Membrane span-kinase fusion protein and the uses thereof

Abstract

The disclosure relates to a recombinant membrane span protein complex, comprising (1) a fusion protein, comprising a membrane span protein fused to a kinase domain, preferably a constitutive kinase and (2) a reporter construct comprising a polypeptide, interacting with the membrane span protein, fused to a reporter phosphorylation domain. The disclosure relates further to the uses of such membrane span protein complex for the detection of compounds that interact with the membrane span protein and for the screening and/or detection of inhibitors of the compound-membrane span protein interactions. In a preferred embodiment, the membrane span protein is a G protein coupled receptor (GPCR) and the method is used for the screening and/or detection of inhibitors of the ligand-receptor binding.

Claims

1. A method for detecting a compound-protein interaction, the method comprising: (a) contacting a eukaryotic cell with a compound, the eukaryotic cell comprising a recombinant protein complex comprising: (i) a first fusion protein comprising a membrane spanning domain fused to a Tky2 tyrosine kinase domain and (ii) a second fusion protein comprising an interaction domain fused to a gp130 reporter phosphorylation domain, wherein the compound mediates the formation of the recombinant protein complex, and wherein the tyrosine kinase domain phosphorylates a tyrosine of the reporter phosphorylation domain upon the formation of the recombinant protein complex, and (b) measuring the phosphorylation of the reporter phosphorylation domain.

2. The method according to claim 1, wherein the tyrosine kinase domain is a mutant tyrosine kinase domain.

3. The method according to claim 2, wherein the mutant tyrosine kinase domain is a constitutively active mutant kinase.

4. The method according to claim 2, wherein the mutant tyrosine kinase domain is an inactive mutant that is activated by addition of an exogenous small molecule.

5. The method according to claim 1, wherein the tyrosine kinase domain is fused to the carboxyterminal end of the membrane spanning domain in the first fusion protein.

6. The method according to claim 1, wherein the membrane spanning domain is a multispan membrane span protein.

7. The method according to claim 6, wherein the multispan membrane span protein is a G protein coupled receptor.

8. The method according to claim 1, wherein the compound is a small molecule.

9. The method according to claim 1, wherein measuring the phosphorylation of the reporter phosphorylation domain is measured indirectly by detection of a reporter gene that is activated by phosphorylation of the reporter phosphorylation domain.

10. The method according to claim 9 wherein the reporter gene is a luciferase gene.

11. The method according to claim 1, wherein phosphorylation of the reporter phosphorylation domain is measured directly.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

(1) FIG. 1: Schematic representation of different embodiments of the recombinant membrane span protein complex, according to the disclosure. “M” depicts a membrane.

(2) A. A membrane span protein (X) is fused to a constitutive kinase (K) and a polypeptide (Y) is fused to a reporter phosphorylation site (R). Interaction between the membrane span protein X and the polypeptide Y results in the reporter phosphorylation site being phosphorylated (P) by the constitutive kinase, leading to a detectable activity.

(3) B. A membrane span protein (X) is fused to a reporter phosphorylation site (R) and a polypeptide (Y) is fused to a constitutive kinase (K). Interaction between the membrane span protein X and the polypeptide Y results in the reporter phosphorylation site being phosphorylated (P) by the constitutive kinase, leading to a detectable activity.

(4) C. A membrane span protein (X) is fused to a constitutive kinase (K) and a second membrane span protein (Y) is fused to a reporter phosphorylation site (R). Interaction between the membrane span proteins X and Y results in the reporter phosphorylation site being phosphorylated (P) by the constitutive kinase, leading to a detectable activity.

(5) FIG. 2: Detection of the ligand-dependent interaction between human somatostatin receptor 2 (SSTR2) and human beta arrestin 2 (ARRB2) in an assay variant that comprises mutant Tyk2 kinase fusion proteins.

(6) A. Schematic overview of the assay. The membrane span protein (X) is fused to the C-terminal region of Tyk2 comprising the kinase domain, whereas the polypeptide interacting with the membrane span protein (Y) is fused to a fragment of gp130 which contains phosphorylation sites. When membrane span protein X and the polypeptide Y interact, the Tyk2 kinase domain phosphorylates the phosphorylation sites of gp130. STAT3 transcription factors are recruited to these phosphorylated sites and are in turn phosphorylated by the Tyk2 kinase domain, which results in their activation. Dimers of activated STAT3 transcription factors are able to bind the specific rPAPI promoter, which drives the expression of a firefly luciferase reporter gene. The activity of this reporter gene is measured as light production in a luciferase detection assay using a luminometer.

(7) B. Application to the analysis of ligand-dependent interaction between SSTR2 and ARRB2. Cells were transfected with the indicated combination of plasmids, and either left untreated (NS) or treated with increasing doses (0.1-1-10 μM) of somatostatin:

(8) a) pMet7-HA-Tyk2(C)+pMG2-ARRB2+pXP2d2-rPAPI-luciferase

(9) b) pMet7-SSTR2-Tyk2(C)−HA+pMG2-SVT+pXP2d2-rPAPI-luciferase

(10) c) pMet7-SSTR2-Tyk2(C)−HA+pMG2-ARRB2+pXP2d2-rPAPI-luciferase

(11) Luciferase activity is shown as fold induction relative to the luciferase activity measured in untreated cells. Error bars indicate standard deviation.

(12) C. Detection of the ligand-dependent interaction between SSTR2 and ARRB2 using an alternative expression vector. Cells were transfected with the indicated combination of plasmids, and either left untreated (NS) or treated with increasing doses (0.1-1-10 μM) of somatostatin:

(13) a) pSVSport-HA-Tyk2(C)+pMG2-ARRB2+pXP2d2-rPAPI-luciferase

(14) b) pSVSport-SSTR2-Tyk2(C)−HA+pMG2-SVT+pXP2d2-rPAPI-luciferase

(15) c) pSVSport-SSTR2-Tyk2(C)−HA+pMG2-ARRB2+pXP2d2-rPAPI-luciferase

(16) Luciferase activity is shown as fold induction relative to the luciferase activity measured in untreated cells. Error bars indicate standard deviation.

(17) D. Dose-response curve of the ligand-dependent interaction between SSTR2 and ARRB2. Cells were transfected with a combination of the plasmids pMet7-SSTR2-Tyk2(C)−HA, pMG2-ARRB2 and pXP2d2-rPAPI-luciferase, and treated with increasing concentrations of somatostatin (SST-14). Luciferase activity is shown as relative light units (rlu). Error bars indicate standard deviation.

(18) FIG. 3: Analysis of the interaction between human angiotensin receptor 1 (AGTR1) and ARRB2.

(19) A. Detection of the ligand-dependent interaction between AGTR1 and ARRB2. Cells were transfected with the indicated combination of plasmids, and either left untreated (NS) or treated with increasing doses (0.1-1-10 μM) of angiotensin II:

(20) a) pMet7-HA-Tyk2(C)+pMG2-ARRB2+pXP2d2-rPAPI-luciferase

(21) b) pMet7-AGTR1-Tyk2(C)−HA+pMG2-SVT+pXP2d2-rPAPI-luciferase

(22) c) pMet7-AGTR1-Tyk2(C)−HA+pMG2-ARRB2+pXP2d2-rPAPI-luciferase

(23) Luciferase activity is shown as fold induction relative to the luciferase activity measured in untreated cells. Error bars indicate standard deviation.

(24) B. Dose-response curve of the ligand-dependent interaction between AGTR1 and ARRB2. Cells were transfected with a combination of the plasmids pMet7-AGTR1-Tyk2(C)−HA, pMG2-ARRB2 and pXP2d2-rPAPI-luciferase, and treated with increasing concentrations of angiotensin II (AngII). Luciferase activity is shown as relative light units (rlu). Error bars indicate standard deviation.

(25) FIG. 4: Evaluation of the effect of GPCR antagonists on the interaction between GPCRs and ARRB2. Cells were transfected with the indicated combination of plasmids, and treated with the indicated combinations of GPCR ligand and antagonist (ligand: 1 μM somatostatin for transfections a and b, 10 μM angiotensin II for transfections c and d; antagonists: 0.05 or 0.5 μM CYN154806; 0.1 or 1 μM losartan or telmisartan):

(26) a) pMet7-SSTR2-Tyk2(C)−HA+pMG1-EFHA1+pXP2d2-rPAPI-luciferase

(27) b) pMet7-SSTR2-Tyk2(C)−HA+pMG2-ARRB2+pXP2d2-rPAPI-luciferase

(28) c) pMet7-AGTR1-Tyk2(C)−HA+pMG1-EFHA1+pXP2d2-rPAPI-luciferase

(29) d) pMet7-AGTR1-Tyk2(C)−HA+pMG2-ARRB2+pXP2d2-rPAPI-luciferase

(30) Luciferase activity is shown as arbitry light units. Error bars indicate standard deviation.

(31) FIG. 5: Dose-dependent effect of GPCR antagonists on the detection of the interaction between GPCRs and ARRB2.

(32) A. Analysis of the effect of GPCR antagonists in an assay measuring the interaction between SSTR2 and ARRB2. Cells were transfected with the following combination of plasmids: pMet7-SSTR2-Tyk2(C)−HA+pMG2-ARRB2+pXP2d2-rPAPI-luciferase. Cells were either left untreated, treated with 10 μM somatostatin or treated with a combination of 10 μM somatostatin and increasing doses (10.sup.−13M up to 10.sup.−6M) of either GPCR antagonist (CYN154806, losartan, telmisartan). Luciferase activity is shown as relative light units (rlu). Error bars indicate standard deviation.

(33) B. Analysis of the effect of GPCR antagonists in an assay measuring the interaction between AGTR1 and ARRB2. Cells were transfected with the following combination of plasmids: pMet7-AGTR1-Tyk2(C)−HA+pMG2-ARRB2+pXP2d2-rPAPI-luciferase. Cells were either left untreated, treated with 10 μM angiotensin II or treated with a combination of 10 μM angiotensin II and increasing doses (10.sup.−13M up to 10.sup.−6M) of either GPCR antagonist (CYN154806, losartan, telmisartan). Luciferase activity is shown as relative light units (rlu). Error bars indicate standard deviation.

(34) FIG. 6: Analysis of ERN1 dimerization.

(35) A. Detection of ERN1 dimerization upon induction of ER (endoplasmatic reticulum)-stress by treatment with tunicamycin. Cells were transfected with the following plasmids:

(36) a) pcDNA5/FRT/TO-ERN1-Tyk2(C)−HA+pMG1+pXP2d2-rPAPI-luciferase

(37) b) pcDNA5/FRT/TO-ERN1-Tyk2(C)−HA+pMG2C-ERN1 pXP2d2-rPAPI-luciferase

(38) After transfection, cells were treated with 0-0.5-1-2 μg/ml tunicamycin, final concentration. Error bars indicate standard deviation.

(39) B. Detection of ERN1 dimerization upon induction of ER-stress by treatment with tunicamycin. Cells were transfected with the following plasmids:

(40) a) pcDNA5/FRT/TO-ERN1-Tyk2(C)−HA+pMG1+pXP2d2-rPAPI-luciferase

(41) b) pcDNA5/FRT/TO-ERN1-Tyk2(C)−HA pMG2C-ERN1 pXP2d2-rPAPI-luciferase

(42) c) pcDNA5/FRT/TO-ERN1-Tyk2(C)−HA pMG2C-ERN1cyt+pXP2d2-rPAPI-luciferase

(43) After transfection, cells were treated with increasing doses tunicamycin. Luciferase activity is shown as fold induction relative to the luciferase signal obtained in cells transfected with unfused gp130 (transfection a) and treated with the same concentration tunicamycin. Error bars indicate standard deviation. Expression of Tyk2(C) and gp130 fusion constructs was evaluated through Western blot applying anti-HA and anti-gp130 antibodies, respectively. Beta-actin expression was stained as a control for equal loading.

(44) C. Analysis of ERN1 structure-function relationship. Cells were transfected with the pXP2d2-rPAPI-luciferase plasmid combined with the indicated Tyk2(C) fusion constructs (pcDNA5/FRT/TO-ERN1-Tyk2(C)−HA, pcDNA5/FRT/TO-ERN1(K599A)-Tyk2(C)−HA or pcDNA5/FRT/TO-ERN1(D123P)-Tyk2(C)−HA) and gp130 fusion constructs (pMG1, encoding unfused gp130 or pMG2C-ERN1 encoding ERN1-gp130), and treated with either tunicamycin or vehicle control (DMSO). Luciferase activity is shown as fold induction relative to the luciferase signal obtained in cells transfected with unfused gp130. Error bars indicate standard deviation. Expression of Tyk2(C) fusion constructs was evaluated through Western blot applying an anti-HA antibody. Beta-actin expression was stained as a control for equal loading.

(45) D. Detection of disruptors of ERN1 dimerization. Cells were transfected with the following plasmids:

(46) a) pcDNA5/FRT/TO-ERN1-Tyk2(C)−HA+pMG1+pXP2d2-rPAPI-luciferase

(47) b) pcDNA5/FRT/TO-ERN1-Tyk2(C)−HA+pMG2C-ERN1 pXP2d2-rPAPI-luciferase

(48) After transfection, cells were treated with tunicamycin or vehicle control (DMSO) combined with increasing doses of Irestatin 9389. Luciferase activity of cells transfected with gp130-fused ERN1 (transfection b) is shown as fold induction relative to the luciferase signal obtained in cells transfected with unfused gp130 (transfection a) and treated with the same concentration of vehicle or tunicamycin with Irestatin 9389. Error bars indicate standard deviation.

(49) FIG. 7: Detection of the interaction between the serotonin transporter (SERT) and synaptobrevins 1 and 2 (VAMP1 and VAMP2). Cells were transfected with the pXP2d2-rPAPI-luciferase plasmid combined with the indicated Tyk2(C) and gp130 fusion constructs. Luciferase activity is shown as fold induction relative to the luciferase signal obtained in cells transfected with unfused gp130 (pMG2). Error bars indicate standard deviation.

EXAMPLES

(50) Materials and Methods to the Disclosure

(51) Plasmids Used in the Examples

(52) A first type of plasmids express chimeric proteins consisting of an HA-tagged C-terminal portion of human Tyk2 fused at its N-terminus to the membrane span protein and are generated in the pMet7 vector, which contains a strong constitutive hybrid SRα promoter (Takebe et al., 1988). A pMet7-dest-Tyk2(C)−HA Gateway destination vector was constructed by first amplifying the Gateway cassette from the pMG1 Gateway destination vector (Lievens et al., 2009) using primers 1 and 2 (see Table below). These primers contained an AgeI and PspOMI restriction enzyme recognition site, respectively, and these enzymes were used to digest the PCR amplicon. Next, the sequence encoding the C-terminal end of human Tyk2 comprising the kinase domain (starting from amino acids 589 and omitting the stop codon) was amplified by PCR on cDNA from HEK293 cells with primers 3 and 4 (see Table below). The former primer contained a NotI restriction site, whereas the latter contained an HA-tag coding sequence as well as an XbaI restriction enzyme recognition site. The PCR amplicon was digested with NotI and XbaI and, together with the AgeI and PspOMI cut fragment described above, ligated in the AgeI-XbaI cut pMet7 vector to generate the pMet7-dest-Tyk2(C)−HA Gateway destination vector. The pMet7-SSTR2-Tyk2(C)−HA and pMet7-AGTR1-Tyk2(C)−HA plasmids were produced by Gateway recombination mediated transfer of the full length sequence of human SSTR2 and AGTR1, respectively, from entry vectors of the hORFeome collection (Lamesch et al., 2007) into the pMet7-dest-Tyk2(C)−HA Gateway destination vector. Using the restriction enzymes EcoRI and MluI, the SSTR2-Tyk2(C)−HA insert (SEQ ID NO:3) of pMet7-SSTR2-Tyk2(C)−HA was subcloned into pSVSport (Invitrogen) to generate pSVSport-SSTR2-Tyk2(C)−HA. The AGTR1-Tyk2-HA construct is depicted in SEQ ID NO:4.

(53) The control plasmids pMet7-HA-Tyk2(C) and pSVSport-HA-Tyk2(C), which are made of the same C-terminal Tyk2 fragment as described above, an HA-tag at the 5′ end and a multiple cloning site at the 3′ end were generated by PCR amplification of the Tyk2 sequence on the pMet7-dest-Tyk2(C)−HA template vector using primers 5 and 6 (see Table below). These primers contain an MfeI site and an HA-tag coding sequence together with an XbaI restriction site, respectively. The MfeI-XbaI digested amplicon was ligated both in the EcoRI-XbaI digested pMet7 vector to result in pMet7-HA-Tyk2(C), and in the EcoRI-XbaI digested pSVSport vector (Invitrogen) to generate pSVSport-HA-Tyk2(C).

(54) pSVSport-HA-Tyk2(C)−RTp66 was produced by transfer of the RTp66 insert from pMG2-RTp66 (Pattyn et al., 2008) to pSVSport-HA-Tyk2(C) using the EcoRI and NotI restriction sites. The HA-Tyk2(C)−RTp66 construct is depicted in SEQ ID NO:28. To generate the pSVSport-HA-Tyk2(C)−SERT plasmid, human SERT was amplified on a SERT containing plasmid template using primers 18 and 19, containing EcoRV and NotI restriction sites, respectively. The amplicon was digested with EcoRV, rendered blunt end by the use of Pfu DNA polymerase and subsequently cut with NotI. This fragment was ligated in pSVSport-HA-Tyk2(C) that was cut with EcoRI, rendered blunt end through Pfu DNA Polymerase treatment and subsequently cut with NotI. The HA-Tyk2(C)−SERT construct is shown in SEQ ID NO:29.

(55) To generate the pcDNA5/FRT/TO-ERN1-Tyk2(C)−HA plasmid, human ERN1 was amplified with primers 9 and 10, containing HindIII and NotI restriction enzyme recognition sites, respectively, using an ERN1 entry clone from the hORFeome collection (Lamesch et al., 2007) as a template. The sequence encoding the C-terminal end of human Tyk2 comprising the kinase domain (starting from amino acids 589 and omitting the stop codon) was amplified by PCR on cDNA from HEK293 cells with primers 11 and 12. The former primer contained a NotI restriction site, whereas the latter contained an HA-tag coding sequence as well as an ApaI restriction enzyme recognition site. The PCR amplicon was digested with NotI and ApaI and, together with the HindIII and NotI cut ERN1 fragment described above, ligated into the HindIII-ApaI cut pcDNA5/FRT/TO vector (Invitrogen) to generate the pcDNA5/FRT/TO-ERN1-Tyk2(C)−HA expression plasmid. The ERN1-Tyk2-HA fusion is depicted in SEQ ID NO:5. The pcDNA5/FRT/TO-ERN1(K599A)-Tyk2(C)−HA plasmid was generated similarly, by amplifying ERN1 from a plasmid containing ERN1(K599A) instead of WT ERN1. The pcDNA5/FRT/TO-ERN1(D123P)-Tyk2(C)−HA plasmid was generated through site-directed mutagenesis of the pcDNA5/FRT/TO-ERN1-Tyk2(C)−HA plasmid using primers 16 and 17. The amino acid sequence of the ERN1 (K599A)-Tyk2(C)−HA en ERN1(D123P)-Tyk2(K)-HA fusion proteins is depicted in SEQ ID NOS:30 and 31, respectively.

(56) The plasmids encoding the fusions with the second interacting polypeptide were of the type also used in MAPPIT, designated pMG2 (WO0190188, Eyckerman et al., 2001; Lemmens et al., 2003). These plasmids encode fusion proteins of the second interacting polypeptide coupled to a fragment of the human gp130 cytokine receptor chain, which contains multiple tyrosine residues that, upon phosphorylation, make up recruitment sites for STAT3. The SV40 large T containing control plasmid pMG2-SVT was generated by transfer of the SVT insert from the previously described pMG1-SVT plasmid (Eyckerman et al., 2001) into the pMG2 vector using EcoRI and NotI restriction enzymes. Human ARRB2 was PCR amplified on an ARRB2 entry clone from the hORFeome collection (Lamesch et al., 2007) using primers 7 and 8 (see Table below) and exchanged with the SVT insert of pMG2-SVT using EcoRI and NotI restriction sites to generate pMG2-ARRB2. pMG1-EFHA1, pMG1-VAMP1 and pMG1-VAMP2 were generated by Gateway recombination mediated transfer of the full length sequences of human EFHA1, VAMP1 and VAMP2, respectively, from entry vectors of the hORFeome collection (Lamesch et al., 2007) into a Gateway compatible version of the pMG1 vector as described earlier (Lievens et al., 2009). The flag tag-gp130-ARRB2, flag tag-gp130-VAMP1 and flag tag-gp130-VAMP2 fusion constructs are depicted in SEQ ID NOS:6, 32 and 33, respectively.

(57) The pMG2C-ERN1 plasmid encoding a fusion protein of the human ERN1 protein N-terminally coupled to a fragment of the human gp130 cytokine receptor chain was generated by PCR amplification of the ERN1 encoding sequence on an ERN1 entry clone from the hORFeome collection (Lamesch et al., 2007) using primers 13 and 14 and cloning this into a MAPPIT vector containing a gp130 encoding sequence at the 3′ end of a Flag-tag encoding sequence and a multi-cloningsite (Pattyn et al., 2008) using EcoRI and XhoI restriction enzymes. The flag tag-ERN1-gp130 fusion construct is depicted in SEQ ID NO:7. The pMG2C-ERN1cyt plasmid encoding a fusion protein of the cytoplasmic portion of the human ERN1 protein fused N-terminally to the gp130 fragment was produced by amplifying the ERN1 cytoplasmic domain on an ERN1 entry clone (see higher) using primers 15 and 14 and cloning this into a MAPPIT vector containing a gp130 encoding sequence using EcoRI and XhoI restriction enzymes, similarly to described above. The flag-tag-ERN1cyt-gp130 fusion construct is depicted in SEQ ID NO:34.

(58) pMG2-RTp51 has been described elsewhere (Pattyn et al., 2008). The flag tag-gp130-RTp51 fusion construct sequence is shown in SEQ ID NO:35. The pMG1 and pMG2 plasmids encoding an unfused gp130 receptor fragment were obtained by cutting out the MAPPIT prey insert of a pMG1 vector using EcoRI and XhoI or of a pMG2 vector using EcoRI and SalI, respectively, blunting the vector backbone through Pfu DNA Polymerase and self-ligation. The amino acid sequence of the polypeptide encoded by pMG1 and pMG2 is depicted in SEQ ID NOS:36 and 37, respectively.

(59) The reporter plasmid pXP2d2-rPAPI-luciferase used in the examples contains the STAT3-dependent rPAPI (rat Pancreatitis-Associated Protein I) promoter driving a firefly luciferase reporter gene as described previously (WO0190188, Eyckerman et al., 2001).

(60) Transfection Procedure

(61) Transfections were carried out using a standard calcium phosphate method. HEK293-T cells were seeded in black tissue-culture treated 96-well plates at 10.000 cells/well in 100 μl culture medium (DMEM supplemented with 10% FCS). Twenty-four hours later, plasmid DNA mixes were prepared that contained plasmids encoding fusion proteins with the first and second interacting proteins and reporter plasmids. The DNA was supplemented with 5 μl 2.5M CaCl.sub.2) and double distilled water to a final volume of 50 μl. This mixture was added drop wise to 50 μl 2×HeBS buffer (280 mM NaCl, 1.5 mM Na.sub.2HPO.sub.4, 50 mM Hepes; pH 7.05) while vigorously vortexing. After incubation at room temperature for 15 min. to allow DNA precipitates to form, the solution was added to the cells at 10 μl/well. Cells were incubated at 37° C., 8% CO2. Twenty-four hours after transfection, cells were treated with the indicated amounts of ligand, either alone or combined with the indicated amount of antagonist. In the case of Irestatin 9389, cells were pre-treated with the antagonist before adding vehicle (DMSO) or tunicamycin. Another twenty-four hours later, luciferase activity was measured using the Luciferase Assay System kit (Promega) on a TopCount luminometer (Perkin-Elmer). Each transfection was done in triplicate and the average of the luciferase activity readings was used in the calculations.

(62) Induction of Dimerization

(63) Tunicamycin (Sigma T7765; 2 mg/ml stock in DMSO) was diluted in culture medium and added to the cells 24 h prior to luciferase signal read-out.

(64) (Ant)Agonists Applied in the Examples

(65) Somatostatin (Sigma 51763) and angiotensin II (Sigma A9525) were solubilized in PBS to make stock concentrations of 500 μM and 10 mM, respectively. CYN154806 trifluoroacetate salt (Sigma C2490) and losartan potassium (Fluka 61188) were dissolved in PBS at a final concentration of 500 μM and 10 mM, respectively. Telmisartan (Sigma T8949) was dissolved in DMSO at a concentration of 10 mM. Irestatin 9389 (Axxon Medchem) was dissolved in DMSO at a concentration of 50 mM.

(66) Western Blotting

(67) Cells were lysed in 1×CCLR buffer (25 mM Tris-phosphate (pH 7.8), 2 mM DTT, 2 mM CDTA (trans-1,2-diaminocyclo-hexane-N,N,N,N-tetra acetic acid), 10% glycerol, 1% Triton X-100). Lysates were centrifuged and supernatants were separated by SDS-PAGE. Proteins were detected by immunoblotting using rat anti-HA (Roche), rabbit anti-gp130 (Santa Cruz Biotechnology) or mouse anti-beta-actin (Sigma) antibodies.

(68) TABLE-US-00001 Oligonu- cleotide primer Sequence (5′ > 3′) 1 CCCACCGGTCCGGAATTGACAAGTTTGTACAAAAAAGC (SEQ ID NO: 9) 2 GGGGGGCCCCAACCACTTTGTACAAGAAAGC (SEQ ID NO: 10) 3 CCCGCGGCCGCTGGCGGTTCGATCACCCAGCTGTCCCACTT GG (SEQ ID NO: 11) 4 TCTAGACTAAGCATAATCTGGAACATCATATGGATACTCGA GGCACACGCTGAACACTGA AGG (SEQ ID NO: 12) 5 CCCCAATTGACCATGTATCCATATGATGTTCCAGATTATGC TTTAATTAAAATCACCCAGCTGTCCCACTTGG (SEQ ID NO: 13) 6 GGGTCTAGAGCGGCCGCACCGGTCTTAATTAAGTCGACGAA TTCGCACACGCTGAACACT GAAG (SEQ ID NO: 14) 7 CCCAAGCTTGAATTCACCATGGGGGAGAAACCCGGGAC (SEQ ID NO: 15) 8 GGGGCGGCCGCCTAGCAGAGTTGATCATCATAG (SEQ ID NO: 16) 9 CCCAAGCTTGGTACCACCATGCCGGCCCGGCGGCTGCTG (SEQ ID NO: 17) 10 CCCGCGGCCGCGCTAGCGAGGGCGTCTGGAGTCACTGG (SEQ ID NO: 18) 11 CCCGCGGCCGCTGGCGGTTCGATCACCCAGCTGTCCCACTT GG (SEQ ID NO: 19) 12 GGGCCCCTAAGCATAATCTGGAACATCATATGGATACTCGA GGCACACGCTGAACACTGA AGG (SEQ ID NO: 20) 13 CCCGAATTCATGCCGGCCCGGCGGCTGCTG (SEQ ID NO: 21) 14 CCCCTCGAGGGGAGGGCGTCTGGAGTCACTGG (SEQ ID NO: 22) 15 CCCGAATTCTTCTGTCCCAAGGATGTCCTG (SEQ ID NO: 23) 16 GGGTAAAAAGCAGCCCATCTGGTATGTTATTGACC (SEQ ID NO: 24) 17 GGTCAATAACATACCAGATGGGCTGCTTTTTACCC (SEQ ID NO: 25) 18 CCCGATATCTATGGAGACGACGCCCTTGAA (SEQ ID NO: 26) 19 GGGGCGGCCGCTTACACAGCATTCAAGCGGA (SEQ ID NO: 27)

Example 1: Detection of the Ligand-Dependent Interaction Between SSTR2 and ARRB2

(69) G-protein coupled receptors (GPCRs) are integral membrane proteins that contain 7 transmembrane domains. Upon binding of the appropriate ligand GPCRs are activated, leading to the recruitment of cytoplasmic beta arrestin proteins. In order to determine whether the assay can detect the somatostatin-dependent interaction between the GPCR SSTR2 and ARRB2, the following combinations of plasmids were transfected (FIG. 2A; 250 ng of the Tyk2(C) fusion construct, 250 ng of the gp130 fusion construct and 50 ng of the luciferase reporter construct) according to the methods described above:

(70) a) pMet7-HA-Tyk2(C)+pMG2-ARRB2+pXP2d2-rPAPI-luciferase

(71) b) pMet7-SSTR2-Tyk2(C)−HA+pMG2-SVT+pXP2d2-rPAPI-luciferase

(72) c) pMet7-SSTR2-Tyk2(C)−HA+pMG2-ARRB2+pXP2d2-rPAPI-luciferase

(73) Transfected cells were either left untreated (NS) or treated with increasing doses (0.1-1-10 μM) of the SSTR2 agonist somatostatin. The fold induction for each sample was calculated as the ratio of the measured luciferase activity relative to the luciferase activity for the untreated sample of the same transfection. The results (FIG. 2B) show a clear ligand dose-dependent signal specifically in the cells co-transfected with both the SSTR2-Tyk2(C) and gp130-ARRB2 fusion constructs (transfection c). No signal was observed when either of the fusion constructs was transfected in combination with a negative control fusion construct (gp130-ARRB2 fusion construct combined with an unfused Tyk2(C) construct in transfection a, or SSTR2-Tyk2(C) fusion construct together with a fusion of gp130 to a fragment of the SV40 large T protein in b).

(74) The assay was further optimized by transferring the Tyk2(C) fusion construct into another vector system (pSVSport) and testing the resulting constructs in a similar experiment as described above. The following combinations of plasmids were transfected (500 ng of the Tyk2(C) fusion construct, 250 ng of the gp130 fusion construct and 50 ng of the luciferase reporter construct) according to the methods described above:

(75) a) pSVSport-HA-Tyk2(C)+pMG2-ARRB2+pXP2d2-rPAPI-luciferase

(76) b) pSVSport-SSTR2-Tyk2(C)−HA+pMG2-SVT+pXP2d2-rPAPI-luciferase

(77) c) pSVSport-SSTR2-Tyk2(C)−HA+pMG2-ARRB2+pXP2d2-rPAPI-luciferase

(78) Transfected cells were either left untreated (NS) or treated with increasing doses (0.1-1-10 μM) of the SSTR2 agonist somatostatin, and signals were calculated as indicated above. The resulting graph (FIG. 2C) shows strong and specific ligand dose-dependent signals up to 30-fold stronger compared to untreated samples.

(79) In another experiment, cells were transfected with 31 ng of the pMet7-SSTR2-Tyk2(C)−HA plasmid, 250 ng of the pMG2-ARRB2 plasmid and 50 ng of the pXP2d2-rPAPI-luciferase plasmid, and transfected cells were treated with a concentration gradient of somatostatin (a ⅓ serial dilution series down from 10 μM). The resulting dose-response curve is shown in FIG. 2D.

(80) Together, these data illustrate that the method is able to detect ARRB2 recruitment to the SSTR2 integral membrane GPCR induced by treatment with the SSTR2 agonist somatostatin.

Example 2: Detection of the Ligand-Dependent Interaction Between AGTR1 and ARRB2

(81) Likewise as in example 1, the ligand-induced recruitment of ARRB2 to another GPCR family member, AGTR1, was tested by transfecting the following combinations of plasmids (250 ng of the Tyk2(C) fusion construct, 250 ng of the gp130 fusion construct and 50 ng of the luciferase reporter construct) according to the methods described above:

(82) a) pMet7-HA-Tyk2(C)+pMG2-ARRB2+pXP2d2-rPAPI-luciferase

(83) b) pMet7-AGTR1-Tyk2(C)−HA+pMG2-SVT+pXP2d2-rPAPI-luciferase

(84) c) pMet7-AGTR1-Tyk2(C)−HA+pMG2-ARRB2+pXP2d2-rPAPI-luciferase

(85) Transfected cells were either left untreated (NS) or treated with increasing doses (0.1-1-10 μM) of angiotensin II, an AGTR1 agonist. The fold induction for each sample was calculated as the ratio of the measured luciferase activity relative to the luciferase activity for the untreated sample of the same transfection. The results (FIG. 3A) show a clear ligand dose-dependent signal specifically in the cells cotransfected with both the AGTR1-Tyk2(C) and gp130-ARRB2 fusion constructs (transfection c). No signal was observed when either of the fusion constructs was transfected in combination with a negative control fusion construct (gp130-ARRB2 fusion construct combined with an unfused Tyk2(C) construct in transfection a, or AGTR1-Tyk2(C) fusion construct together with a fusion of gp130 to a fragment of the SV40 large T protein in b).

(86) In another experiment, cells were transfected with 62 ng of the pMet7-AGTR1-Tyk2(C)−HA plasmid, 250 ng of the pMG2-ARRB2 plasmid and 50 ng of the pXP2d2-rPAPI-luciferase plasmid, and transfected cells were treated with a concentration gradient of angiotensin II (a ⅓ serial dilution series down from 10 μM). The resulting dose-response curve is shown in FIG. 3B.

(87) These results confirm the method's ability to detect the interaction between the AGTR1 integral membrane protein and ARRB2, in a ligand-dependent manner.

Example 3: Effect of GPCR Antagonists on the Detection of the Interaction Between GPCRs and ARRB2

(88) In order to test whether the assay allows evaluating the effect of GPCR antagonists, GPCR ligands were combined with specific antagonists of SSTR2 and AGTR1 in the assay for detection of their interaction with ARRB2. A peptide antagonist that specifically inhibits SSTR2 activation was tested (CYN154806), together with two small molecule AGTR1-selective antagonists (losartan and telmisartan).

(89) Cells were transfected with the following combinations of plasmids (250 ng of the Tyk2(C) fusion construct, 250 ng of the gp130 fusion construct and 50 ng of the luciferase reporter construct) according to the methods described above:

(90) a) pMet7-SSTR2-Tyk2(C)−HA+pMG1-EFHA1+pXP2d2-rPAPI-luciferase

(91) b) pMet7-SSTR2-Tyk2(C)−HA+pMG2-ARRB2+pXP2d2-rPAPI-luciferase

(92) c) pMet7-AGTR1-Tyk2(C)−HA+pMG1-EFHA1+pXP2d2-rPAPI-luciferase

(93) d) pMet7-AGTR1-Tyk2(C)−HA+pMG2-ARRB2+pXP2d2-rPAPI-luciferase

(94) One day after transfection, cells were treated with combinations of GPCR ligand and antagonist (ligand: 1 μM somatostatin for transfections a and b, 10 μM angiotensin II for transfections c and d; antagonists: 0.05 or 0.5 μM CYN154806; 0.1 or 1 μM losartan or telmisartan), and luciferase was measured one day after treatment. The results are shown in FIG. 4 and clearly indicate the specific inhibition by the corresponding antagonist of the GPCR-ARRB2 interactions. The interaction between SSTR2 and ARRB2 (transfection b) can be specifically inhibited by the SSTR2-selective antagonist CYN154806, whereas the AGTR1-specific antagonists losartan and telmisartan have no effect. Conversely, AGTR1-ARRB2 interaction as detected by the assay (transfection d) can be selectively inhibited by the AGTR1-specific antagonists losartan and telmisartan, whereas the SSTR2-selective antagonist CYN154806 has no effect. In both cases, the inhibition through application of the antagonists goes down to background levels observed for cells that had not been treated with GPCR ligand (NS). The inhibitory effect is specific for the GPCR-ARRB2 interaction, as the signal obtained for control interactions of the GPCR-Tyk2(C) fusion construct with a positive control gp130 fusion construct containing EFHA1 (which binds to Tyk2(C) itself), are not affected by the GPCR antagonists.

(95) In a second experiment (shown in FIG. 5), a dose-response curve was generated for the different GPCR antagonists. Cells were transfected with 125 ng of the pMet7-SSTR2-Tyk2(C)−HA or pMet7-AGTR1-Tyk2(C)−HA fusion construct, 250 ng of the pMG2-ARRB2 gp130 fusion construct and 50 ng of the pXP2d2-rPAPI-luciferase reporter plasmid, according to the methods described above. Cells were either left untreated, treated with 10 μM of the appropriate ligand (somatostatin in the case of SSTR2 and angiotensin II in the case of AGTR1) or treated with a combination of the cognate ligand and increasing doses (10.sup.−13M up to 10.sup.−6M) of either GPCR antagonist (CYN154806, losartan, telmisartan). The results are presented in FIG. 5A (for the interaction between SSTR2 and ARRB2) and FIG. 5B (for the interaction between AGTR1 and ARRB2). Again, these data clearly indicate the specific inhibition by the corresponding antagonist of the GPCR-ARRB2 interactions. The interaction between SSTR2 and ARRB2 can be specifically and completely inhibited by the SSTR2-selective antagonist CYN154806, whereas the AGTR1-specific antagonists losartan and telmisartan have no effect. Conversely, AGTR1-ARRB2 interaction as detected by the assay can be selectively and completely inhibited by the AGTR1-specific antagonists losartan and telmisartan, whereas the SSTR2-selective antagonist CYN154806 has no effect. It is of note that the observed stronger effect of telmisartan compared to losartan in this assay corresponds with the reported higher binding affinity of telmisartan versus losartan towards AGTR1 (Kakuta et al., 2005).

(96) Together, these results confirm the specificity of the GPCR-ARRB2 interactions as detected by the assay and indicate that the assay can be applied to identify inhibitors of these interactions.

Example 4: Detection of Context-Dependent Dimerization of a Transmembrane Protein

(97) To support the ability of the method to detect protein-protein interactions under physiological conditions, we studied dimerization of ERN1. ERN1 is a single-span transmembrane protein involved in the cellular response to ER-stress. The ERN1 protein is able to sense unfolded proteins in the ER through its N-terminal domain which is exposed to the ER lumen. This leads to its dimerization and activation of the kinase and endoribonuclease enzymatic domains in its C-terminal moiety exposed towards the cytoplasm. To mimic ER-stress, tunicamycin was applied to the cells, an inhibitor of protein glycosylation which is generally used to induce ER-stress.

(98) In a first experiment, cells were transfected with the following combinations of plasmids (500 ng of the kinase fusion construct, 100 ng of the gp130 fusion construct and 50 ng of the luciferase reporter construct) according to the methods described above:

(99) a) pcDNA5/FRT/TO-ERN1-Tyk2(C)−HA+pMG1+pXP2d2-rPAPI-luciferase

(100) b) pcDNA5/FRT/TO-ERN1-Tyk2(C)−HA+pMG2C-ERN1 pXP2d2-rPAPI-luciferase

(101) After transfection, cells were treated with 0-0.5-1-2 μg/ml tunicamycin, final concentration. The results shown in FIG. 6A show a dose-dependent signal upon addition of tunicamycin, only in cells expressing both ERN1-Tyk2(C) and ERN1-gp130 fusion constructs (transfection b). No signal was observed when the ERN1-Tyk2(C) fusion construct was combined with an unfused gp130 fragment (transfection a).

(102) In a second experiment (FIG. 6B), cells were transfected with the following combinations of plasmids (62.5 ng of the kinase fusion construct, 125 ng of the gp130 fusion construct and 50 ng of the luciferase reporter construct) according to the methods described herein:

(103) a) pcDNA5/FRT/TO-ERN1-Tyk2(C)−HA+pMG1+pXP2d2-rPAPI-luciferase

(104) b) pcDNA5/FRT/TO-ERN1-Tyk2(C)−HA+pMG2C-ERN1 pXP2d2-rPAPI-luciferase

(105) c) pcDNA5/FRT/TO-ERN1-Tyk2(C)−HA pMG2C-ERN1cyt+pXP2d2-rPAPI-luciferase

(106) After transfection, cells were treated with 0-0.04-0.2-1-5 μg/ml tunicamycin, final concentration. The luciferase data are presented as fold induction relative to the signal obtained in cells transfected with unfused gp130 (empty prey; transfection a) and treated with the same concentration tunicamycin. Expression of the different fusion proteins was confirmed using Western blot. These data show that in accordance with the requirement of the ERN1 lumenal domain to sense ER stress, no signal is produced upon overexpression of full length ERN1 kinase fusion and a gp130 fusion containing only the cytoplasmic portion of ERN1 (transfection c).

(107) In a next experiment (FIG. 6C), cells were transfected with combinations of the pXP2d2-rPAPI-luciferase construct (50 ng), a WT or mutant ERN1 kinase fusion construct (62.5 ng) and either unfused or ERN1-fused gp130 construct (125 ng). After transfection, cells were either vehicle (DMSO) treated or treated with 1 μg/ml tunicamycin (final concentration). The mutant ERN1 kinase fusions have mutations in either the luminal domain (D123P) or cytoplasmic ATP-binding pocket (K599A). Both mutations are expected to block ERN1 oligomerization. As evident from FIG. 6C we indeed find that both mutations block the interaction with full length ERN1 gp130 fusion, despite equal expression and similar (aspecific) interaction signals with unfused gp130 constructs.

(108) In another experiment (FIG. 6D), cells were transfected with combinations of the pXP2d2-rPAPI-luciferase construct (50 ng), the ERN1 kinase fusion construct (62.5 ng) and either unfused or ERN1-fused gp130 construct (125 ng). After transfection, cells were treated with tunicamycin (1 μg/ml tunicamycin final concentration) or vehicle (DMSO) combined with increasing doses of Irestatin 9389. This molecule was recently reported to inhibit ERN1 endonuclease activity (Feldman and Koong, 2007). Although the molecular mode of action of Irestatin 9389 was not reported, the molecule induced a dose-dependent disruption of ERN1 dimerization in the assay described herein.

(109) Together, these data indicate that the method is able to specifically detect the ER-stress-induced dimerization of the ERN1 protein and to analyze the structure-function relationship of this protein and pharmacological interference with dimerization of the protein in more detail.

Example 5: Detection of Heterologous Interactions Among Transmembrane Proteins

(110) To further corroborate the ability of the assay to analyze protein-protein interactions involving integral membrane proteins, heterologous interactions between transmembrane proteins were analyzed. Serotonin transporter (SERT) is a multispan integral membrane protein that transports serotonin from the synaptic spaces into presynaptic neurons, this way terminating the action of serotonin and recycling it. In this example, we show its interaction with the synaptobrevins VAMP1 and VAMP2, which are transmembrane proteins involved in fusion of synaptic vesicles with the presynaptic membrane.

(111) Cells were transfected with combinations of the pXP2d2-rPAPI-luciferase construct (50 ng), a SERT or RTp66 kinase fusion construct (1000 ng) and either unfused (pMG2) or one of the indicated gp130 fusion constructs (pMG2-RTp51, pMG1-VAMP1 or pMG2-VAMP2; 500 ng). Luciferase activity is shown as fold induction relative to the luciferase signal obtained in cells transfected with unfused gp130 (pMG2).

(112) The results (FIG. 7) show a clear signal when VAMP1 and VAMP2 gp130 fusion constructs were transfected in combination with the SERT kinase fusion construct, and not when combined with the HIV-1 RTp66 (reverse transcriptase subunit p66) fusion construct. The strong signal obtained for the co-transfection of the RTp66 kinase and the RTp51 gp130 fusion constructs, which has been previously described (WO2012117031), is included as a control for proper expression and functioning of the RTp66 kinase fusion.

REFERENCES

(113) Eyckerman, S., Verhee, A., Van der Heyden, J., Lemmens, I., Van Ostade, X., Vandekerckhove, J., and Tavernier, J. (2001). Design and application of a cytokine-receptor-based interaction trap. Nature Cell Biology 3, 1114-1119. Feldman, D., and Koong, A. (2007). Irestatin, a potent inhibitor of ERN1α and the unfolded protein response, is a hypoxia-selective cytotoxin and impairs tumor growth. J Clin Oncol 25, 3514. Kakuta, H., Sudoh, K., Sasamata, M., and Yamagishi, S. (2005). Telmisartan has the strongest binding affinity to angiotensin II type 1 receptor: comparison with other angiotensin II type 1 receptor blockers. Int J Clin Pharmacol Res 25, 41-46. Lamesch, P., Li, N., Milstein, S., Fan, C., Hao, T., Szabo, G., Hu, Z., Venkatesan, K., Bethel, G., Martin, P., et al., (2007). hORFeome v3.1: a resource of human open reading frames representing over 10,000 human genes. Genomics 89, 307-315. Lemmens, I., Eyckerman, S., Zabeau, L., Catteeuw, D., Vertenten, E., Verschueren, K., Huylebroeck, D., Vandekerckhove, J., and Tavernier, J. (2003). Heteromeric MAPPIT: a novel strategy to study modification-dependent protein-protein interactions in mammalian cells. Nucleic Acids Research 31. Lievens, S., Vanderroost, N., Van der Heyden, J., Gesellchen, V., Vidal, M., and Tavernier, J. (2009). Array MAPPIT: high-throughput interactome analysis in mammalian cells. J Proteome Res 8, 877-886. Pattyn, E., Lavens, D., Van der Heyden, J., verhee, A., Lievens, S., Lemmens, I., Hallenberger, S., Jochmans, D and Tavernier, J. (2008). MAPPIT (Mammalian Protein-Protein Interaction Trap) as a tool to study HIV reverse transcriptase dimerization in intact human cells. J. Virol. Methods 153, 7-15. Takebe, Y., Seiki, M., Fujisawa, J., Hoy, P., Yokota, K., Arai, K., Yoshida, M., and Arai, N. (1988). SR alpha promoter: an efficient and versatile mammalian cDNA expression system composed of the simian virus 40 early promoter and the R-U5 segment of human T-cell leukemia virus type 1 long terminal repeat. Mol Cell Biol 8, 466-472.