Functional bioengineered models of the central nervous system

Abstract

The functional complexity of the central nervous system (CNS) is unparalleled in living organisms. Its nested cells, circuits and networks encode memories, move bodies and generate experiences. Neural tissues can be engineered to assemble model systems that recapitulate essential features of the CNS and to investigate neurodevelopment, delineate pathophysiology, improve regeneration and accelerate drug discovery. In this Review, we discuss essential structure-function relationships of the CNS and examine materials and design considerations, including composition, scale, complexity and maturation, of cell biology-based and engineering-based CNS models. We highlight region-specific CNS models that can emulate functions of the cerebral cortex, hippocampus, spinal cord, neural-X interfaces and other regions, and investigate a range of applications for CNS models, including fundamental and clinical research. We conclude with an outlook to future possibilities of CNS models, highlighting the engineering challenges that remain to be overcome.

Keywords: Biomimetics; Learning and memory.

Fig. 1. Timelines of engineering and biology advances towards neural tissue engineering.

CNS, central nervous system; ES cell, embryonic stem cell; iPS cell, induced pluripotent stem cell; MEA, microelectrode array; NSC, neural stem cell; PSC, pluripotent stem cell^–.

Fig. 2. Anatomy of the central nervous system at multiple scales.

Bioengineered tissue models of the central nervous system (CNS) are designed and built to mimic the structures and functions of in vivo tissues. At the macroscale, the CNS can be subdivided into gross regions, including the cerebrum, brainstem and spinal cord. The cerebrum can be further partitioned into several mesoscale sub-organs, such as the cerebral cortex, basal ganglia and hippocampus — all of which are encased within a dense field of nerve fibres (that is, white matter). At the microscale, cells form highly ordered circuits, interfaces and pathways that reflect complex functions. Cortical and hippocampal circuits display re-entrant or looped activation patterns, which enable cell synchronization and the integration of information associated with perception and memory. However, when left unchecked, re-entrant loops can precipitate seizure activity. Similarly, the neurovascular unit forms a blood–brain barrier (BBB) by tightly joining endothelial cells, pericytes, glial cells and neurons in concentric layers — a selectively permeable gateway that becomes compromised with neurodegenerative disease, cancers, infections and trauma. At finer scales, cells are immersed within complex microenvironments patterned with biophysical and biochemical cues that determine their fate and behaviour. To accurately model CNS function in vitro, the essential elements need to be identified that optimally recapitulate structures and functions of the CNS. CA, cornu ammonis.

Fig. 3. Design elements in functional central nervous system modelling.

Design elements can be selected and combined to recapitulate desired physiological states in bioengineered central nervous system (CNS) tissues. The composition, dimensions, maturity, complexity and interfaces can be customized by selecting different design elements. ECM, extracellular matrix.

Fig. 4. Functional 3D bioengineered central nervous system models.

Engineering-based models combine materials and cell sources to generate biomimetic central nervous system (CNS) tissues. Synthetic materials (for example, polycaprolactone (PCL), poly(3,4-ethylenedixoythiophene) (PEDOT)), biomaterials (such as chitosan, collagen, silk) or combinations thereof are used to generate highly customizable and tractable hydrogels, 3D bioprintable inks, scaffolds and organ-on-a-chip systems. However, engineered models require judicious selection and micromanagement of design elements to avoid generating physiologically aberrant tissues. Cell biology-based models offer greater cytoarchitectonic complexity and ideal 3D microenvironments at the cost of some customizability. Self-assembling organoids, assembloids and spheroids, derived from pluripotent stem cells (PSCs) and primary tissues, do not require support materials, and display conserved neurodevelopmental programmes consistent with in vivo tissues.