The potential of biomaterials for central nervous system cellular repair

The central nervous system (CNS) can be injured or damaged through a variety of insults including traumatic injury, stroke, and neurodegenerative or demyelinating diseases, including Alzheimer's disease, Parkinson's disease and multiple sclerosis. Existing pharmacological and other therapeutics strategies are limited in their ability to repair or regenerate damaged CNS tissue meaning there are significant unmet clinical needs facing patients suffering CNS damage and/or degeneration. Through a variety of mechanisms including neuronal replacement, secretion of therapeutic factors, and stimulation of host brain plasticity, cell-based repair offers a potential mechanism to repair and heal the damaged CNS. However, over the decades of its evolution as a therapeutic strategy, cell-based CNS repair has faced significant hurdles that have prevented its translation to widespread clinical practice. In recent years, advances in cell technologies combined with advances in biomaterial-based regenerative medicine and tissue engineering have meant there is very real potential for many of these hurdles to be overcome. This review will provide an overview of the main CNS conditions that lend themselves to cellular repair and will then outline the potential of biomaterial-based approaches for improving the outcome of cellular repair in these conditions.


Introduction
Although cellular repair for certain central nervous system (CNS) conditions has been the subject of intense research at both preclinical and clinical level for several decades, limitations such as poor cell survival and engraftment, ethical and logistical constraints, and inadequate sources of donor cells, have meant that this approach it not yet part of the armoury for treating conditions such as traumatic brain injury, stroke or neurodegenerative disease. Nevertheless, recent years have seen considerable developments in cell technologies, with the birth of the stem cell era; and parallel developments in regenerative medicine and tissue engineering, with the rational design of biomaterials for CNS compatibility. The first half of this review will provide an overview of CNS conditions that have the potential to be treated using cellular repair, while the second half will then outline the progress that has been made in designing biomaterials for improving the outcome of cell-based CNS repair in these conditions.

Cell-based repair
The CNS can be injured by trauma, stroke and a range of neurodegenerative diseases that can lead to damage and loss of neuronal tissue. Damage to the CNS can be especially detrimental because of its limited capacity for spontaneous regeneration. The burden that CNS injuries cause in our society and the limited success of current pharmacological therapies make the development of new therapeutic interventions that can target the regeneration of CNS tissue necessary. Specifically, cellbased repair is a strategy to replace cells and repair damaged or degenerating tissue of the CNS. The therapeutic potential of cell-based therapies is the reconstruction of damaged circuitry through 1) the direct replacement of the damaged neurons, 2) secretion of therapeutic factors that improve the pathological host environment such as those that promote neuroprotection, anti-inflammatory effects and/or enhanced angiogenesis, and 3) the creation of an environment that promotes host plasticity (Yasuhara et al., 2019). Extensive preclinical data has demonstrated the therapeutic benefits of cell repair and in many cases its efficacy is now being assessed in clinical trials. However, there are many complex hurdles that must be overcome before cell-based repair for the CNS can be translated into a safe, efficacious and widespread clinical practice.

Cell sources
There are several sources of cells that can be used for CNS repair, including primary neural tissue dissected from the developing embryonic or foetal brain, or different types of stem cells including embryonic, mesenchymal, neural and induced pluripotent stem cells. Embryonic stem cells (ESC) are self-renewing pluripotent stem cells derived from the inner cell mass of the blastocysts and have the capacity to differentiate into all three germ layers: endoderm, mesoderm, and ectoderm. Although these cells have the potential to be an effective donor source for cell therapy, they have ethical and logistical limitations, and they carry a risk of teratoma formation after implantation if they are not properly differentiated (Björklund et al., 2002). Mesenchymal stem cells (MSC) are non-hematopoietic, multipotent adult stem cells that have reparative potential in the CNS largely through the beneficial effects of their secretome (Drago et al., 2013). MSCs can be isolated from various tissue, such as bone marrow, amniotic fluid and adipose tissue, making them less ethically controversial than ESCs (Ullah et al., 2015). Neural stem cells (NSC) are the ancestor cells for the CNS, with the ability to give rise to neurons, astrocytes, and oligodendrocytes. NSCs can be taken from either the foetal or adult brain. A benefit of NSCs over pluripotent ESCs or multipotent MSCs is that there is no need to program the cells towards a neural lineage, since NSCs are already neural progenitors. This can also reduce the risk of tumour formation through residual undifferentiated stem cells. Finally, induced pluripotent stem cells (iPSC) are pluripotent stem cells reprogrammed from adult somatic cells, such as skin fibroblasts, by retrovirally introducing gene transcription factors (Oct3/4, Sox2, Klf4, and c-Myc) (Takahashi and Yamanaka, 2006). This eliminates the ethical restrictions of embryonic-derived stem cells and opens the door to patient-specific technologies. However, iPSCs also involve major obstacles, including the reprogramming process itself which requires rigorous pre-clinical assessments before clinical translation (Liang and Zhang, 2013).

Spinal cord injury
Spinal cord injury (SCI) is caused by contusion or partial/complete severance of the spinal cord and can result in a loss of sensory and/or motor function. Cell transplantation is being explored as a therapeutic strategy for SCI through a number of mechanisms; replacing the damaged cells, providing neurotrophic support, modulating the host immune response, and enhancing axonal plasticity are a few examples (Goulão and Lepore, 2016). Transplantation of NPCs in both rodent and non-human primate SCI models have resulted in synaptic formation between the graft and host neurons, and successfully improved motor function (Kadoya et al., 2016;Rosenzweig et al., 2018). The transplantation of MSCs into a rodent SCI model also resulted in amelioration of functional deficits (Chopp et al., 2000). iPSC-derived neurospheres were also transplanted into a rodent SCI model and were found to differentiate into three types of neural cell, neurons, astrocytes, and oligodendrocytes, resulting in remyelination of the host axons and promotion of functional recovery (Tsuji et al., 2010). Another approach for cell repair in SCI is through delivery of Schwann cells, which support axon outgrowth and re-growth of peripheral nerves (Paino and Bunge, 1991).
The first clinical trial of the transplantation of NSCs for the treatment of SCI in 4 patients was conducted in 2018, which found no adverse events from the grafts up to 27 months while sensory and motor function improved in 2 of the subjects (Curtis et al., 2018). While cell-based repair has proved to be a promising therapy for SCI, there a several limitations of cell therapy such as undirected migration, cell death, and limited effectiveness of neurotrophic factors that must be addressed for it to become safe and effective enough for clinical use. The efficacy of cell-based therapy for SCI can be greatly enhanced when used in combination with biomaterial-based approaches such as nerve guidance channels or scaffolds for structural support (Straley et al., 2010).

Traumatic brain injury
Traumatic brain injury (TBI) is damage to the brain caused by sudden impact to the head and is a leading cause of death and disability among young people (Barlow, 2013). Cell-based repair offers a potential therapeutic approach to replace the lost tissue and repair the damage in the brain from TBI. The transplantation of MSCs into the brains of rodents with TBI has been found to improve motor function . It has also been found that when MSCs were administered intravenously to rodents with TBI, the cells successfully migrated to the injury site and the rats displayed improved motor, sensory and reflex function Mahmood et al., 2003). Stem cell therapy has also been tested clinically for the treatment of patients with TBI (Trounson and McDonald, 2015;Schepici et al., 2020). A recent phase 2 clinical trial (NCT02416492) that was completed in 2019 called STEMTRA looked at the effects of intracranial administration of bone marrow-derived cells on patients with chronic motor deficit from TBI, which found that significantly more patients that received the cell treatment showed improved motor status at 6-months compared to patients that received a control sham surgery (McAllister et al., 2020).

Stroke
Stroke is caused by a restriction of blood flow to part of the brain and it is the 2nd leading cause of death worldwide (Feigin et al., 2017). The two main types of strokes are ischemic, which makes up approximately 80% of all strokes and is when the blood supply has stopped due to a blood clot; and haemorrhagic, when a burst blood vessel causes bleeding and damage in the brain (Ingall, 2004). Both murine iPSC-derived neuronal precursors and ESC-derived neuronal precursors demonstrated survival, neuronal differentiation, and improvement of function in stroke lesions of rats (Bühnemann et al., 2006;Chau et al., 2014). It has also been found that human NSCs, both peripherally injected or transplanted directly into the brain of rodent stroke models, survived and differentiated in the brain and improved sensorimotor deficits (Chu et al., 2003(Chu et al., , 2004Ishibashi et al., 2004;Lee et al., 2007). However, NSCs transplanted into a non-human primate stroke model had restricted migration compared to what was generally observed in rodents (Roitberg et al., 2006). Cell-based repair can be a promising therapy for stroke not only through replacement of lost cells but also by creating a neuroprotective environment in the damaged region of the brain (Krause et al., 2019). Treatment with MSCs, intravenously or injected through the internal carotid artery, in a rodent stroke model has been found to reduce the glial scar surrounding the damaged tissue (Li et al., 2005;Shen et al., 2006). Since 2005 there have been several clinical trials that assessed cell therapies for stroke, which have predominantly found them feasible and safe but with limited clinical efficacy (Krause et al., 2019). Although cell-based therapy has the potential to be a successful treatment for stroke, there are many challenges that must first be faced, such as low yields of the transplanted cells and the heterogeneity of the disease and damaged tissue.

Neurodegenerative diseases
2.2.4.1. Parkinson's disease. Parkinson's disease (PD) is the second most common neurodegenerative disease in the world and is associated with the degeneration of dopaminergic neurons in the substantia nigra. The primary clinical characteristics of the disease are resting tremor, hypokinesia, muscular rigidity, and postural instability (Winner et al., 2009;Yasuhara et al., 2017). Dopaminergic cell transplantation has the potential to repair the nigrostriatal pathway and restore dopamine levels in the brain, therefore improving motor function in patients. Since the first cell transplant for PD in 1979 (Perlow et al., 1979), which found that dopamine neuron-rich rat foetal grafts reduced motor abnormalities in a rat model of PD, there have been many animal studies of cell therapy using MSCs, ESCs, NSCs and iPSCs (Yasuhara et al., 2017;Fan et al., 2020). These preclinical studies overwhelmingly support the efficacy of cell replacement therapy for PD thus offering a potential alternative to pharmacological treatments. There are also now several clinical studies using cell repair in PD. For instance, TRANSEURO is a European research consortium that was established in 2010 (NCT01898390) to test the clinical efficacy of using foetal ventral mesencephalic transplants for cell replacement in PD (Barker et al., 2019). In addition to foetal tissue, the first clinical trial began in 2018 using human leukocyte antigen (HLA)-matched human iPSC-derived neurons for patients with PD (Takahashi and Price-Evans, 2019). There was even a recent case study, the first of its kind, in which an autologous transplant of patient-derived iPSCs was used in a patient with PD in order to reduce the risk of graft rejection (Schweitzer et al., 2020). While the homogeneity of the damaged tissue, for the most part, makes PD an ideal candidate for cell-based repair, the poor survival of the transplanted cells hampers its broad translation into the clinic.

Alzheimer's disease.
Alzheimer's disease (AD) is the most common neurodegenerative disorder and is characterised by memory impairment and cognitive decline. AD also involves the degeneration of neurons in the brain, lending itself as potentially suitable for cell replacement therapy. However, unlike PD, there has been little research done on the transplantation of stem cell-derived neurons due to the heterogeneity and widespread distribution of the degenerating cells in AD. One alternative approach taken to slow neuronal degeneration in AD is through the transplantation of astrocytes genetically engineered to secrete nerve growth factor (NGF), which has been found to improve metabolic activity in affected brain areas and slow cognitive decline (Tuszynski et al., 2005). The transplantation of both placenta and adipose-derived MSCs in AD mice have also been found to promote neurogenesis and alleviate cognitive impairment (Yun et al., 2013;Yan et al., 2014). It has been found that the transplantation of NSCs can increase hippocampal synaptic density and improve cognitive function in transgenic AD mice, but this was determined to be mediated by the production of BDNF by the NSCs (Blurton-Jones et al., 2009). There are several clinical trials testing the use of MSC infusions in patients with AD (Cummings et al., 2020). For instance, the Korean company Medipost has completed a phase 2 clinical trial (NCT02054208) to test intra-ventricular administration of a product called NEUROSTEM®, which is umbilical cord blood-derived MSCs, in patients with AD. Another product, AstroStem, uses adipose tissue-derived MSCs and will also soon be tested in patients with AD in a phase 2 clinical trial (NCT04482413). Although the replacement of damaged tissue through cell-based repair may not be feasible, the promotion of a neuroprotective and neuroregenerative environment through the transplantation of therapeutic factor-secreting cells is a potential therapy for AD.

Multiple sclerosis.
Multiple sclerosis (MS) is a chronic inflammatory disease of the CNS, leading to demyelination of axons with consequent functional impairments for patients. Cell-based repair is a potential therapy for MS, through the mechanisms of immunomodulation and remyelination (Martino et al., 2010). For instance, MSCs have been found to interfere with autoimmunity in mice with experimental autoimmune encephalomyelitis (EAE), an animal model of MS, through suppression of effector T cells (Zappia et al., 2005). It has also been found that transplantation of human ESC-derived neural precursor cells into the lateral ventricles of EAE mice reduced physical symptoms (Aharonowiz et al., 2008). While there was no increase in remyelination or evidence that the NPCs differentiated into oligodendrocytes, there was an immunosuppressive and neuroprotective effect of the NPCs in the EAE mice. In humans, there is a phase 3 clinical trial (NCT04047628) just beginning called BEAT-MS (BEst Available Therapy versus autologous hematopoietic stem cell transplant for Multiple Sclerosis) sponsored by the National Institute of Allergy and Infectious Disease that is testing the transplantation of autologous hematopoietic stem cells in patients with treatment-resistant MS.

Limitations of cell-based repair
Cell-based repair has shown the potential to be an alternative treatment to target CNS damage and ameliorate consequent functional deficits through direct replacement of damaged tissue or through other mechanisms, including enhancing neuroprotection and reducing pathological processes. However, cell repair approaches face many obstacles that hamper their broad clinical success. Pre-clinical studies using cell delivery for CNS tissue regeneration have encountered poor cell survival and integration in the host tissue after transplantation (Tam et al., 2014;Assunção-Silva et al., 2015;Praet et al., 2015;Mitrousis et al., 2018). The cells being transplanted are often coming directly from a nutrient and growth factor-rich environment -of either foetal tissue or a controlled cell culture -and put into a diseased or injured adult CNS, which lacks the necessary growth factors for the proliferation and survival of transplanted cells (Praet et al., 2015). In many cases, the transplanted cells also need some form of structural support, especially in spinal cord repair because of the required axonal alignment. Therefore, it appears necessary to refine the delivery of the transplanted cells to enhance their survival and maximise the efficiency of such therapies. Advances in cell replacement therapies, regenerative medicine and tissue engineering offer multiple new approaches for brain repair. Biomaterials -natural or synthetic materials tailored to provide a beneficial effect in a targeted biological system -have the potential to overcome some of the current hurdles of cell replacement therapies and ultimately improve brain repair and regeneration (Orive et al., 2009;Moriarty and Dowd, 2018;Moriary et al., 2019a;Tuladhar et al., 2018;Bruggeman et al., 2019).

Designing biomaterials for cell-based brain repair
Cell-based therapy intends to protect and maintain the viability of damaged host cells or to replace them, while promoting tissue regeneration. Although cell-based therapies have the potential to mitigate the pathological consequences of traumatic CNS injuries, stroke and neurodegenerative conditions, they are nevertheless associated with significant limitations that hamper their translation into clinical practice. Biomaterials have the potential to overcome some of these limitations and they can be harnessed to enhance brain protection, repair and regeneration (Orive et al., 2009;Moriarty and Dowd, 2018;Tuladhar et al., 2018;Bruggeman et al., 2019). Their primary application in this context is as scaffolds to aid with, and improve, cell delivery and engraftment. Biomaterial-based scaffolds serve as a physical structure for cells to adhere to during and after the transplantation process, as a physical barrier against the immune response upon transplantation, and as a reservoir of therapeutic biomolecules at the transplant (Burdick et al., 2016). Essentially, delivering cells in a biomaterial has the potential to provide the implanted cells with a more supportive, protective and trophic microenvironment when compared to the adult, diseased brain into which they are transplanted (see Figs. 1 and 2).
Although biomaterials have successfully been used for reparative and regenerative purposes in many diseases, conditions and tissues for decadessome examples would be in wound healing or in cardiovascular devices (Langer and Vacanti, 2016) their use in the brain is still limited. Designing biomaterials for brain repair has its own unique challenges due to the brain's restricted accessibility and limited capacity for self-repair, the existence of the blood brain barrier (BBB), and the cellular and functional complexity of the brain.
Biomaterials constitute an attractive strategy for the heterogeneous field of cellular brain repair due to their broad diversity and high adaptability to the tissue, disease and intended use (Orive et al., 2009). The recent expansion and success of biomaterials in the medical field is due, in part, to the rational design of biomaterials, specific for each particular tissue, condition and application.

Ideal properties of biomaterials for cell-based brain repair
The inaccessibility, fragility and complexity of the brain challenge the design and development of biomaterials for cell-based brain repair. Although the criteria for designing biomaterials for the brain are predominantly specified by the specific application and target site, some essential requirements emerge.
A key obstacle for the use of biomaterials in the brain is the difficulty in accessing the target site (Newland et al., 2016). Indeed, complex invasive surgery is part-and-parcel of the therapeutic armoury for some neurological conditions, for example, deep brain stimulation for PD (Mansouri et al., 2018;Peng et al., 2018;Harmsen et al., 2020). Nevertheless, if biomaterials are to be used for cell-based brain repair, they should be designed to be delivered in a minimally invasive manner. In this context, although early investigations used implantable scaffolds, the enclosed location of the brain has promoted the development of in situ forming injectable scaffolds, also known as hydrogels, that can be implanted with minimally invasive surgery (Nih et al., 2016). Biomaterials should also be chemically and physically stable long enough to perform their desired biological function and should be completely biodegradable with no remaining residues once they have served their purpose. Furthermore, they should have long-term biocompatibility with the host tissue, and neither the parent material, nor any of its degradative by-products, should generate a host immunogenic response. This is of particular importance when using synthetic biomaterials that are susceptible to more complex degradation patterns. Finally, the developed biomaterials should be scalable to be mass-produced for their introduction to clinical practice.
It is not only the target site where the biomaterial will be used that has to be considered, as the intended application of the biomaterial is of considerable importance too. To support cell-based brain repair, biomaterials should provide a supportive structure that resembles the natural neural tissue, as this facilitates cell survival, proliferation, and differentiation of both transplanted and endogenous cells (Guilak et al., 2009). The ideal scaffold should mimic the chemical, physical and architectural properties of the natural neural extracellular matrix (ECM) (Niemczyk et al., 2018), and its mechanical properties should match that of the neural tissue (Khaing et al., 2014;Zuidema et al., 2014). For this reason, natural biomaterials mainly made of components present in the extracellular matrix such as proteoglycans, non-proteoglycan polysaccharides (such as hyaluronic acid) and proteins (such as collagens) are widely used. Although using natural materials may seem the most practical and straightforward option, synthetic biomaterials offer other interesting advantages such as adding non-naturally occurring properties of interest to the final biomaterial. For instance, synthetic biomaterials can be synthesized to be reactive to the host environment, and their chemical and physical properties can be more extensively modified.

Adapting architecture for cell-based repair
To choose the best biomaterial available for a specific therapeutic approach, its chemical, physical and biological properties as well as its interaction with the target tissue has to be taken into account. In the context of brain cell repair, hydrogels and nanoparticles are, by far, the ones that caught more interest. Both biomaterials can be designed to fulfil the properties outlined previously in this review. It is important to note that this is not an exhaustive review of all the materials that have been used for cell therapy or/and drug delivery, rather it summarises the benefits of using thoughtfully developed biomaterials for regenerative purposes. Some examples of functionalisation of biomaterials for brain repair are listed in Table 1.

Hydrogels
Hydrogels have multiple properties that make them excellent scaffolds to be used in cell transplantation therapies. First of all, hydrogels can provide structural support for the transplanted cells, becoming the neural substrate where cells can attach and mature. Structural support facilitates cell survival and this is of special interest in stroke, where tissue is lost after injury and transplanted cells are required to form de novo tissue. Conveniently, injectable hydrogels can form gels in situ after a pH or temperature change, which allows their intra-cranial delivery (Pakulska et al., 2012). Thus, injectable hydrogels can protect the transplanted cells from the mechanical forces exerted during the injection process, which is known to contribute to cell death (Mitrousis et al., 2018). In addition to damage during the injection process, it has been found that the lifting of the cells prior to the transplantation actually causes the majority of the cell loss. This damage from lifting the cells can be overcome by maturing and transplanting the cells within a 3D culture that encapsulates the cells in a hydrogel (Adil et al., 2017). Moreover, hydrogels can protect the transplanted cells upon transplantation against the host immune response by physically isolating them from the host microenvironment. The reduction of the host immune response by biomaterial-based scaffolds is linked to an improvement of survival of transplanted cells throughout studies in many rodent models such as PD, stroke and traumatic brain injury (Park et al., 2002;Cheng et al., 2013;Hoban et al., 2013).
Hydrogels are a class of highly hydrated biomaterials with hydrophilic three-dimensional polymeric networks bound together by chemical or physical crosslinks. The properties of hydrogels coupled with endless possibilities for tuning and tailoring make them an excellent candidate to be used as scaffolds in brain cell repair. Thus, it is perhaps not surprising that hydrogels are one of the most used approaches in neuroregenerative strategies (Peppas et al., 2006). Hydrogels can be fabricated out of natural or synthetic polymers, and these can be chemically or physically crosslinked to ensure structural stability. The chemical crosslinking can also be used to modulate many aspects of the final biomaterial. For example, it can be adjusted to regulate factors that interfere with cell deployment and the properties of the hydrogel itself like its gelation process, porosity and its degradation pattern (Hennink and Van Nostrum, 2012).
As a biomaterial, hydrogels can be tuned to many shapes and properties and they can be fabricated from natural or synthetic materials. Due to their high content in water and their use of natural polymers, Fig. 2. Impact of a GDNF-loaded collagen hydrogel on primary dopaminergic neuron survival and striatal reinnervation. The survival (a, b) and striatal reinnervation capacity (a, c) of tyrosine hydroxylase immunopositive cells from rat primary ventral mesencephalic (VM) transplants was significantly enhanced when the cells were delivered in a GDNF-loaded collagen hydrogel. Data are represented as mean ± SEM and were analysed by one-way ANOVA with post hoc Bonferroni. **p < 0.01, ***p < 0.001 vs. VM alone; ##p < 0.01 vs. VM in hydrogel; ++p < 0.01, +++p < 0.001 vs. VM & GDNF. GDNF: glial-derived neurotrophic factor. Reproduced from Moriarty et al. (2019a). See also Moriarty et al. (2017).

Table 1
Preclinical in vivo studies of the use of biomaterials for cell-based brain repair in the central nervous system. natural hydrogels have usually good biocompatibility and biodegradability. On the other hand, synthetic hydrogels can be tailored to a higher extent in terms of mechanical strength and biological responses to stimuli and can be modified to acquire special properties. A broadly used synthetic material is polyethylene glycol (PEG) since it is a bioinert material that works well in hydrogels (Peppas et al., 2006). For example, PEG hydrogels were modified with Arg-Gly-Asp (RGD)-containing peptide sequences to modify cell-adhesion (Hern and Hubbell, 1998).
In the context of cell repair, the inherent properties of the hydrogels have many advantages over other scaffold materials. Injectable hydrogels constitute an alternative to complex surgeries as they are introduced in the brain as liquids and polymerase in situ in the transplanted area. The mesh-like morphology of hydrogels makes them ideal to encapsulate cells, as they cannot migrate outside the biomaterial (Pettikiriarachchi et al., 2010). Moreover, the porosity of hydrogels allows long-term cell survival and axon infiltration. In fact, the high permeability of hydrogels supports the diffusion of nutrients and metabolites, which facilitates the viability of encapsulated cells (Nisbet et al., 2008). Furthermore, the mechanical properties of hydrogels are similar to those of the body, which enhances cell survival. They can also be engineered to provide mechanical and biochemical signalling cues to cells to promote cell repair. Interestingly, modifying the mechanical and physical properties of the hydrogels has also been reported to influence cell fate (Aurand et al., 2014).
As mentioned earlier, hydrogels can naturally promote cell survival by providing structural support to the dissociated cells, but more interestingly they can be modified to enhance cell adhesion. For example, the ECM-derived RGD motif is incorporated into synthetic hydrogels to improve cell adhesion (Chien et al., 2012;Zustiak et al., 2013). In the context of PD, Adil et al. (2017) functionalised a hyaluronic acid hydrogel with RGD motifs to tailor the stiffness of the scaffold to encourage survival and maturation of human induced pluripotent stem cells into midbrain dopaminergic neurons. Similarly, the addition of ECM-derived synthetic peptides, from common components of ECM such as fibronectin or laminin, to hydrogels, have been identified as promoters of cell adhesion and viability (Tam et al., 2014). For example, in rodent models of traumatic injury, the addition of these natural compounds to scaffolded cells improved the survival of NSCs (Tate et al., 2002(Tate et al., , 2009Cheng et al., 2013). Additionally, hydrogels can mimic the properties of the ECM to promote cell adhesion, proliferation and differentiation by using natural ECM factors as their main componentsuch as collagen or hyaluronic acid (Hinderer et al., 2016).
Improving viability of the transplanted cells is the main goal that drives the use of scaffolds, nevertheless these biomaterials can also improve other aspects of the microenvironment that can be worth investigating in some fields. Beyond cell carriers, hydrogels can also constitute a reservoir for the controlled release of growth factors, exposing cells to an enriched growth-factor microenvironment. The enrichment of the transplantation site microenvironment using hydrogels as both cell and growth factor carriers can dramatically improve the survival of transplanted cells (see Fig. 2), leading to functional recovery (Moriarty et al., 2017(Moriarty et al., , 2019b. The ideal duration of growth factor release is determined by a balance between the maximum benefits in cell survival and the minimum side effects caused by unnecessary long exposure (Mitrousis et al., 2018). As an example, Fon et al. (2014) used a gelatin-based hydrogel to release glial cell line-derived neurotrophic factor (GDNF) and attract neural progenitor cells towards the implant.
What makes biomaterials so appealing to tissue regenerative strategies is, of course, their adaptability to functionalisation, being highly tuneable in shape, texture, porosity and many other properties. Finding the right biomaterial with all the desired properties is challenging and as described above, the intended use of the material will determine the importance of each of its properties (Tuladhar et al., 2018). Hydrogel design is advancing fast with each new technology that becomes available, improving each aspect involved in cell repair (cell adhesion, biological cues) to provide the best neuroprotective outcome possible to the targeted cells.

Nano-and micro-particles
Nanocarriers such as nanoparticles loaded with relevant biomolecules are an invaluable resource to deliver and release biomolecules to the targeted area in the brain as their small size allows them to cross the BBB. Their ability to encapsulate biomolecules and to deliver them locally has been shown to enhance brain uptake, improve the pharmacological profile and reduce the toxicity of many active compounds (Khan et al., 2017). However, from a regenerative point of view, the delivery of biomolecules alone has a limited regenerative impact in the damaged tissue. For this reason, nanoparticle only approaches for brain repair are not included in this review. However, several studies have shown the potential of microcarriers and other microscale materials for cell-based brain repair in the brain.
Some of the earliest studies in this field involved intra-striatal transplantation of dopamine-releasing rat adrenal chromaffin cells seeded onto Cytodex® (collagen-coated dextran) or glass bead microcarriers resulting in long-term functional recovery of hemi-Parkinsonian rats (Borlongan et al., 1998;Cherksey et al., 1996). Cytodex® beads also improved the survival of both rat allograft and human xenograft foetal ventral mesencephalic transplants in the rat striatum (Saporta et al., 1997). Similarity, colloidal glass microbeads have also been used successfully to mature and transplant embryonic rat hippocampal neurons into the adult rat hippocampus (Jgamadze et al., 2012). Interestingly, one such approach that progressed as far as clinical trial is the use of microcarriers for transplantation of human retinal pigment epithelial cells for PD. These cells can produce and release the dopamine precursor, levodopa, which is the rationale for their use as a cell therapy for PD. Several preclinical studies have shown that seeding and transplantation of these cells on gelatine microcarriers improved their efficacy (Flores et al., 2007;Subramanianet al., 2002;Watts et al., 2003), however this approach produced no benefit when assessed in clinical trials (Gross et al., 2011).
Another approach is to use pharmacologically active microcarriers that are designed to deliver both cells as well as therapeutic molecules. One such example is the use of poly(lactic-co-glycolic acid) (PLGA) microcarriers functionalised with GDNF for in vitro seeding and in vivo transplantation of foetal ventral mesencephalic transplants in the hemi-Parkinsonian rat striatum (Tatard et al., 2007). This approach improved survival and reinnervation capacity of the cells, as well as their functional capability. Pharmacologically active microcarriers have also been used to improve the reparative capacity of a population of MSCs termed marrow-isolated adult multilineage inducible (MIAMI) cells in preclinical models of stroke Quittet et al., 2015), PD (Daviaud et al., 2015;Delcroix et al., 2011) and Huntington's disease (Andre et al., 2019).

Hydrogels containing nano-and micro-particles
Another popular strategy is the combination of injectable cell-loaded hydrogels with nanosystems such as microspheres or nanoparticles (Baumann et al., 2010;Lampe et al., 2011). This synergistic approach has multiple advantages as the hydrogel keeps the cells and particles contained in the area to treat while the particles offer a sustained long-term release of the biomolecules, such as neurotrophic factors to support cell survival. In PD, many studies have shown the beneficial effects of transplanting neuronal cells in GDNF rich scaffolds, improving survival and functional recovery (Wang et al., 2016;Moriarty et al., 2017Moriarty et al., , 2019b. Although cell survival is usually the main target when using nanoparticles, therapeutic moieties can also be used to enhance other aspects such as angiogenesis by adding, for example VEGF in the injured site (Qu et al., 2011;Bible et al., 2012).

Conclusions
The complexity of the brain (and consequently of its disorders) challenges the development of successful treatment strategies. The generalised failure to develop new drug treatments for neurological diseases has prompted the rise of alternative sources such as cell-based therapies to heal and repair the brain. Biomaterials have immensely transformed the potential of brain cell repair. The ability to design biomaterial-based devices to deliver cells and biomolecules in situ in a minimally invasive manner to target neuroprotection and neuroregeneration is an interesting path to treat neurological afflictions without the classic drug delivery limitations encountered when designing therapeutic drugs for the brain. It is clear that we are moving into a specialised rational design of biomaterials to address CNS injuries and that each condition will be exhaustively studied to further design the best possible biomaterial-based therapy. Most likely, it will not be a single solution for each condition, but the contrary, a range of biomaterial-based devices will coexist in their way to clinical practice. The biomaterials field is continuously changing as new developments occur and these novel materials will hopefully improve the efficiency of cell-based therapies in many conditions.