Download Molecular and cellular requirements for the

912
Biochemical Society Transactions (2003) Volume 31, part 5
Molecular and cellular requirements for the
regulation of adenylate cyclases by calcium
D.M.F. Cooper1
Department of Pharmacology, University of Cambridge, Tennis Court Road, Cambridge CB2 1PD, U.K.
Abstract
Calcium-sensitive adenylate cyclases provide a key regulatory device for integrating the activities of the two
major signalling systems, Ca2+ and cAMP. Recent experiments have brought us closer to understanding the
molecular mechanisms whereby Ca2+ either stimulates or inhibits susceptible adenylate cyclases in vitro.
However, in the intact cell an additional layer of sophistication is evident whereby Ca2+ -sensitive adenylate
cyclases are juxtaposed with Ca2+ -entry channels, such that the cyclases respond selectively to capacitative
Ca2+ entry. Part of this dependency is enforced by the placement of Ca2+ -sensitive adenylate cyclases (AC5,
AC6 and AC8) in caveolae, from which at least one Ca2+ -insensitive adenylate cyclase (AC7) is excluded.
However, additional protein–protein interactions are also required to ensure the dependency of these
cyclases on capacitative Ca2+ entry. Recent findings in this area and their implications for ‘local cAMP
signals’ will be discussed.
Introduction
Mammalian adenylate cyclases vary significantly outside of
their largely conserved catalytic domains. This diversity
underlies a broad range of individual regulation [1]. Detailed
structural information is available on the catalytic region of
adenylate cyclase, but the properties conferring individual
regulatory properties are only emerging slowly. The typespecific regulation of adenylate cyclases is augmented by
higher-order assemblies as well as discrete subcellular
targeting. It is also clear that phosphodiesterases – the other
side of the equation that establishes cAMP levels – are
also diverse, compartmentalized and involved in protein–
protein associations [2,3]. A consequence of the organization
of cAMP signalling complexes is the likelihood that the
concentration of cAMP differs in the microdomains where
it is synthesized from the broad cytosol into which it can
diffuse. Newly developed methods for the measurement of
cAMP in microdomains may allow a dissection of the factors
contributing to the organization of these domains.
Diversity and distribution of adenylate
cyclases
Since the first mammalian adenylate cyclase was cloned in
1989 [4], eight other species have been identified. Although
this multiplicity of species was an initial surprise, different
regulatory patterns coupled with selective tissue expression
suggested underlying cellular rationales, which are still being
unravelled [5]. Although most tissues, and indeed many cell
Key words: adenylate cyclase, calmodulin, calcium, cAMP-dependent protein kinase, caveolae,
phosphodiesterase.
Abbreviations used: AC1, AC2, AC3, etc., adenylate cyclase types 1, 2, 3, etc.; CCE, capacitative
Ca2+ entry; [Ca2+ ]i , intracellular [Ca2+ ]; eNOS, endothelial nitric oxide synthase.
1
e-mail [email protected]
C 2003
Biochemical Society
types express more than one adenylate cyclase species, some
selective enrichments are striking. For instance, the Ca2+ stimulated species principally occur in neuronal tissue, as
well as pancreatic tissue [5,6]. In addition, the Ca2+ -inhibited
species, AC5 (adenylate cyclase type 5), is at its highest levels
in striatum and cardiac tissue [5,7]. In cardiac tissue it has been
proposed that the susceptibility to inhibition by Ca2+ of the
cyclases (AC5 and AC6) mediating the actions of sympathetic
neurotransmission provides a key negative feedback device
that contributes to cardiac rhythmicity [7]. Furthermore,
knockout experiments have established an important role
for the adenylate cyclases that are stimulated by Ca2+ acting
by calmodulin (AC1 and AC8) in hippocampal learning and
memory [8,9].
Regulation by calcium
Stimulation
Shortly after the discovery of calmodulin as a calciumdependent regulator of phosphodiesterase, adenylate cyclase
activity in membranes prepared from brain tissue was also
seen to be stimulated by calmodulin (more correctly, of
course, this was a stimulation by Ca2+ in the medium, that
was dependent on calmodulin). The concentrations of Ca2+
eliciting this stimulation in vitro correspond to those achieved
in the cell cytosol upon activation of the phospholipase
C pathway or activation of voltage-gated Ca2+ -channels.
The first adenylate cyclase cloned was the Ca2+ /calmodulinstimulated species, AC1, shortly followed by the similarly
regulated AC3 and AC8 [1].
The Ca2+ /calmodulin-stimulated adenylate cyclases differ
somewhat in their sensitivities to Ca2+ in vitro. In fact, AC3,
while being most closely related to AC1 and AC8 at the
amino acid level, is far from being proven to be stimulated
Calcium Oscillations and the 5th UK Calcium Signalling Conference
Figure 1 Schematic representation of the three types of response
to Ca2+ of the cloned adenylate cyclases
Roman numerals correspond to adenylate cyclases (e.g. V is AC5).
sufficient data have now accumulated from a variety of
systems to show that the anticipated stimulation and inhibition of these cyclases does occur as a direct result of the
elevation of [Ca2+ ]i [10,14–18].
Regulation by CCE (capacitative Ca2+ entry)
unambiguously by Ca2+ /calmodulin. In the absence of other
effectors AC3 is not stimulated by Ca2+ /calmodulin, but
it is stimulated by Ca2+ /calmodulin when the enzyme is
concomitantly activated by either 5 -Gpp(NH)p or forskolin.
The free [Ca2+ ] for half-maximal stimulation of AC1 and
AC3 are 0.05 and 5.0 µM Ca2+ , respectively [10]. AC1
and AC8 can both be stimulated by Ca2+ /calmodulin in the
absence or presence of forskolin with EC50 values of approx.
0.1 and 0.5 µM, respectively [11].
One of the more remarkable properties of Ca2+ -sensitivite
adenylate cyclases is their virtual dependence on CCE for
their regulation. A side effect of establishing the Ca2+
sensitivity of these enzymes in vivo was assessing their
responses to CCE, i.e. the entry of Ca2+ that is triggered by
depletion of Ca2+ stores. This simple but readily interpretable
measure of Ca2+ sensitivity was chosen since it circumvents
many of the steps required to exclude other mechanisms
when hormone is used to elevate [Ca2+ ]i . When expressed in
HEK-293 cells, putatively Ca2+ -sensitive adenylate cyclases
give rise to the anticipated changes in cAMP levels in response
to CCE [16] (AC3, although expressed at adequate levels, is
not stimulated by CCE [11]). Remarkably, however, when
regulation of these cyclases, either endogenously or exogenously expressed, by CCE is compared with other means
of elevating [Ca2+ ]i , i.e. (i) release from intracellular stores,
(ii) ionophore-mediated entry through the plasma membrane
and (iii) arachidonic acid-mediated entry, only CCE is
effective [11,19,20]. Such selectivity has since been confirmed
in a number of other systems [6,14,18,21,22]. The available
evidence suggests a very close apposition between Ca2+ sensitive adenylate cyclases and CCE channels [19,23,24].
Inhibition
Calcium also inhibits the activity of adenylate cyclase in many
tissues. This inhibition takes two forms, either low affinity
(approx. K i value 10–25 µM), which is seen in all tissues, or
both low and high affinity (approx. K i value 0.2–0.6 µM),
which is seen in selected tissues such as cardiac, anterior
pituitary, striatum and a variety of cell lines [12]. Inhibition of
adenylate cyclase by Ca2+ does not require calmodulin (see
below). Cloning of the adenylate cyclases in the early 1990s
yielded what are now distinguished as Ca2+ /calmodulinstimulated (AC1, AC3 and AC8), high- (and low-) affinity
Ca2+ -inhibited (AC5 and AC6) and Ca2+ -insensitive, but
protein kinase C-stimulated (AC2, AC4 and AC7) adenylate
cyclases (Figure 1) [1] (AC9, the ninth known adenylate
cyclase, is a widely distributed, low-activity enzyme that is
inhibited by calcineurin).
Both the high-affinity inhibition by Ca2+ and the
calmodulin-dependent stimulation of adenylate cyclases
occur at concentrations that are achieved in cells upon the
physiological elevation of [Ca2+ ]i (intracellular [Ca2+ ]), as
measured by fura-2. Considerable effort has been put into
proving that hormones that elevate [Ca2+ ]i can regulate such
cyclases by this mechanism – and none other. For instance,
hormones that activate phospholipase C can thereby activate
protein kinase C, which stimulates the activity of certain
cyclases. Furthermore, the elevation of [Ca2+ ]i by carbachol
stimulates phosphodiesterase (type I) and thereby inhibits
cAMP accumulation in 1321 glioma cells [13]. However,
Structural properties
Structural basis for Ca2+ inhibition
A remarkable feature of adenylate cyclases is the fact that
the catalytic domains, which are cytosolic, can be expressed
independently of the rest of the molecule. This has allowed
the expression of considerable amounts of protein, its purification and determination of its crystal structure. Elucidation
of this structure has revealed the atomic basis of adenylate
cyclase catalytic activity [25–27]. Solving the structure also
made it possible to address the structural basis for Ca2+
inhibition. Early experiments had eliminated the obvious
participation of readily dissociable calmodulin in the
inhibition of adenylate cyclases by Ca2+ . More recently,
chimaeras were made between AC5 and AC2 in an attempt at
narrowing down the domains of AC5 that were responsible
for its inhibitory sensitivity. Retention of the first catalytic
domain of AC5 was sufficient to maintain a Ca2+ -inhibitable
activity [28]. This finding focused attention on the mechanism
inferred from the crystal structure. Two aspartate residues
play critical roles in co-ordinating two Mg2+ ions, which
catalyse the nucleophilic attack of the 3 hydroxyl of ribose
on the α-phosphate of ATP. Mutation of the two aspartic
acids in the C1a domain eliminates activity, as expected.
However, mutation of more distal amino acids, which would
be expected to impact on the precise orientation of those
C 2003
Biochemical Society
913
914
Biochemical Society Transactions (2003) Volume 31, part 5
aspartyl hydroxyl groups, had a more benign effect of
reducing the affinity of the adenylate cyclase for Mg2+ . These
mutations turned out to have decreased sensitivity to Ca2+
inhibition. The simplest interpretation of these findings is that
both high- and low-affinity Ca2+ inhibition are mediated in
the Mg2+ -binding domain of the catalytic site [28].
Structural basis for Ca2+ stimulation
Calmodulin mediates the stimulation of both AC1 and AC8.
Removal of calmodulin by EGTA washes eliminates the
stimulation by Ca2+ . Stimulation of AC1 requires amino
acids near the plasma membrane in the C1b region [29,30].
AC8 binds calmodulin at both the N- and the C-terminus.
The latter domain is critical for Ca2+ /calmodulin stimulation
[31]. Deletion of both calmodulin-binding domains from
AC8 results in a super-active enzyme, which suggests that
calmodulin activation of AC8, like other calmodulin targets,
is via a disinhibitory mechanism [31].
Compartmentalization of adenylate
cyclases
Localization of adenylate cyclases in caveolae
or rafts
One factor contributing to the dependence of adenylate
cyclases on CCE is their localization in discrete domains
of the plasma membrane [32,33]. Both AC8 transfected in
HEK-293 cells and endogenous AC6 in C6–2B glioma cells
occur in buoyant membrane fractions. Their presence in
these cholesterol-rich domains is essential for their regulation by CCE, since disruption of these domains by the
cholesterol-extracting agent, methyl-β-cyclodextrin, eliminates the regulation. The Ca2+ -insensitive AC7 is excluded
from rafts. However, even though AC8 must be present in
these domains for regulation by CCE, mutants of AC8 can be
generated that remain in rafts, but are insensitive to CCE [34].
Thus, other presumably protein–protein interactions must
also be required. AC5 also occurs in rafts; it has been proposed
in direct association with caveolin [35–37]. The presence of
AC5/AC6 in caveolae in cardiac myocytes is also believed to
play a critical role in β-adrenergic signalling [38,39].
Although it is not clear which protein–protein interactions
ensure the regulation of AC8 by CCE, some analogies with
the eNOS (endothelial nitric oxide synthase) system are
tantalizing. eNOS is apparently recruited to caveolae by
myristoylation of the C-terminus and retained in an inactive
state through an interaction with caveolin. Upon the triggering of CCE, calmodulin displaces caveolin from eNOS,
resulting in activation of the enzyme [40]. In the case of
AC8, there is an apparently non-essential calmodulin-binding
site at the N-terminus (see above). Deletion of this domain
renders AC8 insensitive to CCE, although it is retained within
caveolae. We wonder whether there may be some parallels
between the eNOS and the cyclase systems in this regard.
C 2003
Biochemical Society
Local cAMP
Isolated reports over the years have suggested that cAMP in
the microdomain of its site of synthesis may not faithfully
reflect cAMP concentrations within the cytosol of cells [41–
43]. As we learn more about the compartmentalization of
adenylate cyclases to sub-domains of the plasma membrane
more physical support for this concept develops. In addition,
the fact that phosphodiesterases are selectively targeted to
different domains of the cell along with cAMP-dependent
protein kinase-anchoring proteins makes an extremely
strong, physical case for cAMP-signalling microdomains [3].
Recent methodologies that are being developed to explore the
dynamics of cAMP in single cells are encountering data that
support this possibility [3,43–49].
Future directions
We have learned a lot about the regulation of Ca2+ -sensitive
adenylate cyclases over the last few years – both at the biochemical and cell-biological levels. From a biochemical viewpoint it will be of considerable interest to know what makes
some cyclases subject to high-affinity inhibition by Ca2+ and
others insensitive, given that adenylate cyclases are at their
most conserved in their catalytic domains. However, there
are significant differences in the amino acids surrounding the
catalytic aspartates between different isoforms. Ideally, crystal structure comparisons of Ca2+ -sensitive and Ca2+ insensitive forms will reveal the underlying mechanisms. It
will also be of considerable interest to determine the role of
the N-terminal binding of calmodulin by AC8 – whether, for
instance, it plays a recruiting role for calmodulin or, perhaps,
induces a conformation in the N-terminus that allows interactions with scaffolding proteins. Newly emerging studies
suggest that higher-order assemblies of adenylate cyclase
molecules can occur. This propensity may be enlightening
in terms of identifying partners in signalling complexes. The
mechanism underlying the apposition of cyclases with CCE
channels remains a burning question, whose solution may
yet provide unsuspected insights into CCE organization
[50]. Continued application of single-cell methods to the
measurement of cAMP is expected to provide novel insights
into the dynamics of the second messenger near the plasma
membrane. It was predicted some time ago with rather conservative modelling, and again more recently, that the interaction between Ca2+ and Ca2+ -sensitive adenylate cyclases
could yield cAMP oscillations [51,52]. If the current technologies support these predictions, the next challenge will be
to develop strategies to search for potential targets of cAMP
spikes.
The work in the author’s laboratory is supported by the National
Institutes of Health and the Wellcome Trust.
References
1 Taussig, R. and Gilman, A.G. (1995) J. Biol. Chem. 270, 1–4
Calcium Oscillations and the 5th UK Calcium Signalling Conference
2 Mehats, C., Andersen, C.B., Filopanti, M., Jin, S.-L.C. and Conti, M. (2002)
Trends Endocrinol. Metab. 13, 29–35
3 Houslay, M.D. and Milligan, G. (1997) Trends Biochem. Sci. 22, 217–224
4 Krupinski, J., Coussen, F., Bakalyar, H.A., Tang, W.J., Feinstein, P.G.,
Orth, K., Slaughter, C., Reed, R.R. and Gilman, A.G. (1989) Science 244,
1558–1564
5 Hanoune, J. and Defer, N. (2001) Annu. Rev. Pharmacol. Toxicol. 41,
145–174
6 Watson, E.L., Jacobson, K.L., Singh, J.C., Idzerda, R., Ott, S.M., DiJulio, D.H.,
Wong, S.T. and Storm, D.R. (2000) J. Biol. Chem. 275, 14691–14699
7 Cooper, D.M.F., Karpen, J.W., Fagan, K.A. and Mons, N.E. (1998)
Adv. Second Messenger Phosphoprotein Res. 32, 23–51
8 Wu, Z.L., Thomas, S.A., Villacres, E.C., Xia, Z., Simmons, M.L., Chavkin, C.,
Palmiter, R.D. and Storm, D.R. (1995) Proc. Natl. Acad. Sci. U.S.A. 92,
220–224
9 Xia, Z. and Storm, D.R. (1997) Curr. Opin. Neurobiol. 7, 391–396
10 Choi, E.J., Xia, Z. and Storm, D.R. (1992) Biochemistry 31, 6492–6498
11 Fagan, K.A., Mahey, R. and Cooper, D.M.F. (1996) J. Biol. Chem. 271,
12438–12444
12 Guillou, J.-L., Nakata, H. and Cooper, D.M.F. (1999) J. Biol. Chem. 274,
35539–35545
13 Harden, T.K., Evans, T., Hepler, J.R., Hughes, A.R., Martin, M.W., Meeker,
R.B., Smith, M.M. and Tanner, L.I. (1985) Adv. Cyclic Nucleotide Protein
Phosphorylation Res. 19, 207–220
14 Watson, E.L., Wu, Z., Jacobson, K.L., Storm, D.R., Singh, J.C. and Ott, S.M.
(1998) Am. J. Physiol. 274, C557–C565
15 Boyajian, C.L., Garritsen, A. and Cooper, D.M.F. (1991) J. Biol. Chem. 266,
4995–5003
16 Cooper, D.M.F., Yoshimura, M., Zhang, Y., Chiono, M. and Mahey, R.
(1994) Biochem. J. 297, 437–440
17 Chabardes, D., Imbert-Teboul, M. and Elalouf, J.M. (1999) Cell Signal. 11,
651–663
18 Burnay, M.M., Vallotton, M.B., Capponi, A.M. and Rossier, M.F. (1998)
Biochem. J. 330, 21–27
19 Fagan, K.A., Mons, N. and Cooper, D.M.F. (1998) J. Biol. Chem. 273,
9297–9305
20 Shuttleworth, T.J. and Thompson, J.L. (1999) J. Biol. Chem. 274,
31174–31178
21 Murthy, K.S. and Makhlouf, G.M. (1998) Mol. Pharmacol. 54, 122–128
22 Cioffi, D.L., Moore, T.M., Schaack, J., Creighton, J.R., Cooper, D.M.F. and
Stevens, T. (2002) J. Cell Biol. 157, 1267–1278
23 Nakahashi, Y., Nelson, E., Fagan, K.A., Gonzales, E., Guillou, J.L. and
Cooper, D.M.F. (1997) J. Biol. Chem. 272, 18093–18097
24 Gu, C. and Cooper, D.M.F. (2000) J. Biol. Chem. 275, 6980–6986
25 Tesmer, J.J., Sunahara, R.K., Gilman, A.G. and Sprang, S.R. (1997) Science
278, 1907–1916
26 Tesmer, J.J.G., Sunahara, R.K., Johnson, R.A., Gosselin, G., Gilman, A.G.
and Sprang, S.R. (1999) Science. 285, 756–760
27 Zhang, G., Liu, Y., Ruoho, A.E. and Hurley, J.H. (1997) Nature (London)
386, 247–253
28 Hu, B., Nakata, H., Gu, C., De Beer, T. and Cooper, D.M.F. (2002) J. Biol.
Chem. 277, 33139–33147
29 Vorherr, T., Knopfel, L., Hofmann, F., Mollner, S., Pfeuffer, T. and
Carafoli, E. (1993) Biochemistry 32, 6081–6088
30 Wu, Z., Wong, S.T. and Storm, D.R. (1993) J. Biol. Chem. 268,
23766–23768
31 Gu, C. and Cooper, D.M.F. (1999) J. Biol. Chem. 274, 8012–8021
32 Huang, C., Hepler, J.R., Chen, L.T., Gilman, A.G., Anderson, R.G.W. and
Mumby, S.M. (1997) Mol. Biol. Cell 8, 2365–2378
33 Gao, T., Puri, T.S., Gerhardstein, B.L., Chien, A.J., Green, R.D. and Hosey,
M.M. (1997) J. Biol. Chem. 272, 19401–19407
34 Smith, K.E., Gu, C., Fagan, K.A., Hu, B. and Cooper, D.M.F. (2002) J. Biol.
Chem. 277, 6025–6031
35 Schwencke, C., Oka, N., Toya, Y. and Ishikawa, Y. (1997) FASEB J. 2,
A323
36 Toya, Y., Schwencke, C., Couet, J., Lisanti, M.P. and Ishikawa, Y. (1998)
Endocrinology 139, 2025–2031
37 Schwencke, C., Yamamoto, M., Okumura, S., Toya, Y., Kim, S.J. and
Ishikawa, Y. (1999) Mol. Endocrinol. 13, 1061–1070
38 Rybin, V.O., Xu, X., Lisanti, M.P. and Steinberg, S.F. (2000) J. Biol. Chem.
275, 41447–41457
39 Ostrom, R.S., Violin, J.D., Coleman, S. and Insel, P.A. (2000)
Mol. Pharmacol. 57, 1075–1079
40 Lin, S., Fagan, K.A., Li, K.X., Shaul, P.W., Cooper, D.M.F. and Rodman, D.M.
(2000) J. Biol. Chem. 275, 17979–17985
41 Buxton, I.L. and Brunton, L.L. (1983) J. Biol. Chem. 258, 10233–10239
42 Sudlow, L.C. and Gillette, R. (1997) J. Gen. Physiol. 110, 243–255
43 Jurevicius, J. and Fischmeister, R. (1996) Proc. Natl. Acad. Sci. U.S.A. 93,
295–299
44 Rich, T.C., Fagan, K.A., Nakata, H., Schaack, J., Cooper, D.M.F. and Karpen,
J.W. (2000) J. Gen. Physiol. 116, 147–161
45 Rich, T.C., Tse, T.E., Rohan, J.G., Schaack, J. and Karpen, J.W. (2001) J. Gen.
Physiol. 118, 63–78
46 Razani, B., Rubin, C.S. and Lisanti, M.P. (1999) J. Biol. Chem. 274,
26353–26360
47 Tasken, K.A., Collas, P., Kemmner, W.A., Witczak, O., Conti, M. and
Tasken, K. (2001) J. Biol. Chem. 276, 21999–22002
48 Zaccolo, M. and Pozzan, T. (2003) Trends Neurosci. 26, 53–55
49 Goaillard, J.M., Vincent, P.V. and Fischmeister, R. (2001) J. Physiol. 530,
79–91
50 Gu, C., Sorkin, A. and Cooper, D.M.F. (2001) Curr. Biol. 11, 185–190
51 Cooper, D.M.F., Mons, N. and Karpen, J.W. (1995) Nature (London) 374,
421–424
52 Gorbunova, Y.V. and Spitzer, N.C. (2002) Nature (London) 418, 93–96
Received 4 April 2003
C 2003
Biochemical Society
915