Final Report Summary - DDIPI (Drug Discovery in Interacting Proteins)
This project was concerned with the discovery of compounds which bind to membrane receptors implicated in human disease, with the identification of proteins which interact with such receptors so as to inhibit these interactions as a therapeutic approach and with the development of a sensor technology based on novel membrane receptors inserted into re-engineered yeast which can be used as a diagnostic tool. The partners were a SME involved in yeast technologies and protein interactions ( Dualsystems), a SME in synthetic organic chemistry (AFchempharm) and an academic group concerned with membrane protein science (NUIM). The principal initial targets were G-protein coupled receptors and the retinol binding protein (RBP) receptor/vitamin A transporter implicated in type 2 diabetes.
l.Anti-diabetic Compounds: Since elevated levels of serum retinol binding protein (RBP) have been deeply implicated in the genesis of insulin resistance and type 2 diabetes, this anti-diabetes drug discovery project was designed to identify non-retinoid compounds which bind to RBP to effect two outcomes, (i) reduction of serum RBP levels by inhibiting formation of the RBP-transthyretin (TTR) complex which maintains serum RBP levels, and (ii) impairment of the RBP:receptor interaction thereby restoring insulin sensitivity. The study was very successful in that two series of effective anti-diabetic compounds were discovered using the high fat-fed model of the disease ( Campos-Sandoval JA, Redondo C, Kinsella GK, Pal A, Jones G, Eyre GS, Hirst SC, Findlay JB. Fenretinide derivatives act by disrupting the interactions of serum retinol binding protein (sRBP) with transthyretin and the sRBP receptor J Med Chem. 2011 Jul 14;54(13):4378-87)
During the structure-activity phase of the project, a derivative of one of these “hits” completely restored normal glucose and insulin tolerance when given to mice previously rendered diabetic. This dramatic result encouraged examination of the effects of the compound in cultured muscle cells in the absence of RBP. The exciting outcome was that the compounds independently of insulin stimulated glucose uptake via the glucose transporter. This result raises the real prospect of an approach to type 1 diabetes as well since it is also a substitute for insulin as regards glucose handling. Further investigation of the mechanism of this stimulated glucose uptake revealed that the compounds were reducing the efficiency of ATP production in the mitochondria by reducing the activity of complex 1 of the electron transport chain. However, ATP levels were not significantly altered. Instead it was likely that the AMP to ATP ratio was increased, thereby stimulating increased phosphorylation of AMPkinase and the activation of this important metabolic enzyme. It was then demonstrated that phosphorylation and stimulation of AS160 occurred followed by increased mobilization of the GLUT4 transporter to the plasma membrane with the resultant increased uptake of glucose. In a variety of tests , the compounds exhibited no off-target or toxic indications. Thus, these compounds are reproducing mechanisms brought into play by exercise, thereby constitute a “natural” anti-diabetic effect ( J. Findlay, G. Kinsella, R. Devine, T. Velasco Torrijos, C.Fishwick and J. C. Stephen, New compounds for the treatment of diabetes, patent pending, , PCT/EP2012/071286. (2012): manuscript in preparation). A proteomics profile of the effect of the drug on muscle cells has been carried out ( manuscript submitted) which reinforces this mechanisms, revealing the widespread associated effects on metabolic processes induced by AMPkinase activation.
2. GPCRs implicated in obesity and Insulin resistance: Recently, two G protein coupled – GPR21 & GPR105 – have been linked to type 2 diabetes. Separate GPR21 and GPR105 knock-out studies in mice on a high-fat diet reported improved glucose tolerance, improved systemic insulin sensitivity, and reduced macrophage infiltration in fat tissue and liver. GPR105 has also been newly proposed to be a regulator of insulin secretion. Therefore, antagonists of these receptors are potentially new therapeutic approaches to type 2 diabetes. GPR21 has no known ligands. Such antagonist will also be valuable for in vivo studies to examine the role of these receptors since knock-out studies can be unreliable indicators of receptor function. The project was concerned, therefore, to develop an in vitro system using cultured cells to investigate the activities of these receptors and for the discovery of agents which influence these activities. This was achieved in both cases, the expression of GPR21, in particular, occurring at very high levels.
Homology models were developed for both GPR21 and GPR105, and validated to aid in the identification of potential ligands with at least 10μM affinity. Virtual screening was pursued to examine the interactions of large compound databases in silico through docking (OpenEye, Fred software). For GPR21, a subset of 28 chemically diverse scaffolds, were subsequently tested in a FRET based assay to monitor IP1 production. From this 3 interesting scaffolds emerged as potential agonists/antagonists. Ongoing work includes SAR exploration of these initial hit compounds. Importantly, the studies reveal that GPR21 is constitutively active and inhibits the action of insulin, suggesting that there is a natural antagonist in blood/tissue extracts which might regulate the system in the likes of the macrophage. Further, two of the above 3 compound hits antagonize the inhibitory effect of GPR21, establishing the proof of concept which underlies the rationale behind the study and effectively de-orphanising the receptor. This is potentially a major new avenue for combating insulin resistance and type 2 diabetes since it addresses the migration of inflammatory macrophages into visceral adipose tissue.
Following a similar approach, 7 novel GPR105 compounds were identified and tested for activity with a transformed Gαi1-coupled yeast assay. Two were found to produce significant agonism in the fluorescence intensity of GPR105-transformed yeast. AfChemPharm synthesized a known antagonist, PPTN(5), for this receptor (not available commercially) for use as a negative control. This work contimues
The melanocortin receptor MC4R is important in energy homeostasis. It regulates food intake and the failure of natural mutants to activate produces an insatiable appetite, leading to gross obesity and type 2 diabetes. More than 80% of over 100 known mutants exhibit a significant failure to traffic to the plasma membrane. Thus, the receptor cannot be influenced by its natural ligands which are located in the extracellular milieu. As above, structural modelling and in silico screening has been used to identify potential “pharmacological chaperones” which will correct the presumed miss-fold and facilitate trafficking. Studies with three mutant receptors and 4 antagonist derivatives proved to be effective in facilitating trafficking, effectively establishing the viability of this new therapeutic approach to natural mutant proteins.
Finally, work was initiated on a member of the adhesion family of GPCRs, a little studied group of 33 receptors. CD97 is deeply implicated in human disease, most clearly thus far in immunity-related conditions such as rheumatoid arthritis. Antibodies to this receptor were found to relieve the symptoms of the condition in mice and to have a positive effect on bone wastage. The recent work was concerned with expressing the receptor in cell systems (yeast and HEK cells) to allow the testing of 8 compounds predicted to be ligands. Two of these showed some indications of being antagonists.This work remains preliminary due to the inconsistent receptor expression
Combating Antibiotic resistance by a new Approach. Antibiotic resistance by bacteria has reached almost crisis proportions and there is a very urgent need to overcome this problem. A good example can be seen in the case of Pseudomonas aeruginosa, responsible for multiples nocosomial infections such as pneumonia and infections of the urinary tract, ear, skin, soft tissues and gastrointestinal tract and for which there are very few treatment options. It is a multidrug-resistant pathogen because in addition to possessing several virulence factors, it has an intrinsic resistance due to the presence of an outer membrane and several multidrug efflux transporters.
The targets for this new therapeutic approach to gram negative organisms are the penicillin binding proteins (PCPs) which are essential for the synthesis of the cell wall. The strategy is to design and synthesis inhibitors which are transition state analogues able to bind covalently to both serine proteases and PCPs. The chemical novelty behind the project is the use of boronic acid derivatives.The boron atom is known to act as an electrophile and is able to mimic the carbonyl functionality of a β-lactam. As a consequence, the boron reacts with the catalytic serine residue of the protease to form a tetrahedral complex, mimicking the transition state of the enzyme-adduct complex. The boronic acid-based inhibitors block access to the active site of β- lactamases, thereby preventing destruction of the β-lactam ring of a drug molecule which can react with the penicillin binding protein to prevent bacterial wall synthesis.
A new generation of boronic acid derivatives targeted at PBP3 Pa was chosen, using ligand-based, structure-based and computer aided drug design (CADD) approaches, partly obtained from the secondment of a scientist from NUIM . To generate a large library of compounds, the fragment combination approach was adopted. Following screening of a potential library, 258 unique molecules were subjected to molecular modelling from which three compounds showing good docking to the target PBP were selected for synthesis.The synthetic routes for these molecules are complex and involve the production of intermediate synthons. Despite considerable difficulty, these have now been
produced in acceptable yield. Once the final molecules are prepared, their ability to bind to the target PCP will be carried out in the NUIM laboratories using protein produced in NUIM by recombinant methods.
Protein Interaction Drug targets. The objective of this project was to use pull-down/mass spectrometry and yeast two hybrid technologies to identify intracellular proteins which interact with the RBP receptor and which may thereby participate in its activity and potentially in the induction of insulin resistance. Such proteins might constitute novel dug targets. We have identified a number of proteins that co-purify with over-expressed STRA6 from HEK cells by an unbiased shotgun proteomic approach as illustrated by a volcano plot. The interactions have been confirmed by direct and reverse immuno-precipitation and by yeast pair-wise interactions. Two of these interacting proteins have no known function but exhibit human mutations which give rise to severe dysfunction, amyotrophic lateral sclerosis (a motor neuron disease: the “ice-bucket challenge”) and blood vessel inflammation both of which can be related to retinol utilization. A third interacting protein is known to affect the activity of critical nuclear retinoic acid receptors. These exciting results not only allow function to be ascribed to these proteins but also indicate new tissue-specific pathways for vitamin A handling and the prospect of developing the new drug strategy of “pharmacological chaperones” to restore the proper folds to the mutant proteins (manuscript in preparation). Knock-out mouse models for one of these proteins is in preparation.
Sensor development. Based on earlier proof of concept work, development of the sensor system has continued despite the withdrawal before the mid-term point of Dualsystems. A yeast strain for the discovery of novel GPCRs for defined analytes has been improved and a new library of 100M genes mutated to the level of 10% in the ligand-binding regions has been generated and validated. This will allow the creation and recovery of new genes by the process of accelerated evolution. The sensor strain is in the process of being made "brighter" by the incorporation of newer fluorescent protein in the yeast genome and by the attachment of strong promoters to increase protein expression. Progress has been severely affected by the withdrawal of Dualsystems who are expert in yeast genetics and manipulation ( patent files).
l.Anti-diabetic Compounds: Since elevated levels of serum retinol binding protein (RBP) have been deeply implicated in the genesis of insulin resistance and type 2 diabetes, this anti-diabetes drug discovery project was designed to identify non-retinoid compounds which bind to RBP to effect two outcomes, (i) reduction of serum RBP levels by inhibiting formation of the RBP-transthyretin (TTR) complex which maintains serum RBP levels, and (ii) impairment of the RBP:receptor interaction thereby restoring insulin sensitivity. The study was very successful in that two series of effective anti-diabetic compounds were discovered using the high fat-fed model of the disease ( Campos-Sandoval JA, Redondo C, Kinsella GK, Pal A, Jones G, Eyre GS, Hirst SC, Findlay JB. Fenretinide derivatives act by disrupting the interactions of serum retinol binding protein (sRBP) with transthyretin and the sRBP receptor J Med Chem. 2011 Jul 14;54(13):4378-87)
During the structure-activity phase of the project, a derivative of one of these “hits” completely restored normal glucose and insulin tolerance when given to mice previously rendered diabetic. This dramatic result encouraged examination of the effects of the compound in cultured muscle cells in the absence of RBP. The exciting outcome was that the compounds independently of insulin stimulated glucose uptake via the glucose transporter. This result raises the real prospect of an approach to type 1 diabetes as well since it is also a substitute for insulin as regards glucose handling. Further investigation of the mechanism of this stimulated glucose uptake revealed that the compounds were reducing the efficiency of ATP production in the mitochondria by reducing the activity of complex 1 of the electron transport chain. However, ATP levels were not significantly altered. Instead it was likely that the AMP to ATP ratio was increased, thereby stimulating increased phosphorylation of AMPkinase and the activation of this important metabolic enzyme. It was then demonstrated that phosphorylation and stimulation of AS160 occurred followed by increased mobilization of the GLUT4 transporter to the plasma membrane with the resultant increased uptake of glucose. In a variety of tests , the compounds exhibited no off-target or toxic indications. Thus, these compounds are reproducing mechanisms brought into play by exercise, thereby constitute a “natural” anti-diabetic effect ( J. Findlay, G. Kinsella, R. Devine, T. Velasco Torrijos, C.Fishwick and J. C. Stephen, New compounds for the treatment of diabetes, patent pending, , PCT/EP2012/071286. (2012): manuscript in preparation). A proteomics profile of the effect of the drug on muscle cells has been carried out ( manuscript submitted) which reinforces this mechanisms, revealing the widespread associated effects on metabolic processes induced by AMPkinase activation.
2. GPCRs implicated in obesity and Insulin resistance: Recently, two G protein coupled – GPR21 & GPR105 – have been linked to type 2 diabetes. Separate GPR21 and GPR105 knock-out studies in mice on a high-fat diet reported improved glucose tolerance, improved systemic insulin sensitivity, and reduced macrophage infiltration in fat tissue and liver. GPR105 has also been newly proposed to be a regulator of insulin secretion. Therefore, antagonists of these receptors are potentially new therapeutic approaches to type 2 diabetes. GPR21 has no known ligands. Such antagonist will also be valuable for in vivo studies to examine the role of these receptors since knock-out studies can be unreliable indicators of receptor function. The project was concerned, therefore, to develop an in vitro system using cultured cells to investigate the activities of these receptors and for the discovery of agents which influence these activities. This was achieved in both cases, the expression of GPR21, in particular, occurring at very high levels.
Homology models were developed for both GPR21 and GPR105, and validated to aid in the identification of potential ligands with at least 10μM affinity. Virtual screening was pursued to examine the interactions of large compound databases in silico through docking (OpenEye, Fred software). For GPR21, a subset of 28 chemically diverse scaffolds, were subsequently tested in a FRET based assay to monitor IP1 production. From this 3 interesting scaffolds emerged as potential agonists/antagonists. Ongoing work includes SAR exploration of these initial hit compounds. Importantly, the studies reveal that GPR21 is constitutively active and inhibits the action of insulin, suggesting that there is a natural antagonist in blood/tissue extracts which might regulate the system in the likes of the macrophage. Further, two of the above 3 compound hits antagonize the inhibitory effect of GPR21, establishing the proof of concept which underlies the rationale behind the study and effectively de-orphanising the receptor. This is potentially a major new avenue for combating insulin resistance and type 2 diabetes since it addresses the migration of inflammatory macrophages into visceral adipose tissue.
Following a similar approach, 7 novel GPR105 compounds were identified and tested for activity with a transformed Gαi1-coupled yeast assay. Two were found to produce significant agonism in the fluorescence intensity of GPR105-transformed yeast. AfChemPharm synthesized a known antagonist, PPTN(5), for this receptor (not available commercially) for use as a negative control. This work contimues
The melanocortin receptor MC4R is important in energy homeostasis. It regulates food intake and the failure of natural mutants to activate produces an insatiable appetite, leading to gross obesity and type 2 diabetes. More than 80% of over 100 known mutants exhibit a significant failure to traffic to the plasma membrane. Thus, the receptor cannot be influenced by its natural ligands which are located in the extracellular milieu. As above, structural modelling and in silico screening has been used to identify potential “pharmacological chaperones” which will correct the presumed miss-fold and facilitate trafficking. Studies with three mutant receptors and 4 antagonist derivatives proved to be effective in facilitating trafficking, effectively establishing the viability of this new therapeutic approach to natural mutant proteins.
Finally, work was initiated on a member of the adhesion family of GPCRs, a little studied group of 33 receptors. CD97 is deeply implicated in human disease, most clearly thus far in immunity-related conditions such as rheumatoid arthritis. Antibodies to this receptor were found to relieve the symptoms of the condition in mice and to have a positive effect on bone wastage. The recent work was concerned with expressing the receptor in cell systems (yeast and HEK cells) to allow the testing of 8 compounds predicted to be ligands. Two of these showed some indications of being antagonists.This work remains preliminary due to the inconsistent receptor expression
Combating Antibiotic resistance by a new Approach. Antibiotic resistance by bacteria has reached almost crisis proportions and there is a very urgent need to overcome this problem. A good example can be seen in the case of Pseudomonas aeruginosa, responsible for multiples nocosomial infections such as pneumonia and infections of the urinary tract, ear, skin, soft tissues and gastrointestinal tract and for which there are very few treatment options. It is a multidrug-resistant pathogen because in addition to possessing several virulence factors, it has an intrinsic resistance due to the presence of an outer membrane and several multidrug efflux transporters.
The targets for this new therapeutic approach to gram negative organisms are the penicillin binding proteins (PCPs) which are essential for the synthesis of the cell wall. The strategy is to design and synthesis inhibitors which are transition state analogues able to bind covalently to both serine proteases and PCPs. The chemical novelty behind the project is the use of boronic acid derivatives.The boron atom is known to act as an electrophile and is able to mimic the carbonyl functionality of a β-lactam. As a consequence, the boron reacts with the catalytic serine residue of the protease to form a tetrahedral complex, mimicking the transition state of the enzyme-adduct complex. The boronic acid-based inhibitors block access to the active site of β- lactamases, thereby preventing destruction of the β-lactam ring of a drug molecule which can react with the penicillin binding protein to prevent bacterial wall synthesis.
A new generation of boronic acid derivatives targeted at PBP3 Pa was chosen, using ligand-based, structure-based and computer aided drug design (CADD) approaches, partly obtained from the secondment of a scientist from NUIM . To generate a large library of compounds, the fragment combination approach was adopted. Following screening of a potential library, 258 unique molecules were subjected to molecular modelling from which three compounds showing good docking to the target PBP were selected for synthesis.The synthetic routes for these molecules are complex and involve the production of intermediate synthons. Despite considerable difficulty, these have now been
produced in acceptable yield. Once the final molecules are prepared, their ability to bind to the target PCP will be carried out in the NUIM laboratories using protein produced in NUIM by recombinant methods.
Protein Interaction Drug targets. The objective of this project was to use pull-down/mass spectrometry and yeast two hybrid technologies to identify intracellular proteins which interact with the RBP receptor and which may thereby participate in its activity and potentially in the induction of insulin resistance. Such proteins might constitute novel dug targets. We have identified a number of proteins that co-purify with over-expressed STRA6 from HEK cells by an unbiased shotgun proteomic approach as illustrated by a volcano plot. The interactions have been confirmed by direct and reverse immuno-precipitation and by yeast pair-wise interactions. Two of these interacting proteins have no known function but exhibit human mutations which give rise to severe dysfunction, amyotrophic lateral sclerosis (a motor neuron disease: the “ice-bucket challenge”) and blood vessel inflammation both of which can be related to retinol utilization. A third interacting protein is known to affect the activity of critical nuclear retinoic acid receptors. These exciting results not only allow function to be ascribed to these proteins but also indicate new tissue-specific pathways for vitamin A handling and the prospect of developing the new drug strategy of “pharmacological chaperones” to restore the proper folds to the mutant proteins (manuscript in preparation). Knock-out mouse models for one of these proteins is in preparation.
Sensor development. Based on earlier proof of concept work, development of the sensor system has continued despite the withdrawal before the mid-term point of Dualsystems. A yeast strain for the discovery of novel GPCRs for defined analytes has been improved and a new library of 100M genes mutated to the level of 10% in the ligand-binding regions has been generated and validated. This will allow the creation and recovery of new genes by the process of accelerated evolution. The sensor strain is in the process of being made "brighter" by the incorporation of newer fluorescent protein in the yeast genome and by the attachment of strong promoters to increase protein expression. Progress has been severely affected by the withdrawal of Dualsystems who are expert in yeast genetics and manipulation ( patent files).