G-protein coupled receptors are the largest class of signaling proteins and are the targets for the majority of current therapeutics. The efficiency with which an extracellular ligand activates the signalling system is dependent upon the amount of activatable proteins available and stoichiometry of these proteins (e. g. a molecular ratio of one type of signalling protein to another). The signalling efficiency is dependent upon protein expression levels, protein compartmentalisation and on chemical alterations of the protein.

Thus, activation of most GPCRs causes them to be moved from the cell surface by a form of vesicular transport termed endocytosis. Activation of most GPCRs also stimulates a process known as rapid homologous receptor desensitization. For many GPCRs this has been shown to reflect ligand stimulated phosphorylation of specific amino acids (typically with C-terminal tail or third intracellular loop of the receptor) which facilitates the binding of B-arrestins, proteins which prevent the receptor from activating its G-protein. Thus rapid homologous receptor desensitization also reduces the amount of receptor protein available for G-protein activation and reduces the efficiency with which the extracellular ligand stimulates the cell.

Thus, for example, a GPCR activating a specific G-protein would be expected to signal inefficiently in a cell in which the G-protein was expressed at a low level, and where a GPCR has the potential to activate two G-proteins, the efficiency with which it activates these distinct pathways could be influenced by the stoichiometry of the receptors to the different G-proteins. Since GPCRs can also cause a reduction in the amount of G-protein a-subunits, and can cause a reduction in the efficiency of G-protein re-cycling, the stoichiometry of receptors to G-protein subunits available for signalling can vary, not only from cell to cell, but also in a given cell type over time. Finally, where the GPCR can activate more than one type of G-protein, receptor phosphorylation may also influence the relative efficiency with which a receptor activates distinct effector systems.

One example of a disease which may be treated by activation of GPCRs is mammary carcinoma. Mammary carcinoma is the most common cause of malignancy in developed countries. One in eight women develop the disease and almost 200,000 new cases are reported each year in the USA alone. It is currently treated by surgery, chemotherapy and, in the case of hormone-dependent cancers, by endocrine therapy. Anti-oestrogens and anti-androgens are the most common endocrine therapy.

GnRH regulates the secretion of luteinising hormone (LH) and follicle-stimulating hormone (FSH) from the pituitary and thereby controls gametogenesis and steroidogenesis in the gonads. It is secreted in a pulsatile fashion from the hypothalamus and each pulse of GnRH elicits a bolus of gonadotrophin (LH and FSH) secretion. This can be blocked with antagonists, or mimicked by agonists, but in the latter case, sustained stimulation causes a paradoxical desensitization. Thus both treatments will ultimately reduce circulating levels of gonadotrophins and gonadal steroids, explaining their efficacy at causing"medical castration"and reducing the proliferation of steroid-dependent mammary carcinoma cells.

For the same reasons, GnRH analogues are increasingly used to treat prostate cancer (the most common form of cancer in males) and a number of other hormone-dependent cancers including those of the ovary and endometrium. However, current therapy of these cancers has important limitations. Metastasis remains a major problem, as does the transition over time from hormone dependence to hormone-independence (e. g.. lack of dependence on gonadal steroids). Patient compliance with hormone therapy can be low, due to the side effects of sterility and/or loss of libido, and the reduction in circulating gonadal steroid levels has serious side effects on the bone and cardiovascular systems. Each of these cancers still has a high mortality rate and a great deal of research is therefore currently focused on improving current therapy. Much of this focuses on the various strategies of gene therapy where DNA is delivered to the cancer cells (e. g.. by packaging into a virus or by encapsulation or complex formation with lipids) where it drives the expression of a protein which either kills, or prevents the proliferation of, the cancer cell.

At the cellular level, GnRH acts via a GnRH-receptor (GnRH-R), which is a GPCR. A schematic representation of a GPCR is shown in Fig. 1 and a schematic representation of a GPCR receptor signalling is shown in Fig. 2. G-proteins bind to guanine nucleotides and act as"molecular switches"in which the GDP bound form is inactive and the GTP bound form is active. GPCRs act primarily via heterotrimeric G-proteins, containing a, P and y subunits. The active conformation of the GPCR acts as a guanine nucleotide exchange factor for its G-protein, stimulating exchange of GDP for GTP on the a-subunit. This liberates the active a-subunit and the (3y subunit complex to regulate their target proteins.

There are multiple forms of each of these subunits, and specific a-subunits activate specific effector proteins. Thus the G-protein G. may contain a-subunits which stimulate adenylyl cyclase, the G-protein Gq may contain aq or al subunits, which activate phospholipase C and the G-protein Gi may contain ai subunits which inhibit adenyly cyclase. The ßy subunits also activate effector proteins but in this case there is less specificity and the physiological relevance of such regulation is not well established. Many GPCRs appear to activate specific G-proteins but some can activate more than one type of heterotrimeric G-protein. The receptor for pituitary adenylyl cyclase activating polypeptide (PACAP, PVR1 receptor) for example, stimulates a Gq mediated activation of phospholipase C and a Gs mediated activation of adenylyl cyclase.

There are at least 1200 GPCR receptors in the human genome, most of which have not yet been cloned. Fig. 18 provides a list of examples of characterized GPCRs. The extracellular regions are involved in ligand recognition whereas the intracellular regions are involved in coupling to heterotrimeric G-proteins. Phosphorylation of serine or threonine residues in the 3rd intracellular loop or C-terminal tail mediates rapid homologous desensitization and agonist-induced internalization of many GPCRs. GPCRs are expressed by all mammalian cells. GnRH-R are also expressed in many hormone-dependent cancers and the proliferation of such cells can be inhibited by GnRH analogues. A therapy could involve blockade of the GPCR (where activation of exogenous GPCRs is thought to support proliferation and/or survival of cancer cells), activation of the GPCR (where this reduces proliferation and/or survival of cancer cells), internalization of cytotoxic ligands via GPCRs (this would include radio-labelled ligands in targeted radiotherapy) and immune attack on the cancer cells by antibodies recognising the receptors.

Extensive research in the field of fertility control has led to the development of thousands of GnRH analogues, including many which selectively activate or block GnRH-R from different species, and several cytotoxic GnRH analogues which selectively destroy GnRH-R bearing cells. Numerous toxins have been prepared by conjugation of peptides (usually via linking groups) to cytotoxic moeities such as ricin; modecoin; abrin; pokeweed antiviral protein, a-amanitin ; gelonin; barley, wheat, corn, rye or flax ribosome inhibiting proteins; diphtheria or shiga toxins; Pseudomonas exotoxin; melphaian; methotrexate; N mustards; doxorubicin; daunomycin; or their modified forms. GnRH analogues conjugated to cytotoxins have already been prepared and shown to inhibit the proliferation of numerous types of cancer cells in vitro and in vivo (Halmos G. et al. (1999) Proc. Natl.

Acad. Sc. USA 93, 2398-2402 ; Kahan Z. et al. (1999) Cancer 85; 2608-2615; Koppan M. et al. (1999) Prostate 38; 151-158; Milovanovic S. R. et al. (1993) Breast Cancer Res.

Treat. 24; 147-158; Miyazaki M. et al (1997) J Natl. Cancer Inst. 89; 1803-1809; Miyazaki M. et al. (1999) Am. J. Obstet. Gynecol. 180; 1095-1103; Nagy A. et al. (1993) Proc. Natl. Acad. Sci. USA 90 ; 6373-6376; Nagy A. et al. (1996) Proc. Natl. Acad. Sci.

USA 93; 7269-7273; Pinski J. et al. (1993) Prostate 23; 165-178; Szepeshazi K. et al.

(1992) Anticancer Drugs 3; 109-116; Szepeshazi K. et al. (1997) Anticancer Drugs 8; 974-987).

Unlike all other known GPCRs, mammalian GnRH-R do not rapidly desensitise or internalise. These responses to GPCR activation are typically dependent upon phosphorylation of sites within the GPCR's C-terminal region, termed"tails"and mammalian GnRH receptors, uniquely, lack such tails (thus the"receptor"tends to remain accessible and activatable, on the cell surface). In contrast, all non-mammalian GnRH-Rs cloned to date do have C-terminal tails and (where investigated) do internalise rapidly.

Fig. 13 shows the amino acid sequences of the C-terminal regions of several cloned GnRH-Rs with the approximate extent of the 7"trans-membrane region (7TM) and C-terminal tails indicated. All cloned mammalian GnRH-R (type I) lack C-terminal tails and do not extend appreciably beyond the 7 TM whereas all cloned non-mammalian GnRH-Rs do have C-terminal tails containing multiple serine (S) and threonine (T) residues. These are potential phosphorylation sites for G-protein receptor kinases or other kinases (such as protein kinases A and C) and such phosphorylation mediates desensitization and agonist-induced internalization of many GPCRs. Mammalian GPCRs are thought to densensitise and internalise slowly (if, at all) because they lack the required C-terminal tail phosphorylation sites (Davidson J. S. et al. (1994) Biochem. J. 300; 299-302; McArdle C. A. et al. (1996) JBiol Chem 271, 23711-23717 ; McArdle C. A. et al.

(1999) Mol Cell Endocrinol 151 ; 129-136 ; Heding A. et al. (1998) J. Biol. Chem. 273; 11472-11477; Pawson A. J. et al. (1998) J. Endocrinol. 156; R9-12).

Further, although the vast majority of work on GnRH-R has focused on those expressed by pituitary cells (gonadotrophs), they are also found (along with GnRH) in some mammary, prostate, endometrial and ovarian cancers (and in cell lines derived from them). Moreover, GnRH analogues can inhibit the proliferation of cell lines derived from such cancers both in vitro and in vivo. Therefore, direct anti-proliferative GnRH treatments on cancer cells may contribute to the efficacy of GnRH analogues in cancer treatment. Indeed the effectiveness of GnRH analogues in treatment of mammary cancer in post-menopausal women and in the treatment of steroid independent cell lines in vivo clearly attest to the importance of such direct effects. Nevertheless, the value of GnRH or GnRH analogues or other compounds which act by binding to GnRH-R (here collectively termed GnRH analogues) as a direct regulator of these cancers remains controversial. This is largely because many hormone-dependent cancers do not express GnRH-R and even where expression of the GnRH-R mRNA can be demonstrated (by PCR), levels of the expressed receptor protein may be too low for detection.

Other work in this area include the selective target cell activation by expression of a GPCR activated superiorly by a synthetic ligand (W097/35478-A), enhancing the sensitivity of tumour cells to therapies by loss of wild-type therapy sensitising gene (W095/30002-A), use of cytotoxic GnRH analogues in hormone-dependent cancers (EP-A-89118460.8), and in destroying gonadotrophs (US-6103881). Modified GnRH-Rs only and their influence on internalization and/or densitisation rates have been disclosed by Heding et al (1998, J. Biol Chem, 273 (19), 11472-11477,), Pawson et al (1998, JEndocrinal, 156 (3), R9-R12) and Lin et al (1998, Mol. Endocrinol 12 (2), 161-171). Modified GnRH-Rs have been expressed in stable cell lines (W099/67292-A and US-5985583). Whilst studying effects of GnRH analogues on proliferation of human mammary cancer-derived cell lines, the inventors found that such analogues did not inhibit the proliferation of cell lines (MCF7) in which GnRH-R were undetectable (in radioligand binding assays) and had only a modest inhibitory effect on proliferation of cell lines in which low levels of endogenous GnRH-R were detectable (T47D). The inventors also found that, although conventional (liposome-based) transfection techniques were inefficient in these cells, it was possible using recombinant adenovirus (Ad) expressing GnRH-R, to transfect high numbers of GnRH-R into these cell lines. Moreover, the expression of GnRH-R on these cells caused by infection with the recombinant Ad enabled GnRH analogues to inhibit the proliferation of these cancer cells. On the basis of these observations, the inventors conclude that gene therapy, used to increase expression of GnRH-R, would increase the effectiveness of therapies based on occupation of the GnRH-R. Therefore this principle would apply to any form of GnRH-R targeted therapy but would be of particular value in treatment of cancers where GnRH or GnRH analogues are known to act directly to inhibit cell proliferation and/or to cause cell death. Therefore, by delivering DNA encoding GnRH-R to cancer cells, increased expression of the receptor protein can be achieved, thereby increasing the effectiveness of therapies targeting these cells via the GnRH-R.

In addition, the desensitization and/or internalization differences between mammalian GnRH-Rs and all other known GPCRs, may be exploited in optimisation of GnRH-R directed cancer therapy. It may therefore be possible to accelerate internalization of a mammalian GnRH-R (and thereby increase the internalization of a GnRH analogue conjugated to a cytotoxic moeity) by addition of a C-terminal tail from a non-mammalian GnRH-R, or to reduce the rate of desensitization (and thereby increase the effectiveness of a GnRH agonist) by removing the C-terminal tails of a non-mammalian GnRH-R or by removing phosphorylation sites from the C-terminal tail of the non-mammalian GnRH-R.

Fig. 14 shows amino acid sequences of the 3rd intracellular loops (e. g.. the sequences between the 5"'and 6"trans-membrane domains) as predicted by SWISS-PROT and by alignment to SWISS-PROT predictions. Phosphorylation of serine (S) and/or threonine (T) residues is important for agonist-induced internalization of, for example, the a2A adrenergic receptor (Liggett S. B. et al. (1992) J. Biol. Chem. 267; 4740-4746) but there is no evidence that these regions play this role in GnRH-R internalization. It may therefore be possible to accelerate internalization of mammalian GnRH-R by exchanging the 3rd intracellular loop of GnRH-R with the corresponding region from the a2A adrenergic receptor. These sequences are all public domain (full amino acid sequences, corresponding nucleic acid sequences, and accession numbers are given below). Thus GnRH-R selected (or engineered) for maximal rates of internalization might be more suitable for delivery of cytotoxic GnRH derivatives to cancer cells, whereas receptors selected (or engineered) for minimal rates of desensitization and internalization would be preferable for therapy based on agonist activity. In either case, the receptors could be selected or engineered to permit or prevent activation by mammalian or non-mammalian GnRH peptides and analogues.

Recombinant Adenovirus expressing human, sheep and type I Xenopus laevis (Ad hGnRH-R, Ad sGnRH-R and Ad XGnRH-R respectively) can be used to effect an extremely efficient transfection and expression of GnRH-R in cell lines derived from mammary cancer (MCF7 and T47D), prostate cancer (PC3), endometrial cancer, ovarian cancer (OVCAR), pituitary cancer (aT4) and cervical cancer (Hela). In each case Ad-mediated expression enables receptor number to be increased to levels well above that of any endogenous GnRH-R. Receptor numbers can be controlled by varying viral titre and transfection efficiencies approach 100% when the ratio of viral particles to transfected cells (the multiplicity of infection, m. o. i.) is 10 or greater.

In MCF7, PC3, aT4 and Hela cells these receptors are functional and mediate GnRH-stimulated activation of phospholipase C and mobilization of Ca2+. Their pharmacological characteristics are indistinguishable from those of endogenous pituitary cell GnRH-Rs in terms of agonist potency, rank order of potency and blockade by antagonists (PLC activation assays and binding experiments). Importantly, when GnRH-Rs were expressed in aT4 pituitary cells, the ligand specificity observed with expressed human or sheep GnRH-Rs differed from that with Xenopus receptors and similar specificity was seen when these receptors were expressed in mammary cancer cells.

Moreover, as anticipated from earlier work with non-mammalian and mammalian GnRH-R (which do and do not, respectively, possess C-terminal tails) the inventors have found that the Xenopus GnRH-Rs desensitize and internalize rapidly, whereas the human GnRH-Rs do not.

Infection of human mammary cancer cell lines (MCF-7 and T47D) with Ad sGnRH-R facilitates or potentiates a direct antiproliferative effect of GnRH (as judged by [3H] thymidine incorporation assays). This effect is mimicked by GnRH-R agonists (Buserelin and chicken GnRH-11 (cGnRH-11)) with the rank order of potency anticipated from radioligand binding studies (Buserelin>GnRH>cGnRH-II). Buserelin is highly potent in these assays (half maximal inhibition of proliferation at 10-100 pM) and this effect is not mimicked but, instead, is competitively blocked by peptide antagonists of pituitary GnRH-R. The inventors have also shown that infection of human prostate cancer derived cell lines with Ad sGnRH-R facilitates a direct anti-proliferative effect of GnRH.

Importantly, the inventors have found that cGnRH-11, which has a relatively high affinity for XGnRH-Rs and a low affinity for sGnRH-Rs does not inhibit proliferation of MCF7 cells infected with Ad XGnRH-R but does inhibit proliferation of cells infected with Ad sGnRH-R (albeit at high concentrations). This clearly implies that the anti-proliferative effect is mediated most efficiently by a GPCR which does not rapidly desensitize or internalize.

GPCR internalization can be important for desensitization, down-regulation and for delivery of ligand-conjugated cytotoxins into receptor expressing cells. To investigate mechanisms of GnRH-R internalization the inventors infected Hela (cervical cancer) cells with Ad hGnRH-R and Ad XGnRH-R and found that the Xenopus GnRH-R is internalised more rapidly than the human GnRH-R. The Hela cells used had been altered to express a dominant negative mutant of the protein dynamin (K44A dynamin) in a regulated manner (the mutant was not expressed when cells were cultured with tetracycline but was expressed when tetracycline was omitted). Using this system, the inventors have found that internalization of the hGnRH-R is dynamin-independent, whereas internalization of the XGnRH-is largely dynamin-dependent. Since GPCRs are internalised via transport vesicles and dynamin regulates the formation of many such vesicles, the implication is that the Xenopus GnRH-R is internalised faster than the human, because it is able to access an internalization route from which the human receptor is excluded. Since C-terminal tails of many GPCRs have been implicated in receptor internalization, the inventors have generated a chimeric protein consisting of the entire human GnRH-R sequence with an additional C-terminal tail from the Xenopus GnRH-R. The inventors have found that this addition caused the hGnRH-R receptor to desensitize more rapidly and confers dynamin sensitivity to its internalization.

Together, these experiments have established 1) that recombinant adenovirus can be used to express GnRH-Rs in hormone-dependent cancer cells 2) that these receptors are functional and retain pharmacological characteristics 3) that activation of such receptors can cause a pronounced anti-proliferative effect 4) that the magnitude of this effect is dependent upon receptor number 4) that the magnitude of the effect can be limited by receptor desensitization and/or internalization and 5) that rates of desensitization and internalization of the GnRH-R are dependent upon receptor structure and can be altered by modification of receptor structure The above strategies are all equally applicable to many other GPCRs and are not limited to GnRH-Rs. Indeed, it was recently shown that when type 2A somatostatin receptors (sst2) were transfected into a pancreatic cancer cell line, this dramatically reduced the growth and metastasis of these cells in vivo and enhanced the inhibitory effect of a cytotoxic derivative of somatostatin on tumour growth (Benali et al. 2000 PNAS, 97: 9180-9185).

According to a first aspect of the invention, there is provided a pro-drug comprising a vector encoding a G-protein coupled receptor (GPCR). The GPCR may be modified from the naturally occurring GPCR, and can be modified in at least one of the following manner: 1. an N-terminal portion of a GPCR from which a part or all of the C-terminal tail has been truncated in order to remove one or more serine or threonine residues, 2. a GPCR in which one or more of the serine and/or threonine residues in the C-terminal tail have been mutated to prevent their phosphorylation e. g.. a serine to alanine, or threonine to alanine mutation, 3. the entire sequence of a GPCR lacking a C-terminal tail with an added C-terminal tail containing potential sites for phosphorylation and p-arrestin binding, 4. a GPCR in which one or more serine and/or threonine residues in the 3rd intracellular loop have been mutated to prevent their phosphorylation (e. g.. a serine to alanine, or threonine to alanine mutation), 5. a GPCR in which the 3rd intracellular loop has been replaced with one from another GPCR, 6. a fusion protein consisting of the entire sequence of a GPCR in tandem with the sequence of a G-protein a-subunit.

Such modification could be made on: 1. a mammalian GnRH-R or somatostatin receptor, 2. a non-mammalian GnRH-R or somatostatin receptor, 3. a mammalian GnRH-R with the C-terminal tail of a non-mammalian GnRH-R added, 4. a GnRH-R or a somatostatin receptor, 7. a mammalian GnRH-R with a third intracellular loop from an a2A adrenergic receptor, 8. a GnRH-R or a somatostatin receptor in tandem with an ai, as or aq subunit.

The term'pro-drug'is conventionally used to refer to a pharmaceutical which is converted in vivo into an active drug. Here the term is extended to encompass vectors encoding GPCRs which are intended to be used in combination with another drug. Specifically the vector-encoding the GPCR is intended to be used in combination with GPCR-targeted therapy, in order to improve the efficiency and or specificity of such therapy or to enable such therapy. In this context, the vector encoding the GPCR is used to increase the amount of target GPCR and the GPCR may be selected or modified to influence pharmacological characteristics such as receptor-ligand specificity, receptor desensitization, receptor internalization and receptor signalling.

Any of the above could be a human GPCR, a non-human GPCR, a mammal GPCR, a non-mammal GPCR, a vertebrate GPCR, a non-vertebrate GPCR, an animal GPCR, a non-animal GPCR.

The pro-drug may be selected from mammalian (e. g. human), amphibian, bird, fish or insect GPCR and may be human or non-human. Where the GPCR is non-human, it may be selected from sheep, cow, rat, mouse, Xenopus, chicken, goldfish or catfish.

Preferably, the GPCR is selected from any one of the nucleotide sequences listed in sequence 2,4,6,8,10,12,14,16,18,20,22,24 and the nucleotide sequence to any one of the amino acid sequences listed in Figure 17, or any variations of those sequences due to substitutions, deletions, and/or additions.

Preferably, the vector is selected from an adenovirus, adeno-associated virus, herpes simplex virus, moloney murine leukaemia virus, minimal or gutless adenovirus, lentivirus, retrovirus, or any non-viral vector such as a lipid.

A preferred GPCR for the pro-drug may be gonadotrophin-releasing hormone receptor (GnRH-R) or a somatostatin receptor.

According to a second aspect of the invention, there is provided a use of a vector encoding a GPCR in the preparation of a medicament for the treatment of diseases. Preferably, the GPCR is modified in relation to the naturally occurring GPCRs according to any one of the following manner: 1. an N-terminal portion of a GPCR from which a part or all of the C-terminal tail has been truncated in order to remove one or more serine or threonine residues, 2. a GPCR in which one or more of the serine and/or threonine residues in the C-terminal tail have been mutated to prevent their phosphorylation e. g.. a serine to alanine, or threonine to alanine mutation, 3. the entire sequence of a GPCR lacking a C-terminal tail with an added C-terminal tail containing potential sites for phosphorylation and p-arrestin binding, 4. a GPCR in which one or more serine and/or threonine residues in the 3rd intracellular loop have been mutated to prevent their phosphorylation (e. g.. a serine to alanine, or threonine to alanine mutation) 5. a GPCR in which the 3rd intracellular loop has been replaced with one from another GPCR, 6. a fusion protein consisting of the entire sequence of a GPCR in tandem with the sequence of a G-protein a-subunit.

Examples of G-protein a-subunits are provided in Fig. 19, but the invention is not limited to these amino acid sequences.

Such modification could be made on: 1. a mammalian GnRH-R or somatostatin receptor, 2. a non-mammalian GnRH-R or somatostatin receptor, 3. a mammalian GnRH-R with the C-terminal tail of a non-mammalian GnRH-R added, 4. a GnRH-R or a somatostatin receptor, 7. a mammalian GnRH-R with a third intracellular loop from an a2A adrenergic receptor, 8. a GnRH-R or a somatostatin receptor in tandem with an ai, as or aq subunit.

Any of the above could be a human GPCR, a non-human GPCR, a mammal GPCR, a non-mammal GPCR, a vertebrate GPCR, a non-vertebrate GPCR, an animal GPCR, or a non-animal GPCR.

The GPCR may be selected from a mammal such as a human, an amphibian such as Xenopus, bird, fish, insect, sheep, cow, rat, mouse, chicken, goldfish or catfish. The GPCR may be selected from any one of the nucleotide sequences listed in sequences 2,4,6,8,10, 12,14,16,18,20,22,24 and the nucleotide sequence to any one of the amino acid sequences listed in Fig. 17 as well as any variations of these sequences due to substitutions, deletions, and/or additions.

Preferably the vector is selected from adenovirus, adeno-associated virus, herpes simplex virus, moloney murine leukaemia virus, minimal or gutless adenovirus, lentivirus, retrovirus, or any non-viral vectors such as a lipid.

Preferably the GPCR is a GnRH-R or somatostatin receptor.

According to a third aspect of the invention there is provided a GPCR modified according to any one of the following:- 1. an N-terminal portion of a GPCR from which a part or all of the C-terminal tail has been truncated in order to remove one or more serine or threonine residues, 2. a GPCR in which one or more of the serine and/or threonine residues in the C-terminal tail have been mutated to prevent their phosphorylation (e. g.. a serine to alanine, or threonine to alanine mutation), 3. the entire sequence of a GPCR lacking a C-terminal tail with an added C-terminal tail containing potential sites for phosphorylation and p-arrestin binding, 4. a GPCR in which one or more serine and/or threonine residues in the 3rd intracellular loop have been mutated to prevent their phosphorylation (e. g.. a serine to alanine, or threonine to alanine mutation), 5. a GPCR in which the 3rd intracellular loop has been replaced with one from another GPCR, 6. a fusion protein consisting of the entire sequence of a GPCR in tandem with the sequence of a G-protein a-subunit.

Such modifications could be made on: 1. a mammalian GnRH-R or somatostatin receptor., 2. a non-mammalian GnRH-R or somatostatin receptor, 3. a mammalian GnRH-R with the C-terminal tail of a non-mammalian GnRH-R added, 4. a GnRH-R or a somatostatin receptor, 7. a mammalian GnRH-R with a third intracellular loop from an a2A adrenergic receptor, 8. a GnRH-R or a somatostatin receptor in tandem with an ai, as or aq subunit.

According to a fourth aspect of the invention there is provided a nucleotide sequence encoding a GPCR modified as described herein.

Preferably the nucleotide sequence is selected from any one of those described herein or any variations of the sequences due to substitutions, deletions'and/or additions. The invention is not limited to these sequences only and any GPCR sequence may be used. For example, the nucleotide sequences relating to the amino acid sequences listed in Fig. 17 may be used.

According to a fifth aspect of the invention there is provided a vector comprising a nucleotide sequence of a GPCR according to any of the sequences described herein.

Preferably the vector is selected from a virus such as an adenovirus, adeno-associated virus, herpes simplex virus, moloney murine leukaemia virus, minimal or gutless adenovirus, lentivirus, retrovirus, a capsid variant with altered cell targeting. This may ensure that the expressed product is targeted directly to the cell, rather than being broken down in the body before reaching the target cell or tissue. Alternatively, non-viral vector such as a lipid, DNA, lipofection and polyfection or receptor mediated gene transfer or means intended to increase their efficiency such as the use of membrane modifying peptides (Wagner. E.

(1999) Advanced Drug Protein 38: 279-289).

According to a sixth aspect of the invention there is provided a method of treating a disease comprising supplying to a patient a vector comprising a GPCR. Preferably the patient is suffering from a condition in which there is GPCR influence. The disease may be cancer, cardiovascular disease, nervous system disorders, digestive system disorders, immune system disorders, respiratory diseases, skeletal and other bone diseases, endocrine disorders, sensory disorders, or muscle diseases. Preferably the vector is as defined as in the fifth aspect of the invention and the GPCR is as defined in the third aspect of the invention and the nucleotide sequence according to the fourth aspect of the invention.

The GPCR may be a GnRH-R or a somatostatin receptor.

Any of the products described herein may be delivered either before, after or simultaneously with other compounds to enhance their action at the target cell or tissue.

Such compounds may be those which act specifically on a GPCR.

According to a seventh aspect of the invention there is provided a method of treating a disease comprising supplying to a patient a pro-drug according to the first aspect of the invention.

According an eighth aspect of the invention there is provided a method of treating a disease comprising supplying to a patient a nucleotide sequence according to the fourth aspect of the present invention.

According to a ninth aspect of the invention there is provide a method of treating a disease comprising supplying to a patient a GPCR and preferably a GPCR according to the third aspect of the invention.

According to a tenth aspect of the invention there is provided a method of increasing GPCR levels in a patient by introducing a vector or a pro-drug according to the invention.

Advantageously the methods of treatment described above influence the signalling, desensitization, internalization or expression in a cell.

According to an eleventh aspect of the invention there is provided a method of screening compounds for antiproliferative or cytotoxic effects comprising the steps of preparing GPCR expressing cancer cells by infection with an adenovirus comprising a GPCR nucleotide sequence into the cells, supplying the compound to the cells and analysing the effects on the cells.

According to a twelfth aspect of the invention there is provided a pharmaceutical composition comprising a vector encoding a GPCR and preferably a vector as defined in the fifth aspect of the invention.

Preferably the GPCR is a GnRH-R or a somatostatin receptor.

A GPCR as defined herein is understood to mean receptor that, upon binding of its natural peptide or nonpeptide ligand and activation of the receptor, transduces a G protein-mediated signal (s) that results in a physiological, cellular response (e. g cell proliferation of secretion). G-protein coupled receptors from a large family of evolutionary related proteins. Proteins that are members of the G-protein coupled receptor family are generally composed of seven putative transmembrane domains, and thus exhibit a structure similar to that shown in Fig. 1. G-protein coupled receptors are also known in the art as "seven transmembrane segment (7TM) receptors"and as"heptahelical receptors" (see e. g Schwartz (1994) Curr. Opin. Biotechnol. 5: 434-444).

A"G protein"as defined herein is understood to mean a protein belonging to a large family of proteins that interact with G-protein coupled receptors to facilitate cellular responses (e. g directly or via cellular second messengers). G proteins are composed of a heterotrimer of three separate amino acid chains (Ga, Gp, Gy). Although the Ga, subunit binds GTP, "G protein"refers to the complete heterotrimer.

Constructs and methods in accordance with the invention will now be described by way of example only, and with reference to accompanying Figures 1 to 19 in which: Fig. 1 is a schematic representation of a GPCR; and Fig. 2 is a schematic representation of a GPCR receptor signalling activity; and Fig. 3 A, B, C are graphs of the levels of specific and non-specific binding of ['25I] Buserelin; and Fig. 4 A, B are graphs of the responsiveness of aT4 pituitary cells (lacking endogenous GnRH-R) to GnRH after infection with replication-deficient recombinant adenovirus expressing GnRH-R ; and Fig. 5 A, B, C, D are graphs demonstrating that pharmocological characteristics of human and Xenopus GnRH-R expressed using recombinant adenovirus in MCF-7 and aT4 cells.

Fig. 6 A, B are graphs of the [3H] thymidine incorporation into newly synthesized DNA in MCF-7 and T47 cells.

Fig. 7 A, B, C are graphs of the incorporation of [3H] thymidine into newly synthesised DNA in MCF-7 cells; and Fig. 8 is a graph of the incorporation of [3H] thymidine into newly synthesised DNA in MCF-7 cells cultured in Buserelin and Antide; and Fig. 9 is a bar chart of the [3H] thymidine incorporation into newly synthesised DNA in MCF-7 cells; and Fig. 10 A, B are graphs showing desensitization and internalization of human and Xenopus GnRH-R in aT4 cells; and Fig. 11 is a graph of-[3H] thymidine incorporation in cells infected with AdGnRH-R or Ad sGnRH-R; and Fig. 12A, B, are graphs demonstrating the effect of GnRH and Buserelin on cell proliferation in the presence or absence of an oestrogen receptor antagonist and agonist; and Fig. 13 shows amino acid sequences of the C-terminal regions of cloned GnRH-Rs; and Fig. 14 shows amino acid sequences of the 3rd intracellular loops (e. g. the sequences between the 5th and 6th trans-membrane domains) as predicted by SWISS-PROT and by alignment to SWISS-PROT predictions; and Fig. 15A is a graph demonstrating internalizing of human GnRH-R in Hela cells and 15B is a graph demonstrating internalizing of Xenopus GnRH-R in Hela cells; and Fig. 16 is a bar chart demonstrating dynamin-dependence of receptor internalization; and Fig. 17 shows amino acid sequences of various GPCRs ; and Fig. 18 provides a list of examples of better characterised GPCRs.

Fig. 19 shows amino acid sequence of various G-protein a-subunits.

1. Expression of wild-type human and non-human G-protein coupled receptor As noted above, GnRH-R are expressed (often along with GnRH) in some mammary, prostate, endometrial and ovarian cancers and GnRH analogues can inhibit the proliferation of cell lines derived from such cancers (in vitro and in vivo) raising the possibility that direct anti-proliferative effects on cancer cells may contribute to the efficacy of GnRH analogues in cancer treatment. Nevertheless, the value of GnRH as a direct regulator of these cancers remains controversial, largely because many hormone-dependent cancers do not express GnRH-R and even where expression of the GnRH-R mRNA can be demonstrated, levels of the expressed receptor protein may be too low for detection. It is therefore likely that the low GnRH-R density in these cancers limits the efficacy GnRH analogues as direct regulators of mammary, prostate, ovarian and endometrial cancer. As a strategy for increasing GnRH-R number we have developed recombinant adenovirus expressing GnRH-R and have found that infection with these does, indeed, facilitate an anti-proliferative effect of GnRH analogues in human mammary and prostate cancer-derived cell lines. The implication is that gene therapy, used to increase GnRH-R expression in such cancers would increase the effectiveness of GnRH-R targeted cancer therapies in vivo.

Somatostatin receptors are a family of 6 GPCRs (sstl, 2A, 2B, 3,4,5) which couple via inhibitory G-proteins (Gi) to adenylyl cyclase but can also activate other effectors such as Ca2+ channels, K+ channels and phosphatases (Schonbrunn A 1999 Ann Oncol 10 Suppl 2: S17-21). Somatostatin analogues which activate these receptors are widely used for cancer therapy and for tumour imaging and the effectiveness of somatostatin analogues in such therapy depends largely upon the number of sst2 receptors expressed in the target tumour (Froidevaux S et al. 1999 Cancer Res 59: 3652-7). Accordingly, strategies circumventing agonist-induced receptor down regulation or favoring agonist-induced receptor up-regulation are being explored as means of improving the efficacy of somatostatin analogue therapy (Froidevaux S et al. 1999 Cancer Res 59: 3652-7). Gene therapy with vectors encoding somatostatin receptors provides an alternative strategy for increasing somatostatin receptor number and thereby increasing the effectiveness of somatostatin receptor targeted therapy.

Pharmacological characteristics (e. g.. ligand specificity) for any given GPCR type, can vary greatly from species to species. An extension of the arguments above is therefore that gene therapy with vectors encoding non-human GPCRs could enable the different functional characteristics of these receptors to be exploited. As an example we have expressed mammalian (sheep, human) and non-mammalian (Xenopus type I) GnRH-R in mammary cancer cells. The rank order of potency for activation of the mammalian GnRH-R (Buserelin>GnRH>cGnRH-II) differed to that seen for activation of the non-mammalian GnRH-R (cGnRH-II>Buserelin>GnRH). Indeed, when Xenopus GnRH-R are expressed, maximal activation can be achieved with a concentration of cGnRH-II which is ineffective at mammalian GnRH-R (Fig. 5). The implication is that gene therapy used to increase the expression of non-mammalian GnRH-R (e. g.. in mammary or prostate cancer) may enable these receptors to be targeted without concomitant activation of endogenous GnRH-R (e. g.. without chemical contraception).

Considerable evidence exists to suggest that somatostatin receptors undergo agonist-induced phosphorylation, P-arrestin binding, desensitization and internalization and that these processes (by limiting the amount of activatable somatostatin receptor at the cell surface) can reduce the effectiveness of somatostatin receptor targeted therapy (Schonbrunn 1999 Ann Oncol 10 Suppl 2 : S17-21 ; Hipkin et al. 1997 J Biol Chem 272: 13869-76; Hukovic et al. 1998 JBiol Chem 273 : 21416-22; Mundell and Benovic 2000 J Biol Chem 275: 12900-8; Kreienkamp et al. DNA Cell Biol 1998 17: 869-78). However, the sst4 subtype is thought not to show agonist-induced desensitization or internalization.

Since sst4, like sst2, signals via Gi to adenylyl cyclase, gene therapy with vectors encoding this non-desensitizing Gi coupled GPCR may facilitate more effective somatostatin receptor targeted therapy than that achieved by gene therapy with a desensitising somatostatin receptor.

2. Expression of GPCRs modified by addition of a C-terminal tail.

Mammalian GnRH-R are unique amongst known GPCRs, in that they completely lack C-terminal tails. Phosphorylation of sites within C-terminal tails of many GPCRs has been implicated in their rapid desensitization and internalization and the lack C-terminal tails (and associated phosphorylation sites) in mammalian GnRH-R is thought to underlie their resistance to desensitization and their slow internalization. The importance of C-terminal tails in this respect is illustrated by work in which incremental truncation of the chicken GnRH-R, incrementally reduced the rate of receptor internalization (Pauson et al. (1998) J.

Endocrinol 156 : R9-R12) Moreover, addition of a C-terminal tail from the catfish GnRH-R to a mammalian GnRH-R was found to increase the rate of internalization and desensitization (Heding et al (1998) J. Biol. Chem 273: 11472-11477). In addition we have recently investigated the dependence of GnRH-R internalization on dynamin, a protein which regulates the internalization of many GPCRs. Human GnRH-R internalization was found to by dynamin-independent whereas internalization of the Xenopus GnRH-R was largely dynamin-dependent. The internalization of a chimeric receptor consisting of the Xenopus C-terminal tail added to the human GnRH-R was also largely dynamin-dependent. Thus it appears that the C-terminal tail of the Xenopus GnRH-R enables it to access an additional (dynamin-dependent) route of internalization and that addition of this tail to a mammalian GnRH-R enables it to target this additional internalization route.

The lack of receptor desensitization may well be advantageous for therapies requiring sustained receptor activation but the low rate of internalization may be disadvantageous for therapies based upon receptor internalization. In recent years, numerous cytotoxic derivatives of GPCR activating ligands have been developed with the rationale that receptor mediated internalization of such compounds may be used to deliver the cytotoxin specifically to receptor expressing cells. Cytotoxic derivatives of GnRH analogues and of somatostatin analogues have already been shown to be inhibit growth and/or metastasis of tumours expressing the corresponding GPCRs in vivo. Assuming that receptor mediated internalization mediates such effects, it may be possible to increase effectiveness of such cytotoxins by increasing receptor internalization rates. Accordingly, addition of a C-terminal tail from (e. g..) a non-mammalian GnRH-R, to a GPCR lacking a C-terminal tail, would be expected to increase the rate of receptor internalization and thereby increase the effectiveness of therapy dependent upon receptor mediated internalization. Similarly, gene therapy with a vector encoding a somatostatin receptor which undergoes agonist-induced internalization (e. g.. the sst2) would be preferable in this regard to gene therapy using a vector encoding a somatostatin receptor which does not show agonist-induced internalization (e. g.. the sst4 receptor).

3. Expression of receptors modified by truncation and/or mutation of C-terminal tails and/or 3rd intracellular loops.

For many G-protein coupled receptors phosphorylation of specific amino acids within their C-terminal tails and/or third intracellular loops mediates receptor desensitization and/or agonist-induced internalization. Receptor desensitization and/or internalization may reduce the therapeutic benefit achieved by GPCR activation so, where gene therapy with a vector encoding a GPCR is used to improve therapy based upon GPCR activation, modifications of the receptor structure which inhibit these processes may be beneficial. For example, we have found that when mammalian (sheep) and non-mammalian (Xenopus type I) GnRH-R are expressed in mammary cancer cells, only the mammalian receptors mediate anti-proliferative effects of GnRH analogues. Indeed, cGnRH-11 can exert an anti-proliferative effect by virtue of cross-reactivity at sheep GnRH-R, but has no such effect when in cells expressing Xenopus GnRH-R, in spite of its higher affinity for the latter. All cloned non-mammalian GnRH-R express C-terminal tails whereas all cloned mammalian type I GnRH-Rs lack these structures and the absence of required phosphorylation sites in the C-terminal. tails of mammalian GnRH-Rs is thought to underlie their atypical resistance to desensitization and low rate of internalization. Accordingly, it appears that the anti-proliferative effect mediated by these receptors in mammary cancer cells, which may ultimately be of benefit for cancer therapy in vivo, is best achieved by activation of a GPCR which is resistant to desensitization and/or internalization. Thus, for the strategy outlined above, where specificity is achieved by expression of a non-mammalian GnRH-R it may be advantageous to modify the receptor, either by truncation of the C-terminal tail in order to remove phosphorylation sites or by point mutation to remove one or more of the phosphorylation sites within the C-terminal tail of the receptor. This strategy would, of course, be equally applicable to other GPCRs including somatostatin receptors. Agonist-induced phosphorylation is implicated in desensitization and internalization of several forms of the somatostatin receptor subtypes and the C-terminal tail of the sst3 receptor has been shown to be crucial for desensitization (Roth et al. 1997 J Biol Chem 272: 23769-74). Similarly, progressive C-terminal truncations of the sst5 have been shown to progressively influence receptor desensitization (Hukovic et al. 1998 J Biol Chem 273: 21416-22). It is therefore likely that modifications (truncation or mutation) which inhibit such phosphorylation will inhibit desensitization and internalization and thereby enhance the beneficial effects of such, somatostatin receptor targeted therapy. The strategy would also be applicable to other GPCRs, including those for which phosphorylated amino acids mediating desensitization and/or internalization are located in the third intracellular loop rather than the C-terminal tail. In this case the appropriate residues in the 3rd intracellular loop would be mutated to prevent their phosphorylation.

4. Expression of GPCR/G-protein fusion proteins GPCRs act primarily by activation of heterotrimeric G-proteins. These consist of a, P and y subunits and the a-subunits are thought to be the major mediators of most effects of GPCR activation. The stoichiometry of receptors to G-proteins is an important determinant of the efficiency of GPCR signalling and alteration in the amount of a-subunit available to the receptor may cause desensitization. Thus, sustained stimulation of GPCRs often causes a loss of G-protein from the activated cell. Different types of a-subunit activate different effector proteins selectively. Thus, a-subunits of the Gq/11 family mediate activation of the enzyme phospholipase C ß, whereas a-subunits of the Gs family mediate activation of adenyly cyclases. Many GPCRs have the potential to activate more than one type of a-subunit and thereby, to activate multiple effector proteins which may well regulate entirely different cellular activities. The molecular determinants of GPCR coupling to different G-proteins are poorly understood but again, the stoichiometry of active receptor protein molecules to activatable G-protein molecules is thought important.

In recent studies, fusion proteins have been engineered consisting of a GPCR sequence, coupled to a G-protein a-subunit sequence. When GPCR activating ligand is added to such fusion proteins the intrinsic a-subunit is activated. In several cases such fusion proteins have been found to show pharmacological characteristics similar to those of the wild-type GPCR but key distinguishing features are that stoichiometry of GPCR to a-subunit is always 1: 1 (e. g.. the a-subunit is not limiting) and responses mediated by the fusion protein may desensitise less than responses mediated by the corresponding unmodified GPCR. Accordingly, where a therapeutic benefit is achieved by gene therapy with a vector encoding a GPCR, the use of a fusion protein consisting of the GPCR and G-protein a-subunit could increase the therapeutic benefit. This would be particularly likely where the normal GPCR has the potential to activate more than one type of a-subunit and the therapeutic benefit reflects activation of a specific subunit, or where limited availability of the a-subunit limits the effect of GPCR activation.

We have shown that stimulation of phospholipase C activating GnRH-R inhibits the proliferation of mammary and prostate cancer cell lines in vitro, but it has been reported that GnRH-R are Gi coupled in endometrial cancer cells and inhibit the proliferation of these cells by activation of Gi. Accordingly, gene therapy with a vector encoding a fusion protein containing a GnRH-R and a G-protein of the aq/11 family might be beneficial for GnRH-R targeted therapy of mammary or prostate cancer, whereas a fusion protein consisting of the GnRH-R and a G-proteins of the ai family might be beneficial for treatment of endometrial cancer.

Recent work has revealed a novel mechanism for desensitization to somatostatin, in which phosphorylation of the Gai is thought to impair their re-cycling, effectively inhibiting signalling. by reducing the receptor's access to the G-protein (Murthy et al. 2000 Am J Physiol Cell Physiol 279: C925-34). Accordingly, gene therapy with a fusion protein containing a somatostatin receptor and a member of the ai family could ensure that the receptor signals via Gi and that the amount of activatable ai is not limiting. Thus, vectors encoding such chimeric G-proteins may provide further benefit for cancer therapy in which Gi coupled somatostatin receptors are targeted.

5. Use of viral vectors as Gene Delivery Systems Recombinant adenoviral vectors are prepared using established techniques (Geddes B. J. et al. (1997) Nature Medicine 3; 1402-1404; Harding T. C. et al. (1998) Nature Biotechnol 16 ; 553-555) and references therein. Recombinant adenovirus have already been shown to be efficient for gene transfection in breast and mammary cells and as efficient vectors for gene therapy of experimental cancers of these (Archer J. S. et al. (1994) Proc Natl Acad Sci USA 91; 6840-6844; Arteaga C. L. et al. (1996) Cancer Res 56; 1098-1103; Asgari K et al.

(1997) Ing. J Cancer 71 ; 377-382; Ficazzola M. A. and Taneja S. S., (1998) Mol Med Today 4; 494-504; Neilsen L. L. et al. (1998) Clin Cancer Res 4; 835-846 and other (Behbakht K. et al. (1996) Am J Obstet Gynecol 175 ; 1260-1265; Bramson J. L. et al.

(1997) Gene Ther 4; 1069-1076; Neilsen L. L. et al. (1998) Clin Cancer Res 4; 835-846; Rosenfeld M. E. et al. (1997) Clin Cancer Res 3; 1187-1194; Tong X. W. et al. (1997) Anticancer Res 17; 811-813 and Tong X. et al. (1998) Anticancer Res 18 ; 719-725 ; von Gruenige et al. (1998) Gynecol Oncol 69 ; 197-204) cells. Such gene therapy may be targeted to specific cells by the use of cell-specific promoter regions or may be made regulatable by the use of regulatable promoter regions. In vitro protocols for adenoviral infection are based on addition of recombinant adenovirus to the culture medium.

6. Assays (a) Radiolisand binding assays.

GnRH-R expression was assessed using whole cell binding assays in which approximately 50,000 cells were incubated in suspension for 30 min at 21 C in 100 ul of physiological salt solution (PSS: 127 mM NaCI, 1.8 mM CaCl2, 5 mM KCI, 2 mM MgC12, 0.5 mM NaH2PO4, 5 mM NaHCO3, 10 mM glucose, 0.1% BSA and 10 mM HEPES, pH 7.4) containing 1 mg/ml bacitracin with approximately 10-10 M radiolabel and 0 or 10-11 10-5 M of the unlabelled competitor peptide, as described (McArdle CA, et al (1992) Mol Cell Endocrinol 87,95-103; McArdle CA et al (1996) J. Biol. Chem. 271,23711-23717). Free and bound peptide were then separated by centrifugation through oil (McArdle CA et al (1992) supra). For human GnRH-R or sheep GnRH-R, the radiolabel was [125I] Buserelin.

For XenopusGnRH-R the radiolabel was [125I] cGnRH 11. For single point binding assays total and non-specific binding were determined by exclusion or inclusion, respectively, of the homologous competitor (Buserelin or cGnRH-11) at 10-7 or 10-6 M. For competition studies, non-linear regression (Graphpad Prism. Graphpad Software inc., San Diego, CA) was used to determine Kd and Bmax values, assuming that the tracer and competitor bind with identical affinity to a single class of receptor. Cell counts performed in parallel enabled calculation of the number of receptors per cell. Receptor internalization was quantified in a modified whole cell binding assay in which approximately 50,000 cells were grown in 24 well plates, were washed in PSS and then incubated at 37C in 200 ml PSS containing approximately 10-10 M radiolabel and 0 (total binding) or 10-6 M (non-specific binding) of Buserelin or cGnRH-11. After the required incubation period (2-120 min) the cells were rapidly rinsed in ice-cold PSS and then incubated for 2 min either in PSS or in 150mM NaCI with 50mM acetic acid (pH 3-4). The cells were then washed again in PSS and solubilized in 0.5 ml of 0.2 M NaOH with 1% SDS. Radiolabel in the solubilized cells was determined by y-counting and specific cell-associated radioactivity was determine by subtraction of non-specific from the total. Total specific binding is defined as the specific binding in cells receiving no acid wash, whereas acid-resistant (internalised) specific binding is defined as that seen in the acid washed cells.

For some of the binding experiments Hela cells expressing a dominant negative mutant of dynamin (K44A dynamin) were used. These cells (Viera AV et al (1996) Science 274, 2086-2089) express K44A dynamin under control of the tet-OFF system and were therefore cultured in medium with 1 mg/ml tetracycline in order to suppress K44A dynamin expression, or without tetracycline, in order to permit its expression (and thereby, to inhibit dynamin).

(b) Total [3H] inositol phosphate CL3ffllPxhaccumulation.

[3H] IPX accumulation was used as a measure of phospholipase C activity using cells labelled by pre-incubation with [3H] inositol and stimulated in the presence of LiCl (McArdle CA et al (1992) supra ; McArdle CA et al (1996) supra). Cells were cultured in 24-well plates in lml of media and 2 pCi [2-3H] inositol (14-16 Ci/mmol) was added to each well for the final 16 hr of incubation. After two washes in PSS, each well was stimulated for the indicated period in 200-250p1 of PSS containing 10 mM LiCl and the indicated concentration of stimulatory peptide. The stimuli were terminated by adding lml of water at 95°C. Inositol phosphates were then extracted and separated from free [3H] inositol using anion exchange chromatography (McArdle CA et al (1992) supra ; McArdle CA et al (1996) supra).

(c) [3Hlthymidine incorporation assays.

The methods used for assessment of [3H] thymidine incorporation into cancer cell lines were adapted from those previously used for pituitary cells (McArdle CA et al (1992) supra). Cells were plated in 96 well plates in culture medium with 10% foetal calf serum (FCS) at a density of approximately 5000 cells per ml. After 24 hours they were transferred to culture medium containing 1% FCS and incubated with test adenovirus at varied m. o. i. (3-300, or 0 in control cells). After 6 hours the cells were transferred to fresh medium (1% FCS) and incubated for approximately 24 hours before addition of test compounds. After a further 4-7 days, 0. 5Ci [3H] thymidine was added to each well and left to incorporate for 4 hours. The cells were then trypsinised in 1001 of trypsin and incubated at 37 C for 30 minutes. The cells were then frozen and thawed and incorporated [3H] thymidine was collected on filter papers using a TOMEC cell harvester and quantified by beta counting.

7. Preparation of Modified GnRH Receptor The methods for introduction of point mutations (replacing one amino acid with another), truncations (removal of all or part of a C-terminal tail) and for generation of chimeric receptors (C-terminal tail of one receptor added to a tail-less receptor, and replacement of a 3rd intracellular loop from one receptor with one from another) are all well established (Arora K. K. et al. 1995 and 1997; Davidson J. S. et al. (1996) J. Biol. Chem. 271; 15510-15514; HedingA. etal. (1998) J Biol. Chem. 273; 11472-11477; PawsonA. J. etal.

(1998) J. Endocrinol. 156, R9-12 ; Sealfon S. C. et al. (1997) Endocr. Rev. 18, 180-205 , and references therein).

Examples of nucleic acid sequences which may be modified are indicated below using their database reference numbers, together with the corresponding amino acid sequence and database reference number: SEQ. 1 : Sequence P 1-human GnRH-R protein sequence SEQ. 2 : Sequence NA1-human GnRH-R nucleic acid sequence SEQ. 3 : Sequence P2-cow GnRH-R protein sequence SEQ. 4 : Sequence NA2-cow GnRH-R nucleic acid sequence SEQ. 5 : Sequence P3-sheep GnRH-R protein sequence SEQ. 6 : Sequence NA3-sheep GnRH-R nucleic acid sequence SEQ. 7 : Sequence P4-pig GnRH-R protein sequence SEQ. 8 : Sequence NA4-pig GnRH-R nucleic acid sequence SEQ. 9 : Sequence P5-rat GnRH-R protein ° SEQ. 10 : Sequence NA5-rat GnRH-R nucleic acid sequence SEQ. 11 : Sequence P6-mouse GnRH-R protein sequence SEQ. 12 : Sequence NA6-mouse GnRH-R nucleic acid sequence SEQ. 13 : Sequence P7-catfish GnRH-R protein sequence SEQ. 14 : Sequence NA7-catfish GnRH-R nucleic acid sequence SEQ. 15 : Sequence P8-amphibian GnRH-R protein sequence SEQ. 16 : Sequence NA8-amphibian GnRH-R nucleic acid sequence SEQ. 17 : Sequence P9-goldfish GnRH-R A protein sequence SEQ. 18 : Sequence NA9-goldfish GnRH-R A nucleic acid sequence SEQ. 19 : Sequence P10-goldfish GnRH-R B protein sequence SEQ. 20 : Sequence NA10-goldfish GnRH-R B nucleic acid sequence SEQ. 21 : Sequence P11-Drosophila (fruitfly) GnRH-R protein sequence SEQ. 22 : Sequence NA11-Drosophila (fruitfly) GnRH-R nucleic acid sequence SEQ. 23: Sequence P12-Horse GnRH-R protein sequence SEQ. 24: Sequence NA12-Horse GnRH-R nucleic acid sequence All the above sequences are available on the Genbank data base, which can be accessed via the National Centre for Biotechnology Information (http ://www. ncbi. nlm. nih. gov/). The Genbank identifier numbers (prefixed gi) are shown at the beginning of each sequence shown in SEQ 1 to 24.

Amino acid sequences of a variety of GPCRs is provided in Fig. 17. Specifically in relation to the human somatostatin receptor 2-human Gai 1 fusion protein, in this construct the C-terminus of the complete human sst2 receptor is fused to the N-terminus of the complete human Gai 1 G-protein subunit (see also Milligan G 2000 Trends in Pharmacological Sciences 21: 24-28; Seifert et al (1999) Trends in Pharmacological Sciences 20: 383-389). The two unidentified amino acids (XX in bold) are included because it may be necessary to include a short linker to facilitate generation of such constructs (see Vorobiov et al (2000), J. Biol. Chem 275: 4166-4170) but also because the optimal length of C-terminal tail for the GPCR (for G-protein coupling and for effector activation by the G-protein) is not known and may therefore be altered by inclusion or omission of a linker sequence.

The above sequences were modified as detailed below: C-terminal truncations: Protein sequences (P7,8,9,10,11) and corresponding NA sequences (NA7,8,9,10,11) in which the C-terminal tails of non-mammalian GnRH-R are removed. The protein sequence is terminated to within a few amino acids of the C-terminal tail/7"trans-membrane region junction shown in fig. 3.

An example of a C-terminal truncation is shown below: C-terminal truncation (truncated goldfish A GnRH-R) MSDNTSLPSVSNASLLPPLTDWRAPSFTPAAQARVAATMVLFLFAAVSNLALLISVSRGR GRRLASHLRP LIISLVSADLMMTFIVMPLDMVWNVTVQWYAGDGLCKLLCFLKLFAMQTSAFILVVISLD RHHAILHPLD SLNAHQRNRRMLLLAWSLSALIASPQLFIFRTVKVKSVDFTQCVTHGSFHERWYETAYNM FHFVTLYVIP LLVMSCCYTCILIEINRQLHKSTEGESLRRSGTDMIPKARMKTLKMTIIIVLSFVVCWTP YYLLGIWYWF QPEMLKVTPEYIHHLLFVFGNLNTCCDPVIYGLYTP C-terminal modifications : This modified receptor involves protein sequences and corresponding NA sequences in which one or more of the serine residues and/or one or more of the threonine residues in the C-terminal tails of non-mammalian GnRH-R (P7,8,9,10,11 and corresponding NA sequences) were mutated to an amino acid which cannot be phosphorylated (e. g.. alanine).

This modification is intended to inhibit receptor desensitization.

An example of a C-terminal modification is shown below: C-terminal mutation (all serines of goldfish A GnRH-R replaced with alanines, underlined) MSDNTSLPSVSNASLLPPLTDWRAPSFTPAAQARVAATMVLFLFAAVSNLALLISVSRGR GRRLASHLRP LIISLVSADLMMTFIVMPLDMVWNVTVQWYAGDGLCKLLCFLKLFAMQTSAFILVVISLD RHHAILHPLD SLNAHQRNRRMLLLAWSLSALIASPQLFIFRTVKVKSVDFTQCVTHGSFHERWYETAYNM FHFVTLYVIP LLVMSCCYTCILIEINRQLHKSTEGESLRRSGTDMIPKARMKTLKMTIIIVLSFVVCWTP YYLLGIWYWF QPEMLKVTPEYIHHLLFVFGNLNTCCDPVIYGLYTPAFRADLARCWRCRTPAEAPRALDR IPHENTAPTRPA C-terminal chimeric receptors : This modified receptor involves protein sequences and corresponding NA sequences in which the C-terminal tail of a non-mammalian GnRH-R is added on to the C-terminus of a mammalian GnRH-R in order to increase the rate of internalization. For example: PI with C-terminal tail of P7, 8,9,10, or 11 P2 with C-terminal tail of P7, 8,9,10, or 11 P3 with C-terminal tail of P7, 8,9,10, or 11 P4 with C-terminal tail of P7, 8,9,10, or 11 P5 with C-terminal tail of P8, 9,10, or 11 P6 with C-terminal tail of P7, 8,9,10, or 11 An example of a C-terminal chimera is shown below: C-terminal chimera (sheep GnRH-R with C-terminal tail of goldfish GnRH-R A, underlined) MANGDSPDQNENHCSAINSSILLTPGSLPTLTLSGKIRVTVTFFLFLLSTIFNTSFLLKL QNWTQRKEKR KKLSKMKVLLKHLTLANLLETLIVMPLDGMWNITVQWYAGELLCKVLSYLKLFSMYAPAF MMVVISLDRS LAITRPLAVKSNSKLGQFMIGLAWLLSSIFAGPQLYIFGMIHLADDSGQTEGFSQCVTHC SFPQWWHQAF YNFFTFSCLFIIPLLIMLICNAKIIFTLTRVLHQDPHKLQLNQSKNNIPQARLRTLKMTV AFATSFTVCW TPYYVLGIWYWFDPDMVNRVSDPVNHFFFLFAFLNPCFDPLIYGYFSLSFRADLARCWRC RTPAESPRSL DRIPHENTSPTRPA 3rd intracellular loop chimeras: This modified construct comprises protein sequences (and corresponding NA sequences) of mammalian GnRH-R (P1-P6) in which the 3rd intracellular loop is replaced with one from another GPCR, in order to increase the potency or effectiveness of a ligand requiring internalization for action. The 3rd intracellular loop used could be that from an a2A adrenergic receptor or any other GPCR in which this domain mediates agonist-induced internalization of the receptor (and ligand).

An example of a 3rd intracellular loop chimera is shown below: 3rd intracellular loop chimera (sheep GnRH-R with 3"* intracellular loop substituted with that from guinea pig a2A receptor) MANGDSPDQNENHCSAINSSILLTPGSLPTLTLSGKIRVTVTFFLFLLSTIFNTSFLLKL QNWTQRKEKR KKLSKMKVLLKHLTLANLLETLIVMPLDGMWNITVQWYAGELLCKVLSYLKLFSMYAPAF MMVVISLDRS LAITRPLAVKSNSKLGQFMIGLAWLLSSIFAGPQLYIFGMIHLADDSGQTEGFSQCVTHC SFPQWWHQAF <BR> <BR> YNFFTFSCLFIIPLLIILVYVRIYQIAKRRTRVPPSRRGPDAHAAAPPGGAERRPNGLGL ERGVGPGGAE<BR> AEPLPTOVNGAPGEPAPAGPRDAEALDLEESSSSEHAERPPGARRPERGLRAKSKARASO VKPGDSLPRR<BR> APGAAGSGTSGSGPGEERGGGAKASRWRGRQNREKRFTFVLAVVIGIWYWFDPDMVNRVS DPVNHFFFLF AFLNPCFDPLIYGYFSL An example of a GPCR-a-subunit fusion protein.

Human somatostatin receptor 2-Human Gail fusion protein MDMADEPLNGSHTWLSIPFDLNGSVVSTNTSNQTEPYYDLTSNAVLTFIYFVVCIIG LCGNTLVIYVILRYAKMKTITNIYILNLAIADELFMLGLPFLAMQVALVHWPFGKAI CRVVMTVDGINQFTSIFCLTVMSIDRYLAVVHPIKSAKWRRPRTAKMITMAVWGV<B R> SLLVILPIMIYAGLRSNQWGRSSCTINWPGESGAWYTGFIIYTFILGFLVPLTIICLCYL FIIIKVKSSGIRVGSSKRKKSEKKVTRMVSIVVAVFIFCWLPFYIFNVSSVSMAISPTP ALKGMFDFVVVLTYANSCANPILYAFLSDNFKKSFQNVLCLVKVSGTDDGERSDS KQDKSRLNETTETQRTLLNGDLQTSIXXMGCTLSAEDKAAVERSKMIDRNLREDG <BR> <BR> EKAAREVKLLLLGAGESGKSTIVKOMKIIHEAGYSEEECKOYKAVVYSNTIQSIIAII&l t;BR> RAMGRLKIDFGDSARADDARQLFVLAGAAEEGFMTAELAGVIKRLWKDSGVQAC<BR > FNRSREYOLNDSAAYYLNDLDRIAQPNYIPTQQDVLRTRVKTTGIVETHFTFKDLH< BR> FKMFDVGGQRSERKKWIHCFEGVTAIIFCVALSDYDLVLAEDEEMNRMHESMKLF DSICNNKWFTDTSIILFLNKKDLFEEKIKKSPLTICYPEYAGSNTYEEAAAYIQCQFE DLNKRKDTKEIYTHFTCATDTKNVOFVFDAVTDVIIKNNLKD CGLF In this construct the C-terminus of the complete human sst2 receptor is fused to the N-terminus of the complete human GailG-protein subunit (see also Milligan G (2000) Trends in Pharmacological Sciences 21: 24-28; Seifert et al. (1999) Trends in Pharmacological Sciences 20: 383-389). The two unidentified amino acids (XX) are included because it may be necessary to include a short linker to facilitate generation of such constructs (see Vorobiov et al. (2000) J, Biol. Chem. 275: 4166-4170) but also because the optimal length of C-terminal tail for the GPCR (for G-protein coupling and for effector activation by the G-protein) is not known and may therefore be altered by inclusion or omission of a linker sequence.

8. Transfection of Pituitary Cells using Recombinant Virus containing a Functional GnRH-R Gene (Fig. 4A and B) Human mammary cancer-derived cells (MCF7) cells were maintained in culture and infected with recombinant adenovirus expressing the sheep GnRH-R (Ad sGnRH-R) at varied titre (m. o. i. = 0,10,30,100 or 300 as indicated). Inositol phospholipid pools were then radiolabeled by incubation in medium containing [3H] inositol before stimulation with the indicated concentration of GnRH in the presence of LiCl. The accumulation of [3H] IPX was then determined as a measure of phospholipase C activity. These data demonstrate that the binding sites expressed after infection of human mammary cancer cells with recombinant adenovirus expressing GnRH-R are functional GnRH-R. Also shown in Hela cells and PC3 cells. aT4 cells provided by Prof. Pamela Mellon (University of California at San Diego, Department of Reproductive Medicine, La Jolla, CA, USA), a gonadotroph progenitor cell line which does not have endogenous GnRH receptors, were cultured for 24 hours with recombinant adenovirus encoding a mammalian (sheep) GnRH-R at 0.3-10 x 106 pfu/ml (equivalent to multiplicities of infection, m. o. i., of 3-100). The infected cells were then used for assessment of GnRH-stimulated [3H] IPx accumulation (where x = 1,2,3 etc.) ([3H] inositol-labelled cells, stimulated for 30 min. in the presence of 10 mM LiCI) as a measure of GnRH receptor activity (phospholipase C activation). GnRH receptor number, measured under similar conditions, was increased over a 70 fold dynamic range from sub-physiological to super-physiological levels (not shown).

Fig. 4B demonstrates that infection with recombinant adenovirus expressing GnRH-R can be used to increase the responsiveness of these cells to an extracellular ligand which acts via GnRH-R and therefore implies that infection with the adenovirus causes the cells to express functional GnRH-R on the cell surface. It also shows that, by varying viral titre (and thereby varying the number of GnRH-R expressed), the efficacy and potency of GnRH can both be controlled. Recombinant adenovirus have also been prepared in a similar manner using human, mouse and Xenopus GnRH-R. Highly efficient transfection and expression of GnRH-R DNA was achieved, resulting in the production of GnRH-R protein which is functional and has pharmacology similar to that of native GnRH-R in these pituitary cells.

9. Demonstration that recombinant adenovirus can be used to express GnRH-R in mammary cancer cell lines.

Two separate human breast cancer-derived cell lines (MCF-7 (MCF7-86012803) and T47D (T47-D-85102201)), obtained from the European Collection of Animal Cell Cultures (http ://fuseii. star. co. uk/camr/) were maintained in culture and infected with recombinant adenovirus expressing sheep GnRH-R at the indicated m. o. i. (See Figs 3A, 3B and 3C) After a further 24 hours in culture they were used in an established whole cell binding assay (McArdle et al. (1992) Endocrinology 130,3567-3574) using ['zSIJBuserelin in the presence (non-specific binding) and absence (total binding) of 1 uM buserelin. The difference between the total and non-specific binding (e. g.. specific binding) is proportional to receptor number in these assays. These specific binding sites are GnRH-R because they mediate appropriate responses to receptor-activating peptides (above) and, in competition binding assays, have similar affinity and pharmacology to the endogenous GnRH-R of aT3-1 cells, Thus, infection with this recombinant adenovirus causes these cells to express GnRH-R at a density dependent upon the amount of adenovirus used and enables GnRH-R expression levels well above any endogenous GnRH-R in these cells. Fig. 3C Human mammary cancer-derived cells (MCF7) cells were maintained in culture and infected with recombinant adenovirus expressing the sheep GnRH-R (Ad sGnRH-R) at varied titre (multiplicity of infection, m. o. i. = 0,10,30,100 or 300 as indicated). They were then used in whole cells radioligand binding studies in which the binding of [25I] Buserelin (a GnRH-R specific agonists) was competed for by unlabelled Buserelin. Specific binding of the radiolabel was not observed in control cell (m. o. i. = 0) but increasing viral titre increased radioligand binding. Statistical analysis of the competition curves revealed high affinity binding sites with Kd values of approx. 1.4 nM. Refitting the data to this pooled Kd revealed that increasing viral titre increased receptor number to over 200,000 sites per cell (Fig. 3C, inset). These data demonstrate that infection of human mammary cancer cells with recombinant adenovirus expressing GnRH-R provides a means of expressing high affinity [125I] Buserelin binding sites in these cells and that varying viral titre provides a means of controlling the number of binding sites expressed. Similar results have been obtained in 7 other cancer cell lines (MCF7 cells from a separate source, T47D mammary cancer cells, PC3 prostate cancer cells, OVCAR ovarian cancer cells, endometrial cancer cells, Hela cervical cancer cells, aT4 pituitary cancer cells).

10. Demonstration that recombinant adenovirus expressing GnRH-R from different species (human and xenopus) can be used to transfect pituitary cells and mammary cancer cells and that the differences in pharmacology of these receptors seen in pituitary cells are also seen in mammary cancer cells. aT4 cells (pituitary gonadotroph progenitor cells) (Fig. SB, D) and MCF-7 (human mammary cancer cells (Fig. SA, C)) were infected with recombinant adenovirus expressing human (Fig. SA, B) or Xenopus (Fig. 5C, D) (African clawed toad) GnRH-R at an m. o. i. of 100. They were then labelled with [3H] inositol and used for measurement of [3H] IPx accumulation using standard methods (McArdle et al (1992) supra) during a 45 min. stimulation with the indicated concentrations of GnRH, chicken II GnRH and Buserelin (See Fig. 5). GnRH and the GnRH agonist, Buserelin, are more potent than chicken II GnRH at the human GnRH-R, whereas chicken II GnRH is more potent than the other two peptides at the Xenopus GnRH-R. Similar order of potencies have been seen in radio-ligand binding studies. Cells (MCF7) cells were maintained in culture and infected with recombinant adenovirus expressing the sheep GnRH-R (Ad sGnRH-R) or the type I Xenopus laevis GnRH-R (Ad XGnRH-R) at an m. o. i. of 100. They were then used in whole cells radioligand binding studies. For cells infected with Ad sGnRH-R, the binding of [25I] Buserelin (a mammalian GnRH-R specific agonists) was competed for by unlabelled Buserelin, GnRH or chicken GnRH-11 (cGnRH-II) as indicated. For cells infected with Ad XGnRH-R, the binding of [1251] cGnRH-11 was competed for by the same unlabeled ligands. Specific binding was not seen for either radiolabel in control cells (m. o. i. = 0). The three peptides competed for [125I] Buserelin binding to the sGnRH-R with the rank-order-of-potency expected for a mammalian GnRH-R (Buserelin>GnRH>cGnRH-II).) whereas the three peptides competed for [125I] cGnRH-II binding to the XGnRH-R with the rank-order-of-potency expected for a non-mammalian mammalian GnRH-R (cGnRH-II>Buserelin>GnRH>cGnRH-II). [3H] IPX accumulation studies performed in parallel revealed that all 3 peptides could activate phospholipase C via both receptors (not shown). In these assays the rank orders of potency for the 3 peptides were identical to those seen in the binding assays. Similar results were obtained when these receptors were exp [ressed by adenovirus infection in aT4 pituitary cancer cells and Hela cervical cancer cells (not shown). These data establish that infection with recombinant adenovirus expressing mammalian and non-mammalian GnRH-R can be used to express functional GnRH in mammary cancer cells and that these receptors retain the pharmacological characteristics of their endogenous counterparts. This difference in pharmacology was predicted from the known pharmacology of the endogenous receptors in mammalian and Xenopus pituitaries. It is important that these differences are also retained within the context of breast cancer cells because it provides a potential means of selectively activating GnRH-R in vivo. Thus, 10-9 M chicken II GnRH would appear to near maximally activate Xenopus GnRH-R without appreciably activating human GnRH-R.

Achieving this in vivo, may allow activation of Xenopus GnRH-R transfected into cancer cells without activating human GnRH in the pituitary-thereby achieving a direct cytotoxic effect on the cancer without the chemical contraception inherent in current therapies with GnRH. The data shown are pooled from 3 separate experiments (mean SEM, n=3), each having duplicate observations, after normalisation as a % of the internal control maximal response (typically that to 10-'-10-6 M Buserelin in cells transfected with hGnRH-R).

11. Demonstration that recombinant adenovirus expressing GnRH-R can be used to facilitate or potentiate an antiproliferative effect of GnRH on human mammary cancer cell lines.

Incorporation of [3H] thymidine into newly synthesised DNA was used as a standard assay for cell proliferation (McArdle et al. 1992; Williams et al. 1999 supra). T47D (Fig. 6B) or MCF-7 (Fig. 6A) cells had been infected either without adenovirus, or with recombinant adenovirus expressing the sheep GnRH-R at an m. o. i. of 30 or 100, and then incubated for 5-6 days in medium with the indicated concentration of GnRH before measurement of [3H] thymidine incorporation on the last day of culture. The proliferation of both of these cells lines has been reported to be inhibited by GnRH but this effect. is modest and is apparently not seen with all clones of these cell lines. As shown in Fig. 6, GnRH did not inhibit the proliferation of control MCF-7 cells but did cause a modest dose-dependent inhibition of the proliferation of T47D. However, when the cells are infected with recombinant adenovirus expressing GnRH-R, GnRH was able to inhibit proliferation of MCF-7 cells, and the inhibitory effect of GnRH was enhanced in T47D cells. GnRH (and agonist or antagonist analogues of GnRH) have been shown to inhibit the proliferation of many cells in vivo and in vitro (Segal et al. 1992; Emons et al. 1993a, 1993b, 1998; Emons and Schally 1994; Jungwirth et al. 1997 supra). The importance of these data is that they show that gene therapy with GnRH-R can be used to increase antiproliferative effects of GnRH on GnRH-responsive cancer cells, and to enable GnRH to do so even in cancer cells which are otherwise not responsive to GnRH. The implication is that gene therapy with GnRH-R may enable GnRH or GnRH analogues to exert these effects in vivo. The data also demonstrate the more general principle that gene therapy with a vector encoding GPCR can be used to increase the sensitivity of breast cancer cells to other antiproliferative treatments.

12. Demonstration that expression of GnRH-R in mammary tumour cells increases the anti-proliferative effect of GnRH and of GnRH analogues. (Fig. 7A, B and C and Fig. 8) Figs. 7A and Fig. 7B shows MCF-7 cells were infected in culture with recombinant adenovirus expressing the sheep GnRH-R, or with control virus expressing no additional protein, at a m. o. i. of 100 or 0. The cells, together with control MCF-7 cells not infected with adenovirus, were then further incubated for 5 or 6 days in medium with the indicated concentration of GnRH (mammalian) or of the agonist analogue, [D-Lys6] GnRH (Sigma-Aldrich Co. Ltd, Poole, Dorset). [3H] thymidine incorporation into newly synthesised DNA was then determined as a measure of the rate of cell proliferation. None of the peptides had a pronounced effect on the proliferation of control cells or of cells infected with the empty adenovirus whereas both peptides caused a pronounced and dose-dependent inhibition of proliferation in cells transfected with the GnRH-R. Figs. 7A and B show that infection with adenovirus expressing the sheep GnRH-R can be used to enable GnRH to inhibit the proliferation of human breast cancer cells, and that this effect is specific, in that it is mimicked by. another activator of GnRH receptors ([D-Lys6] GnRH) and is not seen in cells infected with the empty adenovirus (e. g.. the effect requires GnRH-R expression and is not simply a non-specific consequence of exposure to adenovirus per se). Fig. 7C shows incorporation of [3H] thymidine into newly synthesised DNA was used as a standard assay for cell proliferation (McArdle et al. 1992; Williams et al. 1999). The MCF-7 cells had been infected either without adenovirus, or with recombinant adenovirus expressing the sheep GnRH-R at an m. o. i. of 30 or 100, and then incubated for 5-6 days in medium with the indicated concentration of GnRH, Buserelin or- cGnRH-II before measurement of [3H] thymidine incorporation on the last day of culture.

The proliferation of both of this cell line has been reported to be inhibited by GnRH but this effect is modest and is apparently not seen with all clones of MCF7 cells. In our hands, none of these peptides inhibited the proliferation of control MCF-7 cells (data shown only for Buserelin). However, when the cells are infected with Ad sGnRH-R, All three peptides caused a concentration-dependent inhibition of proliferation. These data demonstrate that infection with recombinant adenovirus expressing GnRH-R can enable GnRH-R activating ligands to exert an anti-proliferative effect in mammary cancer cells.

They also demonstrate that the effect is mediated by the expressed receptor and that pharmacological characteristics of this response are as predicted by binding and [3H] IPX accumulation assays (rank order of potency: Buserelin > GnRH > cGnRH-11). We have observed similar effects in PC3 prostate cancer cells.

This confirms that infection with recombinant adenovirus expressing GnRH-R can be used to induce an antiproliferative effect of GnRH in breast cancer cells and extends this by showing that this holds true for agonist analogues of GnRH. The fact that no such effect is seen in cells infected with empty adenovirus demonstrates that the antiproliferative effect is achieved by virtue of GnRH-R expression, rather than by any non-specific"sensitising" effect of the adenovirus itself. Thus, by adenovirus-mediated transfection of GnRH-R, it is possible to facilitate an anti-proliferative effect of GnRH and GnRH analogues, in breast cancer cell lines.

Therefore, by delivering DNA encoding GnRH-R to cancer cells, increased expression of the receptor protein is achieved, and therefore increased efficiency of therapies targeting these cells via the GnRH-R is possible.

In Fig. 8, cells were plated, incubated, infected with Ad GnRH-R and used for assessment of [3H] thymidine incorporation (on the last day of culture) as described under Fig. 7 except that the cells were cultured in the presence of the indicated concentrations of Buserelin in 'medium with the indicated concentration of Antide. Antide alone did not measurably influence [3H] thymidine incorporation and pooling was achieved by calculating the % inhibition of [3H] thymidine incorporation caused by each concentration of Buserelin.

Similar results were obtained when Cetrorelix was used in place of Antide. This data demonstrates that Antide and Cetrorelix, two peptides which acts as a competitive antagonist at pituitary GnRH-Rs also do so at GnRH-R expressed using recombinant Ad in mammary cancer cells, re-enforcing the conclusion that the receptors expressed in this way retain the pharmacological characteristics of endogenous pituitary GnRH-Rs.

13. Demonstration that infection with recombinant adenovirus expressing GnRH receptors can be used to affect a receptor mediated antiproliferative effect of a cytotoxic-derivative of an agonist analogue of GnRH.

MCF-7 cells were infected in culture with recombinant adenovirus expressing the sheep GnRH-R at a m. o. i. of 100 or with no adenovirus, as indicated. They were then further incubated for 5 or 6 days in medium with no added peptide (control) or with an agonist analogue of GnRH (peptide) or a cytotoxic-derivative of the agonist analogue (peptide-cytotoxin), each at 10-'° M, as indicated. [3H] thymidine incorporation into newly synthesised DNA was then determined as a measure of the rate of cell proliferation, and is shown, in Fig. 9, expressed as a percentage of that seen without peptide. As shown, the agonist analogue had no measurable effect on [3H] thymidine incorporation (at this concentration) irrespective of whether or not the cells had been transfected with GnRH-R.

The cytotoxic conjugate was ineffective in control cells but dramatically inhibited proliferation in cells transfected with GnRH-R. The use of low concentrations of peptides in this experiment (e. g.. concentrations which are ineffective alone) demonstrates that the effect is not just due to agonist activation of GnRH-R (e. g.. the effect is due to receptor-mediated internalization or binding of toxin, rather than receptor activation per se) and implies that this strategy may be used to improve the effectiveness and potency of cytotoxic GnRH derivatives in vivo. The results shown in Fig. 9 establish that infection with recombinant adenovirus expressing GnRH-R can be used to make breast cancer cells sensitive to the antiproliferative effect of a cytotoxic conjugate of a GnRH analogue.

14. Demonstration that human and Xenopus GnRH-R, transfected into aT4 cells by infection with recombinant adenovirus desensitised internalise differentially. (Fig.

10A and B) In Fig. 10A, aT4 cells were infected in culture with Ad hGnRH Ad (open symbols) or with Ad XGnRH-R (filled symbols), each at an m. o. i. of 100-300, and then cultured for 1-2 days. [3H] inositol was added to the medium for the final 16 h of culture, after which the cells were washed and stimulated for the indicated time in medium containing 10 mM LiCl and 10-7 M GnRH (hGnRH-R) or cGnRH-11 (XGnRH-R). [3H] IPX data were expressed as fold increase over basal values. Linear regression analysis revealed comparable initial rates of [3H] IPX accumulation (0.18 and 0.19 fold basal/min during 0-2 min) for XGnRH-R and hGnRH-R respectively, whereas the final rate of [3HIPX accumulation in XGnRH-R expressing cells was only 30% of that in hGnRH-R expressing cells (0.08 and 0.28 fold basal/min respectively during 2-15 min). A radioligand on the cell surface is acid labile whereas that internalised into the cell is acid-resistant. Fig. 10B shows specific acid resistant binding as a % of total specific binding (e. g.. proportional internalization of receptors) after incubation for the indicated period in the presence of'25I-labelled Buserelin or chicken II GnRH in aT4 cells transfected (by infection with recombinant AdV) with human or Xenopus GnRH-R. As shown, radioligand is internalised more rapidly via the Xenopus GnRH-R, implying that this receptor is internalised more rapidly than the human.

Similar comparisons have been carried out using catfish and rat GnRH-R in COS and HEK cells (Heding et al (1998) J. Biol. Chem 273; 11472-11477). The implication is that these differences may be exploited therapeutically. Thus a drug treatment dependent upon receptor activation might be more effective when targeted via a receptor which does not desensitize, whereas a treatment dependent upon receptor-mediated internalization might be more effective when targeted via a rapidly internalised receptor. Since desensitization and internalization of G-protein coupled receptors is often mediated by phosphorylation of serine residues within the C-terminal tail of the receptor, these functional differences most probably reflect the fact that human GnRH-R (like all known type I mammalian GnRH-R) lack C-terminal tails, whereas Xenopus GnRH-R (like all known non-mammalian GnRH-R) possess C-terminal tails with containing serine residues. This raises the possibility that the potency and/or efficacy of drug therapies requiring internalization of cytotoxins via mammalian GnRH-R may be improved by adding C-terminal tails from non-mammalian GnRH-R. Similarly, drug therapies requiring activation of non-mammalian GnRH-R may be improved by truncating the tails or by mutating serines and threonines within the tails (or other intracellular loops) into amino acids which are not phosphorylated.

15. Demonstration that inhibition of mammary cancer cell line proliferation is facilitated by expression and activation of a desensitization resistant GnRH-R but not by expression and activation of a desensitising GnRH-R Incorporation of [3H] thymidine into newly synthesised DNA was used as a standard assay for cell proliferation as described under Figure 7 in MCF7 cells which had been infected with Ad sGnRH-R or with Ad XGnRH-R (both at m. o. i. of 30-100). They were then incubated for 5-6 days in medium with the indicated concentration of GnRH. Buserelin or cGNRH-II before measurement of [3H] thymidine incorporation on the last day of culture.

As expected (Fig. 7) all three peptides inhibited proliferation of the Ad sGnRH-R infected cells but none significantly inhibited proliferation of Ad XGn-RH-R infected cells, in spite of the fact that receptor expression levels were greater in the Ad XGnRH-R infected cells (not shown). These data imply that cancer cell proliferation is more effectively inhibited by activation of non-desensitizing receptors (sGnRH-R) than by activation of desensitising receptors (XGnRH-R).

16. Demonstration that infection with recombinant adenovirus expressing human GnRH increases the anti-proliferative effect of GnRH-R activation in the absence and presence of oestrogen receptor activation.

In Fig. 12A, MCF-7 cells were infected with recombinant adenovirus expressing human GnRH-R or with empty adenovirus, each at an m. o. i. of 100, and then incubated in medium containing an oestrogen receptor antagonist (E antag-raloxifene) at 0 (control) or 10-'M as well as 0 (open bars) or 10-8 M (filled bars) Buserelin. Since the culture medium contains serum with endogenous oestrogens, the plus/minus raloxifene culture conditions represent plus/minus oestrogen receptor activation. The data were normalised as a % of [3H] thymidine incorporation in the control groups receiving no Buserelin. As shown, Buserelin inhibited cell proliferation ([3H] thymidine incorporation) only after infection with Ad GnRH-R and irrespective of whether or not oestrogen receptors were activated.

Similar studies were performed using MCF7 cells infected with Ad sGnRH-R (m. o. i 100) (Fig. 12B) and then incubated with medium containing 10-g M raloxifen and the indicated concentration of 17p-oestradiol and 0 (control) or 10-'M Buserelin. Raloxifen is a competitive antagonist at oestrogen receptors, so it can be included to block effects of endogenous oestrogenic compounds, but the inhibitory effect can be overcome by estradiol and this strategy can therefore be used to reveal the full extent of the effect of estradiol.

The data, which are normalised as a % of [3H] thymidine incorporation on control cells receiving no raloxifen, estradiol or Buserelin, revealed that GnRH-R mediated inhibitory effect of Buserelin occurs, irrespective of the extent of oestrogen-stimulated proliferation.

The implication is that it may be possible to use this strategy to achieve a direct inhibition of tumour cell proliferation in vivo, irrespective of whether or not concomitant chemical contraception is achieved.

17. Demonstration that mammalian and non-mammalian GnRH-R can access functionally distinct routes of internalization, differing in sensitivity to K44A dynamin (Figure 15A, B) K44A dynamin Hela cells cultured in 24 wells in the presence of tetracycline were infected with Ad hGnRH-R (Fig. 15A) or Ad XGnRH-R (Fig. 15B) at m. o. i. values of 50-100.

HeLa (K44A) cells were then cultured with (open symbols) or without (closed symbols) tetracycline for 8 hours, to allow infection, before the medium was changed to fresh containing no Ad with or without tetracycline to allow expression of the mutant dynamin, and then cultured for a further 16 hours. Cells were washed and incubated for the indicated period of time in PSS containing approximately 0.25nM [125I] Buserelin (hGnRH-R) or 0.125nM [1251] cGnRH 11 (XGnRE-R) with 0 or 10-6M Buserelin or cGnRH II (non-specific binding). Transfer to ice-cold PSS and washing in PSS (total binding) or acid (internal binding) terminated the binding. The data shown are mean SEMs (n=3) of acid resistant binding expressed as a percentage of total specific binding pooled from repeated experiments, each having duplicate observations. This data reveals that the Xenopus GnRH-R can be internalised via a dynamin-dependent pathway which is inaccessible to the human GnRH-R, implying that this is why the Xenopus GnRH-R is internalised more rapidly than the human receptor; and 18. Demonstration that modification of a mammalian GnRH-R can enable it to access an additional (K44A dynamin-sensitive) route of receptor internalization (Figure 16) Dynamin-dependence of receptor internalization was determined as described under Fig.

15, except that the receptors used were the human GnRH-R, the Xenopus GnRH-R and a chimeric protein consisting of the entire human GnRH-R sequence with an added C-terminal tail from the Xenopus GnRH-R. As expected, internalization of the Xenopus receptor was largely dynamin-dependent (blocked by K44A dynamin expression) whereas internalization of the human GnRH-R was not. After addition of the Xenopus C-terminal tail, however, internalization of the human GnRH-R was largely dynamin-dependent. This data reveals that these receptors access two functionally distinct routes of internalization and that the C-terminal tail is required for these receptors to access the dynamin-dependent route. The implication is that the lack of C-terminal tail in mammalian GnRH-R will prevent them from accessing the dynamin-dependent internalization route and that this underlies there slow rate of internalization.

Indirect effects of GnRH agonists and antagonists (e. g.. those mediated by'chemical castration') will only work on cancers whose proliferation is dependent upon gonadal steroids. Such cancers often lose steroid-dependence over time so that the indirect effect of GnRH is ineffective. The direct antiproliferative effect can only work on the subset of cancers which express GnRH receptors. Reducing circulating steroids has numerous health/patient compliance problems which might be circumvented by gene therapy with a GnRH-R having different pharmacology to that of pituitary GnRH-R. GnRH-R gene transfer could therefore provide advantages in terms of improved potency, selectivity and range of cancers treatable.

The current mainstays of cancer therapy are surgery, radiation, chemotherapy and biological therapy, and it is likely that gene therapy will be increasingly important in the future. These therapies are often most effectively used in combination. For example, radiation therapy may be used to reduce tumour size before surgery and chemotherapy may be used to treat metastatic cancers after removal of the original tumour by surgery. GnRH analogues have already been shown to be useful in treatment of metastasis (Gonzalez-Barcena D et al, (1995) Urology 45: 275-281; Sanchez-Garrido F et al, (1995) Lancet 345: 868) and to increase the antiproliferative effect of cis-platinum on breast and ovarian cancer cell lines (Ohta H et al, (1998). Therefore, gene therapy with vector encoding a GPCR may be used in conjunction with other forms of therapy, for example, to reduce tumour size before surgery or to sensitise tumour cells to chemotherapy or radiation therapy, thereby reducing the required dosage and side effects of the conventional therapies. Moreover, if exogenous GPCRs were expressed in cancer cells before metastases, this may enable GPCR targeted therapies (including those based upon the use of cytotoxic derivatives) to be used to treat the metastatic cancer.

Gene therapy which increases expression of functional GnRH-R in cancer cells can be used to sensitise these cells to biological therapy targeted via those receptors. Thus, for example, the recombinant adenovirus expressing GnRH-R may be used as a sensitising agent to increase the sensitivity of the tumours to biological therapy. This principle may, of course, be extended to other GPCRs which can have direct beneficial effects on cancer cells but for which usefulness is limited by low receptor number, or undesired side-effects on receptors in other cells and tissues. Extension of the concept to other GPCRs is particularly relevant for GPCRs which are activated by peptides or polypeptides because a) numerous cytotoxic derivatives of such polypeptide ligands have been generated and b) co-evolution of peptides and receptors has ensured that non-mammalian homologues of mammalian GPCRs are often activated selectively by non-mammalian peptides, providing the potential to specifically target exogenously expressed receptors without influencing the endogenous mammalian receptors.