- Email: [email protected]
Microphysiological models of neurological disorders for drug development
Journal Pre-proof Microphysiological models of neurological disorders for drug development Giovanni S. Offeddu, Yoojin Shin, Roger D. Kamm PII:
S2468-4511(19)30085-6
DOI:
https://doi.org/10.1016/j.cobme.2019.12.011
Reference:
COBME 209
To appear in:
Current Opinion in Biomedical Engineering
Received Date: 23 October 2019 Revised Date:
14 December 2019
Accepted Date: 20 December 2019
Please cite this article as: G.S. Offeddu, Y. Shin, R.D. Kamm, Microphysiological models of neurological disorders for drug development, Current Opinion in Biomedical Engineering, https://doi.org/10.1016/ j.cobme.2019.12.011. This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. © 2019 Published by Elsevier Inc.
MICROPHYSIOLOGICAL MODELS OF NEUROLOGICAL DISORDERS FOR DRUG DEVELOPMENT Giovanni S. Offeddu1,+, Yoojin Shin1,+, Roger D. Kamm1,2*
1
Department of Biological Engineering, Massachusetts Institute of Technology, Cambridge MA, USA.
2
Department of Mechanical Engineering, Massachusetts Institute of Technology, Cambridge MA,
USA
+These authors contributed equally to this work. *Correspondence to: Roger D. Kamm, [email protected], Department of Biological Engineering, Massachusetts Institute of Technology, NE47-318, Cambridge MA, USA.
Submitted to Current Opinion in Biomedical Engineering
1
1 2 3 4 5 6 7 8 9 10 11 12 13 14
ABSTRACT Clinical trials of drug candidates intended to treat neurological conditions often fail due to the poor correlation between outcomes in humans and other animal models. Microphysiological systems offer a potential alternative to, or synergetic approach with animal testing by using human cells assembled in three-dimensional morphologies that are representative of human tissues and organs. The recent advances in microphysiological models of neurological disorders are reviewed here in terms of models used to quantify drug distribution and drug engagement and function. We find that the complexity of MPS models continues to increase, but only few neurological conditions have been modeled with the purpose of testing pharmacological traits of candidate drugs. Much more effort appears to have been devoted to models that recapitulate drug distribution, particularly across vascular barriers, and models involving barrier modifications as a result of patient-specific neurological conditions have started to emerge. Based on the current trend, microphysiological models of neurological disorders have the capacity to become a useful tool in the pre-clinical testing of drugs for human use.
Keywords: Microfluidic devices, Organ-on-chip, BBB, Drug delivery, ADME.
2
15
1. INTRODUCTION
16 17 18 19 20 21 22 23
Neurological disorders of the central nervous system (CNS) and peripheral nervous system (PNS) are estimated to cause in excess of 10% of all deaths worldwide [1]. This figure is bound to increase due to the prevalence of neurological conditions with age, coupled with an overall aging population. Yet, while the demand for new therapies increases, development has been limited by unique challenges associated with neurological disorders [2]: The poor accessibility of the organs and tissues involved in the conditions, particularly in the case of the CNS, as well as the lack of complex models that are representative of human biology limit our understanding of neurological disorders and hinder the speedy development of effective treatments.
24 25 26 27 28 29 30 31 32 33 34
Many of the known neurological disorders involve disruption of neuronal networks as a result of the complex interplay between different cell types and extra-cellular matrix (ECM) found in the relevant tissue microenvironments [3]. This is the case, for example, for neurodegenerative disorders such as Alzheimer's disease (AD) and amyotrophic lateral sclerosis (ALS), and for tumors of the CNS like glioblastoma (GB). A model intended for testing the transport, engagement, and functional outcome of candidate drugs should, therefore, recapitulate the intrinsic pathophysiological human tissue complexity. Microphysiological systems (MPS) can meet this requirement by enclosing within microfluidic devices human cells of multiple phenotypes arranged in three-dimensional (3D) morphologies [4]. Through fluidic access ports, these cells can be subjected to controlled chemical and physical stimuli, and critical parameters related to drug efficacy can be measured quantitatively with excellent spatiotemporal resolution [5].
35 36 37 38 39 40
The complexity of new MPS models developed, and consequent relevance for human drug testing, have steadily increased. Here, the recent advances in MPS models of neurological disease are reviewed with particular focus on those models offering the capability to measure parameters affecting drug distribution, as well as on those that allow the quantification of drug effects. Several potential future directions are also suggested, which could accelerate the adoption of MPS models in industry and the clinic.
41 42
2. MICROPHYSIOLOGICAL MODELS OF DRUG DISTRIBUTION
43 44 45 46 47 48 49 50 51
Drug efficacy critically depends on the drug’s local concentration in the vicinity of the intended cellular target [6]. Such targets, in the context of neurological disorders, can often only be accessed through the blood circulation, whereby the capillary endothelium constitutes the major barrier to drug accessibility. Within the CNS, the blood-brain-barrier (BBB), a particularly selective endothelium, has received the greatest attention in terms of new MPS models. Other important endothelial barriers to drug distribution, such as the PNS blood-nerve-barrier (BNB) (Figure 1) or the meningeal blood-cerebrospinal fluid-barrier (BCB), instead, have yet to be modelled through complex MPS. Drug clearance, either by blood and lymphatic vasculature, or by distal organs has begun to be addressed through MPS models developed to capture it.
52
2.1 Models of neurovascular drug transport in the CNS
53 54 55 56 57 58
As the main doorway to the CNS, the BBB constitutes a uniquely impervious barrier to the passage of therapeutic molecules. Most likely for this reason, MPS models of the BBB have seen the largest development efforts in the past several years, as recently reviewed by Gastfriend and co-workers [7]. Other works have now begun to address the important issue of benchmarking these models against the natural human BBB [8,9], offering a framework to analyse the most recent developments in MPS models.
3
59 60 61 62 63 64 65 66 67 68 69 70 71 72
In the first instance, the BBB is made up of multiple cell types, each contributing to the barrier’s function. Brain microvascular endothelial cells (BMECs) form capillaries that are partially covered by pericytes (PCs), as well as astrocyte (ACs) end feet, cumulatively forming the so-called neurovascular unit (NVU). Neuroinflammation, a key factor in many neurological disorders, derives from the concerted production of inflammatory cytokines by the different cells of the NVU [10], and so may only be captured by MPS models that include these cells. Recent models of the BBB have consistently included PCs or ACs, or both [11–15]. In one case, the authors of the model, a hydrogeltemplated vessel coated with primary BMECs in contact with primary PCs or ACs, showed that under induced inflammatory conditions the cytokine release profile was critically altered [11]. As the focus of new BBB models shifts towards capturing aspects of neurological disease, a few groups have also included the contribution of immune cells to the NVU integrity, either modelling the infiltration of the immune cells across the BBB during inflammation or characterizing the disruption of the BBB due to such infiltration [14][16]. These pathophysiological models of the BBB present an important opportunity to characterize and predict the different drug distributions found in health and disease.
73
74 75 76 77 78
Figure 1. Overview of key vascular barriers to neuronal drug transport. The blood-brain-barrier and blood-nerve barrier both consist of multiple cell types and extra-cellular matrices, which can be reproduced within microphysiological models to measure key aspects of drug distribution. This figure was partly created with BioRender.com. Partly adapted from [17].
79 80 81 82 83 84 85 86
Novel therapeutic strategies make use of different types of molecules or carriers that cross the endothelium through different routes [6]. The uniqueness of the BBB extends to mechanisms that prevent most types of therapeutic molecules to cross the endothelium; these include reduced nonspecific paracellular passage by tight junctions, reduced receptor-mediated vesicular transport of large proteins and nanocarriers by low pinocytosis, and reduced membrane diffusion of small molecules by expression of efflux pumps like P-glycoprotein 1 (P-gp) [7]. The importance of characterizing not only the magnitude of passage of a drug across the BBB endothelium, i.e. its 4
87 88 89 90 91 92 93 94 95 96 97 98 99 100
permeability, but also the mode of such passage, is starting to be appreciated as a critical benchmark against the microvasculature in vivo. For example, using a 2D BBB model Park and co-workers characterized the transcellular transport of monoclonal antibodies and nanoparticles directed against Angiopep-2 and the transferrin receptor (TfR), two key BBB markers [18]. Remarkably, MPS models allow the interrogation of how these different routes across the endothelium may change under pathophysiological conditions: the same study by Park et al. suggested that hypoxia is necessary to better recapitulate the native BBB function in terms of P-gp and TfR function [18], while another study also utilizing a 2D MPS model of the BBB found that luminal fluid flow and related shear stress on the endothelium alters the expression of Angiopep-2 [19]. These results may have important implications for neurological conditions deriving from impaired vascular function, like stroke or brain aneurysm. 3D models of the BBB, which present a physiologically relevant capillary morphology through patterning [11] or cell self-assembly [13] may also be advantageous as they can better recapitulate the impact of mechanical stimuli on the pathophysiological BBB marker expression.
101 102 103 104 105 106 107 108 109 110 111 112
Several aspects of drug transport in MPS models of the CNS vasculature remain uncharacterized, in particular (i) transmural fluid flow, (ii) perivascular fluid flow and (iii) additional contributions of the extra-cellular matrix (ECM) to drug transport. Transmural fluid flow (i) across the TJs produces increased convective passage of therapeutic molecules depending on size, which we have recapitulated using a pressurized 3D peripheral microvascular model consisting of primary endothelial and stromal cells [20]. Perivascular fluid flow (ii) is present in larger blood vessel in the CNS and possibly in the BBB, where it may “flush off” therapeutic molecules that managed to cross the vascular barrier [21]. Finally, the NVU also includes a thin layer of ECM on the blood side (the glycocalyx) and the brain tissue side (the basement membrane and glia limitans), which have been found to significantly affect drug transport across the BBB [22], yet have been seldom characterized in existing MPS models [13][20] . These additional contributions to transport are necessary to fully compare, and possibly predict, the trans-BBB distribution of candidate drugs in human populations.
113
126
MPS models of the BBB may be tailored to single patients and their specific neurological disorders. In the wake of such personalized medicine, the use of induced pluripotent stem cell (iPSC)-derived BBB cells in MPS models has recently seen an increase [13][15,16][18][23]. BBB endothelial cells can now be derived from iPSCs [24]. At the same time, even tissue non-specific endothelial cells derived from iPSCs are amenable to be used in BBB models, as we have shown that their phenotype can be made more BBB-like (in terms of decreased permeability and gene expression of multiple key BBB markers) by co-culture with primary ACs and PCs [13]. Very recently, Vatine and co-workers showed that BBB maturation of iPSC-endothelial cells can also be achieved by co-culture with iPSC-neural cells, in a full iPSC-derived MPS model. The authors also produced cells from patients affected by two neurological disorders, Huntington disease and MCT8-specific deficiency, and measured differences in drug distribution across disease-specific BBB MPS models perfused with whole blood so as to capture any protein binding. Such personalized models, coupled with novel strategies to multiplex the culture of MPS [25], may pave the way for accurate and quantitative predictive models of CNS drug distribution in neurological disease.
127
2.2 Models of drug clearance
128 129 130 131 132
The use of MPS models to characterize drug distribution has been so far mostly limited to the characterization of vascular transport, with little attention to the destination of the drug. Past the endothelium, candidate drugs may engage their target or be cleared by the lymphatic vasculature found in the vicinity of the PNS and the periphery of the CNS [26]. Lymphatic vessels prevent accumulation of interstitial fluid and solutes, affecting local concentrations of drugs [27]. The
114 115 116 117 118 119 120 121 122 123 124 125
5
133 134 135 136 137 138 139 140
existing MPS models of the lymphatics are few, but they have made interesting findings in that drainage of interstitial molecules is increased in the presence of tissue-engineered lymphatic vessels [28], and that the co-culture of lymphatic endothelial cells with stromal cells significantly modifies such drainage [29]. Clearance also takes place within the vascular endothelium, where part of the molecules crossing the cells are cleared within lysosomes and prevented from crossing. We have characterized this clearance in a 3D microvascular model, showing that therapeutically relevant plasma proteins like albumin and immunoglobulin-G (IgG) are rescued from degradation by the Fc neonatal receptor (FcRN), although this results in an overall lower endothelial permeability [6].
141 142 143 144 145 146
Ultimately, the clearance of any therapeutic molecule is complex and takes place within multiple organs in the body, each critically affecting the availability of the drug and its capability to engage its target. MPS models of inter-connected organ systems have recently been developed [30][31]. These models can recapitulate aspects of the complex crosstalk between multiple organ tissues and the resulting pharmacokinetic characteristics that determine the absorption, distribution, metabolism and excretion of candidate drugs (ADME).
147 148
3. MICROPHYSIOLOGICAL MODELS OF DRUG ENGAGEMENT AND FUNCTION
149
159
Numerous cases of drugs that proved effective in tests on animal models, yet failed in human clinical trials, have led to the recognition that animal models may not be fully representative of human disease pathology. Largely for this reason, there has recently been a high demand for in vitro systems that can mimic the complexity of human tissues and allow measurement of pharmacological traits. The remarkable development efforts to create MPS related to the brain and nerves have been recently reviewed by Haring and co-workers [32]. Several MPS studies associated with neurological disease have now begun to directly involve quantitative drug screening and evaluation, as well as the study of disease mechanisms. Here, we introduce several recently developed MPS models of the neurological disorders so far addressed that directly involve drug assessment, such as models of AD, ALS, epilepsy, and Parkinson’s disease (PD). Other neurological conditions have yet to be modelled through MPS capable of addressing drug effects on neuronal dysfunction.
160
3.1 Alzheimer’s disease
161
Although several drugs have shown promise in animal models, many clinical trials of therapies for AD have failed, giving rise to tremendous interest in developing an in vitro AD model using human cells or patient-derived iPSCs aimed at drug screening. Recently, in vitro AD models have been developed focusing not only on neuronal loss itself, but also on other factors influencing the disease including the roles of the BBB and microglia. Our group developed a 3D in vitro model of AD by culturing human neural progenitor cells generating extracellular aggregation of amyloid beta (Aβ) and accumulation of hyperphosphorylated tau, with an intact BMEC barrier in a 3D microfluidic system [33]. By using genetically-modified cells that overexpress the different isoforms of amyloid beta [34,35], it was observed that the AD pathological microenvironment exerts a direct effect on BMECs in the BBB. In this 3D AD model, several vascular defects observed in AD patients were successfully mimicked, including an increase in BBB permeability. The studies further demonstrated that neurotoxins passing the leaky BBB can exacerbate neuronal loss in AD. Tests of pharmacological compounds that enhance BBB integrity were also performed, and drug efficacy characterized. Restoration of the disrupted BBB and decreased neuronal loss were observed, supporting the theory that restoration of BBB integrity could be a useful therapeutic target for AD by preventing
150 151 152 153 154 155 156 157 158
162 163 164 165 166 167 168 169 170 171 172 173 174 175
6
182
neurotoxins from entering the brain [33]. Park et al. developed a triculture system of neurons, astrocytes, and microglia to study the role of the latter in neuroinflammation in AD pathogenesis and progression. In this system, microglia recruitment induced by chemokines and cytokines secreted from AD culture and consequent neuron and astrocyte loss were observed. The authors tested IFN-γ or TLR4 blocking antibodies and the TLR4 antagonist LPS-RS, and observed reduced TNFα and NO concentrations and decreased LDH levels in the system, suggesting that neuronal loss depends on microglia activation via IFN-γ and TLR4-dependent mechanisms [36] .
183
3.2 Amyotrophic lateral sclerosis (ALS)
184
204
The recent advances in microfluidic fabrication techniques coupled with biological assembly have widened the horizon of ALS research by better mimicking the nerve-muscle connectivity. Machado et al. developed an in vitro model of a neuromuscular circuit [37]. This system consists of three compartments; a central one for myofiber culture and two outer compartments for motor neuron (MN) and astrocyte (AC) aggregates. Each compartment is connected by microchannel arrays. The system is designed to allow motor axons to extend from the motor neuron soma, project through an array of microchannels toward the central myofibers compartment, and form neuromuscular junctions. The authors demonstrated the capability of the system as an ALS disease model by coculturing MNs with ACs expressing ALS-linked SOD1G93A mutants and observed denervation of myofibers and diminished muscle contractions, both key features of ALS. Also, the authors showed that the candidate ALS drug RIPK1 inhibitor Necrostatin aided in the recovery of both axon growth and myofiber contractions. Our group also developed a 3D ALS motor unit model by co-culturing iPSC-derived MNs from a patient with sporadic ALS with iPSC-derived 3D muscle fiber bundles in a microfluidic system [38]. The system consists of three distinct culture compartments; for MN spheroid culture, neurite growth, and muscle tissue culture. Following muscle fiber bundle formation, MN spheroids derived from iPSCs either from a healthy subject or an ALS patient were introduced and allowed to elongate their neurites toward the muscle fiber bundles and form a functional motor unit with neuromuscular junctions. Significant differences were observed between the two models, in terms of neurite growth rate, force of muscle contraction, and skeletal muscle cell death. The authors tested two drug candidates of ALS, bosutinib and rapamycin, in the system and both proved effective in improving neuronal survival and muscle contraction.
205
3.3 Epilepsy
206
213
Epilepsy is a neurological disorder associated with abnormal neural electrical activity in the brain. Liu et al. developed a microfluidic-based microelectrode array device enabling the support of six organotypic hippocampal slice cultures on a single unit and the detection of separate electrographic recordings simultaneously [39]. Testing of a small molecule inhibitor of receptor tyrosine kinases (RTKs), which change in their expression or phosphorylation in the epileptic brain, was conducted in the system and antiepileptic properties were assessed via biomarker sampling of lactate for seizurelike activity and lactate dehydrogenase (LDH) for seizure-dependent cell death, and via electrical recordings for the electrophysiological verification of the biomarker results.
214
3.4 Parkinson’s disease
215
Systems that have the capability to be used as tools for testing drug efficacy in PD have been recently developed. 3D culture of PD-specific dopaminergic neurons incorporated in hydrogels
176 177 178 179 180 181
185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203
207 208 209 210 211 212
216
7
217 218 219 220 221 222 223 224
within a microfluidic system was found to recapitulate the key cellular defects in PD. Bolognin et al. cultured neuroepithelial stem cell lines derived from PD patients carrying leucine-rich repeat kinase 2 (LRRK2)-G2019S mutations in 3D microfluidic systems implemented with automated image analysis algorithms and observed a time-dependent loss of dopaminergic neurons caused by the mutation. Also, the authors compared 3D culture in microfluidic systems with 2D culture and demonstrated that 3D culture presents more robust PD endophenotypes. They inhibited LRRK2 and found it to effectively alleviate neurodegeneration and the appearance of LRRK2-G2019S-dependent dopaminergic phenotypes [40]. Disease
Alzheimer’s disease (AD)
Amyotrophic lateral sclerosis (ALS)
Cell types Viral-infected human neural progenitor cells (hNPC) with familial AD mutations, human brain endothelial cell line (hCMEC/D3) Viral-infected human neural progenitor cells (hNPC) with familial AD mutations, human iPSCderived neural progenitor cells, immortalized human microglia (SV40 cell line) Mouse embryonic stem cell (ESC)derived motor neurons and astrocytes, primary chick myoblasts
Drug target function Improve BBB integrity and alleviate neuronal loss
Analytical methods Permeability measurement with fluorescent dextrans for BBB integrity, viability test with redfluorescent ethidium homodimer-1 for neuronal loss
References [33]
Reduce neuroinflammation induced by microglia with IFN-γ or TLR4 blocking antibodies and the TLR4 antagonist LPS-RS
Quantification of inflammatory cytokine concentrations by ELISA
[36]
Improve motor neuron survival with RIPK1 inhibitor Necrostatin
Motor neurons differentiated from hESCderived neural stem cells (NSCs) and ALS
Improve motor neuron survival by inhibiting the Src/c-Abl signalling pathway (bosutinib) or inhibiting the mechanistic target of the mTOR
Immunostaining with [37] β3-tubulin antibody to assess axon outgrowth, analysis of myofiber contraction by optogenetic stimulation and quantification by particle image velocimetry (PIV) Measurement of [36] muscle force of contraction through optical stimulation and recording, assessment of
8
Epilepsy
225
patient iPSCderived NSCs, human iPSCderived skeletal myoblasts, iPSC-derived ECs Hippocampi dissected from postnatal day 7–8 SpragueDawley rat pups
pathway (rapamycin)
apoptosis and muscle atrophy by expression of caspase3/7.
Antiepileptic drug of small molecule inhibitors of receptor tyrosine kinases (RTKs)
Biomarker sampling [39] with Lactate and Lactate Dehydrogenase (LDH) Assays, measurement of neural activity with electrical recordings. Parkin’s Neuroepithelial Alleviate Live imaging with [40] disease (PD) stem cell lines neurodegeneration and tretamethylrhodamine, derived from the methyl ester (TKRK) healthy and PD LRRK2-G2019S-dependent and MitoTracker Green patients dopaminergic phenotypes to assess cell death carrying LRRK2- with LRRK2 inhibitor 2 and mitochondrial G2019S (Inh2) activity, mutation morphometric assay to analyze differences in skeleton, perimeter, body pixels, and complexity of the neural networks Table 1. MPS models employed to assess drug effects in neurological disorders.
226 227 228 229 230 231 232 233 234 235 236 237 238 239 240
Besides the MPS models addressed here, there are other MPS models of neurological diseases that have the potential to be used to evaluate drugs, despite the fact that their developers did not yet directly use the models for drug testing. For example, a co-culture platform integrated with microvalves that allow for precise control of cell-cell interactions, particularly via paracrine signaling, was developed to study synucleinopathies such as PD [41]. Also, a Huntington’s disease (HD)-on-a-chip model has been developed by culturing HdhCAG140/+ cortical and striatal neurons from heterozygous mice in a microfluidic system [42]. The microfluidic device consists of three compartments of cortical and striatal chambers that are connected by axonal and dendritic microchannels with an intermediate synaptic chamber. This MPS HD model showed that HD mature networks have pre- and postsynaptic defects and also defective and hypersynchronized corticostriatal networks. Finally, there are a few MPS models of GB addressing drug discovery for the disease [43,44]. However, these models focus only on drug effects on cancer cells, while there is a critical lack of models that address neuronal dysfunctions caused by the cancer cells, through integration of neural cell cultures (e.g. neurons or astrocytes) with GB cultures.
241 242
4. OUTLOOK
243 244 245
MPS models of neurological disease have the potential to play a critical role in drug development through pre-clinical testing of candidate drugs. The past several years have seen increased efforts in developing MPS models of neurological disorder recapitulating selected functions of human tissues 9
246 247 248
and organs; yet, there are still challenges to be considered and overcome, in particular with regard to the balance between complexity and high throughput of MPS models, and how this balance may affect their adoption in industry and the clinic.
249 250 251 252 253 254 255 256 257 258 259 260 261
On one hand, increased model complexity may offer advantages in recapitulating pathophysiological cross talk between different cell types, as well as a more detailed assessment of concurrent drug distribution, target engagement and function. For example, although drug vascular transport and clearance are key aspects to be considered in the evaluation of drug efficacy, currently most MPS models of neurological disorder do not incorporate the functional contribution of vasculature. In particular, models of vascular barriers such as the BNB and BCB are critically unexplored and could be integrated within MPS systems for drug testing in PNS disorders (e.g. ALS) or other peripheral disorders that can affect the CNS [26]. Similarly, since neurological diseases are long-term and progressive, MPSs would benefit from long-term culture to evaluate the exposure to drugs, which may be facilitated by the integration with a perfusion system that continuously provides nutrients to cells. On the other hand, high throughput screening can be a limiting factor in the evaluation of multiple candidate drugs. Current MPS models are inherently lower throughput, high complexity models.
262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278
Standardized, high throughput screening MPSs that incorporate complex, disease-specific 3D cellular microenvironments will likely have increased appeal when compared to animal models. For the purpose of screening drugs with better quality assurance/quality control (QA/QC), well-established and characterized human primary cells or cell lines that can recapitulate physiological or pathological general disease traits may be used. However, the use of multiple types of iPSC-derived cells from the same patient donor can provide more realistic recapitulation of the disease-associated phenotypes for personalized drug efficacy, sensitivity and toxicity screening, although possibly at the expense of high throughput. In addition, MPS model integrated with standardized functional readouts such as permeability test for BBB function (for CNS diseases related to BBB impairment), voltage signaling detection for testing neuronal function, and skeletal muscle contraction for testing muscular function (for ALS models), along with molecular and cellular readouts will provide a deeper insight into the pathophysiological events underlying neurological diseases. Ultimately, MPS models may not entirely replace animal models, but they could offer a strategy to further rank and select candidate molecules in vitro, thus reducing the number of animal models required, limiting costs while also addressing growing ethical concerns. Critically, these models provide drug testing platforms that are more relevant for human patients and may improve drug outcomes in clinical trials.
279 280 281 282 283 284 285 286
ACKNOWLEDGEMENTS
10
287 288 289
G.S.O. is supported by an American-Italian Cancer Foundation Post-Doctoral Research Fellowship. Y.S. is supported by the Cure Alzheimer’s Fund. R.D.K. has funding support from the National Science Foundation (CBET-0939511).
290 291
DISCLOSURES
292 293
R.D.K. is a co-founder of AIM Biotech that markets microfluidic systems for 3D culture. Funding support is also provided by Amgen, Biogen and Gore.
294 295
REFERENCES
296 297 298
1.
Little D, Ketteler R, Gissen P, Devine MJ: Using stem cell-derived neurons in drug screening for neurological diseases. Neurobiol Aging 2019, 78:130–141.
299 300
2.
Danon JJ, Reekie TA, Kassiou M: Challenges and opportunities in central nervous system drug discovery. Trends Chem 2019, 1:612–624.
301 302
3.
Osaki T, Shin Y, Sivathanu V, Campisi M, Kamm RD: In Vitro Microfluidic Models for Neurodegenerative Disorders. Adv Healthc Mater 2018, 7:1700489.
303 304
4.
Truskey GA: Human microphysiological systems and organoids as in vitro models for toxicological studies. Front public Heal 2018, 6:185.
305 306
5.
Hajal C, Campisi M, Mattu C, Chiono V, Kamm RD: In vitro models of molecular and nano-particle transport across the blood-brain barrier. Biomicrofluidics 2018, 12:42213.
307 308 309 310 311
6.
Offeddu GS, Haase K, Gillrie MR, Li R, Morozova O, Hickman D, Knutson CG, Kamm RD: An on-chip model of protein paracellular and transcellular permeability in the microcirculation. Biomaterials 2019, 212:115–125. *The authors separately assessed different physiological modes of endothelial molecular transport and assessed endothelial clearing of pharmacologically-relevant plasma proteins in a 3D model of the microvasculature.
312 313
7.
Gastfriend BD, Palecek SP, Shusta E V: Modeling the blood-brain barrier: beyond the endothelial cells. Curr Opin Biomed Eng 2018, 5:6–12.
314 315 316 317
8.
Helms HC, Abbott NJ, Burek M, Cecchelli R, Couraud P-O, Deli MA, Förster C, Galla HJ, Romero IA, Shusta E V, et al.: In vitro models of the blood--brain barrier: an overview of commonly used brain endothelial cell culture models and guidelines for their use. J Cereb Blood Flow Metab 2016, 36:862–890.
318 319
9.
DeStefano JG, Jamieson JJ, Linville RM, Searson PC: Benchmarking in vitro tissue-engineered blood-brain barrier models. Fluids Barriers CNS 2018, 15:32.
320 321
10.
Oddo A, Peng B, Tong Z, Wei Y, Tong WY, Thissen H, Voelcker NH: Advances in microfluidic bloodbrain barrier (BBB) models. Trends Biotechnol 2019,
322 323 324 325 326
11.
Herland A, van der Meer AD, FitzGerald EA, Park T-E, Sleeboom JJF, Ingber DE: Distinct contributions of astrocytes and pericytes to neuroinflammation identified in a 3D human bloodbrain barrier on a chip. PLoS One 2016, 11:e0150360. **The authors developed a 3D BBB model and assessed cytokine production by different cells of the NVU under induced inflammatory conditions.
11
327 328
12.
Wang YI, Abaci HE, Shuler ML: Microfluidic blood-brain barrier model provides in vivo-like barrier properties for drug permeability screening. Biotechnol Bioeng 2017, 114:184–194.
329 330 331
13.
Campisi M, Shin Y, Osaki T, Hajal C, Chiono V, Kamm RD: 3D self-organized microvascular model of the human blood-brain barrier with endothelial cells, pericytes and astrocytes. Biomaterials 2018,
332 333 334
14.
Lauranzano E, Campo E, Rasile M, Molteni R, Pizzocri M, Passoni L, Bello L, Pozzi D, Pardi R, Matteoli M, et al.: A Microfluidic Human Model of Blood-Brain Barrier Employing Primary Human Astrocytes. Adv Biosyst 2019, 3:1800335.
335 336 337 338 339 340
15.
Vatine GD, Barrile R, Workman MJ, Sances S, Barriga BK, Rahnama M, Barthakur S, Kasendra M, Lucchesi C, Kerns J, et al.: Human iPSC-Derived Blood-Brain Barrier Chips Enable Disease Modeling and Personalized Medicine Applications. Cell Stem Cell 2019, 24:995–1005. **The authors developed full iPSC-derived MPS models of the BBB using cells from healthy patients or patients affected by two neurological disorders, and perfused the models with whole blood showing differences in endothelial barrier properties.
341 342 343
16.
Linville RM, DeStefano JG, Sklar MB, Xu Z, Farrell AM, Bogorad MI, Chu C, Walczak P, Cheng L, Mahairaki V, et al.: Human iPSC-derived blood-brain barrier microvessels: Validation of barrier function and endothelial cell behavior. Biomaterials 2019, 190:24–37.
344 345
17.
Kanda T: Biology of the blood-nerve barrier and its alteration in immune mediated neuropathies. J Neurol Neurosurg Psychiatry 2013, 84:208–212.
346 347 348
18.
Park T-E, Mustafaoglu N, Herland A, Hasselkus R, Mannix R, FitzGerald EA, Prantil-Baun R, Watters A, Henry O, Benz M, et al.: Hypoxia-enhanced Blood-Brain Barrier Chip recapitulates human barrier function and shuttling of drugs and antibodies. Nat Commun 2019, 10:2621.
349 350 351
19.
Papademetriou I, Vedula E, Charest J, Porter T: Effect of flow on targeting and penetration of angiopep-decorated nanoparticles in a microfluidic model blood-brain barrier. PLoS One 2018, 13:e0205158.
352 353 354 355 356
20.
Offeddu GS, Possenti L, Loessberg-Zahl JT, Zunino P, Roberts J, Han X, Hickman D, Knutson CG, Kamm RD: Application of Transmural Flow Across In Vitro Microvasculature Enables Direct Sampling of Interstitial Therapeutic Molecule Distribution. Small 2019, 1902393. *The authors introduced the contribution of transmural fluid flow to drug transport in a 3D model of the microvasculature.
357 358 359
21.
Mestre H, Tithof J, Du T, Song W, Peng W, Sweeney AM, Olveda G, Thomas JH, Nedergaard M, Kelley DH: Flow of cerebrospinal fluid is driven by arterial pulsations and is reduced in hypertension. Nat Commun 2018, 9:4878.
360 361 362
22.
Kutuzov N, Flyvbjerg H, Lauritzen M: Contributions of the glycocalyx, endothelium, and extravascular compartment to the blood-brain barrier. Proc Natl Acad Sci 2018, 115:E9429-E9438.
363 364
23.
Motallebnejad P, Thomas A, Swisher SL, Azarin SM: An isogenic hiPSC-derived BBB-on-a-chip. Biomicrofluidics 2019, 13:64119.
365 366
24.
Shusta E V, Azarin S, Palecek S, Lippmann E: Human blood-brain barrier endothelial cells derived from pluripotent stem cells and blood-brain barrier model thereof. 2018,
367 368
25.
Lee S-R, Hyung S, Bang S, Lee Y, Ko J, Lee S, Kim HJ, Jeon NL: Modeling neural circuit, blood-brain barrier, and myelination on a microfluidic 96 well plate. Biofabrication 2019, 11:35013.
369
26.
Louveau A, Herz J, Alme MN, Salvador AF, Dong MQ, Viar KE, Herod SG, Knopp J, Setliff JC, Lupi AL, 12
et al.: CNS lymphatic drainage and neuroinflammation are regulated by meningeal lymphatic vasculature. Nat Neurosci 2018, 21:1380.
370 371 372 373
27.
Chang C-W, Seibel AJ, Song JW: Application of microscale culture technologies for studying lymphatic vessel biology. Microcirculation 2019,
374 375 376
28.
Thompson RL, Margolis EA, Ryan TJ, Coisman BJ, Price GM, Wong KHK, Tien J: Design principles for lymphatic drainage of fluid and solutes from collagen scaffolds. J Biomed Mater Res Part A 2018, 106:106–114.
377 378 379 380
29.
Gong MM, Lugo-Cintron KM, White BR, Kerr SC, Harari PM, Beebe DJ: Human organotypic lymphatic vessel model elucidates microenvironment-dependent signaling and barrier function. Biomaterials 2019, 214:119225. **The authors developed a 3D lymphatic vessel model that can be separtely perfused to measure lymphatic permeability and drainage.
381 382 383
30.
Edington CD, Chen WLK, Geishecker E, Kassis T, Soenksen LR, Bhushan BM, Freake D, Kirschner J, Maass C, Tsamandouras N, et al.: Interconnected microphysiological systems for quantitative biology and pharmacology studies. Sci Rep 2018, 8:4530.
384 385 386
31.
Prantil-Baun R, Novak R, Das D, Somayaji MR, Przekwas A, Ingber DE: Physiologically based pharmacokinetic and pharmacodynamic analysis enabled by microfluidically linked organs-onchips. Annu Rev Pharmacol Toxicol 2018, 58:37–64.
387 388 389
32.
Haring AP, Sontheimer H, Johnson BN: Microphysiological human brain and neural systems-on-achip: potential alternatives to small animal models and emerging platforms for drug discovery and personalized medicine. Stem cell Rev reports 2017, 13:381–406.
390 391 392 393
33.
Shin Y, Choi SH, Kim E, Bylykbashi E, Kim JA, Chung S, Kim DY, Kamm RD, Tanzi RE: Blood-Brain Barrier Dysfunction in a 3D In Vitro Model of Alzheimer’s Disease. Adv Sci 2019, 1900962. **The authors developed a 3D blood-brain barrier model of AD and observed AD pathological microenvironment directly effects on brain endothelial cells of BBB.
394 395 396
34.
Kim YH, Choi SH, D’avanzo C, Hebisch M, Sliwinski C, Bylykbashi E, Washicosky KJ, Klee JB, Brüstle O, Tanzi RE, et al.: A 3D human neural cell culture system for modeling Alzheimer’s disease. Nat Protoc 2015, 10:985.
397 398 399
35.
Choi SH, Kim YH, Hebisch M, Sliwinski C, Lee S, D’Avanzo C, Chen H, Hooli B, Asselin C, Muffat J, et al.: A three-dimensional human neural cell culture model of Alzheimer’s disease. Nature 2014, 515:274–278.
400 401 402 403 404
36.
Park J, Wetzel I, Marriott I, Dréau D, D’Avanzo C, Kim DY, Tanzi RE, Cho H: A 3D human triculture system modeling neurodegeneration and neuroinflammation in Alzheimer’s disease. Nat Neurosci 2018, 21:941. **The authors developed a 3D AD model to study neural-glial interactions in AD pathology and observed physical interruptions between microglia and nuerons in AD pathological microenvironment.
405 406 407
37.
Machado CB, Pluchon P, Harley P, Rigby M, Sabater VG, Stevenson DC, Hynes S, Lowe A, Burrone J, Viasnoff V, et al.: In Vitro Modeling of Nerve-Muscle Connectivity in a Compartmentalized Tissue Culture Device. Adv Biosyst 2019,
408 409 410 411 412
38.
Osaki T, Uzel SGM, Kamm RD: Microphysiological 3D model of amyotrophic lateral sclerosis (ALS) from human iPS-derived muscle cells and optogenetic motor neurons. Sci Adv 2018, 4:eaat5847. **The authors developed a 3D ALS model by co-culturing skeletal muscle bundles with iPSCderived motor neuron spheroids and observed the ALS motor unit generated fewer muscle contractions and motor neuron degradation compared to a non-ALS motor unit.
413
39.
Liu J, Sternberg AR, Ghiasvand S, Berdichevsky Y: Epilepsy-on-a-chip System for Antiepileptic 13
Drug Discovery. IEEE Trans Biomed Eng 2018, 66:1231–1241.
414 415 416 417 418 419 420
40.
Bolognin S, Fossépré M, Qing X, Jarazo J, Ščančar J, Moreno EL, Nickels SL, Wasner K, Ouzren N, Walter J, et al.: 3D Cultures of Parkinson’s Disease-Specific Dopaminergic Neurons for High Content Phenotyping and Drug Testing. Adv Sci 2019, 6:1800927. **The authors developed a 3D model of PD with patient neurons derived from iPSC carrying the LRRK2-G2019S mutation. They compared 2D culture and 3D culture of dopaminergic neurons and found that 3D culture presents rubust endophenotypes compared to 2D culture.
421 422 423
41.
Fernandes JTS, Chutna O, Chu V, Conde JP, Outeiro TF: A novel microfluidic cell co-culture platform for the study of the molecular mechanisms of Parkinson’s disease and other synucleinopathies. Front Neurosci 2016, 10:511.
424 425 426
42.
Virlogeux A, Moutaux E, Christaller W, Genoux A, Bruyere J, Fino E, Charlot B, Cazorla M, Saudou F: Reconstituting corticostriatal network on-a-chip reveals the contribution of the presynaptic compartment to Huntington’s disease. Cell Rep 2018, 22:110–122.
427 428 429
43.
Yi H-G, Jeong YH, Kim Y, Choi Y-J, Moon HE, Park SH, Kang KS, Bae M, Jang J, Youn H, et al.: A bioprinted human-glioblastoma-on-a-chip for the identification of patient-specific responses to chemoradiotherapy. Nat Biomed Eng 2019, 1.
430 431
44.
Fan Y, Nguyen DT, Akay Y, Xu F, Akay M: Engineering a brain cancer chip for high-throughput drug screening. Sci Rep 2016, 6:25062.
432
14
We wish to draw the attention of the Editor to the following facts which may be considered as potential conflicts of interest and to significant financial contributions to this work: R.D.K. is a co founder of AIM Biotech that markets microfluidic systems for 3D culture. Funding support is also provided by Amgen, Biogen and Gore. We confirm that the manuscript has been read and approved by all named authors and that there are no other persons who satisfied the criteria for authorship but are not listed. We further confirm that the order of authors listed in the manuscript has been approved by all of us. We confirm that we have given due consideration to the protection of intellectual property associated with this work and that there are no impediments to publication, including the timing of publication, with respect to intellectual property. In so doing we confirm that we have followed the regulations of our institutions concerning intellectual property. We understand that the Corresponding Author is the sole contact for the Editorial process (including Editorial Manager and direct communications with the office). He/she is responsible for communicating with the other authors about progress, submissions of revisions and final approval of proofs. We confirm that we have provided a current, correct email address which is accessible by the Corresponding Author and which has been configured to accept email from [email protected]. Signed by all authors as follows: Giovanni Offeddu
10/22/2019
Yoojin Shin
10/22/2019
Roger Kamm
10/22/2019
S2468-4511(19)30085-6
DOI:
https://doi.org/10.1016/j.cobme.2019.12.011
Reference:
COBME 209
To appear in:
Current Opinion in Biomedical Engineering
Received Date: 23 October 2019 Revised Date:
14 December 2019
Accepted Date: 20 December 2019
Please cite this article as: G.S. Offeddu, Y. Shin, R.D. Kamm, Microphysiological models of neurological disorders for drug development, Current Opinion in Biomedical Engineering, https://doi.org/10.1016/ j.cobme.2019.12.011. This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. © 2019 Published by Elsevier Inc.
MICROPHYSIOLOGICAL MODELS OF NEUROLOGICAL DISORDERS FOR DRUG DEVELOPMENT Giovanni S. Offeddu1,+, Yoojin Shin1,+, Roger D. Kamm1,2*
1
Department of Biological Engineering, Massachusetts Institute of Technology, Cambridge MA, USA.
2
Department of Mechanical Engineering, Massachusetts Institute of Technology, Cambridge MA,
USA
+These authors contributed equally to this work. *Correspondence to: Roger D. Kamm, [email protected], Department of Biological Engineering, Massachusetts Institute of Technology, NE47-318, Cambridge MA, USA.
Submitted to Current Opinion in Biomedical Engineering
1
1 2 3 4 5 6 7 8 9 10 11 12 13 14
ABSTRACT Clinical trials of drug candidates intended to treat neurological conditions often fail due to the poor correlation between outcomes in humans and other animal models. Microphysiological systems offer a potential alternative to, or synergetic approach with animal testing by using human cells assembled in three-dimensional morphologies that are representative of human tissues and organs. The recent advances in microphysiological models of neurological disorders are reviewed here in terms of models used to quantify drug distribution and drug engagement and function. We find that the complexity of MPS models continues to increase, but only few neurological conditions have been modeled with the purpose of testing pharmacological traits of candidate drugs. Much more effort appears to have been devoted to models that recapitulate drug distribution, particularly across vascular barriers, and models involving barrier modifications as a result of patient-specific neurological conditions have started to emerge. Based on the current trend, microphysiological models of neurological disorders have the capacity to become a useful tool in the pre-clinical testing of drugs for human use.
Keywords: Microfluidic devices, Organ-on-chip, BBB, Drug delivery, ADME.
2
15
1. INTRODUCTION
16 17 18 19 20 21 22 23
Neurological disorders of the central nervous system (CNS) and peripheral nervous system (PNS) are estimated to cause in excess of 10% of all deaths worldwide [1]. This figure is bound to increase due to the prevalence of neurological conditions with age, coupled with an overall aging population. Yet, while the demand for new therapies increases, development has been limited by unique challenges associated with neurological disorders [2]: The poor accessibility of the organs and tissues involved in the conditions, particularly in the case of the CNS, as well as the lack of complex models that are representative of human biology limit our understanding of neurological disorders and hinder the speedy development of effective treatments.
24 25 26 27 28 29 30 31 32 33 34
Many of the known neurological disorders involve disruption of neuronal networks as a result of the complex interplay between different cell types and extra-cellular matrix (ECM) found in the relevant tissue microenvironments [3]. This is the case, for example, for neurodegenerative disorders such as Alzheimer's disease (AD) and amyotrophic lateral sclerosis (ALS), and for tumors of the CNS like glioblastoma (GB). A model intended for testing the transport, engagement, and functional outcome of candidate drugs should, therefore, recapitulate the intrinsic pathophysiological human tissue complexity. Microphysiological systems (MPS) can meet this requirement by enclosing within microfluidic devices human cells of multiple phenotypes arranged in three-dimensional (3D) morphologies [4]. Through fluidic access ports, these cells can be subjected to controlled chemical and physical stimuli, and critical parameters related to drug efficacy can be measured quantitatively with excellent spatiotemporal resolution [5].
35 36 37 38 39 40
The complexity of new MPS models developed, and consequent relevance for human drug testing, have steadily increased. Here, the recent advances in MPS models of neurological disease are reviewed with particular focus on those models offering the capability to measure parameters affecting drug distribution, as well as on those that allow the quantification of drug effects. Several potential future directions are also suggested, which could accelerate the adoption of MPS models in industry and the clinic.
41 42
2. MICROPHYSIOLOGICAL MODELS OF DRUG DISTRIBUTION
43 44 45 46 47 48 49 50 51
Drug efficacy critically depends on the drug’s local concentration in the vicinity of the intended cellular target [6]. Such targets, in the context of neurological disorders, can often only be accessed through the blood circulation, whereby the capillary endothelium constitutes the major barrier to drug accessibility. Within the CNS, the blood-brain-barrier (BBB), a particularly selective endothelium, has received the greatest attention in terms of new MPS models. Other important endothelial barriers to drug distribution, such as the PNS blood-nerve-barrier (BNB) (Figure 1) or the meningeal blood-cerebrospinal fluid-barrier (BCB), instead, have yet to be modelled through complex MPS. Drug clearance, either by blood and lymphatic vasculature, or by distal organs has begun to be addressed through MPS models developed to capture it.
52
2.1 Models of neurovascular drug transport in the CNS
53 54 55 56 57 58
As the main doorway to the CNS, the BBB constitutes a uniquely impervious barrier to the passage of therapeutic molecules. Most likely for this reason, MPS models of the BBB have seen the largest development efforts in the past several years, as recently reviewed by Gastfriend and co-workers [7]. Other works have now begun to address the important issue of benchmarking these models against the natural human BBB [8,9], offering a framework to analyse the most recent developments in MPS models.
3
59 60 61 62 63 64 65 66 67 68 69 70 71 72
In the first instance, the BBB is made up of multiple cell types, each contributing to the barrier’s function. Brain microvascular endothelial cells (BMECs) form capillaries that are partially covered by pericytes (PCs), as well as astrocyte (ACs) end feet, cumulatively forming the so-called neurovascular unit (NVU). Neuroinflammation, a key factor in many neurological disorders, derives from the concerted production of inflammatory cytokines by the different cells of the NVU [10], and so may only be captured by MPS models that include these cells. Recent models of the BBB have consistently included PCs or ACs, or both [11–15]. In one case, the authors of the model, a hydrogeltemplated vessel coated with primary BMECs in contact with primary PCs or ACs, showed that under induced inflammatory conditions the cytokine release profile was critically altered [11]. As the focus of new BBB models shifts towards capturing aspects of neurological disease, a few groups have also included the contribution of immune cells to the NVU integrity, either modelling the infiltration of the immune cells across the BBB during inflammation or characterizing the disruption of the BBB due to such infiltration [14][16]. These pathophysiological models of the BBB present an important opportunity to characterize and predict the different drug distributions found in health and disease.
73
74 75 76 77 78
Figure 1. Overview of key vascular barriers to neuronal drug transport. The blood-brain-barrier and blood-nerve barrier both consist of multiple cell types and extra-cellular matrices, which can be reproduced within microphysiological models to measure key aspects of drug distribution. This figure was partly created with BioRender.com. Partly adapted from [17].
79 80 81 82 83 84 85 86
Novel therapeutic strategies make use of different types of molecules or carriers that cross the endothelium through different routes [6]. The uniqueness of the BBB extends to mechanisms that prevent most types of therapeutic molecules to cross the endothelium; these include reduced nonspecific paracellular passage by tight junctions, reduced receptor-mediated vesicular transport of large proteins and nanocarriers by low pinocytosis, and reduced membrane diffusion of small molecules by expression of efflux pumps like P-glycoprotein 1 (P-gp) [7]. The importance of characterizing not only the magnitude of passage of a drug across the BBB endothelium, i.e. its 4
87 88 89 90 91 92 93 94 95 96 97 98 99 100
permeability, but also the mode of such passage, is starting to be appreciated as a critical benchmark against the microvasculature in vivo. For example, using a 2D BBB model Park and co-workers characterized the transcellular transport of monoclonal antibodies and nanoparticles directed against Angiopep-2 and the transferrin receptor (TfR), two key BBB markers [18]. Remarkably, MPS models allow the interrogation of how these different routes across the endothelium may change under pathophysiological conditions: the same study by Park et al. suggested that hypoxia is necessary to better recapitulate the native BBB function in terms of P-gp and TfR function [18], while another study also utilizing a 2D MPS model of the BBB found that luminal fluid flow and related shear stress on the endothelium alters the expression of Angiopep-2 [19]. These results may have important implications for neurological conditions deriving from impaired vascular function, like stroke or brain aneurysm. 3D models of the BBB, which present a physiologically relevant capillary morphology through patterning [11] or cell self-assembly [13] may also be advantageous as they can better recapitulate the impact of mechanical stimuli on the pathophysiological BBB marker expression.
101 102 103 104 105 106 107 108 109 110 111 112
Several aspects of drug transport in MPS models of the CNS vasculature remain uncharacterized, in particular (i) transmural fluid flow, (ii) perivascular fluid flow and (iii) additional contributions of the extra-cellular matrix (ECM) to drug transport. Transmural fluid flow (i) across the TJs produces increased convective passage of therapeutic molecules depending on size, which we have recapitulated using a pressurized 3D peripheral microvascular model consisting of primary endothelial and stromal cells [20]. Perivascular fluid flow (ii) is present in larger blood vessel in the CNS and possibly in the BBB, where it may “flush off” therapeutic molecules that managed to cross the vascular barrier [21]. Finally, the NVU also includes a thin layer of ECM on the blood side (the glycocalyx) and the brain tissue side (the basement membrane and glia limitans), which have been found to significantly affect drug transport across the BBB [22], yet have been seldom characterized in existing MPS models [13][20] . These additional contributions to transport are necessary to fully compare, and possibly predict, the trans-BBB distribution of candidate drugs in human populations.
113
126
MPS models of the BBB may be tailored to single patients and their specific neurological disorders. In the wake of such personalized medicine, the use of induced pluripotent stem cell (iPSC)-derived BBB cells in MPS models has recently seen an increase [13][15,16][18][23]. BBB endothelial cells can now be derived from iPSCs [24]. At the same time, even tissue non-specific endothelial cells derived from iPSCs are amenable to be used in BBB models, as we have shown that their phenotype can be made more BBB-like (in terms of decreased permeability and gene expression of multiple key BBB markers) by co-culture with primary ACs and PCs [13]. Very recently, Vatine and co-workers showed that BBB maturation of iPSC-endothelial cells can also be achieved by co-culture with iPSC-neural cells, in a full iPSC-derived MPS model. The authors also produced cells from patients affected by two neurological disorders, Huntington disease and MCT8-specific deficiency, and measured differences in drug distribution across disease-specific BBB MPS models perfused with whole blood so as to capture any protein binding. Such personalized models, coupled with novel strategies to multiplex the culture of MPS [25], may pave the way for accurate and quantitative predictive models of CNS drug distribution in neurological disease.
127
2.2 Models of drug clearance
128 129 130 131 132
The use of MPS models to characterize drug distribution has been so far mostly limited to the characterization of vascular transport, with little attention to the destination of the drug. Past the endothelium, candidate drugs may engage their target or be cleared by the lymphatic vasculature found in the vicinity of the PNS and the periphery of the CNS [26]. Lymphatic vessels prevent accumulation of interstitial fluid and solutes, affecting local concentrations of drugs [27]. The
114 115 116 117 118 119 120 121 122 123 124 125
5
133 134 135 136 137 138 139 140
existing MPS models of the lymphatics are few, but they have made interesting findings in that drainage of interstitial molecules is increased in the presence of tissue-engineered lymphatic vessels [28], and that the co-culture of lymphatic endothelial cells with stromal cells significantly modifies such drainage [29]. Clearance also takes place within the vascular endothelium, where part of the molecules crossing the cells are cleared within lysosomes and prevented from crossing. We have characterized this clearance in a 3D microvascular model, showing that therapeutically relevant plasma proteins like albumin and immunoglobulin-G (IgG) are rescued from degradation by the Fc neonatal receptor (FcRN), although this results in an overall lower endothelial permeability [6].
141 142 143 144 145 146
Ultimately, the clearance of any therapeutic molecule is complex and takes place within multiple organs in the body, each critically affecting the availability of the drug and its capability to engage its target. MPS models of inter-connected organ systems have recently been developed [30][31]. These models can recapitulate aspects of the complex crosstalk between multiple organ tissues and the resulting pharmacokinetic characteristics that determine the absorption, distribution, metabolism and excretion of candidate drugs (ADME).
147 148
3. MICROPHYSIOLOGICAL MODELS OF DRUG ENGAGEMENT AND FUNCTION
149
159
Numerous cases of drugs that proved effective in tests on animal models, yet failed in human clinical trials, have led to the recognition that animal models may not be fully representative of human disease pathology. Largely for this reason, there has recently been a high demand for in vitro systems that can mimic the complexity of human tissues and allow measurement of pharmacological traits. The remarkable development efforts to create MPS related to the brain and nerves have been recently reviewed by Haring and co-workers [32]. Several MPS studies associated with neurological disease have now begun to directly involve quantitative drug screening and evaluation, as well as the study of disease mechanisms. Here, we introduce several recently developed MPS models of the neurological disorders so far addressed that directly involve drug assessment, such as models of AD, ALS, epilepsy, and Parkinson’s disease (PD). Other neurological conditions have yet to be modelled through MPS capable of addressing drug effects on neuronal dysfunction.
160
3.1 Alzheimer’s disease
161
Although several drugs have shown promise in animal models, many clinical trials of therapies for AD have failed, giving rise to tremendous interest in developing an in vitro AD model using human cells or patient-derived iPSCs aimed at drug screening. Recently, in vitro AD models have been developed focusing not only on neuronal loss itself, but also on other factors influencing the disease including the roles of the BBB and microglia. Our group developed a 3D in vitro model of AD by culturing human neural progenitor cells generating extracellular aggregation of amyloid beta (Aβ) and accumulation of hyperphosphorylated tau, with an intact BMEC barrier in a 3D microfluidic system [33]. By using genetically-modified cells that overexpress the different isoforms of amyloid beta [34,35], it was observed that the AD pathological microenvironment exerts a direct effect on BMECs in the BBB. In this 3D AD model, several vascular defects observed in AD patients were successfully mimicked, including an increase in BBB permeability. The studies further demonstrated that neurotoxins passing the leaky BBB can exacerbate neuronal loss in AD. Tests of pharmacological compounds that enhance BBB integrity were also performed, and drug efficacy characterized. Restoration of the disrupted BBB and decreased neuronal loss were observed, supporting the theory that restoration of BBB integrity could be a useful therapeutic target for AD by preventing
150 151 152 153 154 155 156 157 158
162 163 164 165 166 167 168 169 170 171 172 173 174 175
6
182
neurotoxins from entering the brain [33]. Park et al. developed a triculture system of neurons, astrocytes, and microglia to study the role of the latter in neuroinflammation in AD pathogenesis and progression. In this system, microglia recruitment induced by chemokines and cytokines secreted from AD culture and consequent neuron and astrocyte loss were observed. The authors tested IFN-γ or TLR4 blocking antibodies and the TLR4 antagonist LPS-RS, and observed reduced TNFα and NO concentrations and decreased LDH levels in the system, suggesting that neuronal loss depends on microglia activation via IFN-γ and TLR4-dependent mechanisms [36] .
183
3.2 Amyotrophic lateral sclerosis (ALS)
184
204
The recent advances in microfluidic fabrication techniques coupled with biological assembly have widened the horizon of ALS research by better mimicking the nerve-muscle connectivity. Machado et al. developed an in vitro model of a neuromuscular circuit [37]. This system consists of three compartments; a central one for myofiber culture and two outer compartments for motor neuron (MN) and astrocyte (AC) aggregates. Each compartment is connected by microchannel arrays. The system is designed to allow motor axons to extend from the motor neuron soma, project through an array of microchannels toward the central myofibers compartment, and form neuromuscular junctions. The authors demonstrated the capability of the system as an ALS disease model by coculturing MNs with ACs expressing ALS-linked SOD1G93A mutants and observed denervation of myofibers and diminished muscle contractions, both key features of ALS. Also, the authors showed that the candidate ALS drug RIPK1 inhibitor Necrostatin aided in the recovery of both axon growth and myofiber contractions. Our group also developed a 3D ALS motor unit model by co-culturing iPSC-derived MNs from a patient with sporadic ALS with iPSC-derived 3D muscle fiber bundles in a microfluidic system [38]. The system consists of three distinct culture compartments; for MN spheroid culture, neurite growth, and muscle tissue culture. Following muscle fiber bundle formation, MN spheroids derived from iPSCs either from a healthy subject or an ALS patient were introduced and allowed to elongate their neurites toward the muscle fiber bundles and form a functional motor unit with neuromuscular junctions. Significant differences were observed between the two models, in terms of neurite growth rate, force of muscle contraction, and skeletal muscle cell death. The authors tested two drug candidates of ALS, bosutinib and rapamycin, in the system and both proved effective in improving neuronal survival and muscle contraction.
205
3.3 Epilepsy
206
213
Epilepsy is a neurological disorder associated with abnormal neural electrical activity in the brain. Liu et al. developed a microfluidic-based microelectrode array device enabling the support of six organotypic hippocampal slice cultures on a single unit and the detection of separate electrographic recordings simultaneously [39]. Testing of a small molecule inhibitor of receptor tyrosine kinases (RTKs), which change in their expression or phosphorylation in the epileptic brain, was conducted in the system and antiepileptic properties were assessed via biomarker sampling of lactate for seizurelike activity and lactate dehydrogenase (LDH) for seizure-dependent cell death, and via electrical recordings for the electrophysiological verification of the biomarker results.
214
3.4 Parkinson’s disease
215
Systems that have the capability to be used as tools for testing drug efficacy in PD have been recently developed. 3D culture of PD-specific dopaminergic neurons incorporated in hydrogels
176 177 178 179 180 181
185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203
207 208 209 210 211 212
216
7
217 218 219 220 221 222 223 224
within a microfluidic system was found to recapitulate the key cellular defects in PD. Bolognin et al. cultured neuroepithelial stem cell lines derived from PD patients carrying leucine-rich repeat kinase 2 (LRRK2)-G2019S mutations in 3D microfluidic systems implemented with automated image analysis algorithms and observed a time-dependent loss of dopaminergic neurons caused by the mutation. Also, the authors compared 3D culture in microfluidic systems with 2D culture and demonstrated that 3D culture presents more robust PD endophenotypes. They inhibited LRRK2 and found it to effectively alleviate neurodegeneration and the appearance of LRRK2-G2019S-dependent dopaminergic phenotypes [40]. Disease
Alzheimer’s disease (AD)
Amyotrophic lateral sclerosis (ALS)
Cell types Viral-infected human neural progenitor cells (hNPC) with familial AD mutations, human brain endothelial cell line (hCMEC/D3) Viral-infected human neural progenitor cells (hNPC) with familial AD mutations, human iPSCderived neural progenitor cells, immortalized human microglia (SV40 cell line) Mouse embryonic stem cell (ESC)derived motor neurons and astrocytes, primary chick myoblasts
Drug target function Improve BBB integrity and alleviate neuronal loss
Analytical methods Permeability measurement with fluorescent dextrans for BBB integrity, viability test with redfluorescent ethidium homodimer-1 for neuronal loss
References [33]
Reduce neuroinflammation induced by microglia with IFN-γ or TLR4 blocking antibodies and the TLR4 antagonist LPS-RS
Quantification of inflammatory cytokine concentrations by ELISA
[36]
Improve motor neuron survival with RIPK1 inhibitor Necrostatin
Motor neurons differentiated from hESCderived neural stem cells (NSCs) and ALS
Improve motor neuron survival by inhibiting the Src/c-Abl signalling pathway (bosutinib) or inhibiting the mechanistic target of the mTOR
Immunostaining with [37] β3-tubulin antibody to assess axon outgrowth, analysis of myofiber contraction by optogenetic stimulation and quantification by particle image velocimetry (PIV) Measurement of [36] muscle force of contraction through optical stimulation and recording, assessment of
8
Epilepsy
225
patient iPSCderived NSCs, human iPSCderived skeletal myoblasts, iPSC-derived ECs Hippocampi dissected from postnatal day 7–8 SpragueDawley rat pups
pathway (rapamycin)
apoptosis and muscle atrophy by expression of caspase3/7.
Antiepileptic drug of small molecule inhibitors of receptor tyrosine kinases (RTKs)
Biomarker sampling [39] with Lactate and Lactate Dehydrogenase (LDH) Assays, measurement of neural activity with electrical recordings. Parkin’s Neuroepithelial Alleviate Live imaging with [40] disease (PD) stem cell lines neurodegeneration and tretamethylrhodamine, derived from the methyl ester (TKRK) healthy and PD LRRK2-G2019S-dependent and MitoTracker Green patients dopaminergic phenotypes to assess cell death carrying LRRK2- with LRRK2 inhibitor 2 and mitochondrial G2019S (Inh2) activity, mutation morphometric assay to analyze differences in skeleton, perimeter, body pixels, and complexity of the neural networks Table 1. MPS models employed to assess drug effects in neurological disorders.
226 227 228 229 230 231 232 233 234 235 236 237 238 239 240
Besides the MPS models addressed here, there are other MPS models of neurological diseases that have the potential to be used to evaluate drugs, despite the fact that their developers did not yet directly use the models for drug testing. For example, a co-culture platform integrated with microvalves that allow for precise control of cell-cell interactions, particularly via paracrine signaling, was developed to study synucleinopathies such as PD [41]. Also, a Huntington’s disease (HD)-on-a-chip model has been developed by culturing HdhCAG140/+ cortical and striatal neurons from heterozygous mice in a microfluidic system [42]. The microfluidic device consists of three compartments of cortical and striatal chambers that are connected by axonal and dendritic microchannels with an intermediate synaptic chamber. This MPS HD model showed that HD mature networks have pre- and postsynaptic defects and also defective and hypersynchronized corticostriatal networks. Finally, there are a few MPS models of GB addressing drug discovery for the disease [43,44]. However, these models focus only on drug effects on cancer cells, while there is a critical lack of models that address neuronal dysfunctions caused by the cancer cells, through integration of neural cell cultures (e.g. neurons or astrocytes) with GB cultures.
241 242
4. OUTLOOK
243 244 245
MPS models of neurological disease have the potential to play a critical role in drug development through pre-clinical testing of candidate drugs. The past several years have seen increased efforts in developing MPS models of neurological disorder recapitulating selected functions of human tissues 9
246 247 248
and organs; yet, there are still challenges to be considered and overcome, in particular with regard to the balance between complexity and high throughput of MPS models, and how this balance may affect their adoption in industry and the clinic.
249 250 251 252 253 254 255 256 257 258 259 260 261
On one hand, increased model complexity may offer advantages in recapitulating pathophysiological cross talk between different cell types, as well as a more detailed assessment of concurrent drug distribution, target engagement and function. For example, although drug vascular transport and clearance are key aspects to be considered in the evaluation of drug efficacy, currently most MPS models of neurological disorder do not incorporate the functional contribution of vasculature. In particular, models of vascular barriers such as the BNB and BCB are critically unexplored and could be integrated within MPS systems for drug testing in PNS disorders (e.g. ALS) or other peripheral disorders that can affect the CNS [26]. Similarly, since neurological diseases are long-term and progressive, MPSs would benefit from long-term culture to evaluate the exposure to drugs, which may be facilitated by the integration with a perfusion system that continuously provides nutrients to cells. On the other hand, high throughput screening can be a limiting factor in the evaluation of multiple candidate drugs. Current MPS models are inherently lower throughput, high complexity models.
262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278
Standardized, high throughput screening MPSs that incorporate complex, disease-specific 3D cellular microenvironments will likely have increased appeal when compared to animal models. For the purpose of screening drugs with better quality assurance/quality control (QA/QC), well-established and characterized human primary cells or cell lines that can recapitulate physiological or pathological general disease traits may be used. However, the use of multiple types of iPSC-derived cells from the same patient donor can provide more realistic recapitulation of the disease-associated phenotypes for personalized drug efficacy, sensitivity and toxicity screening, although possibly at the expense of high throughput. In addition, MPS model integrated with standardized functional readouts such as permeability test for BBB function (for CNS diseases related to BBB impairment), voltage signaling detection for testing neuronal function, and skeletal muscle contraction for testing muscular function (for ALS models), along with molecular and cellular readouts will provide a deeper insight into the pathophysiological events underlying neurological diseases. Ultimately, MPS models may not entirely replace animal models, but they could offer a strategy to further rank and select candidate molecules in vitro, thus reducing the number of animal models required, limiting costs while also addressing growing ethical concerns. Critically, these models provide drug testing platforms that are more relevant for human patients and may improve drug outcomes in clinical trials.
279 280 281 282 283 284 285 286
ACKNOWLEDGEMENTS
10
287 288 289
G.S.O. is supported by an American-Italian Cancer Foundation Post-Doctoral Research Fellowship. Y.S. is supported by the Cure Alzheimer’s Fund. R.D.K. has funding support from the National Science Foundation (CBET-0939511).
290 291
DISCLOSURES
292 293
R.D.K. is a co-founder of AIM Biotech that markets microfluidic systems for 3D culture. Funding support is also provided by Amgen, Biogen and Gore.
294 295
REFERENCES
296 297 298
1.
Little D, Ketteler R, Gissen P, Devine MJ: Using stem cell-derived neurons in drug screening for neurological diseases. Neurobiol Aging 2019, 78:130–141.
299 300
2.
Danon JJ, Reekie TA, Kassiou M: Challenges and opportunities in central nervous system drug discovery. Trends Chem 2019, 1:612–624.
301 302
3.
Osaki T, Shin Y, Sivathanu V, Campisi M, Kamm RD: In Vitro Microfluidic Models for Neurodegenerative Disorders. Adv Healthc Mater 2018, 7:1700489.
303 304
4.
Truskey GA: Human microphysiological systems and organoids as in vitro models for toxicological studies. Front public Heal 2018, 6:185.
305 306
5.
Hajal C, Campisi M, Mattu C, Chiono V, Kamm RD: In vitro models of molecular and nano-particle transport across the blood-brain barrier. Biomicrofluidics 2018, 12:42213.
307 308 309 310 311
6.
Offeddu GS, Haase K, Gillrie MR, Li R, Morozova O, Hickman D, Knutson CG, Kamm RD: An on-chip model of protein paracellular and transcellular permeability in the microcirculation. Biomaterials 2019, 212:115–125. *The authors separately assessed different physiological modes of endothelial molecular transport and assessed endothelial clearing of pharmacologically-relevant plasma proteins in a 3D model of the microvasculature.
312 313
7.
Gastfriend BD, Palecek SP, Shusta E V: Modeling the blood-brain barrier: beyond the endothelial cells. Curr Opin Biomed Eng 2018, 5:6–12.
314 315 316 317
8.
Helms HC, Abbott NJ, Burek M, Cecchelli R, Couraud P-O, Deli MA, Förster C, Galla HJ, Romero IA, Shusta E V, et al.: In vitro models of the blood--brain barrier: an overview of commonly used brain endothelial cell culture models and guidelines for their use. J Cereb Blood Flow Metab 2016, 36:862–890.
318 319
9.
DeStefano JG, Jamieson JJ, Linville RM, Searson PC: Benchmarking in vitro tissue-engineered blood-brain barrier models. Fluids Barriers CNS 2018, 15:32.
320 321
10.
Oddo A, Peng B, Tong Z, Wei Y, Tong WY, Thissen H, Voelcker NH: Advances in microfluidic bloodbrain barrier (BBB) models. Trends Biotechnol 2019,
322 323 324 325 326
11.
Herland A, van der Meer AD, FitzGerald EA, Park T-E, Sleeboom JJF, Ingber DE: Distinct contributions of astrocytes and pericytes to neuroinflammation identified in a 3D human bloodbrain barrier on a chip. PLoS One 2016, 11:e0150360. **The authors developed a 3D BBB model and assessed cytokine production by different cells of the NVU under induced inflammatory conditions.
11
327 328
12.
Wang YI, Abaci HE, Shuler ML: Microfluidic blood-brain barrier model provides in vivo-like barrier properties for drug permeability screening. Biotechnol Bioeng 2017, 114:184–194.
329 330 331
13.
Campisi M, Shin Y, Osaki T, Hajal C, Chiono V, Kamm RD: 3D self-organized microvascular model of the human blood-brain barrier with endothelial cells, pericytes and astrocytes. Biomaterials 2018,
332 333 334
14.
Lauranzano E, Campo E, Rasile M, Molteni R, Pizzocri M, Passoni L, Bello L, Pozzi D, Pardi R, Matteoli M, et al.: A Microfluidic Human Model of Blood-Brain Barrier Employing Primary Human Astrocytes. Adv Biosyst 2019, 3:1800335.
335 336 337 338 339 340
15.
Vatine GD, Barrile R, Workman MJ, Sances S, Barriga BK, Rahnama M, Barthakur S, Kasendra M, Lucchesi C, Kerns J, et al.: Human iPSC-Derived Blood-Brain Barrier Chips Enable Disease Modeling and Personalized Medicine Applications. Cell Stem Cell 2019, 24:995–1005. **The authors developed full iPSC-derived MPS models of the BBB using cells from healthy patients or patients affected by two neurological disorders, and perfused the models with whole blood showing differences in endothelial barrier properties.
341 342 343
16.
Linville RM, DeStefano JG, Sklar MB, Xu Z, Farrell AM, Bogorad MI, Chu C, Walczak P, Cheng L, Mahairaki V, et al.: Human iPSC-derived blood-brain barrier microvessels: Validation of barrier function and endothelial cell behavior. Biomaterials 2019, 190:24–37.
344 345
17.
Kanda T: Biology of the blood-nerve barrier and its alteration in immune mediated neuropathies. J Neurol Neurosurg Psychiatry 2013, 84:208–212.
346 347 348
18.
Park T-E, Mustafaoglu N, Herland A, Hasselkus R, Mannix R, FitzGerald EA, Prantil-Baun R, Watters A, Henry O, Benz M, et al.: Hypoxia-enhanced Blood-Brain Barrier Chip recapitulates human barrier function and shuttling of drugs and antibodies. Nat Commun 2019, 10:2621.
349 350 351
19.
Papademetriou I, Vedula E, Charest J, Porter T: Effect of flow on targeting and penetration of angiopep-decorated nanoparticles in a microfluidic model blood-brain barrier. PLoS One 2018, 13:e0205158.
352 353 354 355 356
20.
Offeddu GS, Possenti L, Loessberg-Zahl JT, Zunino P, Roberts J, Han X, Hickman D, Knutson CG, Kamm RD: Application of Transmural Flow Across In Vitro Microvasculature Enables Direct Sampling of Interstitial Therapeutic Molecule Distribution. Small 2019, 1902393. *The authors introduced the contribution of transmural fluid flow to drug transport in a 3D model of the microvasculature.
357 358 359
21.
Mestre H, Tithof J, Du T, Song W, Peng W, Sweeney AM, Olveda G, Thomas JH, Nedergaard M, Kelley DH: Flow of cerebrospinal fluid is driven by arterial pulsations and is reduced in hypertension. Nat Commun 2018, 9:4878.
360 361 362
22.
Kutuzov N, Flyvbjerg H, Lauritzen M: Contributions of the glycocalyx, endothelium, and extravascular compartment to the blood-brain barrier. Proc Natl Acad Sci 2018, 115:E9429-E9438.
363 364
23.
Motallebnejad P, Thomas A, Swisher SL, Azarin SM: An isogenic hiPSC-derived BBB-on-a-chip. Biomicrofluidics 2019, 13:64119.
365 366
24.
Shusta E V, Azarin S, Palecek S, Lippmann E: Human blood-brain barrier endothelial cells derived from pluripotent stem cells and blood-brain barrier model thereof. 2018,
367 368
25.
Lee S-R, Hyung S, Bang S, Lee Y, Ko J, Lee S, Kim HJ, Jeon NL: Modeling neural circuit, blood-brain barrier, and myelination on a microfluidic 96 well plate. Biofabrication 2019, 11:35013.
369
26.
Louveau A, Herz J, Alme MN, Salvador AF, Dong MQ, Viar KE, Herod SG, Knopp J, Setliff JC, Lupi AL, 12
et al.: CNS lymphatic drainage and neuroinflammation are regulated by meningeal lymphatic vasculature. Nat Neurosci 2018, 21:1380.
370 371 372 373
27.
Chang C-W, Seibel AJ, Song JW: Application of microscale culture technologies for studying lymphatic vessel biology. Microcirculation 2019,
374 375 376
28.
Thompson RL, Margolis EA, Ryan TJ, Coisman BJ, Price GM, Wong KHK, Tien J: Design principles for lymphatic drainage of fluid and solutes from collagen scaffolds. J Biomed Mater Res Part A 2018, 106:106–114.
377 378 379 380
29.
Gong MM, Lugo-Cintron KM, White BR, Kerr SC, Harari PM, Beebe DJ: Human organotypic lymphatic vessel model elucidates microenvironment-dependent signaling and barrier function. Biomaterials 2019, 214:119225. **The authors developed a 3D lymphatic vessel model that can be separtely perfused to measure lymphatic permeability and drainage.
381 382 383
30.
Edington CD, Chen WLK, Geishecker E, Kassis T, Soenksen LR, Bhushan BM, Freake D, Kirschner J, Maass C, Tsamandouras N, et al.: Interconnected microphysiological systems for quantitative biology and pharmacology studies. Sci Rep 2018, 8:4530.
384 385 386
31.
Prantil-Baun R, Novak R, Das D, Somayaji MR, Przekwas A, Ingber DE: Physiologically based pharmacokinetic and pharmacodynamic analysis enabled by microfluidically linked organs-onchips. Annu Rev Pharmacol Toxicol 2018, 58:37–64.
387 388 389
32.
Haring AP, Sontheimer H, Johnson BN: Microphysiological human brain and neural systems-on-achip: potential alternatives to small animal models and emerging platforms for drug discovery and personalized medicine. Stem cell Rev reports 2017, 13:381–406.
390 391 392 393
33.
Shin Y, Choi SH, Kim E, Bylykbashi E, Kim JA, Chung S, Kim DY, Kamm RD, Tanzi RE: Blood-Brain Barrier Dysfunction in a 3D In Vitro Model of Alzheimer’s Disease. Adv Sci 2019, 1900962. **The authors developed a 3D blood-brain barrier model of AD and observed AD pathological microenvironment directly effects on brain endothelial cells of BBB.
394 395 396
34.
Kim YH, Choi SH, D’avanzo C, Hebisch M, Sliwinski C, Bylykbashi E, Washicosky KJ, Klee JB, Brüstle O, Tanzi RE, et al.: A 3D human neural cell culture system for modeling Alzheimer’s disease. Nat Protoc 2015, 10:985.
397 398 399
35.
Choi SH, Kim YH, Hebisch M, Sliwinski C, Lee S, D’Avanzo C, Chen H, Hooli B, Asselin C, Muffat J, et al.: A three-dimensional human neural cell culture model of Alzheimer’s disease. Nature 2014, 515:274–278.
400 401 402 403 404
36.
Park J, Wetzel I, Marriott I, Dréau D, D’Avanzo C, Kim DY, Tanzi RE, Cho H: A 3D human triculture system modeling neurodegeneration and neuroinflammation in Alzheimer’s disease. Nat Neurosci 2018, 21:941. **The authors developed a 3D AD model to study neural-glial interactions in AD pathology and observed physical interruptions between microglia and nuerons in AD pathological microenvironment.
405 406 407
37.
Machado CB, Pluchon P, Harley P, Rigby M, Sabater VG, Stevenson DC, Hynes S, Lowe A, Burrone J, Viasnoff V, et al.: In Vitro Modeling of Nerve-Muscle Connectivity in a Compartmentalized Tissue Culture Device. Adv Biosyst 2019,
408 409 410 411 412
38.
Osaki T, Uzel SGM, Kamm RD: Microphysiological 3D model of amyotrophic lateral sclerosis (ALS) from human iPS-derived muscle cells and optogenetic motor neurons. Sci Adv 2018, 4:eaat5847. **The authors developed a 3D ALS model by co-culturing skeletal muscle bundles with iPSCderived motor neuron spheroids and observed the ALS motor unit generated fewer muscle contractions and motor neuron degradation compared to a non-ALS motor unit.
413
39.
Liu J, Sternberg AR, Ghiasvand S, Berdichevsky Y: Epilepsy-on-a-chip System for Antiepileptic 13
Drug Discovery. IEEE Trans Biomed Eng 2018, 66:1231–1241.
414 415 416 417 418 419 420
40.
Bolognin S, Fossépré M, Qing X, Jarazo J, Ščančar J, Moreno EL, Nickels SL, Wasner K, Ouzren N, Walter J, et al.: 3D Cultures of Parkinson’s Disease-Specific Dopaminergic Neurons for High Content Phenotyping and Drug Testing. Adv Sci 2019, 6:1800927. **The authors developed a 3D model of PD with patient neurons derived from iPSC carrying the LRRK2-G2019S mutation. They compared 2D culture and 3D culture of dopaminergic neurons and found that 3D culture presents rubust endophenotypes compared to 2D culture.
421 422 423
41.
Fernandes JTS, Chutna O, Chu V, Conde JP, Outeiro TF: A novel microfluidic cell co-culture platform for the study of the molecular mechanisms of Parkinson’s disease and other synucleinopathies. Front Neurosci 2016, 10:511.
424 425 426
42.
Virlogeux A, Moutaux E, Christaller W, Genoux A, Bruyere J, Fino E, Charlot B, Cazorla M, Saudou F: Reconstituting corticostriatal network on-a-chip reveals the contribution of the presynaptic compartment to Huntington’s disease. Cell Rep 2018, 22:110–122.
427 428 429
43.
Yi H-G, Jeong YH, Kim Y, Choi Y-J, Moon HE, Park SH, Kang KS, Bae M, Jang J, Youn H, et al.: A bioprinted human-glioblastoma-on-a-chip for the identification of patient-specific responses to chemoradiotherapy. Nat Biomed Eng 2019, 1.
430 431
44.
Fan Y, Nguyen DT, Akay Y, Xu F, Akay M: Engineering a brain cancer chip for high-throughput drug screening. Sci Rep 2016, 6:25062.
432
14
We wish to draw the attention of the Editor to the following facts which may be considered as potential conflicts of interest and to significant financial contributions to this work: R.D.K. is a co founder of AIM Biotech that markets microfluidic systems for 3D culture. Funding support is also provided by Amgen, Biogen and Gore. We confirm that the manuscript has been read and approved by all named authors and that there are no other persons who satisfied the criteria for authorship but are not listed. We further confirm that the order of authors listed in the manuscript has been approved by all of us. We confirm that we have given due consideration to the protection of intellectual property associated with this work and that there are no impediments to publication, including the timing of publication, with respect to intellectual property. In so doing we confirm that we have followed the regulations of our institutions concerning intellectual property. We understand that the Corresponding Author is the sole contact for the Editorial process (including Editorial Manager and direct communications with the office). He/she is responsible for communicating with the other authors about progress, submissions of revisions and final approval of proofs. We confirm that we have provided a current, correct email address which is accessible by the Corresponding Author and which has been configured to accept email from [email protected]. Signed by all authors as follows: Giovanni Offeddu
10/22/2019
Yoojin Shin
10/22/2019
Roger Kamm
10/22/2019