Periodic Reporting for period 4 - PD UpReg (Gene knock-up via 3’UTR targeting to treat Parkinson’s disease)
Reporting period: 2022-03-01 to 2023-08-31
Human body is built from about 20.000 gene products, mostly proteins. Scientists have gained insight into gene function by analyzing natural gene mutations which often cause inborn genetic disease. In experimental animals scientists can also delete genes and repress their function using various available tools. These approaches have yielded valuable insight into how genes work and how diseases emerge. However, an ability to enhance gene expression-limited to those cells where the gene is naturally expressed - is currently lacking. Such a gene Knock Up technology may be useful to better model and treat human diseases. For example Parkinson's disease PD can rise from duplication or triplication of a gene called alpha-synuclein (aSyn). Scientists have attempted to create animal models for PD by integrating extra copies of aSyn into mouse genome. However, currently this results in aSyn expression in wrong cells and aSyn expression is also devoid of its physiological regulation. As a result, each aSyn transgenic animal line displays a different phenotype or no phenotype at all.
One of the objectives of this project is to create a gene Knock Up tool which allows to increase gene expression limited to naturally expressing cells. We have achieved this via conditional modulation of negative regulation elements in the gene 3' untranslated region or 3’UTR, a sequence in the DNA which controls the stability of the derived RNA copy, a step prior protein synthesis. Using this approach we have created aSyn-cKU animals, which we are currently analyzing for PD. When successful the new PD model may allow discovery of how PD progresses which in turn would allow generation and testing of new treatments. We have also used the cKU approach to create mice where expression of dopaminergic neurotrophic factor GDNF can be increased. GDNF is a secreted protein which supports function of dopamine neurons which die in PD resulting in loss of control over voluntary movement. Injection of GDNF into the brain has been tested in 6 clinical trials in PD, with billions of euros spent to reach that point. However, there are several problems and the conclusion is still elusive. GDNF acts based on chemoattraction. Hence the re-growing dopamine cell extensions – axons - do not grow back to original innervation targets but towards GDNF injection site in the brain where GDNF concentration is the highest, which may contribute to the observed low efficacy and to side effects. Using GDNF cKU mice we were able to ask and answer, does increase in endogenous GDNF expression manifest a better treatment route for PD. Unfortunately, the answer turned out to be a “no”. We found that increase in endogenous GDNF expression before, during or after implementing lesion to midbrain dopamine neurons to induce PD does not influence PD outcome in a meaningful manner, contrary to the expectations. This unexpected result prompted us to ask why there is then GDNF in the first place, what is its function in the striatum, the principle organizing center of dopamine function. We discovered that GDNF levels regulate dopamine levels and that too high GDNF level likely contributes to schizophrenia (SCZ) PMID: 35618883. We have now taken this unexpected discovery further and found that excess GDNF specific response including global gene expression pattern in visible in about 20% of SCZ patients (Runneberger et al submitted).
One of the objectives of this project is to create a gene Knock Up tool which allows to increase gene expression limited to naturally expressing cells. We have achieved this via conditional modulation of negative regulation elements in the gene 3' untranslated region or 3’UTR, a sequence in the DNA which controls the stability of the derived RNA copy, a step prior protein synthesis. Using this approach we have created aSyn-cKU animals, which we are currently analyzing for PD. When successful the new PD model may allow discovery of how PD progresses which in turn would allow generation and testing of new treatments. We have also used the cKU approach to create mice where expression of dopaminergic neurotrophic factor GDNF can be increased. GDNF is a secreted protein which supports function of dopamine neurons which die in PD resulting in loss of control over voluntary movement. Injection of GDNF into the brain has been tested in 6 clinical trials in PD, with billions of euros spent to reach that point. However, there are several problems and the conclusion is still elusive. GDNF acts based on chemoattraction. Hence the re-growing dopamine cell extensions – axons - do not grow back to original innervation targets but towards GDNF injection site in the brain where GDNF concentration is the highest, which may contribute to the observed low efficacy and to side effects. Using GDNF cKU mice we were able to ask and answer, does increase in endogenous GDNF expression manifest a better treatment route for PD. Unfortunately, the answer turned out to be a “no”. We found that increase in endogenous GDNF expression before, during or after implementing lesion to midbrain dopamine neurons to induce PD does not influence PD outcome in a meaningful manner, contrary to the expectations. This unexpected result prompted us to ask why there is then GDNF in the first place, what is its function in the striatum, the principle organizing center of dopamine function. We discovered that GDNF levels regulate dopamine levels and that too high GDNF level likely contributes to schizophrenia (SCZ) PMID: 35618883. We have now taken this unexpected discovery further and found that excess GDNF specific response including global gene expression pattern in visible in about 20% of SCZ patients (Runneberger et al submitted).
I report by my three Objectives
1. Establish whether increasing endogenous neurotrophic factor GDNF and/or BDNF expression in specific brain areas provides side effect‒free improvements in motor and non-motor symptoms in mouse PD models.
Unexpectedly, we found that increase (or deletion) of endogenous GDNF before, during or after implementing lesion to midbrain dopamine neurons has no major effect on PD outcome PMID: 36690469. Rather, we discovered that GDNF levels are important in schizophrenia, where enhanced striatal dopamine is observed in about 50% of SCZ patients PMID: 35618883. We also created putative cKU allele for BDNF, and found that unlike for GDNF, for BDNF similar strategy does not work - BDNF levels in fact were reduced in all tissues except for the lungs (PMID: 36639028), demonstrating that we still lack basic understanding on how 3’UTRs regulate gene expression in vivo.
2. Establish the therapeutic potential of upregulating endogenous Parkin, mitochondrial quality control regulator, expression in PD models together with GDNF and/or BDNF upregulation.
We have created knock-in allele for Parkin cKU in mice. However, for yet unknown reasons, unlike in the above GDNF cKU animals, we do not observe increase in endogenous Parkin expression from the Parkin cKU allele.
3. Create a mouse model to phenocopy PD by generating a conditional Snca knock-up allele and determine the therapeutic potential of the molecular chaperones HSPA1A and proSAAS.
We have successfully created Snca cKU allele and a aSyn cKU allele which contains three independently PD driving mutations (E46K+H50Q+G51D). Analysis of those mice is still continuing.
For HSPA1A and proSAAS we have created 3’UTR targeting lentiviral vectors but similar to BDNF and Parkin, increase in expression was not observed by implementing the proposed 3’UTR shortening. Those results underline that we miss some important pieces of the puzzle in how 3’UTRs regulate gene expression levels, which lead us to our next project, see under Exploitation and Dissemination.
1. Establish whether increasing endogenous neurotrophic factor GDNF and/or BDNF expression in specific brain areas provides side effect‒free improvements in motor and non-motor symptoms in mouse PD models.
Unexpectedly, we found that increase (or deletion) of endogenous GDNF before, during or after implementing lesion to midbrain dopamine neurons has no major effect on PD outcome PMID: 36690469. Rather, we discovered that GDNF levels are important in schizophrenia, where enhanced striatal dopamine is observed in about 50% of SCZ patients PMID: 35618883. We also created putative cKU allele for BDNF, and found that unlike for GDNF, for BDNF similar strategy does not work - BDNF levels in fact were reduced in all tissues except for the lungs (PMID: 36639028), demonstrating that we still lack basic understanding on how 3’UTRs regulate gene expression in vivo.
2. Establish the therapeutic potential of upregulating endogenous Parkin, mitochondrial quality control regulator, expression in PD models together with GDNF and/or BDNF upregulation.
We have created knock-in allele for Parkin cKU in mice. However, for yet unknown reasons, unlike in the above GDNF cKU animals, we do not observe increase in endogenous Parkin expression from the Parkin cKU allele.
3. Create a mouse model to phenocopy PD by generating a conditional Snca knock-up allele and determine the therapeutic potential of the molecular chaperones HSPA1A and proSAAS.
We have successfully created Snca cKU allele and a aSyn cKU allele which contains three independently PD driving mutations (E46K+H50Q+G51D). Analysis of those mice is still continuing.
For HSPA1A and proSAAS we have created 3’UTR targeting lentiviral vectors but similar to BDNF and Parkin, increase in expression was not observed by implementing the proposed 3’UTR shortening. Those results underline that we miss some important pieces of the puzzle in how 3’UTRs regulate gene expression levels, which lead us to our next project, see under Exploitation and Dissemination.
Exploitation and Dissemination
Under ERC PD UpReg we found that contrary to the expectations, endogenous GDNF level is not important in PD. Rather, it is important in schizophrenia, PMID: 35618883. More specifically, we found that about 20% of SCZ display high GDNF and “GDNF response” in the striatum. These results enabled me to create ERANET NEURON Research consortia entitled GDNF UpReg under H2020, which aims to further characterize GDNFhigh SCZ patients, to prepare for future clinical trials. I coordinate this consortia which contains clinical and basic research teams from Germany, Finland and Estonia.
In parallel I have started development of precision medicine to treat specifically GDNFhigh SCZ subgroup. There I have raised further funding and work together with clinical chemists and structural biologists from Finland, Lativa and Sweden on hit-to-lead development of our first drug candidates. I have also applied for ERC Proof-of-Concept grant to take those results further.
Finally, under PD UpReg we made 2 principal observations -our knowledge on 3’UTRs is fragmentary, there is a lack of understanding how they really work thus we are unable to predict the outcome on expression. Second, we found that some 3’UTRs result in better protein expression than the currently “golden” 3’UTRs used for protein expression by industry as well as by the academic research teams. By combining those two observations we successfully rose 700.000 euros R2B grant from Business Finland which creates its own “omics” database on mRNA and protein expression in industrially used cell-lines and works on creating a new machine learning based pipeline which aims to improve protein expression in mammalian cells, eventually aiming to reduce the price tag of recombinant proteins by increasing the efficacy of industrial protein production, pls see https://mammalexpress.com/(opens in new window) for further details on this ongoing project.
Under ERC PD UpReg we found that contrary to the expectations, endogenous GDNF level is not important in PD. Rather, it is important in schizophrenia, PMID: 35618883. More specifically, we found that about 20% of SCZ display high GDNF and “GDNF response” in the striatum. These results enabled me to create ERANET NEURON Research consortia entitled GDNF UpReg under H2020, which aims to further characterize GDNFhigh SCZ patients, to prepare for future clinical trials. I coordinate this consortia which contains clinical and basic research teams from Germany, Finland and Estonia.
In parallel I have started development of precision medicine to treat specifically GDNFhigh SCZ subgroup. There I have raised further funding and work together with clinical chemists and structural biologists from Finland, Lativa and Sweden on hit-to-lead development of our first drug candidates. I have also applied for ERC Proof-of-Concept grant to take those results further.
Finally, under PD UpReg we made 2 principal observations -our knowledge on 3’UTRs is fragmentary, there is a lack of understanding how they really work thus we are unable to predict the outcome on expression. Second, we found that some 3’UTRs result in better protein expression than the currently “golden” 3’UTRs used for protein expression by industry as well as by the academic research teams. By combining those two observations we successfully rose 700.000 euros R2B grant from Business Finland which creates its own “omics” database on mRNA and protein expression in industrially used cell-lines and works on creating a new machine learning based pipeline which aims to improve protein expression in mammalian cells, eventually aiming to reduce the price tag of recombinant proteins by increasing the efficacy of industrial protein production, pls see https://mammalexpress.com/(opens in new window) for further details on this ongoing project.