This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.
Cognitive decline caused by neurodegeneration is emerging as one of the greatest health threats of the twenty first century. As life expectancy of the population increases, so does the prevalence of cognitive impairment and dementia. Tackling the basis of neurodegeneration in the elderly becomes critical. Thus, extensive research has been conducted in the last decades to disclose the molecular mechanisms that lead to neuronal dysfunction and degeneration, yielding to significant breakthroughs in the understanding of several neurodegenerative diseases. As research progresses, many similarities appear which relate these diseases to one another on a sub-cellular level. These include atypical protein assemblies and induced cell death . In the case of Alzheimer's Disease (AD), the most predominant form of dementia worldwide, the neurodegenerative process is caused by accumulation of abnormally folded amyloid-beta (Aï¢) and tau proteins in the brain. Forefront findings also point out to an altered function in the protective role of glial cells (the non-neuronal cell population in the brain) . However, the numerous and intricate signalling pathways that eventually cause neuronal loss are still not fully understood. Despite of intensive efforts in developing drugs to treat Alzheimer's, an effective cure remains still unachieved. Many promising drugs with evident benefits when tested in pre-clinical trials using in vitro cell-based assays or animals, fail when administered to patients. This is partly because reaching the human brain by pharmacological interventions is not an easy task, since few compounds can cross the highly selective tissue barrier that protects the brain, the so-called blood-brain barrier (BBB). Another reason is because the testing conditions differ from the complex human brain physiology. In this regard, the ability to control accurately the cellular environment and intercellular communication of brain cells in vitro while recreating the hallmarks of the condition and monitoring neuronal response is of paramount importance, as it would provide a realistic representation of the in vivo milieu and the means for getting insights into the signalling cascades involved in AD.
The accelerated technological progress is opening new possibilities for handling, controlling, assessing and actuating into biological systems. One clear example is the avenue of soft lithography, a set of microfabrication techniques based on microstructuring elastomeric polymers. Soft lithography-based techniques have tremendously expanded the capabilities in the fabrication of microstructures for biological applications. In particular, these techniques offer the tools needed for mimicking the tissue environment in vitro: they enable the control and patterning of complex molecules relevant to biology, the fabrication of channel structures appropriate for microfluidics, and the patterning and manipulation of cells. Another example is the emerging field of bioelectronics, which exploits the synergy of interfacing biological materials with micro- and nano-electronics. In particular, advances in the control over the microfabrication of conductive polymers has given rise to innovative organic devices, such as the planar organic transistor or the organic electrochemical transistor (OECT), which are excellent bio-interfaces and can sense biological signals with capabilities well beyond the state-of-the-art . Conductive polymers have arisen as more performant electrode materials to interface with neurons, as they offer higher bio-compatibility, ability to measure ions and currents, and capability of functionalization with cell adhesion molecules . The implementation as organic transistors presents the additional advantages of providing local signal amplification, resulting in higher signal-to-noise ratio and the possibility of discriminating neuronal signals that electrodes cannot record. Moreover, higher integration can be achieved thanks to the ability of multiplexing the signals from an array of transistors.
The objective of this project is to combine cutting-edge knowledge in the disciplines of neurosciences, microfabrication and bioelectronics, in order to fabricate a realistic in vitro model of the brain where signalling pathways leading to Alzheimer's Disease can be assessed and breakthrough new therapies can be screened. The MBA-chip is schematically represented in Fig. 1. More in detail, the platform aims at integrating a model of the BBB in the first module, connected by microfluidics to a second module that models the central nervous system (CNS). The CNS will contain several cell compartments with different CNS cells intercommunicated by microfluidics where the flow of endogenous and exogenous factors between cell types will be controlled by on-chip valves. The choice of adequate endpoints is of paramount importance: The platform will be compatible with immunochemistry, a reliable technique to visualize the effect of different environments on cells; however,
Fig.1. Schematic layout of the Micro Brain Analog.
Axonal guidance chamber
for immunolabelling observations, cells must be normally fixed, which makes it impossible to study degeneration and regeneration on the same system. For that purpose, the system will integrate an organic electrochemical transistor (OECT) in the BBB to characterize the barrier integrity, and organic multitransistor arrays (OMTA) in the neuronal chambers to provide real-time measurements of the activity of the neuronal networks in terms of suffering and regeneration. Finally, such configuration aims at investigating the cross-talk between different cell types in response to stimuli and drugs involved in AD, in order to elucidate the mechanisms of neurodegeneration and possible regeneration. For this purpose, the microenvironment and tissue/cell interactions in the frame of AD will be recreated: Endothelium, neurons and glial cells will be challenged by Aï¢ to reproduce AD pathological hallmarks, and drugs that tackle different aspects of the neurodegeneration process will be tested.
This project aims at providing a microfabricated platform that integrates multifunctional modules in a single device and automated recording of neuronal response for high throughput. The development of this more bio-mimetic in vitro model will have numerous impacts in several fields. In the first place, it will tremendously benefit fundamental investigations of the pathogenic roots of AD by allowing to analyse the mechanisms of cell/tissue interactions as well as to test different hypotheses for AD onset and progression. In addition, it can be easily adapted to the study of other neurological disorders. Slight changes to the layout of the chip and the protocols for inducing disease effects could be readily implemented, for instance for studying Parkinson's or Huntington's diseases. Therefore it can be thought of as a new tool for neuroscience research that guides towards the elucidation of the cellular pathways involved in neurological diseases. Moreover, the neuropharmacology research will have a high payoff from the outcomes of this project. The MBA-chip will allow a reliable and high-content testing of promising targets for next generation drug therapies, improving the predictive power in the pre-screening of compounds, reducing dependency on animal studies and prioritising new drug candidates that are more likely to be successful in clinical trials. Such a configuration will speed up and reduce the costs of the drug discovery process by providing an accurate vehicle to assess drug effects, thereby contributing towards the end or reducing the global burden of brain disorders in society. Last but not least, the platform could be used as a valid cell-based toxicology assay for neurotoxicity testing. The mechanisms of neurotoxicity of chemicals, biological agents and nanomaterials would be better understood and evaluated with a non-destructive end-point and the target cell of the agents identified. This would pave the road for innovative miniaturized assays that consume fewer amounts of materials and yield unprecedented realistic information, which could be used as alternative methods to animal testing that are reliable, less time consuming and cost efficient. The platform developed in this project could be used straightforward for neurotoxicity testing, and in future goals could be scaled up to fabricate multiple platforms in the same support and be automated for high throughput screening.
Alzheimer's Disease and neurodegeneration
Neurodegenerative diseases are characterized by dysfunction and death of specific neuronal populations and, most often, by the presence of intracellular and extracellular aggregates of otherwise physiological soluble proteins. This is a common mark observed in post-mortem tissue analysis in several neurodegenerative disorders, such as Alzheimer's Disease (AD), Parkinson and Huntington's diseases and Amyotrophic Lateral Sclerosis. Another recently accepted common feature is the important role that non-neuronal cells, i.e. glial cells such as astrocytes or microglia, play in the evolution of the diseases.
AD is the leading cause of dementia worldwide and afflicted 26.6 million people in 2006 with projections of a fourfold increase by 2050 . Alzheimer's is characterized by a disruption of synaptic function leading to a progressive cognitive decline, memory deficit and behavioural changes. The main hallmarks of the pathology are the extracellular accumulations of amyloid-ï¢ (Aï¢) in neuritic plaques and intracellular neurofibrillary tangles of hyperphosphorylated tau protein, a microtubule-associated protein. Also, inclusions of activated glial cells are present near the amyloid plaques, and significant loss of neurons and synapses has been found in several regions of the diseased brain. Fostered by the strong social impact, a large number of studies have focused on the pathological mechanisms underlying the condition. Neuronal death in AD has been correlated to deficits in neurotrophic support, to alterations in glucose metabolism, and to the toxicity of Aï¢, specially from increased levels of oligomeric Aï¢, which are responsible of synaptic loss even before neuritic plaques and neurofibrillar tangles are formed. Besides, neuronal death is triggered by a huge number of biochemical signals involved in apoptosis, from oxidative stress and mitochondrial dysfunction to inflammatory processes involving dysfunction of glial cells.
The key aspects of AD to be addressed in an in vitro model are the toxicity of Aï¢ and the implications of glial cells in the acceleration of the pathology.
Fig. 2. Brain slice of APP-transgenic mice. Reactive glia surround Aï¢ plaques
MicrogliaAï¢ is produced from the transmembrane Amyloid Precursor Protein (APP). Recent research suggests that the crucial step in developing AD is related to the generation of neurotoxic Aï¢ from sequential proteolysis of the APP by the ï¢- and ï§-secretases. The abnormal overproduction of Aï¢ leads to its aggregation, at first in the form of small, soluble and potentially more neurotoxic oligomers , which eventually are recruited to form ordered, stable amyloid fibrils and deposited in neuritic plaques. The reactivity of Aï¢ is thought to be the main reason of its high ability to disturb brain function. Aï¢ is able to bind to a variety of molecules, including lipids, proteins and proteoglycans. Membrane-bound oligomeric Aï¢ can diffuse laterally and cluster at excitatory synapses inducing the anomalous accumulation of glutamate receptors which elevates the intracellular calcium levels and causes synapse deterioration . Besides, the ability to associate with and affect the integrity of lipid membranes causes membrane invaginations and the formation of pore structures and ion channels , and may also compromise the integrity of the BBB. Moreover, Aï¢ toxicity can also explain other pathological aspects in AD including oxidative stress, inflammation by the local activation of glial cells, and the formation of neurofibrillary tangles by controlling the cleavage and phosphorylation of tau protein .
Glial-cell alterations, such as gliosis in astrocytes and activation of microglia, are increasingly recognized in the pathogenesis of several neurodegenerative diseases. Neuroinflammatory processes, in which reactive glial cells release several proinflammatory mediators, are a common protective response to brain damage. There is substantial evidence that these reactive and inflammatory responses make significant contributions to extend, rather than protect, neuronal damage. In the case of AD, activated microglia and astrocytes are characteristically present in abundance surrounding the amyloid plaques (Fig. 2.) and may play a relevant role in the acceleration of the pathogenesis. As a result of ageing, astrocytes abandon their house-keeping function and senesce, a process in which they stop dividing and start a path of destruction while secreting their inflammatory brew. Recent findings point out that senescent astrocytes peak up to high levels in the presence of Aï¢ plaques . Microglia, on the other hand, is considered the resident innate immune cells in the CNS. Although activated microglia is needed for phagocytosing apoptotic cells, thus clearing toxins and limiting the damage to adjacent tissue, prolonged microglial activation causes a chronic inflammation that contributes to the exacerbation of AD. Several lines of evidence suggest that neurons are directly killed through neurotoxic molecules elaborated from glial activated cells, and that these inflammatory processes are further amplified by toxic Aï¢.
The toxic accumulation of Aï¢ and its link to the glial cell pathology hypothesis are receiving much attention in the AD research community. Aï¢ aggregates not only affect directly to neurons, but also provoke a glial-mediated inflammatory response that contributes significantly to the cell loss and cognitive decline. Aï¢ activated glial cells are not able to recycle the amyloid deposits thus producing the enlargement of the amyloid deposits. Nonetheless, the crucial mechanisms that cause this vicious circle are still unclear. In-depth investigations of neuronal-glial interactions are needed to improve the understanding of the fundamental nature of human neurodegeneration. The innovative recording system that would be implemented in the Micro Brain Analog could provide extraordinary information about the kinetics of neuronal responses and the precise interactions of Aï¢ with the different CNS cell types: the system could record interactions of Aï¢ will cell membranes and subsequent formation of pores and ion channels, thus contributing to the comprehension of the basis of the pathology. Moreover, neuronal-glial interactions and their potential pathways relevant for the study of AD could be investigated in a well-controlled manner with the microfluidic system presented in this proposal.
Microsystems in neurodegeneration
In their normal environment, cells are subject to multiple cues that vary in time and space, including gradients of cytokines and secreted proteins from neighbouring cells, biochemical and mechanical interactions with the extracellular matrix, direct cell-to-cell contacts, and physiological shear stresses from fluid flow. Microfabricated systems can present cells with these cues in a controllable and reproducible fashion that cannot easily be achieved by standard tissue culture . They can also be used to link cell culture with integrated analytical tools that can monitor the biochemical/biophysical processes that govern cell behaviour. Thus, microscale systems are well suited for the development of in vitro models that mimic more closely the human tissue in vivo, with the additional advantage of using reduced amount of cells and reagents, therefore decreasing the impact on the use of animal or human tissue.
The introduction of soft lithography by Whitesides  in the late 90's put microtechnologies within reach for biologists. The technology rapidly matured and, since then, many valuable biological applications have stemmed from it. Soft lithography embraces a collection of techniques based on printing, molding and embossing with an elastomeric stamp, typically poly(dimethyl siloxane) (PDMS). All these techniques can be used and combined to control the physical chemical environment of the cells and their interactions, thus mimicking more closely what happens in a 3D tissue. Fundamental biology research, including neurobiology, largely benefits from the capability of soft lithography based micropatterning and microfluidics to positioning anchorage-dependent cells on a surface with control over their size and spatial arrangement [13,14]. In particular, microfluidics has been widely used for culturing neurons, finding promising applications in neuropharmacology, neuroelectrophysiology, neural stem cell differentiation, and developmental neurobiological studies . Microfluidics enables the manipulation of fluids in channels with dimensions of tens of micrometers. Besides directing the placement of cells and small organisms, the technology also allows creating spatial-temporal controlled environments that mimic the conditions found in vivo. Moreover, microfluidic devices are compatible with high-resolution microscopy, and the introduction of functional modules, such as valves and pumps, have made them more practical for handling complicated processes in a miniaturized and automated fashion. Many microfluidic platforms have been designed around the unique anatomy of cells from the nervous system to study for instance nerve cell activity, or growth on various substrates and microenvironments. These designs offer novel approaches for studying the cellular physiology of the nervous system and the pathophysiology of neurodegenerative diseases. For instance, cell-to-cell contribution to a particular pathology can be studied by allowing the communication of different cells uniquely through the diffusion of soluble factors , or through contact between specific cell parts (axon-glia , axon-neurons ). Also, injury and regeneration of CNS axons can be monitored using compartmentalized microfluidic platforms to guide neuronal growth, isolate distinct neuronal components, i.e. cell bodies, axons, dendrites or synapse, and apply localized treatments to each sub-cellular compartment . In addition, three-dimensional cultures can be implemented by the patterning of hydrogels inside microchannels, where primary cortical neurons are embedded in the hydrogels with the foresight of patterning the layered structure of neural cells in the brain .
Such microfabricated devices have facilitated both applied and basic research into the biology of cells and tissues. However, integration and automation must still be achieved. The rationale for integration includes greater accuracy and reproducibility, smaller sample sizes, and higher throughput. Rather than only monitoring simple phenotypic changes, future integrated systems should be able to gather precise biochemical and mechanistic data from cells and tissues .
Drug screening of neuroactive compounds:
Neuropharmaceuticals is potentially one of the largest sectors of the global pharmaceutical market due to the increase in average life expectancy and the fact that many neurological disorders have been largely refractory to pharmacotherapy. Despite such emerging demands for treatment of CNS diseases, only 7% of CNS drugs in clinical development are successful, as compared to the 12% average across all therapeutic areas . Drugs that treat diseases of the CNS are among the ones with most reduced success partly because traditional screening tools, based on static assays in multiwall microtiter plates, are not good representatives of human metabolism . When a drug is administered, it undergoes complex processes of absorption, distribution, metabolism, and elimination, which determine the amount and nature of the drug that ultimately reaches the target tissues. The interaction of the different cells exposed to the drug also contributes to the result. Therefore, the present trend in pre-clinical evaluation of the efficacy and toxicity of new drugs is to use testing systems in which several cell lines from different tissues interact with one another by the flow of metabolites between cell culture chambers.
One of the causes of the high cost of CNS-targeted pharmaceuticals is the lack of experimental model systems that can reduce or replace costly and time-consuming animal studies. There is a high demand to develop alternative in vitro methods for initial screening to prioritize treatments for further testing. The value of many promising drug candidates is diminished by the presence of barriers between blood and brain, commonly referred to as the blood-brain barrier (BBB). Any drug discovery or development programme involving compounds targeted to the CNS needs to take the properties of the BBB into account to achieve relevant CNS exposure. The BBB is a dynamic interface that regulates the exchange of substances between blood and brain and maintains optimal physiological conditions for neuronal and glial function. The same mechanisms that efficiently protect the brain from harmful compounds are major obstacles in drug development and testing. Existing model systems do not faithfully recreate the active tissue-tissue interface between the microvascular endothelium and neighbouring parenchymal tissues (neurons and glial cells), where critical transport of fluids, nutrients, immune cells, and other regulatory factors occur. The BBB is also implicated in neurodegenerative disorders, in particular in Alzheimer's disease, in which its dysfunction may result in compromised transport and permeability, and can also lead to the release of pro-inflammatory molecules that may cause further CNS damage. Strategies to restore or protect the functionality of the BBB in various CNS pathologies in order to maintain brain homeostasis and normal signalling between brain cells may be of important therapeutic value.
Due to the complexity of the cross-talk between many different signalling cascades triggered in different type of cells that eventually cause neuronal dysfunction and loss in AD, a medication treating the underlying causes of Alzheimer's has not been found yet. The current five FDA-approved Alzheimer's drugs are targeted to inhibit acetyl-cholinesterase or to block the glutamate N-methyl-D-aspartate (NMDA) receptor. They treat only the symptoms and help to temporarily improve cognitive function . New therapeutic approaches targeted directly against Aï¢ toxicity include eliminating its production using secretase inhibitors or enhancing its clearance by Aï¢ immunization. Likewise, controlling glial activation and hence inflammation is a promising therapeutic strategy which opens up a new road for exploring immune system-based therapy of neurodegenerative diseases . Another treatment under investigation is the use of medications to fight insulin resistance, supported by recent studies showing that brain insulin resistance and thus alteration in glucose metabolism contribute to cognitive decline .
The future pharmacological strategies are believed to involve a cocktail of medications that impact on the multiple biochemical alterations involved in the onset and progression of the disease. To speed-up the identification of the most promising drug candidates, innovative and more accurate in-vitro models are needed that can help identifying mechanisms and modes of action, thus speeding up the development process and reducing costly and time-consuming animal experiments. These new models must include the complex intercellular signalling as well as the important contribution of the BBB to reach a better understanding of the modes of action, needed to develop an effective drug activity.
Microsystems in drug screening of neuroactive compounds: the capabilities of organic electronics
Biomimetic microsystems that reconstitute cell-to-cell communication and tissue-tissue interfaces critical to organ function may expand the capabilities of cell culture models and contribute to the reduction of animal testing and clinical studies. In this context, the concept of micro cell culture analogs (ïCCA), i.e. models that replicate aspects of the human body as close as possible, is being developed thanks to microfluidic technology. ïCCA are considered as the physical representation of pharmacokinetic models. Various microscale body-on-chip cell culture models have been fabricated [11,35], and proved the advantage of simulating the exchange of metabolites for drug toxiticy . However, a micro brain analog that includes both the BBB and the CNS for drug screening of neuroactive compounds is still missing. Among the barrier tissues, the BBB is particularly difficult to reproduce in vitro. The transendothelial resistance (TEER) of this decisive restrictive barrier has been measured to be as high as 1000 ï- cm2. In static models, brain endothelial cells are co-cultured with astrocytes in transmembrane wells, since astrocytes provide an important stimulus for the establishment of the barrier function. Furthermore, shear stress influences the differentiation of the barrier. For that reason, new dynamic models culture the cells in a porous tube to provide fluidic shear stress. Recently, a microfluidic in vitro model of the BBB was reported and tested using cell lines . These models, however, lack either achieving a close contact between astrocytes and endothelium or providing fluidic shear stress and physiologically relevant communication with the CNS. The interest of an on-chip fabricated BBB lies on the fact that it enables the integration of in vitro BBB models with other organ analogs to construct a ïCAA that can be used to test drugs that target diseases of the central nervous system. A promising approach that will be explored in this project would be using a biomimetic hydrogel model , where multiple cells can be cultured within and on the surface of hydrogel matrices allowing the contact between different cell types.
In addition, the drug screening of neuroactive compounds cannot be limited to cell viability, since neurons may be damaged and lose their function well before they die. Therefore, electrophysiological recordings, such as the ones provide by multielectrode arrays (MEA), incorporated in the brain model are crucial to provide information of early neuronal events. High-throughput measurement of neuronal activity by MEA provides a powerful strategy to characterize therapeutic molecules in drug discovery and to identify harmful agents in neurotoxicological studies. By means of multiple microelectrodes embedded in a substrate cultured with neurons, MEA allow for simultaneous recordings to be obtained from many cells revealing how the neurons communicate with one another across the neuronal network. They record changes in extracellular electrical potential, more specifically action potentials from nearby units, and have been successfully used to test the pharmacological and toxicological response of CNS primary cultures . Neuronal activity recording is typically achieved by plating the cells on a MEA support made of an array of Gold (Au) or indium-tin-oxide (ITO) electrodes. On the other hand, conductive polymers are emerging as ideal electrode materials for interfacing with neurons. As compared to the metal counterparts, the polymeric electrodes have been shown to reduce the inflammatory response of the brain in implants and to lower the electrical impedance at the interface resulting in recordings of higher quality. Moreover, conductive polymers are able to sense both ionic and electronic currents and can be functionalized with biomolecules that stimulate neural growth. Recently, arrays of PEDOT:PSS, the state-of-the art conductive polymer in bioelectronics, electrodes have been used to measure the neuronal activity in vivo and proved to record the electrophysiological signals with higher accuracy than conventional metal electrodes . It is not unreasonable to speculate that the local signal amplification of PEDOT:PSS, if implemented in a transistor format would improve the sensitivity and monitor neuronal activity with higher fidelity, which could result in unforeseen found neuronal events not recorded before.
How this project will contribute to AD research. Challenges and feasibility:
Among all CNS disorders, AD neuropathology poses the highest heavy economic and social burden, and the impact on health care system will increasingly rise if no efficient therapeutic and early-diagnosis approach becomes available. We have seen the complexity of the problem: the numerous possible mechanisms leading to the development of the disease, and the difficulty reaching the brain by pharmacological interventions. To mimic precisely and controllably in vitro all the mechanisms described above, one has to take into consideration that glial cells are important contributors to disease pathogenesis and any drug administration dedicated to the CNS must effectively cross the BBB. An in vitro platform for culturing multiple cell types linked in a microfluidic system while maintaining their soma compartmental isolation from one another is an evident way forward to tackle these issues. Such platform should be designed to enable the controlled transfer of substances between cells (neurons-glia) and tissues (BBB-CNS), while allowing in situ diagnostics at each stage of the experiment. This research project aims at recreating a brain model on a chip to elucidate the main factors and processes involved in neurodegeneration and nerve regeneration, focusing the neurobiological investigation on the implications of cell communication in Alzheimer's disease. The objective is to model the BBB and the CNS cells communication on a chip to test the balance between neuroprotective agents and neurodegeneration using advanced organic transducers and microfabrication techniques.
The interdisciplinarity of this project, being 50% technological and 50% biological, together with the use of primary cells and the application to the neuroscientific field of neurodegenration in AD, makes this proposal very challenging. However, as expressed in critical reviews from well recognised experts [30,31], the joint work of neurobiologists and microtechnologists is crucial to advance in the knowledge of the nervous system and to create novel neuroengineering tools that will help to reduce the impact of ageing in society. The PI of this project has good practical experience in both fields and the knowledge needed to make the link between neuroscientists, physicists and engineers. This will increase the chances to develop a relevant tool, demonstrate its use, and transfer it effectively to the community of neurobiologists.
An advance needed in microfluidics based cell assays is to make systems easy to operate. Pumpless systems and integrated non-destructive read-outs go in this direction. Technologically, there are three major challenges: the microfabrication of the blood-brain barrier, the integration of organic transistors in the different modules of the device, and the fabrication and operation of on-chip valves, all requiring the fabrication of aligned multilevel photolithography. These challenges and their corresponding alternative plans are described in more detail in the methodology section. To confront these issues, the PI possesses a very strong background in the use of microtechnologies for life sciences, and she will be immersed in the Bioelectronics Department of EMSE, which has sound expertise in organic bioelectronics and the entire infrastructure needed for carrying out advanced microfabrication processes.
To increase the impact and acceptance of micro cell culture analogs, the developed devices must mimic authentic cellular and tissue behaviour. Cell lines do not always satisfy this requirement. The use of primary cells or tissue slices could solve this problem, although tissue slices might limit the throughput if intended to drug screening and toxicity. Primary cells derived from the central nervous system are among the most difficult type of cells to manipulate. Their heterogeneity, biochemical complexity and extreme sensitivity to physico-chemical changes in the environment makes their growth and manipulation cumbersome. Biologically, the main bottleneck will be to define the adequate protocol for growing the different type of cells on the same chip and inside micro-chambers (device pre-treatment and functionalization, programming of dissections to extract the cells, plating protocol, cell density, media). Contamination problems may easily appear due to the frequent handling of the micro-device for cell plating and medium exchange. The PI has worked in the integration of neural cells with microdevices for more than six years, in particular in the fabrication of microfluidic devices that incorporate primary CNS cells to perform toxicity tests of different Aï¢ species (oligomer, fibrils). The PI will count on her previous experience to tackle this challenge, but will also be in contact with experienced groups in neuroscience and stem cells for advising and supporting the studies of neurodegeneration (University of Milan, Prof. Matteoli, and Polish Academy of Sciences, Prof. Buzanska).
These plans show the complexity and difficulty of the project. It requires the access to and support of state-of-the-art microfabrication and engineering facilities and staff, as well as high-level biology and biophysics labs. The integration of the different parts of the work requires the necessary scientific background and environment that can be started only with a medium term perspective and the necessary infrastructure, which is available at EMSE. These aspects can be addressed with the help of the ERC programme. Multidisciplinary groups like the one that will be built in this project are key to solving the societal and financial impact of ageing in our communities.
Section b. Methodology
The envisioned brain analog system (Fig. 1) consist of a BBB unit and a CNS unit. The microfluidic platforms will be fabricated by soft-lithography techniques using replica molding in PDMS, and the whole system will be monitored by organic electrochemical transistors, which will register the formation of the barrier and the activity of the neuronal networks. Several lithographically engraved layers will be designed on PDMS to define microvalves that can be easily controlled by mechanical or pneumatic pressure. The microvalves will control the isolation/ communication between cell compartments by the actuation on the PDMS top layer to open/close the microfluidic channel. Each cell compartment in the CNS unit will host one cell population, i.e. microglia, astrocytes or neurons. The BBB compartment will contain a contact co-culture of endothelial cells and astrocytes. Insults and treatments will be applied on the BBB compartment and the by-products (endogenous and therapeutic substances) diffusing through the BBB layer, will be driven into the CNS compartments. Furthermore, thanks to the gravity-induced flow, it will be possible to reverse the flow direction to study the leakage of the membrane as a consequence of molecules coming from the CNS compartment.
The tasks to achieve the integration of the physical device with the biological entities and proof the concept are divided in 4 work-packages (WP) as follows:
WP1: Microfluidic integration of the BBB model, based on a three-dimensional co-culture and an organic electrochemical transistor read-out.
Objectives: Microfabrication of the BBB unit and integration with microfluidic channels to allow a bi-directional flow circulation.
The BBB unit will consist of a PDMS chamber that will host a suspended porous membrane with a layer of hydrogel on top of it. The hydrogel will contain the entrapped astrocytes and a monolayer of endothelial cells on the top. The barrier formation will be evaluated by TEER measurements using an organic electrochemical transistor (OECT). This compartment will be kept isolated until TEER measurements show that the BBB is fully differentiated. Then, samples will be added to the endothelial side and fluids will drag the diffused molecules towards the CNS area via a microchannel connection (Fig. 3).
Fig. 3. Microfluidic integration of the BBB model. AC = astrocytes, EC=Endothelial cells
AC in hydrogel
FlowTask1.1. Co-culture of BBB cells on hydrogels:
The biological entity of the BBB unit will consist of a contact co-culture of endothelial cells and astrocytes in a transwell culture system where the bottom well is made of PDMS. Current BBB models are based on a porous polymeric membrane with astrocytes grown on the bottom and endothelial cells on the top. One drawback of these membranes is that they do not allow astrocytes to directly contact the endothelial cells. To improve cell-to-cell contact, a layer of agarose-alginate hydrogel containing astrocytes will be deposited in the centre of the porous membrane. Then, a layer of endothelial cells will be grown over the membrane, and the co-culture will be placed inside a PDMS well using the format of transwell cell cultures. Bovine brain microvascular endothelial cells (BBMVEC) and the DI-TNC1 rat astrocytes cell line, both from HPACC, will be first used for setting the parameters of the 3D hydrogel-based co-culture and the operation of the BBB unit. In a later stage, Human Brain Microvascular Endothelial Cells (HBMEC) and primary human Astrocytes (HA), from ScienceCell Research Laboratories, will be tested. These cells are isolated from human brain tissue and can be expanded for at least 15 passages. The hydrogel scaffold will be based on the natural hydrogels agarose and alginate. Agarose is a thermosetting hydrogel that can be gelled thermally within seconds, while alginate is not a thermosetting gel and requires the addition of Ca2+ ions to form the hydrogel structure. Alginate-enriched agarose hydrogels enhance the survival of neural cells, as reported by Renaud et al. . Thermal gelation of the hydrogel mixture produces unbound alginate component that can attach extracellular matrix proteins, such as collagen, which is used to simulate the basal lamina interface between endothelium and astrocytes. Collagen will be added to the liquid hydrogel mixture together with the astrocytes suspension, and the 3D hydrogel scaffold will be solidified by quickly decreasing the temperature to 26Â°C. With the help of Prof. Renaud at EPFL, the concentration of the hydrogel mixture, collagen coating and astrocytes density will be tuned for optimising cell survival. Personnel from this project will do a short training at Prof. Renaud's laboratory in Lausanne to learn the methodology for the construction of 3D hydrogel scaffolds. The knowledge acquired from this training will be applied to the fabrication of the BBB cell model.
Task 1.2. Evaluation of the BBB integrity based on OECTs measurements:
Fig. 4. Principle of OECT recording.The characterization of the barrier function will be performed by an OECT that measures the TEER, which is one of the main indicators of the functional formation of the barrier. Dr. Owens, from the same department that will host this project, has reported for the first time the integration of OECTs with human cells for assessing the integrity of barrier tissues . The innovative measurement method exhibits much higher temporal resolution than conventional measurements of barrier disruption and is able to detect minute dysfunctions of the barrier layer. The principle of operation is depicted in Figure 4. Upon application of a positive gate voltage, the presence of intact barrier cells hinders the diffusion of positive ions into the polymer channel of the transistor, thus decreasing the drain current. The PI will rely on the close collaboration with Dr. Owens for the implementation of the OECT based TEER measurement. The OECT will consist of a conductive polymer channel made of PEDOT:PSS, and gold (Au) source and drain electrodes contacting the extremes of the PEDOT:PSS channel. These elements will be patterned on glass, to which PDMS will be attached. The PEDOT:PSS will be patterned by a photolithography process using parylene as sacrificial layer, as previously described by Malliaras and co-workers . The Au contacts are incorporated to facilitate the connection to the data acquisition system and will be isolated from the culture medium by a lithographically-defined insulating layer. The gate will be made of an Ag/AgCl electrode immersed in the top well of the transwell culture. The dimension of the OECT components will be based on Dr. Owens findings and will be scaled to the microfluidic device. The characterization of the BBB functionality will be done by measuring the TEER and by permeability tests using the impermeable marker Lucifer Yellow. Comparison between the TEER and permeability levels in control conditions (applying only medium) and in barrier disruption conditions (adding nanomolar concentrations of H2O2 to the medium) with reported literature values will be used as confirmation of the validity of the model. Then, different concentrations of oligomeric Aï¢ will be applied to the BBB and the change in TEER will be recorded.
Task 1.3. Microfluidic integration and BBB prototype testing:
The BBB unit will be implemented in a PDMS chamber with a microchannel to communicate with the CNS unit. To test the ability of flowing bypass products from the BBB to the CNS and vice versa, three PDMS wells will be interconnected via microchannels and covalently bonded to a glass substrate (Fig. 3). The transwell culture containing the porous membrane with BBB cells will be placed on top of the central PDMS well, while the second PDMS chamber will be free. The volume and density of hydrogel on the porous membrane will be adjusted to allow the flow of substances across the transwell. The height and volume of the chambers, as well as the z-position of the transwell insert, will be adjusted to achieve a gravity-induced flow from the BBB chamber to the neuronal chamber, or from the neuronal chamber to the BBB one. The bi-directional flow will be visually evaluated through the filling of the empty reservoir. To test that the BBB unit transports molecules properly, insulin will be added to the BBB, a difference of liquid level will be applied between the external PDMS chambers, and the insulin transferred between the BBB and the CNS-featuring well will be detected by ELISA. Insulin is a hormone that easily crosses the BBB in both directions and recent therapeutic strategies for AD are based on insulin administration.
Alternative plan: If the co-culture of endothelial cells and astrocytes cannot be achieved by hydrogel technology, standard cultures directly on the porous membrane will be used. Endothelial cells will by directly grown on the top of the porous membrane and astrocytes on the bottom, as current BBB models are implemented . Alternatively, only endothelial cells will be grown and a BBB inducing medium will be used. The latter approach also yields an accurate BBB model and is currently developed by the University of Artois, which collaborates with EMSE .
WP2: Development and characterization of organic detectors to monitor neuronal networks.
Objective: Fabrication and optimization of a multitransistor array, based on photolithographically patterned conductive polymers, for measuring extracellular action potentials of neurons.
Task 2.1. Fabrication, functionalization and integration of an organic multitransistor array (OMTA):
The chip will incorporate an array or planar organic transducers for continuously monitoring the experiments from the pre-exposed basal state until post-exposure and possible recovery. The integrated electrical read-out will be fabricated on a glass support, onto which the PDMS-based microfluidic chips will be covalently bonded. The centre of the cell chamber will be bare to ensure the compatibility with high-resolution light microscopy, and thus, with non-electrical common neurological end points (immunofluorescence, calcium imaging). The OMTA recording system will be placed around the cell chamber and will consist in an array of 36 organic transistors fabricated as described in Task 1.2, with the particularity that the gate electrode will be a common Au or Ag/AgCl electrode patterned on the side of the cell chamber (Figure 5). The metal electrodes can provoke an induced inflammatory response and consequent neurotoxicity; therefore, all the metal electrodes will be isolated from the culture medium by a PEDOT:PSS or parylene coating to ensure biocompatibility. A customised plug-head will be designed to convert the output currents to voltage and to interface with the data acquisition system: an amplifier chip (Intan), which is also able to multiplex the signals from the transistors array, and an A/D converter. The architecture of the individual transistors, the layout of the array, and the interface to the monitoring station will be optimised to maximise the sensitivity and minimize the signal-to-noise ratio. The optimization work will be based on the work currently carried out at EMSE and will be done previously to the integration of the OMTA in the microfluidic platform. For that, an OMTA will be fabricated on glass and a ring made of PDMS will delimit the array area. Neurons will be plated inside the PDMS ring.
Achieving a good reproducibility of the assays is highly important for drug screening purposes. Electrophysiological tests for screening the behaviour of neuronal networks are based on randomly attached cells to the MEA platform. The reproducibility of the assays could be improved by directing the attachment of neurons into the measurement active area (the PEDOT:PSS conductive polymer in our case). Such preferential attachment can be obtained by functionalizing the active parts with cell adhesion molecules. This project will also explore the benefits of functionalizing PEDOT:PSS with Poly-L-Lysine and Laminin in terms of favouring the organization of the network and improving the reproducibility of the data.
Task 2.2. Definition of neuronal end-points:
In MEA based recordings, the main parameters used to characterize the network activity are the firing rate (average number of spikes per second), the burst frequency (average number of bursts per minute), the duration of the burst, and the number of spikes per burst. These parameters will be recorded by the organic transistor array and use as end-points to characterize the neuronal activity. Due to the signal amplification in organic transistors and the capability of conducting ion and electron currents, differences between the electrophysiology raster plots registered by the organic transistors or the conventional MEA are expected, in particular regarding the improvement of the accuracy in reading the neuronal network thanks to the local signal amplification. To investigate this issue, a double array will be fabricated to register the same neuronal network. An array of bare Au electrodes will be next to a second array of transistors, and a PDMS ring will be placed around the arrays to hold the cell culture. It is possible that new relevant endpoints will be found.
Alternative plan: Multitransistors will be fabricated on glass and interfaced with the microfluidic support. These are more complex to fabricate than multielectrodes and require more electrical contacts to be interfaced with the data acquisition system. If the integration of organic transistors is not successful, a microelectrode array made of conductive polymer will be implemented. This will reduce the complexity of the fabrication process, but still benefiting from the advantages of interfacing organic materials with cells.
WP3: CNS unit and prototype operation via gravity-induced flow.
Objectives: To define the geometry of the CNS cell compartments and to implement on-chip valves to control the flow of substances.
Different cell chambers will be sequentially connected by microchannels. The functioning of the device will be based on gravity-induced flow and the use of valves, therefore avoiding the use of external pumps.
Task 3.1. Configuration of the cell compartments in the CNS unit:
The bypass products coming from the BBB unit will be divided in parallel flows in 4 microchannels and driven into a first line of 4 CNS chambers (labelled as "glial chambers" in Fig. 1). The first line of chambers that receive the stimulus downstream from the BBB will contain different type of cells (astrocytes, microglia), or will be a chamber without cells to be used as control. These chambers will be connected to a second line of CNS chambers that will host neurons. The CNS area will comprise two designs of chambers, represented in Figure 5:
Unstructured diffusion cell chambers: designed to perform cultures with controlled exchange of soluble factors between the different chambers. The nervous system cells will be placed in the centre of the compartment and the perfusion over the whole compartment will be carried out by two side channels, which will ensure that mass transfer between cells and medium is by diffusion only, resembling the in-vivo situation . The side channels will be separated from the central region by 5ïm wide micro-posts with a pitch of 10ïm.
Fig. 5. Diffusion (a) and axonal guidance (b) chambers
(b)Axonal isolation chambers: In which the configuration of the neurons compartment will foresee a region for the attachment of soma and an space consisting of an array of microgrooves (5ïm width) to separate the axonal projections from their soma into the neighbour compartment containing or not other type of cells (microglia, astrocytes). The axonal compartment will be within the group of the first line compartments, while the soma compartment will be considered as the second line. This configuration will be used to stimulate axons independently from soma, and monitor axonal degeneration and regeneration after exposure to the insults (Aï¢, and factors released from non-neuronal cells) applied to the soma or to the axons itself. Moreover, this configuration will suit applications in which neurite outgrowth and nerve injury and regeneration have to be monitored. For example in drug discovery of new treatments for spinal cord injury by testing compounds that promote axons to overcome the growth inhibitory environment of adult CNS .
Task 3.2. Control of the communication between compartments by on-chip microvalves:
Capillary soft valves  are the most appropriate to implement in the prototype, since they are very easy to implement and operate without any peripheral equipment. The valves stop the liquid flow of a microchannel by a capillary barrier that is created simply by abruptly expanding the cross section of the channel. The barrier is suppressed by mechanically pressing the top part of the valve with tweezers. As fabricated, the closed valves open after application of a mechanical pressure. At first, the microchannels connecting the cell compartments will have a cross-section of 50 x 50 ïm (width x height), the channels will be expanded to a width above 500ïm to form the valves, and the thickness of the PDMS microfluidic support will be 2mm in the valves and above 5mm around the cell chambers. The optimal dimensions of the valves and PDMS support will be studied during the course of the project. These valves cannot be closed once they have been opened. To close a flowing channel, pneumatic valves will be implemented. Pneumatic valves consist on a thin PDMS membrane that collapses the microfluidic channel upon applying pneumatic pressure . The PDMS membrane is positioned at the top of the channel to be collapsed, and the pressure to deform the membrane is driven by a top channel designed perpendicularly to the fluid channel direction.
Alternative plan: If the soft capillary valves cannot block the flow in this system, only pneumatic valves will be fabricated. If this is still unsuccessful, syringe pumps connections will be incorporated using a strategy similar to the one presented in our previous works for integrating open cell cultures with microfluidics .
Task 3.3. CNS cell cultures:
To fabricate the prototype and test the control of the microfluidic connections, SH-S5Y5 neuroblastoma, N9 microglia and DI-NTC1 astrocytes will be used first. The cell lines will be plated and grown in their compartments with the valves closed. The cell density and surface functionalization will be adjusted to each cell type. When the cells are mature for experiments, the medium will be unified as neuronal medium and the valves will be opened to put the different wells in communication. To test the cell-to-cell interaction and study the toxicity of Aï¢ on neurons, Aï¢ in the oligomeric and fibrillar form will be prepared and added to all the CNS compartments at increasing concentrations up to few micromolar. In one chip, the neuronal damage will be evaluated by propidium iodide staining with the valves opened. Then, in a different chip, a recovery stimulus will be applied to the first row of chambers and the medium containing the released factors will be transferred toward the second row chambers. Again, the neuronal damage will be evaluated by propidium iodide.
Primary neurons, astrocytes and microglia from rat will be used in a second stage, as described in WP4-Task 4.2.
WP4: Final integration of the BBB and CNS units and test of drugs for Alzheimer's disease.
Objective: To integrate the sub-prototypes in one chip and to test the effects of drugs on Aï¢ degenerated neurons.
To discern what is the communication pathway causing amyloid-induced neurotoxicity, stimuli will be given directly to neuronal cells and indirectly through the stimulation of glial cells and transport of the secreted milieu to the neuronal cells. Also, the efficacy of drugs addressed to different targets will be evaluated in these specific cross-talking situations, applying the drugs to the BBB compartment.
Task 4.1. Final prototype of the micro brain analog:
The MBA chip will be fabricated in one block and the function will be tested using well-known procedures. The cells indicated in Tasks 1.1 and 3.3 will be used for the first biological tests. As a proof-of-concept, H2O2 will be added to disrupt the barrier and glutamate at 100uM concentration will be used to cause neurotoxicity. The response of the network monitored by the OMTA will be used to define the recording profile of neurotoxicity and gradual switch off of the network.
Task 4.2. Definition of the AD on-chip model and drug testing:
Setting of the conditions for inducing an AD microenvironment will be based on previous experience of the PI in testing the toxicity of Aï¢ using microfluidic chips. Previous data (Figure 5) revealed that 2ÂµM concentration of oligomeric Aï¢ causes a significant neuronal loss in microscale cultures of primary cells. However, lower concentrations (500nM) are enough to affect synaptic plasticity. Therefore, oligomeric preparations of Aï¢ (500nM ~ 2ï€ ïM) will be added to the cell compartments and the toxicity recorded by the extracellular electrophysiology parameters learnt from Task 2.2. Aï¢ will be applied directly to neurons, and indirectly through previous incubation of Aï¢ on microglia, astrocytes, isolated axons, or the BBB. These tests will set the basis for the conditions during drug screening. When the toxic response is obtained, neuroprotective agents acting on different targets (neurons, glia or amyloid) will be flown from the BBB compartment and the eventual neuronal regeneration will be recorded. The neuroprotective agents for the initial tests are described below and summarized in Table 1.
Fig. 5. Neurotoxicity of different Aï¢ aggregates. Insert: Microfluidic chip used for testing, unpublished.
FTY720 (Fingolimod, Gilenya): FTY720 is a sphingosine-1-phosphate (S1P) receptor antagonist, commercialized under the trade name Gilenya by Novartis for the treatment of multiple sclerosis. Recent findings in the group of the PI have shown promising properties for treating the toxicity of Aï¢ (Fig. 3). FTY720 is an immunomodulating drug. The mechanism of action in Aï¢ challenged cells is still not know, but could be related to the inhibition of microvesicles shedding from microglia.
Memantine (Namenda, Merz): Memantine is a NMDA-receptor antagonist that acts on glutamate-mediated neurotransmission. It is approved by the European Medicines Agency to treat moderate to severe forms of AD.
Curcumin: Curcumin is a low-molecular weight natural phenol with potent anti-inflammatory and anti-oxidant properties. Administration of curcumin to APPsw transgenic mice has been proved to reduce amyloid plaque burden, while in vitro studies have demonstrated that curcumin is able to reduce the aggregation of Aï¢ and to block the toxicity of oligomers.
Insulin: Insulin is a hormone involved in the regulation of glucose metabolism and can cross the BBB in both directions. Recent studies have directly demonstrate that brain insulin resistance occurs in AD and new treatments are being developed based on the inhalation of repeated doses of insulin.
Primary neurons, astrocytes and microglia isolated from the hippocampus of rat brains will be used to perform these tests. The primary cells will be prepared according to the protocols used in the current laboratory of the PI. An animal facility is foreseen in the host institution, but it will be ready only during the last year of the project. Therefore, the PI considers that a collaboration with her present research group, lead by Prof. Michela Matteoli, is highly recommended. Prof. Matteoli holds more than 20 years of research in the field of neuroscience and is internationally recognized for her work on memory forming and synaptic plasticity. The current activities of the laboratory are pushing strongly towards the study of neurodegeneration in AD. The expertise of Prof. Matteoli's group will boost the relevance of this project and the acceptance in the neuroscience community. For these reasons, a PhD student will be trained during 6 months by the PI in the use of the micro brain analog (the formation of the BBB unit, the flow control, the preparation of Aï¢ and conditioning of the CNS cells, and the data acquisition) and will be sent to the Prof. Matteoli's laboratory to carry out the first drug tests on primary cell cultures and testing the latest molecules under investigations that may appear during the course of this project. In addition, based on the current experience of the PI in interfacing primary cells with microdevices, plating neurons in the micro brain analog for running 20 parallel experiments will only need 1% of the hippocampal neurons obtained from the sacrifice of one animal, producing approximately 10 million of wasted cells. This number is even higher for primary astrocytes. The left over cells could be used for other experiments if the collaboration with Prof. Matteoli's group is established.
Alternative plan: If transferring the prototype to incorporate primary rat cells is not achieved, mainly due to the availability of external complementary instrumentation, the project will use the latest models of human cells to run the Aï¢ and drug tests, such as Primary Human Astrocytes, Microglia and Neurons (ScienCell Research Laboratories) isolated from human brain tissue, or iCell Neurons (Cellular Dynamics) derived from human induced pluripotent stem cells. The use of advanced human cells in this device for real tests is straightforward. To date, these cells are expensive and neurons and microglia cannot be expanded. This is the reason why the prototyping has been planned with immortalized cell lines and primary rat cells.
Task 4.3. Cellular infiltrations:
The BBB is a selectively permeable tissue that prevents most polar compounds with molecular masses greater than 500 Daltons from entering the brain tissue. Only small molecules, lipophilic molecules, or those with specific transport mechanisms can pass. However, in brain metastasis, tumor cells are able to migrate through the endothelial cell layer. As well, infiltrations of macrophages have been found in the brain as a response to neuroinflammation. One could hypothesise that transmigration of neural stem cells is also possible and therefore this project will explore the protective and regenerative potential of cellular infiltrations of human neural stem cells applied to the CNS cells through the BBB. Human Umbilical Cord Blood- Neural Stem Cells (HUCB-NSC) will be used for the tests. The cells will be obtained from Prof. Buzanska at the Polish Academy of Sciences. The PI has extensive experience in the characterization of developmental processes of HUCB-NSC patterned on biofunctional surfaces and its applications to alternative stem cell based neurotoxicology assays [14,39]. HUCB-NSC line has been obtained from the non-hematopoietic mononuclear fraction of human umbilical cord blood. These cells offer an unlimited source of stem cells of human origin and can be kept and harvested for experiments at different stages of neural commitment. Thus, this system is capturing a developmental period from non-differentiated pluripotent stem cells, through committed neural progenitors to differentiated neuronal, astrocytic and oligodendroglial cells. In this project, the cells will be used at the stage of neural progenitors to study the regenerative potential to re-populate and rescue the activity of the damage neuronal networks. The OMTA recordings will give an indication of the effects on the network, and immunocytochemistry using antibodies directed against Nestin (as marker of stem cell), ï¢-TubulinIII (as marker for neuronal commitment) and GFAP (as marker for atrocytic commitment) will be used to characterize the stem cells.
10nM ~ 500nM
Vesicle shedding inhibition
2ïM ï¾ 100ïM
Blocking of NMDA-type glutamate receptors
0.1ïM ï¾ 10ïM
Hindering of Aï¢ aggregation and reducing inflammation
1nM ï¾ 10nM
Compensate the reduced cellular responsiveness to insulin
Table 1. Neuroprotective agents to be tested by the micro brain analog