This one-day introductory workshop assembled fellows from CIFAR’s Molecular Architecture of Life program and pharmaceutical leaders to discuss the current progress, challenges and future opportunities for drug discovery.
Four central themes were explored: the current gaps in industry, the spatio-temporal dynamics of biomolecules, single-cell analysis and systems pharmacology, the future of proteomics and finally the future of structural biology. Presenters reported on topics related to common drug targets, novel approaches for screening drug effects, approaches and tools for monitoring drug target dynamics and disease progression, and novel technologies with clinical utility.
Several factors which are currently limiting drug discovery were highlighted including the availability of tools for assessing spatial-temporal dynamics across different tissue types, the need for increased data sharing, the value in providing more detailed experimental conditions, the lack of a wellness signature and the use of models in the early discovery stage which do not accurately reflect the aspects of human disease. Developing a “healthy” baseline for all types of biomolecules would inform on alterations in disease which could in turn provide potential drug targets. The discussion identified several areas for further exploration. Could the aforementioned limitations account for the observation that currently most drugs fail due to efficacy rather than due to safety? Could the lack of investment in the biological understanding of disease, particularly complex diseases, also contribute to the drug efficacy failure? Would enhanced patient profiling lead to superior drug efficacy as some drugs will only demonstrate applicability in a very specific and limited population (e.g. the presence of a specific mutation)?
The discussion also centered on approaches which can improve upon current drug discovery including the utilization of relevant biological models which recapitulate human disease more accurately (e.g. ex vivo patient samples), combining approaches (structural and computational biology) and combining multiple layers of “omic” data (proteomics, metabolomics, genomics, lipidomics etc.). The development of additional tools is required to facilitate the merging of these large-scale data sets.
Valuable tools for discovering new drugs and monitoring response to therapies were presented and discussed: bioluminescence resonance energy transfer, which enables the study of the dynamics of real-time signaling in living cells, two-photon excited fluorescence, which enables visualisation of subcellular structures in living animal retina, picosecond IR laser which enables surgery precision at the single-cell level with no scar tissue formation, Volta phase plate, a method for phase contrast in transmission electron microscopy, electron paramagnetic resonance (EPR) and double electron-electron resonance (DEER), which are methods that measure the distance between spin labels and ultra-bright electron sources which allow the observation of atomic motions during chemical processes. Improvements to the current technologies including tools which enable their combination are expected to enhance drug discovery. Finally, a high-throughput assay of individual microbiomes to assess drug-microbiome interactions revealed that the response to different FDA-approved drugs differs among individuals and can cause drastic effects on the microbiome thus raising the question of whether new drug candidates should be screened for their effects on the microbiome.
SESSION 1: CURRENT INDUSTRY GAPS
The first session chaired by Michel Bouvier, CIFAR Senior Fellow, PI and CEO, Institute for Research in Immunology and Cancer, focused on current industry gaps and what is required to discover the next generation of drugs. The need to push the limits of existing technology and to create technology which is amenable to a larger community of researchers across different disciplines was highlighted. There is a need to explore and discuss how the scientific community can produce research which is more impactful for drug discovery. Building technology which can inform at the atomic level and creating molecular maps of the cell are approaches which can lead to improved drug design.
INSIGHTS FROM GROUP DISCUSSION ON CURRENT INDUSTRY GAPS
Establishing a standard “healthy” baseline would aid drug discovery.
In order to understand the detailed biological changes occurring in disease, a standard “healthy” baseline is needed for all types of biomolecules. Understanding the detailed mechanisms by which healthy individuals progress to a disease state will provide valuable insight into potential drug targets. While there are copious amount of data for illness, a wellness signature is still lacking. Monitoring individuals over time prior to the onset of disease would enable the establishment of a wellness signature; analyzing human samples currently stored in tissue banks could aid in and establish this protein wellness signature.
Increasing data sharing and further defining sample conditionality can improve drug discovery.
Protein structure is crucial for drug discovery and by combining structural and computational biology, drug development can be improved (e.g. potency). Despite a wealth of data provided with the recent ‘omic’ revolution, additional high quality data is required for computational biologists to generate a complete picture. The importance of providing detailed experimental conditions (conditionality) was highlighted and could aid in addressing the limited ability to reproduce experimental results, known as the “reproducibility crisis”. Publically sharing data in a meaningful manner which enables its interpretation across disciplines was also emphasized. What repository infrastructure is needed to support meaningful data sharing requires further exploration?
Drug discovery would benefit from multi-omic analysis.
Combining several layers of omic data including transcriptomic, genomic, lipidomics and proteomics could provide valuable insights on drug discovery and mechanisms of action; however, approaches in combining omic data is still lacking as are methods which enable meaningful correlation of data. Given that proteins are the ultimate biological functional unit, proteomic data is particularly useful in drug discovery. Currently transcriptomic data is relatively affordable although it is predicted that in the next 5 to 10 years to cost of proteomics will be comparable to that of transcriptomics
SESSION 2: SPATIO-TEMPORAL DYNAMICS OF BIOMOLECULES AND SINGLE-CELL ANALYSIS AND SYSTEMS PHARMACOLOGY
This session was chaired by Robert H Austin, CIFAR Advisor, Professor, Princeton University and included presentations by Michel Bouvier, CIFAR Senior Fellow, PI and CEO, Institute for Research in Immunology and Cancer and by Krzysztof Palczewski, CIFAR Senior Fellow, Professor & Chair, Case Western Reserve University.
Real-Time Detection of Signal Propagation in Live Cells
G-protein-coupled receptors (GPCR) represent a significant portion of drug targets. GPCR signaling is pluridimensional (ability to elicit a variety of signaling events). An example of a physiologically relevant GPCR is the AT1R (angiotensin receptor), for which the ligands selectively activate β-arrestin and demonstrate utility in cardiovascular therapy.
Bioluminescence resonance energy transfer (BRET) is a powerful approach to monitor GPCR trafficking and thus has utility in drug discovery. BRET enables the study of the dynamics of real time signaling in living cells. Clustering drug signaling profiles allows the prediction of drugs side effects.
Systems Pharmacology: Towards Drug Development with the New Understanding of Cell Networks
Approaches which allow for monitoring the progression of disease can benefit drug discovery. Two-photon excited fluorescence (TPEF) enables visualisation of subcellular structures in living animal retina and thus the generation of maps of visual sensitivity. This imaging technique represents a valuable tool to elucidate disease onset, progression and response to therapeutic intervention.
Complex diseases require complex approaches. Systems pharmacology can pave the way for solving the medical needs of visually impaired individuals.
INSIGHTS FROM GROUP DISCUSSION ON SPATIO-TEMPORAL DYNAMICS OF BIOMOLECULES AND SINGLE-CELL ANALYSIS AND SYSTEMS PHARMACOLOGY
Tools for assessing in vitro spatial temporal dynamics are lacking.
On a single-cell level the spatial orientation signature is not captured and is difficult to assess; with the exception of tools specifically for the eye wherein the individual response of rods and cones can be assessed and the same cells can be re-interrogated. When these dynamics are unaccounted for, the context of the environment is lost (e.g. adjacent cells). Do many drugs fail because the spatial temporal dynamics are not assessed? Would the development of effective tools for assessing spatial temporal dynamics which are compatible with various tissues improve drug discovery?
The majority of drugs are not failing for safety, but rather due to lack of efficacy.
This highlights the need for more relevant models for testing early stage drugs (translational issue). Many have transitioned to systems which recapitulate human disease more closely including ex vivo patient samples (e.g. organoids). Given the heterogeneity of certain diseases (e.g. tumor heterogeneity), better patient profiling could also aid in increasing drug efficacy. Could the lack of assessment of spatial temporal dynamics be contributing to drugs failing efficacy?
Would greater investment in understanding the biology lead to increased drug discovery success?
Drug discovery is prominent for monogenic disease (e.g. cystic fibrosis) as the cause is known and a single target can be sought. In the case of complex disease (e.g. Schizophrenia) the detailed mechanisms remain unknown, impeding drug discovery. Again the lack of information on wellness limits our understanding of the biological alterations that occur in disease. Machine learning is also impeded by disease heterogeneity as it requires many consistent data points. The uniqueness of cancer imposes complications for machine learning.
Would altering the drug discovery pipeline increase the success of drug discovery?
For many diseases there are no animal models which fully recapitulate human disease and therefore there is an advocacy movement to proceed directly from preclinical safety in animals directly into Phase I.
SESSION 3:THE FUTURE OF PROTEOMICS
This session was chaired by Krzysztof Palczewski, CIFAR Senior Fellow, Professor & Chair, Case Western Reserve University and included presentations by R.J. Dwayne Miller CIFAR0 Program Co-Director, CIFAR Senior Fellow & Director, Max-Planck Institute and by Daniel Figeys, CIFAR Senior Fellow, Chair of the department of Biochemistry, Microbiology and Immunology, University of Ottawa.
R.J. DWAYNE MILLER
Approaching the Fundamental Space-Time Limits for Cell Imaging: A Technology Roadmap to Making a Molecular Map of the Cell
Advances towards achieving fundamental limits to minimally invasive surgery and bio-diagnostics. Conventional lasers with long pulses cause heating and can burn tissue. Picosecond IR Laser (PIRL) overcomes this tissue burning issue as it energizes water molecules in the tissue for ablation. This technology enables surgery precision at the single-cell level with no scar tissue formation. Less activation of β-Catenin and TGF-β signaling with a PIRL laser is also observed, resulting in quantitatively less inflammation. During the utilization of this technology, it was discovered that fully intact proteins and viruses could be ejected from the samples with fully retained biological function.Given this observation, could this technique then be coupled with other techniques in order to further study these fully functional biomolecules and viruses?
Spatial mapping of tissue now possible.
Discovered new laser ablation process that leaves huge proteins intact in the gas phaseà 100nm core sampling with near single protein detection via mass spectrometry (MS). Future applications could include risk-free cornea transplant.
Through desorption by impulsive vibrational excitation (DIVE), PIRL can eject proteins into the gas phase from tissues which can then be analyzed by MS. This approach provides opportunity for in situ spatial mapping and the possibility of defining the boundary between cancerous and non-cancerous tissue.
A New Platform for Screening Compounds against Individual Microbiomes.
Introducing a high-throughput assay of individual microbiome to assess drug-microbiome interactions entitled RapidAIM (rapid assay of an individual’s microbiome) v2.0. This approach entails the ex vivo growth of an individual’s microbiota exposed to different compounds with assessment by metaproteomics. This method has yielded strong correlation between in vivo and in vitro response in terms of both the bacterial community composition and the affected pathways.
Proposing that all new drug candidates be tested for their effects on the microbiome. The microbiome response to different FDA-approved drugs differs among individuals. Nonsteroidal anti-inflammatory drugs (NSAIDs) can have drastic effects on the microbiome. Pathway changes occur even when the bacterial abundance is not changing.
INSIGHTS FROM GROUP DISCUSSION ON THE FUTURE OF PROTEOMICS
Should microbiome safety be required for drugs discovery?
Given that the microbiome response to different FDA drugs differs among individuals, should new drug candidates be required to undergo microbiome safety testing? Should drugs also be classified according to their effects on the microbiome? After exposure to antibiotics, some microbiomes are unable to recover and initial findings have demonstrated that certain FDA-approved drugs which are not classified as antibiotics can have antibiotic-like consequences on the microbiome. If this testing were to be implemented for drug discovery, the establishment of a “healthy” microbiome would be required.
Picosecond infrared laser (PIRL) has many potential clinical applications
PIRL is expected to cause no damage to the epithelial layer of the cornea as it has been introduced into the eye multiple times and demonstrated no scar tissue. Therefore PIRL may be of potential use for the treatment of cataracts as the current therapeutic method entails the use of a pick, the delivery of shock waves with aspiration which then leads to cornea damage. PIRL may also be amenable for liver biopsies and drug delivery. Could this technology be applicable for metastasis? Given that metastasis is inoperable is there value in removing metastasized tumours? Notably viruses can enter the gas phase when using the technology and thus is now classified as a biohazard; the Tobacco mosaic virus was able to infect a leaf.
SESSION 4: FUTURE OF STRUCTURAL BIOLOGY
This session was chaired by Daniel Figeys CIFAR Senior Fellow, Chair of the department of Biochemistry, Microbiology and Immunology, University of Ottawa with presentations by Oliver Ernst CIFAR Program Co-Director, CIFAR Senior Fellow & Professor, University of Toronto, and by R.J. Dwayne Miller CIFAR Program Co-Director, CIFAR Senior Fellow & Director, Max-Planck Institute
Current progress in understanding GPCR structure
Given that G-protein-coupled receptors (GPCR) represent a significant portion of drug targets, insights on their structure are valuable for improved drug discovery.
Approximately three per cent of the human genome consists of GPCRs and 34 per cent of prescription drugs are targeting ~100 GPCRs. The structures of 50 different GPCRs have been elucidated with only three GPCRs solved for active, inactive and intermediate. Inactive and active rhodopsin structures have been determined, however no structural information is available for the intermediate states.
Tools for studying macromolecular structures. The Volta Phase Plate (VPP), a method for phase contrast in transmission electron microscopy (TEM) consists of a thin (~10nm) continuous carbon film at the back focal plane of the objective lens and enables in focus phase contrast. Electron paramagnetic resonance (EPR) and double electron-electron resonance (DEER) are methods which measure distance between spin labels.
R.J. DWAYNE MILLER
Approaches to High -Throughput Structure Analysis and Imaging in Solution
Mapping atomic motions with ultra-bright electrons: realization of the chemists Gedanken Experiment. The objective is to develop a reaction mode basis for chemistry, beyond pushing arrows and address the missing physics of reduced dimensions. Ultra-bright electron sources have contributed to achieving the long standing goal of observing atomic motions during chemical processes with the ability to achieve 10-femtosecond time scale resolution. During the in situ bacteriophage T4 infection process the full and empty bacteriophage can be discriminated. Throughout the 30-minute imaging of viral release, different orientation of the same particle can be sampled and individual virus structure superimpose well with cryo-EM/tomography data.
INSIGHTS FROM GROUP DISCUSSION ON THE FUTURE OF STRUCTURAL BIOLOGY
Cryo-EM is a valuable tool for studying biological structures and further improvements of the technology and by combining with other technologies will further improve drug discovery. Although improved software and detectors can improve upon the current cryo-EM technology, sample preparation is a limiting factor. Samples must be homogenous and this aspect must be taken into account when relating the results back to biology. To gain a comprehensive picture of drug interaction static pictures are not sufficient. Capturing movies would aid in assessing whether a drug reaches and binds its target.
Download the PDF