Periodic Reporting for period 2 - EVO-NANO (Evolvable platform for programmable nanoparticle-based cancer therapies)
Reporting period: 2019-10-01 to 2022-03-31
Medicine relies on the use of pharmacologically active agents to treat diseases. However, drugs are not inherently effective; their benefit is directly coupled to the manner by which they are administered. It affects drug pharmacokinetics (PK), absorption, distribution, toxicity, duration of therapeutic effect, metabolism and excretion. In the ideal case, drugs would be applied at exactly the therapeutic concentration and would precisely target desired cells or tissues. However, drug delivery, release rate, targeting and stability are not easily controlled and are difficult to predict. To address these limitations, drug delivery systems (DDS) have been designed using a wide array of materials and chemical strategies.
Recently, much progress has been made in developing nanoparticle (NP)- based DDS with the specific aim to improve site-specific targeting of drugs, thus improving efficacy and reducing side-effects. However, in the present state of the art, the rates at which new efficient NP-based DDS are designed are exceedingly small and costly. As a consequence, even though revolutionary new drug delivery strategies can be conceived, throughputs are far too low to approach widespread medical applications. Therefore, our targeted scientific breakthrough is to establish, for the first time, an integrated platform for artificial intelligence (AI)-powered design of novel NP-based DDS, their synthesis and rapid testing, both in vivo and in vitro on a newly developed microfluidic-based platform.
In short, overall goal of EVO-NANO is to create new foundations for rapid development and assessment of new anti-cancer treatments. Since developing new DDSs costs much less and is much shorter than developing new drugs, EVO-NANO’s focus on accelerating production of new DDSs is in line with one of the key strategic challenges defined in the European Commission’s Strategic Plan for Health and Food Safety: (Specific objective 1.3 Achieving greater cost-effectiveness).
Overall objectives of the project are:
Objective 1: To develop a new class of open-ended evolutionary algorithms to creatively assess different cancer scenarios and autonomously engineer effective NP-based solutions to them in a novel way.
Objective 2: To implement a computational platform for the autonomous generation of new strategies for targeting CSC surface receptors using functionalized NPs. In its final form our model will simulate all the main aspects of NP dynamics: their travel via blood streams, extravasation, tumour penetration and endocytosis.
Objective 3: To streamline synthesis of functionalized NPs suggested by the computational platform.
Objective 4: To develop an integrated platform for validation of efficacy of the artificially evolved nanoparticle designs. It will be composed of (i) tumour microenvironments on microfluidic chips that will mimic major physiological barriers for NP tumour delivery and (ii) in vivo pre-clinical tests.
Recently, much progress has been made in developing nanoparticle (NP)- based DDS with the specific aim to improve site-specific targeting of drugs, thus improving efficacy and reducing side-effects. However, in the present state of the art, the rates at which new efficient NP-based DDS are designed are exceedingly small and costly. As a consequence, even though revolutionary new drug delivery strategies can be conceived, throughputs are far too low to approach widespread medical applications. Therefore, our targeted scientific breakthrough is to establish, for the first time, an integrated platform for artificial intelligence (AI)-powered design of novel NP-based DDS, their synthesis and rapid testing, both in vivo and in vitro on a newly developed microfluidic-based platform.
In short, overall goal of EVO-NANO is to create new foundations for rapid development and assessment of new anti-cancer treatments. Since developing new DDSs costs much less and is much shorter than developing new drugs, EVO-NANO’s focus on accelerating production of new DDSs is in line with one of the key strategic challenges defined in the European Commission’s Strategic Plan for Health and Food Safety: (Specific objective 1.3 Achieving greater cost-effectiveness).
Overall objectives of the project are:
Objective 1: To develop a new class of open-ended evolutionary algorithms to creatively assess different cancer scenarios and autonomously engineer effective NP-based solutions to them in a novel way.
Objective 2: To implement a computational platform for the autonomous generation of new strategies for targeting CSC surface receptors using functionalized NPs. In its final form our model will simulate all the main aspects of NP dynamics: their travel via blood streams, extravasation, tumour penetration and endocytosis.
Objective 3: To streamline synthesis of functionalized NPs suggested by the computational platform.
Objective 4: To develop an integrated platform for validation of efficacy of the artificially evolved nanoparticle designs. It will be composed of (i) tumour microenvironments on microfluidic chips that will mimic major physiological barriers for NP tumour delivery and (ii) in vivo pre-clinical tests.
Main Achieved results:
1. Developed evolutionary algorithms based on novelty search and applied to artificial evolution of functionalized nanoparticles
2. Developed initial implementation of the multiscale tumour simulator as a combination of agent based and stochastic based modeling approaches. Tumour model is based on the modified PhysiCell model. The PhysiCell model was modified by introducing heterogeneity on cell division in addition to the introduction of cancer stem cells as a new type of cells. The model was enriched with simulations of the vasculature of the tumour, used to estimate the delivery of nanoparticles into the tumour region. Interaction of nanoparticles and cancer cells was simulated and analysed using the STEPS model, which allowed for the consideration of spatial heterogeneity of cells. We have also developed the protocols for optimisation of sets of functionalized nanoparticles through surrogate-assisted evolutionary algorithms. The algorithms were proven to outperform the standard genetic algorithms. We demonstrated that the use of efficient evolutionary algorithms within a high-throughput computing approach allows us to uncover therapeutic design optima that maximise tumour regression.
3. Developed first two microfluidic microchips:
a. Vascular microchip: The artificial blood vessels were fabricated using microfabrication technologies. A microfabrication process for polymers was implemented for the construction of the vessel scaffold which corresponds to microchannels that reproduce as close as possible the geometry of the tumour capillaries. The channel scaffold is then covered with a monolayer of endothelial cells forming the endothelial barrier of the artificial capillary.
b. Cell microarray chip: Protein micro patterns arrays have been develop as tool to isolate individual cellular entities on the precise locations defined by the patterns for independent observation of the nanomedicine therapeutic effect.
1. Developed evolutionary algorithms based on novelty search and applied to artificial evolution of functionalized nanoparticles
2. Developed initial implementation of the multiscale tumour simulator as a combination of agent based and stochastic based modeling approaches. Tumour model is based on the modified PhysiCell model. The PhysiCell model was modified by introducing heterogeneity on cell division in addition to the introduction of cancer stem cells as a new type of cells. The model was enriched with simulations of the vasculature of the tumour, used to estimate the delivery of nanoparticles into the tumour region. Interaction of nanoparticles and cancer cells was simulated and analysed using the STEPS model, which allowed for the consideration of spatial heterogeneity of cells. We have also developed the protocols for optimisation of sets of functionalized nanoparticles through surrogate-assisted evolutionary algorithms. The algorithms were proven to outperform the standard genetic algorithms. We demonstrated that the use of efficient evolutionary algorithms within a high-throughput computing approach allows us to uncover therapeutic design optima that maximise tumour regression.
3. Developed first two microfluidic microchips:
a. Vascular microchip: The artificial blood vessels were fabricated using microfabrication technologies. A microfabrication process for polymers was implemented for the construction of the vessel scaffold which corresponds to microchannels that reproduce as close as possible the geometry of the tumour capillaries. The channel scaffold is then covered with a monolayer of endothelial cells forming the endothelial barrier of the artificial capillary.
b. Cell microarray chip: Protein micro patterns arrays have been develop as tool to isolate individual cellular entities on the precise locations defined by the patterns for independent observation of the nanomedicine therapeutic effect.
We expect that the platform that will be created within EVO-NANO project will initiate the development of similar platforms that can autonomously evolve their own solutions to other complex nanomedical challenges like in vivo imaging or other drug delivery systems. On a more general scale, EVO-NANO will progress the state-of-the-art of modelling behaviour of changeable agents in dynamic, uncertain and complex environments. Such dynamics is a challenge to team operations in other applications including software engineering, logistics planning, advanced hardware engineering, swarm robotics and intelligence forecasting. It is envisioned that EVO-NANO will provide transferable principles for the development of new, semi-autonomous applications in those fields.