiekey, A. R., Lefkowitz, R. J. 1979. J. Cyclic Nucleotide
Rex 5:251-59
175. Chau, V., Huang,L. C., Romero,G.,
Biltonen, R. L., Huang,C.-H. 1979.
Biochemistry.In press