Human adenylate cyclase and use therefor

The present invention relates to the control of cellular metabolic process by human adenylate cyclase.

Cells of multi-cellular organisms may be metabolically affected by external factors, which are usually chemical. Hormones are a well-known example of such chemical factors. Generally the external chemical factors interact with a specific receptor located on the membrane of the targeted cell. The binding event of the factor to its receptor may induce alterations in cellular metabolism via a "secondary messenger" mediator.

One of the key "secondary messengers" is cyclic AMP (cAMP) (see Sutherland, Science 177:401-407 (1972)). Cyclic AMP is produced from ATP through the action of an enzyme, adenylate cyclase. It is now known that adenylate cyclase activity may be affected by a factor/receptor binding event transmitted through an associated G protein.

Alteration of the intracellular concentration of cAMP affects many cellular reactions. For example, an increase in cAMP intracellular concentration stimulates the activity of protein kinases (enzymes that transfer terminal phosphate groups from ATP to specific sites on targeted proteins). The action of the protein kinases changes the activity or function of its substrate.

For a general review of cAMP and secondary messenger systems reference is made to "Molecular Cell Biology", Darnell et al, 1986, Chapter 16, incorporated herein by reference.

Further investigations of cAMP as a secondary messenger revealed that an alteration in cAMP intracellular concentration was caused by the interaction of several different external factors with their distinct receptors. Further, it was found that different receptors were associated with their own particular G-protein intermediary which was itself associated with adenylate cyclase. More recent investigations have shown that there are in fact different types (isoenzymes) of adenylate cyclase, which display considerable regulatory diversity.

To date eight distinct isoenzymes of adenylate cyclase have been identified and described in the literature. The complete cDNA sequences are known for isoenzymes types 1 to 8. A review of the current understanding and knowledge of the known adenylate cyclase isoenzymes is set out in Pieroni et al, Current Opinion in Neurobiology 3:345-351 (1993); Kerwin Jr in Annual Reports in Medicinal Chemistry, Section VI, Chapter 29, Pages 287-295 (ed Venuti), (1994) and Premont, Methods in Enzymology 238:116-127 (1994).

A summary of the regulation of the known isoenzymes of adenylate cyclase is set out below in Table 1.

TABLE 1 cAMP Isoenzyme Regulated by concentration 1 Ca.sup.2+ /CaM .Arrow-up bold. .beta. .gamma. dimer .dwnarw. 2 G.sub.xi + PKC .Arrow-up bold. .beta. .gamma. dimer .Arrow-up bold. PKC .Arrow-up bold. 3 Ca.sup.2+ /CaM .Arrow-up bold. 4 .beta. .gamma. dimer .Arrow-up bold. 5 Ca.sup.2+ .dwnarw. 6 Ca.sup.2+ .dwnarw. CaM = calmodulin PCK = protein Kinase C

It has now been found, for the first time, that the protein phosphatase calcineurin regulates an adenylate cyclase isoenzyme.

It has further now been found that the isoenzyme regulated by calcineurin is a novel previously uncharacterised adenylate cyclase isoenzyme. The novel isoenzyme of the present invention was originally referred to as adenylate cyclase 10 (AC10), but a review of nomenclature has now caused the novel adenylate cyclase to be referred to as adenylate cyclase 9 (AC9). To avoid confusion with the different isoenzyme known before the nomenclature revision as "adenylate cyclase 9", the novel adenylate cyclase of the present invention will herein simply be referred to as "AC".

The nucleotide sequence encoding for mouse AC has been identified, cloned and sequenced (see Example 2). The nucleotide sequence encoding for mouse AC is given in SEQ ID No 1. The sequence is also accessible in the Genbank.TM. database under accession No. MMU30602 and in the EMBL database under accession No. Z50190.

A nucleotide sequence purporting to be that for human AC is described in WO-A-99/01540. The nucleotide sequence presented in WO-A-99/01540 indicates that the C-terminal portion of the protein is truncated relative to that of the mouse. We have now conducted additional work in this area and, surprisingly, have cloned and sequenced a human AC construct which does not have the truncated C-terminal portion reported in WO-A-99/01540. The sequence obtained by us for human AC is set out in SEQ ID No 98.

The present invention therefore provides a polypeptide encoded by the nucleotide sequence of SEQ ID No 98 or as set out in SEQ ID No 99 (or functional equivalents or parts of those sequences).

The term "functional equivalents" is used herein to refer to any modified version of a nucleotide or polypeptide which retains the basic function of its unmodified form. As an example, it is well-known that certain alterations in amino acid or nucleic acid sequences may not affect the polypeptide encoded by that molecule or the function of the polypeptide. It is also possible for deleted versions of a molecule to perform a particular function as well as the original molecule. Even where an alteration does affect whether and to what degree a particular function is performed, such altered molecules are included within the term "functional equivalent" provided that the function of the molecule is not so deleteriously affected as to render the molecule useless for its intended purpose.

Whilst we do not wish to be bound to theoretical considerations, it is believed that calcineurin regulates AC by removal of phosphate group(s) required for the active form of the enzyme. Thus, the adenylate cyclase activity of AC is believed to decrease in the presence of calcineurin.

In a further aspect, therefore, the present invention provides the use of calcineurin in the regulation of adenylate cyclase activity, in particular in the regulation of AC.

The activity of calcineurin is itself enhanced by the presence of Ca.sup.2+ ions and further enhanced by the additional presence of calmodulin.

In a further aspect, the present invention provides an adenylate cyclase isoenzyme, the activity of which can be regulated by calcineurin.

Desirably the calcineurin regulatable adenylate cyclase isoenzyme is encoded by the nucleotide sequence of SEQ ID No 98, functional equivalents or parts thereof.

In a yet further aspect, the present invention provides a polynucleotide comprising a sequence derived from the sequence set out in SEQ ID No 98 or a part thereof.

The phrase "derived from" includes identical and complementary copies of the sequence of SEQ ID No 1, whether of RNA or DNA and whether in single or double-stranded form. The phrase "derived from" further includes sequences with alterations which (due to the degeneracy of the genetic code) do not affect the amino acid sequence of the polypeptide expressed, as well as sequences modified by deletions, additions or replacements of nucleotide(s) which cause no substantial deleterious affection to function (including the function of the polypeptide expressed).

The polynucleotide of the present invention includes all recombinant constructs comprising a nucleotide sequence of the invention as defined above. Such recombinant constructs may be designed to express only part of AC. The constructs may include expression control sequence(s) which differ to the control sequence(s) naturally adjoining the AC gene. Optionally, the construct may include a non-AC protein encoding region. Thus the recombinant construct includes constructs encoding for chimeric proteins, which comprise at least part of AC or a functional equivalent thereof.

In a particular embodiment, the present invention provides a vector (such as a cloning or expression vector) which comprises a recombinant construct as defined above. Vectors include conventional cloning and expression plasmids for bacterial and yeast host cells as well as virus vectors such as vaccinia, which may be useful for expression in eukaryotic cell lines. Such a vector may be used to transform a suitable host cell (either for cloning or expression purposes) and the transformed host cell also forms a further aspect of the present invention. If the vector produced is comprised only in part of a nucleotide sequence derived from SEQ ID No 1 it may be appropriate to select a host cell type which is compatible with the vector. Mention may be made of prokaryotic host cells such as E. coli as well as eukaryotic host cells, including yeasts, algae and fish, insect or mammalian cells in culture. Insect cells may be especially useful where a baculovirus expression system is used. Suitable host cells will be known to those skilled in the art.

As a general reference to genetic engineering techniques, mention may be made of Sambrook, Fritsch, Maniatis, in "Molecular Cloning, a Laboratory Manual", Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y., 1989; and also to Old and Primrose "Principles of Genetic Manipulation", 5th edition, 1994.

In particular cell lines derived from the human embryonic kidney cell line 293 (HEK293) have been generated and stably express AC. Briefly, the cells were transfected with pcDNA3 containing the full length AC DNA (clone JP 173--see Example 3) and cells were selected on the basis of resistance to G418 antibiotic (0.8 mg/ml) for 4 weeks. Individual clones of resistant cells were expanded and tested for cAMP production in response to CRF and the effects of immunosuppressants were also examined. At least three cell lines exhibited cAMP production in response to CRF which was enhanced by both cyclosporin A and FK506. In the case of FK506, this effect is blocked by L-685,818. Furthermore, the accumulation of cAMP in the presence of inhibitors of phosphodiesterase is about 10-fold higher than in wild type HEK293 cells. This latter finding suggests that the transfected cyclase is constitutively active and thus may be a mutant of the wild type enzyme.

In a yet further aspect, the present invention provides regulators of AC. Such regulators may act directly on AC itself, and may include (but are not limited to) calcineurin and antibodies specific to AC, or may affect production of AC, including (for example) antisense oligonucleotides (which prevent expression of the AC gene by binding to a portion of DNA preventing transcription and/or by binding to the mRNA preventing translation), and agents which bind to receptors and thus affect the activity of AC, for example .beta.-adrenergic agonists and .beta.-adrenergic antagonists. As .beta.-adrenergic agonists mention may be made of salbutamol, clenbuterol, fenoterol and the like, whereas suitable .beta.-adrenergic antagonists include propranolol.

Antibodies (including monoclonal antibodies) specific to AC may be produced using conventional immunological techniques.

In a yet further aspect, the present invention provides a method of regulating cellular metabolism, wherein said method comprises altering the activity or amount of AC. For example, the activity or amount of AC may be influenced by the regulators described above. Alternatively, for example, the amount of AC could be controlled through genetic manipulation of the genome, by providing agents that affect transcription and/or translation of the AC gene.

The distribution of AC varies between different tissues or cell types and it is postulated that the intracellular concentration of AC may fluctuate as a result of expression control of the AC gene in response to particular stimuli.

Studies of AC distribution in tissues suggest that it is prominently present in skeletal muscle and heart (FIG. 17). Its likely function in these tissues is the coordination between contraction and metabolic demand. Intracellular calcium ions promote contraction, and cAMP promotes glycogenolysis. Feedback of calcium ions on cAMP synthesis would delimit the accumulation of calcium and ensure that the contractions are no longer than what can be metabolically tolerated by the cells. In this respect it is plausible that the levels of cyclase will be regulated by the trophic state of the muscle such as that seen in the training of sportsmen and animals such as racing horses. Thus the measurement of AC levels by means of specific probes would constitute a means of predicting optimal regulation of contraction and metabolism of nutrients.

Thus for example AC may be present at abnormal levels in certain disease states or conditions. Mention may be made in this regard of neurological-based disorders (for example Parkinson's disease, epilepsy), psychiatrically-based disorders (for example, anxiety, major depression disorder, mania, schizophrenia, obsessive-compulsive disorder, Tourette's Syndrome and related tics) endocrine-based disorders (for example Cushing's Syndrome and disease, Nelson's Syndrome, Cohn's Syndrome, glucocorticoid resistance, Graves' disease (thyrotoxicosis) with or without exophthalmia, hyper and hypothyroidism; hyperprolactinaemia and its effects); hypertrophy of the prostate; cardiovascular-based disorders (for example, angina pectoris, cardiac infarction, hypertension (benign and malignant)), pregnancy-based disorders (for example, recurrent spontaneous abortion, pre-eclampsia and eclampsia) respiratory-based disorders (for example, asthma, bronchitis, chronic bronchitis emphysema and cor pulmonale), bone-based disorders (for example, hyperparathyroidism, osteomalacia, Paget's disease, osteoporosis), promotion of bone healing, kidney-based disorders (for example, acute and chronic glomerular nephritis, Albright's Syndrome--or the McCune Albright's Syndrome), gut-based disorders (for example, ulcerative colitis, irritable bowel, Crohn's disease, Hirschssprung's disease), tumours (benign and malignant), especially ovarian tumours and prostate tumours, and in the control or promotion of fertility. AC has been found to be of greater prevalence in the brain (in particular the cortex, striatum and hippocampus regions of the brain), in the ovaries and in the lungs.

Finally, substances developed to regulate the activity of this cyclase may be useful as improvers of metabolic balance in ischemic heart muscle and syndromes of skeletal muscle atrophy.

In a further aspect, therefore, the present invention provides a method of treating such conditions by control of AC activity. An example of such control could be the design of compounds that mimic the steric conformation of amino acids in positions 503 to 610, especially 503 to 570, of SEQ ID No 99 which may function as a calcineurin binding-site.

The amino acid sequence of human AC has been analysed (analogously to that of the mouse sequence in Example 3) and a domain corresponding to the immunophilin protein FKBP12 (which potently inhibits calcineurin) has been located in residues 594 to 611. The sequence similarity strongly suggests that this is the calcineurin binding site.

Other adenylyl cyclases were also examined for potential sequence similarities with the FKPB12 and 13 proteins as described in Example 3. A summary of the findings is shown in FIG. 15. Briefly, all of the known adenylyl cyclase sequences appear to show significant similarity with FKBPs in the area that corresponds to the junction of the C1.alpha. and C1.beta. domains (see Example 3 for nomenclature). As shown in FIG. 16 the alignments of individual cyclases with FKBP12 are distinct i.e. different portions of the C1.alpha. high homology region are part of the FKBP like sequence. These alignments are possible because of modular repeats in the cyclase amino acid sequence in this region. The guiding motif for these alignments is the requirement of the presence of an "80s" loop of FKBP12 in a homologous position in the cyclase "FKBP like" sequence (Yang et al, (1993) Journal of the American Chemical Society 115:819-820). It is of note for instance, that ACtypel, a calcium activated adenylyl cyclase contains a significant portion (sequence: FGPLI) of the 80s loop of FKBP12 which is essential for the interaction with calcineurin (Yang et al, 1993 supra).

Taken together with the current knowledge on the regulation of ACtypel it appears that the C1.alpha.-C1.beta. junction domain is an important regulatory site of adenylyl cyclase activity. It is proposed that all adenylyl cyclases are regulated by tightly associated calcineurin bound in this region. This association may eventually prove to be dependent on calcium ions. It is also of note that there are potential protein kinase A, CAM kinase II and casein kinase II phosphorylation sites in this area in most adenylyl cyclases, which could be substrates of calcineurin and/or may influence the avidity of association of calcineurin with the cyclases.

The FKBP12 like domain (residues 594-611) of mouse AC has been expressed as a Glutathione-S-transferase (GST) fusion protein in the expression vector pGEX-2T (Pharmacia). The cyclase portion of this fusion protein can be phosphorylated by cAMP dependent protein kinase and dephosphorylated in a calcium dependent manner by calcineurin. These data support the notion that the region of the cyclase described here is a clacineurin binding site and contains phosphorylated amino acids that are substrates for calcineurin.

Data recently published show a remarkable similarity between a phosphorylation site for myotonic dystrophy protein kinase on the .beta.-subunit of the L type skeletal muscle calcium channel (Timchenko et al, (1995) Proceedings of the National Academy of Sciences of the United States of America 92:5366-5370) and the FKBP-like domain of AC as shown below (wherein .cndot. shows amino acid identity and .vertline. shows functionally conservative subsitutions): ##STR1##

In addition, it has been reported that the calcium .beta.-subunit is selectively dephosphorylated by calcineurin (Lai et al. (1993) Journal of Neurochemistry 61:1333-1339). It is therefore likely that Ser.sub.600 or Ser.sub.601 in AC may be phosphorylated by myotonic dystrophy protein kinase and dephosphorylated by calcineurin. Abnormalities in this process may occur in myotonic dystrophy and contribute to the symptoms of this hereditary disorder, especially with respect to the pathological changes of brain, muscle, heart and anterior pituitary function.

In a yet further aspect, the present invention also provides a diagnostic assay, to determine the presence and/or the amount of AC within a sample, said method comprising contacting an agent specific to AC with said sample and determining the presence and/or amount of complex formed. The agent may be, for example, an anti-sense oligonucleotide which is complementary to the mRNA of AC. Alternatively, the agent may be an antibody specific to AC, or may be calcineurin.

In a preferred embodiment the diagnostic agent is immobilised on a support, for example a membrane. Further, it is also preferred that a labelling moiety is present in the assay, so that measurement of the agent/AC complex is simplified.

The present invention will now be further described with reference to the following examples and with references to the accompanying figures in which:

FIG. 1

Effect of FK506, cyclosporin A (CsA) and MeVal.sup.4 -cyclosporin A (MeVal.sup.4 CsA) (SDZ202-384) on cAMP accumulation induced by 10 nmol/l CRF in AtT20 cells in the presence of 0.5 mmol/l isobutylmethylxanthine. Basal cAMP production was 0.6.+-.0.08 pmol/well. Data are means.+-.SEM, expressed as percentage of the increment caused by 10 nmol/l CRF which was 6.1 pmol/well.

FIG. 2

A: Effect of cyclosporin A on the time-course of basal and CRF-induced cAMP formation. Data are from a representative of 2 experiments. Means.+-.SEM, n=3 for CRF treated groups (.smallcircle.=10 nmol/l CRF+vehicle, .circle-solid.=CRF+1 .mu.mol/l cyclosporin A) and n=1 for basal (.DELTA.=vehicle, .tangle-solidup.=cyclosporin A), 0.5 mmol/l IBM, 0.1 mM rolipram and cyclosporin A were given as pre-treatment for 30 minutes at 37.degree. C. before the application of 3 nM CRF for 10 minutesat 24.degree. C. *P<0.05 compared with respective control group (one way ANOVA followed by orthogonal contrasts).

B: Dependence of the effect of 1 .mu.mol/l FK506 on CRF-induced cAMP formation in the presence of 0.5 mM IBMX. FK506 and IBMX are given as for FIG. 2A. Data are means.+-.SEM, n=4-6/group. (.quadrature.=FK506 group, .DELTA.=control group)*P<0.05 compared with respective vehicle treated group (one-way ANOVA followed by orthogonal contrasts).

FIG. 3

A: The effect of L685,818 on the enhancement of CRF-induced cAMP formation by FK506. Data are means.+-.SEM, the cAMP level in the presence of 10 nmol/l CRF taken as 100% was 12 pmol/well*10 minutes, basal levels were 1.6 pmol/well*10 minutes.

B: The effect of L685,818 on the inhibition of calcineurin activity by 1 .mu.mol/l FK506 in AtT20 cells. Calcineurin activity was measured by the phosphocasein method, the activity measured in the absence of FK506 was taken as 100%.

FIG. 4

Effects of various manipulations that reduce intracellular free calcium concentration on CRF-induced cAMP accumulation in AtT20 cells. IBMX 0.5 mmol/l present throughout. All data are expressed as the percentage of CRF-induced cAMP-formation, means.+-.SEM, n=4/group.

EGTA: cells exposed to 2 mmol/l EGTA in calcium free medium during the preincubation period; nimodipine: cells exposed to 1 .mu.mol/l nimodipine during preincubation period;

BAPTA-AM: cells incubated in 20 .mu.mol/l [1,2-bis-(o-Aminophenoxy)-ethane-N,N,N',N'-tetraacetic acid tetra-(acetoxymethyl)-ester] during the preincubation period.

FIG. 5

Effect of FK506 on the inhibition of CRF-induced cAMP accumulation by extracellular calcium ions in AtT20 cells. Cells were depleted of calcium by preincubation in 2 mmol/l EGTA, calcium free medium, containing 5 .mu.mol/l A23187 and 5 .mu.mol/l nimodipine, and graded amounts of CaCl.sub.2 were added with 10 nmol/l CRF. The values on the abscissa give the nominal free extracellular calcium ion concentration. Data are means.+-.SEM, n=6/group.

FIG. 6

Effect of pertussis toxin (1 .mu.g/ml for 18 hours) on the modulation of CRF-induced cAMP formation by FK506 and somatostatin. AtT20 cells were preincubated with FK506 or somatostatin for 30 minutes. IBMX (0.5 mmol/l) present throughout. Data are means.+-.SEM, n=6/group.

FIG. 7

A: FK506 (0.5 .mu.mol/l) on basal and CRF-induced ACTH secretion in AtT20 cells.

B: Antagonism of the effect of FK506 (0.5 .mu.mol/l) on CRF-induced ACTH release by L685,818, which had no effect on basal ACTH secretion in this system even at 5 .mu.mol/l. n=4/group, means.+-.SEM.

In panel A the values for basal and CRF-stimulated ACTH release taken as 100% were 15.+-.1 and 22.+-.1 fmol/well*30 minutes, respectively.

FIG. 8

Comparison of the sequence (SEQ ID NO: 64) amplified from AtT20 cell RNA using primer set B, with the corresponding sequences with other adenylyl cyclases found in current databases (EMBL, GenBank, SwissProt (SEQ ID NOS: 48 to 64)). The numbers relate to the amino acid sequence of rabbit ACtype5 (ocmradcyv) (SEQ ID NO: 60). The sequence is annotated by .cndot. showing amino acid identity of the novel sequence with at least one previously reported AC, .smallcircle. shows funtionally conservative substitutions, - denotes non-conservative subsitutions.

Abbreviations and Sequence Identification Numbers

hum7--Human ACtype7 (GenBank #D25538) (SEQ ID NO: 48);

mmu12919--mouse ACtype7 (SEQ ID NO: 49);

cya2_rat--rat ACtype2 (SEQ ID NO: 50);

cya4_rat--rat ACtype4 (SEQ ID NO: 51);

hsadencyr8--Human ACtype8 (SEQ ID NO: 52);

ratacviii--rat ACtype8 (SEQ ID NO: 53);

a46187--human ACtype5 (SEQ ID NO: 54);

cya6_mouse--mouse ACtype6 (SEQ ID NO: 55);

a49201--mouse ACtype5 (now renamed as mouse AC6) (SEQ ID NO: 56);

cya6_rat--rat ACtype6 (SEQ ID NO: 57);

cya6_canfa--dog ACtype6 (SEQ ID NO: 58);

s29717--rat ACtype5 (SEQ ID NO: 59);

ocmradcyv--rabbit ACtype5 (SEQ ID NO: 60);

cya5_canfa--dog ACtype5 (SEQ ID NO: 61);

cya1_bovin--bovine ACtypel (SEQ ID NO: 62);

cya3_rat--rat ACtype3 (SEQ ID NO: 63);

AC--AtT20 derived new sequence, ie AC (SEQ ID NO: 64).

FIG. 9

Pileup analysis of adenylyl cyclase sequences found in the EMBL and GenBank data bases.

legend on dendrogram adenylyl cyclase isoform hum7 Human type 7 GenBank #D25538; mmu12919 Mouse type 7; cya2_rat Rat type 2; cya4_rat Rat type 4; cya5_canfa Dog type 5; cya5_rabit Rabbit type 5; s29717 Rat type 5; cya6_mouse Mouse type 6; cya1_bovin Bovine type 1; ratacviii Rat type 8; cya3_rat Rat type 3; Ac Mouse AC 9, ie AC.--

FIG. 10

Hydrophobicity plot of AC according to Kyte et al (1982) J Mol Biol 157:105. The numbers show predicted intramembrane domains.

FIG. 11

Functional Analysis of Cloned AC in Transfected Host Cells

Levels of cAMP in transiently transfected HEK293 (a-c) and COS7 (d) in the presence of phosphodiesterase inhibitors (1 mmol/l isobutylmethylxanthine) and 0.1 mmol/l rolipram). Data shown are means.+-.SEM, n=4/group. Representative data from two similar series of experiments.

(a) HEK293 cells transfected with vector cDNA and pretreated with 0.2% (v/v) ethanol (empty columns) or 2 .mu.mol/l FK506 (striped columns) before being challenged with 0.5 nmol/l CRF.

(b) as (a) except with cells transfected with AC (Data are means.+-.SEM, n=4/group. *P<0.05 when compared with the respective vehicle-treated group, 1-way ANOVA, followed by Newman-Keuls test).

(c) The effect of FK506 on CRF-induced cAMP production was analysed further in (c). The enhancement of the CRF-induced cAMP response by FK506 was blocked by the FK506 antagonist drug L-685,818 (100 .mu.mol/l) (Dumont et al, (1992) J. Exp. Med. 176:751-760), which had no effect on CRF-induced cAMP formation when given alone.

It did, however, similarly to FK506, increase the unstimulated levels of cAMP to 3-fold above basal (not shown). *P<0.05 when compared with the vehicle treated group, 1-way ANOVA, followed by Newman-Keuls test.

(d) Application of cyclosporin A, another calcineurin blocking immunosuppressant enhanced the cAMP response to 1 nmol/l CRF in COS7 cells transfected with AC (*P<0.05, when compared with the vehicle-treated group).

FIG. 12

An FKBP-Like Domain in AC

(A) Schematic drawing of the predicted structure of AC, nomenclature as suggested by Gilman and coworkers (Taussig et al, (1995) J. Biol. Chem. 270:1-4). N=N-terminal intracellular loop; M1 and M2=membrane spanning segments; C1.alpha. and C2.alpha. (thick line)=the highly conserved, putative catalytic cyclase domains; C1.beta. and C2.beta.=the non-conserved, putative regulatory domains of the intracellular loops.

(B) AC.sub.(503-610) (SEQ ID NO: 65) shows approximately 40% overall sequence similarity with FK506 binding-protein 12 (FKBP12) (SEQ ID NOs: 66 to 69). Residue numbers in the top row correspond to AC, in the bottom row to mammalian FKBP12s. A dot indicates sequence identity between AC and at least one of the FKBPs. Vertical lines denote conserved substitutions which were defind as 1) C 2) S T P A G 3) N D E Q 4) H R K 5) M I L V 6) F Y W (see Krupinski et al, (1989) Science 244:1558-1564).

yst=Saccharomyces cerevisiae, hum=human, mus=mouse, ncr=Neurospora crassa.

(C) Alignment of the C1.alpha. and C1.beta. junction region of known mammalian adenylyl cyclase isotypes (SEQ ID NOs: 70 to 78). The underlined region in bovACtp1 (SEQ ID NO: 70) corresponds to 24 residues of the 28 amino acid residue putative calmodulin binding site (495-522) (Vorherr et al, (1993) Biochemistry (U.S.A.) 32:6081-6088; and Wu et al, (1993) J. Biol. Chem. 286:23766-23768), bov=bovine. Abbreviations as for FIG. 15.

FIG. 13

Northern Analysis of AtT20-Cell, HEK293-Cell and Mouse Striatal RNA

.sup.32 P-labelled cDNA probes derived from the AC sequence were used. Note approximately 9 kb band in AtT20 cells that hybridizes with both probes. A similar size RNA species is also intensively hybridizing in mouse striatal RNA and a weak, barely discernible band is found in HEK293 cells. The relative hybridisation intensities (average of the two lanes) with probe jp 164 were: AtT20 2.1; HEK293 0.3; mouse striatum 1.4.

FIG. 14

Localisation of AC mRNA in Mouse Brain

AC mRNA detected by a .sup.35 S-CTP labelled antisense ribonucleic acid probe derived from plasmid JP142 in (A) the hippocampus (pyramidal cell layer of CA1-CA4 (a) and of the subiculum (b), granule cells in dentate gyrus (c) and in various parts of the cereral cortex (posterior cingulate cortex indicated by (d)). (B) .sup.35 S-CTP-sense RNA probe control.

Magnification: 25.times..

FIG. 15

An alignment of adenylyl cyclases with FKBPs (SEQ ID NOs: 79 to 86).

.cndot. indicates identity; .vertline. indicates conserved substitution in at least one of the FKBPs. Conserved substitutions are defined as

1) C

2) STPAG

3) NDEQ

4) HRK

5) MILV

6) FYW

Abbreviations and Sequence Identifiers

YeastFKBP12 Yeast FKPB12 (SEQ ID NO: 79) humFKBP12 human FKBP12 (SEQ ID NO: 80) musFKBP12 mouse FKBP12 (SEQ ID NO: 81) NcrFKBP12 Neurospora crassa FKBP 12 (SEQ ID NO: 83) humFKBP13 human FKBP13 (SEQ ID NO: 84) fkb2_hum human FKBP13 precursor (SEQ ID NO: 85) fkb2_yeast yeast FKBP13 (SEQ ID NO: 86) bovACtp1 bovine adenylyl cyclase type 1 ratACtp3 rat adenylyl cyclase type 3 ratACtp8 rat adenylyl cyclase type 8 rutabaga drosophila calmodulin activated adenylyl cyclase ratACtp5 rat adenylyl cyclase type 5 ACtp5sp1 dog adenylyl cyclase type 5 splice variant ratACtp6 rat adenylyl cyclase type 6 musAC mouse adenylyl cyclase type 9 (AC) ratACtp2 rat adenylyl cyclase type 2 ratACtp4 rat adenylyl cyclase type 4 musACtp7 mouse adenylyl cyclase type 7 humACtp7 human adenylyl cyclase type 7

FIG. 16

Illustrates cyclases extracted from their alignments with FKBPs.

Abbreviations and Sequence Identifiers

1. bovine adenylyl cyclase type 1 (SEQ ID NO: 87)

2. rat adenylyl cyclase type 3 (SEQ ID NO: 88)

3. rat adenylyl cyclase type 8 (SEQ ID NO: 89)

4. drosophila calmodulin activated adenylyl cyclase (SEQ ID NO: 90)

5. rat adenylyl cyclase type 5 (SEQ ID NO: 91)

6. dog adenylyl cyclase type 5 splice variant (SEQ ID NO: 92)

7. rat adenylyl cyclase type 6 (SEQ ID NO: 93)

8. mouse adenylyl cyclase type 9 ie AC (SEQ ID NO: 94)

9. rat adenylyl cyclase type 2 (SEQ ID NO: 95)

10. rat adenylyl cyclase type 4 (SEQ ID NO: 96)

11. mouse adenylyl cyclase type 7 (SEQ ID NO: 97)

FIG. 17

Illustrates the relative hybridisation intensity in different tissues, showing the distribution of AC mRNA in mouse, as measured by RNase protection assay.

FIG. 18

Construction of Complete Human AC9 cDNA

Scheme showing relationship of overlapping human AC9 cDNA clones used in construction of the complete hAC9 cDNA (GenBank/EMBL Accession Number AJ133123) by joining clones HiBBC33, pJP192, pJP230 and 166657 at the Mlu I, Eco RV and Nco I sites indicated as described in materials and methods. The relationship of these clones to human AC9 cDNA clones isolated by other groups (GenBank/EMBL Accession Numbers AF036927 and AB011092) identified in subsequent searches of the GenBank/EMBL DNA sequence database is also represented.

FIG. 19

Immunocytochemical Detection of Human AC9 Protein

Cells cultured on coverslips were briefly fixed with 3% paraformaldehyde and reacted with affinity purified chicken antibodies directed against residues 1340-1353 of the mouse AC9 protein, with a tyrosine added at the N-terminus. Subsequently, rabbit anti-chicken IgG coupled to FITC was applied, and the cells were viewed under a Leica TCS-NT confocal microscope system. Note (a) preferential localisation of immunoreactivity in the vicinity of the plasma membrane and (b) dividing cell with membrane localisation of AC9. (c) Specificity control with antibody preincubated with 10 .mu.M of peptide antigen. The bar represents 10 .mu.m.

FIG. 20

Detection of Multiple Human AC9 Polypeptides by Immunoblotting

Protein extracts from membranes of HEK 293 cells, either untransfected (a & b, lane 3) or stably transfected with human AC9 cDNA (a & b, lanes 1,2,4&5) separated by SDS-PAGE and electroblotted onto PVDF membrane were incubated with antisera directed against mouse AC9 N-terminus (a) or C-terminus (b). Positions of molecular weight markers (KDa) are indicated. A major polypeptide of 175 KDa was detected in two independent stable human AC9 transfectants (a & b, lanes 1 & 2). In addition, a faster migrating polypeptide was also detected. Specificity was confirmed by competition with peptide antigen (a & b, lanes 4 & 5). Increased background staining was apparent when C-terminal peptide was included (b, lanes 4 & 5).

FIG. 21

Analysis of cAMP Synthesis by HEK293 Cells Overexpressing Human AC9

a) Time-course of cAMP accumulation. Blockers of cyclic nucleotide degrading phosphodiesterases (PDE) (isobutylmethylxanthine 1 mM, and rolipram 0.1 mM) and various amounts of Ca.sup.2+ /EGTA were applied at time 0 to cells previously depleted of Ca.sup.2+ in HBSS/EGTA medium containing no added Ca.sup.2+, 1 .mu.M thapsigargin and 10 .mu.M ryanodine. The final extracellular levels of Ca.sup.2+ were (.DELTA.) nominally 0, (.smallcircle.) 2 mM, (n) 5 mM. Data are means.+-.S.E.M. n=4 representative of 2 similar experiments. *P<0.05 when compared with the zero Ca.sup.2+ group, 1-way ANOVA followed by Dunnett's test. b) Effect of FK506 on the inhibitory effect of Ca.sup.2+ in cells pretreated with Ca.sup.2+ free medium containing 5 .mu.M 4-Br-A23187. Blockers of PDE were added in Ca.sup.2+ /EGTA to achive 0.25 and 0.5 mM extracellular Ca.sup.2+. Data are expressed as multiples of the cAMP content measured in the absence of PDE blockers, means.+-.S.E.M. n=4 representative of 3 similar experiments. * P<0.05 when compared with respective vehicle treated controls, +P<0.05 when compared with respective 0 Ca.sup.2+ group. c) Pharmacological analysis of the effect of various immunosuppressant compounds on Ca.sup.2+ inhibition of cAMP production. FK506, cyclosporin A (CsA), and rapamycin were used at 10 .mu.M, L865,818 was present at 50 .mu.M 20 min before the application of PDE blockers and Ca.sup.2+ /EGTA. Data are expressed as multiples of the cAMP content measured in the absence of PDE blockers, means.+-.S.E.M. n=4 representative of 2 similar experiments. * P<0.05 when compared with respective vehicle treated controls.

FIG. 22

Differential Expression of Human AC9 mRNA

a) A probe derived from 5' end of human AC9 cDNA (see materials and methods) detected an 8.5 kb transcript in all tissues represented on a northern blot of mRNAs from human heart (H), brain (B), placenta (Pl), lung (Lg), liver (Lv), skeletal muscle (M), kidney (K) and pancreas (Pn) as well as a 6.3 kb mRNA in heart and skeletal muscle. The control probe (b) confirmed expression of a 2 kb .beta.-actin mRNA in all tissues as well as a 1.8 kb mRNA in heart and skeletal muscle. c) The same northern blot re-probed with a cDNA fragment derived from the extended 3' UTR of human AC9 clone KIAA0520 (see materials and methods) shows exclusive detection of an 8.5 kb MRNA. Positions of molecular weight markers (kb) are indicated.

FIG. 23

Localisation of Human AC9 mRNA in Human Forebrain

a) Autoradiogram showing protection of 314 bases of the single-stranded human AC9 riboprobe derived from clone pJP195 in samples of total RNA from post-mortem human cortex, hippocampus and striatum as well as in the HEK 293 cell positive control (as indicated by the arrow). No protection of probe by yeast RNA (Y) was observed. Probe present prior to single-strand specific nuclease digestion is shown in the first lane (pr). b) In situ hybridisation using .sup.35 S UTP-labelled anti-sense (i) and sense (ii) riboprobes derived from clone pJP195 incubated with 15 .mu.m sections from normal adult post-mortem human brain showing detection of human AC9 mRNA by the anti-sense probe in hippocampal neurones of the CA1-CA3 regions (upper arrow) and dentate gyrus (lower arrow).

EXAMPLE 1

Calcineurin Feedback Inhibition of Agonist Evoked cAMP Formation

Introduction

The major immunosuppressant compounds cyclosporin A and FK506 are potent blockers of the Ca.sup.2+ /calmodulin-regulated protein phosphatase calcineurin (protein phosphatase 2B) in leukocytes (Liu et al, (1991) Cell 66:807-815). This observation has led to the discovery that calcineurin is an essential element of the signal transduction pathway activated by the T-cell receptor (Schreiber (1992) Cell 70:365-368 and Sigal et al, (1992) Ann. Rev. Immunol. 10:519-560). Immunophilins, which are proteins that mediate the effects of immunosuppressants on calcineurin activity in lymphocytes (Schreiber (1992) Cell 70:365-368) have been identified in the brain (Steiner et al, (1992) Nature 358:584-586) providing the plausible molecular targets of the prominent neurological side-effects of FK506 and cyclosporin A (Frank et al, (1993) Transplantation Proceeding 25:1887-1888 and Reyes et al, (1990) Transplantation 50:10434-1081).

While calcineurin is highly abundant in the brain (0.5-1% of total protein) (Klee et al, (1988) Advances in Enzymology 61:149-200) its functions in excitable cells remain to be defined. Calcineurin has been implicated in the control of voltage regulated ion channel activity (Armstrong, (1989) Trends in Neurosciences 12:117-122), particularly with respect to L-type calcium channels (Lai et al, (1993) J. Neurochem. 61:1333-1339). More recent studies have shown that the synaptic vesicular protein dynamin, which is thought to participate in synaptic vesicle recycling in nerve endings, is a prominent substrate for calcineurin (Liu et al, (1994) Science 265:970-973). Furthermore, blockage of calcineurin by immunosuppressants enhanced glutamate release by synaptosomes prepared from rat brain and this appeared to correlate with the state of phosphorylation of dynamin (Nichols et al, (1994) J. Biol. Chem. 269:23817-23823).

With respect to secretory function in other systems, it has been reported (Antoni et al, (1993) Biochem. Biophys. Res. Commun. 194:226-233) that immunosuppressants block calcineurin activity in pituitary corticotrope tumour (AtT20) cells and stimulate Ca.sup.2+ -dependent hormone release in correlation with their calcineurin blocking activity. Cyclic AMP (cAMP) is a cardinal signalling molecule in pituitary corticotrophs that causes an increase of intracellular free Ca.sup.2+ ([Ca.sup.2+ ].sub.i) and triggers the release of adrenocorticotrophic hormone (ACTH) (Antoni, (1986) Endocr. Rev. 7:351-378). Because [Ca.sup.2+ ].sub.i is known to inhibit cAMP formation in a variety of systems (Cooper et al, (1993) Trends Pharm. Sci. 14:34-36) we have analyzed in AtT20 cells the effects of FK506 and cyclosporin A on cAMP production induced by the hypothalamic neuropeptide corticotropin-releasing factor (CRF) and beta-adrenergic stimulation.

Experimental Procedures

AtT20 D16:16 mouse anterior pituitary tumour cells were maintained in culture as previously described (Woods et al, (1992) Endocrinology 131:2873-2880). For measurements of ACTH release, cAMP production or calcineurin activity the cells were plated on 24-well tissue culture plates (5.times.10.sup.4 cells/well), used 4-6 days afterwards. ACTH (Woods et al, (1992) Endocrinology 131:2873-2880) and cAMP (Dufau et al, (1973) Endocrinology 92:6-11) were measured by specific radioimmunoassays. Calcineurin protein phosphatase activity was determined by the .sup.32 P-labelled casein assay (Tallant et al, (1984) Arch. Biochem. Biophys. 232:269-279) or by using the .sup.32 P-labelled RII peptide substrate (Blumenthal et al, (1986) J. Biol. Chem. 261:8140-8145) adapted to measure calcineurin phosphatase activity in AtT20 cell extracts as previously described (Antoni et al, (1993) Biochem. Biophys. Res. Commun. 194:226-233).

Experiments for cAMP were all carried out in Hanks balanced salt solution containing 2 mM CaCl.sub.2 and 1 mM MgSO.sub.4 buffered with 25 mM HEPES pH 7.4 and supplemented with 0.25% (w/v) of bovine serum albumin. The cells were preincubated in serum free medium for 1 hour, after which blockers of phosphosdiesterase (isobutylmethylxanthine 0.5 mM (IBMX) and/or rolipram 0.1 mM) along with various other treatments were applied for 30 minutes at 37.degree. C. Subsequently the cells were cooled to 220.degree. C. for 5 minutes in a water bath and agonists were added for 10 minutes. The reaction was stopped by the addition of 0.2 mol/l HCl to achieve a final concentration of 0.1 mol/l (Brooker et al, (1979) Adv. in Cyclic Nucl. Res. Vol. 10, G. Brooker, P. Greengard and G. A. Robinson, Raven Press, New York, 2-34). In the absence of phosphodiesterase blockers agonist-induced changes of total cAMP content (cells+medium) were small (2-3 fold of basal) and no increment of intracellular cAMP could be detected (Woods et al, (1992) Endocrinology 131:2873-2880). In the presence of IBMX total cAMP content increased linearly with time up to 10 minutes after the addition of CRF and remained constant for up to 20 minutes. In contrast, cellular cAMP content peaked between 2-5 minutes and subsequently declined to basal levels even in the presence of the phosphodiesterase blockers. Hence, after establishing that immunosuppressant drugs had the same effect on peak cellular and total cAMP content under these conditions, all experiments shown here report total cAMP content.

In some experiments cells were preincubated for 30 minutes in medium containing 2 mmol/l EGTA and no added Ca.sup.2+, supplemented with 5 .mu.mol/l A23187 and 2.5 .mu.mol/l nifedipine in order to deplete rapidly mobilized cellular stores of Ca.sup.2+. This treatment also ensured that L-type Ca.sup.2+ channels, the principal avenue of voltage-regulated Ca.sup.2+ influx in AtT20 cells (Luini et al, (1985) Proc. Natl. Acad. Sci. U.S.A. 82:8034-8038; Reisine et al, (1987) Mol. Pharmacol. 32:488-496 and Antoni et al, (1992) J. Endocrinol. 133:R13-R16) were fully blocked and Ca.sup.2+ subsequently added to the extracellular fluid would enter largely through the pores made by the ionophore A23187. The rationale for this pretreatment is that calcineurin reportedly influences L-channel activity (Lai et al, (1993) J. Neurochem. 61:1333-1339; and Armstrong et al, (1987) Proc. Natl. Acad. Sci. U.S.A. 84:2518-2522), whereas the treatment regimen used here would make Ca.sup.2+ entry independent of this regulation.

Immunosuppressant analogues (FK506, courtesy of Fujisawa Ltd, Osaka, Japan; cyclosporin A and SDZ 220-384 (MeVal.sup.4 -cyclosporin A) (Fliri, H. (1993) Antibiotics and antiviral compounds, Chemical synthesis and modification K. Krohn, H. Kirst And H. Maasg, VCN Verlagsgesellschaft mbH, Weinheim, 229-240), courtesy of Sandoz Pharma, Basel, Switzerland, L685,818 (Dumont et al, (1992) J. Exp. Med. 176:751-760) courtesy of Merck&Co, Rahway, N.J.) were also applied during the preincubation period. These compounds were made up in ethanol at 10.sup.-3 mol/l and diluted with the incubation medium to the desired final concentrations. In some cases cells were preincubated with L685,818 the structural analogue of FK506 that binds to FKBP-12 and inhibits prolyl-isomerase activity but is devoid of immunosuppressant activity (Dumont et al, (1992) J. Exp. Med. 176:751-760) for 10 minutes before the addition of FK506 for 30 minutes at 37.degree. C.

Incubations for ACTH secretion were carried out as previously described (Woods et al, (1992) Endocrinology 131:2873-2880) except that the test incubation with CRF was at 22.degree. C.

Amplification and DNA Sequence Analysis of Adenylyl Cyclases cDNAs in AtT20 Cells

Total RNA was prepared from approximately 10.sup.7 cells using Trizol reagent (GIBCO, Paisley, UK) according to the manufacturer's instructions. RT-PCR was carried out using an RNA PCR kit (Perkin Elmer, Warrington, Cheshire, U.K.). Briefly, 0.8 .mu.g of total RNA was denatured at 95.degree. C. for 5 minutes then annealed with 2.5 .mu.M random hexanucleotide primers for first-strand cDNA synthesis which was carried out for 15 minutes at 42.degree. C. in a 20 .mu.l reaction mixture containing 10 mM Tris.HCl, pH8.3, 50 mM KCl, 5 mM MgCl.sub.2, 1 mM each dNTP, 20U RNase inhibitor and 50U MMLV reverse transcriptase. The reaction was terminated at 99.degree. C. for 5 minutes then cooled to 4.degree. C. and stored on ice. PCR was performed using degenerate oligonucleotides, either pair A or pair B corresponding to highly conserved regions within the first (pair A: 5'CTCATCGATGGIGAYTGYTAYTAYTG3' (SEQ ID NO: 5); 3'GGCTCGAGCCAIACRTCRTAYTGCCA5' (SEQ ID NO: 6) expected product size 220 bp) and second (pair B: 5'GAAGCTTAARATIAARACIATIGGI.sup.T /.sub.A.sup.C /.sub.G IACITAYATGGC3' (SEQ ID NO: 7); 3'GGGATCCACRTTIACIGTRTTICCCCAIATRTCRTA5' (SEQ ID NO: 8) expected product size 180 bp) cytoplasmic domains of previously cloned mammalian adenylyl cyclases (Yoshimura et al, (1992) Proc. Natl. Acad. Sci. U.S.A. 89:6716-6720; Krupinski et al, (1992) J. Biol. Chem. 267:24859-24862; and Gao et al, (1991) Proc. Natl. Acad. Sci. U.S.A. 88:10178-10182). For PCR the reverse transcription reaction (20 .mu.l) was expanded to 100 .mu.l and contained 10 .mu.M Tris.HCl, pH 8.3, 50 mM KCl, 2 mM MgCl.sub.2, 200 .mu.M each dNTP, 35 pmol of each primer and 2.5U Amplitaq DNA polymerase. PCR reactions were overlaid with mineral oil (Sigma, Poole, Dorset, U.K.) and denatured at 95.degree. C. for 3 minutes followed by 5 cycles (60 seconds denaturation at 94.degree. C., 60 seconds annealing/extension at 45.degree. C.) then a further 35 cycles (60 seconds denaturation at 94.degree. C., 60 seconds annealing/extension at 55.degree. C.) and finally 7 minutes annealing/extension at 55.degree. C. An aliquot (5%) of each reaction was analysed by agarose gel electrophoresis (3% FMC, Flowgen Instruments Ltd, Sittingbourne, Kent, U.K.). Products within the expected size range for each primer pair were excised from the gel, purified using a Wizard.TM. PCR Prep kit (Promega, Madison, Wis., U.S.A.) and ligated into the vector pGEM-T (Promega). Clones containing an insert of the expected size were identified and their DNA sequence determined by the dideoxynucleotide method (Sequenase 2.0 kit, USB, Amersham International, Aylesbury, U.K.).

Detection of mRNA Expression

Northern analysis was performed using standard procedures. Briefly, 10 .mu.g of total RNA was separated by formaldehyde gel electrophoresis and transferred by blotting onto positively charged nylon membrane (Appligene,) then fixed by baking at 80.degree. C. and prehybridised at 42.degree. C. for 2 hours in 50% deionised formamide, 5.times.SSPE, 0.5.times.Denhardt's, 0.1% w/v SDS, 0.2 mg/ml denatured salmon sperm carrier DNA and 10% Dextran sulphate. Random-primed labelled DNA probe (50 ng; >10.sup.9 cpm/.mu.g) was then added and hybridisation continued overnight at 42.degree. C. The membrane was washed twice for 20 minutes in 2.times.SSC/0.1% SDS, followed by 20 minutes in 1.times.SSC/0.1% SDS at 42.degree. C. and finally 20 minutes in 0.5.times.SSC/0.1% SDS at 50.degree. C. before wrapping in cling-film an exposing to autoradiographic film at -70.degree. C. Ribonuclease protection assays were performed using an RPA II kit (Ambion, AMS Biotechnology, Witney, Oxon, U.K.) according to the manufacturer's instructions. Briefly, 10 .mu.g of total RNA was hybridised overnight at 45.degree. C. to 10.sup.5 cpm of radiolabelled anti-sense riboprobe. Following hybridisation reactions were digested with single-strand-specific RNase and protected fragments resolved on a 6% denaturing polyacrylamide gel which was fixed for 30 minutes in 15% methanol/5% acetic acid, dried and exposed to autoradiographic film at -70.degree. C.

Results

Enhancement of CRF-Stimulated cAMP Production by Immunosuppressants

Immunosuppressant blockers of calcineurin activity, FK506, cyclosporin A and MeVal.sup.4 -cyclosporin A enhanced CRF-induced cAMP production in a concentration dependent manner (FIG. 1). The effect of cyclosporin A (FIG. 2) was statistically significant (P<0.05 or less) at 2, 5 and 10 minutes after the addition of CRF, similar data were also obtained with FK506. Enhancement of cAMP formation by FK506 was apparent at lower concentrations (0.1-10 nmol/l) of CRF, while the maximal response appeared unchanged (FIG. 2B).

The effect of FK506 on CRF-induced cAMP production (FIG. 3) could be antagonized by the FK506 analog L685,818, a potent inhibitor FKBP-12 prolyl isomerase activity which has no immunosuppressant activity and does not block calcineurin (Dumont et al, (1992) J. Exp. Med. 176:751-760) and hence, is a specific antagonist of the calcineurin inhibitory action of FK506. Importantly, this compound also blocked the inhibitory effect of FK506 on calcineurin-mediated dephosphorylation of phosphocasein (FIG. 3) in AtT20 cells.

Receptor-Evoked Synthesis of cAMP is Under Inhibitory Control by Intracellular Ca.sup.2+ and Calcineurin

Lowering of intracellular free Ca.sup.2+ by a variety of methods, such as depletion by EGTA and the calcium ionophore A23187, loading of the cells with the intracellular calcium chelator BAPTA-AM and blockage of calcium channels with the dihydropyridine channel blocker nimodipine, all markedly increased the cAMP response to CRF (FIG. 4). The effect of BAPTA-AM on CRF-induced cAMP formation was statistically significant (P<0.05) by 2 minutes after the addition of CRF and at all subsequent time-points studied up to 20 minutes (not shown).

Addition of graded amounts of Ca.sup.2+ with CRF to cells depleted of Ca.sup.2+ and pretreated with the ionophore A23187 produced a concentration-dependent inhibition of CRF-induced cAMP production to levels seen in non-depleted cells incubated in medium containing 2 mmol/l Ca.sup.2+. The effect of exogenous Ca.sup.2+ could be inhibited by FK506, which, in fact failed to alter cAMP accumulation in the absence of Ca.sup.2+ (FIG. 5).

Site and Specificity of Immunosuppressant Action

The effect of FK506 on CRF-induced cAMP formation was also evident after a 16 hours pretreatment of the cells with pertussis toxin (1 .mu.g/ml) (FIG. 6) which strongly suppressed inhibitory G-protein function as assessed by the attenuation of somatostatin-mediated inhibition of cAMP formation. Pertussis toxin treatment also had no effect on the suppression of CRF-induced cAMP formation by extracelluar Ca.sup.2+ in Ca.sup.2+ -depleted cells (not shown).

FK506 had no significant effect on cAMP accumulation evoked by 10 or 30 .mu.M forskolin, a drug that activates adenylyl cyclase independently of Gs. Loading of the cells with BAPTA-AM caused a small (15%), but statistically significant (P<0.05) enhancement of forskolin-evoked cAMP accumulation (not shown).

Finally, in contrast to the effects of FK506 and cyclosporin A, pretreatment with other blockers of protein phosphatases such as calyculi A (1-30 nmol/l) okadaic acid (0.2-5 .mu.mol/l), caused a concentration-dependent inhibition (up to 80%) of CRF-induced cAMP accumulation (not shown and Koch et al, (1994) Cellular Signalling 6:467-473).

Enhancement of CRF Stimulated ACTH Release by FK506

Blockage of calcineurin activity by FK506 enhanced the release of ACTH evoked by CRF (FIG. 7A), and this action was prevented by L685,818 (FIG. 7B). Note, that the apparent EC.sub.50 s of FK506 to inhibit calcineurin activity in AtT20 cells (Antoni, (1986) Endocr. Rev. 7:351-378), to stimulate ACTH release and to augment cAMP accumulation induced by CRF are all approximately 10 nM.

.beta.-adrenergic Stimulation is Under Similar Regulation by Calcineurin

Isoproterenol induces cAMP accumulation through .beta..sub.2 adrenergic receptors in AtT20 cells (Heisler et al, (1983) Biochemical And Biophysical Research Communications 111:112-119). Isoproterenol induced cAMP formation was also enhanced by BAPTA-AM, FK506 and Ca.sup.2+ -depletion in AtT20 cells (Table 1).

Effect of Immunosuppressants on cAMP Accumulation in AtT20 Cells Correlates With the Expression of a Novel Adenylyl Cyclase mRNA

In order to determine the profile of adenylyl cyclase isoforms present in AtT20 cells two sets of degenerate oligonucleotide primers were used to analyze AtT20 cell total RNA for adenylyl cyclase related sequences by means of RT-PCR. Using primer set B a PCR product of approximately 180 bp was obtained. DNA sequence analysis revealed that approximately 8% of the subcloned 180 bp cDNA fragments amplified proved to be identical to Type 6 adenylyl cyclase. The majority (>90%), however, gave a novel sequence exhibiting a high level of homology with the amino acid sequences of known mammalian adenylyl cyclases present within current databases but which was not identical to any previously reported sequence (FIG. 8). Type 1 adenylyl cyclase was detected in AtT20 cells using primer set A.

Northern blot analysis of total RNA using the novel adenylyl cyclase 180 bp cDNA fragment as a probe indicated hybridisation to an approximately 9 kb mRNA expressed in AtT20 cells. Expression of this message was not detected in NCB20 or HEK293 cell RNA.

As a more sensitive alternative mRNA expression was also assayed by ribonuclease protection of the message using a radiolabelled anti-sense riboprobe transcribed from the novel adenylyl cyclase 180 bp cDNA. The presence of a ribonuclease-protected fragment migrating as a distinct, approximately 155 bp band on a denaturing polyacrylamide gel indicates that the novel adenylyl cyclase mRNA is highly abundant in AtT20 cells whereas only very low levels are present in NCB20 cells and in HEK293 cells the message remains undetected.

Measurements of calcineurin activity in cell extracts from all three cell lines revealed a similar sensitivity of calcineurin protein phosphatase activity (substrate RII subunit peptide Fruman et al, (1992) Proc. Natl. Acad. Sci. U.S.A. 89:3686-3690) towards inhibition by FK506 and cyclosporin A. Furthermore, all three cell lines responded to stimulation by CRF which was enhanced by depletion of intracellular free calcium. FK506 and cyclosporin A enhanced CRF-induced cAMP formation consistently in AtT20 cells; in NCB20 cells only one out of 4 experiments gave a statistically significant effect of cyclosporin A on CRF-induced cAMP accumulation and no effects were observed in HEK293 cells.

Discussion

These data show that receptor-stimulated cAMP formation may be regulated by calcineurin, and that this regulation is associated with the expression of a novel adenylyl cyclase mRNA.

All studies of cAMP formation reported here were carried out in the presence of blocker(s) of phosphodiesterase (IBMX and/or rolipram), and hence the effects observed relate to the synthesis of cAMP rather than its degradation.

Evidence for the involvement of calcineurin in the control of cAMP accumulation is provided by the use of immunosuppressant analogs previously (Antoni et al, (1993) Biochem. Biophys. Res. Commun. 194:226-233; and Fliri (1993) Antibiotics and antiviral compounds, Chemical synthesis and modification K. Krohn, H. Kirst And H. Maasg, VCN Verlagsgesellschaft mbH, Weinheim, 229-240) shown to block calcineurin activity in AtT20 cells and T lymphocytes cells with the same order of potency that they influenced cAMP accumulation (present study). Furthermore, L685,818 an analogue of FK506 (Dumont et al, (1992) J. Exp. Med. 176:751-760) that binds to the prolyl-isomerase FKBP-12 in a manner similar to FK506, but does not give rise to a drug protein complex that inhibits the activity of calcineurin, reversed the effects of FK506 on calcineurin activity, cAMP formation as well as ACTH release. Importantly, when given alone L685,818 had no discernible effect on cAMP formation or ACTH secretion further suggesting that the changes observed upon treatment with FK506 are due to the inhibition of calcineurin. Finally, neither FK506 nor cyclosporin A were effective in cells depleted of Ca.sup.2+.

Taken together, these characteristics justify the conclusion that the effects of immunosuppressants described here are attributable to the inhibition of calcineurin.

The production of cAMP in AtT20 cells is under inhibitory control by [Ca.sup.2+ ].sub.i. Stimulation with cAMP is known to elicit a rise of [Ca.sup.2+ ].sub.i in these cells which is largely derived from the extracellular pool by influx through dihydropyridine sensitive Ca.sup.2+ -channels (Luini et al, (1985) Proc. Natl. Acad. Sci. U.S.A. 82:8034-8038; Reisine et al, (1987) Mol. Pharmacol. 32:488-496 and Antoni et al, (1992) J. Endocrinol. 133:R13-R16), as intracellular pools of Ca.sup.2+ are sparse (Fiekers (1993) Abstracts of the 23rd Annual Meeting of the Society for Neuroscience 1186 Abst 488.3). Thus the [Ca.sup.2+ ].sub.i signal is a measure of electrical activity of the cell and in addition to triggering hormone release provides feedback inhibition to the chemical messenger system that generates it. In the case of CRF-induced cAMP formation this feedback is largely mediated by calcineurin.

Several possibilities have to be considered with respect to the site of action of Ca.sup.2+ /calcineurin in the signal transduction cascade.

An action of calcineurin at the receptor level is conceivable, however, the prevailing concept of G-protein coupled receptors (Sibley et al, (1987) Cell 48:913-922) dictates that receptor-down regulation or uncoupling is largely due to the action of protein kinases while protein phosphatases reverse this process. In contrast, the present data implicate calcineurin as an inhibitor of receptor stimulated cAMP production.

Dephosphorylation of the coupling protein Gs is also a possible site of regulation by calcineurin (Houslay (1994) GTPases in Biology B. F. Dickey and L. Birnbaumer, Springer Verlag, Berlin, Vol 108 Pt2, 147-165). Once more, current evidence in the literature associates protein phosphorylation with down-regulation of G-protein function and implicates protein phosphatases in the restoration of the cellular response (Pyne et al, (1992) Biochem. Biophys. Res. Commun. 186:1081-1086; and Strassheim et al, (1994) J. Biol. Chem. 269:14307-14313).

With respect to the effector adenylyl cyclase, these proteins have lately emerged as dynamic sites of signal integration (Taussig et al, (1995) J. Biol. Chem. 270:1-4). At least two types of cyclase, Types 5 & 6 (Iyengar (1993) Advances in Second Messenger and Phosphoprotein Research B. L. Brown and P. R. M. Dobson, Raven Press Ltd, New York, 28:27-36), are inhibited by Ca.sup.2+ but the mechanism of this effect has not been elucidated (Yoshimura et al, (1992) Proc. Natl. Acad. Sci. U.S.A. 89:6716-6720). The inhibition of Type 5 and 6 cyclase by Ca.sup.2+ is most marked after stimulation by agonists such as isoproterenol in chick heart cells (Yu et al, (1993) Mol. Pharmacol. 44:689-693), or VIP in GH.sub.4 C.sub.1 pituitary tumour cells (Boyajian et al, (1990) Cell Calcium 11:299-307), but much less prominent after activation with forskolin in GH.sub.4 C.sub.1 cells (Boyajian et al, (1990) Cell Calcium 11:299-307). Overall this is analogous to the observations made here, which in the first instance suggest a prominent action of calcineurin at or before the level of G-protein effector coupling. However, multiple types of adenylyl cyclase coexist in all cell types analyzed to date (Yoshimura et al, (1992) Proc. Natl. Acad. Sci. U.S.A. 89:6716-6720; DeBernardini et al, (1993) Biochem. J. 293:325-328 and Hellevuo et al, (1993) Biochem. Biophys. Res. Commun. 192:311-318) and this applies to AtT20 cells as RT-PCR analysis and sequencing of the amplified cDNAs clearly showed coexpression of at least three types of adenylyl cyclase mRNA (type 1 and 6 as well as a novel isotype). Forskolin appears to activate adenylyl cyclase isotypes by different efficacies and mechanisms (Iyengar (1993) Advances in Second Messenger and Phosphoprotein Research B. L. Brown and P. R. M. Dobson, Raven Press Ltd, New York, 28:27-36), hence the relative contribution of individual cyclase isotypes to forskolin-induced cAMP may be different from receptor activated cAMP formation, and as a result forskolin-induced cAMP synthesis and receptor activated synthesis may have different pharamcological profiles.

It is unlikely that type 1 cyclase is involved in the effects of Ca.sup.2+ and calcineurin reported here as it is invariably stimulated by Ca.sup.2+, whereas the predominant effect on CRF and .beta.-adrenergic stimulation was an inhibition by Ca.sup.2+. Type 6 adenylyl cyclase could be implicated because it is strongly inhibited by Ca.sup.2+, however, although the Type 6 isozyme is abundant in NCB20 (Yoshimura et al, (1992) Proc. Natl. Acad. Sci. U.S.A. 89:6716-6720) as well as HEK293 cells (Hellevuo et al, (1993) Biochem. Biophys. Res. Commun. 192:311-318) the stimulation of cAMP accumulation in these cells through endogenously expressed receptors for CRF was not altered upon blockage of calcineurin protein phosphatase activity. Importantly, the novel adenylyl cyclase homologue mRNA was undetectable in HEK293 cells and was found in very low amounts in NCB20 cells, whereas it appears to be the predominant adenylyl cyclase isotype mRNA in AtT20 cells. Work reported elsewhere (Paterson et al, (1995) Biochem. Biophys. Res. Commun. 214:1000-1008) reports isolation of a 4473 bp cDNA from AtT20 cells which contains the complete coding sequence of this novel mouse adenylyl cyclase.

Taken together, these data suggest that the effects of calcineurin inhibitors are associated with specific adenylyl cyclase isotype previously not characterized.

The potential functional significance of calcineurin negative feedback on cAMP formation in pituitary cells is highlighted by previous findings, showing that adrenal corticosteroid inhibition of CRF-induced ACTH release involves the de novo synthesis of calmodulin, the calcium sensor regulatory protein of calcineurin (Shipston et al, (1992) Biochem. Biophys. Res. Commun. 189:1382-1388). Furthermore, the efficiency of corticosteroid inhibition is reduced by approximately 10-fold by immunosuppressant blockers of calcineurin (Shipston et al, (1994) Ann. N. Y. Acad. Sci. 746:453-456; and Antoni et al, (1994) J. Physiol-London 475:137-138).

In summary, calcineurin is a Ca.sup.2+ -operated feedback inhibitor of cAMP production activated by CRF or .beta.-adrenergic receptors in pituitary cortictrope tumour cells. As [Ca.sup.2+ ].sub.i is largely derived through voltage-regulated Ca.sup.2+ channels in AtT20 cells, these cells exemplify a case where calcineurin functions as a link between the chemical and electrical signalling systems of the cell. These findings conform with previous reports (Armstrong (1989) Trends in Neurosciences 12:117-122; and Nichols et al, (1994) J. Biol. Chem. 269:23817-23823) suggesting that in excitable cells calcineurin is a fundamental negative feedback regulator of cellular responses as opposed to its role as a stimulatory element in non-excitable systems (Kincaid et al, (1993) Adv. Prot. Phosphatases 7:543-583). Furthermore, our findings extend the involvement of calcineurin from calcium channels and synaptic vesicle recycling to the adenylyl cyclase signalling pathway that is fundamental for intracellular signalling throughout the CNS.

Summary

The Ca.sup.2+ - and calmodulin-activated protein phosphatase calcineurin is highly abundant in the brain. While a physiological role for calcineurin has been delineated in T lymphocyte activation, little is known about its role in excitable cells. The present study investigated the effects of immunosuppressant blockers of calcineurin on agonist-induced cAMP formation and hormone release by mouse pituitary tumour (AtT20) cells. Inhibition of calcineurin with FK506 or cyclosporin A enhanced cAMP formation and adrenocorticotropin secretion induced by corticotropin releasing-factor (CRF). Further analysis of cAMP production revealed that intracellular Ca.sup.2+ derived through voltage-regulated calcium channels reduces cAMP formation induced by corticotropin releasing-factor or .beta..sub.2 -adrenergic stimulation, and that this effect of Ca.sup.2+ is mediated by calcineurin. Analysis of AtT20 cell RNA indicated the co-expression of at least three species of adenylyl cyclase mRNA encoding types 1 and 6 as well as a novel isotype, which appeared to be the predominant species. In two cell lines expressing very low or undetectable levels of the novel cyclase mRNA (NCB20 and HEK293 respectively), CRF-induced cAMP formation was not altered by FK506 or cyclosporin A. In summary, these data identify calcineurin as a Ca.sup.2+ sensor that mediates a negative feedback effect of intracellular Ca.sup.2+ on receptor-stimulated cAMP production thereby regulating the cellular response to cAMP generating agonists. Furthermore, the effect of calcineurin on cAMP synthesis appears to be associated with the expression of a novel adenylyl cyclase isoform, which is highly abundant in AtT20 cells.

EXAMPLE 2

DNA Cloning and Tissue Distribution of AC

Introduction

Adenylyl cyclases convert ATP to cAMP, one of the earliest recognized intracellular messenger molecules. The family of previously known adenylyl cyclases consists of eight members (Premont (1994) Meth. Enzymol. 238:116-127). Some of these enzymes have been analyzed functionally, and appear to confer unique signal processing capacities to cells (Taussig et al, (1995) J. Biol. Chem. 270:1-4). In particular isotype specific regulation of enzymatic activity by calmodulin, calcium (through as well as independently of calmodulin), protein kinase C as well as G-protein subunits has been demonstrated (for review see Taussig et al, (1995) J. Biol. Chem. 270:1-4). In addition to functional diversity adenylyl cyclase isotypes have distinct tissue distribution profiles (Krupinski et al, (1992) J. Biol. Chem. 267:24859-24862) and particularly marked regional differences of cyclase distribution have been observed in the brain (Xia et al, (1994) Mol. Brain Res. 22:236-244; and Glatt et al, (1993) Nature 361:536-538). Collectively, these observations indicate that the particular adenylyl cyclase isotype profile of a cell is fundamentally important with respect to cellular function.

Pharmacological analysis of the cAMP response to agonist stimulation in mouse corticotroph tumor (AtT20) cells (Antoni et al, (1994) Journal Of Physiology-London 475p:137-138) has shown that calcium inhibition of cAMP accumulation in these cells is mediated by the Ca.sup.2+ /calmodulin activated protein phosphatase calcineurin (protein phosphatase 2B). This novel feature of the cAMP response has led to the examination of the adenylyl cyclase isotype profile of AtT20 cells (Antoni et al, (1995) EMBO Journal submitted), which, amongst two known sequences (types 1 and 6), revealed the presence of a novel isotype. The present study reports the full cDNA sequence and tissue distribution of this novel adenylyl cyclase which is the predominant species of the enzyme expressed in AtT20 cells. The enzyme mRNA is also relatively abundant in the brain as well as certain peripheral endocrine organs, including the ovary and the adrenal gland.

Materials and Methods

Isolation of a cDNA Containing the Complete Coding Sequence of a Novel Mouse Adenylyl Cyclase Identified in AtT20 Cells

AtT20 cells were grown to sub-confluency in DMEM supplemented with 10% foetal calf serum (Woods et al, (1992) Endocrinology 131:2873-2880) and total RNA was isolated from approximately 10.sup.7 cells using Trizol reagent (GIBCO, Paisley, U.K.) according to the manufacturer's instructions. RT-PCR was carried out using an RNA PCR kit (Perkin Elmer, Warrington, Cheshire, U.K.) as previously described (Antoni et al, (1995) EMBO Journal submitted), using degenerate oligonucleotides corresponding to a highly conserved region within the second cytoplasmic domain of previously cloned mammalian adenylyl cyclases (Krupinski et al, (1992) J. Biol. Chem. 267:24859-24862; Yoshimura et al, (1992) Proc. Natl. Acad. Sci. U.S.A. 89:6716-6720; and Gao et al, (1991) Proc. Natl. Acad. Sci. U.S.A. 88:10178-10182) A 180 bp cDNA fragment of a novel adenylyl cyclase amplified from AtT20 cells (Antoni et al, (1995) EMBO Journal submitted) was used as a probe to obtain a cDNA containing the complete coding sequence by screening an oligo dT-primed cDNA library prepared from AtT20 cells and constructed in the vector ZAP II (a generous gift of Dr. M J Shipston, Edinburgh (see Shipston (1992) PhD Thesis, University of Edinburgh)) together with 5' RACE PCR. Screening of approximately 5.times.10.sup.5 clones was carried out according to standard procedures (Sambrook et al, (1989) Cold Spring Harbor Press, U.S.A.). Plaque-purified positive clones were recovered as pBluescript phagemids by excision using the helper phage ExAssist (Stratagene, Cambridge, U.K.). The insert size of these clones ranged from 1-3 kb. DNA sequence was determined from several independent isolates of each clone using Exo III/Mung Bean nuclease digestion and/or clone-specific primers with the Sequenase 2.0 kit (USB, Amersham International, Aylesbury, U.K.). Analysis of the DNA sequence data generated revealed the presence of an open frame which lacked a suitable initiator codon.

Two rounds of 5' RACE-PCR were performed to obtain the remaining of the open reading frame using a 5' RACE System kit (GIBCO, Paisley, U.K.) according to the manufacturer's instructions and summarised in brief as follows. Based on the DNA sequence at the 5' end of the largest cDNA clone isolated by library screening (jp134) a set of three nested anti-sense oligonucleotides were designed. The first of these (most proximal to the 3' end of the cDNA isolated; primer 1: CGTCAATGACCTCAAAGCC (SEQ ID NO: 9)) was used to prime synthesis of first strand cDNA by reverse transcription of 0.8 .mu.g of total RNA isolated from AtT20 cells in a 25 .mu.l reaction containing 20 mM Tris.HCl (pH 8.4), 50 mM KCl, 3 mM MgCl.sub.2, 10 mM DTT, 100 nM primer 1, 400 .mu.M each dNTP and 200U Superscript II Reverse Transcriptase at 42.degree. C. for 30 minutes. First strand cDNA was purified away from this primer using the GlassMax purification system (GIBCO) and one-fifth of the purified cDNA 3'-tailed with dCTP using 10U terminal deoxynucleotide transferase (TdT) in a 25 .mu.l reaction containing 20 mM Tris.HCl (pH 8.4), 50 mM KCl, 1.5 mM MgCl.sub.2, 200 .mu.M dCTP at 37.degree. C. for 15 minutes. Following heat inactivation of the TdT a second strand cDNA synthesis reaction was primed by the anchor primer supplied with the 5' RACE system kit present at 200 nM and carried out as for first strand synthesis. Double-stranded cDNA was subsequently purified using the GlassMax system and one-fifth of this preparation used as a template in PCR containing 20 mM Tris.HCl (pH 8.4), 50 mM KCl, 1.5 mM MgCl.sub.2, 200 mM each dNTP, 200nM each of universal amplification primer (UAP; supplied with the kit) and jp134-specific nested anti-sense primer 2 (SEQ ID NO: 10) (GCCTCTGCACAGCTGCAGTGGGACTCC) and 0.03U/.mu.l Amplitaq DNA polymerase (Perkin Elmer) (parameters for all the PCR reactions can be found in Table II). A second round of PCR using 0.05% of the primary PCR reaction (or 0.05% of size-selected primary PCR products) as a template was performed as described above except jp134-specific nested anti-sense primer 3 (SEQ ID NO: 11) (CCTGGCAGAACTGCTCGATGGCTTTTATCATGC) was substituted for primer 2. This first round of 5' RACE-PCR yielded a product of approximately 1 kb which was purified from an agarose gel and ligated into the plasmid vector pGEM-T (Promega, Madison, Wis., U.S.A.). Although DNA sequence analysis of this fragment (from both strands of two independent clones) extended the open reading frame observed in jp134 by 842 bp no suitable initiator codon was yet apparent. Another set of three anti-sense primers was designed on the basis of sequence data from the 5' end of the 1 kb fragment obtained from the first round and used in a second round of 5' RACE-PCR performed as described above except that 1) the first strand cDNA was synthesized at 50.degree. C. for 30 minutes and 2) substitution of primers 1, 2 and 3 with 4 (SEQ ID NO: 12) (GGAGAAGCTTCCTACTTG), 5 (SEQ ID NO: 13) (GTGGCCGTGAGAGTATGATTGGAGCTGTC) and 6 (SEQ ID NO: 14) (GTCCAAACCTGAAACTGCGCACGCAG), respectively. This yielded a product of approximately 650 bp which, when cloned into pGEM-T and several independent isolates sequenced, completed the open reading frame encoding the novel mouse adenylyl cyclase.

Detection of mRNA Expression

Northern analysis was performed using standard procedures. Briefly, 20 .mu.g of total RNA was separated by formaldehyde gel electrophoresis and transferred by blotting onto positively charged nylon membrane (Appligene, Ullkirch, France) then fixed by baking at 80.degree. C. and prehybridised at 42.degree. C. for 2 hours in 50% deionised formamide, 5.times.SSPE, 0.5.times.Denhardt's, 0.1% w/v SDS, 0.2 mg/ml denatured salmon sperm carrier DNA and 10% Dextran sulphate. Random-primed labelled DNA probe (50 ng; >10.sup.9 cpm/.mu.g) was then added and hybridisation continued overnight at 42.degree. C. The membrane was washed twice for 20 minutes in 2.times.SSC/0.1% SDS, followed by 20 minutes in 1.times.SSC/0.1% SDS at 42.degree. C. and finally 20 minutes in 0.5.times.SSC/0.1% SDS at 50.degree. C. before wrapping in plastic an exposing to autoradiographic film at -70.degree. C. Ribonuclease protection assays were performed using an RPA II kit (Ambion, AMS Biotechnology, Witney, Oxon, U.K.) according to the manufacturer's instructions. Briefly, 10 .mu.g of total RNA was hybridised overnight at 45.degree. C. to 10.sup.5 cpm of radiolabelled anti-sense riboprobe. Following hybridisation, reactions were digested with single-strand-specific RNase and protected fragments resolved on a 6% denaturing polyacrylamide gel which was fixed for 30 minutes in 15% methanol/5% acetic acid, dried and exposed to autoradiographic film at -70.degree. C.

Results and Discussion

Analysis of the cDNA and Amino Acid Sequence

The sequence of the full length cDNA isolated from AtT20 cells (sense strand on top) and the deduced primary structure of AC are shown in SEQ ID No 1. A 4473 bp cDNA containing a 4062 bp open reading frame which encodes a 1352 amino acid protein. A stretch of 10 amino acids near the 5' end of the cDNA strongly resembles the Kozak consensus sequence for initiation of translation and contains the presumed initiator codon of the novel adenylyl cyclase protein. The presence of a translational stop codon upstream of this methionine residue suggests that the complete open reading frame encoding this protein has been cloned. The deduced amino acid sequence between residues 1105 and 1192 appears to be identical (except for a single asparagine to aspartic acid switch at 1191) to a short sequence derived from mouse brain RNA reported in preliminary form (Premont (1994) Meth. Enzymol. 238:116-127) and designated as AC. Although this region of the enzyme contains at least two highly preserved stretches of amino acid sequence common to all cyclases, on the basis of the abundance of AC mRNA in the brain (see below) it seems plausible that the full sequence reported here corresponds to the same enzyme and will subsequently referred to as AC.

While the key adenylyl cyclase consensus sequences thought to be involved in catalytic activity (Taussig et al, (1995) J. Biol. Chem. 270:1-4) are well conserved in AC, overall amino acid sequence homology comparisons indicate that AC is not sufficiently similar to currently known cyclases to be placed within one of the previously suggested subfamilies as illustrated by FIG. 9. Therefore AC represents a sixth adenylyl cyclase subfamily. It cannot be excluded at present that perhaps other members of this subfamily exist.

Hydrophobicity plots (Kyte et al, (1982) J. Mol. Biol. 157:105) return the previously reported adenylyl cyclase structure (FIG. 10) a short N-terminal loop followed by two homologous segments each consisting of a hydrophobic chain with 6 intermembrane helices and a large hydrophilic (putative cytoplasmic) loop. The first cytoplasmic loop appears considerably longer (by about 130 amino acids) in AC than previously reported for other isotypes.

Analysis of putative phosphorylation sites (Pearson et al, (1991) Meth. Enzymol. 200:61-81) indicates several strong consensus protein kinase A and protein kinase C sites as well as two casein kinase phosphorylation sites (Table 3). Elements of the functionally relevant calmodulin binding site identified in ACtypel (Wu et al, (1993) J. Biol. Chem. 286:23766-23768 and Vorherr et al, (1993) Biochemistry (U.S.A.) 32:6081-6088) can be discerned on the basis of conserved amino acid residues at positions 504-505 in the first cytoplasmic domain. However, two important basic residues Lys500and Lys497 which are part of the calmodulin binding site in ACtypel are substituted by Gly in AC making calmodulin binding unlikely in this region of the enzyme.

Distribution of mRNA for AC in Rat Tissues and Mouse Brain

Ribonuclease protection analysis of rat tissues showed hybridizion of total RNA with probe JP114 gave a 125 bp double stranded product that was most abundant in brain, with significant hybridizing activity in the anterior pituitary gland, the ovary, the adrenal gland, the lung and the kidney. Liver, pancreas,spleen, thymus, heart gave no detectable signal, suggesting that the distribution of the enzyme is highly tissue specific.

Gross dissection of the mouse brain suggests that AC mRNA is most abundant in the cortex, striatum and the hippocampus, with lower levels in the cerebellum and much lower but detectable concentrations in the olfactory bulb, the diencephalon, the brain stem and the pituitary gland. It is of note that the RNAse resistant hybrid in mouse tissues and AtT20 cells is larger, about 155 bp, than in rat tissues indicating interspecies sequence variation(s) in this region of the enzyme.

Northern mRNA analysis showed an approximately 9 kb size RNA in AtT20 cells that hybridizes with a JP114 antisense .sup.32 P-DNA probe.

In summary, the present results show the existence of a further member of the adenylyl cyclase family of proteins, which has a restricted tissue distribution and a distinct regional pattern in the brain. The functional properties of this cyclase have yet to be explored in detail. Pharmacological analysis (Antoni et al, (1994) Journal Of Physiology-London 475p:137-138 and Antoni et al, (1995) EMBO Journal submitted) indicates that AC is inhibited by Ca.sup.2+ through calcineurin and that this may be relevant for corticosteroid inhibition of corticotropin secretion in anterior pituitary cells (Shipston et al, (1994) Ann. N.Y. Acad. Sci. 746:453-456). Furthermore, the abundance of AC mRNA in the striatum and hippocampus, where calcineurin is particularly abundant relative to other regions of the brain (Kuno et al, (1992) J. Neurochem. 58:1643-1651) is of interest with respect to synaptic function.

Summary

The cDNA of a novel adenylyl cyclase isotype (AC) has been cloned and sequenced from the mouse pituitary corticotrope tumour cell line AtT20. Adenylyl cyclase cDNA sequences were amplified from AtT20 cells using degenerate primers bracketing a highly conserved motif near the carboxyl terminus region of the protein. The majority of the sublcloned amplified cDNA products revealed a novel sequence which was utilized to screen an AtT20 cell cDNA library, from which a 3120 bp clone was isolated and sequenced. The full length coding sequence was obtained upon two rounds of 5' RACE-PCR, yielding a 4473 bp long cDNA containing an open reading frame of 4062 bp which encodes a protein of 1352 amino acids. The hydrophobicity profile of this protein resembles that of other members of the adenylyl cyclase family in that two sets of six hydrophobic, putatively membrane-spanning regions are predicted as well as a large central cytoplasmic loop and long C-terminal cytoplasmic tail. Amino acid sequence comparisons suggest that this novel enzyme is quite distinct from other known mammalian adenylyl cyclases and cannot be easily assigned to any of the previously observed subfamilies. Tissue distribution of mRNA was examined by RNAse protection assay and indicated the highest abundance of this novel adenylyl cyclase to be in the brain followed by anterior pituitary gland, ovary and adrenal gland, which appeared to express approximately equal levels. Low level expression of this mRNA was detected in lung and kidney while in the heart, liver and pancreas none was apparent. Within the brain, relatively high levels of novel adenylyl cyclase mRNA were detected in cortex, hippocampus, striatum, cerebellum, with much less in diencephalon, olfactory bulb and the brain stem.

EXAMPLE 3

Functional Properties of AC

Method

The complete coding sequence of AC was subcloned into the eukaryotic expression vector pcDNA3 (Invitrogen). The cDNA clone for the CRF.sub.1 receptor cloned into the vector pcDNAI (Chang et al, (1993) Neuron 11:1187-1195) was a gift of Drs. R. V. Pearse II and M. G. Rosenfeld (University of California, San Diego). Both expression plasmids (10 .mu.g) were transfected into SV40-transformed monkey embryonic kidney (COS-7) cells grown in DMEM and 10% fetal bovine serum to 70% confluency in 75 cm.sup.2 flasks following the double DEAE-dextran transfection protocol (Ishikawa et al, (1992) Nucl. Acids Res. 20:4367), while only AC cDNA was transfected into adenovirus-transformed human embryonic kidney (HEK293) cells as these express endogenous receptors for CRF (F. A. Antoni, unpublished data). In control transfections pcDNA3 vector DNA replaced the AC expression construct. Forty-eight hours after the second transfection the cells were harvested in Hank's Ca.sup.2+ and Mg.sup.2+ free balanced salt solution containing 0.1% EDTA and centrifuged at 200.times.g, for 10 minutes. Following resuspension the cells were centrifuged again to remove serum proteins and resuspended in 1 ml of HEPES (25 mmol/l, pH 7.4) buffered Hank's balanced salt solution and a 50 .mu.l aliquot was removed for the measurement of protein content. The cells were then diluted further with 4ml of HEPES buffered Hank's solution containing 0.25% BSA, and preincubated for 1 hour at 37.degree. C. under air. Subsequently the cells were pelleted again and distributed (final concentration 3.times.10.sup.5 cells/ml) into polypropylene vials and processed as AtT20 cells except that 1 mmol IBMX and 0.1 mmol/l rolipram were used as inhibitors of cAMP-degrading phosphodiesterases (see Example 1).

The functional properties of the cloned AC10 were investigated in transiently transfected HEK293 and COS7 cells (FIG. 11). In HEK293 cells which have endogenous receptors for CRF, unstimulated levels of cAMP and the cAMP response to CRF were unchanged upon transfection with AC (FIGS. 11a and b). However, preincubation with FK506 increased unstimulated cAMP levels 3-fold and significantly enhanced the cAMP response to 1 but not to 5 (FIG. 11b) or 25 nmol/l CRF (not shown). These effects were specific to cells transfected with AC (FIGS. 11a and b). Surprisingly, in cells transfected with AC 100 .mu.mol/l L-685,818 increased unstimulated cAMP levels to the same extent as 2 .mu.mol/l FK506, while having no effect in sham-transfected or pcDNA3-transfected cells (not shown). Importantly, L-685,818 had no significant effect on CRF-induced cAMP production but blocked the enhancement of the response to 1 nmol/l CRF by FK506 in cells transfected with AC (FIG. 11c). Slightly different results were obtained in COS7 cells. In this system unstimulated cAMP production in cells transfected with AC cDNA (without or in combination with the CRF receptor) was 1.9-fold higher than in controls receiving vector DNA (unstimulated cAMP [pmol/mg protein] in COS7 cells transfected with CRF receptor and pcDNA3:20.9.+-.0.2; with CRF receptor and AC cDNA: 38.2.+-.3.4; mean.+-.SEM, n=4/group P<0.001, Student's t-test, representative of 4 separate experiments) whereas the increment produced by CRF (1-25 nmol/l) was not consistently altered (FIG. 11d and data not shown). Preincubation with cyclosporin A significantly and selectively enhanced the cAMP response to 1 nmol/l CRF in cells transfected with AC (FIG. 11d).

These data confirm that the cDNA for AC encodes a cAMP generating enzyme. At this point the cause of the differences in unstimulated cAMP production between COS7 cells and HEK293 cells is unknown, however, such discrepancies are not unprecedented in eukaryotic expression systems (Premont (1994) Meth. Enzymol. 238:116-127; and Adie et al, (1994) Biochem. J. 303:803-808). With respect to the paradoxical effect of L-685,818 on unstimulated cAMP levels, it is likely that it has some calcineurin inhibiting activity at a high concentration (100 .mu.mol/l) which may be sufficient to enhance basal cAMP production. However, this small inhibitory effect is not enough to cause a functionally relevant reduction of activity when calcineurin is activated by the cAMP-induced increase of intracellular free Ca.sup.2+ that is known to occur upon a cAMP stimulation in both HEK293 and COS7 (Lin et al, (1995) Mol. Pharmacol. 47:131-139; and Widman et al, (1994) Mol. Pharmacol. 45:1029-1035 (1994)) cells. The consistent finding of the transient transfection experiments is that cAMP production in response to low CRF concentrations is enhanced by immunosuppressant blockers of calcineurin and, in the case of FK506, this is blocked by the antagonist L-685,818. In this respect the findings with cloned and transiently transfected AC are in good agreement with the findings in AtT20 cells and with pilot studies in HEK293 cells stably transfected with AC (Antoni et al, unpublished data).

Sequence Comparison

Method

Comparisons with other adenylyl cylases were made using the GCG package available from the Human Genome Mapping Programme, Cambridge U.K. Initial alignments with FKBP12 were made using GeneJockeyII (Biosoft, Cambridge U.K.), and adjusted visually by the aid of the programme.

Closer examination of the primary sequence of AC revealed that residues 503 to 610 in the first cytoplasmic loop show approximately 40% similarity with FKBP12 (FIGS. 12a and b). The first part of this FKBP12-like segment consists of the C-terminal segment of the C1.alpha. domain which is highly conserved in all adenylyl cyclases (residues 503-570 in FIG. 12c) (Taussig et al, (1995) J. Biol. Chem. 270:1-4). The second part is the beginning of the long non-conserved segment (C1.beta. domain). Importantly, some of the sequence identities correspond to Asp37, Gly86, Phe87, Ile90 in FKBP12 (FIG. 12b) and are amino acid residues of key importance for the high-affinity interaction of the FK506/FKBP12 complex with calcineurin (Aldape et al, (1992) J. Biol. Chem. 267:16029-16032; Yang et al, (1993) J. Am. Chem. Soc. 115:819-820; and Braun et al, (1995) FASEB J. 9:63-72). The best overall sequence similarity is observed with yeast FKBP12, which in fact has been shown to bind calcineurin in the absence of FK506 (Cardenas et al, (1994) EMBO J. 13:5944-5957). Given the prominent actions of calcineurin blocker immunosuppressants on cAMP formation in systems expressing AC it seems reasonable to suggest that AC.sub.(503-610) may be a physiologically relevant docking site for calcineurin.

Adenylyl cyclase type 1 (ACtp1) is also thought to interact directly with a protein co-factor distinct from G-protein subunits as this cyclase is markedly stimulated by Ca.sup.2+ /calmodulin (for review see Taussig et al, (1995) J. Biol. Chem. 270:1-4). Residues 495-522 of ACtp1 (495-518 shown underlined in FIG. 12c) bind calmodulin with nanomolar affinity (Vorherr et al, (1993) Biochemistry (U.S.A.) 32:6081-6088). Moreover, point mutations of Phe.sub.503 and Lys.sub.504 virtually abolish the stimulatory effect of Ca.sup.2+ /calmodulin on this cyclase (Wu et al, (1993) J. Biol. Chem. 286:23766-23768). Intriguingly, the position of residues 495-522 along the cytoplasmic loop of AC corresponds to the C-terminal part of the FKBP12-like sequence in AC (FIG. 12c). It seems therefore that the junction between the C1.alpha. and C1.beta. domains in adenylyl cyclase 1 and in AC is a site of allosteric regulation by calcium binding proteins. In the case of AC a strong candidate for such a regulatory protein is calcineurin.

AC is Enriched in Nerve Cells

Method

Northern analysis of total RNA (20 .mu.g per lane) was performed by transfer onto positively charged nylon membrane (Appligene, Ullkirch, France) then hybridized to radio-labelled probe in 50% formamide and washed according to standard procedures (Sambrook et al, Molecular Cloning: a Laboratory Manual Ed. 2 (Cold Spring Harbor Press, U.S.A., 1989)). The cDNA probes used correspond to amino acid residues 1-195 (JP164) and 1105-1165 (JP114).

Ribonuclease protection assays were performed using an RPA II kit (Ambion, AMS Biotechnology, Witney, Oxon, U.K.) according to the manufacturer's instructions using radio-labelled anti-sense riboprobe derived from JP114 or JP142 (corresponding to amino acids 320-478). The AC probes showed no significant cross hybridisation in Southern analysis with adenylyl cyclase isotype cDNAs 1, 2, 3, 5 or 6. A 250 bp probe hybridising with .beta.-actin was used as the internal standard to correct for differences in RNA loading. Blots and gels were exposed to X-ray film as well as to Molecular Dynamics Phosphorimager casettes and quantified with the ImageQuant software using the 28S RNA band and the .beta.-actin band as standards for RNA loading in Northern and RNase protection analysis, respectively. Division of the integrated volume of pixels of the selected radiolabelled band with the integrated volume of the internal standard band yields the relative hybridisation intensity, which was used to compare the intensity of labelled RNA bands within blots. The relative abundance of the protected RNA species in a given tissue is expressed as the percentage of the relative hybridisation intensity of the whole brain RNA band run in the same assay.

Importantly, we have discovered that AC mRNA is present in relatively high levels in the hypothalamic paraventricular nucleus that contains the major contingent of CRF41 and vasopressin producing neuroendocrine motoneurones and controls the secretion of ACTH by the anterior pituitary gland. In addition, the levels of mRNA are apparently controlled by adrenal corticosteroids, so that lowering of adrenocortical steroid levels causes an increased expression of AC mRNA. This reinforces the notion that AC is expressed in areas of the brain important for the negative feedback action of adrenal corticosteroids, in the hypothalamic-pituitary-adrenocortical axis. This system is a servomechanism where, when corticosteroid feedback is detected as inadequate by the relevant as yet unidentified components of the system, the levels of AC mRNA are increased.

Results

Existence of a calcineurin-regulated adenylyl cyclase is of great potential significance for brain function as calcineurin constitutes about 1% of total brain protein (Klee et al, (1988) Adv. Enzymol. 61:149-200). AC is in fact highly abundant in the brain: Northern blot analysis shows the presence of a 9 kb mRNA in mouse striatum and AtT20 cells (FIG. 13). RNAse protection anaylsis of dissected regions of the mouse brain suggest the relative abundance of AC mRNA to be hippocampus.gtoreq.striatum.gtoreq.cortex.gtoreq.cerebellum.gtoreq.olfacto ry bulb>brain stem>diencephalon.gtoreq.anterior pituitary gland (range 3-fold, see FIG. 13 for methods). These findings indicated a primarily neuronal localisation which was confirmed by in situ hybridisation histochemistry, showing robust AC mRNA hybridisation in hippocampal neurons (FIG. 14).

RNAse protection assay also shows the presence of AC10 mRNA, amongst other tissues, in mouse kidney, thymus and spleen (relative abundance 50, 10 and 8%, respectively if brain=100%, see FIG. 13 for methods) all of which are important targets of immunosuppressants (Schreier et al, (1993) Transplant. Proc. 25:502-507; and Dumont et al, (1992) J. Exp. Med. 176:751-760).

Functional Implications of Calcineurin Regulated Adenylyl Cyclase

In summary, the present results show the existence of a new member of the adenylyl cyclase family of proteins that is abundant in the brain and is inhibited by calcineurin. The intricate control of calcineurin activity by its EF-hand protein .beta.-subunit as well as additional regulation by calmodulin (Schreier et al, (1993) Transplant. Proc. 25:502-507) suggest a potentially fundamental role for AC in situations where fine interplay between intracellular Ca.sup.2+ and cAMP determines cellular function (Cooper et al, (1995) Nature 374:421-424). Furthermore, the abundance of AC mRNA in the striatum and hippocampus, where calcineurin is particularly enriched relative to other regions of the brain (Kuno et al, (1992) J. Neurochem. 58:1643-1651) is of major interest with respect to synaptic function. Finally, it remains to be explored whether or not the currently known Ca.sup.2+ regulated cyclases are also subject to control by calcineurin. A recent study has in fact reported the facilitation of cAMP formation by Ca.sup.2+ /calcineurin suggestive of calcineurin control of a Ca.sup.2+ -stimulated adenylyl cyclase (Baukal et al, (1994) J. Biol. Chem. 269:24546-24549).

EXAMPLE 4

Human AC9

Isolation of a Complete Human AC9 cDNA Clone

The assembly of a full length human AC9 cDNA clone is represented in FIG. 18. The NCBI BLAST computer program (http://www.ncbi.nlm.nih.gov/BLAST/) was used to search the GenBank/EMBL and dbEST (expressed sequence tag) DNA sequence databases for human EST clones exhibiting similarity to mouse AC9 cDNA sequence Z50190. Two entries of interest were initially identified under GenBank/EMBL Accession Numbers R88632 and T08314. The EST cDNA clones, IMAGE 166657 (from adult human brain) and HIBBC33 (from infant human brain), from which these sequences were derived, were respectively obtained from the MRC Human Genome Mapping Project Resource Centre (MRC HGMP-RC, Cambridge, UK) and The Institute for Genome Research (TIGR, Gathersburg, Md., U.S.A., distributed by ATCC, Rockville, Md., U.S.A.). Due to the directed nature of cDNA cloning (Adams et al, (1993) Nat. Genet. 4: 373-80) the database sequence entries for these clones was presumed to represent part of the sense strand of the cDNA. This holds true for sequence R88632 (258 bp) which exhibits 77% sequence identity with nucleotides 3855 to 4113 of mouse AC9 sequence Z50190. However, sequence T08314 (228 bp) exhibited 82% identity with residues 682-909 on the anti-sense strand of mouse AC9 sequence Z50190 suggesting that this cDNA was cloned randomly rather than directionally. This insert may have arisen due to fragmentation of cDNA prior to vector ligation or as a result of mis-priming of cDNA synthesis from a source other than the polyadenylated 3' end of mRNA although a complete sequence representing the primer of cDNA synthesis (Adams et al, (1993) supra) was not found. The inserts of both EST clones were completely sequenced by the dideoxynucleotide method using universal and sequence-specific primers using an ABI Prism dye terminator DNA cycle sequencing kit (Perkin Elmer), resolved on a 4.75% acrylamide denaturing gel and data collected on an Applied Biosystems/Perkin Elmer 373 Stretch automated sequencer. Sequences were analyzed independently using Gene Jockey II program (Biosoft, Cambridge, U.K.) for the assessment of errors and visual confirmation of base calling. IMAGE clone 166657 was found to contain the 3' part of an open reading frame (ORF) encoding human AC9 followed by an untranslated region (UTR) and polyadenylated tail. Clone HIBBC33 contained 5' UTR sequence and the 5' part of the ORF encoding human AC9. To complete cloning of the human AC9 cDNA, the gap between these EST cDNA clones containing the 5' and 3' parts of the ORF were isolated by polymerase chain reaction (PCR) amplification using the High Fidelity Expand PCR system (Boehringer Mannheim/Roche) with sequence-specific primers. A PCR product of 2860 bp (1309-4169; pJP230) was amplified from a denatured human skeletal muscle cDNA library (Clontech) using primers 5'-GAC CTT TGG CTA CCA TTT CCG GGA TG-3' (SEQ ID NO: 100) and 5'-CTT CGC TCA CCT GGA TGC GGC ACT C-3' (SEQ ID NO: 101) respectively. The ends of this PCR product were blunt ended with T4 DNA polymerase (Boehringer Mannheim/Roche) prior to ligation into the SmaI site of plasmid vector pUC9 and subsequently sequenced as above. Additional overlapping clones of PCR products derived from human AC9 cDNA (853-1196; pJP 192 and 853-1167; pJP195), amplified as above or from a pool of denatured human cDNA libraries (Clontech) using primers based on mouse AC9 cDNA (upstream primer 5'-GGA GCG CTG CTT TCC GCA G-3' (SEQ ID NO: 102) with downstream primers 5'-CAT GTT CAC CAT CTC TTT CTT CTC C-3' (SEQ ID NO: 103) and 5'-GTG GCC GTG AGA GTA TGA TTG GAG CTG TC-3' (SEQ ID NO: 104) respectively), were propagated in the TA cloning vector pCRII (Invitrogen) and sequenced as above. The complete human AC9 cDNA (5515 bp, including 5' and 3' Not I cloning sites) was assembled from the above clones using recombinant DNA techniques. Internal PCR products pJP192 and pJP230 were joined at EcoRV site; 1735. The resulting clone was joined to the 5'UTR, initiation codon and 5' part of the ORF in clone HIBBC33 at MluI site; 1053. Finally the insert from clone IMAGE 166657 containing the 3' ORF and UTR was joined to the above clone at NcoI site; 4042 and the complete cDNA cloned as a NotI fragment into pBluescript (Stratagene) and the expression vector pcDNA3 (Invitrogen) for use in subsequent experiments.

Functional Expression of Human AC9

The complete coding sequence of human AC9 was cloned into the expression vector pcDNA3. This constuct and pcDNA3 (10 .mu.g) were each linearised by PvuI digestion prior to electroporation into HEK293 cells. Stably transfected clones were selected for resistance to G418 (Gibco) at 0.6 .mu.g/ml in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% foetal calf serum. A total of 12 clones were expanded and over-expression of human AC9 confirmed in 9 cell lines by immunoblotting using antisera raised against a pentadecapeptide derived from the N-terminus of mouse AC9 (Paterson et al, (1995) Biochem. Biophys. Res. Commun. 214:1000-8). All but one of these 9 clones showed the inhibitory effects of Ca.sup.2+ on cAMP accumulation assayed in the presence of cyclic nucleotide phosphdiesterase (PDE) blockers (Antoni et al, (1998) J. Neurosci. 18:9650-61). Analysis of clone 7 is reported here.

HEK-293 cells display complex intracellular Ca.sup.2+ handling mechanisms that may be controlled by cAMP and calcineurin (Lin et al, (1995) Mol. Pharmacol. 47:131-9; Gromada et al, (1995) FEBS Letts. 373:182-6; Querfurth et al, (1998) Biochem J. 334:79-86), and thus complicate the interpretation of the data. Therefore, intracellular stores of Ca.sup.2+ were depleted to reduce the contribution of intracellular Ca.sup.2+ pools to the intracellular free Ca.sup.2+ signal. All procedures were carried out at 37.degree. C., where required, immunosuppressant drugs FK506 (courtesy of Fujisawa GmbH, Munich, Germany) cyclosporin A (courtesy of Novartis Ltd, U.K.) and L685,818 (Courtesy of Merck & Co, Rahway, N.J.) rapamycin (courtesy of Dr Hans Fliri, Sandoz Pharma, Basel, Switzerland) or ethanol vehicle (0.1% v/v) were added to the medium at the onset of the Ca.sup.2+ depletion period. Two different Ca.sup.2+ depletion protocols were used (Antoni et al, (1998) supra). In the first, cells were incubated in a balanced salt solution containing (mM) 133 NaCl, 5.4 KCl, 0.25 Na.sub.2 HPO.sub.4, 0.44 KH.sub.2 PO.sub.4, 1 MgSO4, 2 EGTA, 5.6 D-glucose, 25 Hepes pH7.40 and 0.1% (w/v) bovine serum albumin (HBSS/EGTA) with 10 .mu.M ryanodine and 1 .mu.M thapsigargin (both from LC Labs, Boston, Mass.). These conditions deplete rapidly exchangeable Ca.sup.2+ pools and activate capacitative Ca.sup.2+ entry mechanisms (Cooper et al, (1994) Cellular Signalling 6:823-40). In the second protocol 5 .mu.M 4-Br-A23187 was used instead of ryanodine and thapsigargin, and Ca/EGTA was added to obtain similar initial levels of [Ca.sup.2+ ].sub.i as in the studies with ryanodine and thapsigargin. Under these conditions intracellular Ca.sup.2+ pools are depleted and, in addition to capacitative Ca.sup.2+ entry, Ca.sup.2+ influx may take place through the pores formed by the ionophore. The cells were incubated in Ca.sup.2+ -depleting medium for 20 min. Subsequently HBSS/EGTA containing various amounts of CaCl.sub.2 (Ca/EGTA) was added to achieve extracellular concentrations of Ca.sup.2+ ranging between 0.125-2 mM. The pH of this solution was 7.65 in order to minimize the change of pH due to the displacement of protons from EGTA--a pH shift from 7.40 to 7.29 occurred upon achieving a free Ca.sup.2+ concentration of 2 mM. After Ca/EGTA was added, the cells were incubated for 5 min after which the cyclic nucleotide PDE inhibitors isobutylmethylxanthine (IBMX, Aldrich, Gillingham, Dorset, U.K.) and rolipram (courtesy of Dr H Wachtel, Schering AG, Berlin Germany), 1 mM and 0.1 mM respectively, were introduced. Unless indicated otherwise, the cells were incubated for a further 10 min and the incubation was then terminated by the addition of 0.2M HCl. The total cAMP content of the cells and the medium was determined after freeze-thawing and acetylation by radioimmunoassay (Antoni et al, (1995) J. Biol. Chem. 270:28055-61).

Analysis of Human AC9 mRNA Expression Northern Blotting

A multiple tissue northern blot (MTN; Clontech) of poly A.sup.+ RNA from human heart, brain, placenta, lung, liver, skeletal muscle, kidney and pancreas (2 .mu.g per lane) was probed with a 528 bp SmaI-MluI fragment from the 5' end of the human AC9 cDNA and subsequently with a human .beta.-actin cDNA control probe supplied with the blot (each labelled with 32P-dCTP by random priming; Pharmacia "Ready-to-Go" system). Hybridization was carried out in Express Hyb solution (Clontech) according to the manufacturers instructions at 68.degree. C. for 1 hour, washed in 2.times.SSC/0.1% SDS at room temperature followed by 0.1.times.SSC/0.1% SDS at 50.degree. C. then wrapped in plastic and exposed to Fuji X-ray film for autoradiography. Following removal of previous probes the filter was hybridised as above with a cDNA fragment derived by PCR amplification from a human skeletal muscle cDNA library (Clontech) corresponding to nucleotides 3816-4487 of clone KIAA0520 (GenBank/EMBL Accession Number AB011092) (26).

Ribonuclease protection assays were performed using a .sup.32 P-CTP labelled riboprobe synthesised by in vitro transcription (T7 RNA polymerase Riboprobe System, Promega) from a HindIII-linearised plasmid (pJP195; FIG. 1) partial human AC9 cDNA clone representing nucleotides 853-1167 according to the kit manufacturer (Ambion). Probe fragments protected from nuclease digestion following incubation with 5 .mu.g total RNA samples from normal adult human brain obtained 4 h after death (JEB; Custodian, MRC Brain Bank, Western General Hospital, Edinburgh) or HEK293 cells (control) were resolved on a 6% denaturing acrylamide gel (8M Urea) and visualised by autoradiography.

In situ hybridisation was performed on 15 .mu.m frozen sections from normal post mortem adult human brain obtained from the MRC Alzheimer Disease Brain Bank, Western General Hospital, Edinburgh--Professor Jeane E Bell, Custodian. Prepared sections were courtesy of Dr Celia Yates and Professor George Fink. .sup.35 S-labelled riboprobes were prepared by in vitro transcription from the human AC9 partial cDNA clone pJP195 (FIG. 18) using T7 (anti-sense) or SP6 (sense) RNA polymerase and used as previously described for mouse AC9 (Paterson et al, (1995) Biochem. Biophys. Res. Commun. 214:1000-8; Rosie et al, (1990) J. Endocrinol. 124:285-9). Following hybridisation, brain sections were exposed to .beta.-max autoradiogaphy film (Amersham-Pharmacia) for 2 weeks. The autoradiogram was developed and photographed under white light illumination.

Immunocytochemistry

HEK293 cells were plated on poly-D-lysine-coated coverslips and cultured as described above. Two to four days after plating, cells were washed in phosphate buffered saline (PBS) and fixed in 3% paraformaldehyde in PBS, pH 7.4, for 20 min at 4.degree. C., treated with chilled (-20.degree. C.) absolute methanol for 10 min (4.degree. C.) and washed in 0.3 M glycine (3 times for 5 min) and 0.2% Triton X-100 (15 min) both in PBS at room temperature. Fixed cells were incubated for 18-24 h at 4.degree. C. with affinity purified chick antibodies (1:400-1:500) raised against the C-terminus of mouse AC9 (Antoni et al, (1998) J. Neurosci. 18:9650-61). FITC-conjugated anti-chicken IgG (Jackson ImmunoRes.Lab.) was applied at 1:400 dilution for 1 hr at RT. After several washes in PBS specimens were mounted in CitiFluor (Sigma) and examined using a Leica TCS-NT confocal microscope. Specificity controls involved preincubation of the antibody with 10 .mu.M antigen overnight at 4.degree. C., after which immunocytochemistry was carried out as above.

Detection of Human AC9 by Immunoblotting

Membrane proteins were prepared from HEK293-derived cell lines as described previously (Antoni et al, (1998) supra). Approximately 10 .mu.g membrane protein per lane were separated by SDS-PAGE and transferred by electroblotting on to PVDF membrane (Bio-Rad) at 220V in tank buffer (25 mMTris, 192 mM glycine, 0.036% SDS, 20% methanol) for 3 hrs at room temperature. PVDF membranes were blocked at room temperature by shaking in PBS containing 5% (w/v) skimmed milk powder and 0.1% Tween 20 for 2 hrs at room temperature, then washed 5.times.5 mins in PBS/0.1% Tween 20 before shaking incubation in PBS/0.1% Tween 20 containing primary antibody (rabbit anti-mouse AC9 N-terminus at 1:4,000 dilution or chicken anti-mouse AC9 C-terminus at 1:10,000 dilution) for 1 hr at room temperature. PVDF membranes were then washed as before and incubated in PBS/0.1% Tween 20 containing sheep anti-rabbit IgG-Horseradish peroxidase (HRP) conjugate (kindly provided by the Scottish Antibody Production Unit) at 1:8,000 dilution or goat anti-chicken IgY-HRP conjugate (Promega) at 1:10,000 dilution respectively) for 1 hour at room temperature then finally washed as above, exposed to reagents for development of chemiluminescence and visualised by exposure to Hyperfilm (ECL system, Amersham-Pharmacia, Aylesbury, UK). To confirm specificity, primary antisera were pre-incubated with mouse AC9-derived pentadeca-peptide antigens at final concentrations 1 .mu.g/ml (C-terminal) and 0.1 .mu.g/ml (N-terminal) for 20 & 45 mins respectively prior to probing of the PVDF membranes.

Results

Cloning and Sequencing of Human AC9 cDNA

Searching of human EST databases for similarity to mouse AC9 identifed two cDNA clones: IMAGE clone 166657 and HIBBC33, which were obtained as described in materials and methods. Oligonucleotide primers based on the DNA sequence of the inserts of these were used to construct a complete cDNA by PCR amplification from human cDNA libraries as described in materials and methods and outlined in FIG. 18. In the cDNA we describe (GenBank/EMBL Accession Number AJ133123) a short open reading frame (ORF) of 72 bp exists near the 5' end (263-334). However, the initiation codon of this ORF does not form part of a Kozak consensus sequence suggesting that translation may not be initiated although this remains to be confirmed. Downstream, a large ORF of 4059 bp is apparent, from 539-4597. This encodes a putative protein of 1353 amino acids, identical in length to the mouse AC9 protein. The deduced amino acid sequence of hAC9 reported here exhibits 92% sequence identity with the mouse AC9 protein and 86.7% identity at the nucleotide level. Structural features predicted from the amino acid sequence using the computer program TopPred 2 (von Heijne (1992) J. Mol. Biol. 225:487-494) are typical of membrane-associated mammalian adenylyl cyclases cloned to date, namely, two sets of six transmembrane regions (M1 and M2), each followed by a large cytoplasmic region (C1 and C2 respectively). These domains are conserved between the human and mouse AC9 proteins along with many motifs for potential post-translational modifications. With N and C termini localised intracellularly, the predicted topology of membrane-bound AC9 indicates conserved sites which may be suitable for modification include those for N-linked glycosylation, located extracellularly (residues 206, 955, 964), as well as intracellular sites suitable for tyrosine phosphorylation (residues 543, 1004, 1225) and serine/threonine phosphorylation by cAMP-dependent protein kinase (residues 373, 374), calmodulin-dependent protein kinase II (residues 1217), casein kinase II (residues 433, 582, 589, 691, 766, 1200) and protein kinase C (residues 27, 87, 112, 357, 365, 535, 620, 661, 762, 838, 1003, 1023, 1179, 1212, 1307).

Expression of Human AC9 Protein Detected by Immunocytochemical Analysis

Immunocytochemical analysis using antisera raised against the C-terminus of the mouse AC9 protein revealed fluorescent labelling in HEK 293 cells stably transfected with cDNA encoding human AC9 (FIGS. 19a & b). No labelling was observed when the antiserum was pre-incubated with peptide antigen (FIG. 19c) indicating specific detection of human AC9 protein. A large majority of human AC9 expressed in HEK 293 cells appeared to be localised to the plasma membrane.

Detection of Multiple Human AC9 Polypeptides by Immunoblotting

Immunoblotting of extracts from pcDNA3 and hAC9 stably transfected HEK293 cells resolved by SDS-polyacrylamide gel electrophoresis (PAGE) showed that hAC9 protein was expressed and specifically recognised by polyclonal antisera raised against peptides derived from the N- or the C-termini of mouse AC9 (FIG. 20). In each case, a broad immunoreactive band migrating between 160-180 kDa was detected, most probably attributable to post-translational modification by phosphorylation and/or N-glycosylation of the human AC9 protein in HEK 293. These observations confirm the predicted conservation of the sequences of the human and mouse AC9 at the N- and C-termini.

Ca.sup.2+ /Calcineurin Inhibition of cAMP Production by Human AC9

In the presence of phosphodiesterase inhibitors, the cAMP content of stably transfected HEK293 cell lines expressing human AC9 increased with time and this was reduced by the presence of Ca.sup.2+ in the medium (FIG. 21a). No statistically significant increase of cAMP was observed in wild type HEK293 cells or HEK293 cell lines transfected with pcDNA3 vector alone. Similar data were obtained in cells treated in 5 .mu.M 4-Br-A23187 containing medium (FIG. 21b) moreover, the inhibitory effect of Ca.sup.2+ was blocked by FK506. L-685,818 and rapamycin, which both interact with FK506-binding-protein-12, but do not inhibit calcineurin (29), failed to block the inhibitory action of Ca.sup.2+ on cAMP production (FIG. 21c). By contrast, cyclosporin A, which blocks calcineurin through a complex formed with cyclophilin A (Schreiber (1992) Cell 70:365-8) also abolished the effect of Ca.sup.2+. These data indicate that the human AC9 expressed in HEK293 cells is able to synthesize cAMP and its activity is inhibited by Ca.sup.2+ via calcineurin.

Tissue Distribution of Two Forms of Human AC9 mRNA

Analysis of hAC9 mRNA by Northern blotting (FIG. 22a) using a radio-labelled probe derived from the 5' end of the hAC9 cDNA (SmaI-MluI fragment; 528 bp) revealed hybridisation to a transcript of approximately 8.5 kb in all tissues represented, including brain, heart, skeletal muscle and pancreas. In addition, an approximately 6.3 kb mRNA was detected which appeared to be restricted to heart and skeletal muscle and gave a more intense signal than the 8.5 kb species. A human cDNA sequence (GenBank/EMBL Accession Number AB011092), described as part of a study reporting the predicted coding sequences of 100 cDNA clones isolated from human brain (Nagase et al, (1998) DNA Research 5:31-9), was retrieved by database searching and found to be identical to nucleotides 2900-5472 of the human AC9 cDNA sequence reported here but continues with a 3' untranslated region extended by 2234 bp. A cDNA fragment derived from this region used in hybridisation with the same MTN Northern blot detected the 8.5 kb human AC9 mRNA exclusively (FIG. 22c). These data are consistent with tissue-specific differential use of alternative polyadenylation signals on completion of mRNA transcription from the human AC9 gene. Subsequent searching of the GenBank/EMBL DNA sequence database for comparison of human AC9 cDNA 3'UTR sequences with the human AC9 gene locus (GenBank/EMBL Accession Number AC005736) confirmed the positions of alternative polyadenylation sites.

Expression of AC9 mRNA in Human Brain

To address the potential role of AC9 within the human brain, expression of AC9 mRNA was investigated by ribonuclease protection assay of total RNA samples isolated from post-mortem hippocampus, cortex and striatum (FIG. 23a). Protection of the anti-sense AC9 riboprobe was observed in each of these regions of human brain, indicating expression of AC9 mRNA in areas important for learning and memory functions. AC9 mRNA distribution in human forebrain was also investigated by in situ hybridisation (FIG. 23b) where enriched expression was detected specifically in neurones of the dentate gyrus and hippocampal CA1-CA3 regions.

Discussion

We have isolated and sequenced a human cDNA encoding an adenylyl cyclase which, distinct from a recently reported predicted human AC9 polypeptide (Hacker et al, (1998) Genomics 50:97-104), encodes a homologue of murine AC9. Cyclic AMP production by this new human AC9 cDNA clone is inhibited by Ca.sup.2+ through the protein phosphatase calcineurin. Analysis of the pattern of the expression of the novel AC9 mRNA indicates differential tissue specificity in the polyadenylation of transcripts. This has not been observed in the mouse and may provide an additional level of control in the regulation of AC9 expression in humans.

Confirmation of the Predicted Primary Sequence of the Human AC9 Protein

The human AC9 cDNA we have cloned predicts a polypeptide of the same length (1353 amino acid residues) as mouse AC9 and is 92% identical with the mouse AC9 polypeptide sequence. The novel human AC9 could be detected with anti-sera raised against the N- or the C-terminal regions of the mouse AC9 polypeptide sequence, confirming the predicted conservation of these regions between the two proteins. The 3' end of the cDNA reported here is identical to part of a 5'-truncated human AC9 cDNA clone KIAA0520 (Nagase et al, (1998) DNA Research 5:31-9). These sequences differ from that of another human AC9 cDNA clone reported recently by Hacker et al. who predicted an alternative C-terminus for the encoded polypeptide (Hacker et al, (1998) Genomics 50:97-104). By comparison of sequences, the prediction by Hacker et al. results from a 2 bp deletion causing a frame shift in the deduced amino acid sequence. Our data, therefore, predict and confirm an amino acid sequence for human AC9 which is distinct from that reported for a previously isolated clone of the human enzyme but which is conserved with mouse AC9.

Inhibition of Human AC9 Activity by Ca.sup.2+ /Calcineurin

Stably transfected HEK293-derived cell lines which over-express human AC9 showed a marked increase in cAMP production in comparison to wild type or pcDNA3-transfected control cells. The synthesis of cAMP was reduced by intracellular Ca.sup.2+ concentrations in the range of 100-150 nM which is similar to previous data with HEK293 cells overexpressing mouse AC9 (Paterson et al, Biochem. Biophys. Res. Commun. 214:1000-8; Antoni et al, (1998) J. Neurosci. 18:9650-61). The inhibitory effect of Ca.sup.2+ was blocked by FK506 or cyclosporin A, which are two mechanistically distinct blockers of the Ca.sup.2+ /calmodulin activated protein phosphatase calcineurin (Sabatini et al, (1992) Cell 70:365-8). In the presence of FK506 or cyclosporin A, cAMP production in hAC9 expressing cells was enhanced. Furthermore, application of L685,818 or rapamycin had no effect on cAMP production. The results indicate that the activity of human AC9, like mouse AC9 (Paterson et al, Biochem. Biophys. Res. Commun. 214:1000-8; Antoni et al, (1995) J. Biol. Chem. 270:28055-61; Antoni et al, (1998) J. Neurosci. 18:9650-61), is under inhibitory control by Ca.sup.2+ which is mediated by calcineurin. The lack of Ca.sup.2+ inhibition of AC9 observed by others may be due to analysis in insect cells (Premont et al, (1996) J. Biol. Chem. 271:13900-7) which may lack the necessary co-factors or may be attributable to a feature of the cDNA and/or transfected cell clone analysed (Hacker et al, (1998) Genomics 50:97-104). Immunological detection of multiple human AC9 polypeptides may represent differential post-translational modifications in HEK293 cells. Correlation of these observations with multiple phosphorylation or other post-translational modifications of human AC9 protein remains to be addressed. Immunocytochemical analysis shown in FIG. 19 suggests that the great majority of human AC9 polypeptides expressed in HEK 293 cells are localised to the plasma membrane.

Differential Human AC9 mRNA Expression

In similarity with the mouse (Premont et al, (1996) J. Biol. Chem. 271:13900-7) and in agreement with a recent study of another human AC9 cDNA clone (Hacker et al, (1998) Genomics 50:97-104), a broad pattern of expression was observed for an 8.5 kb AC9 mRNA in human brain and peripheral tissues when assayed by Northern blotting using a probe derived from the 5' end of the cDNA we have isolated. In the present study, a 6.3 kb human AC9 mRNA also detected by this probe appeared to be restricted to human heart and skeletal muscle. Previously, however, a probe representing a 5'-truncated human AC9 cDNA detected a 6.3 kb transcript at high levels in skeletal muscle, at low levels in kidney, liver, lung and placenta but which was apparently absent from heart (Hacker et al, (1998) supra). Although the probes used in these two studies overlap by approximately 100 bp near the 5' end of the ORF, the hybridisation conditions used may mean that these probes would be capable of detecting differentially spliced human AC9 mRNA transcripts if an intron/exon junction were located within this region. Lack of detection of the 6.3 kb mRNA in heart may potentially reflect mRNA splicing of upstream exons, detectable by the 5' probe, to alternative downstream exons. Similarly, it is possible that hybridisation to the 6.3 kb transcript by the 5'-truncated 4.65 kb partial cDNA probe reflects the presence of alternative 5'-end sequences not detectable by the 5' probe used in this study. However, comparison of our cDNA sequence with human AC9 genomic sequence available in the GenBank/EMBL database (Accession Number AC005736) indicates no evidence of mRNA splicing in this region. The .beta.-actin control in our Northern blotting analysis suggests the presence of proportionally greater amounts of heart and skeletal muscle mRNAs. Therefore, it is possible that the 6.3 kb AC9 mRNA was present in other tissues but not at a level detectable under the conditions we have used and that expression of this transcript is not necessarily restricted to heart and skeletal muscle as our results suggest. A probe derived from the extended 3' untranslated region of the 8.5 kb human AC9 transcript which is present in the cDNA clone KIAA0520 exclusively detected the 8.5 kb mRNA in northern blotting. This indicates the use of alternative polyadenylation signals within the human AC9 gene to be the mechanism accounting for the observation of two mRNAs. Polyadenylation of many eukaryotic mRNAs is understood to contribute to inhibition of mRNA degradation and/or potentiate mRNA translation, reviewed in (Wickers et al, (1997) Curr. Opin. Genet. Dev. 7:220-32). In addition, 5' and particularly 3' untranslated regions (UTRs) have been found to be important for the regulation of mRNA translation, stability and localisation, reviewed in (Wickers et al, (1997) supra; Jackson et al, (1997) Curr. Opin. Genet. Dev. 7:233-41), most significantly during development, reviewed in (Curtis et al, (1995) Cell 81:171-8; Seydoux (1996) Curr. Opin. Genet. Dev. 6:555-61; Macdonald et al, (1996) Curr. Opin. Genet. Dev. 6:403-7). The spatial and/or temporal distribution of trans-acting factors interacting with cis elements within 3'UTRs and form a complex with the 5'UTR and associated translation initiation factors has been proposed a molecular mechanism for the control of gene expression by regulation of translation efficiency (Wickers et al, (1997) supra). Although sequences within the 3'UTRs of human and mouse AC9 mRNAs are highly conserved, the 3'UTRs of human AC9 transcripts extend beyond that of the mouse with the 8.5 kb mRNA possessing considerable more 3' untranslated sequence than the 6.3 kb message. Although the functional significance of 3'UTRs in AC9 mRNAs remains to be addressed but may be subject to differential post-transcriptional regulation providing an additional mechanism of control at the level of mRNA stability and/or translational efficiency facilitating differential regulation of AC9 expression which may be relevant for the appropriate function of the enzyme in humans.

In cardiac myocytes, adenylyl cyclase VI (AC6) expression has been shown to be limiting for transmembrane .beta.-adrenergic signal transduction. Furthermore, all currently known cardiac adenylyl cyclases are under inhibitory control by Ca.sup.2+. Thus precise control of cAMP levels mediated through regulation of AC expression and control of AC activity appears essential for optimal cardiac function (Gao et al, (1998) PNAS U.S.A. 95:1038-43). Alterations to the levels of AC expression in heart, including AC9, may therefore contribute to an impairment of cardiac function related to reduction or loss of cAMP production. A role for post-transcriptional regulation in the control of adenylyl cyclase expression has recently been suggested from studies of ACVIII (AC8) (Muglia et al, (1999) J. Neurosci. 19:2051-8) which have revealed conservation of large 5'UTRs of AC8 mRNAs from mouse, rat and human.

Detection of AC9 mRNA by ribonuclease protection assay revealed expression in hippocampus, neocortex and striatum of post-mortem human brain. In situ hybridisation analysis revealed enriched expression of AC9 mRNA in hippocampus and indicated a neuronal localisation in projection neurones of the CA1-CA3 regions. The Ca.sup.2+ /calmodulin activated protein phosphatase calcineurin, which regulates AC9 activity, is also highly abundant in these regions (Winder et al, (1998) Cell 92:25-37). These findings are likely to be relevant to the function of AC9 in brain, particularly in hippocampus, where precise control of Ca.sup.2+ and cAMP levels is pivotal for synaptic plasticity and suggests potential involvement in functions important for cognitive behaviour (Mons et al, (1995) Trends in Neurosci. 18:536:42; Xia et al, (1997) Curr. Opin. Neurobiol. 7:391-6; Antoni et al, (1998) J. Neurosci. 18:9650-61; Winder et al, (1998) Cell 92:25-27; Mons et al, (1998) Life Sciences 62:1647-52).

Isolation of a human AC9 has enabled our characterisation of the enzyme, allowing comparison with that of the mouse. Differential usage of polyadenylation signals within the gene providing the potential for alternative mRNA post-transcriptional regulation has been identified as a distinct feature of human AC9 likely to be important for precise control of expression. In similarity with mouse AC9, cAMP production by human AC9 is under inhibitory control by Ca.sup.2+ /calcineurin. Post-translational modifications of the human AC9 protein, including possible phosphorylation, now require characterisation and their potential functional consequences investigated. The physiological function of AC9 and its potential involvement in human disease, particularly in view of its broad expression pattern, are complex issues to address. Although important differences have been described, many features are common to the human and murine enzymes. The production of mouse models may, therefore, be useful in the investigation of specific functions of AC9 in vivo.

TABLE I Effect of pretreatment with FK506 (1 .mu.mol/l) and BAPTA/AM (20 .mu.mol/l) on cAMP accumulation induced by .beta.- adrenergic stimulation in AtT20 cells. Data are means .+-.SEM, n = 4/group. IBMX 0.5 mmol/l and rokpram 0.1 mmol/l present throughout. nd - not determined, previous studies showed that there is no effect of either preincubation protocol on basal cAMP levels. Isoproterenol Pretreatment (nmol/l) None FK506 BAPTA/AM None 0.43 .+-. 04 nd nd 0.4 0.7 .+-. 0.04 0.94 .+-. 0.09 0.96 .+-. 0.06 2 2.0 .+-. 0.08 3.3 .+-. 0.7 3.6 .+-. 0.6 10 9.5 .+-. 0.33 15.3 .+-. 2.5 11.4 .+-. 0.9 100 19.5 .+-. .5 31.4 .+-. 1.9 35.5 .+-. 3.4 1000 32.1 .+-. 0.5 45.3 .+-. 9 46.9 .+-. 9

TABLE II PCR protocols for 5' RACE base cDNA cloning of AC from AtT20 cell total RNA Reaction using primers 2 or 3 with UAP 1 cycle 95.degree. C. 5 min 80.degree. 5 min (Taq polymerase added) 5 cycles 94.degree. C. 45 sec 55.degree. C. 30 sec 72.degree. C. 90 sec 30 cycles 94.degree. C. 45 sec 57.degree. C. 30 sec 72.degree. C. 90 sec 1 cycle 72.degree. C. 10 min Reaction using primer 5 with UAP: 1 cycle 95.degree. C. 5 min 80.degree. C. 5 min (Taq polymerase added) 35 cycles 94.degree. C. 45 sec 55.degree. C. 30 sec 72.degree. C. 90 sec 1 cycle 72.degree. C. 10 min Reaction using primer 6 with UAP 1 cycle 95.degree. C. 5 min 80.degree. C. 5 min (Taq polymerase added) 35 cycles 94.degree. C. 60 sec 58.degree. C. 60 sec 72.degree. C. 90 sec 1 cycle 72.degree. C. 10 min

TABLE III Potential sites of phosphorylation by common S/T protein kinases in AC A KINASE: Motif: RXS Position RAS 85 RES 597 RTS 763 RSS 919 Motif: RXXS RASS 85 (SEQ ID NO: 15) RVDS 201 (SEQ ID NO: 16) RSRS 304 (SEQ ID NO: 17) RESS 597 (SEQ ID NO: 18) RPAS 972 (SEQ ID NO: 19) RVLS 1213 (SEQ ID NO: 20) C KINASE Motif: (S/T)X(K/R) S preferred with hydrophobic AA on COOH side of S (23/37 of known sites are like this) SIR 837 SMR 1022 SWR 1310 SVK 356 SVR 27 SYR 761 SYR 1002 SYR 1211 TAK 534 TLR 187 TVK 619 Motif (K/R)XX(S/T) (S preferred with hydrophobic AA) KIKTI 1105 (SEQ ID NO: 21) KTATL 266 (SEQ ID NO: 22) KISTL 436 (SEQ ID NO: 23) KKSSI 370 (SEQ ID NO: 24) KEDSL 732 (SEQ ID NO: 25) KFDSM 94 (SEQ ID NO: 26) KIQSM 1019 (SEQ ID NO: 27) RPASL 972 (SEQ ID NO: 28) CASEIN KINASE II: Motif: 3 acidic aa-s atter phosphorylation site are a strong indication (*), highly S preferring SCAEA 581 (SEQ ID NO: 29) SGFEV 588* (SEQ ID NO: 30) SLCEI 690 (SEQ ID NO: 31) SYQEE 765 (SEQ ID NO: 32) SEFET 885* (SEQ ID NO: 33) SWREP 1310 (SEQ ID NO: 34) TTSET 261 (SEQ ID NO: 35) TKCEK 432 (SEQ ID NO: 36) TKCEK 432 TGVEC 1199 (SEQ ID NO: 37) CAM KINASE II: Motif: XRXX(S/T) Position ERASS 84 (SEQ ID NO: 38) GRVDS 200 (SEQ ID NO: 39) VRSRS 303 (SEQ ID NO: 40) SRESS 596 (SEQ ID NO: 41) RRPAS 971 (SEQ ID NO: 42) YRVLS 1212 (SEQ ID NO: 43) KRHAT 358 (SEQ ID NO: 44) IREKT 719 (SEQ ID NO: 45) SRMDT 1194 (SEQ ID NO: 46) GRSPT 1270 (SEQ ID NO: 47)

TABLE IV Inhibition by 1 .mu.mol/l FK506 of the effect of extracellular calcium ions on 10 nmol/l CRF-induced cAMP formation in calcium-depleted AtT20 cells in the presence of 0.5 mmol/l IBMX (see Legend to Fig 1 and Example 1 for methods). Data are means .+-. SEM, pmol/well, n = 6/group. 2-way ANOVA gave a significant (P < 0.05) interaction between extracellular calcium and FK506. Unstimulated cAMP levels were 0.5 .+-. 0.06 pmol/l. Extracellular free [Ca.sup.2+ ] [mmol/l] Vehicle FK506 0 12.1 .+-. 0.16 10.3 .+-. 0.10 0.5 10.4 .+-. 0.96 10.0 .+-. 0.11 1.0 9.7 .+-. 0.10 10.4 .+-. 0.10 1.5 6.2 .+-. 0.04 11.3 .+-. 0.13 2.0 5.2 .+-. 0.54 9.5 .+-. 0.11